Synergistic activation of mitochondrial metabolism and the glutathione redox couple protects HepG2 hepatocarcinoma cells from palmitoylcarnitine-induced stress

Abstract

Fatty acid stress can have divergent effects in various cancers. We explored how metabolic and redox flexibility in HepG2 hepatocarcinoma cells mediates protection from palmitoylcarnitine. HepG2 cells, along with HCT 116 and HT29 colorectal cancer cells were incubated with 100 μM palmitoylcarnitine for up to 48 h. Mitochondrial H₂O₂ emission, glutathione, and cell survival were assessed in HT29 and HepG2 cells. 100 μM palmitoylcarnitine promoted early growth in HepG2 cells by ~8% after 48 h versus decreased cell survival observed in HT29 and HCT 116 cells. Palmitoylcarnitine increased mitochondrial respiration at physiological and maximal concentrations of ADP, while lowering cellular lactate content in HepG2 cells, suggesting a switch to mitochondrial metabolism. HepG2 cell growth was associated with an early increase in H₂O₂ emission by 10 min, followed by a decrease in H₂O₂ at 24 h that corresponded with increased glutathione content, suggesting a redox-based compensatory mechanism. In contrast, abrogation of HT29 cell proliferation was related to decreased mitochondrial respiration (likely due to cell death) and decreased glutathione. Concurrent glutathione depletion with BSO prevented palmitoylcarnitine-induced growth in HepG2 cells, indicating that glutathione was critical for promoting growth following palmitoylcarnitine. Inhibiting UCP2 with genipin sensitized HepG2 cells to palmitoylcarnitine, suggesting that activation of UCP2 may be a 2nd redox-based mechanism conferring protection. These findings suggest that HepG2 cells possess inherent metabolic and redox flexibility relative to HT29 cells that confers protection from palmitoylcarnitine-induced stress via adaptive increases in mitochondrial respiratory control, glutathione buffering, and induction of UCP2.

INTRODUCTION

Fatty acid challenges induce cell death in some cancers, such as HT29 colorectal (24) and PC3 prostate carcinoma cells (1), while other cells appear resistant. In fact, excess hepatic fat accumulation is associated with increased risk of hepatocellular carcinoma (HCC) development (21). Vendel Nielsen et al. (23) observed that the unsaturated fatty acid, oleic acid, increased growth in HepG2 HCC cells. However, the mechanisms that confer resistance to fatty-acid stress in HCC remain unresolved.

One possibility is the influence of fatty acid metabolism on the cellular redox environment through mitochondrial H₂O₂ emission—a byproduct of oxidative phosphorylation (OXPHOS). This concept is intriguing given that redox conditions can dictate cell fate whereby an excess of H₂O₂ can invoke deleterious effects on cell function and survival (reviewed by Ref. 15), yet low levels of H₂O₂ can act as a hormetic signal that drives cancer growth (10). This suggests that the influence of metabolically derived H₂O₂ will depend on the ability of the cell to maintain redox conditions. In this regard, the antioxidant glutathione is essential in maintaining cell survival. Indeed, pharmacological increases in glutathione stimulated HepG2 growth, whereas glutathione depletion prevented HepG2 proliferation (11).

Evidence also suggests that uncoupling protein 2 (UCP2) attenuates mitochondrial H₂O₂ emission potentially by dissipating membrane potential and preventing increases in superoxide production (the precursor to mitochondrial H₂O₂) from fatty acid oxidation (reviewed in Refs. 2, 6). Although there is controversy over the mechanistic function of UCP2 (4, 7, 17), it appears that UCP2 is able to dissipate H₂O₂ emission. However, it remains to be determined whether UCP2 mediates a redox-dependent compensation in response to fatty acid challenges that prevent oxidative stress and create a progrowth-reduced environment in HCC. Indeed, palmitic acid increased superoxide production in HepG2 cells, which was amplified by genipin, a selective UCP2 inhibitor (14). Likewise, transgenic overexpression of UCP2 in HepG2 cells prevented oxidative modifications of membranes and proteins, which is indicative of attenuated reactive oxygen species production (8).

The degree to which UCP2 determines cell fate in response to fatty acid stress remains uncertain, as does the potential synergistic responses of the glutathione redox couple. Such dynamic relationships make it difficult to predict whether HepG2 cells would fail or succeed at invoking sufficient metabolic and redox adaptations to survive fatty acid stress. In this study, we determined the response of HepG2 cells to palmitoylcarnitine and explored potential metabolic and redox-based mechanisms. The results demonstrate that palmitoylcarnitine stimulates early growth in HepG2 cells that contrasted with decreased survival in HT29 and HCT 116 cells. This protection from palmitoylcarnitine stress in HepG2 cells was dependent on glutathione and UCP2 and was related to rapid changes in mitochondrial bioenergetics. These findings reveal a dynamic metabolic and redox response that ultimately confers protection against palmitoylcarnitine in HepG2 cells.

MATERIALS AND METHODS

Cell culture conditions.

HepG2 hepatocellular carcinoma cells were gifted by Dr. Paul Spagnuolo (University of Guelph, Guelph, ON, Canada). HT29 and HCT 116 colorectal carcinoma cells were gifted by Dr. Samuel Benchimol (York University, Toronto, ON, Canada). HepG2 cells were grown in EMEM, and HT29 and HCT 116 cells were grown in DMEM. Media were supplemented with 10% FBS and 1% penicillin-streptomycin (Wisent, Saint-Jean-Baptiste, QC, Canada).

Palmitoylcarnitine incubations.

Cells were cultured with 2 mM l-carnitine (Sigma-Aldrich, St. Louis, MO) and 0 μM or 100 μM palmitoylcarnitine (Toronto Research Chemicals, Toronto, QC, Canada).

Relative cell survival assay.

Following 24 or 48 h of palmitoylcarnitine incubations, with or without buthionine-sulfoximine (BSO) and genipin (Sigma-Aldrich), cells plated in 96-well optical bottom black walled plates were fixed using 10% formalin and then stained using 0.5% crystal violet (Sigma-Aldrich) in 25% MeOH. Visualization and fluorescent analyses of crystal violet as a measure of relative cell survival were conducted using the Li-Cor odyssey scanner (Li-Cor Biosciences, Lincoln, NE).

NAD(P)H.

The reduction of XTT can serve as an indirect measure of NADH and NADPH, collectively referred to as NAD(P)H (9). During the final 4 h of palmitoylcarnitine incubations, 50 μl of XTT solution (1 mg/mL XTT, 25 μM phenazine methosulfate; BioShop Canada, Burlington, ON, Canada) was added into each well and incubated at 37°C with 5% CO_2. Absorbance was read at 450 nm using the VICTOR³ 1420 Multilabel Counter plate reader (PerkinElmer, Waltham, MA). Cells were digested in-well using 10% RIPA (Sigma-Aldrich) and protein was assessed with BCA (Thermo Fisher Scientific) to normalize NAD(P)H signal.

Live-cell H₂O₂ measurements.

HepG2 cells were seeded in a 96-well optical bottom black-walled plate (Thermo Fisher Scientific, Waltham, MA). Following palmitoylcarnitine incubations, 10 μM Amplex UltraRed and 1 U/mL horseradish peroxidase was added to each well for 10 min and fluorescence (EX568/EM581) was measured using the BioTek Cytation 3 fluorescent plate reader (BioTek, Winooski, VT). A single read after these time points reflected the net accumulated fluorescent resorufin product of oxidized Amplex Ultrared and represents the “net emission” of H₂O₂ that accumulated. Data were normalized to a H₂O₂ standard curve and protein content was determined by BCA as described above.

Glutathione analysis.

Analysis was performed as previously described (22). Briefly, cells were trypsin-lifted, washed with PBS, and resuspended in final buffer (50 mM Tris, 20 mM boric acid, 2 mM l-serine, 20 μM acivicin, and 5 mM N-ethylmaleimide) for reduced (GSH) and oxidized (GSSG) glutathione determination.

Intracellular lactate determination.

Intracellular lactate was determined as previously published (22) and signal was normalized to preperchloric acid protein determination.

Mitochondrial bioenergetic assessments.

Following palmitoylcarnitine incubations, cells were trypsin-harvested from 10 cm dishes, washed in PBS, and resuspended in mitochondrial respiration media (22) supplemented with 20 mM creatine. Cells were permeabilized with 10 µg/mL digitonin (Sigma-Aldrich) for 30 min rocking at room temperature. Following centrifugation (5 min, 500 rcf), cells were resuspended for high-resolution respirometry as previously described (22), with an aliquot removed to determine protein concentration for normalization as outlined above. Figure 1 describes the protocol used to assess mitochondrial respiratory control, with 10 mM cytochrome c added before succinate to test for intactness of the outer mitochondrial membrane. All responses exhibited <15% increase in respiration following this test.

Western blot analysis.

SDS-PAGE was performed as previously described (12) with adaptations for cell culture. Cells were trypsin harvested, PBS washed, and resuspended in lysis buffer (0.5% IGEPAL, 50 mM Tris, 10% glycerol, 0.1 mM EDTA, 150 mM NaCl, and 1 mM DTT) with protease and phosphatase inhibitors (Sigma-Aldrich). Monoclonal anti-UCP2 antibody produced in rabbit (33 kDa, 1:1,000 dilution, D105V; Cell Signaling Technology, Danvers, MA) was used to determine UCP2 protein content and polyclonal anti-TXNRD2 produced in rabbit (1:200, HPA003323; Sigma-Aldrich) was used to determine thioredoxin reductase-2 protein content.

Statistics.

All results are expressed as means ± SE. Significance was determined as P < 0.05 for all measures. Each n signifies an individual experiment (individual culture plates), with each experiment conducted in triplicate wells where appropriate. Unpaired t-tests were conducted with the exception of the mitochondrial respiration data, where a two-way ANOVA was conducted. A Fisher’s LSD post hoc test was conducted following significant interactions in a two-way ANOVA. All statistics were performed using GraphPad Prism 7 (San Diego, CA).

RESULTS

Unique HepG2 cell growth response to palmitoylcarnitine compared with HT29 and HCT 116.

HepG2, HT29, and HCT 116 cells were exposed to 0 µM and 100 µM palmitoylcarnitine (Fig. 1A) for 24 and 48 h. While HT29 and HCT 116 cells displayed varying degrees of decreased relative cell survival (Fig. 1B; P < 0.05) following 24 and 48 h of 100 µM palmitoylcarnitine, in contrast, HepG2 cells displayed an ~8% increase in relative cell growth (Fig. 1B; P < 0.05). This response is notable considering that HepG2 cells have a population doubling time of ~44 h. These responses were related to increased mitochondrial respiration kinetics in HepG2 cells at 24 h (Fig. 1C; P < 0.05) versus decreased respiration in HT29 cells (Fig. 1D; P < 0.05). To gain insight into whether the increased respiration was linked to greater content of electron transport system proteins, we stimulated maximal electron flux by uncoupling the inner mitochondrial membrane with FCCP. Indeed, the greater respiration seen following palmitoylcarnitine in HepG2 cells demonstrates a greater capacity of the electron transport system (Fig. 1E; P < 0.05). Consistent with previous data, HT29 cells displayed a decrease in FCCP-stimulated capacity (Fig. 1F; P < 0.05), likely signifying that HT29 cells were nonviable rather than a direct decrease in mitochondrial respiratory kinetics following palmitoylcarnitine.

Fig. 1.

Palmitoylcarnitine promotes selective growth in HepG2 cells compared with HT29 and HCT 116 cells and increases mitochondrial respiratory capacity in HepG2 cells. A: schematic showing palmitoylcarnitine bypassing CPT-1, whereby it enters the mitochondria and stimulates β-oxidation resulting in ATP and reactive oxygen species (ROS) production. B: relative cell survival was measured in HepG2 (n = 14), HT29 (n = 6), and HCT 116 (n = 6) cells following 24 and 48 h of 100 µM palmitoylcarnitine relative to 0 µM palmitoylcarnitine at the same time points. Data are reported as means ± SE. *P < 0.05, significant decrease relative to 0 µM palmitoylcarnitine of the same cell type within the same time point. #P < 0.05, significant increase relative to 0 µM palmitoylcarnitine of the same cell type within the same time point. C–F: HepG2 and HT29 cells were incubated with 0 µM or 100 µM palmitoylcarnitine for 24 h. Mitochondrial respiration (C and D) was measured following ADP titration supported by 5 mM pyruvate and 2 mM malate (NADH, complex I; not shown), 5 mM glutamate (G; NADH), and 20 mM succinate (S; FADH₂, complex II) (n = 4), and maximal uncoupled rate of respiration (E and F) was measured following FCCP as an index of electron transport chain content (n = 4). Data are reported as means ± SE. #P < 0.05, main effect for palmitoylcarnitine.

Redox responses to palmitoylcarnitine in HepG2 cells.

Considering that fatty acids have been demonstrated to stimulate mitochondrial superoxide and H₂O₂ emission (14, 20) and excess H₂O₂ emission can lead to deleterious effects throughout the cell, such as glutathione depletion and cell death (15), we then measured reduced (GSH) and oxidized (GSSG) glutathione following palmitoylcarnitine for 24 h in HT29 and HepG2 cells. HT29 cells displayed signs of oxidative stress, as there was a palmitoylcarnitine-induced decrease in GSH (Fig. 2A; P < 0.05), an increase in GSSG (Fig. 2B; P < 0.05), and a decrease in the GSH/GSSG ratio (Fig. 2C; P < 0.05). However, HepG2 cells showed an increase in both GSH (Fig. 2A; P < 0.05) and GSSG (Fig. 2B; P < 0.05) and no changes in GSH/GSSG (Fig. 2C; P < 0.05).

Fig. 2.

Redox stress following palmitoylcarnitine exposure: maintenance of overall redox conditions (GSH/GSSG) in HepG2 cells but not in HT29 cells. HepG2 (n = 5) and HT29 (n = 3) cells were incubated with 0 µM and 100 µM palmitoylcarnitine for 24 h. Reduced glutathione (GSH; A), oxidized glutathione (GSSH; B), and the ratio of reduced-to-oxidized glutathione (GSH/GSSG; C) were assessed. Data are reported as means ± SE. *P < 0.05, significant decrease with 100 µM palmitoylcarnitine compared with 0 µM. #P < 0.05, significant increase with 100 µM palmitoylcarnitine compared with 0 µM.

In agreement with HepG2 cells displaying an increase in mitochondrial respiratory kinetics following palmitoylcarnitine, HepG2 cells displayed a decrease in intracellular lactate (Fig. 3A; P < 0.05). The decrease in lactate and increase in mitochondrial respiratory control suggest a shift from glycolysis to aerobic metabolism. However, despite these responses, there was a decrease in NAD(P)H following palmitoylcarnitine (Fig. 3B; P < 0.05).

Fig. 3.

Palmitoylcarnitine alters H₂O₂, glutathione and cell growth. A and B: intracellular lactate (n = 5; A) and NAD(p)H (n = 9; B) were measured in HepG2 cells following 24 h of 0 µM or 100 µM palmitoylcarnitine. C: H₂O₂ was assessed following 10 min and 24 h of 0 µM and 100 µM palmitoylcarnitine (n = 15). D: total glutathione was measured in HepG2 and HT29 cells following 24 h of 0 µM or 100 µM palmitoylcarnitine (n = 3–5). E and F: relative cell survival was assessed in HepG2 (n = 6; E) and HT29 (n = 3; F) cells following 48 h of 0 µM or 100 µM palmitoylcarnitine, as well as concurrent incubations with 50 µM buthionine sufloximine or 100 µM genipin. G and H: HepG2 cells were incubated with 0 µM and 100 µM palmitoylcarnitine for 24 h and uncoupling protein-2 (UCP2) protein content was visualized (G) and determined (H) (n = 5). I: schematic depicting the selective inhibition of UCP2 by genipin and the depletion of glutathione with buthionine sulfoximine (BSO). J: schematic of palmitoylcarnitine acutely triggering an increase in H₂O₂ emission, resulting in UCP2 activation and an increase in GSH, which, in turn, lowers H₂O₂, as well as stimulates an increase in growth. Data are reported as means ± SE. *P < 0.05, significant decrease with 100 µM palmitoylcarnitine relative to 0 µM palmitoylcarnitine. #P < 0.05, significant increase with 100 µM palmitoylcarnitine relative to 0 µM palmitoylcarnitine.

In HepG2 cells, H₂O₂ emission increased transiently by 10 min followed by a decrease at 24 h (Fig. 3C; P < 0.05). This reversal was related to an increase in total glutathione in HepG2 cells contrasted to decreasing total glutathione in HT29 cells (Fig. 3D; P < 0.05), as shown previously (22), suggesting a redox buffering mechanism may have been triggered by the initial H₂O₂ emission. To test this possibility, HepG2 cells were coincubated with palmitoylcarnitine and the glutathione-depleting agent buthionine sulfoximine (BSO). BSO prevented palmitoylcarnitine-induced growth (Fig. 3E; P < 0.05; Fig. 3I) but did not sensitize HepG2 cells to decreased cell survival as was observed in HT29 and HCT 116 cells. We then determined whether the reversal in H₂O₂ emission was related to a second mechanism of mitochondrial uncoupling. The UCP2 inhibitor genipin caused marked sensitization of HepG2 cells to palmitoylcarnitine resulting in a drastic decrease in cell survival (Fig. 3E; P < 0.05; Fig. 3G), despite no change in UCP2 protein content in HepG2 cells (Fig. 3, G and H). Palmitoylcarnitine resulted in significantly decreased cell survival in HT29 cells regardless of BSO and genipin (Fig. 3F, P < 0.05). Collectively, these findings support a model whereby palmitoylcarnitine stimulates acute increases in H₂O₂, which leads to both an increase in total glutathione and UCP2 activation to ultimately decrease H₂O₂ and protect HepG2 cells (Fig. 3H).

DISCUSSION

Unlike HT29 and HCT 116 colorectal carcinoma cells, HepG2 HCC cells were resistant to palmitoylcarnitine-induced decreases in cell survival and demonstrated a small increase in growth by 48 h. This protection in HepG2 was associated with increased mitochondrial respiratory kinetics and decreased intracellular lactate content, suggesting a shift toward oxidative phosphorylation. Furthermore, inhibition of the glutathione redox couple prevented palmitoylcarnitine-induced HepG2 cell growth, whereas UCP2 inhibition sensitized HepG2 cells to palmitoylcarnitine. These findings support a model that HepG2 cells invoke a cytoprotective response to palmitoylcarnitine that is associated with a mitochondrial and redox-based flexibility system (Fig. 4).

Fig. 4.

Proposed model of HepG2 cell adaptation to palmitoylcarnitine. The mitochondrial substrate, palmitoylcarnitine, stimulates an acute increase in H₂O₂. A compensatory increase in glutathione and activation of UCP2 eventually lead to lower H₂O₂ emission. These hormetic responses to palmitoylcarnitine result in an increase in glutathione redox-buffering capacity. An increase in oxidative capacity also improves ATP synthesis. Collectively, the metabolic and redox flexibility of HepG2 cells results in improved proliferation in response to palmitoylcarnitine in contrast to the abrogations observed in HT29 and HCT 116 cells.

Metabolic adaptations in HepG2 responses to palmitoylcarnitine.

A dynamic series of metabolic and redox adaptations occurred within 48 h of palmitoylcarnitine exposure. First, a transient increase in H₂O₂ emission occurred by 10 min. This suggests that reducing equivalents generated by β-oxidation may have exceeded a relatively low rate of OXPHOS in HepG2 cells consistent with the concept of electron slip in situations of excess, reducing equivalent supply (18). The stimulation of H₂O₂ by 10 min may have preceded the increased glutathione at 24 h that coincided with the maintenance of cell survival, in conjunction with the UCP2 dependency observed at 48 h. While evidence suggests that UCP2 may not be a direct uncoupler of membrane potential (18), inhibition of UCP2 by genipin in HepG2 cells was previously shown to increase superoxide generation in response to palmitoylcarnitine in HepG2 cells (14), suggesting that UCP2 attenuates mitochondrial H₂O₂ emission in response to fatty acids (4). Collectively, the present findings suggest that the dissipation of initial H₂O₂ emission at 10 min was related to activation of uncoupling via UCP2, although this latter point is speculative without measures of proton conductance (18).

In support of this proposed model of redox flexibility in HepG2 cells, in mice fed a high-fat diet, inhibition of UCP2 by genipin prevented high-fat diet-induced liver damage (26), suggesting that UCP2 may be involved in the pathogenesis and progression of liver damage toward HCC development. Likewise, Huang et al. (11) demonstrated that glutathione is a critical regulator in HepG2 growth, whereby its depletion by BSO stagnated growth versus increasing glutathione-triggered proliferation. Collectively, these previous findings align with the present study and support the proposed model that early induction of H₂O₂ by palmitoylcarnitine in HepG2 cells mediated cell survival through a UCP2-dependent attenuation of redox stress concurrent with increased glutathione. Although the mechanism for the increased glutathione remains unclear, it is possible that that a UCP2-mediated decrease in H₂O₂ emission may have permitted a higher steady-state maintenance of glutathione or that redox activation of Nrf2 increased total glutathione content (16).

The potential clinical relevance of these findings is intriguing. For example, obesity is associated with a greater occurrence of HCC, particularly stemming from excess adipose deposits in the liver that lead to nonalcoholic fatty liver disease (NAFLD), nonalcoholic steatohepatitis, and cirrhosis (reviewed in Refs. 3, 21)]. It has been previously demonstrated that HepG2 cells cultured in fatty liver matrix had an observable increase in cell growth observed at 9 days (25), which is similar to the observation that the fatty acid, oleic acid, stimulated HepG2 cell growth following 4 days (23). Therefore, it is likely that excess fat provided to mitochondria fosters a robust progrowth environment in HepG2 cells. The present findings suggest that this is mediated through a hormetic mechanism stemming from inherent metabolic-redox flexibility in HCC. This adaptive mechanism does not appear to exist in colorectal cancer cells as shown in the present study whereby palmitoylcarnitine lowered colorectal cancer survival consistent with previous findings (5, 19, 22, 24). As such, these findings suggest that redox-based biomarkers might be identified to predict the relative susceptibility (positive or negative) of cancers to palmitoylcarnitine exposure. Lastly, it should be noted that the effects of fatty acids per se on cancer cell fate may be quite diverse given that high-fat diets are also associated with increased colorectal cancer. The use of palmitoylcarnitine in the present study was intended to isolate the effects of mitochondria specifically given that palmitoylcarnitine bypasses any potential suppression of fatty acid import by CPT-1 that has been reported in some cancers (13).

Another possible mechanism pertains to the increased mitochondrial respiratory capacity observed after palmitoylcarnitine treatment in HepG2 cells. Twenty-four hours of palmitoylcarnitine increased ADP-stimulated mitochondrial respiration in conjunction with lower NAD(P)H content. While NAD(P)H content on its own does not represent rate of generation or oxidation by the mitochondria, the greater respiratory kinetics following palmitoylcarnitine suggests that the lower NAD(P)H was related to a greater rate of its utilization. The lower NADH could, in theory, place less pressure on membrane potential-dependent superoxide generation, thereby explaining the lower H₂O₂ emission observed by 24 h in addition to UCP2 induction discussed above.

Perspectives and conclusions.

The results in HepG2 cells suggest that fatty acid stress survival mechanisms function through a dynamic relationship between oxidant generation and compensatory increases in glutathione redox buffering capacity. The present findings highlight how the interplay between metabolism and redox biology is not necessarily a predictable event of simple oxidative stress, but rather potentially a hormetic system that triggers an advantageous compensation promoting cancer cell growth. As such, this model could be applied to determine how inherent metabolic and redox flexibility through H₂O₂ emission, OXPHOS, UCP2 activity, and glutathione dynamics could confer protection from fatty acid stress.

GRANTS

Funding was provided to C. G. R. Perry by the Natural Science and Engineering Research Council (NSERC; Grant 436138-2013), with infrastructure supported by Canada Foundation for Innovation, the James. H. Cummings foundation, and the Ontario Research Fund. P. C. Turnbull was supported by a NSERC CGS-PhD scholarship.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

P.C.T., C.F.T., M.K.C., and C.G.R.P. conceived and designed research; P.C.T., A.C.D., and C.F.T. performed experiments; P.C.T., A.C.D., M.K.C., and C.G.R.P. analyzed data; P.C.T., C.F.T., M.K.C., and C.G.R.P. interpreted results of experiments; P.C.T. prepared figures; P.C.T. and C.G.R.P. drafted manuscript; P.C.T., A.C.D., C.F.T., M.K.C., and C.G.R.P. edited and revised manuscript; P.C.T., A.C.D., C.F.T., M.K.C., and C.G.R.P. approved final version of manuscript.