Abstract
OBJECTIVE: To understand the chemotherapy response program in ovarian cancer cells at deep transcript sequencing levels. METHODS: Two next-generation sequencing technologies--MPSS (massively parallel signature sequencing) and SBS (sequencing by synthesis)--were used to sequence the transcripts of IGROV1 and IGROV1-CP cells, and to sequence the transcripts of a highly chemotherapy responsive and a highly chemotherapy resistant ovarian cancer tissue. RESULTS: We identified 3422 signatures (2957 genes) that are significantly different between IGROV1 and IGROV1-CP cells (P<0.001). Gene Ontology (GO) term GO:0001837 (epithelial-to-mesenchymal transition) and GO:0034330 (cell junction assembly and maintenance) are enriched in genes that are over expressed in IGROV1-CP cells while apoptosis-related GO terms are enriched in genes over expressed in IGROV1 cells. We identified 1187 tags (corresponding to 1040 genes) that are differentially expressed between the chemotherapy responsive and the persistently chemotherapy resistant ovarian cancer tissues. GO term GO:0050673 (epithelial cell proliferation) and GO:0050678 (regulation of epithelial cell proliferation) are enriched in the genes over expressed in the chemotherapy resistant tissue while the GO:0007229 (integrin-mediated signaling pathway) is enriched in the genes over expressed in the chemotherapy sensitive tissue. An integrative analysis identified 111 common differentially expressed genes including two bone morphogenetic proteins (BMP4 and BMP7), six solute carrier proteins (SLC10A3, SLC16A3, SLC25A1, SLC35B3, SLC7A5 and SLC7A7), transcription factor POU5F1 (POU class 5 homeobox 1), and KLK10 (kallikrein-related peptidase 10). A network analysis revealed a subnetwork with three genes BMP7, NR2F2 and AP2B1 that were consistently over expressed in the chemoresistant tissue or cells compared to the chemosensitive tissue or cells. CONCLUSION: Our database offers the first comprehensive view of the digital transcriptomes of ovarian cancer cell lines and tissues with different chemotherapy response phenotypes.