A sensible approach to making sense of oxygen sensing

The ability to sense and respond appropriately to low $P_{{\mathrm{O}}_2}$ (hypoxia) is critical for the survival of aerobic organisms. In mammals, the carotid body is the prototypic detector of blood $P_{{\mathrm{O}}_2}$ and maintains homeostasis in the cardiorespiratory system. The signalling mechanisms that mediate $P_{{\mathrm{O}}_2}$ chemotransduction have attracted much attention but still remain controversial. In a comprehensive and compelling study published in this issue of The Journal of Physiology, Gao et al. (2017) use a comparative gene‐profiling approach to identify signature features likely to play key roles in acute $P_{{\mathrm{O}}_2}$ sensing.

The carotid body, strategically located at the carotid bifurcation, is a richly vascularized organ consisting of innervated clusters of catecholaminergic glomus cells, i.e. the main chemoreceptive elements, enveloped by glial cell processes. Over the last ∼30 years, a consensus has emerged that acute hypoxia causes glomus cell depolarization via K⁺ channel inhibition, leading to voltage‐gated Ca²⁺ entry and neurotransmitter release. However, the search for the $P_{{\mathrm{O}}_2}$ sensor has long attracted much attention and controversy, though the ‘metabolic hypothesis’ that invokes a key role of components of the mitochondrial electron transport chain (ETC) has received considerable support (Buckler, 2015; Fernandez‐Aguera et al. 2015). Several recent studies using transgenic models (Fernandez‐Aguera et al. 2015) and transcriptome analysis of single mouse glomus cells (Zhou et al. 2016) have shed new light on potential signature features of acute O₂ sensing.

In this new study, Gao et al. (2017) used a subtractive strategy to identify unique genes relevant to O₂ sensing. They compared gene expression profiles among three catecholaminergic cell types that have a common embryonic lineage but display different $P_{{\mathrm{O}}_2}$ sensitivities. In particular, they examined carotid body glomus cells that are exquisitely O₂ sensitive, adrenal chromaffin cells that are weakly‐to‐moderately sensitive, and sympathetic neurons of the superior cervical ganglion (SCG) that are hypoxia insensitive. Using microarrays, real‐time quantitative PCR and immunocytochemical analyses of tissues/cells from adult mice, they compared expression patterns in the carotid body and adrenal gland relative to the SCG. Because these organs contain several cell types besides tyrosine hydroxylase (TH)‐positive sympathoadrenal cells, the authors included an elegant and crucial control. They validated potentially relevant genes using qPCR analysis of sorted, fluorescently labelled cells obtained from TH–green fluorescent protein transgenic mice.

In glomus cells, the authors confirmed high relative expression of certain genes, previously identified in the single cell transcriptome study (Zhou et al. 2016), including hypoxia inducible factor (HIF)‐2α and mitochondrial ETC subunits, namely cytochrome c oxidase subunit IV isoform 2 (Cox4i2) and NADH dehydrogenase (ubiquinone) 1 α subcomplex, 4‐like 2 (Ndufa4l2). A new finding was that another atypical ETC subunit, Cox8b, was also upregulated. Interestingly, these same genes were upregulated in adrenal chromaffin cells, and Cox4i2 and Ndufa4l2 are known to be HIF‐regulated genes during chronic hypoxia. Of particular interest, they observed that Phd3, a member of the prolylhydroxylase family that preferentially hydroxylates HIF‐2α and targets it for proteosomal degradation, was robustly downregulated in O₂‐sensitive cells. This finding, together with the downregulation of transporters/enzymes that regulate cytoplasmic α‐ketoglutarate, a cofactor for the hydroxylation reaction, provides a plausible explanation for the high basal levels of HIF‐2α in O₂‐sensitive sympathoadrenal cells.

The authors also examined genes involved in pyruvate metabolism and the TCA cycle. Their analysis revealed an induction of pyruvate carboxylase (Pcx) together with decreased expression of a pyruvate dehydrogenase subunit (Pdha1) in glomus cells. These data suggest a preferential use of pyruvate to generate oxaloacetate and replenish TCA intermediates. They also fit nicely with the presence of high levels of biotin (a cofactor for carboxylase reactions) and succinate in cells responding acutely to hypoxia, as well as the proposed accumulation of reduced quinone (QH₂) to generate increased ROS and pyridine nucleotides for signalling membrane ion channels (Fernandez‐Aguera et al. 2015).

The study also highlighted particular ion channels linked to acute O₂ sensing. TASK channel subunits (particularly TASK3) that comprise the O₂‐sensitive TASK1/3 background K⁺ channels in glomus cells (Buckler, 2015) were overexpressed in glomus and adrenal chromaffin cells. Also, T‐type (Cav3.2) Ca²⁺ channels, known to facilitate hypoxia sensitivity in both cell types, were overexpressed. Though Trpc5 was also overexpressed, its physiological role remains unknown. Surprisingly, Kcnj11 subunits (Kir6.2) of K_ATP channels that are under transcriptional control by HIF‐2α and regulate O₂ sensitivity in chromaffin cells (Salman et al. 2014) failed to show differential regulation among the three tissues. However, because rodent chromaffin cells are particularly O₂ sensitive in the neonatal period, and show markedly depressed sensitivity in the adult following innervation, a comparative gene expression profile of neonatal chromaffin cells should prove a useful complement to this study.

In summary, this influential paper by Gao et al. (2017) has provided novel insight into potential markers likely to convey O₂ sensitivity in chemoreceptor cells. It reveals the Phd3–HIF‐2α couple linked to upregulation of three atypical mitochondrial ETC subunits, Pcx, and a few ion channels as attractive candidates for conferring acute O₂ sensitivity. Of note, the gene for the putative ‘lactate receptor’ recently proposed as the carotid body O₂ sensor, Olfr78, was upregulated in glomus cells but markedly downregulated in adrenal chromaffin cells, questioning its significance as a major determinant of acute O₂ sensing. The study also serves as a reminder that the holy grail for acute $P_{{\mathrm{O}}_2}$ sensing may not reside exclusively in a single molecular entity, but rather in the cooperative interaction among several players that confer a ‘signature metabolic profile’. How these components, acting together, render glomus cell mitochondria exquisitely sensitive to acute hypoxia remains an exciting challenge for the future.

Additional information

Competing interests

The author declares no competing interests.

Funding

Funding to the author's laboratory was provided by the Canadian Institutes of Health Research and the Natural Sciences and Engineering Research Council of Canada.