Abstract PR1: Genomic discovery of CD44-SLC1A2 gene fusions in gastric cancer

Among the various genomic abnormalities associated with cancers, fusion genes and transcripts are particularly notable due to their cancer-specific nature and translational potential as diagnostic and therapeutic targets. In pediatric acute lymphoblastic leukemia (ALL), assays detecting the AML-ETO and PML-RAR fusion genes are routinely used to diagnose particular clinical subtypes, and treatment of chronic myelogenous leukemia (CML), a disease once associated with extremely poor prognosis, has been revolutionized by imatinib, an inhibitor of the BCR-ABL fusion gene. While fusion genes are well- known in hematological malignancies, our knowledge of recurrent fusion events in solid epithelial tumors is far less mature. To date, only two fusion genes have been identified in solid tumors, including TMPRSS2-ERG in prostate cancer, and EML4-ALK in non-small cell lung cancer. No recurrent fusion genes have been so far identified in common gastrointestinal cancers such as stomach, colon and esophageal adenocarcinoma. In this study, to identify genes affected by genome rearrangements in gastric cancer (GC), which is the second highest cause of global cancer mortality, we analyzed chromosomal imbalances in a cohort of 123 primary GCs and cell lines, and discovered several tumors exhibiting recurrent genomic breakpoints in the SLC1A2/EAAT2 glutamate transporter. 5’ RNA ligase mediated rapid amplification of cDNA ends (RLM-RACE) analysis of a GC cell line with SLC1A2 breakpoints (SNU16) revealed expression of a CD44-SLC1A2 fusion transcript, computationally predicted to result in a truncated but functional SLC1A2 protein. In an independent tumor panel, CD44-SLC1A2 gene fusions were detected in 4% to 5% of GCs (2/43), but not in adjacent matched normal gastric tissues. Using custom-designed fusion-specific siRNAs, we show that silencing of CD44-SLC1A2 in fusion-positive SNU16 cells significantly reduced cellular proliferation, invasion, and colony formation, but not in cell lines lacking CD44-SLC1A2 expression. CD44-SLC1A2 silencing also significantly reduced intracellular glutamate levels and sensitized SNU16 cells to cisplatin. Fusions of SLC1A2 to upstream CD44 regulatory elements likely causes SLC1A2 transcriptional dysregulation, as tumors expressing abnormally high SLC1A2 levels also tended to be CD44-SLC1A2 positive. CD44-SLC1A2 gene fusions may be one mechanism used by cancers to establish a pro-oncogenic metabolic milieu favoring tumor growth and survival. We believe that our study will be of great interest to the gastric cancer community, as it represents the first recurrent fusion transcript identified in this disease (and for that matter any common gastrointestinal cancer). Citation Information: Clin Cancer Res 2010;16(14 Suppl):PR1.