Apoptotic Retinal Neuronal Death by Ischemia–Reperfusion Is Executed by Two Distinct Caspase Family Proteases

Abstract

purpose. To evaluate possible roles of caspase-1 and caspase-3 in retinal ischemia–reperfusion injury.

methods. Retinal ischemia was induced in rats by increasing the intraocular pressure to 110 mm Hg for 60 minutes. Expression of caspase-1 and caspase-3 was studied at the mRNA and protein levels using immunohistochemical staining, western blot analysis, semiquantitative reverse transcription–polymerase chain reaction, and assay of the enzymatic activities. Apoptotic retinal neurons were detected by the TdT-dUTP terminal nick-end labeling (TUNEL) method. To study the roles of the caspases in retinal ischemia–reperfusion injury, an inhibitor of caspase-1, acetyl-tyrosyl-valyl-alanyl-aspart-1-al (Ac-YVAD-CHO; total dose, 10⁻⁷ moles) and that of caspase-3, acetyl-aspartyl-glutamyl-valyl-aspart1-al (Ac-DEVD-CHO; total dose, 10⁻⁷ moles) was injected intravitreally and the number of TUNEL-positive cells was compared with the number in sections not treated with the inhibitors.

results. In the inner nuclear layer (INL), caspase-3-like immunoreactivity was predominantly detected, whereas caspase-1-like immunoreactivity was more predominant in the outer nuclear layer (ONL). Expression of caspase-1 and -3 was upregulated at the protein and gene levels 24 hours after reperfusion. Intravitreal injection of Ac-DEVD-CHO decreased the number of TUNEL-positive cells more significantly in the INL than in the ONL (P < 0.01) at 24 hours, whereas, intravitreal injection of Ac-YVAD-CHO was more effective in decreasing the number in the ONL (P < 0.05).

conclusions. These findings suggest a possibility that cell-type–specific activation of caspases takes place in retinal ischemia–reperfusion injury, and such caspase may induce retinal neuronal cell death.

In contrast to necrosis, apoptotic cell death is a highly ordered process that is characterized by nuclear changes such as chromatin condensation, margination, and fragmentation of the nucleus as well as internucleosomal DNA cleavage. The ultrastructural changes consist of cytoskeletal disruption, cell shrinkage, and membrane blebbing, which then lead to the fragmentation of the dying cell into apoptotic bodies. ¹ This apoptotic process is mediated by transcriptional and translational activation of cell death or cell suicide programs. ^{2 3 4} Although the final step of apoptosis is cell death, various signals contribute to the final common pathway, and intermediate steps can differ depending on the cause of the apoptosis and the cell type. ^{4 5 6}  

Apoptosis plays a major role in the pathogenesis of many ocular diseases and animal models including retinitis pigmentosa, retinal detachment, light injury, glaucoma, age-related macular degeneration, and ischemia–reperfusion injury. ^{7 8 9 10} It has been determined that c-fos is involved in a photoreceptor apoptotic pathway induced by light injury, because this light-induced apoptosis is virtually absent in mice without c-fos. ¹¹ Thus, it is important to know which genes are specifically expressed during apoptosis of retinal cells, because apoptotic cell death can be rescued by inhibiting the expression of essential genes of apoptosis. 

Various gene products play a role in the process, ^{3 9 12 13 14} and among them, several members of the IL-1β-converting enzyme (ICE/caspase-1) family have been implicated in the cascade of apoptotic death in a variety of cell types. ^{4 15 16 17 18 19 20} ICE was initially identified as a ced-3 gene homologue that is required for programmed cell death in the nematode, Caenorhabditis elegans. ^{16 20} At present, at least 11 ICE homologues named caspase-1 through caspase-11 have been identified and are divided into three groups. ^{21 22} Caspase-1 encodes a cysteine protease that cleaves proteins after aspartic acid, ^{23 24} and other caspases are also translated as inactive proenzymes, activated by proteolysis and can cleave a number of cellular proteins. Sometimes the cleavage results in the activation of the protein and sometimes in inactivation. ^{23 24}  

Recently, we and other groups showed that neuronal apoptosis and necrosis play a role in the pathologic mechanisms of ischemia–reperfusion injury in the rat retina in both the inner nuclear layer (INL) and outer nuclear layer (ONL). ^{25 26 27 28} Herein, we present evidence that different caspase families probably play an important role in the apoptotic processes in photoreceptors and retinal neurons in the INL. 

Methods

Animals and Ischemia Model

One hundred four male Sprague–Dawley rats weighing 250 to 300 g were used in this study. The care and maintenance of the rats conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Rats were deeply anesthetized with an intraperitoneal injection of pentobarbital (60 mg/kg) and the pupils dilated with topical phenylephrine hydrochloride and tropicamide. The anterior chamber of the right eye was cannulated with a 27-gauge infusion needle connected to a bottle containing normal saline. The intraocular pressure was raised to 110 mm Hg for 60 minutes by elevating the saline reservoir. Sham-procedure right eyes were treated similarly but without the elevation of the bottle, so the normal ocular tension was maintained. Time-course examination was performed at 6, 12, 24, 48, and 96 hours after reperfusion. 

Antibodies

Antibodies used in this study were obtained from various sources: goat anti-caspase-1 antibody (M-19) ²⁹ and goat anti-caspase-3 antibody (L-18) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA); rabbit anti-bovine s-100 antibody from Immunotech (Marseilles, France); rabbit anti-glial acidic fibrillary protein (GFAP) antibody from Nichirei (Tokyo, Japan); monoclonal anti-α-tubulin antibody from Sigma (St. Louis, MO); rhodamine-conjugated goat anti-rabbit IgG from Dako (Glostrup, Denmark); and rhodamine-conjugated rabbit anti-goat IgG from Jackson ImmunoResearch Laboratories (West Grove, PA). 

The antibody against caspase-1 (M-19) was raised against a peptide corresponding to amino acids 276–294 mapping to the carboxyl-terminus of the 20-kDa subunit of the caspase-1 precursor of mouse origin. This antibody recognizes the 20-kDa subunit of active caspase-1. Anti-caspase-3 antibody was raised against a peptide corresponding to amino acids 157–174 mapping to the carboxyl terminus of the 20-kDa subunit of the 32-kDa cysteine protease precursor of human origin. This antibody also reacts with the active form of caspase-3. The specificity of the antibodies was verified by the manufacturers, and the antibodies were known to cross-react with rat proteins. These antibodies were used for both immunohistochemical studies and western blot analyses. 

TUNEL Staining

DNA TUNEL was performed as described. ³⁰ After rats were anesthetized with intraperitoneal injection of pentobarbital (60 mg/kg), they were perfused transcardially with 100 ml normal saline supplemented with 1 U/ml heparin followed by 200 ml 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). The eyes were enucleated and placed in 4% paraformaldehyde for 24 hours at 4°C. They were rinsed with phosphate-buffered saline (PBS) and then transferred to 10%, 15%, and 20% sucrose in 0.1 M PBS (3 hours for each step) at room temperature. Samples were then snap frozen in Tissue-Tek (Miles Laboratories, Elkhart, IN) on dry ice in hexane and stored at −80°C. Cryostat sections (6–10 μm) were thaw mounted onto glass slides coated with poly-l-lysine and air dried for 2 hours at room temperature. After sections were rinsed three times in 0.02 M PBS (pH 7.4) and reacted with proteinase K, they were incubated with terminal dUTP transferase and biotinylated dUTP in TdT buffer (30 mM Tris [pH 7.2], 140 mM sodium cacodylate, and 1 mM cobalt chloride) for 60 minutes at 37°C in a moisture chamber. After rinsing, the sections were reacted with avidin-fluorescein isothiocyanate and examined with a scanning laser confocal microscope (LSM 410 inverted laser microscope; Carl Zeiss, Oberkochen, Germany). Positive controls were generated using DNase I in TdT buffer (1μ g/ml) before incubation with terminal transferase and biotinylated nucleotides. 

Immunohistochemistry

After rinsing with 0.02 M PBS, cryosections were incubated with 2% normal rabbit or goat serum for 60 minutes at room temperature. Incubation of the primary antibody was performed overnight in a moisture chamber at 4°C. Working concentrations of the primary antibodies (anti-caspase-1, 1:100; anti-caspase-3, 1:100) were determined after applying various concentrations of the antibodies. Rhodamine-conjugated secondary antibody was reacted at room temperature for 2 hours. To investigate whether caspase-1 and caspase-3 are expressed in apoptotic cells, TUNEL staining was performed after immunostaining. To exclude the possibility that TUNEL-positive cells were retinal glial cells and Müller cells, anti-s-100 and anti-GFAP antibodies were reacted before TUNEL staining. Sections were observed with the laser scanning confocal microscope in the transmitted light mode. 

Quantitative Analysis of TUNEL and Immunohistochemistry

Four eyes were used to obtain data for each time point to investigate the time course of TUNEL and immunohistochemical staining. The number of TUNEL-positive and immunostained cells in the INL and ONL was determined in five meridian sections through the optic nerve head. The number of cells was counted in 20 areas of each section at 1 mm from the optic nerve head on the superior and inferior hemisphere using the confocal microscope’s area measure and measure functions. Data are represented as number of cells per square millimeter. ^{25 28} Results are expressed as mean ± SEM. 

Western Blot Analysis

Retinas were dissected from enucleated eyes (n = 4 for each time point) with fine forceps under an operating microscope; the samples were immediately frozen on dry ice and kept at −80°C until used. Samples were homogenized in buffer containing Tris-HCl (pH 8.0), 250 mM NaCl, 0.1 mM phenylmethylsulfonyl fluoride (PMSF), 10μ g/ml pepstatin, 20 μg/ml leupeptin, and 10 μg/ml aprotinin. The protein content was determined using the method of Bradford. ³¹ The mixture was then added to 5× Laemmli sample buffer containing 125 mM Tris-HCl (pH 6.8), 10% (wt/vol) glycerol, 3% sodium dodecyl sulfate, 6% urea, 5%β -mercaptoethanol, and 0.02% bromphenol blue. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis was performed as described ³² after samples were boiled for 5 minutes. A sample of 50 μg protein was loaded on each lane and was electrophoresed on 10% acrylamide gel and then transferred to nitrocellulose membranes. ³³ After rinsing with Tris-buffered saline (pH 7.4) containing 0.25% Tween-20 (TBST), the membrane was incubated with primary antibodies (1:500) in TBST for 1 hour at room temperature. After the membrane was washed three times, it was incubated with the horseradish peroxidase–conjugated secondary antibody diluted 1:500 in TBST for 1 hour at room temperature, washed again, and developed with the enhanced chemiluminescence western blot analysis system (ECL; Amersham, Buckinghamshire, UK). The gel was then exposed to hyperfilm (ECL; Amersham) for 1 to 3 minutes. Loading of approximately equivalent amounts of intact protein content was confirmed by the blot with antibody against α-tubulin protein. 

Semiquantitative RT-PCR

Total RNA was extracted from the frozen stored retina by the guanidium thiocyanate method. PolyA⁺RNA was then purified from the total RNA by an oligo-dT column. Aliquots of polyA⁺RNA (0.1 μg) were used to make the cDNA with a first-strand cDNA synthesis kit (Pharmacia-Biotech, Uppsala, Sweden). PCR was performed by the method of Saiki et al. ³⁴ with slight modifications. The following conditions were used: denaturation at 94°C for 45 seconds; annealing at 55°C for 45 seconds; and extension at 72°C for 90 seconds. The reaction was performed for 33 to 35 cycles. After PCR, 1 μl of PCR product was cloned directly into a plasmid vector (pCRII; Invitrogen, San Diego, CA). Nucleotide sequencing of the cloned DNA was performed by the dideoxynucleotide chain termination method by using a gene analyzer (ABI Prism 310; Perkin–Elmer, Foster City, CA). The sequence data of the PCR products were identical with rat sequences found in GenBank. The optimal PCR cycle number for each primer set was obtained by varying the number of PCR cycles. Primers used to amplify the target genes were: Rat caspase-1 cDNA, 5′-TTCAGGCATGCCGTGGAG (sense) and 5′-AATGTCCCGGGAAGAGGT (antisense), 363 bp ³⁵ ; rat caspase-3 cDNA, 5′-ACATGGAAGCGAATCAATGGACTC (sense) and 5′-AAGGACTCAAATTCTGTTGCCACC (antisense), 698 bp ³⁵ ; and ratβ -actin cDNA, 5′-AGCTGAGAGGGAAATCGTGC (sense) and 5′-CCACAGGAT-TCCATACCTGA (antisense), 297 bp. ³⁶ Expression of β-actin was used as the internal standard. Reaction mixture of 1 picomole/ml caspase-1 primer sets and 0.3 picomoles/mlβ -actin primer sets in a tube was amplified for 33 cycles. Thirty-five PCR cycles were performed for caspase-3 cDNA amplification in a tube including 1 picomole/ml caspase-3 primer sets and 0.2 picomoles/ml β-actin primer sets. PCR products were electrophoresed on a 3% agarose gel and visualized with ethidium bromide. The digital photograph was analyzed (Gel Plotting Macros; NIH Image ver. 1.62; National Institutes of Health, Bethesda, MD). Levels of individual PCR products were expressed as the ratio of individual product optical density to that of the internal standard. 

Intravitreal Administration of Caspase Inhibitors

Selective inhibitors of caspase-1, acetyl-tyrosyl-valyl-alanyl-aspart-1-al (Ac-YVAD-CHO), and caspase-3, acetyl-aspartyl-glutamyl-valyl-aspart-1-al (Ac-DEVD-CHO), ³⁷ were obtained from the Peptide Institute (Osaka, Japan). A 20- mM solution of these inhibitors in dimethyl sulfoxide was prepared, and 2 μl of the solution was injected into the vitreous cavity with a 30-gauge needle 24 hours before ischemic insult. Another 3 μl was injected into the vitreous at the time of reperfusion. As a control, dimethyl sulfoxide (vehicle) was injected in the same way. Twenty-four hours after reperfusion, eyes were enucleated, and TUNEL staining was performed as described earlier, because the time point for peak TUNEL staining was at 24 hours after reperfusion. Eyes that showed vitreous and subretinal hemorrhage from the intravitreal injection were omitted from the study. 

Assay of Caspase-1 and Caspase-3 Activities

Caspase family protease activities were assayed as described ^{19 38} with slight modification. Briefly, frozen stored retinas were homogenized in 10 mM Hepes-KOH (pH 7.2), 2 mM EDTA, 0.1% CHAPS, 5 mM dithiothreitol, 1 mM PMSF, 10 μg/ml pepstatin, 20μ g/ml leupeptin, and 10 μg/ml aprotinin. Homogenates were centrifuged at 15,000g for 30 minutes. Supernatants were transferred to new tubes (Eppendorf; Fremont, CA). Aliquots of extracts (300 μg protein in 100 μl extraction buffer) were preincubated at 37°C for 30 minutes and then mixed with 5 μl 10 mM tetrapeptide substrate, acetyl-tyrosyl-valyl-alanyl-aspartic acidα -(4-methyl-coumaryl-7-amide; Ac-YVAD-MCA) for caspase-1 or acetyl-aspartyl-glutamyl-valyl-aspartic acidα -(4-methyl-coumaryl-7-amide; Ac-DEVD-MCA) for caspase-3 obtained from the Peptide Institute. Free aminomethylcoumarin accumulation, which resulted from cleavage of the aspartate-aminomethylcoumarin bond, was monitored in each sample at 37°C for 180 minutes in a microcubbette using a spectrophotometer (Ultraspec III; Pharmacia, Cambridge, UK). The absorbance of the each sample at 370 nm was plotted against time. Linear regression analysis of the velocity of each curve yielded an activity for each sample. ³⁸ Data were expressed as a percentage of the caspase-family–like activity in the samples from sham-procedure retina. 

Statistical Analysis

The data were analyzed by one-way analysis of variance (ANOVA) followed by Fisher’s post hoc test. P < 0.05 was taken to be statistically significant. 

Results

Immunohistochemistry

No specific staining was detected by the caspase-1 and caspase-3 antibodies in the retina of sham-procedure rats (Fig. 1) . In contrast, there was strong staining of cells in the INL and ONL of the ischemia-injured retinas using antibodies against caspase-1 and caspase-3 (Fig. 1) . Double labeling showed diffuse cytoplasmic staining with anti-caspase-1 and anti-caspase-3 antibodies in shrunken cells that also had weak TUNEL-stained nuclei. Immunostaining for the two caspases was observed in both the nucleus and cytoplasm of the strongly TUNEL-positive cells. There were many cells that were TUNEL-positive but not immunoreactive; the cell type of these cells in both the INL and ONL was difficult to determine, because apoptotic cells showed similar morphologic characteristics. However, immunostained cells were retinal neurons, because they were not stained with anti-s-100 or anti-GFAP antibodies (graphic data not shown). 

Quantitative Analysis of Immunoreactive Cells

Quantitative analyses showed that the number of cells stained with the antibodies against caspase-1 and caspase-3 in the INL and ONL increased at 6 to 24 hours and decreased at 48 hours with a peak at 24 hours after reperfusion. The temporal profiles of TUNEL-positive cells were similar to those of caspase-1–positive cells and caspase-3–positive cells during the reperfusion period (Fig. 2) . In the INL, the number of caspase-3–positive cells was higher than the caspase-1–positive cells at 24 hours after reperfusion and during the follow-up period. Conversely, the number of cells in the ONL that reacted to the anti-caspase-1 antibody was greater than the caspase-3–positive cells during the follow-up period. 

Western Blot Analyses of Caspase-1-like and Caspase-3-like Proteases

The anti-caspase-1 antibody labeled the 20-kDa bands in the rat retina obtained 24 hours after reperfusion (Fig. 3) . Caspase-3-like protease was also expressed in the retina, evidenced by the anti-caspase-3 antibody’s blotting the 22-kDa bands. Other minor binds were not detected by the exposure time used in the study. The apparent molecular weight of the band agreed with that of the active form of both caspase-like proteases. 

The expression level of the 20-kDa subunits of caspase-1 was upregulated at 12 to 48 hours after reperfusion. In contrast, the 22-kDa subunit of the caspase-3–active form was upregulated at 6 to 48 hours after reperfusion. A peak time point of both caspases was at 24 hours after reperfusion. Both caspases were downregulated to basal level at 96 hours after reperfusion. The temporal profile of protein expression levels was similar to the pattern obtained from immunohistochemistry. 

Semiquantitative RT-PCR

Semiquantitative RT-PCR analyses showed a significant (P < 0.01 by ANOVA) upregulation of the mRNAs for caspase-1 and caspase-3 in the experimental retinas compared with that in sham-procedure rat retinas at 12 to 48 hours after reperfusion (Fig. 4) . Such mRNA expression decreased at 96 hours after reperfusion. A peak time point of the two caspases was 24 hours after reperfusion. The expression level of caspase-1 and caspase-3 was 1.48 and 2.18 times higher than that of the sham-procedure controls, respectively. 

Assay of Caspase-1 and Caspase-3 Activity

Caspase activities were assayed by using tetrapeptide substrates. As shown in Figure 5 , both protease activities were elevated in the extract from retinas at 12 to 48 hours after reperfusion. The level of the caspase-1-like activity at the peak time was two times higher than that of the sham-procedure rat retinas and was 1.5 times that in the sham control for caspase-3. Again, the time courses of activity of the two caspases showed a similar pattern in immunohistochemical analysis, protein expression, and gene expression. 

Effects of Caspase Inhibitors

Figure 6 demonstrates the number of TUNEL-positive cells in the INL and the ONL of the retina after intravitreal injection of caspase-1 and caspase-3 inhibitors. The number of TUNEL-positive cells was significantly reduced at 24 hours after reperfusion, compared with the number in sections prepared from the retina that was not treated with the inhibitors (P < 0.01 by ANOVA). In the INL, the number of TUNEL-positive cells was reduced more by the caspase-3 inhibitor than by caspase-1 inhibitor (P < 0.05 by ANOVA). In comparison, in the ONL, the apoptotic cell counts were decreased more by the caspase-1 inhibitor than by the caspase-3 inhibitor (P < 0.05 by ANOVA). 

Discussion

In this study, we demonstrated that expression of caspase-1 and caspase-3 was upregulated at the gene and protein levels after ischemia–reperfusion injury. We have also shown that caspase-1 was predominantly associated with photoreceptor apoptosis, whereas, caspase-3 was more activated in the retinal neurons in the INL than in the ONL. In addition, we found the inhibition of retinal apoptosis by intravitreal injection of caspase inhibitors. 

Caspase-1 was identified as a mammalian homologue of the ced-3 gene, ²⁰ initially identified in Caenorhabditis elegans. Ced-3 is known to promote apoptosis in the nematode during development, and currently, 11 cysteine proteases, ^{21 23 39 40} termed caspases, ²² have been cloned. These family members participate in one of two distinct signaling pathways: activation of proinflammatory cytokines and activation of apoptotic cell death. ^{39 41} All the known caspase family proteases are synthesized as inactive proenzymes that require cleavage to liberate one large and one small subunit to form the active enzyme. ³⁹ These families have been divided into three groups based on homology, ²² and we examined two distinct family members of caspase-1 and caspase-3. 

A number of experimental studies have suggested that activation of the caspase family is important in the execution of apoptotic events in neurons. ^{41 42 43 44 45 46 47 48 49} Our previous studies indicate that neuronal apoptosis and necrosis are the main pathologic mechanism in ischemia–reperfusion injury in the rat retina, based on morphologic (TUNEL staining and ultrastructural studies) and biochemical (DNA fragmentation) evidence, ^{25 26} and other reports have supported our idea. ^{27 50} Furthermore, TUNEL-positive cells in INL were double stained with an amacrine cell marker but not anti-s-100 and anti-GFAP antibodies. ^{25 26} In the present study, the same results were obtained (graphic data not shown). Thus, we conclude TUNEL-positive cells were mainly retinal neurons. Although most of the TUNEL-positive cells were neurons, we have to be careful not to misunderstand that TUNEL positivity directly addresses neuronal death. 

In the present experiment, caspase-1 and caspase-3 were colocalized with TUNEL-positive cells in the INL and ONL, and the temporal profiles of TUNEL staining and immunostaining were similar. Western blot analysis showed activation of caspase-1-like and caspase-3-like proteases during the retinal degenerative process during ischemia–reperfusion. Also, enzymatic activities of caspase were upregulated at 12 to 48 hours after reperfusion and downregulated at 96 hours. The level of these caspase mRNAs increased after reperfusion and decreased at 96 hours. Caspase-1 has been implicated in the death caused by superoxide dismutase downregulation in PC12 cells, but not by withdrawal of trophic factor support. ⁴⁶ Conversely, an antisense constructed to downregulate caspase-2 in PC12 cells inhibites cell death by withdrawal of trophic support, ⁴⁷ but does not inhibit oxidative stress. ⁴⁸ Activation of distinct caspases in the same cells can thus promote the apoptosis initiated by different stimuli. Furthermore, an in vivo study shows that caspase-1 and caspase-3 contribute to neuronal cell death in ischemic brain injury induced by occlusion of the middle cerebral artery in mice, ⁴² and caspase-3, but not caspase-1, is associated with neuronal apoptosis after fluid-percussion–induced traumatic brain injury. ⁴⁵ These studies indicate that different pathologic stimuli can activate a distinct caspase family in the same cell to induce apoptosis. 

Little is known, however, about whether different proteases in the caspase family are activated in apoptotic neurons of different cell types evoked by the same pathologic condition. ⁴⁹ The retina is an appropriate model to answer this question, because there are at least six distinct types of neurons in the retina: retinal ganglion cells, bipolar cells, amacrine cells, horizontal cells, cone photoreceptors, and rod photoreceptors. In the present study, we tried to determine whether different caspases were expressed in apoptotic neurons of the ONL and the INL induced by the ischemia–reperfusion injury. Immunohistochemically, caspase-1-immunoreactive cells in the ONL were more pronounced than those of caspase-3, but in the INL, caspase-3–reactive cells were more numerous than caspase-1 cells. When the caspase-1 inhibitor was injected into the vitreous, the percentage of apoptotic cells in the ONL was decreased significantly more than those in the INL. Conversely, when the caspase-3 inhibitor was administered, the changes in the number of apoptotic cells in the INL was significantly greater than in the ONL. Thus, apoptotic neuronal death in retina after ischemia–reperfusion is executed by two different caspases. In the INL, caspase-3 is the main executor of apoptotic neurons, and in the photoreceptors caspase-1 is predominantly expressed. We thus conclude that the neuronal death evoked in different cell types by the same cause may be mediated by distinct members of the caspase family proteases. Stereologic examination of multiple retinal layers in this model, if possible, would give more satisfactory data. However, in the present investigation such examination was not possible. 

Unfortunately, specific inhibitors of both caspase-1 and caspase-3 have not been discovered, and the inhibitors used in this study were selective rather than specific for caspase-1 and caspase-3. Ac-YVAD-CHO inhibites the enzymatic activity of caspase-4, and also, Ac-DEVD-CHO in high concentrations blocks caspase-7 reactivity. ⁵¹ Therefore, we cannot discriminate between the relationship of caspase-4 and caspase-7 in retinal ischemia–reperfusion injury. Although we have shown the importance of two distinct caspase family proteases in the INL and ONL by a single insult, we have to be careful in drawing a conclusion. Further experiments using specific rather than selective inhibitors or the use of knockout animals will be needed in the future. 

 Supported in part by a Grant-in-Aid 10470567 for Scientific Research from the Ministry of Education, Science, Sports and Culture of the Japanese Government (NY).

 Submitted for publication December 11, 1998; revised April 30, 1999; accepted June 15, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

Confocal microscopic images of TUNEL and immunostaining of rat retina at 24 hours after reperfusion. Double staining with (A) anticaspase-1 antibody and (B) TUNEL. Double staining with (C) anti-caspase-3 antibody and (D) TUNEL. These are digitally overlaid images of immunostaining (red) and TUNEL (green). Sham-procedure retina double stained with (E) caspase-1 and caspase-3 and (F) TUNEL. These are digitally overlaid images of immunostaining (red) and TUNEL (green). TUNEL-stained cells were labeled with both caspase-1 and caspase-3 antibodies. Original digital images were imported on a computer (Macintosh; Apple, Cupertino, CA), into image analysis software (Photoshop 4.0J; Adobe Systems; San Jose, CA) and then transferred to print out (Persuasion 3.0J; Adobe Systems). Bar, 50 μm. GCL, ganglion cell layer; OS, outer segment.

Figure 2.

 

The temporal profile of immunoreactivities for caspase-1 and caspase-3 in the INL (A) and the ONL (B). The number of immunostained cells was counted in 20 areas of each section with a scanning laser confocal microscope. Data are represented as findings per square millimeter. The peak time point of the number of caspase-1– and caspase-3–immunoreactive cells was at 24 hours after reperfusion. Caspase-1 immunoreactivity in the ONL was approximately two times more than that of caspase-3, but in the INL the number of caspase-3–positive cells was approximately two times more than that of caspase-1 (P < 0.01). Data are expressed as the mean ± SEM (n = 4).

Figure 3.

 

Western blot analysis of the rat retina after the ischemia–reperfusion injury. The anti-caspase-1 antibody elicited a 20-kDa band, and anti-caspase-3-antibody revealed a 22-kDa band. The apparent molecular weights of the bands agree with those of caspase-1 and caspase-3. Active forms of caspase-1 (20 kDa) and caspase-3 (22 kDa) were upregulated in the retina at 12 to 48 hours after reperfusion.

Figure 4.

 

The temporal profile of caspase-1 (A) and caspase-3 (B) mRNA expression in retinal ischemia–reperfusion. Levels of mRNA were measured by using semiquantitative RT-PCR. The optical density of the PCR products was expressed as the ratio of the β-actin amplicon of the same cDNA sample. The mRNA level is expressed as a ratio to the sham control. mRNA expression for caspase-1 and caspase-3 increased at 12 to 48 hours, with a peak at 24 hours after reperfusion. Data are expressed as the mean ± SEM (n = 4).* P < 0.05; **P < 0.01.

Figure 5.

 

Enzymatic activity of caspase assayed using the fluogenic tetrapeptide substrates. (A) Caspase-1-like proteases; (B) caspase-3-like proteases. The activities were elevated in the extract from ischemic injured rat retinas at 12 to 48 hours after reperfusion. Data are represented as the percentage of sham-procedure retinas and are expressed as the mean ± SEM (n = 4).* P < 0.01.

Figure 6.

 

Effects of intravitreal injection of caspase inhibitors on the number of TUNEL-positive cells. (A) INL; (B) ONL. These inhibitors reduced the number of the apoptotic retinal neurons. Data are expressed as the mean ± SEM (n = 4).* P < 0.05; **P < 0.01.