Role of PI3K/Akt and MEK/ERK Signalling in cAMP/Epac-Mediated Endothelial Barrier Stabilisation

Abstract

Background and aims: Activation of the cAMP/Epac signalling stabilises endothelial barrier function. Moreover, its activation is accompanied by an activation of PI3K/Akt and MEK/ERK signalling in diverse cell types but their impact on endothelial barrier function is largely unknown. Here the role of PI3K/Akt and MEK/ERK signalling in cAMP/Epac-mediated endothelial barrier stabilisation was analysed.

Methods: Endothelial barrier function was analysed in cultured human umbilical vein endothelial cells (HUVECs) by measuring flux of albumin. A modified cAMP analogue 8-pCPT-2'-O-Me-cAMP (Epac agonist) was used to specifically activate cAMP/Epac signalling.

Results: Epac agonist reduces the basal and attenuates thrombin-induced endothelial hyperpermeability accompanied by an activation of PI3K/Akt and MEK/ERK signalling. The qPCR data demonstrate HUVECs express PI3Kα, PI3Kβ, and PI3Kγ but not PI3Kδ isoforms. The western blot data demonstrate Epac agonist activates PI3Kα and PI3Kβ isoforms. Inhibition of MEK/ERK but not PI3K/Akt pathway potentiates the endothelial barrier protective effects of cAMP/Epac signalling. Inhibition of MEK/ERK signalling in the presence of Epac agonist induces a reorganisation of actin cytoskeleton to the cell periphery, enhanced VE-cadherin localisation at cell-cell junctions, and dephosphorylation of myosin light chains (MLC) but not inhibition of RhoA/Rock signalling. Moreover, Epac agonist promotes endothelial cell (EC) survival via reduction in activities of pro-apoptotic caspases in a PI3K/Akt and MEK/ERK signalling-dependent manner.

Conclusion: Our data demonstrate that the Epac agonist simultaneously activates diverse signalling pathways in ECs, which may have differential effects on endothelial barrier function. It activates PI3K/Akt and MEK/ERK signalling which mainly govern its pro-survival effects on ECs. Inhibition of MEK/ERK but not PI3K/Akt signalling enhances barrier stabilising and barrier protective effects of cAMP/Epac activation.

Chemical compounds used in this study: 8-pCPT-2'-O-Me-cAMP (PubChem CID: 9913268); Akt inhibitor VIII (PubChem CID: 10196499); AS-252424 (PubChem CID: 11630874); IC-87114 (PubChem CID: 9908783); PD 98059 (PubChem CID: 4713); PIK-75 (PubChem CID: 10275789); TGX-221 (PubChem CID: 9907093); Thrombin (PubChem CID: 90470996); U0126 (PubChem CID: 3006531); Wortmannin (PubChem CID: 312145).

Keywords: MEK/ERK; PI3K/Akt; Rac1; VE-cadherin; adherens junctions; cell survival; endothelial permeability; peripheral actin.

FIGURE 1

Effect of the Epac agonist on PI3K/Akt signalling in HUVECs. (A) Time-dependent effect of the Epac agonist 8-CPT-cAMP (8-CPT) on Akt phosphorylation in ECs. Upper panel: representative blots of Akt phosphorylation at S473. Lower panels: densitometric analysis of western blots. n = 3; P < 0.05 for all subfigures; ^∗ vs. control. (B) Effect of Akt inhibitor VIII (Akt inh. VIII) on the Epac agonist 8-CPT-induced Akt phosphorylation. ECs were incubated with different concentrations of Akt inh. VIII as indicated for 30 min and were exposed to 8-CPT for 20 min. (C) ECs were pre-treated with isoform specific PI3K inhibitors or DMSO as indicated before adding 8-CPT for 20 min. and Akt phosphorylation at S473 was analysed by western blot. GAPDH from the same blot was used as loading control. Representative blots from three different experiments. (D) The mRNA expression of PI3K isoforms in HUVEC (Freshly isolated cells, P0; Passage 1, P1; and Passage 2, P2). RNA from human left ventricular tissue (LV) was used as positive control.

FIGURE 2

Analysis of role of PI3K/Akt signalling in Epac-mediated EC barrier stabilisation. (A) Effect of Akt inhibition on Epac-mediated EC barrier stabilisation. ECs were pre-incubated with Akt inh. VIII (5 μM) or DMSO for 30 min before adding the Epac agonist 8-CPT (200 μM) and albumin permeability was measured. ^∗ vs. control. (B) EC were pre-incubated with Akt inh. VIII (5 μM) or DMSO for 30 min before adding Epac agonist 8-CPT (200 μM) and thrombin (Thr; 0.3 IU/ml). ^∗ vs. control; # vs. Thr alone; n.s.: not significantly different. (C) EC were pre-incubated with isoform specific PI3K inhibitors; PI3Kα (PIK; 0.1 μM) and PI3Kβ (TGX; 1 μM) or DMSO for 30 min before adding the Epac agonist 8-CPT (200 μM) and Thr (0.3 IU/ml). ^∗ vs. control; # vs. Thr alone.

FIGURE 3

Effect of PI3K inhibitor Wortmannin on EC barrier function. (A) ECs were pre-incubated with Wortmannin (0.1 μM) or DMSO for 30 min before adding the Epac agonist 8-CPT (200 μM) and albumin permeability was measured. The arrows indicate when the respective drugs were added. ^∗ vs. control. (B) Effect of the Epac agonist on EC actin cytoskeleton and AJs. Representative immunofluorescence images of F-actin labelled with phalloidin-TRITC and VE-cadherin from three experiments of independent cell preparation. The nuclei were stained with DAPI.

FIGURE 4

Effect of Epac activation on MEK/ERK signalling. (A) Time-dependent effect of the Epac agonist 8-CPT-cAMP (8-CPT) on ERK phosphorylation in ECs. Upper panel: Representative blots of ERK phosphorylation. Lower panels: Densitometric analysis of western blots. n = 3; P < 0.05 for all subfigures; ^∗ vs. control. (B) Effect of MEK/ERK inhibition on Epac-mediated EC barrier stabilisation. EC were pre-incubated with U0126 (10 μM) or DMSO for 30 min before adding the Epac agonist 8-CPT (200 μM) and albumin permeability was measured. ^∗ vs. control; # vs. 8-CPT alone. (C) Effect of MEK-inhibition on Epac-mediated EC AJs stabilisation. ECs were treated with agents as indicated. Where inhibitors were used, these were added 30 min before adding the agonist. Representative immunofluorescence images of VE-cadherin from three experiments of independent cell preparation. The nuclei were stained with DAPI. (D) EC were pre-incubated with MEK inhibitor (10 μM) or DMSO before adding the Epac agonist 8-CPT (200 μM) and thrombin (Thr; 0.3 IU/ml). ^∗ vs. control; # vs. Thr alone; §vs. 8-CPT + Thr.

FIGURE 5

Effect of MEK/ERK inhibition on actin cytoskeleton remodelling. (A) Effect of MEK/ERK inhibition on actin cytoskeleton. Representative immunofluorescence images of actin cytoskeleton stained with phalloidin-TRITC (Green) from three experiments of independent cell preparation. Arrows indicate peripheral actin (Scale bar = 20 μm). Nuclei were stained with DAPI (Red). (B) Effect of MEK inhibition on the Epac agonist-mediated Rac1 activation. Rac1 activity was measured by pulldown assay. Representative western blots of active Rac1 (Rac1-GTP), total Rac1, and pERK1/2. ECs were treated with MEK inhibitor (U0126; 10 μM) or DMSO for 30 min followed by treatment with 8-CPT (200 μM) for 20 min. ^∗ vs. control. (C) Effect of MEK/ERK inhibition on RhoA/Rock signalling and MLC phosphorylation. ECs were treated with thrombin in the presence of 8-CPT, U0126 (10 μM), U0126 plus 8-CPT, or DMSO as indicated. Representative western blots of MYPT1 phosphorylation at T850 (Rock target) and MLC phosphorylation at S18/T19 (MLCK target). GAPDH was used as loading control. Representative blots from three experiments.

FIGURE 6

Role of PI3/Akt and MEK/ERK signalling in Epac-mediated EC survival. (A) Effect of the Epac agonist on EC cell number. ECs were treated with Akt inh. VIII, isoform specific PI3K inhibitors, MEK inhibitor U0126, or DMSO as indicated for 60 min, followed by removal of the medium containing inhibitors and treated with 8-CPT (200 μM) for 24 h and the cells were counted afterwards. ^∗ vs. control; # vs. 8-CPT alone; n.s.: not significantly different. (B) Representative immunofluorescence images of cell proliferation marker Ki67 expression from three experiments of independent cell preparation. Growth arrested ECs were treated with 8-CPT or vehicle for 24 h in the EC medium containing reduced growth factors. The cells were fixed with 4% PFA and immunostained using specific antibody to Ki67. The nuclei were stained with DAPI (Bar = 50 μm). (C) Effect of the Epac agonist on EC caspase 3/7 activity. ECs were treated with Akt inh. VIII, MEK inhibitor U0126, or DMSO as indicated for 60 min in growth factors/serum reduced medium, followed by removal of the medium containing inhibitors and treated with 8-CPT (200 μM) for 24 h in growth factors/serum reduced medium and caspase 3/7 activity was measured using Caspase-Glo kit according to manufacturer’s instructions. ^∗ vs. control; # vs. 8-CPT-cAMP alone.

FIGURE 7

Schematic summary of the data. The scheme summarises the effects of the Epac agonist on EC PI3K/Akt and MEK/ERK signalling in relation to EC barrier function. AJs, Adherens junctions; EC, Endothelial cells; MLC, Myosin light chains; MLCK, MLC kinase; MLCP, MLC phosphatase.