American Journal of Respiratory and Critical Care Medicine

The recent explosion in technologies allowing for concomitant examination of gene and protein expression has opened a new and exciting window of opportunity for the identification and characterization of mechanisms underlying the postnatal development of respiratory control. It seemed therefore appropriate to examine critically the current state of knowledge in one area of respiratory control that has fascinated respiratory physiologists for many years, namely the maturation of the mammalian hypoxic ventilatory response (HVR).

The clinical implications of HVR are immediately clear. Apnea and hypoxia are very common problems in the neonate and infant, and often result in the need for prolonged hospitalization and home cardiorespiratory monitoring. Epidemiological and physiological studies suggest a major role for hypoxia in initiating the chain of events that may ultimately lead to apparent life-threatening events (ALTE) and sudden infant death syndrome (SIDS). In addition, hypoxia is a frequent event in infants and children with chronic lung disease or sleep-disordered breathing. Delayed or abnormal development of defense mechanisms within critical brain structures underlying HVR could therefore lead to life-threatening episodes, underscoring the importance of understanding maturation of brainstem mechanisms from an organismal to molecular perspective (1). In this perspective paper, we will examine our current understanding of HVR maturation by exploring some components of the time domains of this response (e.g., early and sustained HVR), the neuroanatomical substrates for HVR in the developing central nervous system (CNS), and some of the intracellular signaling pathways that may underlie the neuronal discharge pattern characteristics of respiratory regions mediating HVR. Future research directions in this important area of neurorespiratory biology will be emphasized when appropriate.

The mammalian ventilatory response to hypoxia is a complex multiphasic phenomenon whose characteristics are highly dependent on the duration of the stimulus, its magnitude, and how the stimulus is presented (2). When an acute hypoxic stimulus is imposed on adult mammals, there is a biphasic ventilatory response as evidenced by an initial rise in ventilation followed by a later ventilatory reduction to a level above prehypoxic values, the latter being termed “ventilatory roll-off” (3).

The early ventilatory enhancement in response to acute hypoxia is comparatively reduced in neonates (4, 5). The magnitude of this response is obviously dependent on the severity of the hypoxic challenge, the level of maturity of the particular mammal at birth, the state of the animal (awake, rapid eye movement [REM], and non-REM sleep), and the postnatal age of the animal. In addition, when the hypoxic stimulus is sustained, ventilation may decrease in developing mammals to stable values that are markedly below those measured during normoxia, particularly during REM sleep, a prominent sleep state in the newborn (4, 5). In other words, the magnitude of the biphasic HVR is particularly prominent early in postnatal life.

It is clear that postnatal changes in carotid chemoreceptor function underlie significant components of the early HVR characteristics (6), and detailed review of this particular area is beyond the scope of this report. However, the neural afferent inputs originating from the carotid body chemoreceptors undergo central relaying and processing that will ultimately result in the efferent respiratory output, that is, the HVR. Considering that the central nervous system displays extensive changes in neuronal circuitry formation, neurotransmitter expression, and overall neural discharge properties, it is reasonable to anticipate that the temporal changes in carotid body hypoxia-sensing properties will be matched by concomitant changes in the central neural regions that receive afferent inputs from these organs. In this context, occurrence of a mismatch between the response properties of peripheral and central components of the HVR neural reflex arc could lead to significant respiratory instability, particularly during sleep states or state transitions, and thereby favor the occurrence of adverse events in vulnerable infants. Notwithstanding such possibilities, neuronal activation within multiple central nervous system structures in response to afferent neural traffic will, in turn, modulate the incoming peripheral chemoreceptor inputs to achieve adequate homeostatic adaptations. In recent years, it has become apparent that in response to acute hypoxia there is a complex recruitment of multiple brain structures, many of which would not be immediately perceived as subserving specific respiratory tasks. For example, a high percentage of neurons in the caudal hypothalamus are stimulated by hypoxia both in vivo and in vitro, and such stimulation is manifested by an increase in firing frequency and membrane depolarization. (7). Similarly, hypoxia-induced activation of cells contained within the red nucleus will elicit significant respiratory inhibition, and this inhibitory activity is particularly prominent in the developing animal, and has therefore been linked to mechanisms underlying the late or inhibitory phase of the biphasic HVR (8). In addition, deep nuclei within the cerebellum exert well-defined modulatory influences on the respiratory responses induced by increases in peripheral chemoreceptor afferent traffic (9). Of note, many other brain regions that modulate sleep state or metabolism have also demonstrated neuronal populations that display excitatory or inhibitory responses to hypoxia. Thus, the conceptual framework has now been expanded to incorporate an increasing number of neural regions that demonstrate functional roles in the HVR, and whose locations extend beyond the brainstem structures traditionally associated with the hypoxic response. At this point in time, very little is known about the postnatal development and function of these regions.

Notwithstanding the role played by nonbrainstem regions in HVR, the nucleus of the tractus solitarius (nTS), a longitudinal structure in the dorsomedial portion of the medulla oblongata, provides the first central relay to cardiorespiratory afferent inputs (10). Indeed, lesions of the commissural nucleus within the nTS will result in marked attenuation of HVR (11). At this level of the caudal brainstem are also located the nucleus ambiguus (nA), the area postrema (AP), the dorsal motor nucleus of the vagus (DMnX), and the hypoglossal nucleus (XII), all of which have roles in respiratory control and the HVR (Figure 1). However, most of the available information on the anatomical projections from vagal afferents to central brainstem structures has primarily been focused on adult mammals; as a result, very little is known about the postnatal development of these central brainstem structures. Davies and Kalia employed horseradish peroxidase neurohistochemistry to examine the transganglionic transport from carotid sinus afferents in the adult cat (12). Dense extraperikaryal labeling within the nucleus of the nTS, AP, and nA was found. The nTS labeling was bilateral, the ipsilateral side being more intense, and another study further showed that there are both central and peripheral sources of tyrosine hydroxylase immunoreactive nerve terminals in this region, suggesting that some of the centrally located neurons may exhibit intrinsic hypoxic chemosensitivity (13). In one of the few anatomical studies available on the ontogeny of vagal afferents, Rinaman and Levitt showed that in the rat embryo vagal motor neurons are first labeled at embryonic Day 13 (E13; rat gestation lasts 21 d), clustered within a region corresponding to the nA, and that by E18 the vagal nuclei appear remarkably mature (14). However, substantial postnatal changes in the DMnX become apparent during the first 10 d of life, indicating that brainstem nuclei undergo substantial developmental changes after birth (15). Unfortunately, the postnatal pattern of synaptic formation within carotid sinus nerve afferent projections to the nTS, XII, and nA has yet to be methodically explored.

During neuronal discharge, a signaling cascade of intracellular events leads to activation of immediate early genes, and to increased formation of c-fos protein in the nucleus. The discovery that c-fos induction can serve as a surrogate marker for neuronal activation (16) prompted some investigators to explore the pattern of c-fos expression in the caudal brainstem in response to an hypoxic stimulus. The overall findings have been thus far quite consistent among the few developmental studies available, and indicate a clear postnatal maturational process in the recruitment of nTS neurons during hypoxia (17, 18). These studies have further reinforced the above mentioned concept that the location of hypoxia-sensitive neurons extends to more rostral brain regions (19). The developmentally regulated increase in neuronal recruitment during acute hypoxia paralleled increases in the expression of synaptic proteins such as synaptophysin, and also the increases in dendritic length and area of influence occurring within brainstem nuclei such as the nTS and the nA (20).

From the relatively scarce published studies on the maturation of the neuroanatomical substrate of HVR, it becomes clear that this system undergoes the majority of its development postnatally, and could therefore be amenable to marked neuronal and functional plasticity, the latter being tightly constrained by the timing at which the perturbations leading to such plasticity are implemented (see below). An exciting avenue that has emerged in recent years incorporates the ability to generate transgenic animals in which the expression of a particular gene or protein has been either reduced or increased. Such transgenic approaches have thus far allowed for important gains in our understanding of HVR. For example, a pathway involving endothelin and the development of the neural crest has provided some insight into critical periods for HVR development. Indeed, Kuwaki and colleagues initially showed that mutant mice in which heterozygosity for a null mutation was generated for the gene corresponding to endothelin 1 demonstrated nearly absent HVR (21). Because endothelin receptors are critically involved in selected aspects of neural crest development (22), it is possible that disrupted neural crest maturation may induce developmentally regulated deficits in HVR. That hypothesis is currently being explored in greater detail in our laboratory (23, 24).

Another intriguing and clinically relevant area of research involves the effect of prenatal and postnatal insults such as those associated with maternal smoking or other recreational drugs on the neuroanatomical pathways mediating the HVR. There is clear evidence that gestational exposure to nicotine induces significant changes in neuronal maturation and circuitry formation, and that it may lead to abnormal hypoxic responses in infants (25). However, a detailed characterization of the mechanisms and consequences of such insults on the development and short-term and long-term adaptations of the HVR integrated neural network is still unavailable.

An additional level of complexity in the postnatal developmental characteristics of HVR occurs when attempts are made to delineate the neurotransmitters, membrane receptors, and downstream signaling pathways mediating HVR. A comprehensive review of this topic is clearly beyond the scope of this report, and therefore an example of a signaling pathway will be presented for illustration purposes.

It has become increasingly clear that glutamate and its receptors, particularly N-methyl-d-aspartate (NMDA) receptors, are critical for the early phase of HVR in the adult rat (26-28). Activation of NMDA receptors during hypoxia will recruit multiple intracellular pathways in nTS neurons that exhibit temporally distinct activation patterns. The timing and nature of the signaling cascade will modulate both the early and late HVR (Figure 2). For example, the NMDA receptor channel opening will activate particular serine-threonine kinases such as protein kinase C (PKC) and also a variety of tyrosine kinases. We have recently shown that activation of PKC-β and PKC-δ mediates important excitatory components of the early HVR whereas administration of selective PKC inhibitors will lead to significant attenuation of the HVR (29). Similarly, pharmacological modification of tyrosine kinase activity within the nTS will be associated with modulation of the HVR (30). It should be stressed that the calcium transients associated with NMDA receptor activation may lead to activation of neuronal nitric oxide synthase (nNOS) and NO production (31). The latter has been particularly implicated in minimization of the ventilatory reductions that occur during the late phase of HVR (31). Interestingly, NMDA receptor activation during hypoxia is also associated with regional release of platelet-derived growth factor (PDGF) within the nTS and activation of PDGF receptors expressed in nTS neurons (32). Activation of PDGF receptors has been shown to reduce NMDA receptor currents and to contribute to the inhibitory mechanisms involved in the ventilatory reduction of the late phase of HVR (Figure 2; 32). Thus, a simplistic model that incorporates both excitatory (a glutamatergic receptor—NMDA) and inhibitory (growth factor receptor—PDGF) components has been delineated in the adult animal.

What happens to this simplistic model in the newborn? One would anticipate reduced expression of the excitatory element and reciprocal enhancements in the inhibitory element. Indeed, the previously suggested NMDA glutamate receptor dependency of HVR is not as immediately apparent in the developing rat (18). Furthermore, examination of other glutamate receptors such as the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) glutamate receptors failed to reveal any substantial role of these receptors in HVR (33). These findings concur with recent data in humans, whereby NMDA receptor binding was found to increase in the brainstem between midgestation and early infancy to moderately high adult levels, while AMPA receptor binding fell over the same time period to low adult levels (34). In addition, when double-labeling immunohistochemical studies for the NMDA glutamate receptor and c-fos were conducted in the rat model, the nTS was only minimally recruited during acute hypoxia in the youngest animals (18, 33). Transgenic alteration of NMDA receptor expression was found to enhance the depression of respiratory activity in the nTS, while postnatal pharmacological NMDA receptor blockade failed to modify respiratory-related activity in the newborn rat (35). Thus, NMDA receptor activity in late fetal life seems to play an important role in the formation of respiratory networks within the nTS that will display excitatory properties when directly stimulated by electrical current. However, these networks are not functionally recruited by glutamate release during acute hypoxia, and their responsiveness to glutamate will increase with advancing postnatal age (18). Similarly, PKC modulation of HVR is absent in the young developing animal, and emerges concurrent with the appearance of NMDA dependency (36). Thus, maturation of the NMDA–PKC pathway primarily occurs postnatally, and suggests that perturbations of this developmental process may impose major changes in the response characteristics of the affected infant. A similar analogy can be derived from the reduced expression of nNOS and the enhanced expression of PDGF receptors found within brainstem neurons of younger animals. Indeed, these heterogeneous expression patterns are closely correlated with the higher degree of ventilatory depression that characteristically develops in the newborn infant (37, 38). Thus, lower nNOS activity in the nTS of developing pups was associated with the inability to sustain the early increases in ventilation induced by hypoxia (37). Similarly, the increased expression of PDGF receptors within the nTS was highly correlated with the magnitude of ventilatory depression found during the late phase of HVR (38). In summary, we propose a model whereby developmental changes in the expression and activity of its constitutive elements may account for some of the typical features of HVR in the young. This model will obviously require the incorporation of functional properties and developmental expression characteristics exhibited by many other known neurotransmitters for which a role in HVR has been ascribed.

The postnatal changes that occur in HVR-related pathways suggest that some of these pathways may exhibit activity-dependent plasticity. This is an exciting possibility because application of plasticity-inducing paradigms to at-risk infants and children could ultimately modify the respiratory behavior properties of such infants, and thereby eliminate, or at least significantly reduce, the risk for the occurrence of the adverse respiratory event being targeted. Initial evidence for the existence of significant plasticity within the HVR pathways in the developing mammal can be inferred from the following experiments: (a) application of episodic hypoxia to 2-d-old rat pups will lead to attenuation of the late HVR depression via enhancement of nNOS expression in the nTS (39); (b) chronic perinatal hyperoxia will induce long-lasting attenuations of the early phase of HVR in the albino rat (40); and (c) prenatal hypoxia impairs the postnatal development of the chemoafferent pathway in rats (41). Thus, it can safely be concluded at this point that the continuum of the various maturational processes underlying HVR at any given developmental stage is far from being understood. Instead, a patchy yet exciting picture has begun to emerge, whereby we may gain insights into not only how the pathways develop, but also which genes regulate such ontogenically regulated processes, and how they can be modified in at-risk individuals or populations. One intriguing example of human populations that may express different manifestations of a genetically determined reduction in HVR at different time points in development was suggested by Tishler and colleagues who identified epidemiological associations between SIDS, sleep-disordered breathing, dimensions of the upper airway, and HVR (42). Therefore, we clearly need to modify our conceptual approach to the developmental process of HVR from one of rigid deterministic development to one of dynamic interactions, such that we incorporate and further understand how system perturbations may modify the response characteristics at any given stage. Such mechanistic understanding may open opportunities for development of novel clinical and epidemiological interventions, and possibly the discovery of pharmacological agents, aiming to foster more physiological adaptations in the at-risk developing infant, and therefore reduce unwarranted morbidity and mortality.

1. Gaultier C. Development of the control of breathing: implications for sleep-related breathing disorders in infants. Sleep 2000;23(Suppl 4): S136–S139.
2. Powell FL, Milsom WK, Mitchell GSTime domains of the hypoxic ventilatory response. Respir Physiol1121998123134
3. Vizek M, Pickett CK, Weil JVBiphasic ventilatory response of adult cats to sustained hypoxia has central origin. J Appl Physiol63198716581664
4. Eden GJ, Hanson MAMaturation of the respiratory response to acute hypoxia in the newborn rat. J Physiol (London)392198719
5. Rigatto H, Brady JP, de la Torre Verduzco R. Chemoreceptor reflexes in preterm infants: I. The effect of gestational and postnatal age on the ventilatory response to inhalation of 100% and 15% oxygen. Pediatrics 1975;55:604–613.
6. Bamford OS, Sterni LM, Wasicko MJ, Montrose MH, Carroll JLPostnatal maturation of carotid body and type I cell chemoreception in the rat. Am J Physiol2761999L875L884
7. Horn EM, Dillon GH, Fan YP, Waldrop TGDevelopmental aspects and mechanisms of rat caudal hypothalamic neuronal responses to hypoxia. J Neurophysiol81199919491959
8. Waites BA, Ackland GL, Noble R, Hanson MARed nucleus lesions abolish the biphasic respiratory response to isocapnic hypoxia in decerebrate young rabbits. J Physiol (London)4951996217225
9. Xu F, Frazier DTModulation of respiratory motor output by cerebellar deep nuclei in the rat. J Appl Physiol8920009961004
10. Finley JCW, Katz DMThe central organization of carotid body afferent projections to the brainstem of the rat. Brain Res5721992108116
11. Housley GD, Sinclair JDLocalization of kainic acid lesions of neurons transmitting the carotid chemoreceptor stimulus for respiration in the rat. J Physiol (London)406198899114
12. Davies RO, Kalia MCarotid sinus nerve projections to the brain stem in the cat. Brain Res Bull61981531541
13. Massari VJ, Shirahata M, Johnson TA, Gatti PJCarotid sinus nerve terminals which are tyrosine hydroxylase immunoreactive are found in the commissural nucleus of the tractus solitarius. J Neurocytol251996197208
14. Rinaman L, Levitt PEstablishment of vagal sensorimotor circuits during fetal development in rats. J Neurobiol241993641659
15. Kalia MEarly ontogeny of the vagus nerve: an analysis of the medulla oblongata and cervical spinal cord of the postnatal rat. Neurochem Int201992119128
16. Morgan JI, Curran TRole of ion flux in the control of c-fos expression, Nature3221986552555
17. White LD, Lawson EE, Millhorn DEOntogeny of the O2-sensitive pathway in medulla oblongata of postnatal rat. Respir Physiol981994123135
18. Ohtake PJ, Simakajornboon N, Fehniger MD, Xue YD, Gozal DNMDA receptor expression in the nucleus tractus solitarii and maturation of hypoxic ventilatory response in the rat. Am J Respir Crit Care Med162200011401147
19. Horn EM, Kramer JM, Waldrop TGDevelopment of hypoxia-induced fos expression in rat caudal hypothalamic neurons. Neuroscience992000711720
20. Rao H, Pio J, Kessler JPPostnatal development of synaptophysin immunoreactivity in the rat nucleus tractus solitarii and caudal ventrolateral medulla. Brain Res Dev Brain Res1121999281285
21. Kuwaki T, Cao WH, Kurihara Y, Kurihara H, Ling GY, Onodera M, Ju KH, Yazaki Y, Kumada MImpaired ventilatory responses to hypoxia and hypercapnia in mutant mice deficient in endothelin-1. Am J Physiol2701996R1279R1286
22. Clouthier DE, Hosoda K, Richardson JA, Williams SC, Yanagisawa H, Kuwaki T, Kumada M, Hammer RE, Yanagisawa MCranial and cardiac neural crest defects in endothelin-a receptor-deficient mice. Development1251998813824
23. Dauger S, Renolleau S, Vardon G, Nepote V, Mas C, Simonneau M, Gaultier C, Gallego JVentilatory responses to hypercapnia and hypoxia in Mash-1 heterozygous newborn and adult mice. Pediatr Res461999535542
24. Renolleau S, Dauger S, Vardon G, Levacher B, Simonneau M, Yanagisawa M, Gaultier C, Gallego JImpaired ventilatory responses to hypoxia in mice deficient in endothelin-converting-enzyme-1. Pediatr Res492001705712
25. Ueda Y, Stick SM, Hall G, Sly PDControl of breathing in infants born to smoking mothers. J Pediatr1351999226232
26. Mizusawa A, Ogawa H, Yoshihiro K, Hida W, Kurosawa H, Okabe S, Takishima T, Shirato KIn vivo release of glutamate in nucleus tractus solitarii of the rat during hypoxia. J Physiol (London)47819945565
27. Ohtake PJ, Torres JE, Gozal YM, Graff GR, Gozal DNMDA receptors mediate cardiorespiratory responses to afferent peripheral chemoreceptor input in the conscious rat. J Appl Physiol841998853861
28. Gozal D, Xue YD, Simakajornboon NHypoxia induces c-fos protein expression in NMDA but not AMPA glutamate receptor labeled neurons within the nucleus tractus solitarii of the conscious rat. Neurosci Lett26219999396
29. Gozal E, Roussel AL, Holt GA, Gozal L, Gozal YM, Torres JE, Gozal DProtein kinase C modulation of the ventilatory response to hypoxia in the nucleus tractus solitarius of the conscious rat. J Appl Physiol84199819821990
30. Czapla MA, Simakajornboon N, Holt GA, Gozal DTyrosine kinase inhibitors within the dorsocaudal brainstem modulate the ventilatory response to hypoxia in the conscious rat. J Appl Physiol871999363369
31. Gozal D, Torres JE, Gozal YM, Littwin SMEffect of nitric oxide synthase inhibition on cardiorespiratory responses in the conscious rat. J Appl Physiol81199620682077
32. Gozal D, Simakajornboon N, Czapla MA, Xue YD, Gozal E, Vlasic V, Lasky JA, Liu JYPlatelet-derived growth factor β receptor activation modulates components of the hypoxic ventilatory response. J Neurochem742000310319
33. Whitney GM, Ohtake PJ, Xue YD, Simakajornboon N, Gozal DAMPA glutamate receptors and respiratory control in the developing rat: anatomical and pharmacological aspects. Am J Physiol2782000R520R528
34. Panigrahy A, Rosenberg PA, Assmann S, Foley EC, Kinney HCDifferential expression of glutamate receptor subtypes in human brainstem sites involved in perinatal hypoxia-ischemia. J Comp Neurol4272000196208
35. Poon CS, Zhou Z, Champagnat J. NMDA receptor activity in utero averts respiratory depression and anomalous long-term depression in newborn mice. J Neurosci 2000;20:RC73.
36. Bandla HPR, Simakajornboon N, Graff GR, Gozal DDevelopmental changes in the hypoxic ventilatory response after systemic protein kinase C (PKC) inhibition in the rat. Pediatr Res431998330A
37. Gozal D, Gozal E, Torres JE, Gozal YM, Nuckton TJ, Hornby PJNitric oxide modulates ventilatory responses to hypoxia in conscious developing rats. Am J Respir Crit Care Med155199717551762
38. Vlasic V, Simakajornboon N, Gozal E, Gozal D. PDGF β receptor expression in the dorsocaudal brainstem parallels hypoxic ventilatory depression in the developing rat. Pediatr Res 2001. (In press)
39. Gozal D, Gozal EEpisodic hypoxia potentiates the late hypoxic ventilatory response in the developing rat: putative role of neuronal nitric oxide synthase. Am J Physiol2761999R17R22
40. Ling L, Olson EB, Vidruk EH, Mitchell GSDevelopmental plasticity of the hypoxic ventilatory response. Respir Physiol1101997261268
41. Peyronnet J, Roux JC, Geloen A, Tang LQ, Pequignot JM, Lagercrantz H, Dalmaz YPrenatal hypoxia impairs the postnatal development of neural and functional chemoafferent pathway in rat. J Physiol5242000525537
42. Tishler PV, Redline S, Ferrette V, Hans MG, Altose MDThe association of sudden unexpected infant death with obstructive sleep apnea. Am J Respir Crit Care Med153199618571863
Correspondence and requests for reprints should be addressed to David Gozal, M.D., Kosair Children's Hospital Research Institute, University of Louisville, 570 S. Preston Street, Ste. 321, Louisville, KY 40202. E-mail:

Dr. Gozal is supported by the National Institutes of Health Grants HL-65270, HL-63912, and HL-66358, The Commonwealth of Kentucky Research Challenge Trust Fund, and the American Heart Association Grant AHA- 0050442N.


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record