The influence of venous CO2 on ventilation in garter snakes

47

Respiration Physiology. 1t3 (1991) 47-60 Elsevier RESP 01730

The influence of

CO 2 on ventilation in garter snakes venous

R.A. Furilla*, E.L. Coates and D. Bartlett Jr. Department of Physiology. Dartmouth Medical School. Hanover. New Hampshire. U.S.A. (Accepted 26 August 1990) Abstract. Garter snakes were used to study the effects ofvenous CO2 loading using the skin as an exchanger. The gaseous environment surrounding the snake's body was isolated by placing the body in a plethysmograph with the head out. While the animal breathed room air, the carbon dioxide concentration within the plethysmograph was varied between 0 and 80~o. Room air was drawn through a funnel placed over the snake's head, thus collecting the exhaled gases, and this gas was analyzed by 02 and CO2 analyzers. The descending aorta was cannulated to measure blood gases. Expired CO2 flow rose linearly with increasing cutaneous CO2. Ventilation increased 3.5-fold at 80% cutaneous CO 2 compared with no cutaneous CO2 load. Neither the mean CO: concentration in exhaled air nor arterial Pco2 changed when the snake was exposed to high levels of CO., at the skin. Thus ventilation increased in proportion to the CO2 load, and was not driven by arterial hypercapnia. Bilateral vagotomy eliminated arterial CO2 homeostasis during cutaneous CO, loading, and ventilation increased with increasing arterial Pco2. Therefore, these snakes respond to extra-arterial elevations in CO2 or to a changing CO2 signal. Furthermore, receptors responsible for the increase in ventilation when venous CO:, is elevated have neurons in the vagus nerves.

Animal, snake; Control of breathing, r0sponse to CO: in snake; CO2, stimulation of respiration; Cutaneous gas exchange in snake; Vagotomy, and response to CO: in snake

in 1960, Yamamoto and Edwards reported that venous loading of COa by an extracorporeal gas exchanger in anesthetized rats caused ventilation to increase in proportion to the CO2 infusion rate, with no change in arterial Pco.,. Since then, the influence of venous CO2 loading has been studied in a variety of mammalian species (Greco et al., 1978; Phillipson et al., 1981; Stremel et al., 1978; Wasserman et al., 1975). Ventilation has been stimulated in each of these experiments, but the results have varied as to whether arterial Pco2 increased, and if so, whether the arterial hypercapnia was sufficient to account for the hyperpnea. Birds and reptiles have intrapulmonary chemoreceptors (IPCs), which decrease their discharge in response to CO2 (Fedde and Peterson, 1970; Fedde et al., 1977; Furilla Correspondence to: R.A. Furilla, Dept. of Biology, New Mexico State University, Las Cruces, NM 88003, U.S.A. * Present Address: Dept. of Biology, New Mexico State University, Las Cruces, NM 88003, U.S.A. 0034-5687/91/$03.50 © 1991 Elsevier Science Publishers B.V. (Biomedical Division)

48

R.A. FURILLA et al.

and Bartlett, 1988). These receptors are capable of influencing ventilation (Osborne et al., 1977), but their normal function is still not understood. Their anatomical location would allow them to monitor systemic venous or pulmonary arterial Pco2. Boon et ai. (1980) found that arterial Pco2 was unchanged in the chicken during venous COz unloading, suggesting an extra-arterial sensory mechanism, but when venous COe was elevated, arterial Pco2 also rose. Similarly, Jones et al. (1985) and Tallman and Grodins (1982) found arterial hypercapnia during venous CO2 loading in ducks. The animals in these studies, however, were either decerebrate, anesthetized or gut ventilated with pure COz. The methods therefore may have obscured the natural response to elevated venous C O 2.

Garter snakes have IPCs similar to those of birds (Furilla and Bartlett, 1988). In addition, they have a cutaneous surface area that is large relative to their body mass. To study the effect of elevated venous and intrapulmonary Pco~ on ventilation and arterial P¢o: in these animals, we have used the skin as a gas exchange organ, thereby minimizing surgical intervention, and eliminating the effects of anesthesia.

Methods

VENTILATION AND GAS EXCHANGE

Eight snakes (99 + 17 g SD) were used to monitor the ventilatory response to CO2 uptake across the skin. These animals underwent no surgical preparation. Each snake was secured to a piece of Plexiglas 1 x 2 x '75 cm with adhesive tape. One strip of tape was placed behind the head, another was placed midway betv, een the head and the tip of the tail, and a loose strip of tape was placed midway between these two anchor points to inhibit lateral movement, without interfering with ventilation. Several other strips of tape secured the posterior end of the snake. The animal was then placed into a Plexiglas cylinder which was used as a plethysmograph (fig. 1). The snake's head was pulled through a small hole in a sheet of medium weight dental dam (Higenic Corp., Akron,

.

From

ii 11" ~ ~ Calibration I I I REC l - - i P T r - l ~ l

~.Dead Space URemoval II

L

To "* 0 = ~ CO~

0 z ' C 0 2 8 Nz ~ Mixer

Analyzers

t Fig. I. Schematic representation of the experimental setup. Air and C02 flowed through the plethysmograph and pneumotachograph. Breathing was measured as the differencefrom this offset flow. Expired gas was collected and analyzed by oxygen and carbon dioxide analyzers. Flow was calibrated by injecting and withdrawing known volumes in the plethysmograph.The hatched area represents a water jacket which was not used in this study.

49

VENOUS CO2 AND VENTILATION IN SNAKES

OH). This rubber membrane was stretched over the end of the plethysmograph, and held in place by a plastic ring secured with wing nuts. This arrangement permitted the snake to breathe room air while most of the body surface was exposed to other gas mixtures in the plethysmograph. Oxygen, nitrogen and carbon dioxide were mixed using calibrated flow meters (Matheson, Secaucacus, NJ) to provide the gases used for cutaneous exposures. The CO2 concentrations used were 0, 10, 20, 40 and 80%. Oxygen concentration was maintained at 20 ~o, and the balance, if any, was made up of nitrogen, The order of exposure to the gas mixtures containing CO2 was varied. The sequence always began and ended with CO2-free air, however, and no more than two hypercapnic mixtures were used without returning to CO2-free air. Each of the above gas mixtures in turn was directed through the plethysm0graph at 100 ml. rain- ~, exiting through a pneumotachograph. This flow was used to r~mintain the gaseous environment within the plethysmograph, and the ventilatoD, flow signal was superimposed on this bias flow. Flow was calibrated with the animal in place by injecting air into and withdrawing air from the plethysmograph while the snake was in a ventilatory pause. The gas w,ixture in the plethysmograph was replaced by removing the rubbel" stopper containing the pneumotachograph, and increasing gas flow to 2 L. min - ~for 5 min. The volume of the plethysmograph was approximately 700 ml, and nearly complete turnover ofthe gas was accomplished during this period. Flow was then reduced to 100 ml. min- ~, and the pneumotachograph was reinstalled. The snake was kept in each new gaseous environment for 30 min to approach a steady state with respect to the respiratory exchange ratio measured at the mouth (fig. 2); ventilation also reached steady state within 30 min. Data were then collected for another 30 min. Struggling occasionally occurred during the recording period, and the results of such trials were discarded. Therefore, paired analyses were not possible. A small funnel was placed over the snake's head, and air was drawn through the funnel at 450 ml. min- ' to collect the animal's expired gases along with an excess of room air. This air was analyzed by a Beckman LB2 CO2 analyzer and a Beckman _

1.5

e:

o o

~

~

o..o~o o

-o

I o

0.5

m

0

0

I

I

I

I

I

500

I000

1500

2000

2500

Time (sec) Fig. 2. Scatter plot of the respiratory exchange ratio (R) as a function o£ time. Zero on the abscissa represents the time at which the skin was exposed to 80% CO2.

50

R.A. FURILLA et al.

OMI1 02 analyzer (Beckman Instruments, FuUerton, CA). The output of these analyzers and that of the pneumotachograph were stored on computer disks using an IBM PC and an analog to digital converter (Data Translation, DT2801A, Marlboro, MA) operating at 20 Hz per channel. The data were analyzed on a breath by breath basis using ASYST (Asyst Software Technologies, Rochester, NY) to determine mean VE, "qo:, ~'co2 and the respiratory exchange ratio (R). A known volume of air containing known concentrations of 02 and CO2 was injected into the funnel over a period of about 2 sec. The recorded output of each gas analyzer was integrated using ASYST. The integrated values which represent known volumes of 0 2 and CO2 were used to convert expired gas concentrations to volume and flow. All measurements were made under BTPS conditions. Comparisons were made using analysis of variance with the level of significance set at P < 0.05. Another 6 snakes (108 + 15 g SD) were ctmsen for blood gas analysis. The snakes were anesthetized with Halothane (Ayerst, New York), and a 3-cm incision was made on the ventral surface to the left of midline approximately 15 cm from the heart posteriorly. The descending aorta was exposed, and two 2-cm lengths of PE90 tubing wer,~ inserted into the aorta, one facing upstream and one downstream. The two cannulae were then attached to a small 'T' connector, and a longer length of PEg0 tubing (approximately 20 cm), through which blood samples could be drawn, was attached to the side arm of the 'T' connector. This arrangement allowed continuous blood flow to the caudal tissues of the snake. The incision was closed and the animal was allowed to recover overnight. The following morning the snake was attached to the Plexiglas strip and placed into the plethysmograph as described above. Blood samples were collected anaerobically in 130-#1heparinized capillary tubes after withdrawal of dead-space blood into a small syringe. In some instances, the sample was withdrawn into a l-mi syringe, then transferred into the capillary tube. In most instances, however, the sampling syringe was removed and the capillary tube was placed snugly into the sampling port of the 3-way stopcock, allowing the snake's blood pressure to fill the capillary tube. After the sample was taken, the dead-space blood was reinjected into the aorta, followed by enough saline to fill the cannula. Oxygen and carbon dioxide pressures and pH were measured using a pH and blood gas analyzer (Radiometer BMS-3, Copenhagen) with the temperature of the electrodes maintained near the snake's body temperature (approximately 26 °C). Corrections for small temperature differences were made using equations presented by Nunn et al. (1965), Ruiz et al. (1975) and Severinghaus et al. (1956). The CO2 and O, electrodes were calibrated using humidified calibration gases, and the pH electrode was calibrated with precision phosphate buffers (Radiometer, Copenhagen). All calibrations were made 5 min before each sample was taken. Hematocrit was measured 3 or 4 times during the day (at the beginning, middle and end of the sequence of runs) using heparinized microcapillary tubes. All blood gas and hematocrit measurements were made immediately after the blood sample was taken. The protocol for blood gas experiments was similar to that described for the experiBlood gases.

VENOUS CO,, AND VENTILATION IN SNAKES

51

ments in which ventilation alone was measured, except that only air containing 0, 40 or 80~o CO2 was delivered to the plethysmograph, and only the last 10 min ofventilatory and expired gas values were analyzed for each condition. Ten minutes was chosen to represent the ventilatory history just before the blood sample was taken; this was a long enough period to reflect the steady state value for ventilation, reducing the effect of inter-breath variability of breathing frequency. Data obtained for runs in which the animal struggled were discarded. Four of the snakes were allowed to breathe air or air containing 2 or 4 ~ CO2 while the plethysmograph contained CO2-free air, and ventilation and blood gas values were measured. Because of the experimental setup, we could not measure oxygen consumption during lung loading with CO2. Three of these four snakes were later vagotomized, and all of the above measurements were made during cutaneous and inhaled CO2 uptake. Surgical preparation for vagotomy was performed with the snake under Halothane anesthesia (Ayerst, New York). The vagus nerves were exposed through an incision on the ventral surface approximately midway along the trachea, and two 2-cm lengths of thread were loosely wrapped around each vagus nerve. The wound was closed with a single stitch, and the animal was allowed to recover overnight. Measurements were made the following day with the vagus nerves intact. In the latter half of the day, the animal was removed from the plethysmograph, the single suture was cut, and the nerves were relocated with the aid of the marker thread and cut. The snake did not struggle or show any signs of discomfort during this procedure. The wound was closed with tape, and the animal was placed back into the plethysmograph, and given time to resettle before measurements were made. All animals were killed at the end of the day with an overdose of sodium pentobarbital followed by an intracardiac injection of KCI. The same statistical procedures were used as described above.

Results

Ventilation and gas exchange. The significantly elevated respiratory exchange ratio (R) measured from the expired air during high levels of cutaneous CO2 exposure suggests that there is a greater volume of CO2 in the expired air than can be accounted for by metabolic CO2 production (fig. 3). Excess carbon dioxide flux (that is, the amount of CO2 exhaled in excess of metabolically produced CO2) was calculated as follows: Excess CO2 flux = V c o 2 - (R'~'02), where R is the respiratory exchange ratio measured when the animal's skin was exposed to air containing no added CO2 (fig..3), ~'co2 is the volume of CO2 exhaled per minute during cutaneous CO2 exposure, and Vo: is the oxygen consumption during cutaneous CO2 exposure. The calculation is based on the assumption that the cellular respiratory quotient (RQ) is not influenced by the level of venous CO2. Figure 4 shows a significant positive correlation between excess CO2 flux and the amount of CO2 present at the skin. The relationship is either linear or slightly curved upward (concaved). The slight curve suggests an increased CO2

52

R.A. FURILLA et al.

2.0 m

m

1.6

1.2

/o/O

m

0.8

_o/° m

0.4 m

m

0 _1

n

I

0

I

20

L

I

I

I

l

40 60 Cutaneous COa (%)

80

Fig. 3. The relationship between pulmonary R and the concentration of CO2 at the skin. The vertical lines associated with symbols in this and all other figures represent + 1 SE. Errors smaller than the symbols are not shown.

production owing to an increased work of breathing at the higher COz concentrations, or enhanced cutaneous uptake of COz at the highest concentrations, possibly reflecting increased cutaneous blood flow. Minute ventilation was also significantly elevated (approximately 3-fold) during cutaneous exposure to 80% CO2 (fig. 5). The increase in ventilation was the result of increases in both frequency and tidal volume (table 1). The concentration of CO2 in the expired air, however, was not affected by that surrounding the skin (fig. 6), suggesting that ventilation was proportional to the CO2 load so that maximum intrapulmonary Pco., remained approximately constant. Although 02 A m

I

.~

0.2

T°' 0.16 0 0

m

0.12¸

~

m

m

mm

0

0.08 "

o ° 0.04 K

UJ

t.

0

i

I

20

I

I

I

I

40 60 Cutaneous CO2 (%)

I

I

80

Fig. 4. The relationship between the minute volume ofexhaled CO2 in excess ofthat produced metabolically and the concentration of CO2 at the skin. See the Methods section for details.

53

VENOUS CO2 AND VENTILATION IN SNAKES

20-

,.E 12

g

_o. a 4 n

0

- I

I

I

0

1

20

I

I

I

I

40 60 Cutoneous C09 (%)

.I

80

Fig. 5. The relationship between minute ventilation and the concentration of CO2 at the skin.

m

coz .~ 2

~9

Q L_ i m i=

X W

0 -I

0

I

I

20

I

I

I

,I

40 60 Cutaneous CO2 (%)

I

]

80

Fig. 6. The mean concentration of CO2 in the expired air and the difference between inspired and expired 02 as a function of cutaneous CO2.

extraction per breath decreased with increasing venous CO2 (fig. 6), total s u m p t i o n rose (table 1).

0 2 con-

Blood gases. Arterial Pco2 was usually higher than expired Pco: (fig. 7), which may

reflect the anatomical dead space, stratification of gases in the avascular air sac, or shunting of venous blood into the arterial circulation. Maximum intrapulmonary Pco: was unchanged during venous CO2 loading (fig. 6), suggesting arterial Pco: homeostasis. Figure 8, moreover, provides direct evidence that the increase in ventilation during cutaneous CO2 loading was not driven by arterial hypercapnia. Adding CO2 to the inspired air elevated arterial CO2 by 12 Tort, but ventilation was little changed compared with the arterially isocapnic ventilatory response to cutaneous

54

R.A. FURILLA et al. TABLE 1 Physiological variables recorded during cutaneous CO2 exposure in intact, non-cannulated snakes.

C02 (%)

0

10

ffmin- ~) ~'T(ml) Vo2(ml.100g-t.min - I ) n(N)

4.3 + 0.3 !.3 + 0.1 0.13+0.01 36(8)

20

4.4 + 0.3 1.3 + 0.1 0.13+0.01 13(7)

40

5.3 + 0.4* 1.4 + 0.1 0.14+0.02 14(7)

80

6.3 + 0.5* 1.6 + 0.1" 0.17+0.01" 15(8)

8.0 + 0.4* 2.2 + 0.2) 0.21 +0.02* 15(8)

n = the total number of trials. N = the number of animals. All values are means + 1 SE. * Significantly different from 0% CO2.

40-

0,-

ooo

.vv' ~ 0 -

•

•

~

•

0,~O •

~

g zo s ~

O-''O

I01

2.0 i

II

,

2.5 J

15

,

3.0 I

I

:5 5 I

19 23 Expired CO z

n

4.0 I

27

4.5 I

i

torr

31

Fig. 7. Scatter plot ofthe partial pressure of arterial carbon dioxide and expired CO=. Open circles represent data from intact snakes. Solid circles represent data collected after vagotomy. Data from all snakes and all cutaneous CO: levels were combined to create he scatter plot. The broken line represents the line of identity.

exposure to 80?/0 CO2. Bilateral vagotomy eliminated the difference in responses to inhaled and cutaneous CO2 uptake (fig. 8). Vagotomy also caused a reduction in breathing frequency and an increase in tidal volume (table 2), but minute ventilation was not significantly altered (fig. 8). Gratz (1984) described a similar ventilatory response to bilateral vagotomy in water snakes.

2,f

VENOUS CO2 AND VENTILATION IN SNAKES

o~ 13t

Intact

4

I

40

' [

oE

hi o~,

4 4k

Vagotomy

l

-~

~

"

0

0

17

21

55

25

i

.'~

29

33

37

41

POco 2 (torr) Fig. 8. The relationship between minute ventilation and mean arterial Pco2. The values associated with the open symbols represent the concentration of CO2 surrounding the skin, and the values associated with the solid symbols represent the concentration of inspired CO2. The errors for 40% cutaneous CO2 after vagotomy are offset.

0.`5 " O camp

0.4-

g 0.~? -

s

0.2-

0

•

...©fo;

. , ~ ' ~ i : b °°°°

o

o.I

OI0

/

o

I

I0

I

1`5

I

20

I

2,5

I

30

VE (ml.lOOg'l.min "1) Fig. 9. The relationship between oxygen consumption and minute ventilation. The solid symbols represent mean values for 0, 10, 20, 40 and 80% cutaneous CO2 from the left, respectively. The open symbols and the associated regression represent individual values.

Oxygen consumption increased linearly with increasing minute ventilation (fig. 9). The average increase in Vo: was nearly two-fold from 0 to 80% cutaneous CO2 with an approximately 3.5-fold increase in ventilation (table 1, fig. 9). Finally, hematocrit was 3(I.2 + 5.3 (mean + st)) at tl~.ebeginning of the experiments, and 19.6 + 1.9 after all blood samples were taken.

56

R.A. FURILLA et al.

+I +I

-I-I -I-I +I

it'l ¢"~I

¢'4

"="

t',l --

~I" r-- q

+I +I

+I +I +I

t(-

'~

¢~'~ ~ '

-d

o

+I +I

+I +I +I

- - ~ +I +I +I

~ o o +I +I +I

~ ~

--~~

4(.

d ~ +I +I

4(-

,~

.-- ¢',I ~ +I +I +I

_=

.=.

o +I +I +I +I +I +I

.,..,

¢,I

~J .J

8 ==

~. ~

~.

+I +I +I

,~. ~. o +I +I +I

+I +I +I +I ÷I ÷I

o

'r. ¢'4

+I +I -H

+I +I +I

+I +I +I -I-I "H -H

o

O ,,.= L.

o

•

o I

.c: llpl} ~

o~'"

t.

o

--,

[_ 0

--, u

II ~

VENOUS CO_, AND VENTILATION IN SNAKES

57

Discussion

During venous CO2 loading, the maximum intrapulmonary concentration of C O 2 appears to be regulated (fig. 6); however, the measured CO2 may not accurately reflect the concentration of CO2 in the faveolar region of the lung. Tracheal volume was measured in 2 snakes, and was approximately 150 ~1. Fresh air will likely fill the trachea at the end of inspiration (the breath-hold begins at end inspiration), and if we also assume, in an extreme case, that there is no passive diffusion of CO2 into the trachea during the breath-hold, then the concentration of CO2 in the faveolar region of the lung would have been underestimated. The percent of the measured tidal volume (table 1) represented by tracheal dead space would be 12, 11, 10, 9 and 7 ~ when exposed to 0, 10, 20, 40 and 80% CO2 at the skin, respectively; therefore, the concentration of CO2 in the faveolar region of the lung at end expiration would be 2.3, 2.3, 2.2, 2.2 and 2.0~o when exposed to 0, 10, 20, 40 and 80 ~ CO2, respectively. As cutaneous CO2 increased, maximum intrapulmonary CO2 may have decreased, suggesting an overcompensation for the venous CO: load. In any event, the maximum intrapulmonary CO2 concentration clearly did not rise with increasing cutaneous (venous) CO2 concentration. Some investigators have suggested that mean arterial Pco: or pH is not the only indicator of ventilatory drive by chemoreceptors, but that changes in the amplitude of the oscillations even around a constant mean may influence ventilation (for a review, see Eldridge and Millhorn, 1986). Grant et al. (1981) showed that arterial CO2 oscillations in mammals increased with venous CO2 load, but could not show a correlation between ventilation and CO2 oscillations. Arterial CO2 oscillations are also inversely related to breathing frequency (Band et al., 1969). Therefore, the tendency for arterial CO, oscillations to increase with venous CO2 loading in this study may have been counteracted by the higher breathing frequency associated with venous loading. Because maximum faveolar Pco., was unchanged or slightly lower during high venous loads (fig. 6), it is unlikely that the magnitude of arterial CO2 oscillations increased with elevated venous CO2. If neither mean arterial P¢o: nor maximum arterial or intrapulmonary Pco, increased with venous CO, load, then what stimulated the increase in ventilation ? If pulmonary arterial Pco: is elevated, then the higher gradient between the blood and intrapulmonary gas at end inspiration will cause an increase in the rate of excretion of CO2 into the lung. The increase in the rate of change of IPC activity at the beginning of the breath-hold brought about by the increase in the rate of rise of intrapulmonary CO2 might stimulate the central controller and cause an increase in ventilation. Furiila (1990) demonstrated that the rate of rise of CO2 early in the breath-hold, not the peak concentration of intrapulmonary CO2, determined the eventual length of that breath-hold. If some or all IPCs lie near small branches of the pulmonary artery, they might be in a position to monitor steady state pulmonary arterial Pco:, even though maximum intrapulmonary CO2 is essentially constant. Tallman and Grodins (1982), however, reported that IPC discharge (normalized to breathing cycle length) was not affected by the level of venous CO2 during any part of the breathing cycle despite a nearly two-fold

58

R.A. FURILLA et al.

change in breathing frequency. They further pointed out that for this to occur the rate of change of IPC discharge must vary with breathing frequency. Breathing frequency, therefore, may depend on the rate of change of IPC discharge. That arterial CO2 loading by way of elevating airway CO2 did not cause an even greater response than venous loading with high levels of CO2 complicates our interpretation of the role of IPCs in the control of breathing. If we ignore for the moment, however, the possible interactions between information from IPCs and systemic arterial or central chemoreceptors, reflex studies have shown that dynamic changes in IPC activity may influence ventilation (Furilla and Bartlett, 1989). Since the phasic pattern of IPC discharge will be reduced during CO2 breathing, the potential positive feedback ventilatory mechanism described by Furilla and Bartlett (1989) will also be reduced. Removal of this excitatory mechanism may explain why lung loading with CO2 did not drive ventilation even higher. That mean arterial Pco., is unaffected by the level of cutaneous CO2 in intact animals (fig. 8) suggests a COz regulatory mechanism located between the skin and systemic arterial circulation. It is unlikely that receptors located in the body wall are responsible for arterial COz homeostasis because that homeostasis was disrupted by vagotomy. Although we have focused on IPCs throughout this discussion, we emphasize that this study does not provide direct evidence that IPCs are responsible for the ventilatory response to venous CO2 loading. Intrapulmonary chemoreceptors are likely candidates for this role, but other receptors with afferents in the vagus nerves could also be responsible. Jackson and Braun (1979) showed that bullfrogs also maintained arterial isocapnia following cutaneous CO2 uptake, eliminating excess CO2 through the lung. Fedde and Kuhlmann (1978) were unable to demonstrate the presence of IPCs in bullfrogs. Finally, an elevation in venous Pco~, is normally the result of an increase in metabolism, as during exercise. During moderate exercise, many animals show no increase in arterial Pco,, and some may even experience a slight decrease in mean arterial Pco, (Faraci et al., 1984). The mechanism that drives ventilation during venous CO2 loading in snakes, whether it be IPCs or not, is likely to be important in the control of ventilation during exercise in these animals.

Acknowledgements.We thank Susan Knuth for her assistance throughout the study, and Dale Ward for his help with the data acquisition. This study was supported by grants HL19827 and HL07449 from the National Heart, Lung and Blood Institute.

References Band, D.M., !. R. Cameron and S.J.G. Semple (1969) Effect of different methods of CO2 administration on oscillations of arterial pH in the cat. J. Appl. Physiol. 26: 268-273. Boon, J. K., W. D. Kuhlmann and M. R. Fedde (1980) Control ofrespiration in the chicken: effects ofvenous CO2 loading. Respir. Physiol. 39: 169-181.

VENOUS CO., AND VENTILATION IN SNAKES

59

Eldridge, F. L. and D.E. Millhorn (1986) Oscillation, gating, and memory in the respiratory control system. In: Handbook of Physiology. Section 3. The Respiratory System, edited by N. S. Cherniack and J.G. Widdicombe. Bethesda, MD, American Physiological Society, pp. 93-114. Faraci, F. M., J. P. Kiley and M. R. Fedde (1984) Chemoreflex drive of ventilation during exercise in ducks. Pfliigers Arch. 402: 162-165. Fedde, M. R. and D. F. Peterson (I 970) Intrapulmonary receptor response to changes in airway-gas composition in Gallus domesticus. J. Physiol. (London) 209: 609-625. Fedde, M. R. and W. D. Kuhlmann (I 978) Intrapulmonary carbon dioxide-sensitive receptors: amphibians to mammals. In: Respiratory Function in Birds, Adult and Embryonic, edited by Piiper, J. Berlin, Springer-Verlag, pp. 33. Fedde, M.R., W.D. Kuhlmann and P. Scheid (1977) Intrapulmonary receptors in the tegu lizard: I. Sensitivity to CO:,. Respir. PhysioL 29: 35-48. Furilla, R.A. (1990)The rate of rise of intrapulmonary CO2 drives breathing frequency in garter snakes. J. Appl. Physiol. (Submitted). Furiila, R.A. and D. Bartlett, Jr. (1988) lntrapulmonary receptors in the garter snake (Thamnophis sirtalis). Respir. Physiol. 74:311-322. Furilla, R.A. and D. Bartlett, Jr. (1989) Intrapulmonary CO2 inhibits inspiration in garter snakes. Respir. Physiol. 78: 207-218. Grant, B.J.B., R.P. Stidwili, B.A. Cross and S.J.G. Semple (1981) Ventilatory response to inhaled and infused CO2: relationship to the oscillating signal. Respir. Physiol. 44: 365-380. Gratz, R. K. (1984) Effect of bilateral vagotomy on the ventilatory responses of the water snake, Nerodia sipedon. Am. J. Physiol. 246: R221-R227. Greco, E. C., Jr., W. E. Fordyce, F. Gonzalez, Jr., P. Reischl and F. S. Grodins (1978) Respiratory responses to intravenous and intrapulmonary CO2 in awake dogs. J. Appl. Physiol. 45:109-114. Jackson, D.C. and B.A. Braun (1979) Respiratory control in bullfrogs: cutaneous versus pulmonary response to selective CO2 exposure. J. Comp. Physiol. 129: 339-342. Jones, D.R., W.K. Milsom and P.J. Butler (1985) Ventilatory response to venous CO2 loading by gut vontilation in ducks. Can. J. ZooL 63: 1232-1236. Nunn, J. F., N.A. Bcrgman, A. Bunatyan and A.J. Coleman (1965)Temperature coefficients for Pco: and Pc)~, of blood ilt vitro. J. Appl. Physiol. 20: 23-26. Osborne, J. L., G. S. Mitchell and F. Powell (1977) Ventilatory responses to CO, in the chicken: intrapulmonary and systomic chemoroceptors. Respir. Physiol. 30: 369-382. Phillipson, E.A., G. Bowes, E. R. Townsend, J. Duffin and J. D. Cooper ( 1981 ) Carotid chemoreceptors in ventilatory responses to changes in venous CO2 load. J. AppL Physiol. 51: 1398-1403. Ruiz, B.C., W. K. Tucker and R.R. Kirby (1975) A program for the calculation of intrapuhnonary shunts, blood gases, and acid-base values with a programable calculator. Anesthesiology 42: 88-95. Severinghaus, J.W., M. Stupfel and A.F. Bradley (1956) Accuracy of blood pH and Pc:o2 determinations. J. Appi. Physiol. 9: 189-196. Stremol, R.W., D.J. Huntsman, R. Casaburi, B.J. Whipp and K. Wasserman (1978)Control ofventilation during intravenous CO2 loading in the awake dog. J. Appl. Physiol. 44:31 i-316. Tallman, R. D., Jr., and F.S. Grodins (1982) Intrapulmonary CO2 receptor discharge at different levels of venous Pco,. J. Appi. Physiol. 53: 1386-1391. Wasscrman, K., B.J. Whipp, R. Casaburi, D.J. Huntsman, J. Castagna and R. Lugliani (1975) Regulation of arterial Pco2 during intravenous CO2 loading. J. Appi. Physiol. 38: 651-656. Yamamoto, W. S. and M. W. Edwards, Jr. (1960) Homeostasis ofcarbon dioxide during intravenous infusion of carbon dioxide. J. Appl. Physiol. 15:807-818.