Volume 177, Issue 13 p. 2932-2946
RESEARCH PAPER
Free Access

Cannabidiol protects against high glucose-induced oxidative stress and cytotoxicity in cardiac voltage-gated sodium channels

Mohamed A. Fouda

Mohamed A. Fouda

Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, Canada

Department of Pharmacology and Toxicology, Alexandria University, Alexandria, Egypt

Search for more papers by this author
Mohammad-Reza Ghovanloo

Mohammad-Reza Ghovanloo

Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, Canada

Search for more papers by this author
Peter C. Ruben

Corresponding Author

Peter C. Ruben

Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, Canada

Correspondence

Peter C. Ruben, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, 8888 University Drive, Burnaby, BC V5A 1S6, Canada.

Email: [email protected]

Search for more papers by this author
First published: 19 February 2020
Citations: 27

Abstract

Background and Purpose

Cardiovascular complications are the major cause of mortality in diabetic patients. However, the molecular mechanisms underlying diabetes-associated arrhythmias are unclear. We hypothesized that high glucose could adversely affect Nav1.5, the major cardiac sodium channel isoform of the heart, at least partially via oxidative stress. We further hypothesized that cannabidiol (CBD), one of the main constituents of Cannabis sativa, through its effects on Nav1.5, could protect against high glucose-elicited oxidative stress and cytotoxicity.

Experimental Approach

To test these ideas, we used CHO cells transiently co-transfected with cDNA encoding human Nav1.5 α-subunit under control and high glucose conditions (50 or 100 mM for 24 hr). Several experimental and computational techniques were used, including voltage clamp of heterologous expression systems, cell viability assays, fluorescence assays and action potential modelling.

Key Results

High glucose evoked cell death associated with elevation in reactive oxygen species (ROS) right shifted the voltage dependence of conductance and steady-state fast inactivation, and increased persistent current leading to computational prolongation of action potential (hyperexcitability) which could result in long QT3 arrhythmia. CBD mitigated all the deleterious effects provoked by high glucose. Perfusion with lidocaine (a well-known sodium channel inhibitor with antioxidant effects) or co-incubation of Tempol (a well-known antioxidant) elicited protection, comparable to CBD, against the deleterious effects of high glucose.

Conclusion and Implications

These findings suggest that, through its favourable antioxidant and sodium channel inhibitory effects, CBD may protect against high glucose-induced arrhythmia and cytotoxicity.

Abbreviations

  • CBD
  • cannabidiol
  • CHO
  • Chinese hamster ovary
  • Nav
  • voltage-gated sodium channel
  • ROS
  • reactive oxygen species
  • SSFI
  • steady-state fast inactivation
  • AP
  • Action potential
  • APD
  • Action potential duration
  • What is already known

    • Cardiac complications, including arrhythmias, are a common cause of morbidity and mortality in diabetes-related hyperglycaemia.
    • Long QT syndrome may be caused by gating defects in the voltage-gated sodium channel, Nav1.5.

    What this study adds

    • High glucose alters gating in Nav1.5, impart via ROS, consistent with that underlie long QT.
    • Antioxidants, including cannabidiol, lidocaine and Tempol, inhibited the Nav1.5 gating defects induced by high glucose.

    What is the clinical significance

    • Cardiac sodium channels are a potential therapeutic target to prevent arrhythmias associated with diabetes.
    • Cannabidiol may have therapeutic potential to prevent cardiac complications associated with diabetes.

    1 INTRODUCTION

    Cardiovascular complications are the main cause of mortality and morbidity in diabetic populations (Matheus et al., 2013). Hyperglycaemia/high glucose levels are considered to be the cornerstone in the development of diabetes-evoked cardiovascular complications (Pistrosch, Natali, & Hanefeld, 2011). The main mechanisms underlying these deleterious effects include oxidative stress, activation of pro-inflammatory and inactivation of pro-survival pathways such as Akt, which eventually culminate in cell death (Rajesh et al., 2010). Moreover, there is a strong correlation between diabetes and long QT syndrome (LQT; Grisanti, 2018).

    Long QT syndrome is a cardiac arrhythmogenic disorder, identified by a prolongation of the QT interval (Napolitano, Bloise, & Priori, 2006). Long QT syndrome increases the risk of sudden death due to ventricular fibrillation (Jones & Ruben, 2008). QT prolongation is most commonly caused by a loss of function in potassium channels, as in long QT syndrome 1 and 2 (Shimizu & Antzelevitch, 1999). In contrast, long QT syndrome 3 is caused by a gain of function in cardiac sodium channels that increases the depolarizing current throughout the action potential plateau (Shimizu & Antzelevitch, 1999). Recently, it was shown that cardiac sodium channels are associated with the pathogenesis of long QT syndrome in diabetic rats (Yu et al., 2018).

    The sodium current passing through Nav channels initiates action potentials in neurons, skeletal muscles and cardiac muscles. Nav channels are hetero-multimeric proteins composed of large ion conducting α-subunits and smaller auxiliary β-subunits (Catterall, 2012). In the heart, Nav1.5 is responsible for the rapid upstroke of the cardiac action potential and for rapid impulse conduction through cardiac tissue (Balser, 1999). Alterations in the biophysical properties of Nav1.5 play an important role in the cardiac arrhythmogenesis (Ruan, Liu, & Priori, 2009). However, the diabetes/high glucose-induced changes in the biophysical properties of Nav1.5 are not well understood.

    Cannabidiol (CBD) is the main non-psychotropic constituent of the Cannabis sativa plant. CBD recently has been approved as an anti-seizure therapeutic (Barnes, 2006; Devinsky et al., 2017). CBD has little to no affinity for endocannabinoid receptors (Thomas, Gilliam, Burch, Roche, & Seltzman, 1998). Recently, we showed that CBD imparts inhibitory effects on Nav1.1–Nav1.7 in vitro (Ghovanloo et al., 2018). CBD is well tolerated and lacks adverse cardiac toxicity (Cunha et al., 1980; Izzo, Borrelli, Capasso, Di Marzo, & Mechoulam, 2009). Interestingly, CBD ameliorated diabetes/high glucose-induced deleterious cardiomyopathy (Rajesh et al., 2010).

    Here, we sought to understand whether the oxidative and cytotoxic effects of high glucose could be, in part, mediated through alterations in the biophysical properties of Nav1.5 and to understand whether CBD could ameliorate the deleterious effects of high glucose via its antioxidant and/or inhibitory effects on Nav1.5 (Ghovanloo et al., 2018; Rajesh et al., 2010). Our results suggest that high glucose modulates the gating properties of Nav1.5. The overall biophysical effect is an induced hyperexcitability. This hyperexcitability can be alleviated by CBD. Lidocaine or Tempol also elicited protection, comparable to CBD, against the deleterious effects of high glucose. We conclude that Nav1.5 could be a molecular therapeutic target for alleviating the deleterious consequences of diabetes/high glucose.

    2 METHODS

    2.1 Cell culture

    CHO (RRID:CVCL_0214) were grown at pH 7.4 in filtered sterile F12 (Ham's) nutrient medium (Life Technologies, Thermo Fisher Scientific, Waltham, MA, USA), supplemented with 5% FBS, and maintained in a humidified environment at 37°C with 5% CO2. Cells were transiently co-transfected with the human cDNA encoding the Nav1.5 α-subunit, the β1-subunit and eGFP. Transfection was done according to the PolyFect (Qiagen, Germantown, MD, USA) transfection protocol. A minimum of 8-hr incubation was allowed after each set of transfections. Then the cells were dissociated with 0.25% trypsin–EDTA (Life Technologies, Thermo Fisher Scientific) and plated on sterile coverslips under normal (10 mM) or elevated glucose concentrations (25–150 mM) for 24 hr prior to electrophysiological or biochemical experiments. Hyperglycaemia is the most important factor in the onset and progress of diabetic complications (Viskupicova et al., 2015). High glucose concentrations are usually used as a model to mimic the in vivo situation of hyperglycaemia in diabetes (Viskupicova et al., 2015). High glucose concentrations (up to 100 mM of d-glucose) have also been previously used to mimic human hyperglycaemia in cell line studies (Viskupicova et al., 2015). 100 mM of glucose was used in many studies and in different cell lines, including human erythrocytes (incubation for 72 hr; Viskupicova et al., 2015), human SH-SY5Y neuroblastoma cell line (Liu et al., 2019) and neuronal PC 12 cells (Chen et al., 2016; Fouda & Abdel-Rahman, 2017). We used the MTS cell viability assay to check the viability of CHO cells at different glucose concentrations. This was done to optimize the glucose concentration that would mimic the diabetic/hyperglycaemia conditions in CHO cells.

    To ensure that there are no confounding effects imposed by high osmolarity, we also performed experiments in the presence of mannitol (100 mM for 24 hr) as osmotic control for high glucose, in accordance with reported studies (El-Remessy, Abou-Mohamed, Caldwell, & Caldwell, 2003; Fouda & Abdel-Rahman, 2017; Sharifi, Eslami, Larijani, & Davoodi, 2009).

    2.2 Determination of cell viability

    To establish the concentration-dependent cytotoxicity caused by glucose in our model system, CHO cells were seeded at 50,000 cells·ml−1 in a 96-well plate for 24 hr (90% confluence) and then treatments were started in normal (10 mM) or elevated (25–150 mM) glucose concentrations for another 24 hr in the presence and absence of different treatments (CBD [1 or 5 μM], lidocaine [100 μM or 1 mM], Tempol [100 μM or 1 mM] or their vehicle). At the end of the incubation period (24 hr), cell viability was measured by MTS cell proliferation assay kit with absorbance measured at 495 nm in accordance with the manufacturer's instructions (Abcam, ab197010, Toronto, Canada).

    2.3 ROS measurement

    Oxidative stress level was measured using 2′,7′-dichlorofluorescein diacetate, a detector of ROS (Korystov, Emel'yanov, Korystova, Levitman, & Shaposhnikova, 2009). Fluorescence intensity was measured 30 min after the reaction initiation using a microplate fluorescence reader set at excitation (485 nm)/emission (530 nm) according to the manufacturer's instructions (Abcam, ab113851). The ROS level was determined as relative fluorescence units of generated DCF using standard curve of DCF (Fouda & Abdel-Rahman, 2017; Fouda, El-Sayed, & Abdel-Rahman, 2018).

    2.4 Electrophysiology

    Whole-cell patch-clamp recordings were implemented using extracellular solution composed of NaCl (140 mM), KCl (4 mM), CaCl2 (2 mM), MgCl2 (1 mM) and HEPES (10 mM). Extracellular solution was titrated to pH 7.4 with CsOH. Pipettes were fabricated with a P-1000 puller using borosilicate glass (Sutter Instruments, CA, USA), dipped in dental wax to reduce capacitance and then thermally polished to a resistance of 1.0–1.5 MΩ. Pipettes were filled with intracellular solution, containing CsF (120 mM), CsCl (20 mM), NaCl (10 mM) and HEPES (10 mM) titrated to pH 7.4. All recordings were made using an EPC-9 patch-clamp amplifier (HEKA Elektronik, Lambrecht, Germany) digitized at 20 kHz via an ITC-16 interface (Instrutech, Great Neck, NY, USA). Voltage clamping and data acquisition were controlled using PatchMaster/FitMaster software (HEKA Elektronik) running on an Apple iMac. Current was low pass filtered at 5 kHz. Leak subtraction was automatically done using a P/4 procedure following the test pulse. Gigaohm seals were allowed to stabilize in the on-cell configuration for 1 min prior to establishing the whole-cell configuration. Series resistance was less than 5 MΩ for all recordings. Series resistance compensation up to 80% was used when necessary. All data were acquired at least 5 min after attaining the whole-cell configuration, and cells were allowed to incubate 5 min after drug application prior to data collection. Before each protocol, the membrane potential was hyperpolarized to −130 mV to insure complete removal of both fast inactivation and slow inactivation. Leakage and capacitive currents were subtracted with a P/4 protocol. All experiments were conducted at 22°C.

    2.5 Activation protocols

    To determine the voltage dependence of activation, we measured the peak current amplitude at test pulse voltages ranging from −130 to +80 mV in increments of 10 mV for 19 ms. Channel conductance (G) was calculated from peak INa:
    urn:x-wiley:00071188:media:bph15020:bph15020-math-0001(1)
    where GNa is conductance, INa is peak sodium current in response to the command potential V, and ENa is the Nernst equilibrium potential. The midpoint and apparent valence of activation were derived by plotting normalized conductance as a function of test potential. Data were then fitted with a Boltzmann function:
    urn:x-wiley:00071188:media:bph15020:bph15020-math-0002(2)
    where G/Gmax is normalized conductance amplitude, Vm is the command potential, z is the apparent valence, e0 is the elementary charge, V1/2 is the midpoint voltage, k is the Boltzmann constant, and T is temperature in K.

    2.6 Steady-state fast inactivation protocols

    The voltage dependence of fast inactivation was measured by preconditioning the channels to a hyperpolarizing potential of −130 mV and then eliciting prepulse potentials that ranged from −170 to +10 mV in increments of 10 mV for 500 ms, followed by a 10-ms test pulse during which the voltage was stepped to 0 mV. Normalized current amplitude as a function of voltage was fit using the Boltzmann function:
    urn:x-wiley:00071188:media:bph15020:bph15020-math-0003(3)
    where Imax is the maximum test pulse current amplitude, z is apparent valency, e0 is the elementary charge, Vm is the prepulse potential, V1/2 is the midpoint voltage of steady-state fast inactivation (SSFI), k is the Boltzmann constant, and T is temperature in K.

    2.7 Fast inactivation recovery

    Channels were fast inactivated during a 500-ms depolarizing step to 0 mV. Recovery was measured during a 19-ms test pulse to 0 mV following −130 mV of recovery pulse for durations between 0 and 1.024 s. Time constants of fast inactivation were derived using a double exponential equation:
    urn:x-wiley:00071188:media:bph15020:bph15020-math-0004(4)
    where I is current amplitude, Iss is the plateau amplitude, α1 and α2 are the amplitudes at time 0 for time constants τ1 and τ2, and t is time.

    2.8 Persistent current protocols

    Late sodium current was measured between 145 and 150 ms during a 200-ms depolarizing pulse to 0 mV from a holding potential of −130 mV. Fifty pulses were averaged to increase signal-to-noise ratio.

    2.9 Action potential modelling

    Action potentials were simulated using a modified version of the O'Hara–Rudy model programmed in Matlab (O'Hara, Virag, Varro, & Rudy, 2011, PLoS Comput. Bio).

    The code that was used to produce model is available online from the Rudy Lab website (http://rudylab.wustl.edu/research/cell/code/Allcodes.html). The modified gating INa parameters were in accordance with the biophysical data obtained from whole-cell patch-clamp experiments in this study for various conditions. The model accounted for activation voltage dependence, SSFI voltage dependence, persistent sodium currents, and peak sodium currents (compound conditions).

    2.10 Drug preparations

    CBD was purchased from Toronto Research Chemicals in powder form. Other compounds (e.g. lidocaine, Tempol, d-glucose or mannitol) were purchased from Sigma-Aldrich (ON, Canada). Powdered CBD, lidocaine, or Tempol was dissolved in 100% DMSO to create stock. The stock was used to prepare drug solutions in extracellular solutions at various concentrations with no more than 0.5% total DMSO content.

    2.11 Data analysis and statistics

    The data and statistical analysis comply with the recommendations of the British Journal of Pharmacology on experimental design and analysis in pharmacology (Curtis et al., 2018). Studies were designed to generate groups of almost equal size (n = 5–9), using randomization and blinded analysis. Normalization was performed in order to control the variations in sodium channel expression and inward current amplitude and in order to be able to fit the recorded data with Boltzmann function (for voltage dependences) or an exponential function (for time courses of inactivation). Fitting and graphing were done using FitMaster software (HEKA Elektronik) and Igor Pro (Wavemetrics, Lake Oswego, OR, USA). Statistical analysis consisted of one-way ANOVA (endpoint data) along with post hoc testing (where F in ANOVA achieved P < 0.05 and there was no significant variance inhomogeneity) of significant findings along with Student's unpaired t-test and Tukey's test using Prism 7 software (GraphPad Software Inc., San Diego, CA, USA). Values are presented as mean ± SEM with probability levels less than 0.05 considered significant. Statistical analysis was undertaken only for studies where each group size was at least “n = 5.” The declared group size is the number of independent values and that statistical analysis was done using these independent values. In the electrophysiological experiments, we randomized the different treatments under the different conditions (e.g. control vs. high glucose concentrations), so that at least n = 5 of cells in each treatment or condition came from, at least, five different randomized cell passages.

    2.12 Nomenclature of targets and ligands

    Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS guide to pharmacology (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2019/20 (Alexander et al., 2019).

    3 RESULTS

    3.1 CBD, Tempol, or lidocaine attenuates high glucose-induced cytotoxicity

    The normal glucose concentration in the F12 (Ham's) nutrient medium for CHO cells is 10 mM. We performed a cell viability assay in which transfected and untransfected cells were incubated in elevated high glucose (range 25–150 mM). This experiment was used to select a glucose concentration that ensures a sufficiently large window to detect readout signals throughout the study. Our results indicate that exposures to higher than normal glucose (i.e. >10 mM) for 24 hr caused a concentration-dependent reduction in cell viability (Figure 1a). Moreover, the cells transfected with Nav1.5 exhibited a greater reduction in viability compared to untransfected cells (P < 0.05) at glucose concentrations of 50, 100 or 150 mM (Figure 1a). These findings suggest that high glucose levels reduce cell viability, with this effect being more pronounced upon Nav1.5 transfection.

    Details are in the caption following the image
    CHO cells viability measured by MTS assay and 2′,7′-dichlorofluorescein biochemical assay of the generation of ROS after 24-hr incubation. (a) Effect of gradual increasing of glucose concentration (10, 25, 50, 100, and 150 mM) on the cell viability of untransfected or Nav1.5-transfected cells. (b) Effect of co-incubation of CBD (5 μM), lidocaine (1 mM) or Tempol (1 mM) or their vehicle on the cell viability of Nav1.5-transfected cells incubated in control (10 mM) or high glucose concentrations (50 or 100 mM). (c) Effect of gradual increasing of glucose concentration (10, 25, 50, 100 and 150 mM) on ROS production of untransfected or Nav1.5-transfected cells. (d) Effect of co-incubation of CBD (5 μM), lidocaine (1 mM),or Tempol (1 mM) or their vehicle on ROS production of Nav1.5-transfected cells incubated in normal (10 mM) or high glucose concentrations (50 or 100 mM). Values are expressed as means ± SEM of five independent experiments. In each experiment, each sample was repeated five times. *P < 0.05 versus corresponding “control” values; ^P < 0.05 versus corresponding “untransfected counterpart.” #P < 0.05 versus corresponding “glucose 50- or 100-mM counterparts”

    To ensure that the reduction in cell viability is indeed due to the presence of Nav1.5 and not a by-product of the stress induced on cells by the transient transfection process, we compared the viability of cells stably transfected with Nav1.5 to blank cells that underwent the transient transfection procedure without adding the cDNA for Nav1.5 (mock transfected). Our results show that the stably transfected Nav1.5 cells exhibited a greater reduction in cell viability compared to mock-transfected cells (P < 0.05) at glucose concentrations of 50, 100 or 150 mM (Figure S1A). This further supports our previous results indicating the role of Nav1.5 in reducing cell viability in high glucose. In addition, mannitol (100 mM) had no significant effect on the cell viability of the stably transfected Nav1.5 cells compared to the untransfected or the mock-transfected cells (Figure S1A).

    To determine if we could pharmacologically attenuate the reduction in cell viability at high glucose, we co-incubated cells at different glucose concentrations with CBD, lidocaine or Tempol. Co-incubation with CBD (5 μM) for 24 hr attenuated the reduction in cell viability at high glucose conditions (50 or 100 mM). However, lidocaine (1 mM) only partially reduced the glucose-elicited cytotoxicity (Figure 1b). Co-incubation with the antioxidant Tempol (1 mM) showed similar results to CBD (Figure 1b). In addition, CBD (1 or 5 μM), lidocaine (100 μM or 1 mM) or Tempol (100 μM or 1 mM) exerted a concentration-dependent cytoprotection in transiently transfected cells with Nav1.5 against high glucose (100 mM; Figure S1C).

    3.2 CBD, Tempol, or lidocaine abolishes ROS formation

    One of the key manifestations of high glucose levels is ROS formation. To determine whether the cell viability data from the previous experiments correlate with increased ROS formation, we measured ROS levels using DCF fluorescence after incubation for 24 hr in elevated glucose concentrations (25–150 mM). DCF fluorescence intensity showed a glucose concentration-dependent increase in the ROS level with no significant difference between the untransfected and Nav1.5-transfected cells (Figure 1c). Similarly, high glucose (25–150 mM) evoked a concentration-dependent increase in ROS production with no significant difference between the stably transfected Nav1.5 cells and the mock-transfected cells (Figure S1B). This suggests that ROS formation is unrelated to the presence of Nav1.5 and may lie upstream of this channel in the cytotoxicity pathway. Next, we found that although co-incubation of CBD (1 or 5 μM) or Tempol (100 μM or 1 mM) for 24 hr in transfected cells concentration dependently abolished the high (50 or 100 mM) glucose-evoked increase in ROS levels (Figures 1d and S1D), our results suggest that higher concentrations of CBD or Tempol exert complete reduction of ROS levels, whereas lidocaine (100 μM or 1 mM, in a concentration-dependent manner) exerts only partial ROS reduction (Figures 1d and S1D). These findings are consistent with ROS formation being upstream of Nav1.5. Moreover, co-incubation with CBD (5 μM) for 24 hr abolished the high (100 mM) glucose-evoked oxidative and cytotoxic effects in untransfected cells, suggesting that the CBD protective effect against high glucose cytotoxicity is Nav1.5 independent (Figure S1E,F).

    3.3 High glucose causes a loss of function in Nav1.5 activation

    Our results from the previous experiments with sodium channel blockers prompted us to test the effects of high glucose on the biophysical properties of Nav1.5. To do this, we performed whole-cell voltage-clamp experiments. We examined the effects of glucose at four concentrations (10, 25, 50 and 100 mM) on Nav1.5 activation by measuring peak channel conductance between −130 and +80 mV. Figure 2 shows the Nav1.5 conductance plotted as a function of membrane potential. High glucose (50 or 100 mM) shifted the Nav1.5 midpoint (V1/2) of activation in the positive direction in a concentration-dependent manner with no significant effect elicited by 25-mM glucose or mannitol (100 mM, osmotic control; Figure S2A and Table 1). This suggests that high glucose concentrations make Nav1.5 less likely to activate at any given membrane potential. Additionally, the slope (apparent valence, z) of the activation curves showed a significant decrease during elevated glucose levels (Figure 2a and Table 1). This decrease in slope suggests a reduction in activation charge sensitivity.

    Details are in the caption following the image
    Normalized conductance plotted against membrane potential with sample currents. (a) Effect of high glucose (50 [number of cells, n = 6, number of cell passages/transfections, N = 6] or 100 mM [n = 9, N = 9]) on the conductance curve of Nav1.5-transfected cells. (b) Effect of CBD (5 μM, n = 6, N = 6), lidocaine (1 mM, n = 5, N = 5) or Tempol (1 mM, perfusion [n = 5, N = 5] or incubation [n = 7, N = 7]) or their vehicle (n = 6, N = 6) on the conductance curve of Nav1.5-transfected cells incubated in control (10 mM) glucose concentration. (c) Effect of CBD (5 μM, n = 6, N = 6), lidocaine (1 mM, n = 8, N = 8) or Tempol (1 mM, perfusion [n = 6, N = 6] or incubation [n = 5, N = 5]) or their vehicle (n = 6, N = 6) on the conductance curve of Nav1.5-transfected cells incubated in 50-mM glucose for 24 hr. (d) Effect of CBD (5 μM, n = 5, N = 5), lidocaine (1 mM, n = 5, N = 5) or Tempol (1 mM, perfusion [n = 5, N = 5] or incubation [n = 5, N = 5]) or their vehicle (n = 9, N = 9) on the conductance curve of Nav1.5-transfected cells incubated in 100-mM glucose for 24 hr
    TABLE 1. Parameters of Steady-state activation
    Midpoints and slopes of steady state activation GVV1/2 (mV) GVz (slope) n
    Control
    Control/vehicle −38.2 ± 0.9 3.7 ± 0.1 6
    Control/CBD (5 μM) −31.1 ± 3.1 3.1 ± 0.2 6
    Control/lidocaine (1 mM) −39.9 ± 5.5 3.6 ± 0.2 5
    Control/Tempol (1 mM) perfusion −36.1 ± 2.2 3.9 ± 0.6 5
    Control/Tempol (1 mM) incubation −41.9 ± 3.2 4.4 ± 0.5 7
    Mannitol (100 mM) −39.2 ± 1.6 5.1 ± 0.8 7
    Glucose (25 mM) −37.4 ± 1.2 3.1 ± 0.2 5
    Glucose (50 mM)
    Glucose (50 mM)/vehicle −30.7 ± 2.2 2.9 ± 0.3 6
    Glucose (50 mM)/CBD (5 μM) −39.1 ± 2.8 4.7 ± 0.9 6
    Glucose (50 mM)/lidocaine (1 mM) −39.4 ± 4.0 5.5 ± 0.9 8
    Glucose (50 mM)/Tempol (1 mM) perfusion −27.6 ± 3.9 2.6 ± 0.2 6
    Glucose (50 mM)/Tempol (1 mM) incubation −43.1 ± 0.5 4.0 ± 0.6 5
    Glucose (100 mM)
    Glucose (100 mM)/vehicle −18.9 ± 2.9 2.8 ± 0.2 9
    Glucose (100 mM)/CBD (5 μM) −39.9 ± 1.0 3.7 ± 0.6 5
    Glucose (100 mM)/CBD (1 μM) −26.9 ± 1.0 2.8 ± 0.2 5
    Glucose (100 mM)/lidocaine (1 mM) −36.3 ± 3.6 4.2 ± 0.8 5
    Glucose (100 mM)/lidocaine (100 μM) −24.7 ± 1.7 3.1 ± 0.1 5
    Glucose (100 mM)/Tempol (1 mM) perfusion −23.1 ± 2.2 2.2 ± 0.1 5
    Glucose (100 mM)/Tempol (1 mM) incubation −41.1 ± 2.4 4.3 ± 1.0 5
    Glucose (100 mM)/Tempol (100 μM) incubation −30.2 ± 1.4 3.4 ± 0.3 5

    3.4 Real-time perfusion of CBD or lidocaine or co-incubation of Tempol restores Nav1.5 activation in high glucose

    To determine whether the change in Nav1.5 activation due to glucose incubation could be restored, we measured channel conductance in the presence of CBD, lidocaine or Tempol. We found that none of CBD (perfusion), lidocaine (perfusion) or Tempol (perfusion or incubation) exerted any significant effect on voltage dependence of activation of Nav1.5 under the control condition (10-mM glucose; Figure 2b and Table 1). However, perfusion of CBD (1 or 5 μM) and lidocaine (100 μM or 1 mM) or co-incubation of Tempol (100 μM or 1 mM; for 24 hr) abolished the high (50 or 100 mM) glucose-elicited shifts of V1/2 and the apparent valence of activation in a concentration-dependent manner (Figures 2c,d and S2B and Table 1). We found that Tempol perfusion had no effects on high glucose-evoked alterations in Nav1.5 activation (Figure 2c,d). This is consistent with Tempol's properties as an antioxidant agent, which requires longer exposures to alleviate high glucose-mediated toxicity. In contrast, CBD and lidocaine may work on the level of Nav1.5 in the membrane and hence, may not need the long exposures required by Tempol. Representative families of macroscopic currents across conditions are shown (Figure 2e). Perfusion of CBD reduced the current density of Nav1.5 with no significant difference between the control condition (from −2.05 ± 0.61 to −0.87 ± 0.23 nA·pF−1) and the high glucose (50 or 100 mM; from −2.40 ± 0.85 to −1.19 ± 0.46 nA·pF−1 or from −2.86 ± 0.76 to −0.95 ± 0.29 nA·pF−1, respectively).

    3.5 High glucose causes a gain of function in Nav1.5 steady-state fast inactivation (SSFI)

    Within a few milliseconds of activation, the DIII–IV linker in Nav1.5 mediates fast inactivation. We show normalized current amplitudes plotted as a function of prepulse potential (Figure 3). High glucose (50 or 100 mM) caused significant positive shifts in the V1/2 obtained from Boltzmann function fits at high glucose (Figure 3 and Table 2). These shifts suggest a gain of function in the voltage dependence of Nav1.5 SSFI and suggest that, at any given membrane potential, Nav1.5 is less likely to inactivate at higher glucose concentrations. In contrast, high glucose (25 mM) or the mannitol (100 mM, osmotic control for high glucose) had no effects on the voltage dependence of Nav1.5 SSFI (Figure S3A and Table 2).

    Details are in the caption following the image
    Voltage dependence of steady-state fast inactivation (SSFI) as normalized current plotted against membrane potential with the insert showing the protocol. (a) Effect of high glucose (50 [n = 6, N = 6] or 100 mM [n = 9, N = 9]) on SSFI. (b) Effect of CBD (5 μM, n = 5, N = 5), lidocaine (1 mM, n = 5, N = 5) or Tempol (1 mM, perfusion [n = 6, N = 6] or incubation [n = 5, N = 5]) or their vehicle (n = 5, N = 5) on SSFI of Nav1.5-transfected cells incubated in control (10 mM) glucose concentration. (c) Effect of CBD (5 μM, n = 5, N = 5), lidocaine (1 mM, n = 7, N = 7), or Tempol (1 mM, perfusion [n = 6, N = 6] or incubation [n = 5, N = 5]) or their vehicle (n = 6, N = 6) on SSFI curve of Nav1.5-transfected cells incubated in 50-mM glucose for 24 hr. (d) Effect of CBD (5 μM, n = 5, N = 5), lidocaine (1 mM, n = 8, N = 8) or Tempol (1 mM, perfusion [n = 5, N = 5] or incubation [n = 5, N = 5]) or their vehicle (n = 9, N = 9) on SSFI curve of Nav1.5-transfected cells incubated in 100-mM glucose for 24 hr
    TABLE 2. Parameters of Steady-state fast inactivation (SSFI)
    Midpoints and slopes of steady state activation SSFI − V1/2 (mV) SSFI − z (slope) n
    Control
    Control/vehicle −94.3 ± 2.5 −2.2 ± 0.2 5
    Control/CBD (5 μM) −113.3 ± 6.2 −1.7 ± 0.1 5
    Control/lidocaine (1 mM) −111.7 ± 5.6 −1.5 ± 0.1 5
    Control/Tempol (1 mM) perfusion −94.1 ± 3.0 −2.2 ± 0.2 6
    Control/Tempol (1 mM) incubation −98.6 ± 2.8 −2.2 ± 0.1 5
    Mannitol (100 mM) −93.6 ± 2.3 −2.1 ± 0.3 5
    Glucose (25 mM) −94.8 ± 3.5 −2.2 ± 0.2 5
    Glucose (50 mM)
    Glucose (50 mM)/vehicle −80.7 ± 3.9 −2.6 ± 0.2 6
    Glucose (50 mM)/CBD (5 μM) −89.6 ± 2.5 −2.5 ± 0.3 5
    Glucose (50 mM)/lidocaine (1 mM) −96.7 ± 5.4 −1.5 ± 0.5 7
    Glucose (50 mM)/Tempol (1 mM) perfusion −76.7 ± 2.4 −2.7 ± 0.1 6
    Glucose (50 mM)/Tempol (1 mM) incubation −94.3 ± 2.7 −2.9 ± 0.2 5
    Glucose (100 mM)
    Glucose (100 mM)/vehicle −66.8 ± 2.6 −2.7 ± 0.2 9
    Glucose (100 mM)/CBD (5 μM) −89.5 ± 2.7 −3.1 ± 0.3 5
    Glucose (100 mM)/CBD (1 μM) −75.0 ± 2.4 −2.4 ± 0.1 5
    Glucose (100 mM)/lidocaine (1 mM) −91.7 ± 2.4 −1.6 ± 0.1 8
    Glucose (100 mM)/lidocaine (100 μM) −73.9 ± 2.0 −2.7 ± 0.3 5
    Glucose (100 mM)/Tempol (1 mM) perfusion −71.8 ± 3.8 −3.0 ± 0.2 5
    Glucose (100 mM)/Tempol (1 mM) incubation −90.5 ± 2.8 −2.8 ± 0.2 5
    Glucose (100 mM)/Tempol (100 μM) incubation −79.9 ± 2.1 −2.3 ± 0.2 5

    3.6 CBD, lidocaine, or Tempol restores the Nav1.5 steady-state fast inactivation (SSFI)

    To determine whether the destabilized SSFI in Nav1.5 could be restored, we measured inactivation in the presence of CBD, lidocaine or Tempol. We first measured compound effects in the control condition. We found that both CBD and lidocaine shifted the inactivation curves to the left. This result is consistent with previous studies on CBD and lidocaine (Ghovanloo et al., 2018; Wang, Mi, Lu, Lu, & Wang, 2015). However, neither perfusion nor incubation with Tempol caused a significant left shift of SSFI of Nav1.5 under the control condition. Next, we performed the same experiments after incubation in 50- or 100-mM glucose. Similar to control, both CBD (1 or 5 μM) and lidocaine (100 μM or 1 mM) shifted the inactivation curves to the left in a concentration-dependent effect (Figures 3c,d and S3B and Table 2). Interestingly, although Tempol perfusion did not change the high glucose-induced effects on SSFI, Tempol (100 μM or 1 mM) incubation concentration dependently shifted the curve to the left (Figures 3c,d and S3B and Table 2). This further suggests that Tempol functions upstream of Nav1.5, in that it only changes the Nav1.5 gating when the channel gating has already been altered due to oxidative stress.

    3.7 Glucose, CBD, lidocaine, or Tempol slows recovery from fast inactivation

    One of the key biophysical features of sodium channels is the kinetics at which they recover from inactivated states. To measure fast inactivation recovery, we held channels at −130 mV to ensure channels were fully at rest, then pulsed the channels to 0 mV for 500 ms and allowed different time intervals at −130 mV to measure recovery as a function of time. We found that incubation in high glucose significantly (though with a relatively small magnitude of difference) increase the slow component of fast inactivation recovery when compared to control (Figures 4a and S4A and Table 3). In addition, CBD, lidocaine or co-incubation with Tempol significantly increased, in a concentration-dependent effect, the time constant of the slow component of recovery from fast inactivation regardless of the glucose concentration (control or high concentration; Figures 4 and S4B and Table 3). However, only lidocaine, but not CBD or Tempol, increased the time constant of the fast component of recovery from fast inactivation regardless of the glucose concentration (Figure 4 and Table 3). These findings suggest that glucose causes a slight loss of function to the fast inactivation recovery of Nav1.5 and the tested compounds further stabilized the inactivated state of the channel (Ghovanloo et al., 2018; Nuss, Tomaselli, & Marban, 1995; Wang et al., 2015)

    Details are in the caption following the image
    Recovery from fast inactivation showing the normalized current plotted against a range of recovery durations (s) with the insert showing the protocol. (a) Effect of high glucose (50 [n = 7, N = 7] or 100 mM [n = 9, N = 9]) on recovery from fast inactivation of Nav1.5-transfected cells. (b) Effect of CBD (5 μM, n = 6, N = 6), lidocaine (1 mM, n = 5, N = 5) or Tempol (1 mM, perfusion [n = 6, N = 6] or incubation [n = 5, N = 5]) or their vehicle (n = 6, N = 6) on the recovery from fast inactivation of Nav1.5-transfected cells incubated in control (10 mM) glucose concentration. (c) Effect of CBD (5 μM, n = 6, N = 6), lidocaine (1 mM, n = 5, N = 5) or Tempol (1 mM, perfusion [n = 6, N = 6] or incubation [n = 5, N = 5]) or their vehicle (n = 7, N = 7) on the recovery from fast inactivation of Nav1.5-transfected cells incubated in 50-mM glucose for 24 hr. (d) Effect of CBD (5 μM, n = 5, N = 5), lidocaine (1 mM, n = 8, N = 8) or Tempol (1 mM, perfusion [n = 5, N = 5] or incubation [n = 5, N = 5]) or their vehicle (n = 9, N = 9) on the recovery from fast inactivation of Nav1.5-transfected cells incubated in 100-mM glucose for 24 hr
    TABLE 3. Time constants for the recovery from fast inactivation
    Midpoints and slopes of steady state activation τfast (s) τslow (s) n
    Control
    Control/vehicle 0.007 ± 0.001 0.007 ± 0.001 6
    Control/CBD (5 μM) 0.007 ± 0.002 0.104 ± 0.025 6
    Control/lidocaine (1 mM) 0.276 ± 0.030 0.307 ± 0.022 5
    Control/Tempol (1 mM) perfusion 0.006 ± 0.001 0.027 ± 0.005 6
    Control/Tempol (1 mM) incubation 0.004 ± 0.001 0.054 ± 0.012 5
    Mannitol (100 mM) 0.020 ± 0.015 0.058 ± 0.016 5
    Glucose (25 mM) 0.006 ± 0.001 0.091 ± 0.019 6
    Glucose (50 mM)
    Glucose (50 mM)/vehicle 0.015 ± 0.011 0.070 ± 0.006 7
    Glucose (50 mM)/CBD (5 μM) 0.006 ± 0.001 0.081 ± 0.009 6
    Glucose (50 mM)/lidocaine (1 mM) 0.160 ± 0.031 0.232 ± 0.029 5
    Glucose (50 mM)/Tempol (1 mM) perfusion 0.008 ± 0.001 0.104 ± 0.021 6
    Glucose (50 mM)/Tempol (1 mM) incubation 0.004 ± 0.001 0.055 ± 0.009 5
    Glucose (100 mM)
    Glucose (100 mM)/vehicle 0.007 ± 0.001 0.082 ± 0.026 9
    Glucose (100 mM)/CBD (5 μM) 0.005 ± 0.001 0.101 ± 0.006 5
    Glucose (100 mM)/CBD (1 μM) 0.020 ± 0.013 0.104 ± 0.019 6
    Glucose (100 mM)/lidocaine (1 mM) 0.178 ± 0.020 0.336 ± 0.042 8
    Glucose (100 mM)/lidocaine (100 μM) 0.032 ± 0.025 0.212 ± 0.013 5
    Glucose (100 mM)/Tempol (1 mM) perfusion 0.005 ± 0.001 0.065 ± 0.032 5
    Glucose (100 mM)/Tempol (1 mM) incubation 0.005 ± 0.001 0.065 ± 0.020 5
    Glucose (100 mM)/Tempol (100 μM) incubation 0.009 ± 0.003 0.118 ± 0.011 5

    3.8 High glucose exacerbates persistent sodium currents and is restored by CBD, lidocaine, or Tempol

    An increased persistent sodium current is a manifestation of destabilized fast inactivation. Large persistent sodium currents are associated with a range of pathological conditions, including long QT3 (Ghovanloo, Abdelsayed, & Ruben, 2016; Wang et al., 1995). To determine the effects of glucose on the stability of Nav1.5 inactivation, we held channels at −130 mV, followed by a depolarizing pulse to 0 mV for 200 ms to elicit persistent currents. Figure 5 shows that incubation in high glucose (50 or 100 mM) significantly (50 mM; 100 mM) increased persistent currents compared to control. On the other hand, neither glucose (25 mM) nor mannitol (100 mM) had any effects on persistent currents compared to control (Figure S5). Although perfusion of CBD, lidocaine, or Tempol (perfusion or incubation) had no effect on the small persistent currents in the control condition, each of the three compounds significantly concentration dependently reduced the high (50 or 100 mM) glucose-induced increase in persistent sodium currents (Figures 5 and S5 and Table 4). In contrast, Tempol perfusion had no effect on high (50 or 100 mM) glucose-elicited increased persistent currents (Figure 5 and Table 4). Reduction of the exaggerated persistent currents at high glucose by CBD is consistent with the previous reports in neuronal sodium channels (Ghovanloo et al., 2018; Patel, Barbosa, Brustovetsky, Brustovetsky, & Cummins, 2016).

    Details are in the caption following the image
    Effect of CBD (5 μM), lidocaine (1 mM), or Tempol (1 mM, perfusion or incubation) or their vehicle on the percentage of persistent sodium currents of Nav1.5-transfected cells incubated in control (CBD [n = 6, N = 6], lidocaine [n = 5, N = 5], Tempol perfusion [n = 5, N = 5], Tempol incubation [n = 5, N = 5], or the vehicle [n = 7, N = 7]), 50-mM glucose (CBD [n = 5, N = 5], lidocaine [n = 5, N = 5], Tempol perfusion [n = 8, N = 8], Tempol incubation [n = 5, N = 5], or the vehicle [n = 7, N = 7]), or 100-mM glucose (CBD [n = 7, N = 7], lidocaine [n = 7, N = 7], Tempol perfusion [n = 7, N = 7], Tempol incubation [n = 5, N = 5] or the vehicle [n = 9, N = 9]) for 24 hr with sample currents and the insert showing the protocol. *P < 0.05 versus corresponding “control” values
    TABLE 4. Persistent current
    Midpoints and slopes of steady state activation Percentage of persistent INa n
    Control
    Control/vehicle 0.97 ± 0.11 7
    Control/CBD (5 μM) 0.72 ± 0.13 6
    Control/lidocaine (1 mM) 0.97 ± 0.17 5
    Control/Tempol (1 mM) perfusion 1.06 ± 0.07 5
    Control/Tempol (1 mM) incubation 1.03 ± 0.08 5
    Mannitol (100 mM) 0.97 ± 0.10 7
    Glucose (25 mM) 1.06 ± 0.12 5
    Glucose (50 mM)
    Glucose (50 mM)/vehicle 3.16 ± 0.42 7
    Glucose (50 mM)/CBD (5 μM) 1.26 ± 0.20 5
    Glucose (50 mM)/lidocaine (1 mM) 1.82 ± 0.11 5
    Glucose (50 mM)/Tempol (1 mM) perfusion 3.04 ± 0.35 8
    Glucose (50 mM)/Tempol (1 mM) incubation 0.90 ± 0.22 5
    Glucose (100 mM)
    Glucose (100 mM)/vehicle 6.41 ± 0.61 9
    Glucose (100 mM)/CBD (5 μM) 1.38 ± 0.20 7
    Glucose (100 mM)/CBD (1 μM) 3.03 ± 0.12 6
    Glucose (100 mM)/lidocaine (1 mM) 1.31 ± 0.25 7
    Glucose (100 mM)/lidocaine (100 μM) 3.50 ± 0.36 5
    Glucose (100 mM)/Tempol (1 mM) perfusion 6.07 ± 0.82 7
    Glucose (100 mM)/Tempol (1 mM) incubation 1.33 ± 0.26 5
    Glucose (100 mM)/Tempol (100 μM) incubation 2.93 ± 0.18 5

    3.9 High glucose prolongs action potential duration in O'Hara–Rudy model of cardiac excitability

    We used the O'Hara–Rudy model to simulate cardiac action potentials and the effect of high glucose incubation and the tested compounds (O'Hara et al., 2011). The control results from the patch-clamp experiments were adjusted to the original model parameters and the subsequent magnitude shifts in the simulations of other conditions were performed relative to the original model parameters. Figure 6 shows that incubation in high glucose caused a concentration-dependent prolongation of the action potential duration from ~300 to ~450 ms in 50-mM glucose and to >600 ms in 100-mM glucose (Figure 6a). This increased action potential duration could potentially lead to the prolongation of the QT interval (Nachimuthu, Assar, & Schussler, 2012). The simulation results suggest that CBD, lidocaine, or incubation (but not perfusion) with Tempol restores the high glucose-elicited prolongation of the action potential duration to nearly that of the control condition (Figure 6b). This reduction in the predicted excitability is consistent with the anti-excitatory effects attributed to the compounds we used, in particular CBD and lidocaine (Ghovanloo et al., 2018; Nuss et al., 1995).

    Details are in the caption following the image
    Action potential model simulation. (a) Action potential duration of Nav1.5-transfected cells incubated in control, 50-mM glucose, or 100-mM glucose for 24 hr. (b) Effect of CBD (5 μM), lidocaine (1 mM), or Tempol (1 mM, perfusion or incubation) or their vehicle on the action potential duration of Nav1.5-transfected cells incubated in 100-mM glucose for 24 hr

    4 DISCUSSION

    The present study suggests, for the first time, that CBD confers protection on Nav1.5 against the high glucose-elicited hyperexcitability and cytotoxicity. This conclusion is based on the following main findings:- (a) although high glucose concentration dependently increased ROS production to the same extent in both Nav1.5-transfected and untransfected cells, cytotoxicity was significantly higher in Nav1.5-transfected cells compared to their untransfected counterparts; (b) high glucose elicited concentration-dependent positive shifts in the voltage dependence of activation and inactivation and exacerbated persistent currents. Increased persistent currents prolong the computational action potential duration. (c) CBD, lidocaine or co-incubation with Tempol (but not Tempol perfusion) restored the effects of high glucose on increased ROS production, cell viability, and biophysical properties of Nav1.5. Our findings implicate the role of Nav1.5 in high glucose-induced hyperexcitability and cytotoxicity, via oxidative stress, which could lead to long QT3 arrhythmia (Figure 7). Also, our findings demonstrate possible therapeutic effects of CBD for high glucose-provoked cardiac dysfunction in diabetic patients.

    Details are in the caption following the image
    A schematic of possible cellular events involved in the protective effect of CBD, lidocaine, or Tempol against high glucose-induced oxidative effects and cytotoxicity via affecting cardiac voltage-gated sodium channels (Nav1.5)

    Diabetes is associated with long QT arrhythmia in clinical (Ninkovic et al., 2016; Pickham, Flowers, & Drew, 2014) and animal studies (Yu et al., 2018). Several studies showed that QT prolongation predisposes to malignant ventricular arrhythmias and sudden death in diabetic patients (Ukpabi & Onwubere, 2017). Also, diabetic patients with long QT syndrome have a threefold increased risk of cardiac arrest (Whitsel et al., 2005). Thus, action potential prolongation in a diabetic individual could be of prognostic importance for sudden cardiac death (Ukpabi & Onwubere, 2017). Interestingly, functional alterations in Nav1.5, the predominant cardiac voltage gated sodium channel (VGSC; Roden, Balser, George, & Anderson, 2002), are associated with long QT3 syndrome (Kapplinger et al., 2015). Thus, our current study used Nav1.5, transiently transfected into CHO cells, to mimic this clinical problem for the following reasons:- (a) transfected CHO cells are extensively used to study mammalian ion channels (Gamper, Stockand, & Shapiro, 2005); (b) high glucose causes oxidative stress and cell death in CHO cells (Selvi, Bhuvanasundar, & Angayarkanni, 2017) and (c) the computer action potential simulations in our current study showed that high glucose prolonged the simulated action potential (Figure 6). This prolongation could contribute to long QT3 arrhythmia (Nachimuthu et al., 2012).

    As a foundation to our study, we showed that incubating CHO cells in elevated glucose caused a concentration-dependent oxidative stress and cell death (Figure 1), which agrees with previous reports (Fouda & Abdel-Rahman, 2017; Fouda et al., 2018; Lamers, Almeida, Vicente-Manzanares, Horwitz, & Santos, 2011). High glucose-induced increase in ROS production is correlated to apoptosis and cell death (Fouda & Abdel-Rahman, 2017; Fouda et al., 2018) due to impairment of pro-survival signalling pathways such as Akt (Van Linthout et al., 2008) and activation of pro-inflammatory and cell death pathways such as NF-κB (Mariappan et al., 2010). In addition, our current findings implicate the role of Nav1.5 as a downstream target for oxidative stress in cytotoxicity provoked by high glucose (Figure 1). This finding is consistent with other studies showing that oxidative stress affects the biophysical properties of Nav1.5 through lipoxidation of the cell membrane and/or the inhibition of Nav1.5 trafficking to the cell membrane (Liu et al., 2013; Nakajima et al., 2010). Moreover, changes in Nav1.5 function are correlated with long QT arrhythmia in diabetic rats (Yu et al., 2018). These findings support our hypothesis that high glucose, at least partly through oxidative stress, alters Nav1.5 function and leads to cytotoxicity and arrhythmia.

    Importantly, our current study shows that incubating Nav1.5 in high glucose is associated with depolarizing shifts in the voltage dependence of activation and inactivation, along with a decrease in the slope (apparent valence) of conductance, indicating a reduction in the relative amount of charge that moves during S4 (voltage-sensor) translocation (Jones & Ruben, 2008). Moreover, high glucose resulted in an increase in persistent current. The increase in persistent current could be caused by destabilizing fast inactivation (Jones & Ruben, 2008). The direction of the voltage sensitivity shifts of both activation and inactivation suggests that high glucose may alter gating of all four Nav1.5 VSDs via a common mechanism. This is similar to how some other physiological changes affect certain sodium channels, such as pH (Ghovanloo & Peters, 2018). However, we cannot rule out other mechanisms, for example, glycosylation. Collectively, incubation in high glucose may alter Nav1.5 gating via oxidative stress (Sovari, 2016).

    Together, these biophysical changes in Nav1.5 could lead to action potential prolongation, as suggested by our simulation results (Figure 6). Action potential prolongation is consistent with a recent study showing that Nav1.5 gating defects contribute to the development of arrhythmia in diabetic rats (Yu et al., 2018). Importantly, our results regarding high glucose-induced increase in persistent current and slowed recovery from fast inactivation are consistent with another report about changes in Nav1.5 in diabetic rats (Yu et al., 2018). However, other aspects, such as the shifts in the voltage dependence of activation and inactivation, were inconsistent with previous findings (Yu et al., 2018). This discrepancy could be attributed to the species chosen for the in vivo animal model (rat vs. heterologous expression of human Nav1.5), the use of pro-diabetic drugs (streptozotocin; STZ) and the different pathogenic stages and the severity of the diabetic heart.

    Interestingly, we found that CBD alleviates these biophysical changes along with reducing the oxidative stress and cytotoxicity. We investigated the possible protective effect of CBD against the deleterious high glucose effects because (a) we have recently shown the inhibitory effect of CBD on Nav1.5 and other sodium channel isoforms. CBD stabilized the inactivated states (Ghovanloo et al., 2018), a result that might explain the reduction by CBD of increased persistent current caused by high glucose. This reduction by CBD might also explain the action potential duration restoration we observed in our simulations. (b) CBD was reported to attenuate diabetes-induced oxidative stress, inflammation and cardiac fibrosis in rats (Rajesh et al., 2010). CBD is documented to prevent oxidative damage in neuronal cultures (Hampson, Grimaldi, Axelrod, & Wink, 1998). Also, CBD exerts antioxidant actions on immune cells (Booz, 2011). Recently, CBD has emerged as a promising treatment for many movement disorders such as Parkinson's disease due to, at least partially, its antioxidant effect (Peres et al., 2018). CBD exerted an antioxidant effect and protected against high glucose-induced cytotoxicity (Figure 1). In addition to the biophysical effects on Nav1.5 (Figures 2-6), our results suggest that CBD may also protect against high glucose-elicited cytotoxicity and action potential prolongation.

    To validate our proposed mechanisms for the protective effects of CBD, we used lidocaine (IC50, 1 mM, Nuss et al., 1995) because of its reported inhibitory effect on sodium channels (Nuss et al., 1995) and its antioxidant effects (Saito et al., 2017). Lidocaine reversed the high glucose-induced shifts in the voltage dependence of activation and inactivation and reduced late sodium current, similar to the effects of, but to a lesser extent than, CBD (Figures 1-6). Interestingly, lidocaine caused a hyperpolarizing shift in the voltage dependence of fast inactivation and prolonged the recovery from fast inactivation without affecting the conductance curve under the control conditions, consistent with another study (Wang et al., 2015). Like lidocaine, CBD affected voltage dependence and persistent currents under the control condition (Figures 2-5). Thus, these findings suggest that CBD may restore biophysical properties impaired by incubation in high glucose, at least partially, through its inhibitory effect on Nav1.5 in addition to its more generalized antioxidant property (Figure 7).

    We used the antioxidant Tempol (1 mM, Saito et al., 2017) to further validate the antioxidant effects of CBD and lidocaine. Although Tempol perfusion had no effect on changes induced by incubation in high glucose, co-incubation of Tempol for 24 hr mitigated the high glucose-elicited depolarizing shifts in the voltage dependence of conductance and fast inactivation and the increased persistent currents (Figures 2-6). In our simulations, the biophysical effects of Tempol abolished the action potential prolongation induced by high glucose. Moreover, Tempol co-incubation reduced high glucose-induced oxidative stress and cytotoxicity (Figure 1). These findings suggest that oxidative stress plays an important role in high glucose-induced biophysical changes in Nav1.5 and cytotoxicity. A plausible explanation for the rapid restoration by CBD or lidocaine, but not Tempol, on high glucose-induced changes in Nav1.5 could be attributed to the inhibitory effect of CBD or lidocaine on sodium channels, in addition to their antioxidant effects (Figure 7).

    Our modelling results were limited by their focus on changes in only Nav1.5. Future studies may focus on the effects of other components of electrical activity to provide a complete description of diabetes-induced pathogenesis in electrical signalling. The use of the CHO immortalized cell line exempts this study from considering the implications of our findings with regard to possible male- and female-specific effects of cannabidiol on high glucose-induced dysfunction of Nav1.5.

    In conclusion, this study implicates Nav1.5 in the high glucose-induced cytotoxicity and prolongation of action potential, which could lead to long QT3 arrhythmia, a clinical complication of diabetes. CBD, through its inhibition of Nav1.5 and its antioxidant effect, mitigates the effects of high glucose and the resultant clinical condition.

    ACKNOWLEDGEMENTS

    This work was supported by grants from Natural Science and Engineering Research Council (NSERC) of Canada to M.-R.G. and P.C.R., a MITACS Elevate grant in partnership with Akseera Pharma, Inc. (IT14449) to M.A.F., and a MITACS Accelerate grant in partnership with Xenon Pharmaceuticals to M.-R.G.

      AUTHOR CONTRIBUTIONS

      M.A.F. collected, assembled, analysed, and interpreted the data. M.-R.G. assisted with analysis and interpretation of data. M.A.F. wrote the first draft of the manuscript. M.-R.G. edited the manuscript. P.C.R. conceived the experiments and revised the manuscript critically for important intellectual content.

      CONFLICT OF INTEREST

      None. The authors declare that this research was conducted in the absence of competing interests. We acknowledge that Akseera Pharma Corp, our MITACS partner, is a pharmaceutical company interested in cannabis, but this fact did not affect our findings.

      DECLARATION OF TRANSPARENCY AND SCIENTIFIC RIGOUR

      This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research as stated in the BJP guidelines for Design & Analysis, and as recommended by funding agencies, publishers and other organizations engaged with supporting research.