The Role of Hypoxia-Inducible Factors in Oxygen Sensing by the Carotid Body

Abstract

Chronic intermittent hypoxia (IH) associated with sleep-disordered breathing is an important cause of hypertension, which results from carotid body-mediated activation of the sympathetic nervous system. IH triggers increased levels of reactive oxygen species (ROS) in the carotid body, which induce increased synthesis and stability of hypoxia-inducible factor 1α (HIF-1α) and calpain-dependent degradation of HIF-2α. HIF-1 activates transcription of the Nox2 gene, encoding NADPH oxidase 2, which generates superoxide. Loss of HIF-2 activity leads to decreased transcription of the Sod2 gene, encoding manganese superoxide dismutase, which converts superoxide to hydrogen peroxide. Thus, IH disrupts the balance between HIF-1-dependent pro-oxidant and HIF-2-dependent anti-oxidant activities, and this loss of redox homeostasis underlies the pathogenesis of autonomic morbidities associated with IH.

Hypoxia-inducible factor 1 (HIF-1) is a transcriptional activator that functions as a master regulator of oxygen homeostasis in all metazoan species (Semenza 2009, 2010). HIF-1 is a heterodimeric protein that is composed of an O₂-regulated HIF-1α subunit and a constitutively expressed HIF-1β subunit (Wang and Semenza 1995; Wang et al. 1995). Proline and asparagine residues in HIF-1α are hydroxylated under aerobic conditions (Kaelin and Ratcliffe 2008), which provides a mechanism for the decreased HIF-1α protein stability and transactivation function that are observed in well-oxygenated, as compared to hypoxic cells (Jiang et al. 1997). Database searches for homologs of HIF-1α led to the discovery of HIF-2α (Ema et al. 1997; Flamme et al. 1997; Hogenesch et al. 1997; Tian et al. 1997), which is also subject to O₂-dependent regulation and heterodimerization with HIF-1β (Wiesener et al. 1998). Whereas HIF-1α is expressed by most nucleated cells of all metazoan species, HIF-2α is present only in the vertebrate lineage and is only expressed in a restricted number of cell types (Loenarz et al. 2011).

Homozygosity for a null (knockout) allele at the Hif1a locus encoding HIF-1α results in embryonic lethality at midgestation with defects in erythrocytosis, vascularization, and cardiogenesis (Iyer et al. 1998; Ryan et al. 1998; Yoon et al. 2006), demonstrating that all three components of the circulatory system are dependent upon HIF-1 for normal development. Mice that are heterozygous for the knockout allele develop normally but have impaired responses to hypoxic and ischemic stimuli (Yu et al. 1999; Shimoda et al. 2001, 2006; Cai et al. 2003, 2008; Li et al. 2006; Bosch-Marce et al. 2007; Whitman et al. 2008; Feinman et al. 2010; Zhang et al. 2010; Kannan et al. 2011; Keswani et al. 2011)

Perhaps the most remarkable phenotype of Hif1a^+/− mice was the demonstration that the carotid bodies (CBs) of these mice do not sense/respond to hypoxia, despite the fact that they are histologically normal, including glomus cell morphometry, and respond appropriately to other stimuli such as hypercarbia and cyanide (Kline et al. 2002; Peng et al. 2006). In contrast to the failure of the CBs from these mice to sense hypoxia, Hif1a^+/− mice manifest normal ventilatory responses to a hypoxic challenge, indicating that other chemosensors compensate for the loss of CB function in vivo. However, hypoxic ventilatory acclimatization responses are defective in Hif1a^+/− mice, suggesting that this is a CB-specific function (Kline et al. 2002; Peng et al. 2006).

The CB is also known to sense and respond to intermittent hypoxia (IH) (Prabhakar et al. 2005). Patients with sleep-disordered breathing (also known as sleep apnea) are subjected to chronic IH, which leads to increased sympathetic nerve activity, hypertension, and its sequelae (Somers et al. 1995; Nieto et al. 2000; Peppard et al. 2000; Shahar et al. 2001). Exposure of rodents to IH for several weeks results in activation of the sympathetic nervous system, leading to elevated plasma catecholamine levels and systemic hypertension (Fletcher et al. 1992; Bao et al. 1997; Lesske et al. 1997; Kumar et al. 2006; Dick et al. 2007; Prabhakar et al. 2007a). Remarkably, Hif1a^+/− mice are completely protected from the elevation of blood pressure and plasma catecholamines that are observed in wild type mice littermates after 10 days of IH (consisting of 15 s of hypoxia, followed by 5 min of normoxia, nine episodes per hour for 8 h per day) (Peng et al. 2006).

IH results in the generation of reactive oxygen species (ROS) in the CB, as reflected in the levels of thiobarbituric acid-reactive substances, and treatment of wild-type mice with the free radical scavenger MnTMPyP blocks the development of IH-induced hypertension (Prabhakar et al. 2007b). The induction of HIF-1α that is observed in the brains of mice exposed to IH is also blocked by radical scavengers, indicating that increased ROS levels are required for HIF-1 activation in response to IH (Peng et al. 2006). When Hif1a^+/− mice are subjected to IH, increased levels of ROS are not detected in the brain, indicating that HIF-1 activation is required for increased ROS in response to IH. These results suggested a feed-forward mechanism, in which the increased generation of ROS induces HIF-1 activity, which in turn leads to increased ROS generation. Studies in PC12 rat pheochromocytoma cells revealed a complex signal transduction pathway by which IH stimulates HIF-1α protein synthesis, stabilization, and transactivation involving: NADPH oxidase; Ca ²⁺ signaling via phospholipase Cγ, protein kinase C, and CaM kinase; prolyl hydroxylases; and mTOR (Yuan et al. 2005, 2008). A critical role for NADPH oxidase in mediating IH-induced oxidative damage has also been demonstrated in mice (Zhan et al. 2005; Peng et al. 2009; Khan et al. 2011).

Recent studies indicate that the transactivation of the Nox2 gene, which encodes NADPH oxidase 2, is critical for the increase in ROS associated with chronic IH (Yuan et al. 2011). Induction of Nox2 mRNA expression was demonstrated in the CB, cerebral cortex, and brainstem, but not in the cerebellum, of wild-type mice, whereas Nox2 mRNA expression was not induced in any of these tissues in Hif1a^+/− mice (Yuan et al. 2011). Analysis of the cerebellar tissues from wild-type mice exposed to IH revealed no activation (by phosphorylation) of phospholipase Cγ or mTOR and no induction of HIF-1α protein, indicating that the IH-induced feed-forward mechanism, which is active in CB, cortex, and brainstem, is not active in the cerebellum (Yuan et al. 2011). Further studies are required to determine whether the response to IH is localized to specific regions of the cortex and brainstem.

Whereas HIF-1α is expressed at low levels in the CB under normoxic conditions and is dramatically induced by IH, HIF-2α is expressed at high levels in the CB under normoxic condition and its levels are dramatically decreased in response to IH, due to calpain-dependent protein degradation (Nanduri et al. 2009). The decreased HIF-2α levels were associated with decreased expression of the Sod2 gene, which encodes superoxide dismutase 2, the enzyme that converts mitochondrial superoxide to hydrogen peroxide. Remarkably, systemic treatment of IH-exposed rats with the calpain inhibitor ALLM blocked HIF-2α degradation, restored SOD2 activity, and prevented oxidative stress and hypertension (Nanduri et al. 2009).

Based on the striking results obtained from Hif1a^+/− mice, CBs from Hif2a^+/− mice were analyzed. Whereas CBs from Hif1a^+/− mice did not respond to hypoxia (Kline et al. 2002), CBs from Hif2a^+/− mice manifested augmented sinus nerve activity in response to hypoxia, whereas responses to hypercarbia or cyanide were normal, as was CB histology including glomus cell morphometry (Peng et al. 2011). Hif2a^+/− mice manifested an augmented hypoxic ventilatory response and instability of breathing. Hif2a^+/− mice were also found to be hypertensive with elevated plasma norepinephrine levels even under normoxic conditions (Peng et al. 2011). Expression of Sod2 mRNA was significantly reduced in the CB of Hif2a^+/− mice. Remarkably, treatment of Hif2a^+/− mice with the superoxide scavenger MnTMPyP normalized the CB response to hypoxia, normalized the hypoxic ventilatory response and blood pressure, and corrected the abnormalities of breathing (Peng et al. 2011). Taken together, the increased oxidative stress associated with IH (as well as all of its downstream sequelae) results both from increased HIF-1 → Nox2 signaling and decreased HIF-2 → Sod2 signaling in response to IH. Thus, an imbalance between pro-oxidant and anti-oxidant activities underlies the pathogenesis of autonomic morbidities associated with IH. More broadly, these results suggest that the functional antagonism between HIF-1 and HIF-2 may play a fundamental role in redox regulation and the maintenance of cardiorespiratory homeostasis.