pmc00040411793Br J DermatolBr J DermatolThe British journal of dermatology0007-09631365-2133406508791227925410.1093/bjd/ljaf277NIHMS2097466ArticleGenomic loci and molecular genetic mechanisms for hidradenitis suppurativaBroadawayK. Alaine*1SunQuan*23EdmistonSharon N45CurrinKevin W1VadlamudiSwarooparani1Miller-FlemingTyne W67ShiYue8910FajgenbaumKristen11Melendez-GonzalezMaria4BuiHelen12BlumFranklin R4WesterkamLinnea4ShamsRayad13MallelaTeja4LevittBrandt14LinLan4HaoHonglin4MemiliAylin13StraubPeter67ZhouWei8915HarrisKathleen Mullan1416MartinAlicia8910CoxNancy J67LiuZhi45ThomasNancy E45LiYun*21MohlkeKaren L*1SayedChristopher J*4Department of Genetics, University of North Carolina, Chapel Hill, NC, USA,Department of Biostatistics, University of North Carolina, Chapel Hill, NC, USA,Center for Computational and Genomic Medicine, Children’s Hospital of Philadelphia, Philadelphia, PA, USA,Department of Dermatology, University of North Carolina, Chapel Hill, NC, USA,Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC, USA,Department of Medicine, Division of Genetic Medicine, Vanderbilt University Medical Center, Nashville, TN, USA,Vanderbilt Genetics Institute, Vanderbilt University Medical Center, Nashville, TN, USA,Program in Medical and Population Genetics, Broad Institute of MIT and Harvard, Cambridge, MA, USA,Stanley Center for Psychiatric Research, Broad Institute of MIT and Harvard, Cambridge, MA, USA,Analytic and Translational Genetics Unit, Massachusetts General Hospital, Boston, MA, USA,Nashville Skin Comprehensive Dermatology Center, Nashville, TN, USA,Department of Dermatology, Howard University College of Medicine, Washington, DC, USA,School of Medicine, University of North Carolina, Chapel Hill, NC, USA,Carolina Population Center, University of North Carolina, Chapel Hill, NC, USA,Psychiatric and Neurodevelopmental Genetics Unit, Center for Genomic Medicine, Massachusetts General Hospital, Boston, MA, USA,Sociology Department, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA

K.A.B. and Q.S. contributed equally to this work; Y.L., K.L.M., and C.J.S. contributed equally to this work

Correspondence: Karen Mohlke. mohlke@med.unc.edu17720251710202521720251935948958Background

Hidradenitis suppurativa (HS) is a common, chronic, and debilitating inflammatory disease that most commonly affects intertriginous skin. Despite its high heritability, the genetic underpinnings of HS remain poorly understood.

Objective

To identify genetic signals associated with HS, determine genetic relationships with other diseases, and investigate potential molecular genetic mechanisms.

Methods

We performed a genome-wide association study meta-analysis of six studies, totaling 4,540 cases and over 1 million controls and identified genetic correlations with other common diseases. We integrated the HS data with expression quantitative trait loci from 10 trait-relevant tissues, epigenomic and transcriptomic data from human scalp, differential expression data from HS lesions versus adjacent skin, and mesenchymal Hi-C chromatin looping data. To identify functional noncoding variants, we performed transcriptional reporter assays for signals near KLF5 and SOX9.

Results

We identified eleven significant HS signals across seven loci: four corresponded to previously reported associations, four represented novel signals within known loci, and three were signals in newly implicated loci. We identified significant genetic correlation between HS and other inflammatory conditions, particularly inflammatory bowel disease, rheumatoid arthritis, type 2 diabetes, and asthma. We prioritized candidate genes for the 11 signals. The risk allele at KLF5 exhibited 10-fold greater transcriptional activity than the non-risk allele, while risk alleles at SOX9 showed significantly reduced transcriptional activity.

Conclusions

Our results provide insights into potential genetic mechanisms underlying HS and suggest potential therapeutic targets for this challenging condition.

Plain Language Summary

Hidradenitis suppurativa (HS) is a chronic skin disease that causes painful, recurring nodules and inflammation, most commonly in areas where skin rubs together. Although HS runs in families, its genetic causes are not well understood. To uncover genetic risk factors, we analyzed data from six studies, including 4,540 people with HS.

We discovered 11 genetic signals linked to HS. Four of these signals had been previously reported, while seven were new. We also found genetic connections between HS and other inflammatory diseases, such as inflammatory bowel disease, rheumatoid arthritis, type 2 diabetes, and asthma.

To better understand how these genetic signals might influence HS, we examined gene activity in different types of skin cells and in HS lesions. Some signals may affect how skin protects the body while others affect the body’s defense system. We also performed experiments to test how HS-associated DNA variants affect genes. We found that genetic changes near KLF5 and SOX9—two important genes for skin development and healing—directly affected gene activity.

These findings provide new insights into the genetic basis of HS and suggest possible biological pathways involved in the disease. By understanding these genetic mechanisms, researchers can explore potential targets for future treatments, offering hope for improved therapies for people with HS.

Introduction

Hidradenitis suppurativa (HS) is a highly heritable1–3 chronic inflammatory skin disease characterized by recurrent painful nodules, abscesses, and tunnels, primarily affecting intertriginous regions4. HS is associated with many comorbidities, including psoriasis5, inflammatory bowel disease (IBD)6, and polycystic ovarian syndrome (PCOS)7. Due to the complexity of HS pathophysiology, the manifestations and progressions of HS phenotypes are highly heterogeneous, as observed by differences across environmental and genetic factors8,9.

Genetic studies have begun to illuminate the HS etiology. Rare variants analyses have pointed to genes related to the γ-secretase complex and autoinflammatory pathways3,10–14. Two genome-wide association studies (GWAS) have identified common variants associated with HS. Our previous meta-analysis of HS Program for Research and Care Excellence (ProCARE) and two biobanks identified associations near SOX9 and KLF515. More recently, a meta-analysis of Nordic, UK, and US-based studies provided further support for these signals and identified four additional common-variant signals: one in WNT10A, one near TMED10, and intergenic signals on chromosomes 6 and 914.

Despite these discoveries, much of HS heritability remains unexplained, and the functional consequences of GWAS associations across populations are unclear. To address these gaps, we expanded ProCARE and conducted a GWAS meta-analysis across six studies with 4,540 cases, identifying eleven association signals. To link these signals to potential genes, we integrated them with transcriptomic and epigenomic data. Lastly, to investigate plausible causal regulatory variants, we analyzed allelic differences in transcriptional activity of variants at KLF5 and SOX9, providing mechanistic evidence of variant effects.

Materials and MethodsGenetic Association and Meta-analyses

We expanded the HS ProCARE study to 986 patients and used METAL16 to conduct a GWAS meta-analysis across six studies (Table S1, Supplementary Methods). To control for population structure, we stratified analyses by population: self-reported black/African participants (AFR) and all other participants (EUR+AMR) (Table S1). GWAS were performed separately in each cohort by population group before meta-analysis. We first performed a GWAS meta-analysis by combining all population groups from all cohorts. Next, we performed analyses stratified by population group, separating AFR participants from all other participants, and then performed analyses stratified by sex (Table S2, Supplementary Methods).

We defined a locus as a lead variant with P-value < 5×10−8 and all variants within 1 Mb of the lead variant, extending to 2 Mb in the HLA region due to extensive linkage disequilibrium (LD). A locus can harbor several genetic signals distinctly associated with HS; to identify conditionally distinct signals within a locus, we performed approximate conditional analysis using GCTA17,18 (Supplementary Methods).

Fine-Mapping

To identify likely causal variants at each locus, we used the Bayesian fine-mapping method SuSiE19 (susieR package 0.12.35) to identify 95% credible sets (Supplementary Methods). To match a signal to credible set, we required that the lead variant be in the credible set.

Genetic Correlation with Comorbid Traits

We applied LD score regression (LDSC)20,21 to estimate the genetic correlation between HS and celiac disease22, IBD23, PCOS24, rheumatoid arthritis (RA)25, type 2 diabetes (T2D)26, atopic dermatitis27, psoriasis28, schizophrenia29, and asthma30 (Supplementary Methods). We reviewed the lead HS variants in the GWAS Catalog31 to identify reported associations with related traits.

Candidate Genes and Regulatory Elements

To prioritize potential candidate genes, we generated variant-level annotation using WGSA Annotator32, tested for colocalization between HS signals and signals from ten sources of expression quantitative trait loci (eQTL) data in HS-relevant tissues33–37, and we queried if genes were differentially expressed between cell types of the scalp38 or differentially expressed in HS lesions versus unaffected skin of HS patients39 (Supplementary Methods).

To investigate potential regulatory elements and their likely effector genes, we examined Hi-C chromatin interaction data from mesenchymal cells40 and compared the HS-associated variants to regions of chromatin accessibility in human scalp cell types38 (Supplementary Methods).

Transcriptional Reporter Assays

We performed transcriptional reporter assays in the HaCaT cell line to test allelic differences in transcriptional activity. DNA fragments spanning rs981625 or a 3-variant haplotype (rs17226067, rs17825774, rs17825799) were cloned into a luciferase reporter vector and transfected into HaCaT cells (Table S3). Luciferase and Renilla activities were measured 48 hours after transfection, and the differences in activity between alleles were tested using two-tailed Student’s t-tests (Supplementary Methods).

ResultsIdentification of HS Association Signals

In a GWAS meta-analysis of all participants, we identified five genome-wide significant loci (P-value < 5×10−8), including two we previously reported near SOX9 and KLF5, two that others previously reported on chromosome 6 and on chromosome 914, and an additional locus near SUCNR1 on chromosome 3 (Figure 1, Table 1, Table S4, Figures S1–S6).

We then performed population- and sex-stratified GWAS meta-analyses. In a EUR+AMR analysis, we found a sixth locus near NIN (P-value = 2.5×10−8, Figure 1, Table 1, Figures S7–8). The lead variant, rs10146136, is rare in African populations (1000G minor allele frequency (MAF) = .005) but common in European populations (1000G MAF = 0.133). While we could detect the variant in three AFR studies, it did not meet the significance threshold in the full meta-analysis (P-value = 1.3×10−7) (Figure S9). In the male-only HS analyses, we found a seventh locus near MXRA7 (rs9901332, P-value = 3.4×10−8) (Figures 1–2, Table 1, Table S4, Figure S10). The lead variant’s effect size in men (beta = 0.43, SE = 0.08) is comparable to the strongest effect of any other lead variant (e.g. rs981625 near KLF5, beta = 0.44, SE = 0.06), but it had no observable effect in women, despite a four-fold larger sample size (beta = −0.04, SE = 0.04, P-value = 0.32). By leveraging population- and sex-stratified analyses, we reinforced the role of common genetic variants in HS pathogenesis while identifying two additional signals that may help elucidate the genetic basis of HS.

To discover additional genetic variants that influence HS risk, we performed approximate conditional analyses on all seven loci17,18. In the all-population meta-analysis, we identified a second signal on chromosome 9 near KLF4 (Table 1, Table S5, Figure S11). The EUR+AMR summary data confirmed this signal and revealed three more: a second signal on chromosome 6 near the HLA region (Figure S12–S13) and two additional signals near KLF5 (Figure S14–S17, Table S5). The HLA and KLF5 non-primary signals narrowly missed the significance threshold in the all-population analysis (conditional P-values ≤ 1.7×10−7). Across all analyses, we found 11 signals in 7 loci.

Fine-Mapping

We identified credible sets of plausibly causal variants for our HS signals. We identified nine credible sets that matched with nine of the eleven HS association signals (Table S6); the two HLA signals did not converge. The primary KLF5 credible set consisted of a single variant (rs981625) that is in low LD with the previous lead (rs17090189 LD r2 = 0.24 in EUR subset of UK Biobank). While the NIN signal and the non-primary KLF5 signals narrowly missed genome-wide significance in the all-population meta-analysis (1.7×10−7 ≤ P-values = 1.3×10−7), they were detected in the all-population meta-analysis using fine-mapping, suggesting that the three signals were not exclusive to the population-stratified results.

Since the extent of LD is typically shorter in AFR populations, we compared credible sets sizes from all-population versus EUR+AMR data. Five of eight credible sets were smaller in the all-population analysis, two were the same size, and only one credible set, near NIN, yielded a smaller credible set with the EUR+AMR data. While differences in variant representation across studies could also influence credible set size, all EUR-specific credible set variants were well represented in the full summary statistics from the all-population analysis. These results highlight the value of multi-population meta-analyses, even when under-represented populations do not yield population-specific signals.

Genetic Correlation with Other Complex Traits

To better understand the genetic relationships between HS and other complex traits, we tested for genetic correlation. We established a significant genetic correlation between HS and IBD (rg = 0.53, P-value = 6.0×10−4)23, RA (rg = 0.31, P-value = 6.0×10−4)25, T2D (rg = 0.37, P-value = 1.1×10−7)26, and asthma (rg = 0.28, P-value = 4.0×10−4)30 (Figure 3, Table S7). As a negative control, schizophrenia showed no evidence of correlation (rg = −0.002, P-value = 0.97). We also found that the lead variant for the primary KLF5 signal, rs981625, is also the lead variant for related traits including eczema, asthma, hay fever, and allergic rhinitis30,41,42 (UK Biobank traits, see web resources).

Candidate Genes and Regulatory Elements

To prioritize candidate genes and to identify potential regulatory elements affected by HS risk variants, we integrated GWAS findings with variant annotations and transcriptomic, eQTL, chromatin accessibility, and Hi-C data (Tables S8–S14). We strengthened evidence for KLF5 and SOX9 as candidate genes. KLF5 was less expressed in HS lesions vs unaffected skin (log2FC = −0.6, P-value = 2.9×10−9) and exhibited high gene activity in scalp keratinocyte clusters (log2 fold-change (log2FC) 0.6 – 2.6, P-values ≤ 4.0×10−166) but no other cell type (Figure 4, Tables S10). The KLF5 primary signal lead variant was in a regulatory element that appears to be only accessible in keratinocytes (Figure 5, Table S13). SOX9 was highly expressed in hair-follicle, spinous, and basal keratinocyte cell clusters (log2FC 0.3 – 5.2, P-values ≤ 5×10−86), but no other scalp cell type (Figure 4, Table S10). The SOX9 lead and LD proxies were also in a regulatory region showing chromatin accessibility in keratinocytes but not seen in other profiled cell types (Figure 5, Table S13). These findings add support that KLF5 and SOX9 signals could act through a keratinocyte-accessible regulatory element not observed in other cell types examined to influence KLF5 and SOX9 expression levels in keratinocytes.

The two chromosome 6 signals were near HLA-DRA and HLA-DRB5. Both genes were highly expressed in immune-relevant cell type clusters, including dendritic cells and macrophages, and were more highly expressed in HS lesion than non-lesion tissue (Figure 4, Table S10). In addition, all proxies for the second HLA signal were in one a regulatory element that appears to be only accessible in lymphoids (Table S13), consistent with the known role of HLA genes in immunity.

Mesenchymal Hi-C chromatin interaction data40 suggested long-range regulatory effects for lead variants on chromosomes 14 and 9. rs10146136 on chromosome 14 interacted with NIN (Figure S18), suggesting a regulatory link to NIN (107 kb from TSS). NIN showed increased expression in HS lesions (Figure 4, Table S10). Similarly, both lead variants on chromosome 9 (rs1994799 and rs1535620) interacted with the promoter of KLF4 (Figure 6, Figure S19), despite 900 kb between the leads and the KLF4 TSS. KLF4 demonstrated decreased expression in HS lesions (log2FC = −0.96, P-value = 1.0×10−10, Figure 4, Table S10). Hi-C links were not observed in mesenchymal for the closer ACTL7B gene (185 kb from TSS), and ACTL7B showed no change in lesion versus non-lesion tissue (log2FC = 0.93, P-value = 0.71) (Figure 4). These findings suggested potential regulatory links between the identified variants and target genes, although further investigation is needed.

For nine of the eleven signals, at least one lead or high-LD proxy variant in the signal’s credible set was located within a scalp regulatory element (Table S13). These data suggest that causal variants reside within these elements, refining the set of most plausible causal variants (Table S6).

Variant Function

To investigate allelic effects of two signals with clear lead variants, we examined regulatory function in transcriptional reporter assays in epidermal HaCaT cells. The KLF5 rs981625-C HS risk allele showed 10-fold greater transcriptional activity than the reference allele (P-value < 2.9 × 10−7) (Figure 5, Figure S20). At SOX9, a three-variant haplotype that included the lead HS risk variant rs17226067-G as well as rs17825774-A and rs17825799-A, showed >2-fold lower transcriptional activity than the reference haplotype (P-value = 0.001) (Figure 5, Figure S20). These data suggest that rs981625 and at least one variant in the haplotype are plausible functional regulatory variants responsible for HS risk at the SOX9 and KLF5 signals.

Discussion

We performed a meta-analysis of HS on 4,540 cases, identifying 11 HS signals in 7 loci. We identified five loci in the all-population meta-analysis, an additional locus in a EUR+AMR analysis and a locus in the male-only analysis. Four loci overlapped known regions (SOX9, KLF5, KLF4, and HLA), while three were novel: SUCNR1, NIN (EUR+AMR-only locus), and MXRA7 (male-only locus). Collectively, these results highlight the complexity of HS genetics, the importance of considering population structure, and the potential of sex differences in HS etiology.

We established a genetic correlation between HS and four common inflammatory and metabolic conditions: IBD23, RA25, T2D26, and asthma30. The correlation with IBD was the strongest; patients with IBD were nine times more likely to develop HS compared to the general population43, suggesting that overlapping immune and inflammatory mechanisms may contribute to both conditions. Rare variants in NOD2, a gene that has been widely implicated in Crohn’s disease44, have been associated with HS45. While many of these variants have been classified as benign or likely benign in silico46, one variant (rs104895452) has a high probability of being deleterious silico45,47. Both HS and IBD are often successfully treated with TNF-alpha inhibitors48,49, highlighting how shared genetic architecture can inform therapeutic strategies. Further, rs981625, the lead variant at KLF5, was also the lead variant for eczema, asthma, hay fever, and allergic rhinitis30,41,42. Understanding these shared genetic bases could offer insights into common pathophysiological mechanisms and potentially guide future research toward more integrated therapeutic strategies.

Previously, we described how dysregulation of KLF5 and SOX9 could be linked to the perturbations in follicular and epidermal differentiation, inflammation and wound healing that are characteristic of HS. Here, we identified likely causal variants driving these non-coding signals. The KLF5 rs981625-C allele was associated with greater transcriptional activity and higher HS risk, aligning with mouse studies where Klf5 overexpression led to hyperkeratosis and follicular occlusions50. Similarly, a SOX9 haplotype including the rs17226067-G allele was associated with reduced transcriptional activity and increased risk of HS, consistent with the observation that mice deficient in Sox9 exhibit follicular degeneration, epidermal hyperplasia, and dermal scarring51.

Chromatin interaction data40 implicated KLF4 as a target gene for two chromosome 9 signals. A transcription factor linked to psoriasis,52 KLF4 regulates wound healing, skin barrier function, proliferation, and follicular/epidermal differentiation53. It is both regulated by and regulates NOTCH1 signaling and is induced by inflammatory cytokines54,55. KLF4 stimulates production of IL-36 and IL-8, both potent inducers of neutrophil-mediated inflammation, IL-6, and IL-10. In human keratinocyte knockdown models of NCST, a ɣ-secretase family member that is mutated in some familial HS cases, KLF4 was upregulated, particularly in the presence of S. aureus, consistent with the role of bacteria in potentiating inflammation in HS56. Overexpression of Klf4 in mice led to impaired differentiation of follicular epithelium with impaired hair growth57. Given its similarities to SOX9 and KLF5 in follicular/epidermal differentiation, wound healing, and inflammation, KLF4 may also influence HS pathogenesis.

We also identified plausible candidate genes at our other loci. HLA-DRA, which is upregulated in HS lesions, regulates antigen presentation and may contribute to inflammation via dysregulated host-microbiome interactions. NIN, a microtubule regulator, is crucial for immune function, barrier function58 and neutrophil extracellular trap regulation59. In sepsis, IL-6 downregulation reduced NIN in monocytes, which allowed for persistence of pathogens and potentiation of other inflammatory bacterial responses60. A similar process could contribute to dysbiosis, biofilm formation, and chronic inflammation in HS tissue61. SUCNR1, the succinate receptor, regulates inflammasome priming and glucose homeostasis. Deficiency in Sucnr1 is linked to obesity, glucose intolerance and enhanced inflammatory signaling62, but also with reduced inflammasome activation in mouse colitis models, suggesting it may also play a pro-inflammatory role in some settings63. In macrophages, KLF4 knockdown has been associated with reduced expression of SUCNR164, potentially linking these risk loci together in HS pathogenesis. Finally, MXRA7 may protect against pro-psoriasis cytokines65, as mouse models deficient in Mxra7 develop more severe disease. Loss of MXRA7 resulted in epidermal proliferation increases, expression of TNF-α, IL-6, and IL-1, and impaired wound healing66, all of which are linked to HS pathogenesis. However, many genes near MXRA7 could contribute to risk. Further studies are needed to investigate these and/or other nearby genes in their roles in HS pathogenesis.

This GWAS meta-analysis has limitations. First, due to sample size constraints, we examined at most two population groups per cohort and analyzed people together from European and admixed American populations. The results cannot be interpreted to infer racial or ethnic differences. Most signals were not significant in the AFR GWAS (0.005 ≤ P-value ≤ .78), highlighting the need for larger AFR HS cohorts to identify additional HS GWAS signals and improve fine-mapping. Allele frequency differences affect association detection; all but one lead variant (SOX9 lead variant rs17226067) was less common in the AFR versus EUR+AMR populations. New signals, especially those with different effects across populations, should be validated. Second, since individuals with HS suffer an average of 7–10 years between manifestation of first symptoms until a formal diagnosis is made67,68, some controls may be affected with HS, biasing results towards the null. Third, we found only one colocalization between HS signals and eQTLs; as eQTL studies expand, we expect to uncover further HS-eQTL colocalizations that could enhance understanding of HS pathogenesis. Fourth, this study focused on common variants (MAF ≥ 1%); HS is known to be associated with rare variants45, and integrating rare variant analyses with future common genetic association studies would enhance our understanding of its genetic basis.

Overall, we identified potential mechanisms and candidate genes for HS by stratifying analyses by population and sex, conducting conditional analyses, and integrating HS signals with complementary GWAS traits, expression and regulatory data, and studies of variant function. This work represents one of the largest HS meta-analyses to date (effective sample size: 17,244) and the largest meta-analysis in AFR participants (effective sample size: 4,212). While further research with larger and more diverse cohorts is needed to fully elucidate the genetic contributors to HS, our findings provide valuable insights into disease pathology and may inform the development of novel therapeutics.

Supplementary MaterialAcknowledgements

We thank the HS ProCARE patients for their contributions. We thank the University of North Carolina BioSpecimen Processing Facility Lab (https://bsp.web.unc.edu), particularly Patricia Basta PhD, for blood/sample processing, nucleic acid extraction, storage, and disbursements, and the University of North Carolina Mammalian Genotyping Core (https://mgc.unc.edu/), particularly Amanda Gerringer, MSFS, for generating genotype data. We acknowledge the participants and investigators of the FinnGen study. This research has been conducted using data from UK Biobank, a major biomedical database (http://www.ukbiobank.ac.uk/), under application number 25953.

Funding sources

We acknowledge support from National Institutes of Health grants R01 AR083790, R01 HD073342 and P01 HD031921. The project described was supported by the National Center for Advancing Translational Sciences (NCATS), National Institutes of Health, through Grant Award Number UM1 TR004406. The UNC Biospecimen Processing Facility core and core staff are supported by Cancer Center Support Grant P30 CA016086 and the Center for Environmental Health and Susceptibility CEHS Grant number P30 ES010126. The Mammalian Genotyping Core is supported by the Lineberger Comprehensive Cancer Center, the CEHS, and facility users. Additional sources of support include the UNC School of Medicine, School of Public Health, University Cancer Research Fund, and the North Carolina Biotechnology Center. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. The BioVU project at VUMC is supported by numerous sources including the NIH funded Shared Instrumentation Grant S10 OD017985, S10 RR025141, and S10 OD025092; CTSA grants UL1 TR002243, UL1 TR000445, and UL1 RR024975; its contents are solely the responsibility of the authors and do not necessarily represent official views of the National Center for Advancing Translational Sciences. BioVU genomic data are also supported by investigator-led projects that include U01 HG004798, R01 NS032830, RC2GM092618, P50 GM115305, U01 HG006378, U19 HL065962, R01 HD074711, and additional funding sources listed at https://victr.vumc.org/biovu-funding/. The Genotype-Tissue Expression (GTEx) Project was supported by the Common Fund of the Office of the Director of the National Institutes of Health, and by National Cancer Institute, NHGRI, National Heart, Lung, and Blood Institute, National Institute on Drug Abuse, National Institute of Mental Health, and National Institute of Neurological Disorders and Stroke.

Conflicts of interest

CS is an investigator for Abbvie, Novartis, Incyte, InflaRx, UCB, and Astrazeneca, a consultant for Abbvie, Novartis, Sanofi, Sandoz, Astrazeneca, Incyte, InflaRx, UCB, Navigator Medicines, Moonlake Immunotherapeutics, and Alumis, a speaker for Abbvie, Novartis, and UCB, and is the secretary of the HS Foundation.

Ethics statement

The UNC Chapel Hill Institutional Review Board approved this study.

Patient consent

Patients with HS presenting to UNC Dermatology outpatient clinics provided written informed consent.

Web resources

GTEx summary results: https://console.cloud.google.com/storage/browser/gtex-resources

Neale lab, GWAS summary statistics of UK Biobank: http://www.nealelab.is/uk-biobank/

Referencesvan StraalenKR, PrensEP, WillemsenG, BoomsmaDI & van der ZeeHH Contribution of Genetics to the Susceptibility to Hidradenitis Suppurativa in a Large, Cross-sectional Dutch Twin Cohort. JAMA Dermatol 156, 1359–1362 (2020).33052394 Kjaersgaard AndersenR Evidence of gene-gene interaction in hidradenitis suppurativa: a nationwide registry study of Danish twins. Br J Dermatol 186, 78–85 (2022).34289077 BruinsmaRL Assessment of familial risk in patients with hidradenitis suppurativa. Br J Dermatol 184, 753–754 (2021).33152101 VinkelC & ThomsenSF Hidradenitis Suppurativa: Causes, Features, and Current Treatments. J Clin Aesthet Dermatol 11, 17–23 (2018).30519375 GargA Comorbidity screening in hidradenitis suppurativa: Evidence-based recommendations from the US and Canadian Hidradenitis Suppurativa Foundations. J Am Acad Dermatol 86, 1092–1101 (2022).33493574 PhanK, TatianA, WoodsJ, CainsG & FrewJW Prevalence of inflammatory bowel disease (IBD) in hidradenitis suppurativa (HS): systematic review and adjusted meta-analysis. Int J Dermatol 59, 221–228 (2020).31631340 PhanK, CharltonO & SmithSD Hidradenitis suppurativa and polycystic ovarian syndrome: Systematic review and meta-analysis. Australas J Dermatol 61, e28–e33 (2020).31261440 van StraalenKR, PrensEP & GudjonssonJE Insights into hidradenitis suppurativa. J Allergy Clin Immunol 149, 1150–1161 (2022).35189127 MaroneseCA, MoltrasioC & MarzanoAV Hidradenitis Suppurativa-Related Autoinflammatory Syndromes: An Updated Review on the Clinics, Genetics, and Treatment of Pyoderma gangrenosum, Acne and Suppurative Hidradenitis (PASH), Pyogenic Arthritis, Pyoderma gangrenosum, Acne and Suppurative Hidradenitis (PAPASH), Synovitis, Acne, Pustulosis, Hyperostosis and Osteitis (SAPHO), and Rarer Forms. Dermatol Clin 42, 247–265 (2024).38423685 PinkAE, SimpsonMA, DesaiN, TrembathRC & BarkerJNW γ-Secretase mutations in hidradenitis suppurativa: new insights into disease pathogenesis. J Invest Dermatol 133, 601–607 (2013).23096707 BalićA, MarinovićB & Bukvić MokosZ The genetic aspects of hidradenitis suppurativa. Clin Dermatol 41, 551–563 (2023).37652193 Morales-HeilDJ Rare missense variants in the SH3 domain of PSTPIP1 are associated with hidradenitis suppurativa. HGG Adv 4, 100187 (2023).37013170 MetpallyRP Identification of Novel Genetic Risk Variants Associated with Hidradenitis Suppurativa in an Exome Sequencing Cohort of 92,455 Individuals. Dermatology 240, 739–749 (2024).39396498 Kjærsgaard AndersenR A genome-wide association meta-analysis links hidradenitis suppurativa to common and rare sequence variants causing disruption of the Notch and Wnt/β-catenin signaling pathways. J Am Acad Dermatol S0190–9622(24)03292–4 (2024) doi:10.1016/j.jaad.2024.11.050.SunQ Genetic Variants Associated With Hidradenitis Suppurativa. JAMA Dermatol 159, 930–938 (2023).37494057 METAL: fast and efficient meta-analysis of genomewide association scans | Bioinformatics | Oxford Academic. https://academic.oup.com/bioinformatics/article/26/17/2190/198154.YangJ, LeeSH, GoddardME & VisscherPM GCTA: a tool for genome-wide complex trait analysis. Am J Hum Genet 88, 76–82 (2011).21167468 YangJ Conditional and joint multiple-SNP analysis of GWAS summary statistics identifies additional variants influencing complex traits. Nat Genet 44, 369–375, S1–3 (2012).22426310 ZouY, CarbonettoP, WangG & StephensM Fine-mapping from summary data with the “Sum of Single Effects” model. PLOS Genetics 18, e1010299 (2022).35853082 Bulik-SullivanB An atlas of genetic correlations across human diseases and traits. Nat Genet 47, 1236–1241 (2015).26414676 Bulik-SullivanBK LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet 47, 291–295 (2015).25642630 JiangL, ZhengZ, FangH & YangJ A generalized linear mixed model association tool for biobank-scale data. Nat Genet 53, 1616–1621 (2021).34737426 LiuJZ Association analyses identify 38 susceptibility loci for inflammatory bowel disease and highlight shared genetic risk across populations. Nat Genet 47, 979–986 (2015).26192919 DayF Large-scale genome-wide meta-analysis of polycystic ovary syndrome suggests shared genetic architecture for different diagnosis criteria. PLoS Genet 14, e1007813 (2018).30566500 SaevarsdottirS Multiomics analysis of rheumatoid arthritis yields sequence variants that have large effects on risk of the seropositive subset. Ann Rheum Dis 81, 1085–1095 (2022).35470158 SuzukiK Genetic drivers of heterogeneity in type 2 diabetes pathophysiology. Nature 627, 347–357 (2024).38374256 Budu-AggreyA European and multi-ancestry genome-wide association meta-analysis of atopic dermatitis highlights importance of systemic immune regulation. Nat Commun 14, 6172 (2023).37794016 SakaueS A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet 53, 1415–1424 (2021).34594039 TrubetskoyV Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature 604, 502–508 (2022).35396580 TsuoK Multi-ancestry meta-analysis of asthma identifies novel associations and highlights the value of increased power and diversity. Cell Genom 2, 100212 (2022).36778051 CerezoM The NHGRI-EBI GWAS Catalog: standards for reusability, sustainability and diversity. Nucleic Acids Research 53, D998–D1005 (2025).39530240 LiuX WGSA: an annotation pipeline for human genome sequencing studies. J Med Genet 53, 111–112 (2016).26395054 GTEx Consortium. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science 369, 1318–1330 (2020).32913098 ShoreCJ Genetic effects on the skin methylome in healthy older twins. Am J Hum Genet 111, 1932–1952 (2024).39137780 KerimovN A compendium of uniformly processed human gene expression and splicing quantitative trait loci. Nat Genet 53, 1290–1299 (2021).34493866 VõsaU Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet 53, 1300–1310 (2021).34475573 XiaoQ Immunosuppression causes dynamic changes in expression QTLs in psoriatic skin. Nat Commun 14, 6268 (2023).37805522 Ober-ReynoldsB Integrated single-cell chromatin and transcriptomic analyses of human scalp identify gene-regulatory programs and critical cell types for hair and skin diseases. Nat Genet 55, 1288–1300 (2023).37500727 FreudenbergJM A Hidradenitis Suppurativa molecular disease signature derived from patient samples by high-throughput RNA sequencing and re-analysis of previously reported transcriptomic data sets. PLoS One 18, e0284047 (2023).37023004 SchmittAD A Compendium of Chromatin Contact Maps Reveals Spatially Active Regions in the Human Genome. Cell Rep 17, 2042–2059 (2016).27851967 JohanssonÅ, Rask-AndersenM, KarlssonT & EkWE Genome-wide association analysis of 350 000 Caucasians from the UK Biobank identifies novel loci for asthma, hay fever and eczema. Hum Mol Genet 28, 4022–4041 (2019).31361310 KichaevG Leveraging Polygenic Functional Enrichment to Improve GWAS Power. Am J Hum Genet 104, 65–75 (2019).30595370 YadavS Hidradenitis Suppurativa in Patients With Inflammatory Bowel Disease: A Population-Based Cohort Study in Olmsted County, Minnesota. Clin Gastroenterol Hepatol 14, 65–70 (2016).25952308 AshtonJJ, SeabyEG, BeattieRM & EnnisS NOD2 in Crohn’s Disease-Unfinished Business. J Crohns Colitis 17, 450–458 (2023).36006803 MuretK Comprehensive Catalog of Variants Potentially Associated with Hidradenitis Suppurativa, Including Newly Identified Variants from a Cohort of 100 Patients. International Journal of Molecular Sciences 25, 10374 (2024).39408704 MintoffD, BorgI & PaceNP NOD2 at the interface of hidradenitis suppurativa and inflammatory bowel disease-An in silico analysis. Exp Dermatol 33, e14928 (2024).37694954 de OliveiraASLE Transcriptome Meta-Analysis Confirms the Hidradenitis Suppurativa Pathogenic Triad: Upregulated Inflammation, Altered Epithelial Organization, and Dysregulated Metabolic Signaling. Biomolecules 12, 1371 (2022).36291580 SavageKT, FloodKS, PorterML & KimballAB TNF-α inhibitors in the treatment of hidradenitis suppurativa. Ther Adv Chronic Dis 10, 2040622319851640 (2019).SouzaRF, CaetanoMAF, MagalhãesHIR & CastelucciP Study of tumor necrosis factor receptor in the inflammatory bowel disease. World J Gastroenterol 29, 2733–2746 (2023).37274062 SurI, RozellB, JaksV, BergströmA & ToftgårdR Epidermal and craniofacial defects in mice overexpressing Klf5 in the basal layer of the epidermis. J Cell Sci 119, 3593–3601 (2006).16912082 VidalVPI Sox9 is essential for outer root sheath differentiation and the formation of the hair stem cell compartment. Curr Biol 15, 1340–1351 (2005).16085486 Ray-JonesH Mapping DNA interaction landscapes in psoriasis susceptibility loci highlights KLF4 as a target gene in 9q31. BMC Biol 18, 47 (2020).32366252 GhalebAM & YangVW Krüppel-like factor 4 (KLF4): What we currently know. Gene 611, 27–37 (2017).28237823 LebsonL Cutting edge: The transcription factor Kruppel-like factor 4 regulates the differentiation of Th17 cells independently of RORγt. J Immunol 185, 7161–7164 (2010).21076063 YinJ Transcription factor Krüppel-like factor 4 upregulated G protein-coupled receptor 30 alleviates intestinal inflammation and apoptosis, and protects intestinal integrity from intestinal ischemia-reperfusion injury. Immun Inflamm Dis 11, e940 (2023).37506161 ZhangY Inflammatory loop involving Staphylococcus aureus, IL-36γ, and cathepsin S drives immunity disorders in familial acne inversa keratinocytes. Heliyon 10, e31509 (2024).38947455 HanH Effect of overexpression of KLF4 on the growth and development of hair follicles in mice. Dev Genes Evol 233, 137–145 (2023).37561178 LeclandN, HsuC-Y, CheminC, MerdesA & BierkampC Epidermal development requires ninein for spindle orientation and cortical microtubule organization. Life Sci Alliance 2, e201900373 (2019).30923192 HaiderP Neutrophil Extracellular Trap Degradation by Differently Polarized Macrophage Subsets. Arterioscler Thromb Vasc Biol 40, 2265–2278 (2020).32673525 AkoumianakiT Uncoupling of IL-6 signaling and LC3-associated phagocytosis drives immunoparalysis during sepsis. Cell Host Microbe 29, 1277–1293.e6 (2021).34214493 ChopraD Innate immunity and microbial dysbiosis in hidradenitis suppurativa - vicious cycle of chronic inflammation. Front Immunol 13, 960488 (2022).35967376 KeiranN SUCNR1 controls an anti-inflammatory program in macrophages to regulate the metabolic response to obesity. Nat Immunol 20, 581–592 (2019).30962591 BausetC SUCNR1 Mediates the Priming Step of the Inflammasome in Intestinal Epithelial Cells: Relevance in Ulcerative Colitis. Biomedicines 10, 532 (2022).35327334 KeiranN SUCNR1 controls an anti-inflammatory program in macrophages to regulate the metabolic response to obesity. Nat Immunol 20, 581–592 (2019).30962591 NingJ Altered expression of matrix remodelling associated 7 (MXRA7) in psoriatic epidermis: Evidence for a protective role in the psoriasis imiquimod mouse model. Exp Dermatol 27, 1038–1042 (2018).29781547 SunZ MXRA7 is involved in monocyte-to-macrophage differentiation. Mol Immunol 171, 12–21 (2024).38735126 SaunteDM Diagnostic delay in hidradenitis suppurativa is a global problem. British Journal of Dermatology 173, 1546–1549 (2015).26198191 KokolakisG Delayed Diagnosis of Hidradenitis Suppurativa and Its Effect on Patients and Healthcare System. Dermatology 236, 421–430 (2020).32610312

HS genetic association signals. A: Manhattan plot of the associations between variants and HS in all populations. Variants within 1 Mb of a previously known HS locus are colored in blue; variants within 1 Mb of a new HS locus (P-value ≤ 5×10−8) are colored in red. Loci with asterisks represent multi-signal loci. B: Variant associations for new HS loci. Left: a signal near SUNCR1 was significantly associated with HS in the all-population analysis; Middle: a signal near NIN was significantly associated with HS in the EUR+AMR stratified analysis; Right: a signal near MXRA7 was significantly associated with HS in males.

HS locus detected in male-only GWAS.

Left: Variants associated with HS in men (top) versus women (bottom). Right: Forest plot of male-only and female-only GWAS results. The MXRA7 locus is strong in men and shows consistent direction and strength of effect across all contributing studies. An effect is not detected in any female-only or sex-combined GWAS results. Studies noted with an asterisk (*) were not included in the meta-analysis due to small sample size.

Genetic correlation between HS and other complex traits.

Green indicates Bonferroni-significant correlations. IBD: inflammatory bowel disease; T2D: type 2 diabetes; PCOS: polycystic ovarian syndrome; RA: rheumatoid arthritis.

Differential expression of prioritized candidate genes in lesion vs non-lesion tissue and in scalp cell types.

We prioritized 20 genes that were within 100 kb of a lead variant, or the nearest protein-coding gene if none were within 100 kb, and genes with external evidence of relevance to HS. Top: the log2 fold change of gene expression in HS lesion versus matched non-lesion tissue. Bottom: the log2 fold change of gene expression in scalp cell types. Only increased expression in a cell type is shown. To be included, the cell type must be more highly expressed in at least one prioritized gene. Circles show the size of the fold change. Orange: keratinocytes; yellow: melanocytes; green: fibroblasts; purple: immune cells; brown: muscle or endothelial cells.

HS loci <italic toggle="yes">KLF5</italic> (left) and <italic toggle="yes">SOX9</italic> (right) may influence expression levels of their target genes in keratinocytes via a regulatory element that appears to be only accessible in keratinocytes.

Top row: Variant associations with HS for the two signals. Middle left: Genome browser plots showing scalp chromatin accessibility by cell type. The lead variant of KLF5 (rs981625) is located within a keratinocyte-accessible regulatory element that was not observed in other examined cell types (chr13_73465564_73466065) and is strongly conserved across vertebrates. No other LD proxies of the lead variant are located within accessible chromatin. KLF5 was much more strongly expressed in keratinocytes than any other cell type (log2fold change = 0.56 – 1.93 for keratinocyte cell clusters, P-values ≤ 4×10−166). Bottom left: the rs981625-C allele, which was associated with higher HS risk, showed a 10-fold greater transcriptional activity than the rs981625-G reference allele (P-value = 2.9 × 10−7). Middle right: the SOX9 lead variant rs17226067, and its LD proxies rs17825774 and rs17825799 (r2 = 1, shown in gold) are located within a region of open chromatin detected in keratinocytes but not in other surveyed cell types (chr17_71519287_71519788); no other LD proxies are located within accessible chromatin. SOX9 was also more highly expressed in hair-follicle (HF) keratinocyte clusters than other scalp cell types (log2 fold-change = 0.28 – 5.19, P-values ≤ 5×10−86). Bottom right: the 3-variant risk haplotype of rs17226067-G, rs17825774-A, and rs17825799-A alleles showed >2-fold reduced transcriptional activity versus the reference haplotype AGG haplotype (P-value = 0.001).

Both conditionally distinct signals on chromosome 9 exhibit a significant Hi-C chromatin interaction peak in mesenchymal data.

A: Variants associated with HS at two conditionally distinct signals. B: Mesenchymal cell Hi-C data40 anchored around rs1994799, the lead variant for the primary HS signal. C: Mesenchymal cell Hi-C data anchored around rs1535620, the lead variant for the secondary HS signal. The target regions immediately flanking the anchor regions and target regions with FitHiC Q-values ≥ 0.01 are not shown. Arc colors are scaled by strength of association measured by −log10(P-value), with the darkest lines showing the link with the strongest P-value. Hi-C data in the red box, centered around KLF4, showed significant interactions; no significant interactions were observed around the more proximal gene ACTL7B (gray box).

Eleven lead variants associated with HS at seven loci

ChrNearby GenePopulationrsIDAllelesEAFORP-valueNeff
Previously reported signals
17 SOX9 ALLrs17226067G/A0.161.285.21E-1317,274
13 KLF5 ALLrs981625C/G0.951.554.68E-1215,521
6 HLA ALLrs6457592C/T0.531.195.18E-1217,274
9 KLF4 ALLrs1994799A/G0.441.171.22E-1017,274
Additional loci
3 SUCNR1 ALLrs2681502C/T0.061.353.87E-0816,933
14 NIN EUR+AMRrs10146136A/G0.131.252.47E-0813,062
17 MXRA7 Men onlyrs9901332T/C0.611.543.42E-082,266
Additional signals at established loci
9 KLF4 ALLrs1535620G/A0.491.177.44E-1016,933
6 HLA EUR+AMRrs9265701A/G0.471.199.50E-098,903
13 KLF5 EUR+AMRrs9796066C/T0.371.181.42E-0913,062
13 KLF5 EUR+AMRrs12100281A/G0.911.311.82E-0813,062

Population indicates samples included in the GWAS analysis. Chr: chromosome. Alleles: effect allele/noneffect allele. EAF: effect allele frequency. OR: odds ratio (exp(beta)). Neff: effective sample size. For non-primary signals, the P-values are from GCTA conditional analyses.

What is already known about this topic?

HS is a chronic inflammatory skin disease with a strong genetic component. Prior common-variant GWAS studies identified six genetic loci associated with HS, but the underlying biological mechanisms remain poorly understood.

What does this study add?

We identified 11 genetic signals, including seven novel signals.

Integrative analyses found genetic links between HS and other inflammatory diseases

Our study prioritized potential causal genes and functional follow-up identified variants near SOX9 and KLF5 that likely drive two non-coding signals.

What is the translational message?

Our findings reveal biological pathways involving keratinocyte function and inflammation, suggesting potential therapeutic targets for HS. Identifying causal genes could inform precision medicine approaches for treatment.