TL;DR
This study reveals how smoking profoundly alters DNA methylation and gene expression in adipose tissue, a key metabolic organ. These changes are not only widespread but also linked to metabolic disease risk and can persist long after quitting, offering new insights into smoking's systemic health impacts.
What Problem Does This Paper Address?
Tobacco smoking is a well-established risk factor for numerous diseases, including cardiovascular and metabolic disorders. While previous research has identified smoking-induced changes in DNA methylation, primarily in blood, the extent to which these epigenetic alterations affect metabolically relevant tissues like adipose tissue and their downstream impact on gene expression and metabolic health remained largely unknown. Understanding these changes in adipose tissue is crucial because smoking cessation is often associated with weight gain and increased visceral fat, hinting at deeper molecular mechanisms at play.
Methods
The researchers employed a multi-pronged approach:
- Epigenome-Wide Association Study (EWAS) and Transcriptome-Wide Association Study (TWAS): They analyzed DNA methylation (using Illumina Infinium HumanMethylation450 BeadChip) and gene expression (RNA-sequencing) in adipose tissue biopsies from a large cohort of female twins. This allowed them to identify specific DNA methylation sites (DMS) and differentially expressed genes (DES) associated with smoking.
- Replication and Tissue Specificity: Identified signals were validated in independent cohorts and across different tissues (blood, skin, lung) to determine if smoking effects are shared or tissue-specific.
- Longitudinal Analysis: The study tracked changes in methylation and gene expression in ex-smokers over time to assess the reversibility of smoking-induced epigenetic marks after cessation.
- Phenotypic Association: They linked smoking-associated epigenetic changes to metabolic health phenotypes like visceral fat mass and android-to-gynoid ratio, and explored their predictive power for future weight gain after smoking cessation.
- Genetic Influence: The study investigated how genetic variations (meQTLs) might influence smoking-associated DNA methylation.
Datasets
- Discovery Cohort: 542 healthy female twins from the TwinsUK cohort (54 current smokers, 197 ex-smokers, 291 non-smokers) provided adipose tissue biopsies for DNA methylation and RNA sequencing.
- Replication Cohorts:
- LEAP cohort (USA): 104 individuals (46 smokers, 58 non-smokers) for adipose tissue methylation replication.
- Finnish Twins: 69 individuals for replication of methylation associations with metabolic phenotypes.
- TwinsUK skin tissue: 195 samples.
- Lung cancer tissue: 168 samples from a multicenter cohort.
- Blood samples: from 567 individuals for methylation and 237 for expression profiling.
- Phenotype Data: Dual X-ray absorptiometry (DEXA) measurements for total fat mass, visceral fat mass, and android-to-gynoid fat ratio were collected from a subset of participants. Longitudinal data on adiposity phenotypes were also used to study changes after smoking cessation.
Key Insights and Ideas
- Adipose Tissue is a Key Target: Smoking doesn't just affect the lungs; it leaves a significant epigenetic footprint in adipose tissue, a critical organ for metabolic regulation.
- Coordinated Changes: Smoking induces coordinated changes in both DNA methylation and gene expression, suggesting a cascade of molecular events.
- Tissue-Shared Signatures: Many smoking-associated methylation changes are not confined to one tissue but are "tissue-shared," indicating a systemic impact.
- Gene Expression is More Dynamic: While methylation changes are more stable and longer-lasting, gene expression shows a more rapid reversal after smoking cessation.
- Predictive Power for Metabolic Health: Specific epigenetic markers in adipose tissue can predict future metabolic health outcomes, such as weight gain after quitting smoking.
Key Results
- 42 Smoking-DMS and 42 Smoking-DES: The study identified 42 differentially methylated sites and 42 differentially expressed genes associated with smoking in adipose tissue.
- Five Overlapping Genes: Five genes (AHRR, CYP1A1, CYP1B1, CYTL1, F2RL3) showed coordinated changes in both DNA methylation and gene expression.
- CYP1A1 as a Strong Biomarker: CYP1A1 gene expression alone could predict current smoking status with 95% accuracy.
- Tissue-Shared Signals: 14 CpG sites were found to be genome-wide differentially methylated in both blood and adipose tissue.
- Reversibility Varies: While gene expression largely reversed within a year of quitting, methylation signals showed slower and less complete reversal, with some persisting for decades.
- Metabolic Health Links: Smoking-DMS in CYP1A1 and NOTCH1 were significantly associated with visceral fat mass and android-to-gynoid ratio.
- Predicting Future Weight Gain: Methylation and expression levels at BHLHE40 and AHRR were predictive of future visceral fat gain after smoking cessation.
Limitations and Future Work
- Cellular Heterogeneity: While the study attempted to account for cell composition in adipose tissue, the authors acknowledge that this may not fully capture the effect of a heterogeneous cell population. They suggest future studies could use histological and immunological staining for a more precise analysis.
- Sample Size for Quitting Group: The sample size of current smokers who quit within the study period was modest, suggesting that findings related to future weight gain after cessation require replication in larger cohorts.
- Adipose-Specific Signals: While some signals appeared adipose-specific, the authors note that limitations in available multi-tissue datasets might have affected the power to detect tissue-shared effects for gene expression. They also point out that some "adipose-specific" signals may have been reported in other tissues in previous studies.
- Future Directions: The authors emphasize the need for further research to fully understand the mechanisms by which smoking impacts metabolic health and to explore the predictive potential of these epigenetic markers in larger, diverse populations. Investigating the role of cell-specific methylation profiles for signals like BHLHE40 is also highlighted.
Summary updated