Introduction

Citri Grandis Exocarpium (CGE, Huajuhong in Chinese) is the immature or nearly ripe, dried outer rind of Citrus grandis’ Tomentosa‘ or Citrus maxima (Burm.) Merr., which belongs to the Rutaceae family (Chinese Pharmacopoeia, 2020). It is a geographical indication product of China, primarily cultivated in Guangdong Province and Guangxi Province, with Huazhou, Guangdong Province, being the most prominent production area1. It has been widely used for thousands of years to treat inflammatory lung disorders such as asthma, cough and phlegm2. The description of CGE in “Supplements to Compendium of Materia Medica” suggests that the longer it undergoes aging process, the better its quality becomes. Likewise, Pericarpium Citri Reticulatae (Citrus reticulata ‘Chachi’, Chenpi in Chinese), another traditional Chinese medicine also belonging to the Rutaceae family, shares the characteristic of improving in quality with longer aging3. The process aging is accompanied by the growth of microorganisms3. The long-term synergistic evolution of endophytic fungi and plants directly or indirectly affects the content and distribution of active ingredients in medicinal plants by influencing or regulating the synthesis and accumulation of secondary metabolites in medicinal plants4. The changes of chemical composition and antioxidant activity of CGE during the storage (aging) process are currently not well understood. Therefore, analyzing the differences in CGE active components across various aging years, as well as identifying key pathways and differential metabolites, are of significant importance for a comprehensive evaluation of its medicinal and nutritional value.

The extracts from citrus genus plants, which are recognized for their dual role in medicine and food, exhibit strong antioxidant capabilities, which primarily attributed to their rich content of natural antioxidants, including flavonoids, phenolic acids, and coumarins, etc.5. These antioxidants can interrupt or reduce the oxidation of components sensitive to the oxidation process by clearing free radicals and reducing oxidative stress, and slow down the process of oxidative damage6. Oxidative damage is associated with various conditions, including the oxidation of some low-density lipoproteins, neurodegenerative diseases, metabolic diseases, and cancer7. Relevant research indicates that the most prominent flavonoid compound in CGE is naringin, which is considered a hallmark of the pharmacological effects related to cardiovascular diseases8. Naringin can significantly enhance the immune system’s effectiveness to avoid internal organs and tissue injury or disease caused by oxidation by increasing the activity of superoxide dismutase, catalase, paraoxonase, glutathione peroxidase and other antioxidant enzy5. Additionally, naringenin, naringin and neohesperidin in CGE relieved liver injury by inhibiting lipid accumulation and iron accumulation in nonalcoholic fatty liver disease (NAFLD)9. The total flavonoids of CGE had a positive therapeutic effect on PM2.5-induced lung injury. CGE extract (CGEE) was found to inhibit lipopolysaccharide-induced damage and nitric oxide release in RAW264.7 cells10. However, the metabolic markers affecting CGE antioxidant activity during aging remain unclear.

Metabolomics can provide important information about the number and types of metabolites present in plants and into changeable patterns in response to various factors such as environmental conditions11. It has been widely utilized to differentiate the compositional changes of medicinal plants at various storage stages, as well as to detect dynamic variations during the aging process, helping to identify effective chemical markers. For instance, it has been utilized to reveal the trends of volatile and non-volatile components in Artemisia argyi, as well as the differential biomarkers associated with varying aging periods12. Metabolomics has also been utilized to effectively distinguish between Citri Reticulatae Chachi Pericarpium (CRCP) and CRBP (Citri Reticulatae Blanco Pericarpium) with different storage years, and 19 metabolites were identified as potential markers for distinguishing between CRBP and CRCP13.

In this study, we employed a comprehensive metabolomics approach based on ultra-high-performance liquid chromatography-quadrupole/time of flight mass spectrometry (UHPLC-QTOF-MS/MS) to analyze the changes in chemical compositions of CGE samples aged for 1, 3 and 5 years. Principal component analysis (PCA), orthogonal partial least squares discriminant analysis (OPLS-DA) and other multivariate statistical methods were used to identify the marker metabolites that dominate the differences, explore potential rules, find key metabolic pathways, and analyze the correlation between differential metabolites, total flavonoids, and antioxidant activity in key pathways. The results will provide a scientific basis for understanding the pharmacological mechanisms of CGE and optimizing its medicinal value.

Results

Color changes in the appearance of CGE with different aging years

During the aging process of CGE, noticeable changes occurred in its appearance (Fig. 1). After one year of aging, the cross-section had a light brown color. As the aging progressed to three years, the outermost contour of CGE turned dark brown. This darkening continued, and by the fifth year of aging, the color of the external contour became even darker, approaching black. Simultaneously, the inner diameter of the light brown area within the cross-section gradually decreased as the aging progressed. These visual changes in CGE’s appearance during the aging process likely reflect alterations in its chemical composition. It is possible that the aging process leads to the degradation, transformation, or formation of new compounds within the CGE3.

Fig. 1: Color changes in the appearance of CGE with different aging years.
figure 1

CG1: CGE aging for 1 year, CG3: CGE aging for 3 years, CG5: CGE aging for 5 years.

Identification of metabolomic compounds

The ion-current diagram for the QC sample extraction represented by the extracted ion chromatogram (EIC) (Fig. 2A) demonstrated that all the target compounds exhibited symmetrical peaks, suggesting effective chromatographic separation of each unique target molecule. The total ion chromatogram (TIC) curve of the QC sample (Fig. 2B) displayed a significant overlap in peak retention time and peak area, suggested excellent stability of the instrument. The PCA-X one-dimensional distribution map (Fig. 2C) indicated that the QC samples were all positioned within ±2 standard deviations (STD), suggested that the data obtained from this experiment exhibited good quality.

Fig. 2: Identification of metabolomic compounds.
figure 2

A Extracted ion flow plots of QC samples. B TIC plots of all QC samples. C One-dimensional distribution of PCA-X of QC samples. D Ring diagram of compound classification statistics.

To measure the variations in metabolite content and metabolic pathways in ECG throughout various years of aging, we employed a comprehensive metabolomics detection technology. The analysis resulted in the detection of a total of 1249 metabolites (Fig. 2D). The substance-level classification showed 15 distinct groups, including 32 amino acid and derivatives (2.56%), 28 nucleotide and its derivatives (2.24%), 63 lipids (5.04%), 10 carbohydrates (0.80%), 8 vitamins (0.64%), 16 organic acids (1.28%), 49 phenylpropanoids (3.91%), 26 quinones (2.08%), 184 flavonoids (14.73%), 128 phenols (10.25%), 150 terpene (12.01%), 28 steroids and steroid derivatives (2.24%), 203 alkaloids (16.25%), 60 lignans and coumarins (4.80%), and 264 miscellaneous (21.14%).

Difference analysis of CGE with different aging years

PCA was conducted on the fingerprint data of CGE with different aging years. The principal components PC1 and PC2 of CG1 vs. CG3, CG1 vs. CG5, and CG3 vs. CG5 groups were 29.9% and 15.5%, 31.1% and 10.6%, and 24.4% and 15.9%, respectively (Fig. 3A, D, G). These values indicated that there were significant differences in the metabolites of different aged CGE. The proximity of certain data points in CG3 and CG5 indicated a high similarity in the types and relative contents of metabolites between these two sample groups. On the other hand, the CG1 group had a more dispersed distribution compared to the other two groups of samples, indicated that CG1 displayed the most substantial variations in total metabolic relative contents among the three groups. These findings indicated that the duration of aging had an impact on the metabolites present in CGE.

Fig. 3: Difference analysis of CGE with different aging years.
figure 3

Scores scatter plot of the PCA model with A CG1 vs. CG3, D CG1 vs. CG5, G CG3 vs. CG5; Scatter plot scores of the OPLS-DA model with B CG1 vs. CG3, E CG1 vs. CG5, H CG3 vs. CG5; In the scatter plot of PCA model, the horizontal value indicates the anticipated principal component score of the first principal component, whereas the vertical coordinate indicates the orthogonal principal component score. Permutation test of OPLS-DA model with C CG1 vs. CG3, F CG1 vs. CG5, I CG3 vs. CG5. In the permutation test of OPLS-DA model, the horizontal coordinate represents model accuracy, the vertical coordinate represents the frequency of occurrence of the model’s classification effect, and the R2Y and Q2 values represent the model’s interpretability of the Y variable and predictability, respectively. CG1: CGE aging for 1 year, CG3: CGE aging for 3 years, CG5: CGE aging for 5 years.

PCA analysis methods do not take into account within-group errors nor do they eliminate random errors11. To address these limitations and to further differentiate among the various groups, as well as to determine the compositional differences in CGE across different aging years, a supervised OPLS-DA analysis was employed (Fig. 3B, E, H). The score plots clearly demonstrated that samples from different aging year groups were effectively distinguished. Unlike the results obtained from principal component analysis (PCA), the OPLS-DA method demonstrated a more pronounced degree of aggregation within each group. Moreover, the CG3 and CG5 groups were distinctly differentiated from each another. This enhanced differentiation can be attributed to the fact that OPLS-DA, as a supervised analytical method, considered the grouping within samples, thereby offering improved discrimination11. The 7-fold cross-validation method was employed to evaluate the quality of the developed OPLS-DA model (Fig. 3C, F, I). The results showed that in the comparison group samples of CG1 vs. CG3, CG1 vs. CG5 and CG3 vs. CG5, the R2Y scores were all >0.99, and the Q2 values were all >0.85. Additionally, all p-values were found to be <0.05, suggested that the developed model effectively captured the differences among the aging time groups. This finding corroborated the response of differential metabolites to aging duration. The OPLS-DA model was further validated through 200 randomized permutation tests, as well as combined permutation tests.

Differential metabolite analysis

Based on the OPLS-DA results, metabolites were selected based on two criteria: VIP > 1 and P-value < 0.05. A total of 517 differential metabolites were screened (Table S1). In the CG1 vs. CG3 group, 166 different metabolites were identified, with 72 metabolites up-regulated and 94 metabolites down-regulated (Fig. 4A). In the CG1 vs. CG5 and CG3 vs. CG5 groups, 163 (42 metabolites up-regulated and 121 metabolites down-regulated) and 127 (46 metabolites up-regulated and 81 metabolites down-regulated) different metabolites were identified, respectively (Fig. 4B, C). A total of 57 differentially expressed metabolites were screened to distinguish CGE samples of different aging years (Fig. 4D), which belonged to various chemical classes, including four amino acids and derivatives (l-valine, l-ornithine, 3-hydroxy-3-methylpentane-1,5-dioic acid, and N-acetyl-l-phenylalanine); four nucleotides and their derivatives (deoxyguanosine, guanine, guanosine, and xanthosine); two lipids (16-hydroxy hexadecanoic acid and 8-methylnonenoate); two vitamins (pyridoxine and riboflavin); ten flavonoids (narirutin, diosmetin, tectorigenin, naringin (1), vestitol, glycinol, chrysoeriol 7-apiosylglucoside, pelargonidin-3-O-glucoside, peonidin-3-glucoside, and quercetin-3-O-sophoroside); eight phenols (gentisic acid, 2’,4’,6’-Trihydroxyacetophenone, myzodendrone, coniferin, benzoic acid, 4-hydroxy-3-methoxymandelate, vanillin, mulberrofuran Q, and 1,7-diphenyl-4-hepten-3-one); five phenylpropanoids (methyl cinnamate, benzyl cinnamate, C-veratroylglycol, angelicin, and cinnamyl cinnamate); five alkaloids (tombozine, piperidine, N-feruloyl agmatine, petasitenine, and N-feruloyl putrescine); three lignans and coumarins (bergamotine, acanthoside B, and esculin); one terpenoid (furanodiene); and thirteen miscellaneous metabolites [(±)-jasmonic acid, trans-zeatin-riboside, raffinose, 5-Thymidylic acid, 2-(methylamino)benzoic acid, palmitoylethanolamide, maltotriose, baldrinal, 2-hydroxy-3-(4-hydroxyphenyl)propenoic acid, N1-methyl-2-pyridone-5-carboxamide,N1-methyl-4-pyridone-3-carboxamide, isonicotinic acid, and benzocaine].

Fig. 4: Differential metabolite analysis of CGE with different aging years.
figure 4

A Volcano plot of CG1 vs. CG3. B Volcano plot of CG1 vs. CG5. C Volcano plot of CG3 vs. CG5. D Venn diagram of differential metabolites from CGEs in multiple pairwise comparisons of CG1 vs. CG3, CG1 vs. CG5, CG3 vs. CG5. E Cluster heat map of 57 differential metabolites. CG1: CGE aging for 1 year, CG3: CGE aging for 3 years, CG5: CGE aging for 5 years.

In order to visualize the variation patterns of metabolites’ relative contents, the raw relative contents of the main candidate metabolites in the three comparison groups CG1 vs. CG3, CG1 vs. CG5, and CG3 vs. CG5 were normalized (Fig. 4E). The results showed that the expression of 22 compounds, including peonidin-3-glucoside, tectorigenin, angelicin, vanillin, gentisic acid, and chrysoberyl 7-apiosylglucoside, was found to be up-regulated in the CG1 vs. CG3 group. Similarly, the expression of 31 compounds, such as naringin (1), diosmetin, peonidin-3-glucoside, tectorigenin, angelicin, and vanillin was up-regulated in the CG3 vs. CG5 group. In addition, a total of 8 potential markers, namely (±)-Jasmonic acid, diosmetin, peonidin-3-glucoside, tectorigenin, angelicin, vanillin, trans-zeatin-riboside, gentisic acid, showed the highest relative contents in CG5 compared to both CG1 and CG3, indicating that these metabolites accumulated with increasing aging time.

Differential metabolite pathway analysis

In order to gain insights into the main pathways involving the 517 mapped metabolites, pathway enrichment analysis was conducted using the KEGG database (Fig. 5A, C, E). The analysis focused on metabolites that showed significant expression differences in CGE samples with different aging years. In the CG1 vs. CG3 group, a total of 58 differential metabolite pathways were identified, which included plant hormone signal transduction (5 differential metabolites), flavonoid biosynthesis (8 differential metabolites), flavonoid biosynthesis (8 differential metabolites), nucleotide metabolism (7 differential metabolites), isoflavonoid biosynthesis (6 differential metabolites), biosynthesis of various alkaloids (6 differential metabolites), and ABC transporters (8 differential metabolites), among others (Fig. 5A). In the CG1 vs. CG5 group, 59 significantly enriched pathways for differential metabolites were observed. The main pathways included flavonoid biosynthesis (13 differential metabolites), isoflavonoid biosynthesis (9 differential metabolites), and nucleotide metabolism (8 differential metabolites), flavone and flavonol biosynthesis (7 differential metabolites), phenylpropanoid biosynthesis (7 differential metabolites), and plant hormone signal transduction (3 differential metabolites), among others (Fig. 5C). In the CG3 vs. CG5 group, 45 pathways were significantly enriched for differential metabolites, mainly flavone and flavonol biosynthesis (7 differential metabolites), flavonoid biosynthesis (7 differential metabolites), nucleotide metabolism (6 differential metabolites), isoflavonoid biosynthesis (5 differential metabolites), anthocyanin biosynthesis (5 differential metabolites), and ABC transporters (7 differential metabolites), among others (Fig. 5E).

Fig. 5: Differential metabolite pathway analysis.
figure 5

KEGG enrichment plot of differential metabolites with A CG1 vs. CG3, C CG1 vs. CG5, E CG3 vs. CG5. The horizontal axis and vertical axis denote the enrichment factor and the corresponding metabolic pathway name for each pathway, respectively. The color of the dot indicates the P-value obtained from the hypergeometric test; The color ranges from blue to red, with a deeper red indicating lower P-values. The size of the dots indicates the number of differential metabolites associated with each pathway. Pathway analysis diagram with B CG1 vs. CG3, D CG1 vs. CG5, F CG3 vs. CG5. Each bubble in the bubble map represents a differential metabolic pathway, with the horizontal coordinates indicating the location of each bubble and the size of the bubbles reflecting the magnitude of the impact factor associated with each pathway in the topological analysis; the vertical coordinates of the bubble represent its location, while the color of the bubble indicates the P-value derived from the enrichment analysis (calculated as the negative natural logarithm, i.e., −ln(p)). A darker color corresponds to a smaller P-value, reflecting a greater significance in terms of enrichment. CG1: CGE aging for 1 year, CG3: CGE aging for 3 years, CG5: CGE aging for 5 years.

KEGG annotation analysis only annotates the pathways involved in differential metabolism. To identify the key pathways with the greatest relevance to metabolite differences in CGE samples with different aging years, we conducted a thorough analysis of the pathways, including enrichment and topology analysis. The differential metabolites were compared to well-established metabolite databases such as KEGG and PubChem. After obtaining the matching information of differential metabolites, we utilized the Oryza sativa japonica (osa) pathway database to search for the corresponding species and conducted metabolic pathway analysis, and the results were represented by bubble map (Fig. 5B, D, F). In the CG1 vs. CG3, CG1 vs. CG5, and CG3 vs. CG5 groups, the bubbles representing the flavonoid biosynthesis route were found to be particularly prominent, characterized by their vivid red color and large size, indicated that this pathway had the most significant impact on the differential metabolites in CGE samples with different aging years. Therefore, the flavonoid biosynthesis pathway was concluded as the key pathway with the highest correlation with the observed differences in metabolites.

Kinetics of flavonoid biosynthesis pathway of CGE during the aging process

The contents of flavonoid metabolites in CGE changed significantly with increasing aging time (Fig. 6). 36 flavonoid metabolites were detected to be involved in the flavonoid biosynthesis pathway, including 14 differential metabolites. Specifically, during the aging process of 1–5 years, pinostrobin (10), (-)-epiafzelechin, galangin and cinnamic acid were up-regulated with the aging time, while 7’,4-dihydroxyflavone (4), chrysin (5), quercetin (7), pelargonidin (12), phloretin, gallocatechin, liquiritigenin, dihydromyricetin, hesperetin and pelargonidin were down-regulated. In addition, the relative contents of naringin (1), naringenin (2), butin (3), dihydrofisetin (8), luteolin (9), phlorizin (14), isosakuranetin, eriodictyol, garbanzol, xanthohumol, cyanidin, and butein decreased first and then increased during the aging process of 1–5 years. The relative contents of apigenin (6), homoeriodictyol (11), delphinidin (13), pinocembrin, tricetin, myricetin, desmethylxanthohumol, kaempferol, pinobanksin, isoliquiritigenin, and sakuranetin exhibited a trend of increased first followed by decreased. Previous study has also suggested that bacteria and fungi can induce changes in flavonoids during the aging process14.

Fig. 6: Dynamic changes of flavonoid biosynthetic pathways during the aging process.
figure 6

The columns from left to right indicate the CG1, CG3, and CG5, respectively. The red and blue colors represent up-regulated and down-regulated metabolites.CG1: CGE aging for 1 year, CG3: CGE aging for 3 years, CG5: CGE aging for 5 years.

Changes of total flavonoid and antioxidant capacity

The variations in total flavonoids and antioxidant capacity, as measured by DPPH, ABTS, and FRAP assays, are presented in Fig. 7. The results suggested that the total flavonoids contents (Fig. 7A) and antioxidant capacity (Fig. 7B–D) of CGEs exhibited the same change trend during the aging period from 1 year to 5 years, with a general trend of decreasing initially and then increasing, which was consistent with the report by Xiong15. Compared to CG1, the flavonoid content in CG3 has decreasing, which may be related to an increase in microbial diversity, and this change leads to a decrease in the levels of secondary metabolites, such as flavonoids, within the plant tissues16. Compared to CG3, the flavonoid content in CG5 has increased, which may be attributed to the enhanced metabolic activity of microorganisms that facilitates the accumulation and transformation of flavonoids. The resulting metabolites possess abundant hydroxyl groups in ring A, thereby enhancing their antioxidant activity17. These influencing factors need further investigation to enhance our understanding of the potential mechanisms underlying CGE, as well as its associated quality and health benefits.

Fig. 7: Total flavonoids and antioxidant capacity of different aging years of CGE.
figure 7

A Total flavonoid content, B DPPH free radical scavenging capacity, C ABTS cation scavenging capacity, and D FRAP reduction capacity. N = 3, different lowercase letters indicated statistical differences (P < 0.05) among three samples by Tukey’s test. CG1: CGE aging for 1 year, CG3: CGE aging for 3 years, CG5: CGE aging for 5 years.

Correlation analysis between flavonoids and antioxidant activity

The relevant network heatmap visually illustrates the correlation between 14 flavonoid metabolites in key metabolic pathways and antioxidant activity (Fig. 8A). The results showed significant positive correlations between total flavonoids, naringin (1), naringenin (2), butin (3), 7’,4-dihydroxyflavone (4), and phlorizin (14) with antioxidant capacity measured by DPPH (r ≥ 0.78, P < 0.05), ABTS (r ≥ 0.81, P < 0.01), and FRAP (r ≥ 0.72, P < 0.05). Apigenin (6), chrysin (5), pelargonidin (12), dihydro fisetin, and quercetin (7) exhibited a strong correlation with antioxidant activity. In contrast, pinocembrin (10), delphinidin (13), and homoeriodictyol (11) showed a weaker association with antioxidant activity. Notably, luteolin (9) demonstrated a stronger correlation with DPPH and FRAP assays, while a weaker relationship with the ABTS assay. It is important to highlight that different flavonoids could exhibit different antioxidant activities in different antioxidant models18. The antioxidant properties of flavonoids depend on their phenolic hydroxyl structure. Although all 14 flavonoids share a similar basic chemical structure, each compound is substituted with a different number and type of functional groups, resulting in different types of flavonoids (Fig. 8B). Naringin (1), 7’,4-dihydroxyflavone (4), chrysin (5), and pinocembrin (10) each contain two hydroxyl groups, while naringenin (2), butin (3), apigenin (6), homoeriodictyol (11), and phlorizin (14) each possess three hydroxyl groups. dihydrofisetin (8), luteolin (9), and pelargonidin (12) are characterized by the presence of four hydroxyl groups, whereas quercetin (7) contains five hydroxyl groups. Delphinidin (13) is notable for having six hydroxyl groups. Additionally, naringin (1), homoeriodictyol (11), and phlorizin (14) not only include hydroxyl groups but also feature other types of functional groups. In these flavonoid compounds, there are differences in the positions of functional groups on the carbon rings. The variation in antioxidant activity among the different flavonoids can be attributed to the number of phenolic hydroxyl O–H groups and the kind and position of substituents on the A-, B-, and C-rings of the flavonoid structures19. These structural differences can influence the ability of flavonoids to scavenge free radicals and exert antioxidant effects.

Fig. 8: Correlation analysis between flavonoids and antioxidant activity.
figure 8

A Correlation network heat map between antioxidant activity and 14 differential metabolites of flavonoids. The red curve represents a positive connection, whereas the blue curve represents a negative correlation. The solid line of the P-value represents a statistically significant correlation, whereas the dotted line represents a connection that is not statistically significant. The color of the heatmap and the size of the square represent the respective correlation coefficients. The intensity of the color and the size of the box directly correspond to the magnitude of the correlation coefficient. The larger the correlation coefficient, the redder the color and the larger the box; Conversely, The smaller the correlation coefficient, the bluer the color and the smaller the box. B Chemical structural formulae for flavonoids. Solid triangular carbon implies that the bond is above the plane of the carbon ring, whereas imaginary triangular carbon suggests that the bond is below it.

Discussion

Comprehensive metabolomics has proven to be valuable in enhancing the reliability and accuracy of metabolite identification11. In this study, metabolites present in CGE under different aging years were characterized using UHPLC-QTOF-MS/MS. The results showed inconsistent trends in the changes of CGE metabolites across different aging periods. This high-throughput detection technique enabled the identification of 1249 differential metabolites in CGE with different aging years. Among these differential metabolites, 57 were identified as potential markers that could be used to distinguish CGEs from different aging years, included ten flavonoids, namely naringin (1), narirutin, diosmetin, tectorigenin, vestitol, glycinol, chrysoeriol 7-apiosylglucoside, pelargonidin-3-O-glucoside, peonidin-3-glucoside, and quercetin-3-O-sophoroside. Flavonoids are considered the main active components of CGE20. Naringin (1), in particular, is the main chemical component used to assess the quality standard of CGE21.

In addition, metabolic pathway analysis indicates that flavonoid biosynthesis was the key metabolic pathway in CGE with different aging years, involving 14 flavonoids differential metabolites. Naringin (1) and naringenin (2), which are metabolites of this pathway, are exactly the main active ingredients of CGE. These findings align with previous metabolomics studies on different strains of CGE22. During the aging process from CG1 to CG5, the content of flavonoids in CGE exhibited a fluctuating trend. This variation may be attributed to microbial metabolism, which produces enzymes such as α-l-rhamnosidase and β-d-glucosidase. These enzymes possess the ability to hydrolyze flavonoid glycosides into aglycones and monosaccharides, thereby enhancing the bioavailability of flavonoids in medicinal plants and their potential health benefits3. Furthermore, the enzymes produced during the growth and metabolic processes of fungi facilitate the degradation of plant cell walls, thereby aiding in the solubilization of flavonoids3. These enzymes include glucosidase, cellulase, chitinase, and xylanase, which hydrolyze glycosidic compounds and compromise the integrity of plant cell walls23. This process enhances the release of flavonoids present in medicinal plants. In the different aging stages of CGE, CG1 exhibited the strongest antioxidant activity, followed by CG5, while CG3 demonstrates relatively weaker antioxidant properties. However, this ranking of antioxidant activity does not necessarily related to other pharmacological effects such as anti-inflammatory, cough and phlegm reduction. Therefore, further investigation is required to explore the additional potential benefits of long-term aged CGE. Correlation analysis indicated that a strong correlation among antioxidant activities (FRAP, ABTS, DPPH), total flavonoids, and 14 different flavonoid metabolites, including naringin (1), naringenin (2), pelargonidin (12), butin (3), apigenin (6), homoeriodictyol (11), and others. The antioxidant capacity of flavonoids is primarily determined by the presence of phenolic hydroxyl groups, particularly on the B ring. The number of phenolic hydroxyl groups, as well as the structural arrangement of the A, B, and C rings can influence the antioxidant activity of flavonoids19.

Flavonoids are well-known for their antioxidant properties and associated with health benefits24. In the CGE, naringin (1), naringenin (2), butin (3), 7’,4-dihydroxyflavone (4), and phlorizin (14) have been identified as potential metabolites associated with antioxidant activity. Contemporary medical studies have shown that these compounds exhibit possess various biological properties with anti-inflammatory, antiviral, and antioxidant effects. For example, naringin (1) effectively reduces excessive inflammatory damage in the lungs induced by Klebsiella pneumoniae (Kpn) and enhances the body’s ability to clear bacteria25. Naringenin (2) have been identified as potent inhibitors of the main protease of COVID-1926,27. Butin (3), recognized for its antioxidant and anti-inflammatory properties, has demonstrated potential in mitigating arthritis in rat models28. The flavonoid 7’,4-dihydroxyflavone (4) inhibits the expression, production, and secretion of MUC5AC through the regulation of NF-κB, STAT6, and HDAC229. The CGE is a traditional Chinese medicine that can be stored for a relative long time, and its aging process is often accompanied by the growth of microorganisms. The variation in microbial community structure is closely related to storage conditions (such as temperature and humidity), aging time, and active components. The interplay among microbial succession, aging time, and active components in CGE is complex and needs further investigation.

In summary, the flavonoid biosynthetic pathway with the highest correlation with differential metabolites, and involved 14 differential metabolites, including naringin (1) and naringenin (2), which were the main active components of CGE. In the different aging stages of CGE, CG1 exhibited the strongest antioxidant activity, followed by CG5, while CG3 demonstrates relatively weaker antioxidant properties. However, this ranking of antioxidant activity does not necessarily related to other pharmacological effects such as anti-inflammatory, cough and phlegm reduction. Therefore, further investigation is required to explore the additional potential benefits of long-term aged CGE. The correlation analysis revealed significant and positive associations between antioxidant activity and total flavonoids, as well as specific metabolites such as naringin (1), naringenin (2), butin (3), 7’,4-dihydroxyflavone (4), and phlorizin (14). This provides valuable insights into the aging mechanisms of CGE and supports the concept of its health benefits.

Materials and methods

Plant materials

Three CGE samples were obtained from the same orchard in Huazhou City, Guangdong Province, China (latitude ranging from 21.29°N to 22.13°N and longitude spanning from 110.21°E to 110.45°E). In the production base, six fruit trees were selected from each of the five sampling points located in the east, south, west, north, and center. On average, 1 kg of fruit was harvested from each tree, resulting in a total of 30 kg. For fresh fruit (young fruit) sampling, fruits with uniform shape, moderate size, and smooth surfaces free from insect damage were chosen. The processing flow of fresh fruit (young fruit) is as follows: harvesting–washing–high-temperature drying (80–90 °C)–fruit pressing–slicing–drying–storage. The samples were collected in May of 2023, 2021, and 2019, respectively, aging for 1 year (CG1), 3 years (CG3), and 5 years (CG5). To maintain consistency, the cultivation methods (fertilization, etc.) and storage conditions for these samples were identical, with all samples being naturally preserved in jute bags at room temperature under the same environmental conditions. This allowed for a natural aging process to occur over the specified durations. All samples were authenticated by Qiongqing Chen from Guangdong Three Cloud Network Technology Co., LTD.

Metabolites extraction

10 mg of the CGE sample was subjected to dispersion in 0.5 mL of an extraction solution (methanol/water = 3:1), which had been pre-cooled to −40 °C and contained an internal standard of 2-chloro-dl-phenylalanine. The samples were vortexed for 30 s, homogenized at a frequency of 40 Hz for 4 min, and subsequently extracted for metabolites using low-temperature sonication in an ice-water bath for 5 min. Following this, centrifugation was performed at 12,000×g for 15 min. This cycle of homogenization, sonication, and centrifugation was repeated three times, after which the supernatant was collected for UHPLC-MS analysis.

UHPLC-QTOF-MS/MS analysis

This project employs the EXION LC System (SCIEX) ultra-high-performance liquid chromatography (UHPLC) instrument, utilizing Waters UPLC columns (1.8 μm, 2.1 mm × 100 mm i.d.) for the chromatographic separation of target compounds. The mobile phase consisted of mobile phase A (water) containing 0.1% formic acid and mobile phase B (acetonitrile). The sample injection amount was 2 μL, and the column temperature was set at 40 °C. The elution gradient of the mobile phase with a flow rate of 0.4 mL/min was set as follows: 0–0.5 min, 98% A; 0.5–10 min, 98–50% A; 10–11 min, 50–5% A; 11–13 min, 5–5% A; 13–13.1 min, 5–98% A; 13.1–15 min, 98% A. The mass spectrometric analysis was conducted using the SCIEX 6500 QTRAP+ triple quadrupole mass spectrometer equipped with the IonDrive Turbo V ESI ion source, employing multiple reaction monitoring (MRM) modes. The parameters for the ion source are as follows: IonSpray Voltage was set to +5500/−4500 V, declustering potential (DP) was set at ±100 V, the temperature of the ion source was set to 400 °C, Curtain Gas pressure was kept at 35 psi, Ion Source Gas 1 was set to 60 psi and Ion Source Gas 2 was set to 60 psi.

Mass spectrometry data and statistical analysis

SCIEX Analyst Work Station Software (Version 1.6.3) was employed for MRM data acquisition and processing. After obtaining the metabolite spectrum analysis data from different samples, the peak area of the mass spectra peaks was integrated for all substances, and the integration was corrected for the mass spectra of the same metabolite in different samples among them11. The metabolites were qualitatively and quantitatively analyzed and annotated by mass spectrometry using a self-built database (San Shu Biotechnology Co. Ltd., Jiangsu, China) and the public database of metabolite information30.

Total flavonoid and antioxidant capacity detection

According to our previous reporting methods30, the following indicators were assessed. The total flavonoid concentration in the CGE extract was determined using a colorimetric method. The absorbance of the supernatant was measured at 510 nm, and the total flavonoid content was expressed as rutin equivalents per gram of dry weight of the ECG sample (mg RE/g). The antioxidant activity of the CGE extracts was assessed by measuring its ability to scavenge free radicals, specifically 2,2-Diphenyl-1-picrylhydrazyl radical (DPPH) 2,2-Azinobis-3-ethylbenzothiazoline 6-sulfonate free radical (ABTS+) and ferric reducing antioxidant power (FRAP). Absorbance readings were taken at wavelengths of 517, 734, and 700 nm respectively. The results were expressed as milligrams of Trolox equivalents per gram of dry weight of the CGE sample (mg TE/g).

Statistical analysis

The metabolomic data were subjected to multivariate analysis, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA), using the SIMCA 16.0.2 software package (Stedim Data Analytics AB, Umea, Sweden). The value of variable importance in the projection (VIP) of the first principal component in OPLS-DA analysis was obtained. It summarizes the contribution of each variable to the model. The metabolites with VIP > 1 and p < 0.05 (Student’s t-test) were considered as significantly changed metabolites. Additionally, pathway enrichment analysis was conducted utilizing the KEGG database (http://www.genome.jp/kegg/) and the MetaboAnalyst platform (http://www.metaboanalyst.ca/). Peak areas were normalized to assess the relative content of metabolites, and heatmap analysis was performed using the OmicStudio tool (https://www.omicstudio.cn). Correlation analysis between relative metabolite content and antioxidant activity was also performed using the OmicStudio tool. Six biological duplicates were used for all tests. The significance analysis was conducted using the SPSS 21.0 program.