DIFFERENTIAL CONTRIBUTION OF C4 AND HLA-DQ GENES TO SYSTEMIC LUPUS-ERYTHEMATOSUS SUSCEPTIBILITY

The particular histocompatability antigen (HLA) gene(s) that may confe r systemic lupus erythematosus (SLE) susceptibility remains unknown. I n the present Study, 58 unrelated patients and 69 controls have been a nalyzed for their class I and class II serologic antigens, class II (D R and DQ) DNA restriction fragment length polymorphism, their deduced DQA1 and B1 exon 2 nucleotide sequences and their cor-responding amino acid residues. By using the etiologic fraction (delta) as an almost a bsolute measure of the strongest linkage disequilibrium of an HLA mark er to the putative SLE susceptibility locus, it has been found that th e strength of association of the HLA marker may be quantified as follo ws: DQA10501 (associated to DR3) or DQB1*0201 (associated to DR3) > n on Asp 57 betaDQ/Arg 52 alphaDQ > DR3 > non Asp 57 betaDQ. Thus, molec ular HLA DQ markers tend to be more accurate as susceptibility markers than the classical serologic markers (DR3). However, dominant or rece ssive non Asp 57 betaDQ susceptibility theories, as previously postula ted for insulin-dependent diabetes mellitus, do not hold in our SLE ne phritic population; indeed, three patients bear neither Arg 52 alphaDQ nor Asp 57 betaDQ susceptibility factors. On the other hand, nonsusce ptibility factors are included in our population in the A30B18CF130DR3 DQ2(Dw25) haplotype and not in A1B8CS01DR3DQ2(Dw24); this distinctive association has also been recorded in type I diabetes mellitus and may reflect the existence of common pathogenic HLA-linked factors for bot h diseases only in the A30B18CF10DR3DQ2(Dw25) haplotype. Finally, the observed increase of deleted C4 genes (and not 'null' C4 proteins) in nephritic patients shows that C4 genes are disease markers, but probab ly without a pathogenic role.