As the increasing incidence of diabetes mellitus (DM), it has currently aroused social concern worldwide.¹ DM, which is a chronic metabolic syndrome with hyperglycemia results in various gastrointestinal (GI) complications, such as diarrhea, gastroparesis, and chronic constipation.^2,3 It has been reported that more than 20% of diabetic patients in Europe and Hong Kong suffer from chronic constipation, which exerted serious effects on patients’ quality of life.^4,5 Given the rising incidence of diabetes-induced chronic constipation and the lack of effective treatments, it is of great significance to elucidate the mechanism of this complication to identify an appropriate treatment. Potential mechanisms for diabetes-induced chronic constipation refer to colonic motor dysfunction.⁶ A previous study by our laboratory has shown that type 3 small-conductance calcium (Ca²⁺)-activated potassium (K⁺) (SK3) channels, expressed in interstitial cells in muscular layers of colons, play a strong part in diabetes-induced constipation.⁷ Thus, the present study focuses on SK3 channels in the colonic muscle layers.

SK channels are coassembled complexes of α-subunits, that forming the pore and the Ca²⁺-binding protein, calmodulin (CaM), which lead to the voltage-independence and calcium gating.^8-10 Indeed, the regulatory subunits of protein kinase CK2 (CK2) and protein phosphatase 2A (PP2A) form a polyprotein complex with the SK channel and regulate the characteristics of calcium gating through phosphorylating or dephosphorylating CaM.^11-13 There are 3 subtypes of SK channels, SK1, SK2, and SK3, which are widely expressed in various tissues and have essential roles in regulating neuronal firing, blood flow, and cell proliferation.^14,15 SK2 and SK3 have more sensitivity to the selective blocker, apamin.¹⁶ Among these channels, SK3 channels are predominantly expressed in platelet-derived growth factor receptor alpha-positive (PDGFRα⁺) cells in colonic muscular layers.¹⁷ Due to the expression of platelet-derived growth factor receptor, one class of interstitial cells is named PDGFRα⁺ cells, which are also named fibroblast-like cells. Normal GI motility is mediated by the coordination of neurons, smooth muscle cells, and interstitial cells, including interstitial cells of Cajal (ICCs) and PDGFRα⁺ cells.¹⁸ Previous studies have shown that PDGFRα⁺ cells have an important function in the transduction of purinergic signals from enteric motor neurons as they express the P2Y1 purinergic receptor and SK3 channels.^19,20 Activation of SK3 channels in PDGFRα⁺ cells causes hyperpolarization, which propagates to smooth muscle cells nearby through gap junctions and leads to the relaxation of SMCs.²¹ Therefore, SK3 channels are fundamental for PDGFRα⁺ cells to conduct signals from neurons. Normal GI motility is regulated by the coordination of the enteric neural system, interstitial cells, and smooth muscle cells.^22,23 Accordingly, to reveal the mechanisms of diabetes-induced constipation, it is of great significance to investigate the properties of the SK3 channels in PDGFRα⁺ cells.

In streptozotocin (STZ)-induced type I diabetic mice, previous study in our lab demonstrated that the proliferative PDGFRα⁺ cells in colonic muscle layers exhibit elevated expression, which is modulated by the upregulation of forkhead box O3 through inhibiting the phosphatidylinositol-4,5-bisphophate 3-kinase/protein kinase B (PI3K/Akt) signalling pathway.²⁴ In addition, we also found enhanced the P2Y1 receptor/SK3 channel signalling pathway in PDGFRα⁺ cells in the colons of diabetic mice. As the P2Y1 receptor/SK3 channel signaling pathway in PDGFRα⁺ cells transduces purinergic signals and inhibits the excitability of smooth muscle cells and colonic motility, the enhanced signaling pathway contributes to colonic slow transit, which eventually leads to diabetes-induced chronic constipation.⁷ Nevertheless, the properties of SK3 channels in PDGFRα⁺ cells from diabetic mice are still ambiguous.

In the present research, experiments are meant to investigate the SK3 currents of PDGFRα⁺ cells in an engineered mouse model, which expresses enhanced green fluorescent protein (eGFP) in PDGFRα⁺ cells. We compared the capacitance and SK3 current density of isolated PDGFRα⁺ cells from control and STZ-treated mice. Furthermore, the SK3 current density of HEK293 cells expressing SK3 channels under normal and high oxidative stress conditions was observed. The protein expression of the PP2A and CK2 regulatory subunits in colonic muscle layers and in HEK293 cells was also detected using western blot analysis. Our data suggested that the CK2 regulatory subunit may lead to changes in the sensitivity of SK3 channels to Ca²⁺ in PDGFRα⁺ cells in the colonic muscular layers from STZ-induced mice.

Ethical Approval

The principles of experiments and animal maintenance all abided by the recommendations of the Guide for the Care and Use of Laboratory Animals of the Science and Technology Commission of PRC (STCC Publication No. 2, revised 1988). The procedures on mice were approved by the Committee on the Ethics of Animal Experiment of Shanghai Jiaotong University School of Medicine (Permit No. Hu 686-2009).

Animals

Pdgfr^{atm11(EGFP)Sor}/J heterozygous mice were obtained from the Jackson Laboratory (Bar Harbour, ME, USA). In Pdgfr^{atm11(EGFP)Sor}/J heterozygous mice, PDGFRα⁺ cells were labelled by the histone 2B-eGFP fusion protein driven by the endogenous, cell-specific Pdgfra promoter. All mice were bred under standard conditions, which is the constant room temperature with a 12-hour–12-hour light-dark cycle and free for water and food.

Diabetic Type 1 Mouse Model With Streptozotocin

Pdgfr^{atm11(EGFP)Sor}/J heterozygous mice were randomly separated into 2 groups as follows: a control group and an STZ-treated group which are fasted for 12 hours. The STZ-treated mice received a single intraperitoneal injection (200 mg/kg) of STZ, which was dissolved in ice-cold 0.1 M citrate buffer. The control mice were intraperitoneally administered the equal volumes of citrate buffer. The level of blood glucose was detected 7 days after the injection of STZ and then tested again 2 months later to confirm the high blood glucose level (the value of blood glucose should be above 16.7 mmol/L).

Isolation of Platelet-derived Growth Factor Receptor Alpha-positive Cells

Pdgfr^{atm11(EGFP)Sor}/J heterozygous mice were executed by the dislocation of cervical vertebra when anaesthetized by isoflurane. After opening the abdomens, the colons were taken out. The mucosa and submucosa layers were peeled away, and the smooth muscular layers were sheared into pieces of approximately 1 mm × 8 mm. The colonic muscle pieces were equilibrated in physiological salt (PSS) solutions with no Ca²⁺ for 30 minutes at 37_℃, and the cells were then dispersed in Ca²⁺-free PSS solutions containing 5 mg/mL collagenase type II (CLS-2; Worthington Biochemical Corporation, Lakewood, NJ, USA), 8 mg/mL trypsin inhibitor (T9128; Sigma-Aldrich, St. Louis, MO, USA) and 8 mg/mL bovine serum albumin (B2046; Sigma-Aldrich). After approximately 20 minutes, the digested muscles were gently rinsed and further digested for 20-30 minutes. After digestion, the colonic muscle layers were gently triturated with a fire-polished glass pipette. The cell suspension was placed on glass coverslips coated with murine collagen in 35 mm culture dishes. The freshly isolated cells were supposed to settle for 30 minutes in SMGM culture medium and cultured in a humidified incubator with 5% CO₂ and the temperature of 37_℃. PDGFRα⁺ cells were used for experiments within 6 to 8 hours after plating.

Patch Clamp Experiments

Freshly isolated PDGFRα⁺ cells and HEK 293 cells were plated in a 0.3 mL chamber on the object stage of an inverted microscope (IX-70; Olympus, Tokyo, Japan) and perfused with external solutions. The patch-clamp technique of whole cell was used for recording membrane currents with an EPC-10 HEKA amplifier (HEKA Instrument, Berlin, Germany). Resistance of the pipette tips was 3 to 6 mΩ for whole-cell recordings. All these patch clamps were carried out at the temperature of 20-25_℃. The time for recording the currents of PDGFRα⁺ cells and HEK293 cells was 20 minutes after the cells were raptured.

Measurement of Malondialdehyde and Superoxide Dismutase

The malondialdehyde (MDA) levels are measured by Lipid Peroxidation MDA Assay Kit-Thiobarbituric Acid (TBA) method (Beyotime Institute of Biotechnology, Nantong, China). Assay Kit-WST method were used to evaluate superoxide dismutase (SOD) activity (Beyotime Institute of Biotechnology). MDA levels and total SOD activity in colonic tissues were normalized to total protein.

Cell Culture

HEK 293 cells were provided by the Institute of Biochemistry and Cell Biology (Chinese Academy of Science, Shanghai, China). For the culture of HEK293 cells, the procedures are in the Song et al.⁷ The transfection of HEK293 cells were performed with Lipo6000 Transfection Reagent (C0526; Beyotime Chemical Co, Jiangsu, China) and used for patch clamp analysis 24 hours after the transfection. To confirm the transfections of SK3 channel, we used the antibody of SK3 to label the SK3 channels transfected in HEK293 cells. After placing the cells on the cover glasses, fixing, blocking and labeling with first antibody (anti-SK3 antibody; 1:500; ab192515; Abcam, Cambridge, MA, USA) were used in immunohistochemistry. Then the secondary antibody (Alexa Fluor 488-labeled goat anti-rabbit IgG; 1:500; A0423; Beyotime Chemical Co) was used. The SK3 proteins labelled by antibodies with fluorescence are the marker for confirming the transfection. HEK 293 cells were treated with hydrogen peroxide (H₂O₂) (100 mM) for 24 hours to induce oxidative stress and then used for patch clamp analysis.

Western Blot

Colonic muscle layers or HEK 293 cells were homogenized in immunoprecipitation assay buffer (1:100; P0013; Beyotime Chemical Co) containing protease inhibitor cocktail (1:100; P1005; Beyotime Chemical Co). The protein samples were 40 mg per lane. The membranes were probed with anti-CK2 (1:1000; 06-873; Merck Millipore, MA, USA), anti-PP2A (1:1000; 05-421; Merck Millipore) antibodies, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:2000; 5174; Cell Signaling Technology).

Solutions and Drugs

PSS with Ca²⁺, which is the external solution for whole-cell recordings contained the following components: 5 mM KCl, 135 mM NaCl, 2 mM CaCl₂, 10 mM glucose, 1.2 mM MgCl₂, and 10 mM HEPES and then use Tris to adjust to pH 7.4. The low Ca²⁺ solution involved the components listed below: 135 mM KCl, 3 mM MgATP, 0.1 mM NaGTP, 2.5 mM creatine phosphate disodium, 0.1 mM EGTA and 10 mM HEPES and then use Tris to adjust to pH 7.2. The high Ca²⁺ pipette solution with the 500 nM of Ca²⁺ contained the following components: 135 mM KCl, 7.74 mM CaCl₂, 3 mM MgATP, 0.1 mM NaGTP, 2.5 mM creatine phosphate disodium, 10 mM EGTA, and 10 mM HEPES and then use Tris to adjust to pH 7.2. The SK3 channel antagonist Apamin and the SK3 channel agonist CyPPA were provided by Tocris Bioscience (Ellisville, MO, USA).

Statistical Methods

Values are the means ± SEM of n cells or n mice. We used P < 0.05 as the threshold to define the statistical significance. The calculations of the statistical analyse are performed by GraphPad Prism software (San Diego, CA, USA). The ANOVA with Bonferroni’s post hoc test or Student’s unpaired t test was used to compare between groups when appropriate.

Blood Glucose and Body Weight

We recorded the body weights of the control and STZ-treated mice, and the data showed that the body weight was significantly reduced in STZ-treated mice (22.87 ± 0.29 g of control mice; 15.04 ± 0.21 g of STZ-treated mice; P < 0.05; n = 50; Fig. 1A). Additionally, the averaged level of blood glucose of the STZ-treated mice 2 months after injection was 25.75 ± 0.79 mmol/L (P < 0.05; n = 50; Fig. 1B), which was much higher than that in the control mice (6.28 ± 0.14 mmol/L).

Figure 1. Streptozotocin (STZ)-induced diabetic mice. (A) Summary of the data which showed the body weights of control and STZ-treated mice (*P < 0.05 compared with control mice; n = 50). (B) Summary of the blood glucose levels between control and STZ-treated mice (*P < 0.05 compared with control mice; n = 50).

Type 3 Small-conductance Calcium-activated Potassium Currents in Platelet-derived Growth Factor Receptor Alpha-positive Cells

As the high expression of SK3 channels in PDGFRα⁺ cells in colonic muscle layers, the activation of this conductance may generate Ca²⁺-activated K⁺ currents, which could decrease the excitability of smooth muscle cells via the gap junctions between PDGFRα⁺ cells and SMCs.²⁵ In PDGFRα⁺ cells, the whole-cell configuration of the patch clamp was used to study the characteristics of SK3 channels. To identify the PDGFRα⁺ cells in the mixed cell dispersions, an engineered mouse model was used, which expresses eGFP protein in PDGFRα⁺ cells.²¹ PDGFRα⁺ cells, which were enzymatically dispersed from the colonic muscle, were round and had bright green nuclei under a fluorescence microscope. We found that the step depolarization from –80 mV to +70 mV caused small amplitude currents in PDGFRα⁺ cells dialyzed with a low Ca²⁺ pipette solution (< 10 nM) in Fig. 2A. Furthermore, dialysis of cells with a high Ca²⁺ pipette solution containing 500 nM Ca²⁺ resulted in large amplitude, time-dependent outward currents (Fig. 2B). Figure 2C shows that ramp depolarizations developed outward currents in PDGFRα⁺ cells dialyzed with the 500 nM Ca²⁺ pipette solution, which was significantly decreased by apamin (300 nM). These data suggested that the PDGFRα⁺ cells dispersed from the colonic muscles generated apamin-sensitive outward Ca²⁺-activated K⁺ currents in response to ramp or step depolarizations.

Figure 2. Type 3 small-conductance calcium-activated potassium current density in platelet-derived growth factor receptor alpha-positive (PDGFRα⁺) cells. (A, B) Representative curves for PDGFRα⁺ cells in response to step potentials from a holding potential of –80 mV to +70 mV with low calcium ion (Ca²⁺) (< 10 nM) pipette solution (containing < 10 nM Ca²⁺, A) and high Ca²⁺ pipette solution (containing 500 nM Ca²⁺, B). (C) Ramp protocols (–80 to +80 mV, 800 msec) were used to PDGFRα⁺ cells in the whole-cell patch clamp with high Ca²⁺ pipette solutions (containing 500 nM Ca²⁺). (D) The I-V curves of net outward current density in PDGFRα⁺ cells with low Ca²⁺ (< 10 nM) pipette solution (black line) and 500 nM Ca²⁺ pipette solution (red line).

Membrane Capacitance of Platelet-derived Growth Factor Receptor Alpha-positive Cells

To study the current density of SK3 channels in PDGFRα⁺ cells, we recorded the membrane capacitance of these cells from the control and STZ-treated mice with the patch clamp under whole-cell configuration. We found that the membrane capacitance of normal PDGFRα⁺ cells was 4.54 ± 0.29 pF and that of diabetic PDGFRα⁺ cells was 4.76 ± 0.18 pF, which suggested that no significant differences were observed in PDGFRα⁺ cells from the control and STZ-treated mice (not shown in the Figure).

Type 3 Small-conductance Calcium-activated Potassium Currents of Platelet-derived Growth Factor Receptor Alpha-positive Cells From STZ-treated Mice

For comparison of properties of SK3 channels in PDGFRα⁺ cells from the normal mice and the diabetic mice, the current density was recorded using a whole-cell patch clamp. Dialysis of PDGFRα⁺ cells using low Ca²⁺ solution resulted in small and steady currents in both the control mouse cells and the STZ-treated mouse cells (Fig. 3A and 3B). The results showed that the outward current density in PDGFRα⁺ cells from control mice was 38.7 ± 4.04 pA/pF at +70 mV. However, the SK3 current density in cells from STZ-treated mice was 16.3 ± 1.36 pA/pF at +70 mV, that was much smaller than that in cells from control mice (P < 0.05; n = 4; Fig. 3C and 3D). These data demonstrated when dialyzed with low Ca²⁺ solution, the SK3 current density is significantly reduced in PDGFRα⁺ cells from diabetic mice, which may result from decreased SK3 expression in PDGFRα⁺ cells or decreased Ca²⁺ sensitivity to SK3 channels.

Figure 3. Comparison of type 3 small-conductance calcium-activated potassium current density in platelet-derived growth factor receptor alpha-positive (PDGFRα⁺) cells from normal and diabetic mice dialyzed with low calcium ion (Ca²⁺) (< 10 nM) pipette solution. (A) Representative current traces in response to step depolarizations in normal PDGFRα⁺ cells with low Ca²⁺ (< 10 nM) pipette solution. (B) Representative current traces of diabetic PDGFRα⁺ cells dialyzed with low Ca²⁺ (< 10 nM) pipette solution when exposed to step depolarizations. (C) Summary of net outward current density in PDGFRα⁺ cells from normal (black line) and streptozotocin (STZ)-treated mice (red line). (D) Summarized data of net outward current density at +70 mV in PDGFRα⁺ cells from normal and diabetic mice. Data are shown as the mean ± SEM (n = 4; *P < 0.05).

It has been reported that SK3 channels are typically activated by the increased concentration of intracellular Ca²⁺ and then regulate membrane potentials. Thus, in our experiment, we dialyzed freshly isolated PDGFRα⁺ cells with a solution containing 500 nM Ca²⁺ to evoke SK3 currents. We found that dialysis of cells from control mice evoked large amplitude currents when the cells were exposed to step depolarization from –80 mV to +70 mV. In contrast, the step depolarizations developed larger amplitude outward currents in PDGFRα⁺ cells from STZ-treated mice (Fig. 4A-C). The SK3 current density in PDGFRα⁺ cells at +70 mV from control mice was 66.8 ± 8.65 pA/pF. However, the SK3 current density at +70 mV in PDGFRα⁺ cells from STZ-treated mice was 90.4 ± 2.24 pA/pF (P < 0.05; n = 4; Fig. 4D). These data demonstrated that the sensitivity of SK3 channels to Ca²⁺ in diabetic PDGFRα⁺ cells was decreased. The increased SK3 current density in diabetic PDGFRα⁺ cells may result from the upregulated expression of SK3 channels.

Figure 4. Comparison of type 3 small-conductance calcium-activated potassium current density in platelet-derived growth factor receptor alpha-positive (PDGFRα⁺) cells from normal and diabetic mice dialysed with 500 nM calcium ion (Ca²⁺) pipette solution. (A) Representative current traces in response to step depolarizations in normal PDGFRα⁺ cells with 500 nM Ca²⁺ pipette solution. (B) Representative current traces of diabetic PDGFRα⁺ cells dialyzed with 500 nM Ca²⁺ pipette solution when exposed to step depolarizations. (C) Summary of net outward current density in PDGFRα⁺ cells from normal (black line) and streptozotocin (STZ)-treated mice (red line). (D) Summarized data of net outward current density at +70 mV in PDGFRα⁺ cells from normal and STZ-treated mice. Data are shown as the mean ± SEM (n = 4; *P < 0.05).

Increased Oxidative Stress in Diabetic Colons

To further explore the mechanism of the decreased sensitivity of SK3 channels to Ca²⁺, we measured and compared the levels of MDA and superoxide dismutase (SOD) activity in the colonic muscular layers from normal and diabetic mice. In normal colonic tissues, the MDA content averaged 0.53 ± 0.02 nmol/mg protein. However, the MDA level in diabetic colonic tissues averaged 0.71 ± 0.05 nmol/mg protein. These data showed that the MDA level had an increasing tendency in diabetic colons compared to normal colons (n = 3; Fig. 5A). Moreover, the SOD activity in diabetic colonic tissues was 0.47 ± 0.03 unit, which was significantly elevated compared to that in normal colonic tissues (ie, 0.31 ± 0.01 unit; n = 5; P < 0.05; Fig. 5B). These data suggested that the diabetic colons were in an increased oxidative stress state, which may result in decreased sensitivity of SK3 channels to Ca²⁺.

Figure 5. Comparison of malondialdehyde (MDA) level and superoxide dismutase (SOD) activity in colonic tissues from normal and streptozotocin (STZ)-treated mice. (A) Summary of the MDA concentration in colonic muscular layers from normal and STZ-treated mice (n = 3). (B) Summary of SOD activity in colonic tissues from normal and STZ-treated mice. Data are shown as the mean ± SEM (n = 5; *P < 0.05). NS, not significant.

Type 3 Small-conductance Calcium-activated Potassium Currents in HEK293 Cells

According to the increased oxidative stress state in the colon tissues of diabetic mice, we investigated whether the increased oxidative stress may result in different SK3 currents. H₂O₂ (100 nM) was used to induce oxidative stress in cultured HEK 293 cells. Figure 6A and 6B show that when dialysed with low Ca²⁺ pipette solution, HEK 293 cells revealed small amplitude currents with an average current density of 13.7 ± 0.65 pA/pF at +70 mV, and H₂O₂-treated HEK 293 cells revealed much smaller amplitude currents with an average current density of 9.0 ± 1.09 pA/pF at +70 mV (n = 4; P < 0.05). Additionally, dialysis of HEK 293 cells with a pipette solution containing 500 nM Ca²⁺ resulted in outward currents with the current density averaging 28.9 ± 1.67 pA/pF at +70 mV. However, dialysis of H₂O₂-treated HEK 293 cells with a pipette solution containing 500 nM Ca²⁺ resulted in outward currents with a current density averaging 43.3 ± 1.73 pA/pF at +70 mV, which was greater than the currents from normal HEK 293 cells (n = 4; P < 0.05; Fig. 6C and 6D). These data showed that when dialyzed with low Ca²⁺ pipette solution, H₂O₂ treatment decreased the current density of SK3 channels in HEK 293 cells. Whereas the H₂O₂ treatment increased the current density of SK3 channels in HEK 293 cells dialysed with high Ca²⁺ pipette solution. These results demonstrated that oxidative stress may contribute to the decreased sensitivity of SK3 channels to Ca²⁺.

Figure 6. Comparison of type 3 small-conductance calcium-activated potassium current density in normal and hydrogen peroxide (H₂O₂)-treated HEK293 cells. (A) Summary of net outward current density evoked by step depolarizations in normal and H₂O₂-treated HEK293 cells with low calcium ion (Ca²⁺) (< 10 nM) pipette solution. (B) Summarized data of the net outward current density at +70 mV in dialyzed normal and H₂O₂-treated HEK293 cells (n = 4; P < 0.05). (C) Summary of net outward current density evoked by step depolarizations in normal and H₂O₂-treated HEK293 cells with high Ca²⁺ (500 nM) pipette solution. (D) Summarized data of net outward current density at +70 mV in PDGFRα⁺ cells from normal and diabetic mice. Data are shown as the mean ± SEM (n = 4; *P < 0.05).

Protein Phosphatase 2A and Protein Kinase CK2 in HEK293 Cells

PP2A and CK2 are regulatory proteins in SK3 channels that form a complex with SK3 channels and modulate Ca²⁺ sensitivity by regulating the characteristics of Ca²⁺ gating through phosphorylating/dephosphorylating CaM. Therefore, it is of great importance to investigate the alterations in PP2A and CK2 in HEK 293 cells in response to oxidative stress stimulation. In the present study, CK2 expression in H₂O₂-treated HEK 293 cells was elevated to 1.37 ± 0.12 (P < 0.05; n = 5; Fig. 7A) compared to control HEK 293 cells. However, no significant differences were observed in the expression of PP2A between the control and H₂O₂-treated HEK 293 cells (P = 0.22; n = 5; Fig. 7B). These results suggested that the changes in the CK2 regulatory proteins may contribute to the decreased Ca²⁺ sensitivity of SK3 channels in H₂O₂-treated HEK 293 cells.

Figure 7. Protein kinase (CK2) and protein phosphatase 2A (PP2A) expressions in colonic muscular layers from normal and streptozotocin (STZ)-treated diabetic mice. (A, B) The relative expressions of CK2 and PP2A in control and hydrogen peroxide (H₂O₂)-treated HEK293 cells (normalized to glyceraldehyde-3-phosphate dehydrogenase [GAPDH]; n = 5, *P < 0.05). (C, D) The relative expressions of CK2 and PP2A in colonic muscle layers from control and STZ-treated mice (normalized to GAPDH; n = 5, *P < 0.05). NS, not significant.

Protein Phosphatase 2A and Protein Kinase CK2 in Streptozotocin-treated Mice

Western blotting was next used to detect the expressions of PP2A and CK2 in the colonic muscular tissues from normal and diabetic mice. According to the data in Figure 7, the expression of CK2 protein in STZ-treated mice was increased to 1.39 ± 0.11 (P < 0.05; n = 5; Fig. 7C) compared to normal mice. In addition, the expression of PP2A was not changed in the colon tissues of STZ-treated mice (P = 0.20; n = 6; Fig. 7D) compared to normal mice. These data showed that in diabetic mice, the upregulated expression of CK2, rather than PP2A, may lead to the decreased Ca²⁺ sensitivity of SK3 channels.

Chronic constipation is a severe GI complication secondary to diabetes mellitus. This study was meant to explore the mechanism of changed SK3 channel characteristics in STZ-treated type I diabetic mice, which may play important roles in colonic dysmotility in diabetes-induced constipation. A previous study has shown that GI interstitial cells, both ICCs and PDGFRα⁺ cells, are postjunctional cells and mediate regulatory inputs from enteric motor neurons.²² The highly integrated behaviours of GI motility require perfect coordination among smooth muscle cells, ICCs and PDGFRα⁺ cells.²⁶ Our previous investigation found up-regulated expression of PDGFRα⁺ cells in the colons from diabetic mice, resulting from the upregulation of forkhead box O3 through inhibiting the PI3K/Akt signalling pathway.²⁴ In addition, our lab also found elevated expression of PDGFRα⁺ cells, SK3 channels and P2Y1 receptors in muscular layers of diabetic colons, leading to the functional upregulation of the P2Y1 receptor/SK3 channel signalling pathway. The enhanced P2Y1 receptor/SK3 channel signalling pathway has stronger effects on inhibiting the colonic motility, eventually resulting in colonic slow transit in diabetic mice.⁷ Moreover, the state of high oxidative stress, which is induced by diabetes, may have an important impact on GI complications.^27,28 Thus, high oxidative stress may contribute to changes in SK3 channel function, but additional investigation is required for confirmation. As SK3 channels play key roles in the inhibitory effects of PDGFRα⁺ cells from the purinergic signals in the GI tract, the electrophysiological characteristics of SK3 channels in diabetic mice should be studied in detail to explore the mechanisms of GI dysmotility.

Our results showed that the SK3 currents in PDGFRα⁺ cells dialyzed with a low Ca²⁺ pipette solution from diabetic colons were dramatically decreased. Moreover, dialysis of diabetic PDGFRα⁺ cells with a pipette solution containing 500 nM Ca²⁺ revealed SK3 currents with larger amplitudes. These results demonstrated that SK3 expression in PDGFRα⁺ cells was upregulated but that the sensitivity of SK3 channels to Ca²⁺ was decreased. Furthermore, the levels of MDA and SOD activity in diabetic colonic tissues were increased, which showed that high oxidative stress occurred in the colons of diabetic mice. Additionally, we used 100 nM H₂O₂ to treat SK3 channel-expressing HEK 293 cells to mimic the enhanced oxidative stress state in diabetes. H₂O₂-treated HEK 293 cells also showed decreased SK3 currents when dialyzed with the low Ca²⁺ pipette solution and increased SK3 current density when dialyzed with the pipette solution containing 500 nM Ca²⁺. These findings suggested that oxidative stress may lead to changed SK3 channel characteristics in STZ-treated type I diabetic mice. Moreover, the CK2 regulatory subunit was upregulated both in H₂O₂-treated HEK 293 cells and colonic tissues from diabetic mice, suggesting that increased CK2 have a strong part in regulating the sensitivity of SK3 channels to Ca²⁺ in colonic PDGFRα⁺ cells from diabetic mice. These data indicated that high oxidative stress leads to increased CK2 expression, thereby mediating the decreased calcium sensitivity of SK3 channels in PDGFRα⁺ cells, which eventually plays a leading role in colonic dysmotility in STZ-induced diabetic mice.

SK channels are highly expressed in many tissues, including neural tissues, vascular endothelium, and smooth muscles.^14,29 In addition, according to the features of SK channels, the voltage-independence and Ca²⁺ gating, and they have a high sensitivity to Ca²⁺.^30,31 In the GI tract, SK3 channels are mainly expressed in PDGFRα⁺ cells. Studies have shown that when purines released by enteric motor neurons act on PDGFRα⁺ cells, they can activate SK3 channels by increasing the concentration of intracellular Ca²⁺, which leads to the generation of transient outward currents that are apamin-sensitive.²¹ The SK3 currents of PDGFRα⁺ cells could diffuse to adjacent smooth muscle cells, hyperpolarize the resting membrane potentials and decrease the excitability of SMCs.³² Regarding the significance of SK3 currents in regulating colonic motility, the present study investigated the changes in SK3 channels in the colons of normal and diabetic mice. First, we dialyzed PDGFRα⁺ cells with solutions containing different concentrations of Ca²⁺. In response to step depolarization protocols from –80 mV to +70 mV, dialysis of PDGFRα⁺ cells with low Ca²⁺ pipette solution revealed small amplitude SK3 currents. When we dialyzed PDGFRα⁺ cells with 500 nM Ca²⁺ pipette solution, step depolarization protocols elicited large amplitude, outward SK3 currents, which were inhibited by apamin. These data showed that the SK3 channels in PDGFRα⁺ cells are activated by the increase in intracellular Ca²⁺.

For the comparison of characteristics of SK3 channels in colonic muscular layers between normal and diabetic mice, the SK3 currents in freshly isolated PDGFRα⁺ cells from normal and diabetic mice were recorded separately in this study. Whole-cell patch techniques showed that dialysis of diabetic PDGFRα⁺ cells with low Ca²⁺ pipette solution resulted in much smaller amplitude SK3 currents than that of normal PDGFRα⁺ cells. These results suggested that the decreased density of SK3 currents in diabetic PDGFRα⁺ cells might be caused by the decreased sensitivity to Ca²⁺ of SK3 channels or the decreased expression of SK3 channels, but further research is required for confirmation. A previous study has suggested that the expression of SK3 channels is elevated in colonic muscle layers in diabetic mice.⁷ In addition, we recorded the SK3 current density of diabetic PDGFRα⁺ cells when dialyzed with a pipette solution containing 500 nM Ca²⁺. The data showed that the dialysis of diabetic PDGFRα⁺ cells with the pipette solution containing 500 nM Ca²⁺ resulted in a larger amplitude of SK3 currents, which indicated that the expression of SK3 channels was not decreased in diabetic PDGFRα⁺ cells. These results demonstrated that the sensitivity of SK3 channels to calcium was decreased in PDGFRα⁺ cells from diabetic mice.

Experimental evidence shows that oxidative stress damage is a leading factor related to the GI complications of diabetes.^27,28 The elevated MDA levels and SOD activities demonstrated that high oxidative stress occurred in the STZ-induced diabetic mice. It has been known that the out-of-balance of the ROS production and antioxidant defences leads to the oxidative stress damage. Furthermore, it has been reported that the addition of low levels of ROS induces apoptosis.³³ An enhanced oxidative stress state has been recognized to be involved in serious cellular damage.²⁷ Hence, we investigated whether enhanced oxidative stress state plays a role in the decreased sensitivity of SK3 channels to calcium in diabetic PDGFRα⁺ cells. In the present study, we used 100 nM H₂O₂ to treat SK3 channel-expressing HEK 293 cells to mimic the increased oxidative stress state in diabetic mice. The present study showed that H₂O₂-treated HEK 293 cells generated smaller amplitude SK3 currents in response to step depolarization when dialysed with low Ca²⁺ pipette solution in comparison with normal HEK 293 cells. Additionally, dialysis of the H₂O₂-treated HEK 293 cells with a pipette solution containing 500 nM Ca²⁺ resulted in larger amplitude SK3 currents compared to control HEK 293 cells. These results showed that high oxidative stress decreases the calcium sensitivity of SK3 channels expressed in HEK 293 cells, suggesting that the enhanced damage of oxidative stress may contribute to decreased sensitivity of SK3 channels to Ca²⁺ in diabetic mice.

Otherwise, very little study is designed to investigate the mechanism by which oxidative stress changes the properties of SK3 channels. It has been reported that the CK2 and PP2A regulatory subunits determine the Ca²⁺ gating properties of the SK channel via phosphorylation or dephosphorylation of CaM.³⁰ CK2 phosphorylates CaM and reduces the calcium sensitivity of SK channels. In addition, PP2A dephosphorylates CaM and increases sensitivity to intracellular Ca²⁺.^13,14,34 Thus, we measured the expression of CK2 and PP2A to explore whether CK2 and PP2A participated in the change of calcium sensitivity of SK3 channels under enhanced oxidative stress conditions. In our study, the expression of CK2 was increased in both HEK 293 cells treated by H₂O₂ and in diabetic mouse colonic muscle layers. However, PP2A expression in H₂O₂-treated HEK 293 cells and in diabetic mouse colonic muscle layers was not significantly changed. These results suggested that the upregulation of CK2 induced by increased oxidative stress may lead to the decreased calcium sensitivity of SK3 channels in diabetic mice.

In summary, SK3 channels in PDGFRα⁺ cells exert a great impact on the resting membrane potential of smooth muscle cells and have key roles in regulating GI motility. In PDGFRα⁺ cells of colonic muscular layers from diabetic mice, calcium sensitivity of SK3 channels was decreased. In addition, H₂O₂-induced high oxidative stress led to the decreased Ca²⁺ sensitivity of SK3 channels in transgenic HEK 293 cells, indicating that high oxidative stress contributes to the decreased calcium sensitivity of SK3 channels in colonic PDGFRα⁺ cells from diabetic mice. Moreover, the regulatory subunit of SK3 channels, CK2, was upregulated in both H₂O₂-treated HEK 293 cells and diabetic mouse colonic muscle layers. According to the CK2 function in decreasing the calcium sensitivity of SK3 channels, these results suggest that the upregulation of CK2 modulates SK3 channel Ca²⁺ sensitivity in the high oxidative stress state. To conclude, this study showed that the high oxidative stress-induced upregulation of CK2 has key roles in the decreased calcium sensitivity of SK3 channels in colonic PDGFRα⁺ cells from diabetic mice, which may have implications for elucidating the mechanisms of GI dysmotility induced by diabetes.

This work was supported by the National Natural Science Foundation of China (31871158 and 31671192) and Foundation of XinHua Hospital (JZPI201708).

None.

Ni-Na Song: conception and design of experiments; collection, analysis, and interpretation of data; drafting of the article; and critical revision of the paper for important intellectual content. Xu Huang, Hong-Li Lu, and Chen Lu: collection, analysis, and interpretation of data. Jie Chen: conception and design of experiments; and analysis and interpretation of data. Wen-Xie Xu: conception and design of experiments; analysis and interpretation of data; and critical revision of the paper for important intellectual content. All authors approved the final version of the manuscript.