Untargeted Metabolomic Analyses and Antilipidemic Effects of Citrus Physiological Premature Fruit Drop

Increasingly globally prevalent obesity and related metabolic disorders have underscored the demand for safe and natural therapeutic approaches, given the limitations of weight loss drugs and surgeries. This study compared the phytochemical composition and antioxidant activity of five different varieties of citrus physiological premature fruit drop (CPFD). Untargeted metabolomics was employed to identify variations in metabolites among different CPFDs, and their antilipidemic effects in vitro were assessed. The results showed that Citrus aurantium L. ‘Daidai’ physiological premature fruit drop (DDPD) and Citrus aurantium ‘Changshan-huyou’ physiological premature fruit drop (HYPD) exhibited higher levels of phytochemicals and stronger antioxidant activity. There were 97 differential metabolites identified in DDPD and HYPD, including phenylpropanoids, flavonoids, alkaloids, organic acids, terpenes, and lipids. Additionally, DDPD and HYPD demonstrated potential antilipidemic effects against oleic acid (OA)-induced steatosis in HepG2 hepatocytes and 3T3-L1 adipocytes. In conclusion, our findings reveal the outstanding antioxidant activity and antilipidemic effects of CPFD, indicating its potential use as a natural antioxidant and health supplement and promoting the high-value utilization of this resource.


Introduction
Obesity has become a major public health concern due to the drastic lifestyle changes of the global population.It is estimated that approximately 500 million people worldwide suffered from metabolic disorders and obesity in 2020 [1].As obesity rates continue to rise, more people risk developing diseases such as diabetes, cardiovascular disease, and fatty liver [2].Reactive oxygen species (ROS) are metabolism byproducts that give rise to cellular oxidative stress when in excess [3], causing obesity-related complications [4].ROS alter the concentrations of molecules involved in inflammation, particularly adipocytes.Changes in the size and abundance of adipocytes promote fat generation and adipogenesis, thus stimulating pre-adipocyte differentiation into mature adipocytes [5,6].Non-alcoholic fatty liver disease (NAFLD) is associated with obesity [7].This emerging chronic liver disease disrupts lipid metabolism and increases the synthesis and accumulation of liver fat, resulting in various liver dysfunctions [8].Therefore, reducing oxidative stress and inhibiting fat generation are promising treatment options for obesity prevention.
Citrus is a plant genus that belongs to the Rutaceae family, comprising various species such as Citrus sinensis (orange), Citrus reticulata (mandarin), Citrus aurantifolia (lime), Citrus limon (lemon), Citrus paradisi (grapefruit), Citrus bergamia (bergamot), Citrus junos (yuzu), and Citrus japonica (kumquat) [9].Citrus plants are extensively cultivated in tropical and subtropical regions across the globe, making them one of the foremost fruit crops worldwide.Citrus fruits are rich in active substances such as carotenoids, essential oils, pectin, flavonoids, and limonoids [10,11], which contribute to their antioxidant, anti-allergic, antiinflammatory, anticancer, blood pressure regulation, and lipid reduction properties [12,13].According to the Food and Agriculture Organization (FAO), the global citrus cultivation area spanned 1.01 million ha in 2020 [14].Generally, the fruit-setting rate of citrus is very low (3-5%), leading to high premature fruit drop [15].
Despite various surgical and pharmacological treatments, no risk-free and effective weight management therapy is currently available.Changes in lifestyle, diet, and reduction in sedentary behavior are considered the best options for obesity prevention [16].Phytotherapy targets health issues by using plant-derived drugs.Additionally, natural plant supplements lead to significant weight loss and improved health by neutralizing ROS.Plant-based products are alternatives to weight management due to the rich active chemical substances, including polyphenols, flavonoids, carotenoids, and catechins.Citrus physiological premature fruit drop (CPFD) is a normal phenomenon during the fruit development process, which is distinct from fruit drop caused by storms, diseases, and pests.Previous studies have shown that the active plant components in dropped citrus fruits are higher than those in mature fruits [17].However, the current research on the utilization of citrus byproducts is mainly focused on processing residual materials from mature fruits, while studies on CPFD utilization are scarce.Currently, only a small proportion of CPFD is collected and used for traditional Chinese medicine, while most are considered agricultural waste.Therefore, repurposing CPFD can reduce waste and environmental pollution.Moreover, this agricultural byproduct is an attractive source of phytochemicals for human dietary supplements.
Increasing evidence suggests that citrus fruits are rich in flavonoids and good regulators of lipid metabolism.In this study, we compared the phytochemical content and antioxidant activity of physiological premature fruit drop and mature fruits of five different citrus cultivars.Subsequently, untargeted metabolomic analysis was performed on two citrus cultivars with higher phytochemical content and stronger antioxidant capacity to further explore their differential metabolites.In addition, we investigated the antilipidemic effect of selected CPFD on HepG2 and 3T3-L1 cells.

Untargeted Metabolomic Analysis of CPFDs
Based on the total phenolic and flavonoid contents and antioxidant activity of CPFD and mature fruit samples of five citrus species, the HYPD and DDPD samples exhibited the best performance for all assessments and, thus, were selected for further analysis.Metabolites in HYPD and DDPD samples obtained 779 features in positive ion mode (ESI+) and 440 features in negative ion mode (ESI−) (Supplementary Materials).Using more metabolic features in chemometric analysis can yield more reliable results in untargeted metabolomics analysis.In this study, principal component analysis (PCA) was introduced to the processed data matrix to reduce data dimensionality and enhance data interpretability.

Principal Component Analysis of CPFD
Unsupervised PCA was employed to observe the overall clustering and distribution trends between HYPD and DDPD.The observed substantial dispersion between HYPD and DDPD suggests noteworthy variances in metabolite profiles.In the ESI+ model, axis 1 (PCoA1) indicated 78.90% of total variability, while axis 2 (PCoA2) demonstrated 4.05% of variability (Figure 3A).In the ESI− model, axis 1 also showed 78.90% of total variability compared to only 3.69% in axis 2 (Figure 3B).

Untargeted Metabolomic Analysis of CPFDs
Based on the total phenolic and flavonoid contents and antioxidant activity of CPFD and mature fruit samples of five citrus species, the HYPD and DDPD samples exhibited the best performance for all assessments and, thus, were selected for further analysis.Metabolites in HYPD and DDPD samples obtained 779 features in positive ion mode (ESI+) and 440 features in negative ion mode (ESI−) (Supplementary Materials).Using more metabolic features in chemometric analysis can yield more reliable results in untargeted metabolomics analysis.In this study, principal component analysis (PCA) was introduced to the processed data matrix to reduce data dimensionality and enhance data interpretability.

Principal Component Analysis of CPFD
Unsupervised PCA was employed to observe the overall clustering and distribution trends between HYPD and DDPD.The observed substantial dispersion between HYPD and DDPD suggests noteworthy variances in metabolite profiles.In the ESI+ model, axis 1 (PCoA1) indicated 78.90% of total variability, while axis 2 (PCoA2) demonstrated 4.05% of variability (Figure 3A).In the ESI− model, axis 1 also showed 78.90% of total variability compared to only 3.69% in axis 2 (Figure 3B).

Screening and Identification of Candidate Differential Metabolites
Identifying candidate differential metabolites in untargeted metabolomics can be challenging, thus requiring accurate molecular weights, retention times, and ion fragments.In order to identify differentially expressed metabolites between HYPD and DDPD, we utilized the partial least squares discriminant analysis (PLS-DA) model to calculate the variable importance for projection (VIP) values.Additionally, to ensure higher accuracy and reliability in the structural elucidation of metabolites, we selected differential metabolites based on the criteria of Fragmentation Score > 50, VIP > 1.00, and p < 0.05.In this study, 87 compounds were identified in ESI+ mode and 10 compounds in ESI-mode as differential metabolites (Fragmentation Score > 50) (Table 1).Among the 97 compounds, DDPD upregulated 41 and downregulated 56 compounds compared to HYPD.These candidate differential metabolites comprised 4 phenylpropanoids, 7 flavonoids, 7 alkaloids, 10 organic acids, 4 terpenes, 27 lipids, and 38 other metabolites.

Screening and Identification of Candidate Differential Metabolites
Identifying candidate differential metabolites in untargeted metabolomics can be challenging, thus requiring accurate molecular weights, retention times, and ion fragments.In order to identify differentially expressed metabolites between HYPD and DDPD, we utilized the partial least squares discriminant analysis (PLS-DA) model to calculate the variable importance for projection (VIP) values.Additionally, to ensure higher accuracy and reliability in the structural elucidation of metabolites, we selected differential metabolites based on the criteria of Fragmentation Score > 50, VIP > 1.00, and p < 0.05.In this study, 87 compounds were identified in ESI+ mode and 10 compounds in ESI-mode as differential metabolites (Fragmentation Score > 50) (Table 1).Among the 97 compounds, DDPD upregulated 41 and downregulated 56 compounds compared to HYPD.These candidate differential metabolites comprised 4 phenylpropanoids, 7 flavonoids, 7 alkaloids, 10 organic acids, 4 terpenes, 27 lipids, and 38 other metabolites.

CPFD Alleviated OA-Induced Steatosis in HepG2
The human liver cancer cell line, HepG2, is commonly used to simulate liver steatosis induced by obesity [21].This study utilized the 3-(4,5-Dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay to assess the effects of various concentrations of HYPD and DDPD on HepG2 cell viability.Furthermore, this method determines the optimal treatment concentration of HYPD and DDPD in HepG2 cells and maximizes their
Based on the above results, it can be observed that two types of citrus premature fruit drop have distinct metabolomic features with respect to their physiological fruit drop.HYFD is rich in terpenoids and lipids, whereas DDPD and HYFD have similar levels of organic acids, flavonoids, and alkaloids, consistent with their total polyphenol and total flavonoid content.The differences in the aforementioned metabolic features can be attributed to varietal characteristics as well as other factors such as growing region, growth conditions, irrigation, and fertilization.

Antilipidemic Effects of CPFD 2.4.1. CPFD Alleviated OA-Induced Steatosis in HepG2
The human liver cancer cell line, HepG2, is commonly used to simulate liver steatosis induced by obesity [21].This study utilized the 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay to assess the effects of various concentrations of HYPD and DDPD on HepG2 cell viability.Furthermore, this method determines the optimal treatment concentration of HYPD and DDPD in HepG2 cells and maximizes their efficacy without compromising cell viability.In this study, HYPD treatment at 0.005 to 0.04 mg/mL did not affect the cell viability of HepG2 after 24 h (Figure 5A).When the HYPD concentration increased to 0.08 mg/mL, the cell viability decreased without significant difference.Nevertheless, cell viability significantly decreased to 82.25% compared to the control group (p < 0.001) when the HYPD concentration was increased to 0.32 mg/mL.There was also a significant decrease in cell viability (80.33%) compared to the control group (p < 0.01) when the cells were treated with 0.64 mg/mL DDPD (Figure 5B).Therefore, we selected two concentrations, 0.16 mg/mL and 0.32 mg/mL, of HYPD and DDPD for subsequent experiments.The accumulation of lipids in hepatocytes is a hallmark of hepatic steatosis.HepG2 cell lines were cultured in a mixture of 0.5 mM OA and CPFD at different concentrations for 24 h to determine the inhibitory effect of different HYPD and DDPD on OA-induced lipid accumulation in the cells.Figure 5C illustrates that the OA treatment increased the triglyceride (TG) content of HepG2 cell lines compared to the control group, but a dose-dependent inhibitory effect was observed in cells that were treated with HYPD.Nonetheless, DDPD exhibited less impact on the TG of HepG2 cells than HYPD.Furthermore, the Oil Red O (ORO) staining confirmed that HYPD and DDPD reduced OA-induced lipid accumulation dose-dependently (Figure 5D).The 3T3-L1 cell line is one of the most commonly used cellular systems to study adipogenesis.When cultured in the presence of differentiation inducers, these cells exhibit characteristics of mature adipocytes in metabolism and lipid accumulation [22].In the present study, the effect of CPFD (0.005-0.64 mg/mL) on the survival of 3T3-L1 cells were assessed via MTT assay to determine the non-toxic concentrations of HYPD and DDPD on undifferentiated 3T3-L1 pre-adipocytes.At 0.64 mg/mL, HYPD slightly decreased the cell viability, but the DDPD was significantly different from the control group (p < 0.005) (Figure 6A,B).The 3T3-L1 pre-adipocytes differentiated within eight days when treated with 0.16 and 0.32 mg/mL HYPD or DDPD, as described in the methodology section.The results revealed that intracellular TG accumulation increased by 314% in differentiated cells compared to undifferentiated cells.It is worth noting that there was a significant dose-dependent decrease in TG content with increasing concentrations of CPFD.Adding 0.64 mg/mL of HYPD and DDPD decreased TG content by 29.06% and 48.73%, respectively, relative to the differentiated cells (Figure 6C).These findings suggest the superior inhibitory effect of DDPD against lipid accumulation in 3T3-L1 pre-adipocytes.In addition, ORO staining confirmed the dose-dependent reduction in lipid accumulation in 3T3-L1 preadipocytes by HYPD and DDPD (Figure 6D).

Discussion
Excessive ROS are closely associated with lipid accumulation; thus, reducing cellular ROS levels and inhibiting adipocyte differentiation are critical preventive strategies for alleviating obesity [23].In this study, CPFD samples were richer in bioactive components and had stronger antioxidant activity than mature fruit of five citrus varieties, which aligned with previous reports [24].Furthermore, HYPD and DDPD samples were further investigated due to their highly abundant plant bioactive components and prominent antioxidant activities.Based on the untargeted metabolomic analysis of differential metabolites, HYPD recorded more terpenoids and lipids than DDPD.Meanwhile, the antilipidemic in vitro experiments demonstrated that HYPD was superior in inhibiting OA-induced HepG2 fat deposition, while DDPD performed better in the adipogenesis suppression of 3T3-L1 cells.
Citrus plants are rich in various bioactive substances, including carotenoids, flavonoids, limonoids, volatile oils, and coumarins.Several factors, including genetics, ripening stage, and different fruit parts, influence the accumulation of bioactive substances in citrus fruits.Moreover, the composition of citrus fruits varies between species.Untargeted metabolomics primarily relies on LC-MS/MS without pre-selecting or screening specific metabolites but rather aims to explore all possible metabolites present in a sample to obtain comprehensive and integrated metabolic information.In the untargeted metabolomics analysis of differential metabolites, several common compounds were identified in all citrus varieties in the present study, including synephrine, limonene-1,2-epoxide, artemetin, limonoate d-ringlactone, syringetin 3-glucoside, neohesperidin, diosmetin, and (R)-meranzin.Synephrine is a widely occurring alkaloid in citrus fruits, particularly in bitter orange Citrus aurantium.This compound exerts a similar effect to adrenaline, such as increasing metabolic rate, promoting energy expenditure, and fat oxidation [25].Similarly, in this study, DDPD exhibited higher levels of synephrine compared to HYPD.High levels of meranzin were also found in the citrus varieties in this study, which was also reportedly abundant in bitter orange essential oil [26].This coumarin compound is commonly found in citrus peel [27].It is noteworthy that among the differential metabolites between HYPD and DDPD, we found that HYPD had a higher abundance of lipids compared to DDPD.However, DDPD accumulated more key products of diacylglycerol, such as PC (16:0/0:0), PC (15:0/0:0), PC (18:1/0:0), PE (16:0/0:0), PE (18:1(9Z)/0:0), MGDG (18:3/18:3), PC (18:2/0:0), and PE (18:2/0:0), following the classical glycerophospholipid biosynthetic pathway, which might contribute to a higher content of essential oil in mature fruits.Interestingly, in another study, we found that the essential oil yields of mature Citrus aurantium L. and Citrus.Changshan huyou.B. Chang were 2.28% and 0.95%, respectively, further supporting the speculation based on the above result [28].
Plant bioactive compounds improve health and prevent chronic diseases in vitro and in vivo by alleviating metabolic disorders, inflammation, and oxidative stress [29,30].These studies suggest that plant-derived bioactive compounds offer new avenues for developing dietary strategies to prevent various diseases [31].Hepatic steatosis refers to the excessive accumulation of lipids in the liver, particularly neutral lipid buildup such as TG within lipid droplets (LDs) in hepatocytes [32].The OA is commonly used to induce hepatic steatosis in HepG2 cells [33].Studies have investigated the impacts of accelerating TG breakdown in reducing hepatocyte lipid accumulation [34].In this study, HYPD and DDPD extracts significantly reduced TG levels, indicating their potential antilipidemic effect.Notably, HYPD performed better than DDPD, possibly attributed to the higher terpenoid content in the former than in the latter [35].Furthermore, neohesperidin exhibited higher expression in HYPD.An earlier study reported that neohesperidin could exert hypolipidemic effects by activating the AMP-activated protein kinase (AMPK) pathway and regulating target genes including Stearoyl-CoA desaturase-1(SCD-1), Fatty acid synthase (FAS), and Acyl-CoA oxidase (ACOX) [36].
Lipogenesis is the process where immature pre-adipocytes differentiate into mature adipocytes, accumulating lipids as fat droplets [37].Differentiating pre-adipocytes into adipocytes increases intracellular lipid accumulation [38].The results of this study provided evidence for the efficacy of HYPD and DDPD in attenuating lipogenesis and lipid accumulation in 3T3-L1 adipocytes, with DDPD demonstrating superior effects compared to HYPD.Furthermore, the HYPD and DDPD treatments significantly reduce the accumulation of TG compared to untreated differentiated cells, suggesting potential anti-obesity effects by inhibiting lipid accumulation.At a high dose (0.32 mg/mL), HYPD and DDPD exhibited anti-lipogenesis effects without cytotoxicity on pre-adipocytes and differentiated 3T3-L1 cells.Therefore, HYPD and DDPD could inhibit TG accumulation and reduce lipogenesis without cellular toxicity.A previous study also demonstrated that immature Citrus sunki HORT.exTANAKA peel extract increased the phosphorylation of AMPK and acetyl-CoA carboxylase (ACC) in mature 3T3-L1 adipocytes, promoted fatty acid β-oxidation, and enhanced lipolysis through the phosphorylation of cAMP-dependent protein kinase (PKA) and hormone-sensitive lipase (HSL) [39].Synephrine is a naturally occurring alkaloid that is widely used in weight loss due to its low toxicity and excellent fat oxidation effects [40].The high content of synephrine in DDPD is also likely an important reason for DDPD's superior ability to inhibit lipid accumulation in 3T3-L1 adipocytes compared to HYDP.In summary, HYPD and DDPD exhibit promising antilipidemic effects in our in vitro cell experiments.However, future studies should investigate the mechanisms underlying the weight-reducing and antilipidemic effects of HYPD and DDPD by incorporating animal models.

Materials
Citrus aurantium 'Changshan-huyou' physiological premature fruit drops and mature fruit drops were obtained from Quzhou, Zhejiang Province, China.Citrus aurantium L. 'Daidai' physiological premature fruit drops and mature fruit drops were obtained from Lianyuan, Hunan Province, China.Citrus reticulate 'Blanco' physiological premature fruit drops and mature fruit drops were obtained from Meishan, Sichuan Province, China.Citrus reticulata cv.Suavissima Ougan physiological premature fruit drops and mature fruit drops were obtained from Lianyuan, Hunan Province, China.Citrus maxima (Burm.)Merr.cv.Jiangyong Yu physiological premature fruit drops and mature fruit drops were obtained from Jiangyong, Hunan Province, China.In this study, citrus physiological premature fruit drops were collected between June and July, and citrus fruits with a diameter between 2.0 and 3.0 cm were selected as experimental materials.

CPFD Samples and Extracts Preparation
In this study, citrus fruit were subjected to drying and grinding processes.The resulting material was sieved using a 60-mesh screen.Dry citrus fruit samples (1 g) were placed in beakers and mixed thoroughly with 10 mL of 80% methanol.The mixture was subjected to ultrasonic extraction at 35 • C and 40% power for 40 min, followed by centrifugation at 5000 rpm for 15 min.The supernatant was collected and stored at −20 • C for the determination of total phenols, total flavonoids, flavonoids, and antioxidant analysis.The freeze-dried powder of different CPFDs was stored in a cool, dry place until use.

Determinations of Phytochemical Composition
The determination of polyphenol content in citrus fruit samples was conducted using the Folin-Ciocalteu colorimetric method with gallic acid as the standard [41].Total flavonoid content was determined by the aluminum chloride-sodium nitrite colorimetric method with rutin as the standard [42].

In Vitro Antioxidant Activity Assays of Citrus Fruit Extract
The antioxidant capacity of citrus fruit extract was evaluated using in vitro antioxidant assays: DPPH radical scavenging assay, ABTS radical scavenging assay, and FRAP assay, following the manufacturer guidelines (Suzhou KeMing Biotechnology Co., Ltd., Suzhou, Jiangsu, China).Trolox, a known antioxidant compound, was used as the reference standard in these assays, and the results were expressed in terms of Trolox equivalent (TE) in µmol/mL.Firstly, all fruit extracts were diluted and used in the DPPH, ABTS, and FRAP assays.Subsequently, the absorbance values were used to calculate the antioxidant capacity based on the corresponding Trolox standard curve, and the results were expressed as µmol/mL TE.

Untargeted Metabolomic Analysis of CPFD
Metabolomic analysis of CPFD was conducted using ultra-high-performance liquid chromatography coupled with a quadrupole-Orbitrap hybrid mass spectrometry (UHPLC-QExactive) system (Thermo Fisher Scientific, Waltham, MA, USA).The CPFD powder was prepared by ultrasonication (5 • C, 40 kHz, 30 min) using an extraction solvent (methanol:water = 4:1, v:v).The resultant solution was filtered using a 0.22 µm membrane filter (Waters, Milford, MA, USA) before analysis.Liquid chromatography was conducted using the ACQUITY UPLC HSS T3 column (100 mm × 2.1 mm i.d., 1.8 µm) (Waters Corporation, Milford, MA, USA).The sample injection volume was set at 2 µL, and the column temperature was maintained at 40 • C. Mass spectrometric signals of the samples were acquired in positive and negative ionization modes.Subsequently, the obtained raw data were preprocessed using Progenesis QI 2.3 software (Nonlinear Dynamics, Waters, USA).The preprocessing steps primarily included filtering, deconvolution, alignment, and normalization.Database searches, conducted using KEGG (http://www.genome.jp/kegg/)(accessed on 7 May 2023), were used to summarize the differential metabolites between the two groups.Metabolite enrichment and pathway analysis were performed to map them onto their respective biochemical pathways.Data analysis was conducted on the online platform of Majorbio Cloud Platform (www.majorbio.com(accessed on 7 May 2023)).

Cell Culture
The hepatocellular carcinoma cell line HepG2 and murine fibroblast-derived preadipocytes (acquired from Shanghai Fuheng Biological Technology Co., Ltd.(Shanghai, China)) were cultured in DMEM medium supplemented with 10% FBS and 1% PS under a humidified atmosphere of 95% air and 5% CO 2 at 37 • C. The HepG2 cells were utilized to investigate the inhibitory effects of CPFD on hepatic lipid accumulation, while the 3T3-L1 cells were employed to evaluate the impact of CPFD on differentiated adipocytes.

Cell Viability Assay
HepG2 and 3T3-L1 cells were seeded into a 96-well plate and incubated for 24 h to assess the impacts of HYPD and DDPD on cell viability.Subsequently, the cells were fixed in MTT reagent, and the absorbance was measured at 490 nm using a microplate reader (Tecan, Männedorf, Switzerland).

OA-Induced Steatosis in HepG2 Cells
To detect the effects of CPFD extracts on OA-induced HepG2 cells, HepG2 cells were incubated with DMEM medium containing 0.5 mM OA and CPFD extracts for 24 h.

Differentiation of 3T3-L1 Cells
After seeding 3T3-L1 cells onto culture plates, the cells were cultured in CS medium until 100% confluence was reached.Subsequently, the induction of differentiation was initiated by culturing the cells for two days in 10% CS medium.Next, the cells were cultured for three days in a differentiation medium containing 10% FBS, 1% PS, 1 µM Dex, 10 µg/mL insulin, and 0.5 mM IBMX.Following this, the cells were further cultured for three days in a medium consisting of 10% FBS, 1% PS, and 10 µg/mL insulin.Finally, the cells were cultured for two days in a medium containing 10% FBS and 1% PS.CPFD extracts at different concentrations were simultaneously added to the culture medium at each time point.

Oil Red O Staining
ORO staining was used to assess the cellular morphology in HepG2 and 3T3-L1 cell lines.The staining procedure was conducted using the ORO staining kit (Solaibao Technology Co., Ltd., Beijing, China), following the manufacturer's guidelines.The cells were washed twice with PBS and fixed using the ORO fixative solution at room temperature.The fixative solution was vertically discarded, and 60% isopropanol solution was added to wash away residual fixative.After gentle agitation, the 60% isopropanol solution was discarded, and freshly prepared ORO working solution was added and incubated at room temperature in the dark for 30 min.The ORO working solution was discarded and rinsed vertically with distilled water 3-4 times to remove residual ORO working solution.Finally, the stained cells were visualized using an inverted microscope (Nikon, Tokyo, Japan).

Cellular Triglyceride Content
After drug treatment, HepG2 and 3T3-L1 cells were washed twice with ice-cold PBS, lysed with lysis buffer, and centrifuged at room temperature at a speed of 2000 rpm for 5 min.The supernatant was collected, and the determination of TG content was conducted strictly according to the instructions provided by the manufacturer of the TG assay kit (Beijing Pulai Gene Technology Co., Ltd., Beijing, China).The cellular protein content was determined using the Bicinchoninic acid protein assay (Shanghai Biyun Tian Biotechnology Co., Ltd., Shanghai, China).The results were expressed as the content of TG per gram of total cellular protein, in units of mmol/g protein.

Statistical Analysis
To compare differences among different samples, we conducted statistical analyses using SPSS 26.0 software.The data were presented as mean ± SD.One-way analysis of variance (ANOVA) was used for comparisons among multiple groups, followed by Tukey tests for analysis.A significance level of p < 0.05 was set for all statistical analyses to determine the statistical significance of the results.PCA and PLS-DA were carried out using the freely available Majorbio Cloud Platform (www.majorbio.com)(accessed on 7 May 2023).The heatmap was generated by the freely available Wekemo Bioinclound Platform (https://www.bioincloud.tech)(accessed on 30 November 2023).

Conclusions
In conclusion, our study elucidates that CPFDs exhibit a profusion of phytochemical constituents and excellent antioxidant activity.Moreover, we found significant differences in organic acids and lipids between HYPD and DDPD.Additionally, HYPD and DDPD

Figure 5 .
Figure 5. CPFD alleviated OA-induced steatosis in HepG2.(A) The effects of HYPD on cell viability in HepG2.(B) The effects of DDPD on cell viability in HepG2.(C) The effects of CPFDs on TG. (D) The

Figure 6 .
Figure 6.CPFD reduced lipid accumulation in 3T3-L1 adipocytes.(A) The effects of HYPD on cell viability in 3T3-L1.(B) The effects of DDPD on cell viability in 3T3-L1.(C) The effects of CPFDs on TG. (D) The morphology of stained intracellular lipid droplets by ORO staining at 200× magnification.Error bars are expressed as means ± SD.The significance levels were denoted as