A dominant form of inherited retinal degeneration caused by a non-photoreceptor cell-specific mutation

Abstract

We have isolated a dominant mutation, night blindness a (nba), that causes a slow retinal degeneration in zebrafish. Heterozygous nba fish have normal vision through 2–3 months of age but subsequently become night blind. By 9.5 months of age, visual sensitivity of affected fish may be decreased more than two log units, or 100-fold, as measured behaviorally. Electroretinographic (ERG) thresholds of mutant fish are also raised significantly, and the ERG b-wave shows a delayed implicit time. These defects are due primarily to a late-onset photoreceptor cell degeneration involving initially the rods but eventually the cones as well. Homozygous nba fish display an early-onset neuronal degeneration throughout the retina and elsewhere in the central nervous system. As a result, animals develop with small eyes and die by 4–5 days postfertilization (pf). These latter data indicate that the mutation affecting nba fish is not in a photoreceptor cell-specific gene.

Mutations that cause retinal degeneration have been found in a variety of vertebrate species including man (1–3). Retinitis pigmentosa (RP), for example, a group of human diseases affecting approximately 1 in 4,000 worldwide, is due primarily to photoreceptor cell degeneration (2, 3). Patients suffering from RP typically become night blind in adolescence and may become completely blind between the ages of 30 and 60 years (2, 3).

RP is inherited in different ways. Approximately 40% of cases are dominantly inherited, 20% are recessively inherited, and 8% are sex-linked (4). The inheritance of the remaining cases is unknown. We know most about dominant RP where mutations have been identified in the rhodopsin gene and in genes coding for another photoreceptor-specific protein, peripherin (5–7). So far mutations in six different loci that cause dominant RP have been identified (8). It is believed, however, that only about half of the cases of dominant RP can be accounted for by the mutations discovered so far (2, 3, 8).

Our laboratory has been seeking visual system-specific mutations in zebrafish (Danio rerio), a freshwater teleost fish that can be readily mutagenized chemically (9–11). To detect slow retinal degenerative diseases similar to those observed in retinitis pigmentosa patients, we have developed a behavioral test, based on the escape response exhibited by fish when challenged by a threatening object, to analyze quantitatively the visual sensitivity of adult zebrafish. Using this assay, we have identified a mutation that causes a dominantly inherited form of retinal degeneration. Genetic studies indicate that this mutation is not in a photoreceptor cell-specific gene and may represent a new form of dominantly inherited retinal degeneration.

MATERIALS AND METHODS

Animals and Maintenance.

AB strain zebrafish (Danio rerio) were obtained originally from the University of Oregon (12) and propagated at our fish facility by inbreeding. Adult zebrafish were kept in 6-liter tanks with recirculating fish water (distilled water with Instant Ocean salt and vitamins added) heated to 28.5°C and fed twice per day with freshly hatched brine shrimp. Embryos were kept in 5-cm Petri dishes in 28.5°C fish water. Animals used in this study were maintained in a normal light–dark cycle (light, 8 a.m.–10 p.m.; light intensity, 1.15–2.45 μW/cm²).

Behavioral Assay.

The apparatus used consisted of an immobilized circular container (10-cm diameter) with transparent sides, surrounded by a rotating drum covered with white paper. A black segment (5 × 5 cm) was marked on the paper as a threatening object. A post (3-cm diameter) that was placed in the center of the container prevented the fish from swimming directly from one side of the container to the other. The drum was illuminated from above with a white light source (unattenuated intensity = 4.25 × 10² μW/cm²) and turned at 10 rpm by a belt attached to an adjacent motor. The fish was viewed by a television monitor attached to an infrared video camera (Fig. 1A).

Figure 1.

(A) Apparatus for the behavioral analysis of visual sensitivity of adult zebrafish. TV, TV monitor; C, infrared video camera; L, light source; P, post; D, drum; M, motor. (B) Dark adaptation curve of adult wild-type zebrafish determined by behavioral testing (see text for details). Data represent the mean ± SD.

Electroretinographic (ERG) Recording.

Zebrafish were dark-adapted for 1 hr before the ERG recording. Under dim red light illumination, zebrafish were anesthetized in 4% 3-amino benzoic acid methylester and immobilized with 10% gallamine triethiodide. Zebrafish were placed on their side on a sponge with one eye facing toward the light source. A slow stream of fish water was directed into the mouth of the fish to keep the animal well oxygenated during the experiment. A small suction pipette (tip diameter of 25 μm) filled with a balanced salt solution was placed on the cornea of the zebrafish, and a silver-ground electrode was placed in the bath. A halogen lamp (unattenuated intensity = 6.75 × 10³ μW/cm²) and shutter were used to illuminate the head of the fish with 0.01-sec flashes of white light. The electrical signals were amplified and recorded conventionally.

Histology.

Histological methods used were similar to those already described (13). In brief, zebrafish embryos and enucleated adult eyes were primary fixed in 4% paraformaldehyde/2.5% glutaraldehyde in 0.06 M sodium phosphate buffer for 2 hr and postfixed in 1% osmium tetroxide in 0.06 M sodium phosphate buffer for 1 hr. After dehydration, specimens were embedded in Epon-Araldite (Polysciences) plastic, cut with an ultramicrotone, stained with 1% methylene blue, and viewed under the light microscope.

Acridine Orange Staining.

Acridine orange staining was used as described (14). In brief, zebrafish embryos were dechorionated and placed in 5 μg/ml of acridine orange (Sigma) in fish water. After 30 min of staining, embryos were washed, anesthetized in 4% 3-amino benzoic acid methylester, and immediately viewed by fluorescence microscopy.

RESULTS

Escape Response.

Zebrafish normally swim slowly in circles when confined in a circular container. However, when challenged by a threatening object, zebrafish exhibit a robust escape response; as soon as the threatening object comes into view, the fish instantly turn and rapidly swim away.

We first tested the effectiveness of our apparatus for eliciting the escape response. With naive wild-type zebrafish (8–10 months old, n = 23), we examined for positive escape responses in two 1-min trials; in one trial the drum rotated clockwise, in the other the drum rotated counterclockwise. The number of times the black segment approached a fish, as well as the number of times a fish showed a positive escape response, was recorded. All of the tested wild-type fish (23 of 23) showed an escape response to the black segment. On average, a fish encountered the black segment 50.26 times in the two 1-min trials, and it responded positively 42.74 times. In other words, approximately 85% of the time wild-type zebrafish responded positively in our apparatus to the approach of the threatening object.

Dark Adaptation.

The escape response is sufficiently robust, so that within 5–10 sec after illuminating the rotating drum, it is usually possible to decide whether a fish is seeing the threatening object. It thus seemed possible, using our behavioral test, to measure the course of dark adaptation of the zebrafish. This would provide data with regard to the absolute visual sensitivity of both the cone and rod systems as well as the time course of dark adaptation.

To test this idea, we did the following experiment. Adult zebrafish (8–10 months old, n = 13) were transferred into transparent containers (one fish in each container). All fish were exposed to a bright light source (intensity = 3.25 × 10³ μW/cm²) for 20 min (light adaptation). The light was turned off and the fish subsequently were kept in complete darkness (dark adaptation). After 2 min in darkness, one fish was placed in the test apparatus to determine the threshold light intensity required to evoke the escape response. Neutral-density filters (in half-log unit steps) were used to change the light intensity on the drum. Initially the light intensity was set at log I = −3.0. If the fish did not show the escape response, the light intensity was increased (i.e., to log I = −2.5) or until the fish showed the escape response. The final light intensity that evoked the escape response was recorded as the threshold for the 2-min point during dark adaptation. Once the threshold was determined, the fish was removed from the apparatus, and the light intensity was reset to a level below the previously measured threshold. Approximately 2 min later, another fish was placed in the apparatus and its threshold was determined. This procedure was repeated every 2 min until dark adaptation was completed (approximately 26 min).

Fig. 1B shows the averaged time course of dark adaptation of wild-type zebrafish. The dark adaptation curve consists of two segments, a cone limb (dashed line) followed by a rod limb (continuous line). The absolute thresholds for the cone and rod systems were −4.3 (2.13 × 10⁻² μW/cm²) and −6.0 (4.25 × 10⁻⁴ μW/cm²) log units, respectively. The cone and rod systems needed approximately 6 and 18 min, respectively, to reach their final thresholds. In the experiment shown in Fig. 1B, each point represents the visual sensitivity of individual fish. It is also possible to determine dark adaptation curves using a single fish, and the results are similar.

We also observed that visual thresholds were significantly lower in the late afternoon hours (4–8 p.m.) compared with early morning hours (4–8 a.m.), suggesting that visual sensitivity in zebrafish is under circadian control (Li and Dowling, unpublished work). Therefore, all visual threshold measurements reported in this paper were done between 4 and 8 p.m.

Isolation of the nba Mutant.

For mutant screening, the light intensity illuminating the drum was set at log I = −5.0, approximately one log unit above the absolute rod threshold of wild-type zebrafish. Individuals that failed to show the escape response under this level of illumination were isolated for rescreening on subsequent days.

In initial experiments, we screened 345 F1 generation fish (8–10 months old) that had been derived from founder fish chemically mutagenized with N-ethyl-N-nitrosourea (ENU). From these 345 fish, we isolated 7 individuals that showed no escape response or an escape response less than 10% of the time when tested under dim light illumination. After rescreening at least five times, the candidate F1 fish were outcrossed to inbred wild-type fish. Two of the candidate fish died before a successful outcross, but the remaining five fish all yielded offspring. The progeny of three crosses showed no detectable visual defects after 11 months of testing and were not examined further. The progeny of two crosses, on the other hand, showed clear visual deficits. In this paper, we describe one of these fish, the night blindness a (nba^da10) fish.

After the outcross of the nba fish with a wild-type zebrafish, we began testing the resulting F2 generation (n = 63) at 2 months of age. At that time all of the F2 generation fish were normal in terms of visual sensitivity. By 4 months of age, 12 of the 63 fish showed no escape response when tested under dim light illumination (log I = −5.0). By 8 months, 27 of the 63 fish showed raised visual thresholds, and by 13 months, 35 of 63 F2 generation fish showed evidence of night blindness. Fig. 2A shows dark adaptation curves of two nba fish at 5.5 months of age. In both cases, thresholds for the cone and rod systems were significantly raised in the affected fish compared with wild-type fish. In the more severely affected fish, there was no evidence of remaining rod function; even after 26 min of dark adaptation, the visual threshold remained above the absolute cone threshold of normal fish.

Figure 2.

(A) Dark adaptation curves of a wild-type fish (circles) and two nba fish (triangles). The horizontal dashed line drawn at log I = −5.0 indicates the light level used for mutant screening. (B) Absolute visual thresholds of nba fish (n = 16) measured behaviorally over a period of 4 months. Note the average thresholds (arrows) become progressively higher with age.

Visual Thresholds in nba Fish Increase with Age.

To measure the visual sensitivity of affected F2 generation nba fish at different ages, we recorded the absolute visual thresholds following 26 min of darkness. We found that the loss of visual sensitivity in nba fish was variable but progressive. At 5.5 months, for example, the average threshold of the affected fish (those whose thresholds were greater than −5.0 log units) was −4.17 ± 0.49 log units, and only 19% of them had a threshold above −4.0 log units. At 7.5 months, the average threshold was −3.98 ± 0.55 log units, and 31% had a threshold above −4.0 log units. By 9.5 months of age, the average threshold was −3.48 ± 0.64 log units, and 75% had a threshold above −4.0 log units. Thus, over a 4-month period, the average visual threshold of the mutant fish rose about 0.7 log units (Fig. 2B).

ERG Recording.

To determine whether the defective visual behavior displayed by the heterozygous nba fish is due to a retinal dysfunction, we recorded the full-field ERG. The vertebrate ERG is characterized by two prominent waves: a corneal negative a-wave, which originates from photoreceptor cells, and a corneal positive b-wave, which originates from second-order cells (15).

We recorded the ERG from both the wild-type (n = 10) and nba (n = 11) fish between 8 and 13 months of age. The ERG waveform of nba fish generally resembled that of the wild-type fish (Fig. 3A). However, the minimum light intensity required to evoke a threshold ERG was significantly higher, by approximately 1.5 log units on average, in the tested nba fish compared with wild-type fish, and the ERG b-wave amplitude of nba fish was reduced over a range of light stimuli compared with that of wild-type fish (Fig. 3B). In addition, the implicit times of the b-wave of nba fish were consistently delayed compared with wild-type fish (Table 1). These data suggest that the defective visual behavior exhibited by nba fish is due to a dysfunction of the outer retina.

Figure 3.

(A) Full-field ERGs of wild-type (Left) and nba (Right) fish to white light (log I = −3) stimuli. a, a-wave; b, b-wave. The vertical dashed lines indicate the onset of the 10-msec light stimulus, and the horizontal lines indicate the implicit times of the b-waves. Calibration bars (Right Lower) signify 200 msec horizontally and 50 μV vertically. (B) V-log curves of wild-type (solid circles) and nba (open circles) fish. The ERG b-wave amplitudes are plotted on a linear scale as a function of log light intensity. Data in B represent the mean ± SD.

Visual thresholds in wild-type animals measured by ERG recording were significantly higher than were visual thresholds measured behaviorally (6.45 × 10⁻³ μW/cm² vs. 4.25 × 10⁻⁴ μW/cm²). The average rise of threshold of the tested mutant animals (13 months old, n = 8) was approximately the same whether measured electrophysiologically or behaviorally (1.56 log units vs. 1.62 log units). However, individual animals showed distinct differences in the extent of threshold increase as measured by the two techniques. In some cases the ERG threshold rise was greater; in others, the rise in behavioral threshold was greater (not shown).

Histological Analysis.

Histological sections of mutant (n = 26) and wild-type (n = 12) retinas were examined at various ages to determine the cellular basis of the retinal defects. We observed structural abnormalities consisting mainly of photoreceptor cell degeneration in all mutant retinas. The degeneration was not uniform over the retinas but tended to be patchy. Areas in which rods had completely degenerated were interspersed between areas where significant numbers of rods remained. Cones generally were preserved better than the rods, but areas of cone degeneration were also observed. The degeneration was also progressive with age. In 5.5-month-old animals, degeneration was seen only in the central-most regions of the retina. By 13 months of age, degeneration was observed across much of the retina, although far peripheral regions tended to be spared.

Fig. 4 shows histological sections from a 13-month-old wild-type fish and a 13-month-old nba fish. In zebrafish the photoreceptors are typically tiered so that in the light-adapted retina the rods are positioned more distally than are the cones (Fig. 4A). In the 13-month-old nba retina, the far peripheral regions were similar to the control retina, but obvious abnormalities were observed throughout the central retina. For example, in some areas only the rods were affected, i.e., the rod outer segments were disorganized and reduced in length, and lipid droplets in the pigment epithelium (PE) were increased in number and size (Fig. 4B). In other areas, the rods were completely absent but the cones appeared quite normal (Fig. 4C). In yet other areas, both the cones and rods were affected, i.e., both were disorganized and reduced in length (Fig. 4D). In addition, some proliferation of the PE was observed in the nba retina (Fig. 4 B and C). Invasion of PE cells into the retina was occasionally seen (not shown). In more degenerated areas, the PE often appeared thin and had lost most of the pigment granules (Fig. 4D).

Figure 4.

Histological sections showing the photoreceptor layer of 13-month-old wild-type and nba retinas. (A) A section from the retina of a wild-type fish. In the light adapted retina, the rods (r) sit distal to the cones (c). Processes from the pigment epithelium (PE) extend between the outer segments of the rods. (B–D) Sections from the central retina of one nba fish showing the variability of degeneration seen in various regions of the affected retina (see text for details). in, inner nuclear layer; ip, inner plexiform layer. Arrows in B indicate the large lipid droplets in PE. [Bar = ≈100 μm (A) and 80 μm (B–D).]

Although the inner layers of the retina usually looked qualitatively normal in affected animals, occasionally some patchy thinning of the inner nuclear layer was observed. To investigate more systematically whether there might be loss of elements other than photoreceptors in the mutant retinas, we measured the thickness of various retinal layers centrally in control (n = 6) and mutant (n = 4) animals at 9.5 months of age. The retinas of mutant animals were clearly thinner than those of wild-type animals (Table 2). Whereas much of this change in thickness could be accounted for by a decrease in the thickness of the photoreceptor layer, a significant difference (P < 0.001, t test) in the thickness of the inner plexiform layer was also noted, suggesting some loss of inner retinal neurons.

Homozygous nba Fish Show Early Neuronal Degeneration in Both Eyes and Brain.

To explore further the nature of the nba mutation, we bred heterozygous nba fish to each other to yield homozygous mutant fish. In initial experiments, we crossed 4–5 pairs of heterozygous nba mutants. In the resulting F3 generation approximately 28% of the fish (39 of 137) showed a small eye phenotype at about 2 days pf (Fig. 5B). To determine whether the small eyes resulted from the loss of retinal cells, we exposed live 2.5-day-old embryos to acridine orange, a vital dye that stains apoptotic cells (14, 16). Such staining revealed prominent apoptotic cell death in both the eyes and brain of homozygous nba fish (Fig. 5D). Histological studies of 2.5-day-old embryos showed that the retinal cells had been generated early in development but then degenerated (Fig. 5F). Cell death was also widespread in the tectum, and scattered dying cells were seen elsewhere in the brain as well. By 3.5 days pf, virtually all of the retinal cells had degenerated and tectal cell death was still obvious (Fig. 4H). By 5 days pf, all of the homozygous nba fish had died.

Figure 5.

Photographs of live embryos (A–D) and histological sections of the brain and retinas (E–H) of wild-type (Left) and homozygous nba (Right) fish. (A and B) Photograph of 2.5-day-old wild-type and nba embryos. Note the smaller eye (arrow) and blood cells that have pooled in the heart (asterisk) in nba fish. (C and D) Acridine orange staining of 2.5-day-old wild-type and nba embryos. Brightly staining apoptotic cells are seen in the retina (arrow) and tectum (arrowhead) in nba fish. Asterisks indicate nonspecific staining of yolk cells. (E and F) Transverse sections through the brain and retina of 2.5-day-old wild-type and nba embryos. Note the darkly stained dying cell in both the retina (arrows) and tectum (arrowheads) in nba fish. (G and H) Transverse sections of 3.5-day-old wild-type and nba embryos. By this time most of the retinal cells have degenerated (arrows), and the tectum continuous to show evidence of cell death (arrowheads) in nba fish. [Bar = ≈250 μm (A–D) and 100 μm (E–H)]. 

DISCUSSION

In this paper, we describe a simple behavioral test based on the escape response that can be used to analyze quantitatively the visual sensitivity of adult zebrafish. Using this assay we have isolated a dominant mutation (nba) that causes a slow retinal degeneration. Heterozygous nba fish display normal visual sensitivity up to 2–3 months of age, but subsequently such fish become night blind. By 9.5 months of age, affected fish show an average absolute threshold rise of about 2.5 log units.

When heterozygous nba fish are bred to homozygosity, retinal degeneration begins by 2 days pf, and by 4 days pf the retina is virtually destroyed. Not only do photoreceptor cells degenerate, but other retinal neurons die as well. Furthermore, tectal and other brain cells also degenerate. By 5 days pf, the animals die. Why they die is not clear, but several defects are apparent in the homozygous animals. For example, although the heart continues to beat at 4 days pf, blood pools in it and does not circulate (Fig. 5B).

If the gene underlying the nba mutation were a photoreceptor-specific gene, we might expect that the photoreceptor cell degeneration would be faster in homozygous fish than in heterozygous fish but that the degeneration would nevertheless be restricted to the photoreceptor cells. Such is the case with the retinal degeneration slow (rds) mouse. In heterozygous rds mice, the photoreceptors degenerate slowly; at 1 year of age about half of photoreceptor cells remain although their outer segments are abnormal. In homozygous rds animals, on the other hand, the photoreceptor outer segments never form properly, and by 1 year of age the photoreceptors have completely degenerated. However, in both heterozygous and homozygous rds animals, the degeneration is limited strictly to the photoreceptors (17).

That the degeneration in homozygous nba fish is not limited to the photoreceptor cells indicates that the nba gene is not photoreceptor specific. Clearly a variety of neurons degenerate early in homozygous nba fish, including retinal, tectal, and other brain neurons. It is perhaps noteworthy that the retinal neurons appear to degenerate faster in homozygous nba fish than the other brain neurons. It also may be the case that degeneration is not limited to the nervous system; both heart and body axis abnormalities are seen in the animals at 3–5 days pf.

A surprising aspect of the nba fish is that even after 13 months, areas containing rods are present throughout the retina. Especially in the periphery, quite normal-looking rods are observed in mutant animals of all ages. This may relate to the fact that rod photoreceptors in fish are generated throughout the life of the animal (18, 19). At the retinal margins in fish, all retinal cells are continuously being generated but there are also precursor cells found scattered throughout the outer and inner nuclear layers that add new rod photoreceptors across the retina. The areas of more normal-looking rods seen throughout the retinas of the mutant animals may represent regions in which rods have been more recently generated.

The continuous generation of rods throughout the life of a fish may also explain the variability in loss of visual sensitivity observed in mutant animals. That is, depending on where and how many new rods are being generated in a retina, the extent of both the behavioral and ERG loss of sensitivity may vary, and the two measures of sensitivity loss may not necessarily go hand in hand, as we have observed.

Finally, what relevance might our findings have for understanding inherited retinal degeneration in man? That mutations in nonphotoreceptor specific genes can cause a retinitis pigmentosa-like retinal degeneration is of obvious significance, and several recessively inherited mutations of this kind have been described in humans and animals (3, 4, 20). Dominantly inherited mutations in non-photoreceptor cell-specific genes that cause retinitis pigmentosa-like degenerations have been observed only rarely (21). Because the photoreceptor cells are primarily affected in heterozygous nba fish, it appears these cells have special susceptibility to mutations of the nba gene. However, clearly in homozygous nba animals and probably in heterozygous nba animals (Table 2), other retinal neurons are affected. It is also the case that neurons elsewhere in the brain degenerate in homozygous nba animals, and the same may hold in heterozygous animals. A search for signs of degeneration in inner retinal layers and elsewhere along the visual pathways may be warranted in cases of dominantly inherited forms of retinitis pigmentosa where the gene defect is not known.

Table 1.

Implicit times of ERG b-wave (msec)

	Wild type	nba/+
Log I = −4	120.18  ±  16.69	190.50  ±  10.92
Log I = −3	112.00  ±  12.08	166.50  ±  17.33
Log I = −2	103.13  ±  10.20	137.50  ±  15.12

Data represent the mean ± SD.

Table 2.

Retinal thickness (μm)

	Wild type	nba/+
Total	566.90  ±  43.71	418.70  ±  37.94
PL	297.29  ±  38.71	213.71  ±  29.99
INL	77.43  ±  15.38	63.07  ±  13.84
IPL	141.02  ±  27.69	89.74  ±  16.02

Data represent the mean ± SD. PL, photoreceptor layer; INL, inner nuclear layer; IPL, inner plexiform layer.

ABBREVIATIONS

nba

night blindness a

ERG

electroretinogram

pf

postfertilization

RP

retinitis pigmentosa

PE

pigment epithelium