Researchers Find New Risk Loci for Systemic Lupus Erythematosus

Jacquelyn K. Beals, PhD

November 10, 2009

November 10, 2009 (Honolulu, Hawaii) — A large-scale replication study focusing on single-nucleotide polymorphisms (SNPs) that were nominally associated with systemic lupus erythematosus (SLE) in a previous genome-wide association study (GWAS) has identified 5 new SLE susceptibility loci.

The study, presented here at the American Society of Human Genetics 59th Annual Meeting and published online October 18 in Nature Genetics, screened variants already associated with other autoimmune diseases and identified additional loci that might influence SLE susceptibility (P < 1 × 10–3).

SLE is highly heritable, with siblings of a proband having roughly 30 times the risk of the general population.

The replication portion of this study analyzed 3735 SNPs (from 2466 regions) selected for their nominal association (< .05) with SLE, as demonstrated in a previous GWAS. From this "enriched" population, the large-scale replication study identified 5 new susceptibility loci for which the association with SLE reached genome-wide significance (< 5 × 10–8). The genes at or near the 5 loci are TNIP1, PRDM1, JAZF1, UHRF1BP1, and IL10:

  • TNIP1 encodes TNFAIP3-interacting protein 1, and might regulate signals from TLR and TNF.

  • An intergenic locus is near PRDM1 (also called BLIMP1), a gene involved in plasma cell differentiation.

  • The JAZF1 region is associated with type 2 diabetes risk and with height.

  • A locus in UHRF1BP1 encodes a nonconservative amino acid change, and the region includes a gene involved in an RNA processing complex frequently targeted by SLE autoantibodies.

  • IL10 is an important cytokine with a role in downregulating immune responses.

"Genome-wide association scans only give you a region of the genome that's associated with the disease," said presenter Robert R. Graham, PhD, from Immunology Biomarkers Group, Genentech Inc, South San Francisco, California, talking with Medscape Pathology & Lab Medicine. "They very rarely know what the actual causal allele is. If we don't know what the causal allele is, it's hard to test whether it's a gain or a loss of function."

"Sometimes there are surprising results," Dr. Graham added. He described PTPN22, a gene associated with multiple autoimmune diseases. "People probably were predicting beforehand that it was a gain-of-function mutation. When they actually did the work, it turned out it was a loss-of-function mutation," he said. "A lot of these regions actually have multiple genes, and it's not always obvious what the real gene is, so that's a definite complication," he said.

In his presentation, Dr. Graham commented on the overlap between autoimmune disease risk loci: "If it's identified in 1 autoimmune disease, it will probably show up in another as well," he said. For example, in screening the 42 alleles associated with other autoimmune diseases, less than 1 SLE association would have been expected by chance. Instead, 5 alleles demonstrated an association with SLE.

The 5 loci that showed a highly suggestive association with SLE risk (P < 1 × 10–3) had previously established associations with a wide variety of autoimmune diseases: type 1 diabetes, Graves' disease, age-related macular degeneration, Addison's disease, multiple sclerosis, psoriasis, inflammatory bowel disease, celiac disease, and SLE.

Despite the identification of many new SLE-related genes, most SLE susceptibility remains unexplained. Asked what percent of SLE risk could be attributed to specific genes, Dr. Graham told Medscape Pathology & Lab Medicine that it "really depends on how you do your calculations. The lower bound is about 10% and the upper bound is about 30% to 40%. . . . It's unclear whether there's copy-number variation, rare variants, epigenetics."

Comoderator Jonathan L. Haines, PhD, professor of human genetics, chief of the Division of Human Genomics, and director of the Center for Human Genetics Research, Vanderbilt University Medical Center, Nashville, Tennessee, agreed. "The vast majority of [genetic risk factors for SLE] still need to be found," he told Medscape Pathology & Lab Medicine.

"Like virtually every other disease we have, it's not going to be a single [gene]. The genes we have so far are not explaining even 50%. It's less than that," said Dr. Haines. "There's a lot to go."

"We're in the very early days, and so there's going to be a lot of work and a lot of hard biology to do to really understand what these genes are doing," acknowledged Dr. Graham. He expects the work to include more functional studies and the next generation of resequencing. "In a few years, I think we'll be able to make a lot of progress," Dr. Graham predicted.

Dr. Graham was a full-time employee of Genentech at the time of this work, and has applied for a patent based on this work. Dr. Haines reports receiving consulting fees and other remuneration from ArcticDx, and has held nonremunerative positions of influence (officer, board member, trustee, or public spokesperson) with ArcticDx and with Optherion, as well as being an inventor/patent owner.

American Society of Human Genetics (ASHG) 59th Annual Meeting: Abstract 271. Presented October 24, 2009.

Nat Genet. 2009;41:1228-1233. Abstract