The extracellular regulated kinases (ERK) 1/2 mediate cannabinoid-induced inhibition of gap junctional communication in endothelial cells

Abstract

1. Cannabinoids are potent inhibitors of endothelium-derived hyperpolarizing factor (EDHF)-mediated relaxations. We set out to study the mechanism underlying this effect and the possible role of cannabinoid-induced changes in intercellular gap junction communication. 2. In cultured endothelial cells, Delta(9)-tetrahydrocannabinol (Delta(9)-THC) and the cannabinoid receptor agonist HU210, increased the phosphorylation of extracellular regulated kinases 1/2 (ERK1/2) and inhibited gap junctional communication, as determined by Lucifer Yellow dye transfer and electrical capacity measurements. 3. Delta(9)-THC elicited a pronounced increase in the phosphorylation of connexin 43, which was sensitive to PD98059 and U0126, two inhibitors of ERK1/2 activation. Inhibition of ERK1/2 also prevented the Delta(9)-THC-induced inhibition of gap junctional communication. 4. Delta(9)-THC prevented both the bradykinin-induced hyperpolarization and the nitric oxide and prostacyclin-independent relaxation of pre-contracted rings of porcine coronary artery. These effects were prevented by PD98059 as well as U0126. 5. In the absence of Delta(9)-THC, neither PD98059 nor U0126 affected the NO-mediated relaxation of coronary artery rings but both substances induced a leftward shift in the concentration - relaxation curve to bradykinin when diclofenac and N(omega)nitro-L-arginine were present. Moreover, PD98059 and U0126 prolonged the bradykinin-induced hyperpolarization of porcine coronary arteries, without affecting the magnitude of the response. 6. These results indicate that the cannabinoid-induced activation of ERK1/2, which leads to the phosphorylation of connexin 43 and inhibition of gap junctional communication, may partially account for the Delta(9)-THC-induced inhibition of EDHF-mediated relaxation. Moreover, the activation of ERK1/2 by endothelial cell agonists such as bradykinin, appears to exert a negative feedback inhibition on EDHF-mediated responses.

Figure 1

Effect of cannabinoids on the intercellular coupling between endothelial cells. (A) Original tracing showing electrical capacity measurements in cultured porcine coronary endothelial cells. Capacitative current transients (depicted) were elicited by applying a 10 mV (10 ms) voltage step and recording the resulting current. Experiments were performed in the absence (CTL) and presence of 18α-glycyrrhetinic acid (GA, 30 μM) and Δ⁹-THC (30 μM). Current measurements on a single isolated cell (one cell) are provided as example for complete uncoupling (B). Statistical analysis showing the effect of Δ⁹-THC (30 μM) HU210 (30 μM), 18α-glycyrrhetinic acid (30 μM, aGA) and palmitoic acid (PA, 10 μM) on electrical capacity measurements in cultured porcine coronary endothelial cells. (C) Effects of Δ⁹-THC (30 μM), HU210 (30 μM), 18α-glycyrrhetinic acid (30 μM, aGA) and palmitoic acid (PA, 10 μM) on the transfer of Lucifer Yellow between cultured human umbilical vein endothelial cells. The data shown represent results obtained in at least five independent experiments, compounds were applied 15 min prior to the determination of intercellular coupling, ^**P<0.01, ^***P<0.001 (by ANOVA).

Figure 2

Effect of cannabinoids on the phosphorylation of connexin 43 and ERK 1/2 in human endothelial cells. (A) Autoradiography of connexin 43 (Cx43) immunoprecipitates (top) and Western blot (bottom) showing changes in the incorporation of ³²P into Cx43 following stimulation with Δ⁹-THC (10 μM, 15 min) in the absence and presence of PD98059. (B and C) Western blot analysis showing the time course of the ERK1/2 phosphorylation elicited by either Δ⁹-THC (10 μM, B) or HU210 (10 μM, C). Experiments were performed in the absence and presence of PD98059 (PD, 50 μM) and U0126 (U0, 10 μM); pc denotes positive control for phosphorylated Cx43. Similar results were obtained in three additional experiments.

Figure 3

Involvement of ERK1/2 in the Δ⁹-THC-mediated inhibition of gap junctional communication. (A) Original recordings showing the effect of PD98059 (50 μM) on the Δ⁹-THC-induced inhibition of dye coupling in human endothelial cells. (B and C) Statistical summary of the effects of PD98089 or U0126 on gap junctional communication assessed as either changes in electrical capacity in porcine coronary endothelial cells (B) or changes in dye transfer between human endothelial cells (C). Experiments were performed in the presence and absence of Δ⁹-THC (10 μM). The data shown represent results obtained in at least five independent experiments, ^**P<0.01 (by ANOVA).

Figure 4

Effects of Δ⁹-THC on the resting membrane potential and the bradykinin-induced hyperpolarization of porcine coronary arteries. (A) The resting membrane potential (resting MP, upper panel) and the bradykinin (1 μM)-induced hyperpolarization (ΔmV, lower panel) of porcine coronary smooth muscle cells in the continuous presence of U46619 (0.1 μM), diclofenac (10 μM) and N^ω-nitro-L-arginine (300 μM). Experiments were performed under control conditions (CTL) and in the presence or absence of Δ⁹-THC (10 μM), PD98059 (PD, 50 μM) and U0126 (U, 10 μM) (B and C) Original tracing (B) and statistical analysis (C), showing the effect of preventing MAP kinase activation on the repolarization speed of vascular smooth muscle cells following the bolus application of bradykinin (Bk, 1 μM). The data shown summarize results obtained in five independent experiments, ^*P<0.05 (t-test).

Figure 5

Effects of Δ⁹-THC and 2′,5′-dideoxyadenosine on the EDHF- and NO-mediated relaxation of pre-contracted rings of porcine coronary artery. Concentration – relaxation curves to bradykinin (A) in the absence and presence of Δ⁹-THC (30 μM) and in the absence and presence of PD 98059 (50 μM), (B) in the absence and presence of Δ⁹-THC (30 μM) and in the absence and presence of U0126 (10 μM) (C) in the absence and presence of PD98059 and U0126 and (D) in the absence and presence of 2′,5′-dideoxyadenosine (50 μM). Experiments were performed in the continuous presence of diclofenac (10 μM) and in the absence (C) and presence (A, B and D) of N^ω-nitro-L-arginine (300 μM). The data shown summarize results obtained in 8 – 10 independent experiments, #P<0.05 THC vs THC+inhibitor of ERK1/2 activation; ^*P<0.05 CTL vs CTL+inhibitor of ERK1/2 activation; ANOVA for repeated measurements.