Breath timing, volume and drive to breathe in conscious rats: comparative aspects

Respiration Physiology 107 (1997) 241 – 250

Breath timing, volume and drive to breathe in conscious rats: comparative aspects Julia K.L. Walker, Barbara L. Lawson, Donald B. Jennings * Department of Physiology, Botterell Hall, 4th Floor, Queen’s Uni6ersity, Kingston, Ontario K7L 3N6, Canada Accepted 2 December 1996

Abstract In conscious animals, respiratory frequency (f) and tidal volume (VT) vary breath to breath. Examining the average value of variables associated with specific bins of another variable, such as breath f, provides a unique tool to examine respiratory behaviour. In conscious Sprague–Dawley rats respiratory breath timing, tidal volume (VT) and drive (VT/TI) were characterized using a plethysmograph. In the majority of rats at low breath f, expiratory time (TE) exceeded inspiratory time (TI) and these times became equal as f exceeded 150 breaths/min; there was no evidence for TI greater than TE at higher f, as observed in cats and dogs. When VT is normalized per kg, rat breath VT and VT/TI, binned by breath f, are continuous with those for the cat and non-panting dog at the lowest breath f. Relative to breath f, breath VT and VT/TI in rats are greater than in normothermic panting dogs (20°C), but only slightly greater than those variables in panting dogs in the heat (30°C). Lower values of breath VT/TI, binned by breath f or V: , in cats and dogs are compensated for by a greater TI relative to the duration of a given breath. This comparative analysis suggests continuities of respiratory pattern generation among species. © 1997 Elsevier Science B.V. Keywords: Respiratory pattern; Conscious rat; Respiratory control; Comparative respiration

1. Introduction We introduced an analytic paradigm to assess respiratory pattern based on the binning of the characteristics of individual breaths. In conscious cats (Jennings and Szlyk, 1985) and dogs (Iscoe et al., 1983), for specific physiological states, inspira* Corresponding author. Tel.: +1 613 5452797; fax: + 1 613 5456880.

tory time (TI) and expiratory time (TE) are highly predictable at a given instantaneous breath frequency (f). Breath f (breath/min) is defined as 60 sec/Ttot (sec) where Ttot=TI + TE. In these species, at low breath f, TE exceeds TI; however with increasing breath f, TI and TE values approach each other and in cats and dogs, eventually ‘crossover’ so that TI exceeds TE (Iscoe et al., 1983; Jennings and Szlyk, 1985). Although the central mechanism responsible for such ‘behavior’

0034-5687/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 3 4 - 5 6 8 7 ( 9 6 ) 0 2 5 2 0 - 0

242

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

is not clear, we speculated that it may reflect hypothalamic input to respiratory centers (Jennings and Szlyk, 1985). In resting dogs, the crossover for breath timing occurs between non-panting and panting breath f (Iscoe et al., 1983), and in resting cats the crossover for breath timing occurs between nonpurring and purring (Jennings and Szlyk, 1985). Chemical stimuli and sensory receptor interventions affect both the timing at a given breath f and the crossover breath f. For example, acute hypercapnia decreases the crossover f in conscious dogs (Iscoe et al., 1983) and cats (Szlyk and Jennings, 1987a); sympathectomy of the carotid bifurcation altered respiratory timing relative to breath f (Szlyk and Jennings, 1987b). In both species, for a given physiological state, variability in breath tidal volume (VT) was also highly predictable relative to breath f. Thus, the timing and volume for varying sequential breaths are ‘packaged’, perhaps to optimize work of breathing or gas exchange. We wished to establish a rat model to pursue studies of central mechanisms for angiotensin II stimulation of ventilation, as observed in dogs (Jennings et al., 1995); stereotaxic lesions can be more precise in rats. The purpose of the present study was to compare respiration in conscious rats under control conditions with that of conscious dogs and cats, as measured in our laboratory. In the present study respiratory pattern, timing and drive in conscious rats was characterized in relation to individual breaths and their variability. Although respiratory pattern has been compared between species, most studies have been constrained to ‘normal’ measurements to provide a single value for respiratory timing and volume. Indeed, allometric analysis assumes each species provides only one average number. We established TI and TE with respect to breath f in conscious rats and compared their timing relations with those of cats and dogs. We also compared VT and VT/TI (normalized for body weight) relative to breath f to those previously reported in the cat and dog. The analyses provide an overview of the differences and continuities of respiratory patterns among these species.

2. Methods Male Sprague–Dawley rats were chronically prepared with a catheter in the abdominal aorta under ketamine anesthesia (Rogarsetic, 70 mg/kg) in combination with xylazine (Rompun, 5 mg/kg). Catheters were drawn subcutaneously and exteriorized dorsally between the scapulae and filled with a 50% heparin-saline solution (Hepalean, Organon Teknika). During surgery, a calibrated VM-FH transmitter (Mini-Mitter, Sunriver, OR) was also placed in the abdominal cavity for measurement of abdominal temperature (Tab). A recovery period of at least 48 h followed surgery before rats were used in experiments. Animals were acclimated to an ambient temperature of 21–22°C and a photoperiod between 06:00–20:00 h; they were fed standard rat chow and allowed free access to water. All experimental procedures conformed to guidelines of the Canadian Council on Animal Care and were approved by the Queen’s University Animal Care Committee. During experiments the rats were studied in a plethysmograph chamber (1.6 l) for respiratory and metabolic measurements. Data from 12 rats were selected whose metabolic measurements were obtained under conditions where the chamber flow rate was 3.3 l/min or less (average flow = 2.090.3 l/min). The fraction of inspired CO2 (FICO2) in the chamber during air flow was less than 0.005. In initial experiments we found that higher flow rates in the chamber resulted in markedly higher metabolic rates. The flowmeter to the chamber was calibrated for each experiment using a dry gas meter (Harvard Apparatus). Chamber temperature (Tch) and humidity were measured using a Vaisala HMP 233 Transmitter (Helsinki, Finland). Temperature measurements made prior to respiratory measurements were used in calculation of metabolic variables while those made at the end of the respiratory period were used in calculation of respiratory variables. Two additional rats were studied, first while conscious, and then while anesthetized (intraperitoneal injection of a-choralose (65 mg/kg) and urethane (800 mg/kg)) and with a tracheal cannula.

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

An extension catheter was attached to the exteriorized aortic catheter of the rat during experiments and connected to a transducer/recorder system (BPA100, Micro-Med, Louisville, KY) outside the chamber to record minute-to-minute mean arterial pressure (MAP) and heart rate (fH). Anaerobic arterial blood samples were obtained in 150 mL heparinized glass capillary tubes for determination of PaO2, PaCO2 and pHa using a Radiometer BMS3 Mk2 blood gas analyzer. Blood gas measurements were corrected to Tab at the time of sampling. For respiratory measurements, flow through the chamber was terminated and the chamber was sealed (Drorbaugh and Fenn, 1955; Jennings and Szlyk, 1985). Breath-to-breath respiratory pressure changes were monitored for 20 – 30 sec using a Validyne differential pressure transducer (model no. DP7) and recorded on a Grass model 7 recorder. The pressure on the reference side of the pressure transducer was stabilized by a connection to a chamber of identical dimensions which had a slow leak to atmosphere. During each measurement period, the pressure-volume relation was established by rapidly injecting 0.1 ml of gas into the chamber during the expiratory phase of the rat and averaging the deflections obtained over the experiment (variability  92.5%). Volume calibration of the plethysmograph was flat to a frequency in excess of 8 Hz. This rapid volume calibration injection (  0.1 sec) was corrected for an overestimation of a true calibration volume by an adiabatic factor and an impact factor related to emptying of the syringe. The correction was established with a ‘dummy rat’, with and without copper sponges in the chamber, by injecting the same gas volume rapidly as well as slowly over 0.2–5 sec; slow injections cannot be done with a rat breathing in the chamber. The adiabatic factor was calculated to be 1.38, comparable to the prediction of 1.4 by Bargeton and Barre`s (1968), and the total multiplication factor to obtain a ‘true’ VT was 1.61. Breaths were analyzed with a graphics tablet (Scriptel) and computer for calculation of VT (ml) (Jennings and Szlyk, 1985). Sequential breaths at each measurement period were analyzed for their Ttot, TI, TE and VT; breath f (60 sec/Ttot), breath ventilation (V: = f×

243

VT) and breath VT/TI were calculated. Average values for sequential breaths over a given measurement period were also calculated. Volume measurements were normalized for body weight in kilograms. Inflow and outflow gases were analyzed immediately before a respiratory period using CO2 and O2 analyzers (Beckman LB2 and OM-11, respectively) which were calibrated with gases of known composition determined by the Scholander technique. Repeat calibrations of the Beckman analyzers were required to be within 0.05% for CO2 and 0.1% for O2. Oxygen consumption (V: O2) and CO2 production (V: CO2) were calculated (STPD) from the differences between inflow and outflow O2 and CO2 concentrations and the known chamber air flow (Sachdeva and Jennings, 1994); R was calculated as the ratio of V: CO2/V: O2.

2.1. Experimental design Experiments were carried out from 10:00–16:00 h. Prior to experiments, rats became accustomed to the chamber for approximately 20 min. An arbitrary zero time was established and measurements were obtained at 10 min intervals over 30–60 min. The individual breath data from rats were compared with data obtained from other species. For dogs, we used previously reported breath data (Iscoe et al., 1983) from five conscious dogs breathing through an endotracheal tube for respiratory measurements under both cool (20°C) and warm ( 30°C) ambient temperatures. Twoway valves attached to the endotracheal tubes of dogs in those studies had inspiratory and expiratory resistances of about 1–2 cm H2O/L per sec. For cats, breath data were compiled from 14 cats studied by the plethysmographic technique in our laboratory; the methodology and analyses were identical to those previously reported for six of the cats (Jennings and Szlyk, 1985).

2.2. Data analyses Statistical analyses were carried out using the computer software package SYSTAT for Windows. Data for each rat were binned by breath f

244

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

and averaged. In the figures we excluded bins where the number of rats contributing average values was two or less. In Table 1, the data represent the average values for rats obtained from the mean values of averaged sequential breaths from each rat at each measurement period. Values presented are means 9 SE. To determine statistical differences in Fig. 4, a paired t-test was used.

3. Results

3.1. Control measurements The chamber temperature during the studies and the physiological characteristics of the rats are outlined in Table 1. The sequential breath-bybreath variability for breath f, TI, TE, VT, VT/TI and V: over 44 sequential breaths (20.6 sec) in one typical resting rat are depicted in Fig. 1. A periodic variability in breath f was most notably reflected by changes in breath V: . Histograms of Table 1 Control measurements in 12 conscious rats Variable

mean9 SE

Weight (g) Tch (°C) Tab (°C) MAP (mmHg) fH (beats/min) PaO2 (torr) PaCO2 (torr) [H+]a (nEq/L) V: (ml/kg per min) VT (ml/kg) fR (breaths/min) V: O2 (ml/kg per min) V: CO2 (ml/kg per min) R

3089 5 23.29 0.2 37.59 0.1 939 2 34299 9592 389 1 379 1 6809 32 7.29 0.3 959 4 25.391.2 20.59 0.8 0.819 0.02

SE, standard error of the mean; Tch, plethysmograph temperature; Tab, abdominal temperature; MAP, mean arterial pressure; fH, heart rate; PaO2, arterial pressure of oxygen; PaCO2 , arterial pressure of carbon dioxide; [H+]a, arterial hydrogen ion concentration; V: , average minute ventilation; VT, average tidal volume; fR, average respiratory frequency; V: O2, oxygen consumption; V: CO2, carbon dioxide production; and R, respiratory exchange ratio.

individual breath f, VT and collected from the 12 rats are depicted in Fig. 2. Distributions are unimodal; the median breath f, in particular, was less than the average values so that breath was skewed to lower values of breath. At low breath f, expiratory flow was associated with the characteristic features associated with expiratory ‘braking’ (Boggs, 1992). The expiratory phase in conscious rats did not exhibit an expiratory pause at any breath f as seen from raw tracings of two typical rats in Fig. 3. As evident from the right part of Fig. 3, when these rats were anesthetized and their tracheas cannulated, there was a more rapid return from peak inspiratory volume to resting end-expiratory volume.

3.2. Breath timing in rats As depicted in Fig. 4, for this group of 12 rats, at the lowest breath f, TE was greater than TI, but TI and TE gradually converged to become ‘equal’ at the highest breath f’s beyond 150–175 breaths/ min. The plots represent the averages of the mean data from each rat for each f bin. The number of rats contributing to each mean value for a bin is given above the x-axis. No crossover of TI and TE, with respect to f, was evident in this most typical pattern for conscious Sprague–Dawley rats; TI did not become greater than TE. It is to be noted that we observed occasional random breaths in the rats to have TI/Ttot exceeding 0.5; but, as evident from Fig. 4 such breaths did not dominate the average timing at any bin of breath f.

3.3. Breath timing in rats compared to cats and dogs Fig. 5A, depicts averages of TE plotted against averages of TI (binned by breath f) from the rats in relation to similar average binned data obtained from resting cats and dogs. In this plot, isopleths for TI/Ttot as well as f are superimposed on the figure. For similar breath f ranges, the ratio of TI/Ttot relative to breath f for rats is less than those for the other species (Fig. 6B inset). As evident from the isolines of TI/Ttot in Fig. 5B, in cats and dogs TI exceeds that in rats for similar

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

245

Fig. 1. Typical data from one rat of breath-to-breath variability of respiratory variables over 20.6 sec.

breath f over the range of f for rats (TI/Ttot equal or greater than 0.5 in cats and dogs and less than 0.5 in rats).

3.4. Breath VT and VT /TI in rats compared to cats and dogs Fig. 6 depicts plots of breath VT/TI (Fig. 6A and B) and breath VT (Fig. 6C and D), normalized by body weight, for rats, cats and dogs versus bins of breath f. For breath VT/TI, isopleths of VT (assuming TI/Ttot = 0.5) are superimposed on panels A and B; for VT, isopleths of V: are superimposed on panels C and D. For Fig. 6A and C, the dotted areas of the figure, where small changes in TI and TE result in very large changes in breath f, are expanded in panels B and D, respectively, for clarity. As shown in Fig. 6A, rat data for average breath VT/TI, relative to breath f for rats extend beyond average breath VT/TI at the lowest breath f for cats and non-panting dogs (under cool ambient conditions). Where the data for rats overlap with bins of breath f of data from the other two species, average breath VT/TI in rats is higher. For the same range of breath f as rats, panting

dogs under cool ambient temperature (nonthermal panting) have much lower breath VT/TI due to lower relative VTs. Interestingly, thermally-induced panting in dogs under warm ambient temperature conditions ( 30°C) is associated with average breath VT/TI values that are intermediate to those in rats for the same range of breath f in both species. In the heat, the VTs of panting dogs are greater than at the lower ambient temperature. In rats, average breath VT remains relatively constant as breath increases with increasing breath f (Fig. 6C). Breath V: in rats is at the upper range of resting V: among these species and compares with ‘panting’ V: in dogs. As with VT/TI, breath VT in rats, relative to bins of breath f, is in continuity with and extends beyond those of cats and non-panting dogs (the latter under cool ambient temperatures). However, as also indicated in Fig. 6C, VT in panting dogs under cool ambient temperatures is relatively less than in rats on a per kilogram basis for the range of breath f observed in rats. It is also apparent from Fig. 6C that VTs (on a weight basis) in rats are only slightly greater, relative to breath f, than those in panting dogs under warm ambient temperatures.

246

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

Fig. 2. Histograms of individual breath data for breath f, breath VT and breath V: from 12 rats.

3.5. Breath VT /TI, relati6e to bins of breath V: , in rats compared to cats and dogs

4. Discussion

Despite the differences in respiratory patterns among species for attaining various individual breath as shown in Fig. 6C, breath VT/TI increases in a linear fashion with breath V: in all of these species (Fig. 7). Moreover, breath VT/TI is higher, relative to V: , in rats than in dogs or cats.

This paper provides the first description of the breath-to-breath pattern of respiration in the conscious Sprague—Dawley rat and puts respiration in this strain of rats in context with other species commonly used for respiratory studies. Although the oscillatory behaviour of respiratory activity has been recognized for many years (for example,

Fig. 3. Typical examples of pressure tracings from the plethysmograph for two Sprague–Dawley rats, both in the conscious state and subsequently with a tracheal cannula following chloralose – urethane anesthesia.

Fig. 4. Average data for TE and TI for bins of breath f from 12 rats. A rat contributed only one average value to each bin. The number of rats contributing data to a bin is shown above the x-axis. * p B0.05 between TI and TE for that bin.

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

Fig. 5. A nomogram (A) in which TE is plotted against TI (both binned by breath f). Data from rats shown in Fig. 4 (“) are plotted against similar data acquired in this lab for resting conscious cats ( ; see Section 2) and resting conscious dogs (; Iscoe et al., 1983). Solid lines superimposed on the diagram represent isolines of TI/Ttot. Dotted lines are isolines of breath f. The data at higher f are expanded for clarity on nomogram (B).

Goodman, 1964; Mead, 1960), most models of central rhythm generation do not incorporate breath variability. In previous papers we discussed in detail differences in interpretation which occur when sequential breaths of different characteristics are averaged, and when a new spectrum of variability, induced by a stimulus, is not taken into account (Iscoe et al., 1983; Jennings and Szlyk, 1985; Szlyk and Jennings, 1987a,b). For other studies in the literature of breath timing in rats (e.g. Boggs and Tenney, 1984) the values of sequential breaths of different breath f have generally been averaged to arrive at conclusions about breath timing. As evident from Fig. 4, conscious resting rats have predictable average timing at a given breath Ttot or f. If, following an experimental intervention, timing of breaths of similar f and Ttot are not compared with control breaths of similar f and Ttot, then the inherent effects of changing f on breath timing (as in Fig. 4) can be obscured. It has been proposed that the Drorbaugh and Fenn (1955) equation underestimates VT as mea-

247

sured in the rat and cat studies (Jacky, 1980). Thus in this comparative analysis, VT for the cat and rat, as measured by plethysmography, may be somewhat low relative to the dog data, where ventilatory parameters were measured by spirometry (Iscoe et al., 1983). However, such potential errors would not appear important to the general conclusions obtained from examining Figs. 6 and 7. In order to compare respiratory variables in different species allometry is employed which allows a physiological variable (Y) to be compared in different size species (Bennett and Tenney, 1982; Boggs, 1992). Body weight (BW) is easily measured and commonly used to examine allometric relations. The general equation Y=aBWb is logarithmically transformed to log Y= log a+ b · log BW, where a is the intercept and b is the slope of the regression (reviewed in Boggs, 1992). In the present comparative analyses in Figs. 6 and 7, where overlapping ranges of respiratory variables are present among species, we used VT per kilogram for the three species. This would seem acceptable for volume measurements since VT has been shown to be an interspecific constant, i.e. VT is related to BW1.0 (Boggs, 1992). The range of breath f of rats corresponds to the range of panting breath f in dogs (Iscoe et al., 1983; Jennings, 1984). However, rats are different from both cats and dogs in which TI exceeds TE at this range of breath f (Fig. 5). It seems unlikely that breath timing over this range of f in panting dogs would be markedly altered by an endotracheal tube for this comparison; at panting f in dogs laryngeal resistance, with potential effects on timing, should normally be minimized. We did not observe crossover of breath timing (TI/Ttot\ 0.5) in relation to breath f in resting rats; a random distribution of a small number of breaths with TI/Ttot greater than 0.5 may have been related to unapparent postural changes. Cats (Jennings and Szlyk, 1985) and dogs (Jennings, 1984; Iscoe et al., 1983) have bimodal distributions of occurrences of breath f associated with the crossover of timing. In resting cats crossover appeared to be associated with purring (Jennings and Szlyk, 1985) and in resting dogs crossover is associated with changing from non-panting res-

248

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

Fig. 6. Average breath VT/TI (A) and breath VT (C) binned by breath f. The boxed areas of A and C are expanded in B and D. The data are for the same rats as shown in Fig. 4. In A and B, dotted lines represent approximate isolines of VT, assuming TI/Ttot=0.5. In C and D, dotted lines represent isolines of breath V: . Data for cats and dogs, and symbols for species, are as in Fig. 5. Breath data obtained from panting dogs exposed to a warm ambient temperature (30°C) (Iscoe et al., 1983) are presented ( ).

piratory frequencies to panting frequencies (Jennings, 1984; Iscoe et al., 1983). We speculated that a hypothalamic mechanism might be involved in crossover in the other species (Jennings and Szlyk, 1985). The only reported data for respiratory timing, binned by breath f, in human subjects were obtained during exercise and in those subjects, like rats, there was no crossover (Caretti et al., 1992); TI became equal to TE at breath f between 39 and 55 breaths/min. The time constant (t) for expiration, as derived from anesthetized or ‘dead’ animals, indicates that t is considerably smaller in rats, than larger animals. Interspecific analyses show t 8BW0.03 (Bennett and Tenney, 1982). For example, measurements of t in rats (excluding the upper airway) gave values from 0.035 sec (Crosfill and Widdicombe, 1961) to 0.0897 sec (Bennett and

Tenney, 1982). These numbers approach the expiratory times of the highest breath f in our rats. In our studies in rats (Fig. 3), at the lowest breath fs the expiratory tracings appear to be associated with expiratory braking, similar to that observed in newborn infants (Fisher et al., 1982). Indeed, when a tracheal cannula was introduced into anesthetized rats the delay in returning to end-expiratory volume was reduced (Fig. 3) so that at least part of the apparent expiratory braking was probably laryngeal in origin. Expiratory activity of both laryngeal adductor muscles (Megirian and Sherrey, 1980) and the diaphragm (Sherrey et al., 1988) have been demonstrated in the unanesthetized rat, similar to the cat (Gautier et al., 1973). In a review, Boggs (1992) points out the commonness and importance of expiratory ‘braking’ among species particularly in the presence of a highly compliant chest wall.

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

It is also apparent that breath timing among species is considerably different for the same f’s, and that a greater amount of time is spent in inspiration than in expiration in conscious dogs and cats compared to conscious rats for similar breath f (Fig. 5 (B)). The latter finding is qualitatively consistent with the calculation, based on allometry, that TI/Ttot 8 BW0.35 (Boggs and Tenney, 1984); however, there is inconsistency for this prediction in the same figure for data between dogs and cats. With the considerable overlap of potential breath f between dogs and cats and between dogs and rats (Fig. 5), it becomes difficult to think in terms of allometric relations for f. As depicted in Fig. 6 resting conscious rats have a range of V: s that is at the high range of ventilation among the three species. The range of V: in rats is equivalent to that observed in panting dogs and the pattern of breathing (breath f and breath VT) is very similar to that of heat stressed panting dogs. As previously described (Iscoe et al., 1983; Jennings et al., 1973; Jennings, 1984), some dogs, while in a steady-state under cool ambient temperatures, have panting behaviour (non-thermal panting). In the heat, respiratory pattern and

Fig. 7. Average breath VT/TI 9 S.D. versus breath V: (binned by V: ) for the same rats as in Fig. 4. The number of rats contributing to data points are indicated in brackets above the data. Symbols and data for other species are as in Figs. 5 and 6.

249

breath timing are different in dogs than during non-thermal panting; VT is greater relative to f in panting dogs in the heat (Fig. 7; Jennings, 1984). A relatively linear relation between breath VT/ TI and breath V: has been observed in dogs (Iscoe et al., 1983), cats (Jennings and Szlyk, 1985) and human subjects also (Caretti et al., 1992). As evident from Fig. 7, this is also true for the rat at lower breath V: , but breath VT/TI is greater in the rat than in the dog and cat which are similar to each other (note from the previous discussion that VT of the cat as well as the rat may be underestimated by the Drorbaugh and Fenn (1955) equation). The muscle fibers of the diaphragm have been shown to have histological characteristics intermediate between ‘fast’ and ‘slow’ muscles, and diaphragmatic fibers of smaller species have a higher proportion of oxidative enzymes and fasttwitch fibers (Blank et al., 1988; Davies and Gunn, 1972; reviewed by Boggs, 1992). Oxidative enzyme activity and fast myosin-related phenotypes are also associated with a species-related respiratory frequency (Blank et al., 1988). Sant’Ambrogio and Saibene (1970) found that respiratory muscle contraction time occurred much faster in rats than in the dog and the cat which is consistent with the greater breath VT/TI in rats than the other two species in Fig. 7. It appears that for a given time available for a breath (Ttot), the slower rate of respiratory muscle contraction in dogs and (cats Sant’Ambrogio and Saibene, 1970) is compensated for by a central adaptation of breath timing generation. In cats and dogs, relative to rats, TI is prolonged at a given breath f ( \TI/Ttot; Fig. 5) which will allow for a greater VT and compensates for the lower breath VT/TI. The finding that the rat has a shorter passive expiratory t for the respiratory system than the other species (Bennett and Tenney, 1982) is also consistent with VT being achieved in the rat with a shorter TI. In conclusion, these analyses demonstrate similarities and continuities in respiratory patterns among resting conscious rats, cats and dogs. Differences of breath characteristics among species are associated with reported contractile differences of respiratory muscle. High frequencies of

J.K.L. Walker et al. / Respiration Physiology 107 (1997) 241–250

250

breathing in rats and dogs are similarly associated with low passive expiratory time constants. Such information could be important in assessing the respiratory response to stimuli in rats and for the understanding of mechanisms for respiratory control in the rat, in comparison with the other species.

Acknowledgements The excellent technical assistance of Miss Heather Lockett was appreciated. The suggestions and comments of Dr John Fisher in assessing the physiological implications of our data were much appreciated. J.K.L. Walker was supported by an Ontario Graduate Award as well as an Ontario Thoracic Society Award. The research was funded by the Medical Research Council of Canada and the Ontario Thoracic Society.

References Bargeton, D. and G. Barre`s (1968). Time characteristics and frequency response of body plethysmograph. In: Progress in Respiratory Research, Vol. 4, Int. Symp. Body Plethysmography, Nijmegan. Basel: Karger, pp. 2–23. Bennett, F.M. and S.M. Tenney (1982). Comparative mechanics of mammalian respiratory system. Respir. Physiol. 49: 131 – 140. Blank, S., V. Chen and C.D. Ianuzzo (1988). Biochemical characteristics of mammalian diaphragms. Respir. Physiol. 74: 115 – 125. Boggs, D.F. and S.M. Tenney (1984). Scaling respiratory pattern and respiratory ‘drive’. Respir. Physiol. 58: 245 – 251. Boggs, D.F. (1992). Comparative Control of Respiration. In: Treatise on Pulmonary Toxicology, Vol. 1: Comparative Biology of the Normal Lung, edited by R.A. Parent. Boca Raton, FL: CRC Press, pp. 309–350. Caretti, D.M., P.C. Szlyk and I.V. Sils (1992). Effects of exercise modality on patterns of ventilation and respiratory timing. Respir. Physiol. 90: 201–211. Crosfill, M.L. and J.G. Widdicombe (1961). Physical characteristics of the chest and lungs and the work of breathing in different mammalian species. J. Physiol. (London) 158: 1 – 14. Davies, A.S. and H.M. Gunn (1972). Histochemical fibre types in the mammalian diaphragm. J. Anat. 112: 41–60.

.

Drorbaugh, J.E. and W.O. Fenn (1955). A barometric method for measuring ventilation in newborn infants. Pediatrics 16: 81 – 87. Fisher, J.T., J.P. Mortola, J.B. Smith, G.S. Fox and S. Weeks (1982). Respiration in newborns. Am. Rev. Respir. Dis. 125: 650 – 657. Gautier, H., J.E. Remmers and D. Bartlett,Jr. (1973). Control of the duration of expiration. Respir. Physiol. 18: 205 – 221. Goodman, L. (1964). Oscillatory behaviour of ventilation in resting man. Trans. Biomed. Eng. 11: 82 – 93. Iscoe, S., R.B. Young and D.B. Jennings (1983). Control of respiratory pattern in conscious dog: effects of heat and CO2. J. Appl. Physiol. 54: 623 – 631. Jacky, J.P. (1980). Barometric measurement of tidal volume: effects of pattern and nasal temperature. J. Appl. Physiol. 49: 319 – 325. Jennings, D.B., C.C. Chen, H.H. Phillips and J. Sparling (1973). Respiration and metabolism in panting and nonpanting resting conscious dogs. J. Appl. Physiol. 35: 490 – 496. Jennings, D.B. (1984). Breathing patterns and drives during thermal and non-thermal panting. In: Thermal Physiology, edited by J.R.S. Hales. New York: Raven Press, pp. 335 – 340. Jennings, D.B. and P.C. Szlyk (1985). Ventilation and respiratory pattern and timing in resting awake cats. Can. J. Physiol. Pharmacol. 63: 148 – 154. Jennings, D.B., J.K.L. Walker and P.J. Ohtake (1995). Central effects and interactions of the renin-angiotensin system and vasopressin on ventilation and PaCO2. In: Lung Biology in Health and Disease, Vol. 82: Ventral Brainstem Mechanisms and Control of Respiration and blood Pressure, edited by C.O. Trouth, R.A. Millis, H.F. Kiwull – Scho¨ner and M.E. Schla¨fke. New York: Marcel Dekker, pp. 203 – 224. Mead, J. (1960). Control of respiratory frequency. J. Appl. Physiol. 15: 325 – 336. Megirian, D. and J.H. Sherrey (1980). Respiratory functions of the laryngeal muscles during sleep. Sleep 3: 289 – 298. Sachdeva, U. and D.B. Jennings (1994). Effects of hypercapnia on ventilation, metabolism and temperature during heat and fever. J. Appl. Physiol. 76: 1285 – 1292. Sant’Ambrogio, G. and F. Saibene (1970). Contractile properties of the diaphragm in some mammals. Respir. Physiol. 10: 349 – 357. Sherrey, J.H., M.J. Pollard and D. Megirian (1988). Proprioceptive, chemoreceptive and sleep state modulation of expiratory muscle activity in the rat. Exp. Neurol. 101: 50 – 62. Szlyk, P.C. and D.B. Jennings (1987a). Effects of hypercapnia on variability of normal respiratory behaviour in awake cats. Am. J. Physiol. 252: R538 – R547. Szlyk, P.C. and D.B. Jennings (1987b). Respiration in awake cats: sympathectomy and deafferentation of carotid bifurcations. J. Appl. Physiol. 62: 932 – 940.