David Cox - US grants

Mathematicians from the Five College Consortium of Amherst, Hampshire, Mount Holyoke, Smith Colleges and the University of Massachusetts will restructure the standard three-semester calculus sequence during the next five years. They will develop a new curriculum in which the four mathematical themes of optimization, estimation and approximation, differential equations, and functions of several variables will be stressed from the beginning. These major mathematical concepts will grow out of exploring significant problems from social, life and physical sciences. Dynamical systems, discrete time models, Fourier series and partial differential equations are some of the concepts which will be explored. The computer will be integrated into the curriculum as a basic conceptual device for structuring the way students think about problems and what it means to solve them. A national Advisory Board of mathematicians, a scientist, and an engineer will contribute ideas, and give overall guidance in the evaluation. Dissemination will be in the form of team-taught courses, weekend retreats, summer workshops for area faculty and high school teachers and publication of the curriculum. These instructional materials will be used at universities, liberal arts colleges, and high schools. The Five Colleges is providing significant cost sharing.

The long term goal of these studies is to determine which chromosome 21 genes result in the characteristic congenital heart disease of Down syndrome when present in three copies. In addition, we wish to determine which chromosome 21 gene result in the presenile dementia and Alzheimer-like neuropathological changes found in Down syndrome. Based on rare patients who are trisomic for a small segment of chromosome 21 in the region of SOD1 and who have congenital heart disease, our working hypothesis is that three copies of a gene or genes in the SOD1 region of chromosome 21 causes heart disease. Based on the tight linkage of the human 21 DNA probe D21S16 to a locus causing autosomal dominant Alzheimer's disease, we hypothesize that trisomy for a gene or genes in the D21S16 region of the chromosome results in the presenile dementia of Down syndrome. To test these hypotheses, we propose to isolate the 10,000 kb regions of human chromosome 21 around SOD1 and D21S16, will be used to define a physical map of each chromosomal region at the 100 kb level of resolution. Analogous physical maps will be constructed for the homologous regions of the mouse genome, and by comparing the position of crosshybridizing probes, the degree of homology between the mouse and human segments will be determined at the molecular level. The isolation and physical maps of these chromosomal segments will provide reagents to test the feasibility of introducing large human chromosome 21 fragments into pluripotent mouse teratocarcinoma cells by protoplast fusion with yeast containing artificial human chromosomes. Ultimately, these cell lines will be used to create "transgenic" mice, which will serve as an in vivo assay system to determine those chromosome 21 regions responsible for heart disease and Alzheimer-like changes in Down syndrome.

The long term goal of these studies is to determine which chromosome 21 genes result in the characteristic congenital heart disease of Down syndrome when present in three copies. In addition, we wish to determine which chromosome 21 gene result in the presenile dementia and Alzheimer-like neuropathological changes found in Down syndrome. Based on rare patients who are trisomic for a small segment of chromosome 21 in the region of SOD1 and who have congenital heart disease, our working hypothesis is that three copies of a gene or genes in the SOD1 region of chromosome 21 causes heart disease. Based on the tight linkage of the human 21 DNA probe D21S16 to a locus causing autosomal dominant Alzheimer's disease, we hypothesize that trisomy for a gene or genes in the D21S16 region of the chromosome results in the presenile dementia of Down syndrome. To test these hypotheses, we propose to isolate the 10,000 kb regions of human chromosome 21 around SOD1 and D21S16, will be used to define a physical map of each chromosomal region at the 100 kb level of resolution. Analogous physical maps will be constructed for the homologous regions of the mouse genome, and by comparing the position of crosshybridizing probes, the degree of homology between the mouse and human segments will be determined at the molecular level. The isolation and physical maps of these chromosomal segments will provide reagents to test the feasibility of introducing large human chromosome 21 fragments into pluripotent mouse teratocarcinoma cells by protoplast fusion with yeast containing artificial human chromosomes. Ultimately, these cell lines will be used to create "transgenic" mice, which will serve as an in vivo assay system to determine those chromosome 21 regions responsible for heart disease and Alzheimer-like changes in Down syndrome.

Funds are requested to support an international workshop to determine the status of genetic and physical maps of human chromosome 21. The workshop, entitled, "Mapping Chromosome 21: Present Status and Future Strategies" will be held in Bethesda, Maryland on April 2-3, 1990 and will include twenty scientists from the USA as well as ten scientists from outside the USA. The primary focus of the meeting will be an analysis of data relating to the chromosome 21 map. The overall objective of the meeting is to highlight inconsistencies in the data, and to discuss strategies to expedite completion of the physical map of chromosome 21. Specific aims of the meeting will be to generate composite genetic and physical maps of chromosome 21 based on both published and unpublished data, and to define chromosome 21 specific "anchor loci." The meeting is intended to broaden collaborative efforts and foster more extensive interactions among the participants.

This three-year teacher enhancement project will provide in- service science content and pedagogy instruction to 80 teachers from 40 middle and high schools. It will emphasize the use of meteorological observations to facilitate the development of scientific concepts and process skills which are relevant to students' everyday lives. Teachers will learn to use and maintain a network of computer linked weather stations for classroom instructions. The project includes academic year in-service courses, annual fall and spring conferences, annual three-week summer workshops and academic year classroom visits by professional meteorologists. The project begins with workshops for 30 teachers in the summer of 1990. Participants may earn six graduate credits from Portland State University for the summer workshops and (1-4) credits for the academic year work. Working scientists are ongoing participants and the resources of public and private schools are combined with those of the university, business community, and professional organizations and agencies. The cooperative pooling of human and fiscal resources from school districts, universities, the business community, professional organizations and other agencies, is a compelling feature of this project. Cost sharing in an amount of $460,219 equals 122% of the NSF award.

The overall goal of this project is to use radiation hybrid mapping to establish a contiguous physical map of the human genome with an average resolution of 100 kb between adjacent markers. Our aim is to determine the order of 30,000 unique DNA landmarks with respect to one another and to obtain an estimate of the physical distance between adjacent markers. Two different sets of radiation hybrid cell lines consisting of approximately 85 independent cell lines per set will be used to construct the map. Set 1 hybrids, which presently exist, will be scored with 6000 selected markers over the first 12 to 18 months of the grant to generate a contiguous physical map of the human genome with an average resolution of 500 kb. Since the Set 1 hybrids do not contain a sufficient number of X-ray breaks to generate a 100 kb average resolution map of the genome, we will generate a second set of radiation hybrids (Set 2) in the first year of the grant. Set 2 hybrids will be scored with 30,000 markers over the remaining four years to generate a contiguous map of the genome with an average resolution of 100 kb. 3000 of the markers on our maps will consist of polymorphic loci and/or genes previously laced on meiotic and/or physical maps by other scientists. These 3000 markers will provide the basis for integration of our maps with preexisting maps. In addition, DNA from Set 1 cell lines, as well as the cell liens themselves, will be made available to other genome mappers, providing a basis for integration of our maps with other maps presently under construction. Since the 30,000 markers scored on Set 2 hybrids will be designed to span the majority of the estimated 20,000 Sfi 1 restriction map of the human genome (Project 3). Such integration will allow for a comparison of order and distance information generated from radiation hybrids with that obtained from uncloned genomic DNA.

Algebraic algorithms discovered in the last three decades, coupled with inexpensive high speed computation have created a minor revolution at the research level in algebraic geometry. The new computational methods have found wide application in a number of other fields (including robotics, computer science, and CAD). These ideas are accessible to undergraduates and, in fact, sometimes simpler to grasp than the material in the standard abstract algebra course. This project is creating materials which introduce and explore these methods. The materials can be used as the basis of a number of useful and accessible first algebra courses for students of science, mathematics and engineering.

Our overall goal is to 1) identify a dense set of single nucleotide polymorphisms (SNPs) spanning the human genome, 2) identify two alleles at each variant site, 3) order these SNPs with high confidence on a single integrated high-resolution map of the human genome containing genes and simple sequence repeat (SSR) markers, and 4) place all of this information in a public database, accessible to a wide range of potential users. We propose a three-year collaborative effort between the Stanford Human Genome Center (SHGC) and Affymetrix, Inc. to screen 24,000 sequence tagged sites (STSs) per year for polymorphic variation using Affymetrix DNA arrays. Each STS will be derived from BAC clones and amplified from the genomic DNA of eight anonymous unrelated individuals. Our preliminary data suggest that we will identify 6,000 such polymorphisms per year, with one in every four STSs containing a common variant. We will use an independent technology (ABI gel-based fluorescent sequencing) on DNA from two informative individuals, identified by chip-based screening, to confirm each SNP. Each polymorphic allele will be placed in a publicly-accessible database within 90 days of demonstrated consistency between gel-based and chip- based analyses. Our final product will be a map of 72,000 STSs, including 18,000 SNPS, integrated with 30,000 previously-mapped genes and simple sequence repeat (SSR) markers, ordered with greater than 95 percent confidence at 100 kb resolution.

The major hypotheses to be tested is that altered levels and patterns of gene expression are present in the brains of patients with depressive illness as compared to normal controls and can be used to define disturbances of the neuronal circuitry of depressed patients involving the dorsal lateral prefrontal cortex (DLPFC), the anterior cingulate gyrus, and the mediodorsal (MD) thalamic nucleus, the paraventricular nucleus (PVN) of the hypothalamus, the hippocampus, and the raphe nuclei. We hypothesize that by simultaneously assessing the levels and expression patterns of thousands of genes from multiple areas of individual brains, we will be able to define a consistent pattern of brain dysregulation in depressed patients that would be difficult if not impossible to recognize by studying the pattern of expression of single genes in single brain regions. The specific research proposed in this application will define the levels and patterns of expression of 10,000 distinct human genes known to be expressed in the brain. These studies will be carried out using samples of five different brain regions (DLPFC; anterior cingulate gyrus, thalamus; hypothalamus; hippocampus; raphe) from a total of 35 matched brain pairs (70 brains total). Ten matched pairs of male and female brains will come from normal individuals. These samples will be studied initially to assess the degree of variability of brain gene expression between normal individuals. The remaining twenty five matched pairs will consist of depressed and control brains. Ten of these pairs of matched depressed/control brains will be used to define changes in gene expression characteristic of depression. The remaining fifteen pairs of matched depressed/control brains will be used in an effort to replicate the gene expression findings from the first ten depressed/control pairs. Specific research aims include: 1) Use the polymerase chain reaction (PCR) to amplify 10,000 different human DNA fragments (Sequence Tagged Sites), each representing a distinct human gene expressed in the brain. 2) Use an automated system to construct micro arrays of the 10,000 brain STSs on glass slides. 3) Use the micro arrays from Aim 2 in conjunction with tissue samples from normal brains to determine the normal range of messenger RNA (mRNA) level for each of the 10,000 genes. 4) Use the microarrays from Aim 2 in conjunction with tissue samples from matched normal and depressed brains to identify consistent alterations in gene expression specific to depressed patients.

Disorders of the cardiovascular (CV) system are frequently due to temporal or quantitative changes in the expression of a large, but finite set of genes. Noncoding cis regulatory sequences play a central role in controlling gene expression and inter-species (i.e., human/mouse) genomic sequence comparisons serve as a rapid and accurate means for identifying such noncoding regulatory elements. The central goal of this PGA will be to use a comparative genomic approach first to identify, and them to determine the function of elements regulating the expression of genes affecting the CV system. The activities of this PGA are not centered on the discovery of new genes, but rather upon using comparative genomics to understand the role of cis regulating elements in the expression of genes already being studied by CV researchers. In this integrated program to "genomically" explore the regulation of CV genes, 200 human genomic intervals (=~200 BACs), each containing a CV gene(s), will be comparatively characterized. The components of this program will include: (1) The acquisition of orthologous human/mouse and other mammalian genomic sequence for a set of prioritized CV genes. Sequences will either be accessed from publicly funded databases or generated by the sequencing component of this PGA. (2) The creation of a cardiovascular comparative genomic database that will contain extensively annotated human and mouse sequences including the localization of conserved noncoding elements in proximity to well studied CV genes. (3) Genome-wide expression profiling to discover genes co-regulated with CV genes and identify shared noncoding regulatory elements, through intra-species analysis. (4) The identification of SNPs within conserved non- coding sequences, and analysis of their effect on CV gene expression in humans. (5) Analysis in genetically engineered mice of a prioritized set of the conserved noncoding elements for their role in CV gene expression. (6) The establishment of an educational program for cardiovascular researchers in the use of genomic databases and tools.