Calpain-Specific Proteolysis in Primate Retina: Contribution of Calpains in Cell Death

Abstract

purpose. One of the leading causes of blindness is retinal damage caused by the high intraocular pressure (IOP) in glaucoma. Previous studies in rats have suggested that the proteolytic enzyme calpain (EC 3.4.22.17) is involved in retinal cell death during ischemia and in acute high IOP. Ubiquitous, calcium-activated calpain-1 and -2 from monkey retina are highly homologous to rat calpains, although expression patterns in variants of tissue-specific calpain-3 are different between monkey and rodent retinas. Thus, the purpose of the present study was to investigate the involvement of calpain-induced proteolysis in retinal cell death in primates.

methods. Calpain involvement in a simulated pathologic condition was examined by incubating monkey retinas in hypoxic conditions (95% N₂ and 5% CO₂) in RPMI medium without glucose. Endogenous tissue calpains were also directly activated in monkey and human retinal soluble proteins by incubating with 2.5 mM calcium. The resultant proteolysis of monkey retinal proteins was assessed by 2D electrophoresis (2-DE).

results. In hypoxic retina, leakage of lactate dehydrogenase (LDH) from retinas into the medium increased, indicating cell death. LDH leakage was partially inhibited by the calpain inhibitor SJA6017. Calpain autolysis was observed, and the calpain-preferred substrate α-spectrin was proteolyzed. In retinal soluble proteins incubated with calcium, a total of 15 spots from 2-DE of retinal soluble proteins were identified by mass spectrometry. Proteolysis of major proteins, vimentin, β-tubulin, α-enolase, and Hsp70 were confirmed by immunoblot analysis. Activation of calpains and proteolysis of these substrates were inhibited by the calpain-specific inhibitor SJA6017.

conclusions. Taken together, these results suggested that calpain activation in primate retinas could play an important role in cell death during hypoxia caused by elevated IOP from glaucoma.

Calpains are a superfamily of structurally related, calcium-activated, cysteine proteases. ^{1 2 3} They include the ubiquitous calpain-1 and -2 and tissue-preferring calpains, such as calpain-3. Activation of calpains has been well documented during neuronal cell death in the ischemic brain. ^{4 5 6 7 8 9 10} Our recent in vivo studies with rats also suggest that calpain isoforms play an important role in retinal ganglion cell death induced by ischemia-reperfusion ¹¹ and by acute ocular hypertension. ¹² The calpain inhibitor SJA6017 partially protects against the loss of ganglion cells. Calpain activation and degradation of known calpain substrates has also been observed during retinal cell damage in an in vitro rat model of hypoxia. ^{13 14} However, the involvement of calpain-induced proteolysis in retinal cell death has not been studied in humans and nonhuman primates. The purposes of the present study were to clarify the role of calpains during cell death in primate retinas and to determine the protein substrates for primate calpains. 

Materials and Methods

Tissues from Human and Monkey Eyes

Eyes from monkeys were obtained from experiments unrelated to the present studies, and experimental animals were handled in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and with the Guiding Principles in the Care and Use of Animals (DHEW Publication, NIH 80-23). For incubation studies, globes were obtained from Macaca mulatta ranging in age from 7 to 15 years at the Oregon National Primate Research Center (Beaverton, OR). Human eyes ranging in age from 27 to 88 years were obtained from the Lion’s Eye Bank of Oregon, using informed consent, after enucleation for corneal transplant or as a research donation. The procurement complied with tenets of the Declaration of Helsinki for Ethical Principles for Medical Research Involving Human Subjects. The present research was approved by the Oregon Health and Science University (OHSU) institutional review board. The average time between death and dissection was 1 hour for monkey eyes and 13 hours for human eyes. 

Retinal Tissue Culture

Monkey retinas were dissected from the eyes in RPMI medium (11875; Invitrogen Corp., Carlsbad, CA) containing 100 U/mL penicillin and 100 μg/mL streptomycin. Each half retina was incubated in a 15-mL centrifuge tube with 5 mL medium for 10 hours. Control, normoxic monkey retinas were incubated in RPMI medium containing 11 mM glucose bubbled with 95% O₂ and 5% CO₂, to supply sufficient oxygen for retinal cell survival at 37°C. To induce hypoxia, retinas were incubated in RPMI medium (11879; Invitrogen) without glucose and bubbled with 95% N₂-5% CO₂ at 37°C. After termination of culture, retinas were washed with PBS and homogenized by sonication (Sonic Dismembrator; Fisher Scientific, Hampton, NH) in 20 mM Tris (pH 7.5), 5 mM EGTA, 5 mM EDTA, and 2 mM dithioerythritol (DTE). 

Measurement of LDH Activity

Leakage of lactate dehydrogenase (LDH) into the medium was used as a marker for membrane breakage and cell death as previously described. ^{15 16} After incubation of the retinal tissue was terminated, the culture medium was centrifuged at 2000 rpm for 5 minutes to remove tissue fragments, and the supernatant was subjected to LDH assay. To assess cellular LDH, the retinas were treated with 1% Triton X-100 (Sigma-Aldrich Corp., St. Louis, MO) in RPMI medium containing 1% bovine serum albumin (Sigma-Aldrich). LDH was measured with an LDH cytotoxicity detection kit (Takara Mirus Bio, Madison, WI) according to the recommendations of the manufacturer, with bovine heart LDH (Sigma-Aldrich) used as a standard. The percentage of leakage of LDH was calculated as [LDH activity in medium/total (medium + retinas)] × 100. Statistical analysis of data was performed with the Dunnett multiple-comparisons test. The significance level was P < 0.05. 

Detection of Calpain Activation

Three methods were used to detect calpain activation in the retinal soluble proteins incubated with calcium: (1) Calpain-specific, α-spectrin breakdown products (SBDPs) were detected by immunoblot analysis. The 145-kDa SBDP is produced by calpains, and the 150-kDa SBDP is produced by calpains and caspase-3. ¹⁷ (2) SJA6017 inhibition of proteolysis (e.g., inhibition of production of the 145-kDa SBDP) was used as a negative control during detection of calpain activation. SJA6017 is a potent inhibitor for calpain-1 and2. Cathepsin-B and -L are also inhibited, but they do not need calcium for activation. SJA6017 does not inhibit other cysteine proteases. ¹⁸ (3) Casein zymography and immunoblot analysis for calpains were also used to determine calpain activation. Contrary to first expectations, zymograms and immunoblots often showed a loss of the intact 80-kDa catalytic subunit of calpain, because after calpain activation, intact calpains rapidly autolyze to active calpain-1 fragments at 78 and 76 kDa (not readily detected on our zymograms) and to a 43-kDa fragment from calpain-2; and then these fragments are further degraded. ¹⁹ Thus, loss of the 80-kDa calpain band was used as indirect evidence of calpain activation. Of course, a decrease in the 80-kDa band could be due to action of other proteases. The concomitant presence of calpain-1 fragments at 78 and 76 kDa ²⁰ and at 43 kDa for calpain-2 ²¹ on immunoblots provided further evidence that calpains were activated in our studies. All experiments were performed at least three times, to confirm reproducibility. 

Immunoblotting and Zymography

SDS-PAGE of soluble proteins was performed on 10% bis-Tris gels (NuPAGE; Invitrogen) with the MOPS (3-(N-morpholino)propanesulfonic acid) buffer system (Invitrogen) for calpain-1 and -2, vimentin, and enolase; and 4% to 12% gels with the MOPS buffer system for α-spectrin, Hsp70, and β-tubulin. Immunoblots were performed by electrotransferring proteins from polyacrylamide gels (NuPAGE; Invitrogen) onto polyvinylidene fluoride (PVDF) membranes (Millipore, Bedford, MA) at 100 V (constant) for 60 minutes at an ice-cold temperature in transfer buffer (NuPAGE; Invitrogen). Antibody sources and dilutions were: rabbit polyclonal antibodies against 1:250 calpain-2 ²² , enolase (1:200; Santa Cruz Biotechnology, Santa Cruz, CA), and mouse monoclonal antibody against calpain-1 (1:500; Affinity Bio Reagents, Golden, CO), α-spectrin (1:1000, nonerythroid; Affiniti Research Product, Exeter, UK), vimentin (1:100; Santa Cruz Biotechnology), Hsp70 (1:100; Santa Cruz Biotechnology), and β-tubulin (1:100; Santa Cruz Biotechnology). Immunoreactivity was visualized with alkaline phosphatase conjugated to anti-rabbit or mouse IgG secondary antibody and BCIP/NBT (5-bromo-4-chloro-3-indoyl phosphate-nitroblue tetrazolium; Sigma-Aldrich). 

Casein zymography was performed according to the method of Raser et al. ²³ Ten percent native gels, copolymerized with 0.1% casein were prerun with buffer containing 25 mM Tris (pH 8.3), 192 mM glycine, 1 mM EGTA, and 1 mM dithiothreitol (DTT) for 15 minutes at 4°C. Thirty micrograms of retinal soluble proteins were loaded and run. After electrophoresis, the gels were incubated with slow shaking overnight at room temperature in 20 mM Tris (pH 7.4), 10 mM DTT, and 2 mM CaCl₂. Gels were stained with a Coomassie G-250 stain (SimplyBlue Safestain; Invitrogen) for 1 hour, and destained with distilled water for 1 hour. For presentation, digital images of the zymography gels were inverted to visualize caseinolysis caused by calpains. Note that SJA6017, used in some of the experiments, is a reversible inhibitor of calpain. This allowed caseinolytic activity to be observed on the zymography gels since, although SJA6017 was added to the preincubation mixtures, SJA6017 was subsequently removed by electrophoresis during zymography. 

Activation of Calpain in the Soluble Proteins from Monkey and Human Retinas by Calcium

Monkey and human retinas were dissected from the eyes and homogenized by sonication (Sonic Dismembrator; Fisher Scientific, Pittsburgh, PA) in buffer containing 20 mM Tris (pH 7.5), 5 mM EGTA, 5 mM EDTA, and 2 mM DTE. Soluble proteins were obtained by centrifugation at 13,000g for 15 minutes. Protein concentrations were measured by the bicinchoninic acid (BCA) assay (Pierce, Rockford, IL), with bovine serum albumin used as the standard. All steps were performed at 4°C. 

Soluble proteins were incubated with or without calcium. Each 500-μL reaction mixture contained 20 mM Tris (pH 7.5), 1 mM DTE, 2.5 mg retinal proteins, and 2.5 mM CaCl₂. Proteolysis was initiated by addition of CaCl₂, allowed to proceed for 2 hours at 37°C, and terminated by the addition of EGTA at 3 mM final concentration. 

2-D Gel Electrophoresis

The first-dimension isoelectric focusing (IEF) for 2-D gel electrophoresis (2-DE) was performed on 11-cm nonlinear pH 3 to 10 immobilized pH gradient strips (ReadyStrip IPG strips; Bio-Rad Laboratories, Hercules, CA). IPG strips were rehydrated in the first-dimension buffer (8 M urea, 2% CHAPS [3-[3-cholamidopropyl]dimethylammonio-2-hydroxy-1-propanesulfonate], 0.5% IPG buffer, 1% dithiothreitol [DTT], with a trace of bromophenol blue), and 200 μg sample. IEF was performed using a rapid ramp up to 8000 V for 10 to 12 hours to give a total of 35,000 V-hours. The current limit was set to 50 μA per strip. The equilibration of each strip was performed in 5 mL SDS buffer (50 mM Tris-Cl [pH 8.8], 6 M urea, 30% vol/vol glycerol, 2% SDS, and trace bromophenol blue) with 2% DTT, followed by 5 mL SDS buffer with 2.5% iodoacetamide. Next, 2-DE was performed on a 12% Bis-Tris gel (Criterion XT; Bio-Rad), with the first-dimension IPG strip embedded at the top in 1% agarose. Proteins were visualized on the gels by Coomassie blue staining (SimplyBlue Safestain; Invitrogen). 2-DE was performed in triplicate for each experiment. Three gel images for each experiment were overlaid gel analysis software (Phoretix 2D Evolution; Nonlinear USA, Inc., Durham, NC) and only consistently obvious changes were chosen for protein identification. 

Protein Identification

Blue-stained proteins resolved on 2-DE gels were excised manually or with a sample-preparation robot (2DiDx; Leap Technologies, Carborro, NC). Gel plugs (1.5 mm) were destained in two changes of 100 mM ammonium bicarbonate (ambic) in 30% methanol. The plugs were dried with acetonitrile, reduced with 10 mM DTT, and alkylated with 50 mM iodoacetamide before being washed and dried again. Approximately 100 ng sequencing-grade trypsin (Sequencing Grade Modified Trypsin; Promega, Madison, WI) in 20 mM ambic were added to each sample along with sufficient buffer to immerse the gel plugs completely. Digestion proceeded overnight at 37°C. In preparation for MALDI (matrix assisted laser desorption ionization mass spectrometry) analysis, 1 μL neat formic acid was added to each sample. MALDI samples were prepared by deposition of 1 μL tryptic digest and 1 μL matrix (10 mg/mL α-cyano-4-hydroxycinnamic acid dissolved in 50% acetonitrile, 0.1% formic acid, and 10 mM ammonium phosphate) on a stainless steel target. 

Spectra for peptide mass fingerprinting were collected on a mass spectrometer (QstarXL; Applied Biosystems, Foster City, CA) equipped with an orthogonal (o)MALDI source. Data from approximately 1000 laser shots (14 μJ) were collected over the range of m/z 800 to 3200. Before data collection, external two-point calibration was performed with angiotensin II (M+H⁺, m/z 1046.5423) and ACTH fragment 18 to 39 (M+H⁺, m/z 2465.1989). After data acquisition, internal calibration of data was performed using the trypsin autolysis peaks (M+H⁺, m/z 842.5099 and 2211.1045), if observed. 

Peptide monoisotopic masses for database searching were generated (Distiller; Matrix Science, London, UK) and submitted to a search engine (Mascot ver. 1.9; Matrix Science) for protein identification. Masses were searched against a nonredundant protein database (nrDB, ver. 20040916; National Center for Biotechnology Information, Bethesda, MD) with the following parameters; taxonomy: metazoa; fixed modifications: cysteine carbamidomethylation; variable modifications: N/Q deamidation, methionine oxidation, one missed cleavage allowed, and mass tolerance of 0.02 Da. 

For samples in which protein identification was not achieved at 99.9% confidence based on peptide mass fingerprinting, MALDI/tandem mass spectrometry (MS/MS) data were collected to enhance confidence. The three most abundant ions (excluding known trypsin autolysis and keratin peptides) in each MS spectrum were sequentially selected, and CID spectra were collected under manual control. Data from approximately 500 to 2000 laser shots (16 μJ) were collected over the range of m/z from a low of 50 to a high of 50 amu (atomic mass units) above the precursor mass. Collision energies were varied during data collection to generate high-quality spectra. As above, peak lists were generated (Distiller; Matrix Science) and submitted to the computer program for protein identity confirmation using the same database as for peptide mass fingerprinting. Parameters were similar to peptide mass finger printing with the exception that mass tolerances were set at 0.1 Da for both precursor and fragment ions. In several cases, MS/MS data were also submitted to Protein Prospector (http://prospector, ucsf.edu/ provided in the public domain by the University of California, San Francisco) for identification. In each case, the data confirmed the primary identification made during peptide mass fingerprinting. One sample (spot 89) was identified by liquid chromatography (LC)/MS/MS with an ion-trap mass spectrometer. Peptide digests were performed similarly, and MS/MS data were submitted to the protein search engine (Mascot; Matrix Science) using parameters similar to those just described, with the exception of mass tolerances of 2.0 Da and 0.4 Da for precursor and fragment ions, respectively. 

Results

Retinal Tissue Culture under Hypoxic Conditions

Monkey retinas were excised and incubated under 95% N₂ and 5% CO₂ to induce ischemia and elevate calcium, and breakdown of the signature calpain substrate α-spectrin was measured. Ten hours after incubation, a dense band for the calpain-specific fragment of α-spectrin appeared at 145 kDa (Fig. 1A , lane 3, open arrowhead). Calpain inhibitor SJA6017 prevented production of the α-spectrin breakdown product (lane 4). 

The active, autolytic fragments of calpain-1 at 78 and 76 kDa demonstrated in human platelets ²⁰ (Fig. 1B , lane 3, open arrowhead) were noted on the immunoblots of the retinas cultured under hypoxia. It is appropriate to assume that calpain-1 autolytic bands from the human and monkey are very similar in molecular weight, because monkey calpain-1 is 98% homologous to human calpain-1. ²⁴ The intact 80-kDa subunit of calpain-2 decreased under hypoxia (Fig. 1C , lane 3). These autolytic losses in calpain-1 and -2 proteins also resulted in losses in the enzymatic activity for calpain-1 (Fig. 1D , lane 3, striped arrow) and for calpain-2 (lane 3, solid arrow). The electrophoretic migration positions of calpain-1 and -2 were confirmed by immunoblotting from the native gels without casein with antibodies against each calpain (data not shown). Addition of calpain inhibitor SJA6017 to the hypoxic culture medium prevented autodegradation of calpain-1 and -2 proteins (Figs. 1B 1C , lane 4) and inhibited the loss of calpain-1 and -2 activities (Fig. 1D , lane 4). 

These data suggest that hypoxia induces cell damage and calcium influx, leading to calpain-induced proteolysis. Indeed, leakage of retinal cytoplasmic LDH into the medium (a measure of cell damage) from hypoxic retinas was 61.5% ± 13.1% compared with that from control normoxic retinas of 11.6% ± 6.2% (Fig. 1E) . The calpain inhibitor SJA6017 caused a statistically significant partial reduction in LDH leakage due to hypoxia (Fig. 1E) . The modest protection by SJA6017 suggests that calpains may contribute to retinal cell death in a limited fashion during hypoxic condition or that the inhibitor does not effectively penetrate the cultured retinas. 

Activation of Calpains in Isolated Monkey and Human Retinal Soluble Proteins by Exogenous Calcium

Incubation with calcium was used to test whether the endogenous levels of calpain-1 and -2 in monkey and human retinal soluble proteins were high enough to proteolyze the calpain substrate α-spectrin. Two hours after addition of calcium, endogenous α-spectrin at 280 kDa decreased (Fig. 2A , lane 3, solid arrow). Dense bands for spectrin breakdown products appeared at 150 and 145 kDa (Fig. 2A , lane 3, open arrowheads). The 145-kDa fragment is specifically produced by calpains. ¹⁷ α-Spectrin proteolysis was also calcium dependent (Fig. 2A , lane 3) and inhibited by the calpain inhibitor SJA6017 (Fig. 2A , lane 4). 

The addition of calcium to monkey retinal soluble proteins decreased the amount of intact 80-kDa catalytic subunit for calpain-1 and -2 (Figs. 2B 2C , lane 3). Further, activated calpain-1 protein fragments at 78 and 76 kDa were found in the monkey retinal soluble proteins incubated with calcium (Fig. 2B , lane 3, open arrowhead). Calcium also produced a 43-kDa fragment from calpain-2 (Fig. 2C , lane 3). This fragment is generated by autodegradative loss of the C-terminal half of the 80-kDa calpain-2 subunit in rats and is associated with activation of calpain-2. ²¹ In soluble monkey retinal proteins, production of the 76/78 and 43 autolytic fragments from calpain-1 and -2, respectively, was also calcium dependent (Figs. 2B 2C , lane 3) and prevented by the calpain inhibitor SJA6017 (lane 4). These experiments were repeated with the soluble proteins from adult human retinas with nearly identical results (Figs. 2D 2E 2F) . Thus, the data from the culture experiments and from calcium-incubated total soluble proteins showed that the levels of endogenous calpain-1 and -2 in soluble retina proteins were high enough in monkey to become active when calcium was elevated. 

Calpain Substrates in Monkey Retina

To detect other potential substrates for activated calpains, 2-DE proteomic maps were made for monkey soluble retinal proteins incubated without (Fig. 3A)and with added calcium (Fig. 3B) . Normal retinas showed numerous protein spots within the molecular weight range of 20 to 80 kDa (Fig. 3A) . Also, there were several horizontal trails of spots at very similar molecular weights but differing PIs. These probably resulted from increasing amounts of posttranslational changes, such as multiple phosphorylations or deamidations to the same parent proteins in these mature monkey retinas. Incubation with calcium caused changes in 21 proteins spots from 15 different proteins that were identified by MS (Table 1) . Spots showing major decreases after calcium are circled in Figure 3A . Proteins showing increased fragments (e.g., vimentin fragments) or modifications after calcium are circled in Figure 3B . Calpain breakdown of vimentin, β-tubulin, enolase, and heat shock protein 70 in vitro was confirmed by immunoblot with substrate-specific antibodies and inhibition of proteolysis by SJA6017 in both monkey and human retinas (Fig. 3C) . Such in vitro data provide tentative assignment of these proteins as potential substrates for calpain in live primate retinas. Note that we have not yet been able to confirm that vimentin, β-tubulin, enolase, and heat shock protein 70 are in fact substrates in calpain-activated, hypoxic retinas. Different results using cultured retinas and isolated total retinal proteins were expected because of the homogenization of mixed cells from different layers. 

Discussion

Calpain Activation in Retinal Soluble Proteins

One of the important findings of the present study was that the endogenous levels of activated calpain-1 and -2 in human and monkey retinal soluble proteins were high enough to proteolyze retinal proteins when the calcium was artificially elevated by incubation with 2.5 mM calcium (Fig. 2) . Six biochemical findings confirmed this conclusion: (1) A calpain-specific, α-spectrin breakdown product at 145 kDa was detected after addition of calcium. (2) The calpain inhibitor SJA6017 prevented production of the 145-kDa SBDP and prevented autolysis of calpains. (3) Production of autolyzed fragments at 78 and 76 kDa for calpain-1 and the 43-kDa autolytic fragment for calpain-2 were detected, along with loss of the intact 80-kDa catalytic subunit. (4) Activation of calpains and proteolysis of α-spectrin were calcium dependent—a well-known property of calpains. (5) Numerous other proteins were shown by proteomics analysis to undergo calcium-dependent proteolysis. One of these, vimentin, is well known to be rapidly hydrolyzed by calpains in other tissues. ²⁴ (6) In addition to the ubiquitous calpains, monkey, and human retinas express a variety of calpain 3 splice variants at lower levels. ²⁵ However, only calpain-1 and -2 were detected on zymography of retinal extracts, suggesting that the major calpains activated by calcium in human and monkey retina were the ubiquitous calpain-1 and -2. 

The ability of calcium to cause activation of endogenous retinal calpains was important to determine because almost all tissues, including retina, ²⁶ contain a potent endogenous inhibitor of calpain-1 and -2 called calpastatin, containing four inhibitory domains. In some tissues, such as aged human lenses, the presence of a high ratio of calpastatin to low levels of endogenous calpains resulted in only minimal calpain activation, even when large amounts of exogenous calcium were added to the lens soluble proteins. ²⁷ The ratios of calpastatin to calpain mRNA are 1.1 and 0.6 in human and monkey retinas, respectively. ²⁸ Our data indicated that human and monkey retinas were apparently able to “escape” the action of calpastatin when calcium was highly elevated. 

Further, various calpain activation mechanisms, including translocation to phospholipid activators such PIP₂ at the membrane, have been invoked to explain calpain activation under the calcium levels found in normal and pathologic tissues. ²⁹ The present in vitro studies with calcium showed that calpain activation occurred even without intact retinal layers and membranes. 

Retinal Tissue Culture

Our in vitro assay measured only the maximum potential calpain activity at saturating calcium levels. Elevated calcium levels have been observed in some human retinal diseases, such as in Müller glial cells from a patient with proliferative diabetic retinopathy ³⁰ and in drusen from the optic nerve head. ³¹ Furthermore, increased retinal calcium and calpain activation were observed in a rat ischemia–reperfusion model, ¹¹ in the WBN/Kob rat model, ³² and in the rat acute high ocular pressure model. ¹² The present study also provided initial data on calpain activation in a model of retinal ischemia in monkey tissue cultured retinas maintained under hypoxic conditions. In this model, strong evidence of calpain-induced proteolysis was observed: production of calpain-specific SBDP at 145 kDa, production of calpain-1 and -2 autolytic fragments, concomitant loss of calpain-1 and -2 activities, and massive cell damage (LDH leakage). These markers of calpain activity were calcium-dependent and reduced by calpain-inhibitor SJA6017. These data indicate that calpains in tissue-cultured retina are activated under hypoxic conditions. 

At the end of the 10 hours of tissue culture, we observed that the retinas in all groups became thin fragile pieces probably by mechanical action of the bubbling O₂ and N₂ gases. Although this adds physical trauma along with hypoxia as a potential initiating factor in the model, biochemical parameters, such as lack of appreciable α-spectrin breakdown and lack of calpain-1 and -2 activation, were similar in the cultured normoxic controls to those in fresh, noncultured retinas (Figs. 1A 1B 1C 1D , lanes 1 and 2). Further, only moderate levels of LDH leaked from normoxic retinas into the culture medium (Fig. 1E , O₂). These data validated our tissue culture model as a reasonable protocol for demonstrating involvement of calpain in neuronal cell damage in hypoxic monkey retina. Our recent studies with rats cultured under hypoxic conditions using the same protocol led to a similar conclusion. ^{13 14} In addition, ischemia–reperfusion and acute ocular hypertension in live rats caused activation of calpains and retinal ganglion cell damage. ^{11 12} However, our retinal hypoxic incubation experiments still leave the problem of determining which specific cells or retinal layers show increased activated calpain. Further in vivo studies will help in understanding which specific retinal cells show calpain activation. 

Administration of the calpain inhibitor SJA6017 reduced retinal cell damage. In monkeys, chronic ocular hypertension over a 20-month period leads to retinal ganglion cell damage (Oka T et al., unpublished data, 2005). These retinas showed the hallmarks of glaucoma, such as disc cupping in the optic nerve head and destruction of the lamina cribrosa. Retinal ganglion cell damage was also found in human glaucoma. ³³ Because our in vitro studies showed calpain activation in monkey and human retinas, calpain may be a factor in retinal cell death in humans, such as occurs in glaucoma, although further studies are needed for verification. Finding partial inhibition of calpains by the calpain inhibitor SJA6017 was provocative because SJA6017 may be a candidate for testing in drug therapy against retinal damage from glaucoma. 

 Supported by Senju Pharmaceutical Co., Ltd.

 Submitted for publication May 25, 2006; revised July 19, 2006; accepted October 10, 2006.

 Disclosure: E. Nakajima, Senju Pharmaceutical Co., Ltd. (E); L.L. David, Senju Pharmaceutical Co., Ltd. (F); C. Bystrom, Senju Pharmaceutical Co., Ltd. (F); T.R. Shearer, Senju Pharmaceutical Co., Ltd. (C, F); M. Azuma, Senju Pharmaceutical Co., Ltd. (E)

 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.

 

(A) Immunoblot for α-spectrin in monkey retinas incubated for 10 hours in normoxia (O₂), hypoxia (N₂), or N₂ and 100 μM calpain inhibitor SJA6017 (SJA; 15 μg soluble protein per lane). Filled arrow: intact α-spectrin band at 280 kDa; open arrowheads: 150- and 145-kDa spectrin breakdown products. Lane 1: fresh retinas (initial). (B) Immunoblot for the intact calpain-1 catalytic subunit at 80 kDa (hatched arrow) and its autolysis fragments at 78 and 76 kDa (open arrowheads; 15 μg soluble protein per lane). (C) Immunoblots for calpain-2 at 80 kDa (filled arrow; 30 μg soluble protein per lane). (D) Casein zymogram (image inverted) showing activated calpain in monkey retinal total soluble proteins after hypoxic treatment of cultured retinas (30 μg soluble protein per lane). Hatched arrow: indicates activated calpain-1; filled arrow: calpain-2. (E) LDH leakage into the medium from incubated retinas. Data are the mean ± SD (n = 8). *P < 0.05, relative to hypoxia.

Figure 2.

 

(A) Representative immunoblot for α-spectrin in monkey retinal soluble proteins incubated with calcium (15 μg proteins per lane). Filled arrow: intact α-spectrin band at 280 kDa; open arrowheads: 150- and 145-kDa spectrin breakdown products. Lanes indicate soluble protein from fresh retinas (initial), proteins incubated for 2 hours without Ca(−), with Ca(+), or with Ca(+) and 100 μM SJA6017 (SJA). (B) Immunoblot for calpain-1 (15 μg proteins per lane). Hatched arrow: the intact calpain-1 band at 80 kDa; open arrowheads: 78- and 76-kDa fragments caused by calpain-1 autolysis. (C) Immunoblot for calpain-2 (30 μg proteins per lane). Filled arrow: intact calpain-2 band at 80 kDa; open arrowhead: autolysis product at 43 kDa. (D–F) the same as (A–C), except that samples were from human soluble retinal proteins incubated with calcium.

Figure 3.

 

Two-dimensional electrophoresis of monkey total retinal soluble proteins incubated for 2 hours without (A) and with (B) calcium (200 μg per Coomassie blue-stained gel). Protein spots were excised and identified by mass spectroscopy (Table 1) . (A, dashed circles) Proteins showing decreases after incubation with calcium; (B) proteins with increased fragments or protein modifications. (C) Immunoblots for vimentin, β-tubulin, enolase, and Hsp70 in monkey retinal soluble proteins (5 μg per lane). Filled arrows: intact proteins; open arrowheads: fragments.

Table 1.

 

Protein Identities for Spots Numbered in Figures 3A and 3B

Spot No.	Protein*	Symbol	Coverage (%)	Accession No.	Molecular Mass Calculated (kDa)	Mascot Score	Predicted Function
24	Hsp70 ATPase domain	Hsp70	46	1S3XA	71.6	84	Chaperone
38	Pyruvate kinase	PK	33	AAQ15274	65.7	82	Glycolysis
48	Transferrin	TF	14	AAAP5055	65.6	50	Iron uptake
62	Collapsin response mediator protein 2	CRMP2	56	NP_001377	59.3	149	Tubulin-binding
75	Vimentin, †	Vim	51	AAH66956	53.7	139	Cytoskeleton
89	α-enolase, †	Eno	59	P06733	47.0	242	Glycolysis
116	Pyruvate kinase	PK	36	AAQ15274	47.1	101	Glycolysis
148	Guanine nucleotide binding protein (G protein)	GNB1	46	AAP35969	37.6	114	Signal transduction
168	Acidic leucine-rich nuclear phospho protein	ANP32A	46	vAAH00608	32.2	43	Inhibitor protein of protein phosphatase 2A
354	β-tubulin 2	βTub	50	AAH24038	36.3	116	Cytoskeleton
358	β-tubulin, †	βTub	37	CAG46756	15.7	67	Cytoskeleton
768	Hsp70 protein 8 isoform 1, †	Hsp70	43	NP_006588	71.9	127	Chaperone
781	Pyruvate kinase	PK	34	S64635	65.1	78	Glycolysis
793	ATP synthase	ATPsyn	25	NP_001001935	60.0	52	Energetic metabolism
815	Vimentin	Vim	38	AAH66956	53.3	93	Cytoskeleton
938	Guanylate kinase 1	GUK	62	AAH07369	19.3	72	GMP kinase
1226	Calmodulin 3	CALM3	50	NP_005175	24.5	50	Calcium binding
1234	Vimentin	Vim	28	AAA61279	21.9	57	Cytoskeleton
1446	Vimentin	Vim	41	AAA61279	49.0	101	Cytoskeleton
1467	Vimentin	Vim	40	AAA61279	39.0	78	Cytoskeleton
1474	Vimentin	Vim	47	AAA61279	38.8	177	Cytoskeleton

*  Protein identification for all spots was confirmed by either peptide mass finger printing or MS/MS.

†  Also confirmed by immunoblott, and proteolysis was inhibited by calpain inhibitor SJA6017 in both monkey and human retinal soluble proteins.