VEGF-C promotes survival in podocytes

【摘要】  Vascular endothelial growth factor (VEGF)-A is an autocrine survival factor for podocytes, which express two VEGF receptors, VEGF-R1 and VEGF-R3. As VEGF-A is not a known ligand for VEGF-R3, the aim of this investigation was to examine whether VEGF-C, a known ligand for VEGF-R3, served a function in podocyte biology and whether this was VEGF-R3 dependent. VEGF-C protein expression was localized to podocytes in contrast to VEGF-D, which was expressed in parietal epithelial cells. Intracellular calcium ([Ca 2+ ] i ) experiments demonstrated that VEGF-C induced a 0.74 ± 0.09-fold reduction in [Ca 2+ ] i compared with baseline in human conditionally immortalized podocytes (hCIPs; P < 0.05, one sample t -test, n = 8). Cytotoxicity experiments revealed that in hCIPs VEGF-C reduced cytotoxicity to 81.4 ± 1.9% of serum-starved conditions ( P < 0.001, paired t -test, n = 16), similar to VEGF-A (82.8 ± 4.5% of serum-starved conditions, P < 0.05, paired t -test). MAZ51 (a VEGF-R3 kinase inhibitor) inhibited the VEGF-C-induced reduction in cytotoxicity (106.2 ± 2.1% of serum-starved conditions), whereas MAZ51 by itself had no cytotoxic effects on hCIPs. VEGF-C was also shown to induce a 0.5 ± 0.13-fold reduction in levels of MAPK phosphorylation compared with VEGF-A and VEGF-A-Mab treatment ( P < 0.05, ANOVA, n = 4), yet had no effect on Akt phosphorylation. Surprisingly, immunoprecipitation studies detected no VEGF-C-induced autophosphorylation of VEGF-R3 in hCIPs but did so in HMVECs. Moreover, SU-5416, a tyrosine kinase inhibitor, blocked the VEGF-C-induced reduction in cytotoxicity (106 ± 2.8% of serum-starved conditions) at concentrations specific for VEGF-R1. Together, these results suggest for the first time that VEGF-C acts in an autocrine manner in cultured podocytes to promote survival, although the receptor or receptor complex activated has yet to be elucidated.

【关键词】  intracellular calcium cytotoxicity lymphatic endothelial cells

VASCULAR ENDOTHELIAL GROWTH factor (VEGF-A), a potent angiogenic factor, is produced in large amounts by mature, healthy podocytes in vivo and in vitro within the glomerulus ( 3 ), which is not a site of overt angiogenesis. The discovery that neuropilin-1, a coreceptor for VEGF-R2, was also expressed by podocytes ( 14 ) led to investigations into the direct effects of VEGF-A on podocytes. These investigations revealed that exogenous and endogenous VEGF-A had an autocrine effect on podocytes resulting in increased survival ( 10 ) and that both human conditionally immortalized podocytes (hCIPs) and primary human culture podocytes (PCPs) expressed VEGF-R1 and VEGF-R3 protein and mRNA. VEGF-A is not thought to be a direct ligand for VEGF-R3; therefore, it was assumed that VEGF-A was signaling through VEGF-R1 in podocytes. It remained to be seen whether VEGF-R3 had a functional role in podocyte biology.

VEGF-R3 is expressed on various tissues, such as the spleen, brain, and epithelial cells of fetal lung ( 32 ), although it is predominantly expressed on lymphatic endothelial cells in the adult (as reviewed in Ref. 21 ). VEGF-R3 is a class III tyrosine kinase receptor of the same family as VEGF-R1 and VEGF-R2 ( 40 ). It consists of seven Ig-like loops, a transmembrane domain, and two tyrosine kinase domains on its cytoplasmic tail ( 16 ). Expression of VEGF-R3 by podocytes ( 10 ) suggests the possibility of local expression of one or more of its ligands, VEGF C and VEGF D. VEGF-C has many similar functions to VEGF, acting via VEGF-R2. It can stimulate proliferation and migration of endothelial cells and induce an increase in vascular permeability, but with a potency four to five times less than VEGF ( 18, 20 ). VEGF-C also has angiogenic properties in in vivo models ( 5, 42 ), but it is a more potent inducer of lymphangiogenesis ( 17, 35 ). Actions of VEGF-C on lymphatic endothelium occur primarily via VEGF-R3 ( 25, 28, 35, 44 ). Low-level generalized expression of VEGF-C is seen in early gestation with expression becoming prominent in perinephric, mesenteric, and cephalic regions during late embryonic development ( 25 ). It is only weakly expressed in adult tissues including lymph nodes, heart, placenta, and intestine ( 19 ).

VEGF-D has similarities in structure and function with VEGF-C and together they form a VEGF subfamily. They both have NH 2 - and COOH-terminal extensions and are proteolytically processed, which accompanies secretion and results in preferential, high-affinity binding to VEGFR-3 ( 20, 41 ). VEGF-D is also a mitogen for endothelial cells ( 1 ) and stimulates both angiogenesis and lympangiogenesis primarily via VEGF-R2 and VEGF-R3, respectively ( 35 ). As with VEGF-C, highest levels of expression are found during fetal development but it is downregulated as development progresses. In the adult, it is expressed primarily in lung ( 8 ), heart, small intestine, and neuroendocrine cells ( 33 ).

VEGF-C mRNA has been detected by RNase protection assay in adult human kidney ( 13 ) and by Northern blotting in rat ( 24 ) and mouse ( 9, 25 ) kidney. Relatively high levels of VEGF-D expression were detected in rat kidney by Northern blotting ( 24 ) and in situ hybridization also revealed renal expression of VEGF-D mRNA in young adult mice ( 8 ). However, no previous studies have sought to define the precise localization of these mediators within the kidney.

VEGF-C induces homodimerization, or heterodimerization of VEGF-R3 with VEGF-R2 ( 7 ), but there is no evidence of heterodimerization with VEGF-R1. VEGF-R2 is not expressed in cultured human podocytes as determined by PCR ( 10 ), focused gene array, and Western blotting ( 38 ), so any downstream effect of VEGF-C might be assumed to be a result of VEGF-R3 activation. VEGF-C induces transphosphorylation of VEGF-R3, which leads to various downstream signaling pathways, similar to that of VEGF-R2 by VEGF-A in vascular endothelial cells including migration and survival of lymphatic endothelial cells ( 28 ).

In this study, we sought to identify potential sources of the ligands for VEGF-R3 within the glomerulus and to decipher the functional effects of VEGF-R3 intracellular signaling on cultured podocytes using exogenous VEGF-C, to assess whether VEGF-R3 signaling promoted survival in a similar manner to VEGF-R2, via VEGF-A. We began by determining the effects of VEGF-C on intracellular calcium concentration ([Ca 2+ ] i ), because this is a common second messenger involved in many cell signaling pathways and it was the first indicator that VEGF-A induced a signaling effect in podocytes ( 10 ). The effects of VEGF-C on cytotoxicity were also determined to assess whether VEGF-C is involved in survival signaling in podocytes, including the use of VEGF-R3 and VEGF-R1 inhibitors, then the effects of VEGF-C on VEGF-R3 phosphorylation were further investigated. In blood vascular and lymphatic endothelial cells, respectively, VEGF-A and VEGF-C survival pathways are very similar but follow different time courses. For example in human microvascular endothelial cells (HMVECs), Akt phosphorylation (implicated in survival signaling) induced by VEGF-A peaked between 20 and 30 min after initial exposure, whereas Akt phosphorylation induced by VEGF-C peaked after only 10 min ( 28 ). VEGF-C was also shown to induce a more prolonged upregulation of phosphorylated MAPK (also implicated in survival signaling) ( 28 ). The effects of VEGF-C on Akt and MAPK phosphorylation were therefore investigated in this paper to determine the survival signaling pathway activated by VEGF-C.

METHODS

Nephrectomy tissue. Human renal tissue was obtained, with consent, from patients undergoing nephrectomy for renal carcinoma. Within 30 min of the nephrectomy, samples of cortex from the normal pole were taken for preparation of frozen and paraffin sections.

Cell culture. hCIPs have been previously described in detail ( 37, 38 ). Briefly, the hCIPs contain a transgene encoding a temperature-sensitive mutant of SV40 large T antigen, which is active at 33°C but inactive at 37°C. In addition, the cell line has been stably transfected with the catalytic domain of the human telomerase gene, hTERT ( 31 ). Cells proliferate at 33°C but when transferred to 37°C they stop dividing and develop a more differentiated phenotype. hCIPs were studied at both 33°C and after 14 days at 37°C (when fully differentiated). All podocytes were cultured in RPMI 1640 medium with insulin, transferrin, selenite (all Sigma, Dorset, UK), and 10% fetal calf serum. HMVEC (derived from neonatal pooled dermal tissue; Clonetics, San Diego, CA) were obtained at passage 3 and used for experiments up to passage 7. HMVEC were cultured in endothelial growth medium 2-microvascular (EGM2-MV; Cambrex, Wokingham, UK), made from endothelial basal medium 2 (EBM2; Cambrex) and fetal calf serum (5%), antimicrobial agents and growth factors as supplied. Cells being prepared for, or being used in, experiments were cultured in EGM2-MV without VEGF-A.

Experimental primary antibodies, secondary antibodies, recombinant proteins, and small molecule inhibitors. Antibodies used for immunofluorescence consisted of 10 µg/ml VEGF-C goat polyclonal antibody (N-19, Santa Cruz Biotechnology, Santa Cruz, CA) and 20 µg/ml VEGF-D monoclonal antibody (MAB 286, R&D Systems). Rabbit anti-goat (61-1611) and goat anti-mouse (62-6611) secondary antibodies, conjugated with fluoroscein isothiocyanate (FITC), were both used at 15 µg/ml (Zymed Laboratories, San Francisco, CA). Antibodies for immunohistochemistry consisted of 1.14 µg/ml VEGF-C goat anti-human polyclonal IgG primary antibody (Santa Cruz Biotechnology, SC-7133) in conjuction with 2 µg/ml horse anti-goat biotinylated secondary antibody. Antibodies used for Western blotting consisted of 0.4 µg/ml rabbit polyclonal anti-human VEGF-C (H190) primary antibody (Santa Cruz Biotechnology) used in conjunction with 0.0014 µg/ml horseradish peroxidase (HRP)-conjugated anti-rabbit secondary antibody (Pierce, Cheshire, UK), 1 µg/ml mouse monoclonal anti-phospho-Akt (p-ser 472/473) primary antibody (BD Bioscience, Oxford, UK) used in conjunction with 0.01 µg/ml HRP-conjugated anti-mouse secondary antibody (Pierce), 1 µg/ml mouse monoclonal anti-Akt antibody used in conjunction with 0.01 µg/ml HRP-conjugated anti-mouse secondary antibody (Pierce), 0.4 µg/ml rabbit polyclonal IgG anti-p44/42 MAPK primary antibody (Cell signaling, Beverley, MA) used in conjunction with 0.005 µg/ml HRP-conjugated goat anti-rabbit IgG secondary antibody (Pierce), and 0.4 µg/ml mouse monoclonal anti-phospho p44/42 MAPK (T202/Y204) (Cell Signaling) used in conjunction with 0.005 µg/ml HRP-conjugated goat anti-mouse IgG secondary antibody (Pierce). For the immunoprecipitation assays, 3 µg per 1,500 µg total protein of rabbit polyclonal IgG human anti-VEGF-R3 antibody (Zymed) were used. The same antibody was used for probing the membranes containing immunoprecipitate of VEGF-C-treated cells at a concentration of 1 µg/ml in conjunction with 0.01 µg/ml HRP-conjugated goat anti-rabbit IgG secondary antibody (Pierce). Membranes were also probed with 0.1 µg mouse monoclonal IgG 2b anti-phospho-tyrosine primary antibody (p-Y99, sc-7020, Santa Cruz Biotechnology) in conjunction with 0.0833 ng/ml HRP-conjugated anti-mouse IgG secondary antibody (Pierce) a protein A/G slurry (Santa Cruz Biotechnology). For the intracellular calcium experiments, cytotoxicity assays, Western blots, and immunoprecipitations cells were treated with 1 nM human recombinant VEGF-A protein (a kind gift from N. Ferrara, Genentec), 1 nM human recombinant VEGF-C protein (a kind gift from K. Alitalo), 0.32 µg/ml neutralizing mouse monoclonal IgG 2A antibody to VEGF-A (R&D Systems), 10 µM MAZ51 (Calbiochem), 100 nM SU-5416 (Calbiochem), or left untreated.

Immunofluorescence. Adjacent 4-µm cryosections of normal renal cortex were air-dried for 15 min and incubated consecutively with primary and secondary antibodies (in blocking solution; 5% fetal calf serum and 0.05% Tween 20 in PBS) before mounting in Vectashield aqueous mountant (Vector Laboratories). Secondary antibodies were applied consecutively. In control sections, the primary antibody was replaced with nonimmune immunoglobulin of the same class, species, and concentration. Sections were examined using a Leitz DMRB fluorescence microscope (Leica, Solms, Germany) and images were captured using a SPOT Slider 2 digital camera (Diagnostic Instruments, Optivision, West Yorkshire) and SPOT computer software (Diagnostic Instruments). Where resulting images are presented, original magnifications are indicated, and where images are compared, identical microscope and camera settings were used.

Cells grown to confluence on 10% collagen/PBS (collagen A, Autogen Bioclear, Calne, Wilts, UK)-coated glass coverslips were fixed in 2% formaldehyde and permeabilized in 0.3% Triton X-100. Cells were incubated with blocking solution and then with primary and secondary antibodies before mounting and viewing as above.

Immunohistochemistry. Sections were dewaxed by immersion in Histoclear (RA Lamb) for 5 min, then rehydrated in reduced quantities of ethanol (99, 90, and 70%) for a total of 2 min each. Slides were washed twice in distilled water for 5 min then twice in PBS for 5 min. Antigen retreival was performed by microwaving sections at 800 W for 8-10 min using 0.1 M Tris/EDTA buffer at pH 9. Samples were allowed to cool, then washed twice in distilled water for 5 min each time. Sections were incubated with freshly prepared 3% hydrogen peroxide for 5 min, which was removed by two 5-min washes in PBS. Tissue sections were blocked in 1.5% normal horse serum for 20-30 min in a humid chamber at room temperature. The sections were then incubated with the primary antibody at 4°C overnight. Adjacent tissue sections cut from the same block as those treated with the antibody were used as a nonimmune IgG negative control, from the same origin as the primary antibody. The following day tissue sections were incubated in blocking solution as before and then incubated with a biotinylated secondary antibody (Vector Laboratories) and left to incubate for 30 min at room temperature. A preprepared avidin-biotinylated enzyme complex kit (Elite, Vector Laboratories) was applied to the sections after two 5-min PBS-T washes. The procedure was completed by performing further two 5-min PBS-T washes and treating the sections with 3,3'-diaminobenzidine substrate (Vector Laboratories) to yield a brown colored product. Slides were counterstained with Mayer's hematoxylin dehydrated by passing through increasing concentrations of ethanol, cleared in xylene for 10 min, and permanently mounted in DPX mountant for histology. Slides were examined using a Nikon E-400 microscope (Nikon Instruments).

VEGF-C and [Ca 2+ ] i measurements in hCIPs. hCIPs were grown on coverslips to confluence, were incubated with Fura 2-AM (10 µM) for 90 min in DMEM at room temperature, and the coverslip was then placed in a coverslip holder. The holder was mounted on a rig consisting of an inverted fluorescence microscope (DM IRB, Leica) equipped with a UV source (Cairn Instruments, World Precision Instruments) with filters for excitation at 340 and 380 nm. Fast switching was achieved using a rotary filter wheel at 50 Hz and a spectrophotometer for photometric measurement (Cairn Instruments). The spectrophotometer received emitted light via a 400-nm dichroic filter and a 510- to 530-nm barrier filter in front of the photometer. Powerlab software was used for analysis and graphic display.

Experiments were conducted in HBSS media containing minimal [Ca 2+ ] o (GIBCO BRL), therefore there was little difference between [Ca 2+ ] i and [Ca 2+ ] o. Tests samples of 1 nM human recombinant VEGF-C protein and HBSS, used as a negative control, were left to wash and record for 5 min. To ensure that [Ca 2+ ] i was effectively measured, 10 µM ionomycin was added to stimulate Ca 2+ entry into the cells. One millimolar manganese chloride (MnCl 2 ), in the continued presence of 10 µM ionomycin, was then used to quench the calcium-sensitive Fura to determine the background (Ca 2+ independent) fluorescence signal.

Emission fluorescent measurements ( I f ) were taken four times a second. The ratio of the I f measured during 340-nm excitation to that during 380-nm excitation (R), proportional to the calcium concentration, was calculated from

where R exp = ( I f340 - B 340 )/( I f380 - B 380 ). I f340 is the I f measured during excitation at 340 nm, I f380 is the I f measured during excitation at 380 nm, and B 340 and B 380 are the background I f values measured during excitations at 340 and 380 nm, respectively (measured as the I f after Mn 2+ quenching). R min is the in vitro ratio for zero [Ca 2+ ] i concentration. The use of this ratio facilitates the detection of small changes in [Ca 2+ ] i, which are independent of the Fura loading.

Cytotoxicity assays. hCIPs were grown to confluency on a 96-well plate (Costar) and then serum starved. After 24 h, 100 µl media were removed from each well and cytotoxicity was assayed using an LDH cytotoxicity detection kit (Roche) and quantified using a Bichrometric Multiscan plate reader (Labsystems). These samples were used for background LDH measurement (T x min). The media were replaced with 100 µl of FBS-free media containing either 1 nM VEGF-C, 10 µM MAZ51 (a cell-permeable ATP-competitive VEGF-A receptor tyrosine kinase inhibitor with greatest specificity for VEGF-R3), 1 nM VEGF-C with MAZ51 100 nM SU-5416, a pan VEGF-R inhibitor, which at this concentration is specific for only VEGF-R1 ( 43 ), SU-5416 and VEGF-C or just FBS free media. Twenty-four hours later, 100 µl of media were again removed from each well and cytotoxicity was assayed and quantified. These samples were used to determine the cytotoxicity (T x exp). Finally, 100 µl of 2% Triton X-100/1 x PBS (final concentration 1%) were added to each well and left for 10 min to completely lyse the cells. One-hundred microliters were removed and cytotoxicity again was assayed and quantified. This enabled determination of the maximum LDH from the well (T x max); % cytotoxicity (T x ) was calculated for each well as:

Min values are divided by two because the control media still left in the 96-well plate were diluted twice by day 2. Exp values are divided by two because the exp media in the plate were diluted twice by day 3. By day 3, the low control was diluted twice more and therefore was a quarter of the original reading. The high and exp readings were combined to allow for differences between treatments in the high control readings. Data were expressed as changes in percent cytotoxicity and significance was tested within experiments using a paired t -test and between experiments using an unpaired t -test (2 samples) or unpaired ANOVA (more than 2 samples).

Western blotting. Serum-starved hCIPs were treated with 1 nM VEGF-A for 20 min, 1 nM VEGF-C for 10 min, or left untreated for the appropriate amounts of time. All samples were lysed in Triton X-100 lysis buffer (containing 1 x PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM PMSF, 1 mM Na 3 VO 4, and 20 µg/ml aprotinin) on ice. The samples were cleared by centrifugation at 13,000 rpm for 15 min at 4°C and the pellet was discarded. Total protein was then quantified using the Bradford Bio-Rad assay (Bio-Rad) according to manufacturer's instructions. Protein samples were separated by SDS-PAGE under reducing conditions and were transferred to PVDF membranes. The membranes were blocked in 5% fat-free milk (or 3% BSA/PBS-T for the anti-p-Y antibody) before incubation with antibodies described above. After incubation with HRP-conjugated secondary antibodies (Santa Cruz Biotechnology), bands were detected by using the ECL chemiluminescence system (Amersham Biotech, Bucks, UK) or the SuperSignal West Femto Maximum Sensitivity Substrate (Pierce).

Immunoprecipitation. To determine the effect of exogenous VEGF-C on VEGF-R3 phosphorylation in hCIPs and HMVECs, cells were serum starved for 16 h and then treated with 1 nM VEGF-C for either 10 or for 2, 7, 20 min for the time response or left in serum-free media. Cells were then lysed, cleared, and protein was quantified as described above. Lysate from each sample was incubated with 3 µg anti-VEGF-R3 antibody per 1,500 µg total protein overnight at 4°C with gentle agitation. Ten microliters washed A/G agarose beads per 100-µl sample were then added to each of the samples and left to incubate for at least 2 h, rocking at 4°C. Samples were then spun for 30 s at 13,000 rpm at 4°C, and the supernatant was removed and retained and the precipitate was washed three times with lysis buffer. Equal quantities of supernatant or the entire precipitate of each protein sample lysate were mixed with 3 x SDS loading buffer (100 mM Tris, pH 6.8, 4% SDS, 20% glycerol, 10% -mecaptoethanol, and 0.2% bromophenal blue) and boiled for 5 min. The precipitate samples were vortexed, then centrifuged for a further 30 s at 13,000 rpm. All samples were run on a 7.5% SDS-PAGE gel at 90 V and were then transferred to a polyvynilidinefluoride (PVDF) membrane over a period of 90 min at 60 V. The membrane was probed with either an anti-VEGF-R3 antibody or an anti-p Y antibody.

RESULTS

VEGF-C and VEGF-D proteins are expressed within the glomerulus. To investigate potential ligands for VEGF-R3 within the glomerulus, VEGF-C and VEGF-D expression was examined using immunofluorescence and immunohistochemistry on human renal cortex tissue sections. Figure 1 A demonstrates positive staining for VEGF-C within the glomerulus on cells on the periphery of the glomerular tuft, indicative of podocytes, and Fig. 1 C demonstrates positive staining for VEGF-D by the parietal epithelial cells, seen in a proportion of glomeruli. VEGF-D was also detected in a proportion of renal cortical vessels ( Fig. 1 E ). These results demonstrate that the two ligands for VEGF-R3, VEGF-C and VEGF-D, are differentially expressed in the glomerulus.

Fig. 1. Glomerular vascular endothelial growth factor (VEGF)-C and VEGF-D protein expression. Immunofluorescence (IF) and immunohistochemistry (IHC) on human renal cortex tissue sections demonstrating expression of VEGF-D and VEGF-C. A : VEGF-C staining in podocytes (arrows), situated on the periphery of the glomerular tuft. B : cells incubated with normal IgG. BS, Bowman?s space; Cap, capillaries; PEC, parietal epithelial cells. Scale bar = 20 µM. C : VEGF-D staining localized to parietal epithelial cell layer of Bowman's capsule (arrows; x 100). D : control IF staining with nonspecific primary antibody ( x 100). E : longitudinal section of renal cortical vessel showing positive VEGF-D expression (green; x 400).

VEGF-C protein expression in hCIPs. VEGF-C expression was examined in hCIPs by Western blotting and immunofluorescence to confirm podocyte expression of VEGF-C. A Western blot probed with an antibody to VEGF-C ( Fig. 2 A ) revealed bands corresponding to different forms of VEGF-C (prepro-protein, proprotein, dimers and COOH-terminal fragments). Immunofluorescence on hCIPs using an antibody against VEGF-C demonstrated cytoplasmic staining ( Fig. 2 B ). These results show that hCIPs express VEGF-C, which suggests an autocrine role of VEGF-C on VEGF-R3 signaling in podocytes.

Fig. 2. VEGF-C protein expression by human conditionally immortalized podocytes (hCIPs). A : bands corresponding to VEGF-C protein (pre-pro-protein (61 kDa), pro-proteins (58 kDa), secreted form (29 kDa) in lysate from untreated hCIPs. B : IF in hCIPs stained for VEGF-C and nuclei counterstained with propidium iodide. C : control IF staining with nonspecific primary antibody.

VEGF-C reduces [Ca 2+ ] i in cultured podocytes. To investigate whether VEGF-C could signal in hCIPs, and to investigate whether it could exert a similar response in hCIPs as previously shown for VEGF-A ( 10 ), the change in fluorescence ratio (R norm, proportional to changes in [Ca 2+ ] i ) in hCIPs was measured in response to VEGF-C. hCIPs were incubated in HBSS containing minimal [Ca 2+ ] o and were treated with 1 nM VEGF-C as described previously ( 10 ). Changes in R norm were measured and compared with previously published effects of VEGF-A on the R norm in hCIPs. Interestingly, VEGF-C reduced the fluorescence ratio in hCIPs compared with baseline recordings ( Fig. 3 B ), as did VEGF-A ( Fig. 3 A ). The change in ratio in response to VEGF-C was immediate and rapid and was an approximately twofold greater reduction than baseline. After 2 min, the response leveled off and remained constant until the end of the experiment. VEGF-C significantly reduced R norm by 0.74 ± 0.09-fold of baseline ( P < 0.05, 1-sample t -test, n = 5; Fig. 3 C ), and similarly, VEGF-A induced a significant reduction by 0.7 ± 0.1-fold of baseline ( P < 0.05, 1-sample t -test; Fig. 3 C ). The function of reduced [Ca 2+ ] i in hCIPs is unknown, yet these results clearly show that VEGF-C can stimulate an effect in cultured podocytes. The reduction in [Ca 2+ ] i in response to both VEGF-C and VEGF-A suggests that the function of VEGF-C may be similar. One effect of VEGF-A on hCIPs was to reduce cytotoxicity induced by serum starvation ( 10 ); therefore, the effects of 1 nM VEGF-C on cytotoxicity induced by serum starvation were also investigated in hCIPs.

Fig. 3. Effects of VEGF-A and VEGF-C on the change in [Ca 2+ ] i in hCIPs incubated in HBSS containing [Ca 2+ ] o. hCIPs were serum starved and loaded with fura-2-AM. VEGF-C was then added to the cells, in the presence of HBSS containing minimal [Ca 2+ ] o. Fluorescent intensity changes were recorded and used to calculate the R norm, which represents changes in [Ca 2+ ] i. Changes in the R norm in hCIPs treated with VEGF-C were compared to historical effects of VEGF-A on changes in R norm in hCIPs. A : example of the effect of VEGF-A on hCIPs, showing a reduction in R norm. B : example of the effect of VEGF-C on hCIPs, showing a reduction in R norm. C : data expressed as means ± SE ratio of treatment/baseline R norm. The R norm was reduced in hCIPs when treated with both 1 nM VEGF and 1 nM VEGF-C. One-sample t -test compared with baseline was used. * P < 0.05, n = 5 and 8, respectively.

VEGF-C reduces cytotoxicity in cultured podocytes in a VEGF-R3 kinase-activated-dependent manner. In addition to the effects of 1 nM VEGF-C on cytotoxicity in hCIPs, the effects of VEGF-C on cytotoxicity induced by serum starvation on HEK293 cells, which do not express any known receptors for the VEGF family of proteins (VEGF-R1, VEGF-R2, or VEGF-R3), were determined to ensure that the effects of VEGF-C on hCIPs were cell specific. MAZ51, a small molecule VEGF-R3 inhibitor, was used to confirm the involvement of VEGF-R3 in VEGF-C signaling. hCIPs and HEK293 cells were serum starved for 16 h, then the media were replaced with fresh serum-free media with 1 nM VEGF-C, 10 µM MAZ51 alone, or in combination with 1 nM VEGF-C or left untreated. The amount of LDH released from the cells into the media was quantified for the low, experimental, and high controls, using a colorimetric assay, as described above. The percentage cytotoxicity was calculated from the densitometry readings of the LDH release. VEGF-C reduced cytotoxicity induced by serum starvation in hCIPs, to a similar extent as that of VEGF-A (VEGF-C; 70.8 ± 3.8% of serum-starved cells compared with VEGF-A; 77.3 ± 4% of serum-starved cells, P < 0.05; Fig. 4 A ). The reduction in cytotoxicity induced by VEGF-C was completely inhibited when preincubated, then coincubated with MAZ51 (106 ± 2.1% cytotoxicity of cells that had been serum starved, ANOVA), whereas MAZ51 by itself was not cytotoxic (88.5 ± 1.9% cytotoxicity of cells that had been serum starved, paired t -test). In contrast, VEGF-C had no significant effect on cytotoxicity in HEK293 cells, which had been serum starved ( Fig. 4 B ). HEK293 cells do not express any VEGF-A receptors, and these results demonstrate that the reduction in cytotoxicity in hCIPs induced by VEGF-C was cell specific. Together, these results suggest that VEGF-C promotes survival in hCIPs in a VEGF-R3-dependent manner. VEGF-C and VEGF-A regulate similar survival pathways in endothelial cells (lymphatic and vascular, respectively), and it was previously shown that VEGF-A also promotes survival in hCIPs ( 11 ). This, however, appears to be through unconventional means because VEGF-A was shown to actually reduce Akt phosphorylation, a component of the PI3-kinase survival pathway, instead of inducing Akt phosphorylation. It was therefore investigated whether VEGF-C followed similar survival signaling cascades in hCIPs to lymphatic endothelial cells (whereby VEGF-C induces Akt phosphorylation) or whether it follows the unconventional route taken by VEGF-A in hCIPs.

Fig. 4. Effects of 1 nM VEGF-C on cytotoxicity in hCIPs (compared with 1 nM VEGF-A) are blocked by MAZ51 and are specific to hCIPs. hCIPs and HEK293 cells were serum starved, then treated with 1 nM VEGF-C, or left untreated for 24 h. hCIPs were also treated with 10 µM MAZ51 in the presence/absence of VEGF-C. LDH release from the cells was quantified and the percentage of cytotoxicity was then calculated. Data were expressed as means ± SE the ratio of the percentage of cytotoxicity (treatment/serum starvation). A : effects of VEGF-C on the percentage of cytotoxicity were compared to previously published effects of VEGF on the percentage of cytotoxicity in hCIPs. Both VEGF-C and VEGF-A significantly reduced the percentage of cytotoxicity (treatment/serum starvation), induced by serum starvation (paired t -test). MAZ51 treatment alone did not significantly reduce cytotoxicity compared with that of serum starvation. Preincubation and cotreatment of MAZ51 with VEGF-C significantly inhibited the effect of VEGF-C on cytotoxicity compared with that of VEGF-C alone (ANOVA). B : there was no significant difference in the effects of VEGF-C compared with serum starvation in HEK293 cells (paired t -test). * P < 0.05, ** P < 0.01, *** P < 0.001, n = 16.

VEGF-C decreases MAP kinase phosphorylation in cultured podocytes. To investigate potential survival pathways induced by VEGF-C in hCIPs and to compare these with VEGF-A signaling in podocytes, the effects of VEGF-C on Akt phosphorylation in hCIPs were examined. hCIPs were serum starved and treated with 1 nM VEGF-C for 10 min and then protein was extracted. The cell lysate was subjected to SDS-PAGE and transferred to a PVDF membrane, which was probed with an anti-phospho-Akt antibody. It was then stripped and reprobed with an anti-Akt antibody and the intensity of the bands corresponding to phospho-Akt was normalized to the intensity of the bands corresponding to Akt. There was no significant effect on Akt phosphorylation, induced by serum starvation, in cells that were treated with VEGF-C (0.87 ± 0.12-fold lower than that in cells that had been serum starved; Fig. 5 B ). In contrast, previously published results ( 11 ) demonstrate that VEGF-A significantly reduced Akt phosphorylation in hCIPs (in VEGF-A-treated cells Akt phosphorylation was 0.37 ± 0.08-fold lower, P < 0.05; Fig. 5 B ). These results indicate that VEGF-C does not activate the Akt survival pathway in hCIPs. VEGF-C survival signaling has also been associated with elevated levels of phosphorylated MAPK in HMVECs ( 28 ). Hence, the effect of both VEGF-C and VEGF-A (and inhibition of VEGF-A with a neutralizing monoclonal antibody to VEGF-A) on MAPK phosphorylation in hCIPs was investigated. hCIPs were serum starved, then treated with VEGF-A for 20 min, VEGF-C for 10 min, a neutralizing monoclonal antibody to VEGF-A for 24 h or left untreated for the appropriate amount of time. Protein was extracted and subjected to SDS-PAGE, then transferred to a PVDF membrane and probed with an anti-phospho-MAPK antibody, then stripped and reprobed with an anti-MAPK antibody. The intensity of the band corresponding to phospho-MAPK was normalized to the intensity of the band corresponding to MAPK. Surprisingly, VEGF-C induced a significant reduction in phosphorylated MAPK levels (0.5 ± 0.13-fold lower in than serum-starved phospho-MAPK, P < 0.05, ANOVA; Fig. 6 B ). This was compared with VEGF-A-treated cells (1.25 ± 0.12-fold greater than serum-starved phospho-MAPK) and cells treated with a neutralizing monoclonal antibody to VEGF-A (0.94 ± 0.17-fold lower than serum-starved phospho-MAPK). The increase in phosphorylated MAPK levels in cells treated with VEGF-A compared with that of serum-starved cells was not significant (1-sample t -test, n = 4, not shown). It was interesting to note that neither VEGF-A nor the neutralizing monoclonal antibody to VEGF-A significantly increased phospho-MAPK in cells that were serum starved ( Fig. 6 B ). These results suggest, therefore, that VEGF-A and VEGF-C activate different signaling pathways in hCIPs to promote survival.

Fig. 5. Effects of 1 nM VEGF-C on Akt phosphorylation compared with VEGF-A in hCIPs. hCIPs were serum starved and treated with VEGF-A for 20 min, VEGF-C for 10 min, or the cells were left untreated for the appropriate amount of time. PVDF membranes containing this protein were probed with an anti-phospho-Akt antibody then stripped and reprobed with an anti-Akt antibody, examples are shown in A. B : data were expressed as means ± SE ratios: phospho-Akt (treatment/serum starvation)/total Akt (treatment/serum starvation). Akt phosphorylation induced by 1 nM VEGF-C was not significantly different from serum-starved levels, but it was significantly greater than in 1 nM VEGF-A-treated hCIPs (unpaired t -test). SS, serum-starved cells. * P < 0.05, n = 3 and 5, respectively.

Fig. 6. Effects of exogenous and endogenous VEGF-A and exogenous VEGF-C on MAPK phosphorylation in hCIPs. hCIPs were serum starved and incubated in VEGF-A, VEGF-Mab, VEGF-C, or left untreated for the appropriate amount of time. PVDF membranes containing this protein were probed with an anti-phospho-MAPK antibody, and then stripped and reprobed with an anti-MAPK antibody, examples of which are shown in A. B : data were expressed as means ± SE ratios: phospho-MAPK (treatment/control)/total MAPK (treatment/control); 1 nM VEGF-C significantly reduced p-MAPK compared with VEGF-A treatment in hCIPs. * P < 0.05, ANOVA, Bonferroni?s post hoc test n = 4.

VEGF-C signals independently of VEGF-R3 phosphorylation. In HMVECs, VEGF-C mediates its effects through VEGF-R2 and VEGF-R3 and it has been suggested that VEGF-C cannot activate VEGF-R3 in the absence of VEGF-R2 ( 2 ). Podocytes do not express VEGF-R2, therefore to confirm that VEGF-C signaling in hCIPs was dependent on VEGF-R3 phosphorylation, VEGF-R3 immunoprecipitations were carried out on VEGF-C-treated hCIPs. hCIPs were serum starved and treated with VEGF-C or left untreated for 10 min. The protein was extracted and immunoprecipitated using an anti-VEGF-R3 antibody and the precipitate and supernatant were subjected to SDS-PAGE, transferred to a PVDF membrane which was then probed with an anti-VEGF-R3 antibody, stripped then reprobed with an anti-p-Y antibody. Figure 7 gives an example of two hCIP experiments and one HMVEC experiment. VEGF-C is known to induce the tyrosine phosphorylation of VEGF-R3 in HMVECs, which express VEGF-R2 and VEGF-R3, therefore these cells were used to demonstrate the effectiveness of the assay. A band at 150 kDa, corresponding to VEGF-R3 was seen in each of the immunoprecipitations of hCIPs and HMVECs both treated with VEGF-C or left untreated ( Fig. 7 A ), demonstrating that the VEGF-R3 immunoprecipitation was successful. Despite stringent controls taken to load equal quantities of total protein (see METHODS ), there are obvious differences in intensity of the bands at 150 kDa, corresponding to VEGF-R3. This may be a consequence of lower overall expression of VEGF-R3 in some cell populations, and in the case of 20-min treatment with VEGF-C, may be due to changes in protein expression. However, when this membrane was stripped and reprobed with an anti-p-Y antibody, a band corresponding to 150 kDa was only seen in the immunoprecipitate of HMVECs treated with VEGF-C. Although demonstrating that the antibody was specific, this also suggests that VEGF-C did not induce VEGF-R3 phosphorylation in podocytes. To test whether this was because VEGF-C promotes survival signaling in podocytes on a different time course, VEGF-R3 immunoprecipitations were carried out on hCIPs treated with VEGF-C for 2, 7, or 20 min as shown in Fig. 8. Distinct bands were seen when probed with anti-VEGF-R3 at 150 kDa, yet no bands were seen in any of the different time courses of VEGF-C treatment that correspond to VEGF-R3. Interestingly, it appears that a molecule of 230 kDa was also immunoprecipitated in cells treated for 2 or 20 min. The only other receptor known to be expressed on podocytes, which may be involved in VEGF-C signaling is VEGF-R1, although it has not been shown to be activated directly by VEGF-C. SU-5416, a small molecule ATP inhibitor, was used to investigate its effects on the VEGF-C-mediated reduction in cytotoxicity. One hundred nanomolar was chosen as the IC 50 for VEGF-R1 has been reported to be 90 nM, whereas for VEGF-R2 and VEGF-R3 the IC 50 is greater than 1 µM. Preincubation then coincubation with SU-5416 significantly blocked the reduction in cytotoxicity seen in response to VEGF-C (106.85 ± 2.8% of serum-starved cells compared with that of VEGF-C alone; 88.2 ± 3.3% of serum-starved conditions, P < 0.001, n = 12, ANOVA; Fig. 9 ) yet had no effect on cytotoxicity when incubated by itself (99.4 ± 2.9% of serum-starved conditions; Fig. 9 ).

Fig. 7. Effects of VEGF-C on VEGF-R3 tyrosine phosphorylation in hCIPs. hCIPs (2 experiments shown) and HMVECs (1 experiment shown) were serum starved and treated with VEGF-C (VC) or left untreated (SS). Protein was immunoprecipitated (IP) using an anti-VEGF-R3 antibody. A : proteins were subjected to SDS-PAGE, transferred to a PVDF membrane, and probed (IB) with an anti-VEGF-R3 antibody. B : membrane was stripped and reprobed using an anti-p-Y antibody. The only band seen was that corresponding to a molecular mass of 150 kDa in the IP of HMVECs treated with VEGF-C ( n = 4).

Fig. 8. Time-dependent effects of VEGF-C on VEGF-R3 tyrosine phosphorylation in hCIPs. hCIPs were serum starved and treated with VEGF-C (VC) or left untreated (SS) for varying amounts of time. Protein was immunoprecipitated (IP) using an anti-VEGF-R3 antibody. A : both the lysate (not shown) and precipitate were subjected to SDS-PAGE, transferred to a PVDF membrane, and probed (IB) with an anti-VEGF-R3 antibody. B : membrane was stripped and reprobed using an anti-p-Y antibody. Bands of 150 kDa corresponding to VEGF-R3 were seen in all samples when probed with anti-VEGF-R3, demonstrating that the assay worked. There were no bands corresponding to phosphorylated VEGF-R3 when probed with the anti-p-Y antibody, although bands at 230 kDa were seen in samples treated for just 2 min and also very faintly in those treated for 20 min.

Fig. 9. Reduction in cytotoxicity induced by VEGF-C is inhibited by SU-5416. hCIPs were serum starved, then treated with 1 nM VEGF-C, 100 nM SU-5416, VEGF-C with SU-5416 or left untreated for 24 h. LDH release from the cells was quantified and the percentage of cytotoxicity was then calculated. Data were expressed as means ± SE for the ratio of the percentage of cytotoxicity (treatment/serum starvation). VEGF-C significantly reduced cytotoxicity to that of control but when cells were preincubated with SU-5416, the effects of VEGF-C on cytotoxicity were blocked. SU-5416 did not induce cytotoxicity by itself. * P < 0.05, ** P < 0.001, ANOVA, Dunnett?s post hoc test, n = 12.

DISCUSSION

Glomerular expression of VEGF-D and VEGF-C. The results have refined previous observations regarding VEGF-C and -D expression in whole kidney by demonstrating glomerular expression of VEGF-C (podocytes) and VEGF-D (parietal epithelial cells). Local production of glomerular VEGF-C and VEGF-D suggests that VEGF-R3 expressed by podocytes is functional. However, as VEGF-D was only detected in some glomeruli and in parietal epithelial cells, "downstream" of VEGF-R3 on podocytes and separated by Bowman's space, we chose to study effects of VEGF-C. It is yet to be confirmed whether VEGF-D can exert a signaling effect in these cells.

VEGF-C as a potential autocrine factor in podocytes. The expression of VEGF-C by hCIPs suggests the possibility that VEGF-C may signal in an autocrine loop, binding and activating VEGF-R3 [as discussed by Jussila and Alitalo ( 21 )]. In support of an autocrine action not only do podocytes express VEGF-R3 but NRP2 mRNA has also been detected (as well as NRP1) ( 14 ), which binds VEGF-C and may be involved in VEGF-R3 signaling ( 22 ). A parallel autocrine action of VEGF-A has been demonstrated in hCIPs where VEGF-A decreases intracellular calcium and acts as a survival factor ( 10 ).

VEGF-C is predominantly expressed in the lymphatics and promotes survival, growth, and migration of lymphatic endothelial cells ( 28 ). The VEGF-C survival pathways via VEGF-R3 are similar to those of VEGF-A via VEGF-R2. This VEGF-R3 activation by VEGF-C leads to phosphorylation of Akt and reduced apoptosis as well as PKC-dependent activation of p42/p44 MAPK, which has also been linked to survival in HMVECs ( 28 ). In a leukemic cell line, which expresses VEGF-R2 and VEGF-R3, VEGF-C also promotes survival, indicated by increased Bcl-2/Bax ratios ( 6 ). This evidence suggests that VEGF-A and VEGF-C may activate similar signaling pathways in podocytes.

Comparing VEGF-C to VEGF-A signaling in podocytes. A comparison between VEGF-C and VEGF-A signaling in hCIPs was tested initially by examining the effect of VEGF-C on the common second messsenger, [Ca 2+ ] i. VEGF-C induced a reduction in R norm, which is proportional to changes in [Ca 2+ ] i. This was proportionally similar to the reduction seen with VEGF-A. Changes in [Ca 2+ ] i may be indicative of numerous responses, many of them detrimental including cell death and actin cytoskeleton reorganization [for example, in response to ANG II ( 30 )], particularly detrimental to glomerular function as podocytes do not regenerate. Although the mechanism for the reduction in [Ca 2+ ] i in response to VEGF-C is not known, the clear increase in activity of calcium extruding or sequestering mechanisms may function to oppose moderate changes in calcium to maintain homeostasis.

VEGF-C induced a reduction in cytotoxicity in hCIPs that had been serum starved, similar to that seen with VEGF-A in hCIPs ( Fig. 4 A ). VEGF-R3 tyrosine kinase inhibition by itself did not increase cytotoxicity above that induced by serum starvation, demonstrating that there was no nonspecific cytotoxic effect of MAZ51 on hCIPs ( Fig. 4 A ). Yet, MAZ51 should inhibit the endogenous VEGF-C autocrine loop through VEGF-R3, thereby inducing an increase in cytotoxicity. It is possible that the majority of VEGF-C detected in the glomerulus and podocytes by immunohistochemistry, immunofluorescence, and Western blotting is the longer, immature form (evident in Fig. 2 A ) and a stimulus is needed to induce the proteolytic activation of VEGF-C. VEGF-R3 tyrosine kinase inhibition prevented the VEGF-C-induced reduction in cytotoxicity suggesting that VEGF-C promotes survival in a VEGF-R3 kinase activation-dependent manner. These results, together with the [Ca 2+ ] i results, suggest that VEGF-C has similar effects to that of VEGF-A, governed by VEGF-R3 and VEGF-R1, respectively. Similarly, in HMVECs it was shown that VEGF-C has a more pronounced effect on signaling than VEGF-A, even though HMVECs express both VEGF-R2 and VEGF-R3 ( 28 ). It was suggested that this was due to the activation of both VEGF-R2 and VEGF-R3 by VEGF-C, unlike VEGF-A, which is not a ligand for VEGF-R3. The signaling pathways of both receptors were therefore suggested to be cumulative. This explanation cannot be applied to podocytes, which do not express VEGF-R2. In fact, it has been suggested that VEGF-C cannot activate VEGF-R3 in the absence of VEGF-R2 ( 2 ). This is clearly not universally the case, because VEGF-C induces an effect in a cell type which does not express VEGF-R2 ( 10 ). It may be possible, however, that VEGF-R3 can only be activated as part of a heterodimer, in which case VEGF-R1, expressed by podocytes ( 10 ), is a potential candidate, although there is no evidence in the literature to date that these two receptors can heterodimerize. Interestingly, it is known that VEGF-R1 can heterodimerize with VEGF-R2 ( 23 ) and VEGF-R2 can heterodimerize withVEGF-R3 ( 7 ).

Potential mechanisms by which VEGF-C promotes survival in podocytes. VEGF-C did not appear to induce Akt phosphorylation at a time course and dose known to stimulate Akt phosphorylation in other cell types using a previously published time course ( 28 ), whereas VEGF-A induces a reduction in Akt phosphorylation. This was surprising, yet it narrows down the possibilities for VEGF-C survival signaling. VEGF-C may act on a different time course to VEGF-A in podocytes, as observed with HMVECs; Akt phosphorylation induced by VEGF-C peaked at least 10 min before that of VEGF-A, and it also led to a more sustained upregulation of phosphorylated MAPK compared with VEGF-A ( 28 ). Therefore, VEGF-C could potentially promote survival in hCIPs through a reduction in apoptosis by activating a signaling pathway, which results in a more rapid response in Akt phosphorylation compared with VEGF-A and a more sustained MAPK phosphorylation response than VEGF-A.

MAPK phosphorylation, usually associated with mitogenesis in VEGF-A signaling ( 45 ), has been implicated in VEGF-C survival signaling in HMVECs ( 28 ). MAPK-activated kinases, such as Rsks (pp90 ribosomal S6 kinase family), are thought to phosphorylate the transcription factor, cAMP response element binding protein (CREB), which increases transcription of prosurvival genes, such as Bcl-2 ( 4 ). Rsks have also been shown to phosphorylate BAD at serine 112, a different serine residue to the one phosphorylated by activated AKT. MAPK phosphorylation can therefore potentially promote survival by upregulating antiapoptotic proteins and deactivating proapoptotic proteins. In contrast to this hypothesis, VEGF-C induced a reduction in MAPK phosphorylation in podocytes ( Fig. 6 ). It is interesting that a similar effect was seen to the effect of VEGF-A on Akt phosphorylation ( Fig. 5 ). It may be that basal MAPK phosphorylation levels are lower in cells that have been serum starved, including hCIPs. This has been demonstrated for basal Akt phosphorylation levels in serum-starved endothelial cells ( 12 ). Conversely, increased MAPK phosphorylation has been associated with podocyte damage in animal and cell models such as a MAPK kinase-dependent increase in MMP-9 enzymatic activities in murine podocytes in response to TGF- ( 27 ), an increased level of ERK1/2 activity in rat podocytes with sublytic exposure to C5b-9 resulting in DNA damage ( 34 ), and increased MAPK activity in a mouse model of collapsing FSGS of HIV-associated nephropathy ( 15 ). It has also been described in biopsies of human diabetic nephropathy ( 36 ). Therefore, a reduction in MAPK phosphorylation in podocytes in response to VEGF-C may actually be a protective mechanism, which supports our hypothesis of VEGF-C acting as a survival factor. In contrast to VEGF-C, neither recombinant VEGF-A protein nor a neutralizing antibody to VEGF-A had a significant effect on MAPK phosphorylation. This suggests that neither exogenous nor endogenous VEGF-A signals through MAPK phosphorylation in hCIPs, which is compatible with the absence of VEGF-R2 ( 26 ).

VEGF-R3 activation by VEGF-C in podocytes. VEGF-R3 immunoprecipitation studies revealed that 1 nM VEGF-C does not appear to induce the phosphorylation of VEGF-R3, in contrast to the cytotoxicity assay that demonstrated that 1 nM VEGF-C induced a reduction in cytotoxicity, which was dependent on kinase activation of VEGF-R3. VEGF-R3 has both a kinase domain and a phosphorylation substrate, and it is assumed to function through autophosphorylation. It is possible that the treatment time of hCIPs by VEGF-C was not conducive to VEGF-R3 phosphorylation in this cell type and that it follows a more acute or more chronic time course. This was not apparent though with the times of VEGF-C treatment chosen in Fig. 8. There seems to be a similar quantity of VEGF-R3 pulled down in the immunoprecipitation assays of the hCIPs compared with the HMVECs, so if VEGF-R3 is phosphorylated by VEGF-C then it would appear to be different due to the biochemistry of the receptor complex. Together, these results suggest that VEGF-C induces the kinase activation of VEGF-R3, yet not its phosphorylation. If VEGF-R3 is forming heterodimers with VEGF-R1, VEGF-C may be capable of binding to VEGF-R3, which would induce the transphosphorylation of VEGF-R1, but may not necessarily induce the phosphorylation of the specific tyrosine residue on VEGF-R3 to induce autophosphorylation. This would explain the lack of VEGF-R3 phosphorylation. Little is known, however, about possible heterodimerization or interaction of these two receptors. In fact, it has only recently been demonstrated that the cytoplasmic tail of VEGF-R3 contains five potential tyrosine residues ( 7 ). We demonstrated that the reduction in cytotoxicity seen in response to VEGF-C is dependent on VEGF-R1 activation ( Fig. 9 ). This is the first demonstration that VEGF-C interacts with VEGF-R1 and raises many questions about the interaction of VEGF-R3 and VEGF-R1, which will need to be pursued in detail and may be relevant in other cell types, which also express VEGF-R1 and VEGF-R3. It also highlights the potential for VEGF-C to signal in a paracrine manner to other cells within the glomerulus, such as the endothelial cells, in a similar manner to VEGF-A. The glomerular endothelial cells are known to express VEGF-R1, VEGF-R2, VEGF-R3, and neuropilins ( 39 ). The effects of VEGF-C on these cells may well be different to those on podocytes because of the presence of VEGF-R2. The effect of both VEGF-A and VEGF-C on podocytes and glomerular endothelial cells may ultimately lie with the expression levels of their VEGF receptors and their efficacy of activation.

In summary, VEGF-C promotes survival in hCIPs in both a VEGF-R3- and VEGF-R1-dependent manner, through a different mechanism to VEGF-A (summarized in Table 1 ). The exact receptor complex biochemistry and downstream signaling events for VEGF-C have yet to be elucidated. These results add to the increasing complexity of signaling pathways, which may maintain the unique phenotypes of glomerular cells, essential for their filtration function.

Table 1. Summary of the comparisons between VEGF-A and VEGF-C signaling in hCIPs

GRANTS

This work was supported by the Richard Bright VEGF Research Trust, the Wellcome Trust (69029), and the British Heart Foundation (BB000003 ).

【参考文献】
  Achen MG, Jeltsch M, Kukk E, Makinen T, Vitali A, Wilks AF, Alitalo K, and Stacker SA. Vascular endothelial growth factor D (VEGF-D) is a ligand for the tyrosine kinases VEGF receptor 2 (Flk1) and VEGF receptor 3 (Flt4). Proc Natl Acad Sci USA 95: 548-553, 1998.

Alam A, Herault JP, Barron P, Favier B, Fons P, Delesque-Touchard N, Senegas I, Laboudie P, Bonnin JM, Cassan C, Savi P, Ruggeri B, Carmeliet P, Bono F, and Herbert JM. Heterodimerization with vascular endothelial growth factor receptor-2 (VEGFR-2) is necessary for VEGFR-3 activity. Biochem Biophys Res Commun 324: 909-915, 2004.

Bailey E, Bottomley MJ, Westwell S, Pringle JH, Furness PN, Feehally J, Brenchley PE, and Harper SJ. Vascular endothelial growth factor mRNA expression in minimal change, membranous, and diabetic nephropathy demonstrated by nonisotopic in situ hybridisation. J Clin Pathol 52: 735-738, 1999.

Bonni A, Brunet A, West AE, Datta SR, Takasu MA, and Greenberg ME. Cell survival promoted by the Ras-MAPK signaling pathway by transcription-dependent and -independent mechanisms. Science 286: 1358-1362, 1999.

Cao Y, Linden P, Farnebo J, Cao R, Eriksson A, Kumar V, Qi JH, Claesson-Welsh L, and Alitalo K. Vascular endothelial growth factor C induces angiogenesis in vivo. Proc Natl Acad Sci USA 95: 14389-14394, 1998.

Dias S, Choy M, Alitalo K, and Rafii S. Vascular endothelial growth factor (VEGF)-C signaling through FLT-4 (VEGFR-3) mediates leukemic cell proliferation, survival, and resistance to chemotherapy. Blood 99: 2179-2184, 2002.

Dixelius J, Makinen T, Wirzenius M, Karkkainen MJ, Wernstedt C, Alitalo K, and Claesson-Welsh L. Ligand-induced vascular endothelial growth factor receptor-3 (VEGFR-3) heterodimerization with VEGFR-2 in primary lymphatic endothelial cells regulates tyrosine phosphorylation sites. J Biol Chem 278: 40973-40979, 2003.

Farnebo F, Piehl F, and Lagercrantz J. Restricted expression pattern of vegf-d in the adult and fetal mouse: high expression in the embryonic lung. Biochem Biophys Res Commun 257: 891-894, 1999.

Fitz LJ, Morris JC, Towler P, Long A, Burgess P, Greco R, Wang J, Gassaway R, Nickbarg E, Kovacic S, Ciarletta A, Giannotti J, Finnerty H, Zollner R, Beier DR, Leak LV, Turner KJ, and Wood CR. Characterization of murine Flt4 ligand/VEGF-C. Oncogene 15: 613-618, 1997.

Foster RR, Hole R, Anderson K, Satchell SC, Coward RJ, Mathieson PW, Gillatt DA, Saleem MA, Bates DO, and Harper SJ. Functional evidence that vascular endothelial growth factor may act as an autocrine factor on human podocytes. Am J Physiol Renal Physiol 284: F1263-F1273, 2003.

Foster RR, Saleem MA, Mathieson PW, Bates DO, and Harper SJ. Vascular endothelial growth factor and nephrin interact and reduce apoptosis in human podocytes. Am J Physiol Renal Physiol 288: F48-F57, 2005.

Gratton JP, Morales-Ruiz M, Kureishi Y, Fulton D, Walsh K, and Sessa WC. Akt downregulation of p38 signaling provides a novel mechanism of vascular endothelial growth factor-mediated cytoprotection in endothelial cells. J Biol Chem 276: 30359-30365, 2001.

Gunningham SP, Currie MJ, Han C, Turner K, Scott PA, Robinson BA, Harris AL, and Fox SB. Vascular endothelial growth factor-B and vascular endothelial growth factor-C expression in renal cell carcinomas: regulation by the von Hippel-Lindau gene and hypoxia. Cancer Res 61: 3206-3211, 2001.

Harper SJ, Xing CY, Whittle C, Parry R, Gillatt D, Peat D, and Mathieson PW. Expression of neuropilin-1 by human glomerular epithelial cells in vitro and in vivo. Clin Sci (Lond) 101: 439-446, 2001.

He J, Husain M, Sunamoto M, Agati VD, Klotman M, Iyengar R, and Klotman P. Nef stimulates proliferation of glomerular podocytes through activation of Src-dependent Stat3 and MAPK1,2 pathways. J Clin Invest 114: 643-651, 2004.

Iljin K, Karkkainen M, Lawrence E, Kimak M, Uutela M, Taipale J, Pajusola K, Alhonen L, Halmekyto M, Finegold D, Ferrell R, and Alitalo K. VEGFR3 gene structure, regulatory region, and sequence polymorphisms. FASEB J 15: 1028-1036, 2001.

Jeltsch M, Kaipainen A, Joukov V, Meng X, Lakso M, Rauvala H, Swartz M, Fukumura D, Jain RK, and Alitalo K. Hyperplasia of lymphatic vessels in VEGF-C transgenic mice. Science 276: 1423-1425, 1997.

Joukov V, Kumar V, Sorsa T, Arighi E, Weich H, Saksela O, and Alitalo K. A recombinant mutant vascular endothelial growth factor-C that has lost vascular endothelial growth factor receptor-2 binding, activation, and vascular permeability activities. J Biol Chem 273: 6599-6602, 1998.

Joukov V, Pajusola K, Kaipainen A, Chilov D, Lahtinen I, Kukk E, Saksela O, Kalkkinen N, and Alitalo K. A novel vascular endothelial growth factor, VEGF-C, is a ligand for the Flt4 (VEGFR-3) and KDR (VEGFR-2) receptor tyrosine kinases. EMBO J 15: 290-298, 1996.

Joukov V, Sorsa T, Kumar V, Jeltsch M, Claesson-Welsh L, Cao Y, Saksela O, Kalkkinen N, and Alitalo K. Proteolytic processing regulates receptor specificity and activity of VEGF-C. EMBO J 16: 3898-3911, 1997.

Jussila L and Alitalo K. Vascular growth factors and lymphangiogenesis. Physiol Rev 82: 673-700, 2002.

Karkkainen MJ, Saaristo A, Jussila L, Karila KA, Lawrence EC, Pajusola K, Bueler H, Eichmann A, Kauppinen R, Kettunen MI, Yla-Herttuala S, Finegold DN, Ferrell RE, and Alitalo K. A model for gene therapy of human hereditary lymphedema. Proc Natl Acad Sci USA 98: 12677-12682, 2001.

Kendall RL, Wang G, and Thomas KA. Identification of a natural soluble form of the vascular endothelial growth factor receptor, FLT-1, and its heterodimerization with KDR. Biochem Biophys Res Commun 226: 324-328, 1996.

Kirkin V, Mazitschek R, Krishnan J, Steffen A, Waltenberger J, Pepper MS, Giannis A, and Sleeman JP. Characterization of indolinones which preferentially inhibit VEGF-C- and VEGF- D -induced activation of VEGFR-3 rather than VEGFR-2. Eur J Biochem 268: 5530-5540, 2001.

Kukk E, Lymboussaki A, Taira S, Kaipainen A, Jeltsch M, Joukov V, and Alitalo K. VEGF-C receptor binding and pattern of expression with VEGFR-3 suggests a role in lymphatic vascular development. Development 122: 3829-3837, 1996.

Lamalice L, Houle F, Jourdan G, and Huot J. Phosphorylation of tyrosine 1214 on VEGFR2 is required for VEGF-induced activation of Cdc42 upstream of SAPK2/p38. Oncogene 23: 434-445, 2004.

Liu S, Liang Y, Huang H, Wang L, Li Y, Li J, Li X, and Wang H. ERK-dependent signaling pathway and transcriptional factor Ets-1 regulate matrix metalloproteinase-9 production in transforming growth factor- 1 stimulated glomerular podocytes. Cell Physiol Biochem 16: 207-216, 2005.

Makinen T, Veikkola T, Mustjoki S, Karpanen T, Catimel B, Nice EC, Wise L, Mercer A, Kowalski H, Kerjaschki D, Stacker SA, Achen MG, and Alitalo K. Isolated lymphatic endothelial cells transduce growth, survival and migratory signals via the VEGF-C/D receptor VEGFR-3. EMBO J 20: 4762-4773, 2001.

Nitschke R, Henger A, Ricken S, Gloy J, Muller V, Greger R, and Pavenstadt H. Angiotensin II increases the intracellular calcium activity in podocytes of the intact glomerulus. Kidney Int 57: 41-49, 2000.

O'Hare MJ, Bond J, Clarke C, Takeuchi Y, Atherton AJ, Berry C, Moody J, Silver AR, Davies DC, Alsop AE, Neville AM, and Jat PS. Conditional immortalization of freshly isolated human mammary fibroblasts and endothelial cells. Proc Natl Acad Sci USA 98: 646-651, 2001.

Pajusola K, Aprelikova O, Korhonen J, Kaipainen A, Pertovaara L, Alitalo R, and Alitalo K. FLT4 receptor tyrosine kinase contains seven immunoglobulin-like loops and is expressed in multiple human tissues and cell lines. Cancer Res 52: 5738-5743, 1992.

Partanen TA, Arola J, Saaristo A, Jussila L, Ora A, Miettinen M, Stacker SA, Achen MG, and Alitalo K. VEGF-C and VEGF-D expression in neuroendocrine cells and their receptor, VEGFR-3, in fenestrated blood vessels in human tissues. FASEB J 14: 2087-2096, 2000.

Pippin J, Durvasula R, Petermann A, Hiromura K, Couser W, and Shankland S. DNA damage is a novel response to sublytic complement C5b-9-induced injury in podocytes. J Clin Invest 111: 877-885, 2003.

Rissanen TT, Markkanen JE, Gruchala M, Heikura T, Puranen A, Kettunen MI, Kholova I, Kauppinen RA, Achen MG, Stacker SA, Alitalo K, and Yla-Herttuala S. VEGF-D is the strongest angiogenic and lymphangiogenic effector among VEGFs delivered into skeletal muscle via adenoviruses. Circ Res 92: 1098-1106, 2003.

Sakai N, Wada T, Furuichi K, Iwata Y, Yoshimoto K, Kitagawa K, Kokubo S, Kobayashi M, Hara A, Yamahana J, Okumura T, Takasawa K, Takeda S, Yoshimura M, Kida H, and Yokoyama H. Involvement of extracellular signal-regulated kinase and p38 in human diabetic nephropathy. Am J Kidney Dis 45: 54-65, 2005.

Saleem MA, Ni L, Witherden I, Tryggvason K, Ruotsalainen V, Mundel P, and Mathieson PW. Colocalization of nephrin, podocin, and the actin cytoskeleton: evidence for a role in podocyte foot process formation. Am J Pathol 161: 1459-1466, 2002.

Saleem MA, O'Hare MJ, Reiser J, Coward RJ, Inward CD, Farren T, Xing CY, Ni L, Mathieson PW, and Mundel P. A conditionally immortalized human podocyte cell line demonstrating nephrin and podocin expression. J Am Soc Nephrol 13: 630-638, 2002.

Satchell S, Anderson K, and Mathieson P. Angiopoietin 1 and vascular endothelial growth factor modulate human glomerular endothelial cell barrier properties. J Am Soc Nephrol 15: 566-574, 2004.

Shibuya M. Vascular endothelial growth factor receptor family genes: when did the three genes phylogenetically segregate? Biol Chem 383: 1573-1579, 2002.

Stacker SA, Stenvers K, Caesar C, Vitali A, Domagala T, Nice E, Roufail S, Simpson RJ, Moritz R, Karpanen T, Alitalo K, and Achen MG. Biosynthesis of vascular endothelial growth factor-D involves proteolytic processing which generates noncovalent homodimers. J Biol Chem 274: 32127-32136, 1999.

Taipale J, Makinen T, Arighi E, Kukk E, Karkkainen M, and Alitalo K. Vascular endothelial growth factor receptor-3. Curr Top Microbiol Immunol 237: 85-96, 1999.

Tille JC, Wang X, Lipson KE, McMahon G, Ferrara N, Zhu Z, Hicklin DJ, Sleeman JP, Eriksson U, Alitalo K, and Pepper MS. Vascular endothelial growth factor (VEGF) receptor-2 signaling mediates VEGF-C(deltaNdeltaC)- and VEGF-A-induced angiogenesis in vitro. Exp Cell Res 285: 286-298, 2003.

Veikkola T, Jussila L, Makinen T, Karpanen T, Jeltsch M, Petrova TV, Kubo H, Thurston G, McDonald DM, Achen MG, Stacker SA, and Alitalo K. Signaling via vascular endothelial growth factor receptor-3 is sufficient for lymphangiogenesis in transgenic mice. EMBO J 20: 1223-1231, 2001.

Wu LW, Mayo LD, Dunbar JD, Kessler KM, Baerwald MR, Jaffe EA, Wang D, Warren RS, and Donner DB. Utilization of distinct signaling pathways by receptors for vascular endothelial cell growth factor and other mitogens in the induction of endothelial cell proliferation. J Biol Chem 275: 5096-5103, 2000.

作者单位：1 Microvascular Research Laboratories, Department of Physiology, and 2 Academic Renal Unit, Southmead Hospital, University of Bristol, Bristol, United Kingdom; and 3 Department of Nephrology, First Affiliated Hospital, Nanjing Medical University, Nanjing, People‘s Republic of China