Role of BKCa in Stretch-Induced Relaxation of Colonic Smooth Muscle

Stretch-induced relaxation has not been clearly identified in gastrointestinal tract. The present study is to explore the role of large conductance calcium-activated potassium channels (BKCa) in stretch-induced relaxation of colon. The expression and currents of BKCa were detected and the basal muscle tone and contraction amplitude of colonic smooth muscle strips were measured. The expression of BKCa in colon is higher than other GI segments (P < 0.05). The density of BKCa currents was very high in colonic smooth muscle cells (SMCs). BKCa in rat colonic SMCs were sensitive to stretch. The relaxation response of colonic SM strips to stretch was attenuated by charybdotoxin (ChTX), a nonspecific BKCa blocker (P < 0.05). After blocking enteric nervous activities by tetrodotoxin (TTX), the stretch-induced relaxation did not change (P > 0.05). Still, ChTX and iberiotoxin (IbTX, a specific BKCa blocker) attenuated the relaxation of the colonic muscle strips enduring stretch (P < 0.05). These results suggest stretch-activation of BKCa in SMCs was involved in the stretch-induced relaxation of colon. Our study highlights the role of mechanosensitive ion channels in SMCs in colon motility regulation and their physiological and pathophysiological significance is worth further study.


Introduction
Mechanical stretch is a basic physiological stimulation. In addition to storage, digestion, absorption, and transport functions, the gastrointestinal (GI) tract is also a stresssensing system and is often subjected to mechanical stretch stimulation [1,2]. Different segments of the GI tract have different responses to stress stimulation: the chyme in the small bowel can lead to intestinal smooth muscle cells (SMCs) contraction by stretch-activation of calcium channels [3] that is conducive to digestion and absorption, while food in the stomach and food residue in the colon can induce a relaxation response, and the dilation of the smooth muscle ensures that increased contents in these cavities do not significantly increase chamber pressure, thus delaying emptying and, consequently, playing a role in storage [4]. The underlying mechanism is unclarified.
Different from the myogenic response of vessels, which refers to the contraction of arterioles initiated by elevations of transmural pressure without the involvement of neural and humoral factors [5], many studies have shown that direct stretch can cause the relaxation of lower esophageal sphincter [6,7], colorectum [8,9], and colon [10]. The relative mechanism of myogenic response of vessels is that the smooth muscle of the blood vessels reacts to the stretch by opening ion channels, which cause depolarization, leading to muscle contraction [5]. It has been reported that the relaxation of GI tract induced by stretch is based on nervous regulation and the localized distension activates enteric reflex to evoke relaxation [8,[10][11][12]. In spite of the neurogenic regulation, it is also reported that there are stretch-activated potassium ion channels in SMCs of GI tract [4]. However, the role of these channels in stretch-induced relaxation of GI tract at tissue level remains uncertain.
The myogenic response of arterioles involves stretchactivation of mechanosensitive ion channels resulting in depolarization of vascular SMCs and calcium influx through L-type voltage-gated calcium channels (VGCC) [13]. Some stretch-dependent potassium (SDK) currents have been found in colonic SMCs [4,10,14]. The activation of SDK outward currents could decrease excitability of SMCs and induce relaxation. Yet the molecular basis is poorly understood.
Large conductance calcium-activated potassium channels (BK Ca ) are widely distributed in various tissues, including the smooth muscle of mammals. They are activated by depolarization and intracellular calcium with high potassium selectivity and high conductance. Potassium outflow mediated by even a mild activation of BK Ca leads to obvious hyperpolarization and relaxation of smooth muscle. Recent studies reported that BK Ca could be activated by membrane stretch [15][16][17]. Our previous study also found that BK Ca in mouse colonic SMCs could be activated by stretch too [18]. Although the physiological significance of the mechanical gating of BK Ca is unclear, it is reasonable to presume that BK Ca in colonic SMCs would be activated when the colon wall is expanded as colonic content increases and result relaxation.
The present study investigated the expression and mechanosensitivity of BK Ca in rat colonic SMCs and examined whether BK Ca are involved in the stretch-induced relaxation of colonic smooth muscle.

Patch-Clamp Experiments.
The SD rats were euthanized by CO 2 according to an IACUC-approved protocol. The distal colon was quickly removed, washed with Ca 2+ -free Hank's solution, and pinned out in a Sylgard-lined dish. The Ca 2+free Hank's solution contained (in mM) 125 NaCl, 5.36 KCl, 15.5 NaHCO 3 , 0.336 Na 2 HPO 4 , 0.44 KH 2 PO 4 , 10 glucose, 2.9 sucrose, and 11 HEPES and was buffered to pH 7.4 with NaOH. After removal of the mucosa and submucosa, strips of smooth muscle layers were cut into small pieces and incu- Patch-clamp studies were performed at room temperature, and cells were used within six hours after enzymatic isolation. The patch pipettes were made from borosilicate glass (BF150-110-10; Sutter Instruments, Novato, CA, USA). The electrode resistance was 8-10 MΩ for single-channel recording.
The single-channel activity was recorded in cell-attached configurations. The pipette solution and bath solution were the same, containing (in mM) 140 potassium aspartate, 5 EGTA, 2 MgCl 2 , and 10 HEPES, with pH adjusted to 7.4. The channel currents were amplified, filtered at 10 kHz, and digitized at 20 kHz using a data acquisition system (Axopatch-200A, Digidata1322A and pClamp 9.2; Axon Instruments, Union City, CA, USA). Suction steps for cell-attached patches were delivered to a side port of the pipette holder with a pressure-generating device (DALE20 pneumatic transducer tester; Luke Biomedical, Liverpool, UK) at a resolution of 1 mmHg. The pressure increased by 10 mmHg in turn and could be manually set in about 1 second. Single-channel activities were analyzed with the QUB software at 2 kHz (http://www.qub.buffalo.edu/; Buffalo, NY, USA). For each recording, 8-second data were carefully analyzed to obtain the channel current amplitude and the open probability. The channel activity in one patch was expressed as , where represents the number of channels in the patch and represents the open probability of a single channel. The outward current was routinely shown to be upward.

Tension Recording of Colonic Smooth Muscle Strips.
The distal colon was isolated from the SD rat and placed in a dish with ice-cold Krebs-Ringer which was gassed with 95% O 2 and 5% CO 2 . The colon has an inner circular and outer longitudinal smooth muscle layers. The longitudinal muscle (LM) strips (2 × 8 mm; width × length) or circular muscle (CM) strips were cut along the direction of the longitudinal or circular axis. The fresh smooth muscle strips were prepared and mounted vertically in Krebs-Ringer's solution maintained at 37 ∘ C. One end of the strip was tied to a hook on the bottom of the organ bath and the other end was connected to a force transducer (MLT0201, Panlab company, Barcelona, Spain). The tissue strips were stretched to 1,000 mg of initial tension and equilibrated for 60 min before experiments were initiated. Parameters of the contraction amplitude and muscle tone of the colon strips were analyzed using the LabChart7 system (AD Instruments Pty Ltd., Bella Vista, New South Wales, Australia). Charybdotoxin (ChTX, 100 nM, Sigma-Aldrich, St. Louis, MO, USA), a nonspecific BK Ca blocker, iberiotoxin (IbTX, 100 nM, Sigma-Aldrich, St. Louis, MO, USA), a specific BK Ca blocker, and tetraethylammonium (TEA, 10 mM, Sigma-Aldrich, St. Louis, MO, USA), a broadspectrum voltage-gated potassium channel (Kv) blocker, were used to block BK Ca . Voltage-gated sodium channel blocker tetrodotoxin (TTX, 100 nM, MedChemExpress, Monmouth Junction, NJ, USA) was used to eliminate the influence of the enteric nervous system.
For the stretch experiments, after pretreatment period, the strips were elongated 1 mm for 10 minutes each time and were stretched five times and the muscle tone was detected to study the response to stretch. The relative change of the muscle tone was determined by measuring the phasic contraction baseline curve.

Statistical
Analysis. All data are expressed as means ± SEM. The data were analysed with GraphPad Prism 5.01 software (GraphPad Software, La Jolla, CA). The Student's paired t-test was used to compare patch-clamp data before and after intervention. For other data, one way ANOVA followed by Newman-Keuls test was used. For all analyses, statistical significance was accepted when P < 0.05.

Expression of BK Ca in Rat Colonic Smooth Muscle.
To confirm that BK Ca play an important role in colonic smooth muscle, the expression levels of BK Ca in smooth muscle layers isolated from SD rat stomach, duodenum, ileum, and colon were detected first. Bar graphs show that BK Ca expression was significantly higher in the colonic smooth muscle than other segments (Figures 1(a) and 1 The single-channel activities of BK Ca on colonic SMCs were identified in the symmetrical K + solution then. The channels were activated by depolarization with a singlechannel conductance of 210 ± 10 pS ( = 10) at +100 mV. The currents were abolished by 100 nM ChTX (data not shown). Multiple open levels could be recorded at +100 mV (Figure 1(c)). The lower diagram is an extension of the underlined part of the upper diagram. Seven open levels were recorded which means at least seven channels existed in the patch. There were 3-7 open levels in each patch (data not shown), which indicated the density of BK Ca was high in colonic SMCs.

Mechanosensitivity of BK Ca in Rat Colonic
SMCs. BK Ca were continuously activated at +60 mV on cell-attached patches and membrane stretch was elicited by applying negative suction pressures ranging from −10 mmHg to −50 mmHg for 30 s with interval of 15 s. A representative patch is shown in Figure 2(a). The expanded data show the amplitude of the single-channel current was not changed while the channel activities increased as the increase of the negative suction pressures (Figure 2(b)). −10 mmHg and −20 mmHg had little effects on channel activities.
was increased obviously in response to −30 mmHg, −40 mmHg, and −50 mmHg negative suctions (Figure 2(c); = 5, P < 0.05). These results suggest that BK Ca in rat colonic SMCs could be activated by membrane stretch in a pressure-dependent manner.

Role of BK Ca in the Myogenic Autoregulation of Colonic
Smooth Muscle Strips Enduring Stretch. We examined the dose-dependent effects of ChTX, TEA, and TTX on the muscle tone and contraction amplitude of colonic smooth muscle strips ( Figure 3). Contraction amplitude increased after blocking BK Ca by ChTX and the affect was dose-dependent (Figures 3(a) and 3(b); = 6, P < 0.05). 100 nM ChTX had the same maximal affects as 200 nM, while blocking BK Ca by ChTX had no effect on the muscle tone (Figures 3(a) and 3(c); = 6, P > 0.05). Contraction amplitude increased after blocking potassium channels by TEA (Figures 3(d) and 3(e); = 6, P < 0.05). Blocking potassium channels with TEA had no effect on the muscle tone (Figures 3(d) and 3(f); = 6, P > 0.05). BK Ca may be activated by both depolarization and increased cytoplasmic calcium during the active contraction process to contribute to repolarization and to regulate the contraction of colonic smooth muscle as a kind of negative feedback. Voltage-gated sodium channel blocker TTX was used to block the effect of enteric nervous system in this study. TTX had no effects on the contraction amplitude or on the muscle tone of colonic smooth muscle strips.
Effects of BK Ca on the tension regulation induced by stretch were investigated. After equilibrating for 60 min, the colonic smooth muscle strips were exposed to ChTX (100 nM) or TEA (10 mM) for 15 min to block BK Ca . Then, the strips were stretched. Specifically the strip was elongated 1 mm for 10 min each time with a total 5 mm extension exerted on the strip. As shown in Figure 4(a), with the application of stretch stimulation, the muscle tone increased to reach a peak value following a gradual decrease to a relatively stable level. The ratio of stable muscle tone and stretched peak tension in the same step was calculated to reflect the capacity of colonic smooth muscle relaxation. A higher ratio reflected a decreased capability to relax. The ratio was significantly increased by ChTX and TEA in both longitudinal muscle strips (Figures 4(b), 4(c), and 4(d)) and circular muscle strips (Figure 4(e)) from the second-time stretch to the fifth-time stretch. These results indicated that BK Ca were involved in the relaxation of GI tract induced by stretch.
To exclude the influence of the enteric nervous system, TTX was used to block nerve activities and then the previous process was repeated (Figures 5(a) and 5(b)). TTX (100 nM) did not affect the ratio of muscle tone/peak tension in longitudinal or circular muscle strips enduring stretch (Figures 5(f) and 5(g)). The muscle strips were exposed to ChTX, IbTX, or TEA after blocking the nervous system by TTX and then stretched ( Figures 5(c), 5(d), and 5(e)). The statistical results showed the ratio significantly increased in both longitudinal and circular muscle strips (Figures 5(h) and 5(i)) from the second-time stretch to the fifth-time stretch in ChTX, IbTX, and TEA groups. As a broad-spectrum blocking agent of Kv, TEA can block a variety of potassium channels including BK Ca channels. ChTX, usually used as a nonspecific BK Ca blocker, can also block other potassium channels such as intermediate conductance calcium-activated potassium channels (IK Ca ) [19]. Because there are many other potassium channels besides BK Ca in colonic SMCs [20][21][22], the BK Ca specific blocker IbTX was used to confirm the role of BK Ca and the results were similar to that of the TTX + ChTX group ( Figure 5). These results indicated that activation of BK Ca in colonic SMCs by stretch may be the underlying mechanism of the stretch-induced relaxation of the colonic smooth muscle strips.

Discussion
Many parts of the GI tract serve a reservoir function. The increase of GI contents does not cause a sustained rise in cavity pressure, which suggests relaxation may be induced by expanding GI wall, thus playing a role of storage [2]. Stretchdependent increase in colon compliance might come from enteric nervous system that inhibits electrical excitability and contractile processes [10], while stretch stimulation applied on the GI wall may cause direct relaxation of GI smooth muscle without the involvement of enteric nervous system. However, the related mechanisms are unclear. Studies have shown that SMCs of GI tract can respond to mechanical stimulation directly [4,9,23]. Mechanosensitive ion channels in the SMCs may play an important role in the process of sensing and transducing of mechanical stimuli. Mechanosensitive ion channels found in GI SMCs include swelling-activated chloride channels [24], stretch-activated nonselective cation channels [25], and calcium channels [26]. These channels are activated to generate inward currents and induce smooth muscle contraction. In the colon, stretch stimulation does not cause obvious contractile response and activation of outward currents may be necessary for stabilizing resting potential to remain in the relaxed state, yet little is known about the molecular basis [4,27].
It has been known that BK Ca exist in colonic SMCs. However, there are few studies comparing the expression of BK Ca in different GI segment. We found that the expression levels of BK Ca in colon, which has the storage function, are significantly higher than other GI segments, especially the small intestine, the main function of which is motion and absorption. Patch-clamp results also indicate the high density of BK Ca in colonic SMCs. The abundant distribution suggests that BK Ca may play an important role in colonic motility regulation.
Mechanosensitivity of BK Ca has been reported [15][16][17][18]. In our previous studies, we have demonstrated that stretch is a BK Ca gating factor independent of transmembrane potential and intracellular Ca 2+ [18]. In this study, BK Ca currents in rat colonic SMCs were recorded in cell-attached mode with 140 mM potassium in pipette solution and extracellular solution making the resting potential at around 0 mV. Suction pressure was applied to a small area of cell membrane contacted with the electrode. This is a conventional method for the study of mechanosensitivity of ion channels [28,29]. We found BK Ca in rat colonic SMCs could be activated by stretch intensity dependently. Although it is difficult to evaluate the physiological relevance of the pressures and holding potentials used, the mechanosensitivity of BK Ca is conclusive. However, as calcium sensitive potassium channels, BK Ca currents with intracellular calcium signals detection together will provide more valuable information. According to our previous [18] and current studies, BK Ca are highly expressed in colonic SMCs both in rats and mice and are of very similar mechanosensitivity. There are no differences between species, suggesting a common and important role of BK Ca in regulating colonic motility. Membrane hyperpolarization induced by mechanical activation of BK Ca may well explain smooth muscle relaxation rather than contract in response to stretch.
It was reported that BK Ca was involved in the mobility regulation of colon and there was an increase in Bethanecholinduced contraction amplitude in colon with impaired BK Ca [30]. To confirm the role of BK Ca in colon motility regulation, muscle tone and contraction amplitude of colonic smooth muscle strips were recorded. In the current study, blocking BK Ca by ChTX enhanced active contraction but had no effects on muscle tone under basic condition. It is reasonable to speculate that BK Ca were activated by both depolarization and increased cytoplasmic calcium as the action potential occurred. BK Ca may play a negative feedback role to repolarize membrane potential and stop calcium inflow. While there were no BK Ca activities at rest, ChTX had no effects on muscle tone.
Even though there are many researchers suggesting that BK Ca was involved in the mobility regulation of colon, the role of BK Ca in the regulation of muscle tension induced by stretch stimulation is seldom reported at organ level. Stretchinduced relaxation but not contraction of colonic smooth muscle is of significance for its storage role. Our previous   study showed that stretch is an independent gating factor for BK Ca activation [18]. However, it is unknown whether BK Ca can be activated at tissue level by stretch independent on the occurrence of active contraction and plays any role in muscle tone regulation of colon enduring stretch. Tension response of colon to stretch and the role of BK Ca in it were studied in the current study. The results showed that stretch induced relaxation of rat colonic strips. The relaxation capability of rat colonic muscle strips enduring stretch was impaired by blocking BK Ca with ChTX. There was no significant difference between the ChTX and TEA groups which indicated BK Ca was the main component of potassium channels involved in the process of tension regulation induced by stretch in colon. This result indicates that BK Ca may be activated by stretch directly to stabilize resting potential and maintain the relaxed state. It is reported that localized distension activates enteric reflex to evoke GI relaxation [8,11,12]. However, our previous study provides evidence that BK Ca in SMCs can be activated by stretch directly without the involvement of enteric nervous system [18]. The responses of colonic smooth muscle strips to stretch were studied in vitro, excluding effects of external venous and humoral factors. TTX was used to block enteric nervous activities. Results showed the relaxation response of colonic strips to stretch was not changed by TTX. Blocking of BK Ca using ChTX or IbTX still increased the muscle tone/peak tension ratio of colonic smooth muscle strips being stretched, suggesting the nervous system did not appear to be necessary and the activation of BK Ca had a role in this process.
In conclusion, we report, for the first time, that BK Ca are abundantly expressed in the colon more than other GI tracts, and there is stretch-induced relaxation in colon and mechanosensitive BK Ca function in it. The sensation and transduction of mechanical signals have important impacts on GI motility functions. Our study highlights a novel mechanism of GI motility controlling. The disturbed mechanical environment may be key pathogenic factors in GI motility disorders involving organ distention and the mechanosensitivity of BK Ca is worthy of more attention.
those who qualify for authorship are listed. Jie Ren and Fang Xin contributed equally to this work.