Erwei Song 14th International Biennial Conference on Metastasis Research 2012

Erwei Song

Studies on non-coding RNAs and breast cancer biology Erwei Song Sun-Yat-Sen Memorial Hospital, Sun-Yat-Sen University, Guangzhou, P.R.China Over the past decade, the central dogma of gene regulation via transcription from DNA to messenger RNA and translation from messenger RNA to protein has been challenged by high-throughput genomic studies demonstrating that the portion of the genome responsible for protein coding constitutes approximately 1.5%, while many noncoding regulatory elements are transcribed into noncoding RNA (ncRNA) that regulate the complexity of higher organisms. Not until recently has it been realized that ncRNAs could play significant regulatory roles in multiple major biological processes involved in development, differentiation, metabolism and diseases. NcRNAs can be classified according to their molecular size into small ncRNAs, including siRNA, miRNAs and piRNAs, and long ncRNAs (lncRNAs) that are greater than 200 nt in length. The formers are highly conserved and responsible for gene silencing via specific base pairing with their targets, while the latter are poorly conserved and regulate gene expression by diverse mechanisms. Since 2002, my research work has been focusing on how ncRNAs participate in the mechanisms of disease progression and their potentials in therapeutic application. Findings of my research work are summarized as follow: 1. The first report demonstrating the therapeutic potential of siRNA in whole animal: When I joined Judy Lieberman’s lab as a postdoctoral research follow at CBR Institute of Biomedical Research of Harvard Medical School in 2002, studies from Tom Tuschl’s group demonstrated for the first time that the mechanism of RNA interference mediated by siRNAs is effective in mammalian cells without inducing interferon response. Introducing siRNAs targeting HIV virus genes into human T lymphocytes protect them from the virus infection in cell cultures. However, at that time, it remains uncertain whether siRNAs can be applied in vivo for disease treatment. To evaluate the therapeutic potential of siRNAs, we employed hydrodynamic injection of siRNAs against a pro-apoptotic Fas gene (Fas-siRNA) into mouse models of hepatitis. Hydrodynamic injection of Fas-siRNAs into the tail veins of the mice efficiently inhibits Fas mRNA and protein expression in mouse hepatocytes. SiRNA-mediated silencing of Fas gene expression protects the animals from acute and chronic hepatitis induced by concanavalin A (ConA). In a more fulminant hepatitis model, injection of Fas-siRNAs dramatically increases survival of the mice challenged by an agonistic Fas-specific antibody. Our findings suggested that silencing Fas expression with RNAi holds therapeutic promise to prevent liver injury by protecting hepatocytes from cytotoxicity. This study was published as an article in Nature Medicine in 2003, in which I am the first author. The paper received reports and comments from Science, Nature Medicine, New England Journal of Medicine, Gastroenterology and so on. In his review on RNA interference, Philip A Sharp, a Nobel Prize winner in 1993 in RNA splicing, considered our study as “the first report that siRNA could be used therapeutically in whole animals”. A panel of figure from our article was chosen as representative result for “Top 10 Scientific Breakthrough of the Year 2003” by Science. Until August, 2012, the paper has been cited in ISI for 672 times. Not only had we shown the therapeutic potential of siRNA in mouse disease models, but we also demonstrated that application of synthetic siRNA duplex can induce potent and long-lasting gene silencing effect in non-dividing cells, such as macrophages. These data were published in Journal of Virology in 2004, where I am the first author, and the ISI citation for this paper was measured up to 142 times until August, 2012. In addition to publishing the above papers, I am also one of the inventors holding an international patent “Inhibition of gene expression using RNA interfering agents”, which was granted in Apr 2010. Besides, I contributed a book chapter “RNA interference in mouse models” for “Gene silencing by RNA interference”, published by CRC Press in 2004. 2. miRNAs involved in the regulation of breast cancer biology In contrast to most siRNAs that are synthetic small RNA duplexes delivered into mammalian cells, miRNAs are endogenous small non-coding RNAs of 19-22 nucleotides in size. Since I came back to Sun-Yat-Sen University to set up my research lab at Department of Breast Tumor Center of Sun-Yat-Sen Memorial Hospital in 2004, I have placed my research focus on how miRNAs contribute to the development of breast cancers. We aimed first to investigate the involvement of miRNAs in regulating the biology of breast “cancer stem cells” (CSCs), a subpopulation of tumor cells with stem-cell like features and strong tumorigenic capability. However, CSCs in solid tumors are so rare that the mechanisms that govern their properties are difficult to study. Based on our observation on clinical samples, we found that breast tumors from chemotherapy-treated patients are highly enriched for cells with properties of CSCs. We exploited the chemotherapeutic resistance of CSCs to generate a highly malignant breast cancer cell line (SK-3rd) with high proportion of CSCs by sequential in-vivo passage in chemotherapy-treated immunocompromised mice. After hurdling this bottle-neck problem of obtaining enough number of CSCs, we profiled the miRNA expression pattern in breast CSCs and identified a number of miRNAs that are differentially expressed from ordinary breast cancer cells without stem-cell like features. Among them, miRNAs from let-7 family are reduced most prominently in breast CSCs. Recapitulating the expression of let-7 in breast CSCs dramatically suppresses their stem-cell like features, such as self-renewal capacity, undifferentiation status and CSC marker expression, and retards tumorigenesis and metastasis of breast tumor xenografts generated by implanting breast CSCs in immunocompromised mice. This study suggested that let-7 reduction not only might serve as a marker to identify breast cancer stem cells, but also let-7 mimics could potentially be used as single agents or combined with conventional chemo/radiotherapy to provide more effective breast cancer treatment by suppressing breast CSCs. The study was published in Cell in 2007 and I am the last communication author. In his preview published in Cell Stem Cell, Curt A Davey pointed out that our study has provided the first glimpse of an answer to the question regarding a causative linkage between the control of miRNA expression and stemness of tumor cells, and this linkage may represent a functional signature of cancer stem cells. Our study was also highlighted by Nature Biotechnology, and was awarded “Top 10 Scientific Breakthrough in China in 2006”. ISI citation for this paper until August 2012 runs up to 466times. Apart from let-7, our further studies demonstrated that miR-30 regulates apoptosis of breast CSCs by targeting to Ubc9 and integrin-3, miR-128 is involved in chemotherapeutic resistance of breast CSCs via Bmi-1 and ABCC5, and miR-34c regulates invasiveness of breast CSCs via notch4. These studies were published in Oncogene, Clinical Cancer Research and Journal of Biological Chemistry respectively and I was the last corresponding author in these papers. Not only did we show that dysregulation of miRNAs play important roles in regulating the biology of breast CSCs, we also investigated how miRNA expression is dystregulated in these cells. For example, we showed that a single CpG island methylation of miR-34c promoter region resulted in loss of the miRNA expression in breast CSCs. Although the presence of CSCs is associated with failure of conventional anti-cancer treatment, targeting critical signaling pathways in tumor cells, such as ErbB2, Wnt and notch pathways, termed targeting therapy, emerges to be effective to eradicate CSCs from the bulk of cancer cells. However, cancer cells may be refractory to such targeting therapy at the beginning, considered as primary resistance, or may develop irresponsiveness in the course of the treatment, named secondary resistance. Therefore, it is necessary to elucidate mechanisms of resistance to targeting therapy and develop strategies to overcome it. To address the involvement of miRNAs in targeting therapy resistance, we compared the miRNA expression profile in paired breast cancer lines sensitive versus resistant to trastuzumab, a therapeutic monoclonal antibody against ErbB2. We found that miR-21 is the most prominently elevated miRNA that targets to PTEN, a known switch for trastuzumab resistance in breast cancers, and blocking the effect of miR-21 with antisense oligonucleotides (ASOs) sensitizes breast cancer cells to treatment with the monoclonal antibody. We published this study in Journal of Biological Chemistry with me as the last corresponding author. Besides, we extended miRNA studies to other cancer types to evaluate their contribution in tumor biology. We found that overexpression of miR-21 in tongue squamous cell carcinomas (TSCC) contributes to reduced apoptosis and increased growth of the tumor cells via targeting TPM1 and PTEN. We also showed that reduction in miR-200 and miR-15b is responsible for chemotherapy induced epithelial-mesenchymal transition (EMT) in TSCCs, leading to a more malignant phenotype and cancer metastasis. Additionally, our study in hepatocellular carcinomas showed that an oncoprotein of hepatitis B virus, HBX, results in reduction in miR-16 family members and is critical for HBX induced transformation of hepatocytes. To support my studies on miRNA and cancer biology, I have played a leadership role in “973 program project” from Ministry of Science in China and “Key project of National Natural Science Foundation of China (NSFC)” from NSFC since 2008. I am also a principal investigator of “A3 Foresight Project” from NSFC, an international collaboration program project in frontier science among China, Korea and Japan. Besides, I was invited to give talks on miRNA and cancer at Harvard Medical School, University California Los Angelus, Cambridge University UK etc. I also chaired international sessions in miRNA in both the 68th and 69th Annual Meeting of Japanese Cancer Association. 3. Identifying a functional receptor for CCL18 on breast cancer cells To further evaluate how the inflammatory microenvironment impacts on miRNA expression and function in breast cancer cells, we studied the interaction between breast cancer cells and tumor associated macrophages (TAMs), the most abundant infiltrated leukocytes in breast cancer stroma. After screening a panel of cytokines associated with M2 macrophages, alternative activated macrophages identified in breast cancer stroma before, we found that CCL18, a chemokine specifically produced by leukocytes from monocytic lineage, is the most abundantly produced cytokine in TAMs. We further showed that CCL18 promotes cancer cell migration and invasion and is associated with clinical metastasis and poor prognosis of breast cancer patients. However, receptors for CCL18 have not been reported before, which hinders further mechanistic studies on CCL18 effects. To identify CCL18 receptor on breast cancer cells, we exploited immunoprecipitation, mass spectrum and isotope competing assay. We found that CCL18 specifically binds to PITPNM3, a membrane protein. Using functional assays for identification of G protein-coupled receptor (GPCR), we found that CCL18-PITPNM3 interaction triggers calcium influx, calcium pathway activation and chemotaxis for PITPNM3 transfected cells. Therefore, we identified that PITPNM3 is a functional CCL18 receptor on breast cancer cells and silencing PITPNM3 abrogates CCL18 effects in breast cancers. These data were published in Cancer Cell in 2011, and the paper was selected as a featured article of the journal issue and was highlighted by Nature. Alberto Mantovani contributed a preview for our paper and indicated that “This finding is provocative, and shed new light on the role of chemokine in cancer and raise important questions. CCL18 may be added to the list of potential therapeutic targets.” Indeed, our patent “Inhibiton of breast cancer invasion and metastasis by targeting CCL18” has been granted by the Chinese Bureau of Patent Right. Our further study demonstrated that CCL18 induces EMT in breast epithelial cells via PI3K/Akt and MAPK pathways. More importantly, TAMs release CCL18 and result in changes of miRNA expression profile in breast cancer cells, which maintains persistent activation of the signaling pathways and EMT in the tumor cells. Other than cancer cells, our further study also characterized that PITPNM3 also serves as a functional CCL18 receptor in leukocytes, including T and B lymphocytes. Our study on the microenvironment of breast cancers suggested that TAMs not only produce cytokines to act on tumor cells, but also release miRNAs that are enwrapped in microvesicles and shuttled into tumor cells. We found that miR-223, a macrophage specific miRNA, can be released by M2 macrophages and shuttled into co-cultured breast cancer cells. By targeting the Mef2c/-catenin pathway, miR-223 promotes invasion of the tumor cells. Therefore, we confirmed that tumor microenvironment may result in miRNA changes in cancer cells and thus regulate their biology. 4. Targeted delivery of siRNA in vivo for therapeutic purpose Our studies on miRNA and breast cancer biology imply tremendous therapeutic potential of small RNAs against the malignancy. However, a major obstacle to developing small interfering RNAs (siRNAs) as cancer drugs is their intracellular delivery to disseminated cancer cells in vivo. To overcome this obstacle, I pioneered the use of fusion proteins of fragmented antibodies and positively charged peptides to deliver siRNAs into specific target cells in vitro and in vivo. To prove the principle, we designed a fusion protein (F105-P) of the heavy chain Fab fragment of an HIV-1 envelope antibody and protamine to deliver siRNA to HIV-infected or envelope-transfected cells. siRNAs bound to F105-P induced silencing only in cells expressing HIV-1 envelope, and siRNAs targeted against the HIV-1 capsid gene inhibited HIV replication in HIV-infected T lymphocytes. This study was published in Nature Biotechnology in 2005, and was selected as a cover story “RNAi Magic Bullet” of the issue and highlighted by Nature. Until August 2012, ISI citation of the paper reached 444 times. Further, we employed a Her2-ScFv -protamine peptide fusion protein (F5-P) to deliver siRNAs targeting to Polo-like kinase 1 (PLK-1), a highly conserved serine-threonine kinase that promotes cell division, into Her2+ breast cancer cell lines and primary human cancers in orthotopic breast cancer models. PLK1-siRNAs transferred by F5-P inhibited target gene expression, reduced proliferation and induced apoptosis of Her2+ breast cancer cell lines and primary human cancer cells in vitro without triggering an interferon response. Intravenously injected F5-P/PLK-1 siRNA complexes concentrated in orthotopic Her2+ breast cancer xenografts and persisted for at least 72 hr, leading to suppressed PLK1 gene expression and tumor cell apoptosis. The intravenously injected siRNA complexes retarded Her2+ breast tumor growth, reduced metastasis and prolonged survival without evident toxicity. Our data suggest that F5-P could be used to deliver siRNAs to treat Her2+ breast cancer. This study has recently been published by Science Translational Medicine and a related patent “The application of a fusion protein of single-chain fragmented antibody and protamine peptide” was granted in 2011. In addition to the fusion protein, we also employed an innovative “two-in-one” micelleplex approach based on micellar nanoparticles of a biodegradable triblock copolymer to simultaneously deliver siRNA and chemotherapeutic drugs into breast cancer cells in vivo. The siRNA and chemotherapeutic drugs carried by the micelleplex can induce a synergistic tumor suppressing effects in cancer xenografts inoculated in immunocompromised mice. 5. Future research plan In the next five to ten years, my research focus will be placed on long noncoding RNAs (lncRNAs) that are more than 200 nucleotides and breast cancer biology. LncRNAs are involved in regulation of multiple major biological processes impacting development, differentiation, metabolism, oncogenesis and so on. In contrast to small ncRNAs such as miRNA and siRNA that are highly conserved across species, lncRNAs are poorly conserved and involved in regulating gene expression and function that are not yet fully understood. The general concept of lncRNA mechanism is that the ncRNAs may act as decoys, guides and scaffolds to control every level of the gene expression program. However, our preliminary data suggest that lncRNAs also play a role in signal transduction of major signaling pathways. We demonstrate that the expression of a lncRNA, NinR1, is upregulated in breast cancer cells by inflammatory stimuli that induce NFB signaling. However, silencing NinR1 that is transcribed by the active NFB promotes inflammation induced NFB activation, leading to enhanced EMT and reduced apoptosis in tumor cells and increased cancer metastasis in vivo. Conversely, ectopic NinR1 expression inhibits NFB activation and the ensuing cancer invasion and metastasis, suggesting NinR1 is a negative regulator of NFB signaling. Indeed, NinR1 prevents nuclear translocation and activation of NFB by interacting with the NFB complex in the cytoplasm and inhibits IKK-induced phosphorylation and degradation of IB. We further identify that NinR1 harbors at its 5’-region a mimicry of NFB binding motif as B DNA that interacts directly with the N-terminal domain (NTD) of the Rel homology region (RHR) in p65, a most abundant NFB subunit. Further, the extending 3’-region of the p65-engaged NinR1 masks the sites of phosphorylation on IB to inhibit its phosphorylation by IKK. These findings indicate a new mechanism by which lncRNAs interact with major signaling molecules to directly modulate their activation and participate in the auto-regulatory feedback circuitry of signaling pathways in cancer cells, thus uncovering a previously unappreciated crosstalk of lncRNAs and cellular signaling that underlies cancer-related inflammation. Based on our findings in lncRNAs, we will further screen for lncRNAs that directly interact with signaling molecules and investigate their roles in regulating the activation of major signaling pathway and their contributions in breast cancer development. We will also evaluate whether lncRNAs may also serve as treatment targets that can be cleaved by therapeutic siRNAs and thus may break new grounds in breast cancer therapies. References 1. Fire, A., Xu, S., Montgomery, M. K., Kostas, S. A., Driver, S. E., and Mello, C. C. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391: 806-811, 1998. 2. Mello, C. C., and Conte, D., Jr. Revealing the world of RNA interference. Nature 431: 338-342, 2004. 3. Song, E., Lee, S. K., Wang, J., Ince, N., Ouyang, N., Min, J., Chen, J., Shankar, P., and Lieberman, J. RNA interference targeting Fas protects mice from fulminant hepatitis. Nat Med 9: 347-351, 2003. 4. Song, E., Zhu, P., Lee, S. K., Chowdhury, D., Kussman, S., Dykxhoorn, D. M., Feng, Y., Palliser, D., Weiner, D. B., Shankar, P., Marasco, W. A., and Lieberman, J. Antibody mediated in vivo delivery of small interfering RNAs via cell-surface receptors. Nat Biotechnol 23: 709-717, 2005. 5. Yu, F., Yao, H., Zhu, P., Zhang, X., Pan, Q., Gong, C., Huang, Y., Hu, X., Su, F., Lieberman, J., and Song, E. let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell 131: 1109-1123, 2007.

Abstracts this author is presenting: