Berberine prolongs the duration of cardiac action potentials without affecting resting membrane potential or action potential amplitude. Controversy exists regarding whether berberine exerts this action by preferential block of different components of the delayed rectifying potassium current, I(Kr) and I(Ks). Here we have studied the effects of berberine on hERG (I(Kr)) and KCNQ1/KCNE1 (I(Ks)) channels expressed in HEK-293 cells and Xenopus oocytes. In HEK-293 cells, the IC50 for berberine was 3.1 +/- 0.5 microM on hERG compared with 11 +/- 4% decreases on KCNQ1/KCNE1 channels by 100 microM berberine. Likewise in oocytes, hERG channels were more sensitive to block by berberine (IC50 = 80 +/- 5 microM) compared with KCNQ1/KCNE1 channels (approximately 20% block at 300 microM). hERG block was markedly increased by membrane depolarization. Mutation to Ala of Y652 or F656 located on the S6 domain, or V625 located at the base of the pore helix of hERG decreased sensitivity to block by berberine. An inactivation-deficient mutant hERG channel (G628C/S631C) was also blocked by berberine. Together these findings indicate that berberine preferentially blocks the open state of hERG channels by interacting with specific residues that were previously reported to be important for binding of more potent antagonists.

Large-conductance Ca2+-activated K+ channels (BK) are believed to underlie interburst intervals and contribute to the control of hormone release in several secretory cells. In crustacean neurosecretory cells, Ca2+ entry associated with electrical activity could act as a modulator of membrane K+ conductance. Therefore we studied the contribution of BK channels to the macroscopic outward current in the X-organ of crayfish, and their participation in electrophysiological activity, as well as their sensitivity toward intracellular Ca2+, ATP, and voltage, by using the patch-clamp technique. The BK channels had a conductance of 223 pS and rectified inwardly in symmetrical K+. These channels were highly selective to K+ ions; potassium permeability (PK) value was 2.3 x 10(-13) cm(3) s(-1). The BK channels were sensitive to internal Ca2+ concentration, voltage dependent, and activated by intracellular MgATP. Voltage sensitivity (k) was approximately 13 mV, and the half-activation membrane potentials depended on the internal Ca2+ concentration. Calcium ions (0.3-3 microM) applied to the internal membrane surface caused an enhancement of the channel activity. This activation of BK channels by internal calcium had a KD(0) of 0.22 microM and was probably due to the binding of only one or two Ca2+ ions to the channel. Addition of MgATP (0.01-3 mM) to the internal solution increased steady state-open probability. The dissociation constant for MgATP (KD) was 119 microM, and the Hill coefficient (h) was 0.6, according to the Hill analysis. Ca2+-activated K+ currents recorded from whole cells were suppressed by either adding Cd2+ (0.4 mM) or removing Ca2+ ions from the external solution. TEA (1 mM) or charybdotoxin (100 nM) blocked these currents. Our results showed that both BK and K(ATP) channels are present in the same cell. Even when BK and K(ATP) channels were voltage dependent and modulated by internal Ca2+ and ATP, the profile of sensitivity was quite different for each kind of channel. It is tempting to suggest that BK and KATP channels contribute independently to the regulation of spontaneous discharge patterns in crayfish neurosecretory cells.

Effects of membrane potential, intracellular Ca2+ and adenine nucleotides on glucose-sensitive channels from X organ (XO) neurons of the crayfish were studied in excised inside-out patches. Glucose- sensitive channels were selective to K+ ions; the unitary conductance was 112 pS in symmetrical K+, and the K+ permeability (PK) was 1.3 x 10(-13) cm x s(-1). An inward rectification was observed when intracellular K+ was reduced. Using a quasi-physiological K+ gradient, a non-linear K+ current/voltage relationship was found showing an outward rectification and a slope conductance of 51 pS. The open-state probability (Po) increased with membrane depolarization as a result of an enhancement of the mean open time and a shortening of the longer period of closures. In quasi-physio- logical K+ concentrations, the channel was activated from a threshold of about -60 mV, and the activation midpoint was -2 mV. Po decreased noticeably at 50 microM internal adenosine 5'-triphosphate (ATP), and single-channel activity was totally abolished at 1 mM ATP. Hill analysis shows that this inhibition was the result of simultaneous binding of two ATP molecules to the channel, and the half-blocking concentration of ATP was 174 microM. Internal application of 5'-adenylylimidodiphosphate (AMP-PNP) as well as glibenclamide also decreased Po. By contrast, the application of internal ADP (0.1 to 2 mM) activated this channel. An optimal range of internal free Ca2+ ions (0.1 to 10 microM) was required for the activation of this channel. The glucose--sensitive K+ channel of XO neurons could be considered as a subtype of ATP-sensitive K+ channel, contributing substantially to macroscopic outward current.