TL;DR
This study reveals that smoking leaves a systemic "epigenetic signature" in adipose tissue, affecting both DNA methylation and gene expression. These changes are not only widespread across tissues but also linked to cardio-metabolic health risks and can even predict future weight gain after quitting.
What Problem Does This Paper Address?
While smoking's detrimental effects on health are well-established, and its impact on DNA methylation in blood is known, the extent to which these changes extend to metabolically relevant tissues like adipose tissue and influence gene expression and cardio-metabolic health has remained unclear. This paper aims to bridge that gap by comprehensively investigating the coordinated effects of smoking on DNA methylation and gene expression in adipose tissue and their subsequent impact on metabolic health.
Methods
The researchers employed a multi-pronged approach:
- Genome-wide Association Studies (EWAS & TWAS): They conducted epigenome-wide association studies (EWAS) for DNA methylation and transcriptome-wide association studies (TWAS) for gene expression in adipose tissue samples.
- Replication and Validation: Findings were replicated in independent cohorts and validated across multiple tissues (blood, skin, lung) to identify tissue-shared effects.
- Functional Analysis: They explored the association of smoking-related epigenetic changes with cardio-metabolic phenotypes (visceral fat, BMI, etc.) and investigated the predictive power of these changes for future metabolic health outcomes.
- Smoking Cessation Analysis: The study examined the reversibility of these epigenetic changes after smoking cessation.
- Genetic Influence Analysis: They investigated the impact of genetic variation on smoking-associated methylation signals.
Datasets
The primary dataset comprised adipose tissue samples from 542 female twins from the TwinsUK cohort (part of the MuTHER study).
Replication and validation datasets included:
- LEAP cohort: 104 participants (USA) for smoking-DMS replication.
- Finnish twins: 69 individuals for cardio-metabolic phenotype analysis.
- Lung cancer subjects: 168 individuals for lung tissue validation.
- Skin tissue samples: 195 individuals for skin tissue validation.
- Peripheral blood samples: from 789 and 362 subjects for tissue-shared effects analysis.
Key Insights and Ideas
- Systemic Epigenetic Footprint: Smoking doesn't just affect the lungs; it leaves a coordinated "epigenetic signature" (changes in DNA methylation and gene expression) across multiple tissues, including metabolically relevant adipose tissue.
- Adipose Tissue as a Key Player: Adipose tissue is not just a passive bystander but actively responds to smoking, with changes that are linked to metabolic health.
- Methylation vs. Expression Dynamics: DNA methylation changes appear to be more long-lasting after smoking cessation, while gene expression changes are stronger but shorter-lived.
- Predictive Power for Future Health: Specific smoking-associated epigenetic markers can predict future weight gain and cardio-metabolic disease risk, even in individuals who quit smoking.
- Genetic Mediation: Genetic variations influencing smoking behavior and nicotine metabolism may exert their effects, at least in part, through DNA methylation.
Key Results
- Identified 42 smoking-associated DNA methylation signals (DMS) and 42 smoking-associated gene expression signals (DES) in adipose tissue.
- Found 5 genes (AHRR, CYP1A1, CYP1B1, CYTL1, F2RL3) with coordinated hypo-methylation and up-regulation in smokers.
- Demonstrated tissue-shared effects for a proportion of these signals across adipose, blood, skin, and lung tissues.
- Observed that smoking-DMS in CYP1A1 and NOTCH1 were significantly associated with cardio-metabolic risk factors like visceral fat mass.
- Showed that BHLHE40 and AHRR methylation and expression levels in current smokers could predict future visceral fat gain after smoking cessation.
- Replication studies confirmed the consistency of these findings.
Limitations and Future Work
- Partial Covariate Correction: The study acknowledges limitations in fully accounting for all potential covariates, particularly alcohol consumption, which co-occurs with smoking.
- Cell Composition Heterogeneity: While efforts were made to account for cell type heterogeneity in adipose tissue, the authors note that this might not fully capture the effect of a heterogeneous cell population as a confounder. Some signals might be cell-subtype specific.
- Modest Sample Size for Quitting Smokers: The sample size of current smokers who subsequently quit was modest, suggesting the need for replication in larger cohorts to confirm predictive findings.
- Future Work: The authors suggest further investigation into the mechanisms underlying these smoking-induced epigenetic changes and their role in mediating environmental effects on cardio-metabolic disease risk. They also highlight the need for future studies to assess potential confounding effects, such as cell-specific methylation profiles.
Summary updated