A Morphometric Study of Light-Induced Damage in Transgenic Rat Models of Retinitis Pigmentosa

Abstract

purpose. To determine relative susceptibility to, and regional variation of, light-induced retinal damage in two rhodopsin-mutant rat models of retinitis pigmentosa, using slow- and fast-degenerating lines.

methods. Transgenic S334ter (lines 4 and 9) and P23H (lines 2 and 3) rats were reared in dim cyclic light or darkness and then exposed to intense green light for 1 to 8 hours. Sections along the vertical meridian were collected for retinal morphology and photoreceptor morphometry 2 weeks later. Unexposed transgenic and normal Sprague-Dawley rats served as the control. Mean outer segment lengths and outer nuclear layer thicknesses were analyzed as a function of position along the vertical meridian and as averages across that vector.

results. Rapidly degenerating S334ter-4 retinas, reared in dim cyclic light, exhibited no light-induced damage, whereas retinas in the other sublines sustained damage within a sensitive region in the superior hemisphere. Light-induced damage always involved loss of outer segment membrane and photoreceptors. In some cases, the retinal pigment epithelium and inner nuclear layer were also affected. Potentiation of light-induced damage by dark-rearing was increased by at least a factor of three, and in some sublines the sensitive region was enlarged to include the entire vertical meridian.

conclusions. A complex pattern of light-induced damage outcomes was identified in S334ter (sublines 4 and 9) and P23H (sublines 2 and 3) rats. The relative susceptibilities of each subline to damage by light were different, even within the same transgene, but consistent factors included a sensitive region in the superior hemisphere and potentiation by dark-rearing.

Exposure of the retina to intense visible light above a certain toxic threshold induces retinal damage, ^{1 2} leading to apoptosis of photoreceptor cells, ^{3 4} followed by apoptotic changes within the pigment epithelial cell layer. ⁵ Light-induced damage is rhodopsin mediated ^{6 7 8} and appears to involve oxidative stress. ^{9 10} There is a dose-response relationship in light-induced retinal damage, so that higher-intensity of light or longer duration of exposure each result in greater retinal damage. ^{9 10}  

The toxic threshold for photoreceptor light-induced damage is variable. Rearing animals in dim or bright cyclic light or in darkness can alter the retina’s content of important visual transduction proteins and its susceptibility to light-induced damage. ^{11 12} The level of environmental light during rearing is inversely related to the extent of light-induced damage, because the retina adjusts photoreceptor cell proteins and metabolites and the length of rod outer segments (ROSs) by the process of photostasis. ^{13 14} Ingestion of photosensitizing drugs can substantially lower the toxic threshold. ¹⁵ Intrinsic factors expressed in the retina in a circadian fashion are also thought to potentiate damage from exposure to light during the normal nighttime dark period, compared with exposure beginning during the day. ^{16 17}  

Retinitis pigmentosa (RP) is a photoreceptor-cell–specific disease caused by many different mutations in rod cell proteins, ^{18 19} including rhodopsin. ^{20 21} Typically, the first degeneration event is shortening of ROSs, ²² followed by apoptosis in rod cells. ²³ Cone cell death normally follows rod cell loss. ²⁴ A variety of hypotheses have sought to explain this rod–cone contagion, including by-products of rod death, ²⁵ loss of rod-derived trophic factors, ²⁶ deficiencies in retinoid transport, ^{22 27} and toxin transfer through the gap junctions that connect rods and cones. ²⁸  

Rhodopsin mutations include the Q344ter and the S334ter truncation mutations, which are missing the last 5 and 15 amino acids, respectively, of the C terminus. Q344ter is associated with functional abnormalities in human autosomal dominant RP, ²⁹ a form of the disease considered to be particularly severe, ³⁰ whereas S334ter is a feature of animal models. These truncated rhodopsins are mistrafficked into photoreceptors ^{31 32 33 34} and result in rod cell apoptosis. ^{23 34 35} Rescue strategies have been tested in Q344ter and S334ter animal models, ^{36 37 38 39 40 41} with mixed results. 

The P23H point mutation, a histidine-for-proline substitution in rhodopsin’s N terminus, also leads to autosomal dominant RP. ^{21 42 43} Because patients with the P23H mutation exhibit variable disease outcomes, it has been proposed that external factors play a role in setting the individual disease course. ⁴⁴ A transgenic P23H rodent animal model is available, ^{45 46 47} as is a triple-mutant model involving two additional opsin mutations (VPP or GHL mice ^{48 49 50} ). Rhodopsin mislocalization ^{50 51} and defective outer segment morphogenesis ⁵² have been reported in P23H mouse mutants, but not in rats. ³² Some P23H sublines degenerate faster than others. ⁴⁶ Rescue strategies have also been tested in P23H animal models, ^{39 47 53 54} again with mixed results. 

There is a well-known genetic predisposition to light-induced damage in mice ^{55 56} and in Royal College of Surgeons rats that otherwise express normal rhodopsins. ⁵⁷ Moreover, the P23H transgenic animal model of RP exhibits a lower phototoxicity threshold, ^{58 59 60} supporting the notion that patients with RP are more susceptible to vision loss when working in high-intensity light environments. ⁶¹ The purpose of this study was to investigate the possibility of variable susceptibility to light-induced damage in the truncation mutant S334ter and the substitution mutant P23H, using fast- and slow-degenerating sublines of each. 

In a companion study published in this issue, in which we used whole-retina assays of rhodopsin and photoreceptor DNA, we found that S334ter and P23H rats exhibit variations in photoreceptor degeneration based on transgene expression and on time-of-day of exposure to intense green light. ⁶² Such studies provide a panretinal estimate of photoreceptor function and survival, but cannot provide direct information about regional variations. Morphologic evidence to date, from Sprague-Dawley ^{16 63} and rhodopsin mutant P23H-3 rats (Coulibaly SF, ARVO Abstract #3372, 2001), strongly indicates a lower threshold to light-induced damage in the superior hemisphere of the retina. Similar findings have been reported in studies of patients with RP. ^{22 64 65 66 67} It was therefore reasonable to hypothesize that rhodopsin-mutant retinas would also exhibit regional variation in susceptibility to light-induced damage. Two morphologic hallmarks of retinal degeneration, whatever the cause, are shortened ROSs and a thinned outer nuclear layer (ONL), ^{22 63} consistent with reduced light-absorption capacity and photoreceptor cell death, respectively. Biochemical measures of changes in photoreceptor DNA content reliably predict parallel changes in mean ONL thickness, whereas changes in rhodopsin content are consistent with changes in mean ROS length, mean ONL thickness, or both. 

In this article, we examine mean ROS length and ONL thickness in the context of the history of exposure to light, transgenic rhodopsin mutations, relative transgene expression, and brief exposure to intense visible light. We found that photoreceptor degeneration was nonuniform among the different mutant sublines and between transgenes; susceptibility to light-induced damage was nonuniform among different mutant sublines, although dark-rearing consistently potentiated it in all lines; and light-induced retinal damage was nonuniform across the vertical meridian, just as we have described in normal albino rats. ¹⁷  

Methods

Animals and Rearing Conditions

All animal procedures conformed to federal and institutional guidelines, including the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Homozygous breeding pairs of transgenic S334ter rats, sublines 4 and 9, and transgenic P23H rats, sublines 2 and 3, and were kindly provided by Matthew LaVail (University of California, San Francisco, CA). Among the transgenic rats, P23H-2 and S334ter-9 exhibit a relatively slow rate of spontaneous retinal degeneration, whereas P23H-3 and S334ter-4 degenerate at a faster rate (LaVail M, unpublished data, February 1998). Heterozygous litters were produced by crossing male homozygotes with normal Sprague-Dawley (SD) females, obtained from Harlan, Inc. (Indianapolis, IN). Additional normal SD rats were used in parallel treatments as control animals. 

Transgenic rats were born and reared under dim cyclic light (20–30 lux, 8 AM to 8 PM). After weaning at age P21, rats either remained in that dim cyclic light or were moved to complete darkness, interrupted by brief periods of dim red light for animal care purposes only. 

Intense Light Treatment

At age P60, rats were dark-adapted for 16 hours to ensure optimum levels of unbleached rhodopsin. ¹⁶ Beginning at 9 AM, rats were then exposed to intense green light (1200–1400 lux, 490–580 nm) in green cylindrical Plexiglas light chambers. ^{6 16} These wavelengths overlap with rhodopsin’s peak light absorbance, which corresponds to the action spectrum peak for light-induced damage. ^{6 68}  

The duration of exposure varied depending on transgene and light-rearing history. Specifically, both S334ter sublines and SD rats received 8 hours of exposure to intense light if cyclic light reared, versus 3 hours if dark reared, whereas both P23H sublines received 4 hours of exposure to intense light if cyclic light reared, versus 1 hour if dark reared. These exposure durations were determined from preliminary experiments and published data. ^{2 16 17} Because P23H rats were found to sustain such extensive visual cell loss from 8- or 3-hour exposure to light, the duration of exposure was reduced in dark- and cyclic-light–reared P23H animals, as described. ⁶² In contrast, cyclic-light–reared SD rats exhibit no significant photoreceptor cell loss from exposures beginning at 9 AM, whereas dark-reared SD rats incur approximately 50% cell loss. ¹⁷  

Other rats from each group received no intense light treatment, serving as unexposed control subjects. Exposed and unexposed rats were maintained in constant darkness for two more weeks, to postnatal day (P)75, to permit removal of photoreceptor debris alongside recovery of viable photoreceptors. All rats were killed by CO₂ overdose. 

Morphology, Morphometry, and Statistical Analysis

Eyes were quickly enucleated for immersion fixation in buffered 2% paraformaldehyde and 2% glutaraldehyde. The cornea was sliced first and the lens removed, to admit fixative to the retina. Fixed eyes were bisected along the superior-to-inferior axis, to access the vertical meridian of each hemisphere. The superior hemisphere was marked, and a piece of cornea was left attached to the inferior aspect for orientation purposes. For consistency, only the temporal hemispheres were used in this study. Osmium postfixation, dehydration, and Spurr resin embedment proceeded by standard methods. ¹⁷ The block surfaces were sectioned only far enough to obtain a complete rim of retinal tissue around each temporal hemisphere, constituting semithin samples from the vertical meridians that were then stained with toluidine blue and examined by light microscopy. 

Before any measurements were made, the expanse of sectioned retina was inspected for proper ROS alignment. ROS and ONL data were collected from all loci in which a clear view of whole ROSs in long section, from connecting cilium to RPE, was afforded in most of the ROSs within the field of view. The block’s sectioning angle was adjusted very slightly to achieve ROS alignment of any unaligned areas, and the remaining measurements were made. In areas where ROSs were absent, ONL measurements were made after ROSs to either side of the lesion were aligned. In areas in which ROSs were uniformly disorganized (as in the S334ter subline 4 samples), we had to rely on our experience with the retina and with these samples to judge when the ONL was properly oriented for measurements. 

Sections of aligned retina were examined qualitatively to assess the integrity of the retinal pigmented epithelium (RPE), the photoreceptors, and the inner retina. For morphometry, 10 random measurements of ROS length, and 5 random measurements of ONL thickness, were made at 0.5-mm intervals extending from the optic nerve head in both superior and inferior directions. Mean ROS lengths and ONL thicknesses for each interval were calculated from three rats per treatment group and plotted as a function of eccentricity from the optic nerve head, producing a morphometric profile across the vertical meridian. Data averaged across the vertical meridian were further analyzed using a homoscedastic Student’s t-test (two-sample, assuming equal variances). 

Results

Light-Induced Damage in S334ter Retinas

A typical cyclic-light–reared, unexposed S334ter-9 retina is shown in Figure 1A . Consistent with of S334ter-9’s designation as “slow degenerating,” the retina appeared normal from the retinal pigment epithelium (RPE) to the outer plexiform layer (OPL) and was in fact indistinguishable from comparable SD retinas. ¹⁷ Two weeks after an 8-hour exposure to intense light (Fig. 1B) , the RPE was intact, but ROSs were shorter than normal, inner segments appeared swollen, and the ONL was less than half its original thickness. The inner retina appeared normal. Dark-rearing did not alter the appearance of the unexposed S334ter-9 retina (Fig. 1C) compared with its cyclic-light–reared counterpart (Fig. 1A) but, as expected, it potentiated light-induced damage (Fig. 1D) . Two weeks after a 3-hour exposure to intense light, neither the RPE nor any photoreceptors were identifiable. Instead, a broad debris zone of disorganized material lay between Bruch’s membrane and the inner nuclear layer (INL). The latter showed signs of swelling, but the inner plexiform layer (IPL) appeared normal. 

A typical cyclic-light–reared, unexposed S334ter-4 retina is shown in Figure 1E . Consistent with its designation as “fast degenerating”, severe degeneration was apparent. The RPE appeared normal, but there was severe shortening of rod inner and outer segments, along with extreme thinning of the ONL. The surviving photoreceptor nuclei included cones and scattered rods. The inner retina appeared normal (Fig. 1E) . Two weeks after an 8-hour exposure to intense light (Fig. 1F) , remarkably little distinguished this retina from its unexposed counterpart. Dark-rearing did not alter the appearance of the unexposed S334ter-4 retina (Fig. 1G) compared with its cyclic-light–reared counterpart (Fig. 1E) , but it did yield measurable light-induced damage. Two weeks after a 3-hour exposure to intense light (Fig. 1H) , the RPE was intact, but not much was left of the photoreceptor inner and outer segments. The ONL seemed a bit thinner. The inner retina appeared normal. 

The morphologic depictions in Figure 1 include evidence for outer segment degeneration and photoreceptor death. We quantified ROS length and ONL thickness in S334ter-9 retinas using morphometry across the vertical meridian (Figs. 2A 2B) . Data from unexposed retinas are indicated by solid symbols and light-induced damage changes by open symbols. A horizontal line on each plot indicates the mean value across the vertical meridians of comparable SD rats. 

In unexposed, cyclic-light–reared S334ter-9 retinas, both ROS length and ONL thickness profiles (dashed lines, solid symbols) were indistinguishable from those obtained in normal SD rats. Dark-rearing (solid lines, solid symbols) did not alter this normal morphometric profile. Light-induced damage in cyclic-light–reared S334ter-9 retinas (dashed lines, open symbols) consisted of moderate reductions in ROS length and ONL thickness that were restricted to the midsuperior retina (Figs. 2A 2B) , a region previously identified in SD rats as a sensitive region. ¹⁷ Damage in dark-reared S334ter-9 vertical meridians (solid lines, open symbols) consisted of devastating reductions in ROS length and ONL thickness (Figs. 2A 2B) . There was complete loss of ROSs and essentially complete loss of photoreceptors over most of the superior half and the central portion of the inferior half. In S334ter-9 rats, dark-rearing potentiation thus acted to increase the area damaged, which nonetheless seemed centered on the midsuperior sensitive region, and to increase the severity of the damage within this broader area. 

Morphometry of S334ter-4 retinas was quite different from that seen in S334ter-9 (Figs. 2C 2D) . Unexposed S334ter-4 retinas were substantially and equally degenerated across the vertical meridian, whether they were cyclic light reared (dashed lines, solid symbols) or dark reared (solid lines, solid symbols). Regarding shortening of the ROSs, the superior S334ter-4 retina appeared more degenerated than its inferior half. This was confirmed by statistical analysis (P < 0.05) and suggests an asymmetry in the rate of S334ter-4 transgene-induced degeneration. Cyclic-light–reared S334ter-4 retinas gave no evidence of light-induced damage, in a sensitive region or elsewhere (Figs. 2C 2D) , dashed lines, open symbols), whereas dark-reared S334ter-4 retinas sustained light-induced damage all across the vertical meridian (Figs. 2C 2D ; solid lines, open symbols). Light-induced photoreceptor death was fairly evenly distributed across the vertical meridian, but light-induced ROS degeneration was statistically more severe in the superior half, the same region in which S334ter-4 transgene-induced ROS degeneration was also accelerated. 

Light-Induced Damage in P23H Retinas

A typical cyclic-light–reared, unexposed P23H-2 retina is shown in Figure 3A . Only mild degeneration was apparent, in the form of fewer than normal rows of photoreceptor nuclei in the ONL. The remainder of the retina, from RPE to INL, appeared normal. Two weeks after a 4-hour exposure to intense light (Fig. 3B) , the RPE appeared somewhat swollen, ROSs were reduced in length, and the ONL had thinned to about half its original thickness. The inner retina appeared normal. Dark-rearing did not alter the appearance of the unexposed P23H-2 retina (Fig. 3C) compared with its cyclic-light–reared counterpart (Fig. 3A) , but, as expected, it potentiated light-induced damage (Fig. 3D) . Two weeks after a 1-hour exposure to intense light, the RPE was no longer identifiable, and a few abnormal-appearing photoreceptor nuclei rested against Bruch’s membrane. The residual nuclei in this dramatically reduced ONL was interrupted by near contact of the INL with Bruch’s membrane. Otherwise, the inner retina appeared normal. 

A typical cyclic-light–reared, unexposed P23H-3 retina is shown in Figure 3E . Mild signs of degeneration at this age were the inverse of the comparable P23H-2 retina, in that the ROS length (not ONL thickness) was reduced. The remainder of the retina, from RPE to INL, appeared normal. The result of 4 hours of exposure to intense light is shown in Figure 3F and shows extensive photoreceptor damage. The RPE was no longer identifiable, and a couple of rows of abnormal-looking photoreceptor nuclei rested against Bruch’s membrane. This outcome is remarkably similar to that found in dark-reared P23H-2 retinas (Fig. 3D) . Dark-rearing did not alter the appearance of the P23H-3 retina (Fig. 3G) compared with its cyclic-light–reared counterpart (Fig. 3E) , but, as expected, it potentiated light-induced damage (Fig. 3H) . Two weeks after a 1-hour exposure to intense light, a thin debris zone abutted Bruch’s membrane, except for one or two remaining photoreceptor nuclei. The inner retina appeared normal. 

Morphometry across the retinal vertical meridian was also plotted for the two P23H sublines (Fig. 4) . In unexposed, cyclic-light–reared P23H-2 retinas, ROS lengths were within normal ranges in the superior and inferior halves (Fig. 4A , dashed line, solid symbols), whereas ONL thickness was slightly reduced in the superior half (Fig. 4B) . This suggests a slight asymmetry in the progress of P23H-2 transgene-induced degeneration, but that was not confirmed by statistical analysis. Light-induced damage was not uniform across the P23H-2 retina’s vertical meridian. Moderate reductions in ROS length (Fig. 4A , dashed line, open symbols) and ONL thickness (Fig. 4B , dashed lines, open symbols) were restricted to the superior half, much as was observed in cyclic-light–reared S334ter-9 retinas. Dark-rearing increased the severity of losses of ROS membranes (Fig. 4A ; solid line, open symbols) and photoreceptors (Fig. 4B) , but the sensitive region did not extend into the inferior retina. 

Uniformly across unexposed, cyclic-light–reared P23H-3 retinas, ROSs were shorter than normal (Fig. 4C ; dashed line, solid symbols), whereas ONL thickness was normal (Fig. 4D) . Dark-rearing did not change these findings (Fig. 4 CD, solid lines, solid symbols). Exposure to intense light of cyclic-light–reared P23H-3 retinas resulted in dramatic damage, but only in the superior half of the vertical meridian (Figs. 4C 4D , dashed lines, open symbols). In contrast, dark-reared P23H-3 retinas sustained near complete damage all across the vertical meridian, amounting to complete loss of ROSs (Fig. 4C , solid line, open symbols) and near-complete loss of photoreceptors (Fig. 4D , solid line, open symbols). In P23H-3 rats, dark-rearing potentiation thus acted to increase the area damaged to include the entire vertical meridian and to increase the severity of the damage within this broader area to complete destruction of the photoreceptors. 

Quantitative Morphometric Analysis

To compare data from the four rhodopsin-mutant sublines in this study, mean ROS length and mean ONL thickness, averaged across the vertical meridian, are shown in Table 1 . Data from age-matched SD rats are also provided. 

By age P75, transgene-based degeneration was nondetectable in S334ter-9 and P23H-2 rats. In both slow-degenerating sublines, ROS length and ONL thickness were not significantly different from those of age-matched SD rats. In contrast, the fast-degenerating S334ter-4 transgene induced a 67% reduction in mean ROS length and a 47% reduction in mean ONL thickness (both P < 0.001). The P23H-3 transgene induced only a 27% reduction in mean ROS length (P < 0.001), with no significant effect on ONL thickness. The same was found in dark-reared rat retinas unexposed to intense light. 

Light-induced degeneration of cyclic-light–reared SD retinas was statistically insignificant but, when SD rats were dark reared, there was a 68% reduction in mean ROS length and a 47% reduction in mean ONL thickness (both P < 0.001). Light-induced damage in cyclic-light–reared S334ter-4 retinas was also insignificant but, after dark-rearing, it amounted to a 59% reduction in mean ROS length (P < 0.01) and a 42% reduction in mean ONL thickness (P < 0.001). When S334ter-9 rats were reared in dim cyclic light, light-induced retinal damage included an 11% reduction in mean ROS length and a 13% reduction in mean ONL thickness (both P < 0.05), but dark-rearing potentiated these reductions to 64% and 52% (both P < 0.001), respectively. 

In P23H-2 retinas after rearing in dim cyclic light, the light-induced reduction in mean ROS length was 13% (P < 0.05). Dark-rearing exacerbated this reduction to 45% (P < 0.001). There was no light-induced change in the P23H-2 retina’s mean ONL thickness after rearing in dim cyclic light, although close examination of the P23H-2 retina’s morphometric plot (Fig. 4B) revealed a focal reduction in the superior retina. When only that sensitive region was analyzed, exposure to light resulted in significant loss of P23H-2 photoreceptor nuclei (P < 0.05). Dark-rearing potentiated photoreceptor loss across the vertical meridians of P23H-2 retinas, resulting in a 30% reduction in mean ONL thickness (P < 0.05). Light-induced damage in P23H-3 retinas was moderate after rearing in dim cyclic light, but was devastating after rearing in darkness. Reductions in mean ROS length were 27% (P < 0.05) versus 92% (P < 0.001), respectively, and reductions in mean ONL thickness were 24% (P < 0.01) versus 74% (P < 0.001), respectively. 

Discussion

Of the S334ter rats reared to age P75 but unexposed to intense light in our study, S334ter-4 rats degenerated considerably faster than S334ter-9 animals. The latter was indistinguishable from normal SD retina, whereas approximately half the rods were lost from S334ter-4 retinas. We further identified an asymmetry in transgene-induced S334ter-4 retinal degeneration, in that visual cell outer segment membrane loss was more severe in the superior hemisphere. In unexposed P23H rats at this age, subline 3 exhibited transgene-induced reduction of ROS length, whereas subline 2 exhibited loss of photoreceptors instead. Overall, however, the P23H-3 degeneration was more significant than the P23H-2 degeneration, consistent with their designations as fast- and slow-degenerating, respectively. 

Recent functional studies on human patients with RP with identical or similar rhodopsin mutations have revealed considerable variability in the spatial distribution of photoreceptor degeneration across the retina (Coulibaly SF, ARVO Abstract #3372, 2001) ^{61 64 65 66 67 69 70 71 72} It was therefore of interest to determine whether these transgenic rats would also exhibit such nonuniform patterns of rod degeneration. Indeed, we found that unexposed S334ter-4 retinas were significantly more degenerated in the superior half of the vertical meridian, in terms of both ROS membrane deterioration and ONL loss. Our data are also suggestive of accelerated superior retinal degeneration caused by S334ter-4 and P23H-2 photoreceptor loss. It may be necessary to rear the animals to a later postnatal age to better observe asymmetries in transgene-induced degeneration. 

Investigators in prior studies reporting on the spatial variability of RP photoreceptor loss have proposed that light-induced damage may play a role in the uneven progression of disease across the retina. ^{61 67} Thus, it was of interest to determine in our rat model the spatial distribution of light-induced damage juxtaposed to the degenerative effects of each transgene. In the current study, our data strongly support the notion of asymmetric light’s effects on the RP mutant retina. We found clear evidence of a sensitive region in the superior hemisphere of all four transgenic sublines, much as previously described in SD rat retinas. ¹⁷  

Compared with normal SD rat retinas, the sensitive regions in three of the four transgenic rats were considerably more susceptible to light-induced damage from exposures beginning at 9 AM. ^{16 17} However, the relative susceptibility of each subline to light-induced damage was quite different, even within the same transgene. When reared in dim cyclic light, S334ter-4 retinas appeared completely resistant to exposure to intense light that killed or damaged photoreceptors in the superior sensitive region of S334ter-9 retinas. This resistance may be attributable to the advanced stage of photoreceptor degeneration and loss that occurred in S334ter-4 retinas by P75. Cyclic-light–reared P23H-2 retinas sustained light-induced damage only in the rod photoreceptors, whereas P23H-3 retinas also lost the RPE in the sensitive region. Thus, in our experiments, the severity of light-induced retinal damage within the same sensitive region was variable. 

In all cases, dark-rearing potentiated light-induced damage, by at least a factor of three (Table 1) . This potentiation always manifested as more complete destruction of the retina in the superior hemisphere. For example, in S334ter-9 and P23H-2, light-induced damage in cyclic-light–reared rats was limited to the photoreceptors, whereas in dark-reared rats it included destruction of the RPE. Dark-reared S334ter-9 retinas also showed signs of swelling in the INL. 

In both S334ter sublines, dark-rearing resulted in an expansion of the light-damaged area into the inferior hemisphere. However, dark-rearing followed by identical exposure to intense light potentiated damage in the two P23H sublines in substantially different ways. In P23H-2, damage to rods was more severe, but remained in the superior hemisphere, whereas in P23H-3, damage to rods was essentially complete and included the entire vertical meridian. 

The spatial variability of light-induced damage observed in our rat experiments is in general agreement with recent studies of human patients with RP (Coulibaly SF, ARVO Abstract #3372, 2001). ^{18 19 22 67 72} These human patterns may be due to the nonuniform spatial distribution of survival or toxic factors, ²² but they are also thought to be influenced by environmental light over the life of the affected individual. ^{1 2 61 67 72} In particular, relative sparing of far peripheral regions of human RP retina is a frequent observation ⁷³ and has been attributed to shading by the pigmented iris and ciliary body. ⁷⁴ We found no evidence of such sparing in either “fast” subline tested (S334ter-4 and P23H-3). It is reasonable to attribute the lack of sparing to the absence of ocular pigmentation in our albino mutant animals. 

With the exception of dark-reared S334ter-9, the INL and IPL of light-damaged transgenic rat retinas appeared normal. Recently, however, Fariss et al. ⁷⁵ have reported substantial remodeling of the inner retina in human RP, something that our study did not address. However, examination of our transgenic specimens for evidence of synaptic remodeling is underway. 

In summary, our study supports the notion that, compared with normal SD rat retinas, transgenic rat retinas are significantly more at risk for regional loss of visual sensitivity after exposure to intense light. If light-induced damage exacerbates mutant rhodopsin gene-based degenerations, the various patterns of RP visual field loss may be explained by the variety of light insults experienced by different patients during their lifetimes. The fact that human retinas receive more intense light from overhead sources has been suggested to explain the altitudinal pattern of degeneration in humans with the P23H mutation. ⁶¹ In that scenario, the incidence of exposure to light sets the boundaries of the hypothesized (inferior) sensitive region. In our rat experiments, intense light was uniformly delivered from circular fluorescent bulbs. ¹⁶ Moreover, exposure was consistent between sublines of the same transgene. With these environmental factors controlled for in our experiments, the variability suggests that intrinsic factors, or genetic predisposition ⁵⁵ related to transgene expression, set the boundaries of the sensitive region. This notion is supported by findings in the companion manuscript that show enhanced susceptibility to light-induced damage of S334ter-9 retinas at all times of the day. ⁶² Together, retinotopic data from morphometric and functional studies of patients with RP and animal models underscore the need for mutation-independent therapies for this genetically ⁷⁶ and spatially (Coulibaly SF, ARVO Abstract #3372, 2001) heterogeneous form of retinal degeneration. 

 Presented in part at the annual meetings of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2001 and 2002.

 Supported by National Eye Institute Grant EY-01959 (DTO); Mary Petticrew, Springfield, OH (DTO); the Ohio Lions Research Foundation (DTO); the University of Wisconsin Oshkosh Faculty Development Program (DKV); and the University of Wisconsin Oshkosh Graduate School (SFC).

 Submitted for publication July 12, 2002; revised August 27, 2002; accepted September 10, 2002.

 Disclosure: D.K. Vaughan, None; S.F. Coulibaly, None; R.M. Darrow, None; D.T. Organisciak, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Representative histology of the midsuperior retina in age P75 mutant S334ter rats, 2 weeks after exposure. In each pair, the unexposed retina is on the left and the exposed retina is on the right. (A, B) Cyclic-light–reared subline 9; (C, D) dark-reared subline 9; (E, F) cyclic-light–reared subline 4; and (G, H) dark-reared subline 4. Arrows: region between the OLM and the RPE. Scale bar, 20 μm.

Figure 2.

 

Morphometry across the vertical meridians of slow- and fast-degenerating retinas in S334ter transgenic rats reared in dim cyclic light (dashed lines) or in darkness (solid lines). Closed symbols: age P75 rats unexposed to intense light; open symbols: rats exposed at P60 and then maintained in darkness for 2 weeks. Sample size was three rats per condition and 10 ROS measurements or 5 ONL measurements per data point. Horizontal line: mean data from normal SD rats. (A) S334ter-9 ROS length; (B) S334ter-9 ONL thickness; (C) S334ter-4 ROS length; and (D) S334ter-4 ONL thickness.

Figure 3.

 

Representative histology of the midsuperior retina in age P75 P23H mutant rats. For each pair, the unexposed retina is on the left and the exposed retina is on the right. (A, B) Cyclic-light–reared subline 2; (C, D) dark-reared subline 2; (E, F) cyclic-light–reared subline 3; and (G, H) dark-reared subline 3. Arrowheads: Bruch’s membrane in severely degenerated specimens. Scale bar, 20 μm.

Figure 4.

 

Morphometry across the vertical meridian of P23H transgenic rats, reared in dim cyclic light (dashed lines) or in darkness (solid lines). Closed symbols: age P75 rats not exposed to intense light; open symbols: rats exposed at P60 and then maintained in darkness for 2 weeks. Sample size was three rats per condition and 10 ROS measurements or 5 ONL measurements per data point. Horizontal line: mean data from normal SD rats. (A) P32H-2 ROS length; (B) P23H-2 ONL thickness; (C) P23H-3 ROS length; and (D) P23H-3 ONL thickness.

Table 1.

 

Morphometric Data Averaged Across the Exposed or Unexposed Vertical meridian of Each Treatment Group, under Both Rearing Conditions.

	Cyclic Light-Reared			Dark-Reared
		Unexposed	Exposed	Unexposed	Exposed
ROS Length
SD	31.7 ± 6.0	32.2 ± 5.1	33.1 ± 7.5	10.8 ± 10.2, †
S334ter-9	31.1 ± 4.0	27.8 ± 4.5, †	30.0 ± 3.0	11.0 ± 11.5, †
S334ter-4	10.5 ± 6.8*	10.6 ± 6.5	12.4 ± 6.3*	5.1 ± 5.4, †
P23H-2	29.0 ± 5.0	25.3 ± 5.2, †	29.1 ± 3.2	16.1 ± 12.5, †
P23H-3	23.3 ± 1.8*	17.4 ± 10.2, †	22.4 ± 3.7*	0.9 ± 2.3, †
ONL Thickness
SD	32.5 ± 4.1	33.3 ± 2.6	33.2 ± 4.2	17.7 ± 11.8, †
S334ter-9	32.9 ± 4.2	28.8 ± 4.6, †	32.4 ± 3.8	15.6 ± 12.0, †
S334ter-4	17.2 ± 4.6*	16.6 ± 3.7	18.4 ± 3.6*	10.8 ± 3.5, †
P23H-2	28.5 ± 3.0	27.8 ± 6.0	30.9 ± 4.3	21.7 ± 9.3, †
P23H-3	32.5 ± 2.5	24.8 ± 9.0, †	29.5 ± 3.6	7.7 ± 4.6, †