2011 Innovative Research Grant 24-Month Progress Reports
Targeting MLL in Acute Myeloid Leukemia
Yali Dou, Ph.D., University of Michigan
Our broad objective in the proposed research is to develop novel chemotherapeutic agents that target the activity of a regulator of a subtype of acute myeloid leukemia, namely the Mixed Lineage Leukemia (MLL) protein. MLL was originally cloned by its direct involvement in a group of distinct human acute leukemia with extremely poor prognosis. MLL gene abnormalities account for 5% to 10% of the disease, and at least 70% of the cases in infants under 1 year old. It is general consensus that MLL mutations disrupt expression of specific genes that are important in early blood cell development. MLL is an enzyme and its activity is essential for leukemia development. Biochemical analyses have shown that MLL activity is tightly regulated by several interacting proteins. Therefore, it is conceivable that disrupting these protein-protein interactions involving MLL will compromise MLL enzymatic activity, which in turn leads to inhibition of leukemogenesis. Using the biochemistry and medicinal chemistry approaches, we have designed a series of inhibitors that target the MLL activity. In the past several months, we have made significant progress in improving our lead compounds in both in vitro and in vivo assays. These results suggest that our approach is valid and is likely to provide new therapeutics for MLL mediated leukemia.
Targeting Genetic and Metabolic Networks in T-ALL
Adolfo A. Ferrando, M.D., Ph.D., Columbia University
Acute lymphoblastic leukemia is the most frequent cancer in children. Despite much progress in the treatment of this disease, leukemia still represents a clinical challenge, particularly in cases diagnosed with T-cell disease. In this project, we aim to elucidate the oncogenic circuitries that control T-cell acute lymphoblastic leukemia. Our ultimate objective is to identify effective new drugs and drug combinations for the treatment of this disease.
In the first two years of funding we have analyzed a highly representative panel of human T-cell leukemia samples to catalog their genetic alterations, genetic programs and metabolic signatures and exploited leukemia specific regulatory circuitries to identify new active drugs and drug combinations. Our results have identified and cataloged two molecular groups of T-cell leukemia characterized by different gene expression programs; identified numerous new genes mutated in T-ALL including ETV6, RUNX1, EZH2 and SUZ12. Most notably these genetic alterations provide new biomarkers for the identification of high risk patients in the clinic. We are currently extending these studies to the analysis of epigenetic lesions and testing the effects of DNA methylation in leukemia development and therapy response.
Following on these results and to gain better understanding of the mechanisms of drug resistance we have extended our mutation analyses to relapsed leukemias. These studies have identified new recurrent mutations that activate NT5C2, a metabolic gene responsible for the inactivation of mercaptopurime, an essential drug in the treatment of T-ALL. This result highlights the importance of cell metabolism in the response to therapy. Ongoing analyses have extended these studies to a broader panel of relapsed tumors uncovering over 180 new mutations associated with leukemia relapse.
A central component of this research is the analysis of genetic and metabolic networks. Using this approach we have uncovered the mechanistic role of TLX1 and TLX3, two major genes driving T-cell leukemia. Moreover, analysis of the circuitries involved in resistance to chemotherapy with glucocorticoids has identified the PI3K-AKT1 pathway as a new therapeutic target for the reversal of resistance to glucocorticoids, a key drug in the treatment of T-ALL. Applying these principles and approaches to the study of NOTCH1 and PTEN, the two most critical factors in T-ALL development, we have uncovered the mechanisms mediating resistance to anti-NOTCH1 therapy in this disease. Strikingly, these results pointed to new drugs and drug combinations for the treatment of leukemia. Ongoing testing of these drugs in highly aggressive leukemias shows strong synergism with anti NOTCH1 therapies, opening the way towards the development of new combinations in the clinic.
Finally, and along this line, we have performed global metabolic profiling of T-cell leukemias and shown that targeted drugs that inactivate NOTCH1 result in dramatic changes in cell metabolism. Strikingly inactivation of PTEN induces reprogramming of cell metabolism and effectively reverses the metabolic shutdown resulting from NOTCH1 inhibition. Most notably these analyses have uncovered cell metabolism as an important Achilles heel in leukemia.
Overall, we have made significant progress towards our goal of using high throughput technologies and network analyses to identify key regulators of leukemic cell growth, and survival and to develop novel and highly effective targeted in this disease.
Targeting Protein Quality Control for Cancer Therapy
Estela Jacinto, Ph.D., University of Medicine & Dentistry of New Jersey - Robert Wood Johnson Medical School
The normal growth and proliferation of cells is orchestrated by a cascade of events that is initiated by binding of a stimulus to a receptor at the membrane. Once triggered, the receptor communicates to the rest of the cell via recruitment of a number of signaling molecules. Depending on the quality and quantity of signals from the receptor, the cellular output can be modified, for example proliferation versus death. Signals from growth receptors on the cell surface can become altered in cancer due to either increased expression of these receptors or mutations that lead to increased activity. In our project, we are addressing how inhibition of the expression of the epidermal growth factor receptors (referred in here as ErbB) can be exploited for cancer therapy. Our lab had initial findings that a protein complex called mTORC2 is involved in protein production and quality control. When mTORC2 is inhibited by pharmacological agents or by genetic manipulation, proteins that are known to become deregulated in cancer such as Akt and ErbB have defects in their synthesis. During the first year of this grant, we have established a role for mTORC2 in controlling the amount and quality of ErbB1 that is expressed in the surface of breast cancer cells. In the second year (July 2012-June 2013), we have identified a possible mediator of the mTORC2 function in ErbB1 quality control. Using protein purification and mass spectrometry, we identified a protein called GFAT1. This protein has been previously characterized in the field of diabetes since it is involved in cellular metabolism. Not much is known about how this protein becomes regulated by nutrients. Our findings now provide a connection between cellular proliferation (via ErbB1 signaling) and metabolism (GFAT1) and that these two pathways could be coupled by mTORC2. In the last six months, we have analyzed how mTORC2 can regulate GFAT1. We found that GFAT1 expression is diminished upon mTORC2 disruption and that GFAT1 is likely phosphorylated by mTORC2 during translation in order to stabilize this protein. We have identified a possible target site in GFAT1 by mTORC2. In the coming year, we will analyze how the regulation of GFAT1 by mTORC2 at this site plays a role in the control of ErbB1 expression. We have also identified another receptor, CD147, that is controlled by mTORC2 in a similar way. CD147 has been indentified in the past as a marker of tumor metastasis and is required for lactate transport in the cell. Therefore, this receptor could play a crucial role in the changes in metabolism that occurs in cancer cells. Since we found that CD147 is highly dependent on mTORC2 for proper expression, we will alternatively examine this receptor, along with ErbB1, to elucidate the function of mTORC2 in protein quality control.
Targeting PP2A and the Glutamine-Sensing Pathway as Cancer Treatment
Mei Kong, Ph.D., City of Hope
Fast-growing cancer cells rely on enhanced nutrient uptake to grow and divide. However, as tumors grow, increased uptake of nutrients and poor vascularization often lead to nutrient deprivation in tumor cells. Understanding the molecular mechanisms that promote cancer cell survival under poor nutrient conditions is important for developing new drugs that could starve tumor cells and block cancer progression. The amino acid glutamine is a major nutrient that supports cell growth and survival. Solid tumors consume glutamine at a rate that outstrips its supply and inevitably end up facing low glutamine conditions. The goal of this project is to determine the molecular basis for tumor cell survival under conditions of glutamine deprivation in order to develop novel drugs targeting this pathway. We have shown that the enzyme PP2A (protein phosphatase 2A) plays a critical role in mediating cell survival upon glutamine deprivation. However, PP2A is a member of a large family of protein complexes that regulate many different cellular functions. In this study, we worked to identify the specific PP2A complex that regulates cancer cell survival upon glutamine deprivation. Our aims are to determine: (1) whether PP2A complexes are regulated by glutamine levels; (2) the mechanism by which PP2A exerts a cell survival effect during glutamine deprivation; (3) whether PP2A contributes to tumor cell survival and whether impairment of PP2Aactivity combined with inhibition of glutamine metabolism can alter cancer cell viability.
During the first eighteen months, we have demonstrated that only the regulatory subunit B55α is strongly and selectively upregulated in response to glutamine withdrawal, thereby triggering the formation of an active PP2A complex consisting of catalytic C, scaffolding A subunits, and the specifically induced B55α subunit. This B55α-containing PP2A complex is critical for cancer cell survival upon glutamine deprivation. We further demonstrated that glutamine deprivation results in activation of p53, an important sensor of metabolic stress, and that B55α-mediated cell survival is p53-dependent. In these six months following the previous progress report in December of 2012, we have successfully fufilled two pre-defined milestones. First, we determined if α4 and B55α promote cell survival upon glutamine deprivation via inhibition of c-Myc activity. Our data shows that B55α functions through regulation of p53, but not c-Myc activity. Second, we identified that B55α enhances cell survival upon glutamine deprivation via a ROS-dependent mechanism. In the next funding period, we will continue experiments outlined in the aims proposed for this funding period in “milestones and deliverables,” which is to identify the specific substrate of the B55α complex upon glutamine deprivation.
Chimeric RNAs Generated by Trans-Splicing and Their Implications in Cancer
Hui Li, Ph.D, University of Virginia
Substantial progress we made in the last 6 months are summarized in the following:
For aim1: identification of additional trans-splicing events in both normal and cancer cells, we have found the presence of PAX3-FOXO1 (FKHR) during stem cell differentiation process. The fusion has been thought to be a unique feature of alveolar rhabdomyosarcoma, a common childhood cancer. We found the same fusion product in fetal muscle samples as well. These findings further challenge the traditional dogma that gene fusions are unique to cancer. In addition to some functional evidence, we have come to the conclusion that such chimeric RNA is not unique to the tumor, it is expressed in normal muscle development process and serves important physiological function, and that its generating mechanism in the normal cells is independent of chromosomal translocation, which is the mechanism for the fusion production in alveolar rhabdomyosarcoma. Unfortunately, the manuscript was not accepted by Nature. We have since submitted half of the story (that PAX3-FOXO1 exists in normal myogensis) to Cancer Discovery and it is under review now. We plan to carry other additional functional study of the chimera in normal muscle development and submit this half to a stem cell journal.
It turns out that PAX3-FOXO1 chimera is uniquely expressed during the myogensis process, as no signal of BCR-ABL fusion (associated with CML) or EWS-FLI1 (associated with Ewing sarcoma) was detected in these samples.
The PAX3-FOXO1 gene fusion is a prominent marker of ARMS and detection of PAX3-FOXO1 fusion RNA by RT-PCR is a standard diagnostic procedure. Our findings of the presence of PAX3-FOXO1 RNA in normal cells raise concerns for false positive diagnoses. Additionally, therapies targeted at the fusion protein may have side effects due to disruption of functions performed by PAX3-FOXO1 in normal developing muscle. Knowledge about the mechanism and the cells that express the fusion products could lead to more specific diagnostic methods with fewer false positives and treatment strategies with less side effects. In addition, knowing the temporal and kinetic expression characteristic and pattern of PAX3-FOXO1 in normal cells will shed light on the etiology of the tumors (ARMS maybe the result of continuous expression of PAX3-FOXO1 in combination with other oncogenic “hits”)
Aim2 is designed to study the implications of chimeric gene fusions in cancer. In the field of endometrial cancer research, a big caveat of using mouse as a model is that mouse as well as most mammals do not have menstrual cycle. By accident, we found that the fusion JAZF1-JJAZ1 we have been studying is unique in species that have menstrual cycle. We have found that the fusion is necessary for the process. To test whether it is sufficient to trigger the process, we used human foreskin fibroblast cells and found that the fusion can induce at least two markers associated with decidualization, a trigger for menstruation. We also found that when combined with EZH2 and EED, two core components of Polycomb Repressive Complex, we can enhance the effect. These surprising findings have led us to hypothesize that by inducing the fusion at right time, we may be able to generate a mouse model that go through menstrual cycle. Such a tool will be extremely useful not only for endometrial cancer or breast cancer research, but also for any research fields related to menstrual cycle. We are now testing the hypothesis in cell culture models. If successful, we will generate a transgenic mouse model. Such a model will allow us to study the fusion’s oncogenic effect when expressed continuously, and test whether the animal will menstruate if the fusion is induced at the right time.
Exome sequencing of melanomas with acquired resistance to BRAF inhibitors
Roger Lo, M.D., Ph.D., Santa Monica-UCLA Medical Center and Orthopaedic Hospital
A small molecule (PLX4032/vemurafenib/Zelboraf) targeting a common melanoma mutation, V600EB-RAF, has shown unprecedented promise in advanced clinical trials (80% of patients respond if their tumors harbor the V600EB-RAF mutation) and confers survival benefit, prompting FDA approval. However, its ultimate success is challenged by so-called acquired drug resistance, which leads to clinical relapse. This type of drug resistance that develops over time occurs within months to years of drug initiation and cuts short the “sudden reprieve” that awakens patients’ hope for a cure (see NY Times stories by Amy Harmon on December 22-24, 2010). Earlier, we reported in Nature the discovery of two means by which melanomas escape from vemurafenib, which suggest new treatment strategies that are testable in clinical trials. This study along with others gave us another insight, that is, melanomas likely use a variety of different ways to escape from B-RAF inhibitors. Discovering other mechanisms of acquired resistance is logically the first step in constructing a therapeutic strategy closer to a cure.
We set forth three research aims centered on this group of V600EB-RAF-positive melanomas treated with B-RAF inhibitors (vemurafenib as well as another competing B-RAF inhibitor, GSK2118436). These aims are based on several premises. First, we need to directly study precious tissues derived from clinical trial patients. Second, we need to enlarge this tissue collection by collaborating among distinct clinical sites. Third, because finding a specific mechanism among the myriad of cancer-related changes is akin to finding a needle in a haystack, we should capitalize on the latest, “high-throughput” genomic technologies. Here, we report assembling a collaboration of multiple clinical sites to study acquired resistance directly in tissue samples from patients. For each patient that participates in this study, we are obtaining a set of normal tissue (e.g., blood), melanoma tissue before drug treatment, and melanoma tissue after an initial shrinkage followed by re-growth. Each set of tumor samples is first studied for the existence of known mechanisms which we have already discovered and characterized with in-depth molecular details in laboratory models. Works along this line have been published recently (Poulikakos et al, Nature, Nov 2011; Shi et al, Nature Communications, March 2012; Shi et al, Cancer Discovery, April 2012) or are currently under peer review for publication likely early in 2014 (Shi, Hugo,…, Lo, 2013; Shi, Hong,…, Lo, 2013). This workflow culls out tumor sample sets or patients for detailed genetic analysis. By harnessing the speed of “next-generation” DNA sequencing technology, we are examining the whole exome or the protein-coding, “business end” of the melanoma genomes for key genetic alterations that account for acquired resistance to B-RAF inhibitors in melanoma. From a patient’s perspective, we can now claim we know how melanomas escape from BRAF inhibitors in over 70% of patients. This knowledge has generated additional hypotheses to improve therapeutic response (both number of responders and duration of response) which are being tested in a clinical trial (1) or will be tested in two additional clinical trials (2-3) currently under review.
1. Safety and efficacy of the AKT inhibitor GSK2141795 in combination with the BRAF inhibitor dabrafenib in patients with BRAF mutant metastatic melanoma and study of the non-MAPK pathway resistance. Phase I/II. 2. A randomized phase II trial of intermittent versus continuous dosing of dabrafenib and trametinib in BRAFV600E/K mutant melanoma (SWOG-sponsored; study chairs: A. P. Algazi, A. I. Daud, & R. S. Lo) 3. Biomarkers of durable response with intermittent therapy with LGX818 and MEK162 combined therapy in patients with BRAF mutant metastatic melanoma (UCLA investigator-initiated; PIs: A. Ribas & R. S. Lo)
Going forward, as we recruit patients for these next-generation clinical trials, we will need to iteratively sample tumor tissues donated by these patients in order to derive further knowledge and improve upon therapeutic outcomes. The SU2C Bud and Sue Selig Innovative Research Grant (IRG) has thus allowed us to take one large step forward in the treatment of 50-60% of all melanoma patients. The success of this research funding and the urgency of the next-step questions should allow us to compete for additional funding to accelerate the next big step forward.
Identification and Targeting of Novel Rearrangements in High-Risk ALL
Charles G. Mullighan, MBBS(Hons), MSc, MD., St. Jude Children’s Research Hospital
Acute lymphoblastic leukemia (ALL) is the commonest childhood cancer, and the leading cause of non-traumatic death in children and young adults. This project has focused on a recently described subtype of ALL termed “BCR-ABL1-like” or “Ph-like” ALL characterized by a range of previously unknown chromosomal changes and mutations that result in activation of cellular growth signals called kinases. Ph-like ALL is common, comprising up to15% of childhood ALL and up to one third of ALL in adolescents and young adults, and associated with a high risk of treatment failure, hence new therapeutic approaches to improve treatment outcomes are required. Work supported by the Stand Up to Cancer Innovative Research Grant has supported genetic analysis of leukemia cells from patients with Ph-like ALL in order to identify the range of genetic alterations in this disorder, to examine the frequency of these changes in large cohorts of ALL patients, and to examine their role in the development of leukemia, and potential responsiveness to therapy
The first aim of this project is to use genomic sequencing and recurrence testing analysis to determine the nature and frequency of kinase activating genetic alterations in children and young adults with ALL. An initial pilot study that used mRNA-sequencing and whole genome sequencing of leukemia cells from 15 children with Ph-like ALL and identified a range of genetic changes activating kinases including CRLF2, ABL1, JAK2, PDGFRB, IL7R, and SH2B3 (LNK). We have now expanded the profiling of cases to define the frequency of Ph-like ALL, and have tested the frequency of each of these changes in cohorts of childhood and adolescent and young adult (AYA) ALL, with current numbers of these cohorts exceeding 1500. The changes identified by sequencing of the first 15 cases were present in 80% of the recurrence cohorts. To identify the kinase-activating alterations in the remaining cases, we are performing mRNA-sequencing, exome sequencing and whole genome sequencing in all cases with suitable genetic material lacking one of the kinase-activating alterations identified in the pilot project (mRNA-seq and whole genome sequencing in almost 100 patients). This second phase of sequencing is complete, and analysis is nearing completion. This has already identified new fusion partners of the known kinases (e.g. ABL1, JAK2 and PDGFRB) and importantly, has identified new kinases as targets for rearrangement, (EPOR, ABL2, AKT2, STAT5B). The next 6 months of the project will witness completion of this sequencing analysis and testing for recurrence of each new alteration.
The second aim of this project was to develop experimental models to examine the way in which the alterations identified in aim 1 contribute to the development of leukemia, and to develop experimental systems to test the potential effectiveness of TKIs. Using laboratory cell lines, I have shown that several of these alterations accelerate cell growth and activate downstream signaling pathways. In addition, alterations such as EBF1-PDGFRB induce leukemia when expressed in mouse bone marrow cells. We have also developed xenograft models in which human leukemia cells are grown in immunodeficient mice. Importantly, growth of these cell lines is inhibited by several TKIs including include imatinib (Gleevec), dasatinib (Sprycel) and ruxolitinib (Jakafi). We are establishing xenografts of additional tumors to test the activity of other targeted agents.
Together, these studies continue to identify the range of lesions underlying BCR-ABL1-like ALL, and show that these alterations directly contribute to the development of leukemia. Importantly, the experimental models show that these alterations are targetable with TKIs. These results have generated tremendous excitement in the ALL field, and efforts to identify patients harboring these lesions at diagnosis and to treat them with these drugs are already underway.
A Systems Approach to Understanding Tumor Specific Drug Response
Dana Pe’er, Ph.D., Columbia University Medical Center
We propose using genomic technologies to track tumor response to potent drug inhibition of critical pathways across a diverse tumor panel. We will develop cutting-edge computational machine learning algorithms to piece these data together and illuminate how a cell’s regulatory network processes signals, and how this signal processing goes awry in cancer. By studying a large panel of diverse tumors we can begin to piece together general principles and patterns in response to drug. These studies should teach us what drives cancers and what part of the networks we should target. For each individual patient, we wish to determine the best drug regime for that individual, informed by a model that can predict tumor response to drugs and their combinations. Treatment that is based not only on understanding which components go wrong, but also how these go wrong in each individual patient, will improve cancer therapeutics.
At the end of the second year we are have made some major progress in our research which include:
1. A potential new combination therapy for melanoma, currently being tested in mice, that can potentially offer response to many patients who are resistant to PLX alone (the current BRAF inhibitor in use), most exciting combination might also offer possible therapeutics for NRAS melanoma patients, who currently have no course of personalized care. 2. We have gained a much better understanding of why and which melanoma cells either die or not follow treatment. This understanding suggests a number of potential avenues to target with drugs. 3. We have made first steps towards understanding a mechanism through which some melanoma cells evade drugs and continue to grow. 4. Importantly, the 3 findings above were discovered using new tools that could be applied to additional cancers and drugs.
Targeting Sleeping Cancer Cells
Sridhar Ramaswamy, M.D., Harvard Medical School
Cancer cells of different types have the very strange ability to go to sleep and then eventually wake up. While cancer cells sleep they are highly resistant to virtually all currently available forms of treatment. However, we do not understand how highly aggressive cancer cells can become dormant. It has proven extremely difficult to study these cells directly in patients and we have lacked suitable model systems to study them in the laboratory. We recently made a remarkable observation, however, that has the potential to open this important area for new investigation. We found that highly aggressive cancer cell lines of various types occasionally produce dormant cells. We went on to develop reliable methods for the prospective identification, isolation, molecular tracking, and experimental study of these “G0-like” dormant cancer cells in human cancer cell lines. Our preliminary results raised the possibility that epigenetic or signaling networks regulate these spontaneously dormant cancer cells. With a SU2C-AACR Innovative Grant Award, we have been using cutting edge molecular and cellular biology and genomic (next-generation sequencing (ChIP-seq / RNA-seq)), proteomic (reverse-phase protein microarrays), and computational technologies to identify and validate 1) genetic and 2) protein signaling networks that might trigger and maintain cancer cell dormancy.
Since the start of the award, we have made tremendous progress (see Dey-Guha, PNAS 108:12845 (2011)). Importantly, we have found that rapidly proliferating cancer cells can divide asymmetrically to produce slowly proliferating “G0-like” progeny that are enriched following chemotherapy in breast cancer patients. Asymmetric cancer cell division results from asymmetric suppression of AKT1 kinase signaling in one daughter cell during telophase of mitosis. Moreover, inhibition of AKT signaling with allosteric small-molecule inhibitors can induce asymmetric cancer cell division and the production of slow proliferators.
Most recently, we have discovered that AKT1 (rather than AKT2 or AKT3) is both necessary and sufficient for entry into the G0-like cell state. Moreover, AKT1 signaling is suppressed by suppression of AKT1 total protein levels via an mTORC2-induced, TTC3 / proteasome-mediated degradation pathway. In addition, RNA-seq studies suggest that G0-like cells actually assume a unique “stem-like” state with activation of the CTTNB1, FOXO1, and NOTCH1 pathways and global alterations in chromatin state. Furthermore, we have found that RNAi-mediated disruption of mTORC2 signaling does not alter the bulk proliferative properties of multiple human cancer cell lines, but completely abrogates the production of “G0-like” cancer cells, which in turn profoundly alters the tumorigeneity of these cell lines as xenografts in nude mice. We have submitted these exciting new mechanistic results for publication (2nd manuscript in preparation).
Cancer cells therefore appear to continuously flux between symmetric and asymmetric division depending on the triggering of a previously unappreciated mTORC2-AKT1-TTC3-proteasome signaling pathway during cancer cell mitosis, and the G0-like cancer cells arising through this mechanism play an important but previously unappreciated role in driving tumorigenesis. This model promises significant implications for understanding how tumors grow, evade treatment, and recur.
Inhibiting Innate Resistance to Chemotherapy in Lung Cancer Stem Cells
E. Alejandro Sweet-Cordero, M.D., Stanford University School of Medicine
Lung cancer is the leading cause of cancer fatalities worldwide. The most common form is non-small cell lung cancer (NSCLC). Platinum-based chemotherapy drugs (such as cisplatin) are commonly used to treat NSCLC, but they only marginally increase survival due to the innate resistance of some tumor cells to chemotherapy. There is an urgent need to develop new ways to increase the effectiveness of chemotherapy for this disease. Past strategies for developing new drug targets have relied almost exclusively on testing cell lines grown directly on plastic culture dishes in “2D”. However, the biology of these cells is very different from that of tumor cells, which survive in a “3D” environment. To address this problem, we have developed methods for growing primary tumor cells in “3D” cultures (suspended in a gel-like material that mimics the tumor environment, rather than attached to plastic).
In our studies, we use tumor cells isolated from a well-characterized mouse model of NSCLC in which tumors carry one of the most frequent genetic mutations found in human lung cancer (a gene called K-ras). We have identified a population of tumor cells from these mice that form spheres in a 3D culture system and can re-initiate tumor growth if transplanted into the lung of an immunocompromised mouse. We call these cells “tumor propagating cells” or “TPCs” We have also determined that these TPC cells are resistant to chemotherapy treatment. In our study, we are trying to determine what genes make these TPCs chemoresistant. To do this, we have used a technique called “gene expression analysis” to find genes that change in their expression when TPCs are treated with chemotherapy either in vivo (in mice) or in vitro (in 3D culture). We have also used gene expression analysis to identify genes that have different gene exrpression in TPCs compared to non-TPC tumor cells. Importantly, we find that human patients that have a more “TPC like” gene expression signature have a worse prognosis. So we feel if we can find ways to target these “TPC and chemoresistanct genes” we may find better ways to treat lung cancer. We have come up with a list of genes that we plan to test for relevance to chemoresistance using a technique called “shRNA knock-down” where we test the effects of inhibiting expression of each of these genes on chemotherapy response. We have carried out preliminary experiments to make sure we have the right conditions to do the “shRNA screen” of chemoresistance factors in TPCs. During the first 24 month funding period, we have completed the preliminary work needed to identify genes most likely to be involved in chemoresistane and tested the conditions for the shRNA screen. Over the next 6 months we plan to carry out the screen and begin to validate the results.
Developing New Therapeutic Strategies for Soft-Tissue Sarcoma
Amy Wagers, Ph.D., Harvard Medical School and Joslin Diabetes Center
Sarcomas are highly aggressive cancers that arise in connective tissues such as bone, fat and cartilage, as well as in muscles and blood vessels embedded within these tissues. Approximately 12,000 Americans are diagnosed with sarcoma each year, and current treatment strategies, especially for advanced forms of the disease, are often ineffective, leading to high rates of mortality among sarcoma patients. To advance sarcoma treatment and develop new approaches to cure these tumors, my lab established a new mouse model of soft-tissue sarcoma in skeletal muscle that introduces disease-relevant genetic modifications into tissue stem cells found normally in the skeletal muscle. We used this model to identify a small group of 141 genes present at increased levels in both mouse and human sarcomas. Our goal in this SU2C Innovative Research Grant is to test this novel set of sarcoma-induced genes to identify new candidate drug targets for these poorly-treatable tumors.
In the past 6 months, we have made substantial progress towards this goal. First, we identified 11 chemical compounds with predicted activity against high priority targets from the custom genetic “knock down” screen we complete earlier in this project. This screen allowed us to select the most promising candidates for follow up from the 141 genes we initially identified. We analyzed the ability of these 11 compounds to inhibit sarcoma cell growth in cell culture, using a panel of cells derived from different mouse and human sarcoma tumors, and confirmed that 8 of the 11 compounds substantially reduced tumor cell growth in culture. These 8 chemicals include Asparaginase (an FDA-approved drug), Amino Sulfoximine 5, Bortezomib, Latrunculin A, UA62784, Aldehyde Erastin, MKE and AKE. We speculate that these chemicals may be useful as anti-sarcoma therapeutic agents, and have begun in vivo testing of these compounds. We are particularly excited about asparaginase, which is already in clinical use for the treatment of some leukemias. Unfortunately, our initial studies have not shown a detectable response of established sarcomas to systemic asparaginase therapy; however, additional dose and delivery optimization is needed before we draw conclusions from this study.
Finally, we have advanced in our analysis of metastatic disease in soft-tissue sarcoma, confirming that our mouse sarcoma model faithfully recapitulates differences in metastatic potential seen in myogenic vs. non-myogenic human pleomorphic sarcomas. This result, together with our ongoing biomarker analyses, further establishes the relevance of our model to discover new candidate therapeutics for human sarcoma.
In summary, our studies pursue a highly integrated strategy to identify novel targets for sarcoma therapy. Ultimately, we believe that this work will help to uncover the root causes of sarcoma formation and identify new strategies to cure these aggressive cancers.
Framing Therapeutic Opportunities in Tumor-Activated Gametogenic Programs
Angelique Whitehurst, Ph.D., University of North Carolina, School of Medicine
The overarching goal of this grant is to identify and nominate new targets for cancer therapy. The focus is on a set (119) of proteins that are only expressed in the male testes but also reactivated in a diverse set of (male and female) tumors. Our work here aims to determine which of these proteins are required for tumor to grow, thus presenting new targets for anti-cancer therapy,whose inhibition would be less detrimental to normal tissues than current therapies. Over the previous 2 years, we have constructed a platform for investigating which of the 119 testes proteins are critical for tumor cells. This study has preliminarily revealed that 75 % of these testes proteins are essential for tumor cells to survive, creating a fresh set of potential targets for therapeutic intervention. In the last phase of this project, we will be elaborating how theses testes protein contribute to tumorigenesis as follows: 1) FATE1 is essential for the survival of nearly every tumor we have tested. Over the last reporting period, we have found that this protein can prevent cell suicide programs from being activated in tumor cells. In the next phase of the project, we will assess if targeting this protein in whole animals also kills tumor cells. If these experiments succeed, FATE1 would be an idea target for therapeutic intervention in a broad range of tumors. 2) This work has revealed that tumor cells required the function of CSAG1 to promote the deflection of growth arrest signals. Continued studies on this protein, will determine if direct targeting would be an entry point to prevent the growth of tumor cells. 3) One common characteristic to all tumors is the ability to grow in low oxygen conditions and we have identified 4 testes proteins that permit survival under these low oxygen conditions. In the next reporting period, we are extending our studies to whole animal analysis to the in vivo roles of these proteins. 4) One of the most aggressive forms of breast cancer is the highly metastatic claudin-low subtype. Our work here has revealed that this type of breast tumor requires a re-expressed testes protein, ZNF165, for survival. We have found that ZNF165 supports the Transforming Growth Factor β signaling pathway, which allows tumors cells to take on the characteristics of migratory cells and move to distant sites. We are currently working to identify which genes ZNF165 regulates in this process. 5) We have also discovered that tumor cells require a testes protein called MAGEA4 to divide. In particular, lung tumor cells express this testes protein and use it to regulate processes involved in both DNA replication and the segregation of genomic material. We have initiated animal studies to determine whether MAGEA4 enhances the growth of tumor cells in vivo. In addition, we have identified a potential binding pocket on MAGEA4 for a small molecule inhibitor, and we are developing assays amenable for screening to identify therapeutics that could inhibit the function of MAGEA4.
Our work here has demonstrated that tumor cells frequently employ testes proteins to support a range of different tumorigenic features. Given that these proteins are not expressed in other adult tissues, they are ideal targets for therapies that may have limited impacts on normal tissues. In fact, mice lacking a number of these proteins are perfectly viable and healthy. These studies have have isolated those testes proteins that are most potently required for tumor cell survival and may present ideal entry points for therapeutic intervention.
Coupled Genetic and Functional Dissection of Chronic Lymphocytic Leukemia
Catherine J. Wu, M.D., Dana-Farber Cancer Institute
The treatment of chronic lymphocytic leukemia (CLL) poses two main challenges: 1) predicting the clinical course in a disease that shows many differences across patients, and 2) overcoming the insensitivity of some patient tumors to chemotherapy. At this time, genetic abnormalities are the best predictors of disease progression, based on gross chromosomal changes. However, an urgent need remains for improved understanding of how disease starts and progresses, which would lead to better predictive markers and potentially more effective (and non-toxic) therapies. Recent advances in genomic technologies provide a unique opportunity to find the genes and molecular circuits that make tumors grow in CLL. We have collected tumor and normal cells from 200 CLL patients and are almost done with sequencing all their genes. We are also looking at how genes are expressed in the same patient tumors using gene microarrays. Most importantly for enabling this project, our laboratory has pioneered the use of silicon-coated nanowires as a method of delivering DNA, RNA to primary CLL and normal B cells, which allows us to genetically manipulate CLL cells for the first time in a high-throughput fashion. Analysis of the first sixty patients has already identified genes that are important for CLL (called ‘driver mutations and pathways’). We have used our nanowires to verify the importance of some of these genes in CLL tumors cells. We now propose to find all the major genes and pathways that control CLL tumor formation. We will use a combination of sequencing technologies with statistical analyses to find the key genes that are important in creating tumors in CLL patients. In addition, we will find out which genes are good predictors of disease progression. Then, we will use our nanowires to place the mutant genes from CLL tumors into normal B cells and see how they affect their behavior. By taking this unique approach of combining different kinds of data collected from patient samples and using nanowires to manipulate the tumor cells in culture, we hope to understand the basic reasons why CLL patients develop cancer. This information will help us predict the progression of disease and provide new strategies for therapy. Finally, our approach can be extended to other tumors, especially leukemias and lymphomas.