INTRODUCTION
Hypoxia, known to be involved in the whole process of liver fibrosis, causes the microenvironmental homeostasis imbalance due to the alterations in cellular metabolism, and the oxidative stress-related autophagy induced by hypoxia is a major contributor to the development of liver fibrosis [
1,
2]. Autophagy performs complex regulatory functions in different cellular contexts [
3]. It is widely recognized that autophagy in hepatocytes could degrade intracellular abnormal proteins and reduce the expression of pro-inflammatory cytokines, thereby alleviating liver fibrosis. On the other hand, autophagy in hepatic stellate cells (HSCs) could promote liver fibrosis by providing energy for the activation of HSCs [
4]. Although much has been done on anti-hepatic fibrosis by inhibiting HSC autophagy in previously reported studies, little progress has been made there. Since hepatocytes account for more than 80% of the liver tissue and function as the main functional cells responsible for liver metabolism and detoxification, hypoxia-induced autophagy mainly occurs in hepatocytes in the initial stage. The change of autophagic flux can deliver inflammatory signals to the liver microenvironment, leading to the progression of liver fibrosis. Therefore, there potentially exists a novel strategy for exploring the mechanism of hypoxia-induced autophagy in hepatocytes so that an efficient and precise regulatory approach can be developed to treat liver fibrosis clinically.
PCSK9, a serine protease, is mainly secreted by hepatocytes binding to low-density lipoprotein receptor (LDLR) and promoting its internalization and lysosomal degradation [
5]. Thus, PCSK9 inhibition increases LDLR expression in hepatocytes, enhancing circulating LDL uptake and reducing plasma cholesterol levels. PCSK9 monoclonal antibody (mAb) that inhibits the combination of PCSK9 and LDLR has been widely used in the treatment of cardiovascular diseases with abnormal lipid metabolism, with no serious adverse reactions seen clinically [
6]. Nevertheless, the role of anti-PCSK9 treatment is not limited to LDL-cholesterol regulation [
7]. PCSK9 inhibitor plays an important role in the clearance of pathogenic lipids including lipopolysaccharide (LPS). Several lines of evidence have suggested that anti-PCSK9 therapy could positively affect sepsis and septic shock [
8‐
10]. In our previously reported study, the evidence showed that anti-PCSK9 treatment ameliorated liver injury by enhancing LPS uptake in hepatocytes [
11]. Additionally, PCSK9 is known to be involved in the regulation of autophagy, playing different roles in a variety of cell types. In the primary mouse cardiomyocytes, PCSK9 inhibition could significantly reduce autophagy via activation of the ROS-ATM-LKB1-AMPK signaling pathway [
12]. Furthermore, anti-PCSK9 therapy has a promising potential of reducing autophagy by inhibiting mammalian rapamycin targeting protein (mTOR) via increasing protein kinase B. However, the role of anti-PCSK9 therapy remains controversial in treating liver fibrosis, liver inflammation, and nonalcoholic steatohepatitis (NASH) [
13‐
15].
In the current study, we discovered the effect of anti-PCSK9 therapy on liver fibrosis when the hypoxia-induced autophagy was inhibited in hepatocytes. In investigating the probable signaling pathway involved in anti-PCSK9 treatment, furthermore, we observed the experimental evidence that anti-PCSK9 therapy could be an effective anti-fibrosis agent.
MATERIALS AND METHODS
Reagents
A list was made of the reagents (Supplementary Table
S1).
Cells
The mouse liver cell line AML12 (Zhong Qiao Xin Zhou Biotechnology Co., Ltd., Shanghai, China) were cultured in DMEM/F12 supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (P/S). The mouse mononuclear macrophage cell line RAW264.7 were purchased from National Collection of Authenticated Cell Cultures and cultured in DMEM supplemented with 10% FBS and 1% P/S. The cells were maintained under a humidified atmosphere of 95% air and 5% CO2 in an incubator at 37 °C. The hypoxic condition was created by incubating hepatocytes in a controlled atmosphere incubator with an atmosphere of 1% O2, 5% CO2, and balance N2 at 37 °C. The cells, seeded in cell culture flasks with serum-containing medium, were routinely passaged every 3 days using trypsin-ethylenediaminetetraacetic acid. All incubations were performed under four passages.
Primary hepatocytes were routinely isolated from male C57BL/6 J mice through in situ enzymatic digestion of the liver with the collagenase I method, followed by density centrifugation, as previously described [
16]. The freshly isolated hepatocytes were seeded into collagen-coated Petri dishes to be grown in RPMI medium 1640 supplemented with 1.0 g/L glucose, 10% FBS, and 1% P/S in a humidified atmosphere of 5% CO
2 at 37 °C. The medium was changed 6 h after plating. PCSK9 inhibitor Evolocumab (Amgen Inc., USA) at a concentration of 100 μg/mL was then incubated for 24 h in hepatocytes.
Cell Transfection
AML12 cells were seeded into 6-well culture plates 24 h before transfection. Reaching 70–80% confluence, the cells were separated into four groups to be transfected with each of the following siRNAs: non-targeting scrambled control siRNA (siNC) group, siRNA1 group, siRNA2 group, and siRNA3 group. AML12 cells were transfected with PCSK9 siRNA and HIF-1α siRNA, respectively. These reagents were transfected into cells via Lipo8000 Transfection Reagent according to the manufacturer’s protocol. The medium was changed 6 h after transfection. Following siRNA transfection and 24-h incubation, the follow-up experiments were conducted. A list was made of the synthesized oligos (Supplementary Table
S2). Knockdown (KD) efficiency was determined by qRT-PCR and Western blot (Supplementary Figure
S1 and Table
S3).
Animals
Male C57BL/6 J mice aged 4–6 weeks, purchased from Slack Laboratory Animal Co., Ltd. (Shanghai, China), were maintained in a light and temperature controlled environment, with free access to water and food. With 1 week of adaptive feeding, the mice were randomly divided into the following 7 groups: negative control (NC), carbon tetrachloride (CCl4)-treated, NC + AAV8-sgPCSK9, CCl4 + AAV8-sgPCSK9, CCl4 + Evolocumab, NC + YC-1, and CCl4 + YC-1. In all NC groups, the mice were injected intraperitoneally with olive oil (5 μL/g) twice a week for 8 weeks. In all CCl4 groups, the mice were intraperitoneally injected with a reagent consisting of 10% CCl4/olive oil (5 μL/g) twice a week. The Evolocumab-treated mice received weekly subcutaneous injections of Evolocumab (10 μg/g), and the AAV8-sgPCSK9-treated mice received CRISPR-Cas9 adeno-associated virus (AAV) via tail vein 2 weeks before intraperitoneal injection of olive oil or CCl4. The AAV injection via tail vein was repeated 1 month later. Those which were treated with YC-1 were intraperitoneally injected with YC-1 (10 μg/g) daily. All mice were euthanized 48 h after the last injection of olive oil or CCl4. The blood and liver samples were collected for subsequent evaluation. All animal procedures were performed in accordance with the guiding principles for the care and use of laboratory animals approved by the Animal Care Committee of Zhongshan Hospital and Fudan University.
Elisa for PCSK9 and TNF-α
In the mouse serum samples or cell culture medium, PCSK9 secretion was measured using the mouse PCSK9 ELISA kit. Similarly, in the mouse serum samples or cell culture medium, tumor necrosis factor-α (TNF-α) was done using the mouse TNF-α ELISA kit according to the manufacturer’s instructions. The linear regression was performed with different concentrations of standards. The samples were combined with peroxidase-labeled IgG anti-PCSK9 or anti-TNF-α antibody attached to the wells of a 96-well plate, which was followed by enzymatic reaction and substrate supplement to be measured at 450 nm with a microplate reader. In each sample, the concentration was calculated according to the linear equation.
Western Blot
AML12 cells and liver tissues homogenized in RIPA buffer underwent a 10-min centrifugation at 4 °C for 10,000 × g. The protein concentration was assessed by BCA method. With the samples heated up to 100 °C for 10 min with loading buffer, 20 μg of protein was resolved by electrophoresis on 10–15% SDS-PAGE gels and transferred to cellulose nitrate filtration membranes. The membranes were blocked for 1 h in double distilled water (ddH
2O) with 5% bovine serum albumin (BSA) powder at room temperature, before incubated overnight at 4 °C with 1:1000 diluted primary antibodies. A list was made of the primary antibodies for Western blot assay (Supplementary Table
S4). Then, washed with TBST (100 mM Tris–HCl, pH 7.5, 0.9% NaCl, 0.1% Tween 20), the membranes were incubated for 1 h with goat anti-rabbit or goat anti-mouse secondary antibodies at room temperature. The expressed proteins were measured with ECLTM Western Blotting Detection Reagents, and the images were recorded using ChemiScope 6000 (Qinxiang Scientific Co., Ltd., Shanghai, China). To determine the optical density of the bands, ImageJ software (NIH, Bethesda, MD, USA) was used.
Immunohistochemistry
When formalin-fixed and paraffin-embedded, the liver tissue sections were sectioned at 4 µm to be deparaffinized, hydrated, and subjected to heat-induced antigen retrieval according to the standard protocols as previously described [
17]. The liver sections were blocked to be incubated overnight at 4 °C with anti-PCSK9 antibody (Abcam, ab31762) diluted (1:200) in double distilled water containing 5% BSA. The sections were subsequently washed before incubated with HRP-conjugated goat anti-rabbit IgG secondary antibodies (Jackson; 1:500), followed by incubation with 3, 3′-diaminobenzidine tetrachloride for 5–10 min and visualization of specific staining by light microscopy. Under a high-power field with BX51 (Olympus, Japan), the images were captured.
Immunofluorescence
After 1-h blocking with 5% BSA at 37 °C, the frozen sections of liver tissue were incubated overnight at 4 °C with HIF-1α antibody (CST, 36169) diluted (1:400) in double distilled water containing 5% BSA and then at 37 °C with Cy3-conjugated goat anti-rabbit IgG (Servicebio, 1:100) for 1 h in the dark. The images were taken under confocal microscopy (FV3000, Olympus, Japan).
Measurement of ROS by Flow Cytometry
Grown in six-well plates, AML12 cells were harvested after hypoxia treatment or mPCSK9 protein stimulation for 24 h. The quantitative measurements of total cellular ROS generation were performed using a Reactive Oxygen Species Assay Kit. The cells were stained with 10 mM DCFH-DA for 20 min at 37 °C. The cellular ROS levels were detected by flow cytometry (Bioscience Aria III, Becton Dickinson, USA). The data were analyzed on FlowJo software (V10.8.1, Becton Dickinson, USA).
Transmission Electron Microscope
The mice samples were perfused with Ringer’s solution before undergoing 0.15 M cacodylate buffer containing 2.5% glutaraldehyde, 2% paraformaldehyde, and 2 mM CaCl2 at 37 °C for 5 min. With post-perfusion, the liver tissue was carefully excised and fixed again in the same fixative overnight at 4 °C. Washed in 0.15 M cacodylate buffer (3 × 5 min), the tissue was post-fixed in 1% osmium tetroxide and 0.3% potassium ferrocyanide in 0.15 M cacodylate buffer with 2 mM CaCl2 for 1 h in the dark, which was followed by 3 × 10 min’ washes in ddH2O. The liver tissue was en bloc stained with 2% aqueous uranyl acetate overnight at 4 °C, before dehydrated in a 30%, 50%, 70%, and 100% Spurr’s resin mixed with 100% ethanol, embedded in fresh 100% Spurr’s resin in silicon molds, and polymerized at 60 °C for 48 h. Once polymerized, the resin blocks were faced, with 70 nm ultrathin sections cut on a Leica EM UC7 ultramicrotome (Leica-Microsystems, Vienna, Austria), picked up on copper formvar/carbon support film grids (PN FCF100H-Cu, Electron Microscopy Sciences, Hatfield, PA), and post-stained with 2% uranyl acetate for 15 min, then with Reynolds lead citrate for 2 min, before imaged on a JEOL JEM-1400 TEM with an AMT XR111 8 Megapixel scintillated CCD camera.
Whole-Transcriptome Amplification and RNA-Sequencing Analysis
Isolated from the liver tissues using Trizol reagent, total RNA was applied to RNA-sequencing (RNA-seq) analysis. Library preparation was performed on an Apollo Library Prep System (Takara, Shiga, Japan) using the TruSeq Stranded mRNA Sample Prep Kit and Liver RNA-seq on an Illumina HiSeq 4000 platform performing the 75-base single-end sequencing. TopHat (v.2.0.13) and hisat2 were used to map the clean reads to each gene, with the raw data normalized to Fragments Per Kilobase of exon model per Million mapped fragments (FPKM) for subsequent analyses. Bioinformatics analyses were performed as previously described [
18,
19]. Differentially expressed genes (DEGs) were identified with the limma package, which implemented an empirical Bayesian approach to estimating gene expression changes using the moderated
t-test. |log FC|> 0.5 and
p < 0.05 were considered as the cutoff criteria for screening DEGs. The functional enrichment analyses of the detected DEGs were performed with the cluster Profiler package, with the terms of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) identified with a cutoff of
p < 0.05.
Furthermore, Gene Set Enrichment Analysis (GSEA) was used to identify the pathways that were upregulated or downregulated in the CCl4-treated mice with/without PCSK9 inhibition. From the MSigDB database, the gene sets were obtained for analysis. And the enrichment analyses of DEGs were performed using Metascape. With the pathway discovered, Cytoscape (Ver3.9.1) was employed to search for differential genes. With all this done, the mechanism diagram was rendered on Biorender software.
Statistical Analysis
All data were presented as mean ± standard deviation (SD). The statistical differences between two groups were determined using the unpaired t-test. The multiple comparisons were analyzed through one-way analysis of variance (ANOVA). Statistical analysis was performed using Prism version 8.0.2 (GraphPad Software, San Diego, CA), all comparisons considered statistically significant when p < 0.05.
DISCUSSION
Anti-PCSK9 therapies have gained prominence in the treatment of hypercholesterolemia, and no serious adverse events have been observed from laboratory to clinical application. At present, however, the data were obscure when the side effect of anti-PCSK9 therapy is considered on liver function, especially when hepatocytes are the main cells producing PCSK9, and cholesterol metabolism is also mainly completed by hepatocytes [
24]. It is of great importance for us to investigate the effect of PCSK9 on hepatic metabolism and biological function.
A previous investigation of mouse plasma, liver, and bile acids revealed that PCSK9-deficient mice had well-maintained cholesterol metabolism and hepatic homeostasis [
25]. PCSK9 mAb therapy significantly reduced hepatic steatosis, liver inflammation, and fibrosis in patients with severe hyperlipidemia [
26,
27]. It appears that anti-PCSK9 therapy not only improves hypercholesterolemia but also treats steatohepatitis. However, some studies have found the unfavorable evidence that high-cholesterol diets in PCSK9 knockout mice showed increasing hepatic free cholesterol and cholesterol crystals, and fibrotic steatohepatitis, even with a higher predisposition to liver cancer as compared with WT mice [
28], and that PCSK9 KO exacerbated NASH, fibrosis, and liver injury in the presence of excess dietary fats in mice [
13]. These findings suggested that anti-PCSK9 could not protect against hepatic steatosis and liver injury.
In view of which, we are not in a position to know whether anti-PCSK9 treatment can improve liver fibrosis. As previously reported, we verified that PCSK9 expression was elevated in the mice and human fibrotic liver samples, and PCSK9 inhibition by curcumin or PCSK9 KO significantly alleviated liver inflammation and fibrosis [
11,
17]. In the current study, we used a CCl
4-induced model of canonical liver fibrosis mice to get a deep insight into the mechanism, with the application of both exogenous PCSK9 inhibition with PCSK9 mAb Evolocumab and endogenous PCSK9 inhibition by CRISPR-Cas9 AAV.
It is well recognized that PCSK9 acts as a cholesterol regulator and a putative cytokine with pro-inflammatory effect. It was reported that the mRNA levels of IL-1, IL-6, TNF-α, CXCL2, and CCL2 increased when macrophages were incubated with recombinant or HepG2-derived PCSK9 protein [
29]. PCSK9 inhibition-caused overexpression of hepatic LDLR promoted LPS clearance, lowering the inflammatory response and increasing survival time in the mice after sepsis [
9,
10]. In the patients with PCSK9 loss-of-function mutations, better clinical outcomes were observed during septic shock [
8]. In our previous study, similarly, PCSK9 inhibition was observed to reduce intestinal endotoxemia, thereby attenuating liver fibrosis [
11].
In the current study, of note were the increase of PCSK9 that paralleled autophagy in the hepatocytes, and the inhibition of PCSK9 that decreased autophagy and alleviated liver fibrosis, which was different from any previously reported evidence. It is generally believed that autophagy in hepatocytes, endothelial cells, and macrophages indirectly protects against liver fibrosis, while autophagy in HSCs aggravates liver fibrosis [
30]. Contrary to the widely accepted theory, however, our study demonstrated that PCSK9 inhibition reduced autophagy in the hepatocytes, thereby attenuating liver inflammation and fibrosis. Similar to our findings, the inhibition of hepatocyte autophagy was reported to ameliorate Fas/FasL-regulated hepatocyte apoptosis, HSC activation, and liver fibrosis [
31]. For this, a plausible explanation is that the occurrence of autophagy is a dynamic process, which may be beneficial at the initial stage and harmful at an advanced stage [
3], where autophagy could even lead to cell death [
32,
33]. Hepatocytes were revealed to own a high level of autophagic flux due to the increased abundance of lysosomes and lysosomal enzymes, and enhanced autophagy could regulate the progression of hepatocyte death [
34]. Therefore, enhanced autophagy is detrimental in hepatocytes, which further supports our findings.
Hypoxia, a common characteristic during the whole process of liver inflammation and fibrosis, affects mitochondrial respiratory chain function in hepatocytes, which in turn may increase intracellular ROS generation through the electron transport chain, an event that may lead to oxidative stress, cell damage, and death [
35]. Studies have proved that hypoxia induces the expression of PCSK9 in cultured cardiomyocytes [
36,
37]; on the other hand, PCSK9 induces mitochondrial dysfunction and ROS generation in endothelial cells in the context of atherosclerosis [
38]. In our study, we found that hypoxia and its generated ROS induced the elevated expression of PCSK9 in the hepatocytes; that PCSK9 inhibition reduced hypoxia-induced ROS production; and that interfering with ROS production also affected the expression of PCSK9 under the hypoxic condition. From our study, both in vitro experiments and RNA-Seq analysis confirmed the evidence that a bidirectional interaction exists between PCSK9 and hypoxia-mediated ROS generation.
Autophagy is recognized as a common mechanism of modulating diverse signaling pathways [
39]. A notable regulator of autophagy is the AMPK signaling pathway, which is activated by hypoxia and metabolic stress [
40]. Metformin was reported to induce autophagy through the AMPK signaling pathway, relieve hypoxic stress, and promote the survival of random skin flaps by increasing neovascularization [
41]. AMPK was indicated to be required for hypoxia-mediated autophagy, which protected cardiomyocytes from ischemia–reperfusion injury [
42]. It is well recognized that AMPK increases autophagy effectively through multiple mechanisms. AMPK has been reported to repress the synthesis of mTOR complex 1 (mTORC1) through phosphorylating the TSC1-TSC2 complex, thereby reducing the activity of mTOR and alleviating the inhibition of mTOR on autophagy [
43,
44]. As an evolutionarily conserved protein kinase, mTOR plays a negative regulatory role in autophagy by inhibiting ULK1 activation [
45]. ULK1 was also phosphorylated and activated by AMPK, leading to the initiation of the autophagic cascade [
46]. In the current study, we observed that PCSK9 promoted the phosphorylation and activation of AMPK, thereby reducing the activity of mTOR, promoting the activation of ULK1, and ultimately enhancing autophagy.
Additionally, it is of more importance to investigate serum PCSK9 levels than hepatic PCSK9 expression. A previously reported multivariable linear regression analysis indicated that non-alcoholic fatty liver disease (NAFLD) score levels were independently related to higher PCSK9 levels [
47]. A further proof was provided that circulating PCSK9 concentrations were not associated with the severity of liver steatosis or histological markers of NASH [
15]. Furthermore, those who had liver cirrhosis produced significantly lower serum PCSK9 concentrations [
48], and those who had chronic hepatitis or liver cirrhosis showed serum PCSK9 levels that were 20–30% lower than the healthy controls [
49]. In order to confirm the reflection effect of PCSK9 on liver biochemical function, further investigations need to be conducted.
In summary, our study demonstrates that the expression of PCSK9 is closely associated with the level of hypoxia-induced autophagy during the development of CCl4-induced hepatic fibrosis. Anti-PCSK9 treatment alleviates liver fibrosis by regulating hypoxia-induced autophagy in the hepatocytes through AMPK/mTOR/ULK1 signaling pathway. Thus, PCSK9 has promising potential of becoming a new biomarker as a therapeutic target for the treatment of hepatic fibrosis.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.