Dual TBK1/IKKε inhibitor amlexanoX mitigates palmitic acid-induced hepatotoXicity and lipoapoptosis in vitro

Keywords: AmlexanoX Palmitic acid HepatotoXicity Lipoapoptosis In vitro


The common causes of Non-alcoholic fatty liver disease (NAFLD) are obesity, dyslipidemia, and insulin resis- tance. Metabolic disorders and lipotoXic hepatocyte damage are hallmarks of NAFLD. Even though amlexanoX, a dual inhibitor of TRAF associated nuclear factor κB (NF-κB) activator-binding kinase 1 (TBK1) and IκB kinase epsilon (IKKε), has been reported to effectively improve obesity-related metabolic dysfunctions in mice models, its molecular mechanism has not been fully investigated. This study was designed to investigate the effects of amlexanoX on in vitro nonalcoholic steatohepatitis (NASH) model induced by treatment of palmitic acid (PA, 0.4 mM), using a trans-well co-culture system of hepatocytes and Kupffer cells (KCs). Stimulation with PA signifi- cantly increased the phosphorylation levels of TBK1 and IKKε in both hepatocytes and KCs, suggesting a potential role of TBK1/IKKε in PA-induced NASH progression. Treatment of amlexanoX (50 μM) showed significantly reduced phosphorylation of TBK1 and IKKε and hepatotoXicity as confirmed by decreased levels of lactate de- hydrogenase released from hepatocytes. Furthermore, PA-induced inflammation and lipotoXic cell death in he- patocytes were significantly reversed by amlexanoX treatment. Intriguingly, amlexanoX inhibited the activation of KCs and induced polarization of KCs towards M2 phenotype. Mechanistically, amlexanoX treatment decreased the phosphorylation of interferon regulator factor 3 (IRF3) and NF-κB in PA-treated hepatocytes. However, decreased phosphorylation of NF-κB, not IRF3, was found in PA-treated KCs upon amlexanoX treatment. Taken together, our findings show that treatment of amlexanoX attenuated the severity of PA-induced hepatotoXicity in vitro and lipoapoptosis by the inhibition of TBK1/IKKε-NF-κB and/or IRF3 pathway in hepatocytes and KCs.

1. Introduction

Non-alcoholic fatty liver disease (NAFLD), a broadly defined term for fatty liver-related chronic liver disease, is one of the most serious health problems in the world (Vernon et al., 2011). NAFLD encompasses a disease spectrum ranging from simple steatosis, non-alcoholic steato- hepatitis (NASH), liver fibrosis, cirrhosis, and finally end-stage hepato- cellular carcinoma (HCC). Among these, NASH is characterized by hepatocyte death, inflammation, steatosis, and fibrosis, which is a sig- nificant risk factor for the development of cirrhosis and HCC. Several LipotoXicity and its related lipoapoptosis are considered as a key element during the progression of NASH (Benedict et al., 2017). EXcess accumulation of non-esterified or free fatty acid (FFA) in hepatocytes overwhelms the capacity of the liver to utilize and store FFAs as tri- glycerides and thereby induces lipotoXic hepatocellular death (Chen et al., 2019). Although accumulating evidence suggests that diet and physical activity can manage the early stage of NAFLD/NASH by attenuating central obesity and insulin resistance (Oliveira et al., 2016), it is necessary to expand pharmacological options for management of patients with NASH.

Activation of TRAF associated NF-κB activator (TANK)-binding kinase 1 (TBK1) and IκB kinase epsilon (IKKε) has been studied extensively concerning their functions in promoting the phosphorylation of inter- feron regulatory factor 3 (IRF3) and IRF7, and subsequent nuclear translocation, leading to transcriptional upregulation of type I interferon (IFN) in the innate immune response (Abe and Barber, 2014; Cai et al., 2014; Wang et al., 2020). Recent evidence has indicated that increased activity of TBK1 and IKKε is associated with the onset of metabolic dysfunction and its related pathology. For example, Cruz et al. reported that IKKε is one of the key factors in the induction of insulin resistance in the hypothalamus, and its inhibition reverses obesity (Weissmann et al., 2014). Consistently, reduced metabolic dysfunction induced by a high-fat diet (HFD) feeding was observed in TBK1 mutant mice (Cruz et al., 2018). Also, IRF3, a downstream target of TBK1 and IKKε, plays an important role in early alcoholic liver disease through endoplasmic re- ticulum stress-mediated hepatocyte apoptosis independent of type I IFN (Petrasek et al., 2013). In addition, a recent study noted that activation of the stimulator of interferon genes (STING)-IRF3 pathway promotes NAFLD development and progression by inducing hepatocyte apoptosis and inflammation, and by disturbing metabolic homeostasis (Qiao et al., 2018). Furthermore, Kumari et al. showed that IRF3 enhances white adipose tissue inflammation and insulin resistance, and represses its browning (Kumari et al., 2016). Furthermore, while mounting evidence revealed that TBK1 and IKKε are not required for NF-κB activation (Guo and Friedman, 2010), many studies have demonstrated that TBK1 and IKKε are involved in the activation of NF-κB signaling pathway (Akira and Takeda, 2004; Hu et al., 2004; Qi et al., 2019).

AmlexanoX is an anti-inflammatory and anti-allergic immunomod- ulatory, used for the treatment of recurrent aphthous ulcers, allergic rhinitis and asthma (Bell, 2005; Makino et al., 1987). Recently, amlex- anoX has been widely used to control metabolic diseases in several studies. Reilly and colleagues indicated that the administration of amlexanoX, an inhibitor of TBK1 and IKKε, decreased the metabolic disorders in obese mice (Reilly et al., 2013). Another recent study indicated that amlexanoX can reverse NASH severity through the inhi- bition of IKKε in hepatic stellate cells (He et al., 2019). However, mo- lecular details underlying cellular levels and the manner by which amlexanoX regulates inflammation and lipid accumulation in hepatic cells remain uncertain. Moreover, the impacts of amlexanoX on the lipoapoptosis of hepatocyte have not yet been determined.

Therefore, we herein investigated the effects of amlexanoX in NASH progression in vitro by using a well-known co-culture system of primary mouse hepatocytes and Kupffer cells (KCs). Using this model, we eval- uated whether the treatment of amlexanoX affects hepatocyte meta- bolism, inflammation, and lipoapoptosis. Furthermore, the impacts of amlexanoX on the activation of KCs were also investigated since their activation contributes to the NASH progression (Park et al., 2016).

2. Materials and methods
2.1. Cell isolation and culture

C57BL/6 N male mice (7-week-old, obtained from Samtako Bio Korea Inc., Osan, Korea) were employed to isolate primary hepatocytes and KCs as previously described (Kim et al., 2018; Roh et al., 2018). Briefly, liver tissues were digested with type I collagenase (1 mL/min) perfusion. Hepatocytes were separated by centrifugation (50 g, 3 min), and nonparenchymal cells (NPCs) fractions were purified by centrifu- gation in Percoll gradient. Then, the KCs were further isolated from NPCs using magnetically activated cell sorting with antibody to F4/80 antigen-conjugated microbeads. Primary hepatocytes were plated in multi dishes (Corning Inc., Corning, NY, USA) coated with rat tail type I collagen (Corning Inc.). In the co-culture system, hepatocytes were plated into 24-well type I collagen-coated plates and KCs were plated in cell-culture inserts with 0.4 μm pore polycarbonate membrane (Corning Inc.) at a 2:1 (Hepatocyte: KC) ratio. Isolated cells were cultured in DMEM with 10 % fetal bovine serum (Welgene, Seoul, Korea) and Antibiotic- Antimycotic solution (Welgene). After cell attachment, cells were incubated under serum starvation overnight. Then, the cells were treated with palmitic acid (PA, 0.4 mM) or vehicle with or without 50 μM amlexanoX for 24 h. The cells and cell supernatants were collected for further experiments. To demonstrate that the PPARγ modulation of amlexanoX results in M2 polarization of KCs, primary hepatocytes and KCs were cultured with amlexanoX and/or SR202 (a PPARγ antagonist, 50 μM) for 24 h.

2.2. Identification of hepatocytes and KCs

The purity of cultured hepatocytes and KCs were determined by morphology in combination with F4/80 staining. Cells were identified by F4/80 staining with immunofluorescence and flow cytometry using monoclonal anti-F4/80 antibody [BM8] conjugated Phycoerythrin (PE) (Thermo Fisher Scientific Inc., Waltham, IL, USA). Immunofluorescence was performed according to the instruction. In detail, Cells were seeded in a 4-well chamber slides at a density of 5 104 cells/mL. Then the cells were fiXed by 4% paraformaldehyde for 10 min and then blocked with
10 % normal goat serum for 30 min. The cells were then incubated with F4/80 antibody (1 μg/mL), overnight at 4 ◦C. After washing with PBS, cells were covered with the mounting medium which contains DAPI that was used to stain the cell nuclei (blue), and the slides were viewed by digital imaging software (analySIS TS, Olympus Corp.).

Supplementary Fig. 1A and B showed the morphological character- izations of cultured hepatocytes and KC. Additionally, the results of immunofluorescence showed that the percentage of F4/80 positive KCs accounted for 96.20 %. Also, we found that F4/80 barely be detected in cultured hepatocyte (Supplementary Fig. 1C).

2.3. PA preparation

Albumin-conjugated PA was prepared as previously described (Zhou et al., 2018). A stock solution of PA (Sigma-Aldrich, St. Louis, Mo, USA) was prepared in 10 % solution of FA-free bovine serum albumin (BSA) (Rocky Mountain Biologicals, Inc., Missoula, USA) and conjugated at 55
◦ C to yield an 8 mM solution. The final molar ratio of free FA to BSA was 5.2:1. The stock solution of PA was deposited at —40 ◦C.

2.4. Lactate dehydrogenase (LDH) assay

Primary mouse hepatocytes (1 105 cells/well) and KCs (5 104 cells/well) were seeded in the trans-well co-culture plate. LDH assay (Sigma-Aldrich) was used to determine cell death according to the manufacturer’s instruction. The absorbance of samples was measured by an EMax spectrophotometry (Molecular Devices, Sunnyvale, CA, USA) at a wavelength of 490 nm.

2.5. Enzyme-linked immunosorbent assay (ELISA)

The protein concentration of pro-inflammatory cytokines in the cell supernatants was measured using commercial ELISA kits (eBioscience, San Diego, CA, USA) according to the manufacturer’s instructions.

2.6. Triglyceride (TG) measurement in hepatocytes

Hepatocytes and KCs were co-cultured in the presence of PA with or without amlexanoX for 24 h. To measure the cellular levels of TG in hepatocytes, the culture medium was removed and hepatocytes were washed with PBS 3 times. After the complete removal of PBS, hepato- cellular lipids were extracted from cell homogenates by using chloro- form/methanol (2:1). The cellular TG content was quantified by using the AM202-K spectrophotometric assay kit (ASAN Pharmaceutical, Hwasung, Korea). The absorbance of samples was measured by the aforementioned EMax spectrophotometry at the wavelength of 550 nm.

2.7. Quantitative real-time PCR (qRT-PCR)

The procedures of real-time PCR were performed as previously described (Song et al., 2019; Zhou et al., 2020b), the brief procedures being as follows. Total RNA was isolated from cells using the Easy-Spin Total RNA extraction kit (GeneAll, Seoul, Korea) according to the manufacturer’s instructions. Following incubation with the DNase I-containing RNase inhibitor (Toyobo, Osaka, Japan), the RNA was converted to complementary DNA sequences by reverse transcriptase using the ReverTra Ace® qPCR RT Master MiX (Toyobo) to perform the qPCR analysis. After completion of the reaction, specificity was verified by melting curve analysis. The glyceraldehyde-3-phosphate dehydro- genase (GAPDH) was used as a constitutive control and the gene expression analysis was performed using Bio-Rad CFX Manager version
3.1 software. Table 1 lists the primers.

2.8. Oil-red O staining

Cultured hepatocytes were washed with cold PBS and fiXed in 4 % paraformaldehyde for 15 min under room temperature (RT). Following 2 washes with ddH2O, the cells were incubated in 60 % isopropanol for 1 min. Subsequently, cells were dipped with Oil Red O (Sigma-Aldrich) for 10 min at RT and washed several times with ddH2O to remove the excess stain. Then, hematoXylin was employed to stain cell nuclei. Pictures were taken using a light microscopy (BX-53 F, Olympus Corp., Tokyo, Japan). The Oil-red O positive area was evaluated using digital imaging software (analySIS TS, Olympus Corp.).

2.9. Immunoblot analysis

As previously described (Zhou et al., 2020a), hepatocytes or KCs were harvested and directly lysed in cold extraction buffer (RIPA,
iNtRON Biotechnology, Inc., Seongnam, Korea) for 30 min. Following centrifugation at 13,000 g at 4 ◦C for 15 min, the protein concentration in the supernatants was measured using Pierce BCA Protein Assay kit (Thermo Fisher Scientific Inc., Waltham, IL, USA) according to the manufacturer’s protocol. Once the equal amount of protein samples were prepared, they were then subjected to 10 % sodium dodecyl sulfate-polyacrylamide gel electrophoresis and subsequently electro-transferred to polyvinylidene difluoride membranes. The mem- branes were blocked with 5 % BSA diluted in Tris-buffered saline (20 Science, Pittsburgh, PA, USA). The protein density was quantified with ImageQuant™ TL software.

2.10. Lipid peroxidation assay

The levels of lipid peroXidation are an important marker of oXidative stress conditions. Malondialdehyde (MDA) contents in hepatocytes were measured by using an OXiSelect TBARS Assay Kit (Cell Biolabs, Inc., San Diego, CA, USA).

2.11. Statistical analysis

All experiments were performed in triplicate using different hepa- tocytes and KCs preparations. Statistical analyses were performed using the one-way ANOVA test followed by Tukey–Kramer’s post hoc test for comparing the control group to the others. P < 0.05 was considered statistically significant. Mean values using different superscripts were statistically significant. Furthermore, differences between the two groups were analyzed by the Student’s t-test and P < 0.05 was considered as statistically significant. All data are represented as the mean ± SEM. 3. Results 3.1. Increased activities of TBK1 and IKKε were found in PA-stimulated hepatocytes and KCs It has been reported that activities of TBK1 and IKKε are elevated in the livers of mice fed with HFD (Chiang et al., 2009; Reilly et al., 2013), and Cho et al. indicated that pTBK1 level is elevated in PA-treated he- patocytes (Cho et al., 2018). Consistent with these findings, the protein levels of pTBK1 and pIKKε were markedly increased in hepatocytes or KCs upon PA stimulation (Fig. 1A and B). These findings suggest that the activation of TBK1 and IKKε might affect the progression of PA-induced hepatotoXicity. 3.2. Inhibition of TBK1 and IKKε ameliorates PA-induced hepatotoxicity in vitro. We sought to evaluate the role of TBK1 and IKKε in the progression of PA-induced hepatotoXicity. Our previous study demonstrated that KCs activation exacerbates PA-induced lipotoXicity and inflammation in hepatocytes (Zhou et al., 2018). We also confirmed that the presence of KCs further increased the PA-induced hepatotoXicity compared with hepatocytes alone (Supplementary Fig. 2A). Consistently, after PA treatment, the protein levels of pTBK1 and pIKKε were significantly higher in the hepatocytes co-cultured with KCs than those in hepatocytes alone (Fig. 2A). Also, increased protein levels of pIKKε but not pTBK1 were observed in co-cultured KCs compared to KCs alone following PA treatment (Fig. 2B). Fig. 1. Increased activities of TBK1 and IKKε are found in PA-stimulated hepatocytes and KCs.Co-cultured primary hepatocytes (1 × 105 cells/well) and KCs (5 × 104 cells/well) were treated with 0.4 mM PA or vehicle for 24 h. The protein levels of pTBK1, TBK1, pIKKε, and IKKε in (A) hepatocytes and (B) KCs were measured by Western blot and the protein ratios of pTBK1/TBK1 and pIKKε/IKKε were quantified (n = 4/ group). (C) The mRNA expression levels of TBK1 and IKKε in hepatocytes and KCs were determined by real-time PCR (n = 8/group). Data are expressed as the mean ± SEM per group. Two-tailed Student’s t-test, *P < 0.05, **P < 0.01. Next, the concentrations of amlexanoX without causing cytotoXicity on primary mouse hepatocytes and KCs were determined. Base on MTT assay, amlexanoX at a concentration of 100 and 200 μM caused sig- nificant cytotoXicity in hepatocytes and KCs, respectively (Supplemen- tary Fig. 2B and C). Thus, several doses of amlexanoX (0, 12.5, 25, and 50 μM) were used and we found that amlexanoX at a dose of 50 μM significantly decreased PA treatment-induced LDH leakage in co-culture supernatants (Fig. 2C). Therefore, we chose 50 μM amlexanoX for further studies and analysis. Furthermore, we found that amlexanoX signifi- cantly decreased the protein levels of pTBK1 and pIKKε in vehicle- and PA-treated hepatocytes and KCs (Fig. 3A and B). Collectively, these findings indicate that amlexanoX could attenuate PA-mediated hepato- toXicity possibly via inhibiting the phosphorylation of TBK1 and IKKε. 3.3. Amlexanox treatment decreased PA-induced inflammatory responses Since inflammation is closely associated with the pathogenesis of NASH (Farrell et al., 2012), we next evaluated the effects of amlexanoX on PA-induced inflammatory responses. The protein levels of TNFα, IL-6, and IL-1β in cell culture supernatants were significantly decreased upon Transcription factor PPARα directly controls various steps of hepatic lipid metabolism, including the expression of CPT-1A (Mandard et al., 2004). Consistent with this result, decreased mRNA level and protein level of PPARα induced by PA treatment were recovered after amlex- anoX treatment (Fig. 6C and D). Furthermore, the gene expression levels of other FA oXidation-related enzymes CYP2e1, CYP3a11, and CYP4a10 were measured. As shown in Fig. 6E, amlexanoX treatment significantly decreased the mRNA expression levels of CYP2e1 in PA-stimulated he- patocytes. Although CYP3a11 was markedly decreased by amlexanoX in hepatocytes without PA treatment, such effects were not observed in PA-treated hepatocytes.The Western blot results further revealed that amlexanoX treatment decreased the protein level of CYP2e1 in PA-treated hepatocytes (Fig. 6F). These findings suggest that amlexanoX improves FA oXidation in PA-treated hepatocytes. 3.5. Amlexanox treatment attenuates PA-induced lipoapoptosis of hepatocytes Since lipid peroXidation and its associated lipotoXic cellular damage contributed to the progression of NASH (Neuschwander-Tetri et al., 2010; Qi et al., 2020), we evaluated the level of lipid peroXidation by measuring MDA contents in hepatocytes. Increased MDA levels induced by PA treatment were significantly reduced by amlexanoX treatment (Fig. 7A). Also, mRNA expression levels of inducible nitric oXide syn- thase (iNOS) were decreased with a concurrent elevation of heme oXygenase-1 (HO-1, a marker of anti-oXidative stress) in PA-treated hepatocytes with amlexanoX treatment, resulting in an overall reduc- tion of oXidative stress (Fig. 7B). These results were further supported by a recent study showing that oXidative stress-related reactive oXygen amlexanoX treatment in PA-treated co-culture system (Fig. 4A). species can induce lipid peroXidation (Barrera, 2012). Consistent with these results, the mRNA expression of these inflamma- Consistent with the above (Fig. 2C) showing reduced PA-induced tory cytokines was changed in the same pattern by amlexanoX treatment hepatocyte death by amlexanoX, the representative figures also (Fig. 4B). These findings indicate that amlexanoX could decrease in- flammatory responses in the in vitro co-culture NASH model. 3.4. Amlexanox-treated hepatocytes were protected from PA-induced intracellular accumulation of lipids To evaluate whether amlexanoX treatment has an impact on PA- induced intracellular lipid accumulation in hepatocytes, we first per- formed Oil-red O staining and quantified the Oil Red O-positive area. Markedly increased lipid accumulation was observed in PA-treated hepatocytes compared with that in vehicle-treated hepatocytes, and such effects were significantly decreased by amlexanoX treatment in PA- treated hepatocytes (Fig. 5A and B). Additionally, amlexanoX signifi- cantly decreased the TG content in PA-treated hepatocytes (Fig. 5C). Next, we evaluated the hepatic mRNA expression levels of de novo lipogenesis-associated genes including FA synthase (FASN), acetyl CoA carboXylase (ACCα), diacylglycerol O-acyltransferase1 (DGAT1), and sterol regulatory element-binding protein-1c (SREBP1c). In PA-treated hepatocytes, significantly decreased levels of these lipogenesis-related genes were observed in the amlexanoX-treated group compared to the vehicle-treated group (Fig. 5D). These findings indicate that amlexanoX suppressed PA-induced lipid accumulation in hepatocytes. Since intracellular TG is either stored or released in combination with very-low-density lipoprotein (VLDL) particles (Browning et al., 2004), we examined the gene expression levels of VLDL synthesis/export-related genes such as apolipoprotein B (ApoB) and microsomal TG transfer protein (MTTP). There was no significant dif- ference between amlexanoX- and vehicle-treated groups in PA-stimulated hepatocytes (Fig. 7D). In line with these results, Western blot analysis showed that the protein level of cleaved caspase 3 was significantly decreased by amlexanoX treatment in PA-stimulated he- patocytes (Fig. 7E). Consistently, amlexanoX significantly decreased the expression level of Bax (pro-apoptosis protein) with a concomitant in- crease of expression level of Bcl2 (anti-apoptosis protein) in hepatocytes treated with PA (Fig. 7E and Supplementary Fig. 3A). Although the protein ratio of Bax/Bcl-2 was markedly elevated by PA treatment, amlexanoX treatment markedly decreased this ratio in PA-treated he- patocytes (Fig. 7E). In line with these results, a similar pattern was observed in the immunofluorescence result of cleaved caspase 3 (Sup- plementary Fig. 3B). These results indicate that amlexanoX treatment protects hepatocytes from PA-induced injury. 3.6. Amlexanox inhibits the activation of KCs and induces polarization of KCs towards M2 phenotype Increasing evidence has revealed that KCs are activated in response PA-stimulated hepatocytes (Fig. 6A). Next, we examined whether amlexanoX enhanced the degradation of FAs through β-oXidation. PA treatment notably decreased the mRNA level of carnitine palmitoyl- transferase 1A (CPT-1A) in hepatocytes, whereas amlexanoX restored the mRNA level of CPT-1A in PA-treated hepatocytes (Fig. 6B). 3.7. Inhibition of TBK1 and IKKε differentially affects the activation of IRF3 and NF-κB in PA-treated hepatocytes and KCs Since IRF3 and NF-κB are downstream targets of TBK1 and IKKε, we determined whether amlexanoX treatment involves the activation of IRF3 and NF-κB in both cell types. In line with a previous study (Qiao et al., 2018), PA stimulation significantly increased the phosphorylation of IRF3 and NF-κB in hepatocytes. AmlexanoX treatment significantly decreased the activation levels of IRF3 and NF-κB in PA-treated hepatocytes (Fig. 9A). Interestingly, amlexanoX treatment substantially reduced the protein level of pNF-κB, not pIRF3 in PA-treated KCs. Fig. 5. Amlexanox treatment protects hepatocytes from PA-induced intracellular accumulation of lipids. (A) To investigate lipid accumulation in co-cultured hepatocytes, hepatocytes from each group were subjected to Oil-red O staining. Magnification: 400 × . (B)The Oil-red O-positive area was analyzed. (C) The effects of amlexanoX on PA-induced TG accumulation were determined (D) The mRNA expressions of lipogenesis- related genes in hepatocytes were measured using real-time PCR. Data are expressed as the mean ± SEM per group. n = 6 for each group. Two-tailed Student’s t-test, *P < 0.05, **P < 0.01. 4. Discussion AmlexanoX is well known to protect against the pathologic effects of decreased mRNA expression level of IFNβ was observed in PA-treated hepatocytes, but not in PA-treated KCs (Fig. 9C and D). Altogether, these findings suggest that the inhibition of NF-κB and/or IRF3 signaling in both cell types might contribute to the protective effects of amlexanoX on PA-induced inflammation and intracellular accumulation of lipids. Fig. 6. Inhibition of TBK1 and IKKε affects lipid metabolism in PA-treated hepatocytes. (A) The mRNA expression levels of VLDL synthesis/export-related genes were determined using real-time PCR. (B) The mRNA expression levels of CPT-1A were determined using real-time PCR. The (C) mRNA expression level and (D) protein level of PPARα were determined by real-time PCR and Western blot, respectively. (E) The mRNA expression levels of CYP2e1, CYP3a11, and CYP4a10 were determined using real-time PCR. (F) The protein level of CYP2e1 was determined by Western blot. Data are expressed as the mean ± SEM per group. real-time PCR: n = 6; Western blot: n = 4. Two-tailed Student’s t-test, *P < 0.05. Fig. 9. Inhibition of TBK1 and IKKε differentially affects the activation of IRF3 and NF-κB in PA-treated hepatocytes and KCs.The protein levels of IRF3, NF-κB, and their phosphorylation in (A) hepatocytes and (B) KCs were analyzed by Western blot and relatively quantified. The mRNA expression levels of IFNβ in (C) hepatocytes and (D) KCs were determined by real-time PCR. Data are expressed as the mean ± SEM per group (real-time PCR: n = 6; Western blot: n = 4. Two-tailed Student’s t-test, *P < 0.05, **P < 0.01). Collectively, the effect of NF-κB on hepatocyte death seems to depend on inclusions in PA-treated hepatocytes. These protein inclusions provoke the damaging agent and may vary among different models of liver hepatocellular oXidative stress (Cho et al., 2018). Furthermore, injury. In the present study, inhibition of TBK1 and IKKε by amlexanoX significantly decreased the activation of NF-κB in PA-treated hepatocytes. The reduced activation of NF-κB resulted in decreased inflam- matory responses in hepatocytes, thus ameliorating hepatocyte death. We investigated the role of TBK1 and IKKε in KCs. Our previous re- sults showed that TBK1 and IKKε are highly expressed in KCs (Zhou et al., 2020c), which suggests a role for TBK1 and IKKε in KCs in the progression of PA-induced metabolic disorders. A previous study showed that activated KCs play a critical mediator in the development of NASH (Yu et al., 2019). KCs were the cells initially responding to the injured hepatocytes, leading to activation of NF-κB and inflammation, involving induction of pro-inflammatory cytokines (TNFα, IL-1β, and IL-6), chemokines, and thereby inflammatory cell recruitment (Tosel- lo-Trampont et al., 2012). Furthermore, recent studies showed that early depletion of KCs prevented the development of NASH in rodents (Yang et al., 2012; Zeng et al., 2015). Thus, we also tried to focus on KCs and found that amlexanoX treatment reduced the activities of TBK1 and IKKε in PA-stimulated KCs. Consistently, decreased activation of NF-κB, but not IRF3, was found in PA-stimulated KCs with amlexanoX treatment. In line with these findings, amlexanoX treatment significantly inhibited the activation of KCs, their inflammatory responses, and induced their po- larization towards the M2 phenotype. Hence, since NF-κB is widely known as a key transcription factor related to M1 macrophage activation (Coscia et al., 2010; Li et al., 2016; Wang et al., 2019; Xu et al., 2019), inhibition of TBK1/IKKε-NF-κB signaling by amlexanoX treatment, at least in part, contributes to suppressing the activation of PA-treated KCs and their inflammation responses. Recently, PPARγ activation was found to be important in M2 macrophages polarization (Kim et al., 2019; Lehrke and Lazar, 2005; Villanueva and Tontonoz, 2010). In the present study, the increased expression level of PPARγ was found in PA-stimulated KCs upon amlexanoX treatment. However, this result is not consistent with the previous finding that shows that amlexanoX administration significantly decreases the expression levels of PPARγ in the livers of mice fed with an HFD (Reilly et al., 2013). Considering that the hepatocyte is the main cellular component in the liver, and significantly decreased PPARγ expression level was found in PA-stimulated macrophage-derived IKKε is involved in the regulation of insulin sensi- tivity and chronic inflammation in obese mice (Chiang et al., 2009). Therefore, in the future study, it would be valuable to investigate the single function or cell type-specific role of each kinase in the patho- genesis of NASH. Conclusively, we provided the evidence that amlexanoX offers pro- tective effects on PA-induced lipotoXicity in vitro. Thus, these findings could provide molecular evidence to elucidate the hepatic benefit of amlexanoX on NASH treatment in clinical trials. Since more effective therapeutic strategies are urgently needed for treating NASH, we hope that the inhibition of TBK1 and IKKε by amlexanoX may prove a promising BAY-985 therapeutic strategy to cure NASH.