View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Vision Research 42 (2002) 527–533 www.elsevier.com/locate/visres Forward and reverse genetic approaches to the analysis of eye development in zebrafish Jarema J. Malicki a,*, Zac Pujic a, Christine Thisse b, Bernard Thisse b, Xiangyun Wei a a Department of Ophthalmology, Harvard Medical School, 243 Charles Street, Boston, MA 02110, USA b IGBMC/CNRS/INSERM/ULP, 1rue Laurent Fries, CU de Strasbourg, Cedex Illkirch 67404, France Received 14 May 2001; received in revised form 25 May 2001 Abstract The zebrafish has been established as a mainstream research system, largely due to the immense success of genetic screens. Over 2000 mutant alleles affecting zebrafish’s early development have been isolated in two large-scale morphological screens and several smaller efforts. So far, over 50 mutant strains display retinal defects and many more have been shown to affect the retinotectal projection. More recently, mutant isolation and characterization have been successfully followed by candidate and positional cloning of underlying genes. To supplement forward genetic mutational analysis, several reverse genetic techniques have also been developed. These recent advances, combined with the genome project, have established the zebrafish as one of the leading models for studies of visual system development. Ó 2002 Elsevier Science Ltd. All rights reserved. Keywords: Zebrafish; Mutagenesis; Screen; Positional cloning; Antisense; Knockdown 1. Forward genetic analysis embryos (Baier et al., 1996; Malicki et al., 1996; Li & Dowling, 1997). A rich repertoire of techniques is The hallmark of forward genetic analysis is a muta- currently available to search for mutant phenotypes genesis screen (Fig. 1). Although exceptionally success- affecting eye development. First, mutant selection ap- ful in invertebrates, this approach has been historically proaches have been developed based on a broad range of limited use in vertebrate model systems. Since its early of morphological, histochemical, and behavioral crite- days as a research organism, the appeal of the zebrafish ria. Similarly, several breeding schemes have been in use. has relied on its potential use in genetic screens. The Some of them take advantage of the ability to produce expectation that the zebrafish model will introduce zebrafish haploid or parthenogenetic diploid embryos screens as a standard tool of vertebrate genetics has on a scale sufficient to support a mutagenesis screen. certainly been fulfilled. Today, the repertoire of zebra- Finally, at least two types of mutagens, chemical and fish mutagenesis tools, breeding strategies and mutant retroviral, have been successfully applied on a large selection approaches has no match in any other verte- scale. The advantages and drawbacks of particular brate. screening tools have been discussed in several recent The complexity and scale of screening experiments in reviews (Malicki, 2000a,b). zebrafish vary greatly. The simplest screening protocols The progress of genetic analysis in zebrafish has relied on morphological inspection of mutant pheno- greatly benefited from a thorough description of the types with a dissecting microscope; the most involved wild-type embryonic and larval eye. Morphological ones employed, for example, sophisticated dye injection transformations leading to the appearance of the optic devices to label axonal projections of a specific cell class. cup have been described in detail (Schmitt & Dowling, While some screens inspected just several hundred adult 1994; Li, Joseph, & Easter, 2000b) and so has the individuals, others analyzed hundreds of thousands of timing of neurogenesis and cell class differentiation (Nawrocki, 1985; Hu& Easter, 1999). These studies revealed that one of the initial events in retinal neu- * Corresponding author. Fax: +1-617-573-4290. rogenesis is the appearance of postmitotic ganglion E-mail address: [email protected](J.J. Malicki). cells in the ‘‘ventral patch’’ area of the retina between 0042-6989/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved. PII: S0042-6989(01)00262-0 528 J.J. Malicki et al. / Vision Research 42 (2002) 527–533 Fig. 1. Forward genetic and reverse genetic approaches in zebrafish. Both approaches consist of selection and analysis steps. Positional and can- didate cloning are used to identify chemically induced alleles. In the case of retroviral mutagenesis, retrovirus itself is a molecular tag that facilitates cloning. Three general approaches to reverse genetic function analysis are available in zebrafish: overexpression of a constitutively active form, overexpression of a dominant negative form, and antisense oligonucleotide interference. 27 and 28 h postfertilization (hpf). This is followed 10 group of mutants is also particularly easy to relate to h later by the emergence of postmitotic interneurons human retinal disorders such as retinitis pigmentosa, which form the inner nuclear layer, and finally at 43 cone dystrophies or macular degenerations. The mutant hpf, the photoreceptor neurons in the outer nuclear phenotypes of elipsa (eli), fleer (flr)andoval (ovl) loci layer. Differentiation of ganglion and photoreceptor display both photoreceptor loss and kidney defects cells has received particular attention. Axons of the bearing a striking resemblance to a syndromic form first ganglion cells leave the eye by 36 hpf and reach of retinitis pigmentosa: the Senior–Loken syndrome the optic tectum by 48 hpf (Stuermer, 1988; Burrill (Warady, Cibis, Alon, Blowey, & Hellerstein, 1994; & Easter, 1995). Similar analysis determined that the Satran, Pierpont, & Dobyns, 1999). Because the zebra- photoreceptor outer segments become distinguishable fish larval retina is cone rich, it may be particularly first by 55 hpf while the photoreceptor synaptic termini useful as a model of macular degeneration in the human differentiate synaptic ribbons first between 60 and 62 retina. This is potentially of great importance as age- hpf (Schmitt & Dowling, 1999). Finally, it has been related forms of macular degeneration account for ap- established by behavioral criteria that the zebrafish proximately 50% of blindness in the western world retina becomes functional between 60 and 80 hpf (Klein et al., 1992; Vingerling et al., 1995; Stone et al., (Easter & Nicola, 1996). All these studies of wild-type 1999). development have been invaluable in the analysis of One particularly informative form of phenotypic mutant phenotypes. study in the zebrafish embryo is mosaic analysis. Mosaic Defects of nearly every aspect of eye development, zebrafish are generated by transplanting blastomeres from optic lobe morphogenesis to the differentiation of from one embryo to another (Ho & Kane, 1990; Halp- photoreceptor synaptic termini, have been found in ern, Ho, Walker, & Kimmel, 1993; Malicki & Driever, mutagenesis screens performed so far (Malicki et al., 1999; Doerre & Malicki, 2001). If the genotypes of host 1996; Brockerhoff, Dowling, & Hurley, 1998; Neuhauss and donor individuals differ, the resulting embryo will et al., 1999). Mutational analysis has been particularly consist of mutant and wild-type cells intermingled to- revealing in four areas: specification of the eye field, gether. In some embryos, mutant cells will be entirely neuronal patterning in the retina, differentiation of surrounded by wild-type tissue. In such cases, if a mu- photoreceptor cells, and retinotectal pathfinding. For tant defect is caused by defective cell–cell communica- example, over 30 mutant alleles produce defects in tion, interaction of mutant cells with wild-type tissue photoreceptor cell development and function alone will restore their normal phenotype. The value of mosaic (Malicki et al., 1996; Fadool, Brockerhoff, Hyatt, & analysis is that it allows one to identify instances of cell– Dowling, 1997; Becker, Burgess, Amsterdam, Allende, cell interactions. & Hopkins, 1998; Brockerhoff et al., 1998; Drummond Mosaic experiments have been performed on selected et al., 1998; Neuhauss et al., 1999). While some of them eye mutants and revealed several cell nonautonomous affect photoreceptor morphogenesis, others do not phenotypes. Interestingly, all four neuronal patterning produce any obvious structural abnormalities. This loci identified so far, glass onion (glo), mosaic eyes (moe), J.J. Malicki et al. / Vision Research 42 (2002) 527–533 529 nagie oko (nok), and oko meduzy (ome), display non- 2. Reverse genetics cellautonomous phenotypes in the retinal neuroepithe- lium (Malicki & Driever, 1999; Jensen, Walker, & Reverse genetic approaches have played a funda- Westerfield, 2001; Pujic & Malicki, 2001). At least in the mental role in the analysis of vertebrate eye develop- case of mosaic eyes, but possibly also in some of the ment. The majority of genes known to regulate retinal other mutants, aberrant retinal neuroepithelial pattern- neurogenesis were first identified using a variant of a ing is caused by defective cell–cell interactions with the reverse genetic approach. Reverse genetic analysis can retinal pigment epithelium. be subdivided into two phases: Molecular cloning of mutant genes is the most in- formative experimental approach to follow mutagenesis 1. Selection of candidate genes that are likely to play a screens. In this area the progress has been breathtaking. role in a given developmental process; Only five years ago, shortly after the completion of the 2. Functional analysis of candidates (Fig.