Novel in-vitro models and correlative experiments with primary tumor/normal tissue specimens have been utilized to identify epigenomic alterations which contribute to initiation and progression of thoracic malignancies. For example, comprehensive analysis of long non-coding RNA (lncRNA) perturbations during pulmonary carcinogenesis demonstrated that cigarette smoke condensate (CSC) significantly increased expression of lncRNA ZFAS1 while decreasing expression of ZNFX1, an uncharacterized, putative tumor suppressor gene. ZFAS1 expression correlated inversely with ZNFX1 expression in primary lung cancers relative to paired normal lung tissues. Over-expression of ZFAS1 or ZNFX1 increased or decreased, respectively, growth and invasion of lung cancer cells in-vitro and in-vivo. Mechanistic studies demonstrated that ZFAS1 functions as a scaffold to recruit DNMTs and polycomb proteins to the ZNFX1 promoter to silence this tumor suppressor gene in respiratory epithelial cells following cigarette smoke exposure. CSC-mediated activation of ZFAS1 coincided with recruitment of miR-31 as well as SP1 to the ZFAS1 regulatory region. Mithramycin, an antineoplastic agent currently in clinical trials in my Section repressed ZFAS1 and induced ZNFX1 expression in lung cancer cells in-vitro and in-vivo. A comprehensive manuscript pertaining to these experiments is presently under peer review. Decreasing cigarette consumption in the U.S. due to public education and tobacco reform has been offset recently by the use of relatively unregulated cigarette alternatives, such as communal smoking of hookah tobacco (shisha) via waterpipe. Although hookah is perceived to be a safe alternative to cigarette smoking, the effects of hookah smoke in respiratory epithelia have not been well characterized. A series of experiments have been conducted to examine the epigenenomic and transcriptomic effects of hookah smoke relative to conventional cigarette smoke in normal and immortalized human respiratory epithelial cells. Briefly, primary normal human small airway epithelial cells (SAEC) from three donors, as well as cdk4/hTERT-immortalized SAEC and human bronchial epithelial cells (HSAEC and HBEC, respectively) were cultured for 5 days (short-term) or 12 months (long term) in normal media with or without cigarette smoke condensates (CSC; 0.05 or 0.1 mg/ml) or water pipe condensates (WPC; 1-2 mg/ml). Short term CSC and WPC exposures mediated dose-dependent growth inhibitory effects in cultured respiratory epithelia cells, which coincided with significantly reduced levels of global H4K16Ac and H4K20me3. CSC and WPC mediated distinct as well as overlapping cancer-associated gene expression signatures that were cell line and dose dependent. 73 genes were commonly regulated by CSC 0.05 mg/ml across three SAEC lines, HBEC, and HSAEC, whereas 157 genes were commonly regulated by CSC 0.1 mg/ml across these five cell lines. In contrast, 34 genes were commonly regulated by WPC 1 mg/ml, and 85 genes were commonly regulated by WPC 2 mg/ml across these five lines. 15 genes were commonly regulated by CSC 0.05 mg/ml and WPC 1 mg/ml across the three SAEC lines, HBEC and HSAEC, whereas 39 genes were commonly regulated by CSC 0.1 mg/ml and WPC 2 mg/ml across the five respiratory epithelial cell lines. Eight genes were commonly modulated by all four condensate exposures across all five cell lines. Notably, one of these genes (epiregulin; EREG) is significantly up-regulated in lung cancers, and correlates inversely with survival in lung cancer patients. Mechanistic studies demonstrated that CSC and WPC mediate demethylation of the EREG promoter in NREC. Results of these short-term exposure experiments, which represent the most comprehensive analysis of either cigarette or hookah smoke in human respiratory epithelial cells are in the final stages of preparation for publication. Additional studies have been conducted to examine the effects of long term WPC relative to CSC exposures in immortalized respiratory epithelial cells. Similar to what we have previously reported for CSC, prolonged WPC exposure mediated time and dose dependent decreases in DNMT1 levels while increasing DNMT3b levels in these cells. These alterations coincided with epigenetic reprogramming, and increased proliferation of cells exposed to WPC. Furthermore, cells that had been chronically exposed to WPC demonstrated significantly increased soft agar clonogenicity. This phenomenon coincided with marked up-regulation of SALL4, a stem cell protein that is over-expressed in human lung cancers and correlates with decreased survival of lung cancer patients. Additional studies are underway to further examine the mechanisms and implications of up-regulation of SALL4 and other stem cell genes in NREC following long-term WPC exposure. Collectively, these findings clearly demonstrate that hookah tobacco is not a safe alternative to cigarettes, and warrant efforts to curtail hookah smoking worldwide. Whereas asbestos burden has been linked to cytogenetic alterations in malignant pleural mesotheliomas (MPM), epigenetic aberrations induced by these fibers have not been fully delineated. To examine this issue, normal human mesothelial cells were cultured with or without crocidolite asbestos fibers for 10 days. Asbestos mediated time and dose dependent repression of RASSF1A, p16INK4a, and p14ARF in normal mesothelial cells; this phenomenon coincided with up-regulation of DNMT1 as well as increased expression of EZH2 the catalytic core component of PRC2, which we have previously shown to be an epigenetic driver of malignancy in pleural mesotheliomas. Up-regulation of EZH2 coincided with repression of miR26A and miR101, which target the 3 UTR of the EZH2 transcript. Asbestos mediated de-novo DNA methylation in the promoter of RASSF1A, but not within the INK4a/ARF locus. In contrast, asbestos induced recruitment of EZH2 to the RASSF1A as well as p16INK4a and p14ARF promoters, which coincided with increases in the PRC-2 mediated repressive histone mark, H3K27Me3, and decreased H3K9Ac (histone activation mark) within these regulatory regions. The fact that other tumor suppressor genes such as RAR-beta, and FHIT which are frequently silenced by epigenetic mechanisms in MPM were not repressed under these exposure conditions, suggests a strong selection pressure to inactivate Hippo, Retinoblastoma, and p53 tumor suppressor pathways early in the pathogenesis of asbestos-induced pleural mesothelioma. A manuscript pertaining to these findings will be submitted for peer review in the near future.
Showing the most recent 10 out of 13 publications
