Two diminutive fish, the zebrafish and Medaka, have recently been swimming into view as important new model organisms for biomedical research. Although both species have been around as experimental models for many years, in the last decade or so they have emerged as a major new players in developmental genetics research, and more recently as models for human disease. The explosive growth of fish as research model organisms has been fueled at least in part by the development of new, powerful, genome-based tools and resources for scientific research in fish, some of which are discussed in the eight reviews in this issue.
Although the zebrafish had been used as a research model since before the end of the Second World War, mainly for teratology and toxicology studies (reviewed in ), the use of the zebrafish as a genetic and experimental embryologic model organism really began with the work of George Streisinger in the 1960s. Streisinger had participated, together with pioneering molecular geneticists, such as Salvatore Luria, Max Delbruck and Seymour Benzer, in laying the foundations for modern molecular biology. After working out the genetic code and other ‘ground rules’ primarily using bacteria and bacteriophage, many members of this group sought new model organisms in which to apply the tools of genetic analysis to more complex problems, such as the development of the body plan during embryogenesis and determination of the functional anatomy of the nervous system. Sydney Brenner and Seymour Benzer went on to carry out their famously groundbreaking work on the nematode Caenorhabditis elegans and the fruit fly Drosophila melanogaster, respectively. Streisinger set his sights on a perhaps even more ambitious target—genetic dissection of development in a vertebrate—and set about looking for an appropriate model organism for this purpose.
He settled on the zebrafish, a vertebrate with high fecundity and an optically clear and externally developing embryo, and together with colleagues at the University of Oregon Institute of Neuroscience he began to develop techniques for genetic manipulation and recovery of mutants in this species. They worked out methods for generation of haploid embryos derived entirely from the maternal genome by fertilization with UV-irradiated inactivated sperm, and then succeeded in producing gynogenetic diploid animals, homozygous diploid ‘clonal’ animals that permitted recessive phenotypes to be revealed in a single generation . This feat made something of a splash for Streisinger in the popular media at the time, but importantly also got the attention of many prominent scientists in other disciplines, notably Drosophila researchers who had been having dramatic success in their efforts to carry out genetic analysis of fly development including future Nobel laureate Christianne Nusslein-Vollhard (see below). In addition to working out ways to easily uncover the phenotypes of recessive mutations, methods were derived for efficiently inducing novel mutations by radiation  or chemical  mutagenesis, and for carrying out genetic mapping in the zebrafish . As the genetic foundations were being laid, parallel efforts were leading to development of some of the experimental embryology techniques that are now part of the standard zebrafish ‘toolkit’, including cell labeling methods for fate mapping and lineage tracing, and transplantation methods for testing cell autonomy of gene function [6–8].
Even after the untimely death of Streisinger in 1984, efforts to promote the fish as an important new model for genetic dissection of vertebrate development continued and accelerated at the University of Oregon and elsewhere, coming to fruition with the initiation and then completion and publication of the results of the first large-scale forward genetic screens for developmental mutations ever accomplished in a vertebrate organism. These ‘big screens’ were carried out by Christianne Nusslein-Vollhard and colleagues in Tubingen (Germany) and by Wolfgang Driever, Mark Fishman and colleagues in Boston, MA, USA, and their results were reported in a single issue of Development published in December 1996 (Volume 123, ‘The Zebrafish Issue’). Thousands of new mutants were generated and characterized, including mutants with specific defects in almost every imaginable embryonic and early larval developmental process. The mutants obtained from these and subsequent genetic screens have provided an unparalleled scientific resource that is still not close to being fully exploited. However, the isolation of all of these mutants meant the beginning of the hard work of determining what genes were actually mutated. Cloning the defective loci for all of these new mutants in a timely fashion necessitated acquiring vastly improved information on the zebrafish genome. Indeed, a great deal of effort has gone into assembling important genomic resources, such as extensive EST libraries, genetic and physical maps, and ultimately, a complete sequence of the genome.
The zebrafish genome is estimated at ~1.6 Gb (the Medaka genome is approximately half this size). A genetic map of the zebrafish containing many thousand markers has been assembled. The genetic map is anchored on a framework physical map of the genome assembled from fingerprinted BAC clones and other linked DNA assemblies, as well as radiation hybrid maps. An effort to sequence the zebrafish genome was initiated by the Wellcome Trust Sanger Institute in the Spring of 2001. The strategy employed is to sequence mapped BAC and PAC clones, complementing this with whole genome shotgun sequence from smaller insert clone libraries. Although ~1.45 Gb of ‘finished’ sequence is currently available from this project, accurate assembly of the sequenced DNA contigs into larger assemblies has been very slow and difficult to achieve despite considerable time and effort. This is in part due to the highly outbred nature of most zebrafish laboratory populations and the fact that the libraries from which most of the initial genome sequencing was carried out contained high levels of haplotypic variation. Currently, most sequencing is being done from a newer library constructed from a single ‘doubled haploid’ fish (obtained using one of Streisinger's methods for generating gynogenetic diploids), and it is hoped that use of this library will aid in linking together some of the more difficult to assemble portions of the genome.
The still-incomplete genome sequence information available has already greatly facilitated the cloning of mutated genes and genome-scale analysis of zebrafish development, and a variety of new tools for manipulating and studying the genome have been devised to further capitalize on the advances to date. The eight articles assembled in this issue of Briefings in Functional Genomics and Proteomics describe some of the tools and methods developed for genome analysis in zebrafish and Medaka, and show how these tools and resources are being brought to bear on novel scientific research topics.
Kobayashi and Takeda describe the current status of the Medaka genome project. Through a combination of historical advantages, foresight and some measure of luck, Medaka researchers have been able to avoid many of the problems encountered in the zebrafish genome project, as noted above. As a result, despite beginning work much later the Medaka genome is currently in a comparatively somewhat more ‘finished’ state. This article details the strategy used to sequence the Medaka genome and discusses a few of the insights into gene evolution that have come from examining the Medaka genome.
The ability to carry out large-scale forward-genetic analysis has been one of the major strengths of both the zebrafish and Medaka models. As described above, in some of the earliest zebrafish work methods were established for efficiently inducing mutations using ‘classical’ mutagens, such as gamma rays and ethyl nitrosourea (ENU). However, even with the availability of a large amount of genomic sequence data, cloning of loci mutated by these traditional mutagens is still difficult and time-consuming. Insertional mutagenesis has been used in non-vertebrate model organisms to generate mutants marked by a ‘molecular tag’, facilitating rapid recovery of the mutated gene, and has recently been applied to vertebrates, such as mice and fish as well. Jao, Maddison, Chen and Burgess review current technology for using retroviruses as insertional mutagens for forward-genetic analysis and as vectors for reverse genetic approaches, such as gene delivery and enhancer or gene trapping. Transposable elements have historically made very important contributions to genetic studies in plants, such as Zea mays, and invertebrates, such as Drosophila melanogaster, and more recently transposons have also found important applications in gene delivery and gene discovery in vertebrates. Ni, Clark, Fahrenkrug and Ekker discuss different classes of transposable element vector systems that have been tested in vertebrates and the relative advantages and disadvantages of some of these systems in fish.
While piscene model systems have been exploited extensively for forward genetic analysis using both traditional non-insertional and insertional mutagenesis methods, until recently there were no viable options available for reverse-genetic targeting of specific genes. ES-cell based methods for introducing targeted alterations at specific genetic loci are still not available in the fish, although there has been some success in obtaining embryonic stem cell—or primordial germ cell-like lines from zebrafish (for example [9, 10]). However, recent advances in two alternative methodologies have made it possible to obtain mutations in particular pre-selected genes of interest in fish. Moens, Donn, Wolf-Saxon and Ma describe the use of Targeting Induced Local Lesions in Genomes (TILLING), a method to identify carriers of mutations in specific genes by high-throughput PCR-based screening of large populations of chemically mutagenized animals. The next review by Amacher describes very recent work demonstrating that zinc-finger nucleases (ZFNs) custom-designed to target specific genes can be used to create mutations with very high efficiency and specificity [11, 12]. Together these methods have now made it possible to readily obtain mutations in selected genes in fish. Eventually, large-scale efforts underway to establish comprehensive ‘libraries’ of insertionally tagged mutants in zebrafish genes will make it unnecessary to screen for mutations in most targets—one will simply be able to order the mutant from a stock center or commercial supplier. However, these efforts are still many years from completion. In addition, the ability to carry out site-specific modification of endogenous genes by homologous recombination has not yet been achieved in the fish, although ZFN technology shows considerable promise in this regard.
The feasibility of moderate- to high-thoughput screening approaches combined with the availability of genomic information has resulted in a variety of creative and powerful applications for zebrafish and Medaka research, some of which are discussed in the remaining reviews of this issue. Pashos, Kague and Fisher discuss findings from comparative genomics studies using fish to examine the function and evolution of cis-regulatory sequences. The zebrafish has proven to be a highly useful system in which to rapidly identify and test potential noncoding regulatory elements by reintroduction of these elements in vivo on transposon vectors with minimal promoters . Some of the newest and most unique research in the zebrafish has been in the area of behavioral genetics. A number of fish researchers have already begun to use mutagenesis screens for behavioral mutants to dissect patterns of connectivity in the brain, and study how these generate and monitor behaviors. Burgess and Granato review some of the considerations and challenges encountered in designing and carrying out screens to identify and characterize mutants affecting complex behaviors in fish. The zebrafish is highly amenable not only to genetic screening for mutants, but also to large-scale phenotype-based screening with structurally diverse small-molecule ‘chemical libraries’. In the final review, Kokel and Peterson discuss how zebrafish can be used for chemical behavioral screens to potentially identify new classes of psychiatric medicines.
Together, the reviews in this issue highlight just a few of the many ways in which the utility of zebrafish and Medaka as model organisms is being expanded though increasing availability of both genomic data resources, new tools for genome manipulation and new avenues of research exploiting the advantages of fish. It is to be expected that as these resources continue to be deepened and broadened in scope the usefulness of these models will only continue to improve.