Background
Non-alcoholic steatohepatitis (NASH) is a dynamic disease. NASH may regress to steatosis and maintain relatively stable, or it can lead to progressive liver fibrosis, cirrhosis, and other advanced liver diseases [
1].
Pyroptosis is a new route of programmed cell death, presenting the characteristics of pro-inflammatory and cytolytic apoptosis, which is mediated by Gasdermin proteins [
2]. In the process of pyroptosis, Gasdermin proteins form holes in cell membranes, causing cells to swell and rupture, releasing inflammatory factors to defend against foreign pathogens [
3]. This process has been generally reported to depend on the activation of Caspase-1[
4,
5], while other studies suggested that the Caspase-1-independent pyroptosis pathway was ubiquitous in mammalian cells, which depended on other activators such as Caspase-4/5/11 [
6,
7]. Pyroptosis has been proven to link with several infectious diseases [
8], autoimmune disorders [
9], and tumors [
10].
Hepatocyte pyroptosis has been reported in mice animal models of NASH and human with obesity-associated NAFLD [
11,
12]. The hepatocyte pyroptosis was initiated by Caspase-1, which was activated by various inflammasomes. The inflammasomes are intracellular epichaperomes activated in liver cells or other non-platelet cells, which can respond to risk factors. The inflammasomes mainly consist of: [
1] pattern recognition receptor (PRR) as sensor molecule, [
2] apoptosis-associated speck-like protein containing CARD (ASC), and [
3] pro-caspase-1 as effector molecule [
13]. As a Caspase-1-dependent programmed cell death, pyroptosis has been proven to make important effects on NASH [
14]. However, until now, only limited studies have been performed on hepatocyte pyroptosis in NASH, as well as the molecular mechanism. Therefore, identifying the key molecules involved in the biochemical cascade activating Caspase-1 inflammasomes is significant, which may provide new candidate strategies for developing gene-mediated therapy.
Annexin A2(ANXA2) belongs to the annexin family, which is a calcium-dependent phospholipid-binding protein. ANXA2 is mainly located on the cell membrane. The abnormal expression of ANXA2 has been detected in various malignancies. It has been reported that the expression of ANXA2 was associated with tumorigenesis, progression, invasion, and metastasis, which can be applied as a biomarker for predicting the prognosis [
15]. Bioinformatics analysis predicted that ANXA2 was a critical event in steatosis. The deregulation of ANXA2 made effects on fat storage in the liver via deregulating the clearance of plasma cholesterol [
16]. The
Anxa2 was also identified as tumor-specific chromatin-accessible regions in hepatocellular carcinoma tissues from NASH [
17]. The experimental results showed that ANXA2 expression was associated with not only liver histological features, but also insulin resistance in NASH [
18]. The increased expression of ANXA2 in hepatocytes can promote hepatic fibrosis in NASH mice, and the mechanism was to increase the expression of osteopontin. The ANXA2-Notch positive regulatory loop was involved in this process [
19]. However, the effects of ANXA2 in NASH-derived hepatocyte pyroptosis have been still unclarified, nor the underlying mechanism.
In this study, the results of both bioinformatic analyses and experiments revealed that, the expression level of ANXA2 was up-regulated in both LPS-treated hepatocytes and NASH liver, and the inhibition of ANXA2 expression can suppress the associated hepatocyte pyroptosis and fibrosis. According to the analysis of upstream transcription factors of ANXA2, it was observed that the highly expressed p-STAT3 could promote the transcription of Anxa2 by binding with its promoter. The p-STAT3/ANXA2 axis can activate the Caspase-1, thus mediating the hepatocyte pyroptosis and fibrosis of NASH.
Methods
Ethics statement
6-week-old C57BL/6J mice were obtained from Charles River Laboratories for constructing the animal model of NASH (Beijing, China) [
20,
21]. The mice were maintained in a light/dark cycle of 12 h/12 h, at 22 ± 2 ℃ and 55-65% humidity. All the manipulation relevant to animals have been reviewed and proved by the Animal Ethics Committee of Shanghai General Hospital, and are conducted according to NIH guidelines (Approval No. 2022AW018).
All the public data were obtained from the Gene Expression Omnibus (GEO) database (
www.ncbi.nlm.nih.gov/geo). GSE55747 mouse gene chip dataset (including 4 healthy liver tissues and 6 CCL4-induced liver fibrosis tissues) was analyzed for screening differentially expressed genes in the development of liver fibrosis. GSE174478 human gene chip dataset (including 94 patients with non-alcoholic fatty liver disease, NAFLD) was analyzed for screening differentially expressed genes related to hepatocyte pyroptosis. The original data were filtered and standardized by log2(x + 1), and Wilcoxon was used for the difference test. Then, each patient was scored for their pyroptotic activity. The REACTOME_PYROPTOSIS gene set of the Molecular Signatures Database (
www.gsea-msigdb.org/gsea/msigdb) was used as the background gene set. The Gene Set Variation Analysis (GSVA) R package was involved in quantifying [
22]. Based on the median value, the patients were divided into groups with high and low pyroptotic activity. The ChIP-seq data of STAT3 binding in inflammatory and healthy mouse hepatocytes were downloaded from GSE96767. The obtained Bedgraph format files were converted into BW format files, enrichment heatmaps were drawn for the upstream and downstream 3 kb regions of Transcription Start Site (TSS) using Deeptools, and differential peak enrichment was observed in IGV browser.
NASH animal model
The NASH mouse model was established according to the previous report [
23]. After adapting for one week, the mice were maintained in an animal room with controlled conditions: light/dark cycle of 12 h/12 h, 22 ± 2 ℃, and humidity of 55-65%. The mice in the NASH group were fed with high-fat diet and the mice in the control group were fed a normal diet. The formula of a high-fat diet and a normal diet was shown in Supplementary file S1. All experiments on mice were approved by the Scientific Research Ethics Committee of our hospital (Approval No. 2022AW018). The animals were handled according to the agency’s guidelines for the care and use of experimental animals.
Animals were randomly divided and there were eight [
8] mice in each group: Control group (NC group), NASH group, Lv-NC group (NASH mice treated with adenovirus Lv-NC), Lv-sh
Anxa2 group (NASH mice treated with adenovirus Lv-sh
Anxa2), APTSTAT3-9R group (NASH mice treated with STAT3-specific inhibitor, APTSTAT3-9R), APTSTAT3-9R + Lv-
Anxa2 group (NASH mice treated with both APTSTAT3-9R and Lv-
Anxa2), Lv-
Anxa2 group (NASH mice treated with adenovirus Lv-
Anxa2), Lv-
Anxa2 + VX-765 group (NASH mice treated with both adenovirus Lv-
Anxa2 and VX-765). The adenovirus was purchased from Sangon Biotech (Shanghai) Co., Ltd. The mice were injected with a total of 1 × 10
9 PFU recombinant adenoviruses via tail vein injection. VX-765 was a specific inhibitor for Caspase-1. After experiments, all the mice were euthanized. The serum and liver tissues were collected for the following experiments.
HE staining
HE staining was applied for detecting pathological changes in the liver. Fresh liver tissue was fixed with 4% paraformaldehyde, embedded in paraffin, and sliced (the thickness was 4 μm). After dewaxing and hydrating, the slices were stained with hematoxylin (Sigma, Shanghai, China) at room temperature (RT), and then rinsed. After differentiating with 5% acetic acid for 1 min, the slices were rinsed and then stained with eosin (Sigma) for 1 min.
Masson staining
Masson staining was applied for detecting liver fibrosis. The kit was purchased from Solarbio (Beijing, China). After dewaxing and hydrating, the slices were stained with iron hematoxylin for 7 min at 20 ℃, and then rinsed. After differentiating with 1% hydrochloric acid alcohol for 30s, the sections were rinsed and then stained with ponceau acid fuchsin for 5 min. After differentiating the slices with 1% phosphomolybdic acid and 1% acetic acid (each for 1 min), the slices were sealed after dehydration and observed under the microscope.
Immunohistochemical staining
After dewaxing and hydrating, the antigen of slices was repaired by incubating with 10 mM sodium citrate and 0.05% Tween-20 (pH 7.4) for 10 min at 100
oC. The endogenous peroxidase was blocked with a mixture of H
2O
2 (3%): methanol (1:1) for 30 min. After blocking with 5% bovine serum albumin for 1 h at RT, the slices were incubated with α-SMA primary antibody (1: 1000, 14-9760-82, Invitrogen) at 4
oC overnight. The HRP labeled 2nd antibody was added and incubated for 1 h at RT. Images were taken with an upright microscope (Zeiss, Oberkochen, Germany) and analyzed quantitatively by ImageJ according to previous literature [
24,
25].
ELISA
ELISA was applied for quantifying the biomarkers for indicating liver damage and inflammation in mouse serum and hepatocyte culture supernatant, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), and pro-inflammatory factors IL-1β, C-reactive protein (CRP). The ELISA kits were obtained from Abcam (Beijing, China). The immunoassay was performed according to the description of the manufacturer.
Western blot (WB)
Total protein was extracted with RIPA lysis buffer (Abcam), and then quantified by BCA kit (Abcam). 20 µg of protein was loaded on the SDS-PAGE gel. The protein was transferred to the PVDF membrane after electrophoresis. After blocking, the membrane was incubated with antibodies for relevant biomarkers overnight at 4
oC, and then incubated with HRP labeled 2nd antibody (1: 10,000, ab205718, Abcam) for 2 h at RT. The immunoreactive stripes were developed with an ECL substrate kit (Abcam). The intensity of stripes was quantified with ImageJ software. The antibodies involved in WB were shown in Table
1.
Table 1
Primary antibodies used for WB.
ANXA2 | 03-4400 | 1:1000 | Invitrogen |
pro-Caspase1 | 14-9832-82 | 1:1000 | eBioscience |
C-Caspase1 | PA5-99390 | 1:1000 | Invitrogen |
ASC | PA5-50915 | 1:1000 | Invitrogen |
GSDMD | ARG41404 | 1:1000 | Arigobio |
GSDMD-N | ab215203 | 1:1000 | Abcam |
α-SMA | 14-9760-82 | 1:1000 | Invitrogen |
p-STAT3 | 00968 | 1:1000 | Genetex |
β-Actin | ab8226 | 1:1000 | Abcam |
RT-qPCR
Total RNA was extracted from liver tissues or hepatocytes and reverse transcribed into cDNA using ReverTra Ace® qPCR RT kit (Toyobo, Japan). RT-qPCR was conducted with a qPCR kit and CFX Connect™ real-time system (BIO-RAD, USA). The relative mRNA expression level of the tested gene was analyzed by 2
−ΔΔCT, with the internal reference of
Gapdh. The sequences of primer pairs were provided (Table
2).
Table 2
Primer pairs used for RT-qPCR
Anxa2
| CACCAACTTCGATGCTGAGAGG | GCACATTGCTGCGGTTTGTCAG |
Gapdh
| CATCACTGCCACCCAGAAGACTG | ATGCCAGTGAGCTTCCCGTTCAG |
In vitro study on mouse hepatocytes.
Hepatocytes were isolated from 4 mice based on the previous report [
26]. The mouse liver was perfused with EGTA buffer (5.4 mM KCl, 0.44 mM KH
2PO
4, 140 mM NaCl, 0.34 mM Na
2HPO
4, 0.5 mM EGTA, and 25 mM Tricine) to drain the blood. Then, the liver tissue was digested with 0.075% collagenase. All the buffers were perfused at a rate of 5 mL min
− 1 for 10 min. After digestion, the liver tissue was filtered and centrifuged three times at 750 rpm for 5 min. The cells were cultured in Williams E medium without phenol red, supplemented by primary hepatocyte maintainer, and cultured overnight at 37 ℃ in 5% CO
2.
The obtained cell suspension was incubated with 0.01% DNase for 15 min, and filtered (pore size: 100 μm). The cells were centrifuged at 200 × g for 10 min for collecting the cell pellet. Hepatocytes were washed with HBSS and cultured in RPMI 1640 medium (Gibco) supplemented with 10% FBS, 100 units mL− 1 penicillin, and 100 µg mL− 1 streptomycin (Sigma). The hepatocytes were cultured overnight at 37 oC and 5% CO2 in 6-well plates coated with collagen. To simulate the hepatocyte under NASH environment, 1 µg mL− 1 LPS(Sigma-Aldrich) was applied to treat cells for priming for 5 h, and then BzATP (100 µM) was applied for inflammasome activation for another 1 h.
The sh-Anxa2 and over-expression DNA plasmid (oe-Anxa2) used for cell transfection were purchased from Origene. Lipofectamine2000 was used for transfection. Cells were treated with 10 µM APTSTAT3-9R or VX-765 for 2 h to inhibit the activation of STAT3 or Caspase-1, respectively.
Luciferase reporter assay
The promoter sequence of
Anxa2 and the potential binding site of p-STAT3 was obtained from the JASPAR database (
http://jaspar.genereg.net/). The
Anxa2 promoter sequence integrating this site was inserted in the pGL4-Luc-Report vector (Promega, Shanghai, China) to construct the promoter luciferase reporter. The above vectors were transfected into hepatocytes. Then, the cells were treated with DMSO or STAT3-specific inhibitor APTSTAT3-9R. After 48 h, the luciferase activity was determined by a double luciferase detection kit (Promega).
ChIP-qPCR
ChIP was performed with Magna ChIP™ A/G Chromatin Immunoprecipitation Kit (Merck Millipore, Burlington, MA, USA) and anti-p-STAT3 (1:50, #9134, CST) antibody or IgG (1:50, #3900, CST). After ChIP, the qPCR was applied for quantifying immunoprecipitated DNA, with the following primers targeted to the Anxa2 promoter sequence (Forward: CCCAGTTCAGAGGAATCCAA; Reverse: CCAGGCCCTACAAGTATCCA).
Statistics
All statistics were carried out by Graphpad Prism 8.02 software. The data were expressed as mean value ± standard deviation (SD) for three independent experiments. The difference was compared with the Unpaired t-test (two groups), One-way ANOVA, or Two-way ANOVA (multiple groups). Tukey was applied for the post hoc test. P < 0.05 indicated statistically significant.
Discussion
Histopathological changes in NASH include hepatic steatosis and inflammation, with characteristic hepatocyte damage (“ballooning-like” degeneration). NASH has been considered as a potential health threat since it may develop to advanced liver diseases [
29]. NASH has become a large and increasing public health problem in worldwide. The growing prevalence may lead to a great disease burden in the future [
30]. Fortunately, NASH is reversible, which can be timely controlled. Several studies have attempted to investigate the molecular mechanism of NASH and develop potential therapeutic strategies [
31]. Here, we provide in vivo and in vitro evidence that
Anxa2 promotes hepatocyte pyroptosis and liver fibrosis. The mechanism is that p-STAT3/ANXA2 axis promotes the activation of downstream Caspase-1, thus inducing hepatocyte pyroptosis in NASH. Importantly, inhibition of the p-STAT3/ANXA2 axis or activation of Caspase-1 has therapeutic implications in the NASH mouse model.
Pyroptosis is a newly characterized type of programmed cell death accompanied by inflammation. Its functions in the development and progression of various diseases remain to be explored. Several studies have tried to apply bioinformatics analysis to explore pyroptosis-related genes, especially in various cancers. One study has compiled 43 pyroptosis-related genes, depicted their expression changes across 31 cancer types, and constructed a novel prognostic risk model to predict cancer patient survival [
32]. In another study on lung adenocarcinoma patients, the pyroptosis-related genes with differential expressions were screened from the RNA sequencing and clinical information datasets. From a total of 33 pyroptosis-related genes, 10 key genes were selected by a machine learning algorithm, assisted in subtyping the patients and predicting the overall survival [
33]. Another comprehensive bioinformatics analysis also performed on lung adenocarcinoma identified a prognostic pyroptosis-related gene signature containing five genes (NLRP7, NLRP1, NLRP2, NOD1, and CASP6). In addition, a lncRNA KCNQ1OT1/miR-335-5p/NLRP1/NLRP7 regulatory axis was predicted based on the ceRNA network [
34]. In a study on breast cancer, “PyroptosisScore” was calculated based on the expression of pyroptosis-related genes. It was generated to quantitatively evaluate pyroptosis in individual patients, which can help clinicians to understand the characteristics of tumor microenvironment infiltration and assist in developing individual immunotherapy [
9]. Some studies have also screened the pyroptosis-related genes associated with liver diseases, mainly liver cancer, such as its diagnosis, prognosis [
35], immune infiltration [
36], immune activity [
37], and so on. Seven pyroptosis-related genes (BAK1, BAX, CHMP2A, CHMP4C, CHMP6, GSDMC, and GSDMD) were identified as diagnostic markers, while another four (TP53, GPX4, GSDMC, BAK1) suggested prognostic prediction significance [
35]. Besides screening markers, other studies have also tried to find the core gene as well as the potential regulatory axis. CASP8 was identified as the core gene in predicting the prognosis of patients with liver cancer. Moreover, the lncRNA MIR17HG/hsa-miRNA-130b-3p/CASP8 regulatory axis was revealed [
38].
Besides various cancers, studies have shown that hepatocyte pyroptosis plays a role in NASH. In a NASH mouse model induced by feeding cholesterol, Sphingomyelin synthase 1 (SMS1) was reported to mediate hepatocyte pyroptosis, thereby inducing NASH [
39]. SMS1 was an enzyme that links ceramide to sphingomyelin synthesis and diacylglycerol generation. Higher levels of free cholesterol in hepatocytes induced an increased level of
Sms1. Under the action of elevated level of SMS1, more diacylglycerol was generated, thus activating KCδ and initiating the DAG-PKCδ-NLRC4 axis. The downstream NLRC4 inflammasome was then activated, thus inducing hepatocyte pyroptosis. In another study performed on alcoholic hepatitis (AH), a Caspase-11-dependent pyroptosis pathway has been performed. GSDMD was activated in AH livers in mice and patients and then inducted pyroptosis downstream of CASP11/4 activation [
40]. The arsenic-induced pyroptosis in NASH involved autophagy, CTSB, and the NLRP3 inflammasome cascade, and that taurine alleviated As
2O
3-induced liver inflammation by inhibiting the autophagic-CTSB-NLRP3 inflammasomal pathway rather than decreasing lipid accumulation [
41]. In a similar study on NAFLD,
Ecklonia cava extracts or dieckol attenuated NAFLD by decreasing the NLRP3 inflammasome and pyroptosis [
42]. The gap junctions mediate intercellular communication and support liver homeostasis, and the connexin hemichannels were preferentially opened by various pathological stimuli, including inflammation and oxidative stress. The latter are essential features of non-alcoholic steatohepatitis. The involvement of connexin32 and connexin43 hemichannels suggested their role as potential drug targets in NASH [
43]. As a major building block of hepatocellular gap junctions, connexin32 was involved in the sequelae of steatosis, which underlined progression of NASH [
44]. However, the experimental evidence for exploring the mechanism of pyroptosis has been limited.
Other studies have demonstrated the mechanism of NASH. The oxidative stress has been considered a common pathogenetic factor [
45]. The hepatotoxicity induced by bupropion resulted from the generated reactive oxygen species (ROS), which can be mitigated by drugs with reactive radical scavenging properties [
46]. Some components, such as Silymarin, a standardized extract of the
Silybum marianum, is believed to make hepatoprotective effects through inhibition of free radicals and inflammation [
47]. The oxidative stress and inflammation have been commonly investigated in the mechanism of NASH.
Different from most of the above studies, our study has been conducted from the molecular level, the role of ANXA2 in pyroptosis and fibrosis in NASH has been investigated from both in vivo and in vitro evidence. In this study, we first involved the GEO database for identifying ANXA2 as a liver fibrosis-related gene, which has so far been rarely discussed. Alcoholic liver disease, alcoholic fatty liver disease, and NASH have all been implicated in the progression of fibrosis. After confirming the upregulation of ANXA2 in liver tissues of NASH mice, the loss-of-function experiment was performed with adenoviral vectors. Knockdown of Anxa2 significantly inhibited hepatocyte pyroptosis (NLRP3, C-CASPASE1, GSDMD-N), fibrosis (α-SMA), and inflammation (IL-1β, CRP). The inflammatory role of the p-STAT3/ANXA2 axis has been innovatively revealed, which exerted its role through the activation of the NLRP3/caspase-1 inflammasome, thus inducing the pyroptosis.
ANXA2 was an important member of the annexin protein family. It exhibited high-affinity binding for calcium ion and phospholipids [
15]. Aberrant expression of ANXA2 was associated with the development of many cancers [
39].
Anxa2 was significantly elevated in both NASH mouse liver and LPS-treated hepatocytes. Meanwhile, inhibiting the overexpression of
Anxa2 can significantly improve hepatocyte pyroptosis and liver fibrosis, either in NASH mice in vivo or LPS-induced hepatocytes in vitro. This suggests a causal link between
Anxa2 and hepatocyte pyroptosis in NASH, at the transcriptional level. In addition, the loss of function experiments can also be applied for screening specific drugs. Some known drugs have been applied for protecting against cell pyroptosis. Metformin protects against intestinal ischemia-reperfusion injury in a TXNIP-NLRP3-GSDMD-dependent manner [
48]. The nature compound Wedelolactone with strong anti-inflammatory and antioxidant activities, which could protect against acute pancreatitis and relevant lung injury against pyroptosis and ferroptosis [
49]. The methane offered a protective effect for septic mice via its anti-inflammation, anti-oxidation, anti-pyroptosis, and anti-apoptosis properties [
50].
To explore the upstream regulators of ANXA2, we used synthetic bioinformatics tools to identify p-STAT3 as a transcription factor responsible for ANXA2 upregulation in NASH. Mice with overexpressed p-STAT3 have been described as a model of systemic sclerosis associated with organ fibrosis [
51]. Nonetheless, the pro-inflammatory effects of p-STAT3 have been validated under different conditions [
52,
53]. Studies have shown that IL-6 played a clear role in liver deterioration and tumor progression, directly affecting the survival of hepatocellular carcinoma patients. Overexpressed IL-6 level was positively correlated with poor liver function, tumor progression, clinical severity, and 6-month mortality. The mechanism of IL-6 biological activity was mainly through the activation of tissue p-STAT3 [
54]. The p-STAT3 has also been a key protein related to epithelial-mesenchymal transition (EMT) in hepatoma cells, and its elevated level indicated the tendency of EMT in inflammatory hepatocytes [
55]. The members of the TNF protein superfamily were among the best-characterized inducers of hepatocyte cell death, a hallmark of liver injury [
56]. The pro-inflammatory cytokine TNF-α was another crucial factor for the activation of caspase-1-dependent hepatocyte pyroptosis in patients with obesity and NASH [
12]. In our study, p-STAT3 promoted the expression of
Anxa2 at the transcription level via binding with the promoter of
Anxa2. The specific inhibition of p-STAT3 can significantly reduce the expression of ANXA2, C-CASPASE1, ASC, and GSDMD-N, thereby improving hepatocyte pyroptosis and fibrosis in NASH. Meanwhile, overexpression of
Anxa2 significantly reversed the effect of inhibiting p-STAT3. Caspase-1 acted as the player of the p-STAT3/ANXA2 axis to promote hepatocyte pyroptosis and fibrosis, while the inhibition of its activity can significantly reduce the pro-fibrosis effect.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.