Retinal cAMP Levels during the Progression of Retinal Degeneration in Rhodopsin P23H and S334ter Transgenic Rats

Abstract

purpose. To test whether high levels of cAMP promote apoptosis and shorten the life of retinal rod photoreceptors, the changes in cAMP levels during retinal degeneration were analyzed in two transgenic rat models that express rhodopsin P23H and S334ter mutations.

methods. Dark- and light-adapted heterozygous P23H (lines 1 and 3; P23H-1 and -3), S334ter line 4 (S334ter-4), and Sprague-Dawley (control) rats were studied at 4 to 8 weeks by cAMP enzyme competitive immunoassay and by cAMP immunocytochemistry.

results. In control animals retinal cAMP content reached a steady state level at 30 days of age. Dark-adapted control retinas had up to 97% higher cAMP content than light-adapted retinas, and photoreceptor cells were the major source of this increase. Dark-adapted photoreceptors in all three lines of transgenic rats at advanced stages of retinal degeneration had cAMP content different from that of the control. In rats that express mutant rhodopsin, the number of photoreceptor cells was progressively reduced, because of retinal degeneration, but dark-adapted cAMP levels did not decline accordingly. P23H transgenic animals of both lines had higher levels of cAMP per photoreceptor cell count than control animals. This elevation was more pronounced as degeneration progressed. S334ter animals showed smaller cAMP elevation than P23H rats at a similar stage of retinal degeneration, but at a point when S334ter rats were undergoing rapid retinal degeneration, whereas in P23H rats retinal degeneration was slowing down.

conclusions. All three lines of transgenic rats carrying rhodopsin mutations show an increase in dark-adapted photoreceptor cAMP levels. A complex relationship exists between cAMP levels and the rate of cell death in the retina. Although initially higher levels of cAMP may promote cell survival and slow down retinal degeneration, ultimately, elevated cAMP levels may become toxic and may contribute to retinal cell death.

Retinitis pigmentosa (RP) is the major form of inherited human progressive retinal disease, with many different underlying genetic origins. Studies have identified causative mutations in a number of photoreceptor-specific genes, such as rhodopsin, α and β subunits of cGMP phosphodiesterase, arrestin, peripherin/RDS, rod outer segment membrane protein (ROM1), and retinal pigment epithelium protein (RPE65) genes (for review, see Refs. ^{1 2} ). All these mutations lead to a common fate, the death of photoreceptor cells by apoptosis, as has been shown in several animal models. ^{3 4} Photoreceptors are more vulnerable to mutations in their proteins and more sensitive to environmental damage than other neurons of the retina and central nervous system (reviewed in Ref. ⁵ ). However, the cellular mechanisms underlying the eventual loss of photoreceptors remain largely unknown. Apoptosis is a complex mechanism that is cell-type and stimulus specific. Cellular interactions are a major factor regulating apoptosis, and there can be a number of different apoptosis pathways within a single cell type. The programs for apoptosis interact and may be modified by various other signaling pathways. ^{6 7}  

The second messenger cAMP (adenosine 3′,5′-cyclic monophosphate) has been implicated in a network of biochemical events leading to the induction or the prevention of apoptosis in a variety of cell types. However, few studies have addressed the role of cAMP in retinal degeneration. Elevated retinal levels of cAMP have been detected in mice expressing the rhodopsin mutant P347S, which is found in patients with autosomal dominant RP (adRP). ⁸ Elevated levels of retinal cAMP have also been reported in rd/rd mice, which have a photoreceptor dystrophy caused by a defect in cGMP phosphodiesterase. ⁹ In rds mice, with retinal degeneration caused by mutation in the gene encoding peripherin, modifications in adenylyl cyclase and phosphodiesterase activities have been observed. ¹⁰ These studies suggest a relationship between intracellular levels of cAMP and RP, whereby high levels of cAMP may promote apoptosis and lead to rod photoreceptor death. Attempts to slow photoreceptor degeneration by treatment with neuromodulator-related factors have shown that those that lower the intracellular levels of cAMP, such as dopamine, D2 receptor agonists, and α₂-adrenergic receptor agonists, appear to convey neuroprotection. ^{11 12} Conversely, neurohormones that increase cAMP level, such as melatonin, increase the susceptibility of photoreceptor cells to light-induced degeneration ^{12 13} whereas melatonin receptor antagonists decrease susceptibility to light damage. ¹⁴ Close relationships exist between dopamine, melatonin, and cAMP in the retina. ^{15 16} Light stimulates dopamine synthesis and secretion from amacrine and interplexiform cells in the inner retina. ¹⁷ Dopamine activates D2-like/D4-subtype receptors in the photoreceptor layer, leading to the inhibition of cAMP production, which results in the inhibition of melatonin synthesis. ^{18 19 20} Melatonin is synthesized and released by photoreceptors in the dark, it inhibits dopamine release and, consequently, increases intracellular levels of cAMP. ^{21 22 23 24} In the retina, dopamine and melatonin are important actors in the modulation of light-dependent phenomena, such as ROS membrane renewal, disc shedding, and retinomotor movement in a circadian and mutually antagonistic fashion. ^{25 26 27} Thus, the maintenance of regulated levels of cAMP in retinal photoreceptors seems to be of major importance in the biology of photoreceptor health. 

In this study we analyzed the changes in cAMP level during retinal degeneration in two transgenic rat models that express rhodopsin P23H and S334ter mutations. More than 80 different rhodopsin mutations have been identified as causing the largest single-gene form of adRP, accounting for 25% of all patients with this disease. ²⁸ The rhodopsin N-terminal point mutation P23H accounts for approximately 12% of families with adRP born in the United States. ^{29 30} Animal models mimicking this mutation have been generated in mice (carrying the murine or human transgene) ^{31 32} and rats (carrying the mouse transgene). ^{33 34} Varying degrees of abnormal localization in outer nuclear layer (ONL) and outer plexiform layer (OPL) of P23H mutant human rhodopsin have been observed in transgenic mice, depending on the relative ratio of transgenic rhodopsin to that of the endogenous murine rhodopsin. ^{31 35} However, no rhodopsin missorting has been detected in P23H transgenic rats ³³ and VPP mice that express the transgene encoding for P23H, V20G, and P27L rhodopsin mutations. ³⁶ Mutations in the C-terminal tail of rhodopsin are known to cause particularly severe forms of human adRP. ^{37 38 39} The C-terminal domain of rhodopsin and, more precisely, the last five amino acids play a major role in sorting of rhodopsin into post-Golgi membranes that deliver it to its normal subcellular location in the rod outer segments (ROS) of the photoreceptor cells. ^{40 41} The P347S transgenic animal model and the Q344ter and S334ter models with truncation of the last five and fifteen C-terminal amino acids, respectively, all show abnormal rhodopsin localizations. ^{33 42 43 44} Failure to properly sort rhodopsin has been postulated as a cause of photoreceptor cell death by the disruption of the regulatory intracellular machinery involved in the maintenance of cell polarity. ⁴¹ Activation of mislocalized rhodopsin has been proposed to contribute to cAMP elevation and thereby to promote cell death in P347S transgenic mice. ⁸  

We assayed the in vivo levels of the intracellular messenger cAMP, in retinas isolated from light-adapted (LA) or dark-adapted (DA) control Sprague-Dawley (Harlan Sprague-Dawley, Indianapolis, IN) and rhodopsin P23H and S334ter transgenic rats. We report that P23H transgenic animals had higher levels of cAMP per photoreceptor cell count than did control animals and that this elevation was more pronounced as degeneration progressed. S334ter animals showed smaller cAMP elevation than P23H rats at a similar stage of retinal degeneration. However, at that point S334ter animals were undergoing rapid retinal degeneration, whereas in P23H rats, degeneration was slowing down. Our data suggest a complex relationship between cAMP levels and the rate of cell death in the retina. 

Materials and Methods

Animals

The studies were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Homozygous transgenic rat breeders were kindly provided by Mathew M. LaVail (University of California San Francisco, San Francisco, CA), and heterozygous animals were produced by mating homozygous breeders with wild-type Sprague-Dawley rats. Experiments were performed using heterozygous P23H-1 and -3 rats which express the Pro23His murine transgene for the rhodopsin molecule, and S334ter-4 rats which carry the murine rhodopsin mutant Ser334ter. All heterozygous transgenic animals and Sprague-Dawley rats were born in our vivarium and were maintained on a 12-hour light–dark cycle at 5-lux intensity at cage level. 

Sample Preparation and cAMP Measurement

Animals were dark or light adapted for 4 to 5 hours and killed by an overdose of pentobarbital sodium (Beuthanasia-D special; Schering-Plough Animal Health, Omaha, NE). The retinas were removed through a central incision in the cornea, after removing the lens and using forceps to gently extrude the retina. Retinas of DA animals were removed under dim red light. DA and LA retinas were snap frozen in liquid nitrogen, thawed, and immediately homogenized in cold 10% trichloroacetic acid (TCA). After incubation on ice for 30 minutes and centrifugation at 4°C for 10 minutes at 10,000g, protein content of the resultant pellet was determined by a bicinchoninic acid assay (Pierce, Rockford, IL). Supernatants were ether washed (to remove TCA), dried under centrifugal vacuum, and analyzed for cAMP content with an enzyme competitive immunoassay (Assay Designs, Inc., Ann Arbor, MI). cAMP concentrations were normalized to the retinal protein content. Transgenic and control animals were compared by using repeated-measures analysis of variance (Proc Mixed in SAS; SAS Cary, NC) to model the effect of light and dark and rat line, accounting for the correlation between animals of the same age. Two-way ANOVA was used to compare the effect of strain and age. One-way ANOVA with Bonferroni post hoc analysis was used to compare transgenic with control animals at each age (Prism software; GraphPad, San Diego, CA). 

Morphometric Analysis

S334ter eyes of 4-, 8- or 15-week-old rats were fixed overnight at 4°C in 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M cacodylate buffer, postfixed in 1% osmium for 1 hour, dehydrated, and embedded in Epon-Araldite resin. One-micrometer-thick sections along the vertical median through the optic nerve were stained with toluidine blue, and measurement of the thickness and cell count were performed as previously described. ⁴⁵ ONL width was specified by the number of photoreceptor nuclei by column and averaged for each retinal section. 

cAMP Immunocytochemistry

cAMP immunocytochemistry was performed on cryostat sections of acrolein-fixed retinas by the procedure developed by Wiemelt et al. ⁴⁶ Anti-cAMP polyclonal antiserum highly specific for cAMP ⁴⁶ was a kind gift of Arthur McMorris (Wistar Institute, Philadelphia, PA). Isolated retinas from DA or LA animals (control and P23H-3, 39-day-old transgenic rats) were incubated for 15 minutes at 37°C in oxygenated Earle’s balanced salt solution containing 1 mM phosphodiesterase inhibitor isobutylmethylxanthine (IBMX) and 10 μM forskolin. Retinas were fixed with 5.5% acrolein in 0.1 M sodium acetate-buffered solution (pH 4.75) for 1 hour at room temperature. The fixative medium was then replaced by 1 mg/mL glycine for 30 minutes and gradually cryoprotected by sucrose. Cryostat sections 10 μm thick were treated for 30 minutes in a quenching solution of 1% glycine, followed by 1% sodium borohydride for 30 minutes before incubation in 50% goat serum and 0.4% Triton X-100 in 50 mM Tris-HCl (pH 7.5). Sections were incubated overnight at 4°C with the cAMP antiserum (1:50) in 5% goat serum and 0.4% Triton X-100 in the same buffer. Primary antiserum was visualized using a Cy3-conjugated goat anti-rabbit IgG (1:400; Jackson ImmunoResearch Laboratory, Inc., West Grove, PA). 

Results

Photoreceptor Degeneration in P23H and S334ter Transgenic Rats

In this study, we compared retinal cAMP levels in three different lines of transgenic rats undergoing retinal degeneration due to mutations in the rhodopsin gene, P23H-1, P23H-3, and S334ter-4. Correlation of retinal degeneration with retinal histopathology in transgenic rats carrying the rhodopsin P23H mutation has been reported previously. ⁴⁵ P23H-1 exhibit faster degeneration than P23H-3 showing approximately 40% cell loss in the ONL and 40% shortened ROS at 4 weeks of age, compared with a minimal cell loss and only slightly shortened ROS in P23H-3. ⁴⁵ To obtain histology studies comparable to those performed on P23H lines, we conducted a similar analysis on S334ter-4 rats at the same ages at which P23H rats had been analyzed (Fig. 1) . Within the first 15 weeks, retinas of animals expressing S334-ter mutation degenerated at a faster rate than those of transgenic P23H-3 rats. At 4 weeks, S334ter-4 rats had approximately eight rows of photoreceptor nuclei (25% reduction) and slightly shortened ROS, similar to P23H-3, whereas at 15 weeks, only three rows of nuclei remained, similar to P23H-1 which initially degenerates faster than P23H-3. Green et al. ³³ observed a similar rate of retinal degeneration in S334ter-4 rats. These authors estimated the levels of transgene expression in this line at approximately 10% of wild-type rhodopsin. 

Changes in cAMP Levels in Transgenic Rats Undergoing Retinal Degeneration

We next determined retinal cAMP content of LA and DA P23H-1, P23H-3, and S334ter-4 rats within the interval of rapid retinal degeneration at approximately 4 to 8 weeks and compared it with the control rats. The retinas of DA animals contained significantly more cAMP than those of LA animals, as shown in Figure 2 . DA control retinas had approximately 75% higher cAMP content than LA retinas. In the control group, retinal cAMP content of LA and DA animals increased between 20 and 30 days of age, when the rat retina is still developing ⁴⁷ but reached a steady state level at 30 days. In S334ter-4 rats cAMP content remained the same from 30 to 62 days of age, although retinal degeneration claimed approximately 25% of photoreceptor cells during the same interval. P23H-3 rats showed a steady increase in DA cAMP over that of the control rats (P < 0.002), although at 62 days, only approximately 75% of photoreceptors remained. Moreover, this increase was more than 25% higher at 62 days than at 30 days (P < 0.01). In P23H-1 a decrease in retinal cAMP content appeared to parallel the rate of photoreceptor loss; however, at 30 days these rats had the same retinal cAMP levels as control rats with only approximately 60% of photoreceptors remaining. Further, at 62 days, this line showed the smallest increase in cAMP in the dark (∼28% over LA animals), probably because of the substantial photoreceptor loss (<40% remaining). At every other age in P23H animals of both lines, the dark–light difference in cAMP was similar to that in the control animals (∼75% dark increase). S334ter-4 rats had smaller dark–light differences than control animals (∼67% dark increase) and significantly lower (∼30%, P < 0.001) cAMP levels in DA retinas during the period of rapid photoreceptor loss. S334ter also differed from P23H rats in cAMP content, and this was most apparent when comparing retinas with roughly equivalent photoreceptor loss. At a point of 50% photoreceptor loss, the light-sensitive cAMP pool in P23H-1 rats was twice that of S334ter-4 animals (increase of 6.45 ± 0.2 vs. 3.04 ± 0.3 pmol cAMP/mg protein, respectively). This indicates that cAMP is regulated differently in these two degeneration models. This may be due, in part, to the effects of the truncated rhodopsin encoded by the S334ter transgene which lacks phosphorylation sites from the C-terminal and this has been shown to cause a decreased rate of inactivation after light activation. ⁴⁸  

cAMP Immunolocalization

To specifically determine the source of cAMP increase in the DA retinas, we performed immunocytochemistry on LA and DA retinas of control and P23H-3 rats undergoing retinal degeneration. We used a novel immunocytochemical approach and an antibody developed by Wiemelt et al. ⁴⁶ to visualize intracellular cAMP. We treated retinas with 10 μM forskolin and 1 mM IBMX to attain a detectable level of cAMP for immunocytochemistry. In the LA retinas, a small amount of cAMP was detected in photoreceptor inner segments, OPL, and ganglion cell layer, and in a subpopulation of cells in the inner nuclear layer (Fig. 3) . In the DA retinas a significant increase in cAMP immunoreactivity was detectable in the inner segments of photoreceptor cells and the ONL containing photoreceptor nuclei, in both control and P23H-3 rats, suggesting that photoreceptor cells are nearly completely responsible for the light-dependent changes in retinal cAMP. Our data are consistent with observations of Cohen ⁴⁹ who found that the light-sensitive pool of cAMP in the retina of DA animals is largely, if not entirely, in the photoreceptor layer. Because cAMP levels in the LA animals changed little during retinal degeneration (Fig. 2) , this suggests that cAMP measurements in DA retinas closely parallel cAMP levels in degenerating photoreceptors. 

Correlation of Retinal Degeneration with Retinal cAMP Levels

Taken together, data presented in Figures 2 and 3 suggest that to evaluate cAMP levels in degenerating photoreceptors properly we must calculate retinal cAMP content to account for photoreceptor cell loss. We have estimated the relative photoreceptor cAMP content of DA control rat retinas (picomoles cAMP per milligram protein per ONL cell count) and compared it with that found in degenerating retinas. P23H-3 and S334ter-4 had significant increases in photoreceptor cAMP with age, between 30 and 62 days (P < 0.0001), whereas P23H-1 had a significantly higher level than the control, as early as 30 days (Fig. 4 ;P < 0.0001). It is apparent from this analysis that all three lines of transgenic rats showed an increase in DA photoreceptor cAMP levels and that in P23H rats, this increase paralleled the degree of retinal degeneration when data from both lines were combined. At 30 days of age, P23H-3 rats had photoreceptor cAMP levels slightly higher than control retinas. At 62 days, cAMP was significantly higher than control (∼60%, P < 0.02) and was comparable to the levels in 30-day-old P23H-1 rat retinas (∼80% increase, P < 0.0001), which were at somewhat more advanced stage of degeneration at that time (68% remaining photoreceptors in P23H-3 vs. 58% in P23H-1). Finally, at 62 days P23H-1 rats had the highest DA photoreceptor cAMP content and the most advanced stage of degeneration. S334ter-4 rats at 62 days had elevated photoreceptor cAMP, but this increase was smaller than in P23H rats at a similar stage of retinal degeneration. Although our study lends support to the conclusion that mislocalization of rhodopsin C-terminal mutants may contribute to the elevated retinal levels of cAMP, we found that P23H lines showed even higher increases in DA cAMP, although they did not show rhodopsin mislocalization at the stages of retinal degeneration analyzed in this study. ³³  

We applied similar analyses to estimate the dark–light differences in cAMP content of degenerating photoreceptors. The most striking increase was observed in P23H-3, which had significantly larger dark–light differences than control animals at 62 days (P < 0.005), at the time when the rate of photoreceptor cell loss appears to be slowing down. ⁴⁵  

Discussion

In this study we showed the existence of different types of alterations in photoreceptor cAMP content in rhodopsin P23H and S334ter transgenic rats, although both retinal degeneration models showed an increase in DA cAMP. Transgenic rats carrying P23H rhodopsin mutations showed a clear correlation between the increase in DA cAMP levels and the degree of retinal degeneration. S334ter animals undergoing rapid retinal degeneration showed lesser cAMP elevation than P23H rats at similar stages of retinal degeneration. Our conclusions are based on the analysis that assumes that the major contribution in cAMP of DA retinas comes from photoreceptor cells, which has been confirmed by our immunocytochemical data. 

The observations that cAMP production is much more increased in P23H than in S334ter rhodopsin transgenic animals may indicate that these two mutations follow a different apoptotic program. In S334ter rats, missorting of the mutant rhodopsin has been observed, ³³ and apoptosis has been described to occur through the caspase cascade pathway. ⁵⁰ However, rhodopsin trafficking in P23H transgenic rats seemed not to be altered at the stages of retinal degeneration analyzed in this study. ³³ In both cases, cAMP may interact with the apoptotic program at multiple points. It is possible that elevated cAMP induces changes in the synthesis of proapoptotic or antiapoptotic factors. Ulshafer et al. ⁵¹ demonstrated in human retinal cultures that inhibition of retinal protein synthesis affects rod photoreceptors to a greater extent than cones or other retinal cells and causes rod cell death. Apoptotic cascades may also differentially interfere with cAMP production and/or hydrolysis in these two models, in that recent data indicate that cAMP-specific phosphodiesterase PDE4A5 is a caspase-3 substrate. ⁵² Therefore, it is possible that cAMP changes that we observed may also occur as a consequence of apoptosis. 

Abnormally high levels of cAMP in the dark in the two retinal-degeneration models that we studied could result from elevated levels of extracellular melatonin or from abnormal intracellular pathways of production or degradation. Melatonin synthesized by photoreceptors inhibits dopamine synthesis and release ²¹ and consequently increases intracellular levels of cAMP. However, some melatonin receptors are negatively coupled to adenylyl cyclase and could modulate cAMP levels independent of the effects on dopamine. ⁵³ There is evidence that photoreceptor cells express different subtypes of melatonin receptors. ⁵⁴ Intracellular levels of cAMP are also modulated through α₂-adrenergic receptors present in the inner segment ⁵⁵ and calmodulin-sensitive adenylyl cyclase. ⁵⁶ Abnormal Ca²⁺ levels could be one mechanism by which cAMP production is increased in the dark. Recent studies have shown, however, that blocking voltage-gated Ca²⁺ channels located on the rod inner segment plasma membrane does not affect the course of degeneration in P23H-1 rats. ⁵⁷ There are other sources of Ca²⁺, such as intracellular stores or the interphotoreceptor matrix through cGMP-gated channels, that could affect intracellular Ca²⁺ levels and thereby alter cAMP levels. 

All transgenic rats in our study showed downregulation of cAMP in the light. Because this light-sensitive pool defines the photoreceptor content of cAMP, we conclude that the intracellular signaling pathways involved in cAMP regulation in response to light are at least partially functional in these transgenic photoreceptors, unlike those of rds/rds mice. ⁵⁸ S334ter retinas have lower total cAMP content than control retinas; therefore, they appear to behave as though exposed to prolonged light. One reason could be the absence of phosphorylation sites in the mutant transgene of S334ter transgenic rats. An additional point at which S334ter may interfere with light response is by inhibiting synaptic transmission, because C-terminally truncated rhodopsin also localizes to the synapse. ³³  

It has been shown that the survival of central nervous system (CNS) neurons, including retinal ganglion cells, depends on physiological levels of electrical activity, as well as cAMP elevation, and these are thought to exert their protective action by recruiting receptors for neurotrophic factors to the plasma membrane. ^{59 60 61} Therefore, lower light regulatable cAMP pool in S334ter retinas may cause retinal cells to become unresponsive to survival factors and thereby promote the death of photoreceptor cells. 

The relationship of the light-sensitive cAMP pool to photoreceptor degeneration within the P23H lines was not straightforward. The light-sensitive cAMP pool in P23H-1 gradually declined with age, whereas in P23H-3, with slower degeneration, it continued to increase, as photoreceptors were progressively lost from the retina. The two lines of P23H differed in the level of cAMP at the comparable stage of degeneration. The 62-day-old P23H-3 animals had a light-sensitive pool of cAMP significantly larger than 30 day-old P23H-1 animals, even though they differed in cell number by only 12%. It is possible that the normal responsiveness of rods to light influenced this cAMP measurement, and indeed the ERG a-wave amplitudes in 4-week P23H-1 animals were approximately one half that in 8-week-old P23H-3 animals. ⁴⁵ The largest dark–light difference in cAMP was observed in P23H-3 animals at 8 weeks when the rate of photoreceptor cell loss seemed to be slowing. ⁴⁵ This suggests that cAMP levels may be more closely related to the rate of photoreceptor loss than to the photoreceptor cell numbers at any given time. Thus, the link between elevated cAMP and photoreceptor degeneration seems to be more complex than has been described for cGMP and photoreceptor death in rd mouse. ⁶²  

Our hypothesis is that high local levels of cAMP in P23H rhodopsin transgenic rats, as well as in other animal models shown to have elevated retinal cAMP (P347S rhodopsin, rd, and rds mice) are toxic for the photoreceptor cells. Initially, elevated cAMP could promote a response to survival factors; therefore, P23H3 that have the largest dark–light difference show the slowest retinal degeneration at that time, whereas S334ter degenerate much faster during the period in which they have the least dark–light difference. However, elevated cAMP in P23H would also be expected to lead to increased melatonin production, and melatonin has been associated with increased susceptibility to retinal degeneration. Therefore, we hypothesize that an increased level of cAMP caused by retinal injury may initially benefit degenerating cells, but that cAMP regulation becomes altered by the increased synthesis of melatonin stimulated by the high intracellular concentration of cAMP. Maintaining low intracellular cAMP under such conditions may interfere with a negative feedback loop and preserve the integrity of the photoreceptor cells. 

 Supported by National Eye Institute Grants EY-12421 and EY-07003, Research to Prevent Blindness, and Foundation Fighting Blindness.

 Submitted for publication August 29, 2001; revised December 26, 2001; accepted January 8, 2002.

 Commercial relationships policy: N.

 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 light micrographs of Epon-embedded S334ter rhodopsin transgenic rat retinas, at the ages of 4, 8, and 15 weeks, showing progressive thinning and loss of photoreceptor cells with age compared with 15-week-old control Sprague-Dawley rat retina. Tissue was sectioned along the vertical meridian through the optic nerve and stained with toluidine blue. Averages of ONL cells, ROS length and outer plus inner segment length in S334ter-4 rats at ages 4, 8, and 15 weeks are also shown. Data are the mean ± SD; n, number of retinas measured.

Figure 2.

 

Comparison of cAMP concentration in DA and LA retinas from S334ter-4 (left) and P23H (right) transgenic rats and Sprague-Dawley control animals. Data are the mean ± SEM of 6 to 35 independent measurements. At 20 days, the immature retina of control animals displayed lower cAMP levels (11.72 ± 0.72 and 6.8 ± 0.36 pmol/mg protein in dark and light conditions, respectively) compared with older animals. No significant changes in total cAMP with age were measured (P > 0.05) between age groups 30-, 39-, and 60-days old in either light condition in control animals. The average total cAMP in control adult rat retinas was 16.90 ± 0.5 pmol/mg in DA and 9.73 ± 0.3 pmol/mg in LA animals. In DA S334ter-4 transgenic animals, a significant difference from control (a 30% decrease, P < 0.001) was observed for age groups 30, 39, and 60 days. Percentage of remaining photoreceptors was estimated from Figure 1 and from Machida et al. ⁴⁵

Figure 3.

 

cAMP immunoreactivity in Sprague-Dawley control (left) and P23H (right) transgenic rats, in DA or LA animals. Isolated retinas were preincubated with forskolin and IBMX, processed for cryosectioning, and labeled with anti-cAMP antiserum. Bright immunolabeling in DA animals (control and P23H) was obtained in the photoreceptor inner segment and in the ONL. In LA animals, staining in the outer retina was nearly undetectable. In the inner retina, a subset of amacrine cells and retinal ganglion cells were strongly labeled in both DA and LA retinas. ROS, rod outer segments; RIS, rod inner segments; ONL, outer nuclear layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform layer; GCL, ganglion cell layer.

Figure 4.

 

Correlation of retinal degeneration with retinal cAMP levels. The relative photoreceptor cAMP content of 30- and 62-day-old DA P23H-1, P23H-3, and S334ter-4 transgenic rats and Sprague-Dawley control animals was calculated by dividing cAMP content (in pmol/mg protein) with ONL cell counts (obtained at 28 and 56 days from Machida et al. ⁴⁵ and Fig. 1 ). The following equation: ς _x ²/x ² = (ς _u ²/u ²) + (ς _v ²/v ²), where x is picomoles cAMP/milligram protein/ONL cell count, u is picomoles cAMP/milligram protein, v is ONL cell count, and ς is the SD of each, was used to estimate the SD of the normalized data. Two-way ANOVA was used to compare the effect of strain and age. One-way ANOVA with Bonferroni post hoc analysis was used to compare transgenic with control animals at each age.