Brazilian Journal of Medical and Biological Research (1998) 31: 863-888 Chemoreceptors and systemic hypoxia 863 ISSN 0100-879X

Chemoreceptors and cardiovascular control in acute and chronic systemic hypoxia

J.M. Marshall Department of Physiology, The Medical School, Birmingham, UK

Abstract

Correspondence This review describes the ways in which the primary bradycardia and Key words J.M. Marshall peripheral vasoconstriction evoked by selective stimulation of periph- • Hypoxia Department of Physiology eral chemoreceptors can be modified by the secondary effects of a • Adenosine The Medical School chemoreceptor-induced increase in ventilation. The evidence that • Chemoreceptors Birmingham B15 2TT strong stimulation of peripheral chemoreceptors can evoke the behav- • Vasodilatation UK • Noradrenaline E-mail: [email protected] ioural and cardiovascular components of the alerting or defence response which is characteristically evoked by novel or noxious Presented at the XII Annual Meeting stimuli is considered. The functional significance of all these influ- of the Federação de Sociedades de ences in systemic hypoxia is then discussed with emphasis on the fact Biologia Experimental, Caxambu, that these reflex changes can be overcome by the local effects of MG, Brasil, August 27-30, 1997. hypoxia: central neural hypoxia depresses ventilation, hypoxia acting on the heart causes bradycardia and local hypoxia of skeletal muscle and brain induces vasodilatation. Further, it is proposed that these Received January 20, 1998 local influences can become interdependent, so generating a positive Accepted March 17, 1998 feedback loop that may explain sudden infant death syndrome (SIDS). It is also argued that a major contributor to these local influences is adenosine. The role of adenosine in determining the distribution of O2 in skeletal muscle microcirculation in hypoxia is discussed, together with its possible cellular mechanisms of action. Finally, evidence is presented that in chronic systemic hypoxia, the reflex vasoconstrictor influences of the sympathetic nervous system are reduced and/or the local dilator influences of hypoxia are enhanced. In vitro and in vivo findings suggest this is partly explained by upregulation of nitric oxide (NO) synthesis by the vascular endothelium which facilitates vasodi- latation induced by adenosine and other NO-dependent dilators and attenuates noradrenaline-evoked vasoconstriction.

Introduction cardiovascular system. Since then, a great deal has been discovered about the respira- The peripheral chemoreceptors have been tory, cardiac and vascular responses evoked known about at least since the beginning of when the chemoreceptors are selectively this century. In the 1920’s and 1930’s it was stimulated and about the ways in which the recognised that stimulation of the carotid responses interact with one another. How- and aortic chemoreceptors could produce ever, still relatively little is known about the reflex effects both on the respiratory and functional importance of peripheral chemore-

Braz J Med Biol Res 31(7) 1998 864 J.M. Marshall

ceptors in cardiovascular control under natu- of chemoreceptor stimulation could have a ral, physiological conditions, or in patho- large impact on the magnitude and direction logical situations. Over the last 10-15 years of the cardiovascular changes (3). The na- we have been investigating their role in regu- ture of these respiratory and cardiovascular lating the cardiovascular system in acute and interactions was largely established by the chronic systemic hypoxia and in particular, carefully controlled work of M. de Burgh we have been trying to establish how the Daly and colleagues (see Figure 1). reflex responses interact with the local influ- ences of hypoxia. This is the major subject Vascular responses matter of this review. However, in order to put this work into context, it is first neces- In a series of studies on dogs anaesthe- sary to briefly review what is known of the tised with chloralose or barbiturates, Daly effects of selective stimulation of the periph- and co-workers perfused the vascularly iso- eral chemoreceptors. lated carotid regions with blood from a donor dog in such a manner that the perfusion pres- Primary and secondary responses sure and therefore the stimulus to the carotid baroreceptors was maintained constant, Between the 1930’s and 1950’s, several whilst the carotid bodies were exposed either studies were published on the cardiovascu- to normoxic, normocapnic blood or to hy- lar changes that could be evoked by selective poxic, hypercapnic blood. In early studies, stimulation of the carotid bodies, either by the systemic circulation was performed by a local injection of substances such as sodium pump at a constant volume per minute so that

cyanide or saline equilibrated with CO2, or any change in perfusion pressure reflected by perfusion with hypoxia and hypercapnic changes in systemic vascular resistance. In blood. The results obtained were somewhat this way they were able to show that when confusing, some studies showing a rise in ventilation was spontaneous, selective stimu- arterial pressure, others a fall, some showing lation of the carotid chemoreceptors induced bradycardia, and others tachycardia (e.g. 1,2). a fall in systemic vascular resistance indicat- However, already it was beginning to be ing peripheral vasodilatation. However, when recognised that the respiratory consequences ventilation was maintained constant during

Figure 1 - Schematic diagram showing the primary and sec- PaCO HR + vasodilatation ondary effects of selective stim- 2 ulation of peripheral chemore- ceptors. + indicates an excita- Lung stretch tory effect. Broken arrow indi- + HR + vasodilata- Respiration receptors cates direct or local rather than tion reflex effects. Open arrow to especially in defence areas indicates a higher skeletal muscle Inspiratory threshold effect. For further de- HR + vasoconstriction drive tails see text. HR, Heart rate. + Peripheral + chemoreceptor Defence HR + vasoconstriction + vasodilatation stimulation areas in splanchnic in skeletal and renal muscle HR + vasoconstriction in skeletal muscle, splanchnic and renal

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 865 artificial ventilation, the same stimulus tion of aortic chemoreceptors evoked only a evoked an increase in systemic vascular re- small ventilatory response. sistance reflecting peripheral vasoconstric- tion (4,5). This suggested that the primary Cardiac responses reflex response to carotid chemoreceptor stimulation was vasoconstriction and that the The heart rate response was also investi- vasodilatation was secondary to the chemore- gated in other early experiments on the dog. ceptor-induced hyperventilation. Other ex- Selective stimulation of the carotid chemore- periments showed that the peripheral vasodi- ceptors was shown to be more likely to evoke latation induced by chemoreceptor stimula- tachycardia when the ventilatory response tion was also greatly reduced or reversed to was large (~200% higher compared to con- vasoconstriction if the dog was allowed to trol); when the ventilatory response was hyperventilate, but with the vagus nerves cut, small, a bradycardia was often seen (6). Care- or if the partial pressure of CO2 in the arterial ful examination of these responses revealed blood (PaCO2) was maintained constant (4,6). that just as with the vascular responses, two Thus, it was concluded that the two major aspects of the hyperventilatory response to factors that contributed to the secondary va- chemoreceptor stimulation affected the car- sodilatation were a reflex triggered by vagal diac response: both pulmonary stretch re- afferent fibres and the hypocapnia that arose ceptor stimulation and hypocapnia contrib- from hyperventilation. Further experiments uted to the tachycardia, while the underlying demonstrated that vasodilatation secondary primary response was apparently bradycar- to hyperventilation occurred in skeletal dia (6-8). However, even when ventilation muscle and skin of the limbs, and in renal and was maintained constant, carotid chemore- splanchnic vasculature, the response being ceptor stimulation still sometimes evoked most pronounced in muscle (7,8). As far as tachycardia in a way that seemed related to the vagal afferents were concerned, the im- the stimulatory drive to ventilation and which portant receptors turned out to be the slowly was most evident when the cardiac vagal adapting pulmonary stretch receptors (6,9), fibres were intact (6,8). This effect was remi- while the efferent arm of the reflex was shown niscent of the sinus arrhythmia that persists to be an inhibition of sympathetic noradre- after lung denervation and which was at first nergic tone (7). The mechanisms underlying ascribed to “irradiation” of the cardiac vagus the vasodilator effects of hypocapnia have with impulses from the respiratory centre received far less attention: they seem to be (13). More recently, it has become clear that mediated mainly by the unloading of the collaterals of central inspiratory neurones central chemoreceptors for CO2 and the re- exert an inhibitory effect on cardiac vagal sulting fall in sympathetic noradrenergic ac- neurones (14,15). This helps to explain the tivity (see 10-12). increase in heart rate that normally occurs It is worth noting that in the early studies with every inspiration (sinus arrhythmia) and of Daly and Ungar (5), selective stimulation reflects a central neural mechanism by which of the aortic chemoreceptors by perfusion of heart rate tends to increase whenever in- the aortic arch with hypoxic hypercapnic spiratory drive increases. There is an addi- blood generally evoked peripheral vasocon- tional effect of central inspiratory drive that striction whether or not ventilation was held tends to increase the excitability of sympa- constant. This was consistent with the evi- thetic neurones, thus facilitating tachycardia dence that the vasodilatation that followed and vasoconstriction (16-18). However, this carotid chemoreceptor stimulation was sec- effect is much smaller than that exerted on ondary to the hyperventilation, for stimula- the cardiac vagus.

Braz J Med Biol Res 31(7) 1998 866 J.M. Marshall

As far as the aortic chemoreceptors are tion or hypocapnia (27-29). concerned, the nature of the primary cardiac response evoked by their stimulation is some- The alerting or defence response what controversial (see 19). Angell-James and Daly (20) demonstrated in their experi- In the middle of the 1960’s a report ap- ments on the dog, in which the aortic bodies peared (30) which suggested that primary were perfused with hypoxic, hypercapnic bradycardia and vasoconstriction, with the blood, that the primary cardiac response was possibility of tachycardia and vasodilatation bradycardia. By contrast, Hainsworth and secondary to hyperventilation, may not be an colleagues (21), who performed similar ex- adequate description of the cardiovascular periments, maintained that the primary re- effects of selective peripheral chemorecep- sponse was tachycardia. tor stimulation. The Italian group led by Zanchetti (30), who were interested in stimuli Species differences that might activate or inhibit the areas of the central nervous system concerned with ag- From all the experiments that have been gressive or defensive behaviour, performed performed on a range of different species, it a series of experiments on cats that had been seems that during selective stimulation of decerebrated at a high level so as to remove peripheral chemoreceptors, vasodilatation the cerebral cortex, but leaving the hypo- and tachycardia secondary to the chemore- thalamus and brain stem intact. Since the ceptor-induced hyperventilation are more usual cortical inhibitory influence over the pronounced in the dog than in other species brain stem has been removed, such prepara- (see Ref. 19). This is probably because the tions show “spontaneous” episodes of sham magnitude of the hyperventilatory response rage that are characterised by the somatic is greater in the dog for the same secondary and autonomic correlates of range or defen- effects can be demonstrated in other species sive behaviour in the cat, i.e., clawing and if the conditions are carefully managed to struggling movements, piloerection, retrac- maximise their effects. Nevertheless, among tion of the nictitating membrane, rapid ven- the commonly used laboratory animals, both tilation and a rise in arterial pressure. the tachycardia and peripheral vasodilata- Zanchetti and colleagues (30) showed that tion secondary to the hyperventilation of such sham rage episodes could be evoked selective chemoreceptor stimulation are weak not only by light mechanical stimulation of or non-existent in the cat (22,23), a small the skin, but by selective stimulation of the tachycardia has been recorded in the rabbit carotid bodies. This led them to suggest that (24,25), although whether or not a second- peripheral chemoreceptors have an excita- ary vasodilatation influence occurs has not tory input into the brain regions that mediate been studied. In the rat, a mild secondary the somatic and autonomic components of tachycardia has been demonstrated, but the defensive and aggressive behaviour. vasodilatation is apparently absent (26). The It must be emphasised that at this stage in limited information available on selective the investigative process, such a proposal chemoreceptor stimulation in primates, in- would have had to be treated with caution, cluding man, suggests that tachycardia and for removal of the normal cortical inhibitory peripheral vasodilatation secondary to hy- influences may have revealed a pathway that perventilation can occur but are weak, the was of no functional significance. However, influence of central inspiratory drive on heart it was very difficult to test the hypothesis rate being more important than the influ- more fully because the common anaesthetics ences of pulmonary stretch receptor stimula- of the time, including chloralose, barbituates

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 867 and halothane were found to block afferent aline. This cardiovascular pattern of response activation of the brain stem defence areas. In was accompanied by other autonomically other words, noxious stimuli that were known mediated changes including pupillary dilata- to evoke defensive or aggressive behaviour tion, retraction of the nictitating membrane, with the characteristic autonomic changes in piloerection and even urination and defaeca- conscious animals could not evoke the auto- tion (33-37). nomic components of the response in ani- Other experiments performed on con- mals anaesthetised with these agents (31,32). scious animals by Folkow and Hilton and Nevertheless, the studies that were per- their research group (33,34,37) showed that formed in the 1950’s, 1960’s and 1970’s by the characteristic cardiovascular pattern of Folkow and Neil (33) in Sweden and by response and the other autonomic changes Hilton (34) in England on anaesthetised cats could be evoked by electrical stimulation in and dogs told us a great deal about what the identified regions, by painful stimuli and became known as ‘the defence response’. by a variety of environmental stimuli rang- They concluded from experiments in which ing from sudden sound to confrontation with brain sites were stimulated electrically that an angry cat or dog. The behavioural changes the regions that integrate the defence re- that accompanied these responses ranged sponse run from the amygdala via an effer- from arousal to mild alerting, or a display of ent pathway to the ventral hypothalamus, defensive behaviour, only reaching overt that a contiguous or further region runs rage, aggression or the full “defence re- through the central grey matter of midbrain sponse” when the stimulus was particularly and medulla and that all of these regions feed strong. Indeed, given that i) the pattern of the through a single pathway that runs through cardiovascular response was consistent irre- the ventral medulla. More recently, this path- spective of the strength of the behavioural way was found to synapse onto the neurones response, ii) defensive or aggressive behav- of the rostral ventrolateral medulla (RVLM) iour was only seen in extremis and iii) the that is now recognised as being of major same cardiovascular pattern was seen in hu- importance in setting the resting level of man subjects when they were exposed to vasomotor tone and in integrating the major relatively mild arousing stimuli, such as sound cardiovascular reflexes (35-37). The pattern and mental tasks, the term “defence response” of response that was evoked in anaesthetised is a misnomer. By emphasising one end of animals by electrical stimulation within the the spectrum, it associates the characteristic integrating regions for the defence response pattern of cardiovascular response with ex- and which was, in fact, crucially important treme circumstances, when in fact this pat- in allowing these regions to be identified, tern of response probably occurs in every characteristically included a rise in arterial one of us many times during every day, pressure, tachycardia, vasoconstriction in whenever we experience a novel or unex- cutaneous renal and splanchnic regions, but pected situation, or become emotionally vasodilatation in skeletal muscle. This muscle stressed by events around us. For these rea- vasodilatation was not secondary to hyper- sons, the term “alerting response” is gener- ventilation or muscle contraction, but was a ally far more useful and accurate (see 37). It true primary component of the ‘defence re- is used in the remainder of this review. sponse’. In cats and dogs it was largely medi- ated by sympathetic cholinergic fibres, but A “new” anaesthetic and a “new” facet of was attributable also to inhibition of sympa- the chemoreceptor response thetic noradrenergic activity to muscle and to the dilator influence of circulating adren- Experiments performed in Birmingham,

Braz J Med Biol Res 31(7) 1998 868 J.M. Marshall

UK, around 1980 demonstrated that the ste- virtually identical with the pattern evoked by roid anaesthetic alphaxalone-alphadalone stimulation in the amygdala or hypothalamic (Althesin or Saffan) does not have the same defence areas, or by stimulation of nocicep- depressant actions on transmission through tor afferent fibres in the radial nerve (39). In the defence areas of the brain as the com- the cat, the muscle vasodilatation evoked by monly used anaesthetics (see 38 and above). chemoreceptor stimulation was mediated by It can be given by intravenous infusion at a sympathetic cholinergic fibres, inhibition of rate that produces anaesthesia and full anal- sympathetic noradrenergic fibres and circu- gesia and yet which does allow the charac- lating adrenaline (39), while in the rat it was teristic cardiovascular and autonomic com- mediated by the last two influences only, as ponents of the alerting response to be evoked is characteristic of the muscle dilatation of by stimulation of known afferent inputs to the alerting response in these species (40). the defence areas, for example nociceptor In view of such evidence, one might ques- afferent fibres in peripheral nerves. This find- tion whether the primary and secondary re- ing was extremely important for it paved the sponses to chemoreceptor stimulation stud- way for detailed studies of the organisation ied under the more commonly used anaes- of the areas of the brain that are responsible thetics (see above) have any functional sig- for the alerting response, from the afferent nificance. The answer to this is undoubtedly, inputs, and including the presumed integrat- yes they do! Experiments on the cat showed ing areas of the amygdala and hypothalamus that when the depth of Althesin anaesthesia and the efferent pathway of the ventral me- was increased by increasing the infusion rate dulla. In particular, it allowed a more com- of the agent, then the tachycardia and muscle plete study of the responses evoked by pe- vasodilatation of the response to carotid ripheral chemoreceptor stimulation. In fact, chemoreceptor stimulation disappeared and all of the research we have performed on were commonly replaced by bradycardia and anaesthetised animals, which is discussed muscle vasoconstriction (39). Furthermore, below, was performed under Althesin or at a light level of anaesthesia, the muscle Saffan anaesthesia and much of it would not vasodilatation and tachycardia that were have been possible without it. Since Althesin evoked by carotid chemoreceptor stimula- was the agent originally introduced for use tion were converted to vasoconstriction and in human patients, but Saffan, which is chemi- bradycardia simply by reducing the concen- cally identical, replaced it as the veterinary tration of the solution used to stimulate the product, both agents are referred to below carotid chemoreceptors. In other words, un- reflecting the year in which the experiments der deep Althesin anaesthesia or with mild were performed. chemoreceptor stimulation, the response that In cats anaesthetised with Althesin and in remained was a mild hyperventilation, rats anaesthetised with Saffan, we were able generalised vasoconstriction in muscle and to show that selective stimulation of the other tissues (41) with bradycardia, just as carotid chemoreceptors evokes the full car- would be expected on selective stimulation diovascular pattern of the alerting response of carotid chemoreceptors under chloralose as described above, including tachycardia or barbiturates, given that the secondary ef- and muscle vasodilatation that are not sec- fects of hyperventilation upon the cardiovas- ondary to the hyperventilation (39,40). This cular system are weak in the cat (see above). response is accompanied by the other auto- Essentially similar results were obtained in nomic features of the alerting response. In the rat under Saffan except that the tachycar- individual animals the pattern of response dia was more persistent, consistent with the evoked by chemoreceptor stimulation was more pronounced tachycardic response to

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 869 hyperventilation in this species (see above may reflect the fact that potassium cyanide and 40). given into the systemic blood stream does On this basis, it seemed reasonable to not provide a simple, selective stimulus to propose that in the absence of anaesthesia, the carotid chemoreceptors. For example, in mild chemoreceptor stimulation would evoke the rat, carotid chemoreceptor stimulation the primary bradycardia and peripheral va- commonly evokes an augmented breath or soconstriction, complicated by the tachycar- respiratory gasp, probably because chemore- dia and/or vasodilator effects of hyperventi- ceptor stimulation facilitates the input of lation in those species in which these sec- rapidly adapting irritant receptors: an aug- ondary influences are evident, but that strong mented breath may be accompanied by a chemoreceptor stimulation would evoke the transient, but pronounced bradycardia full cardiovascular pattern of the alerting (40,43). response (see Figure 1). Correspondingly, The evidence available from studies on behavioural arousal or alerting would be systemic hypoxia in conscious animals and expected to accompany mild chemoreceptor human subjects lends good support to the stimulation, while obvious aggressive or de- general idea that the defence areas can be fensive behaviour might be expected on activated by carotid chemoreceptor stimula- strong chemoreceptor stimulation (39). tion (see below). Unfortunately, these proposals have not yet been fully tested: it is difficult to selec- Functional implications of chemoreceptor- tively stimulate carotid chemoreceptors in a induced activation of the defence areas conscious animal and it is technically diffi- cult to record a full range of cardiovascular If it is accepted that experimental stimu- variables. It is also difficult to establish in lation of the carotid chemoreceptors can ac- conscious animals whether any muscle va- tivate the brain defence areas, then it is rea- sodilatation and tachycardia are integral com- sonable to assume that in everyday life stimuli ponents of the alerting response rather than that activate these receptors, such as hy- secondary effects of hyperventilation, since poxia, hypercapnia and acidaemia, are some this requires the use of paralysing agents. of many inputs, together with visual, audi- Nevertheless, the fragmentary evidence that tory, tactile, noxious and emotional inputs, is available is promising (for a more com- that have the potential to produce the various plete discussion, see Ref. 19). In particular, facets of behavioural arousal and the charac- Franchini and Kreiger (42) showed in ex- teristic cardiovascular and autonomic com- periments on conscious rats that systemic ponents of the alerting response. Since pe- administration of potassium cyanide which ripheral chemoreceptors are tonically active stimulates peripheral chemoreceptors evoked even at normal values of arterial PO2, PCO2 a dose-dependent behavioural response that and pH, it can be argued that their input ranged from arousal to escape behaviour. contributes to the general level of arousal in The accompanying cardiovascular response the awake state (19,40,41). Further, since was a graded rise in arterial pressure, with efferent activity from the defence areas con- tachycardia when the dose of potassium cya- tributes to the tonic level of arterial pressure nide was low and bradycardia when it was through its influence on the RVLM, it can be higher: regional flows were not recorded. argued that peripheral chemoreceptors help The increasing dominance of the bradycar- to set the resting level of arterial pressure in dia is surprising in view of the results dis- the awake state by their influence on defence cussed above but does not necessarily argue areas, as well as by their more direct influ- against the general hypothesis. Rather, it ence on the RVLM or other descending path-

Braz J Med Biol Res 31(7) 1998 870 J.M. Marshall

ways to sympathetic preganglionic neurones stimulation of the baroreceptors this pro- (40,41). On the other hand, since during duces must limit the reflex effects that are sleep, hypoxic stimulation of the peripheral produced by the chemoreceptors. Moreover, chemoreceptors is well known as a stimulus in the presence of any other stimulus to the that can serve to wake the individual, it is individual which is simultaneously raising now reasonable to argue that this reflects arterial pressure, one can infer that this would specific activation of the defence areas by tend to further reduce the reflex vasocon- the chemoreceptors, and that the increased striction that the chemoreceptors can pro- activity in the defence areas contributes to duce. the accompanying rise in arterial pressure However, it is also known, from experi- (see 19). It is then a simple step to propose ments in which the hypothalamic defence that longer term activation of peripheral areas were electrically stimulated, that acti- chemoreceptors, as occurs for example in vation of the defence areas causes inhibition chronic obstructive sleep apnoea, and the of both the bradycardia and peripheral va- hypertrophy and increased sensitivity of the sodilatation evoked by stimulation of the carotid bodies that has been reported in es- baroreceptors (51). In other words, the pat- sential hypertension might contribute to the tern of the defence response predominates hypertension of these conditions by activat- and the baroreceptor reflex is suppressed. ing the defence areas (44,45). This hypoth- This suppression is explained, at least in esis gains weight when one appreciates the part, by a pathway from the hypothalamus effects of defence area activation on the which influences the baroreceptor reflex baroreceptor reflex and realises that chemore- pathway within the nucleus tractus solita- ceptor stimulation can mimic these effects, rius, close to the site of termination of the as is discussed in the next section. baroreceptor afferent fibres (52). In view of this evidence it was reasonable to question Interactions of chemoreceptor stimulation how the effects of chemoreceptor and barore- with the baroreceptor reflex ceptor stimulation would interact under con- ditions in which peripheral chemoreceptors Arterial baroreceptor stimulation, as oc- could activate the defence areas. curs when pressure in the carotid sinus and In fact, experiments on cats anaesthe- aortic arch is raised, produces reflex brady- tised with Althesin, in which the level of cardia and vasodilatation which tends to re- anaesthesia was light enough and the turn arterial pressure to the normal level. chemoreceptor stimulus was strong enough From experiments performed on animals to evoke the full pattern of the alerting re- anaesthetised with the commonly used an- sponse, showed that during this response aesthetics such as chloralose and barbitu- even a supramaximal stimulus to the barore- rates, it has generally been reported that ceptors, an increase in carotid sinus pressure selective stimulation of the baroreceptors to 250 mmHg, produced no detectable effect inhibits the peripheral vasoconstriction on the cardiovascular system. Furthermore, evoked by selective chemoreceptor stimula- if a baroreceptor reflex was initiated first and tion (46-49). This may occur, at least in part, then the carotid chemoreceptors were stimu- by an inhibitory influence of baroreceptor lated, the cardiovascular pattern of the alert- afferent input on chemoreceptor afferent in- ing response completely overcame the car- put within the nucleus tractus solitarius (50). diac and vascular components of the barore- Since chemoreceptor stimulation itself raises ceptor reflex. On the other hand, if the depth arterial pressure by causing vasoconstric- of Althesin anaesthesia was increased, or the tion, this suggests that even the concurrent stimulus to the chemoreceptors reduced, so

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 871 that carotid chemoreceptor stimulation tion of the carotid chemoreceptors as might evoked bradycardia and vasoconstriction occur in sleep apnoea might, by activating the even in muscle, then the reflex effects of defence areas, tend to inhibit the normal abil- baroreceptor stimulation interacted in an ity of the baroreceptor reflex to buffer in- additive way and the bradycardia of chemore- creases in blood pressure and so further contri- ceptor stimulation was accentuated, while bute to the development of hypertension. the vasoconstrictor responses were reduced, or reversed to vasodilatation, as might have Acute systemic hypoxia been predicted from the experiments on ani- mals anaesthetised with the commonly used The information that is reviewed above anaesthetics (53). about the effects of selective stimulation Thus, from these results it can be argued upon the cardiovascular system is the infor- that a strong stimulus to the carotid chemore- mation that provided the background for our ceptors occurring acutely, as for example on studies on the effects of acute systemic hy- sudden exposure to a hypoxic environment, poxia. But it was also important for us to would not only activate the defence areas to recognise that systemic hypoxia can have produce the cardiovascular pattern of the additional effects on the respiratory and car- alerting response, but would suppress the diovascular systems that may interfere with baroreceptor reflex so that the rise in arterial the effects of selective stimulation of the pressure would be greater than otherwise carotid chemoreceptors. All of these effects expected and the perfusion pressure for the are indicated schematically in Figure 2. delivery of blood flow to the skeletal muscles Briefly, there is substantial evidence that would be consequently greater. This, as has hypoxia of the central nervous system can been argued for other stimuli that activate depress ventilation by influences on central the defence areas, may confer a split second respiratory neurones. Further, there is evi- advantage on the individual, in this case dence that central neural hypoxia can in- allowing him or her to run away from the crease sympathetic activity to the heart and hypoxic environment more readily (34,37). circulation, probably in part by acting on Further, more prolonged, but strong stimula- neurones of the ventrolateral medulla (see

Figure 2 - Schematic diagram PaCO2 HR + vasodilatation showing the possible effects of systemic hypoxia. + indicates an excitatory effect; - indicates an + Lung stretch HR + vasodilata- hibitory effect. Arrows as in Fig- Respiration receptors tion ure 1. CNS, Central nervous sys- especially in tem; HR, heart rate. Inspiratory skeletal muscle drive HR + vasoconstriction + Peripheral PaO + Defence 2 HR + vasoconstriction + vasodilatation chemoreceptors areas in splanchnic in skeletal HR + vasoconstriction and renal muscle in skeletal muscle, splanchnic and renal

CNS HR + vasoconstriction

HR + vasodilatation

Braz J Med Biol Res 31(7) 1998 872 J.M. Marshall

19,54). On the other hand, it is well known In a recent study on conscious rats instru- that by direct actions on the heart, hypoxia mented with an ultrasonic Doppler probe for causes bradycardia and myocardial depres- recording hindquarters blood flow and an sion, while by direct actions on the tissues of arterial cannula for recording arterial pres- the systemic circulation hypoxia causes va- sure and heart rate (58), we have attempted a sodilatation. Thus, we were aware that the more complete analysis of the behavioural final or observed response to systemic hy- and cardiovascular changes evoked by expo-

poxia must represent a balance of these indi- sure to 8% O2. Within the first 3 min of vidual effects. It was our intention to estab- hypoxia, we have seen the full repertoire of lish their relative importance. behaviour that is typical of the rat in states ranging from mild arousal to overt fear, or Systemic hypoxia and the alerting response aggression. Concomitantly, arterial pressure, heart rate and hindquarters vascular conduc- In cats and rats anaesthetised with Saffan, tance tended to rise, but it would be difficult it became clear that systemic hypoxia to associate these changes specifically with

achieved by breathing 12, 8 or 6% O2 could the alerting response, rather than with the evoke the full cardiovascular pattern of the exercise that is occurring as well. For the alerting response (55,56). In such experi- remainder of a 20-min period of hypoxia the ments it is difficult to deduce the threshold signs of arousal were less marked and the for the response, but a response pattern that cardiovascular variables returned towards or was recognisable as that of the alerting re- below baseline, consistent with the observa-

sponse was commonly seen when PaO2 tions we have made under Saffan anaesthe- dropped to below 66-65 mmHg. This is not sia (see below). only consistent with the evidence that selec- Interestingly, when the same animal was tive stimulation of carotid chemoreceptors exposed to the same level of hypoxia on the can evoke the cardiovascular pattern of the following day, then the signs of behavioural alerting response under Althesin or Saffan arousal were much less pronounced (58). This anaesthesia, but is consistent with many re- suggests that at least the behavioural compo- ports in the literature that systemic hypoxia nents of the alerting response evoked by elicits behavioural arousal or even escape chemoreceptor stimulation habituate on rep- behaviour in the rat, rabbit, dog and in hu- etition of the stimulus, as is known to occur man subjects (see 19 for references) and with repetition of other alerting stimuli. The with evidence that episodes of behavioural important question, that we cannot yet answer, arousal during hypoxia were associated with is whether when hypoxic stimulation of the tachycardia and a rise in arterial pressure chemoreceptors is repeated, their ability to (57). The fact that signs of behavioural evoke the cardiovascular components of the

arousal were reported when PaO2 reached alerting response and thereby to suppress the relatively low values of below 50 mmHg baroreceptor reflex also habituates. As far as may indicate that the behavioural compo- other alerting stimuli are concerned, such as nents of the alerting response require a more sudden sound, cold water or emotional stress, severe level of hypoxia than the cardiovas- it is known that the cardiovascular compo- cular components. However, the apparent nents of the alerting responses and behaviour- discrepancy may simply reflect differences al alerting do not necessarily habituate to- in the observer’s ability to detect the cardio- gether. Moreover, there may be variation be- vascular pattern of the alerting response in tween individuals, so that some show habitua- an anaesthetised animal and a change in tion of one or more cardiovascular compo- behaviour in a conscious individual. nents of the alerting response on repetition of

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 873

1 the stimulus, while others show no change, or 1 /2 min of hypoxia. This pattern seems to be even sensitisation of one or more components superimposed upon more gradual changes (37). For systemic hypoxia, it is tempting to that are graded with the level of hypoxia speculate that those individuals who show no (55,56). In the cat, there is hyperventilation habituation of the cardiovascular components which is well maintained at least during a 3- of the alerting response to chemoreceptor stim- min period of hypoxia, accompanied by a ulation would be most at risk of suffering the rather small tachycardia and rise in arterial acute effects of a sudden rise in arterial pres- pressure and no change, or slight vasodilata- sure and tachycardia, such as aneurysm or tion in mesenteric, renal and skeletal muscle myocardial infarction during repeated asth- vasculature (56). In the rat, there is also matic attacks and more at risk of developing hyperventilation, but with a more pronounced hypertension as a consequence of sleep ap- tachycardia and substantial vasodilatation in noea. mesenteric, renal, cerebral and skeletal muscle vasculature. Moreover, the hyper- The gradual effects of hypoxia ventilation and tachycardia tend to return towards, or below the baseline, this becom- From our experiments on Saffan anaes- ing more obvious if the period of hypoxia is thetised cats and rats, it is clear that the prolonged to 5 or 10 min (55,59,60; see cardiovascular pattern of the alerting response Figure 3). The net effect of these changes is is most likely to be evoked within the first 1- that arterial pressure falls in the rat to an

Figure 3 - Original traces show- ing the effect of the adenosine 5 receptor antagonist 8-phenyl- VT theophylline (8-PT) on the respi- (ml) 0 ratory and cardiovascular re- 200 sponses evoked in the anaes- thetised rat by a 5-min period of R F breathing 8% O . V , Tidal vol- (breaths/min) 2 T ume; RF, respiratory frequency; 0 FVC, femoral vascular conduc- 0.06 tance; FBF, femoral blood flow; HR, heart rate; ABP, arterial FVC blood pressure. 8-PT was given (ml min-1 mmHg-1) between the left and right hand 0 panels. Note that after 8-PT the 15 VT and HR were better main- tained during hypoxia while the FBF increase in FVC and fall in ABP (ml/min) were reduced. 0 500 HR (beats/min) 250 250 ABP (mmHg)

0

1 min

Braz J Med Biol Res 31(7) 1998 874 J.M. Marshall

extent that is graded with the level of hy- secondary effects of hyperventilation on lung poxia (55). stretch receptors and by the ability of central By performing experiments in which we nervous hypoxia to increase cardiac sympa-

have vagotomised, kept PaCO2 constant, or thetic activity (56,60,62). The late bradycar- paralysed and artificially ventilated the ani- dia that occurs when hypoxia is prolonged mals, we have been able to establish the can be attributed not only to the primary roles of the secondary effects of hyperventi- bradycardia of carotid chemoreceptor stim- lation in these gradual changes. They are ulation, but also to the local effects of hy- relatively weak in both species in agreement poxia on the heart (60,63), while the late with their effects on the responses evoked by decrease in ventilation towards or below the selective stimulation of carotid chemorecep- baseline reflects the central effect of hypoxia tors (see above). In the cat, the lung stretch on respiratory neurones (63). Clearly, this receptors with vagal afferents apparently have later decrease in the ventilatory response no significant influence on the heart rate or means that lung stretch receptors make less vascular responses to systemic hypoxia, but contribution to the tachycardia as the period

if PaCO2 is prevented from falling with the of hypoxia progresses (see 60). The periph- hyperventilation, then there is bradycardia eral vasodilatation in mesenteric and in rather than tachycardia, vasoconstriction in muscle vasculature largely reflects the local the renal and mesenteric circulation, and no dilator effects of tissue hypoxia. These influ- change or slight dilatation in skeletal muscle ences are so strong that they apparently over- (56). This strongly suggests that in the nor- come the primary reflex vasoconstriction

mal situation, the fall in PaCO2, plays a expected from carotid chemoreceptor stimu- major role in counteracting the bradycardia lation and are largely responsible for the fall and vasoconstriction that would be expected in systemic arterial pressure (55). as a primary response to hypoxic stimulation The rat is, therefore, far more dominated of carotid chemoreceptors. Since vasocon- by the local effects of hypoxia on central striction did not occur in muscle, even when respiratory neurones, heart and vasculature

PaCO2 was controlled, it seems likely that in than the cat. This is of particular interest this tissue, the neurally mediated vasocon- because there are reports in the literature striction was opposed by the local dilator indicating that the respiratory and cardiovas- influences of hypoxia. cular responses evoked by systemic hypoxia

By contrast, in the rat, the fall in PaCO2 in neonates are very similar to those we have makes only a minor contribution to the ta- seen in the rat: characteristically, ventilation chycardia and muscle vasodilatation in se- increases and then falls and generally arteri- vere hypoxia (61). On the other hand, the al pressure decreases (see 64,65). We have lung stretch receptors with vagal afferents suggested this may arise because small mam- do contribute to the tachycardia, but not to mals in general, whether the neonates of the peripheral vasodilatation (62). Indeed, large mammalian species or the adults of when ventilation was kept constant by artifi- small mammalian species, have a higher rate

cial ventilation, then hypoxia induced brady- of O2 consumption per gram body weight cardia, but with substantial vasodilatation in than larger mammals: they may therefore be skeletal muscle (60). By considering these more susceptible to the local metabolic ef-

results together with the effects of various fects of hypoxia when O2 supply is reduced antagonists of the sympathetic nervous sys- (55). Unfortunately, the pattern of response tem, we have concluded that the primary that is produced by the local effects of hy- bradycardia of hypoxic stimulation of ca- poxia is potentially life-threatening as is ex- rotid chemoreceptors is opposed both by the plained below.

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 875

A positive feedback loop? most at risk of SIDS are those who have been hypoxic from birth (66). The fact that respiration increases ini- tially in neonates who become hypoxic, but The mediator of the local effects of then falls, has been put forward as an expla- hypoxia - adenosine nation for sudden infant death syndrome (SIDS) or cot death (66). We believe the Although the section above has empha- concomitant cardiovascular changes are cru- sised the importance of the local effects of cially important in this situation. In experi- hypoxia in the rat and other small mammals, ments on rats in which we have recorded it is important to recognise that the local cerebral blood flow, we have demonstrated effects do contribute to the overall response that although there is cerebral vasodilatation to hypoxia in larger mammals as well. It is in response to systemic hypoxia, cerebral simply that in larger mammals the balance is blood flow begins to fall when ventilation tipped more strongly towards the neurally and heart rate drop below their resting values mediated, reflex effects of hypoxia. Thus, a and arterial pressure drops below the auto- fall in ventilation when hypoxia is severe or regulatory range of about 60 mmHg (60). prolonged that can be attributed in part to a We have proposed that the respiratory and depressive effect of hypoxia on the central cardiovascular changes may generate a posi- nervous system (68) is a generally recognised tive feedback loop (Figure 4). Thus, when phenomenon occurring, for example, in the ventilation begins to fall, this exacerbates cat and in human subjects (69,70). Similarly, the fall in PaO2 and the peripheral vasodila- it is known that heart rate and arterial pres- tation and bradycardia that are caused by the sure begin to fall when systemic hypoxia is local effects of hypoxia, so potentiating the severe or prolonged, for example, in both the fall in arterial pressure. Once the arterial cat and human subjects (56,71). It is also pressure is below the autoregulatory range, known that in human subjects systemic hy- then cerebral blood flow falls, the O2 supply poxia induces vasodilatation in splanchnic to the brain is further reduced and the central circulation and forearm muscle (71,72) and neural hypoxia is exacerbated so that the that when the influence of the sympathetic central respiratory neurones are further de- nerve fibres on forearm muscle is blocked, pressed (60). systemic hypoxia still induces substantial Our experiments on newborn piglets in- muscle vasodilatation, as would be consist- dicate that this positive feedback loop is ent with a local dilator influence of hypoxia particularly likely to develop in individuals in whom the stimulatory effect of hypoxia on Figure 4 - The proposed posi- respiration is weak (64). Furthermore, our CNS tive feedback loop that may de- hypoxia studies on rats that have been kept chroni- velop during systemic hypoxia, particularly in small adult mam- cally hypoxic from birth in a hypoxic cham- mals and neonates. CNS, Cen- ber, have shown that the bradycardia fall in tral nervous system; HR, heart arterial pressure and fall in cerebral blood Cerebral Ventilation rate; ABP, arterial blood pres- blood flow sure. flow induced by acute hypoxia is greatly exaggerated in these animals relative to those PaO2 seen in controls (67). Thus, it seems that chronic hypoxia from birth accentuates those ABP HR Vasodilatation very responses that contribute to the positive feedback loop. This is entirely consistent with studies indicating that babies who are

Braz J Med Biol Res 31(7) 1998 876 J.M. Marshall

(72). The question of what produces these addition, when 8-PT was applied topically to local effects is therefore important and may the cerebral cortex, the hypoxia-induced ce- be of general relevance. rebral vasodilatation was attenuated (80). It is generally accepted that adenosine is On the other hand, when 8-sulphophenylthe- produced in tissues when they become hy- ophylline, an adenosine receptor antagonist poxic. When we began our experiments a- that does not cross the blood brain barrier, or denosine concentrations had been shown to adenosine deaminase, which breaks down rise substantially in the brain soon after the adenosine, but does not cross the blood brain onset of hypoxia (73). Moreover, adenosine barrier, were given systemically, they did had not only been implicated in hypoxia- not affect the respiratory response to hy- induced cerebral vasodilatation (74), but it poxia, had less effect on the late fall in heart had been shown to depress ventilation by rate than 8-PT, but still caused a substantial actions on respiratory neurones. In fact, sub- reduction of the muscle vasodilatation (63). stances such as aminophylline and theophyl- These results provide strong evidence that line whose effects include blockade of aden- adenosine released locally in the brain, heart osine receptors have been shown to reduce and skeletal muscle is responsible respec- respiratory depression associated with sys- tively for the respiratory depression and ce- temic hypoxia in neonates (75), as well as in rebral vasodilatation, the bradycardia and adult cats and human subjects (69,70). Fur- the muscle vasodilatation. They also rein- ther, adenosine is well known to be released force our proposal that the respiratory and in the heart under hypoxic conditions and cardiovascular changes are interdependent adenosine not only induces coronary vasodi- in a potentially positive feedback manner latation (76), but can induce bradycardia by (60,63; see Figure 4): if the antagonist used direct actions on the sinoatrial node and by had no effect on the respiratory component interfering with sympathetic and vagal trans- of the response then its effects on the cardiac mission (77). Adenosine had also been shown and vasodilator components were less pro- to be released by skeletal muscle during nounced than for an antagonist that had the muscle contraction when there is a relative potential to affect all components of the hypoxia (78) and by the use of various aden- response. By using specific antagonists for osine antagonists, adenosine had been impli- the subtypes of adenosine receptor, we have cated in the muscle vasodilatation that oc- recently shown that the adenosine that is curs during contraction (e.g. 79). We there- released in systemic hypoxia probably acts

fore hypothesised that adenosine plays a via A2A receptors to produce cerebral va- major role in the cerebral and respiratory and sodilatation (81), but via A1 receptors to cardiovascular changes that occur in the rat produce muscle vasodilatation (82,83). Our during systemic hypoxia. own evidence, together with that of others, To test this hypothesis we have used 8- indicates the respiratory depression and

phenyltheophylline (8-PT) which is a selec- bradycardia are probably mediated by A1 tive adenosine receptor antagonist that does receptors (82,84). not inhibit phosphodiesterase activity like Since adenosine is generally released in aminophylline and theophylline. In full ac- hypoxic tissues and since its ability to de- cord with our hypothesis, 8-PT given sys- press respiration by a central action and to temically greatly reduced the late decrease cause bradycardia and cerebral and muscle in ventilation and bradycardia, the muscle vasodilatation is common to a range of mam- vasodilatation and fall in arterial pressure malian species, it seems reasonable to as- induced by a 5- or 10-min period of breath- sume that adenosine plays a major role in the

ing 12, 8 or 6% O2 (59,63; see Figure 3). In local effects of hypoxia on respiration and

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 877 the cardiovascular system in all mammalian sponses are more common and larger in the species. Experiments already performed on distal, or terminal arterioles, than in the proxi- hypoxia-induced respiratory depression and mal arterioles (85). Thus, the fact that gross on hypoxia-induced bradycardia in isolated muscle vascular conductance increases dur- hearts in a number of species including adult ing systemic hypoxia disguises the fact that man, cat, rabbit and guinea pig are consistent within muscle some vessels constrict while a with that view (69,70,77). majority dilate. When an α adrenoceptor antagonist, Interactions between reflex and local effects phentolamine, was applied to the spinotra- of hypoxia in muscle microcirculation pezius, mean increases in arteriolar diameter produced in particular sections of the arteri- The experiments of the type discussed olar tree by systemic hypoxia were potenti- above in which pharmacological antagonists ated, while mean decreases in arteriolar di- are given systemically are useful in that they ameter were reduced, or reversed to dilator can indicate whether particular nerve-medi- responses. Moreover, the sections of the ar- ated, hormonally mediated or local influ- teriolar tree that were most affected were the ences contribute to the gross change in vas- primary and secondary arterioles of 18-40 cular conductance or resistance that occurs µm that are most strongly constricted by in skeletal muscle or other tissues during direct stimulation of the sympathetic nerves systemic hypoxia. However, they tell us little (81). More detailed analysis of each section of how these factors interact at the level of of the arteriolar tree showed that only ves- tissue microcirculation to determine the O2 sels that constricted in hypoxia were af- supply to the capillaries and thereby to the fected by phentolamine, while those that tissue cells. Further they suffer from the dilated were not (86). These observations disadvantage that many pharmacological are consistent with the occurrence of an antagonists when given systemically affect increase in sympathetic nerve activity to skel- arterial pressure and therefore their effect on etal muscle during systemic hypoxia, as the change in regional vascular conductance would be expected from the primary re- induced by systemic hypoxia may be in part sponse to carotid chemoreceptor stimulation related to their effect on perfusion pressure. (see above) and suggest that the sympathetic To address these issues we have performed nerve activity preferentially constricts the many experiments on the microcirculation primary and secondary arterioles. However, of the spinotrapezius muscle of the rat with they also indicate that many arterioles from the muscle prepared for intravital micros- all sections of the arteriolar tree do not re- copy. Pharmacological agonists and antago- spond to sympathetic activation in hypoxia, nists can then be applied topically to the even though we know they are innervated by muscle in doses that achieve local effects, sympathetic fibres. but which do not affect systemic variables, On the other hand, when the adenosine or the changes induced in them by systemic receptor antagonist, 8-PT, was topically ap- hypoxia. plied to the spinotrapezius muscle it reduced Experiments of this type have shown that mean increases in diameter induced in par- systemic hypoxia induces both increases and ticular sections of the arteriolar tree by sys- decreases in the diameter, dilatation and con- temic hypoxia, or converted them to mean striction, respectively, of individual arteri- decreases in diameter, the responses of the oles of the spinotrapezius muscle (85). On terminal arterioles being particularly affected balance, dilator responses are more common (87). These observations are, of course, fully than constrictor responses and dilator re- consistent with the evidence discussed above

Braz J Med Biol Res 31(7) 1998 878 J.M. Marshall

that adenosine plays a major role in the the major determinant of the response in any hypoxia-induced muscle vasodilation and given arteriole is in fact the local level of indicate that the terminal arterioles, the ves- hypoxia (see 91 and Figure 5). It is known sels that are traditionally thought to be most that there is considerable variation in the

responsive to locally released vasodilator level of tissue PO2 found in different regions metabolites, are particularly affected by lo- of skeletal muscle during normoxia (92). cally released adenosine. Nevertheless, only This reflects several factors including i) re- those arterioles that were dilated by systemic gional differences in the oxygen consump- hypoxia were affected by 8-PT, even though tion of the nearby muscle fibres, ii) distance we were able to show that all arterioles were along the arteriolar tree from the major sup-

responsive to exogenous adenosine (87). plying artery given that O2 diffuses out of Our observations on the effects of an- arterioles along their length, and iii) proxim- tagonists of the actions of circulating hor- ity of arterioles and venules with opposite

mones are comparable with those just de- directions of blood flow, given that O2 may scribed in that a vasopressin receptor an- diffuse from an arteriole to a venule, thus

tagonist affected the arterioles that constricted short-circuiting O2 supply to tissue down- in response to systemic hypoxia, but had no stream of the arteriole. It would be expected effect on those that dilated (88), whereas a ß that adenosine would reach higher concen- adrenoreceptor antagonist affected the arte- trations during systemic hypoxia in regions rioles that dilated in response to systemic of the muscle where the local level of hy- hypoxia, but had no effect on those that poxia is relatively severe, either because of a

constricted (86). These observations were in high level of O2 consumption or a poor “ana- turn consistent with evidence that vasopressin tomical” distribution of O2. The arterioles in is released in response to selective stimula- such regions would be preferentially dilated tion of carotid chemoreceptor and that vaso- by adenosine and this may in turn increase pressin receptor blockade reduces the in- the susceptibility to other dilator influences crease in gross muscle vascular conductance such as the ß adrenoreceptor mediated effect induced by systemic hypoxia (89) and with of adrenaline. On the other hand, the con- evidence that adrenaline is released into the centration of adenosine would be low in blood stream by hypoxic stimulation of ca- regions of the muscle where the local level rotid chemoreceptors and makes a contribu- of hypoxia is relatively mild. The arterioles tion, albeit small, to the hypoxia-induced in such regions may then be particularly increase in gross muscle vascular conduc- vulnerable to constrictor influences such as tance (90). However, we have to conclude those of sympathetic nerve activity and cir- from the microcirculatory studies that some culating vasopressin (91). arterioles “escape” the constrictor influence This heterogeneity in the responses of of vasopressin or “escape” the dilator influ- the muscle arterioles during systemic hy- ence of adrenaline during systemic hypoxia, poxia may be functionally important in al- even though one would expect the concen- lowing a more homogenous distribution of

trations of the two hormones reached in O2 in the various parts of the muscle at a time individual arterioles to be very similar and when the gross O2 supply is reduced. In other even though we could show they were all words, the behaviour we have seen in indi- capable of responding appropriately to ex- vidual arterioles of muscle during systemic ogenous vasopressin or adrenaline (88). hypoxia may explain the finding that the

The most obvious explanation for the variation in the levels of tissue PO2 within heterogeneity of the responses seen amongst muscle is considerably reduced during sys- the arterioles during systemic hypoxia is that temic hypoxia concomitant with the fall in

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 879

Hypoxia Figure 5 - Schematic diagram showing how the balance of the nerve and hormonally mediated constrictor influences and the lo- cally and hormonally mediated dilator influences of systemic Chemoreceptor hypoxia on blood vessels within stimulation skeletal muscle may be differ- ent depending on the local level of hypoxia. In regions where the level of hypoxia is more severe the balance may be more readily tipped towards vasodilatation. Adrenal For further discussion see text. medulla

Adrenaline

Noradrenaline Noradrenaline

Severe Moderate local hypoxia local Adenosine hypoxia Adrenaline Adenosine Adenosine Vasopressin

Angiotensin

K+ K+

Constriction Dilatation

Braz J Med Biol Res 31(7) 1998 880 J.M. Marshall

average tissue PO2 (92). It may also demon- experiments had shown that in cardiac myo- strate how the O2 consumption of resting cytes and coronary artery myocytes, adeno- muscle can be maintained constant during sine receptors are coupled to ATP-sensitive + systemic hypoxia, for a more homogeneous K (KATP) channels in such a manner that distribution of blood flow and therefore O2 adenosine opens the channels (96,97). It was supply through the capillary network would also known that KATP channels are present help muscle fibres to maintain their O2 con- on skeletal muscle fibres (98). As a further + sumption by increasing their O2 extraction (93). part of the background, it was known that K can induce vasodilatation by stimulating Na- Mechanisms of action of adenosine K+ ATPase activity in vascular smooth muscle and by opening inwardly rectifying On the basis of the literature, it is theo- K+ channels (99). Thus, we hypothesised retically possible that during systemic hy- that adenosine released in skeletal muscle

poxia adenosine is released from skeletal during systemic hypoxia opens KATP chan- muscle fibres, vascular smooth muscle, or nels on the skeletal muscle fibres leading to endothelium. Also ATP released from these the release of K+ which then contributes to sites or from sympathetic nerve varicosities the muscle vasodilatation (100). or red blood cells may be hydrolysed extra- In agreement with this hypothesis we cellularly to adenosine by the action of 5’ec- demonstrated that glibenclamide, which in-

tonucleotidase. Further, it is theoretically hibits KATP channels, abolished the increase possible that adenosine induces vasodilata- in the K+ concentration of venous blood + tion by acting on the vascular smooth muscle draining skeletal muscle ([K ]v) that was or endothelium, or more indirectly, by acting induced by systemic hypoxia, without af- on the skeletal muscle fibres, or at prejunc- fecting the increase in the arterial concentra- + + tional sites on the sympathetic nerve vari- tion of K ([K ]a); glibenclamide also re- cosities (91). So far, we have only investi- duced the accompanying muscle vasodilata- gated a few of these possibilities. Firstly, we tion. Similarly, 8-PT not only reduced the found that αß methylene ADP, which inhib- muscle vasodilatation of systemic hypoxia its the action of 5’ectonucleotidase, had no (see above) but also abolished the increase + significant influence on the dilatation in- in [K ]v. Further, glibenclamide reduced duced in hind limb muscles of the rat by muscle vasodilatation evoked by adenosine, systemic hypoxia. This indicates that most while 8-PT abolished both the muscle va- + of the adenosine that is vasoactive is re- sodilatation and the increase in [K ]v evoked leased as such from the intracellular sites by adenosine (100). (94). Thus, it seemed reasonable to conclude Since it was already known that systemic that there are adenosine receptors on skeletal

hypoxia leads to an increase in plasma levels muscle fibres that are coupled to KATP chan- of K+ (95) and since we had shown that the nels: this has since been confirmed by elec- action of circulating adrenaline upon ß adre- trophysiological recordings (101). Our re- noreceptors of skeletal muscle was impor- sults indicate they can be stimulated to open tant in stimulating the re-uptake of K+ during and to allow K+ efflux by adenosine that is

systemic hypoxia, probably by acting on ß2 released in systemic hypoxia and they raise receptors coupled to Na+-K+ATPase (90; see the possibility that this K+ contributes to the Figure 6), we were particularly interested in muscle vasodilatation. However, it seems whether adenosine might also regulate K+ likely that the contribution K+ makes is rela- balance in skeletal muscle. At the time we tively small given that glibenclamide had a began our experiments, electrophysiological much smaller effect than 8-PT on the muscle

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 881 vasodilatation (100). Indeed, the fact that duced dilatation that is mediated by adeno- glibenclamide affected the early and not the sine is also dependent on NO synthesis by late part of the muscle vasodilatation of hy- the endothelium (94). It could be that the poxia (100) suggests that the opening of KATP channels that initiate the muscle va- KATP channels is particularly important in sodilatation are on the endothelium and initiating the vasodilatation rather than main- coupled to adenosine receptors so that their taining it and raises the possibility that the activation triggers synthesis of NO by

KATP channels that are of major importance hyperpolarising the endothelial cells. In in the vasodilatation are on the vascular agreement with this proposal, our recent stud- smooth muscle, or endothelium, rather than ies, which have shown that the hypoxia- on the skeletal muscle fibres. On the other induced muscle dilatation is mediated by A1 hand, the very fact that 8-PT had a much adenosine receptors, have also indicated that larger effect than glibenclamide on hypoxia- muscle vasodilatation mediated by A1 recep- induced dilatation (100) strongly suggests tors is dependent both on the opening of that adenosine also acts in a manner that is KATP channels and on NO synthesis (102,103; independent of KATP channels. Figure 6). Our experiments with nitro L-arginine Thus, we can summarise our results to methyl ester (L-NAME) which inhibits NO date by stating that the adenosine that makes synthesis are consistent with this suggestion the major contribution to the muscle vasodi- (94). For L-NAME greatly reduced both the latation that occurs in the rat during acute muscle vasodilatation induced by systemic systemic hypoxia probably acts by stimulat- hypoxia and that induced by infusion of ing adenosine A1 receptors on the endotheli- adenosine. Thus, it seems that the great ma- um and increasing the synthesis of NO which jority of the component of the hypoxia-in- then induces relaxation of the vascular

Figure 6 - Schematic diagram showing some of the factors that influence the arterioles of skeletal muscle during systemic hy- poxia and the cellular mechanisms by which they may act. Nora- drenaline released from sympathetic nerve fibres and circulating Skeletal muscle fibres in the blood stream exerts a constrictor influence via α receptors. Na+ Circulating adrenaline exerts a dilator influence via ß2 receptors + on the vascular smooth muscle, but may also stimulate the Na - β + 2 K ATPase pump on skeletal muscle fibres via ß2 receptors, thus + increasing the uptake of K . Adenosine that is released from K+ K+ ATP A2A? skeletal muscle fibres may cause the efflux of K+ from those fibres by stimulating adenosine receptors that are coupled to K+ ATP channels. The final concentration of K+ in the interstitial fluid Adenosine is therefore determined by the balance of the influences of + adrenaline and adenosine: K+ is a vasodilator. Adenosine may K also act directly on the vascular smooth muscle. In addition, adenosine that is released from the endothelial cells acts on receptors on the endothelial cells, possible by opening K+ ATP NO channels, to increase the synthesis of NO which exerts a major dilator influence on the vascular smooth muscle. Current evi- A2A? dence indicates that the adenosine receptors on the endothelium A1 β that are stimulated in hypoxia are of the A1 subtype while those 2 Adenosine Noradrenaline on the vascular smooth muscle and skeletal muscle fibres are of Adrenaline α the A2A subtype. Noradrenaline

Braz J Med Biol Res 31(7) 1998 882 J.M. Marshall

smooth muscle. Since the endothelium is a induce the right ventricular hypertrophy, very effective metabolic barrier to the trans- increased pulmonary vascular resistance port of adenosine (84), and since adenosine and pulmonary vascular remodelling that deaminase, which does not cross the vascu- are characteristic of chronic systemic hy- lar endothelium, was just as effective in poxia (104). When adult rats that have been reducing hypoxia-induced vasodilatation as made chronically hypoxic for 3-4 weeks 8-PT, an adenosine receptor antagonist (63), (CH rats) are studied under Saffan anaesthe-

it seems most likely that the majority of the sia while breathing 12% O2, they have in- adenosine is released from the endothelium creased ventilation relative to control, and acts on the endothelium in an autocrine normoxic (N) rats breathing air, which might fashion. Adenosine that is released from skel- be attributed to a tonic drive to respiration etal muscle fibres may contribute in a small exerted by the carotid chemoreceptors and

way to the dilatation by acting on KATP chan- probably to an increased sensitivity of the nels on the skeletal muscle fibres to stimu- central chemoreceptors to CO2 (105). How- late efflux K+ which relaxes the vascular ever, the systemic arterial pressure, heart smooth muscle (Figure 6). rate and gross vascular conductance of hind limb muscle are comparable in CH rats

Chronic systemic hypoxia breathing 12% O2 and N rats breathing air (105). This suggests that cardiovascular ad- Acute systemic hypoxia is of interest in aptations must have occurred: the muscle its own right because it can occur in a num- vasodilatation and reduction in arterial pres- ber of clinical conditions, on immediate ex- sure that might be expected from the re- posure to high altitude and on accidental or sponse to acute systemic hypoxia are appar- experimental exposure to hypoxic gas mix- ently absent, as are the sympathetically me- tures. However, chronic systemic hypoxia, diated tachycardia that might be expected as and the adaptations that occur in this condi- a secondary consequence of hyperventila- tion, is arguably more important because tion and as a result of hypoxia of the central chronic hypoxia is so common in many res- nervous system (see above). This new state piratory and cardiovascular disorders and may be partly explained by the fact that the because it occurs in individuals who increase in haematocrit in the CH rats allows

acclimatise to living at high altitude. The their arterial O2 content when they are breath- number of laboratory studies that have been ing 12% O2 to equal that of the N rats when performed on respiratory and cardiovascular they are breathing air (105). However, since adaptations to chronic systemic hypoxia is the adenosine receptor antagonist, 8-PT, in- still rather small. The results of studies per- creased the control level of ventilation in the formed on chronically hypoxic patients are CH rats it seems there must still be sufficient complicated by the pathological condition hypoxia of the central nervous system to that underlies their hypoxic state, while those cause a tonic release of adenosine and de- performed on healthy individuals who climb pression of central respiratory neurones to high altitude are complicated by many (105). On the other hand, since 8-PT had no factors including the effects of exercise and influence on the control levels of arterial exposure to low temperature. pressure, heart rate or muscle vascular con- Our own experiments have been per- ductance in the CH rats, this indicates there formed on rats housed in a hypoxic chamber are no tonic influences of adenosine upon

at 12% O2 with temperature, humidity and the systemic circulation. This suggests the day/night cycle kept constant. A level of 10- heart and peripheral tissues are no longer

12% O2 has been shown to be sufficient to hypoxic (105).

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 883

Interactions between local and reflex effects There are a number of possible explana- tions for these results. It may be that adeno- When the CH rats were made acutely sine is more readily released by an acute hypoxic by breathing 8% O2 rather than 12% hypoxic stimulus in the CH rats than in the N O2, then they showed a similar pattern of rats and/or the arterioles of CH rats are more response to that which occurs in acute hy- sensitive to adenosine than those of N rats. poxia in N rats. There was an increase in Another explanation, which is not mutually ventilation and heart rate, a fall in arterial exclusive, is that the arterioles of CH rats are pressure and an increase in muscle vascular less affected by the reflex vasoconstrictor conductance with a later fall in ventilation influences of acute hypoxia (see above) than and heart rate towards control levels. More- those of N rats and so they are more readily over, the magnitude of the fall in arterial overcome by the dilator influences. pressure and muscle vasodilatation induced The last possibility seemed to be a par- in the CH rats was almost as great as those ticularly interesting one to us. Firstly, it was that occurred in N rats when they were sub- reported that chronically hypoxic patients jected to the much larger change from air- with respiratory disease showed a reduced breathing to 8% O2 (105). This suggests that ability to maintain their arterial pressure when the vasodilator effects of acute hypoxia may they were subjected to lower body negative be potentiated in CH rats, and/or that the pressure (108): this might be explained if reflex vasoconstrictor responses to carotid they show impaired vasoconstrictor re- chemoreceptor stimulation are impaired. sponses to the sympathetic neurotransmitter Our experiments on the microcirculation noradrenaline. Secondly, it had also been of the spinotrapezius muscle were consistent reported that the dorsal aorta of CH rats with both of these possibilities. We found shows a reduced ability to constrict to phen- that the changes in arteriolar diameter in- ylephrine, vasopressin and angiotensin as duced when CH rats were switched from compared with N rats (109). breathing 12% O2 to 8% O2 were just as great We therefore performed experiments on as those induced in N rats when they were the spinotrapezius muscle and mesentery of switched from breathing air to 8% O2 (106). CH and N rats, to obtain dose-response curves Further, the effects of adenosine receptor for the effects of noradrenaline on the arteri- blockade on these responses were fully com- oles. For both the mesentery and muscle, the parable in the CH and N rats: in each section most obvious difference between the CH of the arteriolar tree, mean increases in di- and N rats was that the maximum vasocon- ameter were reduced or reversed to mean strictor responses evoked in the arterioles by decreases in diameter and quantitatively the noradrenaline were greatly reduced in the sizes of these effects were similar in CH and CH rats and the size of this effect was similar N rats (106). The maximal dilatation pro- in the mesentery and muscle (110). Since duced by topical application of adenosine in arterioles of the intestinal mesentery have each section of the arteriolar tree was also little or no tissue parenchyma around them, similar in the CH and N rats, so there was no there was no reason to argue that the re- reason to suppose that the arterioles of the sponses to noradrenaline were suppressed CH rats were capable of greater maximal by some factor released by tissue cells. dilatation in response to adenosine or any Rather, it seems the constrictor responses to other dilator influence than those of N rats noradrenaline must be reduced by some fac- (106). We have made very similar observa- tor that is intrinsic to the blood vessel wall. tions in the microcirculation of the intestinal As a way of investigating this phenome- mesentery in CH and N rats (107). non, we chose first to study the iliac artery of

Braz J Med Biol Res 31(7) 1998 884 J.M. Marshall

the rat in vitro, this being an artery that of CH rats, the maximum constrictor re- supplies the skeletal muscles of the hind sponse evoked by noradrenaline was greatly limb. When we compared sections of iliac enhanced by topical application of L-NAME artery taken from CH and N rats we found, in to the mesentery, so that it equalled the complete agreement with the results obtained maximum vasoconstrictor response recorded in vivo, that the maximum of the noradrena- in N rats (Marshall JM, unpublished obser- line response-curve arteries from the CH rats vations). was greatly reduced relative to that of the N Thus, our current hypothesis is that rats, but there was no difference between CH chronic hypoxia causes an up-regulation of and N rats in the concentration of noradrena- NO synthesis in the endothelium of the sys- line that produced 50% of the maximum temic circulation. This may be attributed at response. This indicates that there was no least in part to the effect of an increase in difference in the number of noradrenaline shear stress on the endothelium, caused by receptor sites, nor in the binding of nora- the hypoxia-induced increase in haematocrit drenaline to the receptors. However, the dis- for shear stress is known to stimulate NO parity between the iliac arteries of CH and N synthesis (113). As a consequence of in- rats only existed when the endothelium was creased NO synthesis, we propose that the present: when the endothelium was removed, dilator influence of any substance that the maximum response to noradrenaline was achieves its dilator effects in an NO-depend- similar in the arteries from the CH and N rats ent manner, such as adenosine, may be en- (111,112). hanced in chronic hypoxia. This would be An obvious possibility was that NO might expected to lead to an increase in the vasodi- be involved. We therefore tested the effect lator effect of acute hypoxia, just as our of blocking NO synthesis with L-NAME. results have demonstrated. On the other hand, Whereas L-NAME had no effect on the basal up-regulation of NO synthesis would also be tone of iliac arteries of N rats, it substantially expected to reduce the effects of the reflex increased that of CH rats. Furthermore, vasoconstrictor influences of acute hypoxia, whereas L-NAME produced only a small again, just as our results imply. increase in the maximum response to nora- drenaline in the iliac arteries from the N rats, Concluding remarks it substantially increased the maximum re- sponse of the arteries from the CH rats so From a sound foundation of knowledge that their maximum response became com- about the responses that can be evoked by parable to that of the N rats. By contrast, L- selective stimulation of carotid chemorecep- NAME had no significant effect on the nor- tors, we have been able to show that activa- adrenaline-dose response curves of endothe- tion of the defence areas by the carotid lium denuded iliac arteries from either CH or chemoreceptors and that elicitation of the N rats. This provided strong evidence that characteristic pattern of the alerting (defence) the basal synthesis of NO by the endotheli- response is an integral part of the full re- um is increased in the iliac arteries of CH sponse to acute systemic hypoxia. But we rats and suggested that NO was responsible have also shown how this response, as well for the impaired vasoconstrictor responses as the classical primary cardiovascular re- to noradrenaline (112). flex responses to carotid chemoreceptor stim- Having obtained this evidence in vitro it ulation of bradycardia and generalised vaso- was then important to establish whether a constriction and the cardiovascular changes similar effect could be produced in vivo. In that are secondary to chemoreceptor-induced fact, in experiments on mesenteric arterioles hyperventilation can be modified, or even

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 885 overcome, by the local effects of hypoxia on leaves many questions unanswered. To date, the central nervous system, heart and periph- the results suggest that within 3-4 weeks of eral tissues. It seems that these local effects chronic hypoxia, adaptations have taken place of respiratory depression mediated by the such that ventilation is tonically raised, but influence of hypoxia on central respiratory arterial pressure and heart rate are normal. neurones, bradycardia and peripheral vasodi- This alone suggests that the normal ability of latation generally become manifest in se- hyperventilation to induce tachycardia is vere, or longer periods of acute hypoxia, but impaired. In addition, when a further acute are more likely to predominate in small adult hypoxic challenge is superimposed upon the mammals and neonates: they have the poten- chronically hypoxic state, the normal ability tial to form a positive feedback loop that of chemoreceptor-induced stimulation to leads to death. cause tachycardia secondary to hyperventi- Our results suggest that adenosine plays lation and reflex vasoconstriction seems im- a major role in producing these local effects paired (105). These results may be explained, of tissue hypoxia and have shown some of at least in part, by down-regulation of car- the cellular mechanisms by which adenosine diac ß adrenoreceptors, as has been reported achieves these effects. In particular, our ob- in chronic hypoxia (114) and by up-regula- servations on the microcirculation of skel- tion of NO synthesis by the endothelium of etal muscle have demonstrated how adeno- the systemic vasculature, such that func- sine, acting in part in a NO-dependent man- tional responses to the sympathetic neu- ner, can overcome the vasoconstrictor influ- rotransmitter, noradrenaline, are reduced. A ences of chemoreceptor-induced activation question that is of particular interest is of sympathetic noradrenergic fibres and of whether the impairments in the functional circulating hormones on individual arteri- responses to peripheral chemoreceptor stim- oles and so can increase the homogeneity of ulation that occur in chronic hypoxia are the O2 supply through the capillary network. accompanied by increases or decreases in The sum total of experimental studies the cardiac vagal and sympathetic nerve ac- performed on chronic systemic hypoxia tivity that produces them.

References

1. Bernthal T (1938). Chemo-reflex control 4. Daly M De B & Scott MJ (1963). The sis of the primary cardiovascular reflex of vascular reactions through the carotid cardiovascular responses to stimulation effects of stimulation of the carotid body body. American Journal of Physiology, of the carotid body chemoreceptors in chemoreceptors in the dog. Journal of 121: 1-20. the dog. Journal of Physiology, 165: 179- Physiology, 162: 555-573. 2. Heymans CJ, Bouckaert JJ & 197. 9. Angell-James JE & Daly M De B (1978). Dautrebande L (1930). Sinus carotidien et 5. Daly M De B & Ungar A (1966). Compari- The effects of artificial lung inflation on reflexes respiratoires. II. Influences son of the reflex responses elicited by reflexly induced bradycardia associated respiratoires reflexes de l’acidose, de stimulation of the separately perfused ca- with apnoea in the dog. Journal of Physi- l’alcolose, de l’anhydride carbonique, de rotid and aortic body chemoreceptors in ology, 274: 349-366. l’ion hydrogene et de l’anoxemic. Sinus the dog. Journal of Physiology, 182: 379- 10. Lioy F, Hanna BD & Polosa C (1978). CO2- carotodiens et echanges respiratoires 403. dependent component of the neurogenic dans les poumons et au dela des pou- 6. Daly M De B & Scott MJ (1958). The vascular tone in the cat. Pflügers Archives, mons. Archives Internationales de Phar- effects of stimulation of the carotid body 374: 187-191. macodynamie et de Therapie, 39: 400- chemoreceptors on heart rate in the dog. 11. Fukuda YA, Sato A, Suzuki A & Trezebski 448. Journal of Physiology, 144: 148-166. A (1989). Autonomic nerve and cardiovas- 3. Bernthal T, Green W & Revzin AM (1951). 7. Daly M De B & Robinson BH (1968). An cular responses to changes in blood oxy- Role of carotid chemoreceptors in hypoxic analysis of the reflex systemic vasodilator gen and carbon dioxide levels in the rat. cardiac acceleration. Proceedings of the response elicited by lung inflation in the Journal of the Autonomic Nervous Sys- Society for Experimental Biology, 76: 121- dog. Journal of Physiology, 195: 387-406. tem, 28: 61-74. 124. 8. Daly M De B & Scott MJ (1962). An analy- 12. Hanna BD, Lioy F & Polosa C (1979). The

Braz J Med Biol Res 31(7) 1998 886 J.M. Marshall

effect of cold blockade of the medullary Journal of Physiology, 175: 193-202. organisation of central pathways regulat- chemoreceptors on the CO2 modulation 24. Scott MJ (1966). Reflex effects of carotid ing the cardiovascular system. Physiologi- of vascular tone and heart rate. Canadian body chemoreceptor stimulation on the cal Reviews, 74: 323-364. Journal of Physiology and Pharmacology, heart rate of the rabbit. Australian Journal 37. Marshall JM (1995). Cardiovascular 57: 461-468. of Experimental Biology and Medical Sci- changes associated with behavioural alert- 13. Daly M De B (1986). Interactions between ence, 44: 393-404. ing. In: Jordan D & Marshall JM (Editors), respiration and circulation. In: Chemiack 25. Daly M De B & Taton A (1979). Interac- Studies in Physiology: Cardiovascular NS & Widdicombe JG (Editors), Handbook tions of cardiorespiratory reflexes elicited Regulation. Portland Press, London, UK. of Physiology, The Respiratory System, from the carotid bodies and upper airway 38. Timms RJ (1981). A study of the amygda- Control of Breathing. Section 3, Vol. II. receptors in the conscious rabbit. Journal loid defence reaction showing the value American Journal of the Physiological So- of Physiology, 291: 34P (Abstract). of Althesin anaesthesia in studies of the ciety, Washington, DC, 529-594. 26. Marshall JM (1987). Analysis of cardio- function of the forebrain in the cats. 14. Gilbey MP, Jordan D, Richter DW & Spyer vascular responses evoked following Pflügers Archives, 391: 49-56. KM (1984). Synaptic mechanisms in- changes in peripheral chemoreceptor ac- 39. Hilton SM & Marshall JM (1982). The pat- volved in the inspiratory modulation of tivity in the rat. Journal of Physiology, 394: tern of cardiovascular response to carotid vagal cardio-inhibitory neurones in the cat. 393-414. chemoreceptor stimulation in the cat. Journal of Physiology, 356: 65-78. 27. Daly M De B, Korner PI, Angell-James JE Journal of Physiology, 326: 495-513. 15. McAllen RM & Spyer KM (1978). Two & Oliver JR (1978). Cardiovascular-respi- 40. Marshall JM (1987). Analysis of cardio- types of vagal preganglionic motoneu- ratory reflex interactions between carotid vascular responses evoked following rones projecting to the heart and lungs. bodies and upper airway receptors in the changes in peripheral chemoreceptor ac- Journal of Physiology, 282: 353-364. monkey. American Journal of Physiology, tivity in the rat. Journal of Physiology, 394: 16. Katona PG, Debowsky K, Czachurski J & 234: H293-H299. 393-414. Seller H (1989). Chemoreceptor stimula- 28. Daly M De B, Korner PI, Angell-James JE 41. Marshall JM (1986). The role of the gly- tion on sympathetic activity: dependence & Oliver JR (1978). Cardiovascular and cine sensitive area of the ventral medulla on respiratory phase. American Journal of respiratory effects of carotid body stimu- in cardiovascular responses to carotid Physiology, 257: R1027-R1033. lation in the monkey. Clinical and Experi- chemoreceptor and peripheral nerve stim- 17. Lipski J, Coote JH & Trzebski A (1977). mental Pharmacology and Physiology, 5: ulation. Pflügers Archives, 406: 225-231. Temporal patterns of antidromic invasion 511-524. 42. Franchini KG & Kreiger EM (1993). Cardio- latencies of sympathetic preganglionic 29. Jain SK, Subramanian S, Julka DB & Guz vascular responses of conscious rats to neurones related to central inspiratory ac- A (1972). Search for evidence of lung carotid body chemoreceptor stimulation tivity and pulmonary stretch receptor re- chemo-reflexes in man: study of respira- by intravenous KCN. Journal of the Auto- flex. Brain Research, 135: 162-166. tory and circulatory effect of phenyl nomic Nervous System, 42: 63-69. 18. Gilbey MP, Numao Y & Spyer KM (1986). dignanide and lobeline. Clinical Science 43. Marshall JM & Metcalfe JD (1988). Car- Discharge patterns of cervical sympa- 42: 163-177. diovascular changes associated with aug- thetic preganglionic neurones related to 30. Bizzi EA, Libretti A, Malliani A & Zanchetti mented breaths in normoxia and hypoxia central respiratory drive in the rat. Journal A (1961). Reflex chemoreceptive excita- in the rat. Journal of Physiology, 400: 15- of Physiology, 378: 253-265. tion of diencephalic sham rage behaviour. 27. 19. Marshall JM (1994). Peripheral chemore- American Journal of Physiology, 200: 923- 44. Fletcher EC, Lesske J, Qian W, Miller CC ceptors and cardiovascular regulation. 926. & Unger T (1992). Repetitive episodic hy- Physiological Reviews, 74: 543-594. 31. Abrahams VC, Hilton SM & Zbrozyna AW poxia causes diurnal elevation of blood 20. Angell-James JE & Daly M De B (1969). (1960). Active muscle vasodilation pro- pressure in rats. Hypertension, 19: 555- Cardiovascular responses in apnoeic as- duced by stimulation in the brain stem. Its 561. phyxia: role of arterial chemoreceptors significance in the defence reaction. Jour- 45. Trzebski A (1992). Arterial chemoreceptor and the modification of their effects by a nal of Physiology, 154: 491-513. reflex and hypertension. Hypertension, pulmonary vagal inflation reflex. Journal 32. Hilton SM & Zbrozyna AW (1963). Amyg- 19: 562-566. of Physiology, 201: 87-104. daloid region for defence reactions and its 46. Mancia G (1975). Influence of carotid 21. Karim F, Hainsworth R, Sofola OA & efferent pathway to the brain stem. Jour- baroreceptors on vascular responses to Wood LM (1980). Responses of the heart nal of Physiology, 165: 160-173. carotid chemoreceptor stimulation in the to stimulation of aortic body chemorecep- 33. Folkow B & Neil E (1971). Circulation. Ox- dog. Circulation Research, 36: 270-276. tors in the dog. Circulation Research, 46: ford University Press, London, UK. 47. Al-Obaidi M & Karim F (1992). Primary 77-83. 34. Hilton SM (1982). The defence arousal effects of carotid chemoreceptor stimula- 22. Daly M De B, Litherland AS & Wood KN system and its relevance for circulatory tion on gracilis muscle and renal blood (1983). The reflex effects of inflation of and respiratory control. Journal of Experi- flow and renal function in dogs. Journal of the lungs on heart rate and hindlimb vas- mental Biology, 100: 159-174. Physiology, 455: 73-88. cular resistance in the cat. IRCS Medical 35. Hilton SM, Marshall JM & Timms RJ 48. Heistad DD, Abboud FM, Mark AL & Science Library Compendium, 11: 859- (1983). Ventral medullary relay neurones Schmid PG (1974). Interaction of barore- 860. in the pathway from the defence areas of ceptor and chemoreceptor reflexes: modu- 23. McLeod RDM & Scott MJ (1964). The the cat and their effect on blood pressure. lation of the chemoreceptor reflex by heart rate responses to carotid body Journal of Physiology, 345: 149-166. changes in baroreceptor activity. Journal chemoreceptor stimulation in the cat. 36. Dampney RAL (1994). Functional of Clinical Investigation, 53: 1226-1236.

Braz J Med Biol Res 31(7) 1998 Chemoreceptors and systemic hypoxia 887

49. Wennergren G, Little T & Oberg B (1976). ences on the cardiovascular response to Role of adenosine in regulation of cere- Studies on the central integration of exci- graded levels of systemic hypoxia of the bral blood flow: effects of theophylline tatory chemoreceptor influences and in- accompanying hypocapnia in the rat. Jour- during normoxia and hypoxia. American hibitory baroreceptor and cardiac receptor nal of Physiology, 410: 381-394. Journal of Physiology, 253: H165-H175. influences. Acta Physiologica Scandinavi- 62. Marshall JM & Metcalfe JD (1988). Car- 75. Darnell RA (1985). Aminophylline reduces ca, 96: 1-18. diovascular changes associated with aug- hypoxia ventilatory depression: possible 50. Mifflin SW (1993). Inhibition of chemore- mented breaths in normoxia and hypoxia role of adenosine. Pediatric Research, 19: ceptor inputs to nucleus of tractus solita- in the rat. Journal of Physiology, 400: 15- 706-710. rius neurons during baroreceptor stimula- 27. 76. Berne RM (1986). Adenosine: an impor- tion. American Journal of Physiology, 265: 63. Thomas T, Elnazir B & Marshall JM (1994). tant physiological regulator. News in 14-20. Differentiation of the peripherally medi- Physiological Sciences, 1: 163-167. 51. Coote JH, Hilton SM & Perez-Gonzalez JF ated from the centrally mediated influ- 77. Froldi G & Belardinelli L (1990). Species- (1979). Inhibition of the baroreceptor re- ences of hypoxia in the rat during sys- dependent effects of adenosine on heart flex on stimulation in the brain stem de- temic hypoxia. Experimental Physiology, rate and atrio-ventricular nodal conduc- fence centre. Journal of Physiology, 288: 79: 809-822. tion. Circulation Research, 67: 960-978. 549-560. 64. Elnazir B, Marshall JM & Kumar P (1996). 78. Bockman EL & McKenzie JE (1983). Tis- 52. Spyer KM (1994). Central nervous mecha- Postnatal development of the pattern of sue adenosine content in active soleus nisms contributing to cardiovascular con- respiratory and cardiovascular response and gracilis muscle of cats. American trol. Annual Review Prize Lecture. Journal to systemic hypoxia in the piglet - the Journal of Physiology, 230: 1531-1537. of Physiology, 474: 1-19. roles of adenosine. Journal of Physiology, 79. Poucher SM, Nowell CG & Collis MG 53. Marshall JM (1981). Interaction between 492: 573-585. (1990). The role of adenosine in exercise the responses to stimulation of peripheral 65. Gootman PM, Gootman N, Buckley BJ, hyperaemia of the gracilis muscle in chemoreceptors and baroreceptors: the Peterson BJ, Steele AM, Sica AL & anaesthetised cats. Journal of Physiolo- importance of chemoreceptor activation Gandhi MR (1990). Effects of hypoxia in gy, 427: 19-29. of the defence areas. Journal of the Auto- developing swine. In: Acker H, Trzebski A 80. Coney A & Marshall JM (1995). The role nomic Nervous System, 3: 389-400. & O’Regan ARG (Editors), Chemorecep- of adenosine in cerebral cortical vasodila- 54. Sun MK & Reis DJ (1995). Decerebration tors and Chemoreceptor Reflexes. Ple- tation induced by acute systemic hypoxia does not alter hypoxic sympathoexcitato- num Press, New York, 155-163. in normoxic and chronically hypoxic rats ry responses in rats. Journal of the Auto- 66. Naeye RL (1980). Sudden infant death. under anaesthesia. Journal of Physiology, nomic Nervous System, 53: 77-81. Scientific American, 242: 56-62. 487: P174-P175. 55. Marshall JM & Metcalfe JD (1988). Analy- 67. Thomas T & Marshall JM (1995). A study 81. Coney A & Marshall JM (1996). Effect of sis of the cardiovascular changes induced on rats of the effects of chronic hypoxia ZM241385 on the responses to acute sys- in the rat by graded levels of systemic from birth on respiratory and cardiovascu- temic hypoxia and adenosine in the cere- hypoxia. Journal of Physiology, 407: 383- lar responses evoked by acute hypoxia. bral-cortex of the anaesthetised rat. Jour- 403. Journal of Physiology, 487: 513-525. nal of Physiology, 494P: P105-P106. 56. Marshall JM & Metcalfe JD (1989). Analy- 68. Dempsey JA & Foster HV (1982). Media- 82. Bryan PT & Marshall JM (1996). The role sis of factors that contribute to cardiovas- tion of ventilation adaptations. Physiologi- of adenosine A1 receptors in mediating cular changes induced in the cat by graded cal Reviews, 62: 262-346. hypoxia-induced vasodilatation in skeletal levels of systemic hypoxia. Journal of 69. Mullhorn DE, Eldridge FL, Kiley JP & muscle of the anaesthetised rat. Journal Physiology, 412: 429-448. Waldrop TG (1984). Prolonged inhibition of Physiology, 492P: P6 (Abstract). 57. Marshall JM & Metcalfe JD (1990). Ef- of respiration following acute hypoxia in 83. Bryan PT & Marshall JM (1996). The role fects of systemic hypoxia on the distribu- glomectomised cats. Respiration Physiol- of adenosine A2A receptors in vasodilata- tion of cardiac output in the rat. Journal of ogy, 57: 331-340. tion induced in skeletal muscle of the rat Physiology, 426: 335-353. 70. Easton PA & Anthionsen NR (1988). Ven- by systemic hypoxia. Journal of Physiolo- 58. Louwerse AM & Marshall JM (1996). Be- tilatory response to sustained hypoxia af- gy, 494P: P106 (Abstract). havioural and cardiovascular responses ter pretreatment with aminophylline. Jour- 84. Olsson RA & Pearson JD (1990). Cardio- evoked by hypoxia in the unanaesthetised nal of Applied Physiology, 64: 1445-1450. vascular purinoceptors. Physiological Re- rat. Journal of Physiology, 494P: P103- 71. Rowell LR & Blackmon JR (1986). Lack of views, 70: 761-845. P104. sympathetic vasoconstriction in hypox- 85. Mian R & Marshall JM (1991). Responses 59. Neylon M & Marshall JM (1991). The role emic humans at rest. American Journal of observed in individual arterioles and of adenosine in the respiratory and cardio- Physiology, 251: H562-H570. venules of rat skeletal muscle during sys- vascular response to systemic hypoxia in 72. Black JE & Roddie IC (1958). The mech- temic hypoxia. Journal of Physiology, 436: the rat. Journal of Physiology, 440: 529- anism of the changes in forearm vascular 485-497. 545. resistance during hypoxia. Journal of 86. Mian R & Marshall JM (1991). The roles 60. Thomas T & Marshall JM (1994). Interde- Physiology, 143: 226-235. of catecholamines in responses evoked pendence of respiratory and cardiovascu- 73. Winn HR, Rubio R & Berne RM (1981). in arterioles and venules of rat skeletal lar changes induced by systemic hypoxia Brain adenosine concentration during hy- muscle by systemic hypoxia. Journal of in the rat; the roles of adenosine. Journal poxia in rats. American Journal of Physiol- Physiology, 436: 499-510. of Physiology, 480: 627-636. ogy, 241: H235-H242. 87. Mian R & Marshall JM (1991). The role of 61. Marshall JM & Metcalfe JD (1989). Influ- 74. Morii S, Ngai AC, Ko K & Winn HR (1987). adenosine in dilator responses induced in

Braz J Med Biol Res 31(7) 1998 888 J.M. Marshall

+ arterioles and venules of rat skeletal sitive K channels to A1 receptors by G chronically hypoxic, adult rats. Journal of muscle in systemic hypoxia. Journal of proteins in rat ventricular myocytes. Physiology, 501: 439-447. Physiology, 493: 499-511. American Journal of Physiology, 259: 106. Mian R & Marshall JM (1996). The be- 88. Marshall JM, Lloyd J & Mian R (1993). H820-H826. haviour of muscle microcirculation in The influence of vasopressin on the arte- 97. Dart C & Standen NB (1993). Adenosine- chronically hypoxic rats: the role of aden- rioles and venules of skeletal muscle of activated potassium current in smooth osine. Journal of Physiology, 491: 489- the rat during systemic hypoxia. Journal muscle isolated from the pig coronary 498. of Physiology, 470: 473-484. artery. Journal of Physiology, 471: 767- 107. Mian R & Marshall JM (1995). The role of 89. Lioy F, Blinkhorn MT & Garneau C (1984). 786. adenosine in mediating vasodilatation in Regional haemodynamic effects of 98. Spruce AE, Standen NB & Stanfield PR mesenteric circulation of the rat in acute changes in PaCO2 in the vagotomised, (1987). Studies of the unitary properties and chronic hypoxia. Journal of Physiolo- sino-aortic de-afferented rat. Journal of of adenosine-5'-triphosphate-regulated gy, 489: 225-234. the Autonomic Nervous System, 12: 301- potassium channels of frog skeletal 108. Heistad DD, Abboud FM, Mark AL & 313. muscle. Journal of Physiology, 382: 213- Schmidt PG (1972). Impaired reflex vaso- 90. Mian R, Marshall JM & Kumar P (1990). 236. constriction in chronically hypoxemic pa- + Interactions between K and ß2 adrenore- 99. McCarron JG & Halpern W (1990). Potas- tients. Journal of Clinical Investigation, ceptors in determining muscle vasodilata- sium dilates cerebral arteries by two in- 51: 331-337. tion induced in the rat by systemic hy- dependent mechanisms. American Jour- 109. Doyle MP & Walker BR (1991). Alterna- poxia. Experimental Physiology, 75: 407- nal of Physiology, 259: H902-H908. tion of systemic vasoreactivity in chroni- 410. 100. Marshall JM, Thomas T & Turner L cally hypoxic rats. American Journal of 91. Marshall JM (1995). Skeletal muscle vas- (1993). A link between adenosine potas- Physiology, 260: R1114-R1122. culature and hypoxia. News in Physiologi- sium and ATP-sensitive potassium chan- 110. Mian R & Marshall JM (1994). Effect of cal Sciences, 10: 274-280. nels in muscle vasodilatation in the rat in chronic hypoxia on microcirculatory re- 92. Harrison DK, Kessler M & Knauff SK systemic hypoxia. Journal of Physiology, sponses evoked by noradrenaline. Inter- (1990). Regulation of capillary blood flow 472: 1-9. national Journal of Microcirculation: Clini- and oxygen supply in skeletal muscle in 101. Barrett-Jolley R, Comtois A, Davies NW, cal and Experimental, 14: 244. dogs during hypoxemia. Journal of Phys- Stanfield PR & Standen NB (1996). Ef- 111. Bartlett IS & Marshall JM (1995). Com- iology, 420: 431-446. fect of adenosine and intra-cellular GTP parison of the effects of acute hypoxia 93. Davies WR, Thomas T & Marshall JM on KATP channels of mammalian skeletal on iliac artery rings from normal and (1994). The effects of acute and chronic muscle. Journal of Membrane Biology, chronically hypoxic rats. British Journal systemic hypoxia upon muscle oxygen 152: 111-116. of Pharmacology, 116: 203P. supply in anaesthetised rats. Journal of 102. Bryan PT & Marshall JM (1996). Possible 112. Bartlett IS & Marshall JM (1996). Mecha- Physiology, 477: 34P-35P. mechanisms underlying the vasodilata- nisms underlying the depression of nora- 94. Skinner MR & Marshall JM (1996). Stud- tion induced in skeletal muscle of the rat drenaline-evoked contractions induced ies on the roles of ATP, adenosine and by stimulation of adenosine A1 recep- by chronic hypoxia in the rat iliac artery in nitric oxide in mediating muscle vasodila- tors. Journal of Physiology, 497P: 79P. vitro. Journal of Physiology, 491P: P24- tation induced in the rat by acute sys- 103. Bryan PT & Marshall JM (1997). The role P25. temic hypoxia. Journal of Physiology, of KATP channels in adenosine mediated 113. Mekumyants AM, Balasha SA & Khayutin 495: 553-560. vasodilatation in skeletal muscle of the VM (1995). Control of arterial lumen by 95. Paterson DJ, Estavillo JA & Nye PCG rat. Journal of Physiology, 504P: 204P. shear stress on endothelium. News in (1988). The effect of hypoxia on plasma 104. Olsen EB & Dempsey JA (1978). Rat as a Physiological Sciences, 10: 204-210. potassium concentration and the excita- model for human like ventilatory adapta- 114. Voelkel NF, Kegstrand L, Reeves JT, tion of arterial chemoreceptors in the cat. tion to chronic hypoxia. Journal of Ap- McMurty IF & Molinoff B (1981). Effects Quarterly Journal of Experimental Physi- plied Physiology, 44: 763-769. of hypoxia on density of beta-adrenergic ology, 73: 623-625. 105. Thomas T & Marshall JM (1997). The receptors. Journal of Applied Physiology, 96. Kirsch GE, Codina J, Birnbaumer L & roles of adenosine in regulating the res- 50: 363-366. Brown AM (1990). Coupling of ATP-sen- piratory and cardiovascular systems in

Braz J Med Biol Res 31(7) 1998