Electrophysiological evaluation of an anticancer drug gemcitabine on cardiotoxicity revealing down-regulation and modification of the activation gating properties in the human rapid delayed rectifier potassium channel

Abstract

Gemcitabine is an antineoplastic drug commonly used in the treatment of several types of cancers including pancreatic cancer and non–small cell lung cancer. Although gemcitabine-induced cardiotoxicity is widely recognized, the exact mechanism of cardiac dysfunction causing arrhythmias remains unclear. The objective of this study was to electrophysiologically evaluate the proarrhythmic cardiotoxicity of gemcitabine focusing on the human rapid delayed rectifier potassium channel, hERG channel. In heterologous hERG expressing HEK293 cells (hERG-HEK cells), hERG channel current (I_hERG) was reduced by gemcitabine when applied for 24 h but not immediately after the application. Gemcitabine modified the activation gating properties of the hERG channel toward the hyperpolarization direction, while inactivation, deactivation or reactivation gating properties were unaffected by gemcitabine. When gemcitabine was applied to hERG-HEK cells in combined with tunicamycin, an inhibitor of N-acetylglucosamine phosphotransferase, gemcitabine was unable to reduce I_hERG or shift the activation properties toward the hyperpolarization direction. While a mannosidase I inhibitor kifunensine alone reduced I_hERG and the reduction was even larger in combined with gemcitabine, kifunensine was without effect on I_hERG when hERG-HEK cells were pretreated with gemcitabine for 24 h. In addition, gemcitabine down-regulated fluorescence intensity for hERG potassium channel protein in rat neonatal cardiomyocyte, although hERG mRNA was unchanged. Our results suggest the possible mechanism of arrhythmias caused by gemcitabine revealing a down-regulation of I_hERG through the post-translational glycosylation disruption possibly at the early phase of hERG channel glycosylation in the endoplasmic reticulum that alters the electrical excitability of cells.

Introduction

Gemcitabine is an antineoplastic drug, which is commonly used particularly to treat several types of solid organ malignancies such as pancreatic cancer, non-small cell lung cancer, and breast cancer [1]. One of the drug profiles that differentiate antineoplastic pharmacological therapies from other pharmacological therapies is the frequency and severity of adverse effects at clinical application. The most common adverse effects of gemcitabine are symptoms in gastrointestinal, hepatic and hematopoietic disorders including nausea, vomiting, increased alanine aminotransferase (ALT) and aspartate aminotransferase (AST), increased alkaline phosphatase, anemia, and thrombocytopenia [2]. Furthermore, gemcitabine is also known to cause cardiovascular adverse reactions including myocardial ischemia, pericardial diseases, heart failure, and arrhythmias [2]. Among them, supraventricular tachyarrhythmias, especially atrial fibrillation often occur in cancer patients with gemcitabine therapy [3, 4]. There is increasing recognition that the association of gemcitabine with atrial fibrillation is strong and can occur even after the first dose, suggesting that gemcitabine could interact with arrhythmogenic substrates in the heart [5]. Drug interactions with cardiac ion channels are the underlying mechanisms for the occurrence of arrhythmias. Although rare, association of gemcitabine with QT interval prolongations in electrocardiogram (ECG) was also reported in several studies [6–9]. QT interval prolongation is actually not considered to be a toxic substrate, however, this electrical derangement has been observed in human subjects just prior to the onset of cardiac arrhythmias. Co-occurrence of QT interval prolongation and arrhythmias has led to a postulated link between QT and arrhythmogenic ion channel dysregulation. One of the primary causes of drug-induced QT interval prolongation is thought to be blockage of the hERG potassium channel in cardiac myocytes [9]. Of note, multiple recent studies have concluded that prolonged QT interval has an increased risk of atrial fibrillation [10–14].

In the present study, we aimed to electrophysiologically evaluate the action of gemcitabine as a potential modulator of cellular excitability focusing on the hERG channel. We examined effects of gemcitabine on the human embryonic kidney 293 (HEK293) cells stably expressing the hERG channel (hERG-HEK cells) but not on cardiomyocytes because of the following reasons. 1) Freshly isolated cardiomyocytes from adult animals are not suitable for the patch clamp study after the incubation for 24–48 h, 2) Rapid delayed rectifier potassium channel current (I_Kr) are very low in current density in neonatal rat or mouse cardiomyocytes [15], and 3) Our preliminary study revealed that immortalized cells such as H9c2 demonstrate very different electrophysiological features in comparison with cardiomyocytes (data not shown). Since these disparities usually indicate complexed multiple actions of drugs on ion channel transcription, translation, and/or trafficking to the plasma membrane, this study’s purpose was to increase understanding of the arrhythmogenic actions of gemcitabine focusing on the hERG channel at the post-transcription stage or later by use of hERG-HEK cells, exploring possible mechanism for atrial fibrillation and QT interval prolongation as a major cardiotoxicity of this drug.

Materials and methods

The experimental protocol was approved as “a non-animal study” in advance by the Ethics Review Committee for Animal Experimentation of Oita University School of Medicine.

Cell preparation

All experimental protocols were approved in advance by the Ethics Review Committee for Animal Experimentation of Oita University School of Medicine (No. C004003, No. G004006), and were carried out according to the guidelines for animal research of the Physiological Society of Japan to minimize the number of animals used, as well as their suffering.

The human embryonic kidney 293 (HEK293) cells stably expressing the hERG channel (hERG-HEK cells) are generous gifts from Professor Imaizumi, Nagoya City University, Japan. hERG-HEK cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Hyclone, Logan, UT, USA) with 15% fetal bovine serum (FBS, Gibco) at 37°C and exposed to an atmosphere of 5% CO₂. The culture medium was also supplemented with 400 μg·mL^-1 gentamycin (G418, Calbiochem, USA). hERG-HEK cells were treated by gemcitabine for 5 min (short-term effect) in the recording chamber of patch clamp or for 24–48 h (long-term effect) in the cell culture medium for detection of the drug actions on I_hERG and the hERG channel gating properties. Neonatal rat ventricular myocytes were prepared from 1- to 2-day-old Wistar rats as described previously [15]. The cardiomyocytes were maintained at 37°C under 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum for 24 h, then the medium was changed to medium containing gemcitabine (0.01 μM to 50 μM) or vehicle for 24–48 h.

Electrophysiological recording

We used a whole-cell patch clamp technique for the measurement of I_hERG and action potentials according to the previous studies [16–18]. Briefly, I_hERG was recorded by whole-cell patch clamp using an EPC-9 amplifier controlled by Pulse ver.8 software (HEKA Eletronik, Lambrecht, Germany). Patch pipettes were pulled from 75-mm plain capillary tubes (Drummond Scientific Co., Broomall, PA, USA) by Model P-97 (Sutter Instrument Co., Novato, CA, USA), and were heat-polished subsequently to achieve the pipette resistance at 2–4 MΩ when filled with the pipette solution shown below. Series resistance was compensated by at least 80% and was continually monitored throughout the experiment. All the measurements were done at room temperature (20–23°C). For the current recording, the chamber was filled with bath solution contained (mM) NaCl 135, KCl 5, MgCl₂ 1, CaCl₂ 2, HEPES 10, glucose 10 (pH of 7.4 by adjusted with NaOH). The patch electrodes were filled with pipette solution consisting of (mM) KCl 135, MgCl₂ 1, EGTA 5 and HEPES 10 (pH 7.2 by KOH). For the action potential recording, the bath solution (Tyrode’s solution) contained (mM) NaCl 140, MgCl₂ 1, KCl 5.4, HEPES 10, glucose 10 and CaCl₂ 1.8 (pH of 7.4 adjusted with NaOH), and the pipette solution contained (mML) KCl 140, MgCl₂ 2, creatine phosphate 5, HEPES 10, EGTA 0.05 and Mg-ATP 5 (pH of 7.2 adjusted with KOH). For analysis of the conductance-voltage relationship or I_hERG.tail-V relationship, tail current of I_hERG was evoked through a 4-s voltage step to -40 mV, following a 4-s depolarization step with a 10 mV stepwise increase from -80 to 60 mV, which was initiated after a holding potential of -80 mV (Fig 1). The peak of the outward tail current of I_hERG were plotted against the voltage of test pulses, and the data were fitted to the Boltzmann function, as: f(V) = 1/(1 + e[(V_1/2−V)/k]) + C, where the V_1/2 value is the membrane potential when the relationship reaches half level, and the k value is the slope factor. C is a constant component. For analysis of the activation gating properties, we used an envelope of tails procedure, which takes advantage of the fact that although at potentials positive to approximately +30 mV a small and brief transient component of inactivation overlaps with activation [19], recovery from inactivation is rapid [20] and deactivation is relatively slow [21]. The activation kinetics were evaluated by plotting the activating inward tail current envelope as a function of test pulse duration by use of a single exponential equation fits, which provided the time constant (τ_act) for activation at each individual potential, based on the simple Hodgkin and Huxley model [22] (Scheme 1). Namely, hERG-HEK cells were evoked through a depolarization step, from -10 mV to +30 mV, for an incremental duration for 20ms-4,000 ms followed by a holding potential of -100 mV. At each depolarizing potential, τ_act was then plotted against the depolarizing voltage, yielding τ_act-voltage relationship (Fig 3), which were further best fitted to a three-parameter single exponential function: T(x) = T₀ + A x (exp(-x/τ)). For this equations T(x) is the time at voltage x, A is the amplitudes of the exponential, and τ is the voltage constant. For analysis of the inactivation gating properties, the inactivating kinetics were evaluated by the time constant of the tail current; hERG-HEK cells were evoked through a 2.5-s depolarizing steps (test potentials) from -60 mV to +60 mV to inactivate, following a 5-ms repolarization at -120 mV from the holding potential of +60 mV (Scheme 1). At each test potential, onset of each inactivation current or the decay of tail current during each voltage step was fitted by use of a single exponential curve, which provided the time constant for inactivation (τ_inact) at the individual potentials, then τ_inact was plotted against the voltage, yielding τ_inact-voltage relationship (Fig 4). For analysis of the deactivation gating properties, the deactivating kinetics were evaluated by fitting the deactivating inward tail current at each repolarizing voltage step (-180 mV to -110 mV) from the holding potential of +60 mV to a single exponential curve (Scheme 1), providing the time constant for deactivation (τ_deact) at the individual potentials, then τ_deact was plotted against the voltage, yielding τ_deact-voltage relationship (Fig 5). For analysis of the reactivation gating properties namely recovery from the inactivation gating properties, the reactivating kinetics were evaluated by the amplitude of peak tail current corresponding to the hyperpolarizing pulse from -200 mV to +60 mV for 5 ms to allow inactivation to recovery to a steady-state (Scheme 1), followed by a return step to +60 mV (Fig 6).

[Figure omitted. See PDF.]

(A) Representative I_hERG family traces in the control and during the acute (5 min) application of gemcitabine (Gem) (50 μM) are shown. Pulse protocol is shown above. (B) The relationship between terminal I_hERG (I_hERG.term) measured at the end of each test pulse indicated by hollow arrowheads in panel A (▽) and the test pulse potential (V) constructed by using group data are shown as I_hERG.term-V relationship in control (〇) and during the acute application of 50 μM gemcitabine in 5 min (■). The maximum current amplitude of I_hERG.term at +20 mV was945 ± 31 pA in control (〇), and was 972 ± 38 pA with 50 μM gemcitabine (■) in 5 min. (C) The relationship between tail I_hERG (I_hERG.tail) measured at the beginning of the third pulse at -40 mV indicated by filled arrowheads in panel A (▼) and test pulse potential (V) constructed by using group data are shown as I_hERG.tail-V relationship in control (〇) and during the acute application of 50 μM gemcitabine in 5 min (■). (D) Representative I_hERG family traces of the control and long-term (24 h) application of gemcitabine (50 μM) are shown. Pulse protocol is shown above. (E) The I_hERG.term-V relationship constructed by using group data in control and long-term application of 50 μM gemcitabine for 24 h. The maximum current amplitude of I_hERG.terml was significantly reduced by 50% from 1072 ± 34 pA (vehicle) to 540 ± 25 pA (gemcitabine), where the potential for the maximum current was shifted to the hyperpolarized potentials from +20 mV (vehicle) to +10 mV (gemcitabine). (F) The I_hERG.tail-V relationship constructed by using group data in control and long-term application of 50 μM gemcitabine for 24 h. The maximum current amplitude of I_hERG.tail at 60 mV was significantly reduced by 51% from 1709 ± 73 pA (vehicle) to 846 ± 17 pA (gemcitabine). Red line represents Boltzman-fitting curve of gemcitabine data normalized to the I_hERG.tail in vehicle at +60 mV. Insets in B, C, E anf F show I_hERG density at the potentials indicated by each panel graph ordinate. Numbers of cell are shown in parentheses. Values represent the mean ± SE. Unpaired Student’s t test was performed.

Scheme 1.

The peak of the tail currents plotted as a function of the preceding test pulse potentials were fitted with a Boltzmann function, namely the steady-state inactivation curve, was assessed as the reactivation kinetics. At more negative voltages from -170 mV to -200 mV, a proportion of the hERG channels also deactivate during the 5 ms test potential, leading the observed decrease in the peak I_hERG at negative potentials. Therefore, before fitting of the Boltzmann equation, this deactivation was corrected based on the pulse protocol in Fig 5. For analysis of the concentration dependency of gemcitabine on I_hERG, we used six concentrations of gemcitabine (0.01, 0.05, 0.1, 1, 5, and 50 μM) to construct the half-maximal inhibitory concentration (IC₅₀) curve, which was sigmoidal in shape plotted against the log [gemcitabine] fitted with the Hill equation.

Gene expression analyses

The cDNA was synthesized from 1 μg of total RNA using the Transcriptor First Strand cDNA Synthesis Kit (Roche Molecular System Inc, Alameda, CA). Real-time PCR was performed on a LightCycler 96 system (Roche) using THUNDERBIRD Next SYBR qPCR (Toyobo, Osaka, Japan) as a detection reagent. Forward and reverse primer sequences respectively for the rat KCNH2 channel were designed from the sequence in the GenBank database as follows (accession numbers are indicated in parentheses): Kcnh2 (117018), forward 5’-TCAACCTGCGAGATACCAACATG-3’, reverse 5’-CTGGCTGCTCCGTGTCCTT-3’. Data were calculated by 2^-ΔΔCT and are presented as fold change induced in the transcript of each myocyte gene assayed by gemcitabine exposure. Gene expression was normalized to that of GAPDH and compared with the control condition (vehicle defined as 1.0).

Western blotting

Western blotting was performed as described previously [23]. After 48 hours for gemcitabine and/or tunicamycin treatment, cardiomyocytes were lysed on ice 60 min in SDS-modified RIPA buffer containing protease inhibitor cocktails (Cytiva, Marlborough, MA, USA), and then centrifuged at 4 °C 15,000 rpm for 20 min. The resulting cell lysates (30 μg each) were boiled with SDS sample buffer and subjected to SDS-PAGE. The samples were transferred to a polyvinylidene difluoride membrane (Merk Millipore, Darmstadt, Germany), which was blocked for 1 h in Block Ace (DS Pharma, Osaka, Japan) at room temperature, then incubated overnight at 4 °C with primary antibodies against hERG (mouse monoclonal antibody, 1:100, sc-515611, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and GAPDH (rabbit IgG-polyclonal antibody, 1:1000, 10494-1-AP, Proteintech, Rosemont, IL, USA). After washing with 1 × TBS containing 0.1% Tween 20, the membranes were incubated for 1 h at RT with appropriate secondary antibodies, followed by rewashing. Finally, the signals were detected using an ECL Western blotting analysis system (GE Healthcare, Piscataway, NJ, USA) in accordance with the manufacturer’s instructions. For detection of hERG and GAPDH, we performed two or three independent Western blots, and representative data are shown. The band intensities were quantified using Image J software (National Institute of Health, USA).

Immunocytochemistry

The details of the experiments protocol were performed as described previously [15]. Primary antibodies against hERG (1:50, Santa Cruz) and α-actinin (1:200, Cell Signaling, Beverly, MA, USA) were applied following by incubation with the appropriate fluorescence-labeled secondary antibodies (Invitrogen) for 1 h at room temperature. After several washes, the samples were air-dried, mounted with a drop of ProLong Diamond Antifade Mountant with DAPI (Molecular Probes) and subjected to microscopy. Images were acquired using a 100× oil objective (Plan-Apochromat 100× [numerical aperture, 1.46] oil immersion objective for differential interference contrast [DIC]; Carl Zeiss). All sections were analyzed using a confocal laser microscopy system and software (LSM710; Carl Zeiss) that was built around an inverted microscope (Axio Observer Z1; Carl Zeiss, Germany). Images were saved in TIFF format and analyzed by ImageJ software (Wayne Rasband, National Institutes of Health, USA).

Chemicals

All chemicals were purchased from Wako Pure Chemical Industries (Osaka, Japan). Gemcitabine was dissolved in dimethyl sulfoxide (DMSO) as stock solutions (50 mM), and then diluted to final concentrations in cell culture solutions. The final concentration of DMSO or acetic acid in the bathing solution was 0.01% or less.

Statistical analysis

All data are presented as mean ± S.E. Statistical analysis was performed by one-way ANOVA with Tukey-Kramer test, Dunnett’s post hoc test when comparing every experimental group mean to the control group mean, or non-parametric Steel-Dwass test when data was not normal. To assess the differences in two groups, unpaired Student’s t test was performed. EC₅₀ values were estimated by using non-linear least square curve-fitting programs in Sigma plot software ver.10 (SPSS, Chicago, IL, USA). Differences were considered significant when p values were less than 0.05 if nothing else is mentioned.

Results

Acute- and long-term effect t of gemcitabine on I_hERG

We first examined the possible acute effect of gemcitabine on I_hERG. hERG-HEK cells were superfused with normal Tyrode’s solution and the membrane potential was held at -80 mV followed by test pulses with a duration of 4 s, and then voltage-clamped at the potentials of -40 mV for 4 s (Fig 1, inset). Fig 1A illustrates typical acute effects of gemcitabine (50 μM) for 5 min on I_hERG family, demonstrating a negligible action of gemcitabine on I_hERG. Effect of gemcitabine on the terminal I_hERG at the end of the test pulse of 4 s (indicated by hollow arrowheads) were plotted against the membrane potentials imposed (I_hERG.term-V relationship), and summarized in Fig 1B. Effect of gemcitabine on the maximum tail I_hERG at the beginning of the pulse at -40 mV (indicated by filled arrowheads) following the test potentials were plotted against the membrane potentials imposed (I_hERG.tail-V relationship), and summarized in Fig 1C. The amplitude of the maximum outward I_hERG.term, the bell-shaped I_hERG.term-V relationship or the maximum I_hERG.tail was unchanged by gemcitabine. According to these results, it is concluded that gemcitabine has substantially no actions on I_hERG as assessed at 5 min after application.

We then assessed the action of gemcitabine on I_hERG as applied in the culture medium for hERG-HEK cells for 24 h. Fig 1D indicates representative samples of I_hERG family recorded with or without actions of gemcitabine (50 μM) for 24 h, demonstrating a long-term inhibitory effect of gemcitabine on I_hERG. Note that the solution was changed to gemcitabine-free bath solution prior to the patch-clamp study for 1 h. The I_hERG.term-V relationship revealed a long-term reduction of I_hERG by gemcitabine for 24 h; terminal I_hERG was reduced by 53% (50 μM) as assessed by the maximum outward current at +10 mV or +20 mV (Fig 1E). The I_hERG.tail-V relationship revealed a long-term reduction of I_hERG as well by gemcitabine for 24 h; the tail I_hERG was reduced by 52% (50 μM) as assessed by the maximum outward current at +60 mV (Fig 1F). Interestingly, the half-activation potential (V_1/2.act) was significantly changed by gemcitabine treatment; V_1/2.act was 0.1 ± 0.8 mV in vehicle and -5.4 ± 1.4 mV by gemcitabine (50 μM) treatment, which suggests that gemcitabine not only suppresses I_hERG but also modifies the gating properties of the channel.

Concentration-dependent suppression of I_hERG by gemcitabine

Dose-dependent suppression of I_hERG by gemcitabine was constructed by measuring percentage suppression of the maximum tail current of I_hERG with various concentration of gemcitabine (Fig 2). As indicated by representative current families of I_hERG (Fig 2A), the hERG-HEK cells display lower I_hERG amplitude when gemcitabine concentration was increased. The I_hERG.tail-V relationship revealed a clear dose-dependency of gemcitabine actions on inhibition of I_hERG (Fig 2B). Importantly, V_1/2.act was also changed in a concentration-dependent manner; V_1/2.act was 0.1 ± 0.9 mV in vehicle, -2.6 ± 1.7 mV by 0.05 μM gemcitabine, -5.2 ± 1.8 mV by 5 μM gemcitabine and -5.4 ± 1.4 mV by 50 μM gemcitabine. Dose-response curve of I_hERG.tail suppression by gemcitabine indicates that the median effect of concentration (EC₅₀) of gemcitabine to suppress I_hERG was 0.16 μM (Fig 2E).

[Figure omitted. See PDF.]

(A) Representative I_hERG family traces in vehicle, 0.05 μM, 5 μM and 50 μM gemcitabine (Gem) applied for 24 h. Pulse protocol was identical to that shown in Fig 1A and 1D. (B) The I_hERG.tail-V relationship constructed by using group data in vehicle (〇) and long-term application of 0.05 μM (▲), 5 μM (■) and 50 μM gemcitabine (●) for 24 h. I_hERG.tail was assessed by the amplitude of the tail current of I_hERG indicated by the filled arrowheads in panel (A). Red line (Gem 0.05 μM), green line (Gem 5 μM) and blue line (Gem 50 μM) represent Boltzman-fitting curve of each gemcitabine data normalized to the I_hERG.tail in vehicle at +60 mV. (C) Representative I_hERG family traces in vehicle and 200 μM gemcitabine applied for 2 h. In this application protocol, hERG-HEK cells were incubated with 200 μM gemcitabine for 2 h, followed by incubation with gemcitabine-free culture medium for 22 h. (D) The I_hERG.tail-V relationship constructed by group data in vehicle (〇) and 200 μM gemcitabine (■) applied by the application method described in panel (C). The maximum I_hERG.tail was reduced by 38% from 1728 ± 96 pA (vehicle) to 1079 ± 84 pA (gemcitabine). (E) Fractional changes in I_hERG measured at +60 mV in I_hERG.tail were plotted against the concentration of gemcitabine applied for 24 h (〇) and 2 h (■) based on the method mimicking clinical application protocol as described in panel (C). Numbers of cell are shown in parentheses. Values represent the mean ± SE. Statititical significance was determined by ANOVA with Dunnett test (Panel B) and unpaired Student’s t test (Panel D).

According to the standard clinical methods for gemcitabine application protocol by which patient is being treated, 1,000–1,250 mg/m² of gemcitabine is to be injected intravenously for 20–30 min once a week [24]. By this application protocol, the maximum serum concentration of gemcitabine reaches approximately 200 μM immediately after the injection [25], which follows the concentration declining period for nearly 2 h. Thus, we mimicked the application protocol of gemcitabine in this in vitro experiment by culturing hERG-HEK cells with 200 μM gemcitabine for 2 h, followed by incubation with gemcitabine-free medium for 22 h prior to the patch-clamp study. With this experimental protocol, I_hERG was suppressed nearly by 29% in the representative I_hERG family (Fig 2C), and by 38% in the group data (Fig 2D). The result of the normalized I_hERG with this application mimicking clinical protocol is compatibly represented in Fig 2E (filled square).

Actions of gemcitabine on I_hERG activation

Long-term actions of gemcitabine on I_hERG activation were examined using the envelope of the tail test in hERG-HEK cells treated with vehicle or 0.05 μM gemcitabine for 24 h (Fig 3). Fig 3A indicates representative I_hERG in vehicle and in gemcitabine corresponding to the pulse protocol shown above. Then the normalized amplitude of the tail I_hERG was plotted against the pulse duration, yielding the time course of activation process at the test potentials (Fig 3B). In Fig 3B inset, the time course of I_hERG activation at the test potential of +10 mV was fitted to a single exponential function in vehicle and gemcitabine. Activation time constant (τ_act) at the potentials of +10 was 0.66 ± 0.07 s in vehicle and 0.34 ± 0.05 s in gemcitabine. As demonstrated in τ_act-V relationship in Fig 3B, activation of the hERG channel was shifted to the direction of hyperpolarization; the shift was 12.5 mV as assessed at the activation time constant (τ_act) value of 1.0 s.

[Figure omitted. See PDF.]

(A) Representative I_hERG traces elicited by pulse protocol shown above, which displays time course of I_hERG activation measures by envelope of tail test in vehicle and in 0.05 μM gemcitabine (Gem) for 24 h. (B) Activation time constant (τ_act)-V relationship in vehicle (〇) and gemcitabine (■), where τ_act at each test potential was obtained by a single exponential fitting curve in a plot of pulse duration (20–4,000 ms) versus I_hERG.tail. An inset shows a fitting curve of the activating I_hERG at the potentials of +10 mV, yielding τ_act of 0.66 s for vehicle and of 0.34 s for gemcitabine. Data of τ_act were further fitted by a single exponential curve (see Methods Section), which yields I_hERG activation voltage constant of 21.3 mV for vehicle (〇) and 20.2 mV for gemcitabine (■). Numbers of cell are shown in parentheses. Values represent the mean ± SE. Unpaired Student’s t test was performed. *p<0.05 vs. Vehicle.

Actions of gemcitabine on I_hERG inactivation

Long-term actions of gemcitabine on I_hERG inactivation were assessed by applying a conditioning pulse to +60 mV for 1s followed by a brief hyperpolarizing pulse (-120 mV for 5 ms) to allow the hERG channel to recover from inactivation (Fig 4). Depolarizing test pulses were applied to record inactivating I_hERG. Fig 4A indicates representative I_hERG in vehicle and gemcitabine corresponding to the pulse protocol shown above, where the time course of fast inactivating I_hERG was fitted to a single exponential function, yielding a plot demonstrating voltage dependency of the fast time constant of I_hERG inactivation (τ_inact) (Fig 4B). Gemcitabine substantially unaffected τ_inact at all examined potentials, suggesting that long-term modification of the hERG channel by gemcitabine could not regulate gating properties of the channel from the open state to the inactivation state.

[Figure omitted. See PDF.]

(A) Representative I_hERG traces elicited by pulse protocol shown above, which displays voltage-dependent decay or inactivation of I_hERG.tail in vehicle and 0.05 μM gemcitabine (Gem) for 24 h. (B) Relationship between time constant of the decay phase of I_hERG.tail or τ_inact and the test pulse potential in vehicle (〇) and gemcitabine (■). Numbers of cell are shown in parentheses. Values represent the mean ± SE.

Actions of gemcitabine on I_hERG deactivation

Deactivation process of the hERG channel was analyzed by applying long hyperpolarizing test pulses after a conditioning pulse to +60 mV (Fig 5A, inset) to assess the long-term actions of gemcitabine. Fig 5A demonstrates representative I_hERG in vehicle and gemcitabine corresponding to the pulse protocol shown above, where the decay phase of instantaneous I_hERG at the beginning of the test pulse could be fitted to a double exponential function. Although the fast phase could be ascribed to the mixture of both inactivation and the deactivation steps based on this pulse protocol, inactivating I_hERG with an extremely fast gating process may has little impact on the decay of I_hERG [9]. Because of this reason, the fast component of the time constant was compared as a parameter of channel deactivation (τ_deact) between vehicle and gemcitabine treatment. Gemcitabine treatment of hERG-HEK cells did not affect τ_deact of I_hERG suggesting that gemcitabine has no action on deactivation of the hERG channel as a long-term effect. Namely, long-term modification of the hERG channel by gemcitabine could not regulate gating properties of the channel from the open state to the closed state. Analysis for the fast decay phase amplitude (A_fast) also supports the conclusion (Fig 5B, inset).

[Figure omitted. See PDF.]

(A) Representative I_hERG traces elicited by pulse protocol shown above, which displays deactivating phase of I_hERG in vehicle and 0.05 μM gemcitabine (Gem) for 24 h. (B) Relationship between time constant of the decay phase of I_hERG or τ_deact and the test pulse potential in vehicle (〇) and gemcitabine (■). An inset shows a ration of the fast component of the decay amplitude (A_fast) to whole decay amplitude (A_fast + A_slow). Although the decaying phase was fitted by a sum of two exponential components, τ for the fast component or the main component of the decay phase was assessed as τ_deact for the deactivating gating properties. Numbers of cell are shown in parentheses. Values represent the mean ± SE.

Actions of gemcitabine on I_hERG reactivation

Comparison of the process at the recovery from inactivation or reactivation of the hERG channel in vehicle and by gemcitabine treatment was performed by use of the standard three-stage voltage protocol. A 2-s pulse to +60 mV was first applied to attain steady-state inactivation, which was followed by a series of test pulses ranging from -200 mV to +60 mV for 5 ms (Fig 6A inset). This pulse duration is enough to allow recovery of the channel from inactivation, but too short to allow deactivation. Fig 6A demonstrates representative I_hERG in vehicle and gemcitabine treatment corresponding to the pulse protocol shown above, where the peak of the third pulse current was plotted against the potentials of the second pulses, yielding reactivation I-V curves (Fig 6B). At more negative voltages than -170 mV, a proportion of the channels also deactivate during the second pulse, leading to the observed decrease in the peak currents. Before fitting of the Boltzmann equation, this deactivating current was corrected based on the separate protocol such as that shown in Fig 5A to obtain the rates of deactivation at each voltage in the second pulse. With this analysis, voltage dependence of the recovery from steady-state-inactivation or reactivation of I_hERG in vehicle and gemcitabine was nearly identical (Fig 6B), which suggests that modification of the hERG channel by gemcitabine does not affect the reactivation process of the channel. Namely, long-term modification of the hERG channel by gemcitabine could not regulate gating properties of the channel from the inactivated state to the open state.

[Figure omitted. See PDF.]

(A) Representative I_hERG traces elicited by pulse protocol shown above, which displays reactivating I_hERG in response to the conditioning potentials (-200 mV—+60 mV) in vehicle and 0.05 μM gemcitabine (Gem) for 24 h. (B) Relationship between amplitude of reactivating I_hERG.tail and the conditioning pulse potentials in vehicle (〇) and gemcitabine (■). Gray filled circles and gray filled squares represent corrected data set after the replacement of I_hERG deactivating component. Each solid curve was a product of a fit by Boltzmann equation, and dotted line was drawn by Boltzmann fitting curve for vehicle data multiplied by 0.45 that representing the ratio (gemcitabine-I_hERG / vehicle-I_hERG) at -200 mV. Numbers of cell are shown in parentheses. Values represent the mean ± SE.

Actions of gemcitabine under the inhibition of glycosylation

To explore the possible mechanism of gemcitabine on the hERG channel regulation, we examined the long-term actions of gemcitabine under the influence of asparagine (N)-linked glycosylation inhibition, because altered N-glycosylation levels of circulating glycoproteins were reported in cancer patients with gemcitabine treatment [26]. Also because N-glycosylation reportedly stabilizes hERG channel proteins on the plasma membrane [27], we recorded I_hERG in hERG-HEK cells treated with gemcitabine in the presence or absence of the inhibitory signals for N-linked glycosylation. Tunicamycin is widely used as a research tool to block N-linked glycosylation by blocking the transfer of N-acetylglucosamine 1-phosphate to dolichol monophosphate. Fig 7A demonstrates representative I_hERG in vehicle, in gemcitabine alone (24 h), in tunicamycin alone (48 h), and in gemcitabine (24 h) combined with tunicamycin treatment (48 h) in the culture medium; tunicamycin was applied to the culture medium 24 h prior to and during the gemcitabine application. Of note, gemcitabine was without effect on I_hERG when I_hERG was highly suppressed by the co-presence of tunicamycin. When terminal I_hERG at the end of each test pulse were plotted against the voltage applied, a bell-shaped relationship was maintained in treatment of gemcitabine, tunicamycin, and gemcitabine combined with tunicamycin, although the potential for the maximum outward current was shifted by the treatment of gemcitabine (Fig 7B), which is consistent with the result in Fig 1E. It is also notable that the differences of the shape of terminal I_hERG-V relationship (Fig 7B) and the tail I_hERG -V relationship (Fig 7C) between tunicamycin and gemcitabine combined with tunicamycin was negligibly small, which suggests that actions of gemcitabine on the hERG channel and actions for the N-glycosylation disruption overlaps significantly. We also noted that gemcitabine treatment significantly shifted the normalized tail I_hERG toward the hyperpolarized direction by 6.4 mV, while gemcitabine was unable to shift the tail I_hERG when hERG-HEK cells were simultaneously treated with tunicamycin. These results suggest that action site of gemcitabine and tunicamycin on the hERG channel may not be identical.

[Figure omitted. See PDF.]

(A) Representative I_hERG traces in vehicle (48 h), 0.1 μM gemcitabine (Gem) (24 h), 0.5 μg/ml tunicamycin (Tuni) (48 h), and 0.1 μM gemcitabine (24 h) with 0.5 μg/ml tunicamycin (48 h). Tunicamycin was applied to the culture medium 24 h prior to and during the application of gemcitabine for 24 h. (B) The relationship between I_hERG.term measured at the end of the test pulse (indicated by hollow arrowheads in panel A, ▽) and test pulse potential (V) constructed by using group data in vehicle (〇), 0.1 μM gemcitabine (■), 0.5 μg/ml tunicamycin (△), and gemcitabine with tunicamycin (▲). (C) The relationship between I_hERG.tail measured at the beginning of the third pulse at -40 mV (indicated by filled arrowheads in panel A, ▼) and test pulse potential (V) constructed by using group data are shown as I_hERG.tail-V relationship in vehicle (〇), 0.1 μM gemcitabine (■), 0.5 μg/ml tunicamycin (△), and gemcitabine with tunicamycin (▲). (D) Normalized I_hERG.tail–V relationship based on data in panel C for vehicle (〇), 0.1 μM gemcitabine (■), 0.5 μg/ml tunicamycin (△), and gemcitabine with tunicamycin (▲). V_1/2 was 0.1 ± 0.1 mV for vehicle, 6.3 ± 2.0 mV for gemcitabine, 1.2 ± 2.2 mV for tunicamycin and 0.8 ± 1.2 mV for tunicamycin with gemcitabine. Cell culture protocol is shown in inset. Numbers of cell are shown in parentheses. Values represent the mean ± SE. Statititical significance was determined by ANOVA with Tukey-Kramer test.

To further characterize the effect of gemcitabine on the hERG channel glycosylation disruption, a specific class I mannosidase inhibitor kifunensine [28] was applied in combined with gemcitabine. Fig 8A demonstrates representative I_hERG in vehicle (48 h), in kifunensine alone (48 h), and in kifunensine (48 h) combined with gemcitabine treatment (24 h). Importantly, gemcitabine further reduced I_hERG when combined with kifunensine. The tail I_hERG-V relationship revealed that effect of gemcitabine on I_hERG was additive to the effect of kifunensine because the inhibition ratio of I_hERG at the voltage of +60 mV by gemcitabine was nearly identical between the conditions with (-35.8%) or without the presence of kifunensine (-30.5%). On the contrary, kifunensine was unable to reduce I_hERG when hERG-HEK cells were pretreated with gemcitabine for 24 h. These results suggest that I_hERG reduction caused by gemcitabine revealing a down-regulation of I_hERG through the post-translational glycosylation disruption possibly at the early phase of hERG channel glycosylation prior to mannosidase I-dependent process in the endoplasmic reticulum.

[Figure omitted. See PDF.]

(A) Representative I_hERG traces in vehicle (48 h), 100 μM kifunensine (Kif) (48 h), and 100 μM kifunensine (48 h) with 0.1 μM gemcitabine (24 h). Kifunensine was applied to the culture medium 24 h prior to and during the application of gemcitabine for 24 h. (B) The relationship between I_hERG.tail measured at the beginning of the third pulse at -40 mV (indicated by filled arrowheads in panel A, ▼) and test pulse potential (V) constructed by using group data are shown as I_hERG.tail-V relationship in vehicle (〇), 100 μM kifunensine (◇), 100 μM kifunensine with 0.1 μM gemcitabine (◆), and 0.1 μM gemcitabine (■). (C) Representative I_hERG traces in vehicle (48 h), 0.1 μM gemcitabine (48 h), and 0.1 μM gemcitabine (48 h) with 100 μM kifunensine (24 h). (D) The relationship between I_hERG.tail measured at the beginning of the third pulse at -40 mV (indicated by filled arrowheads in panel A, ▼) and test pulse potential (V) constructed by using group data are shown as I_hERG.tail-V relationship in vehicle (〇), 0.1 μM gemcitabine (▽), 0.1 μM gemcitabine with 100 μM kifunensine (▼), and 100 μM kifunensine (◇). Gemcitabine was applied to the culture medium 24 h prior to and during the application of kifunensine for 24 h. Cell culture protocol is shown in inset. Numbers of cell are shown in parentheses. Values represent the mean ± SE. Statistical significance was determined by ANOVA with Tukey-Kramer test.

Actions of gemcitabine on cardiomyocytes

Electrophysiological long-term actions of gemcitabine on neonatal cardiomyocytes were examined by voltage-clamp and current-clamp methods. Gemcitabine treatment for 24 h down-regulated E-4031-sensitive rapidly activating delayed rectifier K⁺ channel current (I_Kr) in neonatal cardiomyocytes in a similar manner in hERG-HEK cells; gemcitabine appreciably reduced the tail current of I_Kr approximately by 30% when assessed by the pulse protocols at +20 mV and +60 mV (Fig 9A and 9B). Also note that the current density of I_Kr was as small as ~1 pA/pF in neonatal cardiomyocytes, which made it nearly unfeasible to assess the gating properties of I_Kr in detail. Meanwhile, the action potential configuration was changed only a little by gemcitabine treatment (Fig 9C–9F), because current density of I_Kr was very small in neonatal cardiomyocyte.

[Figure omitted. See PDF.]

(A, B) Representative I_Kr tail current traces and group data of the control (vehicle) and long-term (24 h) application of gemcitabine (0.1 μM), corresponding to the pulse protocol shown above. Isolation I_Kr was obtained by subtraction of the currents recorded in the presence of E-4031 (10 μM) from the control currents. The maximum outward tail currents of I_Kr (indicated by filled triangles) were assessed as group data. (C) Representative action potentials in vehicle (24 h) and gemcitabine (24 h) of 0.1 μM. (D, E, F) Group data for cardiomyocyte spontaneous beating rate (D), action potential duration at 90% repolarization (APD₉₀) (E), and the maximum diastolic potentials (MDP) (F). The Student’s t test was used to compare the means between two groups. Numbers of cell are shown in parentheses. Values represent the mean ± SE. Unpaired Student’s t test was performed. *p1 = 0.154, and *p2 = 0.260 vs. Vehicle.

In order to further support our conclusion that gemcitabine down-regulates hERG-I_Kr via inhibition of N-glycosylation pathway at N-acetylglucosamine (GlcNAc) phosphotransferase and/or the step(s) between the tunicamycin-sensitive and kifunensine-sensitive site, we additionally examined actions of gemcitabine on hERG mRNA and proteins in neonatal rat cardiomyocytes. It is of importance that gemcitabine substantially exerted no effect on hERG mRNA when applied for 48 h in cardiomyocytes (Fig 10A), suggesting that gemcitabine has no effect on the transcription of hERG. We then performed Western blot analyses to determine the impact of gemcitabine and tunicamycin on hERG protein expression in cardiomyocytes (Fig 10B). In the whole cell lysate of cardiomyocytes treated by tunicamycin, mature form of hERG (127 kDa band), which possibly carries full N-lined sugars, was significantly reduced in comparison with cardiomyocytes treated by vehicle (Fig 10C), whereas the possible immature forms of hERG proteins, indicated by bands near 70 kDa and 90 kDa, were increased by tunicamycin (Fig 10D and 10E). Interestingly, gemcitabine treatment upregulated possible immature forms of hERG protein indicated by bands at 70 kDa and 90 kDa, whereas it unaffected the mature form of hERG protein (127 kDa). Furthermore, immunocytochemistry staining revealed that gemcitabine and tunicamycin suppressed hERG protein expression nearly at the cell surface to a similar extent (Fig 10F and 10G), indicating that gemcitabine down-regulates hERG-I_Kr through post-translational pathway, likely to be involved in N-glycosylation step(s).

[Figure omitted. See PDF.]

(A) Changes of hERG mRNA expressions in vehicle and gemcitabine (0.01–50 μM) applied for 48 h were examined by qRNA. (B) Effects of 0.1 μM gemcitabine (48 h) on hERG protein expression evaluated by Western blot analysis. hERG protein levels of mature form of hERG protein (indicated by an arrow near 127 kDa) and immature/degradative form of hERG proteins (indicated by arrows near 90 kDa and 70 kDa). See the S1 File for the complete gel/blot images. (C, D, E) Normalization of Western blot data for hERG at 127 kDa (C), hERG at 70 kDa (D), and hERG at 90 kDa (E). Non-parametric Steel-Dwass test was applied. *P<0.1. (F) Expression and distribution of hERG as assessed by immunocytochemistry procedure. Cardiomyocytes were exposed to vehicle, gemcitabine (0.1 μM) or tunicamycin (0.5 μg/ml) for 24 h. hERG for stained in red, α-actinin in green, and DAPI staining to visualize nuclei in blue. Scale bar = 20 μm. (G) Quantitative results are shown. The fluorescence intensity of hERG protein with vehicle was set to 1.0. The result for each condition was obtained from different image slides indicated by parentheses. Statistical significance was determined by non-parametric Steel-Dwass test in Panels C, D, and E, whereas ANOVA with Tukey-Kramer test in Panel G. #p < 0.1 versus vehicle. *p < 0.01 versus vehicle. Values represent the mean ± SE.

Discussion

In the present study, we found that an anti-cancer drug gemcitabine suppressed I_hERG in heterologous hERG-HEK cells, as well as I_Kr in neonatal cardiomyocytes, as a long-term effect (24 h). The EC₅₀ of gemcitabine to suppress I_hERG was 0.16 μM, which is within the serum concentration of gemcitabine after intravenous bolus injection based on the standard clinical application protocol [25]. As a long-term effect, gemcitabine modified the voltage dependency of the hERG channel activation, while inactivation, deactivation and reactivation gating properties were unaffected by gemcitabine. Gemcitabine treatment-related arrhythmias associated with QT prolongation or risk of torsade de pointes are possibly caused, at least in part, by the N-glycosylation disruption of the hERG channel.

Gemcitabine-mediated arrhythmias

I_Kr or I_hERG plays a critical role in shaping action potentials in cardiomyocytes, especially the plateau phase and stabilizing action potential durations (APD). Reduction in I_hERG prolongs APD and effective refractory period (ERP), which may result in the destabilization and early termination of reentrant-based arrhythmias [22, 29]. However, cardiac arrhythmias were often reported in 10% of patients or more treated with gemcitabine [4]. Gemcitabine is actually known to cause atrial fibrillation [3, 4]. Indeed, high atrial fibrillation rate caused by gemcitabine was recorded, which may be associated with even the first dose [4]. These observations leave the focus of our attention on the question of how, if gemcitabine prolongs APD, it can also be involved in atrial mechanical function. To put the same question differently: If gemcitabine prolongs atrial contraction time, then how does the mechanism ensure that atrial fibrillation occurs in the case of gemcitabine application? The answer will depend on how atrial contraction regulates atrial/pulmonary vein pressure in human subject. According to human echocardiographic studies, in healthy individuals, left atrial contraction produces forward flow across the mitral valve, as well as a reverse flow in the pulmonary veins [30, 31]. Left ventricular cavity pressure and compliance during atrial contraction affect forward and reverse atrial flow [31, 32]. It is accordingly postulated that the reverse flow in the pulmonary veins will become more prominent when left atrial pressure is increased during atrial contraction. Because left atrial cardiomyocytes have large I_Kr density [33], suppression of I_Kr may participate in the prolongation of contraction time in the atrial tissue, which may cause the increase of atrial pressure and reverse flow in the pulmonary veins. Taken together, a transient suppression of I_hERG by gemcitabine could be a base for the mechanism of atrial fibrillation.

Although the incidence is not high, association of gemcitabine with QT interval prolongations were reported in several studies [6, 7, 34]. The timing of the QT interval at the organism level is related to the APD of ventricular cardiomyocytes acting at the cellular level [35]. Standing on these theoretical viewpoints, we speculate that gemcitabine decreases I_hERG in a long-term application to prolong APD, which could lead to QT prolongation and the risk of torsades de pointes-type arrhythmias. Therefore, caution is highly warranted to avoid conditions that can further lead to QT prolongation such as in hypokalemia, hypothermia, bradycardia, and class III antiarrhythmic drugs etc. in combined with gemcitabine treatment.

Gemcitabine as a glycosylation modulator

Glycosylation of the hERG channels is known to N-linked at position N598 [36]. According to literature, glycosylation of the hERG channel is believed to increase hERG current and stabilize the channel protein on the plasma membrane by decreasing proteolytic susceptibility [37]. In this study, we inhibited N-linked glycosylation of the hERG channel in hERG-HEK cells with tunicamycin (10 mg/ml for 24 h) or kifunensine (100 μM for 24–48 h) to confirm the action of glycosylation on the hERG channel in combined with gemcitabine to explore the mechanism of the action. Because gemcitabine could not affect I_hERG when I_hERG was highly suppressed by tunicamycin (Fig 7C), it is plausible to conclude that gemcitabine acts to suppress a pathways of the channel glycosylation to reduce I_hERG. More specifically, it is suggested that gemcitabine may act on the channel glycosylation step at the tunicamycin-sensitive GlcNAc phosphotransferase and/or the site between the tunicamycin-sensitive and kifunensine-sensitive process, because a mannosidase I inhibitor kifunensine was without effect on I_hERG when hERG-HEK cells were pretreated with gemcitabine for 24 h (Fig 8D). Since tunicamycin disrupts the formation of N-glycosidic linkages by inhibiting the first step in glycoprotein synthesis, and kifunensine blocks endoplasmic reticulum mannosidase I, which halts processing of glycoproteins and the transport to the Golgi, we assume that gemcitabine acts to disrupt the hERG channel glycosylation in the endoplasmic reticulum. Of note, gemcitabine up-regulated immature protein levels of hERG, and down-regulated cell surface expression of hERG. Because mature form of hERG was unchanged in Western blot analyses, and cell surface expression of hERG protein was significantly reduced as assessed by fluorescence intensity by gemcitabine treatment, it is suggested that gemcitabine may also affect the trafficking process of the hERG protein toward the plasma membrane. Taken all together with the result on hERG mRNA, it is likely to conclude that electrophysiological toxicity of gemcitabine on cardiomyocytes can be caused by post-translational disruption of the hERG channel.

We note the fact that disruption in glycan biosynthesis in some tumor cells could profoundly impact multiple molecular events and signaling pathways associated with anticancer drug resistance [38]. Actually, activation of endoplasmic reticulum stress and inhibition of N-glycosylation by tunicamycin enhances susceptibility of some cancer cells to anticancer drugs [39], which implies that some anticancer drugs may act to cancer cells somehow functionally related to protein glycosylation disruption. However, actions of gemcitabine may not be identical to that of tunicamycin on the hERG channel protein based on the fact that gemcitabine shifted the voltage dependency of I_hERG activation toward the hyperpolarized direction whereas tunicamycin was without effect on it (Fig 7D). Since gemcitabine modified the activation gating properties of the hERG channel (Fig 3), unknown long-term actions of gemcitabine on the hERG channel has been postulated accordingly.

Limitation of the study

There are several limitations in the present study. Although we have demonstrated that gemcitabine has a potential to suppress I_hERG as a long term effect in heterologous expression system, we could not confirm the result on I_Kr in cardiomyocytes in detail. It is important to keep in mind that transcriptional, post-transcriptional and post-translation modulation pathways for hERG protein are quite different between heterologous hERG-HEK cells and cardiomyocytes. Also, the structure of glycosylation and its functional implications can vary significantly between different cell types, and between cultured cells and cardiac cells in vivo. Because an incubation technique for the cultivation of adult cardiomyocytes for the duration of 24 h or longer has not been established, we usually apply electrophysiological studies by use of neonatal cardiomyocyte for these purposes. However, neonatal cardiomyocytes only express very small density of I_Kr (Fig 9A and 9B), which is not suitable for the evaluation of the long-term effect of gemcitabine on the hERG channel [15]. Actually, gemcitabine was without effect on action potential configuration in neonatal cardiomyocytes (Fig 9C–9E), probably because I_Kr density was as small as ~1 pA/pF. As such, cardiac cells generated from iPS cells may be suitable for this analysis. Albeit we have obtained EC₅₀ values of gemcitabine to suppress I_hERG, a simple comparison of drug concentrations between in vitro study and human serum sample may not be appropriate. Because gemcitabine is a lipophilic drug, assessment of drug efficacy on I_hERG by an in vitro experiment without serum proteins and/or serum lipids may not represent the effect of gemcitabine in in vivo application. Furthermore, a caution would be needed to extrapolate these single cell experiments to human therapeutics. Regarding several glycosylation steps in the endoplasmic reticulum and in the Golgi apparatus, the first step of N-glycosylation in the assemblage of dolichol-linked precursor oligosaccharide at the cytoplasmic side of the endoplasmic reticulum could be blockaded by tunicamycin [40, 41]. As gemcitabine but not tunicamycin modified the activation gating of the hERG channel, and because a specific class I mannosidase inhibitor kifunensine was unable to reduce I_hERG when hERG-HEK cells were pretreated with gemcitabine, it is speculated that gemcitabine regulates some glycosylation step(s) between the GlcNAc phosphotransferase site andα-1, 2-mannosidase I site in the endoplasmic reticulum, although we were unable to identify the target molecule(s) of gemcitabine in the whole process of hERG protein glycosylation. Because this study is solely dependent upon electrophysiological evaluation of I_hERG, molecular biological analyses for N-glycosylation are obviously needed to address these questions.

Conclusions

Our study revealed that the long-term application of gemcitabine reduces I_hERG possibly through the inhibition of N-linked glycosylation of the channel, which may explain the hitherto unknown pharmacological mechanism of gemcitabine to cause QT prolongation in ECG. Furthermore, excessive reduction of I_hERG by gemcitabine in cardiomyocytes may account for supraventricular arrhythmogenicity of this drug as documented in clinical reports.

Supporting information

S1 File.

https://doi.org/10.1371/journal.pone.0280656.s001

Acknowledgments

The HEK293 cells expressing hERG are a kind gift from Prof. Imaizumi Yuji, Nagoya City University, Graduate School of Pharmaceutical Sciences.

Citation: Wei M, Wang P, Zhu X, Morishima M, Liu Y, Zheng M, et al. (2023) Electrophysiological evaluation of an anticancer drug gemcitabine on cardiotoxicity revealing down-regulation and modification of the activation gating properties in the human rapid delayed rectifier potassium channel. PLoS ONE 18(2): e0280656. https://doi.org/10.1371/journal.pone.0280656

About the Authors:

Mengyan Wei

Contributed equally to this work with: Mengyan Wei, Pu Wang, Xiufang Zhu

Roles: Data curation, Investigation

Affiliations: Department of Cardiology, The First Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, People’s Republic of China, Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Pu Wang

Contributed equally to this work with: Mengyan Wei, Pu Wang, Xiufang Zhu

Roles: Data curation

Affiliations: Department of Cardiology, The First Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, People’s Republic of China, Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Xiufang Zhu

Contributed equally to this work with: Mengyan Wei, Pu Wang, Xiufang Zhu

Roles: Data curation

Affiliations: Department of Cardiology, The First Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, People’s Republic of China, Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Masaki Morishima

Roles: Formal analysis, Investigation, Methodology

Affiliation: Department of Food Science and Nutrition, Faculty of Agriculture, Kindai University, Nara, Japan

Yangong Liu

Roles: Data curation

Affiliations: Department of Cardiology, The First Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, People’s Republic of China, Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Mingqi Zheng

Roles: Formal analysis, Methodology

Affiliation: Department of Cardiology, The First Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, People’s Republic of China

Gang Liu

Roles: Methodology, Project administration

Affiliation: Department of Cardiology, The First Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, People’s Republic of China

Hiroki Osanai

Roles: Formal analysis

Affiliation: Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Kenshi Yoshimura

Roles: Investigation

Affiliation: Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Shinichiro Kume

Roles: Formal analysis

Affiliation: Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Tatsuki Kurokawa

Roles: Formal analysis

Affiliation: Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

Katsushige Ono

Roles: Supervision, Writing – original draft

E-mail: [email protected]

Affiliation: Department of Pathophysiology, Oita University School of Medicine, Yufu, Oita, Japan

ORICD: https://orcid.org/0000-0001-7949-1823

References

1. Plunkett W, Huang P, Xu YZ, Heinemann V, Grunewald R, Gandhi V. Gemcitabine: metabolism, mechanisms of action, and self-potentiation. Semin. Oncol. 1995, 22 (4 Suppl 11), 3–10. pmid:7481842

2. Hilmi M, Ederhy S, Waintraub X, Funck-Brentano C, Cohen A, Vozy A, et al. Cardiotoxicity Associated with Gemcitabine: Literature Review and a Pharmacovigilance Study. Pharmaceuticals (Basel). 2020, 13, 325. pmid:33096756

3. Santini D, Tonini G, Abbate A, Di Cosimo S, Gravante G, Vincenzi B, et al. Gemcitabine-induced atrial fibrillation: a hitherto unreported manifestation of drug toxicity. Ann. Oncol. 2000, 11, 479–481. pmid:10847470

4. Gridelli C, Cigolari S, Gallo C, Manzione L, Ianniello GP, Frontini L, et al; MILES Investigators. Activity and toxicity of gemcitabine and gemcitabine + vinorelbine in advanced non-small-cell lung cancer elderly patients: Phase II data from the Multicenter Italian Lung Cancer in the Elderly Study (MILES) randomized trial. Lung Cancer. 2001, 31, 277–284.

5. Buza V, Rajagopalan B, Curtis AB. Cancer Treatment-Induced Arrhythmias: Focus on Chemotherapy and Targeted Therapies. Circ. Arrhythm. Electrophysiol. 2017, 8, e005443. pmid:28798022

6. Gridelli C, Novello S, Zilembo N, Luciani A, Favaretto AG, De Marinis F, et al. Phase II randomized study of vandetanib plus gemcitabine or gemcitabine plus placebo as first-line treatment of advanced non-small-cell lung cancer in elderly patients. J. Thorac. Oncol. 2014, 9, 733–737. pmid:24722160

7. Rademaker-Lakhai JM, Beerepoot LV, Mehra N, Radema SA, van Maanen R, Vermaat JS, et al. Phase I pharmacokinetic and pharmacodynamic study of the oral protein kinase C beta-inhibitor enzastaurin in combination with gemcitabine and cisplatin in patients with advanced cancer. Clin. Cancer Res. 2007, 13, 4474–4481. pmid:17671132

8. Santoro A, Gebbia V, Pressiani T, Testa A, Personeni N, Arrivas Bajardi E, et al. A randomized, multicenter, phase II study of vandetanib monotherapy versus vandetanib in combination with gemcitabine versus gemcitabine plus placebo in subjects with advanced biliary tract cancer: the VanGogh study. Ann. Oncol. 2015, 26, 542–547. pmid:25538178

9. Kindler HL, Ioka T, Richel DJ, Bennouna J, Létourneau R, Okusaka T, et al. Axitinib plus gemcitabine versus placebo plus gemcitabine in patients with advanced pancreatic adenocarcinoma: a double-blind randomised phase 3 study. Lancet Oncol. 2011, 12, 256–262. pmid:21306953

10. Kirchhof P, Eckardt L, Franz MR, Mönnig G, Loh P, Wedekind H, et al. Prolonged atrial action potential durations and polymorphic atrial tachyarrhythmias in patients with long QT syndrome. J. Cardiovasc. Electrophysiol. 2003, 14, 1027–1033. pmid:14521653

11. Johnson JN, Tester DJ, Perry J, Salisbury BA, Reed CR, Ackerman MJ. Prevalence of early-onset atrial fibrillation in congenital long QT syndrome. Heart Rhythm. 2008, 5, 704–709. pmid:18452873

12. Zellerhoff S, Pistulli R, Mönnig G, Hinterseer M, Beckmann BM, Köbe J, et al. Atrial arrhythmias in long-QT syndrome under daily life conditions: a nested case control study. J. Cardiovasc. Electrophysiol. 2009, 20, 401–407. pmid:19017345

13. Mandyam MC, Soliman EZ, Alonso A, Dewland TA, Heckbert SR, Vittinghoff E, et al. The QT interval and risk of incident atrial fibrillation. Heart Rhythm. 2013, 10, 1562–1568. pmid:23872693

14. Platonov PG, McNitt S, Polonsky B, Rosero SZ, Zareba W. Atrial fibrillation in long QT syndrome by genotype. Circ. Arrhythm. Electrophysiol. 2019, 12, e007213. pmid:31610692

15. Wang Y, Morishima M, Li D, Takahashi N, Saikawa T, Nattel S, et al. Binge alcohol exposure triggers atrial fibrillation through T-type Ca2+ channel upregulation via protein kinase C (PKC) / glycogen synthesis kinase 3β (GSK3β) / nuclear factor of activated T-cells (NFAT) signaling—An experimental account of holiday heart syndrome. Circ. J. 2020, 84, 1931–1940.

16. Ma F, Takanari H, Masuda K, Morishima M, Ono K. Short- and long-term inhibition of cardiac inward-rectifier potassium channel current by an antiarrhythmic drug bepridil. Heart Vessels. 2016, 31, 1176–1184. pmid:26498939

17. Masuda K, Takanari H, Morishima M, Ma F, Wang Y, Takahashi N, et al. Testosterone-mediated upregulation of delayed rectifier potassium channel in cardiomyocytes causes abbreviation of QT intervals in rats. J. Physiol. Sci. 2018, 68, 759–767. pmid:29332211

18. Morishima M, Iwata E, Nakada C, Tsukamoto Y, Takanari H, Miyamoto S, et al. Atrial fibrillation-mediated upregulation of miR-30d regulates myocardial electrical remodeling of the G-protein-gated K+ channel, IK.ACh. Circ. J. 2016, 80, 1346–1355.

19. Liu S, Rasmusson RL, Campbell DL, Wang S, Strauss HC. Activation and inactivation kinetics of an E-4031-sensitive current from single ferret atrial myocytes. Biophys. J. 1996, 70, 2704–1715. pmid:8744308

20. Shibasaki T. Conductance and kinetics of delayed rectifier potassium channels in nodal cells of the rabbit heart. J Physiol. 1987, 387, 227–250.

21. Gintant GA. Two components of delayed rectifier current in canine atrium and ventricle. Does IKs play a role in the reverse rate dependence of class III agents? Circ. Res. 1996, 78, 26–37.

22. Vandenberg JI, Perry MD, Perrin MJ, Mann SA, Ke Y, Hill AP. hERG K+ channels: structure, function, and clinical significance. Physiol. Rev. 2012, 92, 1393–1478.

23. Hirashita Y, Tsukamoto Y, Kudo Y, Kakisako D, Kurogi S, Hijiya N, et al. Early response in phosphorylation of ribosomal protein S6 is associated with sensitivity to trametinib in colorectal cancer cells. Lab. Invest. 2021, 101, 1036–1047. pmid:33911189

24. Sandler A, Ettinger DS. Gemcitabine: single-agent and combination therapy in non-small cell lung cancer. Oncologist. 1999, 4, 241–251. pmid:10394591

25. Veerman G, Ruiz van Haperen VW, Vermorken JB, Noordhuis P, Braakhuis BJ, Pinedo HM, et al. Antitumor activity of prolonged as compared with bolus administration of 2’,2’-difluorodeoxycytidine in vivo against murine colon tumors. Cancer Chemother. Pharmacol. 1996, 38, 335–342. pmid:8674156

26. Miyahara K, Nouso K, Morimoto Y, Kinugasa H, Kato H, Yamamoto N, et al. Prognostic value of altered N-glycosylation of circulating glycoproteins in patients with unresectable pancreatic cancer treated with gemcitabine. Pancreas. 2015, 44, 551–556. pmid:25875794

27. Lamothe SM, Hulbert M, Guo J, Li W, Yang T, Zhang S. Glycosylation stabilizes hERG channels on the plasma membrane by decreasing proteolytic susceptibility. FASEB J. 2018, 32, 1933–1943. pmid:32172531

28. Macharoen K, Li Q, Márquez-Escobar VA, Corbin JM, Lebrilla CB, Nandi S, et al. Effects of kifunensine on production and N-glycosylation modification of butyrylcholinesterase in a transgenic rice cell culture bioreactor. Int. J. Mol. Sci. 2020, 21, 6896. pmid:32962231

29. Jeevaratnam K, Chadda KR, Huang CL, Camm AJ. Cardiac potassium channels: Physiological insights for targeted therapy. J. Cardiovasc. Pharmacol. Ther. 2018, 23, 119–129. pmid:28946759

30. Klein AL, Tajik AJ. Doppler assessment of pulmonary venous flow in healthy subjects and in patients with heart disease. J. Am. Soc. Echocardiogr. 1991, 4, 379–392. pmid:1910836

31. Bukachi F, Waldenström A, Mörner S, Lindqvist P, Henein MY, Kazzam E. Pulmonary venous flow reversal and its relationship to atrial mechanical function in normal subjects—Umeå General Population Heart Study. Eur. J. Echocardiogr. 2005, 6, 107–116.

32. Nishimura RA, Abel MD, Hatle LK, Tajik AJ. Relation of pulmonary vein to mitral flow velocities by transesophageal doppler echocardiography. Effect of different loading conditions. Circulation. 1990, 81, 1488–1497. pmid:2110034

33. Li D, Zhang L, Kneller J, Nattel S. Potential ionic mechanism for repolarization differences between canine right and left atrium. Circ. Res. 2001, 88, 1168–1175. pmid:11397783

34. Vandetanib gemcitabine or placebo plus gemcitabine or vandetanib monotherapy in advanced biliary tract cancer (VANGOGH) Vandetanib gemcitabine or placebo plus gemcitabine or vandetanib monotherapy in advanced biliary tract cancer—tabular view—ClinicalTrials.gov

35. Witchel HJ, Hancox JC. Familial and acquired long QT syndrome and the cardiac rapid delayed rectifier potassium current. Clin. Exp. Pharmacol. Physiol. 2000, 27, 753–766. pmid:11022966

36. Gong Q, Anderson CL, January CT, Zhou Z. Role of glycosylation in cell surface expression and stability of HERG potassium channels. Am J. Physiol. Heart. Circ. Physiol. 2002, 283, H77–H84. pmid:12063277

37. Petrecca K, Atanasiu R, Akhavan A, Shrier A. N-linked glycosylation sites determine HERG channel surface membrane expression. J. Physiol. 1999, 515, 41–48. pmid:9925876

38. Chen R, Lai LA, Sullivan Y, Wong M, Wang L, Riddell J, et al. Disrupting glutamine metabolic pathways to sensitize gemcitabine-resistant pancreatic cancer. Sci. Rep. 2017, 7, 7950. pmid:28801576

39. Ling YH, Li T, Perez-Soler R, Haigentz M Jr. Activation of ER stress and inhibition of EGFR N-glycosylation by tunicamycin enhances susceptibility of human non-small cell lung cancer cells to erlotinib. Cancer Chemother. Pharmacol. 2009, 64, 539–548. pmid:19130057

40. Elbein AD. Glycosidase inhibitors: inhibitors of N-linked oligosaccharide processing. FASEB J. 1991, 5, 3055–2063. pmid:1743438

41. Skrenkova K, Lee S, Lichnerova K, Kaniakova M, Hansikova H, Zapotocky M, et al. N-glycosylation regulates the trafficking and surface mobility of GluN3A-containing NMDA receptors. Front. Mol. Neurosci. 2018, 11, 188. pmid:29915530