A phosphoinositide 3-kinase/phospholipase Cgamma1 pathway regulates fibroblast growth factor-induced capillary tube formation

Abstract

Background: The fibroblast growth factors (FGFs) are key regulators of embryonic development, tissue homeostasis and tumour angiogenesis. Binding of FGFs to their receptor(s) results in activation of several intracellular signalling cascades including phosphoinositide 3-kinase (PI3K) and phospholipase C (PLC)gamma1. Here we investigated the basic FGF (FGF-2)-mediated activation of these enzymes in human umbilical vein endothelial cells (HUVECs) and defined their role in FGF-2-dependent cellular functions.

Methodology/principal findings: We show that FGF-2 activates PLCgamma1 in HUVECs measured by analysis of total inositol phosphates production upon metabolic labelling of cells and intracellular calcium increase. We further demonstrate that FGF-2 activates PI3K, assessed by analysing accumulation of its lipid product phosphatidylinositol-3,4,5-P(3) using TLC and confocal microscopy analysis. PI3K activity is required for FGF-2-induced PLCgamma1 activation and the PI3K/PLCgamma1 pathway is involved in FGF-2-dependent cell migration, determined using Transwell assay, and in FGF-2-induced capillary tube formation (tubulogenesis assays in vitro). Finally we show that PI3K-dependent PLCgamma1 activation regulates FGF-2-mediated phosphorylation of Akt at its residue Ser473, determined by Western blotting analysis. This occurs through protein kinase C (PKC)alpha activation since dowregulation of PKCalpha expression using specific siRNA or blockade of its activity using chemical inhibition affects the FGF-2-dependent Ser473 Akt phosphorylation. Furthermore inhibition of PKCalpha blocks FGF-2-dependent cell migration.

Conclusion/significance: These data elucidate the role of PLCgamma1 in FGF-2 signalling in HUVECs demonstrating its key role in FGF-2-dependent tubulogenesis. Furthermore these data unveil a novel role for PLCgamma1 as a mediator of PI3K-dependent Akt activation and as a novel key regulator of different Akt-dependent processes.

Figure 1. FGF-2 activates PLCγ1 in ECs.

(A) HUVEC were transfected with scrambled siRNA or specific siRNA to downregulate PLCγ1 expression. After 24 h cells were labelled with [³H]myo inositol in M199 containing 2% FBS for a further 24 h and then stimulated with 100 ng/ml FGF-2 for the indicated times. Data indicate the total amount of inositol phosphates generated at the indicated times and are expressed as percentage of the basal inositol phosphates. Data are means±SEM of values obtained from 5 independent experiments. *p<0.05, **p<0.01. Blot shows PLCγ1 expression in lysates of HUVEC transfected with a control scrambled siRNA or siRNA targeting PLCγ1. Equal loading of proteins was assessed by using an anti actin antibody. (B) Intracellular calcium increase in serum starved HUVEC was determined as described in the Materials and Methods session. Cells were stimulated with 100 ng/ml FGF-2 and 1 mM ATP at the indicated times (arrows). Data are mean of values obtained from 4 independent experiments. (C) Expression levels of PLCγ1 or PLCγ2 in HUVEC transfected with the indicated siRNAs. Equal loading was confirmed using anti Akt or anti ERK antibodies. (D) HUVEC were transfected with scrambled siRNA or siRNAs targeting PLCγ1 or PLCγ2. The number of cells was determined by manual counting 24 h or 48 h after transfection. Data are expressed as percentage of cells transfected with scrambled siRNA at each time point and are means±SEM of values obtained from 6 independent experiments. *p<0.05, **p<0.01 vs HUVEC expressing siRNA PLCγ1. In C and D siRNA targeting PLCγ2 were from Dharmacon.

Figure 2. FGF-2 activates PLCγ1 in a mechanism involving PI3K activation.

(A) HUVEC were labelled with [³H]myo inositol for 24 h and then pre-treated with 10 µM LY294002 for 15 min before stimulation with 100 ng/ml FGF-2. Data indicate the total amount of inositol phosphates generated at the indicated times and are expressed as percentages of the basal inositol phosphates. Data are means±SEM of values obtained from 9 independent experiments. *p<0.05. (B) Calcium release from serum starved HUVEC untreated or treated with 10 µM LY294002. Cells were stimulated with 100 ng/ml FGF-2 and 1 mM ATP at the indicated times (arrows). Data are mean of values obtained from 2 independent experiments.

Figure 3. FGF-2 activates PI3K in HUVEC.

(A) TLC analysis of phospholipids extracted from [³²P]-labelled HUVEC stimulated with 100 ng/ml FGF-2 for the indicated times or 1 µM sphingosine-1-phosphate (S1P) for 5 min. Arrow indicates the PtdIns-3,4,5-P₃ spot. (B) HUVEC grown on glass coverslips were incubated overnight in M199+0.5%FBS before stimulation with 100 ng/ml FGF-2 for the indicated times. Coverslips were then incubated with an anti PtdIns-3,4,5-P₃ antibody and analyzed by confocal microscopy. Arrows indicate the plasma membrane localization of the phosphoinositide. Bar = 10 µm.

Figure 4. PLCγ1 is required for FGF-2-induced cell migration.

(A) HUVEC were incubated overnight in M199+0.5%FBS before treatment with 5 µM U73122 or U73343 for 30 min. Migration of cells in the presence of 100 ng/ml FGF-2 was assessed by Transwell assays. Data are means±SEM of values obtained from 7 independent experiments and are expressed as percentage of cells migrating in the absence of FGF-2 (control) ***p<0.001. (B) HUVEC were transfected with scrambled siRNA or siRNA to downregulate PLCγ1 expression. After 24 h cells were incubated overnight in M199+0.5% FBS. Migration of cells in the presence of 100 ng/ml FGF-2 was then assessed by Transwell assays. Data are means±SEM of values obtained from 7-10 independent experiments *p<0.05.

Figure 5. PLCγ1 is necessary for FGF-2-induced tubulogenesis in vitro.

(A) Representative images of HUVEC on Matrigel in the absence or presence of FGF-2 plus 2 µM of the PLC inhibitor U73122 or its inactive analogue U73343. Plot represents a quantitative analysis of the experiments obtained by counting the number of branch points from five fields. Data are means±SEM of values obtained from 12 independent experiments.**p<0.01. NS = not significant. (B) Representative images of HUVEC transfected with the scrambled siRNA or siRNA targeting PLCγ1 on Matrigel in the presence of FGF-2. Quantitative analysis show data as means±SEM of values obtained from 5 independent experiments. NS = not significant.

Figure 6. PLCγ1-dependent PKCα activation mediates FGF-2-induced Akt activation.

HUVEC were incubated overnight in M199+0.5%FBS and then treated with the indicated concentrations of U73122 or U73343 for 30 min (A) or 10 µM of the conventional PKCs inhibitor Gö6976 (B) before stimulation with 100 ng/ml FGF-2 for 10 min. (C) HUVEC were transfected with scrambled siRNA or siRNAs targeting PKCα. After 24 h cells were serum deprived overnight and then stimulated with 100 ng/ml FGF-2 for the indicated times. Phosphorylation of Akt at Ser473 was assessed by using a specific antibody. Membranes were then stripped and re-probed with an anti Akt. Blots are representative of 7 (A) and 3 (B,C) independent experiments. In all cases, quantitative analysis of the band intensities was performed. Data represent values of pSer473 Akt/Akt expressed as fold increase over control.

Figure 7. PKCα is required for FGF-2-induced cell migration.

(A) HUVEC were incubated overnight in M199+0.5%FBS before treatment with 2 µM Gö6796, 50 µM or 10 µM of the MEK inhibitors PD98059 and U0126 respectively or 50 µM of the Akt inhibitor SH5 for 30 min. Migration of cells in the presence of 100 ng/ml FGF-2 was then assessed by Transwell assays. Data are means±SEM of values obtained from 2-7 independent experiments *p<0.05, **p<0.01. (B) HUVEC were transfected with scrambled siRNA or siRNA to downregulate PKCα expression. After 24 h cells were incubated overnight in M199+0.5% FBS. Migration of cells in the presence of 100 ng/ml FGF-2 was then assessed by Transwell assays. Data are means±SEM of values obtained from 5 independent experiments **p<0.01.