Editorial: The primary glomerular filtration barrier--basement membrane or epithelial slits?

In 1961, based on results obtained with the particulate tracer ferritin, Farquhar, Wissig and Palade [15] proposed a functional model for the glomerulus and defined a role for each of its components in the filtration process: a) the basement membrane as the main filter; b) the endothelium as a valve, which by the number and size of its fenestrae, controls access to the filter; c) the epithelium as a monitor which partially recovers proteins that leak through the filter; and d) the mesangium which serves to recondition and unclog the filter by incorporating and disposing of filtration residues which accumulate against it. In 1966, based on results obtained with the histochemically demonstrable tracers, horseradish peroxidase and myeloperoxidase, Graham and Karnovsky [24] questioned the basement membrane as the site of the main filter and proposed instead that it functioned as a crude prefilter with the epithelial slits representing the final critical barrier. While the concept of the "two-barriers-in-series" has enjoyed wide acceptance, the validity of certain of the experimental data used to support the original hypothesis has been questioned [23, 29, 37]. In the meantime, additional experimental evidence obtained largely with the use of particulate tracers (especially dextrans [23, 43, 44]), has provided strong support for the concept that the basement membrane acts as the main barrier to the passage of molecules in the same size range as plasma proteins (32, 000 to 125,000 mol wt). With respect to the function of the other layers in filtration, additional new information that has come to light has supported the roles proposed above. Work with both particulate [16, 60] and enzymatic [24, 29] tracers, as well as studies by Michael et al [61, 62] with aggregated serum ablumin, supported the phagocytic (unclogging) function of the mesangium. There is evidence from work with particulate tracers (particularly that with dextrans in nephrotic animals [43, 44]) which supports the monitoring function proposed for the epithelium. Recognizing that their work with histochemically demonstrable tracers may have certain technical limitations, Karnovsky, Ainsworth and Schneeberger [29, 37, 38] have recently taken the position that there is no definitive answer to the question of which structure-basement membrane or epithelial slits-represents the principal filter, and have suggested that more information is needed in order to make such a decision. But, in fact, the bulk of the evidence available at present favors the basement membrane as the primary filtration barrier in the glomerulus. Substantial evidence based on work with electronopaque tracers (including recent studies with dextrans) indicates retention of a variety of tracers by the basement membrane. On the other hand, unequivocal demonstration of retention of any tracer by the slits is still lacking.