key: cord-0006744-aavpj5r3
authors: Schwarte, L.A.; Zuurbier, C.J.; Ince, C.
title: Mechanical ventilation of mice
date: 2000
journal: Basic Res Cardiol
DOI: 10.1007/s003950070029
sha: ac41df6d6c5bb4ee5f722829c53934fb8badc7e5
doc_id: 6744
cord_uid: aavpj5r3

Due to growing interest in murine functional genomics research, there is an increasing need for physiological stable in vivo murine models. Of special importance is support and control of ventilation by artificial respiration, which is difficult to execute as a consequence of the small size of the animal and the technically demanding breathing pattern. In addition, numerous genetically altered mice show depressed spontaneous ventilation or impaired respiratory responses. After an introduction in murine respiratory physiology we describe options for ventilatory support, its monitoring and the potential side effects. This review will provide an overview on current possibilities in the field of airway support in mouse research.

rate (RR) of 79 min Ð1 , a ventilatory tidal volume (Vt) of 25µl, an inspiratory fraction (inspiratory to total time of the respiratory cycle, Ti/Tt) of 0.44 and a resulting ventilatory minute volume (Ve) of 1.98 mlámin Ð1 was found. Similar values are found for wild type and endothelin-1 knock-out mice (ET (Ð/Ð) ) when birth was achieved by cesarean section (30) . Rather low respiratory values were found in a study comparing wild type neonates with littermates deÞcient for the RET-protooncogene: RR ranged from 28Ð40 min Ð1 , Vt from 7Ð10 µl and VE from 0.26Ð0.32 mlámin Ð1 , with no significant differences between the groups. The authors suggest that a reduced body temperature in their setting contributed to their Þndings (4) . With respect to control of breathing, newborn wild type mice (1 g bodyweight) show similar ventilatory responses to hypercapnia compared to other newborn mammals (34) . Neonates deÞcient in endothelin-1 (129Sv/J x ICR, ET-1 (Ð/Ð) ) have an attenuated ventilatory response to hypoxia and hypercapnia, compared to their wild type littermates (30) . A signiÞcantly reduced ventilatory response to hypercapnia was also found in neonate mice deÞcient in the RET-protooncogene, compared to wild type littermates, whereas no consistent differences were obtained during hypoxic challenge (4) .

Parameters for the spontaneous ventilation of adult mice are given in Table 1 , where the following characteristics of mice ventilation are given: RR = 180Ð270 mlámin Ð1 , Vt = 0.1Ð0.2 ml and Ti/Tt = 0.30Ð0. 35 .

Little information is currently available on the lung mechanics of mice (37) . Murine lung-thorax compliance was measured in pancuronium-relaxed, mechanically ventilated mice (BALB/c, no weight stated) in the range of 31Ð38 µl (cmH 2 O) Ð1 under control conditions, and 12 µl (cmH 2 O) Ð1 after experimental surfactant inactivation (50) . Respiratory system resistance was measured in age matched A/J-and C3H/HeJmice, revealing signiÞcant strain differences under baseline conditions (1 vs. 2 cmH 2 Oám/ Ð1 ás Ð1 ) and especially after i.v. acetylcholine treatment (3 vs. 15 cmH 2 Oám/ Ð1 ás Ð1 ) (11) . In the same study, signiÞcant differences were also found for the respiratory system elastance under both conditions (25 vs. 30 cm H 2 Oáml Ð1 at baseline and 28 vs. 95 cm H 2 Oáml Ð1 after i.v. acetylcholine). Again, one has to realize that marked differences exist between mouse strains in respect to respiratory physiology and pathophysiology.

Divergent data is available on physiologic values for murine arterial blood gasses, regarded as the gold standard for the evaluation of adequacy of ventilation. Information on tissue PCO 2 of the mouse indicates that it Þts within the normal mammalian range (53) . Later it was suggested that mice have considerably lower alveolar and arterial PCO 2 than other mammals, as low as 20 mmHg (at pH 7.4) (31) . Indeed, recent data support these lower PCO 2 (24 mmHg in chronically instrumented 129SV/J x ICR mice (30) , 34 mmHg in chronically instrumented C57BL/6 mice (39)). Although effects of (30, 39) or restrainment cannot completely be excluded in these studies, the data does indicate a lower PaCO 2 for the mouse, e.g., compared to mammals ranging from rats to humans (PaCO 2 33Ð41 and 35Ð45 mmHg respectively).

Mechanical mouse ventilation is indicated unter the following circumstances. First, the ÒanesthesiologicalÓ indication: one of the major goals of anesthesia is achievement of more (39, 52) , mice lacking the brain derived neurotropic factor BDNF (1), RET-protooncogene deÞcient mice with depressed ventilatory response (4), and endothelin-1 deficient mice with altered blood gas-values (e.g., signiÞcantly lower PO 2 than wild type littermates) and impaired respiratory response to hypoxia and hypercapnia (30) . However, it is often unknown when during development and under which conditions alterations occur in the respiratory system or the control of breathing (27) . Also among inbred wild type strains the control of respiration differs substantially, e.g., the C57BL/6-strain was classiÞed as highly responsive to hypercapnia, whereas the DBA/2-strain was highly responsive to hypoxia (51) . Therefore, in these animals mechanical ventilation offers the possibility to study effect of differences in genetic background independent of changes in ventilation. Second, the ÒsurgicalÓ indication: certain surgical procedures have an improved outcome or can only be performed with support of mechanical ventilation. Examples are the bilateral ligation of the carotid artery (36) , where non-ventilated mice often die from respiratory disturbances, or the growing Þeld of cardiac research that needs an open-chest murine model (15, 18, 29, 44) . Hereby opening of the thorax and the adherent interpleural space induces lung collapse with respiratory insufficiency, whereas with support of a mechanical respirator the lungs can be inßated and ventilated.

There are several possibilities of accessing and maintaining the airway in mice. In early studies using a mechanical ventilator with active in-and expiration for mice (28) , the airway was accessed by advancing a tube through the mouth into the pharynx, but not further into the trachea. Although technically less demanding than endotracheal intubation, this attempt risked gastric gas-insufflation (including aspiration of stomach content) and did not provide PEEP capability. A different, nontraumatizing mode to access the airways is performed with the Brady-Newsom apparatus (3): it consists of a gas-chamber around the murine head, sealed at the thorax by a tight neoprenecollar. Elevating the pressure in this chamber leads to increased airway pressures, with respect to ambient air. Thus this concept provides continuous positive airway pressure (CPAP) if the mouse ventilates spontaneously, or mechanical ventilation, if periodic pressure changes in the chamber are generated. Although this apparatus has advantages, e.g., the technically less demanding airway access, it is also hampered with the disadvantage of gastic gas insufflation and aspiration of gastric content. To circumvent this problem, mechanical ventilation of mice today is usually performed after securing the airways by orotracheal intubation or via tracheotomy and consecutive endotracheal intubation. Tracheotomy has been performed in preterm neonate mice (1.15 g bodyweight) with insertion of polyethylene tubes (SP8, Natsume, Tokyo, Japan; tapered 0.4 mm outer diameter (OD)) (30), whereas for adult mice the OD of the endotracheal tube ranges from 0.9 mm (25Ð38 g mice) (29) to 1.1 mm (20Ð35 g mice) (15) . For experiments requiring recovery after surgery, temporary orotracheal intubation is performed, ideally via direct laryngoscopy with a purpose-made laryngoscope (13) . The visualization of relevant airway structures requires a strong small-focus light-source, allowing transillumination of the laryngeal cavity (with vocal cords and tracheal oriÞce), when put on the ventral cervical region from the outside (2). Orotracheal intubation is performed with or without visual conÞrmation of the tubeÕs tip-position (tube position judged by respiratory excursions). Visualization of the trachea (after surgical access to the deeper cervical structures) and controlled advancement of the endotracheal tube allows more accurate positioning of the tube, but is more invasive and has therefore an increased morbidity and mortality.

For animals sacriÞced at the end of the experiment, direct surgical access to the airways is an alternative option. Since most of our mouse surgeries require preparation of deeper cervical structures (e.g., catherization of a carotid artery and jugular vein) performing tracheotomies is convenient, with the additional advantage of minimized dead space (Vd). A reliable surgical procedure for this airway access is brießy described: after induction of anesthesia the mouse is put into the supine position and the head is reclined (silk or rubber band around the upper incision teeth and attached to the operating surface) to allow easy access to the ventral neck. A median cerival skin incision (upper thorax aperture to lower jaw) is performed and layerwise skin, fascia, fat and connective tissue covering the thyroid gland are removed. The thyroid lobes are divided bluntly at their isthmus, pulled aside and kept retracted by bulldogclamps. The pretracheal muscles are spread bluntly and pulled aside to allow access to larynx and trachea. Tracheotomy is performed by a transversal cut between two tracheal rings, usually in the lower third of the trachea. For the endotracheal intubation vascular PE-catheters (Abbocath-T¨, Abbott-Venisystems) are used, cut to a length of about 5 mm with an angled tip (45¡) to allow smooth introduction. The correct position of the endotracheal tube is conÞrmed by judging chest excursions, e.g., if respiration is still visible and symmetric. The tube is secured by ligation (silk 6-0, Davis & Geck¨) around the trachea. If maintenance of spontaneous ventilation is desired, the endotracheal tube is shortened to minimize Vd and airway resistance. For mechanical ventilation, initially applied RR and/or Vt should be sufficiently higher than spontaneous RR and Vt of the anesthetized mouse to decrease respiratory drive (e.g., Òbreathing against the ven-tilatorÓ) by moderate hyperventilation. After connecting the endotracheal tube, anesthesia can be safely deepened for an additional depression of ventilatory drive and facilitation of mechanical ventilation. Alternatively, this goal can be achieved (avoiding additional cardiocirculatory depression) by a non-depolarizing muscle relaxant (vecuronium-bromid 0.25 mgákg Ð1 bodyweight i.p., Norcuron¨, Organon). Using this option it should be realized however that anesthesia depth cannot be judged from parameters like spontaneous movement or limb-withdrawal to paw pinch, but instead on changes in hemodynamics.

A side aspect of murine airway access is the preparation of a ventilated (and perfused) isolated lung. This model has been described for the investigation of cytokines released from a hyperventilated mouse lung (57) . The physiological lung variables obtained from this murine model (female BALB/c, 22Ð30 g body weight) after 60 min of NPV (negative pressure ventilation) are a Vt = 187 µl, dynamic compliance (Cdyn) of 0.022 ml/cmH 2 O and an airway resistance of 0.45 cm H 2 Oás/ml.

There are two basic modes of respiration, spontaneous ventilation and controlled mechanical ventilation (CMV), where work of breathing is taken over by the ventilator. Between these there are a variety of mixed forms, like respiratory assistance, augmenting inspiration if shallow or labored (RSP1002, Kent-ScientiÞc¨, US) or supported spontaneous ventilation with continuous positive airway pressure (CPAP), adopted for mice by Brady and colleagues (3) . However, these mixed modes are not largely established for mice yet, and will therefore not discussed in detail.

Although mechanical ventilation of mice has major advantages, it is confounded by difficulties: obviously, the small size of the animals makes airway access more challenging than in other laboratory animals. For example, in 25Ð30 g mice, the trachea has an accessible length of only 3Ð5 mm and a circumference of about 2.5 mm (49) . Appropriate endotracheal tubes have an outher diameter ranging from 0.4 mm for neonate mice (30) to 1.0 mm for adult mice (13) . The actual mechanical ventilation of the mouse is complicated by the technically demanding respiratory pattern of this species, e.g., the high RR, the low Ti/Tt ratio and the small Vt, with the latter also demanding a minimized deadspace (Vd) to prevent rebreathing. In commercially available rodent ventilators, the relatively large system-Vd allows substantial gas compression, thereby minimizing the Vt delivered to the mouse. Ewart et al. (11) reported that 2 ml of added Vd (compliancẽ 0.002 ml/cmH 2 O) and a peak inspiratory pressure of 40 cm H 2 O reduce the delivered Vt by about 50 %. Thus, mouse ventilators (see Table 2 ) should have a range of technical modalities to enable proper murine ventilation. Since some authors did not report on the type of ventilator used (see Table  3 ) the list might be incomplete. The most obvious difference between spontaneous respiration and usual modes of mechanical ventilation (MV) are inverted pressure relations, with respect to ambient pressure, during the respiratory cycle: during spontaneous inspiration the expansion of the intrathoracic volume (mainly caused by contraction of the diaphragm and extension of the rib cage) generates a negative intrathoracic pressure and allows gas ßow into the lungs. During expiration mainly passive elastic forces of lung and rib cage generate a positive intrathoracic pressure, leading to exhalation of air. Applying usual MV, the air is pushed during inspiration into the miceÕs lungs, leading to a positive intrathoracic pressure. During expiration intrathoracic pressure equilibrates with ambient pressure (intermittent positive pressure ventilation, IPPV) or remains positive (continuous positive pressure ventilation, CPPV). Therefore it is important to notice that mean airway pressure during mechanical ventilation is higher than during spontaneous ventilation. These higher pressures contribute to side effects induced by mechanical ventilation.

Mechanical ventilators are classified according to their method of cycling between the inspiratory and expiratory phases. Small animal ventilators are usually time-cycled: the ventilator switches between the respiratory phases after a certain time has elapsed, depending on the preset respiratory rate.

Although other methods of cycling might be favorable under certain experimental conditions (e.g., volume-cycled to prevent pulmonary volutrauma in murine open chest preparations), they are not established yet. However, although the method of cycling is classiÞed according to a single parameter (e.g., time), some mouse respirators allow a more sophisticated control of the ventilation by limiting relevant parameters: to prevent high inspiratory peak airway pressures, for example, most mouse ventilators allow the limitation of the ventilatory system pressure. Most commercial mouse ventilators are time cycled and pressure limited.

The primary goal of mechanical ventilation is delivery of an adequate respiratory minute-volume (Ve) to the murine 514 Basic Research in Cardiology, Vol. 95, No. 6 (2000) © Steinkopff Verlag 2000 lung. To achieve this, the RR and Vt have to be adjusted to supply the actual ventilatory requirements. RR in most mouse ventilators is the preset cycling parameter, but Vt-lung may vary, depending on the compliance of lungs and tubing and whether respiration is pressure limited. A decrease in compliance of the respiratory system, for example in murine models of interstitial pulmonary diseases (ARDS, pneumonia), increases the airway pressure for a given Vt ventilator . If ventilation is pressure limited, then VT lung will be considerably less than Vt ventilator . In contrast, if higher peak pressures are accepted to maintain Vt lung , the risk to traumatize the delicate murine lung will increase (17) . To prevent the occurrence of these high pressures, inspiration time relative to expiration can be increased for a given Vt (up to inverse ratio ventilation (IRV), with inspiration longer than expiration). Disadvantages of a higher inspiratory fraction during mechanical ventilation are possibly insufficient times for complete exhalation (Òair trappingÓ) and the prolonged phases of elevated intrathoracic pressure, which may cause a decreased cardiac preload and consequently lowered blood pressures (7). A feature provided by some mouse-ventilators is the possible application of external positive end-expiratory airway pressure (PEEP): during expiration, airway pressure is not allowed to equilibrate with ambient pressure (0 cmH 2 O), but kept at a higher level (usually 2Ð10 cmH 2 O). Mechanical ventilation with PEEP prevents the endexpiratory collapse of small airways and possibly recruits atelectatic pulmonary areas for gas exchange by increase of functional residual capacity (FRC). Since pulmonary compliance increases (up to a certain degree) with increased FRC, ventilation with PEEP supports pulmonary inßation. Whether this holds true for mice has not been shown yet. Additionally, ventilation with PEEP might mimic the physiologic end-expiratory tonus of inspiratory muscles and glottis of spontaneous ventilating mice (56) . Typical side effects, demonstrated in a murine model of PEEP, are discussed below.

A list of ventilator settings is given in Table 3 . Most experiments have mechanical ventilation set at RR = 100Ð 150ámin Ð1 , Vt = 0.2Ð0.7 ml, Ti/Tt 0.2Ð0.4. When these values are compared to values of spontaneous ventilating mice (Table  1) , it becomes obvious that lower RR and higher Vt are commonly used to mechanically ventilate mice.

A more rare feature provided by some mouse ventilators is the intermittent sigh-ventilation. Here an inspiration with higher tidal volume (hyperinßation-or sigh-cycle) is administered after a certain number of respiratory cycles (for example every 50 ventilatory cycles) or triggered manually (RSP-1002, Kent-Scientific¨Corporation, Litchfield, CT, USA or 60Ð 1100/1099, Harvard Instruments¨, US). Although not proven, this imitation of physiological sighing is thought to result in recruitment of atelectatic pulmonary areas, without the need for a continuous PEEP, circumventing constant elevation of airway pressures. Intermittent positive pressure ventilation (IPPV) and continuous positive pressure ventilation (CPPV) are the two common modes in murine mechanical ventilation, as provided by commercial ventilators. However, for the ventilation of newborn mice (1.0Ð1.5 g bodyweight, Vt 5Ð20 µl) the concept of the Òiron lungÓ, based on the principle of body surface negative pressure ventilation has been adopted (27) to prolong survival time of NMDA-receptor mutant mice (32) , usually dying within 20 h after birth.

Only a few studies using mechanical ventilation in mice report on additional ventilatory parameters, like inspiratory gas ßow or pressures in the ventilatory system. Using a RR of 140ámin Ð1 , a Vt of 0.7 ml and an inspiratory time of 0.2 s (resulting in a inspiratory fraction (Ti/Tt) of 0.41, Table 3 Data are also available on ventilation regimen for murine open-chest preparations: Guo et al. (18) compared respiratory rates of 95, 105 and 121ámin Ð1 (Harvard rodent ventilator, room air with oxygen (21ámin Ð1 ), Vt = 2.2 ml with endotracheal tube loosely connected) in male ICR mice (8Ð12 weeks old, mean body weight 33.3 g) with respect to the resulting arterial blood gases. Adequate ventilation was obtained with a RR of 105ámin Ð1 , resulting in an arterial PaO 2 of 327 mmHg (177 and 321 mmHg for 95ámin Ð1 and 121ámin Ð1 respectively), an arterial PaCO 2 of 31 mmHg (43 and 18 mmHg for 95ámin Ð1 and 121ámin Ð1 respectively) and an arterial pH of 7.4 (7.3 and 7.5 for 95ámin Ð1 and 121ámin Ð1 respectively). Other regimens for the ventilation of open chest mice are listed in Table 3 .

During controlled mechanical ventilation (CMV), the most frequent applied mode of artiÞcial respiration in mice, the work of breathing is taken over by the respirator. To facilitate this controlled mechanical ventilation (e.g., prevention of Òbreathing against the ventilatorÓ) it is usually performed under deep anesthesia. This ensures that relevant compensatory mechanisms (e.g., onset of spontaneous ventilation induced by hypercapnia or hypoxia) are impaired and underline the importance of monitoring mechanical ventilation. Attempts to monitor mechanical mouse ventilation are complicated by the fact that even standard variables for ventilation are not established yet or show conßicting results (see Table 1 ).

Several possibilities are applicable to monitor adequacy of ventilation. A highly subjective way, requiring experience, is to observe the thorax-excursions with respect to rate, amplitude, symmetry and regularity as indicators of ventilatory rate, tidal volume, endotracheal tube position and attempt to spontaneous ventilatory effort. Dalkara et al. (7) suggest for a physiological mouse ventilation (mean PaCO 2 = 35 mmHg) keeping Òthe volume at a level at which thoracic movements Þrst become noticeably observedÓ. Information on the oxygenation can be obtained from the ear and tailskin color, e.g., with cyanosis indicating poor oxygenation (desoxy-hemoglobin 5 gá100 ml Ð1 ). These measures might serve as a Þrst orientation in the evaluation of the adequacy of mechanical ventilation. More objective and in-depth measures are gas ßows, airway pressures and partial pressures (see below).

The measurement of respiratory gas ßow in mice has been a technical challenge, mainly because pneumotachographs attached to the endotracheal tube increase the ventilatory dead space (Vd). Recently pneumotachographs have been developed for mice that allow both direct airway gas ßow measurement and plethysmography (Kent-ScientiÞc¨Small Animal Pneumotachs, TRN3100, Kent ScientiÞc Corporation). But even modern devices in this setting are suspected to increase dead space to a crucial extent, leading to partial rebreathing and consequently to hypercapnia and eventually hypoxia.

The direct, non-invasive measurement of murine airway pressures is technically difficult; therefore, substitutes have been established to obtain estimates. In commercially available mouse ventilators, the airway pressures are measured inside the ventilator. The advantage of this construction is that it circumvents the need for an additional pressure transducing line, ideally originating from the endotracheal tube itself, to a pressure transducer. A major disadvantage however is that pressures measured in the ventilator need to be extrapolated to airway pressures within the endotracheal tube. This extrapolation is determined by the compliance of the respiratory gases and the tubing system (Òsignal dampingÓ). It is therefore desirable to measure airway pressures as close as possible to the endotracheal tube. With respect to this, we have developed an integrated endotracheal tube within a small aluminum frame (10 x 5 x 2 mm, 5 g), consisting of temperature controlled adapters for in-and expiratory tubing, an inline-capnography cell (Vd = 25 µl) and a liquid-filled airway-pressure transducer. Although airway pressures are usually monitored to maintain safe values for mean-and peak-airway pressure, this measure can also serve to titrate damage in murine models of respirator-induced lung impairment, such as baro-or volutrauma (17) , where tidal volumes (Vt) have been adjusted to achieve traumatizing target airway pressures.

Fresh-gas inßow composition to the mechanical ventilator can be monitored with respect to the concentrations of the different respiratory gases, e.g., oxygen (O 2 ), nitrogen (N 2 ), nitrous oxide (N 2 O) and volatile anesthetics. When rebreathing of exhaled air is excluded in the ventilatory circuit, analysis of fresh gas streaming to the respirator should accurately reßect the composition of inspiratory gases. This attempt is advantageous, compared to sample lines attached to the inspiratory tubing-limb of the system, because it circumvents the need for gas extraction from the inspiratory tubing with the consequence of decreased inspiratory airways-pressures and ventilatory tidal volumes. The actual measurement can be performed online with commercially available multi-gas analyzers (Datex¨, HP¨). It has been argued that ventilation with room air may not be ideal in mice (24) , based on observed high murine P 50 values (PO 2 corresponding to 50 % hemoglobin saturation) of 65 mmHg (46) or even 71 mmHg (48) . However, most studies have found a P 50~ 40 mmHg for the mouse, when measured under standard conditions (pH = 7.4, PCO 2 = 40 mmHg, 37 ¡C) (25, 31, 35, 42, 47) . A P 50 of ~ 40 mmHg is still rightshifted, when compared to humans (P 50 = 28 mmHg). Nevertheless, although even with a high P 50~ 65 mmHg blood is virtually 100 % saturated at room air (46) , increased FiO 2 may be helpful in the mouse when uncertainty exists whether mechanical ventilation optimally expands the lungs.

If volatile anesthetics (halothane, enßurane, isoßurane or more recent ones such as sevoßurane and desßurane) are used, additional considerations have to be taken into account: clinically used vaporizers for volatile anesthetics (Draeger¨, series 19) are usually precalibrated for gas ßows higher than that required for mice and will therefore give unreliable results when used with low gas ßows. Although recalibration of the vaporizer for low air-ßow is the best option, a more common approach is to use an appropriate high gasßow (several hundred mlámin Ð1 ) with the abundant air released via a T-type adapter or valve. Disadvantages of the latter approach are higher costs and the generation of relatively large amounts of waste gas giving environmental contamination and increased toxicity for exposed personnel.

In contrast to maintenance of sufficient arterial oxygenation, which even under conditions of poor ventilatory support can be achieved by an increased fraction of inspiratory oxygen (FiO 2 ), adequate pulmonary elimination of CO 2 is the more challenging aspect of mechanical ventilation. Although the arterial blood gas analysis is regarded as the gold standard for evaluation of proper ventilation, the drawbacks are that it is a discontinuous and invasive method requiring a substantial vol-ume of arterial blood (0.1Ð0.2 ml for a single measurement, 5Ð10 % of the total blood volume of a 25 g mouse (58)), the latter aspect adding new sources of instability to the model. Though often neglected, measurement of respiratory carbon dioxide concentration (capnography) is therefore recommended in mouse research. Especially the determination of the end-tidal CO 2 (etCO 2 ), as a measure for alveolar CO 2 -concentration, is described as essential to maintain stable physiological parameters during mechanical mouse ventilation (7) . However, capnography has rarely been performed in mice due to technical difficulties, like gas dilution artifacts occurring in side stream systems, which prevent accurate detection of endtidal CO 2 -peaks or-plateaus. Therefore, to use the potential of capnography to its full extent, fast responding mainstreamcapnography is required. To this end we have developed an integrated mouse endotracheal tube with an implemented capnography-cell with minimal gas-mixing artifacts. Since this cell is located in the endotracheal tube it allows real-time in-and expiratory mainstream breath-by-breath capnography. A typical capnography curve using this device is shown in Fig.  1 , showing several single breath capnography curves (inner section with high chart speed, RR ~ 100 min Ð1 ).

Capnography is not only a measure for the adequacy of mechanical ventilation, but can also serve to detect changes in hemodynamics. An example for this is given in Fig. 2 : during steady-state anesthesia 0.2 ml blood was withdrawn from an arterial line (black arrow). The end-tidal carbon dioxide concentration drops immediately, likely due to a decrease in pulmonary perfusion (decreased cardiac output). Resuscitation is performed by rapid infusion of 0.2 ml albumin solution (white arrow), which increased capnography values back to baseline, probably by restoring pulmonary perfusion. Since time resolution is enhanced by our breath-by-breath capnography, the relation between expiratory CO 2 and speciÞc events can be described precisely. Additionally, the capnography curves can be analyzed with respect to irregularity (e.g., due to intermittent spontaneous breathing) and shape alterations (e.g., airway obstruction).

A few mouse capnographs are commercially available, mostly using side stream sampling (e.g., Kent ScientiÞc Corporation SC-210 respiratory CO 2 monitor). The only main stream capnograph commercially available for mice to our knowledge is the SC-300 (Kent-ScientiÞc Corporation) with an external CO 2 -probe to be placed in the ventilation tubing. However, we were unable to Þnd studies reporting on this device. A disadvantage of this device, as reported by the manufacturer, is that RR is limited to 150 min Ð1 , which might not meet the demands of all experimental designs (Table 3) .

The enhanced stability of the murine homeostasis is a major goal of mechanical ventilation, but typical side effects have to be considered, possibly adding new sources of instability to the murine model. These side effects can be classiÞed as regional (airways and lungs affected) or systemic (distant organs affected). A typical example for a ventilator-induced regional side effect is the mechanical hyperinßation of the murine lung, e.g., when large tidal volumes or high airway pressures are applied. This induces histological damage of the delicate murine pulmonary structures. This volutrauma has been shown in mice to trigger the onset of inßammation, including expression of cytokines (57) .

The systemic side effects of mechanical ventilation are (at least partly) due to a compromised hemodynamic situation. Mechanical ventilation per se results in larger intrapulmonary pressures as compared to spontaneous ventilation, possibly inhibiting venous return to the heart. In addition, prolonged phases of positive airway pressures, e.g., when the inspiratory fraction (Ti/Tt) is increased or when external PEEP is applied, depresses the circulation by several mechanisms, also via an impaired venous return to the heart. As an example for a global hemodynamic variable, arterial blood pressure decreases immediately after application of higher levels of PEEP: Fig. 3 shows a typical blood pressure trace of a deeply anesthetized mouse (C57BL/6, 25 g bodyweight, anesthesia with ketamine (35 mgákg Ð1 áh Ð1 ) and medetomidine (35 µgákg Ð1 áh Ð1 ) indicating the immediate hypotensive reaction after application of 10 cmH 2 O PEEP. The extent of hypotension induced by mechanical ventilation is further related to the respiratory phase, as demonstrated in Fig. 1 , a simultaneous recording of capnography and arterial blood pressure in a C57BL/6mouse (anesthetized with fentanyl (0.8 mgákgá Ð1 áh Ð1 ), ßuanison (25 mgákgá Ð1 áh Ð1 ) and midazlam (3 mgákgá Ð1 áh Ð1 ), a more hypotensive anesthesia than used in Fig. 3 , explaining the different baseline blood pressures). At end-expiration (white arrow), where intrathoracic pressure is minimal, arterial blood pressure reaches its peak value and vice versa (black arrow).

A second example for side effects of the mechanical ventilation with PEEP distant to the lung is the impairment of the intestinal microcirculation. In anesthetized C57BL/6-mice (terminal ileum exposed) we were able to visualize a marked reduction of perfused capillaries and a decrease in capillary blood flow velocity, using orthogonal polarized spectral (OPS-) imaging (16) . Figure 4 A/B shows a typical example for PEEP-induced microcirculatory impairment. Both images show the same intestinal region, before (4A) and 15 minutes after (4B) onset of 5 cmH 2 O external PEEP, with white arrowheads marking corresponding capillaries (black) in both images. In contrast to Fig. 4A the capillaries in 4B are markedly less numerous and appear dashed, indicating erythrocyte aggregations formed at low capillary blood flow velocities.

Respiratory control of mice is a field of growing interest. Although spontaneous ventilation is the more physiological mode of breathing, only mechanical ventilation maintains physiological values (blood gases, acid-base status) in situations with impaired spontaneous breathing, as occurring during anesthesia, surgery or constitutively in murine strains with impaired spontaneous ventilation. There are currently several commercially available small rodent ventilators feasible for use in mice. If mechanical ventilation is performed, then monitoring of this intervention is required, ideally by means of capnography. Research still has to be performed on how the most physiological ventilation can be achieved in mice, which ventilatory patterns are appropriate and how ventilation related side effects can be limited. 518 Basic Research in Cardiology, Vol. 95, No. 6 (2000) © Steinkopff Verlag 2000 

Brain derived neurotropic factor is required for normal development of the central respiratory rhythm in mice

In vivo electrophysiologic studies in mice

Effect of positive pressure breathing on the vibration tolerance of mice

RET-protooncogene is important for the development of respiratory CO2 sensitivity

Physiological characteristics of the chest and lungs and the work of breathing in different mammalian species

Breathing and pulmonary surfactant function in mice 24 h after ozone exposure

Cerebrovascular responses under controlled and monitored physiological conditions in the anesthetized mouse

The effect of restraint on ventilatory response to hypercapnia and hypoxia in adult mice

Contribution of type I NOS to expired gas NO and bronchial responsiveness in mice

Whole body plethysmography, does it measure tidal volumes of small animals?

Respiratory system mechanics in mice measured by end-inflation occlusion

Measurement of respiratory volume for virus retention studies in mice

Hemodynamic response to anesthesia in pregnant and nonpregnant ICR mice

In vivo left ventricular pressurevolume relations by miniaturized conductance micromanometry

Orthogonal polarization imaging: a new method for study of the microcirculation

A murine model of volutrauma: potential contribution of inßammatory cell proteases to lung injury

Demonstration of an early and a late phase of ischemic preconditioning in mice

Measurement of respiratory volumes of laboratory animals

Various immunological phenotypes are associated with increased airway responsiveness

Invasive hemodynamic studies in open and closed chest mice

(eds) Cardiovascular Physiology in the Genetically Engineered Mouse

Murine cardiac function Ð a cautionary tail

Enhanced oxygenation in vivo by allosteric inhibitors of hemoglobin saturation

Altered respiratory responses to hypoxia in mutant mice deÞcient in neuronal nitric oxide synthase

A miniature mechanical ventilator for newborn mice

Method of artiÞcial respiration in curarized mice

End-systolic pressure-dimension relationship of in situ mouse left ventricle

Impaired ventilatory responses to hypoxia and hypercapnia in mutant mice deÞcient in endothelin-1

Blood oxygen affinity and alveolar ventilation in relation to body weight in mammals

Whisker-related neuronal patterns fail to develop in the trigeminal brainstem nuclei of NMDAR1 knockout mice

Cardiovascular physiology in mice

The ventilatory and metabolic response to hypercapnia in newborn mammalian species

Stable rightward shifts of the oxyhemoglobin dissociation curve induced by encapsulation of inositol hexaphosphate in red blood cells using electrophoration

The development of a new mouse model of global ischemia

Lung tissue behavior in the mouse during constriction induced by methacholine and endothelin-1

Ventilation, CO2-production, and CO2-exposure effects in conscious, restrained CF-1 mice

Leptin prevents respiratory depression in obesity

Determination of ventilatory volume in mice by whole body plethysmography

Cardiovascular reßexes in mice

Regulation of oxygen affinity in blood of fetal, newborn and adult mouse

A computer-controlled research ventilator for small animals: design and evaluation

Reperfusion arrhythmias in the murine heart: their characteristics and alteration after ischemic preconditioning

Scaling: Why is Animal Size So Important

Oxygen dissociation curves of mammalian blood in relation to body size

Temporal change of plasma erythropoetin level in hypobaric hypoxic mice and the influence of an altered blood oxygen affinity

Near simultaneous hemoglobin saturation and oxygen tension maps in the mouse cortex during amphetamine stimulation

Mouse smooth muscle studies using gene-altered mouse models

Lung protein leakage in respiratory failure induced by a hybridoma making monoclonal antibody to the hydrophobic surfactant-associated polypeptide SP-B

Differential control of ventilation among inbred strains of mice

Leptin attenuates respiratory complications associated with the obese phenotype

Tissue gas tension in small wild mammals

Comparative quantitative morphology of the mammalian lung: diffusing area

Depressed heart rate variability and arterial baroreßex in conscious transgenic mice with overexpression of cardiac Gs-alpha

Dynamic mechanisms determine functional residual capacity in mice, Mus musculus

Hyperventilation induces release of cytokines from perfused mouse lung

Direct determination of plasma, cell, and organ blood volumes in normal and hypervolemic mice

Comparison of sensitization to scrude and purified house dust mite allergens