Structural features of nephritogenic lupus autoantibodies.

We have identified monoclonal antibodies derived from MRL-lpr/lpr lupus-prone mice that produced nephritis after passive transfer to normal mice. Our present goal was to elucidate the structural and immunochemical features of nephritogenic Ig that facilitate immune deposition. For this purpose the antigen binding properties, capacity to form immune deposits, and nucleotide sequence of a genetically related autoantibody subgroup were compared. The prototype, H147 (an IgG encoded by 7183/81X VH gene), produced glomerular and tubular basement membrane, mesangial immune deposits, and proliferative glomerulonephritis after passive transfer to normal mice. For comparison three other 7183/81X encoded anti-DNA IgG (H257, H171, and H8a) were evaluated (predicted heavy chain aa homology >75%). H257 produced similar types of immune deposits as H147, and this was associated with nephritis; H8a produced predominantly mesangial deposits, whereas H171 did not produce significant deposits. Although their antigen binding profile to a panel of soluble autoantigens was variable, only H147 and H257 bound to both mesangial and aortic endothelial cell surfaces. V gene sequence analysis of the IgG suggests that individual residues, motifs, and conformations influence the autoantigen binding specificities that contributed to the observed differences in immune deposit formation.

Lupus autoantibodies interact directly with distinct glomerular and vascular cell surface antigens.

We have identified monoclonal anti-DNA antibodies derived from lupus prone MRL-lpr/lpr mice that produce glomerular immune deposits and nephritis after passive transfer to normal mice. Particularly noteworthy is that the location of immune deposition varied among nephritogenic Ig, and this was associated with distinctive histologies and clinical disease profiles. Although their autoantigen binding properties differed, they were highly cross-reactive, in a manner similar to Ig deposited in glomeruli of lupus mice. This antigen binding profile was also typical of other previously described nephritogenic autoantibodies that bound directly to glomerular antigens to initiate immune deposit formation. In this study, we questioned whether ligation of different glomerular antigens by individual autoantibodies could contribute to the observed differences in the location of immune deposits. To examine this possibility, monoclonal anti-DNA antibodies (IgG2a) that produced glomerular immune deposits in different locations were evaluated. H221 produced mesangial, intracapillary (that is, intraluminal or within the capillary lumen) and subendothelial deposits associated with heavy proteinuria, whereas H147 produced mesangial, subendothelial and linear basement membrane deposits associated with proliferative glomerulonephritis. Initially, the capacity of H221 and H147 to bind directly to glomerular and vascular cell surfaces was evaluated. As demonstrated by FACS, H221 bound preferentially to mesangial cells whereas H147 bound preferentially to endothelial cells. To identify possible target cell surface antigens, Western blots, immunoprecipitation of surface labeled cells, and 2D gel electrophoresis were employed. H221 reacted with a 108 kDa protein on mesangial cells not identified by H147, whereas H147 reacted with a 45 kDa protein on endothelial cells not identified by H221. These results support the hypothesis that some nephritogenic lupus autoantibodies initiate immune deposit formation through direct interaction with glomerular antigens. Furthermore, they suggest that the site of immune deposition is determined by both antigen binding properties of the relevant antibody and the location of its target ligand within the glomerulus. In a given individual, therefore, the predominant autoantibody-glomerular antigen interaction may influence the morphologic and clinical phenotype expressed. Variation in the predominant interaction may also contribute to variations in disease expression among individuals with lupus nephritis.