Carbenoxolone effectively blocks myoendothelial gap junctions in rat mesenteric arteries

Christopher Garland, Yasuo Kansui, Kim Dora
University of Bath, Bath, United Kingdom

Structurally related derivatives of glycyrrhetinic acid have been employed as gap junction uncoupling agents in the study of endothelium-derived hyperpolarizing factor (EDHF), yet clearly have other effects. For example, the water-soluble derivative carbenoxolone reduced the amplitude of endothelial cell hyperpolarization to ACh in pinned-out rat mesenteric arteries (Tare et al, 2002). This hyperpolarization reflects activation of SKCa and IKCa channels in the endothelium (see Busse et al 2002). However, ACh evoked increases in endothelial cell calcium are not modified by the presence of carbenoxolone (Mather et al, 2005), so we investigated the effect of carbenoxolene on endothelial cell membrane potential change and dye spread from the endothelium in mesenteric arteries under normalized tension, to attempt to resolve this discrepancy.

Male Wistar rats (200-250g) were killed by cervical dislocation and exsanguination. (Animals (Scientific Procedures) Act 1986, schedule 1). A 2 mm segment of mesenteric artery was mounted in a Mulvany-Halpern myograph in oxygenated (95% O2/5% CO2) Kreb’s buffer at 37°C. A small hole cut on the upper surface enabled access to the endothelium. For dye transfer, artery segments were cannulated at either end with a small glass pipette and pressurized to 70 mmHg. Calcein AM (3 μM) was perfused through the lumen for 9 min selectively to load endothelial cells, and allowed to de-esterify and diffuse for 90 min. Fluoresence emitted from endothelial cells and adjacent smooth muscle was assessed from Z-stacks (Olympus FV500 confocal) followed by 3-D reconstruction (Bitplane).

Endothelial cell resting membrane potential (rmp) recorded with sharp glass-microelectrode was –52.4 ± 1.3 mV, n=11. ACh (1μM) or levcromakalim (LVK, 5μM) hyperpolarized the cells by 9.0 ± 1.4 mV and 19.8 ± 2.1 mV (n=4-5), respectively. During incubation with carbenoxolone (100 μM, 60-180 minutes), rmp was –52.0 ± 1.3 mV (n=6) and hyperpolarization to ACh and LVK 8.1 ± 1.7 mV (n=5) and <1 mV (n=3). After washout for 60-120 minutes, rmp was –50.0 ± 1.1 mV (n=6) and hyperpolarization to ACh and LVK 7.7 ± 1.5 mV (n=3) and 19.7 ± 2.0 mV (n=3). Luminal perfusion with calcein AM uniformly loaded endothelial cells, and slowly stained the adjacent smooth muscle. In the presence of 100μM carbenoxolone, endothelial fluorescence, including endothelial cell projections to the muscle layers, was higher and smooth muscle staining negligible.

In summary, carbenoxolene did not modify the endothelial cell rmp or hyperpolarization to ACh, but reversibly abolished the hyperpolarization to LVK which acts only on the smooth muscle. Together with the block of dye transfer to the smooth muscle and the previous report of unchanged endothelial cell calcium increases with ACh, these data indicate that carbenoxolene reversibly blocks vascular gap junctions, including myoendothelial gap junctions.

Busse et al (2002) Trends Pharmacol. Sci. 23: 374-380
Mather et al (2005) Circ. Res. 97: 399-407
Tare et al (2002) Am. J. Physiol 282: H335-H341



Supported by the Wellcome Trust