Capillary network structure does not affect theoretical analysis of glomerular size selectivity.

Abstract

Anatomical studies have demonstrated that the glomerular capillaries are complex and heterogeneous networks. Conventional models of glomerular size selectivity, however, are based on the assumption of simplified geometries. We developed a theoretical model of glomerular size-selective function based on the geometric data obtained in a previous reconstruction of a glomerular network from a normal Munich-Wistar rat. This heterogeneous model was compared with the homogeneous model conventionally used to calculate membrane selective parameters from the fractional clearance of two test solutes, neutral dextran and Ficoll. For both models we assumed a hypothetical log-normal distribution of pore sizes and calculated optimal membrane pore-size parameters using previously published values of fractional clearances. The difference between the sieving coefficients calculated with the two models was negligible, never exceeding 5.5%. Since the homogeneous model does not consider the pressure drop along the glomerular capillary, we also computed fractional clearances with the homogeneous model, assuming the same pressure drop as in the heterogeneous one. The differences in computed fractional clearances using the homogeneous model with and without a pressure drop were less than 1.2%. We concluded that models based on identical capillary networks can therefore be used for interpreting sieving coefficients for macromolecules.