Gryllus bimaculatus used in the experiments were sexually mature females purchased from a local pet store as food for pet reptiles, and thus, genetic and environmental variability in the specimens were limited. The crickets were housed in plastic containers with cardboard shelters until further analysis. All crickets were bred at 27 ± 2°C, with humidity of 35–60%, in a 12:12-h light:dark cycle. Crickets were provided ad libitum access to feed and water for insects (I, Oriental yeast CO., LTD., Kyoto, Japan).

The adult female crickets were fixed in a chamber via the upper dorsal area under CO₂ anesthesia. The lateral oviducts were exposed by removing connective tissue around the reproductive organs after a dorsal incision in the abdomen in normal saline (in mM): 140 NaCl, 10 KCl, 1.6 CaCl₂, 2 MgCl₂, 44 D-glucose, and 2-[4-(2-hydroxyethyl)-1-piperazinyl] ethanesulfonic acid (HEPES); pH was adjusted to 7.4 with 2-amino-2-hydroxymethyl-1,3-propanediol (Tris(hydroxymethyl)aminomethane) (Tris). The left and right lateral oviducts connected to the common oviduct from the vitellarium were excised. Cell dissociation was performed enzymatically using the protease dispersion method described previously (15). The isolated single lateral oviduct myocytes were maintained in fresh saline at room temperature (23-27°C) and used within 12 h.

The cells were dropped on a glass-bottom dish containing the experimental solution, and the adhered cells were used for measurements. Cells were observed under an inverted microscope (IX70: Olympus, Tokyo, Japan). Currents from cells were amplified using Axopatch 200B (Axon Instruments/Molecular Devices, Union. City, CA, USA) and acquired using the Digidata1440 A/D converter (Axon Instruments/Molecular Devices). Experiments were performed at room temperature (22–27°C) using patch-clamp techniques for the cell-attached and excised inside-out mode. Patch electrodes were prepared from capillary tubes (Hemato-clad capillary; Drummond Scientific Co., Broomall, PA, USA) with a two-stage pipette puller (PC-10 Narishige, Tokyo, Japan) with a tip resistance of approximately 5–8 MΩ when filled with a solution for single-channel recordings. Current signals were filtered at 5 kHz with a four-pole Bessel filter and digitized at 10 or 20 kHz. pCLAMP (version 6, 7, 10, or 11; Axon Instruments/Molecular Devices) software was used for command pulse control, data acquisition, and analysis. A square voltage pulse of 3 mV, 10 ms, 10 Hz was applied as a ‘test pulse’ before the current amplitude measurement to ensure the accuracy of the measurement. A cell membrane resistance >1 gigaohm after application of the test pulse was considered a trial. Trials in which the current amplitude was observed within 10 mins after forming a stable gigaohm-seal were used for further analysis. The amplitude of single-channel currents, number of open channels, and mean open probability (NP_O) were determined by a cursor on Clampfit, Fetchan, pStat, or using the single-channel search mode of the pCLAMP software. Data were also analyzed using the Origin (OriginLab Corp., Northampton, MA, USA), Excel (Microsoft, Redmont, WA, USA), and Sigma Plot (Systat Software, San Jose, CA, USA) software packages. For single-channel recordings to investigate the biophysical properties of ion channels ( Figures 1 , 2 ), the external solution used contained a high potassium solution [(in mM) 140 KCl, 10 NaCl, 1.6 CaCl₂, 2 MgCl₂, 2 HEPES] that maintained the resting membrane potential at zero. The composition of the pipette solution was the same as that of the high-potassium bath solution with or without 1 mM Ba²⁺ used in experiments illustrated in Figures 1 - 4A , but Ca²⁺ was omitted for the experiments in Figures 4B , 5 . To test the K⁺ selectivity of isolated myocytes, 100, 70, and 35 mM KCl solutions were prepared by replacing KCl in the pipette solution with an equal amount of NaCl (11). To investigate the Ca²⁺ dependence of IK channels, Ca²⁺-free solution was obtained by omitting CaCl₂ from the normal bath solution saline and adding 1 mM ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA). The intracellular solution was balanced with 5 mM EGTA to maintain free Ca²⁺ concentrations between 1 nM and 10 μM. A bath solution with Ca²⁺ concentration of ≥100 μM was added by adjusting the amount of CaCl₂ in normal saline. Extracellular effect of Ba²⁺ was tested by application in the pipette solution using the standard backfill method established previously (48). In brief, the tips of the electrodes were first filled with the normal pipette solution, then backfilled with the same solution containing a concentration of its inhibitors; a wait period of at least 10 min was ensured before data recording. Gd³⁺ was dissolved in water to prepare stock solutions, and aliquots were added to the perfusate. All inhibitor reagents used in the experiment were purchased from Sigma-Aldrich Corp. (St. Louis, MO, USA). All other reagents were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan).

All data are expressed as means ± standard error of mean (SEM). We accumulated data for each condition from at least three independent experiments. The numbers of animals, and trials used in each experiment are described in the figure legend. Statistical significance was evaluated by Student’s t-test for comparisons between two mean values to assess statistical significance using Excel (Microsoft Corp., Redmond, WA, USA) or Origin 8 (OriginLab Corp.) software. The data used for statistical analysis passed the Shapiro-Wilk normality test and the Levene equal variance test. For other correlation analyses, least-squares linear regression was performed using Excel (Microsoft Corp.). A P-value of <0.05 was considered significant.

The top trace in Figure 1A shows single-channel activity with large and medium conductance in the steady state of oviduct myocytes. Single-channel currents in the left column in K⁺-rich recording conditions are consistent with our previous observations (11).

As shown in the upper traces of Figure 1A , both large (BK) and intermediate-sized single-K⁺ channel currents (IK) were observed either singly or multiples by trial when the membrane was held at +40 mV. During the course of this experiment, however, we newly found that the activity of this BK channel disappeared when 1 mM Ba²⁺ was added extracellularly. Thus, the extracellular addition of Ba²⁺ enabled selective recordings of the IK channels ( Figure 1A , right column). Therefore, this study performed recordings using Ba²⁺ in a pipette to eliminate BK channel activity.

As shown in Figure 1B , when the membrane potential was held at depolarizing or hyperpolarizing potentials in cell-attached mode, the single-channel properties showed burst-like kinetics. The channel activity did not show voltage dependence (NP_O = 0.44 ± 0.03 and 0.51 ± 0.05 at +60 and −60 mV, respectively; n = 18). Single-channel currents recorded at membrane potentials from −60 mV to +60 mV showed a linear current-voltage (I-V) curve with a slope conductance of 61.5 pS (n = 12) ( Figure 1C ). In addition, even in the absence of extracellular Ba²⁺, IK currents recorded in patch membranes not coexisting with BK did not affect the slope conductance values (59.7 pS, n = 6).

We investigated the effects on single-channel conductance and reversal potential (E_rev) observed in Figure 1 by varying the extracellular K⁺ concentration. I-V relationships constructed from single-channel recordings at membrane potentials from −60 mV to +60 mV ( Figure 2A ) showed linear I-V relationships under three extracellular K⁺ concentrations. Linear analysis of the I-V relationship by least squares showed E_revs of −32.8, −24.5, and +2.4 mV for channel currents recorded under 35-, 70-, and 140-mM conditions, respectively. The slope of the E_rev change plotted against the change in extracellular K⁺ concentration was 58.4 mV per decade change in K⁺ concentration ( Figure 2B ). We next investigated the ion selectivity of the single-channel and recorded the effect on the E_rev by changing the extracellular K⁺ concentration. I-V relationships expressed from single-channel recordings at membrane potentials from −60 mV to +60 mV showed linear I-V relationships under three extracellular K⁺ concentrations. Linear analysis of the I-V relationship by least squares showed E_revs of −32.8, −24.5, and +2.4 mV for channel currents recorded under 35-, 70-, and 140-mM conditions, respectively ( Figure 2A ). The slope of the E_rev change plotted against the change in extracellular K⁺ concentration was 58.4 mV per decade change in K⁺ concentration ( Figure 2B ).

IK channels have the unique property of being activated by increases in [Ca²⁺]_i (28, 49). To directly evaluate the [Ca²⁺]_i dependence of K⁺ channels, we measured single-channel currents from excised inside-out patches of membranes to different concentrations of Ca²⁺-containing bath solutions. Single-channel currents recorded at various concentrations of Ca²⁺ bath solutions at +60 mV were enhanced with increasing [Ca²⁺]_i ( Figure 3A ). The relationship between relative Po and [Ca²⁺]_i was then fitted to the Hill equation, yielding a k-value of 88.5 μM and Hill coefficient of 0.9 ( Figure 3B ).

Activation of Ca²⁺-dependent IK channels in excitable cells is critical for feedback control of both Ca²⁺ influx and cell excitability (27). To investigate the effect of Ca²⁺ on IK channel activity, we examined the effect of extracellular Ca²⁺ removal on IK channel activity.

Extracellular Ca²⁺ abolition suppressed IK channel activity after consistent IK channel activity ( Figure 4A ). We next investigated the pathways of extracellular Ca²⁺ influx and their effects on IK channel activity. Membrane stretching-induced extracellular Ca²⁺ influx pathways play a central role in myogenic muscle contraction (50). Gd³⁺-sensitive- and Ca²⁺ permeable-nonselective cation channels activated by membrane stretching (SA channel) during muscle contraction are functionally expressed in cricket oviduct cells (15). In support of these evidences, Gd³⁺ administration suppressed IK channel activity ( Figure 4B ). We further investigated the proximity of IK and SA channels within microdomains by observing the effect of membrane stretch on IK channel activity.

To eliminate the possibility of Ca²⁺ influx from the bath solution, the bath solution was replaced with a Ca²⁺-free solution before the recording ( Figure 5 , lower panel). Recordings were obtained with and without Ca²⁺ in the pipette. This restricted Ca²⁺ access to the patch membrane only from the pipette solution. In addition, the recording membrane voltage was maintained at −140 mV, allowing simultaneous measurement of low-conductance SA channel currents (~13 pS) (15) and IK channel activity. As shown in Figure 5A , when Ca²⁺ was contained in the pipette, SA channel activity increased in response to membrane stretching caused by mechanical stimulation in response to a negative pressure of −30 cm H₂O in the pipette ( Figure 5A ). This mechano-stimulated single-channel activity is consistent with results in our previous report (15). Around 30 s after SA activity was recorded, IK channel activity was observed. It should be noted that SA and IK channels can be observed individually due to their different conductance and mean opening time. The conductance and mean opening time of the SA channel were 14 pS and 5.4 ms (n = 40) and those of IK channel were 61.5 pS and 1.2 ms (n = 48), respectively. The same SA channel activity as in Figure 5A was observed with −30 cm H₂O suction wherein Ca²⁺ was not contained in the pipette ( Figure 5B ). Under this condition, even if the suction was further increased to −60 cm H₂O, SA channel activity increased, but IK channel activity was not observed (n = 12).

We analyzed the effect of membrane stretching on the activity of SA and IK channels, as observed in Figure 5A . As shown in Figures 6A , B channel activity increased with the strength of membrane stretching, and positive correlation coefficients of 0.987 and 0.996 were calculated for SA and IK channels, respectively. Furthermore, the average value at 10 cm of suction strength was analyzed. It was revealed that SA and IK channels increased NPo by 0.027 and 0.019, respectively, with respect to membrane stretching strength of 1 cm H₂O ( Figure 6C ).