Requirement of Protein Kinase D Tyrosine Phosphorylation for VEGF-A165-induced Angiogenesis through Its Interaction and Regulation of Phospholipase Cγ Phosphorylation*

Vascular endothelial cell growth factor-A165 (VEGF-A165) is critical for angiogenesis. Although protein kinase C-mediated protein kinase D(PKD)activation was implicated in the response, the detailedmechanism remains unclear. In this study, we found that VEGF-A165-stimulated tyrosine phosphorylation of PKD and the dominant negative mutant of PKD, PKD(Y463F), inhibited VEGF-A165-induced human umbilical vein endothelial cell (HUVEC) proliferation. In addition, PKD(S738A/S742A) overexpression inhibited VEGF-induced HUVEC migration. Furthermore, knockdown of PKD by its specific small interfering RNA inhibited VEGF-induced HUVEC proliferation and migration. Moreover transfection of PKD(Y463F), PKD(S738A/S742A), or PKD-small interfering RNA blocked VEGF-induced angiogenesis in vivo. Our signaling experiments show that KDR not Flt-1 mediated PKD tyrosine phosphorylation and KDR tyrosine residues 951 and 1059 were required for VEGF-A165-stimulated PKD serine and tyrosine phosphorylation, respectively. Whereas G protein Gβγ subunits were required for both PKD serine phosphorylation and tyrosine phosphorylation, intracellular Ca2+ mobilization was required for VEGF-A165-stimulated PKD tyrosine phosphorylation and phospholipase C (PLC) activity was required for PKD serine phosphorylation. Surprisingly, the PLC inhibitor did not inhibit PKD tyrosine phosphorylation. Instead, PKD tyrosine 463 was required for VEGF-A165-stimulated PLCγ tyrosine phosphorylation. Moreover, PKD interacted with PLCγ even in unstimulated cells, and PKD tyrosine 463 phosphorylation was not required for this interaction. Together, we demonstrate that PKD interacts with PLCγ and becomes tyrosine phosphorylated upon VEGF stimulation, leading to PLCγ activation and angiogenic response of VEGF-A165.

and growth factors, such as vascular endothelial growth factor (VEGF-A 165 ), 4 basic fibroblast growth factor, platelet-derived growth factor, and transforming growth factor-␣, have an angiogenic activity (1)(2)(3). Among these, VEGF-A 165 stands out because of its potency and selectivity for vascular endothelium. VEGF-A 165 is not only involved in several steps of angiogenesis but is also the only angiogenic factor recognized to date that renders microvessels hyperpermeable to circulating macromolecules (4 -8). VEGF-A 165 extensively reprograms endothelial cell expression of proteases, integrins, and glucose transporters; stimulates endothelial cell migration and division; and protects endothelial cells from apoptosis and senescence (9 -12).
PKD, also known as protein kinase C (PKC), is a serine/ threonine protein kinase with structure, enzymology, and regulatory properties different from the PKC family members. Its most unique feature includes the presence of a Ca 2ϩ -independ-* This work was supported by National Institutes of Health Grant K01 ent catalytic domain, a regulatory pleckstrin homology region, and a highly hydrophobic stretch of amino acids in its N-terminal region (33,34). PKD can be activated in intact cells in response to numerous extracellular stimuli such as growth factors and ligands for G protein-coupled receptors (35)(36)(37)(38)(39)(40)(41)(42)(43). In all these cases, rapid PKD activation is believed to be mediated by PKC-dependent phosphorylation of Ser 738 and Ser 742 within the activation loop of the catalytic domain of PKD (44 -46). PKD activation is associated with its translocation to the plasma membrane and subsequent transient accumulation in the nucleus (44 -46). PKD overexpression markedly potentiates DNA synthesis induced by the G protein-coupled receptor agonists bombesin and vasopressin in Swiss 3T3 cells by increasing the duration of MAP kinase activation (47). On the other hand, PKD can also be tyrosine phosphorylated at tyrosine 463 in response to oxidative stress in HeLa cells, leading to activation of PKD (40, 48 -50). Therefore, we set out to study the role of PKD in VEGF-stimulated angiogenesis.
In this study, we found that VEGF-stimulated tyrosine phosphorylation of PKD in addition to the previously reported serine 738/serine 742 phosphorylation (51). By overexpressing the dominant negative mutants, PKD(S738A/ S742A) and PKD(Y463F), we found that tyrosine 463 phosphorylation was required for VEGF-A 165 -stimulated HUVEC proliferation but not migration and that VEGF-induced HUVEC migration involved PKD Ser 738 /Ser 742 phosphorylation. Both tyrosine phosphorylation and serine phosphorylation of PKD were required for VEGF-A 165 -induced angiogenesis in vivo, but they were regulated by different signaling pathways. Furthermore, our results indicate that PKD tyrosine phosphorylation was required for PLC␥ tyrosine phosphorylation in VEGF-A 165 -stimulated HUVEC. More importantly, immunoprecipitation experiments show that PKD physically interacted with PLC␥ even in non-stimulated HUVEC.

EXPERIMENTAL PROCEDURES
Materials-Recombinant VEGF-A 165 was obtained from R&D Systems (Minneapolis, MN). EGM-MV Bullet Kit, trypsin-EDTA, and trypsin neutralization solution were obtained from Clonetics (San Diego, CA). Vitrogen 100 was purchased from Collagen Biomaterials (Palo Alto, CA). Mouse monoclonal antibody against KDR and rabbit polyclonal antibody against PKD were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Anti-phosphotyrosine antibody (PY20) was obtained from Upstate Biotechnology (Lake Placid, NY). Rabbit anti-phosphoserine PKD and mouse monoclonal antibody against PLC␥ were obtained from Cell Signaling (Beverly, MA) and Transduction Laboratory (San Jose, CA), respectively. [ 3 H]Thymidine was obtained from PerkinElmer Life Sciences. Transwell plate inserts were from Fisher Scientific. CyQuant, Fura-2 AM, and Pluronic F-127 were obtained from Molecular Probes (Eugene, OR).
Cell Culture-Primary HUVEC were obtained from Clonetics (San Diego, CA). Cells were grown on plates coated with 30 g/ml vitrogen in EGM-MV Bullet Kit (5% fetal bovine serum in endothelial basal medium with 12 g/ml bovine brain extract, 1 g/ml hydrocortisone, 1 l/ml GA-1000, and EGF). HUVEC transduced with EGDR, EGLT, or EGDR mutants were grown in the same medium without EGF. HUVEC (passages 3 or 4) that were ϳ80% confluent were used for most experiments. Cells were serum starved in 0.1% fetal bovine serum in EBM for 24 h prior to treatment.
Proliferation Assays-Assays were carried out as described (22,23 Migration Assays-Assays were carried out as described (22,23). Serum-starved HUVECs (infected with retrovirus) were detached from tissue culture plates as described (22,23) and washed twice with endothelial basal medium containing 0.1% fetal bovine serum, and seeded (1 ϫ 10 5 cells per well) into the transwells coated with vitrogen (30 g/ml) and the transwells were inserted in a 24-well plate containing 1 ml of the same medium. Cells over a range of 3 ϫ 10 3 to 1 ϫ 10 5 per well were seeded in a 96-well plate for standard curve. Cells were incubated at 37°C for 1 h to allow the cells to attach, then VEGF-A 165 was added at a final concentration of 10 ng/ml. After incubation for an additional 2 h, cells remaining on the upper surface of the transwell filter membrane were wiped off with a cotton tip. The whole transwell membrane was cut out and placed in an individual well of the 96-well plate, which contains the cells for standard curve. Two hundred l of CyQuant DNA stain was added to each well containing cells or membrane and the plate was kept at 4°C overnight. After warming to room temperature, stained cells were counted in a spectrofluorometer (SpectraFluor; TECAN) with Delta Soft 3 software. Data were expressed as the mean Ϯ S.D. of quadruplicate values. All experiments were repeated at least three times.
Protein Kinase Phosphorylation-Serum-starved HUVECs transduced with different plasmids as indicated were treated with 10 ng/ml VEGF-A 165 , or 10 ng/ml EGF in the case of EGDR, EGLT, or mutants, for different time intervals as indicated. Cell lysates were either subjected to immunoblot analysis using an antibody specific for phosphoserine PKD, or immunoprecipitated with antibodies against PY20 or PKD followed by immunoblotting with antibodies as indicated. All experiments were repeated 3 times.
Quantitative Analysis of Plasma Volumes in Matrigel Assays-These assays were carried out as described (53). Mice (4 per group) implanted with various cell combinations in Matrigel were anesthetized with Avertin (tribromoethanol, 200 mg/kg) and injected intravenously via the tail vein with 0.2 ml of Evans blue dye (5 mg/ml in saline). After 5 min, blood was collected in heparin by cardiac puncture and centrifuged at 14,000 ϫ g for 10 min to obtain platelet poor plasma that was diluted in formamide for measurement of Evans blue dye concentration. Animals were euthanized by CO 2 narcosis and Matrigel plugs were dissected free by cautery to prevent blood loss, weighed, and extracted with 2 ml of formamide at room temperature for 3 days. Dye in plasma or extracted from Matrigels was measured at 620 nM in a Thermo Max microplate reader (Molecular Devices, Menlo Park, CA) using Softmax 881 software. Standard curves were generated by measurement of serial dilutions of Evans blue dye in formamide (g/ml). Intravascular plasma volumes (microliters per g of Matrigel) were calculated on the basis of Evans blue dye concentrations in blood plasma to provide an absolute measure of the volume of plasma in the vascular bed.
Immunohistochemistry-Implanted Matrigel plugs were dissected free, fixed in 4% paraformaldehyde for 4 h, changed to 30% sucrose overnight, and embedded in OCT compound. Frozen sections were then blocked with 5% goat serum and stained with the rat anti-mCD31 antibody (1:50 dilution, BD Biosciences) at room temperature for 1 h. Sections were then washed 3 times with phosphate-buffered saline and incubated for 1 h with biotinylated polyclonal anti-rat IgG antibody (1:500 dilution). Sections were then washed 3 times with phosphatebuffered saline, reacted with the ABC peroxidase kit (Vector Laboratories, Inc. Burlingame, CA) at room temperature for 45 min, and washed twice with phosphate-buffered saline prior to mounting for light microscopy and photography.
Animal Welfare-All animal experiments were performed in compliance with the Beth Israel Deaconess Medical Centers Animal Care and Use Committee.
Statistics-Analysis of variance and the Tukey-Kramer multiple comparisons test were used to determine statistical significance.

VEGF-A 165 -stimulated PKD Phosphorylation at Serine and
Tyrosine Residues-HUVEC was treated with VEGF-A 165 for the indicated times. Equal amounts of cellular extracted proteins were immunoblotted with an antibody against serine 738/ serine 742-phosphorylated PKD. Our data show that VEGF-A 165 induced PKD serine 738/serine 742 phosphorylation (Fig.  1A) as reported most recently (51). To further examine whether VEGF-A 165 induced tyrosine phosphorylation of PKD, HUVEC were treated with VEGF-A 165 for similar period of times. Equal amounts of total cell proteins were immunoprecipitated with antibodies against phosphotyrosine antibody (PY20) or PKD, and the immunoprecipitates were then subjected to immunoblot analysis using an antibody against total PKD or PY20. The results show that VEGF-A 165 -stimulated PKD tyrosine phosphorylation in HUVEC in a time-dependent manner (Fig. 1B). (51) showed that PKC␣-mediated PKD serine 738/serine 742 phosphorylation is involved in VEGF-stimulated HUVEC proliferation. Because PKD tyrosine phosphorylation was shown to be required for PKD activation in response to oxidative stress, we also examined whether PKD tyrosine phosphorylation is involved in VEGF-stimulated HUVEC proliferation. To do this, we overexpressed a dominant negative mutant of PKD, PKD(Y463F), in HUVEC with our retrovirus expression system that gave almost 100% infection yield in HUVEC (22). As expected, VEGF-A 165 -stimulated thymidine incorporation in HUVEC transduced with LacZ ( Fig. 2A, lane 2 versus lane 1, p Ͻ 0.001). Baseline thymidine incorporation was unaffected in HUVEC transduced with PKD (Y463F) (lane 3 versus lanes 1, p Ͼ 0.05) but the response to VEGF-A 165 was strikingly inhibited (lane 4 versus lane 2, p Ͻ 0.001). To further confirm that PKD is required for VEGF-A 165 -stimulated HUVEC proliferation, we used PKD-siRNA to knockdown the expression of PKD in HUVEC (Fig. 2B). Our data show that VEGF-A 165 -stimulated HUVEC proliferation was blocked in HUVEC expressing PKD-siRNA ( Fig. 2A, lane 6 versus lane 2, p Ͻ 0.001). However, negative control siRNA (SiNEG) had no effect (Fig. 2A, lane 8 versus lane 2, p Ͼ 0.05). We further tested whether PKD is required for VEGF-A 165stimulated HUVEC migration, another important feature of angiogenesis. As shown in Fig. 2C, VEGF-A 165 -stimulated HUVEC migration (Fig. 2C, lane 2 versus lane 1, p Ͻ 0.001).
Requirement of PKD Tyrosine Phosphorylation and Serine Phosphorylation for VEGF-A 165 -stimulated HUVEC Angiogenesis in Vivo-We further tested whether PKD played an important role in VEGF-A 165 -induced angiogenesis in vivo with the Matrigel angiogenesis assay (53). To elucidate the mechanisms of angiogenesis, it would be desirable to modulate the expression of individual vascular genes in vivo using the gene overexpression and silencing approaches that have proved to be so powerful in vitro. Recently a novel system has been developed that allowed us to introduce DNA into endothelial cells in vivo (53). SK-MEL-2 tumor cells transfected to overexpress VEGF-A 165 (SK-MEL/VEGF cells) were mixed with PT67 cells packaging retroviruses that expressed PKD(S738A/S742A) and PKD(F463F). The cell mixtures were incorporated into Matrigels that were implanted in the subcutaneous space of nude mice. As previously reported (53), VEGF-A 165 secreted by SK-MEL/VEGF-A 165 cells induces nearby vascular endothelial cells to divide and therefore to become susceptible to infection with retroviruses secreted by PT67 packaging cells.
We next used the intravascular plasma volume of Matrigel plug-associated blood vessels as a novel measure to quantitate the angiogenic response (53). Intravascular plasma volume is an appropriate measure as enlarged mother vessels are a signature property of the early angiogenic response to VEGF-A (54, 55). Evans blue dye was injected intravenously into mice 3 days after implanting Matrigel plugs containing various cell mixtures. Evans blue dye binds to plasma proteins and therefore the amount of plasma within the Matrigel-associated vasculature can be calculated from simultaneous measurements of dye concentration in peripheral blood plasma. Matrigel plugs were harvested 5 min after intravenous dye injection, when blood vessels were filled with dye-plasma protein complexes but before there was time for significant extravasation. In Matrigel plugs con-taining VEGF-A 165 -expressing SKMEL/VEGF cells (alone or with PT67/LacZ cells), intravascular plasma volume as measured by Evans blue dye accumulation increased Ͼ2-fold above baseline levels (Fig. 3B, lane 2 versus lane 1, p Ͻ 0.001). The presence of PT67/PKD(Y463F) cells, PKD(S738A/S742A) cells, or PT67/PKD-siRNA strikingly inhibited the angiogenic response expected from SKMEL/VEGF cells (Fig. 3B, lanes 3-5  versus lane 2, all p Ͻ 0.001). Inclusion of PT67/SiNEG had no effect (Fig. 3B, lane 6 versus lane 2, p Ͼ 0.05). The quantitative measurements of vascular plasma volumes presented in Fig. 3B therefore confirm the qualitative measures of angiogenesis presented in Fig. 3A. In situ hybridization indicated that the expression levels of VEGF-A 165 were similar in Matrigels containing these different cell mixtures (see Ref. 53); Also, transfection of SKMEL and PT67 cells with PKD(Y463F) and PKD(S738A/S742A) has no effect on these cells proliferation (data not shown). Therefore, the results obtained cannot be attributed to effects on VEGF-A expression and the effect on SKMEL and PT67 cells.

Signaling Pathways That Regulated VEGF-stimulated PKD Serine Phosphorylation and Tyrosine Phosphorylation-To
identify which VEGF receptor mediates PKD tyrosine phosphorylation and serine phosphorylation, we used the recently developed receptor chimera (EGDR and EGLT) in which the N-terminal domains of KDR or Flt-1 were replaced with that of EGFR to dissect the signaling pathways mediated by KDR or Flt-1 (22). It was previously shown that HUVECs were not responsive to EGF treatment in the experimental conditions used (22). As expected, in the HUVEC transduced with LacZexpressing viruses, EGF did not stimulate PKD serine phosphorylation (Fig. 4A, first panel). When serum-starved HUVEC transduced with EGDR-or EGLT-expressing viruses were stimulated with EGF for different time intervals, it is EGDR but not EGLT that mediated PKD serine phosphorylation in a similar time course as in VEGF-A 165 -stimulated HUVEC (Figs. 1A and 4A, second and third panels). Recently, we demonstrated that tyrosines 1059 and 951 of KDR are required for VEGF-induced proliferation and migration, respectively  ); B, intravascular plasma volumes (l/g) in Matrigels from A as a quantitative measure of the angiogenic response. Intravascular plasma volumes were determined by accumulation of Evans blue dye administered intravenously 5 min prior to euthanasia (n ϭ 4). All experiments were repeated three times. (23). Therefore, we examined whether the EGDR mutants, EGDR(Y951F) and EGDR(Y1059F), are required for EGFstimulated PKD serine phosphorylation in HUVEC. As shown in Fig. 4A (fourth panel), EGDR(Y951F) cannot mediate PKD serine phosphorylation in response to EGF stimulation. However, EGDR(Y1059F) has no effect (Fig. 4A, fifth  panel). These results indicate that tyrosine 951, not tyrosine 1059, is essential for PKD serine phosphorylation.
It was shown that the G␤␥ subunits of G protein and PLC␥ tyrosine phosphorylation, but not intracellular Ca 2ϩ mobilization were required for VEGF-stimulated HUVEC migration (22, 24, 25, 27-29, 31, 32). Therefore, we examined whether these molecules were required for VEGF-A 165stimulated PKD serine phosphorylation. Serum-starved HUVEC that were transduced with the G␤␥ minigene, h␤APK1(495), or pretreated with U73122 (an inhibitor of the PLC family), U733443 (the negative control of U73122), or BAPTA/AM (an intracellular Ca 2ϩ chelator) for 5 min were stimulated with VEGF-A 165 for different times as indicated. Cellular extracts were immunoblotted with antibody against phosphoserine PKD. The data show that PKD serine phosphorylation was blocked by overexpression of h␤APK1(495) and pretreatment with the general PLC inhibitor U73122, however, pretreatment with BAPTA/AM or U733443 did not inhibit PKD serine phosphorylation in VEGF-A 165 -stimulated HUVEC (Fig. 4B). These results indicate that the G␤␥ subunit of G protein and PLC activity are key intermediary steps for VEGF-A 165 -induced PKD serine phosphorylation. However, intracellular Ca 2ϩ mobilization is not required.
Signaling Pathways That Regulated VEGF-stimulated PKD Tyrosine Phosphorylation-We further studied the signaling pathway that regulated PKD tyrosine phosphorylation. As expected, in the HUVEC transduced with LacZ-expressing viruses, EGF did not stimulate PKD tyrosine phosphorylation (Fig. 5A, first panel). When serum-starved HUVEC transduced with EGDR-or EGLT-expressing viruses were stimulated with EGF for different time intervals, it was EGDR but not EGLT that mediated PKD tyrosine phosphorylation in a similar time course as in VEGF-A 165 -stimulated HUVEC (Figs. 1B and 5A, second and third panels). Furthermore, EGDR(Y1059F) could not mediate PKD tyrosine phosphorylation in response to EGF stimulation as shown in Fig. 5A (fourth panel). However, EGDR(Y951F) had no effect (Fig. 5A, fifth panel). These results indicate that tyrosine 1059, not tyrosine 951, is essential for PKD tyrosine phosphorylation.
Because the G␤␥ subunits of G protein and PLC␥ activity are required for VEGF-A 165 -stimulated HUVEC proliferation (22, 24, 25, 27-29, 31, 32), we then examined whether these molecules were required for VEGF-A 165 -stimulated PKD tyrosine phosphorylation. Serum-starved HUVEC that were transduced with the G␤␥ minigene, h␤APK1(495), or pretreated with U73122 for 5 min, were stimulated with VEGF-A 165 for different times as indicated. Cellular extracts were immunoprecipitated with an antibody against PY20 and then immunoblotted with an antibody against PKD. The data show that PKD tyrosine phosphorylation was completely blocked by overexpres-  sion of h␤APK1(495), however, pretreatment with the PLC inhibitor U73122 did not inhibit PKD tyrosine phosphorylation in VEGF-A 165 -stimulated HUVEC (Fig. 5B). These results indicate that the G␤␥ subunit of G protein is a key intermediary step for VEGF-A 165 -induced PKD tyrosine phosphorylation. However, PLC activity is not required.
Requirement of PKD Tyrosine Phosphorylation for VEGF-A 165 -stimulated PLC␥ Tyrosine Phosphorylation-Our data that PLC activation is not required for VEGF-A 165 -stimulated PKD tyrosine phosphorylation suggested that PLC␥ activation might be downstream of PKD tyrosine phosphorylation. Therefore, we tested whether overexpression of the dominant negative PKD mutants had any effect on VEGF-A 165 -stimulated PLC␥ tyrosine phosphorylation. Serumstarved HUVEC that were transduced with LacZ as a control, PKD(Y463F) or PKD(S738A/S742A), were stimulated with VEGF-A 165 for different times as indicated. Cellular extracts were immunoprecipitated with an antibody against PY20 and then immunoblotted with an antibody against PLC␥. As shown in Fig. 6A, overexpression of PKD(Y463F) almost completely inhibited VEGF-A 165 -stimulated PLC␥ tyrosine phosphorylation in HUVEC (Fig. 6A, panel I). However, overexpression of PKD(S738A/S7482A) had no effect (Fig.  6A, panel I). To further confirm that PKD is involved in PLC␥ activation, we also knocked down PKD expression by the PKD siRNA. The data show that compared with control siRNA, PKD siRNA significantly inhibited VEGF-induced PLC␥ phosphorylation (Fig. 6A, panel I). The membranes were stripped and reprobed with an antibody against KDR. Our data clearly show that either PKD dominant negative mutants or PKD siRNA had no effect on VEGF-A 165 -stimulated KDR tyrosine phosphorylation (Fig. 6A, panel II). The same cellular extracts were immunoblotted with an antibody against MAPK to confirm the equal amount of proteins in all samples (Fig. 6A, panel III).
PKD Physically Interacts with PLC␥ Even in the Quiescent State-To further study the mechanism by which PKD regulates PLC␥ activation, we first tested whether VEGF-A 165 -stimulated PKD interaction with PLC␥. Serum-starved HUVEC were stimulated with VEGF-A 165 for different times as indicated. Cellular extracts were immunoprecipitated with an antibody against PKD and then immunoblotted with an antibody against PLC␥. Surprisingly, PKD interacted with PLC␥ in the absence of VEGF-A 165 stimulation (Fig. 6B). VEGF-A 165 treatment did not further stimulate PKD and PLC␥ interaction (Fig. 6B). To further confirm the PKD and PLC␥ interaction, we tested the effect of PKD dominant negative mutants on this interaction. Serum-starved HUVEC that were transduced with PKD(Y463F) or PKD(S738A/S742A) were stimulated with VEGF-A 165 . Cellular extracts were immunoprecipitated with an antibody against PKD and then immunoblotted with an antibody against PLC␥. As shown in Fig. 6B, overexpression of PKD(Y463F) and PKD(S738A/ S742A) had no effect on the PKD and PLC␥ interaction.

DISCUSSION
The unique serine/threonine protein kinase PKD is typically activated through protein kinase C-dependent phosphorylation of serine 738 and serine 742 at the activation loop in response to stimulation of growth factor receptors and G-protein-coupled receptors (35)(36)(37)(38)(39)(40)(41)(42)(43). Consistent with this pathway of PKD activation, it was very recently shown that PKD can be phosphorylated at Ser 738 /Ser 742 by protein kinase C␣ in VEGF-A 165 -stimulated HUVEC and knockdown of PKC␣ or PKD expression by their respective siRNA inhibited VEGF-A 165 -stimulated HUVEC proliferation (51). Interestingly, Storz et al. (48) reported a novel pathway for PKD activation, in which phosphorylation of PKD tyrosine 463 at the pleckstrin homology domain is essential for its activation by the Src-Abl pathway in response to oxidative stress. In this study, we found that PKD is not only phosphorylated at Ser 738 /Ser 842 , but also tyrosine phosphorylated presumably at tyrosine 463 in HUVEC stimulated with VEGF-A 165 . Importantly, overexpression of PKD(Y463F) markedly inhibited VEGF-A 165 -stimulated HUVEC proliferation, suggesting that in addition to serine phosphorylation, tyrosine 463 phosphorylation of PKD is also important for the VEGF response. Our data further indicate that Ser 738 /Ser 742 phosphorylation, not tyrosine 463 phosphorylation of PKD was required for VEGF-A 165 -stimulated HUVEC migration. These results are consistent with our previous findings that VEGF-A 165 stimulated HUVEC proliferation and migration by different signaling pathways (22,23,32,56,57).
It was previously shown that VEGF-A 165 stimulates HUVEC proliferation through the G␤␥ subunits of G protein and PLC activity (22,23,32,56). Our data clearly show that the G␤␥ subunit of the G protein was required for PKD tyrosine phosphorylation. Surprisingly, the general PLC inhibitor U73122 FIGURE 6. PKD physically interacted with PLC␥ and regulated PLC␥ tyrosine phosphorylation. A, serum-starved HUVEC that were transduced with LacZ as a control, PKD(Y463F) and PKD(S738S/S742A), or PKD siRNA or its control siRNA were stimulated with VEGF-A 165 (10 ng/ml) at different times as indicated. Cellular extracts were immunoprecipitated (IP) with an antibody against phosphorylated tyrosine (PY20) and then subjected to immunoblot analysis with an antibody against PLC␥ (panel I).
Membranes were stripped and immunoblotted (IB) with an antibody against KDR (panel II). Cellular extracts were immunoblotted with an antibody against MAPK to confirm equal protein loading (panel III); B, serumstarved HUVEC that were transduced with LacZ as a control, PKD(Y463F) and PKD(S738S/S742A) were stimulated with VEGF-A 165 (10 ng/ml). Equal amounts of cell protein were immunoprecipitated with the PKD antibody and then subjected to immunoblotting with PLC␥ antibody. The experiments were repeated three times.
could not inhibit VEGF-A 165 -stimulated PKD tyrosine phosphorylation, suggesting that PLC␥ activation may be downstream of PKD tyrosine phosphorylation (Fig. 7). Indeed, overexpression of PKD(Y463F) completely inhibited VEGF-A 165 -stimulated PLC␥ tyrosine phosphorylation. Tyrosine phosphorylation of PLC␥ is known to be necessary for VEGF-A 165 -stimulated MAP kinase activation and endothelial cell growth (28,58). Although PLC␥ tyrosine phosphorylation was believed to be mediated by the tyrosine kinase activity of VEGF receptor 2 (KDR) (58), our new findings that PLC␥ phosphorylation involves tyrosine phosphorylation of PKD suggest that there are interrelationships among KDR, PKD, and PLC␥, and possibly another tyrosine kinase such as Src, as described below. Although tyrosine 1175 of KDR, a putative docking site for PLC␥, was shown to mediate VEGF-A 165 -induced PLC␥ phosphorylation (58), our results show that another KDR autophosphorylation site, Tyr 1059 , mediates PKD tyrosine phosphorylation and subsequent PLC␥ phosphorylation. The detailed mechanism about how these molecules interact with and activate each other remains to be investigated. Further studies indicate that PKD physically interacted with PLC␥ even in the absence of VEGF-A 165 stimulation.
The molecular mechanism by which oxidative stress induces the tyrosine phosphorylation of PKD in HeLa cell was shown to be mediated the Src-Abl pathway (49). This ROS-Src pathway may also be involved in VEGF-A 165 -induced PKD-mediated PLC␥ activation based on several lines of evidence. First, reac-tive oxygen species was known to play an essential role in VEGF-A 165 -stimulated endothelial cell proliferation (59,60). Second, Src is required for VEGF-A 165 -induced PLC␥ activation and angiogenesis (61,62). Because our data also show that there is a constitutive association between PKD and PLC␥, we speculate that PKD might act as a scaffold to recruit SH2 domain-containing Src to this complex to phosphorylate PLC␥.
In summary, we have uncovered a novel mechanism in which PKD interacts with PLC␥ and upon KDR stimulation, PKD become serine phosphorylated and tyrosine phosphorylated by different pathways as described in the legend to Fig. 7. The phosphorylated tyrosine of PKD likely acts as a co-docking site together with the phosphorylated KDR to activate PLC␥ that in turn results in VEGF-A 165 -stimulated HUVEC proliferation and angiogenesis (Fig. 7).