HIF‐1α inhibition in macrophages preserves acute liver failure by reducing IL‐1β production

The development of acute liver failure (ALF) is dependent on its local inducer. Inflammation is a high‐frequency and critical factor that accelerates hepatocyte death and liver failure. In response to injury stress, the expression of the transcription factor hypoxia‐inducible factor‐1α (HIF‐1α) in macrophages is promoted by both oxygen‐dependent and oxygen‐independent mechanisms, thus promoting the expression and secretion of the cytokine interleukin‐1β (IL‐1β). IL‐1β further induces hepatocyte apoptosis or necrosis by signaling through the receptor (IL‐1R) on hepatocyte. HIF‐1α knockout in macrophages or IL‐1R knockout in hepatocytes protects against liver failure. However, whether HIF‐1α inhibition in macrophages has a protective role in ALF is unclear. In this study, we revealed that the small molecule HIF‐1α inhibitor PX‐478 inhibits the expression and secretion of IL‐1β, but not tumor necrosis factor α (TNFα), in bone marrow‐derived macrophages (BMDMs). PX‐478 pretreatment alleviates liver injury in LPS/D‐GalN‐induced ALF mice by decreasing the hepatic inflammatory response. In addition, preventive or therapeutic administration of PX‐478 combined with TNFα neutralizing antibody markedly improved LPS/D‐GalN‐induced ALF. Taken together, our data suggest that PX‐478 administration leads to HIF‐1α inhibition and decreased IL‐1β secretion in macrophages, which represents a promising therapeutic strategy for inflammation‐induced ALF.


| INTRODUCTION
Inflammation-mediated acute liver failure (ALF) refers to a process in which inflammatory factors cause hepatic cell damage. [1][2][3] Prior studies have identified that ALF frequently results from viral hepatitis, drug or chemical toxicity, hepatic ischemia-reperfusion injury. etc. 4 Although the pathogenesis of ALF has been widely studied, the mechanism of liver tissue injury and therapeutic strategies remain to be further explored. 3,5 Cytokine storms caused by overactivation of immune cells (especially macrophages) exacerbate the incidence of ALF. 6 Therefore, inhibition of immune cell activation and cytokine production is expected to be a potential therapeutic strategy.
Macrophages, acting as scavengers through phagocytosis, antigen presentation and cytokine production, play a vital role in immune defense and tissue resolution to maintain homeostasis, while their overactivation leads to cytokine storms and lethal organ dysfunction. 7,8 Most mortality caused by ALF is associated with secondary severe immune response dysregulation. 9 During hepatic inflammation, liver resident Kupffer cells (KCs) are first activated and release chemokines and cytokines, which recruit bone marrow-derived macrophages (BMDMs) and other immune cells to the damaged site of the liver, further amplifying the inflammatory signals and accelerating hepatic cell death. [10][11][12] Proinflammatory cytokines such as IL-1β and tumor necrosis factor (TNFα) act not only on immune cells to amplify inflammation, but also on hepatocytes to induce inflammatory and cell death signals. 13,14 It has been reported that blocking the secretion of IL-1β as well as TNFα effectively improves hepatic cell death and alleviates inflammatory-induced acute liver injury. [15][16][17][18] The hypoxia-inducible factor 1α (HIF-1α) transcription factor is a master regulator of the cellular response to hypoxia and coordinates a transcriptional reprogramming that ensures optimal function and metabolic and vascular adaptation to hypoxia. 19,20 Under normoxic and homeostatic conditions, HIF-1α protein is prolyl-hydroxylated by prolyl hydroxylases (PHDs) and subsequently polyubiquitinated by von Hippel-Lindau tumor suppressor protein (VHL), marking it for proteasomal degradation. 21 Notably, HIF-1α accumulates in all proinflammatory activated immune cells, including macrophages, under normoxic conditions and is involved in regulating the transcription of inflammatory genes such as Il-1β. [22][23][24] HIF-1α inhibition markedly reduces Il-1β gene expression, and myeloidspecific HIF-1α deletion protects against inflammationinduced acute and chronic liver injury. 25,26 A small number of studies have indicated a protective role of HIF-1α inhibitors against inflammation-induced acute liver failure, 25,27 but this effect has not yet been well elucidated.
In this study, we report that the HIF-1α inhibitor PX-478 (S-2-amino-3-[4′-N,N,-bis(chloroethyl) amino]phenyl propionic acid N-oxide dihydrochloride), 28 reduces the gene expression and secretion of IL-1β, but not that of TNFα, in BMDMs. We also demonstrated its potential protective effect in ALF and identified that PX-478 has a synergistic effect with a TNFα neutralizing antibody in the treatment of ALF.

| Animals
Male C57BL/6J mice (RRID:IMSR_JAX:000664) approximately 7-8 weeks old were purchased from Beijing HFK Bio-Technology Co., Ltd (Beijing, China). Animal welfare and animal experimental procedures were performed in accordance with the current guide of the Animal Ethics Committee of the Shanghai Institute of Materia Medica, Chinese Academy of Sciences. Animals were maintained at 22 ± 2°C under 12-h light/dark cycles. Water and food were available ad libitum. Age-matched animals were used for all the experiments.

| LPS-induced acute liver failure
Male mice at the age of 8-10 weeks were randomly divided into two groups according to body weight (the mice with body weight < 24 g were excluded). In the preventive administration study, mice were intraperitoneally (i.p.) injected with or without the HIF-1α inhibitor PX-478 at the indicated dose and frequency and then subjected to i.p. injection of 10 μg/kg body weight LPS with 400 mg/kg D-GalN. 15 min after the last PX-478 dosing. Six hours after injection of LPS and D-GalN, blood was collected from eye sockets for plasma biochemistry, and then mice were killed via anesthesia (CO 2 ). TNFα neutralizing antibody (0.2 mg/kg) or PBS solution was i.p. injection 1 h before LPS/D-GalN injection. In the therapeutic administration study, mice were i.p. injection of 10 μg/kg body weight LPS with 400 mg/ kg D-GalN. PX-478 (100 mg/kg) and TNFα neutralizing antibody (0.2 mg/kg) were i.p. injected 1 h and 3 h after LPS/D-GalN injection, respectively. Liver tissue samples were processed for western blotting, RT-qPCR, histological analysis, and immunofluorescence analysis according to a well-established protocol.

| Cell culture
Bone marrow-derived macrophages (BMDMs) isolation was performed and modified as reported previously. 29 Briefly, 6-to 8-week-old male C57BL/6J mice were killed and disinfected with 75% alcohol. The femur and tibia were removed and placed in 10 mL DMEM cell culture medium containing 10% FBS, and the following steps were completed in an ultraclean table. Both ends of the thigh bone were cut, the bone marrow tissue was blown out with culture medium into a single cell suspension, transferred it to a 15 mL centrifuge tube, and centrifuged at 500g for 5 min. The supernatant was discarded, and 5 mL of red blood cell lysis buffer was added to resuspend the cell pellets. After incubating for 5 min, the same volume of DMEM containing 10% FBS was added, mixed well, and centrifuged at 500g for 5 min. The supernatant was discarded, and the cells were resuspended in fresh culture medium, followed by filtering through a 70 μm filter, counting, and plating at a density of 5 × 10 5 cells/mL. Cells were incubated with 20 ng/mL M-CSF and differentiated for 7 days to generate BMDMs.

| Primary hepatocyte isolation and cell death evaluation
Primary hepatocytes were isolated from 8 to 12-week-old male C57BL/6J mice as pervious described. 30 Cell culture medium from BMDMs pretreated with or without PX-478 for 2 h followed by LPS (10 ng/mL) stimulation for 4 h and ATP (2 mM) for the next 45 min was used to incubate hepatocytes, together with additional D-GalN (5 mg/mL).
After 16 h of cell culture, hepatocytes were incubated with a GreenNuc™ Caspase-3 assay kit for live cells (C1168M) for 15 min. 31 The brightfield of cell morphology and caspase-3-positive cells stained with GreenNuc™ Caspase-3 substrate were photographed with an Olympus IX71 microscope. Three random fields of caspase-3positive areas were selected, and the fluorescence intensity (caspase-3-positive region) in each field was determined using Image-Pro Plus (RRID:SCR_007369). Western blot samples were collected after 24 h of treatment with BMDMs supernatant.

| Histologic analysis
Liver tissue was fixed in 4% paraformaldehyde (PFA) for at least 24 h at 4°C and then desiccated and embedded in paraffin before being cut into 5-μm sections and stained with hematoxylin and eosin (H&E) or analyzed by immunofluorescence (IF). For IF, liver slides were incubated in PBS containing 0.3% Triton X-100 and 5% BSA for 1 h. After incubation with anti-F4/80 (Servicebio Cat# GB11027, RRID:AB_2814687), and anti-CD68 (Servicebio Cat# GB113150, RRID:AB_2924885) at 4°C overnight, liver slides were incubated with fluorescence-labeled secondary antibodies (1:1000) for 1 h at room temperature. Blind microscopy analysis was performed with an Olympus IX71 microscope. Three random fields were selected and photographed, and the fluorescence intensity (F4/80 or CD68 positive area) in each field was counted and analyzed by Image-Pro Plus (IPP) (IPP, RRID:SCR_007369). TUNEL staining was performed on liver sections with a DAB (SA-HRP) TUNEL cell apoptosis detection kit according to the manufacturer's instructions. Three random fields of every slide were photographed. The number of positive cells in each field was counted and analyzed by IPP.

| Relative quantitative PCR
Total mRNA was extracted from cells or mouse liver tissues by Trizol (TaKaRa, 9019, Japan). Then, mRNA (500-1000 ng) was reverse transcribed to cDNA. cDNA was mixed with SYBR Green (2X, ABclonal, RK21205) and gene primers were used to detect gene expression levels by Roche LightCycler 480II real-time fluorescence quantitative PCR (Roche, Basel, Switzerland). β-actin was used as the reference gene. The relative quantity of mRNA was normalized to the control group using the ΔΔCT method, and the primer sequences are shown in the Supplementary material (Table S1).
For the concentration measurement of IL-1β and TNFα in the cell culture medium, BMDMs were treated for the indicated time in 24-well or 48-well plates. To induce IL-1β secretion in BMDMs, the cells were pretreated with compounds for 2 h, and then primed with LPS (10 ng/ mL) for 4 h followed by stimulation with ATP (2 mM) for 45 min. The contents of IL-1β and TNFα in culture medium were detected with commercial HTRF kits (IL-1β, 62MIL1BPEG; TNFα, 62MTNFAPEG. PerkinElmer, USA).
For the concentration measurement of ALT in the cell culture medium, hepatocytes were treated with BMDM culture medium for 16 h. The supernatant was collected, and centrifuged at 12 000g for 10 min to remove cell fragments. The content/activity of ALT in the supernatant was detected with a commercial kit (290703, Sysmex, Shanghai, China).

| Statistical analysis
The results are represented as the mean ± S.E.M. Differences between two groups were determined by a two-tailed unpaired Student's t test. Differences in multiple groups were compared by one-way ANOVA or two-way ANOVA. p < .05 was considered statistically significant.

| LPS induces the expression of Hif-1α and its target gene Il-1β in macrophages
LPS treatment induces massive production of inflammatory cytokines, including IL-1β, in macrophages. To address the important role of HIF-1α in the regulation of IL-1β gene expression and secretion, we first investigated the relationship between HIF-1α transcriptional activity and IL-1β production in LPS-treated BMDMs. Compared with the control group, the gene expression of Hif-1α and its target gene Il-1β was markedly elevated by LPS treatment in a dose-dependent manner ( Figure 1A,B). The Hif-1α gene expression level was strongly positively correlated with that of the Il-1β gene ( Figure 1C, with Pearson's correlation coefficient r 2 = 0.7292, p = .0004). Consistently, we detected a significant increase in HIF-1α and IL-1β protein levels ( Figure 1D), as well as remarkable IL-1β secretion after LPS treatment ( Figure 1E). Next, we demonstrated that LPS-induced HIF-1α and IL-1β gene expression and IL-1β secretion in a time-dependent manner ( Figure 1F-I). Taken together, these data show a clear and strongly positive correlation between HIF-1α and IL-1β expression in LPS-treated macrophages.

| PX-478 reduces LPS-induced expression of IL-1β in macrophages
PX-478, a small molecular HIF-1α inhibitor, reduces HIF-1α transcriptional activity by preventing deubiquitination and increasing proteasome-dependentdegradation of HIF-1α. 28 To investigate the potential effect of HIF-1α inhibition on the expression and secretion of IL-1β, we utilized PX-478 and evaluated its inhibitory effect in macrophages. As predicted, PX-478 treatment dose-dependently reduced HIF-1α responsive element (HRE)-driven transcription reporter activities (Figure 2A). PX-478 treatment did not affect the gene expression level of Hif-1α, but dampened LPS-induced HIF-1α protein accumulation in BMDMs ( Figure 2B,C). Moreover, PX-478 treatment reduced Il-1β gene expression and IL-1β secretion in LPS-induced BMDMs ( Figure 2C-E). LPS treatment also increased the gene expression and secretion of TNFα, another vital inflammatory cytokine but not a target gene of HIF-1α, in monocytes/macrophages during acute inflammation. However, after PX-478 treatment, there was no change in the gene expression and secretion of TNFα, suggesting that PX-478 specifically acts on HIF-1α signaling ( Figure 2F,H). PX-478 treatment had little effect on the cell viability of the 293T cell line and in BMDMs at the indicated concentration ( Figure S1), eliminating the inhibitory effect of PX-478 due to its cytotoxicity. Taken together, these results suggest that inhibition of HIF-1α by PX-478 can reduce the expression and secretion of IL-1β, but not TNFα, in macrophages.

| PX-478 protects against inflammation-induced acute liver failure
Excessive infiltration of macrophages into the liver induced by inflammatory stimulation contributes to the severe pathological development of acute liver failure. 12 A critical involvement of myeloid HIF-1a pathway activation has been determined in a type of inflammatory stimuliinduced liver injury model by lipopolysaccharides (LPS) in D-galactosamine (D-GalN)-sensitized mice. Our results indicated that LPS combined with D-GalN treatment accelerates IL-1β and TNFα secretion from immune cells in ALF ( Figure S2). To evaluate the protective effect of HIF-1α inhibition in vivo, we investigated the effect of PX-478 in an LPS/D-GalN-induced ALF mouse model, as shown in the diagram in Figure 3A. Six hours after LPS/D-GalN injection, severe liver injury occurred and the plasma ALT, AST, and LDH levels were markedly increased in the vehicle (Veh) group, which was improved by PX-478 pretreatment in a dose-dependent manner ( Figure 3B-D). Meanwhile, massive hemorrhage was observed visually in the liver of Veh mice, which was further demonstrated by H&E staining analysis, resulting in increased liver weight and liver index. Hemorrhage and liver weight gain were ameliorated by PX-478 pretreatment, especially at a dose of 100 mg/kg ( Figure 3E-H). H&E staining results indicated that LPS/D-GalN-induced excessive hepatocyte death was also dramatically alleviated by PX-478 pretreatment ( Figure 3H). Furthermore, reduced hepatic cell death levels, as indicated by TUNEL staining of liver sections, were observed in PX-478-treated mice ( Figure 3H,I). Consistently, the cleavage rates of the apoptotic markers poly ADP-ribose polymerase (PARP) and Caspase-3, were significantly reduced after PX-478 treatment ( Figure 3J,K). These findings suggest that PX-478 administration can significantly improve LPS/D-GalN-induced liver dysfunction.

| Pretreatment of PX-478 reduces macrophage infiltration into the injured liver
Excessive interleukin (IL)-1 cytokines produced by macrophages have been proposed as important mediators of LPS-induced liver failure. 32 As indicated above, PX-478   -3), and clevaed-caspase-3 (Cle-Cas-3) in liver tissue (n = 7-9, each lane represents one animal sample). α-Tubulin was used as a loading control. Data are presented as the mean ± S.E.M; *p < .05, **p < .01, and ***p < .001 compared to Veh. treatment reduced IL-1β production in LPS-stimulated BMDMs by preventing HIF-1α protein accumulation. Our data suggested that exogenous pathogens promote a burst of IL-1β production in ALF mice, which was restrained by PX-478 pretreatment ( Figure 4A). Although PX-478 treatment had little effect on Tnfα gene expression and TNFα secretion in BMDMs, the plasma concentration of TNFα in PX-478-treated mice was much lower than that in Veh mice ( Figure 4B). Consistent with the reduced cytokine secretion, the expression of the inflammatory-related genes Il-1β and Tnfα in the injured liver was significantly decreased after PX-478 treatment ( Figure 4C).
Chemokines interact with CCR chemokine receptors to mediate the recruitment of immune cells into injured liver tissues and aggravate hepatic inflammation in ALF. The gene expression of the C-C motif chemokine ligand related genes Ccl2, Ccl3, and Ccl4 in the hepatic cells of PX-478-pretreated mice was much lower than that of Veh-treated mice ( Figure 4C). In addition, vascular cell adhesion protein 1 (VCAM-1) and intracellular adhesion molecule-1 (ICAM-1), which are mainly expressed in endothelial cells, have been shown to mediate monocyte rolling and adhesion to endothelial cells under inflammatory conditions. 33,34 As predicted, the gene expression of Vcam-1 and Icam-1 was dramatically elevated in LPS/D-GalN-induced ALF, which was significantly reversed after PX-478 administration ( Figure 4D). To further confirm the effect of PX-478 on macrophage infiltration in injured livers, immunofluorescence staining was used to quantify F4/80-and CD68-positive cells in liver sections. Consistently, there was less macrophage infiltration in the livers of PX-478-treated ALF mice during LPS/D-GalN exposure ( Figure 4E-G). These results indicate that PX-478 treatment alleviates hepatic inflammatory stress induced by LPS/D-GalN injection.

| PX-478 treatment improves hepatocyte cell death induced by an excessive inflammatory response
To elucidate the contribution of reduced IL-1β production in macrophages to the hepatic protective effect of PX-478, we collected the cell culture medium of BMDMs pretreated with or without PX-478 and incubated primary hepatocytes with the medium for the indicated time to induce cell death, as shown in the diagram in Figure S3. Compared with control (unstimulated) BMDM culture medium, the supernatant of LPS/ATP (L/A)-stimulated BMDMs enhanced obvious hepatocyte cell death, while hepatocytes incubated with PX-478-pretreated BMDM culture medium showed less cell damage ( Figure 5A). Consistently, cleaved caspase-3/7 staining showed significantly lower caspase-3/7 activity when hepatocytes were incubated with culture medium from PX-478-pretreated BMDMs than when hepatocytes were incubated with L/A-stimulated BMDM medium ( Figure 5A,B). L/A medium incubation increased ALT secretion from hepatocytes, which was dampened after culture in PX-478-preconditioned BMDM culture medium ( Figure 5C). Meanwhile, the cleavage rate of PARP and caspase-3 protein in hepatocytes was significantly reduced after PX-478-preconditioned BMDM medium treatment ( Figure 5D).
To eliminate the direct protective effect of PX-478 on hepatocytes, we dissolved PX-478 in L/A-stimulated BMDM medium (PX-hepato) and treated hepatocytes for the indicated times. PX-478-preconditioned BMDM culture medium (PX-BMDM), but not PX-hepato medium, reduced ALT secretion in hepatocytes, cell death, and caspase-3/7 activity ( Figure 5E-G). In addition, the cleavage rate of the apoptotic markers PARP and Caspase-3 was significantly reduced after incubation in PX-BMDM medium, but not in PX-hepato medium ( Figure 5H). These results suggest that the liver protective effect of PX478 treatment is due to the reduction in macrophage secretion of inflammatory cytokines rather than that a direct protective role on hepatocytes.

| PX-478 combined with TNFα neutralizing antibody synergistically protects against LPS/D-GalN-induced liver failure
In the pathological process of acute liver failure, TNFα secreted by immune cells induces remarkable hepatocyte cell death. 35 PX-478 treatment reduced IL-1β, but not TNFα, production in BMDMs. Therefore, we investigated the synergistic protective effect of PX-478 and TNFα neutralizing antibody in the LPS/D-GalN-induced ALF mouse model, as shown in the diagram in Figure 6A. Plasma ALT, AST, and LDH levels in the Veh group were markedly increased, and were improved by PX-478 and TNFα neutralizing antibody separately. The protective effect of PX-478 combined with TNFα neutralizing antibody was much better than that of PX-478 or TNFα neutralizing antibody treatment alone ( Figure 6B-F). Massive hepatic hemorrhage was observed in Veh mice by H&E staining, which was ameliorated in the PX-478 and TNFα neutralizing antibody groups, and further improved in the combination treatment group ( Figure 6G). The expression of Il-1β was significantly decreased in the three treated groups ( Figure 6H). However, the expression of Tnfα was decreased in PX-478 and combination-treated mice, but not in mice treated with TNFα neutralizing antibody. The gene expression of Ccl2, Ccl3, and Ccl4 in liver cells after the combination treatment was much lower than that in the Veh liver or monotherapy group ( Figure 6H), suggesting less immune cell infiltration into the liver. Taken together, our results indicate that PX-478 combined with TNFα neutralizing antibody treatment can synergistically alleviate LPS/D-GalN-induced liver failure.

| PX-478 combined with TNFα neutralizing antibody treatment improves LPS/D-GalN-induced liver failure
To mimic the clinical usage of liver-protecting drugs, we then investigated the direct therapeutic effect of PX-478 combined with TNFα neutralizing antibody in ALF mice, as shown in the diagram in Figure 7A. Our results indicated that liver injury biomarkers (plasma ALT, AST, and LDH levels) were markedly improved after PX-478 combined with TNFα neutralizing antibody treatment ( Figure 7B-D). Similarly, massive liver injury was significantly improved by evaluating liver morphology, liver weight, and H&E staining ( Figure 7E-H). LPS/D-GalNinduced excessive hepatocyte death was indicated by the TUNEL-positive signal in liver sections, which was dramatically alleviated by the combination treatment ( Figure 7H,I). Moreover, PX-478 and TNFα neutralizing antibody-treated ALF mice exhibited significantly downregulated mRNA levels of Il-1β, Tnfα, Ccl2, Ccl3, and Ccl4 in the liver ( Figure 7J). Based on the obtained results, we identified that PX-478 combined with TNFα neutralizing antibody treatment relieves inflammation and blocks hepatocyte cell death in LPS/D-GalN-induced acute liver failure.

| DISCUSSION
Acute liver failure (ALF) is a clinical syndrome of massive hepatocyte cell death, that is usually caused by chemical toxicity, drug toxicity, and inflammatory infection. Studies in humans and mice have shown that ALF disease is manifested by massive proinflammatory activation of macrophages, including resident Kupffer cells and recruited bone marrow-derived macrophages. 1,2 However, the resolution mechanism of acute hepatic inflammation is largely unknown, resulting in a chronic shortage of therapeutic strategies. In this study, we identified that HIF-1α inhibition in macrophages improves inflammation-induced ALF. The HIF-1α inhibitor PX-478 significantly reduced HRE-transcriptional activity, Il-1β gene expression, and IL-1β secretion in macrophages, but had no significant effect on Tnfα gene expression or TNFα secretion. In an LPS/D-GalN-induced ALF mouse model, PX-478 pretreatment resulted in resistance to liver injury and inflammation, characterized by decreased proportion of liver macrophages and less hepatocyte cell death. We also revealed that ALF mice treated with PX-478 in combination with TNFα neutralizing antibody were more likely to experience mild hepatocyte cell death.
Liver resident macrophages, Kupffer cells, and bone marrow-derived macrophages are all involved in the hepatic inflammatory response. 2,6 Although our data suggest that pretreatment of PX-478 dampens hepatic macrophage infiltration to some extent, it needs to be further studied whether PX-478 exhibits antiinflammatory effects on cytokine and chemokine production in Kupffer cells first, thereby preventing the infiltration of innate immune cells, or whether PX-478 directly restricts BMDM migration and IL-1β production at the same time.
Many studies have also determined a critical role of HIF-1α in promoting apoptosis in various cell types, by inducting the expression of several proapoptotic genes, including RTP801/REDD1, B-cell leukemia/lymphoma-2 (BCL2), adenovirus E1B 19-kDa-interacting protein 3 (BNIP3), and BNIP3-like (BNIP3L), [36][37][38] as well as stabilizing p53. 39 Previous studies have also shown that HIF-1α is critically involved in LPS/D-GalN-induced ALF using myeloid-specific HIF-1α null mice. 25 Our data showed that PX-478 attenuated hepatic cell death in this type of inflammation-induced liver injury, which inspired us to speculate that PX-478 also inhibits HIF-1α signaling in hepatocytes, demonstrating its powerful hepatoprotective effect. As shown in our results, further treatment could not prevent hepatocyte apoptosis after incubation in LPS/ ATP-treated BMDM culture medium. Therefore, we assume that the liver protective effect of PX-478 is due to its HIF-1α inhibition and anti-inflammatory effect on macrophages.
Interleukin-1 (IL-1), the prototypic pro-inflammatory cytokine, is a master regulator of inflammation by controlling a variety of immune processes. There are two forms of IL-1, IL-1α, and IL-1β. In most studies, their biological activities are indistinguishable. 40 It has been reported that exogenous IL-1α but not IL-1β, was able to augment TNFαmediated liver injury in a mouse model. 32 Interestingly, in this analysis, we identified that PX-478 treatment decreased IL-1α expression in injured liver tissues as well as in proinflammatory macrophages (data not shown). Considering that the expression of Il-1α is unlikely to be regulated by HIF-1α, the mechanisms of the inhibitory effect of PX-478 on Il-1α expression need further investigation.
Interleukin 1β (IL-1β) and TNFα are critical cytokines secreted by macrophages, that remarkably induce hepatocyte death in ALF. Our results demonstrate that PX-478 reduces Il-1β but not Tnfα expression in BMDMs. Anti-TNFα monoclonal antibodies (adalimumab, certolizumab, golimumab, infliximab, etc) has been shown to be an effective intervention for inflammatory bowel disease (IBD). 41 Thus, we tried to use PX-478 in combination with a TNFα neutralizing antibody to interfere with IL-1β and TNFα signaling simultaneously. The effect of combination therapy on ALF was significantly better than that of a single injection of PX-478 or TNFα neutralizing antibody. In addition, we also identified that PX-478 combined with TNFα neutralizing antibody treatment reversed LPS/D-GalN-induced ALF. Previous studies have indicated that monoclonal antibodies against TNFα may be associated with infections or malignancies. 42 Our novel combination strategy may reduce the dose or frequency of anti-TNFα monoclonal antibody use for ALF treatment. Interestingly, our results also indicated that the content of plasma TNFα in ALF mice was markedly reduced after PX-478 treatment. Previous studies have indicated that TNFα is a paracrine and endocrine cytokine. 43,44 Excess TNFα causes hepatocyte apoptosis and triggers cytokine storms. 45 Thus, PX-478 pretreatment probably turns off cytokine storms and hepatocyte cell death by reducing IL-1β secretion in ALF mice, which may subsequently result in decreased TNFα secretion.
Taken together, our results show that the timing and duration of intervention are key to successful inhibition of the HIF-1α/IL-1β pathway in macrophages in the context of inflammation. The HIF-1α inhibitor PX-478 plays a hepatoprotective role in LPS/D-GalN induced ALF and represents a novel potential treatment for inflammationinduced acute liver failure.

DISCLOSURES
The authors have declared that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available in the methods and/or supplementary material of this article and from the corresponding author upon reasonable request.

SUPPORTING INFORMATION
Additional supporting information can be found online in the Supporting Information section at the end of this article.