Localization of Baroreceptors and Gain of the Baroreceptor-Heart Rate Reflex in the Lizard Trachydosaurus Rugosus

The existence of baroreceptors and baroreflexes in reptiles has been demonstrated by the observation that blood pressure is restored towards normal after being disturbed experimentally in iguanas (Hohnke, 1975) and snakes (Seymour & Lillywhite, 1976; Lillywhite & Seymour, 1978). On the efferent side of the baroreflex, many studies have shown that the heart and blood vessels receive an autonomic innervation which is capable of changing vascular resistance and the rate and force of beating of the heart (Berger & Burnstock, 1979). The afferent and central components, by contrast, have received little attention, so that nothing is known about the sensitivity of the reptilian baroreflex relative to that of mammals. In addition to providing this information, the present investigation sought to localize the baroreceptors of the lizard Trachydosaurus rugosus.

The method selected was to produce changes in arterial pressure by inflation of perivascular cuffs. With this technique the gain of the cardiac component of the baroreflex can be calculated just as it is in mammals (Korner et al. 1972). The technique can also be used to determine the location of the baroreceptors. Inflation of cuffs placed around central arteries reveals, by its effect upon heart rate, whether the baroreceptors lie proximal or distal to the cuffs. Cuffs were strategically placed to test whether baroreceptors lie in any of three different locations chosen on the basis of homology and published reports on baroreceptor nerves in reptiles.

The mammalian carotid body–carotid sinus is homologous with that region in reptiles where the embryonic common carotid artery divides into the internal and external branches (Adams, 1958, 1962). This region was the first to be tested. The second site to be tested was the region in reptiles homologous with the aortic arch baroreceptive region of mammals. In mammals, Nonidez (1935a, 1937) reported that these receptors lie on the persistent aortic arch, near the position of the embryonic ductus arteriosus. The ductus arteriosus or its remnant lies on each lateral aorta in reptiles at quite a distance from the truncus arteriosus (O’Donoghue, 1917, 1920; Burggren, 1976). The third possible baroreceptive region examined was the truncus arteriosus, which is innervated in reptiles by branches of the vagus nerve (Gaskell & Gadow, 1884; Adams, 1962). From previous work it is not clear whether the branches contain baroreceptor afferent fibres. Fedele (1937) observed a consistent depressor response on central stimulation of the truncal nerves, whereas other workers did not (Kazem-Beck, 1885, 1888; Mills, 1885).

Experiments were performed on a total of 14 lizards of either sex weighing 500-730 g. The surgical implantation of inflatable cuffs and blood pressure cannulae was carried out under barbiturate anaesthesia (Nembutal: Abbott Laboratories). The dose of anaesthesia required to prevent a righting reflex, limb withdrawal reflex or arching of the hindlimbs in response to prodding of the vent, ranged from 20–45 mg/kg (i.p.) and was independent of body mass. An initial dose of 20 mg/kg was supplemented after 30 min by a further dose of 10 mg/kg if required. During abdominal surgery, and usually for a short time afterwards, animals were ventilated with a small animal respirator. All experiments were carried out during spring and summer, apart from one which was performed in autumn.

The aorta and posterior vena cava were exposed by a mid-ventral abdominal incision. A cuff was placed around the posterior vena cava just rostral to the liver and its inflation tube was led out through a stab wound in the lateral abdominal wall. A second cuff was installed around the aorta just rostral to the coeliaco-mesenteric artery. About 1 cm rostral to the aortic cuff a non-occlusive polyethylene cannula (i.d. 0·50 mm; o.d. 0·90 mm) was inserted through a very fine puncture in the aorta and advanced a few cm towards the heart. The cannula usually had a closed, angled tip with one or two apertures slightly behind the tip.

A mid-ventral incision was begun at the level of the forelimbs and continued forwards for a distance of about 3 cm. The anterior portion of the pericardium was exposed and the right and left common carotid arteries were separated from the closely adherent lateral aortae at their origin from the truncus. A cuff was placed around each carotid artery in this position (A, Fig. 1). Each ductus caroticus was approached through a lateral incision in the neck and tied firmly. In an attempt to avoid damaging any innervation of the carotid trifurcation the ties were made at the junction of the duct and the lateral aorta. To monitor the effect of cuff inflation the ductus caroticus of one side was cannulated towards the trifurcation.

After exposing the anterior pericardium the lateral aorta of each side was gently separated from the common carotid artery and the pulmonary artery close to the truncus. A cuff was placed around each lateral aorta (B, Fig. 1) and each ductus caroticus was ligated close to the carotid trifurcation. In this way the possibility of damage to any innervation of the lateral aortae was minimized.

The pericardium was opened in the midline and a large cuff was placed around the truncus arteriosus as distal as possible to the ventricle (C, Fig. 1). In one animal a cannula was inserted through a fine stab wound into the carotico-aortic trunk (see Fig. 1). The cannula was advanced towards the ventricle until its tip lay proximal to the cuff and therefore monitored the effect of cuff inflation on intra-truncal pressure.

During an experiment the surgically-prepared animal was placed in a narrow darkened box. Most animals remained quite still for many hours. Room temperature was maintained at about 25 °C. Pressures were measured by connecting the cannulae filled with heparinized saline (100 i.u./ml) to Statham P23DI pressure transducers calibrated against a static pressure head. Heart rate was obtained by using an electro-cardiographic signal as the trigger for a Grass Model 7P44 tachograph. All records were displayed on a Grass Model 79D polygraph. Drugs used were hyoscine hydrobromide (David Bull), atropine sulphate (Sigma) and bretylium tosylate (Burroughs-Wellcome). Each drug was administered by intra-arterial injection over a period of 2–3 min.

In all 14 animals studied, tachycardia and hypertension persisted for several hours following anaesthesia. Moreover, maximal cardiac responses to cuff inflation were usually not obtained until 24 h after anaesthesia. Thus all results presented here were collected at least 24 h after surgery.

Resting heart rate in conscious lizards ranged from 20–30 beats/min and resting systemic arterial pressure was 6·0 kPa systolic (range 5·1–6·6 kPa) over 4·4 kPa diastolic (range 3·9–4·9 kPa).

In all animals studied, inflation of the aortic cuff caused a rise in systemic arterial pressure and a prompt bradycardia which usually reached its maximum within 10 s (Fig. 2 A). Inflation of the posterior vena caval cuff invariably reduced systemic arterial pressure and increased heart rate (Fig. 2B). The time required to reach the maximum increase in heart rate ranged from 90–200 s. Recovery of heart rate to resting levels required up to 10 min, with recovery from tachycardia taking longer than from bradycardia. The mean gain of the baroreceptor-heart rate reflex in 6 lizards was 1370 ms/kPa (see Table 1).

Injection of atropine (2·9 mmol/kg; N = 1) or hyoscine (2·3 mmol/kg; N =2) caused an immediate increase in heart rate of 10% (27–30 beats/min), 25% (20–25 beats/min) and 60% (22–35 beats/min) in the three animals tested. Muscarinic blockade also abolished the often very pronounced beat-to-beat variation in resting heart rate. After either drug, aortic cuff inflation usually still caused a slight reduction of heart rate although never to the level attained during a control response. Posterior vena caval cuff inflation caused heart rate to rise to approximately the same level before and after atropine or hyoscine. In the two animals in which all the necessary measurements were made, muscarinic blockage caused a moderate reduction of the gain of the baroreflex (Fig. 3A; Table 1).

Injection of bretylium (2·4 mmol/kg) caused an immediate increase in blood pressure and heart rate in the three animals tested. After approximately 3 h, blood pressure returned to the control level, while heart rate was still somewhat elevated. Beat-to-beat fluctuations in heart rate persisted after bretylium treatment (Figs. 4A, B), and cardiac responses were still obtained to inflation of the cuffs around the aorta (Fig. 4A) and posterior vena cava (Fig. 4B): but the gain of the reflex was greatly reduced (Fig. 3B; Table 1). In all three animals treated with bretylium, subsequent injection of hysocine (2·3 mmol/kg) elevated heart rate to a level in excess of that achieved during a post-bretylium venous cuff inflation (Fig. 4C). In addition, after treatment with bretylium plus hysocine only very small responses were obtained on inflation of either cuff, and beat-to-beat fluctuations in heart rate were abolished (Figs. 4D, E). The early tachycardia to aortic cuff inflation in Figure 4E is reminiscent of the Bainbridge reflex, and a similar response has been reported after parasympathetic blockade, or after a combination of parasympathetic and sympathetic blockade, in the dog (Glick & Braunwald, 1965).

Two animals were prepared with a cuff around each common carotid artery; in one animal the resting heart rate was within the range quoted above, whereas the second animal had a resting heart rate substantially above that range (about 40 beats/min). In both animals, simultaneous inflation of the cuffs around the right and left common carotid arteries caused a fall in pressure in the ductus caroticus, while having no persistent effect on the pressure measured in the aorta (Fig. 5 A). This procedure had no effect upon heart rate, even though a posterior vena caval cuff inflation produced a substantial tachycardia in the same animals (Fig. 5 B).

Inflation of cuffs around the lateral aortae in two animals caused a fall in aortic pressure and a bradycardia that was similar to that observed when the aortic cuff was inflated. A fall in central arterial pressure induced by vena caval cuff inflation in these animals led to the expected tachycardia.

Inflation of the cuff around the truncus arteriosus caused a marked elevation of systolic pressure measured in the truncus, while diastolic pressure either remained at the control level or fell substantially (Fig. 6). It is difficult to explain the fall sometimes observed in diastolic pressure, but it is possible that the tip of the cannula sometimes interfered with normal closure of the ventricular valves. In any case, a bradycardia invariably occurred during the period of truncal cuff inflation in 21 tests in 2 animals (e.g. Fig. 6). In one animal, heart rate always returned to the control level on deflation of the cuff. In the second animal studied, heart rate either returned to a level close to control (Fig. 6 A), or underwent a lengthy overshoot. This secondary tachy-cardia was always associated with struggling body movements which occurred either near the end of the period of cuff inflation (Fig. 6B) or soon after deflation of the cuff (Fig. 6C). Struggling in this animal always resulted in a substantial and lenghty tachycardia.

Changes in arterial pressure in Trachydosaurus rugosus elicited cardiac responses similar to those observed in mammals (Korner et al. 1972), indicating the existence of pressure-sensitive receptors. Since lizards have an undivided ventricle an increase or decrease in central systemic arterial pressure is probably accompanied by like changes in central pulmonary arterial pressure. Thus, the observation of heart-rate responses to posterior vena caval cuff inflation, and to aortic cuff inflation, could be explained by the presence of baroreceptors in the systemic or pulmonary arterial circuit, or both. It is also possible that the reflexes were mediated by receptors in the heart or central veins. A further point to be borne in mind is that, while cuff inflation experiments can be used to localize the receptor site which has the most powerful influence over heart rate, they cannot eliminate the possibility that subsidiary receptor sites exist else where.

It was shown conclusively by Hering (1923, 1924) that baroreflexes in mammals originate from the region where the common carotid artery divides. The homologous region in lizards with a ductus caroticus, such as T. rugosus, is the area of the carotid trifurcation (Adams, 1958). Inflation of cuffs on each common carotid artery, while producing a fall in pressure in the ductus caroticus, had no effect on heart rate (Fig. 5). Thus, the carotid ‘sinus’ is not the location of the baroreceptors in T. rugosus.

The aortic arch baroreceptors of mammals and birds are reported to lie near where the ductus arteriosus joins the surviving aorta (Nonidez, 1935a, b; Boyd, 1941). However, inflation of a cuff proximal to the homologous region in T. rugosus produced a bradycardia, indicating a pressure rise at the baroreceptors. The arterial regions in which pressure could be expected to rise during lateral aortic occlusion are the carotid arteries, already ruled out as a site of baroreceptors, the truncus arteriosus, and the pulmonary arteries. In view of the bradycardia elicited by inflation of a cuff around the distal truncus arteriosus, the pulmonary arteries can also be ruled out. The evidence therefore points to a truncal location, or to a cardiac or central venous location for baroreceptors in T. rugosus.

The possibility that the baroreceptors of T. rugosus lie in the ventricle, auricles, or central veins can only be assessed on indirect grounds. The atrial mechanoreceptors of mammals produce tachycardia when pressure rises (Ledsome & Linden, 1967), rather than the bradycardia observed in the present study. In addition, the cardiac baroreflex is likely to be involved in the regulation of arterial blood pressure, just as its functional homologue appears to be in mammals. Thus an arterial location for baroreceptors would provide more precise information on arterial pressure and its fluctuations than would receptors in sites isolated from the arteries. Finally, although venous, atrial, and ventricular mechanoreceptors exist in mammals, virtually the entire baroreflex cardiac response is mediated by receptors lying in the arteries (Korner et al. 1973).

A truncal location for baroreceptors is supported by anatomical evidence that nerves similar to the mammalian aortic arch depressor nerves run to the truncus from the vagus in reptiles (Gaskell & Gadow, 1884; Mills, 1885; Kazem-Beck, 1888; Fedele, 1937; Boyd, 1942; Khalil & Malek, 1952; Adams, 1962). Moreover, while there are physiological reports that the truncal nerves of reptiles do not mediate cardiovascular responses (Mills, 1885; Kazem-Beck, 1888), Fedele (1937) showed convincingly that in several reptilian species the truncal nerves contain afferent fibres capable of mediating reflexes on the heart. In an extremely detailed account Fedele traced the truncal nerves to their terminations and stated that the afferent fibres did not reach the ventricle. In view of the arguments outlined above it seems reasonable to conclude that the baroreceptors of T. rugosus are located in the truncus arteriosus.

It is somewhat surprising that the homologue of the mammalian carotid sinus has no baroreceptor role in T. rugosus. This is so, in particular, because the region is innervated in T. rugosus, and in other lizards, by vagal branches and possibly by glossopharyngeal fibres (Terni, 1931; Adams, 1958; Chowdhary, 1950; Oelrich, 1956; Mehra, 1958; Rogers, 1967). And yet the present evidence is clear. It should be remembered that only in mammals is there convincing physiological evidence of a baroreceptive role for the region where the common carotid artery divides; perhaps the carotid sinus baroreceptor region developed de novo in mammals, or in their synapsid reptilian precursors. The present work failed to find evidence for baroreceptors in the lateral aortae of T. rugosus; thus baroreceptors appear to be absent from the position directly homologous with the aortic arch baroreceptor region of birds (Nonidez, 1935b; though see Jones, 1973) and mammals (Nonidez, 1935a, 1937; Boyd, 1941).

The gain of the cardiac component of the baroreflex of T. rugosus is very high when compared with that in the rabbit, cf. a mean value of 1300 ms/kPa (100 ms/mm Hg) v. 5·2±1·0 ms/mm Hg (Korner et al. 1972). Trachydosaurus rugosus has approximately 1/10th the resting heart rate of a rabbit so that given the hyperbolic relationship between heart rate and heart period, it is unreasonable to compare the gain value of Table 1 with that reported by Korner et al. (1972). However, if gain is defined in percentage terms then differences in resting heart rate play no role in the value obtained. Accordingly, experimentally-induced heart rates were expressed as a percentage increase over the minimum heart rate that could be elicited by aortic cuff inflation. Gain was then recalculated as the percentage change in heart rate per unit change in mean arterial pressure. The revised mean gain for the six lizards comprising this study becomes 98%/kPa or 7·2%/mmHg. Similar gain values, dervied from previously published reports on conscious mammals, are comparable to that of T. rugosus (see Table 2).

The physiological significance of gain depends upon how heart rate relates to cardiac output in mammals and birds, and, in lower vertebrates with an undivided ventricle, to total heart output (see Johansen, 1972). Stroke volume is extremely labile in fish and it seems to be the prime determinant of the increased output of the heart during exercise (Jones & Randall, 1979). Little is known of the situation in amphibians,

reptiles or birds. Even in mammals there is no general agreement on the relationship between heart rate and cardiac output. However, recent work suggests that, in unanaesthetized, undisturbed dogs, changes in cardiac output are brought about mostly by changes in heart rate (Vatner & Boettcher, 1978; Boettcher et al. 1978; Anderson, Yingling & Sagawa, 1979). If this finding is true for all tetrapods, then the baroreflex cardiac gain would provide a reasonable index of the comparative ability of animals to increase or decrease the output of the heart in response to arterial pressure disturbances. Concerning the comparative ability of animals to regulate blood pressure, a more useful gain to measure is that relating changes in central arterial pressure and changes in pressure at the isolated baroreceptor site. In this value, total peripheral resistance and heart output both participate. Considering the location of baroreceptors in T. rugosus, this would be a difficult value to acquire.

As in all tetrapods the cardiac pacemaker of T. rugosus is innervated by vagal cholinergic and sympathetic adrenergic fibres (Berger, 1971). Muscarinic blockade with either atropine or hyoscine produced a substantial rise in heart rate, and beat-to-beat fluctuations in heart rate disappeared. Thus, it seems likely that resting heart rate is maintained by modulating efferent vagal discharge in T. rugosus. From the present work it is impossible to judge whether a resting sympathetic discharge occurs in T. rugosus. Lillywhite & Seymour (1978) showed that in the tiger snake, Notechis scutatus, β-adrenoreceptor blockade with propranolol produced a bradycardia. They concluded that the heart of the tiger snake was normally under the influence of excitatory sympathetic tone. However, propranolol blocks the effects of circulating, as well as neurally released, catecholamines. Thus Lillywhite and Seymour’s observations do not necessarily indicate the existence of a resting discharge in cardiac sympathetic nerves.

An attempt was made to determine the contribution of vagal and sympathetic fibres to the gain of the baroreceptor-heart rate reflex by using hyosciene or atropine, and the adrenergic neurone blocking compound bretylium. The results summarized in Table 1 suggest that sympathetic fibres are more important in the cardiac baroreflex than vagal fibres. However, this conclusion requires qualification. It is possible that blockade of one group of fibres leads to a compensating change in the activity of the remaining group. In fact there is evidence that this occurs, particularly after bretylium treatment. Some 3 h after bretylium injection, aortic cuff inflation produced little reduction in heart rate and hyoscine injection led to a large increase in heart rate. The contribution of the vagus to gain might therefore be poorly estimated by the use of bretylium. In spite of this reservation, the combined evidence from muscarinic blockade and adrenergic neurone blockade indicates that in T. rugosus sympathetic fibres play the dominant role in generating the response of the heart to pressure disturbances.

It is a pleasure to acknowledge the perceptive comments on the manuscript of Professor G. D. Campbell. This study was supported in part by grants from the Australian Research Grants Committee. The work was performed while P. J. B. was the holder of a Melbourne University Research Fellowship.