Introduction

Nonalcoholic Fatty Liver Disease (NAFLD) is a common human affliction. Its global prevalence, currently about 25%, has been increasing (Diehl and Day, 2017; Fan et al., 2017; Younossi et al., 2018). NAFLD may progress from nonalcoholic fatty liver to nonalcoholic steatohepatitis, cirrhosis and even hepatic cancer (Friedman et al., 2018; Sheka et al., 2020). Multiple factors may contribute to its pathogenesis (Friedman et al., 2018); prominent among these are the lipopolysaccharides (LPSs, endotoxins) produced by many of the Gram-negative bacteria that inhabit the intestine. Gut-derived LPS may translocate into the portal venous system and traffic to the liver, triggering or exacerbating hepatic inflammation (Albillos et al., 2020; Carpino et al., 2020; Han et al., 2021; Kazankov et al., 2019; Leung et al., 2016; Munford, 1978; Wang et al., 2022).

The LPS molecules that contribute to NAFLD pathogenesis are able to stimulate host cells because their lipid A structure is recognized by MD-2/TLR4 receptors (Ye et al., 2012). Most γ-Proteobacteria such as E. coli produce stimulatory hexaacyl LPS (with six acyl chains) while Bacteroidetes produce non-stimulatory LPS that has four or five acyl chains (Anhe et al., 2021; d’Hennezel et al., 2017). NAFLD is often associated with intestinal dysbiosis produced by increased abundance of γ-Proteobacteria, which leads to tissue inflammation and increased intestinal permeability that allows even more gut-derived LPS to reach the liver (Albillos et al., 2020; Aron-Wisnewsky et al., 2020a; Aron-Wisnewsky et al., 2020b; Mouries et al., 2019; Rodrigues et al., 2024). Although gut-derived LPS is known to induce hepatic inflammation and to exacerbate NAFLD, how LPS influences hepatocyte fatty acid metabolism before NAFLD develops has not been well understood.

Acyloxyacyl hydrolase (AOAH) is a highly conserved animal lipase that is mainly expressed in macrophages, monocytes, neutrophils, microglia, dendritic cells, NK cells and ILC1 cells (Munford et al., 2020). It can inactivate Gram-negative bacterial LPSs by releasing two of the six fatty acyl chains present in the lipid A moiety (Fig 1A). It also can deacylate/inactivate oxidized phospholipids and lysophospholipids, molecules that are also known to contribute to NAFLD (Sun et al., 2020; Zou et al., 2021). We previously reported that AOAH is expressed by gut macrophages and dendritic cells and can inactivate bioactive LPS in feces (Cheng et al., 2023; Janelsins et al., 2014; Qian et al., 2018). Han et al. found that intestine-derived LPS can bind high-density lipoprotein 3 (HDL₃) and be inactivated by AOAH as it traffics to the liver via the portal vein (Han et al., 2021). AOAH also inactivates LPS in the liver, diminishing and shortening hepatic inflammation induced by bloodborne (i.v. injected) LPS (Ojogun et al., 2009; Shao et al., 2011; Shao et al., 2007). The enzyme’s ability to prevent NAFLD by inactivating gut-derived LPS had not been tested (Ojogun, 2008).

Fig 1.

In this study, we found that when mice were fed either normal chow or a high fat diet, AOAH reduced LPS-induced lipid accumulation in the liver, probably by decreasing the expression and activation of Sterol Regulatory Element-Binding Protein 1 (SREBP1), an important transcription factor that promotes fatty acid synthesis (Horton et al., 2002; Shimano and Sato, 2017). AOAH also reduced the expression of CD36 and Fabp3, fatty acid uptake-related genes, and increased that of fatty acid oxidation-related genes (Acot2 and Ppar-α). In addition, AOAH reduced hepatic inflammation and minimized tissue damage. Our results suggest that AOAH plays a regulatory role in ameliorating NAFLD and suggest that measures that increase AOAH abundance in the intestine, liver and/or bloodstream may help prevent this common disease.

Materials and methods

Mice

C57BL/6J Aoah^−/− mice were produced at the University of Texas Southwestern Medical Center, Dallas, Texas (Lu et al., 2003), transferred to the National Institutes of Health, Bethesda, Maryland, USA, and then provided to Fudan University, Shanghai, China. The mutated Aoah gene had been backcrossed to C57BL/6J mice for at least 10 generations. Aoah^+/+ and Aoah^−/−male mice were housed in a specific pathogen-free facility with 12-h light/dark cycle at 22 °C in Fudan University Experimental Animal Center (Shanghai, China). Aoah^+/+ and Aoah^−/− male mice were co-housed for at least 3 weeks before the start and throughout the experiments. All studies used protocols approved by the Institutional Animal Care and Use Committee (IACUC) of Fudan University (2023-DWYY-03JZS). All animal study protocols adhered to the Guide for the Care and Use of Laboratory Animals.

The NAFLD mouse model

Co-housed Aoah^+/+ and Aoah^−/− male mice were fed either a normal diet (NC) or a high fat calorie diet (D12492, Research Diets, USA) that contained protein: carbohydrate: fat (20:20:60, kcal%) plus high fructose (23.1 g/L; F3510, Sigma,USA) and glucose (18.9 g/L; G8270, Sigma, USA) in the drinking water (HFD) for 28 weeks (Liu et al., 2018).

Blood analysis

Total triacylglycerol, total cholesterol, low density lipoprotein (LDL), high-density lipoprotein (HDL), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) in mouse serum were measured in the Department of Laboratory Animal Science, Fudan University using ADVIA Chemistry XPT.

Liver histology

After livers were excised and fixed in 4% paraformaldehyde for 18 h, they were sectioned and stained with hematoxylin and eosin (H & E) and Oil Red O. The samples were examined for steatosis, hepatocyte ballooning degeneration, lobular inflammation and lipid droplets using a Nikon E200 microscope.

NAFLD activity score

Histological analysis of the liver was performed based on the Sheka NAFLD scoring criteria (Sheka et al., 2020). Liver steatosis is an infiltration of hepatic fat with minimal inflammation and is graded based on the fat percentage in hepatocytes: grade 0 (< 5%), grade 1 (5% – 33%), grade 2 (33% – 66%), and grade 3 (> 66%). Inflammatory activity is manifested by two factors: grade 0 (no inflammation), grade 1 (< 2 foci per 200 × field), grade 2 (2 – 4 foci per 200 × field), grade 3 (> 4 foci per 200 × field), and the presence of hepatocyte ballooning degeneration: no ballooned cells (grade 0), a few ballooned cells (grade 1), and many ballooned cells (grade 2).

Liver lipid analysis

Undiluted serum samples and liver homogenates (50 mg/ml) were run in duplicate alongside a standard curve of glycerol (triglyceride assay), cholesterol (cholesterol assay) or palmitic acid (free fatty acid assay) according to the manufacturer’s instructions. Triglyceride Determination Kit and Cholesterol Determination Kit were obtained from Applygen. Free Fatty Acid Quantitation Kit was purchased from Sigma-Aldrich.

Real time-PCR (qPCR)

RNA from livers or isolated hepatocytes was purified using TRNzol Universal Reagent (Tiangen) and reversely transcribed (Tiangen). The primers used for qPCR are listed in Table S1. Actin was used as an internal control and the relative gene expression was calculated using the ΔΔCt quantification method.

Liver immune cell isolation

After mice were exsanguinated, 2 ml PBS containing collagenase (5 mg/ml, type IV, Sigma) were injected into the liver via the inferior vena cava. The liver was cut into small pieces and treated with collagenase (0.5 mg/ml) for 10 min at 37 °C. The liver pieces were mashed by using syringe plungers. The cells were passed through a 70 μm strainer (WHB scientific) and then centrifuged at 50 g for 3 min, 3 times, to pellet hepatocytes. The immune cells in the supernatant were pelleted (500 g for 15 min) and then isolated on a 40% Percoll step gradient (Cytiva). The cells were then stained and analyzed using flow cytometry (Shao et al., 2007).

Flow cytometry

Liver cells were collected by centrifugation and then incubated with Fc blocking antibody (purified anti-mouse CD16/32, BioLegend) on ice for 15 min. After the cells were stained with fluorescence-conjugated antibodies for 30 min on ice, they were washed and subjected to FACS (BD, FACSCelesta). The FACS data were analyzed using Flow Jo software (TreeStar,Inc). All antibodies used for flow cytometry were anti-mouse antigens. Anti-mouse antibodies used for flow cytometry were anti-CD45-BV785 (Clone 30-F11, BioLegend), anti-CD11b-FITC (Clone M1/70, BioLegend), anti-F4/80-BV421 (Clone BM8, BioLegend), anti-Ly6G-PerCP-Cy5.5 (Clone 1A8, BioLegend), anti-Ly6C-APC-Cy7 (Clone HK1.4, BioLegend), anti-MHC II-PE-Cy7 (Clone M5/114.15.2, BioLegend), anti-VSIG4-APC (Clone NLA14, eBioscience), and anti-Tim4-PE (Clone RMT4-54, BioLegend).

Hepatocyte isolation

Primary hepatocytes were isolated from adult mice using a two-step collagenase perfusion method (Charni-Natan and Goldstein, 2020). In brief, the peritoneal cavity was opened and the liver was perfused in situ via the portal vein at 37°C with 20 ml PBS followed by 20 ml DMEM containing 10 mg collagenase (type IV, Sigma). The liver was then removed and gently minced, and the released cells were dispersed in DMEM containing 5% fetal bovine serum (FBS) and 1% penicillin/streptomycin. The solution containing the mixed cells and debris was passed through a 100 µm cell strainer, and then centrifuged at 50 g for 3 min, twice. The hepatocytes were pelleted and then isolated on a 90% Percoll step gradient (Cytiva). Hepatocytes were then resuspended in a TRNzol Universal Reagent (Tiangen) to measure mRNA.

Western blot

Small pieces of liver were lysed with RIPA buffer (Biyotime) containing 1 mM PMSF (Biyotime) and a proteinase inhibitor mixture (Selleck). A commercial cytosol and nucleus Protein Extraction Kit (P0027, Beyotime) was used to separate cytosolic and nuclear proteins in the liver. The following antibodies were used for Western analysis: anti-SREBP1 (SC-17755; Santa cruz), anti-AOAH (HPA021666; Sigma-Aldrich), anti-Lamin B1 (12987-1-AP, Proteintech), anti-Phospho-mTOR (Ser2448) (5536, Cell Signaling Technology), anti-Phospho-p70 S6 Ribosomal Protein (Thr389) (9234, Cell Signaling Technology), anti-Phospho-AKT (Ser473) (4060, Cell Signaling Technology), anti-Phospho-mTOR (Ser2448) (5536, Cell Signaling Technology), anti-mTOR (2983, Cell Signaling Technology), anti-p70 S6 Ribosomal Protein (2217, Cell Signaling Technology), anti-AKT (4691, Cell Signaling Technology), and anti-α-Tublin (HRP-66031; Proteintech). Anti-mouse IgG (7076S; Cell Signaling Technology) and anti-Rabbit IgG (7074S, Cell Signaling Technology) were used as secondary antibodies. ECL substrate (Bio-Rad Diagnostic) was used to detect proteins in Western blotting and the blot bands were quantified by using Image J.

Measurement of TLR4-stimulating activities in mouse feces, liver and plasma

Fresh feces were collected and resuspended in endotoxin-free PBS (0.1 g/ml) and centrifuged at 800 g for 5 min. The supernatant was heated at 70 °C for 10 min. Mice were bled from the eye socket; 5 μl of 0.5 M EDTA was used as an anti-coagulant. Livers were homogenized in PBS, centrifuged and the supernatant was obtained. In some experiments, 200 µg of LPS was resuspended in 200 µl of PBS and then the suspension was slowly administered into the esophagus of mice using a gavage needle. Twenty-four h later, the livers were collected for analysis. All the samples were collected for TLR4-stimulating activity using a cell-based colorimetric endotoxin detection kit (HEK-Blue LPS Detection Kit2, InvivoGen). Diluted samples were added to human embryonic kidney (HEK-293) cells that expressed hTLR4 and a NF-κB-inducible secreted embryonic alkaline phosphatase reporter gene. After 18 h incubation, cell culture media were applied to QUANTI-Blue medium to measure alkaline phosphatase activity. A preparation of E. coli 055:B5 LPS, standardized to FDA-approved control standard endotoxin, which was included in the kit, was used to quantitate TLR4-stimulating activity. Plates were read at a wavelength of 620 nm (Tecan).

Gut permeability analysis

After mice were fasted for 18 h, they were orally gavaged with fluorescein isothiocyanate (FITC)-conjugated 4 kDa dextran (50 mg per 100 g body weight) (46944, Sigma-Aldrich). Four h after gavage, blood was collected from the facial vein and the serum was used for FITC fluorescence measurements (excitation, 490 nm; emission, 520 nm).

RNA-sequencing analysis

Total RNA was isolated using TRNzol from co-housed 6 - 8 weeks old Aoah^+/+ and Aoah^−/− mouse livers. The libraries were sequenced on an Ilumina Novaseq 6000 platform and 150 bp paired-end reads were generated. Differential expression analysis was performed using the DESeq2. Q value < 0.05 and foldchange > 1.5 was set as the threshold for significantly differential gene expression. Gene Set Enrichment Analysis (GSEA) was performed using GSEA software. RNA-seq analysis was conducted by Shanghai OE Biotech. Co., Ltd. China and the results were deposited at PRJNA1022016.

Statistical analysis

Data are presented as mean ± SEM. Differences between groups were analyzed using Mann-Whitney test. To compare kinetic difference, two-way ANOVA test was used. The statistical significance was set at P < 0.05. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Results

AOAH reduces hepatic lipid accumulation

To find out if AOAH prevents NAFLD, we fed co-housed Aoah^+/+and Aoah^−/− mice Normal Chow (NC) or a High Fat Diet (HFD), supplemented with fructose and glucose in the drinking water, for 28 weeks (Fig 1B) (Liu et al., 2018). Aoah^−/− mice fed either NC or HFD gained more weight than did respective Aoah^+/+control mice (Fig 1C). When they were fed either NC or HFD, the livers of Aoah^−/− mice were heavier than those of Aoah^+/+ control mice (Fig 1D). Histological examination and Oil Red O staining revealed that when they were fed NC or HFD, Aoah^−/− mouse livers accumulated more lipid droplets than did the livers of Aoah^+/+ mice (Fig 1E). When we scored NAFLD intensity based on steatosis, hepatocyte ballooning degeneration and inflammation, Aoah^−/− mice developed more severe NAFLD than did Aoah^+/+ mice whether the mice were fed NC or HFD (Sheka et al., 2020) (Fig 1E). When the mice were fed either NC or HFD, Aoah^−/− mouse livers contained more triacylglycerol (TAG) than did Aoah^+/+ mouse livers, while livers from both mouse strains contained a similar amount of total cholesterol (TCHO, Fig 1F). When the mice were fed the HFD, Aoah^−/− mice had higher serum levels of triacylglycerol, total cholesterol, low density lipoprotein (LDL) and free fatty acids than did Aoah^+/+ mice (Fig 1G). Collectively, these findings were evidence that AOAH reduced hepatic triacylglycerol accumulation when mice were fed either NC or HFD.

AOAH prevents hepatic inflammation and tissue injury when mice are fed HFD

Aoah^−/− mice fed HFD had significantly higher serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels than did control Aoah^+/+ mice, suggesting that Aoah^−/− mice experienced more severe liver inflammation and tissue damage (Fig 2A). To assess liver inflammation, we measured pro- and anti-inflammatory cytokine expression. When mice were fed the HFD, Aoah^−/− mouse livers produced more pro-inflammatory IL-6, IFN-γ and anti-inflammatory IL-10 mRNA than did Aoah^+/+ mouse livers, suggesting greater inflammation (Fig 2B). We also measured the expression of genes that are related to fibrosis and found that Aoah^−/− livers had more Timp1 (a pro-fibrosis gene) mRNA and less MMP2 (an anti-fibrosis gene) mRNA than did Aoah^+/+ mouse livers (Kisseleva and Brenner, 2021), indicating that Aoah^−/− mouse livers may be developing more severe fibrosis, although we did not detect fibrosis with Masson staining (Fig 2C). We analyzed the myeloid cells in the liver (Daemen et al., 2021) and found that when the mice were fed the HFD, Aoah^−/−mouse livers contained more neutrophils, monocytes and lipid-associated macrophages (hepatic LAMs) (Remmerie et al., 2020; Su, 2002) than did Aoah^+/+mouse livers (Fig 2D, E). Collectively, when the mice were fed HFD, the livers of Aoah^−/−mice developed significantly greater inflammatory responses and tissue damage.

Fig 2.

AOAH reduces hepatic LPS levels and the expression of fatty acid synthesis genes

We found previously that intestinal macrophages and DCs express AOAH (Janelsins et al., 2014; Qian et al., 2018). To confirm that hepatic cells express AOAH, we first consulted the single cell RNAseq analysis reported by Remmerie et al.(Remmerie et al., 2020), who found that AOAH is expressed in Kupffer cells, monocytes, monocyte-derived cells, NK (circulating NK) and ILC1 (tissue resident NK) cells (Remmerie et al., 2020) (Fig 3A). We used flow cytometry to sort Kupffer cells (CD45⁺NK1.1⁻ F4/80^hiCD11b^mid), monocytes (CD45⁺NK1.1⁻ F4/80^midCD11b^hi), NK cells (CD45⁺SSC^lo NK1.1⁺) (circulating NK and resident ILC1) and we purified the hepatocytes. Using qPCR analysis, we found that AOAH mRNA was present in Kupffer cells, monocytes and NK cells but not in hepatocytes, in keeping with previous findings (Shao et al., 2007) (Fig 3B). Western analysis confirmed that Aoah^+/+ mouse livers but not Aoah^−/− mouse livers had AOAH protein (Fig 3C). As gut-derived LPS can be transported via the portal vein into the liver (Han et al., 2021), we hypothesized that AOAH prevents hepatic inflammation and fat accumulation by inactivating LPS in the gut, portal vein, and liver. We found that Aoah^−/− mouse feces, liver and plasma had higher bioactive LPS levels when Aoah^−/− and Aoah^+/+ mice were fed either NC or HFD (Fig 3D). HFD increased gut permeability, but there was no permeability difference between Aoah^+/+ and Aoah^−/−mice that were fed either NC or HFD (Fig 3E). After the mice were fed HFD for 28 weeks, Aoah^−/− mouse livers increased fatty acid uptake gene Fabp3 mRNA and fatty acid synthesis gene Fasn mRNA, changes that may have contributed to lipid accumulation (Fig 3F).

Fig 3.

AOAH can regulate the expression of hepatic fatty acid metabolism genes

We found that AOAH reduced hepatic lipid accumulation when mice were about 8 months old if they were fed either NC or HFD. To investigate the mechanism, we analyzed the livers of young mice. We co-housed 3 - 4 - week - old Aoah^+/+ and Aoah^−/− mice for 3 - 4 more weeks before removing their livers for RNAseq analysis. The expression of several fatty acid biosynthesis genes, such as Acacb (acetyl-Coenzyme A carboxylase beta), Acss2 (Acetyl-CoA synthetase 2), Pcx (Pyruvate carboxylase), Acly (ATP citrate lyase), Fasn (Fatty acid synthase) and Scd1 (Stearoyl-Coenzyme A desaturase 1), was significantly increased in Aoah^−/− mouse livers (Fig 4A). When we used GSEA (Gene set enrichment analysis) to analyze deferentially expressed genes, we found that fatty acid biosynthesis pathway was enriched (Fig 4B). We then did qPCR and confirmed increases in mRNAs for FA biosynthesis genes described in Fig 4A as well as for Acaca (acetyl-Coenzyme A carboxylase alpha, encoding ACC1, the first and key enzyme on FA synthesis pathway) in Aoah^−/− mouse livers compared with Aoah^+/+mouse livers (Fig 4C). In addition, CD36 (fatty acid uptake (Chen et al., 2022)) mRNA levels increased in Aoah^−/− mouse livers (Fig 4D) while mRNAs for enzymes involved in fatty acid oxidation (Acot2 (Acyl-CoA thioesterase 2) and Ppar-α (peroxisome proliferator-activated receptor α)) (Bougarne et al., 2018; Moffat et al., 2014) decreased (Fig 4E). We found previously that LPS and other TLR agonists increase lipid accumulation in cultured macrophages by increasing expression of Acsl1 (Acyl-CoA synthetase long-chain family member 1) and Dgat2 (Diacylglycerol O-acyltransferase 2) and by reducing the production of Atgl (Adipose triglyceride lipase, Pnpla2) (Huang et al., 2014), yet the livers of Aoah^+/+ and Aoah^−/− mice had similar levels of Acsl1, Dgat2 and Atgl (Pnpla2) mRNA, suggesting that AOAH does not regulate hepatic triacylglycerol metabolism (Fig S1). These results suggest that AOAH reduces liver fat accumulation by diminishing the expression of fatty acid synthesis and uptake genes and increasing that of fatty acid oxidation genes.

Fig 4.

AOAH reduces hepatic SREBP1

As Acaca, Fasn and Scd1 are all target genes for sterol regulatory element-binding protein 1 (SREBP1), a transcription factor for fatty acid biosynthesis, we next analyzed SREBP1 expression in the liver (Horton et al., 2002; Shimano and Sato, 2017). There are two isoforms of SREBP1, SREBP1a and SREBP1c, and the liver predominantly expresses SREBP1c (Horton et al., 2002). The abundance of SREBP-1a and SREBP-1c mRNA increased in the livers of young Aoah^−/− mice (Fig 5A). SREBP1 is synthesized as a 125 KDa precursor (full length, flSREBP1) in the endoplasm reticulum, transferred to the Golgi apparatus, and cleaved sequentially by Site-1 protease and Site-2 protease to generate nuclear SREBP1 (nSREBP1, 68 KDa), which enters the nucleus and activates fatty acid biosynthesis gene transcription (Horton et al., 2002; Shimano and Sato, 2017). The livers mainly had the short form (68 KDa) nSREBP1 protein, which was significantly more abundant in Aoah^−/−mouse livers (Fig 5B). We separated liver cytosol and nuclei and found that the short form SREBP1 was mainly present in nuclei and that Aoah^−/−mouse liver nuclei contained significantly more SREBP1 than did Aoah^+/+ mouse liver nuclei (Fig 5C).

Fig 5.

We isolated hepatocytes from 6 - 8 - week - old co-housed Aoah^+/+and Aoah^−/− mice and found that Aoah^−/− mouse hepatocytes also expressed higher levels of Acly, Acaca, Acacb and Fasn mRNA and lower levels of Acot2 and Ppar-α mRNA, changes that may contribute to lipid accumulation as the mice grow older (Fig S2A - C). When we analyzed the transcript profiles from 6 – 8-week-old co-housed Aoah^+/+ and Aoah^−/− mouse livers, we noticed that the expression of serum amyloid A1 (SAA1), SAA2 and SAA3 was significantly higher in Aoah^−/−mouse livers (Fig S2D). Using qPCR, we found increased Saa1, Saa2, Saa3 and Irak-m mRNA in Aoah^−/− mouse hepatocytes, suggesting that excessive LPS in Aoah^−/− mouse livers induces both inflammatory and regulatory gene expression (Fig S2E). Thus, in young mice, even before hepatic lipid accumulation can be observed, hepatocytes in Aoah^−/− mouse liver have altered expression of genes that may promote lipid storage.

Excessive gut-derived LPS increases hepatic nSREBP1 and mTOR activation

To test whether excessive gut LPS increases liver LPS levels and promotes fatty acid synthesis gene expression, we orally gavaged Aoah^+/+ mice with LPS. We confirmed that orally gavaged (i.g.) LPS increased hepatic LPS levels (Fig 6A). Similar to Aoah^−/− mice, Aoah^+/+ mice that received gavaged LPS had increased levels of Pcx, Acaca, Acacb, Fasn, Scd1 and Cd36 mRNA in their livers (Fig 6B). LPS administered i.g. also increased nSREBP1 in the liver (Fig 6C). AKT-mTOR1-p70 S6-kinase (S6K) activation induces SREBP1c processing in hepatocytes (Jeon et al., 2023a; Owen et al., 2012; Yecies et al., 2011). Consistently, we found that Aoah^−/− mouse livers had elevated AKT-mTOR-S6K activation (Fig 6C). When we treated Aoah^+/+ mice i.g. LPS, hepatic AKT-mTOR-S6K activity increased (Fig 6C). These results suggest that excessive hepatic LPS derived from the Aoah^−/− mouse intestine induces mTOR activity, which increases nSREBP1 abundance and fatty acid biosynthesis gene expression in the liver; AOAH prevents hepatic lipid accumulation by inactivating gut-derived LPS.

Fig 6.

Discussion

The intestine and liver are connected by the portal vein, enabling the transport of gut commensal-derived molecules, including Gram-negative bacterial LPS, to the liver (Albillos et al., 2020; Leung et al., 2016). Much evidence suggests that gut-derived LPS induces hepatic inflammation and therefore exacerbates NAFLD, especially when dysbiosis and intestinal barrier dysfunction have occurred (An et al., 2022; Aron-Wisnewsky et al., 2020a; Aron-Wisnewsky et al., 2020b; Leung et al., 2016). As lipid accumulation in hepatocytes is considered to be the first hit, gut-derived LPS is usually thought to be the second hit in the pathogenesis of NAFLD, mainly inducing inflammation (An et al., 2022), yet the possibility that LPS also has direct effects on hepatocyte lipid metabolism has received little attention. In previous studies we found that when we co-housed Aoah^+/+ and Aoah^−/− mice for 3 or more weeks, they had similar microbiota (Qian et al., 2018), yet we found significantly more LPS in Aoah^−/− mouse feces and livers. Aoah^−/− mice accumulated more hepatic fat than did Aoah^+/+ mice when the mice were fed either normal chow or a high fat diet. Aoah^−/−mouse livers also expressed more inflammation-inducing and pro-fibrosis genes and had more liver damage when they were fed a HFD. Notably, when Aoah^−/− mice were young and had not developed NAFLD their livers already expressed significantly elevated levels of nSREBP1 and its target genes (Fig 7).

Fig 7.

Liver is an important tissue that converts carbohydrates into lipids (Jeon et al., 2023b). SREBP1c, the predominant isoform expressed in the liver, plays an important role in fatty acid synthesis (Shimano and Sato, 2017); SREBP1c mRNA was elevated in NAFLD patient livers (Kohjima et al., 2008) and its chronic activation contributed to NAFLD progression (Kawano and Cohen, 2013); SREBP1 has become a target for NAFLD treatment (Ju et al., 2020; Jump et al., 2013). Intriguingly, SREBP1 levels increased in the livers of 6 - 8 weeks old (i.e., young) Aoah^−/− mice before NAFLD developed; the expression of many SREBP1 target genes, such as those involved in fatty acid biosynthesis (Acly, Acaca, Acacb, Fasn, Scd1, Acss2) also increased (Fig 7). We found that orally gavaged LPS increased hepatic LPS, nSREBP1 abundance, and the expression of nSREBP1’s target genes, suggesting that gut-derived LPS reaches the liver and promotes fatty acid synthesis. Thus, our data suggest that when AOAH is lacking, excessive gut-derived LPS stimulates SREBP1 activation to promote de novo lipogenesis in the liver, contributing to more severe NAFLD.

SREBP1a and SREBP1c mRNA both increased in Aoah^−/− mouse livers, suggesting that regulation occurs at the transcription level or because the Srebf1 gene (encoding SREBP1) promoter contains SREs, increased nSREBP1 induced a feed-forward transcription of SREBP1 (DeBose-Boyd and Ye, 2018). Intriguingly, AKT-mTOR-S6K activity increased in Aoah^−/− mouse livers, which may contribute to increased SREBP1 translocation and processing (DeBose-Boyd and Ye, 2018; Jeon et al., 2023a; Owen et al., 2012). Whether LPS directly stimulates hepatocytes to induce SREBP activation or whether immune cells such as macrophages are involved await further investigation (Barreby et al., 2022; Kazankov et al., 2019; Ye et al., 2012).

In addition to fatty acid biosynthesis gene expression, the expression of CD36, which is involved in free fatty acid uptake (Chen et al., 2022), also increased in the livers of young Aoah^−/− mice while expression of fatty acid oxidation-related genes Acot 2 and Ppar-α decreased (Bougarne et al., 2018; Moffat et al., 2014). In addition to taking up fatty acids, CD36 interacts with INSIG2, a negative regulator of SREBP1, promoting the translocation of SREBP1 from ER to Golgi for cleavage and activation (Zeng et al., 2022). In keeping with our findings, Kim et al., found that LPS suppressed PPAR-α expression via ERK activation and HNF4 phosphorylation in primary mouse hepatocyte culture (Kim et al., 2024). In addition to promoting hepatic FA oxidation, PPAR-α is a transactivating factor that enhances Insig2 expression in hepatocytes, preventing SREBP activation (Lee et al., 2017). These results suggest that hepatic LPS stimulation promotes lipid accumulation via many mechanisms. In a previous study, Huang et al., found that LPS and other TLR agonists promoted fat retention in murine macrophages by increasing triacylglycerol synthesis and reducing lipolysis, yet fatty acid synthesis gene abundance did not change (Huang et al., 2014). In contrast, we found that Aoah^−/− and Aoah^+/+mouse livers had similar levels of Acsl1, Dgat2 and Atgl mRNAs, suggesting that in response to LPS or other PAMPs, hepatocytes and macrophages may accumulate fat via different mechanisms. Notably, we found previously that after LPS stimulation of macrophages in vitro, the culture medium became acidic due to aerobic glycolysis (the Warburg effect). The acidic environment contributed more to increasing fat accumulation than did LPS stimulation (Lu et al., 2014). A sensitive pH indicator is needed to find out if the blood and/or extracellular fluid in the liver become acidic due to excessive LPS stimulation and if the acidity promotes hepatic fat accumulation.

In addition to LPS, AOAH also deacylates and inactivates oxidized phospholipids and lysophospholipids (Zou et al., 2021), DAMP (Danger-Associated Molecular Pattern) molecules that are induced by inflammation and known to contribute to NAFLD (Sun et al., 2020; Wang et al., 2022). By inactivating gut-derived PAMPs and DAMPs, AOAH may decrease hepatic fat accumulation and prevent NAFLD. Increasing AOAH abundance may be a useful way to prevent and/or reduce this common disease.

Data availability

All data are contained within the manuscript and the supplementary files. The RNA-seq data were deposited at PRJNA1022016.

Acknowledgements

We thank She Chen and Yanyong Xu for very helpful discussion.

Funding sources

This work was supported by the National Natural Science Foundation of China (32370979, 32170929, 91742104, 31770993 and 31570910 to M. Lu, 32260189, 32300772 to W. Jiang, 82301980 to B. Zou), the Shanghai Committee of Science and Technology (21ZR1405400 to M. Lu), Guizhou Provincial Science and Technology Projects (ZK2024 - 240 to W. Jiang), and National Institutes of Health (AI18188 and AI44642 to R. Munford).

The authors declare that they have no conflicts of interest with the contents of this article.

Abbreviations

• AOAH: Acyloxyacyl Hydrolase

• LPS: Lipopolysaccharide

• NAFLD: Non-alcoholic Fatty Liver Disease

• HFD: High Fat Diet

• NC: Normal Chow

• PAMP: Pathogen-Associated Molecular Pattern

• DAMP: Danger-Associated Molecular Pattern

• FA: Fatty Acid

• flSREBP1: full-length Sterol Regulatory Element-Binding Protein 1

• nSREBP1: nuclear SREBP1

• LAM: Lipid-Associated Macrophages

• KCs: Kupffer cells

• Mo-KCs: Monocyte-derived Kupffer Cells.