Effect of plasma protein binding on elimination of taurocholate by isolated perfused rat liver: comparison of venous equilibrium, undistributed and distributed sinusoidal, and dispersion models.

In the past, various models have been developed to allow better characterization of the hepatic elimination of substrates from plasma. In this study we investigated the applicability of the venous equilibrium, undistributed sinusoidal, several distributed sinusoidal, and dispersion models to the steady state elimination of sodium taurocholate by the isolated perfused rat liver. Rat livers were perfused with 24-14C-taurocholate (sodium salt) at a concentration of 25 microM (specific activity 500 microCi/mmole) in a single-pass design (n = 7) or at a rate of 0.5 mumol/min (specific activity 40 microCi/mmole) into the portal vein in a recirculating design (n = 5). In single-pass experiments, the changes in hepatic venous outflow concentration (C0) with changes in unbound fraction of taurocholate (fu) from 0.09 to 1.0 were fitted better by the venous equilibrium model, by the dispersion model, and by a distributed model in which heterogeneity in both hepatic blood flow (Q) and intrinsic clearance (CLint) was defined by separate density functions. The very large value of dispersion number (DN greater than 10(7] yielded by the dispersion model is consistent with a high degree of axial mixing of blood within sinusoids. The large coefficients of variation (0.7-232) for the density functions describing the transverse heterogeneity of Q and CLint obtained with the Q/CLint-distributed model were consistent with a large degree of heterogeneity in Q and CLint within the liver. In recirculation experiments, the steady state unbound concentration of taurocholate in the reservoir (Cuss) was independent of fu (range 0.05-0.9). This finding was not predicted by the undistributed sinusoidal model, but was in keeping with the venous equilibrium model, with the dispersion model, and with the Q/CLint-distributed model. Therefore, there is no need to invoke cell surface-mediated dissociation of albumin-ligand complexes in hepatic taurocholate uptake. As the dispersion and Q/CLint-distributed models are conceptually plausible and operationally accurate, it may be time to relinquish the venous equilibrium model, which, though operationally accurate, is conceptually flawed.