Propofol inhibits stromatoxin-1-sensitive voltage-dependent K⁺ channels in pancreatic β-cells and enhances insulin secretion

Cells and cell culture

Mouse insulinoma MIN6 cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Grand Island, NY, USA) containing 450 mg/dl glucose, 10% fetal bovine serum (FBS), 50 µM β-mercaptoethanol, 100 U/ml penicillin, and 0.1 mg/ml streptomycin (Miyazaki, 1990 #103). Rat INS-1 cells were cultured in RPMI1640 (Sigma-Aldrich, St Louis, MO, USA), supplemented with 10% FBS, 10 mM HEPES, 2 mM L-glutamine, 1 mM sodium pyruvate, 50 µM β-mercaptoethanol, 100 U/ml penicillin, and 0.1 mg/ml streptomycin (Asfari et al., 1992).

Isolation of mouse pancreatic islets

Male C57BL/6JJcl mice (8–10 weeks old, n = 8) were sacrificed by cervical dislocation in accordance with protocols approved by the Animal Experimentation Committee, Kansai Medical University (#19-088). Pancreatic islets were isolated by enzymatic digestion from the pancreas with a slight modification (Lacy & Kostianovsky, 1967). The pancreas was removed and digested with collagenase (Type IV, 195 U/ml; Worthington Biochemical, Lakewood, NJ, USA) in a solution containing 2 mM glucose and trypsin inhibitor (0.01%; Sigma-Aldrich) at 37 °C for 30 min with vigorous shaking. The pancreatic tissue was triturated with a pipette and washed two times with enzyme-free solution. Islets were selected with a glass micropipette under a stereomicroscope. Batches of ten islets were used for measurement of insulin concentration.

Reagents

Propofol (2,6-diisopropylphenol) and 2,4-diisopropylphenol were obtained from Sigma Aldrich. Glibenclamide was obtained from Wako Pure Chemical Industries, Ltd. (Osaka, Japan), diazoxide from Abcam (Cambridge, MA, USA), and stromatoxin-1 from Alomone Labs Ltd. (Jerusalem, Israel).

Measurement of insulin concentration

Insulin concentration in the culture medium of MIN6, INS-1 cells, and pancreatic β-cells/islets was assessed using the Mouse/Rat Insulin H-type enzyme-linked immunosorbent assay kit (Shibayagi Co. Ltd., Shibukawa, Japan), following the manufacturer’s instructions (Suzuki et al., 2015). Detailed protocols are available in the Supplemental Information and at protocols.io (10.17504/protocols.io.v63e9gn).

Cell growth assay

Cell growth was assessed using the CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega, Madison, WI, USA) (Sumi et al., 2018b; Suzuki et al., 2015). Cells were seeded at a density of 3 × 10³ cells/well in 96-well plates and incubated for 0, 4, and 12 h. Cell viability was determined by comparing the absorbance values of the treated cells with that of the control cells (MIN6 cells at 24 h incubation), with the latter defined as 100%. All experiments were done in triplicate or quadruplicate. Detailed protocols are available as Supplemental Information and at protocols.io (10.17504/protocols.io.v64e9gw).

Caspase activity assays

The activities of caspase-3 and caspase-7 were determined using an Apo-ONE Homogeneous Caspase-3/7 Assay Kit (Promega), according to the manufacturer’s protocols (Okamoto et al., 2016; Sumi et al., 2018b). In brief, cells were seeded at 2 × 10⁴ cells/well on 96-well plates and incubated overnight. Cells were then treated with the experimental concentrations of camptothecin and propofol for varying lengths of time. Caspase activities were determined by comparing the luminescence values of the treated cells with those of the control cells (incubated without drugs), with the latter defined as 100%. All assays were conducted in triplicate and repeated at least twice. Detailed protocols are available as Supplemental Information and at protocols.io (10.17504/protocols.io.v7je9kn).

ATP assay

Intracellular ATP content was evaluated using the CellTiter-Glo luminescent cell viability assay kit (Promega) (Sumi et al., 2018a). In brief, cells were seeded at a density of 3 × 10³ cells/well on 96-well plates and allowed to grow for 1, 4, and 8 h in the either presence or absence of propofol. The relative ATP levels were determined by comparing the luminescence values of the treated cells to those of control cells cultured in 2 mM glucose. Assays were done in triplicate and repeated at least twice. Detailed protocols are available in the Supplemental Information and at protocols.io (10.17504/protocols.io.v7ke9kw).

ADP assay

Intracellular ADP content was evaluated using the ADP Colorimetric/Fluorometric Assay kit (Abcam, Cambridge, MA, USA). In brief, cells were seeded at a density of 6 × 10⁵ cells/well on 12-well plates and allowed to grow for 1, 4 and 8 h in the either presence or absence of propofol. The relative ADP levels were determined by comparing the luminescence values of the treated cells to those of control cells. Assays were done in triplicate and repeated at least twice. Detailed protocols are available in the Supplemental Information and at protocols.io (10.17504/protocols.io.zk4f4yw).

Measurement of cellular oxygen consumption and extracellular acidification

The cellular oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured using an XFp Extracellular Flux Analyzer (Agilent Technologies, Santa Clara, CA, USA) (Sumi et al., 2018a). MIN6 cells were seeded at a density of 1 × 10⁴ cells/well on the XFp Cell Culture microplate. The XF Cell Mito Stress Test was performed in glucose-containing XF base medium, following the manufacturer’s protocol. Detailed protocols are available as Supplemental Information and at protocols.io (10.17504/protocols.io.v92e98e).

Immunoblotting

Whole-cell lysates were prepared by incubating cells for 30 min in cold radioimmune precipitation assay buffer with Complete Protease Inhibitor Cocktail Tablets (Roche Diagnostics, Tokyo, Japan). Samples were then centrifuged at 10,000× g to allow the cell debris to settle, and 30 µg total protein from the resulting supernatant was separated by 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electro-transferred onto membranes, probed with 1:500 (anti-Kv2.1 Ab) or 1:1000 (anti- β-actin Ab) of the indicated primary antibodies, and then probed with 1:8,000 of donkey anti-rabbit IgG (GE Healthcare, Piscataway, NJ) or sheep anti-mouse IgG (GE Healthcare) conjugated with horseradish peroxidase. Blots were visualized with enhanced Chemi-Lumi One Super (Nacalai Tesque, Kyoto, Japan) (Daijo et al., 2016; Tanaka et al., 2011b). Experiments were performed in triplicate, and representative blots are shown. Detailed protocols are available at protocols.io (10.17504/protocols.io.x9mfr46).

Intracellular insulin concentration assay

Whole-cell lysates were prepared by incubating cells for 30 min in cold radioimmune precipitation assay buffer with Complete Protease Inhibitor Cocktail Tablets (Roche Diagnostics, Tokyo, Japan) after three time washing with PBS. Samples were then centrifuged at 10,000 × g to allow the cell debris to settle to obtain total cell lysates. Then the concentration of insulin protein was assayed using the Mouse/Rat Insulin H-type enzyme-linked immunosorbent assay kit (Shibayagi Co. Ltd., Shibukawa, Japan). The concentrations were compensated with total protein weight.

Quantitative reverse transcriptase-PCR analysis (qRT-PCR)

Total RNA was isolated using RNeasy Mini Kit (Qiagen, Valencia, CA, USA). First-strand cDNA synthesis and real-time PCR were performed as described previously (Sumi et al., 2018a). PCR primers are listed in Table S2. Detailed protocols are available in the Supplemental Information and from protocols.io (10.17504/protocols.io.x9mfr46).

Electrophysiological studies

MIN6 cells were incubated in an extracellular bath solution containing 2 mM glucose at 37 °C for 30 min before patch-clamp experiments (Hayashi et al., 2016; MacDonald et al., 2002; Zhang et al., 2016). Membrane potential measurements and whole-cell current recordings were performed using the EPC 800 patch-clamp amplifier (HEKA Elektronik Inc. Holliston, MA, USA). Experiments were conducted at 23–30 °C. Detailed protocols are available in the Supplemental Information and from protocols.io (10.17504/protocols.io.v68e9hw).

Statistical analysis

Data are presented as means ± standard deviation (SD). Differences between groups were evaluated by one-way analysis of variance (ANOVA) and two-way ANOVA followed by Dunnett’s test or Tukey’s test for multiple comparisons. Statistical analyses were performed with Prism7 (GraphPad Software, Inc. La Jolla, CA). Statistical significance was defined by P-values <0.05 (Sumi et al., 2018a).

Effects of propofol on insulin secretion

The response of insulin secretion to extracellular glucose concentration was examined in mouse MIN6 cells. The cells were cultured in 2 mM glucose and exposed to a range of glucose concentrations (2–20 mM) for 1 h. GSIS was detected in response to increased glucose concentrations (Fig. S1A). The time-dependent profile of GSIS was also investigated. The cells were exposed to 20 mM glucose and the GSIS was assayed at 0, 20, 40, 60, and 120 min (Fig. S1B). To compensate for the unknown influence of cell status on insulin measurement, insulin secretion is usually compensated with total protein weight. We measured the insulin concentration with or without compensation with the protein weight. As shown in Fig. S1C–S1E, the compensation did not impact insulin concentration measurement. Depending on the experimental data, we quantified insulin concentration without the compensation by total protein weight in this study in the case of MIN6 and INS1 cells. The insulin secretion in response to the polysaccharides sucrose and maltose was also assessed (Fig. S1F). Only glucose elicited insulin secretion, indicating other polysaccharides do not largely impact insulin secretion (IS).

Plasma concentrations of propofol in clinical settings during anesthesia and sedation are reported to range between 11 µM and 200 µM (Ludbrook, Visco & Lam, 2002; Vanlander et al., 2015). MIN6 cells were incubated with 5–100 µM of propofol in 2 mM glucose from 30 min to 4 h and then insulin concentrations were assessed. Although 5 µM and 100 µM propofol did not significantly increase basal IS, 10 µM propofol significantly increased IS after 1 h and 4 h incubation, 25 µM propofol significantly increased IS after 1 h incubation, and 50 µM propofol significantly increased IS after 30 min and 1 h incubation (Figs. 1A–1C). When cells were exposed to 20 mM glucose, 5 µM propofol did not significantly increase GSIS after any incubation period, whereas 10 µM propofol significantly increased GSIS after 1 h and 4 h incubation, 25 µM propofol significantly increased GSIS after 30 min and 1 h incubation and 50 µM propofol increased GSIS after 1 h incubation (Figs. 1D–1F). In contrast, 100 µM propofol significantly decreased GSIS compared to the control, after 1 h and 4 h incubation (Figs. 1D–1F).

Next, we tried to examine whether this GSIS induction by propofol was reversible. MIN6 cells were exposed to 25 µM propofol with 20 mM glucose for 1 h followed by incubation in 2 mM glucose without propofol for 6 h prior to re-exposure to 20 mM glucose under indicated concentrations of propofol (Fig. 1G). Statistically significant differences were not observed in GSIS of MIN6 cells with or without pre-treatment with 25 µM propofol. The evidence indicates that the enhancement effect of propofol on GSIS by less than 50 µM propofol is reversible. Next, we examined the effect of the isomer of propofol, 2,4-diisopropylphenol, which does not have hypnotic effects (Tsuchiya et al., 2010). We found that 5–50 µM 2,4-diisopropylphenol enhanced IS and GSIS, in a similar manner to propofol, after 1 h incubation (Figs. 1H and 1I). Therefore, our results demonstrate that the effect of propofol on GSIS is biphasic and is both dose- and time-dependent.

Effects of propofol on insulin secretion induced by glibenclamide or inhibited by diazoxide

Elevation of intracellular ATP concentration ([ATPi]) in response to higher glucose conditions closes ATP-sensitive potassium (K_ATP) channels and depolarizes the plasma membrane (Seino, 2012). The channel closer glibenclamide facilitates insulin secretion in pancreatic β-cells even in low-glucose conditions, while the channel opener diazoxide inhibited insulin secretion in both 2 mM and 20 mM glucose (Seino, 2012).

The 25 µM and 50 µM propofol treatments significantly increased IS induced by 100 µM glibenclamide in 2 mM glucose compared to control with no propofol or glibenclamide treatment, whereas 10 µM propofol had no effect (Fig. 2A). In 10 mM glucose, propofol at doses of 10 µM, 25 µM, and 50 µM enhanced GSIS induced by glibenclamide treatment (Fig. 2B). On the other hand, 10, 25, or 50 µM propofol treatment alleviated 100 µM diazoxide-elicited suppression of GSIS in 10 mM glucose (Figs. 2C and 2D).

The evidence strongly suggested that propofol impacts other pathways that are dependent on K_ATP channels sensitive to glibenclamide and diazoxide.

Effects of propofol on GSIS in rat INS-1 cells and mouse pancreatic β-cells/islets

We next examined the effect of propofol on rat pancreatic β-cell-derived INS-1 cells. Propofol at doses of 25 µM enhanced IS under 2 mM glucose within 30 min, 1 h, and 4 h (Figs. 3A–3C). Under 20 mM glucose conditions, 25 µM propofol enhanced GSIS at all incubation periods (Figs. 3D–3F).

β-cells/islets were incubated with 25 µM of propofol in 2 mM glucose from 30 min to 1 h and then insulin concentration was assessed. In the case of β-cells/islets, insulin concentrations were compensated with total protein weight. As in the case of MIN6 and INS-1 cells, 25 µM propofol significantly increased IS after 1 h incubation (Fig. 3G). When cells were exposed to 20 mM glucose, 25 µM propofol significantly increased GSIS after 1 h incubation. Thus, propofol increased insulin secretion in only cell lines as well as primary mouse pancreatic β-cells/islets (Fig. 3G).

Impact of propofol on proliferation and death of MIN6 cells

MIN6 cells growth under propofol treatment was examined using the MTS assay (Fig. 4A). Propofol treatment at 100 µM significantly suppressed the growth rate of MIN6 cells. The impact of propofol on cell death was also investigated. Camptothecin at 5 µM activated caspase 3/7 within 4 h, whereas 50 µM propofol did not activate caspase 3/7 within the same period. However, 100 µM propofol significantly activated caspase 3/7 at 4 h and 12 h incubation (Fig. 4B). Cell death was also assayed by the trypan blue exclusion method and was found to be significantly induced by 1 mM lidocaine. However, neither 50 µM nor 100 µM propofol induced cell death within the 12 h incubation period (Fig. 4C).

Our findings show that propofol at 50 µM did not affect proliferation or death of MIN6 cells, while propofol at 100 µM did after 12 h incubation.

Effects of propofol on cellular energy metabolism

Intracellular ATP and ADP are generally assumed to play a crucial role in GSIS (Rorsman, 1997; Seino, 2012). Extracellular high glucose increases ATP in pancreatic β-cells (Seino, 2012). We investigated the effect of propofol on ATP under exposure to 2 mM or 20 mM glucose for 1, 4, and 8 h. With 2 mM glucose, 25 µM propofol did not affect ATP. In contrast, 100 µM propofol decreased ATP at 4 and 8 h exposure (Fig. 5A). With 20 mM glucose, 100 µM propofol decreased ATP at all exposure periods (Fig. 5B). Next, we investigated the effect of propofol on ADP (Fig. S2). In conditions where propofol did not affect ATP, the same concentration of propofol did not affect ADP as well.

The impact of propofol on oxygen metabolism of MIN6 cells was investigated. Stimulation with high glucose activated mitochondrial respiration in MIN6 cells. In 5 mM and 20 mM glucose, OCR increased in a time-dependent manner up to 4 h (Fig. S3A). ECAR also increased in a time-dependent manner (Fig. S3B). Neither 25 µM nor 50 µM propofol affected OCR in 20 mM glucose within the 4 h period of study. However, 100 µM propofol suppressed OCR (Fig. 5C). Therefore, our results show that clinically relevant doses of propofol do not affect cellular oxygen or energy metabolism.

Impact of propofol and glucose concentration on expression of GLUT2, Cav1.2, Kir6.2, Kv2.1, SUR1, and insulin

The expression of glucose transporter 2 (GLUT2) channels including the voltage-dependent calcium channel Cav1.2, the Kir6.2 and SUR1 subunit of K_ATP, and the voltage-dependent potassium channel Kv2.1 and insulin was investigated after 1 h and 4 h incubation. The mRNA expression of these channels was not impacted by glucose concentration or by propofol within 1 h or 4 h (Figs. 6A–6F). Additionally, we investigate the mRNA expression at 12 h and 48 h. Incubation with 25 µM propofol did not affected the mRNA expression up to 48 h (Fig. S4).

Because the RT-PCR results clearly indicated that Kv2.2 mRNA was barely expressed in MIN6 cells compared to Kv2.1 (Fig. S5), we focused on evaluating the protein expression of Kv2.1. Immunoblot assay demonstrated the protein expression of Kv2.1 channel was not affected by propofol treatment (Fig. 6G and Fig. S6). Intracellular insulin contents were not affected by the treatment (Fig. 6H). The protein expression of Kv2.1 or insulin was not affected by 12 h or 48 h treatment as in the case of mRNA expression (Fig. S4). These results demonstrate that the expression of molecules that play critical roles in glucose intake or membrane depolarization is not affected by propofol at clinically relevant doses.

Effects of propofol on the membrane potential of MIN6 cells

We measured the membrane potential of MIN6 cells using gramicidin-perforated patch techniques to avoid dialyzing the intracellular components. In this configuration, the intracellular concentration of glucose is preserved, because the area of gramicidin-perforated patch membrane is much smaller than that of the whole-cell membrane. Propofol significantly increased the frequency of the action potential in 10 mM glucose from 0.32  ± 0.15 s⁻¹ to 0.84 ± 0.42 s⁻¹ at 10 µM and 1.08 ± 0.43 s⁻¹ at 25 µM (Fig. 7A; n = 5). Similar results were obtained with glibenclamide (100 µM). Application of additional propofol (25 µM) significantly increased the frequency of the action potential from 0.35 ± 0.34 s⁻¹ with glibenclamide to 0.59 ± 0.23 s⁻¹ (Fig. 7B; n = 4). Treatment with propofol also increased the duration of the action potential: 225.9 ± 104.0 ms with glibenclamide (i in Figs. 7B and 7C; n = 4) and 366.0 ± 179.9 ms with additional propofol (ii in Figs. 7B and 7D; n = 4). We measured whole-cell currents under the same conditions. The application of propofol (100 µM) decreased the slope conductance in a voltage range between −83 and −63 mV from 0.84  ± 0.10 to 0.23 ± 0.12 nS (Fig. 7E; n = 3). However, propofol had a negligible effect on the inward conductance. These results indicated that propofol affected the voltage-dependent outward conductance.

Effects of propofol on voltage-dependent outward potassium currents in MIN6 cells

We measured whole-cell currents with K⁺-rich pipette solution containing 3 mM ATP using voltage-clamp configuration and recorded voltage-dependent outward K⁺ currents in MIN6 cells bathed in 10 mM glucose buffer (Fig. 8A). Typical traces in response to the Kv2.1/2.2-specific antagonist stromatoxin-1 (100 nM) are shown in Fig. 8B. Both peak and sustained K⁺ currents were inhibited by stromatoxin-1 (Fig. S7, n = 6). Propofol at 50 µM significantly decreased K⁺ currents from 137.0 ± 19.4 pA/pF to 91.1 ± 22.4 pA/pF at +52 mV (Fig. 8C, n = 6). The current–voltage relationships of sustained K⁺ currents showed that propofol significantly inhibited the voltage-dependent outward K⁺ currents in a dose-dependent manner (Fig. 8D). The typical traces in responses to propofol are shown in Fig. S8 . The Ki value for the effect of propofol was estimated at 83.6 ± 13.0 µM with a Hill coefficient of 1.3 ± 0.3 (Fig. 8E). The Ki value was not affected by voltage (Fig. 8F). Additionally, similar results were obtained with the isomer 2,4-diisopropylphenol. The Ki value for the effect of 2,4-diisopropylphenol was estimated at 41.1 ± 5.4 µM with a Hill coefficient of 1.5 ± 0.2 (Fig. S9).

We next examined the impact of stromatoxin-1 on IS of MIN6 cells (Figs. 8G–8H) and β-cells/islets (Fig. 8I) with both 2 mM and 20 mM glucose and Stromatoxin-1 affected IS in 2 mM glucose, and enhanced GSIS in 20 mM glucose conditions. Further, 100 µM stromatoxin-1 significantly increased IS even under 2 mM glucose conditions. Thus, in our experimental system using MIN6 cells and β-cells/islets, stromatoxin-1 increased IS as well as propofol.