Skip to main content
NCBI home page

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

bioRxiv logoLink to bioRxiv
[Preprint]. 2025 May 14:2025.05.12.653131. [Version 1] doi: 10.1101/2025.05.12.653131

DNA-methylation markers associated with lung function at birth and childhood reveal early life programming of inflammatory pathways

Priyadarshini Kachroo 1,2,*, Katherine H Shutta 2,3, Enrico Maiorino 2, Matthew Moll 2,4,5, Julian Hecker 2, Vincent Carey 2, Michael J McGeachie 2, Augusto A Litonjua 6, Juan C Celedón 7; National Heart, Lung, and Blood Institute Trans-Omics for Precision Medicine (TOPMed) Consortium, Scott T Weiss 2, Dawn L DeMeo 2,4,*
PMCID: PMC12132251  PMID: 40462996

Abstract

Rationale:

Lung function deficits may be caused by early life epigenetic programming. Early childhood studies are necessary to understand life-course trends in lung diseases.

Objectives:

We aimed to examine whether DNA-methylation at birth and childhood is associated with lung function growth.

Methods:

We measured DNA-methylation in leukocytes from participants in two childhood asthma cohorts (CAMP [n=703, mean-age 12.9 years] and GACRS [n=788, mean-age 9.3 years]) and cord blood from participants in the VDAART study (n=572) to identify CpGs and pathways associated with lung function.

Results:

We identified 1,049 consistent differentially methylated CpGs (608 relatively hypermethylated) across all three studies (FDR-P<0.05). Relatively hypomethylated CpGs were enriched for gluconeogenesis, cell adhesion and VEGF signaling. Relatively hypermethylated CpGs were enriched for Hippo, B-cell and growth hormone receptor signaling. Functional enrichment suggested potential regulatory roles for active enhancers and histone modifications. Additionally, enrichment in PI3K/AKT and Notch pathways in males and enrichment in hormonal pathways in females was identified. Gaussian graphical models identified sex-differential DNA-methylation nodes and hub scores at birth and childhood. Integrating with previously identified polygenic risk scores for asthma and drug-target enrichment identified seven robust genes including MPO, CHCHD3, CACNA1S, PI4KA, EP400, CREBBP and KCNA10 with known associations as biomarkers for asthma severity and drug targets for airway inflammation.

Conclusions:

Epigenetic variability from birth through puberty provides mechanistic insights into fetal programming of developmental and immune pathways associated with lung function. These early life observations reveal potential targets for mitigating risk for lung function decline and asthma progression in later life.

Keywords: Epigenetics, lung function, pathway, sex, childhood

INTRODUCTION

Lung function (LF)(1) impairment often has origins in early life(2) and is influenced by a combination of genetics, prenatal or postnatal exposures, and environmental factors. Life-long consequences may include an early onset of chronic lung diseases including asthma and Chronic Obstructive Pulmonary Disease (COPD)(3) which share clinical features but also demonstrate substantial heterogeneity(4) and impart a major global health burden(5–9).

Asthma progression is associated with reduced lung growth and an accelerated rate of decline in FEV1 and FEV1/FVC(10). Children with reduced baseline FEV1 may be at risk of exacerbations, fixed airflow obstruction and early COPD(11–13). While several genetic variants may influence an individual’s asthma risk(14), LF genetic risk loci account for a fraction of the overall heritability with modest effects(12, 15, 16). Epigenetic marks including DNA methylation (DNA-m) are influenced by genetics and environment(17) and play an important role in lung development(18).

Prior studies have robustly associated DNA-m(19) with LF(20–24) but large-scale studies are needed to comprehensively explore their relationship from early life. A large-scale meta-analysis by Lee et al. expanded our knowledge on ancestry-specific epigenetic associations with LF, but their data mostly included older adults and sex differences were not considered (25). DNA-m levels are particularly impacted by sex during adolescence, and this may further play a role in sex-specific risks of respiratory conditions across the life course(26–28). Previously, DNA-m studies of pre-adolescent participants identified CpGs associated with sex-specific LF trajectories (age 10, 18 and 26 years)(27). Using the same cohorts, sex-specific DNA-m and gene expression patterns were identified at birth that correlated with LF in adolescence (29). Growing evidence suggests that investigating CpGs associated with high-risk LF trajectories (27, 30) may have the potential to identify specific inflammatory markers in relation to distinct asthma heterogeneity and inform preventative interventions for early-onset COPD.

This study examined epigenome-wide associations with multiple LF outcomes in three cohorts: the Genetic Epidemiology of Asthma in Costa Rica Study (GACRS), the Childhood Asthma Management Program (CAMP) and the Vitamin D Antenatal Asthma Reduction Trial (VDAART). LF trajectories associated with COPD were additionally evaluated in CAMP and asthma outcomes at age 6 years were evaluated in VDAART; sex-stratified associations were also assessed. We integrated polygenic risk scores for asthma to capture robust DNA-m signatures independent of genetic risk for asthma. Finally, we used several integrative epigenomic tools to ascertain potential biomarkers adding functional relevance for respiratory disease biology.

METHODS

Overall Study Design

The overall goal of this study was to identify early life predictors of LF in childhood asthma using genome-wide DNA-m from diverse cohorts (Figure 1). We utilized two childhood asthma studies for discovery (CAMP, n=703) and replication (GACRS, n=788) to identify CpGs associated with LF. To test an early origins hypothesis, we further investigated associations in the umbilical cord blood DNA-methylome in association with LF at ages 5 and 6 years in VDAART (n=572). We then meta-analyzed our findings to allow a more comprehensive assessment of the DNA-m associations followed by several downstream integrative analyses (Figure 1).

Figure 1.

Figure 1.

Open in a new tab

Overview of the conceptual framework and study design for the EWAS associations with LF and asthma outcomes in three different studies – CAMP, GACRS and VDAART.

Created in BioRender. Kachroo, P. (2025) https://BioRender.com/502otie

In addition to commonly analyzed LF phenotypes such as Forced Expiratory Volume in 1 second, (FEV1,), Forced Vital Capacity (FVC) and FEV1/FVC, we evaluated the ratio of forced expiratory flow in the mid-portion of vital capacity divided by FVC (FEF25–75/FVC), as a surrogate measure of airway size relative to lung size, an understudied marker of dysanapsis previously associated with LF decline(31, 32). Longitudinal LF trajectories from CAMP(12, 13) were previously defined: Normal Growth (NG), deviations from the canonical NG pattern that can appear as either reduced growth (RG), early decline (ED) or a combination of RG and ED (any RG/any ED). Further, The Global Initiative for Obstructive Lung Disease (GOLD) stages for COPD were used for CAMP subjects who met the criteria for LF impairment at their last spirometry visit (aged 23 to 30 years)(13).’

Written informed consent and assent were obtained from both the parents and the participating children. Details on study populations and spirometry assessments are provided in the online supplement.

Statistical analyses

Association of DNA-methylation with asthma and LF phenotypes

Data preprocessing and quality control methods are detailed in the Online Supplement. Mainly, to analyze DNA-m (hg19 reference genome), we used logit transformed β-values (M-values approximated by log2(β /(1-β)) (31). A multivariable robust regression model implemented in the robustbase R package(32) was used to analyze DNA-m M-values as predictor for LF/asthma outcomes adjusting for known covariates (see online supplement). DMRCate(33) was used to identify differentially methylated regions (DMRs).

To enrich insights into developmental origins, we further performed a meta-analysis using individual study-specific and male- and female-associated differentially-methylated positions (discovery DMPs, P-value<0.05) using inverse variance-weighted fixed-effects models implemented in the METAL software(34). The DMPs with a consistent direction of effect in all three or in at least two of the three studies (FDR<0.05) were retained for downstream analyses.

Functional downstream and integrative epigenomics

To assess whether the significant LF-associated DMPs were attenuated by genetic risk, we performed adjustment for polygenic risk scores for asthma (PRSasthma)(35) and a composite spirometry-based COPD PRS (FEV1, FEV1/FVC; PRSspiro(36)), both which were developed in external training cohorts(37, 38) and calculated in CAMP. We also tested models including the main effects and interaction terms for PRS X CpG. CpG site-based gene ontology and pathway enrichment analysis using KEGG were performed using missMethyl(39). The terms with ≥2 genes at a FDR-P<0.05(40) were regarded as significantly enriched. To get the most functionally relevant network representation, an integrative gene-based enrichment and visualization was performed for the meta-analyzed and male- or female-specific DMPs against five gene-set libraries (Gene Ontology, KEGG, GWAS Catalog, DisGeNET) using the knowledge-graph database and web-server enrichr-KG (https://maayanlab.cloud/enrichr-kg )(41). Further, for the LF-associated CpG DMPs that exist in both males and females but in opposite direction of effect, Gaussian graphical models (GGMs) were constructed to identify differential CpG nodes and hub scores specific to sex-stratified male-female network modules (details in Online Supplement) using the CRAN package huge(42).

Using the experimentally-derived Functional element Overlap analysis of ReGions from EWAS (eFORGE)(43, 44) integrative epigenomics tool, we further explored whether our meta-analyzed LF–associated DMPs were enriched in regulatory elements from the Roadmap Epigenomics Mapping Consortium across more than 20 cell/tissue types. We also applied the drug perturbation gene set enrichment analysis (dpGSEA)(45) (https://github.com/sxf296/drug_targeting) to the CpG-mapped genes and the directionality of effect of their DMPs and identified phenotypically relevant approved or experimental drug targets of clinical relevance derived using the Broad Institute’s Connectivity Map (CMAP) and the Library of Integrated Network-based Cellular Signatures (LINCS) framework. The top 50 drug-targeted genes with the cell-line information were evaluated. The gene-drug targets that showed a significant enrichment as well as target compatibility score at both 90% and 95% confidence interval from both databases CMAP and LINCS(46), were retained.

Biological age may reflect distinct developmental perturbations at birth and during childhood(47, 48). Therefore, we also applied generalized linear models to examine the association between the LF outcomes and measures of epigenetic age acceleration using the R/Bioconductor package methylclock(49). Elastic Net (EN) clock(50) trained on childhood data was applied for CAMP and GACRS and the EPIC clock(51) trained for gestational age predictions was applied for VDAART.

RESULTS

The basic characteristics of each study population have been provided (Table 1 and detailed in Suppl. Tables E1-3. Phenotypic data were available for 703 CAMP participants, 788 GACRS participants and 352 VDAART participants.

Table 1.

Characteristics of the study samples and participants in all three cohorts with data on DNA-methylation and lung function outcomes (N=2,063)

Populations CAMP GACRS VDAART
Number (n) 703 788 572
Age (years) 12.9 (12.1) 9.3 (1.9) 0.0 (0.0)
Gestational age >= 37 weeks, n (%) NA NA 536 (93.7)
Female | Female Child, n (%) 280 (39.8) 321 (40.7) 268 (46.9)
Presence of smoke exposure, n (%) 276 (39.3) 240 (30.5) 13 (2.3)
Longitudinal trajectories, n (%) NA NA
 Normal Growth 183 (26.0)
 Early Decline 287 (40.8)
 Reduced Growth 339 (48.2)
PRE FEV1, mean (SD) 2.6 (0.8) 1.8 (0.5) 1.2 (0.2)
PRE FVC, mean (SD) 3.3 (1.0) 2.1 (0.6) 1.4 (0.3)
PRE FEV1/FVC, mean (SD) 77.9 (8.9) 83.3 (7.5) 89.4 (6.7)
PRE BD FEF25–75, mean (SD) 2.34 (0.95) 2.0 (0.7) 1.6 (0.4)
PRE FEF25–75/FVC, mean (SD) 0.7 (0.2) 0.9 (0.3) 1.2 (0.3)
Asthma at age 6, n (%) NA NA 84 (14.7)
Heightcm, mean (SD) 156 (13.5) 133 (11.7) 115.7 (5.8)
Race and ethnicity, n (%)
 White participants 486 (69.1) NA 202 (35.3)
 African American 90 (12.8) NA 253 (44.2)
 Hispanic 66 (9.4) 788 (100) NA
 Others 61 (8.7) NA 117 (20.5)
Open in a new tab

Abbreviations: CAMP=Childhood Asthma Management Program; GACRS = The Genetic Epidemiology of Asthma in Costa Rica Study; VDAART=The Vitamin D Antenatal Asthma Reduction Trial. Study specific missingness and detailed characteristics for each study are included in Supplementary Tables E1-E3.

Associations between DNA-methylation at birth/childhood and LF

We identified several LF-associated DMPs in the VDAART birth cohort (cord-blood) (Table E4, Figure E1). Between the two childhood cohorts GECRA and CAMP, 803 overlapping CpG DMPs were associated with either FEV1/FVC, FEF25–75 and FEF25–75/FVC, with 99.8% in consistent direction of effect (Tables E5-E7, detailed in Online Supplement). Fewer associations replicated between cord-blood and childhood (Figure E2, Table E8, see online supplement). Nominally significant associations (P<0.05, Table E4, Figure E2-E3) were identified for the RG and ED trajectories in CAMP and for asthma development in VDAART (see Online Supplement).

Based on the exact CpG coordinates (chromosome, start and end) and a stringent threshold of ≥4 CpGs within the associated region, we identified 11 DMRs for FEV1/FVC (mapping to 11 genes [PCYT1A, IL4, EPX, EVL, RASSF2, VTI1A, TLDC2, FBXO7, IL5RA, IGF1R, HS2ST1]) and one significant DMR for FEF25–75/FVC (mapping to URI1) that were replicated between GACRS and CAMP (Table E9). We did not identify any LF-associated DMRs in VDAART (FDR<0.05).

Analyzing LF-associated DNA-m in the sex chromosomes identified mostly X-chromosome associated differential methylation between males and females. Relatively fewer LF-associated X-chromosome DMPs (n=37) were identified in females, that were globally hypo-methylated in gene body and promoter-associated regions compared to male X-chromosome across all studies; 16 DMPs had increased DNA-methylation levels. In males, 652 X-chromosome DMPs were hyper-methylated and 172 DMPs were hypo-methylated (Table E10).

Epigenetic age acceleration was significantly associated with reduced LF in CAMP; cord blood age acceleration was associated with an increased LF in VDAART by age 5–6 (Table E11, see Online Supplement).

Meta-analysis – Recapitulation of LF-associated DMPs with adult COPD

A total of 9,851 meta-DMPs were shared between all LF outcomes (Figure 2). We identified 1,049 (812 unique) meta-LF DMPs (Table 2, FDR<0.05) having consistent direction of effect across all studies, with 338 hypo- (209 genes) and 474 (326 genes) hyper-methylated (Table E12); 47 DMPs were associated with at least three LF traits (Figure 3A, Table E12); the top five consistent DMPs are highlighted in Table 3. Three of the 1,049 meta-DMPs were also associated with the reduced lung growth-trajectory DMPs (cg20981347, cg26657392: MFSD12 and cg17950165: LINC01182). Nine meta-DMPs (flagged in Table E12) were shared across seven blood EWAS studies in adults(52) in association with COPD (cg04637264:KCNIP2, cg09598552:CDH23, cg09646173:PDE6A), FEV1 (cg09646173:PDE6A, cg12077460:MFHAS1) and FEV1/FVC (cg12147622, cg00762550:LIG3, cg01878963:RAP1A, cg16518176:PES1P1) DMPs, while cg00278366:RAD9B was common to all.

Figure 2.

Figure 2.

Open in a new tab

Multi-trait PheWAS graphical representation of the manhattan plot for the meta-analyzed EWAS associations. Only the overlapping DMPs across all tested LF phenotypes were used as input for this plot. Highlighted associations in blue are at a genome-wide threshold (p-value threshold=5.8×10−8). Highlighted associations in red are at a p-value threshold of 1×10−10 for better visualization of the top hits overlapping across multiple phenotypes. The density of CpGs on each chromosome are shown by the legend key with lowest to highest density ranging from green to yellow to red color.

Table 2.

Statistics and Number of epigenome-wide differentially methylated associations from the meta-analyses across three independent study populations: GACRS, CAMP and VDAART

Meta-analyzed phenotype across all three studies FEV1 FVC FEV1/FVC FEF25–75 FEF25–75/FVC
Associations tested
 Overall 132,109 194,765 183,255 152,355 174,109
 Males 157,142 157,759 160,359 159,318 159,659
 Females 139,063 147,010 166,568 153,720 166,407
Associations at FDR<0.05
 Overall 110,919 170,309 165,236 134,015 156,043
 Males 136,910 132,933 141,759 138,233 141,141
 Females 116,869 120,345 146116 133,460 146,614
Consistent at FDR<0.05
 Overall 165 121 261 280 222
 Males 295 83 199 463 277
 Females 52 49 99 72 86
Open in a new tab

Figure 3.

Figure 3.

Figure 3.

Open in a new tab

Upset plot showing the Intersection of differentially methylated CpGs from the meta-analysis in consistent direction of DNA-methylation effect across all three cohorts at FDR < 0.05 A. across LF phenotypes B. across LF phenotypes in males and females.

Table 3.

Top five CpGs with available gene annotations and showing consistent DNA-methylation associations with LF outcomes meta-analyzed between GACRS, CAMP and VDAART cohorts represented by the ‘Direction’ column respectively (FDR < 0.05). The list is sorted by FDR.

Phenotype CpG probe chr pos Gene Effect Direction P.value FDR Relation to Island Gene Context
FEV1 cg12684668 chr5 150,403,466 GPX3 −0.0903 --- 7.56E-08 8.76E-05 S_Shelf Body
cg12233571 chr5 36,613,509 SLC1A3 0.0732 +++ 1.77E-07 0.00016674 OpenSea Body
cg01978458 chr7 2,683,257 TTYH3 0.0894 +++ 3.11E-07 0.00024335 N_Shelf Body
cg07447194 chr3 148,711,966 GYG1 0.0596 +++ 3.66E-07 0.0002705 S_Shore Body
cg12120947 chr6 32,166,652 NOTCH4 0.0594 +++ 3.71E-07 0.0002705 S_Shelf Body
FVC cg01900413 chr11 128,419,356 ETS1 0.1068 +++ 1.44E-06 0.00096312 Island Body
cg13618478 chr19 56,549,544 NLRP5 −0.0751 --- 1.54E-06 0.00100793 OpenSea ExonBnd
cg21817833 chr14 52,118,115 FRMD6 0.1585 +++ 1.62E-06 0.00103473 Island TSS1500
cg09405702 chr8 12,588,030 LONRF1 0.049 +++ 1.79E-06 0.00109532 OpenSea Body
cg07167860 chr13 93,136,776 GPC5 0.0411 +++ 2.58E-06 0.00130417 OpenSea Body
FEV1/FVC cg25627789 chr7 132,694,521 CHCHD3 3.5519 +++ 1.16E-12 3.86E-09 OpenSea Body
cg14978242 chr5 79,501,131 SERINC5 2.2488 +++ 9.02E-11 7.21E-08 OpenSea Body
cg04141008 chr16 3,843,232 CREBBP 3.7667 +++ 2.95E-10 1.73E-07 OpenSea Body
cg05380077 chr11 63,272,225 LGALS12 3.6992 +++ 5.91E-10 2.82E-07 OpenSea TSS1500
cg18460265 chr1 111,062,669 KCNA10 2.7615 +++ 1.09E-09 4.39E-07 OpenSea TSS1500
FEF25–75 cg16107105 chr7 150,646,704 KCNH2 0.1672 +++ 1.32E-10 2.89E-07 N_Shore Body
cg09845476 chr9 127,113,228 NEK6 0.2199 +++ 1.63E-10 3.25E-07 OpenSea 3’UTR
cg26054828 chr10 75,619,500 CAMK2G 0.37 +++ 1.82E-10 3.33E-07 OpenSea Body
cg12120947 chr6 32,166,652 NOTCH4 0.163 +++ 2.60E-10 3.94E-07 S_Shelf Body
cg18337287 chr19 930,871 ARID3A 0.1486 +++ 3.72E-10 4.68E-07 N_Shore Body
FEF25–75/FVC cg13774539 chr2 74,612,706 LOC100189589 0.1213 +++ 1.02E-10 5.82E-08 OpenSea TSS200
cg14978242 chr5 79,501,131 SERINC5 0.074 +++ 2.11E-10 9.96E-08 OpenSea Body
cg18181035 chr3 128,214,460 GATA2-AS1 0.1578 +++ 8.46E-10 2.87E-07 N_Shore Body
cg21271570 chr13 44,961,267 SERP2 0.0897 +++ 3.24E-09 8.53E-07 OpenSea Body
cg12242115 chr12 122,251,166 SETD1B −0.1273 --- 3.78E-09 9.66E-07 S_Shore Body
Open in a new tab

Identification of PRS-robust CpGs after PRS adjustment

PRS integration reduced our sample size in CAMP by almost 38%, yet the regression coefficients remained unchanged for 44/387 DMPs that remained robustly associated with FEV1 (Table E13, Figure E4A). Notably, three of the 44 DMPs were associated in the meta-analysis for CpG-FEV1 associations: cg04266202 (MPO) and cg06070625 (MITF). The regression coefficients remained unchanged for 1,485/3,485 CpGs that remained robustly associated with FEV1/FVC (Table E14, Figure E4B). Notably, 10 were associated in the meta-analysis of CpG-FEV1/FVC associations (cg25627789:CHCHD3, cg12046819:SGMS1, cg00068153:CACNA1S, cg09662086:PI4KA, cg05486260:FAM135B, cg00764582:EP400, cg22330572:AZIN1-AS1, cg04141008:CREBBP, cg18460265: KCNA10). Similar trends were observed when evaluating LF-associated DMPs and PRSspiro models with potential relevance for COPD risk (Figure E4).

Integrative epigenomic analyses identified pathways with functional implications

Of the meta-analyzed LF-DMPs, hypo-methylated CpGs were enriched for growth-related and developmental pathways (Table E15) including Vascular Endothelial Growth Factor (VEGF) receptor signaling (FGF9, PTK2, VAV2, VEGFC), semaphorin-plexin signaling (PLXND1, PLXNA4), glucose metabolism (PGM2, ADPGK, FBP2) and negative Wnt signaling regulation (SOX30, FGF9, LATS2, UBAC2, KREMEN1, NKD1). Hyper-methylated CpGs were enriched for developmental and immune function related processes and pathways (Table E15) including positive regulation of B cell receptor /antigen receptor-mediated signaling (PRKCH, SLC39A10, CD81, LGALS3), insulin receptor signaling (SORBS1, OSBPL8, PTPN1), growth hormone receptor signaling (PTPN1, STAT5A), Wnt signaling (CSNK1E, WNT3), and the Hippo signaling pathway (RASSF1, FRMD6, CSNK1E, RASSF4). The gene-based enrichment identified similar and new integrative associations from KEGG, GWAS Catalog and DisGeNET (Figure 4).

Figure 4.

Figure 4.

Figure 4.

Open in a new tab

Functional integrative gene-based enrichment for genes annotated to the meta-analyzed hypo- and hyper-methylated CpGs consistent across all three studies and across all LF phenotypes: GACRS, CAMP and VDAART. Nodes are the genes; rectangles are the enriched terms. Each library source for the gene-based term enrichment is given a different color: Gene ontology (GO) Biological process (Red), KEGG (Blue), GWAS Catalog (Green) and DisGeNET (Purple), and the gradient is based on the z-score (the darker, the higher z-score). The hypo-methylated sub-network was enriched for biological processes: cell migration and adhesion, embryonic development; KEGG pathways: Calcium signaling, Glycolysis and Rap1 signaling; GWAS associations: post-bronchodilator FEV1/FVC ratio, type 2 diabetes, neurocognitive behaviors. The hyper-methylated sub-network was enriched for biological process: amino acid transport; KEGG pathways: B-cell receptor signaling, ErbB signaling, Aldosterone synthesis/secretion and Notch signaling pathway; GWAS associations: Obesity related traits, insulin resistance and age-related macular degeneration. A. Enrichment analysis for gene list based on hypo-methylated CpGs; B. Enrichment analysis for gene list based on hyper-methylated CpGs

Of all the identified phenotype-specific drug targets (Table E16), 42 drug targets (19 unique genes) and 576 drug targets (27 unique genes) were identified using CMAP and LINCS respectively. Seven of those included the PRS-robust genes for FEV1 (MPO) and FEV1/FVC (CHCHD3, CACNA1S, PI4KA, EP400, CREBBP, KCNA10). Both databases identified 13 overlapping genes with several approved or known drug targets for respiratory and other health outcomes; SERINC5, NOTCH4 and EP400 had multi-phenotype associations (Table 6).

For further functional support of our findings, eFORGE integration for hypo-methylated CpGs showed enrichment of DNAase hotspots in the blood, lung and skin tissues, a strong enrichment of H3K4me1 in blood and transcriptionally active enhancers in several tissues including fetal lung tissue. The hyper-methylated CpGs showed enrichment of DNAase hotspots, a strong enrichment of H3K4me1 and H3K36me3 marks (Figure E5) and a weak transcription with enhancer activity across majority of the tissues including fetal lung.

Sex stratified meta-analysis – Unique sex-specific and sex-divergent LF-DMPs

When stratified by sex, several consistent associations either for males (1,317 DMPs; 1,035 hyper-methylated) or for females (358 DMPs; 1 hyper-methylated) were identified between all studies (Table 2, Table E17, Figure 3B). Male-associated hypo-methylated DMPs were enriched for the PI3K-Akt signaling pathway and hyper-methylated DMPs were enriched for immune cell phenotypes, and Notch signaling. Female-associated hypo-methylated DMPs were enriched for GWAS genes for FEV1 and FEV1/FVC ratio, while the hyper-methylated DMPs were enriched for growth hormone, thyroid hormone, aldosterone and insulin signaling (Table E18).

Further, 76 (68 unique) DMPs exhibited opposite direction of effect between males and females (FDR < 0.05 in either males or females) and had consistent associations in at least two of the three studies for any LF outcome (Table E17). We constructed stratified GGMs on these DMPs and calculated hub scores of each node to assess how influential each DMP is within the female network and the male network (Figure E6A). This analysis yielded 11 DMPs (six with gene annotations) with more than 20% difference in hub scores between males and females from all three study populations (Figure E6B). One CpG annotated to PLTP gene (difference=0.22) was identified in GACRS, four CpGs annotated to STX12 (difference=0.25), LOC101928304 (difference=0.21), KBTBD11 (difference=0.31) and RCCD1 (difference=0.21) were identified in CAMP and 1 CpG annotated to MAP3K7 (difference=0.36) was identified in VDAART. Additionally, eight DMPs were identified that were remarkably different between the birth cohort (VDAART) and either of the two childhood cohorts (GACRS, CAMP; Figure E6C).

DISCUSSION

Epigenetic modifications capture pre- and post-natal environmental exposures, and these signatures may impact lung function and potentially lifelong susceptibility to chronic lung diseases like asthma and COPD(52). Prior large-scale EWAS studies(51) inadequately captured heterogeneity or were limited in scope, motivating us to comprehensively examine the relationship between DNA-m, lung function and asthma from birth to early adulthood. To our knowledge, this is the largest EWAS study of lung function in children at risk for and with asthma, revealing epigenetic dysregulation in key pathways including Hippo, Wnt and VEGF related signaling. In silico analyses have revealed gene-drug associations across three heterogeneous populations spanning birth and childhood and provide potential targets to explore further for primary prevention of obstructive lung disease.

Several population-specific LF-CpG associations replicated between GACRS and CAMP for FEV1/FVC and FEF25–75/FVC phenotypes, while many cord-blood associations were unique to VDAART, suggesting age-related heterogeneity in the epigenome. The LF-associated DMRs replicating between GACRS and CAMP included IL4, EPX, IL5RA, IGF1R, all previously associated with IgE-mediated respiratory diseases (53), allergic asthma(19) and reduced airway hyperresponsiveness in mice exposed to house-dust mite allergen(54). Such candidates could be of interest for epigenetic interventions focused on allergic asthma endotypes.

Our meta-analyzed LF-associated CpGs highlighted novel hypo- and hyper-methylated loci, pathway changes predicting LF decline, and metabolic pathway disruptions which may provide a clearer snapshot of the global DNA-m perturbations impacting asthma pathophysiology and COPD risk. VEGF signaling and glucose metabolism pathways suggest potential intervention targets. Particularly, upregulation of cell adhesion and VEGF signaling plays an important role in Th2 inflammation, regulating airway remodeling and hyper-responsiveness in both asthma and COPD(55). Previously, we identified enrichment of VEGFA-VEGFR2 signaling in fetal lung exposed to in-utero smoke(56) suggesting that some of these marks could be triggered by maternal exposures. Tissue-specific functional enrichment of H3K4me1 marks provided evidence of an additional regulatory role of active enhancers and histone modifications(57). Histone methylation disruption and increased VEGF driven by IL13 has been linked to Th2 inflammation in asthma(58). Moreover, dysregulation in metabolic pathways may be observed during airway inflammation in respiratory diseases(59, 60). Experimental evidence further links VEGF and inhibition of glycolysis to hypoxia in pulmonary hypertension and endothelial dysfunction (61). Epigenetic perturbations in glucose metabolism suggests metabolic reprogramming in the genes mapping to this pathway(60), (62).

Hypermethylated CpGs associated with lung function were enriched for Hippo, B-cell and growth hormone receptor signaling pathways. Several prior studies have demonstrated that alterations in Wnt(63) and Hippo signaling(64) impact lung development and asthma progression(65). Previously, we have shown hyper-methylated CpGs in fetal lung exposed to in-utero smoke recapitulate in adult lung tissue and obstructive lung disease, with enrichment in Wnt and Hippo pathways(66). Balance between Wnt enhancers and inhibitors has demonstrated perturbations in airways from individuals with severe asthma(67). Hyper-methylation of LF-associated CpGs, such as LATS2, a regulator of Wnt-Hippo, can downregulate hippo signaling genes, which could be one potential mechanism to elevate Wnt signaling, disrupting the balance between Th2 and Th17 inflammatory responses in asthma. Positive regulation of B-cell receptor signaling was an enriched pathway for hyper-methylated CpGs, suggesting that B cell regulation could be an immune pathway dysregulated by epigenetic control in childhood and impact future risk for adult lung disease, given recent research supporting the role of the B cell in COPD pathogenesis(68). Genes mapping to the Insulin and growth hormone receptor signaling pathways could point us to the epigenetic regulation of metabolic syndrome affecting the uptake of glucose and lipids associated with asthma pathogenesis. In this pathway, the CpGs for STAT5A and PTPN1 were hyper-methylated across the three studies and were around the transcription start site of the promoter region. Hyper-methylation of the STAT5A promoter has been associated with decreased expression in children with asthma(69). PTPN1 promoter hyper-methylation was associated with type-2-diabetes(70). It is also noteworthy to find interactions between ErbB and aldosterone signaling in the hyper-methylated sub-network. Inhaled corticosteroids (ICS) are recommended to prevent asthma exacerbation in persistent asthma, and aberrant ErbB signaling mediates corticosteroid resistance driven by IL13 in bronchial hyperresponsiveness(71). Further, ICS drugs mainly exert their effects on mineralocorticoids like aldosterone and glucocorticoid hormone receptors, therefore their prolonged exposure can cause defective DNA-binding or receptor mutations as seen in cortisol resistance(72); genes we have identified in the hypermethylated subnetwork may have pharmacoepigenetic relevance.

Existing studies show that LF-associated DNA-m in childhood and adolescence may vary by sex (73–75). We observed male enriched pathways included PI3K/AKT, linked to severe asthma outcomes, while female-specific enriched pathways involved thyroid, aldosterone, and insulin signaling. In contrast, sex-divergent methylation in specific genes may reveal insights into prenatal priming of age-related epigenetic perturbations associated with lung function and asthma in childhood. Exemplar genes with sex variable associations with lung function include MAP3K7, which is a known direct target for IL13 therapy in asthma(76, 77). Reduced PLTP, a phospholipid transport gene, which has been associated with neutrophil degranulation in COPD(78). IL20RB(79) and TRH(80) are critical lung maturation genes with strikingly different hub scores between birth and childhood; previously these genes have been linked with lung fibrosis development. Sex-specific DNA-m differences were also strikingly evident for genes on the X-chromosome, suggestive of potential X-chromosome mechanisms could impact lung function and potentially obstructive lung disease severity(81, 82). Our findings suggest that an in-depth sex-specific investigation should be the standard in large-scale studies. This would help characterize the role of epigenetic marks in sex-related features of lung function and prevalences of obstructive lung diseases across the life-course.

In silico approaches applied to our epigenetic associations identified key genes and potential clinical drug targets for further investigation and validation. SERINC5 and NOTCH4 were enriched for multiple LF phenotypes. Interestingly, NOTCH4, is regulated by Wnt-Hippo signaling(83), its inhibition can suppress Th17-mediated asthma hyper-responsiveness and vascular remodeling in COPD(84). For example, Maraviroc, the first approved CCR5 antagonist and Notch4 drug target, showed promise against severe respiratory conditions(85). SERINC5 regulates viral infections through multiple immune pathways which could mitigate defective host defense mechanisms through targeting (86). Two of our drug targets also included PRS-robust genes MPO(87) and CREBBP(88), considered as biomarker for asthma severity and neutrophilic inflammation; their inhibition could attenuate oxidative damage.

Our study demonstrates that genetic signals may partially attenuate epigenetic signals. However, PRS adjustment did not fully attenuate CpGs associations suggesting these sites may be more susceptible to changes in DNA-m triggered by environmental exposures and potentially better targets for primary prevention through epigenetic plasticity. Only 20 (2.4%) of our 530 consistent meta-LF genes overlap with the lung function genetic associations identified by GWAS(37) suggesting low influence of genetic attenuation on these DNA-m signals. Likewise, Lee et al(25), reported limited overlap of their findings with genetic data. However, several of our identified meta LF-hits overlapped with the COPD and LF-associated CpGs across 17 previously published EWAS studies(52). Future studies should consider investigating causal effects on the epigenome and LF.

Our study has notable strengths in providing a broader understanding of birth and childhood DNA-m associations with lung function in asthma, using three multi-ethnic cohorts and various informatic methods to identify genes and pathways for future clinical study. Several limitations should also be acknowledged. Phenotypic variability, residual population substructure and small sample sizes in certain strata may account for lack of replication of some findings. Polygenic risk scores for asthma were only evaluated for CAMP, limiting replication of signals after adjustment for genetic risk. Further, we did not adjust for smoking-associated CpGs as a measure of second-hand smoke exposure, and this may reflect residual confounding in our findings.

In conclusion, we provide an integrative multidimensional framework demonstrating developmental plasticity and early-life programming of epigenetic mechanisms associated with lung function between birth and adolescence. Our findings inform the identification of genes and pathways for lung function in children with or at risk for asthma and highlight the epigenome as a next-generation target for therapeutic interventions for primary prevention of adult lung diseases and related inflammatory conditions.

Supplementary Material

Supplement 1
media-1.docx (1.5MB, docx)
Supplement 2
media-2.xlsx (18.4MB, xlsx)

Table 4.

Gene-drug targets with significant enrichment and target compatibility score for the meta-analyzed CpG associations specific to each LF outcome

PHENOTYPE CMAP database LINCS database OVERLAP from both databases Number of Drug Targets in CMAP Number of Drug Targets in LINCS
FEV1 2 (GPX3, SLC1A3) 6 (GYG1, NOTCH4, MPO, ABAT, GPX3, SLC1A3) GPX3 2 12
SLC1A3 6 35
FVC 2 (ETS1, LONRF1) 1 (ETS1) ETS1 4 56
FEV1/FVC 8 (CACNA1S, SLC38A10, PPM1H, TINAGL1, CHCHD3, SERINC5, KCNA10, EP400) 7 (CREBBP, PI4KA, RPS6KC1, CHCHD3, SERINC5, KCNA10, EP400) CHCHD3 2 15
SERINC5 1 74
KCNA10 2 2
EP400 1 22
FEF25–75 5 (CAMK2G, NOTCH4, ARID3A, SERINC5, SDC3) 11 (KCNH2, ARID3A, LTBP1, RAB20, SIGLEC8, ITSN1, CAMK2G, NOTCH4, ARID3A, SERINC5, SDC3) CAMK2G 1 18
NOTCH4 4 9
ARID3A 2 11
SERINC5 2 73
SDC3 6 23
FEF25–75/FVC 6 (SERINC5, SETD1B, RAB20, KCNA10, EP400, CACNA1S) 8 (KSR1, SGMS1, CANX, SERINC5, SETD1B, RAB20, KCNA10, EP400) SERINC5 2 90
SETD1B 1 14
RAB20 1 27
KCNA10 2 2
EP400 1 21
Open in a new tab
*

See supplemental Table E16 for drug targets

Key messages:

  • We identified consistent DNA methylation signatures between birth and childhood in critical metabolic, lung development and immune pathways that were associated with lung function and may be influenced by sex and genetics.

  • Our integrative findings provide a deeper understanding for accelerated lung function decline across the life-course and could pave the way for translational interventions for lung diseases based on epigenetic plasticity.

Acknowledgements

The authors thank all the participants, investigators and staff without whom this work could not have been accomplished. We gratefully acknowledge the individuals and studies that provided biological samples and data to TOPMed CAMP and GACRS and VDAART studies. We also appreciate the support of the NHLBI TOPMed initiative in facilitating the 850K EPIC array data generation using the Infinium® MethylationEPIC 850K BeadChip and contributing to overall research work. For a comprehensive list of TOPMed collaborators, refer to https://www.nhlbiwgs.org/topmed-banner-authorship

Grant Funding:

US National Institutes of Health grants: K99HL159234 (P.K.), R00HL159234 (P.K.), P01HL132825 (S.T.W), N01-HC-25195 and HHSN268201500001I (TOPMed), R01HG011393 (D.L.D), R01HL155742 (M.J.M), P01HL114501 (D.L.D, K.H.S), R21HL156122 (D.L.D), U01HL089856 (D.L.D), K01HL169756 (J.H), K01HL166705 (E.M), T32HL007427 (K.H.S), K08HL159318 (M.M)

Footnotes

Conflicts of Interest

J.C.C. received research materials (inhaled steroids) from Merck, to provide medications free of cost to participants in an NIH-funded study, unrelated to this work. AAL contributes to UpToDate, Inc.—author of online education, royalties totaling not more than $3000 per year. STW receives royalties from UpToDate and is on the board of Histolix a digital pathology company. Rest authors declare no conflicts of interest.

This article has an online data supplement.

Data sharing statement

All TOPMed data is person-sensitive, however it can be requested for access and can be made available through the TOPMed consortium after careful review and approval by the TOPMed Data Access Committee (https://topmed.nhlbi.nih.gov/). Participant consent and Data Use Limitations differs within and across TOPMed studies and should be requested individually. Additional documentation, such as of local IRB approval and/or letters of collaboration with the primary study PI(s) may be required.

The CAMP DNA methylation datasets analyzed in the current study are available at the database of Genotypes and Phenotypes (dbGaP) repository (phs001726. v2. p1) here: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs001726.v2.p1. The CRA DNA methylation datasets analyzed in the current study are available at the dbGaP repository (phs000988. v5. p1) here: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000988.v5.p1.

All datasets used and/or analyzed during the current study could be requested from the corresponding author and contact study PIs and made available on reasonable request.

Reference:

  • 1.Jamrozik E, Knuiman MW, James A, Divitini M, Musk AWB. Risk factors for adult-onset asthma: a 14-year longitudinal study. Respirology (Carlton, Vic). 2009;14(6):814–21. doi: 10.1111/j.1440-1843.2009.01562.x. [DOI] [PubMed] [Google Scholar]
  • 2.Sears MR. Lung function decline in asthma. Eur Respir J. 2007;30(3):411–3. doi: 10.1183/09031936.00080007. [DOI] [PubMed] [Google Scholar]
  • 3.Bonnelykke K, Ober C. Leveraging gene-environment interactions and endotypes for asthma gene discovery. The Journal of allergy and clinical immunology. 2016;137(3):667–79. doi: 10.1016/j.jaci.2016.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Carr TF, Bleecker E. Asthma heterogeneity and severity. The World Allergy Organization journal. 2016;9(1):41-. doi: 10.1186/s40413-016-0131-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ivanova O, Richards LB, Vijverberg SJ, Neerincx AH, Sinha A, Sterk PJ, et al. What did we learn from multiple omics studies in asthma? Allergy. 2019. doi: 10.1111/all.13833. [DOI] [PubMed] [Google Scholar]
  • 6.Global Initiative for Asthma. Global strategy for asthma management and prevention, 2021.
  • 7.Masoli M, Fabian D, Holt S, Beasley R. The global burden of asthma: executive summary of the GINA Dissemination Committee report. Allergy. 2004;59(5):469–78. Epub 2004/04/15. doi: 10.1111/j.1398-9995.2004.00526.x ALL526 [pii]. [DOI] [PubMed] [Google Scholar]
  • 8.Becker AB, Abrams EM. Asthma guidelines: the Global Initiative for Asthma in relation to national guidelines. Current opinion in allergy and clinical immunology. 2017;17(2):99–103. doi: 10.1097/ACI.0000000000000346. [DOI] [PubMed] [Google Scholar]
  • 9.gov CDC. CDC - Asthma - Data and Surveillance - Asthma Surveillance Data. [Google Scholar]
  • 10.Strunk RC, Weiss ST, Yates KP, Tonascia J, Zeiger RS, Szefler SJ. Mild to moderate asthma affects lung growth in children and adolescents. The Journal of allergy and clinical immunology. 2006;118(5):1040–7. doi: 10.1016/j.jaci.2006.07.053. [DOI] [PubMed] [Google Scholar]
  • 11.Tantisira KG, Fuhlbrigge AL, Tonascia J, Van Natta M, Zeiger RS, Strunk RC, et al. Bronchodilation and bronchoconstriction: predictors of future lung function in childhood asthma. The Journal of allergy and clinical immunology. 2006;117(6):1264–71. Epub 2006/06/06. doi: 10.1016/j.jaci.2006.01.050. [DOI] [PubMed] [Google Scholar]
  • 12.McGeachie MJ, Yates KP, Zhou X, Guo F, Sternberg AL, Van Natta ML, et al. Genetics and Genomics of Longitudinal Lung Function Patterns in Individuals with Asthma. Am J Respir Crit Care Med. 2016;194(12):1465–74. doi: 10.1164/rccm.201602-0250OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.McGeachie MJ, Yates KP, Zhou X, Guo F, Sternberg AL, Van Natta ML, et al. Patterns of Growth and Decline in Lung Function in Persistent Childhood Asthma. New England Journal of Medicine. 2016;374(19):1842–52. doi: 10.1056/NEJMoa1513737. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Sharma S, Chhabra D, Kho AT, Hayden LP, Tantisira KG, Weiss ST. The genomic origins of asthma. Thorax. 2014;69(5):481–7. doi: 10.1136/thoraxjnl-2014-205166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kachroo P, Hecker J, Chawes BL, Ahluwalia TS, Cho MH, Qiao D, et al. Whole Genome Sequencing Identifies CRISPLD2 as a Lung Function Gene in Children With Asthma. Chest. 2019. doi: 10.1016/j.chest.2019.08.2202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Portas L, Pereira M, Shaheen SO, Wyss AB, London SJ, Burney PGJ, et al. Lung Development Genes and Adult Lung Function. American journal of respiratory and critical care medicine. 2020. doi: 10.1164/rccm.201912-2338OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Yang IV, Lozupone CA, Schwartz DA. The environment, epigenome, and asthma. The Journal of allergy and clinical immunology. 2017;140(1):14–23. doi: 10.1016/j.jaci.2017.05.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Smith ZD, Meissner A. DNA methylation: roles in mammalian development. Nature reviews Genetics. 2013;14(3):204–20. doi: 10.1038/nrg3354. [DOI] [PubMed] [Google Scholar]
  • 19.Cardenas A, Sordillo JE, Rifas-Shiman SL, Chung W, Liang L, Coull BA, et al. The nasal methylome as a biomarker of asthma and airway inflammation in children. Nature communications. 2019;10(1):3095-. doi: 10.1038/s41467-019-11058-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Wang AL, Qiu W, DeMeo DL, Raby BA, Weiss ST, Tantisira KG. DNA methylation is associated with improvement in lung function on inhaled corticosteroids in pediatric asthmatics. Pharmacogenetics and genomics. 2019;29(3):65–8. doi: 10.1097/FPC.0000000000000366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.den Dekker HT, Burrows K, Felix JF, Salas LA, Nedeljkovic I, Yao J, et al. Newborn DNA-methylation, childhood lung function, and the risks of asthma and COPD across the life course. The European respiratory journal. 2019;53(4). doi: 10.1183/13993003.01795-2018. [DOI] [PubMed] [Google Scholar]
  • 22.Imboden M, Wielscher M, Rezwan FI, Amaral AFS, Schaffner E, Jeong A, et al. Epigenome-wide association study of lung function level and its change. The European respiratory journal. 2019;54(1). doi: 10.1183/13993003.00457-2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Jamieson E, Korologou-Linden R, Wootton RE, Guyatt AL, Battram T, Burrows K, et al. Smoking, DNA Methylation, and Lung Function: a Mendelian Randomization Analysis to Investigate Causal Pathways. American journal of human genetics. 2020;106(3):315–26. doi: 10.1016/j.ajhg.2020.01.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Sunny SK, Zhang H, Rezwan FI, Relton CL, Henderson AJ, Merid SK, et al. Changes of DNA methylation are associated with changes in lung function during adolescence. Respiratory research. 2020;21(1):80-. doi: 10.1186/s12931-020-01342-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Lee M, Huan T, McCartney DL, Chittoor G, de Vries M, Lahousse L, et al. Pulmonary Function and Blood DNA Methylation: A Multiancestry Epigenome-Wide Association Meta-analysis. Am J Respir Crit Care Med. 2022;206(3):321–36. doi: 10.1164/rccm.202108-1907OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Koo H-K, Morrow J, Kachroo P, Tantisira K, Weiss ST, Hersh CP, et al. Sex-specific associations with DNA methylation in lung tissue demonstrate smoking interactions. Epigenetics. 2020:1–12. doi: 10.1080/15592294.2020.1819662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Sunny SK, Zhang H, Mzayek F, Relton CL, Ring S, Henderson AJ, et al. Pre-adolescence DNA methylation is associated with lung function trajectories from pre-adolescence to adulthood. Clin Epigenetics. 2021;13(1):5. Epub 20210106. doi: 10.1186/s13148-020-00992-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Solomon O, Huen K, Yousefi P, Küpers LK, González JR, Suderman M, et al. Meta-analysis of epigenome-wide association studies in newborns and children show widespread sex differences in blood DNA methylation. Mutat Res Rev Mutat Res. 2022;789:108415. Epub 20220314. doi: 10.1016/j.mrrev.2022.108415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Mukherjee N, Arathimos R, Chen S, Kheirkhah Rahimabad P, Han L, Zhang H, et al. DNA methylation at birth is associated with lung function development until age 26 years. Eur Respir J. 2021;57(4). Epub 20210415. doi: 10.1183/13993003.03505-2020. [DOI] [PubMed] [Google Scholar]
  • 30.Martino DJ, Bui DS, Li S, Idrose S, Perret J, Lowe AJ, et al. Genetic and Epigenetic Associations with Pre-Chronic Obstructive Pulmonary Disease Lung Function Trajectories. Am J Respir Crit Care Med. 2023;208(10):1135–7. doi: 10.1164/rccm.202306-1025LE. [DOI] [PubMed] [Google Scholar]
  • 31.Du P, Zhang X, Huang CC, Jafari N, Kibbe WA, Hou L, et al. Comparison of Beta-value and M-value methods for quantifying methylation levels by microarray analysis. BMC Bioinformatics. 2010;11:587. Epub 20101130. doi: 10.1186/1471-2105-11-587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.R Development Core Team R. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: R Foundation for Statistical Computing; 2011. p. 409-. [Google Scholar]
  • 33.Peters TJ, Buckley MJ, Statham AL, Pidsley R, Samaras K, V Lord R, et al. De novo identification of differentially methylated regions in the human genome. Epigenetics & chromatin. 2015;8:6-. doi: 10.1186/1756-8935-8-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26(17):2190–1. Epub 2010/07/10. doi: 10.1093/bioinformatics/btq340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Moll M, Sordillo JE, Ghosh AJ, Hayden LP, McDermott G, McGeachie MJ, et al. Polygenic risk scores identify heterogeneity in asthma and chronic obstructive pulmonary disease. J Allergy Clin Immunol. 2023;152(6):1423–32. Epub 20230816. doi: 10.1016/j.jaci.2023.08.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Moll M, Sakornsakolpat P, Shrine N, Hobbs BD, DeMeo DL, John C, et al. Chronic obstructive pulmonary disease and related phenotypes: polygenic risk scores in population-based and case-control cohorts. Lancet Respir Med. 2020;8(7):696–708. doi: 10.1016/s2213-2600(20)30101-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Shrine N, Guyatt AL, Erzurumluoglu AM, Jackson VE, Hobbs BD, Melbourne CA, et al. New genetic signals for lung function highlight pathways and chronic obstructive pulmonary disease associations across multiple ancestries. Nat Genet. 2019;51(3):481–93. Epub 20190225. doi: 10.1038/s41588-018-0321-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Han Y, Jia Q, Jahani PS, Hurrell BP, Pan C, Huang P, et al. Genome-wide analysis highlights contribution of immune system pathways to the genetic architecture of asthma. Nat Commun. 2020;11(1):1776. Epub 20200415. doi: 10.1038/s41467-020-15649-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Phipson B, Maksimovic J, Oshlack A. missMethyl: an R package for analyzing data from Illumina’s HumanMethylation450 platform. Bioinformatics (Oxford, England). 2016;32(2):286–8. doi: 10.1093/bioinformatics/btv560. [DOI] [PubMed] [Google Scholar]
  • 40.Li W, Shih A, Freudenberg-Hua Y, Fury W, Yang Y. Beyond standard pipeline and p < 0.05 in pathway enrichment analyses. Comput Biol Chem. 2021;92:107455. Epub 20210212. doi: 10.1016/j.compbiolchem.2021.107455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Evangelista JE, Xie Z, Marino GB, Nguyen N, Clarke DJB, Ma’ayan A. Enrichr-KG: bridging enrichment analysis across multiple libraries. Nucleic Acids Res. 2023;51(W1):W168–w79. doi: 10.1093/nar/gkad393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Zhao T, Liu H, Roeder K, Lafferty J, Wasserman L. The huge Package for High-dimensional Undirected Graph Estimation in R. J Mach Learn Res. 2012;13:1059–62. [PMC free article] [PubMed] [Google Scholar]
  • 43.Breeze CE, Reynolds AP, van Dongen J, Dunham I, Lazar J, Neph S, et al. eFORGE v2.0: updated analysis of cell type-specific signal in epigenomic data. Bioinformatics. 2019;35(22):4767–9. doi: 10.1093/bioinformatics/btz456. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Breeze CE. Cell Type-Specific Signal Analysis in Epigenome-Wide Association Studies. Methods Mol Biol. 2022;2432:57–71. doi: 10.1007/978-1-0716-1994-0_5. [DOI] [PubMed] [Google Scholar]
  • 45.Fang M, Richardson B, Cameron CM, Dazard JE, Cameron MJ. Drug perturbation gene set enrichment analysis (dpGSEA): a new transcriptomic drug screening approach. BMC Bioinformatics. 2021;22(1):22. Epub 20210112. doi: 10.1186/s12859-020-03929-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Lim N, Pavlidis P. Evaluation of connectivity map shows limited reproducibility in drug repositioning. Sci Rep. 2021;11(1):17624. Epub 20210902. doi: 10.1038/s41598-021-97005-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Peng C, Cardenas A, Rifas-Shiman SL, Hivert MF, Gold DR, Platts-Mills TA, et al. Epigenetic age acceleration is associated with allergy and asthma in children in Project Viva. J Allergy Clin Immunol. 2019;143(6):2263–70.e14. Epub 20190206. doi: 10.1016/j.jaci.2019.01.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Polinski KJ, Robinson SL, Putnick DL, Guan W, Gleason JL, Mumford SL, et al. Epigenetic gestational age and the relationship with developmental milestones in early childhood. Hum Mol Genet. 2023;32(9):1565–74. doi: 10.1093/hmg/ddac302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Pelegí-Sisó D, de Prado P, Ronkainen J, Bustamante M, González JR. methylclock: a Bioconductor package to estimate DNA methylation age. Bioinformatics. 2021;37(12):1759–60. doi: 10.1093/bioinformatics/btaa825. [DOI] [PubMed] [Google Scholar]
  • 50.Zhang Q, Vallerga CL, Walker RM, Lin T, Henders AK, Montgomery GW, et al. Improved precision of epigenetic clock estimates across tissues and its implication for biological ageing. Genome Med. 2019;11(1):54. Epub 20190823. doi: 10.1186/s13073-019-0667-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Haftorn KL, Lee Y, Denault WRP, Page CM, Nustad HE, Lyle R, et al. An EPIC predictor of gestational age and its application to newborns conceived by assisted reproductive technologies. Clin Epigenetics. 2021;13(1):82. Epub 20210419. doi: 10.1186/s13148-021-01055-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Casas-Recasens S, Cassim R, Mendoza N, Agusti A, Lodge C, Li S, et al. Epigenome-Wide Association Studies of Chronic Obstructive Pulmonary Disease and Lung Function: A Systematic Review. Am J Respir Crit Care Med. 2024;210(6):766–78. doi: 10.1164/rccm.202302-0231OC. [DOI] [PubMed] [Google Scholar]
  • 53.Recto K, Kachroo P, Huan T, Van Den Berg D, Lee GY, Bui H, et al. Epigenome-wide DNA methylation association study of circulating IgE levels identifies novel targets for asthma. EBioMedicine. 2023;95:104758. Epub 20230818. doi: 10.1016/j.ebiom.2023.104758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Alfaro-Arnedo E, López IP, Piñeiro-Hermida S, Ucero Á C, González-Barcala FJ, Salgado FJ, et al. IGF1R as a Potential Pharmacological Target in Allergic Asthma. Biomedicines. 2021;9(8). Epub 20210729. doi: 10.3390/biomedicines9080912. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Lee CG, Ma B, Takyar S, Ahangari F, Delacruz C, He CH, et al. Studies of vascular endothelial growth factor in asthma and chronic obstructive pulmonary disease. Proc Am Thorac Soc. 2011;8(6):512–5. doi: 10.1513/pats.201102-018MW. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Kachroo P, Morrow JD, Vyhlidal CA, Gaedigk R, Silverman EK, Weiss ST, et al. DNA methylation perturbations may link altered development and aging in the lung. Aging. 2021;13(2):1742–64. doi: 10.18632/aging.202544. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Sharifi-Zarchi A, Gerovska D, Adachi K, Totonchi M, Pezeshk H, Taft RJ, et al. DNA methylation regulates discrimination of enhancers from promoters through a H3K4me1-H3K4me3 seesaw mechanism. BMC Genomics. 2017;18(1):964. Epub 20171212. doi: 10.1186/s12864-017-4353-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Clifford RL, John AE, Brightling CE, Knox AJ. Abnormal histone methylation is responsible for increased vascular endothelial growth factor 165a secretion from airway smooth muscle cells in asthma. J Immunol. 2012;189(2):819–31. Epub 20120611. doi: 10.4049/jimmunol.1103641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Gan PXL, Zhang S, Fred Wong WS. Targeting reprogrammed metabolism as a therapeutic approach for respiratory diseases. Biochem Pharmacol. 2024;228:116187. Epub 20240330. doi: 10.1016/j.bcp.2024.116187. [DOI] [PubMed] [Google Scholar]
  • 60.Hartsoe P, Holguin F, Chu HW. Mitochondrial Dysfunction and Metabolic Reprogramming in Obesity and Asthma. Int J Mol Sci. 2024;25(5). Epub 20240303. doi: 10.3390/ijms25052944. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Kitagawa A, Jacob C, Jordan A, Waddell I, McMurtry IF, Gupte SA. Inhibition of Glucose-6-Phosphate Dehydrogenase Activity Attenuates Right Ventricle Pressure and Hypertrophy Elicited by VEGFR Inhibitor + Hypoxia. J Pharmacol Exp Ther. 2021;377(2):284–92. Epub 20210323. doi: 10.1124/jpet.120.000166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Chen X, Lin H, Yang D, Xu W, Liu G, Liu X, et al. Early-life undernutrition reprograms CD4(+) T-cell glycolysis and epigenetics to facilitate asthma. J Allergy Clin Immunol. 2019;143(6):2038–51.e12. Epub 20190115. doi: 10.1016/j.jaci.2018.12.999. [DOI] [PubMed] [Google Scholar]
  • 63.Koopmans T, Gosens R. Revisiting asthma therapeutics: focus on WNT signal transduction. Drug Discov Today. 2018;23(1):49–62. Epub 20170907. doi: 10.1016/j.drudis.2017.09.001. [DOI] [PubMed] [Google Scholar]
  • 64.Tang W, Li M, Yangzhong X, Zhang X, Zu A, Hou Y, et al. Hippo signaling pathway and respiratory diseases. Cell Death Discov. 2022;8(1):213. Epub 20220420. doi: 10.1038/s41420-022-01020-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Jia XX, Zhu TT, Huang Y, Zeng XX, Zhang H, Zhang WX. Wnt/β-catenin signaling pathway regulates asthma airway remodeling by influencing the expression of c-Myc and cyclin D1 via the p38 MAPK-dependent pathway. Exp Ther Med. 2019;18(5):3431–8. Epub 20190909. doi: 10.3892/etm.2019.7991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Kachroo P, Morrow JD, Kho AT, Vyhlidal CA, Silverman EK, Weiss ST, et al. Co-methylation analysis in lung tissue identifies pathways for fetal origins of COPD. The European respiratory journal. 2020. doi: 10.1183/13993003.02347-2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Hachim MY, Elemam NM, Ramakrishnan RK, Bajbouj K, Olivenstein R, Hachim IY, et al. Wnt Signaling Is Deranged in Asthmatic Bronchial Epithelium and Fibroblasts. Front Cell Dev Biol. 2021;9:641404. Epub 20210315. doi: 10.3389/fcell.2021.641404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Kliem CV, Schaub B. The role of regulatory B cells in immune regulation and childhood allergic asthma. Mol Cell Pediatr. 2024;11(1):1. Epub 20240104. doi: 10.1186/s40348-023-00174-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Stefanowicz D, Hackett TL, Garmaroudi FS, Günther OP, Neumann S, Sutanto EN, et al. DNA methylation profiles of airway epithelial cells and PBMCs from healthy, atopic and asthmatic children. PLoS One. 2012;7(9):e44213. Epub 20120906. doi: 10.1371/journal.pone.0044213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Huang Q, Han L, Liu Y, Wang C, Duan D, Lu N, et al. Elevation of PTPN1 promoter methylation is a significant risk factor of type 2 diabetes in the Chinese population. Exp Ther Med. 2017;14(4):2976–82. Epub 20170811. doi: 10.3892/etm.2017.4924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.Davies ER, Perotin JM, Kelly JFC, Djukanovic R, Davies DE, Haitchi HM. Involvement of the epidermal growth factor receptor in IL-13-mediated corticosteroid-resistant airway inflammation. Clin Exp Allergy. 2020;50(6):672–86. Epub 20200309. doi: 10.1111/cea.13591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Pretorius E, Wallner B, Marx J. Cortisol resistance in conditions such as asthma and the involvement of 11beta-HSD-2: a hypothesis. Horm Metab Res. 2006;38(6):368–76. doi: 10.1055/s-2006-944530. [DOI] [PubMed] [Google Scholar]
  • 73.Wheatley LM, Holloway JW, Svanes C, Sears MR, Breton C, Fedulov AV, et al. The role of epigenetics in multi-generational transmission of asthma: An NIAID workshop report-based narrative review. Clin Exp Allergy. 2022;52(11):1264–75. Epub 20221006. doi: 10.1111/cea.14223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74.Shah R, Newcomb DC. Sex Bias in Asthma Prevalence and Pathogenesis. Front Immunol. 2018;9:2997. Epub 20181218. doi: 10.3389/fimmu.2018.02997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75.Han L, Zhang H, Kaushal A, Rezwan FI, Kadalayil L, Karmaus W, et al. Changes in DNA methylation from pre- to post-adolescence are associated with pubertal exposures. Clin Epigenetics. 2019;11(1):176. Epub 20191202. doi: 10.1186/s13148-019-0780-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 76.Moynihan B, Tolloczko B, Michoud MC, Tamaoka M, Ferraro P, Martin JG. MAP kinases mediate interleukin-13 effects on calcium signaling in human airway smooth muscle cells. Am J Physiol Lung Cell Mol Physiol. 2008;295(1):L171–7. Epub 20080425. doi: 10.1152/ajplung.00457.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 77.Shi Y, Fu X, Cao Q, Mao Z, Chen Y, Sun Y, et al. Overexpression of miR-155–5p Inhibits the Proliferation and Migration of IL-13-Induced Human Bronchial Smooth Muscle Cells by Suppressing TGF-β-Activated Kinase 1/MAP3K7-Binding Protein 2. Allergy Asthma Immunol Res. 2018;10(3):260–7. doi: 10.4168/aair.2018.10.3.260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 78.Ochieng P, Nath S, Macarulay R, Eden E, Dabo A, Campos M, et al. Phospholipid transfer protein and alpha-1 antitrypsin regulate Hck kinase activity during neutrophil degranulation. Sci Rep. 2018;8(1):15394. Epub 20181018. doi: 10.1038/s41598-018-33851-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 79.Zhu J, Jiang Q, Gao S, Xia Q, Zhang H, Liu B, et al. IL20Rb aggravates pulmonary fibrosis through enhancing bone marrow derived profibrotic macrophage activation. Pharmacol Res. 2024;203:107178. Epub 20240405. doi: 10.1016/j.phrs.2024.107178. [DOI] [PubMed] [Google Scholar]
  • 80.Yu G, Tzouvelekis A, Wang R, Herazo-Maya JD, Ibarra GH, Srivastava A, et al. Thyroid hormone inhibits lung fibrosis in mice by improving epithelial mitochondrial function. Nat Med. 2018;24(1):39–49. Epub 20171204. doi: 10.1038/nm.4447. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 81.Hayden LP, Hobbs BD, Busch R, Cho MH, Liu M, Lopes-Ramos CM, et al. X chromosome associations with chronic obstructive pulmonary disease and related phenotypes: an X chromosome-wide association study. Respir Res. 2023;24(1):38. Epub 20230201. doi: 10.1186/s12931-023-02337-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 82.Camila ML- R, Cho-Yi C, Marieke LK, Joseph NP, Abhijeet RS, Maud F, et al. Sex Differences in Gene Expression and Regulatory Networks across 29 Human Tissues. Cell Reports, ISSN: 2211–1247. 2020;31(12):107795-. doi: 10.1016/j.celrep.2020.107795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 83.Dessie EY, Ding L, Mersha TB. Integrative analysis identifies gene signatures mediating the effect of DNA methylation on asthma severity and lung function. Clin Epigenetics. 2024;16(1):15. Epub 20240120. doi: 10.1186/s13148-023-01611-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 84.Harb H, Stephen-Victor E, Crestani E, Benamar M, Massoud A, Cui Y, et al. A regulatory T cell Notch4-GDF15 axis licenses tissue inflammation in asthma. Nat Immunol. 2020;21(11):1359–70. Epub 20200914. doi: 10.1038/s41590-020-0777-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 85.Gauthier M, Kale SL, Oriss TB, Scholl K, Das S, Yuan H, et al. Dual role for CXCR3 and CCR5 in asthmatic type 1 inflammation. J Allergy Clin Immunol. 2022;149(1):113–24.e7. Epub 20210616. doi: 10.1016/j.jaci.2021.05.044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 86.Shi Y, Simpson S, Ahmed SK, Chen Y, Tavakoli-Tameh A, Janaka SK, et al. The Antiviral Factor SERINC5 Impairs the Expression of Non-Self-DNA. Viruses. 2023;15(9). Epub 20230920. doi: 10.3390/v15091961. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 87.Obaid Abdullah S, Ramadan GM, Makki Al-Hindy HA, Mousa MJ, Al-Mumin A, Jihad S, et al. Serum Myeloperoxidase as a Biomarker of Asthma Severity Among Adults: A Case Control Study. Rep Biochem Mol Biol. 2022;11(1):182–9. doi: 10.52547/rbmb.11.1.182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 88.Chiappara G, Chanez P, Bruno A, Pace E, Pompeo F, Bousquet J, et al. Variable p-CREB expression depicts different asthma phenotypes. Allergy. 2007;62(7):787–94. doi: 10.1111/j.1398-9995.2007.01417.x. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1
media-1.docx (1.5MB, docx)
Supplement 2
media-2.xlsx (18.4MB, xlsx)

Data Availability Statement

All TOPMed data is person-sensitive, however it can be requested for access and can be made available through the TOPMed consortium after careful review and approval by the TOPMed Data Access Committee (https://topmed.nhlbi.nih.gov/). Participant consent and Data Use Limitations differs within and across TOPMed studies and should be requested individually. Additional documentation, such as of local IRB approval and/or letters of collaboration with the primary study PI(s) may be required.

The CAMP DNA methylation datasets analyzed in the current study are available at the database of Genotypes and Phenotypes (dbGaP) repository (phs001726. v2. p1) here: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs001726.v2.p1. The CRA DNA methylation datasets analyzed in the current study are available at the dbGaP repository (phs000988. v5. p1) here: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000988.v5.p1.

All datasets used and/or analyzed during the current study could be requested from the corresponding author and contact study PIs and made available on reasonable request.

Articles from bioRxiv are provided here courtesy of Cold Spring Harbor Laboratory Preprints

RESOURCES