Advanced Glycation End Products Inhibit Tubulogenesis and Migration of Kidney Epithelial Cells in an Ezrin-Dependent Manner
Marisa A Gallicchio*,
E. Anne McRobert*,
Anjali Tikoo*,
Mark E. Cooper and
Leon A. Bach*,
* Monash University, Department of Medicine, and Department of Endocrinology and Diabetes, Alfred Hospital, and Baker Medical Research Institute, Victoria, Australia
Address correspondence to: Dr. Leon A. Bach, Monash University, Department of Medicine, Alfred Hospital, Commercial Road, Prahran, Victoria 3004, Australia. Phone: +61-3-9276-2387; Fax: +61-3-9276-3782; E-mail: leon.bach{at}med.monash.edu.au
Received for publication January 13, 2005.
Accepted for publication November 4, 2005.
Nonenzymatic glycation of proteins to form advanced glycationend products (AGE) is implicated in diabetic complications,including nephropathy. It was shown recently that AGE bind tothe ERM (ezrin, radixin, and moesin) family of membrane-cytoskeletallinker proteins in renal homogenates. Herein is reported theeffects of AGE-BSA on ezrin-dependent LLC-PK1 kidney epithelialcellular functions: migration and hepatocyte growth factor (HGF)-inducedtubulogenesis. LLC-PK1 cells were stably transfected with cDNAfor ezrin sense, ezrin antisense, and N-ezrin. Transfectionof LLC-PK1 cells with ezrin antisense and dominant negativeN-ezrin decreased basal tubulogenesis and migration relativeto vector-only transfection, establishing the ezrin dependencyof these processes. AGE-BSA (20 or 40 µM) significantlydecreased HGF-induced tubulogenesis and basal migration in twovector control lines relative to BSA-treated cells. However,AGE-BSA inhibition of both HGF-induced tubulogenesis and migrationwas overcome by overexpressing ezrin. These results demonstratethat the AGE–ezrin interaction significantly alters cellularfunction. These changes may be relevant to detrimental renalconsequences as a result of diabetes.
Nonenzymatic glycation of proteins, lipids, and nucleic acidsresulting in the accumulation of advanced glycation end products(AGE) is a prominent feature of diabetes. AGE levels correlatewith the development of chronic diabetic complications suchas retinopathy, nephropathy, neuropathy, and vasculopathy (1–3).In the normal rat, injection of AGE leads to renal changes similarto those of diabetic nephropathy, whereas interference withAGE formation reduces diabetic complications in various animalmodels (4–6). The importance of AGE as a pathogenic mechanismin diabetic nephropathy is suggested further by the findingthat AGE formation/accumulation precedes diabetic renal disease(2). Together with a cellular receptor RAGE, AGE accumulatein diabetic renal tissue (2, 7, 8). However, the precise cellularmechanism(s) whereby AGE promote the development of diabeticcomplications remains unknown.
We recently described the novel binding of AGE to the ERM (ezrin,radixin, and moesin) family of proteins (9). ERM proteins arecritical regulators of interactions between the cell membraneand the cytoskeleton (10). As members of the erythrocyte protein4.1 superfamily, ERM proteins are characterized by a conservedN-domain (FERM domain), which associates with the cytoplasmicdomain of specific membrane-associated proteins such as thehyaluronan receptor CD44, CD43, and CD95 and the intercellularadhesion molecules-1, -2, and -3. The C-terminal domain of ERMbinds to F-actin in vitro and in vivo (11, 12). However, theF-actin binding site and binding sites for membrane proteinsare masked by intramolecular association between the N- andC-domains in the inactive cytoplasmic molecule (13). Tyrosinephosphorylation by receptors for EGF and hepatocyte growth factor(HGF) activate ezrin by unmasking the F-actin binding site (14,15). In LLC-PK1, a proximal tubule cell line, tyrosine phosphorylationof ezrin increases cell survival (16).
Tubular injury features prominently in the development of renaldysfunction in diabetes (17, 18). In vivo, epithelial tubulogenesisrequires the migration and repopulation of the tubular conduitwith proliferating cells of the renal proximal tubular epitheliumduring kidney development or after renal injury. In vitro, ezrinis a mediator of HGF-induced LLC-PK1 tubulogenesis because dominantnegative N-ezrin expression and ezrin overexpression inhibitand enhance this process, respectively (15). In this model,tyrosine phosphorylation of ezrin is essential, and this isalso the case with respect to HGF-induced migration of thesecells. As with tubulogenesis, ezrin overexpression in LLC-PK1cells enhances cell migration, whereas N-ezrin expression impairsHGF-induced migration (15). We recently showed that AGE-BSAinhibited LLC-PK1 tubulogenesis and ezrin tyrosine phosphorylation(9).
In this article, we describe the effect of AGE-BSA on LLC-PK1migration and tubulogenesis. We show that inhibition of ezrinactivity or expression decreased migration and tubulogenesis.We also show that ezrin overexpression overcame the inhibitoryeffect of AGE-BSA on migration and tubulogenesis, suggestingthat AGE-BSA inhibited these processes by suppressing ezrinfunction.
Materials
Tissue culture plastics and reagents were purchased from Nunc(Roskilde, Denmark) and Trace Biosciences (Melbourne, Australia),respectively. All other laboratory reagents were purchased fromSigma Corp. (St. Louis, MO) unless otherwise specified.
AGE Preparation
AGE-BSA and AGE-RNAse were prepared as described previouslywith modifications (19). Briefly, BSA or RNAse (10 mg/ml) wasincubated at 37°C for 12 or 9 wk, respectively, with d-glucose(90 g/L) in 0.4 M phosphate buffer. Control BSA or RNAse wasprepared by incubation without glucose. Preparations were lyophilized,resuspended in water, and dialyzed against PBS to remove freeglucose. The characteristic glycation fluorescence (excitation370 nm, emission 440 nm) of AGE-BSA was increased seven- to28-fold compared with BSA and for AGE-RNAse was increased four-foldcompared with RNAse.
Cell Culture
LLC-PK1, a polarized porcine kidney proximal tubule epithelialcell line, was obtained from the American Type Culture Collection(Rockville, MD). Cells were cultured in growth medium (GM) thatconsisted of DMEM that contained 4.5 g/L glucose, 2 mM l-glutamine,5000 IU/L penicillin, 5 mg/L streptomycin, 125 U/L Fungizone,and 2.2 g/L sodium bicarbonate and supplemented with 10% FCSin a 5% CO2 incubator. For experiments under serum-free conditions,cells were cultured in serum-free medium that consisted of DMEMwith antibiotics and glutamine as above supplemented with 0.5g/L BSA.
Ezrin Antisense, N-Ezrin, and Ezrin Transfection
The mammalian expression vector pCR3 (Invitrogen, Carlsbad,CA) with cloned inserts encoding human N-ezrin–green fluorescentprotein (GFP) or ezrin antisense were prepared (20). Full-lengthhuman ezrin was subcloned into mammalian expression vector pMV7(21). Subconfluent LLC-PK1 cells were incubated in serum-freemedium that contained DNA vector construct with insert (1 µg/ml)and lipofectamine (12 µg/ml; Life Technologies, Gaithersburg,MD) for 6 h. Control transfections were done using vector alone.Cells were incubated for an additional 72 h in GM before passagingand selection with Geneticin (900 µg/ml; Life Technologies)in GM. Surviving cell colonies were subcloned over a periodof 2 to 3 mo. Clones were assessed for ezrin under- and overexpressionby Western blotting using an anti-human N-ezrin antiserum andfor N-ezrin expression by N-ezrin ELISA and GFP immunofluorescenceand subcloned further. The presence of relevant DNA insertsin subclones was confirmed by real-time reverse transcription–PCR(RT-PCR) as described below.
Western Blotting
Cell lysates were prepared from parental and transfected cellsin 50 mM HEPES, 0.05 M NaCl, 0.05% Tween 20, and 1% Triton X-100and centrifuged at 12,500 rpm for 1 min, and supernatants werestored at –20°C until used. Proteins (100 µg/sample)were separated by nonreducing SDS-PAGE and transferred to nitrocellulosemembranes. All washes were performed in and antibodies werediluted in 5% skim milk/PBS. Membranes were blocked with 5%skim milk/PBS and probed with a rabbit polyclonal antiserumraised against human N-ezrin at a 1:5000 dilution and incubatedovernight at 4°C. The blot was washed and then incubatedwith biotinylated anti-rabbit antiserum diluted 1:5000 for 1h at room temperature, washed, and incubated with streptavidin-labeledperoxidase at a 1:5000 dilution for 1 h at room temperature.Signal was detected using enhanced chemiluminescence (SupersignalWest Pico Chemiluminescent Substrate; Pierce, Rockford, IL)and exposure to x-ray film for 5 min. Blots were quantifiedusing Image J (National Institutes of Health, Bethesda, MD).
RT-PCR
Total RNA (5 µg) that was extracted from transfected LLC-PK1cell lines was used to synthesize cDNA with the SuperscriptFirst Strand Synthesis system for reverse transcription (LifeTechnologies Invitrogen, Carlsbad, CA). Gene expression wasanalyzed by real-time quantitative PCR performed with the TaqMansystem. Fluorescence for each cycle was analyzed by an ABI Prism7700 Sequence Detection System (PE, Perkin-Elmer, Biosystems,Foster City, CA). For controlling for variation in the amountof DNA that was available for PCR in each reaction, gene expressionof the target sequence was normalized for the amount of an endogenouscontrol, 18S ribosomal RNA (18S rRNA reagent kit; PE Biosystems).Primers and TaqMan probe for human ezrin (forward primer CTGAGACTGCCGTGCTCTTG,reverse primer GTAAGTTTGTGCTGGTCCATCACT, probe 6FAM-CACAAGTCTGGGTACCTC-TAMRA)were constructed with the help of Primer Express (ABI Prism7700). Amplification was performed as follows: 50°C for2 min, 95°C for 10 min, then 40 cycles at 94°C for 20s and 60°C for 1 min. A standard curve was generated usinga human ezrin cDNA standard (1 ng to 0.01 fg), and levels ofezrin cDNA in samples were converted to ezrin copy number (x109/µgRNA). Samples were measured in duplicate.
Immunofluorescence
Cells that were transfected with N-ezrin–GFP were grownon glass coverslips. After fixation in 3.7% paraformaldehyde/PBSfor 10 min, cells were washed in PBS and then water before mountingin Permafluor (Beckman-Coulter, Marseille, France). Cells wereviewed by confocal microscopy using a Leica fluorescence microscopeat an excitation wavelength of 488 nm and emission wavelengthof 568 nm.
Tubulogenesis
Tubulogenesis experiments were performed as described previously(9). Briefly, LLC-PK1 cells were cultured in 24-well platesin 10% FCS/DMEM. Conditioned medium (CM) from 3T3 cells thatwere grown in 10% FCS/DMEM was used as a source of HGF (22).LLC-PK1 cells were washed once in 10% FCS/DMEM and incubatedwith 0.5 ml/well 3T3 CM diluted 1:2 in GM. Alternatively, 100ng/ml recombinant HGF (Sigma) was used in some experiments asindicated. BSA or AGE-BSA was added to a final concentrationof 20 and 40 µM with the equivalent volume of PBS addedto control wells. Collagen type I (6.5 µl/well of 3.66mg/ml stock; Beckman-Coulter) was added, and cells were incubatedfor 48 h, after which tubule numbers were counted in each well.
Migration
Parental or transfected LLC-PK1 cells were grown to confluencein six-well plates. Four wounds per well were made with a pipettetip, and cells were washed twice in 1% FCS/DMEM before adding1 ml/well 1% FCS/DMEM and 1 µg/ml Mitomycin C to inhibitcell proliferation. BSA or AGE-BSA was added to a final concentrationof 20 µM with the equivalent volume of PBS added to controlwells. Photographs of wounds were taken at the start of theexperiment and after 24 h of incubation. Wound areas were calculatedusing Image J (National Institutes of Health). The distanceof cell migration for each treatment was calculated by subtractingthe wound area at T = 24 h from that at T = 0 h and expressingthis as a percentage of PBS-treated control.
Statistical Analyses
Tubulogenesis and ezrin antisense data were analyzed by ANOVA.Migration data were analyzed by repeated measures ANOVA afterlogarithmic transformation to stabilize variance. Correctionfor multiple comparisons was performed using Fisher PLSD test.P < 0.05 was considered significant. Results are expressedas the mean ± SEM. Experiments were repeated betweentwo and nine times as indicated.
Characterization of N-Ezrin–GFP–, Ezrin Antisense–, and Ezrin-Overexpressing LLC-PK1 Cell Lines
Clones were initially screened for N-ezrin–GFP by N-ezrinELISA (results not shown). N-ezrin–GFP–transfectedLLC-PK1 cell clones then were subcloned and selected on thebasis of the presence of GFP as determined by confocal microscopy(Figure 1A). GFP–N-ezrin localized to the cell periphery,and punctate fluorescence staining of the cytoplasm was alsonoted.
Figure 1. Characterization of N-ezrin–, ezrin antisense–, and ezrin-overexpressing LLC-PK1 cells. (A) Confocal microscopic image of green fluorescent protein (GFP)–N-ezrin–transfected cells showing GFP–N-ezrin expression at the cell periphery. (B) Western blot of untransfected (parent) and ezrin-antisense transfected (AS1 and AS2) LLC-PK1 cell lines that were probed with a rabbit anti-human N-ezrin antibody. (C) Quantification of three Western blots of untransfected (parent) and ezrin antisense–transfected (AS1 and AS2) LLC-PK1 cell lines that were probed with a rabbit anti-human N-ezrin antibody (*P < 0.05, **P < 0.01 versus parent). (D) Western blot of untransfected (parent) and ezrin-overexpressing (Ezrin 1 and Ezrin 2) LLC-PK1 cell lines that were probed with a rabbit anti-human N-ezrin antibody. (E) Reverse transcription–PCR analysis of human ezrin gene expression in LLC-PK1 cell lines that were transfected with vector only (Vector 1) or vector-containing ezrin (Ezrin 2) or N-ezrin (N-ezrin 1 and N-ezrin 2). Duplicate samples were prepared and ezrin cDNA levels were quantified against an ezrin standard curve after standardization for 18S levels. Results are expressed as the average ezrin copy number per microgram of RNA (mean ± SEM, n = 2). Magnification, x400 in A.
Ezrin antisense–transfected LLC-PK1 cells were screenedby Western blotting using a rabbit anti-human N-ezrin antibody.Clones were chosen on the basis of reduced expression of ezrincompared with parental LLC-PK1 cells (Figure 1, B and C) Comparedwith the parent LLC-PK1 cell line, ezrin antisense lines AS1and AS2 showed 66% and 88% reductions, respectively, in ezrinexpression. Ezrin-overexpressing LLC-PK1 cells were also screenedby Western blot and shown to produce more ezrin than parentalLLC-PK1 cells (Figure 1D). Transfected cell lines were alsoscreened by real-time RT-PCR for human ezrin mRNA (Figure 1E).Minimal levels of human ezrin (0.096 x 109 copies/µg RNA)were detected in the vector control cell line as a result ofcross-reactivity with pig ezrin cDNA, whereas up to 1000-foldmore ezrin and N-ezrin mRNA (128 and 4.4 x 109 copies/µgRNA, respectively) was measured in the ezrin- and N-ezrin–transfectedcell lines, respectively.
LLC-PK1 Tubulogenesis Is Ezrin-Dependent
We showed previously that LLC-PK1 cells, when treated with HGFin the presence of collagen I, form tubules between domes andthat treatment with AGE-BSA inhibited tubulogenesis (8, 9).This model has been reported previously to be ezrin- dependent(15). This was confirmed in our study, in which tubule formationwas significantly lower in ezrin antisense–and N-ezrin–transfectedcell lines compared with vector control cells (P < 0.05 orP < 0.01; Figure 2A). Inhibition of tubulogenesis by N-ezrinconfirms a dominant negative effect of N-ezrin on endogenousezrin function (15). The cell viability of transfected celllines during the tubulogenesis assay was unaffected as shownin Figure 2B. HGF signaling in transfected ezrin antisense–,N-ezrin–, and ezrin-overexpressing cell lines was shownto be intact as shown by phosphorylation of Akt in responseto HGF (results not shown).
Figure 2. Ezrin dependence of LLC-PK1 tubulogenesis. (A) Vector control (V1), ezrin antisense (AS1 and AS2), and N-ezrin–overexpressing (N-ezrin 1 and N-ezrin 2) LLC-PK1 cell lines were incubated in 24-well plates in 10% FCS/DMEM that contained recombinant hepatocyte growth factor (HGF) and collagen I. After 48 h, tubules were counted. Experiments were performed two to four times in triplicate, and results are expressed as tubule number. *P < 0.05 or **P < 0.01 compared with V1. (B) Photographs of HGF and collagen I–treated V1, V2, N-ezrin 1, N-ezrin 2, AS1, and AS2, respectively. Magnification, x100.
Ezrin Overexpression Overcomes Inhibition of Tubulogenesis by AGE-BSA
When vector control LLC-PK1 cells were incubated in the presenceof 40 µM AGE-BSA, tubulogenesis was inhibited comparedwith BSA-treated cells (Figure 3A). AGE-BSA significantly inhibitedtubulogenesis in vector controls to 68 ± 6% and 58 ±13% of vehicle control, respectively, whereas BSA had no effect(100 ± 3% and 126 ± 15%, respectively; P <0.05 and P < 0.001, respectively). For ezrin-overexpressingLLC-PK1 cell lines, AGE-BSA (40 µM) also significantlyinhibited tubulogenesis to 47 ± 4% and 43 ± 19%of vehicle control, respectively, whereas BSA had no significanteffect (121 ± 25% and 116 ± 8% of vehicle control,respectively; P < 0.01 for both). However, a reduced concentrationof AGE-BSA (20 µM) continued to inhibit tubulogenesisin vector control but not ezrin-overexpressing lines (Figure 3B).AGE-BSA significantly inhibited tubulogenesis in vectorcontrol lines to 51 ± 9% and 43 ± 3% of vehiclecontrol, respectively, compared with BSA-treated cells (103%± 15 and 75 ± 6% of vehicle control, respectively;P < 0.01 and P < 0.001, respectively). In contrast, AGE-BSAtreatment (20 µM) did not inhibit tubulogenesis in ezrin-overexpressingcell lines. Ezrin overexpression thus overcame the inhibitoryeffect of AGE-BSA in this model of kidney epithelial cell tubulogenesis.
Figure 3. Effect of ezrin overexpression on advanced glycation end products (AGE)-BSA inhibition of LLC-PK1 tubulogenesis. (A) 40 µM BSA or AGE-BSA. (B) 20 µM BSA or AGE-BSA. Vector control (V1 and V2) and ezrin-overexpressing (Ezrin 1 and Ezrin 2) cell lines were incubated in the presence of 10% FCS/DMEM that contained 3T3 CM as a source of HGF; collagen I; and PBS vehicle (black), BSA (dark gray), or AGE-BSA (light gray). Cells were also incubated in the absence of HGF with PBS vehicle (white). After 48 h, tubules were counted. Experiments were performed three to six times in triplicate, and results are expressed as a percentage of HGF+PBS control (100% = 56 tubules). **P < 0.01 and ***P < 0.001 versus HGF+PBS control; #P < 0.05, ##P < 0.01, and ###P < 0.001 versus BSA control.
LLC-PK1 Migration Is Inhibited by AGE-BSA
A wound migration model was also established in LLC-PK1 cells(Figure 4A). BSA had no effect on migration (107 ± 6%of control), whereas migration in the presence of AGE-BSA wassignificantly lower than that of BSA and control (86 ±7%, P < 0.05 versus control; P < 0.001 versus BSA; Figure 4B).These experiments were repeated with RNAse, an unrelatedprotein. Unglycated RNAse significantly increased migrationcompared with vehicle control (129 ± 10%; P < 0.05;Figure 4C). Glycation of RNAse produced similar inhibitory effectson LLC-PK1 migration compared with unglycated RNAse (93 ±5% versus 129 ± 10%; P < 0.01; Figure 4C).
Figure 4. Effect of AGE on LLC-PK1 basal migration. (A) Photographs of BSA- (top) and AGE-BSA–(bottom) treated LLC-PK1 cells. LLC-PK1 cells were wounded and then incubated in 1% FCS/DMEM that contained 1 µg/ml Mitomycin C and 40 µM BSA or 40 µM AGE-BSA. LLC-PK1 cells were wounded and then incubated in 1% FCS/DMEM that contained 1 µg/ml Mitomycin C and PBS vehicle (con), 40 µM BSA, or 40 µM AGE-BSA (B) or PBS vehicle (con), 40 µM RNAse (RNAse), or 40 µM AGE-RNAse (C). Wound areas were measured at t = 0 and t = 24 h. Experiments were performed between five and nine times in quadruplicate, and results are expressed as a percentage of vehicle control. *P < 0.05; **P < 0.01; ***P < 0.001.
LLC-PK1 Migration Is Ezrin-Dependent
LLC-PK1 migration was found to be ezrin-dependent as shown inFigure 5, where the basal rate of migration was reduced in N-ezrin–transfected(P < 0.05 for both compared with V1 and V2) and ezrin antisense–transfectedLLC-PK1 cell lines (P < 0.05 for both compared with V1 andV2).
Figure 5. Ezrin dependence of LLC-PK1 migration. Migration of vector control (V1 and V2), N-ezrin (N-ezrin 1 and N-ezrin 2), and ezrin antisense (AS1 and AS2) LLC-PK1 cell lines was studied. Cells were wounded and then incubated in 1% FCS/DMEM that contained 1 µg/ml Mitomycin C. Wound areas were measured at t = 0 and t = 24 h. Experiments were performed five to seven times in quadruplicate, and results are expressed as a percentage of vector control (V1). *P < 0.05 versus V1 or V2.
Ezrin Overexpression Overcomes Inhibition of Migration by AGE-BSA
The same concentration of AGE-BSA (20 µM) that inhibitedtubulogenesis in vector-control but not ezrin-overexpressingLLC-PK1 cell lines had a similar effect on migration (Figure 6).For the vector control clone V1, AGE-BSA treatment (70 ±5%) significantly decreased migration compared with vehiclecontrol (P < 0.01) and BSA (92 ± 6%; P < 0.05).For the vector control clone V2, AGE-BSA (88 ± 7%) significantlyinhibited migration compared with BSA (116 ± 16%; P <0.05) but not vehicle control. However, AGE-BSA treatment didnot significantly inhibit migration in either ezrin-overexpressingcell line. As with parent LLC-PK1 cells, BSA did not significantlystimulate migration in any clone.
Figure 6. Effect of ezrin overexpression on AGE-BSA inhibition of basal migration. Two vector control (V1 and V2) and two ezrin-overexpressing (Ezrin 1 and Ezrin 2) lines were incubated in 1% FCS/DMEM that contained 1 µg/ml Mitomycin C and PBS vehicle (white), 20 µM BSA (dark gray), or 20 µM AGE-BSA (light gray). Wound areas were measured at t = 0 and t = 24 h after the experiment. Experiments were performed four to five times in quadruplicate, and results are expressed as a percentage of PBS vehicle-treated cells. *P < 0.05 versus BSA-treated cells; **P < 0.01 versus control cells.
We showed recently that advanced glycated proteins bind to theN-terminal domain of the ERM family of proteins isolated fromdiabetic kidney as well as purified recombinant N-ezrin in vitro(9). We also demonstrated that AGE-BSA inhibited HGF-inducedtubulogenesis in LLC-PK1 cells, which is an ezrin-dependentphenomenon (15). We therefore hypothesized that AGE bindingto the N-domain of ERM proteins perturbs various cellular functionsby inhibiting ERM actions. If this is the case, we reasonedthat overexpression of an ERM protein should be able to overcomethis inhibition. Indeed, in our study, AGE-BSA inhibited tubulogenesisby up to 57% and migration by approximately 30%, but ezrin overexpressiontotally abrogated these effects. Furthermore, similar to AGE-BSA,inhibition of ezrin action using either antisense or overexpressionof dominant negative N-ezrin dramatically inhibited tubulogenesisand migration. These results strongly suggest that the inhibitoryeffect of AGE-BSA on tubulogenesis and migration in LLC-PK1cells involves interference of ezrin function by the glycatedmoiety, presumably as a consequence of binding to the N-domainof ezrin.
The effect of AGE on migration was tested using two unrelatedproteins: BSA and RNAse. Unglycated BSA did not significantlyaffect migration of LLC-PK1 cells, but RNAse stimulated migration.A similar increase in migration induced by an unglycated protein,when compared with control and glycated protein, was observedby Morita et al. (23). These results indicate the importanceof comparing the effects of glycated proteins with their unglycatedcounterparts to avoid possible confounding as a result of nonspecificeffects of the proteins. Our results clearly show that the effectsobserved with the glycated proteins are due to the glycatedmoiety.
Ezrin plays a significant role in proximal tubule cell migrationand tubulogenesis. Ablation of ezrin impairs membrane rufflingand migration in fibroblasts (24). Crepaldi et al. (15) showedincreased LLC-PK1 migration with ezrin overexpression. Ezrinphosphorylation is essential for its activation and transfectionof LLC-PK1 cells, while ezrin mutated at its tyrosine phosphorylationsites inhibited HGF-induced cell migration and tubulogenesis(15). We previously found that AGE-BSA inhibited tyrosine phosphorylationof ezrin by the EGF receptor (9). Therefore, inhibition of phosphorylationof ezrin by AGE-BSA could explain the dramatic inhibition oftubulogenesis and migration seen in these studies.
Ezrin exerts its cellular actions by using a number of downstreamintracellular signaling pathways. One of these is the phosphatidylinositol(PI) (3) kinase/Akt pathway (16, 25). Ezrin activates PI (3)kinase during tubulogenesis by binding to its p85 regulatorysubunit (16, 26). It therefore is possible that AGE may preventPI (3) kinase activation by inhibiting its interaction withp85. However, we confirmed that HGF stimulates Akt phosphorylation,a downstream target of PI(3) kinase, in LLC-PK1 cells in thisstudy but that AGE-BSA did not inhibit HGF-stimulated Akt phosphorylationin vector control or ezrin-overexpressing cells (results notshown). It therefore is unlikely that the inhibitory effectsof AGE on tubulogenesis and migration are mediated by this pathway.
HGF activates ERK and ERM proteins have been implicated in thisprocess (27). HGF-induced ERK activation is important for theinitiation of tubulogenesis, so we examined the possibilitythat AGE-induced inhibition of this pathway may underlie itseffects in LLC-PK1 cells (28). However, although HGF increasedERK phosphorylation in vector control and ezrin-overexpressingcells, AGE-BSA had no effect in either cell type (results notshown), suggesting that this pathway is not involved in theinhibitory effect of AGE.
ERM actions may also involve a number of other signaling pathways.Ezrin overexpression in LLC-PK1 cells activates focal adhesionkinase, which is involved in cell motility (29). ERM proteinsmay also affect cell motility through their interactions withRho family proteins (30). Rho GTPases, which include the Cdck2,RhoA, and Rac1 subfamilies, are critical for actin cytoskeletalre-organization in epithelial cell morphogenesis (31). Rho GDIis an inhibitory regulator of all of the Rho family members.The N-terminal domain of ERM interacts with Rho GDI and decreasesits inhibitory action, initiating activation of Rho family members(32). However, there is also evidence that ERM proteins mayinhibit Rho activity in some systems (30). ERM proteins arealso implicated in other signaling pathways, including thoseinvolving CD44 or protein kinase C (25, 27, 33, 34). The inhibitoryeffects of AGE on ezrin function could be mediated by one ormore of these pathways. Although beyond the scope of this study,exploring these possibilities is an important direction forfuture studies.
Actin reorganization and changes in cell shape and adhesionare required for ERM-dependent tubulogenesis and migration.Actin disassembly, cell shape alterations, and reduced adhesionhave been described in diabetes (35–37). We thereforepostulate that the high levels of AGE in diabetes and renalfailure may contribute to the development of diabetic nephropathyby inhibiting ERM actions, resulting in actin disorganizationand possibly disruption of signal transduction pathways suchas those described above.
Our studies provide evidence of the cellular consequences ofAGE binding to ERM proteins in proximal tubule cells. The ezrin-dependentprocesses of migration and tubulogenesis are inhibited by AGE-BSA,and ezrin overexpression overcame these effects. These resultssuggest that AGE may induce significant pathologic effects onproximal tubule cell function during the course of diabetesby inhibiting ERM function. Because the formation of AGE isa prominent cause of diabetic complications and ERM have importantmembrane-cytoskeletal functions, the AGE–ERM interactionmay be a novel target for therapies aimed at reducing diabeticnephropathy.
Acknowledgments
This work was supported by grants from the National Health andMedical Research Council of Australia, Servier Laboratories,and the Juvenile Diabetes Research Foundation International.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Vlassara H, Bucala R, Striker L: Pathogenic effects of advanced glycosylation: Biochemical, biologic and clinical implications for diabetes and aging.
Lab Invest 70
: 138
–151, 1994[Medline]
Beisswenger P, Makita Z, Curphey T, Moore L, Jean S, Brinck-Johnsen T, Bucala R, Vlassara H: Formation of immunochemical advanced glycosylation end products precedes and correlates with early manifestations of renal and retinal disease.
Diabetes 44
: 824
–829, 1995[Abstract]
Bierhaus A, Hofmann M, Ziegler R, Nawroth P: AGEs and their interaction with AGE-receptors in vascular disease and diabetes mellitus.
Cardiovasc Res 37
: 586
–600, 1998[Abstract/Free Full Text]
Vlassara H, Striker L, Teichberg S, Fuh H, Li Y, Steffes M: Advanced glycation end products induce glomerular sclerosis and albuminuria in normal rats.
Proc Natl Acad Sci U S A 91
: 11704
–11708, 1994[Abstract/Free Full Text]
Brownlee M, Vlassara H, Kooney A, Ulrich P, Cerami A: Aminoguanidine prevents diabetes-induced arterial wall cross-linking.
Science 232
: 1629
–1632, 1986[Abstract/Free Full Text]
Kihara M, Schmelzer J, Poduslo J, Curran G, Nickander K, Low P: Aminoguanidine effects on nerve blood flow, vascular permeability, electrophysiology and oxygen free radicals.
Proc Natl Acad Sci U S A 88
: 6107
–6111, 1991[Abstract/Free Full Text]
Bendayan M: Immunocytochemical detection of advanced glycation end products in rat renal tissue as a function of age and diabetes.
Kidney Int 54
: 438
–447, 1998[CrossRef][Medline]
Tanji N, Markowitz G, Fu C, Kislinger T, Taguchi A, Pischetsrieder M, Stern D, Schmit A, D’Agata V: Expression of advanced glycation end products and their cellular receptor RAGE in diabetic nephropathy.
J Am Soc Nephrol 11
: 1656
–1666, 2000[Abstract/Free Full Text]
McRobert E, Gallicchio M, Jerums G, Cooper M, Bach L: The amino-terminal domains of the ezrin, radixin and moesin (ERM) proteins bind advanced glycation end products: An interaction that may play a role in the development of diabetic complications.
J Biol Chem 278
: 25783
–25789, 2003[Abstract/Free Full Text]
Tsukita S, Yonemura S, Tsukita S: ERM (ezrin/radixin/moesin) family: From cytoskeleton to signal transduction.
Curr Opin Cell Biol 9
: 70
–75, 1997[CrossRef][Medline]
Pestonjamasp K, Amieva M, Strassel C, Nauseef W, Furthmayr H, Luna E: Moesin, ezrin and p205 are actin-binding proteins associated with neutrophil plasma membranes.
Mol Biol Cell 6
: 247
–259, 1995[Abstract]
Turunen O, Wahlstrom T, Vaheri A: Ezrin has a COOH-terminal actin-binding site that is conserved in the ezrin protein family.
J Cell Biol 126
: 1445
–1453, 1994[Abstract/Free Full Text]
Martin M, Andreoli C, Sahuquet A, Montcourrier P, Algrain M, Mangeat P: Ezrin NH2-terminal domain inhibits the cell extension activity of the COOH-terminal domain.
J Cell Biol 128
: 1081
–1093, 1995[Abstract/Free Full Text]
Krieg J, Hunter T: Identification of the two major epidermal growth factor-induced tyrosine phosphorylation sites in the microvillar core protein ezrin.
J Biol Chem 267
: 19258
–19265, 1992[Abstract/Free Full Text]
Crepaldi T, Gautreau A, Comoglio P, Louvard D, Arpin M: Ezrin is an effector of hepatocyte growth factor-mediated migration and morphogenesis in epithelial cells.
J Cell Biol 138
: 423
–434, 1997[Abstract/Free Full Text]
Gautreau A, Poullet P, Louvard D, Arpin M: Ezrin, as plasma membrane-microfilament linker, signals cell survival through the phosphatidylinositol 3-kinase/Akt pathway.
Proc Natl Acad Sci U S A 96
: 7300
–7305, 1999[Abstract/Free Full Text]
Nath K: The tubulointerstitium in progressive renal disease.
Kidney Int 54
: 992
–994, 1998[CrossRef][Medline]
Gilbert R, Cooper M: The tubulointerstitium in progressive diabetic kidney disease: More than an aftermath of glomerular injury?
Kidney Int 56
: 1627
–1637, 1999[CrossRef][Medline]
Oldfield M, Bach L, Forbes J, Nikolic-Paterson D, McRobert E, Thallas V, Atkins R, Osicka T, Jerums G, Cooper M: Advanced glycation end products cause epithelial-myofibroblast transdifferentiation via the receptor for advanced glycation end products (RAGE).
J Clin Invest 108
: 1853
–1863, 2001[CrossRef][Medline]
Amieva M, Litman P, Laiqiang H, Ichimaru E, Furthmayr H: Disruption of the dynamic architecture of NIH3T3 fibroblasts by the N-terminal domains of moesin and ezrin: In vivo imaging with GFP fusion proteins.
J Cell Sci 112
: 111
–125, 1999[Abstract]
Tikoo A, Varga M, Ramesh V, Gousella J, Maruta H: An anti-RAS function of neurofibromatosis type 2 gene product (Nf2/Merlin).
J Biol Chem 269
: 23387
–23390, 1994[Abstract/Free Full Text]
Montesano R, Schaller G, Orci L: Induction of epithelial tubular morphogenesis in vitro by fibroblast-derived soluble factors.
Cell 66
: 697
–711, 1991[CrossRef][Medline]
Morita K, Urabe K, Moroi Y, Koga T, Nagai R, Horiuchi S, Furue M: Migration of keratinocytes is impaired on glycated collagen I.
Wound Rep Reg 13
: 93
–101, 2005[CrossRef]
Lamb R, Ozanne B, Roy C, McGarry L, Stipp C, Mangeat P, Jay D: Essential functions of ezrin in maintenance of cell shape and lammellipodial extension in normal and transformed fibroblasts.
Curr Biol 7
: 682
–688, 1997[CrossRef][Medline]
Ng T, Parsons M, Hughes W, Monypenny J, Zicha S, Gautreau A, Arpin M, Gschmeissner S, Verveer P, Bastiaens P, Parker P: Ezrin is a downstream effector of trafficking PKC-intergrin complexes involved in the control of cell motility.
EMBO J 20
: 2723
–2741, 2001[CrossRef][Medline]
Keely P, Westwick J, Whitehead I, Der C, Parise L: Cdc42 and Rac1 induce integrin-mediated cell motility and invasiveness through PI(3)K.
Nature 390
: 632
–636, 1997[CrossRef][Medline]
Orian-Rousseau V, Chen L, Sleeman J, Herrlich P, Ponta H: CD44 is required for two consecutive steps in HGF/c-Met signaling.
Genes Dev 16
: 3074
–3086, 2002[Abstract/Free Full Text]
O’Brien L, Tang K, Kats E, Schutz-Geschwender A, Lipschutz J, Mostov K: ERK and MMPs sequentially regulate distinct stages of epithelial tubule formation.
Dev Cell 7
: 21
–32, 2004[CrossRef][Medline]
Poullet P, Gautreau A, Gress K, Girault J, Lovard D, Arpin M: Ezrin interacts with focal adhesion kinase and induces its activation independently of cell-matrix adhesion.
J Biol Chem 276
: 37686
–37691, 2001[Abstract/Free Full Text]
Bretscher A, Edwards K, Fehon R: ERM proteins and merlin: Integrators at the cell cortex.
Nat Rev Mol Biol 3
: 586
–599, 2002
van Aelst L, Symons M: Role of Rho family GTPases in epithelial morphogenesis.
Genes Dev 16
: 1032
–1054, 2002[Free Full Text]
Takahashi K, Sasaki T, Mammoto A, Takaishi K, Kameyama T, Tsukita S, Takai Y: Direct interaction of the Rho GDP disassociation inhibitor with ezrin/radixin/moesin initiates the activation of the Rho small G protein.
J Biol Chem 272
: 23371
–23375, 1997[Abstract/Free Full Text]
Legg J, Lewis C, Parsons M, Ng T, Isacke C: A novel PKC-regulated mechanism controls CD44-ezrin association and directional cell motility.
Nat Cell Biol 4
: 399
–407, 2002[CrossRef][Medline]
Zohar R, Suzuki N, Suzuki N, Arora P, Glogauer M, McCulloch C, Sodek J: Intracellular osteopontin is an integral component of the CD44-ERM complex involved in cell migration.
J Cell Physiol 184
: 118
–130, 2000[CrossRef][Medline]
Kim EK, Geroski DH, Holley GP, Urken SI, Edelhauser HF: Corneal endothelial cytoskeletal changes in F-actin with aging, diabetes and after cytochalasin exposure.
Am J Ophthalmol 114
: 329
–335, 1992[Medline]
Zhou X, Hurst R, Templeton D, Whiteside C: High glucose alters actin assembly in glomerular mesangial and epithelial cells.
Lab Invest 73
: 372
–383, 1995
Cortes P, Mendez M, Riser B, Guerion C, Rodriguez-Barbero A, Hassat C, Yee J: F-actin fiber distribution in glomerular cells: Structural and functional implications.
Kidney Int 58
: 2452
–2461, 2000[CrossRef][Medline]
Related Article
This Month’s Highlights
J. Am. Soc. Nephrol. 2006 17: 315-316.
[Full Text][PDF]
This article has been cited by other articles:
S.-Y. Goh and M. E. Cooper The Role of Advanced Glycation End Products in Progression and Complications of Diabetes
J. Clin. Endocrinol. Metab.,
April 1, 2008;
93(4):
1143 - 1152.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
