Final Report Summary - ONCOGENIC SPLICING (Characterizing the oncogenic properties and target genes of the splicing factor protooncogene SF2ASF) Objectives, results, achievements & list of publications During the past three years we have genetically engineered cells to express a set of mutants of the cancer-promoting gene (oncogene) SF2/ASF (SRSF1). Each of the mutants lacks a specific functional domain and thus is impaired in its abilities to promote a specific biological function. We are evaluating the malignant properties of the different mutants as well as their normal function in gene expression regulation in vitro and in vivo. We found that some mutants such as deletion of RRM2 or a mutant with addition of the RS domain of SC35 which cannot exit from the nucleus (NRS1) still function in the regulation of SF2/ASF splicing target BIN1 while other mutants fail to do so. Our preliminary results also show that some domains of SF2/ASF are required for its cancerous activity (e.g. the RRM1 domain) while deletions of other domains like RRM2 had a surprising effect of generating a 'super oncogene' which converts cells into more violent cancer cells than the normal SF2/ASF. In order to study the direct effects of SF2/ASF and also to be able to examine the dependence of transformed cells on SF2/ASF expression we generated inducible cell lines in which we can turn 'on' and 'off' the expression of SF2/ASF or its mutants and examine early events caused by SF2/ASF up- or down-regulation. Aim 1: To study the mechanisms of SF2/ASF-mediated transformation. Progress made: we have established mouse (NIH 3T3, PHN-1) and human (MCF-10A) non-transformed domains. We are evaluating the oncogenic properties of the different mutants by colony formation assay in soft agar (data not shown) and tumorigenic potential in nude mice. We found that RRM1 and the shuttling activity of SF2/ASF are required for its oncogenic activity and that deletion of RRM2 increased the oncogenic potential of SF2/ASF. These new surprising results suggest that RRM1 alone (with the RS domain) is sufficient for transformation. The NRS1 mutant cannot exit the nucleus and cannot transfors suggesting that the shuttling activity of SF2/ASF is required for transformation. We also examine the function of the mutants in alternative splicing and signaling in vitro and in vivo. We also found that SF2/ASF activates the ERK-MAPK pathway in certain cancers and are examining the contribution of this activation for SF2/ASF-mediated transformation and the mechanism of this activation. In addition, we knocked down SF2/ASF by stable expression of shRNAs in human lung carcinoma cells and examined its effects on transformation in vitro by the soft agar assay and tumorigenicity in vivo in nude mice. We found that SF2/ASF down-regulation inhibited the transformation and tumorigenicity of the cells in vitro and in vivo respectively. Currently, In order to study the direct effects of SF2/ASF and also to be able to examine the dependence of transformed cells on SF2/ASF expression we are generating tet-inducible cell lines (HeLa and MEFs) in which we can turn 'on' and 'off' the expression of SF2/ASF or its mutants and examine early events caused by SF2/ASF up- or down-regulation. We intend to investigate alternative splicing effects as well as the requirement of SF2/ASF to tumor maintenance by 'shutting off' the expression of SF2/ASF after tumors from cells overexpressing SF2/ASF were established. We also compared the oncogenic activity of SF2/ASF to other SR proteins and hnRNP A/B proteins and found that hnRNP A2/B1 acts as a protooncogene in brain cancer. This finding was recently published (see below) in Cancer Research and was partly supported by the ISF. We also investigated two of SF2/ASF's splicing targets we previously identified (Mnk2 and S6K1). We previously showed that SF2/ASF induces the Mnk2b isoform and reduces Mnk2a. We now demonstrated that this splicing switch inactivates the tumor suppressive Mnk2a isoform and elevates an oncogenic isoform Mnk2b. We show that Mnk2a can bind, activate and translocate into the nucleus the stress inducer p38-MAPK inducing the expression of the p38-MAPK stress pathway and leading to apoptosis. Mnk2b cannot bind to p38-MAPK but can still phosphorylate the translation initiation factor eIF4E promoting transformation. These finding were recently submitted for publication. Similar results were obtained for S6K1 were SF2/ASF induces a new isoform we identified which is oncogenic and we found that this isoform can activate mTORC1 inducing eIF4E-BP1 phosphorylation, cellular transformation and increased motility of cells. Inhibiting mTORC1 or blocking 4E-BP1 by expression of phosphorylation-impaired mutant of 4E-BP1 could block all the oncogenic phenotypes of the short isoforms of S6K1. These finding will be submitted soon for publication. Aim 2: To identify and characterize the targets of SF2/ASF regulation. We have decided to change our strategy in identifying the targets of SF2/ASF regulation and use the new technology of deep sequencing to sequence the transcriptome (RNA-seq) of cells with up or down-regulation of SF2/ASF and other factors. We have made progress and established two important collaborations to facilitate the completion of this aim. We are currently sequencing calls with up and down-regulation of splicing factors in collaboration with the group of Prof. Gideon Rechavi, head of the cancer center at Sheba-Tel Hashomer Medical Center using the Solexa technology, and also with Dr Marc Sultan and Prof. Hans Lehrach from the Max Planck Institute for Molecular Genetics in Berlin. We performed part of the sequencing and have already analysed part of the data. A talented postdoctoral fellow, Dr Ilana Lebenthal who joined the the lab is analyzing the data and we already validated some of the analysis by Reverse transcription polymerase chain reaction (RT-PCR) (see the main new application for preliminary results). Publications 1. Regina Golan-Gerst, Michal Cohen, Asaf Shilo, Sung Suk Suh, Arianna Bakacs, Luigi Coppola, and Rotem Karni (2011) Splicing factor hnRNP A2/B1 regulates tumor suppressor gene splicing and is an oncogenic driver in glioblastoma. Cancer Res. 71: 4464-72. Supported in part by the Marie Curie IRG re-integration grant FP7
