Ursolic Acid Inhibits the Activation of Kupffer Cells by Caspase-11/NLRP3 Inammasome Signaling Pathways

Background: Previous studies have indicated that Kupffer cells (KCs) are the main regulatory cells for the activation of hepatic stellate cells (HSCs), and caspase-11/NLRP3 inammasome signaling plays crucial roles in the activation of monocyte-macrophages. Ursolic acid (UA) is a traditional Chinese medicine with antibrotic effects, but the molecular mechanism underlying these effects is still unclear. Methods: A mouse primary Kupffer cell line in vitro and liver brosis mice (including specic gene knockout mice) in vivo were selected as experimental objects. RT-qPCR and Western blotting techniques were utilized to assess the mRNA and protein expression in each group. ELISA and histological analysis were utilized to assess liver injury and collagen deposition. Results: In vitro, caspase-11/NLRP3 inammasome signaling promoted the activation of Kupffer cells, and UA inhibited the activation of Kupffer cells by caspase-11/NLRP3 inammasome signaling. In vivo, UA reversed liver damage and brosis in brotic mice and was related to Kupffer cells; the expression of Caspase-11/NLRP3 inammasome signaling in Kupffer cells of the UA group was inhibited. Even in the CCl4 group, the liver damage and brosis of NLRP3 knockout mice were alleviated, and related experiments also proved that the inhibitory effect of UA on Kupffer cells was related to the activation of the NLRP3 inammasome. Conclusion: Caspase-11/NLRP3 inammasome signal transduction is closely related to the activation of Kupffer cells and the occurrence of liver brosis. Additionally, caspase-11/NLRP3 inammasome signaling serves as a new target for UA antibrosis treatment.


Introduction
Liver brosis is an over repair response caused by various chronic liver injuries characterized by excessive deposition of extracellular matrix (ECM) and dominated by type I collagen in the liver [1] . The continuous development of liver brosis can eventually lead to cirrhosis and even liver cancer, which seriously harms human health [2] . Activated hepatic stellate cells (HSCs) are the main source of ECM, and the activation and transformation process of hepatic brosis is the central event. Overall, inhibition of the activation of HSCs is the key factor controlling the progression of liver brosis [1] . Kupffer cells (KCs) are the main regulatory cells in the process of liver brosis and activate hepatic stellate cells in the resting state to promote liver brosis [3] . Cytokines secreted by activated Kupffer cells can directly affect the activation of HSCs, including TGF-β, IL-1β, INF, and CCL3. Proin ammatory factors promote HSC activation through the NF-κB signaling pathway, such as TNF and IL-1β. CCL3 is a ligand of CCR1 and CCR5, which promote liver brosis [4] . Kupffer cells can also produce IL-1 receptor antagonist (IL-1Ra), IL-10 and other antiin ammatory mediators and can produce matrix metalloproteinase (MMP) to promote the degradation of ECM and improve liver brosis [5] . Kupffer cells (KCs) play a key role in regulating HSC activation and liver brosis progression.
The NOD-like receptor protein 3 in ammasome (NLRP3 in ammasome) consists of pattern recognition receptors (PRRs) and apoptosis-associated speck-like protein containing a caspase-recruitment domain (ASC) and caspase-1 and is a classic intracellular innate immune receptor that can be activated by internal and external danger signals to induce the release of IL-1, IL-18 and other proin ammatory cytokines [6] . Studies have shown that the NLRP3 in ammasome activation mechanism is divided into two kinds: through the NLRP3/ASC pathway, activated Caspase 1 is often referred to as the classic NLRP3 in ammasome pathway; however, caspase 11 can also be recruited and activated to activate caspase 1 through NLRP3, resulting in the release of IL-1β and IL-18 and in ammatory cell death, which is a nonclassical NLRP3 in ammasome pathway [7] . It has been reported that caspase11 −/− mice have stronger resistance to lethal sepsis, and their survival rate is signi cantly higher than that of caspase1 −/− mice and wild-type mice [7] . The activation of caspase 11 is more harmful to the body, which indicates that the nonclassical NLRP3 in ammasome pathway plays a more important role in in ammatory injury progression [8] . Ursolic acid (UA) is a natural monomer compound extracted from traditional Chinese medicine plants and has anti-in ammatory, anti-brosis, and liver protection effects [9,10] . However, whether ursolic acid has an inhibitory effect on Kupffer cell activation and whether the inhibitory effect is related to the nonclassical Caspase-11/NLRP3 in ammasome pathway remain to be further studied. This study mainly explores the potential mechanism of ursolic acid against brosis and provides strong experimental support for the future clinical application of ursolic acid in the treatment of patients with liver brosis.

Reagents and Antibodies
The following reagents were used in this study: CCl4 and olive oil (Shandong Xiya Chemical Industries  [11,12] . The wild-type (WT) C57BL/6 mice used in the experiments were from the Department of Laboratory Animal Science of Nanchang University, and NLRP3 knockout C57BL/6 mice were purchased from the Jackson Laboratory (homozygous: B6.129S6-Nlrp3<tm1Bhk>/J). Based on widely recognized research, carbon tetrachloride (CCl4) was selected to induce liver brosis in mice [13] . According to the principle of random allocation, male C57BL/6 mice weighing 20 to 30 g were randomly divided into the control group [n = 10, gavage with olive oil (2 ml/kg) twice a week for 8 weeks], CCl4 group [n = 10, gavage with CCl4 at 2 ml/kg (20% olive oil dilution) twice a week for 8 weeks], and UA group [n = 10, gavage with CCl4 at 2 ml/kg (20% olive oil dilution) twice a week for 4 weeks and then gavage with CCl4 and UA (40 mg/kg/day) gavage for 4 weeks]. Male NLRP3 knockout mice were randomly divided into the NLRP3 -/group, NLRP3 -/-+CCl4 group and NLRP3 -/-+ UA group (all treatments were consistent with the WT group).
If the mice feel painful during modeling or perfusion, it needs to be killed as soon as possible. Mice were euthanized by inhaling iso urane and then cervical dislocation, and death was con rmed by neck tissue separation. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the First A liated Hospital of Nanchang University (Nanchang, China). All animals received humane care in compliance with institutional guidelines.

Histological analysis
The para n-embedded liver and ileum samples were used to prepare 5 µm thick slices with a microtome. The slices were stained with hematoxylin and eosin using standard methods. Sections underwent hematoxylin and eosin (HE) staining, Sirius red collagen staining and immunohistochemistry (IHC) analysis and were evaluated by microscopy. IHC was used to determine the localization and expression of related proteins. Specimens were incubated with an appropriate antibody and were observed and photographed by confocal microscopy. In immuno uorescence cytochemistry, nuclei were counterstained with DAPI. Finally, images were acquired under a uorescence microscope.

Western blot analyses
Total protein was obtained from tissue lysates or cell supernatant for Western blotting. The protein levels were determined using a BCA assay kit (Tiangen, Beijing, China). Denatured proteins were separated on 10% Tris-glycine polyacrylamide gels by SDS-PAGE and transferred to PVDF membranes. The membrane was treated with a chemical illuminator, and the protein bands were detected with a luminescent image analyzer (Bio-Rad ChemiDoc MP, USA). The relative level of the target protein is the gray ratio between the target protein strip and the GAPDH band.

Extraction of primary Kupffer cells
Mice were anesthetized with iso urane (300-500 ml/min). The abdominal cavity was cut open aseptically, and the inferior vena cava was punctured. The PBS solution was uniformly perfused by the syringe at the same time. The hepatic portal vein was injected with 50 ml of perfusion solution (0.05% collagenase IV) at 37°C and digested for 10 min. In detail, cell sediments were resuspended in 10 ml of RPMI 1640 and centrifuged at 300×g for 5 min at 4°C, the top aqueous phase was discarded, and the cell sediments were reserved. Then, cell sediments were resuspended in 10 ml RPMI 1640 and centrifuged at 50×g for 3 min at 4°C. The top aqueous phase (cleared cell suspension) was transferred into a new 10 ml centrifuge tube and centrifuged at 300×g for 5 min at 4°C, the top aqueous phase was discarded, and the cell sediments were reserved. The cell sediments mainly constrained nonparenchymal cells of the liver, which were KCs, sinusoidal endothelial cells and satellite cells. To purify the obtained cell population further, the method of selective adherence to plastic was used according to Blomhoff et al [14] . KCs were identi ed by immuno uorescence using anti-F4/80 antibody.

Statistical analysis
Statistical analyses were performed using SPSS software version 22.0 (SPSS Inc., Chicago, IL), and image production was performed using GraphPad Prism 6.0 software. Quantitative data are expressed as the means ± standard deviation (SD), and continuous variables were compared using one-way analysis of variance (ANOVA). If positive, multiple comparisons were carried out using the Nemenyi test. All statistical tests were two-sided, and P <0.05 was considered statistically signi cant.

Ursolic acid (UA) inhibits the activation of Kupffer cells in vitro
As shown in Figure  Importantly, the results that the NLRP3 in ammasome expression of LPS +H 2 O 2 +Wedelolactone groups were also lower than LPS+H 2 O 2 groups and LPS+H 2 O 2 +MCC groups had shown that the Caspase-11 is an effective stimulator of NLRP3 in ammasome. Additionally, proin ammatory cytokines ( Figure 2I-J) secreted by Kupffer cells were decreased. The results indicated that the expression of the NLRP3 in ammasome or Caspase-11 was signi cantly inhibited by MCC or wedelolactone, respectively and that the Caspase-11/NLRP3 in ammasome pathway plays a crucial role in the activation of Kupffer cells.
To con rm that UA inhibits the activation of Kupffer cells through the Caspase-11/NLRP3 in ammasome pathway, the relative protein expression of Caspase-11 and the NLRP3 in ammasome pathway (including caspase-11 and NLRP3) in the LPS +H 2 O 2 +MCC group was detected. As shown in Figure 2,

UA reverses liver damage and brosis in brotic mice
To evaluate the effect of UA on liver brosis, liver damage and collagen deposition in mouse livers were measured by HE and Sirius red staining ( Figure 3A). The liver lobule structure, collagen deposition and in ammatory cell in ltration of the CCl4 group were signi cantly enhanced (P<0.05), and the performance of liver brosis was signi cantly improved after UA treatment (P<0.05). The ALT, AST, and hydroxyproline levels in mouse serum were determined to evaluate liver function or liver brosis ( Figure  3B-D). Compared to the control group, the serum levels of ALT, AST and hydroxyproline in the CCl4 group mice were signi cantly increased; however, the levels of ALT, AST and hydroxyproline were inhibited in UAtreated brotic mice (P<0.05). These results indicate that UA can reverse liver damage and brosis in vivo.
Type I collagen (collagen-1), a-SMA, and TIMP-1 often serve as biomarkers of HSC activation, and changes in these biomarkers are often found in the progression of liver brosis. At the mRNA levels, the expression levels of type I collagen (collagen-1), a-SMA, and TIMP-1 in the CCl4 group were signi cantly higher than those in the control group, and this increase in type I collagen, a-SMA, and TIMP-1 was UA reverses liver brosis in liver brotic mice by the Caspase-11/NLRP3 in ammasome pathway in Kupffer cells To con rm the roles of the Caspase-11/NLRP3 in ammasome pathway in liver brosis mice and the UA treatment group, immunohistochemical staining of the whole liver tissue of the three groups of mice was conducted. The expression of Caspase-11 and NLRP3 in ammasome of CCl4 groups were signi cantly higher than control group, and the expression of Caspase-11 and NLRP3 in ammasome were signi cantly decreased by UA treatment ( Figure 4A).
To further analyze the activation of mouse Kupffer cells in vivo, mouse Kupffer cells were isolated. The results of Kupffer cells isolated by CD14 immuno uorescence showed that the purity of Kupffer cells was good (Supplementary Figure 1E). Caspase-11 was more highly expressed in Kupffer cells than in other liver cells (such as hepatocytes and HSCs) after CCl4 induction (Supplementary Figure 1A-D). First, the activation of Kupffer cells isolated from the three groups of mice was determined. The results of proin ammatory cytokine (INF-γ, TGF-β) by ELISA secreted by Kupffer cells in CCl4 group were signi cantly higher than control group (P<0.05); and the pro-in ammatory cytokine of UA group were obviously decreased than CCl4 group (P<0.05) ( Figure 4B-C). To con rm the effect of the Caspase-11/NLRP3 in ammasome pathway on the inhibition of Kupffer cells by UA, the expression of Caspase-11 and the NLRP3 in ammasome in isolated Kupffer cells was detected. Consistent with the vitro results, the expression of Caspase-11, NLRP3 in ammasome in CCl4 group were signi cantly higher than control group; the expression of Caspase-11, NLRP3 in ammasome in UA group were signi cantly lower than CCl4 group (P<0.05) ( Figure 4D-I).
Effect of NLRP3 knockout on liver brosis and Kupffer activation To further con rm the role of the Caspase-11/NLRP3 in ammasome pathway in liver brosis, male NLRP3 knockout mice were randomly divided into the NLRP3 -/group, NLRP3 -/-+CCl4 group, and NLRP3 -/-+ UA group. As shown in Figure 5A, the liver lobule structure, collagen deposition and in ammatory cell in ltration of the NLRP3 -/group, NLRP3 -/-+CCl4 group, and NLRP3 -/-+UA group were signi cantly reversed compared with those of the WT+CCl4 group. Importantly, there was no signi cant difference between the three groups (NLRP3 -/group, NLRP3 -/-+CCl4 group, NLRP3 -/-+ UA group). The results showed that the levels of ALT, AST, and hydroxyproline in the mouse serum of the three groups were signi cantly lower than those in the WT+CCl4 groups, and no obvious change was found between the three NLRP3 -/groups ( Figure 5B-D). The mRNA expression of collagen-1, a-SMA, and TIMP-1 in the three NLRP3 -/groups was signi cantly lower than that in the WT+CCl4 groups, and the differences among the three NLRP3-/groups were not statistically signi cant ( Figure 5E-G). Kupffer cells were isolated from mice, and the proin ammatory cytokines (INF-γ and TGF-β) in Kupffer cells were measured. The mRNA expression of INF-γ and TGF-β in the three NLRP3-/-groups was signi cantly lower than that in the WT+CCl4 groups, and the expression levels of INF-γ and TGF-β in the three NLRP3-/-groups were similar ( Figure 5H-I).
Previous results have indicated that the NLRP3 in ammasome plays a key role in liver brosis.
To assess the expression change of the Caspase-11/NLRP3 in ammasome, the mRNA and protein expression of Caspase-11 and the NLRP3 in ammasome were detected. As shown in Figure 5J, 5°, and 5N, the mRNA and protein expression of the NLRP3 in ammasome were signi cantly decreased after NLRP3 knockout in mice. Importantly, the Caspase-11 expression of NLRP3-/-+CCl4 group after CCl4 induced is still signi cantly higher than NLRP3-/-group, and the Caspase-11 expression of NLRP3-/-+CCl4 group after UA treatment were signi cantly lower than NLRP3-/-+CCl4 group. These results suggest that the NLRP3 in ammasome plays an important role in Caspase-11 in liver brosis progression (P<0.05) ( Figure 5J, 5K, 5M).

Discussion
Liver brosis is a process of extracellular matrix (ECM) deposition or scar formation caused by various factors, including hepatitis viral, nonalcoholic fatty liver, alcoholic fatty liver, biliary or autoimmune liver disease [1] . The continuous development of liver brosis can eventually develop into liver cirrhosis, even liver cancer, which seriously endangers human health [2] . Liver brosis is the early stage of liver cirrhosis.
Effective treatment intervention can effectively prevent the progression of the disease [15] . Therefore, it is of great signi cance to develop anti brosis drugs based on the pathogenesis of liver brosis.
The transformation of quiescent HSCs into proliferative myo broblasts is the central event in the pathogenesis of liver brosis; however, Kupffer cells are the main regulatory cells in the process of liver brosis [3] . The activation of resting HSCs by Kupffer cells can promote the progression of liver brosis; the apoptosis or degradation of activated HSCs by Kupffer cells can promote the progression of liver brosis. In the progressive stage of liver injury, hepatocyte injury or harmful substances (such as bacteria or lipopolysaccharide (LPS)) can trigger damage-related molecular models (DAMPs) or pathogen-related molecular models (PAMPs) to activate Toll-like receptors (TLRs) or tumor necrosis factor receptors (TNFRs) to stimulate Kupffer cell activation [16] . Then, activated Kupffer cells secrete proin ammatory factors, including TGF-β, TNF, IL-1β and IFN-γ. At the remission stage of liver injury, Kupffer cells transform into in ammatory inhibitors and produce anti-in ammatory mediators, such as IL-1Ra and IL-10 [17,18] . Therefore, Kupffer cells play a double-edged sword role in liver brosis. Therefore, the activation of Kupffer cells in vivo and in vitro served as the main observation objects in this study. First, this study found that the activation of primary liver Kupffer cells was signi cantly enhanced after LPS combined with H 2 O 2 stimulation, and the activation could be inhibited by ursolic acid in vitro. The in vitro results indicated that the Caspase-11/NLRP3 nonclassical in ammasome pathway is involved in the activation of Kupffer cells. Next, we demonstrated that UA could reduce CCl4-induced liver brosis and inhibit Kupffer cell activation after Kupffer cells were isolated from mouse livers. The Caspase-11/NLRP3 in ammasome pathway plays an important role in the activation of Kupffer cells in vivo. Finally, the results of the NLRP3 -/mouse experiment showed that the Caspase-11/NLRP3 in ammasome pathway was involved in Kupffer activation.
The NLRP3 in ammasome is an intracellular multiprotein complex that is widely involved in the body's immune response and is related to the pathogenesis of tumors, arteriosclerosis, intestinal in ammation and metabolic diseases [19] . The NLRP3 in ammasome consists of PRRs, ASC, and caspase-1 and is widely distributed in monocytes-macrophages, dendritic cells (DCs), lymphocytes, granulocytes and antigen-presenting cells (APCs) [20] . The NLRP3 in ammasome is a classical receptor of intracellular innate immunity that can be activated by danger signals inside and outside the cell and then induce the release of downstream proin ammatory factors (IL-1β, IL-18) [21] . At present, the NLRP3 in ammasome is expressed in hepatocytes, Kupffer cells, and HSCs in the liver and is activated under certain conditions, eventually leading to the release of IL-1β and IL-18 [22] . However, the expression level of the NLRP3 in ammasome in Kupffer cells was signi cantly higher than that in hepatocytes and HSCs, as shown in Supplementary Figure 1. Therefore, previous studies have indicated that Kupffer cells are the main places for the expression, assembly and activation of NLRP3 in ammasome [23] . There were two pathways involved in the NLRP3 in ammasome activation mechanism: classic NLRP3 in ammasome pathways (NLRP3/ASC/Caspase-1) and nonclassical NLRP3 in ammasome pathways (Caspase-11/NLRP3/Caspase-11) [7] . Hepatocyte death is induced by activated NLRP3 in ammasomes through the pyrolytic pathway and aggravates the proceeding of NASH [24] . Overall, the NLRP3 in ammasome plays a crucial role in the liver in ammation network.
Caspase-11 is an important promoter of the nonclassical pathway of cell pyrolysis. During the progression of liver injury, gram-negative bacteria enter the liver through the portal vein and release lipopolysaccharides (LPS) on the surface to activate Kupffer cells through the TLR pathway [25] . LPS enters Kupffer cells in the form of endocytosis and interacts with intracellular caspase-11 to bind and activate it, thus initiating the nonclassical pathway of cell pyrolysis. LPS enters Kupffer cells through endocytosis and interacts with intracellular caspase-11 and activates it, thereby starting the nonclassical pathway of pyrolysis [26] . On the one hand, activated caspase-11 can activate the downstream NLRP3 in ammasome, releasing IL-1β and IL-18; on the other hand, the cell-breaking membrane protein GSDMD is activated and destroys the cell membrane, releases the cell content, and causes in ammatory damage [27] . The cross-analysis of unbiased RNA sequencing/proteomic analyses identi ed Caspase-11 (Caspase-4 in humans) as a commonly upregulated gene in alcoholic hepatitis (AH) and patients but not in chronic alcoholic steatohepatitis (ASH) mice and healthy human livers [28] . Recent studies have shown that HSPA12A attenuates LPS-induced liver injury by inhibiting caspase-11-mediated hepatocyte pyroptosis via PGC-1α-dependent acyloxyacyl hydrolase expression [28] . Caspase-11 activation is harmful to the body, and its role in the nonclassical NLRP3 in ammasome pathway is critical. In this study, Our results indicated that uric acid can improve liver brosis by inhibiting the Caspase-11/NLRP3 in ammasome pathway of Kupffer cells. In vivo and in vitro, the Caspase-11/NLRP3 in ammasome pathway plays an indispensable role in the activation of Kupffer cells, and uric acid can inhibit the activation of Kupffer cells by the Caspase-11/NLRP3 in ammasome pathway. The results from NLRP3 -/mice showed that the NLRP3 in ammasome had an important impact on the activation progression of Kupffer by Caspase-11. There are also some limitations, such as the lack of results regarding the coculture of Kupffer cells and HSCs and the greater number of Caspase-11 intervention results.
In conclusion, our results indicate that the caspase-11/NLRP3 in ammasome pathway plays an important role in the activation of Kupffer cells. Furthermore, UA may reverse liver brosis by intervening in the Caspase-11/NLRP3 in ammasome pathway in Kupffer cells. The potential mechanism of UA against liver brosis remains unknown. This study aims to clarify the possible molecular targets of UA against liver brosis and provide a reasonable experimental basis for the clinical application of UA in the future. Our results provide new insight into the treatment of liver brosis with UA; however, further in vivo and in vitro studies are needed to con rm these results.

Declarations
Financial support: