Abstract:
To investigate the impact of time varying nature of DNA methylation (DNAm) on the progression of chronic obstructive pulmonary disease
(COPD), we conducted a longitudinal epigenome-wide association study (EWAS). The EWAS findings were integrated with proteomic data to reveal
the interplay between methylation and proteomics in COPD. Methods: Infinium MethylationEPIC BeadChip data was obtained from peripheral
blood leukocyte DNA from 4,493 current and former smokers from the Genetic Epidemiology of COPD (COPDGene) study consisting of two time
points of DNAm and FEV1 data (baseline and 5-year follow-up visits) of non-Hispanic white and African American subjects with a history of
smoking, a median age of 62.2 years at enrollment and 51 percent male subjects. SomaScan blood proteomic data was available from the 5-year
follow-up visit. Linear mixed-e"ects regression (LMER) was used to find the associations between longitudinal DNAm and longitudinal forced
expiratory volume in 1 second (FEV1); adjusted for sex, age, race, cell counts, pack-years of smoking, current smoking status, height and time
since baseline visit. Causal mediation analysis to identify proteins potentially mediating the relationship between DNAm and FEV1 was conducted
using a nonparametric bootstrap procedure from the R package Mediation. Functional overlap analysis on the genome-wide significant CpGs was

performed using the eFORGE tool. Results: From the longitudinal FEV1 and longitudinal DNAm data, LMER identified 66 CpGs having genome-
wide significant associations with FEV1 (p ≤ 5 x 10-8). Intraclass correlation (ICC) of the 66 CpGs calculated using DNAm M-values from both

time points revealed that the majority (40) have ICC between 0.5 and 0.8 and 22 CpGs have ICC greater than 0.8, suggesting that FEV1-
associated CpGs are stable over 5 years. Chromosome 17 demonstrated three significant associations over a 440 base pair region annotated to
the SOCS3 gene (cg13343932 with p = 2.12 x 10-18, cg11047325 with p = 8.77 x 10-16 and cg18181703 with p = 1.07 x 10-13). An X
chromosome CpG near the zinc-finger repressor, BCOR, was also associated with longitudinal FEV1 (cg18422972 with p = 2.14 x10-8).
Functional Overlap analysis of the 66 CpGs revealed significant enrichment in histone methylation (H3K4me1 and H3K4me3) in blood and lung
tissue. Mediation analysis identified ITPRIPL1, NOTCH1 and IGFBP5 as three candidate proteins partially mediating the signal between DNAm and
FEV1. Conclusions: EWAS of longitudinal DNAm data revealed that the CpG sites that are stable over time are associated with longitudinal lung
function (FEV1) and are linked with epigenetic and inflammatory signals.

RATIONALE: Weight loss (WL) in chronic obstructive pulmonary disease (COPD) is associated with increased mortality. Identifying genetic variants associated with WL in COPD may help identify causal pathways as possible therapeutic intervention targets. Our prior genome-wide association study of WL in COPD utilized affinity chip data, but whole genome sequencing (WGS) data is now available for analysis, yielding opportunities for increased cohort size and genomic resolution.

METHODS: We utilized data from seven studies within the Trans-Omics for Precision Medicine initiative: the Cardiovascular Health Study (CHS); Genetic Epidemiology of COPD (COPDGene); Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-points (ECLIPSE); Framingham Heart Study (FHS); Jackson Heart Study (JHS); Multi-Ethnicity Study of Atherosclerosis (MESA); and the SubPopulations and InteRmediate Outcome Measures In COPD Study (SPIROMICS). All analyses were conducted in 8,628 participants with COPD, where COPD was characterized using spirometry and smoking history. Of these participants, 29.6% had WL of 5% or more between study visits or a body mass index (BMI) of less than 20 kg/m2. WGS analyses, adjusted for age, sex, and principal components of ancestry, were performed using the GENESIS BioData Catalyst pipeline, with participants of African-American (AA) and non-Hispanic white (NHW) ancestries analyzed separately. Fixed-effects meta-analysis was performed using METAL.

RESULTS: Across nine ancestry- and study-stratified cohorts of research participants with COPD, we identified eight variants that reached significant association with WL in COPD across five distinct loci. In COPDGene NHW participants with COPD, two chromosome 12 variants intergenic to LINC02426 and CCDC59 were associated with increased WL risk (OR = 5.74, p = 2.0 x 10-8 for both). In ECLIPSE, two intronic variants in SEPTIN7 on chromosome 7 were associated with increased WL risk (OR = 5.35, p = 3.3x10-8 for both). In FHS, one 3’ untranslated region chromosome 1 variant in NR5A2 was significantly associated with WL (OR = 0.136, p = 3.0x10-8). In JHS, one chromosome 2 variant between DOCK10 and NYAP2 was associated with increased WL risk (OR = 15.5, p = 7.7x10-9). Finally, in CHS, two chromosome 3 variants intronic to HHATL were associated with increased WL risk (OR 4.4, p = 2.7x10-8 and OR = 5.1 and p = 3.2x10-8). None of the identified associations remained significant after meta-analysis.

DISCUSSION: We identified five novel loci associated with unintentional weight loss in COPD. Additional efforts are being directed at fine mapping and replication analyses in larger biobanks with genomic data.

Background
Chronic obstructive pulmonary disease (COPD) is associated with age and smoking, but other determinants of the disease are incompletely understood. Clonal hematopoiesis of indeterminate potential (CHIP) is a common, age-related state in which somatic mutations in clonal blood populations induce aberrant inflammatory responses. Patients with CHIP have an elevated risk for cardiovascular disease, but the association with COPD remains unclear.

Methods
We analyzed whole-genome and exome sequencing data to detect the presence of CHIP in 48,835 subjects, of whom 8,444 had moderate-to-very-severe COPD. Cases of CHIP were defined based on somatic mutations detected using the sequencing data that are in a pre-specified list of variants predicted or reported to be pathogenic and drivers of myeloid malignancies. The subjects were selected from four separate cohorts with COPD phenotyping and smoking history, including COPDGene, additional TOPMed cohorts, ICGN-EOCOPD study, and UK Biobank. We applied logistic regression and generalized linear mixed effect models to COPD status adjusting for age, gender, sequencing center, number of pack-years, smoking status, and genetic ancestry. Random-effects meta-analyses were conducted to combine the cohorts. As TET2 was commonly mutated in hematopoietic cells, we measured emphysema in murine models in which Tet2 was deleted in hematopoietic cells and performed single cell RNA-sequencing in lung tissues of the Tet2-wildtype and Te2-knockout mice.

Results
In COPDGene, individuals with CHIP had a risk of moderate-to-severe and severe or very severe COPD 1.6 and 2.2 times greater than non-carriers, respectively (adjusted 95% confidence intervals [CI], 1.1 to 2.2 and 1.5 to 3.2). These findings were consistently observed in three additional cohorts and meta-analyses of all subjects. CHIP was also associated with decreased FEV1% predicted in COPDGene (mean between group difference -5.7%; adjusted 95% CI, -8.8 to -2.6), a finding replicated in additional cohorts. Smoke exposure was associated with a small but significant increased risk of having CHIP (OR 1.03 per ten pack-years, 95% CI 1.01-1.05) in the meta-analysis of all subjects. In cigarette smoke exposure models, inactivation of Tet2 in mouse hematopoietic cells enhanced pulmonary inflammation, increased interferon signaling, decreased TGF- signaling, and increased emphysema development.

Conclusions
Age-associated somatic mutations in blood cells are associated with the development and severity of COPD, independent of age and cumulative smoke exposure.

Background: Chronic Obstructive Pulmonary Disease (COPD), diagnosed by reduced lung function, is a highly heterogeneous disease. Existing polygenic risk scores (PRS) enable early identification of genetic risk for COPD. However, predictive performance of the PRS is limited when the discovery and target populations are not well matched.
Methods: To improve cross-ethnic portability of risk prediction, we introduced a PrediXcan-derived polygenic transcriptome risk scores (PTRS), developed under the hypothesis that the biological mechanisms of disease are shared across ancestry groups. We constructed the PTRS using summary statistics from application of PrediXcan on large scale GWAS of lung function (Forced Expiratory Volume in 1 second (FEV1) and its ratio to Forced Vital Capacity (FEV1/FVC)) from the UK Biobank, which representing primarily European ancestry-based cohort. To examine prediction performance and cross-ethnic portability of the proposed PTRS candidates, we performed smoking-stratified analyses on multi-ethnic training data for 29,381 participants from TOPMed population/family-based cohorts (NHW=14,727, AA=7,025, HIS=7,629). The best risk score candidates were then tested for 11,771 multi-ethnic participants from TOPMed COPD-enriched studies (NHW=8,144, AA=3,627). Analyses were carried out for two dichotomous traits of COPD (Moderate-to-Severe and Severe COPD) and two quantitative lung function traits (FEV1 and FEV1/FVC).
Results: While the novel PTRS had lower prediction accuracy for European ancestry participants than PRS, the PTRS performed slightly better than PRS for predicting COPD in heavy smoking African Americans (OR=1.24 [95%CI: 1.08-1.43] from PTRS and OR=1.10 [95% CI: 0.96-1.26] from PRS for Moderate-to-Severe COPD; for Severe COPD, OR=1.51 [95%CI: 1.04-12.19] from PTRS and OR=1.31 [95% CI: 0.87-1.96] from PRS). In addition, as hypothesized, the cross-ethnic portability was significantly higher for PTRS than for PRS (p<2.2e-16) for both dichotomous COPD traits and across all smoking strata.
Conclusions: Our study demonstrates the value of PTRS for improved prediction of COPD risk in African Americans. Future work will strengthen the PTRS framework by leveraging multi-ethnic gene expression reference data of larger sample sizes and from disease relevant tissues.

Rationale: Idiopathic pulmonary fibrosis (IPF) is characterized by progressive lung scarring and death. Early detection is paramount. Chest computed tomographic imaging can detect a precursor phenotype known as interstitial lung abnormalities (ILA). The MUC5B promoter variant rs35705950 is a common variant of large effect size that confers risk for both IPF (OR 6-13, heterozygotes) and ILA (OR 2-3). The large effect size of this variant may mask identification of polygenic effects. We hypothesized that a polygenic risk score (PRS) excluding the MUC5B region would add complementary predictive value to the MUC5B variant for IPF and ILA.

Methods: Using previously published genome-wide association summary statistics for IPF, we used lassosum to develop an IPF PRS excluding the MUC5B region (±250 kb). We trained a composite risk score of the PRS and MUC5B rs35705950 (PRS-M5B) using 10-fold cross-validation (100 bootstraps) in IPF cases and controls from the external Lung Tissue Research Consortium (LTRC). We tested the associations of the PRS, rs35705950, and PRS-M5B with IPF in LTRC and ILA in the Genetic Epidemiology of COPD (COPDGene) study. Multivariable logistic regressions were performed, adjusting for age, sex, smoking, and genetic ancestry. Area-under-the-curve (AUC) analyses were used to assess model predictive performances. We examined phenotypic variance explained by each risk factor using Nagelkerke R2, and also compared this measure to a traditional PRS including the MUC5B region.

Results: In LTRC (270 controls, 255 IPF cases), the PRS (OR 4.5 [95% CI: 3.3 - 6.1], p=8.0e-22) and rs35705950 (OR 3.6 [95% CI: 2.4 - 5.2], p=6.4e-11) were both associated with IPF, and when considered in a single model, effect sizes were largely unchanged. The cross-product interaction term (PRS X rs35705950) was not significant (p=0.7). PRS-M5B demonstrated a larger effect (OR 18 [95% CI: 10 – 30], p=5.0e-26) compared to either component risk factor. There was a stepwise increase in predictive performances of models including rs35705950 (AUC 0.74), PRS (AUC 0.83), and PRS-M5B (AUC 0.86). In COPDGene (6417 controls, 250 ILA), rs35705950 (OR 1.8 [95% CI: 1.4-2.4], p=2.7e-6), but not the PRS (p=0.08), were significantly associated with ILA. Phenotypic variability in IPF explained by PRS-M5B (R2=0.41) was greater than PRSs including (R2=0.34) and excluding (R2=0.33) the MUC5B region, or rs35705950 alone (R2=0.12).

Conclusions: The MUC5B rs35705950 variant and a PRS excluding the MUC5B region offered complementary predictive value for IPF, but not ILA. For certain traits, such as IPF, separating out common variants of large effect size may aid in identifying polygenic effects.