Autoimmune & ArthritisResearch PaperOpen Access

New Gene-Mapping Tool Reveals How a Single Variant Drives Autoimmune Risk via mTOR

A nanoscale chromatin-mapping platform uncovers how a single noncoding variant creates a rogue DNA motif that silences a key mTOR-regulating gene, raising autoimmunity risk.

Sunday, October 4, 2026 3 views
Published in Nat Genet
A researcher examining a high-resolution 3D chromatin structure visualization on a large monitor in a molecular biology lab, with glassware and a pipette visible on the bench

Summary

Most DNA variants linked to disease sit in noncoding regions, and scientists rarely know which ones matter or which genes they control. This study introduces MCCv, a method that reads the 3D folding of DNA from one copy (allele) of a gene at a time, down to single-basepair resolution. Applying it to 405 regulatory elements tied to inflammatory diseases in human CD4+ T cells, the team nominated 251 likely causal genes and showed that over half of the tested variants altered chromatin contacts. One variant, rs4409785, creates a new CTCF binding site that blocks a super-enhancer from reaching the SESN3 gene. SESN3 senses the amino acid tryptophan to regulate mTOR, and mouse models confirmed its role in autoimmunity. The work gives researchers a way to turn genetic associations into drug-target hypotheses.

Detailed Summary

Most genetic variants linked to common disease lie in the noncoding genome, which makes up over 95% of DNA. Fewer than 5% of disease-associated variants are thought to be functionally active, because each GWAS signal travels with dozens to hundreds of neighboring variants. Immune-mediated inflammatory diseases (IMIDs) affect 5-7% of the population and are chronic and largely incurable. Without knowing which variants are causal and which genes they control, translating genetics into therapies has stalled.

The authors built a Micro Capture-C variant-to-function platform (MCCv). Probes are placed directly over a variant in heterozygous or genome-edited cells, so each allele serves as the internal control for the other. Ultra-short ligation junctions (<800 bp), usually discarded, produce basepair-resolution maps of nucleosome positioning and transcription factor contacts. Long-range junctions (1 kb to 1 Mb) link the element to promoters. About 86.7% of reads were usable for allelic analysis. In Jurkat cells, rs2793109, a rheumatoid arthritis variant, created a RUNX1 motif and raised ZEB1 expression (q = 0.0002) but not ZNF438.

The team then screened IMID loci in primary CD4+ T cells. They imputed 3,104 haplotypes from 400 studies, assigned a candidate mechanism to 47%, and profiled 405 enhancer variants. MCCv captured a mean of 839 ligation junctions per strongest enhancer-promoter contact, versus 2.5 for genome-wide Micro-C, with fivefold better signal-to-noise. It nominated 251 genes. For 60.5% of variants the strongest contact was the nearest gene, and 56% agreed with OpenTargets. The nominated genes were enriched for primary immunodeficiency genes and drug targets.

Single-allele analysis covered 241 heterozygous sites across 3 donors, only about 30% of which conventional 3C methods could analyze. Of these, 54.6% showed at least one significantly altered contact, and disease variants were enriched over controls (up to 7.5-fold; P = 2.73 × 10^-13 for haplotype controls). At rs10950648, loss of PRDM1 binding weakened contact with the AHR promoter. Allelic RNA skew appeared at 35.5% (16 of 45) of variants with changed contacts, a threefold enrichment over background. Direction matched at 64.7% of variant-gene pairs, rising to 88.2% when long-range changes were considered.

The headline example is rs4409785 at 11q21, associated with multiple sclerosis and rheumatoid arthritis. Phenome-wide data also tie it to sepsis risk and protection against Streptococcus and HPV infection. The risk allele creates a neo-CTCF motif, evolutionarily absent in primates, producing a 20-bp footprint. CTCF ChIP-seq and ATAC-seq showed increased binding and accessibility on the alternate allele in 15 heterozygous donors (q = 0.000123). It contacts two convergent CTCF sites. Per the abstract, this blocks super-enhancer contacts with the SESN3 promoter.

SESN3 was shown to regulate mTOR by sensing tryptophan, and mouse models supported a role in autoimmunity. This links immune genetics to the nutrient-sensing pathway central to aging biology. Caveats: allelic analyses used only 3 donors, one cell type (activated CD4+ T cells), and ex vivo conditions. The genome-editing and mouse experiments are not detailed in the text available here.

Key Findings

  • MCCv screened 405 IMID-linked enhancer variants in primary CD4+ T cells, assigning a candidate mechanism to 47% of 3,104 haplotypes from 400 studies
  • Mean of 839 ligation junctions per top enhancer-promoter contact for MCC versus 2.5 for Micro-C, with fivefold greater signal-to-noise
  • 251 causal genes nominated; 60.5% were the nearest gene and 56% matched OpenTargets L2G, with enrichment for primary immunodeficiency genes and drug targets
  • 54.6% of 241 heterozygous disease variants showed at least one altered chromatin contact; enrichment over haplotype controls was 7.5-fold (P = 2.73 × 10^-13)
  • 16 of 45 variants (35.5%) with changed contacts had allelic RNA skew, a threefold enrichment over background; directional concordance 64.7%, rising to 88.2% with long-range direction
  • Only about 30% of heterozygous SNPs could be analyzed with conventional 3C approaches; 19.4% of enhancers had their nearest promoter within 10 kb
  • rs4409785 creates a neo-CTCF site with a 20-bp footprint and increased CTCF binding and accessibility on the risk allele (q = 0.000123, n = 15 donors), blocking super-enhancer contact with SESN3

Methodology

MCCv uses Micro Capture-C with oligonucleotide probes placed directly over variants in heterozygous or genome-edited cells, with six technical replicates per library and an allele-specific pipeline that mitigates mapping bias. It was benchmarked in Jurkat cells (42 gain-of-function variants), then applied to 405 enhancer and CTCF variants in activated primary CD4+ T cells from 3 donors, compared with genome-wide Micro-C. Statistics included paired t-tests with local FDR, Wilcoxon signed-rank tests, chi-squared enrichment tests, and a generalized linear model (basepairC; effect size ≥0.2, Padj ≤0.01) for nanoscale matrices. Phasing linked variants to allele-specific TT-seq and RNA-seq.

Study Limitations

Allelic analyses relied on only 3 donors, a single cell type (activated CD4+ T cells), and cultured primary cells, which may miss tissue- and context-specific effects. Many variants could not be phased reliably, and RNA-skew validation covered a limited subset of variant-gene pairs. Conflict-of-interest statements were not included in the supplied text, and the editing and mouse experiments were truncated from the provided text.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: