Investigations of Photoreceptor Synaptic Transmission and Light Adaptation in the Zebrafish Visual Mutant nrc

Abstract

purpose. To characterize the retinal physiology of the zebrafish visual mutant no optokinetic response c (nrc) and to identify the genetic map position of the nrc mutation.

methods. Electroretinograms were recorded from wild-type and nrc zebrafish larvae between 5 to 6 days postfertilization. Responses to flash stimuli, On and Off responses to prolonged light stimuli, and responses to flash stimuli with constant background illumination were characterized. The glutamate agonist, 2-amino-4-phosphonobutyric acid (APB) was used to examine the photoreceptor specific a-wave component of the electroretinogram. Amplified fragment length polymorphism methodology was used to place the nrc mutation on the zebrafish genomic map.

results. nrc and wild-type zebrafish larvae 5 to 6 days postfertilization have similar threshold responses to light, but the b-wave of the nrc electroretinogram is significantly delayed and reduced in amplitude. On and Off responses of nrc larvae to prolonged light have multiple oscillations that do not occur in normal zebrafish larvae after 5 days postfertilization. Analysis of the b-wave demonstrated a light adaptation defect in nrc that causes saturation at background light levels approximately 1 order of magnitude less than those with wild-type larvae. Application of the glutamate analog, APB, uncovered the photoreceptor component of the electroretinogram and revealed a light adaptation defect in nrc photoreceptors. The nrc mutation was placed approximately 0.2 cM from sequence length polymorphism marker Z7504 on linkage group 10.

conclusions. The zebrafish mutant nrc is a possible model for human retinal disease. nrc has defects in photoreceptor synaptic transmission and light adaptation. The nrc mutant phenotype shows striking similarities with phenotypes of dystrophin glycoprotein complex mutants, including patients with Duchenne/Becker muscular dystrophy. Localization of the nrc mutation now makes it possible to evaluate candidate genes and clone the nrc gene.

Athorough understanding of vision is essential for treatment and prevention of retinal disease. Recently, zebrafish have emerged as a model system for genetic studies of the vertebrate retina. Precedence for genetic analysis of vision lies in Drosophila, where use of classical and molecular genetics has identified novel components of invertebrate vision. ^{1 2 3} However, invertebrate vision and vertebrate vision are fundamentally different at the molecular level. ⁴  

Zebrafish are amenable to genetic screening for visual defects, because they can be bred and maintained with minimal resources and because their sophisticated visual system develops rapidly. Zebrafish have color vision mediated by red, green, blue, and UV-sensitive cone photoreceptors. ⁵ They also have scotopic vision mediated by rod photoreceptors. ⁶ Photoreceptor outer segments appear at 55 hours postfertilization ⁷ and by 3 days postfertilization (dpf) zebrafish have rudimentary visual function. ⁸ By 5 dpf photoreceptors have developed sufficiently to screen for visual behavior, ⁹ although rod function is not distinguishable from cone function until 2 weeks postfertilization. ^{6 10}  

A genetic screen of mutant zebrafish was initiated to identify fish with abnormal visual responses but no obvious gross morphologic defects. ⁹ Zebrafish larvae display an optokinetic response (OKR), characterized by smooth pursuit and saccade eye movements in response to illuminated rotating stripes. The OKR can be used as a highly effective behavioral assay to isolate zebrafish larvae with visual defects. ^{9 11} To select for visual mutants with outer retinal defects, electroretinograms (ERGs) are recorded from larvae with abnormal OKRs. Several recessive zebrafish retinal mutants have been isolated by this method. ¹²  

Here we describe an ERG analysis of a zebrafish visual mutant, no optokinetic response c (nrc), which has no OKR in white light. ^{9 11} Ultrastructural analyses have revealed specific defects in nrc cone photoreceptor pedicles and the nrc retinal pigment epithelium (RPE). ¹³ We show that nrc has an abnormal b-wave and abnormal On and Off responses, suggestive of a synaptic transmission defect. We also show that nrc has a photoreceptor light adaptation defect. As a first step toward determining the primary molecular defect, we identify the genetic locus of the nrc mutation. 

Methods

Generation and Maintenance of Zebrafish Strains

Zebrafish were maintained as described previously ¹⁴ on a 10-hour dark:14-hour light cycle. nrc ^a14 was isolated as described previously. ¹¹ nrc larvae were produced for these studies by crossing nrc heterozygous adults. Therefore, OKR responder siblings, referred to here as wild-type (WT) siblings, are genotypically either heterozygous or homozygous WT fish. We did not detect greater than normal variability in the behavioral phenotype or ERGs in this group of larvae. nrc larvae do not eat, and they die at 10 to 13 dpf. Therefore, for control animals we used unfed OKR responder siblings at 5 to 6 dpf. The lack of food does not affect retinal morphology up to 8 dpf. ¹³  

OKR Screening

OKR screening for visual mutants was done as described previously. ⁹ nrc mutants show no OKR response in white light. However, their eyes occasionally move spontaneously, indicating that the mutation does not interfere with eye movement. 

Electroretinograms

Larvae were positioned on a feather on a sponge in a Petri dish containing Hank’s embryo media ¹⁴ in dim room light. Larvae were immobilized on their side between the ridges of a feather. To record ERGs, a glass micropipette (10- to 30-μm diameter tip) containing a silver wire and filled with Hank’s embryo media was positioned on the cornea, and a ground wire was submerged in the Petri dish. The ERG set up was then placed in darkness. WT larvae, dark adapted for at least 2 hours and positioned under dim red light or infrared light, showed a sensitivity threshold of −4.7 ± 0.4 (mean ± SD) log intensity units (n = 11) using a white light flash that has an unattenuated output at the cornea of 7.82μ J/cm². WT larvae positioned under dim room light showed a similar threshold of −4.3 ± 0.5 log intensity units (n = 5). There is no statistically significant difference between the light- and dark-adapted threshold (P = 0.15). This is consistent with previous reports indicating that rods do not contribute significantly to visual sensitivity at this stage of development. ^{6 10} Visual threshold was chosen to be a 20 μV b-wave response measured from the base line. Light was delivered through fiber optics positioned approximately 10 mm above the eye. Brief flash stimuli (<1 msec) were produced by a Canon 540EZ Speedlite Flash at one-sixteenth power. Prolonged light stimuli or background light were produced with an Orion Fiber Optic Illuminator (Model 77501). Interstimulus time was≥ 5 seconds. Stimuli were attenuated with neutral density or interference filters. Data were filtered with a high-frequency cutoff of 100 Hz and a low-frequency cutoff of 1 Hz. Data were collected in Igor Pro (Wavemetrics, Lake Oswego, OR) with an Instrutech ITC-16 analog to digital interface using a library of custom acquisition routines (written by Fred Rieke, University of Washington). Flash intensity versus wave amplitude data (Figs. 3b 4c) were fit empirically with the following Michaelis-Menten type function: y = k1∗((X) ^k2)/(((X) ^k2)+ (k3) ^k2), where k1, k2, and k3 are constants. A fast positive going electrical artifact was periodically seen at the time of the flash. The artifact is so brief that it did not affect our analysis of these traces. 

Data were included only from larvae that had a consistent response (±5%–10%) to a test flash delivered approximately every 5 to 10 minutes. Unhealthy or damaged WT larvae showed an ERG that decreased in amplitude and changed waveform over the course of the experiment. 

ERGs were recorded both at room temperature, 22 to 24°C and at the fish-rearing temperature of 26 to 28°C by placing the recording dish on top of a glass manifold connected to a circulating water bath. No significant differences were found between high and low temperature recordings. 

For analysis of the ERG a-wave, fish were bathed for 1 minute in 1 mM 2-amino-4-phosphonobutyric acid (APB). ¹⁵ This treatment caused no obvious changes in appearance or swimming behavior. The amplitude of the a-wave was measured from the base line. 

Genetic Map Position

nrc AB strain heterozygotes were outcrossed with WT fish from the polymorphic WIK strain. nrc heterozygotes were identified from progeny of this cross by screening the larvae of pairwise matings. OKR responder and nonresponder larvae were frozen at− 70°C. To prepare genomic DNA, larvae were digested overnight in 200μ g/ml proteinase K, 10 mM Tris, pH 8.0, 0.5% SDS, 0.5 M EDTA at 55°C and then extracted once with phenol and twice with phenol/CHCl₃/isoamyl alcohol, and then EtOH-precipitated. DNA was resuspended in 1× TE, pH 7.6, overnight at 37°C. Linked polymorphic markers were identified using AFLP technology described in Ransom and Zon. ¹⁶  

Markers were placed on the zebrafish physical map using the Loeb/NIH/5000/4000 (LN54) radiation hybrid panel ¹⁷ and the Goodfellow T51 radiation hybrid panel ¹⁸ (Research Genetics, Inc., Huntsville, AL). Reactions were conducted as specified in Hukriede et al. ¹⁷ Z markers were used according to Research Genetics. 

These experiments were in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Results

ERG Waveform and Sensitivity

To determine whether the nrc mutant has a functional defect in the outer retina we compared ERG responses of nrc and OKR responder siblings to a light flash of less than 1-msec duration. Light stimulates a transient negative potential on the cornea, the a-wave. The rising phase of the a-wave is derived from field currents produced by stimulation of photoreceptors. ^{19 20} The a-wave is followed by a corneal positive b-wave. The b-wave is derived from secondary neurons postsynaptic to photoreceptors. ²⁰ In 5 to 6 dpf WT larvae the a-wave is occluded by the b-wave (Fig. 1) . 

The b-wave of the nrc mutant is reduced in amplitude and delayed compared with WT (Fig. 1) . In general in nrc the maximum b-wave amplitude was decreased approximately fourfold, and the b-wave implicit time (flash onset to peak) was increased approximately fourfold (Table 1) . 

nrc and WT larvae have a similar b-wave response threshold. Dark-adapted WT larvae have a sensitivity threshold of− 4.7 ± 0.4 (mean ± SD) log intensity units (n = 11), whereas dark-adapted nrc have a sensitivity threshold of −4.7 ± 0.8 log intensity units (n = 7), using a white light flash that has an unattenuated output at the cornea of 7.82 μJ/cm². The waveforms of WT and mutant larvae ERGs are consistent over the full range of flash intensities (See Fig. 3a , No Background). 

Oscillations in Response to Prolonged Light Stimulus

We used steps of illumination to examine On and Off response signaling pathways. On and Off pathways arise through independent retinal bipolar cells and they are mediated by different glutamate receptors. ²¹ Off bipolar cells are depolarized, and On bipolar cells are hyperpolarized by glutamate released from photoreceptors in the dark. There is growing evidence that the d-wave, a positive change in potential on the cornea when lights are turned off, is generated by off-bipolar cells. ^{22 23} Both the On and the Off responses to prolonged light of nrc mutants are abnormal and oscillatory. 

Figure 2a shows averaged responses from one WT larva and three mutant larvae to 1.6-second light stimuli spanning three orders of magnitude of intensity. The WT On response is the corneal positive b-wave, and the Off response is the corneal positive d-wave. In contrast to WT, the nrc On and Off responses to prolonged light are abnormal and are followed by multiple oscillations. Three nrc larvae examples show the variability between fish. The traces shown in Figure 2a are averages of 5 to 10 individual responses. Often the oscillations are not in phase with one another from one response to the next and are therefore not detected when the data are averaged. Figure 2b shows an example of oscillations in single traces. The number of oscillations in averaged responses to prolonged light stimuli under various conditions ranged from 1 to 5. There was no correlation between the number of oscillations and illumination intensity, illumination duration, or temperature. These studies are summarized in Table 2 . All nrc larvae had oscillations in either the On or the Off response or both. The few On responses that lacked oscillations were similar in waveform to the nrc flash response. 

Light Adaptation

To characterize light adaptation, we measured responses of nrc larvae to flashes of light in the presence of background illumination (Fig. 3) . The nrc and WT waveforms from individual larvae are shown in Figure 3a . Responses of seven WT and eight nrc larvae were averaged and normalized (Fig. 3b) . b-Wave maximum amplitudes were normalized to a 7.8 μJ/cm² flash response with no background illumination. The normalized value instead of the absolute value is presented to focus clearly on the difference in light adaptation instead of the difference in overall b-wave size. nrc responses are more attenuated by background illumination than WT responses. Mutant b-waves in response to a 7.8 μJ/cm² flash are approximately 87% of WT in the presence of 0.012 mW/cm² background light, 46% in the presence of 0.11 mW/cm² background light, and undetectable in 0.96 mW/cm² background light. 

a-Wave Light Adaptation

The b-wave analysis described in the preceding paragraph revealed a light adaptation defect in nrc. To resolve the origin of this defect we examined light adaptation of the photoreceptor specific a-wave. We unmasked the photoreceptor-specific a-wave by suppressing the b-wave with APB. ¹⁵ APB is a glutamate analog that reduces the b-wave of the ERG. ²⁴ Bathing WT or nrc larvae for 1 minute in 1 mM APB unmasks the a-wave of the zebrafish larval ERG (Fig. 4a ). At the a-wave maximum there could be some contribution from the residual b-wave-like component of the ERG. Therefore, we chose to examine the a-wave at a time earlier than the maximum. We chose 15 msec after the a-wave onset since this is the initial part of the a-wave slope that has been attributed to photoreceptors. ²⁵ There is no statistically significant difference between a-wave amplitude at 15 msec after the a-wave onset or implicit time in nrc and WT larvae (Table 1) . 

Background illumination attenuates the nrc a-wave more than that of the WT a-wave. The a-wave waveform of individual nrc and WT larvae is shown in Figure 4b . Averaged a-wave amplitudes 15 msec after a-wave onset are reported in Figure 4c (n = 4). With no background light, there is no statistically significant difference between the WT and nrc responses to the flash series (all in the flash series are P ≥ 0.18). However, in response to a flash of 7.8 μJ/cm² in the presence of 0.012 mW/cm² background, nrc a-waves are reduced approximately 46%, whereas WT are reduced approximately 25%[ nrc = 49.7 ± 10.6 (mean ± SEM), WT = 61.0 ± 17.7]. In the presence of 0.11 mW/cm² background light, nrc are reduced approximately 73%, whereas WT are reduced approximately 41% (nrc = 24.7 ± 9.6, WT = 47.8 ± 18.8). Finally, in 0.96 mW/cm² background light, nrc a-waves are reduced approximately 92%, whereas WT are reduced approximately 49% (nrc = 4.6 ± 4.0, WT = 33.6 ± 6.2). The adaptation defect appears to increase with increasing background light intensity. In the brightest background light, 0.96 mW/cm², the light adaptation defect is statistically significant (P = 0.01; Fig. 4c , n = 4). 

Because nrc has defects in photoreceptor morphology, ¹³ we considered that altered morphology of the photoreceptor layer might generally cause defects in light adaptation. As a control, using the same experimental conditions, we examined light adaptation of another visual mutant, partial optokinetic response b (pob). This mutant has a disrupted photoreceptor layer, because of a selective loss of red cones by 5 dpf. ¹¹ However, we found unlike nrc, pob adapts normally to background light (data not shown). 

nrc Map Position

As a first step in determining the primary molecular defect caused by the nrc mutation we determined its physical map position. We used AFLP technology to identify markers that are closely linked to the mutant gene. ¹⁶ Using 256 primer pair combinations, 21 AFLP markers that are linked to the nrc locus were isolated from duplicate DNA samples, each containing 24 nrc mutant or WT sibling larvae. We next examined recombination frequencies in 60 individual larvae for each of the 21 markers to approximate the genetic distance between the nrc locus and the AFLP marker. Five AFLP markers were within 1 cM of the nrc locus. One of these was subcloned, sequenced, and mapped using the LN54 zebrafish radiation hybrid panel ¹⁷ and the T51 zebrafish radiation hybrid panel. ¹⁸ On the Tuebingen map of the zebrafish genome (http://www.map.tuebingen.mpg.de/) this AFLP marker (named unp1417) is between 54.8 and 57.3 cM from the top of linkage group 10 with a LOD score of 11.8. This approximate map position was confirmed by analyzing SSLP markers ²⁶ present at 56.4 cM (http://zebrafish.mgh.harvard.edu/mapping/ssr_map_index.html). Five recombinants out of 2648 meioses were identified using marker Z7504. This places Z7504 approximately 0.2 cM from nrc. Twenty-five recombinants of 2724 meioses, which were different from Z7504 recombinants, were identified using Z9574. This places Z9574 0.9 cM from the nrc locus on the opposite side from Z7504. 

Discussion

In this study we showed that (1) the b-wave in response to a flash is reduced in amplitude and delayed in the nrc mutant, (2) both the On and Off responses to prolonged stimuli show abnormal oscillations in nrc, and (3) nrc has a light adaptation defect. The physiological defect in the outer retina is likely to be responsible for the mutant’s inability to follow rotating stripes in the optokinetic assay. This study is significant because zebrafish mutants such as nrc are potential models for human retinal disease. Our initial characterization of this mutant provides fundamental functional information about its retinal phenotype. 

Morphologic studies have uncovered structural defects in cone photoreceptor synaptic termini and in the RPE of nrc. ¹³ Proper organization of the photoreceptor terminal is important for visual function. The photoreceptor has a specialized synapse that contains ribbons, vesicle-associated electron dense structures localized to the presynaptic membrane. The role of the ribbon is unknown, but one hypothesis is that it acts as a pipeline to deliver synaptic vesicles to the photoreceptor membrane for tonic neurotransmitter release. ²⁷ Photoreceptor synaptic ribbon structures are present in nrc, but they are not properly anchored to the presynaptic membrane. ¹³ In addition postsynaptic neurons in nrc do not properly invaginate into the photoreceptor, and synaptic vesicles clump in the photoreceptor terminals. ¹³ Secondary neurons that do invaginate nrc photoreceptor terminals appear to be horizontal cells. The RPE in the nrc mutant also has increased phagasomes and lipid droplets. ¹³ However, photoreceptor outer segments are normal in length and the discs are ordered. ¹³  

Delayed and Decreased b-Wave

The delayed and attenuated b-wave in nrc is consistent with a defect in synaptic transmission and is likely the consequence of the morphologic abnormalities in the photoreceptor synapse. An alternative possibility is that there is improper detection of glutamate by secondary neurons. Although the nrc b-wave is abnormal, the nrc threshold response to light is similar to that of WT, suggesting that phototransduction in the nrc mutant is normal. 

Oscillations in Response to Prolonged Stimulus

The initiation or termination of prolonged illumination evokes oscillations in nrc. Photoreceptor synaptic terminals tonically release glutamate in the dark. Light stimulates hyperpolarization of the photoreceptor, which ultimately slows this glutamate release. The onset of darkness returns the photoreceptor to a state of tonic neurotransmitter release and renewed sensitivity to light. Release of neurotransmitter must be carefully controlled so that photoreceptors can properly respond to changes in light stimuli. 

The oscillatory response to a prolonged light stimulus in nrc suggests a defect in modulation of glutamate release. Perhaps glutamate is released in bursts rather than tonically as is normally supported by the ribbon structure. Activity in the stimulated secondary neurons would then generate corneal positive oscillations. This hypothesis is likely given the morphologic defects in the nrc photoreceptor terminals. Alternatively, it is possible that (1) secondary neurons do not properly respond to the change in glutamate release or (2) feedback mechanisms between photoreceptors and secondary neurons that normally dampen the response are defective. 

In a separate study, oscillations in the nrc mutant ERG after a 10-msec pulse of illumination were observed.^13,28 In our studies a flash less than 1 msec consistently showed only a single response. Only 1 in 15 healthy nrc larvae showed an oscillation and only in response to a 7.8μ J/cm² flash intensity. The 10-msec flash duration may be the minimum time necessary to induce oscillations. It should be noted that constant background light did not cause continual oscillations. 

Oscillations in response to a light stimulus have been described in normal ERGs of some vertebrates ²⁰ but not in normal zebrafish ^{6 29 30} after 5 dpf. The normal oscillatory potentials seen in some vertebrates appear as small wavelets superimposed on the b-wave after a short light flash. These oscillations are different from the nrc oscillations in phase, onset, and amplitude. Oscillations similar to the ones in nrc have been observed in three to four dpf zebrafish larvae. ³¹ This similarity suggests nrc has an incompletely developed photoreceptor synapse at 5 to 6 dpf. 

Light Adaptation Defect

Comparison of the b-waves of nrc and WT larvae in the presence of background illumination shows that light adaptation is impaired in nrc. We unmasked the photoreceptor specific a-wave by exposing larvae to APB, an agonist for ON-bipolar cell glutamate receptors ^{15 32 33} and found that light adaptation is defective in nrc photoreceptors. The initial slope of the a-wave has been attributed to photoreceptors. ²⁵ Although APB may also affect other pathways such as On transmission from cones to horizontal cells, ³⁴ it has no known effects on the early kinetics of the photoreceptor derived a-wave. 

WT and nrc a-waves are similar in amplitude and shape in the presence of APB. In conjunction with the normal threshold sensitivity the a-wave provides additional support for normal phototransduction in the mutant. However, background illumination in the presence of APB shows that there is a light adaptation defect in nrc photoreceptors. In the presence of background light, nrc a-wave responses to a flash are attenuated compared with those of WT. Because pob, a mutant with an abnormal photoreceptor layer, ¹¹ has no adaptation defect (data not shown), the light adaptation defect appears to be specific to nrc photoreceptors. 

Light adaptation in photoreceptor outer segments occurs through modulation of phototransduction mediated by Ca⁺². ^{35 36} It seems unlikely that local changes in photoreceptor termini would directly affect adaptation in the outer segment. Another possibility is that the nrc photoreceptor adaptation defect is caused by an abnormality in a protein that has structural or functional roles in both the outer segment and synaptic terminal. The appearance of phagasomes and lipid droplets within the RPE in nrc suggests that photoreceptor disc turnover may be abnormal. However, nrc phototransduction seems normal, and the outer segment morphology is normal. Alternatively, the adaptation defect may reflect a loss of intercellular feedback mechanisms dependent on synaptic transmission. Improper invagination of secondary neurons within the nrc terminal indicates possible abnormal intercellular interactions. ¹³  

In summary, a defect in an element of the photoreceptor synaptic terminal likely explains the abnormalities in nrc electrophysiology and morphology. Intriguingly, the morphologic and electrophysiological phenotypes of nrc share two striking parallels with the phenotypes of mutations associated with the dystrophin glycoprotein complex (DGC). One, disruption of dystrophin in humans and mice ^{37 38 39} and laminin β2 in mice, ⁴⁰ causes an altered b-wave similar to the nrc mutant. A mutation in dystrophin causes Duchenne/Becker muscular dystrophy. Two, disruption of laminin β2, causes morphologic defects within the retina similar to the nrc mutant. ⁴⁰ The synaptic ribbon is mislocalized, and the secondary neurons do not properly associate with the photoreceptor terminal. These studies indicate that the DGC and laminin β2 are important for normal photoreceptor synaptic development and function within the retina. Furthermore, retinal isoforms of dystrophin are important for photoreceptor synaptic function. ^{41 42} The map position of nrc presented here will enable us to evaluate these genes as well as others as candidates. Identification of the nrc gene will give insights into photoreceptor structure and function. 

 Supported by Molecular and Cellular Biology Training Grant 5 T32 6M07270 (HAVE), the Howard Hughes Medical Institute (JBH), and National Institutes of Health Grant EY12373 (SEB).

 Submitted for publication May 31, 2000; revised October 13 and November 3, 2000; accepted November 22, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

The nrc mutant’s response to a flash of 7.8μ J/cm² white light is abnormal in comparison to WT. In nrc the b-wave is delayed and decreased. Each trace is an average of more than five flash responses.

Table 1.

 

ERG a- and b-Wave Implicit Times and Amplitudes in Response to a 7.8 μJ/cm² Flash

	Flash Onset to Peak			15 msec After a-Wave Onset
		a-Wave (msec)*	b-Wave (msec)*	a-Wave Amplitude (μV), †	b-Wave Max Amplitude (μV), †
WT	66 ± 32 (8)	88 ± 8 (9)	81 ± 14 (8)	299 ± 48 (9)
nrc	37 ± 10 (4)	302 ± 54 (8)	92 ± 22 (4)	68 ± 11 (8)
P	0.16	7.9 × 10−20	0.69	0.0004

 The implicit time and maximum amplitude differences between nrc and WT b-waves are statistically significant.

*  Values are means ± SD; numbers in parentheses are the sample numbers.

†  Values are means ± SEM.

Figure 2.

 

(A) The nrc mutant shows abnormal oscillations when prolonged light is turned on and off. A WT sibling shows a large positive going b-wave when the light is turned on and a positive going d-wave when the light is turned off. Light is turned on for 1.6 seconds (□) and then shut off (▪). Light intensity: 1, 0.96 mW/cm²; 2, 0.11 mW/cm²; and 3, 0.012 mW/cm². Each of the three nrc mutant panels shows example traces with varying light intensity from three different larvae. The oscillations do not change significantly with light intensity. Each trace is an average of more than five flash responses. (B) There are numerous oscillations seen in the nrc mutant response. When single traces are averaged, some of the oscillations are no longer apparent because they are not in phase. These traces give an example of the oscillations that are not seen in the average. Light intensity is 0.96 mW/cm².

Table 2.

 

Number of Oscillations in the nrc Larval Response Does Not Correlate with Illumination Duration or Temperature

	No. of Larvae Showing Oscillations at Light Onset	No. of Larvae Showing Oscillations at Light Off
28°C, light on 1.6 sec
WT	0/6	0/6
nrc	5/5	4/5
22°C, light on 1.6 sec
WT	0/2	0/2
nrc	6/9	7/9
22°C, light on 20 sec
WT	0/2	0/2
nrc	4/4	3/4

 Values are presented as number/total number.

Figure 3.

 

(A) Examination of the b-wave reveals an impairment of nrc larvae’s ability to adapt to background illumination. WT and nrc mutant flash responses in no background illumination and constant background illumination are compared. Flash intensities: (A) 7.8 μJ/cm², (B) 1.0 μJ/cm², (C) 0.14μ J/cm², (D) 0.011μ J/cm², and (E) 0.0019μ J/cm². Arrow: the initiation of the flash. (B) The normalized adaptation data of seven WT and eight nrc larvae. Error bars, SEM. The data are normalized as explained in the text to clearly show the light adaptation defect.

Figure 4.

 

(A) To examine the zebrafish a-wave the chemical agonist APB was applied to larvae as described in Methods. Flash responses of untreated WT larvae and APB-treated WT larvae. (B) Comparison of a-waves of WT larvae and nrc larvae flash responses in constant background illumination. All larvae are treated with APB as described in Methods. Flash intensities: (A) 7.8 μJ/cm², (B) 1.0μ J/cm², (C) 0.14μ J/cm², and (D) 0.011μ J/cm². Arrow: the initiation of the flash. (C) The a-wave amplitudes 15 msec after a-wave response onset of four nrc larvae and four WT larvae are compared in the presence and absence of background light. Error bars, SEM.

The authors thank Fred Rieke for providing the Igor data acquisition program and Matthias Seeliger for telling us how to use APB on zebrafish larvae to reveal the a-wave.