key: cord-0008449-lqhs304n authors: Tzipori, S. title: Cryptosporidiosis in Perspective date: 2008-05-05 journal: Adv Parasitol DOI: 10.1016/s0065-308x(08)60353-x sha: 474964ba9cac4aa74fe3a4233c81973302594714 doc_id: 8449 cord_uid: lqhs304n This chapter focuses on cryptosporidiosis in perspective. The significance of cryptosporidiosis in humans, however, depends on the severity of the disease it produces and the incidence in population, both of which are unknown. The frequency of cryptosporidiosis is highest in children aged between six months and three years. The chapter discusses these epidemiological observations and more accurate information regarding the morphology and the life-cycle of Cryptosporidium. Detailed studies on the life-cycle and the ultrastructure have provided a better understanding of the biology of the parasite and have highlighted its unique characteristics. Of these, the existence of sporulated, thin-walled oocysts, and their independence of reducing conditions for excystation are of major importance and may explain the occurrence of autoinfection and persistent infection. Cryptosporidium has numerous characteristics that set it apart from the rest of that subclass. The parasite's ability to maintain persistent infection in certain individuals and its astonishingly stubborn resistance to chemotherapy are serious medical problems that need to be addressed in the future. under one year age groups (Black, 1985) . Cryptosporidiosis, which is emerging as a human disease, is undoubtedly responsible for some of these mortalities. In the concluding remarks made in an earlier review (Tzipori, 1983) , I wrote "Humans are probably susceptible to disease as some evidence exists to suggest that it [cryptosporidiosis] may be a cause of transient, mild, or acute diarrhoea1 illness in immunologically normal individuals. The significance of cryptosporidiosis in humans, however, depends on the severity of the disease it produces and the incidence in the population, both of which are unknown". Within a relatively short time, over 80 communications including letters, commentaries, editorials, case-reports, epidemiological surveys, which included clinical observations, age distribution, seasonality, and many other aspects of the disease in the general population, have since been published in the scientific literature. Consequently, a major portion of this review is devoted to the analysis of these epidemiological observations which, I hope, should provide firm answers to earlier questions about severity and incidence in the human population. More accurate information regarding the morphology and the life-cycle is now available and a detailed outline is included. However, the information available is not yet sufficient to permit full understanding of the nature and the unique and variable behaviour of Cryptosporidium in different classes of vertebrates, different hosts, and their various body systems and organs. Although undoubtedly a coccidian, Cryptosporidium has numerous characteristics which set it apart from the rest of that subclass. The parasite's ability to maintain persistent infection in certain individuals and its astonishingly stubborn resistance to chemotherapy are serious medical problems that need to be addressed in the future. Cryptosporidium is a genus in the family Cryptosporidiidae, suborder Eimeriina, order Eucoccidiida, subclass Coccidia, class Sporozoa, phylum Apicomplexa (Levine, 1980) . At present, the suborder Eimeriina contains 13 families with over 1500 named species. Most of these species, however, belong to the genera Eimeria and, less commonly, Zsospora. Both are intracellular parasites which primarily infect the intestinal tract of vertebrates. Toxoplasma and Sarcocystis, tissue cyst-forming coccidia, are two other important members of the suborder Eimeriina. Cryptosporidium was so named by Tyzzer in 1907 to signify that it is a sporozoon (belonging to the class Sporozoa) in which spores are indistinguishable, absent or concealed (crypto in Latin) in the oocyst. The genus was subsequently established in a family of its own by Lkger in 191 1. The occurence of more than one species within the genus was proposed in 1912 by Tyzzer (I 9 12) on the basis of transmission experiments which he conducted in mice, and the assignment of a new specific name to each additional new animal isolate continued up to 1980. By 1985 Cryptosporidium had been reported in more than 20 species of animals. Cross-transmission experiments, conducted earlier using organisms obtained from guinea-pigs (Vetterling et al., 1971a) and cats (Iseki, 1979) , supported the notion of speciation. But since 1980 evidence has been accumulating which suggests that host species specificity is not a characteristic shared by all, or even most, isolates of Cryptosporidium (Tzipori et al., 1980a; Reese et al., 1982) . Consequently Levine (1984) , in ti review of the taxonomy of the genus Cryptosporidium, tentatively nominated four species representing isolates from mammals (C. muris), birds ( C . meleagridis), reptiles (C. croteli) and fish (C. nusorum), to which he assigned all other known isolates. The nomination of these species is largely based on insufficient information rather than on experimental and extensive epidemiological evidence. Therefore, further studies are required before Cryptosporidium isolated from various sources can be firmly allocated to different species. Because they appear to be morphologically and, from limited serological studies, antigenically (Tzipori and Campbell, 198 1) indistinguishable, assignment to species is probably premature. In other species of coccidia, species within the same genus often show some morphological and biological variation from each other, which has presumably evolved through adaptation to a particular group of hosts. However, Cryptosporidium, presumably because of its ability to exist in different species, has acquired no peculiar features unique to a particular host. Lack of host specificity, at least among domestic, or even all, mammalian species, is one of the major characteristics that sets Cryptosporidium apart from the rest of the coccidia. Unlike Toxoplasma, the only other well known member of Eimeriina that infects a wide range of mammals, which requires two hosts to complete its life-cycle, Cryptosporidium can completely do so in one host. Sarcocystis, on the other hand, consists of numerous species-often three or four within a single host. The lack of specialization-adaptation to a single cell type, location or host-in the evolutionary sense, is usually interpreted as a lack of sophistication. However, adaptation to a broad host range may represent an advanced stage of evolution. Although the case for a monospecific genus (Tzipori et al., 1980a) remains valid for the time being, there is sufficient evidence to indicate that some subtle biological differences exist in terms of preference for a particular host or particular location within the host. However, this may reflect "strain" differences. There is little doubt that the organism which infects the intestine can also infect the trachea, as demonstrated in both immunocompromised and immunocompetent patients with concurrent infections (Forgacs et al., 1983; Kocoshis et al., 1984; Harari et al., 1986) , and by experimental inoculations of animals (Tzipori, 1983; Heine et al., 1984b; Lindsay et al., 1986 ). Yet, under natural conditions, some birds suffer mostly from respiratory infections (Hoerr er al., 1978) , while others have infections of the gut (Tyzzer, 1929) . Some strains of Cryptosporidium appear to infect one site consistently, often in the same host, while others prefer a different site. Tyzzer (1910) distinguished C. muris when he observed the parasite in the gastric mucosa and was able to demonstrate its predilection for this site by transmission experiments. Two years later (Tyzzer, 1912) , he did a similar experiment with an organism which had a preference for the ileal mucosa of the mouse, instead of the stomach. He subsequently called the ileal species, which also had slightly smaller oocysts, C. parvum. Upton and Current (1985) , who also identified two morphologically similar oocysts of different size in cattle, recommended that two distinct species, as described by Tyzzer ( 19 10, 191 2), be recognized. However, they provided no further distinguishing biological features, e.g. site predilection, variation in pathogenicity or antigenic variation. One suspects that Tyzzer (19 12) , in making the distinction, may have been impressed by site predilection as much, if not more, than by variation in the size of oocysts. Upton and Current (1987) warned, appropriately, against naming new species based on host specificity in the absence of adequate cross-transmission data, or careful examination of endogenous development, particularly when oocysts are structurally indistinguishable. There is little doubt that some isolates of Cryptosporidium infect some species of animals more readily than others, suggesting again a degree of adaptability. Cryptosporidium isolated from adult mice readily infected other adult mice (Tyzzer, 1910) ; similarly, isolates from cats (Iseki, 1979) and guinea-pigs (Vetterling et al., 1971a) infected young adults of the same species. In my experience, and that of others, Cryptosporidium isolated from humans, calves, deer, lambs and goat kids could infect infant mice but not adults (Sherwood et al., 1982; Tzipori, 1983) . The nature of the differences between Cryptosporidium isolates requires further investigation. Cryptosporidium from sources other than domestic animals or humans should be examined in newborn laboratory animals, free of specific antibody, and in cell cuiture for evidence of biological and morphological differences. Experimental studies on organisms found in unusual hosts such as fish or snakes, or from peculiar sites such as the conjunctival sac, trachea, or kidney of infected birds, will no doubt help to identify the range of infectivity of Cryptosporidium. Sophisticated studies with monoclonal antibodies, possibly one-dimensional fingerprint analysis, and iso-enzyme studies, will help to define differences between isolates. Cryptosporidium isolates endemic among humans and domestic animals may prove to be more closely related and more interchangeable than those found in wild animals which exist in greater isolation. The life-cycle of Cryptosporidium has been outlined by several investigators. Broadly speaking, it follows closely the pattern characteristic of other coccidia; asexual followed by sexual endogenous stages resulting in production and discharge of oocysts in the faeces (Fig. 1) . The most detailed study was the very first, by Tyzzer in 1910 and 1912 . With minor exceptions he established the outline accepted today without the aid of sophisticated equipment now available. He described the existence of a minute oocyst with four naked sporozoites (without sporocysts) and identified the parasite's u'nique potential for autoinfection, which plays a crucial role in the pathogenesis of the infection in immunologically compromised hosts. For the next 75 years, various investigators have re-examined the lifecycle, latterly with the aid of electron microscopy. Vetterling et al. (1971b) FIG. I . Schematic representation of the life-cycle of Cryptosporidium. 1st. = first generation, 2nd. = second generation. DW = double-walled, Exc = excystation, Mac = macrogamont, Mer = merozoite. Mic = microgamont. Micg = microgamete, 00 = oocyst, Sc = schizont, Sp = sporulated, Spz = sporozoite, SW = single-walled, Tr = trophozoite, Usp = unsporulated, Zy = zygote. described second-generation schizogony in the guinea-pig and suggested that these schizonts were structures previously described by Tyzzer (19 10) as sporulated oocysts. Pohlenz et al. (1978a) confirmed their existence, showing that second-generation schizonts produced four merozoites. They also identified, as oocysts, structures similar to those described by Tyzzer (1 910). Pohlenz et al. (1978a) and Iseki (1979) demonstrated that oocysts in the faeces can be used for diagnosis of cryptosporidiosis in calves and cats. In studies in mice, chicken embryos and cell culture, Current and co-workers (reviewed by Current, 1985) observed two kinds of oocysts which differed in the consistency of the wall; thin-walled oocysts were identified which may hold the key to autoinfectiorl. The issue of whether Cryptosporidium is intracellular or extracellular appears to have been resolved, and the term "intracellular-extracytoplasmic" was coined by Goebel and Braendler (1982) , who, in addition, provided useful details on the gametogony of the parasite. The life-cycle of the parasite begins with the intake of oocysts orally, and possibly by inhalation. Oocysts undergo excystation which, in vitro at least, requires the combined action of trypsin and bile salts (Current and Haynes, 1984) . Excystation releases four naked, non-flagellated sporozoites which are morphologically indistinguishable, as yet, from two types of merozoites. They approach and enter the microvillous epithelial border in the small intestine by flexing and twisting movements to initiate infection (Fig. 2) . The sporozoite (and later merozoites) indents the microvillous membrane, invaginating it in a glove-like fashion (Fig. 3) . The double unit membranes of the host cell extend along the surface of the parasite, finally covering it entirely and forming a parasitophorous envelope which encapsulates the parasite within a parasitophorous vacuole. Electron-dense bands are formed in the host cell cytoplasm opposing the parasite ( Fig. 3(B) ). (a) Host cell invasion and the host parasite interface. This process was studied by Marcia1 and Madara (1986) , employing high resolution electron microscopy and freeze-fracture techniques on ileum from infected guinea pigs. Their study was confined to early and mature trophozoites but presumably similar processes apply to other intracellular forms. The redundant folds of host cell membrane envelop the organism, resulting in the intracellular localization of the parasite within a sac of internalized microvillous membrane. The plasma membrane of the parasite subsequently fuses towards its base with the invaginated host membrane. The two membrane domains isolated by this process subsequently undergo drastic alteration. The host membrane dissolves, and the isolated parasite plasma membrane, which is now in direct contact with the host cell cytoplasm, becomes amplified. During this process, the inner unit membrane of the parasite A Merozoite penetrating mucosal surface by invaginating the microvillous membrane. A constriction by the infolded microvillous membrane, which has not yet been lysed, is evident around the trophozoite. The nucleus is poorly defined but the electron-lucent appearance of the rhoptries (RH) suggests that they have released their contents. (B) Enlargement of marked area, showing other membranes that can be distinguished at this early stage, including the inner microvillous (IM) membrane, and inner (IP) and outer (OP) parasite or pellicle membranes. Three dense bands (DB) in the cell cytoplasm are distinguished which in later stages appear as one band. pellicle disappears, particularly toward the base of the parasite. Membrane invagination is the method of host cell entry for other coccidia (Long and Speer, 1977) . The outer membrane of the pellicle at the attachment zone is thrown up into numerous folds at the base of the parasite to form the "feeder organelle" from which the parasite derives its nutrients directly from the host cytoplasm (Goebel and Braendler, 1982) . The dense bands underlying the parasite attachment site are areas of modified host cell cytoskeleton, the function of which, it is thought, is to help anchor the parasite to the host cell, or resist further invasion into the absorptive cell cytoplasm, or both (Marcial and Madara, 1986) . The vesicles formed next to the feeder organelle have the function of pinocytosis (Goebel and Braendler, 1982) . (6) Trophozoites. After penetration, the merozoite rounds into a spherical body and undergoes differentiation until all organelles are resorbed except the nucleus, Golgi anlagen, pellicle and cytoplasmic ribosomes. Concurrently, the nucleus and nucleolus enlarge. Redifferentiation includes development of the endoplasmic reticulum adjacent to the nucleus, and derivation of ribosomes from the nucleolus. The vacuolated zone with interdigitated membranous folds develops adjacent to the cell cytoplasm ( Fig. 4) . Trophozoites are surrounded by five unit membranes, of which the outer two are of host origin-being the inner and outer membranes of the envelope-and the other three correspond to the pellicle; one is the parasite plasma membrane and the remaining two are the inner, double, unit membrane, intrinsic to the parasite, which is not always distinguishable (Fig. 5) . (c) Schizogony. The trophozoite nucleus divides either three times, to form eight merozoites characteristic of first-generation schizogony or twice, to give rise to four merozoites which is typical of second-generation schizonts. During these divisions the size of the nuclei decreases. As schizogony progresses, the pellicle of the schizont invaginates deeply and covers each of the divided nuclei and the cytoplasmic mass containing the rough endoplasmic reticula. Each merozoite is surrounded by a pellicle composed of three membranes-an outer and two inner (Iseki, 1979; Marcial and Madara, 1986) . Merozoites have an internal structure similar to that described for other coccidia, which includes dense granules, micronemes, nucleus, Golgi complex, conoidal complex, rhoptries, endoplasmic reticulum and polar ring (Vetterling et af., 1971b) . At the end of the process of schizogony, the parasitophorous vacuole contains eight or four free merozoites, depending on the generation of schizont, a small mass of residual cytoplasm of the schizont, a round body or a vacuole, and the attachment zone at the base (Fig. 6) . FIG. 4 . Trophozoites. Four different stages of maturation can be seen, from soon after attachment and rounding up (A), to first nuclear division (D). There is no telling whether they are destined to be firstor second-generation schizonts. (A) Note characteristic nucleus (N) and large nucleolus (NU) within, a double membrane pellicle and not yet fully formed attachment zone. There is no endoplasmic reticulum at this early stage. There is a clear separation zone between the host cell and parasite pellicle, and the parasite membrane has not yet begun to resolve. The inner (IM) and outer (OM) microvillous membranes and inner (IP) and outer (OP) pellicle membranes are apparent; note that IP here appears as a double unit membrane. (B) The rough endoplasmic reticulum (ER) has begun to fill the cytoplasm, the Golgi complex (GC) is visible, and the inner membrane of the pellicle has been absorbed along the area of attachment. The parasitophorous vacuole (P) is enlarged. (C) Fully developed trophozoite before schizogony. The nucleolus has disappeared, a vacuole (V) has appeared, and dense granules, which are later seen in the merozoites, are also present. The attachment zone (AZ) has become vacuolated with interdigitated membranous folds forming the feeder organelle. (D) One nucleus is visible after the first nuclear division, and the ER fills the cytoplasm. Fig. 4(A) , showing an area at the interface below and above the terminal web. Note that the outer microvillous membrane (OM) is covered by glycocalyx of the same consistency as, and continuous with that of, the cell microvilli. The inner microvillous membrane (IM) is internalized (by invagination) and is also covered by glycocalyx; there is a thin layer of cell cytoplasm between OM and IM. At the base, the two units of the inner pellicle (IP) have begun to regress (single arrow), and the IM and outer pellicle (OP) are becoming fused (three arrows). Later, the IM dissolves leaving the OP which forms infoldings and is the only structure separating the cell cytoplasm from the parasite. The "dense band" (DB) is of host origin with no unit membrane structure and therefore is not a true interface between parasite and host cell. Macrogamonts, the female forms of which measure &5pm in diameter, are found in abundance in the brush border, second in number only to trophozoites. The macrogamete is found within a parasitophorous vacuole. It has a large nucleus, situated eccentrically, with a distinct nucleolus; a round membrane-layered lipid vacuole is normally found next to the nucleus. In addition, the macrogamete typically contains in the cytoplasm a large number of polysaccharide granules, electron-dense bodies which are thought to be products of the reduction division (Tyzzer, 1907) or maturation bodies, wall-forming bodies, and rough endoplasmic reticulum. As with all other forms, a vacuolated membranous attachment zone is found at the base (Fig. 7) . Microgamonts are found less frequently than other forms, with the possible exception of sporulated oocysts. They measure 4-5 pm, are similar to schizonts in shape and may contain 14-16 microgametes and a residual body. The microgametes are rod-shaped and the nucleus dominates the small amount of the cytoplasm. During microgametogenesis the nuclei of the microgametes escape into the parasitophorous vacuole by budding. They are thereby covered by the pellicle of the microgamont, which, after the fissions of the nuclei, evaginates into the parasitophorous vacuole. The freed microgamete is wedge-shaped, measuring just under 1 pm in length with a thickened apical pole. The microgametes of Cryptosporidium most closely resemble those of other'Eimeriina, except that they lack flagella. This is interpreted by some (Goebel and Braendler, 1982) to indicate a primitive stage of development. Therefore the microgametes are either immobile and are passed to the macrogamete by intestinal flow, or the intracytoplasmic microtubules (Goebel and Braendler, 1982) may have a role in locomotion by inducing flexing and twisting in a manner similar to that of sporozoites ( Fig. 8 ). Fertilization is achieved by protrusion of parts of the macrogamete membrane towards the microgamete. The adhesion zone of the microgamete is attached to the parasitophorous vacuole of the macrogamont, indenting all the wall membranes. Fertilization is most probably by penetration of the microgamete into the macrogamete (Fig. 9) . (e) Oocyst formation. Formation of the oocyst and often, but not always, sporogony take place in the parasitophorous vacuole. Oocysts are formed from the fertilized macrogamete which undergoes successive changes before and after fertilization (Fig. 7) . Sporulated oocysts, which measure 4 6pm, contain four naked C-shaped sporozoites, which are surrounded by a pellicle and structurally similar to merozoites. Sporozoites contain numerous micronemes, electron-dense bodies, electron-pale vacuoles and highly condensed ribosomes in their cytoplasm. Oocysts with both thin and thick walls have been identified. Both types are sporulated when discharged in the faeces. It has been suggested that the thin-walled oocysts, which excyst mainly within the same host, are responsible for autoinfection (Current, 1985) . I have distinguished single-and double-walled oocysts, which presumably correspond to the thick-and thin-walled forms (Fig. 10 ). More recently, larger oocysts similar in size to those of C . muris (7.4-5.6pm) described by Tyzzer (1907) , have been identified in a small proportion of bovine faeces. Upton and Current (1 985) considered the difference in size of these oocysts sufficient to regard them as belonging to a separate species. The duration of the life-cycle seems to vary from a minimum of 48 hours to as long as 1&14 days before the first appearance of oocysts in the faeces of FIG. 8 . Microgametogony. Microgametogenesis proceeds in 2 phases; a growing phase with multiple fissions of nuclei resulting in 16 microgametes, followed by differentiation during which the gametes evaginate into the parasitophorous vacuole (P). (A), 5 microgametes budding from the microgamont, pellicle of which forms the basis of the pellicle of the microgamete (MIC). Note dense nuclei which are readily distinguished from those of merozoites. (B), free microgametes sectioned in different planes with one in longitudinal section. (C), free wedge-shaped microgamete with a dense nucleus (N) occupying most of the space within 2-unit pellicle. some animals. Oocysts can be detected within 48 hours after inoculation of piglets, lambs, calves and tissue culture. The prepatent period is longer in rodents and in chicken embryos, the shortest being normally 4 days. Duration of the cycle seems therefore to depend on the host, but it is not clear whether some or all steps of the life-cycle are more rapid in a susceptible host, or whether there is a fundamental difference in the sequence. It is also possible that the earlier appearance of oocysts in faeces of clinically affected animals is due to the diarrhoea. It has been established through sequential studies that sporozoites always give rise to first-generation schizonts containing eight merozoites. However, it is not clear what proportion, if any, of firstgeneration merozoites repeat the first-generation schizogony-thus producing more of the same type-and for how long, and how many go on to produce second-generation schizonts containing four merozoites. Nor is it known whether all second-generation merozoites directly form micro-and macrogamonts, or recycle, as do the first generation. Trophozoites of unknown outcome are the predominant forms seen under the microscope in a given infected intestinal section, followed by macrogamonts, and then schizonts containing eight merozoites. This may merely indicate a more prolonged association of these forms with the host rather than their relative prevalence in the gut. However, if the duration of both schizogony generations is similar, it is suggested that first-generation merozoites produce both first-and second-generation schizonts simultaneously, while second-generation merozoites form only micro-and macrogamonts. FIG. 9 . Fertilization. Unlike the description by Goebel and Braendler (1982) , who elegantly demonstrated fertilization of a macrogamete by attachment of the microgamete with the blunt end (A) and (B) depict probable fusion along the long axis; they may represent a more advanced stage. (A) A microgamete is trapped between two macrogametes, indenting the wall of both, with evidence (under higher magnification) that the pellicle of the upper one, which is closer, is interrupted. (B) An oblique section of a macrogamete with what seems to be a microgamete being surrounded by the parasitophorous envelope. (C) Macrogamete in which a probable (blurred) microgamete can be distinguished (six arrows). A thickened area in the parasitophorous envelope opposite suggests a recent "reinforcement" after entry(?) of the microgamete. Penetration and dissolution are rarely seen and must be extremely rapid. FIG. 10 (overleaf). Sporogony. Oocyst formation and sporogony take place in the parasitophorous vacuole shortly after fertilization. (A) Single-walled oocyst containing four newly formed sporozoites (Spz) with a still relatively large residiuum containing a lipid vacuole (LV) and polysaccharide granules (PG). (B) Three fully developed sporozoites within a single-walled oocyst (SW) still inside the parasitophorous vacuole (P). (C) Double-walled sporulating oocyst (DW) with large residiuum within the parasitophorous vacuole. Note rigid wall, in contrast to singlewalled oocysts; a newly fertilized macrogamete containing endoplasmic reticulum (ER) is on the right. (D) A fully sporulated double-walled oocyst showing three sporozoites (Spz) with no apparent trace of the residium. The double wall invariably collapses with processing; AZ = attachment zone. A. EPIDEMIOLOGICAL OBSERVATIONS The history of human cryptosporidiosis is fairly recent, the first two cases having been described by Nime et al. and Meisel et al. in 1976. Bird and Smith (1 980) showed that human Cryptasporidium was morphologically indistinguishable from forms described from other animals. By 1980 further cases had been described in individuals with some clinical evidence of acquired (Weisberger et al.: 1979; Stemmermann et al., 1980) or congenital (Lasser et al., 1979) immunodeficiency. Cryptosporidiosis in these patients was confirmed by histological examination of intestinal biopsies. The first case of cryptosporidiosis in an immunocompetent adult, who suffered severe, watery but self-limiting diarrhoea and vomiting, was described by Tzipori et al. (1980b) . This patient was the first in whom the diagnosis was made by detection of oocysts in faecal smears. The transmission of this human isolate to mice and lambs not only showed that it was biologically indistinguishable from Cryptosporidium *of other mammals, but also indicated indirectly the zoonotic potential of this parasite (Tzipori et al., 1980a (Tzipori et al., , 1982a , as subsequently confirmed by Reese et al. (1982) . The emergence of acquired immunodeficiency syndrome (AIDS) and AIDS-related infections promoted human cryptosporidiosis to the forefront of interest in 1982. The prevalence of the infection in the general population also became apparent with the publication of the first epidemiological studies in 1983 (Jokipii et al., 1983; Tzipori et al., 1983b) , although it had been suspected as early as 1980 from a limited study conducted on patients with gastroenteritis in a hospital in Newcastle, UK (S. Tzipori, R. Madeley and K.W. Angus, unpublished observations) . Ironically, at the time the information was considered by two medical journals not to be of particular relevance. The zoonotic potential of cryptosporidiosis was confirmed by accidental infections of humans who had close contact with infected calves Current et al., 1983; Rahaman et al., 1984) . The early epidemiological studies were followed by numerous reports from around the globe. The following aspects of the infection in humans will be discussed: the frequency of diarrhoea attributed to it in various regions; clinical manifestations; seasonal variation; association with other infections and travel; the mode of transmission; and asymptomatic infections. Tables 1, 2 and 3 summarize the frequency with which Cryptosporidium oocysts have been detected in faeces in various studies of human diarrhoea; they indicate a worldwide prevalence of 1 4 % among patients with diarrhoea in most developed countries, and up to 16% in some less developed countries. ?Studied during a limited season only. OStudied during a limited period. h2 is urban and 3 is rural.