Abstract Chronic alcoholic liver disease has brought great harm to human health. Alcoholic fatty liver disease is the first stage in the progression of all chronic alcoholic liver diseases. At present, there is no cell model that fully matches the etiology (high-fat diet + alcohol) of human alcoholic fatty liver disease. We used 100 mM ethanol +6.25 μM PA to establish the ethanol combined with PA-induced mouse hepatocyte AFLD model (EP-AFLD hepatocyte model) and performed the RNA-seq transcriptome sequencing. Through bioinformatics analysis and comparison, we discovered that the EP-AFLD hepatocyte model was more suitable for studying the pathological mechanism of AFLD than the mouse AFLD hepatocyte model induced by ethanol alone. And through bioinformatics analysis, we further discovered that 77 genes from the differential expression gene set of EP-AFLD hepatocyte model were engaged in the pathological process of mouse AFLD and 40 genes were involved in the pathogenesis of both mouse AFLD and human AFLD. In this study, a novel mouse hepatocyte AFLD model was successfully established by combining ethanol and PA, which can be used to study the molecular mechanism of the pathogenesis of AFLD in mice or humans. This study will provide a brand-new in vitro experimental platform for the in-depth study of AFLD pathogenesis and the screening of AFLD therapeutic drugs. Keywords: Ethanol, Palmitic acid, Alcoholic fatty liver disease, Mouse hepatocyte AFLD model 1. Introduction Alcohol has toxic effects on many important organs of the human body, especially the liver [[39]1]. It was estimated that 2.3 billion people drink alcohol worldwide [[40]2]. Alcohol consumption has become one of the most important inducers of chronic liver disease [[41]3]. It was estimated that 1–2 million people die of cirrhosis and chronic liver diseases every year, and more than half of them might be attributed to alcohol consumption [[42]4]. Alcoholic fatty liver disease (AFLD) is an early liver disease caused by the increase of liver lipid deposition induced by long-term and large-scale intake of alcohol [[43]5]. Studies have shown that about 90% of chronic heavy drinkers will develop AFLD, and long-term alcoholism can aggravate disease progression, deteriorate into alcoholic hepatitis, liver fibrosis, and even cirrhosis [[44]6]. Therefore, prevention and treatment of AFLD can curb the disease process of alcoholic liver disease and have important significance in blocking irreversible liver damage. The most important pathological process of AFLD is hepatocyte steatosis [[45]7]. Alcohol stimulation can lead to disorder of lipoprotein synthesis and metabolism and insufficient fatty acid oxidation in the liver. With the excessive intake of lipid substances, the deposition of fat in the liver is intensified, which leads to AFLD [[46]8]. Researchers' understanding of the molecular mechanism of hepatocyte steatosis mainly comes from the research results of cell models and animal models. Therefore, in order to better simulate the pathological process of AFLD in human body, researchers have established various animal models and cell models of AFLD. The AFLD mouse model was usually established by acute binge ethanol feeding [[47]9], or ad libitum oral ethanol in drinking water [[48]10,[49]11], or chronic ethanol feeding (Lieber-DeCarli model) [[50]12], or intragastric chronic ethanol feeding (Tsukamoto-French model) [[51]13], or alcohol feeding simultaneously with nutritional composition modification, drug factors (such as concanavalin A or carbon tetrachloride), hormones, CYP inducers, TLR ligands, virus infection and gene manipulation (transgenic or gene knockout) (“Second hit” or “Multiple hits” model) [[52]14,[53]15], or chronic + binge feeding (NIAAA model or Gao-binge model) [[54]16]. The cell model of AFLD was usually established by ethanol-treated hepatocyte [[55]17]. Although the in vitro cell model cannot completely replace the in vivo animal model, the in vitro cell model is an important choice for the study of the mechanism of the cells that play a key role in the animal model in the pathological process. Therefore, a cell model more in line with the pathological process in vivo is essential for the study of the pathological mechanism of diseases. Bin Gao et al. fed mice with high-fat liquid feed containing 5% ethanol for 10 days and combined with one binge to induce mice AFLD model (NIAAA model) [[56]16], which highly mimicked the pathogenesis of human AFLD, resulting in significant liver damage in mice, significant increase in liver tissue lipids, and significant increase in serum ALT and TG. In the in vitro cell model of AFLD, researchers mainly use ethanol to stimulate hepatocytes to induce hepatocyte steatosis [[57]18]. Studies showed that 75 mM, 100 mM and 200 mM ethanol can promote hepatocyte steatosis [[58]19,[59]20]. However, this method is not consistent with the pathogenesis of human AFLD [[60]21], which is usually induced by people's preference for high calorie fat diet when drinking alcohol. Palmitic acid (PA) is hexadecanoic acid, which is a saturated higher fatty acid. Studies have shown that PA could promote lipid accumulation in AML12 hepatocytes [[61]22,[62]23]. In this paper, we explored to establish an in vitro hepatocyte AFLD model consistent with the common etiology of AFLD formation in human or mouse (high-fat diet + alcohol) through PA in combination with ethanol, so as to provide an in vitro experimental model more in line with clinical reality for further in-depth study of the common pathogenesis of AFLD and screening of therapeutic drugs for AFLD in the future. 2. Materials and methods 2.1. Cell culture AML12 mouse hepatocyte cell line (ATCC No. CRL-2254) was purchased from the National Collection of Authenticated Cell Cultures (Shanghai, China) and cultured in DMEM/F12 medium containing 10% FBS and Penicillin (100 IU/mL)-Streptomycin (100 μg/mL) solution. The culture conditions were set at 5% CO[2], 37 °C. 2.2. PA stock solution and BSA stock solution preparation 50 mg PA was dissolved in 974.96 μL ethanol, filtered by a 0.22 μm sterile filter to obtain 200 mM PA solution. 7.5 mM PA solution was obtained by adding 375 μL 200 mM PA solution to 9625 μL DMEM/F-12 complete culture solution. Due to the precipitation of PA in the culture medium at room temperature or 37 °C, this step needs to be carried out in a 60 °C water bath to obtain a fully dissolved 7.5 mM PA solution. 3.75 mM PA stock solution (containing 321.52 mM ethanol and 10% BSA) was prepared by mixing 7.5 mM PA solution with 20% BSA solution (10 g BSA dissolved in 50 mL DMEM/F-12 complete culture solution) in equal proportion. 10% BSA stock solution (containing 321.52 mM ethanol) was prepared by mixing 20% BSA solution (containing 643.04 mM ethanol) and DMEM/F-12 complete culture solution in equal proportion. 2.3. Hepatocyte AFLD model induced by ethanol combined with PA 13 mm diameter cell growing-coverslips were disinfected 5 min in 75% ethanol, then they were rinsed three times with sterile PBS and once with serum-free culture solution to remove the residual ethanol. One disinfected-cell growing-coverslip was transferred to one well of a 6-well plate with sterile forceps. 1 mL 1.5 × 10^5 cells/mL AML12 cell suspension solution was added to one well of the 6-well plate, then 1 mL DMEM/F-12 complete culture solution was added to make the cell culture system reach 2 mL/well. 24 h later, the supernatant was replaced with 2 mL fresh DMEM/F-12 culture medium without serum and AML12 cells were starved for next 24 h. Then, the supernatant was discarded, then AML12 hepatocyte steatosis was induced by 100 mM ethanol combined with 3.125, 6.25, 12.5, 25, 50, and 100 μM PA, respectively. 36 h later, the optimal PA + ethanol concentration combination for inducing AML12 hepatocyte steatosis was analyzed by cellular Oil Red O staining. At the same time, total RNA was extracted from a portion of AML12 hepatocytes for RNA-Seq transcriptome sequencing. 2.4. Cellular Oil Red O staining 0.5 g of Oil Red O solid dye was dissolved in 100 mL propylene glycol at 37 °C, then the dye solution was filtered by a 200-mesh cell sieve to make Oil Red O stock solution. The AML12 hepatocytes were washed twice with PBS after the supernatant was discarded, then 4% paraformaldehyde was added to fix the AML12 hepatocytes for 30 min at room temperature. The AML12 hepatocytes were washed 3 times with deionized water after the fixing solution was removed; then 100% propylene glycol was added dropwise for dehydration 10 min at room temperature. Propylene glycol was removed, Oil Red O stock solution preheated at 37 °C was added and AML12 hepatocytes were stained 30 min in a 37 °C constant humidity chamber. 85% propylene glycol was added dropwise for counterstaining 1-2 min after Oil Red O was removed; then 85% propylene glycol was discarded, deionized water was added to wash AML12 hepatocytes twice. Hematoxylin stain was added to stain the cell nuclei for 5 min, then it was removed and deionized water was added to wash AML12 hepatocytes twice. Cell growing-coverslips were taken out from the 6-well plate and immersed into lithium carbonate saturated solution for bluing for 2 s, then it was washed with deionized water, dried at room temperature and sealed with glycerin gelatin. The steatosis in AML12 hepatocytes were observed by imaging with an upright microscope, and the data were processed with Image Pro Plus software. 2.5. Alcoholic fatty liver disease (AFLD) mice model preparation 16 6-week-old female C57BL/6 mice which were purchased from the Experimental Animal Center of the Chinese Academy of Sciences were raised in the SPF-level breeding room of the Experimental Animal Center of Shanghai University of Traditional Chinese Medicine. Control Lieber–DeCarli liquid diet (TP4030C) and ethanol Lieber–DeCarli liquid diet (TP4030D) were purchased from Trophic Animal Feed High-tech Co., Ltd., Nantong, Jiangsu Province, China. 16 mice were randomly divided into two groups and were adaptively fed for 5 days with TP4030C which was ethanol-free. After that, mice in the control group were fed with TP4030C daily for 10 days. Mice in the ethanol group were fed with TP4030D containing 5% ethanol daily for 10 days to induce alcoholic fatty liver disease (AFLD). At the 11th day, all mice were sacrificed. The liver tissues were collected for the RNA-Seq transcriptome sequencing [[63]24]. 2.6. RNA-seq transcriptome sequencing and the screening of differential expression genes The total RNA from hepatocytes and mice liver tissues was extracted and purified. The purified total RNA was subjected to mRNA isolation, fragmentation, first-strand cDNA synthesis, second-strand cDNA synthesis, terminal repair, 3′ end plus A, ligation, and enrichment, to complete the construction of the sequencing sample library. The library concentration was tested and the library size was detected. The cluster generation and the first-stage sequencing primer hybridization were performed. The sequencing reagents were provided, the flow cell with the cluster was loaded onto the machine and the paired-end program was used for double-end sequencing. The sequencing was controlled by the data collection software provided by Illumina and real-time data analysis was performed. The library construction and sequencing were performed at Shanghai Biotechnology Corporation, Shanghai, China [[64]24]. After qualified data were obtained from the raw reads, the reads were converted into FPKM to standardize the gene expression, so then obtain comparable data on the gene expression levels of different genes and different samples. Using edgeR to perform differential gene analysis between samples, the p value was obtained, and a multihypothesis test was performed. The threshold of the p-value was determined by controlling the FDR (false discovery rate). The corrected p value was the q value. At the same time, the differential expression multiple was calculated basing on the FPKM value, which was fold change (FC). The screening conditions were q ≤ 0.05, FC ≥ 1.3, or FC ≤ 0.77. 2.7. Verification of RNA-Seq transcriptome sequencing results by qRT-PCR experiment Total RNA was extracted from hepatocyte or liver tissue, the ratio of RNA A260/280 was 1.9–2.0, indicating that the RNA purity was high. Reverse transcription was performed to obtain cDNA. cDNA, primers, and SYBR green dye were used for quantitative PCR (qPCR) analysis. The relative mRNA expression levels of target genes were calculated by 2^−ΔΔCt method. The amplification parameters were: preheating at 95 °C for 10 s, cycling at 95 °C for 5 s, and 60 °C for 30 s for 40 cycles. 2.8. Data analysis and visualization Intersection analysis of differentially expressed gene sets was completed employing Venny 2.1 online tool ([65]https://bioinfogp.cnb.csic.es/tools/venny/). The plotting of Venn diagrams, Volcano plots and Heatmap, and the KEGG pathway enrichment analysis of differentially expressed genes were completed employing an online platform ([66]https://www.bioinformatics.com.cn, last accessed on July 10, 2023) for data analysis and visualization. The construction of PPI network and screening of core target proteins was completed employing STRING network platform ([67]https://cn.string-db.org/) and Cytoscape 3.8.0 software, the core target proteins were screened based on the criterion that the node degree value exceeded the average value. Human AFLD gene set was obtained from GeneCards website ([68]https://www.genecards.org). 2.9. Statistical analysis Experimental data were processed by using ordinary one-way ANOVA and Šídák's multiple comparisons test with GraphPad Prism 9.0.0 software, and the results were expressed as mean ± standard deviation (MEAN ± SD). p < 0.05 indicates a significant difference between means. 3. Resu1ts 3.1. 100 μM ethanol combined with low concentration PA had no cytotoxicity on AML12 hepatocytes In order to establish a hepatocyte AFLD model of ethanol combined with PA at a low cytotoxic concentration, the cytotoxicity of different PA concentrations combined with 100 mM ethanol on AML12 hepatocytes was studied. AML12 hepatocytes were stimulated by 3.125 μM, 6.25 μM, 12.5 μM, 25 μM, 50 μM, and 100 μM PA combined with 100 mM ethanol, respectively. The CCK-8 assay results showed that, compared with each vehicle group (BSA group), 100 mM ethanol combined with 100 μM or lower concentrations PA had no significant effect on the viability of AML12 hepatocytes, indicating that 100 mM ethanol combined with 100 μM or lower concentrations PA had not cytotoxicity to AML12 hepatocytes ([69]Fig. 1). Fig. 1. [70]Fig. 1 [71]Open in a new tab Effect of 100 mM ethanol combined with different concentrations PA on AML12 hepatocyte activity. Note: P1: 3.125 μM PA, P2: 6.25 μM PA, P3: 12.5 μM PA, P4: 25 μM PA, P5: 50 μM PA, P6: 100 μM PA, E:100 mM ethanol. V1, V2, V3, V4, V5 and V6 (vehicle groups) were the corresponding concentration BSA in corresponding PA groups. Data were displayed in ‾x ± s, and ordinary one-way ANOVA and Šídák's multiple comparisons test were performed. (n = 3 wells/group). 3.2. Establishment of hepatocyte AFLD model of ethanol combined with PA The most important pathological manifestation of hepatocyte steatosis is the accumulation of intracellular fat. Oil red O staining is one of the main staining methods for intracellular fat, and it is a commonly used indicator for the evaluation of hepatocyte steatosis degree [[72]25,[73]26]. AML12 hepatocytes were stimulated by 3.125 μM, 6.25 μM, 12.5 μM, 25 μM, 50 μM, and 100 μM PA combined with 100 mM ethanol, respectively. The intracellular fat was stained with Oil Red O staining. The results showed that 100 mM ethanol combined with different concentrations PA all could promote the steatosis of AML12 hepatocytes, among them, 100 mM ethanol combined with 6.25 μM PA induced the most significant steatosis in AML12 hepatocytes ([74]Fig. 2). Therefore, 100 mM ethanol combined with 6.25 μM PA was used to stimulate AML12 to establish a novel hepatocyte AFLD model in subsequent experiments. Fig. 2. [75]Fig. 2 [76]Open in a new tab Effect of 100 mM ethanol combined with different concentrations PA on the steatosis of AML12 hepatocytes. Note: P1: 3.125 μM PA, P2: 6.25 μM PA, P3: 12.5 μM PA, P4: 25 μM PA, P5: 50 μM PA, P6: 100 μM PA, E:100 mM ethanol. V1, V2, V3, V4, V5 and V6 (vehicle groups) were the corresponding concentration BSA in corresponding PA groups. Data were displayed in ‾x ± s, and ordinary one-way ANOVA and Šídák's multiple comparisons test were performed. Group Vs/E compared with Group Vs, **p < 0.01, ****p < 0.0001; Group Ps/E compared with Group Vs, ####p < 0.0001; Group Ps/E compared with Group Vs/E, Δ p < 0.05, ΔΔΔ p < 0.001, ΔΔΔΔ p < 0.0001. (n = 5 visual fields/group, scale bar: 50 μm). 3.3. The mouse hepatocyte model of AFLD induced by ethanol combined with PA is more suitable for the study of the pathological mechanism of AFLD in vitro than that induced by ethanol alone In order to evaluate mouse hepatocyte model induced by ethanol combined with PA or ethanol alone which one is better for the study of AFLD, we sequenced the hepatocytes of these two models with RNA-seq technology and conducted bioinformatics analysis and comparison. The quality control of FastQ data generated by high-throughput sequencers is performed by FastQC software. All sequencing data in our cell experiment are qualified ([77]Supplementary Material 1). The accuracy of differential expression gene expression changes in RNA-seq transcriptome sequencing results was verified by qRT-PCR, and the results showed that the expression change trend of the selected genes for detection was basically consistent with the RNA-seq sequencing results ([78]Supplementary Material 2). This result indicates that the RNA-seq transcriptome sequencing results of the above cell experiments are reliable, and the bioinformatics analysis based on the results is reliable. In the ethanol combined with PA-induced mouse hepatocyte AFLD model (EP-AFLD hepatocyte model), 78 genes were significantly down-regulated and 241 genes were significantly up-regulated compared to the vehicle group ([79]Fig. 3A and B). In the ethanol-induced hepatocyte AFLD model (E-AFLD hepatocyte model), 93 genes were significantly down-regulated and 260 genes were significantly up-regulated compared to the vehicle group ([80]Fig. 3A and B). Fig. 3. [81]Fig. 3 [82]Fig. 3 [83]Open in a new tab Comparison of RNA-Seq transcriptome sequencing results between the ethanol combined with PA-induced hepatocyte AFLD model (the EP-AFLD hepatocyte model) and the ethanol-induced hepatocyte AFLD model (the E-AFLD hepatocyte model). (A) Volcano plots of the distribution of gene expression fold changes of the EP-AFLD and E-AFLD hepatocyte models. (B) Heatmaps of the differentially expressed genes of the EP-AFLD and E-AFLD hepatocyte models. (C) 49 common signaling pathways of the EP-AFLD and E-AFLD hepatocyte models. (D) 18 unique signaling pathways of the EP-AFLD hepatocyte model. (E) 8 unique signaling pathways of the E-AFLD hepatocyte model. KEGG enrichment analysis was performed on the differentially expressed genes of the EP-AFLD hepatocyte model, and 67 signaling pathways were enriched ([84]Supplementary Material 2). KEGG enrichment analysis was performed on the differentially expressed genes of the E-AFLD hepatocyte model, and 57 signaling pathways were enriched ([85]Supplementary Material 2). The intersection analysis of the signaling pathways enriched from EP-AFLD and E-AFLD hepatocyte models revealed that these two models have 49 common signaling pathways ([86]Fig.s. 3C), 18 unique signaling pathways of the EP-AFLD hepatocyte model ([87]Fig. 3D), and 8 unique signaling pathways of the E-AFLD hepatocyte model ([88]Fig. 3E). In shared signaling pathways, such as Adipocytokine signaling pathway, Cholesterol metabolism, Carbon metabolism, Fatty acid degradation, Insulin resistance [[89]27], MAPK signaling pathway [[90]28], PI3K-Akt signaling pathway [[91]29], PPAR signaling pathway [[92]30], Peroxisome, Bile secretion, Tyrosine metabolism [[93]31], and Drug metabolism are associated with alcohol and lipid metabolism. Among the 18 unique signaling pathways of EP-AFLD hepatocyte model, such as Alcoholic liver disease, AMPK signaling pathway, Tryptophan metabolism [[94]32], Cortisol synthesis and secretion [[95]33], and Steroid biosynthesis [[96]34] are associated with alcohol and lipid metabolism. Among the 8 unique signaling pathways of E-AFLD hepatocyte model, except for the Biosynthesis of unsaturated fatty acids and Fatty acid metabolism pathways, all others are unrelated to alcohol and lipid metabolism. The above results indicated that the EP-AFLD hepatocyte model and the E-AFLD hepatocyte model have significant similarities, but the EP-AFLD model is more suitable than the E-AFLD hepatocyte model for the study of the pathological mechanism of AFLD. 3.4. Ethanol combined with PA-induced hepatocyte AFLD model is applicable to the study of molecular mechanism of mouse AFLD in vitro In order to explore whether the ethanol combined with PA-induced hepatocyte AFLD model (EP-AFLD hepatocyte model) is applicable to the study of the molecular mechanism of mouse AFLD (NIAAA model) in vitro, we sequenced the liver tissues of AFLD mice with RNA-seq technology, and the sequencing results of the liver tissues of AFLD mice and the ethanol combined with PA-induced hepatocyte AFLD model were analyzed and compared by bioinformatics method. The quality control of FastQ data generated by high-throughput sequencers is performed by FastQC software. All sequencing data in our mice experiment are qualified ([97]Supplementary Material 3). The accuracy of differential expression gene expression changes in RNA-seq transcriptome sequencing results was verified by qRT-PCR, and the results showed that the expression change trend of the selected genes for detection was basically consistent with the RNA-seq sequencing results ([98]Supplementary Material 4). This result indicates that the RNA-seq transcriptome sequencing results of the above mice experiment are reliable, and the bioinformatics analysis based on the results is reliable. Comparing AFLD model mice with control mice, 2619 differentially expressed genes were obtained ([99]Supplementary Material 4). Comparing the EP-AFLD hepatocyte model with the vehicle hepatocyte, 319 differentially expressed genes were obtained ([100]Fig. 3A). By performing intersection analysis on these two differentially expressed gene sets, 77 intersection genes were obtained ([101]Fig. 4A). Then, KEGG pathway enrichment was performed on 77 genes, resulting in 19 signaling pathways ([102]Fig. 4B). Among them, Carbon metabolism, Metabolism of xenobiotics by cytochrome P450 [[103]35], Bill secret, Drug metabolism cytochrome P450, Glutathione metabolism [[104]36], Fatty acid association, Valine, Leucine and isoleucine degradation [[105]37], and MAPK signaling pathway are related to alcohol and lipid metabolism. These results showed that the EP-AFLD hepatocyte model was suitable for the study of the molecular mechanism of AFLD mice model (NIAAA model) in vitro. Fig. 4. [106]Fig. 4 [107]Open in a new tab Comparison of RNA-Seq transcriptome sequencing results between the ethanol combined with PA-induced hepatocyte AFLD model (the EP-AFLD hepatocyte model) and the mouse AFLD model. (A) Intersection gene numbers of gene sets of the EP-AFLD hepatocyte model and the mouse AFLD model. (B) Group distribution of the KEGG enrichment terms of 77 intersection genes. (C) The PPI network of 77 intersection genes and the top 18 proteins with the highest Degree value of the core target proteins. We analyzed the protein-protein interaction (PPI) and network topology parameters of the above 77 target proteins, and obtained 35 core target proteins. Among them, the top 18 proteins with the highest Degree value (>mean value) are Glud1, Ephx1, Acaa2, Aldob, Aldh6a1, Idh1, Gstm1, Ugt2b35, Acot1, Arrb1, Ftcd, Sort 1, Itga2, Ppp1r15a, Gstm2, Tgfb3, Psat1, Ppargc1a ([108]Fig. 4C). These proteins may play a key role in the pathological process of mouse AFLD, and their mechanisms can be verified by the EP-AFLD hepatocyte model. 3.5. Ethanol combined with PA-induced hepatocyte AFLD model is applicable to the study of molecular mechanism of human AFLD in vitro Human and mouse are highly homologous at the genetic level, the mouse is widely used to investigate diverse aspects of mammalian biology and pathology [[109]38]. In order to explore whether the ethanol combined with PA-induced hepatocyte AFLD model (the EP-AFLD hepatocyte model) combine AFLD mice model (NIAAA model) is applicable to the study of the molecular mechanism of mouse AFLD in vitro, we conducted an intersection analysis of the differentially expressed gene set of the EP-AFLD hepatocyte model, NIAAA model and the human AFLD gene set, and obtained 40 intersection genes ([110]Fig. 5A). KEGG analysis results showed that these 40 genes were enriched to 24 signal pathways ([111]Fig. 5B), among which Carbon metabolism, Glutathione metabolism, Metabolism of xenobiotics by cytochrome P450, Valine, leucine and isoleucine degradation, MAPK signaling pathway, Drug metabolism - cytochrome P450, Bile secretion, Chemical carcinogenesis - reactive oxygen species are related to ethanol metabolism or fatty acid metabolism. These results indicated that the EP-AFLD hepatocyte model combine NIAAA model could also be used to study the mechanism of human AFLD. Fig. 5. [112]Fig. 5 [113]Open in a new tab Comparison of the differentially expressed gene sets between the ethanol combined with PA-induced hepatocyte AFLD model (the EP-AFLD hepatocyte model), the mouse AFLD model and the human AFLD. (A) Intersection gene numbers of gene sets of the EP-AFLD hepatocyte model, the mouse AFLD model, and the human AFLD. (B) Group distribution of the KEGG enrichment terms of 40 intersection genes. (C) The PPI network of 40 intersection genes and the top 12 proteins with the highest Degree value of the core target proteins. We analyzed the protein-protein interaction (PPI) and network topology parameters of the above 40 target proteins, and obtained 23 core target proteins ([114]Fig. 5C). Among them, the top 12 proteins with the highest Degree value (>mean value) are ALDOB, SORT1, EPHX1, GSTM1, GLUD1, ACAA2, ALDH6A1, FTCD, AQP1, GSTM2, PSAT1 and IDH1 ([115]Fig. 5C). These proteins may play a key role in the pathological process of human AFLD, and their mechanisms can be verified by the EP- AFLD hepatocyte model combine NIAAA model. 4. Discussion In the world, long-term drinking, especially alcoholism, will lead to AFLD, which is one of the main causes of chronic liver disease [[116]2]. It is usually accompanied by the intake of high calorie fat diet when people drink alcohol. In the study of AFLD, researchers usually give ethanol to induce mouse AFLD models on the basis of high-fat diet. Therefore, this study simulates the main pathogeny of human and mouse AFLD, and explores and establishes a novel mouse hepatocyte AFLD model induced by ethanol combined with PA. It has been reported that 100 mM ethanol has no cytotoxicity on AML12 hepatocytes [[117]39], our experiment also obtained the same results, and our experiment also showed that the degree of hepatocyte steatosis induced by 100 mM ethanol concentration was higher than that by low concentration ethanol. High concentration of PA could cause lipotoxicity of hepatocytes and affect cell viability [[118]23], so 100 μM or lower concentrations of PA were selected to combine with 100 mM ethanol to stimulate AML12 hepatocytes, and it was finally discovered that 100 mM ethanol combined with 6.25 μM PA could induce AML12 hepatocyte steatosis most significantly in the absence of cytotoxicity. Therefore, we determined that 100 mM ethanol combined with 6.25 μM PA was the best concentration combination to stimulate AML12 hepatocytes to prepare mouse hepatocyte AFLD model in vitro. In experimental studies, the hepatocyte AFLD model induced by ethanol alone is often used by researchers. We compared the hepatocyte AFLD model induced by ethanol combined with PA with that induced by ethanol alone. Ethanol combined with PA or ethanol alone both could induce hepatocyte steatosis, but hepatocyte steatosis induced by ethanol combined with PA was more obvious than that induced by ethanol alone. Transcriptomic analysis results showed that these two models have certain similarity. They share 49 signaling pathways, a large portion of which are related to alcohol metabolism and lipid metabolism, indicating that both hepatocyte models can be used for studying molecular mechanisms of AFLD. The differentially expressed genes in the EP-AFLD hepatocyte model are enriched in more signaling pathways related to alcohol and lipid metabolism than those in the E-AFLD hepatocyte model, indicating that the EP-AFLD hepatocyte model may bring more benefits than the E-AFLD hepatocyte model in studying the molecular mechanisms of AFLD. As we observed in the hepatocyte steatosis experiment, the lipid deposition in hepatocytes stimulated by 6.25 μM PA+100 mM Ethanol was much higher than that in hepatocytes stimulated solely by 100 mM Ethanol. All these results indicate that, both the EP-AFLD hepatocyte model and the E-AFLD hepatocyte model can be used to study the mechanism of AFLD in vitro, but the EP-AFLD hepatocyte model is more suitable than the E-AFLD hepatocyte model. In further studies, we analyzed the intersection gene sets of the ethanol combined with PA-induced mouse hepatocyte AFLD model (EP-AFLD hepatocyte model) with NIAAA model (mouse AFLD model), and the intersection gene sets of EP-AFLD hepatocyte model combine NIAAA model with human AFLD, respectively. The results showed that the EP-AFLD hepatocyte model is suitable for the molecular mechanism study of hepatocyte steatosis in mouse liver tissue, and EP-AFLD hepatocyte model combine NIAAA model is suitable for the molecular mechanism study of human AFLD. 77 predicted genes (including 35 core target genes) were obtained in the NIAAA model related gene set, and 40 predicted genes (including 23 core target genes) were obtained in the human AFLD related gene set. These genes are related to the hepatocyte steatosis in the pathological process of mouse or human AFLD. Among these genes, such as Gstm1 [[119]40], Psat1 [[120]41], Ppargc1a [[121]42], and IDH1 [[122]43] have been proved to be related to the pathogenesis of alcoholic fatty liver disease. However, there are still a large number of genes whose roles in AFLD have not been experimentally proven, such as Acaa2, Aldh6a1, Idh1, Ugt2b35, Acot1, Arrb1, Ftcd, Itga2, Ppp1r15a, Tgfb3, ALDOB, SORT1, EPHX1, GLUD1, FTCD, AQP1, GSTM2 and so on. This study reveals that these genes may also participate in the pathological process of AFLD, which needs attention in the follow-up AFLD research, so as to find new targets for prevention and treatment of AFLD. 5. Conclusion In summary, there are 77 genes from the differential expression gene set of EP-AFLD hepatocyte model were engaged in the pathological process of mouse AFLD and 40 genes were involved in the pathogenesis of both mouse AFLD and human AFLD. Using the EP-AFLD hepatocyte model or combine it with the NIAAA mouse AFLD model, the action mechanism of these genes in mouse AFLD or human AFLD can be preliminarily clarified through gene overexpression or gene knockout technology. This cell model will provide a convenient in vitro experimental platform for further in-depth study of the pathogenesis of AFLD and the screening of therapeutic drugs for AFLD in the future. Author contribution statement Xudong Hu and Fengjun Qiu conceived and designed the experiments. Fengjun Qiu, Rui Zeng, Du Li, Tingjie Ye, and Hua Li performed the experiments. Xudong Hu, Fengjun Qiu and Wei Xu analyzed and interpreted the data. Xiaoling Wang and Xiaofeng Yan contributed reagents, materials, analysis tools or data. Xudong Hu and Rui Zeng wrote the paper. Funding statement This study was supported by the Scientific Research Project within the Budget of Shanghai University of Traditional Chinese Medicine (2020LK002). Data availability statement The dataset supporting the conclusions of this study are publicly available. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Footnotes Supplementary data to this article can be found online at [123]https://doi.org/10.1016/j.heliyon.2023.e19359. Appendix A. Supplementary data The following are the supplementary data to this article: Multimedia component 1 [124]mmc1.pdf^ (5MB, pdf) Multimedia component 2 [125]mmc2.pdf^ (398.5KB, pdf) Multimedia component 3 [126]mmc3.pdf^ (2.2MB, pdf) Multimedia component 4 [127]mmc4.pdf^ (559.4KB, pdf) References