Advancements in the neurocirculatory reflex response to hypoxia

Abstract

Hypoxia is a pivotal factor in the pathophysiology of various clinical conditions, including obstructive sleep apnea, which has a strong association with cardiovascular diseases like hypertension, posing significant health risks. Although the precise mechanisms linking hypoxemia-associated clinical conditions with hypertension remains incompletely understood, compelling evidence suggests that hypoxia induces plasticity of the neurocirculatory control system. Despite variations in experimental designs and the severity, frequency, and duration of hypoxia exposure, evidence from animal and human models consistently demonstrates the robust effects of hypoxemia in triggering reflex-mediated sympathetic activation. Both acute and chronic hypoxia alters neurocirculatory regulation and, in some circumstances, leads to sympathetic outflow and elevated blood pressures that persist beyond the hypoxic stimulus. Dysregulation of autonomic control could lead to adverse cardiovascular outcomes and increase the risk of developing hypertension.

INTRODUCTION

Physiological systems have evolved over time to reflexively respond to changes in arterial blood gases to maintain homeostasis when oxygen levels are reduced (i.e., hypoxia). However, hypoxia is a major characteristic in the pathophysiology of clinical conditions such as obstructive sleep apnea (OSA) and chronic obstructive pulmonary disease (COPD) (1). Although the pathogenesis of these conditions are distinct and multifaceted, there is increasing recognition that they often coexist and are strongly associated with cardiovascular diseases, including hypertension (2). For example, individuals with OSA exhibit a high prevalence of hypertension (∼50%), and those with hypertension, particularly resistant hypertension, are more likely to have concurrent OSA (∼80%) (3, 4). In addition, the coexistence of COPD and OSA, known as the COPD-OSA overlap syndrome, contributes to higher rates of pulmonary hypertension and is likely mediated by more severe hypoxemia (5). Given that cardiovascular disease is the leading global cause of death and that hypertension is the most preventable modifiable risk factor (6), understanding the integrated mechanisms that contribute to hypertension development in conditions featuring hypoxemia has significant public health implications.

The precise mechanisms linking hypertension to clinical conditions associated with hypoxemia remain incompletely understood; however, data from animal and human models indicate a significant role for exposure to hypoxia or intermittent hypoxia (IH) (7). Independent of known confounding factors, autonomic and cardiopulmonary disturbances related to IH appear to hold clinical implications for the development of hypertension development, especially in OSA (8, 9). Various integrated pathways likely contribute to the pathogenesis of hypertension including, altered sympathetic neural regulation mediated by changes in chemoreflex (10, 11) and baroreflex responsiveness (12–14), as well as altered transduction of sympathetic discharge to the vasculature (15, 16). Even in ostensibly healthy individuals, experimental paradigms inducing disturbances in arterial blood gases demonstrate notable alterations in sympathetic discharge and blood pressure regulation with both acute and chronic hypoxia. Therefore, gaining an understanding of neurocirculatory responses to hypoxia offers valuable insights into early mechanistic factors contributing to the development of hypertension.

The purpose of this review is to offer a thorough overview of the neurocirculatory mechanisms triggered by hypoxia and their effects on arterial pressures. Specifically, this review aims to 1) highlight how evidence from animal and human models has enhanced our understanding of the mechanistic factors contributing to the pathophysiology of hypoxemia-induced hypertension; 2) detail neural reflex systems that govern sympathetically mediated cardiovascular regulation and examine the clinical implications of impaired reflex activation; and 3) consider future research directions to further advance our fundamental understanding of neural regulation and its role in disease progression.

HYPOXIA-MEDIATED REFLEX ACTIVATION OF SYMPATHETIC ACTIVITY: THE CHEMOREFLEX

The autonomic control of the cardiovascular system is, in part, regulated by input from specialized chemoreceptor cells that respond to perturbations in blood gas homeostasis (i.e., oxygen, carbon dioxide, pH). Central chemoreceptors located in the medulla of the brainstem are particularly sensitive to fluctuations in hydrogen ion concentration, primarily driven by increases in the partial pressure of carbon dioxide (i.e., hypercapnia). On the other hand, the carotid bodies, situated at the bifurcation of the carotid arteries, serve as the primary peripheral chemoreceptors for sensing reductions in arterial partial pressure of oxygen (17).

Figure 1 illustrates key neural signaling pathways associated with carotid body-mediated sympathoexcitation. A more detailed review of the sympathetic, parasympathetic, and respiratory networks related to chemoreceptor activation is available elsewhere (18). Briefly, when hypoxemia is detected by oxygen sensors within the carotid bodies, the information is relayed and integrated at the nucleus tractus solitarius (NTS) in the medulla. The NTS then triggers an increase in ventilation and sympathetic outflow by stimulating the central respiratory pattern generator and the rostral ventrolateral medulla (RVLM), respectively (19). In addition, direct anatomical connection from the paraventricular nucleus (PVN) in the hypothalamus can modify the NTS response (20).

Figure 1.

Sympathetic neural signaling pathways activated by hypoxemia. Carotid bodies detect changes in arterial partial pressure of oxygen and reflexively elicit changes in sympathetic activity (description in text). CB, carotid body; CVLM, caudal ventral lateral medulla; IML, intermediolateral nucleus; NTS, nucleus of the solitary tract; PVN, paraventricular nucleus; RVLM, rostral ventral lateral medulla; SFO, subfornical organ.

Excitatory signals from the RVLM provide synaptic input to sympathetic preganglionic neurons in the intermediolateral nucleus (IML) of the spinal cord, though other areas of the brain have also been shown to project to the IML (21). Sympathetic preganglionic neurons synapse in the sympathetic chain or prevertebral ganglia, resulting in postganglionic nerve activation and sympathetic vasomotor outflow (22). The NTS also relays baroreflex-driven inputs to the caudal ventrolateral medulla (CVLM), which tonically inhibits sympathetic premotor neurons in the RVLM (Fig. 2) (22). However, hypoxia has been shown to activate PVN-projecting neurons in the CVLM, potentially influencing the respiratory coupling of sympathetic activity (19, 23).

Figure 2.

Cardiovagal and sympathetic baroreflex response to increased blood pressure. Arterial baroreceptors respond to deformation of the vessel walls with increase in arterial pressure (shown in the blood pressure recording) and relay afferent information to the brain stem (black solid lines). Nucleus of the solitary tract (NTS) projections synapse with excitatory (+) projections to the caudal ventral lateral medulla (CVLM) and the nucleus ambiguous (NA) and dorsal motor nuclei (DMV) of the vagus nerve. Parasympathetic activation decreases cardiac output via decreased heart rate. CVLM inhibits (−) sympathetic premotor neurons within the rostral ventral lateral medulla (RVLM) restraining postganglionic muscle sympathetic nerve activity (MSNA; shown in representative recording) and decreases total peripheral resistance. The coupling of parasympathetic activation and sympathetic inhibition lowers blood pressure back to homeostatic levels (gray dashed lines). IML, intermediolateral nucleus.

Overall, the chemoreflex represents the primary reflex mechanism governing ventilatory and autonomic responses to disturbances in blood gas levels.

Effects of IH on Carotid Body Chemosensory Function

Fletcher et al. (24) were the first to implicate carotid body chemoreceptors in facilitating IH-induced hypertension by demonstrating that carotid body denervation in rats prevented the increased arterial pressures elicited by IH, which has since been confirmed by others (25–29). Sensitization of the sympathetic chemoreflex arc occurs in both acute (30–34) and chronic (25, 26, 35, 36) IH paradigms. Furthermore, recordings of carotid body chemosensory discharge show that IH causes a unique form of sensory plasticity, manifesting as 1) increased basal carotid body discharge; 2) augmented chemosensory and sympathetic response to hypoxia and hypercapnia; and 3) a progressive increase in sensory activity following acute bouts of IH lasting up to an hour (i.e., sensory long-term facilitation) (27–29, 36–40). Thus, chemoreflex control of sympathetic activity is directly affected by IH-induced sensitization of the carotid body and/or altered carotid body inputs to the brainstem.

Increased sensitivity of carotid body chemoreceptors induced by IH appears to be driven by the upregulation of reactive oxygen species (ROS). Common cellular sources of hypoxia-mediated ROS production include nicotinamide adenine dinucleotide phosphate (NADPH) oxidase enzymes and leaky complexes along the electron transport chain in the mitochondria (27, 41). It has been proposed that ROS generated during the reoxygenation phases contribute to heightened sensitivity of the carotid bodies to IH. Carotid bodies exposed to IH show elevated markers of ROS (27, 40), and the administration of a superoxide anion scavenger (40) or antioxidant (28) prevents the increased sensitivity to hypoxia and sensory long-term facilitation. ROS may influence oxygen sensitive potassium channels (28, 42), or induce changes in carotid body calcium signaling (43), leading to augmented membrane depolarization, elevated intracellular calcium levels, neurotransmitter release, and potentiated afferent activity in response to hypoxia. However, the exact mechanism by which ROS alters carotid body function remains to be fully elucidated.

Modulation of carotid body chemosensory function by NADPH oxidase/ROS signaling is multifaceted and may involve both endothelin-1 (40) and angiotensin II (ANG II). Since the effects of endothelin-1 have not yet been replicated in humans, our discussion focuses on ANG II, a key hormone involved in arterial pressure regulation by influencing vascular tone and sympathetic activity (44). ANG II is a potent activator of NADPH oxidase and consequently upregulates ROS production in vascular tissues (45). Evidence supports the existence of a renin-independent, angiotensin-generating system within the carotid body (46) that acts via ANG II type 1 receptors (AT₁R) to increase afferent chemoreceptor neuron activity (47) in response to chronic IH (48). In situ experiments show that acute intermittent application of ANG II induces carotid body sensory long-term facilitation by AT₁R activation of NADPH oxidase/ROS signaling pathways (49), which is comparable with the pathways involved in chronic IH-induced carotid body sensory plasticity and augmented chemoreflex-mediated sympathetic activity (35, 50). Increased carotid body chemoreflex sensitivity with acute and chronic IH is prevented with pretreatment of the AT₁R antagonist losartan (35, 49, 50), further supporting the role of ANG II signaling through the AT₁R in altering chemoreflex control of sympathetic outflow. These data suggest that IH-mediated ANG II generation in the carotid body results in upregulation of NADPH oxidase/ROS signaling via AT₁R activation.

Effects of IH on Central Neural Processing

IH affects the central neural pathways responsible for carotid body sensory information processing and reflex activation of sympathetic responses. Plasticity of the sympathetic neural circuitry occurs within as little as 40 min of IH, and this unique form of IH-induced sympathetic long-term facilitation (sLTF) manifests as progressive increases in sympathetic outflow lasting at least an hour following acute hypoxic exposures (30–32). Higher brain center projections, notably the subfornical organ (SFO) of the third ventricle and the PVN, provide synaptic inputs to the RVLM (Fig. 1) and appear to play a role in the development of hypertension via sympathetic responses elicited by chronic IH (51–55). IH-exposed rats exhibit increased expression of Fos protein markers, an indirect index of neuronal activation, in the NTS, PVN, and the RVLM; thus, it has been speculated that IH-induced activation of neurons in the medullary and hypothalamic regions contributes to sympathetic neuroplasticity (34, 53, 56).

Recent evidence indicates that AT₁R signaling pathways in the carotid bodies and SFO neurons, which are synaptically connected upstream to the PVN, play crucial roles in in the sympathetic activation induced by IH (31). Kim et al. (31) demonstrated that 1) acute IH or intermittent injections of exogenous ANG II resembling IH patterns lead to sustained sympathoexcitation; 2) pretreatment with losartan prevented the elevated sympathetic activity; and 3) both the carotid bodies and SFO were necessary for facilitating the sympathetic effects of IH-induced ANG II pathway activation. Given that the SFO lacks a blood-brain barrier and harbors a high density of AT₁Rs, circulating ANG II can exert hypertensive effects by directly activating neural pathways (57). Thus, it is theorized that circulating ANG II binds to the AT₁R on the surface of the SFO initiating an increase in SFO neuron excitability and transmission of excitatory inputs to the PVN. Recent findings have shown increased neuronal activity directly within the PVN in response to acute IH (34, 58), although further research is required to delineate the precise mechanisms involved. In summary, the neural control of sympathetic activity triggered by IH-induced chemoreflex activation is partially mediated by ANG II binding to the AT₁Rs at the level of the carotid bodies and SFO.

Effects of IH on Sympathetic Outflow to the Periphery

Evidence shows that IH-mediated sympathoexcitation is in part facilitated by excitatory neurotransmitter action in the spinal cord (33). Pituitary adenylate cyclase activating polypeptide (PACAP) is an excitatory neurotransmitter that stimulates the release of catecholamines and is found throughout the sympathetic signaling pathway, including the carotid body, the RVLM, and the sympathetic preganglionic neurons. Data show that the sympathetic response to acute IH is mediated by PACAP receptor activation within the spinal cord to increase sympathetic outflow (33). Furthermore, the existence of a novel oxygen sensing mechanism directly within the spinal cord has been reported. Barioni et al. (59) provide evidence that spinal oxygen sensors respond to a hypoxic bout (1–2 min partial pressure O₂: 60 mmHg) via neuronal nitric oxide synthase 1 and NADPH oxidase mechanisms. Thus, carotid bodies and spinal oxygen sensors may simultaneously contribute to the hypoxia-mediated cardiorespiratory response, though further research is required to determine the clinical implications of this oxygen sensing pathway, particularly in response to IH.

Finally, seminal work by Lesske et al. (26) demonstrated that postganglionic sympathetic neurons were critical in mediating IH-induced hypertension, as chemical destruction of postganglionic sympathetic neurons in rats prevented the rise in arterial pressure following 30 days of IH. Although augmented pressures are not uniformly observed in acute IH interventions (32, 33), neuroplasticity promoting sLTF elicited by acute IH may be an important precursor to hypertension. Collectively, these findings show that acute IH increases sympathetic activity regardless of blood pressure changes, which appears mediated via PACAP receptor signaling pathways and possibly through spinal oxygen-sensing mechanisms and may be a possible mechanism causing hypoxia-induced sLTF.

Experimental Models of Hypoxic Delivery in Humans

Over the past few decades, paradigms causing varying degrees of hypoxemia have been used on healthy humans to study the relationship between changes in arterial oxygen tension and blood pressure. If designed appropriately to mimic the pathological condition (e.g., blood gas fluctuations in OSA), then such experimental models can be used to examine how IH contributes to hypertension development because the effects of coexisting morbidities can be managed and attenuated. Though in other circumstances, IH of milder severity, reduced frequency of IH, and/or longer episodes may have therapeutic potential and this has recently been reviewed elsewhere (60, 61). Laboratory techniques are used to manipulate the percentage of inspired oxygen to mimic the frequency and degree of arterial oxyhemoglobin desaturations observed in OSA (i.e., apnea-hypopnea index ≥ 15/h). When investigating hemodynamic changes to hypoxemia, research groups will often deliver either acute (i.e., minutes to hours) or chronic (i.e., days to weeks) IH exposures. The addition of hypercapnia to IH protocols is less consistent when investigating chemoreflex-mediated cardiovascular responses; however, the use of intermittent hypercapnic hypoxia protocols is increasing, and may better mimic the blood gas disturbances experienced during obstructive apneic events and appears to amplify autonomic and hemodynamic responses. Of note, not all patients with OSA experience significant hypercapnia levels and if hypocapnia is present at the onset of an apnea episode, it may result from hyperventilation induced by a preceding arousal. Nonetheless, combined hypoxia and hypercapnia demonstrate a synergistic effect on sympathetic activation (62, 63) and elicit long-lasting augmentation of sympathetic activity when applied either continuously (63, 64) or intermittently (65).

Although a standardized approach to IH delivery is lacking, there appears to be general agreement among experimental findings supporting the notion that IH elicits autonomic and vascular responses similar to those observed in OSA. Although not an exhaustive list, Table 1 highlights the wide variation in types of experimental IH models, in both animals and healthy humans, and summarizes key neurovascular outcomes. It is important to note that these models are limited in that other key features of OSA, such as hemodynamic adjustments resulting from large swings in intrathoracic pressure or cardiovascular responses to recurrent arousal/sleep fragmentation, are typically not examined.

Table 1.

Experimental models of IH in animals and humans

Experimental Models	Research Question	Subjects	Hypoxia Delivery Paradigm	Hypoxic Intensity and Duration	Neurovascular Outcomes
Animal
Acute IH
Dick et al. (30)	Plasticity of sympathetic nervous system	n = 20 Sprague-Dawley rats	45 s 8% O2, CO2 not controlled and decreased <2 mmHg during exposure; 5 min 100% O2 recovery	Intensity not reported; 10 exposures	Increased splanchnic and phrenic nerve activity 60-min postexposure
Farnham et al. (33)	Sympathoexcitation and spinal cord pathways	n = 110 (F = 6) Sprague-Dawley rats and mice	45 s 10% O2; 5 min room-air recovery	Intensity not reported; 10 exposures	Increased splanchnic sympathetic nerve activity
Chronic IH
Fletcher et al. (24)	Blood pressure with and without peripheral chemoreceptors	n = 52 Wistar rats	12 s 3–5% O2; 2 cycles/min	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir ∼70%; 6–8 h; 35 days	Increased MAP (CIH/CB intact) No change MAP (CIH/CB denervated)
Marcus et al. (35)	Role of angiotensin in resting and chemoreflex stimulated SNA	n = 34 Sprague-Dawley rats	4 min cycles: ∼75 s 10% O2 with N2; remaining time 21% O2	Intensity not reported; 12 h; 28 days	Increased lumbar SNA at baseline, during apnea, and isocapnic hypoxia
Human
Hyperacute IH
Jouett et al. (66)	Role of angiotensin in sympathetic and pressor response	n = 9 (F = 1)	20 s end-expiratory apnea with 100% N2; 40 s room-air recovery	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir 85–90%; 20 min	Increased MAP, MSNA BF
Acute IH
Tremblay et al. (67)	Hemodynamic, endothelial, and carotid baroreflex function following IH	n = 10	1-min hypoxia with 100% N2, CO2 maintain at baseline levels; 1-min room-air recovery	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir ∼80%; 6 h	Increased MAP, carotid baroreflex control of MAP reset to higher pressures
Stuckless et al. (68)	Sympathetic neurovascular transduction responses	n = 9	40 s $\text{P}{\text{ET}}_{{\text{O}}_2}$: 45 mmHg, $\text{P}{\text{ET}}_{{\text{CO}}_2}$: +6 mmHg above baseline; 20-s room-air recovery	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir ∼83%; 40 min	Systemic sympathetic transduction increased following IH, with similar regional forearm sympathetic transduction between IH and sham.
Ott et al. (69)	Characterize sympathetic neural firing patterns	n = 17 (F = 6)	13% O2, CO2 maintain at baseline levels; room-air recovery; 15 cycles	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir ∼92%; 30 min	Increased DBP, MSNA BF, AP frequency, AP clusters
Shafer et al. (70)	Sympathetic and pressor response to handgrip and muscle metaboreflex activation	n = 13	40 s $\text{P}{\text{ET}}_{{\text{O}}_2}$: 60 mmHg, $\text{P}{\text{ET}}_{{\text{CO}}_2}$: +4 mmHg above baseline; 20-s room-air recovery	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir <90%; 40 min	Increased MAP, MSNA BF
Chronic IH
Foster et al. (71)	Ventilatory, cardiovascular, and cerebral tissue O2 responses to hypoxia after short vs. long duration IH	n = 18	5 min 12% O2, CO2 maintained at baseline levels; 5 min room air recovery	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir ∼90%; 1 h; 12 days	Increased hypoxic chemosensitivity that was linearly related to changes in arterial pressure but not heart rate
Pialoux et al. (72)	Oxidative stress and acute hypoxic ventilatory response	n = 10	2 min $\text{P}{\text{ET}}_{{\text{O}}_2}$: 45 mmHg, $\text{P}{\text{ET}}_{{\text{CO}}_2}$ fluctuated normally; 2-min room-air recovery	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir <90%; 6 h; 4 days	Increased MAP, AHVR, Association between AHVR and DNA oxidation
Gilmartin et al. (73)	MSNA and vascular reactivity following 28 days of nocturnal CIH	n = 7	13% continuous O2 with 10–21 s O2 delivery and 150–240 s N2; every 3 min	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir ∼82%; 9 h; 28 nights	Increased MSNA and augmented forearm vascular resistance
Tamisier et al. (74)	Sympathetic and pressure response after 14 days nocturnal IH	n = 12 (F = 2)	13% O2, CO2 fluctuated normally; 30 cycles/h	$\text{S}{\text{p}}_{{\text{O}}_2}$ nadir ∼85%; 8 h; 14 nights	Increased MAP, MSNA BF Decreased baroreflex control of MSNA

AHVR, acute hypoxic ventilatory response; AP, action potential; CB, carotid body; CIH, chronic intermittent hypoxia; F, females; IH, intermittent hypoxia; MAP, mean arterial pressure; MSNA BF, muscle sympathetic nerve activity burst frequency; N₂, nitrogen; $\text{P}{\text{ET}}_{{\text{O}}_2}$, end tidal oxygen; $\text{P}{\text{ET}}_{{\text{CO}}_2}$, end tidal carbon dioxide; SNA, sympathetic nerve activity; $\text{S}{\text{p}}_{{\text{O}}_2}$, oxyhemoglobin saturation.

Autonomic and Cardiorespiratory Response to Chronic and Acute IH in Healthy Humans

In healthy humans, neural and cardiovascular plasticity has been well documented following chronic and acute bouts of IH (66–72, 74–78). Tamisier et al. were the first to explore the pathophysiological effects of nightly IH exposure in healthy participants using a chronic IH model. In several landmark studies, healthy men and women were exposed to 2–4 wk of nocturnal IH sequences mimicking the magnitude and frequency of oxygen desaturations observed in severe OSA (20–30, 10% desaturations/h; 8–9 h/night) (73, 74, 79). Participants exhibited potentiated ventilatory responses to hypoxia, thus demonstrating enhanced chemosensitivity during wakefulness and implicating the role of chronic IH in peripheral chemoreflex plasticity (79). Furthermore, chronic nocturnal IH resulted in sustained increases in ambulatory blood pressures and augmented forearm vascular resistance that were secondary to elevated daytime muscle sympathetic nerve activity (MSNA) (73, 74). Numerous investigations have confirmed that various IH paradigms increase hypoxic ventilatory responses in healthy humans (71, 72, 76, 80, 81), with a strong relation to increased MSNA (76). Taken together, these studies show that IH sensitizes the sympathetic peripheral chemoreflex arc in humans leading to elevated daytime blood pressures. Compelling work has elucidated the mechanistic pathways of IH-induced neurovascular plasticity in humans. Evidence shows that 4 days of chronic IH (6 h daily exposure, poikilocapnic, alternating 2 min normoxia and hypoxia, nadir $\text{P}{\text{ET}}_{{\text{O}}_2}$ = 45 mmHg) elicits increased acute hypoxic ventilatory response, elevated markers of oxidative stress, and augmented mean arterial pressures (72, 82). Furthermore, a single 6-h IH (alternating between 1 min isocapnic hypoxia, nadir $\text{P}{\text{ET}}_{{\text{O}}_2}$ = 45 mmHg and 1 min normoxia) exposure results in increased arterial pressures (75) and markers of oxidative stress (83), an effect blunted by losartan. Interestingly, AT₁R antagonism with losartan also prevents the increased MSNA and hypertension elicited by hyperacute IH (20 min, see Table 1) (66), whereas hyperacute IH markedly elevated MSNA indices, superoxide levels were unchanged (84). It is plausible that longer duration IH may be necessary to elicit measurable changes in ROS; however, the short half-life of superoxide may also complicate its measurement in the blood. Overall, these findings align with data from animal models suggesting that IH-induced upregulation of ROS production via AT₁R activation pathways is likely a key contributor to peripheral chemoreflex sensitization and sympathetic neural plasticity promoting hypertension.

Chemoreflex-Mediated Sympathetic Activation by Continuous Hypoxia and Hypercapnia

Chemoreflex activation elicits hyperventilation and increased sympathetic nerve activity to the peripheral blood vessels, with negative feedback from pulmonary and baroreceptor afferents providing inhibitory influences to restrain chemoreflex-mediated sympathetic and cardiovascular responses (62, 85–87). Early work by Somers et al. (62) demonstrated that in healthy humans, acute exposure (∼5 min) to isocapnic hypoxia (10% O₂) and hyperoxic hypercapnia (7% CO₂ in 93% O₂) independently elicits sympathoexcitation through peripheral and central chemoreceptor pathways, respectively. However, a combined hypercapnic hypoxic stimulus synergistically increases sympathetic outflow, resulting in a noticeable increase in arterial pressure (62, 85). Later investigations expanded on these findings to reveal that short duration (∼20 min) hypercapnic hypoxia causes augmented sympathetic activity that persists on asphyxia termination (64) and that hypoxemia, especially, is necessary to evoke this lasting sympathetic plasticity (88). Indeed, acute continuous isocapnic hypoxia elicits long-lasting sympathetic activation in healthy individuals, which prevails for at least 30 min after blood gases have returned to normal levels (88–90). Interestingly, hypercapnia provided a synergistic effect since 20 min of hypercapnic hypoxia produced sustained increases in MSNA during recovery, whereas 20 min of hypocapnic hypoxia did not (63). Thus, even a short, continuous hypoxic stimulus has a robust influence of chemoreflex activation on sympathetic and pressor responses so long as isocapnia or hypercapnia are maintained.

HYPOXIA-MEDIATED REFLEX ACTIVATION OF SYMPATHETIC ACTIVITY: THE BAROREFLEX

Sympathetic Baroreflex Signaling Arc

The baroreflex tightly governs beat-to-beat mean arterial pressure by adjusting efferent autonomic neural outflow to reflexively modify total peripheral resistance and cardiac output (Fig. 2) (91). In this review, we will focus on the arterial baroreflex mechanisms that regulate arterial pressure through sympathetic activation, although it is worth noting that the baroreflex also influences mean arterial pressure through parasympathetic neural pathways, which impact heart rate and cardiac output (92, 93). Mechanosensitive baroreceptors (possibly Piezo1/2) (94, 95) are primarily located in the aortic arch and carotid sinus, respond to vessel wall deformation, and relay afferent signals to central cardiovascular regulatory centers in the brainstem. Baroreceptor neurons terminate in the NTS, where afferent information is relayed to key central nuclei.

Neurons from the NTS form monosynaptic excitatory connections with inhibitory neurons within the caudal ventrolateral medulla (CVLM), which tonically inhibit sympathetic premotor neurons within the RVLM and restrain postganglionic sympathetic nerve activity during increases in arterial pressure (96). During diastole when arterial pressure falls, decreased baroreflex inhibition of the RVLM presympathetic neurons increases the likelihood of sympathetic bursts.

The baroreflex dynamically adjusts peripheral vascular sympathetic outflow, producing rhythmically entrained bursts of sympathetic activity during transient arterial pressure reductions and inhibited neural activity when arterial pressures rises (96). Although baroreceptor activation promptly reduces efferent sympathetic outflow in response to elevated arterial pressure, afferent activity declines within seconds as the baroreceptors adapts, and within minutes, the baroreflex “resets” to the new prevailing pressure (97).

Baroreflex/Chemoreflex Interactions and Implications for Disease

Both baroreceptor and chemoreceptor reflexes play a significant role in regulating the neural control of circulation and investigating how these opposing autonomic reflexes interact is critical for understanding the mechanisms governing sympathetic responses, especially in the face of physiological stressors (87). Hypoxia-induced activation of peripheral chemoreceptors leads to increased sympathetic activity, causing a rise in arterial pressure, which in turn triggers a compensatory response from the baroreflex to buffer the increase in pressure (98). Over time, chronic reflex activation may disrupt baroreceptor and chemoreceptor reflex interactions, potentially leading to abnormal sympathetic control and elevated arterial pressure.

In patients with OSA, heightened daytime sympathetic activity (99, 100) is likely mediated by increased chemoreflex sensitivity (10, 11) and altered baroreflex function. Although studies on baroreflex sensitivity have yielded inconsistent results (12–14, 101, 102), depressed baroreflex mechanisms have been widely reported. OSA severity likely plays a role in the degree of autonomic dysfunction (103–105) and impairments to cardiac parasympathetic tone may further contribute to cardiovascular morbidity (106, 107). Furthermore, successful treatment of OSA, either through continuous positive airway pressure (CPAP) therapy (103, 105, 108, 109) or weight loss intervention (110), show improvements to baroreflex function, autonomic balance, and blood pressure. Reduced baroreflex sensitivity has also been observed in patients with COPD (111–113). Although supplemental oxygen has been shown to enhance baroreflex gain and improve autonomic metrics (114, 115), our understanding of the underlying pathophysiological mechanisms contributing to abnormal autonomic function in COPD and COPD-OSA overlap syndrome remains incomplete (116). Yet, these findings collectively suggest that a reduction in hypoxic exposure may elicit favorable effects on neurocirculatory control.

Simulated obstructive apneic events in healthy humans show both reduced baroreceptor gain during inspiratory resistance and a rightward resetting of the baroreflex operating point with asphyxia (117), indicating that breathing events occurring in OSA acutely elicit baroreflex dysfunction, which could ultimately promote hypertension development. However, substantial evidence demonstrates that lasting sympathetic activation elicited by hypoxia is mediated by an upward resetting of the baroreflex without altering baroreflex sensitivity, per se (87–90, 118, 119). Therefore, long-term activation of chemoreflexes or other disease-related pathways such as inflammation and oxidative stress may be necessary before clinically meaningful changes to baroreflex function occur.

FUTURE DIRECTIONS

Significant advancements have been made in understanding the intricate mechanisms that govern blood pressure and how dysregulation of autonomic control can contribute to unfavorable cardiovascular outcomes. Animal and human models have provided robust evidence that exposure to bouts of hypoxia uniquely affect the responsiveness and plasticity of the sympathetic nervous system. In healthy humans, chemoreflex-mediated sympathetic activation is facilitated by an upward resetting of the baroreflex despite baroreflex sensitivity remaining unchanged. These experimental approaches have begun to elucidate the neurocirculatory response evoked by hypoxia whether intermittent or sustained; yet several questions remain.

Sympathetic Action Potential Recruitment and Governing Mechanisms with Hypoxia

Advancements in wavelet-based approaches that detect and extract sympathetic action potentials (APs) have begun to elucidate neural regulatory mechanisms governing sympathetic discharge, recruitment, and synchronization strategies of AP subpopulations (90, 120, 121). Whether increased sympathetic outflow is mediated by rate coding (increased AP firing frequency) or latent recruitment of larger sized AP clusters appears to be dependent on various factors, such as severity of the sympathetic stressor (90, 122, 123). Increased burst frequency within the integrated MSNA neurogram has been reported to be mediated by increased AP occurrence and AP incidence of within-burst firing during hypoxic apnea (124) and acute IH (69). Conversely, recent work does not show latent cluster recruitment to meaningfully facilitate sympathetic activation and plasticity with acute isocapnic hypoxia; rather sympathoexcitation appears to be promoted by a resetting of the baroreflex to permit an increase in AP firing frequency, a decrease in asynchronous APs (APs discharging in between sympathetic bursts), and a proportional shift toward larger sympathetic AP activity (90). Although more research is warranted, these findings provide compelling evidence that hypoxia-induced sympathetic activation is in part elicited by a rate coding response (Fig. 3), which over time could have significant implications for end-organ vasoconstriction and peripheral vascular resistance resulting from greater neurotransmitter release for a given hypoxic exposure. More work is also needed to interrogate the influence of biological sex and sex hormones on the sympathetic recruitment strategies, neural reflex control of circulation, and plasticity following chemoreflex activation. Although several studies have incorporated full representation in experimental designs examining sympathetic AP recruitment and governing mechanisms in young, healthy individuals (69, 122, 125), sex differences have not been addressed, likely due to limitations in statistical power (69).

Figure 3.

Representative tracing of sympathetic recruitment and discharge strategies during chemoreflex-mediated stress. Burst size is increased with hypoxia-mediated sympathetic activation (top). In this representative participant, the increase in sympathetic activity is mediated primarily through increased activity (rate coding) of medium-sized sympathetic action potential (AP) clusters, with limited contribution from latent cluster recruitment (bottom).

Finally, continued work is needed to clarify the physiological role of sympathetic AP activity on blood pressure regulation. Asynchronous sympathetic AP activity occurring during periods that were previously considered “neurally silent” has been reported in humans by Klassen et al. (126, 127), and Shafer et al. (90, 128), but our understanding of asynchronous sympathetic discharge and the biological significance of asynchronous AP activity remains in its infancy. Although the origins and regulation of asynchronous AP discharge in humans remains unclear, it has been theorized that sympathetic projections originating from supramedullary regions (e.g., PVN) may provide input to the spinal cord and influence postganglionic discharge patterns (126). Neurons from supramedullary areas have been shown to directly project to the IML in rats (21). Ganglionic neural pathways of asynchronous sympathetic AP activity also merit further investigation. Ganglionic blockage with trimethaphan produces an ordered derecruitment pattern of sympathetic APs, whereby inhibition of larger AP discharge precedes the blockade of medium AP cluster firing. However, in humans, a subpopulation of small AP clusters exhibiting asynchronous behavior remain, suggesting asynchronous AP activity may be governed by non-nicotinic mechanisms at the level of the paravertebral ganglia (126, 127). Identifying the sites of neuronal activity and synaptic mechanisms that regulate sympathetic discharge is key to understanding the mechanisms governing neurocirculatory control, both under homeostatic states and with hypoxia-mediated sympathetic activation.

Sympathetic Transduction and Blood Pressure Maintenance with Hypoxia

The ability of efferent sympathetic outflow to affect vascular tone (i.e., sympathetic vascular transduction) is critical for both short- and long-term blood pressure regulation; however, data regarding sympathetic transduction in OSA is limited, with evidence suggesting that transduction is either maintained (15) or blunted (16) in older (>50 yr) men and women with untreated OSA. Similarly, few investigations have provided mechanistic insight into sympathetic transduction and neurovascular regulation of blood pressure with chemoreflex activation in healthy humans. Stuckless et al. (68) reported that sympathetic transduction of diastolic blood pressure is increased following acute intermittent hypercapnic hypoxia (see Table 1) in young males, while evidence supporting alterations in sympathetic transduction with acute continuous hypoxia is equivocal. Sympathetic transduction has been shown to be increased during isocapnic hypoxia (20 min at 80% $\text{S}{\text{p}}_{{\text{O}}_2}$) (129), though not a universal finding (128), and blunted with poikilocapnic hypoxia (10 min at 80% $\text{S}{\text{p}}_{{\text{O}}_2}$) (130). Although sympathetic activity is augmented following acute continuous hypoxia, parallel changes in vascular resistance have not been observed (64, 130, 131), suggesting that the transduction of sympathetic activity may be attenuated during recovery. Recent evidence supports this hypothesis. After 20-min of isocapnic hypoxia, sympathetic transduction of mean arterial pressure was reduced during early normoxic recovery despite increases in total MSNA activity (128). However, diminished sympathetic transduction has not been observed following poikilocapnic hypoxia (130). With limited evidence, it is challenging to infer the extent to which sympathetic transduction is altered with chemoreflex activation and if acute or plastic changes to sympathetic transduction impact the risk or development of clinically meaningful increases in blood pressure.

Future Directions for Disease Management and Health Promotion

Evidence from animal and human models supports a mechanistic role of chronic exposure to intermittent bouts of hypoxia in the pathogenesis of OSA-induced hypertension (7). Considering that OSA has been identified as an independent risk factor for hypertension development (8, 132) and the global prevalence of OSA is on the rise (133), there is a need for therapeutic strategies that can prevent or mitigate potential negative health consequences of chronic IH and promote positive cardiovascular outcomes. Although CPAP is the gold standard treatment, adherences rates are persistently low (134). Low CPAP compliance may explain why evidence from large population-based studies find that CPAP, which should eliminate IH, does not reduce the rates of cardiovascular disease or adverse cardiovascular outcomes (135–137). Thus, adjunctive therapies (such as antioxidant or angiotensin receptor blockade) and/or lifestyle modification (such as exercise) may help elicit more robust health outcomes (138). Finally, there is a growing interest in the interaction between hypoxia signaling, blood pressure, OSA, and the circadian timing system, which may create novel opportunities to develop personalized treatments particularly in individuals less receptive to CPAP (139, 140).

Perspectives and Significance

Cardiovascular disease (CVD) is the leading cause of morbidity and premature mortality globally and in the United States, with an estimated 82.6 million Americans (1 in 3 adults) being affected. Hypertension is not only the most prevalent form of CVD (76.4 million adults), but it is also the most preventable modifiable risk factor (6). Estimates predict that ∼1.5 billion people will develop hypertension by 2025 (141), which has significant public health implications especially given the increasing prevalence of coexisting multimorbid cardiovascular conditions. In conditions where hypoxemia is a feature, such as OSA, the dysregulation of integrated neural pathways that govern cardiovascular control has significant implications for hypertension risk. Understanding the mechanistic effects of hypoxia on cardiovascular regulation is critical for identifying early hypertension risk factors and mitigating adverse cardiovascular health outcomes.

GRANTS

This work was in part supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health Grant T32 HL083808 (to B.M.S.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

B.M.S., C.R.W., and G.E.F. conceived and the paper; B.M.S. drafted manuscript; B.M.S., C.R.W., and G.E.F. edited and revised manuscript; B.M.S., C.R.W., and G.E.F. approved final version of manuscript.