A Point Mutation (W70A) in the Rod PDEγ Gene Desensitizing and Delaying Murine Rod Photoreceptors

Abstract

purpose. To examine the corneal electroretinogram (ERG) of transgenic mice (W70A mice) carrying a point mutation (W70A) in the gene encoding for theγ -subunit of rod cGMP phosphodiesterase (PDEγ).

methods. The ERG of W70A mice was compared with that of normal mice. Cone responses were separated from rod responses by light adaptation, whereas rod sensitivity was assessed by threshold stimulation with dim light. Spectral sensitivity curves of the ERG were obtained using a constant response criterion.

results. The ERG of the W70A mouse has a desensitized, delayed rod b-wave at threshold, and a prolonged rod b-wave at higher flash intensities. The a-wave is absent even at maximal stimulation. The cone ERG of the W70A mouse is indistinguishable from that of normal mice. The spectral sensitivity of the W70A mouse is maximal in the UV spectrum, in contrast to the normal mouse, which is most sensitive in the green region of the spectrum. This supports the interpretation of the results as normal cone and abnormal rod function in the W70A mouse.

conclusions. The W70A mouse represents new model of stationary nyctalopia that can be recognized by its unusual ERG features.

When light reaches rod photoreceptors, rhodopsin is photoexcited and activates a G-protein (transducin), which activates rod phosphodiesterase (PDE) by removing the inhibitory γ-subunits from the PDE core, so that it can hydrolyze cyclic guanosine monophosphate (cGMP). The lowered cytoplasmic cGMP concentration closes cGMP-gated cationic channels in the plasma membrane, causing the rod to hyperpolarize, which is the adequate stimulus for second-order neurons in the retina. 

We have used targeted gene disruption to eliminate the expression of the rod PDEγ, which leads to photoreceptor degeneration in these mice. ¹ The introduction of a transgene with a point mutation (alanine substitutes for tryptophan at position 70 [W70A] in the 87 amino acids comprising the PDEγ molecule) created a distinct phenotype. ² The retina of the W70A mouse does not degenerate, inasmuch as electroretinograms (ERGs) of 13-month-old W70A mice had the same appearance as those of young ones, and histology revealed no abnormalities at 13 months of age. ² However, the electrophysiology of rod photoreceptors is affected in the W70A mouse. In single rod recordings, the response to light was desensitized and delayed and the recovery of the response was prolonged. In previous experiments, we found the W70A mouse desensitized but could not identify a delay in the ERG corresponding to the delay in the single rod responses. ² We have now identified what we consider to be a delayed rod response in the ERG of the W70A mouse, bringing the ERG into closer agreement with the single photoreceptor electrophysiology. In this report, we characterize in vivo retinal function of the W70A mouse and compare it with disorders found in humans, ^{3 4 5 6 7 8 9 10} and in murine ^{11 12 13} models. 

Methods

The creation of mice lacking PDEγ and of the W70A mouse has been described in detail elsewhere, ^{1 2} as have been the methods used for genotyping and protein analysis. 

ERGs were obtained from mice anesthetized with a mixture of ketamine (50 mg/kg) and xylazine (10 mg/kg) administered intramuscularly. The pupils were dilated with 1% phenylepinephrine and 1% cyclopentolate. The mouse was placed on a heated stage calibrated to maintain the temperature of the body around 37°C. A 30-gauge-needle reference electrode was placed subcutaneously (SC) on the forehead and a similar ground electrode on the trunk. A saline-moistened cotton wick electrode contacted the cornea. The stimuli were obtained from a stroboscope that was removed from its housing and mounted in a metal box with a circular aperture, 3 cm in diameter and placed 9 cm from the center of the pupil. This produces a field of approximately 20°; it is assumed that most of the light stimulus is derived from scattered light. In support of this assumption, ERGs obtained from normal and W70A mice, using a full field dome surrounding the head of the mouse, yielded qualitatively similar responses. ¹³ Neutral density and spectral filters could be placed in front of the aperture of the stroboscope to change the intensity and the wavelength of the flash. The following absorption filters were used: Kodak Wratten gelatin filters 36, 50, 48, 75, 74, 21, and 29 (Eastman Kodak, Rochester, NY) and Corning glass filters 5113 and 5970 (Corning Glass, Corning, NY). The transmission of each filter was measured with a spectrophotometer (Beckman Instruments, Palo Alto, CA), and the wavelength of peak transmission: 410, 458, 471, 488, 538, 593, 633, 360, and 380 nm, respectively, was used for plotting spectral sensitivity curves. The maximum flash intensity (inμ W/cm²) at the cornea delivered through each spectral filter was measured with a digital photometer (J16; Tektronix Instruments, Beaverton, OR), after removing infrared radiation with an appropriate filter. The detector, 1 cm in diameter, was placed at the level of the cornea so that the stroboscope light covered the detecting area completely. The maximal light intensity of the white light (unfiltered) flash was 0.8 × 10³μ W/cm² at the level of the cornea. The duration of the flash was approximately 20 μsec, as stipulated by the manufacturer of the stroboscope (Grass Instruments, Quincy, MA). 

ERG responses were detected with an oscilloscope and an evoked response–detecting computer in parallel (CA 1000; Nicolet Instruments, Madison, WI), which averaged responses at a digitization rate of 1 MHz. The bandpass of the input amplifier was 1 to 250 Hz. The mice were dark-adapted overnight before testing. Stimulation was begun at 4.8 logarithmic units below maximum intensity of the stroboscope and responses were averaged to one flash every second. At high flash intensities, each flash was presented every 20 seconds, which was found long enough to exclude interference of one flash to the next. To determine the spectral sensitivity, we recorded responses to different intensities at each wavelength, from threshold to suprathreshold levels of stimulation. We determined the relative number of quanta per flash to produce a constant criterion response at each wavelength and plotted the reciprocal of these values on a logarithmic scale versus wavelength on a linear scale as a spectral sensitivity function. 

We also examined the light-adapted ERG by exposing the eye to a beam focused on the pupil in Maxwellian view, which illuminated the entire eye of the mouse. The beam was obtained from a slit lamp, entering the eye slightly off the optical axis, to keep the mirror of the slit lamp from blocking the strobe light. The brightness of this field, as seen through the pupil of a human observer, was at maximum 10⁷ candela (cd)/m². The level of brightness of the adapting field was changed by altering the voltage across the bulb of the slitlamp. 

Results

Figure 1 shows ERGs obtained from one representative normal and one W70A mouse. The W70A mouse is less sensitive to light; its ERG shows a prolonged b-wave with strong stimulation (upper traces) and lacks an initial negative a-wave. Arrows highlight a slow positive wave that gradually merges with the b-wave at strong stimulation. Figure 2 gives the b-wave amplitude and the b-wave implicit time as functions of the light intensity; these data are combined from W70A mice derived from different founders (line 1 and line 2). For the b-wave amplitude (Fig. 2A) , the curve of the W70A mice is shifted to the right on the abscissa (light intensity) by approximately two logarithmic units, indicating desensitization of the W70A mouse. With maximal stimulation, the amplitude of the b-wave of the W70A mouse approaches that of the normal mouse. Figure 2B shows the implicit time (time to peak after stimulation) of the b-wave and also shows the implicit time of the late positive wave of the W70A mouse. The implicit time of the b-wave of the W70A mouse is relatively short at all light intensities, including threshold. In contrast, the implicit time of the b-wave recorded from normal mice is short at high light intensities but becomes longer in response to dim stimuli, which is most accentuated at threshold. In the W70A mouse, the implicit time of the late positive wave is much longer but approaches that of normal mice in response to high-intensity stimulation. 

Figure 3 demonstrates that the ERG signature of the W70A mouse was consistent in both transgenic lines, and Figure 4A shows the effect of light adaptation on the ERG of normal and W70A mice. In the dark-adapted state virtually no a-wave is detectable in the ERG of the W70A mouse, whereas it is conspicuous in that of the normal mouse. In the light-adapted state, the a-wave in the normal ERG disappears, and the ERG of the W70A mouse is indistinguishable from that of the normal mouse (Figs. 4A , 4B) . 

We examined the ERG to different parts of the spectrum, including the UV region. Figure 5A shows ERG responses to UV (360 nm), blue (450 nm), and yellow (575 nm) stimuli from threshold to maximal levels for normal (above) and W70A (below) mice. The late positive wave characteristic of the W70A response is apparent at all wavelengths. The difference to the normal ERG can be seen in spectral sensitivity functions (Fig. 5B) . The spectral sensitivity of the dark-adapted W70A mouse is 1 to 2 logarithmic units below that of the normal mouse at all wavelengths except in the UV region, where it reaches similar levels. The maximum sensitivity of the normal mouse is in the greenish region, whereas that of the W70A mouse is in the UV region of the spectrum. The b-wave implicit time at threshold in the normal ERG is between 130 and 145 msec from 400 to 650 nm; in the UV it is shorter. The implicit time of the threshold b-wave of the W70A mouse is shorter at all wavelengths, but the difference between normal and W70A mice is least in the UV region. 

All experiments were conducted on two different transgenic lines of W70A mice, and the results have been found reproducible. No differences could be detected in the ERGs of different lines of W70A mice. 

Discussion

This report describes a unique pattern in the ERG of a transgenic mouse, the W70A mouse. This mouse model is less sensitive and its b-wave implicit time is shorter than that of normal mice. With strong stimuli, the implicit times of the b-waves of normal and W70A mice become similar. In the light-adapted state, the ERG of the W70A mouse is indistinguishable from that of the normal mouse. This picture can be explained by assuming that cones and not rods photoreceptors determine b-wave thresholds in the W70A mouse, because cone b-wave implicit times in the mouse are shorter than those of rods. ¹⁴ Interestingly, the spectral sensitivity curve of dark-adapted W70A mice is virtually identical with that of light-adapted normal mice. ¹⁵ This supports our conclusion that threshold responses are predominantly cone-mediated in the W70A mouse. 

The ERG of the W70A mouse furthermore has an unusual waveform. With strong stimuli, there is virtually no a-wave and the b-wave is more prolonged than the b-wave of the normal mouse. With weaker stimuli a late b-wave-like response is detectable in the ERG of the W70A mouse, which was not appreciated previously. ² This late, insensitive response has an implicit time at threshold that is about three times longer than that of the normal mouse. This is the same order of magnitude at which single rod responses of the W70A mouse are delayed at threshold. ² The abnormal response can be explained by such delay combined with the insensitivity of rod photoreceptors in the W70A mouse. As flash intensity increases, the delayed rod response adds on to the cone response, producing a prolonged b-wave. The lack of an a-wave also is explainable by the insensitivity of W70A rods. A relatively strong rod response is required before an a-wave becomes detectable, which the rods of the W70A are incapable of producing at the flash intensities available. The biochemical reasons for the delay and desensitization of the rod photoresponse have already been discussed. ² In brief, the W70A mutation impairs PDE activation and deactivation, resulting in decreased sensitivity and slowed termination of the photoresponse. 

The W70A mutation appears to leave the cones, including the UV ones, unaffected, which is further support for the finding that rods and cones use different forms of PDEγ. ^{16 17} Because the rods of W70A mice are desensitized, the spectral sensitivity of this mouse is maximal in the UV region of the spectrum. In contrast, dark-adapted normal mice are most sensitive to the greenish part of the spectrum, ^{18 19} where rhodopsin absorbs maximally. Normal mice have a relatively high sensitivity in the UV region of the spectrum. Our results show that in this region, the b-wave implicit time at threshold becomes shorter, as if mediated in part by cones. The relatively high sensitivity of murine UV cones in the light-adapted state has been reported previously, and the presence of UV cones has been demonstrated, ¹⁵ which is supported by our results. 

Recently rod responses have been detected in the Nougaret form of stationary nyctalopia in man. ²⁰ In this case there is a point mutation in the α-subunit of transducin. Transducin binds PDEγ to catalyze the rod photoresponse. This mutation in transducin also desensitizes rod photoreceptors by about two logarithmic units, but there is no delay in the threshold rod ERG because rod a- and b-waves of normal waveform are elicitable. This appears to be a gain of function mutation, which leaves the rods constitutively light-adapted. In the W70A mutant there is a loss of function mutation; the defect is only seen in the homozygous state, it is therefore recessive. 

To summarize, the findings in the ERG of the W70A mouse—desensitized rods in young and old mice, normal functioning cones, absence of retinal degeneration—are characteristic for congenital stationary nyctalopia. Thus, the W70A mouse exhibits a new form of stationary nyctalopia. The unique ERG signature of this mouse should make it easy to detect, if it were to occur in humans. It is also useful in providing another example of a genetic defect in a protein involved in rod phototransduction that does not lead to photoreceptor degeneration. 

 Supported by National Institutes of Health Grant R01 EY 11510, Pro Retina Deutschland e.V. (DJS), Research to Prevent Blindness (SHT), Fight for Sight (SHT), and National Eye Institute (SHT).

 Submitted for publication October 9, 1998; revised June 8, 1999; accepted June 24, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

(A) Dark-adapted ERGs of a normal mouse (left) and a W70A mouse (right) at different flash intensities, which are indicated by the neutral density filtering in logarithmic units at the left of each trace. The vertical line indicates 100 μV, except for the bottom four traces on the left and the bottom six on the right, where it represents 50 and 25 μV for the responses with greater amplification (2× and 4×). The duration of each trace is 300 msec. The arrows point out a slow positive wave seen in the ERG response of the W70A mouse. (B) Responses from a similar pair of mice at a slower time base (700 msec per trace) to reveal the extreme slowness of the late positive b-wave like response in the W70A mouse (arrows).

Figure 2.

 

(A) B-wave amplitude plotted against the logarithm of flash intensity for normal (n = 15) and W70A mice (n = 9). Error bars, ±SEM. (B) B-wave implicit time plotted against the logarithm of flash intensity for normal and W70A mice. The implicit time of the slow positive wave, seen only in the W70A mouse, is also shown. These data represent the average of 15 normal mice, including 7 mice heterozygous for the W70A mutation, which are indistinguishable from normal mice, and 9 W70A mice. The implicit time of the slow response represents results from 6 mice. Error bars, ±SEM.

Figure 3.

 

ERG responses to the same flash intensity (0.9 log units of neutral density filtering) obtained from mice of different transgenic W70A lines. Arrows indicate the late positive wave typical for the ERG of the W70A mouse.

Figure 4.

 

(A) ERG responses of a normal (left) and a W70A mouse (right) in the dark (top trace) and at increasing levels of steady background light, (lower 2 traces). The strength of the adapting light is shown at the left. The vertical line indicates 125μ V for the left and 63 μV for the right column of responses. Each trace is 300 msec in duration. (B) Amplitude of the a-wave (top) and b-wave (bottom) plotted against the strength of the background as the logarithm of candelas per square meter. This data represents the average of five normal and five W70A mice. Error bars, ± SE.

Figure 5.

 

(A) ERG responses of a normal (top) and a W70A mouse (bottom) to flashes from different parts of the spectrum, i.e., at wavelengths of 360 (left), 450 (middle), and 575 (right) nm. Flash intensities are indicated by the neutral density filtering at the left of each trace. The vertical line at the top right represents 125 μV. Each trace is 700 msec in duration. The arrows point out the slow b-wave response in the W70A mouse. (B) Spectral sensitivity functions (top) of normal (closed circles) and W70A (open circles) mice. The ordinate represents the logarithm of the reciprocal of the relative number of quanta in a flash to elicit a constant response of 50 μV. The abscissa represents the wavelength of the flash in nanometers. The implicit time of the near-threshold response is plotted against the wavelength of the flash (bottom). These data are the average of five normal and five W70A mice. Error bars, ±SE.

The authors thank Monica Mendelsohn and members of the Gouras and Goff laboratories for support and discussion of this study.