Vascular endothelial growth factor is a survivalfactor for renal tubular epithelial cells

JOHN KANELLIS, SCOTT FRASER, MARINA KATERELOS, AND DAVID A. POWERImmunology Research Center, St. Vincent’s Hospital, Melbourne, Victoria 3065, Australia

Kanellis, John, Scott Fraser, Marina Katerelos, andDavid A. Power. Vascular endothelial growth factor is asurvival factor for renal tubular epithelial cells. Am J PhysiolRenal Physiol 278: F905–F915, 2000.—Vascular endothelialgrowth factor (VEGF) acts primarily as an endothelial cellmitogen via the ‘‘endothelial cell-specific’’ receptors VEGFR-1(flt-1) and VEGFR-2 (flk-1/KDR). Only a few nonendothelialcells have been shown to possess functional VEGF receptors.We therefore examined the rat renal tubular epithelial cellline NRK52-E. NRK52-E expressed VEGFR-1 and VEGFR-2mRNA and protein by RT-PCR, Northern blotting, Westernblotting, immunofluorescence, and ligand binding. Serum-starved NRK52-E incubated with VEGF showed a significantincrease in [3H]thymidine incorporation compared with con-trol (2.3-fold at 1–10 ng/ml, P , 0.05; 3.3-fold at 50–100ng/ml, P , 0.01). VEGF also protected NRK52-E from hydro-gen peroxide-induced apoptosis and necrosis compared withcontrol (annexin-V-FITC-positive cells, 39 vs. 54%; viablecells, 50.5 vs. 39.7%). Immunohistochemical staining using avariety of antibodies showed expression of both VEGF recep-tors in normal rat renal tubules in vivo. Because VEGFinduced a proliferative and an antiapoptotic response in renaltubular epithelial cells, these data suggest that VEGF mayact as a survival factor for renal tubular epithelium in vivo.

vascular endothelial growth factor receptors; flt-1; flk-1;vascular endothelial growth factor receptor 1; vascular endo-thelial growth factor receptor 2; apoptosis

VASCULAR ENDOTHELIAL GROWTH FACTOR (VEGF) is apotent endothelial cell mitogen that promotes angiogen-esis, increases vascular permeability, and is chemotac-tic for monocytes (13, 14). VEGF has been shown tohave a role in a wide variety of situations, includingembryogenesis, placental growth, tumor growth, diabe-tes, wound healing, inflammatory responses, and tissueremodeling (13, 14). There are two known receptors forVEGF, previously described as flt-1 and KDR/flk-1, nowdesignated VEGF receptor 1 (VEGFR-1) and VEGFreceptor 2 (VEGFR-2), respectively. Other VEGF-likereceptors also exist, such as VEGF receptor 3 (VEGFR-3or flt-4) (26, 27, 35, 39). All of these are type III tyrosinekinases, characterized by seven immunoglobulin-likeloops within their extracellular domain and a splitkinase domain within the cytoplasmic moiety. VEGFand related factors such as placental growth factor and

the recently described novel VEGF molecules VEGF-B,VEGF-C, and VEGF-D are ligands for VEGF receptors(6, 24, 26). VEGF receptors undergo dimerization andautophosphorylation after ligand binding, leading toactivation of intracellular signaling molecules such asMAP kinase and phospholipase C (26, 41).

Until recently, most studies have described VEGFreceptor expression as specific to endothelial cells. Thediscovery of flt-1 (VEGFR-1) on monocytes, and itsability to mediate monocyte chemotaxis in response toVEGF (3), was one of the first examples of nonendothe-lial cells possessing functional VEGF receptors. Thereare now several descriptions of nonendothelial cellsexpressing VEGF receptors, but most of these have notdemonstrated function. For example, VEGF receptorprotein or mRNA has been reported in rat mesangialcells (33), hepatocytes (VEGFR-1 and VEGFR-2) (30),Leydig and Sertoli cells (VEGFR-1 and VEGFR-2) (10),and in endometrial epithelium (VEGFR-2) (7) withoutdemonstration of function. Cell lines that have beenreported to express functional receptors include osteo-blasts (22), human retinal pigment epithelial cells (19),pancreatic duct epithelium (VEGFR-2) (29), and uter-ine smooth muscle cells (5).

This study demonstrates the presence of VEGFR-1and VEGFR-2 protein and mRNA on the renal tubularepithelial cell line NRK52-E, as well as histologicalevidence for VEGFR-1 and VEGFR-2 protein expres-sion on rat renal tubular epithelium in vivo. In addi-tion, this study demonstrates that VEGF can induceproliferation of these cells when serum deprived andprotect against hydrogen peroxide-induced apoptosisand necrosis. These data suggest that VEGF andrelated ligands may function as survival factors forrenal tubular epithelial cells in vivo.

METHODS

Cell culture. NRK-52E, a nontransformed tubular epithe-lial cell line from normal rat kidney (American Type CultureCollection no. CRL-1571), was maintained in Dulbecco’smodified Eagle’s medium supplemented with 15 mM HEPESbuffer (GIBCO-BRL), 10% fetal calf serum, 100 U/ml penicil-lin, and 100 µg/ml streptomycin (Commonwealth SerumLaboratory) at 37°C in standard incubators (95% air-5%CO2). NRK52-E cells have been cloned from a mixed culture ofnormal rat kidney cells and possess characteristics of bothproximal and distal tubular epithelial cells (8). Bovine aorticendothelial cells (BAEC) were grown under identical condi-tions with media further supplemented with cis-hydroxypro-line (20 µg/ml) (Sigma Chemical, St Louis, MO).

Isolation of total RNA. Total RNA was extracted fromcultured cells by using Trizol (Life Technologies, GIBCO-

The costs of publication of this article were defrayed in part by thepayment of page charges. The article must therefore be herebymarked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734solely to indicate this fact.

Am J Physiol Renal Physiol278: F905–F915, 2000.

0363-6127/00 $5.00 Copyright r 2000 the American Physiological Society F905http://www.ajprenal.org

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

BRL, Melbourne, Australia) according to the manufacturer’sspecifications. Sample RNA levels were quantitated by read-ing the absorbance at 260 nm. Final samples were stored at270°C until required for RT-PCR and Northern blot analysis.

Northern blotting. VEGFR-1 and VEGFR-2 cDNA insertswere PCR amplified from reverse transcribed rat kidney RNAby using primer sequences obtained from the known ratreceptor sequences (42, 43) (accession nos. D28498 andU93306, respectively). Primers for the VEGFR-1 insert (for-ward 58-CAAGGGACTCTACACTTGTC-38 and reverse58-CCGAATAGCGAGCAGATTTC-38) resulted in a 240-bpproduct corresponding to a portion of the extracellular do-main (amino acid residues 305–384). Primers for the VEGFR-2insert were as described by Wen and co-workers (42) (forward58-GCCAATGAAGGGGAACTGAAGAC-38 and reverse58-TCTGACTGCTGGTGATGCTGTC-38). These produced a537-bp product corresponding to the intracellular, NH2-terminal end of the tyrosine kinase domain of the receptor(amino acid residues 870–1049). The PCR products werecloned into pGEM-T easy (Promega, Madison, WI), and theDNA sequences were confirmed by sequencing. The VEGFR-1insert was excised from the vector by using the restrictionenzyme EcoR I (Promega), whereas the VEGFR-2 insert wasexcised by using Nco I and Sal I (Promega). A murine GAPDHinsert was obtained as a 1.2-kb Pst I fragment in clonepHcGAP (37). Total RNA samples (,15 µg/well) were fraction-ated on a 1% agarose-formaldehyde gel and transferred toGenescreen Plus membranes (NEN Life Sciences, Boston,MA). Membranes were cross-linked by using a Stratalinker(Stratagene, La Jolla, CA) and then prehybridized for 1 h at65°C by using Rapid Hyb buffer (Amersham International).Inserts were labeled by using the Megaprime DNA labelingsystem (Amersham, Bucks, UK) and added to fresh RapidHyb buffer at 2 3 106 counts/ml hybridization fluid. Mem-branes were hybridized for 2 h at 65°C and then washed threetimes for 20 min each [first wash in 23 standard sodiumcitrate (SSC)/0.1% SDS at 65°C, second wash in 13 SSC/0.1%SDS at 65°C, third wash in 0.13 SSC/0.1% SDS at roomtemperature] before exposure to X-ray film (Kodak).

RT-PCR. First-strand cDNA was synthesized fromNRK52-E cell total RNA by using AMVRT and oligo(dT)(Promega). The subsequent PCR reaction used the sameVEGFR-1 and VEGFR-2 primers described earlier to producecDNA inserts for Northern blotting experiments. PCR prod-ucts were run on 1% agarose gels and analyzed underultraviolet light. NRK52-E cell RNA samples without AMVRTwere used as the negative control in PCR assays.

Laser scanning confocal fluorescence microscopy. All pri-mary antibodies and blocking peptides described were pur-chased from Santa Cruz Biotechnology. Cells were grown on22 3 22-mm glass coverslips until 70% confluent, washedwith PBS, and then fixed in 3.2% paraformaldehyde. Parafor-maldehyde was neutralized with 150 mM glycine in PBS. Thecells were permeabilized with 0.3% Triton X-100 (Bio-Rad),blocked with 5% BSA for 30 min, and then incubated over-night with anti-receptor antibodies in 0.3% Triton X-100 and0.025% 3-[(3-cholamidopropyl)-dimethylammonio]-1-propane-sulfonate (CHAPS; Sigma Chemical). For VEGFR-1, C-17, apolyclonal rabbit anti-human antibody, was used (directedagainst amino acid residues 1312–1328). For VEGFR-2,three different antibodies were used: 1) A-3, a monoclonalmouse anti-mouse antibody directed against amino acidresidues 1158–1345; 2) C-20, a polyclonal rabbit anti-mouseantibody directed against amino acid residues 1326–1345;and 3) N-931, a polyclonal anti-mouse antibody directedagainst amino acid residues 931–997. All primary antibodieswere used at a concentration of 1 µg/ml. Negative controls

were performed by using normal rabbit IgG or an isotype-matched, irrelevant monoclonal antibody at the same concen-trations as primary antibodies. In addition, to further confirmantibody specificity, blocking peptides were used where avail-able (C-17 and C-20). Primary antibodies were incubatedwith their specific blocking peptide or with an irrelevantpeptide for 2 h at room temperature (concentration of peptide10 µg/ml). After incubation with primary antibodies, the cellswere washed once with PBS containing 0.3% Triton X-100and 0.025% CHAPS and then a further two times with PBSalone. Secondary immunofluorescent antibodies (all pur-chased from Molecular Probes, Eugene, OR) were goat anti-rabbit Texas red (to detect C-17), goat anti-rabbit Oregongreen (to detect N-931 and C-20), and goat anti-mouse Oregongreen (to detect A-3). Incubations were for 1 h at roomtemperature. Cells were washed a further three times withPBS, and then the coverslips were mounted with a water-soluble mountant (Aquamount; BDH, Kilsyth, Victoria, Aus-tralia) and analyzed. Images were obtained and generated ona confocal laser scanning microscope (Bio-Rad MRC 1024,Bio-Rad Microscopy Division, Hemel, Hempstead, Herts,UK). BAEC were examined for expression of both receptorsand compared with NRK52-E.

Western blots. Whole cell lysates were obtained from cellsgrown to confluence in 150-cm2 flasks. Whole cell lysis buffer(25 mM HEPES, 0.3 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA,0.5% Triton X-100) was supplemented with 1 mM phenylmeth-ylsulfonyl fluoride (Calbiochem), 1 µM leupeptin (ICN), 0.2µM aprotinin (ICN), and 1 mM 1,4-dithiothreitol (Bio-Rad).Lysates were centrifuged at 18,000 g for 5 min at 4°C, andpellets were discarded. Protein samples mixed in reducingbuffer were resolved on 7.5% SDS-PAGE gels (,40 µg protein/lane) and transferred to nitrocellulose membranes (Trans-Blot Transfer medium; Bio-Rad) by electroblotting. Mem-branes were blocked in 5% wt/vol nonfat milk powder inTris-buffered saline (TBS) for 30 min at room temperature.The VEGF receptor antibodies used were as described forimmunofluorescence. The antibodies were diluted to 2 µg/mlin TBS containing 0.1% azide. For VEGFR-2 membranes, theblocking and primary antibody solutions were supplementedwith 5% rabbit serum. Membranes were incubated withprimary antibody solutions overnight at 4°C and washedthree times for 5 min each in TBS containing 0.05% wt/volTween (Bio-Rad). Secondary antibody incubations were per-formed for 30 min at room temperature. VEGFR-1 mem-branes were incubated with horseradish peroxidase-linkedprotein A (Amersham,) at 1:5,000, and VEGFR-2 membraneswere incubated with horseradish peroxidase-linked rabbitanti-mouse antibody (Dako) at 1:1,000 in 5% rat serum. Aftera further three washes (5 min each in TBS with 0.05% wt/volTween) immunoreactive proteins were detected according tothe enhanced chemiluminescence protocol (Amersham). Blotswere analyzed after exposure to autoradiography film (Hyper-film ECL, Amersham).

Binding assay. NRK52-E were seeded in complete media in24-well plates and grown to 90% confluence. Cells werewashed with cold binding buffer (Dulbecco’s modified Eagle’smedium, 25 mM HEPES buffer, 1% BSA) and then incubatedfor 2 h at 4°C with binding buffer containing 10 pM [125I]-VEGF165 (specific activity 105 counts·min21 ·ng21; NEN). Tocompete with the binding of [125I]-VEGF165, unlabeled ‘‘com-petitor’’ growth factor was also added. Recombinant humanVEGF165 (rhVEGF165; ligand for both receptors), recombinanthuman placental growth factor (rhPlGF; ligand for VEGFR-1only), or epidermal growth factor (EGF; irrelevant control)were used at various concentrations (0, 0.1, 1, 10, and 100ng/ml; R&D Systems, Minneapolis, MN). Supernatants were

F906 VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

subsequently removed, and the cells were washed twice incold binding buffer and incubated for 30 min with 1% SDS in0.4 M NaOH to lyse the cells. [125I]-VEGF165 binding wasmeasured by using a gamma counter (Packard Cobra Auto-gamma 5005, Meriden, CT). Each competing rhVEGF165,rhPlGF, and EGF concentration was assessed in quadrupli-cate. Results were expressed as a percentage of maximum[125I]-VEGF165 binding (where no competitor was added).

Proliferation assay. Cells were seeded into 24-well plates(105/well), serum deprived, and then incubated with rhVEGF165(0, 1, 10, 50, and 100 ng/ml; R&D Systems) in 1% BSA. Theproliferation assay was performed in two different ways. Inthe first group of experiments, cells were seeded in 0.5% fetalcalf serum, incubated overnight, then incubated for 72 h inserum-free media before the addition of rhVEGF165 and[3H]thymidine (1 mCi/well; NEN). In the second group ofexperiments, cells were seeded in serum-free media, leftovernight, and then incubated with rhVEGF165 and [3H]thymi-dine. To adequately control the experiments, an equivalentamount of BSA was added to each well (i.e., the same amountof protein as in the rhVEGF165 wells). As heparin has beenshown to modulate VEGF receptor binding (18, 36), cells withand without heparin (0.1 ng/ml) were also assessed. At theend of the stimulation periods (24-, 48-, and 72-h incubationat 37°C), cells were washed with PBS, lysed with 200 µl of 1 MNaOH, and then filtered through glass-fiber filter paper byusing a cell harvester (Inotech, Dottikin, Switzerland). Spe-cific activity for each well was measured by using a betacounter (Packard Tri-Carb 1600 CA).

Annexin-V-FITC and propidium iodide binding. Apoptosiswas induced by using a modification of previously usedmethods for NRK52-E (34). Briefly, cells were seeded into25-cm2 flasks in media containing 0.1% fetal calf serum andleft overnight. To induce apoptosis, cells were washed andmedia containing 0.1% fetal calf serum was added along withhydrogen peroxide (0.5–1.0 mM) for 6 and 24 h. At the sametime, rhVEGF165 (100 ng/ml) or BSA (control) was added. Theamount of BSA added to the control solution was equivalentto that in the rhVEGF165 solution. Media containing any deadcells was collected and added to cells harvested from the flask.Cells were washed twice and then stained with annexin-V-FITC and propidium iodide (Biosource, Camarillo, CA) accord-ing to the manufacturer’s specifications. Flow cytometry(FACSCalibur, Becton-Dickinson, Sunnyvale, CA) was usedto determine the number of apoptotic, necrotic, and viablecells in each group. Cells positive for annexin-V-FITC wereassessed as apoptotic. Double-stained cells (annexin-V andpropidium iodide) were assessed as late apoptotic or necrotic.Cells negative for both stains were assessed as viable.

Immunohistochemical staining of rat kidney Paraffin-embedded tissue sections (4 µm thick) of normal rat kidneyfixed in 4% paraformaldehyde were analyzed for the presenceof VEGFR-1 and VEGFR-2. The commercial antibodies andblocking peptides already described were used, applyinghorseradish peroxidase immunohistochemical staining meth-ods. Endogenous peroxidase was quenched with 4% hydrogenperoxide in methanol, and nonspecific binding was blocked byusing 10% swine serum before incubation with primaryantibodies overnight at 4°C (concentration of 1 µg/ml). Nega-tive control materials used were normal rabbit IgG (SigmaChemical) in place of the primary antibody for the polyclonalantibodies and an irrelevant, isotype-matched monoclonalantibody in place of the monoclonal antibody. Once again,specificities of the C-17 and C-20 antibodies were furtherverified by using blocking peptides as described for fluores-cence microscopy. A peroxidase kit was used for subsequentsteps (LSAB 2 peroxidase kit, Dako) followed by development

of staining with diaminobenzidine (Dako). Counterstainingin both groups was performed with hematoxylin. To deter-mine the location of tubular staining seen with the receptorantibodies, sequential sections were stained with varioustubular markers as previously described (25). The fluorescein-labeled lectin Arachis hypogea (AH; Sigma Chemical) wasused to identify distal convoluted tubules (DCT) and collect-ing ducts, whereas Phaseolus vulgaris erythroagglutinin(PHA-E; Sigma Chemical) was used for proximal convolutedtubules (PCT). Sheep anti-Tamm-Horsfall protein antibody(gift from Dr. H. Y. Lan, Monash Medical Center, Clayton,Victoria, Australia) was used to identify cortical and medul-lary thick ascending limbs of the loop of Henle.

Statistics. Results from binding and proliferation assayswere analyzed by using Instat 2.01 (GraphPad software). Aone-way ANOVA was performed to determine whether therewas a significant difference between experimental and con-trol groups. Specific statistical tests used were multiplecomparison Bonferroni (binding assay) and Dunnett (prolifera-tion assay) tests. P values of ,0.05 were deemed significant.

RESULTS

RT-PCR and Northern blots. Both RT-PCR and North-ern blotting experiments demonstrated VEGFR-1 andVEGFR-2 mRNA transcripts in the NRK52-E. DNAbands from the RT-PCR were 240 (VEGFR-1) and 537bp (VEGFR-2) as predicted (not shown). Bands ob-tained on Northern blots were of the correct size, ,7.2kb for VEGFR-1 and 6.8 kb for VEGFR-2 (Fig. 1). Asecond mRNA species of ,4.2 kb was shown onVEGFR-1 blots. This may represent an alternativelyspliced isoform and is in keeping with observations byother groups (6, 40).

Laser scanning confocal fluorescence microscopy. Fluo-rescence microscopy showed the presence of both VEGFreceptors in NRK52-E (Fig. 2). VEGFR-1 staining wasevident in the cytoplasm and on the membranoussurface of the cells (Fig. 2A), with some showingprominent staining in vesicle-like structures within thecytoplasm. VEGFR-2 staining was evident with allthree antibodies. The cytoplasmic staining was similarin all groups with a prominent perinuclear pattern,

Fig. 1. Northern blot of total RNA samples from NRK52-E. Northernblot demonstrates mRNA for vascular endothelial growth factorreceptor 1 (VEGFR-1; R1; left lane) and 2 (VEGFR-2; R2; right lane)in normal NRK52-E. A single mRNA species was identified forVEGFR-2 (6.8 kb). Two mRNA species were identified for VEGFR-1(7.2 and 4.2 kb). Total RNA was ,15 µg/well.

F907VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

particularly with the A-3 and C-20 antibodies (Fig. 2, Cand E). In addition, prominent nuclear staining wasseen by using the N-931 antibody (Fig. 2F). Nuclearlocalization of VEGFR-2 has recently been observed inendothelial cells by others (11, 38), but the significance

of our observation is unclear at this stage. The specificblocking peptides inhibited the staining observed (Fig.2, B and D), whereas the use of irrelevant peptides didnot (Fig. 2, A and C). No fluorescence was seen innegative control cells for A-3 and N-931 antibodies (not

Fig. 2. Laser scanning confocal fluorescence microscopyon NRK52-E (A-F) and Bovine aortic endothelial cells(BAEC; G, H). Staining for VEGFR-1 was in red (A, B,G) and for VEGFR-2 in green (C-F, H). Antibodies (C-17ab, C-20 ab, A-3 ab, N-931 ab) and blocking peptides(C-17 p, C-20 p) used are indicated. VEGFR-1 stainingof NRK52-E revealed membranous, cytoplasmic, andslight nuclear staining (A). Some cells had prominentlystained vesicle-like structures in cytoplasm. VEGFR-2staining showed a prominent cytoplasmic and peri-nuclear pattern (C, E, F), with N-931 antibodies alsoshowing strong nuclear staining (F). Specific blockingpeptides to C-17 and C-20 antibodies inhibited staining,whereas use of an irrelevant blocking peptide had noeffect (A-D). Negative controls forA-3 and N-931 antibod-ies (not shown) demonstrated no fluorescence, resem-bling Fig. 2, B and D. BAEC showed similar stainingwith all antibody and peptide combinations. BAECappearances by using C-17 and A-3 antibodies areshown (G, H).

F908 VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

shown). BAEC showed similar staining to the NRK52-Ewith the use of the various antibody and peptidecombinations. VEGF receptor staining in BAEC byusing the C-17 and A-3 antibodies are shown (Fig. 2, Gand H).

Western blots of NRK52-E cell protein. Blots ofNRK52-E whole cell lysates showed bands correspond-ing to the known sizes of VEGFR-1 and VEGFR-2. Bothreceptors are ,200 kDa in size, but this may varydepending on glycosylation and whether the receptor iscomplexed with its ligand (18, 36, 41). Bands of ,200and 170 kDa were demonstrated in the VEGFR-1 blots(Fig. 3A). Whole cell lysates prepared from BAEC wereused as positive controls and demonstrated bands of,200 and 180 kDa. VEGFR-2 blots also demonstratedtwo bands of the expected size in NRK52-E (,200 and

180 kDa), with a similar 180-kDa band found in BAEC(Fig. 3B).

Binding assay. Binding assay for [125I]-VEGF165showed that both rhVEGF165 and rhPlGF bound stronglyto the cells. Both growth factors were able to competeout the binding of [125I]-VEGF165 on the cells (Fig. 4).EGF had no effect on [125I]-VEGF165 binding. Thisfurther supports the presence of VEGFR-1 andVEGFR-2 on NRK52-E. Binding of [125I]-VEGF165 wasdecreased to ,25–40% of normal with the addition of$1 ng/ml of rhVEGF165 or rhPlGF (P , 0.01, Bonferronimultiple comparison test).

Proliferation assay. VEGF significantly stimulatedNRK52-E proliferation in both sets of experiments (Fig.5). Heparin (0.1 ng/ml) did not augment this response.The most significant effect occurred in cells seededwithout serum and then serum deprived overnightbefore incubation with rhVEGF165 (Fig. 5A). After 24 hof incubation with rhVEGF165 at 1 and 10 ng/ml, cellshad a 2.3-fold increase in [3H]thymidine incorporation(P , 0.05, Dunnett multiple comparison test) and a3.3-fold increase at 50 and 100 ng/ml of rhVEGF165 (P ,0.01, Dunnett multiple comparison test). Where thecells were seeded in 0.5% fetal calf serum and thenserum deprived for 72 h before incubation with rhVEGF165,the proliferative effect of VEGF was not evident until72 h of rhVEGF165 incubation (Fig. 5B). At 24- and 48-hincubation, rhVEGF165-incubated cells and control cells(incubated in BSA) had similar levels of proliferation. A1.5-fold increase in [3H]thymidine incorporation wasseen at 72 h with rhVEGF165 concentrations of 50 and100 ng/ml (P , 0.05 and P , 0.01, respectively, Dunnettmultiple comparison test). Results without heparin areshown (Fig. 5).

Annexin-V-FITC and propidium iodide binding. Cellsincubated with rhVEGF165 were protected from hydro-

Fig. 3. Western blots of protein from whole cell lysates of NRK52-Eand BAEC incubated with antibodies for VEGF-R1 (A) and VEGF-R2(B). NRK52-E lysates (right lanes) were compared with BAEC lysates(left lanes), which showed similar-size bands for both receptors.VEGFR-1 blots (A) showed bands of 170 and 200 kDa in NRK52-Elysates (right lane, black arrows) and 180 and 200 kDa in BAEClysates. These are in keeping with known sizes of VEGFR-1. Both celltypes also showed a strong band at ,100–110 kDa, the identity ofwhich is unknown, although this may represent a degradationproduct or truncated receptor isoform. VEGFR-2 blots (B) showedbands in keeping with known sizes of receptor in both NRK52-E (180and 200 kDa, right lane, gray arrows) and BAEC lysates (180 kDa,left lane, black arrow).

Fig. 4. VEGF receptor-binding assay. [125I]-VEGF165 binding toNRK52-E. NRK52-E were incubated with [125I]-VEGF165 (10 pM) andunlabeled growth factor. Various concentrations of recombinant(rh)VEGF165 (dashed line), placental growth factor (rhPlGF; solidline) or EGF (dashed-dotted line) were used to compete for sites inbinding assay. Both rhVEGF165 and rhPlGF significantly inhibited[125I]-VEGF165 binding at concentrations of $1 ng/ml [binding 25–40% of maximum (Max); **P , 0.01]. EGF had no effect on[125I]-VEGF165 binding.

F909VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

gen peroxide-induced apoptosis and necrosis. In cellsincubated with rhVEGF165 for 6 h (Fig. 6), a lowerproportion showed annexin-V-FITC staining comparedwith control cells (39 vs. 54%). The proportion of viablecells (negative for both annexin-V-FITC and propidiumiodide) was also higher in the rhVEGF165-incubatedgroup (50.5 vs. 39.7%). In cells incubated with rhVEGF165for 24 h (not shown), similar results were observedcompared with control (annexin-V-FITC positive cells:28.4 vs. 38.9%; viable cells: 67.5 vs. 57.3%). The major-ity of annexin-V-FITC-positive cells showed double

staining with propidium iodide, indicating the cellswere necrotic or at a late stage of apoptosis. Few cells ineach group showed only single staining for annexin-V-FITC (early apoptosis). Results shown are representa-tive of three separate experiments using the sameconditions.

Immunohistochemical staining of normal rat kidney.Normal rat kidney tubular epithelium showed stainingfor both VEGFR-1 and VEGFR-2. Sequential stainingwith the lectins AH (localizes DCT and collecting ducts)and PHA-E (localizes PCT) was used to localize recep-tor staining where this occurred. VEGFR-1 staining byusing the C-17 antibody was localized to both proximaland distal tubules of the cortex and to S3 segments ofthe PCT in the outer medulla (Fig. 7, A-C). There wasprominent staining of the brush border in cells of the

Fig. 5. Proliferation of serum-deprived NRK52-E in response torhVEGF165. Cells were seeded without serum (24 h; A) or seeded with0.5% serum and then serum deprived (72 h; B). Addition of heparinhad no effect. Only results without heparin are shown. For detaileddescription of conditions, see METHODS. When seeded without serum,rhVEGF165 (24-h incubation) significantly stimulated proliferation ofserum-deprived NRK52-E at all concentrations compared with cellsincubated with BSA alone. rhVEGF165 (1 and 10 ng/ml) produced a2.3-fold increase in [3H]thymidine incorporation (*P , 0.05); 50 and100 ng/ml rhVEGF165 produced a 3.3-fold increase in [3H]thymidineincorporation (**P , 0.01). When seeded with serum and then serumdeprived (B), longer incubation with rhVEGF165 was needed (72 h)before an effect was seen. rhVEGF165 (50 and 100 ng/ml) produced,1.5-fold increase in [3H]thymidine incorporation (*P , 0.05 and**P , 0.01). DPMI, disintegrations/min.

Fig. 6. Survival effect of VEGF. Annexin-V-FITC and propidiumiodide staining of hydrogen peroxide-treated NRK52-E. NRK52-Eseeded in 0.1% fetal calf serum were incubated for 6 h with hydrogenperoxide (0.75 mM) and either rhVEGF165 (100 ng/ml; A) or BSA (B).Proportion of annexin-V-FITC-stained cells was lower in cells treatedwith VEGF compared with control (39 vs. 54%; right). Majority ofannexin-V-FITC-positive cells showed double staining with prop-idium iodide (top right), indicating cells were late apoptotic ornecrotic. Proportion of cells that were viable (negative for both stains;bottom left), stained with annexin-V-FITC alone (bottom right), andpropidium iodide alone (top left) are also shown. Proportion of viablecells was higher in cells treated with VEGF compared with control(50.5 vs. 39.7%).

F910 VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

PCT (Fig. 7, B and C). In the medulla there was mild,generalized VEGFR-1 staining of tubular structuresincluding collecting ducts and loop of Henle (not shown).There was little evidence of staining in glomeruli,peritubular capillaries, and larger vessels. Preincuba-tion of C-17 antibody with C-17p blocking peptideinhibited staining (Fig. 7D).All three VEGFR-2 antibod-ies showed a similar staining pattern in the kidney,with prominent tubular epithelial staining. With themonoclonal antibody (A-3), VEGFR-2 staining wasstrongly localized to the macula densa and DCT of thecortex and to collecting ducts in the inner and outermedulla (Fig. 8, D-F). The localization of VEGFR-2staining was very prominent, with a gradient of stain-ing from the cortex down to the outer and innermedulla (Fig. 8D). Proximal tubules and the loop ofHenle showed minimal staining with the A-3 antibody.The two polyclonal antibodies demonstrated tubularstaining that was more generalized. The C-20 antibodydemonstrated prominent staining of DCT and collect-ing ducts (Fig. 8A). Preincubation with C-20p blockingpeptide inhibited this staining (Fig. 8B). Some nuclearstaining was once again observed with the N-931

antibody, as were glomerular and peritubular capillarystaining (Fig. 8C). Although endothelial cells in thekidney did not show prominent staining with all theantibodies, endothelial cell-specific staining was demon-strated in adult rat heart and lung specimens and avariety of neonatal rat specimens (not shown). Themost prominent endothelial staining was observed byusing the C-17 and N-931 antibodies, confirming thespecificity of these antibodies to VEGF receptors on ratendothelium. All negative controls demonstrated nostaining, in particular, controls using blocking peptidesto the C-17 (VEGFR-1) and C-20 (VEGFR-2) antibod-ies. Preincubation of the C-17 antibody with the C-20peptide, and the C-20 antibody with the C-17 peptide,did not inhibit staining.

DISCUSSION

VEGF is an important angiogenic growth factor thatsignals via VEGF receptors on endothelial cells (12, 14,20). Recent evidence, however, supports a much widerrole for VEGF with reports demonstrating receptors ona variety of nonendothelial cells (5, 19, 22, 29). This

Fig. 7. Immunohistochemical staining for VEGFR-1 in normal rat kidney by using C-17 antibody. A: low-powerview showing staining for VEGFR-1 in proximal and distal tubular structures of cortex (Cx) and in S3 segments ofproximal collecting tubule (PCT) in outer medulla (OM). Magnification 310. B : renal cortex showing prominentstaining for VEGFR-1 predominantly involving PCT, with localization of staining to brush border of cells (w; groupsof proximal tubules). Magnification 325. C : high-power view of renal cortex showing staining for VEGFR-1 onbrush border of PCT cells (arrows). Magnification 3100. D: negative control by using C-17 antibody preincubatedwith C-17 blocking peptide showing no staining. Magnification 380.

F911VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

study is the first to demonstrate functional VEGFreceptors on nonendothelial cells of the kidney.

The rat renal tubular epithelial cell line NRK52-Ewas found to express protein and mRNA for both

VEGFR-1 and VEGFR-2. The sizes of the mRNA spe-cies are in keeping with published studies for bothendothelial and nonendothelial cells (3, 5, 23, 35).Reports vary in terms of the accepted sizes of the

Fig. 8. Immunohistochemical staining for VEGFR-2 in normal rat kidney by using C-20, N-931, and A-3 antibodies.A: C-20 antibody: OM showing prominent collecting duct staining (arrows) and diffuse staining of other tubules.Magnification: 3100. B: C-20 antibody preincubated with C-20 blocking peptide: OM showing inhibition of alltubular staining. Magnification 3100. C: N-931 antbody: renal cortex showing prominent peritubular capillary(arrowheads) and glomerular endothelial cell staining (arrows). Prominent staining of macula densa and DCT isalso shown (T), with only slight staining of remaining cortical proximal tubular structures. Magnification 3180.Diffuse tubular staining was evident in medulla (not shown). D: A-3 antibody: low-power view showing prominentstaining for VEGFR-2 in cells of collecting duct in OM and inner medulla (IM). Magnification 310. E: A-3 antibody:IM showing prominent staining for VEGFR-2 in cells of collecting duct. Magnification 325. F: A-3 antibody:high-power view of normal renal cortex showing prominent staining for VEGFR-2 in cells of DCT (w) and maculadensa (arrows). Magnification 3100. Negative controls for both N-931 and A-3 antibodies demonstrated no staining(not shown), but appearances resembled those where blocking peptides were used (Figs. 7D and 8B).

F912 VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

protein isoforms for the receptors with VEGFR-1, rang-ing from 170 to 210 kDa, and VEGFR-2 ranging from180 to 235 kDa (17, 27, 36, 41). Part of this uncertaintymay relate to the existence of alternatively splicedisoforms. Studies that have performed affinity cross-linking of VEGF receptors generally report the exis-tence of several bands in the range of 170–235 kDa (35,39). In addition, there are reports demonstrating afunctional, truncated form of VEGFR-2 in rat retinaltissue (42) and a soluble variant of VEGFR-1 in humanvascular endothelial cells (2).

VEGF induced a proliferative response in serum-deprived NRK52-E. In cells seeded without serum, theeffect was observed 24 h after incubation with VEGF. Inthis group VEGF appeared to act as a survival factor,allowing the cells to survive and proliferate underconditions of extreme stress. Whether receptor activa-tion mediated the survival and proliferative responsedirectly, or through an effect on other growth factors isnot clear. Studies have demonstrated upregulation ofknown proliferative growth factors in response to VEGF,such as heparin-binding epidermal growth factor-likegrowth factor (HB-EGF) and platelet-derived growthfactor BB (PDGF-BB) (1).

To investigate the potential role of an antiapoptoticor survival response in the proliferative action ofVEGF, the effect of VEGF on apoptosis was examined.Several reports describe the usefulness of annexin-Vbinding as a marker of apoptosis (4, 21). When com-bined with propidium iodide staining, cells can besubdivided into viable, early apoptotic and either lateapoptotic or necrotic, on the basis of their stainingcharacteristics on flow cytometry. VEGF had a small,protective effect on hydrogen peroxide-induced apopto-sis and necrosis, with fewer cells showing staining forannexin-V and propidium iodide when incubated withVEGF, compared with control. Although these resultsmay reflect a proliferative rather than a survival effectof VEGF, the short VEGF/peroxide incubation timeshould have minimized the degree of cell proliferation.In addition, the hydrogen peroxide concentrations weretitrated to obtain a significant degree of apoptosis andnecrosis, making cell proliferation under these condi-tions very difficult. Recent studies have shown a simi-lar response in vascular endothelial cells, using concen-trations of VEGF between 10 and 100 ng/ml (15, 16).This survival effect was shown to be regulated throughVEGFR-2 with stimulation of the phosphatidylinositol38-kinase-Akt signal pathway (16). In the present study,it is difficult to ascertain which of the VEGF receptorsmay have mediated the survival and proliferativeresponse observed in the serum-deprived cells. Bothreceptors have a high affinity for VEGF, although theaffinity for VEGFR-1 is ,40-fold higher than that forVEGFR-2 (Kd values 16 vs. 760 pM) (23, 39, 41). Theconcentration of VEGF required to induce the prolifera-tive response was low ($1 or $260 pM) and in keepingwith signaling through either of the two high-affinityreceptors. Evidence that binding to VEGFR-2 and notVEGFR-1 relies more heavily on heparin modulation(9, 28) would support a role for VEGFR-1 in the

proliferation assay, as there was no additional effectseen when cells were incubated with heparin. However,in endothelial cells and monocytes, VEGFR-1 appearsto be responsible for target cell migration (3, 44), withevidence supporting more complex roles for VEGFR-2in endothelial cells, such as mitogenicity, chemotaxis,actin reorganization, and changes in cell morphology(41).

Immunohistochemical staining demonstrated expres-sion of both receptors in rat renal tubules in vivo. Withtwo of the antibodies (C-17 and A-3), the distributionwas unusual in that each receptor appeared to localizeto specific areas of the nephron. VEGFR-2 staining wasprominent in DCT and collecting ducts, whereasVEGFR-1 staining was more diffuse, involving bothproximal and distal tubules, with more localized stain-ing seen on the brush border of proximal tubules. Thesefindings suggest VEGF may have a specific role in theseparts of the kidney, although the exact nature of thisremains unclear. Histological data presented in thisstudy differs from the distribution of VEGF receptors inthe kidney reported by another group (31, 32). In thesereports, in situ hybridization localized VEGFR-1 andVEGFR-2 mRNA exclusively to renal endothelial cells.Immunofluorescence and in situ [125I]-VEGF bindingwas used to demonstrate receptor protein expressionand this was also localized to endothelial cells. Apartfrom the fact that the techniques used differ, the reasonfor the discrepancy is not clear, although the reportsrefer to human kidney specimens only. A recent study,however, demonstrates evidence for VEGFR-1 on devel-oping renal tubular epithelial cells (38). There are noother studies reporting results in the kidney with theVEGF receptor antibodies used here. The antibodiesused in this study were directed against unique COOH-terminal, cytoplasmic portions of the receptors. Cross-reactivity with other tyrosine kinases has been excluded(Santa Cruz Biotechnology). Simon and co-workers (32)used antibodies raised against recombinant, soluble ex-tracellular portions of the receptors. This discrepancyfurther raises the possibility of the existence of differ-ent receptor isoforms.

In conclusion, this study reports the presence offunctional VEGF receptors on nonendothelial cells ofthe kidney, with VEGF exerting a survival effect on ratrenal tubular epithelial cells in vitro. VEGF maypromote renal tubular epithelial cell survival in vivo insituations associated with cellular stress, for exampleacute ischemia or toxic injury of the kidney. These datasuggest an expanded role for VEGF in pathologicalconditions in the kidney.

The authors are grateful to Dr. Mark Lam, Australian NationalUniversity, Canberra, Australia, for instructing the authors in theuse of the 3-D laser scanning confocal fluorescence microscope.

This work was supported by grants from the Australian KidneyFoundation and the National Health and Medical Research Councilof Australia.

Address for reprint requests and other correspondence: J. Kanel-lis, Immunology Research Center, St. Vincent’s Hospital, 41 VictoriaPde., Fitzroy, Victoria 3065, Australia.

Received 18 June 1999; accepted in final form 24 January 2000.

F913VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

REFERENCES

1. Arkonac BM, Foster LC, Sibinga NE, Patterson C, Lai K,Tsai JC, Lee ME, Perrella MA, and Haber E. Vascularendothelial growth factor induces heparin-binding epidermalgrowth factor-like growth factor in vascular endothelial cells. JBiol Chem 273: 4400–4405, 1998.

2. Barleon B, Siemeister G, Martiny-Baron G, Weindel K,Herzog C, and Marme D. Vascular endothelial growth factorup-regulates its receptor fms-like tyrosine kinase 1 (Flt-1) and asoluble variant of Flt-1 in human vascular endothelial cells.Cancer Res 57: 5421–5425, 1997.

3. Barleon B, Sozzani S, Zhou D, Weich HA, Mantovani A, andMarme D. Migration of human monocytes in response to vascu-lar endothelial growth factor (VEGF) is mediated via the VEGFreceptor flt-1. Blood 87: 3336–3343, 1996.

4. Boersma AW, Nooter K, Oostrum RG, and Stoter G. Quanti-fication of apoptotic cells with fluorescein isothiocyanate-labeledannexin V in Chinese hamster ovary cell cultures treated withcisplatin. Cytometry 24: 123–130, 1996.

5. Brown LF, Detmar M, Tognazzi K, Abu-Jawdeh G, andIruela-Arispe ML. Uterine smooth muscle cells express func-tional receptors (flt-1 and KDR) for vascular permeability factor/vascular endothelial growth factor. Lab Invest 76: 245–255,1997.

6. Clauss M, Weich H, Breier G, Knies U, Rockl W, Walten-berger J, and Risau W. The vascular endothelial growth factorreceptor Flt-1 mediates biological activities. Implications for afunctional role of placenta growth factor in monocyte activationand chemotaxis. J Biol Chem 271: 17629–17634, 1996.

7. Das SK, Chakraborty I, Wang J, Dey SK, and Hoffman LH.Expression of vascular endothelial growth factor (VEGF) andVEGF-receptor messenger ribonucleic acids in the peri-implanta-tion rabbit uterus. Biol Reprod 56: 1390–1399, 1997.

8. De Larco J and Todaro J. Epithelioid and fibroblastic ratkidney cell clones: epidermal growth factor (EGF) receptors andthe effect of mouse sarcoma virus transformation. J Cell Physiol94: 335–342, 1978.

9. Dougher AM, Wasserstrom H, Torley L, Shridaran L, West-dock P, Hileman RE, Fromm JR, Anderberg R, Lyman S,Linhardt RJ, Kaplan J, and Terman BI. Identification of aheparin binding peptide on the extracellular domain of the KDRVEGF receptor. Growth Factors 14: 257–268, 1997.

10. Ergun S, Kilic N, Fiedler W, and Mukhopadhyay AK.Vascular endothelial growth factor and its receptors in normalhuman testicular tissue. Mol Cell Endocrinol 131: 9–20, 1997.

11. Feng Y, Venema VJ, Venema RC, Tsai N, and Caldwell RB.VEGF induces nuclear translocation of Flk-1/KDR, endothelialnitric oxide synthase, and caveolin-1 in vascular endothelialcells. Biochem Biophys Res Commun 256: 192–197, 1999.

12. Ferrara N, Houck KA, Jakeman LB, Winer J, and LeungDW. The vascular endothelial growth factor family of polypep-tides. J Cell Biochem 47: 211–218, 1991.

13. Folkman J. Angiogenesis in cancer, vascular, rheumatoid andother disease. Nat Med 1: 27–31, 1995.

14. Folkman J and Shing Y. Angiogenesis. J Biol Chem 267:10931–10934, 1992.

15. Gerber HP, Dixit V, and Ferrara N. Vascular endothelialgrowth factor induces expression of the antiapoptotic proteinsBcl-2 and A1 in vascular endothelial cells. J Biol Chem 273:13313–13316, 1998.

16. Gerber HP, McMurtrey A, Kowalski J, Yan M, Keyt BA,Dixit V, and Ferrara N. Vascular endothelial growth factorregulates endothelial cell survival through the phosphatidylino-sitol 38-Kinase/Akt signal transduction pathway. J Biol Chem273: 30336–30343, 1998.

17. Gitay-Goren H, Cohen T, Tessler S, Soker S, GengrinovitchS, Rockwell P, Klagsbrun M, Levi BZ, and Neufeld G.Selective binding of VEGF121 to one of the three vascularendothelial growth factor receptors of vascular endothelial cells.J Biol Chem 271: 5519–5523, 1996.

18. Gitay-Goren H, Soker S, Vlodavsky I, and Neufeld G. Thebinding of vascular endothelial growth factor to its receptors isdependent on cell surface-associated heparin-like molecules. JBiol Chem 267: 6093–6098, 1992.

19. Guerrin M, Moukadiri H, Chollet P, Moro F, Dutt K,Malecaze F, and Plouet J. Vasculotropin/vascular endothelialgrowth factor is an autocrine growth factor for human retinalpigment epithelial cells cultured in vitro. J Cell Physiol 164:385–394, 1995.

20. Leung DW, Cachianes G, Kuang WJ, Goeddel DV, andFerrara N. Vascular endothelial growth factor is a secretedangiogenic mitogen. Science 246: 1306–1309, 1989.

21. Martin SJ, Reutelingsperger CP, McGahon AJ, Rader JA,van Schie RC, LaFace DM, and Green DR. Early redistribu-tion of plasma membrane phosphatidylserine is a general featureof apoptosis regardless of the initiating stimulus: inhibition byoverexpression of Bcl-2 and Abl. J Exp Med 182: 1545–1556,1995.

22. Midy V and Plouet J. Vasculotropin/vascular endothelial growthfactor induces differentiation in cultured osteoblasts. BiochemBiophys Res Commun 199: 380–386, 1994.

23. Millauer B, Wizigmann-Voos S, Schnurch H, Martinez R,Moller NP, Risau W, and Ullrich A. High affinity VEGFbinding and developmental expression suggest Flk-1 as a majorregulator of vasculogenesis and angiogenesis. Cell 72: 835–846,1993.

24. Paavonen K, Horelli-Kuitunen N, Chilov D, Kukk E, Pen-nanen S, Kallioniemi OP, Pajusola K, Olofsson B, ErikssonU, Joukov V, Palotie A, and Alitalo K. Novel human vascularendothelial growth factor genes VEGF-B and VEGF-C localize tochromosomes 11q13 and 4q34, respectively. Circulation 93: 1079–1082, 1996.

25. Paizis K, Kirkand G, Polihronis M, Katerelos M, KanellisJ, and Power DA. Heparin-binding epidermal growth factor-like growth factor in experimental models of membranous andminimal change nephropathy. Kidney Int 53: 1162–1171, 1998.

26. Pajusola K, Aprelikova O, Pelicci G, Weich H, Claesson-Welsh L, and Alitalo K. Signalling properties of FLT4, aproteolytically processed receptor tyrosine kinase related to twoVEGF receptors. Oncogene 9: 3545–3555, 1994.

27. Quinn TP, Peters KG, De Vries C, Ferrara N, and WilliamsLT. Fetal liver kinase 1 is a receptor for vascular endothelialgrowth factor and is selectively expressed in vascular endothe-lium. Proc Natl Acad Sci USA 90: 7533–7537, 1993.

28. Roeckl W, Hecht D, Sztajer H, Waltenberger J, Yayon A,and Weich HA. Differential binding characteristics and cellularinhibition by soluble VEGF receptors 1 and 2. Exp Cell Res 241:161–170, 1998.

29. Rooman I, Schuit F, and Bouwens L. Effect of vascularendothelial growth factor on growth and differentiation of pancre-atic ductal epithelium. Lab Invest 76: 225–232, 1997.

30. Sandner P, Wolf K, Bergmaier U, Gess B, and Kurtz A.Induction of VEGF and VEGF receptor gene expression byhypoxia: divergent regulation in vivo and in vitro. Kidney Int 51:448–453, 1997.

31. Simon M, Grone HJ, Johren O, Kullmer J, Plate KH, RisauW, and Fuchs E. Expression of vascular endothelial growthfactor and its receptors in human renal ontogenesis and in adultkidney. Am J Physiol Renal Fluid Electrolyte Physiol 268:F240–F250, 1995.

32. Simon M, Rockl W, Hornig C, Grone EF, Theis H, Weich HA,Fuchs E, Yayon A, and Grone HJ. Receptors of vascularendothelial growth factor/vascular permeability factor (VEGF/VPF) in fetal and adult human kidney: localization and[125I]VEGF binding sites. J Am Soc Nephrol 9: 1032–1044, 1998.

33. Takahashi T, Shirasawa T, Miyake K, Yahagi Y, MaruyamaN, Kasahara N, Kawamura T, Matsumura O, Mitarai T, andSakai O. Protein tyrosine kinases expressed in glomeruli andcultured glomerular cells: Flt-1 and VEGF expression in renalmesangial cells. Biochem Biophys Res Commun 209: 218–226,1995.

34. Takemura T, Kondo S, Homma T, Sakai M, and Harris RC.The membrane-bound form of heparin-binding epidermal growthfactor-like growth factor promotes survival of cultured renalepithelial cells. J Biol Chem 272: 31036–31042, 1997.

35. Terman BI, Dougher-Vermazen M, Carrion ME, DimitrovD, Armellino DC, Gospodarowicz D, and Bohlen P. Identifi-cation of the KDR tyrosine kinase as a receptor for vascular

F914 VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from

endothelial cell growth factor. Biochem Biophys Res Commun187: 1579–1586, 1992.

36. Tessler S, Rockwell P, Hicklin D, Cohen T, Levi BZ, Witte L,Lemischka IR, and Neufeld G. Heparin modulates the interac-tion of VEGF165 with soluble and cell associated flk-1 receptors.J Biol Chem 269: 12456–12461, 1994.

37. Tso JY, Sun X, Koa T, Reece KS, and Wu R. Isolation andcharacterisation of rat and human glyceraldehyde-3-phosphatedehydrogenase cDNAs: molecular complexity and evolution ofthe gene. Nucleic Acids Res 13: 2485–2502, 1985.

38. Tufro A, Norwood VF, Carey RM, and Gomez RA. Vascularendothelial growth factor induces nephrogenesis and vasculogen-esis. J Am Soc Nephrol 10: 2125–2134, 1999.

39. Vaisman N, Gospodarowicz D, and Neufeld G. Characteriza-tion of the receptors for vascular endothelial growth factor. J BiolChem 265: 19461–19466, 1990.

40. Viglietto G, Romano A, Manzo G, Chiappetta G, Paoletti I,Califano D, Galati MG, Mauriello V, Bruni P, Lago CT,Fusco A, and Persico MG. Upregulation of the angiogenicfactors PlGF, VEGF and their receptors (Flt-1, Flk-1/KDR) by

TSH in cultured thyrocytes and in the thyroid gland of thiouracil-fed rats suggest a TSH-dependent paracrine mechanism forgoiter hypervascularization. Oncogene 15: 2687–2698, 1997.

41. Waltenberger J, Claesson-Welsh L, Siegbahn A, ShibuyaM, and Heldin CH. Different signal transduction properties ofKDR and Flt1, two receptors for vascular endothelial growthfactor. J Biol Chem 269: 26988–26995, 1994.

42. Wen Y, Edelman JL, Kang T, Zeng N, and Sachs G. Twofunctional forms of vascular endothelial growth factor receptor-2/Flk-1 mRNA are expressed in normal rat retina. J Biol Chem273: 2090–2097, 1998.

43. Yamane A, Seetharam L, Yamaguchi S, Gotoh N, TakahashiT, Neufeld G, and Shibuya M. A new communication systembetween hepatocytes and sinusoidal endothelial cells in liverthrough vascular endothelial growth factor and Flt tyrosinekinase receptor family (Flt-1 and KDR/Flk-1). Oncogene 9:2683–2690, 1994.

44. Yoshida A, Anand-Apte B, and Zetter BR. Differential endo-thelial migration and proliferation to basic fibroblast growthfactor and vascular endothelial growth factor. Growth Factors 13:57–64, 1996.

F915VEGF IS A SURVIVAL FACTOR FOR RENAL TUBULAR EPITHELIUM

by 10.220.33.3 on April 4, 2017

http://ajprenal.physiology.org/D

ownloaded from