Reactive oxygen metabolites cause massive, reversible proteinuria and glomerular sieving defect without apparent ultrastructural abnormality.

To identify the specific in vivo renal effect of reactive oxygen species (ROS), hydrogen peroxide (H2O2) was infused directly into the left renal artery in Munich-Wistar rats. H2O2 (5 to 50 mumol over 1 h) induced a dose-dependent increase in urine protein excretion rate in infused kidneys, reaching a maximum at the dose of 35 mumol (on average, a 60-fold increase from baseline). The H2O2 (35 mumol)-induced proteinuria peaked over 1 h and completely normalized by 24 h after the infusion. Electrophoresis revealed that the urine protein is primarily of glomerular origin. Fractional clearances of graded-size neutral dextran of larger molecular radii, an index of glomerular size selectivity, were significantly and substantially elevated immediately but normalized by 24 h after the infusion. GFR and RPF rate remained unchanged throughout the entire time course examined. The H2O2-induced proteinuria was largely prevented by pretreatment with catalase (20 mg, iv) or deferoxamine (30 mg/100 g body wt, iv). Thus, iron-dependent metabolites of hydrogen peroxide appear to be involved in this proteinuria and glomerular size-selective defect. Light and electron microscopy, including determination of anionic site density at lamina rara externa of glomerular capillary wall by polyethyleneimine staining, did not reveal any appreciable abnormality throughout the study period, including at the peak of proteinuria. Thus, ROS can cause massive, reversible proteinuria by inducing a molecular size-selectivity defect of the glomerular capillary wall without apparent ultrastructural abnormalities. The results raise the possibilities: (1) that persistent proteinuria of a variety of renal diseases may reflect persistence of pathogenic ROS acting on glomeruli because the potent proteinuric effect of ROS can be transient (2) that the light and electron microscopy abnormalities in glomeruli of ROS-induced renal injuries reported thus far may have no direct causal linkage to proteinuria; and, finally, (3) ROS-induced reversible proteinuria may relate to the mechanism of clinical functional proteinuria, which involves increased oxygen and ROS metabolism, e.g., exercise-induced proteinuria.