Developmental Research Program
The goal of the SPORE Developmental Research Program is to promote new ideas directed toward the SPORE translational melanoma research. SPORE guidelines mandate diligent efforts to identify and fund pilot research projects by providing a formal Developmental Research Program (DRP) that will highlight and support new research endeavors whether collaborative among scientists with one or more SPOREs, or with new scientists outside the SPORE environment, that may eventually reduce the incidence and morbidity or increase the survival of melanoma patients. Year 07 DRP awardees are as follows:
Activators and Activation of the PI3K-AKT Pathway in Advanced Melanoma
Principal Investigator: Michael Davies, M.D., Ph.D.
There is a growing evidence that kinase signaling pathways play a critical role in melanoma. The importance of the RAS-RAF-MEK-MAPK pathway is supported by the high prevalence of mutations in the pathway, and the efficacy of BRAF and MEK inhibitors in BRAF-mutant melanomas. There is also evidence that the PI3K-AKT pathway is frequently activated in this disease, but its clinical associations and significance are less clear. An improved clinical understanding of the PI3K-AKT pathway in melanoma may help further stratify patients prognostically, and will be essential for the rational design of clinical trials with agents targeting it. Previously, we used reverse phase protein arrays (RPPA) to analyze the PI3K-AKT pathway in human melanoma cell lines and frozen melanoma clinical specimens. These studies demonstrated that loss of PTEN expression results inhigh basal activity of the PI3K-AKT pathway and resistance to BRAF and MEK inhibitors. We also found that melanoma cell lines with elevated expression of the insulin-like growth factor-1 receptor (IGF1R) develop compensatory activation of the PI3K-AKT pathway after BRAF or MEK inhibition, resulting in resistance to these agents. Functional studies demonstrated marked synergy for combined inhibition of MEK and IGF1R specifically in the cell lines with high expression of the IGF1R. Recently, we have validated an immunohistochemistry (IHC) test for PTEN expression in FFPE melanomas, including identification of expression levels that correlate with activation of AKT. Further, our collaborators have demonstrated the
feasibility of using RPPA to directly analyze the expression and activation of several signaling pathways in FFPE clinical specimens in breast and renal cancers. These advances, in concert with the repository of annotated clinical specimens available in the MDACC Melcore, provide the unique opportunity to expand our understanding of the clinical significance of the PI3K-AKT and other pathways in this disease. In this study, we will assess the expression of PTEN and the IGF1R in sets of annotated stage III and stage IV melanomas from patients treated at the MD Anderson Cancer Center. The results will be used to determine the rate of PTEN loss and IGF1R overexpression in regional and distant melanoma metastases, and the clinical associations with each of these molecular phenotypes. We will also analyze a panel of tumors with matching frozen and FFPE tissue samples to test the validity of RPPA assays for the assessment of protein signaling networks in archived melanoma samples. These studies will provide critical data for the design of future clinical trials and funding applications on the role and therapeutic targeting of the PI3K-AKT pathway in melanoma. Generation of GD3-specific Chimeric Antigens Receptor Transduced T Cells for the Immunotherapy of Metastatic Melanoma
Principal Investigator: Laurence Cooper, M.D., Ph.D. and Laszlo Radvanyi, Ph.D.
Adoptive -cell therapy (ACT) is emerging to be a powerful method of therapy for late-stage metastatic melanoma with response rates of 50% or more and increased survival of patients receiving expanded tumor-infiltrating lymphocytes (TIL). However, TIL can only be successfully expanded from about 55-60% of patients leaving many without an option for TIL therapy. An alternative option for these patients rapidly gaining momentum in the clinic is to perform ACT using expanded antigen-specific T-cells from autologous peripheral blood mononuclear cells (PBMC) recognizing widely-expressed antigens on melanoma cells. This approach has involved mostly the activation of melanoma-peptide-specific T cells from PBMC followed by their large-scale expansion as well as more recently the transduction of high-affinity TCR chains for melanoma antigens into activated, proliferating T cells. One relatively new approach for ACT is to infuse activated T cells from peripheral blood mononuclear cells (PBMC) of patients transduced with chimeric antigen receptors (CAR) recognizing melanoma-specific antigens. These CARs are advantageous because they are not HLA-restricted and are not susceptible to HLA loss that can occur during tumor progression. The ganglioside GD3 is expressed at high levels in human metastatic melanoma and is not expressed on normal tissues. GD2 has been the ganglioside targeted in melanoma in the past, but GD3 is in fact not only more specific for melanoma tissues, but is also more highly expressed than GD2. The high expression of GD3 in melanoma suggests that it may be an alternative and potent target for immunotherapy using antibodies or with the CAR ACT technology. In this project, we aim to merge our scFv phage cloning technology that has identified a number of high-affinity GD3-specific clones with our established CAR T-cell technology to generate and test human T-cells expressing a GD3-directed CAR against human metastatic melanoma that can then be developed clinically as a method for ACT. We will screen a scFV phage library from isolated melanoma tumor-infiltrating B cells (previously found to have GD3-specificity) to identify scFv clones with a range of avidities for melanoma GD3. We will then clone low, intermediate and high avidity GD3 scFVs into our CAR expression vector system and express in human T cells from melanoma patients. Our CD19-specific CAR structure, which is currently in clinical trials, will be modified by replacing the scFv extracellular motif with scFv regions from the GD3-specific mAbs. The Sleeping Beauty (SB) transposon/transposase system will be used to electro-transfer DNA plasmids coding for the CARs into T cells and CAR+ T cells selectively numerically expanded on our designer artificial antigen presenting cells (aAPC). The electroporated and propagated T cells will be characterized for re-directed effector functions against autologous and allogeneic melanoma cell lines. This approach is clinically-appealing as the SB system and aAPC platforms are already being tested at MDACC in investigator-initiated trials. Our overall hypothesis is that scFv phage clones with a variety of avidities for melanoma GD3 can be isolated from our existing phage library and used to generate a new generation of CAR constructs that can be stably introduced into T cells. This GD3-specific CAR technology represent a novel and powerful approach for ACT that can be used to treat almost all melanoma patients due to the ubiquitous nature of GD3 expression in melanoma. The results of this project will set a foundation for clinical translation to treat melanoma as part of our established CAR T-cell therapy program.
Differentially Expressed miRNAs as surrogates of outcome in stage III melanoma and associated with oncogene mutational status in cutaneous melanoma
Principal Investigator: Michael Tetzlaff, M.D., Ph.D.
Clinical outcomes in melanoma patients remain difficult to predict, despite numerous clinical and pathological parameters derived to guide patient management. In particular, about half of stage III melanoma patients will eventually succumb to their disease, but there is no universally agreed upon prognostic scheme that reliably distinguishes these patients from those who will be cured by surgery alone. There is therefore a critical need to develop new predictive biomarkers that distinguish among these heterogeneous patient groups. In addition, identifying molecular markers that correlate with adverse disease outcomes will likely unveil key biochemical pathways driving these aggressive behaviors and in the process, illuminate novel opportunities for the design of rational therapeutic interventions. An emerging paradigm in the treatment of melanoma is to devise chemotherapeutic agents discretely targeting pathways affected by tumor-specific mutations. This includes inhibitors targeting activating mutations in known oncogenes BRAF, NRAS and CKIT. However, in preliminary studies, clinical response to these inhibitors has been erratic and unsustained. There is therefore an additional critical need to define more completely the molecular-genetic context(s) in which these mutations function. This will establish a foundation for future efforts in which chemotherapeutic strategies might be tailored according to a patientâ€™s genetic signature. An essential ingredient to address each of these critical needs is the identification of a suitable biomarker. microRNAs (miRNA) are short oligonucleotides of non-coding RNA that function in post-transcriptional gene regulatory pathways. miRNAs have emerged as remarkably robust biomarkers for several important reasons. First, each human tissue typeâ€”including cancer tissuesâ€”appears to exhibit a unique miRNA expression profile. Second, the miRNome is comprised of many fewer elements than the mRNome or the proteome, greatly simplifying global assessments of expression. Finally, the small size of miRNAs confers relative stability (compared to mRNA and protein) in a variety of different media. Specifically, it is feasible to obtain high-quality miRNA expression data from formalin-fixed paraffin embedded (FFPE) tissue, including melanomas, and this is often the only tissue available for retrospective analyses of this disease. Nevertheless, clinical specimens remain relatively underutilized in studies examining miRNA expression in melanoma, and no study has systematically incorporated appropriate clinical or pathological predictive factors into their data analysesâ€”an essential consideration to determine the additive value of miRNA profiling in this disease. As one of the largest melanoma treatment centers in the world, The University of Texas MD Anderson Cancer Center (UTMDACC) has FFPE melanoma tissue with comprehensive clinical and pathological annotation available for over 2,000 patients through the MDACC Melanoma Informatics, Tissue Resource, and Pathology Core (MelCore), including a cohort of over 900 patients with known BRAF, NRAS and CKIT mutational status. This unique resource allows for not only definitive molecular profiling of this disease, but also the incorporation of contemporary predictive factors and mutational status to determine the clinical significance of new findings. My career goal is to characterize the functional and predictive significance of miRNAs in melanoma. The specific hypotheses of the current study are that miRNA expression profiles (1) are independent predictive factors among patients with Stage III melanoma and (2) are differentially associated with oncogenic mutations prevalent in melanoma.
Heparanase Mechanisms in Brain Metastatic Melanoma
Principal Investigator: Dario Marchetti, Ph.D.
Patients with brain metastatic melanoma (BMM) have, even with the best available treatments, a median survival of less than six months. Our work has implicated heparanase (HPSE) as a promoter of brain metastasis. In particular, we have demonstrated that heparanase is expressed in human BMM cells and tissues, and functions as a downstream target of neurotrophic receptor pathways affecting cell invasion. Our objective is now to determine how heparanase regulates brain metastasis and use this knowledge to develop new HPSE - based therapies. Underscoring the importance of targeting heparanase, we have made three key discoveries that shed new light on the relevance of this molecule towards the aggressive brain metastatic phenotype. First, we identified microRNA-1258 as a microRNA that inhibits heparanase and suppresses brain metastasis. Second, we discovered that HPSE has functions independent of its enzymatic activity, e.g., regulating Rac and Rho, critical mediators of cytoskeletal dynamics which is a fundamental process in interplays between tumor and microenvironment cells. Third, by investigating circulating tumor cells (CTCs) from blood of brain metastatic cancer patients, we discovered the expression of heparanase in CTCs, and a significant correlation between its presence, epidermal growth factor receptor gene amplification, and the known cancer stem cell marker ALDH1. Based on these discoveries, a logical next step is the formulation of
strategies to inhibit HPSE and suppress BMM. This can be achieved using miR-1258 as well as new HPSE inhibitors, e.g., non-anticoagulant, glycol-split heparins. One of them, SST0001, has emerged as a membrane permeable, potent small-molecule inhibitor of heparanase enzymatic activity and is available to our laboratory. Hypothesis is that heparanase regulates the cross-talk between BMM and brain microenvironment cells, and initiates multiple effects that are critical for the development and progression of BMM. By proposing much broader roles for heparanase, which involve enzymatic and non-enzymatic functions, we aim to identify new roles for this molecule in BMM progression. We consider that inhibiting these functions by injecting BMM cells transduced with miR-1258 in experimental animals, and/or treat these animals with SST0001, will provide one-punch effect as powerful drug combination to inhibit HPSE present in both brain microenvironment and BMM cells, thus affecting the overall BMM onset. Our approaches include human in vitro and in vivo BMM models coupled with lentiviral delivery targeting HPSE with miR-1258, potent HPSE inhibitors, and cutting edge technologies. These studies emphasize the strong translational component of our proposed work. They will provide pre-clinical data fundamental to prove the relevance of HPSE in BMM, and introduce heparanase inhibitors for therapies to treat brain metastasis in general, brain metastatic melanoma in particular.