Abstract

   Nonalcoholic fatty liver disease (NAFLD) has become the most common
   chronic liver disease, which progression is tightly regulated by
   transcription factors (TFs), nuclear receptors, and cellular enzymes.
   In this study, a label-free quantitative proteomic approach was used to
   determine the effect of the high-fat diet on the proteomics profile of
   liver tissue and to identify novel NAFLD related TFs. Mice were fed
   with HFD for 16 weeks to establish a NAFLD mouse model. Mice fed with
   normal chow diet were taken as controls. Liver samples were collected
   from each group for proteomics analysis. A total of 2298 proteins were
   quantified, among which 106 proteins were downregulated, while 256
   proteins were upregulated in HFD-fed mice compared with the controls
   with fold change more than 1.5 and p value less than 0.05.
   Bioinformatic analysis revealed that metabolic-related functions and
   pathways were most significantly enriched. A subgroup of 11 TFs were
   observed to share interactions with metabolic-related enzymes and
   kinases by protein-protein interaction analysis. Among them, 7 TFs were
   selected for verification, and 3 TFs were finally validated, including
   Rbbp4, Tcea1, and ILF2. Downregulating each of the 3 TFs could
   significantly promote lipid accumulation in AML12 hepatocytes, by
   regulating the expression of fatty acid synthesis- or
   β-oxidation-related genes. In contrast, overexpression of Tcea1, Rbbp4,
   and ILF2, respectively, could ameliorate hepatocyte steatosis. These
   findings propose novel lipid metabolism related TFs, which might have
   potential roles in preventing NAFLD.

   Keywords: quantitative proteomics, NAFLD, transcription factor, lipid
   metabolism

Introduction

   Nonalcoholic fatty liver disease (NAFLD), a spectrum of liver disease
   characterized by fat accumulation, affects more than 25% of the global
   population [45]^1. Hepatic steatosis has been recognized as a critical
   trigger for the more advanced disease, nonalcoholic steatohepatitis
   (NASH), which increases the incidence of liver cirrhosis and
   hepatocellular carcinoma [46]^2^-[47]^4. Despite considerable studies,
   the molecular mechanisms of NAFLD remain incompletely understood.
   Currently, no FDA-approved pharmacological strategy has been
   established for NAFLD [48]^5. Thus, further studies are warranted to
   investigate novel therapeutic targets.

   Lipids produced by de novo lipogenesis have been considered to be the
   fundamental step in the development of NAFLD and are caused by the
   imbalance between triglyceride (TG) synthesis and degradation [49]^6.
   The dysregulation of key enzymes involved in fatty acid synthesis or
   oxidation, including ATP citrate lyase (ACLY), fatty acid synthase
   (FASN), acetyl-CoA (ACC), and carnitine palmitoyltransferase-1α
   (Cpt1α), leading to direct TG accumulation [50]^7. Previous studies
   have revealed several protein subtypes that can regulate the expression
   or degradation of these enzymes, such as transcription factors
   (TFs)[51]^8, RNA binding proteins [52]^9, and ubiquitin ligases
   [53]^10. Among TFs, sterol regulatory element-binding transcription
   factor 1 (SREBP1) and peroxisome proliferator-activated receptors
   (PPARs), which have been well-studied, regulate the cellular level of
   lipogenic and fat-consuming enzymes by directly binding to their gene
   encoding elements [54]^11^, [55]^12. Agents targeting SREBP1 and PPARs
   are undergoing clinical trials for NAFLD [56]^13. Considering the
   crucial role of TFs, the development of new TFs involved in regulating
   NAFLD progression has attracted the attention of investigators.

   In recent years, proteomics analysis based on high-throughput mass
   spectrometry (MS) has been used to study different aspects of NAFLD.
   Plasma proteome profiling studies have identified several potential
   biomarkers related to NAFLD stages [57]^14. Quantitative proteomics has
   also been applied to uncover differential expressed proteins (DEPs) in
   NAFLD using clinical liver tissues or mouse models [58]^15^, [59]^16.
   Although hundreds of DEPs have been identified, the precise link
   between DEPs and NAFLD settings is rarely verified. Furthermore, there
   are limited proteomics studies on NAFLD-related TFs screening.

   In the present study, a label-free quantitative proteomics strategy was
   used to identify hepatic DEPs related to NAFLD, followed by the
   construction of protein interaction network and gene ontology
   enrichment through Ingenuity Pathway Analysis (IPA). Among hundreds of
   DEPs, TFs were screened out for in vivo and in vitro verification. The
   potential role of validated TFs including transcription elongation
   factor A protein 1 (Tcea1), histone-binding protein RBBP4 (Rbbp4), and
   interleukin enhancer-binding factor 2 (ILF2) was further investigated
   by gene knockdown and overexpression strategies.

Materials and methods

Animal experiments

   6-week-old male C57BL/6 mice were obtained from the center of
   experimental animals of Shanghai University of Traditional Chinese
   Medicine and were housed in specific pathogen free conditions on a 12 h
   light/dark cycle with free access to water and food. The mice were set
   on the following diets: HFD mice fed with high-fat diet (Dyets, 112252,
   USA, containing 60% kcal fat, 20% kcal carbohydrate, and 20% protein)
   for 16 weeks to establish NAFLD mouse model; NCD mice were fed with
   normal chow diet for 16 weeks. Body weights of mice from both groups
   were recorded every week. The mice were sacrificed at the age of 22
   weeks, and all procedures were conducted in compliance with
   institutional guidelines and protocols. Liver samples after perfusion
   were collected and stored in liquid nitrogen for histological
   evaluation and proteomics analysis. All animal experiments were
   performed with approval from the Institutional Animal Care and Use
   Committee of Shuguang Hospital, Shanghai Chinese traditional medicine
   University.

Liver histological evaluation

   Liver tissues were fixed overnight in 4% PFA in PBS at room
   temperature, embedded in paraffin, and then stained with hematoxylin
   and eosin (H&E) for assessment of liver histology. Liver steatosis was
   calculated by hepatocyte steatosis and ballooning, as described
   previously [60]^17. Oil Red O staining was performed on the frozen
   sections to evaluate lipid accumulation in the liver lobes using
   standard methods [61]^18.

Blood parameters

   The concentrations of the total cholesterol (TC), triglyceride (TG),
   low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein
   cholesterol (HDL-C), hepatic enzymes alanine amino transferase (ALT),
   aspartate amino transferase (AST), and alkaline phosphatase (ALP) were
   measured using kits according to manufacturer's instructions and
   analyzed by using a spectrophotometer (SpectraMax iD5).

Clinical nonalcoholic fatty liver and control liver samples

   Clinical NAFLD and control liver tissues were obtained from liver
   biopsy of NAFLD patients. The severity of hepatic steatosis was
   diagnosed by abdominal ultrasound and was verified according to liver
   histology. These patients were admitted to the ShuGuang Hospital
   Affiliated to Shanghai University of Chinese Traditional Medicine. The
   study protocol was approved by the Ethics Committee of ShuGuang
   Hospital Affiliated to Shanghai University of Chinese Traditional
   Medicine (IRB 2020-878-87).

Western blot

   Equal amounts of proteins were separated on a 10% pre-cast SDS-PAGE
   (Genscript, Nanjing, China), and then transferred to polyvinylidene
   difluoride (PVDF) membranes (PALL Life Sciences, USA). The membranes
   were blocked with 5% milk in TBST (TBS with 0.1% Tween-20) for 1 h at
   room temperature, and then incubated overnight at 4 ℃ with primary
   antibodies: anti-Rbbp4 antibody (Proteintech, 20364-1-AP), anti-Tcea1
   antibody (Proteintech, 17825-1-AP), anti-ILF2 antibody (Proteintech,
   14714-1-AP), anti-NPM1 antibody (Proteintech, 10306-1-AP), anti-Acc1
   antibody (abcam, ab45174), anti-Fasn antibody (abcam, ab128870),
   anti-Srebp1c antibody (abcam, ab28481), anti-Acox1 antibody (abcam,
   ab184032), anti-Cpt1α antibody (abcam, ab128568), anti-Flag antibody
   (Proteintech, 20543-1-AP), anti-GAPDH antibody (abcam, ab8245),
   anti-β-actin antibody (Proteintech, 66009-1-Ig). The membranes were
   washed and incubated with horseradish peroxidase-conjugated secondary
   antibodies for 1 h. Protein bands were visualized using enhanced
   chemiluminescence kit (UUbio, Suzhou, China) with a Tanon 5200
   multi-imaging system (Tanon Technology Co., Ltd, Shanghai, China).

Sample preparation for label-free proteomics

   The total proteins were extracted from perfused liver tissues by
   homogenizing in RIPA lysis buffer (Thermo scientific, Waltham, USA).
   After centrifugation at 12,000g for 10 min at 4 °C, the supernatant was
   collected and the protein concentrations were determined using the BCA
   kit (UUbio, Suzhou, China), according to the manufacturer`s
   instructions. 100 μg of protein samples were reduced using 20 mM
   dithiothreitol (DTT, Sigma-Aldrich) at 60 °C for 60 min, and 100 mM
   iodoacetamide was used for alkylated at room temperature for 40 min.
   Alkylated protein was precipitated overnight at -20℃ by adding 5 times
   the volumes of acetone. The protein precipitates were resuspended in
   100 mm ammonium bicarbonate, followed by digested overnight with
   trypsin at a trypsin-to-protein mass ratio of 1:50. Finally, the
   peptide was desalted using the MnoSpin C18 column (GL Sciences, Japan),
   vacuum-dried, reconstituted in 0.1% formic acid.

LC-MS/MS analysis and bioinformatics analysis

   The peptides were analyzed using LC-MS/MS on Q Exactive™ HF-X (Thermo
   Fisher Scientific) coupled to an EASY-nLC 1000 UPLC system (Thermo
   Fisher Scientific). MS1 was set at a resolution of 70,000 and then
   selected for MS/MS with an NCE setting of 28. The MS2 was set at a
   resolution of 17,500. Quantitative data analysis was performed by
   MaxQuant (v1.6.1.2) using raw data files of MS/MS spectra searched
   against the UniProtKB mouse database (release 2021_04, 17090 entries)
   concatenated with reverse decoy database. Trypsin/P was used as the
   cleavage enzyme allowing up to two missing cleavages. The mass
   tolerance for precursor ions in First search and in Main search was set
   as 20 ppm and 6 ppm, respectively, and for fragment ions was set to
   0.02 Da. Carbamidomethylation of cysteines was defined as fixed
   modification, while protein N-terminal acetylation and methionine
   oxidation were defined as variable modifications for database searches.
   Maximum FDR, calculated by employing a reverse database strategy, was
   set to 1% for peptides and proteins. Label-free quantification was
   performed by Intensity-based absolute quantification (iBAQ). Protein
   expression from each sample was compared using normalized iBAQ value,
   based on at least two unique peptides, and data processing was aligned.
   The 1.5 folds cut-off was set up for the screening of differentially
   expressed proteins.

   Gene ontology (GO) annotation, KEGG pathways were performed using
   Omicsolution online tool ([62]https://omicsolution.org/wkomics/main/).
   P-value < 0.05 was considered to be statistically significant.
   Quantitative data of differentially expressed proteins was submitted to
   Ingenuity Pathway Analysis (IPA, QIAGEN, Redwood City,
   [63]http://www.ingenuity.com/) for analysis of protein-protein
   interaction.

Immunohistochemistry analysis

   Formalin-fixed liver tissue samples were analyzed by
   immunohistochemistry. Briefly, mouse liver sections were
   deparaffinized, repaired with 0.01 M sodium citrate-hydrochloric acid
   buffer solution, and incubated with primary antibodies, including
   Rbbp4, Tcea1, and ILF2 (1:100 dilution) at 4 °C overnight. The
   incubated liver sections were treated with peroxidase-conjugated
   secondary antibody (111-035-003, Jackson ImmunoResearch, USA) at room
   temperature, stained with diaminobenzidine for 3-5 min, and
   counterstained with hematoxylin. The liver sections were then observed
   under an optical microscope.

siRNA transfection and knockdown

   AML12 and HepG2 cells were transfected using RNAiMAX transfection kits
   (Thermo scientific, USA) with negative control siRNA or siRNA
   oligonucleotides specific to Rbbp4, Tcea1, and ILF2 according to
   manufacturer's instructions. After 48 h of transfection, cells were
   collected and used for downstream applications. The siRNA sequences
   were listed in [64]Table S1.

Plasmid construction and transfection

   Plasmids encoding full-length Rbbp4, Tcea1, and ILF2 were constructed
   by cloning the indicated coding regions of mouse Rbbp4 cDNA, Tcea1
   cDNA, or ILF2 cDNA into the pEnCMV-Flag vector. AML12 cells were
   transfected using lipofectamine 3000 transfection kits (Thermo
   scientific, USA) with negative control plasmid or overexpression
   plasmid specific to Rbbp4, Tcea1, and ILF2 according to manufacturer's
   instructions. After 48 h of transfection, cells were collected and used
   for downstream applications.

Statistical analysis

   Data were expressed as mean ± standard deviation. Two tailed student's
   t test was used to analyze the differences between two groups. P value
   less than 0.05 was considered statistically significant. Statistical
   analysis was performed using Graph Pad Prism 8 (version 8.2.1).

Results

Construction and evaluation of steatohepatitis by high-fat diet in mice

   To induce hepatic steatosis, we fed mice a high-fat diet (HFD) for 16
   weeks, whereas the control group maintained a normal chow diet (NCD).
   The body weight of the mice in the HFD group increased significantly
   compared with that in the NCD group, which was nearly two times that of
   the NCD group at the 16th week (Figure [65]1A). Firstly, we performed
   glucose tolerance test (GTT) and insulin tolerance test (ITT) on HFD
   and NCD-fed mice, HFD mice displayed enhanced glucose intolerance and
   insulin resistance compared to NCD mice ([66]Figure S1). All mice were
   sacrificed 16 weeks after HFD. Liver weight and liver/body ratios of
   the HFD group were significantly higher than those of the NCD group
   (1.65 ± 0.09 g vs. 3.14 ± 0.65 g, n = 7, p < 0.001; 5.6 ± 0.28% vs. 6.7
   ± 0.43%, n = 7, p < 0.05), and the fatty liver had almost 100% increase
   in hepatic triglyceride (TG) (Figure [67]1B-[68]1E). What's more, the
   HFD group also displayed elevated serum levels of alanine transaminase
   (ALT), aspartate transaminase (AST), total cholesterol (TC), TG, LDL-C,
   and HDL-C compared with the NCD group (Figure [69]1F-[70]1K). Liver HE
   staining results showed steatosis of hepatocytes, slight inflammatory
   cell infiltration, and ballooning degeneration in HFD-fed mice (Figure
   [71]1L). Excessive fat droplets accumulation in the hepatocytes of
   HFD-fed mice was then confirmed by the Oil red O staining (Figure
   [72]1M).

Figure 1.

   [73]Figure 1
   [74]Open in a new tab

   Mice model of high-fat diet induced nonalcoholic fatty liver. (A) Body
   weight gain of HFD- and NCD-fed mice (n=7 mice per group). (B) Gross
   morphology of livers from HFD- and NCD-fed mice. (C-K) Liver weight,
   liver/body weight ratio, liver TG, serum ALT, serum AST, serum TC,
   serum TG, serum LDL-C, and serum HDL-C of the HFD and NCD groups (n=7
   mice per group). (L) H&E staining of livers from HFD- and NCD-fed mice.
   Magnification: 100x and 400x, scale bar: 50 μm. (M) Oil red O staining
   of livers from HFD- and NCD-fed mice. Scale bar: 50 μm. All data
   represent the mean ± SEM. Student's t test,* p < 0.05, ** p < 0.01, ***
   p < 0.001.

Identification and quantification of differentially expressed proteins (DEPs)
related to steatohepatitis

   We then analyzed the total proteome alteration of these mice by
   label-free quantification based on LC-MS/MS. Equal amounts of liver
   proteins from the NCD and HFD groups were reduced, alkylated and
   digested with trypsin, with 5 biological replicates per group.
   Principle component analysis (PCA) of protein expression profiles of
   all samples showed that the global patterns of the proteome differed
   between the NCD and HFD groups (Figure [75]2A). However, individual
   differences were noted within each population. A total of 2298 proteins
   were identified with at least two unique peptides at a false discovery
   rate of < 1%. After quantification, 362 proteins were screened out as
   DEPs, with fold change more than 1.5 and p value less than 0.05, among
   which 106 DEPs were downregulated, while 256 DEPs were upregulated in
   HFD-fed mice compared with the controls ([76]Table S2, Figure [77]2B),
   and the heatmap of DEPs was shown in Figure [78]2C.

Figure 2.

   [79]Figure 2
   [80]Open in a new tab

   Label-free quantification proteomics of livers from HFD- and NCD-fed
   mice. (A) Principle component analysis (PCA) for proteome of all liver
   samples. (B) Volcano plot of DEPs. Red and blue dots represent
   upregulated and downregulated proteins, respectively. (C) Heatmap of
   DEPs. (D) Biological process enrichment of DEPs. (E) Cell component
   enrichment of DEPs. (F) KEGG pathway enrichment of DEPs. (G)
   DEPs-upstream regulators-biological processes interaction network
   constructed by IPA. Blue represents inhibition of protein/biological
   process, Orange represents activation of protein/biological process.

Gene ontology and KEGG pathway enrichment analysis of DEPs

   Further bioinformatics analysis was performed on the 362 DEPs
   identified above. Among the altered proteins, 62.7% were annotated by
   Gene Ontology (GO) as involved in metabolic processes ([81]Figure S2A).
   The highly enriched proteins were associated with lipid metabolic
   process, fatty acid metabolic process, organic acid metabolic process,
   and oxidation-reduction process (Figure [82]2D). As expected, DEPs were
   significantly enriched in mitochondrion, peroxisome, and microbody,
   where organelles metabolic catalysis took place (Figure [83]2E). KEGG
   pathway enrichment analysis demonstrated that metabolism-related
   pathways including PPAR signaling, AMPK signaling, insulin signaling,
   and fatty acid metabolic signaling were mainly enriched pathways
   (Figure [84]2F). Based on the DEPs, we constructed DEPs-upstream
   regulators-biological processes network by ingenious pathway analysis
   (IPA), showing that lipid oversupply could activate PPAR-α, PPAR-γ,
   SREBPs, ESRR, and INSR in hepatocytes, resulting in activation of fatty
   acid oxidation and lipid droplets formation (Figure [85]2G).

   In the molecular function category, 69% of DEPs were proteins with
   catalytic activity (Figure [86]3A), among which Acly, Acox1, Acc1,
   Hadha, and Acadm were top 5 high-weight catalytic enzymes in weight
   distribution analysis results ([87]Figure S2B). Other highly enriched
   molecular functions included ion binding, ATP binding, and nucleic acid
   binding (Figure [88]3A), indicating that metabolism disturbance of
   hepatocytes was the major trigger of NAFLD.

Figure 3.

   [89]Figure 3
   [90]Open in a new tab

   Bioinformatics analysis of DEPs revealed metabolic dysfunction related
   TFs. (A)Molecular function enrichment of DEPs. (B) Heatmap of
   differential abundance TFs. (C) Protein-protein interaction network
   contructed by IPA. Green represents downregulated DEPs, red represents
   upregulated DEPs, gray represents upstream reguators of DEPs.

Screening and validation of candidate proteins in vivo and in vitro

   We noticed that 56 DEPs had nucleic acid binding capacity, including
   transcription factors (TFs) and RNA binding proteins. Considering the
   crucial role of TFs in regulating gene expression and signaling
   transduction, 11 differential expressed TFs were selectited for further
   analysis (Figure [91]3B and Table [92]1). PPI network was then
   constructed using IPA, which revealed the interactions of these 11 TFs
   with key proteins in lipid metabolism, such as SREBP1, AMPK, PPARa,
   FASN, and Acc1 (Figure [93]3C). The result suggests that these 11 TFs
   may regulate genes implicated in fatty acid metabolism. Among the 11
   TFs, Ruvbl2, Pc, Sra1, and Preb have been reported to regulate lipid
   metabolism [94]^19^-[95]^22, we therefore chose the remaining 7 TFs for
   further validation.

Table 1.

   Differentially expressed transcription factors quantified by LC-MS/MS.
   Protein IDs Gene names Protein names Unique peptides Sequence
   coverage [%] Mol. weight [kDa] Fold change
   (HFD/NC) p value
   [96]P10711 Tcea1 Transcription elongation factor A protein 1 2 9 33.88
   0.01 3.77E-02
   [97]Q9WTM5 Ruvbl2 RuvB-like 2 6 14.5 51.112 0.34 1.44E-02
   [98]Q61937 Npm1 Nucleophosmin 9 31.8 32.56 0.65 1.62E-02
   [99]Q02819 Nucb1 Nucleobindin-1 12 32.9 53.408 1.51 2.06E-02
   [100]Q05920 Pc Pyruvate carboxylase 54 48.8 129.68 1.53 9.08E-05
   [101]P42669 Pura Transcriptional activator protein Pur-alpha 5 18.7
   34.883 2.94 2.63E-02
   [102]Q80VJ2 Sra1 Steroid receptor RNA activator 1 2 9.1 25.557 3.50
   1.40E-02
   [103]Q9WUQ2 Preb Prolactin regulatory element-binding protein 7 22.1
   45.437 3.72 2.41E-02
   [104]Q9CR86 Carhsp1 Calcium-regulated heat stable protein 1 4 26.4
   16.062 11.78 1.95E-02
   [105]Q60972 Rbbp4 Histone-binding protein RBBP4 3 19.1 47.655 0.47
   1.86E-02
   [106]Q9CXY6 Ilf2 Interleukin enhancer-binding factor 2 4 12.1 43.062
   0.38 4.02E-03
   [107]Open in a new tab

   We first verified the expression of 7 TFs at transcription level in
   liver tissues, which were significantly downregulated in the HFD group
   compared to the NCD group (Figure [108]4A). The same result was also
   observed in AML12 hepatocytes treated with palmitic acid/oleic acid
   (PAOA, 0.25 mM) for 24 h (Figure [109]4B). 4 TFs whose transcription
   level were consistent with the mass spectrometry result, including
   ILF2, Npm1, Rbbp4, and Tcea1, were chosen for validation in primary
   hepatocytes. PAOA treatment caused a significant decrease in the
   transcriptional expression of these 4 TFs in primary hepatocytes
   (Figure [110]4C). We then confirmed the changes in the protein levels
   of these 4 TFs in liver tissues by western blot, which revealed the
   downregulation of ILF2, Rbbp4, and Tcea1 in HFD-feeding mice (Figure
   [111]4E and [112]4F). In addition, immunohistochemical staining showed
   the same result, and expression of ILF2, Rbbp4, and Tcea1 were mainly
   observed in the nucleus of hepatocytes, instead of immune cells, which
   suggested the potential role of these 3 TFs in the regulation of NAFLD
   progression (Figure [113]4D and [114]4G). To verify these findings, we
   further evaluated the 3 TFs expression in clinical liver samples. The
   expression of Rbbp4, Tcea1, and ILF2 were significantly downregulated
   in significant hepatic-steatosis NAFLD patients compared to
   non-significant hepatic-steatosis controls (Figure [115]4H and
   [116]4I). Due to individual differences, decreased trend was observed
   in ILF2, although difference did not reach significance (Figure
   [117]4I).

Figure 4.

   [118]Figure 4
   [119]Open in a new tab

   Verification of candidate TFs in vivo and in vitro. (A) Verification of
   candidate genes mRNA expression in liver tissues of HFD- and NCD-fed
   mice (n=5). (B) Verification of candidate genes mRNA expression in
   AML12 cells after PAOA treatment for 24 h (n=3). (C) Verification of
   candidate genes mRNA expression in primary hepatocytes after PAOA
   treatment for 24 h (n=3). (D) Representative images of
   immunohistochemistry analysis of Tcea1, Rbbp4, ILF2 expression and
   localization in liver sections from HFD- and NCD-fed mice.
   Magnification: 200x and 400x, scale bar: 50 μm. (E) Western blot
   analysis of ILF2, Rbbp4, Tcea1, and Npm1 expression in liver tissues
   from HFD- and NCD-fed mice. (F) Quantification of western blot analysis
   of ILF2, Rbbp4, Tcea1, and Npm1 protein expression (n=4). (G)
   Quantification of immunohistochemistry analysis of Tcea1, Rbbp4, and
   ILF2 expression (n=3). (H) Western blot analysis of ILF2, Rbbp4, and
   Tcea1 expression in liver tissues from NAFLD patients and control. (I)
   Quantification of western blot analysis of ILF2, Rbbp4, and
   Tcea1protein expression (n=5). All data represent the mean ± SEM.
   Student's t test, * P < 0.05, ** P < 0.01, *** P < 0.001.

Rbbp4, Tcea1, and ILF2 regulate lipid metabolism related genes expression

   To investigate the role of Rbbp4, Tcea1, and ILF2 in cellular lipid
   metabolism, these 3 TFs were knocked down in AML12 and HepG2
   hepatocytes by siRNA silencing strategy. 3 siRNA oligos targeting
   different mRNA regions were constructed for each gene. The siRNA oligo
   achieved more than 90% knockdown efficacy of mRNA expression compared
   to the siRNA control ([120]Figure S3). The siRNA with the highest
   knockdown efficacy was used for further gene functional study in AML12
   and HepG2 cells. Based on the potential interactions between Rbbp4,
   Tcea1, and ILF2 with key regulators in lipid metabolism (Figure
   [121]3C), we first evaluated the effects of Rbbp4, Tcea1, and ILF2
   knockdown on the expression of key genes involved in lipid metabolism.
   Fatty acid synthesis genes Acc1, Fasn, Srebp1, and β-oxidation related
   genes Cpt1α, Acox1 were detected by qPCR and western blot analysis.
   Total RNA and proteins were extracted 48 h after siRNA transfection in
   AML12 and HepG2 cells. As shown in Figure [122]5A and [123]5B, the mRNA
   expression of Acc1 in Tcea1 knockdown cells was siginificantly
   incresed, and ILF2 downregulation significantly promoted Acc1, Fasn,
   and Srebp1 expression. In addition, the expression of Cpt1α and Acox1
   expression was significantly decreased when Rbbp4 was downregulated.
   Protein expression detected by western blot in AML12 cells showed the
   same result (Figure [124]5C-[125]5H). The results demonstrated that
   Tcea1 and ILF2 knockdown promoted fatty acid synthesis, and Rbbp4
   knockdown inhibited fatty acid β-oxidation, which may lead to lipid
   accumulation in hepatocytes.

Figure 5.

   [126]Figure 5
   [127]Open in a new tab

   Tcea1, Rbbp4, and ILF2 regulated fatty acid synthesis or β-oxidation
   related gene expression. (A-B) RT-qPCR analysis of fatty acid synthesis
   related genes Acc1, Fasn, and Srebp1, and β-oxidation related genes
   Cpt1α, Acox1 expression in AML12 and HepG2 cells after Tcea1, Rbbp4,
   and ILF2 knockdown, respectively, compared with the siNC group. (C)
   Western blot analysis of Acc1, Fasn, Srebp1, Cpt1α, and Acox1
   expression in AML12 cells after Tcea1, Rbbp4, and ILF2 knockdown,
   respectively. (D-H) Quantification of relative protein expression
   level. n=3 biological replicates. All data represent the mean ± SEM.
   Student's t test, * P < 0.05, ** P < 0.01, *** P < 0.001.

Rbbp4, Tcea1, and ILF2 deficiency promotes lipid accumulation in hepatocytes

   AML12 cells were transfected with siRNA for 48h and then treated with
   PAOA for another 24 h, and the control group was treated with
   corresponding amount of BSA. Protein expressions of Rbbp4, Tcea1, and
   ILF2 in AML12 cells were detected by western blot (Figure [128]6B).
   Through Oil red O staining and cellular TG concentration measurement,
   we observed that compared with the control, Rbbp4, Tcea1, and ILF2
   knockdown significantly increased cellular lipid accumulation (Figure
   [129]6A ,6C and 6D). In addition, Rbbp4, Tcea1, and ILF2 deficiency
   significantly promoted basal levels of lipid contents in HepG2 cells
   ([130]Figure S4).

Figure 6.

   [131]Figure 6
   [132]Open in a new tab

   Tcea1, Rbbp4, and ILF2 deficiency promoted lipid accumulation of
   hepatocytes, respectively. (A) Representative images of Oil red O (ORO)
   staining of AML12 cells treated with or without PAOA for 24 h after
   Tcea1, Rbbp4, and ILF2 knockdown, respectively. Scale bar: 100 μm. (B)
   Western blot analysis of Tcea1, Rbbp4, and ILF2 expression in AML12
   cells used for ORO staining. (C) Quantification of relative ORO
   positive area. (D) Measurement of cellular TG concentration in gene
   knockdown AML12 cells after PAOA treatment for 24 h. (E-I) RT-qPCR
   analysis of Acc1, Fasn, Srebp1, Cpt1α, and Acox1 gene expression in
   AML12 cells treated with or without PAOA after Tcea1, Rbbp4, and ILF2
   knockdown, respectively. siNC represents cells transfected with
   negative control siRNA. n=3 biological replicates. All data represent
   the mean ± SEM. Student's t test, ** P < 0.01, *** P < 0.001 compared
   with the siNC group; ## P < 0.01, ### P < 0.001 compared with the
   siNC+PAOA group.

   We further detected the expression of above-mentioned genes in Tcea1,
   ILF2, and Rbbp4 knockdown cells under fatty acid excess conditions.
   AML12 cells with or without gene knockdown were treated with PAOA for
   24 h. Cells transfected with control siRNA and treated with BSA were
   used as control. PAOA treatment significantly increased fatty acid
   synthesis and β-oxidation related gene expression compared with the
   control (Figure [133]6E-[134]6I). After PAOA exposure, Acc1 mRNA
   expression increased notably in all 3 gene knockdown cells compared
   with cells transfected with control siRNA (Figure [135]6E). Fasn mRNA
   expression level in ILF2 and Rbbp4 knockdown cells was nearly two-fold
   higher than that in control cells after PAOA treatment (Figure
   [136]6F). In addition, ILF2 knockdown, rather than Teca1 and Rbbp4
   knockdown, significantly increased Srebp1 expression compared with
   control cells (Figure [137]6G). However, no significant change were
   observed in Acox1 and Cpt1α expression in all 3 genes konckdown cells
   after PAOA treatment (Figure [138]6H and [139]6I).

Rbbp4, Tcea1, and ILF2 play beneficial role in alleviating hepatocyte
steatosis

   To validate the role of the 3 TFs in regulating lipid metabolism, we
   generated mouse Rbbp4, Tcea1, and ILF2-overexpressing AML12 hepatocytes
   by transduction with a Flag-vector carrying corresponding gene (Figure
   [140]7B). After 48 h of transfection, cells were treated with PAOA for
   another 24 h. Compared to the Flag control, the Rbbp4, Tcea1, and
   ILF2-overexpression vectors significantly attenuated lipid accumulation
   in hepatocytes (Figure [141]7A and [142]7C). Overexpression of Tcea1
   significantly inhibited the expression of Acc1, and Rbbp4
   overexpression significantly promoted Cpt1α and Acox1 expression. In
   addition, the expression of Acc1, Fasn, and Srebp1 expression was
   significantly decreased when ILF2 was overexpressed. In summary,
   hepatocyte Rbbp4, Tcea1, and ILF2 regualted lipid metabolism genes
   expression, and exerted beneficial effects by suppressing lipid
   accumulation.

Figure 7.

   [143]Figure 7
   [144]Open in a new tab

   Overexpression of Tcea1, Rbbp4, and ILF2, respectively, in AML12 cells
   ameliorated hepatocyte steatosis. (A) Representative images of ORO
   staining of AML12 cells treated with PAOA after 48 h of transfection
   with Flag-NC, Flag-Tcea1, Flag-Rbbp4, and Flag- ILF2 plasmids,
   respectively. Scale bar: 100 μm. (B) Protein expression analysis of
   Flag-tagged Tcea1, Rbbp4, and ILF2 expression in AML12 cells after 48 h
   of transfection with corresponding plasmids. (C) Quantification of ORO
   positive area per field. n=3 biological replicates. (D-F) RT-qPCR
   analysis of Acc1, Fasn, Srebp1, Cpt1α, and Acox1 expression in AML12
   cells after Tcea1, Rbbp4, and ILF2 overexpression, respectively,
   compared with the Flag-NC group. All data represent the mean ± SEM.
   Student's t test, * P < 0.05, ** P < 0.01.

Discussion

   In the present study, we conducted the proteomic profiling of liver
   tissues of HFD-induced fatty liver mice by a label-free quantitative
   proteomics approach to elucidate the molecular mechanisms of fatty
   liver development. 2298 proteins were identified, of which 362 proteins
   were quantified as DEPs between the HFD and NCD groups. Further
   bioinformatics analysis of DEPs demonstrated that metabolism-related
   processes were mainly dysregulated by HFD. A subset of TFs having PPI
   network with key regulators in lipid metabolism was significantly
   enriched within DEPs. 7 TFs were further verified in vivo and in vitro,
   and 3 TFs were finally validated. Gene knockdown and overexpression
   experiments indicated that the 3 TFs played a suppressive role in
   hepatocyte steatosis through regulating the expression of genes related
   to lipid synthesis or oxidation. Thus, our study provides description
   of a potential role of Tcea1, Rbbp4, and ILF2 in the regulation of
   NAFLD for the first time.

   High-fat diet induced metabolic disorders can lead to insulin
   resistance and hyperglycemia [145]^23. Numerous proteomic studies of
   NALFD have shown that metabolism-related processes and pathways,
   including PPAR, AMPK, insulin signaling, are highly enriched through
   Gene ontology enrichment analysis, which is consistent with our results
   [146]^15^, [147]^24. In addition, KEGG analysis also enriched other
   signaling pathways, such as HIF-1 signaling, steroid biosynthesis, and
   glucagon signaling. These pathways may also contribute to the
   development of NAFLD, and could be investigated in future studies. We
   also found that nearly 60% of DEPs were metabolism-related proteins,
   mainly enriched in fatty acid and organic acid metabolism processes.
   Furthermore, molecular function enrichment results show that more than
   70% of DEPs have catalytic activity, including transferase activity,
   hydrolase activity, and oxidoreductase activity. Collectively, our
   results demonstrate that excessive lipid supply lead to the
   dysregulation of catalytic enzymes expression, which results in lipid
   metabolism disorders and promotes NAFLD.

   A previous study has shown that during prolonged HFD, changes in global
   protein levels revealed a strong metabolic reprogramming toward lipid
   consumption and accumulation in response to excessive dietary lipid
   exposure, which suggests that prolonged HFD could coordinate metabolism
   in reponse to the lipid challenge [148]^24. We noticed that 15% of DEPs
   were nucleic acid binding proteins, composed of RNA binding proteins
   and DNA binding proteins, playing crucial role in regulating gene
   expression. Upstream regulators were predicted based on the DEPs
   identified using the IPA analysis, we constructed a crucial interaction
   network consisting of protein-upstream regulator-biological processes,
   which revealed that SREBPs, PPAR-α, PPAR-γ, INSR, and ESRRA were key
   upstream regulators modulating lipid oxidation and lipid droplets
   formation. Of note, 4 of the 5 key regulators including SREBPs, PPAR-α,
   PPAR-γ, and ESRRA were TFs, suggesting aberrant activation of TFs might
   be a critical factor in regulating catalytic enzymes expression and
   driving liver metabolic disorders.

   Sterol regulator element binding proteins (SREBPs) are a family of
   transcription factors involved in the biogenesis of cholesterol, fatty
   acids, and triglycerides [149]^11^, [150]^25^, [151]^26. In the liver,
   SREBPs are involved in the pathogenesis of NAFLD, NASH, hepatitis and
   hepatocellular carcinoma [152]^27. Nowadays, researches on SREBP
   inhibitory agents are being conducted. Inhibitors of SREBP processing
   can be helpful in reducing the risk of atherosclerosis, metabolic
   syndrome, and obesity, such as 27-HC and botulin [153]^28^, [154]^29. A
   recent phase 2a clinical trial revealed that a DGTA2 inhibitor, which
   could reduce lipogenesis through downregulating SREBP1 expression,
   could enhance the therapeutic efficacy and overcome the side-effect of
   ACC inhibitors for NAFLD [155]^30. The results demonstrate that TFs
   could be potential therapeutic targets for NAFLD.

   Transcription factors are main factors that regulate gene expression,
   which are generally low-abundance proteins. Due to the technical
   difficulty of identifying low-aboundance regulatory proteins by mass
   spectrometry, few studies have sought to identify and characterize
   novel NAFLD-related TFs. In recent years, several nucleic acid binding
   proteins and TFs enrichment approaches have been established
   [156]^31^-[157]^33. With the advancement of mass spectrometry
   (MS)-based proteomics, substantial progress has been made in the
   quantitative measurement of TFs on a proteomic scale [158]^33.
   Recently, Liu et al. have quantified hundreds of TFs with differential
   abundance of DNA-binding levels in the liver of obese mice compared
   with normal mice by using the TF response element-affinity enrichment
   method coupled with LC-MS/MS, which identified PU.1 as a novel TF in
   regulating NASH progression [159]^34. In the present study, although
   only 11 DEPs were characterized as TFs by the whole proteome
   quantification, we observed their protein-protein interactions with key
   regulators and enzymes of lipid metabolism, suggesting their potential
   role in modulating NAFLD development. Finally, the expression of Rbbp4,
   TCEA1, and ILF2 were validated in both in vivo and in vitro models,
   especially in NAFLD clinical samples.

   As demonstrated from our results, the identified candidate
   NAFLD-related TFs: Tcea1, Rbbp4, ILF2, Npm1, Ruvbl2, Npm1, Nucb1, Pc,
   Pura, Sra1, Preb, and Carhsp1 (Table [160]1) are found in the same
   clusters with already well-known NAFLD-related proteins (Figure
   [161]3F), enhancing their potential implication in NAFLD. Notably,
   Ruvbl2, Pc, Sra1, and Preb which were found upregulated in NAFLD liver
   tissue from our results, are literally confirmed for their regulation
   in promoting lipid accumulation of hepatocytes. However, our study
   provided a description of a role for Tcea1, Rbbp4, and ILF2 in the
   regulation of inhibiting lipid accumulation. The results demonstrate
   the high-confidence of our MS results, and also imply the complexity
   and heterogeneity of signaling networks interaction caused by metabolic
   disorders. To our knowledge, there are no any studies that have
   implicated these three proteins in the modulation of lipid metabolism
   and NAFLD.

   Tcea1 is a transcript elongation factor that stimulates RNA cleavage by
   RNA polymerase II by allowing backtracked enzymes to resume
   transcription elongation [162]^35. In human cells, Tcea1 affects
   transcript elongation and cell viability [163]^36. Recent studies have
   unveiled the function of Tcea1 during stress responses, which
   demonstrated that tcea1 was an essential component for thermal
   adaptation in plants [164]^37, preventing transcription stress [165]^38
   and oxidative stress [166]^39. Herein, we observed that Tcea1 was also
   essential for fatty acid excess adaptation in hepatocytes. However,
   little is known about its regulation during stress responses,
   especially in pathological states. It has been reported that the
   expression of Tcea1 could be regulated by YEATS4, a NAFLD-related
   proteins, in HCC cell lines [167]^40^, [168]^41. We therefore speculate
   that the downregulation of Tcea1 in NAFLD may be regulated by YEATS4,
   which should be investigate in future study.

   Rbbp4 is a histone-binding subunit that regulate chromatin metabolism.
   As a core component of several complexes, Rbbp4 plays important role in
   histone modification, chromatin assembly and remodeling, which can
   regulate gene transcription, cell cycle, and proliferation [169]^42.
   Moreover, RBBP4 has gained attention for its potential involvement in
   the mechanism of tumorigenesis, such as liver cancer [170]^43, lung
   adenocarcinoma [171]^44 and acute myeloid leukemia [172]^45. It has
   been reported that Rbbp4, a subunit of NURF complex, is involved in the
   regulation of lipid droplet size by transcriptional repression of
   target genes [173]^46. In the present study, we observed the protection
   role of rbbp4 in hepatocyte steatosis through promoting fatty acid
   β-oxidation related gene expression. However, the mechanism of its
   regulation in pathological states have not been reported.

   ILF2 encodes NF45, the regulatory subunit of NF90/NF110 complexes that
   are involved in mitotic control and various aspects of RNA metabolism,
   including transcription, RNA transport, mRNA stability, and translation
   [174]^47^, [175]^48. Depletion of the NF45 subunit is accompanied by a
   dramatic decrease in the levels of NF90 and NF110. Likewise, depletion
   of NF90 greatly reduces the level of NF45. NF45 is
   post-transcriptionally regulated by NF90, which is unstable and
   degraded in the absence of NF90. Beyond interacting with its known
   binding partners NF90 and NF110, ILF2 also interacts with numerous
   RNA-binding proteins (RBPs), including YB-1, NPM1, and various
   heterogeneous nuclear ribonucleoproteins (hnRNPs), which are directly
   involved in DNA repair and genome stability maintenance. However, the
   role of ILF2 in the regulation of lipid metabolism/NAFLD has not been
   reported. A recent study has demonstrated that YB-1, a ILF2 interacting
   protein, regulates liver lipid metabolism by regulating the
   Wnt/β-catenin signaling pathway [176]^49, which implied the possibility
   of ILF2 in the regulation of NAFLD. Thus, this study reports that ILF2
   is a potential positive modulator of lipid metabolism by inhibiting
   fatty acid biosynthesis genes expression.

   However, the role of these 3 TFs in liver and NAFLD is still unknown.
   To better understand the functions of Rbbp4, TCEA1, and ILF2 in
   lipogenesis, we studied the effects of gene knockdown in hepatocytes on
   lipid accumulation and the expression of genes involved in adipogenesis
   and β-oxidation. We found that the downregulation of these 3 TFs
   significantly increased lipid accumulation in hepatocytes, and the
   phenotype could be rescued when they were overexpressed. In addition,
   TCEA1 and ILF2 mainly affected the expression of fatty acid
   synthesis-related genes, and several β-oxidation-related genes were
   downregulated after RBBP4 knockdown. However, the above-mentioned genes
   expression patterns were different under excessive lipid exposure,
   which indicated that in the context of NAFLD, diverse mechanisms might
   underlie the lipid accumulation reduction effect of the 3 TFs. The
   present study has limitations, one is lack of verification of Tcea1,
   Rbbp4, and ILF2 function in vivo by gene knockout/overexpression mice,
   another is lack of target gene screening for Tcea1, Rbbp4, and ILF2
   which will be explored in future studies.

   In summary, our study provides a comprehensive analysis of protein
   expression patterns and biological processes perturbations involved in
   the development of NAFLD by quantitative proteomics. Of particular
   translational interest is the finding that we identified three novel
   NAFLD-related TFs, Tcea1, Rbbp4, and ILF2, which were validated in the
   NAFLD model mice and clinical liver samples. Knockdown of the 3 genes
   promoted hepatocyte steatosis by dysregulating the expression of genes
   involved in adipogenesis and β-oxidation, respectively. These findings
   proposed novel lipid metabolism-related TFs, which might have potential
   roles in preventing NAFLD.

Supplementary Material

   Supplementary methods, figures and table.
   [177]Click here for additional data file.^ (992.4KB, pdf)

Acknowledgments