ZEBRAFISH GENOMICS INITIATIVE GRANTEES MEETING

Natcher Center, National Institutes of Health

MEETING REPORT

Background
The February 1997 Boston workshop, Current Advances in Defining the Zebrafish Genome, precipitated recommendations promoting the use of zebrafish as a model organism for the study of vertebrate development. Program staff from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Center for Research Resources (NCRR), and National Human Genome Research Institute (NHGRI) reviewed the report. A report was presented to the Director of the NIH, in spring, 1997, with the recommendation to develop the zebrafish system for genetic studies of vertebrate embryogenesis and disease. In response to these recommendations, the Director of the NIH formed the Trans-NIH Zebrafish Coordinating Committee , which first met in the fall of 1997. This working group is composed of representatives from most of NIH's Institutes and Centers having an interest in promoting zebrafish as a research model. The Committee is co-chaired by Dr. Josephine Briggs of NIDDK and Dr. Tyl Hewitt of the National Institute for Child Health and Human Development (NICHD).

The initial effort of this group resulted in 13 Institutes sponsoring a Request for Applications (RFA) entitled, "Genomic Resources for the Zebrafish" (DK-98-006). The purpose of the RFA was to solicit applications aimed at creating resources to facilitate the mapping and positional cloning of genes in the zebrafish, specifically the (1) generation of a genetic map with a resolution of 0.3 cM or better, (2) development of expressed sequence tags (ESTs) from existing and new cDNA libraries featuring specific developmental time points and tissues, and (3) creation of a physical or radiation hybrid (RH) map. Currently, 18 Institutes support the initiative and will participate in the oversight of funded projects. The principal awards were made through NIDDK and will be managed by NIDDK, NICHD, NHGRI, and other participating Institutes through the Trans-NIH Zebrafish Coordinating Committee. Four awards were granted in summer, 1998 and a fifth award was made in fall, 1998.

Meeting Objectives

The meeting, chaired by Dr. Briggs and Dr. Hewitt, brought together the five grantees, representatives from the Trans-NIH Zebrafish Coordinating Committee, informatics experts, and the Chair of the Executive Zebrafish Advisory Committee (EZAC). The meeting objectives included (1) an explanation of the structure of the zebrafish genomics initiative, (2) presentation of short-term and long-term goals for the five funded projects, (3) a review of informatics resources and limitations, and (4) planning and policy discussions.

Structure of Initiative

The zebrafish genomics initiative is currently funded at $4.7 million per year for 3 years. The principal investigators (PIs) comprise the Steering Committee. During the meeting, Dr. Leonard Zon was elected Chair of the Steering Committee. This committee will determine the agenda for annual meetings. Additional meetings may occur by teleconference as needed. An external advisory committee will advise the Trans-NIH Zebrafish Coordinating Committee in setting goals for the zebrafish genome initiative and examining research progress in relation to those goals. Until membership is finalized, Dr. Nancy Hopkins has agreed to chair the committee. The meeting participants were asked to review a tentative list of EZAC members and suggest additional expertise. Dr. Richard Myers was recommended. Tascon, Inc. has been contracted to provide logistical support for meetings and teleconferences.

Dr. Briggs announced that the Trans-NIH Coordinating Committee will sponsor a zebrafish workshop on May 10-13, 1999 at NIH. She suggested that the Steering Committee hold its annual meeting at this time.

Aims of Individual-Funded Projects

Dr. Stephen Johnson, Washington University School of Medicine will use an integrated approach to genomic resource development. EST sequencing will be utilized to develop sequence tagged sites (STSs) that will be used to identify single nucleotide polymorphisms (SNPs) between inbred strains. SNPs will be mapped on RH and meiotic mapping panels. Genetic and physical mapping of 10,000 zebrafish ESTs will provide candidate genes for zebrafish mutations and a dense framework for initiating positional cloning projects. Comparison of zebrafish gene map positions with human homologues will reveal roles for human genes and assist in reconstructing the evolutionary history of the vertebrate genome.

Oligonucleotide hybridization fingerprint analysis of 278,000 independent zebrafish cDNA clones from various tissues and developmental stages will be analyzed to identify 50,000 clusters likely to represent each of the zebrafish genes. Approximately 10,000 STS markers, generated from 3' ESTs, will be used in collaboration with Dr. Zon, to generate a 7,500-marker RH map. Sequencing and map data will be updated weekly on a publicly accessible website.

As the mapping strains SJD and C32 have residual polymorphism (less than 5%), they will be bred to complete homozygosity. Polymerase chain reaction (PCR)-based methods will be used to identify individuals, homozygous at residual polymorphic loci, for propagation. As the SJD and C32 strains are male-biased (typically less than 5% female), ingression of dominant female-promoting genes will be achieved by serial backcrossing.

Discussion focused on two topics -- stock centers and avoiding redundancy in mapping efforts. To avoid mapping the same genes repeatedly, publicly accessible websites will be monitored regularly, particularly between Dr. Johnson and Dr. William Talbot. All the PIs agreed that maintaining large fish stocks, due to imbalanced sex ratios, is problematic. Issues include the demand for increased physical space for fish tanks and the possibility of strain contamination. The need for a stock center that maintains test strains was emphasized.

Dr. Leonard Zon, Children's Hospital Corporation, Boston, will utilize amplified fragment length polymorphism (AFLP) technology to define genetic markers within the zebrafish genome. A RH map will be developed by (1) comparing four RH panels for retention frequency and resolution, (2) constructing an anchored framework map of microsatellite markers and cloned cDNAs on the RH panel with the most appropriate resolution, and (3) positioning 5,000 to 10,000 EST markers on the selected RH panel. Data will be posted on the Zebrafish Information Network (ZFIN) at regular intervals.

Currently, two RH panels are available -- the Ekker panel and Goodfellow panel (named after the investigators). Characterization of the panels involves the following criteria: (1) retention rate -- the percentage of zebrafish genome in each hybrid, (2) connectivity -- the ability to detect linkage between distant markers, (3) resolution -- the ability to determine the order of nearby loci, and (4) typical fragment size or average breakpoint frequency. The two panels have similar retention rates, but the Goodfellow panel produces a larger number of smaller fragments (approximately 66 kb), while the Ekker panel produces a smaller number of larger fragments (100 kb to 150 kb). Resolution and connectivity comparisons are not yet completed.

RH and genetic maps yield similar results, but the genetic map is more accurate. The advantage of RH mapping is the ability to position many markers rapidly. Remaining RH mapping issues for this project concern choosing a panel for EST mapping and information technology coordination. The latter issue involves choosing a software program for analysis (RH Mapper or SA Mapper), data storage format, and Internet publishing format. Selection of a RH panel for EST mapping will be based on resolution, connectivity, and the cost and availability of the panel. An advisory panel will help in the selection process.

Dr. Mark Fishman, Massachusetts General Hospital (MGH), plans to increase the density of microsatellite markers on the zebrafish genetic map by achieving an average intermarker distance of 0.3 cM. Currently, a low-density map exists, consisting of 705 markers covering all 25 zebrafish chromosomes at an average interval of 5 cM. Microsatellite markers are based on simple sequence length polymorphism (SSLPs) repeats found throughout the genome. These markers are the standard starting point for positional cloning and serve as anchors for other types of maps.

Microsatellite markers present several advantages over other types of markers, including (1) their abundance and relatively uniform distribution throughout the genome, (2) polymorphism between outbred strains, (3) their reliability and ease in use, which makes them ideal tools for small laboratories or high-throughput facilities, (4) a high incidence of codominance permits tracking of all four alleles, (5) rapid progress in positional cloning projects as they are single locus amplification products. The latter characteristic also facilitates anchoring a RH-based EST map to the genetic map.

Data will be posted on the local MGH website (http://zebrafish.mgh.harvard.edu ), ZFIN (http://zfin.org/ZFIN/ ), and the National Center for Biotechnology Information (NCBI) database for STSs (http://www.ncbi.nlm.nih.gov/ ). Data information will vary depending on the database, but will include marker name, primer sequence, PCR conditions, and product sizes in multiple strains.

Subsequent discussion questioned the optimal number of markers, based on cost-effectiveness. Participants agreed with Dr. Fishman's approach to achieve an average intermarker distance of 0.3 cM as this resolution facilitates chromosome walks. The added expense for increased resolution produces minimal research gains. Instead, individual investigators with a mutation or tag between two markers can expand the local map. Dr. Zon suggested that a central broker, such as Dr. John Postlethwait at the University of Oregon, notify investigators mapping the same region or bin. This process would expedite determining the order of local markers. The participants agreed that incentives are needed to encourage submission of prepublished data to centralized databases.

Dr. William Talbot, Skirball Institute, New York University Medical Center, will meiotically localize 3,000 zebrafish genes on an integrated genetic map to create a resource that will enhance the comparative and functional analysis of the vertebrate genome. The genes will be mapped by scoring single-strand conformational polymorphisms (SSCPs) in 3' untranslated regions (UTRs) and other nonconserved regions. As genes are uniquely suited markers for comparative genomics, the project will allow construction of a map of chromosomal segments conserved between zebrafish and human. This will accelerate gene mapping in zebrafish by defining boundaries of conserved segments and will facilitate comparison with the gene-rich maps of mouse and human.

A framework map will be constructed by assigning 500 publicly available CA repeat markers in a homozygous diploid (HD) mapping panel. Genes will be mapped to this panel and the data made available to the zebrafish community. SSLP markers in the framework map will serve as a standard of comparison to physical maps and genetic maps made with different mapping panels (e.g., mutation mapping crosses). The HD mapping panel is expected to achieve a resolution of 2 cM on the genetic map. A World Wide Web interface will allow rapid public access to map information and streamline data management.

Discussion focused on DNA libraries. Dr. Marco Marra explained that libraries are not equivalent. Libraries physically suitable for hybridization may not perform well at the sequencing stage. Additionally, cell types may be inappropriate for particular tasks. For example, lambda-based cDNA libraries are not well suited for high-throughput sequencing efforts. Technologies used with libraries must also be considered. Fingerprinting produces single representatives of a particular gene and eliminates redundancy, but focuses on a small subset of libraries. EST-based techniques support more complex libraries, but target fewer genes.

Dr. Marnie Halpern, Carnegie Institute of Washington, will produce a comprehensive deletion panel for the zebrafish genome. The panel will integrate locations of specific deficiencies with the evolving zebrafish genetic map. The project will involve collecting, preserving, and cataloging existing deficiencies and translocation strains. Strains will be maintained by freezing sperm and storing samples in a centralized repository. The status and properties of each deficiency strain will be recorded in a database. Additionally, recently isolated and newly gamma-ray-induced deficiency strains will be recovered and characterized.

Deficiency genome DNA samples will be typed using mapping primers derived from cloned genes, ESTs, or anonymous DNA markers (e.g., SSLPs). High-quality DNA samples from mutant embryos, preselected for the loss of specific chromosomal regions, will be arrayed in multiwell plates for high-throughput mapping and distribution to the zebrafish community. Lastly, expressed sequences will be correlated to mutant phenotypes. Preliminary descriptions and video images of mutant phenotypes corresponding to each deficiency will be stored in an online database. This will enable researchers to immediately correlate deleted gene sequences with a candidate mutant phenotype.

Requested deficiency mutation strains will be thawed for in vitro fertilization of wildtype eggs and shipped as live embryos. An estimated 200 embryos will be recoverable from each deletion strain, providing enough DNA for 500 PCR reactions per embryo. Aliquots sufficient for 1,000 PCR reactions will be provided upon request.

Dr. Fishman stated that his team could help select microsatellite markers to be used as anchors for the genetic map. Ideally, the anchors should be evenly distributed, with a marker every 5 cM. Crude mapping would require approximately 100 markers, with more refined projects requiring about 500 markers. The markers selected as anchors could be listed on ZFIN and refined over time. His suggestion was well received.

Informatics

Dr. Monte Westerfield presented information about ZFIN. ZFIN began after the 1994 Cold Spring Harbor meeting on zebrafish development and genetics. At that time, meeting participants recognized that much of the generated zebrafish data would not be published in traditional journals, and a centralized database would provide access to that data. The ZFIN project, originally funded by a National Science Foundation starter grant, is now funded by NCRR.

ZFIN offers a broad range of information, such as an annotated atlas of embryonic and larval anatomy, meiotic and RH maps, descriptions of wildtype and mutant strains, links to related websites, commercial sources for reagents, and publications on research methods and technologies. Currently, ZFIN contains descriptions of approximately 2000 mutants, with images for 85% of the mutants. A simple dictionary of 30 to 40 terms allows a search for mutants based on five criteria. Future plans call for an expanded dictionary to include standardized descriptors for anatomical structures, physiological processes, and developmental stages. Development of interfaces for user-friendly access to data is also planned. A September 1998 meeting at the University of Oregon on zebrafish mapping bioinformatics discussed the status and future plans for ZFIN. A summary of that meeting was provided to the participants of the Grantees Meeting.

Discussion centered on the compatibility of mapping software, the compatibility of the different types of maps, and the expeditious availability of new data. As mapping laboratories need to develop and maintain their own in-house databases for organizing and curating locally generated map data, Dr. Westerfield is concerned that each laboratory will develop its own specialized software. He recommended that mapping laboratories coordinate their local software resources. A common software program (e.g., RH Mapper) would facilitate data sharing and ease the work involved in exporting data to ZFIN.

The second issue discussed at the meeting involves integrating data from different mapping panels to allow direct comparisons of different maps. Dr. Westerfield explained that a fully automated computer generation of a single, definitive, integrated map is not practical due to ambiguities and conflicts inherent in data from different mapping experiments. Producing such a map would require extensive curation by a chromosome committee. Dr. Fishman suggested the use of an anchored set of maps. While an anchored map would not provide perfect colinear alignment, it would ease comparison between different maps. On a related topic, participants discussed cross-species genomic analysis. Dr. Westerfield explained that future plans call for ZFIN links to records of orthologous genes in other species databases and links to external databases (e.g., GenBank). A mechanism for displaying and comparing syntenic relationships between zebrafish and other organisms is also planned.

Participants agreed that new data should be made publicly available when it is generated. Dr. Westerfield suggested that unpublished data (e.g., mapped genes and mutations) be held for a waiting period after submission to ZFIN, but before journal publication. The waiting period would allow researchers to register and reserve mutant and gene names before publication. Additionally, journals should be encouraged to require ZFIN accession numbers for genes, mutations, and map markers by the time of publication. This requirement would provide incentive for researchers to submit their data to common databases.

Planning and Policy Discussion

Data Sharing and Release Policies

The investigators agreed that sharing sequence data is not problematic; standards exist for other projects, and data can be updated on a timely basis. Mapping data is problematic as it must be error-checked in a time-consuming manner. Dr. Halpern stated that deletion data could be promptly posted on ZFIN if researchers understand the considerable lag time in obtaining live embryos. Dr. Briggs requested that the five grantees draft a written policy for data release.

Dissemination of Proposed Data and Resource-Sharing Plans

The participants discussed mechanisms to encourage collaboration within the zebrafish community and move the genome project forward. The following suggestions were made: (1) list the names of the grantees and contact information on ZFIN, (2) on ZFIN note how often updates will be posted, (3) allow investigators to anonymously post map locations of unpublished genes and mutations for a 6-month period prior to journal publication; this would allow researchers to establish collaborations with one another while protecting their data before publication, (4) make a large-scale mapping effort part of the initiative project so that 100 to 200 mutations could be mapped quickly and made available, (5) submit a letter or brief article to Science or Nature describing the dissemination process and spirit of collaboration.

Development and Oversight of Database

All the grantees are committed to using ZFIN as a centralized resource. Dr. Westerfield reiterated the recommendations made at the September 1998 Zebrafish Mapping Bioinformatics meetingC(1) only mapped ESTs should be stored in ZFIN with a link to access a complete listing in NCBI's dbEST, (2) storing primer sequences in ZFIN would be useful, even though this may duplicate data stored elsewhere, (3) information about allele size should be included for each strain in which a marker has been mapped, (4) information about the quality of each marker should be added, e.g., the percent of successful amplification to indicate reliability, (5) a list of linked markers to each mutant should be added, (6) information on deficiency chromosomes should be added including the mutant phenotype, linkage group, missing markers, and flanking markers that are present.

International Relations

Several participants expressed the need to offer funds for additional staffing of the stock center in Tuebingen, Germany. This center provides fish globally, but is understaffed. Dr. David Badman remarked that Dr. Harold Varmus, NIH Director, is interested in this issue. Funds may be available through the Fogerty International Center. On a related topic, the grantees voiced concern on the restrictive nature of transfer agreements with the Max Planck Institute. Dr. Briggs stated that these issues would be pursued.

Future Resource Needs

The grantees were asked to list and prioritize resources that would augment their current projects should additional funds become available. The investigators agreed to the following priorities: (1) informatics -- increase support for development and eight additional personnel, (2) EST sequencing -- increase the goal of 50,000 sequences to 100,000 and represent organogenesis through 36 hours of development, (3) EST mapping -- increase the number of mapped ESTs from 10,000 to 25,000 over 3 years, (4) physical map -- add support for fingerprinted contig maps, (5) new equipment, such as improved devices for genotyping and robotics for 384-well PCR machines, (6) facilitatory projects such as development of better vectors for transgenics, promoter identification, marked chromosomes, and targeted mutation.

Participants

Dr. David Badman National Institute of Diabetes and Digestive and Kidney Diseases 301 594-7717 David_Badman@nih.gov	Dr. Josephine Briggs, Co-Chair National Institute of Diabetes and Digestive and Kidney Diseases 301 496-6325 briggsj@extra.niddk.nih.gov
Dr. Jill Carrington National Center for Research Resources 301 435-0776 jillc@ep.ncrr.nih.gov	Dr. Hemin Chin National Institute of Mental Health 301 443-1706 hemin@codon.nih.gov
Dr. Mark Ekker Ottawa Civic Hospital 613 798-5555 ext 6033 ekker@civich.ottawa.on.ca	Dr. Elise Feingold National Human Genome Research Institute 301 496-7531 feingole@odder.nhgri.nih.gov
Dr. Adam Felsenfeld National Human Genome Research Institute 301 496-7531 adam_felsenfeld@nih.gov	Dr. Mark Fishman, Grantee Massachusetts General Hospital 617 726-3738 fishman@cvrc.mgh.harvard.edu
Dr. Andreas Fritz Department of Biology Emory University 404 727-9012 afritz@biology.emory.edu	Dr. Maria Giovanni National Eye Institute 301 496-0484 myg@nei.nih.gov
Dr. Judith Greenberg National Institute of General Medical Sciences 301 594-0943 greenbej@nigms.nih.gov	Dr. Marnie Halpern, Grantee Carnegie Institute of Washington 410 554-1218 halpern@mail1.ciwemb.edu
Dr. Deborah Henken National Institute of Child Health and Human Development 301 496-5541 dh50g@nih.gov	Dr. Tyl Hewitt, Co-Chair National Institute of Child Health and Human Development 301 496-5541 hewittt@hd01.nichd.nih.gov
Dr. Nancy Hopkins Massachusetts Institute of Technology 617 253-6414 nhopkins@mit.edu	Dr. Don Jackson Massachusetts General Hospital 617 726-6902 jackson@cvrc-taco.mgh.harvard.edu
Dr. Steve Johnson Washington University 314-362-0362 sjohnson@genetics.wustl.edu	Dr. Robert Karp National Institute of Alcohol Abuse and Alcoholism 301 443-2239 rkarp@willco.niaaa.nih.gov
Dr. Carol Letendre National Heart, Lung, and Blood Institute 301 435-0080 letendrc@gwgate.nhlbi.nih.gov	Dr. John McPherson Washington University 314 286-1848 jmcphers@watson.wustl.edu
Dr. Grace Shen National Cancer Institute 301 435-5226 gs35r@nih.gov	Dr. William Talbot, Grantee New York University Medical Center 212 263-1932 talbot@saturn.med.nyu.edu
Dr. Monte Westerfield University of Oregon 541-346-4607 monte@uoneuro.uoregon.edu	Dr. Leonard Zon, Grantee Children's Hospital Corporation, Boston 617 355-7707 zon@rascal.med.harvard.edu