Aberrant Expression of ROS1 is Uncommon in Ovarian Cancer

Abstract

Background: The receptor tyrosine kinase ROS1 is overexpressed in multiple cancer types and is demonstrated to drive tumorigenesis by fusing with a number of genes including FIG, CD74, and SLC34A2. However, in the context of ovarian cancer, the frequency and the clinicopathologic significance of ROS1 expression and gene fusion are largely undetermined. Herein, we aim to examine the expression profile of ROS1 and to detect the presence of known reported ROS1 fusion genes in ovarian cancer. Methods: Immunohistochemistry for ROS1 was carried out on 94 formalin-fixed, paraffin-embedded ovarian carcinoma tissues (41 clear-cell, 24 serous, 19 endometrioid, and 10 mucinous); 10 benign and 9 borderline tumor cases were included for comparison. RT-qPCR was also performed to examine the expression profile of ROS1 at the mRNA level in a panel of 13 ovarian cancer cell lines and 2 normal ovarian epithelial cell lines. Given that ROS1 is often fused with other genes in cancer, fusion-specific qPCR was further conducted by designing primers targeting three most common ROS1 fusion genes, FIG-ROS1, CD74-ROS1, and SLC34A2-ROS1. To evaluate ROS1 expression at the protein level, immunoblotting of cell lysates was also performed in the above cell lines. Subcellular localization of ROS1 was determined by immunofluorescence. Results: In our cohort of ovarian tumor clinical samples, ROS1 exhibited negative staining in all the cancer tissues. Interestingly, in a serous carcinoma cell line, SKOV3, we observed a higher expression of ROS1 transcripts compared to the other ovarian cancer cell lines. However, no expression of the three known ROS1 fusion transcripts was detected by RT-qPCR, suggesting that SKOV3 cells might not display genomic rearrangement leading to ROS1 gene fusion. Our immunoblotting analysis revealed that ROS1 protein products were barely detectable in all the cancer cell lines including SKOV3. In concordance with the immunoblotting results, fluorescent signals could not be detected in SKOV3 cells probed with the ROS1 antibody. Although the undetectable protein expression is contradictory to the expectation from the elevated ROS1 mRNA level in SKOV3, we believe this discordance may be caused by underlying post-transcriptional mechanisms. Conclusion: This study shows that our cohort of ovarian tumor tissue and panel of ovarian cancer cell lines do not express ROS1 proteins and any common ROS1 fusion products. The role of ROS1 in ovarian carcinogenesis still remains to be elucidated.