Small conductance Ca²⁺-activated K⁺ channels are regulated by Ca²⁺–calmodulin-dependent protein kinase II in murine colonic myocytes

Abstract

1. Ca²⁺ regulates the activity of small conductance Ca²⁺-activated K⁺ (SK) channels via calmodulin-dependent binding. We investigated whether other forms of Ca²⁺-dependent regulation might control the open probability of SK channels.

2. Under whole-cell patch-clamp conditions, spontaneous openings of SK channels can be resolved as charybdotoxin-insensitive spontaneous transient outward currents (STOCs). The Ca²⁺–calmodulin-dependent (CaM) protein kinase II inhibitor KN-93 reduced the occurrence of charybdotoxin-insensitive STOCs.

3. The charybdotoxin-insensitive STOCs are related to spontaneous, local release of Ca²⁺. KN-93 did not affect spontaneous Ca²⁺-release events.

4. KN-93 and W-7, a calmodulin inhibitor, decreased the open probability of SK channels in on-cell patches but not in excised patches.

5. Application of autothiophosphorlated CaM kinase II to the cytoplasmic surface of excised patches increased the open probalibity of SK channels. Boiled CaM kinase II had no effect.

6. We conclude that CaM kinase II regulates SK channels in murine colonic myocytes. This mechanism provides a secondary means of regulation, increasing the impact of a given Ca²⁺ transient on SK channel open probability.

A significant portion of the inhibitory regulation of gastrointestinal (GI) smooth muscle is mediated by ATP released from enteric inhibitory motor neurons (see Hoyle & Burnstock, 1989). Post-junctional responses to purinergic inhibitory inputs occur via stimulation of P_2Y purinoceptors and activation of small conductance Ca²⁺-activated K⁺ (SK) channels (Koh et al. 1997; Vogalis & Goyal, 1997). SK channels were cloned from rat and human brain, and they constitute a unique family of potassium channels (Kohler et al. 1996). SK channels have been identified in GI smooth muscle cells by RT-PCR (A. Epperson & B. Horowitz, personal communications) and characterized in native cells by electrophysiological techniques (Koh et al. 1997; Vogalis & Goyal, 1997). SK channels are voltage independent, Ca²⁺ dependent and have a slope conductance of 5.3 pS in symmetrical K⁺ concentrations. These channels transduce fluctuations in intracellular Ca²⁺ concentration into changes in membrane potential (Xia et al. 1998) and can therefore regulate membrane excitability and, most importantly, the open probability of voltage-dependent Ca²⁺ channels. In the case of GI muscle cells localized Ca²⁺ release from IP₃ receptor-operated stores is responsible for activation of SK channels in response to ATP stimulation (Kong et al. 2000). At present, it is unknown whether SK channels are regulated solely by local changes in Ca²⁺ concentration or whether there is secondary regulation of these channels via Ca²⁺-dependent protein kinases.

Studies of cloned channels (SK1-SK3 isoforms) show that Ca²⁺ regulation occurs by binding of Ca²⁺ to calmodulin, which forms heteromeric complexes with SK channels (Xia et al. 1998). Ca²⁺ binding is thought to induce channel gating by causing conformational changes in calmodulin that are conveyed to the α subunit of SK channels. Expression of SK isoforms in Xenopus oocytes resulted in channels that were activated by Ca²⁺ but not affected by calmidazolium or calmodulin inhibitory peptide. Thus strong binding occurs between calmodulin and SK α subunits. The lack of effect by calmodulin inhibitory drugs suggested that SK channels are not directly regulated by calmodulin-binding enzymes, such as Ca²⁺-calmodulin-dependent (CaM) protein kinases (e.g. CaM kinase II) or calcineurin, although inhibitors of CaM kinase II and calcineurin were not used in these studies.

CaM kinase II is expressed in smooth muscles and has been reported to regulate cell migration (Abraham et al. 1997), Ca²⁺ currents (McCarron et al. 1992), Ca²⁺-activated Cl⁻ currents (Wang & Kotlikoff, 1997), rapidly inactivating delayed rectifier K⁺ currents (Koh et al. 1999), SR Ca²⁺-ATPase (Xu & Narayanan, 1999), and the Ca²⁺ sensitivity of smooth muscle myosin light chain kinase (Edwards et al. 1998). SK2, one of the predominant isoforms of SK channels expressed by GI smooth muscles, has four potential sites for CaM kinase II phosphorylation: one in the N-terminal and three in the C-terminal ends of the protein. Ca²⁺-dependent protein kinase C (PKC) isoforms are also expressed in smooth muscles (Andrea & Walsh, 1992) and SK channels also contain consensus sequences for PKC. Regulation of SK channel open probability by CaM kinase II or PKC could represent an unrecognized feature of the Ca²⁺ dependence of this conductance. Therefore, we have tested the hypothesis that part of the regulation of SK channels in native smooth muscle myocytes is provided by CaM kinase II or PKC.

METHODS

Preparation of isolated myocytes

Colonic smooth muscle cells were prepared from BALB/c mice. Briefly, mice were anaesthetized with chloroform, and after cervical dislocation the colon was removed as approved by the Institutional Animal Care and Use Committee. Colons were cut along the longitudinal axis, pinned in a Sylgard-lined dish, and washed with Ca²⁺-free, phosphate-buffered saline (PBS) containing (mM): 125 NaCl, 5.36 KCl, 15.5 NaOH, 0.336 Na₂HPO₄, 0.44 KH₂PO₄, 10 glucose, 2.9 sucrose, 11 N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid) (Hepes). Mucosa and submucosa were removed using fine-tipped forceps. Pieces of muscle were incubated in a Ca²⁺-free solution supplemented with 4 mg ml⁻¹ fatty acid-free bovine serum albumin, 2 mg ml⁻¹ papain, 1 mg ml⁻¹ collagenase, and 1 mM dithiothreitol. Tissues were incubated at 37°C in enzyme solution for 8–12 min, washed with Ca²⁺-free solution and single cells dispersed with gentle agitation. Dispersed cells were stored at 4°C in Ca²⁺-free solution supplemented with S-MEM media (Sigma) and 0.5 mM CaCl₂, 0.5 mM MgCl₂, 4.17 mM NaHCO₃ and 10 mM Hepes. Cells were allowed to adhere to the bottom of a recording chamber on an inverted microscope for 5 min prior to recording. Myocytes from both the longitudinal and circular smooth muscle layers were used. Because the circular layer is thicker than the longitudinal muscle it is likely that the majority of cells were from the circular layer.

Voltage-clamp methods

The whole-cell patch-clamp technique was used to record membrane currents from dispersed myocytes. Currents were amplified with an Axopatch 200B amplifier (CV-4 headstage; Axon Instruments, Foster City, CA, USA) and digitized with a 12 bit A/D converter (Axon Instruments). Data were filtered at 1 kHz with pCLAMP software (Axon Instruments). Frequency and amplitude of spontaneous transient outward currents (STOCs) during 1 min of recording were analysed using the Mini Analysis program (Synaptosoft Software, Leonia, NJ, USA). For the single channel recordings, data were sampled at 5 kHz and filtered at 1 kHz. All-points amplitude histograms were constructed and open probabilities were determined from 2 min recordings. All recordings were performed at room temperature.

Smooth muscle myocytes were bathed in a solution containing (mM): 5 KCl, 135 NaCl, 2 CaCl₂, 10 glucose, 1.2 MgCl₂ and 10 Hepes adjusted to pH 7.4 with tris(hydroxymethyl)aminomethane (Tris). STOCs were recorded in the majority of cells and these events were highly sensitive to Ca²⁺ as previously shown in studies of guinea-pig colonic myocytes (Kong et al. 2000). Replacement of external Ca²⁺ with Mn²⁺ gradually abolished STOCs (within 20 min). Dialysis of cells with 10 mM BAPTA (free Ca²⁺∼10 nM) using the conventional whole-cell configuration of the patch-clamp technique abolished STOCs. Dialysis of cells resulted in run-down of STOCs within several minutes even when 0.1 mM EGTA (free Ca²⁺∼100 nM) was used in the pipette solution. To avoid run-down of STOCs, experiments were performed with the perforated patch technique. For perforated patch experiments, the pipette solution contained (mM): 140 KCl, 0.5 EGTA and 5 Hepes adjusted pH 7.2 with Tris. Amphotericin B (90 mg ml⁻¹, Sigma) was dissolved with dimethyl sulfoxide, sonicated and diluted in the pipette solution to give a final concentration of 270 μg ml⁻¹.

For recording K⁺ channels in cell-attached or excised patches, the bath solution contained (mM): 140 KCl, 1 EGTA, 0.61 CaCl₂ and 10 Hepes adjusted to pH 7.4 with Tris. Two concentrations of Ca²⁺ were added to bath solutions buffered by 1 mM EGTA to create Ca²⁺ activities from 10⁻⁷ to 10⁻⁶ M. The pipette solution was identical to the whole-cell bath solution and included 200 nM charybdotoxin to inhibit BK.

KN-93, KN-92, KN-62 and W-7 (Calbiochem, San Diego, CA, USA) were added to bath solutions. In a previous study we tested a range of concentrations of KN-93 and found that 1 μM was effective in blocking effects attributable to CaM kinase II in murine colonic myocytes (Koh et al. 1999). Therefore, we have used this concentration in the present studies. Autophosphorylated and boiled CaM kinase II were gifts from Drs Debra A. Brickey and Tom R. Soderling (Vollum Institute; see Lledo et al. 1995 for details of preparation). Autothiophosphorylated CaMKII was stored in 1 mM EGTA and diluted in bath solution. The ratio of dilution from stock autothiophosphorylated CaMKII was 10000-fold. Calphostin C and chelerythrine (both from RBI) were diluted in the external bath solution.

Confocal microscopy

Cell suspensions were placed in a specially designed 0.5 ml chamber with a glass bottom. The cells were incubated for 35 min at room temperature in Ca²⁺-free buffer containing fluo-3 acetoxymethyl ester (AM) (10 μg ml⁻¹; Molecular Probes, Eugene, OR, USA) and pluronic acid (2.5 μg ml⁻¹; Teflabs, Austin, TX, USA). Cell loading was followed by a 25 min incubation in a solution containing 2 mM Ca²⁺ to restore normal extracellular Ca²⁺ concentration and to complete de-esterefication of fluo-3. All measurements were made within 15–45 min after restoring extracellular Ca²⁺.

An Odissey XL confocal laser scanning head (Noran Instruments Inc., Middleton, WI, USA) connected to a Nikon Diaphot 300 microscope with a × 60 water immersion lens (NA, 1.2) was used to image the cells. The cells were scanned using Intervision software (Noran Instruments Inc.) running on an Indy workstation (Silicon Graphics, Inc., Mountain View, CA, USA). Changes in the fluo-3 fluorescence (indicating fluctuations in cytosolic Ca²⁺) were recorded for 20 s test periods using T-series acquisition and a laser wavelength of 488 nm (excitation for fluorescein isothiocyanate, FITC). Six hundred frames were acquired per test period (one frame every 33 ms), creating 20 s movie files.

Image analysis was performed using Interactive Data Language software (Research Systems Inc, Boulder, CO, USA). Baseline fluorescence (F_o) was determined by averaging 10 images (out of 600) in which no activity was observed. Ratio images (F/F_o) were constructed and replayed to detect regions of interest where spontaneous changes in F/F_o occurred. F/F_ovs. time traces were analysed by Microcal Origin (Microcal Software, Inc., Northampton, MA, USA) and AcqKnowledge Software (Biopac Systems Inc., Santa Barbara, CA, USA). Fluorescence traces shown in this paper represent averaged F/F_o from a box region of 2.2 μm × 2.2 μm centred within an area of interest. This box size (4.8 μm²) was determined empirically to be the best compromise between temporal and spatial precision of Ca²⁺ release events and the signal to noise ratio.

Statistical analysis

All statistical analyses were performed using SigmaStat 2.0 (Jandel Corporation, San Rafael, CA, USA). Data represent means ± standard error of the mean. Values of n are the number of cells. Student's paired t tests were performed and a P value of less than 0.05 was considered significant.

RESULTS

Effects of inhibitors of CaM kinase II and PKC on spontaneous transient outward currents

Holding cells at -30 or -40 mV under whole-cell voltage clamp with perforated patches resulted in spontaneous transient outward currents (STOCs; Fig. 1A, B and E), which have previously been reported to be due to openings of large and small conductance Ca²⁺-activated K⁺ channels in GI smooth muscles (Kong et al. 2000). Addition of charybdotoxin (ChTX, 200 nM) reduced the amplitude and frequency of STOCs and revealed small-amplitude ChTX-resistant STOCs (Fig. 1A, C and F) that were reduced by apamin (10⁻⁶ M; see Kong et al. 2000; O. Bayguinov & K. M. Sanders, unpublished observation). The CaM kinase II inhibitor KN-93 (1 μM) reduced the frequency of STOCs in a reversible manner from 76 ± 13 to 14 ± 5 events min⁻¹ (n = 4; Fig. 1A, D and G). KN-62 (10 μM) had similar effects to KN-93. KN-92 (1 μM), an inactive form of KN-93, had no effects on the frequency of STOCs (not shown). Inhibitors of PKC, calphostin C (100 nM; n = 2) and chelerythrine (1 μM; n = 2) also had no effect on STOCs (not shown).

Figure 1. Effects of KN-93 on spontaneous transient outward currents (STOCs) in murine myocytes.

Cells were studied with the amphotericin-premeabilized patch technique (see Methods). A, representative STOCs recorded at holding potentials of -30 mV. Charybdotoxin (ChTX, 200 nM for 15 min) reduced the amplitude and frequency of STOCs. KN-93 reduced the amplitude and frequency of ChTX-resistant STOCs. B-D, periods denoted by the horizontal bars in A shown on an expanded time scale, STOCs in control (B), in the presence of ChTX (C) and in the presence of ChTX and KN-93 (1 μM) (D). E–G, amplitude-frequency histograms (corresponding to the STOCs in B–D, respectively) show the inhibition of STOCs by ChTX (200 nM) and KN-93 (1 μM).

The occurrence of STOCs in smooth muscles has been attributed to localized release of Ca²⁺ (Benham & Bolton, 1986; Nelson et al. 1995). Therefore, the reduction in STOCs caused by KN-93 could be due to effects on spontaneous Ca²⁺ release events. Cells loaded with fluo-3 were examined with confocal microscopy. Spontaneous, localized Ca²⁺ release events in murine colonic myocytes have previously been shown to be blocked by inhibitors of IP₃ receptors but not by ryanodine (Bayguinov et al. 1999). Application of KN-93 had no measurable effect on spontaneous Ca²⁺ release events (Fig. 2), suggesting that the effects of KN-93 on whole-cell currents may be mediated by effects on the ionic conductances that underlie ChTX-independent STOCs. Therefore, additional experiments were performed using single channel recording techniques to directly assess the regulation of SK channels by CaM kinase II.

Figure 2. Effects of KN-93 on Ca²⁺ transient activity.

A, fluorescence from a small region of an isolated myocyte that demonstrated spontaneous Ca²⁺ transients. The data are from a 20 s scan under control conditions. B, a 20 s scan of the same cell as in A 5 min after addition of KN-93 (1 μM). The CaM kinase II inhibitor had no effect on Ca²⁺ transients. C, summary of data from 7 experiments. There was no significant change in the area under the Ca²⁺ transients before and after KN-93 (n = 7, P > 0.5). The method for determining F/F_o is described in Methods. The scale bar for F/F_o= 2.

Effects of inhibitors of CaM kinase II and calmodulin on SK channels

We have previously characterized SK channels in murine colonic myocytes. We tested the effects of KN-93 on the open probability of SK channels in cell-attached patches using asymmetrical K⁺ gradients (5 mM/140 mM). To exclude contamination from other conductances (see Koh et al. 1997), the patch pipettes contained ChTX (200 nM) and cells were held at 0 mV to inactivate delayed rectifier channels. The mean amplitude of currents attributed to SK channels was 0.35 ± 0.05 pA at a holding potential of 0 mV (n = 12; Fig. 3A and D). KN-93 (1 μM) decreased the open probability (P_o)from 0.18 ± 0.05 to 0.02 ± 0.01 (n = 4; Fig. 3B and E). NP_o (where N is the number of active channels in the patch) decreased within 5 min of exposure to KN-93 and recovered after KN-93 was removed (Fig. 3C, F and G). Exposure to KN-92 (1 μM) did not affect the open probability of SK channels (data not shown). The calmodulin inhibitor W-7 had similar effects on the open probability of SK channels to KN-93. W-7 (10⁻⁵ M) reduced NP_o from 0.26 ± 0.1 to 0.01 ± 0.01 (n = 4; Fig. 3H and I).

Figure 3. Effects of KN-93 and W-7 on SK channels in on-cell patches.

A, control recordings from an on-cell patch. The patch was held at 0 mV in asymmetrical K⁺ concentrations (5 mM in pipette and 140 mM in bath solution). B, KN-93 (1 μM) decreased SK channel openings. C, the effects of KN-93 were reversible. Traces below main recordings in A–C show the periods denoted by horizontal bars on an expanded time scale. D–F, all-points amplitude histograms from recordings in A-C, respectively. G, a plot of the time course of the effects of KN-93 on SK channel open probability. H, control recordings from an on-cell patch held at 0 mV in asymmetrical K⁺ concentrations (5 mM in pipette and 140 mM in bath solution; 10⁻⁶ M free Ca²⁺ in bath). o, channel open; c, channel closed. I, W-7 (10 μM) reduced SK channel open probability. Traces below main recordings in H and I show the periods denoted by horizontal bars on an expanded time scale. All-points histograms from recordings before and after W-7 are shown under the traces.

Treatment of excised patches with constitutively activated autophosphorylated Ca²⁺-calmodulin kinase II and W-7

Experiments were also performed to test the Ca²⁺ sensitivity of SK channels in excised (inside-out) patches. Patches were excised into bath solutions containing 10⁻⁷ M Ca²⁺. The NP_o of SK channels was 0.01 under these conditions (Fig. 4A and C). Raising bath Ca²⁺ from 10⁻⁷ to 10⁻⁶ M increased the open probability from 0.01 to 0.4 (holding potential 0 mV; Fig. 4B and D), demonstrating the Ca²⁺ dependence of these channels as previously reported (Koh et al. 1997). We tested the effects of autothiophosphorylated CaM kinase II (Lukas et al. 1998) on the open probability of SK channels by exposing the cytoplasmic surface of the patches to the enzyme. In 10⁻⁷ M Ca²⁺ addition of activated CaM kinase II increased the open probability of SK channels from 0.006 ± 0.001 to 0.10 ± 0.02 (n = 3, P < 0.001) (Fig. 4E and F). Boiled CaM kinase II was without effect on NP_o (n = 3; data not shown). In contrast to the effects of W-7 on NP_o of SK channels in on-cell patches, W-7 (up to 10⁻⁵ M) had no effect on SK channels in excised patches.

Figure 4. Ca²⁺ sensitivity and effects of CaM kinase II on SK channels in inside-out patch.

A, the patch was held at 0 mV in asymmetrical K⁺ concentrations (5 mM in pipette and 140 mM in bath solution). The intracellular surface of the membrane was exposed to 10⁻⁷ M free Ca²⁺. The Ca²⁺ was increased to 10⁻⁶ M in B. Traces under the main recordings in A and B show the periods denoted by horizontal bars on an expanded time scale. C and D, all-points histograms from recordings in 10⁻⁷ and 10⁻⁶ M Ca²⁺, respectively. E, an excised patch held at 0 mV was exposed to 10⁻⁶ M Ca²⁺ and then the solution was changed to 10⁻⁷ M Ca²⁺, causing a decrease in NP_o of SK channels. Application of autothiophosphorylated CaM kinase II (CaMKII, 10 nM) to the cytoplasmic surface of the patch increased openings of SK channels. Traces below the main recording in E show the periods denoted by horizontal bars on an expanded time scale. All-points amplitude histograms in the presence of 10⁻⁶ M and 10⁻⁷ M free Ca²⁺ and autothiophosphorylated CaM kinase II are shown in F.

DISCUSSION

We previously showed with immunoblots from colonic muscle homogenates that CaM kinase II is expressed in the tunica muscularis of the murine colon, and immunohistochemistry confirmed the presence of CaM kinase II in smooth muscle cells (Koh et al. 1999). CaM kinase II slowed the rate of inactivation of the fast-inactivating component of the delayed rectifier in colonic myocytes (Koh et al. 1999), and others have shown that CaM kinase II modulates voltage-dependent Ca²⁺ currents in amphibian gastric smooth muscle (McCarron et al. 1992, 1997). In the present study we found that CaM kinase II also regulates the open probability of SK channels, which are important in mediating responses to enteric inhibitory neural inputs. Taken together, these findings suggest that CaM kinase II contributes to both myogenic and neurogenic regulation of colonic motility.

Previous studies reported that cloned SK channels expressed in Xenopus oocytes depend upon a tight association between SK channels and calmodulin for Ca²⁺-dependent regulation (Xia et al. 1998). The same study concluded that SK channels are not affected by calmodulin-dependent enzymes. This study was elegant in demonstrating the integral role of calmodulin in the regulation of SK channels, but direct tests of the role of CaM kinase II and other calmodulin-dependent enzymes were not performed. The negative conclusion about regulation by calmodulin-dependent enzymes came from the observation that application of calmidozolium and calmodulin inhibitory peptide to the cytoplasmic surface of excised patches was without effect. We also observed that W-7 had no effect on SK channels in excised patches, but W-7, a general calmodulin inhibitory drug, and KN-93, a specific inhibitor of CaM kinase II, both inhibited the open probability of SK channels in on-cell patches. Ledoux and coworkers (1999) recently reported that blockers of CaM kinase II (KN-62 and KN-93) and the inactive analogue of KN-93, KN-92, produced non-specific inhibitory effects on voltage-dependent K⁺ currents of vascular smooth muscle cells. It is unlikely that the effects of KN-93 on SK channels and STOCs attributed to SK channels in the present study were non-specific in the present study, because: (i) KN-92 had no effect on STOCs or single channel currents, and (ii) exposure of the cytoplasmic surface of excised patches to constituitively activated CaM kinase II activated SK channels. Taken together, these data suggest that CaM kinase II regulates SK channels in GI smooth muscle cells.

Recent studies have suggested that SK channels in colonic myocytes are activated by purinergic stimulation of P_2Y receptors (Koh et al. 1997). Coupling between P_2Y receptors and SK channel activation occurs via G-protein-dependent activation of phopholipase C and production of IP₃ (Boyer et al. 1989; Kong et al. 2000). Localized Ca²⁺ release due to IP₃ receptor-operated stores appears to be the prime stimulus for SK activation (Bayginov et al. 1999; Kong et al. 2000). Activation of SK channels leads to outward current and reduced membrane excitability. The present study indicates that the impact of localized enhancement in [Ca²⁺] may be amplified by activation of CaM kinase II. Thus, a given rise in [Ca²⁺] may result in an increase in the open probability of SK channels that is greater than would be caused by the facilitating role of Ca²⁺ alone. This may increase the sensitivity of SK channels to transient, localized Ca²⁺ release events and allow the cell to achieve functional increases in SK without large amplitude or sustained Ca²⁺ transients, which are inherently contradictory to the inhibitory nature of enteric purinergic inputs in GI muscles. The findings of the present study provide an important new dimension to the regulation of SK channels in GI muscles, and may contribute to phenomena such as facilitation or summation of junction potentials during repetitive stimulation of enteric motor neurons (Burnstock et al. 1963).

In summary, Ca²⁺ release from IP₃ receptor-operated stores is the means by which SK channels are regulated and purinergic enteric inhibitory responses are mediated in GI muscles. The present study describes a new mechanism in which Ca²⁺-dependent enhancement in SK channel open probability is amplified by CaM kinase II. The properties of this enzyme are such that this form of regulation could lead to amplification of SK open probability in response to a given increase in [Ca²⁺] and facilitation of channel openings and inhibitory responses with repetitive stimulation.