Modeling the oxyhemoglobin dissociation curve at the cerebral capillary level in humans

Hitherto, changes in the configuration of the oxyhemoglobin dissociation curve triggered by the increase in PCO2 and reduction in pH that occurs, as the red blood cell passes through target tissue capillaries, have largely been ignored in most human-experimental studies of the oxygen transfer from capillary blood to brain tissue mitochondria. In the present study, we developed a method that permits the calculation of the mean capillary P50 and the Hill slope, h, from paired arterial-jugular venous PO2 and SO2-values. By mathematical formalism, and by assuming a linear relationship between the oxygen tension and saturation in the Hill plot within the capillary range, we derived the following equation for the Hill slope (Figure 1): h = (ln(SartO2/(1-SartO2)) - ln(SjugO2/(1-SjugO2)))/(ln(PartO2)-ln(PjugO2)), The corresponding P50 was found using the Hill equation: P50 = PartO2 ((1-SartO2)/SartO2)^(1/h) From this, we found that the P50 was 3.7 (0.2) kPa and the Hill slope was 2.4 (0.2) in 30 healthy volunteers during resting breathing. These values differed from the standard arterial values of 3.5 kPa and 2.8 (p < 0.001 for both), which are often used when modeling the cerebral capillary transfer of oxygen in humans. The use of the calculated capillary P50 and Hill slope resulted in systematically higher capillary, PcapO2, and mitochondrial, PmitO2, oxygen tensions than when using the standard arterial values, with a bias of 1.4 kPa and limits of agreement of 0.5 kPa to 2.3 kPa for both PcapO2 (Figure 2) and PmitO2. The method presented here for modeling the oxyhemoglobin dissociation curve at the cerebral capillary level provides physiologically relevant estimates of PcapO2 and PmitO2, which are systematically higher than when assuming a ‘fixed’ oxyhemoglobin curve.