The composition of glomerular crescents was examined on the frozen kidney sections obtained from 10 patients (5 patients with IgA nephropathy, two with Henoch-Schönlein purpura nephritis and three with glomerulonephritis due to undetermined etiology) using well-defined monoclonal and polyclonal antibodies to coagulation proteins, extracellular matrices, intermediate filament proteins and immune cells. Fibrinogen/fibrin related antigens (FRA), which were stained with anti-fibrinogen serum, were positive in the crescents of all the patients, but monoclonal antibody to crosslinked fibrin or von Willebrand factor (factor VIII related) antigen did not bind to the crescents. This suggests that the FRA deposited in the crescents is fibrinogen or its degradation products rather than fibrin. Staining for intrinsic components of renal basement membrane, including type IV and V collagens, laminin and fibronectin, were consistently positive in all stages of the crescents. Cytokeratin, showing cytoplasmic staining of the glomerular parietal epithelium and tubular epithelium in the normal kidney, was demonstrated in three patients with cellular crescents. Vimentin, which is normally distributed in parietal and visceral epithelial cells in the glomeruli and interstitial cells, was found at all stages of the crescents. These findings suggest that in the early stage of crescent formation, glomerular epithelial cells play an important role, and that the accumulation of intrinsic basement membrane constituents is associated with the formation and progression of the crescents. None of the crescent cells reacted with either of two monoclonal antibodies (Mo2 and FMC 32) to monocytes/macrophages or with nonspecific esterase staining. It seems that, at least in our patients, monocytes are a minor factor contributing to the formation of glomerular crescents.

Tubular basement membrane (TBM) was prepared from normal human kidneys and solubilized with various enzymes. Collagenase digestion released antigenic moieties from the TBM. All four anti-TBM antibodies we studied, three from patients with idiopathic tubulo-interstitial nephritis (TIN) and one from a renal allograft recipient, distinctively reacted with collagenase-digested (CD) TBM during enzyme-linked immunoassay and could discriminate among sera of normal controls or of other nephritis patients, including anti-glomerular basement membrane (anti-GBM) nephritis. When digested with pronase, trypsin, or pepsin, antigenicity of the TBM decreased. We studied the TBM antigens with immunoprecipitation and immunoblotting. After incubation of radio-iodinated CDTBM with anti-TBM sera, immunoprecipitates were identified by single-dimension SDS polyacrylamide gel electrophoresis or two-dimension gel electrophoresis, followed by autoradiography. All four antibodies had identical results on immunoprecipitation; under nonreducing conditions, they gave two protein bands with m.w. of 54,000 and 48,000 and with pI 7.0 to 8.0 and 6.5 to 7.0. Electrophoresis performed under reducing conditions disclosed only one band at the m.w. of 48,000 and pI of 6.5, suggesting that the 54-kDa component is composed of peptides linked by interchain disulfide bonds. Immunoblot analysis showed that the anti-TBM antibodies were heterogeneous; three antibodies from the idiopathic TIN patients reacted with the 54-kDa band, but the one from the renal allograft recipient reacted with neither band. This finding suggests that there are two antigenic determinants on the 54-kDa component. One such determinant that was resistant to denaturation with SDS was detected by the first three antibodies, and the other that was sensitive to such denaturation bound to the last antibody. The 48-kDa component seemed not to be immunoreactive after incubation with SDS. We studied TBM antigens reactive with anti-GBM antibodies. By immunoblotting, all four sera from patients with anti-GBM nephritis stained TBM proteins of 45 to 50 kDa and 25 to 27 kDa at pH 8.0 to 9.0; this was similar to the staining pattern of CDGBM with the same sera, but the highly cationic (pH greater than 9.0) components were specifically detected in the CDGBM. By inhibition ELISA, the binding of the anti-GBM sera to denatured CDTBM decreased with preincubation of the sera with CDGBM, suggesting that the anti-GBM antibodies recognize the same epitope(s) on the GBM and the TBM.

Fibrinogen fragments (X, Y, D and E) and fibrin fragments (D-dimer and E) were examined in the urine of 52 patients with various types of renal diseases. One or more of the fibrinogen fragments were detectable in all the urine specimens. D-Dimer together with fibrinogen fragments was found in 38 of the 52 patients. The clearance ratio of D-dimer to IgG, which indicates D-dimer generated in the kidney, was lower than 1 in all the patients with the minimal changes nephrotic syndrome, and was greater than 1 in the majority of patients with acute glomerulonephritis, rapidly progressive glomerulonephritis, mesangial proliferative glomerulonephritis and membranoproliferative glomerulonephritis. Our results suggest that urinary fibrin/fibrinogen degradation products (FDP) in renal diseases are derived primarily from increased filtration of FDP from the plasma through a damaged glomerular basement membrane, and that the mechanism of lysis of cross-linked fibrin deposited in the glomeruli occurs simultaneously in some types of glomerulonephritis. It seems that the determination of the clearance ratio of D-dimer to IgG may be useful in assessing the activation of the coagulation and fibrinolysis systems in the kidney in patients with renal diseases.