Bax Is Not the Heterodimerization Partner Necessary for Sustained Anti–photoreceptor-Cell-Death Activity of Bcl-2

Abstract

purpose. Ectopic expression of Bcl-2 in photoreceptors of certain mouse models of retinitis pigmentosa (RP) temporarily slows disease progression. The temporary effect produced by Bcl-2 may result from insufficient levels of functional complexes between Bcl-2 and additional proteins necessary for maintaining the anti-apoptotic activity of Bcl-2. Although the overexpression of Bax generally induces apoptosis, Bax exerts anti-apoptotic properties when complexed with Bcl-2 in certain cell culture systems. These studies were designed to determine whether coexpression of Bcl-2 and Bax would improve the neuroprotective effect provided by Bcl-2 alone in photoreceptors of mice with autosomal dominant RP (adRP).

methods. Transgenic mice were produced that overexpressed Bax and Bcl-2 specifically in photoreceptor cells, using the murine opsin promoter to drive transgene expression. These mice were crossed with an adRP mouse model to assess the effect of coexpression of Bax and Bcl-2 on retinal degeneration. Morphologic analysis was performed on retinas isolated at various developmental times to monitor disease progression.

results. Ectopic expression of Bax in photoreceptors resulted in extensive rod cell death dependent on the level of Bax transgene expression. Although Bcl-2 was able to inhibit Bax-induced photoreceptor cell death, the coexpression of Bcl-2 and Bax in photoreceptors of mice with adRP did not enhance the protective effect against photoreceptor cell death exerted by Bcl-2 alone.

conclusions. Coexpression of Bax and Bcl-2, at the levels produced in the transgenic lines, does not extend the temporary neuroprotective effect produced by Bcl-2 in photoreceptors of mice with adRP.

Recent evidence indicates that genetic mutations within several components of the visual transduction pathway may be involved in the pathogenesis of a variety of retinopathies. ^{1 2} Mutations within the genes encoding rhodopsin, the α and β subunits of rod cyclic guanosine monophosphate (cGMP) phosphodiesterase, and the α subunit of cGMP-gated cation channel have been linked to various forms of retinitis pigmentosa (RP). ^{3 4 5 6} In addition, mutations within genes encoding structural components of the photoreceptor cell, such as peripherin and rom-1, lead to degeneration of the retina. ^{7 8 9} Retinal degeneration may also result from cellular insults such as exposure to constant light. ^{10 11 12} Regardless of the nature of the primary genetic defect or the triggering damaging event, retinal degeneration generally occurs by a process that has the characteristics of programmed cell death. ^{13 14 15}  

Programmed cell death, or apoptosis, is an inherent physiological process that normally controls and stabilizes cell populations in multicellular organisms. ¹⁶ Programmed cell death occurs during normal embryonic development, yet can be triggered in response to certain pathologic states in an adult organism. For example, physiologic programmed cell death occurs in the photoreceptors during the development of the neural retina, ¹⁷ whereas pathologic cell death occurs in photoreceptor cells in response to various genetic mutations ¹ and to excessive light exposure. ^{10 11 12} Therefore, photoreceptor apoptosis results from the activation of an intrinsic mechanism by either internal or external signals. The molecular event, or events, that activates the programmed cell death pathway in photoreceptor cells remains to be determined. 

Various members of the Bcl-2 family of proteins can regulate the programmed cell death pathway. ^{18 19} The Bcl-2 gene product, for instance, has been shown to protect several cell types, including neurons, from cell death induced by a variety of stimuli, ^{20 21} and increased expression of Bax has most often been implicated as a cause of cell death. ^{22 23} However, Bax has been shown to exert either pro- or anti-apoptotic activity, depending on the cellular background in which it is expressed. For example, the overexpression of Bax in embryonic neurons has been shown to promote cell survival in response to withdrawal of nerve growth factor in vitro. ²⁴ In addition, Bax-deficient mice show tissue hyperplasia or hypoplasia, depending on the cellular context. ²⁵ Although many of the Bcl-2 family members, such as Bcl-2 and Bax, combine to form both homodimers and heterodimers, it is questionable which complex serves as the functional moiety in regulating apoptosis. ^{22 26 27 28 29 30} Because mutations that affect the ability of Bcl-2 to heterodimerize with Bax abrogate its ability to counter apoptosis, ³¹ the possibility remains that the Bcl-2/Bax heterodimer could serve as an anti-apoptotic functional complex. Therefore, the relative levels of these various proteins within a cell, as well as the cellular context, determine the susceptibility of a cell to a given apoptotic stimulus. 

Ectopic expression of Bcl-2 in several mouse models of RP has been shown to delay progression of the degenerative process temporarily without correcting the underlying defect. ^{32 33 34} It is not yet known how the Bcl-2 gene product performs such a protective function in photoreceptor cells or why the effect is transient. The temporary effect produced by Bcl-2 may result from insufficient levels of the proteins necessary for persistent and complete anti-apoptotic activity of Bcl-2, such as BAG-1 or Bax, which are thought to form functional complexes with Bcl-2. ^{22 35} For example, it has recently been shown that the protective effect provided by Bcl-2 was markedly improved after coexpression with BAG-1 in photoreceptor cells of mice expressing the S334ter mutant form of rhodopsin. ³⁶ In addition, coexpression of Bax and Bcl-2, such that either Bax or Bcl-2 were present in excess, has been found to promote the survival of neurons deprived of neurotrophic factors. ²⁴ The present study was designed to determine whether the basis for the temporary protective effect produced by Bcl-2 is related to insufficient levels of Bax in the retina and whether the coexpression of Bax with Bcl-2 would improve the neuroprotective effect provided by Bcl-2 alone against retinal degeneration. 

Methods

Construction of Transgene Expression Vector

The Bax expression vector used to create transgenic mice was constructed by inserting the murine Bax coding sequence between the 4.4-kb fragment of the rod opsin 5′ flanking sequence and the mouse protamine 1 polyadenylation sequence and cloning into the multiple cloning site of the vector (pBluescript II KS(+); Stratagene, La Jolla, CA). The fusion gene was released from the vector by digesting with BssHII, gel purified, and used for microinjection of fertilized eggs. Similar expression vectors have been used to direct expression of Bcl-2 and BAG-1 specifically in photoreceptor cells. ^{32 36}  

Generation of Transgenic Mice

All protocols involving the use of mice adhered to the regulations set forth in the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Transgenic mice were generated and screened as previously described. ³⁶ Four Bax transgenic founder lines (designated rhBax) were identified and mated to wild-type animals to expand the colonies for future analysis. 

Genotype Analysis

PCR was performed on genomic DNA samples prepared from tail biopsy samples to screen for the presence of the Bax transgene as follows: initial denaturation at 94°C for 5 minutes, followed by 35 cycles of 94°C for 1 minute, 63°C for 1.5 minutes, and 72°C for 2 minutes. DNA primers used to amplify the Bax transgenic sequence were Bax OP1 3′ seq 5′-TGG TGA GCG AGG CGG TGA GGA C-3′ and Rh1.1 5′-GTG CCT GGA GTT GCG CTG TGG G-3′. Genotype analyses for the Bcl-2 and S334ter rhodopsin mutant transgenic animals were performed as previously described. ³⁶  

Western Blot Analysis

Protein immunoblot analysis was performed using retinal extract preparations to assess expression levels of the transgenes, as previously described with minor modifications. ³⁶ Extracts were prepared by homogenizing a single retina in 100 μL RIPA buffer containing 1% Nonidet P-40 (NP40), 0.5% sodium deoxycholate, and 0.1% SDS in phosphate-buffered saline (PBS) containing 1 tablet of a protease inhibitor cocktail (Complete Mini Protease Inhibitor Cocktail set; Roche Molecular Biochemicals, Mannheim, Germany) per 10 mL RIPA lysis buffer. Approximately equivalent amounts of protein for each sample were subjected to SDS-PAGE using 15% Tris-HCl acrylamide gels. Nylon membranes, after protein transfer, were probed with a Bax rabbit polyclonal primary antibody (product no. 06-499; Upstate Biotechnology, Lake Placid, NY). Filters were then stripped and reprobed with a primary antibody raised against recombinant nonacylated recoverin (1:5000 dilution; kindly provided by James Hurley, PhD, Department of Biochemistry, University of Washington HHMI, Seattle, WA, and Ching-Kang Chen, PhD, Department of Ophthalmology, University of Utah, Salt Lake City, UT). ³⁷ Purified m-Bax-his protein and His6-huBcl-2 purified proteins were used as controls (generously provided by John C. Reed, MD, PhD, The Burnham Institute, La Jolla, CA). 

Retinal Tissue Processing for Light Microscopy

Retinal tissues for morphologic analysis were processed as previously described, with minor modifications. ³⁸  

Coimmunoprecipitation Assay

Extracts were prepared by homogenizing a single retina in 175 μL lysis buffer (0.2% NP40 in PBS containing 1 tablet of the protease inhibitor cocktail per 10 mL). Aliquots, containing approximately equivalent amounts of extracted protein, were incubated with either 1 μg of Bax antibody or a rabbit polyclonal cytochrome c antibody (product no. SC7159; Santa Cruz Biotechnology, Santa Cruz, CA) as a nonspecific control. Protein A Sepharose (Amersham Pharmacia Biotech, Piscataway, NJ) was used to precipitate the immune complexes. Samples were centrifuged and the supernatant recovered. Equal volumes of pellets were subjected to SDS-PAGE on 12% polyacrylamide gels. Aliquots of the supernatant were run to assess efficiency of the precipitation. Antibodies used in the immunoprecipitation (1 μg) and either m-Bax-his– or His6-huBcl-2–purified proteins (10 ng) were run as the control. Proteins were transferred to nylon membranes and processed for immunodetection with a Bcl-2 antibody (1:500 dilution; product no. SC783; Santa Cruz Biotechnology). Filters were stripped after initial antibody detection and then reprobed with either the Bax polyclonal antibody or the cytochrome c antibody at 1 μg/mL final concentration. 

Morphometric Analysis

Morphometric analysis was performed by counting photoreceptor cells in 200-μm spans from four different quadrants of either wild-type or transgenic retinas oriented from the superior to the inferior hemisphere of the eye. Counts were taken from at least two representative animals for each genotype at each time point. 

Results

Downregulation of Endogenous Bax Expression during Retinal Development

The temporary nature of the neuroprotective effect produced by Bcl-2 in photoreceptors of mouse RP models ³² could result from insufficient levels of proteins, such as Bax, that may be needed to maintain the anti–cell-death function of Bcl-2. ^{22 31} The level of endogenous Bax expression was analyzed to determine whether decreasing levels of the Bax protein would correlate with the decreasing anti-apoptotic activity of Bcl-2 in photoreceptor cells. Establishing the endogenous level of Bax during development was also necessary, to determine the contribution of transgene expression to the overall level of Bax expression in the RhBax transgenic retinas (described later). Protein immunoblot analysis of retinal extracts obtained from wild-type animals demonstrated that endogenous expression of Bax was downregulated during early development (Fig. 1 , compare lanes 1 through 8). A significant decrease in expression of Bax occurred between postnatal day (P)16 and P24, with the low level of expression persisting through adulthood (Fig. 1 , compare lanes 5 through 8). A known amount of purified recombinant m-Bax-his protein was run as a control to estimate size and quantity of Bax protein (Fig. 1 , lane 9). The larger band in some of the retinal samples may represent Bax homodimer formation. ²² The expression level of recoverin was used to assess the relative amounts of photoreceptor cells present in each retinal extract, because expression of recoverin is restricted to photoreceptor cells and some cone bipolar cells. ^{39 40} Immunolocalization of the Bax protein in wild-type retinas at P16 demonstrated faint immunoreactivity in the ganglion cell layer, the inner nuclear layer, and the rod inner segment region. The expression pattern of Bax in the retina is consistent with histogenic cell death occurring in these cell types during this developmental period (data not shown). The observed decrease in expression of Bax may therefore result from the downregulation of Bax in photoreceptor cells as well as in other cell types within the retina. Regardless, the progressively decreasing level of Bax expression with age correlates with the decline in the anti–photoreceptor-cell-death activity of Bcl-2 against adRP, and it is possible that overexpression of Bax in adult photoreceptor cells extends the neuroprotective effect produced by Bcl-2 in mice with RP. 

Generation of Transgenic Mice Expressing Bax Specifically in Photoreceptor Cells

Transgenic mice that ectopically express Bax specifically in photoreceptor cells were generated to cross with Bcl-2 transgenic animals to determine whether expression of Bax would prolong the temporary neuroprotective effect produced by Bcl-2 in the rhodopsin S334ter adRP mouse model. ³² The Bax expression vector was constructed by fusing the 4.4-kb 5′ regulatory region of the murine opsin gene to the Bax cDNA sequence with the polyadenylation sequence of the mouse protamine 1 gene placed immediately 3′ to the Bax sequence. The rod opsin promoter has been shown to direct gene expression specifically to the photoreceptors in an appropriate spatiotemporal manner during retinal development, with higher expression levels in the superior hemisphere of the eye and lower levels in the inferior hemisphere. ^{38 41 42} Four RhBax transgenic founder lines were identified and analyzed for the expression of the Bax transgene. 

Expression of the Bax transgene was assessed by protein immunoblot analysis using retinal extract preparations (Fig. 2) . Because a murine Bax sequence was used to create the transgenic construct, the Bax protein produced from the transgene could not be distinguished from the endogenous Bax protein. Therefore, expression of the transgene was assessed by comparing the overall level of Bax expression in transgenic retinas with the level in nontransgenic retinas. The Bax transgene was expressed at discernible levels in two of the four founder lines. Expression levels of Bax in both RhBax-C and RhBax-F transgenic retinas were greater than the endogenous level of expression in a wild-type littermate at P11, demonstrating expression of the Bax transgene (Fig. 2 , compare lane 1 with lanes 2 and 3). The lower level of recoverin in the RhBax-F retinal sample indicates that fewer photoreceptor cells were present in the RhBax-F retina (Fig. 2 , compare lanes 1 and 3). Known amounts of purified recombinant m-Bax-his protein were run as a control to estimate approximate size and quantity of Bax protein produced in the transgenic retinas (Fig. 2 , lanes 4 through 6). The total expressed Bax protein was estimated to be 10 to 50 ng in the RhBax-C transgenic retina at P11. A higher level of Bax expression would be expected in the RhBax-F transgenic retina, because photoreceptor degeneration appears to be induced in this line as early as P11 (also see Fig. 3 ). 

Analysis of Morphologic Ultrastructure of Transgenic Retinas for Effects of Bax Overexpression

Overexpression of Bax and the formation of Bax homodimers have been shown to promote cell death in certain cell types. ²² Results from the protein immunoblot analysis indicate that Bax overexpression in photoreceptor cells may lead to photoreceptor cell death. Therefore, before coexpressing the Bax and Bcl-2 transgenes in an adRP mouse model and assessing the combined transgene effect on the process of retinal degeneration, it was important to determine the effect of overproduction of Bax alone on photoreceptor cell morphology. The retinal morphology of Bax transgenic mice was assessed at various times after birth, using light microscopy. Morphologic analysis of RhBax-C and RhBax-F transgenic retinas at P11 and at 6 weeks of age confirmed that these two lines expressed the Bax transgene at different levels, as judged by the different degrees of photoreceptor cell loss in these retinas compared with a nontransgenic retina (Figs. 3 4) . Expression of the Bax transgene in both lines created a gradient of photoreceptor cell death that progressively decreased from the superior to the inferior regions of the eye. This phenomenon is most likely the result of production of a Bax expression gradient by the opsin promoter, which in turn, decreases the expression levels of the Bax transgene from the superior to the inferior hemisphere of the transgenic retina. ³⁸ Therefore, the extent of degeneration appears to depend on the expression level of the Bax transgene, as illustrated by the gradients of cell death induced by Bax, as well as by the different degrees of Bax-induced cell death in the two transgenic lines. These results are consistent with other reports showing that Bax homodimers induce programmed cell death. ^{22 43}  

Coexpression of Bax and Bcl-2 in Photoreceptor Cells

Although expression of the Bax transgene induces photoreceptor cell death, it is possible that Bcl-2 coexpression would sufficiently rescue the photoreceptors and allow a subsequent assessment of the combined effect of Bcl-2 and Bax on photoreceptor cell death in the rhodopsin S334ter mutant adRP mouse model. Therefore, the RhBax transgenic lines were crossed with the Bcl-2B transgenic line, previously generated using the same opsin promoter, to create transgenic animals that coexpress both Bax and Bcl-2 in photoreceptor cells. ³² Morphology of the retinas expressing both Bax and Bcl-2 was assessed to determine whether Bcl-2 expression prevents photoreceptor cell death induced by Bax expression. When the RhBax-C line was crossed with Bcl-2B (RhBax-C/Bcl-2B), Bcl-2 expression suppressed Bax-induced cell death in a gradient manner (Fig. 5) . The superior region with presumably higher levels of Bax was severely degenerated, whereas the inferior region with the lower levels of Bax had degenerated less and had retained the photoreceptor outer segments. This rescue persisted over time, with two to three layers of photoreceptor nuclei showing residual outer segments still present in the inferior region of the retina at 10 weeks of age (data not shown). Thus, the ability of Bcl-2 to inhibit Bax-induced photoreceptor cell death was significantly greater in the inferior region of the retina, where less Bax was expressed, which emphasizes the importance of the ratio of Bax to Bcl-2 in the regulation of cell death. 

Interaction of Bax and Bcl-2 In Vivo

Before introducing the two transgenes into the adRP mouse model and assessing the effect of coexpressing Bax and Bcl-2 on photoreceptor cell death, immunoprecipitation experiments were performed to determine whether the ability of Bcl-2 to inhibit Bax-induced cell death was related to complex formation between these two proteins. Immunoprecipitation using a Bax antibody resulted in coprecipitation of Bcl-2 with Bax protein in transgenic retinas expressing Bcl-2, either alone or together with Bax, but not in retinas expressing only the transgenic Bax protein (Fig. 6 , left: compare lanes 1 through 3). Coprecipitation of Bcl-2 with Bax in the Bcl-2B transgenic sample is most likely a result of complex formation between the low level of endogenous Bax protein and the Bcl-2 transgene product (lane 2). The large amounts of Bcl-2 protein present in the immunodepleted lysates is consistent with a higher level of expression of Bcl-2 transgene than of Bax transgene, leading to excess Bcl-2 protein not complexed with Bax (lanes 5 and 6). Approximately 400 to 800 ng of Bcl-2 was expressed, whereas only 20 to 100 ng of Bax is estimated to be expressed per adult transgenic retina. ^{36 44} The failure of a cytochrome c antibody to precipitate Bcl-2 confirms the specificity of complex formation between Bax and Bcl-2 (Fig. 6 , right: lanes 1 through 3). Known amounts of purified recombinant m-Bax-his and His6-huBcl-2 protein were run as the control (lanes 7 and 8). Nonspecific reactivity between the Bcl-2 antibody with either the Bax or the cytochrome c antibody used in the immunoprecipitation step is also shown (lane 9). Therefore, immunoprecipitation experiments showed that Bcl-2 and Bax formed a complex when coexpressed in photoreceptor cells, suggesting that the ability of Bcl-2 to inhibit Bax-induced photoreceptor cell death may be caused by direct interaction between these two proteins. 

Effect of Bcl-2 and Bax Coexpression on the Ability of Bcl-2 to Prevent Photoreceptor Cell Death in an adRP Mouse Model

To test the hypothesis that Bcl-2/Bax is the functional complex that inhibits apoptosis, the RhBax/Bcl-2B transgenic line was then crossed with a transgenic line expressing the S334ter mutant form of rhodopsin to determine whether the combined expression of Bax and Bcl-2 would enhance and prolong the protective effect against photoreceptor cell death afforded by Bcl-2 alone. ^{32 45} The effect of expressing Bax, alone and together with Bcl-2, on the progression of retinal degeneration was assessed in animals at P24 and P31. In contrast to results obtained with Bcl-2, expression of Bax alone did not prevent photoreceptor cell death in animals expressing the mutant rhodopsin transgene (data not shown). When Bax was expressed together with Bcl-2 (RhBax-C/Bcl-2B/S334ter), the ability of Bcl-2 to slow the degenerative process was not improved. The morphology was similar between Bcl-2B/S334ter and RhBax-C/Bcl-2B/S334ter transgenic retinas at P24, with approximately three to four rows of photoreceptor cells remaining (data not shown). However, the Bcl-2B/S334ter retinas displayed two to three rows of sparse photoreceptor cells, and the RhBax-C/Bcl-2B/S334ter retina contained one to two rows of photoreceptor cells at P31 (Fig. 7) . Although, only single RhBax-C/Bcl-2B/S334ter and Bcl-2B/S334ter transgenic retinas were analyzed at the P24 and P31 time points, respectively, morphology of retinas taken from animals with similar genotypes and at slightly different ages are consistent with the general outcome and conclusions derived from this study. These data show that Bcl-2/Bax heterodimers did not serve the function of preventing apoptosis. Rather, the presence of Bax titrated away the anti-apoptotic activity of Bcl-2 to prolong photoreceptor cell survival in S334ter mice. 

Discussion

Ectopic expression of genes that regulate programmed cell death in photoreceptors may serve as a means of treating retinal degenerative diseases. It has been shown that ectopic expression of Bcl-2 in photoreceptor cells temporarily delays programmed cell death in several mouse models of RP, including a transgenic mouse line that overexpresses the S334ter mutant form of rhodopsin and a mouse line that expresses a nonfunctional phosphodiesterase protein (rd mouse). ³² The temporary nature of the protective effect produced by Bcl-2 may be caused by insufficient levels of additional proteins needed to sustain the anti–cell-death function of Bcl-2 in photoreceptor cells. The synergistic effect against photoreceptor cell death in the S334ter rhodopsin mutant mouse produced upon coexpression of BAG-1 with Bcl-2 confirms this notion. ³⁶ The success of BAG-1 in enhancing the anti-apoptotic activity of Bcl-2 in photoreceptors suggests that other proteins, such as Bax, that are thought to form functional complexes with Bcl-2 in certain cell types may also effectively enhance the ability of Bcl-2 to prevent photoreceptor cell death. The present study was undertaken to test the hypothesis that Bcl-2/Bax heterodimer is a functional complex that could delay photoreceptor cell death. 

Endogenous levels of Bax expression were analyzed during early retinal development to determine whether decreasing levels of the Bax protein would correlate with the decreasing ability of the Bcl-2 protein to prevent photoreceptor cell death. Indeed, Bax expression has been found to decrease continuously during the first month of age with a dramatic decrease occurring between P16 and P24, the time during which the neural retina undergoes differentiation. ¹⁷ Bax has also been found to be downregulated during the development of other neuronal tissues such as the cerebral cortex and cerebellum. ⁴⁶ This downregulation of Bax in the retina correlates with the decreased ability of Bcl-2 to prevent photoreceptor cell death during this developmental period in mice that either express the S334ter mutant rhodopsin or are homozygous for the rd mutation. ³² Therefore, if the Bcl-2/Bax heterodimer is the anti-apoptotic functional complex, it is possible that the decreasing level of Bax expression becomes rate-limiting in the anti–photoreceptor-cell-death activity of Bcl-2 in the S334ter and rd mouse models of RP. To test the hypothesis that Bax expression is necessary for the sustained anti–photoreceptor-cell-death activity of Bcl-2, transgenic mice that expressed Bax specifically in photoreceptor cells were produced and crossed with Bcl-2 transgenic animals to produce a transgenic line that expressed both Bax and Bcl-2 in rod cells. A preliminary analysis of the RhBax single and RhBax/Bcl-2 double transgenic retinas performed before introducing the genes into an adRP model showed that the Bax transgene was expressed in a gradient across the transgenic retina as indicated by the progressively decreasing gradient of photoreceptor cell death from the superior to the inferior regions of the eye. It is not known whether the pattern of Bcl-2 transgene expression is uniform or in a gradient similar to the Bax expression pattern. Regardless, coexpression of Bcl-2 with Bax suppressed Bax-induced cell death in a gradient fashion, with the superior region of the retina being severely degenerated, whereas the inferior region had degenerated less and retained photoreceptor outer segments. The gradient of Bax expression produced by the opsin promoter may have created an optimal ratio of expression of Bcl-2 to Bax that significantly prevented the Bax-induced retinal degeneration in the inferior region of the transgenic retina. 

Unlike the synergistic protective effect produced upon coexpression of BAG-1 and Bcl-2 against photoreceptor cell death induced by expression of the S334ter mutant rhodopsin transgene, ³⁶ coexpression of Bax and Bcl-2 did not produce an enhanced neuroprotective effect. In fact, the ability of Bcl-2 to delay the process of photoreceptor cell death was impaired in the presence of Bax, in that the retinal degeneration was more severe in Bcl-2/Bax/S334ter than in Bcl-2/S334ter. Therefore, expression of Bax in the transgenic retinas appeared to titrate away effective concentrations of Bcl-2, leading to a slight acceleration of photoreceptor cell death. These data support a pro-apoptotic function of Bax in the retina that can be inhibited by Bcl-2. 

The domain for Bcl-2 homodimerization is distinct from the heterodimerization domain, and Bcl-2 homodimers containing mutated heterodimerization domains do not demonstrate anti-apoptotic properties. ³¹ Because Bcl-2 heterodimerization appears to be a requirement for its activity, perhaps it is not surprising that overexpression of Bcl-2 alone was not completely effective in protecting photoreceptors undergoing apoptosis. In the current study, Bax did not function in the context of an anti-apoptotic dimerization partner for Bcl-2 in the retina, and the temporal protective effect of ectopic Bcl-2 in the photoreceptors was probably through its ability to heterodimerize with another partner that is present in limited amounts in the photoreceptors. 

Therefore, transgenic mice that overexpress Bax, a death-promoting gene, in photoreceptor cells displayed extensive loss of the retina that was significantly inhibited by Bcl-2 expression, providing a useful in vivo model system to study the mechanism, or mechanisms, by which these two proteins may interact to regulate photoreceptor cell death. Although, much is already known regarding the important role that Bcl-2 and Bax play in regulating the programmed cell death pathway, most of the current knowledge has come from studying the behavior of these two proteins in vitro, either in a test tube or in cultured cells. Understanding how these proteins function in vivo, within a living animal, is a prerequisite to designing viable therapeutic strategies for the treatment of degenerative diseases, such as RP, that result from inappropriate regulation of programmed cell death. 

 Supported by National Eye Institute Grant EY-06702-03 (PE-C, MIS) and EY12155 and Core Facility Grant EY-03040 to Doheny Eye Institute; The Ruth and Milton Steinbach Fund; Research to Prevent Blindness; and the Beckman Foundation (JC). JC is a Research to Prevent Blindness James S. Adams Scholar and a Beckman Investigator.

 Submitted for publication July 3, 2001; revised November 29, 2001; accepted December 18, 2001.

 Commercial relationships policy: N.

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

 

Figure 1.

 

Bax expression was developmentally regulated in photoreceptor cells. Western immunoblot analysis using retinal extracts prepared from wild-type retinas taken at P2, P9, P11, P13, P16, P24, 10 weeks, and approximately 5.5 months of age (lanes 1 through 8, respectively). Approximately equivalent amounts of protein were loaded for each sample. Purified recombinant m-Bax-his protein (1 ng) was run as a control sample (lane 9). Arrows: bands corresponding to Bax and Recoverin proteins.

Figure 2.

 

Expression levels of Bax transgene in RhBax transgenic retinas. Protein immunoblot analysis using retinal extracts prepared from nontransgenic and transgenic littermates at P11. Samples were taken from a nontransgenic littermate (lane 1) and from animals representing transgenic lines RhBax-C (lane 2) and RhBax-F (lane 3). Approximately equivalent amounts of protein were loaded for each sample. Increasing amounts of purified m-Bax-his protein were run as control samples (lanes 4 through 6). Arrows: protein bands corresponding to Bax and Recoverin.

Figure 3.

 

Overexpression of Bax induced a gradient of photoreceptor cell death in transgenic retinas. Light photomicrographs of retinal cross sections taken from a wild-type (A, B), an RhBax-C transgenic animal (C, D), and an RhBax-F transgenic animal (E, F) at P11 and 6 weeks of age. Superior and inferior regions of the retinas are labeled. The dark bands correspond to photoreceptor nuclei stained with a dye and demonstrate the gradient of Bax-induced photoreceptor cell death.

Figure 4.

 

Ectopic expression of Bax affected the morphology of photoreceptor cells and induced retinal degeneration. One-micrometer Epon-embedded sections from wild-type, an RhBax-C, and an RhBax-F transgenic retina at P11. Sections are from the superior (A, C, E) and the inferior (B, D, F) regions of the retinas. Note the differences in the thickness of the ONL between the wild-type and the transgenic retinas.

Figure 5.

 

Expression of Bcl-2 prevented Bax-induced photoreceptor cell death. (A) One-micrometer Epon-embedded sections of various quadrants from retinal sections taken from a wild-type (A–D), an RhBax-C (E–H), and an RhBax-C/Bcl-2B (I–L) transgenic retina at P31 demonstrate the gradient of rescue from cell death provided by Bcl-2 in RhBax-C/Bcl-2B double-transgenic retinas. Photomicrographs were taken in four areas of the retina, from the superior to the inferior regions of the eye that correspond approximately to areas labeled 1 through 4 in the photomicrograph in (B, right). (B) Summary of morphometric analysis. Photoreceptor cell counts were taken within 200-μm spans of four quadrants of the retina, corresponding to the numbers shown on the photomicrograph at right. Counts were taken from two representative animals for each genotype. Morphology of retinas from additional transgenic animals was similar to those used for the analysis. Graph demonstrates the slight inherent variability between retinal samples.

Figure 6.

 

Bax and Bcl-2 formed a complex in RhBax-C/Bcl-2B transgenic retinas. Retinal lysates prepared from an RhBax-C, a Bcl-2B, and an RhBax-C/Bcl-2B transgenic animal at P16, P15, and P18, respectively, were subjected to immunoprecipitation using an anti-Bax antibody (left) or an anti-cytochrome c antibody (right), size-fractionated by SDS-PAGE, and analyzed by Western blot with a Bcl-2 antibody (lanes 1–3). Immunoprecipitates were collected by centrifugation. and aliquots of the resultant supernatants were run as a control, to assess the efficiency of the coimmunoprecipitation (lanes 4–6). Purified m-Bax-his protein (right, lane 7; left, lane 8) or Bcl-2 protein (right, lane 8; left, lane 7) were run as a control. An aliquot of the Bax antibody equivalent to the amount used in the immunoprecipitation step was also used as a control to show nonspecific binding (lane 9). Immunoppt., immunoprecipitate samples; Immunodep. lysate, immunodepleted lysate samples.

Figure 7.

 

Coexpression of Bax and Bcl-2 did not enhance the ability of Bcl-2 to prevent retinal degeneration. Light photomicrographs of retinal sections taken from a wild-type (A), an S334ter (B), an RhBax-C/Bcl-2B (C), and an RhBax-C/Bcl-2B/S334ter (D) transgenic animal at P31. Photomicrographs are taken in the inferior regions of the retinas.

The authors thank Nancy Hong (University of California, Berkeley, CA) for supplying the murine Bax coding sequence and Jean Edens for technical assistance.