Calpain inhibitor-1 reduces renal ischemia/reperfusion injury in the rat

Transcription

1 Kidney International, Vol. 59 (2001), pp Calpain inhibitor-1 reduces renal ischemia/reperfusion injury in the rat PRABAL K. CHATTERJEE, PAUL A.J. BROWN, SALVATORE CUZZOCREA, KAI ZACHAROWSKI, KEITH N. STEWART, HELDER MOTA-FILIPE, MICHELLE C. MCDONALD, and CHRISTOPH THIEMERMANN Department of Experimental Medicine and Nephrology, The William Harvey Research Institute, and the Royal London School of Medicine and Dentistry, London, England, and Department of Pathology, University Medical Buildings, Aberdeen, Scotland, United Kingdom; Institute of Pharmacology, University of Messina School of Medicine, Messina, Italy; and Laboratory of Pharmacology, Faculty of Pharmacy, University of Lisbon, Lisbon, Portugal Calpain inhibitor-1 reduces renal ischemia/reperfusion injury Renal ischemia is one of the most common causes of in the rat. acute renal failure (ARF), initiating a complex and inter- Background. Activation of the cysteine protease calpain has related sequence of events, resulting in injury to and the been implicated in renal ischemia/reperfusion (I/R) injury. The aim of this study was to investigate the effects of calpain inhibi- eventual death of renal cells [1, 2], with proximal tubule tor-1 (Cal I-1) in an in vivo model of renal I/R injury. (PT) cells demonstrating particular susceptibility [2, 3]. Methods. Male Wistar rats were administered Cal I-1 (10 Although reperfusion is essential for the survival of ischmg/kg, IP) 30 minutes before undergoing bilateral renal isch- emic tissue, there is good evidence that reperfusion itself emia (45 minutes) followed by reperfusion (6 hours). Plasma causes additional cellular injury (reperfusion injury) [3], concentrations of urea, creatinine, Na, -glutamyl transferase which has been attributed to the generation of reactive ( GT), aspartate aminotransferase (AST) and urinary Na, glutathione S-transferase (GST), and N-acetyl- -d-glucosamidepletion, neutrophil infiltration, phospholipase activa- oxygen species (ROS), adenosine 5 -triphosphate (ATP) nidase (NAG) were measured for the assessment of renal dysfunction and I/R injury. Creatinine clearance (C Cr ) and frac- tion and membrane lipid alterations, cytoskeletal dystional excretion of Na (FE Na ) were used as indicators of function, and intracellular Ca 2 accumulation [3 5]. Nuglomerular and tubular function, respectively. Kidney myeloper- merous studies have demonstrated a role for Ca 2 in oxidase (MPO) activity and malondialdehyde (MDA) levels the pathophysiology of renal injury associated with both were measured for assessment of neutrophil infiltration and lipid peroxidation, respectively. Renal sections were used for hypoxia and ischemia/reperfusion (I/R) [5, 6]. Intracelluhistologic grading of renal injury and for immunohistochemical lar accumulation of Ca 2 under ischemic conditions re- localization of inducible nitric oxide synthase (inos) and sults in the activation of Ca 2 -dependent enzymes such cyclooxygenase-2 (COX-2). as calpain and nitric oxide synthase (NOS), which are Results. Cal I-1 significantly reduced I/R-mediated increases implicated in the pathophysiology of ARF [6, 7]. in urea, creatinine, GT, AST, NAG, and FE Na and significantly Calpains are nonlysosomal neutral cysteine proteases improved C Cr. Cal I-1 also significantly reduced kidney MPO activity and MDA levels. Cal I-1 also reduced histologic eviin that are activated by Ca 2 and that appear to be involved dence of I/R-mediated renal damage and caused a substantial Ca 2 signaling in mammalian cells [8]. To date, two major reduction in the expression of inos and COX-2, both of which isoforms have been identified: calpain I (or -calpain) and involve activation of nuclear factor- B (NF- B). calpain II (or m-calpain), which require low ( mol/l) Conclusions. These results suggest that Cal I-1 reduces the and high (mmol/l) Ca 2 concentrations for activation, renal dysfunction and injury associated with I/R of the kidney. respectively [8]. Following activation, calpain selectively We suggest that the mechanism could involve the inhibition of I/R-mediated activation of NF- B. cleaves a specific subset of cellular proteins, including cytoskeletal and membrane proteins, several enzymes, and transcription factors [9], including nuclear factor- B Key words: kidney injury, inducible nitric oxide synthase, COX-2, nuclear (NF- B) [10], which is involved in the expression of factor- B, acute renal failure. including inducible NOS (inos) [11] and cyclooxygen- ase-2 (COX-2) [12]. Received for publication August 11, 2000 and in revised form November 30, 2000 Calpain inhibitor-1 (Cal I-1) is a cell-permeable pep- Accepted for publication December 26, 2000 tide aldehyde that blocks the active site of calpain 2001 by the International Society of Nephrology [13, 14]. Along with other inhibitors of calpain such as 2073

2 2074 E64 or PD and the naturally occurring calpain surgically for renal I/R as described previously [29]. inhibitor protein calpastatin, Cal I-1 has been used to Briefly, anesthetized rats were placed onto a thermostati- investigate the role of the excessive activation of calpain cally controlled heating mat (Harvard Apparatus Ltd., in the pathophysiology of a variety of disorders, including Kent, UK), and body temperature was maintained at 38 cataract, restenosis, arthritis, stroke, and myocardial in- 1 C by means of a rectal probe attached to a homeothermic farction [13, 14]. Several studies have demonstrated that blanket. A tracheotomy was performed to maintain calpain activation is implicated in the pathophysiology airway patency and to facilitate spontaneous respiration. of I/R injury in several organs, including the brain [15], The right carotid artery was cannulated (PP50, I.D heart [16], and liver [17]. Interestingly, I/R has also been mm; Portex, Kent, UK) and connected to a pressure reported to attenuate calpastatin activity [18]. In the transducer (Senso-Nor 840, Horten, Norway) for the kidney, there is evidence from in vitro studies using rat measurement of mean arterial blood pressure (MAP) and rabbit PT suspensions that calpain activation is involved and derivation of the heart rate (HR) from the pulse in the cellular injury/death associated with hyp- waveform, which were displayed on a data-acquisition oxia [19, 20] and exposure to toxic agents [21, 22], and system (MacLab 8e, AD Instruments, Hastings, UK) that calpain inhibitors may be beneficial under these installed on an Apple Macintosh computer. MAP and conditions [19, 20, 22]. Although there is some evidence HR were monitored for the duration of each experiment. that calpain is activated in the ischemic rat kidney in The jugular vein was cannulated (PP25, I.D mm; vivo [23], the beneficial actions of calpain inhibitors in Portex) for the administration of drugs. A midline lapa- animal models of renal I/R remain unclear [24]. rotomy was performed and the bladder was cannulated Several studies have demonstrated that Cal I-1 can (PP90, I.D mm; Portex). Both kidneys were located, reduce the proteolytic cleavage of I B- / from its com- and the renal pedicles, containing the artery, vein, and plex with the NF- B/Rel A unit and, hence, inhibit the nerve supplying each kidney, were carefully isolated. activation of NF- B and its subsequent translocation from the cytosol into the nucleus [25 27]. Thus, Cal I-1 Renal ischemia/reperfusion can reduce the expression (for example, after exposure As described previously [29], rats were allowed to to endotoxin or inflammatory cytokines) of many NF- B stabilize for 30 minutes before they were subjected to dependent genes, including inos [26 28] and COX-2 bilateral renal occlusion for 45 minutes using artery clips [28]. However, to date, such an effect has not been demonstrated to clamp the renal pedicles. Reperfusion commenced once in renal tissues in vitro or in vivo. the artery clips were removed (control animals). Occlusion Therefore, the aim of this study was to investigate the was verified visually by change in the color of the kidneys effects of Cal I-1 on renal dysfunction/injury caused by to a paler shade and reperfusion by a blush. Other rats renal I/R in the anesthetized rat. In order to gain a better were subjected to sham operation (sham operated) where insight into the mechanism of action of Cal I-1, we also identical surgical procedures to control animals were investigated the effects of Cal I-1 on the expression of used, except that they did not undergo bilateral renal inos and COX-2 in kidneys of rats subjected to renal clamping and were maintained under anesthesia for the I/R. Furthermore, the actions of Cal I-1 were compared duration of the experiment. At the end of all experiwith that of the serine protease inhibitor chymostatin. ments, animals were killed by an overdose of sodium thiopentone. METHODS Experimental protocol Animal preparation Upon completion of surgical procedures, the animals In vivo studies were carried out using 56 male Wistar were randomly allocated into seven groups. rats (Tuck, Rayleigh, Essex, UK) weighing 200 to 250 g, (1) I/R control group. These animals underwent renal that received a standard diet and water ad libitum and ischemia for 45 minutes followed by reperfusion for six were cared for in accordance with both the Home Office hours (N 12). Guidance in the Operation of the Animals (Scientific Procedures) (2) I/R Cal I-1 group. These animals were adminisand Act 1986, published by H.M.S.O. (London, UK) tered Cal I-1 (10 mg/kg IP) 30 minutes prior to I/R as the Institutional Animal Research Committee Guide described previously (N 12). Cal I-1 was dissolved in for the Care and Use of Laboratory Animals published by 50% (vol/vol) EtOH and 50% (vol/vol) saline. We have the U.S. National Institutes of Health (NIH publication previously shown this dose and method of administration number 85-23, revised 1996). All animals were anesthe- of Cal I-1 to (a) inhibit an increase in calpain activity [28], tized with sodium thiopentone (Intraval Sodium, 120 (b) reduce the expression of inos and COX-2 [28, 30], mg/kg IP; Rhone Merieux Ltd., Essex, UK), and anesthe- and (c) to attenuate the activation of NF- B [30] associated sia was maintained by supplementary intravenous infusions of sodium thiopentone. Animals were prepared with endotoxic and hemorrhagic shock in the rat [28, 30].

3 2075 (3) I/R chymostatin group. These animals were admin- described method [35]. Briefly, at the end of the experiments, istered chymostatin (10 mg/kg IP) 30 minutes prior to kidney tissue was weighed and homogenized in I/R (N 6). Chymostatin was dissolved in 50% (vol/ a solution containing 0.5% (wt/vol) hexadecyltrimethyvol) EtOH and 50% (vol/vol) saline. lammonium bromide dissolved in 10 mmol/l potassium (4) I/R vehicle group. These animals were adminisat phosphate buffer (ph 7.4) and centrifuged for 30 minutes tered the vehicle for Cal I-1 and chymostatin [50% 20,000 g at 4 C. An aliquot of supernatant was then EtOH/50% saline (vol/vol) IP] 30 minutes prior to I/R removed and added to a reaction mixture containing 1.6 (N 6). mmol/l tetramethylbenzidine and 0.1 mmol/l hydrogen (5) Sham group. Sham-operated rats were subjected peroxide (H 2 O 2 ). The rate of change in absorbance was to identical surgical procedures described previously in measured spectrophotometrically at 650 nm. MPO activ- this article, except for renal I/R, and they were maindegrade ity was defined as the quantity of enzyme required to tained under anesthesia for the duration of the experimiu/100 1 mol of H 2 O 2 at 37 C and was expressed in ment (that is, 45 minutes 6 hours, N 12). mg wet tissue. (6) Sham Cal I-1 group. These were identical to the Determination of malondialdehye levels sham-operated animals except for the pretreatment with Cal I-1 (10 mg/kg IP) 30 minutes prior to surgery (N 4). Levels of malondialdehyde (MDA) in kidneys were (7) Sham vehicle group. These were identical to shaming a protocol described previously [36]. Briefly, kidney determined as an indicator of lipid peroxidation followoperated animals except they were pretreated with vehitissue was weighed and homogenized in a 1.15% (wt/ cle [50% EtOH/50% saline (vol/vol) IP] 30 minutes prior to the surgery (N 4). vol) KCl solution. A 100 L aliquot of homogenate was All animals in these seven groups received a continutaining 200 L 8.1% (wt/vol) lauryl sulfate, 1.5 ml 20% then removed and added to a reaction mixture conous infusion of 0.9% (wt/vol) saline (4 ml/kg/h, IV) throughout the I/R period. (vol/vol) acetic acid, 1.5 ml 0.8% (wt/vol) thiobarbituric acid, and 700 L distilled water. Samples were then Measurement of biochemical parameters boiled for 1 hour at 95 C and centrifuged at 3000 g for 10 minutes. The absorbance of the supernatant was At the end of the reperfusion period, blood (1 ml) measured spectrophotometrically at 650 nm. MDA levsamples were collected via the carotid artery. The samels were expressed as mol/l MDA/100 mg wet tissue. ples were centrifuged (6000 r.p.m. for 3 minutes) to separate plasma. All plasma samples were analyzed for bio- Histologic evaluation chemical parameters within 24 hours after collection At postmortem, a 5 mm section of kidney was removed (Vetlab Services, Sussex, UK). Plasma concentrations and placed in formalin and processed through to wax. of urea and creatinine were measured as indicators of Five micrometer sections were cut and stained with heimpaired glomerular function [31]. Plasma concentramatoxylin and eosin. Histologic assessment of tubular tions of -glutamyl transferase ( GT) and aspartate aminecrosis was determined semiquantitatively using a notransferase (AST), enzymes that are both located in method modified from McWhinnie et al [37]. Random the PT [32], were used as indicators of renal reperfusion cortical fields were observed using a 20 objective. A injury (Discussion section). graticule grid (25 squares) was used to determine the Urine samples were collected during the reperfusion number of line intersects involving tubular profiles. One period, and the volume of urine produced was recorded. hundred intersections were examined for each kidney, Urine concentrations of creatinine and Na were mea- and a score from 0 to 3 was given for each tubular profile sured (Vetlab Services) at the end of the reperfusion involving an intersection: 0 normal histology (Fig. 6A); period and were used in conjunction with plasma concen- 1 tubular cell swelling, brush border loss, nuclear contrations to estimate creatinine clearance (C Cr ) and frac- densation, with up to one third of the tubular profile tional excretion of Na (FE Na ) using standard formulae. showing nuclear loss (Fig. 6B); 2 same as for score 1, These were used as respective indicators of glomerular but greater than one third and less than two thirds of and tubular function during the reperfusion period. Con- the tubular profile show nuclear loss (Fig. 6C); and 3 centrations of urinary glutathione S-transferase (GST) greater than two thirds of the tubular profile showing and N-acetyl- -d-glucosaminidase (NAG), specific indi- nuclear loss (Fig. 6D). cators of tubular damage [33, 34], were also measured The total score for each kidney was calculated by the (Laboratory of Pharmacology, University of Lisbon, Lis- addition of all 100 scores with a maximum score of 300. bon, Portugal). Immunohistochemical analysis of inos and Determination of myeloperoxidase activity COX-2 expression Myeloperoxidase (MPO) activity in kidneys was used The expression of inos and COX-2 proteins was evaluated as an indicator of neutrophil infiltration using a previously in kidneys from untreated (sham-operated), con-

4 2076 trol, and animals administered Cal I-1 using immunohistochemical protocols described previously [38]. At the end of the experiment, kidneys were bisected, snap frozen in liquid nitrogen, and stored at 70 C. When required, samples were thawed, dehydrated using graded ethanol, and embedded in Paraplast, after which 8 m sections were cut. After deparaffinization, endogenous peroxidase was quenched using 0.3% (vol/vol) H 2 O 2 in 60% (vol/vol) methanol for 30 minutes. Sections were permeablized using 0.1% (vol/vol) Triton X-100 in phosphate-buffered saline (PBS; 0.01 mol/l, ph 7.4) for 20 minutes. Nonspecific adsorption was minimized by incubating the section in 2% (vol/vol) normal goat serum (DBA, Milan, Italy) in PBS for 20 minutes. Sequential incubation for 15 minutes with avidin and biotin (DBA) was used to block endogenous binding sites. The sections were then incubated overnight with 1:1000 dilution of primary anti-inos or anti-cox-2 antibody (DBA) or with control solutions. Controls included buffer alone or nonspecific purified rabbit IgG (DBA). Specific labeling was detected with a biotin-conjugated goat anti-rabbit IgG (DBA) and avidin-biotin peroxidase complex (DBA). To verify the binding specificity for inos or COX-2 (negative controls), some sections were also incubated with only the secondary antibody (no primary antibody). Positive controls for inos and COX-2 binding were performed on sections of kidney obtained from mice that were administered lipopolysaccharide (LPS; 50 mg/kg, IV). Materials Unless otherwise stated, all compounds used in this study were purchased from Sigma-Aldrich Company Ltd. (Poole, Dorset, UK). Cal I-1 was purchased from Calbiochem Novabiochem (Nottingham, UK). All solu- tions used for in vivo infusions were prepared using nonpyrogenic saline [0.9% (wt/vol) NaCl; Baxter Healthcare Ltd., Thetford, Norfolk, UK]. Statistical analysis Fig. 1. Alterations in (A) plasma urea and (B) plasma creatinine concentrations during renal ischemia/reperfusion (I/R) in the presence of Cal I-1 (10 mg/kg IP), chymostatin (Chymo, 10 mg/kg IP), or vehicle [50% (vol/vol) EtOH/saline, IP]. *P 0.05 vs. sham-operated group; P 0.05 vs. control group (45 minutes of ischemia and 6 hours or reperfusion). RESULTS Effect of Cal I-1 and chymostatin on ischemia/ reperfusion-mediated glomerular dysfunction All values described in the text and figures are expressed as mean SEM for N observations. For in vivo studies, each data point represents biochemical measurements obtained from 6 to 12 separate animals. For his- tologic scoring, each data point represents analysis of kidneys taken from 6 to 12 individual animals. For immunohistochemical analysis, the figures shown are representative of at least three experiments performed on different experimental days. Statistical analysis was carried out using GraphPad Prism/InStat (GraphPad Software, Inc., San Diego, CA, USA). Data were analyzed using one-way analysis of variance (ANOVA) followed by Dunnett s post hoc test. A P value of less than 0.05 was considered to be significant (NS, nonsignificant). Animals that underwent renal I/R exhibited significant increases in the plasma concentrations of urea and creatinine compared with sham-operated animals (Fig. 1) and a significant reduction in C Cr (Fig. 2A), suggesting a sig- nificant degree of glomerular dysfunction. Pretreatment of rats with Cal I-1 prior to I/R produced relatively small, but significant, reductions in the plasma levels of urea and creatinine (Fig. 1) and increased C Cr significantly (Fig. 2A). The serine protease inhibitor chymostatin, administered prior to I/R, did not have a sig- nificant effect on the increased plasma urea and creatinine concentrations or the reduced C Cr associated with I/R (Figs. 1 and 2A). The administration of vehicle for

5 2077 Fig. 3. Alterations in (A) urinary glutathione S-transferase (GST) and (B) urinary N-acetyl- -D-glucosaminidase (NAG) levels during renal I/R in the presence of calpain inhibitor-1 (Cal I-1, 10 mg/kg IP), chymo- statin (Chymo, 10 mg/kg IP), or vehicle [50% (vol/vol) EtOH/saline, IP]. *P 0.05 vs. sham-operated group; P 0.05 vs. control group (45 minutes of ischemia and 6 hours of reperfusion). Fig. 2. Effect of renal I/R on (A) creatinine clearance and (B) fractional excretion of Na in the presence of calpain inhibitor-1 (Cal I-1) (10 mg/kg IP), chymostatin (Chymo, 10 mg/kg IP), or vehicle [50% (vol/ vol) EtOH/saline, IP]. *P 0.05 vs. sham-operated group; P 0.05 vs. control group (45 minutes of ischemia and 6 hours of reperfusion). Cal I-1 and chymostatin [50% (vol/vol) EtOH] to rats prior to I/R did not result in any significant alterations on plasma levels of urea or creatinine or in C Cr compared improvement in tubular function (Fig. 2B). The adminiswith control animals (Figs. 1 and 2A). The administration tration of chymostatin or vehicle did not produce signifi- of Cal I-1 or vehicle to sham-operated rats did not result cant alterations in FE Na compared with control animals, in any alteration in plasma levels of urea or creatinine and administration of Cal I-1 to sham-operated rats did or in C Cr in comparison with sham-operated animals adsham-operated not alter FE Na in comparison with values obtained from ministered saline only (data not shown). animals administered saline only (Fig. 2B). Renal I/R produced a significant increase in the uri- Effects of Cal I-1 and chymostatin on ischemia/ nary concentrations of both GST and NAG (Fig. 3), reperfusion-mediated tubular dysfunction/injury suggesting significant tubular injury. The administration Fractional excretion of sodium (FE Na ), calculated us- of Cal I-1 significantly reduced both urinary GST and ing plasma and urinary concentrations of Na in associainjury. NAG concentrations, suggesting attenuation of tubular tion with urine production (urine flow, ml/min), was The administration of chymostatin or vehicle did used as an indicator of tubular function. I/R produced not produce significant alterations in urinary GST or a significant increase in FE Na, suggesting tubular dysfunc- NAG concentrations compared with control animals tion (Fig. 2B); however, administration of Cal I-1 prior to (Fig. 3). The administration of Cal I-1 or vehicle to the I/R produced a significant reduction in FE Na, suggesting sham-operated animals did not alter GST or NAG con-

6 2078 Fig. 4. Alterations in (A) plasma -glutamyl transferase ( GT) and (B) plasma aspartate aminotransferase (AST) concentrations during renal I/R in the presence of calpain inhibitor-1 (Cal I-1, 10 mg/kg IP), chymostatin (Chymo, 10 mg/kg IP), or vehicle [50% (vol/vol) EtOH/ saline, IP]. *P 0.05 vs. sham-operated group; P 0.05 vs. control group (45 minutes of ischemia and 6 hours of reperfusion). Fig. 5. Effect of renal I/R on kidney (A) myeloperoxidase (MPO) activity and (B) malondialdehyde (MDA) levels in the presence of calpain inhibitor-1 (Cal I-1, 10 mg/kg IP), chymostatin (Chymo, 10 mg/kg IP), or vehicle [50% (vol/vol) EtOH/saline]. *P 0.05 vs. sham-operated group; P 0.05 vs. control group (45 minutes of ischemia and 6 hours of reperfusion). centrations in comparison with values obtained from sham-operated animals administered saline only (data not shown). Effects of Cal I-1 and chymostatin on ischemia/ reperfusion-mediated reperfusion injury Rats subjected to renal I/R exhibited a substantial increase in kidney MPO activity and MDA levels, sug- gesting increased neutrophil infiltration and lipid peroxidation, respectively (Fig. 5). However, pretreatment of rats with Cal I-1 prior to I/R produced a significant reduction of MPO activity in comparison with the activity obtained from control rat kidneys (Fig. 5A). Similarly, pretreatment of rats with Cal I-1 produced a significant reduction of the MDA levels associated with I/R (Fig. Renal I/R produced significant increases in the plasma concentrations of GT and AST in comparison with values obtained from sham-operated animals (Fig. 4). Plasma concentrations of GT and AST, which were used as markers of reperfusion injury, were significantly reduced subsequent to pretreatment with Cal I-1 prior to I/R (Fig. 4). The administration of chymostatin or vehicle did not produce significant alterations in plasma levels of GT or AST in comparison with control animals (Fig. 4). The administration of Cal I-1 or vehicle to the shamoperated animals did not alter plasma concentrations of GT or AST in comparison with values obtained from sham-operated animals administered saline only (data not shown). Effects of Cal I-1 and chymostatin on kidney MPO activity and MDA levels

7 2079 Fig. 6. Histologic evaluation of renal I/R. Histology of a normal proximal tubule (PT) (A) is compared with tubules revealing brush border loss, nuclear condensation, and cytoplasmic swelling and loss of nuclei in up to one third of the tubular profile (B). More severe cellular loss, with between one third and two thirds of the tubular profile denuded of nuclei (C), and greater than two thirds of nuclei lost (D) are also represented. HandE, 700. (E) Effect of renal I/R on total severity score in the presence of calpain inhibitor-1 (Cal I-1, 10 mg/kg IP), chymostatin (Chymo, 10 mg/kg IP), or vehicle [50% (vol/vol) EtOH/saline]. *P 0.05 vs. sham-operated group; P 0.05 vs. control group (45 minutes of ischemia and 6 hours of reperfusion). 5B). The administration of chymostatin or vehicle prior to I/R did not have a significant effect on MPO activity or MDA levels in comparison with values obtained from control animals (Fig. 5). Effects of Cal I-1 and chymostatin on I/R-mediated renal histopathology In comparison with the normal tubular histology ob- served in kidneys taken from sham-operated rats (Fig. 6A), animals that underwent renal I/R demonstrated the recognized features of severe acute tubular damage (Fig. 6D) [39]. These features included brush border loss, nu- clear condensation, cytoplasmic swelling, and a consis- tent loss of significant numbers of nuclei from tubular profiles (Fig. 6D). When compared with the total severity score mea- sured from kidneys obtained from sham-operated animals, I/R produced a significant increase in total severity score, which was significantly reduced by administration of Cal I-1 prior to I/R (Fig. 6E). The administration of chymostatin or vehicle to rats prior to I/R did not have a significant effect on total severity score (Fig. 6E). Immunohistochemical localization of inos and COX-2 formation When compared with kidney sections obtained from sham-operated rats (Fig. 7A), immunohistochemical analysis of sections obtained from rats subjected to renal I/R revealed positive staining for inos (Fig. 7B). In contrast, substantially reduced staining was observed in the kidney sections obtained from rats administered Cal I-1 (Fig. 7C). The positive control for inos staining, performed on sections of kidney obtained from mice subjected to endotoxemia, demonstrated marked posi- tive staining for inos protein (Fig. 7D). No positive staining for inos protein was observed in kidney sec- tions from rats subjected to renal I/R, which were incubated with the secondary antibody only (no primary anti- body negative control; Fig. 7E). In comparison with kidney sections obtained from sham-operated rats (Fig. 8A), sections prepared from rat kidneys subjected to I/R demonstrated marked staining for COX-2 (Fig. 8B). Kidneys obtained from rats admin- istered Cal I-1 demonstrated markedly reduced staining for COX-2 upon comparison with kidneys obtained from control animals (Fig. 8C), suggesting a reduction in the expression of COX-2 subsequent to pretreatment with Cal I-1 prior to I/R. Staining for endogenous biotin was not detected in any sections. The positive control for COX-2 staining, performed on sections of kidney ob- tained from mice subjected to endotoxemia, demon- strated marked positive staining for COX-2 protein (Fig. 8D). In contrast, no positive staining for COX-2 protein could be observed in kidney sections from rats subjected to renal I/R that were incubated with the secondary antibody only (no primary antibody negative control; Fig. 8E). DISCUSSION There is good evidence from both in vivo and in vitro studies that calpain activation plays an important role in the pathophysiology of renal injury mediated by hypoxia and I/R [23, 40]. We demonstrate here, to our knowledge for the first time, that the administration of Cal I-1 prior

8 2080 Fig. 7. Immunohistochemical localization of inducible nitric oxide synthase (inos) in rat kidney following I/R. (A) Kidney section from a sham animal (original magnification, OM, 80) and (B) control group (renal I/R only, OM 32), with typical areas of inos immunoreactivity indicated by arrows. In comparison, the inos immunoreactivity of kidneys from rats treated with Cal I-1 (10 mg/kg IP) (C) was markedly reduced (OM 80). (D) Positive control for inos immunoreactivity: Kidney sections were prepared from mice subjected to endotoxemia (LPS, 50 mg/kg IV). Typical areas of inos immunoreactivity are indicated by arrows (OM 80). (E) Negative control for inos immunoreactivity: Kidney sections were prepared from rats subjected to I/R only and incubated with secondary antibody only (no primary antibody, OM 80). Figures are representative of at least three experiments performed on different days. to renal I/R significantly reduces the renal dysfunction and injury caused by I/R of the rat kidney. This conclusion is supported by four key findings: In a rat model of renal I/R, Cal I-1 significantly reduced the I/R-mediated increases in (1) the plasma levels of urea and creatinine, (2) the FENa and urinary concentrations of GST and NAG, (3) the plasma levels of GT and AST, and (4) MPO activity and the levels of MDA in the kidney. Cal I-1 also significantly reduced the histologic evidence of I/R-mediated tubular injury and substantially reduced the immunohistochemical evidence of inos and COX-2 expression, suggesting that the protective mechanism provided by Cal I-1 may involve, to some degree, the inhibition of the activation of NF- B. Ischemia/reperfusion of the kidney causes both glo- merular and tubular dysfunction [41]. In this study, the moderate but significant increases in the plasma concentrations of urea and creatinine subsequent to I/R suggest the impairment of glomerular function [31], and this was reflected by a significant reduction in CCr observed subsequent to I/R. However, I/R also resulted in significant increases in urinary NAG and GST concentrations, which can be regarded as markers for tubular injury and possibly tubular function [33, 34]. This was reflected by a significant increase in FENa. Taken together, these markers suggest a significant reduction in tubular function and increased tubular injury in this model of renal I/R. Although there are recognized limitations in the use clearance techniques to measure CCr and FENa in renal models that involve extensive tubular injury, the compar-

9 2081 Fig. 8. Immunohistochemical localization of cyclooxygenase-2 (COX-2) in rat kidney following I/R. (A) Kidney section from Sham animal (OM 80). (B) Control group (renal I/R only, OM 32) with typical areas of COX-2 immunoreactivity indicated by arrows. COX-2 immunoreactivity of kidneys from rats treated with Cal I-1 (C) was markedly reduced in comparison with staining obtained from the control group (OM 80). (D) Positive control for COX-2 immunoreactivity: Kidney sections were prepared from mice subjected to endotoxemia (LPS, 50 mg/kg IV). Typical areas of COX-2 immunoreactivity are indicated by arrows (OM 80). (E) Negative control for COX-2 immunoreactivity: Kidney sections were prepared from rats subjected to I/R only and incubated with secondary antibody only (no primary antibody, OM 125). Figures are representative of at least three experiments performed on different days. atively larger increase in the markers of tubular dysfunction/injury (FENa, NAG, GST) suggests that there is a greater dysfunction and injury to the tubules rather than the glomeruli in this model of renal I/R. Marked tubular damage in models of I/R (such as the one used here) has been demonstrated previously [42] and was reflected, in this study, in the histopathological analysis of kidneys subsequent to I/R obtained from control animals. Renal I/R also produced significant increases in plasma concentrations of GT and AST, both of which are regarded as nonspecific markers of extensive cellular disruption or necrosis [31] and were used in this study as markers of renal I/R. Both enzymes are present within the PT [32] and can be released into the urine subsequent to renal injury [43, 44]. Furthermore, a recent study has demonstrated that GT is significantly elevated subsequent to short-term ischemia of the rat kidney [45]. Therefore, it is certainly feasible that the increased plasma levels measured subsequent to renal I/R originated from the kidney after extensive damage to the tubular architecture. Renal I/R also caused a significant increase in both MPO activity, indicating neutrophil accumulation, and in MDA levels, indicating increased lipid peroxidation, and taken together, increased oxidative stress. In this study, Cal I-1 produced a significant reduction of renal dysfunction and injury mediated by I/R of this kidney. Cal I-1 appears to provide a greater beneficial action against tubular, rather than glomerular, dysfunction. This is supported by our findings that compared

10 2082 with rats subjected to I/R only, Cal I-1 produced rela- expression of inos or COX-2 protein (data not shown). tively small, but significant, decreases in plasma urea and Although this is the first report to our knowledge of such creatinine levels. This moderate effect on glomerular an effect of Cal I-1 in renal tissues, the mechanism(s) function was reflected by the small, but also significant, by which Cal I-1, but not chymostatin, potentially inhibits increase in C Cr. In contrast, Cal I-1 had a marked effect the activation of NF- B and subsequently the attenuaon markers of tubular dysfunction (FE Na ) and injury (uri- tion of inos (and COX-2) expression certainly warrants nary NAG and GST). This was further supported by further investigation. the renal histopathology that revealed the recognized In conclusion, this study demonstrates that Cal I-1 features of severe acute tubular damage in rat kidneys reduces the renal dysfunction and injury caused by renal subjected to I/R only, which were markedly reduced by I/R in the anesthetized rat. Cal I-1 also reduces the I/R- Cal I-1. mediated expression of inos and COX-2 in the rat kid- Calpain inhibitor-1 readily permeates cell membranes ney. Thus, we speculate that the inhibition by Cal I-1 of and subsequently inhibits calpain activation [13, 14]. Al- the activation of NF- B may contribute to the beneficial though not measured in this study, one potential mecha- effects of Cal I-1 in I/R injury of the rat kidney. nism of action by which Cal I-1 could produce its beneficial effects is via the inhibition of cysteine protease ACKNOWLEDGMENTS (calpain) activity. However, this is unlikely to provide a P.K.C. and M.C.McD. are funded by the Joint Research Board of significant mechanism of protection in this model of renal St. Bartholomew s Hospital (Grants XMLA/G7Z4). K.Z. is supported I/R as (1) other cysteine protease inhibitors such as leupara by the Deutsche Herzstishung, H.M.F. by the Portuguese Fundacao a Ciencia e Technologia (Praxis XXI/BPD/16333/98), and C.T. peptin and calpeptin do not protect against toxicantby a British Heart Foundation Senior Research Fellowship (FS/96018). mediated renal cell death and ischemic ARF, respec- This work was supported in part by the Clinica Medica e Diagnostico, tively [22, 24] and (2) serine protease inhibitors such as Dr. Joaquim Chaves, Lisbon, Portugal. Parts of this study were pre- antipain have also been shown not to reduce ischemia- sented to the British Pharmacological Society meeting (Cambridge, UK) and were subsequently published (abstract; Chatterjee et al, Br mediated ARF [22]. In this study, chymostatin, another J Pharmacol 129:34P, 2000). potent inhibitor of serine proteases such as chymotrypsin and papain [46], did not affect I/R-mediated renal dysperimental Medicine and Nephrology, The William Harvey Research Reprint requests to Prabal K. Chatterjee, Ph.D., Department of Ex- function and injury. This is in keeping previous reports Institute, St. Bartholomew s and the Royal London School of Medicine from our laboratory that chymostatin, at the same dose and Dentistry, Charterhouse Square, London, EC1M 6BQ, England, as that used in this study, does not protect against the United Kingdom. circulatory failure and multiple organ dysfunction mediated p.k.chatterjee@mds.qmw.ac.uk by endotoxic [28] or hemorrhagic shock [30]. Taken together, this supports the view that an inhibition of APPENDIX protease activity is unlikely to account for the beneficial Abbreviations used in this article are: ARF, acute renal failure; AST, action of Cal I-1 observed in this study. aspartate aminotransferase; ATP, adenosine 5 -triphosphate; Cal I-1, Another putative mechanism for the beneficial actions calpain inhibitor-1; C Cr, creatinine clearance; COX-2, cyclooxygenase-2; observed with Cal I-1 in this study is related to its ability FE Na, fractional excretion of sodium; GT, gamma-glutamyl transferase; GST, glutathione S-transferase; HR, heart rate; inos, inducible nitric to attenuate the activation of NF- B [25 28, 30]. NF- B oxide synthase; I/R, ischemia/reperfusion; MAP, mean arterial pressure; is a member of a family of dimers belonging to the Rel/ MDA, malondialdehyde; MPO, myeloperoxidase; NAG, N-acetyl- NF- B family of polypeptides, and the most frequently -d-glucosaminidase; NF- B, nuclear factor- B; NOS, nitric oxide synthase; PT, proximal tubule; ROS, reactive oxygen species. observed form of NF- B is a dimer composed of two DNA-binding proteins, namely NF- B (or p50) and RelA (or p65), although other dimeric combinations also REFERENCES exist [47]. In this study, Cal I-1 reduced the renal I/R- 1. Thadhani R, Pascual M, Bonventre JV: Acute renal failure. N Engl J Med 334: , 1996 mediated expression of both inos and COX-2 protein, 2. Lieberthal W, Levine JS: Mechanisms of apoptosis and its potenas demonstrated using immunohistochemistry, and it is tial role in renal tubular epithelial cell injury. Am J Physiol our hypothesis that Cal I-1 mediates this effect via inhibi- 271:F477 F488, Weight SC, Bell PR, Nicholson ML: Renal ischemia-reperfusion tion of the expression of NF- B. This notion is supported injury. Br J Surg 83: , 1996 by direct evidence obtained in this laboratory demon- 4. Kloner RA, Przyklenk K, Whittaker P: Deleterious effects of strating that (1) Cal I-1 inhibits LPS-stimulated NF- B oxygen radicals in ischemia/reperfusion: Resolved and unresolved issues. Circulation 80: , 1989 DNA binding activity in RAW macrophages [30] 5. Paller MS: The cell biology of reperfusion injury in the kidney. and that (2) Cal I-1 inhibits LPS-mediated degradation J Invest Med 42: , 1994 of I B, I B, and I B in rat vascular smooth muscle 6. Yaqoob MM, Edelstein CL, Schrier RW: Identification of the cell cultures [30]. Furthermore, in our hands and in keepoxia-induced proximal tubular injury. Nephrology 2(Suppl):S15 novel calcium mediated cellular events in the pathogenesis of hyp- ing with findings that we have previously reported [28, 30], S21, 1996 chymostatin did not have any effect on the I/R-mediated 7. Edelstein CL, Yaqoob MM, Schrier RW: The role of calcium-

11 2083 dependent enzymes nitric oxide synthase and calpain in hypoxia- inhibitor of the proteolysis of I B, on the circulatory failure and induced proximal tubule injury. Renal Fail 18: , 1996 multiple organ dysfunction caused by endotoxin in the rat. Br J 8. Sorimachi H, Ishiura S, Suzuki K: Structure and physiological Pharmacol 121: , 1997 function of calpain. Biochem J 328: , Chatterjee PK, Zacharowski K, Cuzzocrea S, et al: Inhibitors of 9. Watt F, Molloy PL: Specific cleavage of transcription factors by poly (ADP-ribose) synthetase reduce renal ischemia-reperfusion the thiol protease, m-calpain. Nucleic Acids Res 21: , 1993 injury in the anesthetized rat in vivo. FASEB J 14: , Liu ZQ, Kunimatsu M, Yang JP, et al: Proteolytic processing of 30. McDonald MC, Mota-Filipe H, Paul A, et al: Calpain inhibitor nuclear factor- B by calpain in vitro. FEBS Lett 385: , 1996 I attenuates the activation of nuclear factor B and the multiple 11. Xie Q, Kashiwabara Y, Nathan C: Role of transcription factor organ injury/failure caused by hemorrhagic shock. FASEB J NF- B/Rel in induction of nitric oxide. J Biol Chem 269: , 15: , Stogdale L: Correlation of changes in blood chemistry with pathological 12. Crofford LJ, Tan B, McCarthy CJ, Hla T: Involvement of nuclear changes in the animal s body. II. Electrolytes, kidney func- factor- B in the regulation of cyclooxygenase-2 expression tion tests, serum enzymes, and liver function tests. J S Afr Vet by interleukin-1 in rheumatoid synoviocytes. Arthritis Rheum 40: Assoc 52: , , Guder WG, Ross BD: Enzyme distribution along the nephron. 13. Mehdi S: Cell-penetrating inhibitors of calpain. Trends Biol Sci Kidney Int 26: , : , Cuzzocrea S, Sautebin L, De Sarro G, et al: Role of IL-6 in 14. Wang KKW, Yuen P-W: Development and therapeutic potential the pleurisy and lung injury caused by carrageenan. J Immunol of calpain inhibitors. Adv Pharmacol 37: , : , Rami A, Krieglstein J: Protective effects of calpain inhibitors 33. Harrison DJ, Kharbanda R, Scott Cunningham D, et al: Distriagainst neuronal damage caused by cytotoxic hypoxia in vitro and bution of glutathione S-transferase isoenzymes in human kidney: ischemia in vivo. Brain Res 609:67 70, 1993 Basis for possible markers of renal injury. J Clin Pathol 42: , 16. Iwamoto H, Miura T, Okamura T, et al: Calpain inhibitor I reduces 1989 infarct size and DNA fragmentation of myocardium in ischemic/ 34. Bosomworth MP, Aparicio SR, Hay AWM: Urine N-acetyl- reperfused hearts. J Cardiovasc Res 33: , D-glucosaminidase: A marker of tubular damage? Nephrol Dial 17. Kohli V, Gao W, Camargo CA, Clavien PA: Calpain is a mediator Transplant 14: , 1999 of preservation-reperfusion injury in rat liver transplantation. Proc 35. Mullane KM, Kraemer R, Smith B: Myeloperoxidase activity as Natl Acad Sci USA 94: , 1997 a quantitative assessment of neutrophil infiltration into ischemic 18. Sorimachi Y, Harada K, Saido TC, et al: Downregulation of myocardium. J Pharmacol Methods 14: , 1985 calpastatin in rat heart after brief ischemia and reperfusion. J 36. Ohkawa H, Ohishi N, Yagi K: Assay for lipid peroxides in animal Biochem 122: , 1997 tissues by thiobarbituric acid reaction. Anal Biochem 95: , 19. Edelstein CL, Weider ED, Yaqoob MM, et al: The role of cysteine 1979 proteases in hypoxia-induced rat renal proximal tubular injury. 37. McWhinnie DL, Thompson JF, Taylor HM, et al: Morphometric Proc Natl Acad Sci USA 92: , 1995 analysis of cellular infiltration assessed by monoclonal antibody 20. Edelstein CL, Yaqoob MM, Alkhunaizi AM, et al: Modulation labelling in sequential human renal allograft biopsies. Transplantaof hypoxia-induced calpain activity in rat renal proximal tubules. tion 42: , Hottendorf GH: Renal morphologic-functional correlations and Kidney Int 50: , 1996 drug development, in Acute Renal Failure Correlations Between 21. Edelstein CL, Ling H, Gengaro PE, et al: Effect of glycine on Morphology and Function, edited by Solez K, Wheton A, New prelethal and postlethal increases in calpain activity in rat renal York, Marcel Dekker, 1984, pp proximal tubules. Kidney Int 52: , Edelstein CL: Calcium-mediated proximal tubular injury: What is 22. Schnellmann RG, Williams SW: Proteases in renal cell death: the role of cysteine proteases? Nephrol Dial Transplant 15: , Calpains mediate cell death produced by diverse toxicants. Renal 2000 Fail 20: , Paller MS: Pathophysiologic mechanisms of acute renal failure, 23. Doctor RB, Bennett V, Mandel LJ: Degradation of spectrin and in Mechanisms of Injury in Renal Disease and Toxicity, edited by ankyrin in the ischemic rat kidney. Am J Physiol 264:C1003 C1013, Goldstein RS, Ann Arbor, CRC Press, 1994, pp Venkatachalam MA, Bernard DB, Donohue JF, Levinsky NG: 24. Takaoka M, Itoh M, Hayashi S, et al: Proteasome participates in Ischemic damage and repair in the rat proximal tubule: Differences the pathogenesis of ischemic acute renal failure in rats. Eur J among the S 1,S 2 and S 3 segments. Kidney Int 54:31 49, 1978 Pharmacol 384:43 46, Donadio C, Tramonti G, Lucchesi A, et al: -Glutamyltransferase 25. Lin YC, Brown K, Siebenlist U: Activation of NF B requires is a reliable marker for tubular effects of contrast media. Renal proteolysis of the inhibitor I B- : Signal-induced phosphorylation Fail 20: , 1998 of I B- alone does not release active NF B. Proc Natl Acad Sci 44. Feilleux-Duche S, Garlatti M, Aggerbeck M, et al: Cell specific USA 92: , 1995 regulation of cytosolic aspartate aminotransferase by glucocorti- 26. Griscavage, JM, Wilk S, Ignarro LJ: Inhibitors of proteosome coids in the rat kidney. Am J Physiol 265:C1298 C1305, 1993 pathway interfere with induction of nitric oxide synthase in macro- 45. Cutrin JC, Zingaro B, Camandola S, et al: Contribution of phages by blocking activation of transcription factor NF- B. Proc glutamyl transpeptidase to oxidative damage of ischemic rat kid- Natl Acad Sci USA 93: , 1996 ney. Kidney Int 57: , Milligan SA, Owens MW, Grisham MB: Inhibition of I B- and 46. Umezawa H: Structures and activities of protease inhibitors of I B proteolysis by calpain inhibitor I blocks nitric oxide synthase microbial origin. Methods Enzymol 45: , 1976 synthesis. Arch Biochem Biophys 335: , Siebenlist U, Franzoso G, Brown K: Structure, regulation and 28. Ruetten H, Thiemermann C: Effect of calpain inhibitor I, an function of NF- B. Annu Rev Cell Biol 10: , 1994