A, Fraction of outer nuclear layer (ONL) cells positive on terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL) staining reached a peak at day 3 after separation of the retina from the retinal pigment epithelium (RPE) in both the rat and mouse. The percentage of TUNEL-positive cells in the ONL was significantly reduced (asterisks, P = .05) when Fas signaling was inhibited by the injection of a Fas-neutralizing antibody (Fas-NAb) or when the separation was induced in the Fas mutant mouse strain LPR. At day 7 postseparation, the fraction of TUNEL-positive cells decreased, as expected,4,7 to near baseline levels. Fas inhibition did not affect this decrease, suggesting that Fas inhibition resulted in fewer cells entering the apoptotic cascade rather than just being delayed in their entrance. Representative photomicrograph showing the reduction in TUNEL-positive staining of photoreceptor cell nuclei (arrow) when retina/RPE separation was induced in the LPR mouse (C) as compared with the control (C57BL mouse) (B). Nuclear staining was done in control (D) and LPR (E) mice with propidium iodide (original magnification × 20).
Outer nuclear layer (ONL) cell counts (A) and thickness (B) before and after separation of the retina from the retinal pigment epithelium (RPE). Histologic analysis of an attached section (C), a section in which the retina and RPE are separated (D), a separated section treated with a Fas-neutralizing antibody (Fas-NAb) (E), and a separated section treated with small inhibitory RNA against the Fas transcript (siFas) (F) demonstrating the ability of reduced Fas-receptor activity to preserve ONL cell counts and thickness after retina/RPE separation (toluidine blue, original magnification × 40).
A, The increase in Fas-receptor transcript levels induced by the retina/retinal pigment epithelium (RPE) separation can be reduced by the subretinal injection of small inhibitory RNA against the Fas-receptor transcript (siFAS). B, The reduced transcript level caused by an siFAS injection also results in a reduced level of Fas-receptor protein as detected on immunoblot (retina/RPE separation, left; retina/RPE separation treated with siFas, right). There was equal loading across both lanes.
Immunohistochemical analysis demonstrating the decrease in rhodopsin (A) and cone opsin (D) staining after retina/retinal pigment epithelium separation. Inhibition of Fas signaling by treatment with either Fas-neutralizing antibodies or small inhibitory RNA against the Fas-receptor results in rhodopsin (B and C) and cone opsin (E and F) staining that approaches the levels seen in intact attached retinas. This suggests that the preserved photoreceptors are capable of producing normal photopigment.
Zacks DN, Boehlke C, Richards A, Zheng Q. Role of the Fas-Signaling Pathway in Photoreceptor Neuroprotection. Arch Ophthalmol. 2007;125(10):1389-1395. doi:10.1001/archopht.125.10.1389
To determine whether inhibiting the Fas proapoptosis pathway will result in increased photoreceptor survival after separation of the retina from the retinal pigment epithelium (RPE).
Retina/RPE separation was induced in rat and mouse eyes by the subretinal injection of hyaluronic acid, 1%. Fas-pathway signaling was inhibited by the concomitant injection of a Fas receptor–neutralizing antibody, small inhibitory RNA against the Fas-receptor transcript (siFAS), or the use of the Fas-receptor defective mouse strain LPR. Indices of photoreceptor death included terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL) staining, cell counts, and retinal thickness measurements. Retinas were immunostained with antibodies against rhodopsin and cone opsin to evaluate rod and cone photopigment production, respectively.
Inhibition of Fas signaling using Fas receptor–neutralizing antibody, siFas, or LPR mice resulted in a significant reduction in the number of TUNEL-positive photoreceptor cells as well as in a significant preservation of outer nuclear layer cell counts and thickness as compared with retina/RPE separation in eyes with intact Fas signaling. Fas-pathway inhibition resulted in preservation of both rhodopsin- and cone opsin–positive cells.
Inhibition of the Fas proapoptosis pathway results in significant photoreceptor preservation after retinal separation from the RPE.
Fas-pathway inhibition might serve as a novel mechanism for preserving photoreceptor cells during retinal disease.
Photoreceptors, the primary transducers of visual stimuli, receive the majority of their nutritional and metabolic support from the underlying retinal pigment epithelium (RPE).1,2 As part of many retinal disease processes, however, the photoreceptors become separated from the RPE. A primary consequence of this separation, and the subsequent disruption of normal homeostasis, is the apoptotic death of the photoreceptor.3- 7 Even as the underlying disease is being treated, the interval of photoreceptor/RPE separation can result in significant amounts of photoreceptor deaths.8- 13 This makes the development of adjunctive photoreceptor-protective treatments of paramount importance for improving patients' visual outcomes.
During periods of stress, multiple signaling events can act concurrently within a cell to control apoptosis.14- 18 Numerous signaling pathways have been described in a variety of different tissues and cell types. We have shown previously that within the retina there is separation-induced activation of the Fas proapoptotic pathway as well as activation of the intrinsic (mitochondrial) pathway and caspase 9, a serine protease and key member of the intrinsic pathway that serves as an initiator of apoptosis.6 We also showed that Fas-pathway activation precedes and can act as an upstream initiator of intrinsic pathway activation.6 Interestingly, retina/RPE separation induces transcription of Fas-pathway genes but not of genes encoding proteins in the intrinsic pathway, suggesting that gene transcription serves as an additional point of regulation of photoreceptor apoptosis.6
Based on these results, we designed this study to test the hypothesis that the inhibition of Fas signaling will result in enhanced photoreceptor survival during periods of retinal separation from the RPE. This hypothesis led to 3 predictions: first, inhibition of Fas-pathway signaling at the time of retina/RPE separation will result in fewer terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL)–positive cells in the outer nuclear layer (ONL) as compared with the control eyes where separation is induced in the absence of Fas inhibition. Second, inhibition of Fas signaling will result in increased ONL cell counts and increased ONL thickness—both markers of the number of photoreceptors—after chronic separation as compared with control eyes. Finally, in experimental eyes, there will be enhanced preservation of rhodopsin and cone opsin immunolabeling, histologic markers of rod and cone cell viability, respectively. In this study, we confirmed all 3 of these predictions, providing strong support for our hypothesis that inhibition of the Fas-signaling pathway results in significant photoreceptor preservation during periods of retina/RPE separation.
All animal experiments were performed in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Visual Research and the guidelines established by the University Committee on Use and Care of Animals of the University of Michigan. Retina/RPE separations were created as previously described.5 Adult brown Norway rats were anesthetized with a 1:1 mixture of ketamine (100 mg/mL) and xylazine (20 mg/mL), and pupils were dilated with topical phenylephrine, 2.5%, and tropicamide, 1%. A sclerotomy was created approximately 1 to 2 mm posterior to the limbus with a 20-gauge microvitreoretinal blade, with special caution to avoid damaging the lens. A Glaser subretinal injector with a 32-gauge tip (BD Ophthalmic Systems, Sarasota, Florida) that was connected to a syringe filled with 10 mg/mL sodium hyaluronate (Healon; Pharmacia and Upjohn Co, Kalamazoo, Michigan) was introduced through the sclerotomy into the vitreous cavity. The tip of the subretinal injector was introduced into the subretinal space through a peripheral retinotomy and the sodium hyaluronate was slowly injected, thus separating the neurosensory retina from the underlying RPE. In all experiments, approximately one-third of the retina became separated, with a total volume injection of approximately 80 to 100 μL of hyaluronic acid. To create retina/RPE separation in the mouse, essentially the same procedure was performed. The brown Norway rats were retired breeders obtained from Charles River Laboratories (Wilmington, Massachusetts). The 2 mouse strains used were C57BL and LPR (on a C57BL background; Jackson Laboratory, Bar Harbor, Maine). Mice were between the ages of 3 and 6 weeks at the time of injection.
We employed 3 different techniques to inhibit signaling by the Fas receptor. The first method was to inject a Fas receptor–neutralizing antibody (Fas-NAb) into the subretinal space at the time of retina/RPE separation. Ten micrograms of Fas-NAb was delivered as previously described6 (clone ZB4; Upstate Biotechnology, Lake Placid, New York) in 10 μL of phosphate-buffered saline injected into the already detached subretinal space using a syringe (Hamilton Corp, Reno, Nevada) passed through the same sclerotomy and retinotomy used to create the retina/RPE separation. Control eyes had retina/RPE separation induced but then received a subretinal injection of phosphate-buffered saline without the Fas-NAb.
The second technique for inhibiting Fas-pathway signaling was to inject small inhibitory RNA against the Fas receptor (siFas) to prevent the separation-induced increase in transcription of that gene. Sequences for the small inhibitory RNA were provided by Invitrogen Life Technologies (Carlsbad, California). A small inhibitory RNA dose of 0.5 × 10−9 mol was injected into the subretinal space of the detached retina in a volume of 10 μL of phosphate-buffered saline. Injection of siFas was also performed using a Hamilton syringe (Hamilton Corp), which passed through the same sclerotomy and retinotomy used to create the retina/RPE separation.
Finally, retina/RPE separation was induced in the LPR mouse strain (bred on a C57BL background), which contains a null mutation for the Fas-receptor gene and are thus unable to transduce the Fas signal. As a control, retina/RPE separation was induced in C57BL mice.
Extraction of RNA and quantitative real-time polymerase chain reaction (qRT-PCR) was performed as previously described.6 Briefly, total RNA was isolated with Trizol reagent (Invitrogen, Carlsbad, California). First-strand complementary DNA was synthesized from total RNA using Superscript III reverse transcriptase and oligo(dT)20 (Invitrogen). Real-time PCR was performed by incubating complementary DNA, primers, and iQ SYBR Green Supermix in a Bio-Rad iCycler (Bio-Rad Laboratories, Hercules, California). Primers specific for the rat hypoxanthine phosphoribosyl transferase gene were used as an internal control to allow for normalization and direct comparison between multiple samples. All primers were designed to span intron/exon boundaries to distinguish between transcripts and any contaminating genomic DNA. Samples lacking reverse transcriptase or a complementary DNA template served as negative controls. For each primer set, qRT-PCR was performed on samples derived from 3 different animals and repeated 3 times per animal. Quality analysis of the PCR products was performed by running an aliquot of each sample on a TAE, 2%, gel. In addition, melting curves for each PCR reaction were analyzed to ensure that dimerization of PCR primers did not provide a false-positive response. Each reaction was performed 3 times on 3 independent samples derived from separate animals. The mean change in transcript levels between eyes with retina/RPE separation vs eyes without retina/RPE separation was then calculated.
Immunoblot analysis of the Fas-receptor protein was performed as previously described.6 The anti-Fas primary antibody (Santa Cruz Biotechnology, Santa Cruz, California) was used at a dilution of 1:1000. Equal loading was verified by Ponceau-S staining and densitometry analysis of a nonspecific band present across all lanes as previously described.6
Preparation of eyes for histology was performed as previously described.6 The eyes were enucleated and placed in paraformaldehyde, 4%, overnight at 4°C. Whole eyes were then placed in a Tissue-Tek II tissue processor (Sakura, Tokyo, Japan) for standard paraffin embedding. Eyes were then sectioned at a width of 6 μm on a standard paraffin microtome. Toluidine blue staining of sections was performed using a standard histological technique.
TUNEL staining was performed on paraffin sections using the ApopTag Fluorescein In Situ Apoptosis Detection Kit (Chemicon International, Temecula, California) according to the manufacturer's instructions. Propidium iodide (Chemicon International) was used for nuclear counterstaining.
Immunohistochemistry was performed on sections obtained from paraffin-embedded retinas. Epitope unmasking was performed by incubating the sections in a boiling 10-mM citrate buffer (pH = 6.0) for 10 minutes. Slides were then washed with phosphate-buffered saline with Tween 20, 0.5%, incubated for 10 minutes in hydrogen peroxide, 3%, at room temperature, followed by another phosphate-buffered saline with a Tween 20, 0.5%, wash. The incubation with primary and secondary antibodies was performed with a Tyramide Signal Amplification–Fluorescein Detection Kit (PerkinElmer, Boston, Massachusetts) following the manufacturer's instructions. The primary antibodies used were a monoclonal anti-rhodopsin (No. MAB5356; Chemicon International) and a polyclonal anti-opsin, red/green (No. AB5405; Chemicon International) antibody. Secondary antibodies used were Alexa Fluor 488 and Alexa Fluor 546, respectively (Invitrogen).
Three main outcome measures were used in this study: TUNEL-positive cells in the ONL, total ONL cell counts, and ONL thickness as a percentage of total retinal thickness. TUNEL counts were performed by selecting 3 random high-power (40 ×) fields per section and counting the number of TUNEL-positive cells and dividing by the total number of ONL cells present in the same field. Counts of ONL cells were performed in a similar manner on 3 randomly selected high-power fields per section. Normalization of ONL counts—performed to allow for intersample comparison and to account for possible differences in angles of sectioning between eyes—was done by dividing the cell count by the total retinal thickness. Outer nuclear layer thickness (in microns) was measured in the same 3 positions along the section, and the mean ONL thickness was normalized against the mean total retinal thickness for that section. Total retinal thickness (in microns) was measured from the outer edge of the ONL to the inner edge (closest to the vitreous) of the ganglion cell layer. Photoreceptor inner and outer segments were not included in the total retinal thickness, as there is variable retraction of these elements postseparation and does not necessarily reflect the postreattachment viability of the cell.19,20 The thickness of the inner retina (total retina minus ONL) did not vary significantly with time after separation, regardless of treatment group, suggesting that differences in normalized ONL counts were because of the number of cells present in the ONL and not variability in the denominator used for normalization. For each high-power field, a mean retinal thickness was obtained by averaging the thickness at 3 equally spaced points along the section. For all experiments, measurements were done on 3 sections from 10 eyes, each eye from a separate animal. All counts were done in a masked fashion. Comparison between mean counts was performed using a Fisher exact t test.
The first prediction that we tested was whether inhibition of Fas-pathway signaling will prevent photoreceptors from entering the apoptotic cascade, as measured by TUNEL staining. Previously published work in rats and mice has shown that the peak of TUNEL staining occurs at 3 days after retina/RPE separation, with a rapid decline in TUNEL-positive cells to near-preseparation levels by day 7.4,7 These results were confirmed in our study, with a peak of approximately 5% of ONL cells displaying TUNEL-positive staining at day 3 postseparation, with a rapid decline to less than 1% by day 7 postseparation (Figure 1). In our rat model of retina/RPE separation, injection of Fas-NAb into the subretinal space resulted in 60% fewer TUNEL-positive photoreceptors detected at day 3 postseparation as compared with separated retinas not injected with Fas-NAb (P = .05) (Figure 1). Similarly, approximately 75% fewer TUNEL-positive photo receptors were detected when the retina/RPE separation was created in the LPR mouse (Figure 1). We did not detect an increase in TUNEL-positive cells at day 7 postseparation in Fas-NAb–treated eyes or LPR mice as compared with controls (Figure 1), which was consistent with the conclusion that inhibition of Fas-receptor activation results in fewer cells entering the apoptotic cascade rather than just delaying activation of apoptosis.
We next sought to test whether the decrease in TUNEL-positive photoreceptor cells observed at 3 days corresponds to greater numbers of photoreceptor cells present after retina/RPE separation of longer duration, as predicted by our hypothesis. Figure 2 shows that rat retinas separated for 2 months showed a 55% reduction in ONL cell counts (P = .001) and a 47% reduction in ONL thickness as compared with attached retinas (P = .001). In contrast, eyes that received a single injection of Fas-NAb at the time of retina/RPE separation had only a 15% reduction in ONL cell counts and 16% reduction in ONL thickness at 2 months. That is, treated retinas had 90% more ONL cells (P = .001) and the ONL was 60% thicker (P = .001) than untreated retinas.
Consistent with our previously published results,6 separation of the retina from the RPE resulted in a marked increase in the transcription of the Fas receptor (Figure 3). However, a single dose of siFAS injected into the subretinal space at the time the retina/RPE separation was created resulted in a significant reduction in the level of transcript seen, as well as in the amount of Fas-receptor protein produced (Figure 3). As seen in Figure 2, the injection of siFAS at the time of separation also resulted in a 68% (P = .001) and 73% (P = .001) increase in ONL cell counts and ONL thickness at 2 months, respectively, as compared with untreated retinas separated from the RPE. Essentially, the same result was also observed when the retina/RPE separation was performed in the LPR mouse, with a 74% (P = .001) and 79% (P = .001) increase in ONL cell counts and ONL thickness at 2 months, respectively, as compared with retina/RPE separation in C57BL mice.
Finally, we performed immunostaining of retinal sections using antibodies against rhodopsin and cone opsin as a measure of cell density and viability (Figure 4). Though qualitative in nature, there appeared to be much less staining for both proteins in the retinas of eyes with 2-month separations, consistent with the significantly reduced number of photoreceptor cells present. For eyes treated with Fas-NAb or siFAS, both of which resulted in significant preservation of photoreceptors in these 2-month separations, we observed an increased level of staining that approaches that seen in the attached control retinas. These data suggest that the preserved cells are viable and capable of producing photopigments.
The results presented in this report confirm the importance of Fas-pathway signaling in controlling the separation-induced death of photoreceptors, with inhibition of this pathway resulting in greatly enhanced numbers of photoreceptors that survive retina/RPE separation. Regardless of how inhibition was accomplished—by blocking the Fas receptor, preventing translation of the Fas-receptor messenger RNA, or using a mouse with a defective Fas receptor—significantly more photoreceptors survived than in retinas with intact Fas signaling. These data strongly support the possibility of using Fas-pathway blockade as a therapeutic modality to prevent photoreceptor death during periods of retinal separation from the RPE.
Injury-induced transcription of the Fas receptor is not unique to the retina. In animal models of cerebral ischemia, there is an increase in Fas transcription, particularly in the penumbral region.21- 23 In other words, cells that survive the initial injury increase their level of the Fas receptor. In our experimental model of retina/RPE separation, we also noted an increase in the Fas-receptor transcript in the cells that survive the initial separation.6 In this study, we show that inhibiting this transcription appears to prevent the death of these photoreceptors, suggesting 2 phases of separation-induced apoptosis. The first phase is an immediate death of a portion of the photoreceptors, using preexisting stores of proapoptosis proteins (both Fas dependent6 and Fas independent4). TUNEL assays and cell counts in experimental models of retina/RPE separation suggest that approximately 5% of cells enter apoptosis in this initial phase.4,7 This would then be followed by the second phase of cell death, which our data suggest is predominantly Fas dependent and depends on the increased transcription of Fas-pathway intermediates. This phase is most likely the one affected by our small inhibitory RNA injection and could help explain why our photoreceptor counts with this intervention never exceeded 90% of preseparation levels. This also suggests a potential role for combination therapy aimed at both the Fas-dependent and Fas-independent pathways.
Although we demonstrate significant levels of photoreceptor preservation, the effect was incomplete. This could represent a dose effect or an inability of the injected agents to adequately access their target cells. Another possibility, however, is that some cells enter the apoptosis pathway in a Fas-independent manner. From our previous work, we know that the separation-induced activation of the intrinsic pathway is only partially controlled by the activation of Fas,6 consistent with the possibility of an alternate activator of the intrinsic pathway. One such alternate activator of separation-induced photoreceptor apoptosis was proposed by Hisatomi and colleagues,4 who demonstrated the activation of the apoptosis-inducing factor–dependent apoptosis pathway in an experimental rat model of retina/RPE separation similar to the one used in this study. Our results, however, show that a preponderance of the cell death is driven by the Fas proapoptosis pathway.
When Hisatomi et al4 created retina/RPE separation in LPR mice, however, they were unable to detect decreased TUNEL staining in the ONL and concluded that the Fas proapoptotic system does not play a role in photoreceptor death in this type of injury. In their experiments, however, the LPR mutation was carried on a C3H background,4 whereas we used the C57BL background. While we cannot state with certainty why they did not observe the same results that we did, the difference in background strain serves as a major confounding variable. The C3H background is a carrier for the retinal degeneration 1 (rd1) allele.24,25 Retinal degeneration in rd1 mice is predominantly driven by apoptosis-inducing factor and caspase 12,26 suggesting that perhaps photoreceptor apoptosis after retina/RPE separation in the C3H background, even if the mouse is only heterozygous for the rd1 allele, is driven more by apoptosis-inducing factor than by Fas.
Our results suggest that treated eyes continue to produce components of the visual transduction pathway, namely rhodopsin and cone opsin. Yang and colleagues7 previously reported that retina/RPE separation causes relatively equal rates of apoptosis in rods and cones. They were also able to demonstrate an equal reduction in TUNEL staining for both cell types when retina/RPE separation was induced in a Bax-deficient mouse.7 Bax is a proapoptosis protein that can be activated by Fas and can result in mitochondrial permeabilization and intrinsic pathway activation.27 Our results are consistent with those of Yang et al,7 with Fas inhibition preventing the decrease in separation-induced staining for both rhodopsin and cone opsin, markers of rods and cones, respectively.
The potential clinical relevance for anti-Fas therapy extends across a broad spectrum of retinal diseases that exhibit a component of retina/RPE separation. The utility of such treatment would most likely be as an adjunct aimed at preserving the photoreceptors while the underlying condition is being addressed. Though additional apoptosis pathways may be activated after retina/RPE separation, our data suggest that a large number of photoreceptors can be preserved by inhibiting Fas-pathway activation. Further work is required to more clearly define the therapeutic parameters within which modulation of Fas-pathway activity will allow for increased photoreceptor survival, but adjunctive therapy with such a photoreceptor-protective agent will potentially allow for improved visual outcomes in patients.
Correspondence: David N. Zacks, MD, PhD, Department of Ophthalmology and Visual Sciences, University of Michigan Medical School, 1000 Wall St, Ann Arbor, MI 48105.
Submitted for Publication: January 31, 2007; final revision received May 14, 2007; accepted May 30, 2007.
Author Contributions: Dr Zacks had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Financial Disclosure: None reported.
Funding/Support: This study was supported by grant K08-EY-14705 from the National Eye Institute (Dr Zacks).