Methotrexate-Induced Liver Injury Is Associated with Oxidative Stress, Impaired Mitochondrial Respiration, and Endoplasmic Reticulum Stress In Vitro

Abstract

Low-dose methotrexate (MTX) is a standard therapy for rheumatoid arthritis due to its low cost and efficacy. Despite these benefits, MTX has been reported to cause chronic drug-induced liver injury, namely liver fibrosis. The hallmark of liver fibrosis is excessive scarring of liver tissue, triggered by hepatocellular injury and subsequent activation of hepatic stellate cells (HSCs). However, little is known about the precise mechanisms through which MTX causes hepatocellular damage and activates HSCs. Here, we investigated the mechanisms leading to hepatocyte injury in HepaRG and used immortalized stellate cells (hTERT-HSC) to elucidate the mechanisms leading to HSC activation by exposing mono- and co-cultures of HepaRG and hTERT-HSC to MTX. The results showed that at least two mechanisms are involved in MTX-induced toxicity in HepaRG: (i) oxidative stress through depletion of glutathione (GSH) and (ii) impairment of cellular respiration in a GSH-independent manner. Furthermore, we measured increased levels of endoplasmic reticulum (ER) stress in activated HSC following MTX treatment. In conclusion, we established a human-relevant in vitro model to gain mechanistical insights into MTX-induced hepatotoxicity, linked oxidative stress in HepaRG to a GSH-dependent and -independent pathway, and hypothesize that not only oxidative stress in hepatocytes but also ER stress in HSCs contribute to MTX-induced activation of HSCs.

Keywords: ER stress; HepaRG; in vitro model; liver fibrosis; methotrexate; mitochondria; oxidative stress; stellate cells.

Figure 1

The effect of MTX treatment on HepaRG viability and functionality. HepaRG were exposed to MTX for 72 h. Viability was assessed using the CellTiter-Glo^® Luminescent Cell Viability Assay and expressed as relative ATP content (% of control), N = 6–9 (A). The HepaRG were fixed and stained for albumin (green) and counterstained with DAPI (blue). Images were taken using the Zeiss Colibri 7 LED system (B). Scale bar representing 50 µm. Graphs represent means ± SD; statistical analysis based on one-way ANOVA; ***, p ≤ 0.001.

Figure 2

Expression of folate transporter SLC19A1 and DHFR in HepaRG. HepaRG were treated as previously described with or without MTX for 72 h. Gene expression of folate transporter SLC19A1, responsible for intracellular uptake of MTX, was detected using q-RT-PCR. Data are expressed as fold change, N = 3 (A). HepaRG were treated with 30,000 nM MTX for 7 days and processed for Western blot analysis. Anti-DHFR antibody was used to detect DHFR, β-actin was used as loading control (B). Semi-quantification of DHFR normalized to β-actin band intensity, N = 3 (C). Quantitative intensity staining value (QISV); Bar graphs represent means ± SD; statistical analysis based on one-way ANOVA (A) and Student’s unpaired t-test (C); *, p ≤ 0.05; **, p ≤ 0.01.

Figure 3

NAC treatment reduces superoxide formation in MTX-treated HepaRG. HepaRG were pre-exposed for 2 h to 125 µM NAC or left untreated, then exposed to MTX without NAC or MTX with NAC for 72 h. The HepaRG were stained live with Celltracker violet BMQC (blue) and MitoSOX^TM (red) to investigate superoxide production following MTX exposure with and without NAC. Images were taken using the Zeiss Colibri 7 LED Fluorescence system (A). Scale bar representing 200 µm. The quantity of the MitoSOX^TM was measured using ImageJ and normalized to the quantity of Celltracker violet BMQC. Data are expressed as QISV, N = 3 (B). Viability was assessed using the CellTiter-Glo^® Luminescent Cell Viability Assay and expressed as relative ATP content (% of control), N = 3 (C). For better illustration, dashed lines have been added showing the mean value of the solvent control DMSO, 0.06% (B,C). Bar graphs represent means ± SD; statistical analysis based on two-way ANOVA; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001.

Figure 4

Measurement of the oxygen consumption rate in MTX-treated HepaRG. Mito Stress test profile (A). The oxygen consumption rate (OCR) of HepaRG exposed to MTX for 72 h was measured using the Seahorse XF96 analyzer (B). The HepaRG were sequentially exposed to three treatments: Oligomycin (1 μM), FCCP (2.5 μM), and a mix of Rotenone (0.5 μM) + Antimycin A (0.5 μM). OCR was measured 6 times per treatment step and expressed as pmol/min, N = 3 (B). Spare respiratory capacity was calculated for each condition as the difference from the basal to the maximal OCR (C). Graphs represent means (B) or means ± SD (C); statistical analysis based on one-way ANOVA; *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001.

Figure 5

Effect of NAC on oxygen consumption rate in MTX treated HepaRG. The oxygen consumption rate (OCR) of HepaRG exposed to MTX for 72 h (A) or HepaRG preincubated with 125 μM NAC and then exposed to MTX for 72 h (B), was measured using the Seahorse XF96 analyzer. The HepaRG were sequentially exposed to three treatments: Oligomycin (1 μM), FCCP (2.5 μM), and a mix of Rotenone (0.5 μM) + Antimycin A (0.5 μM). OCR was measured 6 times per treatment step and expressed as pmol/min. Spare respiratory capacity was calculated as the delta maximal respiration to basal respiration (C). Graphs represent means (A,B), and means ± SD (C); statistical analysis based on two-way ANOVA; *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001.; N = 3.

Figure 6

MTX triggers the activation of 2D hTERT-HSC in mono- and co-culture with HepaRG. Monocultures of 2D hTERT-HSC (A), or co-cultures with hTERT-HSC and HepaRG (B), were exposed to varying concentrations of MTX or the positive control TGF-β1 for 7 days. Cells were stained for stress marker αSMA (green), hepatocyte marker albumin (red), and counterstained with DAPI (blue). Scale bar representing 100 µm.

Figure 7

MTX treatment induces gene expression of ER-stress marker DDIT3 in mono- and co-cultures. hTERT-HSC monocultures or co-cultures with HepaRG were exposed to different MTX concentrations or TGF-β1 for 7 days. Q-RT-PCR was performed to detect expression levels of DDIT3. For a better visual representation, the dashed line showing the baseline expression of DDIT3 in the solvent control, was added. Bar graphs represent means ± SD; statistical analysis based on one-way ANOVA; ****, p ≤ 0.0001; N = 3.