CrossTalk opposing view: the hypoxic ventilatory response does not include a central, excitatory hypoxia sensing component

Combining state-of-the-art (opto)genetic, molecular, viral transfection, imaging and staining techniques, Gourine and coworkers have constructed a framework with astrocytes as brain interoceptors particularly focusing on a putative role as central respiratory CO2 and O2 sensors (Gourine et al. 2005; Gourine & Kasparov, 2011; Angelova et al. 2015; Gourine & Funk, 2017; Rajani et al. 2018). In the opinion of this author, this work does not provide conclusive evidence for their hypothesis of an involvement of astrocytes as central O2 sensors in the ventilatory response to hypoxia (HVR) especially in awake animals and humans. Several animal species show immediate hypoventilation, hypercapnia and substantial reduction or entire absence of their HVR after bilateral carotid body denervation (CBD). Partial or complete restoration of the HVR is observed in rat, cat, piglet, goat and pony, all with different rates. This results from up-regulation of aortic bodies, release from cortical inhibition, recruitment of accessory glomus tissue in the trunk, neuroplastic changes consisting of axon regeneration, building alternative circuitries and recruiting central (including astrocytic) O2 sensors that may be silent in carotid body-intact conditions (Tenny & Ou, 1977; Martin-Body et al. 1986; Olson et al. 1988; Roux et al. 2000; Teppema & Dahan, 2010; Mouradian et al. 2012). Surprisingly, Angelova et al. (2015) did not challenge CBD animals with acute hypoxia or NaCN infusion shortly (a few days) after surgery; after 10 weeks, hypoxic responses in these animals may thus be due to neuroplastic adaptations. In the conscious goat, isolated central hypoxia does not lead to a progressive increase in ventilation as occurs during chronic carotid body (CB) hypoxia (Weizhen et al. 1992). The central hypoxia-induced rise in ventilation in the awake dog depends on peripheral chemoreceptor integrity (Curran et al. 2000). Up to two decades after CB removal or carotid endarterectomy, humans show no HVR (Wade et al. 1970; Swanson et al. 1978; Honda et al. 1979; Dahan et al. 2007). Exceptionally, on a hypercapnic background, a small response develops that can be ascribed to an involvement of the aortic bodies (Swanson et al. 1978; Timmers et al. 2003) but not to central mechanisms (Honda et al. 1979). Gourine and Funk (2017) propose an important stimulatory role of mitochondrial ROS in the HVR. To support their hypothesis involving a role of electron flow in the mitochondria in the HVR, they refer to patients with a mutation in the gene encoding succinate dehydrogenase (SDHD). These patients, however, do indeed show increased mitochondrial ROS (Cerecer-Gil et al. 2010) but do not have an abnormally large HVR but rather one at the lower end of normal (Dahan et al. 2007). Carotid body type I cells from SDHD+/− mice show an unaltered response to hypoxia, and mitochondrial complex II is not involved in oxygen sensing in these mice (note that total SDHD gene knockout is lethal in these mice; Piruat et al. 2004). Utilizing barometric plethysmography, Gourine and coworkers tested their hypothesis in vivo. Even if air humidity and temperature in the plethysmograph are controlled, body temperature measured and the inspired CO2 concentration maintained constant, the pressure signal may be influenced by frequency and airway resistance (Enhorning et al. 1998). Percentages of oxygen in the inspired air as the independent variable (i.e. the chamber O2%) is not the same as the inspired (Fig. 5 in Angelova et al. 2015), and has little predictive value as to the actual stimulus, in this case, the . Using arbitrary or relative units is not meaningful and can even be misleading. In other words, what would be of interest is the following: (1) to show blood gases or at least oxygen saturation (e.g. by using a tail probe); (2) to employ a useful index of the HVR such as / or / (normalized to body weight; see also Olson et al. 1988; Morgan et al. 2014); and (3) precise control of the Useful quantitative comparisons between groups require exposure to equal stimulus levels that reach their final values at equal rates. The hypoxic challenges in Angelova et al. (2015) were poikilocapnic, which is a confounding factor, given the CO2 sensitivity of astrocytes (Gourine et al. 2005) and the known O2–CO2 interaction in the rat (Wilson & Teppema, 2016). Finally, Angelova et al. (2015) claim lower respiratory activity in PINK1-deficient mice, but failure of astrocytes to sense low oxygen in these mice does not lead to a reduced response to hypoxia (Fig. 7E in Angelova et al. 2015). In conclusion, the data of Gourine and coworkers convincingly show O2 sensitivity of brain astrocytes and provide details of the resulting stimulus-tranduction cascade, involving ROS, the spread of Ca2+ waves and a role of ATP. Concerning the role of Ca2+, could gap-junctional Ca2+ exchange between glial cells and retrotrapezoid (RTN) neurons excite the latter? And what is the effect of glial depolarization (previously qualified as ‘glial impairment’ by Holleran et al. 2000) and ATP on the release of H+ by glia? Whether the reported increase in respiratory activity in CBD animals (10 weeks after surgery) is due to O2 sensing by astrocytes remains to be determined, as is the case with an astrocytic role in the normal poikilocapnic (and consequently hypocapnic) and isocapnic HVR. The human poikilocapnic response to mild acute hypoxemia (saturation ∼80%) is biphasic, consisting of an initial increase in ventilation, followed by a secondary fall and a rise in cerebral blood flow (CBF; Steinback & Poulin, 2007, 2016). The modest rise in CB activity will induce little central depression other than that caused by a local fall in and may account entirely for the small sustained rise in ventilation. Consequently, astrocytes are unlikely to be involved in maintaining an appropriate minute ventilation. If a modest decrease in brainstem (which is smaller than in the arterial blood) would result in astrocytic release of ATP, then, apart from directly impacting RTN neurons, the lower would tend to reduce it. If, as claimed by Gourine et al. (2005) and Rajani et al. (2018), a prolonged astrocytic ATP release in the anaesthetized rat maintains phrenic activity in the depressing phase, why then is a similar release after CBD not able to augment respiratory activity (Gourine et al. 2005)? The data from the poikilocapnic studies in Angelova et al. (2015) were collected in the last 5 min of a 10 min lasting exposure, but the time courses of both stimuli and responses are not shown, so it is unclear to what extent these data include those from the depressing phase. In the laboratory setting the HVR is often measured isocapnically to quantitatively estimate hypoxic sensitivity and O2–CO2 interaction effects. The isocapnic HVR is also biphasic, and the secondary fall (HVD) is related to a rise in cerebral blood flow combined with central depression (Teppema & Dahan, 2010). During HVD ventilation reaches a level ∼30% above control, maintained by the carotid bodies that operate at lower gain but do not show a biphasic response (Teppema & Dahan, 2010). In anaesthesia, however, HVD is uncoupled from carotid body activity (see Teppema & Dahan, 2010) and may involve a different role of ATP, as suggested in the rat (Rajani et al. 2018). If the O2 sensitivity of astrocytes is of less relevance to the awake ventilatory response to mild hypoxia, what then is its physiological significance? In the event of local brain hypoxia or ischaemia, an increase in ventilation would be counterproductive: the resulting fall in would cause vasoconstriction and thus impede blood flow to the affected areas and other brain regions. A crucial role of astrocytes is sensing (and adapting) neural activity and metabolism, and translating this also into an adaptation of local vasomotor activity to defend the supply of oxygen and nutrients (Gordon et al. 2016; Mukandala et al. 2016). That O2 sensitivity is a general property of astrocytes, rather than a distinctive feature of those located in the ventral medullary surface may suggest that although the latter are located in close association with the RTN, their primary role is to guard the supply of oxygen and nutrients by other means than by stimulating ventilation. Readers are invited to give their views on this and the accompanying CrossTalk articles in this issue by submitting a brief (250 word) comment. Comments may be submitted up to 6 weeks after publication of the article, at which point the discussion will close and the CrossTalk authors will be invited to submit a ‘Last Word’. Please email your comment, including a title and a declaration of interest, to jphysiol@physoc.org. Comments will be moderated and accepted comments will be published online only as ‘supporting information’ to the original debate articles once discussion has closed. Luc J. Teppema received his MSci (1978) and PhD (1984) degrees from the Catholic University of Nijmegen, The Netherlands (currently Radboud University). As a staff member at the Department of Anaesthesiology, Leiden University Medical Centre, his research focuses on the effects of hypercapnia and hypoxia on the control of breathing and the effects of anaesthetics, analgesics and carbonic anhydrase inhibitors. Disclaimer: Supplementary materials have been peer-reviewed but not copyedited. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. None declared.