Selective glomerular thrombosis in rats induced by combined injections of nephrotoxic antiserum and lipopolysaccharide.

Experimental glomerular thrombosis was induced in rats by combined injections of nephrotoxic antiserum and lipopolysaccharide. For the development of glomerular thrombosis, administration of nephrotoxic antiserum (greater than or equal to 0.1 ml pooled material) was required as a preparatory agent and greater than or equal to 100 ng lipopolysaccharide as a provoking agent. The severity of renal lesions was not parallel with the amounts of nephrotoxic antiserum and lipopolysaccharide injected. Transient clamping of a unilateral renal artery for 10 to 20 minutes at the time of the nephrotoxic antiserum injection partially prevented the development of glomerular thrombosis in the clamped side. Intervals between the preparatory and provoking injections were found to be -4 to 72 hours for the development of renal lesions. With the preparatory injection of 0.1 to 0.3 ml nephrotoxic antiserum a thrombotic lesion developed exclusively in glomerular capillary walls greater than or equal to 2 hours after the lipopolysaccharide injection. No thrombotic lesion was observed in other tissues such as lung, liver, or intestine, but a generalized Shwartz-manlike phenomenon was observed with the preparatory injection of 0.5 ml nephrotoxic antiserum. When rats were pretreated with nephrotoxic antiserum and 3 hours thereafter transfused with 1 to 3 X 10(8) polymorphonuclear leukocytes, which had been incubated with lipopolysaccharide for 30 minutes in vitro and washed three times with buffered physiologic saline solution, a marked glomerular thrombosis was also induced. The result indicates that lipopolysaccharide plays a role in the development of thrombosis by a direct effect on leukocytes. The development of glomerular thrombosis was prevented in a leukocytopenic state when leukocyte count was less than 600/microliter, but not in thrombocytopenic rats with a platelet count 8.7 to 30 X 10(3)/microliter. Leukocyte count and plasma fibrinogen level decreased, and prothrombin time and activated partial thromboplastin time were prolonged significantly during the pathologic course. Platelet count and FDP did not change significantly. This experimental model has a basic similarity to the generalized Shwartzman reaction, but the lesions develop exclusively in glomeruli.