Using high-density arrays, Dr. Jerome Brody and colleagues from Boston University are collaborating with Affymetrix to establish a baseline of normal gene expression in human airway epithelial cells. Their goal is to define the transcriptome changes that occur in these cells after exposure to inhaled smoke, and amongst these changes, identify those that are reversible upon smoking cessation. In addition, they are attempting to define how variables such as gender, race, and cumulative smoke exposure affect the epithelial cell transcriptome. Results from this study, and others to follow, will enable the development of a link between transcriptome signatures and a patient's risk profile, prognosis, or treatment option.

This study used human airway epithelial cells obtained by fiberoptic bronchoscopy, a medical technique that uses a thin, flexible instrument that is inserted into the lungs, where a tissue sample is taken by airway brushing. This procedure was performed on healthy, non-smoking patients; their smoking counterparts; and former smokers.

The Normal Airway Transcriptome

Lung tissue samples from healthy individuals were used to assess the normal airway transcriptome. Epithelial cells were obtained by as described above and RNA was extracted, labeled, and hybridized to GeneChip® U133A Arrays to quantitate the expression of more than 22,000 genes.

Transcription profiling of normal genes expressed in airway epithelial cells serves several purposes, Including:

• Providing insight into the function of these cells
• Functioning as a baseline for defining how cigarette smoking alters gene expression of these cells
• Identifying the effect of variables such as age, sex, degree of exposure on these cigarette smoke-induced changes in gene expression
• Identifying genes that revert to their normal levels after discontinued cigarette usage

The researchers found that over 2,000 genes were expressed in the epithelial cells of non-smokers. These genes were generally unaffected by parameters such as race, age, or gender. Most of the genes expressed in these cells, fall into several functional categories: genes related to oxidative stress, to ion/electron transport, to chaperon activity, and to vesicular transport. This implies that the function of these epithelial cells in normal non-smoking individuals is related to mechanisms of oxidant and toxic defense.

Smokers and Outliers

The researchers then assessed the transcription profile of smokers and compared these results to those of non-smokers. The most significant changes in gene expression were reflected in 97 genes. Genes expressed in epithelial airway cells that showed an increase in their expression profile include those that are related to cell adhesion, oxidant stress, glutathione metabolism, electron transport, xenobiotic metabolism and secretion, and several oncogenes. Genes whose expression was decreased were those related to inflammation regulation and tumor suppression.

Ironically, the profile of three of the smokers was similar to that of never smokers, despite their substantial and continuous exposure to cigarette smoke. These individuals did not have the high expression of a number of genes that characterize the "smoker airway transcriptome", such as the protective detox genes. The absence of genes related to detoxifying functions may be a factor in putting these "outliers" at high risk for more severe smoke-induced damage. Interestingly, one of these outlier patients subsequently developed cancer, which was diagnosed one year after the sample extraction. In addition, the expression profile of these 97 genes in one patient who had never smoked was much more similar to that of the smoker population. Whether this means that this individual is at higher risk for lung cancer is yet to be determined.

Cancer risk for smokers increases as a function of pack years. Results from this study show that the expression of selected putative oncogenes correlates positively with pack years, whereas that of some tumor suppressor genes correlates negatively. Only a small fraction of the signature variation could be explained by variables such as gender, age of initial smoke exposure, age of cessation, and environmental smoke exposure. More samples for each of the clinical parameters are required to understand their relationship to differential expression patterns.

Race seems to have an effect on the airway transcriptome of smokers: even though this study includes a limited sample size, it appears that some genes (not included in the original 97-gene group) were differentially expressed in Caucasian current smokers and healthy individuals compared to their African American counterparts. This difference may give insight into the increased risk of lung cancer among African-American smokers.

Former Smokers

There appears to be a threshold of two years after which the expression profile of former smokers segregates with that of healthy individuals. When the epithelial cell transcriptome of former smokers was assessed, the expression of some genes had reverted to normal levels. These reversible genes, associated with smoking cessation, were found to be predominantly drug-metabolism genes and antioxidant genes.

There were also genes whose level did not revert to normal levels two years after smoking cessation. These included potential tumor suppressor genes whose expression decreased, as well as several putative oncogenes, whose expression remained increased. The fact that some changes that occurred as a function of cigarette smoking persisted even after the patients stopped smoking may explain why some individuals are at higher risk to lung cancer even after they have stopped smoking.

Conclusions

Collaborators at Boston University and Affymetrix studied the global effects of smoking on the transcriptome of airway epithelial cells using GeneChip Arrays, which enabled them to:

• Correlate many transcribed genes not previously associated with smoking
  
• Describe the normal function of specific epithelial cells from a complex pulmonary region among healthy individuals
  
• Identify smokers that respond differently to cigarette exposure — a difference that might explain their increased risk to lung cancer
  

These gene expression profiles provide powerful insight into the inner workings of both healthy and smoke-exposed airway epithelial cells and help us understand, at a molecular level, the changes that persist in the lungs of former smokers. Ultimately, these compelling gene expression profiles could be applied clinically as predictive biomarkers for lung cancer.