INH (Lot number MKBV9475V, analytical standard ≥99%), LPS (Lot number 025M4128V, derived from Escherichia coli 0128:B12 serotype, source strain is CDC 2440-69), were purchased from Sigma-Aldrich (St. Louis, MO, USA). DEX as sodium phosphate ready-made injections (Lot number 5160303) was obtained from Hubei Chang Tian Pharmaceutical Company (Hubei, China). TRIzol reagent bought from Invitrogen Life Technologies (CA, USA) while PrimeScript^TM RT Master Mix from Takara Biotechnology (Dalian, China) and SYBR Green Supermix from Vazyme Biotech (Nanjing, China). Other reagents were from high-analytical grade commercially available or as specified in the relevant contexts.

Male, Sprague-Dawley rats weighing 200–220 g were obtained from Shanghai Lingchang Biological Technology Co., Ltd. (Shanghai, China). All experimental procedures were conducted in accordance with the guide for Institutional Animal Care and Use Committee at China Pharmaceutical University and the National Institutes of Health (NIH) guidelines for the care and use of laboratory animals. Rats were housed in controlled environmental conditions (23 ± 1°C, 55 ± 5% relative humidity, 12 h light–dark cycle) with free access to food and water ad libitum. Animals were acclimatized for 1 week before conducting the experiments. All experimental procedures were approved by China Pharmaceutical University, National Drug Screening Centre and Jiangsu Institute of Materia Medica ethical committees, Nanjing, China.

Rats (n = 8 per group) were randomly divided into five groups: control group (group I), INH 200 mg/kg (ig plus intravenous LPS 2 mg/kg (group II), INH 400 mg/kg (ig) plus intravenous LPS 2 mg/kg (group III), while both group IV and group V received the same INH and LPS doses in addition to intraperitoneal DEX (4 mg/kg).

INH was administered for 14 consecutive days, whereas LPS was given as bolus dose at day 14, 2 h before the last INH dose. DEX was also given as a single dose at day 14, 1 h prior LPS administration. These selected concentrations were in accordance with previous research on INH/LPS combinations (Clayton et al., 2007; Metushi et al., 2012; Su et al., 2014; Hassan et al., 2016) and DEX (Dandona et al., 1999; Eum et al., 2003; Cui et al., 2015). Rats were sacrificed after INH last dose; blood was collected, allowed to clot at room temperature and centrifuged for serum. Liver sections were isolated, frozen in liquid nitrogen then stored for further experiments.

For hepatotoxicity determination, following the standard enzymatic techniques, serum alanine transaminase (ALT), aspartate transaminase (AST), total bile acids (TBA), total bilirubin (TBil), gamma-glutamyl transferase (GGT), triglyceride (TG), and total cholesterol (TC) levels were measured by HITAC7170A automatic analyzer (Hitachi, Japan). Both liver TG and TC were determined following the manufacturer’s instructions of corresponding detection kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China).

Immediately following sacrifice of the rats, histological examinations were conducted on rat liver slices. Liver slices were directly fixed in 10% paraformaldehyde solution and embedded in paraffin wax, then stained with hematoxylin and eosin (H&E). Slides were coded, randomized, and assessed by pathologists who during the evaluation of the slides were blinded to the treatment groups.

In order to verify accumulation of lipids in hepatic tissues following INH/LPS co-treatment, fresh frozen liver sections were treated with oil red O staining following the standard protocol at Jiangsu Provincial Hospital of Integrated Traditional and Western Medicine (Nanjing, China).

The levels of total superoxide dismutase (SOD), reduced glutathione (GSH), malondialdehyde (MDA), and hepatic total-anti-oxidant capacity (T-AOC) contents were detected with their corresponding assay kits provided by Nanjing Jiancheng Bioengineering Institute (Nanjing, China), according to manufacturer instructions.

Total RNA was extracted from rat’s liver tissues by TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) following manufacturer instructions provided, 1 μg RNA was quantified using 2000 Nanodrop Spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA) then reverse-transcribed into cDNA with PrimeScript^TM RT Master Mix (Takara Biotechnology, Dalian, China). Quantitative polymerase chain reaction (PCR) was executed in 20 μL volume containing 10 μL SYBR Green Supermix (Vazyme Biotech, Nanjing, China), 1 μL of cDNA, 7 μL of RNase/DNase-free water and 500 nM each primer. The thermal cycler conditions were as follows: 30 s at 95°C, followed by 40 cycles of 5 s at 95°C and 10 s at 60°C. A melting curve analysis was carried out for each reaction from 65 to 95°C. The threshold cycle at which the fluorescent signal reached an arbitrarily set threshold near the middle of the log-linear phase of amplification for each reaction was calculated, and relative quantities of each mRNA were determined. Primers used for PCR were listed in Supplementary Table S1. Gene expression was evaluated by ΔΔCT method in which glyceraldehyde-3-phosphate dehydrogenase (GAPDH) served as a reference gene.

Protein samples were extracted from liver tissues with radioimmunoprecipitation assay (RIPA) buffer with phosphatase and protease inhibitors (Beyotime, China). The extracted proteins (70 μg) were quantified by bicinchoninic acid protein assay kit (Beyotime, China) and prepared with 4× sample buffer (BD Biosciences, USA). Following their separation in sodium dodecyl sulfate polyacrylamide gel electrophoresis, proteins were transferred to the polyvinylidene difluoride membranes, blocked with 5% bovine serum albumin (BSA) for 1 h and further incubated with specific primary antibodies overnight. The proteins in bands were then incubated with secondary antibodies and detected on a Bio-Rad Gel Doc XR⁺ Imaging System (CA, USA). Primary and secondary antibodies used for western blotting were listed in Supplementary Tables S2,S3, respectively.

Paraffin-embedded liver sections were stained by transferase dUTP nick-end labeling (TUNEL) assay in order to identify apoptotic hepatocytes using TUNEL detection kit (KeyGEN BioTECH, Nanjing, China) following the manufacturer’s guidelines. TUNEL-stained liver samples were captured with light microscope (Olympus IX81). Immunohistochemistry testing was performed for CYP2E1 and peroxisome proliferator-activated receptors alpha (PPARα) using corresponding antibodies on hepatic tissues fixed in formaldehyde and embedded in 3 μ paraffin slices.

The data in all experiments are expressed as mean ± SD. Statistical comparisons were performed by one-way analysis of variance; difference between two groups was analyzed by Student’s two-tailed t-test using GraphPad Prism 6.0 (Graph-Pad, San Diego, CA, USA). The level of significance was set at P-value <0.05.

INH/LPS co-treated animals showed reduction in both total body weight and liver weight, while DEX-pretreated rats revealed insignificant partial increment in total body weight with improvement in liver weight nearly up to the control group ones (Figures 1A,B). Meanwhile, serum markers of liver injury including TBA, TBil, and GGT showed significant elevation in the animal groups treated by INH and LPS combination, with highest increase within INH dose 400 mg/kg. Pre-addition of DEX reduced serum TBil and GGT levels to those in the control group (i.e., significantly indifferent from control group measurement) while TBA level remained elevated, especially in the rat group that received INH 400 mg/kg (Figures 1C,D). In line with our previous results, serum ALT level showed significant reduction in INH/LPS-treated rats, while those receiving DEX revealed marginal increment but still beneath the control levels. Furthermore, AST levels insignificantly fluctuated in the INH/LPS co-treated rats, while administration of DEX resulted in normal control levels (Figure 1D). Histopathological evaluation revealed massive hepatocellular damage, necrosis, micro-and macro-vesicular steatosis accompanied with severe inflammation and inflammatory infiltration in INH/LPS-treated rats (Figure 2A). However, in DEX-administered rats there were no signs of inflammation, inflammatory infiltration, or steatosis rather than minor hepatocellular necrosis as observed and indicated in the histological toxicity score (Figure 2B).

Since micro- and macro-vesicular lipid droplets had been obviously accumulated and easily detected under microscope after H&E staining, we carried out oil red O staining. As shown in Figure 3A, both 200 and 400 mg/kg INH plus LPS groups displayed intense dose-dependent micro- and macro-vesicular steatosis as confirmed by the strong staining, while liver slides from DEX pre-treated rats showed minor staining of lipid droplets, indicating that DEX minimizes the steatotic-induction abilities of INH/LPS co-treatment.

To validate the changes in hepatic lipid profile, both hepatic TG and TC were quantified using their corresponding quantification kits. Hepatic TC level was significantly increased in INH/LPS-treated animals, with high elevation observed in rats receiving INH 400 mg/kg (Figure 3B). Meanwhile, hepatic TG level also showed significant elevation in rats administered INH/LPS combination, but in rats pre-treated with DEX, their liver TG levels rendered almost to the normal control levels. Moreover, serum TC and TG levels were significantly elevated after administration of INH/LPS combination, while serum TG level significantly reduced in rats given DEX whereas serum TC levels were kept raising (Figure 3B).

Oxidative stress created from the imbalance between free radicals’ generation and removal following INH treatment has being proposed to be one of the causative mechanisms by which INH could produce hepatic damage (Shen et al., 2007; Rao et al., 2012). As indicated in Figure 4, INH/LPS combination showed a significant reduction in both SOD and T-AOC levels with maximum reduction observed at animal group receiving INH 400 mg/kg, which were significantly improved after addition of DEX. Furthermore, MDA is one of the oxidative stress factors being produced as a result from lipid peroxidation, MDA determination showed significant elevation in INH/LPS-treated rats with highest elevation observed in INH dose 400 mg/kg, while DEX pre-administered groups, their MDA levels returned to the normal control ones.

Liver GSH levels exhibited significant inhibition especially in INH 400 mg/kg plus LPS. Meanwhile, rats pre-treated with DEX revealed improvement in GSH levels. These results indicated that, hepatic anti-oxidant abilities have been raised due to the DEX pretreatment which led to minimum hepatocellular damage and reduced the liver injury.

Due to the elevation of serum bile acids parameters mainly TBA, TBil, and GGT in rats receiving INH/LPS combination, we carried out further investigation of the relevant genes associated with bile acids. As indicated in Figure 5A, both farnesoid X receptor (FXR) and small heterodimer partner (SHP) expressions were significantly repressed in INH/LPS-treated groups, DEX pretreatment did not restore their corresponding control levels but it clearly affected expressions of CYP7A1, CYP27A1, and CYP8B1, the genes responsible for bile acid synthesis, which were significantly over-expressed in animal groups receiving DEX (Figure 5B). These genes significant over-expressions were positively correlated with the observed serum TBA elevation.

In the meantime, the repression of bile acid transporters bile salt export pump (BSEP), Na⁺-taurocholate co-transporting polypeptide (NTCP), organic anion-transporting polypeptide 1 (OATP1) and multi-drug resistance protein (MRP₂) seen following INH/LPS treatment also indicated the accumulated levels of toxic bile acids in the hepatocytes. Surprisingly, DEX pretreatment significantly elevated BSEP and MRP₂ expression indicated the reduction of the hepatocyte toxic bile acids accumulation, but DEX showed no improving signs on NTCP and OATP1 expressions (Figure 5C). Meanwhile, DEX administration also caused MRP₃ over-expression (data not shown) which could further explained the elevated serum bile acids level noticed in rats received both DEX and INH/LPS combination.

Although immune system’s role have been neglected in the pathogenesis of INH-induced liver toxicity due to the absence of fever, rash, eosinophilia, and rapid start of INH toxicity upon dose resuming (Metushi et al., 2016), but more recent findings support the relationship between INH toxicity and immune system (Salazar-Paramo et al., 1992; Metushi et al., 2011; Metushi et al., 2014). Consistently with these recent findings, our results demonstrated that, the inflammatory mediator cytokines as interleukin 6 (IL-6), tumor necrosis factor alpha (TNF-α), IL-1α, IL-1β, and interferon gamma (INFγ) were significantly over-expressed in INH/LPS-treated rats whereas DEX pre-treatment significantly reversed their levels to their corresponding control group values (Figure 5D) therefore, eliminating the augmented role of the LPS through prevention of the inflammatory stress.

Many researchers, including our previous report, focused on the core role played by CYP2E1 in INH toxicity (Yue et al., 2004; Cheng et al., 2013; Shayakhmetova et al., 2015; Hassan et al., 2016), we found a positive correlation between INH hepatotoxicity and CYP2E1. Spectacularly, CYP2E1 at its gene expression level almost dwindled in all INH/LPS-treated rats, even those already receiving DEX (Figure 5E). Meanwhile, hepatic lipid accumulation as evidenced by analysis of hepatic TC and TG and red oil O staining led us to study the change in the PPARα, which is highly expressed in liver and accumulating evidences report the vital role of PPARα in the regulation of hepatic lipid profile (Silva et al., 2015). Figure 5F also displayed the significant repression of PPARα in INH/LPS groups while in rats pre-treated with DEX; their PPARα expression was significantly elevated. Furthermore, the enzymes responsible for synthesis of lipids, mainly fatty acid synthase (FAS) and hydroxymethylglutaryl-CoA synthase (HMGCS) revealed significant reduction in INH/LPS groups, which might be attributed to the severe accumulation of hepatic lipids, while those receiving DEX their FAS and HMGCS expression levels were nearly normal (Figure 5F).

Associated with their relevant expressed genes, protein levels of FXR, SHP, and NTCP were reduced in INH/LPS-treated rats whereas their expression levels were elevated in rats received DEX, except for FXR (Figure 5G). Meanwhile, CYP7A1 protein expression was reduced in the absence of DEX, while in the presence of DEX, INH/LPS-treated animals showed supreme protein level, which was clearly in match with its gene expression level and further explained the elevated serum TBA observed in these rats.

CYP2E1 protein level showed pronounced elevation in animals receiving INH/LPS combination with highest increment observed in rats received LPS/INH 400 mg/kg. Unexpectedly DEX pre-treatment did not completely block CYP2E1 protein expression, but only caused marginal reduction in CYP2E1 protein expression, fortifying the fact that, LPS only over-activated and sensitize the liver toward CYP2E1 effects, therefore blocking of LPS will not lead to CYP2E1 inhibition. Interestingly, CYP2E1 immunohistochemistry results revealed that, in line with the protein levels, INH/LPS-treated rats had more intense coloration reflected the hepatocellular CYP2E1 accumulation. In contrast, rats received DEX pre-treatment, their hepatocyte CYP2E1 level was significantly reduced as indicated by more fading colors (Figures 6A,B).

We also focused on the protein level of PPARα in liver after INH/LPS treatment in order to verify the PCR result. Western blot and immunohistochemistry experiments showed that, PPARα protein expression was significantly down-regulated, but its expression was up-regulated in the DEX pre-administered groups (Figures 7A,B).

As experimentally evident, most of INH liver damage came through apoptotic induction in hepatocytes and breaks of DNA strands (Lee et al., 2013; Perwitasari et al., 2015). As displayed in Figure 8A after TUNEL assay application, no observation of hepatocyte apoptosis in the control group, whereas following INH/LPS co-treatment, rat hepatocytes were stained TUNEL positive, however, addition of DEX led to a decrease in this apoptotic activity, revealing that DEX suppressed the apoptotic-stimulating effect of LPS on INH toxicity. This finding was further fortified by analysis of the indicator of cellular apoptosis, hepatic cleavage caspase-3 protein levels, in which groups received INH/LPS co-therapy manifested higher protein expression which was not present after DEX addition (Figure 8B).