Effects of cigarette smoke on the human oral mucosal transcriptome

Abstract

Use of tobacco is responsible for approximately 30% of all cancer-related deaths in the United States, including cancers of the upper aerodigestive tract. In the current study, 40 current and 40 age- and gender-matched never smokers underwent buccal biopsies to evaluate the effects of smoking on the transcriptome. Microarray analyses were carried out using Affymetrix HGU133 Plus 2 arrays. Smoking altered the expression of numerous genes: 32 genes showed increased expression and 9 genes showed reduced expression in the oral mucosa of smokers versus never smokers. Increases were found in genes involved in xenobiotic metabolism, oxidant stress, eicosanoid synthesis, nicotine signaling, and cell adhesion. Increased numbers of Langerhans cells were found in the oral mucosa of smokers. Interestingly, smoking caused greater induction of aldo-keto reductases, enzymes linked to polycyclic aromatic hydrocarbon-induced genotoxicity, in the oral mucosa of women than men. Striking similarities in expression changes were found in oral compared with the bronchial mucosa. The observed changes in gene expression were compared with known chemical signatures using the Connectivity Map database and suggested that geldanamycin, a heat shock protein 90 inhibitor, might be an antimimetic of tobacco smoke. Consistent with this prediction, geldanamycin caused dose-dependent suppression of tobacco smoke extract-mediated induction of CYP1A1 and CYP1B1 in vitro. Collectively, these results provide new insights into the carcinogenic effects of tobacco smoke, support the potential use of oral epithelium as a surrogate tissue in future lung cancer chemoprevention trials, and illustrate the potential of computational biology to identify chemopreventive agents.

Figure 1

A, Unsupervised hierarchical clustering of the expression of probesets differentially expressed in the oral mucosa of smokers vs. never smokers. Smokers and never smokers cluster primarily into two distinct groups. Each column corresponds to the expression profile of an oral mucosal biopsy, and each row corresponds to an mRNA. The color in each cell reflects the level of expression of the corresponding mRNA relative to its mean level of expression in the entire set of biopsy samples. In this heatmap, the increasing intensities of red signify that a specific mRNA has a higher expression in the given sample whereas the increasing intensities of blue mean that this mRNA has lower expression. White indicates mean level of expression. B, Direct interaction network of differentially expressed genes generated using IPA and other known interactions. The white nodes represent genes with no significant expression change that potentially contribute to the effects of smoking.

Figure 2

Increased numbers of Langerhans cells were found in the oral mucosa of smokers. Non-neoplastic oral mucosae from never smokers (A) and smokers (C) were morphologically similar, but samples from never smokers showed relatively few Langerhans cells (B) compared to those from smokers (D), which contained numerous Langerhans cells in the peripapillary (arrow) and interpapillary mucosa (arrowhead). [Magnification 100X for panels A–D; panels A and C stained with hematoxylin and eosin; panels B and D stained with CD1a immunostain and hematoxylin]. E, Intraepithelial cells that displayed moderate to strong cytoplasmic staining for CD1a in dendritic-type cellular processes were quantified in the peripapillary, interpapillary and superficial epithelium. A statistically significant increase in the number of CD1a positive cells was found in all three regions in smokers compared to never smokers (P<0.001, <0.001 and =0.032 for peripapillary, interpapillary and superficial areas, respectively). Panel E reflects the total number of CD1a positive cells in the three regions. Columns, means; bars, S.E.; n = 27/group. *, P<0.001. F, MSK-Leuk1 cells were pretreated with vehicle or the indicated concentration of geldanamycin for 2 h. Subsequently, cells received vehicle or TS for 5 h and were then harvested for Western blot analysis. Cellular lysate protein (100 μg/lane) was loaded onto a 10% SDS–polyacrylamide gel, electrophoresed and subsequently transferred onto nitrocellulose. Immunoblots were probed with antibodies specific for CYP1A1, CYP1B1 and β-actin.