Robert Darnell - US grants

This proposal integrates interests and training in clinical neurobiology and basic molecular biology; the long term objective is an understanding of neural trophism and plasticity, and their relevance to neuro-oncology. Previous work demonstrated that autoantibodies present in patients with neurologic disease and occult malignancy bind antigens (termed 'onconeural') that are coordinately expressed in neurons and tumor cells; these autoantibodies can be used to clone onconeural genes. It is hypothesized that there are specific signal transduction pathways shared by tumor cells and neurons. Two candidate pathways, not necessarily exclusive, are suggested: proto-oncogene and onconeural pathways. An experimental protocol is suggested to identify onconeural autoantibodies from patients with neurologic disease, and use them to clone genes from lambda gt11 expression vectors. DNA sequence data will give preliminary information about the function of these genes. To test the relationship between neural and growth control pathways, full length cDNA's will be transfected into various cell lines and assayed for transforming activity. Conversely, proto-oncogenes and onconeural genes will be tranfected into neural cells and assayed for a role in neural trophism and plasticity. Clinical information regarding the patients neurologic disease and malignancy will be correlated with the study of their autoantibodies. These experiments will be related both to basic mechanisms of neurophysiology and growth control, and a number of clinical problems. These include the regulation of neuron-specific genes, the regulation of growth control pathways in tumors, the identification of tumor specific antigens of potential diagnostic and therapeutic importance, the identification of factors and pathways that may be involved in neural trophism and neurodegenerative disease, and mechanisms of autoimmune disease.

The goal of this proposal is to determine the extent to which neuronal RNA binding proteins (n-RBPs) are present in the nervous system, and to explore the role they play in both the development and function of the brain and in paraneoplastic neurologic disease (PND). PNDs are autoimmune disorders that are believed to be triggered when tumor cells express proteins that are normally made only in neurons. This is associated both with effective anti-tumor immunity and neurologic degeneration. A number of groups have identified n-RBPs by expression vector cloning of brain libraries utilizing antiserum from PND patients. Two distinct groups of n- RBPs have emerged from this work. The Hu family of genes are human homologues of the Drosophila Elav gene, and are also related to the Drosophila sex-lethal gene. Elav is essential for neurogenesis in Drosophila, and sex-lethal mediates alternative splicing in a sequence specific manner. The Nova family of genes are related to the hnRNP K, FMR- l (fragile-X) and yeast MER- l RBPs. FMR- l loss of function mutations result in mental retardation, and the MER-l gene regulates alternative splicing. However, aside from sequence homologies vaguely suggesting a possible role in alternative splicing, very little is known regarding the role of the n-RBP proteins in brain or tumors. We propose a three-fold approach to the study of n-RBPs associated with paraneoplastic neurologic disease. Descriptive studies will include a careful evaluation of the extent and expression of the n-RBP gene families. We will specifically correlate these findings with the neurologic symptoms of PND patients. Functional studies will be performed with the aim of identifying specific target RNAs bound by n-RBPs. Biologic studies will evaluate the role of these genes in mice. These studies will contribute to our understanding of the role of n-RBPs in the development and function of the nervous system, and will clarify their role in the pathophysiology of the PNDs, including their association with autoimmune neurologic disease.

Patients with the paraneoplastic neurologic syndromes develop immunity to neural antigens expressed in their tumors, leading to effective anti-tumor immunity and autoimmune neurologic disease. Typically patients die from severe neurologic degeneration, for which there is no available treatment. We believe that treatment aimed at suppressing cellular immunity may have been overlooked as a treatment option in the paraneoplastic neurologic disorders, and are studying the effects of such immune suppression, using cyclosporin and prednisone, on clinical outcome.

The purpose of this study is to explore a role for cell-mediated immunity in the pathogenesis of patients with paraneoplastic neurologic disorders. These patients manifest anti-tumor immunity clinically that is associated with antibodies to onconeural antigens that we have cloned. We are studying whether these patients harbor T cells in their serum and spinal fluid that recognize and kill autologous cells expressing onconeural antigens.

DESCRIPTION (provided by applicant): We propose to identify functional interactions between Ago-miRNA and neuron-specific RNA binding proteins (nRNABPs) in the normal and stressed mouse brain. RNA regulation in vivo involves the superimposed and likely combinatorial action of many RNABPs, but most investigations of RNABPs have focused on the role of individual proteins in vitro. This has confined our understanding of conventional regulation by RNABPs and how they intersect with Argonaute (Ago) mediated miRNA regulation; while it is suspected that these two systems intersect, there is almost no understanding of where they do so or what the functional consequences are. The expanding list of RNABPs and miRNAs associated with neurologic disease has underscored large gaps in our understanding of their roles in disease pathogenesis. We developed HITS-CLIP during the last funding period, a revolutionary approach to the study of protein-RNA regulation, and will apply it for the first time to studying coordinate and dynamic regulation by Ago-miRNA and neuronal RNABPs. These studies will be applied to normal brain, using homologous recombination in mice to establish robust and relevant in vivo study in neurons. They will then be applied to stress brain, using three brain-disease relevant paradigms.

Study of the paraneoplastic neurologic disorders (PNDs) offers a unique opportunity to study cellular mechanisms of naturally occurring tumor immunity in humans. Patients with these neurologic disorders harbor systemic tumors that express proteins whose normal expression is entirely restricted to neurons. Perhaps in part because such proteins are normally concealed in immune privileged sites, their ectopic expression in tumors is believed to result in a potent tumor immune response that is associated with clinically evident tumor suppression. These patients come to clinical attention, however, when their immune response begins, by unknown mechanisms, to recognize and destroy not only the tumor cells, but also neurons that normally express the same antigen. Among patients with a specific PND, paraneoplastic cerebellar degeneration (PCD), which develops in women with evidence of tumor immunity to occult breast or ovarian cancers, cerebellar Purkinje neurons are destroyed. PCD patients harbor serum and CSF antibodies specific to a protein termed cdr2. Both PCD tumors and Purkinje neurons express cdr2. We have also found that the cdr2 antigen is expressed in a wider range of gynecologic tumors than the rare co-incidence of tumor immunity and cerebellar degeneration present in PCD might suggest. In preliminary studies of tumors obtained from neurologically normal tumor patients, we have found cdr2 antigen expression in 60% of ovarian tumors and 25% of breast tumors. In preliminary experiments we have found consistent evidence that PCD patients harbor cytotoxic T-cell lymphocytes (CTLs) able to kill cells presenting cdr2 peptide. We detected these CTLs using a simple recall assay and autologous dendritic cells (DCs) to activate their T cells.

The paraneoplastic neurologic diseases (PND's) are a diverse group of neurological disorders arising in the setting of cancer that are believed to have an immunological basis. It is believed that the disorders are triggered when systemic tumors express proteins normally made only in neurons, leading to effective anti-tumor immunity and autoimmune neuronal degeneration. We have utilized high titer antibodies present in the serum and CSF of PND patients to classify the disorders and identify the target antigens. However, treatment protocols that reduce antibody titers show no clinical response, and passive transfer of antibody or induction of antibody responses in animals do not reproduce the disorders. We propose to explore the contribution of B and T cells to the clinical disorder; in particular, we suspect and will specifically search for a previously unidentified T cell component to the disease. In preliminary studies, we have been able to find evidence for a specific cytotoxic T cell immune response in the serum and spinal fluid of several PND patients. We will attempt to modify the autoimmune response in order to improve clinical outcome, monitoring both clinical and immunological parameters.

DESCRIPTION (provided by applicant): The long-term objective of this application is to rigorously demonstrate the role RNA binding plays in the function of the paraneoplastic opsoclonus-myoclonus ataxia (POMA) antigen Nova within neurons. We aim to characterize the structural domains in Nova, called KH domains that are capable of high-affinity RNA binding and amenable to analysis by X-ray crystallography. We then aim to identify high-affinity RNA targets capable of cocrystallizing with these Nova KH domains, through the use of RNA selection methods. Nova KH domain-RNA cocrystals will be grown and the structures determined through X-ray diffraction. These structures will allow us to predict and engineer specific mutations in the Nova KB domains that either abrogate RNA binding completely or abrogate sequence-specific RNA recognition. We will test the activity of these mutants using previously established assay systems in heterologous transfected neuroblastoma cells. From these experiments, we will choose "designer" mutants for in vivo expression in mouse brain. These experimental models will be generated by crossing mutant Nova transgenes into (previously generated) Nova-null mice. These studies will allow us to test the hypothesis that RNA binding plays a critical role in generating the phenotype of Nova-null mice, which includes motor neuron degeneration in the brainstem and spinal cord, tissues that are specifically affected in patients with the POMA syndrome.

Study of the paraneoplastic neurologic disorders (PNDs) offers a unique opportunity to study cellular mechanisms of naturally occurring tumor immunity in humans. PND patients harbor systemic tumors expressing proteins normally made only in neurons. Perhaps in part because of this restriction to immune privileged sites, ectopic expression of these proteins in tumors is believed to result in a potent tumor immune response and clinically evident tumor suppression. We have studied the tumor immune response in patients with paraneoplastic cerebellar degeneration (PCD). PCD patients develop tumor immunity to breast or ovarian cancers that express a Purkinje neuronal protein termed cdr2. In preliminary experiments we have identified cdr2-specific CTLs in the peripheral blood of 5/5 HLA- A2.1+PCD patients. These data suggest a cellular basis for the tumor immunity in PCD. We plan to confirm and extend these observations by using a variety of methodologies to detect antigen-specific CTLs in PND patients. We have been able to detect cdr2-specific CTLs using apoptotic cdr2 expressing cell line as the source of antigen, and will examine whether apoptotic antigen expressing tumor cell lines can efficiently serve as the source of antigen for presentation to autologous DCs. We will extend these studies to a new antigen, Hu, targeted in PND patients with small cell lung cancer. We will also explore the role that CD4 cell help plays in triggering effective tumor immunity in PND. Finally, we have found that PND antigens are expressed in a high percentage of tumors obtained from neurologically normal cancer patients. We will examine whether some such patients harbor antigen-specific CTLs and have tumor immunity. Moreover, we will test the hypothesis that three different clinically important sets of cancer patients with PND antigen-expressing tumors can be distinguished in the laboratory; those without tumor immune responses, who we hypothesize have tolerized CTL responses and absent CD4 helper responses, those with tumor immunity in the absence of neurologic disease, who have both CTL and CD4 helper responses, and those with PND, who have broken immune tolerance to neuronal antigens expressed in the brain. These experiments offer the possibility of developing new cancer treatments and advancing the understanding of autoimmune disease of the nervous system.

DESCRIPTION (provided by applicant): Prostate cancers that have spread beyond the confines of the gland regress following the withdrawal or blockade of androgens, and at relapse, following chemotherapy. In both settings the outcomes are similar: most cells undergo growth arrest, but only few undergo apoptotic death. We hypothesize that an immune based approach can eliminate the non-proliferating yet viable cells, particularly after tumor mass has been de-bulked. Prostate cancer offers several advantages in testing new tumor immunotherapies. Serum prostate specific antigen (PSA) levels provide a simple, yet excellent marker of response to therapy. Patients with rising PSA, who have a poor prognosis, can be identified while they are still functionally healthy. As elimination of growth-arrested cells is essential in such patients, they are ideally suited for tumor immunotherapy. Our objective is to demonstrate that immunization of prostate cancer patients with autologous dendritic cells (DC's) cross-presenting apoptotic prostate tumor cells safely induces cytolytic T cell responses to tumor antigens. We will establish a system for the detection of tumor-specific T cell responses in prostate cancer patients that parallels methods established in our laboratory for influenza-specific T cell responses in normal individuals. Apoptotic prostate tumor cells will be co-cultured with DC's, allowing uptake and presentation of multiple tumor antigens on all MHC I molecules. These DC's will then be used to immunize patients. We will monitor patients for acute toxicity and T cell responses to established (e.g. prostate-specific membrane antigen) and new marker antigens present in the apoptotic prostate tumor cells to determine activity of our immunization. Our immunization method and key assays for antigen-specific T cell response are independent of patient HLA haplotype, allowing patients of all haplotype to enter the study. Our strategy is distinct in the breadth of antigen presented and potency resulting from use of the cross-priming pathway--up to 10,000 times more efficient than peptide pulsed DCs. Our results should indicate whether this new approach to DC based, immunotherapy has a potential role in the treatment of prostatic cancer as well as other malignancies.

vaccine development; apoptosis; neoplasm /cancer vaccine; neoplasm /cancer immunology; dendritic cells; monocyte; brain neoplasms; therapy design /development; lymphoma; nervous system neoplasms; glioma; patient oriented research; clinical research; human tissue; human subject;

DESCRIPTION (provided by applicant): This is a proposal to explore the role of RNA dysregulation in the pathogenesis of amyotrophic lateral sclerosis (ALS). The work stems from recent findings that rare patients with familial ALS have mutations in RNA binding proteins. We will use newly developed methods to look at normal and abnormal RNA-protein interactions in tissues obtained from mouse models of ALS and from human ALS samples obtained at autopsy. This work has the potential to uncover new insights, as well as diagnostic markers and therapeutic targets for ALS. In addition, the use of new methods and paradigms will serve as a model for how to approach a growing list of neurologic disorders associated with RNA dysregulation. PUBLIC HEALTH RELEVANCE: This work is a study of a currently fatal human neurologic disorder that currently affects over 20,000 Americans. There is no means of diagnosing the disease short of clinical examination, and there is no treatment. The work offers the hope of understanding the cause, and of developing diagnostic markers and therapeutic targets for the disorder.

DESCRIPTION (provided by applicant): We propose to change the current approach to understanding the molecular basis of memory. Our approach challenges the current focus on quantitative identification of synaptic RNAs by developing new technologies to address protein synthesis-dependent synaptic plasticity: single- synapse-CLIP and synaptic translational profiling. These will allow us to redefine the problem from two new superimposed perspectives: the need to identify regulated RNA-protein complexes in specific synapses, and the need to define their role in translational regulation. Specific synapses will be studied: the apical dendrites of cerebellar Purkinje neurons (a site of motor learning), of CA1 pyramidal neurons in the stratum moleculare of the hippocampus (a site of associative memory), and of layer V pyramidal neurons of the visual cortex (a site of activity- dependent plasticity). Key RNA-protein complexes to be studied in these synapses will be Argonaute (Ago)-mRNA-miRNA ternary complexes, translationally regulated FMRP-mRNA complexes, and neuron-specific RNA regulatory protein-mRNA complexes known to be present in the dendrite and to bind 3' UTRs (nElavl (Hu proteins), Nova). These complexes will be compared with a delineation of all ribosome-mRNA synaptic complexes present in the same dendrites, allowing us to validate interactions by identifying translationally regulated synaptic mRNAs (synaptic translational profiling). Regulated dendritic RNAs will be further validated by assessing for their translational state in two well-studied paradigms of protein synthesis- dependent synaptic plasticity: that in the hippocampus, and in the visual cortex after dark rearng and subsequent light exposure. These studies will revolutionize our understanding of the nature and regulation of local synaptic mRNAs that underlie memory, setting the stage for new insight into neurologic diseases of memory such as Alzheimer's and other neurodegenerative diseases. PUBLIC HEALTH RELEVANCE: Dysregulation of RNA-the key molecule between genomic DNA and proteins-is increasingly recognized to lie at the root of human neurologic disease. More precisely, memory, and complex cognitive function in general, is thought to depend on the regulation of protein synthesis at the sites of neuronal connections. Understanding the mechanisms governing this regulation will be studied here in an entirely innovative set of experiments, and they hold the key to understanding disorders of memory and cognition.

The key challenge addressed in this proposal is to develop a means to harness the power of molecular biology to define therapeutic targets for brain disease. This treatment-oriented approach combines the urgency of a practicing neurologist with the knowledge and technology modern science brings to neuroscience. From the basic science perspective, understanding the fundamental root mechanism of disease is an uncompromising goal. From the neurologist's perspective, the perfect cannot be the enemy of the good, leading to five basic points: · Success to date in the treatment of brain disease offers a key lesson in focus. Treatments target accessible molecules, and this will dictate how we focus big data analysis. · Human neurologic disease is complicated. The best model system for understanding neurologic disorders is the human; studies of human brain material must be integral to developing new treatments. · Regardless of the cause of brain disease?and current neuroscience is appropriately focused on tracing ?genetic? (DNA) etiologies?the manifestations of such defects are mediated by the stoichiometry, distribution and variability of cell-specific RNA regulation and its consequent effects on proteins within affected cells. · Different cell types contribute to different brain disorders, but the difference between individual cells of any one type is unknown. The differences between them are manifest at the level of RNA, not DNA. · The quantity, quality (isoforms) and distribution of targets (e.g., receptors) are enormous. The unique spectrum of diversity of an individual cell type?neurons, astrocytes or microglial cells?is unknown, more so when comparing diseased and normal brain. Using a variety of new strategies, we will study RNA regulation in individual cell types. · Bridging these points together requires new methods and computational approaches. · The net result of contrasting RNA regulation in individual human cell types in health and disease will uncover otherwise hidden cell type-specific targets for therapeutics.