Evidence for Potentiation of M-Type Potassium Current by Flavonoid Corylin (3-(2,2-dimethylchromen-6-yl)-7-hydroxychromen-4-one)

1. Introduction

Corylin (3-(2,2-dimethylchromen-6-yl)-7-hydroxychromen-4-one) is a major bioactive flavonoid isolated from the fruit of Psoralea corylifolia Linne (Fabaceae), also known as buguchi or Bo-Gol-Zhee. It has been increasingly demonstrated to exert a wide range of pharmacological actions, including direct free radical scavenging, inhibition of biomolecules, suppression of lipopolysaccharide-induced inflammation, and modulation of antioxidant defense [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]. This compound might also have potential in treating brain inflammation and attenuating the progression of neurodegeneration. For example, the extract from the seeds of Psoralea corylifolia has been demonstrated to be against palmitate-induced neuronal apoptosis in PC12 cells [21] and to exert neuroprotective and anti-neuroinflammatory effect in hippocampal cells, microglia and retinal ganglion cells [4,22]. Notably, in addition to stimulating L-type Ca²⁺ currents [23,24], quercetin—another flavonoid—has been reported to enhance the activity of M-type K⁺ channels [25,26,27]. Several botanical medicines have also been recently shown to activate KCNQ channels [26,27]. However, the extent to which corylin or other related compounds can alter the magnitude, gating properties, and voltage-dependent hysteresis (Hys_(V)) of plasmalemmal ionic currents in excitable cells remains largely unresolved.

The KCNQ (K_V7) family of K⁺ channels consists of five members, designated KCNQ1-KCNQ5 (K_V7.1-K_V7.5). Among these, KCNQ2, KCNQ3, and KCNQ5 encode the principal subunits of K_V7.2, K_V7.3, and K_V7.5 channels, respectively, which are broadly expressed in both nervous and endocrine tissues. Notably, heteromeric KCNQ2/KCNQ3 channels closely replicate the native M-type K⁺ current (I_K(M)), sharing its biophysical properties and sensitivity to pharmacological inhibitors such as linopirdine. The designation “M-type” reflects the current’s regulation by muscarinic acetylcholine receptors [27,28,29]. The magnitude of these currents can widely regulate membrane excitability in varying types of excitable cells that include endocrine cells [28,29,30]. Once activated by membrane depolarization, they exhibit slow activation and deactivation kinetics [31,32,33,34]. It has also been shown that by enhancing Na⁺-current recovery, the I_K(M) amplitude induced by high frequency stimulation can expedite action potential (AP) firing with stable waveforms and reliable synaptic transmission [34,35]. Moreover, pharmacological targeting of I_K(M) (or KCNQ/K_V7 channel-mediated currents) has been recently recognized as a potential adjunctive strategy for a range of neurological disorders characterized by neuronal hyperexcitability, including cognitive dysfunction, epilepsy, major depression, and neuropathic pain [29,35,36,37].

Building on these considerations, the present study investigated whether and how corylin modulates the magnitude, gating kinetics, and Hys_(V) behavior of I_K(M) in pituitary tumor (GH₃) cells. Importantly, our findings demonstrate that, beyond its previously described anti-inflammatory, anti-neoplastic, and anti-oxidative actions, corylin can interact with K_M (KCNQ or K_V7) channels to enhance I_K(M) in a concentration-, time-, voltage-, and Hys_(V)-dependent manner in excitable cells such as GH₃ lactotrophs.

2. Results

2.1. Stimulatory Effect of Corylin on M-Type K⁺ Current (I_K(M)) Recorded from Pituitary GH₃ Cells

In the initial set of experiments, we examined whether I_K(M) in GH₃ cells can be modified by corylin. Cells were bathed in a high-K⁺, Ca²⁺-free solution containing 1 μM tetrodotoxin (TTX) and 0.5 mM CdCl₂, while the recording pipette was filled with a K⁺-based internal solution. After establishing whole-cell current recordings, a 1-sec depolarizing voltage pulse from −50 to −10 mV was applied to evoke inward I_K(M), which displayed a slowly activating, non-inactivating time course, consistent with previous reports [26,31,32,34,38,39,40]. However, of additional interest, as cells were exposed to corylin, the amplitude of I_K(M) activated by long-lasting step depolarization became progressively increased (Figure 1A). For example, the I_K(M) during the exposure to 10 μM corylin evidently arose from 194 ± 25 to 348 ± 31 pA (n = 8, P < 0.05). Following the removal of corylin, the current amplitude returned to 202 ± 26 pA (n = 8). Concurrently, the value of activation time constant (τ_act) of I_K(M) became shortened, as evidenced by a significant reduction in τ_act from 144.6 ± 9.6 to 99.7 ± 7.4 msec (n = 8, P < 0.05) by adding 10 μM corylin (Figure 1B,C). Furthermore, as cells were continually exposed to 10 μM corylin, subsequent addition of 10 μM linopirdine could attenuate current amplitude as well decrease τ_act value effectively (Figure 1C). Linopirdine was reported to be an inhibitor of I_K(M) [31,34,40,41]. These results suggest that corylin is capable of modulating the activation kinetics of I_K(M).

We subsequently established the concentration-dependent relationship of corylin-induced stimulation of I_K(M) and the findings are presented in Figure 1D. Notably, this compound can increase the amplitude of I_K(M) in a concentration-dependent manner. Based on a least-squares fit to the modified equation (as indicated in Materials and Methods), the data yielded a half-maximal effective concentration (EC₅₀) for I_K(M) stimulation of 3.8 μM and a Hill coefficient of 1.2. The results lead us to indicate that the addition of corylin would exert a stimulatory effect on I_K(M) in these cells.

2.2. Effect of Corylin on the Steady-State Current Versus Voltage (I-V) Relation and Activation Curve of I_K(M)

Next, we wanted to test if I_K(M) activated by different levels of membrane potential can be altered by the existence of corylin. As illustrated in Figure 2A,B, when a series of voltage steps ranging from −60 to −10 mV was applied to the test cells from a holding potential of −50 mV, the absolute amplitude of I_K(M) increased progressively, particularly at membrane potentials more depolarized to −40 mV. In addition, the quasi-steady-state activation curve of I_K(M), in the absence and presence of corylin, was constructed and the results are presented in Figure 2C. The relationship between membrane potential and normalized I_K(M) amplitude was also fitted with a Boltzmann function (see Materials and Methods), and the fit was evaluated using goodness-of-fit analysis. In control (i.e., corylin was not present), V_1/2 = −25.1 ± 1.5 mV and q = 4.9 ± 0.2 e (n = 8), while during cell exposure to 10 μM corylin, V_1/2 = −31.8 ± 1.7 mV and q = 5.1 ± 0.2 e (n = 8). Moreover, corylin increased the absolute amplitude of the deactivating tail I_K(M). For instance, during membrane depolarization from −50 to −10 mV, the tail I_K(M) measured upon repolarization to −80 mV increased from 252 ± 14 to 321 ± 19 pA (n = 8, P < 0.05). The results indicated that, in the presence of corylin, the quasi-steady-state I_K(M) activation curve was shifted leftward (toward more hyperpolarized potentials) by approximately 7 mV in the presence of corylin, without any noticeable change in the gating charge associated with channel activation.

2.3. Corylin Effect on I_K(M) Amplitude Evoked During a Train of Depolarizing Command Voltages in GH₃ Cells

Recent work has demonstrated the effectiveness of the train of depolarizing pulses in modifying the I_K(M) magnitude [26,34,35]. For this reason, we continued to explore whether corylin-mediated stimulation of I_K(M) in these cells can be modified during pulse-train stimulation. In this set of whole-cell current measurements, we applied a 20 Hz train of depolarizing pulses from −50 to −10 mV to the tested cell. As demonstrated in Figure 3A,B, the current activation and deactivation evoked by responding to such pulse-train stimulation was robustly observed. Furthermore, exposure of cells to corylin increased I_K(M) during a train of depolarizing pulse (Figure 3C). Concomitantly, the time course of current activation became faster in the presence of corylin. For instance, exposure to 10 μM corylin significantly increased the absolute amplitude of I_K(M) measured at the end of pulse-train stimulation, rising from 254 ± 21 to 369 ± 28 pA (n = 7, P < 0.05). In parallel, the amplitude of deactivating I_K(M) was markedly elevated from 623 ± 43 to 983 pA (n = 7, P < 0.05). Additionally, the τ_act value for I_K(M) during a train of command voltages was measurably shortened to 53 ± 9 msec (n = 7, P < 0.05) from a control value of 114 ± 14 msec (n = 7). It is therefore clear that exposure to corylin produced a considerable increase in the amplitude of I_K(M), along with a reduction in the τ_act value during a train of depolarizing pulses.

2.4. Augmentation of the Strength in Voltage-Dependent Hysteresis (Hys_(V)) of I_K(M) Caused by Corylin

It has been demonstrated that V_ramp-induced Hys_(V) of I_K(M) could contribute to AP firing in various types of excitable cells [26,32,34]. Hys_(V) refers to a pronounced lag in current magnitude when the membrane potential is linearly ramped in the opposite direction. Accordingly, the experiments were designed to determine whether V_ramp-induced Hys_(V) was functionally active in GH₃ cells and to evaluate how cell exposure to corylin may influence the Hys_(V)’s strength in I_K(M). In whole-cell voltage-clamp recordings, the cell was held at −50 mV, after which a series of double triangular V_ramp pulses (ranging between −60 and 0 mV, 3.6 sec in duration, delivered at 0.05 Hz) was applied via a digital-to-analog conversion. In accordance with previous observations [26,32,34], when a double V_ramp was applied to the examined cells, the current amplitude at a given membrane potential evoked during the ascending (forward or upsloping) limb of V_ramp was markedly smaller than that measured at the same voltage during the descending (backward or downsloping) limb of the voltage (Figure 4A). For example, in the control period (i.e., in the absence of corylin), the absolute amplitude of I_K(M) at −20 mV differed significantly between the ascending (upsloping) and descending (downsloping) phases of V_ramp, measuring 198 ± 22 and 418 ± 28 pA (n = 8, P < 0.05). These findings strongly indicate the presence of I_K(M)’s Hys_(V) in response to an upright isosceles-triangular V_ramp in GH₃ cells [34].

Of additional notice, as GH₃ cells were continually exposed to corylin, the Hys_(V) strength of I_K(M) became accentuated. For example, in the presence of 10 μM corylin, I_K(M) amplitude (in absolute value) at the level of −20 mV evoked during the ascending or descending end of V_ramp became measurably raised to 218 ± 23 pA (n = 8, P < 0.05) or 682 ± 38 pA (n = 8, P < 0.05), respectively. It was therefore observed that the current magnitude at the descending end of V_ramp increased more significantly than that at the ascending phase of V_ramp. We then quantify the strength of Hys_(V) loop by estimating the total area (∆area, shaded region) which encircles the current amplitude between the ascending and descending end of V_ramp. The experimental data were compiled and are summarized in Figure 4B. These observations clearly demonstrate that corylin (3 or 10 μM) increased Hys_(V)’s ∆area of I_K(M) in GH₃ cells. Additionally, during continued presence of corylin (10 μM), the subsequent addition of linopirdine (10 μM) was able to attenuate corylin-increased ∆area of I_K(M) evoked during isosceles-triangular V_ramp. Therefore, it is amenable to reflect that the corylin presence can increase I_K(M) in a concentration- and Hys_(V)-dependent fashion in these cells.

2.5. Comparison Among Effect of Corylin, Corylin Plus Carvedilol (Carv), Corylin Plus Iberiotoxin (Iber), 17β-Estradiol, and Corylin Plus Dapagliflozin (Dapa) on I_K(M) Amplitude Observed in GH₃ Cells

Recent reports have demonstrated the ability of corylin to bind to β₃-adrenergic receptors in adipocytes [11,18,42]. The induction of osteoblastic differentiation caused by corylin was mediated through its binding to and interaction with estrogen receptors [18,42]. Pituitary lactotrophs have been previously reported to express the activity of estrogen receptors [43]. It would thus be important to examine whether corylin-stimulated I_K(M) presented herein could be a result of its binding to β-adrenergic or estrogen receptors. Under our experimental conditions, as cells were continually exposed to 10 μM corylin, the addition of neither carvedilol (Carv, 10 μM) nor iberiotoxin (200 nM) was able to have any modifications on corylin-stimulated I_K(M), as summarized in Figure 5. Carvedilol, a non-selective β-adrenergic blocker, was shown previously to antagonize the activity of β₃-adrenergic receptors expressed in cardiac tissue [44]. It is thus thought that carvedilol might block I_K(M)-stimulating effects of corylin via β-adrenergic receptors. Iberiotoxin was an inhibitor of large-conductance Ca²⁺-activated K⁺ channels [45]. Moreover, the addition of 17β-estradiol (10 μM) was found to have no effect on I_K(M). In the continued presence of 10 μM corylin, subsequent addition of dapagliflozin (10 μM) could attenuated corylin-induced stimulation of I_K(M). Dapagliflozin (Dapa), an inhibitor of Na⁺-dependent glucose co-transporters, has been reported to suppress I_K(M) directly [39]. Based on the present observations, it is therefore unlikely that the corylin-stimulated I_K(M) observed in GH₃ cells results primarily from its binding to β-adrenergic or estrogen receptors.

2.6. Effect of Corylin on M-type K⁺ (K_M) Channels Recorded from GH₃ Cells

The corylin-induced enhancement of whole-cell I_K(M) may result from several mechanisms. Specifically, alterations in channel open probability, single-channel conducance, gating kinetics (e.g., mean open time), or a combination of these factors could underlie the observed stimulation of I_K(M). To further investigate single-channel activity in K_M channels in the presence or absence of corylin, additional measurements were performed. In these cell-attached recordings, cells were maintained in a high-K⁺, Ca²⁺-free solution, while the recording electrode was filled with a low-K⁺ (5.4 mM) solution. As illustrated in Figure 6, when the tested cell was voltage-clamped at +20 mV relative to the bath, the activity of single K_M channels was robustly observed [31,39,40]. When corylin was applied to the bath, the channel open probability became progressively increased. In continued presence of corylin, further addition of linopirdine attenuated corylin-induced increase of channel activity. For example, the presence of 10 μM corylin conceivably increased the probability of K_M-channel openings from 0.023 ± 0.004 to 0.041 ± 0.006 (n = 8, P < 0.05); however, no modification in the single-channel amplitude was found (3.3 ± 0.2 pA [control] versus 3.4 ± 0.3 pA [in the presence of corylin]; n = 8, P > 0.05).

We further examined and analyzed the kinetic properties of K_M channels obtained with or without addition of corylin. As demonstrated in Figure 6B, the distribution of open durations was least-squares fitted by a single exponential. The mean open time of K_M channels arose from 1.5 ± 0.2 to 2.8 ± 0.3 msec (n = 8, P < 0.05). The corylin-induced modulation of K_M-channel activity likely reflects an increase in channel open duration rather than any change in single-channel amplitude.

Moreover, in the continued presence of 10 μM corylin, further addition of linopirdine (10 μM) or dapagliflozin (10 μM) was able to attenuate corylin-enhanced channel activity effectively (Figure 6C). Linopirdine can suppress the activity of K_M channels, while dapagliflozin, known to be an inhibitor of Na⁺-dependent glucose co-transporter, was reported to inhibit I_K(M) amplitude [34,39]. Therefore, corylin-stimulated I_K(M) could be reasonably explained by the increased channel open probability as well as by its prolongation in mean open time of K_M channels.

2.7. Effect of Corylin on erg-Mediated K⁺ Current (I_K(erg)) Recorded from GH₃ Cells

In another set of measurements, we attempted to examine if another types of K⁺ current (e.g., I_K(erg)) would be sensitive to any perturbations by the presence of corylin. To measure I_K(erg) [39,46], we put GH₃ cells in high-K⁺, Ca²⁺-free solution, and we then filled up the recording pipette with K⁺-enriched solution. As the whole-cell configuration was established, we held the examined cell at −10 mV and a series of command voltage steps ranging between −90 and 0 mV with varying durations was imposed on it. As shown in Figure 7, under cell exposure to 10 μM corylin, the I_K(erg) amplitudes measured throughout the entire voltage-clamp voltages imposed were conceivably reduced. For example, upon exposure to 10 μM corylin, the absolute peak amplitude of deactivating I_K(M), elicited by membrane hyperpolarization from −10 to −90 mV, decreased from 729 ± 70 to 607 ± 38 pA (n = 8, P < 0.05). After washout of the compound, current amplitude was returned to 724 ± 68 pA (n = 8). Therefore, unlike its stimulation of I_K(M), the I_K(erg) observed in GH₃ cells [30,39,46] was susceptible to being mildly inhibited by adding corylin.

2.8. Docking Results of the Molecular Interactions Between Corylin and the KCNQ2 or KCNH2 Channel

In this study, we extended to examine how the protein of KCNQ2 could be auto-docked by corylin through PyRx software (https://pyrx.sourceforge.io, accessed on 27 March 2026/. The predicted binding sites of corylin are presented in Figure 8. It needs to be emphasized that corylin can form hydrophobic contact with several amino acid residues, including Thr276, Thr277, Val302, Ala306, Ala309, Gly310 and Gly313, while it forms hydrogen bonds with residue Ser 314 with the distance of 2.70 and 2.76 Å, and that the estimated binding affinity among this interaction was −8.4 kcal/mol. In keeping with the experimental observations made above, these results can thus be interpreted to mean that the corylin molecule may bind to the intracellular domain adjacent to transmembrane segment of the channel (S6 region). The activity of K_M (KCNQ or K_V7) channels occurring in excitable cells is thus expected to confer the susceptibility to modifications by corylin or its structurally similar compounds [26,27].

Because corylin can inhibit the amplitude of I_K(erg)_, the KCNH2 (HERG, human ether-à-go-go-related gene) protein was further docked with corylin using PyRx software. The predicted binding sides of corylin on this channel protein are illustrated in Figure 9. This compound was observed to establish hydrophobic interactions with several residues, including Val3(A), His402(A), Val476(A), and Ala478(C). The corylin molecule can form hydrogen bonds with residues Arg4(A), Lys407(A), Asp411(A), and Arg541(A), with bond lengths of 2.81, 3.12, 3.01, and 2.85 Å, respectively. The results estimate a strong binding affinity of −7.8 kcal/mol. The predicted interaction thus suggests that corylin-mediated inhibition of I_K(erg) in GH₃ cells is due to its binding to KCNH2 channels.

3. Discussion

The striking findings demonstrated in this work are that corylin, a bioactive flavonoid currently recognized as a potential life prolonging agent [47], produces a stimulatory action on I_K(M) in a concentration, voltage-, and Hys_(V)-dependent fashion in GH₃ lactotrophs. A leftward shift in the steady-state activation curve of I_K(M) was observed in the presence of this compound with no change in the gating charge of the curve. Cell exposure to it can elevate the probability of K_M channels that would be open, in combination with a measurable lengthening in mean open time of the channel. However, the I_K(erg) in GH₃ cells was slightly suppressed by the presence of corylin. Docking analysis revealed specific atomic-level interactions between the corylin molecule and the KCNQ2 or KCNH2 channel structure.

The magnitude of Na⁺ currents can rapidly decline in an exponential manner during high-frequency stimulation, as previously demonstrated in GH₃ cells [48]. However, it needs to be emphasized that because of its slow activation and deactivation kinetics, the I_K(M) amplitude can progressively arise during repetitive firing of APs. During high-frequency stimulation, the accumulation of I_K(M) is hence allowed to hyperpolarize the afterpotential and hence to speed the recovery of Na⁺ channels from inactivation. As a corollary, the augmentation of I_K(M) magnitude caused by corylin during high-frequency activity is of particular significance and thus capable of facilitating the firing of neuronal APs with stable waveform and high-fidelity synaptic signaling [26,34,35].

Like the Hys_(V) behavior existing in solar cells [34,49], the current investigations clearly observed the appearance of the overall behavior in I_K(M)’s Hys_(V) evoked by a long-lasting upright isosceles-triangular V_ramp [30,32,39]. That is, the magnitude of these instantaneous currents measured between ascending and descending ends of double V_ramp turned out to be strikingly distinguishable. Alternatively, as the membrane potential becomes depolarized (i.e., upward ramp of triangular V_ramp), the voltage dependence of K_M channels may shift the mode of Hys_(V) to one which occurs at less negative potentials with smaller current magnitude, leading to a minor effect on the rising phase of AP. However, as the membrane potential is hyperpolarized (i.e., during the downward limb of V_ramp or repolarizing phase of AP), the voltage dependence of I_K(M) activation would switch to more hyperpolarized voltages with a higher current magnitude, thereby having the tendency to increase membrane repolarization as well as to increase recovery of Na⁺ currents. Furthermore, findings from these observations led us to unravel that the triangular V_ramp-induced I_K(M) did undergo striking Hys_(V) change in the voltage dependence, and such Hys_(V) loops were subjected to being accentuated by adding corylin. In other words, the ∆area (indicated in shaded area of Figure 4A) of I_K(M) loop evoked in response to long-lasing triangular V_ramp significantly arose following the application of corylin. As such, the existence of corylin may increase I_K(M) magnitude in a concentration- and Hys_(V)-dependent fashion. However, further work needs to be conducted to examine if corylin-perturbed modifications in Hys_(V) behavior of I_K(M) are tightly linked to conformational changes or docking interactions in the voltage sensors of the K_M channel.

Earlier investigations have revealed the ability of corylin to bind to and then to activate β ₃-adrenergic receptors present in adipocytes [11,42]. It has been also demonstrated that corylin might interact with estrogen receptors to induce osteoblastic differentiation [18,50]. Estrogen receptors was noticed to express in pituitary lactotrophs [43]. However, in our study, during the continued exposure to corylin, further application of carvedilol, known to block β₃ adrenoceptors in cardiac tissue, failed to have any adjustments on corylin-stimulated I_K(M). Moreover, the presence of 17β-estradiol alone did not cause any pertubations on I_K(M) observed in GH₃ cells. In this scenario, it appears unlikely that under our experimental conditions, corylin-mediated stimulation of I_K(M) or K_M-channel activity is attributed to its high-affinity binding to either β-adrenergic or estrogen receptors

The EC₅₀ of corylin required to stimulate I_K(M in GH₃ cells was determined to be approximately 3.8 μM. This value aligns closely with the concentration rages (1-300 μM) previously reported for its anti-oxidative, anti-inflammatory, and antineoplastic effects [4,7,8,9,11,12,15,16,47,51,52]. It should also be noted that the perturbations of corylin on membrane excitability may be strongly influenced by several confounding factors, including the concentration of corylin, the baseline resting potential, the firing patterns of APs, or a combination of these variables. It is, therefore, anticipated that the K_M channel is an important target for the action of corylin. The concentrations used to affect magnitude, gating kinetics and Hys_(V) behaviors of I_K(M) presented herein would be pharmacologically significant in body fluids and tissues. The corylin molecule may have the propensity to exercise a higher affinity to the open state than to the resting (closed) state in the K_M channel, thereby de-stabilizing the open conformation, while the detailed ionic mechanism of corylin actions on ionic currents is not thoroughly understood.

A previous paper [53] reported that the single-channel amplitude of K_M channels was lower than that observed in the present study. The reason for this discrepancy remains unclear. One possible explanation is that single-channel conductance of K_M channels may vary among different tissue preparations. Our findings are consistent with those reported earlier studies [31,38,54]. Because single-channel conductance of KCNQ2 and KCNQ3 channels is higher than that of KCNQ4 and KCNQ5 channels. It remains to be determined whether corylin differentially regulates distinct populations of K_M (KCNQ/K₇) channels.

4. Materials and Methods

4.1. Chemicals, Drugs, Reagents and Solutions

Corylin (IUPAC name: 3-(2,2-dimethylchromen-6-yl)-7-hydroxychromen-4-one, 53947-92-5, SCHEMBL1096083, CHEMBL1271888, C₂₀H₁₆O₄, CAS: 53947-92-5; PubChem CID: 5316097) was supplied by MedChemExpress (GeneChain, Kaohsiung, Taiwan). Iberiotoxin was purchased from Alomone Labs (Asia Bioscience, Taipei, Taiwan), dapagliflozin (Dapa) was from Cayman (Ann Arbo, MI), carvedilol (Carv) was from Tocris (Union Biomed, Taipei, Taiwan), while 17β-estradiol, linopirdine (Lino), and tetrodotoxin (TTX) were from Sigma-Aldrich (Merck, Taipei, Taiwan). Corylin, carvedilol, dapagliflozin, and linopirdine were dissolved in dimethyl sulfoxide (DMSO) as 20 mM stock solution, and there were thereafter diluted in extracellular solution to the final concentration achieved, while iberiotoxin was dissolved in 0.9% NaCl. For cell preparations, all culture media, horse and fetal calf sera, L-glutamine, and trypsin/EDTA were acquired from HyClone^TM (Thermo Fisher, Logan, UT), while other chemicals or reagents were of laboratory grade and taken from standard sources.

The extracellular solution (normal Tyrode’s solution buffered with HEPES) contained the following ionic composition (in mM): NaCl 136.5, KCl 5.4, MgCl₂ 0.53, CaCl₂ 1.8, glucose 5.5, and HEPES 5.5, adjusted to pH 7.4 with NaOH. For recording of macroscopic K⁺ currents (I_K(M) or I_K(erg)), the pipette solution consisted of (in mM): KCl 140, MgCl₂ 1, Na₂ATP 4, Na₂GTP 0.1, EGTA 0.1, and HEPES, titrated to pH 7.2 with KOH. To measure I_K(M), I_K(erg), or K_M-channel activity, the bath solution contained a high K⁺ solution (in mM): KCl 130, NaCl 10, MgCl₂ 3, glucose 6, and HEPES 10, adjusted to pH 7.4 with KOH. To record the activity of single K_M channel, the pipette solution was composed of the following (in mM): NaCl 136.5, KCl 5.4, MgCl₂ 0.53, and HEPES-NaOH buffer 5 (pH 7.4).

4.2. Cell Preparations

GH₃ pituitary tumor cells (BCRC-60015; Bioresources Collection and Research Center, Hsinchu, Taiwan) were cultured in Ham’s F-12 media supplemented with 15% horse serum (v/v), 2.5% fetal calf serum (v/v), and 2 mM L-glutamine [23,48,55]. To induce differentiation, cells were transferred to a serum-free, Ca²⁺-free medium. Under these experimental conditions, cell viability typically remained at 80-90% for up to two weeks. Cultures were maintained at 37 °C in a humidified incubator with a CO₂/air mixture (1:19).

4.3. Electrophysiological Measurements

Before each experiment, GH₃ cells were carefully dispersed with 1% trypsin/EDTA solution, and we thereafter quickly put an aliquot of cells suspension in a recording chamber mounted on the stage of a CKX-41 inverted microscope (Olympus; Yuan Yu, Taipei, Taiwan). Cells were immersed at room temperature (20-25 °C) in HEPES-buffered normal Tyrode’s solution that contained 1.8 mM CaCl₂. When they were left to adhere to the chamber’s bottom for several minutes, the measurements were performed. Ionic currents were recorded with patch electrodes in the cell-attached or whole-cell configuration of a modified patch clamp technique, as described elsewhere [34,38,54,55]. GΩ-seals were typically formed in an all-or-none manner, leading to an improvement in signal-to-noise ratio. The recording pipette was connected to the input stage of an RK-400 (Bio-Logic, Claix, France) or an Axopatch-200B patch-clamp amplifier (Molecular Devices, Bestgen Biotech, New Taipei City, Taiwan). Patch electrodes (3-5 MΩ in bathing solution) were made from Kimax^®-51 borosilicate capillary tubes (#34500; Merck, Taipei, Taiwan) using a two-step vertical puller (PB-7; Narishige, Taiwan Instrument, Tainan, Taiwan) and their tips were heat-polished in an MF-83 microforge (Narishige). All potentials were corrected for the liquid junction potential that would develop at the pipette tip in situations where the composition of the pipette internal solution was different from that in bath medium. Tested compounds were applied by perfusion or added to the bath to obtain the final concentration indicated. In the experiments with corylin plus linopirdine, linopirdine was applied after addition of corylin. When high-frequency stimuli were needed, we used an Astro-Med Grass S85X dual output pulse stimulator (Grass; Zhong Yan, Kaohsiung, Taiwan) [26,48,56].

The current and voltage signals were monitored in real time and recorded onto a laptop computer. The recorded data were low-pass filtered at 2 kHz using an FL-4 four-pole Bessel filter (Dagan, Minneapolis, MN) and digitized at 10 kHz or more with a Digidata 1440A interface (Molecular Devices). The device was connected to either an RK-400 or Axopatch-200B patch-clamp amplifier, which was controlled via a universal series bus (USB) connection using the pClamp 10.6 software (Molecular Devices). Ionic currents obtained from whole-cell and single-channel recordings were analyzed offline using pClamp 10.7, OriginPro^® (OriginLab; Scientific Formosa, Kaohsiung, Taiwan), and custom-written macros in Excel^® 2021 (Microsoft, Redmond, WA) running on Windows 11. Capacitive transients following repolarization were commonly observed; therefore, the tail deactivating K⁺ currents were measured after the capacitive currents had settled, typically between 10 and 20 msec after the end of voltage pulse.

4.4. Data Analyses

To determine the concentration-dependent stimulatory effect of corylin on the amplitude of I_K(M), GH₃ cells were bathed in a high-K⁺, Ca²⁺-free solution, while the recording electrode was filled with a K⁺-containing solution. To measure I_K(M) amplitude, we voltage-clamped each tested cell at −50 mV, and the depolarizing pulse up to 1 sec in duration to −10 mV was imposed on it. The I_K(M) amplitude measured at the end of depolarizing pulses in the presence of 300 μM corylin was defined as 1.0 (i.e., 100%). The corresponding amplitudes obtained during the control period (in the absence of corylin) and during exposure to different concentrations of corylin (1-300 μM) were measured and compared. The concentration required to stimulate 50% of the current amplitude was determined according to a modified Hill function. That is,

$$percentageincrease={\displaystyle \raisebox{1ex}{$E_{max}\times {\left[C\right]}^{n_H}$}\!\left/ \!\raisebox{-1ex}{$\left({EC}_{50}^{n_H}+{\left[C\right]}^{n_H}\right)$}\right.}$$

(1)

In this equation, EC₅₀ = the concentration required for 50% stimulation; n_H = the Hill coefficient; [C] = the corylin concentration applied; and E_max = maximal stimulation. This formula enables optimal convergence, providing the best fit line and accurate parameter estimates (e.g., EC₅₀ and n_H).

The activation time constants (τ_act) of I_K(M) in response to prolonged membrane depolarization, obtained in the absence or presence of corylin, were determined by fitting the digitized current traces with a single-exponential function, as indicated in Figure 1B.

To characterize the stimulatory action of corylin on I_K(M) amplitude, we constructed the quasi-steady-state activation curve of the current. The relationships between the membrane potentials and the normalized amplitudes of I_K(M) with or without the existence of this compound were established and thereafter fitted with a Boltzmann function given by:

$${\displaystyle \raisebox{1ex}{$I$}\!\left/ \!\raisebox{-1ex}{$I_{max}$}\right.}={\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$\left\{1+e^{\left[\raisebox{1ex}{$-(V-V_{1/2})qF$}\!\left/ \!\raisebox{-1ex}{$\left(RT\right)$}\right.\right]}\right\}$}\right.}$$

where I_max = the maximal activated current of I_K(M); V = the membrane potential; V_1/2 = the voltage for half-maximal stimulation; q = the apparent gating charge of the activation curve; F = Faraday’s constant; R = the universal gas constant; and T = the absolute temperature.

Linear (e.g., single-channel conductance) or nonlinear (e.g., concentration-dependent relationships and voltage-dependent activation curves) curve fitting to data sets was performed using an interactive least-squares method. The analysis was conducted using tools such as the Solver add-in in Excel^® 2021 (Microsoft) and OriginPro^® 64-bit (OriginLab).

4.5. Single-Channel Analysis of the K_M Channel

Single K_M-channel currents in GH₃ cells were recorded and analyzed by pClamp 10.7 (Molecular Devices). To evaluate single-channel opening events, amplitude distributions were fitted with multi-Gaussian adjustments. Channel open probabilities were determined through an iterative process to minimize the χ ² values across a sufficiently large set of independent observations. Open lifetime distributions of K_M channels (i.e., mean open time), obtained with or without corylin presence, were fitted using least-squares analysis with logarithmically scaled bin widths.

4.6. Statistical Analyses

The presented data are expressed as mean ± standard error of the mean (SEM), with n representing the number of cells from which the samples were obtained. Comparisons between two groups were performed using paired or unpaired Student’s t-tests, as appropriate. For multiple-group comparisons, one-way analysis of variance (ANOVA) was conducted, followed by Fisher’s least significant difference (LSD) post hoc test to assess individual group differences. Statistical significance was defined as P < 0.05 and is indicated in the figure by ^*, ^**, or ⁺