Sorafenib induces variations of the DNA methylome in human hepatocellular carcinoma cells.

Sorafenib is currently the standard treatment for advanced hepatocellular carcinoma (HCC). Epigenetic alterations such as DNA methylation, play a decisive role in the development and progression of HCC. To our knowledge, there are no studies that have analyzed the global DNA methylation changes in HCC cells treated with sorafenib. Using MeDip-chip technologies we analyzed sorafenib effects on the methylome of human HCC cells. We found 1230 differentially methylated genes in HA22T/VGH cells treated with sorafenib compared to untreated cells. Gene ontology and pathway analysis found enriched several GO terms related to transcription factors and different pathways involved in tumorigenesis and cancer progression. Among the genes differentially methylated we found genes related to apoptosis (FOXO3, SMAD2, p21), angiogenesis (EPAS1) and invasion (MMP3, MMP7, RAC1, RHOC), genes belonged to pathways deregulated in HCC such as RAF/MEK/ERK (MAPK3, MAP2K2), JNK (MAPK8, MAP2K7), JAK-STAT (JAK1, STAT3, STAT5, CCND3), PI3K/AKT/mTOR (TSC2, PRKCZ) and NF-κB (IKBKG, MALT1, MAP3K14) and cancer-associated non-coding genes such as MALAT1, miR-149 and miR-675. We found a general trend where oncogenes were hypermethylated and tumor suppressor genes were hypomethylated after sorafenib treatment. Finally, we validated MeDip-chip results on several differentially methylated genes using COBRA and direct bisulfite sequencing and for these genes we evaluated the relationship between methylation level and mRNA expression.Our results suggest that in HCC cells sorafenib could affect the methylation level of genes associated to cancer-related processes and pathways related to sorafenib mechanism of action. These results identified novel sorafenib targets genes which could be targets in a new design of multi-target and combined therapies, and to better understand the emergence of resistance in HCC. In this work, HA22T⁄VGH HCC cells were maintained in RPMI-1640 (Life Technologies) supplemented with 10% fetal bovine serum at 37°C in a 5% CO2 incubator. Sorafenib (Nexavar®, Bayer Corporation) was dissolved in 100% DMSO (Sigma-Aldrich) and diluted in DMEM (Life Technologies) to the required concentration with a final DMSO concentration of 0.1%. Cells were cultured in 10 cm Ø plates up to almost confluence and were treated with sorafenib (5 µM) or solvent (0.1% DMSO) for 24 hours. Genomic DNA was extracted from treated and untreated cells using Wizard Genomic DNA Purification Kit (Promega) according to the manufacturer’s instructions and quantified using NanoDrop (Thermo Fisher Scientific). Global DNA methylation profiles of sorafenib treated and untreated cells were obtained by methylated DNA immunoprecipitation coupled with Affymetrix Human Promoter 1.0R Tiling Arrays (MeDip-chip) using a modification of the Affymetrix chromatin immunoprecipitation assay protocol. Genomic DNA from sorafenib treated and untreated cells was fragmented using micrococcal nuclease (New England Biolabs) to obtain DNA fragments ranging from 200 to 500 nucleotides and used them as input (IN) DNA. Agilent Bioanalyzer with the RNA 6000 Nano LabChip Kit was used to check the size of fragmented DNA. Part of IN DNA (4 µg) was immunoprecipitated (IP) with 10 µl of anti-5-MethylCytosine Antibody (Eurogentec, BI-MECY-0100) using Dynabeads® Protein G immunoprecipitation kit (Life Technologies). The methylated enriched DNA was purified by standard phenol/chloroform procedure and isopropanol precipitated. The MeDip was performed in duplicate for each DNA sample starting with initial fragmentation step. A total of 200 ng of IN and IP DNA was amplified with the Affymetrix Chromatin Immunoprecipitation Assay Protocol and hybridized on Human Promoter 1.0R arrays. Totaling 3 arrays per each cellular condition (2 IP DNA and 1 IN DNA) were performed. The CEL files were imported into Partek® Genomics Suite® (PGS). The differences between sorafenib treated and untreated cells were analyzed using ANOVA. Significantly differentially methylated regions (DMR) were obtained using the MAT algorithm (model-based analysis of tiling arrays algorithm). Using PGS function, each DMR was associated to its nearest gene of human genome. The genomic coordinates of each DMR were calculated based on UCSC human genomic assembly version 18.