Elucidating the mechanism of EWS-FLI1 induced oncogenesis
- Michelle Renee Marques.
- Sept. 2012.
- Physical description
- online resource (xv, 215 pages) : illustrations (some color)
- Marques, Michelle Renee.
- Cleary, Michael L. thesis advisor.
- Kuo, Calvin Jay. thesis advisor.
- Longaker, Michael T. thesis advisor.
- Sweet-Cordero, Eric thesis advisor (primary).
- Stanford University. Program in Cancer Biology.
- Stanford University. Committee on Graduate Studies. degree grantor.
- Includes bibliographical references.
- Ewing's sarcoma is the second most common malignant bone cancer in children. The prominent defining feature of Ewing's sarcoma is a translocation event between a member of the FET family of RNA binding proteins and a member of the Ets transcription factor family. The majority of patients have a translocation event between the EWSR1 gene and the FLI1 gene. The EWS-FLI1 translocation was first discovered in 1992 and to date, the mechanism by which EWS-FLI1 induces the formation of Ewing's sarcomas remains unclear. Understanding the role of EWS-FLI1 in oncogenesis is critical for Ewing's sarcoma and would have broad implications for other cancers as well. Translocations involving members of the FET or Ets families are also found in leukemia, prostate cancer and other sarcomas. A primary goal of my graduate work has been to develop tools to express EWS-FLI1 in primary human cells as well as in genetically engineered mice to understand how EWS-FLI1 induces oncogenesis and determine the cell of origin in Ewing's sarcoma. As recent work suggested that Ewing's sarcomas arise from a mesenchymal stem/progenitor cell (MSC), we examined the effects of EWS-FLI1 expression in primary human MSCs. We isolated MSCs from pediatric patients at Lucile Packard Children's Hospital to establish human bone marrow derived MSC lines (which we call HBMs). Through a series of experiments, we learned that the precise expression levels of EWS-FLI1 were critical in determining the effect of this oncogene on primary cells. High expression of EWS-FLI1 was not tolerated in HBMs. In contrast, when expressed at lower levels, stable EWS-FLI1 expression was maintained in HBMs. To elucidate transcriptional targets of EWS-FLI1 in HBMs, we used next-generation sequencing (RNAseq) to identify genes dysregulated by EWS-FLI1. Using this approach we identified 170 targets that constitute an EWS-FLI1 expression signature, including novel target genes. Expression of a subset of these genes was dependent on EWS-FLI1 expression in Ewing's sarcoma cell lines, validating their regulation by EWS-FLI1. The majority of these target genes were required for growth in soft agar of Ewing's sarcoma cell lines and some also showed an effect on cell growth. Among these EWS-FLI1 target genes we focused on a novel long non-coding RNA, lnc277, which is induced and regulated by EWS-FLI1 in Ewing's sarcoma cell lines and in other human cell lines ectopically expressing EWS-FLI1. Expression of lnc277 is highly specific to Ewing's sarcoma and is required for cell growth and transformation by EWS-FLI1. To decipher a mechanism for how lnc277 functions in Ewing's sarcoma cells, we have used protein arrays to identify interacting proteins. Lnc277 appears to interact with several proteins involved in transcription, splicing, RNA stability and translation, including STAU1, HNRPK1 and several others. Additionally, we performed RNAseq analysis of lnc277 knock-down to identify specific genes whose expression is altered upon depletion of lnc277. To elucidate the cell of origin for Ewing's sarcoma and create a model that can be used to test novel strategies for treatment, we have genetically engineered mice to conditionally express the EWS-FLI1 translocation from the endogenous EWSR1 locus. We have generated mice that contain lox sites within both the EWSR1 locus and the FLI1 locus such that upon Cre recombinase expression, some cells will undergo a reciprocal recombination event, generating both the EWS-FLI1 and FLI1-EWS chromosomes. We have genomic DNA and mRNA confirmation that this recombination occurs in vitro and in vivo after expression of Cre recombinase. This is the first example to our knowledge of a mouse model that faithfully recapitulates a translocation mechanism in a solid tumor. The reciprocal translocation model relies on two chromosomes recombining with each other, an event that we have found to be highly rare with these two chromosomes in the mouse. Therefore, we focused our efforts on a second mouse model where the recombination event occurs much more efficiently, our EWS-FLI1-V5 mouse model. The EWS-FLI1-V5 mouse model expresses a V5-epitope tagged version of EWS-FLI1 also from the EWSR1 locus. To create this model, a FLI1 cDNA was introduced downstream of the EWSR1 gene on the same chromosome. The expression of Cre recombinase results in the formation of the translocation by splicing the N-terminal EWSR1 exons to a FLI1 cDNA containing the C-terminal exons. This model leads to expression of EWS-FLI1-V5 in the majority of cells where Cre is expressed. We have carried out in vitro studies expressing EWS-FLI1-V5 in mouse embryo fibroblasts (MEFs) and mouse MSCs. Whereas EWS-FLI1-V5 expression inhibits proliferation in MEFs, MSCs expressing EWS-FLI1-V5 continue to proliferate. We have demonstrated that several of the new target genes identified in the human system were also regulated by EWS-FLI1-V5 in mouse cells. We have crossed both our Ewing's sarcoma mouse models to four Cre strains that express Cre recombinase in mesenchymal tissues as well as one that expresses Cre recombinase in the neural crest lineage. Mice from the reciprocal translocation model failed to develop tumors, most likely because the translocation event was so rare either no cell recombined the EWSR1 and FLI1 loci or that EWS-FLI1 expression was not tolerated in the cells that did recombine the loci. The EWS-FLI1-V5 mice expressing EWS-FLI1 in the mesenchymal lineage using Dermo1-Cre, Col1[alpha]2-Cre, Prx1-Cre or Sox9-Cre died embryonically. Interestingly, we only obtained mice that could potentially be expressing EWS-FLI1-V5 in the neural crest lineage using P0-Cre, suggesting the expression of EWS-FLI1-V5 in these cells was not toxic or that other cells can compensate for loss of the cells expressing EWS-FLI1-V5. Whether these adult mice actually express EWS-FLI1-V5 in the tissues derived from the neural crest lineage and whether these mice are tumor prone are areas for future study. Through this thesis work, we have used a combined approach that leverages both human and mouse model systems to create an in vivo model of Ewing's sarcomagenesis. These models could be used to define the cell of origin for Ewing's sarcoma and gain an understanding of the genetic requirements for oncogenesis downstream of EWS-FLI1. Through our studies of pediatric human mesenchymal stem cells expressing EWS-FLI1 in Chapters 2 and 3, we have discovered a number of novel EWS-FLI1 target genes and identified a lncRNA that is highly specific to and required for EWS-FLI1 mediated oncogenesis. In Chapters 4 and 5, two novel transgenic mouse strains were generated to express the EWS-FLI1 gene fusion from the endogenous EWSR1 locus in a way that is physiologically relevant to Ewing's sarcoma. These tools should help define the effects of EWS-FLI1 expression in primary and cancer cells and hopefully result in new therapies to benefit children diagnosed with this disease.
- Publication date
- Submitted to the Program of Cancer Biology and the Committee on Graduate Studies of Stanford University.
- Thesis (Ph.D.)--Stanford University, 2012.