Identification of membrane-bound CR1 (CD35) in human urine: evidence for its release by glomerular podocytes.

Complement receptor 1 (CR1) is present on erythrocytes (E-CR1), various leucocytes, and renal glomerular epithelial cells (podocytes). In addition, plasma contains a soluble form of CR1 (sCR1). By using a specific ELISA, CR1 was detected in the urine (uCR1) of normal individuals (excretion rate in 12 subjects, 3.12 +/- 1.15 micrograms/24 h). Contrary to sCR1, uCR1 was pelleted by centrifugation at 200,000 g for 60 min. Analysis by sucrose density gradient ultracentrifugation showed that uCR1 was sedimenting in fractions larger than 19 S, whereas sCR1 was found as expected in fractions smaller than 19 S. The addition of detergents reduced the apparent size of uCR1 to that of sCR1. After gel filtration on Sephacryl-300 of normal urine, the fractions containing uCR1 were found to be enriched in cholesterol and phospholipids. The membrane-association of uCR1 was demonstrated by analyzing immunoaffinity purified uCR1 by electron microscopy which revealed membrane-bound vesicles. The apparent molecular mass of uCR1 was 15 kD larger than E-CR1 and sCR1 when assessed by SDS-PAGE and immunoblotting. This difference in size could not be explained on the basis of glycosylation only, since pretreatment with N-glycosidase F reduced the size of all forms of CR1; however, the difference in regular molecular mass was not abrogated. The structural alleles described for E-CR1 were also found for uCR1. The urine of patients who had undergone renal transplantation contained alleles of uCR1 which were discordant with E-CR1 in 7 of 11 individuals, indicating that uCR1 originated from the kidney. uCR1 was shown to bind C3b-coated immune complexes, suggesting that the function of CR1 was not destroyed in urine. A decrease in uCR1 excretion was observed in 3 of 10 patients with systemic lupus erythematosus, corresponding to the three who had severe proliferative nephritis, and in three of three patients with focal sclerosis, but not in six other patients with proteinuria. Taken together, these data suggest that glomerular podocytes release CR1-coated vesicles into the urine. The function of this release remains to be defined, but it may be used as a marker for podocyte injury.