Preventing Retinal Detachment–Associated Photoreceptor Cell Loss in Bax-Deficient Mice

Abstract

purpose. To characterize photoreceptor cell apoptosis and cell loss in a mouse model of experimental retinal detachment (RD), and to use the technology of mouse genetics to study the molecular mechanisms underlying RD-associated photoreceptor degeneration.

methods. Retinal detachments were created in adult wild-type and Bax-deficient mice by subretinal injection of 1.4% sodium hyaluronate. At 1, 3, 7, and 28 days after injection, animals were killed, eyes enucleated, and retinal sections studied by histochemistry, immunofluorescence labeling, and confocal microscopy. Rods and cones were labeled, and apoptotic cells were identified with terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL). Photoreceptor cell apoptosis and cell loss were assessed quantitatively by counting both surviving and TUNEL-positive rods and cones.

results. TUNEL-positive cells were found within the outer nuclear layer (ONL) of the detached portions of the retina. They were detected in the detached retina on day 1, peaked on day 3, and dropped precipitously after day 7 after RD. Photoreceptor cell loss of both rods and cones followed a similar time course after RD. Moreover, deletion of the proapoptotic gene Bax in a knockout mouse model abolished the RD-associated photoreceptor cell degeneration.

conclusions. Apoptosis is a major mechanism leading to photoreceptor cell death after RD. Blockage of the activity of the proapoptotic molecule Bax in a knockout mouse model prevents photoreceptor cell apoptosis and cell loss. These data suggest that the Bax-mediated apoptotic signaling pathway plays a critical role in RD-associated photoreceptor cell death.

Retinal detachment—the separation of the neurosensory retina from the underlying retinal pigment epithelium (RPE)—is one of the most common causes of photoreceptor cell loss and resultant blindness in young adults. It can be caused by full-thickness retinal breaks, vitreoretinal traction, and/or subretinal exudation. ¹ Final visual prognosis after RD repair is associated with involvement of the macula and the duration of the RD. ² For example, patients with a macula-off RD have a 70% chance of returning to 20/40 or better vision if the RD is repaired within 10 days; however, that chance is reduced to 35% or 20% if the macular detachment lasts more than 10 days or 6 weeks, respectively. ² To date, there are no effective treatments for this RD-induced permanent loss of vision. 

Over the last decade, multiple lines of evidence have indicated that apoptosis is a major cause of photoreceptor degeneration and vision loss in experimental RD ^{3 4 5 6} and, similarly, is a central mechanism associated with neuronal cell loss after trauma, ischemia, and neurodegeneration in the central nervous system (CNS). ^{7 8} Apoptosis represents physiological or pathologic cell death that occurs when a cell activates its suicide program in response to internal or external signals. ⁹ A currently prevailing view is that, as a cell receives death-inducing stimuli, the mitochondrial outer membrane loses its integrity, resulting in the release of cytochrome c into the cytoplasm. ¹⁰ As a consequence, released cytochrome c binds to Apaf-1, which recruits caspase-9 to form a multimeric complex, leading to the initiation of the caspase cascade and, in turn, apoptosis. ¹⁰  

Mitochondrial permeability is maintained by various molecules that interact in a complex system to regulate the release of cytochrome c and other proapoptotic proteins, such as apoptosis-inducing factor (AIF) and nucleases. ¹¹ The Bcl-2 family member proteins are thought to be central regulators of this process. ¹² The Bcl-2 family is divided into two classes that exert opposing effects on cell death. Antiapoptotic members, such as Bcl-2 and Bcl-x_L, protect cells from apoptosis; proapoptotic members, such as Bax and Bak, trigger or sensitize cells for apoptosis. ¹³ When Bax and/or Bak are activated, they trigger mitochondrial outer membrane perforation by a mechanism that has yet to be identified. This leads to the release of cytochrome c and apoptotic regulatory proteins into the cytoplasm and resultant activation of the executioner caspases. In contrast, Bcl-2 and/or Bcl-x_L block this process and thus inhibit programmed cell death. Much evidence has indicated that the pro- and antiapoptotic members of Bcl-2 family can neutralize each other by heterodimerization or form homodimers, but it remains unclear which complex of these serves as the functional moiety in regulating apoptosis. ¹⁴  

Bcl-2-related gene products have been shown to be critically involved in numerous CNS diseases and degeneration ¹⁵ and thus may constitute prime targets for therapeutic interventions in RD. Mice lacking certain Bcl-2 family-member proteins have been generated. For example, animals deficient in Bcl-x_L die during embryogenesis because of massive cellular apoptosis in the CNS, liver, and hematopoietic system. ¹⁶ In contrast, mice without the proapoptotic member Bax are viable ¹⁷ and have more neurons in the peripheral and CNS than normal. ¹⁸ In culture, neurons from Bax-deficient mice are resistant to cell death induced by trophic factor deprivation as well as various other death-inducing stimuli. ¹⁹  

Involvement of Bax in developmental and pathologic retinal cell death has also been reported. ²⁰ A low level of Bax expression has been detected in photoreceptor cells of adult mice. ²¹ In the absence of Bax, there is a profound increase in the survival of retinal ganglion cells and a decrease of cell apoptosis in the inner nuclear layer (INL). However, deletion of Bax alone does not protect photoreceptor cells from naturally occurring cell death or degeneration induced by rd mutation. ²⁰ Recently, it has been shown that concurrent knockout of Bax and Bak is required for rescuing photoreceptor cells from developmental death. ²² Given that expression of Bax protein is induced by retinal ischemia ²³ —a situation that occurs in the outer retina after an RD—it is likely that Bax is involved in RD-associated photoreceptor cell loss. 

In the present study, we sought to characterize and delineate the time course of photoreceptor cell apoptosis and cell loss in a mouse model of RD. Experimental models of RD have been established in rat, cat, and other animals ^{1 24 25} ; however, little work in this area has been performed in the mouse. Recent advances in mouse genetic technology have made the mouse an appealing animal model for the study of eye diseases. The availability of genetically mutated mouse lines provides a unique opportunity for the study of molecular mechanisms underlying photoreceptor degeneration after RD and thus for identification of therapeutic targets for drug development. After establishment of an experimental model of RD in mice, in the present study, we report utilization of Bax-deficient (Bax^−/−) mice ¹⁷ to examine the involvement of Bcl-2 family proteins in RD-induced photoreceptor cell degeneration and dysfunction. 

Materials and Methods

Animals

Adult (8–12 weeks) C57Bl/6J wild-type and Bax-deficient ¹⁷ mice were used in this study. All experimental procedures and use of animals followed the protocol approved by the Animal Care and Use Committee of the Schepens Eye Research Institute and conformed to the standards set forth in the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Retinal Detachments

Mice were anesthetized by intraperitoneal injection of a mixture of ketamine (62.5 mg/kg) and xylazine (12.5 mg/kg; Phoenix Pharmaceutical, Inc., St. Joseph, MO). Left pupils were dilated with topical application of 1% cyclopentolate (Akorn, Inc., Buffalo Grove, IL) and 2.5% phenylephrine hydrochloride (Akorn, Inc.). A scleral puncture was made at the supernasal equator of the left eyeball with a glass micropipette, to lower intraocular pressure. A glass micropipette was then inserted into the subretinal space, and 1 to 2 μL of 1.4% sodium hyaluronate (Healon GV; Pharmacia & Upjohn, Uppsala, Sweden) was injected between the retina and the RPE. Sodium hyaluronate is a viscous material commonly used during intraocular surgery in humans and is not associated with any known ocular toxicity. ²⁶ Mice receiving scleral punctures without sodium hyaluronate injection served as the control. 

Retinal Histology, Immunofluorescence Staining, and TUNEL

Mice were killed on 1, 3, 7, and 28 days after surgery. Eyes of the control and experimental groups were collected and sectioned at 14 μm. Retinal sections were fixed with 4% paraformaldehyde for 1 hour before staining with cresyl violet or hematoxylin and eosin to reveal general retinal morphology and laminar structure. The retinas that were found to have reattached on the day of death were excluded from further analysis. To detect apoptotic cells, a TUNEL assay was performed. Retinal sections were permeabilized with a solution containing 0.1% Triton X-100 and 0.1% sodium citrate for 2 minutes at 4°C and then labeled with an apoptosis detection kit (In Situ Cell Death Detection Kit; Roche Molecular Biochemicals, Mannheim, Germany) according to the manufacturer’s guidelines. 

For immunofluorescence staining, retinal sections were preblocked with phosphate-buffered saline (PBS) containing 5% bovine serum albumin (Sigma-Aldrich, St. Louis, MO) and 0.2% Triton X (Sigma-Aldrich), followed by a reaction with a primary antibody against rod/rhodopsin (1:2500) or cone opsin (1:2000) ²⁷ overnight at 4°C. Retinal sections were then incubated with secondary antibody, 7-amino-4-methyl-coumarin-3-acetic acid (AMCA)–conjugated donkey anti-chicken (1:50) or Cy3-conjugated goat anti-mouse (1:200; both from Jackson ImmunoResearch Laboratories, Inc., West Grove, PA), for 2 hours at room temperature. For some retinal reactions, 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, 1 mg/mL in PBS; Sigma-Aldrich) stain was used to reveal the nuclear structure in retinal sections. 

Quantification of Photoreceptor Cell Loss and Apoptosis

To assess photoreceptor cell loss and apoptosis quantitatively, we used three methods: measurement of the outer nuclear layer (ONL) thickness, ²⁸ counts of TUNEL-positive cells or TUNEL-positive rods and cones, and assessment of the densities of surviving rods and cones. These measurements and cell counts were performed in the detached portion of the retina or the control retina that received sclera puncture. Photomicrographs of retinal sections were taken by microscope (model TE300; Nikon, Tokyo, Japan) equipped with epifluorescence illumination, a color digital camera (Spot; Diagnostic Instruments, Sterling Heights, MI), and a computer (Gateway, Sioux City, SD). Measurements of ONL thickness were made using photomicrographs taken from cresyl violet–stained retinal sections. Using NIH Image software (available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD), at least five separate measurements of ONL thickness were obtained from each retinal section. For the experimental groups, only measurements taken from the detached portions of the retinas were used, and the numbers were averaged. To quantify cell apoptosis, retinal sections reacted with TUNEL were photographed. TUNEL-positive cells were counted, and areas of the detached portions of the retinas or control retinal sections were measured using NIH Image. The number of TUNEL-positive cells was expressed as cells per unit area of the retinal section. To identify the types of photoreceptor cells undergoing apoptosis, retinal sections were counterlabeled with antibodies against rod/rhodopsin and cone opsin. ²⁷ Surviving rods and cones or TUNEL-positive rods and cones were counted, respectively, and the numbers were presented as cells per square millimeter of the retinal section measured. In each group, four to five mice were studied at each time point. 

Results

A Mouse Model of Experimental RD

To generate a mouse model of RD, we injected 1.4% sodium hyaluronate (Healon GV; Pharmacia & Upjohn) into the subretinal space of adult mice. Successful RD was confirmed ophthalmoscopically and generally involved one third of the retina. At 1, 3, 7, and 28 days after RD, mice were killed and the eyes sectioned and processed for staining with cresyl violet or hematoxylin and eosin to reveal general retinal morphology (Figs. 1A 1B) . In most cases, retinal sections from mice that received subretinal injection of sodium hyaluronate revealed detachment of the retina from the underlying RPE layer (Fig. 1B) . The laminar structure of the detached retina appeared normal. Starting at day 1, many photoreceptor cells in the detached retina revealed a collapsed inner segment (Figs. 1E 1F . Pyknotic nuclei, indicating apoptotic cell death, were noticeable in the ONL (not shown), and visible thinning of the ONL was observed as the duration of RD increased. In some mice, reattachment of the retinas occurred spontaneously. In general, subretinal injection of 1.4% sodium hyaluronate successfully created an RD that would last for 28 days. 

Time Courses of Photoreceptor Cell Apoptosis and Degeneration

Photoreceptor cell apoptosis has been observed in experimental RD created in animal models such as rat, ^{5 6} rabbit, ²⁹ cat, ³ and ground squirrel. ²⁵ To understand whether RD induces similar photoreceptor cell apoptosis and degeneration in mice, we performed TUNEL assays on the mouse retinal sections. In control retinas, which received only a scleral puncture without subretinal injection, no TUNEL-positive cells were observed (Fig. 1C) . In contrast, numerous TUNEL-positive cells were seen in the detached retinas of the experimental group, which received a sodium hyaluronate injection (Fig. 1D) , beginning at 1 day after creation of the RD. Most TUNEL-positive cell nuclei were present in the ONL of the detached portion of the retina (Fig. 2) . Some were seen across the outer limiting membrane into the outer segment layer and the subretinal space, implying active phagocytosis of apoptotic cells by macrophages. ³⁰ No TUNEL-positive cells were detected in the adjacent attached retina. 

To assess quantitatively and to delineate the time course of photoreceptor cell apoptosis, we counted TUNEL-positive cells in retinal sections on days 1, 3, 7, and 28 after RD (Fig. 3A) . We found that apoptotic cells appeared on day 1 after RD and peaked on day 3. At day 7, the number of apoptotic cells decreased significantly and remained unchanged until day 28. Thus, RD-induced photoreceptor cell apoptosis was most prominent during the first week after injury. 

To determine whether apoptosis contributes primarily to photoreceptor cell loss associated with RD, we quantified the survival of photoreceptor cells at various time points after RD by measuring ONL thickness (Fig. 3B) . On day 1, ONL thickness of the detached retina declined to 57.5 ± 4.8 μm compared with 83.3 ± 3.3 μm in the control retina, representing a 30% loss of cells in the ONL. On day 3, ONL thickness of the detached retinas decreased further to 47.5 ± 5.1 μm or a 43% total cell loss. By day 7, ONL thickness decreased to 42.5 ± 2.5 μm and remained unchanged through day 28 (Fig. 3B) . The data indicate a close correlation between the peak of photoreceptor cell apoptosis and cell loss during the first 3 days of RD and suggest that apoptosis is a key phenomenon leading to photoreceptor cell degeneration associated with RD. 

Specific Responses of Rods and Cones after RD

There are two types of photoreceptor cells in the mouse retina: rods, which make up approximately 97.2% of the photoreceptor cell population, and cones, which form the remaining 2.8%. ³¹ To determine whether these two types of photoreceptor cells respond similarly to RD damage, we labeled retinal sections with TUNEL and primary antibodies against rod/rhodopsin and cone opsin (Fig. 4) . Staining of rods revealed robust labeling in the entire ONL of the retina, whereas cone-opsin–positive cells were present only at the outer layer of the ONL lining the outer limiting membrane (Figs. 2A 2B) . After counting TUNEL-positive rods and cones in retinal sections (Fig. 4) , we noted that these two types of photoreceptor cells had a similar pattern of apoptosis after RD (Figs. 5A 5B) and that this pattern paralleled the rate of total cell apoptosis in the detached retina. Apoptotic rods and cones became detectable on day 1, peaked on day 3, and declined thereafter. On day 28, very few apoptotic rods or cones were observed. Consistent with the previously reported ratio of rods to cones in the mouse retina, 2.3% ± 0.8% of the apoptotic cells were cone-opsin–positive and 90.3% ± 3.9% were rod-positive. Of the TUNEL-positive cells, 7.6% ± 3.8% could not be identified. Rods and cones, therefore, appear to exhibit a similar susceptibility to cell apoptosis after RD. 

To confirm that rods and cones followed a similar time course of cell loss, we counted surviving rods and cones in retinal sections from mice at various days after the induction of RD (Figs. 5C 5D) . Again, after correlating the dramatic increase in rod and cone apoptosis, we noted that the number of surviving rods and cones decreased significantly from day 1 to day 3 after RD. By day 3, there was a 36.5% ± 7.9% loss of rods and a 46.3% ± 6.6% loss of cones. On day 7, the total number of surviving rods (6892.0 ± 613.3 cells/mm²) and cones (283.9 ± 54.1 cells/mm²) in the detached retinas represent a 39.2% ± 5.4% and 63.0% ± 7.1% accumulated loss of rods or cones, respectively. The percentage of rod-positive cells surviving after RD was significantly higher than that of cone-opsin–positive cells (P < 0.05). From days 7 to 28, no further loss of rods and cones was detected. Thus, although RD induced a similar pattern of cell apoptosis in rods and cones, it caused a more profound loss of cone-opsin–positive cells than rod/rhodopsin-positive cells in the detached retina, which may suggest different functional changes of rods and cones after the RD. 

Blockage of Photoreceptor Cell Degeneration in Bax-Deficient Mice

The Bcl-2 family member protein Bax is thought to be a central regulator of ischemia- and trauma-associated neuronal apoptosis. The expression of Bax has also been found to be upregulated in the retina after ischemic injury. ²³ Taking advantage of mouse genetic technology and the availability of Bax^−/− mice, ¹⁷ we sought to determine whether the Bax-mediated pathway also plays a role in RD-associated cell death. We compared photoreceptor cell apoptosis and cell loss in wild-type and Bax^−/− mice within the first week of RD, when photoreceptor cell death peaked in the retinas of wild-type mice. In contrast to the results observed in wild-type mice, we detected no TUNEL-positive cells in the detached retinas of Bax^−/− mice (Fig. 6A) . Moreover, ONL thickness in retinal sections of the detached retinas of Bax^−/− mice remained unchanged in all ages examined (73.3 ± 0.3 μm measured on day 1 and 65.0 ± 0.5 μm on day 3 in the RD retinas and 66.7 ± 0.9 μm in the control retinas). In contrast, ONL thickness in the wild-type mice in the RD group declined significantly between days 1 and 3 in comparison with the control retinas. In addition, we stained retinal sections with hematoxylin and eosin to reveal more clearly the morphologies of the outer and inner segments of photoreceptor cells. We noted that by day 28 after RD, inner and outer segment remnants were rarely seen on surviving photoreceptor cells of wild-type mice but were intact in the detached retinas of Bax^−/− mice (not shown). Thus, Bax deficiency not only protects the retina against RD-induced cell death, it also prevents the degeneration of the inner and outer segments of photoreceptor cells. The lack of photoreceptor cell death was also confirmed by counting the numbers of surviving rods and cones in retinal sections of Bax^−/− mice before and after RD (Figs. 6B 6C) . We therefore conclude that Bax is essential for RD-associated photoreceptor cell apoptosis and degeneration. 

Discussion

Experimental models of RD have been well established in animals such as cats, rats, and ground squirrels, ^{1 3 25} but the study of RD in mice is just beginning. ³² The unique anatomy of the mouse eye—for example, its smaller vitreous cavity and bigger lens—presents special surgical challenges and results in a higher rate of spontaneous retinal reattachment. In this study, we were able to overcome these anatomic limitations and produce RD that lasted up to 28 days. This is the first systematic study of the time course and pattern of photoreceptor cell apoptosis and cell loss associated with RD in mice. 

Consistent with reports regarding other animal and human studies, ^{1 4 5 33} in the current study RD induced apoptosis that was limited to the ONL of the detached portion of the retina, an area that becomes ischemic because of the separation of the retina from the underlying choroidal blood supply. There is recent evidence that oxygen supplementation during RD reduced photoreceptor cell death in a cat model of RD, which supports the notion that RD-associated retinal degeneration is related to oxygen deficiency. ^{24 34}  

In our study, most photoreceptor cell death associated with RD took place from days 1 to 3 after detachment and declined rapidly over the next few days. For example, by 28 days the detached mouse retina had lost approximately 50% of its photoreceptor cells, and 86% of this loss had occurred by day 3. Furthermore, rod and cone photoreceptor cells followed a similar time course of apoptosis and cell loss. Our results are similar to those reported by others in humans and ground squirrels, showing that apoptosis of photoreceptor cells can appear as early as 8 hours and diminishes rapidly after 1 week of RD. ^{25 33} These data suggest that, to minimize RD-induced permanent damage to the retina, repair of the RD should be performed as soon as possible. This may be particularly important for restoring vision in patients with macula-involved RD. In addition, consideration should be given to clinical procedures that induce only a transient RD—for example, as a result of retinal or RPE cell transplantation ^{35 36} or injection of substances into the subretinal space. ³⁷ With recent progress in stem cell and gene therapy research, these procedures may emerge as promising therapies for retinal diseases in the near future. However, it should be kept in mind that, although the procedure results in a detachment of the retina that remains for only a few days, significant and irreversible damage to the retina may still occur. 

We notice that while the most dramatic loss of ONL nuclei occurs on day 1 in the detached retina, detection of TUNEL-positive, apoptotic cells peaks on day 3. Previous studies have shown that detection of TUNEL-positive cells in situ reflects not only the rate of cell apoptosis in the retina, but also the rate of the disposal of apoptotic nuclei. It has been suggested that apoptotic cells are removed by phagocytosis by macrophages as well as by their neighboring cells, including viable rods, in the retina. ³² Thus, it is possible that on day 1, TUNEL-positive cells disappear rapidly in the detached retina because there are more viable cells around. As more rods are lost by day 3, residual phagocytotic capacity is reduced, leading to a transient prolongation of the lifetime of apoptotic nuclei on day 3 and an increase in the number of apoptotic photoreceptor cells in the detached retina. 

An important finding of this study is that, after RD, the percentages of apoptotic rods (90.3%) and cones (2.3%) detected remained similar to the percentages of those photoreceptors in the control retina. No significant differences between the rates or patterns of rod and cone cell apoptosis were observed. In contrast, we noticed a more profound loss of cone-opsin–labeled cells (63%) than of rods (39%) after RD. The data suggest that, although RD results in similar apoptosis of rods and cones, surviving cones may also lose certain aspects of their functional properties to escape apoptosis. It has been reported that rods and cones behave differently in response to RD. Most surviving rods continue to be positive for the proteins they are expressing, whereas cones rapidly lose the expression of several proteins, including cone opsin and calbindin D. ^{25 38} One interpretation is that, because cones have a normal metabolic rate estimated to be 15 times that of rods, to survive the hypoxia resulting from RD, these cells downregulate the production of many proteins that are not critical for survival and enter a state that conserves metabolic energy. ³⁸ Our results support the notion that RD induces a similar rate of cell death in rods and cones, but it may result in different functional changes in these two cell types. This finding may be important for the understanding of the recovery process of vision after RD and for the development of corresponding treatment strategies for RD. 

Our mouse model of RD was created by injecting 1.4% sodium hyaluronic acid into the subretinal space. In human RD, subretinal fluid normally contains a lower concentration of hyaluronic acid (<1%) than that used in the mouse model. It is unclear how this difference may affect oxygen or nutritional diffusion from the choroids to photoreceptor cells and therefore the outcome of visual function recovery after RD in humans. It would be interesting to study whether human photoreceptor cells follow a similar time course of apoptosis and cell loss after RD, and whether the retinas of mice that receive a subretinal injection of a lower concentration of hyaluronic acid would exhibit a similar pattern of photoreceptor degeneration. 

Finally, in this study, we provide evidence suggesting that the proapoptotic protein Bax plays an essential role in RD-associated photoreceptor cell apoptosis. Involvement of caspase activation and mitochondrial-dependent cell death pathways in RD-associated photoreceptor cell apoptosis and dysfunction have already been reported. ^{5 6} Bax, as a proapoptotic member of the Bcl-2 family, acts as a central player regulating cytochrome c release from the mitochondria and in turn, causing subsequent caspase activation. ¹² Activation of caspase cascades and Bax is known to play an important role in execution of neuronal cell death after CNS trauma and ischemia—conditions that are associated with RD. Although under normal oxygen conditions, expression of Bax is low in the adult retina, its expression is induced by retinal ischemia, ³⁹ suggesting that Bax may be involved in photoreceptor cell degeneration under conditions such as RD. Much recent evidence has indicated that overexpression of Bax leads to the death of rod photoreceptor cells. ²¹ Translocation of Bax to the rod mitochondria has been found to associate with rod apoptosis in the lead-exposed mouse model. ⁴⁰ In the present study, photoreceptor cell loss associated with RD was abolished in the absence of Bax. However, Bax deficiency is insufficient to rescue the developmental loss of rods; rather, deletions of both Bax and Bak are required. ²² These results suggest that developmental death of photoreceptor cells is mediated by multiple, parallel pathways that involve both Bax and Bak; whereas, Bax alone plays a more important role in the pathologic death of photoreceptor cells. 

However, we should not rule out the possibility that deletion of Bax may alter the microenvironment of the retina, which in turn contributes to the protective effect of Bax deficiency on RD damage. The absence of Bax, although it alone does not rescue the photoreceptor cells from developmental death, enhances cell survival in the INL. ²⁰ It remains unclear how this change may alter the functional property of retinal glial cells or supplies of various neurotrophic factors in the Bax^−/− retina and influence the survival of photoreceptor cells after RD. Alternatively, deletion of Bax may result in changes in the activities of parallel signaling pathways that regulate apoptosis, such as Bak, p53, and DHA. Subsequently, these changes could affect photoreceptor cell sensitivity to RD-induced apoptosis signals and the survival of photoreceptor cells after RD in Bax-knockout mice. In any case, our results indicate that apoptosis is an essential component in RD-associated photoreceptor degeneration and that deletion of Bax protects the photoreceptor cells from loss. These data suggest that Bax may be a potential target for pharmacologic therapy to treat retinal damage after RD. 

 Supported by grants from the National Eye Institute (Grant EY012983), the Massachusetts Lion’s Eye Research Fund, the Department of the Army, and the Joint Clinical Research Center at the Schepens Eye Research Institute and the Massachusetts Eye and Ear Infirmary of Harvard Medical School.

 Submitted for publication August 4, 2003; revised October 21 and November 4, 2003; accepted November 12, 2003.

 Disclosure: L. Yang, None; D. Bula, None; J.G. Arroyo, None; D.F. Chen, None

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

 

Figure 1.

 

Photomicrographs of retinal sections taken from control (A, C, E) or RD (B, D, F) mice. The sections were stained with either cresyl violet (A, B), TUNEL (C, D), or hematoxylin and eosin (E, F). Arrowheads: TUNEL-positive cells; ( amp;) subretinal space under the detached portion of the retina. GCL, ganglion cell layer. Scale bar: (A, B) 50 μm; (C–F) 25 μm.

Figure 2.

 

Epifluorescence photomicrographs of retinal sections from an RD mouse labeled with primary antibodies against rod/rhodopsin (A, F), cone opsin (B), or secondary antibodies only as negative controls (C, D) and TUNEL (E, F). Arrowheads: TUNEL-positive cells. Note that most cones localized at the outer layer of the ONL, lining up along the inner limiting membrane. All TUNEL-positive cells were found in the ONL, which can be recognized by its robust fluorescence labeling by anti-rod/rhodopsin. Scale bar: 50 μm.

Figure 3.

 

Quantification of photoreceptor cell apoptosis and cell loss at various days after the induction of RD. (A) Counts of TUNEL-positive cells in retinal sections of mice at various days after RD. The numbers are expressed as TUNEL-positive cells per unit area of the retinal section of the detached portion of the retina. (B) Measurements of ONL thickness in cresyl violet–stained retinal sections from mice at various days after RD. Data are expressed as the mean ± SEM.

Figure 4.

 

Confocal images with orthogonal projections of TUNEL-positive rods (A–C) and cones (D–F) found in the detached mouse retina. Retinal sections are double labeled with TUNEL and primary antibodies against either rod rhodopsin or cone opsin. Scale bar: 4 μm.

Figure 5.

 

Patterns of rod and cone photoreceptor cell apoptosis and cell loss after RD. Quantification of apoptotic rods (A) and cones (B) in retinal sections of mice taken at various time points after the induction of RD. Retinal sections were double labeled with TUNEL and primary antibody against rods or cones. Quantification of surviving rods (C) and cones (D) in retinal sections of the detached portions of the retinas taken from mice killed at various time points after RD and the surviving rods and cones counted. Data are expressed as the mean ± SEM.

Figure 6.

 

Comparison of RD-associated photoreceptor cell apoptosis and cell loss in retinal sections of wild-type and Bax-deficient mice. (A) Quantification of TUNEL-positive cells in the detached portions of the mouse retinas. (B) Measurements of ONL thickness in the detached and control retinas at various days after RD. (C) Counts of survival cones in the control and detached retinas at various days after RD. Data are expressed as the mean ± SEM *P < 0.05; **P < 0.01; ***P < 0.005 by two-tailed t-test.

The authors thank Tiansen Li from the Massachusetts Eye and Ear Infirmary, Harvard Medical School, for providing rod/rhodopsin and cone opsin antibodies.