Light-Induced Retinopathy: Comparing Adult and Juvenile Rats

Abstract

purpose. To investigate the effect of chronic exposure to a bright, luminous environment, starting at the opening of the eyes, on the retinal structure and function of the suckling rat.

methods. Juvenile Sprague–Dawley rats were exposed to 10,000 lux, for varying lengths of time between postnatal day (P)14 and P34. Results were compared with those obtained from adult rats exposed to the same light intensity and for the same duration. Electroretinograms (ERGs) were recorded at 1 month and 2 months of age, after which the retinas were harvested for histologic analysis.

results. In juvenile rats, the severity of light-induced retinopathy (LIR) depended not only on the duration of the exposure but, more important, on the age of the rat at the onset of exposure. For example, in adult rats, 6-day exposure reduced the thickness of the outer nuclear layer (ONL) to less than 18% of normal, whereas in juvenile rats, 6-day exposure between P14 and P20 reduced it to 50% compared with 27% after exposure between P28 and P34. An adultlike effect could only be evidenced in rats exposed at the end of the first postnatal month (P28-P34). A similar age-dependent effect was also noted on the electroretinogram.

conclusions. These results show that, compared with the mature retina, the developing retina appears to be relatively preserved from the devastating consequences of exposure to bright light. This window of resistance to light damage gradually weakens as the juvenile rat approaches its 1-month anniversary.

Numerous studies ^{1 2 3 4 5 6 7 8 9} have demonstrated the deleterious effects of exposure to bright light on adult retinas. Severity of the retinal damage depends essentially on four factors: the duration of the exposure, the intensity and the wavelength of the luminous environment, as well as the type of exposure (cyclical or chronic). Although each method used to generate this retinopathy has its own particularities, a general degenerative pattern can nevertheless be described. ¹⁰ Light damages are rhodopsin mediated and, thus, are initiated in the outer segments of the photoreceptors, where the light stimulus is transduced. ⁹ The first signs of degeneration are detected by electron microscopy as early as 1 hour after light exposure. ^{11 12 13} The lamellar structure of the saccules at the most external tip of the outer segments becomes disorganized, leading to membrane separation and vacuolization. The slender outer segments become rounded and then detach from the atrophied inner segments. Some pyknotic photoreceptor nuclei appear in the outer nuclear layer (ONL) along with anomalies, such as an increase in the number of lysosomes, in the retinal pigment epithelium. In the most extreme cases, the ONL completely disappears, whereas the inner nuclear layer (INL), the inner plexiform layer (IPL), and the retinal ganglion cell layer (RGC) of the retina do not appear to be affected. ^{10 11 13 14} These structural alterations are also associated with functional losses. For example, after exposure to a constant light of 1700 lux for as few as 24 hours, 8-week-old rats showed almost complete abolition of the electroretinogram (ERG) response. This was accompanied by marked thinning of the ONL. ^{1 2 4 15} The processes of light damage described here are those reported for adult animals. Little is known about juvenile rats exposed to the same bright light. 

The development of the visual system (retinal neurogenesis and synaptogenesis) begins early in embryonic life and lasts until the first postnatal days of life. ^{16 17} In rodents, the somas of rods and cones are already present at birth; in rats, retinogenesis is achieved near postnatal day (P)12. ¹⁶ In rats, rhodopsin is detectable for the first time at P7, and an ERG can be recorded at approximately P12. ^{18 19} In a previous study, we showed that exposure to a bright, luminous environment before eye opening in neonatal rats caused only subtle functional and structural retinal damage (Joly S, et al. IOVS 2003;44:ARVO E-Abstract 1870). In comparison, adult rats subjected to the same luminous environment for as few as 6 days had a nearly abolished ERG, and their retinas were devoid of photoreceptors and ONL. These results suggest that the developing retina might benefit from relative immunity to exogenous stress agents such as bright light (Joly S, et al. IOVS 2003;44:ARVO E-Abstract 1870). 

The purpose of this study was, therefore, to determine the consequences of chronic exposure to a bright, luminous environment, starting at eye opening, on the retinal structure and function of juvenile “suckling” rats. Our results suggest that the retina of the juvenile rat has remarkable resistance to light damage but that this gradually wears off by the end of the first month of life. 

Methods

Animals

Experiments were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the McGill University–Montreal Children’s Hospital Animal Care Committee. Pregnant Sprague–Dawley rats (Charles River Laboratories, St-Constant, Quebec, Canada) were ordered at day 15 of gestation and were kept under a normal lighting environment of 80 lux (12 hours dark/12 hours light). During the exposure period, pups and their mothers (juvenile rats were all weaned at the end of the exposure) were put into clear polymethyl methacrylate cages and were exposed for 12 hours each day to a bright, luminous environment of 10,000 lux, as measured at the floor level of each of the cages with a light meter (IL 1700 Research; International Light, Newburyport, MA). The setup consisted of a dozen 48-inch long fluorescent tubes (Sylvania, 34 W, cool-white neon light; Mississauga, ON, Canada) covering the top and two sides of each cage. Rats were exposed from eyelid opening (P14) until P17 (n = 11), P20 (n = 4), and P28 (n = 17) or from P20 to P28 (n = 10) or P28 to P34 (n = 10). After exposure, the rats were removed from the bright, luminous environment and were maintained in the normal lighting conditions of the animal care facility (12-hour dark/12-hour light at 80 lux) until P60. Similarly, adult female albino rats were also exposed to the same bright, luminous environment (12-hour dark/12-hour light at 10,000 lux) for 1 (n = 9), 3 (n = 3), 6 (n = 4), or 14 (n = 2) consecutive days or were kept as controls (n = 10) in the normal lighting conditions of the animal care facility. Female rats were chosen so that the data they generated could be added to those obtained from the mothers that had been exposed with their pups. After exposure, the rats were returned to the normal luminous environment of the animal care facility. Control age-matched juvenile (n = 30) and adult female (n = 6) rats were raised under the normal luminous environment (12-hour dark/12-hour light at 80 lux) of the animal care facility. 

Electroretinography

The ERGs were first recorded between P30 and P37 and were repeated at P60. For adult rats, ERGs were obtained once 1 month after the beginning of the light exposure. ERG recordings were performed according to a protocol previously reported. ^{20 21} Briefly, after a 12-hour period of dark adaptation, rats were anesthetized, under dim-red light illumination, with intramuscular injection of ketamine (85 mg/kg) and xylazine (5 mg/kg). Pupils were dilated with drops of 1% cyclopentolate hydrochloride. The rats were then positioned, lying on their sides, in a recording chamber of our design ²² in which the top part housed the rod desensitizing background light and a photostimulator (PS 22; Grass Instruments, Quincy, MA). The ERGs were recorded with a fiber electrode (DTL; 27/7 X-Static silver-coated conductive nylon yarn; Sauquoit Industries, Scranton, PA) maintained on the cornea with 2% methylcellulose, which also prevented desiccation of the cornea. Reference (E5 disc electrode; Grass Instruments) and ground (E2 subdermal electrode; Grass Instruments) electrodes were inserted in the mouth and subcutaneously in the tail, respectively. Broadband ERGs (bandwidth, 1–1000 Hz; 10,000×; 6-db attenuation; P-511 amplifiers; Grass Instruments) and oscillatory potentials (bandwidth, 100-1000 Hz; 50,000×; 6-db attenuation) were recorded simultaneously with a data acquisition system (AcqKnowledge; Biopac MP 100 WS; Biopac System Inc., Goleta, CA). Scotopic luminance-response functions were obtained in response to progressively brighter stimuli spanning a 6 log-unit range (PS 22 photostimulator [Grass Instruments]; interstimulus interval, 10 seconds; average, 5 flashes; 0.3 log-unit increment; maximal intensity, 0.6 log cd · s/m²). This was followed by the recording of the photopic (cone-mediated) signal (photopic background, 30 cd/m²; flash intensity, 0.9 log cd · s/m²; interstimulus interval, 1 second). The recording of the photopic responses was performed 20 minutes after the opening of the background light to avoid the previously reported light-adaptation effect. ^{23 24}  

Retinal Histology

At the end of the last ERG session (at P60), the rats were killed with anesthesia overdose followed by intracardiac perfusion of 3.5% glutaraldehyde in 0.067 M Sorensen phosphate buffer (pH 7.4). Eyes were enucleated and immersed for 1 hour in 3.5% glutaraldehyde. Corneas and lenses were then removed, and the remainder of each eye was fixed in 1% osmium tetroxide solution for 4 hours. After sequential dehydration in baths of 50%, 70%, 90%, 95%, and 100% ethanol, the specimens were embedded in Epon resin (Mecalab, Montreal, Canada). Semithin sections, 0.7 μm thick, were mounted on glass slides and stained with 0.1% toluidine blue. Images were examined and captured with a 1300 monochrome 10-bit CCD camera (Retiga; QImaging, Burnaby, BC, Canada) mounted on a microscope (Eclipse E-600; Nikon, Tokyo, Japan) with image analysis software (Northern Eclipse; Empix Imaging Inc., Mississauga, ON, Canada). Except for the data shown (see Fig. 5 ), thicknesses of the retinal layers (measured with Image-Pro Plus Imaging Software; Media Cybernetics, Silver Spring, MD) were obtained from measurements made on eight retinal sections taken in the superior hemisphere of the retina (control juveniles, n = 3; P14-P20; n = 3; P14-P28, n = 4; P20-P28, n = 3; P28-P34, n = 3; control adults, n = 4; adults exposed for 1 day, n = 6; for 3 days, n = 3; for 6 days, n = 2; and for 14 days, n = 2) because the superior retina was previously shown by others to be the most affected by exposure to bright light. ²⁵ On each section, retinal layer thicknesses were measured in five randomly chosen sectors. 

Data Analysis

Analysis of the ERG waveforms was performed according to standard practice. ^{20 21 26} The amplitude of the a-wave was measured from baseline to the most negative trough, whereas that of the b-wave was measured from the trough of the a-wave to the most positive peak of the retinal response. Peak times were measured from flash onset to peak. Scotopic luminance-response function curves were obtained by plotting the amplitude of the b-wave against the intensity of the flash used to evoke the response. From this curve, the V _max value (maximum rod-mediated response) was defined as the amplitude of the b-wave measured when an a-wave appeared clearly on the ERG (e.g., a-wave threshold) and the (mixed) rod a- and b-waves referred to the ERG evoked to the highest flash luminance available. The amplitude of each oscillatory potential (OP) was measured from the preceding trough to the peak. OP amplitudes were reported as the sum of the OPs (SOPs; OP₂ + OP₃ + OP₄ + OP₅ in scotopic conditions and OP₂ + OP₃ + OP₄ in photopic conditions). OP₁ was not included in the analysis because it is often difficult to identify. Statistical analysis was performed using one-way ANOVA and Tukey post hoc test (GraphPad Instat Software Inc., San Diego, CA). 

Results

Effects of Exposure to Bright Light on the Function of Adult and Juvenile Retinas

Representative scotopic and photopic ERG and OP responses obtained from adult and juvenile rats exposed or not to the bright, luminous environment are presented at Figures 1(scotopic) and 2(photopic). Group data are reported in Table 1 . 

In adults, 1-day exposure (Fig. 1B)already caused significant (P < 0.001) reductions in mixed (rod-cone) a- and b-wave amplitudes to 31% and 41% of normal, respectively, compared with 48% (P < 0.001) for the rod V _max. Similarly, the cone b-wave was also reduced to 64% of control amplitude (P < 0.05; Table 1 ; Fig. 2A ). Increasing the exposure to 3 consecutive days (Figs. 1C 2A ; Table 1 ) did not significantly add to the amplitude attenuation already measured after 1 day of exposure. The scotopic mixed (rod-cone) a-wave completely disappeared after 6 days of exposure to bright light (Fig. 1D) , whereas residual mixed (rod-cone) b-wave (6% of normal; Table 1 ) and photopic b-wave (14% of control, P < 0.001; Fig. 2A ; Table 1 ) could still be obtained after more than 14 days of exposure (Figs. 1E 2A , arrows). 

Contrary to the reaction in adult rats, juvenile rats exposed to the same bright environment for 3 (P14–P17 exposure interval) consecutive days did not develop significant ERG anomalies, irrespective of the ERG component considered (Figs. 1G 2B , second tracing; Table 1 ; P > 0.05). Increasing the light exposure to 6 days (P14-P20; Fig. 1H ) reduced the amplitude of the mixed (rod-cone) a- and b-waves to 51% (P < 0.001) and 67% (P < 0.01) of control, respectively (Table 1) , but the rod V _max and the cone b-wave (Fig. 2B)were not significantly attenuated. As the exposure interval lengthened, there was a gradual attenuation of the ERG signal (scotopic ERG, Figs. 1I IJ 1K ; photopic ERG, Fig. 2B ; Table 1 ). After the longest exposure regimen tested (P14–P28), the amplitude of the mixed (rod-cone) a-wave was less than 16% of normal whereas that of the mixed (rod-cone) b-wave, rod V _max and cone b-wave were 44%, 48%, and 54% of normal, respectively (P < 0.001; Table 1 ). Interestingly, compared with adults exposed for 6 days to the bright, luminous environment, juvenile rats exposed for the same duration still produced ERG responses, albeit significantly smaller than normal. The extent of these reductions depended on the age at which the exposure started, as exemplified in Figure 1and in Table 1 . For example, 6-day exposure between P28 and P34 generated significantly more severe retinopathy (according to ERG measurements) than that obtained after 6-day exposure between P14 and P20 (reductions to 3% and 51% of the control value for the a-wave and to 11% and 67% for the b-wave during P28–P34 and P14–P20, respectively). Similarly, 6-day exposure between P28 and P34 caused more severe retinopathy than that obtained after 8-day exposure between P20 and P28 (reduction to 17% of the control value for the a-wave and to 31% for the b-wave after P20–P28 exposure). Irrespective of the duration of exposure (adults) or the exposure regimen considered (juvenile), scotopic ERGs were always significantly (P < 0.05; compare mixed rod-cone b-wave and cone b-wave measurements at Table 1 ) more attenuated than photopic ERGs, suggesting that rods are significantly more susceptible to bright-light insult and that this retinopathy mostly (or initially) affects rod function. 

The oscillatory potentials (scotopic OPs, Fig. 1L ; photopic OPs, Fig. 2C ) were also affected after exposure to the bright, luminous environment. Six days of continuous exposure completely abolished the scotopic and photopic SOPs of adult rats (Table 1) , whereas there were gradual decreases in the amplitude of the scotopic (Fig. 1L)and photopic OPs (Fig. 2C)in juvenile rats, effects also measurable with the data found in Table 1 . However, for an equal duration of exposure, the OPs of juvenile rats were better preserved than those of adult rats (Table 1) . For instance, 3 days of exposure in adult rats reduced the scotopic and photopic SOPs to 50% and 83% of normal amplitude, respectively. Similarly, 6-day exposure between P14 and P20 reduced the rod SOPs to 71% of normal amplitude but had no measurable impact on the photopic SOPs, whereas 6-day exposure between P28 and P34 reduced the scotopic and photopic SOPs to less than 8% and 19% of normal amplitudes, respectively. Again, irrespective of the age of the rat, rod-mediated OPs were always significantly more affected than cone-mediated OPs. 

Effects of Exposure to Bright Light on the Structure of Adult and Juvenile Retinas

In adult rats, all outer retinal layers rapidly and significantly decreased after as little as 1 day of exposure (outer segment [OS], 15% of normal; inner segment [IS], 14%; ONL, 32%; outer plexiform layer [OPL], 10%; P < 0.001; Figs. 3B 4A ). Increasing the exposure to 3 days (Fig. 4A)did not significantly add to what was readily observed after 1 day of exposure, but increasing it to 6 days (Figs. 3D 4A)led to the complete disappearance of the OS, IS, and OPL. The ONL was reduced to 17% of normal (P < 0.001), a value that was further attenuated to 13% of control thickness after 14 days of exposure (Figs. 3E 4A) . Layers below the OPL were never (except for a slight but significant increase in thickness of the ganglion cell layer [GCL] after 6 days of exposure) altered after exposure to the bright, luminous environment (Fig. 4A) . 

In juvenile rats, most of the damage resulting from exposure to bright light was also concentrated in the outer retinal layers. Although the thinning of the outer retinal layers (OS, IS, ONL, OPL) measured after the P14 to P20 exposure was not significantly enhanced (Figs. 3G 4B)with longer exposure intervals (P14–P28, P20–P28), a different picture emerged after the P28 to P34 exposure—the nearly complete destruction of all the retinal layers above the inner nuclear layer (Fig. 3J) . As indicated in Figure 4B , OS, IS, and OPL completely disappeared, but a residual (13% of normal thickness; P < 0.001) ONL remained. This roughly represented a one nucleus–thick layer, as illustrated in Figure 3J . 

Results presented in Figure 4Balso suggest that the more aggressive exposure regimens significantly increased the thickness of the INL. This phenomenon could not be attributed to the normal maturation process of the retina because, as illustrated in Figures 3K -M and Figure 4C , the thickness of the retina normally decreases as the rat ages. This was most noticeable for the ONL, OPL, INL, and GCL. In contrast, the OS lengthens during the same period, and the content of the ONL nuclei appears fragmented in rats aged 14 days (Fig. 3N)compared with older rats (Fig. 3O) . As the ONL becomes thinner, so does the INL (Fig. 4C) , contrary to what is observed in retinas of juvenile rats exposed to the bright-light source (Fig. 4B) . To explore this further, we compared (Figs. 4D 4E)the thickness of the ONL with that of the INL taken from the retinas of juvenile rats after the different exposure regimens (Fig. 4B) . Similar correlations were obtained from the data taken from the adult retinas (Fig. 4A ; normal and exposed) resulting from the normal maturation of the juvenile rat retinas (Fig. 4C) . As expected from the data shown (Fig. 4C) , a high correlation coefficient associates the two events (Fig. 4E , solid line), indicating that as the juvenile retina matures, the thicknesses of both nuclear layers are reduced. This relationship was strikingly different from that evidenced by the data obtained from the juvenile rats exposed to bright light, in which the gradual thinning of the ONL appears to be “compensated” by a proportional increase in thickness of the INL. Again, the two events were highly correlated (Fig. 4E , dotted line). A similar “compensation phenomenon” could not be documented with the adult (exposed) retina (Fig. 4D)in which the thickness of the INL remained the same despite the progressive disappearance of the ONL, thus explaining the poor correlation coefficient obtained. 

Previous reports ²⁵ have documented a regional difference in the reaction of the retina to bright light by which the superior (dorsal) retina was always significantly more affected (at histologic examination) than the inferior (ventral) retina. We examined whether a similar differential effect could also be documented in our juvenile rats (Fig. 5) . If one only takes into consideration the thickness of the ONL, a striking difference in reaction to light between the superior and inferior retinas clearly stands out. In adults, after 6 consecutive days of exposure, complete destruction of the superior ONL was observed (Fig. 5D) , but a two nuclei–thick ONL persisted in the inferior retina (Fig. 5H)compared with 12.5 ± 0.7 rows of nuclei in normal retina. In comparison, after exposure from P14 to P20, the superior ONL of the juvenile retina (Fig. 5B ; also normally composed of 12 rows of nuclei) is reduced to 5.3 ± 0.6 rows of nuclei (42% of normal) compared with 8.7 ± 1.2 rows (70% of normal) for the inferior ONL (Fig. 5F ; n = 3; Student t-test, P < 0.05). Further reductions to 3.8 ± 0.5 (30% of normal) and 6.5 ± 0.6 (52% of normal) rows of nuclei for the superior (Fig. 5C)and inferior ONL (Fig. 5G) , respectively (n = 4; Student t test, P < 0.001), are obtained after the P14 to P28 regimen. The hemispheric difference in the kinetics of this light-induced retinal degeneration is summarized at Figure 5I , in which the numbers of nuclei in the inferior and superior ONL of juvenile (black squares) and of adult (white squares) retinas are compared. Each data point represents the cellular count obtained after light exposure of predetermined duration, as indicated numerically above each data point. Light-induced retinal degeneration was slower (i.e., lower slope value) in young (dashed line) than in older (solid line) retinas. Moreover, the y-intercept values reveal that a completely destroyed superior ONL corresponds to four layers of nuclei inferior to the ONL in the juvenile retina compared with approximately two rows of nuclei in the adult retina, stressing once more the increased resistance of the juvenile retina to the light-induced retinopathy. 

Maturation of the Light-Damaged Retina

In a previous study (Joly S, et al. IOVS 2003;44:ARVO E-Abstract 1870), we presented ERG evidence suggesting that the retinopathy induced after postnatal exposure to a bright, luminous environment was degenerative rather than destructive. With time away from the end of the exposure period, the retinal potentials became more severely attenuated. To examine whether a similar trend could be observed in the present study, we devised the ERG maturation index parameter, which represents the ratio between ERG amplitudes measured at P60 compared with the same ERG component measured at P30 (i.e., P60/P30 × 100%). Results from this data manipulation are illustrated in Figure 6 . As we showed previously, ^{20 21} the normal maturation of the rat retina (P30–P60) is accompanied by a reduction in ERG amplitudes. For exposure regimens that were terminated before P28, the maturation indexes were similar (albeit smaller in most cases) to those measured in normal rats. In contrast, for exposure regimens that ended at or after P28 (especially P14–P28, P20–P28, and P28–P34), the maturation indexes were significantly larger than those measured in normal rats, suggesting that the normal course of the retinal maturation process was significantly altered as a result of late postnatal exposure to a bright, luminous environment. 

Our findings thus clearly suggest that compared with those of adult rats, the retinas of juvenile rats exhibit remarkable resistance to light damage. This is best summarized at Figure 7 , in which the effect of exposure to bright light on the structure (thickness of the ONL) and function (average of ERG parameters) of the adult (Fig. 7A)and juvenile (Fig. 7B)retinas are compared. Although in adult rats there is clear evidence of a dose–response effect (structure and function), the latter is less obvious in data obtained from the juvenile rat. In juvenile rats, the effect of exposure to bright light on retinal structure or function depends not only on the duration of exposure but also on the degree of retinal maturation at the onset of exposure. 

Discussion

Results presented here support the claim that after exposure to a bright, luminous environment, the retina of the juvenile rat is significantly less damaged (structure and function) than that of the adult rat retina. Of all the ERG components considered, the rod a-wave (adult and juvenile responses) is most affected by the intense luminosity of the environment (Table 1) , in agreement with a previous claim that light-induced retinopathy (LIR) is rhodopsin mediated. ⁹ The next most affected (juvenile and adult responses) are the rod b-wave, rod V _max, and the cone b-wave, suggesting that the sequence of events leading to the full expression of this LIR is the same irrespective of the age of the rat. Unfortunately, for reasons that remain to be explained, rat retinas do not generate recordable photopic (cone-mediated) a-waves (Fonteille VL, et al. IOVS 2005;46:ARVO E-Abstract 2246). ²⁷ Consequently, it is impossible, based on these data, to state whether the cone photoreceptor (structure, function, or both) itself is also directly impaired by the bright, luminous environment. However, the fact that the cone b-wave is significantly reduced strongly suggests that this might be the case, though an indirect effect (such as by way of the rod pathway) at the retinal site where the cone b-wave is generated (e.g., ON-depolarizing bipolar cells receiving rod and cone input ²⁸ ) cannot be ruled out. We propose another possible mechanism. 

Presumably, as a result of this gradual loss in photoreceptor cells, the OPL thins significantly (Fig. 4B) . No other retinal layers (of adult or juvenile retinas) were reduced as a result of exposure to the bright, luminous environment, suggesting that death of the photoreceptor cells alone explains the functional deficits (attenuation in ERG a- and b-waves and OPs) obtained after exposure. Conversely, our results also suggest that compared with those of the adult retina, photoreceptors of the juvenile rat retina are partially protected from the deleterious effects of bright light. This is in sharp contrast to previous results of ours obtained after the postnatal exposure to hyperoxia that showed the functional consequences (reduced b-wave and OPs with spearing of a-wave) were tied to the specific alterations of the deeper retinal layers while the photoreceptor layer remained unaffected. ^{20 21}  

Strikingly, rod and cone photoreceptors were not similarly affected by the bright, luminous environment; rods appeared to be more sensitive than cones to light damage. As shown in Figure 7Band Table 1 , cones required significantly more light (i.e., exposure of longer duration) than rods to yield equivalent b-wave reductions, supporting previous reports indicating that, at least in rodents, LIR is rhodopsin mediated, ^{6 9 29 30} thus explaining the response discrepancy of rods and cones after exposure to bright light. If indeed light-induced retinal abnormalities (in structure and function) are rhodopsin mediated, how can we explain our demonstration of cone b-wave anomalies? Previous studies suggest that the degeneration of cones might be induced by the massive release of neurotransmitters or potentially toxic ions by the dying rods ³¹ or that it might be the consequence of the loss of some trophic supports usually provided by the rods. ^{31 32} A rod-secreted surviving factor, the rod-derived cone viability factor (RdCVF), was recently identified in an in vitro model of rd mouse retina, supporting the latter concept. ^{33 34} That the light-induced anomaly is observed fairly early in the process (rather than cone function starting to degenerate only when rod function is completely abolished) also supports this concept. 

Comparative analysis of the recordings taken at P30 and at P60 revealed a small (but nonsignificant) enhancement (i.e., P30 amplitude measurements > P60 amplitude measurements) of the ERG anomalies evidenced at P30 (Table 1)for rats exposed before P28, suggesting that the retinopathy induced after light exposure (measurements taken at P30) could deteriorate with time (measurements at P60). Again, this is in sharp contrast to the oxygen-induced retinopathy in which recordings obtained at P60 were not different from those obtained at P30, with the only differences attributable to the normal maturation process. ^{20 21} As illustrated in Table 1and Figure 6 , during the normal maturation of the retina, there is a gradual decrease in amplitude of all the ERG parameters (rod a- and b-waves, rod V _max, cone b-wave) such that the ERG amplitudes measured at P60 are significantly smaller than those measured at P30. This is consistent with previous reports that compared ERG responses taken at P30 and P60 in normal rats. ^{20 21 35} However, the exact reverse (i.e., P60 measurements > P30 measurements) was observed for exposure intervals including P28 and later (P14–P28, P20–P28, P28–P34). It could be that the first ERG recording (i.e., between P30 and P37) was too close to the end of the bright-light exposure and, thus, the photoreceptors did not have sufficient time to partly recover from its possible reversible effect. ^{36 37 38} Clearly, this phenomenon must be further investigated. 

Our results also showed that the superior retina was more affected than the inferior retina after exposure to bright light (Fig. 5) , a finding consistent with previous reports obtained with the adult rat. ^{9 25 39 40 41} Nevertheless, the kinetics of the LIR are much slower in juvenile rats than in adult rats (Fig. 7) , a feature that further distinguishes adult from juvenile LIR. Similarly, these kinetics are much slower in the inferior retina than in the superior retina (Fig. 5I) . 

Finally, our results also demonstrated an interesting phenomenon of the juvenile rat retinas: progressive thinning of the ONL (resulting from the exposure) was accompanied by a gradual increase in thickness of the INL (Fig. 4D 4E) , a phenomenon not observed when the adult retina was subjected to the same stress or during the normal maturation process of the retina. This could suggest a compensation phenomenon by which the INL would increase its synaptic contacts to amplify the lower signal generated by the impaired outer retinal layers to generate close to normal retinofugal output, a phenomenon that could only be evidenced when using a developing retina model. Of interest, we did not observe a similar compensation phenomenon with our model of oxygen-induced retinopathy. ^{20 21} Supportive of this compensation hypothesis are previous studies reporting some neural remodeling after retinal degeneration, ^{3 42} such as migration of RPE cells to the retina, hypertrophy and migration of Müller cells, and hypertrophy of the soma of horizontal cells. A similar remodeling process has been described in an animal model of retinitis pigmentosa, namely the rd/rd mutant mouse. ^{43 44} The increase in thickness of the INL we observed in our model of light-induced degeneration could have followed one of the remodeling scenarios suggested here. 

In summary, our results show that in contrast to the retina of adults, the retina of juvenile rats is relatively preserved from the deleterious effect of exposure to bright light exposure until approximately 1 month of age. Further investigations are necessary to determine the mechanisms at the origin of this endogenous resistance. 

 Supported by Canadian Institutes of Health Research Grants MOP-13383 and MOP-14639, from FRSQ-Réseau Vision, and from the McGill University–Montreal Children’s Hospital Research Institute.

 Submitted for publication November 29, 2005; revised February 14 and March 13, 2006; accepted May 18, 2006.

 Disclosure: S. Joly, None; V. Pernet, None; A.L. Dorfman, None; S. Chemtob, None; P. Lachapelle, 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.

 

Table 1.

 

Group Data Reporting Amplitudes of Scotopic a- (Mixed Rod-Cone a-Wave) and b- (Mixed Rod-Cone b-Wave) Waves, Rod V _max, Photopic b-Wave (Cone b-Wave), Scotopic SOPs, and Photopic SOPs Measured in Control and Exposed Rats

Light Exposure Regimen	Age, d (no. rats)	Mixed Rod-Cone a-Wave	Mixed Rod-Cone b-Wave	Rod V max	Photopic Cone b-Wave	Scotopic SOP	Photopic SOP
Adults, healthy	30 (10)	243.7 ± 47.8	645.7 ± 143.3	484.1 ± 112.6	157.1 ± 35.7	266.7 ± 57.2	41.8 ± 12.8
Adults, 1 d	30 (9)	76.3 ± 44.2*	267.3 ± 146.8*	234.9 ± 123.1*	100.4 ± 52.0‡	153.6 ± 58.7	32.2 ± 13.3
Adults, 3 d	30 (3)	73.8 ± 30.0*	271.9 ± 60.6†	244.7 ± 56.1‡	89.8 ± 24.3§	129.3 ± 14.6	34.8 ± 12.1
Adults, 6 d	30 (4)	—	45.5 ± 21.8*	—	21.7 ± 12.2*	—	—
Adults, 14 d	30 (2)	—	38.1 ± 10.8*	—	16.6 ± 1.5*	—	—
Control	30 (30)	305.2 ± 65.3	752.9 ± 163.5	490.3 ± 102.3	165.2 ± 39.7	373.2 ± 82.9	47.9 ± 22.4
	60 (23)	238.2 ± 74.9	609.1 ± 173.5	420.4 ± 129.5	138.2 ± 42.8	267.1 ± 83.8	39.6 ± 15.9
P14–P17	30 (11)	293.9 ± 47.8	727.8 ± 125.8	450.7 ± 70.5	175.8 ± 36.5	351.3 ± 68.8	54.4 ± 16.5
	60 (11)	194.1 ± 80.2	528.1 ± 215.5	351.8 ± 136.7	125.7 ± 40.0	266.1 ± 117.2	41.0 ± 12.6
P14–P20	30 (4)	155.3 ± 23.7*	506.9 ± 116.5†	372.3 ± 48.6	146.7 ± 35.2	264.8 ± 59.5	48.8 ± 13.7
	60 (4)	124.9 ± 6.4†	459.2 ± 22.1	314.1 ± 40.1	124.4 ± 26.4	239.9 ± 27.5	41.8 ± 10.3
P14–P28	30 (17)	48.6 ± 20.2*	332.9 ± 106.0*	234.1 ± 76.4*	89.0 ± 29.6*	165.2 ± 70.2*	26.9 ± 9.0†
	60 (10)	73.7 ± 24.6*	402.8 ± 127.8†	286.4 ± 99.1‡	98.6 ± 37.6	198.5 ± 62.2	36.3 ± 11.9
P20–P28	30 (10)	53.1 ± 14.3*	233.6 ± 55.0*	232.4 ± 46.6*	68.2 ± 13.1*	175.9 ± 68.2*	30.4 ± 14.6
	60 (10)	133.5 ± 21.5*	502.9 ± 99.9	454.7 ± 84.2	152.4 ± 22.5	245.4 ± 37.5	51.9 ± 6.1
P28–P34	30 (10)	8.1 ± 10.7*	79.8 ± 27.1*	38.6 ± 62.7*	33.9 ± 11.1*	29.6 ± 7.5*	9.0 ± 5.0*
	60 (10)	46.3 ± 15.6*	251.6 ± 88.1*	255.3 ± 63.9†	96.9 ± 29.2‡	117.1 ± 28.6*	33.4 ± 6.7

 Amplitudes are given as mean (μV) ± 1 SD.

 Statistical analysis was performed with one-way ANOVA followed by post hoc Tukey test (* P < 0.001; † P < 0.01; ‡ P < 0.05) between controls and exposed rats at 30 and 60 days of age. Student t test was also used to compare control adults and adults exposed for 3 days (§ P < 0.05).

Figure 1.

 

Representative scotopic ERGs (A–K) and OPs (L) recorded in adults (A–E) and in juvenile rats (F–L). ERGs were obtained from control (unexposed) adults (A), in 30 day-old control (unexposed) juvenile rats (F), in adults exposed for 1 day (B), 3 days (C), 6 days (D), and 14 consecutive days (E), and in rats exposed at different time intervals during the first postnatal month (G–K) to a bright, luminous environment. Regimens of exposure are indicated at the top of each column (F–K). In adults, ERG recordings were performed 1 month after the beginning of the light exposure and between P30 and P37 in juvenile rats. Progressively brighter flashes (bottom to top) from −6.3 to 0.6 log cd · s/m² (as indicated at the left of the tracings in A) were used to establish the luminance-response function curves. Letters a and b on the control tracing (A, F, intensity 0.6) identify the ERG a- and b-waves, respectively. (Oblique arrows): Remnants of the b-wave recorded after exposures of 6 (D) and 14 (E) consecutive days in duration. (L) Representative scotopic OPs obtained from a normal (unexposed) rat (C, tracing) and from rats exposed to the bright, luminous environment during the different exposure regimens, as indicated at the left of each tracing. Flash intensity: 0.6 log cd · s/m². OPs are numerically identified, as shown in the control tracings. Horizontal calibrations and vertical calibrations: 40 msec and 500 μV (ERGs, A–K); 20 msec and 50 μV (OPs, L), respectively. A 20-msec prestimulus baseline is included in all tracings. (Vertical arrows): Flash onset.

Figure 2.

 

Representative photopic ERGs (A, B) and OPs (C) recorded from adult (A) and juvenile (B, C) rats. The duration of exposure (in days; A) and the different exposure regimens (exposure intervals in postnatal days; B) are indicated to the left of each tracing. In both groups, C (first tracing) identifies the response recorded from the control (unexposed) rat. In the adult rats, residual cone b-waves are identified with oblique arrows (tracings 6 and 14). (C) Representative photopic OPs obtained from a normal (unexposed) rat (tracing C) and juvenile rats exposed to the bright, luminous environment during the different exposure regimens, as indicated to the left of each tracing. Flash intensity: 0.6 log cd · s/m². OPs are numerically identified, as shown in the control tracings. Horizontal calibrations and vertical calibrations: 20 msec (ERGs [A, B] and OPs [C]) and 50 μV (ERGs [A, B]) or 5 μV (OPs [C]), respectively. A 20-msec prestimulus baseline is included in all tracings. (Vertical arrows): Flash onset.

Figure 3.

 

(A–E) Representative photomicrographs of retinas obtained from a normal (unexposed) adult rat (A) and from adult rats exposed for 1 day (B), 3 days (C), 6 days (D) or 14 consecutive days (E) to the bright, luminous environment. (F–J) Representative photomicrographs of retinas obtained from a normal (unexposed) juvenile rat (F) and from juvenile rats exposed during the different exposure regimens (as indicated on the top of each retina). (K–O) Representative photomicrographs of retinas obtained at P14 (K), P20 (L), and P28 (M) illustrating the normal development of the retina in juvenile rats. (N, O) Magnified views of the ONL at P14 (N) and at P28 (O). FL indicates nerve fiber layer. Calibration bars: (A–M) 50 μm; (N, O) 25 μm.

Figure 4.

 

(A, B) Measurements of the retinal layer thicknesses (ordinate, μm ± 1 SD) in adult (A) and in juvenile (B) rats. Given the difficulty in accurately dissociating the different retinal layers, the ONL data of the 28- to 34-day group includes what is left of the OS, IS, ONL, and OPL (B). (C) Measurements of the different retinal layers obtained at the different postnatal ages (see key in upper right corner). Statistical analysis was performed with one-way ANOVA followed by post hoc Tukey test comparing data obtained from all the exposure regimens with controls (*P < 0.05; **P < 0.01; ***P < 0.001) and Student t test (†P < 0.05) between nonexposed adults and adults exposed for 3 days (A). Retina refers to thickness of the entire retina. (D, E) Correlation between the thicknesses of the ONL and the INL (μm ± 1 SD) of retinas obtained from exposed adult (A) and juvenile (B) rats after exposure to bright light (dotted line and square data points) and during development (solid line and triangular data points). Numerals next to each data point indicate the duration of exposure (in days) for adult (A) and juvenile (B, dotted line) rats or the age of the juvenile rat at the time the retina was excised (B, solid line).

Figure 5.

 

Representative photomicrographs of the superior (A–D) and inferior (E–H) retinas of control (unexposed) rats (A, E) and from rats exposed from P14 to P20 (B, F), P14 to P28 (C, G), and for 6 consecutive days (D, H) to the bright, luminous environment. (I) Comparison of the linear regression curves obtained when correlating the superior (abscissa) and inferior (ordinate) ONL thicknesses (in number of nuclei units) of juvenile (black squares, dashed line) and adult (white squares, solid line) retinas. Each data point represents the cellular count obtained after light exposure of predetermined duration, as indicated numerically above each data point. Correlation coefficient (r) and slope are indicated next to each curve. Scale bar, 50 μm.

Figure 6.

 

Retinal maturation measured for control (unexposed) and exposed juvenile retinas. The index of retinal maturation index represents the average (percentage) of the ratios (P60/P30 × 100) for each ERG parameter (rod a-wave, rod b-wave, rod V _max, cone b-wave). Values are given as mean ± 1 SD. Statistical analysis was performed with one-way ANOVA followed by post hoc Tukey test (*P < 0.05; ***P < 0.001).

Figure 7.

 

Summary of findings illustrating the kinetics of the light-induced retinal degeneration in adult (A) and juvenile (B) rats. (A) In adult rats, the gradual loss in amplitude of the ERG parameter (average percentage of control of the mixed rod-cone a- and b-waves, photopic b-wave and rod V _max) and thickness of ONL (ordinate, percentage of control values) are plotted against the duration of exposure (indicated in days on the abscissa), whereas (B) in juvenile rats, it is plotted against the age at offset of exposure (in days on the abscissa). In juvenile rats, the duration of the exposure is indicated in days on top of each data point.

The authors thank Hakima Moukhles for use of the ultramicrotome and Louise Pelletier for technical assistance with electron microscopy.