Hif1α-dependent mitochondrial acute O2 sensing and signaling to myocyte Ca2+ channels mediate arterial hypoxic vasodilation

Article https://doi.org/10.1038/s41467-024-51023-3 Hif1α-dependent mitochondrial acute O 2 sensing and signaling to myocyte Ca2+ channels mediate arterial hypoxic vasodilation Alejandro Moreno-Domínguez 1,2,3 , Olalla Colinas 1,2,3 , Ignacio Arias-Mayenco 1 , José M. Cabeza 1 ,JuanL.López-Ogayar 1 , Navdeep S. Chandel 4 , Norbert Weissmann 5 , Natascha Sommer 5 ,AlbertoPascual 1,3 & José López-Barneo 1,2,3 Vasodilation in response to low oxygen (O 2 ) tension (hypoxic vasodilation) is an essential homeostatic response of systemic arteries that facilitates O 2 supply to tissues according to demand. However, how blood vessels react to O 2 deficiency is not well understood. A commonbeliefisthatarterialmyocytes are O 2 -sensitive. Supporting this concept, it has been shown that the activity of myocyte L-type Ca2+channels, the main ion channels responsible for vascular contractility, is reversibly inhibited by hypoxia, although the underlying molecular mechanisms have remained elusive. Here, we show that genetic or pharmacological disruption of mitochondrial electron transport selectively abolishes O 2 modulation of Ca2+ channels and hypoxic vasodilation. Mitochondria function as O 2 sensors and effectors that signal myocyte Ca2+ channels due to constitutive Hif1α-mediated expression of specificelectron transport subunit isoforms. These findings reveal the acute O 2 -sensing mechanisms of vascular cells and may guide new developments in vascular pharmacology. It has been known for over a century that oxygen (O 2 ) is a major regulator of blood flow1–6; (see7,8for reviews and references). O 2 is a potent vasoconstrictor6,9–12 and in most systemic arterial territories, particularly in skeletal muscle, coronary, and cerebral vessels, a decreaseinO 2 tension (hypoxia) leadstorapid vasodilation8,13–17.Inthis manner, O 2 supply to tissues (e.g., the working heart) is swiftly adapted to demand. This essential vasomotor response also contributes to neurovascular coupling, a dilation of small blood vessels whereby O 2 availability is increased in the most active brain regions18–20. Details of the cellular mechanisms governing hypoxic vasodilation (HVD) remain controversial and the underlying molecular processes are largely unknown7,8,21. The vascular endothelium and neighboring tissues can release relaxing factors that act on smooth muscle8,22–29. In addition, hemoglobin desaturation and release of vasodilators from red blood cells has been proposed to induce increases in blood flow in hypoxic capillaries30,31. However, numerous studies have suggested that myocytes are a primary site for vascular acute O 2 sensing8,21,32.HVDhas Received: 15 January 2024 Accepted: 23 July 2024 Check for updates 1 Instituto de Biomedicina de Sevilla (IBiS), Hospital Universitario Virgen del Rocío/CSIC/Universidad de Sevilla, Seville, Spain. 2 Departamento de Fisiología Médica y Biofísica, Facultad de Medicina, Universidad de Sevilla, Seville, Spain. 3 Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Madrid, Spain. 4 Division of Pulmonary and Critical Care Medicine, Department of Medicine, Northwestern University, Chicago, IL, USA. 5 Excellence Cluster Cardiopulmonary System, University of Giessen and Marburg Lung Centre (UGMLC), German Centre for Lung Research (DZL), Justus-Liebig-University, Giessen, Germany. e-mail: [email protected] Nature Communications | (2024) 15:6649 1 1234567890():,; 1234567890():,; been recorded in arteries cleaned of surrounding tissue, denuded of endothelium, and bathed in cell-free external solutions equilibrated with low O 2 tension (PO 2 )33–35. Moreover, hypoxic relaxation is observed in arteries depolarized and precontracted with high extracellular K+14,35–37, indicating that PO 2 directly modulates transmembrane Ca2+ influx in vascular smooth muscle (VSM)36,38. In support of this concept, hypoxia, within physiological values, has been shown to elicit fast, and highly reversible inhibition of L-type voltage-gated Ca2+channels (the main ion channel type responsible for vascular contractility) in myocytes from several systemic rodent, porcine, and human arterial models9,35,37,39–41. However, the molecular mechanism underpinning Ca2+ channel regulation by PO 2 has been elusive42. Here, we report that a constitutively expressed hypoxia inducible factor 1α (Hif1α)-dependent mitochondrial acute O 2 sensing and signaling pathway, in systemic arterial myocytes, mediates the hypoxia-induced inhibition of Ca2+ channel activity and HVD. Results Disruption of mitochondrial complex I in arterial myocytes abolishes acute O 2 modulation of Ca2+channel activity and hypoxic vasodilation Inhibition of mitochondrial complex (MC) I activity in carotid body chemoreceptor glomus cells, the prototypical acute O 2 sensors in mammals, suppresses responsiveness of these cells to hypoxia43–45. Hence, we investigated whether a similar phenomenon occurred in arterial myocytes. Disruption of MCI was achieved by targeting the gene encoding Ndufs2, a core subunit essential for ubiquinone (CoQ) binding and NADH dehydrogenase activity46. Cells lacking Ndufs2 show almost complete abolition of MCI function44,47.Inthe present study, mice (of both sexes) carrying Ndufs2 floxed and minus alleles45 were bred with mice carrying the tamoxifen (TMX)-activated Cre recombinase gene driven by the smooth muscle-specificSM22 promoter48,49. These adult mice, with conditional deficiency of MCI in VSM cells (MCI-SM mice), were studied in comparison with wild type (WT) mice. Two-month-old mice received a TMX diet for one month andwerelefttorecoverforatleastoneadditionalmonthonanormal diet (Fig. 1a). TMX-treated MCI-SM mice ( ≥4-months old) showed a strong and selective decrease in Ndufs2 mRNA obtained from the femoral artery and aorta (Fig. 1b, c). Ablation of Ndufs2 in VSM cells resulted in an ~80% decrease in MCI activity in the aorta (Fig. 1d). Except for an increase in spleen size, MCI-SM mice did not show any alteration in survival, body weight or anatomical features (Supplementary Figs. 1a–c). They had a normal hematocrit and arterial histology (Supplementary Figs. 1d–g), with normalheart rate and hypoxic ventilatory responses to hypoxia, a homeostatic reflex largely dependent on O 2 -sensing by carotid body glomus cells44,50,andhypercapnia (Supplementary Figs. 1h–j), However, in comparison with WT mice, MCI-SM mice showed a marked hyperlactatemia (Fig. 1e), probably as a consequence of the up-regulation of glycolysis in MCI-deficient myocytes45. They also showed a significant decrease in systolic and diastolic blood pressures (Fig. 1f), a phenomenon that could be due to the decreased vascular sensitivity to O 2 tension and the blunted vasoconstrictor effect of O 2 on aorta and peripheral arteries (see below). These changes indicated that even under basal physiological conditions the loss of MCI activity in VSM cells alters fundamental homeostatic mechanisms. Next, we studied acutely dispersed femoral myocytes, as a representative example of conduit systemic myocytes39,51, from MCISM mice which, despite having a selective absence of Ndufs2 immunoreactivity, appeared healthy and with the characteristic elongated shape of WT VSM cells (Fig. 1g). However, they showed an increase in mitochondrialnumber,similartowhatwehavepreviouslyseeninMCI- deficient carotid body glomus cells44, and aberrant ultrastructural changes that resemble those reported in skeletal muscle with mitochondrial myopathies52 (Supplementary Figs. 2a–i). Patch clamped MCI-deficient myocytes exhibited normal passive electrical parameters and density of voltage-gated Ca2+ and K+channels (Fig. 1h; and Supplementary Figs. 3a–e), as well as responses to vasoactive agents (see below), indicating that as it occurs with O 2 -sensitive carotid body glomus cells44 arterial myocytes can survive mitochondrial dysfunction possibly due to compensatory up-regulation of glycolysis45 (see Fig. 1e). WT mouse femoral myocytes were exposed to hypoxia by switching from PO 2 of 150 mmHg to ~15 mmHg in the solution bathing the cells39, covering in this manner the full range of O 2 tensionsexisting in most tissues during normoxia and physiological hypoxia53.As observed previously in several types of systemic myocytes39,51,this treatment produced a fast, progressive, and reversible decrease in the amplitude of whole-cell L-type Ca2+ channel currents without altering their activation or deactivation kinetics (Fig. 1i–k and Supplementary Figs. 4a–d). However, in contrast to their normal histological appearance and basic electrophysiology, VSM cells from MCI-SM mice exhibited a practically abolished modulation of Ca2+ channel activityby PO 2 (Fig. 1l, m). In these patch clamp experiments cells were dialyzed with intracellular MgATP to prevent activation of ATP-dependent K+ channels, despite the participation of these channels in VSM responses to physiological hypoxia is questioned21,32,35,41,54,55. Voltage-dependent K+currents were unaffected by hypoxia (Supplementary Figs. 4e–g), suggesting that in femoral vascular myocytes voltage-gated K+channels lack sites of interaction with the mitochondrial signals. Femoralmyocytesdidnotshowanyevidenceofvoltage-gatedNa+ currents, despite we used a high-Na+external solution in our patch clamp experiments. In addition, the size and time course of inward currentswereunaffectedbycompletereplacement ofextracellularNa+ with the impermeant cation N-methyl-D-glucamine (Supplementary Figs. 5a, c). Modulation of the Ca2+ currents (recorded without Na+in the internal solution) by hypoxia was also unaltered in the absence of extracellular Na+(Supplementary Figs. 5b, c), thus making it unlikely any relevant role of Na+control of mitochondrial signaling56 in the response to hypoxia of myocyte Ca2+ channels. In agreement with the electrophysiological data, a strong inhibition of transmembrane Ca2+ influx triggered by either methyl 2,5- dimethyl-4-[2-(phenylmethyl)benzoyl]-1H-pyrrole-3-carboxylate (FPL), an L-type Ca2+ channel activator57,58, or depolarization with high extracellular K+15,36 was produced by hypoxia in intact Fura 2-loaded femoralmyocytes.These effectsofhypoxiawereeliminatedinMCI-SM myocytes (Fig. 2a–c and Supplementary Figs. 6a–d). Rotenone, a highly membrane-permeant MCI blocker, mimicked hypoxia and reversibly inhibited FPL-induced transmembrane Ca2+ influx, an effect that was markedly attenuated in femoral myocytes from MCI-SM mice (Fig. 2d, e). Preincubation of WT myocytes with rotenone not only produced a dose-dependent inhibition of the amplitude of FPL- induced Ca2+ signals but it occluded any further effect of hypoxia (Supplementary Figs. 7a–d). Caffeine-evoked Ca2+ release from the sarcoplasmic reticulum59 was unaltered in MCI-deficient myocytes (Supplementary Figs. 8a, b) and unaffected by rotenone (Supplementary Figs. 8c, d). Similar to hypoxia, blockers of the MCI (rotenone or piericidin A), MCIII (antimycin A) and MCIV (cyanide) decreased the amplitude of the Ca2+ currents and prevented any further effect of hypoxia (Fig. 2f left, g; and Supplementary Fig. 9a–f). Inhibition of the current with rotenone also occluded any further effect of cyanide (Fig. 2f, right, g). As responsiveness to hypoxia may change in different arterial segments and territories8,17, we also studied cerebral arterial VSM cells, as a representative example of resistance myocytes, which have a smaller size (Fig. 3a)and capacitance (Supplementary Figs. 3a and 10a) than femoral myocytes but similar O 2 -sensitive Ca2+ currents (Fig. 3b, c). Passive electrical parameters and Ca2+ current amplitudes of cerebral myocytes were unaffected by MCI deficiency (Supplementary Fig. 10a); however, the modulation of Ca2+ channel activity by hypoxia was abolished (Fig. 3b, c; Supplementary Fig. 10b, c). We also Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 2 tested that modulation of FPL- or high extracellular K+-induced transmembrane Ca2+ influx by hypoxia was abolished in MCI-deficient myocytes from middle or posterior cerebral arteries (example of resistance VSM cells) (Fig. 3d–f) and from aorta (example of myocytes from conduit and more elastic artery) (Supplementary Figs. 11a–c). Consistent with the findings in single femoral and cerebral myocytes, endothelium-denuded femoral arteries resected from MCI-SM mice showed a normal passive diameter and contractile responses to FPL, high K+and caffeine, although they had strongly inhibited hypoxiainduced vasodilation (Fig. 3g–j). Taken together, these data indicate that mice can survive MCI disruption in VSM, although they exhibit hyperlactatemia, decreased systemic blood pressure, and strong inhibition of HVD. Remarkably, MCI-deficient arterial myocytes have a normal morphology, electrical ba Ndufs2 CRESM22 Age (months) TMX EXP 12 4 34.5 MgATP (4 mM) Molecular characterization of MCI-SM mice Systemic phenotype Single cell Ndufs2 expression 2+ Acute inhibition of L-type Ca channels by hypoxia - femoral arterial myocytes - mRNA (relative to WT) mRNA (relative to WT) MCI activity (% of WT) Femoral Aorta cd WT MCI-SM WT MCI-SM g f e basal lactate (mmol/L) blood pressure (mmHg) 25 Pm DSMA Ndufs2 DAPI h -70 mV 10 mV 100 pA 5 ms Hypoxia -70 mV 10 mV Hypoxia 100 pA 5 ms i j (1 min) (3 min) (1 min) (3 min) WT MCI-SM lk t activation at 10 mV ½ (ms) tau fast deactivation at -70 mV (Ps) WT MCI-SM MCI-SM mouse model m WT MCI-SM % current inhibition in hypoxia Vm (mV) patch pipette myocyte Fig. 1 | Acute hypoxic inhibition of Ca2+ channel currents is abolished in mitochondrial complex I-deficient femoral arterial myocytes. a MCI-SM mouse model and time course of tamoxifen (TMX) treatment. b,cNdufs2 and Sdhd mRNA levels in femoral artery and aorta from WT mice (n= 5, blue dots) and MCI-SM mice (n= 5, orange dots). dMitochondrial complex I (MCI) activity in aortas of WT and MCI-SM mice (n=3).eLactate plasma levels of WT (n= 8) and MCI-SM (n=8)mice. fSystolic (n= 10) and diastolic (n= 11) blood pressure of WT and MCI-SM mice. gFemoral myocytes from WT(upper panel) and MCI-SM (lower panel)mice stained with antibodies against Ndufs2 (green) and alpha-smooth muscle actin (α-SMA, red). hSchematic drawing of a whole-cell patch-clamped myocyte dialyzed with 4 mM MgATP. iReversible inhibition of Ca2+ currents recorded in acutely dispersed WT femoral myocytes upon exposure to hypoxia (O 2 tension ~15 mmHg). Currents were generated by 20-ms depolarizing pulses to 10mV. Time after switching from normoxia to hypoxia is indicated. jTime course ofactivation (time to reachthe half maximal value; t 1/2 ) of currents recorded at 10 mV, measured in WT femoral myocytes under normoxic (Nx, n= 13/5 cells/mice) or hypoxic (Hx, n= 11/5 cells/mice) conditions. kTime course of deactivation at −70mV (tau of fast tail component) of currents recorded under normoxia (Nx, n= 13/5 cells/mice) or hypoxia (Hx, n= 11/5 cells/mice). lLoss of the hypoxic inhibition of Ca2+ currents in femoral myocytes from MCI-SM mice. mEffect of hypoxia expressed as percentage inhibition of the current elicited at −10, 0, 10 and 20 mV in femoral arterial myocytes from WT mice (n= 11/5 cells/mice) and MCI-SM mice (n=15/5 cells/mice). Data are presented as mean ± SEM with statistically significant Pvalues (<0.05) indicated in each panel. In b–f,j,kand mcomparison of values was performed by two-tailed unpaired t tests. Source data are provided as a Source Data file. Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 3 parameters, voltage-gated ionic currents, and response to caffeine but lack the hypoxic inhibition of transmembrane Ca2+ influx. These functional alterations appear to be due to the loss of L-type Ca2+ channel modulation by O 2 , a physiological process whose underlying mechanism has remained unknown and is now shown to be strictly dependent on MCI function. Interaction of O 2 -dependent mitochondrial signals with Ca2+ channels The production of intracellular reactive oxygen species (ROS) of mitochondrial origin has been reported in several VSM cell types exposed to hypoxia60–62. On the other hand, in O 2 sensing carotid body glomus cells, hypoxia induces rapid and reversible increases in NADH and H 2 O 2 , which are abolished by MCI-deficiency44.Ithasrecentlybeen proposed that acute O 2 sensing in glomus cells relies on a slowdown of the mitochondrial electron transport chain (ETC) in hypoxia and a subsequent backlog of electrons with the accumulation of NADH and ROS (H 2 O 2 ) in the mitochondrial intermembrane space (IMS) (Fig. 4a). After equilibration with the cytosol, thesemitochondrial signals, which are also produced by the rotenone-induced blockade of MCI, modulate membrane ionchannelsto induceglomuscelldepolarization and transmitter release44,47. To test whether similar mitochondrial-to- membrane signaling mechanisms operate in VSM cells, we studied the effects of the intracellular application of NADH and H 2 O 2 ,aswellas the reducing agent N-acetylcysteine (NAC), on Ca2+ currents in dialyzed patch-clamped WT myocytes (Fig. 4b). We also assessed the effect of pharmacological modulation of the intracellular NADH/NAD+ ratio by extracellular application of highly membrane permeable αhydroxybutyrate or α-ketobutyrate (αHB/αKB)47,63 (Fig. 4b). Both internal NADH and H 2 O 2 at concentrations near estimated physiological levels produced a clear dose-dependent inhibition of Ca2+ currents, with the effects of the two agents found to be additive (Fig. 4c,d; and Supplementary Fig. 12a, b). Similar to hypoxia (see Figs. 1i–l), NADH and H 2 O 2 consistently decreased Ca2+ current amplitude without affecting their activation or deactivation kinetics (Fig. 4e, f). Modulation of the Ca2+ current by hypoxia was blunted or even abolished in cells already loaded with relatively low doses of NADH plus H 2 O 2 (Fig. 4c, g). Dialysis of myocytes with NAC (0.5 mM) did not appreciably change the Ca2+ current amplitude (Fig. 5a, b), but did strongly inhibit the hypoxic modulation of these channels (Fig. 5a, c). Moreover, the application of cell-permeant αHB, which increases intracellular NADH levels, decreased, as seen with hypoxia, the amplitude of the Ca2+ current (Fig. 5d, e). In contrast, αKB, which reduces intracellular NADH47,63, had little effect by itself but decreased the hypoxic modulation of the current (Fig. 5d, f). These data indicate that the increase in intracellular NADH and H 2 O 2 levels produces a robust inhibition of myocyte Ca2+ currents similar to that elicited by hypoxia or rotenone, and that lowering PO 2 does not have any further effect on the current once NADH and H 2 O 2 are interacting with the Ca2+ channels. Loading the cells with NAC can prevent the action of hypoxia-generated H 2 O 2 , while the increase in NADH alone (application of extracellular αHB) partially mimics hypoxia. These findings strongly suggest that the fast, reversible, and MCI-dependent % decrease 'ratio 340/380 hypoxia 2+ Single myocyte cytosolic Ca 'ratio 340/380 FPL (1 PM) Hypoxia 0.05 ΔRatio F/F 340 380 5 min FPL (1 PM) FPL (1 PM) 0.05 ΔRatio F/F 340 380 5 min Hypoxia MCI-SMWT % decrease 'ratio 340/380 RTN (1 PM) WT MCI-SM ab ce FPL (1 PM) RTN (1 PM) FPL (1 PM) RTN (1 PM) 0.05 ΔRatio F/F 340 380 0.05 ΔRatio F/F 340 380 5 min 5 min dfg MCI-SMWT 5 ms 100 pA 5 ms -70 mV 10 mV -70 mV 10 mV RTN+Hx RTN Control RTN+CN RTN Control 100 pA current density at 10 mV (relative to control) Control RTN (0.75 PM) RTN (0.75 PM) + Hx RTN (0.75 PM) + CN (300 PM) 2+ Effect of rotenone and hypoxia on Ca influx Fig. 2 | Acute hypoxic inhibition of transmembrane Ca2+ influx is lost in mice with mitochondrial complex I-deficient arterial myocytes. a Changes in the FPL- evoked increase in cytosolic [Ca2+] elicited by hypoxia (O 2 tension ~15 mmHg) in Fura-2-loaded femoral myocytes from WT and MCI-SM mice. bBox plot showing the increase in cytosolic [Ca2+] elicited by 1 μM FPLin dispersed cells from WT mice (n= 31/4 cells/mice, blue dots) and MCI-SM mice (n= 38/4 cells/mice, orange dots). cEffect of hypoxia on the FPL-evoked increase of cytosolic [Ca2+]inFura-2-loaded dispersed myocytes from WT mice (n= 31/4 cells/mice) and MCI-SM mice (n= 38/4 cells/mice). dChanges in the FPL-evoked increase of cytosolic [Ca2+]inducedby 1μM rotenone (RTN) in WT and MCI-SM femoral myocytes. eEffect of rotenone on the FPL-evoked increase in cytosolic [Ca2+]inWT(n= 24/4 cells/mice) and MCI-SM (n = 17/3 cells/mice) femoral myocytes. fInhibition of Ca2+ currents by 0.75 μM rotenone (RTN) in acutely dispersed femoral arterial myocytes from WT mice. Note that rotenone prevented any further effect of hypoxia (RTN+Hx, left) or 300 μM cyanide (RTN + CN, right). gPeak Ca2+ current density in WT myocytes at 10 mV under control conditions (n=11/3 cells/mice) and in the presence of 0.75 μM rotenone (n= 11/3 cells/mice), 0.75 μM rotenone+hypoxia (n= 11/3 cells/mice), or 0.75 μM rotenone+300 μM cyanide (n= 9/3 cells/mice). In c, e and gdata are presentedas mean ± SEM with statistically significantPvalues (<0.05) indicated in each panel. The box plot in b represents median (middle line), 25th, 75th percentile (box), and largest and smallest values range (whiskers). In bPvalues were calculated with the two tailed Mann–Whitney test. In cand ecomparison of values was done by two-tailed unpaired t tests. In gcomparisons were performed by one-way ANOVA followed by Tukey’s multiple comparisons post hoc test. Source data are provided as a Source Data file. Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 4 hypoxic modulation of myocyte L-type Ca2+ channel activity is mediated by mitochondrial NADH and H 2 O 2 . Transgenic expression of yeast NADH dehydrogenase in MCI- deficientarterialmyocytesrestoresO 2 modulation of Ca2+ channels Given the essential role played by MCI in acute hypoxia signaling in arterial myocytes, we tested whether transgenic expression of Ndi1, a single-molecule yeast NADH dehydrogenase, can rescue Ca2+ channel sensitivity to hypoxia in MCI-deficient VSM cells. Ndi1 is attached to the inner mitochondrial membrane and transfers electrons from NADH to CoQ, but unlike MCI this activity is not coupled to a proton pump45 (Fig. 6a). In this way, Ndi1 expression allowed us to assess if the recovery of NADH dehydrogenase activity by itself, without the electrochemical gradient generated by MCI, is sufficient to induce functional recovery in MCI-null myocytes. To this end, we generated two Femoral artery ex vivo % decrease in force force (mN) IC (Pm) 100 hypoxia caffeine hij g basal FPL KCl FPL (1 PM) Hypoxia MCI-SM WT FPL (1 PM) Hypoxia 1 mN 1 min WT MCI-SM 100 pA 5 ms -70 mV 10 mV -70 mV 10 mV 100 pA 5 ms Hypoxia Recovery Normoxia 2+ Acute inhibition of L-type Ca channels by hypoxia - cerebral arterial myocytes - ba c 25 Pm Femoral myocyte Cerebral myocyte % current inhibition in hypoxia Vm (mV) WT MCI-SM WT MCI-SM de Hypoxia FPL (1 PM) 0.05 ΔRatio F/F 340 380 2 min 'ratio 340/380 FPL (1 PM) MCI-SM WT hypoxia % decrease 'ratio 340/380 WT MCI-SM f Hypoxia FPL (1 PM) 0.05 ΔRatio F/F 340 380 2 min Single myocyte cytosolic 2+ Ca Fig. 3 | Mitochondrial complex I disruption abolishes hypoxic inhibition of transmembrane Ca2+ influx in cerebral myocytes and arterial hypoxic vasodilation. a Representative photomicrographs comparing acutely dispersed myocytes obtained from femoral (upper panel) and cerebral (lower panel) arteries. bDifferential modulation of Ca2+ channels by hypoxia (O 2 tension ~15 mmHg) in cerebral arterial myocytes from WT and MCI-SM mice. Current traces are superimposed to facilitate comparison. cEffect of hypoxia expressed as percentage inhibition of the current elicited at −10, 0, 10 and 20mV in cerebral arterial myocytes from WT (blue dots) and MCI-SM (orange dots) mice (n= 10/3 cells/mice). dChanges in the FPL-evoked increase of cytosolic [Ca2+]byhypoxia,inFura-2- loaded cerebral arterial myocytes from WT and MCI-SM mice. eIncrease in cytosolic [Ca2+]elicited by1 μM FPL in dispersedcells from WT mice (n= 21/3cells/mice) and MCI-SM mice (n= 27/3 cells/mice). fEffect of hypoxia on the FPL-evoked increase in cytosolic [Ca2+], shown as percentage of the peak FPL-evoked signal in Fura-2 loaded dispersed cells from WT mice (n= 21/3 cells/mice) and MCI-SM mice (n= 27/3 cells/mice). gChanges in the FPL-evoked increase in arterial wall tension induced by hypoxia in femoral arterial rings from WT and MCI-SM mice. hBasal internal vessel circumference in WT and MCI-SM femoral rings (n= 14/4 rings/ mice). iBox plot of isometric force generated in arterial ringsfrom WT and MCI-SM mice in response to either 1 μMFPL(n= 21/7 rings/mice), 70 mM KCl (n=18/7and 19/7 rings/mice, respectively) or 5 mM caffeine (n= 13/4 rings/mice). jEffect of hypoxia on the FPL-induced contraction of WT and MCI-SM arterial rings (n= 14/4 rings/mice). In c,e,f,hand jdata are presented as mean ± SEM with statistically significant Pvalues (<0.05) indicated in each panel. Pvalues were calculated with two-tailed unpaired t tests. In ithe box plots represent median (middle line), 25th, 75th percentile (box), and largest and smallest values range (whiskers). Pvalues were calculated with the Kruskal-Wallis test followed by Dunn’s multiple comparisons post hoc test. Source data are provided as a Source Data file. Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 5 new conditional adult mouse models (named WT/NDI1-SM and MCI/ NDI1-SM) in which TMX treatment induced Ndi1 expression in WT and MCI-SM mice, respectively (Fig. 6b). We showed immunocytochemical evidence of Ndi1 expression in VSM cells (Fig. 6c) and the appropriate selective down-/up-regulation ofNdufs2 andNdi1 mRNAs,respectively, in arteries of the various mouse strains (Fig. 6d, e). Ndi1 expression did not induce any major alteration in survival, body weight or organ size in WT mice; however, the presence of Ndi1 rescued practically all the systemic structural and functional alterations observed in MCI-SM mice (Supplementary Figs. 13a–d). Notably, increased lactatemia and low systolic and diastolic blood pressures, which were characteristic features of MCI-SM mice (see Fig. 1e, f), showed full recovery in MCI/ NDI1-SM mice (Fig. 6f, g). While Ndi1 expression in WT myocytes did not produce any change in Ca2+ current amplitude or modulation by O 2 (Supplementary Fig. 13e), Ndi1 did induce an almost complete recovery of the inhibition of Ca2+ current by hypoxia in MCI/NDI1-SM myocytes (Fig. 6h, i, and Supplementary Fig. 13f), which was abolished in MCI- null VSM cells (see Figs. 1i, m). In agreement with these data, inhibition of FPL-induced transmembrane Ca2+ influx by hypoxia, which was blunted by Ndufs2 deficiency (see Fig. 2a–c), was also rescued in Ndi1-expressing myocytes (Fig. 7a, left, b, c). In contrast, the responsiveness of FPL-induced transmembrane Ca2+ influx to rotenone (see Fig. 2d, e) was not recovered in MCI/NDI1-SM myocytes, where it is well known that Ndi1 is insensitive to this drug45 (Fig. 7a right, d). a 100 pA 5 ms NADH/ HO/ 22 NAC DKB/DHB MgATP -70 mV 10 mV Hypoxia Normoxia Recovery Hypoxia Normoxia Control NADH (200 PM) H O (25 PM) 22 NADH (200 PM) + H O (25 PM) 22 current density at 10 mV (relative to control) Control NADH (200 PM) H O (25 PM) 22 NADH (200 PM) + H O (25 PM) 22 100 pA 5 ms 100 pA 5 ms 100 pA 5 ms b c d t activation at 10 mV ½ (ms) tau fast deactivation at -70 mV (Ps) CC ef HO 22 NADH + % current inhibition in hypoxia gControl NADH (200 PM) + H O (25 PM) 22 0 MCI CoQH2MCIII MCIV NADH ROS ROS MCII Other O2 H2O (Hx) Matrix IMS RTN Mitochondrial electron transport chain - effect of hypoxia - HO 22 NADH + patch pipette myocyte 2+ Modulation of Ca channels by NADH and H O 22 Vm (mV) CN Fig. 4 | Effect of changes in intracellular levels of NADH and/or H 2 O 2 on Ca2+ currents in arterial myocytes. a Scheme of the mitochondrial electron transport chain andacutedynamicchangesinducedbyhypoxia (Hx,redlinesandsymbols)in acute O 2 -sensing cells. IMS (mitochondrial intermembrane space), CoQH 2 (reduced ubiquinone), RTN (rotenone), CN (cyanide). bSchematic drawing of a whole-cell patch-clamped myocyte dialyzed with 4 mM ATP plus variable concentrations of NADH, H 2 O 2 or N-acetylcysteine (NAC). Membrane-permeable α-ketobutyrate (α- KB) and α-hydroxybutyrate (α-HB) were applied extracellularly. cInhibition of femoral myocyte Ca2+ currents by hypoxia (O 2 tension ~15 mmHg) and by intracellular dialysis with NADH (200 μM), H 2 O 2 (25 μM), or both. dPeak current density at 10 mV under the experimental conditions indicated in (c). Number of cells/mice: Control (blue dots, 16/7); NADH (yellow dots, 9/3); H 2 O 2 (orange dots, 8/3); NADH + H 2 O 2 (brown dots, 10/4). eBox plot illustrating the time course of activation (time to reach half the maximal value; t 1/2 )ofCa 2+ currents recorded from WT myocytes at 10 mV. C, control conditions (n= 10/4 cells/mice); NADH + H 2 O 2 ,cells dialyzed with 200μM NADH plus 25 μMH 2 O 2 (n= 10/4 cells/mice).fTime course of deactivation (tau of fast component of tail current) of Ca2+ currents recorded from WT myocytes at −70 mV. C and NADH + H 2 O 2 as indicated in (e). gEffect of hypoxia expressed aspercentage inhibition of the current elicited at−10,0,10and20mVin femoral myocytes from WT mice, under control conditions (n=5/4cells/mice)or dialyzed with 200μMNADH+25μMH 2 O 2 (n = 10/4 cells/mice). In d,f,andgdata are presented as mean ± SEM with statistically significant Pvalues (<0.05) indicated in each panel. In dcomparisons were performed by one-way ANOVA followed by Tukey’s multiple comparisons posthoctest. In fandgcomparisonswerecalculated using two-tailed unpaired t tests. In ethe box plot represents median (middle line), 25th, 75th percentile (box), and largestand smallest values range(whiskers). Pvalue wascalculated with the two tailed Mann–Whitney test. Source data are provided as a Source Data file. Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 6 To further demonstrate the recovery of hypoxia signaling in Ndi1 expressing MCI-deficient myocytes, we recorded reversible changes in NADH autofluorescence induced by hypoxia in single dispersed myocytes using two photon microfluorimetry. As shown previously in O 2 - sensitive glomus cells47, hypoxia induced rapid and reversible increases in NADH levels in WT myocytes. This signal was practically abolished in cells from MCI-SM mice and was fully recovered in MCI/NDI1- SM myocytes (Fig. 7e, f). In all these genetically different cell types, a clear decrease in NADH levels was observed after addition of αKB (1 mM) (Fig. 7e), which permeates to the cell interior and consumes NADH during its conversion to αHb (see Fig. 5d). These data further support the view that the essential role played by MCI during hypoxia is to generate the signaling molecules (NADH and H 2 O 2 ) that regulate Ca2+ channels. In this regard, Ndufs2 deficiency can be fully rescued by the transgenic expression of yeast NADH dehydrogenase. Acute O 2 sensitivity of arterial myocytes relies on the constitutive activity of Hif1αand expression of atypical cytochrome c oxidase subunits The hypoxia inducible transcription factors Hif1αand Hif2αare key mediators of cellular transcriptional responses to hypoxia64. We have shown that Hif2α, which is expressed at constitutively high levels in the carotid body65–67, isrequiredforinductionofa transcriptionalprogram that determines the expression of atypical MCIV subunit isoforms conferring acute responsiveness of chemoreceptor glomus cells to physiological hypoxia68–70.GiventhatHif2αis expressed in systemic arteries, although at lower levels than in the lung (Supplementary Fig. 14a), we investigated whether conditional deletion of the Hif2α gene (Epas1) also alters myocyte acute O 2 sensing. To this end, we analyzed the EPAS1-ESR mouse model with a TMX-induced loss of Epas1 alleles in adulthood68 (Supplementary Fig. 14b). We found a marked selective decrease of Epas1 mRNA in the aortas of EPAS1-ESR mice, however no changes in mRNA levels of relevant MCIV subunit isoforms were observed (Supplementary Fig. 14c). Modulation of Ca2+ channel activity (Supplementary Figs. 15a–d)andFPL-mediatedCa 2+ influx (Supplementary Figs. 15e–g) during hypoxia were also unaltered in Hif2α-deficient myocytes. These findings indicated that Hif2αis not responsible for the acute O 2 sensitivityof systemicVSMcells.Theyalso led us to hypothesize that the expression of MCIV subunit isoforms presumably involved in acute O 2 sensing could be under the transcriptional control of Hif1α, which is expressed in arterial smooth muscle at higher levels than those seen for Hif2α(Supplementary Fig. 16a). To address this question, we generated adult HIF1α-SM mice with a conditional loss of Hif1αalleles in VSM cells (Fig. 8a) following the same breeding protocol and TMX treatment described for MCI-SM mice (see Fig. 1a). We confirmed a marked selective decrease in Hif1α mRNA levels in aortic tissue (with unchanged Epas1 mRNA levels) paralleled by a manifest mRNA down-regulation of relevant atypical MCIV subunit isoforms, such as Cox4i2 or Cox8b68,aswellasPgk1,a well-known Hif1α-dependent gene encoding a glycolytic enzyme71,72. Other mRNA species encoding MCIV subunits (Ndufa4l2, Ndufa4), enzymes (Pcx), or transcription factors suggested to regulate the expression of MCIV subunits (Rbpj, Chchd2, Cxxc5)73 were either less affected or unaltered (Fig. 8b, c). Remarkably, HIF1α-SM mice did not show any major alteration in body weight or gross modification of organ size, hematocrit or heart rate (Supplementary Figs. 16b–e). However, similar to MCI-SM mice (see Fig. 1f), HIF1α-SM mice showed a DHB+Hypoxia DHB Control DKB+Hypoxia DKB Control current density at 10 mV (relative to control) current density at 10 mV (relative to control) Control DHB (2.5 mM) DHB (2.5 mM) + Hypoxia Control DKB (2.5 mM) DKB (2.5 mM) + Hypoxia d ef 100 pA 5 ms 100 pA 5 ms 2+ Modulation of Ca channels by NADH and redox reagents bca current density at 10 mV (relative to control) Control NAC (500 PM) 100 pA 5 ms Control NAC (500 PM) -70 mV 10 mV NAC (500 PM) NAC+Hypoxia 100 pA 5 ms % current inhibition in hypoxia Control NAC (500 PM) Vm (mV) 0 Fig. 5 | Effect of NADH and redox reagents on Ca2+ currents in arterial myocytes. aCa2+ current traces recorded at 10 mV from dispersed femoral arterial myocytes. Left, Superimposed traces obtained in control conditions and in cells dialyzed with N-acetylcysteine (NAC). Right, Superimposed traces obtained in cells dialyzed with NAC before and after exposure to hypoxia. bCurrent density at 10mV in control conditions (n= 5/4 cells/mice) and in the presence of 500 μM NAC (n= 11/4 cells/ mice). cEffectof hypoxia, expressedas percentage inhibition of the current elicited at −10, 0, 10 and 20 mV on myocyte Ca2+ currents recorded in control conditions (n= 9/6cell/mice) or incellsdialyzed with 500 μMNAC(n= 11/4 cells/mice). dInset. Scheme representing the reversible conversion of αKB into αHB in cells with parallel consumption or generation of NADH. LDH (lactate dehydrogenase), HBDB (βhydroxybutyrate dehydrogenase). Modulation of Ca2+ channels by extracellular application of 2.5 mM αHB or 2.5 mM αKB in femoral myocytes (see Fig. 4b). The effects of hypoxia are also shown. eCa2+ current density at 10 mV recorded in myocytes under control conditions and in the presence of 2.5 mM αHB or 2.5 mM αHB+hypoxia (n= 14/4 cells/mice). fCa2+ current density at 10 mV recorded in femoral myocytes in control conditions and in the presence of 2.5 mM αKB or 2.5 mM αKB+hypoxia (n= 15/3 cells/mice). Data in b,c,eand fare presented as mean ± SEM with statistically significant Pvalues (<0.05) indicated in each panel. Comparisons were calculated using two-tailed unpaired t tests. Source data are provided as a Source Data file. Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 7 clear decrease in systolic and diastolic blood pressures (Fig. 8d). Hif1α- deficient femoral artery myocytes appeared normal, with an unchanged density of L-type Ca2+ channels (Supplementary Figs. 16f, g). However, in these myocytes, modulation of Ca2+ channel activity by hypoxia, as evidenced both by direct measurements of whole-cell currents and by FPL-induced transmembrane Ca2+ influx in Fura-2- loaded isolated cells, were practically abolished (Fig. 8e–i, Supplementary Figs. 16f, g). In contrast, the rise in cytosolic [Ca2+] induced by either caffeine or phenylephrine (an α 1 receptor agonist)74, was unaffected (Fig. 8j). These data indicated that O 2 modulation of Ca2+ 2+ Ndi1 expression rescues O Ca channels 2modulation of % current inhibition in hypoxia Matrix IMS MCII ECA MCI Ndi1 Q NADH NADH NAD NAD ++ 2e e - - Ndufs2 CRESM22 Age (months) TMX EXP 124 34.5 MCI/NDI1-SM mouse model STOP Ndi1 WT MCI-SM MCI/NDI1-SM WT/NDI1-SM Ndufs2 (Aorta) mRNA (relative to WT) blood pressure (mmHg) basal lactate (mmol/L) Systemic phenotype Molecular characterization Replacement of MCI with yeast Ndi1 Ndi1 (Aorta) systolic diastolic mRNA (relative to WT/NDI1) defg ab -70 mV 10 mV 5 ms Hypoxia Recovery Normoxia Hypoxia Normoxia Recovery 100 pA 100 pA 5 ms WT/NDI1-SM MCI/NDI1-SM hi WT/NDI1-SM MCI -SM/NDI1 -70 mV 10 mV H+ Vm (mV) c DSMA GFP DAPI WT/NDI1-SM 25 Pm WT Fig. 6 | Transgenic expression of yeast Ndi1 rescues O 2 modulation of Ca2+ channel activity in MCI-deficient arterial myocytes. a Scheme illustrating the replacement of mitochondrial complex I (MCI) with Ndi1 in the MCI/NDI1-SM mouse model. bExperimental MCI/MDI1-SM mouse model and time course of tamoxifen (TMX) treatment. cImmunostaining illustrating α-smooth muscle actin (α-SMA, red) and Ndi1 expression (green fluorescent protein, GFP) in freshly dispersed femoral arterial myocytes from WT (upper panel) and WT/NDI1-SM (lower panel) mice. d,eNdufs2 and Ndi1 mRNA levels in aortas from WT (blue dots), MCISM (orange dots), MCI/NDI1-SM (green dots) and WT/NDI1-SM (black dots) mice (n=5 replicates per group, with each replicate consisting of one mouse of the corresponding genotype). fBasal plasma lactate levels in WT mice (n= 8), MCI-SM mice (n= 8) MCI/NDI1-SM mice (n= 11) and WT/NDI1-SM mice (n=9).gSystolic and diastolic bloodpressureinWTmice(n= 11, 10respectively), MCI-SMmice (n=11,10 respectively) MCI/NDI1-SM mice (n= 12, 12 respectively) and WT/NDI1-SM mice (n= 10, 10 respectively). hMacroscopic current traces illustrating the reversible inhibition of L-type Ca2+ channel currents by hypoxia in femoral arterial myocytes from WT/NDI1-SM and MCI/NDI1-SM mice. iEffect of hypoxia expressed as percentage inhibition of the Ca2+ current elicited at −10, 0, 10 and 20mV in femoral arterial myocytes from WT/NDI1-SM mice (n= 12/4 cells/mice) and MCI/NDI1-SM mice (n= 11/4 cells/mice). In d–g,andi, data are presented as mean ± SEM with statistically significant Pvalues (<0.05) indicated in eachpanel. In d–gcomparisons were performed by one-way ANOVA followed by Tukey’s multiple comparisons post hoc test. In icomparison of values were performed by two-tailed unpaired t tests. Source data are provided as a Source Data file. Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 8 channel activity in arterialmyocytes depends on Hif1α, possibly due to the expression of Hif1α-dependent mitochondrial MCIV subunit isoforms. Among the MCIV subunits whose mRNA expression was modulated by Hif1αin arterial myocytes (see Fig. 8b, c), we focused on Cox4i2, a tissue-specific Cox4 isoform that is highly expressed in the lung75 and the carotid body66,67, and which is induced by hypoxia in several cell classes in a Hif1α-dependent manner76. Expression of Cox4i2 instead of Cox4i, the most common subunit isoform, alters cytochrome c oxidase (CCO) activity75–77 and is essential for ROS production and depolarization of cultured pulmonary VSM cells in response to hypoxia78. Mice lacking Cox4i2show abolition of the acute phase of hypoxic pulmonary vasoconstriction78 and strong inhibition of the acute O 2 -sensing properties of CB glomus cells68. To analyze the participation of Cox4i2 in the acute O 2 -modulation of Ca2+ channel activity in systemic arterial myocytes, we generated adult conditional Cox4i2-deficient mouse models using either TMX-activatable generalized Ubiquitin (UBC)-Cre or smooth muscle-specificSM-Cre (COX4I2-UBC and COX4I2-SM mice, respectively) (Fig. 9a). In the two mouse models, we observed a clear selective decrease in Cox4i2 mRNA in aortic myocytes (Fig. 9b, c). In addition, we showed a decrease in Cox4i2 immunostaining in Cox4i2-deficient femoral myocytes (Fig.9d). COX4I2-SM miceshowed normal body weight, organ size and hematocrit (Supplementary Figs. 17a–c). However, we observed a strong inhibition of the hypoxic modulation of Ca2+ currents measured in dispersed Cox4i2-deficient femoral VSM cells (Fig. 9e, f; Supplementary Figs. 17d, e). Moreover, endothelium-denuded femoral arteries removed from these mice had normal passive diameters and contractile responses to FPL and high K+,butwere,nevertheless, practically insensitive to hypoxia (Fig. 9g–j). These observations suggest that the Hif1α-dependent expression of Cox4i2 in vascular myocytes plays a fundamentalrole in the acute modulation of Ca2+ channel activity by PO 2 and in HVD. Discussion Two main conclusions can be drawn from the studies presented in this paper. The firstisthatacutemodulationofL-typeCa 2+ channels by PO 2 in systemic arterial myocytes depends on the integrated functions of mitochondria working as O 2 sensors and effectors. The O 2 modulation of Ca2+ channels is abolished in MCI-deficient cells, but can be rescued by the transgenic expression of a yeast NADH dehydrogenase, and mimicked by the intracellular application of NADH and H 2 O 2 . Remarkably, modulation of Ca2+ channels by O 2 depends on dynamic changes in the mitochondrial ETC, and requires a permanent, constitutive activation ofHif1α, which in turn determines the expression of atleast one atypicalMCIVsubunit isoform.Itmustbe alsostressedthat our electrophysiological studies, done on VMS cells dialyzed with 4 mM MgATP, strongly suggest that changes in intracellular ATP levels21,32,35,41,54,55 are not involved in inhibition of Ca2+ channel activity and depolarization-induced Ca2+ influx during hypoxia. These findings support a mitochondrial-to-membrane signaling (MMS) model of acute O 2 sensing44,68 in which the decrease in MCIV activity under hypoxia leads to a backlog of electrons along the ETC and an increase in the reduced ubiquinone (CoQH 2 ) pool. This results in a slowdown of MCI activity and the accumulation of NADH and H 2 O 2 which interact with Ca2+ channels (Fig. 10). This model provides a mechanistic 5 min FPL (1 PM) 0.05 ΔRatio F/F 340 380 5 min Hypoxia MCI/NDI1-SM Hypoxia Hypoxia 1 mM DKB 1 arb. units Hypoxia Hypoxia 1 mM DKB Hypoxia Hypoxia 1 mM DKB WT MCI-SM MCI/NDI1 -SM 0.05 ΔRatio F/F 340 380 RTN (1 PM) FPL (1 PM) a 2+ Single myocyte cytosolic Ca NADH autofluorescence e 1 min b RTN (1 PM) % decrease 'ratio 340/380 'ratio 340/380 FPL (1 PM) % decrease 'ratio 340/380 hypoxia NADH (arb. units) hypoxia cd f MCI/NDI1-SM WT MCI-SM MCI/NDI1-SM Fig. 7 | Transgenic expression of yeast Ndi1 rescues O 2 modulation of cytosolic Ca2+ and NADH signals in MCI-deficient arterial myocytes. a Changes in the FPL- evoked increase of cytosolic [Ca2+] by hypoxia (left) and 1 μMrotenone(right),in intact Fura-2-loaded femoral arterial myocytes from MCI/NDI1-SM mice. bBox plot of increases in cytosolic [Ca2+] elicited by 1 μM FPL in dispersed myocytes from WT (n= 31/3 cells/mice), MCI-SM (n= 38/4 cells/mice) and MCI/NDI1-SM (n= 21/3 cells/ mice) mice. cBox plot representing the effect of hypoxia on the FPL-evoked increase of cytosolic [Ca2+], shown as percentage decrease of the peak FPL-induced signal in Fura-2-loaded dispersed myocytes from WT (blue dots, n=31/3cells/ mice), MCI-SM (orange dots, n= 38/4 cells/mice) and MCI/NDI1-SM (green dots, n= 21/3 cells/mice) mice. dQuantification of the effect of rotenone on the FPL- evoked increase of cytosolic [Ca2+] in dispersed myocytes from WT (n = 24/4 cells/ mice), MCI-SM (n= 17/3 cells/mice) and MCI/NDI1-SM (n= 16/3 cells/mice) mice. eHypoxia and αKB-induced changes in NADH autofluorescence (AF in arbitrary units, arb. units) recorded in arterial myocytes from WT, MCI-SM and MCI/NDI1-SM mice. fBox plot showing the maximal increase in NADH autofluorescence (arb. units)inresponsetohypoxia,inarterialmyocytesfromWT(n= 19/7 cells/mice), MCI-SM (n =22/9 cells/mice) and MCI/NDI1-SM (n= 23/9 cells/mice) mice. In ddata are presentedasmean ± SEMwithstatistically significantPvalues(<0.05) indicated. Comparisons were performed by one-way ANOVA followed by Dunnett’s multiple comparisons post hoc test. Box plots in b,cand frepresent median (middle line), 25th, 75th percentile (box), and largest and smallest values range (whiskers). P values were calculated with the Kruskal-Wallis test followed by Dunn’s multiple comparisons post hoc test. Source data are provided as a Source Data file. Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 9 37. Herrera, G. M. & Walker, B. R. Involvement of L-type calcium channels in hypoxic relaxation of vascular smooth muscle. J. Vasc. Res. 35,265–273 (1998). 38. Pearce,W.J.,Ashwal,S.,Long,D.M.&Cuevas,J.Hypoxiainhibits calcium influx in rabbit basilar and carotid arteries. Am. J. Physiol. 262,H106–H113 (1992). 39. Franco-Obregon, A., Ureña, J. & Lopez-Barneo, J. Oxygen-sensitive calcium channels in vascular smooth muscle cells and their possible role in hypoxic arterial relaxation. Proc.NatlAcad.Sci.USA92, 4715–4719 (1995). 40. Franco-Obregon, A. & Lopez-Barneo, J. Differential oxygen sensitivity of calcium channels in rabbit smooth muscle cells of conduit and resistance pulmonary arteries. J. Physiol. 491, 511–518 (1996). 41. Quayle, J. M., Turner, M. R., Burrell, H. E. & Kamishima, T. Effects of hypoxia, anoxia, and metabolic inhibitors on K ATP channels in rat femoral artery myocytes. Am.J.Physiol.HeartCirc.Physiol.291, H71–H80 (2006). 42. Moreno-Domínguez, A., Colinas, O., Smani, T., Ureña, J. & López- Barneo, J. Acute oxygen sensing by vascular smooth muscle cells. Front. Physiol. 14, 1142354 (2023). 43. Ortega-Sáenz, P., Pardal, R., García-Fernández, M. & López-Barneo, J. Rotenone selectively occludes sensitivity to hypoxia in rat carotid body glomus cells. J. Physiol. 548,789–800 (2003). 44. Fernández-Aguera, M. C. et al. Oxygen sensing by arterial chemoreceptors depends on mitochondrial complex I signaling. Cell. Metab. 22,825–837 (2015). 45. Jiménez-Gómez, B. et al. Transgenic NADH dehydrogenase restores oxygen regulation of breathing in mitochondrial complex I-deficient mice. Nat. Commun. 14, 1172 (2023). 46. Kashani-Poor, N., Zwicker, K., Kerscher, S. & Brandt, U. A central functional role for the 49-kDa subunit within the catalytic core of mitochondrial complex I. J. Biol. Chem. 276,24082–24087 (2001). 47. Arias-Mayenco, I. et al. Acute O 2 sensing: Role of coenzyme QH2/Q ratio and mitochondrial ROS compartmentalization. Cell. Metab. 28,145–158 (2018). 48. Solway, J. et al. Structure and expression of a smooth muscle cellspecificgene,SM2.2 alpha J. Biol. Chem. 270,13460–13469 (1995). 49. Li, L., Miano, J. M., Mercer, B. & Olson, E. N. (1996). Expression of the SM22 alpha promoter in transgenic mice provides evidence for distinct transcriptional regulatory programs in vascular and visceral smooth muscle cells. J. Cell. Biol. 132,849–859 (1996). 50. Cabello-Rivera, D. et al. Oxygen regulation of breathing is abolished in mitochondrial complex III-deficient arterial chemoreceptors. Proc.NatlAcad.Sci.USA119, e2202178119 (2022). 51. Franco-Obregón, A. & López-Barneo, J. Low PO 2 inhibits Ca2+ channel activity in arterial smooth muscle cells. Am. J. Physiol. 271, H2290–H2299 (1996). 52. Vincent, A. E. et al. The spectrum of mitochondrial ultrastructure defects in mitochondrial myopathy. Sci. Rep. 6, 30610 (2016). 53. Li, L., Shen, S., Bickler, P. & Jacobson, M. P. Searching for molecular hypoxia sensors among oxygen-dependent enzymes. Elife 12, e87705 (2023). 54. Dart, C. & Standen, N. B. Activation of ATP-dependent K+channels by hypoxia in smooth muscle cells isolated from the pig coronary artery. J. Physiol. 483,29–39 (1995). 55. Jackson, W. F. Hypoxia does not activate ATP-sensitive K+channels in arteriolar muscle cells. Microcirculation 7,137–145 (2000). 56. Hernansanz-Agustín, P. et al. Na+controls hypoxic signaling by the mitochondrial respiratory chain. Nature 586,287–291 (2020). 57. Zheng, W., Rampe, D. & Triggle, D. J. Pharmacological, radioligand binding, and electrophysiological characteristics of FPL 64176, a novel nondihydropyridine Ca2+ channel activator, in cardiac and vascular preparations. Mol. Pharmacol. 40,734–741 (1991). 58. Rampe,D.&Dage,R.C.FunctionalinteractionsbetweentwoCa 2+ channel activators, (S)-Bay K 8644 and FPL 64176, in smooth muscle. Mol. Pharmacol. 41,599–602 (1992). 59. Watanabe, C., Yamamoto, H., Hirano, K., Kobayashi, S. & Kanaide, H. Mechanisms of caffeine-induced contraction and relaxation of rat aortic smooth muscle. J. Physiol. 456,193–213 (1992). 60. Michelakis,E.D.,Thebaud,B.,Weir,E.K.&Archer,S.L.Hypoxic pulmonary vasoconstriction: redox regulation of O 2 -sensitive K+ channels by a mitochondrial O 2 -sensor in resistance artery smooth muscle cells. J. Mol. Cell. Cardiol. 37, 1119–1136 (2004). 61. Sommer, N. et al. Mitochondrial cytochrome redox states and respiration in acute pulmonary oxygen sensing. Eur. Respir. 36, 1056–1066 (2010). 62. Waypa, G. B. et al. Hypoxia triggers subcellular compartmental redox signaling in vascular smooth muscle cells. Circ. Res. 106, 526–535 (2010). 63. Sullivan, L. B. et al. Supporting aspartate biosynthesis is an essential function of respiration in proliferating cells. Cell 162, 552–563 (2015). 64. López-Barneo, J. & Simon, M. C. Cellular adaptation to oxygen deficiency beyond the nobel award. Nat. Commun. 11, 607 (2020). 65. Tian, H., Hammer, R. E., Matsumoto, A. M., Russell, D. W. & McKnight, S. L. The hypoxia-responsive transcription factor EPAS1 is essential for catecholamine homeostasis and protection against heart failure during embryonic development. Genes Dev. 12, 3320–3324 (1998). 66. Zhou, T., Chien, M. S., Kaleem, S. & Matsunami, H. Single cell transcriptome analysis of mouse carotid body glomus cells. J. Physiol. 594, 4225–4251 (2016). 67. Gao, L. et al. Gene expression analyses reveal metabolic specifications in acute O 2 -sensing chemoreceptor cells. J. Physiol. 595, 6091–6120 (2017). 68. Moreno-Dominguez, A. et al. Acute O 2 sensing through HIF2α- dependent expression of atypical cytochrome oxidase subunits in arterial chemoreceptors. Sci. Signal. 13, eaay9452 (2020). 69. Colinas, O., Moreno-Domínguez, A., Ortega-Sáenz, P. & López- Barneo, J. Constitutive expression of Hif2αconfers acute O 2 sensitivity to carotid body glomus cells. Adv. Exp. Med. Biol. 1427, 153–162 (2023). 70. Bishop, T. & Ratcliffe, P. J. Genetic basis of oxygen sensing in the carotid body: HIF2αand an isoform switch in cytochrome c oxidase subunit 4. Sci. Signal. 13,eaba1302(2020). 71. Semenza,G.L.,Roth,P.H.,Fang,H.M.&Wang,G.L.Transcriptional regulation of genes encoding glycolytic enzymes by hypoxiainducible factor 1. J. Biol. Chem. 269,23757–23763 (1994). 72. Rankin, E. B. et al. Hypoxia-inducible factor-2 (HIF-2) regulates hepatic erythropoietin in vivo. J. Clin. Invest 117,1068–1077 (2007). 73. Aras, S. et al. Oxygen-dependent expression of cytochrome c oxidase subunit 4-2 gene expression is mediated by transcription factors RBPJ, CXXC5 and CHCHD2. Nucleic Acids Res. 41, 2255–2266 (2013). 74. Lee,C.H.etal.(2001).Themechanismofphenylephrine-mediated Ca2+ oscillations underlying tonic contraction in the rabbit inferior vena cava. J. Physiol. 534,641–650 (2001). 75. Hüttemann, M. et al. Cytochrome c oxidase subunit 4 isoform 2-knockout mice show reduced enzyme activity, airway hyporeactivity, and lung pathology. FASEB J. 26,3916–3930 (2012). 76. Fukuda, R. et al. HIF-1 regulates cytochrome oxidase subunits to optimize efficiency of respiration in hypoxic cells. Cell 129, 111–122 (2007). 77. Pajuelo-Reguera, D. et al. Cytochrome c oxidase subunit 4 isoform exchangeresultsinmodulationofoxygenaffinity. Cells 9, 4432020 (2020). Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 16 78. Sommer, N. et al. Mitochondrial complex IV subunit 4 isoform 2 is essential for acute pulmonary oxygen sensing. Circ. Res. 121, 424–438 (2017). 79. Prange-Barczynska, M. et al. Hif-2αprogrammes oxygen chemosensitivity in chromaffincells.PreprintatbioRxiv,https://doi.org/ 10.1101/2023.07.14.548982 (2023). 80. Weir,E.K.,Lopez-Barneo,J.,Buckler,K.J.&Archer,S.L.(2005). Acute oxygen-sensing mechanisms. N. Engl. J. Med. 353, 2042–2055 (2005). 81. Scortegagna, M. et al. Multiple organ pathology, metabolic abnormalities and impaired homeostasis of reactive oxygen species in Epas1−/−mice. Nat. Gen. 35, 331–340 (2003). 82. Gruber, M. et al. Acute postnatal ablation of Hif-2αresults in anemia. Proc.NatlAcad.Sci.USA104,2301–2306 (2007). 83. Chiamvimonvat, N. et al. Functional consequences of sulfhydryl modification in the pore-forming subunits of cardiovascular Ca2+ and Na+channels. Circ. Res. 76,325–334 (1995). 84. Fearon, I. M. et al. Modulation of recombinant human cardiac L-type Ca2+ channels α1C subunits by redox reagents and hypoxia. J. Physiol. 514,629–637 (1999). 85. Muralidharan,P.,Szappanos,H.C.,Ingley,E.&Hool,L.Evidencefor redox sensing by a human cardiac calcium channel. Sci. Rep. 6, 19067 (2016). 86. Kilfoil, P. J., Tipparaju, S. M., Barski, O. A. & Bhatnagar, A. Regulation of ion channels by pyridine nucleotides. Circ. Res. 112,721–741 (2013). 87. Jackson, W. F. K v channels and the regulation of vascular smooth muscle tone. Microcirculation 25, https://doi.org/10.1111/micc. 12421 (2018). 88. Raph,S.M.,Dwenger,M.M.,Hu,X.&Nystoriak,M.A.BasalNAD(H) redox state permits hydrogen peroxide-induced mesenteric artery dilatation. J. Physiol. 601,2621–2634 (2023). 89. Yang, Q. et al. Clustering of L-type calcium channels by Shank3. J. Neurochem. 167,16–37 (2023). 90. Olson, K. R. A case for hydrogen sulfide metabolism as an oxygen sensing mechanism. Antioxid. (Basel) 10,1650(2021). 91. Segal, S. S. Regulation of blood flow in the microcirculation. Microcirculation 12,33–45 (2005). 92. Ryan, H. E. et al. Hypoxia-inducible factor-1αis a positive factor in solid tumor growth. Cancer Res.60,4010–4015 (2000). 93. Hayashi, S. & McMahon, A. P. Efficient recombination in diverse tissues by a tamoxifen-inducible form of Cre: A tool for temporally regulated gene activation/inactivation in the mouse. Dev. Biol. 244, 305–318 (2002). 94. Ruzankina, Y. et al. Deletion of the developmentally essential gene ATR in adult mice leads to age-related phenotypes and stem cell loss. Cell Stem Cell 1, 113–126 (2007). Acknowledgements This research was supported by the Spanish Ministries of Science and Innovation and Health (Grants SAF2016-74990-R, JL-B. and PID2019- 106410RB-I00, J.L.-B. funded by MCIN/AEI/10.13039/501100011033, and the European Research Council (ERC Advanced Grant PRJ201502629, J.L.-B.). We thank Dr. Juan Luis Ribas (CITIUS, University of Seville) for help with electron microscopy. We also thank the staff of IBiS and “Centro de Producción y Experimentación Animal Oscar Pintado”for technical assistance. Author contributions A.M.-D., O.C., I.A.-M., J.M.C. J.L.L.-O., A.P. and J.L.-B. performed the experiments and/or analyzed the data. N.C., N.W. and N.S. provided essential material for the experiments. All authors contributed to the design of the study and to the discussion of the data. A.M.-D. and J.L.-B generated the draft of the paper. J.L.-B. coordinated the project and the writing of the paper. Competing interests The authors declare no competing interests. Ethics approval This study was performed following the recommendations set out in the Global Code of Conduct for Research in Resource-Poor Settings. Additional information Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41467-024-51023-3. Correspondence and requests for materials should be addressed to José. López-Barneo. Peer review information Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available. Reprints and permissions information is available at http://www.nature.com/reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by-nc-nd/4.0/. © The Author(s) 2024 Article https://doi.org/10.1038/s41467-024-51023-3 Nature Communications | (2024) 15:6649 17