Breaking down the barrier: evidence against a role for heparan sulfate in glomerular permselectivity.

T he glomerular capillary wall is thought to function as both a sizeand charge-selective barrier. The concept of charge selectivity emerged from a series of now classic studies that used tracers such as dextran, peroxidase, and ferritin to evaluate the influence of molecular charge on glomerular filtration (1–5). The permeability of anionic derivatives of each tracer was lower than their neutral counterparts of comparable size, whereas the permeability of cationic forms was enhanced. This led to the theory that the passage of endogenous circulating polyanions, notably albumin, would likewise be impeded by the “fixed” or intrinsic negative charge of the glomerular capillary wall. These types of studies have since been challenged on many fronts, particularly on the basis of deformation, degradation, or selective uptake of the tracers. Whether charge selectivity actually exists and is important for glomerular function is a subject of intense debate (6). Despite the controversy, the concept remains a cornerstone of renal physiology. Considerable work has gone into localizing and defining the biochemical nature of the charge barrier. Glomerular anionic sites can be detected on the basis of their affinity for cationic probes (e.g., polyethyleneimine, cationized ferritin, lysozyme) and have been found in association with each layer of the capillary wall. The anionic glycocalyx of podocytes and endothelial cells, formed largely by the sialoprotein podocalyxin (7), may contribute to the barrier. Indeed, podocalyxin was shown to play a critical role in dictating podocyte foot process architecture, presumably through charge-related repulsive effects (8). However, the glomerular basement membrane (GBM) is generally considered to be of primary importance for the charge barrier. GBM charge is imparted by the sulfated glycosaminoglycan (GAG) side chains of proteoglycans, and to a lesser extent by carboxyl and sialyl groups of glycoproteins. The former were classified as heparan sulfate (HS) proteoglycans (HSPG) in an influential study in which in vivo perfusion of heparinase, but not other GAG-degrading enzymes, led to a dramatic increase in the permeability of rat GBM to native (anionic) ferritin (9). Other sulfated GAG are not found in the GBM in significant amounts, but hyaluronic acid is present. Charge barrier dysfunction has long been touted as an underlying cause of human glomerular disease (10–12). This may be brought about by decreased expression or undersulfation of GBM-HSPG (13,14). Segmental or global loss of GBM-HS has been reported in human membranous nephritis, lupus nephritis, minimal change disease, and diabetic nephropathy (13,15), as well as in rat models of adriamycin and Heymann nephritis (16,17). The intensity of GBM labeling inversely correlates with severity of disease, which supports the theory that reductions in GBM-HS contribute directly to a loss of barrier function. However, in a recent study, GBM-HS was reported to be normal in diabetic humans and rats with microalbuminuria, and it was concluded that its loss may be a secondary event occurring at advanced stages of this disease (18). In this issue of JASN, Wijnhoven and colleagues provide evidence that challenges the notion that HS is an important constituent of the glomerular filtration barrier (19). Here they report on a short-term study of renal function in rats following intravenous administration of heparinase to degrade glomerular HS. As a control, rats were injected with neuraminidase, which cleaves neuraminic acid from glomerular sialoproteins such as podocalyxin and causes proteinuria. Careful analysis of the specificity and efficiency of deglycosylation was performed using lectin histochemistry and immunostaining with a panel of well-defined HS antibodies. It is this approach, combined with the simple determination of albumin excretion (as opposed to the use of a surrogate tracer), that distinguishes this study. Injection of heparinase was followed quickly by urinary excretion of HS. After 24 h there was complete loss of glomerular HS as assessed by immunostaining, which indicated there was no regeneration of HS over the course of the experiment. Staining for the core protein agrin, the major GBM-HSPG, was not disrupted. Labeling with the cationic probe cuprolinic blue, which detects proteoglycans with their anionic GAG side chains as a filamentous network in the GBM, was virtually abolished after heparinase injection, yet glomerular and tubular architecture was otherwise normal. The key finding of this study is that the glomerular filtration barrier of heparinasetreated rats was functionally intact, despite the loss of HS and a disruption of GBM charge. The authors conclude that the loss of GBM-HS does not lead to increased permeability to protein. Their findings are in accordance with recent work using isolated perfused kidneys that have pointed to other GAG (chondroitin sulfate and hyaluronic acid) present in the endothelial Published online ahead of print. Publication date available at www.jasn.org.