SEARCH

SEARCH BY CITATION

Keywords:

  • behavioral visual test;
  • mutation;
  • visual system development;
  • retinal degeneration;
  • olfactoretinal centrifugal pathway;
  • zebrafish

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION
  5. DOMINANT VISUAL SYSTEM MUTATIONS
  6. PERSPECTIVES
  7. Acknowledgements
  8. REFERENCES

Zebrafish are a promising model for behavioral and genetic studies of vertebrate visual system development and retinal degeneration. In the past few years, numerous studies on zebrafish vision have been published. While most of the studies focus on the molecular and cellular characterization of mutations that disrupt zebrafish visual system structure in early development, others examine the mechanisms that underlie inherited visual system disorders in adults. Behavioral assays, along with morphologic and electrophysiological methods, are powerful tools for functional analyses of zebrafish visual development and performance. © 2001 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION
  5. DOMINANT VISUAL SYSTEM MUTATIONS
  6. PERSPECTIVES
  7. Acknowledgements
  8. REFERENCES

Dominant mutations that cause retinal degeneration have been found in many species including flies, mice, rats, dogs, and man (Dowling and Sidman, 1962; Benzer, 1967; Harris and Stark, 1977; Hafen et al., 1987; Steele and O'Tousa, 1990; Dryja et al., 1990, 1991; Olsson et al., 1992; Suber et al., 1993; Berson, 1993; Dolph et al., 1993; Roof et al., 1994; Farber, 1995). In humans, approximately 40% of retinitis pigmentosa (RP, characterized by progressive photoreceptor cell degeneration) cases are shown to be dominantly inherited (Bunker et al., 1984; Enevoldson et al., 1994; Kemp et al., 1994). Patients suffering from RP lose their peripheral and night vision in adolescence and become completely blind between the ages of 30 and 60 years (Berson, 1993). Mutations that cause dominant RP have been identified in several photoreceptor cell–specific genes, for example, in genes that encode rhodopsin and peripherin (Sanberg et al., 1990; Schmidt and Berson, 1990; Stone, 1998). However, it has been estimated that only approximately half of dominant RP cases can be attributed to mutations discovered thus far (Berson, 1993). The genetic basis for the rest of RP cases, on the other hand, remains completely unknown.

Zebrafish (Danio rerio) have recently become a mainstream model for vertebrate vision research and have been used for mutational screening of visual system genes (Nawrocki, 1985; Kljavin, 1987; Larison and Bremiller, 1990; Burrill and Easter, 1994, 1995; Raymond et al., 1995; Schmitt and Dowling, 1994, 1996, 1999; Brockerhoff et al., 1995; Malicki et al., 1996; Fulwiler et al., 1997; Li and Dowling, 1997; Becker et al., 1998; Hu and Easter, 1999; Marcus et al., 1999; Li et al., 2000a,b; Rick et al., 2000; Malicki, 2000; Baier, 2000; Li, 2001). Screening for zebrafish visual system mutations is carried out mostly by using three methods, including morphologic, dye-trace, and behavioral tests (Brockerhoff et al., 1995; Baier et al., 1996; Malicki et al., 1996; Li and Dowling, 1997). Behavioral tests are especially useful in studying zebrafish visual performance and the inherited visual disorders affecting adults. In this article, I review briefly some of the behavioral assays, the usefulness of these assays in mutational screening, and the major findings from the characterization of two dominant mutations, night blindness a (nba) and night blindness b (nbb).

BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION
  5. DOMINANT VISUAL SYSTEM MUTATIONS
  6. PERSPECTIVES
  7. Acknowledgements
  8. REFERENCES

Behavioral tests have long been used to assess visual development and to screen for visual system mutations (Benzer, 1967; Harris and Stark, 1977). So far, five behavioral visual tests based on the visual startle response, phototactic response, optokinetic response (OKR), optomotor response (OMR), and escape response have been used to analyze zebrafish visual system development and function (Kimmel et al., 1974; Clark, 1981; Brockerhoff et al., 1995; Easter and Nicola, 1996; Li and Dowling, 1997; Neuhauss et al., 1999; Bilotta, 2000). Some of these assays are used for examining the visual development of larvae, whereas others are more suitable for evaluating visual system function in adults. I review briefly each of them (Table 1; Fig. 1).

Table 1. Behavioral Assays for Measuring Visual System Functionsa
Behavioral assayAgeReliabilityMutant screen
  • a

    dpf, days postfertilization; OKR, optokinetic response; OMR, optomotor response.

Start response<4 dpf+No
Phototaxis>3 dpf+/−No
OKR3–7 dpf+Yes
OMR>4 dpf+Yes
Escape response>2 months+Yes
thumbnail image

Figure 1. Behavioral assays based on zebrafish visual startle response (A), phototactic response (B), optokinetic response (C), optomotor response (D), and escape response (E). Shadows in A and B indicate dark illumination. Vertical stripes and the black segment, shown in C–E, are moved from right to left (arrows).

Download figure to PowerPoint

Startle Response

Kimmel et al. (1974) reported that the visual startle response is useful for analyzing zebrafish visual development. In response to a sudden decrease in illumination, for example, the fish initiate a body movement to flee from the shadow, or the “looming predator.” This visual startle response is not observed in blind fish nor in zebrafish in which the retinal photoreceptor cells have not been developed. In a recent study using the startle response assay, Easter and Nicola (1996) evaluated systematically the time course of larval zebrafish visual development. They reported that the visual startle response is not seen before 68 hr postfertilization (hpf), although at this stage of development the mechanical-induced body movement is observed. The first visual startle response is seen between 68 and 70 hpf, when synapses between retinal neurons start to form. Previous anatomic studies have shown that, by 68–70 hpf, the outer segments of retinal photoreceptor cell are developed, and synaptic ribbons in both outer and inner retinal terminals are formed (Branchek and Bremiller, 1984; Schmitt and Dowling, 1999).

The visual startle response assay can be used to test simple vision, for example, to distinguish light from darkness, or vice versa. Complex visual functions such as form vision, on the other hand, cannot be evaluated by using this assay. Furthermore, the startle response test can only be used to measure early visual system development, i.e., before 96 hpf. Older embryos often display spontaneous movements, thereby making it difficult to judge whether the movement is solely elicited by an abrupt change in illumination (Kimmel et al., 1974). This test has not been used in the behavioral screening of visual system mutations in zebrafish.

Phototactic Response

This assay tests the phototaxis of zebrafish (Brockerhoff et al., 1995). The test is conducted in a rectangular acrylic box with a sliding partition separating two chambers. Both chambers are equally illuminated. At first, the fish are allowed to freely disseminate between the two chambers. The number of fish in each chamber is counted. Then, one of the two chambers is covered, and the other remains illuminated. After a few minutes, the partition is inserted, and the number of the fish in each chamber is recounted. Normally, the fish will swim toward the illuminated chamber. The experiment can also be done in another way. For example, the fish are first constrained in the dark chamber. Then, the partition is removed, and the fish are allowed to swim freely to the other brightly illuminated chamber. After a few minutes, the partition is reinserted, and the number of fish in each chamber is recounted. It is expected that the fish that have normal vision will swim toward the light source, whereas those with visual system defects will be unable to do so. Both larval and adult zebrafish can be tested by using this assay.

The phototaxis assay is somewhat useful in assessing the light response of zebrafish. However, it is not robust enough to distinguish the visual system mutants from normal animals. For example, after 1–5 min of free distribution, only 85% of the fish migrate to the illuminated chamber (Brockerhoff et al., 1995). The phototaxis assay has not been used in mutational screening of visual system genes.

Optokinetic Response

The optokinetic response (OKR) assay, a commonly used visual test in both clinical and basic research, measures eye movements in response to moving objects across the visual field (Clark, 1981). OKR is characterized by two components: a smooth pursuant eye movement in the direction of the moving objects followed by a rapid saccadic eye movement when the images leave the view. In contrast to the startle and phototactic responses, OKR is used for measuring sophisticated visual performance, such as form vision, of developing zebrafish.

The apparatus used for the OKR test consists of a plastic Petri dish surrounded by a rotating drum covered with black and white stripes. The fish are immobilized in the Petri dish by exposure to low concentrations of methylcellulose. This prevents the fish from swimming around but allows their eyes to move freely. The eye movements of the fish are observed by using a video camera. By using the OKR test, Clark (1981) evaluated the development of both cone and rod systems of larval zebrafish. More recently, using a similar approach, Easter and Nicola (1996) studied the time course of the visual development of larval zebrafish. They reported that form vision in zebrafish develops between 73 and 80 hpf, when retinal neurons are fully differentiated and synaptically connected.

The OKR test has been used for mutational screening of visual system genes (Clark, 1981; Brockerhoff et al., 1995; Neuhauss et al., 1999). By screening 266 mutagenized genomes, Brockerhoff et al. (1995, 1997, 1998) identified several mutants that have normal eye morphology but fail to show OKRs. Some of the mutants are completely blind (i.e., noa, nrc) and some are partially blind (poa, pob). Most of the mutations cause abnormalities in the outer retina (Brockerhoff et al., 1998; Allwardt et al., 2001; Van Epps et al., 2001). By using a similar approach, Neuhauss et al. (1999) screened 340 previously identified loci (Haffter et al., 1996). Mutations that cause defects at various sites of the visual pathway were discovered, for example, mutations that affect the development of the lens (bum), retina (lak, drp, nir), and ganglion cell axon pathfinding (gup, sly, pic).

The OKR test is excellent for measuring zebrafish visual function between the ages of 3–7 days postfertilization (dpf). Fish outside this age range, on the other hand, cannot be tested by using this assay. This is because larvae younger than 3 dpf do not display OKR, whereas embryos older than 7 dpf cannot be readily restrained in the Petri dish (Clark, 1981; Brockerhoff et al., 1995).

Optomotor Response (OMR)

The optomotor response (OMR) assay also examines motion detection. The classic OMR apparatus is similar to the one used for the OKR test, except that the Petri dish is replaced by a circular container filled with water (Clark, 1981; Bilotta, 2000). The fish are allowed to swim freely in the container. Normally, the fish swim in the direction of the moving objects, for example, the rotating black and white stripes.

The OMR apparatus has been recently modified for mutational screening (Neuhauss et al., 1999). Instead of using the circular container and rotating stripes, Neuhauss et al. (1999) tested the OMR of larval zebrafish in elongated rectangular chambers. Computer-generated images (black and white stripes) are presented below the chambers. The fish are first placed in one end of the chamber. A few minutes after the stripes are presented, the distribution of the fish in the chamber is determined. Wild-type fish swim with perceived motion and congregate in the other end of the chamber. Individuals that cannot detect the moving stripes due to visual defects, on the other hand, stay where they were originally placed or swim only randomly in the chamber. In a recent study, Neuhauss et al. (1999) tested the OMR of 411 previously identified loci (Haffter et al., 1996). They uncovered several mutations that cause eye-specific defects, such as pic, blu, and mao.

The OMR test has been used for testing visual function in both larval and adult zebrafish (Clark, 1981, Easter and Nicola, 1996; Neuhauss et al., 1999; Bilotta, 2000). However, the usefulness of OMR tests for mutational screening in adult animals has not been explored. Not every adult fish swims in the direction of the stripes; sometimes they swim against the direction of the rotating stripes (Maaswinkel and Li, unpublished data).

Escape Response

A behavioral test based on the visually mediated escape response has been developed that permits quantitative analyses of adult zebrafish visual sensitivity (Li and Dowling, 1997). Normally zebrafish swim slowly in circles when confined to a circular container. The fish swim randomly in either direction, clockwise or counterclockwise. However, when challenged by a threatening object, such as a black segment that rotates outside the container, the fish display a robust escape response. As soon as the black segment comes into view, the fish immediately turn and rapidly swim away from the black segment. In two 1-min trials, for example, in one trial the black segment rotates clockwise and in the other it rotates counterclockwise, the fish encounter the black segment approximately 50 times. On average, the fish show an escape response 40–45 times.

The escape response test is robust, so that within 5–10 sec after the black segment is rotated, a judgment as to whether the fish “see” the black segment can be made. By varying the light intensity illuminating the rotating black segment, the light threshold required to evoke an escape response is measured. By using this assay, Li and Dowling (1997, 1998, 2000a,b) evaluated a variety of visual aspects in adult zebrafish, for example, the time course of cone and rod dark adaptation, the absolute visual sensitivity, the circadian control of visual function, and the effect of dopamine on vision. They also screened the F1 generation of mutagenized zebrafish for dominant visual system genes. Two of these mutations (nba and nbb) are discussed below.

This escape response evoked by the rotating black segment does not habituate. It is measured in both light- and dark-adapted animals and in experiments carried out during the day or night. This test is suitable for measuring visual sensitivity in adult animals after 2–3 months of age.

DOMINANT VISUAL SYSTEM MUTATIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION
  5. DOMINANT VISUAL SYSTEM MUTATIONS
  6. PERSPECTIVES
  7. Acknowledgements
  8. REFERENCES

By using the behavioral assay based on the visually mediated escape response, Li and Dowling (1997) screened 345 F1 generation fish for dominant mutations that cause visual system defects. During the screening, the light that illuminates the rotating black segment is set at a level approximately 1.0 log unit above the threshold level of wild-type fish. Under this level of illumination, animals that show the escape response are considered normal, whereas the individuals that fail to show escape responses are isolated as night-blind. The night-blind fish are rescreened, outcrossed, and further characterized by using histologic and physiological assays.

The nba Mutation

Heterozygous nba mutants (nba+/−) have normal vision before 3 months of age, but thereafter, they become night blind. By 10 months of age, the visual sensitivity of nba+/− fish decreases to levels approximately 3.0 log units, or 1,000-fold, below those of wild-type fish. The nba+/− mutants undergo slow photoreceptor cell degeneration involving initially the rods but later the cones as well. Interestingly, degeneration of the photoreceptor cells in nba+/− fish is not uniform over the retina but tends to be patchy. In the central retina, for example, areas where rods completely degenerated are interspersed between areas in which significant numbers of rods remained (Fig. 2).

thumbnail image

Figure 2. Histologic sections of wild-type (A) and nba+/− (B–D) retinas at 13 months of age. A: A transversal section of a wild-type retina. Note the length of the cone (c) and rod (r) outer segments. B–D: Central retinal sections from a single nba+/− retina. Note the variability of cone and rod degeneration. In some areas, only rods are affected (B). Large lipid droplets in the pigment epithelium are observed (arrow in B). In other areas, rods are completely degenerated (C) or both rods and cones are affected (D). in, inner nuclear layer; ip, inner plexiform layer. (Li and Dowling 1997; printed with permission from the publisher) Scale bar in D = 100 μm in A, 80 μm in B–D.

Download figure to PowerPoint

Whereas heterozygous nba fish show progressive retinal degeneration, homozygous nba (nba−/−) fish display an early onset of neural degeneration in the central nervous system (CNS) that begins approximately 2 dpf. By 2.5 dpf, the nba−/− embryos can be readily distinguished from normal siblings by their small eyes. In nba−/−, extensive cell death is observed in the retina, as well as in the tectum where retinal ganglion cell axons terminate. By 3.5 dpf, the nba−/− retinas and the tectum are virtually destroyed. By 5 dpf, homozygous nba fish die.

Li and Dowling (1997) suggest that nba represents a novel type of dominant retinal degeneration that has not been described previously. This conclusion is based on two observations. First, nba is not a photoreceptor cell–specific gene. In several dominant retinal degeneration cases, such as in retinal degeneration slow (rds) mice, the degeneration is restricted to the photoreceptor cells. In heterozygous rds mice, for example, at 1 year of age, approximately half of the photoreceptor cells are degenerated. When rds is bred to homozygosity, the photoreceptor cell outer segments do not form properly, and by 1 year of age the photoreceptor cells have completely degenerated. Nevertheless, the degeneration is restricted to the photoreceptor cells (Sanyal et al., 1984; van Veen et al., 1988). This is not the case for the nba mutation. In nba+/− fish, not only the photoreceptors but also the inner retinal neurons are degenerated (Li and Dowling, 1997). Second, nba causes homozygous lethality. In nba−/− embryos, the degeneration is found all over the CNS. In addition, other organs such as the heart are affected. Embryos tend to develop abnormally and die by 5 dpf. Dominantly inherited mutations in nonphotoreceptor cell–specific genes that cause RP-like degeneration have been reported only rarely (Mullen and LaVail, 1975; To et al., 1993).

The nbb Mutation

The nbb mutation causes visual threshold fluctuation under scotopic conditions (Li and Dowling, 2000a). It seems that the loss of visual sensitivity in nbb+/− fish is due to a dysfunction in the rod system. For example, after 2 hr of dark adaptation, rod signals measured by phototopically matched colored light are no longer detected. Cone functions, on the other hand, are essentially normal, although slight cone threshold elevations are observed. In nbb+/−, the outer retinal function determined by electroretinographic recordings is indistinguishable from that of wild-type animals. However, inner retina neural activities such as ganglion cell–firing are somewhat abnormal; the threshold light required to fire ganglion cell action potentials is significantly higher in nbb+/− than in wild-type fish.

At the present time, the only known anatomic abnormalities in nbb+/− are found in the olfactoretinal centrifugal pathway. In most fish species, terminalis neurons (TNs) that are located in the anterior/ventral part of the olfactory bulb project axons to the brain (Munz et al., 1982; Demski and Northcutt, 1983; Springer, 1983; Stell et al., 1984; Walker and Stell, 1986; Stell and Walker, 1987; Umino and Dowling, 1991). While most of the TN axons terminate in the forebrain and midbrain, some project to the retina where they extend along the border of the inner nuclear layer and interplexiform layer and synapse onto dopaminergic interplexiform cells (DA-IPCs) (Zucker and Dowling, 1987). In wild-type retinas, TN axons first extend out radially and then branch. Secondary and tertiary branches are often observed. In nbb+/− fish, the number and appearance of TNs in the olfactory bulbs as well as the distribution of TN axons in the forebrain and midbrain appear normal, and the number of TN axons that enter the retina is similar to that of wild-type fish. However, the distribution of TN axons in the retina is altered (Fig. 3). In addition, the number of DA-IPCs in the nbb+/− retina is reduced, particularly in aged animals (Fig. 3).

thumbnail image

Figure 3. Olfactoretinal centrifugal pathway in zebrafish. A: A schematic drawing of the forebrain and midbrain of zebrafish (dorsal view). Terminalis neurons (TNs, red circles) are found in the olfactory bulb. OE, olfactory epithelium; ON, olfactory nerve; OB, olfactory bulb; OP, optic nerve; TE, telencephalon; TC, tectum; RE, retina. B, C, and D highlight the olfactory bulb, the optic nerve, and the retina, respectively. B: A whole-mount olfactory bulb (outlined by the dashed line, anterior is up) stained with an antibody against FMRFamide, a neuropeptide released by TN. Both TN cell bodies and axons are stained (arrows). C: FMRFamide immunostaining of a control retinal section showing TN axons in the optic nerve (vertical arrows) and in the retina (horizontal arrows). Photoreceptor cells (top left) are nonspecifically labeled. D: A double-labeled control retinal section showing the TN axons (red) and dopaminergic interplexiform cells (DA-IPCs; green, arrows). Photoreceptor cells (top) are nonspecifically labeled. E,F: Flat-mounted, double-labeled, wild-type (E) and nbb+/− (F) retinas. Note the disruption and reduction of TN axons (red) and DA-IPCs (green) in the nbb+/− retina. (Li and Dowling 2000a; printed with permission from the publisher) Scale bar in F = 100 μm in B,E,F, 30 μm in C,D

Download figure to PowerPoint

It is worth noting that homozygous nbb are embryonically lethal. In nbb−/− embryos, the retina develops normally during the first 2 dpf, but thereafter, they start to degenerate. The degeneration is found all over the CNS, including the retina, forebrain, and midbrain. By 7 dpf, the retina as well as tectum are destroyed. The nbb−/− embryos die at 8 dpf.

PERSPECTIVES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION
  5. DOMINANT VISUAL SYSTEM MUTATIONS
  6. PERSPECTIVES
  7. Acknowledgements
  8. REFERENCES

Behavioral tests based on CNS-controlled visual responses have been used in functional analyses of zebrafish visual system development and visual performance. While the behavioral startle and phototactic response assays test simply the light response, OKR, OMR, and escape response assays examine directly visual acuity and visual sensitivity (Kimmel et al., 1974; Clark, 1981; Brockerhoff et al., 1995; Easter and Nicola, 1996; Li and Dowling, 1997; Neuhauss et al., 1999). The OKR, OMR, and escape response assays are sensitive and robust, allowing for the screening of visual system mutations without the preconception of gross eye morphologic abnormalities. It is expected that, in the near future, other behavioral or behavioral-psychological assays will be developed to study high-level visual functions and to isolate mutations that cause subtle defects in the visual pathway, for example, mutations that affect spatial and frequency sensitivity or contrast sensitivity (Baier, 2000; Orger et al., 2000).

To date, mutations in many genes that affect zebrafish visual system development and function have been identified. However, most of these mutations have only been preliminarily studied (Karlstrom et al., 1996; Malicki et al., 1996; Trowe et al., 1996; Link et al., 2000; Malicki and Driever, 1999). In the future, these mutations will be further characterized. By examining cell morphology or by using cell-specific antibodies, for example, one would be able to determine the cell types that are affected by the mutated genes (Kimmel et al., 1995; Westerfield, 1995; Talbot et al., 1995; Connaughton and Dowling, 1998; Connaughton et al., 1999; Vihtelic et al., 1999; Kennedy et al., 2001). In addition, physiological assays that permit the direct analyses of retinal neuronal properties should be further developed. To date, several in vivo electrophysiological assays, such as the electroretinographic and ganglion cell spike recordings, have been used in the analysis of zebrafish outer and inner retinal functions (Branchek, 1984; Brockerhoff et al., 1995; Li and Dowling, 1997, 2000a; Hughes et al., 1998; Saszik et al., 1999). However, intracallular or tectum recording techniques have not been explored in the mutants (Sajovic and Levinthal, 1982a,b, 1983; McMahon, 1994; Fan and Yazulla, 1997; Connaughton and Nelson, 2000). These assays may very likely provide tools that allow one to narrow down the defects to a cellular level in the mutant retinas. The characterization of zebrafish visual system mutations will provide insights into the molecular genetics of vertebrate vision.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION
  5. DOMINANT VISUAL SYSTEM MUTATIONS
  6. PERSPECTIVES
  7. Acknowledgements
  8. REFERENCES

Research in the author's laboratory is supported by a start-up fund from the University of Kentucky College of Medicine and the NIH.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. BEHAVIORAL ANALYSIS OF VISUAL DEVELOPMENT AND VISUAL FUNCTION
  5. DOMINANT VISUAL SYSTEM MUTATIONS
  6. PERSPECTIVES
  7. Acknowledgements
  8. REFERENCES