Selonsertib, a Potential Drug for Liver Failure Therapy by Rescuing the Mitochondrial Dysfunction of Macrophage via ASK1-JNK-DRP1 Pathway


 BackgroundAcute liver failure (ALF) is associated with high mortality rate and there are still no effective treatments expect liver transplantation and artificial liver therapies. This study was aimed to determine the effects, therapeutic window and mechanisms of selonsertib, a selective inhibitor of ASK1, for ALF therapy.ResultsLipopolysaccharide and D-galactosamine (LPS/GalN) was used to simulate ALF. We found that selonsertib pretreatment significantly ameliorated ALF as determined by reduced hepatic necrosis and serum alanine aminotransferase, aspartate aminotransferase and inflammatory cytokine levels. However, selonsertib is only effective early after LPS/GalN administration and the limited therapeutic window was related to the activation and mitochondrial translocation of JNK and DRP1. Further experiments revealed that selonsertib could alleviate LPS-induced mitochondrial damage in macrophages by evaluating the mitochondrial membrane potential and mitochondrial permeability transition pore opening in macrophages. Selonsertib also suppressed the release of inflammatory cytokines from macrophages by reducing DRP1-mediated mitochondrial dysfunction, which was confirmed by using mdivi, the specific DRP1 inhibitor.ConclusionsSelonsertib protected against LPS/GalN-induced ALF by attenuating JNK-mediated DRP1 mitochondrial translocation and then rescuing mitochondrial damage in macrophages and may have therapeutic potential for early ALF patients.

Cytokine detection by ELISA The cytokine levels in supernatants from cell cultures were quanti ed with an enzyme-linked immunosorbent assay (ELISA) with a commercial mouse Tumor Necrosis Factor-α (TNF-α) and Interleukin-1α (IL-1α, MultiSciences, China) ELISA kit according to the manufacturer's instruction.
Cytometric Bead Array (CBA) analysis Cytokine levels in the serum samples were detected by BD LSR Fortessa cytometer (Becton Dickinson Holdings Pte Ltd, Singapore) using commercially available kit BD™ Cytometric Bead Array (CBA) Mouse In ammation Kit (BD Biosciences, USA) following manufacturer's instructions. In brief, cytokine standards were prepared using a vial of lyophilized mouse in ammatory cytokines, including IL-6, IL-10, Monocyte Chemoattractant Protein-1 (MCP-1), Interferon-γ (IFN-γ), TNF-α, and IL-12p70, and assay diluent by the method of serial dilutions. Capture bead was added into each tube that is samples, standards, and negative control and was incubated for 30 min at room temperature in the absence of light. The ow cytometer was calibrated using cytometer setup beads and the assay was performed.
Mitochondrial Permeability Transition Pore (mPTP) opening detection and mitochondrial membrane potential observation The mitochondrial membrane potential (Δψm) was observed by staining with the JC-1 probe (Beyotime Biotechnology, Jiangsu, China) according to the manufacturer's instruction. After JC-1 incubation for 20 min at 37 °C, monomeric green uorescence emissions and aggregate red uorescence intensities in cells were observed by two-photon confocal laser scanning microscopy (Lei TCS SPB-MaiTai MP). The mPTP opening was evaluated through the mPTP Assay Kit (Beyotime Biotechnology) according to the manufacturer's instruction and observed with the above microscopy. A decrease in green uorescence indicated that the increased opening of mPTP.

Determination of Mitochondrial oxidative stress
Cells were pelleted by centrifugation and resuspended and incubated in pre-warmed growth medium containing 250 nM MitoSOX™ Red Mitochondrial Superoxide Indicator (Thermo Fisher Scienti c, Inc.) for 30 min at 37 °C in a 5% CO 2 atmosphere. After incubation, the cells were washed with PBS and the uorescence was then assessed by uorescence microscopy.

Mitochondria isolation
The mitochondria from macrophages and liver tissues were isolated by using Cell Mitochondria Isolation Kit (Beyotime Biotechnology) and Tissue Mitochondria Isolation Kit (Beyotime Biotechnology), respectively, according to the manufacturer's instruction. The protein expression levels in the isolated mitochondria were then examined by Western blot analysis.

Western blot analysis
To determine protein expression levels, whole-cell, cell mitochondrial cytoplasmic fractions, tissue extracts, and tissue mitochondrial fractions were lysed with RIPA peptide lysis buffer (Beyotime Biotechnology) containing 1% protease inhibitors (Pierce). Samples were quanti ed using a Pierce BCA Protein Assay Kit according to the manufacturer's instructions, and 30 μg protein was resolved by SDS-PAGE and transferred to PVDF membranes. Membranes were blocked in 5% nonfat milk and incubated overnight at 4°C with primary antibodies, including anti-ASK1 (1:1000; Cell Signaling Technology), anti-p- with HRP-tagged secondary antibodies at room temperature for 1 hour. The protein bands were visualized using Enhanced Chemiluminescence System and detected in ChemiScope Western Blot Imaging System (Clinx Science Instruments Co., Ltd.). Image J software was used for gray-scale value assay.

Statistics
All data are presented as mean ± SD. Comparisons were made between two groups using independent samples t-test, or between three or more groups using one-way analysis of variance (ANOVA) with Tukey's post-hoc test. P < 0.05 was considered signi cant.

Selonsertib protects against LPS/GalN-induced ALF by inhibiting the activation of JNK pathway
To investigate the effect of selonsertib on LPS/GalN-induced ALF, mice were treated with 15, 30, or 60 mg/kg of selonsertib, 30 min prior to administration of LPS/GalN and then examined 6 h later. As shown in Figure 1A and B, LPS/GalN induced signi cant liver injury as indicated by dramatically elevated serum ALT, AST, and TBiL levels accompanied with massive hepatic necrosis, while pretreatment with all of these three dosages of selonsertib resulted in remarkable reduction in serum ALT, AST, and TBiL levels ( Figure 1A) and less hepatic necrosis ( Figure 1B). CBA cytokine analysis also showed that selonsertib pretreatment could signi cantly reduce the elevated serum levels of in ammatory cytokines and chemokines, such as TNF-α, IFN-γ, IL-12p70, IL-6, and MCP-1, whereas enhance the serum level of IL-10, an anti-in ammatory cytokine, in LPS/GalN-administered mice, especially at the dosage of 30 and 60 mg/kg ( Figure 1C). These data indicate that selonsertib can effectively prevent LPS/GalN-induced ALF.
As selonsertib is a speci c ASK1 inhibitor, we then determined the protective mechanism of selonsertib in this model by focusing ASK1-mediated signaling pathway. JNK activation has been implicated in variety of liver injuries and ASK1 is suggested to function upstream of JNK pathway [3,12]. As we expected, LPS/GalN administration induced remarkable elevation of phosphorylated JNK (p-JNK) levels in liver tissues, while selonsertib pretreatment markedly reduced the phosphorylation of JNK ( Figure 1D). Quantitative densitometric analysis also con rmed the changes of p-JNK/total JNK levels by pretreatment with all of these three dosages of selonsertib. These results con rmed that inhibition of ASK1-mediated JNK activation by selonsertib may contribute to the prevention of LPS/GalN-induced ALF.
Selonsertib e cacy is related to JNK mitochondrial translocation Since sustained JNK activation and translocation to mitochondria has been implicated in liver injury [12,13], we then assessed the effect of selonsertib on the activation and mitochondrial translocation of JNK. As shown in Figure 2A and B, the LPS/GalN-induced elevation in p-JNK ( Figure 2A) was accompanied by increases in total JNK and p-JNK in mitochondria ( Figure 2B) and this was prevented by administration of selonsertib.
We then analyzed the therapeutic window of selonsertib in this model. Mice treated with selonsertib 0.5 h or 1 h after LPS/GalN also showed signi cant protection against liver injury when compared to vehicletreated mice but when selonsertib administration was delayed to 2 h or 4 h post LPS/GalN, no protection against liver injury were evident as determined by serum ALT, AST and TBiL levels ( Figure 2C), HE staining ( Figure 2D), and CBA cytokine analysis ( Figure 2E). We further investigated whether the limited therapeutic window of selonsertib on ALF was related to the aberrant JNK activation. Limited LPS/GalNinduced JNK activation was observed as early as 0.5 h and 1 h with almost no JNK and p-JNK translocation to the mitochondria. However, by 2 h, massive JNK activation and extensive mitochondrial translocation had occurred ( Figure 2F and G). The above data suggested that the protective effect of selonsertib on LPS/GalN-induced ALF was depended on suppression of early JNK activation and mitochondrial translocation.

Selonsertib suppresses DRP1 mitochondrial translocation and recti es abnormal mitochondrial dynamics
As DRP1 translocation to mitochondria has been suggested to be downstream of JNK activation and mediate mitochondrial ssion [14,15], we then investigated the effect of selonsertib on JNK-mediated DRP1 activation and mitochondria translocation in the whole and mitochondrial fraction of liver tissues from the mice following 0.5, 1, 2, 4 h LPS/GalN exposure. As shown in Figure 3A, LPS/GalN administration induced gradually elevation of DRP1 phosphorylation at Ser616 (p-DRP1) in the whole tissue samples and there was a dramatic increase of p-DRP1 was detected at 2 h after LPS/GalN injection. Further analyses of the levels of DRP1 and p-DRP1 proteins in mitochondrial fraction showed that LPS/GalN injection promoted the translocation of both DRP1 and p-DRP1 to mitochondria, especially from 2 h after LPS/GalN injection (Fig. 3B). Interestingly, selonsertib pretreatment could effectively suppress the elevation of DRP1 phosphorylation and its translocation to mitochondria by LPS/GalN induction ( Figure 3C and D).
As altered DRP1 activity was known to result in abnormal mitochondrial dynamics [16], we then investigate the effect of selonsertib on mitochondrial dynamics by determining the expression of mitochondrial ssion-and fusion-related proteins. Compared with those in the normal control (NC) group, the expression of ssion-related proteins FIS1, and MFF were signi cantly upregulated whereas the fusion-related proteins MFN1, MFN2, and OPA1 were concurrently signi cantly decreased in the liver tissue from LPS/GalN injected mice ( Figure 3E). While, the dysregulated expressions of mitochondrial dynamics-related proteins were markedly rescued by selonsertib pretreatment. These data suggest that selonsertib may prevent ALF through modulating mitochondrial dynamics by suppressing DRP1 mitochondrial translocation.

Selonsertib prevents mitochondrial damage of macrophages in ALF through suppressing DRP1 activation
Macrophages play a pivotal role in modulating hepatic immune microenvironment and in ammatory response in ALF [17]. As previous studies showed that LPS exposure usually induced mitochondrial dysfunction in macrophages [18,19], we further determine the role of selonsertib in protecting mitochondrial function in the LPS-primed mouse macrophage cell line, RAW264.7. As we expected, LPS exposure induced a marked increase in the mitochondrial oxidative stress in macrophages as indicating by MitoSox, while selonsertib treatment signi cantly weakened the red uorescence intensity in the LPSprimed macrophages ( Figure 4A-left). In addition, LPS priming caused progressive opening of the mPTP, as inferred from the apparent decrease of the uorescent probe, calcein, which had been preloaded into mitochondria ( Figure 4A-right). LPS priming also decreased mitochondrial membrane potential (Δψm), as indicated by the increased conversion of red uorescence to green uorescence of JC-1 dye ( Figure 4Amiddle). While, selonsertib pretreatment effectively prevented the LPS-induced the mPTP opening and Δψm dissipation of macrophage as shown by mitochondrial accumulation of calcein uorescence and reduced the ratio of red to green uorescence of JC-1 dye in macrophages, respectively ( Figure 4A). Moreover, LPS exposure also induced the activation and mitochondrial translocation of JNK and DRP1 in macrophages just as we observed in the liver tissue of LPS/GalN-induced ALF, while, selonsertib pretreatment effectively suppressed those activation and mitochondrial translocation by LPS priming (Figure 4B and C). Meanwhile, LPS stimulation induced signi cant release of in ammatory cytokines, such as TNF-α and IL-1α, whereas selonsertib treatment alleviated the release of these cytokines ( Figure  4D).
By using mdivi, a DRP1 inhibitor, we further found that inhibition of DRP1 activation could effectively inhibit its mitochondrial translocation and then alleviated the LPS-primed abnormal mitochondrial dynamics and mitochondrial dysfunction, which were indicated by the alleviation of the abnormal expressions of mitochondrial dynamics-related proteins and the reduction of mPT pore opening and Δψm dissipation in the LPS-primed macrophages ( Figure 5A and B). We further con rmed the protective effects of mdivi on LPS/GalN-induced ALF by the observation of reduced serum ALT, AST, and TBiL levels and less hepatic necrosis in those mice by mdivi pretreatment (Figure 5C and D). Thus, JNK-mediated DRP1 mitochondrial translocation might be an important link that mediated the protective effect of selonsertib on ALF and mitochondrial damage.

Discussion
ALF, also known as fulminant hepatic failure, is a life-threatening clinical syndrome induced by massive hepatic necrosis. Its medical management is associated with high rates of morbidity and mortality and there are still no effective treatments expect liver transplantation and arti cial liver therapies [1]. Thus, it is important to identify a potential drug and further clarify its treatment mechanism and determine its therapeutic window for ALF. In this study, we showed that selonsertib, a selective inhibitor of ASK1 and currently in a phase III clinical trial on NASH [20], could effectively protect against LPS/GalN-induced ALF by inhibiting the activation of hepatic ASK1-JNK-DRP1 pathway and alleviating mitochondrial damage of macrophages. Unfortunately, selonsertib is only effective early after LPS/GalN administration and the limited therapeutic window may be related to the time point of JNK and DRP1 mitochondrial translocation.
ASK1, a mammalian MAP3K, is activated in response to various cytotoxic stresses, including serum withdrawal, ROS, tumor necrosis factor, microtubule interfering agents, and cancer chemotherapeutic agents and it has been implicated in a variety of cellular functions, including cell survival, and in ammatory response [5,21,22]. Activated ASK1 then activates downstream kinases such as JNK and p38 pathways, resulting cell apoptosis and in ammatory cytokine expression [3,23]. Recently, several reports showed that ASK1 is over-activated in variety of liver diseases including drug-and toxin-induced ALF [6,24,25]. In this study, we found that ASK1 was signi cantly activated in the liver tissues of LPS/GalN-induced ALF (Supplementary Figure S1A). Thus, as a speci c inhibitor of ASK1, selonsertib might be a potential drug for ALF therapy. As we expected, pretreatment with selonsertib could remarkably lessen hepatic necrosis and reduce serum levels of ALT, AST, TBiL, and in ammatory cytokines and chemokines in a dose-dependent manner. In addition, selonsertib pretreatment also suppressed the activation of JNK ( Figure 1) and p38 (Supplementary Figure S1A) in the live tissue from the ALF mouse model. Other studies also supported the above ndings as ASK1 inhibition protected against acetaminophen-induced liver injury and high-fat diet-induced hepatic steatosis by attenuating JNK and p38 activation [6,26]. Unfortunately, further experiments showed that selonsertib is only effective early after LPS/GalN administration (within 1h in the model) and loses its e cacy relatively rapidly after that time. The limited therapeutic window of selonsertib on ALF may be due to the time point of JNK activation and mitochondrial translocation as limited LPS/GalN-induced JNK activation and mitochondria translocation was observed as early as 1 h, whereas a massive JNK activation and an extensive mitochondrial translocation were occurred from 2 h later. While, there is no obvious regularity of the activation of hepatic p38 after LPS/GalN exposure (Supplementary Figure S1B), suggesting that in this model ASK1-midiated JNK activation and mitochondrial translocation might be the key factor determining selonsertib therapeutic e ciency. A previous study also showed that JNK, but not p38, contributed to ASK1-mediated cellular damage in direct APAP-induced hepatocyte injury [24].
The primary pathophysiological event in LPS/GalN-induced ALF is the release of pro-in ammatory cytokines derived from macrophages [11,27]. Emerging evidence now indicates that mitochondrial dysfunction is a possible mechanism of LPS/GalN-induced ALF, which causes abnormal mitochondrial dynamics, ROS over-production, and the release of pro-apoptotic proteins and other factors, resulting in cell death [28]. Mitochondrial quality control and dynamics are essential for regulating mitochondrial homeostasis. Excessive mitochondrial fragmentation with increased ssion or impaired fusion is a hallmark of many degenerative diseases and contributes to their pathophysiology [29,30]. It was also reported that hepatitis B virus induces mitochondrial ssion, which contributes to mitochondrial dysfunction and cell injury [31]. The processes of mitochondrial dynamics are regulated by conserved dynamin-related GTPases, including the ssion proteins DRP1, FIS1, and MFF and the fusion proteins MFN1, MFN2, and OPA1 [32]. Phosphorylation of DRP1 at Ser616 and translocation from cytoplasm into mitochondria seem to be the key upstream event of mitochondrial fragmentation and dysfunction [33]. In the present study, we also found that selonsertib could effectively suppress DRP1 activation and mitochondrial translocation and then rescue the mitochondrial dysfunction and suppress the mitochondrial oxidative stress in the liver tissue of ALF. Since JNK activation is thought to be the critical controlled in the processes of DRP1-mediated ssion, the protective role of selonsertib on ALF may through modulating mitochondrial dynamics by JNK-DRP1 pathway.
Macrophages play a pivotal role in the pathophysiology of in ammatory diseases [34]. Several studies proved that the danger of ALF was dependent on the liver immune homeostasis and in ammation microenvironment mainly modulated by macrophages [17,35]. As LPS exposure was known to induce macrophage mitochondrial dysfunction in acute tissue injury [19], we then used LPS-primed macrophages as in vitro model to determine whether selonsertib could protect the mitochondrial function of macrophages in ALF. We found that selonsertib pretreatment effectively prevented the LPS-induced mitochondrial dysfunction of macrophage and the release of pro-in ammatory cytokines TNF-α and IL-1α, which were known as major cytokines causing liver injury and in ammation [36,37]. By using mdivi-1, a selective inhibitor of DRP1, we further con rmed the role of DRP1 in mediating LPS-induced mitochondrial dysfunction in macrophages and LPS/GalN-induced fulminant liver injury and in ammation. These data indicate that DRP1 may be the key molecule contributing to the protective effect of selonsertib on ALF by affecting the mitochondrial function of macrophage. Although previous studies mainly focused on the role of ASK1 in modulating cytotoxic stress-induced hepatocellular necrosis and apoptosis [14,25,38], our study showed that the therapeutic effect of ASK1 inhibition on ALF may also through protecting macrophage mitochondrial function.

Conclusion
In summary, our data identi ed the speci c ASK1 inhibitor selonsertib as a potential therapeutic drug for the early treatment of ALF by inhibiting JNK-mediated DRP1 mitochondrial translocation and then rescuing mitochondrial damage. In addition, our study further con rmed the important role of mitochondrial function in the pathogenesis of ALF by determining the interaction between ASK1-JNK-DRP1 axis-mediated mitochondrial shape and the abnormal secretion of in ammatory cytokines in macrophages, suggesting restoring mitochondrial homeostasis may also offer a novel strategy for ALF therapy development.

Consent for publication
Not applicable.

Availability of data and materials
All data and materials are available upon request.

Figure 2
The therapeutic e cacy of selonsertib is related to JNK mitochondrial translocation (A, B) Western blot analysis and gray value assay on p-JNK and JNK levels in murine whole liver samples (A) and the mitochondria fraction of liver samples (B), 1 or 2 from 6 samples were shown in the blots. (C) Serum levels of ALT, AST, and TBIL in different groups with selonsertib pretreatment (Pre) or treatment at 0.