Cellular and Mitochondrial Toxicity of Tolcapone, Entacapone, and New Nitrocatechol Derivatives

Abstract

Nitrocatechols are the standard pharmacophore to develop potent tight-binding inhibitors of catechol O-methyltransferase (COMT), which can be used as coadjuvant drugs to manage Parkinson’s disease. Tolcapone is the most potent drug of this class, but it has raised safety concerns due to its potential to induce liver damage. Tolcapone-induced hepatotoxicity has been attributed to the nitrocatechol moiety; however, other nitrocatechol-based COMT inhibitors, such as entacapone, are safe and do not damage the liver. There is a knowledge gap concerning which mechanisms and chemical properties govern the toxicity of nitrocatechol-based COMT inhibitors. Using a vast array of cell-based assays, we found that tolcapone-induced toxicity is caused by direct interference with mitochondria that does not depend on bioactivation by P450. Our findings also suggest that (a) lipophilicity is a key property in the toxic potential of nitrocatechols; (b) the presence of a carbonyl group directly attached to the nitrocatechol ring seems to increase the reactivity of the molecule, and (c) the presence of cyano moiety in double bond stabilizes the reactivity decreasing the cytotoxicity. Altogether, the fine balance between lipophilicity and the chemical nature of the C1 substituents of the nitrocatechol ring may explain the difference in the toxicological behavior observed between tolcapone and entacapone.

Catechol-O-methyltransferase (COMT, EC 2.1.1.6) is an intracellular magnesium-dependent methyltransferase responsible for the metabolization/methylation of catechol substrates such as endogenous neurotransmitters (dopamine, epinephrine or norepinephrine), hormones, and xenobiotics with a catecholic structure substrates.^1,2 In Parkinson’s disease (PD), COMT inhibitors are used as adjunctive therapy to improve motor-fluctuation related to levodopa therapy: inhibition of COMT avoids premature levodopa metabolization by COMT and increases the dopamine pool in the brain.³ Tolcapone (Figure 1) was the first FDA-approved COMT inhibitor to reach the market. During clinical trials, up to 3.7% of the individuals tested showed increased levels of aspartate and alanine aminotransferase in the serum, but no serious hepatotoxic events were reported. As so, tolcapone was considered safe and its approval for commercialization came right after.^4,5 Upon market introduction in 1998, three fatal cases of fulminant hepatitis following tolcapone treatment were reported,⁴ leading to its withdrawal. Later on, adjustments on dosages were done and due to the lack of alternatives, tolcapone was reintroduced in the market with several warnings and guidelines regarding its use.⁵ Ever since, two other COMT inhibitors were approved for PD management: entacapone (1999) and opicapone (2020). While providing a significant increase in safety, these drugs are still outperformed by tolcapone in terms of clinical effectiveness in the control of motor fluctuations during levodopa therapy, providing an off-time reduction of about −0.61 and −1 h/day, respectively.⁶

Figure 1.

Chemical structures of nitrocatechol-based catechol O-methyltransferase inhibitors: tolcapone, entacapone, and opicapone.

The clinical and histological findings in patients hospitalized after tolcapone treatment were consistent with drug-induced hepatotoxicity.⁷ The exact mechanism of tolcapone-induced hepatotoxicity is unclear, but previous work suggests the formation of toxic reactive metabolites and protein adducts,⁸ redox cycling and oxidative stress related events, and uncoupling of the mitochondrial respiratory chain.⁷ All of these reports flag the nitrocatechol moiety as an effector of tolcapone-induced toxicity and identify it as a toxicophore. However, the lack of toxicity observed so far for entacapone and opicapone – also nitrocatechol derivatives – and our recent work on nitrocatechol-based COMT inhibitors⁹ does not support this hypothesis. Recently, we hypothesized that the lipophilicity of the chemical structure may be a significant molecular descriptor of drug-induced toxicity. To address this hypothesis, we first performed an array of cell-based assays for tolcapone and entacapone (Figure 1) using different cellular systems (primary rat hepatocytes, Caco-2 and HepG2 cells) exposed to increasing concentrations of tolcapone and entacapone (0, 1, 10, and 50 μM) for 24 h. To compare the mechanisms of cytotoxicity of tolcapone vs entacapone, we also assessed the potential role of oxidative stress, P450 metabolism, P-gp efflux, and mitochondrial function. Following this first comparison, we next investigated the influence of lipophilicity in the cytotoxicity, by repeating the same screening protocols using our in-house nitrocatechol library.⁹ Our overarching aim was to understand the chemical descriptors that are behind the cytotoxic differences of tolcapone and entacapone. To the best of our knowledge, this is the first report focusing on the interplay between nitrocatechol chemistry and toxicity.

Methods

Chemicals and Reagents

All reagents used were of analytical grade or the highest grade available. Tolcapone and entacapone were acquired from Amadis Chemical Co., Ltd. (Hangzhou, China). Caffeic acid (1), neutral red (NR) solution, (4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium (MTT) bromide, trypan blue solution [0.4% (w/v)], and Dulbecco’s modified Eagle’s medium (DMEM) with 4.5 g/L glucose were obtained from Sigma-Aldrich (St. Louis, MO, USA). Reagents used in cell culture, including nonessential amino acids (NEAA), heat inactivated fetal bovine serum (FBS), 0.05 or 0.25% trypsin/1 mM EDTA, antibiotic (10000 U/mL penicillin 10000 μg/mL streptomycin), phosphate-buffered saline solution (PBS), and Hank’s balanced salt solution (HBSS) were purchased from Gibco Laboratories (Lenexa, KS, USA). Dimethyl sulfoxide (DMSO), absolute ethanol, and acetic acid were obtained from Merck (Darmstadt, Germany). Compounds 1–12 were obtained using straightforward synthetic methodologies and were characterized by nuclear magnetic resonance spectroscopy and mass spectrometry. The full description of the synthetic methodologies used and the spectroscopic data have been described elsewhere.⁹

Cell Culture

Primary Rat Hepatocytes

Rat hepatocytes were isolated from 8- to 9-week-old male Wistar Han rats (200–250 g) purchased from Charles-River Laboratories (Barcelona, Spain). Surgical procedures were performed under anesthesia by inhalation of isoflurane in an isolated system and carried out between 10.00 and 11.00 a.m. Cells were isolated through collagenase perfusion. Briefly, a cannula was inserted in the hepatic portal vein, and the liver was perfused initially with Hank’s washing buffer containing BSA and the chelating agent EGTA, followed by a solution of collagenase supplemented by its cofactor calcium. The liver capsule was then gently disrupted to release isolated liver cells into a Krebs–Henseleit buffer. The cell suspension was subsequently purified through three cycles of low-speed centrifugation (300 rpm for 2 min). The final suspension was then incubated with penicillin/streptomycin (500 U/mL/500 μg/mL), at 4 °C, for 30 min. Cell viability was estimated by the trypan blue exclusion test and was always higher than 80%. Then, a suspension of 500000 viable cells/mL was cultured in 96-well plates at approximately 100000 cells/cm², in William’s E medium, supplemented with 10% FBS (100 U/mL), penicillin (100 U/mL), streptomycin (100 μg/mL), insulin (5 μg/mL), dexamethasone (50 μM), gentamicin (100 μg/mL), and fungizone (2.5 μg/mL) and incubated overnight at 37 °C, with 5% CO₂, to allow cell adhesion. 24 h after seeding at 100000 cells/cm², the cells were exposed to the test compounds (0, 1, 10, and 50 μM) in cell culture medium without FBS for 24 h. At the selected time point, cytotoxicity was evaluated by MTT reduction and NR uptake assays.

Caco-2 Cells

Caco-2 cells (ATCC-HTB-37, lot: 70046148) were routinely cultured in 75 cm² flasks using DMEM with 4.5 g/L glucose, supplemented with 10% heat inactivated FBS, 100 μM NEAA, 100 U/mL penicillin, 100 μg/mL streptomycin, 2.5 μg/mL fungizone, and 6 μg/mL transferrin. The cells were maintained in a 5% CO₂–95% air atmosphere at 37 °C, and the medium was changed every 2 days. Cultures were passaged weekly by trypsinization (0.25% trypsin/1 mM EDTA). The cells used in all the experiments were taken between the 58th and 63rd passages. In all experiments, cells were seeded onto 96-well plates (60000 cells/cm²) and used 3 days after seeding, when confluence was reached. The cells were exposed to the test compounds (0, 1, 10, and 50 μM) in cell culture medium without FBS for 24 h. At the selected time point, cytotoxicity was evaluated by MTT reduction and NR uptake assays.

HepG2 Cells

Human HepG2 cells (ATCC, ATCC-HB-8065, lot: 70047955) were routinely cultured in 75 cm² flasks using DMEM with 25 mM glucose, supplemented with 6 mM of glutamine, 5 mM HEPES, 44 mM sodium bicarbonate, 1 mM sodium pyruvate, 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin, and 2.5 μg/mL Amphotericin B. For the experiments performed in oxidative phosphorylation-forcing media, HepG2 cells were previously adapted to the OXPHOS media, composed by basal DMEM without glucose, supplemented with 10 mM galactose, 6 mM glutamine, 5 mM HEPES, 44 mM sodium bicarbonate, 1 mM sodium pyruvate, 10% FBS, penicillin (100 U/mL), streptomycin (100 μg/mL), and Amphotericin B (2.5 μg/mL). For that, during three passages of HepG2 cells, the amount of galactose in the cellular medium was progressively increased until residual amounts of glucose were present in the culture media. The cells used in all of the experiments were between the 20th and 30th passages. For the experiments with the Seahorse XF96 extracellular flux analyzer, cells were seeded in specific plates at a density of 20000 cells/well and used 24 h after seeding. The cells were exposed to the test compounds (0, 10, and 50 μM) in cell culture medium, for 24 h.

Cell Viability Assays

Incubation with Test Compounds

The cells were exposed to the test compounds at three different concentrations (1, 10, and 50 μM) for 24 h, and cellular viability was evaluated by the MTT reduction, and the NR uptake assays, compared to control untreated cells (% of control, n ≥ 4). Stock solutions of the test compounds (50 mM) were prepared in DMSO, and each compound solution was diluted in cell medium to reach the desired well concentration. The concentration of DMSO per well was always lower than 0.1%.

2,5-Diphenyl-2H-tetrazolium Bromide (MTT) Reduction Assay

The MTT reduction assay was used to measure mitochondrial dysfunction (decrease in mitochondrial dehydrogenase activity) in cells exposed to the test compounds. The change in absorbance generated is dependent on the degree of reduction of the MTT tetrazolium salt (water-soluble) to MTT formazan (water-insoluble) by cellular dehydrogenases within metabolically active cells. At the selected time point, the culture medium was replaced by fresh cell culture medium containing 0.5 mg/mL MTT, and the cells were incubated at 37 °C in a humidified, 5% CO₂ atmosphere for 2 h. After this incubation period, the cell culture medium was removed, and the formed formazan crystals were dissolved in 100% DMSO. The absorbance was measured at 550 nm in a multiwell plate reader (BioTek Instruments, Vermont, USA). The percentage of MTT reduction relative to that of the control cells was used as a viability measure [MTT reduction (% of control)].

Neutral Red (NR) Uptake Assay

The NR uptake assay is based on the ability of viable cells to incorporate and bind the supravital dye neutral red in lysosomes, thus providing a quantitative estimation of the number of viable cells. 24 h after exposure, the cell culture medium was removed, and the cells were incubated with neutral red (50 μg/mL in cell culture medium) at 37 °C, in a humidified, 5% CO₂–95% air atmosphere, for 2 h. After this incubation period, the cell culture medium was removed, the dye absorbed only by viable cells extracted [with absolute ethyl alcohol/distilled water (1:1) containing 5% acetic acid], and the absorbance measured at 540 nm in a multiwell plate reader (BioTek Instruments, Vermont, USA). The percentage of NR uptake relative to that of the control cells was used as a viability measure [NR uptake (% of control)].

Co-incubations with P450 and P-gp Inhibitors

Cell culture and seeding were performed for rat hepatocytes and Caco-2 cells, as described above. When confluence was reached, the cells were preincubated with the P450 inhibitor 1-aminobenzotriazol (ABT, 100 μM) or the P-gp inhibitor zosuquidar (ZOS, 5 μM) in cell culture medium without FBS for 1 h at 37 °C and 5% CO₂ and then treated with the test compounds (final concentration: 50 μM). Within the same plate, parallel column wells were treated with either medium with vehicle, ABT (100 μM), ZOS (5 μM), or the test compound (50 μM). The cells were then incubated for 24 h at 37 °C and 5% CO₂, and cellular viability was determined by the MTT reduction and NR uptake assays.

Oxidative Stress Measurements

2′,7′-Dichlorofluorescein diacetate (DCFH-DA) is a nonfluorescent and membrane-permeable compound. Once in the cytoplasm, esterases remove the acetate groups to produce 2′,7′-dichlorodihydrofluorescein (DCFH) which, because of its polarity, is not cell-permeable. DCFH is easily oxidized to 2′,7′-dichlorofluorescein (DCF), a highly fluorescent compound (excitation 485 nm: emission 530 nm), by several reactive oxygen species (ROS), including hydrogen peroxide, hydroxyl radicals, and nitrogen dioxide.¹⁰ On the days of the experiment, the cells seeded at the appropriate density were preincubated with 10 μM DCFH-DA for 30 min. At the end of the incubation period, excess dye was rinsed twice with HBSS and the cells were incubated with the test compounds. For each plate, blanks (vehicle and cell culture medium without the cells) and negative controls (cells treated with medium without test compounds) were run to measure background fluorescence and basal dye oxidation, respectively. DCFH-DA fluorescence was measured at the selected time-point using a fluorescence plate reader (BioTek SynergyTM HT) set at 485 nm excitation and 530 nm emission.

Oxygen Consumption Rate Measurements in HepG2 Cell Line

Cellular oxygen consumption rate and extracellular acidification was measured at 37 °C using the platform Seahorse XF96 extracellular flux analyzer (Seahorse Bioscience, Agilent). For each experiment, HepG2 cells were seeded at a density of 20 × 10³ cell/well and incubated at 37 °C in a 5% CO₂ humidified atmosphere for 24 h prior the treatment. Following, the cells were exposed to the test compounds (0, 10, and 50 μM) in cell culture medium for more 24 h. The day prior the assay, the XF96 sensor cartridge was hydrated in 200 μL/well of calibration buffer at 37 °C. For each assay, the cell culture medium was replaced by 175 μL of prewarmed, low-buffered serum-free DMEM, supplemented with 25 mM glucose, or 10 mM galactose, 1 mM pyruvate, and 6 mM glutamine, with pH adjusted to 7.4. Just before the assays, HepG2 cells were incubated at 37 °C for 60 min in the assay medium to allow for temperature and pH to reach equilibrium before the first-rate measurement. For mitochondrial respiration, oligomycin, FCCP, and rotenone/antimycin were prepared as an 8x, 9x, and 10x stock in assay medium. The inhibitors were preloaded into the ports of each well in the XF96 sensor cartridge. Half of the plate was used to evaluate the effects of the compounds in cells cultured in high-glucose medium, and the other half was used to test the effects of the compounds in HepG2 cells cultured in. The sensor cartridge and calibration buffer were loaded into the XF96 extracellular flux analyzer to calibrate the cartridge. When the calibration was complete, the calibration plate was replaced with the cell plate, and the assay started. Three baseline rate measurements of the oxygen consumption rate (OCR) of cells were made using a 3 min mix and 3 min measure cycle. The inhibitors were then injected by the XF96 analyzer into each well and mixed. OCR measurements were made using the 3 min mix and 5 min measure cycle. Results were analyzed by the XF Stress Test Report Generators from Agilent. Cellular normalization was performed by measuring cellular protein, using the SRB method as described by our group.¹¹

Evaluation of Drug Lipophilicity

The lipophilicity of the drugs is directly correlated with the chromatographic hydrophobicity index (CHI) under study and was determined on a NEXERA-i LC-2040C ultrahigh-performance liquid chromatograph (RP-UHPLC) (Shimadzu, Kyoto, Japan) equipped with a diode array detector and controlled by the LabSolution system (version 5.90, Shimadzu).

The chromatographic hydrophobicity index (CHI) values at pH 2.3 were determined using an experimental protocol described elsewhere.¹² The acidic conditions used were relevant to prevent ionization of the phenolic function and to avoid peak tailing. The CHI values were assessed from experimental retention times (t_R) of the samples and a mixture of reference compounds obtained using a Luna C18 (2) column (150 × 4.6 mm, 5 μm, Phenomenex, CA, USA). Stock solutions of compounds in DMSO (10 mM) were diluted in acetonitrile:water (1:1) mixture to obtain a final concentration of 250 μM. The mobile phase A was aqueous acetic acid 1% (v/v) (pH 2.3), and the mobile phase B was acetonitrile. The following gradient program was applied: 0–6 min 0–100% B, 6–14 min 100% B, 14–16 min 100–0% B. The flow rate was 1 mL/min, and the injection volume was 20 μL. The system was calibrated using known standards with reported values of CHI (see Figure S4 and Table S1). The results are expressed as mean values of three experiments.

Ethics

All animal experiments were approved by the Ethics Committee of the Faculty of Pharmacy of the University of Porto (Porto, Portugal) regarding the welfare of experimental animals and performed in accordance with the national legislation. All procedures performed in studies involving animals were in accordance with the ethical standards of the national legislation and approved by the Ethics Committee of the Faculty of Pharmacy of the University of Porto (Porto, Portugal).

Statistical Analysis

For each assay, all experiments were performed in triplicate and the coefficient of variation was always <10%. The data obtained are expressed as mean ± standard deviation (SD) of at least four biologic independent experiments. For cellular studies, normality of the data distribution was assessed by three different tests: KS normality test, D’Agostino and Pearson omnibus normality test, and Shapiro–Wilk normality test. Simple statistical comparisons between groups were made using the parametric method of one-way ANOVA, followed by the Dunn’s post hoc test. Two-way ANOVA followed by Tukey post hoc test was used for multigroup comparison. Statistical significance was set at p < 0.05. All statistical analysis was performed with GraphPad Prism 6 for Windows (GraphPad Software, San Diego, CA, USA).

Results and Discussion

Tolcapone and Entacapone Toxicity Profiles in Primary Rat Hepatocytes

Considering the structural similarity between tolcapone and entacapone, we first studied the toxicity profile of both of the COMT inhibitors in freshly isolated primary rat hepatocytes. As cellular models for liver toxicity studies, primary rat hepatocytes show a good compromise in terms of cell viability and handling, quality of the results, and hepato-specific differentiation and function.¹³

The screening of tolcapone in primary rat hepatocytes showed an abrupt decline (>50%, p < 0.001) in both viability end points for the highest tested concentration (Figure 2). Despite this, we observed that entacapone (50 μM) also decreased MTT reduction by 49% (p < 0.001) and NR uptake by 20% (p < 0.01) (Figure 2). A clear difference in the cytotoxic profiles of the two compounds was noticed at 10 and 50 μM.

Figure 2.

Cytotoxic effects of tolcapone (black bars) and entacapone (red bars) in primary rat hepatocytes, at three different concentrations (1, 10, and 50 μM) and after 24 h of exposure, by measuring the metabolic activity (A) and lysosomal activity (B) by MTT and NR uptake assays, respectively. The hepatocyte metabolic activity was also measured after a pretreatment with P450 inhibitor ABT (100 μM) followed by the exposure with tolcapone (50 μM) (C). The data are expressed as the means of four independent experiments together with the standard error mean (mean ± SEM). Black and orange dot lines represent the control data and the threshold of 85% of cell viability, respectively. Statistical calculations were performed by ANOVA. In all cases, p values lower than 0.05 were considered significant (*p < 0.05, and ****p < 0.0001 compared with control cells; ^##p < 0.01, and ^####p < 0.0001 compared with the data from cells treated with tolcapone).

The observed tolcapone hepatotoxicity led us to hypothesize that a bioactivation process mediated by P450 specifically related to the reductive metabolism of nitrocatechols can occur with the formation of reactive and toxic intermediates.⁸ In fact, it has been proposed that the bioreduction of tolcapone nitro group catalyzed by CYP reductase is a relevant process that leads to the formation of reactive o-quinone and quinone-imine species that further interact with mitochondrial components, leading to hepatocellular damage.¹⁴⁻¹⁶

For that, we preincubated primary rat hepatocytes with a nonselective P450 inhibitor (1-aminobenzotriazole, ABT, 100 μM) for 1 h followed by the treatment with tolcapone 50 μM for 24 h. The cellular metabolic activity at the final end point was measured (Figure 2). Under our experimental conditions, no statistically significant difference was observed between hepatocytes treated with 50 μM tolcapone and ABT 100 μM + 50 μM tolcapone. This result strengthens our previous hypothesis related to a putative direct toxic effect of unmetabolized tolcapone, which does not require metabolic bioactivation by the P450. These results are in line with a preceding observation that the inhibition of the reductive pathway in primary rat hepatocytes did not prevent tolcapone-induced toxicity.¹⁷ In addition, the analysis of tolcapone metabolism and excretion reported that glucuronidation is the main route for tolcapone metabolism and oxidative reactions mediated by P450 enzymes are of small significance.¹⁶ Indeed, in a clinical report of two cases of tolcapone-induced asymptomatic hepatic dysfunction, patients exhibited abnormal activity of UDP-glucuronosyltransferase (UGT), suggesting that specific genotypes of the UGT1A9 gene may be a predisposing factor for tolcapone-induced hepatotoxicity.¹⁸ Together, the results point toward an intrinsic tolcapone toxicity or at least, its toxicity does not depend on P450 or UGT metabolic pathways.

Tolcapone and Entacapone Toxicity Profiles in Caco-2 Cells

To gain insight into tolcapone and entacapone toxicity, an intestinal cellular line (Caco-2) was further used to complement the previous data. In Caco-2 cells, tolcapone induced a decrease in MTT reduction (50 μM, viability = 68.0 ± 6.7%) and lysosomal activity (50 μM, viability = 86.5 ± 3.5%), which was not observed for entacapone at the same concentration (viability in both assays >96%, Figure 3).

Figure 3.

Cytotoxic effects of tolcapone (black data) and entacapone (red data) in Caco-2 cells, at three different concentrations (1, 10, and 50 μM) and after 24 h of exposure, by measuring the metabolic activity (A) and lysosomal activity (B) by MTT and NR uptake assays, respectively. The metabolic activity of Caco-2 cells was also measured after a pretreatment with P450 inhibitor ABT (100 μM) or P-gp inhibitor ZOS (5 μM) followed by exposure with tolcapone (50 μM) (C). The data are expressed as the means of four independent experiments together with the standard error mean (mean ± SEM). Black and orange dotted lines represent the control data and the threshold of 85% of cell viability, respectively. Statistical calculations were performed by ANOVA. In all cases, p values lower than 0.05 were considered significant (*p < 0.05, ***p < 0.001 and ****p < 0.0001 compared with control cells; ^##p < 0.01, ^###p < 0.001, and ^####p < 0.0001 compared with the data from cells treated with tolcapone; ⁺⁺p < 0.01 compared with tolcapone treatment alone).

Tolcapone cytotoxic effects observed in this cell system were less evident than those obtained using hepatocytes. Even though we check in this cell system if P450 metabolism can act as detoxifying mechanism,¹⁷ a pretreatment of cells with ABT (100 μM) was also performed. In accordance with data from hepatocytes, pretreatment of the cells with ABT (100 μM) (Figure 3) did not affect the effect of tolcapone at 50 μM (tolcapone = 68.0 ± 6.8% vs tolcapone + ABT = 68.9 ± 7.7%), which clearly indicates that P450 activity does not modulate tolcapone-induced toxicity at the enterocyte level.

As Caco-2 cells are considered a suitable cell line model for P-glycoprotein (P-gp)-mediated transport studies,¹⁹ we also assessed the potential role this transmembrane glycoprotein in reducing intracellular concentration of tolcapone, by using a P-gp inhibitor (ZOS at 5 μM). The inhibition of P-gp with ZOS (5 μM) led to a significant decrease in metabolic activity (tolcapone = 68.0 ± 6.7% vs tolcapone + ZOS = 54.1 ± 9.1%, p < 0.01). The results suggest that tolcapone is likely to be a P-gp substrate in Caco-2 cells. P-gp substrates are known to be small lipophilic, basic nitrogen-containing compounds capable of forming hydrogen bonds.²⁰ The data flagged the potential role of P-gp efflux as a detoxifying mechanism for tolcapone, which is similar to entacapone which is a substrate of breast cancer resistance protein.²¹ The results are in agreement with the hypothesis proposed by other authors that tolcapone is a substrate of efflux transporters.^21,22

Tolcapone and Entacapone Toxicity Profiles in HepG2 Cell Line

The evaluation of the toxicity of tolcapone and entacapone was next complemented by using the human HepG2 cell line, being an in vitro protocol frequently used in safety assessment of drugs hepatotoxicity²³ aiming to identify the relevance of mitochondria in this process. With this objective, a screening was conducted using HepG2 cells, cultured in glucose- or galactose-containing medium, by measuring their metabolic and lysosomal activities as well as the amount of ROS produced after drug treatment (Figure 4).

Figure 4.

Screening of cytotoxic effects of tolcapone (black bars) and entacapone (red bars) in HepG2 cells after 24 h of treatment with three different concentrations (1, 10, and 50 μM) in medium enriched with galactose versus glucose. Cells grown in a high-glucose-containing medium use glycolysis for ATP generation while cells grown in a galactose-containing medium rely almost exclusively on mitochondria for their ATP production (A). The cell viability was determined by measuring the metabolic activity (B and E) and lysosomal activity (C and F) by MTT and NR uptake assays, respectively. The measurement of intracellular reactive species (RS) in HepG2 cells after 24 h treatment with the compounds was normalized with the metabolic (D and G) for both medium. The data related to cell cultured in high-glucose- and galactose-containing medium is presented in panels (B–D) and (E–G), respectively. In panels (B–G), the control data is represented by a black dotted line, while an orange dotted line represents the 85% threshold of cell viability. The data are shown as the means of four independent experiments together with the standard error mean (mean ± SEM). Statistical calculations were performed by ANOVA. In all cases, p values lower than 0.05 were considered significant (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 compared with control cells; ^##p < 0.01, ^###p < 0.001, and ^####p < 0.0001 compared with the data from cells treated with Tolcapone).

We used two different cell culture media because cells cultured in galactose are unable to generate sufficient ATP from glycolysis; therefore, they are forced to rely on mitochondrial oxidative phosphorylation (OXPHOS) for ATP generation. This makes those cells more sensitive to mitochondrial perturbation than cells grown in glucose media (Figure 4A).²⁴ Fundamentally if a test compound impairs the mitochondrial capacity to synthesize ATP, its toxicity shifts toward significantly lower concentrations in this modified media.²⁵ It is assumed that the cost of additional ATP consumption in the Leloir pathway renders glycolysis less efficient with galactose, and this places greater demands on mitochondrial ATP production. The increased reliance on mitochondrial ATP synthesis is presumed to explain the enhanced sensitivity for detecting inhibition or impairment to mitochondria in HepG2 incubated with the galactose-modified media.²⁵

Starting with the HepG2 cells cultured with a medium with a high concentration of glucose (25 mM), it was possible to observe a depletion of cell viability after tolcapone’s treatment in a dose-dependent manner (Figure 4B,C). When HepG2 cells were exposed to tolcapone for 24 h, a significant decrease of metabolic activity was observed for all concentrations tested, which was more severe for the concentrations 10 μM (80.9 ± 4.1%) and 50 μM (49.3 ± 2.7%). The same tendency was observed in the data obtained from NR uptake. All concentrations significantly decreased the lysosomal activity of cells, reaching as low as 54.7 ± 5.7% (p < 0.0001) when tolcapone was tested at 50 μM (Figure 4). For entacapone, a reduction of the metabolic activity at all tested concentrations (metabolic activity >85.9%) and a reduction in the lysosomal activity at the highest concentration tested (50 μM, NR uptake >91.1%) was observed. Nevertheless, the results of the measured cell viability parameters were higher than 85%, indicating a putative safety profile.

Furthermore, we investigated the role of oxidative stress in the toxicity of both drugs using the fluorescent ROS dye DCFH, usually administered in its diacetate (DCFH-DA) form, which permeates cells (Figure 4D). Therefore, the DCFH fluorescence was normalized with the metabolic and lysosomal activities. Generally, tolcapone led to an increase of oxidative stress, mainly for the concentration of 50 μM (158.0 ± 21.7% or 143.6 ± 25.1%, values normalized with MTT or NR results, respectively), which did not occur with entacapone (p < 0.0001). In fact, the treatment with this drug did not change the oxidative stress levels compared with the control cells for all concentrations tested.

These findings were even more drastic when HepG2 cells were cultured with galactose medium (Figure 4E–G) and treated with tolcapone or entacapone under the same conditions as described above. In OXPHOS conditions, the harmful effect of tolcapone in HepG2 cells was increased, with a clear decrease in metabolic activity at 10 μM (to 45.2 ± 6.4%) and 50 μM (to 5.6 ± 1.1%). This type of effect was also observed in the lysosomal activity with a decrease of cell viability to values of 68.4 ± 5.6% and 15.0 ± 4.1%, respectively. Interestingly, entacapone treatment showed cell viability values higher than 86%. These data are significantly different when compared with tolcapone mainly for the two highest concentrations, suggesting that tolcapone per se induced mitochondrial toxicity.

After incubation of cells with the tolcapone, an increase (>114%) in the dye fluorescence was measured for all concentrations tested, which was mainly observed at 50 μM with a 12 times higher ROS levels normalized with metabolic activity (Figure 4G) when compared with control cells. From the data, we can conclude that tolcapone can produce a change in the cellular redox state. Contrarily, the treatment of HepG2 cells with entacapone caused a decrease in oxidative stress levels compared with that of control cells for the concentrations of 10 and 50 μM.

From the data obtained using HepG2 as an in vitro model, it is important to highlight the significant difference between the results from cells exposed to tolcapone with those from the cellular treatment with entacapone mainly for the two highest concentrations (10 and 50 μM) tested. Moreover, a clear difference of the data obtained after tolcapone treatment in the two different mediums used allowed us to suggest a clear role of mitochondria in tolcapone hepatotoxicity events.²⁶

Modulation of Mitochondrial Bioenergetics

The data obtained so far pointed to a direct mitochondrial toxicity of tolcapone as the main pathway. To study the effects of tolcapone and entacapone (50 μM) in mitochondria, we used human HepG2 cells, due to their increased susceptibility to drug-induced mitochondrial alterations.²⁷ We evaluated key parameters of mitochondrial function by measuring the cellular oxygen consumption rate (OCR) after inducing mitochondrial stress with specific respiration modulators.²³

The results depicted in Figure 5A,B are in accordance with the data collected so far showing that tolcapone causes more harmful effects at the mitochondrial level than entacapone.²⁸

Figure 5.

Mitochondrial parameters of HepG2 cells exposed to tolcapone and entacapone 50 μM for 24h, cultured in high-glucose (A) and galactose media (B) and related analyses of bioenergetic parameters (C). In panel (C), solid and squared bars represent data from experiments performed in high-glucose and OXPHOS media, respectively, while black and red data represent results from the cell treatment with tolcapone and entacapone, respectively. OCR was measured using a Seahorse XF96 extracellular flux analyzer. The data are expressed as the means of four independent experiments together with the standard error mean (mean ± SEM). The black line represents the control data. Statistical calculations were performed by two-way ANOVA. In all cases, p values lower than 0.05 were considered significant (*p < 0.05 and ****p < 0.0001 compared with control cells; ^##p < 0.01 compared with the data from cells treated with tolcapone for the same culture medium used; ⁺p < 0.05, ⁺⁺⁺p < 0.001 comparing cells treated with same drug but with different cell culture medium).

Tolcapone significantly decreased maximal respiration (by 55%, p < 0.0001), spare respiratory capacity (by 80%, p < 0.0001), and nonmitochondrial respiration (by 25%, p < 0.5) when incubated at 50 μM for 24 h in HepG2 cells cultured in high-glucose medium (Figure 5C). The significant decrease in the maximal respiration induced by tolcapone suggests an inhibition of mitochondrial respiratory complexes or substrate supply/oxidation, consequently reducing the capacity of cells to respond to an energetic demand. This effect was not so accentuated for entacapone under the same experimental conditions. In terms of maximal respiration, HepG2 cultured in a high-glucose medium and treated with entacapone only presented a decrease of 19% compared with control cells, which is a clear difference for tolcapone data (p < 0.01). The inhibition of complexes I, II, and IV for tolcapone and of complexes I and IV for entacapone was directly demonstrated in permeabilized mouse liver mitochondria by Grunig and colleagues, which ultimately led to a higher toxicity of tolcapone.²⁸ A slight difference in the OCR related with nonmitochondrial respiration was noted for cells treated with tolcapone in high-glucose medium (73.9 ± 12.8%), which could be related with a possible inhibition of NADPH oxidases.²⁹

The major difference from control cells was observed in spare respiratory capacity of the cells mainly when treated with both COMT inhibitors. It has been suggested that spare respiratory capacity is the most sensitive parameter affected by a mitotoxicant.³⁰ Compared with the decrease of 80% caused by tolcapone in spare respiratory capacity, the treatment with entacapone only led to a depletion of 54% (p < 0.0001) of this bioenergetic parameter in the HG medium. A reduction of the spare respiratory capacity may occur when the ATP demand exceeds the maximal ATP supply from glycolysis and oxidative phosphorylation, thereby implicating the spare respiratory capacity critical for maintaining ATP generation under conditions of increased demand.³¹

In fact, despite that nonsignificative when compared with control cells, tolcapone (50 μM) led to a decrease of the OCR associated with ATP production and basal respiration in both media tested (Table S2). Contrarily, entacapone (50 μM) treatment led to a slight increase of the level of OCR related to basal respiration and ATP production in high-glucose medium and a decrease of the level of proton leak-associated OCR when cells were treated in galactose medium (Table S2).

Meanwhile, we also measured the OCR of HepG2 with both drugs at 10 μM and the data presented in Figure S2 (supporting information) showed no significant alteration when compared with control cells.

Under the current experimental conditions (replacing glucose with galactose as the major source of carbon and supplementing the media with an excess of pyruvate and glutamine), we expected a higher susceptibility of HepG2 to mitochondrial toxicants; however, that was not observed in the Seahorse experiments (Figure 5). We suggest that this occurred by one of two mechanisms, or both: (a) another mechanism besides mitochondrial toxicity could play a role in the tolcapone hepatotoxicity or (b) culturing HepG2 cells in the modified galactose media may result in increased mitochondrial biogenesis and lower accumulation of drug per mitochondrial mass. In fact, we observed an increased in mtDNA copy number, an increase in the NDUFB8 subunit of complex I, an increase in the subunit of MTCO1 of complex IV, and an increase in the subunit. As such, the Seahorse data do not rule out that tolcapone directly causes mitochondrial toxicity, which may be observed at earlier time points.

Tolcapone, Entacapone, and Derivatives Thereof Lipophilicity

Several groups have shown tolcapone, but not entacapone, is a potent uncoupler of oxidative phosphorylation in in vitro and in vivo.^7,32 Although the uncoupling effect was directly linked to the nitrocatechol moiety,^9,33 which works as a protonophore that dissipates the proton gradient in mitochondria,³⁴ no effect was observed for entacapone. In general, the uncoupling activity of protonophores increases with lipophilicity.³⁵ After the first mechanistic studies, we investigated lipophilicity as a potential mechanism for the different toxicity profile of this class of compounds. For this, we studied a small in-house library of nitrocatechol-based COMT inhibitors, which share the same nitrocatechol moiety and have different substituents on the side chain (Table 1) under the same experimental conditions described thus far (Tables S3 and S4). To determine the lipophilicity, we calculated the value of CHI log D_2.3 for each nitrocatechol-based compound at pH 2.3. It is well known that catecholic group in physiological conditions (pH 7.4) could be deprotonated,³⁶ so to maintain the neutral form and peak shape, the measurements of CHI evaluation were performed under acidic conditions.³⁷

Table 1. In-House Nitrocatechol Library: Code, Chemical Structure, CHI log D_2.3, and COMT IC₅₀ Inhibitory Values^ab.

^aValues obtained from reference.⁹

^b* Inactive.

As it was observed, the insertion of the nitro group increases the value of CHI log D_2.3 when compared with related compounds with hydrogen in position R1 (1 vs 2; 7vs8 and 10vs11). Meanwhile, by increasing the length and the insertion of the aryl group in the R3 position, an increase of lipophilicity was observed (by comparison of nitro-based compounds 2, 5, 8, and 11).

Considering the values of CHI log D_2.3 of tolcapone (2.547) and entacapone (1.750) previously obtained by our group, we suggested that the difference in lipophilicity between these two drugs could explain the difference between their cytotoxicity in the previous cell lines tested. The increased lipophilicity of tolcapone enables faster diffusion across biological membranes, particularly in mitochondria, which may explain the toxic effects we observed in hepatocytes and HepG2 cells.

To establish this trend, we try to correlate the cell viability obtained in both glucose and galactose medium, measured by both metabolic and lysosomal activities (Figures S4 and S5) of all compounds tested under this work with their CHI log D_2.3 values (Figure 6). We found a proportional correlation between lipophilicity (estimated by CHI log D_2.3) and the cell death of HepG2 cells treated with 50 μM of the test compounds, which was particularly clear when metabolic activity was measured using the OXPHOS conditions (R² = 0.9303) (Figure 6B).

Figure 6.

Cell viability estimated by measuring both metabolic and lysosomal activities of HepG2 cells cultured in high-glucose (A) and galactose (B) medium as a function of the lipophilicity of synthesized compounds 1–12 as well as tolcapone and entacapone calculated by CHI log D_2.3 values obtained at pH 2.3.

Lipophilic nitrocatechols may diffuse rapidly into cells and, in the absence of detoxification processes and transporter-mediated efflux, trigger cytotoxicity in a dose-dependent manner, which is in accordance with previous reports.³⁸ Despite this, the question related to the possibility of the nitro group as an essential model of toxicity remains. To answer this question, we repeated the screening with three similar catechols lacking the nitro group–caffeic acid (1), caffeic acid benzyl ester (7), and caffeic acid phenetyl ester (10) (Figure 7A). The decrease of viability in HepG2 cells exposed to the test compounds was a direct function of lipophilicity, showing that toxicity does not depend on the presence of a nitro group. It is important to note that carboxylic acid-based compounds, such as compounds 1–3, due to their ability to be deprotonated under physiological conditions, normally are not able to permeate cellular membranes and in that order are not capable of exerting cytotoxic effects.

Figure 7.

Rational design and comparison between the insertion of nitro or/and cyano groups in the catecholic scaffold and the related cellular effects. The cell viability of HepG2 cells cultured with glucose medium after 24 h treatment with compounds 1–3 and 7–12 (10 μM) by measuring both metabolic and lysosomal activities (A) and the measurement of RS normalized with the metabolic activity of the HepG2 cells after 24 h treatment with the compounds (B). Mitochondrial parameters of HepG2 cells exposed to compounds 9 and 12 (50 μM) for 24 h, cultured in high-glucose (C) and OXPHOS media (D) and the comparison of OCR related with different bioenergetics parameters (E) measured using a Seahorse XF96 extracellular flux analyzer. The data are expressed as the means of four independent experiments together with the standard error mean (mean ± SEM). The control data is represented by a black dotted line. Statistical calculations were performed by ANOVA. In all cases, p values lower than 0.05 were considered significant (*p < 0.05, ***p < 0.001, ****p < 0.0001 compared with control cells; ^#p < 0.05 compared with the data from cells treated with compound 9 for the same culture medium used; ⁺p < 0.05 comparing cells treated with same drug but with different cell culture medium).

Although compounds 7 and 10 already showed some cytotoxicity, when a nitro group was added to the catecholic ring, the cellular treatment of the cells with resulting compounds (2, 8, and 11) led to an increase of cell death when compared with those without nitrocatecholic groups. This occurs because chemical modification of catechol with an electron-withdrawing group (e.g., nitro group) lowered the pK_a hydroxyl group and was demonstrated to increase catechol’s resistance to oxidation for surface bonding.³⁹ The higher ability of the catechol group to be deprotonated could be the reason why an increase of ROS production was observed when cells were treated with compounds 9 and 11 (Figure 7B). Surprisingly, those harmful effects were reverted when a cyano group was introduced to compound scaffold presenting higher cell viability and lower ROS production (Figure 7B).

In order to assess the role of the cyano group in the toxicity of the nitrocatechols, a mitochondrial toxicity screening, presented in Figure 7C,E, was performed using compounds 9 and 12 under the same conditions described already for tolcapone and entacapone (Figure 5).

In Figure 7C, it was observed that the most lipophilic compound 12 showed a more prone harmful effect on the mitochondria of HepG2 cells cultured in high-glucose medium, which did not occur when cells grew in galactose medium (Figure 7D).

Both compounds did not interfere in nonmitochondrial respiration -related OCR (Figure 7E) when compared with control cells, which when compared with the results obtained for tolcapone showed an increase in mitochondrial viability. Despite that, compound 12 caused a decrease in maximal respiration (∼23%) and spare respiratory capacity (∼42%) when compared with control cells in a glucose medium. The slight effects caused by compound 9 in maximal respiration (%OCR = 85.6%) and spare respiratory capacity (%OCR = 78.7%) were noticed only in cells cultured in galactose medium.

Thus, although tolcapone, entacapone, and compounds 9 and 12 are nitrocatechols (Table 1), there is a major difference in the substituents located at the C1 position of the nitrocatechol ring. Only tolcapone displays a carbonyl group directly attached to this position, while all the other derivatives bear a double bond with trans isomerism. The presence of this chemical feature extends the π-system and delocalizes the electrons of the aromatic ring, which may confer increased stability and lower reactivity to the nitrocatechol moiety. Accordingly, this structural feature could contribute to the lower mitochondrial toxicity of entacapone and compounds 9 and 12 when compared to tolcapone.

Conclusions

The present study pointed toward a clear difference in the toxicological behavior of tolcapone and entacapone. Considering that after oral administration of the highest dosage of tolcapone (200 mg), the plasma concentration of the drug is approximately 20–25 μM, in most in vitro studies, tolcapone using hepatic cells only demonstrates cytotoxicity in concentrations above 50 μM. This exercise is of course oversimplistic but gives us some hints that tolcapone-induced hepatic damage might involve other mechanisms.

Meanwhile, the evaluation of the lipophilicity and cytotoxicity of compounds for a small library gave us a glimpse of the required approaches to fine-tune of the structure of nitrocatechols to decrease their cytotoxicity. In summary, the current study suggests that tolcapone-induced toxicity occurs due to the direct mitochondrial effect of the unmetabolized compound. Our findings suggest that the toxic effects of nitrocatechols are enhanced by their lipophilicity due to increased diffusion and/or accumulation in biological membranes. In addition, the presence of a carbonyl group directly attached to the nitrocatechol ring may increase the reactivity of the nitrocatechol toward its intracellular targets. To counteract this fact, the presence of cyano group in double bond is found to be a productive chemical feature as it enhances the stability of the nitrocatechol group decreasing its reactivity, which is ultimately translated in amelioration of cell viability.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.4c00124.

• Data from CHI measurements using a calibration set of compounds; mitochondrial parameters analyzed by Seahorse XF96 extracellular flux analyzer of tolcapone, entacapone, compounds 9 and 12; and cell viability of compounds 1–12 using primary rat hepatocytes, HepG2, and Caco-2 cell lines (PDF)

The authors declare no competing financial interest.