LAUSR

In Response: This response letter addresses discussion initiated by Hüttemann et al in their Letter to the Editor, in response to our article “Ndufs2, a Core Subunit of Mitochondrial Complex I, Is Essential for Acute Oxygen-Sensing and Hypoxic Pulmonary Vasoconstriction”.1 We stand by the results presented in the original publication and wish to expand our rationale about the screening of COX4i2 (cytochrome c oxidase subunit 4 isoform 2) in vitro, and further the discussion of preclinical models of the complex pulmonary vascular oxygen-sensing mechanism. Ndufs2 (NADH dehydrogenase [ubiquinone] ironsulfur protein 2) was identified as a candidate oxygen-sensor as a result of over 3 decades of research demonstrating that the Complex I inhibitor rotenone, mimics hypoxic pulmonary vasoconstriction and recapitulates the effects of hypoxia in other oxygen-sensing tissues, including type 1 glomus cells of the carotid body2 and adrenomedullary chromaffin cells.3 As the quinone binding site within Complex I, Ndufs2 is a primary driver of Complex I activity. It is also the rotenone binding site4 and is both redox-sensitive5 and a target of posttranslationalmodification.6 Our finding that Ndufs2 is essential for acute oxygen-sensing in the pulmonary vasculature reinforces previously published work by FernándezAgüera et al7 identifying Ndufs2 as the oxygen-sensor of the carotid body, as well as the hypothesis that oxygensensing tissues have common sensor(s).8 We also made every effort to examine the theories of other groups by screening multiple subunits implicated in pulmonary vascular oxygen-sensing beyond our own hypothesis, including the Riekse Fe-S subunit9 and COX4i2.10 We do not believe the COX4i2 data in our original article was misinterpreted, given the information available at the time of publication. While we cannot comment on unpublished data from other groups, we found that an ≈50% knockdown of COX4i2 (verified at the gene and protein level with appropriate statistical replicates) did not result in any reduction of hypoxic pulmonary vasoconstriction, dose-dependent or otherwise. It is also pertinent to address the importance of modelling robust hypoxic pulmonary vasoconstriction as well as significant differences in experimental design used by various groups. While Sommer et al presents a thorough examination of COX4i2 in oxygen-sensing, it cannot be ignored that the range for ∆PAP in their acute hypoxia studies spans only 0 to 1 mm Hg.10 ∆PAP of 1.0±0.11 mm Hg for wild-type mice is far below the previously published hypoxia-induced ∆PAP of ≈10 mm Hg reported by Schwenke et al,11 and thus it remains difficult to interpret the physiological relevance of this model with such small hypoxic changes in mPAP. Further, since in vitro and in vivo models as well as experimental conditions, including degree of hypoxia, duration of hypoxic exposure, and numerous measurement techniques/ assays are not identical, somewhat divergent results with respect to the exact identity of the oxygen-sensor, are not entirely surprising. In the fullness of time, it is likely to be confirmed that all proposed subunits play different roles in pulmonary vascular oxygen-sensing, providing a degree of redundancy similar to that found in other biological systems.12 The evolution of the vascular oxygen-sensing field has resulted in a large degree of consensus: it is widely accepted that the mitochondrion is the oxygen-sensor, that at least one (and likely multiple) electron transport chain protein(s) sense changes in PO2 in a redoxdependent manner, and that mitochondrially derived reactive oxygen species are the secondary messenger, altering the cellular milieu to stimulate the closing of voltage-gated potassium (Kv) channels, opening of voltage-gated L-type calcium channels, and subsequent vasoconstriction. In conclusion, we appreciate the concerns raised by Hüttemann et al and encourage other groups to consider