(1)卵巣癌における化学療法感受性バイオマーカーの開発と組織型による分子標的の同定 : 治療個別化に向けたBioinformaticsの応用(<特集>第61回学術講演会シンポジウム3「卵巣癌の新たな治療戦略-基礎から,そして臨床から-」)

概要

論文の詳細を見る
Expression of the genomic genes, not a single gene, in cancer cells may be relevant to clinical features of individual cancers. In order to link findings of basic research with the treatment of ovarian cancer, genome-wide analysis using bioinformatics is thought to be necessary. In this research, we analyzed the following three subjects using bioinformatics to improve drug treatment against ovarian cancer. 1) Biomarker of chemo-sensitivity A transcript factor YY1 was one of genes that discriminate between good and poor prognosis in patients with serous ovarian cancer. In the microarray dataset of 88 serous ovarian cancers, genes having both YY1 binding sites and E2F binding sites positively correlated with YY1, which is consistent with a previous report that YY1 increases E2F activity. Hierarchical clustering by the genes having both YY1 and E2F binding sites which positively correlate with YY1 classified ovarian cancers into the two clusters, "YY1 Low" cluster and "YY1 High" cluster. In serous ovarian cancer patients who received paclitaxel, YY1 High showed better survival than YY1 Low, though in patients who did not receive paclitaxel, this classification did not correlate with patient's survival. NCI60 cell lines, derived from cancers of various organs, were analyzed in a similar way, and YY1 Low cells were resistant to paclitaxel and docetaxel than YY1 High cells, while this classification did not correlate with platinum sensitivity. The siRNA knock-down of YY1 in BG1 ovarian cancer cell line resulted in resistance to paclitaxel and docetaxel, while it did not affect on the sensitivity to cisplatin. Therefore, we conclude that YY1 increases E2F activity and confers taxane sensitivity in ovarian cancer. 2) Novel treatment of ovarian clear cell carcinoma This study aims to find a novel therapeutic drug against ovarian clear cell carcinoma (CCC) based on the biologic features detected by gene profiling. Microarray analysis using both ovarian cancer cell lines (13 CCC and 25 non-CCC) and ovarian cancer tissues (8 CCC and 91 non-CCC) identified "Clear Cell Signature" which discriminates between CCC and non-CCC. The Clear Cell Signature, comprising of 303 upregulated and 39 downregulated genes, was reproducible in the two other microarray datasets. The Clear Cell Signature contained HNF1β and genes having HNF1 binding sites. Mutation of HNF1β is known to cause renal hypoplasia. Hierarchical clustering of the merged dataset of 38 ovarian cancer cell lines and NCI60 indicated that CCC was strikingly similar to renal cell carcinoma. Because Sorafenib is effective against renal cell carcinoma, we examined the in vivo antitumor effect of Sorafenib against RMG-2 CCC cell line inoculated on nude mice. Sorafenib caused the RMG-2 tumor to shrink significantly. We conclude that Sorafenib may be useful against CCC. 3) DNA methylation and signal pathway DNA methylation is relevant to carcinogenesis and cancer progression. However, there are few studies that analyzed genome-wide DNA methylation in ovarian cancer. In this research, we aimed to analyze how the genome-wide DNA methylation affects on the signal pathway and phenotypes in ovarian cancer. Forty-three ovarian cancer cell lines were treated by 5Aza-dC, followed by microarray analysis. Reactivated 360 genes were assumed as methylation candidate genes, and they enriched genes belonging to TGF-beta signal pathway. Demethylation by 5Aza-dC significantly upregulated the TGF-beta pathway activity detected by Binary Regression and SMAD3-luciferase assay in ovarian cancer cell lines. Because cancer metastasis is supposed to be associated with epigenetic changes and TGF-beta pathway deregulation, expression of the 360 methylation candidate genes and TGF-beta activity were analyzed in primary sites and ometntal metastasis of ovarian cancer samples. Hierarchical clustering by the 360 genes identified a definite gene cluster of approximately 100 genes, which contain multiple genes known to activate TGF-beta signal pathway. TGF-beta activity detected by Binary Regression positively correlated with the expression of the gene cluster. Furthermore, immunohistochemical analysis for the paired samples of primary site and omental metastasis showed increased expression of TGFBR2 and phospohorylated SMAD2 in omental metastasis. Treatment with A-83-01, an inhibitor of TGFBR1 kinase, suppressed expression of phospohorylated SMAD2 in HM-1 mouse ovarian cancer cell line, and also significantly inhibited migration and invasion of HM-1. Furthermore, in vivo treatment with A-83-01 prolonged survival of mice in the HM-1 peritoneal metastasis model. Therefore, we conclude that inhibition of TGF-beta signal pathway may be useful in the treatment of ovarian cancer metastasis. In summary, using bioinformatics, we found that 1) ovarian cancers with activated YY1/E2F signal are sensitive to taxane, 2) Sorafenib is effective against CCC, 3) TGF-beta inhibitor may be useful to treat ovarian cancer metastasis. These results show that a genome-wide approach using bioinformatics may be useful to develop a novel diagnostic and therapeutic modality against ovarian cancer.
2009-11-01

(1)卵巣癌における化学療法感受性バイオマーカーの開発と組織型による分子標的の同定 : 治療個別化に向けたBioinformaticsの応用(<特集>第61回学術講演会シンポジウム3「卵巣癌の新たな治療戦略-基礎から,そして臨床から-」)

スポンサーリンク

概要

著者

関連論文

スポンサーリンク