Erythrocytes: surveyors as well as purveyors of oxygen?

FOR MORE THAN A CENTURY a large body of evidence has accumulated in the physiological and pathophysiological literature to support the close relationship between the demand of oxygen by the tissues and the supply of oxygen by the cardiorespiratory system. When the demand for oxygen changes, how is its supply altered to match the increased demand and what are the signals by which the increased demand is recognized and then communicated to the appropriate effector cells? In particular, how does the cardiovascular system know how to respond to an imbalance between the demand for and supply of oxygen? A number of different mechanisms have been proposed to explain the response of the cardiovascular system to local imbalances in oxygen supply and demand, under the general heading of the local regulation of blood flow (14). One such mechanism, the metabolic hypothesis for local control of blood flow, posits that the concentration of vasodilator products of cellular metabolism in the interstitial fluid which bathes the resistance vessels is the chemical signaling link between blood flow and metabolism. A long list of vasodilator mediators have been proposed over the years, with adenosine being one of the most widely investigated. Although this regulatory mechanism works from a qualitative vantage point, critical quantitative tests have not provided definitive support. Searches for a specific oxygen sensor, including its location and sensitivity to changes in oxygenation [oxygen tension (PO2) or hemoglobin oxygen saturation (SO2)?], have not yielded firm support either (2, 11). More recent entries into this field, such as nitrosothiols (S-nitrosohemoglobin, 20) and nitrite (8), have garnered some support but so far have not passed all the necessary tests, and their roles remain uncertain. The involvement of nitric oxide is particularly intriguing since nitric oxide can modulate both the supply of oxygen through its vasodilatory action on resistance vessels as well as the demand for oxygen through its inhibitory action on oxygen consumption through cytochrome c oxidase in the mitochondria (1). Ellsworth and colleagues (7) proposed the novel idea that red blood cells (RBCs) act as mobile sensors for local oxygenation by the oxygen-dependent release of the vasodilator ATP from RBCs (12). One way to quantify oxygen transport is to employ Fick’s principle to relate the variables of oxygen consumption (V̇O2), convective oxygen supply (QO2 Q[O2]), and oxygen extraction {EO2 ([O2]in [O2]out)/[O2]in} by the tissue (14). This “black box” approach works particularly well at the level of the whole organism and individual organs, although defining an appropriate reference volume of tissue involved in oxygen consumption becomes more problematic at the level of microvascular networks and single microvessels. When oxygen demand changes, adjustments in oxygen supply and/or oxygen extraction take place in an attempt to match the altered demand, to the extent that they can. Oxygen delivery or supply is determined mostly by the arterioles, whereas the site of oxygen extraction is distributed primarily between the arterioles and capillaries. Thus the local regulation of arteriolar and capillary perfusion must play a key role in matching oxygen supply to altered demand. Looking into the “black box”—the microcirculation—using current methods of intravital microscopy reveals the presence of significant heterogeneity in hemodynamics (i.e., RBC velocity, lineal density, and supply rate) and oxygenation (i.e., PO2 and SO2) in arterioles and capillaries (3–5, 9, 13, 17). Since the hemoglobin in RBCs carries about 98% of the oxygen in blood and its ability to release oxygen to the tissues is modulated by allosteric effectors such as temperature, pH, and PCO2 through shifts in the oxygen dissociation curve, it is clear that the RBCs play a dominant role in the supply of oxygen to tissues. Mathematical models of oxygen transport have played key roles in contributing to our understanding of oxygen transport and its regulation (19). Ninety years ago August Krogh published his famous model of oxygen transport that involved a single capillary surrounded by a cylinder of oxygen-consuming tissue (15). Krogh’s original idea was that the oxygen supply to tissues was modulated by alterations in the number of bloodperfused capillaries linked to tissue activity (i.e., changing surface area available for oxygen diffusion by changing capillary density). However, Poole, Hudlicka, and others (18) have reported that most capillaries are perfused all the time, so that changes in the surface area for oxygen diffusion are more closely tied to the surface area corresponding to RBCs in contact with the capillary wall, as proposed by more recent studies (16, 21). The article by Ellis et al. (6) in this issue of the American Journal of Physiology-Heart and Circulatory Physiology addresses a number of the issues raised above, using a multifaceted approach in an animal model of prediabetes that combines in vitro work on ATP release by RBCs, in vivo measurements of oxygen transport in capillary networks, and mathematical modeling that incorporates the experimental findings. The idea tested was that prediabetes is associated with elevated insulin levels that might lead to the development of microvascular dysfunction. Since RBCs possess insulin receptors, the hyperinsulinemia in these animals might inhibit the release of ATP from the RBCs (10), preventing this mechanism from matching oxygen supply with oxygen demand. Two groups of male Zucker diabetic fatty rats were used as subjects, either leptin receptor knockouts (fa/fa) or controls (fa/ ). Hypoxia-induced ATP release from RBCs from each group of rats was similar, and prior incubation with insulin similarly prevented hypoxiainduced ATP release from RBCs in both groups. Thus the role of RBCs as oxygen sensors linked to ATP release appeared to be compromised by insulin. In vivo measurements of oxygen transport in capillaries of the extensor digitorum longus muscle further supported the hypothesis that the oxygen supply was significantly lower in prediabetic animals compared with conAddress for reprint requests and other correspondence: R. N. Pittman, Dept. of Physiology and Biophysics, School of Medicine, Virginia Commonwealth Univ., Richmond, VA 23298 (e-mail: [email protected]). Am J Physiol Heart Circ Physiol 298: H1637–H1638, 2010; doi:10.1152/ajpheart.00285.2010. Editorial Focus