Transcriptomic epidemiology of smoking: the effect of smoking on genome-wide gene expression
We recently published a study showing the massive impact of cigarette smoking, as an environmental influence, on whole blood gene expression (Charlesworth et al. 2010). In the initial study we obtained genome-wide quantitative transcriptional profiles from 1,240 individuals from the San Antonio Family Heart Study, including 297 current smokers. Using lymphocyte samples, we identified 20,413 transcripts with significantly detectable expression levels, including both known and predicted genes. Correlations between smoking and gene expression levels were determined using a regression model that allows for residual genetic effects.
With a conservative false-discovery rate of 5% we identified 323 unique genes (342 transcripts) whose expression levels were significantly correlated with smoking. These genes showed significant over-representation within a range of functional categories that correspond well with known smoking-related pathologies, including immune response, cell death, cancer, natural killer cell signalling and xenobiotic metabolism. The results indicate that not only individual genes but
entire networks of gene interaction are influenced by cigarette smoking.
We now have a second set of gene expression profiles and smoking data from this same cohort, 15 years after the original study. We need to determine whether the initial results are consistent in this cohort, whether the gene expression alterations revert to 'normal' in those individuals who have quit smoking and whether there are other detectable longitudinal effects.
- Dr Jac Charlesworth (Research Fellow)
- Dr John Blangero, Texas Biomedical Research Institute