Intermedin (adrenomedullin2) stabilizes the endothelial barrier and antagonizes thrombin-induced barrier failure in endothelial cell monolayers

Abstract

Background and purpose: Intermedin is a member of the calcitonin gene-related-peptide (CGRP) family expressed in endothelial cells and acts via calcitonin receptor-like receptors (CLRs). Here we have analysed the receptors for intermedin and its effect on the endothelial barrier in monolayers of human umbilical vein endothelial cells (HUVECs).

Experimental approach: We analysed the effect of intermedin on albumin permeability, contractile machinery, actin cytoskeleton and VE-cadherin in cultured HUVECs.

Key results: Intermedin concentration-dependently reduced basal endothelial permeability to albumin and antagonized thrombin-induced hyperpermeability. Intermedin was less potent (EC(50) 1.29 ± 0.12 nM) than adrenomedullin (EC(50) 0.24 ± 0.07 nM) in reducing endothelial permeability. These intermedin effects were inhibited by AM(22-52) and higher concentrations of αCGRP(8-37), with pA(2) values of αCGRP(8-37) of 6.4 for both intermedin and adrenomedullin. PCR data showed that HUVEC expressed only the CLR/RAMP2 receptor complex. Intermedin activated cAMP/PKA and cAMP/Epac signalling pathways. Intermedin's effect on permeability was blocked by inhibition of PKA but not of eNOS. Intermedin antagonized thrombin-induced contractile activation, RhoA activation and stress fibre formation. It also induced Rac1 activation, enhanced cell-cell adhesion and antagonized thrombin-induced loss of cell-cell adhesion. Treatment with a specific inhibitor of Rac1 prevented intermedin-mediated barrier stabilization.

Conclusion and implications: Intermedin stabilized endothelial barriers in HUVEC monolayers via CLR/RAMP2 receptors. These effects were mediated via cAMP-mediated inactivation of contractility and strengthening of cell-cell adhesion. These findings identify intermedin as a barrier stabilizing agent and suggest intermedin as a potential treatment for vascular leakage in inflammatory conditions.

Figure 1

Intermedin (IMD) reduces permeability of HUVEC monolayers to macromolecules. (A) HUVEC monolayers were exposed to intermedin (1, 5 and 10 nM) or vehicle (control) as indicated. (B) Concentration-response curve for intermedin in experiments described in A on permeability after 10 min. HUVEC monolayers were exposed to intermedin as in A or vehicle (Control) as indicated. Mean ± SD of three experiments of independent cell preparations, *P < 0.05 significantly different from versus control. (C) Effect of intermedin on thrombin-induced hyperpermeability. HUVEC monolayers were exposed to (1, 5 and 10 nM) in the presence of thrombin (Thr; 0.2 IU·mL⁻¹) or vehicle (Control) as indicated. (D) Concentration-response curve of intermedin in experiments described in C on permeability after 10 min. HUVEC monolayers were exposed to intermedin as in C or vehicle (Control) as indicated. Mean ± SD of three experiments of independent cell preparations; ^#P < 0.05 vs. control, *P < 0.05 significantly different from thrombin alone.

Figure 2

Role of CGRP and AM₁/AM₂ receptors in intermedin-mediated reduction in permeability. (A) Concentration-response curve of αCGRP_8–37 effect on intermedin (IMD)-induced reduction in permeability. HUVEC monolayers were exposed to intermedin (10 nM), in the absence or presence of increasing concentrations of αCGRP_8–37 or vehicle (Control) as indicated. αCGRP_8–37 was added 30 min before addition of intermedin (mean ± SD of three experiments of independent cell preparations). *P < 0.05 significantly different from intermedin alone. (B) Concentration-response curve of αCGRP_8–37 effect on intermedin-mediated reduction in thrombin (Thr)-induced hyperpermeability. HUVEC monolayers were exposed to thrombin (0.2 IU·mL⁻¹) plus intermedin (10 nM), in the absence or presence of increasing concentrations of αCGRP_8–37 as indicated. αCGRP_8–37 was added 30 min before addition of intermedin (mean ± SD of three experiments of independent cell preparations). *P < 0.05 significantly different from intermedin plus thrombin. (C) Concentration-response curve of AM_22–52 effect on intermedin-induced reduction in permeability. HUVEC monolayers were exposed to intermedin (10 nM), in the absence or presence of increasing concentrations of AM_22–52 or vehicle (Control) as indicated. AM_22–52 was added 30 min before addition of intermedin (mean ± SD of three experiments of independent cell preparations). (D) Concentration-response curve of AM_22–52 effect on intermedin-mediated reduction in thrombin-induced hyperpermeability. HUVEC monolayers were exposed to thrombin (0.2 IU·mL⁻¹) plus intermedin (10 nM), in the absence or presence of increasing concentrations of AM_22–52 as indicated. AM_22–52 was added 30 min before addition of intermedin (mean ± SD of three experiments of independent cell preparations). (E) Combined effect of αCGRP_8–37 and AM_22–52 on intermedin-induced reduction in permeability. HUVEC monolayers were treated with intermedin (10 nM) in the absence or presence of αCGRP_8–37 (1 µM), AM_22–52 (10 and 100 nM) or αCGRP_8–37 (1 µM) plus AM_22–52 (10 and 100 nM) as indicated. αCGRP_8–37 and/or AM_22–52 were added 30 min before addition of intermedin [mean ± SD of three experiments of independent cell preparations; *P < 0.05 significantly different from intermedin alone; ^#P < 0.05 significantly different from αCGRP_8–37 plus AM_22–52 (10 nM); n.s., not significantly different from AM_22–52(100 nM)]. (F) Expression of mRNA encoding RAMP1, 2, 3, CLR and β-actin in HUVEC (P0-P2) by RT-PCR. Human testis mRNA was used as positive control. MW; molecular weight marker.

Figure 3

Effect of αCGRP_8–37 and AM_22–52 on intermedin (IMD) and adrenomedullin (AM)-mediated reduction in endothelial permeability. (A) Concentration-response curve of adrenomedullin on HUVEC permeability. HUVEC monolayers were exposed to increasing concentrations of adrenomedullin and the reduction in permeability after 10 min is shown. (B) Concentration-response curve of AM_22–52 on AM-mediated reduction in permeability. HUVEC monolayers were exposed to adrenomedullin (10 nM), in the absence or presence of increasing concentrations of AM_22–52 or vehicle (Control) as indicated. AM_22–52 was added 30 min before addition of adrenomedullin (mean ± SD of five experiments of independent cell preparations; *P < 0.05 significantly different from adrenomedullin alone). (C) Permeability response of HUVEC to intermedin (increasing concentrations) in the absence or presence of αCGRP_8–37 (1 µM) or AM_22–52 (1 µM). (Mean ± SEM of three experiments of independent cell preparations). (D) Permeability response of HUVEC to adrenomedullin (increasing concentrations) in the absence or presence of αCGRP_8–37 (1 µM) or AM_22–52 (1 µM). (Mean ± SEM of three experiments of independent cell preparations).

Figure 4

Intermedin (IMD) activates the cAMP/PKA pathway. (A) Intermedin induces cAMP production in HUVEC. HUVECs were treated with increasing concentrations of intermedin for 10 min and the cAMP concentrations were measured by a colorimetric method. Data are means ± SD of three separate experiments with independent cell preparations. *P < 0.05 significantly different from control. (B) Effect of intermedin on CREB phosphorylation. Representative Western blots of CREB phosphorylation. HUVECs were exposed to increasing concentrations of intermedin for 10 min. The lower band represents phospho-ATF, another target of PKA, which is recognized by the antibody used. (C) Effect of PKA inhibition on intermedin-induced reduction in permeability. HUVEC monolayers were treated with intermedin (10 nM), thrombin (Thr; 0.2 IU·mL⁻¹), PKI (20 µM), intermedin plus thrombin, PKI plus intermedin plus thrombin or vehicle (Control), as indicated. Data are means ± SD of three separate experiments with independent cell preparations. *P < 0.05 significantly different from control; ^#P < 0.05 significantly different from thrombin alone, §P < 0.05 significantly different from intermedin plus thrombin. (D) Effect of intermedin on Rap1 activation. HUVECs were exposed to increasing concentrations of intermedin for 10 min.The Western blots are representative of three separate experiments with independent cell preparations. *P < 0.05 significantly different from control. (E) Effect of eNOS inhibition on intermedin-induced reduction in permeability. HUVEC monolayers were treated with intermedin (10 nM) in the presence or absence of L-NAME (100 µM) and L-NNA (100 µM) or vehicle (Control), as indicated. Data are means ± SD of three separate experiments with independent cell preparations. *P < 0.05 significantly different from control; n.s., not significantly different from intermedin alone.

Figure 5

Effect of intermedin on VE-cadherin and actin cytoskeleton. HUVECs were treated with intermedin (10 nM), thrombin (Thr; 0.2 IU·mL⁻¹), intermedin plus Thr or vehicle (C, control) for 10 min and immunostained for VE-cadherin and F-actin (TRITC-labelled phalloidin). Arrows denote VE-cadherin localized at cell borders and arrowheads denote gaps between cells (scale bar 20 µm; representative immunostaining of five experiments of independent cell preparations).

Figure 6

(A) Effect of intermedin on thrombin (Thr)-induced MLC and MYPT1 phosphorylation. Representative Western blots of MLC and MYPT1 phosphorylation. HUVECs were treated with intermedin (10 nM), thrombin (0.2 IU·mL⁻¹), intermedin plus thrombin, Y27632 (10 µM; a Rock inhibitor was used as positive control) or vehicle (Con, control) for 10 min. In HUVEC, phospho-MLC appears as a double band which indicates two isoforms of the protein and both are phosphorylated in response to thrombin. Y27632 in the concentration used is specific inhibitor of Rock (Ishizaki et al., 2000). (B) Effect of intermedin on thrombin-induced RhoA activation. HUVECs were treated with intermedin (10 nM), thrombin (0.2 IU·mL⁻¹), intermedin plus thrombin or vehicle (Con, control) for 10 min and active RhoA was detected by an ELISA-based assay. The levels of total RhoA in the lysates were analysed by immunoblot and were used to normalize for loading. The active RhoA is given as x-fold of control. (mean ± SD of three experiments of independent cell preparations, *^#P < 0.05). (C) Effect of intermedin on Rac1 activity. Representative Western blots of Rac1-GTP and Rac1 total. HUVECs were treated with intermedin (10 nM), thrombin (0.2 IU·mL⁻¹), intermedin plus thrombin or vehicle (Con, control) for 10 min and active Rac1 was detected by pulldown assay. The active Rac1 is given as x-fold of control. Mean ± SD of three experiments of independent cell preparations, *^#P < 0.05. (D) Effect of the Rac1 inhibitor, NSC23766, on intermedin-mediated reduction of macromolecule permeability. HUVEC monolayers were treated with intermedin (10 nM), thrombin (0.2 IU·mL⁻¹), NSC23766 (10 µM), intermedin plus thrombin, NSC plus intermedin plus thrombin or vehicle (Con), as indicated. Data are means ± SD of three separate experiments with independent cell preparations. *P < 0.05 significantly different from control; ^#P < 0.05 significantly different from. thrombin alone, §P < 0.05 significantly different from intermedin plus thrombin.

Figure 7

Summary of the results presented in this study. Intermedin (IMD), via CLR/RAMP2 receptors, activates cAMP signalling in HUVEC monolayers which on the one hand inactivates RhoA/Rock pathway leading to inactivation of contractile machinery. On the other hand, cAMP signalling activates Rac1, thus mediating adherens junction stabilization. The combined effect of this phenomenon is barrier stabilization. AJ, adherens junctions; MLC: myosin light chain; MLCK, MLC kinase; MLCP, MLC phosphatase.