Anat Erdreich-Epstein - US grants

DESCRIPTION (provided by applicant): Brain tumors are an important cause of cancer-related death. Malignant brain tumors are highly angiogenic and their microvessels express the integrin cell surface receptors alphavbeta3 and alphavbeta5. Clinical trials using RGDfV, an integrin-alphavbeta3/beta5 inhibitor, are promising. It is therefore critical to better understand the molecular mechanism of alphav-integrin inhibition in order to improve its therapeutic efficacy in future clinical trials. Our goal in this proposal is to examine ceramide as a molecular target in inhibition of endothelial integrins alphavbeta3/beta5. Inhibition of integrins alphavbeta3/beta5 induces apoptosis of endothelial cells (EC) and prevents brain tumor growth in an intracranial mouse xenograft model. Our previous work shows that inhibition of EC alphavbeta3/beta5 increases levels of the intracellular pro-apoptotic lipid second messenger ceramide. Our more recent results suggest this is by hydrolysis of sphingomyelin due to activation of acid sphingomyelinase. Importantly, integrins alphavbeta3/beta5 and p53 act in concert in neovascularization, but the mechanism by which integrins regulate p53 is not known. Our preliminary data, that inhibition of EC integrins alphavbeta3/beta5 suppressed phosphorylation of the pro-survival kinase PKB/Akt, and that Akt regulated p53 activity, are very interesting in this regard, especially since Akt is a potential target for inhibition by ceramide. And lastly, we have shown that the synthetic retinoid, fenretinide, also induces ceramide-mediated EC apoptosis. However, this ceramide generation is by de novo synthesis, a mechanism that is different from RGDfV. Based on these data we hypothesize that ceramide is critical in the mechanism of endothelial cell apoptosis that is induced by inhibition of integrins alphavbeta3/beta5. Important questions stemming from this are: 1) Is RGDfV-induced ceramide required for integrin alphavbeta3/beta3-mediated apoptosis; 2) Do integrins alphavbeta3/beta5 control p53 activity via their regulatory effect on Akt, and does ceramide mediate this function; 3) Does combination of RGDfV + fenretinide have better therapeutic efficacy against brain tumors, compared to each alone; and last, 4) Does the in vivo effect of RGDfV target the endothelial cells, the tumor cells, or both. Our Specific Aims to investigate these questions and examine our hypothesis are: 1.To determine molecular interactions between ceramide, integrins alphavbeta3/alphavbeta5 and endothelial cell apoptosis. 2. To elucidate the signaling mechanism of alphavbeta3/alphavbeta5 integrin blockade downstream of ceramide generation. 3.To examine the combination of fenretinide + RGDfV in an intracranial mouse brain tumor model. These studies will provide critical mechanistic data on the function of alphavbeta3/beta5 integrin inhibition in antiangiogenic cancer therapy, and will be essential for design of improved treatments for patients with malignant brain tumors.

DESCRIPTION (provided by applicant): Angiogenesis, the sprouting of new capillaries, is essential for cancer growth. Inhibition of integrins av-beta3/av-beta5, cell surface receptors critical in angiogenesis, induces endothelial apoptosis, disrupts angiogenesis and inhibits brain tumor growth in a mouse model. The potential utility of integrin av-beta3/av-beta5 inhibition in cancer therapy is becoming apparent with the encouraging responses to RGDfV, an integrin-function-blocking peptide that completed a phase-I clinical trial in adults with recurrent gliomas. Despite the emerging use of av-beta3/av-beta5 integrin inhibition in cancer therapy, especially against brain tumors, the molecular mechanism of this inhibition is poorly understood, precluding rational design of combination therapies for future trials. We have shown that integrin inhibition increases intracellular ceramide, a pro-apoptotic lipid second messenger. It is not known whether this ceramide increase is required for the anti-angiogenic action of integrin av-beta3/av-beta5 inhibition. Our recent data suggest that acid sphingomyelinase is critical in this mechanism. We hypothesize that the ceramide increase, generated by acid sphingomyelinase and induced by integrin- av-beta3/av-beta5 inhibition is required for endothelial apoptosis and for the effect of RGDfV against brain tumors in vivo. Our Specific Aims are: 1) To determine molecular interactions between ceramide, integrins av-beta3/av-beta5 and apoptosis, and 2) To compare the inhibitory effect of RGDfV in WT and ASMase deficient mice using an intracranial mouse brain tumor model. In vitro we will use ASMase-deficient or -competent endothelial cells plated on vitronectin and compare their ceramide and apoptotic responses to RGDfV. In-vivo we will implant intracranial brain tumors in ASMase knockout mice and wild type littermates and determine differences in survival, tumor growth and angiogenesis in RGDfV- or control-treated mice between the groups. Demonstration of a requirement for acid sphingomyelinase in av-beta3/av-beta5 integrin-mediated endothelial apoptosis will be the first evidence for a causal role of ceramide in integrin signaling. Requirement for ceramide metabolism in integrin signaling has potentially far-reaching implications by providing a molecular rationale for combining integrin-inhibiting drugs with agents geared to optimize the ceramide response, resulting in improved anti-cancer efficacy.

DESCRIPTION (provided by applicant): This project focuses on the role of PID1 (Phosphotyrosine Interaction Domain containing 1) in gliomas. PID1 is a phosphotyrosine binding (PTB) domain-containing protein that was discovered in 2006 and which has never been studied in the context of cancer. Ectopic expression of PID1 has opposite effects on proliferation in different cell types. Using microarrays and quantitative RT-PCR, we found that high expression of PID1 correlates with better outcome in two types of brain tumors, gliomas and medulloblastomas. Our in vitro preliminary data further show that ectopic expression of PID1 has growth-inhibitory effects in brain tumor cell lines from three different types of brain tumors: glioblastomas, medulloblastomas, and atypical teratoid rhabdoid tumors (ATRT). We chose to focus this PID1-centered R21 proposal on gliomas since they constitute the most common primary brain tumor in humans and have grim prognosis when malignant. The goal of this proposal is to uncover the molecular mechanisms by which PID1 mediates its inhibitory effect in gliomas. Specifically, we will test the hypothesis that PID1 exerts its inhibitory effect in malignant gliomas in part through its PTB domain. We propose the following two Specific Aims: 1) To determine the critical domain(s) of PID1 that mediate(s) its inhibitory function in GBM cell lines. In this Aim, we will make mutants of known and predicted domains of PID1 and test their ability to decrease proliferation and induce apoptosis. 2) To examine the effect of PID1 expression on growth of intracranial human cell line-derived GBM tumors in NOD-SCID mice. In this Aim, we will use regulated expression of PID1 and examine the effects of PID1 expression on tumor growth, survival, proliferation, and apoptosis. This work is novel and of high impact, as it will further our knowledge on the molecular mechanisms and functions of PID1 in gliomas, forming the first basis of knowledge about PID1 in any brain tumor and any cancer. The clinical correlation we uncovered in our Preliminary Data highlights the relevance and significance of this topic. In addition, our finding of an inhibitory function for PID1 in three different types of brain tumors suggest that PID1 may also be important in other cancers, and possibly in other diseases, potentially expanding the impact of our work.

DESCRIPTION (provided by applicant): Medulloblastomas (MBs) are the most common malignant brain cancers in children and frequently disseminate to the leptomeninges. Leptomeningeal MB carries poor prognosis and thus constitutes an unmet clinical need. Its therapy is suboptimal since drugs frequently do not effectively enter the cerebrospinal fluid (CSF) or are washed out rapidly. Thus, the leptomeninges are a sanctuary from systemic therapy. It is therefore imperative to develop novel approaches for drug delivery into the CSF to treat leptomeningeal MB. Our goal is to develop a new clinical approach to effectively treat leptomeningeal MB. We propose metronomic (multiple small frequent infusions) drug administration directly into the brain ventricles to achieve maxi- mal therapeutic efficacy. Therapy will be via a novel, wirelessly-operated, programmable, implantable, re- fillable micropump developed by Dr. Meng, which is now ready for preclinical testing in this R21. The main innovations of this pump are the combination of wireless control, programmability, and miniaturized packaging. Our pilot data show that this refillable pump can reliably deliver hourly infusions into the CSF over prolonged periods. In a small pilot experiment, topotecan given metronomically into the CSF via this pump induced complete remission in a mouse with leptomeningeal MB, whereas similar dose delivered intraperitoneally had no benefit. Thus, metronomic delivery of chemotherapy into the CSF may prove more effective and less toxic than systemic delivery. This R21 will test our pump in treatment of leptomeningeal MB in mice. Treatments such as chemotherapy, anti-GD2-mediated immunotherapy and others may benefit from metronomic intra-ventricular delivery via our novel wireless programmable implantable refillable pump. We hypothesize that metronomic therapy delivered directly into the CSF by our pump will be more effective against leptomeningeal MB in mice compared to the same agents delivered via other schedules/routes. We will test this using our pump in two Specific Aims: 1) To test pre-clinical safety and efficacy of metronomic intra-ventricular chemotherapy via our micro-pump in mice carrying leptomeningeal MB compared to drug delivery via other routes and schedules; 2) To test safety and efficacy of intra-ventricular anti-GD2-based immunotherapy via our micro-pump in mice carrying leptomeningeal MB compared to anti-GD2 delivery by other routes and schedules. Results from this R21 will establish principles for dosing regimen(s) for phase I clinical trials using a scaled-up version of this pump and provide the pre-clinical data fo this novel clinical approach. This work is of high significance as our novel pump will allow effective drug delivery into the leptomeninges. The clinical impact of this approach is wide, since it may also apply to other diseases or other cancers affecting the brain and spinal cord for which sustained delivery of drugs is a challenge. Thus, our application is highly novel and has high significance and potential for strong impact in an area where there is a great unmet need.

? DESCRIPTION (provided by applicant): Our research centers on the biology of brain tumors and on the role of PID1, which we recently reported to be a candidate tumor suppressor in gliomas and medulloblastomas. In this exploratory R21 proposal we will focus on glioblastoma (GBM), the most common primary malignant brain tumor in humans. We showed that PID1 overexpression was growth-inhibitory in cell lines from GBM and that higher PID1 mRNA levels were correlated with longer patient overall survival. Here we will examine the role of PID1 as a sensitizer of GBM to therapy and examine its molecular mechanism. PID1 is a phosphotyrosine binding (PTB) domain-containing protein discovered in 2006. PID1 functions in obesity-mediated insulin resistance via inhibition of insulin signaling, is linked to Alzheimer's disease, and as we showed, has tumor-suppressive effects in glioma and medulloblastoma brain tumors by an unknown mechanism. Ongoing experiments find that PID1 sensitizes glioma cells to chemotherapy and that it is a novel binding partner to a receptor with critical roles in GBM. In this Exploratory R21 proposal we aim to understand the mechanism by which PID1 sensitizes GBM to chemotherapy and inhibits GBM growth. Our Aims will 1) determine the mechanism and characteristics of PID1 interactions with the receptor and effect on signaling and proliferation in GBM, and 2) investigate the sensitization of GBM to chemotherapy in vitro and in vivo. Our recent work highlights the significance of PID1 in brain tumors, pointing to a potential inhibitory and/or therapy-sensitizing role of PID1. The significance of our mechanistic findings will likely extend to other diseases, as PID1 is relevant in obesity, insulin resistance, and Alzheimer's disease. The main innovative aspects of this work include the finding of PID1 as a sensitizer of GBM to chemotherapy and the novel mechanism we will uncover. These results will provide the basis to design modalities that will sensitize GBM to therapy.

Medulloblastoma (MB) is the most common malignant brain tumor in children. We were the first to report on a growth-suppressive effect of PID1 (Phosphotyrosine Interaction Domain containing 1) in MB, constituting the first reported link of PID1 to MB and to cancer overall. We showed that PID1 slowed growth and induced apop- tosis of cell lines from three types of brain tumors, MB, glioma, ATRT, and that lower tumor PID1 mRNA level in MB and glioma tumors correlated with shorter patient survival. Recently we showed that PID1 sensitizes cell lines of MBs and gliomas to chemotherapy, suggesting a potential therapeutic value. Interestingly, PID1 was required for the anti-MB effect of cisplatin, a mainstay of MB chemotherapy. Although cisplatin increased PID1 mRNA, it decreased PID1 protein level in a manner restored by proteasome inhibitors, suggesting that cisplatin promoted proteasomal degradation of PID1. We therefore hypothesize that stabilization of PID1 protein will enhance the anti-MB effect of cisplatin. This hypothesis will be examined in two Specific Aims: 1) To determine PID1 amino acids that are post-translationally modified to mediate cisplatin-dependent deg- radation of PID1 and test if mutating these amino acids to stabilize PID1 will augment the anti-MB effect of cisplatin in culture and in vivo. 2) To determine if proteasome inhibitors augment MB response to cisplatin, if this is via PID1 stabilization, and if the combination is safe and effective against MB in vivo in brains of BarTeL mice. This project will be in cultured MB cells and in our innovative transgenic mouse model for MB, BarTeL and will also test a new brain permeable proteasome inhibitor, marizomib, against intracranial MB in mice. At its con- clusion we will have defined the cisplatin-induced post-translational modifications of PID1, focusing on those that affect protein stability of PID1 and ways to stabilize PID1 in order to enhance efficacy of cisplatin. The potential impact of the knowledge gained is that it will support future translation for patient benefit. In the short term our findings may lead to pre-clinical testing of combination of the new brain-permeable proteasome inhibitor in combination with cisplatin in patients with recurrent MB. In the long term, the knowledge gained will inform our planned design of PID1-based therapeutics to exploit its tumor growth-suppressive effect in MB and other brain tumors.

PROJECT SUMMARY TITLE: USC/CHLA Summer Oncology Research Fellowship (SORF) Program for medical students Physician-scientists working in cancer research provide unique insights that are critical for continued advances in cancer therapy. There is a dire shortage of physician-scientists and their numbers continue to decline. The Summer Oncology Research Fellowship (SORF) Program at Norris Comprehensive Cancer Center, Keck School of Medicine, University of Southern California (USC) and Children?s Hospital Los Angeles (CHLA) pro- vides a unique oncology research experience by immersing outstanding medical students of diverse back- grounds from across the United States in cancer research through mentored hands-on research projects and a cutting edge interactive educational curriculum. The goal is to encourage these motivated students to pursue a medical career that actively involves cancer research. This program, by increasing the number and diversity of students entering the physician-scientist education pipeline, is designed to address the severe shortage of such professionals, a shortage that stifles advances in pediatric and adult oncology. This highly competitive SORF Program has been continuously active for over 40 years, with demand steadily increasing: 303 students (top 5-20% of applicants) completed the Program since 2001. Of SORF graduates in the 10 years between 2001-2010, 18% are in oncological specialties and 11% are involved in research, compar- ing favorably with overall USA medical school graduates (1.98% in oncology, 0.13% in research; AAMC data). Thus, our Summer Oncology Research Fellowship Program has great potential to enhance the ?supply? of stu- dents who enter the pipeline leading to biomedical careers that will involve active cancer research. This Multiple-PI R25 proposal seeks support for the SORF Program to bring it to its next level by providing students with a better understanding of how science is conducted in the 21st century and interesting them in pursuing careers that involve cutting edge research related to oncology. Program activities include hands-on individual mentored research projects in adult and pediatric oncology, interactive cancer research-focused edu- cational activities, professional development in scientific communication, career development activities towards oncology research, and exposure to other trainees, scientists, physicians, patients and cancer survivors. Other goals of the program are to increase diversity of participants, provide instruction on responsible conduct of re- search, and institute short and long term follow-up to obtain feedback and outcomes data on the Program. Thus, by increasing the ?inflow? at the beginning of the pipeline and maintaining long-term contact with participants we anticipate to strengthen the oncology research workforce.