Carbon dioxide is a triple vasodilator

Carbon dioxide (CO₂) can regulate blood flow and is applied therapeutically in intensive care units to treat brain injury, as well as in balneotherapy for peripheral arterial disease (PAD) and diabetic angiopathy; however, its mode of action remains unclear.

The vasoactive CO₂ effects were tested in arteries of healthy C57BL/6J mice, hypertensive apolipoprotein E-deficient mice, and soluble guanylyl cyclase (sGC) knockout mice in a small vessel myograph with and without pharmacologically intervening in endothelium- and/or vascular smooth muscle-mediated vasodilation. CO₂-based near-infrared spectroscopy (NIRS-CO₂) was developed to assess vasoreactivity of the skin microcirculation to CO₂ in healthy individuals, PAD and coronary artery disease (CAD) patients, and was compared with flow-mediated dilation (FMD). We identified CO₂ as a triple vasodilator mimicking the actions of endothelium-derived relaxing factor (nitric oxide, NO), endothelium-derived hyperpolarization factor, and direct myogenic vasodilators. CO₂ engaged endothelial NO/sGC, endothelial small-/intermediate-conductance calcium-activated potassium channels (SK_Ca/IK_Ca), and myogenic voltage-gated (K_V) and IK_Ca potassium channels, respectively, acting as a triple vasodilator. CO₂-evoked vasodilator responses were blunted and delayed in diseased human and murine arteries. In the human cohort, the NIRS-CO₂-derived time-to-intersection of the HbO₂ and HHb curves, capturing the delay phenotype, showed a strong association with PAD/CAD status and, in exploratory analyses, also distinguished young individuals with cardiovascular risk factors, supporting NIRS-CO₂ as a physiological readout that integrates endothelial and myogenic components of microvascular reactivity. Duration and extent of CO₂ vasodilation were coupled to tissue metabolism through vascular carbonic anhydrases (CAs), providing a mechanism for vasculometabolic coupling and one for clinically approved CA inhibitors.

NIRS-CO₂ provides a feasible readout of CO₂-evoked microvascular responsiveness and shows disease-associated alterations in our PAD/CAD cohort. Larger studies will validate generalizability across vasculopathies and clarify the relative contributions of NO-sGC vs. K⁺ channel-linked mechanisms for future therapeutic translation.