You are here:
COMPARISON OF IN VITRO AND IN VIVO RESPONSES TO ARSENIC: GENE EXPRESSION PROFILING IN NORMAL HUMAN EPIDERMAL KERATINOCYTES AND HYPERKERATOSES FROM ARSENIC-EXPOSED HUMANS
Citation:
BAILEY, K., Y. XIA, J. MO, R. D. OWEN, J. S. MUMFORD, AND S. Y. THAI. COMPARISON OF IN VITRO AND IN VIVO RESPONSES TO ARSENIC: GENE EXPRESSION PROFILING IN NORMAL HUMAN EPIDERMAL KERATINOCYTES AND HYPERKERATOSES FROM ARSENIC-EXPOSED HUMANS. Presented at AACR special conference - Oncogenomics 2007: Dissecting Cancer Through Genome Research, Phoenix, AZ, January 31 - February 03, 2007.
Impact/Purpose:
To better understand the carcinogenic potential of trivalent arsenicals, we compared gene expression profiles of RNA isolated from two sample types.
Description:
Chronic exposure to arsenic is positively associated with skin, urinary bladder, lung, liver and kidney cancer development in humans. Elucidating the mode of action of arsenic carcinogenesis is a complicated issue as target cells are exposed to different toxic species of arsenic. Humans are generally exposed to inorganic arsenic in the drinking water, which is metabolized to several trivalent and pentavalent methylated intermediates that are distributed throughout the body and then excreted. The contribution of each of these arsenicals towards arsenic carcinogenesis is unknown. To better understand the carcinogenic potential of trivalent arsenicals, we compared gene expression profiles of RNA isolated from two sample types: (1) commercially-obtained normal human epidermal keratinocytes (NHEKs) exposed to arsenicals in vitro and (2) skin biopsies from humans exposed in vivo to arsenic via drinking water. In the in vitro study, NHEKs were exposed to non-cytotoxic, environmentally relevant concentrations of trivalent inorganic arsenic (iAsIII), monomethylarsonous acid (MMAIII) and dimethylarsinous acid (DMAIII) for 24 h. These profiles were compared to the in vivo profiles we obtained from arsenic-related hyperkeratoses (HKs) isolated from 7 Inner Mongolians who were chronically exposed to high levels of arsenic (212-950 µg/ml) in their drinking water for¿20 years. RNA prepared from untreated NHEKs and from normal skin from 4
individuals living near the endemic area (exposed to 7 µg/ml arsenic in drinking water) served as control in vitro and in vivo samples, respectively. Gene expression profiling was performed on all samples using Affymetrix® GeneChip Human U133 Plus 2.0 arrays and altered genes were determined using Genespring software (1-way ANOVA, p¿0.05;¿1.5-fold change in expression relative to controls). The HK samples had keratin gene expression patterns characteristic of hyperproliferative skin and exhibited increased expression of oncogenes and squamous cell carcinoma antigens. In addition, altered genes in HK are consistent with several proposed mechanisms of arsenic carcinogenesis, including oxidative stress, disruption of DNA repair processes, apoptosis effects and cell cycle modulation. The in vitro gene expression profiles after exposure to iAsIII, MMAIII and DMAIII had several commonalities, including the induction of stress response genes and modulation of apoptosis-related genes in a dose-dependent manner. However, each trivalent arsenical also had unique expression profiles: iAsIII-treated cells demonstrated strong induction of phase II-response genes controlled by transcription factor Nrf2; MMAIII-treated cells demonstrated an induction of inflammatory and proliferative genes and modulation of cytokines; DMAIII treatment resulted in the fewest number of modulated genes, with several being pro-apoptotic. Comparison of the in vitro and in vivo transcriptional responses to arsenic exposure suggest that all three trivalent arsenicals studied, and in particular MMAIII and iAsIII, may play critical roles in arsenic skin carcinogenesis and may provide useful biomarkers of exposure and response.