rs3789604 — RSBN1

RSBN1 rs3789604 — A Psoriasis Haplotype Tag at the PTPN22 Locus

The PTPN22 gene region on chromosome 1p13 is one of the most studied loci in autoimmune genetics. Most attention focuses on the R620W missense variant (rs2476601), which alters T-cell signaling threshold and drives risk for rheumatoid arthritis, type 1 diabetes, and lupus. But haplotype analyses have revealed that the locus harbors multiple independent signals¹¹ multiple independent signals
Deep sequencing and haplotype studies of 37+ PTPN22 variants identified secondary association signals that are statistically independent of R620W and tag different disease patterns. rs3789604 is one such secondary variant — a synonymous change in the neighboring RSBN1 gene that marks a distinct haplotype block associated specifically with early-onset psoriasis and Graves' disease, rather than with the RA/T1D pattern driven by R620W.

RSBN1 (round spermatid basic protein 1) encodes a protein involved in chromatin regulation, but rs3789604's disease associations operate through linkage disequilibrium with nearby PTPN22 regulatory elements²² linkage disequilibrium with nearby PTPN22 regulatory elements
rs3789604 is a synonymous coding change — it doesn't alter the RSBN1 protein. Its clinical significance comes from being a tag SNP that marks a chromosomal haplotype block also covering PTPN22 regulatory regions. Changes in the haplotype's regulatory architecture modulate PTPN22 expression or splice isoform ratios in immune cells. PTPN22 encodes LYP (lymphoid tyrosine phosphatase), a potent suppressor of T-cell receptor signaling that sets the activation threshold for both T and B lymphocytes.

The Mechanism

rs3789604 sits at GRCh38 chr1:113,812,320 within an exon of RSBN1, which is transcribed on the minus strand. The plus-strand alleles are T (reference, ~83% frequency) and G (alternate, ~17%). The nucleotide change is synonymous in RSBN1 — it does not alter the arginine at codon 31 — but rs3789604 resides in the same chromosomal neighborhood as PTPN22 and is in linkage disequilibrium³³ linkage disequilibrium
LD means alleles co-occur more frequently than chance because of their proximity on the chromosome — rs3789604 and nearby variants travel together through generations, so any one serves as a proxy for the others with multiple PTPN22 intronic and upstream variants. The G allele marks a specific haplotype block (documented in scleritis studies as the "TTATACGCG" haplotype) that is enriched in several inflammatory conditions.

The functional effect is inferred to operate through non-coding regulatory changes in the PTPN22 gene: altered promoter activity, enhancer function, or splicing signals that modulate how much LYP protein is produced in immune cells. Lower LYP activity (from reduced expression) lowers the activation threshold for T cells, increasing the likelihood of autoreactive responses. Higher LYP activity (from elevated expression) may conversely impair normal immune responses. The exact direction depends on the specific haplotype and disease context — which explains why the G allele shows risk in psoriasis but some Asian studies report different directional effects for Graves' disease.

The Evidence

The strongest evidence comes from psoriasis. Smith et al. 2008⁴⁴ Smith et al. 2008
Polymorphisms in the PTPN22 region are associated with psoriasis of early onset. Br J Dermatol 2008 found rs3789604 significantly associated with Type I (early-onset) psoriasis (P=0.0002) in a UK cohort of 647 cases and 566 controls, replicated in a combined dataset of 900 cases and 2,590 controls. Crucially, R620W showed no association with psoriasis in the same samples — demonstrating that rs3789604 tags a distinct functional signal, not the classic R620W haplotype.

This finding was independently replicated by Li et al. 2009⁵⁵ Li et al. 2009
Further genetic evidence for three psoriasis-risk genes: ADAM33, CDKAL1, and PTPN22. J Invest Dermatol 2009 in a combined meta-analysis of 2,823 psoriasis cases and 4,066 controls (P=3.45×10⁻⁵), one of the largest psoriasis genetics datasets assembled at that time. The consistent signal across two independent study designs in different populations substantially strengthens the association.

For Graves' disease, data show population heterogeneity. Ichimura et al. 2008⁶⁶ Ichimura et al. 2008
Associations of PTPN22 gene polymorphisms with susceptibility to Graves' disease in a Japanese population. Thyroid 2008 found the A allele (T on plus strand, the reference) significantly elevated in Japanese GD patients (OR=1.45, P=0.0085; 414 patients, 231 controls). In contrast, a Chinese study (Gu et al. 2010, PMID 19438904) found the AA genotype correlated with reduced GD risk. This apparent contradiction likely reflects differences in haplotype background⁷⁷ haplotype background
The same rs3789604 allele may sit on chromosomes with different flanking PTPN22 variants in different populations, meaning the allele tags different regulatory configurations in Japanese vs Chinese vs European genomes.

For rheumatoid arthritis, Wesoly et al. 2007⁸⁸ Wesoly et al. 2007
The 620W allele is the PTPN22 genetic variant conferring susceptibility to RA in a Dutch population. Rheumatology 2007 found no independent RA association for rs3789604 (P=0.134), confirming that at least in Europeans, R620W is the dominant signal and rs3789604 does not independently tag RA risk.

Practical Implications

The clinical takeaway from rs3789604 is primarily about psoriasis and related spondyloarthropathy risk, operating through the PTPN22 locus independently of R620W. Carriers of one or two G alleles face modestly elevated risk for early-onset psoriasis. Psoriasis affects about 2–3% of the global population and can transition to psoriatic arthritis in approximately 30% of cases — a destructive inflammatory arthropathy that may present years after skin disease onset. Catching psoriasis early enables topical and systemic interventions that reduce the progression to joint involvement.

The scleritis and Graves' disease signals are secondary findings with more limited or population-specific evidence. Carriers should be aware that autoimmune conditions at the PTPN22 locus tend to cluster — having one PTPN22-region variant increases the probability of other inflammatory conditions in the same pathway.

Interactions

rs3789604 is in close proximity to rs1217414 (an intronic PTPN22 variant also associated with psoriasis and ankylosing spondylitis) and was found to interact in the psoriasis signal. Smith et al. found that carrying risk alleles at both rs1217414 and rs3789604 showed a stronger combined association (P=0.002) than either alone, consistent with a shared haplotype effect. Both variants are largely independent of R620W (rs2476601), which primarily drives RA and T1D risk. Together, rs3789604 and rs1217414 define a psoriasis/spondyloarthropathy-specific PTPN22 haplotype distinct from the R620W haplotype that governs RA and T1D risk.