rs689 — INS INS VNTR (Type 1 Diabetes Susceptibility)

Tag SNP for the insulin gene VNTR promoter — shorter Class I repeats raise T1D risk by reducing thymic insulin expression

See your personal result for INS

Upload your DNA data to find out which genotype you carry and what it means for you.

Upload your DNA data

Works with 23andMe, AncestryDNA, and other DNA test exports. Results in under 60 seconds.

INS VNTR — The Thymic Tolerance Switch

The insulin gene (INS) on chromosome 11p15.5 controls more than blood sugar regulation. A polymorphic repeat region in its promoter — the Variable Number Tandem Repeat11 Variable Number Tandem Repeat
A stretch of short DNA sequences repeated between 26 and ~200 times just upstream of the insulin coding region (VNTR) — determines how much insulin is expressed in the thymus, the organ that trains the immune system to ignore the body's own tissues. The rs689 A/T SNP is a tag variant in near-complete linkage disequilibrium22 linkage disequilibrium
LD means the A/T allele at rs689 reliably predicts which VNTR class a person carries, without sequencing the repeat itself with the VNTR class: the A allele marks the shorter Class I repeats (26–63 copies), and the T allele marks the longer Class III repeats (~140–210 copies). This single locus — IDDM2 — accounts for roughly 10% of the familial clustering of type 1 diabetes (T1D) and is the strongest non-HLA genetic risk factor for the disease.

The Mechanism

The rs689 A allele weakens the polypyrimidine tract33 polypyrimidine tract
A pyrimidine-rich sequence just upstream of the intron 1 splice acceptor site that U2AF35 protein must bind to initiate splicing at the 3′ splice site of INS intron 1. In carriers of the A allele, intron 1 retention increases and proinsulin output drops — the INS transcript is less efficiently processed. In medullary thymic epithelial cells44 medullary thymic epithelial cells
mTECs are the cells that present self-antigens including insulin to developing T cells; T cells that react too strongly are deleted — a process called central tolerance (mTECs), this reduced insulin expression means fewer insulin-reactive T cells are deleted during immune education. Cai et al.55 Cai et al.
Cai CQ et al. Both polymorphic VNTR and AIRE modulate differential expression of insulin in human thymic epithelial cells. Diabetes, 2010 measured a threefold difference in thymic insulin mRNA between Class III and Class I carriers in an AIRE-dependent manner. The surviving auto-reactive T cells can later reach the pancreas and destroy insulin-producing beta cells.

The Evidence

The most direct evidence comes from Kralovicova & Vorechovsky66 Kralovicova & Vorechovsky
Kralovicova J, Vorechovsky I. Allele-specific recognition of the 3′ splice site of INS intron 1. Hum Genet, 2010, who showed that introducing the A allele into a protective Class III haplotype was sufficient to reduce proinsulin levels to those of a risk Class I construct. This positions rs689 — not the VNTR repeat count itself — as the functional effector77 functional effector
Earlier models assumed the VNTR repeat length was the causal element; the 2010 work demonstrated the flanking SNPs controlling splicing are the true functional variants.

Fan et al.88 Fan et al.
Fan Y et al. Thymus-specific deletion of insulin induces autoimmune diabetes. EMBO J, 2009 deleted insulin expression specifically in mTECs in mice and found 100% diabetes penetrance within 3 weeks postnatal — the most direct demonstration that thymic insulin is required for central tolerance to insulin-producing beta cells.

GWAS evidence is exceptionally strong: the GWAS Catalog records associations with T1D at p = 5 × 10⁻¹⁹⁶ for the A allele. The protective T allele (Class III) carries an OR of approximately 0.45 for T1D (OR ≈ 2.21 per T allele for protection in one large study), with risk frequency of the A allele at ~28% in Europeans. In African populations the A allele is markedly more common (~76%), but HLA alleles dominant in T1D susceptibility are rarer in African-ancestry populations, so the overall T1D rate does not simply track the A allele frequency.

Practical Actions

The INS VNTR locus primarily informs T1D risk stratification, not treatment. There are no approved interventions that modify VNTR function. For AA homozygotes, the main value is awareness: earlier monitoring for T1D-associated autoantibodies (IAA, GADA, IA-2A, ZnT8A) substantially improves the chance of catching pre-symptomatic T1D before diabetic ketoacidosis occurs. At-risk individuals can also evaluate eligibility for T1D prevention trials such as teplizumab (anti-CD3 antibody), which has FDA approval to delay T1D onset in at-risk individuals with 2+ autoantibodies.

Interactions

The IDDM2 effect is additive with HLA class II risk alleles. Individuals carrying both the AA genotype here and high-risk HLA-DR3/DR4-DQ8 haplotypes have substantially higher lifetime T1D risk than either locus predicts alone. rs3842753 (the −23HphI variant, also in the INS 3′ UTR) is a second INS tag SNP in high LD with rs689 and is often co-analyzed. rs7903146 (TCF7L2) primarily affects T2D risk through beta-cell mass, a distinct mechanism from the central tolerance pathway here.

Genotype Interpretations

What each possible genotype means for this variant.

TT Class III Homozygous (Protective) Beneficial

Both alleles tag the longer Class III VNTR — lowest T1D risk at this locus

You carry two copies of the T allele, tagging the longer Class III VNTR in the INS promoter. Class III homozygotes have the highest thymic insulin expression at this locus, leading to more complete deletion of insulin-reactive T cells. About 52% of people of European descent share this genotype. This is the most protective genotype for T1D risk at the IDDM2 locus, with approximately 2-fold lower T1D risk compared to AA carriers.

AT Class I/III Heterozygous Intermediate

One risk and one protective allele — moderately elevated T1D susceptibility

Class III VNTR alleles show partial dominance: even one copy increases thymic insulin mRNA above the Class I baseline. Heterozygotes retain more robust central tolerance induction compared to Class I homozygotes, but the A allele still reduces overall INS splicing efficiency relative to TT carriers.

AA Class I Homozygous (High Risk) High Risk

Both alleles tag the shorter Class I VNTR — highest T1D risk

The A allele at rs689 weakens the 3′ splice site of INS intron 1, reducing proinsulin output in medullary thymic epithelial cells approximately threefold compared to Class III carriers. This reduced presentation of insulin peptides during T cell education allows more insulin-reactive T cells to escape deletion, creating a precondition for beta-cell autoimmunity.

In African-ancestry individuals this genotype is considerably more common (~58% are AA based on A allele frequency ~0.76), yet T1D rates are lower in this population — illustrating that HLA risk alleles and environmental factors modulate the absolute risk this locus contributes.