Trans-NIH Mouse Initiative
Reports and Publications


Report from the Workshop on Genomic and Genetic Tools for the Zebrafish
May 10-11, 1999

Executive Committee:
Igor Dawid (Chair)     Nancy Hopkins
Geoffery Duyk     John Postlethwaite
Mark Fishman     Bernard Thisse
Pascal Hafter     Monte Westerfield
Marnie Halpern     Leonard Zon

Preface: The NIH Planning ProcessAdvocate for Zebrafish Research

Workshop Participants' Report

  1. Introduction: The Zebrafish in the Context of the Human Genome Project
  2. Summary of Current Genomic Resources and Activities
  3. Genomic Infrastructure Needs
    1. Mapping tools, ESTs and cDNAs
    2. Physical map construction
    3. Sequencing the zebrafish genome
    4. Budget estimates
  4. Technologies to Advance the Analysis of Gene Function
    1. Genetic screens
    2. Mapping of existing and anticipated mutants
    3. Technologies for genome manipulation
    4. Gene expression
    5. Stock center-related issues
    6. Budget estimates
  5. Informatics Issues
    1. Anatomical Atlases
    2. Budget estimates

Special Applications of the Zebrafish Model

  1. Studies of Musculoskeletal and Skin Developmental Diseases
  2. Zebrafish and Cancer Research
  3. Eye Development and Disease
  4. Kidney Development and Disease
  5. Cardiovascular and Other Organogenesis in Zebrafish
  6. Research on Aging
  7. Child Health and Human Development
  8. Hematopoiesis
  9. Research on Ear Morphogenesis and Hearing Disorders
  10. Developmental Toxicology in Zebrafish
  11. Genome Research
  12. Neurogenesis and Neurological Disorders
  13. Early Developmental Programs in the Zebrafish
  14. Research on Allergies and Infectious Diseases
  15. Mutants that Affect Mental Behavior
  16. Craniofacial and Dental Research

Up to Top

The NIH PLANNING PROCESS for Zebrafish Research

The following report is from the Workshop, "Genomic and Genetic Tools for the Zebrafish", held May 10-11 at Lister Hill Auditorium on the NIH campus. The meeting was under the sponsorship of the Trans-NIH Zebrafish Coordinating Committee (ZFCC), which is comprised of representatives from eighteen of the NIH’s Institutes and Centers. The morning of the first day was devoted to ‘state of the art’ talks by leading zebrafish investigators on genetic and genomic issues followed in the afternoon by three break-out meetings to discuss genomic infrastructure needs, technologies related to the functional analysis of gene expression, and related informatics issues. The second day was devoted to a plenary session to hear and debate the break-out session recommendations. The report presented here is a summary of these deliberations, prepared by the Workshop Planning Committee and Break-out Group Chairs.

This was the third in a series of NIH-sponsored activities to address issues surrounding the zebrafish as a model for development and disease research, particularly the issue of genomics. It was structured to build upon the previous planning activities. The first of these meetings "Current Advances in Defining the Zebrafish Genome", was held in Boston in February 1997. The report of that workshop contained a set of recommendations that provided the basis for a Request for Applications resulting in the funding of the five current zebrafish genomics projects. These projects were initiated in 1998 with cooperative funding from thirteen institutes, and are scheduled to run three years. They are overseen by the ZFCC, with the advice of an external oversight committee. Also in response to the 1997 report, a Program Announcement was published which highlighted the goals and interests of eighteen participating Institutes and Centers in the use of the zebrafish model.

In February 1999, Dr. Varmus convened the Workshop on Non-Mammalian Models, the second meeting at which planning for the zebrafish was addressed. The zebrafish was one of the model organisms considered in depth at that meeting. A panel of zebrafish investigators, chaired by Leonard Zon and John Postlethwaite formulated a report outlining the needs for genetic and genomic tools for the model. This report was used as a basis for planning the May meeting. The recommendations in the present report are quite similar to those developed in February, although they amplify and build upon the recommendations of the earlier panel in important ways.

Also included in the present report are a series of brief essays that outline the power of this model organism to address important questions in specific areas of biology and medicine. Zebrafish investigators prepared these essays after the May meeting through the organizational efforts of Leonard Zon. The ZFCC believes these excellent, short reports may be useful to the NIH Institutes in clarifying the utility of the zebrafish model for their own individual goals.

Up to Top


I. Introduction: Zebrafish in the Context of the Human Genome Project

Vertebrate model organisms complement the human genome project in important ways. As the sequencing of the human genome nears completion, assigning functions to large numbers of sequences has taken on increased importance. Linkage analysis has proved relatively ineffectual in analyzing complex diseases with small contributions from individual genes. Association studies have the power to find even small contributions of a gene, but require candidate genes. Large-scale screens in model organisms provide the perfect next step to identify candidate genes for human diseases and will help focus the Human Gene Project (HGP).

Phenotype-based genetic screens begin with function. Large-scale phenotypic screens can define genes as critical to developmental, physiological, or pathological processes because they identify genes in the living organism. If facile enough and cost-effective enough, they can be driven to saturation, isolating all genes informative with regard to any given process. They are limited in such circumstances only by the ingenuity of investigators. Although yeast, flies, and worm genetics meet these criteria, they do not speak to many important developmental decisions that characterize the vertebrate, including man. For example, there is no logical framework, and only few genes known, for patterns of organ assembly, function, and disease, and it is difficult to approach genetics of complex behavior.

Zebrafish are vertebrates, and tractable and inexpensive enough for screens in many labs. Although many important processes may be extrapolated from invertebrate organisms, including cell fate assignment, many others, such as organ form and function, are new to vertebrates, and the genes involved not known. Within vertebrates, however, these processes have been well conserved. Therefore, the zebrafish is an appropriate and relevant organism for screens applied to these vertebrate-specific questions. Efficient, saturation phenotypic screens are possible in zebrafish because the embryo does not develop hidden inside the mother, as in mammalian embryos, and because the embryos themselves are translucent allowing the direct visualization of individual developing cells. Mutations are currently available in nearly 1000 genes essential for embryonic development, and directed screens are adding more. Methods to store sperm facilitate the storage of mutant strains, and methods to generate haploid and homozygous diploid embryos expedite phenotypic screens. American and European stock centers for the distribution of mutants and a centralized database for genetic and other types of information have been established.

The zebrafish system, therefore, will open new fields of biology and medicine

Beyond the obvious assets in exploring vertebrate early developmental processes, genetic screening in zebrafish provides access to these genes for other vertebrate-specific processes, including:

  • Organ system (integrative) biology and physiology
  • Complex diseases
  • Behavior
  • Vertebrate evolution

Rapid gene discovery depends upon rapid cloning of mutations.

For gene discovery to be relevant to other organisms, and to permit biochemical characterization of pathways involved, these mutant genes must be cloned. If the zebrafish really is to provide an important biological complement to the HGP, this must be accomplished in a rapid, efficient, and cost effective way.

Zebrafish mutations can be cloned. Existing maps and BAC and YAC libraries have sufficed for cloning more than two dozen mutations from chemical mutagenesis screens, identifying homologs of known genes in addition to novel genes. The recent successes of viral insertion and gene trap methods show promise in making rapid cloning of tagged sites feasible, an advance that will make this strategy a highly attractive approach to genetic screening. In addition, it may prove feasible to produce a library that could be screened for insertions into any given gene. Thus, two complementary routes exist to link genes and functions. One begins with phenotype, most effectively induced by chemical mutagenesis, and uses positional cloning methods to identify the gene. The second, still under evaluation for frequency of mutagenesis, begins with a defined gene of interest, identified as tagged with an insertion, and seeks the mutant phenotype.

The first set of recommendations from this workshop focuses upon infrastructure needs to go from phenotype to gene -- i.e., to expedite positional or candidate gene cloning. These are detailed in the chapter entitled "Genomic Infrastructure Needs." The second group of recommendations focuses on the need for further targeted strategies to identify specific genes and understand their function, and are described in the chapter entitled, "Technologies to Advance the Analysis of Gene Function." Lastly, recommendations on "Informatics Needs" are described.

Up to Top

II. Summary of Current Genomic Resources and Activities

Mapping resources: Extensive mapping efforts by individual laboratories have produced a genetic map, currently anchored with over 2000 independent simple sequence length polymorphisms (SSLP) and 600 random amplified fragment length polymorphisms (RAPDs). Over 400 genes have been placed on the map using restriction fragment length polymorphisms or single strand conformational polymorphisms. Through the recently funded NIH initiative, over the next two years we anticipate that enough micro-satellite markers will be placed on the map to yield a one cM average interval (being done by Mark Fishman’s laboratory). Additional mutant mapping can be accomplished by establishing linkage with micro-satellite markers. An important advantage of this organism is the ability to perform half tetrad analysis, which makes use of the ability to produce gynogenetic diploid fish. When the high-density map becomes available, this will provide a quick method for establishing mutant gene localization.

Radiation hybrid panels: Two independent zebrafish-mammalian radiation hybrid panels have been developed for mapping of genes or EST sequences. One panel was made by Peter Goodfellow and another by Marc Ekker. These two panels have been typed with 1000 to 2000 SSLP markers, and both have good retention of zebrafish DNA and high resolution. RH panel mapping under the current initiative will be performed in the laboratories of Leonard Zon and of Steve Johnson. The panels are extremely useful for positional cloning approaches and can be utilized to establish the location of thousands of genes or ESTs. These reagents are particularly useful for establishing whether a candidate gene or EST is represented by a zebrafish mutant. In addition to the radiation hybrid panel mapping, there is another zebrafish genome initiative project that will position 3000 independent genes on the meiotic map (Wil Talbot’s lab). Some of these genes also will be also placed on the radiation hybrid panel, anchoring the RH map to the genetic map and establishing syntenic relationships between zebrafish and humans.

Genomic DNA libraries: With regard to genomic DNA libraries, there are two YAC libraries for the zebrafish (Mark Fishman’s and Leonard Zon’s laboratories). These are useful for positional cloning or chromosomal walking. A PAC and a BAC library are available from commercial resources. Currently, Cclone inserts averages are currently 100kb; in the future, it would be helpful to have a larger insert genomic PAC or BAC libraries. This will be critical for the generation of a physical map of the genome. The NIH has funded an EST sequencing project for 50,000 independent ESTs (Steve Johnson’s lab). These ESTs are derived from oligo-fingerprinted libraries available in the community and should produce an excellent resource for candidate genes.

Other technologies: Another funded project is to develop deletion mutants representing the entire genome (Marnie Halpern’s laboratory). This will be useful to establish the null phenotype for mutant genes and will also evaluate gene function. Apart from those projects supported by the zebrafish genomics initiative, other important resources and technologies are being developed. Transgenic technology for the zebrafish has been mainly developed by Shuo Lin at the Medical College of Georgia The laboratory of Shuo Lin at the Medical College of Georgia was the first to show consistent developmental regulation of transgenesis in zebrafish and continues to develop enhancements such as the technology to express green fluorescent protein (GFP) in various tissues. Insertional mutagenesis has been developed by Nancy Hopkins’ laboratory, and recent exciting findings regarding gene trap events suggest they will yield an excellent resource for genomics and genetics of the zebrafish.

In summary, over the past several years, the zebrafish community has amassed a significant set of reagents and resources to enhance the genetics and genomics of the system. The community has created many of the tools necessary for positional and candidate cloning of mutant genes, thus establishing the basic infrastructure necessary to exploit the genetic power of this model organism.

Up to Top

III. Genomic Infrastructure Needs

  1. A. Mapping tools, ESTs and cDNAS

SNPs: Single nucleotide polymorphism maps are needed as the next and more efficient mapping tool, as they have been in human genetics. New technologies based on SNP markers offer the potential for rapid, automated genotyping. The EST projects will provide a large pool of sequences that are already paid for and can be used to identify SNPs. Although the frequency of SNPs will be lower within ORFs than in non-coding regions, it will carry the added benefit of targeting known and/or candidate genes.

ESTs and cDNA libraries: ESTs provide candidate genes, useful in positional cloning (during walks and for recognizing ORFs) and for ORF recognition in cloning of insertion sites. New cDNA libraries are needed as EST sources, and for BAC hybridization during positional cloning. These libraries should be from different representative stages and organs, normalized and gridded, and constructed in appropriate vectors for sequencing and, of course, made publicly available.

From these libraries, an additional 100,000 ESTs should be sequenced (divided between 5'- and 3'-end reads). These could include the sequencing of insertion sites from retroviruses.

RH panel mapping of ESTs is critical to correlate candidate genes with mutations at the highest rate, and to examine conserved synteny with other model organisms and with the human. An additional 30,000 ESTs should be mapped on RH panels, in addition to the 10,000 currently supported.

Up to Top

B. Physical map construction

A physical map is critical crucial to expedite positional cloning, and may provide a critical step towards sequencing the genome. Having such a map will expedite walks, and thus reduce reducing the costs and increase increasing the speed with which individual laboratories can clone their mutations of interest. There are several ways to assemble all of the DNA fragments into contigs. The current standard is fingerprinting by enzymatic digestion. There will be gaps left after fingerprinting. These are best closed (as will be needed to sequence the genome) by end-sequencing of BACs, themselves anchored to the genetic map.

Currently, because of the YAC cloning techniques utilized, telomeres are under-represented, and telomeric YAC libraries need to be constructed. Similarly, current BAC coverage is insufficient, so new BAC libraries need to be constructed, with 200 kb inserts.

The key needs here are:

    • High quality BAC library.
    • BAC fingerprinting
    • BAC anchoring by EST to genetic maps.
    • Half YAC (telomere) library construction.

Up to Top

C. Sequencing the zebrafish genome

Because the zebrafish is the only vertebrate now amenable to large-scale saturation genetic screens, it can provide the single strongest functional underpinning to the Human Genome Project. Relating the zebrafish and human genomes will facilitate gene discovery and reveal connections to human biology, but this requires that zebrafish mutations be cloned rapidly. This is the most important reason to sequence the genome. The cost saving to the many small laboratories involved would more than outweigh the expense of the project. Just reducing the cost of sequencing BACs for the several hundred genes that will be cloned by walking over the next few years will go along way to offset the cost of the project. Obviously, the accomplishment of genome sequencing depends upon, but dwarfs, the other infrastructure projects, and is more properly compared only with efforts now underway to sequence the human and mouse genomes. The value of both of those projects will be greatly enhanced by the international zebrafish program for functional gene discovery.

The recommendation of this workshop is that immediate steps should be taken to sequence enough of the genome to establish the quality of the libraries and maps, and to acquire insight into gene structure and repeats, both of which would aid current mutation cloning efforts. These goals could be achieved by sequencing approximately 5% of the zebrafish genome. There are two options for such a project: either focus on sequences scattered through the genome (designated by investigator interest) or attempt to identify one long contiguous piece, preferably an entire chromosome. The recommendation of the breakout group on genomics at this workshop was for the latter option, but it was recognized that this issue should be the topic of further discussion. It was also recommended that the quality for current purposes need not be "fully polished".

Recommendations for sequencing:

    • BAC end sequencing to close contig gaps.
    • Sequence 5% of the genome.

Up to Top

D. Budget estimates

Workshop participants developed the following direct cost estimates:

1. Maps

ESTs: map 30,000 additional on RH panels, over 3 years
[$6,000,000 total (= 30,000 ESTs x $200/EST)]

Single nucleotide polymorphisms (SNPs): map 10,000 SNPs over 3 years [$240,000 total (= [10,000 reads x 4 strains x $5/read] + [$1/pcr x 4 strains x 10,000 ESTs] )]

2. Expressed Sequence Tags and Libraries

ESTs (3 years)
[$1,000,000 (= 100,000 clones x 2 reads/clone x $5 per read)]

Construction of new cDNA libraries (normalized, gridded, many stages and organs, in the best vectors, publicly available) (3 years)
[$450,000 (= $150,000/y x 3 y)]

3. Physical map

High quality BAC libraries of 200 kb inserts (1 year)
[$250,000 (2 libraries@10xcoverage x $125,000/library)]

Fingerprinting BACs. (1 year)
[$1,000,000 (=200,000 clones x $5/clone)]

Anchoring BACs to the genetic map. (2 years)
[$1,000,000 (= 10,000 markers x $100/marker)]

Half-YAC (telomere) library (1 year)

4. Genome sequence

BAC end sequencing (3 years)

Full shotgun sequence 5% of genome (3 years)
[$12,000,000 (=$0.15/base x 80,000,000 bases]


Genome infrastructure:
Year 01= $4,480,000
Total (3 years) = $10,090,000

Sequencing 5% of genome
Year 01 = $5,000,000
Total (3 years) = $15,000,000

Up to Top

IV. Technologies to Advance the Analysis of Gene Function in Zebrafish

Genetic tools. This portion of this report deals with the continued potential of mutagenesis screens and with new technologies that will greatly accelerate the analysis of gene function in normal and mutant phenotypes, and the identification of genes affected in mutants. Expenditure of relatively small amounts of money on the tools described here has the potential to advance the field significantly by facilitating more sophisticated analysis of gene function, sometimes at reduced cost and effort, and to make the field accessible to more laboratories.

Up to Top

A. Genetic screens

The need for continued emphasis on mutagenesis screening and phenotype. The most powerful and unique feature of the zebrafish is that it is a vertebrate model organism in which large-scale forward mutagenesis screens can be performed with relative ease. Participants in this workshop were unanimous in expressing the conviction that work to date had only begun to exploit the power of mutagenesis screening to detect developmental genes and pathways of interest. The ultimate goal should be saturation mutagenesis of the zebrafish genome. The community felt that such screens are best performed in individual laboratories, and that centralized mutagenesis centers offered few advantages.

To date, screens have focused exclusively on phenotypes involved in embryonic development. The potential of phenotypic screens that focused on the adult was emphasized by many participants, with potential possible applications envisaged in a variety of disease areas as well as in response to environmental toxins and drugs.

It was also emphasized that despite decades of research that have proven over and over again the phenomenal power of the genetic screen for opening new and unexpected avenues of research, the approach is often looked upon unfavorably by study sections. Genetic screens, even when carefully focused on defined developmental pathways, may be seen as ‘fishing expeditions’ relative to more tightly hypothesis-driven science. For this reason, although genetic screens are the raison d'etre of the fish field, and receive much favorable lip service, they are difficult to fund via the R01 mechanism. Thus, there was considerable support at the workshop for special mechanisms for that would increasing increase funding for this type of research. Particularly hard to fund is the acquisition of fish tank systems for mutagenesis screens. It is interesting to note that no large-scale genetic screen in the zebrafish has ever been supported entirely or even primarily by NIH funds. With the advent of promising insertional mutagenesis strategies, it is more pressing than ever to develop a strong support base for creative and rigorous mutagenesis screens.

High priority is given to the following:

  • Support of additional genetic screens in individual laboratories focusing on identifying mutations that affect the structure and function of specific tissue/organ systems.
  • Development of phenotypic tools and new genetic screens to analyze the genetic basis of adult phenotypes including behavior, aging, heart disease, cancer, response to environmental toxins, etc.
  • Genetic screens utilizing altered gene expression patterns, as a tool to identify components of genetic pathways.
  • Support of screens using new strategies for insertional mutagenesis.
  • Support for expanding institutional fish facilities and purchasing fish tank systems.

Up to Top

B. Mapping of existing and anticipated mutants

Previously and newly isolated mutations need to be linked to candidate genes. If all chemically induced mutations were placed on the zebrafish genetic map, then as an ever-increasing number of gene sequences become mapped, the probability of cloning mutants by matching them to candidate genes will continue to increase. It was felt that obtaining rough map positions for a large number of mutants was valuable for this reason. This will be an ongoing endeavor, as mutant screens will continue to be a central activity of the fish field.

Up to Top

C. Technologies for genome manipulation

To In order to exploit the power of the zebrafish to clarify gene function, we need to rapidly expand the application of several novel techniques that have been developed over the past five years. As we have learned in other model organisms, a handful of molecular genetic tricks can determine the ease, speed, and cost-effectiveness of genetic studies in each organism. DNA-mediated or transposable element transgenesis, insertional mutagenesis, knockouts, embryonic stem cells, the availability of cell lines, methods for freezing and storing organisms, have all transformed fields in the case of specific model organisms. The zebrafish still lags in this regard.developing and taking advantage of these technologies.

Technologies and methods that could greatly accelerate the pace of analyzing gene function in the fish include the following:

  • Generation of transgenic fish expressing GFP, in specific lineages and on each chromosome, in order to i) facilitate better and more rapid analysis of mutant phenotypes; and ii) facilitate carrying mutant lines of fish; . crossing Crossing mutations into the appropriate line bearing a (linked) GFP-marked chromosome, would permit rapid identification of progeny carrying the mutation.
  • Development of other types of transgenics including the UAS-gal4 system, flp, cre, etc.
  • Full exploitation of existing, or additional, libraries of fish in which gene trap viruses have integrated into large numbers of genes, creating screenable collections of ‘knock-out libraries.’

Up to Top

D. Gene expression

i. Array technology: Array methods for systematic assessment of patterns of gene expression were felt by many participants to be important tools for the zebrafish community.

Specific recommendations:

  • Making new cDNA libraries from specific tissues and at multiple developmental stages.
  • Getting more expressed sequence tags (ESTs) from these libraries.
  • Making these reagents available in a suitable form to use in microarrays of 10,000-20,000 ESTs to begin with. Ultimately the goal is that expression of all cDNAs can be systematically assessed.

ii. Gene expression patterns (Also, see section on Informatics Issues): There was substantial support for the development of systematic information about spatial patterns of gene expression. Taking advantage of the transparency of the zebrafish embryo, high-resolution in situ hybridization can be performed on whole mount embryos, using probes for known genes and for ESTs.

The goal of this project would be:

  • Providing a large collection of cell and/or tissue specific molecular markers for analysis of wild-type development (establishment of a "molecular anatomy" of the embryo) as well as for the phenotypic description of mutants obtained from genetic screens.
  • Providing starting points for functional analysis (using overexpression studies as well as the development of numerous Gal4-UAS lines).
  • Providing a powerful way (candidate gene approach) to clone genes disrupted in mutant embryos based on the correlation between mutant phenotypes, expression pattern and data concerning sequence and map position of these genes.

All these expression data should be released to the public via the Zebrafish Information Network [ZFIN] in an integrated database (see anatomical atlases, below) in a standardized format allowing the search by names, sequences, genetic map positions, and anatomical structures. The database will include keyword descriptions of all spatially restricted expression patterns as well as the corresponding pictures for all developmental stages analyzed.

Up to Top

E. Stock Center and Training -- Related Issues

The U. S. Stock Center is a major concern of the zebrafish research community. There are a large number of mutant lines currently spread among many laboratories. Moreover, many laboratories are generating new mutant and transgenic lines at an increasing pace. The community is concerned about whether the this Stock Center can adequately house all these lines safely. The current plan for the Stock Center calls for space to maintain a maximum of about 5,000 lines in tanks; the remainder will be kept as frozen sperm. The mutant collection in Germany will not be expanded, meaning 400 or fewer lines will be maintained there. To In order to ensure that all lines are available, it is suggested that the Stock Center concentrate on providing lines quickly and efficiently from frozen sperm. This will require additional personnel.

For many laboratories, in vitro fertilization and recovery of mutant lines is difficult. The Stock Center should provide these services. Additional staff will also be required to meet these goals because both services are more labor intensive than simply maintaining lines.

The community felt that putting all mutants in one Stock Center, without some backup was not a good idea. Funding should be provided for banking sperm at a second site that would serve as a safety backup. This would require money for liquid nitrogen freezers and at least a part time salary.

Concern was expressed that the number of individuals with expertise in fish pathology is very limited. Expanded training in this field was a recommendation of this workshop.

Specific recommendations:

  • Expanded support for existing stock center for providing mutant fish from frozen sperm and in providing training opportunities in techniques needed to recover mutant lines.
  • Support for an additional back-up center in the U.S. for sperm preservation to prevent catastrophic loss of mutants.
  • Expanded training of individuals with expertise in fish pathology.

Up to Top

F. Budget estimates

Workshop participants developed the following direct cost estimates:

1. Genetic screens (including adult phenotypes) $3.0 M/yr

2. Technologies for genome manipulation Total/yr: $3.5 Million

  • Methods improvement and novel developments: $1.5 M/yr
  • GFP-marked chromosomes: $1 M/yr
  • KO-viral insert approach: $1 M/yr

3. Stock centers $ 0.35 M/yr

4. Array technology $1 M/yr

5. Mapping mutants $0.5 M/yr

It is anticipated that each of the above initiatives will be funded for 3 years.


Year 01 = $8,350,000
Total (3 years) = $25,050,000

Up to Top

V. Informatics Issues

With the increasing complexity of biological, sequence and mapping data being generated, the need to develop and maintain a community-based database becomes paramount. The recommendations in this chapter clearly impact on the research activities described in the preceding two chapters since the integration of information at a common site will provide a valuable resource to the zebrafish community. This important role is to be fulfilled by ZFIN, currently funded through a grant to the University of Oregon. However, the cost of maintaining such a sophisticated database is high and one that appropriately should be shared by Institutes and organizations supporting research using zebrafish. It is a complicated issue and the recommendations cited below look at both long-term needs of development and maintenance of the system as well as the more short-term need for developing an anatomical atlas and dictionary.

The utility of genomic, mapping and biological information to the community of investigators interested in the zebrafish is heavily dependent upon the existence of mechanisms that facilitate access to this information. The sharing of materials and data in a timely manner is an essential element if rapid progress is to be made in the construction and use of genetic maps. The personnel who develop and maintain ZFIN are crucial to its success in implementing the goals of the zebrafish community. The data editors who annotate expression and phenotypic data and the informaticists who are needed to expand the capabilities of the website are essential to ensure that ZFIN is able to fulfill its mission of providing the entire community with ready access to genomic and developmental data.

One of the important needs is the development of standard terminology and staging criteria as well as standardized formats that will allow for the rapid transfer and submission of sequence, mapping, and biological data to ZFIN. This will also be important for the integration of the zebrafish database with those of other model organisms. Community standards for data acquisition as well as rules regarding intellectual properties and full disclosure of potential restrictions on the use of data need to be developed. Many of these activities will be covered in the costs associated with the maintenance of the informatics infrastructure indicated below. Recommendations include:

    • Data editors to annotate expression and phenotypic data
    • Bioinformaticists to expand capabilities of database
    • Develop "data model" for minimal criteria for submission and acceptance of data to ZFIN
    • Establish community standards for access and use of database
    • Interface ZFIN with stock center database so that requests for orders can be input through ZFIN
    • Define terminology and staging criteria for the establishment of a controlled vocabulary and anatomical dictionary (see below)
    • Integrate ZFIN with the databases of other organisms
    • Copy ZFIN so that private contractors can’t resell database information

Up to Top

A. Anatomical atlases

Anatomical atlases will be important for the rapid development of the fish field. Because the fish has only recently become a popular model organism, anatomical descriptions of the development of structures from embryo to adulthood are incomplete. For this reason, the zebrafish community needs, with some urgency, to build an anatomical atlas and establish an anatomical dictionary. Development of uniform terminology is a critical first step in the larger problem of providing systematic annotated information about gene expression.

The specific goals of this project are to:

    • Establishment of a common language so that the same key words are used by the whole zebrafish community. Thus, we need to build a zebrafish anatomical dictionary.
    • Development of a set of images that illustrate these terms throughout zebrafish development, and make these available through ZFIN.
    • Use of this dictionary to assess and annotate in a systematic fashion the gene expression patterns coming from large-scale in situ hybridization screens and phenotypic data coming from mutagenesis screens, including that available in the literature.
    • Making annotated high quality anatomic images of gene expression available in ZFIN. The long-term goal would be a four-dimensional — space and time - description of the pattern of expression of key genes.
    • Making compatible links between the zebrafish anatomic database information and other databases.

Up to Top

B. Budget Estimates

1. Informatics infrastructure                 $1.0 M/yr

Data editors          $0.5 M/yr
Bioinformaticists      $0.5 M/yr

2. Gene expression atlas                    $1.5 M/yr

Total Estimate of Direct Costs

Year 01              $2,500,000
Total (3 Years)         $7,500,000


Up to Top



Studies of Musculoskeletal and Skin Developmental Diseases

Stephen Johnson
Washington University School of Medicine
Sharon Amacher
University of Oregon

The zebrafish offers a number of advantages for studying the biology of the skin and skeletal/muscle systems. Because embryonic development is external to the mother and the embryos are transparent, development of skin, pigment cells, bones and muscles are easily seen in the living embryo.

One major area of research in zebrafish is development of the melanocyte pigment pattern in zebrafish. Abnormal melanocyte development or survival in humans leads to a variety of syndromes or diseases, including piebaldism, vitiligo, moles, and skin cancer. Studies of development of the adult zebrafish melanocyte stripes reveals that these cells are recruited from a population of undifferentiated precursors or stem cells that persist throughout the life of the fish. Mutations have been identified that affect how the precursor cells are established or later recruited to proliferate and differentiate. Other mutations affect migration of melanocytes, or the distance between stripes. More than one hundred mutations have been identified that affect the development of embryonic melanocytes.

The control of bone growth and regeneration has been investigated in zebrafish. Dozens of mutations affecting jaw cartilage or bone formation have been identified. Mutations affecting the regeneration of the adult fin may help us better understand the genetic control of bone remodeling and healing. For example, when adult fins are amputated, normally post-mitotic differentiated bone cells at the amputation plane are recruited to divide and form the regeneration blastema. Screens for mutations that affect fin regeneration have identified seven temperature-sensitive mutations that affect aspects of the regeneration or morphogenesis of the regenerating bone in the fin. Other mutants have been identified that affect the rate of growth of the fin ray, either causing abnormal morphology of the bone, or bones that grow too slow or too fast. Molecular characterization of these mutations may reveal previously unknown players in these processes shared among all vertebrates.

Important recent work in zebrafish has advanced our understanding of vertebrate muscle development and function. The isolation of a large collection of mutants affecting the development and function of muscles provides an important resource for investigation. These mutants affect a variety of processes including: somite formation (5 genes), horizontal myoseptum formation (11 genes), myoblast differentiation (3 genes), myofiber development and/or organization (8 genes), muscle tissue maintenance (4 genes), and locomotion (>40 genes). Although preliminary characterization has been done on these mutants, further analysis and molecular characterization is important. For example, double mutant combinations must be made to understand gene interactions and to order the genes identified by mutation into a regulatory hierarchy. Furthermore, we must discover the molecular nature of the encoded gene products.

Significant advances have been made in understanding the organization of the segmental plate and the origin of somites in zebrafish. For example, slow and fast muscle precursors become specified very early in development (long before overt muscle differentiation) and then undergo dramatic morphogenetic movements to assume their final positions in the embryo. Indeed, some zebrafish mutations, particularly those that affect horizontal myoseptum formation, may identify components involved in specifying these two muscle cell types.

A wide variety of molecular markers that identify different regions of the developing somites and segmental plate have been characterized in zebrafish. One such gene is her-1, a homolog of the Drosophila pair-rule gene hairy. As in Drosophila, this gene marks the developing segments in zebrafish in a pair-rule fashion. In fact, her-1 "prefigures" somite development by several hours, making it the earliest known marker of specified somites. Mutations have been isolated that disrupt the expression pattern of her-1 in zebrafish embryos, and thus identify genes important for the initial specification of somites.

The molecular characterization of genes responsible for mutant phenotypes may identify genes involved in human disease. For example, zebrafish mutants that display muscle-specific degeneration may identify components involved in the destructive degeneration-regeneration process that occurs in human muscle dystrophies. Besides identifying genes involved in these diseases, these mutant lines would then become useful model systems to test potential therapies. Other mutants, such as the large number of genes affecting motility and locomotion, may help us identify genes involved in proper neuromuscular function and in human neuromuscular diseases.

Up to Top

Zebrafish and Cancer Research

Keith C. Cheng
Penn State University College of Medicine
June 2, 1999

The fundamental phenotypes of the human cancer cell include abnormal growth control, abnormal cell differentiation, invasion, metastasis, and genomic instability. The identification of the genes controlling these processes among the 50,000-100,000 genes in the human genome will be one of the major charges of NIH-funded research in the next several decades. We propose that the zebrafish can play a key role in defining those genes, through the generation and study of mutants.

A variety of screens can yield information about cancer. Among the hundreds of genes identified in the large scale developmental mutant screens, many may be expected to affect growth control, cell communication, differentiation and and cell migration. Some of the temperature-sensitive fin regeneration mutations generated at the University of Oregon may lead to discovery of basic cell growth control mechanisms. Among the hematopoiesis mutants are some whose phenotype is reminiscent of myelodysplasia or pre-leukemia.

Two Penn State screens may yield tumor suppressor genes. One, currently funded by the National Science Foundation, has yielded mutants that show elevated frequencies of somatic loss of heterozygosity. These genomic instability mutations are expected to have defects in chromosome segregation, recombination, and gene regulation. Preliminary data suggest potential tumor susceptibility, consistent with the hypothesis that some of these mutations will function as tumor suppressor genes. The second screen is based upon the high sensitivity of light microscopy to detect defects in cell differentiation, as is done routinely in human pathology. A particularly interesting mutant affects multiple organs and shows cytological abnormalities that are common in cancer, including pleomorphism, nuclear atypia, and hyperchromatic nuclei. This work shows that a histology-based screen can produce potentially important mutants.

Large scale zebrafish screens have required funding by companies or private research institutes. The corresponding difficulty of individual labs to gain funding for small-scale screens suggests that "small labs" need a mechanism for funding of large facilities for collaborative research endeavors. The building of new zebrafish facilities around new areas of investigation not covered in developmental biology screens is one potential approach. To facilitate the exploration of cancer-related research, it will be best to maintain a cross-Institute effort in the support of zebrafish.

Up to Top

Eye Development and Disease

John Dowling
Harvard Biological Labs

The zebrafish visual system, especially the eye, is particularly amenable for genetic analysis. Development of the eye is rapid; within 24 hours post-fertilization (pf) a well-formed eye is present. Differentiation of the neural retina occurs between 1 and 3 days pf; so that by three days pf the retina appears functional. Visual responses can first be elicited at this time, and by 5-6 days pf robust optokinetic reflex responses can be obtained from 98% of normal animals. Initially the retina is cone dominated; abundant rods are not evident until the second week of life. In adult fish, the retina contains four morphologically distinct cone types. Short single cones are ultraviolet-sensitive whereas long single cones are blue-sensitive. The principal and accessory members of the double cones are red- and green-sensitive respectively. In addition, the cones are arranged in a precise mosaic pattern across the retina, aiding in their identification.

In genetic studies carried out so far, numerous mutations affecting retinal development have been observed (Malicki et al., 1996; Heisenberg et al., 1996; Fadool et al., 1997). Most of these mutants were detected because of a small eye phenotype. Many show both retinal and brain defects histologically, but some show only retinal deficits. Mutants with abnormal retinal lamination have been observed, as well as mutants showing selective retinal cell loss. In some cases cells fail to form; in other cases, specific cell types rapidly degenerate after differentiation. Visual behavioral studies at 5-7 days (pf) (Brockerhoff et al., 1995; Brockerhoff et al., 1997) have revealed functional defects in morphologically normal fish. One mutant, for example, appears to have a defect in synaptic transmission between photoreceptor and second-order cells. Another mutant loses all its red-cones between 3 and 5 days pf, resulting in an animal deficient in red-sensitive vision. This latter mutation does not involve the opsin gene, suggesting that this mutation represents a new form of cone-specific color blindness. Finally, a dominant mutation causing slow photoreceptor cell degeneration has been detected behaviorally in adult fish (>4 months pf) (Lei and Dowling, 1997). When homozygous, this mutation causes an early (~2 day pf) massive retinal and tectal cell degeneration and death of the animal by 5 days of age, suggesting that the gene involved is not photoreceptor cell-specific. Again, this mutation appears to represent a new type of inherited retinal degeneration. Another promising approach is the study of retinal-tectal projections. In Friedrich Bonhoeffer has been able to detect over 100 mutants which have defects of the ganglion cell axons finding their way to the tectum.

Future studies will include isolation of visual mutations using more subtle and sophisticated behavioral tests, as well as the further characterization of both behavioral and morphological mutants. So far the focus has been on retinal mutations, but mutations affecting higher visual processing and eye movement mechanism are likely to be found. Molecular genetic studies to isolate the mutated genes represent an important next step in the enterprise.


Brockerhoff, S. E., Hurley, J. B., Janssen-Bienhold, U., Neuhauss, S. C., Driever, W. and Dowling, J. E. A behavioral screen for isolating zebrafish mutants. Proc. Natl. Acad. Sci., 92:10545-10549, 1995.

Brockerhoff, S. E., Hurley, J. B., Niemi, G. A. and Dowling, J. E. A new form of inherited red-blindness in zebrafish, J. Neurosci. 17:4236-4242, 1997.

Fadool, J. M., Brockerhoff, S. E., Hyatt, G. A. and Dowling, J. E. Mutations affecting eye morphology in the developing zebrafish, Danio rerio. Developmental Genetics. 20:288-295, 1997.

Heisenberg, C.-P., Brand, M., Jiang, Y.-J., Warga, R. M., Beuchle, D., van Eeden, F. J. M., Furutani-Seiki, M., Granato, M., Haffter, P., Hammerschmidt, M., Kane, D. A., Kelsh, R. N., Mullins, M. C., Odenthal, J. and Nüsslein-Volhard, C. Genes involved in forebrain development in the zebrafish, Danio rerio. Development, 123:179-190.

Li, L., and Dowling, J. E. A dominant form of inherited retinal degeneration caused by anon-photoreceptor cell-specific mutation, Proc. Natl. Acad. Sci., 94:11645-11650, 1997.

Malicki, J., Neuhauss, S. C. F., Schier, A. F., Solnica-Krezel, L., Stemple, D. L., Stainier, D. Y. R., Abdelilah, S., Zwartkruis, F., Rangini, Z. and Driever, W. Mutations affecting development of the zebrafish ear. Development, 123:263-273, 1996.

Up to Top

Kidney Development and Diseases

Iain Drummond

Mass. General Hospital - East, Boston

The zebrafish, as a model system for vertebrate development, offers distinct experimental advantages for studies of kidney development. The zebrafish pronephros is a remarkably simple kidney, consisting of only two nephrons with glomeruli fused at the midline, pronephric tubules connecting directly to the glomeruli via a neck segment, and paired bilateral pronephric ducts which convey the altered blood filtrate outside the animal (1). Large scale mutant screens taking advantage of the optical transparency of zebrafish embryos have isolated many mutants with obvious visible kidney defects (1,2,3). In addition, a recent mutagenesis screen in the laboratory of Dr. Mark Fishman employing wt1 as an in situ probe for glomerular development has suceeded in isolating new mutants with more subtle kidney defects. The combined yield of all screens has provided mutants affecting each of the principle stages of nephrogenesis including 1) specification of the nephrogenic mesoderm, 2) nephron patterning, 3) the development of epithelial polarity, and 4) vascularization of the glomerulus. In addition, fifteen loci have been isolated which result in cystic maldevelopment in the pronephros.

Formation of the nephrogenic mesoderm.

Dorsalized mutants swirl and somitobun lack signals necessary for ventral mesodermal development and show a reduction or elimination of pax2.1-positive, presumptive kidney intermediate mesoderm (4). Conversely, the ventralized mutant chordino lacks a dorsal determinant and shows an expansion of the pax2.1 expression domain (5). The dorsalized mutants swirl and somitobun carry mutations in zebrafish BMP2b and SMAD5 respectively (6-8), while the ventralized mutant chordino carries mutations in zebrafish chordin, an inhibitor of BMP activity (9). The data from these mutants suggests that BMP signaling, most likely in concert with additional signals, is required to specify kidney mesoderm, as well as other ventral mesodermal derivatives.

Nephron patterning

Zebrafish pronephric podocytes and tubules differentiate from a single primordium, suggesting that nephron patterning events must occur to restrict podocyte and tubule cell differentiation to their appropriate cell fates and positions in the nephron (1). In the zebrafish pax2.1 mutant no isthmus (noi) we have observed a failure to develop pronephric tubules and an expanded expression of podocyte markers wt1 and vegf in cells normally destined to differentiate into the pronephric duct. These results shed new light on the role of pax2.1 in restricting podocyte development and suggest that pronephric tubule and podocyte cell differentiation may be regulated by antagonistic signals.

The development of epithelial polarity

Several mutants affecting epithelial polarity and the terminal differentiation of pronephric duct epithelial cells have been found (1). The function of the duct cells in recovering ions and metabolites from the glomerular filtrate is dependent upon the establishment of apical-basal membrane polarity and the asymmetric distribution of membrane transporters. The driving force for many transport processes is the sodium gradient established by the Na+/K+ ATPase, a membrane bound ion pump normally located basolaterally in kidney tubule and collecting system epithelia (10). In several mutants including double bubble, fleer and dizzy, antibody staining for the Na+/K+ ATPase is strongly apical and diminished on basolateral membranes (1), suggesting a failure to properly target or retain the Na+/K+ ATPase on the basolateral cell surface. Studies of these mutants are likely to contribute to our understanding of mechanisms underlying the generation of cell polarity and terminal epithelial cell differentiation.

Vascularization of the glomerulus and podocyte morphological differentiation

In zebrafish the nephron primordia rest directly on the ventral surface of the dorsal aorta where the glomerulus will form. Between 40 and 48 hours post-fertilization (hpf) glomerular formation occurs starting with the extension of podocyte foot processes and the interdigitation of the glomerular basement membrane (GBM) with endothelial cells (1). This infolding of the GBM correlates with capillary loop formation and the onset of glomerular filtration. The zebrafish mutant cloche survives into the larval period desptie the fact that it is nearly completely lacking in endothelial cells (11). Zebrafish cloche embryos thus provide a unique opportunity to investigate the role of endothelial cell derived signals in the differentiation of pronephric podocytes. We find that the differentiation of podocytes in cloche proceeds normally despite the complete absence of glomerular endothelial cells (12). cloche podocytes express wt1 and vegf and form extensive foot processes arranged as pedicels along a basement membrane (12). These findings indicate that once initiated, podocyte differentiation can proceed independently of endothelial cells or endothelial cell derived signals. The idea that podocytes may initiate glomerulogenesis by recruiting endothelial cells is substantiated in studies of the mutant floating head. floating head embryos lack a notochord and as a consequence also lack a dorsal aorta (13), the normal blood supply for the pronephros. We have found that in the absence of the dorsal aorta, clusters of podocytes develop at ectopic lateral locations and appear to recruit flk-1-positive endothelial cells from alternate sources. Remarkably, these clusters of podocytes proceed to form two separate vascularized nephrons, fully capable of blood fitration. The availablity of these mutants has allowed us to demonstrate that podocytes play a major role in glomerular morphogenesis, independently of associated cells and tissues.

Pronephric mutants as models of cystic maldevelopment.

Fifteen different zebrafish genetic loci have been identified which when mutated result in the development of pronephric cysts followed by general edema (1). The total number of mutants having a kidney cyst phenotype is now thirty-eight and includes mutants from the original screens in Tübingen and Boston as well as a recently completed in situ hybridization screen in Mark Fishman's laboratory at MGH. This large number of mutants and our results that additional mutants are falling into the original fifteen complementation groups suggests that a genetic pathway responsible for proper epithelial development may be close to saturated with mutations. The mutant double bubble, develops glomerular cysts at the stage when blood filtration is getting established and later, defects in the glomerular basement membrane are evident (1). As noted above, double bubble mutants also show defects in Na+/K+ ATPase targeting in the pronephric ducts which may result in altered ion gradients across pronephric epithelia. Despite the cloning of several genes associated with human polycystic kidney disease, the mechanism of cyst formation remains unknown. In human disease, developmental or cellular defects in epithelia caused by cystic mutations are often obscured by compensatory tissue responses, making it difficult to define the primary cause of cyst formation. The mechanism of cyst development in zebrafish cyst mutants is also unknown but could be due to either 1) an inability of the forming glomerulus to withstand the onset of filtration pressure or 2) to altered osmotic forces in the mutant pronephroi. To distinguish among these possibilities we have created a double mutant homozygous for a the dbb cystic mutation and the myocardial function mutation silent heart. These double mutants have no blood pressure to drive filtration and would be not be expeccted to develop cysts if glomerular filtratration were the primary force driving cyst formation but presumably would develop cysts if osmotic forces within the pronephros were the cause of glomerular and tubule distension. These types of experiments underscore the utility of zebrafish as a genetic system to perform types of analysis that cannot reasonably be performed in any other model organism. It is interesting to note that both glomerulocystic kidneys and apical mislocalization of the Na+/K+ ATPase have been associated with mutations in the PKD1 gene which accounts for the majority of cases of autosomal dominant polycystic kidney disease (ADPKD) in humans (14,15). In a candidate gene approach to cloning the zebrafish cystic genes, we have isolated zebrafish homologs of the alpha1 subunit of the Na/K ATPase which is mislocalized in dbb, as well as homologs of polycystin (PKD1) and the mouse cystic gene tg737. Two zebrafish cyst mutants, fleer and elipsa, display a combined renal-retinal dysplasia (1,16), a disease in humans referred to as Senior-Loken syndrome (17). No human loci or candidate genes have yet been identified for Senior-Loken syndrome. Complementation tests of fleer and elipsa as well as additional cyst mutants isolated in the Tübingen screen have shown that fleer and elipsa are distinct loci with one allele for fleer and two alleles for elipsa. Current efforts to map and positionally clone these zebrafish genes will potentially speed the isolation of human genes responsible for similar heritable defects.


1. Drummond IA, Majumdar A, Hentschel H, Elger M, Solnica-Krezel L, Schier AF, Neuhauss SC, Stemple DL, Zwartkruis F, Rangini Z, Driever W, and Fishman MC (1998) Early development of the zebrafish pronephros and analysis of mutations affecting pronephric function. Development 125:4655-67.

2. Haffter P, Granato M, Brand M, Mullins MC, Hammerschmidt M, Kane DA, Odenthal J, van Eeden FJ, Jiang YJ, Heisenberg CP, Kelsh RN, Furutani-Seiki M, Vogelsang E, Beuchle D, Schach U, Fabian C, and Nusslein-Volhard C (1996) The identification of genes with unique and essential functions in the development of the zebrafish, Danio rerio. Development 123:1-36.

3. Driever W, Solnica-Krezel L, Schier AF, Neuhauss SC, Malicki J, Stemple DL, Stainier DY, Zwartkruis F, Abdelilah S, Rangini Z, Belak J, and Boggs C (1996) A genetic screen for mutations affecting embryogenesis in zebrafish. Development 123:37-46.

4. Mullins MC, Hammerschmidt M, Kane DA, Odenthal J, Brand M, van Eeden FJ, Furutani-Seiki M, Granato M, Haffter P, Heisenberg CP, Jiang YJ, Kelsh RN, and Nusslein-Volhard C (1996) Genes establishing dorsoventral pattern formation in the zebrafish embryo: the ventral specifying genes. Development 123:81-93.

5. Hammerschmidt M, Pelegri F, Mullins MC, Kane DA, van Eeden FJ, Granato M, Brand M, Furutani-Seiki M, Haffter P, Heisenberg CP, Jiang YJ, Kelsh RN, Odenthal J, Warga RM, and Nusslein-Volhard C (1996) dino and mercedes, two genes regulating dorsal development in the zebrafish embryo. Development 123:95-102.

6. Nguyen VH, Schmid B, Trout J, Connors SA, Ekker M, and Mullins MC (1998) Ventral and lateral regions of the zebrafish gastrula, including the neural crest progenitors, are established by a bmp2b/swirl pathway of genes. Dev Biol 199:93-110.

7. Kishimoto Y, Lee KH, Zon L, Hammerschmidt M, and Schulte-Merker S (1997) The molecular nature of zebrafish swirl: BMP2 function is essential during early dorsoventral patterning. Development 124:4457-66.

8. Hild M, Dick A, Rauch GJ, Meier A, Bouwmeester T, Haffter P, and Hammerschmidt M (1999) The smad5 mutation somitabun blocks Bmp2b signaling during early dorsoventral patterning of the zebrafish embryo. Development 126:2149-2159.

9. Schulte-Merker S, Lee KJ, McMahon AP, and Hammerschmidt M (1997) The zebrafish organizer requires chordino [letter]. Nature 387:862-3.

10. Drubin DG, and Nelson WJ (1996) Origins of cell polarity. Cell 84:335-344.

11. Stainier DY, Weinstein BM, Detrich HW, 3rd, Zon LI, and Fishman MC (1995) Cloche, an early acting zebrafish gene, is required by both the endothelial and hematopoietic lineages. Development 121:3141-50.

12. Majumdar A, and Drummond IA (1999) Podocyte differentiation in the absence of endothelial cells as revealed in the zebrafish avascular mutant, cloche [In Process Citation]. Dev Genet 24:220-9.

13. Fouquet B, Weinstein BM, Serluca FC, and Fishman MC (1997) Vessel patterning in the embryo of the zebrafish: guidance by notochord. Dev Biol 183:37-48.

14. Torra R, Badenas C, Darnell A, Bru C, Escorsell A, and Estivill X (1997) Autosomal dominant polycystic kidney disease with anticipation and Caroli's disease associated with a PKD1 mutation. Rapid communication. Kidney Int 52:33-8.

15. Wilson PD, Sherwood AC, Palla K, Du J, Watson R, and Norman JT (1991) Reversed polarity of Na(+) -K(+) -ATPase: mislocation to apical plasma membranes in polycystic kidney disease epithelia. Am J Physiol 260:F420-30.

16. Malicki J, Neuhauss SC, Schier AF, Solnica-Krezel L, Stemple DL, Stainier DY, Abdelilah S, Zwartkruis F, Rangini Z, and Driever W (1996) Mutations affecting development of the zebrafish retina. Development 123:263-73.

17. Gusmano R, Ghiggeri GM, and Caridi G (1998) Nephronophthisis-medullary cystic disease: clinical and genetic aspects. J Nephrol 11:224-8.

Up to Top

Cardiovascular and Other Organogenesis in Zebrafish

Mark C. Fishman
Mass General Hospital and Harvard Medical School

Zebrafish mutations discovered in the two original, and several subsequent, large screens have already revealed an organizing logic to organ development, physiology, disease and evolution. Each of the genes discovered is an entrance point to pathways not previously imagined.

Screens for organ mutants are of biologic and medical importance

1. Organs can be dissected as genetic "modules". Single gene mutations in the zebrafish can remove or disrupt components of organs with relative selectivity, leaving the remainder of development to proceed with impunity. In this way, they reveal separable units, as did screens in the fly for units of body form. These include genes for formation of the ventricle of the heart, for heart valves, or for the glomerulus of the kidney. 2. Mutations reveal candidates for disease. Single gene heritable arrhythmias and heart failure mutants speak to common adult disorders. Single gene-related disorders of early organ form or laterality are models of common congenital abnormalities. Complex diseases have not proved readily decipherable by whole genome linkage studies. However, the power of association studies, using candidate genes, is such that it can reveal the role of genes contributing even minimal components of risk. Screens in zebrafish provide such candidates. 3. The mutations perturb "new" vertebrate evolutionary additions. The global form of heart, kidney, pancreas, etc. are new on the evolutionary scene, not present in primitive chordates. Interestingly, the modules perturbed by mutations are precisely those for these new evolutionary additions, and therefore speak to how large-scale innovations may have occurred by single gene mutations during evolution.

The zebrafish is the most tractable and relevant genetic system for organ screens

Unlike Drosophila or C. elegans, the organs of zebrafish are similar to those of man. Whereas Drosophila lacks vessels and the heart has no ventricle, endocardium, valves, or conduction system, the zebrafish heart is essentially identical to that of the three week gestation human. The mouse embryo is not as good for organ screens because of its relative opacity. More importantly, and in contrast to the zebrafish, the mouse embryo depends upon ongoing delivery of nutrients and removal of waste products, leading to rapid deterioration of mutants which affect these processes.

Some examples of zebrafish organ mutants which have proved important

1. Models of disease.

a) Congenital

The gridlock mutation lacks circulation to the tail, because of failure to assemble one branch point, embryologically identical to the region affected in congenital coarctation of the aorta (Weinstein et al., 1995). The asymmetry mutants, in which the right-left axis of the heart is reversed, affect the process believed perturbed in ventricular inversion (Chen et al., 1997). The mutation with a diminutive ventricle (Stainier et al., 1996) affects processes relevant to the hypoplastic heart syndromes.

b) Adult

Many common disorders are impossible to model in animals, and are approachable in humans only by total genome scanning. For example, heart failure is a common and polygenic disorder, frequently "idiopathic" in nature. Several mutants evidence the same dilated poorly contractile physiology as do these patients. Other individual mutants model each common rhythm disorder, including heart block, fibrillation, and bradycardia (Baker et al., 1997). Mutants which cause disarray of epithelial patterning in the gut (Pack et al., 1996) may be relevant to cancer, as are the mutants which affect angiogenesis. Those which perturb one pancreatic lineage (Pack et al., 1996) speak to issues important in tissue replacement for cystic fibrosis and diabetes.

2. New paradigms in biology

These mutants have revealed steps of organ assembly, steps we could not have predicted to exist. In terms of generation of heart form, for example, there are genes needed specifically to generate one chamber but not the other, to assemble a single heart tube rather than two, to elicit valve formation or to generate the endocardium (Fishman and Olson, 1997). Mutants separate lineages, such as those for pancreatic exocrine and endocrine cells. In each case, the mutation is discrete and other aspects of development are relatively intact. These mutants provide an entrance to pathways to understand development of the ventricle, or valves, or endocardium, etc. In addition, the mutants provide evidence that discrete genetic pathways regulate important physiological parameters, such as heart size (Stainier et al., 1996).

Future programs

Biological and molecular work-up of current mutants

In vitro assay systems have been useful. Patch clamping of isolated cardiocytes, for example, has identified the channel defect in pacemaker mutants. Cell transplantation between mutant and wild-type fish has revealed the existence of unexpected signals, for example, how tissues regulate assembly of neighboring vessels. The movements of cells in the heart field, and essential control over regulation in the field to perturbants, can be analyzed at a single cell level by caged fluorescent lineage markers (Serbedzija et al., 1998). Positional cloning is underway. Using the microsatellite map , there is linkage for about 20 organ mutants, and positional cloning has delimited the region, for several, to a BAC.

New screens

There are opportunities for new types of screens with regard to organs. Screens based on molecular probes can identify mutations in organ systems which are difficult to visualize directly, such as the kidney, or in the precursor populations which generate the organs. Insertional screens, taking advantage of sequenced sites around insertions to reveal tissue-related ESTs, could provide mutants in defined genes. Dominant mutations, in adults, will allow correlation with physiology and discovery of modifier mutations.

Candidate genes for the human genome

It goes without saying that, once cloned, the genes for mutants with medically relevant phenotypes become candidates for disease genes. The phenotypes suggest that these might be especially useful in the study of complex and common diseases, for which uniform human populations may be elusive, such as heart failure, sudden death, and diabetes.

Selected References

Baker, K., Warren, K. S., Yellen, G., and Fishman, M. C. (1997). Defective "pacemaker" current (Ih) in a zebrafish mutant with a slow heart rate. Proc. Natl. Acad. Sci. USA 94, 4554-4559.

Chen, J.-N., van Eeden, F. J. M., Warren, K. S., Chin, A. J., Nusslein-Volhard, C., Haffter, P., and Fishman, M. C. (1997). Left-right pattern of cardiac BMP4 may drive asymmetry of the heart in zebrafish. Development 124, 4373-4382.

Fishman, M. C., and Olson, E. N. (1997). Parsing the heart: genetic modules for organ assembly. Cell 91, 153-156.

Pack, M., Solnica-Krezel, L., Malicki, J., Neuhauss, S. C. F., Schier, A. F., Stemple, D. L., Driever, W., and Fishman, M. (1996). Mutations affecting development of zebrafish digestive organs. Development 123, 321-328.

Serbedzija, G. N., Chen, J.-N., and Fishman, M. C. (1998). Regulation in the heart field of zebrafish. Development 125, 1095-1101.

Stainier, D. Y. R., Fouquet, B., Chen, J. N., Warren, K., Weinstein, B., Meiler, S., Mohideen, M. A. P. K., Neuhauss, S. C. F., Solnica-Krezel, L., Schier, A. F., Zwartkruis, F., Stemple, D. L., Malicki, J., Driever, W., and Fishman, M. C. (1996). Mutations affecting the formation and function of the cardiovascular system in zebrafish embryos. Development 123, 285-292.

Weinstein, B. M., Stemple, D. L., Driever, W., and Fishman, M. C. (1995). gridlock, a localized heritable vascular patterning defect in the zebrafish. Nature Med. 1, 1143-1147.

Up to Top

Research on Aging

David Grunwald
University of Utah

The most outstanding feature of the recent studies with the zebrafish is the broad range of developmental and physiological characteristics that can be probed by mutational analysis in this organism. In addition to the well-known isolation of mutants with embryonic developmental phenotypes, ongoing studies that are yielding palpable results include the recovery of mutations that affect cardiovascular physiology and the rate at which somatic mutations arise spontaneously. Hence it is evident that zebrafish research can be applied toward the study of genetic, cell biological, and developmental processes that influence or correlate with aging. We foresee four areas of research that may contribute significantly to our understanding of aging:

1) research concerning the accumulation of somatic mutations, including the study of genetic variants with altered somatic mutation rates

2) research concerning the biological effects of alterations in DNA metabolism, including the study of genetic variants that affect DNA replication, repair, or recombination

3) research concerning the biological effects of altered metabolic rates, including the study of genetic variants that display alterations in the accumulation of specific metabolite species, such as free radicals

4) research concerning factors that influence the rate of development, including the study of the potential correlation between the timing with which developmental landmarks are attained and other markers of aging or lifespan.

The areas of research highlighted here reflect the priorities of ongoing research in the zebrafish field. We anticipate that as the field matures and expands, a broader range of developmental and physiological characteristics will come under study, and some of these additional areas of research will be of significance to aging processes.

Up to Top

Child Health and Human Development

Charles Kimmel
University of Oregon

Fundamental developmental genetic mechanisms have been conserved in all vertebrates. Because of this fact and because zebrafish have marked advantages for both genetic and developmental studies, the zebrafish provides a useful model for understanding many aspects of human development.

At the genetic level, conservation is apparent from mutant phenotypes. For example, our earlier work revealed that the zebrafish no tail gene is the homologue of the Brachyury gene in the mouse, and the zebrafish valentino gene is the homologue of the kreisler gene in the mouse. Both genes were initially identified in both species by their mutant phenotypes. Brachyury/no tail mutants lack tail and notochord development in both zebrafish and mouse. kreisler/valentino mutants lack development of hindbrain segments r5 and r6 in both zebrafish and mouse. These were the first two cases in which two vertebrate species have been connected by functional (mutational) analyses of orthologous genes. In both of these cases the phenotypic analysis available in zebrafish has pointed to new understanding of the defects in the mouse. These initial two cases demonstrate that zebrafish is a useful system for understanding how genes encode early development in all vertebrates, including humans.

Since these studies, thousands of mutations have become available through continued and expanded genetic screens. My lab has collaborated with others in the study about 10 of these mutations, and during the past two years we have identified 5 of them moleularly. They include genes of the PBX family that control hindbrain segmentation, a T-box that specifies muscle development in dorsal mesoderm, and transcription and cell signaling factors that control craniofacial skeletal patterning. All of these areas are crucial ones to understand with respect to our eventual abilities to deal with human disorders and birth defects

Our excellent progress has depended on advances in zebrafish genomics. All of the initial identification of the genes relied on our new mapping abilities, including making use of the radiation hybrid panels that have become recently available. Cloning the genes depended on new libraries, and one of them depended on the very newly available EST databases. This progress is only a taste of what we will witness in the future, because we have only begun to scratch the surface in terms of both mutational analyses and utilizing genomics. We will most certainly continue to learn that many of the genes now known only by their mutant phenotypes in zebrafish have direct counterparts in humans. This will help us understand what is likely happening when these human genes function incorrectly or not at all, as in many types of congenital diseases.

Up to Top


Leonard I. Zon
Children's Hospital, Boston

The zebrafish is an excellent system to study hematopoiesis. Blood is formed within 24 hours and the optical clarity of the embryo allows visualization of mutants that affect the number and differentiation of circulatory blood cells. Normal hematopoiesis occurs in a region above the yolk sac called the intermediate cell mass. This region yields primitive embryonic cells initially and then subsequently c-myb positive cells in the ventral wall of the dorsal aorta, similar to the aorta-gonad-mesonephros (AGM) region of higher vertebrates. The posterior region of the ICM contains cells that are poorly differentiated and have a blast-like morphology. These cells are likely to be stem cells or progenitors. Zebrafish globins have been characterized and embryonic to adult hemoglobin switching occurs after day 21 of development. Over 80 genes have been cloned that are homologs of human or mouse factors that participate in hematopoiesis. cDNA homologs have been isolated for every gene that has been knocked-out in murine embryonic stem cells which have a defect in hematopoiesis. This includes SCL, LMO2, GATA-1, GATA-2, TTG2, PU.1, myb, flk1, flt4, and fli1. Thus, the hematopoietic program has been largely conserved throughout vertebrate evolution.

We have collected over 26 complementation groups of hematopoietic mutants and have recently derived 8 complementation groups of mutants affecting the T cell program. Mutants have defects in either dorsal-ventral patterning, hemagioblast formation, hematopoietic stem call formation, proliferation or differentiation. A number of the mutants have difficulty in making hemoglobin. Recently, the gene for two of the mutants with defects in hemoglobin production were isolated. These encoded the ALAS-E gene and UROD gene, two enzymes involved in heme biosynthesis. Thus, the fish is an excellent model for studying hemoglobin production. In addition, the fish model has a relevance to human disease in that missense mutations in these genes cause congenital sideroblastic anemia or phorpheryia, respectively. The sauternes gene was isolated by positional cloning approaches, thereby demonstrating the success of this technique. Isolation of many of the mutant genes are in progress and should provide interesting novel genes for regulation of hematopoiesis as well as disease models for understanding the pathophysiology of the disorder.

Genetic epistasis is possible in the zebrafish. cloche is a defect in which there is no blood or blood vessel formation. Recent evidence has demonstrated that the cloche mutant can be rescued by the overexpression of the helix-loop-helix protein SCL. Overexpression of SCL leads to rescue of hematopoiesis and vasculogenesis. Thus, overexpression of downstream genes can rescue mutant phenotypes i.e. SCL acts downstream of cloche. In the future, it should be possible to do focus screens for mutants that affect other hematopoietic tissues or to envision studying earlier the process of stem cell formation in vivo in the embryo. With this respect, transgenic lines have been made in which green fluorescent protein has been driven by either GATA-1 or GATA-2 promoters. This allows an early definition of the hematopoietic program in vivo and should allow for isolation of interesting cell populations and characterization of the mutants. Finally, dominant suppressor screens may be possible in the zebrafish, thus developing genetic pathways surrounding mutant phenotypes.

Up to Top

Research on Ear Morphogenesis and Hearing Disorders

Eric S. Weinberg
University Of Pennsylvania
Jarema Malicki
Mass Eye and Ear, Boston

Animal models are extremely valuable for learning the genetic basis of inner ear disorders. Because of obvious similarities in the pathology of humans and mice with inner ear defects, the murine model is particularly important model for genetic deafness. It is now evident from recent work with the zebrafish that a second powerful animal model is now available for studying the genetic basis of inner ear development and function. In particular, the ability to trace the fate and lineage of early embryonic cells, the feasibility of large scale screens of mutants with early embryonic phenotypes, the ease of assaying the developmental effects of gene products produced ectopically from injected RNAs, and the possibility of constructing transgenic and chimeric animals all make the zebrafish a potentially valuable animal model.

Recent mutant screens have yielded over 90 mutations (affecting approximately 40 loci) that primarily affect development of the inner ear. Mutations in at least nine genes affect morphology and patterning of the inner ear epithelium, including semicircular canal development and development of maculae and cristae. Other mutants affect ear size, otolith formation, or initial specification of the otic placode. Many of the ear mutants also exhibit abnormal behaviors such as swimming in circles or upside down. Work on these mutants is only just beginning, and undoubtedly, further analysis will increase our understanding of inner ear development. One particularly well studied case is the valentino mutation, which is a homolog of the mouse kreisler gene. In both zebrafish and mouse, a malformation of the inner-ear is accompanied in mutant individuals by abnormalities in the hindbrain, the site at which the gene is active, suggesting conserved signaling mechanisms in inner-ear formation.

The vestibular part of the ear of teleosts and mammals is remarkably conserved. Inner ears of both taxa have semicircular canals, a utriculus, a sacculus, and types of otolith structures; the main auditory organ in mammals (and birds), however, the cochlea, is absent in the teleost ear. Nevertheless, for auditory functions, teleosts appear to use the sensory epithelium of the sacculus and other structures that are primarily vestbular in mammals. Since the sensory epithelia of both vestibular and auditory structures of mammals and other vertebrates share a common basic tissue architecture of hair cells and supporting cells, mutants defective in such structures in fish ARE relevant to studies of human sensory epithelial defects. At least two mutants with deffective hair cell cillia have been recently identified in zebrafish: sputnik (spu) and mariner (mar). Related to this previous point is the finding that some types of mammalian deafness are accompanied by impairment of vestibular function; such syndromes (e.g., Usher syndrome type 1B) are caused by defects in components of the sensory epithelium common to the vestibular and auditory regions of the ear. Moreover, different alleles of the same gene can cause loss of vestibular function without impairment of hearing, or loss of hearing without loss of vestibular function. Therefore, identification of mutants with purely vestibular defects (easily identifiable in zebrafish mutant screens) may ultimately be useful for identifying genes that have both vestibular and auditory functions in humans.

Even at first glance, many of the recently identified zebrafish phenotypes resemble human inherited auditory disorders. For example, branchio-oto-renal syndrome (BOR), mandibulofacial dystosis, craniofacial dystosis, Alpert syndrome and some others are associated with craniofacial abnormalities. Similar combination of phenotypes is obvious in the zebrafish mutations quadro and little richard which affect both otic vesicle and branchial arches. Likewise, selected genetic defects of both human and zebrafish auditory system are associated with pigmentation defects. In humans, wardenburg syndrome, piebaldness, vitiligo and universal dyschromatosis involve depigmentation. Among fish mutants golas, piegus, mizerny, punktata and others affect both auditory system and pigmentation. The current collection of genetically defective zebrafish strains, although already impressive, is likely to expand even more. New, creative screening methods offer opportunities to search for mutations affecting more and more specific developmental processes. In the area of ear research, particularly promising are screens for defects of hair cell development. Zebrafish lateral line hair cells can be visualized by immersing fish larvae in solution of a fluorescent dye for a few minutes allowing for efficient detection of many aberrations of this important cell type. As hair cells are essential for auditory perception of both fish and humans, this type of screening may have substantial medical importance.

The zebrafish is rapidly becoming a valuable model system for the studies of human hearing disorders. In the immediate future, we expect progress in two areas: identification of additional mutations of zebrafish hearing and molecular characterization of mutant genes. The current and future mutagenesis screens will involve increasingly more sophisticated screening approaches, such as behavioral tests. One of the currently performed screens, for example, takes advantage of the startle response to identify hearing deficient fish. The second area of future growth is molecular characterization of mutant genes. Rapid development of genomic tools, such as genome maps, large insert libraries and est databases, has already produced benefits in that several mutant genes have been characterized via positional cloning and many others through the candidate approach. These prospects should be a major factor in encouraging research on the basic development and sensory physiology of the zebrafish inner ear, and hence, the middle ear of all vertebrates.

Up to Top

Developmental Toxicology In Zebrafish

R.E. Peterson and W. Heideman
University of Wisconsin, Madison, WI

The zebrafish has been extensively utilized for toxicologic studies, allowing rapid testing of various agents on normal embryogenesis. The zebrafish system has the potential to genetically dissect pathways that regulate response to toxins.

The polychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), and biphenyls (PCBs) are global environmental contaminants and potent developmental toxicants. Human exposure occurs during pregnancy and breast feeding when these chemicals are transferred transplacentally to the embryo and fetus and lactationally to the neonate posing a human health risk. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is the prototype compound used to study this type of toxicity. In all vertebrates, the developing embryo/fetus is significantly more sensitive to TCDD than the adult, suggesting that TCDD has the ability to perturb critical developmental events. Therefore, a thorough mechanistic understanding of TCDD action during early development is necessary to assess the risk that such exposure poses to children. Toxic effects of TCDD occur by TCDD binding with high affinity to the aryl hydrocarbon receptor (AhR) followed by changes in gene expression. After TCDD binding, the activated AhR translocates to the nucleus and dimerizes with its partner ARNT. AhR-ARNT heterodimers recognize and bind dioxin response elements found in promoters of responsive genes to alter gene expression. AhR and ARNT are members of the basic helix-loop-helix/PAS family of transcription factors that include drosophila PER and SIM, murine ARNT-2, and human hypoxia inducible factor 1a. Members of this family are master developmental regulators, and it is intriguing to speculate similar roles for AhR and ARNT.

The rationale for studying AhR and ARNT function in zebrafish is that much of our current understanding of TCDD action involves characterizing endpoints of TCDD exposure, and very little is know about the actual mechanism of toxicity. It is largely accepted that TCDD developmental toxicity in fish like mammals is AhR-mediated, but the gene targets of activated AhR-ARNT that are causally related to the expression of toxicity are poorly understood. Furthermore, the understanding of the physiological role of AhR and ARNT in normal development is meager at best. The zebrafish are being used to investigate the developmental toxicity of TCDD and the physiological function of the AhR signaling pathway. Zebrafish embryos exposed to graded doses of TCDD for 1 hour, from 2-5 hours post fertilization, are responsive to TCDD (Toxicol. Appl. Pharmacol. 142: 56-68, 1997). The developmental toxicity is characterized by pericardial and yolk sac edema, craniofacial malformations, and mortality. Importantly, severe hemodynamic changes, manifested as slowed blood flow in vascular beds of the trunk, head, and gills and slowed heart rate, occur prior to or coincident with the onset of lesions. Visual inspection of the transparent embryos shows that TCDD does not inhibit initial development of vasculature or prevent initiation of blood flow to the above structures. Rather it appears to interfere with the maintenance of peripheral vascular beds after they are formed.

Future research will be directed at determining whether the hemodynamic changes are due to a direct effect of TCDD on maintenance of the vasculature and/or are secondary to cardiac insufficiency. Mutagenesis screening will be used to identify key genes involved in the developmental cardiovascular toxicity of TCDD and for this purpose a zebrafish genetic map would be helpful. The human health significance of this research is that development and maintenance of a vascular supply is a fundamental requirement for organogenesis in the embryo. TCDD exposure does not appear to affect vasculogenesis. Rather the subsequent process of angiogenesis which includes sprouting, growth, migration, and remodeling of endothelial cells seems altered. Therefore, our future research will focus on the hypothesis that the developmental cardiovascular toxicity of TCDD is caused by an interference with angiogenesis. Angiogenesis is also implicated in the pathogenesis of a variety of human diseases such as proliferative retinopathies, age-related macular degeneration, tumors, rheumatoid arthritis, and psoriasis. Findings in zebrafish may ultimately demonstrate whether or not the AhR signaling pathway is an important regulator of angiogenesis with the degree of AhR activity in vascular endothelial cells correlating with physiological and/or pathological regulation of blood vessel growth.

Up to Top

Genome Research

John Postlethwait
University of Oregon

Knowing the sequence of the human genome will revolutionize our understanding of human biology, human development, and human health. But how can we go quickly from human genome sequence to human gene function? The thesis of the paragraphs below is that research on zebrafish is the most efficient way to understand the functions of genes necessary for the development and physiology of all vertebrates, and thus zebrafish research can efficiently illuminate the function of human genes known only by sequence and map position.

The schedule for a complete, polished human genome sequence has recently been revised to the end of 2003, two years ahead of earlier projections. And a "working draft" will be available already by the end of 2001, barely two years away. We are faced with a surfeit of information and a deficit of understanding. A major challenge is the interpretation of the immense amount of sequence data -- among many questions, one reigns supreme: What does each gene do?

An efficient way to find out what genes do is to isolate mutations that disrupt the gene's abilities to direct the construction and operation of the body's organs and organ systems. Screens for such mutations must be driven by phenotype, because it is the phenotype that one wants to understand: How is the phenotype constructed? What makes the phenotype function? Such screens are difficult and expensive to conduct in mammals, where embryos are opaque and hidden inside the mother. Zebrafish embryos, however, develop outside the mother and embryos are optically translucent so that each individual cell is visible until late in development. Therefore, one can conduct efficient, cost effective, phenotype driven screens at near saturation levels for mutations essential for zebrafish development and physiology. In addition to providing an understanding of the molecular genetic mechanisms of development and physiology, such mutations become models for disease in humans and other animals.

Mutations are currently available in nearly 1000 genes essential for embryonic development in zebrafish, and more are being added by directed screens. For example, by staining for the expression of specific genes, researchers have found mutations that block a single segment of the hindbrain, or a specific valve of the heart. Because the early development of human embryos, including the development of the heart, the brain, the body axis, the ear, the liver, and so on, are highly similar in zebrafish and human embryos, the genes that actually effect zebrafish development will have close relatives that will direct similar processes in human development and organ function.

An important addition to our understanding of human biology will come through the identification of human gene functions suggested by the phenotypes of zebrafish mutations. The ability to move from zebrafish mutation to human gene, however, depends on the ability to move from zebrafish genome to human genome. Fortunately, recent results have shown that zebrafish and humans share large chromosome segments which have been conserved intact during the 420 million years of evolution since the divergence of the two phylogenetic lineages. For example, the apparent orthologues of 14 genes localized along the full stretch of the long arm of human chromosome 2 (about 5% of the human genome) have been mapped in zebrafish. Eleven of these genes reside on one zebrafish chromosome, and the remaining three on one other chromosome. This same group of genes is also on two chromosomes in mouse. Analysis of the full set of mapped zebrafish genes suggests that the number of chromosome segments shared between zebrafish and human genomes may be just slightly more than the number of segments shared between the genomes of mouse and humans.

The implication of the extensive conservation of chromosome segments between human and zebrafish genomes is that comparative mapping should be almost as effective between zebrafish and humans as between mouse and humans. The pressing need now is to place enough orthologous markers on the zebrafish genetic map (which is currently at about 1 cM (about 600kb) average resolution due to microsatellite markers) to clearly define the borders of conserved segments. This could be achieved by an expanded EST project, using normalized libraries from a variety of tissues and developmental stages, and the funding of projects to map the ESTs in the genome, and to systematically catalogue their expression patterns in embryos, larvae, and adult fish. Furthermore, we need to establish a physical map of the zebrafish genome, both to improve the ability of workers to molecularly characterize the sequences disrupted by the mutations they study, but also to provide a framework for subsequent sequencing of the complete zebrafish genome.

Because of the similarity of development and physiology in zebrafish and humans, because of the ability to conduct efficient, near saturation, phenotype driven screens in zebrafish, because of the ease of single cell investigation in zebrafish embryos, and because of the similarity of genomes in zebrafish and humans, the zebrafish should be the next organism on the list of those to be sequenced in their entirety after mouse. This would enormously facilitate the cloning of mutations, and hence speed suggestions for functions of human genes.

Furthermore, the phylogenetic distance of zebrafish from humans is sufficient that sequences that are not highly constrained by natural selection will have been randomized when corresponding parts of the two genomes are compared. This will greatly facilitate the identification of regulatory elements and other functional features of the genome. Because of the ability to work with zebrafish embryos, the functions of these sequences can be directly tested in transgenic animals.

The unique feature that zebrafish brings to the analysis of vertebrate genomes is the ability to perform an efficient mutational -- and hence a functional -- analysis of a vertebrate genome. If, for example, a zebrafish mutation maps to a chromosome segment known to be conserved between zebrafish and humans, then the human genome can suggest candidate genes for the zebrafish mutation. Molecular genetic experiments can then test whether the candidate gene is in fact disrupted by the mutation. Reciprocally, and importantly, the phenotype of the zebrafish mutation can suggest functions for the human genes, which might otherwise be known only by sequence and position. In this way, a greater knowledge of the zebrafish genome could help move the human genome project from descriptive genomics to functional genomics.

Up to Top

Neurogenesis and Neurological Disorders

Alexander F. Schier Skirball Instutute, NYU School of Medicine

The study of neural development and function in the zebrafish is facilitated by the translucence and accessibility of the embryo, the relative simplicity of its nervous system, and the availability of a large number of mutations disrupting genes that are essential for normal development and physiology. Despite its simplicity, many of the features of the zebrafish nervous system are conserved in higher vertebrates and humans, including the patterning of the neural tube, the differentiation of neurons, the routes of axon tracts, neuronal signaling, and many aspects of behavior. Large-scale mutant screens in zebrafish have led to the identification of more than 50 genes affecting various aspects of neural development and function. The identified genes are involved in neural induction and patterning, axon pathfinding, neuronal differentiation, and behavior. Another 100 genes were found to be required for neural survival. To name a few examples,

  • cyclops and one-eyed pinhead mutants lack the ventral regions of the neural tube, resulting in cyclopia.
  • mind bomb mutants display supernumerary neurons.
  • unplugged mutants lead to the abnormal axon pathfinding of motorneurons.
  • who-cares mutants affect the topographic mapping of retinal axons on the tectum target area.
  • accordion mutants display abnormal locomotion.
  • nic mutants are immotile and affect acetylcholin receptor function.
  • space cowboy mutants display degeneration of the central nervous system.

In addition to their importance in the analysis of basic neurobiological processes, the identified genes might also have relevance to the study of disease states: defects in cyclopic mutants resemble the human congenital disorder holoprosencephaly, degenerative mutants might serve as models for neural degeneration conditions in humans, and retinal axon pathfinding mutants might serve as models for human defects such as ocular miswiring.

The recent molecular cloning and phenotypic characterization of zebrafish mutations has already led to some exciting new insights into neural development and function. The isolation of cyclops as a nodal-related gene and of one-eyed pinhead as an EGF-CFC protein has identified a previously unexpected role for TGF-beta signaling in neural patterning. The cloning of sonic you as sonic hedgehog and you-too as gli2 has shown that the hedgehog pathway plays an important role in forebrain development and axon pathfinding. In vivo imaging techniques have allowed monitoring the neuronal activity during locomotion and escape behavior with single-cell resolution. The on-going molecular isolation of the disrupted genes and their phenotypic characterization using modern imaging approaches promises unprecedented insights into the molecular mechanisms underlying neural function and development in vertebrates.

Up to Top

Early Developmental Programs in the Zebrafish

William S. Talbot
Stanford University School of Medicine
Pascal Haffter
Max-Planck-Institut fur Entwicklungsbiologie

The zebrafish has emerged as an important model organism for the study of early development in vertebrates. Zebrafish and other vertebrates share many fundamental similarities in the organization of their body plans and developmental pathways. The experimental advantages of the zebrafish make it uniquely suited for the investigation of dynamic processes such as axis formation and gastrulation. For example, the optical clarity and external development of the embryo allow one to trace the movements of individual cells in the living embryo as the animal develops. Moreover, large-scale genetic screens have identified more than two thousand mutations that collectively define the functions of hundreds of genes essential for many aspects of early development. For example, dino and swirl disrupt neural induction and the dorsal-ventral axis, bozozok and squint eliminate the Spemann organizer, trilobite alters the morphogenetic movements that reshape the embryo during gastrulation, and spadetail, no tail, floating head and numerous others perturb the development of specific cell types and organs that are established during embryogenesis. The elegant phenotypic analysis of these mutants that is possible in zebrafish will provide an unparalleled understanding of the functions of the disrupted genes. For example single-cell labeling experiments demonstrated that the floating head mutation alters the fates of a specific population of cells in the Spemann organizer, such that they form muscle instead of notochord. This and many other examples show that cellular analysis in zebrafish provides an understanding of gene function that cannot be matched in other vertebrates.

Over the last two years, advances in the molecular analysis of zebrafish mutations have demonstrated that zebrafish genetics can discover new genes and new gene functions. One striking example is the one-eyed pinhead gene, first described in 1996 as a mutation disrupting endoderm, Spemann organizer, and floor plate development. In 1998, positional cloning identified the oep gene as a novel membrane-associated protein of unknown function. Homologs were known in mammals and frog, but not Drosophila or C. elegans. Thus oep provided the first evidence for an essential function of a new gene family that is apparently specific to vertebrates. Recent evidence shows that the Oep protein is a co-factor for signaling by nodal-related proteins, a sub-class of the TGF-beta superfamily of secreted growth factors. The analysis of oep illustrates that mutational analysis in zebrafish can reveal the functions of novel, essential genes.

There are many other recent examples of new gene functions identified by analysis of zebrafish mutations. In one case, a previously unsuspected role for the nodal-related signals in patterning of the neural tube was established from the study of the cyclops mutation. Characterization of the extensive collection of other mutations that perturb neural patterning will allow a detailed understanding of the pathways that establish the floor plate and other specialized cell types in the ventral neural tube. Two additional genes essential for neural patterning have been shown to encode components of the hedgehog signaling pathway. Isolation of the remaining genes will likely identify novel components of the hedgehog and nodal signaling pathways and elucidate the interactions between these pathways in the development of the neural tube and other organs.

Many of the genes defined by zebrafish mutations have human homologs implicated hereditary diseases and cancer. For example, loss of hedgehog signaling can result in Grieg cephalopolysyndactyly or holoprosencephaly whereas ectopic hedgehog signaling results in basal cell carcinoma. In addition to the hedgehog and TGF-beta signals, zebrafish mutations define components of other signaling pathways with important roles in human disease, including members of the Wnt and FGF families of secreted growth factors. Thus zebrafish mutations provide powerful tools for understanding and modulating the cell signaling events that underlie human disease and vertebrate biology.

In summary, recent advances have established the value of the zebrafish as a system for the discovery of new genes and new gene functions. The molecular analysis of mutations will identify factors regulating fundamental processes conserved in vertebrates. New screens will discover more mutants with developmental and behavioral phenotypes that have not been scored previously, and genetic modifier screens will define genes that interact with key players in established pathways. Further development of transgenic technology and refinement of imaging techniques will take functional analysis of these genes to a new level, allowing the in vivo detection of gene expression.

Up to Top

Research on Allergies and Infectious Diseases

Nikolaus Trede
Children's Hospital, Boston

Similar to higher vertebrates, the teleost zebrafish has T and B lymphocytes and is thus an excellent system to study lymphopoiesis. Using degenerate oligonucleotide polymerase chain reaction (PCR) and low stringency hybridization, zebrafish homologs of early hematopoietic transcription factors, such as c-myb, PU.1 and tal-1/scl, as well as the immunoglobulin heavy chain gene (cm), the lymphocyte specific marker ikaros, the T-cell tyrosine kinase lck, and rag-1 and rag-2 were obtained. Expression of the above genes in wild-type zebrafish assayed by in situ hybridization shows presence of T cells in the bilateral thymi as early as day 3 of life. The analysis of lymphocyte development is currently being extended to include 22 complementation groups of zebrafish with defects in blood production. These mutants potentially represent genes required for lymphopoiesis and are of particular interest as they are expected to have defects in stem cell differentiation at various stages. One of these mutants, cloche, appears to produce neither erythrocytes nor lymphocytes.

By analogy to erythropoiesis, where at least 22 complementation groups have been defined, multiple developmental steps are likely to exist between the uncommitted HSC and the mature lymphocyte. To derive a panel of mutants with defects in lymphopoiesis, we have carried out chemical and radiation mutagenesis of the zebrafish genome. We screened these mutated zebrafish larvae with the erythroid and lymphoid markers, alpha-globin and rag-1, respectively. Mutants with defects in both markers are likely to have a defect in hematopoietic stem cell induction, proliferation, or differentiation. Mutants with defects in rag-1 alone are likely to solely affect lymphopoiesis.

In this screen 110 clutches of sufficient size and quality derived from ENU mutagenized F1 offspring were analyzed. We identified 17 clutches with a defect in Rag-1 expression, nine with a defect in globin expression and three with a defect in both. F1 females that these clutches were derived from were outcrossed, and to date we have confirmed the mutant phenotype in five Rag-1 deficient and one globin deficient mutants by F2 incrossing. Two of the Rag-1 mutants (CZ-3 and CZ-5) fall into the same complementation group. Phenotypic analysis of the mutants shows total absence of Rag-1 expression in CZ-3/5 and CZ-26, while CZ-18 and CZ-32 have faint Rag-1 expression in the thymi of mutant offspring. Further analysis demonstrates normal neural crest cell development and endodermal marker expression in the mutants. Pharyngeal arch development and thymic development are abnormal in CZ-3/5. We have used half tetrad analysis to map these mutants. Investigation of zebrafish with defective lymphopoiesis will further our understanding of hematopoiesis and will have therapeutic impact for bone marrow transplantation, stem cell gene therapy, and leukemia.

Up to Top

Mutants That Affect Mental Behavior

Monte Westerfield
University of Oregon

The small size, optical transparency and rapid development of zebrafish make them good subjects for studying the mechanisms that regulate development of nervous system function and behavior. Several labs are currently funded to study retinal physiology including John Dowling, Harvard University, who is recording from cone photoreceptors and Douglas McMahon, University of Kentucky, who is investigating mechanisms of retinal synaptic plasticity by electrical recording. Joe Fetcho, SUNY Stony Brook, is recording membrane potentials in hindbrain neurons during escape behaviors using optical imaging techniques. John Schmidt, SUNY Albany, is studying activity and trophism in synaptic stabilization of retinal arbors. In principle, the results of these types of studies can provide not only important information about normal physiology, but also can be used to design genetic screens for mutations in genes that regulate these physiological processes.

The first screen for behavioral mutants was carried out by Kimmel and Westerfield, University of Oregon, in the 1980's and was based on a simple test of the response to touch. A number of mutations affecting neural crest cells (sensory neurons), motoneurons, and muscles were discovered. Additional alleles of some of these mutations as well as several new mutations were obtained in the recent large scale screens which looked for paralyzed or weak movement phenotypes. During the past year, at least two labs have begun new genetic screens for mutations that affect behavior. The James Hurley lab, University of Washington, is screening for mutations affecting visual behavior and Kate Whitlock in the Westerfield lab, University of Oregon, is screening for olfactory mutants on responses to olfactants. The olfactory screen has already uncovered several new mutations. Further elaboration of behavioral mutant screens and molecular characterization of the disrupted genes will help further out understanding of the genetic basis of vertebrate behavior.

Up to Top

Craniofacial and Dental Research

James A. Weston
Institute of Neuroscience
University of Oregon

Considerable progress has been made in identifying and analyzing mutations that affect craniofacial development in the embryonic zebrafish. This work will provide noteworthy opportunities for understanding the etiology of vertebrate craniofacial abnormalities. For example, 33 alleles of 19 genetic loci that affect the development of the zebrafish jaw and branchial arches were revealed by the Tübingen screen and an additional 48 alleles of mutations at 34 loci have been found by the Boston screen. Work in a number of other labs has revealed a variety of mutations that affect the patterning of the vertebrate head and the development of neural crest derived structures. There are also a number of pigmentation mutants with pleiotropic effects on other neural crest derivatives that will undoubtedly provide important insights into the mechanisms of lineage diversification within the neural crest, and the regulation of development of various crest derivatives. Finally, a number of mutants affecting motility that express pleiotropic effects on neural crest derivatives suggest that the functions of a number of other genes in crest development will be uncovered when analysis of these mutations proceeds.

The special attributes of the zebrafish embryo (e.g., its transparency and rapid development outside of the mother or an eggshell, the ability to label individual cells with lineage tracers and to follow them in living embryos by time-lapse, DIC- fluorescence- and confocal microscopy), combined with the growing library of mutants that affect craniofacial structures or the neural

crest from which such structures are thought to arise, provide an outstanding resource for research on the mechanisms of normal development of specific craniofacial structures and other neural crest derivatives. Thus, the zebrafish embryo provides a tractable experimental system to analyze specification of developmentally distinct lineages derived from the neural crest and the development of jaw and branchial arch structures.

Although the zebrafish lack dentition in the mandibullar and maxillary bones, they do possess masticatory teeth. These teeth are embedded in cellular bone and project into the oral cavity at the level of the pharynx. They oppose the dorsal hyperkeratinized pad. The teeth are composed of acellular dentin and have anatomy similar to that seen in humans. The supporting bone is highly cellular, and shows evidence of rapid turnover. The maturation of these teeth occurs within the first week of development, and precedes ossification of the major chondroid bones. A screen has been initiated to identify mutants that affect the normal formation of these teeth. These mutations will represent interesting models for identifying precise molecular and cellular mechanisms of tooth development. The study of these mutants will be complimentary to work done in higher vertebrates, and will likely provide novel genes that would not have been identified in those higher species.

Up to Top