The SLE variant Ala71Thr of BLK severely decreases protein abundance and binding to BANK1 through impairment of the SH3 domain function.

The B-lymphocyte kinase (BLK) gene is associated genetically with several human autoimmune diseases including systemic lupus erythematosus. We recently described that the genetic risk is given by two haplotypes: one covering several strongly linked single-nucleotide polymorphisms within the promoter of the gene that correlated with low transcript levels, and a second haplotype that includes a rare nonsynonymous variant (Ala71Thr). Here we show that this variant, located within the BLK SH3 domain, is a major determinant of protein levels. In vitro analyses show that the 71Thr isoform is hyperphosphorylated and promotes kinase activation. As a consequence, BLK is ubiquitinated, its proteasomal degradation enhanced and the average life of the protein is reduced by half. Altogether, these findings suggest that an intrinsic autoregulatory mechanism previously unappreciated in BLK is disrupted by the 71Thr substitution. Because the SH3 domain is also involved in protein interactions, we sought for differences between the two isoforms in trafficking and binding to protein partners. We found that binding of the 71Thr variant to the adaptor protein BANK1 is severely reduced. Our study provides new insights on the intrinsic regulation of BLK activation and highlights the dominant role of its SH3 domain in BANK1 binding.

The dual effect of the lupus-associated polymorphism rs10516487 on BANK1 gene expression and protein localization.

Numerous loci have been found genetically associated with complex diseases, but only in a few cases has the functional variant and the molecular mechanism behind it been identified. Recently, the association of the BANK1 gene with systemic lupus erythematosus (SLE) was described. Here, we investigated the role of the associated polymorphisms on gene function and found that SNP rs17266594 located in the branch point consensus sequence has negligible effect on splicing or gene expression. The non-synonymous SNP rs10516487 located in exon 2 influenced splicing efficiency by creating an exonic splicing enhancer site for the SRp40 factor. Further, this same SNP generates protein isoforms with differential and measurable self-association properties. The full-length protein isoform containing the R61 variant forms larger protein scaffold complexes in the cell cytoplasm compared with the protective BANK1-61H variant. We also observed that, contrary to the full-length isoforms, the short Δ2 isoform of BANK1 displays a homogeneous cytoplasmic distribution, underscoring the potential role of the exon 2-coded protein domain in the scaffolding function of BANK1. We provide evidence that the non-synonymous SNP rs10516487 (G>A; R61H) shows a dual nature by first, influencing mRNA splicing and consequently the quantity of protein, and, second, by producing a risk variant-containing protein isoform with increased potential for multimerization.

Fine mapping of the SLEB2 locus involved in susceptibility to systemic lupus erythematosus.

We have previously reported linkage of systemic lupus erythematosus to chromosome 2q37 in multicase families from Iceland and Sweden. This locus (SLEB2) was identified by linkage to the markers D2S125 and D2S140. In the present study we have analyzed additional microsatellite markers and SNPs covering a region of 30 cM around D2S125 in an extended set of Nordic families (Icelandic, Swedish, and Norwegian). Two-point linkage analysis in these families gave a maximum lod score at the position of markers D2S2585 and D2S2985 (Z = 4.51, PIC = 0.65), by applying a "model-free" pseudo-marker linkage analysis. Based on multipoint linkage analysis in the Nordic families, the most likely location of the SLEB2 locus is estimated to be in the interval between D2S125 and the position of markers D2S2585 and D2S2985, with a peak multipoint lod score of Z = 6.03, assuming a dominant pseudo-marker model. Linkage disequilibrium (LD) analysis was performed using the data from the multicase families and 89 single-case families of Swedish origin, using the same set of markers. The LD analysis showed evidence for association in the single-case and multicase families with locus GAAT3C11 (P < 0.0003), and weak evidence for association was obtained for several markers located telomeric to D2S125 in the multicase families. Thirteen Mexican families were analyzed separately and found not to have linkage to this region. Our results support the presence of the SLEB2 locus at 2q37.

Association analysis with microsatellite and SNP markers does not support the involvement of BCL-2 in systemic lupus erythematosus in Mexican and Swedish patients and their families.

We have described suggestive linkage between microsatellite markers within the cytogenetic region 18q21-23 and SLE, a region where linkage with other autoimmune diseases has also been detected. The Bcl-2 gene located within this region, is a candidate gene because of its role in apoptosis, a physiological mechanism that could be deregulated in autoimmune disease. Furthermore, several studies have found abnormalities of Bcl-2 expression in SLE patients. We therefore sought to determine if the Bcl-2 gene is involved in SLE by studying members of a large cohort of Mexican SLE patients (n = 378) and 112 Swedish simplex families. Using a microsatellite marker and two single nucleotide polymorphisms located within the gene, we were unable to detect association between Bcl-2 and SLE in either population. We also tested whether combinations of alleles of the Bcl-2 and IL-10.G microsatellites would increase the risk for SLE. Our results do not support such hypothesis. Our findings suggest that linkage between SLE and the 18q21-23 region is due to a gene other than Bcl-2.