Mechanism of a methylxanthine drug theophylline-induced Ca²⁺ signaling and cytotoxicity in AML12 mouse hepatocytes

Abstract

Theophylline is a methylxanthine drug used in therapy for respiratory diseases. However, the impact of theophylline on Ca²⁺ signaling has not been explored in liver cells. This study examined whether theophylline affected Ca²⁺ homeostasis and its related cytotoxicity in AML12 mouse hepatocytes. Cell viability was measured by the cell viability reagent (WST-1). Cytosolic Ca²⁺ concentration ([Ca²⁺]_i) was measured by the Ca²⁺-sensitive fluorescent dye fura-2. Theophylline (25–125 μM) induced [Ca²⁺]_i rises and cause cytotoxicity in AML12 cells. This cytotoxic response was reversed by chelation of cytosolic Ca²⁺ with BAPTA/AM. In Ca²⁺-free medium, treatment with the endoplasmic reticulum Ca²⁺ pump inhibitor thapsigargin abolished theophylline-induced [Ca²⁺]_i rises. Conversely, treatment with theophylline also abolished thapsigargin-induced [Ca²⁺]_i rises. However, inhibition of PLC failed to alter theophylline-evoked [Ca²⁺]_i rises. In Ca²⁺-containing medium, modulators of store-operated Ca²⁺ channels inhibited 30% of the [Ca²⁺]_i rises, whereas the PKC modulators had no effect. Furthermore, theophylline-induced Ca²⁺ influx was confirmed by Mn²⁺-induced quench of fura-2 fluorescence. Together, in AML12 cells, theophylline caused Ca²⁺-associated cytotoxicity and induced Ca²⁺ entry through PLC-independent Ca²⁺ release from the endoplasmic reticulum and PKC-insensitive store-operated Ca²⁺ channels. BAPTA-AM with its protective effects may be a potential compound for prevention of theophylline-induced cytotoxicity.

Introduction

Theophylline has long been regarded as a major bronchodilator in the treatment of human asthma [1]. As a member of the xanthine family, it bears structural and pharmacological similarity to theobromine and caffeine [2]. Theophylline has been used to treat airway diseases for more than 80 years. It has been shown that theophylline has anti-inflammatory effects in asthma and chronic obstructive pulmonary disease (COPD) [3]. Theophylline is given orally as slow-release preparations for chronic treatment and intravenously for acute exacerbations of asthma [4]. Regarding effect of theophylline on physiological responses in various models, theophylline was shown to inhibit DNA synthesis and cell viability in MDA-MB-231 human breast cancer cells [5], induce growth inhibition and apoptosis in human rhynopharyngeal KB and lung H1355 epidermoid carcinoma cell lines [6], and partially revert cachexia in tumor-bearing rats [7]. Therefore, in addition to treating respiratory disease, theophylline is a prospective drug to examine the impact of cytotoxicity in cultured cell models.

A change and fluctuation in cytosolic Ca²⁺ concentration ([Ca²⁺]_i) is a pivotal second messenger of numerous cellular responses including gene expression, protein activation, secretion, plasticity, proliferation, and death [8]. Some diseases have been shown to be caused by abnormal Ca²⁺ homeostasis, such as Alzheimer diseases [9], cardiac diseases [10], and cancers [8]. A Ca²⁺ signal can occur by Ca²⁺ entry from extracellular medium or be released from internal organelles [11,12]. There are different Ca²⁺ channels and receptors responsible for the Ca²⁺ influx on the plasma membrane [12]. These channels or receptors usually stimulate phospholipase C (PLC) rendering Ca²⁺ release from intracellular stores such as the endoplasmic reticulum, which evoked Ca²⁺ influx across the plasma membrane [12,13]. Thus, the exploration of the mechanisms of the drug-induced [Ca²⁺]_i rises is very important for understanding the drug’s effect on the cell.

In terms of Ca²⁺ signaling, a few papers have reported that theophylline interacts with [Ca²⁺]_i. It has been shown that theophylline released Ca²⁺ from intracellular stores of pig oocytes [14], and attenuated Ca²⁺ sensitivity in porcine tracheal smooth muscles [15]. Moreover, a previous research has shown that Ca²⁺ signaling plays a major role in theophylline-induced toxicity and death in Sprague–Dawley rats and beagle dogs [16]. However, the effect of theophylline on Ca²⁺ homeostasis and its related underlying mechanisms in liver cells has not explored.

Because theophylline is widely used anti-asthma drug and may provide pharmacological perspective for human, the effect of use of theophylline on cytotoxic responses should be cautioned in liver cells. The goal of this study was to examine the effect of theophylline on [Ca²⁺]_i and cell viability in AML12 (alpha-mouse-liver-12) mouse hepatocytes. AML12 cells were established from the normal liver of a 3-month-old male mouse and were often used as a normal control in cancer studies [17]. Fura-2 was used as a fluorescent Ca²⁺-sensitive dye to measure [Ca²⁺]_i changes in this study. Theophylline-induced [Ca²⁺]_i rises in cell models were characterized, the concentration-response plots were established, and the pathways underlying the [Ca²⁺]_i rise were explored. The action of theophylline on cell viability was also elucidated.

Materials and Method

Reagent

Theophylline (Fig. 1A) was from Sigma-Aldrich® (St. Louis, MO, USA). Aminopolycarboxylic acid/acetoxy methyl (fura-2-AM) and 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid/acetoxy methyl (BAPTA-AM) were from Molecular Probes® (Eugene, OR, USA). Thapsigargin (TG, an endoplasmic reticulum Ca²⁺ pump inhibitor), U73122 (a PLC inhibitor), adenosine triphosphate (ATP, a PLC-dependent agonist), modulators of store-operated Ca²⁺ channels (2-APB, econazole and SKF96365), phorbol 12-myristate 13 acetate [PMA; a protein kinase C (PKC) activator], and GF109203X (a PKC inhibitor) were from Sigma-Aldrich® (St. Louis, MO, USA) unless otherwise indicated. The reagents for cell culture were from Gibco® (Gaithersburg, MD, USA).

Figure 1.

Effect of theophylline on [Ca²⁺]_i in fura-2-loaded AML12 cells. (A) The chemical structure of theophylline. (B) Theophylline was added at 25 s. The concentration of theophylline was indicated. The experiments were performed in Ca²⁺-containing medium. (C) Effect of theophylline on [Ca²⁺]_i in Ca²⁺-free medium. Theophylline was added at 25 s in Ca²⁺-free medium. (D) Concentration-response plots of theophylline-induced [Ca²⁺]_i rises. Y axis is thepercentage of the net (baseline subtracted) area under the curve (25–250 s) of the [Ca²⁺]_i rises induced by 100 μM theophylline in Ca²⁺-containing medium. Data are mean ± SEM of three independent experiments. ^*P < 0.05 compared to open circles.

Cell culture

AML12 (alpha-mouse-liver-12) mouse hepatocytes (BCRC® 60326™) were purchased from Bioresource Collection and Research Center (BCRC, Taiwan). AML12 cells were cultured in 90% 1:1 mixture of DMEM and Ham’s F12 medium supplemented with 10% heat-inactivated fetal bovine serum, 100 units mL⁻¹ penicillin and 100 μg mL⁻¹ streptomycin. All these cell models were maintained at 37°C in a humidified 5% CO₂ atmosphere.

Solutions used in [Ca²⁺]_i measurements

Ca²⁺-containing medium (pH 7.4) had 140 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), and 5 mM glucose. Ca²⁺-free medium contained similar chemicals as Ca²⁺-containing medium except that CaCl₂ was replaced with 0.3 mM ethylene glycol tetraacetic acid (EGTA) and 2 mM MgCl₂. Theophylline was dissolved in dimethyl sulfoxide (DMSO) as a 0.1 M stock solution. The other chemicals were dissolved in water, ethanol, or DMSO. The concentration of organic solvents in the experimental solutions was <0.1%, and did not alter viability and basal [Ca²⁺]_i.

[Ca²⁺]_i measurements

Confluent cells grown on 6 cm dishes were trypsinized and made into a suspension in culture medium at a concentration of 1 × 10⁶ cells mL⁻¹. Cell viability was determined by trypan blue exclusion. The viability was >95% after the treatment. Cells were subsequently loaded with 2 μM fura-2-AM for 30 min at 25°C in the same medium. After loading, cells were washed with Ca²⁺-containing medium twice and were made into a suspension in Ca²⁺-containing medium at a concentration of 1 × 10⁷ cells mL⁻¹. Fura-2 fluorescence measurements were performed in a water-jacketed cuvette (25°C) with continuous stirring; the cuvette contained 1 mL of medium and 1 × 10⁶ cells. Fluorescence was monitored with a Shimadzu RF-5301PC spectrofluorophotometer immediately after 0.1 mL cell suspension was added to 0.9 mL Ca²⁺-containing or Ca²⁺-free medium, by recording excitation signals at 340 and 380 nm and emission signal at 510 nm at 1-s intervals. During the recording, reagents were added to the cuvette by pausing the recording for 2 s to open and close the cuvette-containing chamber. For calibration of [Ca²⁺]_i, after completion of the experiments, the detergent Triton X-100 (0.1%) and CaCl₂ (5 mM) were added to the cuvette to obtain the maximal fura-2 fluorescence. Then the Ca²⁺ chelator EGTA (10 mM) was added to chelate Ca²⁺ in the cuvette to obtain the minimal fura-2 fluorescence. Control experiments showed that cells bathed in a cuvette had a viability of 95% after 20 min of fluorescence measurements. [Ca²⁺]_i was calculated as previously described [18].

Cell viability assays

The measurement of cell viability was based on the ability of cells to cleave tetrazolium salts by dehydrogenases. The intensity of developed color directly correlated with the number of live cells. Experiments were conducted according to manufacturer’s instructions specifically designed for this assay (Roche Molecular Biochemical, Indianapolis, IN, USA). Cells were seeded in 96-well plates at a density of 1 × 10⁴ cells per well in culture medium for 24 h in the presence of 0–125 μM theophylline. The cell viability detecting reagent 4-[3-[4-lodophenyl]-2-4(4-nitrophenyl)-2H-5-tetrazolio-1,3-benzene disulfonate] (WST-1; 10 μL pure solution) was added to samples after treatment with theophylline, and cells were incubated for 30 min in a humidified atmosphere. In experiments using BAPTA-AM to chelate cytosolic Ca²⁺ to prevent [Ca²⁺]_i rises, cells were treated with 5 μM BAPTA-AM for 1 h before treatment with theophylline. The cells were washed once with Ca²⁺-containing medium and treated with or without theophylline for 24 h. The absorbance of samples (A450) was measured using a multiwall plate reader. Absolute optical density was normalized to the absorbance of unstimulated cells in each plate and expressed as a percentage of the control.

Mn²⁺ quenching assays

For Mn²⁺ quenching experiments, Mn²⁺ quenching of fura-2 fluorescence was performed in Ca²⁺-containing medium containing 50 μM MnCl₂. MnCl₂ was added to cell suspension in the cuvette 30 s before the fluorescence recording was started. Data were recorded at excitation signal at 360 nm (Ca²⁺-insensitive) and emission signal at 510 nm at 1-s intervals as described previously [19].

Statistical analysis

Data are reported as mean ± standard error of the mean (SEM) of three independent experiments (n = 3), and were analyzed by one-way analysis of variances (ANOVA) using the Statistical Analysis System (SAS®, SAS Institute Inc., Cary, NC, USA). Multiple comparisons between group means were performed by post hoc analysis using the Tukey’s HSD (honestly significantly difference) procedure. A P value <0.05 were considered significant.

Results

Theophylline concentration-dependently induced [Ca²⁺]_i rises in a Ca²⁺-containing or Ca²⁺-free medium in AML12 cells

The effect of theophylline on basal [Ca²⁺]_i was examined. The resting [Ca²⁺]_i concentrations in the AML12 cells were 51 ± 3 nM (n = 3) (Fig. 1B). Theophylline at concentrations between 25 and 100 μM evoked [Ca²⁺]_i rises in a concentration-dependent fashion in Ca²⁺-containing medium. At a concentration of 100 μM, theophylline caused [Ca²⁺]_i rises that reached a net increase of 125 ± 3 nM (n = 3) followed by a slow decay (Fig. 1B). The Ca²⁺ response saturated at 100 μM theophylline because at a concentration of 125 μM, theophylline induced a similar response as that induced by 100 μM (not shown). Figure 1C shows that in Ca²⁺-free medium, 50–100 μM theophylline evoked concentration-dependent rises in [Ca²⁺]_i. In the absence of extracellular Ca²⁺, 100 μM theophylline induced [Ca²⁺]_i rises of 45 ± 3 nM (n = 3). Figure 1d shows the concentration–response relationships of theophylline-evoked [Ca²⁺]_i rises. The EC₅₀ value was 45 ± 3 μM or 75 ± 3 μM in Ca²⁺-containing medium or Ca²⁺-free medium, respectively, by fitting to a Hill equation. Ca²⁺ removal reduced the Ca²⁺ signal by ~30%.

A Ca²⁺ chelator BAPTA/AM treatment partially reversed theophylline-induced cytotoxicity in AML12 cells

Given that acute incubation with theophylline induced substantial [Ca²⁺]_i rises, and that unregulated [Ca²⁺]_i rises often alter cell viability [8,12], experiments were performed to examine the effect of theophylline on cell viability. The intracellular Ca²⁺ chelator BAPTA/AM [20] was used to prevent [Ca²⁺]_i rises during theophylline treatment. After treatment with 5 μM BAPTA/AM, 0–125 μM theophylline failed to evoke [Ca²⁺]_i rises (Fig. 2A). This suggests that BAPTA-AM effectively chelated cytosolic Ca²⁺. Cells were treated with 0–125 μM theophylline for 24 h, and the tetrazolium assay was performed. In the presence of 25–125 μM theophylline, cell viability decreased concentration-dependently (Fig. 2B). The next question was whether the theophylline-induced cytotoxicity was caused by preceding [Ca²⁺]_i rises. Figure 2b shows that 5 μM BAPTA-AM loading did not alter the control value of cell viability. In the presence of 25–125 μM theophylline, BAPTA-AM loading partially reversed theophylline-induced cell death in AML12 cells by 6.2 ± 0.5%, 18.5 ± 0.6%, 25.3 ± 0.8%, 14.2 ± 0.3% or 8.2 ± 0.3% (P < 0.05) (n = 3), respectively (Fig. 2B). Therefore, the findings implicate that theophylline-induced cytotoxicity was caused by [Ca²⁺]_i rises in AML12 cells.

Figure 2.

Cytotoxic effect of theophylline in AML12 cells. (A) Following BAPTA/AM treatment, cells were incubated with fura-2/AM as described in “Materials and Method”. Then [Ca²⁺]_i measurements were conducted in Ca²⁺-containing medium. Theophylline (20–100 μM) was added as indicated. (B) Cells were treated with 0–125 μM theophylline for 24 h, and then cell viability assay was conducted in AML12 cells. Data are mean ± SEM of three independent experiments. Each treatment had six replicates (wells). Data are expressed as percentage of control that is the increase in cell numbers in theophylline-free groups. Control had 11 567 ± 325 cells/well before experiments, and 13 123 ± 333 cells/well after incubation for 24 h. ^*P < 0.05 compared to control. In each group, the Ca²⁺ chelator BAPTA-AM (5 μM) was added to cells followed by treatment with theophylline in Ca²⁺-containing medium. Cell viability assay was subsequently performed. ^#P < 0.05 compared to the pairing group.

Theophylline induced Ca²⁺ release from endoplasmic reticulum in AML12 cells

In non-excitable cell types including AML12 cells, the endoplasmic reticulum has been thought to be the primary Ca²⁺ depot [11,12]. Thus, the role of the endoplasmic reticulum in theophylline-evoked Ca²⁺ release in AML12 cells was examined. The experiments were performed in Ca²⁺-free medium to exclude the participation of Ca²⁺ influx. Because Ca²⁺ response induced by theophylline was saturated at 100 μM, in the following experiments 100 μM theophylline was used as control. Figure 3A shows that after 100 μM theophylline-induced [Ca²⁺]_i rises, addition of 1 μM TG [21], an endoplasmic reticulum Ca²⁺ pump inhibitor, failed to induce [Ca²⁺]_i rises. Conversely, Fig. 3B shows that addition of 1 μM TG induced [Ca²⁺]_i rises of 45 ± 3 nM (n = 3). Addition of 100 μM theophylline afterwards failed to induce Ca²⁺ release. This suggests that theophylline induced [Ca²⁺]_i rises by causing Ca²⁺ release from the endoplasmic reticulum.

Figure 3.

Effect of TG on Ca²⁺ release induced by theophylline in AML12 cells. (A and B) TG (1 μM) and theophylline (100 μM) were added at time points indicated. Experiments were performed in Ca²⁺-free medium. Data are mean ± SEM of three independent experiments.

Lack of a role of PLC in theophylline-evoked [Ca²⁺]_i rises in AML12 cells

Among the many cytosolic enzymes that trigger the release of Ca²⁺ from the endoplasmic reticulum, PLC plays a key role [11,12]. Because theophylline released Ca²⁺ from the endoplasmic reticulum, the role of PLC in this process was examined. The PLC inhibitor U73122 [22], was applied to explore if the activation of this enzyme was required for Ca²⁺ release induced by theophylline. First, ATP was used to test the activity of U73122. Figure 4A shows that ATP (10 μM) induced [Ca²⁺]_i rises of 63 ± 3 nM (n = 3). ATP is a PLC-dependent agonist of [Ca²⁺]_i rises in most cell types [23], and therefore is the reason to be used as a tool to examine whether U73122 effectively inhibited the activity of PLC. Figure 4B shows that incubation with 2 μM U73122 did not change basal [Ca²⁺]_i but abolished ATP-induced [Ca²⁺]_i rises. However, incubation with U73122 did not inhibit 100 μM theophylline-induced [Ca²⁺]_i rises. Furthermore, treatment with U73122 also did not reduce the combination of ATP and theophylline-induced [Ca²⁺]_i rises (n = 3). U73343 (2 μM), a U73122 analog, failed to have an inhibition on ATP-induced [Ca²⁺]_i rises (not shown). U73343 is a PLC-insensitive structural analog of U73122 and is often used as a control for U73122 activity. Therefore, it suggests that Ca²⁺ release induced by theophylline was not associated with PLC activity.

Figure 4.

Lack of an effect of U73122 on Ca²⁺ release induced by theophylline in AML12 cells. Experiments were performed in Ca²⁺-free medium. (A) ATP (10 μM) was added as indicated. (B) From left to right (the time zone of the area under the curve), first column is 100 μM theophylline-induced [Ca²⁺]_i rises. Second column shows that 2 μM U73122 did not alter basal [Ca²⁺]_i. Third column shows ATP-induced [Ca²⁺]_i rises. Fourth column shows that U73122 pretreatment for 200 s completely abolished ATP-induced [Ca²⁺]_i rises. Fifth column shows that U73122 pretreatment for 200 s did not inhibit 100 μM theophylline-induced [Ca²⁺]_i rises. Sixth column shows that U73122 (incubation for 200 s) and ATP (incubation for 50 s) pretreatment did not inhibit 100 μM theophylline-induced [Ca²⁺]_i rises. Data are mean ± SEM of three independent experiments. ^*P < 0.05 compared to first bar (control). Control is the area under the curve of 100 μM theophylline-induced [Ca²⁺]_i rises (25–220 s).

Theophylline-induced 2-APB-sensitive but PKC-insensitive store-operated Ca²⁺ entry in AML12 cells

Experiments were conducted to explore the Ca²⁺ entry mechanism of theophylline-induced [Ca²⁺]_i rises. Store-operated Ca²⁺ entry modulators (2-APB, 20 μM, econazole, 0.5 μM, and SKF96365, 5 μM) [24–27], a PKC activator (PMA, 1 nM) [28], or a PKC inhibitor (GF109203X, 2 μM) [29] were added 1 min before 100 μM theophylline. Except PMA and GF109203X, 2-APB, econazole and SKF96365 inhibited theophylline-evoked [Ca²⁺]_i rises by 30 ± 6%, 26 ± 6%, or 28 ± 6%, respectively (P < 0.05) (n = 3) (Fig. 5). This suggests that PKC-insensitive or 2-APB-sensitive store-operated Ca²⁺ entry were involved in theophylline-induced [Ca²⁺]_i rises.

Figure 5.

Effect of Ca²⁺ channel modulators on theophylline-induced [Ca²⁺]_i rises in AML12 cells. In modulator-treated groups, the reagent was added 1 min before theophylline (100 μM). The concentration was 20 μM for 2-APB, 0.5 μM for econazole, 5 μM for SKF96365, 10 nM for phorbol 12-myristate 13-acetate (PMA), 2 μM for GF109203X, Data are expressed as the percentage of control (1^st column) that is the area under the curve (25–200 s) of 100 μM theophylline-induced [Ca²⁺]_i rises, and are mean ± SEM of three independent experiments. ^*P < 0.05 compared to the first column.

Theophylline induced Mn²⁺ influx in AML12 cells

Experiments were conducted to confirm that theophylline-induced [Ca²⁺]_i rises involved Ca²⁺ influx. Mn²⁺ enters cells through similar pathways as Ca²⁺ but quenches fura-2 fluorescence at all excitation wavelengths [19]. Based on this rational, quenching of fura-2 fluorescence excited at the Ca²⁺-insensitive excitation wavelength of 360 nm by Mn²⁺ implicates Ca²⁺ entry. Figure 6 shows that 100 μM theophylline induced an instant decrease in the 360 nm excitation signal that reached a maximum value of 103 ± 3 arbitrary units (n = 3). This implies that Ca²⁺ influx participates in theophylline-induced [Ca²⁺]_i rises.

Figure 6.

Effect of theophylline on Ca²⁺ influx by measuring Mn²⁺ quenching of fura-2 fluorescence in AML12 cells. Experiments were performed in Ca²⁺-containing medium. MnCl₂ (50 μM) was added to cells 1 min before fluorescence measurements. The Y axis is fluorescence intensity (in arbitrary units) measured at the Ca²⁺-insensitive excitation wavelength of 360 nm and the emission wavelength of 510 nm. Trace a: control, without theophylline. Trace b: theophylline (100 μM) was added as indicated. Data are mean ± SEM of three independent experiments.

Discussion

Ca²⁺ signaling plays a pivotal role in triggering and regulating physiological responses in most cell types [8,12]. Previous studies have shown that theophylline affects Ca²⁺ signaling and its related physiology in pig oocytes [14] and porcine tracheal smooth muscles [15]. However, the effect of theophylline on Ca²⁺ homeostasis in cultured liver cell models is still unknown. This is the first study to explore whether theophylline induced [Ca²⁺]_i rises in AML12 cells. The data show that theophylline (25–100 μM) evoked concentration-dependent [Ca²⁺]_i rises in AML12 cells. In terms of Ca²⁺ homeostasis, theophylline evoked [Ca²⁺]_i rises by causing Ca²⁺ influx from extracellular medium and depleting intracellular Ca²⁺ stores because removing extracellular Ca²⁺ partly reduced theophylline-induced [Ca²⁺]_i rises. Removal of extracellular Ca²⁺ decreased theophylline-induced response throughout the measurement period, suggesting that Ca²⁺ influx occurred during the recording period of 220 s.

In terms of cytotoxic response, our findings show that theophylline (25–125 μM) caused cytotoxicity in AML12 cells at a concentration range comparable to that of evoking [Ca²⁺]_i rises. This cytotoxic effect appears to be associated with preceding [Ca²⁺]_i rises in AML12 cells because BAPTA/AM pretreatment inhibited theophylline-induced [Ca²⁺]_i rises by partially reversing cytotoxicity. A link has been established between Ca²⁺ overloading and cytotoxicity [8,12]. Since theophylline induced Ca²⁺-associated cell death in AML12 cells, it might interfere with numerous Ca²⁺-sensitive processes that orchestrate to alter physiology of AML12 cells [8,12].

The mechanism of Ca²⁺ release induced by theophylline was explored in AML12 cells. Regarding the Ca²⁺ stores involved in theophylline-induced Ca²⁺ release, the TG-sensitive endoplasmic reticulum stores seemed to be the dominant one, because TG abolished theophylline-induced Ca²⁺ release. One possible mechanism is that theophylline acts similarly to TG by inhibiting the endoplasmic reticulum Ca²⁺-ATP pump [21]. Furthermore, our results show that U73122 most likely suppressed ATP-induced [Ca²⁺]_i rises via inhibiting PLC activity. However, the data show that incubation with 2 μM U73122 did not alter 100 μM theophylline-induced [Ca²⁺]_i rises. Therefore, it suggests that theophylline-induced Ca²⁺ release was through a PLC-independent mechanism, given the Ca²⁺ release was not altered when PLC activity was inhibited by U73122. PLC-independent Ca²⁺ release pathways may include NADPH oxidase pathway and phospholipase A₂ pathway [30]. Previous studies showed that phospholipase A₂/NADPH oxidase may collaborate to regulate Ca²⁺ release [31]. Therefore, theophylline-induced Ca²⁺ release in AML12 cells deserves further assessment.

Previous researches have shown that the store-operated Ca²⁺ channels played a key role in cultured cell models [32,33]. In our study, theophylline appears to cause Ca²⁺ entry via store-operated Ca²⁺ entry which is induced by depletion of intracellular Ca²⁺ stores [33], based on the inhibition of theophylline-induced [Ca²⁺]_i rises by 2-APB, econazole, and SKF96365. These three compounds have often been used as blockers of store-operated Ca²⁺ entry in different cells [24–27], although none of them exerts selective inhibition. Because theophylline-induced [Ca²⁺]_i rises were significantly inhibited by ~30% by 2-APB, econazole, or SKF96365, which is also similar to the magnitude of theophylline-induced Ca²⁺ influx, it suggests that store-operated Ca²⁺ entry was involved in theophylline-induced Ca²⁺ influx in AML12 cells.

Because activation of PLC produces IP₃ and diacylglycerol, which activates PKC [11,12], the effect of modulation of PKC activity on theophylline-induced [Ca²⁺]_i rises was examined. Activation or inhibition of PKC failed to affect theophylline-induced [Ca²⁺]_i rises. This suggests that PKC did not play a role in theophylline-induced [Ca²⁺]_i rises. The data further show that theophylline induced Mn²⁺ entry. Mn²⁺ enters cells through similar mechanisms as Ca²⁺ but quenches fura-2 fluorescence at all excitation wavelengths [19]. Therefore, quenching of fura-2 fluorescence excited at the Ca²⁺-insensitive excitation wavelength of 360 nm by Mn²⁺ implicates Ca²⁺ influx.

In previous studies, theophylline at concentrations between 500 and 2500 μM caused cytotoxicity in MDA-MB-231 cells [5], KB cells, and H1355 cells [6]. Our study shows that 25–125 μM theophylline concentration-dependently inhibited cell viability in AML12 cells. Therefore, the effect of theophylline on cytotoxicity may depend on cell types and concentrations. Several studies were performed to explore the plasma level of theophylline in animal or adults [34–36]. The plasma level of theophylline may reach ~ 50 μM [34–36]. Furthermore, the recommended plasma level of theophylline is 55–110 μM in humans [16]. This level may go much higher in patients with liver disorders or taking higher doses. Previous studies suggested that because theophylline in the blood requires active transport into the liver, it is metabolized by the cytochrome P450 (CYP) family [35,36]. Therefore, theophylline might be excreted through liver in human. Our data show that theophylline at a concentration of 25 μM induced slight cell death and the IC₅₀ (half maximal inhibitory concentration) value of theophylline was ~62.5 μM in AML12 cells. Thus, our data may be clinically relevant in some groups of patients.

There are some limitations in this study. First, the evaluation of the toxicological effect of theophylline was restricted to AML12 mouse hepatocytes, and thus other human hepatocytes or liver tissues could also be employed to examine the toxicological effect. Second, the toxicity of theophylline to liver function was not evaluated in an animal model. Because an in vitro study cannot perfectly mimic an in vivo exposure, in the future, our research will expand to the in vivo toxicity of theophylline.

Conclusions

In AML12 cells, the results show that theophylline induced Ca²⁺-associated cytotoxicity, released Ca²⁺ from endoplasmic reticulum via a PLC-independent fashion, and induced Ca²⁺ entry via a PKC-insensitive, store-operated Ca²⁺ entry. This study provides a basis for further studies of the role of theophylline in human health management, and a strong rationale for clinical evaluation of this drug. Because our data show that BAPTA-AM significantly prevented theophylline-induced cytotoxicity in AML12 cells, future studies are needed to explore various biological or synthetic compounds to effectively chelate intracellular Ca²⁺ for maintaining human liver function in in vivo studies.

Author contributions

G.C.S. and W.Z.L. wrote the manuscript. W.Z.L. performed the experiments. All authors reviewed and edited the manuscript before submission.

Conflict of interest

There are no conflicts of interest to declare.