Making sense out of oxygen sensor.

Autoregulatory mechanisms that adjust the diameter of a vascular bed and the blood flow through it to the intrinsic parameters of tissue oxygen demand have become a cornerstone of vascular physiology.1–4 Although examples of hypoxia-induced vasoconstriction do exist (such as pulmonary microvasculature), most blood vessels respond to hypoxic or ischemic environment with vasodilation. Coronary arteries are not an exception to this rule; in fact, autoregulation in this vascular bed is of paramount importance. Increased myocardial demand for oxygen cannot be met by increasing oxygen extraction, because coronary arteriovenous difference is already high under normal conditions. Therefore, matching the supply depends almost entirely on the ability to increase coronary blood flow. Although the occurrence of hypoxiaand ischemia-induced coronary vasodilation has been well established, the number of theories attempting to explain the phenomenon has multiplied. Proposed physiological models include (1) direct sensing of reduced PO2 by different cellular elements of the vascular wall5; (2) production and release of vasodilatory metabolites by the oxygen-deprived myocardium; (3) changes in intracellular calcium or proton metabolism and distribution, or a rapidly developing deficiency in high-energy phosphates, which suppress the contractile apparatus of the smooth muscle cells5,6; and (4) a shift in the affinity of hemoglobin for nitric oxide (NO).7 In this issue of Circulation Research, Shimizu et al8 provide further insight into the mechanisms of hypoxic vasorelaxation in porcine coronary arteries. Previous studies by Daut et al,9 performed in guinea pigs, demonstrated the existence of glibenclamide-inhibitable coronary vasodilation in response to hypoxia or ischemia. On the basis of these findings, Shimizu et al8 launched a systematic investigation of the effect of K channel inhibitors on hypoxic vasorelaxation. The authors used inhibitors of several classes of K channels, specifically, Ca-dependent (tetraethylammonium, apamin, and charybdotoxin), voltagedependent (4-aminopyridine), ATP-sensitive (glibenclamide), and inward rectifier (BaCl2), only to conclude that this mechanism is not operant in porcine coronary vessels. In addition, studies using calciumand pH-sensitive fluorophores showed the lack of a clear correlation between changes in these parameters and hypoxic relaxation. Furthermore, hypoxic vasorelaxation did not appear to be related to the loss of endogenous phosphagens by porcine coronary arteries. Paradoxically, the force of smooth muscle contraction, but not ATP utilization, was increased under hypoxic conditions. To complete the picture, it is important to note that previous studies from the same laboratory using various bona fide pharmacological inhibitors have investigated the validity of several additional candidates for the role of oxygen sensor, including the Na pump, eicosanoids, hydrogen peroxide, superoxide, and protein kinases C and G, all of which have failed the selection process for an unambiguous factor responsible for hypoxic vasorelaxation.10 This exhaustive analysis is impressive, and the derived long list of trial-and-error eliminated candidates is palpably helpful, because it provides an ample opportunity to revisit the problem of oxygen sensing and vascular adaptation to oxygen deprivation. Under the circumstances, it may be time to peer down the evolutionary road to oxygen sensing in bacteria and yeast in an attempt to discern the silhouettes of modern oxygensensing systems, such as NAD(P)H oxidase, heme proteins, and P-450 enzymes, as has been done in a recent article by Bunn and Poyton.11 Perhaps the most recent addition to this family of oxygen sensors or effectors, which I find intriguing, has been described in Drosophila.12 Hypoxia produced a series of rapid behavioral changes and induced cell cycle arrest, processes that were diminished by an inhibitor of NO synthase (NOS) and by inactivating mutations in the gene encoding for the cGMP-dependent protein kinase. Moreover, both shortand long-term responses to oxygen deprivation in Drosophila were dependent on NO. Although these examples of hypoxic adaptation are evolutionarily remote, I strongly believe that they may contribute fundamentally to our understanding of the problem of hypoxic relaxation of mammalian arteries. This optimism is based on the following circumstantial evidence. Tissue responses to hypoxia are orchestrated by NO at different levels: (1) SNO-Hb[Fe(II)]O2 hemoglobin releases NO at decreased PO2; (2) glibenclamide-insensitive component of vasodilation could be due to the effect of low PO2 on the vascular endothelium to generate NO, as suggested by Daut et al9; (3) NO can be displaced by carbon monoxide from the preexisting, probably heme-bound, cellular pool, where it is constantly replenished by the functioning endothelial NOS, thus exerting a rapid vasodilatory effect independent on the enzymatic machinery, as shown by Thorup et al13; and (4) reduction in myocardial O2 consumption, so important in preconditioning, appears to be dependent on NO.14 All of these data point toward the possibility of NO mediation of adaptive responses to hypoxia. The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association. From the Departments of Medicine and Physiology and Biophysics, State University of New York, Stony Brook, NY. Correspondence to Michael S. Goligorsky, Health Science Center, Division of Nephrology, State University of New York, Stony Brook, NY 11794-8152. E-mail [email protected] (Circ Res. 2000;86:824-826.) © 2000 American Heart Association, Inc.