rs17696736 — NAA25 NAA25 G/A

NAA25 rs17696736 — The 12q24 Autoimmune Locus Tag

The rs17696736 variant sits inside an intron of NAA25 (N-alpha-acetyltransferase 25) on chromosome 12q24, but its disease associations are better understood through the wider 12q24 locus¹¹ 12q24 locus
A cluster of genes on chromosome arm 12q, including SH2B3/LNK and ATXN2, that shows unusually broad association with autoimmune, metabolic, and cardiovascular conditions. rs17696736 acts as a tag SNP for this region, meaning it travels in linkage disequilibrium²² linkage disequilibrium
Physical co-inheritance of nearby variants; when one is measured the other can be inferred without direct genotyping with functional variants in neighboring genes. In genome-wide association studies, carrying the G allele has been consistently associated with elevated risk for type 1 diabetes and juvenile idiopathic arthritis, and minor associations with serum lipids and urate have also been reported.

The Mechanism

The principal causal gene at this locus is thought to be SH2B3 (also called LNK), which encodes an adaptor protein that suppresses cytokine receptor signaling downstream of JAK2³³ suppresses cytokine receptor signaling downstream of JAK2
LNK physically binds JAK2 and thrombopoietin receptor, acting as a brake on white blood cell proliferation and cytokine amplification loops. When SH2B3 function is reduced through regulatory variants tagged by rs17696736-G, the brake on immune cell proliferation weakens, predisposing toward excessive lymphocyte and myeloid cell activation. A second gene, ATXN2 (ataxin-2), lies immediately adjacent to SH2B3 and regulates RNA processing; variants in this gene are hypothesized to contribute to the metabolic arm of the 12q24 phenotype — insulin resistance and lipid dysregulation — while SH2B3 variants primarily explain the autoimmune arm.

NAA25 itself is the auxiliary (scaffolding) subunit of the NatB N-terminal acetyltransferase complex, which co-translationally acetylates the N-terminus of roughly 20% of all human proteins⁴⁴ acetylates the N-terminus of roughly 20% of all human proteins
N-terminal acetylation modulates protein stability, localization, and protein-protein interactions; substrates include Bax, actin, and multiple metabolic enzymes. Whether the intronic variant directly alters NAA25 expression or is purely a surrogate for nearby functional variation remains unresolved. The available evidence points to SH2B3 as the primary driver of autoimmune association at this locus.

The Evidence

The 12q24 locus was first confirmed as a T1D susceptibility region by Todd et al. in a 2007 Nature Genetics GWAS⁵⁵ Todd et al. in a 2007 Nature Genetics GWAS
Four new chromosome regions confirmed: 12q24, 12q13, 16p13, 18p11; P_overall ≤ 1.15×10⁻¹⁴ for 12q24 spanning >4,000 T1D cases. The rs17696736 SNP specifically was tested in an Estonian case-control study by Douroudis et al.⁶⁶ Douroudis et al.
154 T1D patients vs. 230 controls; rs17696736 G allele: OR=1.53, 95% CI 1.14–2.04, p=0.0046, which confirmed that G allele carriers had a 53% higher T1D risk compared to AA homozygotes. A follow-up Estonian study by Kisand and Uibo⁷⁷ Kisand and Uibo
65 LADA, 154 T1D, 260 T2D patients and 229 controls found rs17696736 remained part of an optimized T1D prediction model alongside HLA-DQB1, insulin gene, PTPN22, CTLA4, and CD226 variants, but notably did NOT contribute to LADA prediction — underscoring that this locus tags classic type 1 diabetes autoimmunity rather than the more slowly progressive autoimmune diabetes spectrum.

In juvenile idiopathic arthritis, rs17696736 was independently replicated across two studies: Prahalad et al. 2009⁸⁸ Prahalad et al. 2009
445 JIA cases, 643 controls; OR=1.20, p=0.041 and Thompson et al. 2010⁹⁹ Thompson et al. 2010
>6,000 participants; OR=1.19, p=2.59×10⁻⁵. This supports the concept that the 12q24 locus confers shared autoimmune susceptibility across multiple conditions — a pattern now recognized for several immune regulatory loci. A 2010 childhood stroke study found no association between rs17696736 and ischemic stroke in children, placing this variant squarely in the autoimmune rather than vascular domain.

Secondary GWAS associations include modest effects on HDL, LDL, and total cholesterol (beta ~0.02 standard deviations per allele), and an inverse association with serum urate levels — G carriers have marginally lower urate, potentially reflecting SH2B3/ATXN2 effects on purine metabolism. Effect sizes are small and clinical relevance is modest compared to the autoimmune associations.

Practical Implications

For G allele carriers, the primary actionable implication is awareness of autoimmune disease susceptibility — specifically type 1 diabetes and inflammatory arthritis. The G allele frequency in European populations (~40–43%) means heterozygosity (AG) is actually the most common genotype in Europeans, whereas GG homozygosity (~13–18%) carries the largest additive risk increment. East Asian populations are almost entirely AA at this locus, consistent with lower T1D rates in East Asian populations.

Unlike strongly autoimmune-specific HLA variants, the 12q24 signal has a modest per-allele OR (~1.2–1.5 per study), meaning it contributes as one piece of a polygenic T1D risk picture rather than dominating risk individually. It is most useful in the context of a multi-locus T1D genetic risk score.

Interactions

rs17696736 operates in the same 12q24 locus as rs653178 (ATXN2/SH2B3), which is the lead GWAS SNP for multiple autoimmune conditions including T1D, celiac disease, hypertension, and asthma. These two SNPs are in strong linkage disequilibrium in European populations. The combined effect of SH2B3 dysregulation (autoimmune predisposition) and ATXN2 variation (metabolic predisposition) is hypothesized to jointly explain why T1D patients commonly show features of both autoimmunity and metabolic dysfunction.

Epistatic interactions have been noted with PTPN22 (rs2476601) — the tyrosine phosphatase that acts in the same T-cell signaling pathway. rs2476601 is one of the strongest non-HLA T1D risk variants (OR ~1.7–2.0 for heterozygotes); when combined with 12q24 risk variants, the cumulative probability of T1D rises meaningfully above either variant alone, though the specific compound genotype data for rs17696736 + rs2476601 is not available in the literature.