Tandem mass tag–based quantitative proteomics analysis reveals the effects of the α-lactalbumin peptides GINY and DQW on lipid deposition and oxidative stress in HepG2 cells

The objective of this study was to investigate the mechanism by which the α-lactalbumin peptides Gly-Ile-Asn-Tyr (GINY) and Asp-Gln-Trp (DQW) ameliorate free fatty acid–induced lipid deposition in HepG2 cells. The results show that GINY and DQW reduced triglyceride, total cholesterol, and free fatty acid levels significantly in free fatty acid–treated HepG2 cells. Based on proteomic analysis, GINY and DQW alleviated lipid deposition and oxidative stress mainly through the peroxisome proliferator-activated receptor (PPAR) pathway, fatty acid metabolism, oxidative phosphorylation, and response to oxidative stress. In vitro experiments confirmed that GINY and DQW upregulated the mRNA and protein expression of fatty acid β-oxidation–related and oxidative stress–related genes, and downregulated the mRNA and protein expression of lipogenesis-related genes by activating peroxisome proliferator-activated receptor α (PPARα). Meanwhile, GINY and DQW reduced free fatty acid–induced lipid droplet accumulation and reactive oxygen species generation, and enhanced the mitochondrial membrane potential and ATP levels. Furthermore, GINY and DQW enhanced carnitine palmitoyl-transferase 1a (CPT-1a) and superoxide dismutase activities, and diminished acetyl-coenzyme A carboxylase 1 (ACC1) and fatty acid synthase (FASN) activities in a PPARα-dependent manner. Interestingly, GW6471 (a PPARα inhibitor) weakened the effects of GINY and DQW on the PPARα pathway. Hence, our findings suggest that GINY and DQW have the potential to alleviate nonalcoholic fatty liver disease by activating the PPARα pathway


INTRODUCTION
Nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease in the world and is becoming a serious clinical and social problem because of its prevalence among the obese population (Zhao et al., 2021). Notably, NAFLD is closely associated with various diseases, including obesity, type 2 diabetes, dyslipidemia, and cardiovascular disease (Zhang et al., 2020). Although NAFLD is not a critical illness, the persistent presence of steatosis in the liver may lead to liver fibrosis, cirrhosis, liver failure, and even liver cancer (Ni and Wang, 2016). It is well known that the liver is a major organ associated closely with glucose and lipid metabolism and it plays a central role in lipid homeostasis. Excessive intake of dietary fat leads to an increase in serum free fatty acids (FFA), which will accelerate the accumulation of triglycerides (TG) in hepatocytes and further promote the occurrence of NAFLD (Xu et al., 2021). Recently, it has been shown that the onset and progression of NAFLD are associated with multiple factors, including lipid deposition, lipid peroxidation, oxidative stress, and apoptosis (Awad et al., 2016). Therefore, pharmacological interventions to improve hepatic lipid metabolism and oxidative stress, and to promote the metabolism of hepatic lipids, glucose, and other substances are of great importance for the prevention and alleviation of NAFLD.
Although synthetic drugs are available for the treatment of metabolic diseases, they are often accompanied by undesirable adverse effects. Accordingly, there is a growing interest in exploring food-derived natural products for the alleviation of metabolic diseases. Dietary proteins are known for their rich nutritional and biological properties. Nutritionally, dietary proteins are rich in amino acids, which are essential for physical development (Friedman, 1996). Furthermore, their biological properties are related to their induction of satiety and the bioactive peptides released in the gastrointestinal tract (Corrochano et al., 2019). During the past 20 years, numerous studies have shown that food-derived bioactive peptides can affect major body systems positively, particularly the cardiovascular, endocrine, and immune systems, and minimize the risk of developing chronic diseases (Manikkam et al., 2016). Moreover, food-derived bioactive peptides have been verified to be safer, milder, and more easily absorbed than synthetic drugs (Gong et al., 2020). Among the numerous protein sources, milk protein is an excellent source of bioactive peptides. Many bioactive peptides derived from milk protein have been indicated to have multiple biological functions, such as anti-inflammatory, antiobesity. and antioxidant activities (Mati et al., 2017;Ballatore et al., 2020).
α-Lactalbumin, the second major protein in whey protein, has been demonstrated to be a source of multiple bioactive peptides (e.g., anti-inflammatory, antioxidant, and angiotensin converting enzyme-inhibitory activities; Ma et al., 2016;Corrochano et al., 2019;Worsztynowicz et al., 2020). In our previous study (Chen et al., 2020), α-lactalbumin improved hepatic lipid metabolism and reduced body weight gain in obese mice. Notably, bioactive peptides are specific fragments of protein and are inactive within the parent protein. In order for the bioactive peptides to exert their health benefits, the proteins need to be hydrolyzed in the gastrointestinal tract, where the peptides are released and then enter the blood circulation system to reach the site of action (Vermeirssen et al., 2004;Möller et al., 2008). Therefore, we inferred that α-lactalbumin-derived peptides play a key role in improving hepatic lipid metabolism in obese individuals. As expected, we discovered that the highly abundant peptides Gly-Ile-Asn-Tyr (GINY) and Asp-Gln-Trp (DQW) released from α-lactalbumin by simulated gastrointestinal digestion improved FFA-induced lipid deposition in HepG2 cells and may have the potential to alleviate NAFLD . However, lipid homeostasis is a complex metabolic network in which a variety of proteins are involved in regulating lipid metabolism throughout the biological system . To our knowledge, no studies have examined the effects of GINY and DQW on FFA-treated HepG2 cells using proteomics. Hence, proteomics analysis is of great significance to reveal the protective mechanisms and therapeutic targets of the peptides GINY and DQW against hepatic steatosis.
Quantitative proteomics technology is a method to reveal differentially expressed proteins (DEP) accurately in life processes or diseases and to predict drug therapeutic targets and potential mechanisms (Tang et al., 2021). Compared with label-free methods, the tandem mass tag (TMT) has the advantages of high throughput and resolution, accurate protein quantification, and good reproducibility (Kan et al., 2020). Therefore, in our study, TMT-based quantitative proteomics and bioinformatics analyses were used to reveal the mechanism by which GINY and DQW modulated FFA-induced lipid metabolism disorder in HepG2 cells. Further validation was performed using multiple experiments, and the proteomic results were verified by quantitative real-time (qRT)-PCR and western blotting. This study proved the potential role of α-lactalbumin peptides GINY and DQW in preventing and alleviating NAFLD.

MATERIALS AND METHODS
No human or animal subjects were used, so this analysis did not require approval by an Institutional Animal Care and Use Committee or Institutional Review Board.

Cell Culture and Treatment
HepG2 cells were provided by the Stem Cell Bank of the Chinese Academy of Sciences. The cells were cultured in Dulbecco's modified Eagle's medium (Hy-Clone) with 10% fetal bovine serum (Gibco) and 1% penicillin-streptomycin at 37°C under 5% CO 2 ambient conditions. To simulate hepatocyte steatosis, HepG2 cells (1 × 10 5 /mL) were inoculated in 6-well plates, and upon reaching 80% to 90% confluence, HepG2 cells were exposed to 0.5 mM mixed FFA (oleic acid: palmitic acid = 2:1) with or without the addition of 250 mg/mL peptide GINY or DQW for 24 h . For the PPARα inhibition experiments, HepG2 cells were exposed to 0.5 mM FFA with 10 μM GW6471 and

Protein Digestion and TMT Labeling
Protein digestion and TMT-labeling was performed as described previously . The experimental groups were divided into 4 groups: the control group, the FFA group, the FFA + GINY group (GINY), and the FFA + DQW group (DQW). There were 3 parallel samples in each group. In brief, the total protein was extracted from HepG2 cells by sodium dodecylsulfate-dithiothreitol (4% SDS, 100 mM Tris-HCl, 1 mM dithiothreitol, 1% protease inhibitor, pH 8.0). After centrifugation, the protein concentration in the supernatant was determined using a BCA protein assay kit. A protein sample (200 μg) was incorporated into 30 μL sodium dodecylsulfate-dithiothreitol buffer (4% SDS, 100 mM dithiothreitol, 150 mM Tris-HCl, pH 8.0). The protein suspensions were then digested with 4 μg of trypsin (Promega) in 40 μL of NH 4 HCO 3 buffer overnight at 37°C. The resulting peptides were desalted on C18 cartridges and the peptides were collected as a filtrate. For labeling, trypsin-digested peptide samples were labeled using TMT reagent (Thermo Fisher Scientific).

Liquid Chromatography-Electrospray Ionization Tandem MS Analysis and Data Analysis
Liquid chromatography (LC)-tandem MS (MS/ MS) was performed as described previously . Briefly, the LC-MS/MS analysis was performed on a Fusion orbitrap mass spectrometer coupled to Easy nLC (Thermo Fisher Scientific). The peptide mixtures were injected onto a C18 reversedphase column (Thermo Fisher Scientific) and separated with a 90-min gradient at a flow rate of 250 nL/min. Buffer A and B were 0.1% formic acid and 80% acetonitrile with 0.1% formic acid, respectively. The mass spectrometer was operated in positive ion mode, and the peptides were recognized using a data-dependent method. Then, MS/MS spectra were recognized using the MASCOT engine embedded in Proteome Discoverer 1.4 against the UniPort Homo sapiens database; the FASTA database contained 194609 protein sequences (UniPort_Homo_sapiens_194609_20210223, http: / / www .UniPort .org). For protein identification, the search options were as follows: peptide mass tolerance, 20 mg/kg; fragment mass tolerance, 0.1 Da; enzyme, trypsin; maximum missed cleavages, 2; fixed modification, carbamidomethyl (C), TMT 10plex (N-term), TMT 10plex (K); variable modification, oxidation (M), false discovery rate (FDR) ≤ 0.01.

Bioinformatic Analysis
Bioinformatic analysis was performed as described previously (Duan et al., 2020. The fold change (≥ 1.2 or ≤ 0.83) and FDR (Benjamini-Hochberg method)-adjusted P-value (< 0.05) were used to identify DEP between 2 groups. The Gene Ontology (GO) terms of the DEP were annotated using the software Blast2GO (https: / / www .blast2go .com), including the biological process, cell component, and molecular function. For pathway analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the DEP was performed using the DAVID database (https: / / david .ncifcrf .gov). Protein-protein interactions (PPI) were analyzed using STRING (http: / / string -db .org).

Mitochondrial Membrane Potential Assay
The mitochondrial membrane potential (MMP) in HepG2 cells was assessed using a JC-10 assay kit (Solarbio). After treatment, 1 mL Dulbecco's modified Eagle's medium containing 10 μM JC-10 was added to each well and incubated for 30 min at 37°C in an incubator. Then, the cells were washed 3 times using PBS, and images were captured on an immunofluorescence microscope (Olympus).

Reactive Oxygen Species Analysis
Reactive oxygen species (ROS) analysis was performed as described previously . Briefly, HepG2 cells were stained with 10 μM 2′,7′-dichlorodihydrofluorescein diacetate (Beyotime) for 30 min. After staining, cells were washed twice using PBS. Then, the fluorescence signals were captured using immunofluorescence microscopy.

Transmission Electron Microscopy
Transmission electron microscopy was performed as described previously (Lei et al., 2019). Briefly, HepG2 cells were fixed with 3% glutaraldehyde at 4°C for 12 h and then fixed in 1% osmium tetroxide for 2 h. The cells were dehydrated with a graded acetone series and finally embedded in Epon. Ultrathin sections were stained with uranyl acetate and lead citrate, and examined using a Zeiss 900 electron microscope operated at 80 kV.

Real-Time Quantitative PCR
The qRT-PCR analysis was performed as described previously . Briefly, total RNA was extracted using TRIzol reagent (, and cDNA was synthesized using the cDNA Synthesis Kit (TaKaRa). Last, qRT-PCR was performed using a CFX96 Real-Time PCR Detection System with SYBR Green qPCR Mix. Relative gene expression levels were normalized to β-actin using the 2 −ΔΔCt method. The gene primer sequences are listed in Table 1.

Western Blotting
Western blotting was performed as described previously . Briefly, proteins were extracted from HepG2 cells using radioimmunoprecipitation assay buffer lysis buffer with phenylmethylsulfonyl fluoride and phosphatase inhibitors. Equal amounts of protein (50 μg) were separated by SDS-PAGE, followed by transfer onto a nitrocellulose filter membrane. The membrane was blocked and incubated with the primary antibody in 5% BSA in Tris-buffered saline Tween at 4°C overnight. Next, the membrane was incubated with alkaline phosphatase-conjugated secondary antibody and then incubated with alkaline phosphatase. The protein bands were detected using a FluorChem Imaging System, and the relative density of each band was quantified using ImageJ software.

Statistical Analysis
All experiments were performed at least 3 times. The data are expressed as the means ± standard deviation. Student's t-test was performed to test for significant differences between 2 groups. One-way ANOVA followed by Duncan's test was used to determine significant differences between multiple groups. A P-value < 0.05 or an FDR-adjusted P-value < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS 20.0.

Effects of Peptides GINY and DQW on Lipid Deposition
HepG2 cells originate from human hepatocytes and have been demonstrated to retain various biochemi-  (Javitt, 1990). Oleic and palmitic acids can induce intracellular lipid deposition in HepG2 cells (Willebrords et al., 2015). Therefore, HepG2 cells are an excellent model for investigating NAFLD. In our study, the effects of GINY and DQW on lipid deposition in HepG2 cells were evaluated. As shown in Figure 1a through 1c, compared with the control group, FFA increased intracellular TG, TC, and FFA levels 2.31-, 3.98-, and 4.09-fold, respectively (P < 0.05). As reported previously (Xu et al., 2021), when excess FFA enter the liver, they cause lipid metabolism disorders in hepatocytes, accelerating the development of NAFLD, which is in agreement with our results. Encouragingly, the GINY and DQW treatment decreased intracellular TG, TC, and FFA levels remarkably (Figure 1a-1c, P < 0.05; vs. the FFA group), indicating the potential of GINY and DQW to alleviate lipid accumulation. Staining with Nile Red, which binds to lipids and emits red fluorescence, is an inexpensive and reliable method to detect intracellular lipid droplets. As shown in Figure 1d, FFA treatment enhanced significantly the intracellular fluorescence intensity relative to the control group, whereas the fluorescence intensity was decreased markedly in the GINY group and DQW group (vs. the FFA group). Thus, Nile Red staining results further demonstrated that GINY and DQW alleviated effectively FFA-induced lipid deposition in HepG2 cells.

Identification of DEP
With the development of MS, quantitative proteomics has emerged as a promising strategy to predict therapeutic targets and potential mechanisms (Tang et al., 2021). To explore the underlying mechanism of GINY and DQW further in alleviating hepatic steatosis, TMT-based quantitative proteomics was performed. Overall, 3,808 quantifiable proteins were identified from the 4 groups, and the quantitative results are presented in Supplemental Table S1 (https: / / doi .org/ 10 .17632/ t9c2xms74b .2; Chen et al., 2022). Of them, 213 proteins were identified as DEP ( Figure  2a and Supplemental Tables S2-S4, https: / / doi .org/ 10 .17632/ t9c2xms74b .2; Chen et al., 2022). Further analysis showed that 37 proteins were the mutual DEP among the 4 groups (Figure 2a and Supplemental Table  S5, https: / / doi .org/ 10 .17632/ t9c2xms74b .2; Chen et al., 2022). As shown in Figure 2b and Supplemental Table S2, 155 proteins were identified as DEP (60 upregulated and 95 downregulated) in the FFA group (vs. the control group). Compared with the FFA group, 99 proteins were identified as DEP (61 upregulated and 38 downregulated) in the GINY group (Figure 2b and   Supplemental Table S3), and 89 proteins were identified as DEP (50 upregulated and 39 downregulated) in the DQW group (Figure 2b and Supplemental Table S4). Furthermore, the heatmaps of hierarchical clustering showed that the trends in DEP between the 3 replicates of the same treatment showed better consistency (Figure 2c). In addition, the hierarchical clustering analysis clustered the DEP that may be involved in similar functions (Mohallem and Aryal, 2021).

GO Analysis of the DEP
The identified DEP from HepG2 cells were categorized further using GO annotation. In terms of molecular function, DEP were mainly related to binding, lipid binding, antioxidant activity, and enzyme regulator activity in the FFA group (Figure 3a, vs. the control group). According to the cellular component category, these DEP were chiefly associated with organelles, intracellular components, and cells in the FFA group (Figure 3a, vs. the control group). Moreover, the biological process categorization showed that these DEP primarily participated in lipid metabolic processes, small molecule biosynthetic processes, responses to stress, and metabolic processes in the FFA group (Figure 3a, vs. the control group). Chienwichai et al. (2019) found that the DEP in FFA-treated HepG2 cells were enriched mainly in lipid metabolism and energy metabolism pathways, leading to intracellular hepatic steatosis, which agreed with our results. Encouragingly, the GO annotations of DEP in the GINY group and DQW group were roughly consistent compared with FFA-treated cells (Figure 3b and 3c). Based on the molecular function category, the most abundant DEP in both groups were involved mainly in antioxidant activity, enzyme binding, enzyme regulator activity, and lipid binding (vs. the FFA group). In terms of the cellular component, the results showed that the DEP in the GINY group and DQW group were distributed primarily in the cytoplasm, organelle, cell, and mitochondrion (vs. the FFA group). Lipid metabolic process, oxidation-reduction process, response to stress, and catabolic process were the predominant 4 functions of DEP in the GINY group and DQW group (vs. the FFA group). Consistently, Guan et al. (2021) demonstrated that, based on proteomic analysis, MFG-E8 ameliorated rotenone-induced cell damage mainly by regulating lipid metabolism and oxidative stress-related pathways in L6 cells. Therefore, our GO enrichment analysis revealed that GINY and DQW may ameliorate NAFLD by modulating lipid metabolism, energy metabolism, and oxidative stress.

KEGG Pathway Enrichment Analysis of DEP
The KEGG pathway analysis helped us to understand more fully the metabolic and signal transduction pathways of the identified DEP. It has been noted (Rui, 2014) that when excessive FFA in the blood enter the liver, this disrupts hepatic lipid metabolism and energy metabolism. In our study, the results of KEGG pathway analysis indicated that these DEP in the FFA group were concentrated mainly in fatty acid metabolism, NAFLD, oxidative phosphorylation, and metabolic pathways compared with the control group, demonstrating the adverse effects of FFA on glucose and lipid metabolism (Figure 4a). Remarkably, the DEP in the GINY group and DQW group were enriched mainly in fatty acid metabolism, the PPAR signaling pathway, oxidative phosphorylation, and metabolic pathways (Figure 4b and 4c, vs. the FFA group). Similar to another proteomic study , DEP enriched in the KEGG pathway after acupuncture treatment were found to be involved mainly in the PPAR signaling pathway, fatty acid biosynthesis, and fatty acid metabolism, effectively improving hepatocyte steatosis. In conclusion, these enriched pathways are of great importance for target selection in future studies. Among them, PPARα, a nuclear hormone receptor encoded in the human genome, has been demonstrated to be a therapeutic target for NAFLD (Montaigne et al., 2021). A battery of studies has clearly established that the activation of the PPARα pathway can regulate fatty acid metabolism, energy metabolism, and oxidative stress (Kondo et al., 2009;Hao et al., 2021). Our previous study  also indicated that GINY and DQW attenuated FFA-induced hepatic steatosis in HepG2 cells by activating PPARα. Hence, these results based on prediction and in vitro experiments proved that PPARα-mediated metabolic pathways may play a critical role in alleviating lipid deposition and oxidative stress.

PPI Enrichment Analysis of DEP
Proteins generally do not execute their functions individually but rather cooperate with each other to carry out various biological functions . Among the 213 DEP, 37 proteins were mutual DEP among the 4 groups (Supplemental Table S5). To elucidate more completely the interactions among the mutual DEP, we subsequently built a PPI network using the STRING database. Consistent with the KEGG pathway results, the PPI results show that these protein clusters are closely related to the PPAR signaling pathway, fatty acid metabolism, oxidative phosphorylation, and response to oxidative stress (Figure 4d). Notably, 16 proteins (including PPARα, CPT-1a, ACOX1, ACC1, FASN, FABP1, APOA1, APOC3, SOD1, and ATP5E) were located at the center of the PPI network, which indicates they interact more strongly with other DEP (Figure 4d). Hence, these 16 proteins are key proteins among these 37 mutual DEP. Detailed information on these key DEP is listed in Table 2.
PPARα, a member of the PPAR family in the liver, is a therapeutic target in NAFLD (Mandard et al., 2004, Montaigne et al., 2021. PPARα alleviates lipid deposition mainly by upregulating the expression of fatty acid β-oxidation-related genes (such as FABP1, CPT-1a, and ACOX1) and downregulating the expression of lipid synthesis-related genes (such as ACC1 and FASN; Cui et al., 2021). In our study, compared with the control group, the protein levels of PPARα in the  FFA group were downregulated significantly by 35.00% (Table 2, P < 0.05). In contrast, GINY and DQW upregulated markedly the protein levels of PPARα 1.23-and 1.27-fold, respectively (Table 2, P < 0.05; vs. the FFA group), revealing the possible mechanism by which GINY and DQW improve NAFLD by activating PPARα.
ACC1 and FASN are key rate-limiting enzymes in fatty acid synthesis. ACC1 catalyzes the carboxylation of acetyl-CoA to generate malonyl-CoA, which can be used by FASN for fatty acid synthesis (Yeh et al., 2018). FASN is thought to be one of the major ratelimiting enzymes in regulating de novo lipid synthesis (Yeh et al., 2018). In contrast, CPT-1a and ACOX1 are the key enzymes responsible for fatty acid β-oxidation, in which ACOX1 initiates the oxidation of long-chain fatty acids (LCFA), and CPT-1a is the transport enzyme that transports fatty acids to mitochondria for degradation . Compared with the control group, the FFA treatment upregulated the protein levels of ACC1 and FASN 1.98-and 1.56-fold, respectively, whereas it downregulated the protein levels of CPT-1a and ACOX1 by 55% and 47%, respectively, causing intracellular lipid accumulation (Table 2, P < 0.05). Encouragingly, GINY and DQW downregulated markedly the protein levels of ACC1 and FASN, and upregulated the protein levels of CPT-1a and ACOX1 (Table 2, P < 0.05; vs. the FFA group), revealing that GINY and DQW improved NAFLD by reducing fatty acid biosynthesis and enhancing fatty acid β-oxidation.
FABP1 is a 14-kDa cytoplasmic protein in the liver (Ballester et al., 2017). Interestingly, knockout of the FABP1 gene inhibits the oxidation of LCFA, diverting LCFA to be stored in adipose tissue, resulting in weight gain in mice (Petrescu et al., 2013). Fibrate upregulates hepatic FABP1 expression by activating the PPARα pathway, promoting its roles in the uptake and cellular disposition of FFA (Chuang et al., 2009). Consistent with previous studies, GINY and DQW treatment upregulated markedly the protein levels of FABP1 1.76-and 1.73-fold, respectively (Table 2, P < 0.05; vs. the FFA group). Moreover, APO play a key role in lipid transport and metabolism. APOA1 is the predominant high-density lipoprotein and plays a key role in lipid transport and metabolism by activating lecithin cholesterol acyl-transferase and promoting cholesterol excretion from peripheral tissues (Sorci-Thomas et al., 2009). In contrast, APOC3 can inhibit lipoprotein lipase-mediated lipolysis and reduce the clearance of TG lipoprotein, thereby increasing plasma TG levels . A previous studies indicated that knockdown of the APOC3 gene reduces the risk of cardiovascular disease . Studies have shown that activation of PPARα could upregulate the expression of APOA1 and downregulate the expression of APOC3, thereby improving dyslipidemia in obese mice (Mandard et al., 2004). In our study, GINY and DQW upregulated markedly the protein levels of APOA1 1.60-and 1.59-fold, respectively, and downregulated the protein levels of APOC3 by 36% and 38%, respectively (Table 2, P < 0.05; vs. the FFA group). Based on these results, GINY and DQW may exert their lipid-lowering effects by regulating APOA1 and APOC3.
Superoxide dismutase 1 (SOD1) can catalyze the conversion of superoxide radicals into oxygen and hydrogen peroxide . Su et al. (2021) found that baicalin upregulated the protein expression of antioxidant enzymes (SOD) and increase the ATP  content in mitochondria, protecting the liver against oxidative stress. Moreover, ATP synthase subunit e (ATP5E) encodes the structural subunit ε of ATP synthase, and mutations in ATP5E inhibit ATP synthase biogenesis (Hejzlarova et al., 2014). ATP synthase is a key enzyme in mitochondrial energy conversion that is involved in catalyzing ATP synthesis during mitochondrial oxidative phosphorylation (Song et al., 2021). In our study, the protein levels of SOD1 and ATP5E in the FFA group were downregulated remarkably by 34% and 23%, respectively, compared with the control group, whereas GINY and DQW upregulated the protein levels of SOD1 and ATP5E significantly ( Table 2, P < 0.05; vs. the FFA group), proving that they could modulate energy metabolism and oxidative stress. Overall, according to the KEGG and PPI analyses, we speculate that GINY and DQW may ameliorate FFA-induced lipid accumulation, energy metabolism disorders, and oxidative stress mainly via the PPARα signaling pathway.

Validation of the Improvement of Lipid Metabolism by Peptides GINY and DQW
As observed by transmission electron microscopy, the number and area of lipid droplets in the FFA group were increased significantly, whereas the GINY and DQW treatments reduced the area and number of lipid droplets (Figure 5a, vs. the FFA group). Mitochondria are essential for cellular energy metabolic processes, including the production of ATP, β-oxidation, and apoptosis. It has been well-established that FFA-induced lipotoxicity and oxidative stress can cause mitochondrial swelling (mitochondrial ultrastructural damage) and are implicated in a variety of metabolic diseases (Zhang et al., 2021b). In our study, the mitochondrial morphology in FFA-treated cells was rounder and more swollen than tubular mitochondria in normal cells, whereas GINY and DQW effectively suppressed mitochondrial damage (Figure 5a). Consistent with our findings, Lei et al. (2019) demonstrated that sulforaphane reduced the number and area of lipid droplets and improved mitochondrial function in FFA-treated human hepatocyte line 5 cells.
Because PPARα can maintain lipid homeostasis, synthetic PPARα agonists have been used in the treatment of NAFLD. For example, fenofibrate (a PPARα agonist), has proved effective in the alleviation of dyslipidemia and hepatic lipid accumulation in animal models (Lefere et al., 2020). In our study, FFA treatment markedly downregulated the mRNA expression levels of PPARα, CPT-1a, ACOX1, and FABP1 by 45.08%, 35.50%, 55.82%, and 26.77%, respectively, and upregulated the levels of FASN 2.14-fold ( Figure   5b, P < 0.05; vs. the control group). In addition, the FFA treatment downregulated the protein expression levels of PPARα, CPT-1a, ACOX1, and FABP1 by 33.44%, 21.55%, 50.81%, and 31.00%, respectively, and upregulated the levels of FASN 2.13-fold ( Figure  5c-h, P < 0.05; vs. the control group). As expected, compared with the FFA group, the GINY and DQW treatments remarkably upregulated the mRNA and protein expression levels of PPARα, CPT-1a, ACOX1, and FABP1, but downregulated the mRNA and protein levels of FASN (Figure 5 b-h, P < 0.05). Similar observations have been reported in which PPARα activation enhanced the expression of CPT-1a, ACOX1, and FABP1; and inhibited the expression of SREBP-1c, FASN, and ACC1, resulting in a reduction in hepatic and serum lipid accumulation in obese mice (Kondo et al., 2009;Petrescu et al., 2013;Yeh et al., 2018). Furthermore, UCP play a critical role in the control of energy homeostasis, and polymorphisms in uncoupling protein 1 (UCP1) and UCP2 genes are linked to obesity (Tutunchi et al., 2020). Grav et al. (2003) discovered that PPARα-stimulated UCP2 expression was linked to altered hepatic energy status and an increased mitochondrial fatty acid oxidation rate in the liver of rats. As shown in Figure 5i, the mRNA and protein expression levels of UCP2 in the FFA group were decreased by 40.00% and 35.00%, respectively, compared with the control group (P < 0.05). However, DQW treatment upregulated markedly the mRNA and protein expression levels of UCP2 in FFA-treated cells (Figure 5b and 5i, P < 0.05), indicating the potential of DQW to improve energy metabolism.
Taken together, the expression trends of PPARαmediated genes were consistent with their expression trend in the proteomics analysis, and therefore we speculate that GINY and DQW may activate the PPARα pathway, altering the expression of key DEP and ultimately attenuating hepatic steatosis.

Validation of the Improvement of Oxidative Stress by Peptides GINY and DQW
Excessive accumulation of FFA in hepatocytes stimulates the production of ROS in mitochondria, leading to a decrease in MMP and disturbing mitochondrial homeostasis (Begriche et al., 2006;Niu et al., 2020). In turn, mitochondrial dysfunction inhibits fatty acid oxidation and causes hepatic metabolic disturbances, which creates a vicious cycle (Zhang et al., 2021b). Consistent with previous reports, FFA treatment reduced MMP and ATP levels significantly, inhibited SOD activity, and exacerbated ROS production, but GINY and DQW alleviated the adverse effects of FFA on HepG2 cells (Figure 6a-d, P < 0.05). Hence, from the perspective of these indicators, GINY and DQW could improve oxidative stress and mitochondrial function. Li et al. (2012) certified that PPARα activation could upregulate the activities and protein expression of hepatic SOD and catalase (CAT), thereby protecting hepatocytes from oxidative stress, which is consistent with our findings.
Nrf2 is an oxidative stress-mediated transcription factor, and has been considered a target for NAFLD therapy. Upon exposure to oxidative stress, Nrf2 translocates into the nucleus, where it can activate the expression of antioxidant enzymes (such as SOD, CAT, and HO-1; Li et al., 2021). Hence, Nrf2 can balance mitochondrial ROS production and maintain mitochondrial transmembrane potential and ATP production, protecting cells from injury (Lei et al., 2019). In our study, FFA downregulated significantly the mRNA and protein expression levels of Nrf2 and HO-1 ( Figure  6e-h, P < 0.05; vs. the control group). However, GINY and DQW reversed these trends significantly, showing the ability to alleviate oxidative stress (Figure 6e-h, P < 0.05). Consistent with these results, formononetin has been shown to attenuate malonaldehyde levels and enhance CAT activity in rat kidney tissue by activating the PPARα/Nrf2/HO-1 pathway, thereby protecting against kidney injury .
Moreover, PGC-1α enhances mitochondrial oxidative phosphorylation through interactions with multiple transcription factors and triggers mitochondrial gene transcription and mitochondrial DNA replication driving mitochondrial biogenesis (Lei et al., 2019). Ping et al. (2015) found that salidroside ameliorated myocardial injury and increased mitochondrial respiratory function by activating the PGC-1α/Nrf2 signaling pathway. In our study, compared with the control group, FFA downregulated significantly the mRNA and protein expression of PGC-1α by 29.20% and 42.37%, respectively, whereas GINY and DQW reversed these trends markedly (Figure 6e and 6i, P < 0.05; vs. the FFA group). Similarly, Ni et al. (2022) demonstrated that the seed oil of Rosa roxburghii Tratt activated the PPARα/PGC-1α pathway and subsequently improved mitochondrial function, thereby ameliorating lipid metabolic disorders and oxidative stress. Therefore, we speculate that GINY and DQW could increase ATP and MMP levels and SOD activity by activating the PPARα pathway, thereby protecting mitochondrial function and biogenesis.

Peptides GINY and DQW Alleviated FFA-Induced Hepatic Steatosis in a PPARα-Dependent Manner
Considering the modulatory effects of peptides GINY and DQW on PPARα and the salutary effects of PPARα agonists on NAFLD, we wondered whether GINY and DQW attenuated lipid accumulation and oxidative stress in a PPARα-dependent manner in HepG2 cells. GW6471 is a typical PPARα inhibitor (Ni et al., 2022). To verify our hypothesis, we used the PPARα inhibitor GW6471 to examine the effects of the peptides GINY and DQW on the PPARα signaling pathway and lipid accumulation. Compared with the FFA group, the TG and TC levels were decreased significantly in the GINY and DQW groups, whereas GW6471 diminished the lipid-lowering effects of GINY and DQW remarkably (Figure 7a and 7b, P < 0.05), which is consistent with a previous study .
Moreover, we explored the effects of GW6471 on the protein expression of PPARα and its target gene. As shown in Figure 7c and 7d, compared with the FFA group, GINY and DQW significantly upregulated the protein expression levels of PPARα, CPT-1a, ACOX1, and HO-1, and downregulated the levels of FASN (P < 0.05). However, GW6471 significantly reduced the regulatory effects of GINY and DQW on the protein expression levels of PPARα, CPT-1a, ACOX1, HO-1, and FASN (Figure 7c and 7d, P < 0.05). In addition, GINY and DQW enhanced markedly the activities of CPT-1a and SOD, and suppressed the activities of ACC1 and FASN (Figure 7e-h, P < 0.05; vs. the FFA group), indicating GINY and DQW could improve lipid metabolism and oxidative stress. Interestingly, we found that GW6471 significantly diminished the regulatory effects of GINY and DQW on the activities of CPT-1a, ACC1, FASN, and SOD (Figure 7e-h, P < 0.05). These observations further confirmed our previous findings that GINY and DQW alleviated FFA-induced hepatic steatosis in a PPARα-dependent manner in HepG2 cells . In addition, other studies showed that GW6471 intervention inhibited the activation of PPARα, downregulated the expression of PGC-1α and CPT-1a significantly, and upregulated the expression of ACC1 and FASN in the liver, weakening the hepatoprotective effect of natural products against NAFLD, which is consistent with our results , Ni et al., 2022. In summary, these results prove that GINY and DQW may attenuate lipid deposition and oxidative stress in a PPARα-dependent manner.

CONCLUSIONS
This study reveals the protective effects of the α-lactalbumin peptides GINY and DQW against FFAinduced hepatic steatosis. Proteomics and bioinformatics results indicate that GINY and DQW alleviate lipid deposition and oxidative stress mainly via the PPAR signaling pathway, fatty acid metabolism, oxidative phosphorylation, and oxidative stress. Notably, 16 proteins (including PPARα, CPT1a, ACOX1, FASN, FABP1, APOA1, APOC3, and SOD1) were identified as key proteins regulated by GINY and DQW in the 4 pathways outlined herein. Furthermore, the results of in vitro experiments demonstrate that GINY and DQW attenuate FFA-induced lipid accumulation and oxidative stress, and these effects were diminished by the PPARα inhibitor GW6471. These results prove that GINY and DQW alleviated intracellular lipid deposition and oxidative stress in a PPARα-dependent manner. Notably, although GINY and DQW normalized hepatic lipid accumulation and oxidative stress in HepG2 cells, the cell model of NAFLD does not fully reflect human NAFLD (Tarantino et al., 2021). Therefore, future clinical studies are needed to confirm the clinical safety and side effects of GINY and DQW, and their potential to alleviate NAFLD.

ACKNOWLEDGMENTS
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The authors have not stated any conflicts of interest.