SPS2-01 Epigenome-wide Association Study of Endotoxin Exposure and Respiratory Outcomes in Textile Workers

Rationale

Textile manufacturing employs approximately 430 million people and is among the largest traditional industries worldwide, particularly in low- and middle- income economies. Endotoxin is the primary bacterial toxicant in cotton dust found in textile manufacturing and has been linked to respiratory diseases and declines in pulmonary function. However, little is known about DNA methylation (DNAm) changes, molecular pathways, and impacts on respiratory outcomes associated with endotoxin exposure. To investigate further, we conducted an epigenome-wide association study (EWAS) in textile workers with occupational exposure to endotoxin.

Methods

The Shanghai Textile Worker Study has followed 919 cotton and silk workers since 1981, with repeated assessments of on-site endotoxin exposure, spirometry, and symptom questionnaires. The primary exposure of interest is ever exposed to occupational endotoxin. Endotoxin-free silk workers were recruited as controls. Blood samples were collected in 1996 from 281 cotton and 253 silk workers and assayed for DNAm using the Illumina EPICv2 (940k) array. Multivariable linear mixed-effects regression models were used to identify differentially methylated positions (DMPs) associated with occupational endotoxin exposure with nested random intercepts to account for technical variables. CpGs below the epigenome-wide significance threshold of p < 5 × 10⁻⁸ were enriched for KEGG pathways using an over-representation analysis (ORA) approach. Associations between significant CpGs and selected respiratory outcomes were also evaluated.

Results

We identified 36 CpGs significantly associated with occupational endotoxin exposure. The top signals when comparing cotton to silk workers were cg22740441(PPP3CA, p = 4.76 × 10⁻11) and cg11248361(AQP4-AS1, p = 1.75 × 10-10). The immune-related “antigen processing and presentation” (KEGG: 04612) was the top enriched pathway. Fifteen CpGs were associated with years since exposure cessation, suggesting that the effect of endotoxin on DNAm could be reversible over time. With respect to respiratory outcomes, we identified three CpGs associated with forced expiratory volume in 1 second (FEV-1) (cg26491213 [SIX3], cg00495557[lncRNA coding gene related to genomic instability], and cg11248361[AQP4-AS1]), 12 associated with self-reported dyspnea (e.g., cg05408620 [FSTL1]), and 17 associated with self-reported chronic cough (e.g., cg05408620 [FSTL1]). Our results were largely consistent in sensitivity analyses stratified by sex (males and females), smoking status (smokers and non-smokers), and alternative exposure assessment (cumulative endotoxin concentration and cumulative dust concentration).

Conclusion

We found evidence that occupational exposure to endotoxin is associated with alterations in DNAm with biological pathways likely related to immune function. These DNAm alteration may be partially reversible after exposure cessation and play a mechanistic role in respiratory outcomes among endotoxin-exposed populations.

This abstract is funded by: None