James M. Olson - US grants

DESCRIPTION (Adapted from applicant's abstract): Several lines of evidence suggest that mutant huntingtin affects gene transcription by sequestering transcription activating and repressing proteins. This does not explain the gene expression changes that occur before mutant huntingtin is detectable in the nucleus, nor does it account for transcription changes caused by abnormal signaling from damaged afferent neurons. The overall aim of this application is to test the hypothesis that the earliest gene expression changes in Huntington's disease (HD) reflect a response of the neuron to misfolded huntingtin protein and to abnormal signaling between afferent and target neurons. To accomplish this the investigator proposes two Specific Aims 1) use mice and cell culture models that express mutant huntingtin protein in eitherthe nucleus orthe cytoplasm to determine how cells transcriptionally respond to each and 2) use mice that express mutant huntingtin protein in eitherafferent neurons ortarget neurons to determine transcriptional responses in neurons. The transcriptional responses will be correlated with pathogenic changes that occur in response to the transgenes. This will generate information and tools needed to further model and test early processes in Huntington's disease. Our long-term goal is to identify early pathogenic events in Huntington's disease in order to provide rational targets for the development of prophylactic drugs.

DESCRIPTION (provided by applicant): Medulloblastoma is a malignant pediatric brain cancer that arises from developing cerebellum. Our preliminary data show that medulloblastoma cell survival critically depends on persistent activity of neurodevelopmental pathways that favor proliferation and block differentiation. Small molecule inhibition of the sonic hedgehog or notch pathway kill medulloblastoma cells more effectively than standard chemotherapy agents. Our broad hypothesis is that both hedgehog and notch signaling are necessary for medulloblastoma, but neither is sufficient. Our long term goal is develop and utilize genetically precise mouse models of medulloblastoma to provide insight into disease pathogenesis and understand which molecular pathways are appropriate for therapeutic targeting. We have developed the ND2:SmoA1 mouse model that faithfully recapitulates human medulloblastoma with a 48% tumor incidence and a mean age of diagnosis of 25.7 weeks. In this model, Nmyc and notch signaling are increased in cerebellar tumors but not in non-neoplastic areas of cerebellum. The specific aims of this project are to 1a) complete a full neuropathologic analysis of ND2:SmoA1 medulloblastomas, 1b) determine by magnetic resonance imaging the limit of detection and rate of growth of ND2:SmoA1 mouse medulloblastoma, 2a) determine whether Nmyc is necessary for the sonic hedgehog component of mouse medulloblastoma development, 2b) determine whether Nmyc is sufficient to induce medulloblastomas in the absence of other shh target induction, 3a) determine whether Hes5 activation by notch2 is necessary for ND2:SmoA1 medulloblastoma and 3b) determine whether transgenic expression of constitutively active notch in GNPs is sufficient to generate medulloblastomas or whether it increases tumor incidence in ND2:SmoA1 or ND2:Nmyc mice when bred onto these backgrounds. The significance of this proposed work is that it will identify the essential elements of sonic hedgehog and notch pathway signaling required for medulloblastoma genesis and provide genetically precise models for testing and prioritizing candidate pathway inhibitors prior to human trials.

[unreadable] DESCRIPTION (provided by applicant): Children with medulloblastoma/PNET are currently treated with surgery, radiation, and chemotherapy. Survivors often suffer severe long-term toxicity from these treatments. Our laboratory has shown in ex vivo specimens from human medulloblastoma and mouse medulloblastoma models that targeted therapies including 13-cis retinoic acid (RA), cyclopamine, notch pathway inhibitors, histone deacetylase inhibitors and combinations of these agents induce medulloblastoma cell death as well as more toxic chemotherapy agents that are currently used for these patients. Based on our findings related to 13-cis RA, the Children's Oncology Group has developed a national Phase III clinical trial to assess efficacy of this agent. The broad long term goals of the biology correlative studies to this clinical trial are to 1) identify biomarkers with prognostic and predictive value for future clinical trials and 2) prioritize candidate targeted therapies for future clinical trials. The specific aims of this proposal are to utilize ex vivo surgical specimens to 1) identify biomarkers predicting therapy failure in high-risk medulloblastomas/SPNETs and 2) prioritize targeted therapies for future clinical trials. The significance of this work is that it is a direct means toward replacing current pediatric brain tumor treatment modalities with more effective and less toxic alternatives. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Cancer therapy efficacy is often limited by toxicity in normal tissues. The ability to target chemotherapy to cancer cells, largely bypassing normal tissues, would fundamentally improve the efficacy of cancer treatment. We recently showed that a scorpion-derived peptide, chlorotoxin, targeted the near infrared fluorescent molecule, Cy5.5 to cancer cells. This new molecular imaging conjugate permitted sensitive detection of cancerous foci as small as 200 cells. Unlike most targeting agents, CTX appears to target many types of solid tumors rather than one particular type. Whereas our previous work showed that CTX:Cy5.5 is a sensitive diagnostic agent, the current proposal seeks to determine whether CTX conjugates can be used as selective therapeutic agents. The broad, long-term goal of this research is to learn how to safely use CTX to deliver cytotoxic drugs to cancer cells, thus lessening chemotherapy side effects. The specific aims of this proposal are to 1) identify the molecular target of CTX;2) identify normal tissues that bind CTX and define CTX binding heterogeneity in selected cancers;and 3) establish the efficacy and toxicity of CTX conjugated to a potent cytotoxic drug using mouse cancer models. The significance of this work is that CTX targeting has the potential to markedly increase the therapeutic index of conjugated chemotherapy. For cancer patients, delivering higher doses of therapeutics to cancer cells while sparing most normal tissues, could result in improved survival and quality of life. PUBLIC HEALTH RELEVANCE: The relevance of the proposed work is that chlorotoxin appears to be a molecule that has high potential to specifically deliver cancer therapies to cancer cells, largely bypassing normal tissues. For cancer patients, this means more therapy could be delivered to cancer cells while the patient experiences fewer side effects. Of particular interest is that chlorotoxin appears to bind to most types of malignant solid tumors rather than just one or two types.

DESCRIPTION (provided by applicant): Over 90% of cancer patients that enroll in Phase I or II clinical trials experience no benefit from the experimental therapies, yet are exposed to drug toxicity and other challenges related to treatment. For over 50 years, physicians have used patient-specific information about drug resistance and sensitivity to select antibiotics for patients with infections, but this personalized approach has evaded the oncology community because cancer cell behavior in vitro drug sensitivity assays does not correlate with in vivo response to therapy in most cases. We have developed technologies that enable oncology drug sensitivity/resistance testing of multiple drugs or drug combinations in vivo during the days prior to surgical resection of a tumor. This approach allows drugs to interact with cancer cells while the latter are in their native tumor microenvironment. Our broad long-term goal is to develop reliable in vivo-based oncology drug sensitivity/resistance assays for patients with many types of solid tumors. Our overall goal of the STTR Phase I and II projects are to develop and test devices that are suitable for human lymphoma patients and to initiate human clinical trials. Our Specific Aim for the Phase I portion is to develop a single use (disposable) porous needle array and demonstrate that it meets drug delivery precision specifications. Provided that quantitative milestones are met in Phase I, Phase II will proceed with the following Aims: Aim 1) To develop a prototype suitable for use in human lymphoma patients;and Aim 2) to conduct a pilot "first in humans" clinical trial. The significance of the proposed work is that it will reduce the frequency of cancer patients being exposed to drugs that cause toxicity but offer no clinical benefit. The commercialization potential is described in a comprehensive business plan. We provide letters from highly respected individuals in the biotechnology, life sciences and personalized medicine fields to attest to the commercial potential of this technology. PUBLIC HEALTH RELEVANCE: Project Narrative It is estimated that approximately 1.4 million new cases of cancer will be diagnosed in the United States in 2008. Improved methods for prioritizing cancer therapeutics based on patient-based indicators of efficacy are needed. We are proposing to develop a device which enables comparison of multiple drugs or combinations in vivo, with the tumor micro-environment intact. The long-term goal of this research is to develop reliable in vivo- based oncology drug sensitivity/resistance assays for patients with many types of solid tumors. This technology will reduce the frequency of cancer patients being exposed to drugs that cause toxicity but offer no clinical benefit. This personalized treatment approach will improve patient outcome and enhance the quality of care for millions of individuals that suffer from cancer.

DESCRIPTION (provided by applicant): Despite aggressive treatment with surgery, combination chemotherapy with stem cell rescue, and in some cases radiation, young children with malignant brain tumors have a 5-year event-free survival rate of 15-30%. A critical barrier to more effective and less toxic therapies for these children is the paucity of functional knowledge about the molecular pathways that are critical for survival of these age-specific tumors. Furthermore, the incidence of infant and toddler brain tumors is low enough that they are orphan diseases, which attract no appreciable industry interest. In this proposal, we bring together some of the world's leading experts on high throughput RNAi assays to identify candidate therapeutic targets with an innovative new approach to test and prioritize potentially synergistic combination therapies and a highly experienced brain tumor translational research team to solve the specific clinical problem that highly aggressive therapies are failing to improve outcomes in infants and toddlers with brain tumors. Our broad, long-term goal is to double the cure rate for infants and toddlers with brain cancer. Our specific aims are 1) To assess the efficacy of Cdk 4/6 inhibition in clinically relevant mouse models of ATRT and medulloblastoma;2) To identify novel therapeutic targets in infant and toddler brain tumors;3) To advance one highly effective drug combination to the point of human clinical trials for infants and toddlers with brain tumors. The expected outcome is a combination therapy regimen that produces durable remission in established, bulky, clinically relevant mouse models of infant and toddler brain cancer. The significance of this work is that pediatric neuro-oncologists will abandon the highly toxic and ineffective therapeutic regimens that we are currently using in favor of a targeted approach that has higher efficacy and less toxicity. PUBLIC HEALTH RELEVANCE: This proposal integrates drug target identification and an innovative approach to prioritizing highly effective combinations of cancer drugs to advance more effective and less toxic therapy regimens for infants and toddlers with brain tumors.

? DESCRIPTION (provided by applicant): Glioblastoma Mutiforme (GBM) causes death in 95% of patients within two years even with the current standard of care. The heterogeneity within tumors and across molecular subclasses is a major challenge and there is a significant clinical need for new therapeutic agents that target GBM cells regardless of their oncogenic drivers. We recently identified PHF5A as a target that, when perturbed, causes GBM cells from multiple molecular subclasses to die, but does not harm normal neural stem cells or astrocytes. In vivo studies revealed that targeting PHF5A in mature GBM caused tumor regression, suggesting that these cancer cells rely on PHF5A for tumor maintenance. PHF5A is a member of the spliceosome machinery involved in the recognition of uncommon C-rich sequences in 3' splice acceptor sites. Our studies revealed that normal cells can compensate for loss of PHF5A activity, but GBM cells and MYC-transformed astrocytes experience mis-splicing of many essential genes resulting in cell cycle arrest. Our long-term goal is to discover and develop drugs that effectively treat GBM. Our short-term goal is to develop assays to discover candidate therapeutics that target PHF5A. This application seeks funding to develop the appropriate assays to efficiently identify candidate PHF5A inhibitors in collaboration with the NIH National Center for Advancing Translational Science (NCATS). We will then validate the screening hits in a battery of secondary and tertiary assays eliminating false positives, compounds with DMPK liabilities, and those that broadly inhibit RNA splicing. This will be accomplished through the following specific aims/phases: Phase 1- Develop assays to identify and characterize inhibitors of PHF5A; Phase 2- Conduct an HTS campaign to identify small molecule inhibitors of PHF5A; Phase 3- Validate active molecules in secondary and tertiary assays. Innovation stems from the design of minigene reporters that induce red fluorescent protein or luciferase signal only when PHF5A- mediated mis-splicing occurs. In contrast to most splicing screens which have signals of 3-5-fold, the assay we have developed produces a signal of 80-200-fold over background. Innovative designs of secondary and tertiary screens create a clear path to effectively identify selective PHF5A inhibitors and not simply general splicing inhibitors (which we believe would have on-target toxicity and hence be unsuitable as cancer therapeutics). The significance of the proposed work is that successful identification of cell penetrating PHF5A inhibitors creates a clear path to preclinical in vivo testing and potential promotion to human clinical trials. Because MYC transformation of normal cells causes exquisite sensitivity to PHF5A inhibition, it is also likely that the molecules identified will be effective for MYC-driven cancers other than GBM, which is particularly important for PHF5A inhibitors that are effective but fail to cross the blood brain barrier. The NCATs screening program is top tier in the world and Dr. Olson's experience in discovering and developing drugs that are now in human clinical trials integrate the experience needed to be successful in such an endeavor.

? DESCRIPTION (provided by applicant): Improvements in cancer diagnostics and therapeutics have positively impacted outcomes for oncology patients within select clinical niches. However, significant therapeutic uncertainty awaits the majority of today's oncology patients who present in the clinic with cancer that will relapse after exhibiting an initial response or whose cancer is already refractory to standard of care (SOC) therapy. Unfortunately, the absence of technology that accurately predicts a patient's non-response to SOC means many patients will experience toxic side effects to systemic therapy without receiving clinical benefit. Such patients will also lose valuable time and money that could be used to identify and finance more effective treatments and, in some cases under the selective pressure induced by suboptimal therapy, will develop pan resistance rendering alternate treatments ineffective. Successful translation of Presage's innovative CIVO technology into the oncology clinic will provide a personalized system that identifies patients whose cancer is not going to respond to SOC; thereby limiting unwarranted exposure to costly, ineffective therapy and its associated toxicity while increasing time available for evaluation of new therapeutic modalities at earlier stages of disease. Specifically, CIVO is a hand-held microinjection device engineered to deliver minute amounts of up to eight drugs or drug combinations directly into a patient's solid tumor. CIVO is complemented by an automated analytical package that measures tumor cell specific and microenvironment alterations at the single cell level, revealing comprehensive profiles of an individual's tumor response to each injected agent. Furthermore, CIVO is superior to alternative platforms developed to predict tumor response to chemotherapy due to CIVO's enablement of evaluating drug response in vivo, in the patients living tumor, in the natural context of native tumor stroma and the host immune system. The CIVO clinical trial detailed in this application is designed to establish the negative predictive value (NPV) of CIVO in diffuse large B cell lymphoma (DLBCL) patients who are candidates for one of three second line SOC therapies. The initial training stage of the trial will be a non- blinded, retrospective study used to establih CIVO quantitative thresholds for determining if a patient's lymphoma is non-responsive to microinjected drug in relation to the patient's response to systemic therapy. The second stage, or Test phase, of the trial will employ the quantitative threshold(s) obtained from the training stage to determine if CIVO can predict non-response to systemic second line therapy. SBIR support in establishing CIVO clinical NPV in DLBCL will build the clinical foundation required to expand translation of CIVO into additional tumor types, serving to drive more efficient and cost effective patient segregation and therapeutic selection in the oncology clinic.

Brain tumors cause more deaths in children than any other form of cancer. Most pediatric brain tumor patients receive surgery and radiation as key elements of treatment. To help surgeons maximally and safely remove brain tumors, we previously discovered and developed Tumor Paint, which delivers fluorescent signal to brain tumor cells in pediatric clinical trials. Chlorotoxin (CTX), the scorpion-derived tumor targeting peptide, crosses the blood brain barrier (BBB) and specifically binds to cancer cells. Because chlorotoxin can deliver fluorescent molecules to the cytoplasm of brain tumor cells, we hypothesized that it could carry therapeutic molecules as well. As we focus on developing therapeutic candidates that use CTX or CTX pharmacophores, it becomes essential to understand the mechanism of BBB penetration. In addition to work on CTX-based brain tumor therapies (e.g., delivery of chemotherapy or immunotherapy to brain tumors), we have made significant progress on a candidate drug that could potentially help every child who undergoes radiation therapy for brain tumors. Because brain irradiation causes severe and irreversible neurocognitive damage in children, we aspire to engineer a therapeutic agent that blocks the toxic respiratory burst of microglia in normal brain following radiation. Blockade of the Kv1.3 potassium ion channel on microglia has been shown to block radiation damage to normal brain in mice. We have engineered an optide (optimized peptide) that specifically blocks Kv1.3 but unfortunately does not, in its current form, cross the BBB. The gap in knowledge that we intend to address is that the mechanism by which CTX and some other optides penetrate the BBB is unknown. Because the Lys27 face of CTX is sterically hindered by a fluorophore in the Tumor Paint clinical candidate that crosses the BBB in children, we hypothesize that the pharmacophore responsible for BBB penetration lies on a different face than the face that contains Lys27. The key hurdle that prevents clinical development of an optide that blocks Kv1.3 to alleviate radiation-induced brain damage is that it does not cross the BBB and therefore fails to reach its target. We hypothesize that we can engineer the candidate Kv1.3 blocker in a manner that fosters BBB penetration. Our Specific Aims are: Aim 1: To identify the pharmacophore of chlorotoxin responsible for BBB penetration Aim 2: To identify the transporter responsible for optide penetration of the BBB Aim 3: To create an optide that has a therapeutic pharmacophore and a BBB-penetrating pharmacophore The significance of this work is that we will produce a clinical development candidate that could alleviate severe brain damage caused by irradiation in children. The foundational knowledge could be applied to a new generation of drugs for many brain disorders.

PROJECT SUMMARY A solid Consortium project should contribute a unique and compelling approach to neuroblastoma and supply tools or components that enable other members of the Consortium to make their best work even better. The project detailed here commences with the selection and careful characterization of a dozen fully human antibodies to Glypican-2, the most promising antigen in neuroblastoma, from a pre-existing hybridoma stock. These antibodies, along with their full sequences, binding affinities, cellular internalization, epitope, and (where possible) crystal structures, will be made available to Consortium members within the first year for use in their modality of choice. This project will explore their utility as components in double bispecific antibody therapy: this double bispecific approach targets two cancer antigens and two T cell receptors simultaneously, inducing T cell activation and co-activation only in the presence of cancer cells expressing both antigens. The double targeting strategy should allow for a much higher level of selective engagement and killing than has heretofore been possible with therapies that target only a single cancer antigen, and should it prove effective in neuroblastoma the approach is likely to have utility in a wide variety of cancers. One historical weakness for the development of immuno-oncology approaches such as this one is that the animal models have had very little predictive value. For this reason, the best of the humanized mouse models, ?MISTRG? mice, will be used both for the development of these molecules and for testing the other consortium members? approaches, where they could be useful. CIVO multi- needle array technology will be employed to simplify the combinatorial challenge associated with testing pairs of bispecific molecules; this technology will also be open for use collaboratively with the other Consortium members. In short, the proposal provides the Consortium with human antibodies against a validated neuroblastoma target, access to the current state-of-the-art in humanized mice, a CIVO multi-needle array device to facilitate testing therapeutics, and a double bispecific antibody therapeutic approach that promises a high level of T cell killing and selectivity and a likely broad applicability beyond neuroblastoma.