Proc. Natl. Acad. Sci. USA Vol. 92, pp. 7662-7666, August 1995 Medical Sciences

The role of in hypoxia-induced rat renal proximal tubular inJury (/calcium/) CHARLES L. EDELSTEIN, ERIC D. WIEDER, MUHAMMAD M. YAQOOB, PATRICIA E. GENGARO, THOMAS J. BURKE, RAPHAEL A. NEMENOFF, AND ROBERT W. SCHRIER* Department of Medicine, University of Colorado School of Medicine, Denver, CO 80262 Communicated by Carl Gottschalk The University of North Carolina, Chapel Hill, NC, April 18, 1995 (received for review February 6, 1995)

ABSTRACT The role of the lysosomal proteases cathep- specific lysosomotropic detergent C12-imidazole demon- sins B and L and the calcium-dependent cytosolic strated that killing can be caused by activation and/or calpain in hypoxia-induced renal proximal tubular injury was release into the cytoplasm of cysteine proteases (6). Studies of investigated. As compared to normoxic tubules, B cysteine proteases and proximal renal tubular damage have and L activity, evaluated by the specific fluorescent substrate been confined to cyclosporine. In cultured tubules. exposed benzyloxycarbonyl-L-phenylalanyl-L-arginine-7-amido-4- acutely to cyclosporine, calpain activity increased, but cathep- methylcoumarin, was not increased in hypoxic tubules or the sin activity was not increased and were intact; these medium used for incubation ofhypoxic tubules in spite ofhigh results suggested that calpain but not cathepsins are involved lactate dehydrogenase (LDH) release into the medium during in acute cyclosporine-induced renal tubular injury (7). Thus, hypoxia. These data in rat proximal tubules suggest that using freshly isolated rat proximal tubules, the aims of the cathepsins are not released from lysosomes and do not gain present study were to address the following questions: access to the medium during hypoxia. An assay for calpain (i) Do the lysosomal proteases (cathepsins) play a role in this activity in isolated proximal tubules using the fluorescent hypoxic injury? substrate N-succinyl-Leu-Tyr-7-amido-4-methylcoumarin (ii) To what extent does hypoxia increase calpain activity in was developed. The calcium ionophore ionomycin induced a renal proximal tubules? dose-dependent increase in calpain activity. This increase in (iii) Do cell permeable cysteine protease inhibitors protect calpain activity occurred prior to cell membrane damage as against hypoxia-induced renal proximal tubular injury by assessed by LDH release. Tubular calpain activity increased inhibiting calpain and/or cathepsin activity in isolated proxi- signifi'cantly by 7.5 min of hypoxia, before there was signifi- mal tubules? cant LDH release, and further increased during 20 min of hypoxia. The cysteine protease inhibitor N-benzyloxycar- MATERIALS AND METHODS bonyl-Val-Phe methyl ester (CBZ) markedly decreased LDH release after 20 min of hypoxia and completely prevented the Reagents. Digitonin, N-succinyl-Leu-Tyr-7-amido-4-methyl- increase in calpain activity during hypoxia. The increase in coumarin (N-succinyl-Leu-Tyr-AMC), hyaluronidase type III calpain activity during hypoxia and the inhibitor studies with from sheep testes, polyoxyethylene 23 lauryl ether (Brij 35 CBZ therefore supported a role for calpain as a mediator of solution), and N-benzyloxycarbonyl-Val-Phe methyl ester hypoxia-induced proximal tubular injury. (CBZ) were purchased from Sigma. (2S,3S)-trans-Epoxysuc- cinyl-L-leucylamido-3-methylbutane ethyl ester (E-64-d) was provided by Taisho Pharmaceutical, Tokyo. Benzyloxycar- The mechanisms responsible for hypoxia-induced tubular ep- bonyl-L-phenylalanyl-L-arginine-7-amido-4-methylcoumarin ithelial cell injury and death are controversial (1). Intracellular (Z-Phe-Arg-AMC) and AMC were made by Peptide Institute, calcium has been suggested to be important in mediation of Osaka, and obtained from Peptides International. Ionomycin this hypoxic injury in renal proximal tubules (2). Our labora- was purchased from Calbiochem; Pluronic F-127 was from tory recently reported that cytosolic free Ca2+ [(Ca2+)i] is Molecular Probes; collagenase type B (lot number OFAA133) significantly increased by 2 min of hypoxia in proximal tubules, was from Boehringer Mannheim; and fatty acid-free bovine and that (Ca2+)i levels at 10 min of hypoxia correlated with albumin was from ICN. subsequent damage at 20 min (3). Thus, prompt increases in Preparation of Tubules. Proximal tubules were isolated (Ca2+), may play an initiating role in hypoxic injury. This from kidney cortex ofmale Sprague-Dawley rats (bodyweight, conclusion was supported by the protection afforded by pre- 200-300 g) using collagenase digestion and Percoll centrifu- vention of the increase in (Ca2+), (3, 4). The mechanisms gation (8, 9), and hypoxia was created as described (10). whereby increases in (Ca2+), lead to cell membrane injury, Measurement of Lactate Dehydrogenase (LDH) Release however, remain to be defined. An increase in (Ca2+)i may and . LDH release was measured to evaluate cell activate calcium-dependent , which could provide a damage as described (9, 11). The percentage of LDH released mechanism for cell injury. The increase in (Ca2+)i during from tubules was calculated by determining the ratio of LDH hypoxia may activate phospholipase A2 and thus be partly in the supernatant compared to that in the lysed tubule pellet responsible for phospholipid degradation in membranes ob- plus the supernate. Protein was measured by the Lowry served during hypoxia (5). Another potential mechanism for method with bovine serum albumin used as the standard (12). calcium-dependent cell injury is activation of the calcium- dependent cytosolic protease calpain. In certain circum- Abbreviations: N-succinyl-Leu-Tyr-AMC, N-succinyl-Leu-Tyr-7- stances, proteases have been shown to play a role in cellular amido-4-methylcoumarin; Z-Phe-Arg-AMC, benzyloxycarbonyl-L- injury. For example, studies with cultured fibroblasts and the phenylalanyl-L-arginine-7-amido-4-methylcoumarin; LDH, lactate de- hydrogenase; CBZ, N-benzyloxycarbonyl-Val-Phe methyl ester; (Ca2+)i, cytosolic free calcium; E-64-d, (2S,3S)-trans-epoxysuccinyl- The publication costs of this article were defrayed in part by page charge L-leucylamido-3-methylbutane ethyl ester; DMSO, dimethyl sulfoxide; payment. This article must therefore be hereby marked "advertisement" in NS, not statistically significant. accordance with 18 U.S.C. §1734 solely to indicate this fact. *To whom reprint requests should be addressed. 7662 Downloaded by guest on September 23, 2021 Medical Sciences: Edelstein et al. Proc. Natl. Acad. Sci. USA 92 (1995) 7663

Cathepsin Assay. The cathepsin assay used in this study is separated by centrifugation. The calpain assay was then based on the assay for developed by Barret and performed on this supernatant (cytosolic extract) as follows: Kirschke (13) and the assay developed by Olbricht et al. (14) 0.5 ml of supernatant was preincubated with imidazole for cathepsins B and L in isolated tubule segments. Z-Phe- buffer with or without 5 mM CaCl2 for 10 min at 37°C in a Arg-AMC was used as a specific substrate for cathepsins B and shaking water bath. After 10 min of preincubation, 10 ,ul of L (15). Cathepsins B and L cleave the substrate and release the substrate N-succinyl-Leu-Tyr-AMC was added. The total AMC, which is highly fluorescent at 380-nm excitation and volume of the assay solution was made up to 2 ml with the 460-nm emission. imidazole buffer. Incubation was continued for a further 30 The tubules were homogenized in a 10 mM phosphate buffer min at 37°C in the shaking water bath. In the assay per- (pH 7.4) containing 10 mM Na2HPO4 and 10 mM NaH2PO4. formed without CaC12, the imidazole-HCl buffer containing The incubation solution contained 11.5 mM Na2HPO4, 55.2 1 mM EDTA and 10 mM EGTA was used. After the 30-min mM KH2PO4, and 4 mM EDTA; it also contained 0.2% Triton incubation, fluorescence at 380-nm excitation and 460-nm X-100 and 0.05% bovine serum albumin. The pH of the emission was determined. Calpain activity was determined as incubation solution was 6.0. Cathepsins B and L have an the difference between the calcium-dependent fluorescence optimum pH for proteolytic activity of 5.5-6.0. They are and the non-calcium-dependent fluorescence. An AMC stan- irreversibly inactivated at pH > 7 (13). The stop solution dard curve was determined for each experiment. Calpain contained 100 mM iodoacetate in a buffer containing 30 mM activity was expressed in pmol of AMC released per min of sodium acetate and 70 mM acetic acid. The pH was 4.75. A 10 incubation time per mg of tubule protein. mM stock solution of the substrate Z-Phe-Arg-AMC was The balpain substrate used in our assay, N-succinyl-Leu- prepared in dimethyl sulfoxide (DMSO). The stock solution Tyr-AMC, has been shown to be proteolyzed in vitro by the was freshly diluted daily to a 1 mM solution with 0.1% Brij cathepsin (16). Therefore, to ensure the calpain selec- solution. Cysteine (8 mM) was added to the 1 mM stock tivity of our assay and exclude any possible effect of lysosomal solution just before use. leakage of cathepsins in our calpain assay, the following After 15 or 25 min of hypoxia, 1 ml of tubule suspension was measures were taken: (i) digitonin, which releases cytoplasmic separated into a pellet and supernatant by centrifugation. The contents while keeping lysosomes intact, thus limiting release tubule supernatant was kept on ice. The pellet was sonicated of lysosomal cathepsins, was used to lyse the cells; (ii) the assay for 30 sec in 4 ml of the phosphate buffer to release lysosomal was perfotmed at pH 7.3 at which cathepsins are known to be enzymes and then centrifuged at 10,000 x g for 10 min to inactivated (13); and (iii) calpain activity was defined and remove fragments of cell membranes. The cathepsin assay was calculated as calcium-dependent activity, thus excluding ca- then performed on the tubule supernatant and the tubule thepsin activity, which is strictly calcium independent (13). homogenate. Cysteine Protease Inhibitor Studies. E-64-d, an oxirane The assay itself was performed as follows: 8 ,ul of tubule inhibitor, and CBZ, a peptide aldehyde inhibitor, are chemi- homogenate or 48 ,ul of tubule supernatant was added to 100 cally dissimilar, noncharged, membrane-permeable inhibitors Al of the incubation solution. After 10 min of preincubation in of the cysteine proteases calpain and the cathepsins (18). A a shaking water bath at 370C, 50 ,u1 of the substrate was added. stock solution of 100 mM E-64-d inhibitor in DMSO was used. The incubation was continued for a further 30 min at 37°C in The solution was stored at -20°C. A solution of 100 mM CBZ a shaking water bath. At the end of the incubation period, 2 ml in DMSO and Pluronic F-127 (50 mg/ml) in DMSO was of the stop solution was added to stop the reaction. freshly prepared before use. Fluorescence at 380-nm excitation and 460-nm emission was After the recovery period, the tubules were preincubated determined with a Hitachi F2000 spectrophotometer. An with the inhibitor for 10 min at 37°C in the shaking water bath. AMC standard curve was determined for each experiment. Statistical Analysis. Normally distributed and nonnormally Activity of cathepsins B and L in the tubules was expressed in distributed data were analyzed by the paired Student t test and nmol of AMC released per min of incubation time per mg of the Wilcoxon rank sum test, respectively. Multiple group tubule protein. comparisons were done by ANOVA with the posttest accord- Calpain Assay. The calpain assay used in this study is based ing to Newman-Keuls. A P value of <0.05 was considered on that described by Sasaki et al. (16) for purified porcine statistically significant. Values are expressed as means ± SEM. kidney calpain. N-Succinyl-Leu-Tyr-AMC was used as a sus- ceptible substrate for calpain (16). A stock solution of 25 RESULTS mg/ml was prepared in DMSO and was stored at -20°C. The imidazole-HCl buffer used contained 63.2 mM imida- Cathepsin Activity After 15 and 25 min of Hypoxia. Tubule zole and 10 mM 2-mercaptoethanol (pH 7.3). After exposure cathepsin activity was similar in both normoxic and hypoxic to normoxia, ionomycin, or hypoxia, the tubule pellet was tubules (Fig. 1). A very small amount of cathepsin activity was separated from a 2-ml tubule suspension by centrifugation. detected in the medium of normoxic and hypoxic tubules (Fig. The tubule pellet was immediately resuspended in a calcium- 1). If a leak of cathepsin activity from the damaged hypoxic free imidazole-HCl buffer containing in addition 1 mM EDTA tubules into the medium had occurred, one would have and 10 mM EGTA (pH 7.3). The suspension was then incu- expected higher cathepsin activity in the medium of the bated with digitonin (10 AM) at 37°C in a shaking water bath hypoxic tubules and lower tubular cathepsin activity. Neither for 5 min. Digitonin (10 ,uM) selectively permeabilizes the event was observed, despite high LDH release in hypoxia vs. plasma membrane but does not destroy lysosomal or mito- normoxia at 15 mim: 33.9% ± 3% vs. 11.1% ± 0.9%, respec- chondrial membranes of hepatocytes (17) or mitochondrial tively (P < 0.0001; n = 8). Although cathepsin activity in the membranes of rat proximal tubules (9). In separate experi- tubules after 25 min of normoxia or hypoxia was higher than ments, using the lysosomal dye Lucifer yellow and a video at 15 mia, this difference was not statistically significant (NS). imaging microscope previously described (3), our laboratory Validation of Calpain (Calcium-Dependent Cytosolic Pro- has confirmed that 10 ,tM digitonin selectively permeabilizes tease) Activity Assay. Calpain activity measured after 2 min of the plasma membrane of isolated rat proximal tubules without incubation with the calcium ionophore ionomycin increased in affecting the lysosomal membrane. Thus, 10 ,uM digitonin a dose-dependent fashion (Fig. 2). In the same tubules, 2 min releases cytosolic enzymes, including calpain, while keeping of preincubation with ionomycin did not damage the tubules the lysosomal membrane intact. compared to the control (Fig. 2). The enzyme activity mea- After incubation with digitonin, the tubule pellet and su- sured was calcium dependent (assay was done with and without pematant containing released cytosolic calpain were again calcium). Downloaded by guest on September 23, 2021 7664 Medical Sciences: Edelstein et al. Proc. Natl. Acad. Sci. USA 92 (1995)

200 r mg ofprotein (P < 0.001; n = 6) (Fig. 3). At 7.5 min ofhypoxia, NS calpain activity was 19.7 ± 4.3, which was significantly in- . C creased compared to 14.4 ± 3.2 in normoxic controls (P < ' 2 150 0.04; n = 6). The increase in calpain activity at 7.5 min O occurred before there was co0. iLT significant plasma membrane dam- CE age as assessed by LDH release (8.5% ± 0.9% in normoxic *j n..100 controls vs. 10.7% ± 0.8% after 7.5 min of hypoxia) (NS; n = .: 6). Cysteine Protease Inhibitor Studies with E-64-d. After the oOc o) E 501F recovery period, the tubules were preincubated for 10 minwith C E-64-d (600 ,uM) at 37°C in a shaking water bath to allow the inhibitor to enter the tubule cell. The tubules preincubated o1 /////' _NS with E-64-d showed a 99.7% suppression of calpain activity, NS from 23 ± 3 to 0.06 ± 0.003 pmol min-1 mg-1 after 15 min of 200 hypoxia (n = 6; P < 0.01). However, this dose of E-64-d significantly damaged normoxic controls after 15 min: LDH release increased from 12.1 ± 0.9 in normoxic controls to 18.8 S2t0150- + 1.2 in normoxic controls preincubated with E-64-d (n = 6; P < 0.05). In spite of this effect, E-64-d still provided partial CE cytoprotection from hypoxic injury. E-64-d (600 ,M) reduced *j , 100 aL C the LDH release after 15 min of hypoxia, from 43.8% ± 2.6% to 33% ± 2.2% (P < 0.01; n = 8). In separate experiments, 50 E-64-d reduced the LDH release after 25 min of hypoxia from C. 61.7% ± 1.6% to 49% ± 3.7% (P < 0.01; n = 5), thus also NS partially protecting against 25 min of hypoxic injury. 0 CBZ Markedly Protects Against Hypoxia-Induced LDH Tubules Media Release. After the recovery period, the tubules were. preincu- FIG. 1. Cathepsin activity in normoxic and hypoxic tubules and the bated for 10 min with CBZ (1 mM, 2 mM). CBZ induced a medium used for incubation. Hatched bars, normoxia; solid bars, dose-dependent decrease in calpain activity (Fig. 4). This was hypoxia. (Upper) Cathepsin activity after 15 min of hypoxia (n = 8). accompanied by dose-dependent protection against hypoxia- (Lower) Cathepsin activity after 25 min of hypoxia (n = 5). induced LDH release (Fig. 4). CBZ (2 mM) completely inhibited the increase of calpain induced by hypoxia, reducing Calpain Activity After 7.5 and 20 min of Hypoxia. Tubule the calpain activity to the same as that in normoxic controls. calpain activity in normoxic controls was 10.3 + 3 and in- This decrease in calpain was accompanied by a marked creased after 20 min of hypoxia to 27 ± 1.9 pmol per min per 30 p<0. mm 40 p<0.01 4J 0 ._L 00)% p a 20 > 6) 00 E 0j %%0CE20 a0-5M._ 0 .C .Q E C ) C 610 co_1 10 0. aE 0 7.5 min 20 min 60r 60 r

x NS 6) 40o x 0 6) NS 6) 40 F I to L. 0 z 20 6) 0 [ NS L.0 -J I I 0 20 F -i NS I 0 3 5 Ionomycin (lwM) 0 7.5 min min... FIG. 2. Dose-dependent effect of ionomycin on calpain activity precedes cell membrane injury. (Upper) Calpain activity after 2 min of FIG. 3. Time course of the effect of hypoxia on calpain activity and exposure to increasing doses of ionomycin (n = 5). (Lower) Membrane LDH release in freshly isolated rat proximal tubules. Tubules were damage as assessed by LDH release does not increase after 2 min of exposed to 7.5 and 20 min of normoxia or hypoxia. Open bars, incubation with increasing doses of ionomycin (n = 5). normoxia; solid bars, hypoxia (n = 6). Downloaded by guest on September 23, 2021 Medical Sciences: Edelstein et aL Proc. Nati. Acad. Sci. USA 92 (1995) 7665

p<0.01 a 9% reduction from 123.8 ± 7.7 to 112.9 ± 7.2 pmolFmin-' * mg- 1 by 500 ALM CBZ (n = 4; NS) and an 18% reduction to P<0.01 p<0.05 101.4 ± 7.3 by 2 mM CBZ (n = 4; P < 0.01).

L.z0- DISCUSSION 4-' 06 The rationale for studying calpain, a calcium-dependent cys- 0.5 2 teine protease, in hypoxic tubular injury is based on the E 20 .C following observations suggesting a role of calcium in medi- ating hypoxic injury: (i) hypoxic damage in rat proximal tubules is accompanied by an increased Ca21 influx from the 0E 0. medium into the cells (9); (ii) chemically dissimilar calcium channel blockers decrease the rate of Ca2+ influx and damage after 10 min of hypoxia (9); (iii) in primary cultures of anoxic rabbit renal tubules, calcium plays a role in renal cell death and p<0.001 tubular necrosis during the early periods of reoxygenation (19) and channel blockers increase cell 60 p(0.001 p<0.001 calcium viability (20); (iv) when (Ca2+)i and hypoxic injury were studied concurrently in the same tubules, the 10-min (Ca2+), increase correlated x F significantly with subsequent cell damage observed at 20 min 40 (3); and (v) lowering extracellular calcium attenuates proximal tubule cell hypoxic injury as reflected by LDH release (3, 4). 20- Calpain is the major cytosolic protease and the only calcium- dependent cytosolic protease so far described (21). Renal brush border membranes are rich in proteases (22) and calpain has been demonstrated in proximal tubules in culture (7) and - in both distal and tubules in the rabbit collecting (23, 24). The potential role of cysteine proteases in renal tubular injury has been suggested by Wilson and Hartz (7). In cultured Normoxia Hypoxia Hypoxia Hypoxia renal tubules, cyclosporine-induced tubule cell death was (20 Min) + + inhibited by a reduction in extracellular calcium in the medium, CBZ (1mM) CBZ (2mM) suggesting a role for calcium-dependent processes in cyclo- FIG. 4. Effect of CBZ on LDH release and calpain activity in sporine toxicity. Indeed, calpain activity was increased in normoxic and hypoxic tubules (20 min). (Upper) Effect of CBZ on tubules exposed acutely to cyclosporine. hypoxia-induced activation of calpain [n = 8 for normoxia, hypoxia (20 On this background, calpain-mediated degradative proteo- min), and hypoxia + CBZ (2 mM); n = 5 for CBZ (1 mM)]. (Lower) lysis may be an important mechanism contributing to lethal Effect of CBZ on plasma membrane damage as assessed by percentage anoxic cell injury. For example, ischemia of hippocampal LDH release [n = 10 for normoxia, hypoxia (20 min), and hypoxia + neurons triggers of cytoskeletal spectrin, a pre- CBZ (2 mM); n = 5 for CBZ (1 mM)]. ferred substrate of calpain, and inhibition of this proteolysis protects hippocampal neurons from ischemia (25, 26). pH- decrease in hypoxia-induced LDH release. This is compared dependent nonlysosomal proteolysis contributes to lethal an- to E-64-d, which completely inhibited all calpain activity in oxic injury of rat hepatocytes (27). Calcium-dependent non- both controls and hypoxic tubules but also damaged normoxic lysosomal proteases may contribute to cell injury during anoxia controls. Compared to E-64-d, CBZ did not damage normoxic in myocardium (28, 29). However, the role of these proteases, controls: LDH release was 14% ± 0.5% in normoxic controls specifically calpain, in hypoxia-induced proximal renal tubular compared to 14.3% 0.6% in normoxic controls preincubated damage has not been studied. with 1 mM CBZ (n = 5; NS) and 14.5% ± 0.8% in normoxic In the present study, lysosomal cathepsin B and L activity controls preincubated with 2 mM CBZ (n = 5; NS). was measured in isolated proximal tubules and the medium Calpain Activity in the Medium. Calpain activity measured used for incubation. Cathepsins were not elevated after 15 and in the medium and expressed as a percentage of tubular 25 min of hypoxia as compared to normoxic controls. These activity was negligible: 1.8% in normoxia, 0.35% after 20 min results are in agreement with the following observations that of hypoxia, 4% after treatment with 5 ,M ionomycin, and the non-calcium-dependent lysosomal cathepsins do not play a undetectable after preincubation with CBZ (2 mM). role in lethal injury. These observations are that lysosomes are CBZ Protects Against lonomycin-Induced Tubule Damage. highly resistant to disruption during lethal injury (30), lysoso- After the recovery period, normoxic tubules were incubated mal proteolysis is inhibited in lethal anoxic injury of rat with 5 ,uM ionomycin for 5 min. LDH release was 8% ± 0.9% hepatocytes (27), lysosomal proteolysis is suppressed by ATP in normoxic controls, 18.5% ± 2.5% in ionomycin-treated depletion (31), and lysosomal cathepsins do not play a role in tubules (n = 4; P < 0.001 vs. normoxic controls), and 13% ± acute cyclosporine toxicity (7). 1% in ionomycin-treated tubules preincubated with 2 mM To study the role of calcium-dependent cytosolic cysteine CBZ (n = 4; P < 0.01 vs. ionomycin alone). CBZ significantly proteases in cell injury, an assay for calpain in freshly isolated decreased LDH release induced by ionomycin, thus indicating rat proximal tubules was then developed. Calpain exists in the that ionomycin-induced tubular damage is also mediated in cytosol as the inactive proenzyme procalpain, which translo- part by calpain. cates from the cytosol to the cell membrane in the presence of Effect of E-64-d and CBZ on Tubular Cathepsin Activity. micromolar levels of calcium. Autocatalytic activation of pro- After the recovery period, the tubules were preincubated for calpain to active calpain occurs at the membrane in the 10 min with E-64-d (600 ,M) or CBZ (500 ,uM, 2 mM). The presence of physiological levels of calcium and phosphatidyl- tubules preincubated with E-64-d showed a 100% suppression inositol (21, 32). In the present study, the activation of calpain of cathepsin activity after 15 min of hypoxia. However, CBZ was calcium dependent, as indicated by the ionomycin studies. was a much less potent inhibitor of tubular cathepsin activity Specifically, increasing doses of ionomycin also increased than E-64-d. Cathepsin activity after 15 min of hypoxia showed calpain activity in a dose-dependent manner. We then studied Downloaded by guest on September 23, 2021 7666 Medical Sciences: Edelstein et al. Proc. Natl. Acad. Sci. USA 92 (1995) calpain activation during hypoxia, as we have previously found 3. Kribben, A., Wieder, E. D., Wetsels, J. F. M., Yu, L., Gengaro, that the increase in (Ca2+), precedes the hypoxic damage (3). P., Burke, T. J. & Schrier, R. W. (1994) J. Clin. Invest. 93, We found that calpain activity increased during 20 min of 1922-1929. hypoxia; this increase was significant at 7.5 min, a time prior 4. Wetzels, J. F. M., Yu, L., Wang, X., Kribben, A., Burke, T. J. & to evidence of plasma membrane damage. These findings Schrier, R. W. (1993) J. Pharmacol. Exp. Ther. 267, 176-180. suggested that in 5. Choi, K. H., Gengaro, P., Schrier, R. W. & Nemenoff, R. A. calpain may play a role hypoxic injury of the (1992) J. Am. Soc. Nephrol. 3, 704 (abstr.). rat proximal tubule rather than being the result of the injury. 6. Wilson, P. D., Firestone, R. E. & Lenard, J. (1987) J. Cell Bio. To further determine whether calpain plays a role in this 104, 1223-1229. hypoxic injury, inhibitor studies were undertaken. In the past, 7. Wilson, P.D. & Hartz, P.A. (1991) Cell Biol. Int. Rep. 15, specific cysteine protease inhibitor studies were difficult to 1243-1258. interpret because the charged protease inhibitors did not 8. Wetzels, J. F. M., Wang, X., Gengaro, P. A., Nemenoff, R. A., penetrate cells. Synthetic uncharged cell-penetrating inhibitors Burke, T. J. & Schrier, R. W. (1993)Am. J. Physiol. 264, F94-F99. of cysteine proteases have recently been developed and are 9. Almeida, A. R. P., Bunnachak, D., Burnier, M., Wetzels, J. F. M., proving useful in examination of the role of calpain in many Burke, T. J. & Schrier, R. W. (1992)J. Pharmacol. Exp. Ther. 260, cellular processes (18). In the present study, E-64-d and CBZ, 526-532. chemically dissimilar, uncharged, specific cysteine protease 10. Joseph, J. K., Bunnachak, D., Burke, T. J. & Schrier, R. W. inhibitors that penetrate cell membranes were used. E-64-d (1990) J. Am. Soc. Nephrol. 1, 837-845. completely inhibited both cathepsin and calpain activity in 11. Bergmeyer, H. U. (1974) Methods in Enzymatic Analysis (Aca- demic, New York), 2nd Ed., pp. 574-589. hypoxic tubules. Normoxic tubules demonstrated significant 12. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. membrane injury with E-64-d, suggesting that E-64-d has (1951) J. Biol. Chem. 193, 265-275. nonspecific effects or that complete inhibition of calpain 13. Barret, A. J. & Kirschke, H. (1981) Methods Enzymol. 80, 535- activity in normoxic tubules is damaging. The possibility that 565. 100% inhibition of calpain may be damaging to normoxic 14. Olbricht, C. J., Cannon, J. K., Garg, L. C. & Tisher, C. C. (1986) tubules is suggested by the regulatory role of calpain in Am. J. Physiol. 250, F1055-F1062. cytoskeletal or cell membrane modeling during normal cell 15. Kirschke, H., Wood, L., Roisen, F. J. & Bird, J. W. C. (1983) development (33). In spite of this deleterious effect in nor- Biochem. J. 214, 871-877. moxic tubules, E-64-d elicited cytoprotection against both 15 16. Sasaki, T., Kikuchi, T., Yomuto, N., Yoshimura, N. & Murachi, and 25 min of hypoxia. Nevertheless, because of the cytotoxic T. (1984) J. Biol. Chem. 259, 12489-12494. effects of E-64-d on normoxic tubules, studies were under- 17. Gores, G. J., Niemenen, A. L., Wray, B. E., Herman, B. & taken with CBZ, another cell-permeable cysteine protease Lemasters, J. L. (1989) J. Clin. Invest. 83, 386-396. 18. Mehdi, S. (1991) Trends Biochem. Sci. 16, 150-153. inhibitor. CBZ produced minimal inhibition of cathepsins 19. Wilson, P. D. & Schrier, R. W. (1986) Kidney Int. 29, 1172-1179. (9-18%) but completely abolished the increase in calpain 20. Schwertschlag, U., Schrier, R. W. & Wilson, P. D. (1986) J. activity during hypoxia. Moreover, this inhibitor exhibited no Pharmacol. Exp. Ther. 238, 119-124. deleterious effect on normoxic tubules, thus explaining the 21. Suzuki, K., Saido, T. C. & Hirai, S. (1992) Ann. N.Y Acad. Sci. more marked cytoprotection against hypoxic plasma mem- 674, 218-227. brane damage than was observed with E-64-d. 22. Scherberich, J. E., Wolf, G., Stuckhardt, C., Kugler, P. & Schoeppe, In summary, in freshly isolated renal proximal tubules, an W. (1988) Adv. Exp. Med. Biol. 240, 275-282. assay for calpain has been developed and used to show that 23. Hayashi, M., Kasau, Y. & Kawashima, S. (1987) Biochem. calpain activity is increased during hypoxia and that an early Biophys. Res. Commun. 148, 567-574. increase in calpain activity precedes hypoxia-induced cell 24. Yoshimura, N., Hatanaka, M., Kitahara, A., Kawaguchi, N. & membrane damage. An important role for calpain as a medi- Murachi, T. (1984) J. Bio. Chem. 259, 9847-9852. ator of proximal tubular hypoxic injury is supported by studies 25. Seubert, P., Lee, K. & Lynch, G. (1989) Brain Res. 492, 366-370. with 26. Lee, K. S., Frank, S., Vanderklish, P., Arai, A. & Lynch, G. (1991) CBZ, which inhibits the increase in calpain activity during Proc. Natl. Acad. Sci. USA 88, 7233-7237. hypoxia and markedly protects against hypoxic tubular dam- 27. Bronk, S. F. & Gores, G. J. (1993) Am. J. Physiol. 264, G744- age. The present results thus indicate that, in addition to G751. phospholipases (5) and nitric oxide (34), calpain also plays a 28. Lizuka, K., Kawaguchi, H. & Yasuda, H. (1991) Biochem. Med. role in hypoxia-induced proximal tubular injury in the rat. It Metab. Biol. 46, 427-431. will be interesting to elucidate the cellular targets of these 29. Tolnadi, S. & Korecky, B. (1986) Can. J. Cardiol. 2, 442-447. calcium-dependent enzymes during hypoxia. 30. Hawkins, H. K., Ericsson, J. L. E., Biberfeld, P. & Trump, B. F. (1972) Am. J. Pathol. 68, 255-288. E-64-d was a generous gift from Taisho Pharmaceutical, Tokyo. This 31. Plomp, P. J. A. M., Gordon, P. D., Meijen, A. J., Hoyvik, H. & work was supported by a grant from the National Institutes of Health Seglen, P. 0. (1989) J. Biol. Chem. 264, 6699-6704. (DK 35098). C.L.E. was supported by grants from the University of 32. Saido, T. C., Sorimachi, H. & Suzuki, K. (1994) FASEB J. 8, Stellenbosch and the National Kidney Foundation of South Africa. 814-822. 33. Kamatsu, K., Inazuki, K., Hosoya, J. & Satoh, S. (1986) Exp. 1. Bonventre, J. V. (1993) Kidney Int. 43, 1160-1178. Neurol. 91, 23-29. 2. Schrier, R. W., Arnold, P. E., Van Putten, V. J. & Burke, T. J. 34. Yu, L., Gengaro, P., Niederberger, M., Burke, T. J. & Schrier, (1987) Kidney Int. 32, 313-321. R. W. (1994) Proc. Natl. Acad. Sci. USA 91, 1691-1695. Downloaded by guest on September 23, 2021