P38 MAPK Inhibits Autophagy and Promotes Hepatic Stellate Cells Inammatory Responses via Atg13 During Acute-On-Chronic Liver Failure

Inammation plays a critical role in disease pathogenesis of acute-on-chronic liver failure (ACLF). Atg13 is a vital regulatory component of ULK1 complex, which plays an essential role in initiation of autophagy, and autophagy has been connected to hepatic inammation. The aim of this study is to evaluate how autophagy regulates hepatic stellate cells (HSCs) inammatory responses via Atg13 during ACLF. Clinical data were collected from ACLF patients, and surgical resected paran embedded human ACLF liver tissues specimens were collected. Inammation and autophagy were investigated by immunoblot analysis in HSCs treated with lipopolysaccharide (LPS). Co-immunoprecipitation were used to investigate interaction of Atg13 and ULK1. Our data exhibit that serum LPS is positively associated with disease severity in ACLF patients, and p38 MAPK is overexpressed in ACLF liver tissues. Inammatory factors are up-regulated via the activation of p38 MAPK and the inhibition of the autophagy in LX-2 cells. Furthermore, we illuminate in the vitro study that LPS triggers p38 MAPK activity, resulting in phosphorylation of Atg13, inhibition of Atg13-ULK1 interaction and autophagy. This study highlights a molecular mechanism that LPS promotes inammation through inhibition of autophagy in HSCs via Atg13, and provides a new understanding into the mechanistic of inammatory process of severe hepatitis and a novel strategy for ACLF treatment. growth phase were treated with 1µg/ml LPS in a time-dependent manner (0-8h); (E, F) LX-2 cells in the logarithmic growth phase were treated with 1µg/ml LPS together with or without 50nM Rapamycin or 20μM SB203580 for 4h for the detection of autophagy; (G, H) LX-2 cells in the logarithmic growth phase were treated with 1µg/ml LPS together with or without 200nM Balomycin A in a time-dependent manner (0-8h) for the detection of autophagy. Then, autophagy and inammation were evaluated by immunoblotting with specic antibodies, as indicated. The P62, LC3, NLRP3 and pro-IL1 levels were quantied by densitometry and normalized to the actin level. Secretion levels of IL-1β were detected by ELISA.


Introduction
Liver failure, including acute, chronic, and acute-on-chronic liver failure, is a rare but dramatic clinical syndrome characterized by massive hepatocyte death and overactivation of hepatic in ammation [1]. Acute-on-chronic liver failure (ACLF), characterized by an acute deterioration of liver function in patients with pre-existing chronic liver disease, usually results in hepatocellular dysfunction and carries a high mortality rate [2]. Apart from liver transplantation, few effective therapies are available, and ACLF continues to be a huge therapeutic challenge [3]. The precise molecular mechanisms for the pathogenesis of ACLF have not been clari ed, exploring ACLF-associated molecules may enable the development of strategies to improve the prognosis for patients with ACLF.
Hepatic stellate cells (HSCs) are resident mesenchymal cells that have features of broblasts and pericytes, and account for 15% of total resident cells in normal human liver [4]. HSCs are key nonparenchymal components in the sinusoid with multiple functions [5]. Some recent studies have revealed that activated HSCs may release in ammatory cytokines like interleukin (IL)-1β and IL-18, and HSCs in ammation has been shown participated in the pathogenesis of several liver diseases: HSCs of murine or human origin produce in ammatory cytokines promoting hepatocellular carcinoma and immune-mediated hepatitis[6], our previous study has shown that HSCs in ammation participates in disease pathogenesis of ALF [7].
Autophagy is a conserved process by which cytoplasmic components, including damaged proteins and organelles, are degraded by lysosomes [8]. Autophagy and in ammation are highly intertwined cellular process. Autophagy suppresses proin ammatory process and in ammasome activity [9]. Decreased autophagy was found leading to elevated hepatic in ammation promoting the progression of alcoholic liver disease [10], and LPS & D-GalN induced liver injury in mice [11]. In ammatory cytokines also function to reciprocally control autophagy [9]. However, the mechanisms by which in ammatory signals speci cally relieve the negative suppression by autophagy on HSCs in ammation during ACLF remains unknown.
Mitogen-activated protein kinase (MAPK), including p38, ERK and c-JNK are members of a ubiquitous protein serine/threonine kinase family responsible for signal transduction in eukaryotic organisms [12].
MAPK activation is implicated in the production of many in ammatory mediators [13]. Mammalian genomes encode four distinct p38 MAPK isoforms, including α, β, γ and δ. MAPK p38α and β are commonly activated by stressful or pro-in ammatory stimuli [14]. Most studies of p38 MAPK have focused on its functions in in ammatory cells during the pathogenesis of in ammation dependent diseases, including rheumatoid arthritis, Crohn's disease, psoriasis and asthma [15,16].
In the present study, we show that p38 MAPK plays an essential role in relieving autophagic control in response to in ammatory signal. We found elevated expression of p38 MAPK in liver tissues from patients with ACLF. Stimulation of LPS inhibits autophagy via p38 MAPK, this inhibition is necessary for LPS-induced in ammasome activation in LX2 cells. We show that p38 MAPK directly interacts with Atg13, phosphorylates Atg13 and reduces Atg13-ULK1 interaction.

Patients
From January 2016 to September 2017, a total of 56 patients diagnosed with ACLF were enrolled in our study at the First A liated Hospital of Xi'an Jiaotong University, Shaanxi, China. All participants provided written informed consent, depending on the patient's altered mental status, and the study was approved by the Research Ethics Committee of the First A liated Hospital of Xi'an Jiaotong University. Patients were diagnosed with ACLF based on the criteria of Asian Paci c Association for the Study of the Liver (APASL): 1) serum bilirubin ≥ 85 mol/L; 2) INR ≥ 1.5 or prothrombin activity ≤ 40%; 3) any degree of encephalopathy and/or clinical ascites within 4 weeks; 4) and an evidence of ongoing chronic liver diseases. Patients who were diagnosed with ACLF and aged 18 to 75 years were included. We calculated the Model for End-Stage Liver Disease (MELD) score using the standard formula: 11.2*ln (INR) + 9.57*ln (creatinine, in mg per decilitre) + 3.78*ln (bilirubin, in mg per decilitre), with a lower limit of 1 for all variables. During the same period, age-and sex-matched healthy and cirrhotic participants were recruited as controls.

Estimation of LPS and IL-1β
Serum LPS levels were measured by using a Limulus Amebocyte Lysate (LAL) commercial test kit (Xiamen Houshiji, Ltd, Xiamen, China), and Secretion IL-1β levels were detected by utilizing Human IL-1 beta Quantikine ELISA Kit (SLB50, R&D Systems) according to the manufacture's protocol. Samples and standards were run in duplicate.

Histological sampling
We collected surgical resected para n-embedded human ACLF liver tissues specimens (5 cases) and cirrhotic liver tissue specimens (5 cases) from the Department of Pathology, the First A liated Hospital of Xi'an Jiaotong University, with the approval of the Institutional Review Board. Immunoreactions were performed on selected liver sections. Antigens were detected by one of the following primary antibodies, followed by appropriate secondary antibodies: anti-p38 (#8690, Cell Signaling Technology, Danvers, MA, USA). The slides were then observed under a Nikon Eclipse microscope (Tokyo, Japan) coupled to a digital camera.

Cells culture
The human hepatic stellate cells LX2 cells, HEK-293 cells were cultured in Dulbecco's modi ed Eagle's medium and THP-1 cells were cultured in RPMI-1640 medium supplemented with 10% fetal bovine plasma and 2mM L-glutamine at 37°C in a 95% air, 5% CO 2 -humidi ed atmosphere. Cells were trypsinized, and 5×10 5 cells were seeded onto plastic dishes and then treated with LPS, SB203580, Rapamycin or Ba lomycin A1.

Co-immunoprecipitation assay for protein binding
Cells lysates were prepared in lysis buffer, and incubated with ProteinA/G-Sepharose beads at 4 ℃ for 3 hours. Pre-cleared lysates were incubated with appropriate antibody at 4 ℃ for 12 hours with gentle rotation. Protein-A/G-Sepharose beads were added and incubated for 3 hours. Immunoprecipitants were collected by centrifugation, washed ve times with lysis buffer, heated at 100 ℃ for 5 min and subjected to SDS-PAGE.

Immunoblotting
Protein extracts were prepared from cells by RIPA Lysis Buffer supplemented with Complete EDTA-free protease inhibitor cocktail tablets (Roche Applied Science, Basel, Switzerland) and phosphatase inhibitor cocktails (Sigma-Aldrich). Protein samples (50 µg) were loaded onto SDS-PAGE gels and transferred onto PVDF membranes. After blocking in 5% evaporated milk at room temperature for 2 h, the membranes were then incubated with the indicated primary antibodies in 5% evaporated milk in TBS plus 0.1% Tween 20 overnight at 4°C. Signals were developed using a chemiluminescent substrate and visualized through X-ray lms.

Statistical analysis
The results are expressed as the means ± standard deviation. Statistical analysis was performed using SPSS software 13.0 (SPSS, Inc., Chicago, IL, USA). The Shapiro-Wilk test and Levene statistic were used to evaluate the normality and homogeneity of the variance, respectively. According to the situation, t-tests or Mann-Whitney U tests were used to evaluate differences between two groups; correlations between two quantitative groups were analyzed with Pearson or Spearman correlation tests. The χ2 test was used for comparisons between two groups. The reported P-values are two-sided, and P-values < 0.05 were considered statistically signi cant.

Serum LPS level increased in ACLF
Liver failure is an in ammation-mediated hepatocellular injury process, LPS levels in serum is elevated in patients with ALF and ACLF due to increased gut permeability. Here we investigated serum LPS levels in patients with ACLF (Table 1). We found a signi cant increase in serum LPS concentration in patients with ACLF compared to cirrhotic patients (177.5 ± 64.52 vs 5.830 ± 0.3019, P < 0.05) (Fig. 1A). Previously, we found that MELD score > 25 was associated with short-term mortality in ALF patients. We then divided these ACLF patients into a low-risk group (MELD score ≤ 25) or high-risk group (MELD score > 25). Patients in the high-risk group presented with higher LPS levels compared to patients in the low-risk group (217.0 ± 70.56 vs 83.68 ± 31.79, P < 0.05) (Fig. 1B). Furthermore, a positive correlation was observed when the LPS level was correlated with the MELD score (R 2 = 0.5164, P < 0.01) (Fig. 1C).

Expression of p38 MAPK in ACLF
LPS is known to engage p38 MAPK in inducing in ammatory response. We wondered whether this pathway participating in the pathogenesis of ACLF and tested in vivo expression of p38 MAPK in liver tissues from ACLF patients. Specimens displayed positive immunoreactions for p38 MAPK in patients with ACLF, and expressions of α-SMA were also found in specimens ( Fig. 2A-H).

LPS promotes in ammation in HSCs through p38 MAPK
LPS is known to engage in ammatory response in THP-1 cells. We showed here that SB203580, a chemical inhibitor of p38 MAPK effectively suppressed LPS induced phosphorylation of p38 MAPK and NLRP3 in ammasomes activation in THP-1 cells (Fig. 3A, B, C). In the present study, we found that LPS could induce time-dependent activation of p38 MAPK, and SB203580 suppressed this phosphorylation ( Fig. 3D, E) in LX2 cells. We also found that LPS induced a time-dependent activation of in ammatory response in LX2 cells, and the usage of SB203580 signi cantly attenuated LPS induced NLRP3 activation and production of IL-1β (Fig. 3F, G). In addition, we found a reversal of LPS induced accumulation of p62, which means autophagy inhibition, in SB203580 co-treated LX2 cells (Fig. 3F).

LPS inhibits autophagy in HSCs through p38 MAPK
To understand the mechanism of how signals control and release autophagic suppression of in ammation, we rst tested whether LPS modulates autophagy in HSCs. We treated LX2 cells with LPS and found a and dose-dependent decrease in the level of autophagy marker LC3-II, the lipidated form of microtubule-associated protein 1A/1B-light chain 3 (LC3), and an increase in the level of autophagy adapter protein p62, respectively (Fig. 4A, B). Next, we evaluated the time course of autophagy in response to LPS in LX2 cells. LX2 cell were treated with 1µg/ml of LPS, the LC3-II expression decreased, whereas p62 expression rstly increased and then decreased after 4 hours (Fig. 4C, D).
To further con rm that LPS inhibits autophagy in LX2 cell, we used rapamycin, an initiator of autophagy, and Ba lomycin A1 (Baf A1), which is known to block the fusion of autophagosomes with lysosomes, together with LPS. Our data showed that rapamycin inhibits LPS-induced changes in LC3-II and p62 (Fig. 4E). Exposure to Baf A1 signi cantly promoted LPS-induced accumulation of p62, and NLRP3 activation and production of IL-1β (Fig. 5G, H). Furthermore, SB203580 successfully alleviated LPSinduced inhibition of autophagy (Fig. 4F).

p38 MAPK phosphorylates Atg13 and reduces Atg13-ULK1 interaction
In ammatory signal activates p38 MAPK, but its role has not been fully de ned. Atg13 is a key regulator that functions upstream in the autophagic cascade and is regulated via phosphorylation, we tested the possibility that p38 MAPK may regulate autophagy by directly targeting Atg13. Here, we rstly found p38α MAPK may directly interact with Atg13 as we overexpressed MYC-Atg13 and FLAG-p38α MAPK in HEK293 cells, and immune-precipitated with an anti-MYC antibody and blotted the precipitate with an anti-FLAG antibody or performed the immunoprecipitation in reverse. This analysis showed that Atg13 and p38α MAPK directly associated with each other (Fig. 5A). Consistent with the possibility that p38α MAPK may regulate Atg13 activity, we found that co-expression with p38α MAPK causes Atg13 to migrate at a higher molecular weight position. This change in Atg13 migration could be reversed by pretreatment the lysates with SB203580 or λ phosphatase (Fig. 5B, C). In LX2 cells, treatment of LPS led to phosphorylation and increased expression of Atg13 and this phenomenon was reversed by the usage of SB203580 (Fig. 5D).
Atg13 together with Atg101, FIP200 and ULK1 constitute the ULK1 complex, which plays an essential role in the initiation step of autophagy. Atg13 binds to ULK1 when receiving signals of nutrient status, and then recruits downstream autophagy related proteins. We tested whether p38 MAPK regulates the Atg13/ULK1 interaction. After treatment of the p38 MAPK activator anisomycin, the binding of Atg13 to ULK1 was decreased in HEK293 cells expressing MYC-tagged Atg13 and FLAG-tagged ULK1, suggesting that the inhibitory effects of p38 MAPK on autophagy were due to a decreased association of Atg13 with ULK1 that disrupted the initiation of autophagy. (Fig. 5E).

Discussion
It has been known for years that autophagy negatively control in ammasome activity, a decrease in autophagic activity correlates positively with in ammatory response [9]. ACLF is a life-threatening disease, characterized by over activation of hepatic in ammation [17]. Our data show that serum LPS level was signi cantly higher in ACLF due to increased gut permeability, meanwhile, serum LPS was found correlated with MELD score and associated with disease severity in patients with ACLF.
HSCs activation is the central step during liver brogenesis, our previous study revealed that HSCs activation participated in maintaining liver architecture during ALF [18]. Some recent studies have found that HSCs appear to replay in ammation signaling from the sinusoid to parenchyma: HSCs from both humans and rodents produce in ammatory cytokines that promote hepatocellular carcinoma and immune-mediated hepatitis [6]. However, few studies examined the roles of HSCs in hepatic in ammation during liver failure. Our recent work showed that during the pathogenesis of ALF, reactive oxygen species activate the NLRP3 in ammasome and promote in ammation in HSCs. We also revealed that LPS treatment induced reactive oxygen species (ROS) generation in HSCs via mitophagy inhibition [7]. In the present study, we found elevated expression of p38 MAPK in liver tissues from patients with ACLF.
Stimulation of LPS inhibits autophagy via p38 MAPK, and this inhibition is necessary for LPS-induced in ammasome activation in LX2 cells. Our data suggested that usage of SB203580 successfully inhibits LPS induced autophagy inhibition and in ammasome activation.
We found in this present study a central mechanism by which proin ammatory signals exempli ed by LPS relieves the tight inhibitory control exerted by autophagy on the in ammatory process. However, the mechanism through which this proin ammatory signal regulate autophagy in HSCs during ACLF remain elusive. We showed in the present study that in response to proin ammatory signal, p38 MAPK directly interacts with Atg13, phosphorylates Atg13 and reduces Atg13-ULK1 interaction, and then reduces autophagy [19]. Atg13 together with Atg101, FIP200 and ULK1 constitute the ULK1 complex, which plays an essential role in the initiation step of autophagy: receiving signals of nutrient status, recruiting downstream autophagy related proteins and governing autophagosome formation [20]. Because the Atg13-Ulk1 complex is the key upstream regulator of the autophagy pathway, our data suggests that proin ammatory signal engages autophagy at one of the earliest steps of entire process to allow more e cient removal of the inhibition during HSCs in ammation.

Conclusion
In conclusion, the present study revealed increased serum LPS level and overexpression of p38 MAPK in patients with ACLF. We also found LPS activates p38 MAPK, and then inhibits autophagy by disrupting    8h); (E-G) LX-2 cells in the logarithmic growth phase were treated with or without 20 μM SB203580 for 30min, then cells were treated with 1µg/ml LPS for 15min for the detection of p38 MAPK or in a time-dependent manner (0-8h) for the detection of in ammation. Then, in ammation was evaluated by immunoblotting with speci c antibodies, as indicated. The NLRP3 and pro-IL1 levels were quanti ed by densitometry. Secretion levels of IL-1β were detected by ELISA. Figure 4 LPS inhibits autophagy via p38 MAPK in HSCs. (A, B) LX-2 cells in the logarithmic growth phase were treated with different concentrations of LPS (0-10 μg/ml) for 4h; (C, D) LX-2 cells in the logarithmic treated with or without 20μM SB203580 for 30min before the usage of 1µg/ml LPS for 1h, the expression of Atg13 was then detected. (E) Myc-tagged Atg13 and Flag-tagged p38 were expressed in HEK-293 cells with or without anisomycin. 36h following transfection, cell lysate was immunoprecipitated with anti-Myc and anti-Flag antibodies. Then, expression of p38 MAPK, Atg13 and ULK1 were evaluated by immunoblotting with speci c antibodies, as indicated.