Focal adhesion signaling in acute renal failure Alderliesten, M.C.

Alderliesten, M.C.

Citation

Alderliesten, M. C. (2009, February 11). Focal adhesion signaling in acute renal failure. LACDR, Division of Toxicology, Faculty of Science, Leiden University.

Retrieved from https://hdl.handle.net/1887/13803

Version: Corrected Publisher’s Version

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden Downloaded from: https://hdl.handle.net/1887/13803

Note: To cite this publication please use the final published version (if

applicable).

4

Chapter 4

Focal Adhesion Kinase Signaling is Involved Ischemia/Reperfusion Induced Acute Renal Failure

Maaike C. Alderliesten¹, Yu Qin¹, Bernd Dworniczak², Petra Pennenkamp², Hilary Beggs³, Takaharu Ichimura⁴, Joseph Bonventre⁴, Bob van de Water¹

1Division of Toxicology, Leiden Amsterdam Center for Drug Research, Leiden University, Einsteinweg 55, P.O. Box 9502, 2300RA, Leiden, the Netherlands.

2Human Molecular Genetics, Institute of Human Genetics, WWU Muenster, Germany.

3Department of Ophthalmology and Physiology, University of California, San Francisco, USA.

4Department of Medicine, Renal Division, Brigham and Women’s Hospital, Harvard Medical School, Boston,USA.

Abstract

Focal adhesion kinase (FAK) is a non-receptor protein tyrosine kinase located at the focal adhesions. During renal ischemia/reperfusion the focal adhesions show dynamic reorga- nization. In addition FAK is differentially phosphorylated on its tyrosine residues during reperfusion indicating a role for FAK activation and downstream signaling events dur- ing renal ischemia/reperfusion. In this study we used a conditional renal proximal tubule specific FAK knockout mouse model to investigate the role of FAK during renal ischemia/

reperfusion injury. Mice with a floxed fak gene (FAK^loxP/loxP) were crossed with transgenic mice expressing tamoxifen-inducible Cre recombinase under control of the γGT-promoter (γGT-CreER^T2), thereby generating FAK^loxP/loxP//γGT-Cre-ER^T2 mice. Tamoxifen treatment caused fak recombination (FAKΔloxP/ΔloxP) followed by a reduction in FAK protein levels at the focal adhesions in the proximal tubules. Mice were subjected to unilateral ischemia by renal pedicle clamping followed by reperfusion for 24 h. Scoring renal injury in H&E- stained sections indicated that the FAKΔloxP/ΔloxP mice are significantly less susceptible to I/R injury than the FAK^loxP/loxP littermates. Similar findings were observed after immunostaining for a predictive marker of renal injury, kidney injury molecule-1 (KIM-1). Interestingly, the protection against I/R coincided with decreased immune cell infiltration and endothelium activation of the injured renal tissue. Our data suggest a model whereby FAK drives a cel- lular stress response in the reperfusion phase after an ischemic period that culminates in immune cell infiltration and renal tissue injury.

Introduction

Acute renal failure (ARF), often induced by ischemia/reperfusion (I/R), is associated with a high morbidity and mortality that has not changed over the past fifty years1. Detailed insight in the cellular and molecular mechanisms of ARF is required to develop novel therapeutic approaches that minimize injury and hasten recovery.

During I/R injury subsets of proximal tubule cells detach from the extracellular matrix (ECM) and are expelled into the lumen^2-4. Adhesion of cells to the ECM provides the cells with environmental signals necessary to control cellular processes like survival, prolif- eration, differentiation and migration^5;6. These processes regulate normal development, maintenance and recovery from injury of renal tissue. Therefore over the last years at- tention has been directed to the role of cell-ECM adhesion-mediated signaling during I/R. Cell-ECM adhesions are mediated mainly via engagement of transmembrane integ- rin molecules by their extracellular ligands. Subsequent clustering of the integrins leads to their activation and results in recruitment and tyrosine phosphorylation of an array of protein s^5;6. Multiple protein-protein interactions have been defined at the focal adhesions (FAs) and most proteins have several interacting partners allowing generation of divergent signaling complexes. Proximal tubule cell detachment from the ECM during I/R preceded alterations in the F-actin cytoskeleton and redistribution of cell adhesion molecules and their signaling partners located at the cell-ECM adhesion complexes. This includes redis- tribution of α3 and ß1 integrins from the basolateral membrane to the apical membrane in in vitro and in vivo I/R models^7;8. This is associated with the distruption of FAs in close relationship with the dephosphorylation of FA proteins during hypoxia in isolated tubules and during I/R in vivo^9;10.

Focal adhesion kinase (FAK) is a non-receptor protein tyrosine kinase that localizes at the FAs. Integrin clustering results in autophosphorylation of FAK on tyrosine residue 397 (Y397)¹¹. Upon autophosphorylation FAK can be phosphorylated on other tyrosine residues by its binding partner Src. FAK interacts with many other kinases like Src as well as adaptor proteins such as paxillin and cytoskeletal structural proteins and is involved in signaling processes during survival, FA dynamics, cell adhesion and F-actin cytoskeletal organization as well as stress signaling by activation of ERK¹¹. FAK is also implicated in the inflammatory response to I/R injury by regulating the activation of JNK, ERK and NFκB pathways, which can affect tissue damage by attracting immune response factors^12;12-15. Recently we have reported that in a renal I/R rat model FAs undergo drastic restructur- ing. FA proteins including FAK are dephosphorylated directly after ischemia, followed by rephosphorylation during reperfusion. FAK was phosphorylated on different tyrosine resi- dues during reperfusion. We also reported that inhibition of ERK by using MEK1/2 inhibi- tor U0126, reduced FAK restructuring and phosphorylation of different tyrosine residues

during I/R. In addition inhibition of ERK attenuated renal injury induced by I/R⁹. The role of FAK and its dynamic phosphorylation during I/R remains unclear. Here we investigated the role of FAK in I/R induced ARF in vivo. Since FAK knockout mice are embryonically lethal^16;17 we set up a conditional proximal tubule specific FAK knockout mouse model using the Cre/LoxP system by crossing floxed

fak

fak

Renal proximal tubular cell specific FAK knockout mice are significantly less susceptible to I/R injury than the FAK^loxP/loxP littermates. This was associated with reduced immune cell infiltration and endothelium activation of the injured renal tissue. Our data suggest a model whereby FAK in the reperfusion phase after an ischemic period drives a cellular stress response that culminates in immune cell infiltration and renal tissue injury.

Material and methods

Mice and genotyping

Mice were maintained and bred at the animal facility of the Gorlaeus Laboratories of Le- iden University in accordance with institutional guidance and national health standards.

The mice were regularly monitored and had free access to water and standard mice chow.

All experiments using mice were approved by the local animal experimental committee of the Leiden University. Mice were genotyped using PCR. Genomic DNA was isolated from earcuts and FAK^loxP and FAK^wt bands (400bp and 290bp respectively) are detected us- ing 5’-GAGAATCCAGCTTTGGCTGTTG-3’ and 5’-GAATGCTACAGGAACCAAATAAC-3’

primers and the following conditions: 5 min at 95˚C (one cycle), 30s at 94˚C, 30s at 62˚C and 1 min at 72˚C (35 cycles), and 5 min at 72˚C. Cre-ER^T2 bands are detected using 5’- GTT CAG GGA TCG CCA GGC G-3’ and 5’-GCT GGC TGG TGG CAG ATG G -3’ primers and the following conditions: 5 min at 95˚C (one cycle), 1 min at 94˚C, 1 min at 65˚C and 2 min at 72˚C (30 cycles), and 10 min at 72˚C. PCR products were seperated at 2% agarose gels.

Isolation and culture of primary mouse renal cells

Mouse renal cells were isolated from the kidneys of male mice (25gram) with Rosa- Cre-ER^T2//FAK^LoxP/LoxP genotype. The kidneys were minced and digested in a collage- nase solution in Hank’s Hepes balanced salt solution (137mM NaCl, 5mM KCl, 0.8 mM MgSO4.7H2O, 0.4mM Na2HPO4.2H2O, 0.4mM KH2PO4, 1.3mM CaCl2, 4mM NaHCO3, 25mM HEPES, 5mM D-glucose, pH 7.4) for half hour at 37^oC. The cell suspension was washed three times in Hank’s Hepes. After the washing steps cells were resuspended in Dulbecco’s Eagles medium-Ham F12 1:1, 1% FBS, 0.5 mg/ml bovine serum albumin, 10 ng/ml EGF, 10 ng/ml cholera toxin, 50nM hydrocortisone, 15mM HEPES, 2mM glutamine, 1x Insulin-transferrin sodium selenite supplement and 1% (v/v) penicillin, streptomycin and amphotericin B (PSA) and plated on collagen coated dishes. The cells were maintained at 37^oC in a humidified atmosphere of 95% air/5% CO2 and fed every other day. Cells were exposed to 4-hydroxy-tamoxifen (4-OHT) to induce FAK recombination every other day of culturing together with refreshing the medium. Cells were used for experiments when they reached a confluent monolayer after 9 days of culturing.

Tamoxifen treatment of mice

300 mg tamoxifen powder (Sigma cat. No T5648) was dissolved in 900μl EtOH and sus- pended in 5.1 ml sunflower oil. This suspension (50mg/ml) was sonicated 2 times 30 seconds for a clear solution and stored at -20˚C in aliquots. The solution was thawed at

37˚C and administered orally to mice (100μl) by a feeding needle for 4 consecutive days followed by a period of 4 days without treatment, before analysis and the onset of the I/R injury.

Recombination PCR

For each mice used in the experiment 3 cryosections (10 μm) of the right, unaffected kid- ney were lysed and DNA was extracted. The presence of FAK^ΔloxP (recombination PCR) was detected using 5’-GACCTTCCAACTTCTCATTTCTCC-3’ and 5’-GAATGCTACAG- GAACCAAATAAC-3’ primers and the following conditions: 5 min at 95˚C (one cycle), 30s at 94˚C, 30s at 62˚C and 1 min at 72˚C (35 cycles), and 5 min at 72˚C. PCR products were separated on a 2% agarose gel.

Renal ischemia/reperfusion injury

A unilateral mouse model of renal I/R was used. Male mice (12 to 14 weeks of age) were anesthetized with isoflurane in a mix of N₂O and O ₂. A small incision was made over the left flank; the left renal pedicle was prepared and clamped with a B1 hemostatic clamp (Fine Science Tools, Heidelberg, Germany) for 35 minutes, while right kidneys were unaffected and served as internal controls. After removal of the clamp, the kidney was reperfused for 24 hours. After reperfusion, both left and right kidneys were harvested and prepared for further analysis as described under ‘Tissue preparation’. Additional control kidneys were obtained from animals that underwent sham surgery without I/R. Each group consisted of a minimum of 8 animals.

Tissue preperation

After harvesting, both left and right kidney was briefly washed in ice cold PBS to remove excess blood, and sliced into two equal halves. One half of each kidney was frozen in liq- uid nitrogen and stored at -80ºC for Western blot analysis or immunohistochemistry. The other half was fixated in ice-cold Carnoy’s solution (60 % (v/v) absolute ethanol, 30 % (v/v) chloroform and 10 % (v/v) glacial acetic acid) for 3 hours and thereafter transferred to 70%

(v/v) ethanol and stored at 4 oC and used for histopathological evaluation.

Histopathologic evaluation

Kidneys were embedded in paraffin and sectioned (3μm) onto APES coated slides. Paraf- fin sections were deparaffinized and rehydrated before staining for hematoxylin and eosin (H&E) and examined for tubular injury resulting from I/R injury using light microscopy (Leica DM6000B, 400x magnification). To assess tubulointerstitial injuries, kidney sec- tions were divided into three regions, i.e., cortex, outer medulla and inner medulla. Using

semiquantitative indices sections were analyzed for the evaluation of acute tubulointer- stitial damage. In each region, extents of tubular cast formation, tubular dilatation and tubular degeneration (vacuolar change, loss of brush border, detachment of tubular epi- thelial cells and condensation of tubular nuclei) were scored according to following criteria by two independent observers: 0, normal; 1<30%; 2, 30%-70%; 3, >70% of the pertinent area. After scoring all individual scores were summed to define the overall tubular damage in the kidney.

Immunohistochemistry of Kidney Injury Molecule 1 (KIM-1)

Paraffin sections of the injured kidney were stained for KIM-1 as described previously (Ichimura etal. 1998). Briefly, after antigen retrieval the sections were blocked with goat serum followed by incubation with affinity-purified rabbit polyclonal anti-peptide R9 anti- body at a concentration of 5 μg/ml. After 1 h, the sections were washed in PBS and incu- bated with biotinylated goat anti-mouse IgG. After further washes with PBS, the sections were incubated with an avidin-biotinylated horseradish peroxidase complex and counter- stained with 0.01% toluidine blue. Sections were examined for KIM-1 staining resulting from I/R injury using light microscopy (Leica DM6000B, 400x magnification). The sections were scored according to following criteria by two blinded observers: 0, no staining; 0.5, one tubule; 1, very few tubules; 2, several tubules but not wide spread; 3, wide spread staining or entire OSOM; 4, many tubules stained beyond OSOM.

Immunohistochemistry of focal adhesion proteins

Frozen sections were stained for focal adhesion proteins as described previously (Alderli- esten etal. 2007) Briefly, sections were blocked in 5% (v/v) normal goat serum (NGS, Vec- tor Laboratories) for 1 hour and incubated overnight at 4ºC in a humidified chamber with primary antibody; FAK (Upstate), P_Y397-FAK (BioSource), and P_Y118-paxillin (BioSource).

Thereafter, slides were washed and incubated for 1 h with secondary antibody; Alexa488- labeled goat anti-rabbit (Invitrogen). Rhodamin/Phalloidin (Molecular Probes) was used for F-actin staining. Images were made using a Bio-Rad Radiance 2100 confocal system with a 60x Plan Apo (NA 1.4:Nikon) objective lens. All images were processed with Image- Pro® Plus (Version 5.1 Media Cybernetics).

Immunohistochemistry of immune response factors

Frozen sections (10μm) were air dried and fixated in acetone for 10 min. After washing with PBS, the sections were incubated with the primary antibody, F4/80 (Rat MoAb, 1:8), CD45 (Ly2 Rat MoAB 1:1000), CD8 (Ly2 Rat MoAB 1:500), VCAM (CD106, Rat MoAB 1:400) and ICAM (Rat MoAb, 1:200), for 1 h except for F4/80 which was incubated overnight.

Thereafter slides were washed and incubated with secondary antibody Rabbit anti-rat Ig/

HRP for 30 min. After removing the secondary antibody; slides were washed, developed for 10 min and counterstained with haematoxylin. Sections were examined for immune response factor staining resulting from I/R injury using light microscopy (Leica DM6000B, 400x magnification).

Western blot analysis

Western blot samples were processed as described previously (Alderliesten etal. 2007).

Briefly, frozen sections were lysed in Triton lysisbuffer (20 mM Tris pH 7.4,137 mM NaCl, 2 mM EDTA, 1% Triton X-100, 25 mM ß-glycerophosphate, 10% glycerol) with inhibitors.

Lysates were syringed four times though a 26 G needle, centrifuged and immediately boiled in sample preparation buffer (125 mM Tris-HCl, pH 6.8, 20% glycerol, 4% SDS, bromophenol blue). Protein concentrations were determined using a Bradford assay with IgG as a standard. Equal amounts of protein were separated by SDS-PAGE and trans- ferred to PVDF membrane (immobilon-P, Millipore). Blots were blocked in 5% (w/v) BSA in TBS-T (0.5 M NaCl, 20 mM Tris-HCl and 0.05 % (v/v) Tween-20 pH 7.4) for 1h. Primary antibody incubation was performed overnight at 4 ºC in anti-P_Y397-FAK (1 μg/ml, polyclonal, Biosource), anti-FAK (monoclonal, 1 μg/ml, Transduction Laboratories), anti-P_Y118-paxillin (0.75 μg/ml, polyclonal, Biosource), anti-paxillin (monoclonal, 0.5 μg/ml, Transduction Laboratories). Thereafter blots were incubated with horseradish peroxidase conjugated secondary antibody (GE Healthcare) or fluorescent CY5 conjugated secondary antibody (Jackson laboratories) in TBS-T for 1 h at room temperature. Protein signals were detected with ECL plus method (GE Healthcare) or using fluorescence of the secondary antibody followed by scanning of the blots with the Typhoon 9400 (GE Healthcare).

Statistics

All values are provided as mean ±SEM. Statistical analysis was performed by student T- Test. Significant difference was set when P-value was <0.05.

Figure 1. Tamoxifen successfully recombines loxed fak alleles in the kidney cells. Mice were geno- typed for FAK^loxP/loxP, γGT-Cre-ERT2 and FAK by PCR (A). To show recombination renal cells of FAK^loxP/

loxP//γGT-Cre-ERT2, FAK^loxP/loxP//Rosa-Cre-ERT2 mice were isolated and treated with 4-OHT, genomic DNA was extracted and checked for recombination by PCR (FAKΔloxP/ΔloxP). Recombination band appears at 327 bp(B-C). Protein levels of FAK were detected by Western blot (B-C). FAK^loxP/loxP//γGT-Cre-ERT2, FAK^loxP/loxP//Rosa-Cre-ERT2 mice were treated with 5 mg tamoxifen for 4 days, 4days after the last treat- ment kidney, liver and spleen were harvested and checked for recombination. (D). To determine the protein levels of FAK frozen sections were prepared for Western blot analysis and stained for P_Y397-FAK and FAK (E). To determine the location of the FAKΔloxP/ΔloxP renal cells, frozen sections FAKΔloxP/ΔloxP

and FAK^loxP/loxP kidneys were stained for FAK and F-actin (F). * shows the proximal tubule. All sections were imaged using confocal laser scanning microscope. Sections are representative of proximal tubules in 3 different mice.

Results

Generating conditional proximal tubule specific FAK knockout mice

To study the role of FAK during I/R induced ARF we used a conditional proximal tubule specific FAK knockout mouse model making use of cre-LoxP technology. Mice with LoxP sites flanking the second exon of the fak gene (FAK^loxP/loxP) (Beggs etal. 2003) were crossed with transgenic mice that expressed Cre recombinase under control of the γGT promoter (γGT-Cre-ER^T2 ) that is only active in the proximal tubular cells (Dworniczak etal. 2007) to generate FAK^loxP/loxP//γGT-Cre-ER^T2 mice. Genomic DNA was prepared from earcuts and mice were genotyped for FAK, FAK^loxP/loxP and Cre-ER^T2 using PCR (Fig 1A). Genotype analysis of the newborn pups showed a mendalian distribution (data not shown), suggest- ing that both constructs have no effect on embryogenesis and/or viability of the mice.

First we checked recombination of the fak in vitro in isolated primary renal cells from both FAK^loxP/loxP//γGT-Cre-ER^T2 and FAK^loxP/loxP//Rosa-Cre-ER^T2 mice, in which every cell ex- presses Cre recombinase. Activation of Cre recombinase by addition of 4-hydroxy-tamox- ifen (4-OHT) showed clear recombination of the fak allele coinciding with a decrease in FAK protein level. FAK^loxP/loxP//γGT-Cre-ER^T2 renal cells showed only a partly recombined

fak allele which is according to expectations since only the proximal tubules express Cre.

On protein level we do not see a decrease in FAK levels after 4-OHT treatment, possibly because the reduction s too small to visualize on Western blot (Fig. 1B-C).

Male FAK^loxP/loxP//γGT-Cre-ER^T2 mice and their FAK^loxP/loxP littermates were treated with 5 mg tamoxifen for 4 consecutive days to induce Cre recombinase-mediated recombination of the fak alleles. Four days after the last treatment, kidneys, liver and spleen were harvested and checked for recombination using a non quantitative PCR. The tissue of FAK^loxP/loxP//

Rosa-Cre-ER^T2 mice shows recombination of fak alleles (FAKΔloxP/ΔloxP) by the appearance of a recombination band at 327 bp in all isolated tissues only after tamoxifen treatment (Fig. 1D). The tamoxifen treated FAK^loxP/loxP//γGT-Cre-ER^T2 mice show only recombination of the fak alleles in kidney tissue and not in spleen or liver and only a small portion of the total floxed FAK (Fig. 1D). The PCR results show no complete fak recombination; this can be explained by the fact that the γGT promoter is only active in the proximal tubule cells and not in the other renal cells. Control FAK^loxP/loxP mice showed no recombination in the kidney and in any other tissue (data not shown). Therefore we can conclude that the FAK knockout mouse model FAK^loxP/loxP//γGT-Cre-ER^T2 is both inducible and kidney specific. On protein level we do see a decrease in FAK in the kidney after tamoxifen treatment in the FAK^loxP/loxP//Rosa-Cre-ER^T2 mice but not in the FAK^loxP/loxP//γGT-Cre-ER^T2 mice similar to the in vitro results. This is not unexpected since proximal tubule cells make up only a small portion of the cells in kidney tissue; in addition we also do not observe a decrease

FAK protein almost disappeared from the basolateral side of the proximal tubule 4 days after the last tamoxifen treatment in FAK^loxP/loxP//γGT-Cre-ER^T2 mice compared to their FAK^loxP/loxP littermates (Fig. 1F). Using P_Y397-FAK staining on frozen sections of the kidney we show that a proximal tubule has an extremely light FAK staining at the basolateral side compared to the distal tubule in FAK^loxP/loxP //γGT-Cre-ER^T2 mice whereas in FAK^loxP/loxP mice normal P_Y397-FAK staining was observed.

These data show that the FAK^loxP/loxP//γGT-Cre-ER^T2 mouse is a successful inducible proxi- mal tubule specific FAK knockout mouse model.

FAK

mice are less susceptible for I/R induced renal injury

Male FAK^loxP/loxP //γGT-Cre-ER^T2 mice and their FAK^loxP/loxP littermates were treated with tamoxifen resulting in FAKΔloxP/ΔloxP and FAK^loxP/loxP mice. Both groups were subjected to left renal pedicle clamping for 35 min followed by 24 hours of reperfusion to induce I/R- induced acute renal failure or sham surgery.

Figure 2. I/R-induced injury in FAKΔloxP/ΔloxP and FAK^loxP/loxPkidneys. FAK^loxP/loxP//γGT-Cre-ERT2 mice and their FAK^loxP/loxP littermates were treated with 5mg tamoxifen for 4 days. 4 days after the last treatment mice were subjected to 35 minutes of ischemia and 24 hours of reperfusion or sham surgery.

Paraffin sections were stained for hematoxylin and eosin to determine tubular damage (A-B). Pictures were made using light microscopy (Leica DM6000B, 25x and 50x magnification, zoom 400x magnifi- cation) and are representative for the average injury of the experimental group. * shows casts in the tubules,  indicates dilatation and w degeneration of the tubules.

The indicated time of I/R resulted in clear but not severe injury in the left kidney. The control, sham operated, as well as the contralateral kidneys did not show any hallmarks of ischemic injury. They also did not show any differences between the FAKΔloxP/ΔloxP and FAK^loxP/loxP group suggesting that the loss of FAK in the proximal tubule cells on the short term has no impact on renal tissue histology (Fig. 2A). The kidneys exposed to ischemia followed by reperfusion in both experimental groups show clear histopathological markers of I/R injury. In the cortex of the kidney mainly dilatation and to some extent cast formation caused by loss of brush borders and detached cells was visible (Fig. 2B). Only few cells of the tubules also showed degeneration marked by fragmented nuclei and vacuole forma- tion. The medulla, both the inner and outer stripe, showed the highest degree of injury in both experimental groups. In this region dilatation was severe, cells were detached and situated in the lumen and the tubules were partly denuded. Cells showed clear degenera- tion markers as seen in the cortex but to a greater extent (Fig. 2B). Using semi-quantitative analysis of all the aforementioned histopathological markers of I/R-induced renal injury showed a significant difference in total renal injury between the FAKΔloxP/ΔloxP and

FAK^loxP/loxP group. The FAKΔloxP/ΔloxP mice showed about half the total renal injury score compared to their FAK^loxP/loxP littermates (Fig. 3A-B). Scoring of histopathological markers per region further shows that primarily dilatation in all regions is different between the two experimental groups. The FAK^loxP/loxP group shows significantly more dilatation than the FAKΔloxP/ΔloxP group. Cast formation in the inner stripe of the medulla is also scored signifi- cantly lower in the FAKΔloxP/ΔloxP kidneys (Fig. 3C).

FAK

mice show less kidney injury molecule 1 staining after I/R-induced

renal injury

Kidney injury molecule 1 (KIM-1) is a known biomarker of ischemic injury which is only found in kidney tissue after ischemia in the renal tubules. Staining kidney sections for KIM-1 showed that in the FAK^loxP/loxP kidneys was generally found in the proximal tubules of the OSOM and cortex whereas in the FAKΔloxP/ΔloxP kidneys KIM-1 was found in mainly in the proximal tubules in the OSOM (Fig. 4A). KIM-1staining was scored from 0 to 4 and showed similar scores to the histopathological renal injury. FAKΔloxP/ΔloxP kidneys have sig- nificantly less KIM-1 staining after I/R than the FAK^loxP/loxP kidneys (Fig. 4B-C) and thus less renal injury. Pictures shown in figure 4 are representative for the amount of injury scored in the particular group and are derived from the same kidney as the H&E stainings shown in figure 3. These data show that FAK is important in the development of I/R-induced renal injury.

Kidneys of FAK

mice are less infiltrated with immune response factors after I/R-induced renal injury

I/R-induced renal injury coincides with an immune response and infiltration of immune cells. We stained for immune response markers F4/80, a marker for murine macrophages, ICAM, an intracellular adhesion molecule involved in leukocyte binding, VCAM, a vascular adhesion molecule often induced by cytokines, CD8, a marker for cytotoxic T lympho- cytes and CD45, a marker for T and B lymphocytes.I/R does not cause infiltration of Figure 3. Scored I/R-induced injury of FAKΔloxP/ΔloxP and FAK^loxP/loxP kidneys. FAK^loxP/loxP//γGT-Cre- ERT2 mice and their FAK^loxP/loxP littermates were treated with 5mg tamoxifen for 4 days. 4 days after the last treatment mice were subjected to 35 min of ischemia and 24 h of reperfusion or sham sur- gery. Paraffin sections were stained for hematoxylin and eosin to determine tubular damage. Sections were scored double-blind and semi-quantitatively for casts, dilatation and degeneration per region (i.e.

cortex, OSOM, ISOM) to assess tubule-interstitial injury. Data are shown as total tubular injury score (A), as box-whiskers plot to show spreading of the total injury score(B) and as average score per histo- pathological marker per region (C). All data are presented as mean ±SEM (n = 8-10 mice per group), * p < 0.05 compared to FAK^loxP/loxP littermates.

CD8 positive cells however the other markers showed increased staining after I/R in both experimental groups. Macrophages were found mainly in the pericapillary areas or at the severely damaged parts of the tissue. CD45 staining, showing T and B lymphocytes, was localized mainly in the OSOM and some in the cortex, the same parts of the tissue as the F4/80 staining, specific for macrophages. ICAM staining was seen throughout the kidney (data not shown) whereas VCAM was mainly organized in array like structures beginning in the papilla moving outward following the endothelium of the blood vessels. VCAM Stain- ing was seen more in the OSOM but scarcely in the cortex (Fig.: 4). The overall trend seen was that the amount and intensity of the staining correlated with the overall injury score.

Meaning that in the FAKΔloxP/ΔloxP kidneys the infiltration of immune cells and the activation of endothelium, shown by VCAM, is less. Pictures shown in figure 5 are representative for the amount of injury scored in the particular group and are derived from the same kidney as the H&E stainings shown in figure 2.

Figure 4. FAKΔloxP/ΔloxP mice show less KIM-1 after I/R compared to their FAK^loxP/loxP littermates.

Paraffin sections of the kidneys of FAKΔloxP/ΔloxP mice and their FAK^loxP/loxP littermates were stained for KIM-1. Pictures were made using light microscopy (Leica DM6000B, 25x and 100x magnification) and are representative for the average injury of the experimental group (A). B shows the amount of KIM-1 staining that corresponds with the score between 0 and 4. Sections were scored blind for KIM-1 stain- ing. The distribution of the scores of the two groups of mice is shown in percentages of total mice per group (D). The correlations between total renal injury and KIM-1 score is shown in E. Data are repre- sented as mean +SEM (n = 8-10 mice per group), * p < 0.05 compared to FAK^loxP/loxP.

Figure 5. FAKΔloxP/ΔloxP mice show less I/R-induced immune response compared to their FAK^loxP/loxP littermates. Frozen sections of the kidneys of FAKΔloxP/ΔloxP mice and their FAK^loxP/loxP littermates subjected to 35 minutes of ischemia and 24 hours of reperfusion or sham surgery were stained for F4/80, CD45 and VCAM. Pictures were made using light microscopy (Leica DM6000B, 25x and 100x magnification) and are representative for the average injury of the experimental group.

Discussion

In this study we deleted fak specifically in the renal proximal tubule cells in vivo to study the role of FAK during I/R induced ARF. Our main conclusion of this study is that condi- tional and proximal tubule specific deletion of FAK prior to renal I/R leads to a decrease in renal injury. In addition, this coincides with less infiltration of immune cells in response to I/R

These results suggest that FAK is involved in the development of I/R induced renal injury possibly by activation of signaling pathways that cause the infiltration of immune cells to the damaged areas in the kidney, resulting in aggrevation of renal tissue injury.

FAK knockout mice are embryonically lethal (E8.5)^16;17. In addition, knocking out FAK in adulthood in the whole organism can result in a complex knockout phenotype because FAK is an ubiquitously expressed protein involved in many cellular processes. To study the role of FAK during renal I/R an inducible tissue specific FAK knockout mouse model is an ideal tool to bypass these drawbacks. The proximal tubule cells are the primary target of I/R injury, therefore we chose a proximal tubule specific FAK knockout mouse model.

The proximal tubule cells make up only a small portion of the total renal cells. In this study we show that administration of tamoxifen to the FAK^loxP/loxP//γGT-Cre-ER^T2 mice resulted in recombination of the fak alleles and deletion of FAK protein selectively in the proximal tubule cells. This is conform the results published by Dworniczak etal. 2007 where they show that Cre-ER^T2 can only be activated after tamoxifen treatment and that Cre activity is restricted to the proximal tubule¹⁹. From these data we can conclude that the inducible proximal tubule specific FAK knockout mouse model is fully functional and useful to study the role of FAK during I/R.

No differences were found in the histology of the kidney between the FAKΔloxP/ΔloxP and FAK^loxP/loxP group. Taking into account that only a small portion of the nephron is affected by FAK deletion and the kidney has not been exposed to injury, these data suggest that the structural organization of and/or molecular environment in the native tissue may over- come the lack of signaling through FAK in FAKΔloxP/ΔloxP cells. Histopathological markers of I/R injury show that the kidneys are mainly affected in the OSOM region where the proxi- mal tubules are located. The FAKΔloxP/ΔloxP mice show significantly less renal injury than their FAK^loxP/loxP littermates.

Several in vitro studies showed that deletion of FAK induced apoptosis because of loss of adhesion to the ECM. In addition our previous studies showed that dominant negatives of FAK potentiated nephrotoxicant-induced apoptosis in renal cells^20;21. However other studies report that FAK deletion does not induce apoptosis. It is possible that the cellular context accounts for these opposing effects. Furthermore it has been well established that

in vitro and in vivo studies can give opposite results. In vivo study on cell-ECM adhesions during renal I/R by Zuk etal. showed that ß1-integrins translocated from the apical to the lateral membrane perturbing cell-ECM adhesion⁸. Maintaining active ß1-integrins during I/R by using HUTS-21, an antibody that recognizes only the active epitope of ß1-integrins, preserves the function of the tissue after renal I/R and stabilized FAK phosphorylation during ischemia maintaining FA functionality⁷. Since FAK binds to integrins and is im- portant for the downstream signaling we expected FAK deletion to worsen I/R induced renal injury. These data seem in contradiction to our results. However, although FAK and ß1-integrin are often part of the same pathway they also have distinct functions. FAK is a tyrosine kinase with many binding partners and involved in many cellular processes in addition adhesion mediated survival signaling.

FAK is implicated in the immune and inflammatory response signaling by regulating cy- tokine production¹². We show that the exposed kidneys of the FAKΔloxP/ΔloxP group have less proinflammatory mediators and infiltrated immune cells than the kidneys of the

FAK^loxP/loxP mice. Many studies have shown that during I/R the inflammatory response is a prominent contributor to renal injury. Besides the circulating leukocytes, tubular and glom- erular mesengial cells are a major source of proinflammatory products like TNFα, TGFß and interleukins 1, 2 and 6. FAK is important for interleukin-6 production after TNF-α stimulation¹². Furthermore FAK is necessary in the activation of NFκB an important media- tor of TNF-α signaling and quickly activated after ischemic insults. FAK deletion prevents NFκB activation and subsequent signaling to induce proinflammatory mediators and im- mune cell infiltration^12;22;23. This suggest that a reduced immune response because of a lack of FAK in the proximal tubule cells can lead to reduction in renal cell injury after I/R in FAKΔloxP/ΔloxP mice compared to their FAK^loxP/loxP littermates.

Another pathway implicating FAK in the immune response after I/R is the MAPK pathway.

FAK deletion has reported to inhibit MAPK/ERK pathway activation and subsequent pro- duction of IL-6 and other cytokines, thereby inhibiting the immune cell infiltration^22;23. We have previously shown that phosphorylation of ERK1/2 is upregulated in rat kidneys very quickly after ischemia. We see the same upregulation in injured mouse kidneys (data not shown). The ERK1/2 pathway is placed upstream of TNF-α mediated inflammation and im- plicated in the expression of death ligands and proinflammatory cytokines after ischemia.

Inhibition of ERK1/2 with the MEK inhibitor before cisplatin induced renal injury decreased TNF-α gene expression and inflammation inhibition of ERK protects the kidney to I/R injury²⁴. Furthermore in a previous study we show that ERK1/2 inhibition in rats before I/R attenuates I/R induced injury. However the interaction between FAK and ERK1/2 activa- tion needs further investigations especially in vivo.

As we showed previously (Alderliesten etal.) FAK is differentially phosphorylated on pTyr

residues during reperfusion⁹. The different pTyr residues bind different sets of proteins and as such are involved in different signaling pathways. Therefore it can be suggested that FAK has different roles during the several stages of reperfusion after renal ischemia.

The first 24 h might involve signaling to stress pathways involved in the immune response whereas during the later stages of reperfusion FAK is involved in signaling that leads to regeneration of the injured tissue. However this should be further investigated.

In conclusion, we propose a model where FAK contributes to cytokine production by acti- vating one or more cellular stress pathways like the NFκB and/or the MAPK/ERK pathway.

This results in an immune response that leads to the infiltration of immune cells to the damaged area and activation of the endothelium potentiating renal injury.

KIM-1 is a type 1 transmembrane protein whose expression is undetectable in normal renal tissue and is markedly upregulated with injury of proximal tubule epithelial cells in rats. In addition it has been reported to be upregulated in a limited number of native human biopsies where acute tubular injury was diagnosed histologically. Kim-1 staining has proved to be a more sensitive marker for ARF than creatinine levels in serum or BUN measurements as well as histological assessment with H&E or PAS staining. Furthermore it is a very sensitive marker for early kidney tubular injury. KIM-1 expression has been as- sociated with the potential for regeneration of the proximal tubules and recovery of kidney function. Recently, Ichimura etal. have proposed a model where KIM-1 mediated phago- cytosis of apoptotic and necrotic debris by epithelial cells via binding to the cell surface and triggering internalization. By doing so, KIM-1 may play an important role in limiting the autoimmune response to injury. Phagocytosis of apoptotic cells mediated by KIM-1 may result in the generation of anti-inflammatory cytokines as occurs with phagocytosis by macropahges. Another process KIM-1 might be contributing besides clearance of the lumen is regeneration of the proximal tubules.

Acknowledgements

We thank Emile de Heer for his suggestions, Annemieke van der Wal and Reshma Lalai for staining of the immune response factors and the members of the Division of Toxicology of the Leiden/Amsterdam Center for Drug Research for valuable discussion and support.

This work was supported by grants from the Netherlands Organization for Scientific Re- search (grant 908-02-107).

References

Hoste EA, Schurgers M: Epidemiology of acute kidney injury: how big is the problem?

1.

Crit Care Med 2008, 36:S146-S151.

Racusen LC, Fivush BA, Li YL, Slatnik I, Solez K: Dissociation of tubular cell detach- 2.

ment and tubular cell death in clinical and experimental “acute tubular necrosis”. Lab Invest 1991, 64:546-556.

Racusen LC, Solez K: Ideas in pathology. Exfoliation of renal tubular cells. Mod Pathol 3.

1991, 4:368-370.

Racusen LC: Tubular injury in human kidneys: pathologic findings and pathogenic 4.

mechanisms. Clin Investig 1993, 71:858-860.

Reddig PJ, Juliano RL: Clinging to life: cell to matrix adhesion and cell survival. Can- 5.

cer Metastasis Rev 2005, 24:425-439.

Horwitz AR, Parsons JT: Cell migration--movin’ on. Science 1999, 286:1102-1103.

6.

Molina A, Ubeda M, Escribese MM, Garcia-Bermejo L, Sancho D, de Lema GP, Liano 7.

F, Cabanas C, Sanchez-Madrid F, Mampaso F: Renal ischemia/reperfusion injury:

functional tissue preservation by anti-activated {beta}1 integrin therapy. J Am Soc Nephrol 2005, 16:374-382.

Zuk A, Bonventre JV, Brown D, Matlin KS: Polarity, integrin, and extracellular matrix 8.

dynamics in the postischemic rat kidney. Am J Physiol 1998, 275:C711-C731.

Alderliesten M, de Graauw M, Oldenampsen J, Qin Y, Pont C, van Buren L, van de 9.

WB: Extracellular signal-regulated kinase activation during renal ischemia/reperfu- sion mediates focal adhesion dissolution and renal injury. Am J Pathol 2007, 171:452- 462.

Weinberg JM, Venkatachalam MA, Roeser NF, Senter RA, Nissim I: Energetic deter- 10.

minants of tyrosine phosphorylation of focal adhesion proteins during hypoxia/reoxy- genation of kidney proximal tubules. Am J Pathol 2001, 158:2153-2164.

Mitra SK, Hanson DA, Schlaepfer DD: Focal adhesion kinase: in command and con- 11.

trol of cell motility. Nat Rev Mol Cell Biol 2005, 6:56-68.

Huang D, Khoe M, Befekadu M, Chung S, Takata Y, Ilic D, Bryer-Ash M: Focal adhe- 12.

sion kinase mediates cell survival via NF-kappaB and ERK signaling pathways. Am J Physiol Cell Physiol 2007, 292:C1339-C1352.

Funakoshi-Tago M, Sonoda Y, Tanaka S, Hashimoto K, Tago K, Tominaga S, Kasa- 13.

hara T: Tumor necrosis factor-induced nuclear factor kappaB activation is impaired in focal adhesion kinase-deficient fibroblasts. J Biol Chem 2003, 278:29359-29365.

Ishibe S, Joly D, Liu ZX, Cantley LG: Paxillin serves as an ERK-regulated scaffold 14.

for coordinating FAK and Rac activation in epithelial morphogenesis. Mol Cell 2004, 16:257-267.

Webb DJ, Donais K, Whitmore LA, Thomas SM, Turner CE, Parsons JT, Horwitz AF:

15.

FAK-Src signalling through paxillin, ERK and MLCK regulates adhesion disassembly.

Nat Cell Biol 2004, 6:154-161.

Furuta Y, Ilic D, Kanazawa S, Takeda N, Yamamoto T, Aizawa S: Mesodermal defect 16.

in late phase of gastrulation by a targeted mutation of focal adhesion kinase, FAK.

Oncogene 1995, 11:1989-1995.

Ilic D, Furuta Y, Kanazawa S, Takeda N, Sobue K, Nakatsuji N, Nomura S, Fujimoto 17.

J, Okada M, Yamamoto T: Reduced cell motility and enhanced focal adhesion contact formation in cells from FAK-deficient mice. Nature 1995, 377:539-544.

Beggs HE, Schahin-Reed D, Zang K, Goebbels S, Nave KA, Gorski J, Jones KR, 18.

Sretavan D, Reichardt LF: FAK deficiency in cells contributing to the basal lamina results in cortical abnormalities resembling congenital muscular dystrophies. Neuron 2003, 40:501-514.

Dworniczak B, Skryabin B, Tchinda J, Heuck S, Seesing FJ, Metzger D, Chambon 19.

P, Horst J, Pennekamp P: Inducible Cre/loxP recombination in the mouse proximal tubule. Nephron Exp Nephrol 2007, 106:e11-e20.

van de WB, Nagelkerke JF, Stevens JL: Dephosphorylation of focal adhesion kinase 20.

(FAK) and loss of focal contacts precede caspase-mediated cleavage of FAK during apoptosis in renal epithelial cells. J Biol Chem 1999, 274:13328-13337.

van de WB, Houtepen F, Huigsloot M, Tijdens IB: Suppression of chemically induced 21.

apoptosis but not necrosis of renal proximal tubular epithelial (LLC-PK1) cells by focal adhesion kinase (FAK). Role of FAK in maintaining focal adhesion organization after acute renal cell injury. J Biol Chem 2001, 276:36183-36193.

Guijarro C, Egido J: Transcription factor-kappa B (NF-kappa B) and renal disease.

22.

Kidney Int 2001, 59:415-424.

Meldrum KK, Hile K, Meldrum DR, Crone JA, Gearhart JP, Burnett AL: Simulated 23.

ischemia induces renal tubular cell apoptosis through a nuclear factor-kappaB de- pendent mechanism. J Urol 2002, 168:248-252.

Jo SK, Cho WY, Sung SA, Kim HK, Won NH: MEK inhibitor, U0126, attenuates cispl- 24.

atin-induced renal injury by decreasing inflammation and apoptosis. Kidney Int 2005, 67:458-466.