Our study identified essential epigenetic markers pinpointing CpG regions likely to impact COPD more than PRISm significantly. Although previously well-established COPD-associated genes were not identified in our study, significant methylation changes associated with COPD were observed. Importantly, we can cross-reference with RNA sequencing results to identify the functional impact of methylation-induced changes. Further validation in the KoGES cohort, representing a general population, confirmed the consistency of our findings between PRISm and COPD groups, thereby reinforcing the significance of our results.

Smoking is a crucial etiological factor in COPD, and it is worth highlighting that smoking can trigger cellular stress and has been associated with various DNA methylation patterns. This connection between smoking, cellular stress, and DNA methylation patterns plays a significant role in understanding the development and progression of COPD (Eriksson Strom et al. 2022). DNA methylation profiles in peripheral blood have emerged as potential biomarkers for assessing disease risk, and these profiles hold promise for translation into clinical applications. Previous studies on DNA methylation, particularly in blood samples, have reported hypomethylation in regions associated with COPD, such as SERPINA1, FUT7, DIP2C, and the rare variant rs140198372 of SERPINA12 (He et al. 2020). In addition to COPD, DNA methylation analysis has also been reported for PRISm and COPD in a paper published by A. Domingo-Relloso et al. In this paper, grouping was conducted into ‘Airflow limitation (FEV1/FVC < 70)’ and ‘restriction (FEV1/FVC > 70, FVC < 80%) ‘. Next, DNA methylation analysis was performed in blood, and FEV1, FEV1/FVC, and airflow limitation were analyzed as outcome variables, and duplicate genes were organized. They selected EGFR, MAPK1, and PRPF8 genes. There were no genes that overlapped with our results. To clarify the reason for this result, we defined the PRISm and COPD groups by FEV1/FVC ratio and FEV1 and selected genes by comparing DMPs and DEGs between each group, so it appears that previously reported genes did not show significance (Domingo-Relloso et al. 2022). Referring to these results, we also conducted a study to identify epigenetic CpG sites using peripheral blood samples.

This study identified seven genes as epigenetic site regions, including those exhibiting differential DNA methylation patterns and gene expression changes (DEGs). These seven genes—EEF1A2, EMP2, EPCAM, MTSS1L, ARHGEF10, HYDIN, and FADS2—are pivotal in transitioning from PRISm to COPD. Furthermore, these genes were recognized as DMPs and DEGs closely associated with COPD and the progression of lung-related diseases.

Specifically, we observed CpG sites in COPD patients, cg11878698 and cg03074225 located in the body region of the EEF1A2 gene were hypermethylated, but cg26186239 located in the 3’UTR was hypomethylated. The CpG sites selected in ARHGEF10 and HYDIN were hypomethylated, and the CpG sites located in the 5’UTR region of the EPCAM gene were confirmed to be hypomethylated. In the EMP2 gene, all CpG sites in the TSS1500 region were hypermethylated, and CpG sites in the 5’UTR were hypomethylated.

EEF1A is involved in translational processes and displays non-canonical functions. EEF1A also plays a role in viral infections, interacting with viral proteins to enhance viral replication, reduce apoptosis, and increase cell transformation (Wefers, et al. 2022). Viral infections are a common and significant factor in COPD exacerbations, responsible for 50–66% of cases. Approximately 25% of exacerbations involve viral and bacterial infections, often leading to greater severity (Frickmann et al. 2012). While not universally confirmed, the connection between COPD and viral infections suggests a plausible link. EEF1A2 is a previously unreported gene in open databases, potentially holding novel insights for COPD research.

EMP2, a member of the tetraspan superfamily of membrane proteins, is thought to regulate cell adhesion and signaling (Lin et al. 2020; Wadehra et al. 2003). It has been implicated in various cellular functions in cancer cells. EMP2 is highly expressed in the lungs and may affect polymorphonuclear leukocyte entry into the airspace. Notably, EMP2 has been negatively correlated with other COPD-related genes in genome-wide association studies (GWAS), explaining its decreased expression in COPD patients (Lin et al. 2020). Indeed, while the exact relationship between the DEGs and the specific CpG sites may be challenging to elucidate, their combined findings suggest a consistent pattern of reduced expression of the EMP2 gene in individuals with COPD.

EPCAM, an evolutionarily conserved type 1 transmembrane glycoprotein, is expressed in epithelial cells of various tissues, including the lungs. Its primary function is the regulation of cell–cell interactions and the modulation of cell adhesion molecules. EPCAM has also been identified as a biomarker for stem-/progenitor-like cells in the human lung (Vries et al. 2022; Huang et al. 2018; Wang et al. 2020). While EPCAM displayed differential methylation patterns based on gender, age, and smoking, it did not reach statistical significance in our analysis.

ARHGEF10 plays a crucial role in cell signaling events as a member of the Rho GTPase family, which regulates the actin cytoskeleton, influencing various cellular processes such as morphological changes, migration, and cytokinesis across different cell types (Khan et al. 2021). It has been widely expressed in various tissues, notably high spinal cord, and dorsal root ganglion expression levels. ARHGEF10 is involved in neuronal growth and axonal guidance processes and has implications for human hypomyelination (Niftullayev and Lamarche-Vane 2019). Our findings align with previous reports, as reduced expression of ARHGEF10 has also been documented in individuals with COPD (Singh et al. 2014).

FADS2 is primarily associated with cholesterol, sphingomyelin, phospholipids, and fatty acid metabolism, with implications in respiratory diseases like asthma, COPD, and pneumonia. Abnormal FADS2 expression disrupts cell membrane phospholipid balance, affecting membrane fluidity and signal transmission, leading to proinflammatory factor production and arachidonic acid metabolite generation. Increased glucose uptake in adipose tissues has also been associated with FADS2 expression. As metabolic syndrome worsens inflammatory conditions, there may be an association between decreased FADS2 expression and COPD progression (Li et al. 2020; Fujii, et al. 2021).

HYDIN is mainly studied in lung-related diseases, particularly in the context of respiratory infections. Mutations in HYDIN can cause primary ciliary dyskinesia, a genetic disease characterized by chronic respiratory infections, highlighting its potential importance in COPD, where respiratory infections play a critical role (Xu et al. 2022; Cindrić et al. 2020). HYDIN, relatively unexplored in previous research, emerged as a significant gene in our analysis. This gene suggests that HYDIN may have disease specificity or relevance to COPD beyond the infection-induced COPD reported in previous studies (Xu et al. 2022; Cindrić et al. 2020).

MTSS1L has been linked to neonatal respiratory distress syndrome (Bhattacharya et al. 2021). While its precise role remains unclear, our findings suggest its association with COPD, especially considering the measurement of DNA methylation expression in children with the disease and the reported connection between bronchopulmonary dysplasia and COPD (Priyadarshini et al. 2020).

It revealed correlations with various factors related to COPD, including negative regulation of gonadotropin secretion, steroid catabolic processes, and positive regulation of tau-protein kinase activity (Laghi et al. 2009). These findings align with previous reports of associations between COPD and endocrine hormone disorders, steroid metabolism, and the potential involvement of tau-protein kinase activity. Glucocorticoids and TNF-α have been linked to gonadotropins, suggesting a connection between these factors and COPD. Steroid catabolism, specific to COPD, can reflect the disease’s severity, and tau-protein kinase activity may be associated with COPD, particularly considering the involvement of transcription factor EB (TFEB), known to be linked to COPD (Bodas et al. 2017; Creutzberg and Casaburi 2003; Bodas and Vij 2017).

Continuous exposure to harmful substances in cigarette smoke or e-cigarette/nicotine vapor induces inflammatory oxidative stress in the airways, leading to irreversible lung damage (Creutzberg and Casaburi 2003). This may explain the correlation between disease aggravation through smoking or external stress and the identified gene functions.

Given the increasing importance of PRISm, as defined by the GOLD criteria, and its association with COPD, additional studies focusing explicitly on PRISm are warranted. According to W. Sara Renata Alex et al., approximately one-third of PRISm patients progress to COPD (Alex, et al. 2020). PRISm patients experience varying degrees of lung function decline over time. While some maintain PRISm status, others transition to normal lung function or develop COPD. Unlike COPD, PRISm is characterized by a restrictive pattern on pulmonary function tests, with preserved FEV1/FVC ratios. PRISm patients tend to be younger, female and have less smoking history compared to COPD patients. Lung parenchymal diseases, rather than airway diseases, are common in PRISm, such as tuberculous traces, bronchiectasis, and mild interstitial lung disease. Although PRISm patients transitioning to COPD have risk factors similar to those for COPD, such as old age, smoking, and asthma history, they exhibit a slower decline in lung function compared to COPD-only patients but faster than normal individuals. The presence of multiple COPD risk factors in PRISm patients, along with severe emphysema and small airway dysfunction, contributes to their progression toward COPD. Comparing PRISm and COPD, identifying genes that show differences, and investigating genetic factors related to COPD can help reveal the differences, relationships, and associations between COPD and PRISm. Additionally, comprehensive studies investigating the mechanisms underlying symptom progression associated with the up or down-regulation of these genes are essential. The present study has significant implications in investigating the mechanisms underlying symptom progression associated with the up or down-regulation of the gene.

This study is subject to several limitations. Firstly, it included relatively small sample size, with only 572 selected DNA methylation and 60 RNA sequencing data from the KOCOSS-NIH registry and 98 selected DNA methylation patients from the KoGES cohort for the replication study, potentially introducing bias. The DNA methylation data was also derived from whole blood samples, which may not fully capture the tissue-specific DNA methylation patterns relevant to lung diseases such as COPD and PRISm. Studies demonstrating a relationship between DNA methylation and lung disease derived from blood samples have been reported despite this (Domingo-Relloso et al. 2022; Mulder et al. 2021; Robert, et al. 2023).

The key strength of this study lies in its comprehensive observational data sourced from the KOCOSS-NIH registry, which specifically targeted Koreans for COPD research. Moreover, the genes identified through analysis of this registry exhibited consistent patterns when evaluated in the KoGES cohort, enhancing the robustness of the gene selection approach.

However, this study was conducted in Koreans and has racial limitations. In addition, EWAS and RNA sequencing analyses were conducted in some patients in KOCOSS-NIH, and DNA methylation data were produced from 2019 to 2021 to refine the data comprehensively. It’s important to note that the number of patients included in the analysis is small. However, the genes selected from the KOCOSS-NIH registry showed similar patterns in KoGES, a cohort of different characteristics, confirming that these genes are associated with COPD specific to Koreans. This study’s findings could have a significant impact on future research and the understanding of COPD.