Michael Wigler - US grants

This laboratory is involved in the isolation and characterization of oncogenes. In particular, we are conducting a molecular analysis of action of the ras proteins. Both the immediate and long-range targets of action of these proteins, the "ras pathway," are being determined and the interconnection of this pathway with known pathways is being studied. The yeast S.\cerevisiae has ras genes which are structurally, biochemically, and functionally closely related to the mammalian ras genes. Preliminary work strongly suggests that the yeast ras genes exert their effects by modulating cyclic AMP levels. Studies of the yeast ras will help guide studies of the mammalian ras. In addition, we propose to utilize recent advances in techniques of gene transfer to search for genes which function to restrain cellular proliferation. (Y)

The goal of this program is to understand cellular transformation at the molecular level: both the genetic events which induce transformation and the alteration in biochemical events which ensue. There are five component sections: Human Oncogenes (Wigler) focuses on new human oncogenes which have either a novel structure or an interesting pattern of expression in neoplasia. Oncogenesis in Transgenic Mice (Hanahan and Field) utilizes a relatively new technology to explore the multistep nature of carcinogenesis in specific tissues. c-fos Transcriptional Activation (Gilman) explores the molecules which induce fos oncogene expression in response to signal transduction. Tropomyosins and Transformation (Helfman) analyses changes in the expression of the tropomyosin isoforms and their effects on the transformed phenotype. Membrane Signal Transduction (Feramisco and Bar-Sagi) describes the interactions of various second messenger systems related to growth proliferation signals which originate at the cellular membrane. The project brings together five diverse approaches to a common problem.

The RAS genes were first discovered as the transforming principles of oncogenic retroviruses, but have subsequently been found to be important normal constituents of the mammalian genome. They have been highly conserved in evolution and are members of a larger family of genes which encode small molecular weight guanine nucleotide binding proteins. Although the function of RAS remains unknown, there has been recent, striking progress in understanding of the RAS proteins, and much more can be anticipated in the near future, including: an understanding of effector function for RAS in micro-organisms, discovery of a large number of additional, RAS-related genes and their modes of action, the regulation of RAS and RAS-related proteins, their processing, their structure, and their physiological role. It is becoming clear that a vast array of cellular processes are controlled by these proteins, and hence, an understanding of RAS is not only essential for an understanding cellular growth control, but for much of cell biology as well. About 300 investigators representing cell biology, molecular biology, pharmacology, and structural biochemistry are expected to participate. The proposed meeting on "The Function and Evolution of RAS Proteins" will bring together investigators in the field of RAS research in order to accelerate the exchange of information, promote cooperation and catalyse the development of fresh approaches to the understanding of this important cell biological problem.

The RAS gene were first discovered as the transforming principles of oncogenic retroviruses, but are normal constituents of the mammalian genome. They have been highly conserved in evolution and are members of a larger family of genes which encode small molecular weight guanine nucleotide binding proteins. Mutant, oncogenic cellular RAS genes have been found in a high percentage of human cancers, suggesting a common underlying metabolic defect in a large proportion of human malignancy. Yet the function of oncogenic RAS, or even normal RAS, remains unknown. There has been recent and striking progress in our understanding of the RAS proteins, and much more can be anticipated in the coming year, including: an understanding of effector function for RAS in micro-organisms, discovery of the large number of RAS related genes and their modes of action, the regulation of RAS and RAS related proteins, their processing, their structure and their physiological role. It is becoming clear that a vast array of cellular processes are controlled by these proteins, and hence an understanding of RAS is not only essential for an understanding of cancer, but for much of cell biology as well. The purpose of the proposed meeting is to bring together the leaders in the field of RAS research in order to accelerate the exchange of information, promote cooperation among scientists and catalyse the development of fresh approaches. We expect 300 participants from various fields including cell biology, molecular biology, pharmacology, structural biochemistry, and signal transduction. Registration will be open, provided that the meeting is not oversubscribed. In the latter case, selection of participants will be made according to their field of research. There will be about fifty speakers. The meeting will be international. There will be no publication of the proceedings.

cAMP is a second messenger involved in many of the normal physiological responses of tissues. Levels of cAMP can be modulated by production, through adenylyl cyclases, or destruction, through cAMP phosphodiesterases. A large number of biochemically and pharmacologically distinct forms of cAMP phosphodiesterases (PDEs) have been found in mammalian tissues. Most tissues express more than one form of the enzyme, yet there are clear tissue specific patterns of expression. Several forms of cAMP PDE can be specifically targeted pharmacologically, and the therapeutic modulation of cAMP levels can beneficially alter the course of a variety of disorders. Because of the complexity of expression of different forms of cAMP PDE, and because of their potential as targets for pharmacological intervention, it is desirable, and probably necessary, to isolate and characterize the genes encoding cAMP PDEs. We have been exploring the RAS/adenylyl cyclase pathways in the yeast S. cerevisiae. These studies have led us to clone and characterize the yeast genes encoding cAMP PDE. As a consequence we have developed a general method for isolating mammalian cAMP phosphodiesterase genes using mammalian cDNAs cloned into yeast expression vectors and genetic selection in yeast. At least two distinct forms of mammalian PDEs have been cloned by us and expressed in yeast entirely lacking endogenous PDEs. One of these is a member of the dunce-like family of PDEs, and one is novel. We propose to characterize these and other human PDEs expressed in brain, heart and other tissues. This work has potential application for the understanding and control of a wide range of conditions including heart, circulatory and pulmonary dysfunction, mental disorders, and perhaps immunologic disturbances.

cAMP is a second messenger involved in many of the normal physiological responses of tissues. Levels of cAMP can be modulated by production, through adenylyl cyclases, or destruction, through cAMP phosphodiesterases. A large number of biochemically and pharmacologically distinct forms of cAMP phosphodiesterases (PDEs) have been found in mammalian tissues. Most tissues express more than one form of the enzyme, yet there are clear tissue specific patterns of expression. Several forms of cAMP PDE can be specifically targeted pharmacologically, and the therapeutic modulation of cAMP levels can beneficially alter the course of a variety of disorders. Because of the complexity of expression of different forms of cAMP PDE, and because of their potential as targets for pharmacological intervention, it is desirable, and probably necessary, to isolate and characterize the genes encoding cAMP PDEs. We have been exploring the RAS/adenylyl cyclase pathways in the yeast S. cerevisiae. These studies have led us to clone and characterize the yeast genes encoding cAMP PDE. As a consequence we have developed a general method for isolating mammalian cAMP phosphodiesterase genes using mammalian cDNAs cloned into yeast expression vectors and genetic selection in yeast. At least two distinct forms of mammalian PDEs have been cloned by us and expressed in yeast entirely lacking endogenous PDEs. One of these is a member of the dunce-like family of PDEs, and one is novel. We propose to characterize these and other human PDEs expressed in brain, heart and other tissues. This work has potential application for the understanding and control of a wide range of conditions including heart, circulatory and pulmonary dysfunction, mental disorders, and perhaps immunologic disturbances.

[unreadable] DESCRIPTION (provided by applicant): [unreadable] We will continue developing and applying tools for the genomic analysis of cancer. Specifically, [unreadable] 1. We have developed a representational array based method for the measurement of gene copy levels in cells, and propose a high resolution analysis of tumor cell lines to develop a database of commonly amplified and deleted loci. [unreadable] 2. We will use this database to identify candidate tumor suppressor genes, that will be validated by the analysis of somatic mutations in tumors. [unreadable] 3. We will perform a functional analysis of lead candidates, utilizing [[among other tools]] microarray expression analysis and tumorigenicity experiments [unreadable] [unreadable]

Nestler 9728062 Encoded combinatorial chemistry has created a new paradigm in the quest for synthetic low molecular weight receptors. Branched peptidic molecules that can bind peptides in a sequence-specific manner have been found in encoded combinatorial libraries. The potential of this system is used in this study to derive low molecular weight receptors that are capable of binding to and modulating RAS proteins, which play central roles in signal transduction, and single amino acid mutations render the wild-type protein a potent oncogene. Experiments are designed to elaborate a new strategy for blocking biochemical processing. Instead of inhibiting the enzyme responsible for a biochemical modification, the substrate is rendered inaccessible for the transforming enzyme by small synthetic receptor molecules that bind to the domains to be modified. More specifically, small molecules that bind to the carboxy-terminal domain of RAS are derived from encoded combinatorial libraries of branched peptidic receptors. Stepwise screenings are used to select for tightly binding and highly selective "molecular forceps" with ability to interfere with RAS processing through preventing the attack of the farnesyl transferase to the carboxy-terminus of RAS. The results of this study and the methodologies developed to achieve them, should lay the groundwork for the discovery of small molecular agents that can modulate signal transduction pathways and other cellular processes of physiological importance.

In Phase One we will test if application of representational techniques can enable the development of high resolution, global genomic micro- arrays for measuring changes in gene copy number (amplification and deletion), loss-of-heterozygosity (LOH), and point mutation frequency in tumor DNA derived from biopsies. In particular, we will assay for these changes in low complexity representations (LCRs) of tumor and normal genomes. In Phase Two we will scale up the process of collecting LCR probes and their associated map positions. We will fabricate arrays of DNA probes using the conventional capillary arrayer format to achieve densities between 3,000 and 30,000 probes. These arrays will be tested under a variety of controlled and field test conditions to monitor their utility for detecting changes throughout the tumor genome. We will develop database and analysis tools so that the results of experiments can be readily accessed in a positionally ordered manner.

We propose to expand a pilot project to develop a high density gene copy number micro-array based on low complexity genomic representations. Such a tool will lead to improved classification of cancers, which will likely impact all areas of cancer diagnosis and treatment, and be an enormous boon to the discovery of cancer causing genes. We are poised to scale up from arrays of a thousand probes to sets of probes in excess of 30,000 that can be rapidly mapped to very high resolution in array format. Our method will be able to resolve changes in the genome with a resolution of every 50 to 100 kilobases. Moreover, we believe that we can significantly enhance the closure of human genome sequencing project by providing independently derived BAC contigs and probes for gaps in the existing BAC maps.

DESCRIPTION (provided by applicant): The general aim of this proposal is to develop and verify micro-array based genomic mapping tools. These tools are based on representational approaches to genomic sampling, coupled with powerful statistical and algorithmic methods for data analysis. The tools are designed to enable integration of genetic, physical and transcript maps. Specifically, we describe: 1. Correspondence Mapping. This method makes assignments of arrayed probes to BACs that contain them, on a very large and parallel scale. It uses algorithms based on binary partitions. Among its uses are finding BACs corresponding to unfinished regions in genome assembly projects, finding BACs that correspond to expressed regions of the genome, and relating probes with annotation (for example, relating probes that have been physically or genetically mapped to probes that correspond to expressed sequences by finding BACs to which both probe sets belong.) 2. Linear Mapping. This method clusters probes into contigs, and within contigs establishes a linear ordering with pairwise probe distances. It uses algorithms based on Hamming metrics. It can be applied to fine mapping of the genome using a set of probes that are dense within the genome and a library of BACs, or to coarse mapping of a sparse set of probes and radiation hybrids. By using overlapping sets of probes, the coarse map can be used to orient the contigs from fine mapping. 3. Genetic Mapping. This method allows massive parallel array-based genotyping for a set of probes, VLCR SNPs, that derive from a "sliver" of the genome that can be amplified en masse. These probes are a subset of the probes that can be mapped by the other methods.

DESCRIPTION (provided by applicant): Autistic Spectrum Disorders (ASD) are characterized by delay in or absence of language acquisition, deficits in social interactions and repetitive behaviors. ASD are largely genetic in origin, and occur either sporadically (simplex) or in a familial (multiplex) pattern, are far more commonly in males (4:1 ratio over females), and have an overall incidence of ~1 in 150 births. Identification of genes responsible for ASD has been complicated by many factors, such as the small number of autistic pedigrees-which may reflect the sporadic incidence-as well as the lack of consistent clinical data among existing samples, particularly with respect to the less severe (and more phenotypically heterogeneous) portion of the autistic spectrum. This project is intended to address the genetic basis of ASD through a high-throughput, candidate-gene sequencing approach. The proposed approach is designed to complement existing comparative genomic hybridization (CGH) analyses. The combination of identifying novel ASD candidate genes through the ongoing CGH study and detailing the mutational spectra of 100 known/predicted and novel ASD candidate genes through this work will reveal a major fraction of the genetic variation underlying ASD. For these experiments, the Simons Simplex Collection (SSC) will be utilized as the sample population. The SSC is the largest high- quality set of ASD families assembled so far, and it is specifically designed to compensate for the shortcomings of existing ASD family populations by ensuring comprehensive and consistent clinical analyses. More importantly, it is a collection of simplex families, and as a result the proportion of cases of autism due to spontaneous mutation-as opposed to inheritance-is maximized. A likely outcome of this work will be significant advances in molecular screening of ASD among young children. This impact would be felt in many ways: more precise diagnosis with respect to clinical subtypes of ASD;assessment of ASD severity based on genetic markers;and treatment more specifically tailored to the needs of affected individuals. Given the heterogeneity of ASD and current lack of markers, this study stands to provide significant progress in deconvoluting the complex phenotypes associated with autism. This work will also provide new avenues for future studies, from new mouse models to elucidation of genetic pathways required for language acquisition, social interactions and behavior. PUBLIC HEALTH RELEVANCE: This study will yield significant insight into the molecular basis of autistic spectrum disorders, which are largely genetic in origin, highly prevalent among all populations, and difficult to diagnose and classify. This work will lead to advances in molecular screening of ASD among young children, which will in turn result in improved diagnoses and more precisely targeted treatment regimes.

? DESCRIPTION (provided by applicant): Recent studies of cancer genomes on a single-cell level have revealed the complexity of the disease and the presence of multiple genealogically related cell populations in a tumor. Detailed knowledge of the clonal structure of a cancer potentially is of high clinical value: multiplicity of clones or of lesions in most advanced clonesis a possible measure of progression; spatial pattern of clone dispersal in a tumor may signal elevated propensity to invade; lesions observed in individual clones but not in the bulk tissue may point to targets for therapy. DNA copy number profiling of cells from low-coverage sequencing is an accurate, economically feasible technological approach to the study of cancer sub-population structure. Novel multiplex sequencing techniques, developed by the Wigler lab at CSHL, permit simultaneous sequencing of hundreds of single-cell DNA specimens and their subsequent copy-number profiling at up to 50kb resolution. Optimal use of this data type for robust reconstruction of cancer cell phylogenies is a challenging computational problem requiring new informatic and statistical tools. The dual purpose of this project is, first, to deveop, optimize, test and deploy such tools; and, secondly, integrate the results of phylogenetic analysis with other forms of genomic, biological, pathological and clinical data. The centerpiece of the proposed phylogenetic analysis pipeline is to use a recently developed computational approach termed CORE (Cores Of Recurrent Events) for transforming copy number profiles of single cells into a form suitable for phylogeny. CORE is a flexible method, with multiple options allowed for several of its components. This freedom of choice will be exploited for optimization of performance. The entire pipeline will be tested using both simulated data with known underlying phylogeny and genomic profiles from pathologically characterized tumor tissues we have already generated. Preliminary application of CORE to the phylogenetic reconstruction problem showed ample promise to justify further study. In parallel, we propose to design an interactive user interface to single-cell genomic data integrated with their phylogenetic interpretation and several other forms of genomic, clinical and pathological annotation. The potential utility of such a comprehensive integrated interface is strongly suggested by a preliminary study. All software products to be developed in the course of the project will be made available to the community, for use and further development.