Adenosine kinetics in canine coronary circulation

Adenosine kinetics in the coronary circulation were investigated in anesthetized closed-chest dogs by analysis of multiple-indicator dilution experiments. During simultaneous intracoronary bolus injections of 125I-labeled albumin, [14C]sucrose, and [3H]adenosine, dilution curves were measured by automated sampling of coronary venous blood, using high-performance liquid chromatography and isotope detection techniques. Under control conditions, only 1% of the injected [3H]adenosine was detected in coronary venous samples, compared with the reference tracer, [14C]sucrose, and the peak of the adenosine dilution curve preceded those of the reference tracers by 2-4 s. Optimized model fits to the control adenosine curves required the combination of high ratios of the capillary endothelial cell membrane permeability-surface area product to flow (5.3) and endothelial cell consumption capacity to flow (23), in addition to a broad heterogeneity of flow. Dilution curves were measured under control conditions and during increased coronary flow (nitroglycerin), inhibition of the enzyme adenosine kinase (iodotubercidin), and blockade of membrane adenosine transport (dipyridamole). Reliability of the parameter estimates was confirmed using residual analysis, sensitivity function analysis, and Monte Carlo simulation techniques. Accuracy of the model was confirmed in separate experiments in which nontracer adenosine was infused into the coronary artery and the model prediction of coronary venous adenosine concentrations was compared with measured values. Using published measurements of coronary blood flow and arterial and coronary venous plasma adenosine concentrations in open-chest dogs, the estimated in vivo interstitial adenosine concentration is 100-220 nM.