Introduction
Oxaliplatin is widely used for the treatment of colorectal cancer.1,2 However, studies have shown that systemic chemotherapy containing oxaliplatin causes hepatic sinusoidal dilatation in up to 78% of patients.3 In the liver, oxaliplatin can also cause hepatic stellate cell (HSC) activation and liver endothelial cell apoptosis.4–6 In addition, increased generation of reactive oxygen species (ROS) associated with mitochondrial dysfunction is a primary cause of oxaliplatin toxicity in the liver.7,8 Protecting the liver while treating the tumor with oxaliplatin is the key to successful treatment.9
Clinical studies have revealed that prophylactic administration of magnesium isoglycyrrhizinate (MgIG) significantly reduced the incidence of liver injury induced by oxaliplatin.7,10,11 MgIG is a novel α-isomer compound synthesized from 18 β-glycyrrhizic acid by isomerization and salification, and has been used in the treatment of chronic hepatitis. Moreover, MgIG protects against liver injury induced by various chemotherapy drugs, including oxaliplatin, paclitaxel, and epirubicin.8 Studies of MgIG’s metabolism showed that it was mainly distributed in the liver tissue through liver and intestine circulation.12 The current understanding of MgIG’s hepatoprotection includes preventing or reducing hepatotoxicity by scavenging free radicals induced by toxins, such as concanavalin A and cyclophosphamide.13,14
Connexin 43 (Cx43) is an important gap junction protein that allows intercellular communication between adjacent cells. In addition to its canonical role at the plasma membrane, connexin can produce small fragments that are present in different cellular compartments, including the nucleus, and act as transcriptional factors to regulate gene expression.15 Cx43 protects the liver from acute-on-chronic liver failure, cirrhosis, and acetaminophen hepatotoxicity,2,16 but the effectiveness of Cx43 in oxaliplatin-based chemotherapy is controversial. In in vitro cell models, Cx43 enhanced oxaliplatin cytotoxicity and oxaliplatin inhibited Cx43 expression.17 Studies have shown that Cx43 mediates the development of resistance to oxaliplatin-based chemotherapeutic agents in colorectal cancer.18 Zou et al.8 have described the hepatoprotective effects of MgIG in an oxaliplatin-induced model of liver injury, which were associated with effects of MgIG on oxidative stress and the proinflammatory cytokine interleukin (IL)-6. The mechanistic link between Cx43 and MgIG in oxaliplatin-induced liver injury remains elusive.
In this study, we established a mouse model of oxaliplatin-induced liver injury to demonstrate HSC activation in oxaliplatin-induced liver injury. The hepatoprotective effects of MgIG were associated with inhibition of HSC activation mediated by metabolite of MgIG on mitochondria Cx43. This study describes the molecular mechanism underlying the hepatoprotective activity of MgIG beyond the oxidative stress pathway and provides evidence in support of Cx43 as a potential therapeutic target in oxaliplatin-induced liver injury.
Methods
Chemicals and reagents
MgIG (Lot: 00215027, purity > 98%) was provided by Chia Tai Tianqing Pharmaceutical Group Co., Ltd. (Lianyungang, Jiangsu Province, P. R. China) and oxaliplatin (Item no. O124003, CAS No. 61825-94-3) was provided by Shanghai Aladdin Biochemical Technology Co. Ltd. (Shanghai, China). Lipopolysaccharide (LPS) was purchased from Sigma (Sigma, St. Louis, MO, USA).
Mice and xenograft model establishment
The mice were housed in the specific pathogen-free barrier facility at Nanjing University of Chinese Medicine. Ten-week-old female C57BL/6 mice were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd (Beijing, China). The tumor-bearing model was established by subcutaneously inoculating 5×105 MC-38 cells into the right flank of the mice. MC-38 cells were derived from chemically induced colon cancer tissue of C57BL/6 female mice.19 Therefore, female mice were used in this study. To determine the protective effectiveness of MgIG, C57BL/6 mice were treated by intraperitoneal injection of MgIG 40 mg/kg once daily followed 1 h later by oxaliplatin 6 mg/kg) for 5 weeks. After 5 weeks, the mice were fasted 12 h and euthanized for blood collection and harvesting liver and tumor tissue.
Cell lines and treatment
The MC-38 colon cancer cell line was kindly provided by Professor Zhang Xiaohong (School of Medicine, Ningbo University, China), and the immortalized LX-2 HSC line was gifted by Professor Zheng Shizhong (Nanjing University of Chinese Medicine). MC-38 and LX-2 were cultured in Dulbecco’s modified eagle medium containing 10% fetal calf serum, 100 U/mL penicillin, 100 µg/mL streptomycin, and 2 mM L-glutamine. LX-2 cells were treated with 2 µg/mL oxaliplatin for 24 h, and 40 µg/mL MgIG was given 1 h before oxaliplatin. Cx43-knockdown short hairpin (sh)RNA constructs were lentivirally packaged and transduced into LX-2 cells according to the manufacturer’s instructions (Genepharma, Shanghai, China). The target region sequences were CCAAACTGATGGTGTCAAT. Lentivirally transduced cells were selected with complete medium containing 1 µg/mL puromycin (Sigma).
Determination of MgIG and 18 α-glycyrrhetinic acid in liver tissue
MgIG was administered to mice intraperitoneally at a dose of 40 mg/kg/day. After 5 weeks, the mice were euthanized. Mitochondria and nuclei from mouse liver were collected with tissue mitochondria isolation and nuclear and cytoplasmic protein extraction kits (Beyotime Biotechnology, Shanghai, China). The protein supernatant was processed and determined as previously described.20 Mitochondria and nuclei were collected from LX-2 cells with cell mitochondria isolation and nuclear and cytoplasmic protein extraction kits (Beyotime Biotechnology, Shanghai, China). MgIG and 18 α-glycyrrhetinic acid were assayed by ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS; Acquity UPLC and Xevo TQ tandem quadrupole mass spectrometer (Waters Corporation, Milford, MA, USA).20 Separation was performed on Shim-pack XP-ODS II columns (75 mm×2.0 mm, 2.2 µm). The mobile phase was 0.1% formic acid in water as phase A and acetonitrile as phase B. Gradient elution was performed as: initial to 35% B (0–0.5 m), 35% to 72% B (0.5–1.0 m), 72% B (1.0–1.6 m), 72–80% B (1.6–3.5 m), 80% B (3.5–3.7 m), 80–35% B (3.7–3.9 m), and stop at 5 m. The flow rate was 0.4 mL/m and the column temperature was 40°C. The retention times were 1.66 m for MgIG and 3.00 m for 18 α-glycyrrhetinic acid. The mass spectrometer was equipped with an electrospray ionization source. The source parameters were a desolvation gas flow of 1,000 L/h at a temperature of 550°C, a cone gas flow of 50 L/h, and a source temperature of 150°C. The capillary voltage was 3,000 V. Analyte detection was performed by multiple reaction monitoring (MRM). The cone voltage setting depended on the specific MRM for each compound. Data was collected in MRM mode by simultaneously screening parent and daughter ions. Dwell time was automatically set by the software. MgIG and 18 α-glycyrrhetinic acid standards were obtained from Chia Tai Tianqing Pharmaceutical Group Co., Ltd.. The standard curve was calculated as y=2E−07x + 8E−06, R2=0.9999 (MgIG); y=2E−05x − 0.003, R2=0.9998 (18 α-glycyrrhetinic acid). Quantification of the target compounds concentrations was based on the peak areas of the sample and the standards.
Histopathology and immunohistochemical (IHC) staining
Small pieces of liver tissue from each study group were fixed in 4% paraformaldehyde for 24–48 h and embedded in paraffin. The blocks were cut into 3 µm slices, processed for hematoxylin-eosin (H&E) and IHC staining using standard protocols. The antibodies used for IHC were rabbit anti-Cx43 (Cell Signaling Technology, Danvers, MA, USA; #3512S, 1:500) and rabbit anti-α-SMA (Proteintech, Singapore; # 14395-1-AP, 1:200) as previously described. The percentage of Cx43-positive cells was determined as the ratio of diaminobenzidine-stained / nuclear hematoxylin-stained cells. Staining results were captured with a Mantra quantitative pathology workstation (Mantra; PerkinElmer Inc., Hopkinton, MA, USA) and quantified by ImageJ (Bethesda, MD, USA).21
Immunofluorescence
Tissue sections of 3 µm thickness were baked in oven at 62°C for 6 h before dewaxing and rehydration. The slides were washed with phosphate buffered saline (PBS) three times before and after heat-induced antigen retrieval. LX-2 cells seeded on coverslips were fixed in methanol for 20 m and permeabilized with 0.1% Triton X-100 (Genview Life Scientific Inc, Tampa, FL, USA) for more than 10 m. Endogenous peroxidases were blocked with 3% H2O2 for 20 m. Nonspecific staining was preblocked with PBS containing 2.5% bovine serum albumin. The primary antibodies were rabbit anti-Cx43 (Cell Signaling Technology; Danvers, Massachusetts, USA, #3512S, 1:500), rabbit anti-TOM20 (Proteintech; #11802-1-AP, 1:1,000), rabbit anti-α-SMA (Proteintech; #14395-1-AP, 1:200) or chicken anti-glial fibrillary acidic protein (GFAP) (Abcam, Cambridge, UK; #ab4674) at 4°C overnight. The secondary antibodies were goat anti-rabbit IgG (Santa Cruz Biotechnology, Dallas, TX, USA; 1:200), Alexa Fluor 594 goat anti-mouse IgG (H+L); Proteintech; #SA00006-3) or Goat anti-chicken IgY H+L (Alexa Fluor 647; Abcam; ab150171) or Hoechst 33342 (Beyotime Biotechnology, Shanghai, China) for 10 m. Stained sections were scanned by fluorescence optical microscopy (Mantra; PerkinElmer).
Transmission electron microscopy
Fresh liver samples were fixed in 2.5% glutaraldehyde for no more than 2 m and kept at 4°C. Tissue was post-fixed in 1.5% osmic acid for about 1 h and washed with 0.1 M PBS and then dehydrated in a 50–100% alcohol series, transferred to pure acetone, embedded, sliced into 70 nm sections, and stained with uranyl acetate and lead citrate. The sections were observed with a Hitachi H-7650 electron microscope.
Western blot assay
Protein extracts of liver tissues were loaded on sodium dodecyl sulfate–polyacrylamide gels and transferred to polyvinylidene difluoride membranes (EMD Millipore, Burlington, MA, USA) after electrophoresis. Protein concentration was determined with a bicinchoninic acid protein assay kit (Beyotime Biotechnology, Shanghai, China). The proteins were marked with anti-Connexin 43 (Cell Signaling Technology; Danvers, Massachusetts, USA, #3512S, 1:1,000), N-cadherin (Cell Signaling Technology; Danvers, Massachusetts, USA, #13116S, 1:1,000), GAPDH (Cell Signaling Technology; Danvers, Massachusetts, USA, #2118s, 1:1,000), TOM20 (Proteintech; Wuhan, China, #11802-1-AP, 1:1,000), or collagen I (Abcam; Cambridge, CB2 0AX, UK, #ab117119,1:1,000). Enhanced chemiluminescence was used for final detection. Secondary antibodies were used at a dilution of 1:10,000. Gels were imaged using a Bio-Spectrum Gel Imaging System (Bio-Rad Laboratories, Hercules, CA, USA), and band intensity was measured with Photoshop CS6 software (Adobe, San Jose, CA, USA). Cx43 protein bands were normalized to TOM2. Mitochondria from LX-2 cells were collected with a mitochondria isolation kit (Beyotime Biotechnology, Shanghai, China).
Real-time PCR
Total RNA was extracted from LX-2 cells after MgIG treatment using RNAiso Plus (Takara, Dalian, China) according to the manufacturer’s instructions. Total RNA was reverse transcribed with Fast Quant RT kits (Tiangen Biotech Co., Ltd, Beijing, China) and quantified by SYBR green PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA). Reactions were run on a 7500 Fast Real-Time PCR System (Applied Biosystems Inc., Waltham, MA, USA). The oligonucleotide sequences of the primers (Sangon Biotech, Shanghai, China) were 5′-CAATCTCTCATGTGCGCTTCT-3′ (sense) and 5′-GGCAACCTTGAGTTCTTCCTCT-3′ (antisense) for Cx43(GJA1) and 5′-CTTCTTTTGCGTCGCCAGCCGA-3′ (sense) and 5′-ACCAGGCGCCCAATACGACCAA-3′ (antisense) for glyceraldehyde-3-phosphate dehydrogenase (GAPDH).
Oil red O staining
Liver tissue sections or LX-2 cell-seeded coverslips were fixed with formaldehyde-calcium solution for 10 m and soaked in 60% isopropanol after washing with distilled water. After 10 m of staining with Oil Red O solution (Solarbio, Beijing, China), differentiation with 60% isopropanol, and washing with distilled water, the preparations were counterstained with Mayer’s hematoxylin. Tissue sections and cells were imaged with a microscope and digital camera (Mantra; PerkinElmer). Evaluation and quantification were carried out as previously described.22,23
Measurement of ROS and mitochondrial membrane potential
ROS levels and the mitochondrial membrane potential of LX-2 cells were measured with commercially available kits (Beyotime Biotechnology, Shanghai, China) and Mitochondrial Membrane Potential Assay Kit were used (Beyotime Biotechnology, Shanghai, China). LX-2 cells were plated in 6-well plates at a density of 105/well, treated with 2 μg/mL oxaliplatin for 24 h, with 40 µg/mL MgIG given 1 h in advance of oxaliplatin. For ROS determination, dichloro-dihydro-fluorescein diacetate (DCFH-DA) was diluted with basal medium to a final concentration of 10 µmol/L. Cells were collected and resuspended in diluted DCFH-DA at a concentration of 1×106/mL, and incubated at 37°C for 20 m. The probe and cells were thoroughly mixed. The cells were washed three times with serum-free cell culture medium, after digestion, and assayed by flow cytometry (Accuri C6; BD Biosciences, Franklin Lakes, NJ, USA). To measure mitochondrial membrane potential, cells were resuspended in 0.5 mL of culture medium and 0.5 mL of JC-1 staining working solution. The cells and the probes were incubated 20 m at 37°C and then centrifuged at 600 g for 4 m. The precipitated cells were washed twice with 1× PBS, resuspended in JC-1 staining buffer (1×), and assayed by flow cytometry (Accuri C6; BD Biosciences).
Statistical analysis
Statistical analysis was conducted with GraphPad software (GraphPad Software Inc., La Jolla, CA, USA). All results were obtained from independent triplicate assays unless otherwise indicated and were reported as means ± standard deviation (SD). Multiple group comparisons were done by one-way analysis of variance. P<0.05 was considered statistically significant.
Results
MgIG alleviates hepatotoxicity induced by oxaliplatin in tumor-bearing mice
MgIG 40 mg/kg/day in mice was determined to be equivalent to a clinical dose in humans. Compared to the vehicle group, robust increase was found in serum AST and ALT in the oxaliplatin treatment group (p<0.01 and p<0.001, respectively). Prophylactic administration of MgIG before treatment with oxaliplatin showed significant attenuation in AST and ALT compared with oxaliplatin-treated animals (Fig. 1, p<0.01 and p<0.001, respectively). No significant differences in AST and ALT level were observed between the control and MgIG-treatment groups (Fig. 1A, B). The inhibitory effect of oxaliplatin on tumor size was not affected by MgIG (Fig. 1C). Histological analysis of liver tissue in vehicle and MgIG groups showed normal patterns. Liver tissue in the oxaliplatin treatment group demonstrated infiltration of inflammatory cells, loss of cellular integrity, and necrosis. Prophylactic administration with MgIG before oxaliplatin markedly suppressed the histological anomalies compared with the oxaliplatin group (Fig. 1D). Transmission electron micrographs of liver tissue in the oxaliplatin treatment group showed significant lysosome aggregation and mitochondrial damage. MgIG pretreatment markedly ameliorated these phenomena (Fig. 1E).
Distribution of MgIG and its metabolite 18 α-glycyrrhetinic acid in liver tissue
The distribution of MgIG and its metabolites 18-α-glycyrrhetinic acid (Fig. 2A, B) in liver cell nuclei, mitochondria, and the cytoplasm external to the mitochondria was were determined by UPLC-MS/MS. Normalization against protein concentration, showed that mitochondria had higher concentrations of MgIG and 18-α-glycyrrhetinic acid than the nuclei, indicating accumulation of the compounds in mitochondria in the liver (Fig. 2C, D). Next, we explored potential targets that mediated MgIG and its metabolite’s protective function in liver mitochondria. Cx43 is a protein in gap junctions, which are involved in liver damage caused by various drugs, as well as diseases such as liver fibrosis and cirrhosis.24 Therefore, we determined whether Cx43 expression was altered in the livers of mice treated with oxaliplatin or MgIG. The results showed that oxaliplatin induced the expression of Cx43 in the liver tissue of tumor-bearing mice, and MgIG inhibited the abnormal increase caused by oxaliplatin (p<0.001, Fig. 2E, F).
MgIG attenuates the expression of Mito-Cx43 in HSCs induced by oxaliplatin
GFAP was used as the marker of HSCs, and the expression of Cx43 and its colocalization with HSCs were determined by immunofluorescence. The results showed that the expression of Cx43 was increased by oxaliplatin, and colocalization with HSCs was observed. Intervention with MgIG inhibited the expression of Cx43 in liver tissue and reduced the colocalization of Cx43 with HSCs (Fig. 3). We stimulated LX-2 cells with 2 µg/mL oxaliplatin and assayed Cx43 by immunofluorescence. The results showed that oxaliplatin induced the expression of Cx43 and its localization in mitochondria in LX-2 cells (Fig. 4A). We extracted mitochondria to detect changes in expression of Mito-Cx43 in LX2 cells by western blotting. The results showed that MgIG significantly reduced the level of Cx43 in mitochondria (p<0.001, Fig. 4B, C). Oxaliplatin induced Cx43 expression in LX-2 cells and colocalization with the mitochondria marker TOM20. However, when MgIG was given before oxaliplatin, colocalization was reduced. The results indicate that MgIG reversed the increased expression of Cx43 in HSCs caused by oxaliplatin. Oxaliplatin induced Cx43 expression in the cytoplasm, mitochondria, and nuclei. MgIG inhibited the expression of Cx43 (Fig. 4C–E). The inhibitory effects did not rely on translocation of Cx43.
MgIG inhibits oxaliplatin-induced liver fibrosis and lipid deposition
Liver fibrosis and abnormal accumulation of lipids are induced by damage of mitochondria. Collagen deposition and α-SMA expression are recognized markers of liver fibrosis. Collagen deposition in liver tissues was observed by Masson staining, and α-SMA expression was assayed by immunohistochemistry. The results of Masson staining and immunohistochemistry of α-SMA both showed that MgIG had effects on oxaliplatin-induced liver fibrosis (Fig. 5A, B). Oil red O staining also showed that MgIG alleviated oxaliplatin-induced liver lipid deposition (Fig. 5C). The results indicate that MgIG relieved oxaliplatin-induced liver fibrosis. The activation of HSCs is the main cause of liver fibrosis. We determined the protein level of α-SMA by immunofluorescence in vitro, finding that the protein level of α-SMA was enhanced in LX-2 cells induced by oxaliplatin (Fig. 5D). The expression of collagen I in LX2 cells was also significantly increased by oxaliplatin (p<0.001, Fig. 5E, F). After MgIG intervention, the expression of α-SMA (Fig. 5D) and the expression of collagen I were significantly reduced compared with administration of oxaliplatin alone (p<0.01, **, Fig. 5E–F). Oil red O staining also showed that MgIG reduced oxaliplatin-induced lipid deposition in LX-2 cells (Fig. 5G).
MgIG attenuates Mito-Cx43 and activation of HSCs
To further confirm the effect of MgIG on Mito-Cx43 expression in HSCs, LPS was used to activate Cx43 as a positive model control in LX-2 cells. Immunofluorescence staining showed that Cx43 expression was enhanced after LPS treatment compared with the blank group, and colocalization occurred between the green fluorescence-labeled TOM and the red fluorescence-labeled Cx43. Red fluorescence intensity was decreased after the intervention of MgIG, and colocalization of green and red fluorescence was significantly reduced (Fig. 6A, E). In addition, MgIG reduced LPS-induced lipid droplet deposition and α-SMA fluorescence expression levels in LX-2 cells (Fig. 6B–D, F). After Cx43 was knocked down in LX-2 cells by lentiviral transduction (Fig. 7A, B). Although oxaliplatin induced lipid deposition in Cx43-depleted LX-2 cells, the inhibitory effect of MgIG on oxaliplatin-induced Cx43-depleted LX-2 cells, lipid deposition and the fluorescence intensity of α-SMA was reduced (Fig. 7C, D, J). Moreover, the inhibition of LPS-induced HSC activation, lipid deposition, and α-SMA expression were attenuated in Cx43-depleted LX-2 cells (Fig. 7D, C–J). The results suggest that Cx43 was involved in the inhibition of oxaliplatin-induced LX-2 activation by MgIG.
MgIG attenuates the activation of HSCs by reducing ROS generation, mitochondria dysfunction, and transcription of N-cadherin
In drug-induced liver injury, mitochondria are highly susceptible to damage that results in dysfunction.25 An increase of ROS is the main manifestation of mitochondrial dysfunction and is an important factor for inducing HSC activation and hepatic fibrosis.26 In addition, an increase in ROS causes destruction of the mitochondrial membrane potential.27 The results showed that oxaliplatin exposure induced oxidative stress in LX-2 cells, and that ROS were decreased after intervention of MgIG (Fig. 8A, B). Quantification of JC-1 fluorescence intensity (red/green fluorescence intensity) indicated that MgIG reduced mitochondria dysfunction that was induced by oxaliplatin (Fig. 8C, D) and that the protection of mitochondria by MgIG may be related to reduction of Cx43.28
Previous immunofluorescence and western blot data showed that MgIG not only attenuates the level of Cx43 in mitochondria, but also inhibits oxaliplatin-induced Cx43 entry into the nucleus (Fig 4A–D). Because Cx43 has been reported to function as transcription factor to activate N-cadherin,15 we assayed the expression of the downstream protein N-cadherin by western blotting. The results revealed that oxaliplatin stimulated the expression of N-cadherin in LX-2 cells, while MgIG intervention reduced N-cadherin (Fig. 8E, F). As a key regulator and indicator of HSC activation, N-cadherin drives epithelial-mesenchymal transition (EMT) and inhibits the apoptosis of HSCs.29,30 The results indicate that MgIG inhibited oxaliplatin-induced HSC activation, which was closely related to its regulation of Cx43 in mitochondria and nuclei.
Discussion
Oxaliplatin is successfully used for management of colorectal cancer and is the standard treatment of metastatic disease. However, it may cause liver injury, particularly sinusoidal obstruction syndrome, which is characterized by histologic changes such as sinusoidal dilatation and congestion, fibrosis, and obstruction. The histopathological changes result from drug-induced damages to endothelial cells and activation of HSCs.31 Our study focused on HSCs and used in vitro and in vivo models to investigate the mechanisms of MgIG in protecting liver from oxaliplatin-based chemotherapy. We found that MgIG inhibited oxaliplatin- or LPS-induced HSC activation and collagen and lipid deposition. Gap junction protein alterations are closely related with liver diseases including drug-induced liver toxicity. Cx43 is the major gap junction protein expressed on HSCs. Results of ours and previous studies demonstrated that oxaliplatin induced Cx43 expression in liver cells, which indicates that Cx43 mediated oxaliplatin-induced liver injury.32 MgIG treatment attenuated Cx43 expression induced by oxaliplatin or LPS in tumor-bearing mice or LX-2 cells. By performing shRNA knockdown experiments, we verified that Cx43 mediated the inhibitory effects of MgIG on HSC activation and lipid deposition induced by oxaliplatin or LPS.
Cx43 is the major gap junction protein in HSCs and mediates liver damage induced by acetaminophen,33,34 oxaliplatin,35 cisplatin,36 cadmium,37 and sevoflurane.38 It also has a role in maintaining liver hemostasis. Hepatocyte cell proliferation was reduced and fewer necroinflammatory lesions were present in the liver parenchyma of Cx43 deficient mice (PMID: 20830702). Our study and those of others have shown that Cx43 is highly expressed in HSCs (Fig. 3).39 The data indicate that Cx43 is a potential target for diseases and pathologies related to activation of HSCs. Inhibition of the gap junction protein may reduce drug toxicity by preventing the propagation of inflammatory or death stimuli to neighboring cells.16 Apart from its localization at the cytomembrane and endoplasmic reticulum, Cx43 is also expressed in mitochondria and nuclei. Our current understanding of mitochondrial Cx43 is based on heart disease scenarios. Cx43 translocates into mitochondria and regulates apoptosis and oxidative stress in terms of ROS production, calcium overload, and mitochondrial membrane depolarization in myocardial hypoxia.40 However, whether mitochondrial Cx43 is involved in hepatotoxicity is unknown. Presently, we found that MgIG reduced the expression level of Cx43 in the mitochondria and nuclei of LX2 cells induced by oxaliplatin (Fig. 4). Besides its functions in mitochondria, Cx43 increased the transcriptional regulation of N-cadherin.15 As an important regulator of HSC activation, N-cadherin promotes EMT and inhibits the apoptosis of HSCs.41,42 We showed that oxaliplatin stimulated the expression of N-cadherin in HSCs, while MgIG intervention reduced N-cadherin (Fig. 8). Cx43 has a critical role in drug-induced organ injury, especially in the liver, kidney, and gastrointestinal tract.35 Therefore, the clinical implications of our study may be related to increased susceptibility to drug-induced organ injury.
The study has some limitations. We did not provide direct evidence of the binding of MgIG to Cx43 protein. Several reports have described the use of 18 α-glycyrrhetinic acid to block intercellular communication and inhibit the expression of Cx43.43–45 We found that MgIG and its metabolite 18 α-glycyrrhetinic acid could be detected in mitochondria in vivo (Fig. 2A–C). It might be possible confirm the direct binding of MgIG or 18 α-glycyrrhetinic acid to Cx43 in mitochondria by microscale thermophoresis.
In China, MgIG has been used in combination with oxaliplatin to treat tumors, but the mechanism of MgIG protection against liver injury induced by oxaliplatin and its effect on tumors has not been elucidated. Our study shows that MgIG was hepatoprotective against oxaliplatin-induced liver injury in tumor-bearing mice, and did not affect the therapeutic effectiveness of oxaliplatin against tumors and Cx43 mediated the protection of MgIG against liver injury. Our study may shed some light on the clinical application of MgIG for patients with Cx43-disturbed expression under pathological conditions.