The necessity for defining hypoxia as O2-limited energy flux rather than low partial pressure is explored from a systems perspective. Oxidative phosphorylation, the Krebs cycle, glycolysis, substrate supply, and cell energetics interact as subsystems; the set point is a match between ATP demand and aerobic ATP production. To this end the transport subsystem must match the transcapillary and mitochondrial O2 fluxes. High transcapillary O2 flux requires intracellular PO2 in the range 1-10 Torr. In this range the O2 drive on electron transport must be compensated by adaptive changes in the phosphorylation and redox drives. Thus the metabolic subsystem supports diffusive O2 transport by maintaining O2 flux at intracellular partial pressures required for O2 release from blood. Since responses to stress are distributed according to the state of the entire system, several simultaneous metabolic measurements, including intracellular PO2 (or a known direction of change in intracellular PO2) and the O2 dependence of a measurable function are required to judge the adequacy of O2 supply. ATP demand and aerobic capacity must also be evaluated, because the hypoxic threshold depends on the ratio of ATP demand to aerobic capacity. The application and limitation of commonly used criteria of hypoxia are discussed, and a more precise terminology is proposed.

Erythrocyte-containing capillaries were counted in dog gracilis muscles freeze-clamped at rest and after twitch contraction at 4/s. In each of 21 muscles, 6--8 blocks were examined at -70 degrees C without fixation or staining; 15 fields were counted per block. Frequency analysis of capillaries per field based on the negative binomial distribution indicated that capillary density at rest was controlled by arterioles. Active vasomotion of these arterioles was "switched off" within 5 s after onset of exercise. Capillary density was then determined passively by stochastic rheologic factors acting at the individual capillaries. Thus exercise changes the site and the mechanism of capillary control. Recruitment occurred first where capillary density was lowest, and was complete in 15 s; this greatly decreased the heterogeneity of capillary spacing. Mean capillary density increased 1.5- to 3-fold, whereas flow increased almost 7-fold. Calculated mean velocity and mean transit time of erythrocytes in capillaries were 1.1 mm/s and 920 ms at rest and 4.2 mm/s and 215 ms after 3 min of exercise.

Capillary control was quantified in dog gracilis muscles from in situ. About 550 capillaries/mm2, one-third the total number present, were perfused with erythrocytes simultaneously at rest; two-thirds the total could be perfused during maximal vasodilation. The functional capillary reserve was about 600/mm2. Capillary distribution at rest reflects a passive, random process at individual capillaries and an active process that coordinates perfusion of small groups of capillaries. The latter creates long diffusion distances. These are unaltered by denervation, or flow per se, but are abolished by adenosine. Twitch contraction at 4/min recruited about 400 capillaries/mm2 without any change in flow. Capillaries opened selectively where diffusion distances were longest. The same changes occurred within 5 s during work at 4/s, even if flow was held constant. If flow could increase, about 200 additional capillaries/mm2 were slowly recruited, without change in capillary distribution. Conclusions are that 1) hemodynamics and active vasomotion contribute equally to capillary density at rest; 2) active papillary control in exercise is ungraded and solely responsible for eliminating metabolically significant diffusion paths; 3) flow and capillary density can be controlled independently by proximal and terminal arterioles, respectively.

The intracellular distribution of O2 in cross sections of dog gracilis muscles was determined by myoglobin (Mb) cryospectrophotometry. The volume sampled by the photometer was approximately 30 micron3 and contained 1-2 mitochondria. Measurements could be made to within 3 micron of capillaries without interference from hemoglobin. Mb saturation was uniform at all loci examined when respiration was blocked with cyanide. During twitch contraction at maximum O2 consumption, saturations within a cell cross section varied by up to 20%. The corresponding difference in partial pressure of O2 (PO2) was 1.5 Torr. Circumferential O2 gradients parallel to and 5 micron from the sarcolemma were greatest near capillaries. They did not exceed 0.1 Torr/micron and were dissipated within 25 micron of the sarcolemma. Gradients perpendicular to the sarcolemma were less than 0.02 Torr/micron. Saturation was not significantly correlated with cell diameter. Minimum PO2 was seldom located at the center of the cell cross section. Differences in saturation between contiguous cells often exceeded 10%. The distribution of O2 within cells appeared to reflect both an intercellular O2 flux and and an O2 flux from adjacent capillaries. Data agree qualitatively and quantitatively with mathematical models that take account of the particulate nature of blood and facilitated diffusion by Mb.

Myocardial tissue pressure increases from epicardium to endocardium, and in the deeper layers exceeds ventricular blood pressure during one-third of the cardiac cycle (21). The effect of this tissue pressure gradient on local blood flow was studied using the depot clearance technique. Blood flow was found to be at least 25% lower in the deep regions as compared with superficial ones. With total coronary inflow held constant, vagal arrest of the heart removed the tissue pressure gradient, and simultaneously redistributed flow from superficial to deeper layers. We conclude that the gradient in tissue pressure, and hence in the extravascular component of coronary resistance, is at least in part, the cause of the nonhomogeneous blood flow across the wall. By use of the oxygen cathode, a gradient of oxygen tensions was observed which paralleled the blood flow gradient; mean oxygen tension in the subepicardium averaged twice that in the subendocardium. The gradient in oxygen tension appears to be of sufficient magnitude to determine a transmural gradient in aerobic metabolism.