Fish strains, mutagenesis, and screening.
We used fish from the TL strain for mutagenesis and crossed them to fish from the WIK strain for linkage mapping (see below). Embryos and larval fish were kept in E3 solution (egg water): 5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl2, and 0.33 mM MgSO4 supplemented with 1:107 w/v methylene blue. Mutations in the zebrafish genome were induced in the spermatogonia of 41 founder males (F0) by three to five treatments with ENU (3 mM for 1 h each, at weekly intervals) and bred to homozygosity over two generations, as previously described [31, 32]. Details of the screen statistics and the specific-locus test used to measure the mutation rate are given in Results.
Genetic linkage mapping.
We used microsatellite-based linkage mapping methods to locate the mutation in the zebrafish genome . Heterozygous carriers of the mutation (in the TL background) were crossed to the highly polymorphic WIK strain. Carrier pairs were identified from this hybrid progeny and mated repeatedly. Clutches were sorted for mutants and nonmutant siblings using behavioral assays (often a combination of OMR to quickly enrich for mutants, followed by OKR of the enriched population for unambiguous identification of mutants). Bulk-segregant analysis was performed using pooled DNA from siblings and mutants. This method involves PCR with a set of 192 polymorphic simple-sequence repeat markers (oligonucleotide primers targeted to unique sequences flanking dinucleotide repeats of variable length ). The markers were selected to cover the entire zebrafish genome (25 linkage groups) at roughly even intervals (K. F.-B., unpublished data). Candidate markers showing co-segregation with the mutant pool were confirmed by PCR of single-fish DNA. Map position was further verified by demonstration of linkage to additional markers located in the presumed chromosomal region.
We completed classical complementation crosses among all mutants with similar phenotypes (Table 1) or with reported mutants with similar phenotypes or similar map position (if available). Heterozygous nof carriers were obtained from S. Brockerhoff (University of Washington). Heterozygous bel carriers were obtained from C. B. Chien (University of Utah). Complementation tests for nok were carried out by S. Horne (UCSF). Complementation tests for bru were carried out by J. Malicki (Harvard).
Assessment of VBA.
Fish were kept on the fluorescent illuminator (950 cd/m2) for at least 20 min to light-adapt. The pigmentation of the fish was visually scored in four grades to determine the VBA score, with 1 = normal (WT), 0.7 = slightly dark, 0.3 = intermediate dark, and 0 = strongly dark. In this scoring system, the previously discovered, RGC-deficient lakritz mutant scored 0  and served as a reference to calibrate the index. The VBA score for variably dark mutants was estimated by averaging over at least ten individuals.
Recording of the OMR.
The OMR assay was conducted as described previously . Visual stimuli were displayed on a flat-screen CRT monitor that faced upward. The stimuli, which consisted of moving sinusoidal gratings, were generated in MATLAB (MathWorks, Natick, Massachusetts, United States), using the Psychophysics Toolbox extensions (http://psychtoolbox.org). The gamma function of the CRT was measured using a Minolta LS-100 (Tokyo, Japan) light meter, and corrected using MATLAB. The images of the fish before and after each stimulus were captured by a digital still camera (Nikon CoolPix [Tokyo, Japan]), which was triggered by MATLAB using a set of serial commands. These images were downloaded from the camera offline and analyzed using custom macros in Object-Image (http://simon.bio.uva.nl/object-image.html). Ten to 40 larvae (routinely 25) were placed in custom-built acrylic tanks, or “racetracks,” which allowed the larvae to swim in only two directions. Twelve racetracks were placed side by side on the monitor. After subtracting two consecutive images to remove the background, the position of each fish was determined by using the Analyze Particles function of Object-Image. The average position of the fish in each tank before a stimulus was then subtracted from the average position after 30 s of exposure to a standard motion stimulus. The OMR index of a recessive mutant was calculated for stimuli of 100% and 75% contrast by measuring the average distance swum by the 25% weakest responders in a clutch, divided by the distance swum by the 75% best responders. Each stimulus contrast and stimulus direction were repeated four times and the average OMR score was calculated offline.
Recording of the OKR.
The OKR assay was conducted as described previously . An animation of sine-wave gratings was projected on the internal wall of a drum (height, 6 cm; inner diameter, 5.6 cm), using an LCD projector (InFocus LP755 [Wilsonville, Oregon, United States]) . To focus the image at close distance, a wide-angle conversion lens (Kenko VC-050Hi [Tokyo, Japan]), a close-up lens (King CU+1 [Tokyo, Japan]), and a neutral density filter (Hoya ND4 [Tokyo, Japan]) were placed in front of the projector. Twelve zebrafish larvae were immobilized in 2.5% methylcellulose in E3 egg water with their dorsal sides up in the inverted lid of a 3.5-cm diameter petri dish and placed into the center of the drum. The fish were imaged using a dissecting microscope (Nikon SMZ-800) and a CCD camera (Cohu MOD8215–1300 [Tokyo, Japan]) to observe horizontal eye movements. Sinewave gratings with a spatial frequency of 20° per cycle moving at 10°/s were used. Image-J (http://www.rsb.info.nih.gov/ij/) was used for both stimulus generation and image analysis. Images were captured via an LG-3 video capture board (Scion; http://www.scioncorp.com) at two frames per second with Scion Java Package 1.0 for Image-J Windows. A custom-programmed Image-J plug-in (A. M., unpublished data) was used to calculate the changes in eye angles. The OKR index of a mutant was defined here as the saccade number per minute divided by the saccade number per minute observed in WT.
Surrogate light adaptation assay.
The dynamics of OKR in response to sudden changes in illumination was measured as described previously . Fish larvae were put in the dark for 45 min to let them dark-adapt, then subjected to the OKR recording at 2, 8, 15, and 30 min after return to a bright environment (2,400 cd/m2 underneath the larvae; 400–600 cd/m2 at the internal drum wall, where the visual stimulus was projected).
Recording of spontaneous swimming activity.
Spontaneous swimming activity was measured as described . Larvae at 7 dpf were tested in groups of six fish in a rectangular compartment (3 cm × 7.5 cm) of a four-well, clear acrylic plate (12.8 cm × 7.7 cm [Nunc, Roskilde, Denmark]). Fish images were captured by a digital camcorder (Sony TRV-9 [San Diego, California, United States]) at a rate of 0.5 Hz for 20 min in Adobe Premiere. Recorded movies were analyzed using Image-J. Each frame was subtracted (pixel by pixel) from the previous frame to extract the fish that moved during the inter-frame interval. Spontaneous activity was quantified by counting the number of moving fish across all frames. The SSA index was calculated by dividing the number of movement episodes seen in mutants by that seen in WT siblings.
Histology and immunohistochemistry.
Zebrafish larvae were fixed in 4% paraformaldehyde in PBS at 4 °C for 2–16 h, transferred to 30% sucrose in PBS plus 0.02% NaN3 for 16 h or more, mounted in O. C. T. Compound (Sakura Finetek USA, Torrance, California, United States), frozen, and sectioned at 10–12 μm. In some cases, after fixation, the sample was dehydrated in an ethanol series followed by xylene, embedded in paraffin, and sectioned at 6 μm. For immunohistochemistry, the section was incubated with primary antibodies, fluorescent dye-conjugated secondary antibodies (Molecular Probes, Eugene, Oregon, United States), counterstained with 4′,6-diamidino-2-phenylindole (DAPI), and mounted with Fluoromount-G (Southern Biotechnology Associates, Birmingham, Alabama, United States).
Fluorescent axon tracing of the optic tract.
Zebrafish larvae were fixed in 4% paraformaldehyde in half-strength PBS at 4 °C overnight. The fish eye was injected with 1% 1,1′-dioctadecyl-3,3,3′,3′- tetramethylindocarbocyanine (DiI), DiD, or DiO dissolved in chloroform . Fluorescent images were observed with a confocal laser-scanning microscope (BioRad MRC 1024 [Hercules, California, United States] or Zeiss LSM [Oberkochen, Germany]).