Calcium-induced Calpain Mediates Apoptosis via Caspase-3 in a Mouse Photoreceptor Cell Line*

The rd mouse, an accepted animal model for photoreceptor degeneration in retinitis pigmentosa, has a recessive mutation for the gene encoding the β-subunit of the cGMP phosphodiesterase. This mutation results in high levels of cGMP, which leaves an increased number of the cGMP-gated channels in the open state, thus allowing intracellular calcium (Ca2+) to rise to toxic levels, and rapid photoreceptor degeneration follows. To delineate the events in rd photoreceptor degeneration, we demonstrated an increase in calpain and caspase-3 activity, hypothesizing that Ca2+-mediated apoptosis in photoreceptors is mediated by calpain, involving mitochondrial depolarization and caspase-3 activation. To examine this hypothesis further, a murine photoreceptor-derived cell line (661W) was treated with the Ca2+ ionophore A23187, cGMP-gated channel agonist 8-bromo-cGMP, or phosphodiesterase inhibitor isobutylmethylxanthine to mimic the increased Ca2+ influx seen in the rd photoreceptors. Ca2+-induced cell death in 661W cells was found to be mediated by calpain and caspase-3 and could be completely inhibited by the calpain inhibitor SJA6017, implicating both calpain and caspases in the apoptotic process. The apoptotic events correlated in an SJA6017-inhibitable manner with bid cleavage, mitochondrial depolarization, cytochrome c release, and caspase-3 and -9 activation. We concluded that Ca2+ influx in the rd model of photoreceptor degeneration leads to the activation of the cysteine protease calpain, which executes apoptosis via modulation of caspase-3 activity.

The rd mouse, an accepted animal model for photoreceptor degeneration in retinitis pigmentosa, has a recessive mutation for the gene encoding the ␤-subunit of the cGMP phosphodiesterase. This mutation results in high levels of cGMP, which leaves an increased number of the cGMP-gated channels in the open state, thus allowing intracellular calcium (Ca 2؉ ) to rise to toxic levels, and rapid photoreceptor degeneration follows. To delineate the events in rd photoreceptor degeneration, we demonstrated an increase in calpain and caspase-3 activity, hypothesizing that Ca 2؉ -mediated apoptosis in photoreceptors is mediated by calpain, involving mitochondrial depolarization and caspase-3 activation. To examine this hypothesis further, a murine photoreceptor-derived cell line (661W) was treated with the Ca 2؉ ionophore A23187, cGMP-gated channel agonist 8-bromo-cGMP, or phosphodiesterase inhibitor isobutylmethylxanthine to mimic the increased Ca 2؉ influx seen in the rd photoreceptors. Ca 2؉ -induced cell death in 661W cells was found to be mediated by calpain and caspase-3 and could be completely inhibited by the calpain inhibitor SJA6017, implicating both calpain and caspases in the apoptotic process. The apoptotic events correlated in an SJA6017-inhibitable manner with bid cleavage, mitochondrial depolarization, cytochrome c release, and caspase-3 and -9 activation. We concluded that Ca 2؉ influx in the rd model of photoreceptor degeneration leads to the activation of the cysteine protease calpain, which executes apoptosis via modulation of caspase-3 activity.
Inherited retinal degeneration diseases such as retinitis pigmentosa result in photoreceptor degeneration in which night blindness and loss of peripheral vision are the initial symptoms finally culminating in a progressive loss of vision (1). At least 30 genes have been implicated in the genetics of retinitis pigmentosa, many of which encode photoreceptor-specific proteins such as the structural protein peripherin (2), rod outer segment membrane protein-1 (3), rod cGMP phosphodiesterase (4), and rhodopsin (5). Extensive studies of retinal degeneration in animal models such as the rd (6) and the rds (7) mice, as well as several other knock-out and transgenic mice, suggest that apoptosis is the common feature of photoreceptor cell death in all the models (8).
Apoptosis is an active mode of cell death that is induced by a variety of physiological and pathological stimuli. Photoreceptor apoptosis resulting in visual deficits occurs in humans and animals secondary to inherited, chemical-, disease-, and injuryinduced retinal degeneration. Calcium (Ca 2ϩ ) overload is suggested to play a fundamental role in the process of photoreceptor apoptosis in chemical-induced and inherited retinal degenerations. Sustained increases in intracellular Ca 2ϩ trigger apoptosis in a diverse array of in vivo and in vitro systems (9,10). Various studies have indicated that elevated rod photoreceptor Ca 2ϩ plays a key role in the process of apoptotic cell death in humans and animal models during inherited retinal degeneration, retinal diseases and injuries, and chemical exposure. These include patients with retinitis pigmentosa and cancer-associated retinopathy (11), retinal degeneration mice (12), and rats with light-induced damage (13) and hypoxicischemic injury (14).
The genetic defect in the rd mouse is a mutation of the ␤-subunit of the cGMP phosphodiesterase, which is the same gene affected in human recessive retinitis pigmentosa. The recessive mutation in rd mice leads to rapid photoreceptor degeneration, which completely eliminates photoreceptors by postnatal day 21 (P21) (15). The absence of phosphodiesterase activity leads to an increased accumulation of cGMP in the photoreceptors compared with those in wild type mice (6). In turn, this leads to an increase in Na ϩ and Ca 2ϩ influx through the cGMP-gated cation channels, which subsequently causes a metabolic overload and toxicity to the photoreceptors. Photoreceptor Ca 2ϩ levels in the rd mouse are elevated, starting at P5 and have been shown to be increased to ϳ190% by P15 compared with photoreceptors from wild type mice (12). This increase in free intracellular Ca 2ϩ may directly activate apoptosis in photoreceptor degeneration, which is postulated to be the cause of cell death in the rd mouse (8). Although the molecular mechanisms underlying the apoptotic pathway in the rd model remain unknown, there are two possible (but not mutually independent) triggering mechanisms: one is Ca 2ϩ overload, and the other is the generation of reactive oxygen species.
There is converging evidence to suggest that mitochondria and caspases play a fundamental role in the execution of Ca 2ϩinduced rod apoptosis (16). A crucial link is understanding how increased Ca 2ϩ leads to mitochondrial depolarization and/or caspase-3 activation; however, this remains unclear. The two events could be independent (caspase-3 is a Ca 2ϩ -activated protease) or dependent (events downstream of mitochondrial dysfunction are know to lead to caspase-3 activation). Interestingly, in unpublished gene expression array experiments comparing retinas from rd and wild type mice, we have observed a significant increase in the levels of mRNA for m-calpain in the rd mouse. 1 Based on that and other observations in various systems, we propose the following pathway of apoptosis in photoreceptor degeneration ( Fig. 1): an increase in intracellular Ca 2ϩ activates calpain, a Ca 2ϩ -dependent cysteine protease (17), which can cleave the proapoptotic Bcl-2 family protein bid. Interaction of truncated bid (t-bid) with the mitochondrial permeability transition pore (PTP) 2 causes mitochondrial membrane potential loss (⌬⌿ m ), leading to the release of cytochrome c. The increase in cytoplasmic cytochrome c causes the assembly of the apoptosome (18), leading to the activation of caspases, which cleave downstream death substrates and activate endonucleases that cleave genomic DNA into fragments resulting in the apoptotic nuclear morphology (19).
In this study, we first establish that both calpain and caspase-3 are active in the rd mouse retina. To investigate further the Ca 2ϩ -activated apoptotic pathways in photoreceptors, we used 661W photoreceptor cells exposed to the Ca 2ϩ ionophore A23187, the cGMP-gated channel agonist 8-Br-cGMP, or the phosphodiesterase inhibitor 3-isobutylmethylxanthine (IBMX). The 661W cells, which express several markers of photoreceptor cells such as cone opsins, transducin, and cone arrestin (20), are sensitive to photooxidative stress similar to normal retinal photoreceptor cells (21) and showed an increase in intracellular Ca 2ϩ after treatment with A23187, 8-Br-cGMP, or IBMX, making them an excellent in vitro model to study the potential pathways of photoreceptor degeneration of the rd mouse retina. Using appropriate enzyme inhibitors for calpain (SJA6017) and caspase-3 (z-devd-fmk), we were able to confirm the molecular mechanisms involved in the Ca 2ϩ influxactivated apoptotic pathway outlined in the hypothesis.

MATERIALS AND METHODS
Animals-The rd mouse line was generously provided by Dr. Debra Farber (Jules Stein, UCLA), and wild type animals with the same genetic background (C57BL/6) were purchased from Taconic Farms (Germantown, NY) and maintained in the Animal Facility at the Medical University of South Carolina with food and water ad libitum. Animals were handled in accordance with institutional guidelines and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.
Enzymatic Assay for Calpain Activity-Calpain activity was measured using Ac-LLY-AFC as the substrate provided in the calpain activity assay kit (Calbiochem). 661W cells (1 ϫ 10 6 ) were resuspended in 100 l of extraction buffer and centrifuged at 10,000 ϫ g for 1 min. The cell lysate was transferred to a 96-well plate to which 10 l of 10ϫ reaction buffer and 5 l of calpain substrate were added. After incubation at 37°C for 1 h, the samples were read in a fluorometer equipped with a 400 nm excitation filter and a 505 nm emission filter. In the case of retinas dissected from rd and wild type mice, the tissue was resuspended in 100 l of lysis buffer (50 mmol/liter B-glycerophosphate, 20 mmol/liter EDTA, 15 mmol/liter MgCl 2 , 1 mmol/liter dithiothreitol), kept on ice at 4°C for 15 min, and then homogenized by tissue grinding and sonication on ice. The supernatant was collected after centrifugation at 20,000 ϫ g for 15 min. Protein content was measured using Bradford Folin's reagent method (Bio-Rad Laboratories), and enzyme activities were expressed as relative fluorescence units/mg of protein of each sample. The arbitrary values were presented as the mean Ϯ S.E. of three to five experiments or conditions.
Enzymatic Assay for Caspase Activity-The activities of caspase-3, -8, and -9 were measured using the following fluorogenic enzyme substrates: z-DEVD-AFC (Molecular Probes), IETD-AFC (Clontech), and LEHD-AFC (Alexis Biochemicals), respectively. After the desired duration of treatments, the cells were collected as a pellet and resuspended in 50 l of cell lysis buffer and incubated on ice for 10 min. Subsequently, the cells were centrifuged at 3,000 ϫ g for 5 min, and the supernatant was transferred to a 96-well plate to which 50 l of reaction buffer (50 mmol/liter PIPES, pH 7.4, 10 mmol/liter EDTA, 0.5% CHAPS) containing 10 mmol/liter dithiothreitol and 5 l of the respective substrate was added. The plate was incubated at room temperature for 1 h, and the samples were read in a fluorometer equipped with a 400 nm excitation filter and 505 nm emission filter. In the case of retinas dissected from rd and wild type mice, the tissue supernatant was prepared and collected as described above. The protein content was measured using Bradford Folin's reagent method (Bio-Rad Laboratories), and enzyme activity was expressed as relative fluorescence units/mg of protein.
The arbitrary values were presented as the mean Ϯ S.E. of three to five experiments or conditions. FIG. 1. Hypothetical pathway of calpain activation in the Ca 2؉ -induced apoptotic mechanism. Ca 2ϩ enters the photoreceptor outer segment via the cGMP-gated channels and activates calpain, promoting the cleavage of the Bcl-2 family protein, bid, which targets the mitochondria. Cytochrome c release from the mitochondria into the cytosol is triggered by the preceding events, promoting the interaction of Apaf-1 and caspase-9 subsequently leading to caspase-3 activation, which cleaves the DNA into high molecular weight fragments and results in morphological apoptosis. PARP, poly-(ADP)-ribose polymerase.
Measurements of Intracellular Calcium-The intracellular levels of Ca 2ϩ in 661W cells were measured using the Fluorometric Imaging Plate Reader System (FLIPR) as described in the assay protocol (Molecular Devices). The cells were seeded in a black wall 96-well plate with an optimal cell seeding density to have a uniform, confluent monolayer after an overnight incubation. The next day, cells were loaded with a Ca 2ϩ -sensitive fluorescent dye Fluo-3 (4 mol/liter) in a dye loading buffer (Hanks' buffer without phenol red ϩ 20 mmol/liter HEPES and 1% bovine serum albumin) and incubated at 37°C for 1 h. After incubation, the cells were washed three to four times with the wash buffer (Hanks' buffer and 20 mmol/liter HEPES), loaded with the ionophore A23187, 8-Br-cGMP, or IBMX with varying concentrations in different wells, and intracellular Ca 2ϩ levels were assayed in the FLIPR system equipped with an 488 nm excitation filter and 570 nm emission filter.
MTT Assay-Cells were washed twice with cold phosphate-buffered saline (PBS) after appropriate treatment periods and incubated in culture medium with 0.5 mg/ml MTT dye (Sigma) for 2 h. After aspiration of the medium, the dark blue crystals formed were dissolved with 0.1 N HCl in isopropyl alcohol, and absorbance was measured at 570 nm (background wavelength 630 nm) using a spectrophotometer. Results are presented as percentage of survival, taking control as 100%.
Detection and Quantification of Apoptosis-To detect and quantify the number of cells undergoing apoptosis, cells were harvested after appropriate treatments and washed in cold PBS, recentrifuged, and resuspended in annexin binding buffer (50 mmol/liter HEPES, 700 mmol/liter NaCl, 12.5 mmol/liter CaCl 2 , pH 7.4) and stained with Alexa Fluor 488 annexin V and propidium iodide dyes (Apoptosis Assay Kit, Molecular Probes). Cells were then followed by flow cytometric analysis on a FACScan instrument maintained by the Medical University of South Carolina core facility, measuring the fluorescent emission at 530 nm and 575 nm to detect and quantify the population of live and apoptotic cells.
Immunocytochemistry Analysis-Cells were fixed in 4% paraformaldehyde in PBS, pH 7.4, on chamber slides after the appropriate treatment periods. Primary antibodies were added in PBS-TX (0.4% Triton X-100 (TX)) for 2 h at room temperature. The primary antibodies were: rabbit polyclonal anti-activated caspase-3 (1:200, BD Biosciences), goat polyclonal anti-calpain (200 g/ml, Santa Cruz Biotechnology), rabbit polyclonal t-bid (1:100, BIOSOURCE International), and a sheep polyclonal anti-cytochrome c (1:100, Oncogene Research Products). The primary antibodies were visualized as appropriate using goat antirabbit IgG (Jackson Immunoresearch) or rabbit anti-sheep/goat IgG conjugated to fluorescein (Vector Laboratories) at concentrations in accordance with the manufacturer's recommendation. Cells were photographed using a Nikon microscope equipped for fluorescence and a digital camera driven by Axioscope software.
Determination of ⌬⌿ m -The ⌬⌿ m was determined by using JC1 dye (Molecular Probes), which accumulates in a potential-dependent manner in the mitochondria indicated by a fluorescence emission shift from the J aggregate (hyperpolarized mitochondria) exhibiting red fluorescence (525 nm), to J monomer (depolarized mitochondria) showing green fluorescence (590 nm). The population of cells undergoing the shift from J aggregate to monomer signifying mitochondrial depolarization is indicated by a decrease in the red/green fluorescence intensity ratio. Cells were grown on chamber slides, and after treatment with ionophore for 6 h with and without the presence of the appropriate inhibitors, cells were washed with cold PBS and analyzed immediately using confocal microscopy (Leica laser scanning confocal imaging system).
Detection of Cytochrome c Release-To determine the release of cytochrome c from the mitochondrial to cytosolic fraction, a quantitative determination of cytochrome c concentrations in cell lysates and subcellular fractions was performed. To isolate the mitochondrial and cytosolic fractions, cells were fractionated as described in the Apoalert cell fractionation protocol (Clontech). The pelleted cells were resuspended in fractionation buffer mix (including 500ϫ protease inhibitor mixture and 1 mol/liter dithiothreitol), incubated on ice for 10 min, homogenized, and centrifuged at 10,000 ϫ g for 25 min at 4°C. The supernatant was collected, and the protein concentration was determined using Bradford reagents. The total amount of cytochrome c, as well as that in the different subcellular compartments, was then assessed using a cytochrome c immunoassay (R&D Systems). Equal amounts of protein from the cytosolic and enriched mitochondrial fractions were plated in microplate wells containing 100 l of substrate solution. The plate was incubated for 30 min at room temperature, 100 l of stop solution was added, and the plate was read at A 450 nm (correction wavelength at 540 nm). Results are expressed as a measure of cytochrome c concentration (ng/ml) as arbitrary units compared with the untreated controls, in the mitochondrial and cytoplasmic subcellular fractions of the cells.
Statistical Analysis-All experiments described were performed at least in triplicate. Statistical significance was determined using Student's t test, with a significance level of p Ͻ 0.05.

RESULTS
Calpain and Caspase-3 Activity in the rd Mouse Retina-To establish the cornerstones of our hypothesis of the apoptotic pathway in the rd mouse photoreceptors, calpain and caspase-3 activities were measured in rd and wild type mouse retinas. As shown in Fig. 2A, calpain activity in the rd mouse retina was elevated at P7 compared with wild type retinas; activity levels peaked by P10; and after completion of degeneration of photoreceptors (P21), the activity levels were comparable with those found in the wild type mouse retina. Similarly, caspase-3 levels (Fig. 2B) were elevated in the rd mouse retina starting at P7, peaked at P15, and dropped to wild type levels by P21. The temporal pattern of these two enzymes is indicative of a sequential activation. In addition, these results correlate with an earlier report that mouse Ca 2ϩ levels in the rd retinas were increased significantly compared with wild type retinas beginning at P5 and stayed elevated until P17 (12). The markedly increased activity levels of calpain and caspase-3 in the rd mice, in parallel with the massive Ca 2ϩ influx, suggest a correlation of these two events.
Calcium-induced Apoptosis in 661W Cells-After having established in vivo that both calpain and caspase-3 are activated in the rd retina, we sought to establish a photoreceptor cell culture model to investigate the mechanisms of Ca 2ϩ -induced cell death. For this purpose, 661W cells were treated with A23187, 8-Br-cGMP, or IBMX, all of which were used to mimic the massive Ca 2ϩ influx observed in the rd mouse model. Intracellular Ca 2ϩ levels were measured by loading the 661W cells with the Ca 2ϩ indicator Fluo-3 and measuring the fluorescence changes after treatment with the compounds by the FLIPR method. 661W cells showed a multifold increase in the intracellular Ca 2ϩ after treatment with A23187, 8-Br-cGMP, or IBMX compared with the untreated controls (Fig. 3A). Cell viability in response to the prolonged rise in intracellular Ca 2ϩ was measured using the colorimetric MTT assay. 661W were found to undergo cell death in a time-and dose-dependent manner after treatment with A23187 (Fig. 3B). The 5 mol/ liter dose of ionophore A23187, which caused 40 Ϯ 5.8% cell death after 24 h, was chosen for all further experiments. We will refer to the 5 mol/liter calcium ionophore treatment as A23187 or ionophore treatment. Likewise, when 661W cells were subjected to 1 mmol/liter 8-Br-cGMP or 1 mmol/liter IBMX for different time periods, a pattern of cell death was observed similar to that seen for the Ca 2ϩ ionophore (57.7 Ϯ 2.3% and 66.7 Ϯ 4.8% cell death, respectively) (Fig. 3C). To confirm that ionophore-induced cell death in 661W cells was associated with apoptosis, flow cytometry analysis of fluorescently labeled annexin V binding (plotted on each x axis of Fig.  4) and propidium iodide uptake (plotted on each y axis of Fig. 4) was performed. The untreated cells remained viable as observed in the lower left panel (R1) after 18 h (Fig. 4A) and 24 h (Fig. 4B). Ionophore-treated cells underwent apoptosis as seen by a shift toward the lower and upper right panels (R2 and R4) of

. Measurement of intracellular Ca 2؉ levels and cell viability in 661W cells.
A, levels of intracellular Ca 2ϩ in 661W cells were measured by Fluo-3 dye using the FLIPR method. Cells were loaded with Fluo-3 dye, incubated for 1 h, and treated with 5 mol/liter A23187, 1 mmol/liter 8-Br-cGMP, or 1 mmol/liter IBMX. There was a significant increase in the levels of intracellular Ca 2ϩ in 661W cells after treatment with any of the three compounds compared with the untreated controls. B, A23187 treatment caused a loss of cell viability in a time-and dose-dependent manner as indicated by the spectrophotometric analysis of 661W cells after a 2-h incubation at 37°C with 0.5 mg/ml MTT dye. Cell viability was decreased 40% (Ϯ5.8) by 24 h and 60% (Ϯ4.1) by 30 h after a 5 mol/liter ionophore treatment. C, a significant decrease in cell viability was also observed after treatment with 8-Br-cGMP or IBMX (57.7 Ϯ 2.3% and 66.7 Ϯ 4.8% after a 24-h treatment, respectively). D, the effect of SJA6017 and z-devd-fmk pretreatment on cell viability was measured at different time intervals. Pretreatment with 100 mol/liter SJA6017 completely attenuated the ionophore-induced cell death (92.2 Ϯ 3.4%), whereas 2 mol/liter z-devd-fmk offered only partial protection (72 Ϯ 5.8%). A similar increase in cell viability after 24 h was observed in the case of 8-Br-cGMP (88.8 Ϯ 4.5% and 72.3 Ϯ 3.4%) and IBMX (90.4 Ϯ 5.6% and 73 Ϯ 4.6%) with SJA6017 and z-devd-fmk pretreatments, respectively. The results are shown as a percentage of control values (arbitrarily set at 100%). All experiments were performed in triplicate; p Ͻ 0.05, and data are presented as the mean Ϯ S.E. also activated by Ca 2ϩ influx in 661W cells, pretreatment with both the enzyme inhibitors was used for the cell survival assays, enzymatic activity assays, and immunocytochemistry.
To examine the effects of the calpain inhibitor SJA6017 and the caspase-3 inhibitor z-devd-fmk on the Ca 2ϩ influx-induced cell death, cells were pretreated with these inhibitors, and the cell viability was measured by MTT assay. Inhibiting calpain by SJA6017 significantly increased the cell survival (Fig. 3D) compared with z-devd-fmk pretreatment (92.2 Ϯ 3.4% and 72 Ϯ 5.8% cell viability, respectively) after A23187 treatment for 24 h. The cell death triggered by 5 mol/liter A23187 could be attenuated by SJA6017 in a dose-dependent manner (data not shown), with a maximal effect at 100 mol/liter, but only a partial rescue could be achieved with z-devd-fmk pretreatments (Fig. 3D), with a maximal effect at 2 mol/liter. These two doses for both inhibitors were subsequently used for all experiments and will therefore only be referred to by name, excluding the dose. Likewise, cell viability was significantly higher after SJA6017 pretreatment (88.8 Ϯ 4.5% and 90.4 Ϯ 5.6%) than z-devd-fmk (72.3 Ϯ 3.4% and 73 Ϯ 4.6%), when cells were exposed to 8-Br-cGMP or IBMX for 24 h. Similarly, using annexin V binding and propidium iodide uptake, pretreatment with SJA6017 (Fig. 4, E and F) was found to offer complete protection (6.8 Ϯ 1.8% and 8 Ϯ 2.2% apoptotic cells, respectively) from the ionophore-induced apoptotic cell death, whereas z-devd-fmk pretreatment (Fig. 4, G and H) offered only a partial rescue at both the time points (17.7 Ϯ 2.3% and 24 Ϯ 4.0% apoptotic cells, respectively). Because caspase-3 activation is a relatively late event in the apoptotic cascade, a significant number of cells may already be in the late stage of apoptosis, detectable by annexin V staining (Fig. 4, G and H). On the other hand, SJA6017 presumably blocks the events upstream of mitochondrial changes and the caspase cascade activation, thereby completely blocking the proapoptotic events induced by the ionophore. These findings demonstrate that raising intracellular Ca 2ϩ by A23187, 8-Br-cGMP, or IBMX induces apoptosis in 661W cells, and prevention of calpain activation is a key event to rescue the photoreceptor cells from Ca 2ϩ influx-induced cell death.
Enzymatic activity of the two protease systems in the cells was monitored further in the presence of inhibitors to establish the activation cascade. Fluorometric measurement of the calpain substrate Ac-LLY-AFC was used to measure calpain activity, which was activated by the ionophore, 8-Br-cGMP, or IBMX treatment and could only be inhibited by SJA6017 but not with z-devd-fmk (Fig. 5A), suggesting an upstream role of the Ca 2ϩ -activated cysteine protease, calpain. Conversely, analysis of the cleavage of the caspase-3 substrate, Ac-DEVD-AFC, induced by A23187, 8-Br-cGMP or IBMX, demonstrated an increase in caspase-3 activity, which could be inhibited by both SJA6017 and z-devd-fmk (Fig. 5C). These results suggest that the Ca 2ϩ -induced calpain activation mediates cell death by regulating caspase-3 activity in an SJA6017-inhibitable manner.
Immunocytochemistry was performed using specific antibodies for calpain and activated caspase-3 after ionophore induction of 661W cells. The untreated cells showed minimal fluorescence compared with a markedly increased signal for calpain after ionophore treatment (Fig. 6A). A similar observation was made for the presence of activated caspase-3, which, as expected, had a relatively weak fluorescent signal for untreated cells compared with the bright intense fluorescence seen in ionophore-treated cells (Fig. 6D).
Taken together, our results indicate that (i) Ca 2ϩ -induced cell death is caused by apoptosis, involving the activation of both calpain and caspase-3; and (ii) the acute rise in Ca 2ϩ can be reproduced by different means (Ca 2ϩ ionophore, 8-Br-cGMP, or IBMX) resulting in comparable levels of cell death and enzyme activation. Thus, the remainder of the pathway was analyzed using Ca 2ϩ ionophore treatment.
Analysis of the Caspase Cascade Involved in Ionophore-induced Apoptosis-Subsequent to the initial trigger of massive Ca 2ϩ influx, there could be two possible pathways for the initiation of the apoptotic mechanism. One of them is a Ca 2ϩactivated cysteine protease calpain; the other way could be caspase-8 activation. As hypothesized in Fig. 1, the effector caspase-3 could be initiated by bid cleavage-mediated mitochondrial events, which in turn are triggered by either calpain or caspase-8. To delineate this part of the pathway involved in Ca 2ϩ -induced apoptosis in 661W cells, caspase-8 and -9 enzyme activities were examined (Fig. 5, B and D). The proposed alternate route of Ca 2ϩ -induced caspase-8 activation was not detected in our model because no changes in the Ca 2ϩ -induced caspase-8 activity were detected in the 661W cells (Fig. 5B) or in the rd mouse retinas (data not shown). However, ionophore treatment leads to an increase in caspase-9 activity, which was attenuated by SJA6017, but z-devd-fmk did not change the activity levels (Fig. 5D).
Calpain Cleaved Bid-induced Cytochrome c Release from the Mitochondria-So far, the results demonstrated that iono-FIG. 5. Ca 2؉ -induced caspase-3 and calpain activation in 661W cells. A, activation of calpain was induced in 661W cells after A23187, 8-Br-cGMP, or IBMX treatments, as determined by the use of membrane-permeant fluorogenic calpain-specific substrate Ac-LLY-AFC. The fluorescence (400/505 nm; excitation/emission) was measured as relative fluorescence units (RFU)/mg of protein and is expressed as arbitrary units determining the change in calpain activity, compared with untreated cells. Preincubation with SJA6017 completely inhibited the Ca 2ϩ -induced calpain activity, whereas z-devd-fmk had no effect on calpain activity (p Ͻ 0.02). B, caspase-8 activity was measured by assessing the fluorometric cleavage of IETD-AFC after treatment with A23187, which demonstrated no significant change in 661W cells. For all experiments, the arbitrary values represent the mean Ϯ S.E. for triplicate experiments per treatment condition (p Ͻ 0.01). C, the induction of caspase-3 activity in 661W cells after treatment with A23187, 8-Br-cGMP, or IBMX was measured by cleavage of the fluorogenic substrate DEVD-AFC (relative fluorescence units/mg of protein and expressed as arbitrary units). Preincubation with either z-devd-fmk or SJA6017 completely blocked the Ca 2ϩ -induced caspase-3 activity (p Ͻ 0.01). D, the ionophore induction of caspase-9 activity was measured by the cleavage of LEHD-AFC substrate. SJA6017 pretreatment completely attenuated the ionophore-induced caspase-9 activity, whereas z-devd-fmk offered no change (n ϭ 3, p Ͻ 0.01).
phore-induced calpain triggers the apoptotic cell death machinery involving caspase-9 and the effector caspase-3. The role of intermediate events involving the intrinsic (mitochondrial) pathway was investigated further to link more tightly the observed events of Fig. 1.
The Ca 2ϩ -activated calpain is presumed to target bid, a proapoptotic Bcl-2 family member protein, which can alter the mitochondrial PTP. Upon cleavage, the t-bid allows for the opening of the PTP, leading to a collapse of the mitochondrial membrane potential and ultimately the release of cytochrome c (18). Analysis of Western blots of ionophore-treated cells revealed the presence of the 15-kDa t-bid protein (Fig. 7), which was absent in the untreated cells and could be blocked by the calpain inhibitor SJA6017 but not by z-devd-fmk. Similarly, an increase in t-bid signal was observed using immunofluorescence analysis of ionophore-treated cells compared with untreated cells (Fig. 6B).
The ⌬⌿ m in response to the ionophore-induction was investigated by the use of JC1 dye with confocal microscopy. Ionophore treatment induces a shift from the JC1 aggregate (red) to the JC1 monomeric (green) form of the dye, signifying mitochondrial depolarization because of the loss of membrane potential (Fig. 8, C and D) compared with the untreated cells (Fig.  8, A and B), which had an abundance of JC1 aggregate-labeled cells. By blocking calpain activation via SJA6017 (Fig. 8, E and  F), the mitochondrial membrane potential remained polarized as evidenced by an increase of the population of cells exhibiting the JC1 aggregate form, whereas z-devd-fmk pretreatment only had a marginal effect (Fig. 8, G and H). 10 mol/liter cyclosporin A, a PTP antagonist, was used as a positive control that attenuated the ionophore-induced mitochondrial depolarization (Fig. 8, I and J).
The subsequent translocation of cytochrome c from the mitochondria to the cytosol in response to the ionophore was analyzed using immunocytochemistry (Fig. 6C) and enzyme-linked immunosorbent assays (Fig. 9). A dramatic increase in the near homogeneous staining for cytochrome c was observed in cultures treated with the ionophore for 4 h. In untreated (control) cultures, staining for cytochrome c was light and appeared to be localized in puncta distributed throughout the cytoplasm, consistent with its mitochondrial localization. Likewise, subcellular fractionation of the ionophore-treated cells demonstrated a shift of cytochrome c from the mitochondrial to the cytosolic compartment, which could be prevented by SJA6017 and cyclosporin A, but not by z-devd-fmk (Fig. 9). DISCUSSION Caspases and calpain represent two cytosolic proteolytic systems, which are capable of producing cleavage of various endogenous proteins. The role of caspase-3 in the execution of apoptosis in two models of photoreceptor degeneration (i.e. rd mouse and lead-induced photoreceptor apoptosis) has been generally accepted (16,22). Yet, a recent report postulates a caspase-independent, but possibly calpain-dependent, mechanism in the rd mouse model of photoreceptor degeneration (23). By contrast, however, in both the in vivo and in vitro models used in our studies we observed the activation of both calpain and caspase-3, using a variety of techniques. It is well estab- FIG. 6. Immunofluorescence analysis of ionophore-induced apoptotic components in 661W cells. Fluorescence microscopy was used to analyze the effect of ionophore on components of the apoptosis cascade (i.e. the presence of activated caspase-3, expression of calpain, bid cleavage, and cytochrome c release) in 661W cells. Cells were treated with A23187 for appropriate time periods, fixed in paraformaldehyde on chamber slides, incubated with antigen-specific polyclonal antibodies followed by fluorescein isothiocyanate-conjugated secondary antibodies, and observed under a fluorescence microscope. The untreated cells showed a very weak fluorescent signal for calpain (A) compared with the cells treated with the ionophore for 1 h. SJA6017 blocked the increase in calpain, whereas z-devd-fmk failed to do so. The antibody specific for truncated bid (B) revealed a significant increase in the intensity of fluorescence in the cells treated with ionophore for 3 h compared with the untreated cells. This increase in t-bid was attenuated by SJA6017 pretreatment, but not by z-devd-fmk. Cytochrome c (C) was localized to mitochondria in the untreated cells (punctate staining) but was found to be distributed throughout the cytosol (diffuse fluorescence in the cells) when treated with the ionophore for 4 h. SJA6017 prevented the cytochrome c translocation, whereas z-devd-fmk could not prevent this event. Caspase-3 activation (D) was induced by ionophore treatment for 6 h and was attenuated by both SJA6017 and z-devd-fmk pretreatments.
lished that several caspases are involved in, if not always essential for, executing apoptosis, but the cross-talk between calpain and caspases remains unclear. Various studies have shown that activation of calpain induced by different apoptotic stimuli precedes cell death in different cell systems. Calpains can presumably activate the apoptotic phenotype by utilizing the existing latent death machinery, mainly via involvement of the caspase cascade. Hence, we set out to investigate the cross-talk between calpain and caspase-3 via the crucial endogenous regulators of the intrinsic apoptotic pathway (e.g. the proapoptotic factors such as cytochrome c) as represented in Fig. 1.
Pharmacological manipulations of the rd mouse retina in vivo are difficult to achieve because animals would have to be treated at a very early age (prior to P6; see Fig. 2) and are difficult to interpret because of complex interactions. For this reason, we determined the cornerstones of the apoptotic pathway in vivo and followed up the pharmacology of the pathway analysis in vitro. Using enzymatic assays, we established that both calpain and caspase-3 are in a highly activated state during the phase of photoreceptor degeneration (P7-21) in the rd mouse, and the time course is highly indicative of a sequential activation of the two respective enzymes. To be able to analyze the apoptotic pathway in vitro, it was necessary to recapitulate Ca 2ϩ -induced apoptosis in a photoreceptor cell line. For this purpose, 661W cells, which are derived from an ocular tumor and exhibit several markers of differentiated photoreceptors (20), were exposed to compounds known to increase intracellular Ca 2ϩ , i.e. A23187, 8-Br-cGMP, and IBMX. Subsequent to the multifold increase in Ca 2ϩ induced by these compounds (Fig. 3A), the 661W cells were found to undergo apoptosis in a calpain-and caspase-3-dependent manner. After having established the model system, we present evidence that ionophore-induced apoptosis involves the sequential activation of these two enzyme systems and not two parallel pathways. The obvious candidates to link the upstream protease (i.e. calpain) with the effector caspase (i.e. caspase-3) are the mitochondria (see Fig. 1). Activation of calpain is known to cleave the proapoptotic Bcl-2 protein bid, which targets the mitochondria. Truncated bid has been shown to bind to BAK, a member of the Bax subfamily of the proapoptotic Bcl-2 family members, resulting in its oligomerization and formation of pores in the mitochondrial outer membrane, enabling the release of cytochrome c (24). Once released into the cytosol, cytochrome c is postulated to promote the formation of the caspase-9-Apaf-1 complex, which forms the apoptosome, which in turn activates caspase-3 (18). In the experiments reported here, ionophore was found to induce an increase in bid cleavage, which could be inhibited by the calpain inhibitor, SJA6017. Likewise, Ca 2ϩ ionophore treatment resulted in a collapse of the mitochondrial membrane potential, in cytochrome c release from the mitochondria, and in caspase-3 activation, all three of which could be blocked by the calpain inhibitor SJA6017. Thus, it appears that in the cell culture model of photoreceptor degeneration, caspase activation proceeds through the intrinsic (mitochondrial) apoptotic pathway. A similar model of ionophore-induced calpain activation triggering intracellular changes involving the release of cytochrome c from the mitochondria has been postulated in a large cell carcinoma cell line (25). Although massive Ca 2ϩ influx may target the mitochondria directly, in our model we observed that inhibition of calpain activation offers protection by blocking the downstream mitochondrial events and activation of the caspase cascade, thereby ruling out a direct mechanism for Ca 2ϩ influx affecting the mitochondria. An alternate route of bid cleavage and mitochondrial depolarization is presumed to be via caspase-8 activity (26). However, in our model the ionophore induction did not induce caspase-8 activity in either the cell culture model (Fig. 5B) or the in vivo model of the rd mouse retina (data not shown).
The results from this study identified selective targets of increased Ca 2ϩ influx in the rd model of photoreceptor degeneration and delineated the biochemical mechanisms and temporal sequence of the apoptotic signaling cascade events. Our study demonstrates that extrinsic agents can block the upstream events of the apoptotic pathway regulating the progres-FIG. 8. Ionophore-induced ⌬⌿ m . JC1 dye-stained 661W cells were analyzed with confocal microscopy in the absence (control) and presence of A23187. The untreated cells showed a significant population of cells with the J aggregate (red) form of the JC1 dye (A) compared with the J monomer (green) form (B) indicative of a normal mitochondrial membrane potential. A significant shift of J aggregate (red) to J monomer (green) was observed in the population of cells treated with A23187 for 6 h (C and D) signifying a loss of mitochondrial membrane potential. However, pretreatment with SJA6017 (E and F) or 10 mol/liter cyclosporin A (CsA) a PTP antagonist (I and J), attenuated the mitochondrial depolarization induced by the ionophore; whereas no change in the J monomer (green) staining was observed after pretreatment with z-devdfmk (G and H).
FIG. 9. Determination of cytochrome c release. The release of cytochrome c was analyzed by quantitatively evaluating the mitochondrial and cytoplasmic subcellular fractions and analyzing them fluorometrically. 661W cells were treated by A23187 for 4 h with and without 10 mol/liter cyclosporin A (CsA), SJA6017, or z-devd-fmk pretreatments. A23187 triggered the release of cytochrome c from the mitochondria into the cytosol as the cytochrome c concentration in the mitochondrial subcellular fraction was decreased to 34 Ϯ 1.1% compared with the untreated control. SJA6017 pretreatment blocked the cytochrome c translocation with a considerable increase in the cytochrome c concentration of the mitochondrial fraction (93 Ϯ 1.3%). Pretreatment with z-devd-fmk did not block the release of cytochrome c in response to A23187 treatment as the mitochondrial cytochrome c concentration (30 Ϯ 1.4%) was comparable with the ionophore treatment alone. The PTP antagonist cyclosporin A was used as a positive control. All experiments were done in triplicate; p Ͻ 0.01. sion of retinal degeneration. The ability to modulate key regulators of apoptotic cell death in the context of a mammalian photoreceptor-derived cell line, and more importantly, in the context of a known human retinal degenerative disorder, will be useful for developing novel therapeutic strategies.