Silencing of functional p53 attenuates NAFLD by promoting HMGB1-related autophagy induction

p53 heterozygous knockout mice (C57BL/6 background) were purchased from Biocytogen Co., Ltd. (Beijing, China). Animals were bred and maintained at 23 ± 2 °C with a 12-h light/12-h dark cycle at the Medical Science Institution of Zhejiang Province (Hangzhou, China). The p53 genotypes of the offspring mice were identified by the methods used by the facility (Supplementary Figure 1). Male adult wild-type (p53^+/+) and knockout (p53^−/−) mice weighing 19–23 g were randomly separated into two groups. For mice of both genotypes, the mice were randomly divided into two groups that were fed either a standard chow diet (SCD) provided by the Medical Science Institution of Zhejiang Province (Hangzhou, China) or a high-fat diet (D12492; Research Diets, New Brunswick, NJ, USA) for 8 weeks (n = 6). During the study, mice were given free access to water. The mice were killed, and the liver samples were separated and fixed in 10% neutral formalin or rapidly frozen in liquid nitrogen for further analysis. The fixed liver samples were embedded in paraffin, sectioned and then stained with hematoxylin and eosin (H&E) for histological examination according to the morphological criteria described previously [29]. The liver samples were analyzed with a transmission electron microscope (HT7700, Hitachi, Japan) after being fixed in 2.5% malondialdehyde. All institutional and national guidelines for the care and use of laboratory animals were followed.

For GTT, the mice were fasted for 16 h in advance in the 6th week of modeling. For ITT, the mice were fasted for 6 h in advance in the 7th week of modeling. A blank blood sample was taken from the tail end at time 0 to test the glycemic index using commercial test strips. The mice were then intraperitoneally injected with glucose (1 g/kg) for GTT or insulin (1 U/kg) for ITT, and blood samples were collected and tested to determine the glycemic index at 15, 30, 60, 90, and 120 min after injection. Feeding was resumed immediately after the experiment.

HepG2 and Huh7 cells (American Type Culture Collection, Manassas, VA) were maintained in Dulbecco’s modified Eagle medium (DMEM) containing 10% fetal bovine serum (FBS) and 100 U penicillin/streptomycin under 5% CO₂ at 37 °C. The cells were seeded overnight to allow them to adhere to the plate. PA was dissolved in phosphate-buffered saline (PBS) containing 25% BSA and added to the cell culture medium, followed by 24 h of incubation before harvest. For the evaluation of autophagic flux, the cells were treated with 10 μM chloroquine for 4 h.

Lipofectamine RNAiMAX reagent (Thermo Fisher Scientific, Waltham, MA USA) was used for siRNA transfection according to the manufacturer’s instructions. The siRNAs were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China), and the cells were transfected with the siRNAs at a final concentration of 50 μM and maintained for 48 h until harvest. The sequences of the siRNAs are as follows: NC (forward: UUCUCCGAACGUGUCACGU dTdT, reverse: ACGUGACACGUUCGGAGAA dTdT); human p53 (forward: CACUACAACUACAUGUGUA dTdT, reverse: UACACAUGUAGUUGUAGUG dTdT); human HMGB1 (forward: CUGAUGCAGCUUAUACGAA dTdT, reverse: UUCGUAUAAGCUGCAUCAG dTdT); mouse p53 (forward: GUAAACGCUUCGAGAUGUU dTdT, reverse: AACAUCUCGAAGCGUUUAC dTdT); mouse Hmgb1 (forward: AUGCAGCUUAUACGAAGAUAA dTdT, reverse: UUAUCUUCGUAUAAGCUGCAU dTdT). Lipofectamine 3000 reagent (Thermo Fisher Scientific) was used for plasmid transfection, and 2 μg plasmids were transfected into each well (6-well plate). The pEGFP-LC3 plasmid was a generous gift from Professor Min Zheng (Zhejiang University, Hangzhou, China).

Hepatocytes from male p53 wild-type or null mice (8 weeks) were isolated using a two-step perfusion and collagenase digestion method as described previously [30]. The isolated cells were seeded on plates coated with rat tail collagen (Yisheng Biotechnology, Shanghai, China) and cultured in DMEM (supplemented with 10% FBS and 100 U penicillin/streptomycin) for the first 4 h, and then the medium was exchanged (DMEM/F12 without FBS) until harvest.

The hepatic and intracellular TG levels were analyzed using a commercial kit (E1013-105, Applygen Technologies, Beijing, China) according to the manufacturer’s instructions. Briefly, the liver tissues or collected cells were treated with lysis buffer on ice. The homogenates or lysates were incubated at 70 °C for 10 min and then centrifuged at 2000 rpm for 5 min at room temperature. The supernatant was assessed with the relevant working solution. The protein concentration in the resulting lysates was determined using a bicinchoninic acid protein assay kit (P1513, Applygen Technologies Inc.). The TG values were normalized according to the total protein levels.

Cryosections of liver tissues or cell slides were stained with oil red O according to a standard protocol (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). The results were examined using a Leica DM5000B microscope (Leica, Heidelberg, Germany). The data shown were obtained from one representative experiment with three independent replicates.

Paraffinized liver sections were deparaffinized with xylene and rehydrated, followed by incubation with 3% hydrogen peroxide. Heat epitope retrieval was performed in target retrieval solution at pH 7.5 for 20 min. The sections were preblocked with 10% goat serum (ZSGB-BIO, Beijing, China) and then incubated with HMGB1 antibody (Cell Signaling Technology, Beverly, MA, USA) at a dilution of 1:100 overnight at 4°C. Tissue sections were stained with HPR secondary antibody (dilution: 1:100, ZSGB-BIO) for 1 h at 37 °C in an incubator. Immunoreactivity was detected using a DAB kit (ZLI-9017, ZSGB-BIO, Beijing, China) and visualized as brown staining. Slides were counterstained with hematoxylin. The data shown were obtained from one representative experiment with three independent replicates.

Cells were seeded onto rat tail collagen-coated cover glasses (in a 24-well plate). The cell slides were fixed with 4% paraformaldehyde and incubated with 0.3% Triton X-100. Then, the cells were incubated in goat serum solution (ZSGB-BIO, Beijing, China) for 30 min at room temperature, followed by incubation with primary antibodies at 4 °C overnight and a secondary antibody (goat anti-rabbit IgG H&L Alexa Fluor 488/594 ab150077/ab150080, Abcam, Cambridge, UK) at room temperature for 1 h. DAPI staining (1 μg/ml, Thermo Fisher) was performed immediately before mounting. Images were captured with an Olympus IX81-FV1000 confocal microscope. For the liver slides, the first few steps were the same until the overnight primary antibody incubation was performed, and the subsequent steps were the same as those used for the cell slides described above. The data shown were obtained from one representative experiment with three independent replicates.

Total RNA was extracted from the liver tissues or cultured cells using RNAiso Plus reagent according to the manufacturer’s instructions (TaKaRa Biotech, Kyoto, Japan), followed by random reverse transcription to cDNA using the PrimeScript RT reagent kit DRR036A (TaKaRa Biotech). Quantitative reverse-transcription polymerase chain reaction (PCR) analysis was performed using the SYBR Premix Ex Taq II kit RR820A (TaKaRa Biotech) on a CFX96 real-time PCR system (Bio-rad, Hercules, CA, USA). The glyceraldehyde-3-phosphate dehydrogenase gene was included with each run for the normalization expression. The gene-specific primer sequences were as follows (5′–3′): human GAPDH (forward: TCAACGACCACTTTGTCAAGCTCA, reverse: GCTGGTGGTCCAGGGGTCTTACT); human p53 (forward: GAGGTTGGCTCTGACTGTACC, reverse: TCCGTCCCAGTAGATTACCAC); human HMGB1 (forward: TATGGCAAAAGCGGACAAGG; reverse: CTTCGCAACATCACCAATGGA); mouse Gapdh (forward: AGGTCGGTGTGAACGGATTTG, reverse: GGGGTCGTTGATGGCAACA); mouse p53 (forward: AGGTCGGTGTGAACGGATTTG, reverse: GGGGTCGTTGATGGCAACA); and mouse Hmgb1 (forward: AGGTCGGTGTGAACGGATTTG, reverse: GGGGTCGTTGATGGCAACA).

Total proteins were extracted from liver tissues or cultured cells using RIPA assay lysis buffer (Applygen Technologies, Beijing, China) and quantified using the BCA Protein Assay (Beyotime, Jiangsu, China). Nuclear and cytoplasmic proteins were extracted using a commercial kit (P1201, Applygen Technologies, Beijing, China). Forty micrograms of protein extract was separated by 12% SDS-PAGE and electrophoretically transferred onto polyvinylidene fluoride membranes (0.2 μM, Millipore, MA). Then the membranes were blocked with 5% nonfat milk in Tris-buffered saline at room temperature and incubated at 4 °C overnight with primary antibodies, including GAPDH (#2118), ACTB (#4970), LC3 (#4108), Beclin1 (#3495), HMGB1 (#6893), and LaminB1 (#13435) from Cell Signaling Technology, p53(ab26) from Abcam, p21(556430) from BD (Franklin Lakes, NJ, USA), and P62 (#PM045) from MBL International (Woburn, MA, USA). Incubation with HRP-conjugated anti-rabbit or anti-mouse IgG secondary antibody (Dawen Biotec, Hangzhou, China) was performed for 1 h at room temperature. Specific bands were visualized using an ECL detection kit (P0018, Beyotime, Jiangsu, China) and photographed with a ChemiScope 6000 Pro Touch (Clinx Science Instruments, Shanghai, China). The data shown were obtained from one representative experiment with three independent replicates.

The data were expressed as the mean ± SEM and analyzed using GraphPad Prism 6 (GraphPad Software Inc., San Diego, CA). One-way ANOVA followed by Dunnett’s multiple comparison post hoc test or an unpaired Student’s t test was used for the statistical analysis. Only the comparisons indicated above the bars were performed, and differences were considered significant if the probability (p value) was less than 0.05 (p < 0.05).

To investigate the effect of p53 loss on autophagy in hepatocytes, we performed immunofluorescent staining of mouse liver sections. As indicated by LC3 immunostaining, the number of punctate autophagosomes was significantly increased in p53-null mouse livers (Fig. 1a). Western blotting analysis revealed an increase in the LC3II/LC3I ratio, as well as a decrease in the p62 expression in p53-null mouse livers, which demonstrated the upregulation of autophagic flux (Fig. 1b). Furthermore, we found that p53 loss-induced autophagy was associated with an increase in BCL1 protein expression (Fig. 1b).

We next investigated how autophagy was regulated when p53 was silenced in human liver-derived HepG2 and Huh7 cells. We first confirmed p53 function in these two cell lines by measuring the mRNA level of the p53 target gene p21 after treating cells with the p53 activator Nutlin3a and inhibitor PFTa. As illustrated in Supplementary Figure 2a–d, nutlin3a treatment time-dependently induced p21 transcription in HepG2 cells, while the same treatment did not induce p21 in Huh7 cells. Furthermore, PFTα treatment displayed a rapid and significant inhibition effect on p21. However, no such inhibition effect was observed in Huh7 cells. These results suggest that p53 is functional in HepG2 cells, but not in Huh7 cells. Next, cells were transfected with siRNA to knockdown p53 mRNA levels with a silencing efficiency of 91.4% for HepG2 cells and 89.3% for Huh7 cells (Supplementary Figure 3a, b). Confocal analysis showed that p53 knockdown increased the number of punctate autophagosomes in HepG2 cells (Fig. 2a). In contrast, knockdown of p53 did not increase the number of punctate autophagosomes in green fluorescent protein (GFP)-LC3-transfected Huh7 cells (Fig. 2b). A protein expression assay also found increased p62 degradation, an increased LC3II/LC3I ratio and an increased level of BCL1 expression in HepG2 cells with p53 silencing (Fig. 2c). However, in Huh7 cells, p53 knockdown did not affect p62 expression or the LC3II/LC3I ratio (Fig. 2d). In addition, we observed an increase in the LC3II/LC3I ratio in cells with p53 silencing followed by chloroquine (a well-known autophagy inhibitor) treatment versus that in cells treated with vehicle alone, indicating that p53 silencing indeed induced autophagic flux (Supplementary Figure 4).

To investigate the association between autophagy and lipid degradation in hepatocytes, we performed colocalization analysis of autophagosomes and lipids. In PA-treated HepG2 cells, immunolabelled LC3 was found to colocalize with BODIPY 493-labeled lipids (Supplementary Figure 5a). Furthermore, when the cells were treated with chloroquine, the intracellular TG level was significantly elevated in the absence of PA (Supplementary Figure 5b), which was then confirmed by ORO staining (Supplementary Figure 5c). These results suggested a strong association between autophagosomes and intracellular lipids in the liver.

To explore the effect of p53 loss on HFD-induced NAFLD in mice, male p53 WT and null mice were fed either a SCD or HFD for 8 weeks. No significant difference in food or water intake was observed in any of the animal groups (data not shown). H&E staining as well as ORO staining suggested obvious hepatic steatosis in mice fed a HFD compared to mice fed a SCD (Fig. 3a, b). Additionally, ITT and GTT indicated impaired insulin sensitivity in mice fed a HFD (Fig. 3c, d). Meanwhile, mice fed a HFD exhibited higher serum insulin level (Fig. 3e). Furthermore, HFD feeding also significantly elevated hepatic TG levels (Fig. 3f). Interestingly, in HFD-fed mice, p53-null mice exhibited the attenuation of hepatic steatosis and lipid and TG accumulation as well as lower insulin and improved ITT/GTT performance compared with wild-type mice (Fig. 3a–e). These results suggest a protective effect of p53 loss on NAFLD in vivo.

We confirmed the protective effect of the knockdown of functional p53 on NAFLD in vitro. Firstly, we confirmed the function or dysfunction of p53 by detecting the induction effect of p53 agonist on p21 transcription in primary hepatocytes isolated from wild type or knockout mice (supplementary Figure 2e, f). We established a NAFLD model in mouse primary hepatocytes via 400 μM PA treatment for 24 h. Elevated intracellular TG levels and obvious lipid staining demonstrated steatosis in both cell lines (Fig. 4a, b). In hepatocytes from p53-null mice, reduced TG levels and attenuated lipid staining were observed. Similarly, 400 μM PA treatment for 24 h also resulted in obvious TG level elevation and strong lipid staining, while p53 knockdown resulted in reduced TG levels and attenuated lipid staining (Fig. 4c, d). However, p53 silencing had no significant effect on steatosis in Huh7 cells at multiple PA concentrations (Fig. 4e, f). Interestingly, p53 silencing exacerbated TG accumulation caused by 400 μM PA treatment in Huh7 cells.

HMGB1 is an autophagy regulator that activates BCL1 in the cytoplasm. We examined the distribution of HMGB1 in mouse livers via immunohistochemical analysis. Interestingly, HMGB1 existed mainly in the nucleus zone in hepatocytes derived from wild-type mice, but was more highly distributed in the cytoplasm in hepatocytes from p53-null mice (Fig. 5a). Additionally, the nuclear, cytoplasmic and total HMGB1 protein expression in mouse livers as well as in HepG2 and Huh7 cells was determined. As a result, in hepatocytes from p53-null mice and in p53-silenced HepG2 cells, HMGB1 expression was decreased in the nucleus but elevated in the cytoplasm, while no significant change was observed in the total level (Fig. 5b, c). Such phenomena were not observed in p53-silenced Huh7 cells (Fig. 5d). These results indicated that loss/silencing of functional p53 could cause the translocation of HMGB1 from the nucleus to the cytoplasm.

To further investigate how HMGB1 participates in the regulatory mechanism of p53 in autophagy and NAFLD, we knocked down HMGB1 expression in HepG2 cells and primary hepatocytes with siRNAs with silencing efficiencies of 77.3% and 70.3%, respectively (Supplementary Figure 3d, e). In both cell lines, HMGB1 silencing counteracted the elevation of the LC3II/LC3I ratio, the upregulation of BCL1, and the degradation of p62 induced by p53 silencing (Fig. 6a, b). What’s more, we found upregulation of BCL1, inhibition of p62 and increase of LC3II/I ratio, following HMGB1 knocking down both in primary hepatocytes and HepG2 cells (Fig. 6a, b). Furthermore, we tested the effect of HMGB1 knockdown on PA-induced steatosis in p53-silenced cells. In HepG2 and primary hepatocytes from wild-type mice, HMGB1 silencing attenuated TG and lipid accumulation induced by PA administration (Fig. 6c–f). Interestingly, in both cell lines, HMGB1 knockdown offset the alleviating effect of p53 silencing on hepatic TG accumulation and steatosis. These results collectively suggested that HMGB1 contributed to the protective effect of p53 silencing on NAFLD by regulating autophagy. Furthermore, we also silenced p53 by siRNA transfection in primary hepatocytes from wild-type mice with a silencing efficiency of 92.5% (Supplementary Figure 3c) and found that Hmgb1 knockdown also blocked autophagy induction and alleviated NAFLD symptoms via p53 knockdown (Supplementary Figure 6a–c).