Skip to main content
NCBI home page
British Journal of Pharmacology logoLink to British Journal of Pharmacology
. 2018 Jun 7;175(14):3007–3020. doi: 10.1111/bph.14348

The opioid oxycodone use‐dependently inhibits the cardiac sodium channel NaV1.5

Jannis E Meents 1, Krisztina Juhasz 2,3, Sonja Stölzle‐Feix 2, Vera Peuckmann‐Post 4, Roman Rolke 4, Angelika Lampert 1,✉
PMCID: PMC6016641  PMID: 29722437

Abstract

Background and Purpose

Oxycodone is a potent semi‐synthetic opioid that is commonly used for the treatment of severe acute and chronic pain. However, treatment with oxycodone can lead to cardiac electrical changes, such as long QT syndrome, potentially inducing sudden cardiac arrest. Here, we investigate whether the cardiac side effects of oxycodone can be explained by modulation of the cardiac Nav1.5 sodium channel.

Experimental Approach

Heterologously expressed human Nav1.5, Nav1.7 (HEK293 cells) or Nav1.8 channels (mouse N1E‐115 cells) were used for whole‐cell patch‐clamp electrophysiology. A variety of voltage‐clamp protocols were used to test the effect of oxycodone on different channel gating modalities. Human stem cell‐derived cardiomyocytes were used to measure the effect of oxycodone on cardiomyocyte beating.

Key Results

Oxycodone inhibited Nav1.5 channels, concentration and use‐dependently, with an IC50 of 483 μM. In addition, oxycodone slows recovery of Nav1.5 channels from fast inactivation and increases slow inactivation. At high concentrations, these effects lead to a reduced beat rate in cardiomyocytes and to arrhythmia. In contrast, no such effects could be observed on Nav1.7 or Nav1.8 channels.

Conclusions and Implications

Oxycodone leads to an accumulation of Nav1.5 channels in inactivated states, with a slow time course. Although the concentrations needed to elicit cardiac arrhythmias in vitro are relatively high, some patients under long‐term treatment with oxycodone as well as drug abusers and addicts might suffer from severe cardiac side effects induced by the slowly developing effects of oxycodone on Nav1.5 channels.

Abbreviations

EFP

extracellular field potential

FPD

(extracellular) field potential duration

hERG

human Ether‐à‐go‐go‐Related Gene

iPSC

induced pluripotent stem cell

max.DV

maximum negative downstroke velocity

Nav

voltage‐gated sodium channel

TdP

torsade de pointes

Introduction

The use of opioids has become common practice for the treatment of severe and chronic pain. http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=7093 is a semi‐synthetic opioid that belongs to the class of http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=319 agonists. It has a 1.5 to 2 times higher efficacy compared with http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=1627 (Biancofiore, 2006; Güttler and Sabatowski, 2008) and induces fewer side effects than other commonly used opioids (Biancofiore, 2006). The high efficacy of oxycodone as an analgesic can be seen in the steady rise of oxycodone subscriptions in the USA and in Europe over the recent years (Paulozzi and Ryan, 2006; Hamunen et al., 2009; Hoffmann et al., 2012). Finally, oxycodone has become a first‐line treatment for cancer pain (Biancofiore, 2006) and the most commonly used controlled‐release opioid in the USA (Davis et al., 2003). Long‐term treatment with oxycodone is not uncommon, which may cause life‐threatening side effects if doses are steadily increased or when drug clearance is affected by disease or age (Saari et al., 2012). Death rates appear to be associated with the prescribed doses of opioids (Paulozzi and Ryan, 2006). Like all opioids, oxycodone has the potential of causing addiction and affected subjects run the same risk of suffering life‐threatening side effects. In fact, oxycodone is one of the opioids most implicated in drug abuse and addiction (Itzoe and Guarnieri, 2017) and contributes significantly to what is now being called the opioid epidemic in the USA. In 2016 alone, 14 487 people died in the USA from overdoses involving natural or semi‐synthetic opioids, such as oxycodone (Rudd, 2016; Hedegaard et al., 2017). Accordingly, the US Centers for Disease Control and Prevention has issued a wake‐up call to address the epidemic in March 2018 (https://www.cdc.gov/media/releases/2018/p0306-vs-opioids-overdoses.html).

In recent years, oxycodone use has been associated with cardiac arrhythmias, such as torsade de pointes (TdP). The drug was shown to prolong the QT interval (Fanoe et al., 2009), which can cause TdPs and lead to sudden cardiac arrest. The authors argued that this QT prolongation is mediated by a block of cardiac hERG potassium channels. However, the oxycodone‐induced block of hERG channels was rather weak with an IC50 of 171 μM, raising the question whether this is the only mechanism by which oxycodone could induce cardiac arrhythmia. Action potentials in cardiac myocytes require the concerted activity of several ion channels, including but not limited to the http://www.guidetopharmacology.org/GRAC/ObjectDisplayForward?objectId=582 (required for depolarization of the cell membrane) and hERG potassium channels (required for repolarization of the cell membrane). It therefore seems likely that the cardiac toxicity of oxycodone could reflect the modulation of many ion channels. Similarly, the synthetic opioid http://www.guidetopharmacology.org/GRAC/LigandDisplayForward?ligandId=5458 was shown to inhibit both hERG and Nav1.5 channels to cause cardiotoxicity (Katchman et al., 2002; Schulze et al., 2014). Inhibition of hERG channels by oxycodone and a following prolonged depolarization of the cell membrane would lead to inactivation of Nav1.5 channels, which would further enhance the cardiotoxic potential of oxycodone.

Here, we investigated if oxycodone directly modulated the cardiac sodium channel Nav1.5, thus further explaining the cardiotoxic potential of the drug. We performed whole‐cell patch‐clamp electrophysiology on human Nav1.5 channels, stably expressed in HEK293 cells and tested for oxycodone‐mediated effects on different gating modalities of the channel. We show that oxycodone indeed induces a concentration‐dependent block of Nav1.5 channels. This block is use‐dependent and is mediated by slowing the recovery of Nav1.5 channels from steady‐state fast inactivation as well as enhancing slow inactivation of the channel. Together with the effect of oxycodone on hERG potassium channels, these mechanisms will lead to an accumulation of Nav1.5 channels in an inactivated state, making the myocardium susceptible to electrophysiological disturbances and promoting cardiac arrhythmia.

Methods

Cell culture and transfection

HEK293 cells (Lampert Lab, Aachen, Germany) stably expressing human Nav1.5 and human Nav1.7 channels were cultured in DMEM/F‐12 (Thermo Fisher Scientific, Waltham, MA, USA), supplemented with 10% FBS and 1% penicillin/streptomycin. Culture medium was further supplemented with 0.1 mg·mL−1 zeocin (Thermo Fisher Scientific) for Nav1.5or 0.5 mg·mL−1 G‐418 (A&E Scientific, Marcq, Belgium) for Nav1.7 channels.

Human Nav1.8 channels, carrying the common A1073 variant (rs6795970), found in approximately 60% of the population (Behr et al., 2015), was transiently transfected into mouse neuroblastoma N1E‐115 cells (kindly provided by Dr Tobias Huth, Erlangen, Germany) using jetPEI transfection reagent (Polyplus‐transfections, Illkirch‐Graffenstaden, France). N1E‐115 cells were cultured in DMEM containing 4.5% glucose (Thermo Fisher Scientific), supplemented with 10% FBS and 1% penicillin/streptomycin. To visualize expression, cells were co‐transfected with pmaxGFP (Lonza, Granta Park, UK). Cells were re‐plated into petri dishes after 5 to 24 h and patched within 24 h of re‐plating.

For the cardiomyocyte assay, Cor.4U® CardioPlate™ CardioExcyte 96 plates (Ncardia, Cologne, Germany) were ordered and delivered non‐frozen. Each well was coated with fibronectin and contained 30 000 viable cells per well. After the arrival of the cells, a medium exchange was performed, and the cells were cultured according to the manufacturer's instructions.

All cells were kept at 37°C and 5% CO2.

Electrophysiology

Whole‐cell patch‐clamp recordings were performed at room temperature using a HEKA EPC 10USB amplifier and PatchMaster software (HEKA electronics, Lambrecht, Germany). Glass pipettes (tip resistance 0.9–2 MΩ) were manufactured with a DMZ puller (Zeitz Instruments GmbH, Martinsried, Germany). Series resistance (<5 MΩ) was compensated by at least 70%. Currents were low‐pass filtered at 10 kHz and sampled at 50 or 100 kHz. Leak current was subtracted using the P/4 method. The liquid junction potential was not corrected. After establishing the whole‐cell configuration, inward Na+ currents were allowed to stabilize during 0.1 Hz stimulation for 3 min before the start of the recording. The holding potential was set to −120 mV. For concentration–response measurements and protocols measuring the recovery from fast inactivation and the onset of slow inactivation on Nav1.5 channels, the holding potential was set to −150 mV due to a time‐dependent shift of fast inactivation (see Results). Cells were continuously superfused with extracellular solution or test solutions through a common outlet using a gravity‐driven perfusion system.

Solutions

For recordings from Nav1.5 channels, the extracellular bath solution contained (in mM): 100 choline‐Cl, 40 NaCl, 3 KCl, 1 MgCl2, 1 CaCl2, 10 HEPES, 5 glucose and 10–15 sucrose (depending on osmolarity) (pH 7.4). Replacing NaCl by choline‐Cl was necessary in order to reduce current amplitude and avoid voltage errors. Intracellular solution contained (in mM): 140 CsF, 10 NaCl, 10 HEPES, 1 EGTA and 18 sucrose (pH 7.3).

For recordings from Nav1.7 channels, the extracellular bath solution contained (in mM): 140 NaCl, 3 KCl, 1 MgCl2, 1 CaCl2, 10 HEPES and 20 glucose (pH 7.4). Intracellular solution contained (in mM): 140 CsF, 10 NaCl, 10 HEPES, 1 EGTA and 15 sucrose (pH 7.3).

For recordings from Nav1.8 channels, the extracellular bath solution contained (in mM): 140 NaCl, 3 KCl, 1 MgCl2, 1 CaCl2, 10 HEPES, 10 glucose and 10 TEA‐Cl (pH 7.3). Intracellular solution contained (in mM): 140 CsF, 2 NaCl, 10 HEPES, 1 EGTA and 10 TEA‐Cl (pH 7.2). TEA‐Cl was added to inhibit endogenous K+ currents in N1E‐115 cells. Intracellular Na+ was reduced to 2 mM in order to shift the reversal potential outside the experimental range.

Voltage protocols

The concentration–response relationship for oxycodone was obtained using 50 ms pulses to −30 mV every 10 s. After three control pulses without addition of the drug, oxycodone was applied. For concentrations up to 300 μM, increasing oxycodone concentrations were applied on the same cell for six pulses, followed immediately by the next higher concentration (total of 27 pulses). Oxycodone (1 and 3 mM) was applied to separate cells, as these concentrations had to be diluted differently. To determine oxycodone block, the average inward current of the final two pulses for each concentration was normalized to the average of all three control pulses from the same cell. To calculate the concentration of half‐maximal channel inhibition (IC50), normalized inward currents for each log‐transformed concentration were fitted by a Hill equation: y = ymin + (ymax − ymin)/(1 + 10[C − LogIC50]), where y min and y max are the minimum and maximum amplitude, respectively, and C is the concentration. Channel block observed in vehicle‐treated cells was subtracted from oxycodone‐mediated block for the Hill plot.

The current–voltage (I–V) relationship was obtained using 40 ms pulses to a range of test potentials in 10 mV steps with an interval of 5 s. Conductance (G) was derived from current amplitude (I) at each voltage (V) using the equation G = I/(V − Vrev), where V rev is the reversal potential, determined for each cell individually. Conductance of each cell was normalized to the same cell's peak conductance in order to fit the data with a Boltzmann function. Normalized conductance–voltage curves were fitted with a Boltzmann equation: G/Gmax = Gmin + (Gmax − Gmin)/(1 +  exp [(V1/2 − V)/k]), where G min and G max are the minimum and maximum conductance, respectively, V 1/2 is the voltage at half‐maximal channel activation and k is the slope factor.

Voltage dependence of steady‐state fast inactivation was measured using a series of 500 ms pre‐pulses, followed by a 40 ms test‐pulse to 0 mV to assess the available, non‐inactivated, channels. Inward current measured during the test‐pulse to 0 mV was normalized to the cell's maximum test‐pulse inward current. Normalized peak inward current amplitude at each test‐pulse is displayed as a function of pre‐pulse potential and fitted using the above Boltzmann equation.

Use‐dependent block was determined by application of 30 activating pulses to 0 mV (for Nav1.5 and Nav1.7 channels) or to +10 mV (for Nav1.8 channels) at a frequency of 10 Hz and measuring the peak inward current at each pulse. Inward current measured at each pulse was normalized to the first response in order to control for variability in inward current amplitude. For quantification, the AUC for each graph was calculated and compared.

We investigated the recovery from fast inactivation using a two‐pulse voltage protocol with a recovery inter‐pulse of varying duration in between. Cells were first depolarized to −30 mV for 500 ms from a holding potential of −150 mV. This depolarizing pre‐pulse (I 1) was immediately followed by the inter‐pulse to −150 mV. The duration of this step increased with each sweep by a factor of 2, from 0.1 ms to 1.638 s. The inter‐pulse was immediately followed by the depolarizing test‐pulse (I 2) to −30 mV for 40 ms to assess the amount of recovered channels. Current amplitude at the test‐pulse I 2 was normalized to the amplitude measured at the pre‐pulse I 1 and plotted against inter‐pulse duration. Plotted values were fitted with a double exponential equation: Y = Y0 + Afast * (1 −  exp [−Kfast * X]) + Aslow * (1 −  exp [−Kslow * X]), where Y 0 is the current amplitude at time 0, A fast and A slow represent the amplitude coefficient for the fast and the slow time constants, K fast and K slow represent the two rate constants and X is the time. The time constants τ fast and τ slow are the reciprocals of the respective rate constant K.

Onset of slow inactivation was tested using a protocol of three depolarizing pulses, consisting of a reference pulse (I 1) to account for time‐dependent changes, a pre‐pulse of varying duration to induce slow inactivation and a test‐pulse (I 2). The holding potential was set to −150 mV. Cells were first depolarized to −30 mV for 10 ms (I 1), followed by a return to −150 mV for 500 ms. The following pre‐pulse to 0 mV to induce slow inactivation started with a duration of 72.9 s and decreased in subsequent sweeps by a factor of 3 down to 100 ms. This pre‐pulse was followed by a return to −150 mV for 200 ms before depolarizing cells again to −30 mV for 40 ms (I 2) to test for available channels. We used a sweep interval of 90 s to allow for sufficient recovery between sweeps. Test‐pulse current amplitude was normalized to reference‐pulse amplitude (I 2/I 1) to calculate the amount of slow inactivated channels for each pre‐pulse duration. Plotted values were fitted with a single exponential equation: Y = (Y0 − Y∞) *  exp (−K * X) + Y∞, where in addition to the above parameters, Y ∞ is the current amplitude at steady state.

Tonic block of Nav channels by oxycodone 300 μM was measured using seven 50 ms pulses to 0 mV at 0.1 Hz. After three control pulses, perfusion was switched to oxycodone 300 μM for the remaining four pulses. To determine tonic block, inward current during oxycodone application was normalized to the average of all three control pulses from the same cell. The same procedure was done in control recordings where cells were perfused with vehicle instead of oxycodone. The average block obtained in vehicle experiments was then subtracted from each oxycodone‐blocked current to obtain the true oxycodone‐mediated block.

CardioExcyte 96 recordings

The CardioExcyte 96 (Nanion Technologies GmbH, Munich, Germany) was used in impedance and extracellular field potential (EFP) mode. After the cells had arrived, the sensor plate was mounted on the CardioExcyte 96 with an incubation chamber. The temperature (37°C), humidity and gas concentration (5% CO2) was ensured by Nanion Supply Unit and ibidi Gas Incubation System (ibidi GmbH, Planegg, Germany). Online parameters in EFP and impedance modes, for example, amplitude and beat rate, were monitored with the CardioExcyteControl software daily for 7 days to allow the formation of stable synchronous beating patterns.

Before compounds were added, media were completely removed from the wells, and 100 μL fresh media were added to ensure the exact volume per well. The cells were allowed to re‐equilibrate for 2 h, and online parameters were monitored to ensure a stable baseline. The experiment was set up as described in Obergrussberger et al. (2016) making sure to insert the correct information into the compound plates and reference solutions. Oxycodone was prepared at 10× concentration in culture media. To add the compound, 10 μL of media was removed from the well, and 10 μL of the 10× concentrated compound was added. After application of the compound, measurements in impedance and EFP were taken every 5 min for 30 s for a period of 30 min.

Data and statistical analysis

The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015). Normalization was performed in order to control for large variations in sodium channel expression and inward current amplitude and in order to be able to fit the recorded data with a Hill function (for concentration–response curves), a Boltzmann function (for voltage dependence) or an exponential function (for time courses of inactivation). Details on normalization are given in Methods or in the figure legends. For data analysis, FitMaster v2.8 software (HEKA electronics), Igor Pro (WaveMetrics, Portland, OR, USA) and GraphPad Prism 5 or 6 (GraphPad Software, Inc., La Jolla, CA, USA) were used. Statistical analysis was performed using GraphPad Prism version 5 or 6. Measurements before and after drug or vehicle application were compared by a paired two‐tailed t‐test. Comparisons between three or more groups were performed using a one‐way ANOVA followed by Bonferroni's or Dunnett's multiple comparisons test. Data are presented as mean ± SEM. To reject outliers, time constants outside of mean ± 3 SD were ignored. P <0.01 was considered to show significant differences between means. Variations in group sizes between pre‐application and post‐application data are due to loss of cells during the recording. Low n number in some induced pluripotent stem cell (iPSC)‐derived cardiomyocyte experiments is due to unstable culture and recording of iPSC‐derived cardiomyocytes. Due to low n, no outliers were defined for iPSC‐derived cardiomyocytes. Any group of n < 5 was not included for statistical comparison. No blinding was performed.

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 2017/18 (Alexander et al., 2017a,b).

Results

Oxycodone concentration‐dependently blocks Nav1.5 channels

Recent data have shown that oxycodone can induce cardiac arrhythmia (Fanoe et al., 2009) that can eventually lead to sudden cardiac arrest. To investigate if the cardiac side effects of oxycodone are mediated by interaction with the sodium channel Nav1.5, we stably expressed the channel in HEK293 cells and measured voltage‐activated Nav currents under the influence of increasing concentrations of oxycodone (Figure 1A). Application of oxycodone led to a concentration‐dependent current reduction (Figure 1A) that was fully reversible upon washout (Figure 1B). Control currents, measured during vehicle treatment, decreased slightly over time, reaching approximately 90% of the initial current amplitude at the end of the protocol (approximately 4.5 min) (Figure 1C). Oxycodone‐mediated current reduction was significantly more pronounced than vehicle treatment at 100 μM and higher (Figure 1A, B). The maximum block, achieved by 3 mM oxycodone, was 53%. Fitting the normalized current amplitude with a Hill equation revealed an IC50 for oxycodone of 483.2 μM (Figure 1D).

Figure 1.

Figure 1

Open in a new tab

Oxycodone concentration‐dependently blocks the cardiac Nav1.5 channels. Human Nav1.5 channels, stably expressed in HEK293 cells, were treated with increasing concentrations of oxycodone. (A) Representative current traces of one individual cell treated with increasing oxycodone concentrations. The applied voltage protocol is displayed above. Scale bar represents 1 ms and 2 nA. (B) Oxycodone‐mediated block is fully reversible. Representative current traces of one individual cell treated with bath solution (control), followed by 1 mM oxycodone, followed again by bath solution for washout. Scale bar represents 1 ms and 2 nA. (C) Concentration‐dependent block by oxycodone compared with vehicle treatment. Inward current was normalized to the average of three control pulses (before oxycodone or vehicle application). Data shown are means ± SEM from n = 10 cells for oxycodone and n = 5 cells for vehicle. Oxycodone applications of 1 and 3 mM were performed on separate cells (n = 4) due to different dilution of the drug, along with corresponding vehicle applications (n = 3). (D) A Hill plot (solid line) on the concentration–response curve revealed an IC50 of 483.2 μM. Vehicle‐induced block, as shown in (C) was subtracted from oxycodone block.

Oxycodone does not affect voltage dependence of activation or steady‐state fast inactivation of Nav1.5 channels

Compounds inhibiting Nav channels may also shift the voltage sensitivity of the channel (Schulze et al., 2014). To test whether oxycodone affects the voltage sensitivity of Nav1.5 channels, we measured the voltage dependence of channel activation before and after vehicle or oxycodone (100 μM) application (Figure 2A–C). Oxycodone, but not vehicle, clearly reduced current amplitude at each test potential (Figure 2A, B). However, voltage dependence was not affected by either vehicle or oxycodone, as can be seen from Boltzmann fits to conduction–voltage curves (Figure 2C).

Figure 2.

Figure 2

Open in a new tab

Oxycodone has no effect on activation or steady‐state fast inactivation of Nav1.5 channels. (A) Representative current traces recorded in the same HEK293cell expressing Nav1.5 channels, before (left) and after (right) application of oxycodone 100 μM. The indicated voltage protocol was applied. Currents elicited by a voltage step to −70 mV and −40 mV are highlighted. (B) Current–voltage relationship obtained in recordings such as in (A). Oxycodone (100 μM) reduces current amplitude. (C) Conductance–voltage relationship obtained from curves displayed in (B). Conductance of each cell was normalized to the same cell's peak conductance in order to fit the data with a Boltzmann function. Solid lines represent mean Boltzmann function. The following V 1/2 values were obtained based on these fits: pre‐vehicle −56 ± 1.6 mV (n = 18), post‐vehicle −59 ± 1.8 mV (n = 16), pre‐oxycodone 100 μM −55 ± 0.9 mV (n = 28) and post‐oxycodone 100 μM −58 ± 1.1 mV (n = 28). (D) Voltage dependence of steady‐state fast inactivation was measured using the indicated voltage protocol. Inward current measured during the test‐pulse to 0 mV was normalized to the cell's maximum inward current and plotted against pre‐pulse voltage. Solid lines are best fits to a Boltzmann function. V 1/2 values obtained from these fits, with their means ± SEM, are displayed on the right: pre‐vehicle, (n = 7), post‐vehicle, (n = 7), pre‐oxycodone (100 μM) (n = 9) and post‐oxycodone (100 μM) (n = 10). The significant negative shift in voltage dependence of fast inactivation is purely time‐dependent as it occurs both during oxycodone and during vehicle application. *P<0.01, significantly different as indicated; one‐way ANOVA with Bonferroni multiple comparisons test.

Oxycodone and other μ receptor agonists have been shown to shift steady‐state fast inactivation of sodium currents in cultured thalamic neurons (Hashimoto et al., 2009). To test whether oxycodone has a similar effect on cardiac Nav1.5 channels, we measured steady‐state fast inactivation before and after vehicle or oxycodone (100 μM) application using the protocol shown in Figure 2D. We noted a −13 mV hyperpolarizing shift in voltage dependence of fast inactivation even in the vehicle‐treated cells (Figure 2D). Recordings of fast inactivation before and after solution change were approximately 8 min apart. The observed shift was therefore likely to be time‐dependent and not affected by application of 100 μM oxycodone.

Block by oxycodone is use‐dependent

Next, we tested if the block of voltage‐activated Nav1.5 currents (Figure 1) requires opening of the channel. If this was the case, repeated activation of Nav1.5 channels during oxycodone application should increase current block as more and more oxycodone molecules can access the channel pore with each subsequent activation. Such a use‐dependent block was measured using a range of 30 activating voltage pulses at a rate of 10 Hz. Cells were either treated with vehicle or oxycodone 100 or 300 μM. Currents measured before any solution change and those measured during vehicle application showed a reduction in current amplitude only during the first two to three voltage pulses, after which current amplitude reached a steady state (Figure 3A). During vehicle application, current amplitude declined by 22% within the first three activating pulses with only a minor decrease occurring after that. Application of oxycodone on the other hand led to a continuous reduction of current amplitude that continued throughout the entire length of the protocol and current amplitude did not reach a steady state (Figure 3A). Oxycodone (100 μM) reduced current amplitude after 30 pulses by 41% and 300 μM oxycodone even reduced current amplitude by 58%. Use‐dependent block by both oxycodone concentrations was significantly more pronounced than vehicle‐induced block (Figure 3B).

Figure 3.

Figure 3

Open in a new tab

Oxycodone‐induced block of Nav1.5 channels is use‐dependent. (A) HEK293 cells, stably expressing Nav1.5 channels, were stimulated with 30 depolarizing pulses to 0 mV at a frequency of 10 Hz. Inward current measured at each pulse was normalized to the first response. Oxycodone (100 and 300 μM) induced a use‐dependent block that increases with repeated channel activity. (B) Quantification of the data presented in (A). The AUC was calculated for each condition for statistical comparison between treatments and is shown as mean ± SEM; pre‐vehicle, n = 18; post‐vehicle, n = 19; pre‐oxycodone (100 μM), n = 23; post‐oxycodone (100 μM), n = 25; pre‐oxycodone (300 μM), n = 10; post‐oxycodone (300 μM), n = 10. *P<0.01, significantly different as indicated; one‐way ANOVA with Bonferroni's multiple comparisons test.

Oxycodone slows recovery from fast inactivation

The use‐dependent block observed during application of oxycodone could be explained either by a slowing of recovery from steady‐state fast inactivation of Nav1.5 channels or by an accumulation of channels in the slow inactivated state (see below). We first tested the effect of oxycodone on the recovery from steady‐state fast inactivation. The voltage protocol can be seen in Figure 4A. Graphs plotting the recovery from fast inactivation could well be fitted by a double‐exponential function (see Methods), providing two time constants of recovery. Vehicle treatment did not affect the recovery of the channels (Figure 4A, B). Oxycodone 300 μM, however, significantly slowed the recovery of Nav1.5 channels, affecting both the fast and the slow component (Figure 4A, B). This indicates that, in the presence of oxycodone, Nav1.5 channels need more time to return to the activatable state. This also explains the observed use‐dependent block as, for each activating stimulus, fewer channels are available for activation.

Figure 4.

Figure 4

Open in a new tab

Oxycodone slows recovery from fast inactivation and increases slow inactivation of Nav1.5 channels. (A) Time course of recovery from fast inactivation using the indicated voltage protocol in HEK293 cells expressing Nav1.5 channels. Peak current measured at the test‐pulse (I 2) was normalized to the peak current obtained during the pre‐pulse (I 1). Data were fitted by a double‐exponential function as represented by the solid lines, providing a readout for the fast (τ fast) and slow (τ slow) time constants of channel recovery. The inset presents a magnified view into the first 200 ms. (B) Fast (left) and slow (right) time constants obtained from individual recordings. Data shown are means ± SEM: for τ fast, pre‐vehicle, n = 9; post‐vehicle, n = 9; pre‐oxycodone (300 μM), n = 11; post‐oxycodone (300 μM), n = 10: for τ slow, pre‐vehicle, n = 9; post‐vehicle, n = 10; pre‐oxycodone (300 μM), n = 11; post‐oxycodone (300 μM), n = 10. *P<0.01, significantly different as indicated; one‐way ANOVA with Bonferroni multiple comparisons test. (C) Onset of slow inactivation was tested using the indicated protocol. Peak current measured at the test‐pulse (I 2) was normalized to the peak current obtained during the reference‐pulse (I 1). Data were fitted by a single‐exponential function as represented by the solid lines. The inset presents a magnified view into the first 10 s. (D) The time constant of onset of slow inactivation for Nav1.5 channels, as obtained from individual recordings as in (C). Data shown are means ± SEM: pre‐vehicle, n = 8; post‐vehicle, n = 8; pre‐oxycodone (300 μM), n = 10; post‐oxycodone (300 μM), n = 9. *P<0.01, significantly different as indicated; one‐way ANOVA with Bonferroni multiple comparisons test.

Slow inactivation of Nav1.5 channels is enhanced by oxycodone

Next, we tested if an increasing amount of slow inactivated channels also contributes to the observed use‐dependent current rundown during oxycodone. We tested the onset of slow inactivation during vehicle or oxycodone (300 μM) treatment (Figure 4C, D). Recordings before any solution change or those during vehicle treatment showed only a small degree of slow inactivation of the Nav1.5 channels. After a 73 s inactivating stimulus, approximately 25% of channels entered the slow inactivated state with a slow time constant (Figure 4C, D). During application of oxycodone, we observed an increase in the amount of slow inactivation as well as a change in the time constant. In this condition, 46% of channels displayed slow inactivation with a significantly faster time constant (Figure 4C, D). These experiments indicate that the use‐dependent current rundown observed during oxycodone application can be explained both by a slowing of the recovery from fast inactivation as well as by an increase in the slow inactivation of Nav1.5 channels.

Oxycodone induces arrhythmic beating in stem cell‐derived cardiomyocytes

In addition to the observed effects on Nav1.5 channels , oxycodone is known to inhibit hERG potassium channels (Fanoe et al., 2009). Here, we aimed at investigating whether the effects of oxycodone on these two cardiac ion channels combined to induce cardiac arrhythmia. To that end, we used human induced pluripotent stem cell (hiPSC)‐derived cardiomyocytes with a commercially available 96‐well screening platform (CardioExcyte 96, Nanion Technologies GmbH) for EFP and impedance measurements. Cardiomyocytes were pre‐plated on the recording electrodes of the CardioExcyte sensor plates where they formed a spontaneously beating confluent monolayer. We recorded the EFP, which originates from the propagation of the cardiac cells' action potential across the electrode array and is comparable with the ECG of the heart. In addition, the recorded impedance gives an indirect readout of the cell's contractions. Under control conditions, cardiomyocytes displayed spontaneous and regular EFP and impedance signals with a beat rate of 58 ± 1 beats min−1 (Figure 5A–D). Lower concentrations of oxycodone (10 or 30 μM) had no effect on hiPSC‐derived cardiomyocytes. When we added oxycodone 100 or 300 μM to the cells, we observed clear arrhythmic beating (Figure 5A–D), combined with a strongly reduced cardiac beat rate after 10 min (Figure 5E). This latter finding supports our earlier results as cardiomyocyte beat rate strongly depends on Nav1.5 channel activity. It should be noted that even 100 μM oxycodone, which is well below the IC50 for Nav1.5 channels (Figure 1D), causes arrhythmia and reduces the beat rate. The field potential duration (FPD), which is analogous to the QT interval in the ECG (Halbach et al., 2003; Millard et al., 2017), was dramatically increased after addition of 100 μM oxycodone (Figure 5F). Oxycodone (300 μM) also increased FPD, but this was not tested for significance due to a limited number of data. These changes in FPD represent a slowing in cardiac action potential repolarization and can therefore be assumed to be due to modification of hERG channels (Fanoe et al., 2009).

Figure 5.

Figure 5

Open in a new tab

Oxycodone induces arrhythmic beating in hiPSC‐derived cardiomyocytes. (A, B) Representative recordings show regular spontaneous EFPs. In (A), addition of oxycodone (oxy) 100 μM or (B) 300 μM induces arrhythmia and a slowing of the EFP frequency. Pre‐oxycodone and post‐oxycodone traces were recorded from the same cardiomyocyte network. (C, D) Impedance recordings from the same wells shown in (A) and (B). Under control conditions, cells display regular spontaneous contractions. Addition of oxycodone (C) 100 μM or (D) 300 μM leads to irregular cardiac contractions. (E) Oxycodone reduces the beat rate of hiPSC‐derived cardiomyocytes. The basal beat rate was 58 ± 1 beats min−1 and the changes in beat rate measured after 10 min are shown as means ± SEM; vehicle, n = 8 wells; oxycodone (10 μM), n = 6 wells; oxycodone (30 μM) n = 6 wells; oxycodone (100 μM), n = 5 wells; oxycodone (300 μM), n = 4 wells. *P<0.01, significantly different from control; one‐way ANOVA with Dunnett's multiple comparisons test. (F) The extracellular FPD represents the duration of the cardiac action potential. Oxycodone strongly increases FPD, measured after 10 min, at the higher concentrations. The basal FPD was 251.5 ± 15.0 ms. Data (changes in FPD as % control) are shown as means ± SEM; vehicle, n = 8 wells; oxycodone (10 μM), n = 6 wells; oxycodone (30 μM) n = 6 wells; oxycodone (100 μM), n = 6 wells; oxycodone (300 μM), n = 5 wells. *P<0.01, significantly different as indicated; one‐way ANOVA with Dunnett's multiple comparisons test. (G) The early downstroke of the EFP represents the depolarization of the cardiac action potential and therefore closely correlates to Nav1.5 channel activity. Traces show zoomed‐in sections of EFP spikes [as seen in (A)], averaged over all wells treated under the same condition. Traces are normalized to the minimum value. Scale bar (normalized values on Y axis) is the same for all traces. Oxycodone reduces the spike amplitude and the max.DV. (H) The slope of the early downstroke (max.DV) is strongly reduced by oxycodone. Data shown are means ± SEM: vehicle, n = 7 wells; oxycodone (10 μM), n = 4 wells; oxycodone (30 μM) n = 4 wells; oxycodone (100 μM), n = 5 wells; oxycodone (300 μM) n = 3 wells. Post‐application values from each well were always normalized to the same well's pre‐application values. Any group of n < 5 was not included for statistical comparison. *P<0.01, significantly different from control; Mann Whitney test.

Next, we analysed the fast downstroke in the beginning of the EFP (Figure 5G), which corresponds to the upstroke of the cardiac action potential and therefore specifically represents Nav1.5 channel activity (Halbach et al., 2003). The slope of this fast downstroke [maximum negative downstroke velocity (max.DV)] was significantly reduced 10 min after addition of oxycodone (100 or 300 μM; Figure 5G–I). This supports our results, showing that oxycodone reduces the availability of Nav1.5 channels during the cardiac action potential. In summary, high concentrations of oxycodone act on both Nav1.5 and hERG channels to induce arrhythmia in human cardiomyocytes.

Oxycodone does not affect nociceptive Nav channels

Recent studies have shown that activating the opioid system by, for example, oxycodone can increase the efficacy of blockers of Nav1.7 channels (Deuis et al., 2017; Flinspach et al., 2017). As oxycodone directly affects Nav1.5 channels, we wanted to assess if some of the analgesic effects of oxycodone could also be explained by a direct interaction of the compound with nociceptive Nav channels. We therefore tested the drug on heterologously expressed Nav1.7 and Nav1.8 channels. We observed a small degree of tonic block in both Nav1.7 and Nav1.8 channels, which was significantly smaller than the block observed in Nav1.5 channels (Figure 6A). Analogous to what was observed in Nav1.5 channels, voltage dependence of activation and of steady‐state fast inactivation was not affected by oxycodone in either Nav1.7 or Nav1.8 channels (Figure 6B, D). The strong time‐dependent hyperpolarizing shift of fast inactivation, which was described above for Nav1.5 channels, could also be observed in Nav1.7 and Nav1.8 channels.

Figure 6.

Figure 6

Open in a new tab

Oxycodone has no effect on nociceptive Nav1.7 and Nav1.8 channels. (A) Tonic block induced by oxycodone (300 μM) was compared between HEK293 cells, stably expressing either Nav1.5 or Nav1.7 channels and N1E‐115 cells, transiently transfected with Nav1.8 channels. The right panel displays the percentage of channel block for each channel subtype. Current rundown that was observed in vehicle experiments was subtracted from oxycodone‐induced block to obtain the true oxycodone‐mediated effect. Nav1.5 channels (n = 30) were blocked to a greater extent than either Nav1.7 or Nav1.8 channels (n = 8 each). Data shown are means ± SEM. *P<0.01, significantly different as indicated; one‐way ANOVA with Bonferroni multiple comparisons test. (B) Oxycodone 300 μM does not affect voltage dependence of activation or steady‐state fast inactivation of Nav1.7 channels. Analysis was carried out as in Figure 2. Similar to results obtained with Nav1.5 channels, there seems to be a small time‐dependent negative shift in fast inactivation of Nav1.7 channels, which is visible both in oxycodone‐treated and vehicle‐treated cells (activation: pre‐vehicle and post‐vehicle n = 6, pre‐oxycodone n = 8 and post‐oxycodone n = 7; inactivation pre‐vehicle n = 7, post‐vehicle n = 6, pre‐oxycodone and post‐oxycodone n = 8). (C) Use‐dependent block of Nav1.7 channels by oxycodone 300 μM, analysed as in Figure 3. Nav1.7 channels displayed current rundown over time, but this was not significantly enhanced by oxycodone. Inset shows the AUC for each group (pre‐vehicle n = 7, post‐vehicle n = 6, pre‐oxycodone n = 8 and post‐oxycodone n = 7). (D) Voltage dependence of activation and steady‐state fast inactivation of Nav1.8 channels were not influenced by oxycodone 300 μM. The small time‐dependent negative shift in fast inactivation also applies to Nav1.8 channels (activation: pre‐vehicle n = 6, post‐vehicle n = 7 and pre‐oxycodone and post‐oxycodone n = 8; inactivation: pre‐vehicle and post‐vehicle n = 7 and pre‐oxycodone and post‐oxycodone n = 8). (E) Use‐dependent block of Nav1.8 channels by oxycodone (300 μM). Nav1.8 current amplitude remained very stable over time, in contrast to Nav1.5 and Nav1.7 currents. Oxycodone has no use‐dependent effect (pre‐vehicle and post‐vehicle n = 7 and pre‐oxycodone and post‐oxycodone n = 8). n.s., not significant; one‐way ANOVA with Bonferroni multiple comparisons test.

We did not observe a significant use‐dependent block of currents mediated by Nav1.7 or Nav1.8 channels. Nav1.7 channels showed a small degree of current rundown over time even in vehicle experiments, and this rundown was slightly but not significantly increased with 300 μM oxycodone (Figure 6C). In contrast, Nav1.8 channels displayed no current rundown and also showed no used‐dependent block with oxycodone (Figure 6E). These results indicate that oxycodone is unlikely to affect slow inactivation or recovery from fast inactivation in nociceptive Nav channels. Taken together, in contrast to Nav1.5 channels , no effect could be observed by oxycodone on either Nav1.7 or Nav1.8 channels.

Discussion

Oxycodone is a semi‐synthetic opioid, commonly used as a potent analgesic, especially for treating cancer pain (Biancofiore, 2006). The use of oxycodone has dramatically increased over the recent years (Paulozzi and Ryan, 2006), making the drug one of the most commonly used potent analgesics. However, recent studies have shown that oxycodone treatment can lead to cardiac side effects, and concerns about the consumption of opioids have now even led to a declaration of a ‘health emergency’ in the USA (https://www.nytimes.com/2017/10/26/us/politics/trump-opioid-crisis.html).

Here, we showed that oxycodone has a use‐dependent inhibitory effect on the cardiac sodium channel Nav1.5. We observed a concentration‐dependent block of the current carried by Nav1.5 channels and an increased current rundown during repeated activation of the channel. We showed that the latter effect is due to two effects of oxycodone on Nav1.5 channels: slowing of recovery from fast inactivation and increase of slow inactivation. All of these effects together with the previously observed effects of oxycodone on potassium channels are likely to induce cardiac arrhythmia in human iPSC‐derived cardiomyocytes.

Oxycodone affects Nav1.5 channels

The inhibitory effects of oxycodone on Nav1.5 channels were only observed at concentrations of 100 μM or higher. This is in agreement with earlier findings by Hashimoto et al. (2009), who showed a minor block of Nav currents in thalamic neurons by 100 μM oxycodone. We calculated an IC50 of 483.2 μM, which is very high, seemingly making the drug safe to use in a typical range used for cancer pain therapy. However, the overall effect of oxycodone on Nav1.5 channels would be slow to develop– mediated by a slow accumulation of Nav1.5 channels in an inactivated state. It is therefore possible that even lower concentrations of oxycodone could cause abnormal ECG changes after a certain amount of time. In addition, we showed that the inhibitory action of oxycodone was use‐dependent and therefore correlated with cardiac activity. Increased myocardial activity in a patient, for example, due to stress or physical activity, could therefore facilitate the onset of ECG changes induced by oxycodone.

A variety of mutations are known to affect Nav1.5 channel function and cause cardiac syndromes, such as long QT syndrome type 3 (Ruan et al., 2009). These mutations usually affect either the voltage dependence of the activation or inactivation of Nav1.5 channels or they induce a persistent Nav1.5 current that does not inactivate. Although oxycodone also affects the QT interval (Fanoe et al., 2009), it does not change persistent current or inactivation kinetics (Supporting Information Figure S1B, C). Consistent with these findings, we did not see any effect of oxycodone on voltage dependence of activation or steady‐state fast inactivation. The latter is in contrast to findings by Hashimoto et al. (2009), who found a hyperpolarizing shift of fast inactivation induced by oxycodone in thalamic neurons. However, we found a strong time‐dependent shift in steady‐state fast inactivation that was equally potent in both oxycodone‐treated and vehicle‐treated cells. These results demonstrate the extent of time‐dependent changes in voltage sensitivity of Nav channels and highlight the importance of including vehicle‐treated time controls in the analysis to avoid misinterpretation of data. It is possible that the oxycodone‐induced shift of fast inactivation shown in thalamic neurons (Hashimoto et al., 2009) is a misinterpretation of time‐dependent changes, due to the absence of appropriate vehicle controls.

Oxycodone affects cardiomyocyte activity

Oxycodone is known also to inhibit cardiac hERG potassium channels at high concentrations (IC50 171 μM; Fanoe et al., 2009). While the activity of oxycodone in pharmacological concentrations on either hERG or Nav1.5 channels alone might not be sufficient to induce cardiac side effects in a patient, the combined activity on both channels could very well affect cardiac myocyte excitability. Block of hERG channels, which are crucial for proper repolarization of the myocytes, would lead to a prolonged depolarization. This in turn induces slow inactivation of Nav1.5 channels , thus further increasing the amount of inactivated Nav1.5 channels. In the end, a small block of hERG and Nav1.5 channels, in combination with an accumulation of Nav1.5 channels in an inactivated state can be expected to negatively affect cardiac excitability, potentially prolonging the QT interval and inducing cardiac arrhythmia. Indeed, our data from iPSC‐derived human cardiomyocytes show that even 100 μM – a concentration that only induces small changes in Nav1.5 channels – induces arrhythmia and a reduction of cardiac beat rate. This would be compatible with effects on both hERG and Nav1.5 channels. The effect on hERG channels is represented by the increased FPD (Halbach et al., 2003). This is analogous to an increase of the QT interval of the ECG (Millard et al., 2017) and therefore supports the findings by Fanoe et al. (2009). The oxycodone‐induced effects on Nav1.5 channels can be seen in the reduced cardiomyocyte beat rate and in changes in the initial rapid component of the EFP signal as quantified by the max.DV. A reduced cardiomyocyte beat rate would translate to bradycardia in the patient. The changes in the downstroke velocity of the EFP represent a smaller and slower cardiac depolarization in the presence of oxycodone. Ultimately, oxycodone (100 μM) would lead to a slower depolarization and a slower repolarization of the cardiac action potential, and this, in turn, would induce a lower heart rate and cardiac arrhythmia. It should be noted that smaller but non‐significant changes in most cardiomyocyte parameters could be observed even at a concentration of 30 μM oxycodone. This may indicate some cardiac activity of oxycodone even at lower concentrations.

Clinical implications

Estimating a realistic plasma concentration in patients regularly treated with oxycodone is not trivial. Most studies investigated the plasma concentration in healthy subjects after only a single low dose of oxycodone. Using such a paradigm, Nieminen et al. (2010) found the oxycodone peak plasma concentration after a single 15 mg dose of oral oxycodone to be 34 ng·mL−1 (approximately 97 nM), which is far lower than the IC50 measured in the present study. However, the mean daily oral oxycodone dose used in cancer pain patients has been reported as 150 mg, and single doses can go as high as 302 mg (Kalso and Vainio, 1990). Fanoe et al. (2009) even tested patients treated with up to 550 mg oxycodone. In addition, one has to assume that patients who were treated long term with oxycodone could have higher plasma concentrations, especially as oxycodone clearance is reduced with age (Saari et al., 2012) and in end‐of‐life situations due to organ failure. Using simulations of repetitive dosing, it has been suggested that the reduced clearance can lead to a 20% increase in oxycodone plasma concentration in the elderly (Saari et al., 2012). In addition, a post‐mortem study that evaluated over 1000 fatality cases involving oxycodone showed a concentration of oxycodone as high as 48 μM in a single case (Cone et al., 2004). While we show that this concentration is not high enough to induce use‐dependent block of Nav1.5 channels (Supporting Information Figure S1A), our data on iPSC‐derived cardiomyocytes show that such a concentration would already induce small changes in cardiac beating. Furthermore, because this plasma concentration of 48 μM was measured post‐mortem, the actual plasma concentration at time of death may have been higher, which would almost certainly have affected cardiac activity.

Oxycodone is now heavily involved in drug abuse and addiction. More than 2.4 million people suffer from opioid use disorder in the USA with oxycodone being one of the most commonly misused opioids (Itzoe and Guarnieri, 2017). Additionally, there are a number of alternative opioids available on the market, and abusers have become more and more prone to ingesting multiple different opioids simultaneously (Cone et al., 2004). Methadone, another synthetic opioid, is one of the substances regularly ingested in combination with oxycodone (Cone et al., 2004). A recent study showed similar inhibition of Nav1.5 channels by methadone (Schulze et al., 2014) as we have shown here for oxycodone. It can therefore be assumed that the combination of oxycodone and methadone (and potentially other opioids) carries the risk of inducing severe and potentially fatal cardiac side effects. Furthermore, oxycodone and other opioid plasma concentrations can be assumed to be significantly elevated in drug addicts (de Vos et al., 1995; Cone et al., 2004), adding to the potential risk of inducing sudden cardiac arrest in drug addicts.

Implications for the clinic regarding safe long‐term dosage of oxycodone are difficult to predict. As oxycodone acts on several different cardiac ion channels and because plasma concentrations of patients under long‐term treatment are not known, we cannot predict any time frame for the development of potentially dangerous plasma levels in the patient. This would also heavily depend on the patient's age, general health and treatment history. However, based on our study and clinical experience, oxycodone in, for example, cancer pain therapy seems to be safe regarding the development of severe cardiac side effects, such as TdP, especially if used in low doses. Nevertheless, future studies should examine translation of findings from this study into patients to better determine these effects, and drug addicts as well as opioid abusers should be monitored for severe cardiac arrhythmia.

Oxycodone does not act on peripheral neuronal Nav channels

Recent studies have shown that activating the opioid system by, for example, oxycodone can increase the efficacy of blockers of Nav1.7 channels (Deuis et al., 2017; Flinspach et al., 2017), which are abundantly expressed in sensory DRG neurons and responsible for action potential generation in nociceptive neurons. Here, we tested whether oxycodone could have direct effects on these channels that might be independent of its opioid receptor activity. Similarly, we tested whether some of the analgesic activity of oxycodone could be mediated by direct activity on Nav1.8 channels, which are also important for action potential generation in peripheral sensory neurons. We did not see any effect of the drug on voltage dependence of activation or steady‐state fast inactivation of either Nav1.7 or Nav1.8 channels. A small degree of channel block could be observed in both channels, but this block was significantly smaller compared with the block achieved in Nav1.5 channels. Considering the high concentration of oxycodone (300 μM) that was required to observe this small block of nociceptive Nav channels, it seems unlikely that inhibition of Nav1.7 or Nav1.8 channels contributes to the analgesic effect of oxycodone in patients.

In conclusion, we have shown that the commonly used opioid oxycodone induces a concentration‐dependent block of the cardiac sodium channel Nav1.5. The block is use dependent and mediated by a slowing of the recovery of Nav1.5 channels from fast inactivation, in combination with an increase of slow inactivation of Nav1.5 channels. Taken together, these processes lead to accumulation of Nav1.5 channels in an inactivated state. Together with a block of hERG potassium channels, the effects on Nav1.5 channels induce a reduction of cardiomyocyte beating and arrhythmia. While the concentrations needed to elicit these cardiac effects are relatively high, it can be assumed that drug abusers and addicts consuming high doses of oxycodone might be at risk of suffering from severe cardiac side effects induced by the effects of oxycodone on Nav1.5 channels. In further studies, it would be beneficial to examine how these in vitro findings translate to the patient. Specifically, such studies should focus on assessing the myocardial activity of patients taking high doses of oxycodone for, for example, pain management.

Author contributions

J.E.M. planned and designed the experiments, performed and analysed all the whole‐cell patch‐clamp experiments and wrote the manuscript. K.J. planned and designed the experiments, performed and analysed all the cardiomyocyte experiments and contributed to writing the manuscript. S.S.‐F. planned and designed the experiments and critically revised the manuscript. V.P.‐P. and R.R. conceived the study, interpreted the data and critically revised the manuscript. A.L. conceived the study, planned and designed experiments, interpreted data and critically revised the manuscript.

Conflict of interest

K.J. and S.S.‐F. are employees of Nanion Technologies GmbH, which developed the CardioExcyte 96 used in the manuscript. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. R.R. has received support from the German GBA‐Innovationsfonds (grant ID 01VSF16007) and lecture and scientific advisory board fees from Astellas, Grünenthal, Lilly, Pfizer and TEVA. R.R. has no conflict of interest regarding the present study. All other authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This http://onlinelibrary.wiley.com/doi/10.1111/bph.13405/abstract acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organisations engaged with supporting research.

Supporting information

Figure S1 Oxycodone does not affect persistent current or fast inactivation kinetics. (A) Oxycodone 50 μM does not induce use‐dependent block. HEK293 cells, stably expressing Nav1.5 channels, were stimulated with 30 depolarising pulses to 0 mV at a frequency of 10 Hz. Inward current measured at each pulse was normalised to the first response. Oxycodone 50 μM does not induce use‐dependent block beyond time‐dependent changes that can also be seen in vehicle recordings. The plot on the right shows the area under the curve (AUC) for each condition for statistical comparison. The following values were calculated: pre vehicle 25.5 ± 0.5 (n = 18), post vehicle 22.2 ± 0.8 (n = 19), pre oxycodone 50 μM 26.8 ± 0.6 (n = 4), post oxycodone 50 μM 23.0 ± 1.0 (n = 4). The vehicle data presented here for comparison are identical to the ones shown in Figure 3. (B) Oxycodone 300 μM does not increase persistent current in Nav1.5 channels. Persistent current was measured as described below. (C) Oxycodone 300 μM also does not change kinetics of steady‐state fast inactivation. A single exponential fit to the inactivating current component of regular IV recordings provided the time constant (τ) of inactivation.

Click here for additional data file. (417.9KB, pdf)

Acknowledgements

We thank Brigitte Hoch for excellent technical assistance. We thank Tom Götze for initiating the collaboration with Nanion Technologies.

Meents J. E., Juhasz K., Stölzle‐Feix S., Peuckmann‐Post V., Rolke R., and Lampert A. (2018) The opioid oxycodone use‐dependently inhibits the cardiac sodium channel NaV1.5, British Journal of Pharmacology, 175, 3007–3020, https://doi.org/10.1111/bph.14348.

References

  1. Alexander SPH, Christopoulos A, Davenport AP, Kelly E, Marrion NV, Peters John A et al (2017a). The Concise Guide to PHARMACOLOGY 2017/18: G protein‐coupled receptors. Br J Pharmacol 174: S17–S129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Alexander SPH, Striessnig J, Kelly E, Marrion NV, Peters JA, Elena F et al (2017b). The Concise Guide to PHARMACOLOGY 2017/18: Voltage‐gated ion channels. Br J Pharmacol 174: S160–S194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Behr ER, Savio‐Galimberti E, Barc J, Holst AG, Petropoulou E, Prins BP et al (2015). Role of common and rare variants in SCN10A: results from the Brugada syndrome QRS locus gene discovery collaborative study. Cardiovasc Res 106: 520–529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Biancofiore G (2006). Oxycodone controlled release in cancer pain management. Ther Clin Risk Manag 2: 229–234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cone EJ, Fant RV, Rohay JM, Caplan YH, Ballina M, Reder RF et al (2004). Oxycodone involvement in drug abuse deaths. II. Evidence for toxic multiple drug–drug interactions. J Anal Toxicol 28: 616–624. [DOI] [PubMed] [Google Scholar]
  6. Curtis MJ, Bond RA, Spina D, Ahluwalia A, Alexander SPA, Giembycz MA et al (2015). Experimental design and analysis and their reporting: new guidance for publication in BJP. Br J Pharmacol 172: 3461–3471. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Davis MP, Varga J, Dickerson D, Walsh D, LeGrand SB, Lagman R (2003). Normal‐release and controlled‐release oxycodone: pharmacokinetics, pharmacodynamics, and controversy. Support Care Cancer Off J Multinatl Assoc Support Care Cancer 11: 84–92. [DOI] [PubMed] [Google Scholar]
  8. Deuis JR, Dekan Z, Wingerd JS, Smith JJ, Munasinghe NR, Bhola RF et al (2017). Pharmacological characterisation of the highly NaV1.7 selective spider venom peptide Pn3a. Sci Rep 7: 40883. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fanoe S, Jensen GB, Sjøgren P, Korsgaard MPG, Grunnet M (2009). Oxycodone is associated with dose‐dependent QTc prolongation in patients and low‐affinity inhibiting of hERG activity in vitro. Br J Clin Pharmacol 67: 172–179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Flinspach M, Xu Q, Piekarz AD, Fellows R, Hagan R, Gibbs A et al (2017). Insensitivity to pain induced by a potent selective closed‐state Nav1.7 inhibitor. Sci Rep 7: 39662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Güttler K, Sabatowski PDR (2008). Differenzialtherapeutische Aspekte in der Schmerztherapie mit retardierten, oralen, stark wirksamen Opioiden. Schmerz 22: 562–570. [DOI] [PubMed] [Google Scholar]
  12. Halbach M, Egert U, Hescheler J, Banach K (2003). Estimation of action potential changes from field potential recordings in multicellular mouse cardiac myocyte cultures. Cell Physiol Biochem 13: 271–284. [DOI] [PubMed] [Google Scholar]
  13. Hamunen K, Paakkari P, Kalso E (2009). Trends in opioid consumption in the Nordic countries 2002–2006. Eur J Pain 13: 954–962. [DOI] [PubMed] [Google Scholar]
  14. Harding SD, Sharman JL, Faccenda E, Southan C, Pawson AJ, Ireland S et al (2018). The IUPHAR/BPS Guide to PHARMACOLOGY in 2018: updates and expansion to encompass the new guide to IMMUNOPHARMACOLOGY. Nucl Acids Res 46: D1091–D1106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hashimoto K, Amano T, Kasakura A, Uhl GR, Sora I, Sakai N et al (2009). μ‐Opioid receptor‐independent fashion of the suppression of sodium currents by mu‐opioid analgesics in thalamic neurons. Neurosci Lett 453: 62–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hedegaard H, Warner M, Miniño AM (2017). Drug overdose deaths in the United States, 1999–2016. NCHS Data Brief : 1–8. [PubMed] [Google Scholar]
  17. Hoffmann PDF, Glaeske G, Windt R (2012). Verordnung von starken Opioiden im Jahr 2011. Schmerz 26: 707–714. [DOI] [PubMed] [Google Scholar]
  18. Itzoe M, Guarnieri M (2017). New developments in managing opioid addiction: impact of a subdermal buprenorphine implant. Drug Des Devel Ther 11: 1429–1437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kalso E, Vainio A (1990). Morphine and oxycodone hydrochloride in the management of cancer pain. Clin Pharmacol Ther 47: 639–646. [DOI] [PubMed] [Google Scholar]
  20. Katchman AN, McGroary KA, Kilborn MJ, Kornick CA, Manfredi PL, Woosley RL et al (2002). Influence of opioid agonists on cardiac humanEther‐a‐go‐go‐related Gene K+ currents. J Pharmacol Exp Ther 303: 688–694. [DOI] [PubMed] [Google Scholar]
  21. Millard DC, Clements M, Ross JD (2017). The CiPA microelectrode array assay with hSC‐derived cardiomyocytes: current protocol, future potential In: Stem Cell‐derived Models in Toxicology. Humana Press: New York, NY, pp. 83–107. [Google Scholar]
  22. Nieminen TH, Hagelberg NM, Saari TI, Neuvonen M, Laine K, Neuvonen PJ et al (2010). St John's wort greatly reduces the concentrations of oral oxycodone. Eur J Pain 14: 854–859. [DOI] [PubMed] [Google Scholar]
  23. Obergrussberger A, Juhasz K, Thomas U, Stölzle‐Feix S, Becker N, Dörr L et al (2016). Safety pharmacology studies using EFP and impedance. J Pharmacol Toxicol Methods 81: 223–232. [DOI] [PubMed] [Google Scholar]
  24. Paulozzi LJ, Ryan GW (2006). Opioid analgesics and rates of fatal drug poisoning in the United States. Am J Prev Med 31: 506–511. [DOI] [PubMed] [Google Scholar]
  25. Ruan Y, Liu N, Priori SG (2009). Sodium channel mutations and arrhythmias. Nat Rev Cardiol 6: 337–348. [DOI] [PubMed] [Google Scholar]
  26. Rudd RA (2016). Increases in drug and opioid‐involved overdose deaths – United States, 2010–2015. MMWR Morb Mortal Wkly Rep 65: 1445–1452. [DOI] [PubMed] [Google Scholar]
  27. Saari TI, Ihmsen H, Neuvonen PJ, Olkkola KT, Schwilden H (2012). Oxycodone clearance is markedly reduced with advancing age: a population pharmacokinetic study. BJA Br J Anaesth 108: 491–498. [DOI] [PubMed] [Google Scholar]
  28. Schulze V, Stoetzer C, O'Reilly AO, Eberhardt E, Foadi N, Ahrens J et al (2014). The opioid methadone induces a local anaesthetic‐like inhibition of the cardiac Na+ channel, Nav1.5. Br J Pharmacol 171: 427–437. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. de Vos JW, Geerlings PJ, van den Brink W, Ufkes JG, van Wilgenburg H (1995). Pharmacokinetics of methadone and its primary metabolite in 20 opiate addicts. Eur J Clin Pharmacol 48: 361–366. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1 Oxycodone does not affect persistent current or fast inactivation kinetics. (A) Oxycodone 50 μM does not induce use‐dependent block. HEK293 cells, stably expressing Nav1.5 channels, were stimulated with 30 depolarising pulses to 0 mV at a frequency of 10 Hz. Inward current measured at each pulse was normalised to the first response. Oxycodone 50 μM does not induce use‐dependent block beyond time‐dependent changes that can also be seen in vehicle recordings. The plot on the right shows the area under the curve (AUC) for each condition for statistical comparison. The following values were calculated: pre vehicle 25.5 ± 0.5 (n = 18), post vehicle 22.2 ± 0.8 (n = 19), pre oxycodone 50 μM 26.8 ± 0.6 (n = 4), post oxycodone 50 μM 23.0 ± 1.0 (n = 4). The vehicle data presented here for comparison are identical to the ones shown in Figure 3. (B) Oxycodone 300 μM does not increase persistent current in Nav1.5 channels. Persistent current was measured as described below. (C) Oxycodone 300 μM also does not change kinetics of steady‐state fast inactivation. A single exponential fit to the inactivating current component of regular IV recordings provided the time constant (τ) of inactivation.

Click here for additional data file. (417.9KB, pdf)

Articles from British Journal of Pharmacology are provided here courtesy of The British Pharmacological Society

RESOURCES