It is well known that macromolecules like albumin are markedly restricted in their passage across the glomerular capillary wall. However, the relative importance of solute size, charge and shape is currently debated since much of the previous work is based on dextran in neutral or charge-modified forms. These polymers have certain drawbacks that make them less suitable for analysis of capillary permeability and the notion of a glomerular charge barrier has therefore been questioned. Moreover, macromolecules larger than albumin (mol. wt. 69 000) have been suggested to pass through nonselective 'shunt' pathways. In order to study glomerular permeability, isolated rat kidneys were perfused with albumin solutions containing trace amounts of two differently radiolabelled isoenzymes of lactate dehydrogenase (LDH) at low temperature to inhibit tubular function. The isoenzymes have similar size (mol. wt. 140 000) and shape but differ in charge, one carrying a negative net surface charge (LDH1, − 19) and the other being slightly cationic (LDH5, + 2). The urine and perfusate samples were subjected to high pressure liquid chromatography (HPLC) gel-filtration to allow for measurements of intact LDH. The fractional clearance was 0.11% ± 0.04% for the anionic LDH1 and 0.56% ± 0.07% for LDH5, whereas that for albumin was 0.21% ± 0.03% at a glomerular filtration rate of 0.11 ± 0.01 mL min−1 g−1 kidney wet weight. The results were analysed using a homogenously charged membrane model and are compatible with a charge density of 35 mEq L−1, with 95% confidence interval of 26-41 mEq L−1. These findings suggest a significant glomerular charge selectivity for proteins substantially larger than albumin.
The charge density is, however, far less than estimated from dextran studies.