- Email: [email protected]
Episodic breathing in alligators: Role of sensory feedback
Respiration Physiology, 87 (1992) 77-90 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0034-5687/92/$03.50
77
RESP 01846
Episodic breathing in alligators: Role of sensory feedback M.A. Douse* and G.S. Mitchell Department of Comparative Biosciences. University of Wisconsin. Madison. Wisconsin. U.S.A. (Accepted 26 August 1991) Abstract. The episodic breathing pattern in many reptiles consists of two or more clustered breaths separated by variable non-ventilatory periods. This pattern is commonly postulated to result from oscillations in lung and/or blood Pc2 or Pco2 via chemoreceptor feedback. We tested this hypothesis by monitoring breathing pattern in: (1) awake, undisturbed alligators and (2) sedated alligators (approx. 25 mg/kg pentobarbital, i.p.; 3 days prior to data collection)~in sedated alligators, measurements were made: (1)before and after bilateral cervical vagotomy, a procedure that removes peripheral arterial chemoreceptors, CO,sensitive intrapulmonary chemoreceptors and pulmonary stretch receptors (n -- 6); and (2)during unidirectional ventilation (UDV) at high flow rates (> 2 L/min), thereby minimizing oscillations in lung and blood Pc2 and Pco2 (n -- 6). Measurements on sedated alligators were made at 30 and 20 °C in each ofthese conditions, in awake, undisturbed alligators, breathing was typically episodic with 2-7 breaths/cluster, although the pattern was easily altered (increased breaths/cluster) by even seemingly minor disturbances. In sedated alligators, episodic breathing was still evident after vagotomy, but only at increased inspired CO2; at 5 % CO2 four of six alligators exibited episodic breathing consisting of 2-3 breaths/cluster interspersed with occasional single breaths. An episodic breathing pattern was also evident during UDV; at low levels of CO2,2-4 breaths/cluster interspersed with occasional single breaths were evident in four alligators, while two had 6-8 breaths/cluster. Increasing CO: in the UDV gas stream generally increased the number of breaths/cluster. After vagotomy, all six alligators could manifest an episodic breathing pattern during UDV in at least one CO2 condition (> 2 breaths/cluster interspersed with occasional single breaths). The episodic breathing pattern was very labile, sometimes changing to single breaths without apparent cause. The results suggest: (!)episodic breathing requires neither feedback from vagal sensory receptors nor oscillations in respiratory gases; and (2) changes in arterial Pea, modulate, but do not initiate episodic breathing. Episodic breathing in alligators may be due to complex interactions of higher brain structures with the central rhythm generator.
Control of breathing, CO2 sensitivity, reptile; Pattern of breathing, periodic breathing, reptile; Reptile, Alligator mississipplensis; Vagus nerve, respiratory pattern, reptile
A fundamental question in comparative respiratory physiology concerns the genesis and control of episodic breathing patterns in reptiles and other lower vertebrates. Episodic Correspondence to: G.S. Mitchell, Department of Comparative Biosciences, University of Wisconsin-Madison, 2015 Linden Drive West, Madison, WI 53706, U.S.A. * Present address: Department of Physiology, University of Toronto, Toronto, Ontario, Canada.
78
M.A. DOUSE AND G.S. MITCHELL
breathing consists of two or more clustered breaths, separated by longer but highly variable apneas or non-ventilatory periods (cf Shelton et aL, 1986). Concomitant with episodic breathing are oscillations in lung and blood Po2 and PEG2. During a breath cluster, lung and blood Po: increase while Pco, decreases; the reverse occurs during a non-ventilatory period (Lenfant et al., 1970; Burggren and Shelton, 1979; Glass and Johansen, 1979). It has been postulated that the episodic breathing pattern results from these oscillations in Po: (Burggren and Shelton, 1979; Shelton and Croghan, 1988) or PeG, (West et al., 1989) acting via chemoreceptor feedback. According to this hypothesis, as the non-ventilatory period progresses, chemoreceptor drive progressively increases, eventually reaching a threshold that initiates breathing. Conversely, during a breathing bout, chemoreceptor drive diminishes until the ventilatory period terminates (Shelton and Croghan, 1988). We tested this hypothesis by monitoring the breathing pattern of alligators: (1) before and after bilateral cervical vagotomy, a procedure that removes peripheral arterial chemoreceptors (Benchetrit et al., 1977; Douse and Mitchell, 1992), CO2-sensitive intrapulmonary chemoreceptors and pulmonary stretch receptors (Powell et ai., 1988; Douse et al., 1989); (2)during ~midirectionai ventilation (UDV) at a high flow rate, thereby minimizing oscillations in lung and blood Pc>, and PEG,; and (3) during combined UDV and bilateral vagotomy. The results indicate that, although vagal receptors and oscillations in respiratory gases modulate quantitative aspects of the episodic breathing pattern, they are not essential to the genesis of episodic breathing.
Methods
Animals Except for the data recorded from unanesthetized, quiet alligators (n = 4; Series I), data described in this paper were collected during the same experiments as described in the preceding paper (Douse and Mitchell, 1992). Briefly, experiments were performed on 19 juvenile American alligators (,4. mississippiensis), ranging from 1.0 to 4.2 kg body mass (mean = 1.9 + 0.8 kg ( + SD)). The alligators were obtained from the Rockefeller Wildlife Reserve in Grand Chenier, LA. Animals were fed a diet of mice, with water readily available, and housed at 30 °C. Animals were fasted for > 1 week prior to experiments. Series i: Effects of anesthesia. In Series !, breathing movements were recorded from four unanesthetized alligators by inductive plethysmography (Respitrace Corp.). The alligators were placed in a dark chamber maintained at 30 °C, and at least 12 h were allowed to pass before data collection commenced. Measurements were made for I h periods, three times per day. Care was taken to remain quiet during measurements. The effects of the anesthetic protocol on ventilatory pattern were studied in three additional alligators at 30 °C without other experimental manipulations. General
EPISODIC BREATHING IN ALLIGATORS
79
anesthesia was induced with pentobarbital (approx. 25 mg/kg; i.p.). This anesthetic protocol resulted in a surgical level of anesthesia on the preparation day, and a stable but light anesthesia level on the two subsequent days of data collection. Further pentobarbital was administered as required to maintain stable anesthesia during data collection. The animals were studied in ventral recumbancy while body temperature (Tb) was monitored with a cloacal thermistor and regulated with a temperature controlled table, The femoral artery and vein were cannulated and the animals were intubated with an endotracheal tube. Series H: Effects of vagotomy. The effect of bilateral cervical vagotomy on breathing pattern was determined in six spontaneously breathing, sedated alligators. Breathing pattern was monitored during control conditions (room air) and with inhaled CO2 (5 % CO2 in 20-30% 02, balance N2). At least 30 min were allowed to achieve steady state at. each CO2 level; ventilatory measurements and an arterial blood sample were obtained during the subsequent 30 min. The vagi were bilaterally sectioned following application of local anesthetic (2% lidocaine), and at least 2 h were allowed to pass before the protocol was repeated. Vagal denervation was confirmed by the absence of response to intravenous injection of NaCN (50-100/~g/kg), a dose that elicits a brisk ventilatory response in intact alligators (Douse and Mitchell, 1992). Series 111: Unidirectionalventilation. The effects of unidirectional ventilation on breathing pattern were determined in six sedated alligators. Three heat-flared cannulae were inserted at the base and sides of each lung and secured with purse-string sutures. The body wall and skin were closed separately. Incisions were infiltrated with local anesthetic (2% lidocaine). The alligators were unidirectionallyventilated (UDV) by passing a humidified gas stream (4% CO2; 30-40% 02 in N2; >2 L/rain) through each cannula, into each lung and out via the endotracheal tube. Water temperature in the humidifier was maintained at the alligator's Tb, ensuring that lung temperature approximated Tb. The breathing pattern during U DV was observed at different levels of Flco~ in the UDV gas stream. At least three different COz levels were delivered to the lungs. The first and last levels were always FIco: = 0.04. The sequence of administration of the remaining levels was random and consisted of two or more of the following: Flee., = 0.02; 0.06; 0.08; and/or 0.09 (in 30-40% 02, balance N2). Different CO2 levels were used as necessary to elicit breathing movements in at least two conditions. At least 30 rain were allowed to pass at each Flco~ level to achieve steady state before ventilatio, i was measured (30 min) and an arterial blood sample was obtained. On the second day of Series Ill experiments, the vagi were sectioned and the protocol was repeated. Efficacy ofvagal denervation was assessed by lack ofventilatory response to NaCN (50-100/tg/kg; i.v.). In both Series II and IIl experiments, protocols were conducted at body temperatures of 30 and 20 ° C. Changes in body temperature took 30-60 rain.
80
M.A. DOUSE AND G.S, MITCHELL
Measurements Arterial blood pressure in the femoral artery and airway pressure in the endotracheal tube were monitored with pressure transducers (Statham, P23-id). Levels of CO2 and 02 in the gas stream were measured with an infrared CO 2 analyzer (Anarad, model pro-20) and an oxygen analyzer (Teledyne, model 33 IXB). Ventilation was monitored in the lightly anesthetized alligators with a pneumotachograph attached to the endotracheal tube (Flcisch, type 00). The pressure drop across the pncumotachograph was measured with a differential pressure transducer (Validyne, MP45) and integrated to give tidal volume (Gould). The continuous airflow during UDV in Series IIl (bias flow) was electronically offset; the pneumotachograph was calibrated during the apneic periods. The response of this system was linear over the ventilation range observed. Arterial blood was sampled by withdrawing 0.2-0.5 ml blood into a syringe and then withdrawing 0.3 ml blood anaerobically into a heparinized 0.5 ml glass syringe. The blood-saline mixture was then returned to the animal and the cannula was flushed with fresh heparinized saline. Analysis of arterial Pco2 and Po_, began within I rain of sampling using a micro gas analyzer (Radiometer, model BMS 3 MKII, PHM73) maintained at the animal's body temperature. Electrodes were calibrated before and after each measurement: the Po: electrode was calibrated with air-saturated water; the P¢o.~ electrode was calibrated with saturated gases provided by a gas-mixing device
(5.61 ~o and 11.22% CO2; Radiometer, GMA2) and a precision-calibrated gas tank (2.53% CO2). Accuracy of the calibration gases was confirmed with a WOsthoff gas mixing pump. A#aS;vis It is difficult to develop a formal and objective definition of episodic breathing due to the variable and labile breathing patterns observed. For our purposes, episodic breathing is functionally defined as consisting of two or more breaths in a breath cluster (breaths/cluster), separated by longer non-ventilatory periods. Intermittent breathing is also a useful term to describe breathing patterns in reptiles and amphibians, and connotes any breathing pattern where the breath hold between breaths is longer than the active ventilatory cycle (Milsom, 1990). Intermittent breathing is therefore a more general term, refering to episodic breathing, episodic breathing interspersed with single breaths, and single breaths separated by breath holds of longer duration. The distinction between a breath hold within a breath cluster and a non-vvntilatory period between breath clusters in episodic breathing is somewhat arbitrary (Milsom and Jones, 1980). Precise quantification is further confounded by the effects of experimental manipulations such as vagotomy and unidirectional ventilation on the duration of the ventilatory cycle, Therefore, we restricted attention to gross, qualitative changes in the overall breathing pattern, and did not attempt precise quantitation of ventilatory periods vs non-ventilatory periods (e.g. Naifeh etal., 1971a,b). Our interest was to assess qualitatively whether episodic breathing patterns (as defined above) persist after systematic removal of relevant sensory feedback mechanisms.
EPISODIC BREATHING IN ALLIGATORS
81
Results
Series I: Effects of anesthesia. Respiratory activity in unanesthetized and unrestrained alligators showed considerable variability. This variability was observed in the same alligator on the same day (Fig. IA), or between alligators (Fig. 1A and B). Typically the pattern consisted ofbreath clusters interspersed with inspiratory breath holds ofvarious durations. Three of the four alligators had a quiet, resting breathing pattern consisting of 2-7breaths/cluster. One alligator, however, consistently had episodes of 5-16 breaths/cluster. It was extremely easy to disturb the animals. Although animals were kept in a dark chamber separated from the rest of the room, disruption of the ongoing breathing pattern was often evident when experimenters quietly entered the opposite end of the room (Fig. IB). Over 30 breaths/cluster were observed in excited animals. Thus, unanesthetized alligators are extraordinarily sensitive to small disturbances resulting in large breathing pattern alterations. Anesthesia 3 days prior to data collection resulted in sedated, but spontaneously breathing alligators. In general, the anesthetic protocol decreased the number of breaths/cluster, and prolonged the respiratory cycle time (cf Figs. I and 2). There were also time-dependent changes in breathing pattern over the duration of an experiment. All three alligators maintained in constant conditions at 30 °C had an episodic breathing pattern by the end of the experiment (Fig. 2), consisting of 2-3 breaths/cluster interspersed with occasional single breaths. However, only one alligator had a clear episodic breathing pattern at the beginning of data collection (2 breaths/cluster; Fig. 2A). The other two animals exhibited intermittent breathing, which progressed to episodic breathing (2-3 breaths/cluster) by the end of data collection (Fig. 2B). Changes in
Be
All
mmmn i
W
=. .
.
.
cmomm .
," -
i~
!
t'W
Fig. !. Traces showing variability and sensitivity of the breathing pattern to a minor disturbance in two unanesthetized alligators at 30 °C (Series !). in A: top trace is breathing movements recorded from one alligator < 30 rain after placement of Respitrace; middle trace was 5 h and bottom trace was 9 h later. In B: top trace represents breathing movements in a different alligator 2 h after Respitraee placement; m~ddle trace = 5 h and bottom trace = 9 h later. At arrow, the experimenter quietly entered the room, approximately 8 m distant. Even though there was no obvious noise associated with this disturbance and the alligator was separated by 8 m and two visual or sound barriers, a disturbed ventilatory pattern was consistently evident. The ventilation monitor was not calibrated, and the traces provide only a qualitative index of respiratory amplitude.
82
M.A. DOUSE AND G.S. MITCHELL
A.
v,
B, otOOr
i|O0~
.8|
-|co
.
i.°oi
-----'u
v,
Vr
ili
-
-
Vt -80
-
-80 !
(;I) e I
Fig. 2. Effects of anesthetic protocol, tracheal intubation and time on breathing pattern in two spontaneously breathing alligators. In A, an episodic breathing pattern was evident throughout the experiment, in contrast, the alligator in B progressively developed a greater degree of episodic breathing. In both A and B, the top trace was recorded < 30 min after experiment began; middle trace was 9 h and bottom trace was 33 h later. Presence of the tracheal tube prevents an inspiratory breath hold. VT = tidal volume.
breathing pattern with time were not associated with significant changes in breathing frequency (f), tidal volume (VT), minute ventilation (Qi) or Pace,. Series H: Effects of mgotomy. The anesthetic protocol also tended to decrease the number of breaths/cluster in Series II compared to unanesthetized animals, although this effect was less marked than in Series I (Fig. 3). Five of six intact alligators breathing room air had an episodic breathing pattern consisting of 2-4 breaths/cluster interspersed with occasional single breaths (Fig. 3). One alligator had a stable intermittent pattern of single breaths. Increasing inhaled CO2 to 5% increased VT, f and V]. Increased f was due to decreases in the non-ventilatory period and an increase in the number of breaths/cluster. In four alligators, 3-8 breaths/cluster were observed while two animals had between 12 and 30 breaths/cluster (Fig. 3). Altar bilateral cervical vagotomy, intermittent but not episodic breathing was observed in five of six alligators breathing room air (Fig. 4A). The remaining alligator had a breathing pattern that is difficult to characterize (Fig. 4B). Increasing inhaled CO2 to 5?/o post-vagotomy reverted four of six alligators to an episodic breathing pattern A.
e.
VT
i,
J,
i
I
I
{|
VT
i i
'°° -tOO
-II
II
i J l
i, L ,ll,tll q--
! 5 mm I Fig. 3, CO~ effects on the episodic breathing pattern of two intact, spontaneously breathing alligators (Series !!), In both A and B, the top trace is breathing 0% CO2 and the bottom trace is breathing 5% CO.~.
83
EPISODIC BREATHING IN ALLIGATORS
B.
AI
I s=,~l Fig. 4. Effects of bilateral cervical vagotomy on the episodic breathing pattern in two spontaneously breathing alligators (Series If). Same alligators as in Fig. 3A and B. In both A and B, the top trace is breathing 0% CO, and the bottom trace is 5 ~ CO2.
(Fig. 4). The episodic breathing pattern usually consisted of two and sometimes three breaths/cluster interspersed with occasional single breaths. Two alligators exhibited intermittent single breaths, regardless of CO2 level. There were no significant effects of CO2 on mean f, VT or ~'I post-vagotomy. Episodic breathing was evident in all six alligators during unidirectional ventilation (Figs. 5 and 6). Typically, the breathing pattern consisted of 2-3 breaths/cluster, separated by non-ventilatory periods of several minutes duration (Fig. 5). The pattern of breathing, however, was extremely variable both within and between alligators (Fig. 6). Increasing CO2 levels in the UDV gas stream increased mean f, VT and VI. The increase in fwas through a decrease in the non-ventilatory period and, in three alligators, through an increase in the number of breaths/cluster. At low CO2 levels (2-4 ~o), three alligators did not breathe. At mid CO2 levels (4-6 ~), the bre:'~thing pattern consisted of 2-4 breaths/cluster, interspersed with occasional single breaths in Series 111: Unidirectional ventilation.
VT
Po,
*|001 *4
.o ~ - - - ~ t - ~
. . . . . . . .
:
0.10r Flco2 I 0.06
0 u.
Fig. 5. Episodic breathing despite unidirectional ventilation, minimizing oscillations in respiratory gases in a sedated alligator (Series Ill). FIco2 =0.04 and Flo2 =0.35 in the unidirectional gas stream {Paco~ = 36 mmHg and Pao2 = 205 mmHg). Paw = airway pressure.
84
M.A. DOUSE AND G.S. MITCHELL
All
B. VT
':I
A
-IIOIIt.
C, VT "a°°I
.nl v,.
D. ,,
Fig, 6. Variability of episodic breathing patterns during unidirectional ventilation in four sedated alligators (Series III). In A, FIco2 -- 0.04 (Patio, = 36 mmHg; Pao, = 205 mmHg), in B, Fuco, = 0.08 (Paco~ = 63 mmHg; Pao, = 198mmHg). In C, Flco2 =0.04 (Paco,= 44mmHg; Pao, = 207mmHg). In D, Flco, = 0.08 (Paco, = 69 mmHg; Pao: = 134mmHg).
four alligators (Fig. 6A and C). The two remaining alligators had 6-8 breaths/cluster. At high COa 10vels(6-10%), tht'¢."alligators had 2-4 breaths/cluster (Fig. 6B), two had 6-12 breaths/cluster (Fig. 6D), and one alligator went from an episodic breathing pattern to a virtually continuous breathing pattern with few, if any, non-ventilatory periods. After bilateral vagotomy, all six alligators could still manifest an episodic breathing pattern in at least one condition of CO2 during UDV (Fig. 7). This pattern typically consisted of 2-3 breaths/cluster separmed by variable periods of apnea and interspersed with occasional single breaths. Although all alligators could manifest an episodic breathing pattern during UDV after vagotomy, this pattern was very labile and could change to single breaths despite apparently constant conditions (Figs. 7C and 8A). Increasing the CO 2 level in the UDV gas stream increased mean f, VT, and VI. At low CO 2 levels (4'~o), five of six alligators did not breathe. At higher CO 2 levels (6-9%), four of six alligators had a breathing pattern consisting of two breaths/cluster interspersed with occasional single breaths (Fig. 7A and C). The remaining two alligators had an episodic breathing pattern of 2-5 breaths/cluster (Fig. 7B and D).
Effects of temperature. Episodic breathing was evident at both 30 and 20 °C in all experimental series (Fig. 8). The results were generally similar at both temperatures,
EPISODIC BREATHING IN ALLIGATORS
85
A.
LLIt
J.
I
-20
BII
~200~ VT
11
e
-2on
11
o -800
l l
4t tt
t¸
CI v,
1111II
Dill
,t200F
w
o
Ill
~[1
LL
.,.oE
llr
lJl
II
Fig, 7. Effects 0fbilaterai vagotomy and unidirectionalventilation on the episodic breathing pattern in four sedated alligators (Series II1). These are the same alligators as in Fig. 6. In A, Flee2 = 0.06 (Pace2 ffi 50mmHg; Pao2 = 158mmHg). In B, FIco~ ffi 0.08 (Pace2 ffi 59mmHg; Pao~-216mmHg). In C, Free: ffi0.08 (Pace2 = 7S mmHg; Pao2 ffi 110 mmHg). In D, Flee2 = 0.09 (Paco: = 70 mmHg; • Pao, ffi 88 mmHg).
although the pattern was slightly different at 20 ° C. Often there were longer breath hold periods within a breath cluster and longer non-ventilatory periods between clusters at 20 °C than at 30 °C, even in the same animal (Fig. 8). Both of these factors increased the difficulty of characterizing the breathing pattern as episodic v s intermittent. In general, there was a decrease in the number of breaths/cluster at 20 °C in any given condition (range = 2-7 breaths/cluster).
86
M.A. DOUSE AND G.S. MITCHELL
A. 30 °C
"2001
ill II j J
VT
B. 20 "C
'~200r... ] VT
,
i
Fig. 8. Breathing pattern recorded from the same alligator as in Fig. ?A after bilateral vagotomy and during unidirectional ventilation at body temperatures of 30 °C (A) and 20 °C (B) (Series Ill). In both A and B, Flee= = 0.06 and Fie, = 0.3. In A, Pace= = 47 mmHg and Pao2 -- 178 mmHg. In B, Pace2 = 50 mmHg and Pao, = 158 mmHg.
Diseussion
The fundamental question addressed in this paper is: Can an episodic breathing pattern be observed after removing important sources of (vagal) chemoreceptor feedback or with minimal oscillations in respiratory gases? The results suggest that: (1)episodic breathing does not require feedback from vagal sensory receptors (arterial chemoreceptors, CO2-sensitive intrapulmonary receptors and pulmonary stretch receptors) and (2)steady-state CO 2 levels affect episodic breathing patterns quantitatively, but oscillations in respiratory gases do not appear to be necessary for the expression of episodic breathing.
Critique ofmethods. A formal, quantitative analysis of the episodic breathing pattern was not attempted due to the large variation within and between alligators, the labile nature of the episodic breathin~ pattern and definitional concerns (see Analysis). Instead, our approach was to qaalitatively assess the effects of bilateral vagotomy, unidirectional ventilation, anesthesia, C02 and temperature on the episodic breathing pattern. The episodic breathing pattern observed after anesthesia, vagotomy or UDV, for example, was in many respects different from that observed in intact, unanesthetized alligators. However, our effort was to test hypotheses concerning the role of sensory feedback in the initiation and control of episodic breathing in a qualitative sense. Our
EPISODIC BREATHING IN ALLIGATORS
87
objective was to elucidate the minimal mechanism(s) necessary for expression of an episodic breathing pattern in alligators. Arterial Pco2 was not continuously monitored during UDV. The crocodilian lung has a multicameral architecture (Perry, 1988) that could allow air flow to stream through central bronchi without thoroughly ventilating some chambers. Thus, breathing efforts may be necessary to ventilate these areas, potentially changing regional lung and blood gas composition with each breath. Since 4% CO 2 in the UDV gas stream was used to establish normal blood gas values, the high inspired CO2 would minimize Pace2 oscillations despite possible changes in regional lung ventilation with each breathing effort. Indeed, serial samples indicated that Pace, varied less than 2 mmHg. Given the low ventilatory sensitivity to Pace, in these alligators (Douse and Mitchell, 1992), such changes in Pace2 would not have a noticeable effect on ventilation (see below), Further elevations in inspired CO2 would minimize CO2 oscillations even more. Nevertheless, elevated CO, levels increased rather than decreased the incidence of episodic breathing. Arterial Po2 measurements in Series III indicated that ventilation-perfusion mismatch and/or cardiopulmonary shunting may have occasionally been present in some alligators (see Fig. 7D). Changes in cardiopulmonary shunt during UDV could conceivably lead to oscillations in blood gases, resulting in an episodic breathing pattern. However, it is unlikely that changes in cardiopulmonary shunt would occur over the same, short time scale necessary to initiate and terminate an episode of breathing in a breath cluster. Moreover, most alligators could manifest an episodic breathing pattern despite high background levels of Pace,, when ventilation-perfusion mismatch or cardiopulmonary shunt was not indicated.
Series i: Effects of anesthesia. Unanesthetized alligators in conditions of minimal environmental stimulation have an episodic breathing pattern typically consisting of 2-3 breaths per cluster (Huggins et aL, 1969). A pattern of 2-7 breaths/cluster was observed in the unanesthetized alligators in this study. However, an increase in the number of breaths per cluster was observed in excited animals (see Fig. 1), emphasizing the considerable variability and labile nature of the episodic breathing pattern. Light levels of anesthesia were used in order to reduce discomfort to the animal and to maintain a constant 'state'. Under conditions of light anesthesia, breathing pattern closely mimicked the pattern recorded from unanesthetized alligators after prolonged periods with minimal environmental stimulation (this study, Fig. 1; Huggins etal., 1969). The anesthetic protocol did tend to decrease the number of breaths per cluster and prolong the breath hold period within a breath cluster, as has been noted previously in crocodilians (Naifeh et aL, 1971a). Collectively, the results suggest that both light anesthesia and arousal state affect breathing pattern in alligators in a quantitative rather than qualitative manner. Series H: Effects of vagotomy. An episodic breathing pattern was not generally observed during spontaneous breathing after bilateral vagotomy, except when respiratory drive was increased with inhaled CO2. This may suggest that peripheral arterial
88
M.A. DOUSE AND G.S. MITCHELL
chemoreceptors and/or the pulmonary receptors have a role in the manifestation of episodic breathing in alligators. However, these sensory pathways are not necessary to generate episodic breathing as demonstrated by the return of episodic breathing at elevated CO2. The effect of CO 2 on breathing pattern following vagotomy contrasts with the nearly complete aboiition of the ventilatory response to hypercapnia in the same experimental conditions (Douse and Mitchell, 1992). In a previous study, bilateral vagotomy did not abolish episodic breathing and only slightly decreased the number of breaths per breath cluster in unanesthetized alligators and caimans (Naifeh et al., 197 lb). This difference in results may be due to the residual anesthetic effects in our study, decreasing respiratory drive and suppressing episodic breathing. Although not specifically commented on by the authors, an episodic breathing pattern may also be observed in turtles after bilateral vagotomy (Frankel et a!., 1969; Milsom and Jones, 1980). There were still oscillations in Pace_, after vagotomy in spontaneously breathing alligators, which may have affected non-vagal CO2-sensitive areas, leading to an episodic breathing pattern. Although increasing inhaled CO2 would decrease Paco~ oscillations, increasing inspired CO2 increased rather than decreased the incidence of episodic breathing (Fig. 3). Episodic breathing was still evident despite unidirectional ventilation. Since oscillations in lung and blood CO2 and 02 were minimized by UDV, oscillations in chemoreceptor feedback from any source is unlikely to have initiated episodic breathing. Episodic breathing has also been observed recently in intact toads during unidirectional ventilation (West et ai., 1987). We also observed episodic breathing during U DV after bilateral vagotomy, but the pattern was very labile and could change to single breath intermittent breathing despite apparently constant conditions and with no obvious state transitions. Nevertheless, qualitatively similar episodic breathing patterns were observed in this study despite vagotomy, unidirectional ventilation, unidirectional ventilation with vagotomy and changes in body temperature. Models of mammalian periodic breathing may be useful in elucidating the mechanism(s) initiating episodic breathing in reptiles. These models suggest that respiratory in stability and periodic or episodic breathing may arise from: (1) increases in circulatory dela~ from lungs to chemoreceptors; (2) increased chemoreceptor sensitivity; and (3) changes in a chemical threshold for ventilatory acti~,ity (i.e. apneic threshold; Khoo eta l., 1982; Cherniack and Longobardo, 1986). A fundamental assumption in all models of unstable breathing is that the system is operating as a closed chemoreceptor feedback loop (el Cherniack and Longobardo, 1986). Further, all of these models rely on Pace, or Pao, oscillations either to initiate or to maintain respiratory instability (Khoo etaL, 19821 Cherniack and Longobardo, 1986). in our study, UDV opens chemoreceptor feedback loops and minimizes changes in Pace, or Pao:. In addition, episodic breathing was observed after vagotomy despite decreases in both 02 and CO2 ventilatory sensitivity and at Pace, levels well above the apneic threshold (Douse and Mitchell, 1992). Thus, even if there are small changes in CO2/H + in the microSeries Ill: Unidirectional ventilation.
EPISODIC BREATHING IN ALLIGATORS
89
environment of central chemosensitive areas during UDV (post-vagotomy), it is unlikely that such changes would lead to respiratory instability and episodic breathing since the response to large changes in Paco: during UDV after vagotomy is sluggish and relatively small (Douse and Mitchell, 1992). Collectively, the results suggest that episodic breathing in alligators does not require vagal sensory feedback from peripheral arterial chemoreceptors, CO2-sensitive intrapulmonary chemoreceptors, or pulmonary stretch receptors. Further, it appears that tonic stimulation of both vagal and non-vagal CO2-sensitive chemoreceptors modulates, but does not initiate the ongoing rhythm. Finally, the episodic breathing pattern does not require oscillations in respiratory gases or oscillations in afferent vagal feedback and, therefore, may be inherent to brainstem mechanisms producing the respiratory motor output. Although the experimental manipulations used in this study have not eliminated all possible sources of oscillatory sensory feedback (e.g. chest wall receptors), the results strongly suggest that the essential mechanisms generating the episodic pattern are intrinsic to the central neural networks associated with ventilatory control. Indeed, strong support for this hypothesis has been provided by our recent observation that an in vitro brainstem-spinal cord preparation from adult turtles is capable of producing an episodic motor output similar to breathing in the whole animal, without peripheral inputs and despite constant CO, and 02 conditions (Douse and Mitchell, 1990). Thus, the episodic breathing pattern in reptiles appears to be inherent to brainstem mechanisms generating respiratory rhythm. This does not rule out a role for oscillatory afferent feedback in modulating the ongoing episodic pattern. In intact, unanesthetized alligators, 3 or more breaths per cluster are characteristic of episodic breathing (Huggins et al., 1969; Naifeh et al., 197 la,b; this study). However, 2-3 breaths per cluster is more commonly observed after mild anesthesia, bilateral vagotomy (Naifeh et al., 197 la,b; this study), or unidirectional ventilation (this study), suggesting that vagal feedback and state have roles in modulating quantitative details of episodic breathing in reptiles. Acknowledgements. We are grateful to Ted Joanen and Larry McNease of the Rockefeller Wildlife Refuge
for their help in procurement of alligators and to Dr. W.K. Milsom for loan of the Respitrace. (3.S. M. was supported by a Research Career Development Award from NIH (HL01494).
References
Benchetrit, G., J. Armand and P. Dejours (1977). Ventilatory chemoreflex drive in the tortoise, Testudo horsJleldi. Respir. Physiol. 3I: 183-191. Burggren, W.W. and (3. Shelton (1979). Gas exchange and transport during intermittent breathing in
chelonian reptiles. J. Exp. Biol. 82: 75-92. Cherniack, N.S. and (3.S. Longobardo (1986). Abnormalities in respiratory rhythm. In: Handbook of Physi~logy, Section3: The Respiratory System, Vol.ll; edited by N.S. Cherniack and J.(3. Widdicombe. Bethesda, MD: American Physiology Society, pp. 729-749. Douse, M.A., F.L. Powell, W.K. Milsom and (3. S. Mitchell (1989). Temperature effects on pulmonary receptor responsesto airwaypressure and CO2 in Alligator mississippiensis. Respir. Physiol. 78:331-344.
90
M.A. DOUSE AND G.S. MITCHELL
Douse, M. A. and G. S. Mitchell (1990). Episodic respiratory related discharge in turtle cranial motoneurons: In vitro and in rive studies. Brain Res. 536: 297-300. Douse, M.A. and G. S. Mitchell (1992). Effects of vagotomy on ventilatory responses to CO2 in alligators. Respir. Physiol. 87: 63-76. Frankel, H. M., A. Spitzer, J. Blaine and E. P. Schoener (1969). Respiratory response of turtles (Pseudemys scripta) to changes in arterial blood gas composition. Comp. Biochem. Physiol. 31: 535-546. Glass, M. L. and K. Johansen (1979). Periodic breathing in the crocodile, Crocodylus niloticus: Consequences for the gas exchange ratio and control of breathing. J. Exp. Zool. 208: 319-326. Huggins, S.E., H.E. Hoff and R.V. Pena (1969). Heart and respiratory rates in crocodilian reptiles under conditions of minimal stimulation. Physiol. Zool. 42: 320-333. Khoo, M.C., R.E. Kronauer, K.P. Strohl and A.C. Slutsky (1982). Factors inducing periodic breathing in humans: A general model. J. Appl. Physiol. 53: 644-659. Lenfant, C., K. Johansen, J.A. Petersen and K. Schmidt-Nielsen (1970). Respiration in the fresh water turtle, Chelys fimbriata. Respir. Physiol. 8: 261-275. Milsom, W.K. and D.R. Jones (1980). The role of vagal afferent information and hypercapnia in control of the breathing pattern in Chelonia. 3. Exp. Biol. 87: 53-63. Miisom, W. K. (1990). Mechanoreceptor modulation ofendogenous respiratory rhythms in vertebrates. Am. J. Physiol, 259: R898-R910. Naifeh, K.H., [i.E. Huggins and H.E. Heft (1971a). Study of the control of crocodilian respiration by anesthetic dissection. Respir. Physiol. ! 2:251-260. Naifeh, K. H., [i. E. Huggins and H.E. Heft (1971b). Effects of brain stem section on respiratory patterns of crocodilian respiration. Respir. Physiol. 13: 186-197. Perry, [i.F. (1988). Functional morphology of the lungs of the nile crocodile, Crocodylus niloticus: nonrespiratory parameters. J. Exp. Biol. 134: 99-117, Powell, F. L., W, K, Milsom and G.S. Mitchell (1988). Effects of intrapulmonary CO2 and airway pressure on pulmonary regal afferent activity in the alligator. Respir. Physiol. 74: 285-298. [ihelton, O., D,R. Jones and W,K, MUsom (1986), Control of breathing in ectothermic wrtebrates. In: Handbook of Physiology, [iection 3: The Respiratory System, Vol. 2; edited by A.P. Fishman, N. $. Cherniack, J.O. Widdieombe and [i. R. Geiger. Washington, DC: American Physiological Society, pp. 857-909. Shelton, G, and P,C. Croghan (1988), Gas exchange and its control in non-steady-state systems: The consequences of evolution from water to air breathing in the vertebrates, Can. J, Zool. 66: 109-123, West, N. H., Z, L, Toper and ii, N, Van Viler 0987), Hypoxemie threshold for lung ventilation in the toad. Resplr, Physlol, 70: 377-390. West, N.H., A.W. Stairs and W,W, Burggren (1989~ Factors terminating nonventilatory periods in the turtle, Chelydra serpenthea, Resplr, Physiol. 7"/: 337-350.
77
RESP 01846
Episodic breathing in alligators: Role of sensory feedback M.A. Douse* and G.S. Mitchell Department of Comparative Biosciences. University of Wisconsin. Madison. Wisconsin. U.S.A. (Accepted 26 August 1991) Abstract. The episodic breathing pattern in many reptiles consists of two or more clustered breaths separated by variable non-ventilatory periods. This pattern is commonly postulated to result from oscillations in lung and/or blood Pc2 or Pco2 via chemoreceptor feedback. We tested this hypothesis by monitoring breathing pattern in: (1) awake, undisturbed alligators and (2) sedated alligators (approx. 25 mg/kg pentobarbital, i.p.; 3 days prior to data collection)~in sedated alligators, measurements were made: (1)before and after bilateral cervical vagotomy, a procedure that removes peripheral arterial chemoreceptors, CO,sensitive intrapulmonary chemoreceptors and pulmonary stretch receptors (n -- 6); and (2)during unidirectional ventilation (UDV) at high flow rates (> 2 L/min), thereby minimizing oscillations in lung and blood Pc2 and Pco2 (n -- 6). Measurements on sedated alligators were made at 30 and 20 °C in each ofthese conditions, in awake, undisturbed alligators, breathing was typically episodic with 2-7 breaths/cluster, although the pattern was easily altered (increased breaths/cluster) by even seemingly minor disturbances. In sedated alligators, episodic breathing was still evident after vagotomy, but only at increased inspired CO2; at 5 % CO2 four of six alligators exibited episodic breathing consisting of 2-3 breaths/cluster interspersed with occasional single breaths. An episodic breathing pattern was also evident during UDV; at low levels of CO2,2-4 breaths/cluster interspersed with occasional single breaths were evident in four alligators, while two had 6-8 breaths/cluster. Increasing CO: in the UDV gas stream generally increased the number of breaths/cluster. After vagotomy, all six alligators could manifest an episodic breathing pattern during UDV in at least one CO2 condition (> 2 breaths/cluster interspersed with occasional single breaths). The episodic breathing pattern was very labile, sometimes changing to single breaths without apparent cause. The results suggest: (!)episodic breathing requires neither feedback from vagal sensory receptors nor oscillations in respiratory gases; and (2) changes in arterial Pea, modulate, but do not initiate episodic breathing. Episodic breathing in alligators may be due to complex interactions of higher brain structures with the central rhythm generator.
Control of breathing, CO2 sensitivity, reptile; Pattern of breathing, periodic breathing, reptile; Reptile, Alligator mississipplensis; Vagus nerve, respiratory pattern, reptile
A fundamental question in comparative respiratory physiology concerns the genesis and control of episodic breathing patterns in reptiles and other lower vertebrates. Episodic Correspondence to: G.S. Mitchell, Department of Comparative Biosciences, University of Wisconsin-Madison, 2015 Linden Drive West, Madison, WI 53706, U.S.A. * Present address: Department of Physiology, University of Toronto, Toronto, Ontario, Canada.
78
M.A. DOUSE AND G.S. MITCHELL
breathing consists of two or more clustered breaths, separated by longer but highly variable apneas or non-ventilatory periods (cf Shelton et aL, 1986). Concomitant with episodic breathing are oscillations in lung and blood Po2 and PEG2. During a breath cluster, lung and blood Po: increase while Pco, decreases; the reverse occurs during a non-ventilatory period (Lenfant et al., 1970; Burggren and Shelton, 1979; Glass and Johansen, 1979). It has been postulated that the episodic breathing pattern results from these oscillations in Po: (Burggren and Shelton, 1979; Shelton and Croghan, 1988) or PeG, (West et al., 1989) acting via chemoreceptor feedback. According to this hypothesis, as the non-ventilatory period progresses, chemoreceptor drive progressively increases, eventually reaching a threshold that initiates breathing. Conversely, during a breathing bout, chemoreceptor drive diminishes until the ventilatory period terminates (Shelton and Croghan, 1988). We tested this hypothesis by monitoring the breathing pattern of alligators: (1) before and after bilateral cervical vagotomy, a procedure that removes peripheral arterial chemoreceptors (Benchetrit et al., 1977; Douse and Mitchell, 1992), CO2-sensitive intrapulmonary chemoreceptors and pulmonary stretch receptors (Powell et ai., 1988; Douse et al., 1989); (2)during ~midirectionai ventilation (UDV) at a high flow rate, thereby minimizing oscillations in lung and blood Pc>, and PEG,; and (3) during combined UDV and bilateral vagotomy. The results indicate that, although vagal receptors and oscillations in respiratory gases modulate quantitative aspects of the episodic breathing pattern, they are not essential to the genesis of episodic breathing.
Methods
Animals Except for the data recorded from unanesthetized, quiet alligators (n = 4; Series I), data described in this paper were collected during the same experiments as described in the preceding paper (Douse and Mitchell, 1992). Briefly, experiments were performed on 19 juvenile American alligators (,4. mississippiensis), ranging from 1.0 to 4.2 kg body mass (mean = 1.9 + 0.8 kg ( + SD)). The alligators were obtained from the Rockefeller Wildlife Reserve in Grand Chenier, LA. Animals were fed a diet of mice, with water readily available, and housed at 30 °C. Animals were fasted for > 1 week prior to experiments. Series i: Effects of anesthesia. In Series !, breathing movements were recorded from four unanesthetized alligators by inductive plethysmography (Respitrace Corp.). The alligators were placed in a dark chamber maintained at 30 °C, and at least 12 h were allowed to pass before data collection commenced. Measurements were made for I h periods, three times per day. Care was taken to remain quiet during measurements. The effects of the anesthetic protocol on ventilatory pattern were studied in three additional alligators at 30 °C without other experimental manipulations. General
EPISODIC BREATHING IN ALLIGATORS
79
anesthesia was induced with pentobarbital (approx. 25 mg/kg; i.p.). This anesthetic protocol resulted in a surgical level of anesthesia on the preparation day, and a stable but light anesthesia level on the two subsequent days of data collection. Further pentobarbital was administered as required to maintain stable anesthesia during data collection. The animals were studied in ventral recumbancy while body temperature (Tb) was monitored with a cloacal thermistor and regulated with a temperature controlled table, The femoral artery and vein were cannulated and the animals were intubated with an endotracheal tube. Series H: Effects of vagotomy. The effect of bilateral cervical vagotomy on breathing pattern was determined in six spontaneously breathing, sedated alligators. Breathing pattern was monitored during control conditions (room air) and with inhaled CO2 (5 % CO2 in 20-30% 02, balance N2). At least 30 min were allowed to achieve steady state at. each CO2 level; ventilatory measurements and an arterial blood sample were obtained during the subsequent 30 min. The vagi were bilaterally sectioned following application of local anesthetic (2% lidocaine), and at least 2 h were allowed to pass before the protocol was repeated. Vagal denervation was confirmed by the absence of response to intravenous injection of NaCN (50-100/~g/kg), a dose that elicits a brisk ventilatory response in intact alligators (Douse and Mitchell, 1992). Series 111: Unidirectionalventilation. The effects of unidirectional ventilation on breathing pattern were determined in six sedated alligators. Three heat-flared cannulae were inserted at the base and sides of each lung and secured with purse-string sutures. The body wall and skin were closed separately. Incisions were infiltrated with local anesthetic (2% lidocaine). The alligators were unidirectionallyventilated (UDV) by passing a humidified gas stream (4% CO2; 30-40% 02 in N2; >2 L/rain) through each cannula, into each lung and out via the endotracheal tube. Water temperature in the humidifier was maintained at the alligator's Tb, ensuring that lung temperature approximated Tb. The breathing pattern during U DV was observed at different levels of Flco~ in the UDV gas stream. At least three different COz levels were delivered to the lungs. The first and last levels were always FIco: = 0.04. The sequence of administration of the remaining levels was random and consisted of two or more of the following: Flee., = 0.02; 0.06; 0.08; and/or 0.09 (in 30-40% 02, balance N2). Different CO2 levels were used as necessary to elicit breathing movements in at least two conditions. At least 30 rain were allowed to pass at each Flco~ level to achieve steady state before ventilatio, i was measured (30 min) and an arterial blood sample was obtained. On the second day of Series Ill experiments, the vagi were sectioned and the protocol was repeated. Efficacy ofvagal denervation was assessed by lack ofventilatory response to NaCN (50-100/tg/kg; i.v.). In both Series II and IIl experiments, protocols were conducted at body temperatures of 30 and 20 ° C. Changes in body temperature took 30-60 rain.
80
M.A. DOUSE AND G.S, MITCHELL
Measurements Arterial blood pressure in the femoral artery and airway pressure in the endotracheal tube were monitored with pressure transducers (Statham, P23-id). Levels of CO2 and 02 in the gas stream were measured with an infrared CO 2 analyzer (Anarad, model pro-20) and an oxygen analyzer (Teledyne, model 33 IXB). Ventilation was monitored in the lightly anesthetized alligators with a pneumotachograph attached to the endotracheal tube (Flcisch, type 00). The pressure drop across the pncumotachograph was measured with a differential pressure transducer (Validyne, MP45) and integrated to give tidal volume (Gould). The continuous airflow during UDV in Series IIl (bias flow) was electronically offset; the pneumotachograph was calibrated during the apneic periods. The response of this system was linear over the ventilation range observed. Arterial blood was sampled by withdrawing 0.2-0.5 ml blood into a syringe and then withdrawing 0.3 ml blood anaerobically into a heparinized 0.5 ml glass syringe. The blood-saline mixture was then returned to the animal and the cannula was flushed with fresh heparinized saline. Analysis of arterial Pco2 and Po_, began within I rain of sampling using a micro gas analyzer (Radiometer, model BMS 3 MKII, PHM73) maintained at the animal's body temperature. Electrodes were calibrated before and after each measurement: the Po: electrode was calibrated with air-saturated water; the P¢o.~ electrode was calibrated with saturated gases provided by a gas-mixing device
(5.61 ~o and 11.22% CO2; Radiometer, GMA2) and a precision-calibrated gas tank (2.53% CO2). Accuracy of the calibration gases was confirmed with a WOsthoff gas mixing pump. A#aS;vis It is difficult to develop a formal and objective definition of episodic breathing due to the variable and labile breathing patterns observed. For our purposes, episodic breathing is functionally defined as consisting of two or more breaths in a breath cluster (breaths/cluster), separated by longer non-ventilatory periods. Intermittent breathing is also a useful term to describe breathing patterns in reptiles and amphibians, and connotes any breathing pattern where the breath hold between breaths is longer than the active ventilatory cycle (Milsom, 1990). Intermittent breathing is therefore a more general term, refering to episodic breathing, episodic breathing interspersed with single breaths, and single breaths separated by breath holds of longer duration. The distinction between a breath hold within a breath cluster and a non-vvntilatory period between breath clusters in episodic breathing is somewhat arbitrary (Milsom and Jones, 1980). Precise quantification is further confounded by the effects of experimental manipulations such as vagotomy and unidirectional ventilation on the duration of the ventilatory cycle, Therefore, we restricted attention to gross, qualitative changes in the overall breathing pattern, and did not attempt precise quantitation of ventilatory periods vs non-ventilatory periods (e.g. Naifeh etal., 1971a,b). Our interest was to assess qualitatively whether episodic breathing patterns (as defined above) persist after systematic removal of relevant sensory feedback mechanisms.
EPISODIC BREATHING IN ALLIGATORS
81
Results
Series I: Effects of anesthesia. Respiratory activity in unanesthetized and unrestrained alligators showed considerable variability. This variability was observed in the same alligator on the same day (Fig. IA), or between alligators (Fig. 1A and B). Typically the pattern consisted ofbreath clusters interspersed with inspiratory breath holds ofvarious durations. Three of the four alligators had a quiet, resting breathing pattern consisting of 2-7breaths/cluster. One alligator, however, consistently had episodes of 5-16 breaths/cluster. It was extremely easy to disturb the animals. Although animals were kept in a dark chamber separated from the rest of the room, disruption of the ongoing breathing pattern was often evident when experimenters quietly entered the opposite end of the room (Fig. IB). Over 30 breaths/cluster were observed in excited animals. Thus, unanesthetized alligators are extraordinarily sensitive to small disturbances resulting in large breathing pattern alterations. Anesthesia 3 days prior to data collection resulted in sedated, but spontaneously breathing alligators. In general, the anesthetic protocol decreased the number of breaths/cluster, and prolonged the respiratory cycle time (cf Figs. I and 2). There were also time-dependent changes in breathing pattern over the duration of an experiment. All three alligators maintained in constant conditions at 30 °C had an episodic breathing pattern by the end of the experiment (Fig. 2), consisting of 2-3 breaths/cluster interspersed with occasional single breaths. However, only one alligator had a clear episodic breathing pattern at the beginning of data collection (2 breaths/cluster; Fig. 2A). The other two animals exhibited intermittent breathing, which progressed to episodic breathing (2-3 breaths/cluster) by the end of data collection (Fig. 2B). Changes in
Be
All
mmmn i
W
=. .
.
.
cmomm .
," -
i~
!
t'W
Fig. !. Traces showing variability and sensitivity of the breathing pattern to a minor disturbance in two unanesthetized alligators at 30 °C (Series !). in A: top trace is breathing movements recorded from one alligator < 30 rain after placement of Respitrace; middle trace was 5 h and bottom trace was 9 h later. In B: top trace represents breathing movements in a different alligator 2 h after Respitraee placement; m~ddle trace = 5 h and bottom trace = 9 h later. At arrow, the experimenter quietly entered the room, approximately 8 m distant. Even though there was no obvious noise associated with this disturbance and the alligator was separated by 8 m and two visual or sound barriers, a disturbed ventilatory pattern was consistently evident. The ventilation monitor was not calibrated, and the traces provide only a qualitative index of respiratory amplitude.
82
M.A. DOUSE AND G.S. MITCHELL
A.
v,
B, otOOr
i|O0~
.8|
-|co
.
i.°oi
-----'u
v,
Vr
ili
-
-
Vt -80
-
-80 !
(;I) e I
Fig. 2. Effects of anesthetic protocol, tracheal intubation and time on breathing pattern in two spontaneously breathing alligators. In A, an episodic breathing pattern was evident throughout the experiment, in contrast, the alligator in B progressively developed a greater degree of episodic breathing. In both A and B, the top trace was recorded < 30 min after experiment began; middle trace was 9 h and bottom trace was 33 h later. Presence of the tracheal tube prevents an inspiratory breath hold. VT = tidal volume.
breathing pattern with time were not associated with significant changes in breathing frequency (f), tidal volume (VT), minute ventilation (Qi) or Pace,. Series H: Effects of mgotomy. The anesthetic protocol also tended to decrease the number of breaths/cluster in Series II compared to unanesthetized animals, although this effect was less marked than in Series I (Fig. 3). Five of six intact alligators breathing room air had an episodic breathing pattern consisting of 2-4 breaths/cluster interspersed with occasional single breaths (Fig. 3). One alligator had a stable intermittent pattern of single breaths. Increasing inhaled CO2 to 5% increased VT, f and V]. Increased f was due to decreases in the non-ventilatory period and an increase in the number of breaths/cluster. In four alligators, 3-8 breaths/cluster were observed while two animals had between 12 and 30 breaths/cluster (Fig. 3). Altar bilateral cervical vagotomy, intermittent but not episodic breathing was observed in five of six alligators breathing room air (Fig. 4A). The remaining alligator had a breathing pattern that is difficult to characterize (Fig. 4B). Increasing inhaled CO2 to 5?/o post-vagotomy reverted four of six alligators to an episodic breathing pattern A.
e.
VT
i,
J,
i
I
I
{|
VT
i i
'°° -tOO
-II
II
i J l
i, L ,ll,tll q--
! 5 mm I Fig. 3, CO~ effects on the episodic breathing pattern of two intact, spontaneously breathing alligators (Series !!), In both A and B, the top trace is breathing 0% CO2 and the bottom trace is breathing 5% CO.~.
83
EPISODIC BREATHING IN ALLIGATORS
B.
AI
I s=,~l Fig. 4. Effects of bilateral cervical vagotomy on the episodic breathing pattern in two spontaneously breathing alligators (Series If). Same alligators as in Fig. 3A and B. In both A and B, the top trace is breathing 0% CO, and the bottom trace is 5 ~ CO2.
(Fig. 4). The episodic breathing pattern usually consisted of two and sometimes three breaths/cluster interspersed with occasional single breaths. Two alligators exhibited intermittent single breaths, regardless of CO2 level. There were no significant effects of CO2 on mean f, VT or ~'I post-vagotomy. Episodic breathing was evident in all six alligators during unidirectional ventilation (Figs. 5 and 6). Typically, the breathing pattern consisted of 2-3 breaths/cluster, separated by non-ventilatory periods of several minutes duration (Fig. 5). The pattern of breathing, however, was extremely variable both within and between alligators (Fig. 6). Increasing CO2 levels in the UDV gas stream increased mean f, VT and VI. The increase in fwas through a decrease in the non-ventilatory period and, in three alligators, through an increase in the number of breaths/cluster. At low CO2 levels (2-4 ~o), three alligators did not breathe. At mid CO2 levels (4-6 ~), the bre:'~thing pattern consisted of 2-4 breaths/cluster, interspersed with occasional single breaths in Series 111: Unidirectional ventilation.
VT
Po,
*|001 *4
.o ~ - - - ~ t - ~
. . . . . . . .
:
0.10r Flco2 I 0.06
0 u.
Fig. 5. Episodic breathing despite unidirectional ventilation, minimizing oscillations in respiratory gases in a sedated alligator (Series Ill). FIco2 =0.04 and Flo2 =0.35 in the unidirectional gas stream {Paco~ = 36 mmHg and Pao2 = 205 mmHg). Paw = airway pressure.
84
M.A. DOUSE AND G.S. MITCHELL
All
B. VT
':I
A
-IIOIIt.
C, VT "a°°I
.nl v,.
D. ,,
Fig, 6. Variability of episodic breathing patterns during unidirectional ventilation in four sedated alligators (Series III). In A, FIco2 -- 0.04 (Patio, = 36 mmHg; Pao, = 205 mmHg), in B, Fuco, = 0.08 (Paco~ = 63 mmHg; Pao, = 198mmHg). In C, Flco2 =0.04 (Paco,= 44mmHg; Pao, = 207mmHg). In D, Flco, = 0.08 (Paco, = 69 mmHg; Pao: = 134mmHg).
four alligators (Fig. 6A and C). The two remaining alligators had 6-8 breaths/cluster. At high COa 10vels(6-10%), tht'¢."alligators had 2-4 breaths/cluster (Fig. 6B), two had 6-12 breaths/cluster (Fig. 6D), and one alligator went from an episodic breathing pattern to a virtually continuous breathing pattern with few, if any, non-ventilatory periods. After bilateral vagotomy, all six alligators could still manifest an episodic breathing pattern in at least one condition of CO2 during UDV (Fig. 7). This pattern typically consisted of 2-3 breaths/cluster separmed by variable periods of apnea and interspersed with occasional single breaths. Although all alligators could manifest an episodic breathing pattern during UDV after vagotomy, this pattern was very labile and could change to single breaths despite apparently constant conditions (Figs. 7C and 8A). Increasing the CO 2 level in the UDV gas stream increased mean f, VT, and VI. At low CO 2 levels (4'~o), five of six alligators did not breathe. At higher CO 2 levels (6-9%), four of six alligators had a breathing pattern consisting of two breaths/cluster interspersed with occasional single breaths (Fig. 7A and C). The remaining two alligators had an episodic breathing pattern of 2-5 breaths/cluster (Fig. 7B and D).
Effects of temperature. Episodic breathing was evident at both 30 and 20 °C in all experimental series (Fig. 8). The results were generally similar at both temperatures,
EPISODIC BREATHING IN ALLIGATORS
85
A.
LLIt
J.
I
-20
BII
~200~ VT
11
e
-2on
11
o -800
l l
4t tt
t¸
CI v,
1111II
Dill
,t200F
w
o
Ill
~[1
LL
.,.oE
llr
lJl
II
Fig, 7. Effects 0fbilaterai vagotomy and unidirectionalventilation on the episodic breathing pattern in four sedated alligators (Series II1). These are the same alligators as in Fig. 6. In A, Flee2 = 0.06 (Pace2 ffi 50mmHg; Pao2 = 158mmHg). In B, FIco~ ffi 0.08 (Pace2 ffi 59mmHg; Pao~-216mmHg). In C, Free: ffi0.08 (Pace2 = 7S mmHg; Pao2 ffi 110 mmHg). In D, Flee2 = 0.09 (Paco: = 70 mmHg; • Pao, ffi 88 mmHg).
although the pattern was slightly different at 20 ° C. Often there were longer breath hold periods within a breath cluster and longer non-ventilatory periods between clusters at 20 °C than at 30 °C, even in the same animal (Fig. 8). Both of these factors increased the difficulty of characterizing the breathing pattern as episodic v s intermittent. In general, there was a decrease in the number of breaths/cluster at 20 °C in any given condition (range = 2-7 breaths/cluster).
86
M.A. DOUSE AND G.S. MITCHELL
A. 30 °C
"2001
ill II j J
VT
B. 20 "C
'~200r... ] VT
,
i
Fig. 8. Breathing pattern recorded from the same alligator as in Fig. ?A after bilateral vagotomy and during unidirectional ventilation at body temperatures of 30 °C (A) and 20 °C (B) (Series Ill). In both A and B, Flee= = 0.06 and Fie, = 0.3. In A, Pace= = 47 mmHg and Pao2 -- 178 mmHg. In B, Pace2 = 50 mmHg and Pao, = 158 mmHg.
Diseussion
The fundamental question addressed in this paper is: Can an episodic breathing pattern be observed after removing important sources of (vagal) chemoreceptor feedback or with minimal oscillations in respiratory gases? The results suggest that: (1)episodic breathing does not require feedback from vagal sensory receptors (arterial chemoreceptors, CO2-sensitive intrapulmonary receptors and pulmonary stretch receptors) and (2)steady-state CO 2 levels affect episodic breathing patterns quantitatively, but oscillations in respiratory gases do not appear to be necessary for the expression of episodic breathing.
Critique ofmethods. A formal, quantitative analysis of the episodic breathing pattern was not attempted due to the large variation within and between alligators, the labile nature of the episodic breathin~ pattern and definitional concerns (see Analysis). Instead, our approach was to qaalitatively assess the effects of bilateral vagotomy, unidirectional ventilation, anesthesia, C02 and temperature on the episodic breathing pattern. The episodic breathing pattern observed after anesthesia, vagotomy or UDV, for example, was in many respects different from that observed in intact, unanesthetized alligators. However, our effort was to test hypotheses concerning the role of sensory feedback in the initiation and control of episodic breathing in a qualitative sense. Our
EPISODIC BREATHING IN ALLIGATORS
87
objective was to elucidate the minimal mechanism(s) necessary for expression of an episodic breathing pattern in alligators. Arterial Pco2 was not continuously monitored during UDV. The crocodilian lung has a multicameral architecture (Perry, 1988) that could allow air flow to stream through central bronchi without thoroughly ventilating some chambers. Thus, breathing efforts may be necessary to ventilate these areas, potentially changing regional lung and blood gas composition with each breath. Since 4% CO 2 in the UDV gas stream was used to establish normal blood gas values, the high inspired CO2 would minimize Pace2 oscillations despite possible changes in regional lung ventilation with each breathing effort. Indeed, serial samples indicated that Pace, varied less than 2 mmHg. Given the low ventilatory sensitivity to Pace, in these alligators (Douse and Mitchell, 1992), such changes in Pace2 would not have a noticeable effect on ventilation (see below), Further elevations in inspired CO2 would minimize CO2 oscillations even more. Nevertheless, elevated CO, levels increased rather than decreased the incidence of episodic breathing. Arterial Po2 measurements in Series III indicated that ventilation-perfusion mismatch and/or cardiopulmonary shunting may have occasionally been present in some alligators (see Fig. 7D). Changes in cardiopulmonary shunt during UDV could conceivably lead to oscillations in blood gases, resulting in an episodic breathing pattern. However, it is unlikely that changes in cardiopulmonary shunt would occur over the same, short time scale necessary to initiate and terminate an episode of breathing in a breath cluster. Moreover, most alligators could manifest an episodic breathing pattern despite high background levels of Pace,, when ventilation-perfusion mismatch or cardiopulmonary shunt was not indicated.
Series i: Effects of anesthesia. Unanesthetized alligators in conditions of minimal environmental stimulation have an episodic breathing pattern typically consisting of 2-3 breaths per cluster (Huggins et aL, 1969). A pattern of 2-7 breaths/cluster was observed in the unanesthetized alligators in this study. However, an increase in the number of breaths per cluster was observed in excited animals (see Fig. 1), emphasizing the considerable variability and labile nature of the episodic breathing pattern. Light levels of anesthesia were used in order to reduce discomfort to the animal and to maintain a constant 'state'. Under conditions of light anesthesia, breathing pattern closely mimicked the pattern recorded from unanesthetized alligators after prolonged periods with minimal environmental stimulation (this study, Fig. 1; Huggins etal., 1969). The anesthetic protocol did tend to decrease the number of breaths per cluster and prolong the breath hold period within a breath cluster, as has been noted previously in crocodilians (Naifeh et aL, 1971a). Collectively, the results suggest that both light anesthesia and arousal state affect breathing pattern in alligators in a quantitative rather than qualitative manner. Series H: Effects of vagotomy. An episodic breathing pattern was not generally observed during spontaneous breathing after bilateral vagotomy, except when respiratory drive was increased with inhaled CO2. This may suggest that peripheral arterial
88
M.A. DOUSE AND G.S. MITCHELL
chemoreceptors and/or the pulmonary receptors have a role in the manifestation of episodic breathing in alligators. However, these sensory pathways are not necessary to generate episodic breathing as demonstrated by the return of episodic breathing at elevated CO2. The effect of CO 2 on breathing pattern following vagotomy contrasts with the nearly complete aboiition of the ventilatory response to hypercapnia in the same experimental conditions (Douse and Mitchell, 1992). In a previous study, bilateral vagotomy did not abolish episodic breathing and only slightly decreased the number of breaths per breath cluster in unanesthetized alligators and caimans (Naifeh et al., 197 lb). This difference in results may be due to the residual anesthetic effects in our study, decreasing respiratory drive and suppressing episodic breathing. Although not specifically commented on by the authors, an episodic breathing pattern may also be observed in turtles after bilateral vagotomy (Frankel et a!., 1969; Milsom and Jones, 1980). There were still oscillations in Pace_, after vagotomy in spontaneously breathing alligators, which may have affected non-vagal CO2-sensitive areas, leading to an episodic breathing pattern. Although increasing inhaled CO2 would decrease Paco~ oscillations, increasing inspired CO2 increased rather than decreased the incidence of episodic breathing (Fig. 3). Episodic breathing was still evident despite unidirectional ventilation. Since oscillations in lung and blood CO2 and 02 were minimized by UDV, oscillations in chemoreceptor feedback from any source is unlikely to have initiated episodic breathing. Episodic breathing has also been observed recently in intact toads during unidirectional ventilation (West et ai., 1987). We also observed episodic breathing during U DV after bilateral vagotomy, but the pattern was very labile and could change to single breath intermittent breathing despite apparently constant conditions and with no obvious state transitions. Nevertheless, qualitatively similar episodic breathing patterns were observed in this study despite vagotomy, unidirectional ventilation, unidirectional ventilation with vagotomy and changes in body temperature. Models of mammalian periodic breathing may be useful in elucidating the mechanism(s) initiating episodic breathing in reptiles. These models suggest that respiratory in stability and periodic or episodic breathing may arise from: (1) increases in circulatory dela~ from lungs to chemoreceptors; (2) increased chemoreceptor sensitivity; and (3) changes in a chemical threshold for ventilatory acti~,ity (i.e. apneic threshold; Khoo eta l., 1982; Cherniack and Longobardo, 1986). A fundamental assumption in all models of unstable breathing is that the system is operating as a closed chemoreceptor feedback loop (el Cherniack and Longobardo, 1986). Further, all of these models rely on Pace, or Pao, oscillations either to initiate or to maintain respiratory instability (Khoo etaL, 19821 Cherniack and Longobardo, 1986). in our study, UDV opens chemoreceptor feedback loops and minimizes changes in Pace, or Pao:. In addition, episodic breathing was observed after vagotomy despite decreases in both 02 and CO2 ventilatory sensitivity and at Pace, levels well above the apneic threshold (Douse and Mitchell, 1992). Thus, even if there are small changes in CO2/H + in the microSeries Ill: Unidirectional ventilation.
EPISODIC BREATHING IN ALLIGATORS
89
environment of central chemosensitive areas during UDV (post-vagotomy), it is unlikely that such changes would lead to respiratory instability and episodic breathing since the response to large changes in Paco: during UDV after vagotomy is sluggish and relatively small (Douse and Mitchell, 1992). Collectively, the results suggest that episodic breathing in alligators does not require vagal sensory feedback from peripheral arterial chemoreceptors, CO2-sensitive intrapulmonary chemoreceptors, or pulmonary stretch receptors. Further, it appears that tonic stimulation of both vagal and non-vagal CO2-sensitive chemoreceptors modulates, but does not initiate the ongoing rhythm. Finally, the episodic breathing pattern does not require oscillations in respiratory gases or oscillations in afferent vagal feedback and, therefore, may be inherent to brainstem mechanisms producing the respiratory motor output. Although the experimental manipulations used in this study have not eliminated all possible sources of oscillatory sensory feedback (e.g. chest wall receptors), the results strongly suggest that the essential mechanisms generating the episodic pattern are intrinsic to the central neural networks associated with ventilatory control. Indeed, strong support for this hypothesis has been provided by our recent observation that an in vitro brainstem-spinal cord preparation from adult turtles is capable of producing an episodic motor output similar to breathing in the whole animal, without peripheral inputs and despite constant CO, and 02 conditions (Douse and Mitchell, 1990). Thus, the episodic breathing pattern in reptiles appears to be inherent to brainstem mechanisms generating respiratory rhythm. This does not rule out a role for oscillatory afferent feedback in modulating the ongoing episodic pattern. In intact, unanesthetized alligators, 3 or more breaths per cluster are characteristic of episodic breathing (Huggins et al., 1969; Naifeh et al., 197 la,b; this study). However, 2-3 breaths per cluster is more commonly observed after mild anesthesia, bilateral vagotomy (Naifeh et al., 197 la,b; this study), or unidirectional ventilation (this study), suggesting that vagal feedback and state have roles in modulating quantitative details of episodic breathing in reptiles. Acknowledgements. We are grateful to Ted Joanen and Larry McNease of the Rockefeller Wildlife Refuge
for their help in procurement of alligators and to Dr. W.K. Milsom for loan of the Respitrace. (3.S. M. was supported by a Research Career Development Award from NIH (HL01494).
References
Benchetrit, G., J. Armand and P. Dejours (1977). Ventilatory chemoreflex drive in the tortoise, Testudo horsJleldi. Respir. Physiol. 3I: 183-191. Burggren, W.W. and (3. Shelton (1979). Gas exchange and transport during intermittent breathing in
chelonian reptiles. J. Exp. Biol. 82: 75-92. Cherniack, N.S. and (3.S. Longobardo (1986). Abnormalities in respiratory rhythm. In: Handbook of Physi~logy, Section3: The Respiratory System, Vol.ll; edited by N.S. Cherniack and J.(3. Widdicombe. Bethesda, MD: American Physiology Society, pp. 729-749. Douse, M.A., F.L. Powell, W.K. Milsom and (3. S. Mitchell (1989). Temperature effects on pulmonary receptor responsesto airwaypressure and CO2 in Alligator mississippiensis. Respir. Physiol. 78:331-344.
90
M.A. DOUSE AND G.S. MITCHELL
Douse, M. A. and G. S. Mitchell (1990). Episodic respiratory related discharge in turtle cranial motoneurons: In vitro and in rive studies. Brain Res. 536: 297-300. Douse, M.A. and G. S. Mitchell (1992). Effects of vagotomy on ventilatory responses to CO2 in alligators. Respir. Physiol. 87: 63-76. Frankel, H. M., A. Spitzer, J. Blaine and E. P. Schoener (1969). Respiratory response of turtles (Pseudemys scripta) to changes in arterial blood gas composition. Comp. Biochem. Physiol. 31: 535-546. Glass, M. L. and K. Johansen (1979). Periodic breathing in the crocodile, Crocodylus niloticus: Consequences for the gas exchange ratio and control of breathing. J. Exp. Zool. 208: 319-326. Huggins, S.E., H.E. Hoff and R.V. Pena (1969). Heart and respiratory rates in crocodilian reptiles under conditions of minimal stimulation. Physiol. Zool. 42: 320-333. Khoo, M.C., R.E. Kronauer, K.P. Strohl and A.C. Slutsky (1982). Factors inducing periodic breathing in humans: A general model. J. Appl. Physiol. 53: 644-659. Lenfant, C., K. Johansen, J.A. Petersen and K. Schmidt-Nielsen (1970). Respiration in the fresh water turtle, Chelys fimbriata. Respir. Physiol. 8: 261-275. Milsom, W.K. and D.R. Jones (1980). The role of vagal afferent information and hypercapnia in control of the breathing pattern in Chelonia. 3. Exp. Biol. 87: 53-63. Miisom, W. K. (1990). Mechanoreceptor modulation ofendogenous respiratory rhythms in vertebrates. Am. J. Physiol, 259: R898-R910. Naifeh, K.H., [i.E. Huggins and H.E. Heft (1971a). Study of the control of crocodilian respiration by anesthetic dissection. Respir. Physiol. ! 2:251-260. Naifeh, K. H., [i. E. Huggins and H.E. Heft (1971b). Effects of brain stem section on respiratory patterns of crocodilian respiration. Respir. Physiol. 13: 186-197. Perry, [i.F. (1988). Functional morphology of the lungs of the nile crocodile, Crocodylus niloticus: nonrespiratory parameters. J. Exp. Biol. 134: 99-117, Powell, F. L., W, K, Milsom and G.S. Mitchell (1988). Effects of intrapulmonary CO2 and airway pressure on pulmonary regal afferent activity in the alligator. Respir. Physiol. 74: 285-298. [ihelton, O., D,R. Jones and W,K, MUsom (1986), Control of breathing in ectothermic wrtebrates. In: Handbook of Physiology, [iection 3: The Respiratory System, Vol. 2; edited by A.P. Fishman, N. $. Cherniack, J.O. Widdieombe and [i. R. Geiger. Washington, DC: American Physiological Society, pp. 857-909. Shelton, G, and P,C. Croghan (1988), Gas exchange and its control in non-steady-state systems: The consequences of evolution from water to air breathing in the vertebrates, Can. J, Zool. 66: 109-123, West, N. H., Z, L, Toper and ii, N, Van Viler 0987), Hypoxemie threshold for lung ventilation in the toad. Resplr, Physlol, 70: 377-390. West, N.H., A.W. Stairs and W,W, Burggren (1989~ Factors terminating nonventilatory periods in the turtle, Chelydra serpenthea, Resplr, Physiol. 7"/: 337-350.