blood-and-gut-cell-analyses-match-140-inflammatory-bowel-disease-risk-loci
Blood and Gut Cell Analyses Match 140 Inflammatory Bowel Disease Risk Loci

Blood and Gut Cell Analyses Match 140 Inflammatory Bowel Disease Risk Loci

A comprehensive multi-tissue genomic investigation combining sorted circulating immune cells and single-cell intestinal transcriptomics has uncovered matching cis-acting expression quantitative trait loci for 140 inflammatory bowel disease risk loci, implicating hundreds of candidate regulatory genes and highlighting the oncology drug entrectinib as a potential repurposing candidate. Genome-wide association studies have identified over two hundred distinct genomic regions associated with susceptibility to Crohn’s disease and ulcerative colitis, yet the vast majority of these loci reside within non-coding regulatory sequences rather than protein-altering open reading frames. Pinpointing the precise causative genes perturbed in cis has historically been impeded by standard tissue-level bulk sequencing, which obscures the subtle transcriptional shifts occurring within rare or specialized cell populations. By resolving expression phenotypes across twenty-seven purified circulating immune cell subsets and forty-three distinct mucosal cell populations isolated from intestinal biopsies, the investigative team established the CEDAR-2 reference atlas to systematically bridge statistical disease associations with primary cell-type-specific transcriptional regulation across human health and chronic inflammatory pathology.

To capture circulating immune variation with rigorous cellular resolution, the researchers collected peripheral blood from 251 healthy European donors and isolated twenty-six distinct immune fractions alongside total peripheral blood mononuclear cells through fluorescence-activated and magnetic cell sorting. The sorted lineages encompassed conventional, intermediate, and non-classical monocytes, natural killer subsets, dendritic cells, innate lymphoid cells, multiple stages of B cells, and finely divided T-lymphocyte subsets including mucosal-associated invariant T cells, regulatory T cells, and helper lineages. Genome-wide genotyping on Illumina OmniExpress BeadChips yielded approximately 6.3 million quality-controlled variants following imputation against the TOPMed reference panel. Transcriptomic sequencing generated thousands of high-depth mRNA libraries, revealing an average of 16,615 detected genes per cell population. Across these sorted blood populations, cis-expression quantitative trait locus mapping with QTLtools identified 60,113 significant associations affecting 11,874 unique genes at a within-cell-type false discovery rate of 0.05 or lower, with each detected signal explaining an average of 19.7 percent of local expression variance.

Because inflammatory bowel diseases directly manifest within the gastrointestinal tract, the team simultaneously investigated mucosal tissue architecture by collecting mucosal biopsies from the terminal ileum, transverse colon, and rectum of sixty healthy adult individuals undergoing screening colonoscopies. Tissue specimens were enzymatically and mechanically fractionated into epithelial and lamina propria cell suspensions, barcoded with hashtag antibodies, and profiled using high-throughput droplet-based single-cell RNA sequencing. Quality filtering recovered 293,801 high-quality cells across fifty-seven donors, averaging 48,628 reads and 2,035 detected genes per single cell. Unsupervised graph-based clustering and batch-balanced integration across anatomical locations resolved 276 discrete clusters structured within a unified 551-node hierarchical cluster tree. These branches mapped comprehensively to thirteen epithelial lineages, fourteen lymphoid populations, six myeloid subsets, and ten stromal cell identities, providing an expansive cellular framework that captured both shared mucosal lineages and location-restricted identities such as ileal Paneth cells across healthy gastrointestinal anatomy.

Applying pseudo-bulk quantitative trait locus mapping across every leaf and node of the intestinal lineage tree revealed 35,010 significant cis-associations influencing 3,007 genes at a false discovery rate of 0.05 or lower. The yield of discovered regulatory signals scaled directly with cell abundance within individual nodes, plateauing near 1,100 detections per node due to the cohort sample size of fifty-seven donors. By clustering shared expression association patterns using the directional theta metric, the researchers grouped significant intestinal signals into 3,345 gene-specific regulatory modules. These modules exhibited marked overdispersion across the hierarchical tree, demonstrating that mucosal regulatory activity is rarely distributed uniformly. Instead, 524 modules operated strictly within a single mucosal cell type, with 429 demonstrating complete anatomical specificity, predominantly distinguishing small intestinal enterocyte lineages from large intestinal colonocytes. In rare instances, regulatory variants produced opposing transcriptional effects across differentiation states, reversing directional slope as progenitor cells transitioned into mature absorptive enterocytes.

When regulatory modules from circulating immune cells and mucosal biopsies were merged into a joint catalog, the resulting dataset encompassed 24,745 cross-gene regulatory modules, of which 2,172 were active across both blood and mucosal compartments. Strikingly, the cross-tissue comparison illuminated 132 regulatory modules that were entirely undetectable in circulating blood populations despite higher statistical power, yet manifested robustly within mucosal-resident lymphoid or myeloid cells. Further scrutiny showed that seventy-five percent of these compartment-specific intestinal signals occurred within lymphoid cells, while twenty-five percent characterized myeloid lineages. For tissue-resident myeloid cells, the appearance of unique regulatory effects was largely driven by the selective expression of previously dormant genes, such as CCL24 in intestinal macrophages, or by specialized resident mast cells absent from circulation. Conversely, gut-resident memory T and B lymphocytes demonstrated module switching, wherein cells with shared lineage identities utilized fundamentally distinct regulatory variants under the localized biochemical influence of the gut microenvironment.

The researchers subsequently cross-referenced the unified expression quantitative trait locus atlas against 206 established genetic risk loci for inflammatory bowel disease, which comprised 173 general loci, 157 Crohn’s disease signals, and 125 ulcerative colitis signals derived from European genome-wide association studies. Evaluating colocalization between disease association patterns and cellular expression profiles using the directional theta statistic and empirical permutation testing identified significant concordances matching 119 risk loci at a false discovery rate of 0.05 or lower, expanding to 140 loci and 556 candidate genes under a ten percent false discovery rate threshold. This represents a 2.3-fold increase in resolved loci compared to earlier bulk-tissue reference panels. Matching associations were detected in both blood and gut for fifty-five percent of these loci, exclusively in circulating cells for twenty-four percent, and solely within intestinal biopsies for twenty-one percent, underscoring the critical necessity of profiling primary mucosal tissue alongside systemic immune populations.

Although mucosal-specific regulatory modules accounted for only ten percent of the broader expression catalog, they contributed nearly half of all identified disease-matching candidate genes, emphasizing the localized origin of pathological predisposition. Individual risk loci frequently matched multiple distinct candidate genes, averaging four putative targets per locus and reaching up to thirty-four in complex genomic regions. Interestingly, adjacent disease association peaks within single genomic regions frequently resolved into cell-type-divergent mechanisms for the exact same gene. For example, within chromosome 5, risk associations for ulcerative colitis colocalized with IRGM expression in circulating CD8 positive memory T cells and CD74 expression in colonocyte precursors, whereas an adjacent peak for Crohn’s disease colocalized with IRGM in circulating naive CD4 positive cells and CD74 in mature small intestinal enterocytes, illustrating how distinct regional signals can govern shared loci through divergent cell types.

To evaluate whether the candidate genes identified by expression colocalization genuinely contribute to disease causality, the authors examined multiple independent functional and genetic benchmarks. A blinded systematic literature review conducted by independent scientists revealed that 88 of the 481 protein-coding disease-matching genes altered susceptibility to experimentally induced colitis when genetically modified in rodent models, representing a 2.3-fold enrichment over expression-matched control genes with non-colocalizing regulatory variants. Furthermore, the magnitude of the colocalization theta metric correlated significantly with reported rodent phenotype alterations. In human genomics, evaluating exome sequencing from approximately 30,000 inflammatory bowel disease patients showed that rare coding variant burden was generally absent across most loci, indicating that common complex disease heritability is predominantly mediated through cell-type-restricted expression changes rather than systemic protein-disrupting mutations, with specific exceptions such as protective rare variants in TAGAP.

To bridge genetic predisposition with clinical applicability, the investigators compared the 556 candidate genes against transcriptomic profiles from fifty-five active Crohn’s disease patients, published mucosal single-cell datasets, and large-scale plasma proteomic surveys. A total of 152 candidate genes exhibited altered expression during active disease that aligned directionally with their genetically inferred disease risk, fulfilling key criteria for actionable drug targets. Eleven candidate genes coincided with therapeutics previously tested or approved for inflammatory bowel disease, including integrin subunit alpha 4 targeted by vedolizumab, interleukin-6 receptor, and Janus kinase 2. Furthermore, the catalog nominated forty-five candidate genes targeted by existing pharmacological agents that had never been evaluated in inflammatory bowel disease, including twenty-two where the mechanism of drug action directly matched the protective regulatory direction indicated by the genetic theta values.

Among these novel opportunities, the small-molecule tyrosine kinase inhibitor entrectinib emerged as a particularly compelling repurposing candidate. Entrectinib is an approved anti-neoplastic therapy targeting NTRK, ROS1, and ALK oncogenes with an established clinical safety profile. Beyond its annotated kinase targets, biochemical studies demonstrate that entrectinib binds directly to arginine 121 of NEK7, an essential regulator that binds and activates the NLRP3 inflammasome complex. Genetic variants increasing NEK7 expression in circulating naive B cells were found to elevate inflammatory bowel disease risk with high statistical confidence, while elevated NEK7 expression was confirmed across circulating immune populations and uninflamed colonic epithelium from Crohn’s disease patients. Previous experimental studies established that shRNA-mediated knockdown of NEK7 markedly ameliorates dextran sulfate sodium-induced colitis in rodents, providing strong pre-clinical corroboration for evaluating entrectinib’s anti-inflammatory mechanism in human disease.

The study authors nonetheless emphasized important biological caveats that must govern translational interpretation and downstream drug development. Colocalization analyses revealed that in at least ten instances, regulatory variants exerted opposite directional effects on the same target gene depending on the specific cell lineage examined, or displayed contrasting associations between Crohn’s disease and ulcerative colitis. In the case of NEK7, non-significant downward transcriptional trends in mucosal lymphocytes highlight the risk that systemic therapeutic inhibition might exert conflicting effects across different anatomical compartments. Similarly, genetic downregulation of known pro-inflammatory mediators such as interleukin-18 receptor components paradoxically correlated with increased disease risk, suggesting that hypomorphic signaling might predispose to disease onset while active inflammation demands alternate therapeutic strategies. By providing a cell-resolved genomic framework across human blood and intestine, this study establishes a foundational resource for navigating these tissue-specific complexities and advancing targeted interventions for inflammatory bowel disease.

Subject of Research: Cell-type-resolved blood and gut expression quantitative trait locus mapping in inflammatory bowel disease

Article Title: Cell-type specific analyses in blood and gut identify cis-eQTL matching 140 IBD risk loci and entrectinib as repurposing candidate

Article References: Perée, H., Petrov, V. A., Tokunaga, Y., Kvasz, A., Farnir, F., Vieujean, S., Regimont, S., Mni, M., Wéry, M., Azarzar, S., Jacques, S., Fouillien, N., Karim, L., Deckers, M., Detry, E., Mayer, A., Stephan, R., Harshman, K., Mizoro, Y., … Rahmouni, S. (2026). Cell-type specific analyses in blood and gut identify cis-eQTL matching 140 IBD risk loci and entrectinib as repurposing candidate. Nature Communications, 17(1), Article 10031. https://doi.org/10.1038/s41467-026-76672-4

Image Credits: AI Generated

DOI: 10.1038/s41467-026-76672-4

Keywords: Inflammatory Bowel Disease, cis-eQTL, Single-Cell RNA Sequencing, Crohn’s Disease, Ulcerative Colitis, Entrectinib, Colocalization, Intestinal Biopsies, Peripheral Blood Mononuclear Cells, Transcriptomics, Regulatory Variants, Genome-Wide Association Studies, NEK7, Immune Cell Populations, Lamina Propria, Epithelial Cells, Drug Repurposing