Gene expression signature-based prediction model for synergistic effect of histone deacetylase inhibitor and chemotherapy

Abstract

Purpose We evaluated the cytotoxic effects of combining of suberoylanilide hydroxamic acid (SAHA), a histone deacetylase inhibitor, with taxanes in human gastric cancer cell lines, and studied the pre-treatment difference of gene profile to identify synergism-related genes that could potentially mediate the cytotoxic response. Next we developed predicting model for the chemosensitivity of taxane/SAHA combination. To understand the roles of synergism-related genes we investigated the change of protein expression related with cell cycle and HDAC. We also analyzed genetic variants by whole exome-sequencing to identify possible mechanisms that may be responsible for synergism. Methods Twenty five gastric cancer cell lines with 22K gene expression data were treated with SAHA and paclitaxel or docetaxel, and the synergistic interaction between the drugs was evaluated in vitro using the combination index (CI) method. We performed significance analysis of microarray (SAM) to identify chemosensitivity-related genes in gastric cancer cell lines that were concomitantly treated with SAHA and taxane. We generated a correlation-matrix between gene expression and CI values to identify genes whose expression correlated with a combined effect of taxanes and SAHA. On the basis of expression profiles of genes selected from microarray, we developed ‘Molecular Diagnosis Score (MDS)’ prediction model with quantitative RT-PCR (qRT-PCR). To identify the mechanism of synergism, mRNA expression of selected genes with RT-PCR, protein expression of cell cycle-, apoptosis-related genes, histone acetylation with western blot and cell cycle analysis with flow cytometry were investigated. We performed whole-exome sequencing on 40 gastric cancer cells using Illumina HiSeqTM. The sequencing data were examined for genes which were involved in histone acetylation, cell cycle regulation and

tubulinsResults Taxane and SAHA combination had a synergistic cytotoxic effect against taxane-resistant gastric cancer cells. We selected 49 chemosensitivity-related genes, which were commonly identified in paclitaxel and docetaxel combined with SAHA, via SAM analysis. Among them, nine common genes were extracted from the subsequent correlation-matrix analysis. These genes were validated with qRT-PCR and used for developing MDS. Application of MDS scoring system for evaluating tumor tissue of gastric cancer patients who were treated with taxane predicted the sensitivity of taxane with 100% sensitivity, 43% specificity. The selected nine genes of gastric cancer cells were modulated by SAHA and taxane. The protein expression of p21 and acetylated histone H3 were significantly increased with paclitaxel/SAHA in synergism-inducing cell line. The frequency of OR4L1Gly109Ser, GBP4G2366R, KCNN3L66H , IGSF9G34E mutated cells was significantly higher in synergism-induced cells than other cells, respectively. HDAC9G366V mutation was exclusively identified in synergism- induced YCC-16. Conclusions Collectively, we discovered biomarkers for predicting chemosensitivity of SAHA/taxane combination to further aid the development of personalized chemotherapy regimen for gastric cancer patients. Prospective study is needed to fully define the clinical utility of the biomarkers.