Skip to main content

Molecular mechanisms of acute oxygen sensing.

Periodic Reporting for period 4 - OxygenSensing (Molecular mechanisms of acute oxygen sensing.)

Reporting period: 2020-05-01 to 2021-10-31

Oxygen (O2) is essential for life, however, the mechanisms whereby cells detect changes in O2 tension are not well understood. The lack of O2 (hypoxia) induces hyperventilation and sympathetic activation, which in seconds increase the uptake of O2 from the atmosphere and its distribution to the tissues in the organism. These responses are mediated by the carotid body (CB), which contains neuron-like O2-sensitive cells (glomus cells) that in response to hypoxia release transmitters to activate sensory nerve fibers impinging upon the brain respiratory and autonomic centers. The mechanisms underlying how the specialized glomus cells are able to “sense” rapid changes in O2 in the environment have remained elusive. The objective of the ERC “OxygenSensing” proposal was to unravel the molecular mechanisms of acute O2 sensing. We have reported a gene expression profile characteristic of acute O2 sensing cells. This includes atypical mitochondrial complex IV (MCIV) subunits, metabolic enzymes, hypoxia inducible factor 2 alpha (HIF2a) and several ion channel subtypes. We have generated several genetically modified mouse models with conditional ablation of genes coding for MCI (NDUFS2) and MCIV (COX8B) subunits as well as pyruvate carboxylase (PCX). In addition, we have used mouse with conditional ablation of Epas1 (coding HIF2a) and COX4I2 (coding a MCIV subunit isoform). The experimental data have allowed us to propose a complete and robust model of acute O2 sensing. It is based on the fact that CB glomus cells contain specialized mitochondria with a very active electron transport chain (ETC) but a cytochrome c oxidase that has a low apparent affinity for O2. As a consequence of these properties, hypoxia (within physiological ranges) produces a marked slowdown of the mitochondrial ETC, which leads to the accumulation of reduced ubiquinone and the production of NADH and reactive oxygen species (ROS) in MCI and MCIII. These signaling molecules modulate ion channels in the membrane that induce cell depolarization, Ca2+ influx and transmitter release. We have also advanced in the understanding of the properties of the CB neurogenic niche that helps to explain CB plasticity during exposure to chronic hypoxia. We have discovered a new population of CB “neuroblasts” that in few hours can divide and differentiate into mature O2-sensitive glomus cells. Finally, we have generated a genetic mouse model of CB developmental atrophy, which exhibits lack of responsiveness to hypoxia.
The main results achieved are:

a)Sub-project 1 in the application (Determining the role mitochondria, particularly the role of mitochondrial complex I, in acute O2 sensing by arterial chemoreceptors and other cells that respond acutely to hypoxia).
The objectives in this “sub-project”, which are the foundation of the general research proposal, have been practically completed. Coincident with the onset of the project, in October 2015, we published a model of acute oxygen sensing proposing a fundamental role of MCI in acute responses to hypoxia by arterial chemoreceptors cells in the CB and the adrenal medulla (AM) (Fernández-Agüera et al., Cell Metabolism, 2015). During the duration of the ERC grant we have made important advances in this sub-project and performed experiments on genetically modified mice with conditional ablation of the Ndufs2 gene, which encodes a protein forming the ubiquinone-binding site at MCI. Our data have served to postulate a comprehensive model of acute O2 sensing in which accumulation of reduced ubiquinone during hypoxia determines the slow down (or even reversion) of MCI electron flow and the production of ROS and accumulation of NADH. These molecules signal the membrane K+ channels to induce cell depolarization and release of transmitters, which activate the respiratory center to elicit the hypoxic ventilatory response (HVR). These data, which were by themselves a main goal of the project, was published in Cell Metabolism (Arias-Mayenco et al., Cell Metabolism, 2018).
As Hif2a is constitutively expressed in CB glomus cells (see Sub-project 2 below), we have studied whether this transcription factor has any role in acute O2-sensing. The experiments were done using conditional Epas1 (coding for HIF2a) knockout mice. We have found that inducible HIF2a-deficiency in adult mice results in abolition of the HVR as well as the cellular responses to hypoxia in CB glomus cells, although responsiveness to hypercapnia remains unaltered. The loss of HIF2a leads to a selective down-regulation of the atypical mitochondrial subunits (NDUFA4L2, COX4I2 and COX8B) in glomus cells. To clarify the role of these atypical mitochondrial subunits in acute oxygen sensing we have studied genetically modified mice with conditional ablation of the genes coding for ether NDUFA4L2 and COX4I2. Mice without NDUFS4L2 are normal but ablation of the Cox4i2 gene in glomus produces an abolition of the hypoxic ventilatory responses and glomus cell responsiveness to hypoxia. These data have completed the “mitochondria to membrane signaling model” of CB acute oxygen sensing in which a cytochrome c oxidase (MCIV) with low affinity for O2 acts as an O2 sensor and MCI as an effector and generator of the signaling molecules (ROS and NADH). The complete model of acute O2 sensing has been published in Science Signaling (Moreno-Dominguez et al., 2020) as well as in several review articles of our group. We have also studied mice deficient in pyruvate carboxylase (PCX) and showed that, as expected, the lack of this enzyme results in partial inhibition of the systemic (HVR) and cellular responsiveness to hypoxia (see Sub-project 2, below).

b) Sub-project 2 in the application (Elucidating the properties of mitochondrial metabolism and biogenesis in O2-sensitive cells).
In the context of this sub-project we have performed a comparative gene expression analysis of CB, AM and superior cervical ganglion (SCG) cells. These three cell types have a common embryological origin but different sensitivity to hypoxia, being the CB and AM oxygen sensitive (CB>AM) and SCG oxygen insensitive. We have found several genes that define a “metabolic signature profile” in acute oxygen sensing cells. These genes encode three atypical mitochondrial subunits (NDUF4L2, COX8B and COX4I2), HIF2a, pyruvate carboxylase (PCX), and several ion channel subunits (TASK1, TASK3, a1H T-type Ca2+ channel, and TRPC5) which are up-regulated, and PHD3 which is down-regulated (see Gao et al., Journal of Physiology, 2017). We have also found that biotin, a cofactor necessary for PCX activity, is highly concentrated in CB cells (Ortega-Saenz et al., Journal of Physiology, 2016). Although none of these genes may be absolutely required for acute O2-sensing, we think that, together, they generate the appropriate metabolic status to confer upon the chemoreceptor cells special sensitivity to changes in PO2. One of the most important consequences of the identification of the “genetic profile” characteristic of acutely O2-sensing cells, is that it allowed the functional role of these genes to be tested experimentally using genetically modified animal models (see Sub-project 1 above).
In addition to the genes mentioned in the preceding paragraph, one of the genes more highly expressed in the CB glomus cells is OLFR78, encoding an atypical G-protein coupled olfactory receptor expressed in several tissues outside the olfactory epithelium. Surprisingly, OLFR78 has been suggested to be a lactate receptor essential for the CB-dependent oxygen regulation of breathing. However, this proposal has been challenged by our experiments in collaboration with the groups of Peter Mombaerts (Max-Plank Institute, Frankfurt, Germany) Randy Johnson (University of Cambridge, UK) and Hiro Matsunami (Duke University, USA) showing that OLFR78 is not necessary for sensing hypoxia or lactate (see Torres-Torrelo et al., Nature, 2018). We have found that, actually, lactate is by itself is a potent activator of CB glomus cells that potentiate the effects of hypoxia (Torres-Torrelo et al., Nature Communications, 2021).

c) Sub-project 3 in the application (Investigating the role of the adult neural stem cell niche in CB over- activation during sustained and intermittent hypoxia).
We have made important advances in this sub-project thanks to the identification in the adult rat CB of a subpopulation of TH-positive cells, which are “immature” (electrically excitable but O2-insensitive). Upon exposure to sustained hypoxia these immature cells are rapidly converted to mature O2-sensitive glomus cells (see Sobrino et al., EMBO Reports, 2018). In most aspects, this subpopulation of premature glomus cells behaves in a way similar to neuroblasts (precursors of mature neurons) in the central germinal centers. Interestingly, we have seen that similar changes occur in the CB of rats subjected to chronic intermittent hypoxia. These observations help to explain CB over-activation (increased sympathetic outflow) in sleep apnea patients undergoing repeated episodes of intermittent hypoxia.

d) Sub-project 4 in the application (Determining the impact of chemoreceptor inhibition -hypoxia intolerance- on adult brain neurogenesis).
In collaboration with Dr. David Macias (group of Dr. Randy Johnson at the University of Cambridge, UK) we have developed a model of CB atrophy (lack of CB development) based on the embryonic ablation of the HIF2a gene. These animals show intolerance to hypoxia (Macias et al., eLife, 2018).
In the context of this project, we have also studied the effect of genetic inhibition of MCI activity on the proliferation and differentiation of central neural progenitors. We have found that MCI is not needed for the maintenance of neural stem cells but it is required for their proliferation and differentiation to oligodendrocytes and neurons.
Acute oxygen sensing is essential for life in mammals. However, understanding the underlying molecular mechanism has remained elusive. The progress reached in the current ERC grant based on the comparative gene expression analysis between oxygen sensitive (cells in CB and AM) and oxygen insensitive (SCG neurons) sympatho-adrenal cells, has allowed us to propose the existence of a “signature metabolic profile” in acute O2-sensing cells. These data, together with the functional analyses at the systemic, cellular and mitochondrial levels performed on several genetically modified mouse models, have resulted in the elaboration of the most advanced and comprehensive hypothesis of acute oxygen sensing available so far. We think that the articles published in Cell Metabolism, Journal of Physiology, EMBO reports, eLife, Science Signaling, Nature, and Nature Communications collectively represent a breakthrough in the field regarding an issue that has been investigated by us, and others, for the last 30 years.
fig-oxygensensing.jpg