Degradation of renal basement membrane by neutral proteinases from rabbit polymorphonuclear leucocytes [proceedings

Abstract

Studies of the pathogenesis of glomerular disease have suggested that polymorphonuclear leucocytes promote glomerular damage. The pattern of events observed in the autologous phase of nephrotoxic serum nephritis in rabbits indicates that the fixation of antibodies to glomerular basement membrane initiates the activation and binding of complement. This results in the chemotaxis of polymorphonuclear leucocytes t o the glomerulus where they become closely associated with the basement membrane and basement-membrane antigens are liberated into the urine; there is a concomitant alteration in the permeability characteristics of the basement membrane (Hawkins & Cochrane, 1968). These changes are prevented if the leucocytes are depleted prior to antibody administration. The cytoplasmic granules of human polymorphonuclear leucocytes are known to contain a number of neutral proteinases (Barrett, 1977), two of which (elastase and cathepsin G) can degrade renal basement membrane (Davies er al., 1978), and it has been proposed that release of these enzymes results in damage to the basement membrane in vivo. In the case of rabbits however, the granules contain mainly acidic proteinases and, although neutral proteinases with low activity against histones have been described (Davies et al., 1971), there have been no direct studies of enzymes that might hydrolyse basement membrane. Here we report the presence of neutral proteinase activity in the granules of rabbit polymorphonuclear leucocyges that act on renal basement membrane. Rabbit polymorphonuclear leucocytes were prepared from peritoneal exudates (Cohn & Hirsch, 1960) and the leucocytes were washed in 0 .34~-sucrose and disrupted by homogenization. Intact cells and nuclei were removed by centrifugation at 600g for lOrnin and the granules were harvested by centrifugation at 12000g for 20min. These were suspended in 0.01 M-sodium phosphate buffer, pH7.4, and the suspension was subjected to freezing in acetone/solid CO, and thawing at room temperature three times before centrifuging at 30000g for 30min to pellet the granular membranes. This pellet was then extracted at 0°C with 0.01 M-sodium phosphate buffer, pH7.4, containing 1 M-NaCI and 0.1 % (v/v) Triton X-100, for 20min with shaking. The suspension was centrifuged at 30000g for 30min and the pellet was extracted twice more before combining the supernatants as the membrane extract. Renal basement membrane was prepared by the method of Ligler & Robinson (1977). Hydroxyproline was measured by the method described by Blumenkrantz & Asboe-Hansen (1975), protein was determined by using the Fohn procedure described by Lowry et al. (1951). Each fraction was examined for basement-membrane-degrading activity by using a protein ratio (fraction/basement membrane) of 1 :20 incubated in 0.01 M-Tris/HCI buffer, pH7.4, containing 0.15M-NaC1, for 18 h a t 37°C with constant shaking. After incubation the undigested basement membrane was removed by centrifugation a t 2000g for lOmin and the supernatant was examined for basement-membrane breakdown products. Since collagen-like proteins are major structural components of basement membrane the release of hydrocyproline was taken as a measure of basement-membrane breakdown. Only the granular-membrane extract showed any basement-membrane-degrading activity and this could be stimulated 3-fold by increasing the NaCl concentration to 0 . 5 ~ . Under these conditions up to 15% of the basementmembrane hydroxyproline was released. The neutral proteinase activity against basement membrane was optimally active slightly above neutral pH (Fig. 1). Polyacrylamide-gel electrophoresis with 7.5 % acrylamide gels containing 1 % (w/v) sodium dodecyl sulphate revealed that high-molecular-weight (>90000) proteins had been degraded. The effects of a number of inhibitors on the enzymes were examined