One of the many unresolved questions regarding the permeability of the glomerular filtration barrier is the reason behind the marked difference in permeability between albumin and polysaccharide probe molecules such as Ficoll and dextran of the same molecular size. Although the differences in permeability have been mainly attributed to charge effects, we have previously shown that this would require a highly charged filtration barrier, having a charge density that is ~10 times more than that on the albumin molecule. In this article, the classic two-pore model was extended by introducing size distributions on the solute molecules, making them conformationally flexible. Experimental sieving data for Ficoll from the rat glomerulus and from precision-made silicon nanopore membranes were analyzed using the model. For the rat glomerulus a small-pore radius of 36.2 Å and a geometric standard deviation (gSD) for the Ficoll size-distribution of 1.16 were obtained. For the nanopore membranes, a gSD of 1.24 and a small-pore radius of 43 Å were found. Interestingly, a variation of only ~16% in the size of the polysaccharide molecule is sufficient to explain the difference in permeability between albumin and Ficoll. Also, in line with previous data, the effects of applying a size distribution on the solute molecule are only evident when the molecular size is close to the pore size. Surely there is at least some variation in the pore radii, and, likely, the gSD obtained in the current study is an overestimation of the "true" variation in the size of the Ficoll molecule.
Keywords: heteroporous model; pore model; sieving coefficient.