James Sikela - US grants

Protein phosphorylation systems of the brain are extraordinarily more active then those of other tissues and appear to represent a key brian mechanism, effecting a multitude of diverse neuronal functions. However, in contrast to the especially prominent role protein phosphorylation appears to play in brain function, very little is known regarding the genes encoding brain protein kinases, and even less is known about the true physiological role(s) such kinase fulfill in the brain. Therefore the first major objective of this proposal is to obtain fundamental information regarding the genes encoding brain protein kinases in two specific ways: 1) A detailed molecular characterization of the gene encoding a new brain Ca++/calmodulin-dependent protein kinase (CaM kinase IV) will be developed using a partial-length cDNA as a starting point. Emphasis will be placed on obtaining basic information with regard to the sequence, structural organization, expression, and regulation of this gene. We have recently determined that the human CaM kinase IV gene maps to the same chromosome position (5q22) as the gene for familial polyposis coli (FPC), and inherited form of colon cancer. Therefore, the possibility that the genes are closely linked will be investigated by a) identification of RFLPs associated with the CaM kinase IV gene and b) use os such RFLPs in linkage analyses with FPC families. 2) A novel molecular cloning strategy, homology probing will be utilized to identify genes encoding previously uncharacterized protein-serine kinases present in the brain. In this approach brain cDNA libraries will be screened with two sets of degenerate oligonucleotide probes derived from conserved regions of the catalytic domain of known protein-serine kinase. Once a clone has been identified by sequence as encoding a new serine kinase, a detailed characterization of the corresponding gene, in the manner described above for the CaM kinase IV gene, will be initiated, with priority given to those kinases that appear to be brain-specific. In addition, a second approach, i.e. he use of 125I-calmodulin as a probe to screen brain expression libraries, will also be explored. The second major objective of this proposal is to obtain information about the true physiological function(s) of these new kinases in the brain. Initial studies will utilize synthetic peptides, derived from the deduced amino acid sequence of the cloned kinase cDNAs, to generate antibodies of predetermined specificity. Such antibodies would be used to isolate the corresponding kinase which, in turn, would then be subjected to biochemical and functional characterization, including, for example, analysis of catalytic activity, second-messenger dependence, and substrate specificity. The antibodies would also be used in immunocytochemical analyses of brain at the light and electron microscopic levels to obtain information about the regional, cellular, and subcellular distribution of the kinase. As a longterm strategy, the use of transgenic nice will be explored as a possible approach to obtaining a more comprehensive understanding of the in vivo function of these kinases. While technically challenging, the ability to produce, and then analyze at multiple levels, mice in which specific kinase genes are under- or over-expressed would represent a powerful approach to understanding the biological role(s) such proteins fulfill in the brain.

The goals of this component are to identify the mouse gene underlying Lore2, a QTL for alcohol sensitivity located on mouse chromosome 2, and to provide congenic mouse strains for the use of other components of the University of Colorado Alcohol Research Center (CU ARC). The focus of our work is the ISS and ILS strains of mice which differ markedly in the length of loss of righting reflex (LORR) after ethanol administration. This differential LORR has been shown to be associated with several distinct chromosomal regions (QTLs), each of which explains part of the difference. Lore 2 explains about 14.2 % of the genetic variance in sleep time (a 25 minute difference between the inbreds). This QTL has been detected has been detected both in RIs (p less than 0.001) and in F2's (lodmax=6.6) and has been confirmed using directed genome selection methodologies. Lore2 is located near 85 cM of chromosome 2 (l-lod-support interval is from 78-95 cM. Component 3 will simultaneously carry out mouse breeding efforts that will narrow the Lore2 interval to less than 1 cM (Dr. Johnson) and gene identification strategies that will candidate genes in the interval for sequence differences (Dr. Sikela). We will also interact with other ARC investigators to establish functional confirmation of the gene(s) identified by this component and provide congenic strains for the other components. The mouse genetic component will perform two different functions. First, a series of congenics will be constructed carrying each short-sleep QTL on the ILS background (the complementary congenics, L alleles on an ISS background, are being developed by an RO1). Second, the Lore2 QTL will be localized to 1cM or less using markers spanning the 78-95 cM interval on chromosome 2. Both general aims use speed congenic approaches to backcrosses of ISS to ILS and the reciprocal together with typing of SSLP markers. Three rounds of backcrosses will be performed using the assistance of a speed congenic analysis followed by assessment of phenotype if mice in the fourth round and subsequent completion of seven more rounds of backcrosses. Mice will be assessed for recombination events in this interval on chromosome 2 to localize Lore2 to approximately 1cM at better than a 50% probability. For gene identification, the primary strategy will be to utilize the rapidly emerging mouse and human genome resource relating to ESTs/genes and gene maps. Through the use of these resources, mouse cDNAs that lie in the Lore2 interval will be identified and used for PCR-based direct sequence comparisons between ILS and ISS mice. Once sequence differences are obtained they will be tested by appropriate functional assays and other confirmatory methods though collaborations with other ARC laboratories. The human counterpart of the resulting Lore2 gene will be a candidate for contributing to genetic predisposition to alcoholism. (For those unfamiliar with the specialized genetics terminology, a lexicon has been provided as Table 1.)

DESCRIPTION (provided by applicant): The primary goal of this revised proposal is to identify genes underlying quantitative trait loci (QTLs) related to alcohol action. Considerable progress has been made toward this goal in the first five-year funding period and now important new mouse genomics resources and strategies will be utilized that should significantly accelerate this process even further. The research plan for the coming funding period will be composed of two primary areas of focus: The first (Specific Aims 1-2) will be on four Lore QTLs related to initial sensitivity to ethanol in the ILS and ISS mice. Considerable prior work has been carried out on the Lore2 QTL which will now be added to this proposal. The proposed gene identification efforts will involve the continued characterization and fine mapping of three altered Lore QTL genes we have identified, comparative sequencing of additional, newly identified Lore candidate genes, and identification of regulatory sequence variations within differentially expressed QTL genes. Lore-specific custom gene chips have been created that together with standard mouse chips will allow the expression level of all Lore genes to be compared between ILS and ISS strains. The second (Specific Aims 3-4) will be on application of a novel in silico method we have recently developed for rapidly identifying gene variants in alcohol-related QTLs identified using the C57BL/6J (B6) and DBA/2J (D2) strains. This method represents a powerful new tool that will be used to identify gene variants in two alcohol-related B6xD2 QTLs. These four specific aims have been designed to be comprehensive, and will be used to search for potentially important interstrain changes by surveying both the protein coding and regulatory regions of QTL genes. These analyses will utilize a combination of high throughput comparative DNA sequencing, high density microarray studies and the new in silico methods that take advantage of strain-specific whole genome sequences that have just become available. In response to the last review, Specific Aim 5 has been significantly modified and reduced in scope from the previous proposal. This aim will now focus on the development of BAC genomic libraries for the ILS and ISS mice, which will set the stage for functional studies using transgenic animals. These experiments represent a logical extension of previous work from this laboratory and provide a comprehensive approach to the identification of QTL genes involved in alcohol action.

The unprecedented pace at which genetic technology and knowledge is advancing threatens to overwhelm society's ability to provide sound ethical guidance regarding how such advances should be used. Given the complexity of the science and the difficult ethical choices society is faced with, the development of individuals who have expertise in both areas would be highly desirable. With this goal in mind, this research proposal seeks to provide Dr. Sikela, an established genome scientist, with a rigorous and effective training experience in biomedical ethics, with a specific focus on the ethical and societal implications of genetic enhancement technologies. Drs. Mark Yarborough and Eric Juengst will serve as co-sponsors of the fellowship. The training experience will focus primarily on research (80 percent), with some time also devoted to coursework (15 percent) and teaching (5 percent). The research will investigate several critical issues related to genetic enhancement technology, including an investigation of the ethical challenges that might be posed by the ability to alter the genetic makeup of future generations. The research will be comprised of discussions with sponsors and other experts in the field, reading of relevant literature, and active involvement in conferences, seminars, courses, and teaching that relate to biomedical ethics and genetic enhancement. It is the intent of this training experience to provide Dr. Sikela with the expertise that will allow him to make significant contributions to society's effort to deal with these emerging ethical issues.

DESCRIPTION (provided by applicant): The primary goal of this application is to identify genes and other genomic features that underlie the action of alcohol. While we have made considerable progress toward this objective through the use mouse quantitative trait loci (QTL) mapping, gene expression profiling and comparative DNA sequencing, these studies were limited and complicated by two newly discovered factors related to the analysis of complex traits 1) mammalian genomic variation, including copy number variation (CNV), is much more prevalent than previously expected , and 2) mouse genome sequence contains many gaps and complex, duplication-rich regions that are likely to be misassembled. We plan to apply two recently developed genomic strategies that circumvent many of these limitations. The first strategy will be to carry out very high density array-based comparative genomic hybridization (aCGH) of several mouse strains (ILS, ISS, LXS RIs and HS4) known to differ in alcohol-related phenotypes, and identify CNVs that are contributing to differences in alcohol action between these strains. CNVs will be independently confirmed, CNV boundaries fine-mapped using custom aCGH arrays, and genes within the CNV identified and characterized. To our knowledge this study will be the first genome-wide investigation of the role of CNVs in alcohol action. The second strategy will be to obtain and compare the complete DNA sequences of four QTL regions (Lores1, 2, 4, and 5) between strains that have been selected for differential sensitivity to alcohol (ILS and ISS). This will be carried out by array-capture of QTL- specific DNAs from each strain, followed by ultra-high throughput DNA sequencing of the complete QTL interval using next generation DNA sequencing technology. This study will identify all DNA sequence variations within each Lore QTL and, as such, will represent the most complete assessment of genomic variation with these alcohol-related QTLs. PUBLIC HEALTH RELEVANCE:The proposed experiments have the potential to identify important new genes and pathways that underlie the action of alcohol in mammalian systems. Such new insights could lead to improved strategies for the treatment and prevention of alcoholism and alcohol abuse.

SUMMARY We have established that human genome sequences encoding a novel protein domain, DUF1220, are highly amplified in the human lineage (>200 copies vs. 1 in mouse/rat) and may be important in human-specific cognitive function. The majority of DUF1220 domains are located at 1q21.1, one of the most complex regions of the human genome and filled with gaps and segmental duplications. Copy number variations (CNVs) in the 1q21.1 region have now been implicated in numerous diseases associated with cognitive dysfunction, e.g. autism, mental retardation, schizophrenia, microcephaly and macrocephly. These findings may be indicative of a novel recurrent rearrangement and reflect a new cognition-related syndrome specific for the 1q21.1 region. In order to more precisely identify the CNV boundaries and causal disease genes in these patients, a haploid BAC library will be used to generate a finished sequence map of the region. Sequencing of 1q21.1 BACs from a Hydatiform (haploid) mole library will be carried out in collaboration with Dr. Rick Wilson at the Washington Univ. at St. Louis Genome Center to generate a single haplotype path across the region. The finished 1q21.1 sequence will be used for fine mapping of already identified disease-associated CNVs for autism, mental retardation, microcephaly and macrocephaly, through collaborations with the laboratories of Drs. Evan Eichler and James Lupski. High-density custom tiling arrays will be generated for the finished 1q21.1 region and used for array CGH to fine map CNV breakpoints in these patients and identify candidate genes. In addition the role of DUF1220 domain copy number in autism will be investigated by QPCR analysis of individuals with autism using DUF1220-specific primers. To investigate the function of DUF1220 domains in a living mammal, DUF1220-minus mice we have generated (the first animal model for DUF1220 function) will be subjected to behavioral testing to assess the affect of DUF1220 domain loss on learning and memory.

DESCRIPTION (provided by applicant): DUF1220 protein domains are highly amplified in the human lineage, show signs of positive selection and, in brain, are highly expressed in regions associated with higher cognitive function. Virtually all DUF1220 domains are primate-specific, with the highest copy number (212) found in humans;mice have only one copy and none of the primate type. The majority of DUF1220 domain sequences are found at 1q21.1, and several reports have found copy number variations in 1q21.1 associated with a number of cognitive diseases including autism, schizophrenia and microcephaly. Given the potential importance of DUF1220 to human cognitive disease and brain evolution, the goal of this application is to develop the first animal model encoding human DUF1220 domains. Two human BAC clones each encoding a different human DUF1220 gene and multiple DUF1220 domains, will be used for the production of transgenic mice. Additional sequences (tauGFP and tau/Tomato) will be added to each construct to allow expression of the gene to be followed in both neuronal cell bodies and axonal projections. The mice will be initially characterized by Western analysis using the anti- DUF1220 antibody we have developed that sees the primate/human-specific DUF1220 domains but not the single DUF1220 domain found in mice. For follow-up characterization, those mice expressing the two DUF1220 transgenes will be used 1) for cross-breeding experiments to increase the DUF1220 copy number to a more human- like level, 2) for high-resolution immunocytochemical analysis of brain expression patterns, and 3) for a series of behavioral tests designed to monitor learning and memory abilities. PUBLIC HEALTH RELEVANCE: This work should generate the first animal model to study the neuronal and behavioral function of DUF1220 domains, sequences that may be important to higher brain function in humans. As a result, these models should increase our understanding of the molecular mechanisms that underlie certain cognitive diseases such as mental retardation and autism.