Skip to main content
Download PDF

Integration of functional genomics and statistical fine-mapping systematically characterizes adult-onset and childhood-onset asthma genetic associations

Genome Medicine volume 17, Article number: 35 (2025) Cite this article

Abstract

Background

Genome-wide association studies (GWAS) have identified hundreds of loci underlying adult-onset asthma (AOA) and childhood-onset asthma (COA). However, the causal variants, regulatory elements, and effector genes at these loci are largely unknown.

Methods

We performed heritability enrichment analysis to determine relevant cell types for AOA and COA, respectively. Next, we fine-mapped putative causal variants at AOA and COA loci. To improve the resolution of fine-mapping, we integrated ATAC-seq data in blood and lung cell types to annotate variants in candidate cis-regulatory elements (CREs). We then computationally prioritized candidate CREs underlying asthma risk, experimentally assessed their enhancer activity by massively parallel reporter assay (MPRA) in bronchial epithelial cells (BECs) and further validated a subset by luciferase assays. Combining chromatin interaction data and expression quantitative trait loci, we nominated genes targeted by candidate CREs and prioritized effector genes for AOA and COA.

Results

Heritability enrichment analysis suggested a shared role of immune cells in the development of both AOA and COA while highlighting the distinct contribution of lung structural cells in COA. Functional fine-mapping uncovered 21 and 67 credible sets for AOA and COA, respectively, with only 16% shared between the two. Notably, one-third of the loci contained multiple credible sets. Our CRE prioritization strategy nominated 62 and 169 candidate CREs for AOA and COA, respectively. Over 60% of these candidate CREs showed open chromatin in multiple cell lineages, suggesting their potential pleiotropic effects in different cell types. Furthermore, COA candidate CREs were enriched for enhancers experimentally validated by MPRA in BECs. The prioritized effector genes included many genes involved in immune and inflammatory responses. Notably, multiple genes, including TNFSF4, a drug target undergoing clinical trials, were supported by two independent GWAS signals, indicating widespread allelic heterogeneity. Four out of six selected candidate CREs demonstrated allele-specific regulatory properties in luciferase assays in BECs.

Conclusions

We present a comprehensive characterization of causal variants, regulatory elements, and effector genes underlying AOA and COA genetics. Our results supported a distinct genetic basis between AOA and COA and highlighted regulatory complexity at many GWAS loci marked by both extensive pleiotropy and allelic heterogeneity.

Background

Asthma is a common, complex lung disease with a significant genetic component [1,2,3]. The largest genome-wide association study (GWAS) of asthma to date leveraged data from more than 1.5 million individuals across multiple ancestries and identified more than 150 significant loci [4]. Genetic associations, however, do not reveal the molecular mechanisms underlying the pathogenesis of asthma. Moreover, while most people with a diagnosis of asthma share a similar set of symptoms, there are many subtypes in which individuals display unique sets of clinical characteristics and biomarker measurements [5,6,7,8,9,10,11,12,13]. Age of onset is an important criteria used in differentiating asthma subtypes [14], and recent GWAS of adult-onset asthma (AOA) and childhood-onset asthma (COA) suggested that the clinical heterogeneity between AOA and COA reflected differences in their underlying genetics [15, 16].

Yet, challenges remain in translating GWAS results into biological insights [17, 18]. Causal variants at individual loci are often elusive due to complex linkage disequilibrium (LD) structures. Furthermore, most associated variants map within non-coding regions of the genome, making it difficult to know their functional effects. It is generally assumed that non-coding GWAS variants exert their effects through changes in the expression of nearby genes. However, identifying the causal genes targeted by GWAS variants is not straightforward due to complexities of gene regulation, with cis-regulatory elements (CREs) often regulating more than one gene in more than one cell type or tissue and often over long distances [19, 20]. Efforts have been made to address these post-GWAS challenges by jointly analyzing GWAS results with expression quantitative trait loci (eQTLs), using techniques such as colocalization [21,22,23,24,25] and transcriptome-wide association studies [24, 26,27,28,29] (TWAS). Nonetheless, eQTLs only explain a small fraction of the heritability for most complex traits [30, 31], limiting their utility for identifying effector genes. Some studies have explored additional molecular phenotypes such as alternative splicing (s)QTLs [32], chromatin accessibility (ca)QTLs [33, 34], and DNA methylation (me)QTLs [35], but these QTLs are not widely available and some (caQTLs and meQTLs) do not point to their target genes directly. Importantly, colocalization and TWAS are prone to false positive findings [36,37,38]. In addition, to our knowledge, no studies to date have systematically explored the genetic underpinnings of AOA and COA in fine-mapping studies or by integrating functional data from a diverse range of cell types and modalities.

In this study, we used a combination of computational and experimental approaches to systematically fine-map genetic loci associated with AOA and COA [15]. Our innovative pipeline [39] identified putative causal variants at these loci, nominated and validated candidate CREs that are likely disrupted by putative causal variants, and prioritized effector genes supported by multiple lines of genetic evidence (Fig. 1). Collectively, our analyses highlighted distinct genetic bases of AOA and COA while revealing pervasive pleiotropy and allelic heterogeneity at asthma GWAS loci.

Fig. 1
figure 1

Study workflow. For each step, the input data and assay are shown in brackets. UKB: UK Biobank, OCR: open chromatin region, CRE: cis-regulatory element, MPRA: massively parallel reporter assay, eQTL: expression quantitative trait locus

Full size image

Methods

Study subjects

Genetic association studies were conducted using individuals from the UK Biobank (UKB), a population-based prospective study which has collected phenotypic information, biological samples, health records, and genetic data from over 500,000 participants [40]. We identified 20,702 AOA cases, 9613 COA cases, and 308,537 shared controls following the same inclusion/exclusion criteria and phenotype definitions described in our previous study [15] (Additional file 1: Table S1). Briefly, we excluded related individuals with poor-quality genotypes and ambiguous sex assignments. AOA cases were subjects diagnosed of asthma between 26 and 65 years of age, COA cases were subjects diagnosed of asthma before 12 years of age, and controls were subjects aged older than 38 years with no reported asthma. Individuals with chronic obstructive pulmonary disease (COPD), chronic bronchitis, and emphysema were further excluded from the AOA cases and controls.

GWAS of AOA and COA

We performed GWAS of AOA and COA [41] using an updated version of the imputed genotypes from the UKB following the same quality control and analysis pipeline described in our previous study [15]. Briefly, associations between 9,738,009 autosomal single-nucleotide polymorphisms (SNPs) and AOA/COA status were tested using logistic regression implemented in Hail 0.2 [42], with sex and the first 10 genetic principal components included as covariates.

Heritability enrichment analysis

We performed ATAC-seq in airway smooth muscle cells [43] (ASMCs) and harmonized chromatin accessibility data in 19 lung and seven blood cell types from three published studies [44,45,46] (Additional file 1: Table S2; Additional file 2: Supplementary Methods). Using stratified LD score regression [47] (S-LDSC), we estimated the heritability enrichment of open chromatin regions (OCRs) in individual cell types and cell lineages. We adjusted for annotations included in the baseline LD model [48] in all S-LDSC analyses.

Statistical fine-mapping

Functional fine-mapping was implemented through a two-step procedure. In the first step, we used an empirical Bayesian model, TORUS [49], to estimate a prior probability (functional prior) for each SNP using GWAS summary statistics and a set of input functional annotations (i.e., OCRs in cell lineages significantly enriched for AOA/COA heritability). Next, we used the summary statistics version of “sum of single effects” (SuSiE) [50, 51] model to fine-map LD blocks at GWAS loci. For each SNP, SuSiE estimates a posterior inclusion probability (PIP), which reflects the strength of evidence supporting it as a causal variant. To integrate functional annotations with fine-mapping, we specified the prior probabilities of individual SNPs using the functional priors computed by TORUS in the first step. For comparison, we also performed fine-mapping without using functional information (i.e., using uniform priors). See Additional file 2: Supplementary Methods for details.

We considered AOA and COA credible sets at the same LD block as shared if they shared more than half of their SNPs or if the total PIP of the shared SNPs was greater than 50% of the PIP of either credible set. If these criteria were not met, the credible set was considered to be AOA- or COA-specific.

Mapping candidate CREs and computing ePIPs

To map candidate CREs, we merged the ATAC-seq peaks across 20 lung and seven blood cell types [44, 45] using the bedtools [52] version 2.30.0 merge command with the -d option set to −1. The output was a set of non-overlapping OCRs, each of which was considered as a candidate CRE. The element PIP (ePIP) of a candidate CRE was defined as the sum of the PIPs of the credible set SNPs within the CRE.

Linking candidate CREs to target genes

We used four features to nominate likely target genes for each candidate CRE: (1) the nearest gene to the candidate CRE; (2) if the SNP with the highest PIP in the candidate CRE was an eQTL [21, 25, 53,54,55] for that gene; (3) if the candidate CRE interacted with the promoter of a gene by Promoter Capture Hi-C (PCHi-C) [46, 56,57,58], defined as ≥ 50% physical overlap between the candidate CRE and the non-promoter end of PCHi-C loops; and (4) if a regulatory element in the activity-by-contact (ABC) dataset [59] had ≥ 50% physical overlap with the candidate CRE, in which case the gene with the highest ABC score was considered as a putative target gene. For each candidate CRE, we only used eQTL, PCHi-C, and ABC data collected from tissues and/or cell types matching the CRE’s cell lineage(s) to identify likely target genes (Additional file 1: Table S2). Gene annotations were curated using mapgen R package [60] version 0.5.9, and we restricted our analysis to protein-coding genes.

Prioritizing effector genes

To prioritize candidate genes, we utilized both candidate CREs and exonic variants, which may disrupt protein-coding sequences. The derived gene scores summarized the total genetic evidence supporting the role of a gene in AOA or COA. The score of a gene \(g\), \({S}_{g}\), was the sum of the contributions of all variants that support \(g.\) We defined genes with gene score ≥ 0.95 as high-confidence candidate causal genes for AOA or COA. See Additional file 2: Supplementary Methods for details. To attribute a gene’s score to individual credible sets, we grouped the variants linked to the gene by credible sets and calculated the total contribution of each credible set to the gene.

Results

Leveraging functional annotations to fine-map AOA and COA GWAS loci

To identify causal variants of asthma, we statistically fine-mapped AOA and COA loci from GWAS in UKB (Methods; Additional file 2: Fig. S1). We employed a functional fine-mapping approach, which leveraged chromatin accessibility data from blood and lung cell types relevant to asthma pathogenesis to improve the resolution of fine-mapping. Using ATAC-seq peaks (Additional file 1: Table S2), we first mapped OCRs of each cell type. Because OCRs in cell types from the same cell lineage shared similar heritability enrichments (Additional file 2: Fig. S2), we pooled OCRs by lineage for assessing heritability enrichment and for all subsequent analyses (Methods). Both AOA (p = 6.98 \(\times\) 10−5) and COA (p = 6.13 \(\times\) 10−9) risk variants were significantly enriched in OCRs of lymphocytes (Fig. 2A). COA risk variants were also significantly enriched in OCRs of epithelial cells (p = 0.02) and mesenchymal cells (p = 0.02).

Fig. 2
figure 2

A Heritability enrichment estimates for OCRs in asthma-relevant cell types. Lymphoid: lung B cells, lung T cells, lung NK cells, blood B cells, blood T cells, blood NK cells; Myeloid: lung macrophage, blood myeloid dendritic cells, blood plasmacytoid dendritic cells, blood monocytes; Epithelial: alveolar type 1 cells, alveolar type 2 cells, pulmonary neuroendocrine cells, lung basal cells, lung ciliated cells, lung club cells, bronchial epithelial cells; Mesenchymal: lung matrix fibroblasts, lung myofibroblasts, lung pericytes, airway smooth muscle cells; Endothelial: lung arterial cells, lung capillary cells, lung lymphatic cells. Confidence intervals are ± 2 standard errors. B PIPs for SNPs in adult-onset asthma (left panel) and childhood-onset asthma (right panel) fine-mapping, with SNPs weighted by functional annotations (y-axis) or by uniform weights (x-axis). C Distribution of the number of SNPs in the adult-onset asthma and childhood-onset asthma credible sets. D Distribution of the number of shared and specific credible sets

Full size image

Next, we integrated OCRs of enriched cell lineages using a Bayesian hierarchical model [49] (Additional file 1: Table S3) and performed functional fine-mapping [50, 51] for all LD blocks harboring at least one genome-wide significant SNP (p < 5 \(\times\) 10−8) (Methods; Additional file 2: Supplementary Methods). For comparison, we also performed fine-mapping [50, 51] under the default setting that assumes all SNPs are equally likely to be causal a priori (Fig. 2B). To quantify the effect of incorporating functional information in fine-mapping, we assigned variants into one of three categories by their PIP: low-confidence (0.1 < PIP ≤ 0.5), mid-confidence (0.5 < PIP ≤ 0.8), and high-confidence (PIP > 0.8) (Additional file 1: Table S4). For AOA, functional fine-mapping led to a 30% increase in low-confidence variants, no change in mid-confidence variants, and a 50% increase in high-confidence variants. For COA, we observed 7, 120, and 33% increase in low-confidence, mid-confidence, and high-confidence variants, respectively.

Fine-mapping identifies groups of variants (i.e., credible sets) that contain at least one causal variant with 95% confidence. We discovered 21 and 67 credible sets among the 16 and 48 LD blocks that were fine-mapped for AOA and COA, respectively (Additional file 1: Table S5). About one-third of the LD blocks (5 AOA and 16 COA) had more than one credible set (Additional file 2: Fig. S3), suggesting multiple independent causal signals within these blocks. The number of SNPs within a credible set varied widely, ranging from 1 to 136, with median values of 10 for AOA and 4 for COA (Fig. 2C). Among all credible sets, only 16% were shared between AOA and COA (Fig. 2D).

Identifying cell types and CREs mediating genetic risk of asthma

The enrichment analysis above revealed genome-wide cell type heritability enrichment patterns but did not provide information on the relevant cell types for individual credible sets. To assess the evidence supporting a cell lineage for each credible set, we calculated the proportion of PIPs of variants that were within OCRs of each lineage (Additionalfile 2: Supplementary Methods). This proportion can be understood as the probability that the causal variant acts on the phenotype through a lineage. Using this strategy, we found that among AOA credible sets, 75% of the PIPs on average were attributed to lymphocytes (Fig. 3A). In contrast, the COA credible sets had higher proportions of PIPs attributed to epithelial cells (19%) and mesenchymal cells (17%), in addition to lymphocytes (46%) (Fig. 3B).

Fig. 3
figure 3

A Cellular contexts of adult-onset asthma credible sets based on OCRs. The proportion of the total PIP in each credible set is attributed to OCRs of each of the five cell lineages or to none. Each horizontal bar corresponds to a credible set, which is labelled in parentheses by the nearest gene to the SNP with the highest PIP; the length of bars of different colors shows the proportion of PIPs assigned to each lineage. Because not all SNPs in the credible sets overlapped with an OCR, some were not assigned to a cell lineage (gray bars). B Cellular context of childhood-onset asthma credible sets. C Adult-onset asthma and childhood-onset asthma candidate CRE ePIP distributions. D Distribution of the number of cell lineages underlying candidate CREs. E Adult-onset asthma and childhood-onset asthma ePIP distributions of candidate CREs (from panels C and D) that overlapped with bronchial epithelial cells MPRA+ and MPRA− sequences. The p-values were computed using Wilcoxon rank-sum test

Full size image

We next sought to nominate specific CREs that may be mediating the genetic effects of causal variants. We ranked candidate CREs by their ePIPs, which can be interpreted as the expected number of causal SNPs targeting the CRE (Methods). Using this approach, we identified 62 AOA and 169 COA candidate CREs with nonzero ePIPs (Fig. 3C; Additional file 1: Table S6). Notably, 64% of these candidate CREs were defined by OCRs of multiple cell lineages (Fig. 3D), indicating potential pleiotropic effect of asthma risk variants.

To experimentally assess regulatory activities of candidate CREs and variants, we performed massively parallel reporter assay (MPRA) in a human bronchial epithelial cell [61] (BEC) line, 16HBE14o-, to examine enhancer activities and allele-specific effects of 2034 SNPs chosen from AOA and COA GWAS loci (Additional file 2: Supplementary Methods). Among those, 438 SNPs were in sequences that tested positive for enhancer activity in a bronchial epithelial cell line, and 34 of those showed allele-specific effect on enhancer activity (Additional file 1: Table S7). We then used the MPRA results to assess the CREs selected by our ePIP strategy (Additional file 2: Supplementary Methods). The validated sequences in MPRA (MPRA+ set) had significantly higher ePIPs than sequences tested negative in MPRA (MPRA− set) for COA (p = 5.74 \(\times\) 10−5, Fig. 3E). In contrast, we did not observe a significant difference in ePIPs for AOA (p = 0.21, Fig. 3E). Taken together, the MPRA results suggest that the COA candidate CREs were distinctly enriched for enhancer activities in BECs.

Linking candidate CREs to target genes and prioritizing asthma risk genes

We next aimed to link candidate CREs to their target genes and identify likely causal genes for AOA and COA. Using functional genomics data that capture long-range regulation, including chromatin interactions and eQTLs from asthma-relevant tissues and cell types (Methods; Additional file 1: Table. S2), we nominated 107 putative target genes for 62 AOA candidate CREs and 253 putative target genes for 169 COA candidate CREs. Notably, 53 and 118 candidate CREs were linked to at least two different genes in AOA and COA, respectively (Additional file 2: Fig. S4), thus underscoring the challenge of precisely identifying target genes for specific CREs.

To address this challenge, we developed a scoring strategy to prioritize asthma effector genes by aggregating causal evidence from variants linked to the same gene (Methods). Using this strategy, we identified 76 AOA genes and 203 COA genes with nonzero gene score (Additional file 1: Table S8). Of these, 10 and 35 genes were considered as high-confidence candidate causal genes for AOA and COA, respectively (Fig. 4A, B). The most significant genes were often supported by multiple lines of evidence, with the greatest contributions coming from variants potentially targeting their nearest genes and variants linked to the genes by ABC models. All AOA candidate causal genes were targeted by SNPs from a single credible set, while 11 COA candidate causal genes were supported by SNPs from more than one credible set (Additional file 2: Fig. S5).

Fig. 4
figure 4

A Adult-onset asthma high-confidence candidate causal genes (N = 10), listed in descending order. The intensity of color shows the score contributed by each category. Nearest: variants whose nearest gene is the candidate gene; ABC: variants linked to the candidate gene by the ABC model; PCHi-C: variants linked to the candidate gene by PCHi-C; eQTL: variants linked to the candidate gene by eQTL; Exon: variants in the candidate gene’s exonic regions. The number in the parentheses indicates the number of variants linked to the corresponding gene. B Childhood-onset asthma high-confidence candidate causal genes (N = 35), listed in descending order. C Top Biological Processes GO terms enriched among AOA (top) and COA (bottom) high-confidence candidate causal genes, generated by WebGestalt’s weighted set cover algorithm

Full size image

To understand the biological functions of the prioritized genes, we assessed the enrichment of Biological Process Gene Ontology (GO) terms [62, 63] for candidate risk genes (Additional file 2: Supplementary Methods). A total of nine and 56 Biological Process GO terms were significantly enriched for AOA and COA candidate genes (FDR < 0.05), respectively, with eight of these shared between AOA and COA (Additional file 1: Table S9). The top enriched GO terms in both AOA and COA were associated with cytokine production and inflammatory response (Fig. 4C).

Functionally assaying candidate CREs and causal variants

Based on our integrative analyses, we selected six candidate CREs at both shared and specific loci for further functional validation using luciferase assays in 16HBE14o- cells (Additional file 2: Supplementary Methods): three were high-confidence candidate enhancers, two were candidate enhancers whose top SNP overlapped with an MPRA+ sequence, and one was in a promoter region (Additional file 1: Table S10). The results are described in the following sections.

Candidate enhancers at a COA-specific locus at chromosome 1q25.1

We identified one credible set containing two SNPs at a COA-specific locus at chromosome 1q25.1 (Fig. 5A; Additional file 1: Table S5, COA cs5). The most significant SNP in COA GWAS and the most likely causal SNP, rs11811856 (PIP = 0.95), mapped within an intron of TNFSF4 and overlapped with an OCR in epithelial, endothelial, mesenchymal, and myeloid cells in lung and lymphocytes in blood. The distance from the OCR midpoint to the TNFSF4 transcription start site (TSS) was 4966 bp. This OCR physically contacted the promoters of several genes based on PCHi-C of blood immune cells [56]: TNFSF18, CENPL, and DARS2. To complement the PCHi-C results, we checked ABC scores in relevant cell types (Additional file 1: Table S2). Interestingly, TNFSF4 is the most likely target of the OCR based on the ABC scores in immune cells, fibroblasts, and endothelial cells [59]. We also identified a distal candidate enhancer that looped to the promoter of TNFSF4 in PCHi-C of blood immune cells, BECs, and ASMCs (distance = 460,760 bp) harboring a high-PIP SNP rs78037977 (PIP = 0.92; Additional file 1: Table S5, COA cs4). This OCR overlapped with an MPRA+ sequence, supporting its regulatory activity. While this enhancer also contacted dozens of other gene promoters according to PCHi-C, TNFSF4 was a top gene by ABC scores in immune cells, epithelial cells, and fibroblasts. Taken together, these observations suggested that TNFSF4 is likely a COA risk gene, possibly with two independent causal signals targeting this gene.

Fig. 5
figure 5

A A childhood-onset asthma-specific locus at chromosome 1q25.1. From top to bottom, the first two tracks show the −log10 p-values from GWAS and PIPs from fine-mapping, respectively. Each point is a SNP, and assayed SNPs are denoted by larger squares. Different colors are used in the PIP track to represent different LD blocks. The two SNPs in candidate enhancers, rs78037977 and rs11811856, are highlighted in red. The next five tracks display chromatin accessibility from (sn)ATAC-seq of different cell lineages, with each dark blue vertical bar showing the location of an OCR. The next three tracks show chromatin interactions from PCHi-C of different cells, where the loops from the distal candidate enhancer to TNFSF4 promoter in all three cell types are highlighted in red. The last track shows the genes at the locus. BICs: blood immune cells, ASMCs: airway smooth muscle cells, BECs: bronchial epithelial cells. B A childhood-onset asthma-specific locus at chromosome 19q13.11. The PCHi-C loops from the candidate enhancer to the CEBPA promoter in blood immune cells and bronchial epithelial cells are highlighted in red

Full size image

Luciferase assay showed enhancer activity only for the rs11811856-C allele, the non-risk allele (Additional file 2: Fig. S6, left), for the candidate enhancer in the TNFSF4 intron, suggesting allele-specific effects of rs11811856-C vs. rs11811856-G (p = 0.02). In contrast, the distal candidate enhancer did not show regulatory effect in the luciferase assay (Additional file 2: Fig. S6, right), possibly due to cell type-specific regulation in a different cell type.

A candidate enhancer at a COA-specific locus at chromosome 19q13.11

One credible set (Additional file 1: Table S5, COA cs67) at a COA-specific locus at chromosome 19q13.11 (Fig. 5B) had two putative causal SNPs, rs118013485 (PIP = 0.46) and rs117710327 (PIP = 0.52). The two SNPs were in nearly perfect LD, with only two haplotypes (rs118013485-G/rs117710327-C and rs118013485-A /rs117710327-A) observed in the 1000 Genomes European populations [64, 65]. Both SNPs resided in the same OCRs in cell types from all five lineages. The ePIP of this candidate sequence was 0.99, suggesting that it mostly likely mediates the effect of the underlying causal variant(s). Although none of the eQTL datasets identified any target gene(s) for this candidate enhancer, PCHi-C in BECs [46] indicated that this candidate CRE only contacted the promoter of CEBPA (distance = 66,853 bp). In blood immune cells, this candidate enhancer looped to the promoters of CEBPG and CEBPA, both of which are CCAAT enhancer-binding proteins. Moreover, CEBPA had the highest ABC score in immune cells. We observed different levels of enhancer activities (rs118013485-A/rs117710327-A vs. rs118013485-G/rs117710327-C, p = 0.04) in luciferase assays between the two haplotypes (Additional file 2: Fig. S7), with decreased activity associated with the haplotype carrying the COA risk alleles.

Candidate enhancers at an AOA and COA shared locus at 5q31.1

At a shared locus at 5q31.1, we discovered two credible sets shared by AOA and COA (Additional file 1: Table S5, AOA cs7 and COA cs24, AOA cs8 and COA cs25), and one credible set that was specific to COA (Additional file 1: Table S5, COA cs23). Nominating the true causal SNPs from these credible sets was difficult, as none of the SNPs in the five credible sets had a PIP > 0.5. Therefore, we selected the AOA credible set containing the fewest SNPs for functional validation studies (AOA cs7; Fig. 6A). Among the four SNPs in cs7, rs1023518 (PIP = 0.35) and rs3857440 (PIP = 0.30) were captured by one candidate enhancer, while SNP rs3749833 (PIP = 0.27) resided in a separate candidate enhancer, both of which were represented by OCRs in all five blood and cell lineages. These three SNPs together accounted for 97% of the total PIP in the credible set. Furthermore, rs1023518 and rs3749833 overlapped with different sequences that each demonstrated enhancer activity in MPRA, whereas the sequence containing rs3857440 did not show enhancer activity in MPRA.

Fig. 6
figure 6

A A shared locus at chromosome 5q31.1. See Fig. 5 figure legend. The PCHi-C loop from the candidate enhancer to IRF1 promoter is highlighted in red in blood immune cells. B A shared locus at chromosome 12q13.2

Full size image

In luciferase assays, the sequence harboring rs1023518 and rs3857440 tested negative (Additional file 2: Fig. S8, left), but the sequence containing rs3749833 was validated as an allele-specific enhancer (Additional file 2: Fig. S8, right). Moreover, although another genome-wide significant SNP rs11748326 was located within the same luciferase construct as rs3749833, only rs3749833 showed significant allelic effects (Additional file 2: Fig. S8, right). These observations indicated that rs3749833 is likely a causal variant exerting its effect in BECs. We also looked at eQTLs and chromatin interaction data to determine the likely target genes of this enhancer. While the GTEx eQTL data nominated PDLIM4 as the putative target gene in skin and SLC22A5 as the putative target gene in whole blood, lung, and skin, this enhancer only interacted with the promoter of IRF1 in PCHi-C of blood immune cells. Additionally, IRF1 had the highest ABC score in immune cells. These findings suggest that one or more of these genes are regulated by this enhancer.

A candidate promoter at an AOA and COA shared locus at 12q13.2

We evaluated an OCR located 2 kb upstream of RPS26 that was characterized by ATAC-seq peaks in all 27 blood and lung cell types. The AOA and COA ePIPs of this OCR were 0.71 and 0.51, respectively, attributed to two SNPs in a pair of shared credible sets: rs705704 (AOA PIP = 0.41; COA PIP = 0.38) and rs705705 (AOA PIP = 0.29; COA PIP = 0.13) (Additional file 1: Table S5, AOA cs19 and COA cs50). We observed extensive chromatin interactions at this locus in BECs and ASMCs, potentially indicating a high level of regulatory activities in these cell types (Fig. 6B).

We performed luciferase assays for the two haplotypes comprised of the two SNPs in 1000 Genomes European populations (rs705704-G/rs705705-G and rs705704-A/rs705705-C). In line with our expectations for a promoter, we observed strong regulatory effect of both haplotypes on the luciferase activity, with fold change compared to the control ranging from ~ 50 times to > 300 times across experimental replicates (Additional file 2: Fig. S9). In addition, the asthma-associated rs705704-A/rs705705-C haplotype showed significantly lower luciferase activity than the rs705704-G/rs705705-G haplotype (rs705704-A/rs705705-C vs. rs705704-G/rs705705-G, p = 0.01), suggesting haplotype-specific regulation.

Discussion

Personalized risk prediction and treatment strategies for common, complex diseases are the aspirations of precision medicine [66]. The extraordinary heterogeneity of asthma makes these goals particularly challenging. Having a more refined understanding of the shared and distinct molecular genetic mechanisms underlying different asthma subtypes could lead to the discovery of new therapeutic targets as well as identifying individuals who would most likely benefit from therapies. Indeed, a recent study [67] estimated that drugs targeting genes with genetic support had 2.6 times greater probability of achieving clinical success than those targeting genes without genetic support. Our GWAS of AOA and COA [15] serves as a first step toward characterizing the underlying molecular mechanisms and nominating causal genes for these two important asthma subtypes.

In this study, we coupled computational and experimental methods to systematically uncover putative causal variants, candidate CREs, and effector genes for AOA and COA. Utilizing chromatin accessibility data in multiple cell types, we showed that both AOA and COA GWAS signals were concentrated in regulatory regions of immune cells. In contrast, the enrichment of GWAS loci in lung structural cells was a distinctive feature of COA, consistent with results in our previous GWAS [15] based on gene expression data. Leveraging functional fine-mapping, we uncovered a plethora of causal signals that further highlighted the distinct genetic bases underlying risk for AOA and COA. Our multi-level analyses revealed two broad patterns: first, most candidate CREs of asthma were in OCRs across multiple cell lineages, suggesting that most genetic variants of asthma have pleiotropic effects across cell types. Second, allelic heterogeneity was common. This was supported by both the presence of more than one credible set at many loci, and by the fact that many of the candidate genes were supported by two independent causal signals. Overall, our results underscore the complexity of the molecular mechanisms linking genetic variants of asthma pathogenesis.

Many of the genes prioritized by our scoring system had strong prior evidence supporting their roles in asthma. To highlight a few examples: (1) BACH2, the highest-ranked AOA risk gene, is a key regulator of T-cell and B-cell differentiation [68,69,70]. (2) SMAD3, the highest-scoring COA effector gene, is a crucial transcription factor in the transforming growth factor beta signaling pathway [71], a central mediator of airway remodeling in asthma [72]. (3) GATA3, the third highest-scoring AOA effector gene and a COA candidate risk gene, is a master regulator [73, 74] that modulates the expression and production of the type 2 cytokines that play a prominent role in both AOA and COA [75, 76]. (4) The prioritized causal genes of COA included several epithelial function-related genes. OVOL1, a transcription factor involved in epithelial cell differentiation [77], regulates the expression of FLG in normal human epidermal keratinocytes [78]. The OVOL1-FLG axis contributes to the pathogenesis of atopic dermatitis, an allergic condition that is often comorbid with asthma and has shared genetics with asthma [79], likely mediated by disrupting barrier function [80]. (5) The two prioritized toll-like receptor (TLRs) genes, TLR1 and TLR10, are both expressed in airway epithelium [81]. Together with other TLRs, they orchestrate response against microbes through the activation of TLR signaling pathways in epithelial cells [82, 83]. (6) Finally, HDAC7, the fifth highest scoring gene for AOA, resides at an AOA-specific GWAS locus. HDAC7 is a histone deacetylase involved in transcriptional regulation. This gene plays a key role in the function of regulatory T cells [84] and has been shown to potentially play a role in asthma and allergic diseases through epigenetic modifications [85].

The effectiveness of our analytical strategy was supported by the successful validation of selected candidate CREs by luciferase assays: four out of the six selected candidate CREs displayed regulatory activity and allelic effects in vitro. The genes likely regulated by these validated CREs were among the top genes prioritized by our gene scores and have known relationships to the pathogenesis of AOA and/or COA. For example, TNFSF4 encodes OX40 ligand (OX40L). By binding to the OX40 receptor, OX40L on antigen-presenting cells activates the OX40 costimulatory molecule on T cells. Importantly, the OX40-OX40L pathway has been shown to play a key role in the differentiation of Th2 cells and the activation of memory Th2 cells [86]. A recent Phase 2b clinical trial, STREAM-AD, showed that amlitelimab, a non-T cell depleting monoclonal antibody that blocks OX40L on antigen-presenting cells, exhibited sustained treatment effects on patients with atopic dermatitis [87]. A Phase 2 study examining the efficacy of amlitelimab on asthma patients is underway [87]. Another candidate effector gene, CEBPA, is a key regulator of lung epithelial cell development [88, 89]. Previous studies showed that CEBPA expression was absent in cultured ASMCs from subjects with asthma [90, 91]. The lack of CEBPA expression was associated with the failure of glucocorticoids to inhibit ASMC proliferation [90], suggesting that it could play a role in steroid-resistant asthma. Consistent with this observation, the enhancer potentially regulating CEBPA displayed lower activity in luciferase assays in the 16HBE14o- cell line with the risk allele for COA (rs117710327-C) compared to the non-risk rs117710327-A allele. Another putative causal gene, IRF1, encodes a transcription factor that regulates the activity of interferon and is involved in various aspects of adaptive and innate immune responses to pathogens [92,93,94]. IRF1 is upregulated by rhinovirus (RV) in epithelial cells [95]. RV-associated wheezing illness in early life is one of the most significant risk factors for COA [96, 97] and is strongly associated with asthma exacerbations and hospitalizations throughout life [98]. In addition, IRF1 has been identified as a key driver of lipopolysaccharide (LPS)-induced interferon responses at birth [99]. Based on these and other data [100,101,102], impaired immune responses to microbial infections is thought to be a key mechanism underlying asthma onset and severity [103]. These findings are also consistent with our observation of reduced enhancer activity associated with the asthma risk allele (rs3749833-C).

We recognize several limitations of our study. First, our analysis primarily relied on open chromatin as functional annotations, missing other mechanisms of gene regulation (e.g., alternative splicing). Second, in regions with extensive LD, many SNPs will receive small PIPs, often making it impossible to distinguish likely causal SNP(s). Both possibilities may explain why the most significant COA locus at chromosome 17q12-q21 was not among those prioritized by our pipeline. This locus is characterized by extensive LD over ~ 150 kb in populations of European ancestry and the likely causal SNPs affect splicing (rs11078928) of GSDMB and/or encode a missense mutation (rs2305480) in GSDMB [104]. Third, the epigenomic and gene expression data used in our studies were in unstimulated cells. It is possible that cells stimulated with asthma-promoting cytokines or viruses, as examples, will induce context-specific CREs that were missed by the current study. Fourth, we applied a heuristic method to score and rank the genes and different kinds of evidence linking a CRE with a gene were weighted equally. A better approach may be to assign weights differently, putting more emphasis on datasets more likely to support functional relationships. Fifth, to maximize detection power, we used summary statistics from AOA and COA GWAS for fine-mapping. We therefore were not able to include individual-level comorbidities as covariates in our analyses, nor were we able to stratify our analyses by clinical features. As a result, we may have missed signals that are specific to severe asthma [105] or asthma associated with other disorders [106, 107]. Finally, our study included only UKB individuals who self-identified as White British and, therefore, our fine-mapping results reflect the specific LD and allele frequency patterns of this population. A recent study [108] has shown that fine-mapping can greatly benefit from including individuals of diverse genetic backgrounds, which can uncover putative causal variants that are more frequent in non-European populations. Additionally, the distinct haplotype structures among different populations can help disentangle SNPs that are in high LD in one population but not in others. Taken together, increasing the genetic diversity of future fine-mapping studies, along with rigorous analytical approaches and more precise phenotype definitions, is critical for expanding our understanding of the genetic architecture of complex traits across various populations.

Conclusions

By combining experimental and computational approaches, our study provides the most thorough follow-up of AOA and COA GWAS discoveries to date. We identified numerous risk variants, regulatory elements, and candidate genes and uncovered key insights into the genetic architecture of AOA and COA. Our datasets provide a valuable resource for future functional studies to understand the biological mechanisms underlying the genetics of asthma with onset in both childhood and later in life.

Data availability

This study uses genotype and phenotype data from the UK Biobank under application number 44300. Access to UK Biobank resource is available with application at http://www.ukbiobank.ac.uk. Summary statistics of AOA and COA GWAS performed with UK Biobank version 3 genotypes can be accessed through Zenodo: https://doiorg.publicaciones.saludcastillayleon.es/10.5281/zenodo.14842007. The sequencing data generated in this study were deposited in EMBL-EBI's Array Express database (https://www.ebi.ac.uk/biostudies/arrayexpress) under accession numbers E-MTAB-14267 (ASMCs ATAC-seq), E-MTAB-14273 (BECs MPRA), and E-MTAB-14295 (ASMCs PCHi-C). The snATAC-seq data from 18 lung cell types were downloaded from https://www.lungepigenome.org. The ATAC-seq data from seven blood cell types were downloaded from https://github.com/caleblareau/singlecell_bloodtraits/tree/master/data/bulk/ATAC/narrowpeaks. The BECs ATAC-seq data are available as supplementary material from the original publication: https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s42003-020-01411-4. The PCHi-C datasets are available as supplementary material from the original publications: https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.cell.2016.09.037 (blood immune cells), https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s42003-020-01411-4 (BECs), and https://doiorg.publicaciones.saludcastillayleon.es/10.1038/s42003-020-01411-4 (bulk lung). The ABC models are available on Engreitz Lab’s website: https://www.engreitzlab.org/resources/. GTEx V8 eQTLs were downloaded from https://www.gtexportal.org/home/downloads/adult-gtex/qtl. DICE eQTLs were downloaded from https://dice-database.org/downloads#eqtl_download. OneK1K single-cell eQTLs in peripheral blood mononuclear cells were downloaded from https://onek1k.org/. Single-cell eQTLs in lung were downloaded from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE227136. All analysis code is available on GitHub: https://github.com/ez-xyz/asthma_finemapping.

Abbreviations

GWAS:

Genome-wide association study

AOA:

Adult-onset asthma

COA:

Childhood-onset asthma

LD:

Linkage disequilibrium

CRE:

Cis-regulatory element

eQTL:

Expression quantitative trait locus

TWAS:

Transcriptome-wide association study

sQTL:

Splicing quantitative trait locus

caQTL:

Chromatin accessibility quantitative trait locus

meQTL:

DNA methylation quantitative trait locus

UKB:

UK Biobank

COPD:

Chronic obstructive pulmonary disorder

SNP:

Single-nucleotide polymorphism

ASMC:

Airway smooth muscle cell

S-LDSC:

Stratified LD score regression

OCR:

Open chromatin region

SuSiE:

Sum of single effects

PIP:

Posterior inclusion probability

ePIP:

Element PIP

PCHi-C:

Promoter Capture Hi-C

ABC:

Activity-by-contact

MPRA:

Massively parallel reporter assay

BEC:

Bronchial epithelial cell

GO:

Gene Ontology

TSS:

Transcription start site

TLR:

Toll-like receptor

OX40L:

OX40 ligand

RV:

Rhinovirus

LPS:

Lipopolysaccharide

snATAC-seq:

Single-nucleus ATAC-seq

CPM:

Counts per million

References

  1. Ober C. Asthma genetics in the post-GWAS Era. Ann ATS. 2016;13:S85-90.

    Article  Google Scholar 

  2. Vicente CT, Revez JA, Ferreira MAR. Lessons from ten years of genome-wide association studies of asthma. Clin Transl Immunol. 2017;6:e165.

    Article  Google Scholar 

  3. Willis-Owen SAG, Cookson WOC, Moffatt MF. The genetics and genomics of asthma. Annu Rev Genom Hum Genet. 2018;19:223–46.

    Article  CAS  Google Scholar 

  4. Tsuo K, Zhou W, Wang Y, Kanai M, Namba S, Gupta R, et al. Multi-ancestry meta-analysis of asthma identifies novel associations and highlights the value of increased power and diversity. Cell Genomics. 2022;2:100212.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  5. Fitzpatrick AM, Teague WG, Meyers DA, Peters SP, Li X, Li H, et al. Heterogeneity of severe asthma in childhood: confirmation by cluster analysis of children in the National Institutes of Health/National Heart, Lung, and Blood Institute Severe Asthma Research Program. J Allergy Clin Immunol. 2011;127:382-389.e13.

    Article  PubMed  Google Scholar 

  6. Wenzel SE. Asthma phenotypes: the evolution from clinical to molecular approaches. Nat Med. 2012;18:716–25.

    Article  PubMed  CAS  Google Scholar 

  7. Wesolowska-Andersen A, Seibold MA. Airway molecular endotypes of asthma: dissecting the heterogeneity. Curr Opin Allergy Clin Immunol. 2015;15. Available from: https://journals.lww.com/co-allergy/fulltext/2015/04000/airway_molecular_endotypes_of_asthma__dissecting.10.aspx.

  8. Carr TF, Bleecker E. Asthma heterogeneity and severity. World Allergy Organization J. 2016;9:41. https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s40413-016-0131-2.

    Article  CAS  Google Scholar 

  9. Gelfand EW, Schedel M. Molecular Endotypes Contribute to the Heterogeneity of Asthma. Immunol Allergy Clin North Am. 2018;38:655–65.

    Article  PubMed  Google Scholar 

  10. Kaur R, Chupp G. Phenotypes and endotypes of adult asthma: moving toward precision medicine. J Allergy Clin Immunol. 2019;144:1–12.

    Article  PubMed  Google Scholar 

  11. Kuruvilla ME, Lee FE-H, Lee GB. Understanding asthma phenotypes, endotypes, and mechanisms of disease. Clin Rev Allergy Immunol. 2019;56:219–33.

    Article  PubMed  PubMed Central  Google Scholar 

  12. McDowell PJ, Heaney LG. Different endotypes and phenotypes drive the heterogeneity in severe asthma. Allergy. 2020;75:302–10.

    Article  PubMed  Google Scholar 

  13. Foppiano F, Schaub B. Childhood asthma phenotypes and endotypes: a glance into the mosaic. Mol Cell Pediatr. 2023;10:9.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Trivedi M, Denton E. Asthma in children and adults—what are the differences and what can they tell us about asthma? Front Pediatr. 2019;7. Available from: https://www.frontiersin.org/articles/10.3389/fped.2019.00256.

  15. Pividori M, Schoettler N, Nicolae DL, Ober C, Im HK. Shared and distinct genetic risk factors for childhood-onset and adult-onset asthma: genome-wide and transcriptome-wide studies. Lancet Respir Med. 2019;7:509–22.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Ferreira MAR, Mathur R, Vonk JM, Szwajda A, Brumpton B, Granell R, et al. Genetic architectures of childhood- and adult-onset asthma are partly distinct. American J Human Genet. 2019;104:665–84.

    Article  CAS  Google Scholar 

  17. Gallagher MD, Chen-Plotkin AS. The post-GWAS Era: from association to function. American J Human Genet. 2018;102:717–30.

    Article  CAS  Google Scholar 

  18. Tam V, Patel N, Turcotte M, Bossé Y, Paré G, Meyre D. Benefits and limitations of genome-wide association studies. Nat Rev Genet. 2019;20:467–84.

    Article  PubMed  CAS  Google Scholar 

  19. Schoenfelder S, Fraser P. Long-range enhancer–promoter contacts in gene expression control. Nat Rev Genet. 2019;20:437–55.

    Article  PubMed  CAS  Google Scholar 

  20. Claringbould A, Zaugg JB. Enhancers in disease: molecular basis and emerging treatment strategies. Trends Mol Med. 2021;27:1060–73.

    Article  PubMed  CAS  Google Scholar 

  21. Schmiedel BJ, Gonzalez-Colin C, Fajardo V, Rocha J, Madrigal A, Ramírez-Suástegui C, et al. Single-cell eQTL analysis of activated T cell subsets reveals activation and cell type–dependent effects of disease-risk variants. Science Immunol. 2022;7:eabm2508.

    Article  CAS  Google Scholar 

  22. Soskic B, Cano-Gamez E, Smyth DJ, Ambridge K, Ke Z, Matte JC, et al. Immune disease risk variants regulate gene expression dynamics during CD4+ T cell activation. Nat Genet. 2022;54:817–26.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  23. Nathan A, Asgari S, Ishigaki K, Valencia C, Amariuta T, Luo Y, et al. Single-cell eQTL models reveal dynamic T cell state dependence of disease loci. Nature. 2022;606:120–8.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  24. Wei J, Resztak JA, Ranjbaran A, Alazizi A, Mair-Meijers HE, Slatcher RB, et al. Functional characterization of eQTLs and asthma risk loci with scATAC-seq across immune cell types and contexts. American J Human Genet. 2025;112:301–17.

    Article  CAS  Google Scholar 

  25. Natri HM, Del Azodi CB, Peter L, Taylor CJ, Chugh S, Kendle R, et al. Cell-type-specific and disease-associated expression quantitative trait loci in the human lung. Nat Genet. 2024;56:595–604.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  26. Jiang Y, Gruzieva O, Wang T, Forno E, Boutaoui N, Sun T, et al. Transcriptomics of atopy and atopic asthma in white blood cells from children and adolescents. Eur Respir J. 2019;53:1900102.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  27. Forno E, Zhang R, Jiang Y, Yan Q, Han YY, Acosta-Perez E, et al. Transcriptome-wide association study (TWAS) of nasal respiratory epithelium and childhood asthma. Eur Respir J. 2019;54:OA4943.

    Google Scholar 

  28. Valette K, Li Z, Bon-Baret V, Chignon A, Bérubé J-C, Eslami A, et al. Prioritization of candidate causal genes for asthma in susceptibility loci derived from UK Biobank. Commun Biol. 2021;4:700.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  29. Sajuthi SP, Everman JL, Jackson ND, Saef B, Rios CL, Moore CM, et al. Nasal airway transcriptome-wide association study of asthma reveals genetically driven mucus pathobiology. Nat Commun. 2022;13:1632.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  30. Yao DW, O’Connor LJ, Price AL, Gusev A. Quantifying genetic effects on disease mediated by assayed gene expression levels. Nat Genet. 2020;52:626–33.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  31. Mostafavi H, Spence JP, Naqvi S, Pritchard JK. Systematic differences in discovery of genetic effects on gene expression and complex traits. Nat Genet. 2023;55:1866–75.

    Article  PubMed  CAS  Google Scholar 

  32. Mu Z, Wei W, Fair B, Miao J, Zhu P, Li YI. The impact of cell type and context-dependent regulatory variants on human immune traits. Genome Biol. 2021;22:122.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  33. Bossini-Castillo L, Glinos DA, Kunowska N, Golda G, Lamikanra AA, Spitzer M, et al. Immune disease variants modulate gene expression in regulatory CD4+ T cells. Cell Genomics. 2022;2:100117.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  34. Benaglio P, Newsome J, Han JY, Chiou J, Aylward A, Corban S, et al. Mapping genetic effects on cell type-specific chromatin accessibility and annotating complex immune trait variants using single nucleus ATAC-seq in peripheral blood. PLoS Genet. 2023;19:e1010759.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  35. Soliai MM, Kato A, Helling BA, Stanhope CT, Norton JE, Naughton KA, et al. Multi-omics colocalization with genome-wide association studies reveals a context-specific genetic mechanism at a childhood onset asthma risk locus. Genome Med. 2021;13:157.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  36. Wainberg M, Sinnott-Armstrong N, Mancuso N, Barbeira AN, Knowles DA, Golan D, et al. Opportunities and challenges for transcriptome-wide association studies. Nat Genet. 2019;51:592–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  37. Hukku A, Pividori M, Luca F, Pique-Regi R, Im HK, Wen X. Probabilistic colocalization of genetic variants from complex and molecular traits: promise and limitations. American J Human Genet. 2021;108:25–35.

    Article  CAS  Google Scholar 

  38. Zhao S, Crouse W, Qian S, Luo K, Stephens M, He X. Adjusting for genetic confounders in transcriptome-wide association studies improves discovery of risk genes of complex traits. Nat Genet. 2024;56:336–47.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  39. Zhong X, Mitchell R, Billstrand C, Thompson EE, Sakabe NJ, Aneas I, et al. Integration of functional genomics and statistical fine-mapping systematically characterizes adult-onset and childhood-onset asthma genetic associations. GitHub repository. 2025. https://github.com/ez-xyz/asthma_finemapping.

  40. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK Biobank resource with deep phenotyping and genomic data. Nature. 2018;562:203–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  41. Zhong X, Mitchell R, Billstrand C, Thompson EE, Sakabe NJ, Aneas I, et al. Integration of functional genomics and statistical fine-mapping systematically characterizes adult-onset and childhood-onset asthma genetic associations. AOA and COA GWAS summary statistics. Zenodo. 2025. https://doiorg.publicaciones.saludcastillayleon.es/10.5281/zenodo.14842007.

    Article  Google Scholar 

  42. Hail Team. Hail 0.2. Available from: https://github.com/hail-is/hail.

  43. Zhong X, Mitchell R, Billstrand C, Thompson EE, Sakabe NJ, Aneas I, et al. Integration of functional genomics and statistical fine-mapping systematically characterizes adult-onset and childhood-onset asthma genetic associations. ASMCs ATAC-seq data. EMBL-EBI ArrayExpress Database. Accession number: E-MTAB-14267. 2025.

  44. Ulirsch JC, Lareau CA, Bao EL, Ludwig LS, Guo MH, Benner C, et al. Interrogation of human hematopoiesis at single-cell and single-variant resolution. Nat Genet. 2019;51:683–93.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  45. Wang A, Chiou J, Poirion OB, Buchanan J, Valdez MJ, Verheyden JM, et al. Single-cell multiomic profiling of human lungs reveals cell-type-specific and age-dynamic control of SARS-CoV2 host genes. Morrisey EE, De Langhe S, editors. eLife. 2020;9:e62522.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  46. Helling BA, Sobreira DR, Hansen GT, Sakabe NJ, Luo K, Billstrand C, et al. Altered transcriptional and chromatin responses to rhinovirus in bronchial epithelial cells from adults with asthma. Communications Biology. 2020;3:678.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  47. Finucane HK, Bulik-Sullivan B, Gusev A, Trynka G, Reshef Y, Loh P-R, et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat Genet. 2015;47:1228–35.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  48. Gazal S, Finucane HK, Furlotte NA, Loh P-R, Palamara PF, Liu X, et al. Linkage disequilibrium–dependent architecture of human complex traits shows action of negative selection. Nat Genet. 2017;49:1421–7.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  49. Wen X. Molecular QTL discovery incorporating genomic annotations using Bayesian false discovery rate control. Ann Appl Statistics. 2016;10:1619–38.

    Article  Google Scholar 

  50. Wang G, Sarkar A, Carbonetto P, Stephens M. A simple new approach to variable selection in regression, with application to genetic fine mapping. J R Stat Soc Ser B Stat Methodol. 2020;82:1273–300.

    Article  Google Scholar 

  51. Zou Y, Carbonetto P, Wang G, Stephens M. Fine-mapping from summary data with the “Sum of Single Effects” model. PLoS Genet. 2022;18:e1010299.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  52. Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26:841–2.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  53. The GTEx Consortium, Aguet F, Anand S, Ardlie KG, Gabriel S, Getz GA, et al. The GTEx consortium atlas of genetic regulatory effects across human tissues. Science. 2020;369:1318–30.

    Article  PubMed Central  Google Scholar 

  54. Schmiedel BJ, Singh D, Madrigal A, Valdovino-Gonzalez AG, White BM, Zapardiel-Gonzalo J, et al. Impact of genetic polymorphisms on human immune cell gene expression. Cell. 2018;175:1701-1715.e16.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  55. Yazar S, Alquicira-Hernandez J, Wing K, Senabouth A, Gordon MG, Andersen S, et al. Single-cell eQTL mapping identifies cell type–specific genetic control of autoimmune disease. Science. 2022;376:eabf3041.

    Article  PubMed  CAS  Google Scholar 

  56. Javierre BM, Burren OS, Wilder SP, Kreuzhuber R, Hill SM, Sewitz S, et al. Lineage-specific genome architecture links enhancers and non-coding disease variants to target gene promoters. Cell. 2016;167:1369-1384.e19.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  57. Jung I, Schmitt A, Diao Y, Lee AJ, Liu T, Yang D, et al. A compendium of promoter-centered long-range chromatin interactions in the human genome. Nat Genet. 2019;51:1442–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  58. Zhong X, Mitchell R, Billstrand C, Thompson EE, Sakabe NJ, Aneas I, et al. Integration of functional genomics and statistical fine-mapping systematically characterizes adult-onset and childhood-onset asthma genetic associations. ASMCs PCHi-C data. EMBL-EBI ArrayExpress Database. Accession number: E-MTAB-14295. 2025.

  59. Nasser J, Bergman DT, Fulco CP, Guckelberger P, Doughty BR, Patwardhan TA, et al. Genome-wide enhancer maps link risk variants to disease genes. Nature. 2021;593:238–43.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  60. Selewa A, Luo K, Wasney M, Smith L, Sun X, Tang C, et al. Single-cell genomics improves the discovery of risk variants and genes of atrial fibrillation. Nat Commun. 2023;14:4999.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  61. Zhong X, Mitchell R, Billstrand C, Thompson EE, Sakabe NJ, Aneas I, et al. Integration of functional genomics and statistical fine-mapping systematically characterizes adult-onset and childhood-onset asthma genetic associations. BECs MPRA data. EMBL-EBI ArrayExpress Database. Accession number: E-MTAB-14273. 2025.

  62. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene Ontology: tool for the unification of biology. Nat Genet. 2000;25:25–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  63. The Gene Ontology Consortium, Aleksander SA, Balhoff J, Carbon S, Cherry JM, Drabkin HJ, et al. The gene ontology knowledgebase in 2023. Genetics. 2023;224:iyad031.

    Article  Google Scholar 

  64. Machiela MJ, Chanock SJ. LDlink: a web-based application for exploring population-specific haplotype structure and linking correlated alleles of possible functional variants. Bioinformatics. 2015;31:3555–7.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  65. Auton A, Abecasis GR, Altshuler DM, Durbin RM, Abecasis GR, Bentley DR, et al. A global reference for human genetic variation. Nature. 2015;526:68–74.

    Article  PubMed  Google Scholar 

  66. Denny JC, Collins FS. Precision medicine in 2030—seven ways to transform healthcare. Cell. 2021;184:1415–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  67. Minikel EV, Painter JL, Dong CC, Nelson MR. Refining the impact of genetic evidence on clinical success. Nature. 2024;629:624–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  68. Roychoudhuri R, Hirahara K, Mousavi K, Clever D, Klebanoff CA, Bonelli M, et al. BACH2 represses effector programs to stabilize Treg-mediated immune homeostasis. Nature. 2013;498:506–10.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  69. Roychoudhuri R, Clever D, Li P, Wakabayashi Y, Quinn KM, Klebanoff CA, et al. BACH2 regulates CD8+ T cell differentiation by controlling access of AP-1 factors to enhancers. Nat Immunol. 2016;17:851–60.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  70. Laidlaw BJ, Cyster JG. Transcriptional regulation of memory B cell differentiation. Nat Rev Immunol. 2021;21:209–20.

    Article  PubMed  CAS  Google Scholar 

  71. Mullen AC, Orlando DA, Newman JJ, Lovén J, Kumar RM, Bilodeau S, et al. Master transcription factors determine cell-type-specific responses to TGF-β signaling. Cell. 2011;147:565–76.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  72. Halwani R, Al-Muhsen S, Al-Jahdali H, Hamid Q. Role of transforming growth factor–β in airway remodeling in asthma. Am J Respir Cell Mol Biol. 2011;44:127–33.

    Article  PubMed  CAS  Google Scholar 

  73. Tindemans I, Serafini N, Di Santo JP, Hendriks RW. GATA-3 function in innate and adaptive immunity. Immunity. 2014;41:191–206.

    Article  PubMed  CAS  Google Scholar 

  74. Fang D, Zhu J. Dynamic balance between master transcription factors determines the fates and functions of CD4 T cell and innate lymphoid cell subsets. J Exp Med. 2017;214:1861–76.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  75. Lambrecht BN, Hammad H, Fahy JV. The cytokines of asthma. Immunity. 2019;50:975–91.

    Article  PubMed  CAS  Google Scholar 

  76. Harker JA, Lloyd CM. T helper 2 cells in asthma. J Exp Med. 2023;220:e20221094.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  77. Saxena K, Srikrishnan S, Celia-Terrassa T, Jolly MK. OVOL1/2: drivers of epithelial differentiation in development, disease, and reprogramming. Cells Tissues Organs. 2020;211:183–92.

    Article  PubMed  Google Scholar 

  78. Tsuji G, Hashimoto-Hachiya A, Kiyomatsu-Oda M, Takemura M, Ohno F, Ito T, et al. Aryl hydrocarbon receptor activation restores filaggrin expression via OVOL1 in atopic dermatitis. Cell Death Dis. 2017;8:e2931–e2931.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  79. Ferreira MA, Vonk JM, Baurecht H, Marenholz I, Tian C, Hoffman JD, et al. Shared genetic origin of asthma, hay fever and eczema elucidates allergic disease biology. Nat Genet. 2017;49:1752–7.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  80. Furue K, Ito T, Tsuji G, Ulzii D, Vu YH, Kido-Nakahara M, et al. The IL-13–OVOL1–FLG axis in atopic dermatitis. Immunology. 2019;158:281–6.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  81. Renkonen J, Toppila-Salmi S, Joenväärä S, Mattila P, Parviainen V, Hagström J, et al. Expression of Toll-like receptors in nasal epithelium in allergic rhinitis. APMIS. 2015;123:716–25.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  82. McClure R, Massari P. TLR-dependent human mucosal epithelial cell responses to microbial pathogens. Front Immunol. 2014;5. Available from: https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2014.00386.

  83. Wenger M, Grosse-Kathoefer S, Kraiem A, Pelamatti E, Nunes N, Pointner L, et al. When the allergy alarm bells toll: The role of Toll-like receptors in allergic diseases and treatment. Front Mol Biosci. 2023;10. Available from: https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2023.1204025.

  84. Axisa P-P, Yoshida TM, Lucca LE, Kasler HG, Lincoln MR, Pham GH, et al. A multiple sclerosis–protective coding variant reveals an essential role for HDAC7 in regulatory T cells. Sci Transl Med. 2022;14:eabl3651.

  85. Morin A, Thompson EE, Helling BA, Shorey-Kendrick LE, Faber P, Gebretsadik T, et al. A functional genomics pipeline to identify high-value asthma and allergy CpGs in the human methylome. J Allergy Clin Immunol. 2023;151:1609–21.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  86. Croft M, So T, Duan W, Soroosh P. The significance of OX40 and OX40L to T-cell biology and immune disease. Immunol Rev. 2009;229:173–91.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  87. Sanofi. Press Release: New Phase 2b results for amlitelimab support potential for best-in-class maintenance of response in atopic dermatitis. 2024. Available from: https://www.sanofi.com/en/media-room/press-releases/2024/2024-03-11-06-00-00-2843456.

  88. Lourenço AR, Roukens MG, Seinstra D, Frederiks CL, Pals CE, Vervoort SJ, et al. C/EBPɑ is crucial determinant of epithelial maintenance by preventing epithelial-to-mesenchymal transition. Nat Commun. 2020;11:785.

    Article  PubMed  PubMed Central  Google Scholar 

  89. Hassan D, Chen J. CEBPA restricts alveolar type 2 cell plasticity during development and injury-repair. Nat Commun. 2024;15:4148.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  90. Roth M, Johnson PRA, Borger P, Bihl MP, Rüdiger JJ, King GG, et al. Dysfunctional Interaction of C/EBPα and the Glucocorticoid Receptor in Asthmatic Bronchial Smooth-Muscle Cells. N Engl J Med. 2004;351:560–74.

    Article  PubMed  CAS  Google Scholar 

  91. Borger P, Matsumoto H, Boustany S, Gencay MMC, Burgess JK, King GG, et al. Disease-specific expression and regulation of CCAAT/enhancer-binding proteins in asthma and chronic obstructive pulmonary disease. J Allergy Clin Immunol. 2007;119:98–105.

    Article  PubMed  CAS  Google Scholar 

  92. Honda K, Takaoka A, Taniguchi T. Type I inteferon gene induction by the interferon regulatory factor family of transcription factors. Immunity. 2006;25:349–60.

    Article  PubMed  CAS  Google Scholar 

  93. Jefferies CA. Regulating IRFs in IFN Driven Disease. Frontiers in Immunology. 2019;10. Available from: https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2019.00325.

  94. Feng H, Zhang Y-B, Gui J-F, Lemon SM, Yamane D. Interferon regulatory factor 1 (IRF1) and anti-pathogen innate immune responses. PLoS Pathog. 2021;17:e1009220.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  95. Djeddi S, Fernandez-Salinas D, Huang GX, Aguiar VRC, Mohanty C, Kendziorski C, et al. Rhinovirus infection of airway epithelial cells uncovers the non-ciliated subset as a likely driver of genetic risk to childhood-onset asthma. Cell Genomics. 2024;4. https://doiorg.publicaciones.saludcastillayleon.es/10.1016/j.xgen.2024.100636.

  96. Jackson DJ, Gangnon RE, Evans MD, Roberg KA, Anderson EL, Pappas TE, et al. Wheezing Rhinovirus Illnesses in Early Life Predict Asthma Development in High-Risk Children. Am J Respir Crit Care Med. 2008;178:667–72.

    Article  PubMed  PubMed Central  Google Scholar 

  97. Çalışkan M, Bochkov YA, Kreiner-Møller E, Bønnelykke K, Stein MM, Du G, et al. Rhinovirus Wheezing Illness and Genetic Risk of Childhood-Onset Asthma. N Engl J Med. 2013;368:1398–407.

    Article  PubMed  PubMed Central  Google Scholar 

  98. Holgate ST, Wenzel S, Postma DS, Weiss ST, Renz H, Sly PD. Asthma nature reviews. Dis Prim. 2015;1:15025.

    Google Scholar 

  99. Read JF, Serralha M, Armitage JD, Iqbal MM, Cruickshank MN, Saxena A, et al. Single cell transcriptomics reveals cell type specific features of developmentally regulated responses to lipopolysaccharide between birth and 5 years. Front Immunol. 2023;14. Available from: https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2023.1275937.

  100. Wark PAB, Johnston SL, Bucchieri F, Powell R, Puddicombe S, Laza-Stanca V, et al. Asthmatic bronchial epithelial cells have a deficient innate immune response to infection with rhinovirus. J Exp Med. 2005;201:937–47.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  101. Contoli M, Message SD, Laza-Stanca V, Edwards MR, Wark PAB, Bartlett NW, et al. Role of deficient type III interferon-λ production in asthma exacerbations. Nat Med. 2006;12:1023–6.

    Article  PubMed  CAS  Google Scholar 

  102. Jackson DJ, Makrinioti H, Rana BMJ, Shamji BWH, Trujillo-Torralbo M-B, Footitt J, et al. IL-33–dependent type 2 inflammation during rhinovirus-induced asthma exacerbations in vivo. Am J Respir Crit Care Med. 2014;190:1373–82.

    Article  PubMed  PubMed Central  Google Scholar 

  103. Jackson DJ, Gern JE. Rhinovirus infections and their roles in asthma: etiology and exacerbations. J Allergy Clin Immunol iIn Pract. 2022;10:673–81.

    Article  CAS  Google Scholar 

  104. Ober C, McKennan CG, Magnaye KM, Altman MC, Washington C 3rd, Stanhope C, et al. Expression quantitative trait locus fine mapping of the 17q12–21 asthma locus in African American children: a genetic association and gene expression study. Lancet Respir Med. 2020;8:482–92.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  105. Shrine N, Portelli MA, John C, Soler Artigas M, Bennett N, Hall R, et al. Moderate-to-severe asthma in individuals of European ancestry: a genome-wide association study. Lancet Respir Med. 2019;7:20–34.

    Article  PubMed  PubMed Central  Google Scholar 

  106. Cazzola M, Rogliani P, Ora J, Calzetta L, Materna MG. Asthma and comorbidities: recent advances. Pol Arch Intern Med. 2022;132(4):16250.

    Article  PubMed  Google Scholar 

  107. Jia G, Zhong X, Im HK, Schoettler N, Pividori M, Hogarth DK, et al. Discerning asthma endotypes through comorbidity mapping. Nat Commun. 2022;13:6712.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  108. Verma A, Huffman JE, Rodriguez A, Conery M, Liu M, Ho YL, et al. Diversity and scale: Genetic architecture of 2068 traits in the VA Million Veteran Program. Science. 2024;385:eadj1182.

Download references

Acknowledgements

The authors would like to acknowledge William Wentworth-Sheilds for assistance with data curation, Dr. Kevin Luo for assistance with the mapgen R package, Dr. Milton Pividori for assistance with creating the Miami plot, and Dr. Caleb Lareau for helpful discussions about the ATAC-seq data from blood cells. The authors gratefully acknowledge the Gift of Hope Organ and Tissue Donor Network and their donors and their families for providing tissues used for this study. The authors further acknowledge the anonymous reviewers for their helpful comments.

Funding

This study was supported by U19 AI62310 (C.O., M.A.N., A.I.S.), R01 MH110531, R01 MH116281, R01 HG010773, R01 HL163523 (X.H.), and UG3/UH3 OD023282 and UM1 AI160040 (C.O.). I.M.S. was supported by T32 HL007605. N.S. was supported by K08 HL153955.

Author information

Author notes

  1. Marcelo A. Nóbrega, Xin He and Carole Ober jointly supervised the project.

Authors and Affiliations

  1. Department of Human Genetics, University of Chicago, Chicago, IL, 60637, USA

    Xiaoyuan Zhong, Robert Mitchell, Christine Billstrand, Emma E. Thompson, Noboru J. Sakabe, Ivy Aneas, Isabella M. Salamone, Jing Gu, Marcelo A. Nóbrega, Xin He & Carole Ober

  2. Division of Pulmonary and Critical Care Medicine, Department of Medicine, University of Virginia, Charlottesville, VA, 22908, USA

    Anne I. Sperling

  3. Section of Pulmonary and Critical Care Medicine, Department of Medicine, University of Chicago, Chicago, IL, 60637, USA

    Nathan Schoettler

Authors
  1. Xiaoyuan Zhong
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Robert Mitchell
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Christine Billstrand
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Emma E. Thompson
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Noboru J. Sakabe
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Ivy Aneas
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Isabella M. Salamone
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Jing Gu
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Anne I. Sperling
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Nathan Schoettler
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Marcelo A. Nóbrega
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. Xin He
    View author publications

    You can also search for this author inPubMed Google Scholar

  13. Carole Ober
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

X.Z. harmonized the datasets, performed the computational analyses, interpreted the results, and prepared the manuscript; R.M and E.E.T. performed ATAC-seq and PCHi-C in ASMCs; C.B. performed luciferase assays in BECs; I.A. performed MPRA in BECs; N.J.S. processed the ATAC-seq, PCHi-C, and MPRA data; J.G. assisted with data harmonization and computational analyses; I.M.S., A.I.S., and N.S. assisted with results interpretation; M.A.N. supervised the functional genomics experiments; X.H. and C.O. supervised the computational analyses; M.A.N., X.H., and C.O. designed the study and interpreted the results. All authors contributed to writing the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Xiaoyuan Zhong, Marcelo A. Nóbrega, Xin He or Carole Ober.

Ethics declarations

Ethics approval and consent to participate

Institutional Review Board (IRB) approval was waived because this research only used publicly available data. The biological samples were from deceased donors and therefore are not considered human subjects by the University of Chicago IRB.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhong, X., Mitchell, R., Billstrand, C. et al. Integration of functional genomics and statistical fine-mapping systematically characterizes adult-onset and childhood-onset asthma genetic associations. Genome Med 17, 35 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13073-025-01459-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13073-025-01459-z

Keywords

Genome Medicine

ISSN: 1756-994X

Contact us