Proc. Natl. Acad. Sci. USA Vol. 92, pp. 10545-10549, November 1995 Neurobiology A behavioral screen for isolating zebrafish mutants with visual system defects SUSAN E. BROCKERHOFF*, JAMES B. HURLEYt, ULRIKE JANSSEN-BIENHOLDt, STEPHAN C. F. NEUHAUSS§, WOLFGANG DRIEVER§, AND JOHN E. DOWLING* *Department of Molecular and Cellular Biology, The Biological Laboratories, Harvard University, 16 Divinity Avenue, Cambridge, MA 02138; tHoward Hughes Medical Institute, Box 357370, University of Washington, Seattle, WA 98195; tDepartment of Neurobiology, University of Oldenburg, Postfach 2503, 26111 Oldenburg, Germany; and §Cardiovascular Research Center, Massachusetts General Hospital and Harvard Medical School, 13th Street, Building 149, Charlestown, MA 02129 Contributed by John E. Dowling, August 8, 1995 ABSTRACT Optokinetic and phototactic behaviors of ze- genetic dissection of the zebrafish visual system should be brafish larvae were examined for their usefulness in screening applicable to other vertebrates. for recessive defects in the visual system. The optokinetic Recently, two groups developed chemical mutagenesis pro- response can be reliably and rapidly detected in 5-day larvae, cedures and methods for efficiently growing large numbers of whereas the phototactic response of larvae is variable and not zebrafish (9-12). These procedures have made it possible to robust enough to be useful for screening. We therefore mea- conduct large-scale genetic screens in which zebrafish larvae sured optokinetic responses ofmutagenized larvae as a genetic from the third generation are analyzed for recessive mutations. screen for visual system defects. Third-generation larvae, Furthermore, a genetic linkage map in zebrafish is now representing 266 mutagenized genomes, were examined for available so mutant genes can be isolated by positional cloning abnormal optokinetic responses. Eighteen optokinetic-defective (13). mutants were identified and two mutants that did not show We first characterized two visual behaviors-phototaxis and obvious morphological defects, no optokinetic response a (noa) and optokinetic responses-in wild-type zebrafish larvae (3-19 response a were studied further. We days pf). Preliminary experiments on wild-type larvae (4) partial optokinetic (poa), suggested that both of these assays would be useful. We then recorded the electroretinogram (ERG) to determine whether analyzed the optokinetic responses of mutagenized larvae as a these two mutations affect the retina. The b-wave of noa larvae primary screen for detecting recessive defects in the visual was grossly abnormal, being delayed in onset and significantly system. As a secondary screen, we recorded the electroreti- reduced in amplitude. In contrast, the ERG waveform of poa nogram (ERG) from larvae 5-7 days pf to identify mutations larvae was normal, although the b-wave was reduced in ampli- that specifically affect the retina. We describe here the feasi- tude in bright light. Histologically, the retinas of noa and poa bility of this approach for identifying mutations affecting the larvae appeared normal. We conclude that noa larvae have a visual system and describe two mutants isolated on the basis of functional defect in the outer retina, whereas the outer retina of their abnormal optokinetic response. poa larvae is likely to be normal. MATERIALS AND METHODS Benzer (1) was the first to report that mutant Drosophila could be identified by their phototactic behavior. Subsequently, a Animals. AB strain zebrafish were obtained originally from number of nonphototactic mutants were found to have specific Oregon (14) and propagated at Harvard University by in- molecular defects in their photoreceptors (2). A phototaxis breeding. The AB strain maintained at the Massachusetts mutant that failed to respond to UV light, sevenless, lacks General Hospital was also originally obtained in Oregon and UV-sensitive photoreceptor cells (3); analysis of this mutant was then selected over several generations to be free of lethal has defined the role of cell-cell interactions in ommatidial mutations (9). In this study, zebrafish between 3 and 19 days development (for review, see ref. 6). pf are referred to as larvae. The water used for fish was Because there are significant differences between vertebrate reverse-osmosis distilled and then reconstituted for fish com- and invertebrate eyes, genetic analysis of the Drosophila eye patibility by addition of salts (2 g of Instant Ocean per gal; 1 has provided only limited information about the vertebrate gal = 3.785 liters) and vitamins (Fritz, Dallas). visual system. To apply a genetic analysis to the vertebrate Mutagenesis. The procedures for mutagenesis and for con- ducting crosses to identify recessive mutations in the third visual system, we have turned to zebrafish (Danio rerio). generation of mutagenized fish have been described (9). Zebrafish are visual and exhibit be- highly vision-dependent Briefly, male AB fish were mutagenized with N-ethyl-N- havior as early as 3 days postfertilization (pf) (4). They possess nitrosourea (Sigma) and outcrossed with wild-type females. four types of cones and are tetrachromatic. Short single cones The resulting F1 generation fish were crossed with each other contain a UV-sensitive photopigment, whereas long single or with wild-type fish to generate F2 families. Pairs of F2 cones contain a blue-sensitive pigment; a green-sensitive pig- siblings were then crossed to uncover recessive mutations in ment is in the short member of the double cones and a the F3 generation. The total number of genomes screened was red-sensitive pigment is in the long member of the double determined from the total number of F2 families and the extent cones (5). Rod photoreceptors are also present, so scotopic to which each F2 family was examined. The probability of and photopic vision can be analyzed in this organism (7). finding a mutation in a given F2 family depends on the number Furthermore, early eye morphogenesis and organization of the of crosses performed from that family and the number of zebrafish visual system are well characterized and similar to larvae examined from each cross. The number of mutagenized other vertebrates (8). Thus, information obtained from a genomes screened per family = (1 - 0.75s) x a, where S is the sum of fractions of crosses screened per family = x1 + X2 + .. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in Abbreviations: pf, postfertilization; ERG, electroretinogram; OKN, accordance with 18 U.S.C. §1734 solely to indicate this fact. optokinetic nystagmus. 10545 Downloaded by guest on September 27, 2021 10546 Neurobiology: Brockerhoff et al. Proc. Natl. Acad. Sci. USA 92 (1995) + x, and a is the number of mutagenized genomes crossed into a given F2 family (value of 1 or 2); x, is the fraction of cross n that was screened (1 - 0.751^), where L, is the number of if larvae screened from cross n. One N-ethyl-N-nitrosourea-induced allele of noam631 and IAI one allele ofpoam724 were isolated. Thirty-eight of 122 larvae : 1 examined from crosses between noa-carrying Go fish (F2 fish in original screen) showed no optokinetic response in white light. Go nba-carrying fish were outcrossed pairwise with AB a~- - fish and 13 noa carriers were identified in the F1 generation. The optokinetic response of 442 larvae from pairwise crosses of noa-carrying F1 fish were analyzed and 119 gave no opto- kinetic response. Of these 119 nonresponders, only 3 were noted as not having expanded melanophores (see Results). One pair of poa-carrying Go fish was identified. The opto- kinetic responses of 57 larvae from crosses between these two fish were analyzed and 10 larvae had an abnormal partial optokinetic response. All 10 of these larvae had expanded melanophores. Go poa-carrying fish were outcrossed with AB fish and four F1 carriers have been identified. The optokinetic ..... responses of 26 larvae were analyzed from crosses between FIG. 1. Apparatus for measuring optokinetic responses. D, drum; these F1 fish; 6 larvae had a partial optokinetic response and F, fiber optic; L, monochromatic light source; M, microscope; V, all 6 were darker than wild-type larvae. Additionalpoa larvae videocamera. were selected from additional crosses between the poa- carrying F1 fish based on their unusual swimming behavior and drum was rotated in two directions and the eye movements darker pigmentation (see Results). Of "50 larvae selected in were analyzed by watching the larva through the microscope. this way, all had a partial abnormal optokinetic response. A response was considered positive if a single smooth pursuit Finally, the optokinetic responses of 32 larvae from a cross and saccade eye movement in the proper direction was ob- between noa- and poa-carrying fish were analyzed and all served after starting drum rotation in each direction (see showed a normal optokinetic response, suggesting that the noa Results). A larva was considered abnormal if it showed no eye and poa mutations are in different genes. Furthermore, the movements at all or if the eye movements were unusual, such above data suggest that both mutations are recessive. as too fast, too small, or in the wrong direction. On average, Behavioral Assays. A useful behavioral screen must be 1 min was required to analyze each larva including time spent reliable and fast because recessive defects are detected in only arranging it. Thus, we could screen '60 larvae per h or '500 25% of a given population and thus large numbers of animals larvae per day. Optimally, at least 10 larvae were examined must be analyzed. Also, the assay should be conducted on from each cross and larvae from at least 6 crosses were young fish so that the cost and labor associated with raising examined from each line. Two hundred and forty-one genomes many larvae is minimal. were screened with white light several log units above thresh- Phototaxis. The phototactic response of AB larvae was old. The stripes were illuminated from above and below with measured using a 10.5 x 3 x 4 cm (length x width x height) a fiber optic light source.