key: cord-0005454-hrxxdqei authors: nan title: Abstracts: LVII Annual Meeting of the Italian Society for Veterin Ary Sciences (SISVET), Ischia 2003 date: 2004 journal: Vet Res Commun DOI: 10.1023/b:verc.0000044434.75094.b5 sha: 0ba08d35d8c3d827a79950dbadc0e76a6f6da45a doc_id: 5454 cord_uid: hrxxdqei nan advances in information and other technologies, remaining aware and responsive to users and the need to aggressively globalize their teaching, research and outreach programs. We must learn to share resources, ideas and, most importantly, knowledge. Generally speaking, people have yet to appreciate learning as a lifelong process. In many countries, economic difficulties are affecting or postponing the implementation of important measures to further the aims of lifelong learning (LLL). Furthermore, LLL will become a mainstream preoccupation for our faculties since it is essential for economic and social development. The clear message is that traditional systems must be transformed to become much more open and flexible, so that learners can have individual learning pathways, suitable to their needs and interests, and thus genuinely take advantage of equal opportunities throughout their lives. Creating a culture of learning depends ultimately on increasing learning opportunities, raising participation levels and stimulating demand for learning. Today, curriculum reform is an on-going process in most schools, increasing direct self-learning periods, reducing formal teaching, increasing practical and extra-mural services, introducing electives/optional oriented to animal species and future specialization areas. Resources are declining significantly and this is likely to continue, meaning that one needs to be more imaginative and entrepreneurial in attracting the resources required to serve the public interest. The formal system is still rigid and based on very old compulsory European Directives (EEC 1027/78, 1028/78) pointing to old-fashioned omni-competence. Despite many projects and programs initiated by the EU, and other bodies such as EAEVE, drawing up benchmarking for Veterinary Science, most curricula are not in modules, and there is little interplay between paths of learning. The project VET2020 (www.fmv.utl.pt/vet2020), a market study on the future profile of a veterinarian according to society demands, that we have coordinated for the past 2 to 3 years among 22 European partner countries, has presented clear evidence that there are specific reasons why veterinary education needs reforming. First of all, the agricultural/animal industry is declining in importance in most European countries while growing in developing countries, where population increases. Society in general, and gender of the veterinarian, is changing rapidly. The changing role of the veterinarian and the occurrence of new emerging diseases of animal origin and human implications, are strong reasons to justify a set of additional subjects or program adjustments. This could be the most important specific reason for changes in veterinary education, through its implications in Food Science, Animal/Public Health and Environmental Protection. There is a drastic need to 11 address food supply and security issues in the veterinary curriculum. Concerns about the environment and natural resource development must also be taken into consideration. Nevertheless, despite the needs of common policies and legislation within the EU, it is very difficult to point out a model example to be copied by all schools in this process of change. One can learn from all processes and from all examples, but solutions have to be local. They must be studied and suited to local conditions, regularly updated, yet fulfilling the minimum requirements set up by specific guidelines/directives. The process of change needs devoted and appropriate institutional management and leadership, teamwork, real definition of aims and goals, avoiding isolation, inbreeding, provincialism and inadequate quality control. There has been much discussion on various different fora concerning accreditation. This includes discussion with the European Commission concerning accreditation in the sense of the legal recognition of veterinary diplomas issued by establishments that meet a certain standard. The European Commission has explicitly stated that the current legal basis for recognition and free movement is that set out in the veterinary training directives. Any changes to the current system would require modification of EU law. Establishing a legally-based system of accreditation would require at least two major changes: -The existing directives on veterinary training and the mutual recognition of veterinary diplomas/degrees would have to be revoked; -A legal and professional framework would need to be established under new legislation whereby automatic recognition is given to diplomas delivered by accredited establishments, but whereby also further requirements can be imposed on individuals with diplomas from other establishments. The European Commission has made it clear that it has no plans to embark on such changes without the political will of the Member States. The Commission has no competence to intervene in this way in higher education. In this phase, a methodology was developed which permitted the defining of the most important indicators of the veterinary sector and established all the necessary elements with which to implement the successive phase of research. Through careful bibliographic research and close collaboration with the main faculties of veterinary medicine of the 20 European countries involved in the survey, which supplied and validated the data making up this report, the following indicators have been defined: 1. The number of veterinarians and students. 2. The distribution of veterinarians according to fields of work. 3 . The change in the number and gender of students over time. 4 . The number of main livestock species. The VET2020 project puts the number of veterinarians in these countries at 173,427; approximately three-quarters of these are found in the European Union (139, 893) and the remainder reside in the Non EU Countries (33, 534) . These data were provided by the Faculties of Veterinary Medicine involved in the project (see Fig. 1 ). Figure 2 illustrates the distribution of veterinarians by region. For a better understanding of the importance of the veterinary profession in each of the countries involved in the survey, Table I shows the number of veterinarians per 1,000 inhabitants in each country. Four of the five most populous European countries -Germany, Italy, the United Kingdom and France -have values more or less similar, ranging from 0.309 per 1,000 inhabitants for France to 0.366 for Germany. Spain is the only big European country with a higher value, 0.620. The countries with the highest absolute values are the Slovak Republic (0.700) and Slovenia (0.651). Ireland, Greece and Belgium have values above 0.5. The countries with the lowest ratio of veterinarians are Sweden (0.196), Finland (0.281) and Poland (0.285) (see Table I ). In order to evaluate the prospects of the veterinary sector, it is opportune to verify not only the number of veterinarians, but also the number of veterinary students. The total number of veterinary students in the 20 countries covered in the survey is 55,469, 83.2% of whom reside in the EU; the remainder being from Non EU Countries. Here 16 it is also opportune to compare the number of veterinary students to the total population of each country under consideration. This is illustrated in Table II , where the data is presented as veterinary students per 1,000 inhabitants (see Table II ). For a complete analysis of the university system, Fig. 3 illustrates the total number of veterinary faculties present in each country. The results emerging from the analysis of Phase A, although carried out in an in-depth and critical manner, did not provide all of the elements necessary to gain an understanding of, for example, the demand for veterinary services in 2020, new fields of specialisation or the veterinary fields with the best employment prospects, nor did they help in establishing the strengths and weaknesses of the current system. The methodology of the survey was found to be the best method to compensate for this lack of specific information. This also allowed the successive phases of the project to be based on less extemporaneous knowledge. So the second phase of the project entailed an original sample survey focusing on principals with direct experience in the veterinary field or those with a thorough knowledge of the main characteristics and changes affecting the future of the veterinary profession. In particular, the following were carried out (numbers in brackets refer to completed questionnaires): Total 100.0 100.0 100.0 *Multiple answers refer to the sum of responses given by respondents and indicate the proportion of respondents that gave a specific answer, independently of the order of priority. With reference to the above table for example, 60.1% of veterinarians in countries of the European Union, that consider the university systems in their respective countries to be very efficient, indicated a ''good basic education'' as a strong point of the system. This factor is indicated as the most important strong point by 53% (1st choice). In all of the following tables marked with an asterisk (*), multiple answers have been analyzed. The tables must be interpreted taking this into consideration. Source: Nomisma VET2020 Survey. (823) The VET Survey focused on veterinary professionals operating in different fields: practitioners, government veterinarians, professors or university researchers, veterinarians employed in industry. For the Employers Survey, companies involved in the food sector, feed and pharmaceutical industries as well as veterinarian organisations were surveyed; these partici-18 Figure 3 . Faculties of Veterinary Medicine in Europe by Country -2001 pants represented those who directly employ or can furnish employment to veterinarians. Organisations Survey -(111) The Consumer Organisations survey was aimed at associations whose members are habitual users of veterinary services. For all three surveys, data was collected through questionnaires distributed by e-mail and was filled in by accessing a web page dedicated to the VET2020 survey (e.g., http://www2.nomisma.it:3110/Vet/index.htm). The survey was conducted according to the following process: 1. Initial contact -request for collaboration with VET2020 project by e-mail, with attached Survey presentation letter. 2. Reminder to non-respondents sent at regular intervals, every 10 days, up to a maximum of 5 solicitations. 3. For the fourth and fifth requests to respond, the country partner sent a personal request for collaboration to the non-respondents. In general, response to the survey was very good. Contact by e-mail was a less intrusive way to conduct the survey, providing more freedom for the respondent to choose the best time in which to complete the questionnaire. This proved to be very 19 useful to the veterinarians and companies and organisations responding. Finally, using e-mail as the method for contacting respondents greatly reduced the time necessary for conducting the survey. In order to understand the value of collecting data in this manner, it is important to point out that approximately 1,600 veterinarians were contacted for the VET Survey and 823 completed questionnaires were returned, giving a response rate of approximately 50%. A total of 3,600 e-mails were sent, including initial contacts and further requests for collaboration. It was deemed important to evaluate perceptions of the efficiency of the current university system. The results obtained indicate a clear cut prevalence of positive evaluations and a high level of efficiency (see Fig. 4 ). 59.7% regarded the university system of veterinary education to be efficient or very efficient. This value is greater than 65% in Non EU Countries. 59% of the respondents in EU Countries had perceptions of efficiency. Differences appear in the analysis by region. In fact in some cases, the current university system is perceived better in respect to the evaluation given by the respondents of all the 20 countries involved. In particular, Region 1 (United Kingdom and Ireland), Region 3 (France, Belgium and Netherlands), Region 5 (Austria, Czech Republic, Slovenia, Slovak Republic and Hungary) and Region 7 (Finland, Norway and Sweden) demonstrate higher efficiency evaluations, while in Region 2 (Portugal and Spain), Region 4 (Germany, Poland and Romania) and Region 6 (Greece and Italy), the proportion of respondents that gave a not efficient Figure 4 . What is your evaluation of the efficiency of the veterinary training offered by the current university system in your country? 20 evaluation is superior to those who gave an efficient one. Furthermore, United Kingdom and Ireland (Region 1, 18.9%) and Finland, Sweden and Norway (Region 7, 16.2%) are distinguished by the highest evaluations, declaring that the university systems of their countries are ''very efficient''. Tables III and IV permit a further evaluation, showing strong and weak points of the current university system. Of all the respondents who perceived their institutions to be highly efficient, the most important point in the European university system was that of a ''good basic education'' guaranteed to students (48% as the first choice and 55% overall). Further Total 100.0 100.0 100.0 *Multiple answers refer to the sum of responses given by respondents and indicate the proportion of respondents that gave a specific answer, independently of the order of priority. With reference to the above table for example, 60.1% of veterinarians in countries of the European Union, that consider the university systems in their respective countries to be very efficient, indicated a ''good basic education'' as a strong point of the system. This factor is indicated as the most important strong point by 53% (1st choice). In all of the following tables marked with an asterisk (*), multiple answers have been analyzed. The tables must be interpreted taking this into consideration. Source: Nomisma VET2020 Survey. 21 elements of efficiency which were well appreciated are ''highly qualified teachers'' (15.5% and 22.1% respectively) and ''good practical training staff '' (10.7% and 28%) . When comparisons are made between EU and Non EU Countries, no real differences emerge, with the exception of the importance that Non EU Countries veterinarians assign to the ''qualification of teachers and staff ''. Also, the method for ''student selection'' appears to be more important for EU countries. With regard to weak points, the responses seem to be concentrated on a limited number of factors. The lack of clinical experience is considered by far as the most important weak point (30.2% and 50.6%) at a group level and this is even more pronounced in the Non EU Countries (54.5% and 67.9%). Excessive numbers of students (17.6% and 21.6%) and the lack of coherence in educational programs to meet market demands (15.8% and 26.1%) appear to be quite problematic, particularly in countries of the European Union. Common to both groups is the perception of a poor professionalism of teachers in the teaching staff. The shared opinions on weak and strong points permit a better understanding of information relative to modality and tools for improving the educational training offered by veterinary faculties in the European university system (Table V) . By far the most important factor cited (22% of first choices and 39% overall) was a deeper practical and clinical experience. This was even more strongly suggested by veterinarians from the Non EU Countries. Following this, in descending order of importance are a limited number of students (above all in EU countries), more availability of funds (Non EU Countries), teaching motivation by professors and more cooperation with the private sector (EU countries). In addition to adjustments in the European university system, it was further requested to develop the tools necessary to guarantee to the students a level of competence capable of meeting the challenges and demands of the next twenty years. Another important objective of the VET2020 Survey was to identify, according to the different categories of respondents, the opinion on the changes to be seen in 2020, in the number of veterinarians employed in the respective countries. The survey enables us to estimate the number of working veterinarians in 2020. The first important finding, confirmed by all three surveys, regards the increase in the number of veterinarians employed in 2020. 58.3% of the responses received from all the 20 countries covered in the VET Survey indicated that they expected growth in the number of employed veterinarians in 2020. Analogous predictions emerged in the other two surveys: 76.1% of respondents to the Employers Survey and 54.1% of the participants in the Consumer Organisations Survey in the 20 countries indicated growth as well. Growth predictions are confirmed for both the European Union and Non EU Countries, with some differences in intensity, as can be seen in the following table. The balance represents the difference between those who indicated an increase and those who predicted a decrease, net the proportion of those who foresaw no change at all. The prediction of an increase in professional opportunities for veterinarians in 2020 22 Total 100.0 100.0 100.0 *Multiple answers refer to the sum of responses given by respondents and indicate the proportion of respondents that gave a specific answer, independently of the order of priority. With reference to the above table for example, 60.1% of veterinarians in countries of the European Union, that consider the university systems in their respective countries to be very efficient, indicated a ''good basic education'' as a strong point of the system. This factor is indicated as the most important strong point by 53% (1st choice). In all of the following tables marked with an asterisk (*), multiple answers have been analyzed. The tables must be interpreted taking this into consideration. Source: Nomisma VET2020 Survey. 23 with respect to the present constitutes an important result of the survey, confirmed not only in the different geographical areas but also reinforced by the expectations of those who utilise professional veterinary services (see Table VI ). In the VET Survey it was possible to verify the intensity of forecasted levels of growth in various areas of the continent. The following figure shows the balance in the seven identified regions. To summarize the opinions expressed by individual regions, Fig. 5 proposes, for each region, a ''Balance Value'', the value that represents the difference between those predicting an increase in the number of working veterinarians and those foreseeing a decrease, net of the proportion of respondents that expressed the opinion that the number of working veterinarians would remain stable. It is possible to immediately note that for all the regions the balance value is positive. This signifies that the prevalent prediction of growth in the number of employed veterinarians is transversal and shared by all the Regions. It is easy to verify how the results of the entire survey (Total) are strongly influenced not only by Region 7 (Sweden, Norway and Finland), but also by Region 3 (France, Belgium and Netherlands) as well as countries of Region 2 (Spain and Portugal), all more important in numerical terms in relation to the number of veterinarians present. All other areas appear to be under the total European value and in particular, the estimates of growth appear especially low in Region 2 (9.3% of balance value), in Region 1 (United Kingdom and Ireland, 12.4%) and Region 4 (Germany, Romania and Poland, 10.7%) . To better understand the directions of such growth and spell out the implications for training and educational requirements, the investigation conducted an in-depth analysis into the opportunities Non EU with individual professional fields of veterinary medicine. Source: Nomisma VET2020 Survey. The following three tables synthesize the results of the forecasts referring to veterinarians employed in different fields of activity in 2020. The forecasts for veterinary employment in each field were formulated taking into consideration the prevalence of the responses furnished by the respondents. They are listed in decreasing order in terms of the percentages of respondents who, respectively, expected growth, stability and decrease in job opportunities. Furthermore, the diagrams illustrate the differences between results in European Union Countries and the Non EU Countries. All three groups surveyed shared the opinion that the number of veterinarians who will find work in the field of ''Food Quality and Safety'' in 2020 will increase. There are, however, some differences at a geographical level in terms of the growth expectations for Other fields of professional activity, which can be identified by observing the preceding diagrams. Furthermore, it should be noted that among the professional fields where a decrease in job opportunities has been predicted over the long-term, ''Experimentation on Animals'' was prevalent (only the Employers predicted stability in the number of jobs in this field). Although having different proportions, analogous situations seem to be present in the different geographical areas taken into consideration. Figure 5 shows the most important fields in terms of growth, stability or decrease for each of the seven regions. The trend for each geographical area was identified by the prevalence of the responses by veterinarians of each region. Only the field ''Food Quality and Safety'' appears among the first three choices for every region and seems to be the most highly considered element in Region 1 (Ireland and United Kingdom), The free movement of persons, goods and services within the European Union (EU) is one of the major principles established by the European treaties. This free movement shall now be reinforced through the full application of the new general system for the mutual recognition of professional qualifications, in which veterinary medicine is included. The success of this measure for internal market development imposes availability of professionals with the highest possible basic training and opportunities for continuing education and specialisation. Such benchmark definition requires the establishment of veterinary training throughout the EU to focus on the qualitative aspects of the basic training they impart. New production forms, new labour markets and a higher degree of consciousness of the producers and the consumers, together with an ever-increasing load of new information and knowledge in most veterinary fields had forced changes in veterinary education strategies. These changes have led to the adaptation of curricula and the application of new pedagogical concepts ultimately leading to the design of new, exciting programmes of veterinary training. Some of them use a combination of basic education and elective terms while others have focused training in species-oriented tracks already by the time students enter the clinical level. There is general consent that the quality of basal training must enable the student to achieve a level of confidence in life-long learning so he/she would be able to follow relevant CPD's and, eventually, pursue specialisation. At the same time, veterinary establishments are concerned with their ability to achieve these goals, mostly due to the usual high costs of veterinary training that constrain their chances to maintain equality of training levels through the EU. We need to find tools to harmonise veterinary training among the establishments of veterinary education in Europe, beyond the compulsory subject and training minimum requirements laid down by the Directive 78/1027. Harmonisation requires regulations but also awareness. Establishments of veterinary education must not only comply with regulations but also become aware of the advantages of quality assurance of their basic training. The present paper is a series of personal reflections by the author who ultimately addresses veterinary educators and interest organisations such as the European Undergraduate veterinary education shall lead to formal qualification to pursue the activities of a veterinarian. The classical view is that a veterinary professional is to be in his core a well-trained clinician, able to take care of diseased animals. However, the professional field for a veterinarian is widening. Two major factors can be identified: the even-increasing bulk of knowledge in veterinary medicine and, perhaps more importantly, the request from society that the veterinary profession undertakes roles relevant to the re-assurance of human well-being, in terms of public health, but also regarding the increasing consciousness in issues of animal welfare, sustainable animal production and environmental protection. In 1999/2000, an EU-financed project (VET2020; Development of European Educational Strategies: Design of Veterinarian Profiles Identified by Market Needs for the Year 2020, SOKRATES Thematic Network project 10042-CP-1-(99)2000-1-PT-ERASMUS-ETN), hereby named ''Vet2020'' was launched, whose aim was to explore the market needs for different profiles of veterinarians for the year 2020. The project ultimate aim was to use these data to signal ways for further development and, eventually, the harmonisation of the veterinary curricula among the European establishments of veterinary education. As a major component of the project, a survey was conducted by a sub-contractor (Nomisma Spp, Italy), to determine the presence of trends indicative of different market needs for some veterinary profiles. The survey attempted to gather the opinion of a randomised sample of professionals, in as many different working fields as possible, to express their opinion on curricular issues and areas of expected development by 2020. Other important actors were also included such as employers of veterinarians, consumers of veterinary services, professional associations, etc. The results, although emanating from an overall rather small sample which might be considered by some countries/regions as insufficient or of low reliability, yield a series of interesting trends of opinions whose further analysis has been presented at the 5th General Meeting of VET2020 held in Toulouse on May 24, 2003 (Seminarium on ''Future directions for veterinary education''). The analysis of the data emanating from the survey shall constitute the basis for the Final Report of the VET2020 project, to be ready for deliverance to the Commission by the project Co-ordinator following the 6th General Meeting (Lisbon, September 20, 2003) . From the Nomisma report, the overall opinion (by all target groups surveyed) signalled the veterinary profession is not expected to decrease its present level of engagement among European professions, neither in terms of needs (employers and consumer organisations) nor expectations (consumer organisations, professional organisations, professionals, etc.) . Professionals regarded veterinary education as dynamic, and could foresee -according to the survey -changes in the curricula that better adapt to expected needs of the market. Among these, the survey identified the continuation of a good, yet perhaps diversified and much more specialised clinical training; as well as an increase in the training of particular issues of utmost importance for the protection of consumers' health, such as food quality and safety or veterinary public health. Among the employers and, particularly, the consumers' associations (the ones the market listen carefully to), areas such as veterinary public health, including herd-health and surveillance were among those predicted to harbour the highest opportunity for employment by 2020. As a whole, even noting the presence of regional and national differences, the survey clearly marked the need for adaptations of the curricula in veterinary training so that they best convey to the expectations of the responders, eventually leading to a stronger harmonisation among the establishments of veterinary education in Europe. The present paper is a series of personal reflections by the author who ultimately addresses veterinary educators and interest organisations such as EAEVE and FVE to focus on strategies of quality assurance as the basis for claims of amendments of the EU-Directive/s regulating veterinary training in Europe. The paper includes the rationale behind initiatives undertaken by the VET2020 Steering Committee, in this respect. Veterinary undergraduate education was established in Europe about 250 years ago, as an answer to the needs of sectors of the society that needed skilful craftsmen, able to cope with the diseases of the moment. Noteworthy is to remember that veterinarians were, from the beginning, engaged in the establishment of preventive measures for the animal plagues that were a threat for the animal population and the humans that either consumed their products, cohabited with them, or both. From being initially directed towards the equine species for its role in transportation, haulage and for their role in less noble causes for mankind (warfare), more attention was later paid to buiatrics, due to the recurring epizootics among cattle, followed by other food-producing animals. Veterinary education establishments, the educators that shape the impact of these establishments, and the veterinary profession as a whole have a lot to be proud of since two and a half centuries ago. The veterinary profession continues to play leading roles in alleviating the suffering of animals kept for recreational, social or production reasons, on the control of zoonotic diseases, as well as in increasing the animal health status of food-producing animals to such an extent that their potential productivity of human foodstuff has in many countries reached top levels and contributes to better 32 human health. Veterinarians are also taking part in the development of science, often at the very cutting edge, in areas involving genetics, microbiology, nutrition and reproduction. We continue to produce highly qualified and talented graduates who contribute to society in many different ways. It is, therefore, our duty to see that veterinary undergraduate education continues to develop to ensure the best possible training of academic professionals. In that duty is included our need to face changes with a positive attitude, making the necessary critical analyses considering our strengths and weaknesses and formulating imaginative plans to meet these challenges. Educators have always faced the increasing volume of scientific knowledge and formulated their teaching strategy accordingly. Today we also face the increasing expectations of a public (as producers, consumers or decision-makers) that in a more critical manner, often related to the higher flow of popular information, questions the quality of the veterinary profession. The above factors lead to an increasing interest of the professionals to pursue specialisation after graduation. Educational establishments are confronted by two major needs: new curricula of veterinary education that can deliver academic professionals trained for a life-long learning as well as the training of new educators and postgraduates that would ensure the function and future development of veterinary education and research. Obviously, these two needs become tasks which can only be tackled if the quality of our function as training institutions is high, and sustainable. Veterinary undergraduate education in Europe (and perhaps worldwide) is probably still considered one of the most traditional university educations, with extensive formal teaching, based upon rigid curricula, most often teaching facts rather than principles. Another characteristic, owing to its legal professional status, is the in-built requirement for a ground of basic training that should be broad enough to warrant the graduate to practise, in principle, in any field of veterinary medicine. Such equation, of getting qualified training in every field within a limited timeframe, is not solvable. Possible solutions have been, and are being, tested among establishments that recognise this dilemma. In most places throughout Europe, efforts are now made to train students to develop basic skills but also to make them aware of the urgent need to seek information and to learn how to learn to make use of the information gathered. They should be, one would hope, better trained in problem-identification, problem-solving and decision-making. Basically, educators nowadays aim to engage veterinary education in the worldwide trend of life-long learning of our students instead of allowing them to memorise facts in order to pass examinations at the end of their courses. As well, throughout Europe, there is an increasing awareness for implementing pedagogical approaches such as problem-based learning (PBL)-strategies in the pre- 33 clinical years in order to train the students to study in a different manner than what has been praxis until now. An early exposure to veterinary working fields, to the clinical subjects and a larger degree of interaction between educators from different institutions are also among the implemented strategies. Inclusion of new subjects, attempting the elimination of the sacred limits between disciplines are now more common, and we use more extramural training and elective courses or activities to be able to cope with the requirements of society. Accompanied with proper didactics, the clinical subjects have evolved from treating individuals to think more globally, involving societal needs, animal welfare, economics and epidemiological aspects. Classical subjects, such as food hygiene, are now widened into Veterinary Public Health for instance, and population medicine, herd health and surveillance are now developing away from the classical clinics of individual food-producing animals. More and more establishments of veterinary education assume that there is major need for defining a core curriculum; either because of the inherent claim that all graduates should receive the same legal status to practise the profession, because it contributes to the integration of course components and also because it aids the development of elective periods, diversification or even the adoption of restricted individual programmes. Defining a core curriculum provides larger benefits when a PBL-learning strategy has been applied at the beginning of the veterinary programme. Problem-based learning involves the use of clinical problems (sometimes cases, but these can very well be of a population nature rather than individual) in order to create an active, student-centred learning environment. Considering that most veterinary problems often require an interdisciplinary approach, students learn how to study basic and clinical sciences in an integrated fashion. This training of how to learn should be easily made permanent by the student, since it does not have to be maintained in the programme scheme along the curriculum, but rather be applied by the students themselves after passing the pre-clinical period. Particular advantages of an early application of PBL-strategies are the self-direction of learning, obtaining interdisciplinary knowledge, as well as the development of personal skills, including personal communication skills. Both analytical and technical/clinical skills need to accompany this approach of learning, particularly with proper hands-on training. However, educators must be convinced that the PBL-strategy is not just another fancy method of instruction but a new philosophy of education, focusing on the students' own work, so that they can grow as independent learners. Veterinary knowledge increases, as well as the flow of information rising beyond our capability of pressing it within a formal curriculum of 5 or 5.5 years of instruction. The development of new technologies in many para-clinical and clinical subjects (including artificial reproduction technologies, specialised surgical sub-disciplines, imaging technologies, etc.) in the undergraduate syllabus continues to grow. There is a society-driven increment in the diversity of species which veterinarians working as clinicians are to be consulted about, and farm animal clinics evolve towards preventive medicine, including issues of herd-health surveillance and risk assessment rather than the classical handling of individual illnesses. All these changes diminish our chances of reducing the already overcrowded veterinary curriculum. Counter-measures are already being tested, some of them successfully. Provision of electives and/or tracks can individualise the programmes of veterinary undergraduate education. Elective periods, often at the end of a common part of the curriculum and lasting a few weeks to 1 or even 2 semesters, help the students to identify an area of particular interest and deepen into it. The area can include courses (either of narrow or broad nature) or activities (primarily skill-training in a particular subject or series of subjects). In some cases, elective periods are associated (with advantage) to the task of pursuing a project work, also in a subject the student selected by her/himself. The introduction of elective periods is a good measure aiming at reducing the omnicompetence we have discussed above, but it has shown deviations (add-ons) and it does not necessarily limit the core per se. Tracking implies a new type of curriculum design, where the student opts to emphasise a specific area of veterinary medicine, so that a considerably enhanced level of knowledge is achieved at graduation. Even not being a specialisation (a level that would be obtained after graduation) tracking takes the student to a more useful level of competence in that particular chosen field, limiting redundant information and facilitating a more focused learning. Its major argument is that it limits information overload in the curriculum. However, tracking is not common yet, and many concerns have been raised in establishments that have opted to consider this option. Firstly, it complicates an already rather complex veterinary curriculum and forces the decision by young students (maturity is not an issue to be discussed here, but it has been named) to opt for a track, sometimes rather early in the curriculum. Tracking is at the very end a decision to be made by a certain establishment in the light of the socio-economic values and needs for veterinary professionals. In some areas (as is the case of University of California at Davis, USA) the practising veterinarian has clearly evolved from an omni-competent, primary healthcare provider for several species, to a single species health specialist, a disciplinary specialist or even a species-specific discipline specialist. However, there are many regions of the world where the predominant need still is for omni-competent, multispecies veterinarians. Tracking or path-differentiation will no doubt be discussed further and some establishments in The Netherlands, Belgium, Italy and the Nordic countries are already approaching a more differentiated form of curriculum. In Utrecht for instance, the new curriculum (evolved from the study plan from 1995, and initiated 2001) includes separate educational paths (individual animals, farm animals, veterinary public health, veterinary scientific research and veterinary administration and management), enrolling on a 6-year long curriculum that attempts an increasing species-differentiation in practice as major output. 35 Changes in curricula among veterinary schools in Europe have improved training. Many students are more prone towards academic training, more interested in developing communicative skills, and are also better skilled in problem-solving. However, not all students see the necessity for improving communication skills, nor do they see the need of accomplishing academic skilfulness, since they want to become a veterinarian as fast as possible. Development of problem-solving skills is often seen as a slow process that goes against the students' inbuilt interest in gathering as many study credits as fast as possible. Last but not least, most students want to know and do all within the veterinary field. Is this a student-restricted problem? Not really. Educators are a part of the problem. Students find it comfortable and safe to receive information via theoretical lectures where the educator tries her/his best to cover everything in the subject, with the false conviction that providing all information is the right thing to do. With such a conviction, the educator is reluctant to reduce the content of theoretical classes, and ultimately, of their own discipline in the curriculum, unless they are involved in multidisciplinary teaching. Although already repeated in as many opportunities as provided by conferences and courses at national, regional and even at EAEVE's level, the ideal of full coverage is still a major component of the problem. Students must at any given course be provided with tools for seeking knowledge and further information than what is given in the syllabus. Not least important, students must have enough time to convert information into knowledge. In less and less schools throughout Europe, but still in too many, the number of students is too large. There must be a certain educator:student ratio in order to be able to provide the best possible training. The larger the number of students related to the number of teachers the worst would happen; lectures will be the consequence, and demonstrations will replace hands-on practicals ... Furthermore, low educator numbers are usually connected to few academic staff numbers per institution, with the logical consequence that these are tied to teaching hours, and where any curricular change, even of a small nature, is seen as a threat and a burden impairing other relevant University activities such as research and research education, for instance. Sometimes this matter must be handled in two directions, firstly, assuming that numerus clausus must be discussed and accepted as a need and, secondly, understanding that sometimes an academic structure based on larger departments is to be preferred. Are these scenarios working against core curricula? Yes, having teachers (and professionals) with such a conservative view is translated into students feeling they are not provided with the broadest possible training, which they assume must be the best. In establishments where core curricula have been imposed in a top-down approach, educators often responded with a document (often very, very large) of a core syllabus, where almost everything is listed for the student to understand she/he 36 must know a bit of everything, so that treating ''all creatures large and small'' can be warranted ... This misunderstanding simply contributes to maintain the bad circle. Educators are themselves, as a whole, not happy. Frustration is not infrequent, especially for those subjects that would be considered more ''academic'' than ''craftsmanship-oriented''. A visible consequence among veterinary establishments worldwide is the worrying lowering number of veterinary graduates enrolling in postgraduate veterinary training. Whether this is caused by low salaries among university staff members compared to professional positions is always a matter of discussion, but a fact is that academic positions in the pre-and para-clinical departments are being filled by scientists without veterinary degrees, most often against the clear strategic wishes of the institutions. The major problem we are facing nowadays is the disparity, still present after so many years of genuine work with curriculum development, among establishments of veterinary training in the EU. The free movement of persons, goods and services within the European Union (EU) is amongst the major principles established by the European treaties. The veterinary profession is fully included in this movement through the automatic recognition of the professional diplomas granted by the EU establishments of veterinary education (Council directives 89/48/EEC and 78/1026/EEC). This recognition relies on the assumption that an equivalent level of veterinary training is provided throughout, as required by the Directive 78/1027 (Council Directive 78/1027/EEC), which establishes a minimum of subjects and overall timeframes to be included in the veterinary curriculum. Both Directives enshrined into Community law the automatic recognition of veterinary diplomas across the EU, as well as the freedom for establishment of professionals and the freedom to provide a professional service wherever in the EU. These basic instruments for a functioning open market are now to be reinforced through the full application of the New Directive of mutual recognition of professional qualifications, where a benchmark of availability of high-quality professionals has been launched for the EU. Such a benchmark definition is of utmost importance for the EU as a whole, and particularly for the Establishments that impart the basic training for these professionals. However, if there is a dichotomy, as it is experienced at present, between the expected high-quality of the professionals and the qualitative aspects of the basic training, the system will no doubt crack in the long-term. It is a logical expectation that this is not the intention of the European Commission, and that provisions are to be taken to ensure this will never happen. However, a major responsibility lies on the establishments themselves and on the national authorities. The Directive 78/1027 harmonises, as defined above, the minimum compulsory 37 subject and training requirements for the training of a veterinary surgeon within the EU but has, despite its 25 years of age, no description of neither qualitative nor quantitative aspects for these requirements, of utmost relevance for issues of quality assurance. There is, however, a very important statement in the EC-council Directive 78/1027/EEC (Annex of the Study programme for veterinary surgeons) where it is stated that ''The distribution of the theoretical and practical training among the various groups of subjects shall be balanced and coordinated in such a way that the knowledge and experience listed in Article 1 (1) of this Directive may be acquired in a manner which will adequately enable veterinary surgeons to perform all their various duties''. Such a statement in the Directive lays the basis for quality assurance of veterinary training in the European Union at present. However, in order to warrant that the level of quality is the most appropriate, and to eventually achieve its highest possible and most relevant level among establishments in the EU, an analysis should be undertaken. Such analysis of the Directive has already been regarded as most relevant for the basic veterinary training and the development of the profession, by both the EAEVE and the FVE. Owing to this apparent lack of clear internal demands for quality within the frame of the Directive 78/1027, and in order to be able to ensure that different establishments of veterinary education (among those EU-schools and those of candidate countries) are not only complying with the minimum requirements laid down by the Directive but also are aware of quality issues, a system of voluntary evaluation of teaching capabilities has been running since 1986. The system was firstly run by the EAEVE with participation from veterinary practitioners in the visitation teams and, since June 2000 by the joint efforts of the EAEVE and the FVE, following the adoption of a new set of Standard Operation Procedures (SOP, 2000) . This SOP has been revised in 2001. Basically all the EU-schools (exception made of those recently established) have been evaluated at least once. A total of 65 Faculties has to date been evaluated under the European system. This figure does not include the evaluations performed by the American Veterinary Medical Association (AVMA) or the Royal College of Veterinary Surgeons (RCVS). Evaluations are to be carried out at 7-10 year intervals for each school and some (6) are already to be scrutinised now for the second time. Most schools (>90%) have expressed satisfaction with the present system of evaluation. The evaluation system has rendered rectification of major deficiencies (those defined as not conforming with the minimum requirements of the Directive 78/1027), and it has enabled schools to enact curricular changes derived from the recommendations, to acquire additional academic and support posts, refurbish buildings and 38 facilities, undertake new management structures and substantially augment their numerus clausus or otherwise, to ensure possible best-quality training. One wonders if these changes would have been possible without the aid of the voluntary evaluation system, whose recommendations actually help the responsible national authorities to understand the value of the veterinary training in a European perspective. The schools themselves should remember the evaluation has the ultimate aim to help the training establishment to improve, not to jeopardize, its existence. Running this voluntary evaluation system, that has been perceived as very effective in providing useful assessment reports, has made possible the identification of not only deficiencies among schools but also shortages in the Directive 78/1027, particularly in issues of quality assurance. Therefore, for the sake of the European harmonisation everyone is willing to reach, a better and more transparent system is to be devised and applied. The public, including the pertinent national authorities and the potential students, need to know the contents of these assessment reports. The better knowledge of strengths and weaknesses are of importance to encourage staff and student mobility, also a major goal of the European Commission. Teacher mobility is far too low among establishments of veterinary medicine in Europe, and curricular development, to cite one reason, benefits from teachers' interaction. This mobility should be, and in some schools has already become policy, a pre-requisite to facilitate sustainable student mobility within the Erasmus programme. Various amendments to the Directive 78/1027 have been considered and proposed during the past 25 years, mostly concerned with the differences among the various establishments of veterinary education. These differences have been substantiated, as expected, following voluntary evaluation of teaching capabilities since 1986. The evaluation of the schools, standardised with the SOP, and the participation of both organisations EAEVE and FVE is a lengthy procedure that ends with the agreement between parties at the Joint Education Committee (JEC) and Executive Committees. Obvious differences exist among establishments, inherent of a European perspective, both in terms of facilities, input/output (students/graduates), numbers of teaching and support staff, accessibility to teaching material etc, all of which can cause eventual differences among the graduates. Control mechanisms to ensure that the quality of the veterinary training is acceptable have been proposed by the existing Advisory Committee of Veterinary Training (ACVT, Internal Market III/F/5171/7/92-EN, Brussels 2-5-1993) and FVE (FVE/2000/011) but were overruled by the Commission for not fully considering the free-movement of professionals. The FVE, as a professional organisation, has earlier focused on issues of veterinary demography (Document FVE 00/011) as a major threat to the quality of veterinary education in Europe. However, such a focus on 39 issues of veterinary students or numbers of graduates among the EU has not impacted per se on the quality of training of the establishments, since the latter has not been taken direct account of, a matter for lengthy, not always constructive, discussions between EAEVE and FVE. The FVE, again acting as the professional organisation it is, focused on threats such as higher rates of veterinary under-or unemployment and the question of an over migration of veterinarians, as those causing damage to the profession. The first mentioned issue is a reality in some countries, where the number of establishments of veterinary training exceeds the expectations of the internal market. Their existence sometimes follows short-sighted political agendas (statistical treatment of unemployment, excessive offer of undergraduates without the demand of a market) or expression of regional autonomy, etc. The second issue is unsubstantiated for most of the EU, exception made of few countries, where large numbers of veterinarians cross the borders to provide services or establish themselves, following EU established laws (Rodriguez-Martinez, unpublished survey). Other matters, such as changes in gender balance, are a worldwide reality, and not necessarily related to the quality of undergraduate teaching. Despite disagreement being present for some of the above aspects between the organisations, there is a large list of agreed points, where the FVE and the EAEVE have repeatedly declared their intention to join efforts for the best development of the quality of veterinary training in the EU. Among these points is the design of a curriculum that provides (a) basic veterinary knowledge (curriculum core), (b) differentiation in later areas (clinics, research, veterinary public health, etc) and (c) a suitable system for continuing professional development (CPD's) and specialisations within the umbrella of the European College system (see below). The EAEVE has repeatedly declared its wish to update the educational standards of the EU-Directive 78/1027, following considerations -in agreement with the professional organisations FVE and RCVS -that they need to be expanded and upgraded to best reflect the benchmarks stated in the SOP. Discussions with representatives from the DG15 Internal Market earlier this year have resulted in the suggestion by the officials that EAEVE and FVE should need to make a formal submission to present possible amendments to the Directive/s. The Vet2020 Steering Committee and the Regional representatives of the project have recognised the inherent value of these signals and the need of an analysis to disclose the best way to attempt adaptations of curricula that follow the trends identified in the Vet2020 European survey. It was also the opinion of the Steering Committee and the Regional representatives of the VET2020 project (Meeting in Lisbon, March 1st 2003) , that such an in depth analysis should be carried out to produce a document listing a series of proposals, embracing earlier initiatives from both EAEVE and FVE and related to the outcome 40 of the Vet2020 project, that would be considered as a starting point for expected, suitable upgrading of the EU-Directive 78/1027 and the New Directive on recognition of professional qualifications. Such upgrading (as amendments of the Directive 78/1027 and therein) must be considered as a way to consolidate the requirements of veterinary quality that best define the benchmarks for a well-functioning open internal market for veterinary services and the establishment of veterinarians anywhere in Europe. For such a task, an ad-hoc group was created (R. Leiser, J. Leibetseder, W. Swannet, S. Cinotti, J. Soriano, D. Noakes & H. Rodriguez-Martinez, chairman) which was to co-opt a representative from EAEVE and another from FVE, to draft a proposal (to be ready by September 20th, 2003) intended to stimulate the preparation, by EAEV E and FV E, of a joint formal document to be submitted to the European Commission aiming at the evolvement, with particular emphasis on quality assurance, of the educational standards set out by the Directive 78/1027 (Annex V). This document should necessarily incorporate the benchmarks provided by the EAEVE/FVE Standard Operations Procedure (SOP) for visits of the establishments of veterinary education in Europe. As pointed out already, this visitation system has provided an invaluable first step in the harmonisation of veterinary education in Europe. However, more instruments, issued by the Commission, are needed to further continue on this path of quality assurance among the establishments of veterinary education in the EU, including those present in the countries that are candidates for imminent or later inclusion in the Union. Some of these items have been included in the Vet2020 project, where an exploration of the market needs for different profiles of veterinarians the next 20 years, would help both the establishments of veterinary education and the veterinary profession in focusing on issues of training quality while developing harmonised veterinary curricula in Europe. The primary goal for any amendments of the Directive/s is to ensure the best possible quality of the basic training, as well as an evaluation system that makes possible the surveillance (and thus quality control) of this training. Such a system is already present (Joint EAEVE/FVE SOP) but there is a need for a series of updates of the EU regulations so that the organs running this voluntary evaluation system (Joint Education Committee and Executive Committees) have a stronger ability to force schools to follow their recommendations as well as providing the EU member-states with the power to enforce rectification. In order to ensure transparency of the system, a system of publication of the final reports from each evaluation performed is desirable, eventually leading to the publication of lists of approved (or even not yet approved) establishments. The General Assembly of EAEVE, the highest body for 41 decisions within the organisation, has already taken a very important first step in this direction at its meeting in Toulouse, France (May 22nd, 2003) approving, unanimously, the following declaration: ''T he present Directive 78/1027 is considered as a starting point for upgrading this Directive, including the SOP and accreditation of veterinary establishments''. Such changes can be difficult to make under the current exercise of consolidating the new Directive of mutual recognition of professional qualifications. The Legal Affairs Committee of the European Parliament voted on June 17th 2003 to accept the principle of the Commission's proposed directive to integrate into a single text, establishing a general system thought to simplify and modernise all seven current sector directives governing the mutual recognition of professional qualifications, including veterinary surgeon, general nurse, dental practitioner, midwife, architect, pharmacist and doctor. Noteworthy, although both EAEVE and FVE, as well as the rapporteur (Mr. S. Zappalá, EPP-ED-I) marked their belief that the current system of sector directives and corresponding specific systems have been successful and efficient and should therefore be maintained intact, the Committee voted in favour of the New Directive. Under this factual scenario, there is a clear need to prepare for amendments, see that a qualified majority vote is present, being indispensable that a certain number of Member States agree on the amendments, under the period of discussion of the issue in substance, that ought to start by September 2003. The veterinary profession has a clear preventive role for human health going hand in hand with issues of animal health per se. Amendments to the new Directive might be difficult to achieve and Vet2020 is of the opinion that EAEVE and FVE should concentrate efforts on issues of quality assurance of training in the first place, focusing on convincing as many governments of the Member States to support such a change. There is, therefore, an urgent need for agreement on a minimum set of items to be considered, and ultimately agreed, by the EAEVE and the FVE in terms of quality assurance for the evolvement of veterinary education. The primary item to be considered is the attainment of three levels of veterinary education, as already contemplated in the recommendations by ACVT on February 10th, 1993: Initial training (5-6 years of veterinary training) divided into a basic (core) curriculum for all students, allowing in the final years, a period of in-depth studies. The latter can be done by in-depth studies associated with a final thesis work and/or by the creation of differentiation in the final (clinical) years or by any other devised programme, following school-, national-or even regional initiatives. These steps do not exclude each other, and allow for in-depth studies in ANY area of basic veterinary training, without detriment of a general license to practice veterinary medicine (Full diploma). Complementary training (specialisation). This system, based on resident-positions (inward) or open training (outward) is linked to the European Colleges (ECVS) or National organisations, ruled by National legislation. Continuing education (CPD's) within the frame of the Schools, and/or associated bodies (FVE, Colleges, other associations or even enterprises). The here defined basic, initial training, is the primary target for a system of evaluation that must warrant a certain degree of quality assurance. Such basic training is to be followed by specialist studies (or postgraduate studies of academic nature) for three or four more years. The basic training period could be divided into two levels, in order to, eventually, best allocate the intentions of the Bologna Declaration, provided the extent of the Declaration would include Veterinary training. However, considering the presence of clear national or regional differences with regard to the need, or even the value of such division of levels, this issue should be handled cautiously. At present, there is no need to introduce a degree of Bachelor of Veterinary Medicine other than for academic purposes (as an intermediate degree WITHIN the veterinary basic training programme), since such graduates are not needed on the present labour market, a pre-requisite for the enforcement of the Bologna declaration. Both EAEVE and FVE should jointly stand behind a series of initiatives, primarily as a follow-up of the recommendations on the modification of Directive 78/1027 proposed by the ACVT on February 10, 1993 (Internal Market III/F/5171/7/92-EN, Brussels 2-5-1993) . In particular, areas of public health, herd health and surveillance and clinical subjects should be priority, following the identification made through the project VET2020. Adoption and control of quali-and quantitative minimal training has to be issued, with the full adoption of proper indicators (as stated in the SOP) of practical training with adequate teaching supervision. Focus should, however, be put on quality of teaching rather than solely on arithmetical ratios. More in detail, EAEVE and FVE should work together to: Fully implement the SOP (Joint EAEVE/FVE), using the modification of Directive 78/1027 proposed by the ACVT on February 10, 1993 (Internal Market III/F/5171/7/92-EN, Brussels 2-5-1993 ) as a basic, albeit modifiable model. Evolve this SOP, by ensuring quality control of the teaching of every subject of the initial veterinary training with an extra-commitment on the areas identified as expanding areas in the veterinary market (VET2020). Implement a transparent system of publication of the Evaluation report after scrutiny of the SOP-guided evaluation by the EAEVE/FVE Joint Education Committee (JEC) and the Executive committees and thus considered as Final Reports of each establishment. Request from the Commission and pertinent bodies that a provision of power by the pertinent authorities is issued, to enforce that minimum resources (economic or otherwise) are ensured for the best-quality of training and that rectification is possible, when the quality of training is not satisfactory for a period passing two years from the SOP-guided inspection (application of derogation clause [Article 56]). Reclaim the need to have a SOP-inspection elicited before a new establishment of 43 veterinary education is allowed to operate in the EU (pre-requisite of the SOPguided visitation). As well, written consent is to be required for open publication of the final report by the JEC. Negotiate the development of a system of oYcial recognition of specialisation in veterinary medicine, in the new EU-Directive (SLIM Directive, 2001/19/EC) pending approval. The above set of points has to be seen as positive for the profession as such, and by the establishments of veterinary education in particular. In the first case, it has to be viewed without detriment in the EU-New Directive considering the (i) provision of services or (ii) establishment in a host state. However, it is pertinent to consider that host states should be able to require completion of areas, when quality assuranceprovided by a revisited SOP -is not proven. Obviously, this marks the needs for the schools to improve, or maintain, high levels of training quality. Without this prerequisite, any system of full mobility will always be shadowed by disbelief and, ultimately, counteract the goal of full recognition of veterinary qualifications through the EU. 9 pp, 1990. ANMVI (Associazione Nazionale Medici Veterinari Italiani), the National Association of Italian Veterinarians, is a professional association which was established just four years ago, but which already includes 15 Italian cultural and scientific associations, is an active part of the UEVP, and represents the impressive total of 10,000 Italian practitioners. We have carefully assessed the VET2020 European study, but we have not stopped there -ANMVI commissioned Nomisma to further elaborate the situation in Italy via an additional survey of data on veterinary practice in Italy in 2003, thus updating the data from the preliminary survey included in VET2020 by a further four years. This was done above all to determine the current situation of our profession in Italy in particular, and to identify possible future developments in the sector and therefore occupational developments for the coming years with reference to 2020. For a number of years now, our association has continued to express great concern over the crises and difficulties that our veterinary world experiences, and has urged the competent authorities and different components of the sector to act urgently in order to avoid situations that will only lead to the qualitative and ethical degeneration of our profession. Unfortunately, the results of this survey -conducted in great detail and very comprehensively by Nomisma -confirm that our concerns have always been correct and based on real issues. We need only add some data to realise that the veterinary world is in an extremely dangerous high-risk situation that requires immediate intervention. It is unacceptable that in the EU one veterinarian out of six is Italian, that 26.6% of veterinary medicine students are Italian, and that almost 20% of veterinary medicine faculties are located in Italy, but that in spite of this we are not even in the top rankings in Europe for zootechnic resources, for either pets or farm animals. In addition, a comparison of the 1999 data with the 2003 data reveals a 13% increase in total members of the professional rolls in just four years! Occupational growth forecasts for 2020 are low in percentage terms for our sector, whereas total members of the professional rolls are set to double by 2020 unless we take immediate action. In this extremely difficult situation, we consider it unthinkable that the Italian university system be augmented by another degree course in veterinary medicine, in a university such as Catanzaro which does not achieve the minimum 46 faculty requirements provided for by the Ministry in terms of its existing faculties, and which totally lacks all the equipment necessary for a veterinary medicine faculty. It is sad to say that this opening has been made possible through the collaboration of three other Italian veterinary medicine faculties, which will undoubtedly do no favours either for the students of this new phantom faculty, or for veterinary practice at the national level. As a professional association, we have no choice but to expose the absurdity of this situation in Italian university education, however here it is our intention to address mainly those aspects concerned with what follows training -i.e. professional access. The right to study is both fundamental and guaranteed by our constitution, but the right to study cannot necessarily tally with the right to carry out a profession. This is precisely the situation in Italy at present -the traineeship and state examinations have become a formality, such that 99% of candidates pass these examinations (FNOVI data) after a six month traineeship which can be done before graduation, and which in many cases is reduced to a mere list of attendance signatures with no effective carrying out of practical work. A profession such as the veterinary profession, which requires unflagging dedication, constant effort in terms of ongoing training, considerable investment in facilities, apparatus and equipment, and continual regulatory compliance -and all this without the financial reward found in other liberal professions -requires training for professional activities which differs from the current training programme. Students who access the profession need to be sufficiently motivated and trained, whereas at the moment the veterinary sector is currently choked with professionals who have not been adequately trained to face up to reality, and who in many cases are prepared to trade off their professionalism in order to pick up scraps of work, or who after months or even years of underemployment are destined to seek employment in another sector in which they can use their basic biology training. We are therefore committed to having an effective practical traineeship included in the pending reform of the professional rolls, which is soon to be debated in parliament. This practical traineeship should be of adequate duration and performance, with some form of financial reward if possible, and at its conclusion there needs to be a real examination, in line with the other professions. The purpose of all this is not so much corporative, to protect veterinarians who are already working, so much as to ensure that the students who graduate from our faculties in the future will have a real chance of finding work that is worthy of their efforts. FNOVI (Italian Federation of Veterinary Orders), after considering the results of the survey carried out by Nomisma, aimed at designing future scenarios for the veterinary sector in the year 2020, on behalf of the European Medical Veterinary Schools, cannot but confirm its envisaged perceptions. In fact, in 2020 the professional scenario will be characterised by an excess of professionals who will inevitably have to compete in a market characterised by limited demand (i.e. decreasing employment growth). This is in stark contrast to the professional position currently on offer. Italian professionals will further suffer from the following disadvantages, as listed below: $ A high number of professionals (the number of vets is likely to increase by 100% by 2020); $ Underemployment and unemployment (a stable employment scenario in Italy is likely to be achieved in an averagely longer timeframe as compared with other European countries); $ Envisaged market barriers for new professional profiles; $ A slight contraction of a few professional sectors (Vet sector growth in Italy in 2020 is likely to be less than 3%) due to a decrease in zootechnical production, and a lower production of animal-based food; $ Professional education which hardly matches market needs, and is also characterised by an overall lack of knowledge and specialisation. I invite you all to read that document, as it represents what should never be done. A careful analysis of the text highlights: the lack of an innovative project, the unavailability of a project aimed at restructuring the ''Condoleo Farm'' gifted by the Province to the University, the lack of a zootechnic Farm, of a pound, of labs, of a teaching hospital, and of a slaughter-house. However, all these negative circumstances were, amazingly, considered the basis for a favourable decision. In other words, the new Catanzaro Faculty was to be equipped with all the necessary staff by means of conventions with other Italian faculties. This communication was followed by the D.M. of the 3rd of September aimed at authorising the Catanzaro Graduation Course as a not well identified ''matter of urgency''. We are aware of the resignation of the President of the Dean Conference (many thanks to Prof. Girardi for his seriousness and coherence), and the protest of the professional world. From our perspective there are at least three issues to reflect on: $ it is necessary to create an independent evaluation Authority as soon as possible; $ instead of improving the existing Faculties (only 2 out of 14 actually reach European standards) by turning some of them into specialty schools, personal and local interests have been favored; $ the Catanzaro experience may give rise to other similar initiatives to the detriment of quality (something the Catanzaro Graduation Course surely cannot guarantee). This will eventually contribute to the dispersion of Italian professionals to the rest of Europe. As to the issue of professional access, also reinforced by the political intention to update the qualifying examination by the end of this year, the Federation of Orders has envisaged the setting up of an exam aimed at certifying, in addition to adequate knowledge, professional ability or at least the attitude to reach a professional level. The Federation of Orders does not envisage nor encourage the use of internal barriers or obstacles aimed at postponing or forbidding the access of newly qualified graduates into the market, but simply aims at guaranteeing the natural objectives of 49 a qualifying examination. Such objectives involve a system to monitor professional ability based on the application of identical standards throughout the national territory, and managed equally by universities and professionals. This exigency is perceived by FNOVI as a current necessity despite the statement in the MIUR outlines that the veterinary sector is not characterised by any system distortion, provided the coherence of the national data (2718 qualified out of 2760 candidates, equal to 98.5%, and five faculties reporting qualification of 100% students). As for the junior graduation courses, the professional world wonders what a graduate in ''Science of breeding, hygiene and well-being of dogs and cats'' is meant to do. The answer is supposedly provided by the course advertisement brochure, which states that: ''S/he will support the practitioner vets in her/his professional activity, either in-office, at vet hospitals, dog-pounds, and public and/or private cat recovery houses''. We cannot avoid noting that: $ there is no need for such a professional profile, nor will the market ever be able to integrate it; $ there is an attempt to create new professions with no market requirement for them; $ there is no need to create additional unnecessary ghostly professional profiles; $ it is necessary to think about new professional profiles from an ethical perspective, provided that addressing people's energies toward pure illusions with no serious occupational perspectives is an injustice. In conclusion, first of all it is necessary to establish a university system which is able to match educational needs, and to guarantee its adherence to European standards aimed at offering a wider and more diversified range of options and competence as compared to those currently available. We strongly wish to create a stable agreement between the university and the professional world in Italy, with the aim of achieving long-term professional co-operation. In Italy, about 269 companies are more or less actively involved in shellfish farming, using different methods and often having plurispecific productions. These companies employ nearly 4,000 people (Prioli, 2001) . This system is characterized by a complex structure where the legacy of old traditions coexists with modern intensive farming techniques. The transformation process leading to new farming methods, which enabled farmers to go beyond the local sphere of activity and handmade production, was triggered in the second half of the 1980s by the introduction of a new species and a new technique. The introduction into lagoon areas of the mid-Adriatic Sea of Manila clam (T apes philippinarum) resulted in an excellent adaptation of this new species to the local environment, to the extent that it spread spontaneously and allowed farmers to move from rearing practices to management of production areas in a more or less controlled way. This allowed farmers to benefit from a resource that completely changed the production structure of the areas involved in this process, with significant social repercussions. The advent of new offshore mussel farming technologies allowed farmers to conquer new spaces, in addition to the traditional farms, which were mostly located in coastal and lagoon areas, such as the Veneto lagoon, the Gulf of Trieste, the Gulf of Taranto, etc. This led to the setting up of numerous offshore production facilities, no longer constrained by environmental and health problems. So, today shellfish farming has become the main aquaculture activity in Italy, although production is almost exclusively limited to mussels (Mytilus galloprovincialis), Manila clams (T apes philippinarum), and only small quantities of Grooved carpet shells (T apes decussatus) and oysters (Crassostrea gigas and Ostrea edulis) (Prioli, 2001) . The aim of this work is to analyse the most recent data, thereby contributing to a wider knowledge of the present status of Italian shellfish farming through a short overview of the two most representative sectors, namely mussel farming and clam farming. In the year 2000, there were 204 companies owning mussel farms (Table I) , mostly concentrated in some main production poles whose location depends on the hydrologic and trophic features of the settlements. Some of the main production areas have a long-standing tradition; others have become more important after the introduction of offshore farms and the ''conquest'' of new areas at sea. The most important production areas of longer tradition include the Gulf of Taranto (Puglia), La Spezia (Liguria), the Veneto lagoon, the Phlegraean coast (Campania), and, in more recent times, the Trieste coastline (Friuli-Venezia Giulia), the Gulf of Olbia (Sardinia), the Emilia-Romagna region and along the Adriatic coast of Puglia. There is currently a positive trend towards the settlement of new production areas along the coast of those Italian regions offering more favourable conditions for mussel growing. At the moment, only two coastal regions are still totally lacking in mussel farms, namely Calabria and Basilicata. In Italy, the mussel farming sector employs about 1,800 people, around 1,500 as fixed workers and around 300 as temporary workers, hired occasionally or on a seasonal basis. The average number of workers per company is therefore nine ( Table I) . The majority of workers are obviously concentrated in the main production areas previously mentioned. The regions of Veneto, Emilia-Romagna, Sardinia and Puglia are considerably above the national average, which is mainly due to the presence of a large number of cooperative businesses. Abruzzi 1 7 5 2 7 Campania 12 62 62 0 5 Emilia-Romagna 19 367 268 99 19 Friuli-Venezia Giulia 24 50 50 0 2 Lazio 4 26 22 4 7 Liguria 68 107 107 0 2 Marche 6 32 13 19 5 Molise 2 8 8 0 4 Puglia 31 375 348 27 12 Sardinia 16 270 125 145 17 Sicily 1 2 0 2 2 Veneto 20 558 558 0 28 Total 204 1864 1566 298 Average=9 53 Apart from some relatively small farms based on extensive use of artificial reef units and management of shoals of mussels on the seabed, Italian mussel farming essentially relies on three rearing systems, namely the fixed system, the single ventia long-line and ''Trieste'' long-line or multi ventia. The fixed system is employed in lagoon areas or in sheltered coastal areas, and can be traced back to more ancient settlements. However, facilities have been through a gradual process of modernization, which in some cases coincided with the adoption of other systems. The regions with the highest number of settlements of this kind are Puglia, Emilia-Romagna and Liguria, although in Emilia these facilities, situated inside the Sacca di Goro Lagoon, are gradually being abandoned and progressively replaced by offshore surface long-lines. Single ventia plants are relatively new, since most of them have come into use over the last 15 years, but they have quickly become the strong point of Italian mussel farming, accounting for nearly 75% of the lm presently available for cultivation. Indeed, this parameter has been identified as an indicator of production capacity, since the number of strings of mussels cultivated is basically related to the linear meters of long-lines available. As previously mentioned, these kinds of facilities are located at sea, since they offer excellent resistance against marine weather events, even those of significant magnitude. The multi ventia plants are very common in Friuli-Venezia Giulia, where nearly all rearing systems are of this type, and in Puglia, Liguria and Sardinia, as well. Multi ventia long-lines originated along the Trieste coastline, where they experienced their greatest development in the 1980s, and are used in partially or totally sheltered areas. In Italy there are about 2,000,000 lm of long-line available, with each company managing about 10,000 lm on average. The regions with the largest number of linear meters are Emilia-Romagna, Puglia, the Veneto, Friuli-Venezia Giulia and Sardinia, where the largest companies are also based (Table II) . In absence of any other official and reliable sources, the data regarding production in 2000 and presented herein were derived from a set of evaluations mainly based on the following: values reported by producers; values obtained on the basis of production capacity of farms; values obtained on the basis of the characteristics of the production area and the staff employed; estimates by local experts derived from trade flows; assessment of produce transferred to other farms. The result is about 85,000 tons of production, well below the value emerging from official statistics, namely 130,000 tons (Prioli, 2001) . Apart from the figure itself, the data highlight the presence of production poles, although in some cases the values are overestimated because of some adult mussels that are put on the market after a short period of immersion in these farms. Puglia, the Veneto and Emilia-Romagna account for 80% of national production, followed by Friuli-Venezia Giulia and Sardinia. In this case, the figures do not include the production of spat, which is not always adequately reported by farmers and comes from three main areas, namely Puglia, the Veneto and Emilia-Romagna. The mussels produced in the various farms are not placed uniformly on the market, 54 since there are seasonal peaks, which can be more or less dramatic. At a national level, the large majority of farms market their produce between May and September, while between November and February the number of farms selling mussels is lower (Prioli, 2001) . More than 70% of the national production is sold by producers to wholesalers, while about 11% is sold directly to consumers or retailers. The amount intended for the processing industry is very low, being lower than 1%. Around 14% of the national production is destined for other farms, although the amount that is sent to other farms with the intermediation of wholesalers, which cannot be quantified, should also be added to this percentage (Prioli, 2001) . Clam farming companies are concentrated in the principal Italian lagoon areas. It has been estimated that there are 54 of them, generally located in the Po plain and the Veneto area (Emilia-Romagna and Veneto regions) and in Sardinia (Table III ). In the lagoons of the High Adriatic, clam farming is mainly based on the production of Manila clam (T apes philippinarum), and some modest quantities of Grooved carpet shell (T apes decussatus) coming from the ''pialasse'' of the Ravenna area (brackish lagoons connected to the sea). In Sardinian ponds the autochthonous species of T apes decussatus is extensively harvested. The clam farming sector employs quite a high number of people, for a total of 3020 workers, also including temporary and seasonal staff, mostly employed in the regions with the highest level of production, namely the Veneto, with 1677 workers and Emilia-Romagna, with 1053 (Table III) . Clams are cultivated on the seabed, over a total surface area of about 112,000,000 m2, 50% of which is in the Veneto, 36% in Sardinia and about 9% in Emilia-Romagna (Prioli, 2001) . It should be pointed out, however, that the surface area reported here is the one that can be directly ascribed to fish farming companies, which do have some kind of resource management practice. So, if we are also to consider public areas, which are not directly linked to any company, but where clams are collected anyway, often by unauthorized fishermen, then the figures will increase considerably, up to about 220,000,000 m2 . Clams are currently reared using naturally-produced juveniles, and only in some special cases by purchasing spat from artificial breeding. Since the Manila clam has adapted so well to the Adriatic lagoon environments, this species is now very widespread in those areas and farming activities basically consist of managing shoals and leased areas. This task includes soil cleaning and preparation, transfer of spat, thinning, assessment of present biomass, fishery rotation and surveillance. At present, there is only one commercial hatchery used for Manila clam seed production, situated in the Grado lagoon, which is connected with other areas employed for the successive growing stages. A recent study has shown that the national production of T . philippinarum in 1999 was about 61,800 tons, also including the product harvested outside farming areas; 50,700 tons come from the Veneto areas, 9,300 from Emilia-Romagna and 1,800 from the lagoons in Friuli-Venezia Giulia. In 1999, production of T . decussatus, almost exclusively from Sardinian ponds, was as high as 135 tons. Grooved carpet shell (T . decussatus) and Manila clam (T . philippinarum) are marketed in different ways, due to their respective production and market values. The 56 former is generally destined for the local market, mostly for various forms of direct sale, with a more limited involvement of different players. The latter has a larger market and a broader trading process, whose main protagonists are wholesalers and retailing through big department stores. In recent years, molluscs and their defensive mechanisms have been investigated with great interest, both for the ecological, social, economic, nutritional role of these organisms and for their many pathological microrganisms (Arzul et al., 2001) . In addition, bivalves represent very important carriers of some human pathogens such as hepatitis A virus or V ibrio species. The effectiveness of the mollusc defense system can modulate the properties of infective microorganisms, so mollusc-microorganism interactions assume importance as human health risk factors (Genthner et al., 1999; Lacoste et al., 2001) . Mollusc defense systems and their state of integrity also represent a suitable biomarker in the monitoring of polluted sites; for example, several studies indicated the possibility of a correlation between high metal levels and low rates of phagocytosis and, possibly, between release of high levels of oxygen radicals and low background of organic contaminants (Pipe et al., 1995) . Finally, characteristics of the immune system of molluscs could represent a simplified model for understanding more complex and phylogenetically more advanced systems. Invertebrates possess aspecific and innate immune mechanisms, and they are lacking in immune memory following the first encounter with a pathogen. Molluscs have humoral and cellular immunity, and the humoral system is constituted by lysosomal enzymes, agglutinins, lectins and antimicrobial peptides. Nevertheless, cellular immunity seems to perform the main role in shellfish immune processes (Roch, 1999) . Lysosomal enzymes (b-glucuronidase, acid and alkaline phosphatase, lipase, 58 aminopeptidase and lysozime) are comprised especially in granular hemocyte lysosomes and their release into serum depends on cell degranulation during phagocytosis (Pipe, 1990) . Lysosomal enzymes also possess digestive function as indicated by their abundance in the stylus and digestive gland. In fact, digestive and defense functions could act contemporaneously because filtered bacteria represent nourishment for marine bivalves, so enzymes present in the digestive tract hydrolyze microorganisms performing both functions at the same time. Agglutinins have been found in many mollusc tissues and these glycoproteins act as opsonins against different types of erythrocytes (hemagglutinins) and other cells such as bacteria, protozoa and algae (Chu, 1988) . Lectins represent agglutinin-like molecules and are present in the hemolymph and in the hemocyte membrane; they have a specific opsonizing function in mollusc defense mechanisms such as hemocyte aggregation and foreign cell agglutination. In particular, lectins are specific for hemocyte and bacteria glycoconjugates, so promoting recognition of foreign cells. However, the biological significance of the different carbohydrate specificities is not yet known, in fact most invertebrate lectins are heterogeneous and may also bind other ligands (Tunkijjanukij et al., 1998) . Antimicrobial peptides have been discovered in hemolymph and mussel hemocytes, and have been classified as defensins, mytilins, myticins and mytimycin. These compounds act on a rather broad spectrum of microbial organisms and their activity has been shown outside hemocyte or in phagocytosed bacteria (Mitta et al., 2000) . Hemocyte phagocytosis represents the main process of the entire cell defense system and is constituted by different phases including recognition, adhesion, ingestion, destruction and elimination of foreign cells. Mollusc hemocytes also have important functions in transport of nutrients, wound repair, and removal of catabolic products or pollutants (Feng, 1988) . Hemocytes can be generally subdivided into two main types: hyalinocytes, small cells with poor cytoplasm containing few or no granules; and granulocytes, large cells with a very abundant granular cytoplasm. Granulocytes are those most responsible for phagocytosis and they can be classified into acidophilic, basophilic and neutrophilic cells. The presence of these two hemocyte types and, possibly, of their subpopulation has been demonstrated through microscopic, immunological and enzymatic analysis in several molluscs such as Mytilus edulis, M. galloprovincialis, Mercenaria mercenaria, Ruditapes decussatus, Ostrea edulis, Crassostrea virginica, Cr. gigas (Hine, 1999) . Different hemocyte types isolated from many molluscs have the common characteristic of phagocytosing foreign cells or particles using lysosomal enzymes and respiratory burst. Mollusc hemocytes respond to appropriate stimuli with a respiratory burst resembling that of mammalian phagocytes. The respiratory burst activity consists of ROS synthesis. Superoxide, the hydroxyl radical, singlet oxygen, hydrogen peroxide, hypohalides, halidamines, nitric oxide and the peroxynitrite anion represent 59 the most important ROS and all are very strong oxidants. ROS production during phagocytosis of immune stimulant yeast (Saccharomyces cerevisiae), laminarin, phorbol miristate acetate and microbial LPS has been demonstrated for many different mollusc hemocytes such as Patinopecten yessoensis, Pecten maximus, Cr. virginica, Cr. gigas, O. edulis, Mya arenaria, M. mercenaria, M. edulis (Roch, 1999) , M. galloprovincialis (Arumugam et al., 2000) , and R. decussatus (Tafalla et al., 2003) . However, the exact role of ROS production remains to be clarified in natural and experimental infections with important mollusc pathogens (Roch, 1999) . In fact, O. edulis and Cr. gigas hemocyte do not show LuCL during Bonamia ostrae ingestion. In vitro studies have shown that Cr. virginica hemocytes fail to produce LuCL during Perkinsus marinus phagocytosis, but the addition of osmotically killed and structurally intact Perkinsus sp. cells triggered a large LuCL response, comparable to that shown with other foreign particles such as zymosan (Anderson, 2001) . M. galloprovincialis hemocytes do not produce LuCL during stimulation with V ibrio tapetis and Perkinsus-like organisms (Pseudoperkinsus tapetis) (Ordas et al., 2000) . The inability of hemocytes to inactivate pathogen organisms has negative effects on the health condition of molluscs and on the economic or commercial activity connected with marine farmed molluscs. Lysosome-phagosome fusion was not observed in hemocytes infected with Bonamia sp. and the lack of fusion may be due to membraneous vesicles originating from parasite lipoid bodies (Hine et al., 1994) . Bonamia sp. is mainly phagocytosed by agranular hemocytes and the higher resistance shown by some oyster strains to the disease is attributed to a reduced number of these target cells, thus suggesting the possibility of producing less susceptible oyster broodstock (Naciri-Graven et al., 1998) . The parasite P. marinus is responsible for widespread mortalities in oysters Cr. virginica farmed along most of the U.S. Atlantic coast. Oyster mortality usually peaks in late summer-autumn when combined high temperature (>25°C) and high salinity (at least 12%) favor development of heavy infections (Oliver et al., 1998) . High environmental temperatures affect and modify cellular and humoral activities and, at the same time, increase the metabolic activity of the parasite; so, the rate of killing may not be sufficient to cope with Perkinsus sp. multiplication (Chu et al., 1993) . Some authors report that P. marinus actively produces factors that act as free radical scavengers or as inhibitors of ROS generation by Cr. virginica hemocytes (Anderson, 2001) . In contrast Cr. gigas is not susceptible to P. marinus disease and it is hypothesized that the capacity of Cr. gigas to increase its protease inhibitors represents the key event in its resistance to parasite infection by neutralization of proteases secreted by P. marinus, thus preserving the oyster hemagglutinins from degradation. Such hemagglutinins will be ready to act as opsonins stimulating phagocytosis of parasites (Romestand et al., 2002) . ROS produced by the host during inflammation are highly oxidising compounds against host and microbial macromolecules. In particular, the hydroxyl radical, hydrogen peroxide and peroxynitrite anion have the most important role in protein denaturation. Damages on polypeptide chains evoke the synthesis of HSPs, both in host and in pathogen to repair denaturated proteins and to protect their respective 60 proteic pool from possible denaturation. A significant 70 kDa HSP synthesis in M. galloprovincialis hemocytes during V ibrio alginolyticus phagocytosis has been demonstrated, while Escherichia coli determines only a small increase in HSP 70. These results are explained by the different phagocytosis rates derived from the different mollusc behaviour on interaction with bacteria coming from dissimilar environments and confirm the existing relationship between phagocytosis and HSPs production (Tiscar et al., 1998) . Furthermore, the diversity in HSPs expression at conditions of high environmental temperatures seems to be an important factor promoting pathogenicity in the P. marinus-Cr. virginica relationship (Tirard et al., 1995) . Moreover, among the possible pollution biomarkers, the HSPs expression in molluscs has been proposed because the synthesis of these proteins is likely to be induced by a large number of chemicals (Bierkens, 2000) . A new interesting approach in invertebrate immunity studies in recent years involves research into the molecules that act as chemical mediators between immune system cells. These compounds also seem to possess an important role in bilateral information exchanges between the immune and neuroendocrine systems. These relationships are mediated by many types of molecules such as corticotropinreleasing hormone, ACTH, monoamines (epinephrine, norepinephrine and dopamine), glucocorticoids, free radicals, cytokines (IL-1, IL-6 and TNF-a), opioid peptides and opiates. Opioids are produced through proteolysis of large precursor molecules such as POMC, proDyn and proEnk. Effects of POMC derived peptides have been particularly noted in the vertebrate immune system, concerning antibodies synthesis, chemotaxis and phagocytosis (Stefano and Salzet, 1999) . M. edulis hemocytes contain mammalian-like POMC and its derivates such as methionine-enkephalin-like peptides, c-melanocyte-stimulating hormone, a-melanocyte-stimulating hormone, ACTH, b-endorphin and c-lipotropin hormone (Stefano et al., 1999; Ottaviani et al., 1997) . ACTH has been revealed in M. edulis hemolymph stimulated with LPS and selected ACTH fragments seem to modulate the mobility of hemocytes and, in parallel, to contrast stimulation of hemocytes with TNF-a. ProDyn and proEnk-like proteins have been characterized in M. edulis and proEnk, in particular, contain antimicrobial peptides such as enkelityn or peptides similar to peptide B and bovine amidorphin. The endocannabinoids receptor system has also been characterized in the mussel, together with opioids system. In fact, mussel hemocytes seem to possess a cannabinoid 1 receptor-like related to nitric oxide release. Briefly, peptide opioids give an amoeboid morphology to hemocytes, while the endocannabinoid anandamide inhibits this phenomenon through a nitric oxide dependent mechanism (Salzet and Stefano, 2002) . The cytokines represent the most important immune mediators, playing an essential role in mechanisms such as cell multiplication and differentiation, apoptosis and cytotoxicity. IL-1a, IL-1b, IL-6, IL-8 and TNF-a have 61 been isolated in M. edulis. IL-1 and TNF-a influence the mobility of M. edulis hemocytes and an increase in IL-1-like molecules has been shown during contact of hemocytes with a synthetic opioid analogous to methionine enkephalin (Stefano et al., 1999; Ottaviani et al., 1997) . In conclusion, comparative studies on immune invertebrates systems could facilitate understanding of the receptors involved in innate immunity. These receptors evolved in higher organisms with the main function of influencing the ''pain'' signal originating from danger. In addition, understanding the mechanisms of the mollusc immune system represents an important research objective in the protection of shellfish production from pathogens and help us to better understand the main innate biological processes that have maintained their characteristics through the evolution of different organisms. Lacoste, A., Malham, S., Cueff, A. and Poulet S., 2001 . Stress and stress-induced neuroendocrine changes increase the susceptibility of juvenile oysters (Crassostrea gigas) to V ibrio splendidus. Applied Environmental Microbiology, 67, 2304-2309 Mitta, G., Vandenbulke, F., Noel, T., Romestand, B., Beauvillan, J., Salzet, M. and Roch, P., 2000. Differential distribution and defence involvement of antimicrobial peptides in mussel. Journal of Cell Science, 113, 2759-2769 Naciri-Graven, Y., Martin, A., Baudc, J., Renault, T. and Gerard, A., 1998 In the 1990s EEC countries drafted several directives concerning the aquaculture sector, including Directive 91/67/EEC from 28.01.1991, accepted into Italy by DPR n. 555 on 10.12.1992 (''concerning the animal conditions governing the placing on the market of aquaculture animals and products'') and Directive 95/70/EC from 22.12.1995, accepted into Italy by DPR n. 395 on 20. 10.1998 ('' introducing minimum Community measures for the control of certain diseases affecting bivalve molluscs''). These norms have been and are still being continuously reviewed to meet the needs of this dynamically evolving sector. DPR 555 includes among its provisions a list of the diseases that are so serious for production that they must be eradicated from the EEC territory (Annex A, List I), as well as a list of other diseases that also affect production (Annex A, List II). The countries must become aware of the diseases that are present in their territories, to implement sanitary measures to limit their spread as much as possible and to implement eradication plans (acknowledging the areas or farms that are free from one or more diseases). To avoid the propagation of these diseases on EEC territory, the directive regulates how to reintroduce and market live animals according to the sanitary qualification of the area or farm. Special attention is given to aquaculture animals and products imported from third countries, for reintroduction, which must offer similar sanitary guarantees to those existing in EEC countries. Annex A, List II of DPR 555 derived from Annex A of 91/67/EEC, includes two mollusc diseases, namely Bonamiosis and Marteiliosis for flat oyster (Ostrea edulis) ( Table I) . Bonamiosis (also known as microcell disease and haemocyte disease of flat oysters), present in the Northern hemisphere, is caused by Bonamia ostreae. Besides Ostrea edulis, it can also infect other oyster species (O. conchaphila, O. puelchana, O. angasi and T iostrea chilensis). Bonamiosis is present in Europe along the Atlantic coast from Spain to the United Kingdom and in the USA. Most infected oysters generally look normal. The pathological alterations are represented by haemocyte infiltrations in various organs, in the digestive gland, the gonads, the gills and the mantle. The parasite multiplies in the haemocytes that phagocytosed it; when the haemocytes are affected by necrosis, most parasites can be released. The disease is transmitted from oyster to oyster. The initial phase can affect oysters of all ages. In the presence of an endemic disease, bigger oysters are most likely to be affected. The disease is present all year round and the induced death rate can reach 90%, increasing in the summer period. Stress conditions on the molluscs (turbidity, manipulation, etc.), as well as high concentration of oysters affect the appearance of the disease and its consequent mortality rate. No eradication systems have yet been developed. The mortality rate may be reduced by using a floating farming system and/or decreasing the population density. To market the product, in acknowledged areas and farms, EEC norms require a thorough sanitary control of the introduced material (to ensure it is free from disease). If Crassostrea gigas is introduced into acknowledged areas or farms, it must be carefully checked, especially if small, to avoid the presence of O. edulis. In Italy, where there are currently no flat oyster farms (apart from a few mainly experimental attempts), Bonamia ostreae is found sporadically and only in the wild. This situation is probably caused by farming attempts performed in the 1970s-1980s, that made use of infected foreign material, which infected the subjects present in the wild in the Adriatic Sea (Tiscar et al., 1991; Tiscar et al., 2003) . Flat oyster Marteiliosis is caused by Marteilia refringens, a Phylum Paramyxea protozoan. This disease is also known as Aber disease. It is present in Europe along the Atlantic coasts from Portugal to the United Kingdom. The infected oyster may look emaciated, will consequently slow down its growth and possibly die. The pathological changes include the presence of the parasite at various developmental stages, namely in the stomach epithelial cells, in the intestine, in the gills, but especially in the digestive gland tubules. The disease is transmitted naturally in oysters. The most favourable period for infection occurs in the summer, when the water temperature averages 17°C. Since no eradication system is devised, it is recommended not to transfer infected material to acknowledged areas or farms. In Italy the situation is comparable to that previously described for Bonamiosis. DPR 395 includes among its provisions the requirement that farm owners register their farm with the National Health Service and possess a loading and unloading register of the material introduced or destined for another farm for re-introduction. When the National Health Service finds a case of unusual death in the wild or in farms, it will envisage a sanitary monitoring programme to ascertain or exclude the presence of the diseases included in Annex A to 91/67/EEC or Annex D to 95/70/EC as the cause of death. On February 5th 2003 the EU Commission modified Annex D (2003/83/EC) (Table II) . Annex D originally included serious diseases not present in Europe. EU countries had to pay great attention not to introduce them, considering how difficult it is to carry out eradication plans in mussel culture. In the event of an introduction, it is compulsory to notify the EU of the centre of infection. Annex D was reviewed to adapt EU norms to OIE norms. In the annex to DPR 555 and DPR 395 all notifiable or meaningful diseases present in the OIE norms have now been included. As a consequence, many of these pathogens are present in Europe since some species are widely farmed or present in the wild (Ostrea edulis, Crassostrea gigas, Ruditapes philippinarum and Ruditapes decussatus). In this paper, only Perkinsiosis, present in Italy, will be briefly illustrated. Perkinsiosis is caused by a protozoan of the Perkinsus genus that, according to recent studies, belongs more to the Dinoflagellata class than Apicomplexa. In Europe it was reported as Perkinsus atlanticus in several mollusc species, especially Veneridae (T apes decussatus, T apes philippinarum, Venerupis pullastra, Venerupis aureus). In Italy it was detected in farmed T apes philippinarum (Ceschia et al., 1991) and in several wild mollusc species: T . decussatus and V. aureus (Da Ros et al., 1985 , 1986 Breber, 1985) ; O. edulis (Tiscar et al., 1992) , Chamelea gallina (Rubini et al., 1996) and Callista chione (Canestri-Trotti et al., 2000) . Recent genetic studies have shown that P. atlanticus is similar to P. olseni which infects several Haliotis species farmed in Australia, New Zealand, etc. The mollusc can probably be infected by every developmental stage of the parasite, although the most effective stages are probably trophozoites and zoospora. The parasite reaches the host's connective tissue by lysis of the basal lamina of the external epithelia (gills, mantle, digestive gland, as well as kidney, foot, gonads, intestine) or may be spread through the hemolymph. Infected molluscs do not normally present any clear external clinical characteristics, but in the case of a serious infection where the subjects may be present pale pulp colouring and weak consistency, greyish gills due to the presence of whitish nodules (parasitical cysts) may be observed. The presence of Perkinsus negatively affects numerous structures and physiological activities of the host causing alterations in the gonads, tissue degradation and disorganization, influencing tolerance to salinity, causing acidosis and limiting growth. At present researchers pay particular importance to this parasitosis. A large infection, in Europe, may cause mortality by itself. Its presence, even if limited, when the molluscs are stressed, increases the death rate. Korean researchers believe that P. atlanticus has damaged and drastically reduced clam production in Korea. High temperatures and salinity increase the damage and the intensity of the infestation. The host responds with an increased inflammatory reaction represented by a granular-type hemocyte infiltration. The parasite may be isolated (free or intrahemocytal) or enclosed in cysts made of a positive PAS substance. The fishing activity in the Campania region has its roots in a very old tradition and is favoured by a large combination of factors: the sea, the coast and especially the geological phenomena of the Flegrei lakes. In Roman times the Flegrei lakes belonged to the State. In that period the Lucrino and Averno lakes were fused together and because of their location they were chosen as the port of the imperial fleet. At that time the first fishing activities started and fishponds were built. Fishing activities achieved high levels in productivity and wealth very quickly, but in 1538 everything suddenly stopped because of a volcanic eruption. Only in 1770 under Ferdinando IV of Borbone did the fishing activity receive a vital new impetus. Now these lakes are thoroughly exploited for aquaculture and oyster and mussel cultures. In 1973 the mussel culture suffered an abrupt interruption because of a cholera epidemic. All fishponds of the Naples gulf were destroyed as the source of the epidemic was attributed to mussels. Subsequently this activity recovered slowly and today there is a large number of fishponds on the east coast of Naples (Torre del Greco, Torre Annunziata, Castellammare di Stabia), on the west coast (Baia, Cuma, Bacoli, Villa Literno) and on the Naples gulf. The most important fishponds are located: (a) in Baia, with an area of around 16 square kilometers. The production declared for the last year was of about 9500 quintals mussels; (b) in Castellammare di Stabia with an extension of 1.6 square kilometres and declared production of about 1000 quintals; (c) in Torre del Greco where the fishponds cover an area of about 0.8 square kilometers with a declared production of 250 quintals; (d) in Fusaro lake the surface dedicated to mussel culture is around 0.3 square kilometers and the official production is of about 2500 quintals. 70 The official average annual mussel productivity of the Campania region is of about 15,000 quintals which are mostly absorbed by the local market. However assessments by the ASSOMITILI (the most important producers association) show a real production of around 30,000 quintals. These data highlight the notable difference between the potentiality of the fittings and the declared production. Therefore a notable quantity of product is introduced into the market illegally without sanitary controls. Neither the sanitary authorities nor the policy authorities are able to control such massive abuse. In Campania there are nine bivalve mollusc dispatch centres and five dispatch/ purification centres and the species of bivalve molluscs commonly commercially exploited include Mediterranean blue mussels (Mytilus galloprovincialis), various clams including the native clam (T apes decussatus), manila clam (T apes philippinarum), razor shell clam (Ensis spp.), flat oyster (Ostrea edulis), donax clam (Donax trunculus) and the striped venus (Venus gallina). Bivalve molluscs can be the vehicles or agents of different parasitic, bacterial and viral diseases as well as biotoxins and contaminants (Jaksic et al., 2002; Chironna et al., 2002) . Although indigenous marine viruses are very abundant, only human viruses have been associated with illness due to seafood consumption. Viral problems are limited to the role of food in recycling human viruses back to humans. The viruses most adapted for this purpose are those transmitted by the fecal-oral route and include viral agents causing human gastrointestinal diseases such as hepatitis A virus (HAV) and the polio virus. These viruses contaminate shellfish through water contamination (i.e sewage pollution of the marine environment), or inadequate hygiene practices during processing and handling (Lees, 2000) . Virus accumulation by shellfish is a complex process and depends on several factors among which temperature is one of the most important. Viral and bacterial accumulation under natural conditions changes significantly through the year. In clams the greatest accumulation of microorganisms occurs at temperatures ranging from 11.5 to 21.5°C (Jofre, 1992) . The quality of shellfish depends on the quality of the growing waters and on the process of the shellfish self-cleaning mechanism (depuration). These factors are considered to be effective for the reduction or elimination of bacteria but they are not very successful in preventing viral transmission (Jofre, 1992) . Mussels are the most important disease vectors in Campania region. The association of infectious illnesses with this species reflects the traditional consumption of raw or lightly cooked mussels. Shellfish bacterial contamination is currently evaluated by means of coliforms, E. coli and Salmonella spp. counts according to the EU shellfish microbiological standards. However, bivalve molluscs classified as moderately polluted according to EU standards have been found to be contaminated with HAV and enterovirus. In England an outbreak of viral gastroenteritis (Norwalk agent) was caused by the 71 ingestion of oysters that had previously been depurated according to EU sanitary regulations. In the Naples area, since 2000 the Veterinarian Services have introduced strong controls against the illegal sale of molluscs as a consequence of an increase of HAV outbreaks observed by the Epidemiological and Prophylaxis Service. In that year 15 samples of bivalve molluscs were examined for the presence of HAV. Ten samples were negative for HAV while the remaining 5 samples (3 mussels, 1 donax clam and 1 manila clam) were found to contain viral RNA coming from a whole virus. One of these samples came from the Campania region. A similar sampling was carried out during 2001. Three samples (one mussel and two clams) out of nine were found to contain viral RNA coming from a whole virus. At the same time the Veterinarian Service carried out a microbiological screening on shellfish. One sample from the Lazio region out of a total of 29 was found not to conform to all EU microbiological standards. These results reinforce once again the inadequacy of bacteriological standards in the assessment of viral contamination. In the revised draft Regulation of the European Parliament and of the Council laying down specific hygiene rules for food of animal origin the bivalve molluscs chapter is one of the most important. This Regulation lays down general rules for food business operators on the hygiene of foodstuffs, taking into particular account the fact that the primary responsibility for food safety rests with the food business operator and that he needs to ensure food safety throughout the food chain, starting with primary production. The provisions of this Regulation apply to primary production and the associated operations (transport, storage and handling of primary products at the place of production, provided that this does not substantially alter their nature). As far as possible, food business operators, by the means of guides to good practice and the application of HACCP principles, have to ensure that primary products are protected against contamination (taking into consideration any processing that primary products will subsequently undergo). In Chapter Two (Hygiene requirements for the production and harvesting of live bivalve molluscs) several requirements were established for (a) production areas, for (b) harvesting and handling after harvesting and (c) for relaying. In the case of production areas the requirements are similar to those fixed by Directive CE 492/91. In addition permitted treatment methods are specified, such as sterilisation in hermetically sealed containers and heat treatment, that live bivalve molluscs must undergo to eliminate pathogenic microorganisms, when they have not been submitted to purification or relaying. In addition food business operators must not produce live bivalve molluscs in, or harvest them from, areas that the competent authority has not classified, or which are unsuitable for health reasons. Food business operators must take into account any relevant information concerning areas suitable for production and harvesting, including information on environmental and weather conditions. They must use this information to determine the appropriate treatment to apply to harvested batches. For harvesting and handling after harvesting food business operators 72 have to conform to other requirements, such as adequate protection of live bivalve molluscs from crushing, abrasion or vibration and against extreme temperatures and preventing their re-immersion in water that could cause additional contamination. When carrying out conditioning in natural sites, they have to use only areas that the competent authority has classified as being of class A. There must be a minimum distance between relaying areas, and also between relaying areas and production areas, so as to minimise any risk of the spread of contamination. The food business operator must use techniques for handling live bivalve molluscs intended for relaying that permit the resumption of filter-feeding activity after immersion in natural waters. The relaying period must be fixed depending on the water temperature. This period must be of at least two months' duration unless the competent authority agrees to a shorter period on the basis of the food business operator's risk analysis. The 'all in, all out' system must be used, so that a new batch of molluscs cannot be brought in before the whole of the previous batch has been removed. Finally, the food business operators managing relaying areas must keep permanent records of the source of live bivalve molluscs, relaying periods, relaying areas used and the subsequent destination of the batch after relaying, for inspection by the competent authority. Another important chapter is the Fifth (health standards for live bivalve molluscs) in which it was fixed that, in addition to ensuring compliance with the microbiological criteria adopted in accordance with the Regulation on the hygiene of foodstuffs, the food business operators must ensure that live bivalve molluscs placed on the market for human consumption meet the known standards. Many of these standards are the same as those in the current legislation, such as an adequate response to percussion and normal amounts of intravalvar liquid. In addition molluscs must not contain marine biotoxins in total quantities (measured in the whole body or separately in any edible part) that exceed the following limits: for Paralytic Shellfish Poison (PSP), 800 micrograms per kilogram; for Amnesic Shellfish Poison (ASP), 20 milligrams of domoic acid per kilogram; for okadaic acid, dinophysistoxins and pectenotoxins together, 160 micrograms of okadaic acid equivalents per kilogram; for yessotoxins, 1 milligram of yessotoxin equivalent per kilogram and for azaspiracids, 160 micrograms of azaspiracid equivalents per kilogram (Proposal for a Regulation of the European Parliament and of the Council laying down specific hygiene rules for food of animal origin, 2002). In our opinion, and in agreement with other authors (Jofre, 1992) , to reduce public health hazards associated with the consumption of shellfish, two kinds of measures should be considered: those regarding consumer education and information and those aimed at reducing the contamination to the lowest possible levels. It is important to educate the public (consumers, food manipulators) and high risk groups (i.e. diabetics, immunodeficients) about the risks deriving from consumption of raw or undercooked shellfish. This custom is one of the most frequent risk factors for acquiring HAV for all age groups in our country (Mele et al., 1997) . But it is also important to implement the procedures based on the HACCP principles, which, together with the application of good hygiene practice, should reinforce the responsibility of food business operators. Moreover, the present regulations must be enforced, for example through the introduction of new indicator organisms. New techniques, like molecular biology, for virological analysis of shellfish have to be adopted to establish the safety of growing water and the efficacy of the depuration process. These techniques must be adequately validated before routine use. Finally interdisciplinary collaboration between different sectors of food hygiene is important to guarantee an effective control. With regard to this it is interesting to notice that in the year 2000 the researchers of the Zoological Station in Naples identified a saxitoxin and a neosaxitoxin in a microseawood (genus Alexandrium) ( Fig. 1 ) as well as domoic acid in a diatom (genus Pseudo-Nitzschia) (Fig. 2 ) collected in the Naples Gulf. Since then the veterinarian Service has been checking for the presence of these toxins in samples of bivalve molluscs sold in town. At the moment all results are negative. . So the CNITV deals with all questions related to veterinary toxicology except for veterinary drugs but including drugs for humans. The Laboratory of Toxicological Analyses performs various types of analyses on samples sent by veterinarians from the French territory, often after a contact with the Center. There are strong relations with the other disciplines of the Veterinary School, allowing for a rapid and quite efficient multidisciplinary approach of the cases and with the research activities in ecotoxicology of the department (bio-indicators, relation between biochemical toxicology and genetics). The main objectives of the CNITV are: -to answer phone calls (24 h a day) or mail regarding any toxicologic question (diagnosis, treatment, prognosis evaluation, toxicological, pharmacological and environmental information on chemicals or noxious agents (plants, animals), 76 including differential diagnosis and help in case solving (samplings, contacts with specialists). -to generate a field-cases database gathering all the questions, suspected or confirmed poisoning cases. -to maintain scientifically accurate and up-to-date literature and manage a specific bibliographic database in order to analyze the cases recorded by phone or by mail. -to participate in the initial and permanent education of veterinarians in the fields of toxicology (clinical, environmental). In order to achieve its objectives, the Center has been developing various resources: -human resources: a permanent veterinarian retributed by the center, a part time documentalist, a part time secretary, teachers and veterinarians acting benevolently, students of 2nd and 3rd cycle. -computer databases: 3 specific systems: V-Tox is used to manage all the cases recorded by the Center (120,000 cases), Sentinel-vetA only deals with the management of veterinary drug related cases and Toxvet is concerned with the bibliographic references. -a specific library of toxicology. Collaborations are diverse: -with other veterinary toxicological centers in France such as the one at the Veterinary School of Nantes or abroad (Pisa and Parma in Italy, IAPIC in the USA) -with the veterinary professional organisations -with human poison centers and the network of toxicovigilance -with public or private organisations of animal welfare, food safety and environment protection -with the industry of plant protection, animal feed. We will present here and discuss some statistical data describing the activity of the Center based on the present volume of cases (about 10,000 cases in year 2002 introduced in the database on a total number of 15,000 calls) and on the evolution observed in those recent years which shows a constant rise in the number of calls (from 4000 in 1991 to 15,000 in 2002) and the computerized cases. About 75% of the calls come from veterinary practitioners. Territorial services (Laboratories and Veterinary Offices) represent about 5%, animal owners (pet owners, farmers) about 20%. Human Poison Centers call often, especially for cases concerning children exposed to veterinary products or pollution affecting together human and animal health. T ype of calls ( Fig. 1) . Calls mostly are considered as emergencies or occur following an intoxication. In the first case the veterinarian is handling an animal which is, or may be, intoxicated and he asks about the treatment that should be instituted (60%). In the second case (27%), the problem previously occurred and the veterinarian wants to get more information about it (for example what samples to confirm the intoxication, what inquiry on the field). 8% of the calls ask for information on a product, often in Figure 1 . Motives for calls to the CNITV 78 order to prevent a risk of intoxication. Other questions concern residues or assistance in an expertise. Species involved (Fig. 2) . Small animals represent the highest proportion, dogs mostly (68%) due to the important canine population and to their propensity to chew and swallow almost anything. Cats (18%), despite their natural mistrust, increase regularly in number of calls, due probably to the expanding medicalisation. The numbers of cases concerning new companion animals (rabbits, snakes) are sharply increasing. Bovine (4%) and other farm animals represent a smaller percentage; despite the regular decrease in percentage, the number of calls is quite constant from one year to another. Besides animal health considerations, there are often concerns on chemical residues in animal products (milk, meat, eggs). Calls concerning humans are quite numerous (about 5%). Cases on wild terrestrial or aquatic animals are less numerous but often complex; they are often investigated by the laboratory. The presentation of the different toxic agents involved or suspected in the diverse animal species is not an easy task: there are obviously great differences of exposure and sensitivity between species; moreover, the causality assessment is often difficult. We developed for several years a probability scoring (or imputability assessment) of the cases in 6 categories based on criteria of exposure, chronology and clinical features, sometimes laboratory analyses; this allows us to classify the cases as quite probably linked to the toxic agent (which represents approximately 30% of the calls, Droguet, 2002) or not. Still the causality assessment is not an ''exact science'' due to the uneven reliability of the data collected during or after the call and to the weak knowledge on the toxicity of many agents in the field conditions; it should regularly be reviewed on the basis of recent information. Quite often, a toxicological possibility is proposed by the Center on the basis of symptoms described, which obviously may lead to an uncertain assessment of the case. For the different types of agents, the probability may be very unequal. For example, the implication of pesticides is less frequent than generally suspected by the callers. The current knowledge by the veterinarian of a toxicological issue may affect greatly the number of calls: for example, they call rarely for strychnine or metaldehyde intoxications or for toxic animals (snakes, caterpillars). So the statistical representation is certainly not to be considered as an exact picture of the intoxications in the field. However, the consideration of the data from the laboratory or specific surveys help to build a more exact representation. Only large trends shall be presented here. Globally, the repartition between the principal groups of agents is indicated in Fig. 3 . Pesticides represent about 45% of the calls. The proportion is higher for herbivores than for carnivores and the different types of pesticides have various occurences in different species. Circumstances of intoxications are diverse; accidental via ingestion Insecticides cases (about 45% of pesticide calls) are dominated by organophosphates and carbamates, followed by pyrethroids. Herbicides are more frequently involved in cattle, although commonly used herbicides such as glyphosate are frequently encountered in small animals as well (Burgat et al, 1998) . Rodenticides and molluscicides are very often involved in small animal intoxications (in about 50% of pesticides calls) especially antivitamins K (22% of pesticides calls) such as difenacoum and bromadiolone which appear most frequently. Old products are still frequently involved such as chloralose (4-5% of total small animal calls) or metaldehyde (about 1.5%). Wild animals are often affected by pesticides as shown by the data from the center, but with a better accuracy from the laboratory. There are specific programs of surveillance on terrestrial fauna (Sagir program) and on fish. The category ''Pollutants'' represents about 20% of the calls, with various types of compounds such as metallic or organic substances, household products, exposure to wastes etc. Environmental pollutions more often concern production animals, especially bovines (lead, copper, dioxins, aromatic hydrocarbons, etc.); they often raise questions on residues in the animal products which implies their quantification and evaluation. A specific surveillance of cases related to exposure towards epuration waste recycling in agriculture (Ademe program). Wild animals, especially fish, are also concerned; a special surveillance may be set up with environmental organisations. Small animals are often affected by domestic products (detergents, disinfectants especially hypochlorite, paintings, anti freeze). Heavy exposure to petroleum is frequent in cats falling in a tank. There is an interesting correspondence with the risks observed in young children to the domestic chemical environment. Plants represent around 11% of the calls. Small animals often ingest domestic plants (Ficus, Dieffenbachia) or garden plants (Laurel, Yucca) . In cattle, predominant toxic plants are Taxus, Mercurialis, Galega especially in sheep, Amaranthus, acorn, bracken fern. Some plants are quite typical of a region such as Ferula in the South, especially Corsica. Mycotoxins are often suspected especially in subacute episodes, but difficult to assess. Toxic animals represent a great number of intoxications, essentially in dogs. But the number of cases registered is probably not representative, due to the good knowledge of the problem by the veterinarian. Toads, Vipera, venomous insects are the most frequently cited. Caterpillar Thaumatopea represents a major problem in the dog; it may also affect humans and shows an interesting geographical evolution from the South to the North of France. Up to year 2000, veterinary drugs were included in our statistics altogether with ''human drugs'', the total representing around 18% of the calls. Since 2001, the official system of veterinary pharmacovigilance implies a specific management of the cases concerning veterinary drugs (about 2000 cases in year 2001). So toxicovigilance deals only with drugs for humans involved in domestic animals. The exposure may be via accidental ingestion of badly stored pills, which is very frequent in dogs with neuroleptics (especially benzodiazepines such as bromazepam), anti-inflammatory agents (ibuprofen), contraceptive pills. Obviously, the risks are very different for those different categories; countermeasures including antidotes were developed especially for diazepines intoxications (Bertini et al., 1995) . Voluntary administration of drugs such as paracetamol or aspirin in cats, or loperamide in dogs of collie breed is another important source of drug toxicity, which may find interesting relations with pharmacovigilance. Pet food tolerance represents a growing concern with several cases registered. These are ususally quite difficult to assess and relate often to individual (or family) susceptibility. In dogs, chocolate ingestion is a common intoxication. Production animal feed problems are probably underscored in our data because they are managed by nutritionists; they may relate with ionophore additives (monensin) for example in horses or in turkeys, but the incidence decreased due to an improvement in the incorporation procedures. In about 10% of the calls, the Center proposed an hypothesis of a non toxic disease on the basis of chronology and semiology. The caller was usually directed to a specialised laboratory to confirm the hypothesis. The activities of a Veterinary Poison Center can initiate early help for diagnosis and appropriate treatment of animal intoxications. They also contribute to the initial and permanent education of veterinarians. Careful analysis of the clinical and epidemiological experience is useful to determine current and future trends in field toxicology. This approach of toxicovigilance may also be used to investigate the impact of recently marketed chemical formulations The Italian Veterinary Toxicologic Assistance Service (Servizio di Assistenza Tossicologica Veterinaria, S.A.T.V.) was established in mid 1996 at Turin. It differs from other European Poison Centres in that it has been promoted and is operated by a cultural association of veterinarians devoted to small animal practice (Società Culturale Italiana Veterinari Animali da Compagnia, S.C.I.V.A.C.) under the scientific advice of the Pharmacology and Toxicology Unit of the Department of Animal Pathology of the University of Turin. Other special characteristics of the Service are that it may be accessed only by licensed veterinarians and is completely free of charge. Of course, this means that it needs to be sponsored and given grants by private companies. The main aims of S.A.T.V. are the following: -to collect relevant toxicological information from different sources (assessed field cases, scientific literature) and to store it in a computerised database; -to provide real-time telephonic assistance to colleagues facing toxicological emergencies in dogs and cats; -to perform epidemiological studies based on the telephone enquiries; -to promote the exchange of information with other national or international Poison Centres. The S.A.T.V. is active 24 h a day; the staff is made up of group 12 colleagues with special training in veterinary clinical toxicology who have access to a dedicated computerized database which is periodically updated on behalf of the staff of the Pharmacology and Toxicology Unit of the Department of Animal Pathology of the University of Turin. Xenobiotics have been divided into the following classes: veterinary and human drugs, agrochemicals (insecticides, herbicides, fungicides), rodenticides, molluscacides, miscellaneous, halogens, metals and antidotes. Each file bears the commercial trade names of drugs, pesticides or products containing the active principle(s), lethal and/or toxic doses for both dogs and cats (or laboratory animals), clinical signs, clinicopathology, laboratory diagnosis and antidotal and/or supportive 84 treatment. Phone calls are recorded and stored in the database and a file is automatically generated for subsequent retrieval containing information regarding the (geographical) region of the caller, the species involved (cat or dog), as well as the age, sex, breed and environment (urban or rural) of the affected animal. The present work shows all results collected from the SATV from May 1996 to May 2003. It is necessary to specify that calls are stored in the database only when they are not too generic and refer to a known active principle. Calls for information queries have not been considered in the present report. Since May 1996 the SATV has stored 4297 calls coming from all over Italy on its data base. Calls come especially from the North (66%), probably because of a better knowledge of the service by veterinary practitioners. Enquiries from Central and Southern Italy make up the remainder of calls (34%). The SATV has mostly been contacted for problems involving dogs (80% of total calls). The lower incidence of calls concerning cats may be attributed to the more ''prudent'' feeding habits characterising this species together with its tendency to stay by itself when feeling ''sick'': this happens especially for individuals living in country areas or in a semi-domestic state. According to our data, poisonings appear to be relatively more frequent in urban (57%) than in rural areas, not only because of the large array of potentially toxic household products, but also because poisonings affecting animals living in a restricted area such as a flat are more easily detected. A breakdown of all enquiries by class of chemicals is depicted in Table I . Veterinary and, to a far lesser extent, human drugs are responsible for the greatest percentage of calls in both species: namely 23% for dogs and 35% for cats. Within this class, antiparasitic compounds account for (as many as) 64% and 79% of the calls in dogs and cats, respectively, followed by non steroidal antiinflammatory agents (10% for dogs and 7% for cats). The majority of the enquiries concerning antiparasitic drugs refer to principles active against external parasites, such as pyrethroids and organophosphorus compounds or carbamates. In most cases misuse by the animal's owners or the accidental ingestion of commercial preparations such as collars, powders or sprays containing such compounds has been the cause of poisoning. In this respect it is worth noting that, on a percentage basis, pyrethroids (mostly permethrin) account for nearly half of the calls concerning this subclass in the cat, thereby confirming the particular sensitivity of this species toward these derivatives (Martin and Campbell, 2000) . According to our data, the improper use of ivermectin (49 calls) may still represent a problem not only for Collie dogs, which are reported to be most susceptible to this drug but also for other dog breeds as well as cats (Nelson et al., 2003) . Chlorpyriphos (47 calls for dogs and 12 for cats), Amitraz (38 calls for dogs), and Fipronil (28 calls for dogs and 21 for cats) are the subject of a significant number of inquiries; the latter is normally very well tolerated but multiple applications of very Continued high dosages (particularly to puppies and kittens) may give rise to nervous stimulation which is related to the blocking of GABA receptors brought about by this drug (Hainzl et al., 1998) . As regards the NSAIDs, ibuprofen, paracetamol, and aspirin account for the majority of the inquiries for dogs, and paracetamol and aspirin for cats, which are highly sensitive to these drugs. In most cases animals were poisoned as the result of the accidental ingestion of tablets intended for human therapy. Finally, hormones and psychoactive drugs are implicated in a small percentage of calls in dogs (4%), the former class being entirely represented by oestrogen-based oral contraceptives, the latter by hashish (15 calls) and Cannabis indica (11 calls). Agrochemicals are responsible for a significant percentage of enquiries for both species (19% of the total calls for dogs, 14% for cats). Insecticides are the most represented class (40% for the dog, 55% for the cat). For dogs the most implicated compounds are carbamates, especially Methomyl (39 calls) and Carbaryl (22 calls), organophosphorus compounds with particular regard to Diazinon (28 calls) and Malathion (21 calls), and organochlorine compounds, with 31 calls regarding Endosulfan. For both species, fungicides account for nearly one-fourth of the calls of the agrochemical class and the majority of these (43% for dogs and 52% for cats) are related to the accidental ingestion of copper sulphate. As regards herbicides, the most represented active principles for both species are Paraquat (47 calls for dogs and 5 for cats), mostly as the result of malicious poisonings, and Glyphosate (85 calls for dogs and 14 for cats), which is usually reported to affect animals ingesting grass and vegetation recently sprayed with a glyphosate-based commercial preparation. In this respect it should be noted that a surfactant present in many liquid preparations may be responsible for most of the reported toxic effects, including vomiting, diarrhoea, and nervous symptoms (Burgat et al., 1998) . Miscellaneous agents account for one-fifth of the calls in both species (20% in dogs and 23% in cats). The high percentage of enquiries is probably due to the great number of compounds included in this class (household products, solvents, hydrocarbons, zootoxins, and so on). It is important to stress the considerable role played 87 by zootoxins (136 calls), particularly as regards venomous bites by V ipera spp.; over 95% of these cases involve dogs. Household products (particularly sodium hypochlorite and detergents) are also responsible for a large number of enquiries, thereby confirming that companion animals (like human beings) may be daily exposed to serious dangers even in a relatively ''safe'' environment like a flat. In addition, our centre has received 20 enquiries regarding exposure to petroleum distillates (petrol, diesel, etc.) in cats, which are more prone to ingest considerable amounts of these derivatives through grooming. Rodenticides represent another class of xenobiotics heavily involved in pet toxicoses (19% of the total calls for dogs and 5% for cats). The higher incidence displayed by the canine species over the feline one is most likely attributable to the less selective feeding habits of dogs coupled with the malicious use of these compounds as baits. Anticoagulants account for the majority of these calls, especially in dogs (77%), with an extremely significant role played by the so called ''second generation'' molecules, particularly Bromadiolone (109 calls for dogs and 11 for cats) and Brodifacoum (89 calls for dogs and 4 for cats), which are far more dangerous than those belonging to the ''first generation'' (e.g. Warfarin) because they may cause fatal poisonings even after a single exposure (Petrus and Henik, 1999) . Older but still largely employed rodenticides like zinc phosphide (52 calls) and strychnine (39 calls) still represent a considerable threat to dogs. Despite the relatively low incidence of enquiries concerning molluscicides (8% and 3% of the total calls for dogs and cats, respectively), this class includes very toxic active principles like Metaldehyde and Methiocarb (Puyt, 1995) . Indeed, Metaldehyde represents the first cause of enquiries for dogs (183) and a significant number of calls (103) were also concerned with Methiocarb, suggesting the low efficacy of the repellents that are usually added to commercial preparations. As far as toxic plants are concerned, our data point to a greater sensitivity of the cat (11%) as compared to the dog (5%). Unlike dogs, felines seem to be more prone to be poisoned by house-or ornamental plants, which are involved in nearly 70% of all enquiries related to phytotoxins, with Euphorbia pulcherrima (Poinsettia) and Ficus spp. accounting for the majority of the calls. According to our data, dogs are more frequently affected by wild plants (60% of the calls in this class), although a significant number of cases have involved the accidental ingestion of tobacco. Finally, metals (especially lead and iron salts) no longer seem to be a primary cause of poisoning in domestic carnivores, representing 4% of the total enquiries for both species. Interestingly, mercury (35 calls) accounts for more than one third of the calls pertaining to this class in the dog as the result of the accidental ingestion of thermometers, although the metal content is normally too low to pose relevant clinical problems. When considering the total number of stored calls, it should be stressed that, unlike similar Centres, the work of S.A.T.V. is currently restricted to cases of poisoning 88 concerning domestic carnivores and the team is not allowed to address enquiries from non licensed veterinarians. In addition, the Centre is not yet equipped with its own facilities for performing the toxicological analyses. As regards the reliability of the generated epidemiologic data, one should take into account that the practitioner is usually more likely to contact the Centre when facing a case involving a ''new'' or an unusual poison rather than a known active principle. This may at least partly explain the relatively low percentage of the calls concerned with ''classic'' poisons such as for instance strychnine, Paraquat, organophosphorus or lead and perhaps the relatively high frequency of the enquiries about (some) ''emerging'' poisons such as Glyphosate, Methomyl and Endosulfan. Nonetheless, the collected data indicate how dangerous the misuse of the veterinary drugs can be (in particular that of external antiparasitic agents), since they represent as many as one third and one fourth of the total calls for cats and dogs, respectively. The potential danger of certain agrochemicals and of many cleaning products also has to be stressed. Finally, one should also emphasize the high incidence of calls related to second generation anticoagulant rodenticides. Since most cases arise from the ingestion of baits, in our opinion the free sale of commercial preparations containing these very toxic active principles should be restricted or even no longer be allowed. In contrast with France, where the Centre National d'Informations Toxicologiques Veterinaires (CNIVT), has been operating for several years to prevent poisoning, giving information about treatments, performing toxicological analysis and, above all, giving epidemiological predictions to vets, associations and/or citizens, Italy has not yet developed such a structure able to coordinate knowledge in the toxicological field. So, to date, in our country, incidence of poisoning in domestic animals has only been partially evaluated in isolated studies. This deficiency is reflected by the exact knowledge of these phenomena. In fact, in Italy the incidence of poisonings in domestic animals has only been partially evaluated through isolated studies and the data reported are not representative of reality as they generally refer only to analytical results obtained from IIZZSS, from Toxicological Veterinary laboratories of Faculties of Veterinary Medicine or referred to on-phone surveys results (SATV). Moreover, these studies particularly consider Central and Northern Italy, and the data relating to Southern Italy are scarce. Therefore, there is a critical lack of information for practitioners in this field and a real risk of misevaluating this important phenomenon. So, the present paper tries to define, in a better way, the incidence of poisonings in companion animals in Italy and to improve knowledge of this phenomenon. The presentation is divided in three parts. In the first one F. Caloni In the period between January 2000 and December 2002, our active collaboration permitted the recording and classification of data, and also a ''follow up'' of the veterinary cases. Our epidemiological investigation takes into consideration animal intoxications from Northern Italy, considering a total of 342 contacts. 34.7% of the calls received by CAV were from owners of animals and 65.3% from vets. 90.5% of the total consulting in the CAV was related to dogs and cats, and only a low percentage related to other species like bovine, equine, birds, rabbits, swine, ovine and exotics. 55% of dogs and cats had an age of less than 2 years, 34% from 2 to 8 years, and 11% over 8 years. 43% of the sample were males while 57% were females. For the categories of toxics responsible for poisoning pesticides are the highest class, followed by household products, drugs, plants, industrial products and animal poisonings. The class ''other'' includes abuse substances, food, and cosmetics. Rodenticides (36%) are most responsible for intoxication in the class of pesticides, followed by insecticides (25.6%), herbicides (24.3%), fungicides (8.1%) and molluscicides (6%). The ''households'' class consists of detergents, but also batteries, solvents and thermometers. Industrial chemicals including fuels, heavy metals and ethylene glycol represents the causative agent most involved in the dog. 74% of the cases of the toxic agent happen in the domestic environment, either in the house or outside (garden). 13.5% of poisonings happen outside the domestic environment (open space). Only a small percentage of intoxications have been verified in the place of work or in public places. Respect to the frequency of exposure, 71% of the subjects came into contact with the toxin only once, 3.5 more than once, and the frequency was unknown for the remaining 25.5%. The occurrence reliability, that is the assurance of the owner of exposure to the toxin, was certain for 51.4% of cases, probable for 27.3% and unknown for 21.3% of the intoxications. The cause of intoxication was accidental for 81.92%, iatrogenic for 3.32%, fraudulent for 2.58% and unknown for 12.18% of cases. The most frequent route of intoxication was by ingestion and was not reported on the schedule for 6.9%. 52% of the subjects were not tested by laboratory analysis; and we did not receive information for 29.5%. The result was reported for 46.8% of cases, with recovery for 42%, permanent disability for 1.3% and in 3.5% of the cases the animals died. The toxicological risk was assessed as ''high'' for 14.5%, ''medium'' for 16%, ''low'' for 8.2%, ''not excluded'' 28%, ''excluded'' 10.6%, while for 22.7% this was not assigned. Comparing the incidence of the type of poisoning in this investigation with the previous survey, where pesticides were the most frequent poisoning class, it has been observed that this is now the rodenticides. From the analyzed data it appears that zinc phosphide is a frequent causative agent of intoxication in 91 dogs. Poisoning by plants is also increased as compared with the last survey of the CAV. Intoxication by plants took place in the garden, and oleander was the most responsible plant, in the flat Euphorbia Pulcherrima was the most involved, once again in dogs. Under the class ''drugs'' we intend only products for human use, imprudently left by the owner within reach of animals. Some of these intoxications showed a seasonal trend, like the increase in intoxications by antihistaminic drugs in spring. The comparison between dogs and cats demonstrates that for the cat the domestic environment, in the house, is the place were intoxications most easily happen (62.5%). This could be related to the life-style of these animals. In Italy cats living in urban areas normally do not have the possibility to go outside from the house. The most frequent route of exposure to toxins is by ingestion. It is very important to underline that for 53.2% of cases the result of the intoxications remained unknown, despite our strict follow-up. We can conclude that our investigation for 2000-2002 related to the intoxications in Northern Italy, gave important and useful results for operators and demonstrated the utility of ongoing collaboration with the Poison Control Center, University, Veterinarians, and other organizations to give rapid and efficient therapeutic aid, and also to create a up-to-date database of intoxications in domestic animals. Since 1996, when it was founded, the Laboratory of Toxicology of the Veterinary Clinic Department (University of Pisa) has carried out a service of veterinary toxicological diagnostics and annually receives many requests for consultancy from selfemployed veterinarians, government corporations and private citizens, regarding suspicious individual or general poisonings in domestic or wild animals. In the period from 1 January 1999 to 30 June 2003 the Laboratory received 746 samples for a total of 1248 requests for analyses, 492 of which were found to be positive (40%). Based on clinical symptoms or anatomo-pathological exam, the veterinary or Laboratory personnel suggest looking for one, two or more compounds, suspected to be the cause of the toxicological event, for a medium value of 1.7 analyses of samples. The number of different toxic substances for which analysis has been requested and the percentage of those which really resulted positive from 1999 to the present have revealed how cholinesterase inhibitors play a role of primary importance. As a matter of fact, these substances are easily available in many formulations and it is possible to buy them without any particular restriction. Rodenticide poisonings caused by anticoagulants are less frequent; however they still show a high positivity. With these accidental poisoning seems more probable, due to struggles against rodents. As regards the frequency of poisoning due to zinc phosphide, its incidence is not particularly high and remains the same during the entire period taken into consideration. 92 Metaldehyde poisoning (percentage of positivity of 11% in 1999) has increased over the years, reaching 31% in the first six months of 2003. Metaldehyde is a much-used molluscicide and is easily available, whose use has increased to prepare poisoned baits. A particular case is represented by strychnine which is known as a very dangerous poison widely used in the 80s: this now justifies the high number of requests. However, its employment is nowadays under strict legal restriction and limits its real effects on the total number of analyses found positive ( Table I) . Investigation of the positive results for toxicants examined obtained for biological material from different species has underlined the definite predominance of the dog as the animal species that risks intoxication the most. In second position, considerably less at risk, is the cat, followed by avian species, the horse, the fox and cattle. One of the most important causes of the high frequency of intoxication in the dog is its tendency to eat everything around it, unlike the cat which is extremely suspicious towards any unfamiliar compound. Another important source of poisoning in the dog is malicious poisoning, which seems to be related to the hunting season when animals are taken to areas where they are not welcome. The low number of reports for the horse and cattle is probably connected to their conditions of life: there are few occasions that permit these animals to feed themselves freely and without controlled fodder. It is important to note the high percentage of positivity found for baits discovered by chance (74%), for which it has been possible to find a remarkable heterogeneity of preparation, always characterized by the presence of elevated concentrations of toxin. The substances mainly used in these formulations are cholinesterase inhibitors (organophosphorates and carbamates), whose high toxicity makes them particularly suitable for malicious employment. The differences in the incidence of positive samples in the different periods of the year deserves particular attention. It is in fact possible to note how, uniformly for the four years considered, the poisonings are more frequent during spring (from March to June), during which the number of animals living in the open air increases and, consequently, the people annoyed by them. Even intoxications in the month of September are very frequent, coinciding with the opening of the hunting season. In this period poisoned baits are usually spread to exterminate the natural predators of the game (foxes etc) or to damage competitors. The low values obtained for the months of December and January are due to the fact that in this period animals live mostly indoors and are therefore less subject to intoxications. Finally, it is important to underline that some malicious poisonings are very dangerous even for humans, who can easily come into contact with toxic substances used in baits spread in public and private gardens. The aim of this study was to provide poisonings information for dogs and cats in Southern and Insular Italy and to determine the poisons more frequently involved. For this purpose were created survey cards which were proposed to practitioners through on-phone survey. The regions considered are: Abruzzo, Molise, Lazio, Puglia, Campania, Calabria, Basilicata, Sicily and Sardinia. For the Abruzzo Region it was possible to correlate data obtained through on-phone survey with the analytical results of toxicological analysis performed by the Istituto Zooprofilattico dell'Abruzzo e del Molise ''G. Caporale''. 63.2% of 310 practitioners contributed to the survey by answering our questionnaire. The highest interest and participation in the survey was observed in the regions of Abruzzo, Molise and Lazio with a percentage of 79.1%, 87.5% and 82.1%, respectively. The total number of poisonings considered were 17,324; 12,012 concerning dogs and 5,312 for cats. 45.1% of cases were suspected to be intentional. Only 56.1% of the veterinarians interviewed sent material to confirm their diagnosis to a laboratory. For pets that were still alive (81% of cases), a blood sample was the material preferably sent to perform hematochemical and coagulative evaluation. In 48.6% of cases, the gastric content was analysed, while urine and faeces analyses occurred respectively in 16.2% and 10.6% of cases. For dead animals or those that died during hospitalisation, 54.3% of vets sent the whole carcass to the lab, while the others performed an autopsy and sent organs or tissues to the laboratory for specific toxicological analysis. Some practitioners related that they usually contact the Anti-poisoning Centre of Torino (SATV) to have phone-support for diagnosis. As regards the classes of poison implicated, 84% of cases were related to anticoagulants (AC) and Organophosphate (OP), while 7.6% and 6.2% were associated to metaldehyde (MET) and strychnine (STR), respectively. Other toxic substances causing poisonings were also signalled but with an incidence less than 1%. As reported in Table II , the high number of requests for specific toxicological analysis is closely related to a wider range of toxic compounds found. Particularly in Abruzzo-Molise and Lazio where 60% of vets sent samples to a lab, the number of 94 TABLE II Percentage of vets who sent samples to a laboratory to request specific toxicological analysis for toxic compounds causing poisoning in companion animals. The regions considered are: Abruzzo (A), Molise (M), Lazio (L), Puglia (P), Campania (Camp), Calabria (Cal), Basilicata (B), Sicily (Sic) and Sardinia (Sar) toxic substances revealed were 12 versus 7 for the other regions of southern Italy. The average percenrage of vets requesting specific analysis in the other southern regions was about 50%, falling to 35.3% for Sicily and 9.1% for Sardinia. Such results are probably related to the lack of specific laboratories in these areas or to the tendency of vets to use only their own experience to perform diagnoses. However, it is important to note that many toxic substances produce similar neurological symptoms, highlighting the very real importance of specific analysis in differentiation of causes of poisoning. Without analysis, there is a particularly frequent confusion between OP poisoning and other neurotoxic substances, which are, thus, less frequently signalled and revealed. While we can observe a seasonal incidence for OP poisonings, for AC, MET and STR poisoning this was not observed. Even if MET is currently used in agriculture during spring and autumn, the constant presence and referral of MET poisoning is related to bait frequently prepared to criminally poison dogs and cats. We also found a correlation between agricultural practices and incidence of some poisonings. Particularly in Puglia the diffusion of permanent cultivation justifies the 1.3% of paraquat poisoning observed. Conversely, less intensive soil utilization in Molise is in accordance with the low frequency of OP poisoning. In Sicily, where 95 permanent cultivation is widely practised, OP poisoning is very frequent, representing the highest percentage observed (48.2%). For Abruzzo region, following previous studies, we were able to benefit from the data from the lab of the Istituto Zooprofilattico dell'Abruzzo e del Molise ''G. Caporale''. Toxic substances more often reported by vets were AC and OP while other compounds were reported less. These data confirm the trend observed in a previous work performed during the period 1997-2000 in the Abruzzo region where the percentages of poisonings signalled were 79.2% for AC, 69.4% for OP, 38.9% for MET and 37.5% for STR. Interestingly it can be observed that only 0.8% of samples were positive for AC. This result is probably related to the ability of veterinarians to diagnose AC poisoning directly from anamnesis and clinical symptoms, without the necessity to refer to the lab, while for the other compounds the veterinarians usually sent samples to IZS to confirm the poisoning. These data are similar to those obtained during the period 1999-2001 in the Tuscany region and reported by other authors indicating a similar trend. In contrast, considering the data obtained in a study performed in Friulia Venezia Giulia during the period 1997-2000 by IZV delle Venezie and the University of Padua, the compounds most commonly identified were plant protection products (49%): particularly OC, while OP were found in 21% only of cases. STR was also an important cause of poisoning in Friuli Venezia Giulia while AC and MET seemed to be less important. The recent decree of the President of the Council of Ministers of 28 February 2003, which acknowledges the agreement between the Ministry of Health, the Regions, and the Autonomous Provinces of Trento and Bolzano on the subject of the well-being of pets and pet therapy, provides that the Regions adopt -within the framework of their own competencies -provisions aimed at: 1. guaranteeing the well-being of pets 2. avoiding both direct and indirect reprehensible uses 3. the institution of a system of identification by means of subcutaneous microchips 4. utilising pet therapy with the elderly and children in need As far as the well-being of pets is concerned, the problems linked to poisoning are one of the aspects which has most greatly involved public opinion, and in the past has led the Regions to introduce new legislation. Article 21 of Law No. 157 of 11 February 1992 established ''regulations for the protection of warm-blooded wild fauna and for game sampling'', prohibiting the use of poisoned lures or food scraps and the application of penalties: namely, a fine of up to € 1,549.37. Furthermore, depending on the particular case, the legislation allows for further hypotheses of a criminal offence as provided for by the Penal Code. The sale of poisonous substances, utilised above all in the extermination of rats, is regulated by Royal Decree No. 1965 of 27 July 1934, which obliges vendors to keep a register in which are noted the personal particulars of the purchaser, the job or profession exercised, the substance purchased, and the relative quantity. Recently, in addition to sales, it has become necessary to extend the regulations to include the preparation, possession, and distribution of poisoned lures or food scraps. Until now, only two Regions -namely, Tuscany and Umbria -have promulgated laws dealing with the prohibition of the possession and use of poisoned lures, integrated in the 1992 Law (see References) . The aims of the laws are to safeguard animals and human health, to ensure public hygiene, and to protect the environment. 98 Article 1 ratifies the prohibition to possess and/or prepare any sort of food containing toxic substances with the exception of that required for the extermination of rats. Article 2 describes the procedures for the extermination of rats, already described in regulations in force at present but are better defined as regards to areas and information. In the Region of Tuscany, an explicit consent on the part of the owner or of those having the right, must exist before instituting procedures to rid premises, buildings, habitations, deposits, offices or private work-yards of rats. In the Region of Umbria, rat extermination is only permitted in private places subject to a communication being made by the owner (or others having the right) to the Commune and to the Local Health Unit at least 15 days before the event. Any other non-private and non-delimited area to be submitted to rat extermination must be authorised by the Commune, indicating the substances to be used and the duration of the treatment. Furthermore, the areas must be indicated by means of special Tables which detail the intervention and the person in charge of the treatment. A system of Registers of the treatments carried out in the territory of jurisdiction has also been instituted. Article 3 describes the administrative sanctions for those who fail to comply with the first two articles. In the Region of Tuscany, these sanctions amount to € 1,549.37, while they fluctuate between € 103.29 and € 619.74 in the Region of Umbria. To the fines is added the precautionary seizure of the poisoned lures or food scraps. In addition, if the person violating the regulation holds licences or permits in spheres of faunal and/or agro-silvo-pastoral activities or the collection of spontaneous products of the woods, suspension of the licences or authorisations for one year is provided for. In the case of repeated violations, such licences will be revoked permanently. If the lures are scattered by special security guards or voluntary guards, in view of the seriousness of the action and the institutional role that these hold, the fine is doubled and the appointment is definitively revoked. Administrative penalties are applied by the Province, which also transmits the documents for the possible revocation of the licences, authorisations, or appointments to the interested organisations for subsequent execution. Interventions of surveillance, prevention, and control of the territory are provided for in Article 5, by means of a decontamination of the areas affected by episodes of poisoning under the care of the Communes in collaboration with the local Health Units and the Provincial Police. These activities may avail themselves of the aid of the Voluntary Environmental Guards (in Tuscany) or Ecological Guards (in Umbria), as well as of the owners/managers of the properties by means of a co-ordination between the Provincial and Municipal Police forces. If the episodes continue in time, at the request of the Communes, under the care of the Provinces, delimitation of the perimetric area and the points of access is provided for, with notices to warn of the danger of poisoning without interrupting the faunal and/or agro-silvo-pastoral activities and collection of the spontaneous products of the woods. The important role of veterinarians is described in Article 6. If veterinary surgeons 99 come to learn of a case of poisoning during the exercise of their private practice or of surveillance in public health, they must report the episode to the Provincial Police or to the Commune within 24 h, using a special form, and -for the Region of Umbriaalso to the Local Health Unit competent for the territory. For the Region of Tuscany, the report form is attached to the law in question and distributed to veterinarians by the Provinces, while preparation of the form for the Region of Umbria was provided for in other legislative regulations. The veterinarian must also send samples of the organs, tissues or excrement to a regional laboratory for analysis, as well as the poisoned lures or food scraps discovered, for further diagnostic tests capable of identifying the poisonous substance. Should the aforesaid procedures not be carried out, the Communes for the Region of Umbria and the Provinces for the Region of Tuscany may apply a pecuniary penalty of from € 25.82 to € 103.29 which, in the case of repetition, may be transformed into a report to the competent Order of Veterinary Doctors, for possible disciplinary measures. The obligatory nature of the laboratory tests has made it necessary for the Regional Councils to identify, within the framework of the regional health service, at least one reference laboratory which is able to investigate a primary list of 14 poisons, as reported in Article 7 (Table I) . The choice fell on the Experimental Institutes of Zooprophylaxis, so called Istituti Zooprofilattici Sperimentali (IZS), competent for the territory. The Region of Umbria has established that samples can be transferred to the laboratory directly by the veterinarian or through the local Health Unit. It has also arranged for special forms to be used for sending and reporting them to the Provinces, Communes and Local Health Units, as well as Tables to be utilised in rat extermination activities and the delimitation of the territories during decontamination. The laws also provide for response times relative to the laboratory analyses and the procedures for sending test reports, which the Region of Tuscany has inserted in a special article, while the Region of Umbria makes reference to a subsequent act of the Regional Council. Thus, for the Region of Tuscany, the terms for the analyses are 10 days from the arrival of the sample, but are flexible depending on the type of investigation to be carried out. The Region of Umbria states the times with greater precision: these fluctuate from 7 to 15 days, in relation to the typology of the substance to be investigated. The test reports are transmitted via fax or e-mail to the veterinarian who sent the sample, to the Commune, to the Provincial Police, and -for the Region of Umbriaalso to the Local Health Unit. As far as coverage of the costs for shipping and analyses is concerned, the Region of Tuscany has inserted the analysis laboratory at the IZS of Florence into the Regional Health Plan in force. For the moment, it has not determined the cost of the toxicological analyses. For collection of the samples, in consideration of the complex nature of the territory, a network has been set up that avails itself of the organisation of police, environmental and animal-rights support associations, veterinary clinics and surgeries. The samples collected are transferred by the districts of the Veterinary Services of the Local Health Units to the Sections of the IZS. From here, the Institute completes the transport with its own vehicles to the chemical laboratory of the Florence Section. For the cost of the analyses, the Region of Umbria refers to the price list for services of the IZS, while shipment of the carcasses or samples is made free of charge through the competent veterinary services of the Local Health Units. Upon request, also those coming from veterinary doctors in private practice. The subsequent two Articles deal with two related aspects, which are aimed above all at prevention by means of the preparation and publication, under the care of the Provinces, by January 31st of every year, of an epidemiological map of the episodes of poisoning relative to the preceding year, and the establishing of a list of poisonous substances which must only be on sale under a regimen controlled by means of special registers. This list must be updated every two years and published in the official bulletins of the respective Regions. Through cartography, monitoring can be carried out for a better knowledge of the distribution of episodes in the territory. Furthermore, if there are several substances widely used in the preparation of poisoned food stuffs, these must be sold under strict supervision, in order to ensure their correct use. The Region of Tuscany provides for the setting up of a special technical consultative board of policy and surveillance for the control of the law; while the Region of Umbria avails itself of the pre-existing Regional Committee for the Protection of Animals. The number of participants on the boards varies from region to region, but has several elements in common. The Region of Tuscany also establishes the minimum three-monthly frequency for the Board's convocation. The final Article regards the financial regulations. In 2001 the Region of Tuscany 101 provided financing of € 15,493.71, to be divided between the Provinces in relation to the agro-silvo-pastoral area; however, in successive years, this will have to be transformed into self-financing on the part of the Provinces, with the collection of the revenues deriving from fines. Instead, the Region of Umbria has provided for the allocation under the regional chapter ''Prophylaxis, veterinary restoration to health'', and the IZS, to cover the costs of shipment and analysis of the samples, providing for an annual determination of the amount of the costs with the Regional financial law. Also provided for were resources for covering the costs of the Regional Committee for the Protection of Animals. Now at a distance of two years from the issuing of the laws, a first review must be made. In Tuscany, application of the regulation enabled a comparative evaluation to be carried out, during the first six months of 2002, through meetings realised by the Order of Veterinary Doctors at a provincial and inter-provincial level between veterinarians, provincial administrations, the state Corps of Forestry Service rangers, and associations of volunteers. The meetings revealed a sensitivity to the problem and interest in collaborating for application of the law; nevertheless, several problems linked to a uniform application of the regulations in the territory emerged. A fundamental aspect is that of verifying the appropriateness of the demand and, thus, the subsequent availability of reliable epidemiological data through the necessary filter of the veterinary doctor. It was sought to streamline the route for reporting and sending the material to the laboratories, by means of the use of a single report form and the sending of samples through Internet. To this end, the Region of Tuscany has activated a special space on its own web site. One problem seems to be the requirement to make a report for uncertain cases or for incidents in which intention to act with malice is not recognised. The possibility that the owner refuses to have an autopsy performed can represent an obstacle to the sending of pathological material to be submitted to toxicological investigations. Consequently, the Region of Tuscany has initiated an information campaign to sensitise the public regarding the application of the law. In several cases, difficulties have been registered regarding carrying out the autopsy in the surgeries; moreover, transportation of the samples must be effected according to the recent MINSAL (Ministry of Health) circular No. 5 of 8/05/2003, ''Recommendations for safety in transporting infectious materials and diagnostic samples''. The role of the Veterinary Services of the Local Health Units, to which the Region of Umbria has recognised specific functions, must be better clarified in Tuscany, especially in the link with the Communes for the flow of information and epidemiological surveillance. Also, no uniform criteria have been adopted on the part of the Provinces for preparing the annual cartography, with the consequent difficulty in collecting data at a regional level. As far as the analytical laboratory activity and training are concerned, availing itself of a protocol of agreement stipulated with the Region of Tuscany, an integration has been initiated between the IZS and the Faculty of Veterinary Medicine of Pisa, in order to make the services rendered more efficient and efficacious. In Tuscany, the link-up with the Bill on the protection of animal rights through the setting up of the regional National Veterinary Health Service can be interesting: in addition to identifying the recipients of the service and the criteria and procedures for the conventions between private structures and public administrations, it must also define the minimum levels of the services. One of the main activities in a laboratory of veterinary toxicology is to detect, in organic and inorganic matrices, substances suspected to be responsible for individual and collective poisonings of domestic and wild animals . The veterinary surgeon who asks for a toxicological analysis should keep to some elementary rules, which allow quick and cheap identification of the toxic substance. There are two main steps: 1) the proper collection, packaging and delivery of samples; 2) the compilation of a detailed and accurate clinical-toxicological schedule. Unfortunately, our experience has shown that the above-mentioned rules are almost always unknown. Consequently, samples are frequently sent to the laboratory in unsuitable conditions, which make toxicological analysis impossibile. The experiences of the laboratories of Veterinary Toxicology of the Universities of Naples and Pisa are discussed in this communication. In particular, correct ways for collecting, packaging and delivering samples and an example of a clinical-toxicological schedule are treated. Some of the baits recently recovered in the Tuscany region are also illustrated. Organic matrices to be collected are different, depending on which toxic substance is responsible for poisoning. If the nature of the chemical is known, it is possibile to consult data about its toxicokinetic properties and to collect the most suitable samples. In the absence of this knowledge, specimens such as vomitus, blood, urine, liver, stomach or rumen contents, kidney and fat may be particularly useful for toxicological analyses. Samples should be submitted in suitable quantities: 5 ml of serum; 10 ml of whole blood; 50 ml of urine; 200 g of bait, vomitus, gastric or rumen content; 100 g each of liver, kidney and fat. The entire brain should be submitted. With the exception of only very small species such as canaries, small rodents etc, the whole carcass does not need to be delivered. When the bait is found, it must always be sent to the laboratory. Each specimen should be put into a single plastic container (if possible, suitable for contact with food) that can be tightly sealed. One must avoid: $ the use of plastic bags or glass containers, which break easily with consequent loss of sample; $ the addition of antiseptics, preservatives and fixatives, which may interfere with the chemical analysis; $ contact of the sample with adsorbent substances, such as cotton-wool, gauze, cloth or paper, which may cause its desiccation. Indications (owner's name, animal name or number, and type of specimen) must be reported on the container on a label or written directly on it by means of an indelible pen. In order to avoid the degradation of chemicals, tissue specimens should be quickly frozen, whereas serum and blood should be refrigerated. Samples must be delivered by an express carrier, in an insulating container, such as a polystyrene box, where they should be surrounded with adsorbent materials such as sawdust or newspaper. In any case, samples must be kept cool during the delivery. A clinical-toxicological schedule, not in direct contact with the samples, must be sent to the laboratory together with them Peterson and Talcott, 2001) . The compilation of the clinical-toxicological schedule is very important, in order to allow the laboratory to carry out the analysis in a proper way. This should contain all information, which may address the toxicologist to intelligently select which toxicant to analyse for, especially when the toxic identity is unknown. In fact, analytical methods for the research of xenobiotics in organic and inorganic matrices are highly specific and they aim at the detection of only one chemical or, at least, of a group of substances having a very similar chemical structure. In addition, the correct compilation of the clinical-toxicological schedule by the veterinary surgeon allows responsible bodies to gather information about modalities, types and places of poisonings in order to create an updated database of intoxications in domestic animals. Figure 1 shows the clinical-toxicological schedule usually given by the Veterinary Toxicology Section of Naples to veterinary surgeons who attend the Specialization Schools of the Faculty of Veterinary Medicine in Naples. As previously mentioned, when the bait suspected to be responsible for intoxication is found, it should always be sent to the laboratory of veterinary toxicology: it could be essential for resolution of the case. All the poisoning baits submitted to the laboratory of Veterinary Toxicology of Pisa have been used with malicious intention. Some of these caused a great concern both to animal and human health because they were found in public and private areas (i.e. gardens). Unfortunately, this practice has become widespread in the Tuscany region, so much so that the police have instituted a special team to fight against the producers of poisoning baits. A lot of poisoning baits are sent to our laboratory every year, thus we have appreciated a trend of development both in their preparation and toxicological power. Classical baits were simply prepared, using discard organic matrix blended with very dangerous toxic substances (strychnine, cholinesterase inhibitors, metaldehyde etc.). More recently, time-consuming baits have been found such as emptied chicken necks, filled up with a toxin substance and tied up with a string, or hollowed out cheese, filled up with a toxic and covered with a lid of cheese. Successively, we received baits that contained two different toxic substances. This expedient may have been carried out in order to hide pathognomonic symptoms due to a single toxic substance and, consequently, to confuse the veterinarian in detoxification therapy practice. The most evolved bait was found a few months ago: an original in vivo bait (Meucci et al., 2003) . A clay-like matrix (analysed to be a mixture of pirimicarb and endosulfan) was inserted in the subcutaneous tissue of the neck region of a female pheasant and the animal was tied to a tree. The skin close to the subcutaneous implant contained several stitches. It is important to note that whoever prepared the above-mentioned bait is able to use anaesthetic drugs (if they have been used) and to apply stitches. In conclusion, we can note that more toxic and novel baits are tending to replace classical baits. The follicle plays a fundamentally important role in reproduction and hence understanding the mechanisms of its growth, development and ovulation are important to improve and control reproductive function in farm and companion animals. Follicles become responsive to FSH at the antral stage through the development of FSH receptors on granulosa cells. FSH causes an increase in granulosa cell numbers, oestradiol and inhibin synthesis, insulin-like growth factor (IGF) production and growth of the follicle. The secretion of FSH is constitutive, resulting in its secretion into the blood stream following synthesis. GNRH plays a role in FSH secretion because immunization of heifers against it suppresses its secretion (Crowe, 1999) . The balance of supply of inhibins and activin within the anterior pituitary gland is important in determining FSH output as well as the concentration of oestradiol, which has a negative feedback on FSH. There are recurrent transient increases in FSH during the oestrous cycle, in postpartum anoestrus, prior to puberty and up to the last 1-2 months of pregnancy. However, the patterns of LH and FSH are divergent, pointing to the fact that GNRH plays an important role in determining FSH concentrations. Following the LH/FSH pre-ovulatory surge in cattle, oestradiol concentrations decline and progesterone concentrations in follicular fluid increase as granulosa cells develop LH receptors and begin the transition to luteal cells. This peri-ovulatory decrease in oestradiol and inhibins results in an increase in FSH concentrations (Adams et al., 1992; Sunderland et al., 1994) and the emergence of the first follicle wave on days 1 to 3 of the oestrous cycle. The follicles that emerge and grow to 4-6 mm are all oestrogen active follicles and through their secretion of oestradiol and inhibins, FSH concentrations begin to decline and reach basal concentrations by day 4 to 5 of the cycle. FSH is the key hormone that regulates the regular occurrence of two or three follicle waves during the oestrous cycle of heifers. Each wave is about 7 to 10 days in duration. The first 3 days of a wave are characterized by a transient rise in basal serum FSH concentrations, a selection phase that results in a reduction in the number of growing follicles in the initial cohort from 15-25 3-mm follicles to 6-7 follicles ≥5 mm in diameter, and a decline in serum FSH concentrations coincident with onset of atresia of all but a single~8-mm dominant follicle. The remainder of a nonovulatory wave (postselection) is characterized by growth and differentiation of a single dominant follicle to potential ovulatory size during a period of relatively low serum FSH concentrations. Specifically, the dominant follicle has an enhanced capacity to produce oestradiol as it reaches 8-9 mm in diameter, an increased amount of FSH receptors and/or its mRNA in granulosa cells by 11-12 mm in diameter. Loss of dominance marks the end of a wave and is characterized by decreased production of oestradiol, atresia or ovulation of the dominant follicle depending on stage of the oestrous cycle the wave develops, and emergence of a new follicular wave. Current evidence indicates that follicle cohort members all grow at similar rates following emergence up to 6 mm or so, but there may be a hierarchy in size of the follicles with the largest member more likely to become the future dominant follicle (Ginther et al., 2001) . There is a distinct change in growth rate of the largest two follicles (F1 and F2) at around 7 mm with the F1 becoming the dominant follicle while the F2 eventually succumbs to atresia. Although the end of selection and attainment of dominance, where growth deviation occurs, can be detected by ultrasound scanning of ovaries in cattle (Mihm et al., 1997) , the biochemical signals that cause only one follicle to continue growth and development are now being elucidated. Differential responsiveness of follicles to FSH may be regulated by intrafollicular factors such as oestradiol, progesterone, inhibins activin-A, follistatin and insulin-like growth factor binding proteins (IGFBP). These factors are synthesized locally in follicles, are regulated at least partially by FSH and may have important autocrine or paracrine roles in modulation of FSH action and hence, follicular growth and function. Potential intrafollicular factors involved in follicle selection were identified by aspirating follicular fluid from multiple individual follicles from the same heifer. The future dominant follicle was associated with highest oestradiol content in follicular fluid and the absence of detectable levels of IGFBP-4 (Mihm et al., 2000) . FSH plays a key role in inducing the aromatase enzyme system in bovine granulosa cells. This suggests that the future dominant follicle is more sensitive to FSH due perhaps to increased FSH receptors or enhanced cell signal transduction mechanisms required for continued oestrogen synthesis. In addition, the lower IGF-B4 4 levels should result in increased bioavailability of IGFs within the follicular fluid, which should lead to proliferation of and increased FSH receptors on granulosa cells, increased oestradiol synthesis and maintain the health of the follicle. Rhodes et al. (2001) showed that selection of the dominant follicle was related to increased aromatase activity, increased cycle AMP in response to LH, increased oestradiol synthesis and decreased concentrations of IGFBP-4 and 5 within two days of ovulation. Thus, the bioavailability of IGFs is a key determinant of follicle development and survival in cattle. The regulation of IGFBPs plays an important role in follicular development which occurs either through changes in gene expression or proteolysis (Nicholas et al., 2002) . Concentrations of IGFBP 2 and 5 seem to be regulated mainly by gene expression while changes in IGFPB 2 and 4 are affected by intrafollicular levels of 109 specific proteases (Monget et al., 2002) . One of the proteases identified that degrades IGFBP-4 is the pregnancy specific protein A (PAPP-A) which is responsible for the complete degradation of intrafollicular IGFBP-4 in preovulatory follicles (Monget et al., 2002) . In contrast, Spicer et al. (2001) reported that levels of IGFBP-4 and 5 were decreased by proteolysis in preovulatory follicles coincident with a reduction in levels of IGFBP-2, 4 and 5. The relative roles of gene expression and proteolysis of IGFBPs in determining the bioavailability of IGF-I within the follicle require further clarification. In summary, our work (Austin et al., 2001) shows that selection of a dominant follicle from a pool of growing antral follicles is a dynamic process regulated by interactions of the gonadotrophins and intraovarian factors. Following the increase in circulating FSH concentrations, follicles of 2.5-5 mm in diameter are stimulated to grow. At this time, the five largest follicles per heifer have similar amounts of oestradiol, IGFBPs, activin-A, follistatin, and percentage of apoptotic granulosa cells. By 33 h after the peak in FSH, the two largest follicles increase in size and continue to grow while FSH concentrations decline. In addition, these follicles gain an enhanced capacity to produce oestradiol, inhibin precursors, and activin-A, but begin to lose the ability to produce follistatin while maintaining low amounts of IGFBP-2, -4, and 5 and a low percentage of apoptotic granulosa cells compared with the other follicles in the cohort. By 53 h after the peak in FSH, a dominant-like follicle can be distinguished in the cohort based on its largest size, highest oestradiol, and lowest amount of IGFBPs, though functional dominance, i.e., suppression of growth of all other cohort follicles has not yet been established. At 84 h after the peak in FSH, the selected dominant follicle is the largest follicle in the cohort, it is approximately 10 mm in diameter and contains the highest amounts of ostradiol and the lowest amount of IGFBPs and follistatin compared with the other follicles in the cohort. In contrast, the remaining follicles in the cohort (F2, F3, F4 and F5), though continuing to grow until 84 h after the peak in FSH, reach smaller maximum diameters of 5.5-8.5 mm. Thus, follicular growth is characterized by 1) increases in follicular diameter, oestradiol, and at least initially in inhibin and activin-A production; 2) a decrease in follistatin; and 3) maintenance of low amounts of low molecular weight IGFBPs and a low percentage of apoptotic granulosa cells. Follicles in the wave destined to become atretic are characterized by loss of capacity to produce oestradiol and enhanced production of low molecular weight IGFBPs. These changes not only precede atresia but also occur before the cessation of follicular growth. Based on these results the earliest intrafollicular changes that distinguish a follicle destined to become dominant from the other follicles in growing cohort are enhanced capacity to produce oestradiol and maintenance of low levels of IGFBPs. Herd reproductive efficiency is a major component leading to the economic success of the commercial dairy production unit. The modern day lactating dairy cow is considered to be sub-fertile in which the physiological stresses of high production and current management systems have a negative impact on both estrus detection and fertility. The high levels of milk production, dry matter intake, nutritional status, and metabolism of the cow appear to have altered both ovarian follicle development and luteal status. These alterations have contributed to low rates of estrus detection, increased multiple ovulations, and reduced embryo survival. With our current understanding of the factors controlling ovarian follicular dynamics and CL function, systems to manipulate the ovary for timed insemination, re-synchronization of timed inseminations for cows that do not conceive, and strategies to improve embryo survival have been developed to improve reproductive performance. Pharmacological control of the estrous cycle involves synchronization of follicular development that is coupled with the timely induction of corpus luteum (CL) regression, and synchronization of ovulation to improve pregnancy rate. Veterinarians and producers need to understand how these systems work in order to obtain compliance and optimal results. In lactating dairy cows the tools available for reproductive management are somewhat limited due to regulated use of such hormones as estradiol, GnRH, progestins, gonadotrophins and PGF 2a that varies among countries. Any use of a hormone should be in a manner that mimics normal physiological controls and responses of the reproductive system. Pregnancy rate (PR) of the herd is the proportion of pregnant cows relative to all eligible cows to become pregnant at every estrous cycle past the voluntary waiting period. Thus pregnancy rate is a function of estrus detection rate and conception rate. Ferguson and Galligan (1993) reported that PR at the first postpartum AI explained 79% of the variation in calving interval. It is also clear that after first insemination strategies need to be developed to identify and re-synchronize the non-pregnant cow in an efficient manner to achieve a 112 good herd pregnancy rate. Development of timed insemination programs with normal fertility has the ability to optimize service rate and thus improve PR in herds with low estrus detection rates. The challenge is to manipulate ovarian and embryo development in a manner to have an impact on conception rate and pregnancy losses. These two components that influence herd pregnancy rate are much more difficult to impact. Since the mid 1970s, PGF 2a has been the foundation system used in most dairy herds to synchronize estrus. Single or multiple injections of PGF 2a or its analogues causes a responsive CL to regress and cows return to estrus in 2 to 7 days. However, presence of anestrous cows and low rates of estrus detection will have a major impact on the number of cows responding to a PGF 2a program with detected estruses. Corpora lutea that are less than 5 days of the estrous cycle are non-responsive to PGF 2a . Therefore, a single injection of PGF 2a given at random should induce estrus in approximately 70% of the cows. Approximately, 90% of the cows will have an induced estrus response when two injections of PGF 2a are given 10 to 14 days apart. Our experience is that approximately 25% of lactating dairy cows approaching a voluntary waiting period of approximately 60 days are anovulatory and of course would be non-responsive to a PGF 2a program. The estrous synchronization program with PGF 2a does not lend itself to a singled timed insemination because there is no control over follicle development. This contributes to the variability in detected estruses between 2 to 7 days after PGF 2a injection. Injection of GnRH at random stages of the estrous cycle followed 7 days later by an injection of PGF 2a increased the number of dairy heifers synchronized within a 5-day period and enhanced the precision of synchrony during days 2 and 3 after injection of PGF 2a (Thatcher et al., 1989) . Treatment with GnRH results in a LH surge and ovulation of a dominant follicle (≥10 mm) leading to recruitment of a new cohort of follicles. GnRH induced ovulation is critical for recruitment of a new follicle wave, and the highest response to an GnRH injection is observed when it is given between days 5 and 9 of the cycle when there is a first wave dominant follicle that is responsive to LH (Vasconcelos et al., 1999) . The system of giving a GnRH injection followed 7 days later by PGF 2a (Selectsynch) results in acceptable estrus (67.2%), conception (41.5%) and pregnancy (30.2%) rates with lactating dairy cows (Burke et al., 1996; Santos et al., 2001) . 113 With the ability to synchronize follicular wave development coupled with PGF 2a induced regression of the CL, it was possible to implement a precise synchronization of ovulation for a Timed Insemination with an acceptable conception rate to first service. The Ovsynch protocol was developed as a breeding strategy to eliminate the need for estrus detection. The protocol is composed of an injection of GnRH at random stages of the estrous cycle to induce ovulation of the dominant follicle and synchronize a new follicle wave emergence. Seven days later, PGF 2a is given to regress the original and the newly formed CL, followed by a second GnRH injection 48 h later to induce a synchronous ovulation 24 to 32 h later. A timed insemination is carried out at 12 to 16 h after the second GnRH injection (Fig. 1) . This protocol has been implemented very successfully in many commercial dairy farms world wide as a strategy for AI during the first postpartum service, as well as for re-insemination of non-pregnant cows. Although the Ovsynch protocol allows for TAI without the need for estrus detection, approximately 10 to 15% of the cows will display signs of estrus during the protocol and they should be inseminated promptly if maximum PR is to be achieved. Lactating dairy cows from three commercial herds (n=333) were randomly assigned to either the Ovsynch protocol or AI based on estrus detection with periodic use of PGF 2a (Pursley et al., 1997) . Pregnancy was diagnosed 32 to 38 d after AI, and non-pregnant cows were re-inseminated using the original treatment. Median days postpartum to first AI (54 vs 83; p<0.001) and days open (99 vs 118: p<0.001) were lower for cows in Ovsynch compared to cows inseminated following estrus detection. Therefore, implementation of TAI with the Ovsynch protocol for the first postpartum AI improved reproductive performance in dairy cows, and subsequent re-insemination of open cows with TAI resulted in higher PR. When measuring PR, Burke et al. (1996) demonstrated that an Ovsynch program for a first service timed insemination was as effective as inseminating lactating dairy cows at detected estrus following a synchronization program of GnRH and PGF 2a given 7 days apart. Response to the Ovysnch protocol is optimized when cows ovulate to the first GnRH injection of the program, and when a responsive CL is present at the moment of the PGF 2a treatment. Vasconcelos et al. (1999) noted that initiation of the Ovsynch protocol between days 5 and 9 of the cycle resulted in the highest ovulation rate in response to GnRH. Ovulation to the first GnRH injection and initiation of a new follicular wave should improve PR because an ovulatory follicle with a reduced period of dominance is induced to ovulate (Austin et al., 1999) . Furthermore, initiating the Ovsynch protocol prior to day 12 of the estrous cycle should minimize the number of cows that come into estrus prior to the 2nd GnRH injection and ovulate prior to the completion of the program. Moreira et al. (2001) designed a pre-synchronization protocol to optimize response to the Ovsynch program by giving two injections of PGF 2a 14 days apart, with the second injection given 12 days prior to the first GnRH of the TAI protocol (Fig. 2) . The Presynch-Ovsynch program increased pregnancy rates 18 percentage units (i.e., 25% to 43%); in lactating cyclic cows. Additional strategies to pre-synchronize in order to optimize the Ovsynch program are possible but need to be well thought out in order to minimize cows being in early metestrus or late diestrus when the Ovsynch program is initiated. Future programs for further optimization of fertility likely will consider programs that manipulate ovarian function such that follicular turnover via ovulation or induced follicular atresia occurs in all cows, and luteal phase like progesterone concentrations are sustained until the time of induced CL regression. These future systems in lactating dairy cows may include insertion of intravaginal devices containing progesterone and acute injections of estrogens. Success of the OvSynch program is dependent on whether lactating dairy cows are anestrus or cycling. Pregnancy rates were less in cows that were not cycling at the time the OvSynch program was initiated (e.g., 22% versus 42%). Overall, the TAI protocol was able to induce cycles in 75% of anestrous cows, based upon the number of anestrous cows which were classified as ovulating to either the first and/or second injection of GnRH. If anestrous cows ovulate to the first and second GnRH injections of the OvSynch program then pregnancy rates appeared to be normal (e.g., 39%). Insertion of progestin devices as part of the Ovsynch program (i.e., between GnRH and PGF 2a injections) may also benefit anestrous animals. An alternative strategy to control the time of ovulation is the ability of exogenous estradiol to induce a LH surge by stimulating hypothalamic secretion of GnRH when given in a low progesterone environment during late diestrus and proestrus. An estradiol induced LH surge lasts for approximately 10 h, which is comparable to a spontaneous LH surge and longer than the LH surge induced by GnRH. Estradiol cypionate (ECP), an esterified form of estradiol-17b that is available commercially for use in cattle, has been used as part of a timed insemination program in lactating dairy cows. The ECP is used to replace the second GnRH injection of an OvSynch program. Cows are pre-synchronized with two injections of PGF 2a given 14 d apart with Heatsynch beginning 14 d after the second injection of PGF 2a . Cows are then injected with GnRH followed by PGF 2a 7d later. The ECP (1 mg, i.m.) is injected 24 h after PGF 2a , and cows are inseminated 48 h later. Pregnancy rates did not differ between Heatsynch and OvSynch programs (Pancarci et al., 2002; Fig. 3) . Cows detected in estrus after ECP had a higher fertility than those not detected in estrus at the timed insemination. In lactating dairy cows, the frequencies of detected estrus and ovulation after ECP were 75.7% and 86.5%, respectively (Pancarci et al., 2002) . Estrus occurred at 29.0±1.8 h (n=28) after ECP, lasted for 12.5 h and was associated with an average of 20.3 mounts. Mean intervals to ovulation were 55.4±2.7 h after ECP and 27.5±1.1 h after onset of estrus. Since 75% of the ovulations occurred between ≥48 h to ≤72 h after ECP, it is recommended that any cow detected in estrus by 1.5 days after ECP injection be inseminated at detected estrus, and all remaining cows be inseminated at 48 h. Based on synchronization of ovulation and pregnancy rates, ECP can be utilized as an alternative to induce ovulation in place of GnRH for a 116 Figure 3 . Heatsynch protocol for timed AI timed insemination. Since lactating dairy cows have reduced concentrations of plasma estradiol in the preovulatory period and reduced intensity of estrus, the elevation of estradiol following ECP injection supplements for a lactational induced deficiency, and our experience indicates that cows expressing estrus are fertile. The secretion of LH is regulated directly by GnRH whereas estradiol induces LH secretion indirectly via a positive feedback stimulation of hypothalamic GnRH secretion that then stimulates LH secretion. If cows are anovulatory (e.g., anestrus or have not developed positive estradiol feedback) the Heatsynch program may not be as effective as the GnRH based OvSynch program in which GnRH causes the direct secretion of LH. Greater uterine tone, ease of insemination and occurrence of estrus with the use of the Heatsynch program are well received by inseminators. Alternatively, in facilities with concrete flooring, the reduced estrous expression associated with the OvSynch program may be preferred. Since fertility between the two programs appears to be comparable, producers have a choice, which also includes relative costs of drugs (i.e., ECP35.6%) and pregnancy rate (48.5>23.6%) at day 30 after insemination than the cows detected in heat and inseminated with the Presynch-Selectsynch program. Pregnancy losses were similar (12%) between day 30 and 58 after insemination. Apparently, increased estradiol concentration in the periovulatory period of lactating dairy cows at first service enhanced conception rate and warrants additional investigation. Re-synchronization of non-pregnant cows is required if optimal herd PRs are to be achieved. Only 30 to 45% of inseminated cows are pregnant at 40 d after insemina-117 tion, and the non-pregnant cows need to be re-inseminated as quickly as possible. Strategies to accomplish this with the use of timed insemination can be rather aggressive with re-synchronization of follicle development prior to pregnancy diagnosis and early pregnancy diagnosis with the use of ultrasound. A study was conducted to determine the effects of re-synchronization with GnRH on day 21 after AI on PR and losses of pregnancy in lactating dairy cows . Holstein cows, 585, on two dairy farms were assigned to one of two treatments. On day 21 after the pre-enrolment AI, animals assigned to the re-synchronization group received an injection of 100 mg of GnRH, whereas animals in the control group received no treatment. Pregnancy was diagnosed on day 28 and reconfirmed 14 d later. Non-pregnant cows on day 28 were timed inseminated after the completion of the Ovsynch protocol 10 and 17 d after the enrolment in the study for re-synchronized and control groups, respectively. For re-synchronized and control cows, PR at days 28 (33.1 vs 33.6%; p<0.80) and 42 (27.0 vs 26.8%; p<0.98) after the pre-enrolment AI did not differ. Administration of GnRH on day 21 after AI had no effect on the losses of pregnancy between re-synchronized and control groups from 28 to 42 d (17.9%; p<0.74) after AI. Pregnancy rate after the re-synchronization period was similar for both treatment groups and it averaged 29.4%. Therefore, re-synchronization with Ovsynch beginning on day 21 after the initial insemination and prior to pregnancy diagnosis at day 28 did not adversely affect the initial PR and pregnancy loss in lactating dairy cows, and it can potentially expedite re-insemination of non-pregnant cows. With this approach, the re-synchronization and control groups were re-inseminated at 31 and 38 days after the previous service. In our previous studies we evaluated the interval from AI to return to estrus in cows that were nonpregnant (Fig. 4 ). Based on this distribution of intervals to estrus, it is feasible to inject GnRH at day 23 (i.e., 22-24 days) after insemination to synchronize the follicular wave and ensure that corpora lutea are present at day 30 that will be responsive to PGF 2a . Currently, a study is underway (J. Bartolome and W.W. Thatcher, personal communication) in which GnRH is administered between days 22 and 24, and pregnancy diagnosis using US is conducted between days 29 to 32 (i.e., mean 30 days). Cows diagnosed nonpregnant receive PGF 2a , and ovulation is synchronized with either ECP or GnRH (Fig. 4) . The timing of the Ovsynch protocol is standard with the ovulatory dose of GnRH given 48 h after injection of PGF 2a and a TAI at approximately 16 h after GnRH. Our experiences with ECP for re-synchronization are such that ECP (1 mg) is given 24 h after injection of PGF 2a , and we have 118 Figure 4 . Future strategy for re-synchronization chosen to time-inseminate all of the cows at approximately 36 h after injection of ECP. Results evaluating 756 cows indicate the following distribution according to stages of the estrous cycle at the time of pregnancy diagnosis: diestrus 75%, metestrus 5.8%, proestrus 9.6%, ovarian cysts 7.9% and anestrus 1.6%. In addition, 50% of the ovarian cysts showed luteinization at ultrasound. For the 567 diestrus cows, PR for re-synchronization were 28.7% (60/209) for cows subjected to PGF-ECP-TAI and 26.9% (60/223) for cows subjected to PGF-GnRH-TAI. Pregnancy losses between days 30 and 55 averaged 10% and did not differ between groups. Since we still have cows without a corpus luteum despite the fact they received a GnRH injection 7 d prior to pregnancy diagnosis, alternative protocols are being evaluated for cows in metestrus, proestrus and with an ovarian cyst. Early detection of nonpregnant cows is crucial for enhancing reproductive performance. Cows with a CL at the time of diagnosis of nonpregnancy which were treated with GnRH 7 days earlier can be re-synchronized with PGF 2a followed by either ECP or GnRH. Cows are then TAI within 3 days of the non-pregnant ultrasound diagnosis. However, cows with CL at diagnosis of nonpregnancy which were not pre-treated with GnRH can be re-synchronized with protocols for TAI (Heatsynch and Ovsynch). In addition, ultrasound and rectal palpation allow for detection of reproductive tract abnormalities that will contribute to treatment and culling decisions. Future cow-side pregnancy tests may allow detection of nonpregnant cows at an early stage (e.g., day 23) so that re-synchronization protocols can be initiated only in cows known to be nonpregnant. Choosing the proper stage to initiate the program with GnRH (e.g., day 23) takes advantage of the re-occurring follicular wave and CL to reduce the time for re-insemination. In recent years modern methods of assisted reproduction have been widely applied to cattle breeding (Galli and Lazzari, 1996; Galli et al., , 2003 . From the very beginning, the IVF technique in the bovine was mainly considered the method that could allow production of large numbers of cheap embryos from the ovaries of slaughtered females for commercial purposes. While this represented an obvious and immediate application it was also an instrumental step to fundamental studies by making available large numbers of embryos for research use. Therefore, a considerable body of fundamental knowledge has accumulated and this progress has increased not only the applications but also the understanding of the basic physiological processes involved in oocyte and embryo development. In this paper we will summarise the state of the art in assisted reproduction in bovines, including a brief analysis of the environmental, physiological and technical factors that affect the successful completion of all the steps that bring the immature oocytes to the stage of a good quality viable embryo. Any female after puberty is a potential donor of oocytes. The only requirement is the presence of antral follicles on the ovaries. Therefore, only animals with hypoplasic ovaries or in the immediate post-partum or late in pregnancy should be considered 122 unsuitable donors. Interestingly, certain ovarian pathological conditions such as ovarian cysts that often compromise the reproductive functions do not represent major obstacles for in vitro embryo production. Also pregnant animals can be good oocyte donors at least in the first months of pregnancy. General body conditions are important in determining the number of oocytes recovered and their subsequent successful development into viable embryos (see Table I ). The variability of the oocyte developmental competence observed amongst donors or batches of donors is great but probably not different from that observed in conventional ET. Certainly this variability is mostly dependent on nutrition and management of the animals (Kruip et al., 1995) and on the correlated physiological status. The common habit of feeding cattle diets high in degradable crude protein or in excess of requirements can reduce fertility and lower uterine pH (see Butler as review, 1998; Ocon and Hansen, 2003) . Recent studies have shown that high plasma levels of ammonia and urea correspond to increased concentrations in the reproductive tract and especially in the oviduct (Kenny et al., 2002) . The altered oviductal milieu is expected to cause profound disturbances to the developing conceptus since it is known that bovine embryos exposed to different culture environments can have altered level of expression of developmentally important genes . Over the years it has become clearly evident that the developmental capacity of the oocyte is also dependent on intrinsic ovarian factors. The most relevant is the follicular environment in which the oocyte has developed before collection. Primarily, the size of the follicle is very important, being correlated to the stage of oocyte development along the growth phase. The status of the follicle along the follicular wave, growing or regressing, and its consequent degree of atresia greatly affect oocyte development capacity (Galli and Moor, 1991) . A recent study in which oocytes were collected from cycling donors at precise stage in the cycle, has reiterated this concept by showing that oocyte quality is in fact the major factor that determines the efficiency of embryo production (see as review Lonergan et al., 2001) . On the other hand, more extensive studies on cycling animals with or without prestimulation with gonadotropins have shown that in general, but not always, prestimulated donors provide oocytes with higher developmental competence (see as review Hendriksen et al., 2000) . At the beginning of the IVP technology all oocytes were collected from slaughtered donors but in recent years the Ovum Pick Up technique (OPU) has become a routine procedure . This technological advance has considerably widened the potential application of IVP by making it possible to couple assisted reproduction and genetic improvement in cattle. It consists of the transvaginal aspiration of follicular oocytes with the aid of a scanner with an adequate endovaginal (or adapted for the vaginal use) sector probe. A guided needle secured to the probe is connected to a test tube and to a vacuum pump to aspirate the follicular fluid and the oocyte contained in it. The donor is confined in a crush, mildly sedated and given an epidural anaesthesia just before collection. OPU has virtually no drawbacks for the donor and can even have a therapeutic effect in some infertile donors affected by ovarian cystic syndrome or similar pathologies that compromise reproductive function. Virtually any female starting from 6 months of age up to the third month of pregnancy and also soon after calving (2-3 weeks) , is a suitable donor (9). Young calves 2-3 monthsold can also be oocyte donors although collection requires surgical laparotomy. In Table II we present data from our laboratory to illustrate the influence of the age of donors on embryo production. Unlike conventional superovulation, OPU does not interfere with the normal reproduction and production cycles of the donor. OPU can be performed sporadically or on a regular basis such as two times a week for many weeks or months. The twice a week protocol is the one that yields the maximum number of competent oocytes in a given period of time. If irregular intervals of oocytes retrieval are used or if OPU is done once a week a dominant follicle develops (Garcia and Salaheddine, 1998) and this causes the atresia of subordinate follicles making them less suitable for embryo production because of the lower quality of the oocytes. A practical tip to overcome this problem or to avoid having a donor in estrus at the time of OPU, is to administer 3-4 days prior to collection a Gn-RH analogue to pharmacologically ablate the dominant follicle and to initiate a new follicular wave (our unpublished observations). Another advantage of OPU is that it is not necessary to treat the donor with gonadotropins with the inevitable side effects. This is a very important advantage especially for young heifers in which gonadotropin-stimulation can cause mammary oedema and ovarian cystic syndrome, and for show cows where repeated superovulation can cause relaxation of the udder ligament. However, a Canadian study has demonstrated that a combination of mild superovulatory treatment with FSH and OPU can increase the number of embryos produced in a given time, therefore representing a valid alternative to conventional superovulation 124 (Bousquet et al., 1999) . A final advantage of OPU is the possibility of using over a short time, or even on the same collection (when many oocytes are recovered), more sires to achieve in a short time several different dam-sire combinations. This is especially relevant for those donor cows of high genetics that are requested as sire's dams. Following collection the oocyte must undergo the processes of in vitro maturation and fertilization to initiate embryo development. A considerable body of knowledge has accumulated over the years on the biology of the bovine oocyte and reliable procedures are available in the literature to mimic the process of bovine oocyte maturation and fertilization in vitro. In general over 80% of the immature oocytes transferred in the commonest maturation medium, TCM 199 with 10% FCS and gonadotropins, complete maturation and around 70% of them are successfully fertilised. Following cleavage several different protocols can be employed to develop the early embryos up to the blastocyst stage. Co-culture systems and cell-free systems are used depending on the different laboratories and also an in vivo culture system, in the oviduct of sheep temporary recipients, is successful. As recently demonstrated by Lonergan et al. (2001) , the culture system applied from cleavage to blastocyst is the critical factor that affects embryo quality while, as mentioned above, it is the intrinsic oocyte quality that determines embryo yield. An interesting approach has been developed in the last few years to improve the quality of the oocytes through a prematuration culture period of 24-48 h, in presence of molecules capable of blocking the oocyte in the germinal vesicle stage. This method, generally referred to as two-step culture system, was expected to help the oocyte to complete its cytoplasmic capacitation during the prematuration step therefore improving its ability to undergo the following maturation and to develop into a viable embryo. Unfortunately, no improvement of developmental capacity has been demonstrated in oocytes subjected to prematuration. Interestingly however, this method has proven valuable to manipulate the timing of the maturation process during complicated procedures such as nuclear transfer (Lagutina et al., 2002) given that the embryos produced by the two-step culture system have normal viability (Ponderato et al., 2002) . The successful application of the in vitro embryo production technology requires a highly trained technical team and a fully equipped assisted reproduction laboratory. It is therefore a demanding technique and consistent success is the result of scientific knowledge, dedication and attention to details. In the future, a more widespread use of assisted reproduction techniques in cattle will be facilitated by continuous support 126 for basic studies and also by the ability of the cattle industry to fully exploit the opportunities offered by scientific advances. The water buffalo (Bubalus bubalis) does not yet have such characteristics as to warrant the label ''dairy animal''. The buffalo bred in Italy produce almost exclusively milk with a good income while buffalo meat has always been considered a by-product and only in recent years is being considered as for co-primary production. The milk of this species accounts for over 50% of drinking milk, in certain developing countries, such as India, Pakistan, Egypt and Nepal, while in Italy buffalo milk is almost exclusively used for mozzarella cheese production. Since the establishment of a Herd book, the ''Italian Mediterranean Buffalo'' has shown increasing production (from 1977 to 2002) both in terms of quantity ( Fig. 1 ) and quality (Fig. 2) , which may be chiefly attributed to a change in feeding techniques (Table I) . Indeed, comparisons between findings in 1967 (Intrieri and Minieri, 1967 (Zicarelli et al., 1977 and 2000 show that there has been an increase in the fat (Fig. 3 ) and protein content of milk ( Fig. 4 ) both in the catabolic and anabolic phases of lactation. This may be attributed to the increase in energy of the diet (0.75 vs 0.9 UFL/kg D.M.) and protein density (11% vs 15%) obtained chiefly with the decrease in the amount of forage (90% vs 45-50%). Part of the increase in protein content occurring in recent years may also stem from the increase in milk urea content, although no correlations have emerged between proteins and milk urea (r=−0.08) or between urea and cheese yield (r=− 0.10). In recent studies, there was a good correlation between urea (Fig. 6 ) and milk freezing point -MFP -(r=0.47). The latter result, besides confirming our previous findings (Campanile et al., 1998) which indicated that MFP values may depend on dietary factors (wide NSC/protein ratio and/or low protein values), also showed that the lactation phase , regardless of the type of diet adopted, may alter both the urea and MFP values. This assumes major importance for buffalo farming as the herd, due to buffalo reproductive seasonality, changes the mean number of days in milk during the year and hence the milk may sometimes show higher values of MFP that do not stem from the addition of water. The increase in milk quality during recent years has also caused an increase in cheese yield as a significant correlation with milk protein content (r=0.47) has been demonstrated (Fig. 5) in experimental trials . Compared with cow's milk, buffalo milk has a higher percentage of all components (Table II) . In spite of its higher fat percentage, milk and mozzarella cholesterol content is lower for buffalo than for cow's milk (275 mg vs 330 mg and 1562 mg vs 2287 mg respectively). This is of major interest, together with some studies that report 131 a larger number of small fat globules in buffalo milk as compared to bovine and sheep milk. It is well known that small fat globules are rich in polyunsatured fatty acids (Martini et al., 2003) . Thanks to its higher dry matter content, buffalo milk has a higher cheese yield that, however, does not seem to be due solely to the different milk quality. For example, for each protein percent point more than 5 cheese yield points are obtained for buffalo and about 3 for cows even when milk is adjusted for the same caloric content. The difference is attenuated although it still persists if the cheese yield is adjusted for fat content (Table II) . Unlike bovines, as shown by the researches of Addeo et al. (1977) , no polymorphisms have been detected for a s1 , b or the k casein, except for differences in the electrophoretic mobility of a s2 casein fraction. The k casein is similar to the cow's kB casein and has 7 main fractions (Addeo et al., 1977) , of which k4 and k5 show two more (a and b) and k7 four more (a, b, c and d). Fraction k 1 represents respectively 40% of the total k-casein in buffalo and is very similar to kB 1 -casein in the cow, where it accounts for only 25% of total k-casein (Addeo et al., 1977) . The larger quantity (Table III) of k-casein as compared with the cow speeds up the enzymatic curding phase which requires a smaller quantity of renin. Optimal curd elasticity is obtained for the buffalo at pH 4.9, and for the cow at pH 5. 2-5.0 (Addeo et al., 1996) . The amino acid composition of the k-casein of the two species differs in the quantity (mole/mole protein) of N-acetylgalactosamine (0-4.3 and 0-6.7 respectively in buffalo and cow) and sialic acid (5.5-8.5 and 3.5-4.3 In an alkaline environment the electrophoretic mobility of the three components of buffalo a s1 casein is lower than that of bovine. This is a property which enables the presence of cow's milk to be detected (over to 5%) in the event of sophistication (Ferranti et al., 1996) . Moreover, due to a mutation, there is a phosphorylate residue instead of hydrophobic one, which is responsible for the enhanced non-polar nature of the buffalo protein. In contrast, in cow's milk a s1 casein has a phosphoseryl group in position 115 which is surrounded by hydrophobic amino acids. The absence of phosphoserin 115 in the a s1 casein of buffalo strengthens the non-polar nature of this protein (Ferranti et al., 1998) . The interaction of hydrophobic groups is probably responsible for the characteristics of the a s1 casein in the context of casein micelles. Finally, the loss of a phosphate group, the increase in density and higher sensitivity of buffalo casein micelles to chymosin may partly explain the shorter clotting time and the higher cheese yield of water buffalo as compared to bovine milk (Addeo et al., 1980) . The cheese-making potential of buffalo milk is also better than that of the cow due both to the higher casein content and the phosphorylation characteristics of a s2 casein: 10 and 11 phosphate groups/mole (in that it consists of two fractions) vs four respectively in buffalo and cow (Ferranti et al., 1996; Addeo et al., 1996) . Buffalo's b casein is similar to that of the cow and has two variants. Of these, A has been found only in Venezuelan buffalo and differs from B in three amino acids (Ferranti et al., 1998) . The two variants A and B closely resemble cow's bA 2 casein, differing from the latter by four and five amino acids respectively. The b and a s1 casein fractions make up 70% of the micellar network of proteins. Hydrolysis of buffalo b casein with plasmin produces the fractions c 2 and c 3 . Thanks to the identification of fraction c 2 , it is possible to detect the presence (<1%) of cow's milk in buffalo mozzarella (Addeo et al., 1989) . Finally, the protein pattern of the two species differs in the smaller quantity of a s1 and b found in buffalo, as well as the larger quantity of k-casein and a s2 casein (Table III) . However, the absence of polymorphisms for a s1 , b 133 and the k casein, that is found for bovine milk, may suggest that there are no differences between milk of different buffalo populations and/or subjects. No differences have been found between buffalo and bovine variant B of b-lactoglobulin while some differences exist between the two species in the B variant of a-lactoalbumin. Recently a novel variant of a-lactoalbumin has been found, called A, that differs from the analogous bovine A variant in that it seems to contain less glucides. The frequency of the AA genotype is very low, about 0.5% as is the AB one (2.4%) (Chianese et al., 2001) . In a recent trial carried out over two consecutive years , it emerged that by using the formula of Altiero et al. (1989) , which calculates cheese yield by taking milk fat and protein content into account, it is possible to identify buffalo (Table IV) which, given the same calculated cheese yield and hence the same milk lipoprotein content, have a different experimental cheese yield. For these buffaloes it has been calculated that one point more of milk protein percentage supplies 57 g of curd for buffaloes with an experimental cheese yield higher than that calculated and 53 g for those in which the experimental cheese yield is lower than the calculated one. Buffaloes with the higher yield were those that had produced less milk per lactation (<3000 kg) in the first year of the trial. In the second year, buffaloes were monitored between 130 and 160 days of lactation, insofar as the yield in this phase had correlated strongly (r=0.93, p<0.001) with the average cheese yield over the whole period of lactation. In this trial, no differences emerged concerning the cheese yield among the animals that had produced more or less than 3000 kg of milk/lactation, while buffaloes with different experimental cheese yield also showed a different value for theoretical cheese yield. A higher value of experimental cheese yield was found for 3 out of 14 (21.4%) of buffaloes yielding more than 3000 kg of milk and 18 out of 33 (54.5%) of buffaloes yielding less than 3000 kg of milk per lactation. In the first year no differences emerged in the rheological parameters between buffaloes with high and low experimental cheese yields. By contrast, the more productive animals (more than 3000 kg of milk) showed a shorter enzymatic phase and higher SH values over the two years. No relationship was found between milk characteristics over the two years of the trial, while the cheese yield in the fist year was related to the cheese yield at 4 (r=0.765; p<0.01) and 24 h (r=0.732; p<0.05) from milking, in the second year. Further studies are required to identify buffaloes whose casein retains more whey and more fat during the clotting process and to ascertain whether there are protein fractions associated with the phenomenon. Moreover, also with respect to the dairy cow, the genetic variants detected up to now do not fully justify the differences in yield. Recently we ascertained (in press) that Bruna Italiana cows have a higher milk yield than Italian Friesians, having the same k-casein variant and milk fat content. This is particularly important in that a higher cheese yield is also synonymous with an enhanced cheese-making potential and hence with milk able to provide cheese without recourse to particular technological stratagems. This feature has enabled the ''Mozzarella di Bufala Campana'' DOP (MBC) to achieve its well-known success: of the DOP cheeses it ranks third according to sales which have continually risen since the establishment of the MBC consortium. Albeit difficult to achieve, selection should aim to identify those higher-cheese-yielding animals which are not always those that produce milk with a higher protein content. By selecting only for higher milk production, one risks increasing the number of animals within the buffalo population producing milk which is not always optimal for processing. Selection made for higher-cheese-yielding animals is even more likely to characterize the ''Italian Mediterranean Buffalo'', since it cannot be done with the genetic variants of milk. This would further validate the existing regulations of MBC that restrict the product to exclusive processing of milk from the ''Italian Mediterranean Buffalo'' and that pursue the aim of binding the product to the breed and the geographical area. The reference to a link with a certain geographic area is represented by the use of local forage for animal feeding and is greatly emphasized by the amendment to the MBC production regulation regarding milk processing, which stipulates that ''acidification of milk must be achieved by adding whey, derived from previous buffalo milk processing''. This technique does not enhance product shelf-life but is a characteristic which has distinguished MBC from other stretched curd cheeses and has underpinned its success. If changes are to be made, they should aim to increase the frequency of supply to large distributors as this would even further guarantee the label ''fresh product''. To avoid any confusion, it should be pointed out that this paper will examine the question: (a) of buffalo milk and mozzarella, but not that of Buffalo Mozzarella from Campania (Mozzarella di Bufala Campana) nor that of cow's milk, unless expressly indicated; (b) of the buffalo in its entirety, despite dealing separately with its components. The discussion will necessarily be brief, albeit with references to the main issues of consumption, processing, distribution and milk production, as well as general prospects. According to ISMEA, continuing the trend of previous years, in 2000 domestic consumption of buffalo products rose further, although the same year saw a contraction across the board in the consumption of other DOP cheeses, as occurred across the whole cheese sector. By contrast, demand for DOP cheeses in 2001 showed no variations in volume, although it increased in value by 5.5%. In particular, Mozzarella di Bufala Campana experienced increases, for 2001, both in quantity (5.9%) and in value (12.4%) (Fabbro, 2003; Osservatorio Consumi, 2003) . These consumption trends in buffalo products occurred within a price framework which is on average 50% higher than that of mozzarella made from cow's milk. Consumption is chiefly located in Southern Italy. However, there seems to be a clear trend towards greater distribution in other parts of Italy. In other words, despite being more concentrated in the areas where the raw material is produced, buffalo mozzarella consumption is spreading throughout Italy and abroad to the same extent as other, longer-established products. In the geographical areas far removed from the production area, percentage variations in the quantities consumed are three times higher in north-western and central Italy than in the south. Finally, it is precisely this 138 expansion of the buffalo mozzarella market and its greater product penetration which afford the greatest short-term prospects for overall sales expansion, and not the increase in per capita consumption. The supply of buffalo milk, and hence that of mozzarella, is also continually expanding, both in terms of quantity and quality. All this suggests generally positive forecasts for sectoral development, although there are still doubts concerning the sector's actual ability to match, or even promote further growth, with appropriate improvements in processing and distribution. Several years ago it was thought ''that buffalo mozzarella has a potential market which is perhaps difficult to reach, but whose theoretical size may be valued at about three times that of its current market'' (de Stefano, 1999) . This forecast still holds, although subject to certain conditions. The processing of buffalo mozzarella is structurally different from that of cow's milk mozzarella. The dairies are smaller, they use less advanced techniques and usually operate on geographically very limited markets, which often coincide with the production areas of the raw material (de Stefano, 1996) . The degree of competition between the dairies is high, the tendency to specialize is low, and investments in the distribution and marketing sector are almost non-existent. The distribution of buffalo mozzarella has specific features that distinguish it from the rest of the cheese-making sector. These concern the greater importance on total sales of specialised and traditional firms as compared with large distributors, assisted retail as opposed to self-service, and bulk purchase over packaged purchase, while the distributor's brand name has only recently appeared upon sale of the product (de Stefano et al., 2001) . Distribution outside the local market is generally not effected by the dairies, but is entrusted to third parties. In this situation, hardly any dairies have up until now been involved in labelling policies. The product is usually sold in traditional low-hygiene packages of varying weight which are ill-suited to modern requirements of handling and display, especially those of supermarkets. It is thus no surprise that much buffalo mozzarella is sold via traditional assisted retail or that it is limited to the segment of gastronomic products. One of the factors slowing down the adoption of more modern sales approaches is undoubtedly the limited shelf life of the product, on average shorter than that of the cow's milk product. This is also why the latter has a higher service content, and is marketed to a greater extent in fixed weight packaging. The buffalo farm sector is expanding both in typical areas in Italy and in other areas of central-northern Italy, and even abroad. Yield expressed in milk is rising, breeding techniques and animal welfare are improving and livestock quality and milk hygiene have also shown considerable signs of improvement (Cerrato, 1999 : Flora, 2002 de Stefano, 1999) . However, in this period of great progress throughout the farming sector in Europe, the improvements recorded in this specific sector may be described as limited. They chiefly concern the search to contain production costs, while insufficient attention is paid to improving quality. This is particularly applicable to the organoleptic characteristics of milk and even, in some areas, herd welfare and the techniques adopted. For example, improvements introduced to animal diets explicitly aim to reduce production costs, and have little to do with the need to guarantee milk quality. Economic globalisation is also affecting this sector, which is why buffalo farms in other countries, especially in Europe, are drawing ever closer to Italian and international markets. Buffalo mozzarella still experiences considerable competition from the cow's milk product, due both to high substitutability between the two products in the consumer market, and to the effects caused by distribution. Thus, in order to assess the prospects of the sector, we should not overlook trends due to changes in distribution channels. Indeed, large-scale distribution was in the past oriented towards a preference for 140 cow's milk mozzarella, especially due to its greater adaptability to the requirements of modern sales strategies and techniques. Despite attracting a specific consumer segment, buffalo mozzarella ended up occupying market spaces left free by its cow's milk counterpart and by the firms producing it, rather than winning its own independent share. In other words, buffalo mozzarella has seen its own market grow in recent years because the market for the cow's milk product has grown. Expansion of the latter has been mainly due to: (a) changes in consumption styles and tastes on the part of Italian consumers; (b) growth in modern distribution and promotion of the cow's milk product by the large cheese-making industries. Buffalo product sales have grown continuously in the recent past, although to a lesser extent overall than might have been expected, given its favourable reception by consumers. The key reasons for this lie in: (a) the structural and organisational limits of the industrial processing sector; (b) the notable shortcomings currently found in the distribution of the end product; (c) inadequate strategies for improvement of milk quality. Moreover, the marked growth in the supply of milk and processed cheese in recent years, which should continue in the immediate future, means that the markets for the end product also have to expand at a similar rate. As the wholesale demand elasticity of the end product is not high and the demand elasticity of milk on farms is low, there only needs to be a modest surplus in supply for there to be severe reductions in milk selling prices on the part of farmers. Thus, in the future the market will play an extremely important role. In particular, the distribution sector will become increasingly demanding with regard to terms and conditions of supply, requiring consistently high quality standards of goods and product safety, and generally all those features that allow the adoption of more modern techniques of packaging, promotion, merchandising, and management (Del Giudice and de Stefano, 2002) . Such requirements are ill-matched by a structure which is little more than artisanal, found in most of the buffalo milk processing sector. The above-mentioned factors, which affected the whole mozzarella market in the past, have ended up almost spontaneously benefiting consumption of the buffalo product as well. These factors could well decelerate in the future. Thus, it is necessary to adopt specific development strategies for the processing of buffalo mozzarella, which should have two different objectives. (a) On one hand, dairies, once their technology has been upgraded, will have to continue to target precise market niches, in which to offer a consistent, extremely differentiated, easily recognisable high-quality product, with a high degree of added service and perhaps a semi-artisanal appearance. (b) On the other hand, the more dynamic and ambitious firms will have to explicitly target modern distribution channels and more distant markets, not being 141 satisfied only with existing niches, but launching themselves to win large market shares, in open competition with the cow's milk product. Finally, we should especially focus on the relationship that may be established between buffalo mozzarella, intended as a typical product with high development potential, and the modern strategies of large-scale distribution. In general terms, ''typical products with high development potential'' are those whose production on the one hand shows a non-negligible economic and productive dimension, combined with good market growth prospects, and on the other an organisational level of the filière with considerable limits and shortcomings. This definition is perfectly suited to buffalo mozzarella. To develop such production, stress has been repeatedly laid on the ever-diminishing importance of price-based competitive strategies and the growing importance of strategies based on improvement and diversification of services and assortment. Increasingly, therefore, the operative and strategic choices of the large Italian and European distributors will be translated into an increase in available customer services, the creation of a more direct relationship between employees and consumers, and into planning a product range guided by new criteria. Of these, the regional product range is assuming considerable importance. Therefore, taking due account of further growth in the importance of large distribution and hence the greater importance of the rationale of modern trade, the sectors producing typical foodstuffs in traditional manner have to rethink their own strategic choices so as to take advantage of the chances offered by the new strategies of large distributors. This opens up even more needs and opportunities for market development of buffalo mozzarella, which will need to be taken into account in the short term. The situation briefly outlined above indicates many concluding considerations. Without wishing to list them one by one, it would be worth focusing on at least two major points. These considerations also dictate specific strategies for the buffalo milk filière, both for the distribution sector and for farms. The first point is that buffalo mozzarella differs from most of the other typical products in Italy, especially because: (a) it has already reached high levels of supply and demand; (b) its supply continues to increase sharply. In other words, buffalo mozzarella can no longer be considered a niche product. If we wish to seek a parallel, then we should be careful not to commit the mistake typically made by many parents, who fail to realise that their children have grown up, and continue to treat them as children even when they have become adults! Going back to buffalo mozzarella, it should be realised that traditional markets, although these 142 should not be neglected, may well have become insufficient to absorb the foreseeable future increase in supply, and that wider markets must therefore be conquered. This means that the increase in distribution must necessarily pass through large distributors, something that in turn will force the processed product to learn to adapt to the many technical and organisational requirements of these distributors, without relinquishing its typical organoleptic characteristics. It is chiefly up to the industrial processing sector and scientific and technological research to ensure that this fundamental and urgent task is accomplished. The second point is that the typical nature of Campanian Buffalo Mozzarella is, all things considered, an attribute that could also prove somewhat transient. Such typicality, apart from the origin of the raw material, currently lies almost exclusively in the fact that the product is obtained from buffalo milk, irrespective of its other technical and organoleptic characteristics. Yet with the increase in end demand for the finished product, the expansion of buffalo farming into other areas in Italy and the rest of Europe, the increase in interest from multinational economic lobbies in the cheese-making sector, along with the intensification of globalisation processes and their relative market ''philosophy'', the current typicity of buffalo mozzarella could become increasingly difficult to uphold. The product therefore needs to also become differentiated by other traits, to be linked to the production area and to farming techniques. In this regard, the organoleptic traits of mozzarella play a fundamental role. Of particular importance is the quality of the raw material. This means that the focus must be on livestock technology, especially diet, perhaps by the rediscovery and diffusion of techniques which are traditional to typical buffalo farm areas and have recently been overlooked as they were unsustainable in terms of production costs. This time the task falls chiefly on the farming sector and once again the research sector. The economic situation of the buffalo filière is still favourable, although clouds are already gathering on the horizon. There is no reason to be complacent and let the chance to set up the sector's future pass by, especially with the likely prospect of more difficult times just around the corner. In Italy, commercialization of milk for mozzarella cheese production represents the main income of buffalo breeding. This is a critical point for farm economy. Meat production could represent a good collateral or main activity. In addition, the market interest in buffalo meat has progressively increased due to the negative trend towards bovine meat consumption, which has very recently worsened due to the BSE scare. In the past much research was carried out in the Veterinary (Ferrara et al., 1969) and Agricultural Faculties (Matassino et al., 1976) of the University of Naples and at the Istituto Sperimentale per la Zootecnia of Rome demonstrating both the modality of breeding buffalo as a meat producer and the good nutritional characterics of buffalo meat. More recently, Di Lella et al. (1998) , in a trial aiming to verify the influence of the feeding programme on the growth dynamics of young buffalo bulls, divided 24 sevendays-old buffalo calves according to a 2×2 factorial design: two weaning ages (63 d vs 84 d) and two weaning concentrates (CP 17% vs 14%; starch 37% vs 29.6%, as fed). The calves of groups A and B, weaned at 63 d, received 6 l/head/d of acidified milk replacer (in the ratio of 180 g/l of water) until 42 d. Subsequently, the replacer amount was gradually decreased, whilst administering the same volume. The animals of groups C and D received 8 l/head/d of the same milk replacer, but in the ratio of 140 g/l, until 56 d. Also in this case the amount of replacer was gradually decreased, leaving the volume unchanged. Roughly chopped alfalfa hay and weaning concentrate were available from the fifth week; corn silage was administered starting from 70 d. 144 After weaning, the animals were fed ad libitum hay and corn silage; the concentrates were administered in the amount of 12 kg/d/group. The weaning age did not influence performance. By contrast, the weaning concentrate strongly affected the growth dynamics in the first 6 months: in this phase the concentrate with higher protein and starch content produced better results. However, in the subsequent period (6-16 months) due to compensative growth, improved performance was obtained using concentrates with less favorable characteristics. These observations suggest that concentrate with higher protein and starch content be administered during weaning, when the calves are destined to be slaughtered at lower weights compared to those reached in this trial (400 kg); otherwise it could be opportune to administer the protein-poor concentrate in order to limit feeding costs. In order to increase the knowledge about buffalo growth, characterised by periods of very scarce weight gain followed by compensatory phases, Infascelli et al. (2001) carried out research on 12 six-month-old male buffalo, equally divided into group A (111.0±6.9 kg LW) and B (116.7±7.6 kg LW). Group A was fed 100 g DM/kg LW0.75 of diet (CP/DM 14.1%, UFV/kg DM 0.84) based on rye-grass silage, ryegrass hay and concentrate. Group B received the same diet, but every 2 months the DM was alternatively 20% more or less than group A. From 300 kg LW until the end of trial (477 d of age), group B received diet (CP/DM 14%, 0.88 UFV/kg DM) obtained by varying the proportions of feeds. Between 180-238 d of age, group B showed a higher daily weight gain (DWG: g 480±0.46 vs 510±0.34, for A and B); however the conversion index (CI) was worse due to the higher amount of DM and UFV intake. In the interval 238-303 d, with a decreasing DM intake, group B showed lower DWG and worse CI. In the following interval, notwithstanding the increase in the DM intake, group B had similar results. On the contrary, between 380-415 d, group B had more favourable CI (UFV/kg weight gain 5.195 vs 4.452 for A and B) with slightly lower DWG (g 1143.0±0.28 vs 1023.0±0.14, for A and B). Group A showed the best results for the period 180-415 d, even if differences were not significant. Increasing the energy density of the diet in the interval 415-477 d slightly improved the total (180-477 d) DWG of group B, however less than that of group A (855.2±0.58 vs 769±0.25). Consequently, the total CI of group B was worse than that of the interval 180-415 d. Considering the whole trial, the buffalo growth curve showed a progressive increase in DWG, reaching its peak at 303 d of age (180 kg LW) until 415 d (300 kg LW). Successively the CI considerably worsened. An innovative method to obtain ''biologic'' young buffalo bulls could be the cowcalves system adopted in Brazil . As reported in Table I, due to the high price of buffalo milk, the utilization of dairy cow at the end of her productive career to wean two buffalo calves should be financially more favourable. This should produce economically more favourable results also compared with the administration of milk replacers, thanks to the compensation for the nurse by the EU. Because of sanitary reasons (inter-species transmission of infective diseases) the system should be utilised in stalls outside dairy farms. are higher than those obtained for bovines (Failla et al., 2001) . Fat content and dry matter percentage (Table III) increase for aged buffaloes (+2.1 and +2.8 percent respectively) while the protein and ash percentages do not change very much (Cutrignelli et al., 1996; Failla et al., 2001) . All colour parameters decrease with age so the young buffalo bulls show a darker coloured meat, with a hue closer to magenta, but with less intensity (Table IV) . The measured tenderness (as Warner Bratzle shear force) on cooked meat, which is the best parameter for consumer appreciation, does not show differences according to age. The water holding capacity of raw meat, that strongly affects the choice of consumer at the moment of sale, is better for veal, while the difference is not important for cooked meat (Failla et al., 2001) . The fatty acid composition (Table V) is significantly different between two typologies. The aged animals show more saturated fatty acids, in particular C 16 and C 18, more C 18:1 and less C 18:2. All the data presented confirm that buffaloes have good meat quality characteristics both at six and at 15 months of age, in comparison with bovines (Gigli et al., 1978; Cutrignelli et al., 1996) . In addition, the difference between In spite of the higher costs of buffalo meat production, due to their slower growth compared to specialised bovine breds, there are sound reasons to exploit buffalo meat. Its very good nutritional characteristics and the low number of subjects (considering a mean intercalving period of 400 days, the 120,000 cows actually bred in Italy could produce 96,000 calves of which 52% are males) are well appreciated by the consumer from a dietetic point of view as well as the fact that the sector is more controllable, in terms of food security. In our opinion, in order to further promote buffalo meat production and safeguard the consumer, a regulation providing only commercialization of subjects grown in the range of physiological weight gain should be prepared. The animal register of births could represent a good way of verifying if the slaughtering weight agrees with the aforementioned weight gains. The birth concentration in 6-7 months (using or not the out-of-breeding-mating strategy) could cause a discontinuous availability of meat on the market. This problem could be solved by slaughtering the subjects at different ages and modulating growth through the adoption of proper rationing schemes. In this case, the contemporary presence of animals with different ages should be taken into account planning the stalls in order to guarantee a good available space/head. In a multiple study between veterinarians and psychologists the owner-pet relationship was evaluated, focusing attention on the problems and the benefits due to the relationship (Bono, 2000; Camperio Ciani 2000; Bono et al., 2003) and a multiple approach method was validated to assess the quality of life of the pet in its relationship with the owner (Marinelli et al., 2001) . The aim of this study was to analyse the relationship between cats and their owners and identify factors that influence the quality of life of the cat. Participants were recruited as volunteers from the staff of the University of Padua and their relatives. Cats younger than six months of age and cats that did not live with their owners were excluded from the experiment. The cat-owner relationship was assessed using a multiple approach: four questionnaires and a simple physical examination (Marinelli et al., 2001) . The questionnaires investigated the owner's features (age, gender, education, marital status, job, features of the family, place of living, size of dwelling and social relations); the cat's features (age, gender, breed, neutered or spayed, age of adoption, source, whether it is living with other animals or not); the care given to the cat and the cat's behaviour (attachment to the owner, house soiling, behaviour towards strangers, cats and other animals). A veterinarian evaluated the cat's physical condition assessing only the nutritional status and ear condition, because in the previous study they were found to be objective indexes of physical condition. The aspects of the physical examination and the answers to the questionnaires regarding care and behaviour were codified on a score scale. The sum of the scores (total score) represented the index of the aspects considered (care given to the cat, cat behaviour and physical condition). The total scores were subjected to Spearman's Test to determine the correlation with the features both of the owner and the cat. A sample of 65 owners and 114 cats was analysed. 84.6% of the people were women. In 49.2% of the cases the age ranged from 21 to 30 years. 37% of the owners had degrees and 55.3% were employed. Most of them lived in the city (63%) and had had previous experience with a cat or a dog (89.2%). Most of the cats were cross breeds (95%). 60.5% were females (88.4% spayed) and 39.5% males (62.2% castrated). The cats, acquired in 22% of the cases in the most suitable period of their life (7th-10th week), had been given by friends (33.3%) or had been found by chance (43.8%). They lived with other pets (74.5% with cats and 27% with dogs) and 57.8% of them had had previous diseases. The results of statistical analysis showed that the total scores of the care given to the cats, the cat behaviour and the cat's physical condition correlated with the features both of the owner and the cat ( Table I ). The total score of the care given to the cats correlated positively with the sex (female) and the education of the owner and with the neutering of the cat, and negatively with the position of student. The total score of cat behaviour correlated positively with: old age and the sex (female) of the owner, the presence of children in the house, participation in organised social activities and care given by all family members, while it correlated negatively with the number of family members and the presence of other animals in the house. The total score of the physical examination correlated positively with the number of affective bonds and negatively with the age of the cat, its origin (given by friends) and neutering. The main obstacle we encountered in carrying out this study was that of finding a representative sample of volunteers. This was due to the need to respect people's privacy and also to the availability only of people sensitive to and very interested in this topic. The results showed that some features both of the owner and of the cat influence the quality of life of the cat. In particular, care given to the cat seems to depend mainly on the cultural and economic situation of the owner (factors such as education and employment position). Cat behaviour seems to be affected not only by its own social environment, but also by that of its owner (social activities, number of family members, presence of children and of other animals in the house). Physical condition does not appear to depend on specific aspects. This is due to the fact that physical condition is influenced by many factors which are not exclusively linked to the cat-owner relationship. Urinary incontinence is defined as unconscious urine loss during filling, the persistence of which could result in cutaneous disease and lower urinary tract infection for the bitch, in addition to causing sanitary and management problems for the owner. Spaying is one of the various mechanisms involved in urinary incontinence in the canine species (Gregory, 1994; Holt et al., 1993) . Among surgical (colposuspension, cystourethropexy, collagen or TeflonA prosthesis) or pharmacological (a-adrenergic agonists, oestrogens) treatments, we focused our attention on oestrogen replacement therapy using conjugated oestrogens. Literature suggests that oestrogens are able to improve urethral closure by the presence of specific local receptors able to modulate the number and sensitivity of alpha-2 receptors in the distal urinary tract (Creed, 1983; Miodrag et al., 1988) . In our work we investigate the incidence of urinary incontinence related to spaying in the bitch and evaluate the efficacy of oestrogen replacement therapy. Data are for the years between 1995 and 2001, and include 430 bitches of different ages, breeds and sizes. Follow ups were made every three months after ovariectomy or ovariohysterectomy to monitor the effect of spaying. Acquired incompetence of the ureteral sphincter was diagnosed after eliminating every other possible aetiology by means of diagnostic investigation (abdominal palpation, colposcopy, radiography, echography, urine analysis, etc. ...). The owners were advised about the possible sideeffects of long-term administration of oestrogens and about the importance of continuous medical treatment. A daily onset dosage able to block urinary incontinence 154 was found by the administration of natural oestrogens at different concentrations. After two weeks at the onset dosage with no urinary incontinence observed, we passed to a maintenance dosage, eventually reducing the frequency of administration. The bitches with no symptoms were examined once every three months and blood haemocytometric and biochemical analyses were performed to check bone marrow depression and hepatic function. The data concerning the type of surgery performed, the size and the body weight of the bitches were analysed by the use of a chi-square test and odds-ratio to evaluate their possibile correlation with urinary incontinence. 22 of 430 subjects observed (5.1%) presented symptoms of urinary incontinence, with a variable onset period between 1 to 60 months after surgery. Incidence was 72.8% within 12 months. 7 of 201 ovariectomized bitches and 15 of 229 ovariohysterectomized bitches presented urinary incontinence. Statistical analysis shows no difference between the two groups, with a chi-square value of 1.49 (p=0.222) . Correlation between the size of the subjects and the onset of urinary incontinence produces a chisquare value equal to 16.81 ( p<0.001), showing a tendency of large subjects (BW>20 kg) to aquire the disease. Concerning obesity, the possibile correlation between obese subjects before and after surgery and not-obese subjects (x2=3.647; p=0.056) is not significant, while the odds ratio value of 3.63 (1.13-11 .64) appears meaningful. The therapeutic plan performed determined the complete disappearance of symptoms by the administration of a posology adapted for the specific requirements of each subject (Table I) . Blood analysis did not show any modification. To date, the longest duration of replacement oestrogenic therapy is 49 months. Incidence of after-spaying urinary incontinence incidence in our data was 5.1%, in agreement with Thrusfield et al. (1998) , but less than the 20% reported by Arnold et al. (1989) . The relationship between acquired urinary incontinence and the type of surgery performed did not show meaningful values, as also reported by Okkens et al. (1997) , but in contrast with Osborne et al. (1999) , who reported a possible greater frequency of urinary incontinence in ovariohysterectomized subjects. In agreement with literature (Gregory, 1994; Holt et al., 1993) , it is clear that large subjects (>20 kg), more than those of small or medium size, showed a greater predisposition to the disease. It was also found that bitches that were overweight before spaying had 3.5 times more risk of developing the disease as compared with subjects that were not obese before and after surgery, confirming that obesity could be a predisposing factor for urinary incontinence. Oestrogen replacement therapy as treatment for after-spaying urinary incontinence, once the exact posology was found for each subject, induced a complete regression of symptoms, without the evidence of oestrogen-related side effects. Actin, a major cytoskeletal protein involved in many functions in mammalian cells, supports the process of spermatogenesis by sustaining the morphological changes occurring in the germ cells as well as the acquisition of motility by the spermatozoon (Lecuyer et al., 2000) . Moreover, the biological transformations that take place at a molecular level during the process of capacitation and leading to acrosome reaction also seem to be dependent on both actin localization and its polymeric status (F-actin). In the pig, as for other species, cytoskeletal reorganization has been related to either release of the acrosomal content or the fertilization rate (Castellani-Ceresa et al., 1993; Liu et al., 2002) . It can be hypothesized that in the sperm the cytoskeleton acts as an organizer of specific domains on the plasma membrane. This arrangement occurs during the process of sperm maturation by interaction of the actin cytoskeleton with enzymatic or receptor proteins and/or fusogenic lipids in the membrane, allowing the translocation of specific molecules with antigenic properties to the sites where they are more directly involved in functional processes (AR, fusion etc.). On the basis of this knowledge, the authors have studied the process of capacitation in boar spermatozoa by modulating actin polymerization with cytochalasin D, a specific inhibitor of filament formation. In particular the effect of the actin cytoskeleton on a) acquisition of the capacitation pattern by CTC, and b) incidence of acrosomal exocytosis elicited in the sperm after exposure to zonae pellucidae proteins has been evaluated. Finally, the study has focussed on the major events taking place in the lag time from zona stimulation to acrosomal exocytosis, in particular i) the Ca waves triggered by the exposure of single sperm cells to solubilized ZP and ii) the localization of PLCc 1 , a membrane linked enzyme acting in the Ca-dependent transduction pathways of AR. 158 The ejaculates of two boars of proven fertility were weekly recovered and immediately diluted in an extender type Modena to a final concentration of 2×107 sperm/mL, either with or without cyto D (20 mM). After at least 6 h of cyto D activity in the solutions, capacitation of each sample was carried out in a humidified incubator at 5% CO 2 at 38.5°C for 4 h, as previously described (Barboni et al., 1995) . The semen samples were then submitted to the following procedures: the capacitation levels achieved by the sperm were recorded using CTC, which identifies a specific fluorescent pattern depending on the functional status of the cells . Briefly, the sperm were incubated for 30 s with CTC 750 mM on a warm stage, then fixed in glutaraldehyde (0.4% v/v) and inspected under a fluorescent microscope. The percentage of AR recorded in the sperm exposed to solubilized ZP was detected after treatment of the samples with 95% ethanol by using FITC-conjugated PSA, a specific agglutinin which reacts with the acrosomal content of intact cells. The Ca levels after exposure to ZP were evaluated using a confocal microscope equipped with an argon laser on single cells with a kinetic program. In brief, after coincubating the samples with the membrane permeant, Ca sensitive dye Fluo4-AM, a drop of washed spermatozoa was placed into a registration chamber coated with poly-L-lysine to allow adhesion of the cells. After a few scans, solubilized ZP (1ZP/mL) were added to the chamber and the Ca signalling recorded. The localization of PLCc 1 was obtained by previously fixing (paraformaldeyde, 0.2%) and permeabilizing the cells (triton X-100, 0.1%); the sperm suspension was then incubated with anti PLCc 1 (1:250) raised in rabbit and the secondary antibody GaR conjugated with Cy3. The CTC staining carried out after four hours capacitation either with or without cyto D showed a similar proportion of cells that achieved the capacitated pattern (36.33±6.03 vs 34.87±2.2% respectively) characterized by a uniformly distributed fluorescence on the acrosomal region (Fig. 1) . In spite of the attainment of this capacitated pattern, when the sperm were challenged with solubilized ZP, for samples capacitated under control conditions 42-6% of the cells underwent AR, but only 12-0.5% of cyto D-treated sperm were able to complete the exocytotic event (Fig. 2) . The discrepancy of the obtained data suggests that actin polymerization is essential in order for the sperm to complete activation induced by the exposure to ZP with release of the acrosomal content. The measurement of intracellular Ca for single cells exposed to solubilized ZP did not show any significant differences in the kinetics of Ca signalling, or in the concentration achieved by sperm treated with cyto D, with respect to controls. In both groups, intracellular Ca levels peaked 20-30 s after ZP exposition, and gradually decreased to reach the basal concentration after 800 s. On the contrary, cyto D treatment influenced the localization of PLCc 1 , the enzyme From the data obtained it can be assumed that actin polymerization is the sine qua non condition that allows the sperm to successfully perform the acrosome reaction after being engaged by ZP proteins. Nevertheless, by blocking actin filament formation, some modifications that take place during the process of capacitation can be prevented. Sperm treated with cyto D, even if they are able to gain the capacitated pattern with CTC, are not able to complete the process through the exocytosis of the acrosomal content after exposure to ZP. It can be supposed that the lack of PLCc 1 expression at the equatorial level, as revealed by immunohistochemistry, could be responsible for this inability to undergo AR. This de-localization of the enzyme could cause an absent or limited release of intracellular Ca in specific subdomains, which may be pivotal for fusion of the plasma with the outer acrosomal membrane (Spungin et al., 1995) . This was impossible to ascertain by evaluation of the whole calcium response of sperm as in our study. Experiments are in progress in order to verify both the changes of localized Ca fluxes in distinct regions of the sperm head and the co-localization of PLCc 1 with actin filaments. These investigations could lead to a better understanding of the correlation between the actin polymerization status and the late events of the acrosome reaction, which take place after the engagement of sperm ZP receptors. Ultrasonographic identification and evaluation of foetal and extrafoetal structures allow determination of the gestational age and prediction of the delivery day for the bitch (Cartee and Rowles, 1984; England et al., 1990; Yeager et al., 1992; Gonzales-Bulnes et al., 1993; Moriyoshi et al., 1996; Luvoni and Grioni, 2000; Son et al., 2001) . This represents a useful tool for canine reproductive programs, especially in the case of multiple or uncertain mating times. The research on new parameters to improve accuracy of the prediction of the day of parturition for the bitch prompted our previous investigation, which aimed at identifying encephalic foetal structures during pregnancy using ultrasonography. The deep portion of telencephalic vesicle (DPTV), represented by thalamus and basal nuclei primordia, was visualized between the 30th and 8th day before parturition as an ovoidal anechoic area with clearly defined margins (Beccaglia et al., 2003) . The aims of the present study were the ultrasonographic evaluation of DPTV growth during pregnancy in bitches of different sizes and the assessment of the suitability of DPTV measurement in comparison with measurement of the inner chorionic cavity diameter (ICC) and biparietal diameter (BP) which are usually adopted to predict the day of parturition for this species. For the present study serial ultrasonographic examinations were performed three times a week (from 18th-20th day after last mating to the day of parturition) for seven bitches, four Jack Russell terriers of 5-7 kg bodyweight (small size) and three mixed breed bitches of 18-25 kg bodyweight (medium size). The ultrasonographic examinations were made using a Medison SonoAce 8800 equipped with a 7.5 MHz convex 162 probe. Depending on the pregnancy period, ICC, or BP and DPTV were identified during each examination. The results were expressed as days before parturition (day 0). A linear regression model was adopted to analyse the relationship between the growth of DPTV and the days remaining to parturition. The equation obtained relating to the DPTV diameter was further applied to predict the day of parturition for bitches with unknown breeding dates. The results were compared to those obtained using ICC and BP measurements by the application of the equations described in a previous study (Luvoni and Grioni, 2000) . The distribution of the data derived from DPTV measurement had a different trend for small (Fig. 1 ) and medium size (Fig. 2) dogs. For small dogs DPTV growth tended to reach a plateau, as shown from a significant ( p<0.001), but low regression coefficient (R2=0.70). For medium size dogs linear growth was observed (R2=0.91). The different first order coefficients denoted a different growth velocity of the DPTV in the two sizes of dog. The prediction of the delivery day derived from ICC and BP measurements, with an accuracy of ±1 day, was significantly more reliable than that obtained using DPTV ( p<0.05), irrespective of bitch size (Table I) . Extending the interval to ±2 days, these differences were not observed for the small dog group and were smaller in the medium size dogs. Although both DPTV and BP were detected in the same gestational period, the low number of DPTV observations was due to its small diameter in the first period of pregnancy and to the almost complete calcification of the skull in the last week of pregnancy, both of which limit its measurement. The results of this study showed that ultrasonographic evaluation of DPTV growth is possible for dogs of different sizes. This structure cannot be considered the optimal parameter to predict the day of parturition, but could be usefully combined with BP measurements in order to increase examination reliability. The endometrial cellular infiltrate has been the subject of various studies aimed at determining and clarifying its physiological and pathological importance. Gonzales et al. (1985) classified the endometria of non-pregnant cows into four categories and showed how, as the grade of severity of endometrial damage increases, the number of mast cells, which are considered fundamental in beginning and propagating the inflammatory process, increased notably. The correlation between endometrial mast cells and the quantity of some inflammatory mediators was studied by Belluzzi et al. (1994) who found a close correlation between the elevated values of leukotriene B 4 (LTB 4 ) and the granulocytic tissue infiltrate. No correlation was found for mast cells. The difference in the average values of LTB 4 and thromboxane B 2 (TXB 2 ) found in the endometria of cows having clinically evident endometritis and in healthy cows having normal reproductive activity was statistically significant for both mediators. A precise evaluation of the distribution of mast cells in the endometria of cows was carried out by Galeotti et al. (1997) , leukotriene C 4 and interleukin-6 (IL-6), all of which are inflammatory mediators potentially released by mast cells. 12 Holstein-Friesian cows between three and eight years of age culled for various reasons were studied. 24 h before slaughtering a bioptic endometrial sample and a blood sample were taken. The bioptic samples were dosed with LTB 4 and LTC 4 after column purification and with IL-6 without purification employing an EIA kit which uses human monoclonal antibodies. After slaughtering, five samples were taken from each uterus: one from the body, one from the base of each horn and one from the apex of each horn. The samples were fixed in buffered formalin and were then processed using normal histological techniques and stained with hematoxylin-eosin and toluidine blue in order to highlight the mast cells. The count of these mast cells was carried out in five optic fields (400×) for each of the five sample sites, according to a method described by Galeotti et al. (1997) . The hematic progesterone was dosed in order to evaluate the phase of the estral cycle. The statistical comparison between the values of the mediators and the number of mast cells counted at the base of the horn was carried out using linear regression. The comparison between the values of the mediators and mast cells in healthy endometria with respect to pathological ones was carried out using the non-parametric test according to Wald-Walfowitz. The values of the mast cell infiltrate in the five sample sites were compared with the non-parametric test for dependent samples according to Wilcoxon. The histological evaluation of the degree of total cellular infiltrate (granulocytes, lymphocytes, plasmacytes) allowed us to identify four uteri affected by endometritis and eight healthy ones. The dosage of IL-6 furnished values all below the calibration curve (5 pg) and were therefore not considered valid. The average values of LTB 4 and LTC 4 , taken from the bioptic sample, and the average value of the mast cells at the base of the horn and in the five samples taken post-mortem from healthy cows and those with endometritis are reported in Table I. 25% of the cows examined had different histological reports in the different sites of evaluation. This result could be a cause of disagreement regarding the representativeness of a single biopsy. The count of mast cells carried out using toluidine blue stain permitted us to confirm that the values found in the endometria deemed healthy are in agreement with the normal value defined by Galeotti et al. (1997) . The differences between the average values of mast cells present in healthy endometria with respect to the pathological ones are not significant however, nor are the differences in average concentrations of the two mediators in the two categories. However, a slight increase of LTC 4 in uteri with endometritis was observed, a category where the average count of mast cells was higher. The distribution of mast cells in the uterus does not seem to have distinctive characteristics in the five sites examined, since only the differences between the average values of the mast cells in the body with respect to the apex of the right horn were significant. This result is difficult to interpret and could be correlated with the number of pregnancies sustained by the horn and/or the site of the most recent pregnancy. Our research did not show any significant relationship between the number of mast cells present at the base of the horn and the value of the eicosanoids LTB 4 and LTC 4 obtained from bioptic sample at the same site. We need to consider that the staining using toluidine blue carried out by us shows only nondegranulated mast cells and therefore does not give an indication of the relationship between degranulated and non-degranulated mast cells in the optic field under examination. Other staining techniques (Kü ther et al., 1998) showed mast cells which do not coincide with those shown by toluidine blue in bovine endometria populations and, therefore, the technique used to visualize the mast cells can condition the result. A second consideration regards the mediators investigated in this study, that is LTB 4 , LTC 4 and IL-6, which are indicated in the literature as also being released by mast cells. However, at this time, there is no study that demonstrates their release in animal endometria. On the other hand, the relationship between the LTB 4 values and the neutrophil leukocyte infiltrate seems to be verified (Belluzzi et al., 1994) . The test results of the dosage of IL-6 suggest that the human monoclonal kit does not have a cross-reaction with bovine endometrial tissue. In conclusion, this study did not show any strict correlation between the inflammatory mediators examined and the mast cells highlighted with toluidine blue. Histochemical and immunohistochemical staining could furnish more reliable data on the number of mast cells present in the endometrium. The possibility that the mediators investigated and correlated with the 168 mast cells are also released by other cells constituting the endometrial infiltrate suggests that we turn our attention to substances which are more specifically attributable to mast cells, such as histamine, tryptase and chymase. The difficulties in the prediction of ovulation time in the mare, together with the increasing use of cooled and frozen semen for artificial insemination, have stimulated investigators to search for methods for induction of ovulation. These methods are essentially based on exogenous administration of one of three hormones directly or indirectly involved in the mechanism of ovulation: hCG (Human Chorionic Gonadotrophin), CEG (Crude Equine Gonadotrophin) (Duchamp et al., 1987) , or GnRH (Gonadotrophin-releasing hormone). Repeated administration of hCG, which is a heterologous glycoprotein for mares, may stimulate an immune response able to neutralise the effect of the hormone (Roser et al., 1979) . Although CEG has not shown this negative side effect, it is not found on the market, and purified extracts are produced and utilised only by few research groups. GnRH, physiologically released in a pulsatile manner, and its analogues have been tested for induction of ovulation in oestrous mares either by repeated injections at given time intervals (Palmer and Quellier, 1988) or as slow-release subcutaneous implants. For this purpose, a slowrelease subcutaneous implant containing the synthetic GnRH analogue deslorelin acetate (OvuplantA) is successfully used in United States and Australia (Meinert et al., 1993; Meyers et al., 1997) . The response to OvuplantA administration was similar to that obtained using hCG. However, this formulation is currently not available in the Italian and European market. The efficacy of buserelin, a GnRH analogue also commercialised in Italy, has been recently investigated for induction of ovulation in the mare. The administration of a single dose has not yielded encouraging results (Vidament et al., 1992) . On the other hand, Battut et al. (2001) observed that most mares treated twice daily with intravenous (IV) administration of 20 or 40 mg of buserelin ovulated within 48 h. The aim of our study was to evaluate the clinical efficacy of repeated administration of 40 mg buserelin for induction of ovulation in the mare, as compared to a single administration of 2500 IU hCG or a placebo. 170 During three breeding seasons, 37 Haflinger mares aged 2-5 years were kept in open paddocks and fed hay ad libitum. Their ovarian activity was monitored daily by transrectal ultrasonography (Dynamic Imaging System 2000, and a 5 MHz linear probe) for 83 oestrous cycles. When a dominant follicle equal to or larger than 32 mm in diameter was detected, together with a uterus showing the typical signs of oestrus, ultrasonographic examinations were intensified to one every 12 h, and mares were assigned to one of the three following treatments: Group C (control), treated with a placebo, consisting of 2 ml distilled water IV, twice daily until ovulation (n=19); Group Bu40, treated with buserelin acetate, 40 mg IV, twice daily until ovulation (n= 22); and Group hCG, treated with a single IV administration of 2500 IU hCG (n= 42). Mares were artificially inseminated in 29 cycles every 48 h from the day following the onset of treatment until ovulation, using at least 500×106 motile spermatozoa obtained from a stallion of proven fertility. At day 7 post-ovulation, 20/29 mares were subjected to uterine flushing for embryo recovery, while for the remaining 9 mares pregnancy diagnoses were carried out ultrasonographically at 14 days post-ovulation. Data are shown as means -standard deviation, and were analysed using the Student's t-test and Chi-squared-test. Over 83 oestrous cycles, 86 ovulations were observed: 78 single and 4 double ovulations, and one luteinization. The number of ovulations per cycle and the fertility of treated mares were not influenced by treatment. The mean time from onset of treatment to ovulation was significantly anticipated for both Group Bu40 and Group hCG, as compared to the control, and ovulation occurred earlier for mares belonging to Group hCG. Likewise, the proportion of ovulations occurring within 48 h of treatment was significantly lower for the control group, as compared to treated mares, while of those occurring between 36 and 48 h after treatment the proportion was 88.1% for Group hCG, significantly higher than for Group Bu40 (22.7%) and Group C (5.3%), which had no significant differences between them. Results are shown in detail in Table I . The results of this study showed that, although the buserelin protocol hastened ovulations in mares without interfering with their fertility as compared to the control group, its efficacy was clearly lower than that of the hCG protocol. These results are only partially in agreement with those found in the literature: Squires et al. (1988) showed a mean time from the beginning of twice daily administration of 40 mg buserelin and ovulation of 45.6±15.2 h, much shorter than the 74.9±41.9 h found in our study. Similarly, Battut et al. (2001) , when comparing the twice daily administration of 40 and 20 mg buserelin, reported a higher proportion of ovulations within 48 h of the start of the treatment (83% and 90% respectively), as compared to that observed in our study (43%). The aim, when inducing ovulation in mares, is not so much to hasten it, but to ensure that it occurs in a short and predictable time interval, in order to plan when the mare should be artificially inseminated. Contrary to what has been described in the literature, in our study treatment with buserelin did not seem to fulfil these criteria, as it was not able to induce ovulation in the mare in a repeatable and predictable way. prostaglandin analogue for induction of ovulation in mares. Proceedings of the 11th International Congress Animal Reproduction and Artificial Insemination, Dublin, 5, 460-462 Vidament, M., Arnaud, G., Trillaud-Geyl, C., Duchamp, G. and Palmer, E., 1992 . Analogue of GnRH (buserelin) and of PGF2a do not induce ovulation in mares. Proceedings of the 12th International Congress Animal Reproduction and Artificial Insemination, The Hague, 4, [1927] [1928] [1929] The storage of ovarian cortical strips has been proposed as a strategy for preservation of fertility, especially for young cancer patients or for women at risk of premature ovarian failure. In fact, the possibility of restoring fertility and achieving pregnancy after transplantation of cryopreserved ovarian cortical strips has been reported for large mammals (Gosden et al., 1994) , rodents (Cox et al., 1996) and recently for humans (Oktay, 2001) . In effect, primary/secondary follicles survive cryopreservation and grafting procedures in a large number, and can undergo subsequent development. However, the applicability of this procedure is affected by two main problems: firstly, the chemical and physical damage induced by the cryopreservation protocol, which results in a dramatic loss of potentially developing follicles; secondly, reimplantation of a stored ovarian cortex after chemotherapy does not exclude an eventual cancer relapse. Thus, an alternative option to preserve fertility relies on the possibility of obtaining mature oocytes from in vitro cultured small follicles isolated from frozen/thawed tissue. Also in this case it is necessary to develop cryopreservation protocols able to preserve follicle functionality and to identify culture conditions capable of successfully supporting full follicle development. The aim of our study was to compare the in vitro development of intact sheep preantral follicles isolated from cryopreserved ovarian cortical strips stored with two widely used CPAs, DMSO and EG. Prepubertal sheep ovaries were collected from the local abattoir and transported to the laboratory in a thermostatic container at 30°C within 1 h of collection. The ovaries were cut with surgical blades into fragments of about 0.5×0.5 cm that were 174 used to isolate unfrozen preantral follicles or to cryopreserve ovarian strips. Dimethyl sulfoxide (DMSO) or ethylene glycol (EG) were used at the final concentration of 1.5 M in M2 medium supplemented with 10% FCS, antibiotics (75 mg/l penicillin-G, 50 mg/l streptomycin sulfate) and 0.1 M sucrose. Ovarian pieces were firstly aspirated into plastic insemination straws (IMV, France), kept on ice for 15 min and transferred to a programmable freezer (Cryo-Logic, IVM, France). The straws were cryopreserved according to a slow cooling protocol developed by Gosden et al. (1994) then plunged into liquid nitrogen and stored for up to 3 months. Thawing was carried out by rapidly warming the straws to 38°C. Dilution of the cryoprotectants was performed at room temperature with HEPES-buffered TCM199 supplemented with decreasing concentrations of DMSO or EG and 0.1 M sucrose (3 steps, 15 min). Preantral follicles were mechanically isolated from either frozen or fresh ovary fragments in Hepes-buffered TCM199. Preantral follicles were then measured, and selected follicles of a mean diameter of 170±20 mm were individually transferred to 96-V-well microtiter plates and cultured in 25 ml of a-MEM supplemented with 2% FCS, 1% ITS, antibiotics and 1 mg/ml ovine FSH (oFSH). Culture was carried out for 10 days at 38°C in a O 2 5%, CO 2 5%, N 2 90% gas mixture (Cecconi et al., 1999) . Culture media were changed daily and 15 ml of recovered conditioned media was stored at -80°C for estradiol determination. At the end of culture follicles were analysed to record: 1) diameter; 2) percentage of differentiated antral-like cavities; 3) morphological aspect (by histological analysis) and 4) estradiol production. Cumulus oocyte complexes (COCs) were isolated from in vitro growth antral follicles and analysed to record: 1) percentage of isolated healthy oocytes presenting a surrounding layer of cumulus cells and 2) the metabolic coupling between germinal and somatic compartment (by evaluation of the 3H-uridine uptake). All parameters were evaluated in parallel for early antral follicles isolated from fresh ovaries. The first series of experiments was performed in order to compare the morphological architecture of preantral follicles isolated from fresh or frozen ovarian cortex. The majority (94%; n=33) of preantral follicles isolated from the fresh cortex showed the presence of oocytes with agranular cytoplasm and a central GV, surrounded by several layers of granulosa cells and a well organized theca layer. In contrast, 50% of EG (n=15) and 34% of DMSO (n=15; EG vs. DMSO: p<0.01) cryopreserved follicles revealed a damaged structure, frequently characterized by loss of intercellular communication between the oocyte and the surrounding granulosa cells, and signs of pyknosis in the GV as well as in some granulosa cells. Regarding growth, freshly isolated and frozen preantral follicles grew from a mean diameter of about 170 mm to a final diameter of approximately 300-310 mm after 10 days of culture, without significant differences between the experimental groups tested. A translucent antrallike cavity was recognizable starting from day 6 and, by the end of culture, the percentage of follicles forming antral-like cavities was similar for fresh and frozen follicles. This result was confirmed by histological analysis, which evidenced a similar morphological aspect between IVG antral follicles derived from fresh or cryopreserved tissue. Fresh follicles released increasing amounts of E 2 with a sharp rise from day 6 to day 10 (p<0.01). A similar final level of this steroid was found for EG cryopreserved follicles, although lower than for the group control, up to day 6 (p<0.05). E 2 production by DMSO treated follicles was comparable to that of those treated with EG up to day 6, but from that day onwards it remained significantly lower as compared with fresh and EG groups (p<0.01). On day 10, COCs were recovered from IVG antral follicles and their morphological aspect was assessed. As shown in Table I , the percentage of healthy COCs derived from IVG unfrozen follicles was significantly higher as compared to that of cryopreserved follicles, even though EG appeared to more efficiently protect COC quality. The extent of metabolic co-operation between the somatic and germinal compartments was measured by comparing coupling index (CI) values between the different experimental groups. CI values were loosely sorted into two classes, termed ''high'' (range 10-30) and ''low'' (<10). 100% and 80±10% of COCs obtained from control and IVG fresh follicles, respectively, displayed high CI values in comparison with 51±6% of the EG and 35±7% of the DMSO group. In this study we demonstrated that cryopreservation of sheep cortical tissue allows the recovery of viable preantral follicles, which can be grown in vitro with differing efficiency depending on the cryoprotective agent employed. Our results confirm previous observations indicating that the freezing/thawing process severely damages preantral follicles, which are considered less resistant than primordial ones to the injuries caused by cryopreservation (Gosden et al., 2002) . The increased complexity of the growing follicles makes them more sensitive to cryodamage. It is well known that the physiological differentiation of ovarian follicles in mammalia is strictly dependent on the existence of a bi-directional regulative dialogue between the germinal and somatic compartments, exerted through the production of paracrine/autocrine factors, or through the presence of functional gap junctions (Cecconi, 2002) . From our results, evaluation of the post-thaw viability of the follicles using morphological analysis shows that cryopreservation can cause disruption of gap junctional communication, thus impairing follicle survival. The extent of such damage is dependent on the cryoprotectant used, as demonstrated by the finding that the percentage of healthy follicles is higher in EG than DMSO. Our results also demonstrated that the in vitro culture of sheep preantral follicles is a useful test for evaluating the damage induced by CPA and/or freezing procedures, as previously proposed for the mouse (Newton et al., 2001) and the sheep (Amorin et al., 2003) . However, a comprehensive evaluation of cryodamage cannot be exclusively based upon morphological analysis but also needs to be assessed by more detailed biochemical investigations. (Platt, 2002) . The first electromyographic studies were performed by Adrian and Bronk in 1929 , by means of a specially manufactured electrode called concentric needle. In this method the electromyographers assess the traces empirically, by viewing them on an oscilloscope screen and by listening to their sound on an audio monitor, providing a semiquantitative evaluation of the muscle activity (Adrian and Bronk, 1929) . Recently, Wijnberg et al. (2002) performed, in horses, an offline analysis of the MUPs in a semiautomatic way. This method, where the MUPs were not automatically selected and were manually analyzed, is quite reliable and easy to use. However, to observe the results in real time, we used a multi-MUP EMG analysis, based on a ''turns'' analysis described by Willison (1964) and refined by Stalberg et al. (1995) . The main advantages of this method are that it is an online method, which permits an automatic expression of more than one MUP, it is fast enough for routine use, it does not require special electrodes and it is reproducible. The aim of our study was to create a normative database that can be used for electromyographic evaluation of muscular function in clinical studies and to compare the MUP parameters between the two groups of animals which are characterized by a different muscular mass. Electromyographic examination was performed in 10 healthy and not sedated horses, 4 male and 6 female, and in 9 healthy and not sedated ponies, 3 male and 6 female, using portable commercial EMG equipment (Keypoint, Dantec) with dedicated software. The animals were grounded using a subcutaneous needle electrode (ground electrode), positioned near a bony prominence. The EMG examination was performed by inserting a disposable concentric needle electrode into the muscle, advancing or withdrawing it in small steps and maintaining it in each new position with a firm grip, to sample different areas of the muscle. The EMG examination was performed in splenius, trapezius, and triceps muscles. Each muscle was evaluated at rest and during different degrees of muscle contraction. The software proposes a grouping of the MUPs, and many different MUPs were usually obtained from each recording site. First of all, the MUPs were divided into monophasic and polyphasic, according to the number of phases, and thus, for each subset of MUPs, the program analysed duration, amplitude, area/amplitude ratio and firing rate. All values are expressed as mean -standard deviation (s.d.). The statistical significance of differences between the two sets of data was assessed using the two tailed t-test for unpaired data. A p value <0.05 was regarded as a statistically significant difference. The mean values of the main parameters evaluated in splenius, trapezius and triceps muscles of the two groups of animals are reported in Table I . Our data shows that the number of MUPs is significantly greater for ponies, showing that it is easier to perform the EMG analysis for ponies. Therefore, our results demonstrate that the percentage of the polyphasic MUPs and the area/amplitude ratio are significantly lower for ponies, while the duration and the firing rate are significantly higher than those for horses. In contrast, no difference between the two groups was detected for the amplitude logarithm. This study presents normative data for healthy animals, which can be used for electromyographic evaluation of muscular function in clinical studies. Our results show that the parameters characterizing the MUPs in the two groups of healthy animals are significantly different, indicating that EMG can be influenced by different muscular mass and by the different composition of motor units. (6) Mammals react in a variable manner when stimulated by a stress or a novel situation through a behavioural coping reaction. Two main kinds of coping are known: a sympathetic reaction, with high general activity (active or proactive subjects), and a parasympathetic activity, with a general avoidance reaction, immobility and passivity (passive or reactive subjects) (Ruis et al., 2000) . In porcine species, individual responses can be related to genetics or direct learning, and the piglet's ability to react to a slight acute stress is easily determined by a backtest (BT) (Hessing et al., 1994) . In the BT the subject is put on its back and constrained for one minute, counting the escape attempts. The relationship between BT and production vocation or immunology is controversial (Hessing et al., 1994; Jensen, 1995) and has not yet been fully clarified. In order to evaluate the relationships between reactivity and somatic growth, a backtest trial was performed taking into account some factors that could influence the subject's growth and relating these to body weight in the period from 3 to 90 days. The trial was performed in two Northern Italian piggeries in March-September on Landrace×Large White piglets from 38 litters. All piglets of each litter were considered. The total number of subjects was 342 at the start of the trial (3 days) and 121 at the end (3 months), with an equal proportion of males and females. The difference in number of piglets between the starting period and the end period is due to the fact that 200 animals were sold and 11 died during the trial. Piglets were treated with iron dextran (200 mg iron) and males were castrated at the age of 5-10 days. All piglets 182 were weaned at the age of 25-28 days. All piglets were individually weighed at day 3, 10, 17, 30 and 90 of life. Individual reactivity was assessed by a backtest (BT) (Hessing et al., 1994) on day 10 and 17 of life: briefly, the piglet is held on its back by an operator, who grabs the head (extended), and the hindlegs quite firmly for one minute. The escape attempts (wriggles separated by a pause) are counted. The total number of attempts for each test is employed to calculate a ''reactivity class'', as reported by van Erp-van der Kooij et al. (2000) . Three classes are formed: high reacting piglets (HR), intermediate reactivity piglets (IR), and low reactive piglets (LR). In order to relate the reactivity class with the live weight, an ANOVA mixed model was performed: farm, sex, day and class were considered as fixed factors, whereas litter and piglet were considered as random effects, as reported below (random effects in bold): Weight ijklmn =Farm i +Sex j +Day k +ReactClass l +Litter m +Piglet n +e ijklmn Multiple comparisons between groups were performed using a Tukey post-ANOVA test. On day 10 of life, the mean BT score was 2.09±1.23, whilst on day 17, the mean BT score was 1.87±1.19, with a total range of 0-6 attempts of escape/min. The percentage of piglets included in the LR class was 39.18%, while the IR class had the lowest percentage (24.56%). The HR class has an intermediate value, with a percentage of 36.26%. Descriptive statistics (expressed as mean value−standard deviation) and mixed model ANOVA results are reported in Table I . Results show that in the first three months of life there is an effect of the reactivity class on weight, in addition to that due to sex and time of weighing. HR piglets (with high reactivity) had a significantly lower ( p<0.05) weight with respect to the other two groups (5.8±6.6 kg for the HR class versus 6.9±8.3 kg for the LR class and 7.2±9.2 kg for the IR class). It should be stressed that on day 3, the weight of the three groups did not differ. These results confirm those obtained in a previous study (Faustini et al., 2003) , where pre-weaning HR piglets showed lower body weight and growth. than LR or IR subjects. These results are not in agreement with those reported by Van Erp-Van der Kooij et al. These results contribute to a better characterization of a piglet's reactivity in response to a stressor. If the relationship between reactivity-growth versus composition were also confirmed for the latter phases of fattening, BT could be assumed as one of the parameters to permit the selection of animals with different growth-indices. In any case, a detailed analysis of the genetic effects of this aspect is necessary to confirm these preliminary results. DNA microsatellites or Short Tandem Repeats (STRs) are short, tandemly repeated sequences of a bi-, tri-or tetranucleotide unit with a random distribution throughout the genome (Ostrander et al., 1995) . They have been used extensively in applications as diverse as diagnosis of inherited diseases and forensic medicine for DNA fingerprinting and parentage testing. The success of this last application is due to the fact that STRs are highly polymorphic and, at the same time, they are sufficiently stable to be inherited unaltered from one generation to the next (Zajc et al., 1994; Ostrander et al., 1995; Muller et al., 1999) . The aim of this work is to show an STR application to solve a doubtful paternity case for a boxer litter. Five puppies of the same litter, their mother and three putative sires were subjected to parentage testing. All subjects were boxer dogs. Among the three stud dogs, only one was seen to mate with the bitch two different times. Despite this, the breeder could not exclude undesired couplings involving the other two stud dogs. DNA was extracted from whole blood samples in K 3 EDTA using a commercial available kit (Perfect Blood gDNA Mini Isolation Kit Eppendorf ); genomic DNA was PCR amplified on Mini Cycler MJ Research thermalcycler. Ten microsatellites were selected from the literature; their flanking sequences or their specific amplification primer pairs were available alternatively. To our knowledge, allele frequencies and polymorphism specific for boxer breed microsatellites have never been investigated; therefore microsatellites were chosen based on high Heterozygosity (He) and Polymorphism Information Content (PIC) values, previously calculated for other dog breeds, and product length suitable for the successive resolution on polyacrylamide gel (80-500 bp). The STRs selected were: AHT125, AHT140, AHT128, AHTf12, AHT121, Ren 02K21, Ren 05C07, FH 2001, FH2006, FH 2152 (Holmes et al., 1994; Holmes et al., 1995; Francisco et al., 1996; Holmes et al., 1998; Jouquand et al., 2000) . Primer pairs from the literature were employed in all cases. Moreover, a primer pair for the microsatellite Ren 02K21 was designed by the authors on the basis of the known flanking sequences. Optimal conditions were established for each microsatellite by adjusting MgClh, primer and template concentrations. The thermalcycling conditions were the same for all STRs tested. PCR products were run on 6% vertical polyacrylamide gel (NovexA TBE gel 6% Invitrogen; XCell Sure LockTM MiniCell, Invitrogen) and revealed on a U.V. transilluminator after EtBr staining. Gel images were acquired using a digital camera. For each microsatellite, PCR products for all tested dogs were evaluated on the same gel while the size of the amplicons was estimated using a 50 bp ladder as reference standards (Step Ladder 50 bp, Sigma). With the exception of microsatellite AHT f12, all STRs tested were correctly amplified with the conditions applied. Nevertheless, in the case of microsatellite Ren 02K21 only the primer pair designed by the authors, not those from literature, gave a specific PCR product. Six microsatellites, separated on polyacrylamide gel, did not show polymorphisms useful for the assignment of parentage. Three STRs (FH 2001, AHT121, Ren 02K21) revealed a full allele compatibility of all puppies with putative father no. 1, the only one who certainly mated the bitch, and excluded a possible superfecundation for puppies no. 1, 2 and 3. Furthermore, the exclusion of sire no. 2 from the paternity of puppy no. 4 was shown, while none of the possible fathers could be excluded from the paternity of puppy no. 5 because its allelic patterns were always compatible with those of all stud dogs (Fig. 1) . DNA polymerase slippage made the allelic pattern interpretation more difficult in two cases. The thermalcycling conditions applied caused formation of a very large number of aspecific PCR products during amplification of microsatellite Ren 05C07. Parentage testing interpretation relies on the fact that STRs are inherited in a true Mendelian fashion and express a codominant nature of allelic variants. Thus, a stud dog can be excluded as father of a litter when the puppies present one allele which could not have been inherited either from the bitch or the same stud dog. In the parentage testing presented, only putative sire no. 1 had full allele compatibility with all puppies. The visualisation of PCR products using polyacrylamide gel can reveal, with some limitations, allele polymorphisms able to solve doubtful paternity cases. Nevertheless, in canine forensic medicine the case to solve often concerns possible mixed paternity, which means that puppies of the same litter may have been fathered by different sires (i.e. superfecundation). On this basis, the technology applied in this study allowed us to partially assign the paternity, leaving it unsolved for two of the five puppies. This may be explained by two considerations: the absence of literature data about microsatellite polymorphism in boxer dogs may have caused the poor choice of markers for solving parentage in this breed; moreover, the visualisation technology applied only allows us to distinguish alleles that differ greatly in size, at the same time reducing the potential information content of each microsatellite. The family of trophic factors consisting of neurotrophins and their specific Trk receptors has been studied by many authors for their role at the level of the central and peripheral nervous system (see review by Huang and Reichardt, 2001) . It has recently been demonstrated that these proteins also play an important role in the development and maintenance of the immune system. Data are present in the literature regarding the expression of neurotrophins and their receptors in the thymus of mammals, both at the level of mRNA and protein (Maroder et al., 1996) , but data about the expression and localization of these factors in lymphoid organs still remain incomplete. Transgenic animals provide the most powerful way to demonstrate the effect of a protein in vivo. Mice with mutations targeted in genes codifying for the neurotrophins and Trk receptors provided important information about the role of these proteins in the development of the nervous system (Fariñ as, 1999) . Previous studies carried out using this experimental model demonstrated that the neurotrophins and Trk receptors play an important role in the development and maintenance of the thymic homeostasis (Aloe et al., 1999; Garcia-Suarez et al., 2000) . Therefore, the aim of this study is to present the most recent data obtained in our laboratory and to clarify the possible functional meaning of the neurotrophins as modulators of the thymic microenvironment. Thymus of C57B1/6 mice, removed under deep chloral hydrate anesthesia (350 mg/Kh, i.p.) was processed for light microscopy, single and double immunohistochemistry, western-blot, RT-PCR and northern-blot analysis. The antibodies used for immunohistochemical procedures were: anti-TrkA and anti-TrkB (Santa Cruz Biotechnology, CA, USA), anti-cytokeratins (DPC, Los Angeles, USA), anti-S100 (Dako, Copenhagen, Denmark), and anti-F4/80 (Serotec, Oxford, UK). For transmission electron microscopy analysis knock-out for trkA and trkB mice of the same stock from the colony of the Department of Molecular Biology, Bristol-Myers-Squibb Inc., Princeton, NJ (kindly provided by Dr. Silos-Santiago) was perfused transcardially with 2.5% glutaraldehyde. The thymuses removed were fixed in the same perfusion liquid, post-fixed in 1% osmium tetroxide, dehydrated and embedded in DurcupanA ACM (Fluka). Ultrathin sections were examined under a Jeol Jem 100 SX electron microscope. The results obtained by Western-blot showed the presence of a proteinic band in the mouse thymus with a molecular weight of 140 kDa, corresponding to the full length TrkA receptor. The single and double immunohistochemistry showed that this protein was present in cells with cytoplasmic prolongations, localized at subcapsular and medullary levels. These cells were positive to cytokeratins, which are specific markers for epithelial cells, thus confirming their epithelial origin. The ultrastructural analysis of mouse thymus knock-out for TrkA showed clear morphological modifications at the medullary level, consisting of cysts of different sizes, lined by a monostratified epithelium which was often covered by cilia and containing scarce muciparous cells. The presence of these cysts reveals the persistence of a structure with embryonic origin. The results obtained by Northern-blot demonstrated expression of the mRNA for TrkB in thymus of mice at different ages (0 and 15 days pn). An age-dependent decrease of the mRNA for thymus TrkB was observed, while this protein remained constant at the brain level used as a positive control. The analysis of the Western-blot carried out on samples of whole thymus, isolated lymphocytes, isolated stromal cells and brain of 15 day-old animals demonstrated, in all samples examined, the presence of a proteinic band of 145 kDa corresponding to the full length TrkB receptor. The immunohistochemical analysis revealed a strong immunoreactivity in stromal cells localized on the cortico-medullary border. The double immunohistochemistry for TrkB and F4/80 identified the non lymphoid cells positive to TrkB as macrophages. The ultrastructural study carried out on the thymus of TrkB knock-out mice demonstrated clear modifications in lymphocytes and stromal cells. The modifications of 191 lymphocytes consisted of fragmentation or absence of the nucleus and accumulation of cytoplasmic electrondense bodies (apoptotic bodies), as well as evident signs of apoptosis. Neurotrophins are well known for their action on nervous tissues (Fariñ as, 1999) but recent data have suggested that these proteins also act in the regulation of the immune system (Otten and Gadient, 1995; Aloe et al., 1999; Ruberti et al., 2000) . Mice deficient in neurotrophins or those having non-functional neurotrophin receptors have added substantial information to the knowledge of the role of neurotrophins in nervous system development (Fariñ as, 1999) . However, the non-neuronal phenotype of these mouse strains has only been poorly analyzed, and remains basically unknown. Regarding TrkA deficient mice, we have demonstrated in this study that the lack of this protein provokes an altered organogenesis of the murine thymus. The thymus of TrkA knock-out mice showed the presence of cysts of different sizes lined by a single line of cells, cilia and mucous cells, probably due to the absence of trka functional genes. The mechanism through which the TrkA receptor could act on the development of thymic epithelial cells remains unclear. Regarding TrkB, the present study demonstrates that the thymus of mice lacking functional TrkB develops thymic alterations, at least in the cell population that normally expresses this receptor. The homozygous mice carrying a mutation in the trkb gene resulting in a non functional TrkB protein, showed structural and ultrastructural changes mainly consisting of massive cell death of T-lymphocytes for apoptosis. These data suggest that TrkB is involved in the prevention of the death of cortical lymphocytes. Our results are in agreement with previous data reported by Maroder et al. (1996) , who observed that BDNF prevented cell death of immature thymocytes. Taken together, these results lend support to the idea that in the mammalian thymus both the NGF/TrkA and the BDNF/TrkB systems play important roles in the development and maintenance of epithelial cells and thymocytes, respectively, and that both ligand-receptor complexes are involved in the intercellular communication between these two main cell types of the thymus. This study on transgenic mice greatly contributes to the understanding of the possible role of neurotrophins in the immune system and the importance of these growth factors in health and disease. The recombinant form of bovine ST (rBST) is used to increase milk yield in cows. This kind of treatment, which is authorised in many countries, has been definitively banned by the E.U. because of its potential hazard to animal health (Kronfeld, 2001; Macrì, 1999) . There are currently no analytical methods able to reveal rBST administration. However, there are circulating molecules which can be considered as analytical targets and indirect indicators for the abovementioned treatment of bovines, since their concentration is proportional to the stimulative action of ST. The most famous of these is Insulin-like Growth Factor 1 (IGF-1), a hormone mainly secreted by the liver (Prosser et al., 1989 ) that mediates the biological action of ST on several tissues. Almost all seric IGF-1 circulates bound to specific IGF-binding proteins (IGFBPs); for this reason the current analytical methods used to measure it are still imprecise (Breier et al., 1991; Twigg et al., 2000) . On the other hand, the direct analytical approach is very difficult because the primary sequence of recombinant BST differs from the native form for only one aminoacid at the N-terminal position. The aim of our work is to evaluate the possibility of producing antibodies capable of direct discrimination, between BST and rBST. We have used the murine immune system as our model. We produced polyclonal antibodies against a synthetic peptide that mimics the N-terminal region of rBST. Immunoenzymatic tests of the sera derived from immunised animals allowed us to evaluate their eventual different reactivity against the two forms of the protein, suggesting the possibility of producing monoclonal antibodies 194 with a higher affinity for the recombinant form. Immunoglobulins with these characteristics are determinants for the development of analytical methods able to reveal the rBST pharmacological treatment. With the aim of obtaining antibodies against the N-terminal region of rBST, balb-c mice were immunised with a complex formed from a peptide conjugated to BSA or KLH, used as a carrier protein (BSA-Peptide and KLH-Peptide). The peptide sequence corresponds to the first ten rBST N-terminal amino acids, followed by a tail of lysines and serines to augment water solubility: H 3 Five mice were treated with BSA-peptide and another five with KLH-peptide. Two immunisation cycles were performed using 50 mg of the peptide conjugated in Freund's adjuvant (complete for the first cycle, incomplete for the following 3 boosts) administered every two weeks subcutaneously. Then, after confirmation of the antibody titre, mice were sacrificed for bleeding. Antigens and analytical methods: BST, rBST, BSA-peptide, KLH-Peptide, and BSA, KLH and myoglobin, used a as control antigen, were utilised at different concentrations to evaluate the specific antibody titre of the sera. ELISA and Western blots were utilised as immunoenzymatic tests. Direct ELISA were performed by incubating sera at various dilutions with the antigen bound to the microtitration plate and polyclonal anti-mouse total IgG conjugated with alkaline phosphatase was used for the binding revelation. The electroblotting first required a SDS-Page antigen separation and their subsequent transfer onto nitrocellulose membranes. Membranes were there incubated with immune sera and the final revelation was performed as previously described for ELISA. In both groups of mice, the immunoenzymatic tests revealed a much higher antisera avidity toward carrier-peptide complexes than toward the two species of ST, even when the peptide was conjugated to a different carrier than the one employed for the immunization (Fig. 1) . This evidence can be explained by considering that only a fraction of the antibodies developed against the peptide is also able to recognize the BST head, namely those directed against the first amino acids in the N-terminal position of the protein: especially those presenting an intrinsic steric adaptability. Specifically comparing the reactivity of the whole sera on the two types of BST, we observed a positive reaction for both of them in eight out of the ten analyzed mice. The fact that some mice did not produce an appreciable immune response against the BST N-terminal portion could be explained by considering the remarkable similarity 195 Figure 1 . Mean values of ELISA performed with the sera of the three best mice immunised with KLH which showed a higher avidity to rBST as compared to BST between the ST of bovines and that of rodents. Within the group of responsive mice, the reaction of the antibodies was comparable for both molecular forms in four animals (three of which had been immunized with BSA-peptide), whereas the other four (three of which had been immunized with KLH-peptide) showed visibly higher reactivity against the recombinant form. These results suggest a possible proliferation of lymphocytic clones producing immunoglobulins with a higher affinity for the recombinant form of ST. Consequently, we consider the efforts aimed at producing monoclonal antibodies with these characteristics, using the lymphopoietic tissues derived from animals with a higher discriminative response, to be a valid approach to this problem. Prosser, C.G., Fleet, I.R. and Corps, A.N., 1989. Increased secretion of insulin-like growth factor I into milk of cows treated with recombinantly derived bovine growth hormone. Journal of Dairy Research, 56, 17-26. Twigg, S.M., Hardman, K.V. and Baxter, R.C., 2000. A purified bovine serum albumin preparation contains an insulin-like growth factor (IGF) binding protein-3 fragment that forms ternary complexes selectively with IGF-II and the acid-labile subunit. Growth Hormone and IGF Research, Aug., 10, 215-223. Programmed cell death (PCD) is an essential process in development that often occurs through activation of an apoptotic mechanism. Apoptosis can be detected in numerous areas of the central nervous system (CNS) under normal conditions but also in several neurodegenerative diseases. In the postnatal cerebellum (Lossi et al., 2002) apoptosis affects mainly pre-migratory granule cells (CGCs), their precursors within the external granular layer (EGL), and post-mitotic, post-migratory CGCs within the internal granular layer (IGL). Apoptosis is characterized morphologically by a series of nuclear alterations including fragmentation and chromatin condensation eventually leading to the appearance of the so called apoptotic bodies, consisting of highly condensed cellular and nuclear fragments. Numerous proteins are involved in the execution of apoptosis, among which caspases play a pivotal role (Allsopp et al., 2000) . Caspases are synthesized as inactive precursors that are activated following proteolysis. In CGCs, caspase 3 is activated at the endpoint of a cascade of events which affects other members of this protein family. The main substrate of active caspase 3 is poly-ADP-ribosepolymerase-1 (PARP-1 - Fig. 1A) , a protein normally involved in genomic surveillance and DNA reparation (Smith, 2001) . Following caspase 3 cleavage, a 85 kD fragment (p85) is generated that is known to be responsible for at least some of the morphological changes of apoptosis. Other proteins involved in the regulation of apoptosis are related to cell cycle progression and checkpoint: these include the retinoblastoma protein (Rb) and several kinases that control the transition between the various phases of the cell cycle depending upon their degree of phosphorylation (Padmanaban et al., 2001) . In the current work we have studied the distribution of p85 and some cell cycle- Cerebella were obtained from New Zealand rabbits at postnatal day 5 (P5) after vascular perfusion with 4% paraformaldehyde. Parasagittal sections were labeled using the TUNEL method for in situ detection of DNA fragmentation and the ABC procedure for immunocytochemical detection of cleaved caspase 3, p85, phospho-Rb (ser780), phospho-Rb (ser795), phospho-Chk1, and phospho-cdc2 (Tyr15). The presence of p85 was also assessed in Western blotting from cerebellar extracts. Within the EGL there is no positive stain for cleaved caspase 3 or p85. In this layer of the forming cerebellar cortex we observed a positive immunoreaction for phospho-Rb, which is responsible for blocking the G 1 /S transition of the cell cycle, at the level of premigratory CGCs. Within the EGL there is also a positive stain for phospho-Chk1, which blocks the cell cycle at G 2 /M. Cells with the cytoplasmic stain for cdc2 correspond to proliferating neurons. Within the IGL there is no expression of these cell cycle-related proteins, whereas a positive signal for cleaved caspase 3 and p85 was detected, with a cell distribution overlapping that of TUNEL positive nuclei ( Fig. 1B-D) . This work demonstrates that in the postnatal cerebellum two different apoptotic pathways occur, affecting the CGCs at different stages of their maturation. In the IGL the apoptotic pathway is ''classically'' executed by caspases, with cleavage of caspase 3 and PARP-1 as the major endpoint events. In the EGL apoptosis appears to be caspase-independent and involves the activation/inactivation of certain checkpoint proteins and, most likely, some mitochondrial factors such as the apoptosis-inducing factor (AIF) (Susin et al., 1999) . AIF is a recently discovered molecule that is normally localized inside mitochondria, but, following an apoptotic stimulus, it translocates to the nucleus, where it triggers some of the morphological alterations typical of programmed cell death. Sarcomeric myosin is characterised by a high structural heterogeneity which is essential to achieve different functional performances. This variability is based on the presence of different myosin heavy chain isoforms (MHC) which are considered to be the marker of fibre types. Different genes that codify for type 1, 2A, 2X and 2B myosin heavy chain isoforms (paralogous isoforms) present in skeletal muscle have been identified. Furthermore, there are also less frequent isoforms which are expressed in some special muscles or transitionally expressed during development. When corresponding isoforms from different species are compared (orthologous isoforms) it is possible to detect a low nucleotidic variability that determines the different contractile performances of muscle fibres. In muscle, ''pure'' fibres express a single myosin heavy chain isoform whereas ''hybrid'' fibres express, in the same fibre, more than one MHC isoform: for example type 1/2A, (conventionally called 2C), 2A/X and 2X/B. The isoform type 2B (2B fibre type) has been detected only in small mammals (mouse, rat, rabbit and marsupials) and is the most represented isoform in some muscles. This isoform is not expressed in humans although the gene is present. Moreover, the fibre type originally classified histochemically as 2B in humans shows a higher similarity to the 2X of rat and mouse than to the 2B of the same species. However, Bottinelli and Reggiani (2000) have shown that in man, where only three MHC isoforms (type 1, 2A and 2X) are present, the functional plasticity of muscle is not lost when compared to small mammals in which the 2B isoform is also expressed. As far as other species (dog, cat, horse, monkey, goat) are concerned, recent studies have identified only three type of fibres or three isoforms as for humans. Two of these isoforms (type 1 and type 2A) are very similar to the rat and mouse isoforms while the third one seems to be more similar to 2X. The general opinion among researchers 202 is that the 2B type would be specific only for small mammals and marsupials. The latter hypothesis has not been confirmed for the pig where the 2B isoform is present and is expressed in several muscles and, as shown by our research, is in association with the isoform 2X (hybrid fibres 2X/B). In bovine, histochemical, immunohistochemical and electrophoretic data have identified three isoforms, indicated as type 1, 2A and 2B (Duris et al., 2000) ; these data have been only partially confirmed by gene expression studies where the 1, 2A and 2X isoforms have been identified (Tanabe et al., 1998) . Since the isoform 2B might also be expressed in large animals, the aim of this study was to assess the presence of the gene codifying the 2B isoform, its expression by RT-PCR and the presence of the corresponding protein by gel-electrophoresis. Samples were dissected from bovine muscles with different composition of fibre types, such as masseter, longissimus, extensor carpi radialis, diaphragma, pectoral, extraocular muscles (rectus lateralis) and retractor bulbi. Genomic DNA was extracted from bovine blood and two regions of 2B isoform gene were amplified using degenerate primers, designed on the basis of known sequences (man and pig). The two isolated fragments were cloned and sequenced. Total RNA was extracted from dissected muscles and reverse transcribed using a reverse transcriptase enzyme. Based on our genomic results, bovine 2B MHC specific primers were designed, while specific oligonucleotides to amplify the type 1, 2A and 2X MHC isoforms were designed on bovine sequences data present in literature. Fragments of the above described sampled muscles were solubilised in Laemmli solution. Proteins were separated on 9% polyacrylamide slab gels following a procedure developed by Blough et al. (1996) by modifying electrophoretic running parameters. Electrophoresis was run at constant 50 Volt for 2 h followed by 40 h at 200 V at 4°C. Electrophoretic bands in the migration zone of the myosin heavy chain 203 (200 kD) were identified by comparison with samples previously analysed by RT-PCR. Two fragments of 2B isoform were sequenced from genomic bovine DNA. The first one (123 nt, GenBank AY135646) corresponding to the human exon 40, shows a strong similarity (96%) to the 2X and 2B and therefore seems inadequate for a precise identification of 2B gene. The second fragment (515 nt GenBank AY227972) was amplified with a forward primer designed on the loop2 exon 16 (a region that presents the higher variability between paralogus isoforms) and with a reverse primer on exon 21. The sequence obtained from the second fragment is more similar to 2B than to 2X published sequences. The nucleotidic identity score is 93% with 2B isoform of man and pig, while the nucleotidic identity score with 2X isoform is 90% (man) and 91% (pig and bovine). The phylogenetic tree confirms the position of our sequence in the orthologous 2B group. RT-PCR analysis shows the expression of the type 1 isoform in masseter muscle, of type 1 and 2A in diaphragma, and of type 1, 2A and 2X in longissimus, pectoral, extensor carpi radialis and extraocular muscles. In all muscles examined only the rectrator bulbi and the rectus lateralis express the type 2B isoform together with type 1, 2A and 2X. Protein electrophoretic separation shows the same pattern of expression observed by RT-PCR. Comparing muscles with different composition of fibre types allowed precise identification of the electrophoretic bands of type 1, 2A and 2X. The band that corresponds to 2B isoform is difficult to identify since its electrophoretic migration is similar to that of the extraocular band, a MHC typical of rotator muscles (such as the rectus dorsali) but not or scarcely present in retractor bulbi (Sartore et al., 1987) . Our results show that the 2B MHC gene is present in the bovine genome. However, skeletal muscles express only type 1, 2A and 2X MHC. The pig is therefore the only species where 2B isoform is expressed in skeletal muscles (Chikuni et al., 2001) although often in association with the isoform 2X (hybrid fibres). In man, recent data indicate the expression of 2B mRNA in some particular muscle such as the masseter but the corresponding protein has not been detected. In bovine, only a very peculiar muscle such as the retractor bulbi, shows expression of 2B and the corresponding protein shows an electrophoretic migration velocity very similar to the extraocular isoform. Therefore, it will be important to improve the SDS-gel electrophoresis technique in order to distinguish these two bands. It is difficult to explain the presence of this isoform in the retractor bulbi in view of its poor function and the disappearance of this muscle in man. Embryo development is dependent on the complex control of gene expression that is strictly regulated through specific rearrangements of maternal and paternal chromatin after fertilization. Most of the epigenetic changes of the genome structure are realized at an early stage after fertilization and include mechanisms of DNA methylation and histone acetylation. This process of genome remodelling seems to be possible thanks to the cytoplasmic machinery prepared by the oocyte during its maturation, since incomplete oocyte maturation, as often occurs following in vitro protocols, results in defects in gene expression and leads to abnormal embryo development (Young, 2000) . On this basis the present research has been designed to describe and compare the epigenetic modifications (methylation/acetylation) that take place in pig zygotes obtained by fertilizing oocytes matured in vivo or in vitro. For this research zygotes produced in vivo or in vitro were utilized. For in vivo experiments prepubertal gilts were injected with 1250 IU of eCG, to induce follicular growth and maturation and, 60 h later, with 750 IU of hCG to induce ovulation within 40-44 h. 30 h later the hCG gilts were inseminated with semen obtained from boars of proven fertility and the zygotes were collected after 10-12 h by uterine flushing carried out under anaesthesia. For embryos produced in vitro, COCs isolated from healthy follicles, 4-5 mm in diameter, were matured in tissue culture medium (TCM) 199 with the support of both the gonadotropins (1 mg/ml USDA-pLH-B1 and 1 mg/ml USDA-pFSH-B1) for 40 h, as previously described (Mattioli, 1994) . The in vitro fertilization (IVF) protocol was carried out by co-incubating denuded oocytes for 2 h with capacitated spermatozoa at a concentration of 2×106 spz/ml. Fresh 206 semen from boars of proven fertility was used for the IVF procedure. Oocytes were then transferred to modified TCM 199 (Mattioli, 1989) and incubated for a further 8-10 h to reach the pronuclei stage. In order to evaluate and compare the pattern of DNA methylation and histone acetylation, the zygotes, produced either in vivo or in vitro at the same developmental stage, were utilized for the immunofluorescence protocol. To this end the zygotes were briefly incubated in Tyrode's solution and transferred to 0.1% pronase to remove the zona pellucida, washed with 0.05% Tween-20 in PBS, fixed for 1 h in 4% paraformaldehyde in PBS and permeabilized with 0.2% Triton X-100 in PBS for 30 min at room temperature. For the detection of global DNA methylation, zygotes were treated with 2M HCl at room temperature for 30 min to effect DNA denaturation and a specific primary antibody, mouse monoclonal anti-5-methyl-cytosine (dilution 1:500), was used. For the detection of global histone acetylation fixed and permeabilized zygotes were immediately incubated with the primary antibody, an anti-hyperacetylated histone H4 (dilution 1:300). For both the methylation and acetylation protocols the primary antibody was detected by the specific secondary antibody coupled with Cy-3 (dilution 1:400). Negative control samples, in which the primary antibody was omitted, were also evaluated for each experiment. At the end of the immunofluorescence protocol zygotes were analyzed using a Bio-Rad confocal laser scanning microscope equipped with a krypton/argon ion laser. Digital optical sections were obtained by scanning a z series of 0.2 mm slices throughout the planes of focus containing maternal and paternal pronuclei and the images were later projected. Some zygotes were fixed in acetic acid/ethanol (1:3) for 24 h, permeabilized and processed both for staining with lacmoid and observation under a microscope equipped with phase contrast, in order to identify the male pronucleus by the presence of the sperm tail nearby, and then for the immunofluorescence protocol and confocal microscopy analysis for evaluation of DNA methylation and histone acetylation. The evaluation of the global methylation pattern of DNA in pig zygotes obtained in vivo showed that in most of the zygotes (79%, n=39) maternal and paternal pronuclei showed a different pattern of methylation. The male pronucleus resulted frequently demethylated (Table I , panel A), in agreement with results obtained for other species (mouse, rat and bovine; Mayer, 2000; Dean et al., 2001) . The global methylation pattern of DNA evaluated in pig zygotes produced in vitro showed that only 40% (n=45, p<0.05) revealed demethylation of the paternal genome (Table I) Zygotes produced in vitro 40%*, n=45 Zygotes produced in vitro 67%*, n=45 Panel A: DNA methylation in parental pronuclei: male genome (arrow) more demethylated as compared to the female one. Panel B: histone acetylation in parental pronuclei: male genome (arrow) hyperacetylated and its pattern of fluorescence characterized by the presence of clusters, in contrast with the more homogeneous distribution observed in the female pronucleus. *Indicates data significantly different ( p<0.05). The analysis of global histone acetylation in zygotes produced in vivo demonstrated that, in addition to a lower degree of DNA methylation, most of the male pronuclei (87%, n=41) also showed a global condition of hyperacetylation of histone H4 as compared to the female pronuclei (Table I, panel B) , with the male pronucleus characterized by clusters of chromatin asymmetrically distributed inside it, as compared to the female pronucleus where the fluorescence was homogeneously distributed. The same evaluation carried out for zygotes produced in vitro showed that only 67% of male pronuclei were hyperacetylated (n=45; p<0.05), also in this case showing a reduced and slowed ability of oocytes to remodel the paternal genome. In addition, the morphology of the pronuclei and the pattern of global DNA acetylation was often more irregular as compared to that of oocytes matured in vivo. The results presented in the present paper show that a process of chromatin remodelling takes place in pig oocytes during the early hours following fertilization. This is 208 mainly due to epigenetic modifications, such as DNA methylation and histone acetylation, and was found to be differentiated between maternal and paternal pronuclei. The analysis at a global level of DNA methylation, first carried out in zygotes produced in vivo, has in fact demonstrated that the male genome is characterized by a lower level of DNA methylation as compared to the female one. Moreover, the observation that the paternal pronucleus was also characterized by a higher level of histone acetylation in most of the zygotes analysed seems to indicate an early transcription of the male chromatin component. The same analysis of DNA methylation and histone acetylation subsequently carried out in pig zygotes obtained from oocytes matured and fertilized in vitro, demonstrated that the ability to correctly remodel the sperm chromatin is markedly reduced when compared to that of zygotes obtained in vivo. This demonstrated that the process of oocyte maturation is crucial for the oocyte to develop the mechanisms necessary for complete chromatin rearrangement. In fact, when the process of maturation doesn't occur under physiological conditions, such as in the case of maturation in vitro, this results in a reduced ability to correctly remodel the paternal genome that may compromise the success of embryo development. It is known that carbon monoxide (CO) is a toxic gas that rapidly leads to death by means of the action of carboxyhemoglobin (COHb). Recent work has shown that low doses of CO induce anti-inflammatory and anti-apoptotic effects in endothelial cells. Moreover, in experimental animal models of cardiac transplant, CO administration was found to prolong the survival time of the graft (Bach et al., 1997; Brouard et al., 2000; . CO is generated by the action of constitutive (HO-2) and inducible (HO-1) heme oxygenase, that are widespread in the organism. The expression of HO-1 is enhanced by cytokines and endotoxins and both HO-1 and CO suppress the pro-inflammatory response and promote the anti-inflammatory response of monocytes. Because some cytokines may induce the vascular leakage that characterizes inflammatory and septic diseases, the aim of our work was to evaluate the possible protective effect of low doses of inhaled CO in an experimental model of swine endotoxic shock. In our model, we have studied the main functional and molecular parameters related to the onset of lung oedema, such as the compliance of the respiratory system (C rs ), the mean pulmonary arterial pressure (MPAP), the extravascular lung water (EVLW) and endothelial vascular growth factor (VEGF). 210 The study was performed on eight Large White pigs, weighing 22.0±2.69 kg (Mean±d.s.). The animals, sedated with propionylpromazine hydrochloride 1% (0.05 ml/kg i.m.) and anaesthetised with thiopental sodium (bolus of 15 mg/kg, followed by an infusion of 9 mg/kg/h), were tracheostomised and mechanically ventilated with 100% O 2 . The respiratory flow and tidal volume were measured using a pneumotachograph (Fleisch n.2) inserted between the endotracheal tube and the ventilator, while tracheal pressure was measured by connecting a pressure transducer to a side port of the endotracheal cannula. A polyethylene catheter was inserted into the right femoral artery to monitor systemic arterial pressure and a Swan-Ganz catheter was introduced into the pulmonary artery to evaluate pulmonary arterial pressure (MPAP) and cardiac output. Intrathoracic blood and extravascular lung water volumes were evaluated using a thermodilution technique (PiCCO, 22G-5F; Pulsion Medical System, Germany). All signals were recorded every 30 min on a multichannel pen recorder polygraph, while the compliance (C rs ) and the resistance (R rs ) of the respiratory system were calculated using standard formulas. Blood samples were collected for the ELISA test of VEGF. The animals of group 1 (n=4) were treated with LPS of E. coli (40 mg/kg/h i.v. for 240 min) after 60 min of O 2 inhalation, while those of group 2 (n=4) inhaled CO for 60 min (250 ppm in air), and then received LPS, as group 1. At the end of the experiment, some samples of cardiac and pulmonary tissues were collected for HO-1 Western Blot analysis. The statistical significance of the differences between baseline values and posttreatment conditions were evaluated using the T-Test for paired measurements. A p value of <0.05 was considered to be statistically significant. Our results show that the administration of LPS induced a biphasic increase of mean pulmonary arterial pressure. The pre-treatment with carbon monoxide did not alter the early LPS-induced hypertensive response, while it did cause a greater increase of the MPAP in the last phase of experimental time. Despite the strengthening of pulmonary hypertension, CO administration reduced the changes in C rs induced by administration of LPS. These data, that indicate a protective mechanism exerted by CO on respiratory mechanics, are corroborated by the EVLW changes. Our results showed that, in animals treated with CO, the volume of extravascular lung water significantly decreased, particularly in the early phase of endotoxic shock. The plasma levels of VEGF in animals treated only with LPS were 416±120 pg/ml, and those in animals pre-treated with carbon monoxide were between 0 and 15 pg/ml, confirming the hypothesis of a protective mechanism CO-induced on the altered capillary permeability due to LPS (Karmpaliotis et al., 2002) . In our experimental model, the expression of HO-1 in cardiac and lung tissues was not modified, confirming that activation of HO-1 requires a longer experimental period, as demonstrated in a previous ''in vitro'' study. Therefore, the results that we obtained in our hyper-acute model of endotoxic shock are due to inhaled CO, without any interference from neo-synthesized molecules in the tissue. Ghrelin is a 28 amino acid growth hormone-releasing peptide isolated from the rat stomach (Kojima et al., 1999) . This molecule possesses a unique serine residue at the third position (Ser3) that is modified by n-octanoid acid. Acylation is essential for ligand binding to the receptor GHS-R and subsequent ghrelin activity. In mammals ghrelin is expressed in a restricted hypothalamic area, the arcuate nucleus (Lu et al., 2002) . This indicates that ghrelin may play a role in the regulation of feeding behavior and energy metabolism in the central nervous system (Date et al., 2000) . Ghrelin is widely expressed in a variety of peripheral tissues, i.e. the stomach, intestine, placenta, kidney and pancreas (Kojima et al., 1999; Date et al., 2000; Mori et al., 2000; Dornonville de la Cour et al., 2001; Gualillo et al., 2001) . Although little is known about ghrelin in non-mammalian vertebrates, ghrelin has been identified in the brain and stomach of a frog (Rana esculenta), showing the same acylation at the third aminoacidic residue as for mammals (Galas et al., 2002) . Recently, Kaya et al. (2002) purified chicken ghrelin from the stomach of a chicken (Gallus gallus) and characterized its structure. Chicken ghrelin consists of 26 amino acids showing acylation at Ser3 and the same sequence in the N-terminal region as for ghrelin from humans and rats. In the present study we have used antisera against rat ghrelin to investigate the distribution of ghrelin-like peptide in the gastroenteric system of the chicken. Three-month-old chickens Gallus domesticus of both sexes were used in this study. The gastrointestinal tract was sectioned in small samples obtained from the following segments: proventriculus, ventriculus, duodenum, jejunum, ileum and rectum (colon) (Baumel J). The tissues were immediately fixed in Bouin solution for 24 h, dehydrated 214 and embedded in paraffin. Sections of 6/7 mm in length were processed according to the peroxidase-antiperoxidase method, and incubated with a rabbit ghrelin against the N-terminal sequence of the rat ghrelin (ghrelin [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] ) diluted 1:750. The specificity of the reaction was tested by omitting both primary or secondary antibodies, and utilizing primary antibody preabsorbed with an excess of antigen. The sections were examined using a Leitz Aristoplan microscope equipped with a Leica DC 300F digital camera. Ghrelin immunoreactive cells (ghrelin-ic) were mainly found in the mucosal layer of the proventriculus and also in the small and large intestine. In the proventriculus ghrelin-ic were generally found at the base of glandular lobuli and in the basal zone of the plicae. Most of the ghrelin-ic were small and rounded in shape. These cells, intensely immunoreactive (ir) especially in the basal cytoplasm and named ''closedtype cells'', were not in contact with the lumen. A few ghrelin-ic, elongated in shape, showed an apical process that contacted the lumen and were the so-called ''open-type cells''. Ghrelin immunoreactivity decreases from the pilorus toward the enteric distal tracts. In the duodenum in particular the two types of ghrelin-ic, ''closed'' and ''open'' cells, were still observed along villi intestinales. Ghrelin-ic were less numerous in the jejunum and ileum, and were generally of the ''closed'' type. In the colon rare triangular-shaped ''closed'' ghrelin-ic were detected. In this study we have clearly demonstrated that ghrelin producing cells exist in the chicken gastrointestinal tract. The distribution pattern and morphological characteristics of ghrelin-ic suggest that they correspond to endocrine type cells. However, further investigations are required in order to identify the specific cell types according to the current classification (Solcia et al., 2000) . In the frog stomach ghrelin endocrine cells have been observed in the mucosa and in pit-associated glands (Galas et al., 2002) . In the rat and human stomach ghrelin-producing endocrine cells have been identified as X/A like cells of the oxyntic mucosa (Kojima et al., 1999; Date et al., 2000) . These results indicate that chicken ghrelin is also evolutionally conserved both in its distribution and in its aminoacidic sequence. High yielding dairy cows are particularly vulnerable, during the transition period, to any event able to stimulate the immune system (viruses, bacteria, many kind of stress, endotoxins, trauma, etc.). Pro-inflammatory cytokines (i.e. IL-1, IL-6, TNFa), mainly released from monocyte-macrophage cells, are the common mediators of these events. Cytokines are able to activate other cellules like those of hypothalamus, liver, reproductive organs etc. (Elsasser et al., 1997) . As previously stressed (Bertoni et al., 2000; Trevisi et al., 2002) , the consequent endocrine-metabolic changes can cause marked negative effects: anorexia, lipomobilization, reduction of milk yield and reproductive traits. Therefore, the modulation of cytokine release and/or the inhibition of some related effects could represent a noticeable target in the transition period. The aim of our study concerned the evaluation of the effects of a preventive administration of a well-known anti-inflammatory drug (acetyl-salicylate, able to inhibit both cyclooxygenases -COX-1 and COX-2 -the key enzymes of inflammatory response) in post-calving dairy cows. The research was carried out in a commercial herd where dry cows were fed with 3 kg of alfalfa hay and grass hay ad libitum, while lactating cows received TMR ad libitum (28.3% maize-silage, 21.2% grass hay, 3.8% alfalfa hay, 46.7% concentrate all on dry matter basis). Twenty-two pluriparous cows were allocated in two homogeneous groups, CTR and LAS, according to their calving number, BCS and time from calving. LAS received daily i.m. injections of lysine acetyl-salicylate (Gellini Farmaceutici, S.p.a., LT, Italy) in the first 5 days of lactation (DIM): 15 g/d in the first 3 DIM and 7.5 g/d in the following 2 days. The CTR group was used as a control. The milk yield (3, 7, 10, 14, 21, 28, 42, 56, 70, 90, 120 DIM), 14, 28, 42, 56 DIM) , metabolic profile (3, 7, 10, 14, 21, 28, 46, 56 DIM) , disease occurrence and fertility were monitored. In addition, representative feed samples were collected for proximate analysis 218 and the evaluation of rations. The statistical analysis was performed by SASTM (8 TS M0 version) MIXED procedure, using cow within treatment, DIM, treatment and DIM-treatment interaction as factors. The lactation curve of each cow was estimated according to Wood equation by Proc. NLIN of SASTM. During the trial, the diet composition did not change either in dry (0.70 UFL/kg, d.m., 11.6% CP d.m.) or in lactating groups (0.92 UFL/kg, d.m., 15.1% CP d.m.) . Dry matter intake (DMI) was also acceptable in dry and lactation periods, 11.3 and 22.9 kg/d respectively; but fresh cows showed an energy (-6%) and a protein (-15%) deficiency, according to their milk yield level. LAS cows showed: a higher milk yield ( p<0.10 at the peak; Fig. 1 ); a more marked (−1.08 vs −0.92 score of CTR) and more prolonged reduction of BCS; a lower frequency of anorexia at the beginning of lactation (10 vs 45% of CTR). Conversely, LAS was affected by higher frequency of metritis (30 vs 18% of CTR), although the reproductive troubles that were observed during the trial were similar (about 50%) in the two groups. Moreover, LAS showed better fertility traits: more cows pregnant after 1st insemination (50 vs 27%), less services per conception (2.11 vs 2.33 n°), a slight reduction of open days (126 vs 133 d) and fewer cows culled for infertility (1 vs 2) . During the 1st week of lactation, LAS showed a lower level of glucose ( p<0.05 at 3 DIM) and higher levels of NEFA and GOT in comparison to CTR, suggesting a more marked energy deficiency explained by the higher milk production. Nevertheless, the pattern of changes of plasma bOHB did not differ between groups, while urea reached higher levels (but within reference interval) in LAS. Furthermore, LAS showed more favourable indices of inflammatory status: higher Zn ( p<0.05 at 7 DIM), lower acute phase proteins (haptoglobin and particularly ceruloplasmin, as shown in Fig. 1) , total proteins and globulins ( p<0.05). The less severe inflammatory Figure 1 . Behaviour of milk yield and plasma ceruloplasmin in dairy cows treated (LAS) or not (CTR) with lysine acetyl-salicylate in post-calving. Different letters showed a significant difference between groups (a,b: p<0.05) 219 processes observed in LAS had considerably improved the levels of proteins (sometimes in a significant manner) of common liver synthesis: albumin, lipoproteins (total cholesterol) and retinol-binding protein (vitamin A). The treatment with acetyl-salicylate seems to minimize the undesired consequences of the inflammatory events around calving, which occurred also in treated cows, as confirmed by the increase of acute phase proteins. The positive effects of LAS treatments could also be seen in the quicker increase of DMI that was suggested by the lower frequency of anorexia and the similar levels of bOHB, despite a more pronounced energy deficiency (confirmed by the more marked reduction of BCS). Moreover, the higher DMI of LAS also seems supported by the higher plasma urea level, mainly related to the total protein intake when diet composition is unchanged. The better health status (i.e. more favourable indices of inflammation) and perhaps the higher feed intake of LAS allowed a prompt recovery of common metabolic conditions, a better liver activity, a rise in milk yield and an improvement in fertility. In conclusion, the use of molecules able to modulate the eicosanoids after calving seems promising, but further research is required. The last month of pregnancy and suckling are two critical periods in the sheep production cycle. Moreover, in both cases, the nutritive requirements increase while the dry matter (DM) intake is not adequate. Therefore, energy deficits may well occur with negative consequences on the body weight of lambs at birth and on subsequent milk production. Thus the aim of our paper was to study the effect of two diets differing in energy value on the productive performance of Sarda ewes during the last month of pregnancy and suckling. The study was carried out on a Sarda sheep farm in the province of Caserta in which the animals were bred in the stall system. The study concerned 30 pregnant sheep, for which single births had been diagnosed by echograph. One month before lambing, they were divided into two groups, whose diets consisted of (% on dry matter): maize silage (35), cut oat hay (30) and concentrate (35), the latter two differing for the two groups. Group A was fed a commercial concentrate and group B received a mix of the concentrate (14), maize (13.4) and soybean meal (7.6) in order to diversify the NSC contents of the two diets, maintaining the same percentage of crude protein. Table I shows the chemical characteristics of the diets (ASPA, 1980; Martillotti et al., 1987) . The nutritive value of the diets was estimated following INRA (1988) indications. Under the rationing scheme, different quantities of DM according to body weight (BW) were administered twice daily as TMR (2% BW at the 5th gestation month and colostral period, 3.7% from the 5th to 7th day; 4.5% from the 8th to 14th day and 5% from the 15th). After lambing, due to complications, two sheep from group A and three from group B were eliminated. The male/female ratio between lambs born was 54:46 in group A and 50:50 in group B. The lambs were weighed at birth and at sale (30 days old) in order to calculate their weight gain during suckling. With such measurements we were able to estimate the average amount of milk intake from lambs during suckling using the equation developed by Poujardieu (1969) . All the results obtained were processed 223 using the GLM procedure of SAS (2000) following the model: y ijk =m+ a i +a j +ab ij +e ijk , where y ijk is the single observation, a i is the diet effect (i=1, 2), a j =is the sex effect ( j=male or female), ab ij is the interaction between diet and sex, and e ijk is the error. Table II shows the results relative to: weight at birth and at sale, weight gain (WG) and the milk consumed by lambs (MC) according to mothers' diet and sex. The different NSC content of the diets induced statistically significant differences (p<0.05) only for weight at sale. Interestingly, WG and MC were higher for the group that received the diet with the highest nutritive value. The results, in agreement with those of Cannas (2001) , could be due to the high level of somatotrope hormone (SH) present in the blood during the first period of lactation: this would also direct energy use (and hence NSC use) towards milk production instead of the deposition of body fat (Peel et al., 1987) . Regarding the differences between the sexes, the males showed higher values ( p<0.01) for all the examined parameters. Diet×sex interaction was never found to be statistically significant. The results obtained show that use of the diet with a higher NSC content could improve the productive performance of primiparous Sarda ewes during the last month of pregnancy and suckling. The absence of statistically significant differences between body weight and milk consumed by lambs could be justified by the high variability of the results obtained. Feeding of marine fish widely uses the strategy which consists of the partial substitution of crude protein with lipids (protein-sparing effect). However, high dietary lipid content modifies whole body chemical composition, resulting in a higher adipogenetic deposition which influences the organoleptic quality of fish products (Segato et al., 2003a) . This influence on quality can be assessed by sensory analyses such as a discriminative taste panel. The aim of this study was to evaluate the sensory quality of fish in relation to dietary fat content on organoleptic parameters of European sea bass reared in mariculture. European sea bass (initial body weight: 304±30 g) were reared in two floating cages (density: 9 kg/m3) placed in the Gulf of Trieste. During summer and autumn of 2001 (97 experimental days) fish were fed (1% BW) two isoproteic diets with different crude fat level (Low vs. High). Proximate composition (AOAC, 2000) and gross energy of the two experimental diets are reported in Table I . At the end of the trial, 30 fish (final body weight: 412±26 vs 407±22 g) from each experimental group were caught and killed on ice. After 48 h conservation at 2±1°C, discriminative sensory analysis was carried out on fish (n=8) by 10 trained tasters. Gutted trunks were wrapped in tinfoil and cooked in an electric oven at 165°C until the fillet core temperature reached 60°C. Jury evaluation considered the following attributes: trunk compactness, exudation and skin detachment; fillet (white epiaxial muscle) odour, lightness, tenderness, juiciness, marine and salty flavour and taste. The score was expressed on a 9-point scale (1=low ... 9=high) using the attributes schedule in Table II . For tasting each judge considered the same dorsal muscular portion, avoiding intermuscular adipose deposits. Data were submitted to one-way ANOVA according to a mono-experimental design (diet effect) by PROC GLM of SAS (1999) . Trunk compactness and significant visible exudation were lower for fish fed with the High diet (Table II) . This result is probably due to the higher fillet fat content observed in High diet fed fish (7.7% vs. 8.9%; p<0.1 -data not tabulated). The fillet structure is influenced by intra-and perimuscular fat content and its location which modify texture and water hold capacity (Torrissen et al., 2001) . The higher exudative quantity observed in the fish fed the Low diet trunk was probably due to moisture losses caused by more intensive protein denaturation during heating treatment. On the contrary, fish fed the higher fat content diet showed higher values of fillet juiciness and tenderness. These quality traits also are dependent on the higher level of muscular fat. With regard to juiciness, suitable marbling limits cellular liquid losses during heating and induces chewing salivation. Lipids also limit the cohesion between muscular fibres and between muscular and collagen fibres, reducing firmness of the cooked flesh (Segato et al., 2003b; Torrissen, ibidem) . Data of sensory tenderness were confirmed by maximum shear force determination (7.8 vs. 6.0 N; p<005 -data not tabulated). Panellists did not detect any differences in fillet lightness while instrumental measurements using the Hunter-L*a*b* method showed significant differences (L*:77.6 vs. 81.4; p<0.01 -data not tabulated). Some sensory parameters (i.e. colour) showed thresholds which are not perceptible even by trained panellists. Organoleptic traits such as odour, flavour and taste were not influenced by diet any factors. Considering the dietary fat range tested (18-25%), this result shows that differences detected by smell and taste of cooked products could mainly be related to the properties of the raw materials used in the fattening diet rather than to the level of crude fat administered. In conclusion, feeding a high-fat diet to mariculture European sea bass leads to higher adipogenesis both as lipidic deposits and as intramuscular fat; flesh from this rearing condition is characterized by lower cooking losses and higher juiciness and tenderness. Ochratoxin A (O.A), a secondary metabolite with teratogenic, hepatotoxic, carcinogenic and nephrotoxic activity in many animal species, including human beings (Hohler, 1998) , is mainly produced by Aspergillus ochraceus and Penicillium viridicatum. Being ubiquitous, these moulds can easily contaminate foodstuffs. O.A is predominantly found in cereal grains, cereal products, legumes, oilseed, coffee beans and feed. Moreover, O.A has been identified in the tissue and organs of animals (pigs, chicken) fed contaminated feed. The presence of O.A has also been reported in human milk. The IARC (International Agency for Research in Cancer) has classified O.A as Group 2B, a possible human carcinogen (IARC, 1993) . In several areas of Eastern Europe, where chronic exposure to O.A occurs, involvement of this mycotoxin in the cancer aetiology of the urinary system, and in kidney pathologies typical of Balkan Endemic Nephropathy (BEN) has been suspected. Studies into the correlation between O.A and BEN (Puntaric et al., 2001) have shown higher O.A contamination levels in cereals from endemic areas with respect to cereals from non-endemic areas. The high risk for Public Health represented by such a mycotoxin and the insufficient epidemiological data relevant to Italian regions, which neighbour the Balkan area, suggested to us the present survey aimed at evaluation of overall status of O.A contamination levels in some cereals produced in Puglia and Basilicata, which are regions where the economy is primarily based on agriculture. 260 cereal samples (95 durum wheat, 80 maize, and 85 barley) from Puglia and Basilicata were examined. Just before analysis, each sample was finely ground with a 230 Cemotec 1090 mill (Tecator) to a particle size of <1 mm, 25 g of homogenized sample was taken for analysis, using the procedure described by Nesheim (Nesheim et al., 1992) . O.A was extracted with chloroform-aqueous phosphoric acid, and isolated by liquid-liquid partitioning into an aqueous bicarbonate solution. The extract was transferred to a C 18 solid-phase extraction (SPE) column, and O.A was eluted with ethyl acetate-methanol-acetic acid (95/5/0.5). O.A was identified by reversed-phase Liquid Chromatography (LC) and quantified by fluorescence detection. The performance of this method has been evaluated and described by Palermo (Palermo et al., 2002) in a previous review. Apparatus: The chromatographic system used was a Waters Model 600 E (Waters Milford, MA, USA) equipped with a Waters Model 579 fluorescence detector and a Waters Model 5200 integrator-recorder. Reagents: Ethyl acetate, chloroform, benzene, methanol and water were HPLC grade, and acetic acid, sodium bicarbonate and phosphoric acid were ACS grade (J.T.Baker, Koeln, Germany). Celite 545 (Merck, Darmstadt, Germany), SPE-Isolute-C 18 microcolumns (Intern.Sorbent Techn., Mid Glamorgan, U.K.). Standard O.A was purchased from Sigma (Sigma Chem. Co., St Louis, Missouri, USA). Chromatographic conditions: The LC column used was a Supelcosil C 18 reversedphase packing, 4.6×250 mm, 5 mm film thickness (Supelco, Bellefonte, PA, USA). The mobile phase was a water-acetonitrile-acetic acid (99/99/2) mixture and the flow rate was 0.8 ml/min. The temperature of the column was 30°C and the injection loop was 20 ml. The fluorescence detector was regulated at 330 nm for excitation and 470 nm for emission wavelengths. The results on the O.A levels in various cereal samples are reported in Table I . O.A was detected (c>LOD) in 20.4% of examined samples. Only one maize sample from Trinitapoli (Foggia) had an O.A content (5.2 ng/g) above the maximum tolerance limit for O.A in cereals (5.0 ng/g), recently fixed by the European Community Regulation no. 472/2002. The highest frequency of contamination was in the range of LOD -0.9 ng/g for 13.5% of examined samples, while only 6.5% had an O.A value between 1.0 to 4.9 ng/g. The highest mean level of contamination (30.0%) was found for maize. Overall, the obtained data are reassuring. In fact, only 1.3% of samples were contaminated with a value higher than the maximum limit of tolerance of Ochratoxin A in cereals as fixed by the European Community. The distribution of the O.A positive samples, according to the areas where the sampled cereals were produced, is reported in Table II from Puglia and Basilicata with an O.A concentration above the EC legal limit. This is a reassuring finding on the overall status of the O.A contamination level in the cereals produced in these regions of Italy. O.A contamination was in the range of 1.0 to 4.9 ng/g in 6.5% of examined samples, a fact that indicates that the surveillance of O.A in cereals and other matrices should be continued, in compliance with the EC programs on food quality assurance. The application of Ovum Pick-up (OPU) technology, together with multistep embryo production in vitro (IVEP), represents a valid procedure for the recovery of oocytes from live donors and the attainment of a large number of embryos (Looney et al., 1994; . In fact, collection of oocytes by means of ultrasound guided follicular aspiration is routinely performed with success in large prepubescent and adult ruminant species . In the buffalo species, because of their low response to superovulation treatments (Karaivanov, 1986) , OPU technology, together with an improvement in the multistep process of in vitro embryo production, seems to be the best way to enhance genetic progression through the maternal lineage Neglia et al., 2003) . However, in vitro embryo production in buffalo has always been associated with a lower efficiency as compared to the bovine (Totey et al., 1992) , although recently some encouraging results have been reported both from slaughterhouse recovered oocytes (Gasparrini et al., 2000) , and from OPUderived oocytes (Neglia et al., 2003) . Obviously, in order to optimize the procedure, an efficient method for embryo freezing should be developed. In fact, until now the only pregnancies recorded for the buffalo species were obtained from the transfer of fresh embryos produced in vitro. In 1991 Madan et al. (Madan et al., 1991) reported the birth of the first buffalo calf after the transfer of a fresh blastocyst obtained through in vitro maturation (IVM), in vitro fertilization (IVF) and in co-culture with oviductal cells (IVC). Four other calves were produced in India in 1994, by transfer of fresh blastocysts obtained in vitro (Madan et al., 1994) , using the same method. In 1997, Boni et al. (Boni et al., 1997) , reported two pregnancies obtained from the transfer of two fresh morulae, produced after OPU/IVEP procedures. Only three calves have been produced from frozen blastocysts (Galli et al., 1998) ; in this case zygotes (2 days after IVF) were transferred into ligated sheep oviduct before freezing. The possibility of cryopreserving buffalo embryos by vitrification with good efficiency in terms of in vitro survival rate, has recently been reported (Neglia et al., 2001) . On the basis of these results a one month trial was carried out on a buffalo farm in order to verify the efficiency of this freezing method, by transferring vitrified and fresh embryos produced using OPU-IVEP technology. In March 2002 12 lactating pluriparous buffalo cows underwent repeated transvaginal follicular aspiration twice-weekly for seven sessions. OPU was carried out by using a portable ultrasound unit (Aloka SSD-500) with a 5 MHz sector scanner and a metal guide to fit 17 gauge needles both allocated in a properly designed vaginal guide. A vacuum pressure of -40 mm Hg was constantly maintained and all visible antral follicles were punctured. The Cumulus Oocyte Complexes (COCs) were searched immediately after filtering the aspirated follicular fluid and aspiration medium and classified under 5 categories (A, B, C, D, E), according to a standard classification (Neglia et al., 2003) . COCs were placed in 25 mM hepes buffered TCM 199 supplemented with 0.5 mg/ml FSH, 5 mg/ml LH, 1 mg/ml estradiol and 50 mM cysteamine (Gasparrini et al., 2000) in a portable incubator at 38.5°C and moved to the lab within 4 to 6 h. In the laboratory COCs were transferred into 50 ml droplets of a final maturation medium consisting of bicarbonate-buffered TCM 199 (B 199) with hormones and cysteamine in the same concentration previously described under medical oil. The droplets were incubated at 38.5°C for 22-24 h under a controlled gas atmosphere consisting of 5% CO 2 in humidified air. The day after IVM the oocytes were in vitro fertilized in 50 ml droplets of Fert-TALP, supplemented with 0.2 mM/ml penicillamine, 0.1 mM/ml hypotaurine and 0.01 mM/ml heparin. Sperm was treated using the swim-up procedure in a modified version of Ham's medium and resuspended in a final concentration of 2×106/ml. Fertilizing droplets were then incubated under the same gas atmosphere as for in vitro maturation. After 20-22 h presumptive zygotes were cultured in 20 ml droplets of SOFaaBSA in a modulation chamber with a gas atmosphere of 5% CO 2 , 7% O 2 and 88% N 2 . On day 5 (day 0=IVF day) the cleavage rate was assessed and embryos were transferred into fresh droplets of the same medium for two further days of culture. Final embryo output (morulae +blastocysts) was evaluated on day 7 of culture. Some embryos were transferred fresh into opportunely synchronized buffalo heifers, by using a double injection of PGF 2a 12 days apart. Other embryos were vitrified using a method previously described (Neglia et al., 2001) and transferred after warming. Confirmation of pregnancy was obtained by rectal palpation of the genital apparatus at 45 and 120 days (after embryo transfer). The cleavage rate (48.1%) as well as total morulae+blastocyst rate (19.6%) were lower than our standard results (Neglia et al., 2003) , probably because a hemorrhagic 235 diarrhoea was diagnosed in the animals during the trial. A total number of 15 embryos were transferred: 7 blastocysts were transferred fresh and 8 after vitrification and warming. No pregnancies were assessed from fresh transferred embryos, whereas three pregnancies were found at 45 and 120 days after transfer of vitrified embryos (37.5%). Unfortunately one pregnancy was interrupted at 180 days, while the other two reached full term. In February 2003 the first two buffalo calves from vitrified embryos entirely produced in vitro were born. As mentioned above, only three calves have been produced from frozen embryos (Galli et al., 1998) , when zygotes were cultured into ligated ewe oviduct. In this way the sheep represents a temporary host, in which embryos can develop until day 7: this step improves the quality of the embryo, which grows under ''normal'' conditions and is more resistant to the cryopreservation process. In the buffalo species, the entire in vitro embryo production system has been developed by extrapolating information acquired in more studied species (cattle and sheep): for this reason the embryo nutritional requirements are unknown. Further studies are therefore necessary in order to optimize the in vitro development conditions of buffalo embryos. We underline the lack of pregnancies established by fresh embryos: this surprising finding may be explained by the higher sensitivity of buffalo embryos to environmental stress as compared with bovine (Neglia et al., 2003) . However, the results obtained in this trial, demonstrate the possibility of using vitrified embryos produced in vitro in order to produce buffalo calves. This may in turn allow the application of OPU/IVEP technology to accelerate genetic progression through maternal lineage. The replacement of fish meal with plant protein sources is a major issue of interest in the aquaculture of Teleostei (Tacon, 1994) . Although several studies have been undertaken over the last two decades, many questions remain about the feasibility of the total substitution of traditional protein sources. Moreover, the effects of the protein source on the quality characteristics of the product obtained have been little investigated (Médale, 2003) . The aim of this study was to evaluate the effect of a diet based only on plant protein sources on the behaviour of some biochemical and physical parameters during the first 24 h after death. Sterile triploid (to avoid possible effects of reproduction) rainbow trout (Oncorhynchus mykiss, Walbaum) at an average initial body weight of 19 g were divided into two groups, reared under the same conditions (water temperature 17±1°C; dissolved oxygen 8.5 ppm) and fed two isoproteic (49% crude protein) and isoenergetic (21 kJ/g digestible energy) diets (two replicate for diet), different for the protein source. One diet contained fish meal (FM) as its protein source, while the second one contained only a mixture of plant protein ingredients (PP). The diets were adequately balanced for amino acidic content. After 157 days of rearing, the fish (having an average final body weight of 832 and 670 g, respectively for FM and PP) were killed by a blow to the head, after a cold narcosis in water and ice. 40 fish from each diet group were analysed during the first 24 h after death (10 fish at 0 h and 6 fish at each of the next 238 sampling times: 2, 3, 4, 5 and 24 h), while the remaining trout were retained for analysis performed each 24 h over the whole shelf life to determine other parameters (not presented here). Immediately after death, all fish were weighed and total and standard length were measured. 30 trout from each group were covered in ice and stored at 1°C until the time of measurement. Immediately after death a sample of cranial epaxial muscle was withdrawn for 10 fish per diet and immediately frozen in liquid nitrogen for determination of ATP and its catabolites (Burns and Ke, 1985) . The adenylate energy charge [AEC: (ATP+0.5ADP)/(ATP+ADP+AMP)] was then computed from the concentration of ATP and its catabolites. The eye liquor and epaxial muscle pH and the dielectric properties (Fish Tester VI) were also measured for the same animal. The same parameters were also measured at each of the following sampling times. In addition, the rigor index (Bito et al., 1983) was measured. On the right fillet, withdrawn from the fish analysed immediately after death, the total length and colour in three different sites (cranial epaxial, cranial ipaxial, middle caudal) were measured using a Chromameter (Minolta CR-200; CIE L*, a*, b* system). Those measurements (fillet length and colour) were repeated on the same fillets stored at 1°C in a closed aluminium box at 2, 3, 4 and 24 h after death. The data were analysed with a one way ANOVA (diet), while the data concerning the measurements obtained in the subsequent samplings on the same fillets (fillet length and colour) were analysed by analysis of variance using PROCMIXED by the SAS statistical package, considering diet, time of sampling and site of measurement (only for colour) as fixed factors and the subject as random factor. The total replacement of fish meal by a mixture of plant protein sources over a long duration led to significant differences in trout growth. Preliminary statistical analyses showed that body weight did not have a strong influence on the behaviour of the physical and biochemical parameters considered. Evolution of the rigor index showed different behaviour between the two groups, with the trout fed the PP diet already showing a value of nearly 100% at 2 h after death, while the FM trout attained the whole rigor 3 h later (Table I) . The observed behaviour is consistent with the AEC value (significantly lower in PP trout at 2 h after death) and with the evolution of fillet shrinkage. In fact, although the values obtained at 24 h for the fillet shrinkage were similar for the two groups, this parameter showed a more gradual behaviour in trout fed the FM diet. No differences were observed for the pH values, either in the eye liquor or in the epaxial muscle, or for the dielectric properties. As regards the colour, the site of measurement strongly ( p<0.01) affected all parameters. The lightness was modified during the 24 h, with higher values at the start and at the end of the evaluation period, but did not reveal any differences due to diet. The yellowness (b*) was lower for FM than PP trout fillets, with more significant differences at 4 h after death (12.56 vs 14.56), justifying the differences recorded for the Hue values. The protein source influenced the rigor index, AEC and fillet shrinkage behaviour, probably in relation to a difference in the efficiency of cell energy recovery in the hours immediately after death, induced by the diet composition. The differences observed in fillet colour seem to demonstrate the presence of an interaction between the protein source and the pigment used in the diet formulation. The results obtained suggest a possible effect of the quality of the dietary protein source on the biochemical processes during early post mortem stages, with potential consequences on the shelf-life and quality characteristics of the final product. Conjugated linoleic acids (CLA) is the collective name for a group of geometric and positional isomers of linoleic acid (C18:2) in which the double bonds are separated by a single carbon-carbon bond instead of a methylene group. CLA are known to have many favourable biological effects: anti-carcinogenic properties , antioxidant properties (Ip et al., 1991) , as well as a nutrient partitioning effect (Mersmann, 2001) . They have been shown to protect against atherosclerosis in rabbits (Lee et al., 1994) due to the reduction of plasma HDL cholesterol and triglycerides (Corino et al., 2002a) . Moreover, CLA isomers have also been shown to modulate the immune system (Sugano et al., 1998; Bassaganya-Riera et al., 2001; Corino et al., 2002b) . The aim of the present study was to determine the effects of dietary CLA supplementation in sows during late gestation and lactation on colostrum immunoglobulins of classes G, A and M (IgG, IgA, IgM) and on the piglet serum immunoglobulin title. A total of 21 gestating sows were allotted according to weight and parity to one of three dietary treatments: 0% CLA (C) or 0.5% CLA from seven days before parturition until seven days postpartum (T1) and 0.5% CLA from seven days before parturition until weaning (T2). CLA powder contains 50% CLA isomers (CLA ONE powder, PharmaNutrients, Inc., 918 Sherwood Drive Lake Bluff, IL 60044, USA). During parturition, samples of colostrum were collected from each sow and were 242 immediately frozen at −20°C pending analysis of immunoglobulin concentrations. Blood samples were taken randomly from two piglets per sow at 21 d postpartum and 15 d post weaning for the determination of the immunoglobulin titre. The immunoglobulin concentrations (IgG, IgA, IgM) were determined based on the quantification of specific antigens according to the radial immunodiffusion method of Mancini et al. (1965) . A commercial kit specific for swine (Bethyl Laboratories Inc., Montgomery, TX, USA) was used for the assay. The results on the colostrum immunoglobulin concentration were analyzed by ANOVA using the SPSS (2001) . Differences between means were tested using the Student-Newman-Keuls test. Data on serum parameters were analyzed by repeated measurements (ANOVA). The results on colostrum immunoglobulin concentration are shown in Table I . Colostrum IgG concentration was significantly higher (p<0.05) for treated groups than control group (+30%). The concentration of IgA and IgM tended to be higher ( p<0.10) in the colostrum of sows given CLA. There were no differences between the T1 and T2 groups because sows were given CLA treatment at the same time, 7 d before farrowing. The serum IgG concentration in piglets at 21 and 35 d post-partum is reported in Figure 2 . Serum IgG concentration at 21 d postpartum was significantly higher (p<0.05) for treated groups (T1 and T2) than control group (C). At 35 d postpartum serum IgG was higher only for the T2 group as compared to control group. There were no significant differences in serum IgA and IgM in relation to the diet of sows or age of piglets, however the control group showed lower values than the treated groups. The effects of dietary CLA supplementation on colostrum and serum immunoglobulin concentration confirmed the immunomodulatory effects of these isomers as previously reported by Corino et al. (2002) and Bontempo et al. (2002) . In particular we observed a positive effect at 21 d postpartum for piglets born from treated sows (T1 and T2) as compared to those born from non-treated sows (C). At 15 d post weaning there was a positive effect only for piglets born from sows fed CLA until weaning (T2). The T2 groups presented higher values for all the classes of immunoglobulins as compared to the control group in agreement with the results reported by Sugano et al. (1998) . Rooke et al. (2002) suggest that IgG synthesis by piglets is positively correlated to the amount of maternal IgG absorbed and thus reinforces the importance of an adequate IgG intake from colostrum. Also Krakowski et al. (2002) reported that high immunological values of colostrum determine the immunity of piglets not only in the suckling period, but also after weaning. In conclusion, dietary CLA supplementation seems to favorably influence immune responses in piglets, suggesting an enhancement of host defenses against invading organisms. This enhancement is likely to be particularly important in young animals that have not yet developed a specific immune response repertoire. Our findings suggest the dietary CLA should be further investigated as a possible alternative to the administration of growth-promoting antibiotics. It has been demonstrated in previous reports (Potena et al., 2001a; Potena et al., 2001b; Zicarelli et al., 2001) that several subjects of ''Bufala Mediterranea Italiana'' showed a variable clot yield of between 22 and 30%, despite having the same milk chemical composition, the same diet and the same days open. Correlations between genetic variants and milk yield have not yet been observed for the buffalo species, but the possibility that one animal shows similar milk yield values in successive lactations, may be indicative of the utility of selecting for that characteristic. Currently, PKM supplies information on the quantity of mozzarella cheese produced/lactation, but this does not take into account the milk's cheese-making properties. Therefore, we thought it worthwhile to investigate the repeatability of the ''milk yield'' character and the possibility of classifying subjects into the categories of ''high'' or ''low'' milk yield. This may avoid monthly controls during lactation, because, in the case of a positive result, this character could be included in the selection programs for the species, whose milk is only used for cheese-making. During the first year of the trial, 1.5 litres of milk samples from the morning milking were collected monthly from 26 buffaloes. For 10 of these, other milk samples were also collected between 120 and 150 days open two times during the following year. The buffalo diet was kept constant throughout the trial (0.92 UFL/kg DM, 16% CP, 38% NDF, 31% NSC, 0.68% Ca and 0.38% P). Milk clot yield was evaluated by heating the milk at 42°C and adding 5 ml liquid clot (1:10000 titre) for each sample, according to the method described by Intrieri et al. (Intrieri et al., 1986) . The clot was weighed after 4 h, stored at 4°C, and weighed again 24 h from its constitution, upon whey elimination. Statistical analysis was performed by using ANOVA and analysis of the correlation coefficients (SPSS 11.0.1, 2002) . Significant values for the correlation coefficients were observed between the first and second year at 120 and 150 days for milk production (0.809; p<0.01), milk clot yield at 4 (0.765; p<0.01) and 24 (0.732; p<0.05) hours after clot production and quantity of clot for each percent point of proteins (0.672; p<0.05). If the value recorded in the first month is excluded ( Table I ) the grams of clot obtained from 1 litre of milk increased progressively during lactation (r=0.534 and r=0.527 respectively; p<0.01). Between 121 and 180 days open the values were constant and not statistically different from those registered in other periods; the only variation was represented by a slight lowering registered between 241 and 270 days open. This variation was probably due to the coincidence with the period in which buffaloes were milked once daily and were near to dry themselves. The values in clot weight, both at 4 and 24 h, and at the different days open were always significantly correlated with the average of total lactation, excluding those at the end of lactation (>270 days). The highest correlation coefficients were obtained at 115 and 150 days. The mean of these values was highly correlated with the milk clot yield of total lactation (r=0.79 and r=0.93, respectively at 4 and 24 h; p<0.000). The grams of clot from 1 litre of milk were repeatable in the same buffalo during the two successive years, although the data need to be confirmed by further experiments. The value of the correlation coefficient between the average of the clot weight at 115 and 150 days open and the average of milk clot yield of each buffalo indicates that it is possible to reduce the number of controls for identification of the individual milk clot yield from nine to two. Enterohaemorrhagic (EHEC) Escherichia coli are recognised as human pathogens responsible for haemorrhagic colitis and in some cases, particularly in children under 6 years of age, cause of complications such as haemolytic uraemic syndrome and/or thrombotic thrombocytopenic purpura. The VTEC serotype O157 is most frequently reported in many countries and is frequently isolated from cattle and sheep, although other serotypes have been isolated from human cases. Human infection can also cause non-bloody diarrhea or result in no symptoms. Incidence of infection has presented a seasonal trend in many countries, with the summer months presenting higher prevalence of infection in both animals and humans. Many cases of infection have been reported in European and non-European countries and up to now 245 cases of VTEC infection have been reported to Enter-net Italia surveillance system. The aim of this paper is to evaluate the application of an official method for detecting typical (sorbitol-negative, glucuronidase negative) viable Escherichia coli O157 in food to microbial testing of beef carcasses and faecal samples at slaughterhouse. In detail, immuno-complexes were spread over one half of the plate with a sterile swab and were diluted further by streaking with a loop on the remaining two unstreaked quadrants. All isolated colonies with typical morphology were tested for ability to produce indole (ISO 7251:1993, 5.5 and 9.2.5-8; 9.2 .6 modified about incubation temperature: 37°C instead of 45°C), fermenting cellobiose and O157 antigen using latex agglutination kit (Dry Spot E. coli O157, Oxoid, S.p.A., Garbagnate Milanese, Italia). Typical colonies (i.e. indole positive, cellobiose negative and positive to latex agglutination test) were further tested using the API 20E kit (BioMérieux Italia, Roma). The key for identification was that reported in Table 1 of Laboratory Procedure MFLP-80-2001. The results obtained from the forty-five animals show that nineteen were positive for Escherichia coli O157. Eight of these were found positive only in faecal samples, four were found positive in both carcass and faecal samples and seven were found positive only in the carcass samples. Chance of detecting suspect colonies on differential media was affected negatively in the first 30 samples (four out of fifteen faecal samples and three out of fifteen carcass samples) by outgrowth of background micro-organisms, resulting confluent growth and/or diffusion of fluorogenic pigment. The problems was solved by reducing the incubation time in pre-enrichment broth to 6-8 h and improving the method of spreading the bacteria immuno-bead complexes. Failure to recover sorbitol-negative and/or glucuronidase-negative colonies on differential agar plates may occur, because selectivity of differential agar plates is not 251 able to prevent outgrowth of confounding micro-organisms, which are common in beef carcasses and faeces. Major difficulties were found with the use of differential agar with MUG, because the water soluble pigment can diffuse rapidly in the agar. The choice of 6-8 h incubation time for pre-enrichment and careful spreading of immunobeads-bacteria complexes onto differential and selective agar, allowed very good recovery of isolated suspect colonies of Escherichia coli O157 from both beef carcasses and faeces. In conclusion the official laboratory procedures MFLP-80 (March 2001) and MFLP-90 (April 1997) adopted by Canada's Directorate of Health for Food can also be used for successful detection of typical Escherichia coli O157 from beef carcasses and faeces. The identification of food species is becoming a very important issue for the assessment of food composition, necessary for the provision of proper consumer information. The increasing demand for transparency in the food industry derives either from socio-religious or economic reasons and has provoked a strong demand for appropriate detection methods that allow identification of the different components in processed food. The conventional methodologies used for the determination of species origin in food products are mainly based on immunochemical and electrophoretic analysis of proteins and on liquid chromatography techniques. However, immunological and electrophoretic methods cannot distinguish between closely related species and are often not suitable for food products with complex matrixes. Chromatographic techniques allow detection of differences in the percentages of fatty acids but are rather laborious. Recently, DNA-based techniques such as the polymerase chain reaction (PCR) or DNA hybridisation have been applied to identify meat species. These methods are sensitive and can still be used if the meat has been cured or autoclaved. Therefore, the aim of this study was either to optimise the extraction procedure of genomic DNA or to evaluate the suitability of a PCR method employing mitochondrial primers in detecting the authenticity of processed meat products. The tests were conducted with species-specific primers; in particular a region encoding tRNA e ATPase8 genes in the mitochondrial DNA was selected for bovine (forward: 5∞-CACAATCCAGAACTGACAC-3∞; reverse: 5∞-GTAGGCTTGGGAA TAGTACGA-3∞), the gene coding for subunit 1 of NADH dehydrogenase was selected for Gallus-Gallus (forward: 5∞-CAAACATTGTGGGCCCTTTT-3∞; reverse: 5∞-AGTGGGAGGGGGACTCAAAT-3∞) and two primer pairs were selected for swine; the first nuclear pair capable of amplifying the porcine endogenous retroviruses (FORWARD: 5∞-CCCAGTACCCCTTCTAGGT-3∞; reverse 5∞-TCATCTA ATTGGAGGGTCAA) and the second amplifying a segment of the mitochondrial cytochrome b gene (forward: 5∞-AAATCTCCCCTCAATGGTAT-3∞; reverse: 5∞-GGTAGTCCTATTATCGTGGG-3∞). The specificity of the selected primers was assessed using genomic DNA obtained from cell lines derived from specific and closely related species (turkey, sheep, goat). Moreover, the DNA extraction efficiency was evaluated (QIAamp DNA stool Mini Kit Qiagen S.p.a. -MI) related to different samples of processed meat products, starting from 200 mg of each sample. In particular the following products were selected: 15 samples of mortadella (5 known as Bologna-IGP), 20 samples of gelatine (10 known as having certified bovine origin) and 25 samples of sausages (10 of certified swine origin, 5 of certified poultry origin and 5 of certified bovine origin). The extraction was measured using a spectrophotometer (260/280 nm). Specific PCR reactions were performed on the extracted DNA. The results were analysed using standard agarose gel electrophoresis with ethidium bromide stain and then photographed in the UV region (254 nm). The results were compared with those obtained using the conventional ELISA test. The results of the tests carried out to evaluate the selected primers and performed on selected cell lines have shown their suitability and the complete absence of crosscorrelations. Moreover, the efficiency of selected extraction kit was assessed since all samples showed a value of 3 mg±1 DNA/ml. The results related to the authentication of food samples by PCR technique, compared to those obtained by ELISA test are reported in Table I . The results suggest the higher overall sensitivity of the PCR method for processed meat products. The gelatines, for example, were not suitable for the ELISA test as they were highly degraded. Furthermore, the PCR method appeared to be able to detect very low levels of other species different from those expected. Among the primers used, those specific to porcine endogenous retroviruses (PERV) appeared to be unsuitable for food analysis since they are only able to detect swine origin for unprocessed meat. The results obtained showed the high suitability of PCR-based methods in detecting the authenticity of origin of processed meat products. This technique, which is able to reveal very low levels of species-contamination, appeared to be more suitable than the conventional ELISA test, which is highly affected by thermal treatments. This diagnostic approach could become a useful tool in detecting possible frauds, offering more safety and security to the final consumers. Organophosphorus pesticides are esters of phosphoric acid with different substituents; they are a heterogeneous group of compounds widely used within agriculture as insecticides for their broad spectrum of action. Because of their short persistence in the environment, these compounds have replaced the organochlorine pesticides, however their toxicity, together with the possibility of bioaccumulation along the trophic chain, may represent a risk for human health (Tiecco, 2000) . For these reasons we have developed an analytical method to identify and quantify the residues of organophosphorus pesticides in milk samples. There are already several methods available in the literature to determine the organophosphate contamination in vegetables, while very few have been developed regarding milk or other fatty matrices (Erney, 2000; Juhler, 1997) . Problems due to the complexity of these matrices, along with the presence of interferents, make the analyses of samples of animal origin difficult. For this reason extraction and purification are crucial, especially considering the difficulty in characterisation of organophosphorus pesticides having very different chemical structures and physical properties. The aim of this work was to determine eight residues in milk: acephate, chlorpyriphos, chlorpyriphos-methyl, diazinon, methamidophos, methidation, phorate, pirimiphos-methyl. We have focused on these compounds because of their wide use in agriculture and their toxicity for human health. 1 mL of acetone was added to 20 g of milk warmed to at least 20°C before the analysis. A mixture of 25 mL 1:4 acetone:acetonitrile was then added to the sample, 258 without shaking, and it was allowed to stand for 20 min. The sample was then shaken for 30 s, followed by centrifuging at about 4000 r.p.m. for 5 min. The resulting liquid was decanted into a flask, and 2 mL of water was added to the milk solid followed by mixing with a spatula. Then 20 mL of acetone:acetonitrile mixture was added, and the sample was mixed again and extracted as before. The liquid phase was added to the previous one, and the extraction was repeated a third time. The liquid phases were shaken with 50 mL of dichloromethane in a separatory funnel and the phases were allowed to separate. The extraction was repeated two further times with 50 mL of dichloromethane. After drying the extracted solution over anhydrous sodium sulphate for 30 min, the fraction was filtered using a paper filter and evaporated to dryness using a rotary evaporator. The sample was dissolved in 1 mL of acetonitrile and loaded onto a SPE C18 Monofunctional cartridge (500 mg/3 mL Phenomenex) previously conditioned with 5 mL of acetonitrile. The flask was then washed with another 2 mL of acetonitrile and 1 mL of 2-propanol. The resulting 4 mL of sample was taken to dryness and redissolved in 200 mL of acetone. 2 mL of the obtained mixture was injected into the GLC system. Gas chromatograph analyses were carried out using a HRGC Mega 2 8560 equipped with two NPD detectors (Nitrogen-Phosphorus Detector), combined with two capillary columns of different polarities: Zebron ZB-5 (30 m×0.32 mm, i.d., 0.25 mm d.f.), and Zebron ZB-50 (30 m×0.32 mm, i.d., 0.25 mm d.f.). The helium carrier gas was set at 1 cc/min. The chromatographic temperature programme was first varied from 100°C to 130°C at 5°C/min, maintaining this temperature for 10 min, followed by a second increase of 5°C/min up to 220°C maintained for 7 min, then a third increase at 6°C/min to 274°C. The identification of pesticides was obtained through comparison of retention times from the two different columns. The method was validated and tested on samples of raw milk for human consumption. Before extraction, 1 mL of parathion-ethyl (concentration: 0.2 ppm) was added as an internal standard for quantitative determination. The method was developed in a UNI EN ISO 9001:2000 certified laboratory. This method proved to be effective for the simple and rapid quantitative determination and identification of eight organophosphorus pesticide residues in milk. The method was validated considering precision, accuracy and the limits of determination and quantification. Precision and accuracy were determined for milk spiked with three different concentrations of analytes. The values of relative standard deviation (RSD %) ranged between 0.10% and 0.38% while recoveries were more than 70%, except for methamidophos with a value of about 50%. This is probably due to the pesticide volatility which may result in a loss of residue during concentration (Erney, 259 2000). For each compound the limit of quantification (LOQ) is 5 ng/g (ppb) referred to milk, while the limit of detection (LOD) is 1 ng/g (ppb), estimated as the concentration of analyte which yields a signal-to-noise (S/N) ratio of at least 3/1. These values permit quantification of the residues below the following legal limits: 10 ppb for chlorpyriphos-methyl, diazinon and methamidophos; 20 ppb for methidation and phorate; 50 ppb for pirimiphos-methyl (The European Commission, 2003) . Parathion-ethyl was chosen as the internal standard because it should not be present in food since it was removed from the list of permitted bioactive substances in 2001 (Decreto Ministeriale 6 agosto 2001); moreover, this compound is stable and gives a good response to the NDP detector (Juhler, 1997) . The method was tested on some raw milk samples. Among the samples analysed (more than 100) 30% were positive to acephate and chlorpyriphos contamination, but always below legal limits. Detection of positive samples confirms the importance of the control of OPPs contamination in milk, and illustrates the efficiency of this method. As for several other species of interest in meat production, fish can suffer endogenous contamination of muscle as a consequence of a number of different environmental or catching stressors. This can be deduced from the studies of Herborg and Villadsen (1975) on the higher bacterial load and the lower quality and shelf-life of stressed trout, as well as from the research of Panebianco et al. (1993) on the isolation of Clostridium perfringens from the muscle of Boops boops. In order to probe this issue we carried out research into the occurrence of bacteraemia in reared fish subjected to different catching methods. In previous work we showed the high incidence of bacterial isolation from muscle and blood of marinecultured fish (Dicentrarchus labrax, Sparus aurata) and isolation of the same bacterial genus from muscle and blood in 10.81% of cases (Giuffrida et al., 2003) . In this work we report the results for cultured specimens of Salmo gairdnerii. 39 specimens of Salmo gairdnerii were used for this study. Fish were caught with a small net and divided into three groups (A, B and C) on the basis of three different slaughtering methods: 11 fish (Group A) were slaughtered by resection of the medulla oblongata; 16 fish (Group B) were electrocuted (12 V per 15∞∞) and 12 fish (Group C) were slaughtered by simple anoxia. Once the fish did not show any more vital behaviour, a blood sample was collected from the heart using a sterile syringe, after disinfection of the skin, and inoculated into three different blood-culture systems: Signal Blood Culture System (SBCS, OXOID); Hémoline Performance Anaérobie (BIOMÉ RIEUX); IsolatorTM 1.5 (OXOID); and into tubes of Salt Polimixin Broth (SPB, OXOID). SBCS blood-cultures were pre-incubated at 36°C for 1 h and then at 36°C for 24 h in a thermostatic agitator. The blood-cultures showing growth were streaked onto plates of 3 different media: i) MacConkey agar (OXOID) plates which were incubated at 35°C for 24 h; ii) Pseudomonas agar base (OXOID plus Pseudomonas CFC Supplement) plates which were incubated at 30°C for 24 h; iii) Sulphite Polimixin Sulphadiazine agar (SPS, OXOID) plates which were incubated at 37°C for 48 h using anaerobic jars with a Gas Generating Kit (bioMerieux). The same media were used for the IsolatorTM 1.5 blood-culture system, while the Hémoline Performance Anaérobie blood-cultures were only streaked onto SPS agar plates, incubated as indicated above. SPB enrichments were incubated at 30°C for 24 h and streaked onto Thiosulfate Citrate Bile Salts Sucrose (TCBS) agar (OXOID) plates, then incubated at 30°C for 24 h. A muscle sample was also aseptically drawn from each fish and divided into three sub-samples which were homogenised (1:9), respectively, with Peptone Buffered Water (PBW, OXOID), Reinforced Clostridium Medium (RCM, OXOID) and Salt Polimixin Broth (SPB, OXOID). PBW enrichments were streaked onto MacConkey agar (OXOID) plates which were incubated at 35°C for 24 h and onto Pseudomonas agar base (OXOID, plus Pseudomonas CFC Supplement) plates which were incubated at 30°C for 24 h. RCM enrichments were streaked onto SPS (OXOID) plates incubated at 37°C for 48 h using anaerobic jars with a Gas Generating Kit (BIOMERIEUX). SPB enrichments were streaked onto TCBS agar (OXOID) plates, incubated at 30°C for 24 h. After incubation all morphologically distinct colonies were picked, purified by restreaking onto tubes of Trypticase Soy Agar (TSA) (Oxoid), and identified using common bacteriological techniques and the Api 20E and API 20A systems (bioMerieux). Each sample was also subjected to necroscopy and a sterile swab was rubbed over each phlogistic focus and streaked directly onto plates of TCBS agar (OXOID) and Aeromonas agar (OXOID), which were incubated at 30°C for 24 h, as well as onto plates of Chromocult Coliform agar, which were incubated at 37°C for 24 h. The identification of morphologically distinct colonies was carried out as described above. Isolations were carried out in 82.05% of blood samples and in 64.10% of muscle samples. The isolation from blood regarded the genera Erwinia (31.03%), Enterobacter (24.14%), Shigella (12.07%), Serratia (8.62%), Flavibacterium (6.90%), Hafnia (3.45%), Acinetobacter (5.17%), Pseudomonas (5.17%), Sphingobacterium (1.72%) and Aeromonas (1.72%). The isolations from muscle regarded the genera Enterobacter (48.00%), Aeromonas (28.00%); Pseudomonas (16.00%), Chryseomonas (16.00%), Sphingobacterium (16.00%), Erwinia (8.00%), Hafnia (8.00%), Flavibacterium (4.00%) and Acinetobacter (4.00%). In three cases (7.69%) the same bacterial genera were isolated from both blood and muscle; the genera were Citrobacter freundii, Pseudomonas fluorescens/putida and Enterobacter cloacae. Necroscopic examination showed three subjects with oral haemorrhagic foci from which Enterobacter agglomerans, Aeromonas hydrophila, Pseudomonas paucimobilis, Sphingobacterium multivorum and Pseudomonas cepacia were isolated. According to our results for marine cultured fish, the occurrence of bacteraemia in trout seems to be very frequent. The percentage isolation of the same bacterial genus from blood and muscle (7.69) is similar to that observed in Dicentrarchus labrax and Sparus aurata (10.81%) (Giuffrida et al., 2003) . In these cases no correlation was observed with slaughtering methods, while other stressors (fish density, physical and chemical modifications of water, feed competition, etc.) could be influential in this regard. Furthermore, it is well known that these stressors can predispose fish to several pathologies sustained by opportunistic bacteria. The isolation of Pseudomonas paucimobilis and Aeromonas hydrophila from observed oral lesions, could demonstrate the occurrence of these stressors. Foodborne infections are increasingly being recognized as being responsible for many human illnesses (Croci et al., 2001) . Previously more attention was directed towards the detection of bacteria or parasites; however, the importance of viral infections is still underestimated by routine surveillance (Sair et al., 2001) . Some types of food, such as shellfish, seem to play a major role in viral outbreaks, due to their high capacity of water filtration. Viruses are able to persist in mussels for a long period, even after their immersion in cleaned stalling water, and it has been demonstrated that they can survive decontamination processes used for bacterial agents (Richards et al., 1998) . Italian law by decree no. 530 of 12/30/1999, regarding the sanitary rules for the trade of living mussels, does not provide for control of the absence of viral particles as sanitary inspection; it is only based on the quantitative determination of fecal bacteria. Nevertheless, it has been demonstrated that there is no correlation between bacterial and viral infections. Accordingly it is very important to establish new parameters capable of determining the tolerance of shellfish towards enteroviruses. For this reason, the D.C. no. 199/313/CE of 4/29/1999 indicates different procedures for the depuration of stalling water for molluscs which can eliminate enteric viruses. The purpose of this study was to evaluate the inactivation kinetics of enteroviruses with sodium hypochlorite, peracetic acid, chlorine dioxide, and hydrogen peroxide tested on viral dilutions and then inoculated in cell cultures. The efficacy of the inactivation process was evaluated by calculating the infectious titre. Inactivation kinetics were also studied using RT-nested PCR in contaminated live shellfish. Tests were performed using poliovirus type 1 (vaccine strain), coxsackievirus B5 and echovirus 30. Viruses were propagated in the BGM (Buffalo green monkey kidney) cell line, which was amplified in MEM (Eagle's minimum essential medium Earle BSS) enriched with 10% foetal calf serum and 1% antibiotics (potassic penicillin G-1000 UI/ml, streptomycin sulphate-500 mg/ml and amphotericin B-2.5 mg/ml). The following disinfectant solutions were prepared: sodium hypochlorite (0.2[1 mg/l), chlorine dioxide (0.5[2 mg/l), hydrogen peroxide (0.1[0.5 g/l), peracetic acid (160[640 mg/l). 0.5 ml of each virus was added to 4.5 ml of disinfectants at the above concentrations. The solutions were then serially diluted by 10−1 and inoculated on a BGM monolayer which was incubated for five days. The infectious titre (TCID 50 ) was then determined using the Reed and Muench method . Experiments in the aquarium were conducted using poliovirus 1 in the presence of hydrogen peroxide which is known for its low toxicity towards shellfish. Forty molluscs were contaminated with the virus (final titre: 104.50 TCID 50 ) for 24 h and were then transferred to a cleaned water pool containing H 2 O 2 (0.375 g/l). Samples were then taken at pre-arranged time intervals and the residual titre was calculated by infecting the BGM cells and performing RT-nested PCR on shellfish tissue. The tests carried out using virus and sodium hypochlorite at different concentrations showed a clearly reduced infectious titre. In particular, concentrations of 0.6, 0.8 and 1 mg/l reduced respectively to 2, 3 and 4 Log. Inactivation kinetics of the virus with chlorine dioxide allowed detection of the fall in the titre from 106.50 after 15 min to 104.50 after 60 min at 1 mg/l. Peracetic acid reduced the infectious titre to 105.24 after 10 min and 104.24 after 60 min. Hydrogen peroxide had good inactivation efficacy under neutral pH conditions, while temperature did not influence its activity. The use of sodium hypochlorite allowed detection of a decrease in virus viability 30 and 60 min after the first contact. The same behaviour was observed with chlorine dioxide and peracetic acid over a range of concentrations from 0.6 to 2.0 mg/l. Hydrogen peroxide (0.5 and 1 g/l) reduced the infectious titre by 3-4 logarithmic units after 6 h. RT-nested PCR performed on water artificially contaminated with poliovirus type 1 confirmed the presence of virus in shellfish tissue 24 h after the addition of hydrogen peroxide and its complete absence after 48 h. In contrast, the inoculation of mussel homogenate on BGM after 24 h of disinfection did not allow detection of any cytopathic effect. The results showed that the three types of viruses had similar kinetics when treated with sodium hypochlorite while, in the presence of chlorine dioxide, poliovirus 1 was more resistant than coxsackievirus B5 and echovirus 30. This difference was also confirmed on treatment with peracetic acid, while hydrogen peroxide, considering its low concentrations, was the most effective inactivator. The results showed that enteroviruses are resistant towards the most common disinfectants used in depuration systems. Following an initial decrease in the first titre count, viruses persist at a low concentration for a long period which is possibly affected by other factors such as pH and temperature. Positive samples found using RT-nested PCR could be due to the higher sensitivity of this method which can detect as few as 10 viral particles. Nevertheless, it has not yet been established whether this positivity can be attributed to viral infecting particles or to nucleic acid residues. Negativity revealed through PCR is very important for evaluation of depuration processes, but this could be affected by the large number of inhibitors that are potentially present within a product. For this reason it is necessary to test the inactivation kinetics of various viral serotypes and to prepare new experimental techniques that can more efficiently guarantee the complete depuration of these products. Meat preparations are exposed to microbiological risk due to their chemical-physical characteristics and the processing steps employed (Pizzin et al., 1998) . Pathogens such as Staphylococcus aureus, Escherichia coli and Salmonella spp. are frequently found in these products (Pezzotti et al., 2001) . The aim of this study is to assess the presence of these pathogens in meat preparations exclusively or mainly made with ground beef meat (hamburgers and meat patties) and raw pork sausages taken from the local market (Umbria region, Italy). The study also aims to evaluate the antibiotic-resistance of the eventually isolated Staphylococcus spp., Staphylococcus aureus, Enterococcus spp. and Salmonella spp. bacteria as well as to contribute to evaluation of the possible role of food of animal origin in the trasmission of antibiotic-resistant bacteria (WHO, 1997). Samples from 10 batches of beef hamburgers and 19 batches of meat patties (beef patties and beef/pork patties) were analysed. All of these products were produced at the meat department of various supermarkets. Samples from 25 batches of raw pork sausages were also analysed: four of these batches were produced in a semi-industrial plant and the remaining 21 in the meat department of different supermarkets. All beef and pork meat preparations sampled were taken directly from the display cases and transported in a refrigerated container to the laboratory within two hours. The samples were analysed using the following methods: Pathogen isolation: a) Staphylococcus spp.: spread onto Baird-Parker+EGG (Oxoid) and incubated at 37°C for 48 h; b) Staphylococcus aureus: spread onto Baird-Parker+RPF (Oxoid) and incubated at 37°C for 48 h; c) Escherichia coli: spread onto Coli ID (bioMèrieux) and incubated at 37°C for 48 h; d) Salmonella spp.: 25 g of each sample was pre-enriched in 225 ml of buffered peptone water and incubated for 18 h at 37°C. Pre-enrichment broth culture (0.1 ml) was then inoculated into 10 ml of Rappaport-Vassiliadis broth and incubated at 42°C for 24 h. Having completed the enrichment process, a loop (0.1 ml) of broth culture was then spread onto an XLD plate (Oxoid) and incubated at 37°C for 24 h. The suspect Salmonella spp. colonies were further tested (Kligler and urease) and serologically typed. Antibiotic resistance evaluation: a) Enterococcus spp. vancomycin-resistant: 25 g of sample was put into 225 ml of VRE Broth Base+Meropenem (Oxoid) and incubated at 37°C for 24 h. A loop (0.1 ml) of broth culture was then inoculated into VRE Agar Base+Meropenem+Vancomycin (Oxoid) and incubated at 37°C for 24 h; b) 93 of the detected strains of Staphylococcus spp. and 18 strains of Staphylococcus aureus were randomly selected and tested for meticillin and vancomycin resistance; c) isolated strains of Salmonella spp. were tested against the following antibiotics by using the ''disk diffusion method'' according to Kirby- tables (1997a, 1997b) . The results of the analyses performed for evaluation of the presence of pathogens are shown in Table I . Staphylococcus aureus was isolated in 21.1% of the hamburgers and meat patties (all samples containing less than 5×102 cfu/g). Escherichia coli was detected in 88.9% of the hamburgers and meat patties, with a concentration ranging from 5×102 to 5×103 cfu/g in 14.8% of the samples and higher than 5×103 cfu/g in 3.7% of the samples. Salmonella spp. was not detected in any of the above samples. Staphylococcus aureus was isolated in 47.1% of the sausages sampled (all samples containing less than 5×103 cfu/g and 11.77% containing more than 5×102 cfu/g). Escherichia coli were detected in 78.3% of sausages, with a concentration ranging from 5×102 to 5×103 cfu/g in 16% of samples and no samples had a concentration above 5×103 cfu/g. Salmonella typhimurium and Salmonella rissen were detected in two sausages samples, respectively. Regarding the antibiotic resistance, no Enterococcus spp. vancomycin resistant strains were detected and all the Staphylococcus aureus and Staphylococcus spp. strains tested proved non-resistant to vancomycin and meticillin. Salmonella typhimurium was resistant to (AMP), (STR), (SUL), and (TET), and Salmonella rissen to (STR), (SUL), (TET) and (TRI). The microbiological characteristics of the products considered in this study were similar to those reported by other authors (Del Gaudio et al., 2001 , Pezzotti et al., 2001 . The detection of Salmonella spp. in two of the examined pork sausages cannot be considered an unusual event, in agreement with other author (Cantoni et al., 1999; Mioni et al., 2000; Pezzotti et al., 2001) especially regarding Salmonella typhimurium (Bonardi et al., 2002) , in that this is one of the serotypes most frequently detected in pig carcasses (Swanenburg et al., 2001 ). An interesting result of the study is the isolation of two Salmonella spp. strains that were resistant to four antibiotics. The risk that these micro-organisms can cause salmonellosis that cannot be treated with commonly used antibiotics is of great concern. Although the data are preliminary, it is reassuring to know that 100% of Staphylococcus aureus and Staphylococcus spp. proved non-resistant to vancomycin and meticillin and that no Enterococcus spp. vancomycin resistant were detected. The isolated strains of Escherichia coli were also tested for antibiotic resistance. However, due to the complexity of the results obtained, the data will be published at a later time. Finally, with regard to the consequence for the consumer, it must be stressed that the microbiological risk represented by meat preparations can only be dramatically reduced when these products are properly cooked. Neoplasms of the urinary bladder have a low incidence in domestic animals. In dogs they account for less than 1% of all reported neoplasms (Confer and Panciera, 2001) . Cattle grazing on bracken fern-infested lands are at increased risk of suffering from tumours of the urinary bladder (Galati, 1997) . It has been suggested that long term urine storage plays an important role in urothelial carcinogenesis since it could enhance the action of carcinogenic agents as it increases urine contact time with tissue (Caywood et al., 1980) . It has been suggested that primary epithelial tumours of the urinary bladder occur more frequently in female than male dogs with an average age of 10 years (Crow, 1985) . About 76% of all urinary bladder tumours develop from the urothelium, and most of them are malignant (Caywood et al., 1980; Strafuss et al., 1975) . The recent WHO Histological classification of urinary bladder tumours in humans (Mostofi and Davis, 1999) describes some new variants of urothelial carcinoma that have not yet been reported in animals. In an ongoing research program about the incidence of urinary bladder tumours at the Department of Veterinary Pathology at Naples Veterinary School, two new cases were observed: a lipid cell variant of urothelial carcinoma in a cow and a carcinosarcoma in a dog. Samples of the neoplastic urothelium were collected from bovine and canine urinary bladders. The samples were fixed in 10% buffered formalin and embedded in paraffin; 274 5 micron-thick sections were stained with haematoxylin-eosin (HE), periodic acid-Schiff (PAS) and Mayer mucicarmine stainings for neutral and acidic mucopolysaccharides. In addition, 8-10 mm thickened frozen sections were stained with saturated Sudan black B in 70% isopropyl alcohol and Oil Red O to detect intracytoplasmic lipids. The lipid cell variant of urothelial carcinoma of the bovine urinary bladder was histologically characterized by clusters of urothelial cells exhibiting the appearance of signet-ring cells. Sudan Black B and Oil Red O stainings carried out on frozen sections showed intracytoplasmic lipids in many of these signet-ring cells. Signet-ring cells were also seen in the neoplastic proliferations. The carcinosarcoma was characterized by two different components: multiple and diffuse nests of urothelial neoplastic cells were scattered over sarcomatous lesions. The lipid cell variant of urothelial carcinoma in which clusters of lipocyte-like cells were scattered over the neoplastic urothelium suggests that they could be of epithelial origin. It is well known that the potentiality of the malignant urothelium is reflected in the wide range of variant morphologies of carcinomatous lesions. The lipid cell variant, which is reported to be a rare variant of human urothelial carcinoma, has now been reported in veterinary medicine for a cow. Vesical carcinosarcoma is also very unusual for the dog. The authors' previous investigations showed that daily administration of 300 ppb of aflatoxin to sows from day 42-45 of gestation until weaning time, produced indirect intoxication of their piglets, leading to the exhibition of mild clinical signs of disease and histological patterns of altered cell-mediated immunity . Harvey et al. (1995) reported deficient levels of retinol and tocopherol in the tissue and serum of pigs affected by aflatoxicosis; they suggested that greater amounts of vitamins E and A are required by the body during the course of this disease and that both vitamins play a protective and detoxifying role in the presence of aflatoxins. Based on these considerations, the aim of our research was to assess wherether vitamin A and E supplements in AF contaminated feedstuff given to sows during gestation and lactation had a similar protective and detoxifying effect on their offspring. To verify this, the young animals' immune status was examined and signs of any adverse effects were sought by means of haematological and morphostructural examinations. Six polyhybrid [Duroc×(LW×L)], clinically healthy, pluriparous sows that had been artificially inseminated with semen from the same boar were used in the study. Two of them were kept as controls and the other four (AF+VitA/E) were kept under conditions of chronic aflatoxicosis by adding aflatoxins (B1,B2,G1,G2) to their daily rations from day 42/45 of gestation until day 23 of lactation. All the animals were fed the same basic feedstuff that had been previously tested for mycotoxins. Aflatoxin intoxication was achieved by cultures of Aspergillus flavus var. flavus and Aspergillus flavus var. parasiticus seeded onto maize and rice grains. The intoxicated sows were supplemented with 60 UI of vitamin E and 11.000 UI of vitamin A per kg of feed. Before euthanizing the animals, peripheral blood samples were collected from 4 piglets/sow on day 23 postpartum. The immunophenotypical evaluation of the main T/B lymphocyte subpopulations (CD4, CD8, CD21, CDc/d) was performed by flowcytometry (De Angelis et al., 2001) . Blood counts and serum biochemistry were also measured following traditional methods . Portions of the piglet's liver, kidneys, spleen, thymus and lymph-nodes were fixed in formalin, processed and stained following the routine procedure (Luppi et al., 2002) for post mortem examination. Samples of milk were taken from the sows on day 3 and 23 post-partum. The residual aflatoxins were measured by HPLC (Silvotti et al., 1997) . The piglets born from the two groups of sows considered (control and AF+VitA/E) were alive and viable, exhibiting clinical conditions throughout the trial. Post mortem examination revealed no patient alterations. Histological examinations showed structural changes of the liver, thymus, spleen and lymph-nodes of the AF+VitA/E sows. Signs of chronic disease were found in the liver with endocytoplasmic vacuolization and a greater amount of connective epimentation surrounding the lobules. Moderate lymphocyte depletion was observed in the splenic cortex and in the lymph-nodes. The malpighian splenic follicles showed signs of poor activations. Apart from slight individual differences, the blood counts were not particularly different between the two groups of piglets. Only MetHb was significantly lower in the AF+VitA/E group than in the controls. As shown in Table I , the CD4 and CD8 subpopulations were significantly reduced and so was the ratio between the intoxicated piglets and the controls. The residual AF measurements by HPLC showed traces of AF-M1 and 73 ng/l of AF-M2 on day 3 of lactation and 77 ng/l of AF-M1 and 22 ng/l of AF-M2 on day 23 of lactation. On the whole, the results of the study: 1) confirm the findings of our previous investigations that the administration of a pool of 300 ppb of AF to sows during pregnancy and lactation induces toxic effects on their piglets ; 2) show that the factor causing intoxication of the piglets was the great amount of AF-M1 and AF-M2 they had received through their mothers' milk during their 23 days of life; 3) underscore the poor protective and detoxifying effects of vitamins E and A given to the sows as supplements in their daily rations contaminated by AF. The supplements did not prevent or restrict the morphostructural damage observed in the liver, thymus and lymph-nodes of the piglets that had been indirectly exposed to AF nor did they limit the abnormal cell-mediated immuno response. This may engender a greater susceptibility to disease in the animals. The serum prevalence of Coxiella burnetii found in herds tested in the South of Italy (Landolfi et al., 2000; Capuano et al., 2001; Capuano et al., 2002) encouraged us to improve the performance of the PCR test to detect the C. burnetii pathogen in milk. Setting cow milk samples: one cow milk sample of each group was contaminated in activated cells infected with Coxiella burnetii (Dade Behring, Marburg GmbH, Germany). Each sample was serially diluted to ascertain the reliability of PCR, combined with other different techniques. After contamination, group I specimens were packed by centrifugation at room temperature at 2000 G for 20 min. All the pellets were washed three times with PBS (phosphate-buffered saline, pH 7.4) and resuspended in PBS. To detect DNA C. burnetii from group II specimens, the positive serum for Q fever (Dade Behring, Marburg GmbH, Germany) was diluted 1:1000 and added to artificially infected milk specimens. After one hour of incubation at room temperature, all samples were centrifuged at 10,000 G for 10 min, washed three times and resuspended in PBS. Immunomagnetic beads (IMB) covered with sheep antibody against rabbit IgG were added to the immunoprecipitates (Dynabeads M280) at the final concentration of 1.2×108 beads/ml. Lastly, the specimens were determined by rotation for two hours at room temperature (Muramatsu et al., 1996) . The IMB belonging to group II samples were caught by trapping them against the surface of the Eppendorff like tube, using a magnetic particle concentrator (MCP-M, 280 DYNAL A.S.): then after removing the surplus fluid, all IMB were washed three times with PBS and lastly, resuspended in PBS. Group III specimens were prepared for group II and then IMB was added and the samples were rotated for 18 h at 4°C to promote interaction between all the immunoprecipitates and the antibodies covering the IMB. Before addition at the usual concentration to the group IV samples, all IMB were rotated for 18 h at 4°C after addition of the positive serum for C. burnetii. After the addition of IMB, the group IV specimens were rotated for two hours at room temperature and they were then determined as for those of III group to collect all the IMB. All the precipitates, resuspended in PBS, were boiled for 10 min and one portion of each sample was used for PCR assay. PCR was performed according to the method designed for the amplification of htp-AB gene and one repeated region of the C. burnetii genome by Berri et al. (2000) using Trans-1 and Trans-2 primers (Fournier et al., 1998) . We did not observe any non-specific amplification fragments in each group, although the direct IMB (group IV - Fig. 3 ) was the most reliable and rapid for detection of the pathogen C. burnetii as compared with classical and indirect (group II - Fig. 1 ) or modified IMS (group III - Fig. 2 ) methods. The PCR test is very useful for the detection of C. burnetii in milk, but it is not always a diagnostically sensitive and reliable method, especially in milk collections, due to the presence of some substances that inhibit DNA amplification. The right choice of technique for sample preparation is fundamental for the improvement of detection PCR. This study is so important to stress that both indirect and modified IMS and direct IMS are the only diagnostic methods able to detect C. burnetii in fully diluted samples (over 10−11 dilutions) and this may be because of the extended interaction time between IMB and the positive serum for Q fever in these techniques allows determination of the bacterial cells contaminating all the samples. Thus, direct IMS is really the most reliable method to detect C. burnetii by PCR, which is why only five hours are required to observe amplification. Viral Encephalopathy and Retinopathy (VER), otherwise known as Viral Nervous Necrosis (VNN), is responsible for frequent outbreaks with high mortality in a wide variety of marine fish species all over the world. The disease mainly affects the larval and juvenile stages which show neurological symptoms due to the typical vacuolating lesions in the brain, spinal cord and retina (Munday et al., 2002) . The disease is caused by a virus of the Betanodavirus genus; seven viral species have been classified in this genus (van Regenmortel et al., 2000) . This classification is based on the fish species from which the virus was isolated, but recently the isolation of the same virus from different host species has become very frequent (Ciulli et al., 2002; Castric et al., 2001) . Cross-infection tests highlighted that the capacity of the virus to infect different species is mainly due to the environmental temperature at which the fish are farmed, thus at which the infection occurs, more than the host species itself . Using capsid protein gene analysis it was possible to classify four genotypes of Betanodavirus: Striped Jack Nervous Necrosis Virus (SJNNV), Tigger Puffer Nervous Necrosis Virus (TPNNV), Barfin Flounder Nervous Necrosis Virus (BFNNV) and Redspotted Grouper Nervous Necrosis Virus (RGNNV; Nishizawa et al., 1997) . Iwamoto et al. (1999) demonstrated the different infection temperature ranges for the four genotypes. Both the infection and disease seem to be very temperature-dependent, in fact most of the outbreaks happen in the summer season. This preliminary study elucidates the in vitro effects of temperature on Betanodavirus infection in the SSN-1 cell line. NNV-infected SSN-1 cells incubated at 10-30°C after viral adsorption were observed for cytopathic effect and viral growth in the culture media supernatant and cell pellets were evaluated by titration. The nodavirus strain used in the experiments was isolated from symptomatic juvenile white seabass (Atractoscion nobilis) cultured in southern California. SSN-1 cells were grown in 12.5 cm2 flasks until the monolayers were confluent. The growth medium was removed and cells were inoculated with 250 ml of viral suspension (106,5 TCID 50 /0.05 ml). After 1 h adsorption at room temperature the cells were washed twice with L-15 medium, then 3 ml of L15 medium containing 2% Foetal Bovine Serum (FBS) were added. Double flasks were incubated at 10°C, 15°C, 20°C, 25°C and 30°C and daily observed under a microscope to observe cytopathic effects. At 0, 12, 24, 48, 72, 96, 120, 168, 240 and 360 h post-infection both the free virus in the growth medium and the virus associated with the cells was titrated. Supernatants and cells were collected by scraping the bottom of the flasks. Cells were separated by centrifuging at 1800 g for 15 min and resuspended in 3 ml of L15 medium containing 2% FBS. Cells were homogenised in order to lyse them using syringes with a 25G needle. The titrations were performed in 96-well plates and TCID 50 /0.05 ml were calculated according to . A cytopathic effect was observed at 25°C two days post-infection and at 20°C three days post-infection. The observed cytopathic effect consisted of the typical cytoplasmatic vacuoles and it led to complete disintegration of the monolayer three days post-infection at 25°C and four days post-infection at 20°C. At 15°C we observed some alteration of the monolayer without the typical cytoplasmatic vacuoles. At 10°C we observed the death of all cells ten days post-infection. At 30°C, although in the first instance the monolayer was altered we didn't observe any relevant changes in the cells, even if they looked old and suffering from the high temperature. The viral titre of cell samples incubated at 20°C and 25°C increased from 104,376 TCID 50 /0.05 ml to 106,260 TCID 50 /0.05 ml and to 106,365 TCID 50 /0.05, respectively and we observed a very similar increase in growth medium samples incubated at the same temperature. The viral titre remained constant for all days tested for samples incubated at 10°C. At 15°C we observed a small increase in the viral titre >105,3 TCID 50 /0.05 ml only for cell samples seven days post-infection. At 30°C the viral titre decreased progressively during the test days, particularly in the supernatant where we could not detect the virus after 96 h (Fig. 1) . The viral titre remained constant for all days tested for samples incubated at 10°C, and this showed that no viral growth occurs at this temperature. 10°C was an unsuitable temperature for growth of the SSN-1 cell line, so cell death interfered with possible viral growth and we could not obtain any information about viral growth at this temperature. The high viral titre observed at the end of the experiment is probably due to the high resistance of the virus at this temperature. 15°C seemed to permit only partial virus replication in the cell, only a small amount of virus is produced and no typical vacuolisation can be observed in the cell monolayer. We suppose that even 15°C was an unsuitable temperature for growth of the SSN-1 cell line and this could interfere with viral growth. The optimum temperature for growth is thus 25°C even if the virus also grows and gives a cytopathic effect at 20°C. 30°C does not seem to permit viral replication and causes a rapid viral inactivation. We also observed that the virus present inside the cell is more resistant to heat inactivation as compared to the virus free in the growth medium. The current shortage of human allografts for transplantation may be overcome by using animal organs and tissue. Among the various donor species, pigs seem to be the best choice because they are available in a large number, can be housed in controlled specific pathogen free (SPF) environments and have anatomic and physiological features similar to humans. Despite these favourable conditions, xenotransplantation has two main problems. The first is the immunological incompatibility between pigs and humans which results in hyperacute rejection, and the second is the potential transmission of pathogens, especially endogenous retroviruses. The development of transgenic pigs with no or low expression of Gal a(1-3)Gal antigen and with human cell surface complement regulators may overcome hyperacute rejection (Hammer et al., 1985) . However, the potential transmission of pig endogenous retroviruses from the xenograft cannot be eliminated by modern gene knock-out technology and is still a potential risk. Therefore, the purpose of this study is to find some means of inhibiting PERV replication and production from two pig cell lines following treatment with some antiretroviral drugs commonly used in human medicine, as well as evaluating the potential tumorigenicity of these viruses for nude mice. The investigations were carried out on two cell lines derived from newborn pig kidneys (NSK; newborn swine kidney) and trachea (NPTr, newborn pig trachea), 288 respctively (Ferrari et al., 2003) both infected and able to produce A and B endogenous retrovirus classes. This infection was also transmitted, in vitro, to human cells. The following drugs were selected: nelfinavir (a protease inhibitor) (NFV, 0.030 mM), nevirapine (a non-nucleoside reverse transcriptase inhibitor) (NVP, 0.075 mM) and the four nucleoside analogs: stavudine (d4T, 0.042 mM), dideoxyinosine (ddI, 0.042 mM), lamivudine (3TC, 0.087 mM) and zidovudine (AZT, 5 mM). The NSK and NPTr cell lines were cultivated for eight serial passages in culture medium containing a mixture of the above compounds. Viral production was evaluated in the supernatant medium of the last serial passage. The control used was culture medium from NSK and NPTr cells cultivated in absence of the drugs. Detection of the viral sequences in cell DNA and virus expression in the supernatant culture medium were evaluated using PCR and RT-PCR, respectively. The analyses were conducted employing primers specific for the conserved PERV pol region (Paradis et al., 1999) This was carried out according to Barre-Snoussi et al., 1983 . Briefly, the supernatants from NSK and NPTr cells were centrifuged at 200 g and the supernatants centrifuged again at 250,000 g. The viral pellets were resuspended in 5 ml of 2.5% Triton-X100 in order to disrupt viruses and were added to 50 ml of RT-cocktail solution (100 mM Tris HCl pH 7.9, 150 mM KCl, 1.2 mM MnCl 2 , 4 mM DTT, 0.25 U/ml poly (rA) p(dT) 12-18 , 60 mCi/ml 3H TTP). After incubation at 37°C for 60∞ a 120 mM NaPP in 60% trichloracetic acid (TCA) was added for 15∞ at 4°C in order to induce precipitation. Samples were then transferred to a cellulose nitrate filter disc (0.45 mm) previously embedded with TCA (5%), rinsed, dried and re-suspended in scintillation fluid. The quantity of incorporated radioactivity was measured in a scintillation counter. T umorigenicity NSK and NPTr filtered supernatants (0.45 mm) were centrifuged as above; the viral pellets were diluted in 400 ml of culture medium and inoculated by the subcutaneous 289 route to four nude mice (two for each supernatant type) at volumes of 200 ml for each animal (100 ml in each side). The mice were observed for 60 days, after which time they were euthanized. Tissue samples were collected from the inoculation areas at necropsy, as well as from organs and tissue of the abdominal, thoracic and neurocranic cavities. All samples were analysed using PCR for proviral integration of PERV. Furthermore, 5 mm paraffin inclusion sections were fixed in buffered formalin (pH 7.4) and then stained with the usual methods of histology and immunohistochemistry. Treatment with the selected anti-retroviral drugs did not completely inhibit virus replication. In particular, the RT analysis did not show a significant difference in virus production between the NSK cells exposed to anti-retroviral treatment and the control group (reduction in virus production of 3 fold). In contrast, virus replication appeared to be considerably lower (20 times) in the NPTr treated cells than in the un-exposed controls. Furthermore, the RT-PCR showed the persistence of viral particles in the supernatants of all treated and non-treated cells, and proviral sequences were also detected in the cell DNA by PCR. The inoculation of retroviruses in nude mice did not cause tumor growth and no malignant cells were observed in the organs or tissues of the injected animals on histological examination. Similarly, no viral sequences were detected in the tissues tested. It has been shown that pig endogenous retroviruses can be transmitted to other celltypes (Soncini et al., 2001) . However, their expression by human cells in vitro can be inhibited by treatment with drugs which are commonly used for retrovirus infection treatment in humans (Powell et al., 2000; Qari et al., 2001) . In contrast, the results of this study showed that the expression of endogenous retroviruses from pig cells was reduced, but not completely inhibited, by antiretroviral treatment, suggesting some receptorial differences between the two species. On the other hand, the absence of tumorigenicity in the nude mice injected with retrovirus particles indicated that the 290 malignant features previously observed following inoculation either with NSK or NPTR cells were probably due to the cells themselves. In conclusion, while the hyperacute rejection which may follow organ transplantation may be overcome by the use of transgenic pigs expressing human cell surface complement regulators, in contrast, retroviruses may incorporate the factors which regulate complement activation during their budding throughout the cell membrane with consequent prevention of virolysis induced by the natural antibodies in human serum. As a consequence, the potential transmission of pig endogenous retroviruses during xenotransplantation may really represent a risk for public health. Contagious ecthyma is an infectious disease that commonly affects sheep, goats and occasionally humans. The disease is serious enough worldwide to be a cause of economic loss in sheep and goats, although this is poorly quantified (Haigh and Mercer, 1998) . The causative agent of the disease is the orf virus, an epitheliotropic double-stranded DNA poxvirus of the genus Parapoxvirus. The importance of orf to immunologists and virologists lies in the fact that this virus can repeatedly infect sheep and goats in spite of the presence of a vigorous host immune and inflammatory response to infection. As a result of the immunomodulating activity of the virus no effective vaccine has been developed yet. Recent progress in analysis of the orf virus genome, the identification of its genes, the construction of deletion mutants as well as the generation of recombinants is leading to the selection of viral components that could mediate immuno-stimulating activity. In this study we experimentally tried to assess the immunogenicity of the F1L protein in order to use it as a subunit vaccine. The full-length F1L gene of an Italian orf virus strain isolated from chamois (IT-TO) was amplified by PCR and cloned into the pCR T7/CT-TOPO vector fused with a six-histidine tag at the C terminus. The recombinant plasmid was propagated in E. coli TOP10F' grown in a selective medium. The expression was performed in 50 ml of E. coli BL21 transformed with 10 ng of the purified recombinant plasmid. Cells were grown overnight at 37°C with shaking and were then induced with 1 mM IPTG for 3 h at 37°C. The induced culture was centrifuged at maximum speed and the resulting pellet was analysed to detect expression of the recombinant fusion protein. Western blot analysis was carried out using antibody against 6xHis epitope, polyclonal antibody from convalescent sera and monoclonal antibody 4D9 which binds to the viral neutralising epitope (Czerny et al., 1997) . The full length F1L protein was purified using a metal-chelate affinity chromatography under denaturing conditions. The eluates were analysed using pre-cast polyacrylamide gels and tested by Dot blot using monoclonal antibody 4D9. Two rabbits were inoculated subcutaneously with 13.6 mg of the purified recombinant full-length F1L protein in Gerbu's adjuvant and were boosted 3 and 5 weeks after the first inoculation. The third inoculation was performed intravenously. The animals were bled at various intervals to check their antibody response against the subunit antigen by Western immunoblot analysis. Virus neutralization assays were performed by adding two-fold dilutions of preimmune and postimmune sera mixed with 103TCID 50 of wild type orf virus to TFO (ovine fetal testicular cells) cultured in Eagle's medium. The antibody titre was determined as the reciprocal of the serum dilution that neutralized more than 50% of CPE of the respective virus. Western immunoblot analysis of induced culture pellets showed a band in the expected size range for the recombinant protein (42-46 kDa), moreover the recombinant peptide was detected using the 4D9 monoclonal antibody. Eluates obtained from protein purification and analysed by polyacrylamide gel electrophoresis showed a band in the expected size range for the recombinant protein which reacted against the monoclonal antibody 4D9 in a Dot blot. Western blot assay using sera against the full-length recombinant F1L protein showed that the two rabbit seroconverted and the immunodominant protein of the IT-TO strain was recognized at a serum dilution of 1/100 and 1/1000 after the first inoculation and at 1/400 and 1/800 after the third immunization. Virus neutralization assays showed that both animals developed neutralizing antibodies against the viral strain IT-TO and the titers were 1:100 and 1:1000 respectively. The results indicate that the method of choice for the production of the recombinant F1L protein is suitable for evaluation of the immunogenicity of the structural peptide. In fact, immunodetection of the recombinant peptide performed using pooled polyclonal sera from convalescent sheep and the 4D9 monoclonal antibody show that the neutralising epitope is conserved even after expression in BL21 cells. These results demonstrate that rabbit immunization with the full-length F1L recombinant protein was able to stimulate the production of neutralizing antibody and suggest that the F1L protein is important for infection of the orf virus in cell Respiratory infections are widespread in the cat population and they are often associated with several pathogens such as FHV-1, FCV and Chlamydophila spp. (Gaskell and Dawson, 1998) . The definitive diagnosis of feline herpesvirosis and feline chlamydophilosis requires the isolation of their aetiological agents, since serology alone does not allow us to distinguish infected animals from vaccinated ones. FHV-1 isolation using cell cultures is easily performed, even if some cats can show negative results because of (i) inadequate amount of virus present in the ocular and/or pharyngeal swabs, (ii) presence of antibodies in the extra-cellular liquids able to inhibit viral replication in vitro, (iii) contemporaneous presence in the swabs of FCV that could disguise the cytopathic effect induced by FHV-1. As regards Chlamydophila felis these problems are increased by the objective difficulty of the isolation that can be performed through inoculation of embryonate eggs of chicken. To remove these disadvantages it is useful to utilize the techniques of molecular biology, such as the polymerase chain reaction, which is able to detect specific sequences of DNA. Furthermore, it is possible to proceed from the amplicons to the identification of the chlamydofiles after treatment with restriction enzymes. The aims of the current research are (i) to report the results of a study performed by duplex-PCR specific for FHV-1 and Chlamydophila spp. in a population of 40 cats with U.R.T.D.-related symptoms in order to appraise the prevalence of two of the most common agents associated with respiratory infections, and (ii) to identify Chlamydophila felis using restriction enzymes. 296 From October 2002 to January 2003 forty mucosal samples, each composed of a pharyngeal and conjunctival swab, were collected from cats with respiratory symptomatology related to U.R.T.D. The samples were collected from private and public Italian veterinary clinics in the Isernia, Ascoli Piceno and Teramo areas. At the time of collections an anamnestic card was filled in with information about the type of swabs collected, the clinical signs and the possible pharmacological or vaccinal treatments for each cat examined. It was not possible to collect pharyngeal swabs from three of the cats. The remaining 77 samples were collected using suitable sterile swabs which were dipped immediately after collection into DMEM medium with added antibiotics. After this the samples were kept at +4°C during transport to the laboratory and were then stored at −80°C until testing. Nucleic acid extractions were performed using a commercial kit (QIAmp UltraSens Virus, Qiagen), useful for the extraction of DNA and RNA from liquids containing few cells. The method followed was the one recommended by the manufacturer. The treated samples were first submitted to single PCR specific for FHV-1 and Chlamydophila spp. and then to the duplex-PCR. The FHV-1 target sequence was 321 bp in length and included the TK gene coding for thymidine-kinase enzyme. The following pair of primers was applied: FHV-F: 5∞-T GT CCG CAT T TA CAT AGA T GG-3∞ sense primer; FHV-R: 5∞-GGG GT G T T C CT C ACA TAC AA-3∞ antisense primer. Single PCR specific for Chlamydophila spp. For Chlamydophila spp. the target sequence was 590 bp in length and it included the OMP2 gene encoding for the outer membrane protein (Hartley et al., 2001) . The following pair of primers was applied: Chla AF: 5∞-ATG TCC AAA CTC ATC AGA CGA G-3∞ sense primer; Chla AR: 5∞-CCT TCT TTA AGA GGT TTT ACC CA-3∞ antisense primer. Duplex-PCR for Chlamydophila spp. and FHV-1 Simultaneous amplification reactions were performed in 50 ml of total volume containing 4 ml of viral and bacterial DNA, HotMaster Taq 1X buffer, 200 mM each of dATP, dTTP, dGTP and dCTP, 2.5 U HotMaster Taq DNA polymerase (Eppendorf ), 0.04 mM FHV-F and FHV-R, 0.2 mM Chla-AF and Chla-AR. The reaction mixture was subjected to 40 cycles of 1 min at 94°C, 1 min at 58°C and 1 min at 72°C followed by final extension of 7 min at 72°C. PCR products were subjected to electrophoresis through a 2% agarose gel. After staining with ethidium bromide the visible bands of expected size were visualized with ultraviolet transillumination. PCR products resulting from amplification of the OMP2 gene of Chlamydophila spp. were subjected to restriction endonuclease analysis using the HINDIII enzyme (Biolabs, New England) to proceed to species identification. HINDIII enzyme recognizes the AiAGCTT sequence and is able to cut the PCR product positive for Chlamydophila felis into two fragments of 122 bp and 468 bp in length, thus discriminating Chlamydophila felis from other species. Digested fragments were analyzed using 3% agarose gel electrophoresis and visualization with ethidium bromide staining and ultraviolet transillumination. Twenty-four conjunctival swabs (60%) and twenty-one pharyngeal swabs (56.70%) were positive for FHV-1 and only nine conjunctival swabs (22.50%) were positive for Chlamydophila spp. The swabs positive using single PCR were also positive using duplex-PCR and all swabs positive for Chlamydophila spp. were positive for FHV-1. Furthermore, eight animals were positive for both pathogens. Only fourteen cats had been vaccinated against FHV-1 and none against Chlamydophila felis. The results divided according to the vaccination against FHV-1 reveal that most of the positive swabs originated from animals that had not been submitted to vaccination. As regards Chlamydophila spp. the highest number of positive swabs originated from animals treated with non-specific antibiotic therapy (7/27). Another interesting aspect concerns the type of swabs collected. In fact only the ocular swabs were positive for Chlamydophila spp., while the results were more complicated for FHV-1. Indeed, three animals had positive pharyngeal swabs, four had positive ocular ones and eighteen had positive results for both. Finally, the restriction analysis performed using HIND III enzyme on the amplicons positive for Chlamydophila spp. showed the same pattern and therefore was concluded to belong to the Chlamydophila felis species. 298 The data coming from this investigation underline the fact that FHV-1 and Chlamydophila spp. are present in mucosal swabs collected from cats with respiratory symptomatology living in the different geographical areas subject to investigation. Another interesting aspect is that related to the type of sample sent to the laboratory. In fact, it must be underlined that while for Chlamydophila spp. positive results were found only in swabs from conjunctiva, for FHV-1 the infection was diagnosed on swabs collected from both conjunctiva and/or tonsils. Restriction analysis applied to positive amplified sequences highlighted a substantial uniformity in the nine strains, and therefore, all are identifiable as belonging to the Chlamydophila felis species. The correspondence of the results obtained using single specific PCRs towards each pathogen and those from duplex-PCR shows that this last method has identical diagnostic features to the specific PCR. The use of a single test which is able to identify both pathogens simultaneously, is particularly important, since it permits early diagnosis, using a test that has a higher sensitivity and specificity than the traditional techniques. Angiogenesis, the formation of new vessels from the pre-existing vascular bed, is a key event in physiological processes, such as embryogenesis and cyclic changes of the ovary and endometrium (Dvorak, 1995) , as in pathologic processes, in particular in oncology. Neoplastic cells, having a high proliferative rhythm and increased metabolic requirements, are able to induce the formation of new vessels by angiogenesis (Folkman, 1990) . Therefore, a correlation between neoplastic proliferation and angiogenesis has been demonstrated in several tumoral types (Vartanian et al., 1995) . In addition, the formation of new vessels is also correlated with the metastatic potential of a neoplasm. For these reasons, the microvessel density (number of microvessels per mm2 of neoplastic tissue) which is the direct expression of the angiogenic potential of a neoplasm, together with the evaluation of vascular parameters (area and perimeter, in particular) are considered powerful prognostic tools in both human and in veterinary medicine (Fox, 1996; Weidner, 1998; Maiolino et al., 2000 Maiolino et al., , 2001 Restucci et al., , 2002 . 30 samples of gingivas, six normal and 24 pathologic, were examined; these were classified, using the Pindborg criteria (1997), as medium dysplasia (7), severe dysplasia (9) and carcinomas (8). All samples were fixed in 10% neutral buffered formalin and embedded in paraffin wax. 4 mm thick sections were immunohistochemically stained (streptavidin-biotin-peroxidase) with a monoclonal mouse anti human Von Willebrandt factor, which labels the endothelial cells (Dako, Denmark). 300 The microvessel density was assessed by selecting at 200× magnification (20 objective and 10 ocular) the more immunolabelled areas, located in or under the gingival epithelium using a microscope (Nikon Eclipse E-600, Tokyo, Japan) coupled with a video camera (JVC TK-C1380E Tokyo, Japan). At least 20 areas were selected for each sample. The images were then captured at higher magnification (400×), and manual outlining of the vessels was performed; areas, perimeters and number of vessels per square millimetre were calculated, using an image analysis system (MONO, Imagine e Computer System, Milan). Each immunolabelled endothelial cell separate from adjacent microvessels was counted as a single microvessel, according to standards of evaluation (Weidner et al., 1993) . Data obtained expressed as mean and standard deviation were correlated with histological grades of dysplasias and carcinomas, by analysis of variance. The count of immunolabelled vessels per 0.304 mm2 on image analysis showed a progressive increase in number of vessels proceeding from normal gingivas, to dysplasias, to carcinomas. A decrease in the area and perimeter of vessels was observed in carcinomas as compared with dysplasias and normal gingivas. These differences were not statistically significant (Table I) . In human medicine, some clinical studies have demonstrated that 10-20% of oral dysplastic lesions progress to carcinoma and the risk of neoplastic transformation is 43% for severe dysplasia (Pazouki et al., 1997) . In this study a progressive increase in the number of vessels was observed from normal tissues, to dysplasias to carcinomas, suggesting an implication of angiogenesis in neoplastic progression also for oral cavity lesions of the dog. These results are in agreement with previous studies, in which the so-called ''angiogenic switch'', by which a cellular phenotype commutes and becomes able to activate angiogenesis, is fundamental for the acquisition, by neoplastic cells, of features of malignity such as fast growth and metastatic capability. The decrease in the areas and perimeters of vessels in severe dysplasias and carcinomas reflects an increased angiogenic rhythm, which can give rise to numerous, but small and often abnormal, new vessels. In conclusion, angiogenesis may be involved in the neoplastic progression of gingival lesions in dogs, and microvessel density and vascular parameters may be sensitive prognostic parameters. Transmissible Spongiform Encephalopathies (TSE) are a group of disorders of the CNS that alarmed the scientific community especially because of the strictly correlated recent epidemics of BSE in cattle and vCJD in humans. The possible transmission of the disease from animal to man, the inability of the detection of the pathologic form of the prion protein (PrPres) before the appearance of symptoms and the lack of effective therapies against the infection make the development of prophylactic measures against TSE urgent, even if these are not easy to achieve due to the very low immunogenicity of the prion protein (Hill et al., 1997; Mabbott et al., 2001) . Encouraging results have been obtained in the prevention of Alzheimer's disease and experimental scrapie in murine models using repeated injections of synthetic peptides or recombinant proteins: therefore, it seems to be possible to bypass the immune tolerance toward self proteins by using proper immunization protocols (Schenk et al., 1999; Sigurdsson et al., 2002) . In this work we investigated whether active immunization of hamsters using synthetic peptides that reproduce different parts of the sequence of the homologous PrP and a fragment of human PrP can prevent or influence the evolution of the experimental prion disease. Hemocyanin) . The first vaccination and two boosters (at 10 day intervals) were performed one month before challenge. The first group of animals was infected by the intra-cerebral route with 103 LD 50 /30 ml and the second group by the intra-peritoneal route with 105 LD 50 /500 ml of 263 K hamster-adapted scrapie strain. Five boosters with the same antigens were given at 20 day intervals after the challenge. The brains of the animals that died during experimentation were removed and cut midsagittally; one cerebral hemisphere was kept in 20 ml of 10% formalin for histopathological and immunohistochemical analysis. The presence and spread of typical neurological lesions and PrPres deposits in the brain were evaluated using hematoxylin-eosin staining, the deposits of pathologic prion protein was detected using a rabbit polyclonal antibody, anti PrP (kindly provided by the Health Ministry) and revealed using a peroxidase labelled anti-rabbit serum. The remaining cerebral hemisphere was kept at −80°C for Western blotting analysis. The animals immunized and infected by the intra-cerebral route died at the same time, or even before, the positive controls (Table I) . The typical histopathological scrapie lesions were more evident in the immunized groups as compared with the controls: the medulla is the encephalic district in which the lesions appear more diffuse for all experimental groups. The immuno-histochemical analysis showed a slightly smaller amount of pathological protein only for animals immunized with the heterologous peptide PrP 82-146. The distribution of PrPres appeared quite uniform in the different encephalic districts, with less involvement of the hippocampus. Differences in survival curves were observed in immunized and intra-peritoneally infected animals; hamsters immunized with peptides PrP 119-146 and PrP 142-179 survived 25 days longer than the controls (Table I) with no significant differences between them (p values 0.2517 and 0.2867, respectively). The amount of pathological protein in the brain was lower than for controls and was generally proportional to the extension of the histopathological lesions, even some exceptions were observed. With the aim of explaining the early mortality of the i.-c. infected immunized animals as compared to the controls we hypothesized that an auto-immune response was The identification of suitable therapies against TSE (Transmissible Spongiform Encephalopathies) is a pressing problem to address. Different anti-prionic compounds have been shown to be effective in vitro, and some of the molecules tested are able to prolong the incubation period of the disease in experimentally infected animals (Brown, 2002) . Alzheimer's Disease and prion diseases share a number of common features: both are characterized by the generation of misfolded proteins (PrPres in prion disease), causing amyloid deposits in the CNS. The capability of Clioquinol (5-chloro-7-iodo-8-hydroxyquinoline) to dissolve amyloid plaques in post-mortem brain tissue from Alzheimer's patients, and also to dissolve Alzheimer-like plaques in transgenic mice has recently been shown (Melov, 2002) . Clioquinol acts as a zinc and copper chelator: the interaction with metals seems to be involved in the deposition and stabilization of amyloid plaques, therefore chelating agents can interfere with abnormal protein deposition by preventing metal/A-b amyloid interaction, (Cuajungco et al., 2000) . Since copper and zinc binding is probably also involved in PrPres conversion, we have tested the anti-prionic activity of Clioquinol on hamsters experimentally infected with scrapie. Different groups of Golden Syrian hamsters were infected either using the intracerebral (i.c.) or intra-peritoneal (i.p.) route, with a suspension of pooled brains (prion titer of the inoculum=103 LD 50 /30 ml and 105 LD 50 /500 ml respectively), derived from hamsters sacrificed at the final stage of experimental infection with the 263K scrapie strain. These animals were orally treated with 7.5 mg/kg/day of Clioquinol (EnteroviorformioA) dissolved in drinking water. One group of non-infected hamsters was also treated with Clioquinol (control of chronic toxicity). The animals were housed in single cages (IVC, Tecniplast) supplied with HEPA filtered air through a self-closing valve. Experimental procedures followed the EU International Guidelines for animal experimentation. The procedures for evaluation of the anti-prionic activity of Clioquinol include General Behaviour Observations (GBO), which consists of analysis of passive avoidance and spontaneous motor activity (Braida et al., 2000; Sala et al., 1997) , performed twice (61-65 days and 73 days post-infection for i.c.-and 150 days post-infection for i.p.-infected hamsters). Passive avoidance: an apparatus consisting of a box divided by a wall into two compartments (one of which is dark) of the same size, in which the floor has a grid of steel rods, was used. Retention of the avoidance response was tested 24 h after the electric shock trial (2 mA/3 s). The latency of re-entry into the dark compartment was recorded up to a maximum of 180 s. Spontaneous motor activity: this parameter was evaluated in an activity cage (U.Basile, Italy), placed in a sound-attenuating room. The cage was fitted with two parallel horizontal infrared beams located 2 cm from the floor. Cumulative horizontal movement counts were recorded for 10 min. Animals chronically treated with Clioquinol showed no toxicity symptoms until the end of the experiment. I.c. scrapie-infected animals treated with Clioquinol survived only 5.5 days longer than untreated controls. For Clioquinol-treated i.p. infected animals survival (D= 40.7 days) was significantly ( p=0.0347) longer than for untreated controls. The results of GBO are shown in Fig. 1 . Scrapie i.c.-infected hamsters showed a strong impairment of mnemonic function and a slight, but not significant, reduction of spontaneous motor activity at 63 days post-infection. No change in these parameters was recorded 73 days post-infection. Treatment with Clioquinol only produced a slight improvement in mnemonic function. Scrapie i.c.-infected hamsters showed a strong impairment of memory function and a moderate reduction of spontaneous motor activity at 150 days post-infection. Treatment with Clioquinol produced a slight, but not significant, improvement in mnemonic function, while it completely restored the spontaneous motor activity. The early memory impairment observed in scrapie i.c.-infected animals suggests that the developing neuronal damage is in brain areas associated with learning. Therefore, Figure 1 . Result of general behaviour observation in animals infected by the i.c. route (upper graphs) and by the i.p. route (lower graphs). p<0.05; **p<0.01 vs controls; § §p<0.01 vs infected the evaluation of memory deficit can be a useful tool in the early detection of infection before development of the classical clinical signs. The slight delay in memory and motor activity detriment and the improvement of motor activity observed for i.p.-scrapie infected hamsters chronically treated with Clioquinol, as well the significant elongation of the survival curve, suggests a possible therapeutic effect of the drug. The i.p. route of scrapie infection is closer to the natural infection, which is assumed to happen through the intake of infected food. These results suggest that Clioquinol can interfere with the development of the disease, as the GBO indicate. Therefore, this antibiotic possesses some of the ideal properties for a candidate agent against TSE, also because its low toxicity can allow experimentation into use of higher doses of the drug, which may be more active. With few exceptions, the parasitological profile of cats has been poorly researched. This is probably because of the objective difficulties experienced during the sampling stage, such as blood sampling or exact identification of faecal samples taken from subjects living in colonies. For these reasons, and considering that no research has been conducted on this topic in Sardinia, except for the paper by Arru and Lai on Ancylostomosis back in 1965, we decided to report here the results of an epidemiological survey conducted on our island. The aim was to assess the main parameters of the principal parasitoses present, some of which are also of zoonosic interest (T oxoplasma spp., T oxocara spp., Uncinaria spp., Dipylidium spp.). 160 cats aged between 2 months and 16 years belonging to different breeds, some of which lived exclusively in flats, and others more or less regularly outside, were coprologically monitored over the period between April 2000 to October 2001. In addition, IgG and IgM anti-T oxoplasma gondii were sought in the serum of 54 subjects using indirect immunofluorescence (IFI). The faecal samples were first examined macroscopically to check for the possible presence of proglottides which were observed without any staining and/or by colouring with hydrochloric carmine and which were then isolated and classified. The taxonomic keys normally used were those indicated by Khalil et al. (1994) . The method used in the subsequent copromicroscopic examinations is one that provides for sedimentation and flotation with resolution of specific gravity 1450. 54 of these samples were additionally processed to check for the presence of larvae of Aelurostrongylus abstrusus using Baermann's methodology. A necroscopic examination was carried out on 23 subjects which had died of natural causes (or because they 312 were run over by motor vehicles) with the aim of isolating and identifying the parasitic agents present at the gastro-intestinal, broncho-pulmonary and cardiac level. The propagation forms found and the samples of parasites isolated on autopsy were classified according to the morphometric keys proposed by Sloss and Kemp (1985) and Thienpont et al. (1986) . A serum prevalence of 37% of IgG and no positive results for IgM were found for T oxoplasma gondii; and no oocysts of this protozoan were found in the faeces. It is thus evident that we were faced exclusively with the chronic stages of this infection, and this confirms that the role of the cat in the spread of this parasitosis is marginal and/or exclusively linked to very young subjects and/or those in bad immunitary conditions. Furthermore, the serological data do not differ too much from those noted in other national districts (33% Tenter et al., 2000) or international data (42% Rhode Island, U.S. A. -De Feo et al., 2002; 26.3% Sao Paolo, Brasil -Silva et al., 2002) . Table I shows the results obtained both from the coprological examination and from autopsies. Total percentage of infestations and/or mono-specific infections were 26.9%. The stratification of results obtained for cats according to sex did not show any (Poglayen et al., 1985) . However, infestation by A. abstrusus, a parasitosis that is generally underestimated, is another matter altogether. In the present research, we have shown that normal copromicroscopic analyses carried out on live animals can generally only reveal the presence of this nematode in a limited number of cats, in contrast with results obtained using Baermann's methodology or an autoptic examination. It is clear, then, that if parasitoses are present in over 20% of autopsies carried out on cats, it is to be hoped that any routine examination carried out in veterinary surgeries would include Baermann's methodology, bearing in mind the pathogenic power generally attributed to this nematode (Marcato, 1988) . Moreover, research unfortunately confirms, once again, the limits of parasitological analyses, particularly when confronted with Cestoda, which were found for 39.1% of cats undergoing an autopsy, as compared to 10.6% of those analysed copromicroscopically ( p<0.001). In no case was there evidence of Filariid nematodes in the cardiac cavities or pulmonary arteries. Mycobacterioses, also known as Tuberculosis, are the most common chronic bacterial diseases in ornamental fish, affecting both temperate and tropical species both in freshwater and the marine environment (Ghittino, 1985; Bozzetta et al., 1995; Chinabut, 1998; Prearo et al., 2002) . Moreover, they may be found both in wild fish and those cultured for human consumption (Colorni, 1992; Dos Santos et al., 2002) . Besides inducing mortality in fish, they represent a potential hazard for man, being included among zoonoses (Ghittino and Bozzetta, 1994) . Mycobacterioses are caused by ubiquitous bacteria that are highly resistant in the aquatic environment and are difficult to control. Mycobacterium fortuitum, M. marinum and M. chelonae are the most commonly isolated species. They are all capable of inducing human cutaneous infections, with development of nodular lesions that are generally localized in the limbs (Ghittino, 1985) . In fish, symptoms appear late, are non-specific and include slow growth, lethargy, anorexia and starvation. Lesions in the skin and typical whitish nodules in the viscera may be detected at necropsy (Ghittino, 1992) . Therapy is rather difficult, both for the fact that early diagnosis is seldom achieved and for the lack of effective drugs (Ghittino, 1985) . In the present work, the results from an 18 month survey on cultured and wild ornamental fish are reported, with the aim of improving knowledge on the distribution of Mycobacterioses in Italy. 312 aquarium fish, both of domestic and foreign origin, were sampled by importation and distribution centers and by private centers, over a period from July 2001 to February 2003. 266 freshwater and 46 marine fish were subjected to necropsy and, when nodular lesions were observed, Ziehl-Neelsen stained imprints were performed. Portions of liver, spleen and kidney were fixed in a 10% formalin solution, processed and stained with hematoxylin-eosin and Ziehl-Neelsen for histological examination. A specific bacteriological examination was carried out on all fish. The entire specimen, in the case of small fish, or the viscera, for larger fish, were minced, diluted in deionized sterile water, homogenized and filtered. This solution was decontaminated with HPC (1:1), centrifuged and the pellet obtained was inoculated into selective media (Loewenstein-Jensen and Stonebrink), and incubated at 30°C until colonies developed. Colonies that were positive at Ziehl-Neelsen were subcultured to perform biochemical identification. The tests utilized were nitrates reduction, urease, growth on 5% NaCl medium, growth at 37°C, growth on McConkey agar without crystal violet, niacin, Tween hydrolysis, arylsulfatase, cold catalase and catalase at 68°C for 20 min. 135 samples, out of 312 analyzed, were found to be positive for mycobacteria on bacteriological examination (45%). 75 of these had already been found positive by imprint (55%). Among these clear macroscopic findings of Tuberculosis were observed only for 38 fish (50%). 34 specimens with tubercular lesions belonged to freshwater species and 4 to marine species. Exophthalmos and keratitis were observed, but not skin lesions reported in literature (Ghittino, 1985; Chinabut, 1998) . Visceral nodules had a size of 0.5-12 mm and were variable in number. 17 out of the remaining 37 fish without tubercular lesions but identified as being positive by imprint, were positive on histological examination. Granulomas, at different stages of development, were found in the liver, spleen and kidney. These were made of epithelioid cells containing acid-fast bacteria, surrounded by a thick connective capsule. Out of 135 isolates of mycobacteria, 109 were phenotypically identified as Mycobacterium fortuitum, 16 as M. chelonae, 5 as M. marinum, 3 as M. terrae and 2 as M. gordonae. Mixed infections, caused by different mycobacterial species, were never observed in the same fish sample. The most common freshwater fish affected by Mycobacterioses were the gold fish (Carassius auratus), the Siamese fighting fish (Betta splendens), Colisa lalia and Barbus tetrazona. Among marine fish the most affected species was found to be Zebrasoma veliferum. Our results confirm that Mycobacterioses are widely spread among ornamental fish in Italy and may be considered as one of the main bacterial diseases. For this reason, these should always be suspected in weak and stressed fish. Unlike what is reported in the literature, where the most frequent mycobacterial species in freshwater fish is considered to be Mycobacterium marinum (Ghittino, 1985; Chinabut, 1998) , we found a higher prevalence of M. fortuitum and M. chelonae. The exact role played by M. terrae and M. gordonae as fish pathogens is still unknown, since these mycobacteria were isolated from fish without macroscopic or microscopic signs of disease. Considering the increasing importance of aquarium fish as pet animals and the lack of specific laws on this topic, a focused study into the health risks due to mycobacteria in ornamental fish is recommended. Small ruminant lentiviruses (SRLV) are a heterogeneous group of viruses causing persistent infection and chronic debilitating diseases in sheep and goats worldwide. Economic losses are related to the drop in milk production due to progressive pneumonia, lameness, and mastitis. Phylogenetic studies have been carried out using ovine or caprine viral strains obtained from different geographical regions and recent data suggest that a differentiation of SRLV based on host species origin may not be accurate (Zanoni, 1998) . This is consistent with the identification of Maedi-Visna Virus (MVV)-like strains in goat and Caprine Arthritis-Encephalitis Virus (CAEV)like strain in sheep. Recently a new cluster of ovine lentivirus isolates, related to CAEV has been described in France and Italy (Leroux et al., 1997; Grego et al., 2002) . Phylogenetic studies have mainly been carried out using pol gene fragments, while limited information is available for gene fragments encoding major structural proteins involved in the antibody response and used for the development of serological tests. The use of a single strain-based immunoassay in detecting sheep and goat lentivirus infection is based on the finding that immunodominant regions are mainly crossreactive among ovine isolates, as well as between ovine and caprine lentiviruses (Gogolewski et al., 1985) . However, recent studies suggest that the heterogeneity of the immunodominant epitope of the capsid antigen may affect sensitivity of serological tests based on a single strain (Grego et al., 2002) . In this study the gag gene of two genetically distinct ovine lentiviruses was characterized and epitope variability in the matrix protein (p16) and in the maior capsid antigen (p25) was evaluated. Two ovine lentiviruses isolated in Italy were used for this study. The former, It-561, has been previously classified in the MVV cluster while the latter, It-Pi1, has been 320 clustered in the new genotype named CAEV-like. Partial sequences of the pol and gag genes of the two strains have been previously determined (Grego et al., 2002) . On the basis of the available sequences coding for the pol gene (GenBank accession number AY044811 and AY044815) and the highly conserved LTR region, specific primers were selected to amplify a 4 kb genomic fragment encompassing the entire gag gene of the two lentiviruses. Amplicons of the expected length were cloned into the pDRIVE vector and sequenced. The gag gene sequences of strain It561 and It-Pi1 were aligned with the consensus sequence of the SRLV sequences available in the GenBank database. Genetic distances were computed using MEGA (Kumar et al., 2001) and were used to construct a neighbour-joining tree (Saitou et al., 1987) . Statistical support for the tree was evaluated by bootstrapping (Felsenstein, 1985) . Immunodominant regions of p25 have been previously characterized (Rosati et al., 1999) , while relevant epitopes of p16 were determined in this study using recombinant subunits expressed in frame with glutathione S-transferase. Fusion proteins were purified using affinity chromatography and used in indirect ELISA to characterize the reactivity of sheep sera collected from different Italian flocks. The nucleotide sequence data reported in this paper have been submitted to GenBank (Accession number AY265455 for It-561 and AY265456 for It-Pi1). Phylogenetic analysis carried out on the gag gene confirmed that strain It-561 was genetically similar to MVV Icelandic prototype K1514 while strain It-Pi1 was more similar, although distinct from, the CAEV strain CO (Fig. 1) . Two immunodominant regions of p25 (previous study) (Rosati et al., 1999) and p16 (this study) were identified in the two strains and a moderate variability in the aminoacid sequence was observed (see Table I ). Ovine sera used in the present study mainly exhibited reactivity against the antigens derived from It-Pi1. A previous phylogenetic study has been conducted in Italy on SRLV isolated in the last decade. Most ovine isolates (8 out of 9) belonged to the same cluster of It-Pi1 named CAEV-like. Thus, in Italy, many of the infected ovine population harbour viral strains which are different from those employed in the currently available diagnostic immunoassays. In fact, in different commercially available ELISAs, recombinant antigens have been developed from the classical ovine lentivirus genotype (i.e. SAOVV, EV1, ZZV1050) (Saman et al., 1999; Zanoni et al., 1991) . This genotype is still circulating in our country, as demonstrated by isolation of strain It-561. The genetic variability of the field isolated raises a question regarding the validity of the single genotype-based serological test. This study clearly shows that immunodominant regions of two structural proteins are quite variable in the two ovine genotypes. A more accurate epitope characterization using overlapping fragments revealed a common or less variable N-terminal epitope and a distinct C-terminal epitope. Thus, the cross-reactivity between different genotypes is given by these and other conserved sequences. However, a subset of infected population may be partially reactive against common epitopes and not be detected using a diagnostic test based on the heterologous genotype. For CAEV infected goat flocks we observed that the homologous antigen is able to detect an additional 7% of seropositive animals as compared with heterologous antigen (Grego et al., 2002) . The systematic use of a single strain immunoassay in an infected population, followed by slaughtering of seropositive animals may select for ''diagnostic test escape'' viral mutants, thus interfering with eradication programs. The recombinant subunits employed in the present study could be used in parallel to evaluate the genotype circulating in the flock under study and subsequently apply the homologous antigen for diagnostic purposes. Alternatively, a combination of antigens derived from the two genotypes in a single test may increase the sensitivity of serological diagnosis. Physical examination and standard electrocardiography have been used for several years to check the cardiac electrical activity and in particular to detect cardiac arrhythmias. Routine electrocardiography is the gold standard for the diagnosis of cardiac arrhythmias in practice. However, as a standard ECG lasts 1 or 2 min, that is to say 0.07-0.14% of 24 h, its sensitivity to cardiac arrhythmias is correlated with the frequency at which they occur. ECG sensitivity is high for continuous arrhythmias, such as atrial fibrillation in the dog, or very frequent arrhythmias, such as certain cases of ventricular premature complexes (VPCs), but is very low for detection of sporadic arrhythmias such as parossistic tachycardia. Furthermore, standard ECG requires the patient to be restrained in right lateral recumbency, influencing the autonomic nervous system and potentially inducing or suppressing cardiac arrhythmias (Doxey et al., 1985) . In 1963 N.J. Holter proposed monitoring of cardiac rhythm for 24-48 h using a portable recorder to look for arrhythmias during the normal activity of the patient in order to increase the sensitivity of the ECG to cardiac arrhythmias. Nowadays 24-hour ambulatory electrocardiography (Holter monitoring) is widely used in human medicine, and less commonly in veterinary medicine, to detect cardiac arrhythmias in patients affected by syncope, sudden weakness or sudden dyspnoea, and to assess the efficacy of antiarrhythmic treatment. Furthermore, Holter monitoring can be useful for an early diagnosis of cardiac arrhythmias in particular dog breeds in which arrhythmogenic cardiomyopathies are inherited or where modifications of the cardiac rhythm are associated with extracardiac disease. Holter recordings were obtained using a Braemer DL 700 digital recorder (15.24-8.89-2.41 cm; 283 gr) that stored the data in a 10 MB flash card. We shaved the thorax bilaterally over the middle and ventral third of the ribs, between the 5th and the 8th rib on the right side and between the 3rd and 7th rib on the left. We cleaned the skin using gauze with alcohol in order to remove sebum and dirt from the surface and to ensure a good contact between electrodes and the skin. We used selfadhesive electrocardiographic patches with 2.4 cm diameter. Each patch was covered with self-adhesive tape to avoid their displacement or detatchment. Seven electrodes were placed in order to obtain 3 leads with a V1, V3, V5 modified system (Gershwin, 1975; Doxey et al., 1985; Elie and Hoenig, 1995) . Electrocardiographic patches and their wires were secured using a strip of self-adhesive tape that fixed the recorder between the scapulae on the withers region. The recorder was further secured to the patient using a 3M VetrapTM elastic bandage that also covered the wires and the patches. Two loops of elastic bandage were crossed between the forelimbs to give more stability to the recorder and electrodes and to avoid them moving caudally and laterally. Each dog went back home with its owner who had to check the patient during recording to avoid damage to the recorder and to note all activity of the patient. The recording was then processed using the Holter Win P-V 5.40 Analyser installed on a notebook. The analysis was based on the general morphologic ECG features of the analyser but it was possible to modify and to upgrade them constantly before, during and after the automatic analysis. The tables with maximum, minimum and mean heart rate, supraventricular and ventricular ectopies and full disclosure of the recording were printed out for each patient. We performed the Holter monitoring as an ancillary diagnostic test for 8 dogs of different breeds and sex, with an age range of 1 to 10 years and with a weight ranging from 5 to 72 kg. Each dog was conducted to the Section of Internal Medicine of the Pathology, Diagnostic and Veterinary Clinic Department of the Veterinary Faculty, University of Perugia for a cardiologic check up. We also performed a Holter recording in 3 clinically normal dogs as a control group. Each recording was obtained without any kind of restriction or of change in the daily activity of the patient. Holter recordings were clear enough to monitor the cardiac rhythm during the examination, even though in some cases they lasted less than 24 h, and to allow the analyser to work properly. It was possible to identify cardiac arrhythmias undiagnosed by physical examination or standard ECG and to have an electrocardiographic diagnosis in each case. Some of the arrhythmias we found were considered pathological and requiring proper antiarrhythmic treatment. Physiological arrhythmias such as marked sinus bradycardia or sinus arrest, even lasting 3.5 s, or a brief period of sinus tachycardia, with the heart rate over 250 bpm, which can be considered the result of the autonomic nervous system influence (Davies et al., 1984) were detected, as well as definitely pathological arrhythmias: ventricular INTRODUCTION Plasma thrombomodulin (TM) is a membrane glycoprotein in the vascular endothelium which shows high affinity to thrombin. The thrombomodulin-thrombin complex induces coagulation by activation of protein C that inactivates the intrinsic pathway plasma factors V and VIII, finally blocking the formation of new thrombin. Plasma TM is mainly produced by endothelial cells and to a lesser extent by lymphocytes, neutrophils, monocytes, etc. Several studies performed, both in humans and laboratory animals, showed that TM levels were increased during haemostatic disorders (disseminated intravascular coagulation, thrombosis, etc) and during several diseases that induce endothelial damage (such as diabetes mellitus, lupus erythematosus and atopia), and renal and hepatic involvement, rickettsiosis, etc. The main characteristics of such diseases are micro and macroangiopathies due both to direct action of infective agents or to immune activity. The presence of vasculites has also been reported during leishmaniasis and it can be responsible for a very severe and often fatal evolution of the disease (Pumarola et al., 1991) . Antigen-antibody immunocomplexes deposit in the endothelium inducing complement activation and inflammation. In canine leishmaniasis clinical signs such as renal failure, uveitis, arthritis, etc. are generally due to this autoimmune process. When these clinical signs are evident it is generally too late for a therapeutic approach, therefore, identification of markers for early detection of endothelial damage can be of great help to the veterinarian. The object of this study was to evaluate plasma TM concentrations in dogs affected by leishmaniasis at different grades of clinical involvement, in order to verify if plasma TM could also be utilized in dogs, as in humans, as a marker for endothelial injury. Moreover, the study of such a marker could lead to a better understanding of the mechanisms involved in this zoonosis. Haemostatic abnormalities such as epistaxis, haematuria and haemorrhagic diarrhoea, have been reported in canine Leishmaniasis (CanL) . Thrombocytopenia, thrombocytopathy, prolonged thrombin time (TT), prolonged activated partial thromboplastin time (APTT) and increased fibrinogen/fibrin degradation products have been found on L eishmania infection, and indicate that this parasite affects primary haemostasis, coagulation, and fibrinolysis (Moreno, 1999) . The effect on haemostasis is complex and not all the haemostasis alterations reported in the literature can be found at once in a dog with leishmaniasis because of the typical clinical polymorphism of the disease. The pathogenesis of these disorders is not well known but, in human leishmaniasis, it has been attributed to necrotising vasculitis, renal and/or hepatic failure and immune-mediated disorders. The object of this study was to evaluate platelet function and secondary haemostasis in dogs naturally infected by L eishmania infantum with different grades of clinical involvement, in order to improve knowledge of the haemostatic disorders found for this protozoan disease. 30 dogs naturally infected by L eishmania infantum and a control group of 10 healthy dogs were examined. None of the infected dogs had ever undergone any specific anti-Leishmania therapy. Sick animals were divided into three groups of 10 dogs on the basis of the clinical examination : first group -oligo-symptomatic dogs; second group -symptomatic dogs; third group -markedly symptomatic dogs. The clinical diagnosis was always confirmed by direct observation of the protozoan in Giemsa stained fine needle aspirates of bone marrow and lymph nodes, and serologically using the immunofluorescent antibody test (IFAT). The threshold titre for positivity was 1/160. A full blood count and measurement of total plasma proteins (TP), serum protein electrophoresis, alanine transaminase (ALT), creatinine and urea serum levels were obtained for each dog. Prothrombin time (PT) and activated partial thromboplastin time were determined twice using a semi-automatic coagulometer. The aggregation response of platelet-rich plasma was measured using a turbid metric method with a Chronolog aggregometer after addition of collagen type I calf skin and adenosine 5∞-diphosphate (ADP) suspension. The collagen and ADP suspension were used individually at doses within a range of 2.5-200 mg/ml and 2.5-10 mM. Mean values with their standard deviations were calculated from the data obtained and then statistically evaluated using one-way analysis of variance. Significant results were subjected to the Student-Newman-Keuls Multiple Comparisons Test. Compared with the control group, the maximum relative aggregation response decreased significantly in each of the groups at the various doses employed. Decrease in platelet aggregation using collagen was more evident than that observed with ADP agonist (ADP-vs-collagen 83.7 vs 78.4%). When higher doses of collagen were employed the third group showed a significant decrease in aggregation response to collagen as compared to the first and second groups (Table I) . A reduction in haemoglobin plasma levels and an increase in urea and creatinine plasma levels were observed for the third group; the same group showed moderate thrombocytopenia, while no difference in platelet count was found in the other groups of sick dogs (Table I) . A longer APTT was observed for the second and third groups as compared with the control and the first groups. Finally, the A/G ratio dropped progressively in all three groups of infected dogs. As previously reported under natural and experimental conditions (Moreno et al., 1998; Valladares et al., 1998) , our results also show that a significant reduction in platelet aggregation can be found in CanL. In agreement with previous preliminary work (Ciaramella et al., 2002) , the reduction in platelet aggregation produced by the two agonists was different. The greater sensitivity of collagen as compared with ADP can be explained by considering their It is known that serum levels of cytokine, including interleukin-I (IL-I) and interleukin 6 (IL-6) are very high in cancer patients and are related to oxidative stress. These cytokines play a key role in the pathogenesis of cancer-related anorexia-cachexia syndrome (CACS) induced both by chronic administration of chemotherapeutic drugs and the tumour itself. Recently, Mantovani et al. (1998) demonstrated that high doses of Medroxyprogesterone acetate (MPA), a semisynthetic progestin, and Megestrol acetate (MA) reduce serum levels of interleukin-I (IL-I) and interleukin 6 (IL-6) in vivo for cancer patients with CACS leading to beneficial therapeutic effects. Moreover, clinical use of anthracyclines for the treatment of several malignant tumours has been limited both by a dose-dependent cardiotoxicity and drug resistance. Therefore, the clinical use of such chemosensitizers remains questionable and the identification of more active but less toxic compounds is needed. Our recent studies, in chronic lymphatic leukaemia cells, have demonstrated that MPA reduces Malondialdehyde (MDA) levels increased by anthracyclines (Pagnini et al., 2000) . In this study we evaluated the activity of two myocardial antioxidant enzymes, copperzinc superoxide dismutase (Cu-Zn SOD) and catalase (CAT) in the plasma of rats treated per os for 15 days with Idarubicin (IDA) and Medroxyprogesterone acetate (MPA), either used alone or in combination. Two hours after the last treatment, rats were killed by decapitation, blood was collected in heparinized tubes and was centrifuged at 13,000 g for 3 min. The spectrophotometric assay, as described by Ewing and Janero (1995) , was applied to evaluate Cu-Zn SOD activity. The assay is based on spectrophotometric assessment of O 2 − mediated nitro blue tetrazolium reduction by an aerobic mixture of NADH and phenazine methosulphate, which produces superoxide chemically at non-acidic pH. Absorbance was read at 520 nm using a Perkin Elmer UV Spectrophotometer. The colorimetric assay, described by Sinha (1971) (4) was applied to evaluate Catalase activity. This method is based on the fact that dichromate is reduced to chromate acetate when heated in acetic acid in the presence of H 2 O 2 , with the formation of perchromic acid as an unstable intermediate. The chromic acetate thus produced is measured colorimetrically and absorbance is read at 570 nm. Table I shows the effects of IDA (0.25 mg/kg/day) and MPA (25 mg/kg/day) used alone or in combination, on Catalase activity. The calalase decreased by 42.7±8.4% in plasma of rats treated with IDA but MPA alone did not modify this parameter. MPA used in combination with IDA reduced the inhibition activity of anthracyclines on Catalase. The inhibition shifted from 42.7±8.4% (IDA alone) to 25.5±9.47% (IDA+MPA).The effects of IDA (0.25 mg/kg/day) and MPA (25 mg/kg/day) used alone or in combination, on Cu-Zn SOD activity are shown in Table II . MPA reduced the inhibition activity of the anthracyclines on Cu-Zn SOD. The inhibition shifted from 31.5±4.41% (IDA alone) to 20.5±3.34% (IDA+MPA). Clinical and experimental data suggest that high concentrations of reactive oxygen species (ROS) such as superoxide (O 2 −) and hydrogen peroxide (H 2 O 2 ) can cause cellular damage and may play a key role in drug-induced toxicity (Lesko et al., 1980) . Such oxidative stress is not only responsible for the cancer genesis but also for both an anthracycline-induced toxicity effect and anorexia/cachexia. It has been reported that, as the cancer-related anorexia/cachexia syndrome (CACS) is correlated with cytokine levels, it is consequently also correlated with oxidative stress. It has been recently demonstrated that high doses of Medroxyprogesterone acetate (MPA) and Megestrol acetate provide protection from CACS by modulating cytokines and serotonin production and that such an effect is related to oxidative stress. Our results showed that high doses of MPA reduce the inhibition of Cu-Zn SOD and Catalase induced by idarubicin. In conclusion, these data are in agreement with those of other authors, showing that IDA-induced oxidative stress can be reduced by administration 338 of high doses of MPA. In this way, both reduction in cardiotoxicity and beneficial therapeutic effects can be achieved. In human medicine the genetic components of insulin-dependent diabetes mellitus (IDDM) have been investigated, and it is well known that many genes are involved in the pathogenesis of IDDM (Davies et al., 1994) . These genes, mainly belonging to the major histocompatibility complex (MHC), seem to enhance the development of antipancreatic T lymphocytes (Zamani et al., 1998) . Although there is evidence for the inheritance and immune mediated pathogenesis of IDDM in dogs, the genes involved in the spontaneous disease have not been identified and may vary from breed to breed (Gershwin, 1975; Kramer et al., 1988; Elie and Hoenig, 1995; Klimmer et al., 2002) . A familial predisposition for IDDM has been described for Samoyeds (Kimmler et al., 2002) , Keeshonds (Kramer et al., 1988) , Golden Retrievers (Williams et al., 1980) and Poodles (Gershwin, 1975) . Epidemiological studies on the risk of developing diabetes mellitus (DM) in dogs have mainly been focused on North American canine populations (Marmor et al., 1982; Hess et al., 2000) . To the best of our knowledge only one epidemiological survey has been carried out in Europe (Doxey et al., 1985) . Unfortunately this study did not show the results obtained in detail and probably for this reason very common Italian breeds were not mentioned. From anecdotal reports, some breeds such as the English Setter and Irish Setter may be at high risk of developing DM in Italy. The aim of this study was to evaluate the breed distribution and the risk of developing DM in an Italian canine population. The identification of these breeds may further direct studies concerning pedigree analyses, genetic aspects and mode of inheritance of this disease. By running a computer search of all dogs admitted to the Department of Veterinary Clinical Sciences of the Alma Mater Studiorum-University of Bologna between January 1995 and December 2002, it was possible to identify 225 dogs diagnosed with DM. Medical records regarding the physical examination, urinalysis and biochemistry were reviewed in detail for each dog and 186 dogs with DM were used for the study. Inclusion criteria consisted of persistent hyperglycemia with glycosuria and at least two common clinical signs of DM (polyuria, polydipsia, weight loss or polyphagia) or glucosuria with ketonuria. 39 dogs were excluded from the study because they had insufficient data for diagnosis (a single result of either hyperglycemia or glycosuria without clinical information about the dog), or transient (diestral) or secondary DM. At the end of this selection, the case series included 186 dogs with DM that had been admitted to our Teaching Hospital between January 1995 and December 2002. 13857 dogs without DM had been admitted to the Department of Veterinary Clinical Sciences during the same period and this was considered the comparison group. Of 135 different breeds (including the mixed-breed category) admitted to the Clinic over the period of the study, 23 (including the mixed-breed category) had at least one dog with DM. Statistical analysis was restricted to those breeds where the sum of observed and expected values for diabetic subjects (calculated by multiplying the number of subjects in each breed by the percentage of diabetics in the whole population studied) was>4. This procedure ensured that a sufficient numbers of dogs was considered in order to document breeds at low and high risk of developing DM. The odds ratio was calculated for each breed by choosing mixed breed dogs as the reference breed (odds ratio=1). A 95% confidence interval was also calculated. A value of p<0.05 was considered significant and <0.001 highly significant. 13 breeds showed a sum of observed and expected values>4. The following 11 breeds of diabetic dogs did not meet the criteria and were consequently excluded from the statistical analysis: Samoyeds (2 dogs), Lagotto (2), Maltese (2), Pug (1), Collie (1), Corsican (1), Fox Terrier (1), Miniature Schnauzer (1), Shih Tzu (1), Newfoundland (1), Pomeranian (1). The Boxer breed was included because it was highly represented (666) and also showed an expected count >4 (8.82), despite the fact that no diabetic dogs were present (observed count=0). In our survey, the Irish Setter was the breed with the highest risk of developing DM, having an odds ratio 3.91 times greater than that of the control group (mixed breed dogs). The following breeds were also found to be at high risk Among the breeds that had sufficient numbers of dogs to meet our sample size criteria for analysis, our results showed that four dog breeds are at high risk and three dog breeds are at low risk for developing DM. While our data for the Yorkshire Terrier and Poodle are substantially in agreement with the results obtained by other authors (Doxey et al., 1985; Hess et al., 2000) , our study also suggested that Irish and English Setters are respectively 3.9 and 2.6 times more likely to develop DM than mixed breed dogs. The low risk documented for German Shepherd, Boxer and Doberman Pinscher is in agreement with previous reports (Hess et al., 2000; Marmor et al., 1982) . The reason for the differences between the results of our study and those of other previously published reports, could be due either to the different distribution/representation of canine breeds in different geographic locations or to the presence of different genetic components of IDDM within the same breed (Gershwin, 1975) . A further limitation of our survey may be that the study population (percentage of diabetic subjects over the entire population examined=1.33%) may not accurately reflect the population of diabetic dogs at large. This may be because the Department of Veterinary Clinical Sciences of the University of Bologna is considered as a referral centre for DM. To the best of our knowledge, this is the first large European epidemiological study conducted for evaluation of the risk of DM development in canine breeds spread over the Italian territory. In conclusion our study suggests that Irish Setter and English Setter dogs should be added to the list of breeds at high risk of developing DM, at least while considering the Italian situation. In our opinion, these results constitute an important starting point for further research regarding pedigree analyses, genetic aspects and mode of inheritance of DM in the dog. Tetralogy of Fallot (TOF) is a congenital cardiopathy of man and domestic animals. Fallot was a cardiologist in human medicine of the late 19th-early 20th century. This cardiac malformation is characterized by ventricular septal defect, dextroposition of the aorta, pulmonic stenosis and consequent right ventricular hypertrophy. Cardiac morphological alterations in TOF derive from a defective fetal development of both pulmonary and aortic arteries, resulting from troncus arteriosus, and right and left ventricular outflow tracts, resulting from conus arteriosus. There is also an incomplete development of the conotruncal septum which causes a skew between dorsal and apical parts of interventricular septum. As shown by Goodwin and Cooper (1992) , pulmonary stenosis causes an increased resistance to blood flow ejected from the right ventricle. This leads to hypertrophy and hypertension of this ventricle, and leads to a right-to-left shunt through interventricular septum defect. The first consequence of the shunt is an inflow in aorta of bad oxygenated blood, that can easily induce hypoxia. Symptoms are failure to grow, exercise intolerance, cyanosis and syncope. Cyanosis depends on the entity of the shunt and it is not always present, but it is typical of most cases. For this reason, in the past the illness was called ''blue disease'' because of the colour assumed by mucosae. Cyanosis is always present when the subject gets excited or during physical effort, when even asphyxia can occur. TOF is more common for dogs than for cats (Patterson et al., 1993) . For this species diagnosis is performed by anatomo-pathological examination or angiography which is an invasive intra vitam examination involving a lot of risks (Eyster et al., 1977) . The aim of this report is to furnish a contribution to the study of this pathology in cats, especially for the non-invasive diagnostic methods such as ultrasound. 344 The subject of the study is a three-year old male European cat. Anamnesis was of depression, hypokinesia, increased respiratory effort, even with light physical activity, and lipothymic crises characterized by loss of standing position and unconsciousness. The lipothymic crises and dyspnea were related to the intensity of physical activity. The cat underwent physical, electrocardiographic, radiographic and ultrasonographic examination. It was impossible to collect a blood sample because of the occurrence of respiratory crisis during sampling. Electrocardiography was performed with the cat in right lateral recumbency, using a Nihon Kohden -Eclaps 12 -ECG 8110R instrument. Radiography was performed with a latero-lateral projection (left to right). M-mode, b-mode and Doppler echocardiographic evaluations were performed through the clipped right parasternal window and clipped apical and caudal left parasternal windows. The ultrasound machines used were a Kontron ''Vetson Color'', with a micro-convex 6.5 MHz probe and a Medison AS-8800, provided with a microconvex multi-frequency (8-5 MHz) transducer. Three months after the first visit the subject was visited and examined again. This time it was possible to take a blood sample with anticoagulant which was processed using an automatic cell counter (Abbot ''Cell-Dyn 3500'') to obtain the hemogram. During the first observation, clinical examination led us to identify dyspnea and cyanosis which was particularly evident on the nose, ears and mucosas. The intensity of the dyspnea and cyanosis increased as soon as the subject became more excited because of the inevitable handling during the visit. It was possible to identify a weak and frequent pulse (240/1), and a diffuse thrill, synchronous with the cardiac apex beat. A systolic harsh murmur, of grade 4-5/6, best heard in the pulmonic valve area and also on the right side of the chest (third intercostal space) was also evident. Clinical findings were almost the same during the second visit, although the systolic murmur was less loud (grade 3-4/6). The electrocardiogram showed a right ventricular enlargement and the radiography showed a slight reduction of the left atrium volume. Mono-and bi-dimensional echocardiography showed a large ventricular septal defect (Fig. 1) , a dextroposition of the aorta (''overriding aorta'') ( Fig. 2) producing stenosis of the right ventricle outflow tract (pulmonic stenosis) and, lastly, a hypertrophy of the interventricular septum and of the right ventricle wall. Color Doppler examination showed a turbulent blood flow in the pulmonary artery and through the ventricular septal defect with a right-to-left shunt. Hemogram results showed polycythemia (HCT=70.4%, RBC=16.8×106/mm3, HGB=23.6 g/dL) as compared to the normal limits (Feldman et al., 2000) . The congenital cardiopathy suspected during physical examination was confirmed by instrumental and laboratory tests. The test that appeared to be most useful for confirmation of the diagnosis of tetralogy of Fallot was echocardiography. In fact ultrasonography gave a correct intra vitam diagnosis, despite the fact that the probes normally used in veterinary medicine are not very adaptable to small acoustic windows. The fact that the murmur was also heard on the right side of the chest is probably due to the radiation of the murmur caused by pulmonic stenosis and/or to the blood flow through the ventricular septal defect or through the ascending tract of the ''overriding aorta''. This kind of murmur is less loud when the ventricular septal defect is so large that turbulences are absent. Another reason is when haematic hyperviscosity, due to polycythemia consequent to persistent hypoxia, is also present (Kirby and Gillick, 1974) , like in our case. During the second half of last century, the wolf (Canis lupus Linnaeus, 1758) was an endangered species in many European countries. In recent decades specific protection programs have resulted in a progressive increase of live wolves in different countries, Italy included. Accordingly, many scientific studies have been carried out to investigate different aspects of wolf ecology. The vast majority of veterinary studies on this wild canid are mainly related to the anaesthetic protocols for its capture and restraint, as well as to its susceptibility to infectious and parasitic diseases. Little is known regarding the prevalence of non-infectious diseases in the wolf and, among them, of cardiovascular disorders. In the present paper, the clinical, echocardiographic, and echo-Doppler findings of mitral regurgitation in a captive arctic wolf are described. A 10-year-old, captive male arctic wolf, housed at the National Wildlife Refuge of Popoli, in central Italy, was examined during the annually scheduled health check. The subject was immobilized with an intramuscular injection of xilazine and ketamine. Anaesthesia was maintained with isofluorane in oxygen after endotracheal intubation. A physical examination, including cardiac auscultation, and blood sampling for complete blood count (CBC) and biochemical profile were carried out. The echocardiographic examination (real-time B-mode and M-mode) and the echo-Doppler examination were conducted using a portable ultrasound machine equipped with a 3.5 MHz mechanical sector transducer and Doppler frequency of 2.5 MHz, with simultaneous electrocardiograms. The standard echocardiographic scan planes described in domestic carnivores (Thomas et al., 1993) were employed. Measurements of the cardiac chambers and walls, as well as spectral cardiac Doppler flow profiles were obtained. 348 On physical examination, a loud systolic murmur (IV/VI grade) heard best over the mitral area was appreciable. Results of CBC and the serum biochemical profile were normal. Two-dimensional echocardiographic examination revealed a mild thickening of the mitral valve leaflets (Fig. 1 ) associated with a mildly dilated left atrium (left atrial diameter/aortic diameter ratio=1.8) (Fig. 2) . On echo-Doppler examination, a systolic regurgitant flow across the mitral valve was noted. No other echocardiographic and echo-Doppler abnormalities were found. Thus, mitral regurgitation, most likely due to mitral endocardiosis (ME), was diagnosed. Mitral regurgitation is the most common and clinically important valvular disorder in the domestic dog (Bonagura and Luis Fuentes, 2000) . Furthermore, mitral degeneration (endocardiosis) is the most common cardiac disease in dogs (Buchanan, 1992) . Aged small and middle-size canine breeds are usually affected (Thrusfield et al., 1985; Whitney, 1974) , although ME is also diagnosed in large breed dogs (De Madron, 1992; De Madron, 1998) . Mild thickening of the mitral valve leaflets, without other appreciable echocardiographic abnormalities, and absence of clinical and clinical pathology evidence of an inflammatory process, strongly suggested chronic degenerative mitral valve disease as the inciting cause of mitral regurgitation in the present case. Mild left atrial enlargement (Rishniw and Erb, 2000) was the only cardiac consequence of mitral insufficiency for the wolf in the present report. Mitral regurgitation was also found in three out of seven aged Italian grey wolves undergoing a diagnostic protocol similar to that of the arctic wolf described in the present paper (Guglielmini et al., 2003) . To the best of the author's knowledge, this is the first report of the in vivo diagnosis of mitral insufficiency in a wolf. Mitral endocardiosis seems to be a relatively common cardiac disorder in the aged wolf, as for the domestic dog. The green Iguana is one of the most popular reptile pets, so it is very important to find anaesthetic protocols useful for the anaesthetization of these potentially violent patients which are also particularly sensitive to handling stress. According to the bibliography (Addis et al., 1990; Bennet et al., 1998; Barten, 1993; Heard et al., 2002; Mazzi, 2000; Nevarez et al., 2002) reptiles respond to anaesthetic drugs differently from mammals and birds. Injectable agents have long induction times, poor muscle relaxation and rough recovery. Inhalation anaesthesia is not always useful because lots of species can hold their breath even for several minutes or hours. Six green iguanas (Iguana iguana) of different sexes, ages and weights (mean: 1233.3±251.6 g; range: 1000-1500 g) received a dorsal i.m. injection of a commercial combination of tiletamine/zolazepam (Z), dissolved in the standard way. Four animals were intubated because diagnostic manipulations or surgery were too long. They inhaled Isoflurane/O 2 by means of a T-Ayre. For endotracheal intubation a shortened defluxion tube made from a ''Butterfly'' was used and a cat urine catheter was placed in it so it was stiffer to slide. Each iguana was carefully weighed. Posologies were reduced in the event of serious wasting or dehydration. All animals were checked during anaesthesia and recovery. Heart and respiratory rates were recorded 5 min from inoculation and then at 10 minute intervals throughout the anaesthetic period. These rates were noted at the end of Isoflurane/O 2 use and 10 min 352 after extubation. The time taken to reach the surgical plane of anaesthesia and the adverse effects during anaesthesia and awakening were recorded. Statistical analysis was performed (mean and range of variation). Posology -Each iguana received i.m. 10.53±4.2 mg/kg of Z and immediately all animals became excited if not restrained, then anaesthetic effects started and their muscles relaxed a few minutes later. The iguana placed itself in sternal decubitus recumbency then the reptile could be easily handled. Induction time was defined as no response pressing between digits (mean: 6.5±0.7 min) and anaesthesia was considered as achieved when were no righting or toe reflexes. With Isoflurane/O 2 the patients lost their palpebral reflexes after 5 min. Intubation -Good relaxation of the jaw muscle, even without application of local anaesthetic to the surface of the glottis, facilitated intubation. Monitoring -Cardiac frequency increased slightly in these animals (Addis et al., 1990; Bennet et al., 1998) ; both patients that only received Z and those who received Isoflurane/O 2 had 100 beats/min on average. Respiratory frequencies decreased, their mean values were 10 breaths/min in the first case and 6 in the second one. Anaesthetic concentration: mean 1.8±0.4% of Isoflurane. Awakening -The endotracheal tube was removed 12 min from the end of Isoflurane administration. Recovery from anaesthesia was longer than 45 min for some of these patients. There were no particular side effects. Cardiolipin (CL) is an acidic phospholipid, first discovered in beef heart tissue but not specific to the heart. It is found both in bacteria and mitochondria of mammals. Mitochondria contain CL as an essential component of the intricate structure of the mitochondrial cristae membrane. CL is more abundant in heart tissue only because of the high concentration of mitochondria in cardiac myocytes. In recent years, anti-CL antibodies (ACA) have been extensively studied in human medicine and their role in antiphospholipid syndrome (APS), a potentially life-threatening autoimmune coagulation disorder, has been well established. Clinical features that should alert the physician to consider APS include recurrent foetal loss, arterial or venous thrombosis, thrombocytopenia and high positivity to the ACA test (Harris et al., 1999) . However, searching in the literature, we found a lack of investigation on ACA in veterinary medicine. Therefore, we aimed to carry out a preliminary research to verify the presence of IgG ACA levels in dog serum. In this preliminary study we tested IgG ACA levels in healthy and diseased dogs. All the dogs tested were randomly selected among those referred to the clinic of our veterinary faculty. There were 116 males and 81 females, ranging from 2 months to 14 years of age, and of different breeds. 134 of them were healthy and were referred to our clinic for regular vaccinations. These animals were used to determine the reference range values. 63 were referred because of a variety of pathologies. These were used to determine IgG ACA levels for diseased dogs. For our purpose, the ELISA technique was used as routinely performed for diagnosis of APS in humans (Tincani et al., 1986) . Commercially available antidog IgG conjugated with alkaline-phosphatase 356 (Sigma-Aldrich, S.r.l., Milano, Italy) were used to partially modify the technique for dogs. Optical density was read using a spectrophotometer at 450 nm and was expressed as mean±standard deviation of IgG/ml units (GPL). ACA levels were present for our group of healthy dogs, with a mean value of 5.40±2.66. A high prevalence of ACA has also been documented for populations of healthy adults and children (Cabiedes et al., 2002; Mannoussakis et al., 1987) . Moreover, based on the different levels obtained, dogs with various pathologies could be divided into five groups corresponding to the human model (Harris et al., 1987) . Diagnosis of APS in humans requires the presence of both clinical and laboratory abnormalities. It can be made when the presence of an elevated ACA level occurs in addition to a thrombotic event associated with vascular thrombosis, thrombocytopenia, and recurrent foetal loss, or with haemolytic anaemia, livedo reticularis, and neurologic disorders Olmstead, 2001; Quintero del Rio, 2002) . It is worth noting that at least two of the four dogs in group V had clinical manifestations, history and laboratory abnormalities which would be suggestive of APS for humans. An eight-year-old female boxer was referred because of spondylarthrosis with a history of infertility and recurrent foetal loss; laboratory abnormalities showed serious thrombocytopenia (platelet count 101×103/ml; reference range 240-400×103/ml) as well as a very high ACA level. A three-year-old male crossbreed was referred because of progressive blindness and retinal thrombosis was diagnosed, together with a very high ACA level. Therefore, our report shows that ACA are present in both healthy and diseased dogs, can be easily detected using the ELISA technique, and can reach very high levels, suggesting a possible role in the pathogenesis of some disorders of dogs. However, at present, it is not possible to affirm that APS can exist in dogs. The present findings are too preliminary and still too limited to allow such a conclusion. Further investigation is certainly needed to clarify the role, if any, of ACA in dogs. Interest has recently been increasing as to the role cytokines play in modulating the immune response. However, in the veterinary field, this line of research is still in its infancy. Few studies have been made on the quantification of the cytokines present in dairy milk and these have been conducted only on non-mastitic animals (Leutenegger et al., 2000) . The evaluation of the cytokine profiles in dairy milk has previously been achieved using the reverse transcriptase-polymerase chain reaction (RT-PCR) method, both for healthy and mastitic cows (Peli et al., 2003) . In this study, cytokine quantification in dairy milk was determined using the relatively newly developed realtime TaqManA reverse transcriptase polymerase chain reaction (real-time TaqManA RT-PCR) method. A comparison was made between healthy cows and animals affected by mastitis, the latter having a high (>5.0×105) or a low (<1.5×105) number of milk somatic cells as measured by the Somatic Cell Count (SCC) test. All milk samples were analysed for the presence of three cytokines, chosen because of their key role in the immunological regulatory process: IL-8 (chemotactic activity), TNF-a (pro-inflammatory activity) and IL-10 (regulatory/anti-inflammatory activity). Milk was collected from 15 Italian Friesian dairy cows at mid-lactation. Two 50 ml aliquots of milk were manually and aseptically collected from each udder quarter. One aliquot was used for the standardised bacteriology and SCC procedures and the second was used to extract the mRNA coding for the three cytokines TNF-a, IL-8 and IL-10. For each animal only the milk sampled from one of the udder quarters was used, which was selected on the basis of the bacteriological and SCC test results. All milk cells were concentrated by centrifugation at 700G for 10 min at 4°C. The supernatant and the fat layer were discarded and the resulting pellet was washed twice in cold PBS (pH 7.4), and then re-suspended in 250-500 ml of PBS. The mRNA was extracted from a 15 ml starting volume of each milk sample, using the Micro Fast-Track mRNA isolation kit (Invitrogen), and re-suspended in 50 ml of nucleasefree H 2 O (Promega). All the extracted mRNAs were kept at −80°C until use. To quantify the expression of the cytokines coding for mRNA in the samples, the realtime TaqManA PCR method, capable of detecting less than 1 mg of mRNA, was used. A housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), was simultaneously processed in Real-time TaqManA PCR for each sample, as an endogenous control. This was done to normalise the process for the variability in the number of cells in the original samples, for differences in extraction efficiency, for mRNA degradation in the starting material, and for reverse transcriptase reaction efficiency. The primers and the probe used for the detection and quantification of the IL-10 cytokine were designed by us using sequence data available in GenBankA, and were achieved with the aid of the Primer Express software (PE Applied Biosystems). The forward and reverse primers are located in two adjacent exons, while the probe lies in-between them to hybridise at the junction of the two exons. The primers and probes for the IL-8 and TNF-a cytokines, and for the endogenous control (GAPDH), were obtained from the literature (Leutenegger et al., 2000) . Each TaqManA probe was labelled at the 5∞ end with fluorochrome FAM (6-carboxyfluorescein) and with fluorochrome TAMRA (6-carboxytetramethyl-rodamine) at the 3∞ end (PE Applied Biosystems). The Real-time PCR and reverse transcription of the mRNA target was performed using a ''one-step'' reaction method, in a final volume of 25 ml. The primers were used at a concentration of 400 nM each, while the probe was used at a concentration of 80 nM, in the reaction buffer (One-Step PCR MasterMix; PE Applied Biosystems), which also contained the reverse transcriptase enzyme and the AmpliT aqGoldA enzyme. Finally, 1 ml of extracted mRNA was added to the mixture. The reverse transcription, and real-time TaqManA PCR, was executed in the ABI PrismA 7700 Sequence Detection System (PE Applied Biosystems), and was done as follows: a 30 min cycle at 48°C, a 10 min cycle at 95°C, then 50 alternating cycles each of 15 s at 95°C, and of 60 s at 60°C. Cytokine quantification was achieved using the Ct (Cycle threshold) comparative method, and is expressed as ''n-fold upregulation of cytokine transcription'' in relation to a calibrator which is represented by the smallest signal detectable for that specific cytokine. Based upon the results obtained from the bacteriological and SCC tests the Friesians fell into three groups of five animals each: Group A showing an average SCC of 361 4.097×103, and the presence of Corynebacterium sp., Staphylococcus sp., and Streptococcus sp. in two, one and two animals, respectively; Group B, showing an average SCC of 74×103 and the presence of E. coli, and Streptococcus sp. in two and three animals respectively; Group C, showing an average SCC of 59×103, and no bacteria detected in the milk. The analytical method was able to quantify all the cytokines in all 15 samples examined, and thus confirms the utility of this method for the study of cytokine gene expression in milk. The results, expressed as group mean values of the three cytokines analysed, are shown in Fig. 1 . The levels of TNF-a were significantly higher in animals with positive bacteriology (Groups A and B) than in Group C (negative bacteriology), and are not linked to SCC. This result confirms that a bacterial infection, independent of any chemotactic effect, stimulates an increase in Figure 1 . Mean values of the three cytokines analysed in milk cells from groups A, B and C. Cytokine transcription is expressed as the n-fold differences to the calibrator, which was the lowest detectable cytokine signal. Note the semilogarithmic scale 362 the production of TNF-a (which causes a local inflammatory response). To the contrary, in samples with high SCC (Group A), levels of IL-8 were higher than seen in the low SCC Groups B and C. This confirms the chemotactic role (mostly on the neutrophils) of this cytokine the synthesis of which does not seem to be induced by the presence of bacteria, as is the case for TNF-a, and as deduced from the lower levels of this cytokine detected for Group B (Barber and Yang, 1998) . These data confirm recent findings which show that an increase of TNF-a is not necessary to stimulate the synthesis of IL-8 in the mammary gland (Personn et al., 2003) . As regards IL-10, the results seem to show a correlation with SCC, but the lack of published data on the precise role and origin of this cytokine in the ruminant mammary gland indicates the need for further studies for the explanation of our results. Much research remains to be done on the interactive role of these bacteria and their influence on the quantitative profiles of cytokines in the ruminant udder. The cat is an unusual host of L eishmania spp. and only a few cases of leishmaniasis have been reported for this species, mostly in areas where the microorganism and several species of phlebotomine sandflies are endemic (Pennisi, 2002) . About twenty clinical cases have been reported in the literature, sometimes described only briefly with regard to both the clinical course and the therapy employed. Because of this, we wish to describe the clinical findings of four cases of feline leishmaniasis observed between December 1997 and May 2001. Four domestic cats were involved (three short-haired, one long-haired); two males (cases 2 and 3) and two females (cases 1 and 4). The cats were six (case 3), 10 (case 4) and 14 (case 1) years of age, in case 2 the age was unknown but the subject was an adult cat. All cats came from the Messina district and they were free-roaming. Diagnosis was achieved by microscopic (smears from cutaneous ulcers and cysts, and from lymph node needle-aspiration) and serological investigations (IFAT) with anti-L eishmania antibody titres ranging from 640 (cases 1, 2, and 4) to 1280 (case 3). Further confirmation came from PCR and isolation of L eishmania spp. in cases 2, 3 and 4. The cats were tested for FIV and FeLV (Duospeed, BVT, La Seyne sur Mer, France), T oxoplasma gondii (IFAT), Bartonella henselae (IFAT) and given blood tests (complete blood cell count, biochemistry profile, serum protein electrophoresis, urinalysis). Two cats did not undergo specific treatment (cases 1 and 3) because the owner The owner opposed any further investigation and asked for euthanasia. The cutaneous lesions, lymph node enlargement, serum protein electrophoresis and antibody titre did not change over the course of the disease, and when the cat was put down the anti-L eishmania titre was 640, with positive lymph node needle-aspiration (PCR and microscopy). Case 4 (treated with allopurinol, erythropoietin, ferrous sulphate and rHuIFNa-2a) showed a rapid improvement with recovery in appetite and vitality within the first two weeks, while a dramatic improvement of the severe haematological data was seen after three weeks. Five months later a weight gain (3 kg) was observed, along with the regrowth of hair, the liver and lymph nodes enlargements disappeared, and the hypergammaglobulinemia and anti-L eishmania antibody titre were reduced. The drug regimen was modified and the cat was given allopurinol for ten further months. At the end of the treatment (15 months in all) L eishmania serology and PCR were negative, while neutropenia and hypergammaglobulinemia were still present. Two months later the cat exhibited anaemia, leucopenia, hypergammaglobulinemia and positive antibody (titre 80) once again. In the following months the uveitis recurred, anaemia worsened and the IFA titre rose to 1280. Parasites were isolated once again from the submandibular lymph node. At this time allopurinol (20 mg/kg SID) was recommenced, and the treatment remains ongoing. For the first time feline leishmaniasis has been reported in FIV+ cats (three cases out of four) and treated with oral drugs, which generally meet with greater compliance from the owners, especially for long-term therapy. Clinical signs of feline leishmaniasis are quite similar to those observed for dogs: lymph node enlargement, cutaneous ulcers, alopecia, weight loss, pale mucous membranes, hypergammaglobulinemia and CRF. On the contrary, cutaneous bloody cysts (with evidence of amastigotes inside) have not been reported for canine leishmaniasis. The ocular lesions cannot be considered specific because of the concurrent FIV infection (cases 1, 3 and 4) and a high anti-T oxoplasma titre (case 4). Despite the fact that the literature reports the efficacy of fluconazole (Gangneux et al., 1999 ), itraconazole (Gangneux et al., 1999 and spiramycin+metronidazole (Gangneux et al., 1999; Pennisi et al., 2001) , these drugs did not work for the FIV− cat (case 2) which developed CRF. The same happened in case 3 which, in spite of the FIV infection and the lack of therapy, lived for five more years after diagnosis. Allopurinol was given to cats for the first time and it was well tolerated in case 4 which had a symptomatic (stage IV) FIV infection. The allopurinol treatment was clinically efficient but, as reported for dogs (Baneth, 2002) , did not succeed in eradicating the infection. 366 Blood pressure can be evaluated either by using a direct invasive method (arterial puncture) or an indirect noninvasive method (oscillometric or doppler measurement). However, both methods are rarely used in practice: arterial puncture is potentially dangerous and hard to achieve in a non-sedated patient, while the noninvasive techniques are not widely used because of the lack of materials and methods effective enough to obtain proper measurements (Bodey et al., 1996; Edwards, 1992) . Because of the few references in the bibliography about arterial blood pressure measurement we have focused our attention on the oscillometric technique as it is the easiest to perform in practice and the only method that can be used to assess the diastolic value. Arterial blood pressure was assessed using an oscillometric technique in dogs (49) and horses (33) referred to the Section of Internal Medicine of the Department of Pathology, Diagnostic and Veterinary Clinic of the University of Perugia for clinical evaluation. All horses were clinically normal, while dogs were divided into a control group of healthy dogs and a group of patients with diseases that could affect arterial blood pressure. Measurements were performed using MemoprintTM, an electronic device product by Medvet, that assesses the diastolic and systolic blood pressure and the mean heart rate. Four cuffs of different size were available, two for large and medium size dogs, one for small dogs and cats and one for horses. For the dog measurements were performed with the patient in sternal or lateral recumbence with the cuff constricting the distal third of the left forelimb, attempting to maintain the cuff on the same horizontal plane as the heart in order to evaluate the radial artery pulse. At least five measurements were performed for each dog and the weighted mean was then calculated. For horses the cuff was placed around the base of the tail (coccigeal artery) and three measurements and the arithmetical mean were recorded. We divided the animals into 9 groups, ordered as follows: Gr A: two-year-old foals, 8 males and 7 females; Gr B: 10 Standardbreds, 6 male, 1 neutered male and 3 females; Gr C: 8 non-athletic horses, 4 castrated males and 4 females; Gr D: 11 clinically healthy dogs, 8 males and 3 females; Gr E: 10 dogs affected by chronic renal failure, 5 males and 5 females; Gr F: 6 dogs affected by diabetes, 2 males and 4 females; Gr G: 7 dogs with heartworm disease, 4 males and 3 females classes II-III; Gr H: 9 dogs with chronic mitral regurgitation, 5 males and 4 females, 7 in treatment for congestive heart failure; Gr I: 6 dogs affected by atrial fibrillation, 5 males and 1 female, 2 undergoing treatment for arrhythmia. The statistical significance of the differences between the groups was evaluated using the Mann-Whitney U test. The repeatability of measurement was evaluated by calculating the coefficient of intraoperator variation. We assumed as acceptable a variability of less than 10%. The repeatability of measurement, expressed in mmHg, was high for the horse (DS= 3.65 and 2.60 for systolic and diastolic pressure respectively with mean values of 113 and 72) and good in the dog (DS=11.15 and 9.01 with mean values of 143 and 79). The pressure measured for different groups of horses was similar (mean systolic pressure=113 and diastolic=72). This is interesting if we consider that in human medicine it is well known that athletes' blood pressure is lower than that of normal people and that this decrease in mean arterial pressure can be used as an index of performance (Ellestad, 1987) . By the statistical analysis of our results we conclude, on the contrary, that athletic training is not able to modify the arterial pressure for horses; therefore the arterial pressure cannot be used as a performance index. In normal dogs the mean systolic pressure was 143 mmHg and the diastolic pressure was 79 mmHg. These values are not so different from those obtained in other studies using the direct technique (Mishina et al., 1997) . Also for the dog we did not find any difference between males and females as described in the literature of human and veterinary medicine (Mishina et al., 1997) . Negative and positive differences, and their significance, between pathological groups and the control group Gr D are noted (n.s.=not significative) in Table I . Arterial blood pressures, both systolic and diastolic, were statistically higher for dogs affected by chronic renal failure and dirofilariasis than control dogs, while for diabetic dogs only the diastolic values were higher. No differences were evident between dogs with cardiac disease, whether undergoing treatment or not, and normal dogs. Our study highlights the clinical usefulness and applicability of blood pressure INTRODUCTION Ochratoxin A is a naturally occurring mycotoxin produced by Aspergillus ocraceus, mainly in tropical regions, and Penicillum verrucosum mainly in temperate areas. Ochratoxin A consists of a polyketide-derived dihydroxy-coumarin moiety linked through the 12-carboxy group to phenylalanine (Beretta, 1998) . Research in Ochratoxin A (abbreviated as OTA) has increased since 1994, when The International Agency for Research on Cancer classified OTA as a possible human carcinogen, based on sufficient evidence of carcinogenicity in experimental animal studies (IARC, 1994) . P. verrucosum grows in cereals in cold climates and, as a result, the presence of OTA is larger in Scandinavia and other cool temperate zone areas of Central Europe. It has been suggested that 50% of OTA intake can be attributed to cereal and cereal products. OTA has a potent toxicity and its nephrotoxic and carcinogenic effects in all mammalian species have been demonstrated. OTA's toxic action seems to be more dangerous than that of Aflatoxin (Muller, 2003) . Daily intakes of 1 mg/kg BW for 5-6 days can be lethal. OTA mainly affects protein synthesis and subsequently DNA and RNA synthesis. It is a powerful competitive inhibitor on phenylalanine-tRNA ligase, reducing protein synthesis as a result of its inhibitory action. It seems certain that the supply of phenylalanine can reduce OTA's toxicity. The interest in OTA in zootechny regards only monogastric species as ruminants can convert Ochratoxin A into Ochratoxin a which is less toxic (Abarca, 2003) . Pigs and chickens are the animals most exposed to the toxicity of OTA. It is well known that OTA plays a major role in the aetiology of nephritis in pigs, causing serious kidney lesions, while 372 in poultry OTA can cause neuronal anomalies. In addition, daily intake of OTA can reduce Feed Conversion Rate with subsequent lower growth, and it results in a lower efficiency of the immune system; as a result of kidney damage the animals increase their water consumption. The target organ of toxicity in swine is the kidney, but also the liver which has function of detoxification (Beretta, 1998) . The aim of the present work was to evaluate the effects of OTA in the diet of growing pigs, concerning liver damage. These effects were evaluated using proteomic methods, monitoring the effects of the administration of OTA on the soluble protein fraction of the liver. The proteomic tools used were high-definition two dimensional electrophoresis coupled with MALDI-TOF mass spectrometry. 64 Large White x Landrace pigs, half of which were castrated males and half females, from 8 litters, with a mean live weight of 41 kg, were divided into 2 groups of 32 animals each and housed in pens each of which held 8 animals. The control group was given a commercial pig feedstuff (Control diet). The treated group was fed the same feedstuff with the addition of 25 mg/kg Ochratoxin A from Aspergillus ochraceus (OTA 25 mg/kg), obtained from Sigma-Aldrich s.r.l. Milano. The diets were provided ad libitum and the animals had free access to drinking water. The pigs were individually weighed at the start of the trial and after 119 days, before slaughtering. Data were recorded regarding Live Body Weight (BW) and Average Daily Weight Gain (ADG). Feed intakes of each pen were recorded at the end of the trial to calculate the Feed Efficiency. The liver of four pigs from each group was collected at slaughtering, immediately stored at −20°C and then submitted to proteomic investigation. 100 mg of liver tissue were homogenised and the soluble proteins were extracted in 1 ml of lysis buffer: 9.5 M urea, 2% CHAPS, 0.8% Pharmalyte 3-10, 1% DTT and 2 mM PMSF in an ice bath (Rabilloud, 2000) . The solution was then centrifuged (18000 g) and filtered. The samples contained about 10 mg/ml of proteins and the quantity utilized for electrophoresis was about 1 mg. The IPG strips (home made; range 3-10 and 4-7) were reswelled for 6 h (active: sample was directly put into the buffer), 3 h at 30 V and 3 h at 50 V. The total volume of rehydration was 400 mL, 100 mL of sample for each strip. The run was carried out overnight under a layer of paraffin oil at 15°C, under a voltage gradient up to 8000 V (IPG-phor, Amersham Biosciences). After equilibration steps with DTT and IAA, the strips were put onto second-dimension gel, which was a classical SDS-PAGE, 10%T. The analytical staining was performed using silver staining protocol, while the preparative staining for the identification of proteins in mass spectrometry was coomassie blue. The spots of interest were excised with a cutter, de-stained with an acetonitrile/ammonia bicarbonate solution and dehydrated with acetonitrile, then incubation was carried out overnight with trypsin at 37°C. The reaction was stopped at 4°C and the tryptic digests were analyzed using a MALDI-TOF spectrometer (Tof-Spec 2E, Micromass, Manchester, UK). The comparison of digests was performed in a data bank using a free-on line software, ProFound (www.proteometrics.com). Figures 1 and 2 show two dimensional maps of the soluble protein fraction (liver) of swine (control, left) and added OTA in the diet (right). The matching of the electrophoretograms indicates that there is a greater expression of the a and b chains of haemoglobin in swine undergoing OTA treatment which is probably due to a haemorrhagic status of the liver. In the animals treated with OTA there is an increased expression of different enzymes with related isoforms: carboxypeptidase A, catalase, fumarate-hydratase, phosphatidyl ethanolamine binding protein and inducible NO synthase. These are hepatic enzymes that are indicators of hepatic and energetic stress. This investigation shows some important differences between the proteins expressed by hepatic tissue under stress conditions; such proteins are not identified with routine analysis. Proteomic investigations are a useful tool in the identification of diagnostic markers. Equine Cushing-like syndrome or Hyperadrenocorticism is caused by excessive secretion of ACTH by the pituitary gland, secondary bilateral adrenocortical hyperplasia and associated hypercortisolemia. Equine Cushing-like syndrome is due to hyperplasia or adenoma formation of the pituitary pars intermedia. A possible cause is the reduction of negative regulation of dopamine on ACTH secretion because of a hypothalamic disease (Dybdal, 1997) . Considering this etiopathogenesis a more appropriate description of the disease is probably Pituitary Pars Intermedia Dysfunction (PPID) (Schott, 1997) . PPID is relatively frequent in old horses (>18 years) with no apparent sex predisposition although authors have suggested that females are afflicted more than males (Boujon et al., 1993; Heinrichs, 1990) . The characteristic clinical signs include: weight loss secondary to muscle wasting, exercise intolerance, pendulous abdomen, hirsutism and curly hair coat, polyuria and polydipsia, hyperhydrosis, delayed wound healing (Thompson et al., 1995) and bulging supraorbital fat pads as a result of fat redistribution. Less frequent symptoms are skin infections, urinary tract infections, gingivitis, periodontal infections, bronchopneumonia and chronic recurrent laminitis (Hillyer et al. 1992; Reed, 1998) . Persistent hypercortisolism induces insulin resistance and associated glucose intolerance, thus hyperglycemia and hyperinsulinemia are possible (Beech e Garcia, 1985; Beech, 1987; Reed, 1998) . Furthermore, infertility seems to be a common consequence of releasing inhibition of gonadothropic hormones (Love, 1993) . 18-years old, mare, Italian saddle horse. Veterinary case histories exist to document the condition of the animal at the time of the visit and during the course of any illness; they furthermore constitute a particularly useful element in the case of subsequent or periodic controls, allowing the veterinary doctor to access data from previous clinical notes, and on the basis of these to be able to make more useful diagnoses. Since this official record is able to provide an accurate picture of the clinical situation of the animal, the choices made by the doctor and the evolution of the illness, on one hand it constitutes an important source of information for historical health research and the aims of scientific research and statistical analysis, and on the other hand allows any parties suffering damage to reconstruct and evaluate the behaviour of the veterinary doctor and ascertain any professional responsibility. In order to fulfil these tasks correctly, it is considered that case histories must contain at least the following essential information, naturally in addition to the identification data concerning the structure and the veterinary doctor responsible for the case: a) general details of the animal's owner; b) species, race, sex and age of the animal; c) reason for the visit or hospitalisation; d) anamnesis; e) clinical inspection and general controls carried out, and their results; f ) collateral exams carried out; g) diagnosis; h) prognosis; i) therapy; j) course of the illness. From a regulative point of view, there are no specific instructions on how to keep veterinary case histories. Although concise, some guidelines can however be found in the instructions governing human medicine. In particular, article 35 of the Italian Ministerial Law (D.M.) of the 27th of June 1986 -determining the requirements for private clinics -sets out that a case history must be completed for every patient, from which the following should be documented: ''complete personal data, diagnosis on arrival, family and personal anamnesis, objective examination, laboratory and specialist examinations, diagnosis, therapy, results and after-effects''. This law also foresees that ''... the regular updating of the case history during hospitalisation is the responsibility of the chief consultant or head of service, while the filing and issue of copies to all those who are entitled to receive them is the responsibility of the medical director''. Clearly the above regulation is designed for the management of case histories in the field of human medicine and, more particularly, in the case of private clinics; however the rules contained in the law also appear to be a useful model for use in veterinary practice, providing helpful indications for correct and suitable behaviour. In law, the function of case histories is described as follows: they are ''a diary of the course of an illness and other relevant clinical factors''(1) and ''certify the therapy undertaken'' (2) . More particularly, it states that such records ''confirm their function as a diary, and the facts must be written down as they happen'' (3) . Concerning the content, it is specified that ''in the records the diagnosis, progression of the illness and therapies administered on each occasion are noted''(4), ''as well as any operations carried out, the results of therapy and the length of the patient's stay in hospital'' (5) . It follows that, above all in university clinics and veterinary hospitals, where it is almost inevitable that such notes are made by more than one person, it is good practice that each note be initialled by the person making it. Concerning the legal nature of case histories, and in particular whether they should be considered a public or private document, and whether their contents should be given a privileged legal validity as evidence or not, contrasting views can be found in existing legal doctrine or, in any case, views that do not help the operator to clarify his ideas on this issue. Concerning this matter Piccini (1993) states: ''The legal nature of case histories has not yet been firmly defined, meaning whether they should be considered a real public deed that can be used as privileged evidence, or as mere hospital documents with technical and medical relevance, or as documents that can be compared to administrative certification'' (6) . Pugliatti (1999) , on the other hand, more peremptorily states that ''the value of the case history as an official record is indisputable''(7). We feel that the issue of the legal nature of case histories and the legal validity of their contents as evidence is a very delicate matter, and one which should be handled starting from the general meaning of the term oYcial records, and the value of these records as evidence. Article 2699 of the Italian Civil Code defines an official record as ''the document drawn up in the required formal manner, by a notary or other Public Official (P.O.) authorised to officially certify the place in which the document was made''. The following article (2700) then specifies the validity of the document, stating that the ''official record constitutes full evidence unless subject to a summons challenging its origin by the P.O who has drawn it up, as well as the declarations from the 383 parties and other events that the P.O. declares to have taken place in his presence, or to have been carried out by him''. Of particular interest to this study, an official record is deemed to be any document drawn up by the P.O. for any purposes inherent to his functions and this even includes purely internal records drawn up with the aim of documenting matters relating to the activities he carries out and the regularity of the administrative operations for which he is responsible (8). As far as clinical records are concerned, the Supreme Court has deemed that they ''must be considered an official record, as they constitute the execution of certificatory power and are part of the public nature of the medical activity to which they refer, not only the case histories kept by a public hospital but also, by virtue of the delegation of public functions conferred upon private institutes by the national health service, records kept by a private clinic that has an arrangement with the national service (specifically on the subject of ideological forgery of documents committed by doctors in medical records obtained from private clinics)'' (9) . Having therefore established, albeit briefly, what should be considered an official record, and when a veterinary case history can be considered official, we now need to examine how, and which parts of, an official record (and more particularly a case history) can be considered privileged evidence (as established in the mentioned article 2700 of the Italian Civil Code), meaning that it takes precedence over other sources of evidence. On this matter, legal doctrine deems that privileged evidence cannot concern the whole of an official record, but needs to be reserved and limited to only some of its parts, and more specifically: a) to the declarations by the parties and other events that the P.O. declares to have taken place in his presence; b) to the actions that the P.O. declares to have carried out personally. As far as the remaining parts are concerned, opinions (or judgements) expressed by the P.O. in the document, these cannot be considered privileged evidence and therefore these declarations may be proved to the contrary (10). By applying the general concepts outlined above on this subject, we may conclude, in the context of case histories drawn up for example in a university clinic, that privileged evidence may be considered the date and the checks carried out, but not the diagnosis, prognosis and any other matter conditioned by the doctor's own evaluation. The Supreme Court has in fact clarified that ''case histories drawn by a doctor in a public hospital cannot, for this matter alone, completely be considered an official document with validity as privileged evidence, as this particular validity as evidence must be limited to their certification by the P.O. and to the events the P.O. declares to have taken place in his presence or to actions that have been carried out by him''(11). Together with tropomyosin, the troponin complex participates in the contraction of skeletal and cardiac muscles. Troponin is composed of three subunits, namely troponin C (TnC), troponin I (TnI), and troponin T (TnT), which binds tropomyosin. There are three different isoforms of TnT and TnI, two of which are specific for skeletal muscle, while the third is specific for the myocardium (cTnT and cTnI). As the homology between troponins is about 95% among mammals, commercial diagnostic kits designed for use in humans also provide excellent results in other animals. . In human medicine, the measurement of cTnT or cTnI levels is now used together with measurement of the levels of creatine kinase (CK) and lactate dehydrogenase (LDH) and their isozymes for assessment of myocardial damage. For diagnostic purposes, one advantage is that these proteins are normally absent in blood in subjects without myocardial damage. Even minimal increases in the serum levels of these proteins are indicative of myocardial damage and, in the case of cTnT, it is detectable in serum from 2 h to 14 days following the onset of myocardial damage. Moreover, in contrast to the isozymes of CK and LDH, the results are not influenced by use of different clinical samples (hemolytic, lipemic, bile samples). The degree of increase in the level of cTnT is relatively proportional to the degree of myocardial damage, with prognostic implications as well. At the same time, the extreme sensitivity of this technique also allows for its use in other nonischemic pathologies including chronic cardiac insufficiency, pulmonary emboli, acute myocarditis, systemic hypertension, arrhythmias, and toxicity from doxorubicin (De Francesco, 2002) . In veterinary medicine, O'Brien et al. (1997) have demonstrated the utility of measuring the levels of cTnT in the diagnosis of experimentally induced myocardial damage. cTnT levels were determined using an immunochemical test developed for humans using whole, heparinized venal blood (Cardiac Reader, Roche) within 2 h of blood collection. Values of cTnT less than 0.05 were considered negative (expressed in ng/ml). Values of cTnT between 0.05-0.1, 0.1-2, and >2 were considered as low, intermediate, and high, respectively. The cTnT levels were measured for a control group of 20 dogs that were judged as healthy on the basis of history and objective clinical examination with particular attention to the cardiovascular system. This group of animals comprised subjects with different races and gender with an age between 1 to 12 years (mean 6.1). Another group of 30 animals was defined as cardiomyopathic on the basis of clinical and instrumental examination (e.g., echocardiograph and ecoDoppler, radiography and EKG). These all had cardiac insufficiencies between 1b and 3a, according to the ISACHC classification. On the basis of echocardiographic examination, the myocardiopathic group was further divided into two subgroups: non-hypokinetic myocardiopathic (12 dogs aged between 3 months and 16 years, mean 7.5) and hypokinetic myocardiopathic (global or segmental hypokinetic; 18 dogs aged between 7 months and 15 years, mean 6.0). Statistical analysis using Fisher's test was employed to evaluate the correlation between a positive cTnT test and cardiac insufficiency, as well as ventricular kinetics. The aim of the present work was to determine the value of cTnT for healthy dogs and for dogs affected with different cardiac pathologies, evaluating eventual correlations between cTnT levels, the ISACHC classification for cardiac insufficiency, the index of systolic volume (ESV-I), and ventricular kinetics. The control group of healthy dogs had values of serum cTnT that were below detectable levels (i.e., <0.05 ng/ml). Ten of the 30 myocardiopathic dogs (33%) tested positive for cTnT. However, there was no significant correlation between a positive result in the cTnT assay and the type of cardiac insufficiency or with the ESV-I. In the non-hypokinetic group, only one of the 12 dogs (8.3%) which had an intermediate level of cTnT was positive. In the hypokinetic myocardiopathic group, 9 out of 18 animals tested positive, three with low levels of cTnT, four with intermediate levels, and two with high levels. The probability of a positive cTnT value was significantly higher for hypokinetic animals with respect to non-hypokinetic dogs (Fisher's exact test, OR=10, p<0.05). It was possible to carry out a second examination for 4 out of 10 dogs at follow-up visits during the course of therapy. In three animals, the cTnT level was below detectable levels by the second control visit and was associated with an improvement in general clinical conditions in all cases. The remaining dog, diagnosed with dilative myocardiopathy, had levels of cTnT that had increased over a two week interval, although they still fell in the intermediate value category. During this period, the dog's conditions progressively worsened in spite of therapy and the animal died three days after the last visit. For the three dogs with dilative myocardiopathy, the value of cTnT was below detection limits. 387 The absence of detectable cTnT (>0.05 ng/ml) in healthy dogs confirms the utility of this assay in assessing acute myocardial damage, similar to that reported both for both human and veterinary subjects (De Francesco, 2002) . The results of the present work indicate the usefulness of the cTnT test for the identification of cardiopathic subjects for whom disease is associated with myocardial damage. Furthermore, our data are also interesting in relation to the lack of correlation between a positive cTnT result and the degree of systolic insufficiency. In addition, the observation that only 50% of the hypokinetic group was positive for cTnT suggests that the assay may have particular utility in discriminating animals with myocardial damage in course from those in which hypokinesis is most likely indicative of a previous injury. On the other hand, the positive test results observed in non-hypokinetic, myocardiopathic subjects implies that the measurement of cTnT is a diagnostic tool that is valuable for the identification of myocardial damage, even in animals that would not otherwise be suspected to have myocardial damage on the basis of clinical and instrumental examination. The determination of cTNT during the follow-up period may also have prognostic implications. During this period, clinical improvement was seen in all three 3 cases with negative cTnT, while in the subject for whom increased levels of cTnT were observed, clinical deterioration was noted and resulted in sudden death. Lastly, for the three dogs with dilative myocardiopathy serum cTnT was not detected. This may be explained by the low concentrations of troponin T at the myocardial level, as reported by O'Brien (1997) for Dobermans affected with idiopathic dilative myocardiopathy. Our results are in contrast to those found for humans, in which a relatively high percentage of patients were found to be positive for cTnT during the course of dilative myocardiopathy. It has been reported that the return to normal levels of cTnT following therapy also has a beneficial impact on prognosis, with survival times that are significantly higher . However, in the present study the number of animals with dilative myocardiopathy, all with negative cTnT, is too small to make any conclusions in this regard. Physiotherapeutic analysis, including postural evaluation of the patient, is very important in rehabilitative medicine (Basaglia, 2000) . Patients with locomotory problems often present altered body and movement patterns. Body Pattern (SC) is defined as: ''the sensation that everyone has of their body position in space and of its movement, formed on the basis of proprioceptive information integrated by visual afferences'' (Buzzi et al., 1996) . In physiology this concept is defined as kinaesthesia (Mountcastle, 1977) or somatic sensation (Ruch, 1976) . Posture is defined as the position the body adopts in space by means of the tonic activity of muscles acting against gravity (Villeneuve, 1996) . Modifications of this body pattern happen, for example, when there is pain in the locomotory apparatus or elsewhere, which leads to a reduction of the range of the movements, to anthalgic and compensatory postures and to substitutive movements or ''tricks''. This situation induces incorrect motory patterns and new body representations which are memorised by the nervous system. The aim of ''neuromotory reprogramming'' is to re-establish the normal body and global kinetic patterns by means of active proprioceptor stimulation and suitable postural exercises. In this paper we wish to present a rehabilitating postural methodology, in which global postural exercises, both static and dynamic, are added to the normal rehabilitation program. The exercise technique needs to be ''global'' as in nature there is no such thing as an isolated single muscle contraction (muscular chains mechanism), as the Nervous System regulates movement as a whole (Souchard, 1994) . Case 1=Male (M), 4 years (y), diagnosis: Hansen 1 disk rupture in T13-L1; haemilaminectomy 3 days after the appearance of para-paresis; beginning of rehabilitation program 14 days after surgery (PR+14d). Case 2=F6y, lumbar disk rupture, no surgery; beginning of rehabilitation program 60 days after the appearance of the symptoms. Case 3=F7y, multiple pelvis fracture. Reconstructive surgery using bone plates 4 days after trauma, PR+35d. Case 4=F6m, bilateral hip dysplasia. Bilateral triple pelvic osteotomy (TPO) 30 days after the appearance of symptoms, PR+20. After examining the postural and body pattern alterations, we made the patients perform postural exercises which led the patient to regain the ability to stand on all fours with the correct body pattern. Moreover, these exercises aimed to strengthen the so-called postural or gravity-opposing muscles, to regain normal mobility and to enhance muscle strength and coordination. Coordination, static-dynamic and functional balance exercises are performed mimicking normal movements. They are performed very slowly to facilitate memorisation through repetition, following the action-repetition principle of rehabilitative medicine, always taking care to make the patient always adopt a correct postural and motory pattern. Thus, during the working session postural errors are continuously corrected and neurophysiological responses are activated by following simple and instinctual postural and motory patterns (sit/stand up). Several working phases can be identified: a phase in which the dog lies on one side, another in which the dog is sitting in a ''sphinx'' or a ''sit dog'' position, another in which it stands on all fours, and a locomotion phase. The phases used to quantify progress are coded as in Vallani et al., 2002 . Case 1=on the first visit the dog was in F1/F2; after one month it was in F4, at 45d in F5, at 55d in F6 (all time values are calculated from the beginning of the rehabilitation program). Follow up: at two months it was in F6. Case 2=on the first visit it was in F1; at two and a half months in F4; at three months in F5 and at three and a half months in F6. Follow up: at 5 months it was in F6. Case 3=(due to the severity of the symptoms the neurological numerical phases classification was used in this case) on the first visit the dog was in F1; at 30d in F2/3; after 45d in F4; after three months in F6 (i.e. F5 for orthopaedic patients). Follow up: at four months the dog was in F6. Case 4=on the first visit the dog was in F1, it dragged its feet and rested the dorsal part of the feet on the ground; at 30d it was in F4; at 50d in F5. Follow up: at 5 months the dog was in F6. We conclude that the use of global postural techniques thus appears to be a useful addition to the rehabilitation programs of neurological and orthopaedic canine patients. 391 Censimento nazionale sulla molluschicoltura Elementi di valutazione ecologica, economica e sociale per fronteggiare la flessione produttiva di vongole filippine nell'Alto Adriatico Reactive Oxygen Species and antimicrobial defenses of invertebrates: a bivalve model In vitro production of superoxide and nitric oxide (as nitrite and nitrate) in M. galloprovincialis hemocytes upon incubation with PMA or laminarin or during yeast phagocytosis Evidence for interspecies transmission of oyster herpesvirus in marine bivalves Applications and pitfalls of stress proteins in biomonitoring Perkinsus marinus susceptibility and defense-related activities in eastern oysters C. virginica: temperature effects Humoral defense factors in marine bivalves Cellular defense mechanisms of oysters and mussels Factors influencing in vitro killing of bacteria by hemocytes of the eastern oyster (Crassostrea virginica) Interaction of phagocytosed Bonamia sp. with hemocytes of oysters T iostrea chilensis The inter-relationships of bivalve hemocytes On-growing of the carpet-shell clam (T apes decussatus) (L.): two years experience in Venice Lagoon Monitoring of infections by Protozoa of the genera Nematopsis, Perkinsus and Porospora in the smooth venus clam Callista chione from the North-Western Adriatica Sea (Italy) Presenza di Perkinsus in vongole veraci (Ruditapes philippinarum) allevate nel Nord-Est Italia /83/EC of 5th February 2003 amending Council Directive 95/70/EC introducing minimum Community measures for the control of certain diseases affecting bivalve molluscs Council Directive 91/67/EEC of 28th January 1991 concerning the animal health conditions governing the placing on the market of aquaculture animals and products Council Directive 95/70/EC of 22nd December 1995 introducing minimum Community measures for the control of certain diseases affecting bivalve molluscs. OJ No. L. 332, 30 Perkinsus, a protistan threat to bivalve culture in the Mediterranean basin Regolamento recante norme di attuazione della direttiva 95/70/CE in materia di misure minime di lotta contro talune malattie dei molluschi bivalvi. Gazzetta UYciale della Repubblica Italiana Perkinsus sp. nei bivalvi dal Reno a Cesenatico Su alcune patologie riscontrate in ostriche piatte (Ostrea edulis) provenienti da banco naturale Pathologies found in molluscs from Apulian coasts, Southern Italy. 5th International Colloquium on Pathology in Marine Aquaculture Presenza di Bonamia ostreae in ostriche piatte (Ostrea edulis) del Nord Adriatico Detection of hepatitis A virus in mussels from different sources marketed in Puglia region (South Italy) Viruses and bivalve shellfish Occurrence of Vibrio spp. In sea fish, shrimps and bivalve molluscs harvested from Adriatic sea Bivalve molluscs as vectors of human enteric viruses. Microbiologie Aliments Nutrition Incidence of and risk factors for hepatitis A in Italy: Public Health indications from a 10-year surveillance Proposal for a Regulation of the European Parliament and of the Council laying down specific hygiene rules for food of animal origin Benzodiazepine poisoning in companion animals Glyphosate toxicosis in domestic animals: a survey from the data of the Centre National d'Informations Toxicologiques Vétérinaires Les intoxications des animaux domestiques: bilan du CNITV 1999. T hèse de doctorat vétérinaire Loperamide poisoning in the dog General survey of the activity of the National Veterinary Antipoison Center Etude épidémiologique et toxicologique d'après les données du CNITV de Lyon concernant les anti-inflammatoires non stéroïdiens, le paracétamol et les benzodiazépines. T hèse de doctorat en pharmacie Mechanisms for selective toxicity of fipronil insecticide and its sulfone metabolite and desulfinyl photoproduct Ivermectin toxicity in an Australian Shepherd dog with the MDR1 mutation associated with ivermectin sensitivity in Collies Permethrin toxicity in cats Pericardial effusion and cardiac tamponade secondary to brodifacoum toxicosis in a dog Sémiologie et diagnostic différentiel des intoxications par les convulsivants chez les carnivores domestiques, Recueil de Médecine Vétérinaire Avvelenamenti nel cane nella regione Abruzzo nel quadriennio Avvelenamenti negli animali: casistica riscontrata dall'IZS delle Venezie nel quadriennio Intossicazioni veterinarie: casistiche del Centro Antiveleni di Milano anni Incidence of intentional poisoning of dogs in the Abruzzo Region of Italy. Veterinary and Human T oxicology Intossicazioni in medicina veterinaria Poisonings in animals: the 1993-1994 report of the American Association of Poison Control Centers. Veterinary and Human T oxicology T ossicologia Clinica Veterinaria Avvelenamenti da pesticidi negli animali da compagnia nell'isola di Ischia Intossicazione in medicina veterinaria: casistiche del laboratorio di tossicologia della facoltà di medicina veterinaria di Pisa nel triennio Tossicosi esogene dei carnivori domestici Legge 11 febbraio 1992, n. 157. Norme per la protezione della fauna selvatica omeoterma e per il prelievo venatorio,Gazzetta UYciale 25 febbraio 1992 Norme sul divieto di utilizzo e detenzione di esche avvelenate, Bollettino UYciale della Regione T oscana n. 27 del 27 agosto 2001, Parte I, Sezione I Legge Regione Umbria 22 ottobre 2001, n. 27. Norme in materia di divieto di detenzione ed utilizzazione di esche avvelenate, Bollettino UYciale della Regione dell'Umbria n. 53 del 26 ottobre 27/01 concernente norme in materia di divieto di detenzione ed utilizzazione di esche avvelenate Intossicazioni in medicina veterinaria T ossicologia Clinica Veterinaria Tossicologia veterinaria: descrizione di un caso in un fagiano Association between surges of follicle-stimulating hormone and the emergence of follicular waves in heifers Effect of duration of dominance of the ovulatory follicle on onset of estrus and fertility in heifers Evaluation of timed insemination using a gonadotropin-releasing hormone agonist in lactating dairy cows Timed artificial insemination (TAI) with estradiol cypionate (ECP) or insemination at detected estrus in lactating dairy cows Effect of resynchronization with GnRH on day 21 after artificial insemination on conception rate and pregnancy loss in lactating dairy cows Reproductive programs in dairy herds Effects of presynchronization and bovine somatotropin on pregnancy rates to a timed artificial insemination protocol in lactating dairy cows Use of estradiol cypionate in a pre-synchronized timed artificial insemination program for lactating dairy cows Reproductive management of lactating dairy cows using synchronization of ovulation Effect of human chorionic gonadotropin on luteal function and reproductive performance of high producing lactating Holstein dairy cows Concepts for regulation of corpus luteum function by the conceptus and ovarian follicles to improve fertility Synchronization rate, size of the ovulatory follicle, and pregnancy rate after synchronization of ovulation beginning on different days of the estrous cycle in lactating dairy cows Review: effect of protein nutrition on ovarian and uterine physiology in dairy cattle Development of immature bovine oocytes into viable embryos in vitro Practical aspects of IVM-IVF in cattle Embryo production by ovum pick-up from live donors Bovine embryo technologies. T heriogenology Effects of repeated ultrasound-guided transvaginal follicular aspiration on bovine oocyte recovery and subsequent follicular development Bovine follicular development and its effect on the in vitro competence of oocytes Effect of elevated systemic concentrations of ammonia and urea on the metabolite and ionic composition of oviductal fluid in cattle Kinetics of oocyte maturation and subsequent development of IVF, parthenogenetic, and NT bovine embryos after meiotic inhibition with roscovitine Cellular and molecular deviations in bovine in vitro-produced embryos are related to the large offspring syndrome Factors influencing oocyte and embryo quality in cattle Disruption of bovine oocytes and preimplantation embryos by urea and acidic pH Embryonic and foetal development of bovine oocytes treated with a combination of butyrolactone I and roscovitine in an enriched medium prior to IVM and IVF Market situation for buffalo milk products Ottime potenzialità per l'allevamento bufalino in Italia Sviluppo del mercato della Mozzarella di Bufala. Proceedings I Congresso Nazionale sull A new role for typical products in modern retails La filiera della mozzarella di bufala: un laboratorio dello sviluppo locale. L a Questione Agraria La spesa domestica per i prodotti tipici nel Considerazioni sulla produzione del vitellone bufalino da carne. Agrisole, suppl Chemical-nutritional characteristics of meat produced by young buffalo bulls. Proceeding of the XXXI International Symposium of Zoothecny ''Role of animal products on animal health Influence of the feeding programme on growth dynamics of buffalo young bulls until 16 months of age Produzione del vitello bufalino a carne bianca alimentato con due diversi tipi di latte. Nota II: qualità della carne La produzione della carne bufalina in Italia Prove comparative fra vitelli bovini e bufalini. V. Composizione acidica del grasso di varia localizzazione in animali di 28 settimane di età. Annali dell 'Istituto Sperimentale per la Zootecnia Influence of different rationing scheme on the growth performance of young buffalo bulls Studio comparativo fra bufali e bovini sulla qualità della carne. I. caratteristiche mioreologiche all'età di 20 settimane Prove comparative di allevamento di bovini e bufali fino a 20 settimane di età. II. Composizione acidica del grasso, prove di assaggio delle carni I problemi relazionali della convivenza tra animale da compagnia e proprietario. In: Il contributo della Fisiologia al miglioramento della convivenza tra animale e uomo La relazione cane-proprietario: Indagine nel contesto urbano e rurale di Padova Pet Therapy: terapia o arricchimento ambientale? In: Il contributo della Fisiologia al miglioramento della convivenza tra animale e uomo Metodi per la valutazione della qualità della vita dell'animale da compagnia nella relazione con il proprietario Incontinentia urinae bei der kastrierten hü ndin: häufigkeit und rassedisposition Effect of hormones on urethral sensitivity to phenylephrine in normal and incontinent dogs Developments in the understanding of the pathophysiology of the urethral sphincter mechanism incompetence in the bitch Association in bitches between breed, size, neutering and docking, and acquired urinary incontinence due to incompetence of the urethral sphincter mechanism. T he Veterinary Record Sex hormones and the female urinary tract Comparison of long-term effects of ovariectomy versus ovariohysterectomy in bitches Nef rologia ed urologia del cane e del gatto Acquired urinary incontinence in bitches: its incidence and relationship to neutering practices Influence of progesterone on boar sperm capacitation Actin polymerization in boar spermatozoa: fertilization is reduced with use of cytochalasin D. Molecular Reproduction and Development Actin-binding properties and colocalization with actin during spermiogenesis of mammalian sperm calicin An antiactin monoclonal antibody inhibits the zona pellucida-induced acrosome reaction and hyperactivated motility of human sperm Identification of capacitation in boar spermatozoa by chlortetracycline staining Sperm exocytosis reconstructed in a cell-free system: evidence for the involvement of phospholipase C and actin filaments in membrane fusion Ultrasonographic identification of fetal brain structures during pregnancy in the dog Preliminary study of the ultrasonographic diagnosis of pregnancy and fetal development in the dog Estudio ecográfico de la gestatió n en cánidos Studies on canine pregnancy using B-mode ultrasound: Development of the conceptus and determination of gestational age Determination of gestational age in medium and small size bitches using ultrasonographic fetal measurements Observation of the growth process of a beagle embryo and fetus by ultrasound Establishment of the prediction table of parturition day with ultrasonography in small pet dogs Ultrasonographic appearance of the uterus, placenta, fetus, and fetal membranes throughout accurately timed pregnancy in Beagles Raffronto tra parametri istologici e livelli di mediatori chimici della flogosi in biopsie endometriali di bovine ipofertili Evaluation of mast cells in calf and heifer uteri Morphometric studies of the bovine uterus: microscopic lesion and retrospective reproductive history Bovine mast cells: distribution, density, heterogeneity, and influence of fixation techniques The equine endometrial mast cell during the puerperal period: evaluation of mast cell number and types in comparison to other inflammatory changes Use of buserelin to induce ovulation in cyclic mares Alternative solutions to hCG induction of ovulation in the mare Advancing the time of ovulation in the mare with a short-term implant releasing the GnRH analogue deslorelin Use of the GnRH analogue desloreline acetate in a slow-release implant to accelerate ovulation in oestrus mares Uses of LHRH and analogues in the mare The development of antibodies to human chorionic gonadotrophin following its repeated injection in the cyclic mare Use of hCG, GnRH agonist or REFERENCES Isolated ovine primordial follicles cryopreserved in different concentrations of ethylene glycol Growth and differentiation of small ovarian follicles in mammals: problems and perspective In vitro development of sheep preantral follicles Transplantation of cryopreserved fetal ovarian tissue to adult recipients in mice Restoration of fertility to oophorectomized sheep by ovarian autografts stored at −196 degrees Current perspective on primordial follicle cryopreservation and culture for reproductive medicine In-vitro growth of murine pre-antral follicles after isolation from cryopreserved ovarian tissue Ovarian tissue cryopreservation and transplantation: preliminary findings and implications for cancer patients Discharge of impulses in motor nerve fibres. The frequency of discharge in reflex and voluntary contraction Caratterizzazione della reazione al backtest nel suinetto presvezzamento: correlazioni con l'accrescimento corporeo Individual behavioural and physiological strategies in pigs Individual variation in the behaviour of pigs -Noise or functional coping strategies? Personalities in female domesticated pigs: behavioural and physiological indications Adaptation to social isolation. Acute and long-term stress responses of growing gilts with different coping characteristics Individual behavioural characteristics in pigs and their impact on production Dog parentage testing using canine microsatellites A class of highly polymorphic tetranucleotide repeats for canine genetic mapping Three polymorphic canine microsatellites Eighteen canine microsatellites Nine canine microsatellites Canine parentage testing based on microsatellite polymorphisms Identification and characterization of a set of 100 tri-and dinucleotide microsatellites in the canine genome Use of canine microsatellite polymorphisms in forensic examinations One hundred and one new simple sequence repeatbased markers for the canine genome A new method of paternity testing for dogs, based on microsatellite sequences, T he Veterinary Record Nerve growth factor: a neurotrophin with activity on cells of the immune system Neurotrophin actions during the development of the peripheral nervous system Interleukin-6 and interleukin-6 receptor mRNA expression in rat central nervous system TrkA is necessary for the normal development of the murine thymus Neurotrophins: roles in neuronal development and function Expression of TrkB neurotrophin receptor during T cell development. Role of brain derived neurotrophic factor in immature thymocyte survival Phenotypic knockout of nerve growth factor in adult transgenic mice reveals severe deficits in basal forebrain cholinergic neurons, cell death in the spleen, and skeletal muscle dystrophy Radioimmunoassay for insulin-like growth factor-I: solutions to some potential problems and pitfalls Concerned about use of BST in dairy cattle La posizione dei comitati scientifici della U.E. sulla somatotropina bovina ricombinante Caspase 6 activity initiate caspase 3 activation in cerebellar granular cell apoptosis Synapse-independent and synapse-dependent apoptosis of 200 cerebellar granule cells in postnatal rabbits occur at two subsequent but partly overlapping developmental stages Role of cell cycle regulatory proteins in cerebellar granule neuron apoptosis The world according to PARP Molecular characterization of mitochondrial apoptosis-inducing factor Human skeletal muscle fibres: molecular and functional diversity Enhanced electrophoretic separation and resolution of myosin heavy chains in mammalian and avian skeletal muscles Differences in molecular structure among porcine myosin heavy chain 2a, 2x and 2b isoforms Specificity of different anti-myosin heavy chain antibodies in bovine muscle Fibre types in extraocular muscles: a new myosin isoform in the fast fibres Sequencing of the 2a, 2x and slow isoforms of the bovine myosin heavy chain and the different expression among muscles Disruption of imprinting caused by deletion of the H19 gene region in mice Developmental competence of pig oocyte matured and fertilized in vitro Concentration of cyclic AMP during the maturation of pig oocyte in vivo and in vitro Identification of capacitation in boar spermatozoa by chlortetracycline staining Demethylation of the zygotic paternal genome Reprogramming of genome function through epigenetic inheritance Improving the safety of embryo technologies: possible role of genomic imprinting Accommodation of vascularized xenografts: 212 expression of ''protective genes'' by donor endothelial cells in a host Th2 cytokine environment Carbon monoxide generated by heme oxygenase 1 suppresses endothelial cell apoptosis Angiogenic growth factors in the pathophysiology of a murine model of acute lung injury Carbon monoxide has anti-inflammatory effects involving the mitogen-activated protein kinase pathway Carbon monoxide generated by heme oxygenase-1 suppresses the rejection of mouse-to-rat cardiac transplants Ghrelin, a novel ghrowth hormone-releasing acylated peptide is 215 synthesized in a distinct endocrine cell type in the gastrointestinal tracts of rats and humans A-like cells in the rat stomach contain ghrelin and do not operate under gastrin control Immunoistochemical localization and biochemical characterization of ghrelin in the brain and stomach of the frog Rana esculenta Ghrelin, a novel placental-derived hormone Ghrelin and the regulation of energy homeostasis: a hypothalamic perspective Purification and characterization of rat des-Gln14-Ghrelin, a second endogenous ligand for the growth hormone secretagogue receptor Chicken ghrelin: purification, cDNA cloning, and biological activity Ghrelin is a growthhormone-releasing acylated peptide from stomach Immunocytochemical observation of ghrelin-containing neurons in the rat arcuate nucleus Kidney produces a novel acylated peptide, ghrelin Gastric endocrine cells: types, function and growth Alterazioni della funzionalità epatica nel puerperio e conseguenze sulla fertilità della bovina Nutritional modulation of somatotropic axiscytokine relationships in cattle: a brief review Dry matter intake reduction before calving in dairy cows: relationship with immune system and metabolism conditions. 53rd Meeting European Association Animal Production Commissione valutazione alimenti. Valutazione degli alimenti di interesse zootecnico. Analisi chimica Tecniche di alimentazione delle pecore e stima del valore energetico e proteico degli alimenti Alimentation des bovins, ovins et caprins. Institut National de la Recherche Agronomique Metodi di analisi per la valutazione degli alimenti di impiego zootecnico Recherche d'une méthode d'estimation de la production laitière des femelles ovines et bovines pendant la phase d'allaitement SAS/STATA Software: Changes and Enhancements through Release 8.1 Ochratoxin A in food and feed: occurrence, legislation and mode of action International Agency for Research on Cancer) 1993. In IARC Monographs on the Evaluation of REFERENCES Pregnancies established after transferring embryos yielded by Ovum Pick-up and in vitro embryo production in Italian buffalo cows Embryo production by Ovum pick-up in water buffalo Embryo production by Ovum Pick-up from live donors Effect of cysteamine during in vitro maturation on buffalo embryo development Comparative studies on the superovulatory effect of PMSG and FSH in the water buffalo (Bubalus bubalis). T heriogenology Commercial aspects of oocyte retrieval and in vitro fertilization (IVF) for embryo production in problem cows In vitro fertilization in buffalo and birth of first ever IVF buffalo calf In vitro production and transfer of embryos in buffaloes Cryopreservation of in vitro produced buffalo and bovine embryos by vitrification Bovine and buffalo in vitro embryo production using oocytes derived from abattoir ovaries or collected by transvaginal follicle aspiration In vitro maturation, fertilization and development of follicular oocytes from buffalo (Bubalus bubalis) Effects of dietary conjugated linoleic acid in nursery pigs of dirty and clean environments on growth, empty body composition, and immune competence Dietary 244 conjugated linoleic acid (CLA) influence immune response in suckling and weanling piglets Influence of dietary conjugated linoleic acid on growth, meat quality, lipogenesis, plasma leptin and physiological variables of lipid metabolism in rabbits Effects of dietary conjugated linoleic acid on some aspecific immune parameters and acute phase protein in weaned piglets Mammary cancer prevention by conjugated dienoic derivative of linoleic acid The influence of nonspecific immunostimulation of pregnant sows on the immunological value of colostrum Conjugated linoleic acid and atherosclerosis in rabbits Immunological quantitation of antigens by single radial immunodiffusion Mechanisms for conjugated linoleic acid-mediated reduction in fat deposition Formation and action of anticarcinogenic fatty acids The acquisition of passive immunity in the new-born piglet Conjugated linoleic acid modulates tissue levels of chemical mediators and immunoglobulins in rats Proposta di un modello matematico per il pagamento del latte di bufala secondo qualità Resa alla caseificazione e parametri reologici del latte bufalino: risultati preliminari su latti individuali Andamento di alcune componenti del latte di bufala in funzione della distanza dal parto: risultati preliminari Confronto tra bufale con differente resa alla caseificazione REFERENCES EC Commission Decision N. 471 of 08/06/2001, GUCE L 165 ) Food Directorate's-Health Canada's Liquid chromatographic identification of meats Double antibody ELISA for detection of trace amounts of pig meat in raw meat mixtures Characterization and identification of raw beef, pork, chicken, and turkey meats by electrophoretic patterns of their sarcoplasms proteins Determination of organophosphorus pesticides in whole/chocolate/skim-milk and infant formula using SPE with capilary GC-NPD. Food and Drug Administration L aboratory Information Bulletin ORA n Optimized method for the determination of organophosphorus pesticides in meat and fatty matrices The European Commission Ispezione degli alimenti di origine animale Indagine sulle batteriemie in spigole e orate d'allevamento. Atti Aassociazione Italiana VeterinariIgienisti Cit Ricerca di Clostridium Perfringens in Pesci Marini Virus enterici veicolati da alimenti Simple method of estimating 50% end points Microbial purification of shellfish: a review of depuration and relaying Improved detection of human enteric viruses in foods by RT-PCR Microbiological hazard of fresh pork sausages: result of a 1-year study Isolamento di salmonelle da alimenti di origine animale nel biennio 1997-1998 Valutazioni microbiologiche di carni trite e preparazioni di carni ai sensi del D.P.R. 309/98 Salmonella contamination of minced meat and meat products. Atti X Congresso Nazionale Associazione Italiana Veterinari Igienisti Performance standards for antimicrobial disk and dilution susceptibility tests for bacteria isolated from animals: tentative standard, Waine, PA National Committee for Clinical Laboratory Standards Valutazione microbiologica di carni macinate e preparazioni di carne Valutazioni microbiologiche di carni bovine e suine macinate Salmonella in slaughter pig: prevalence, serotypes and critical control points during slaughter in two slaughterhouses T he medical impact of antimicrobical use in food animals T homson's Special Veterinary Pathology Urinary tract neoplasms in dogs and cats. T he Compendium on Continuing Education for the Practicing Veterinarian Tumori dell'apparato urinario Histological T yping of Urinary Bladder T umours -W HO International Histological Classification of T umours Neoplasms of the canine urinary bladder La contaminazione delle razioni per scrofe con aflatossine B1 e/o G1. Effetti sulle performances produttive della prole nelle fasi di accrescimento e di ingrasso. L a Rivista di Scienza dell'Alimentazione Aflatoxicosis in pregnant sows: cell-mediated response in their new-born piglets Sottopopolazioni linfocitarie nel sangue periferico di suinetti da 1 a 10 settimane di età Effect of aflatoxins on tocopherol and retinol concentrations in growing barrows Immunotossicità da aflatossine in scrofe gravide Immunotoxicological effects on piglets of feeding sows diets containing aflatoxins. T he Veterinary Record The detection of Coxiella from ovin genital swabs, milk and fecal samples by the use of single touchdowm polymerase chain reaction Influence of three types of farm management on the seroprevalence of Q fever as assessed by an indirect immunofluorescence assay Epidemiological survey of Q fever on dogs in an area of the Southern Italian Appennines and correlations with leishmaniosis and rickettsiosis Diagnosis of Q fever Atti Società Italiana Patologia Allevamento Ovini Caprini Sea bream Sparus aurata, an asymptomatic contagious fish host for nodavirus Analisi del gene della proteina del capside in Nodavirus proveniente da un cefalo (L iza ramada) di allevamento Trasmission of viral encelopathy and retinopathy (VER) to yolk-sac larvae of the Atlantic halibut Hippoglossus hippoglossus: occurrence of nodavirus in various organs and a possible route of infection High permissivity of the fish cell line SSN-1 for piscine nodaviruses Betanodavirus infections of teleost fish: a review Genomic classification of fish nodaviruses by phylogenetic analysis of the coat protein gene A simple method of estimating fifty per cent end-points Pathogenicity of nodavirus strains from striped jack Pseudocarnax dentex and Atlantic halibut Hippoglossus hippoglossus, studied by waterborne challenge of yolk-sac larvae of both teleost species Family Nodaviridae Establishment and characterization of two new pig cell lines for use in virological diagnostic laboratories Production of transgenic rabbits, sheep and pigs by microinjection Search for cross species transmission of porcine endogenous retrovirus in patients treated with living pig tissue Antiretroviral agents inhibit infection of human cells by porcine endogenous retroviruses Susceptibility of the porcine endogenous retrovirus to reverse transcript and protease inhibitors Presenza di retrovirus endogeni in colture cellulari di origine suina e determinazione della eventuale trasmissione in sistemi cellulari di primati Identification of three distinct antigenic sites in parapoxvirus. Archives of V irology Characterisation of the H3L envelope protein: topology and posttranslational membrane insertion via the C-terminal hydrophobic tail Infectious Diseases of the dog and cat PCR detection and molecular identification of Chlamydiaceae species Vascular permeability factor/vascular endothelial growth factor, microvascular hypermeability, and angiogenesis Induction of angiogenesis during the transition from hyperplasia to neoplasia What is the evidence that tumors are angiogenesis-dependent? Tumour angiogenesis Patterns and emerging mechanism of the angiogenic switch during tumorigenesis Expression of vascular endothelial growth factor in basal cell tumours and in squamous cell carcinomas of canine skin Angiogenesisi in squamous cell carcinomas of canine skin: an immunohistochemical and quantitative analysis The association between tumour progression and vascularity in the oral mucosa Histological typing of cancer and precancer of the oral mucosa. In: World Health Organization International Histological Classification of T umours II Expression of vascular endothelial growth factor in canine mammary tumors Endothelial cell proliferation in prostatic carcinoma and prostatic hyperplasia: correlation with Gleason's score, microvessel density, and epithelial cell proliferation Tumoral vascularity as a prognostic factor in cancer patients: the evidence continues to grow The same prion protein cause vCJD and BSE The immunobiology of TSE diseases Immunization with amyloid-beta attenuates Alzheimer disease-like pathology in the PDAPP mouse Immunization delays the onset of prion disease in mice CP 55,940 protects against ischemia-induced electroencephalographic flattening and hyperlocomotion in Mongolian gerbils Drug therapy in human and experimental transmissible spongiform encephalopathy Metal chelation as a potential therapy for Alzheimer's disease and C is for Clioquinol' -the AbetaCs of Alzheimer's disease. T rends in Neurosciences Polydeoxyribonucleotide (defibrotide) protects against post-ischemic behavioral, electroencephalographic and neuronal damage in the gerbil Ancylostoma tubaeforme (Zender, 1800) in gatti selvatici e domestici della Sardegna Epidemiologic investigation of seroprevalence of antibodies to T oxoplasma gondii in cats and rodents Keys to the Cestode Parasites of Vertebrates Patologia respiratoria animale Fauna parassitaria gastro-intestinale del gatto nelle Città di Bologna Prevalence of T oxoplasma gondii antibodies in sera of domestic cats from Guarulhos and San Paulo Parassiti in medicina veterinaria L 'esame coprologico nella diagnosi delle elmintiasi, seconda edizione T oxoplasma gondii: from animals to humans Isolamento e tipizzazione di micobatteri in pesci tropicali d'allevamento Mycobacteriosis and Nocardiosis A systemic mycobacteriosis in the European sea bass Dicentrarchus labrax cultured in Eilat (Red Sea) Mycobacterial infection in farmed turbot Scophthalmus maximus Indagine sulla diffusione delle Micobatteriosi ittiche e sui relativi reperti anatomoistopatologici. T esi di Specializzazione in Allevamento, igiene, patologia delle specie acquatiche e controllo dei prodotti derivati Profilassi delle zoonosi di origine ittica Casi di micobatteriosi in pesci rossi d'acquario Confidence limits on phylogenies: an approach using the bootstrap Antigenic crossreactivity between caprine arthritis-encephalitis, visna and progressive pneumonia viruses involves all virion-associated proteins and glycoproteins Genetic heterogeneity of small ruminant lentiviruses involves immunodominant epitope of capsid antigen and affects sensititity of single-strain-based immunoassay MEGA2: molecular evolutionary genetics analysis software Genomic heterogeneity of small ruminant lentiviruses: existence of heterogeneous populations in sheep and of the same lentiviral genotypes in sheep and goats. Archive of V irology Characterization of the immunodominant crossreacting epitope of visna maedi virus and caprine arthritis-encephalitis virus capsid antigen The neighbor-joining method: a new method for reconstructing phylogenetic trees A new sensitive serological assay for the detection of lentivirus infections in small ruminant Phylogenetic analysis of small ruminant lentiviruses Expression in Escherichia coli and sequencing of the coding region for the capsid protein of Dutch maedi-visna virus strain ZZV 1050: Application of recombinant protein in enzyme-linked immunosorbent assay for the detection of caprine and ovine lentiviruses Holter monitoring and cardiac event recording A retrospective clinical study of canine leishmaniasis in 150 dogs naturally infected by L eishmania infantum Thrombosis and nephrotic syndrome in a dog with a visceral leishmaniasis A role of thrombomodulin in the pathogenesis of thrombin-induced thromboembolism in mice Santoro1 and P. Lombardi2 1Department of Veterinary Clinical Science -Section of Internal Medicine paociara@unina.it Keywords: Leishmaniasis, platelet, aggregation, dog Abbreviations: CanL A retrospective clinical study of canine leishmaniasis in dogs naturally infected by L eishmania infantum Platelet aggregation response in canine leishmaniasis Leishmania immune adherence reaction in vertebrates Evaluation of secondary haemostasis in canine leishmaniasis Evaluation of primary haemostasis in canine leishmaniasis Study of haemostatic disorders in experimentally induced leishmaniasis in Beagle dogs Funzioni e T ecnologie Biologiche e; 2Dipartimento di Patologia e Sanità Animale -Università di Napoli IL-I, interleukin-I CAT, catalase; Cu-Zn SOD, copper-zinc superoxide dismutase Microplate superoxide dismutase assay employing a nonenzymatic superoxide generator Role of superoxide in deoxyribonucleic acid strand scission Cytokine activity in cancer-related anorexia/cachexia: role of megestrol acetate and medroxyprogesterone acetate Medroxyprogesterone acetate increases anthracyclines uptake in chronic lymphatic leukemia cells: role of nitric oxide and lipid peroxidation Colorimetric assay of catalase Bergamini1* 1Department of Veterinary Clinical Sciences FAMIGL I@vet.unibo.it Keywords: dog, diabetes mellitus, epidemiology, breed A genome-wide search for human type 1 diebetes susceptibility genes Canine diabetes mellitus: a retrospective survey Canine immune-mediated diabetes mellitus: a case report Familiar canine diabetes mellitus Breed distribution of dogs with diabetes mellitus admitted to a tertiary care facility A qualitative assay for beta cell antibodies. Preliminary results in dogs with diabetes mellitus Familial insulin-dependent diabetes mellitus in Samoyed dogs Inheritance of diabetes mellitus in Keeshond dogs Epizootiologic patterns of diabetes mellitus in dogs Diabetes mellitus in a colony of golden retrievers Reevaluation of the importance of polymorphic HLA class II alleles and amino acids in the susceptibility of individuals of different populations to type I diabetes Tetralogy of Fallot in a cat Schalm's veterinary haematology Understanding the pathophysiology of congenital heart defects Polycythemia and tetralogy of Fallot in a cat A single major-gene defect underlying cardiac conotruncal malformations interferes with myocardial growth during embryonic development: studies in the CTD line of keeshond dogs Echocardiography Causes and prevalence of cardiovascular disease Primary acquired mitral insufficiency in adult large breed dogs Unusual aspects of mitral valve disease in the dog Echocardiographic and echo-Doppler examination in the European wolf (Canis lupus L Evaluation of four 2-dimensional echocardiographic methods for assessing left atrial size dogs Recommendations for standards in transthoracic two-dimensional echocardiography in the dog and cat Observation on breed and sex in relation to canine heart valve incompetence Observation on the effect of age on the severity of heart valve lesion in the dog Anestesia degli animali ''insoliti'' -II Anestesia nei rettili Cardiopulmonary and anesthetic effects of propofol administered intraosseously to green iguana The medical care of iguanas and other common pet lizards. Veterinary Clinics of North America British Small Animal Veterinary Association: Manual of Exotic Pets Elementi di anestesia degli animali esotici e selvatici Anticardiolipin Antibodies in Dogs R. Papini1*, O. Pellegrini2, C. Rossini1 and G. Guidi1 1Department of Veterinary Clinics Dipartimento di Clinica Veterinaria, V iale delle Piagge 2 -56124 Pisa -Italy E-mail: rpapini@vet.unipi.it Keywords: anticardiolipin, antibody, antiphospholipid syndrome, dog, Italy Abbreviations: CL, Cardiolipin; ACA ELISA, enzyme-linked immunosorbent assay; IgG, immunoglobulin G; GPL, IgG/ml units REFERENCES Anti-cardiolipin anti-cardiolipin plus bovine, or human beta(2)glycoprotein-I and anti-human beta(2)glycoprotein-I antibodies in a healthy infant population Thrombosis, recurrent fetal loss and thrombocytopenia. Predictive value of the anticardiolipin antibody test Evaluation of the anti-cardiolipin antibody test: report of an international workshop held 4 High prevalence of anti-cardiolipin and other autoantibodies in a healthy elderly population A review of antiphospholipid syndrome Antiphospholipid antibodies in pediatrics Overview on anticardiolipin ELISA standardization Chemotactic activities in non-mastitic and mastitic mammary secretions: Presence of Interleukin-8 in mastitic but not non-mastitic secretions Quantitation of bovine cytokine mRNA in milk cells of healthy cattle by real-time TaqMan PCR Detection of proinflammatory and regulatory cytokines in bovine millk using RT-PCR Cytokines in mammary lymph and milk during endotoxin-induced bovine mastitis Giudice1 1Department of Medical Veterinary Sciences -Faculty of Veterinary Medicine Semel in die Blood urea nitrogen; CFR, Chronic renal failure; PCR, Polymerase chain reaction 364 was not compliant with the drug regimen. Cases 2 and 4 were treated by oral administration. Case 2 was given three different oral treatments without any success: fluconazole (5 mg/kg SID for 2 months), metronidazole (25 mg/kg) and spiramycin (150,000 UI/kg) SID for 35 days and itraconazole (50 mg SID) for 2 months. Seven months later the owner decided to suspend all treatment drugs. Case 4 was given allopurinol (20 mg/kg SID) for 15 months. Case 4 was a FIV+ cat also suffering from pancytopenia. On this occasion erythropoietin (50 UI/kg/48 h SC) and ferrous sulphate were administered (50 mg SID PO) for ten weeks and rHuIFNa-2a (30 UI SID topically on the oral mucosa) for 5 months All cats showed hyperproteinemia and hyperglobulinemia with monoclonal hypergammaglobulinemia. Cases 1, 3 and 4 were FIV+; case 2 was positive for Coronavirus (titre 25). Cases 1 (titre 100), 2 (titre 400) and 4 (titre 6400) were IgG positive for T oxoplasma, while cases 2 (titre 256) and 3 (titre 256) were IgG positive for B. henselae. Case 1 (FIV+) was also affected with CRF complicated by non-regenerative anaemia, bronchopneumonia and pyothorax and died 35 days after diagnosis. In cases 2 (unsuccessful treatment) and 3 (no treatment), development of the disease led to the progressive development of CRF with severe non-regenerative anaemia. Euthanasia was performed 22 months (case 2) and 5 years (case 3) after diagnosis. In case 2 CRF was evident 5 months after diagnosis with consequent cachexia and severe anaemia A review of the treatment of canine leishmaniasis A high prevalence of feline leishmaniasis in southern Italy Efficacy of metronidazole-spiramycin combined therapy in the treatment of canine leishmanisis Comparison of direct and indirect (oscillometric) measurements of arterial blood pressure in conscious dogs Monitoraggio della pressione sanguigna con metodi non invasivi e trattamento dell'ipertensione A clinical evaluation of blood pressure through noninvasive measurement using the oscillometric procedure in conscious dogs Aspergillus carbonarius as the main source of ochratoxin A contamination in dried vine fruits from the Spanish market T ossicologia Veterinaria Effects of the mycotoxin ochratoxin A and some of its metabolites on the human cell line THP-1. T oxicology Ochratoxin A. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans Two dimensional gel electrophoresis and identification methods Evaluation of thyroid, adrenal and pituitary function Disease of the pituitary gland Hormonal response to Thyrotropin-releasing hormone in healthy horses and horses with pituitary adenoma Equine pituitary adenoma: a functional and morphological study Plasma adrenocorticotropin concentration in healthy horses and horses with clinical signs of hyperadrenocorticism Pituitary Pars Intermedia Dysfunction (Equine Cushing's like Disease) L aboratory Profiles of Equine Diseases Equine pituitary tumors Equine Cushing's disease Pergolide treatment for Cushing's syndrome in a horse Plasma adrenocorticotropin (ACTH) concentrations and clinical response in horses treated for equine Cushing's disease with cyproheptadine or pergolide Pituitary Adenomas: Equine Cushing's Diseases The Michigan Cushing's project Problems in the diagnosis of pituitary adenoma (Cushing's syndrome) in horses Equine Cushing's disease V 10/01/1987 in Cass. Pen V 21/04/1983 in Cass. Pen 11/07/1992 in Giust Manuale di Diritto Penale, p.s., II, Milano 11. Cass. Pen. Sez. V, 24/10/1980 in Cass Prospective clinical evaluation of serum cardiac troponin T in dogs admitted to a veterinary teaching hospital Deficiencies of myocardial troponin-T and creatine kinase MB isoenzyme in dogs with idiopathic dilated cardiomyopathy Cardiac troponin T is a sensitive, specific biomarker of cardiac injury in laboratory animals Persistently increased serum concentrations of cardiac Troponin T in patients with idiopathic dilated cardiomyopathy are predictive of adverse outcomes L e Metodologie Posturali in Funzione Educativa e Rieducativa Sensazione Somatica Basi del Metodo di Rieducazione Posturale Globale Uso di una piattaforma biomeccanica nel programma di riabilitazione di pazienti canini ortopedici e neurologici Pied, É quilibre & Posture. Editions Frison-Roche This work was supported by grants provided by the Italian Ministry for Agricultural and Forest Politics (VI Piano Triennale della Pesca e dell'Acquacoltura). The authors are indebted to Dr. Federica Gusson, Dr. C. Scotti and Sig. Gianluca Maggiora. The work was partly supported by grants from Nestlé Purina PetCare Italia S.p.A. The authors thank Martina De Zanche for her assistance with data collection and all the volunteers who took part in the study. The Authors wish to thank Simona Normando and the colleagues who referred the patients. stimulated. We will verify this hypothesis by looking for complement fractions in CNS. In fact, autoimmune encephalitis occurred during clinical immunization trials with b-amyloid performed on Alzheimer's disease patients; for this reason, the trials were interrupted. We can conclude that the immunization can only induce a certain rate of resistance to experimental infection if the infection is performed via the intraperitoneal route, which is more comparable to natural infection. The more effective immunogens in this experiment comprise the prion protein sequence probably involved in the PrPres conversion process. Immunization performed using the described protocols is to be considered protective as it can delay the onset of the symptoms and the death of the animals, but it cannot stop the disease.Plasma TM levels were measured in 38 animals: 24 dogs naturally infected by L eismania infantum (none of the animals had ever undergone any specific anti-Leishmania therapy) and 14 healthy dogs (control group). Sick dogs were divided into two groups on the basis of the clinical examination . The first group comprised 10 oligo-symptomatic dogs while the second group comprised 14 symptomatic or markedly symptomatic dogs. The clinical diagnosis was always confirmed by direct observation of the protozoan in Giemsa stained fine needle aspirates of bone marrow and lymph nodes, and serologically using the immunofluorescent antibody test (IFAT). The threshold titre for positivity was 1/160. A full blood count and measurement of total plasma proteins (TP), serum protein electrophoresis, alanine transaminase (ALT), creatinine and urea serum levels were obtained for each dog. Prothrombin time (PT), activated partial thromboplastin time (APTT) and fibrinogen were also determined using a semi-automatic coagulometer. TM concentration in the plasma was measured using ELISA with a commercially available kit (American Diagnostic Inc., USA). Mean values with their standard deviations were calculated from the data obtained and were then statistically evaluated using one-way analysis of variance. Significant results were subjected to the Student-Newman-Keuls Multiple Comparisons Test. The mean±SD of plasma levels of thrombomodulin in healthy dogs was 2.51±0.5 ng/ml, while in dogs with leishmaniasis it was 3.27±0.8 ng/ml ( p<0.01): group I: 2.9±0.4 ng/ml, group II: 3.63±0.9 ng/ml ( p<0.001 vs control group and 0.05 vs group II). The results of coagulation tests were in the normal range (Table I) .Normocytic and normochromic anaemia, slight thrombocytopenia, increased activities of serum ALT, increase in beta and gamma globulins and moderate hypoalbuminaemia were observed for the second group. Plasma TM levels were significantly higher for dogs with leishmaniasis, especially those with severe clinical signs (Group II). In this study a significant relationship between high plasma TM levels and beta and gamma globulins plasma levels was observed. These elevated concentrations of gamma-globulins can be explained as a polyclonal response of B lymphocytes. These antibodies are not all directed against Leishmania but also reflect an autoimmune reaction. Deposition of circulating immuno-complexes (type III hypersensitivity) in blood vessels of different organs such 329 as those of the liver, spleen, skin, kidney, etc cause diffuse microvasculites and consequently, damage to the endothelium. The skin, ocular and haemopoietic involvement observed for animals of the second group is suggestive of an immunomediated process of the vessels with endothelial damage and TM increase. In agreement with Kumada et al. (1988) , the increase of ALT in the second group suggests a decrease in hepatic clearance and, consequently, an increase in plasma TM levels. In human internal medicine, high levels of TM have also been reported for the thromboembolic condition and for renal failure which have also been reported for canine leishmaniasis (Font et al., 1993; . However, in our research, the basal levels of PT, APTT, fibrinogen, urea and creatinine excluded renal failure and/or hemocoagulopathy. In conclusion, despite the low number of dogs used, our results suggest that plasma thrombomodulin can also be used as a non-invasive marker for endothelial injury in dogs as well as in humans. Further studies, including immunohistochemistry and the value of other endothelium markers, are needed to assess the role 330 played by TM in the genesis of endothelial lesions in dogs suffering from physiological and pathological conditions. different affinities for the membrane receptors: ADP agonist interacts with the GPIIb/IIIa membrane receptor, while collagen fastens especially to the GPIa/IIa membrane receptor. It is possible that the parasite causes direct damage to the different platelet membrane receptors which cannot then be recognised as ''self '' and consequently activate the immune-mediated process. Recently Dominguez and Torano (2001) have demonstrated that in the first phases of infection L eishmania is able to interact directly with the platelets by a specific mechanism termed ''immune adherence'', with the formation of large aggregates. Our data suggest that in dogs with severe clinical symptoms there is involvement of the intrinsic pathway as demonstrated by the prolonged APTT levels. This involvement could be due to a reduction in synthesis together with an increase in consumption of one or more clotting factors following hepatic damage and the chronic inflammatory state. Similar results were reported by Moreno (1999) for 26 dogs with Leishmaniasis although this author has not related his results to clinical conditions. In conclusion, this study demonstrates that in natural CanL primary and secondary haemostatic alterations are related to the severity of clinical symptoms. Collagen agonist could be used as a non-invasive first choice marker for the evaluation of platelet aggregation in a L eishmania infection. 334 In animals with very severe clinical conditions, the compromised immune state, the chronic inflammatory condition and the progressive liver and renal damage influence platelet function and the synthesis and metabolism of the clotting factors and can cause future bleeding disorders. The bibliography regarding reptilias anaesthesia and especially the green iguana is poor. However, according to many authors (Heard et al., 2002; Mazzi, 2000; Nevarez et al., 2002) inhalation agents are the best in spite of the limitations caused by voluntary apnoea, even if some prefer direct intubation without a pharmacological preventive treatment. Injectable anaesthetics have some counter indications (longer induction and recovery, less muscle relaxation etc.). It was possible to use Z with posologies indicated by other authors (Addis et al., 1990; Barten, 1993) but lower than those reported by Mazzi (2000) . This anaesthetic allows a quicker induction and the presence of benzodiazepine gives the muscle relaxation required for handling and orotracheal intubation. No other analgesic is needed as Nevarez et al. (2002) suggest, because cyclohexamine provides sufficient action during diagnostic and surgical procedures. Using Zoletil i.m. in the cranial dorsal muscles there is no dosage problem related to 353 the presence of the renal portal circle and its use is easier than that of other anaesthetics which require an i.v. or intraosseous injection (Bennet et al., 1998) . The heart rate increase caused by cyclohexamine is useful to avoid bradycardia, a very common side effect when using propofol (Bennet et al., 1998) . Cyclohexamine also prevents side effects caused by the use of propofol. Zoletil did not cause either inducted apnoea or hypoxia. Orotracheal intubation allows use of a lower concentration of Isoflurane than in anaesthetic chambers or under maintenance using a mask (Addis et al., 1990; Barten, 1993) . Long recovery times are necessary, not 12 h as Mazzi (2000) because of Ketamine.Use of Zoletil does not cause excessive excitement in the green iguana, not similarly to effects noted for mammals and birds. These experiences show that this anaesthetic method is valid for anaesthesia of the green iguana. Differences in the mean systolic and diastolic pressure observed between pathological groups and the control group for the dog measurement using the oscillometric technique as an easy and reliable diagnostic tool, especially for some common diseases of the dog. The horse suffered from chronic recurrent laminitis and a one year history of infertility as well as progressive hirsutism, hyperidrosis, pendulous abdomen, lethargy, polyuria and polydipsia (90 L of water every day) and several months of anorexia. At the time of the visit the subject was also febrile (39.8°C) and the respiratory rate was slightly increased (30 min). Chest auscultation demonstrated harshness of expiratory sounds and inspiratory crackles, particularly in antero-ventral areas. Anoestrus was confirmed by a routine gynecological examination. Hyperglycaemia (197.7 mg/dl, normal range 62-134 mg/dl) (Eades e Bounous, 1997) and hypostenuria (SG 1002) were the only abnormalities detected. A presumptive diagnosis of PPID was made and a ACTH stimulation test was carried out using esacetate tetracosaetide (SynacthenA, Novartis Farma) at the dose of 100 IU/kg IV. Basal and two hours post-ACTH cortisol plasma concentrations (RIA method, DPC, USA) were 146 nmol/l (normal range: 25-155 nmol/l) and 366 nmol/l (normal range: <400 nmol/l), respectively. ACTH (chemiluminescent enzyme immunoassay, DPCA, USA) and insulin (chemiluminescent enzyme immunoassay, DPCA, USA) plasma concentrations were also measured and were 155 pg/ml (normal range: ≤35 pg/ml - Perkins et al., 2002) and 241 mUI/ml (normal range: 5.40-36.00 mUI/ml), respectively (Beech e Garcia, 1985; Beech, 1987; Reed, 1998) . In order to establish the normal range of equine ACTH plasma concentration we analysed plasma samples obtained from 5 healthy horses ranging from 17 to 22 years of age. Our range was 13.4±2.23 pg/ml, basically similar to the values reported by others (Couetil et al., 1996; Perkins et al., 2002) . The ACTH and insulin plasma concentrations were dramatically above normal values and were compatible with the diagnosis of PPID. The horse was treated with a dopamine agonist: pergolide mesylate (NoparA, Lilly, Florence, Italy) initially administered PO at a low-dose: 0.5 mg/24 h to avoid adverse effects of the drug, and gradually increased by 0.5 mg/24 h every 3 days to a final dosage of 3 mg/24 h (Munoz et al., 1996) . Antibiotics were administered to treat the respiratory problems. Most of the clinical signs (lethargy, anorexia, PU/PD) resolved in about 6-7 weeks and ovarian activity in 8 weeks as previously observed by others (Reed, 1998) . ACTH 379 plasma concentration was monitored after 45 days and 14 months and was 26.8 pg/ml and 27.7 pg/ml, respectively. At the current time the mare is still in good health and has normal ovarian activity. 25-years old mare, French saddle horse. The history reported recurrent laminitis lasting several months, hirsutism, pendulous abdomen, bulging supraorbital fat pads and a recent severe progressive polyuria and polydypsia (120 L of water every day). The haematology and biochemical profile were unremarkable and urinalysis showed hypostenuria (SG 1007). A presumptive diagnosis of PPID was made on the basis of characteristic clinical signs. A definitive diagnosis of PPID was made by measuring the ACTH plasma concentration: the value was found to be more than twenty fold above the normal range: 370 pg/ml, while the insulin plasma concentration was within the normal range: 15.5 mUI/ml. The mare was treated with pergolide mesylate using the same protocol reported before. The definitive dosage was increased up to 4 mg/24 h after one month of therapy because no improvement was observed after four weeks. Clinical signs ameliorated after about six weeks despite a slight persistent PU/PD. The ACTH plasma concentration was monitored after two and four months and was 91.9 pg/ml and 59.1 pg/ml, respectively. At the current time the mare is still in good health, despite a slight persistent PU/PD. Equine PPID is a frequent disease in old horses, although many presumptive diagnoses are probably not confirmed because of the owners' poor compliance and sometimes due to the limited number of laboratories that perform horse ACTH plasma concentration analysis. In the reported cases PPID was suspected at the time of the 380 visit due to the presence of most of the classical clinical signs of the disease. The definitive diagnosis was strongly supported by the increased ACTH plasma concentration found above >35 pg/ml which is recognised as the critical value with a sensitivity of 100% (Perkins et al., 2002) .In contrast with the ACTH-stimulation test and dexamethasone suppression test, only a single blood sample is necessary to measure the basal ACTH concentration, which seems to be more practical and time saving.Pergolide mesylate treatment was successful for both horses and no side effects were noticed even though anorexia, sweating, dyspnoea, dry mouth, diarrhoea, colic and laminitis are reported as a consequence of higher doses of this drug (Beech, 1999; Munoz et al., 1996; Van der Kolk et al., 1997) . Furthermore, the restoration of fertility in ''case 1'' permitted successful use of the mare as embryo donor. In both cases the improvement of clinical signs seemed to be associated with reduced levels of ACTH plasma concentration that for the ''case 1'' mare returned within the normal range.In conclusion, the ACTH chemiluminescent immunoassay provides a good opportunity for veterinarians to confirm the suspicion of PPID and pergolide is the first drug of choice for its treatment, especially considering that only 1-4 tablets/day have to be given.