Quercetin Inhibits Pacemaker Potentials via Nitric Oxide/cGMP-Dependent Activation and TRPM7/ANO1 Channels in Cultured Interstitial Cells of Cajal from Mouse Small Intestine

Quercetin (3,3',4',5,7-pentahydroxyl-flavone) is one of the most common flavonoids and is present in various foods, such as, onions, apples, broccoli, tea, and red wine [1,2,3,4]. Quercetin has a variety of activities in man. For example, it has been reported to have antioxidant [5,6,7], anti-inflammatory [8], anti-anaphylactic [9], anti-aging [10], antiplatelet aggregatory [11], radioprotective [12], immunomodulatory [13], antidiabetic [14], and anticancer [15] effects, and to improve lipid metabolism [16,17] and endurance exercise capacity [18]. Quercetin also reduces gastrointestinal (GI) intestinal transit in mice [19], induces intestinal smooth muscle relaxation [20], and prevents castor oil-induced diarrhea [21]. These properties are consistent with the use of flavonoids as antidiarrheal drugs in traditional medicine. However, the cellular and molecular mechanisms underlying the GI motility effects of quercetin are not well understood.

Interstitial cells of Cajal (ICCs) are the GI pacemaker cells that generate rhythmic oscillations in membrane potentials known as slow waves [22,23], and ICC loss has been implicated in several motility disorders, which suggests that they play an important role in the regulation of GI motility [24]. In addition, ICCs mediate inhibitory and excitatory signals from the enteric nervous system to smooth muscle [25,26,27], and evidence indicates that endogenous agents, such as, neurotransmitters, hormones, and paracrine substances, modulate GI tract motility by influencing ICCs [28,29,30,31]. Furthermore, it has been shown that the pacemaker activities of ICCs in the murine small intestine are mainly due to periodic activations of nonselective cation channels (NSCCs) [32,33] or Cl^- channels [34,35,36]. Kim et al. [24] suggested transient receptor potential melastatin (TRPM) 7 is required for ICC pacemaker activity in the murine small intestine, and that a Ca²⁺ -activated Cl^- channel (CaCC) is involved in the slow waves generated by ICCs; this Cl^- channel was later identified as transmembrane protein 16A (Tmem16A; anoctamin1 (ANO1)) [35]. The role played by ANO1 in ICC pacemaker activity was identified in ANO1 knockout mice [37], and several other authors have since shown that ANO1 participates in pacemaker activity [38,39]. Kim et al. [24,40] showed that TRPM7 mRNA and TRPM7 currents are expressed in cultured ICCs, and Shahi et al. [40,41] showed that ANO1 mRNA and ANO1 currents are also expressed in cultured ICCs. Accordingly, it has been proposed that TRPM7 and ANO1 be considered potential targets for the pharmacological treatment of GI motility disorders.

Quercetin plays a vital role in the GI system [42,43,44,45,46], but relatively little is known about its effects on the pacemaker activities of ICCs in the GI tract or on GI motility-related ion channel activities. In the present study, we investigated the inhibitory effects of quercetin on the pacemaker potentials of cultured ICCs, and characterized quercetin-mediated signaling pathways. In addition, we investigated the inhibitory effects of quercetin on GI motor functions by measuring the intestinal transit rates (ITR) of Evans blue in normal mice.

Animal care and experiments were conducted in accordance with the guidelines issued by the ethics committee of Pusan National University (Republic of Korea). Balb/c mice were used throughout the study. Small intestines (from 1 cm below the pyloric ring to the cecum) were removed and opened along the mesenteric border. Luminal contents were removed using Krebs-Ringer bicarbonate solution, and the tissues obtained were pinned to the base of Sylgard dishes. Mucosae were removed by sharp dissection and small tissue strips of intestine muscle (consisting of circular and longitudinal muscles) were equilibrated for 30 min in Ca²⁺-free Hank's solution (containing (in mM); KCl 5.36, NaCl 125, NaOH 0.34, Na₂HCO₃ 0.44, glucose 10, sucrose 2.9 and HEPES 11; pH 7.4). Cells were then dispersed in an enzyme solution containing collagenase (Worthington Biochemical, Lakewood, NJ, USA; 1.3 mg ml^-1), bovine serum albumin (BSA, Sigma-Aldrich, St Louis, MO; 2 mg ml^-1), trypsin inhibitor (Sigma-Aldrich; 2 mg ml^-1), and ATP (0.27 mg ml^-1), and plated onto sterile glass coverslips coated with murine collagen (2.5 µg ml^-1; Falcon/BD, Franklin Lakes, NJ, USA) in 35 mm culture dishes. Cells were then cultured at 37°C in a 95% O₂-5% CO₂ incubator in smooth muscle growth medium (SMGM; Clonetics, San Diego, CA, USA) supplemented with 2% antibiotics/antimycotics (Gibco, Grand Island, NY, USA) and murine stem cell factor (SCF; 5 ng ml^-1; Sigma-Aldrich). All experiments on ICC clusters were performed after culture for 12 hrs. ICCs were identified immunologically using an anti-c-kit antibody (phycoerythrin (PE)-conjugated rat anti-mouse c-kit monoclonal antibody; eBioscience, San Diego, CA) at a dilution of 1:50 for 20 min (Fig. 1). ICCs were morphologically distinct from other cell types in the culture and thus it was possible to identify the cells by phase contrast microscopy once they had been verified with anti c-kit antibody.

The physiological salt solution used to bathe cultured ICC clusters (Na⁺-Tyrode) contained (in mM): KCl 5, NaCl 135, CaCl₂ 2, glucose 10, MgCl₂ 1.2, and HEPES 10 (adjusted to pH 7.4 with NaOH). The pipette solution used to examine pacemaker activity contained (in mM): KCl 140, MgCl₂ 5, K₂ATP 2.7, NaGTP 0.1, creatine phosphate disodium 2.5, HEPES 5, and EGTA 0.1 (adjusted to pH 7.2 with KOH). Patch-clamp techniques were conducted in whole-cell configuration to record membrane currents (voltage clamp mode) and potentials (current clamp mode) from cultured ICCs using Axopatch I-D and Axopatch 200B amplifiers (Axon Instruments, Foster, CA). Command pulses were applied using an IBM-compatible personal computer and pClamp software (version 6.1 and version 10.0; Axon Instruments). Data were filtered at 5kHz and displayed on an oscilloscope, a computer monitor, and/or a pen recorder (Gould 2200; Gould, Valley View, OH, USA). Results were analyzed using pClamp and Origin software (version 6.0, Microcal, USA). All experiments were performed at 30-33°C.

A mammalian expressible plasmid for mouse ANO1/TMEM16A (pEGFP-N1-mANO1, ac isoform) was generously provided by Dr. Min Goo Lee (Yonsei University, Korea). Plasmids were transiently transfected into human embryonic kidney (HEK) 293T cells using Lipofectamine Plus Reagent (Life Technologies, USA), according to the manufacturer's instructions. Briefly, HEK293T cells were cultured in DMEM supplemented with 10% fetal bovine serum and 0.5% penicillin-streptomycin at 37°C in a 90% O₂-10% CO₂ incubator. One day before transfection, cells were transferred on glass coverslips. For electrophysiological experiments, HEK293T cells were transfected with 0.9 μg of mANO1 plasmid and 0.1 μg of a plasmid expressing green fluorescent protein (pEGFP-N1). Cells transfected with 1 μg of pEGFP-N1 were used as negative controls. HEK-293T cells were routinely studied at 21-25°C, 24 h after transfection. The bath solution contained 146 mM HCl, 10 mM HEPES, 10 mM glucose, 1 mM MgCl₂, 1 mM CaCl₂, and 150 mM N-methyl-D-glucamine (NMDG), adjusted to pH 7.4. The pipette solution contained 150 mM NMDG-Cl, 1 mM MgCl₂, 3 mM MgATP, 10 mM EGTA, 5 mM CaCl₂ and 5 mM HEPES at pH 7.2 (titrated with NMDG). The free calcium concentration was fixed at 200 nM using WEBMAX-C software (C. Patton, Stanford University, www.stanford.edu/~cpatton/maxc.html).

HEK-293T cells transfected with the Flag-murine LTRPC7/pCDNA4-TO construct were grown on glass coverslips in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum, blasticidin (5 μg/ml), and zeocin (0.4 mg/ml). TRPM7 (LTRPC7) expression was induced by adding 1 μg/ml tetracycline to culture medium. Whole-cell patch-clamp experiments were performed at 21-25°C 24 h after induction using cells grown on glass coverslips. The internal pipette solution for whole cell current recordings contained (in mM) 145 Cs-methanesulfonate, 8 NaCl, 10 Cs-2-bis(2-aminophenoxy)-ethane-N,N,N′,N′-tetraacetic acid, and 10 HEPES-CsOH adjusted to pH 7.2 with CsOH. The standard extracellular Tyrode's solution contained (in mM) 145 NaCl, 2.8 KCl, 2 CaCl2, 1.2 MgCl2, 10 HEPES, and 10 glucose, adjusted to pH 7.4 with NaOH. All experiments were performed at 21-25°C.

ICCs were preincubated with 100 μM IBMX for 30 min at 37°C to inhibit cAMP and cGMP degradation and then incubated with quercetin (200 μM) for 10 min. Cells were then homogenized in buffer containing 4 mM EDTA to prevent the degradations of enzymatic cAMP or cGMP and homogenates were then heated for 5 min in a boiling water bath to coagulate protein. After centrifugation at 3,000 rpm for 5 min, supernatants were transferred to new tubes and stored at 4°C. Samples were assayed for cAMP or cGMP using cAMP or cGMP ELISA kits (Enzo Life Science, Farmingdale, NY, USA).

The effects of quercetin on intestinal motility were assessed by measuring the intestinal transit rates (ITRs) of Evans blue solution (5 %, w/v, in DW). Briefly, 30 min after the intragastric (i.g.) administration of quercetin to normal mice, 0.1 ml of Evans blue solution was administered i.g. through an orogastric tube. Animals were sacrificed 30 min after this administration, and intestinal transit was determined by measuring the distance that the Evans blue had migrated in intestines from the pylorus. ITR (%) were calculated by expressing the distance traveled by the dye as a percentage of the total length of small intestine (from the pylorus to the ileal end). To minimize possible inter-day variations, all measurements were performed on same days.

Quercetin was purchased from Alexis Biochemicals (Vinci Biochem, Florence, Italy) and naloxone from TOCRIS (USA). All other drugs were obtained from Sigma. Stock solutions were prepared and stored, according to the manufacturer's instructions. Chemicals were dissolved in Na⁺-Tyrode solution to their final concentrations immediately before use.

Results are expressed as means ± SEMs. Statistical analysis was performed using the student's t test or by analysis of variance (ANOVA) followed by Tukey's multiple comparison test, as appropriate. The analysis was performed using GraphPad Prism version 6, and p values of <0.05 were considered statistically significant. The n values reported in the text refer to the number of cells used in patch-clamp experiments.

Cultured ICC clusters had a mean resting membrane potential of -51 ± 5 mV and produced electrical pacemaker activity at a frequency of 17 ± 2 cycles per minute and an amplitude of 25 ± 4 mV (n = 40) at 30°C in current clamp mode. Quercetin at concentrations from 100 to 200 μM decreased pacemaker amplitude and frequency in a concentration-dependent manner with an IC₅₀ of ∼88.3 μM (Fig. 2).

To investigate which potassium channels mediate quercetin-induced pacemaker activity, we used various potassium channel blocker types. The treatment of ICCs with tetraethylammonium chloride (TEA) (10 mM; a Ca²⁺ -activated K⁺ channel blocker) or BaCl₂ (50 μM; an inward rectifying K⁺ channel blocker) had no effect on pacemaker activities, and in the presence of TEA or BaCl₂, quercetin still inhibited pacemaker activities (n = 5; Fig. 3Aa, 3Ab and 3B). In addition, treatments with glibenclamide (10 μM; an ATP-sensitive K⁺ channel blocker), 4-aminopyridine (5 mM; a transient voltage-dependent K⁺ channel blocker), or apamin (1 μM; a Ca²⁺-activated K⁺ channel blocker) had no effect on pacemaker activity, and quercetin still inhibited pacemaker activities when co-treated with glibenclamide, 4-aminopyridine, or apamin (n = 5; Fig. 3Ac, 3Ad, 3Ae and 3B).

Opioid receptor antagonists, such as, naloxone, are used to treat bowel diseases [47], and numerous studies have described the pharmacological effects of naloxone on GI motility in vitro in guinea pigs, rodents, and humans [48,49,50,51,52]. Therefore, we investigated the relationship between quercetin and opioid receptors by examining the effect of naloxone (a pan-opioid receptor antagonist) on pacemaker activity inhibition by quercetin. In the absence of naloxone, quercetin decreased the pacemaker activity amplitude and frequency (Fig. 2), but naloxone prevented pacemaker activity inhibition by quercetin (Fig. 4A) (pacemaker activity amplitude was 24 ± 4 mV in the presence of naloxone (Fig. 4B)).

To investigate the signaling mechanisms involved and the roles played by G proteins during pacemaker activity inhibition by quercetin, we applied GDPβS (1 mM; a non-hydrolysable guanosine 5′-diphosphate analogue that permanently inactivates GTP binding proteins [53,54]) to the patch pipette solution, and this resulted in the prevention of pacemaker activity inhibition by quercetin (n = 5; Fig. 5A). Mean depolarizations were 24.6 ± 1.2 mV and in the presence and absence of GDPβS (Fig. 5B).

To investigate whether pacemaker activity inhibition by quercetin is mediated by a cyclic nucleotide-dependent pathways, we used SQ-22536, an inhibitor of adenylate cyclase, and 1H-[1, 2, 4]oxadiazolo[4, 3-a]quinoxalin-1-one (ODQ), an inhibitor of guanylate cyclase. Preincubation of ICCs with SQ-22536 (100 μM) for 10 min had no effect on pacemaker activity, and in the presence of SQ-22536, quercetin (100 μM) still inhibited pacemaker activity (n = 6; Fig. 6Aa and 6B). However, ODQ (100 μM) blocked pacemaker activity inhibition by quercetin (n = 6; Fig. 6Ac and 6B). Also, to see whether the pacemaker activity inhibition by quercetin was mediated by cAMP-dependent protein kinase A (PKA) or cGMP-dependent protein kinase G (PKG), we examined the effects of myristoylated PKA inhibitor (an inhibitor of PKA) and RP-8-CPT-cGMPS (an inhibitor of PKG). Preincubation of ICCs with myristoylated PKA inhibitor (1 μM) had no effect on pacemaker activities, and in the presence of myristoylated PKA inhibitor, quercetin inhibited pacemaker activities (n = 5; Fig. 6Ab and 6B). However, in the presence of RP-8-CPT-cGMPS (10 μM), quercetin failed to inhibit pacemaker activities (n = 5; Fig. 6Ad and 6B). Nitric oxide (NO) activates soluble guanylyl cyclase, which results in the formation of cGMP and in the activation of PKG [55]. To investigate whether pacemaker activity inhibition by quercetin is mediated by NO, we added L-NAME (100 μM; a non-selective NO synthases (NOS)) inhibitor, and found that it blocked pacemaker activity inhibition by quercetin (n = 5; Fig. 6Ae and 6B). In addition, to determine whether changes in cAMP and cGMP contents are involved in the inhibitory effect of quercetin on ICCs, intracellular cAMP and cGMP contents were measured under basal and quercetin-stimulated conditions. Quercetin did not stimulate cAMP production (Fig. 7A; control: 11.1 ± 1.2 vs. quercetin: 12.3 ± 1.4 pmol/mg-1 protein), but stimulated cGMP production (Fig. 7B; control: 11.3 ± 1.1 vs. quercetin: 13.8 ± 1.3 pmol/mg-1 protein).

It has been reported ANO1 and TRPM7 are involved in slow wave generation and that they are required for ICC pacemaker activity in the murine small intestine [24,35]. Furthermore, TRPM7 and ANO1 mRNA have been reported in ICCs and TRPM7 and ANO1 currents have been described [24,35,40,41]. Thus, we examined the effects of quercetin on the expressions of ANO1 and TRPM7 channels in HEK 293 cells, which overexpress both channels. To investigate whether ANO1 channels are involved in the pacemaker activity of ICCs, we performed whole-cell patch clamp recordings on hANO1-transfected HEK293T cells using a 200 nM free Ca²⁺ pipette solution to activate ANO1 currents (I_ANO1). Fig. 8A shows representative I_ANO1 recordings and its inhibition by different concentrations of quercetin. After membrane break-in, an outward rectifying Cl^- current was immediately activated by diffusion of the pipette solution. After confirming a steady-state current level, we applied depolarizing step pulses from -100 mV to + 100 mV (Fig. 8A (a)). Using step depolarization, delayed activation of outward currents and a deactivation tail current were observed, which are typical characteristics of I_ANO1 (Fig. 8D) Fig. 8F (a) shows the steady-state I-V relationship from step depolarization of the untreated control (black arrow). Next, we serially perfused cells with quercetin-containing bath solution (100, 200, or 300 μM). Quercetin significantly inhibited I_ANO1 currents. The associated I-V relationship curves at peak I_ANO1 (1) and at quercetin concentrations of 100 (2), 200 (3), or 300 μM (4) were obtained using a ramp-like pulse protocol from -100 mV to + 100 mV (Fig. 8B). After treatment with 300 μM quercetin, we again applied the step-pulse, and found that 300 μM quercetin almost completely inhibited I_ANO1 current (Fig. 8E). Relative steady-state I-V relationships are presented in Fig. 8F (b). To compare inhibitory effects on ANO1, we treated cells with 30 μM T16Ainh-AO1 (a selective ANO1 inhibitor) at the end of the experiment as a positive control. To determine inhibitory percentages at each concentration, the normalized amplitudes of the currents (I_ext/I_con × 100%) of quercetin-treated cells were measured at +100 mV. As shown in Fig. 8C, quercetin significantly inhibited I_ANO1 in a dose-dependent manner. TRPM7 currents were elicited by voltage ramps ranging from -100 to +100 mV from a holding potential of -60 mV. After break-in, 2-3 min was allowed for current amplitudes to reach the steady state. The currents displayed a profound outwardly rectifying I-V relationship in Ca²⁺-containing normal bath solution (normal bath), presumably due to the relatively small inward current carried by Ca²⁺ compared to the large outward current carried by Cs⁺ [56,57]. In order to investigate the regulatory effects of quercetin on TRPM7 currents, we applied quercetin at different concentrations. Application of quercetin (100-300 μM) to HEK cells transfected with TRPM7 elicited a rapid decrease in outward and inward whole-cell currents (Figs. 9Aa, 9Ba and 9Ca). The time course of outward current is plotted at +100 mV and inward current at -100 mV for respective cells (n = 7; Figs. 9Ab, 9Bb and 9Cb). These effects were reproducible and fully reversible, and repeated application of quercetin yielded similar reductions in current amplitude. Mean peak inward and outward current densities before and after quercetin application were -2.5 ± 0.5 pA/pF and -1.0 ± 0.3 pA/pF at -100 mV and +155.2 ± 3.2 pA/pF and +74.5 ± 2.1 pA/pF at +100 mV for quercetin at 100 μM, -4.1 ± 0.3 pA/pF and -1.7 ± 0.2 pA/pF at -100 mV and +153.4 ± 1.5 pA/pF and +60.3 ± 2.5 pA/pF at +100 mV for quercetin at 200 μM, and -2.6 ± 0.5 pA/pF and -1.0 ± 0.5 pA/pF at -100 mV and +161.4 ± 2.3 pA/pF and +41.4 ± 2.2 pA/pF at +100 mV for quercetin at 300 μM, respectively (n = 8, mean ± SEM; Fig. 9).

The ITRs (%) of Evans blue over 30 min in normal mice are shown in Fig. 10. The ITR for untreated normal mice (control, n = 25) was 65.3 ± 2.1 %. To examine GI motility regulation by quercetin, the aqueous extract of the dried immature fruits of Poncirus trifoliata Raf. (PF) was used. PF is one of the most popular traditional folk medicines in Korea as is derived from the Rutaceae fruits. It has been shown to have unique and potent prokinetic activities in normal rodents and in rodent models of GI motility dysfunction (GMD) [58,59]. PF significantly increased ITR (%) at a dose of 1 g/kg (by 75.1 ± 1.8 %, p<0.01), which is consistent with previous reports [60,61]. However, the ITR values for quercetin administered at 10 and 100 mg/kg doses were 64.5 ± 1.9 and 49.8 ± 1.7 %, respectively (n = 10; Fig. 10), that is, quercetin reduced the ITR.

In the present study, we found that quercetin decreased the amplitude and frequency of ICC pacemaker activity via a series of opioid receptor signaling pathways, including G protein-cGMP-PKG-NO activation. As ICC pacemaker activities are mainly due to periodic activations of TRPM7 channels [24,33] or ANO1 channels [35,37,38], TRPM7 and ANO1 modulators are potential targets for the treatment of GI motility disorders. In the present study, quercetin inhibited TRPM7 and ANO1 currents, indicating that the TRPM7 and ANO1 channels are involved in the quercetin-mediated inhibition of pacemaker activity in ICCs. In addition, in vivo, quercetin (10-100 mg/kg, p.o.) dose-dependently decreased ITR in normal mice.

With regard to intestinal motility, quercetin has been reported to inhibit the peristalsis of guinea pig ileum in vitro [62] and to depress intestinal motility in a concentration-dependent manner by facilitating inhibitory enteric pathways [63]. Furthermore, quercetin was found to be associated with decreases in some inflammation markers and changes in gut microbiota when administered to healthy mice [64]. In the present study, quercetin inhibited ICCs pacemaker activity and suppressed ITR in mice.

In a previous study, we found that ANO1 was well expressed in ICCs within 12 hours of culture, but subsequently decreased with time [40]. Therefore, in the present study, we used ICCs that had been cultured for least than 12 hours (Fig. 1). Previous reports have shown TRPM7 and ANO1 mRNA are present and TRPM7 and ANO1 currents are expressed in cultured ICCs [24,35,40,41], and thus, we examined the effects of quercetin on ANO1 and TRPM7 channels in HEK 293 cells, which overexpress both channels. Quercetin is the most abundant flavonoid and decreases the ICC pacemaker activity (Fig. 2) and induces G protein-cGMP-PKG-NO sensitive pacemaker potentials (Fig. 5 and 6). We are of the opinion that quercetin might decrease GI motility at the tissue level based on the observations that; 1) quercetin inhibited TRPM7 and ANO1 channels and 2) decreased the pacemaker activities of cluster type ICCs.

We believe that TRPM7 and ANO1 are involved in the upstroke of pacemaker potentials. In ICCs, inhibitory neurotransmitters induce hyperpolarization and reduce the amplitude and frequency of pacemaker activity. In the present study, quercetin also reduced the amplitude and frequency of pacemaker activity. Therefore, we believe that quercetin may have the same functions as inhibitory neurotransmitter. Specific details of ion channel mechanisms during the upstroke and plateau phase of pacemaker potentials have not been elucidated. Therefore, it is difficult to determine with precision the effects of quercetin on pacemaker potentials, additional studies are required to identify the ion channels involved. In addition, we cannot explain the exact mechanism underlying the reduction of motility rate induced by quercetin. However, we suppose because quercetin inhibited TRPM7 and ANO1 channels and decreased pacemaker activity in cluster type ICCs that this inhibition might be due to inhibition of smooth muscle or the enteric nervous system.

In conclusion, we found that quercetin reduced the amplitude and frequency of ICC pacemaker activity in a G protein-, cGMP-, PKG-, and NO-dependent manner via opioid receptor activation. Furthermore, TRPM7 and ANO1 were found to be involved in the quercetin-induced regulation of pacemaker activity in cultured murine ICCs. Finally, our findings suggest that quercetin should be viewed as a novel drug development candidate for the treatment of GI spasm, pain, transit disturbances, and other symptoms related to GI motility disorders.

This work was supported by a National Research Foundation of Korea (NRF) Grant funded by the Korea government (MSIP) (#2014R1A5A2009936).

The authors have no potential conflict of interest to declare.

H. Gim and J.H. Nam contributed equally to this work and should be considered joint first authors.