| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1From the James A. Baker Institute for Animal Health, College of Veterinary Medicine, Cornell University, Ithaca, New York; and the 2School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
METHODS. The retinas of 25 XLPRA2-affected dogs (age range, 2–40.6 weeks) and age-matched control subjects were collected, fixed, and embedded in epoxy resin for morphologic evaluation or in optimal cutting temperature (OCT) medium for TUNEL assay and immunohistochemistry. Cell-specific antibodies were used to examine changes in rods and cones and to evaluate the effects of the primary photoreceptor degeneration on inner retinal cells.
RESULTS. Abnormal development of photoreceptors was recognizable as early as 3.9 weeks of age. Outer segment (OS) misalignment was followed by their disorganization and fragmentation. Reduction in length and broadening of rod and cone inner segments (IS) was next observed, followed by the focal loss of rod and cone IS at later time points. The proportion of dying photoreceptors peaked at approximately 6 to 7 weeks of age and was significantly reduced after 12 weeks. In addition to rod and cone opsin mislocalization, there was early rod neurite sprouting, retraction of rod bipolar cell dendrites, and increased Müller cell reactivity. Later in the course of the disease, changes were also noted in horizontal cells and amacrine cells.
CONCLUSIONS. XLPRA2 is an early-onset model of XLRP that is morphologically characterized by abnormal photoreceptor maturation followed by progressive rod–cone degeneration and early inner retina remodeling. The results suggest that therapeutic strategies for this retinal degeneration should target not solely photoreceptor cells but also inner retinal neurons.
To understand the retinal function of the RPGR protein, as well as the pathogenic mechanisms that link mutations in RPGR with the death of photoreceptor cells, several animal models have been used. These comprise two transgenic murine models (RPGR knockout mouse,8 and a dominant gain-of-function mutant9 ), and two naturally occurring canine mutations in exon ORF1510 that cause two forms of X-linked progressive retinal atrophy (XLPRA).
In XLPRA1, a five-nucleotide deletion (del1028-1032) in exon ORF15 causes an immediate premature stop codon that results in a protein truncated of its 230 C-terminal amino acids. This mutation causes a loss of function of RPGR. Morphologic characterization showed that photoreceptor cells develop and function normally, but then undergo progressive rod–cone degeneration. The earliest histologic signs of rod degeneration are detected at 11 months of age, and are followed at later stages by cone death.10 11
In XLPRA2, preliminary results from our group have shown that the disease is a much more severe and earlier form of retinal degeneration than XLPRA1.10 A two-nucleotide deletion (del 1084-1085) in exon ORF15 results in a frameshift that changes the deduced peptide sequence by the inclusion of 34 additional basic residues and increases the isoelectric point of the truncated protein.10 In addition, the mutant ORF15 protein was also shown to accumulate in the endoplasmic reticulum of transfected COS7 cells. These results suggest that the mutation in XLPRA2 causes a toxic gain of function and is comparable to the severe human phenotype resulting from microdeletions that cause a frameshift.12
Although the precise subcellular location of the RPGR ORF15 protein in photoreceptor cells is still debated,13 14 findings in the null mutant mouse suggest that it may play a role in maintaining a polarized distribution of proteins between inner (IS) and outer (OS) segments.8 Yet, it is still unclear how the loss of function of RPGR or the expression of a toxic mutant RPGR protein in XLPRA2, may initiate a cascade of molecular events that ultimately lead to photoreceptor cell death. To begin to address this question, we examined the retinal structural alterations that occur in XLPRA2 and characterized the time course of photoreceptor disease, degeneration, and death and the subsequent alterations that occur in the inner retina. We found an early onset of photoreceptor disease leading to cell death, as well as early inner retina remodeling. Our results identify the critical stages in the pathogenesis of the disease and define the time windows for testing novel therapies.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
Sections from both the superior and inferior meridians were examined with a 40x objective on a light microscope (Axioplan; Carl Zeiss Meditec GmbH Oberkochen, Germany). The sections were examined in contiguous fields from the optic disc to the ora serrata. This included evaluation of the retinal pigment epithelium (RPE), the rod and cone OS and IS, and the thickness and density of the outer (ONL) and inner (INL) nuclear layers. For each dog, a single section from both quadrants was used for quantitative evaluation of the photoreceptor cells, and INL cells at three specific locations: S1, 2000 ± 500 µm from the optic nerve; S3, 2000 ± 500 µm from the ora serrata; and S2, midway (±500 µm) between these two points. At each of these sites, the number of rows of nuclei in the ONL and INL were counted in at least three areas of a 40x field and averaged. For the same areas, the thickness (in micrometers) of the ONL and INL were measured on digitally captured images (Spot 4.0 camera, Diagnostic Instruments, Inc., Sterling Heights, MI).
The kinetics of photoreceptor cell loss were analyzed by fitting the ONL thickness data to solutions of the following differential equations reported by Clarke et al.17 We are also reporting below the integral equations used for the statistical analysis, since we found typographical errors confirmed by the authors (Geoff Clarke, personal communication, November 29, 2005) in the integral equations provided in the supplementary information that accompanied their paper (http://www.nature.com/nature/journal/v406/n6792/suppinfo/406195a0.html).
 | (1) |
 | (2) |
 | (3) |
Phagocytic cells present in the photoreceptor layer were counted throughout the entire length of both the superior and inferior retinal meridians and expressed as the number of phagocytic cells per unit length of retina. The unit length was set as 10,000 µm. Pyknotic photoreceptor nuclei were counted in the ONL in both the superior and inferior meridians and expressed as the number of pyknotic nuclei per unit area of ONL. The unit area was set as 1 million µm2 (1 M µm2) of ONL. A similar count was used to quantify TUNEL-positive photoreceptor cells.
TUNEL Assay
In 18 XLPRA2-affected dogs, one eye was processed immediately after enucleation for TUNEL assays and/or immunohistochemistry (Table 1 : TUNEL, IHC). After enucleation, a slit was made through the globe at the level of the ora serrata, and the entire globe was fixed for 3 hours in 4% paraformaldehyde in 0.1 M phosphate-buffered saline at 4°C. The posterior segment then was isolated, the vitreous gently removed, and the eye cup fixed for an additional 24 hours at 4°C in 2% paraformaldehyde in 0.1 M phosphate-buffered saline. The tissue was then trimmed, cryoprotected sequentially for 24 hours in a solution of 15% and 30% sucrose in 0.1 M sodium phosphate and 0.15 M sodium chloride (pH 7.2; BupH, phosphate-buffered saline; Pierce, Rockford, IL; referred in the text as PBS) at 4°C, and embedded in OCT medium.
Cryosections (7 µm thick) along the superior meridian of 17 XLPRA2 dogs (age range, 3.9–40.6 weeks) were used for TUNEL assay, according to the manufacturer’s protocol (In situ cell death detection kit; Roche) and stained with 4',6'-diamino-2-phenylindole (DAPI). Sections along the superior meridian of three normal beagles (ages, 4, 5, and 6 weeks) were also used. Positive control specimens included sections pretreated with DNase I (3 U/mL in 50 mM Tris-HCl [pH 7.5] and 1 mg/mL BSA for 10 minutes at room temperature). For negative control subjects, the terminal transferase enzyme was omitted from the TUNEL reaction mixture. Sections were examined from the optic disc to the ora serrata by epifluorescence microscopy with the 40x objective. TUNEL-labeled cells in the ONL were counted throughout the entire length of the section (i.e., from disc to ora serrata). In determining the proportion of photoreceptor cells that undergo cell death as a function of time, we express our results as the number of TUNEL-labeled photoreceptor cells per 1 M µm2 of ONL. The area of the ONL of each section was obtained by measuring the entire length of the ONL from optic disc to ora serrata, and multiplying it by the average thickness of the ONL throughout the section (mean value of the thickness measured in the three locations S1, S2, and S3). This method may slightly underestimate, in areas of decreased photoreceptor density, the proportion of cells that are TUNEL positive. Yet, it was selected because individual cell count could not be determined on 7-µm-thick DAPI-stained cryosections. For each dog, this procedure was performed in triplicate with sequential sections from the superior meridian. The values were averaged and reported as the mean ± SD.
Immunohistochemistry
Sections along the superior retinal meridian of nine XLPRA2 dogs (age range, 2–40.6 weeks) and six normal dogs (age range, 2–24 weeks) that were processed as described earlier were used for fluorescent immunohistochemistry. We used a battery of cell-specific primary antibodies,9 18 19 20 21 22 23 24 25 26 27 28 29 30 of which more than half worked on canine retina (see details in Table 2 ). Because of the lack of specific antibodies directed against all subpopulations of ganglion cells in the dog, our study did not include assessment of their density or morphology. Because previous testing of RPGR and RPGRIP antibodies conducted in our laboratory failed to show any cross-reactivity or specificity on canine retina, we did not include them in this study.10 Cryosections (7–10-µm thick) were incubated overnight with the primary antibodies after a blocking step with 10% normal serum from the appropriate species. The antigen–antibody complexes were visualized with fluorochrome-labeled secondary antibodies (Alexa Fluor, 1:200; Invitrogen, Carlsbad, CA). DAPI stain was used to detect cell nuclei. Slides were mounted with a medium composed of polyvinyl alcohol and DABCO (1,4 diazobizyklo-[2.2.2]oktan) (Gelvatol; Sigma-Aldrich, St. Louis, MO), and examined with an epifluorescence microscope (Axioplan; Carl Zeiss Meditec). Images were digitally captured (Spot 4.0 camera; Diagnostic Instruments, Inc.) and imported into a graphics program (Photoshop; Adobe, Mountain View, CA) for display.
|
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Normal retinal development in the dog is complete at approximately 7 to 8 weeks of age.15 16 At birth, photoreceptor cells have not completely differentiated and are located in the sclerad portion of the outer neuroblastic layer, which then give rise to the ONL. Maturation of photoreceptors occurs in waves from the central to peripheral retina.31 IS are first seen as short bulges of cytoplasm protruding from the external limiting membrane between 1 day and 1 week after birth.32 At approximately 2 weeks of age, IS are visible throughout the entire length of the retina, and OS formation is underway centrally (Fig. 1A) . By 5.4 weeks of age, OS are formed and begin to elongate (Fig. 1B) . Full maturation of photoreceptors is reached at approximately 8 weeks of age, at which time the retina resembles that of the adult (Fig. 1C) . Pyknotic figures in the ONL were extremely rare (0 to 1 per 1 M µm2 of ONL) at all ages examined. Concurrent with the maturation of the photoreceptors, there are changes in the inner retinal layers which are most prominent in the INL. These primarily consist of a decrease in the number of nuclei, presumably because of an increase in eye size (see Figs. 1A 1B 1C ).
|
|
At 3.9 weeks of age, there was a moderate increase in the number of pyknotic cells in the ONL (Table 3) and very subtle changes in the morphology of the OS. This early stage of OS disruption and misalignment was better seen when dogs were slightly older (
5 weeks of age; stage 1; Fig. 1E ). At that age, nuclear pyknosis had increased in the ONL (Table 3) .
By 7.9 weeks of age (stage 2; Fig. 1F ), there was severe OS disintegration, with abundant disorganized and disoriented membranous material persisting in the photoreceptor layer. ONL thickness was moderately reduced, and the number of pyknotic nuclei in the ONL had decreased (Table 3) . A similar decline in INL thickness, as seen in the normal retina, was observed between 2 and 7.9 weeks of age (see Figs. 1A 1B 1C 1D 1E 1F ) and reached approximately three to four rows of nuclei (see Table 3 ).
By 11.9 and 16 weeks, there was narrowing of the subretinal space with shortening or loss of rod IS, and these abnormalities were comparable at both time points (stage 3; Fig. 1G ). Distorted OS persisted in the photoreceptor layer. There was marked rod loss, with the ONL reduced to approximately 60% of its original thickness. Pyknosis in the ONL was further reduced. A few phagocytic cells located in the subretinal space in close apposition to the RPE were first seen in the 11.9-week-old retina. Their number increased at 16 weeks (Table 3) . Disease-associated thinning of the INL was observed at both ages and was most pronounced in the peripheral retina (Table 3) .
There was a clear decrease in the density of both rod and cone photoreceptors by 26 weeks of age (stage 4; Fig 1H ). This decrease was seen at the level of the photoreceptor layer, where there were areas devoid of any rod and cone IS and also in the ONL where internuclear spacing was increased. ONL thickness was less than 50% of its original thickness. In the photoreceptor layer, the remaining rod and cone IS appeared broader and, although the subretinal space was severely narrowed, shortened, and distorted, OS were still present. Cone nuclei displaced into the IS and extruding into the subretinal space were first seen at this age. Phagocytic cells persisted in the subretinal space.
The latest stage of disease examined was at 40.6 weeks (stage 5; Fig. 1I ). At that time point, there were approximately two to three rows of nuclei left in the ONL, and the remaining cone and rod IS were short and broad. Numerous photoreceptor cells maintained a shortened and misaligned OS, and RPE cytoplasmic processes extended toward them. Phagocytes remained in the subretinal space, but were not migrating in the ONL or inner retina. Nuclear pyknosis persisted in the ONL but was not observed in the INL, despite a decrease in its thickness (Table 3) .
To illustrate the rate of photoreceptor cell loss that occurs during the course of the disease along both the superior and inferior retinal meridians, we plotted the average thickness of the ONL (expressed as either the number of nuclei per column, or in micrometers) as a function of age. These graphs (Fig. 2) show a major and rapid early cell loss occurring from 4 to 12 weeks of age. Subsequently, the number of remaining photoreceptors continued to decrease but at a slower rate. The kinetics of photoreceptor cell loss was best described by a model of constant risk of cell death when the ONL thickness was measured as the number of nuclei. When the ONL thickness was measured in micrometers, the data were best fit by both a model of constant risk and of decreasing risk of cell death.
|
Cell Death
Because of the uniform distribution and rate of disease along the superior and inferior meridians, we determined cell death by using TUNEL assays in sections from the superior meridian. The earliest age when cell death was examined by TUNEL assay was 3.9 weeks, because this was the age when the first morphologic signs of disease were detected. Approximately 31 to 48 TUNEL-labeled cells per 1 M µm2 of ONL were counted. The proportion of photoreceptors undergoing cell death was higher at 5 and 6 weeks, and reached a peak of more than 300 TUNEL-positive cells per 1 M µm2 of ONL at 6.7 weeks of age. In normal retinas of 4, 5, and 6 week-old beagles, the number of TUNEL-positive cells per unit area of ONL was significantly lower and did not exceed six per 1 M µm2 of ONL. At 8 weeks of age, the proportion of photoreceptors undergoing cell death in the mutant retina had decreased to approximately half that occurring at 6.7 weeks (
150 TUNEL-positive cells per 1 M µm2 of ONL). At any given time after 12 weeks of age, the proportion of dying photoreceptors was significantly reduced and close to 80 cells/1 M µm2 of ONL (Fig. 3) . TUNEL-labeled cells were equally distributed throughout the length of the retina, but it appeared that at the earlier ages, there were more dying cells located in the vitreous half of the ONL (Fig. 4A1) . Yet, although, TUNEL-positive photoreceptors were seen in the outer half of the ONL, it was extremely rare before 26 weeks of age to detect any labeling in the outermost row of ONL nuclei, where cone somas are located (Figs. 4A1 4A2) . In 26- and 40.6-week-old XLPRA2 retinas, it was frequent to observe, particularly in the retinal periphery, cone nuclei that were ectopically located in the IS (Fig. 4A3) . Double fluorescence labeling showed that a few displaced cone nuclei were TUNEL positive (Fig. 4A4) . Rare TUNEL-positive cells were also present in the INL and GCL in both mutant and normal young retinas (4–6 weeks), and this was therefore considered to be a normal finding not associated with disease. At later time points in the mutants, there was a relative absence of TUNEL-labeling in those layers.
|
|
The integrity of the RPE was evaluated with an antibody directed against RPE65. There was no loss in RPE65-immunoreactivity in 16- and 26-week-old mutant retinas and the numerous phagocytic cells present in the subretinal space were not labeled by the RPE65 antibody (data not shown). Double-fluorescence immunolabeling with rod opsin and human cone arrestin antibodies showed partial mislocalization of the rod photopigment to the ONL as early as 2 weeks of age in mutant retinas (data not shown). Although rod opsin labeling was restricted to the OS in normal subjects (Fig. 4B1) , distinct staining of the plasma membrane around the rod somas also was visible throughout the entire length and thickness of the ONL in affected dogs at all ages (Fig. 4B2) . In the 7.9-week-old mutant retina, short rod-opsin–positive neurites originating from rod somas extended into the inner retina (Fig. 4B2) , and, in older animals, the sprouting was more prominent and extended deeper into the inner retina, reaching the inner plexiform layer (IPL; Fig. 4B3 ; 40.6 weeks of age). In contrast, cone neurite sprouting was not observed at any stages of the disease examined.
To confirm further that these opsin-positive projections were rod neurites, we performed double immunofluorescence labeling with rod opsin and synaptophysin antibodies (Figs. 4C) . Anti-synaptophysin labeled both plexiform layers in normal subjects (Fig. 4C1) , but, in mutant retinas, outer plexiform layer (OPL) labeling was thinned, and punctuate staining in the INL colocalized with the rod-opsin–positive neurites (Fig. 4C2) . Colocalization occurred at beaded varicosities along the neurites and at their terminals. These had either a bulb-shaped appearance or that of a typical rod spherule (Fig. 4C3) . Although rod opsin and cone arrestin labeling persisted in the 40.6-week-old affected retina, we observed a decrease in cone arrestin immunoreactivity at the level of the cone axons and pedicles (Fig 4B3) that was first visible at 26 weeks of age. Because of the thinning of the ONL at that age (approximately three rows of nuclei), the lengths of the remaining rod and cone axons were significantly shorter than in a normal adults. Even though thinning and disruption of the photoreceptor layer caused occasional retinal separation artifacts during tissue fixation and processing in older retinas, distinct rod opsin and arrestin labeling was observed, respectively, in some of the remaining rod and cone OS (Fig. 4B3) .
To characterize better the two subpopulations of cone photoreceptor cells during the course of disease, we used antibodies raised against short (S)- and medium (M)/long (L)-wavelength cone opsin on retinas at various ages and disease stages. Both S and M/L cone opsin labeling were present in young and older (40.9 weeks) affected dogs. In normal retinas, labeling was restricted to the cone OS (Figs. 4D1 4D3) , but in mutant retinas there was partial mislocalization of the two types of cone opsins to the IS, perinuclear area, axon, and pedicles (Figs. 4D2 4D4) . Mislocalization of S opsin was observed in some S cones, distributed throughout the entire length of the retina, as early as 3.9 weeks of age. By 6 weeks, S opsin mislocalization was found mainly in some peripheral cones (Fig. 4D2) . At later ages S opsin localization was normal. Although the M/L opsin antibody that we used caused some nonspecific background staining of the INL and faint labeling of cone somas and axons, we observed a similar transient and partial mislocalization of the photopigment, particularly in the peripheral retina of 3.9 and 6-week-old affected dogs (Fig. 4D4) . At 8 weeks of age, M/L opsin mislocalization was essentially restricted to the perinuclear area of the cones and did not extend into the axons and pedicles; in older animals, M/L opsin localization was normal.
Because we observed OPL thinning and rod neurite sprouting, we decided to use several inner retina cell markers to determine whether photoreceptor disease and degeneration were also associated with inner retinal changes. A variety of antibodies that label subpopulations of horizontal, bipolar, and amacrine cells, as well as Müller cells were tested in either single or double immunofluorescence analysis.
Anti-calbindin antibody labeled horizontal cells and, to a lesser extent, amacrine and RPE cells in both affected and normal retinas at 4 weeks of age. At later time points in disease, immunostaining was predominantly found in horizontal cells somas and processes. This pattern persisted throughout the course of the disease. Because of the variability in the intensity of calbindin-labeling in the dog, we were not able to quantify with certainty the number of horizontal cells present throughout the length of a retinal section and compare these counts at different stages of the disease. Nevertheless, we were able to observe in older affected retinas a flattening of their axonal arborization associated with the thinning of the OPL (Figs. 5A1 5A2 5A3) .
|
staining of rod bipolar cells showed that these second-order neurons developed normally (Fig. 5B1) in the diseased retina (data not shown). At 11.9 weeks of age there was a mild reduction in the density of their dendritic arborization, which was followed by progressive shortening and total atrophy at later stages (Figs. 5B2 5B3) . By performing double-immunofluorescence experiments with PKC and Go (a cell-marker for ON bipolar cells) antibodies, we were able to distinguish rod bipolar cells that coexpressed PKC and Go from ON-cone bipolar cells that were only Go immunoreactive (Fig. 5C1) . We confirmed that rod bipolar cell dendrites underwent retraction with time, but did not observe a similar change in ON-cone bipolar cells (Figs. 5C2 5C3 5C4) . Indeed, even at the latest time-point examined (40.6 weeks), there was distinct labeling of the dendrites of cone bipolar cells that appear as a continuous layer in the OPL (Fig. 5C4) .
Although we tested a variety of antibodies (anti-ChAT, anti-TH, anti-Dab1, anti-
-aminobutyric acid [GABA], see Table 2 ) reported to label different subpopulations of amacrine cells in rodents, we were successful only in detecting GABAergic amacrine cells in the canine retina. Labeling of amacrine cell bodies with the GABA antibody was limited to the central retina in the normal adult, yet intense staining of the IPL laminae was seen throughout the entire retina (Fig. 5D1) . A similar pattern was seen in the mutant retina until 11.9 weeks of age. Thereafter (26 and 40.6 weeks), there was a significant increase in the number of GABA-immunoreactive amacrine cells. These were located both at the inner border of the INL as well as displaced into the ganglion cell layer (GCL), and they were found all along the length of the retina. In addition, there was a thinning of the IPL and a loss of its normal lamination (Fig. 5D2 5D3) The GABA antibody also labeled the somas and processes of horizontal cells in the normal canine retina (Fig. 5D1) . This was also observed in young affected retinas until 8 weeks of age. In the 11.9-week-old mutant retina there was decrease in the intensity of the labeling, and, after 26 weeks, no staining of any horizontal cell was observed, although it was distinct in the normal retina (Fig. 5 , compare D1 with D2, D3).
Because the cellular retinaldehyde-binding protein (CRALBP) antibody that we used to examine Müller cells did not label canine retina, we used instead an antibody directed against glutamine synthetase. With this antibody, we found a decrease in Müller cell length in mutant retinas associated with the thinning of the ONL; however, there was no apparent reduction in their density at 11.9 weeks of age. By 26 weeks of age, there was a significant reduction in Müller cell immunoreactivity, and by 40.6 weeks, immunolabeling was almost absent (data not shown). This precluded assessment of glial cell loss at later stages of the disease. GFAP immunolabeling was used to evaluate the level of glial reactivity in Müller cells. GFAP staining was limited to astrocytes and Müller cell end feet in normal retinas of all ages (Fig. 5E1) as well as in the youngest (3.9 weeks) retina. A gradual increase in GFAP immunoreactivity began at 5 weeks and peaked at 8–12 weeks of age (Fig. 5E2) . Minimal GFAP reactivity was seen in older retinas when outer retinal atrophy was more advanced (Fig. 5E3) .
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Evaluation of mutant retinas showed that this was a very early-onset disease characterized by visible abnormalities in both rods and cones before their maturation. At 3.9 weeks of age, the earliest signs of OS disruption were detectable on 1-µm-thick plastic-embedded sections, and mislocalization of rod and cone opsins was seen by immunohistochemistry. At this same age, there was clear evidence of rod photoreceptors undergoing cell death. We confirmed that this differed from the normal developmental cell death process, which was found at 2 weeks of age in both normal and mutant retinas (data not shown). In the mutant retina the number of TUNEL-labeled photoreceptor cells was significantly higher than in age-matched control subjects. Over the ensuing weeks, there was an increase in the proportion of TUNEL-labeled photoreceptors that resulted in a peak of cell death between 6 and 7 weeks. After this early burst of cell death, the rate was significantly decreased after 12 weeks of age. Our findings are similar to that reported in the Rdy cat, a model of autosomal dominant rod–cone dysplasia, in which an early onset of photoreceptor death begins at 5 weeks of age and peaks at approximately 9 weeks of age.22 Our results clearly suggest that the risk of death in a single photoreceptor cell is not the same at all ages and that the mathematical model of a constant or decreasing risk of photoreceptor death, as suggested by Clarke et al.,17 may not be applicable to this class of RPGR mutations, at least during the initial course of the disease.
Although the data on the ONL thickness (Fig. 2 , Table 4 ) are best fit with an exponentially declining curve that would be consistent with the kinetics of a constant or decreasing risk of photoreceptor death in XLPRA2, the TUNEL data (Fig. 3) do not support this model. We saw (Fig. 3) two distinct phases of cell death. An initial phase, from 3.9 to 7 weeks of age, showed an increased risk for photoreceptors to die (peak of TUNEL-positive cells at 6 to 7 weeks of age). This was followed by a second phase of a rapid (from 7 to 12 weeks) and then a more gradual decrease (after 12 weeks) in the number of TUNEL-positive cells per unit area of ONL. It is therefore possible that the limited number of observations between 3.9 and 7 weeks of age fails to show an early sigmoidal decline in photoreceptor cell number, which we would have expected to observe with an increased risk of cell death. This initial phase of photoreceptor death does not appear to be unique to XLPRA2. Indeed, a group of investigators used TUNEL assay to look at the kinetics of cell death in the rd mouse and recorded a similar increase in photoreceptor death that reached its peak by 16 days of age.34 Not surprisingly, in their studies, Clarke et al.17 35 reported that the rd mouse was the only animal model for which an increased risk of cell death could not be excluded, since the data fit equally well to mathematical models of constant and exponentially increasing risk.
|
The mislocalization of opsins in both rods and cones at the early stages of the disease suggests that RPGR is expressed in both populations of photoreceptors, and lends additional support to the hypothesis that RPGR is involved in the trafficking of proteins from the IS to the OS, or in their retention in the OS.13 A similar mislocalization of both rod and cone opsins has been reported recently in a human carrier (Adamian M, et al. IOVS 2005;46:ARVO E-Abstract 3400), yet, in the mouse, only cone opsins appear to mislocalize.8 In the present study, we found early signs of cone disease both in plastic-embedded sections and in cryosections treated for immunohistochemistry; however, we did not detect any features of cone cell death before the age of 26 weeks. This raises the question of whether the mechanism of cone death is directly caused by the RPGR mutation or is non–cell-autonomous and secondary to rod degeneration. In animal models of retinitis pigmentosa caused by mutations in rod-specific genes, the delayed cone loss is thought to be caused by the death of rods that induces structural alterations in the photoreceptor layer and ONL, the release of toxic byproducts, and/or a decrease in the secretion of cone survival factors.38 39
Contrary to implications of the results in immunocytochemical studies on human retinas with advanced stages of retinitis pigmentosa,23 40 inner retinal remodeling is not, at least in the dog, a late response to photoreceptor degeneration. Our observations showed that synaptic connectivity in the OPL was altered at early stages of the disease. Rod neurite sprouting was first apparent within 1 to 2 weeks after the peak of photoreceptor cell death that occurred at approximately 7 weeks of age. Although this is the first report of rod neurite sprouting in the canine retina, it has been observed in two other animal models of retinitis pigmentosa: the rhodopsin transgenic pig41 and the Rdy cat with autosomal dominant rod–cone dysplasia.22 Neurite sprouting of cones has been described in the rd mouse as an early-onset change that starts at P8, when rod degeneration begins.42 Neither this abnormality, nor elongation and branching of cone axons as previously described in humans40 were observed in the XLPRA2 retina at any studied age, but these changes may occur in much older animals with more advanced disease. Concomitant to the onset of rod neurite sprouting, there was a loss in synaptophysin immunoreactivity of the OPL as early as 7.9 weeks of age. This illustrates the possible rapid and early disorganization in synaptic connectivity that could seriously hamper retinal function at a stage of disease when photoreceptor loss is limited.
Another early change that occurred throughout the entire thickness of the neuroretina was the increased reactivity of Müller cells. An increase in GFAP labeling of Müller cells occurred approximately 2 weeks before the peak of photoreceptor cell death and reached its highest level in the following weeks. These glial cells, with somas located in the INL, have apical radial processes extending into the external limiting membrane (ELM). It is therefore possible that Müller cells detect early structural or chemical modifications in the photoreceptor layer and/or ONL caused by the degeneration of photoreceptor cells and relay this change in the outer retina environment to deeper retinal layers. GFAP reactivity in Müller cells had decreased at 26 weeks of age and was nearly normal by 40.6 weeks. These findings are different from those reported in human retinas with advanced disease in which GFAP reactivity persists.43 This difference may be explained by the fact that we did not collect retinas from the very advanced stages of disease. Thus, it may be that GFAP reactivity in Müller cells occurs both at the onset of photoreceptor degeneration and during the terminal stages of retina atrophy.
After these early changes in the inner retina, there was progressive retraction of rod bipolar cell dendrites that began by approximately 12 weeks of age. Contrary to what has been described in the rd mouse,44 rod bipolar cells developed dendritic arborizations. This may be because in XLPRA2, unlike in the rd mouse, most photoreceptor cells reach a stage of functional and structural maturation, albeit abnormal, that allows the formation of synapses with second-order neurons. We did not observe any significant loss in the arborization of ON-cone bipolar cells, as occurs in the rd mouse at later stages of the disease, when cones undergo cell death.25 The most probable explanation of this difference is that, at the latest time point that we examined ON cone bipolar cells (40.6 weeks), there was still a significant number of cone photoreceptors in the ONL that had not degenerated. This may suggest that the cone-mediated pathway remains functional at more advanced stages of the disease. Although more extensive electroretinographic testing is needed to verify this hypothesis, our previous study has shown the presence of robust cone signals at a time when there is significant loss of rod-mediated responses.10
Information on the histopathologic changes that occur in retinas of human patients with RPGR exon ORF15 mutations has currently only been reported in a carrier.12 Yet, it appears that frameshift mutations in RPGR exon ORF15 have a comparably severe phenotype in both canine and human retinas.
The findings reported in this study have several important implications for the development of therapeutic approaches for retinal degeneration in humans. Among these, the use of survival factors (e.g., ciliary neurotrophic factor [CNTF] and brain-derived neurotrophic factor [BDNF]) as well as a gene-silencing approach, via ribozymes or siRNA, are currently being investigated in our laboratory. The results of this present study suggest that a first strategy should consist in initiating therapy before the burst of photoreceptor cell death that occurs in the dog between 6 and 7 weeks of age. Delivering the therapeutic agent at approximately 4 weeks of age may prevent or delay the onset of photoreceptor degeneration. Intraocular injection of survival factors that are immediately biologically active is feasible at such a young age. Alternatively, another potential strategy would be to target the rescue of the photoreceptor cells that survived the initial burst of cell death. Indeed, we have shown that after this event, the rate of cell death is considerably slowed down, and that even in the most advanced stages of disease, the morphology of the remaining photoreceptor cells is reasonably preserved. We have observed that some photoreceptors maintain their IS and OS and continue to express proteins involved in the phototransduction pathway unlike what has been reported in humans with more advanced retinitis pigmentosa caused by rhodopsin gene mutations.30 From this perspective, 12 weeks of age may be an optimal time-window to initiate therapy because, at that age, approximately 60% of photoreceptor cells remain.
Another important aspect that future studies should address is whether rescuing photoreceptor cells also allows reversal of inner retinal changes that occur secondary to rod and cone disease and degeneration. Modification of the inner retina after photoreceptor degeneration does not appear to be dependent on the genetic cause of the disease. Indeed, in addition to the two-nucleotide deletion in RPGR ORF15 that occurs in the XLPRA2 dog, recent studies have shown inner retina remodeling in the rd and crx–/– mice, two models of retinal degeneration caused, respectively, by a mutation in the PDE6B and CRX genes.24 44 In addition to these morphologic changes, a switch in neurotransmitter sensitivity of rod bipolar cells has been demonstrated in the rd mouse.45 All these recent findings suggest that, if a similar phenomenon occurs in the human with early stages of retinal degeneration, then novel therapeutic approaches must be evaluated, not solely on their protective effect on photoreceptor cells, but also on their capacity in maintaining functional synaptic connections between photoreceptor cells and inner retinal neurons.
In conclusion, we have characterized the structural changes that occur in the XLPRA2 retina, an early-onset canine model of X-linked retinitis pigmentosa caused by a microdeletion in RPGR exon ORF15 with resultant frameshift. This is a valuable spontaneous animal model that may provide a better understanding of the retinal function of the RPGR protein and the pathogenic mechanisms that lead to photoreceptor death. It also may provide a tool to assess the in vivo efficacy of novel therapies.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Submitted for publication June 30, 2005; revised December 2, 2005; accepted February 23, 2006.
Disclosure: W.A. Beltran, None; P. Hammond, None; G.M. Acland, None; G.D. Aguirre, 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.
Corresponding author: Gustavo D. Aguirre, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA 19104; gda{at}vet.upenn.edu.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  X. Shu, Z. Zeng, P. Gautier, A. Lennon, M. Gakovic, E. E. Patton, and A. F. Wright Zebrafish Rpgr is required for normal retinal development and plays a role in dynein-based retrograde transport processes Hum. Mol. Genet., February 15, 2010; 19(4): 657 - 670. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Brunner, S. Skosyrski, R. Kirschner-Schwabe, K.-P. Knobeloch, J. Neidhardt, S. Feil, E. Glaus, U. F. O. Luhmann, K. Ruther, and W. Berger Cone versus Rod Disease in a Mutant Rpgr Mouse Caused by Different Genetic Backgrounds Invest. Ophthalmol. Vis. Sci., February 1, 2010; 51(2): 1106 - 1115. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Sakai, M. Kondo, S. Ueno, T. Koyasu, K. Komeima, and H. Terasaki Supernormal ERG Oscillatory Potentials in Transgenic Rabbit with Rhodopsin P347L Mutation and Retinal Degeneration Invest. Ophthalmol. Vis. Sci., September 1, 2009; 50(9): 4402 - 4409. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Beltran, G. M. Acland, and G. D. Aguirre Age-Dependent Disease Expression Determines Remodeling of the Retinal Mosaic in Carriers of RPGR Exon ORF15 Mutations Invest. Ophthalmol. Vis. Sci., August 1, 2009; 50(8): 3985 - 3995. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Gu, W. A. Beltran, S. Pearce-Kelling, Z. Li, G. M. Acland, and G. D. Aguirre Steroids Do Not Prevent Photoreceptor Degeneration in the Light-Exposed T4R Rhodopsin Mutant Dog Retina Irrespective of AP-1 Inhibition Invest. Ophthalmol. Vis. Sci., July 1, 2009; 50(7): 3482 - 3494. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Tuntivanich, S. J. Pittler, A. J. Fischer, G. Omar, M. Kiupel, A. Weber, S. Yao, J. P. Steibel, N. W. Khan, and S. M. Petersen-Jones Characterization of a Canine Model of Autosomal Recessive Retinitis Pigmentosa due to a PDE6A Mutation Invest. Ophthalmol. Vis. Sci., February 1, 2009; 50(2): 801 - 813. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Specht, S.-B. Wu, P. Turner, P. Dearden, F. Koentgen, U. Wolfrum, M. Maw, J. H. Brandstatter, and S. tom Dieck Effects of Presynaptic Mutations on a Postsynaptic Cacna1s Calcium Channel Colocalized with mGluR6 at Mouse Photoreceptor Ribbon Synapses Invest. Ophthalmol. Vis. Sci., February 1, 2009; 50(2): 505 - 515. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Gu, W. A. Beltran, Z. Li, G. M. Acland, and G. D. Aguirre Clinical Light Exposure, Photoreceptor Degeneration, and AP-1 Activation: A Cell Death or Cell Survival Signal in the Rhodopsin Mutant Retina? Invest. Ophthalmol. Vis. Sci., November 1, 2007; 48(11): 4907 - 4918. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. S. Aleman, A. V. Cideciyan, A. Sumaroka, S. B. Schwartz, A. J. Roman, E. A. M. Windsor, J. D. Steinberg, K. Branham, M. Othman, A. Swaroop, et al. Inner Retinal Abnormalities in X-linked Retinitis Pigmentosa with RPGR Mutations Invest. Ophthalmol. Vis. Sci., October 1, 2007; 48(10): 4759 - 4765. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. G. Jacobson, T. S. Aleman, A. V. Cideciyan, E. Heon, M. Golczak, W. A. Beltran, A. Sumaroka, S. B. Schwartz, A. J. Roman, E. A. M. Windsor, et al. Human cone photoreceptor dependence on RPE65 isomerase PNAS, September 18, 2007; 104(38): 15123 - 15128. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
