Renal Ischemia/Reperfusion Injury: Functional Preservation by Anti-Activated ␤1 Therapy

Ana Molina,* Marı´a Ubeda,* Marı´a M. Escribese,* Laura Garcı´a-Bermejo,* David Sancho,† ʈ Guillermo Pe´rez de Lema,‡ Fernando Lian˜o,§ Carlos Caban˜as, Francisco Sa´nchez-Madrid,† and Francisco Mampaso* *Department of Pathology, Hospital Ramo´n y Cajal, Universidad de Alcala´, Madrid, Spain; †Department of Immunology, Hospital de la Princesa, Universidad Auto´noma de Madrid, Madrid, Spain; ‡Medizinische Poliklinic der Ludwig Maximillians-Universitat, Munich, Germany; §Department of Nephrology, Hospital Ramo´n y Cajal, ʈ Universidad de Alcala´, Madrid, Spain; and Institute of Pharmacology and Toxicology (Centro Mixto CSIC-UCM), Facultad de Medicina, Universidad Complutense, Madrid, Spain

Renal ischemia/reperfusion injury (IRI) is an important cause of acute renal failure. Cellular and molecular responses of the kidney to IRI are complex and not fully understood. ␤1 localize to the basal surface of tubular interacting with components of the basal membrane, including IV. Whether preservation of tubular epithelium integrity could be a therapeutic approach for IRI was assessed. The effects of HUTS-21 mAb administration, which recognizes an activation-dependent epitope of ␤1 integrins, in a rat model of IRI were investigated. Preischemic HUTS-21 administration resulted in the preservation of renal functional and histopathologic parameters. Analyses of activated ␤1 integrins expression and focal adhesion kinase phosphorylation suggest that its deactivation after IRI was prevented by HUTS-21 treatment. Moreover, HUTS-21 impaired the inflammatory response in vivo, as indicated by inhibition of proin- flammatory mediators and the absence of infiltrating cells. Ex vivo adhesion assays using reperfused kidneys revealed that HUTS-21 induced a significant increase of epithelial cell attachment to collagen IV. In conclusion, the data provide evidence that HUTS-21 has a protective effect in renal IRI, preventing tubular epithelial cell detachment by preserving activated ␤1 integrins functions. J Am Soc Nephrol 16: 374–382, 2005. doi: 10.1681/ASN.2004070528

schemia/reperfusion injury (IRI) is an important cause of the basement membrane (BM) (15). Recently, attention has fo- acute renal failure in native kidneys and allografts (1). cused on the role of ␤1 integrins in the loss of tubular cells and I Cellular and molecular responses of the kidney to acute tubular lumen occlusion by cellular debris during acute IRI (2). ischemic injury are complex and not fully understood (2–4). Using an in vitro model of renal epithelial cells that were Several studies have examined the role of leukocytes and their exposed to oxidative stress, it was demonstrated redistribution surface adhesion molecules in the pathogenesis of postischemic of the ␣3␤1 integrin-subunit from the basolateral to apical cell renal damage. Leukocyte adhesion molecules seem to facilitate membranes (16). Also, in an in vivo model of IRI, it has been polymorphonuclear neutrophils recruitment during reperfu- proposed that the decrease of ␤1 integrins on basal surfaces sion, being implicated as mediators of renal IRI (5,6). Comple- contributes to tubular epithelia detachment into the lumen. mentary studies using mAb, antisense oligonucleotides, or Such observations provide a potential mechanism to explain ␤ gene “knockout” indicate that blockade of 2 integrins and/or tubular obstruction and backleak of glomerular filtrate after intercellular adhesion molecule-1 attenuates IRI in some exper- acute IRI (17). Administration of synthetic arginine-glycine- imental models (7–13). Furthermore, it was proposed recently aspartic acid peptides that block integrins ameliorates renal IRI, that T cells might also mediate IRI through the interaction with suggesting that arginine-glycine-aspartic acid peptides inhibit renal tubular epithelial cells (14). cell– in the tubular lumen, thus reducing intratu- ␤ In adult kidney, members of the 1 integrin subfamily are the bular obstruction (18–20). most common and localize to basal surfaces of tubular epithe- Integrin activation and deactivation are not fully understood lia, interacting with extracellular matrix (ECM) components of in in vivo processes. It is conceivable that after IR, ␤1 integrins, which are basally localized in the stressed tubular cells, are ␤ Received July 6, 2004. Accepted October 15, 2004. deactivated. Conversely, as a new epithelium reforms, 1 inte- grins again become activated and return to an exclusively basal Published online ahead of print. Publication date available at www.jasn.org. location. The state of integrin activation can be assessed by a Address correspondence to: Dr. Francisco Mampaso, Hospital Ramo´n y Cajal, group of mAb (HUTS) that selectively recognize ␤1 integrins in Departamento de Patologı´a, Carretera de Colmenar, Km 9.1, 28034 Madrid, Spain. Phone: 34-91-3368052; Fax: 34-91-3369016; E-mail: fmampaso.hrc@salud. their active form (21). In fact, an activated epitope has been madrid.org defined in the rat using the anti-human HUTS-21 mAb, which

Copyright © 2005 by the American Society of Nephrology ISSN: 1046-6673/1602-0374 J Am Soc Nephrol 16: 374–382, 2005 ␤1 Integrin in Ischemia/Reperfusion Injury 375 recognizes ␤1 integrins on rat lymphocytes after activation with concentration was quantified by Bradford colorimetric assay (24). Equal ϩ ϩ divalent cations Mn2 and Hg2 . This mAb confers protective amount of proteins (30 ␮g) were separated by 10% SDS-PAGE and effects in an in vivo autoimmune nephritis model (22). immunoblotted onto polyvinylidene difluoride membranes (Bio-Rad, ␤ ␤ We assessed herein the effects of HUTS-21 systemic admin- Hercules, CA). 1 integrin expression was detected using anti- 1 anti- body (M-106; Santa Cruz) and ECL Western blotting detection system istration, infused before the renal artery clamp in an IRI rat (Amersham Pharmacia Biotech., Buckinghamshire, UK). Results were model. A single administration of this anti-activated ␤1 mAb expressed as percentage of ␤1 integrin expression of sham-operated rat resulted in the preservation of renal function and prevention of kidneys. tissue damage after ischemic insult. Quantitative Reverse Transcriptase–PCR Materials and Methods Total RNA was obtained from snap-frozen kidney tissue using the Animals Ultraspec RNA isolation system (Biotecx, Houston, TX). Two micro- Male Sprague-Dawley rats (IFFA-Credo, Paris, France) that weighed grams of DNaseI-treated RNA were reverse-transcribed with MuLV 180 to 200 g were used throughout the experiments. Animals were reverse transcriptase (Roche, Indianapolis, IN). mRNA expression was treated according to the institutional guidelines that are in compliance quantified by real-time PCR following the manufacturer’s instructions with the National Institutes of Health Guide for the Care and Use of (Lightcycler rapid thermal cycler; Roche) using the primers specific for Laboratory Animals. exon sequences (5Ј33Ј): ATG CCA TCC TGC GTC TGG ACC TGG C (␤-, sense), AGC ATT TGC GGT GCA CGA TGG C (␤-actin, Surgery and Experimental Protocol antisense), CAC CTC TCA AGC AGA GCA CAG (IL-1␤, sense), GGG Rats were anesthetized with an inhaled anesthesia mixture of isoflu- TTC CAT GGT GAA GTC AAC (IL-1␤, antisense), CCA GGA GAA rane 2% (Abbott Laboratories Ltd., Queenborough, Kent, England) and AGT CAG CCT CCT (TNF-␣, sense), TCA TAC CAG GGC TTG AGC oxygen 1 L/min and placed on a temperature-regulated table (39°C) to TCA (TNF␣, antisense), CCC TTC CGA AGT TTC TGG CAG CAG maintain body temperature. Renal ischemia was induced by clamping (inducible nitric oxide synthase [iNOS], sense), GGG CTC CTC CAA both renal pedicles during 45 min. Vehicle group (n ϭ 36), received an GGT GTT GCC C (iNOS-antisense). intravenous injection of phosphate buffer (pH 7.4) 5 min before clamp- ϭ ing. The HUTS-treated group (n 36) received a single intravenous Isolation of Proximal Tubular Epithelial Cells ␤ injection of 0.28 g/100 g body wt anti-activated 1 integrin HUTS-21 Proximal tubular epithelial cells (PTEC) were obtained from rat ϭ mAb (19) 5 min before clamping. The sham-operated group (n 8) kidney cortices that were subjected to 45 min of ischemia and3hof underwent the same surgical procedure, except that the clamp was not reperfusion or sham operation, following the process described by applied. Blood samples and kidneys were collected at different reper- Vinay et al. (25) with some modifications. Briefly, renal cortices were fusion times and processed for different studies. incubated in gassed Krebs-Henseleit saline (pH 7.4) that contained 0.1 mg/ml collagenase type I (Sigma, St. Louis, MO) at 37°C for1hand Functional and Histopathologic Studies sieved throughout 0.250- to 0.035-mm mesh. Cell populations were Serum creatinine was determined using a modified Jaffe’s reaction, separated by gradient centrifugation in 50% Percoll (Sigma) solution and blood urea nitrogen (BUN) was measured on the AEROSET System (Krebs-Henseleit saline:Percoll, 1:1, gassed for1honice). PTEC that (Abbott Laboratories, Abbott Park, IL). Kidneys were processed for were obtained from the appropriate band were immediately used for light microscopy examination according to standard procedures, and cell adhesion experiments. sections were stained with hematoxylin and eosin and periodic acid- Schiff. Histopathologic changes were analyzed for tubular epithelial Cell Adhesion Assay cell necrosis, tubular dilation, proteinaceous casts, and medullary con- Cell adhesion assay was performed as described (26,27). PTEC (3 ϫ gestion as suggested by Racusen (23). The presence of interstitial leu- 104 per well), prepared as above, were incubated with 5 ␮g/ml kocyte cell infiltrates was assessed by immunohistochemistry. HUTS-21 or irrelevant mAb (anti–N- [H-63; Santa Cruz]; anti– I-A [Serotec]) or without mAb in 1 ␮g/ml collagen IV (Sigma) or Immunohistochemistry Studies BSA-coated wells. Nonadherent cells were removed, and the percent- Immunostaining was performed using an indirect immunoperoxi- age of attached cells was calculated by measuring absorbance at 540 nm dase technique. Snap-frozen renal tissues were staining with anti-␣1 after fixation and staining with 0.5% crystal violet in 20% methanol. integrin (Chemicon, Temecula, CA) and HUTS-21 mAb. Paraffin-em- bedded renal sections were stained with anti-CD68 (ED1), anti-CD3 Statistical Analyses (1F4; Serotec, Oxford, UK), anti-focal adhesion kinase (FAK; C903), and Results are expressed as arithmetic means (Ϯ SD). Statistical com- anti-phosphorylated FAK (pFAK [Y397]-R; Santa Cruz Biotechnology, parisons between groups were performed by t test. The difference was Santa Cruz, CA). Enumeration of interstitial infiltrating macrophages Ͻ ϩ ϩ considered to be significant at P 0.05. (CD68 ) and T cells (CD3 ) was determined by counting the total number of positive-labeled cells examined in 10 randomly chosen areas of interstitial infiltrates. Results IRI Diminishes Expression of Activated ␤1 Integrins on the Basal Cell Surface of PTEC Western Blot Analysis ␤ Pieces of snap-frozen kidneys were homogenized in lysis buffer that Previous studies suggest a role for 1 integrins in the tubular contained 100 mM buffer phosphate (pH 7.6), 150 mM NaCl, 10 mM cell detachment and tubular lumen occlusion after IRI (17–20). ␤ EDTA, 0.5% Triton X-100, and a cocktail of protease and phosphatase Here, we analyzed the expression pattern of activated 1 inte- inhibitors and then were incubated for 60 min on ice and centrifuged at grins in a rat model of bilateral renal ischemia. As shown in 13,000 ϫ g for 30 min at 4°C. Supernatants were collected, and protein Figure 1, activated ␤1 integrins (HUTS-21 mAb staining) local- 376 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 374–382, 2005

Figure 1. Effect of HUTS-21 administration on ␤1 integrin expression in ischemic renal tissue. (A) Immunohistochemical detection of activated ␤1 and ␣1 integrins. Activated ␤1 integrin localized exclusively to basal cell surface of proximal and distal tubule in sham-operated rat kidney (a), decreasing the immunostaining after renal ischemia (0 h reperfusion; b) and 24 h of reperfusion (c). Immunodetection of ␣1 integrin in sham-operated rats (d), after renal ischemia (e) and 24 h of reperfusion (f), showed no changes. (B) Densitometric quantification (as a percentage of control, sham-operated) of ␤1 expression under the different reperfusion times. Results represent the mean Ϯ SD; n ϭ 4. Representative Western blot analysis of ␤1 integrin in total ischemic kidney lysates is also shown (bottom). Magnification, ϫ500 in A, immunoperoxidase (PO). ize exclusively to the basal cell surface of proximal and distal tubules in the cortex and in the outer and inner medulla in normal kidneys (sham operated; Figure 1A, a). In contrast, much weaker immunostaining was observed after renal isch- emia in the basal cellular layer of the same tubular structures (Figure 1A, b). After 24 h of reperfusion, activated ␤1 integrins that were basally localized in the stressed tubular cells were deactivated (Figure 1A, c), as indicated by the loss of HUTS-21 staining. Anti-␣1 integrin positivity, which is found associated with ␤1 chain in PTEC, revealed no change in sham-operated animals after renal ischemia and after 24 h of reperfusion (Figure 1A, d through f). Quantitative analyses of ␤1 control expression by Western blot showed a slight but not statistically significant decrease of ␤1 levels during the reperfusion process (Figure 1B).

Preischemic HUTS-21 Administration Confers Renal Protection Animals that underwent renal ischemia showed marked de- terioration of renal functional parameters with significant in- crease in both creatinine and BUN levels, compared with basal levels from sham-operated and normal rats (Figure 2). Serum Figure 2. Effect of HUTS-21 mAb administration on renal func- creatinine and BUN levels exhibited a steady increase from 0 h tion after ischemia. Serum creatinine (A) and blood urea nitro- of reperfusion, reaching the maximal values after 24 h and gen (B) levels were measured. Rats that received intravenous f declining thereafter to reach background levels after 7 d (data injection of HUTS-21 mAb ( ) showed an accelerated recovery e not shown). Preischemic administration of HUTS-21 produced as compared with vehicle rats ( ). Control, dotted bar; Sham, Ϯ P Ͻ n ϭ a remarkable renal functional protection, as shown in the serum hatched bar. Values are mean SD. * 0.001; 6. creatinine (0.40 Ϯ 0.098 versus 1.60 Ϯ 0.324 mg/dl; P Ͻ 0.001) and BUN levels (25 Ϯ 11.0 versus 120 Ϯ 11.79 mg/dl; P Ͻ 0.001) at 24 h of reperfusion (Figure 2). Administration of HUTS-21 to logic examination revealed patchy distributed renal lesions in sham-operated and normal rats produced no changes in renal the vehicle group (Figure 3). Renal lesions were evident as early functional parameters (data not shown). as1hofreperfusion: Signs of sublethal injury (Figure 3a, Parallel to the deterioration of functional parameters, histo- arrows) and altered cell adhesion, such as tubular cell loss/ J Am Soc Nephrol 16: 374–382, 2005 ␤1 Integrin in Ischemia/Reperfusion Injury 377

circulating leukocytes into renal interstitium. Although poly- morphonuclear neutrophil infiltration has been widely studied, less is known about the role of macrophages and T cells after IR insult. Here we studied the contribution of both cell popula- tions to IRI (Table 1). Interstitial macrophages were first de- tected as early as 12 h of reperfusion, reaching the maximal number at 24 h (Figure 4a). In addition, a discrete number of T cells began to be detected at 24 h of reperfusion, reaching a prominent number at 72 h (Figure 4b). It is interesting that HUTS-21 administration showed a drastic reduction on both macrophages and T cells (Figure 4, c and d).

HUTS-21 Administration Reduced Proinflammatory Cytokines and iNOS mRNA Levels in Renal IRI Because proinflammatory cytokines are keys for tissue in- flammatory cell infiltration and its expression during IRI re- sults in iNOS upregulation, we studied the effects of HUTS-21 administration on its local mRNA expression (Figure 5). Real- time quantitative RT-PCR analysis of renal tissue revealed sim- ilar kinetic profiles for TNF-␣ and IL-1␤ mRNA expression in ischemic kidneys (vehicle group), showing a maximum value at 12 h after ischemia, declining thereafter, and reaching basal levels at3dofreperfusion. By contrast, the analysis of this proinflammatory cytokine mRNA expression in the HUTS- treated group showed no changes, maintaining the basal values along reperfusion times. Local expression of iNOS mRNA showed a delay compared with proinflammatory cytokines. HUTS-21 administration sig- nificantly prevented the increase in local iNOS mRNA expres- Figure 3. Effect of HUTS-21 administration on the time course of sion. the histopathologic features of ischemic kidneys. Kidney sam- ples from the vehicle group showed evident renal lesions at the HUTS-21 Increases Adhesiveness of Ischemic PTEC early time point being maximal at 24 h of reperfusion. The To investigate the mechanisms of effects of HUTS-21 admin- HUTS-treated kidneys revealed the absence of renal tissue le- sions. (a and e) One hour of reperfusion (b and f);8hof istration over IRI, we examined its effects on cell adhesion to reperfusion (c and g); 12 h of reperfusion (d and h); 24 h of ECM. PTEC from kidneys of IRI and control rats were incu- reperfusion. Arrows, sublethal injury (a), detached epithelial bated with HUTS-21, with irrelevant antibody, or without an- cells (b and c). *Intratubular casts. Magnification, ϫ400, peri- tibody, and PTEC adhesion to collagen IV was analyzed. Cell odic acid-Schiff. adhesion from ischemic or control PTEC on the BSA matrix in the presence or absence of all antibodies showed basal values (data not shown). As shown in Figure 6, adhesion of control pauci-cellular epithelium with flattening as well as detached PTEC was not further enhanced in the presence of HUTS-21. By cells in tubular lumina, were observed. At8hofreperfusion, a contrast, adhesion of ischemic PTEC to collagen IV in the massive PTEC detachment (Figure 3b, arrows) was observed. presence of HUTS-21 was threefold higher compared with isch- Renal lesions reached a peak between 12 and 24 h of reperfu- emic PTEC, either in the absence or in the presence of control sion (Figure 3, c and d, respectively). At that time, more severe mAb. and lethal injury (tubular epithelial cell necrosis) was found, being most prominent in the outer medullary stripe, but also Ischemia-Induced FAK Dephosphorylation Is Prevented by with patchy involvement of the cortical proximal segments. In HUTS-21 Administration addition, tubular dilation with proteinaceous and epithelial FAK activity is regulated by phosphorylation after activation cellular cast-filled lumen in the inner zone of the medulla of integrins, including ␤1 integrins, among other factors. To (Figure 3, c and d, asterisk) and blood congestion in the outer investigate further in vivo ␤1 activation in the outcome of IRI, zone of the inner stripe of the medulla were observed. The we evaluated the phosphorylation level of FAK (pFAK) during preischemic treatment with HUTS-21 showed complete ab- IR and upon administration of HUTS-21. Figure 7 shows pFAK sence of all renal lesions at 12 and 24 h of reperfusion (Figure 3, levels in the tubular epithelium using a specific antibody that g and h, respectively), which paralleled the accelerated recov- recognizes FAK phosphorylation at Y397 residue. Ischemia in- ery of renal function. duced FAK dephosphorylation. During reperfusion (3 to 24 h), An additional important aspect of IRI is the recruitment of pFAK levels remained lower than the basal ones (sham-oper- 378 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 374–382, 2005

Table 1. Macrophage and T cell infiltration in the kidney from rats that were subjected to renal ischemia followed by reperfusion

Treatment Group Reperfusion Time (h) Macrophages (ϫ 10) T Cells (ϫ 10)

Sham operated 3.75 Ϯ 0.96 3.00 Ϯ 0.82 Vehicle 0 4.25 Ϯ 0.96 1.75 Ϯ 1.71 3 4.00 Ϯ 1.63 3.00 Ϯ 1.63 8 3.25 Ϯ 2.06 4.25 Ϯ 1.26 12 52.50 Ϯ 18.48 3.50 Ϯ 2.08 24 204.75 Ϯ 41.53 10.00 Ϯ 2.58 72 117.50 Ϯ 19.23 113.00 Ϯ 13.93 HUTS treated 0 2.50 Ϯ 2.08 2.75 Ϯ 1.71 3 3.00 Ϯ 1.41 3.25 Ϯ 0.96 8 4.50 Ϯ 1.29 4.25 Ϯ 2.22 12 5.50 Ϯ 1.29a 3.75 Ϯ 1.50 24 4.75 Ϯ 2.36a 3.00 Ϯ 1.41b 72 4.50 Ϯ 2.38a 4.00 Ϯ 1.82a

aP Ͻ 0.001, bP Ͻ 0.05 versus vehicle at the corresponding time.

Figure 4. Cellular infiltration into the postischemic kidney. Mac- ϩ ϩ rophages (CD68 cells) and T lymphocytes (CD3 cells) were detected in kidneys from the vehicle group (a and b, respec- tively), whereas no inflammatory cells were observed in HUTS- treated kidneys (c and d). Magnification, ϫ600 in a and b, PO; ϫ400, c and d, PO.

ated group). HUTS-21 administration preserved FAK phos- phorylation from the early times of reperfusion, even at higher levels than basal pFAK. No changes in total FAK expression were detected by immunohistochemistry, neither upon IRI nor upon HUTS-21 administration (data not shown). These results suggest that IRI induces deactivation of ␤1 integrin that was restored after HUTS-21 administration.

Figure 5. Proinflammatory cytokines and inducible nitric oxide Discussion ␤ synthase mRNA analysis. Real-time quantitative reverse tran- 1 integrins function mainly as cellular receptors for ECM scriptase–PCR analysis of mRNA from sham-operated (‚), ve- ␤ proteins. The importance of 1 integrins in the architectural hicle (F), and HUTS-treated (E) kidneys. Results represent the and functional maintenance of the tubular compartment of mean Ϯ SD, n ϭ 4, and are normalized to ␤-actin expression kidney tissue is known (28). ␤1 integrin–ECM interaction fa- measured in parallel in each sample. *P Ͻ 0.001; **P Ͻ 0.01; vors cell adhesiveness and epithelium integrity (15). Moreover, ***P Ͻ 0.05. J Am Soc Nephrol 16: 374–382, 2005 ␤1 Integrin in Ischemia/Reperfusion Injury 379

the ␣4␤1 integrins are implicated in cell–cell interaction be- tween circulating leukocytes and activated endothelium (29). Here, we report the protective effects of systemic HUTS-21 administration in a rat model of renal IRI. HUTS-21 recognizes an activation-dependent epitope on the human ␤1 integrin chain and cross-reacts with a similar activated epitope in the rat (21,22). We have focused our studies on both the PTEC detach- ment from the and the adhesion and extravasation of circulating leukocytes into the renal interstitium. Figure 6. Effect of HUTS-21 mAb on proximal tubular epithelial Animals that underwent renal ischemia showed a marked cells (PTEC) adhesion to collagen intravenously. PTEC from deterioration of renal functional parameters with significant control (e) and ischemia/reperfusion injury (IRI; f) kidneys increase in both creatinine and BUN levels. Histopathologic were treated with HUTS-21 mAb or with irrelevant antibodies renal lesions showed a massive PTEC detachment, with the (anti–I-A, anti–N-cadherin). Adhesion of control PTEC to BSA renal lesions reaching a peak at 24 h of reperfusion. Preischemic in the absence of antibody was considered as basal adhesion, HUTS-21 administration resulted in an accelerated improve- and the percentage of adhesion was calculated according to ment of renal function and complete absence of renal tissue these values in each independent experiment. PTEC that were lesions at 24 h of reperfusion. cultured in the absence of antibody served as reference. *P Ͻ Cellular adhesive properties are regulated by selective ex- 0.001. pression of the integrin repertoire and by the modulation of their binding properties. A distinctive feature of integrins is the ability to modulate the level of adhesiveness rapidly and re- versibly (30). The process of activation is a direct consequence of integrin conformational changes resulting in an increased ligand binding affinity. Consequently, a potential strategy to maintain tubular cell integrity would be to avoid ␤1 integrin deactivation. It is interesting that binding of HUTS-21 to par- tially activated ␤1 integrins results in freezing active conforma- tion of these receptors and enhancement of cell adhesion to ␤1 integrin ligands (21). Our experiments addressed the role played by activated ␤1 integrins in the course of renal IRI. Our data showed that during IRI, ␤1 integrins are deactivated as reflected in the loss of HUTS-21 staining on the basal cellular layer of stressed PTEC as well as FAK dephosphorylation. These results could not be explained on the basis of a loss of ␤1 integrin expression, because Western blot analysis showed no significant changes in its expression levels during IRI. ␤1 integrins act as cell surface receptors for ECM proteins and extracellular signal transducers into the cell, including actin reorganization. Thus, integrins clustered to the basolateral surface of adherent cells to form focal adhesion complex (FAC) (31). One important component of this adhesion complex is the p125 FAK that can be activated by autophos- phorylation, and this activation is required for the FAC forma- tion and function (revised in 32). Phosphorylated FAK can bind the cytoplasmic domain of ␤1 integrins and activate actin cy- toskeleton assembly (33–35). Activation/deactivation of ␤1 in- tegrins can be estimated indirectly by FAK phosphorylation, as assessed in this study. Previous reports have demonstrated that ischemia/hypoxia induced FAK dephosphorylation in vivo (36) and in vitro (37). Our results demonstrate that renal ischemia by Figure 7. Effect of HUTS-21 administration on expression of itself induces FAK dephosphorylation, correlating with ␤1 in- pFAK. Immunohistochemical staining of kidneys from sham- tegrin deactivation. Moreover, reperfusion induced slight ␤1 operated, vehicle, and HUTS-treated groups. Note the in- creased intensity of staining with anti-pFAK mAb in the HUTS- integrin activation but was not sufficient to increase FAK phos- treated group in all reperfusion times compared with vehicle. phorylation. However, HUTS-21 that recognizes partially acti- ␤ (a) Negative control (b); sham operated (c and d);0hofreper- vated 1 integrin “locked” the integrin in the active form, fusion (e and f);3hofreperfusion (g and h); 24 h of reperfusion. resulting in a strong FAK activation by phosphorylation. Magnification, ϫ400 in a through g, PO; ϫ600 in h, PO. Administration of HUTS-21 could maintain the integrity and 380 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 374–382, 2005 functionality of FAC as determined by FAK phosphorylation, tion was able to block circulating leukocyte extravasation, thus resulting in enhanced adhesiveness of PTEC. abrogating the development of interstitial nephritis. Although The diminished immunostaining of HUTS-21, together with we do not know the exact mechanism/s responsible for this the decreased FAK phosphorylation in ischemic kidneys, sug- phenomenon, it is conceivable that the blockade of proinflam- gests the functional disruption of FAC and, consequently, the matory cytokine expression could contribute to the abrogation loss of adhesion. PTEC require anchorage to the BM for normal of the inflammatory response amplification and consequently function, which is mediated by integrins’ binding to ECM to leukocyte cell infiltration. Although kidney interstitium in- proteins, being collagen IV its major component. It was dem- filtration by circulating leukocytes has been considered an im- onstrated previously (38) that ␤1 integrins’ binding to collagen portant source of cytokines, evidence suggests that glomerular IV and not to other ECM proteins elicits mesangial cells and tubular epithelial cells are additional major events in injured PTEC that stimulate cell survival and are cellular sources of TNF-␣ (45). Moreover, the expression of important for the recovery of physiologic functions. Our results proinflammatory cytokines during IRI gives rise to the upregu- revealed that ex vivo HUTS-21 was able to increase the adher- lation of iNOS, leading to tissue injury (46–49). We found that ence of ischemic PTEC to collagen IV, suggesting that in vivo HUTS-21 administration prevents expression of proinflamma- HUTS-21 most probably increases the adhesiveness of ␤1 inte- tory cytokines and iNOS from renal tissue, thus abrogating the grins to the basal lamina, thus contributing to prevent renal inflammatory response. damage. It is conceivable that HUTS-21 administration in- We conclude that in vivo administration of HUTS-21 likely creases the adhesion and focal contact between PTEC and basal keeps ␤1 integrins in an activated state, as occurs in in vitro lamina, thus diminishing cell detachment and casts formation. studies (50). Consequently, the sustained adhesion of PTEC to Renal injury after ischemia seems to be a consequence of ␤1 integrin ligands on BM could attenuate tubular epithelial tissue hypoxia from renal blood flow deficiency but also from detachment into the lumen and tubular obstruction. The pres- the process of reperfusion, leading to an active inflammatory ervation of tubular epithelial integrity could explain the pro- response (4). Sublethal or even lethal injured proximal tubular tective effect of HUTS-21, supporting renal function improve- and endothelial cells trigger this process through the release of ment, proinflammatory cytokine and iNOS suppression, and a proinflammatory mediators that will promote cellular infiltra- decreased T cell and monocyte/macrophage extravasation into tion. As compared with vehicle rats, HUTS-21 administration the renal interstitium. showed a drastic inhibitory effect on the recruitment of both macrophage and T cell leukocyte populations, as well as on the Acknowledgments expression of proinflammatory cytokines, thus abrogating the This work was support by grants from Spanish Ministerio de Ciencia inflammatory response. Although it is well demonstrated that y Tecnologı´a (MCyT) (SAF2001-1048-C03-02), Comunidad Auto´noma leukocyte adhesion molecule blockade plays a tissue protective de Madrid (CAM) (08.3/0011.1/2001), Fondo de Investigacio´n Sanitaria role in IRI in muscle and heart, it is still a matter of controversy (FIS) (PI02-0346), and Fundacio´n Mutua Madrilen˜a del Automovilista in the kidney. CD11/CD18 and intercellular adhesion mole- de Investigacio´n Me´dica to F.M.; from FIS (PI01-3001) and CAM (08.3/ cule-1 blockade are usually protective in experimental renal IRI 0001.1/2003) to L.G.-B.; and from Fundacio´n Juan March to F.S.- (9,12). In contrast, induction of systemic neutropenia and selec- M. A.M. is funded by a grant from CAM (08.3/0011.1/2001) and M.E. tin function blockade do not have a protective effect, suggesting from Spanish MCyT (SAF2001-1048-C03-02). We thank Dr. Herra´ez Domı´nguez for excellent technical assistance a neutrophil-independent mechanism for renal protection (6). with the renal functional parameters measurements. Although neutrophil infiltration has been widely studied (re- vised in 39), less is known about the role of macrophages and ϩ T cells after IR insult. Macrophages and CD4 cell infiltration References within a few days after IRI was described after warm ischemia 1. Lian˜o F, Pascual J: The Madrid Acute Renal Failure Group. (40) and after experimental cold ischemia (41), with limited Epidemiology of acute renal failure: A prospective multi- information, however, as to its precise topographic localization center community based study. Kidney Int 50: 811–818, and kinetics. Ysebaert et al. (42) reported that only at the later 1996 time of regeneration did a sequential infiltration of macro- 2. Eldestein CL, Ling H, Schrier W: The nature of renal cell phages and T cells become prominent. Nevertheless, it has been injury. Kidney Int 51: 1341–1351, 1997 reported (43,44) that the blockade of CD29-B7 co-stimulatory 3. Thadhant R, Pascual M, Bonventre JV: Acute renal failure. pathway, important for lymphocyte activation, significantly N Engl J Med 334: 1448–1460, 1996 abrogates postischemic renal dysfunction in the rat. Rabb et al. 4. Bonventre JV: Mechanisms of ischemic acute renal failure. (14) recently proposed that T cells might also mediate IRI Kidney Int 43: 1160–1178, 1993 5. Rabb H, O’Meara Y, Maderna P, Coleman P, Brady H: through the interaction with PTEC. These observations concur Leukocytes, cell adhesion molecules and ischemic acute with our findings regarding the early presence of mononuclear renal failure. Kidney Int 51: 1463–1468, 1997 cells in the renal interstitium, supporting their important role in 6. Lauriat S, Linas SL: The role of neutrophils in acute renal renal IRI. failure. Semin Nephrol 18: 498–504, 1998 In a previously reported study, we demonstrated the role of 7. Tajra LC, Martin X, Margonari J, Blanc-Brunat N, Ishibashi ␤1 activated integrins in the development and progression of M, Vivier G, Panaye G, Steghens JP, Kawashima H, Mi- an autoimmune model of nephritis (22). HUTS-21 administra- yasaka M, Treille-Ritouet D, Dubernard JM, Revillard JP: J Am Soc Nephrol 16: 374–382, 2005 ␤1 Integrin in Ischemia/Reperfusion Injury 381

In vivo effects of monoclonal antibodies against rat beta(2) relevance of activated ␤1 integrins in mercury-induced integrins on kidney ischemia-reperfusion injury. J Surg Res nephritis. J Am Soc Nephrol 11: 1075–1084, 2000 87: 32–38, 1999 23. Racusen LC: Biology of disease. Alteration in tubular epi- 8. Haller H, Dragun D, Miethke A, Park JK, Weis A, Lip- thelial cell adhesion and mechanisms of acute renal failure. pouldt A, Gross V, Luft FC: Antisense oligonucleotides for Lab Invest 67: 158–165, 1992 ICAM-1 attenuate reperfusion injury and renal failure in 24. Bradford M: A rapid and sensitive method for the quanti- the rat. Kidney Int 50: 473–480, 1996 fication of microgram quantity of protein utilizing the prin- 9. Kelly KJ, Williams WW, Colvin RB, Bonventre JV: Anti- ciple of protein-dye binding. Anal Biochem 72: 248–254, body to intercellular adhesion molecule 1 protects the kid- 1976 ney against ischemic injury. Proc Natl Acad Sci U S A 91: 25. Vinay P, Gougoux A, Lemieux G: Isolation of a pure sus- 812–816, 1994 pension of rat proximal tubules. Am J Physiol 241: F403– 10. Kelly KJ, Williams WW, Colvin RB, Meeham SM, Springer F411, 1981 TA, Gutierrez-Ramos JC, Bonventre JV: Intercellular adhe- 26. Arroyo AG, Sa´nchez-Mateos P, Campanero MR, Martı´n- sion molecule 1-deficient mice are protected against isch- Padura I, Dejana E, Sa´nchez-Madrid F: Regulation of the emic renal injury. J Clin Invest 97: 1056–1063, 1996 ␤1 integrin-ligand interaction through the ␤1 subunit. J Cell 11. Linus SL, Whittenburg D, Parsons PE, Repine JE: Ischemia Biol 117: 659–670, 1992 increases neutrophil retention and worsens acute renal 27. Luque A, Sa´nchez-Madrid F, Caban˜as C: Functional regu- failure: Role of oxygen metabolites and ICAM-1. Kidney Int lation of the human integrin VLA-1 (CD49a/CD29) by 48: 1584–1591, 1995 divalent cations and stimulatory ␤1 antibodies. FEBS Lett 12. Rabb H, Mendiola CC, Dietz J, Saba SR, Issekutz TB, 346: 278–284, 1994 Abanilla F, Bonventre JV, Ramirez G: Role of CD11a and 28. Pozzi A, Zent R: Integrins: Sensors of extracellular matrix CD11b in acute renal failure in rats. Am J Physiol 267: and modulators of cell function. Nephron Exp Nephrol 94: F1052–F1058, 1994 77–84, 2003 13. Rabb H, Mendiola CC, Saba SR, Dietz J, Smith CW, 29. Lim YC, Wakelin MW, Henault L, Goetz DJ, Yednock T, Bonventre JV, Ramı´rez G: Antibodies to ICAM-1 protect Caban˜asC,Sa´nchez-Madrid F, Lichtman AH, Luscinskas kidneys in severe ischemic reperfusion injury. Biochem Bio- FW: ␣4␤1-Integrin activation is necessary for high-effi- phys Res Commun 211: 67–73, 1995 ciency T-cell subset interactions with VCAM-1 under flow. 14. Rabb H, Daniels F, O’Donnell M, Mahmud H, Saba SR, Microcirculation 7: 201–214, 2000 Keane W, Tang WW: Pathophysiological role of T lympho- 30. Sanchez-Mateos P, Caban˜asC,Sa´nchez-Madrid F: Regula- cytes in renal ischemia-reperfusion injury in mice. Am J tion of integrin function. Semin Cancer Biol 7: 99–109, 1996 Physiol Renal Physiol 279: F525–F531, 2000 31. David FS, Zage PE, Marcantonio EE: Integrins interact with 15. Korhonen M, Ylanne J, Laitinen, L, Virtanen I: The ␣1-␣6 focal adhesion through multiple distinct pathways. J Cell subunits of integrins are characteristically expressed in Physiol 181: 74–82, 1999 distinct segments of developing and adult human 32. Mahdi MH, Andrieu V, Pasquier C: Focal adhesion kinase nephron. J Cell Biol 111: 1245–1254, 1990 regulation by oxidative stress in different cell types. Life 50: 16. Gailit J, Colflesh D, Rabiner I, Simone J, Goligorsky MS: 291–299, 2000 Redistribution and disfunction of integrins in cultured re- 33. Schaler MD, Otey CA, Hildebrand JD, Parsons JJ: Focal nal epithelial cells exposed to oxidative stress. Am J Physiol adhesion kinase and bind to peptides mimicking 264: F149–F157, 1993 beta integrin cytoplasmic domains. J Cell Biol 130: 1181– 17. Goligorsky MS, Lieberthal W, Racusen L, Simon EC: Inte- 1187, 1995 grin receptors in renal tubular epithelium: New insights 34. Chen LM, Bailey D, Ferna´ndez-Valle C: Association of ␤1 into pathophysiology of acute renal failure. Am J Physiol integrin with focal adhesion kinase and paxillin in differ- 264: F1–F8, 1993 entiating Schwann cells. J Neurosci 20: 3776–3784, 2000 18. Noiri E, Gailit J, Sheth D, Magazine H, Gurrath M, Muller 35. Tahiliani PD, Singh L, Auer KL, LaFlamme SE: The role of G, Kessler H, Goligorsky MS: Cyclic RGD peptides ame- conserved amino acids motifs within the integrin ␤3 cyto- liorate ischemic acute renal failure in rats. Kidney Int 46: plasmic domain in triggering focal adhesion kinase phos- 1050–1058, 1994 phorylation. J Biol Chem 272: 7892–7898, 1997 19. Goligorsky MS, Nori E, Kessler H, Romanov V: Therapeu- 36. Zalewska T, Ziemka-Nalecz M, Sarnowska A, Domanska- tic potential of RGD peptides in acute renal failure. Kidney Janik K: Transient forebrain ischemia modulates signal Int 51: 1487–1492, 1997 transduction from extracellular matrix in gerbil hippocam- 20. Noiri E, Romanov V, Forest T, Gailit J, DiBona GF, Miller F, pus. Brain Res 977: 62–69, 2003 Som P, Oster ZH, Goligorsky MS: Pathophysiology of renal 37. Wachtel M, Frei K, Ehler E, Bauer C, Gassmann M, Gloor tubular obstruction: Therapeutic role of synthetic RGD SM: Extracellular signal-regulated protein kinase activa- peptides in acute renal failure. Kidney Int 48: 1375–1385, tion during reoxygenation is required to restore ischemia- 1995 induced endothelial barrier failure. Biochem J 367: 873–879, 21. Luque A, Go´mez M, Puzo´n W, Takada Y, Sa´nchez-Madrid 2002 F, Caban˜as C: Activated conformations of very late activa- 38. Nony PA, Schnellmann RG: Mechanisms of renal cell re- tion integrins detected by a group of antibodies (HUTS) pair an regeneration after acute renal failure. J Pharmacol specific for a novel regulatory region (355–425) of the Exp Ther 304: 905–912, 2003 common ␤1 chain. J Biol Chem 271: 11067–11075, 1996 39. De Greef KE, Ysebaert DK, Ghielli M, Vercauteren S, Nou- 22. Escudero E, Martı´n A, Nieto M, Nieto E, Navarro E, Luque wen EJ, Eyskens EJ, De Broe ME: Neutrophils and ischemia A, Caban˜asC,Sa´nchez-Madrid F, Mampaso F: Functional reperfusion injury. J Nephrol 11: 110–122, 1998 382 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 374–382, 2005

40. Takada M, Nadeau KC, Shaw GD, Marquette KA, Tilney 46. Sekhon CS, Sekhon BK, Singh I, Orak JK, Singh AK: At- NL: The cytokine-adhesion molecule cascade in ischemia/ tenuation of renal ischemia/reperfusion injury by a triple reperfusion injury of the rat kidney. Inhibition by a soluble drug combination therapy. J Nephrol 16: 63–74, 2003 P-selectin ligand. J Clin Invest 99: 2682–2690, 1997 47. Chatterjee PK, Patel MS, Sivarajah A, Kvale EO, Dugo L, 41. Azuma H, Nadeau K, Takada M, Mackenzie HS, Tilney Cuzzocrea S, Brown PA, Stewart KN, Mota-Filipe H, Britti NL: Cellular and molecular predictors of chronic renal D, Yaqoob MM, Thiemermann C: GW274150, a potent and dysfunction after initial ischemia/reperfusion injury of a highly selective inhibitor of iNOS, reduces experimental single kidney. Transplantation 64: 190–197, 1997 renal ischemia/reperfusion injury. Kidney Int 63: 853–865, 42. Ysebaert DK, De Greef KE, Vercauteren SR, Ghielli M, 2003 Verpooten GA, Eyskens EJ, De Broe ME: Identification and 48. Kielar ML, Jeyarajah DR, Zhou XJ, Lu CY: Docosahexae- kinetics of leukocytes after severe ischaemia/reperfusion noic acid ameliorates murine ischemic acute renal failure renal injury. Nephrol Dial Transplant 15: 1562–1574, 2000 and prevents increases in mRNA abundance for both TNF- 43. Chandraker A, Takada M, Nadeau K, Peach R, Tilney N, alpha and inducible nitric oxide synthase. J Am Soc Nephrol Sayeghi M: CD28–B7 blockade in organ dysfunction sec- 14: 389–396, 2003 ondary to cold ischemia/reperfusion injury. Kidney Int 52: 49. Ling H, Edelstein C, Gengaro P, Meng X, Lucia S, Knotek 1678–1684, 1997 M, Wangsiripaisan A, Shi Y, Schrier R: Attenuation of renal 44. Takada M, Chandraker A, Nadeau K, Sayeghi M, Tilney N: ischemia-reperfusion injury in inducible nitric oxide The role of the B7 costimulatory pathway in experimental synthase knockout mice. Am J Physiol 277: F383–F390, cold ischemia/reperfusion injury. J Clin Invest 100: 1199– 1999 1203, 1997 50. Go´mez M, Luque A, del Pozo MA, Hoog N, Sa´nchez- 45. Donnahoo KK, Shames BD, Harken AH, Meldrum DR: The Madrid F, Caban˜as C: Functional relevance during lym- role of tumor necrosis factor in renal ischemia-reperfusion phocyte migration and cellular localization of activated ␤ 1 injury. J Urol 162: 196–203, 1999 integrins. Eur J Immunol 27: 8–16, 1997