Photoreceptor and Post-Photoreceptoral Contributions to Photopic ERG a-Wave in Rhodopsin P347L Transgenic Rabbits

Abstract

Purpose.: The a-wave of the photopic electroretinogram (ERG) of macaque monkeys is made up of the electrical activities of cone photoreceptors and post-photoreceptoral neurons. However, it is not known whether the contributions of these two components change in retinas with inherited photoreceptor degeneration. The purpose of this study was to determine the contributions of cones and post-photoreceptoral neurons to the a-wave of the photopic ERGs in rhodopsin Pro347Leu transgenic (Tg) rabbits.

Methods.: Ten Tg and 10 wild-type (WT) New Zealand White rabbits were studied at 4 and 12 months of age. The a-waves of the photopic ERGs were elicited by xenon flashes of different stimulus strengths before and after the activities of post-photoreceptoral neurons were blocked by intravitreal injections of a combination of 0.2 to 0.4 mM of 6-cyano-7-nitrouinoxaline-2,3(1H,4H)-dione, disodium (CNQX) and 2 to 4 mM of (±)-2-amino-4-phosphonobutyric acid.

Results.: The percentage contribution of the cone photoreceptors to the photopic ERG a-waves increased with increasing stimulus strength, and the percentage ranged from 54% to 75% in 4-month-old WT rabbits. In contrast, the percentage contribution of the cone photoreceptors in 4-month-old Tg rabbits ranged from 32% to 51% (P < 0.05). The mean percentage contribution of cone photoreceptors became still smaller at 11% to 48% in 12-month-old Tg rabbits.

Conclusions.: These results suggest that the relative contribution of cone photoreceptors to the photopic ERG a-wave is smaller in retinas with inherited photoreceptor degeneration. This indicates that the a-waves of the photopic ERGs in patients with retinitis pigmentosa must consider this lower contribution from the cone photoreceptors.

The electroretinogram (ERG) is a mass electrical potential change of the retina that is elicited by light stimulation and is easily recorded noninvasively with a corneal electrode. ¹ The ERG arises from the neural activity of the different types of retinal cells, and it can be used to perform a layer-by-layer study of retinal function in patients and animals. ²  

The origins of the photopic or light-adapted a-wave of the ERG in macaque monkeys was studied by Sieving et al. ^{3 –5} They injected glutamate agonists and antagonists intravitreally to dissect the retinal circuits. They found that the a-wave of the photopic ERG received contributions not only from the cone photoreceptors but also from post-photoreceptoral neurons (e.g., OFF-bipolar cells and horizontal cells) ^3,4 because cis-2,3-piperidine dicarboxylic acid (PDA) or kynurenic acid reduced the a-wave amplitude. A later study by Robson et al. ⁶ showed that the PDA-sensitive post-photoreceptoral a-wave component started at much earlier times of approximately 5 ms in macaques. Frieberg et al. ⁷ also estimated the time course of the cone photoreceptor response in normal human ERGs using the paired-flash technique, in which an intense “probe” flash was delivered at different times after a “test” flash. Their results showed that the photopic ERG a-wave of the human ERG contains an appreciable postphotoreceptoral component, similar to that reported in monkeys. ^{3 –6}  

These studies, which were designed to determine the origins of the photopic ERG a-wave, have been performed primarily on normal macaque monkeys and human eyes. ^{3 –7} It is not known whether the contributions of photoreceptors and post-photoreceptoral neurons are altered in retinas with inherited photoreceptor degeneration (e.g., retinitis pigmentosa [RP]) because the most commonly used RP animals are mice and rats, whose amplitude of photopic ERG a-wave is very small. This makes it difficult to quantify the changes in the a-wave amplitude before and after intravitreal injection of pharmacologic agents. ^{8 –12}  

We have recently succeeded in generating a rabbit model of retinal degeneration. ¹³ This animal has the rhodopsin Pro347Leu mutation, which is one of the major mutations in autosomal dominant retinitis pigmentosa in humans. ¹⁴ These animals have a slowly progressive photoreceptor degeneration, as do human RP patients with this mutation, ^{13,15 –18} though it is still unclear whether the retinal degeneration is due to a point mutation of the rhodopsin gene or to an overexpression of rhodopsin in these animals. We believed that this rhodopsin transgenic (Tg) rabbit can be an excellent animal model in which to study the retinal origins of the photopic ERG a-wave in RP because rabbits have a large photopic a-wave. In addition, the large size of the rabbit's eye enabled us to perform reliable intravitreal injections of pharmacologic agents. ^15,16,18  

Thus, the purpose of this study was to compare the contributions of cone photoreceptors and post-photoreceptoral neurons with the a-wave of the photopic ERGs between wild-type (WT) and Tg rabbits. To accomplish this we examined the postphotoreceptoral neural activity before and after they were blocked by pharmacologic agents. 

Materials and Methods

Animals

The experiments were performed on 10 Tg and 10 littermate WT New Zealand White rabbits. Our techniques for generating Tg rabbits have been described in detail. ¹³ This study was conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. All protocols were approved by the Animal Research Review Board of Nagoya University Graduate School of Medicine (no. 23005). 

ERG Recordings

Each animal was anesthetized with an intramuscular injection of 25 mg/kg ketamine and 2 mg/kg xylazine. ERGs were recorded with a bipolar contact lens electrode (GoldLens; Doran Instruments, Littleton, MA). Animals were placed in a Ganzfeld bowl and stimulated with stroboscopic stimuli (model SG-2002; LKC Technologies, Gaithersburg, MD). The full-strength stimulus was attenuated with neutral density filters in 0.5-log unit steps. Photopic ERGs were recorded after 10 minutes of light adaptation, and the stimulus strength ranged from 0.2 to 2.2 log cd-s/m² (photopic unit), and they were presented on a rod-suppressing white background of 3.3 log scot td. Signals were amplified, band pass-filtered between 0.3 to 1000 Hz, and averaged using a computer-assisted signal analysis system (MEB-9100; Neuropack, Nihon Kohden, Tokyo, Japan). 

Drug Injections

Drugs and techniques for the intravitreal injections have been described in detail. ^15,16,18 The drugs were dissolved in sterile PBS, and the pH was titrated to 7.4 with hydrochloric acid or sodium hydrate. The drugs were injected into the vitreous with a 30-gauge needle inserted through the pars plana approximately 1 mm posterior to the limbus. 

Two types of glutamate analogs—(±)-2 amino-4-phosphonobutyric acid (APB; Sigma-Aldrich Japan, Tokyo, Japan) and 6-cyano-7-nitroquinoxaline-2,3(1H,4H)-dione (CNQX, Sigma-Aldrich Japan)—were used. APB is an agonist of the type 6 metabotropic glutamate receptor, and it blocks signal transmission between the photoreceptors and depolarizing or ON-bipolar cells. ¹⁹ CNQX is an antagonist of the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid/kainic acid (AMPA/KA) class of ionotropic glutamate receptors and is known to block the light responses of the hyperpolarizing or OFF-bipolar cells, horizontal cells, and all third-order retinal neurons. ²⁰ Thus, the combination of APB and CNQX is expected to isolate the photoreceptor responses. We could not use PDA, ²¹ another type of antagonist of the AMPA/KA class of ionotropic glutamate receptors, because PDA was not commercially available. Intravitreal concentrations were 2 to 4 mM for APB and 0.2 to 0.4 mM for CNQX, assuming that the vitreous volume of the NZW rabbit is 1.5 mL. ²² The drugs were dissolved in 0.05 mL saline. 

Recordings were begun approximately 60 to 90 minutes after the drug injections, and studies were completed within 3 hours. Although the drug effects were reversible, we only used the rabbits that had not been used for any previous experiments. 

Measurement of a-Waves

To determine the photoreceptor and post-photoreceptoral contributions to the a-wave of the photopic ERGs quantitatively, we measured the amplitude of the a-wave before and after drug administration. Before drug administration, the a-wave amplitude was measured from the baseline to the first negative trough; after it, the a-wave amplitude was measured from the baseline to the potential at the time of the a-wave peak before the drugs (Fig. 1A). Then the percentage cone photoreceptor contribution was calculated by the expression (a-wave amplitude after APB and CNQX)/(a-wave amplitude before drugs) × 100. This method has been used to determine the degree of cone photoreceptor contribution to the a-wave. ³  

Figure 1.

 

Photopic ERGs of WT and rhodopsin P347L Tg rabbits. (A) Method of measuring a-wave amplitude. The a-wave amplitude before drug administration was measured from the baseline to the first negative trough. (B) The a-wave amplitude after drug administration was measured from the baseline to the negative value at the time of the a-wave peak before drug administration. Representative photopic ERGs recorded from WT and Tg rabbits at 4 and 12 months of age. ERG waveforms to five different stimulus strengths of 0.2 to 2.2 log cd-s/m² are shown. (C) Plots of the a-wave amplitude to five different stimulus strengths. Results of WT and Tg rabbits at 4 and 12 months of age are shown. Bars indicate the SE of the means of five animals. (D) Plots of the a-wave implicit times to five different stimulus strengths. Results of WT and Tg rabbits at 4 and 12 months of age are shown. Bars indicate the SE of the means of five animals.

Statistical Analysis

Because the data were normally distributed, unpaired Student's t-tests were used to determine whether the amplitude of the a-wave of WT rabbits was significantly different from that of Tg rabbits. Differences were considered to be significant when P < 0.05. 

Results

Photopic ERGs of WT and Tg Rabbits

Representative photopic ERGs recorded from WT and Tg rabbits at 4 and 12 months of age are shown in Figure 1B. The ERG waveforms elicited by five different stimulus strengths from 0.2 to 2.2 log cd-s/m² are shown. We found that all the ERG components of Tg rabbits decreased progressively with increasing age; the a-wave was more affected than the b-wave. These general ERG findings agree with the results reported in our earlier publications. ^13,15  

The amplitudes of the a-waves of the photopic ERGs of Tg rabbits were significantly smaller than those of WT rabbits at 4 months of age and even smaller at 12 months of age (Figs. 1B, 1C). The implicit times of the photopic ERG a-wave were not significantly different between the Tg and WT rabbits when they were 4 months of age, but the Tg rabbits had severely delayed implicit times when they were 12 months of age (Fig. 1D). 

Effect of APB or CNQX Alone on Photopic ERG a-Wave

To confirm that the a-waves of the photopic ERG in rabbits originated from the same neurons as macaque monkeys, we examined the effect of APB or CNQX alone on the a-wave of the photopic ERGs in WT and Tg rabbits when they were 4 months of age (Fig. 2). We found that intravitreal injection of APB did not alter the leading edge of the photopic a-wave, and the maximal a-wave amplitudes were nearly the same before and after the APB injection for both WT and Tg rabbits (Fig. 2, upper trace). In contrast, an intravitreal injection of CNQX significantly changed the leading edge of the a-wave, and the maximal a-wave amplitude was significantly reduced in both types of rabbits (Fig. 2, lower trace). These results were comparable to the results in primates ^{3 –6} and support the belief that the photopic ERG a-wave receives significant contributions from post-photoreceptoral neurons, including OFF-bipolar cells and horizontal cells in both WT and Tg rabbits. 

Figure 2.

 

Representative waveforms of photopic ERG a-wave before and after APB or CNQX alone in WT and Tg rabbits of 4 months of age. ERG waveforms to five different stimulus strengths of 0.2 to 2.2 log cd-s/m² are superimposed.

Amplitude Changes of Photopic ERG a-Wave after Pharmacologic Drug Administration

We next examined the contribution of the cone photoreceptors to the photopic ERG a-wave at the time of the a-wave peak. The black lines in Figure 3 show the photopic ERG a-waves before drugs, and the color lines (blue, WT; red, Tg) show the ERG waveforms after intravitreal injection of a solution of combined APB and CNQX (i.e., the cone photoreceptor response). The vertical dotted lines show the timing of the a-wave peaks before drug administration. As reported in primates, ^{3 –6} the a-wave amplitude is greatly reduced after blocking all post-photoreceptoral neurons by glutamate analogs. 

Figure 3.

 

Representative waveforms of photopic ERG a-wave before (black) and after (blue and red) intravitreal injection of combination of APB and CNQX in WT and Tg rabbits at 4 and 12 months of age. Vertical dotted lines: timing of the a-wave peaks before drug administration.

Mean amplitudes of the a-wave before and after drug administration at the time of the a-wave peak (Fig. 1A), are plotted in Figure 4. The a-wave amplitude decreased after injection of both APB and CNQX for all stimulus strengths in both WT and Tg rabbits. 

Figure 4.

 

Plots of the a-wave amplitude before (black) and after (blue and red) intravitreal injection of combination of APB and CNQX in WT (left) and Tg (right) rabbits at 4 and 12 months of age. Bars indicate the SE of the means of five animals.

Relative Contributions of Cone Photoreceptors to Photopic a-Wave

We next compared the relative contributions of the cone photoreceptors with the photopic ERG a-wave for the two types of rabbits. For this, we calculated the percentage contribution of the cone photoreceptors; that is, we divided the a-wave amplitude after APB+CNQX by the a-wave amplitude before drug administration (Fig. 5). We found that the percentage contribution of the cone photoreceptors became greater with increasing stimulus strengths in both WT and Tg rabbits, which is consistent with the findings in normal macaque monkeys. ³ The percentage contribution of the cone photoreceptors ranged from 32% to 51% in Tg rabbits, which was significantly smaller than that in WT rabbits at 54% to 75%, at 4 months of age (P < 0.01; Fig. 5, left). 

Figure 5.

 

Plots of the percentage contribution of cone photoreceptors to the photopic ERG a-wave in WT (blue) and Tg (red) rabbits at 4 and 12 months of age. Bars indicate the SE of the means of five animals. *P < 0.05; **P < 0.01.

We also calculated these values when the animals were 12 months of age. The percentage contribution of the cone photoreceptors ranged from 11% to 48% in Tg rabbits, which was also significantly smaller than that in WT rabbits at 41% to 70% (P < 0.05; Fig. 5, right). The percentage contribution of cone photoreceptors to the photopic a-wave in 12-month-old Tg rabbits was <50% for all stimulus strengths. 

Comparison of Postreceptoral Components

The smaller contributions of cone photoreceptors to the photopic a-waves in Tg rabbits can be explained simply by a decrease in cone photoreceptor responses caused by the photoreceptor degenerations, which can be clearly seen in Figure 4. However, it can also be caused by an increase in neural activities of the post-photoreceptoral neurons. To investigate whether the latter explanation was the cause, we calculated the amplitudes of post-photoreceptoral components at the time of the a-wave peak by subtracting the post-APB+CNQX waveform from the predrug waveform. Results are plotted in Figure 6. 

Figure 6.

 

Plots of the amplitude of post-photoreceptoral component in the photopic ERG a-wave in WT (black) and Tg (red) rabbits at 4 and 12 months of age. Bars indicate the SE of the means of five animals. *P < 0.05.

Although the maximal amplitudes of the post-photoreceptoral components were not significantly different in WT and Tg rabbits, the intensity amplitude function for the two types of rabbits were different when they were 4 months of age (Fig. 6, left). The amplitude of the post-photoreceptoral component was nearly saturated at lower stimulus strengths of 0.7 to 1.2 log cd-s/m² in Tg rabbits. In contrast, this value increased gradually and reached maximum amplitude at the highest stimulus strength of 2.2 log cd-s/m² in WT rabbits. The amplitudes of the post-photoreceptoral components in Tg rabbits were significantly larger than those in WT rabbits at lower stimulus strengths of 0.2 and 0.7 log cd-s/m² (P < 0.05). 

A similar tendency of the stimulus strength-amplitude functions of WT and Tg rabbits was also seen when they were 12 months of age, but the overall amplitudes of post-photoreceptoral components of Tg rabbits were greatly reduced, probably because of advanced retinal degeneration. These results indicated that the smaller contribution of cone photoreceptors to the photopic a-wave in young Tg rabbits occurred partially because of the enhanced post-photoreceptoral responses at lower stimulus strengths. 

Discussion

It is unknown whether the contributions of photoreceptors and post-photoreceptoral neurons are altered in retinas with progressive photoreceptor degeneration. Our present results clearly demonstrated that the percentage contribution of the cone photoreceptors to the photopic a-wave was significantly lower in rhodopsin P347L Tg rabbits than in WT rabbits over a 2 log unit range of stimulus strengths at both 4 and 12 months of age. We found that especially in the retina of 12-month-old Tg rabbits, the percentage contribution of cone photoreceptor to the photopic ERG a-wave was less than half, irrespective of the stimulus strength (Fig. 5, right). 

Our results showed that the effects of stimulus strength on the cone photoreceptors and post-photoreceptoral contributions to the photopic a-wave of normal retinas were similar to those in primates reported by Bush and Sieving. ³ They measured the degree of cone photoreceptor and post-photoreceptoral contribution to the photopic a-wave at the time of the a-wave peak in normal macaque monkeys before and after APB and PDA. They did not report the exact percentage values, but they showed ³ that it was relatively low at 55% at the lowest stimulus strengths and that it gradually increased to a maximum of 92% at the highest stimulus strength. They interpreted these findings that the post-photoreceptoral contribution to the photopic a-wave was primarily responsible for the initial 1 to 1.5 log units of strength, whereas cone photoreceptor contribution progressively dominated the photopic a-wave at higher stimulus strengths. We also observed a similar pattern in our WT rabbits (Fig. 5), but the percentage contribution of cone photoreceptor at the highest stimulus strength was higher in macaque (92%) than in our WT rabbits (75%). This difference might have been due to the difference in the type of stimulus (200-ms long-flash stimuli in their study vs. xenon brief-flash stimuli in our study) or difference in species. 

We found that the percentage contribution of cone photoreceptors to the photopic a-wave in Tg rabbits was significantly lower than in WT rabbits (Fig. 5). These results are reasonable because the cone photoreceptor is gradually attenuated whereas the middle and inner retinas are still well preserved in Tg rabbits. ^13,15 Additional analyses demonstrated that the smaller percentage contribution of cone photoreceptors in young Tg rabbits can be explained, in part, by the enhancement of the amplitudes of the post-photoreceptoral component, especially at lower stimulus strengths (Fig. 6, left). Such enhanced amplitudes of the post-photoreceptoral component in Tg rabbits were no longer present at 12 months in Tg rabbits, probably because of advanced retinal degeneration. 

We do not know the exact mechanism for the enhanced amplitudes of the post-photoreceptoral components elicited by weaker stimulus intensities in young Tg rabbits. This enhanced post-photoreceptoral response may be due to secondary functional changes in the OFF-bipolar/horizontal cells or their synapses after progressive photoreceptor degenerations. 

Using computational molecular phenotyping, we have recently shown that during the course of rod photoreceptor degeneration, rod ON-bipolar cells switch their phenotype by expressing ionotropic glutamate receptors (iGluRs). ¹⁷ We also found that the rod bipolar cells effectively lose rod contacts and make ectopic cone contacts and express iGluRs. ¹⁷ This secondary retinal remodeling may contribute to the enhanced post-photoreceptoral responses in our Tg rabbits. Similarly, detailed ERG studies in rhodopsin P347L Tg pigs and rabbits have demonstrated that the electrical activities of the cone ON-pathway were also enhanced at a relatively early stage of retinal degeneration. ^18,23 In addition, an increase in the ERG responses from the inner retina (e.g., scotopic threshold response) was also reported in the retina of the aged Royal College of Surgeons rat, a rodent model of retinal degeneration. ^24,25  

Taken together, inherited retinal diseases associated with progressive photoreceptor degeneration may lead to different types of functional changes in the post-photoreceptoral retinal circuits, including the ON- and OFF pathways, during a relatively early stage of retinal degeneration. 

We believe our results have important clinical implications. The a-wave of the photopic ERG is believed to be shaped primarily by electrical activities of cone photoreceptors in patients. However, the results of this study suggest that the cone photoreceptor function may be overestimated when the amplitude of the cone ERG a-wave is used as an indicator of residual cone photoreceptor functions in patients with progressive photoreceptor degeneration such as RP. Thus, when the standard stimulus strength (3.0 cd-s/m² = 0.48 log cd-s/m²) recommended by the International Society of Clinical Electrophysiology of Vision ¹ was used, contributions of the cone photoreceptors to the photopic a-wave was only 34% at the time of the a-wave peak, and the other 66% originated from post-photoreceptoral neurons (Fig. 5, left). Our results suggest that the lower contribution of the cones to the a-waves of the photopic ERGs must be considered in patients with RP. 

There are limitations to this study. One was that we assessed the contribution of photoreceptors and post-photoreceptoral components only at the time of the a-wave peak before the drugs. However, the peak time of the a-wave depends on not only the stimulus strength but also on the presence of retinal degeneration (Fig. 1D). In addition, the a-wave can be truncated by the b-wave. To overcome this, we measured the a-wave amplitude at specific times before the b-wave intrusion (10.5 ms for 0.2 log cd-s/m², 9.5 ms for 0.7 log cd-s/m², 8.5 ms for 1.2 log cd-s/m², 7.5 ms for 1.7 log cd-s/m², and 6.5 ms for 2.2 log cd-s/m²), and calculated the percentage cone photoreceptor contribution when the animals were 4 months of age. We found that the cone photoreceptor contribution still tended to be smaller in Tg rabbits than in WT rabbits, and the differences were significant at the two lower stimulus strengths (P < 0.01, Supplementary Fig. S1A). We also measured the a-wave amplitude at a single constant time of 7 ms and calculated the percentage cone photoreceptor contribution. Again, the cone photoreceptor contribution tended to be smaller in Tg rabbits than in WT rabbits, but the difference was significant only at the highest stimulus strength (Supplementary Fig. S1B). 

In summary, our results indicate that the relative contribution of cone photoreceptors to the photopic ERG a-wave is smaller in retinas with inherited photoreceptor degeneration. These results suggest that care must be taken in interpreting the a-wave amplitudes of photopic ERGs in patients with progressive photoreceptor degeneration. 

Supplementary Materials

Figure sf01, TIF - Figure sf01, TIF 

Footnotes

 Supported by Grant-in-Aid for Scientific Research B (203904480) and Grant-in-Aid for Scientific Research C (20592075) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

Footnotes

 Disclosure: R. Hirota, Astellas Pharma Inc. (E); M. Kondo, None; S. Ueno, None; T. Sakai, None; T. Koyasu, None; H. Terasaki, None

The authors thank Duco I. Hamasaki for editing the manuscript and Michael Bach for helpful discussions.