Methylene Blue Inhibits Cromakalim-Activated K+ Currents in Follicle-Enclosed Oocytes

The effects of methylene blue (MB) on cromakalim-induced K+ currents were investigated in follicle-enclosed Xenopus oocytes. In concentrations ranging from 3–300 μM, MB inhibited K+ currents (IC50: 22.4 μM) activated by cromakalim, which activates KATP channels. MB inhibited cromakalim-activated K+ currents in a noncompetitive and voltage-independent manner. The respective EC50 and slope values for cromakalim-activation of K+ currents were 194 ± 21 µM and 0.91 for controls, and 206 ± 24 µM and 0.87 in the presence of 30 μM MB. The inhibition of cromakalim-induced K+ currents by MB was not altered by pretreatment with the Ca2+ chelator BAPTA, which suggests that MB does not influence Ca2+-activated second messenger pathways. K+ currents mediated through a C-terminally deleted form of Kir6.2 (KirΔC26), which does not contain the sulfonylurea receptor, were still inhibited by MB, indicating direct interaction of MB with the channel-forming Kir6.2 subunit. The binding characteristics of the KATP ligand [3H]glibenclamide are not altered by MB in a concentration range between 1 μM-1 mM, as suggested by radioligand binding assay. The presence of a membrane permeable cGMP analogue (8-Br-cGMP, 100 µM) and a guanylate cyclase activator (BAY 58-2667, 3 µM) did not affect the inhibitory effects of MB, suggesting that MB does not inhibit cromakalim-activated K+ currents through guanylate cyclase. Collectively, these results suggest that MB directly inhibits cromakalim-activated K+ currents in follicular cells of Xenopus oocytes.


Introduction
Methylene Blue (MB), a cationic dye belonging to the phenothiazine-class, is utilized in several clinical conditions, e.g., treatment of methemoglobinemia and malaria, and has been suggested to have beneficial effects in the management of pain, depression, and neurodegenerative diseases [1][2][3]. One of the most established pharmacological actions of MB is its vasoconstrictive effect in various vascular beds. MB induces vasoconstriction in both preclinical [4,5] and clinical studies [6,7]. Thus, vasoconstrictive actions of MB are utilized currently to elevate blood pressure in numerous critical clinical conditions, e.g., cardiovascular and septic shock, and cardioplegia [8]. However, the mechanisms mediating these therapeutic effects remain largely unknown. It is assumed that inhibition of the cGMP pathway is the main mechanism underlying the vasoconstrictive actions of MB. However, several studies suggest that MB has also cGMP-independent direct actions on ion channels [9][10][11], neurotransmitter receptors [12,13], transporters [14], and enzymes [15,16].
K ATP (ATP-sensitive potassium) channels are involved in the regulation of vasoconstriction, heart rhythm and insulin secretion. Low intracellular ATP levels or K ATP channel openers, e.g., cromakalim, pinacidil, diazoxide [17,18], activate K ATP channels; while glibenclamide and other antidiabetic sulfonylureas suppress their activity [19][20][21]. A broad range of diseases, e.g., diabetes, cardiac arrhythmias, hypertension, is related to their dysfunction [22][23][24]. K ATP channels play an important role in controlling vasoconstriction [22,23,25]; it is therefore likely that their modulation may contribute to some of the previously described vasoconstrictive actions of MB.
Xenopus laevis oocytes are surrounded by follicular cells expressing endogenous K ATP channels and are connected to oocytes through gap-junctions [26][27][28]. These channels can be utilized to characterize the influence of pharmacological compounds upon K ATP channels, even though they may not exhibit exactly the same pharmacological characteristics as those of pancreatic β-cells or of vascular beds. In this manuscript, using two-electrode voltage clamp techniques on follicle-enclosed oocytes, we describe how MB modifies the function of K ATP channels.
For inside-out patch experiments, oocytes were chemically defolliculated by collagenase 1A treatment (Sigma, St. Louis, MO, USA; 2 mg/mL, 2 h) and injected with 2 ng cRNA encoding Kir6.2∆C26 mutant. After 2 days of incubation in modified Barth's solution, oocytes were placed in hypertonic solution containing 200 mM K + -aspartate at pH 7.0, and the vitelline layer was removed with sharpened watchmaker's forceps. The pipette (external) solution contained (in mM) 140 KCl, 1.2 MgCl 2 , 2.6 CaCl 2 , and 10 HEPES (pH 7.4). Intracellular (bath) solution contained (in mM) 107 KCl, 10 EGTA, 2 MgCl 2 , 1 CaCl 2 , and 10 HEPES (pH 7.2 with KOH; final K + ∼ =140). Patch pipettes had a resistance of 250-400 kΩ when filled with the pipette solution. Currents were recorded at 20-22 • C from giant inside-out patches using Axopatch 200B amplifier (Axon Instruments Inc., Burligame, CA, USA) at a holding potential of 0 mV, sampled at a rate of 2 kHz and filtered at 1 kHz. Currents were evoked by 2 s voltage ramps from −100 mV to +100 mV at a pulse frequency of 0.5 Hz. In each inside-out patch, the efficacy of 1 mM K 2 ATP to block the K ATP current was tested before applying MB. Leak currents recorded at 10 mM K 2 ATP at the end of the experiments were subtracted from recordings. The recording chamber had a volume of 250 µL and the perfusion rate was 2 mL/min. Due to light sensitivity of MB, experiments were conducted in the dark and MB containers and perfusion lines kept in the dark by aluminum foils.
Statistical significance at the level of 0.05 was analyzed using the Student's t-test, paired t-test or ANOVA. Concentration-response curves were obtained by fitting the data to the logistic equation, where x and y are concentration and response, respectively, E max is the maximal response, EC 50 is the half-maximal concentration, and n is the slope factor. For data analysis, calculations, and fits of the data, the computer software Origin 8.5 (Microcal Software-OriginLab Corp., Northampton, MA, USA) was used. For radioligand binding experiments, follicle-enclosed oocytes were suspended in 300 mL of buffer containing 50 mM HEPES, 300 mM sucrose and 1 mM EDTA at 4 • C on ice as described earlier [32,33]. Oocytes were homogenized using a motorized Teflon homogenizer (six strokes, 15 s each at high speed). This was followed by sequential centrifugations at 1000× g (10 min) and 10,000× g (20 min); each time the pellet was discarded and the supernatant was used for the subsequent step. The final centrifugation was at 60,000× g for 25 min. The microsomal pellet, which contains the membranes of follicular cells [34], was resuspended in 50 mM HEPES buffer and used for the binding studies.
The radioligand binding experiments were carried out at room temperature (20-22 • C) for 1 h. Oocyte membranes were incubated in 1 mL of 50 mM HEPES, pH 7.5, at a protein concentration of 200-500 µg/mL. [ 3 H]glibenclamide was dissolved in ethanol/dimethyl sulphoxide (1:1). For each experiment, freshly made glibenclamide solution was used. For the analysis, calculations, nonlinear curve fitting and regression fits of the radioligand binding data, the computer software Origin 8.5 (Microcal Software-OriginLab Corp., Northampton, MA, USA) was used.

Results
A bath application of 100 µM cromakalim for 5 min changed the mean value for the resting membrane potential from -38 ± 4 mV (mean ± SE, n = 12) to -85 ± 5 mV (n = 12), which is close to the reversal potential for K + in oocytes [35]. In line with earlier investigations [26,27,36], bath application of cromakalim activated a slowly developing outward current ( Figure 1A) and maximal amplitudes of these currents did not change significantly during consecutive administrations of cromakalim every 10 min for up to 80 min. Glibenclamide (1 µM), a specific blocker of K ATP channels, reversibly inhibited cromakalim-activated current (51 ± 4% inhibition; data not shown, n = 7), and the directions of these currents reversed by increasing the external K + concentration to 100 mM (data not shown, n = 4).
The administration of MB (30 µM) for 20 min caused a significant inhibition of the slow-outward current induced by 100 µM cromakalim with incomplete recovery during the washout period ( Figure 1B). Increasing the MB administration time to 30 min (n = 3) did not cause further inhibition, suggesting that the effect of MB reached a steady state within 10 to 20 min. Figure 1C shows concentration-dependent inhibition of cromakalim-induced outward current by MB. The minimum concentration of MB causing a significant inhibition of outward current was 3 µM (11 ± 4% inhibition; n = 4-5; p < 0.05, paired t-test). The IC 50 (a fifty percent of maximal MB inhibition) and slope value (n) were 22.4 µM and 1.1, respectively, and the maximum inhibition was reached at the concentrations of 300 µM and above (84 ± 5% inhibition, n = 5-7).
1.1, respectively, and the maximum inhibition was reached at the concentrations of 300 μM and above (84 ± 5% inhibition, n = 5-7).  The current-voltage (I-V) relationship of cromakalim-induced currents in the absence and presence of MB (30 µM) is presented in Figure 2A. The effect of MB on the cromakaliminduced net outward current (cromakalim-activated current minus resting current at given voltage) did not show voltage dependence i.e., the extent of MB inhibition was not changed significantly in the voltage range studied ( Figure 2B). In addition, in the absence and presence of 30 µM MB, the reversal potentials of the cromakalim-activated currents were -89 ± 4 mV and -92 ± 3 mV, respectively (p > 0.05, paired t-test; n = 5), indicating MB did not alter the ionic selectivity of the K ATP channels.  Cell bodies of oocytes are coupled to follicular cells through gap junctions (for reviews, [28,35]), therefore, MB may affect gap junctions and alter the resistance of the ionic pathways. For this reason, the resistances in follicle-enclosed (to investigate the involvement of R o ) and enzymatically defolliculated oocytes (to investigate the involvements of R j and R f ) in the presence and absence of MB were measured (without involvement of cromakalim-induced conductances in follicular cells). There was no significant change in resistances measured from defolliculated or follicle-enclosed oocytes, in the presence and absence of MB (n = 14-16; Student's t-test, p > 0.05, Figure 2C).
MB has been demonstrated to cause changes in intracellular Ca 2+ homeostasis [37][38][39] and modulate the functions of Ca 2+ -activated K + channels [11,[40][41][42]. Thus, the effects of MB on Ca 2+ -dependent second messenger systems or Ca 2+ -activated K + and/or Cl − channels may interfere with the effect of MB on cromakalim-induced currents. In order to investigate the involvement of intracellular Ca 2+ in the MB effect, follicle-enclosed oocytes were incubated in BAPTA-AM (5 mM) overnight (12-h) and injected with BAPTA (50 nL, 100 mM) 15 min before the recordings to ensure the chelation of intracellular Ca 2+ in both follicular cells and oocytes. Percent inhibitions of cromakalim-induced currents by MB were not significantly different between BAPTA-treated oocytes and controls injected with 50 nL distilled-water ( Figure 2D, p > 0.05, Student's t-test; n = 5). In addition, current-voltage relationships in the absence and presence of MB indicated no significant alterations in reversal potentials in BAPTA-treated oocytes (−89 ± 3 versus −91 ± 4; p > 0.05, Student's t-test; n = 5).
The cromakalim binding site(s) may be involved in MB inhibition of K ATP channels. For this reason, we examined concentration-response curves of cromakalim activation in the absence and presence of MB (30 µM). MB did not cause significant changes in the EC 50 values but inhibited the maximal cromakalim-activated currents (54 ± 5 % of controls; n = 5-6). Respective EC 50 and slope values in the absence and presence of MB (30 µM) were 194 ± 21 µM and 0.91 vs. 206 ± 24 µM and 0.87, suggesting that MB inhibition of cromakalim-induced K + currents occurs in a noncompetitive manner.
The K ATP channel is comprised of four Kir6.2 subunits and each subunit is associated with a larger regulatory sulfonylurea receptor (SUR) subunit (for a review, [43]). Therefore, we investigated the effect of MB on the specific binding of [ 3 H]glibenclamide, a sulfonylurea class drug, in the microsomal fraction of Xenopus oocytes. Figure 3B shows equilibrium curves for the binding of [ 3 H]glibenclamide, in the absence and presence of MB (30 µM). Maximum bindings (B max ) of [ 3 H]glibenclamide for controls and MB-treated membranes were 5.78 ± 0.39 and 5.54 ± 0.41 pmol/mg, respectively (p > 0.05, Student's t-test; n = 5). The affinities (K d ) of [ 3 H]glibenclamide for controls and MB-treated membranes were 1.21 ± 0.14 and 1.32 ± 0.22 nM, respectively. Similarly, the specific binding of [ 3 H]glibenclamide was not altered by the incubation of microsomal membranes with increasing concentrations (1 µM to 1 mM) of MB ( Figure 3C).
The C-terminally truncated form of Kir6.2 (Kir∆C26) lacks the last 26 amino acids but forms functional ATP-sensitive channels in the absence of the SUR subunit [44]. This mutant channel was employed to investigate if MB can act on the channel-forming Kir6.2 subunit in the absence of SUR. The application of MB (30 µM) for 40 s caused a significant inhibition of currents mediated through Kir∆C26 subunits in a voltage-independent manner ( Figure 4A). The reversal potential was not altered in the presence of MB; reversal potentials were 2 ± 3 mV and 0 ± 2 mV in the presence and absence of MB, respectively. MB caused a 48 ± 5% (n = 4) inhibition of controls (at −100 mV), and recovery was incomplete within the time course of the experiments ( Figure 4B). Membranes 2023, 13, x FOR PEER REVIEW 7 of 12   MB is a known inhibitor of soluble guanylate cyclase [45,46], and KATP channels h been shown to be modulated by cGMP [47,48]. We have tested the involvement of guan ate cyclase activity by investigating the effects of BAY 58-2667 (3 µ M), a potent activa MB is a known inhibitor of soluble guanylate cyclase [45,46], and K ATP channels have been shown to be modulated by cGMP [47,48]. We have tested the involvement of guanylate cyclase activity by investigating the effects of BAY 58-2667 (3 µM), a potent activator of soluble guanylate cyclase [49], and 8-Br-cGMP (100 µM), a membrane-permeable cGMP analogue, on the MB inhibition of cromakalim-activated K + currents in follicle-enclosed oocytes. Application of BAY 58-2667 alone or 8-Br-cGMP alone for 20 min. caused 14 ± 4% (p < 0.05, n = 5, Student's t-test) and 16 ± 4% (p < 0.05, n = 6, Student's t-test) potentiation of cromakalim-activated K + currents, respectively. However, after 20 min preincubation with BAY 58-2667 or 8-Br-cAMP, the extent of MB inhibition was not significantly different from MB alone (p > 0.05, n = 5-7, Student's t-test; Figure 4C).

Discussion
Our results indicate that cromakalim-activated K + currents in follicle-enclosed oocytes were inhibited by MB in a non-competitive (with respect to cromakalim and glibenclamide binding sites) and cGMP-independent manner. Follicular cells are electrically coupled to oocytes through gap junctions (for reviews, [28,35]), and, thus, an effect of MB on gap junctions would alter membrane resistance (through oocyte, gap junctions, and follicular cells) and interfere with MB inhibition of K + currents. MB, however, did not cause a significant alteration in the cell input resistances in either follicle-enclosed or defolliculated oocytes, suggesting that, when cromakalim-activated channels are closed, other ionic conductances are not significantly affected by MB in either defolliculated or follicle-enclosed oocytes.
Changes in the intracellular Ca 2+ levels would alter the activities of Ca 2+ -activated Cl − and K + channels and, thus, could interfere with the effect of MB on cromakalim-activated K + currents. MB inhibited the cromakalim-activated K + currents in BAPTA-treated oocytes to the same extent as in untreated oocytes. Similarly, the reversal potential of the cromakaliminduced current was not changed significantly, suggesting that Ca 2+ -activated Cl − conductances are not significantly involved in the effect of MB on K + currents. In addition, since the cromakalim-activated currents were recorded near the reversal potential (−20 mV) for Ca 2+ -activated Cl − channels in oocytes [50], Cl − currents are not likely to interfere with the effect of MB on K ATP channels. In line with earlier investigations [26], the currentvoltage relationship for cromakalim-activated currents was linear within the voltage range studied (−120 to 20 mV); neither the linear characteristics nor the reversal potential for cromakalim-activated K + currents was altered by MB ( Figure 2A) and MB inhibition of cromakalim-activated K + currents was voltage-independent ( Figure 2B).
Analysis of cromakalim concentration-response curves in the presence and absence of MB indicated that MB inhibits K + -currents in a non-competitive manner, suggesting that MB does not act by inhibiting the cromakalim binding site on the channel. Similarly, radioligand binding studies with [ 3 H]glibenclamide, a sulfonylurea-class antidiabetic drug, showed that MB does not significantly affect the glibenclamide binding site in follicle-enclosed oocyte membranes either. The pore of the K ATP channel is formed from four Kir6.2 subunits, each of which is associated with a larger regulatory sulfonylurea receptor (SUR) subunit, which is the primary target for K ATP blockers [43]. Importantly, MB also inhibited ion currents in oocytes expressing the C-terminally deleted mutant of Kir6.2 (Kir∆C26), which forms functional channels but does not contain a sulfonylurea binding site (SUR) [44]. Thus, the results of electrophysiological studies on Kir∆C26 mutant channel and radioligand binding studies suggest that MB does not suppress the function of K ATP channels by inhibiting the known sulfonylurea binding site, but rather by acting at a different site on the K ATP channel. Collectively, these results indicate that neither cromakalim nor sulfonylurea binding sites are involved in the inhibition of K ATP channels by MB.
Similar to our findings, MB has been suggested to have direct inhibitory effects on the functions of K + [11,[40][41][42] and Na + [9] channels and α 7 nicotinic receptors [13] with IC 50 values ranging from 10 to 100 µM (for a review, [3]). MB is a phenothiazine-based molecule, and structurally similar compounds, such as promethazine [51], chlorpromazine [27,52] mefloquine, quinine, quinidine, and quinacrine [53,54], have been shown to directly inhibit K ATP channels. Collectively, these earlier results support our findings suggesting that MB directly inhibits the function of K ATP channels.
Based on our findings, we speculate that some of the cardiovascular effects of MB are mediated by inhibition of K ATP channels. In previous investigations, MB, in concentration ranges used in the present study, was reported to induce depolarizations in various cell types in a cGMP-independent manner [40,[55][56][57]. Depolarization by suppression of K ATP channels has been demonstrated to cause activation of voltage-gated Ca 2+ channels and to increase intracellular Ca 2+ levels, and a subsequent vasoconstriction in vascular beds (for reviews, [17,22]). Therefore, the suppression of K ATP channels may cause depolarizations of the cell, activate voltage-gated Ca 2+ channels, and subsequently contribute to MB-induced vasoconstriction. In conclusion, our results suggest that inhibition of K ATP channels by MB may be one of the mechanisms contributing to the vasoconstrictive effects of this compound. However, the elucidation of the mechanisms underlying MB's vasoconstrictive effects will require further studies analyzing the roles of K ATP channels in vasoconstriction of a specific vascular bed.