Skip to main content
NCBI home page
Foods logoLink to Foods
. 2026 Mar 4;15(5):865. doi: 10.3390/foods15050865

Unraveling Variations in Primary Metabolites of Longjing Green Tea During Processing and from Different Geographical Origins

Zhiyuan Lin 1,2,3,†, Mei Chen 1,†, Bo Zhou 4, Junfeng Tan 1, Liang Zeng 2, Zhi Lin 1, Jinchi Tang 4,*, Weidong Dai 1,*
Editor: Jong-Sang Kim
PMCID: PMC12984178  PMID: 41829138

Abstract

(1) Background: Primary metabolism is essential for tea quality formation, however systematic analysis of tea primary metabolites remains limited; (2) Methods: An orthogonal design was used to optimize the pre-column derivatization reaction conditions for the tea matrix. Using the optimized method, gas chromatography–mass spectrometry (GC–MS) was employed to analyze Longjing green tea (LJGT) samples from different processing stages and regions; (3) Results: Optimal derivatization was achieved with 75 μL of methoxamine hydrochloride pyridine solution at 30 °C for 1.5 h. A total of 52 primary metabolites were identified in LJGT. Processing analysis showed that 8 metabolites associated with carbohydrate metabolism significantly decreased during spreading. Five reducing sugars significantly decreased, while sucrose, turanose, and quinic acid significantly increased due to the thermal action during pan-fixation. Additionally, 15 key differential compounds were identified among three regions (Xihu, Qiantang, and Yuezhou) of LJGT. Quantitative analysis revealed that shikimic and quinic acid contents were significantly higher in the Xihu region compared to other regions; (4) Conclusions: This study established a pre-column derivatization GC-MS method for primary metabolite profiling, elucidated metabolic regulation during LJGT processing, and identified differences in primary metabolite content among LJGT from different geographical origins.

Keywords: longjing green tea, primary metabolites, derivatization, GC-MS, geographical origin, tea processing

1. Introduction

Primary metabolites are the fundamental basis for plant survival, growth, and development [1]. As the central hub of cellular metabolism, they provide energy currencies and carbon skeletons for physiological processes and serve as direct precursors of secondary metabolites such as polyphenols, alkaloids, and terpenoids [2,3]. Beyond their physiological functions, primary metabolites play key roles in shaping tea sensory attributes. For instance, soluble sugars contribute significantly to the sensory profile by modulating the balance of bitterness and astringency [4], while reducing sugars interact with amino acids during processing to generate volatile aroma compounds [5]. Furthermore, the biosynthesis and accumulation of primary metabolites are highly sensitive to environmental factors, this environmental dependence makes them promising chemical markers for the geographical authentication of tea [6,7].

Longjing green tea (LJGT), one of China’s most prestigious Geographical Indication (GI) products, is highly acclaimed for its roasted bean-like aroma and fresh, mellow taste [8]. Although extensive chemical characterization has been conducted to elucidate the key compounds defining LJGT quality, most studies have focused on secondary metabolites. Regarding aroma, key roasted notes such as 3-ethyl-2,5-dimethylpyrazine and 2,3-diethyl-5-methylpyrazine have been confirmed to originate primarily from the Maillard reaction between amino acids and soluble sugars during roasting [9,10]. In terms of taste, glutamate, a critical contributor to umami, has been shown to increase significantly during spreading due to protein hydrolysis [11]. Furthermore, factors such as cultivar, geographical origin, and processing conditions profoundly influence LJGT quality [12,13]. For instance, the degree of spreading determines the taste and color of the final tea infusion by modulating the metabolic transformation of organic acids and glycosides [14]. High-altitude environments enhance the sweet and mellow flavor profile by upregulating metabolism of flavonoids, amino acids, and carbohydrates [15]. Crucially, the occurrence of these critical biochemical reactions depends directly on the supply and transformation of primary metabolites. However, despite their central role as substrates, systematic reports on the primary metabolite profile of LJGT remain scarce. This gap limits our ability to fully understand the molecular mechanisms underlying quality formation in LJGT from a metabolic flux perspective.

Analysis of primary metabolites is challenging because of their high polarity, low volatility, and lack of suitable chromophores, which hinder direct detection by conventional liquid chromatography with ultraviolet detection (LC-UV) or gas chromatography (GC) [16,17]. Gas chromatography-mass spectrometry (GC-MS) combined with derivatization has emerged as an effective strategy to overcome these obstacles. Specifically, a two-step derivatization approach is commonly employed: first, carbonyl groups (e.g., aldehydes, ketones) are stabilized via oximation using reagents such as methoxyamine hydrochloride to form methoxime derivatives, thereby preventing ketone isomerization and improving chromatographic peak shape; subsequently, silylation reagents such as N-methyl-N-(trimethylsilyl) trifluoroacetamide are used to convert polar functional groups (e.g., -OH, -COOH, -NH2) into volatile and thermally stable trimethylsilyl derivatives for enhanced volatility and detection [18,19]. However, the derivatization efficiency is highly sensitive to reaction conditions and matrix effects, which can compromise the accuracy and reproducibility of quantitative analysis in complex matrices such as tea.

In this study, we systematically optimized methoxamine hydrochloridederivatization conditions to minimize matrix interference and enhance detection sensitivity. Using this optimized method, we analyzed LJGT samples from different processing stages and regions for their primary metabolites. This work offers new insights into the metabolic mechanisms that drive the formation of quality and the geographical differentiation of green tea.

2. Materials and Methods

2.1. Chemicals and Reagents

Citric acid, quinic acid, galactose, glucose, and sucrose were purchased from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Oxalic acid, inositol, and maltose were purchased from Taizejiaye Technology Development Co., Ltd. (Beijing, China). Shikimic acid and gallic acid were purchased from Puyihua Science and Technology Co., Ltd. (Beijing, China). Malic acid and phosphoric acid were purchased from Shuopu Biotechnology Co., Ltd. (Guangzhou, China) and Qichuang Scientific Equipment Co., Ltd. (Hangzhou, China), respectively. All standards met chromatographic purity ≥ 95%. Pyridine, methoxyamine hydrochloride, and N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA) of chromatographic grade were purchased from Sigma-Aldrich (St. Louis, MO, USA). Deionized water was produced by a Milli-Q water purification system (Millipore, Billerica, MA, USA).

2.2. Tea Samples and Preparation

To trace metabolic dynamics during manufacturing, a controlled processing experiment was conducted. Fresh leaves (FL) from the Longjing 43 tea plant, harvested as one bud and two leaves in early April, were collected from tea gardens in Xihu District, Hangzhou City, Zhejiang Province. As shown in Figure 1, LJGT is processed through three stages: spreading, pan-fixation, and pan-drying in accordance with GB/T 18650-2008 [20]. Samples were collected immediately after each stage and designated as spreading leaves (SL), pan-fixation leaves (PL), and finished tea (FT). All samples were cryopreserved in liquid nitrogen, freeze-dried under vacuum, ground into fine powder using a tube mill (IKA, Staufen, Germany), and stored at −20 °C. In addition, 49 commercially available LJGT samples were collected from three geographical origins: Xihu, Qiantang, and Yuezhou. These samples, all produced from fresh leaves plucked in the spring (from mid-March to early April) and comprising single buds or one bud with one newly unfolded leaf, were either provided by institutions such as the Tea Research Institute of the Chinese Academy of Agricultural Sciences and the Chinese Tea Society or independently procured. Detailed information on the samples is presented in Table S1.

Figure 1.

Figure 1

Open in a new tab

Diagram of the process of Longjing green tea (LJGT) processing.

For sample preparation, 40 mg of tea powder was mixed with 2 mL of boiling water. The mixture was homogenized and extracted for 5 min. The supernatant was then separated by centrifugation (Eppendorf 5810R, Hamburg, Germany) at 9000 rpm for 10 min. A 0.5 mL aliquot of the supernatant was lyophilized to remove all moisture. Quality control (QC) samples were prepared by mixing equal amounts of each sample.

2.3. Optimization of Pre-Column Derivatization Conditions for Primary Metabolites in LJGT

To optimize the derivatization conditions for the tea matrix, an L9 orthogonal array design with three factors at three levels was used based on the previous literature [21,22]. The factors investigated included the methoxamine hydrochloride pyridine solution (MOX) volume (A), derivatization temperature (B), and reaction time (C). The levels of each factor are shown in Table S2, and the orthogonal design table is shown in Table S3.

To the freeze-dried QC sample, add corresponding volume of a 20 mg/mL methoxamine hydrochloride pyridine solution, vortex for 1 min, and derivatize at the specified temperature and time according to the orthogonal table. Then add the 60 μL of MSTFA, derivatize at 39 °C for 30 min Finally, centrifuge the sample at 9000 rpm for 10 min, and collect the supernatant for GC-MS analysis. Optimal derivatization conditions were selected using the total response area in chromatograms as the evaluation criterion.

2.4. GC–MS Analysis

A 7890B gas chromatography system coupled with a 7000C mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) was used for primary metabolite analysis. Chromatographic separation was performed on an HP-5MS capillary column (30 m × 0.25 mm × 0.25 μm, Agilent, Santa Clara, CA, USA). The injection volume was 1 μL. The column temperature program was as follows: initial temperature held at 60 °C for 3 min, increased to 290 °C at 7 °C/min, and held at 290 °C for 5 min. The inlet temperature was maintained at 280 °C with a split ratio of 10:1. High-purity helium (>99.999%) was used as the carrier gas at a constant flow rate of 1 mL/min in constant flow mode. Mass spectrometry was conducted in electron impact (EI) mode at 70 eV. The ion source and transfer line temperatures were set to 220 °C. Data were acquired in full scan mode with a mass range of 33–600 m/z. A solvent delay of 6.3 min was applied.

2.5. Identification and Quantification of Compounds

Compound identification was performed using the Mass Hunter Workstation Software Unknowns Analysis (Version 10.1). Compounds were screened based on mass spectra matches to NIST11 library entries, with a positive or negative correlation score exceeding 600. Selected compounds were validated using chemical reference standards. The absolute concentrations of 6 organic acids and 4 soluble sugars were determined using linear calibration curves, which were established by plotting peak areas against concentrations of mixed standard solutions ranging from 0.5 to 125 μg/mL.

2.6. Methodology Validation

A series of concentration gradients of 10 standards were prepared by diluting the stock solutions, and then the linearity of each standard was examined and expressed through regression coefficients (R2). The limit of quantitation (LOQ) was defined as the lowest standard solution concentration that produced a signal-to-noise (S/N) ratio of 10. Intra-day precision was assessed by performing six independent derivatization and analytical replicates of the standard solution within a single day. Inter-day precision was evaluated by analyzing the standard solution on three consecutive days, with one complete derivatization and analysis performed each day. Precision was expressed as the relative standard deviation (%RSD).

2.7. Statistical Analysis

Partial least squares discriminant analysis (PLS-DA) was performed using SIMCA 14.1 (Umetrics Corp., Umeå, Sweden). The data were scaled using Unit Variance (UV) scaling in the SIMCA software prior to PLS-DA modeling. Key differential compounds were then selected based on a variable importance in projection (VIP) score greater than 1 and p < 0.05 [23]. One-way analysis of variance (ANOVA), range analysis, independent samples t-test, and the Kruskal–Wallis test were performed using SPSS Statistics 20.0 (IBM Corp., Armonk, NY, USA). To control the family-wise error rate, Tukey’s HSD test was used following ANOVA, and Dunn’s test was applied after the Kruskal–Wallis test. Pathway analysis was performed using the online tool MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/, accessed on 26 January 2026) [24]. The Arabidopsis thaliana reference species library was selected for functional mapping and visualization of the metabolic pathways. Heatmap visualization and hierarchical cluster analysis (HCA) were conducted using the CNSknowall platform (https://cnsknowall.com, accessed on 17 January 2026). Bar graphs and violin plots of the results were generated using GraphPad Prism 9.0 (GraphPad software, San Diego, CA, USA).

3. Results and Discussion

3.1. Optimized Conditions and Analytical Performance of the Derivatization Method

As shown in Table S4, analysis of variance (ANOVA) revealed that MOXvolume, derivatization temperature, and reaction time had significant effects on the total response area (p < 0.01), indicating the need to optimize these factors. To determine the optimal combination of reaction conditions, an orthogonal experimental design was used, and the results were visually analyzed. As shown in Figure 2a and Table S5, the highest k-values are observed with an MOXvolume of 75 μL (A3), a derivatization temperature of 30 °C (B1), and a reaction time of 1.5 h (C2). Analysis of the R-value indicates that MOXvolume and derivatization temperature are the two factors that exert the most significant influence on the total response area (Figure 2b and Table S5). Under the optimized conditions, derivatization of QC samples resulted in a significant increase in the number of detectable peaks in the chromatograms (p < 0.05) and improvements in both total response area and reproducibility compared with the pre-optimization method (Figure 2c–e).

Figure 2.

Figure 2

Open in a new tab

Optimization and validation of pre-column derivatization conditions using orthogonal experiments. (a) k-values reflecting the average effects of factor levels on the total response area; (b) R-values indicating the influence of factors by range analysis; (c–e) comparison of the total response area (c), total peak number (d), and repeatability of compounds (e) between unoptimized and optimized derivatization conditions. Note: Significance determined by independent samples t-test, * p < 0.05.

3.2. Methodology Validation Results of the Optimized Pre-Column Derivatization GC-MS Method

The optimized derivatization GC-MS method was systematically validated and evaluated for performance (Table 1). The results showed excellent linearity, with regression coefficients (R2) ranging from 0.9907 to 0.9971 and a linear range spanning 2–3 orders of magnitude. The LOQ was within the range of 0.5–30 μg/mL. The inter-day and intra-day precision ranged from 1.47% to 9.3% and from 4.4% to 12.98%, respectively. These results indicate that the newly developed rapid metabolomics method is highly reliable and robust, making it suitable for high-throughput sample analysis.

Table 1.

Linear range, R2, LOQ, and repeatability of 10 primary metabolites in tea using optimized pre-column derivatization GC-MS method.

No. Compound Standard Curve Linear Range
(μg/mL)		 R2 LOQ Repeatability (%)
Intra-Day Inter-Day
1 Malic acid Y = 206,087X – 193,758 2–90 0.9953 2 4 9.95
2 Shikimic acid Y = 1,404,325X – 499,300 0.5–90 0.9953 0.5 2.33 11.98
3 Galactose Y = 1,912,872X – 402,366 1–60 0.9912 1 3 12.98
4 Glucose Y = 822,253X + 139,750 1–40 0.9928 1 2.39 8.67
5 Gallic acid Y = 336,120X – 96,173 2–120 0.9952 2 5.45 9.45
6 Oxalic acid Y = 31,117X – 14,855 30–125 0.9926 30 4.97 11.26
7 Citric acid Y = 22,640X – 58,475 5–100 0.9917 5 9.30 4.71
8 Quinic acid Y = 1,427,722X – 1,374,798 1–40 0.9936 1 7.13 5.35
9 Sucrose Y = 288,063X + 63,646 0.5–60 0.9907 0.5 1.47 4.40
10 Fructose Y = 615,350X – 73,612 0.5–100 0.9971 0.5 1.53 11.37
Open in a new tab

3.3. Dynamic Changes of Primary Metabolite During LJGT Processing

To investigate the dynamic changes in primary metabolites during LJGT processing, this study employed non-targeted metabolomics analysis based on an optimized derivatized GC-MS method for tea samples at different processing stages. After preprocessing the raw data through feature ion extraction, peak detection, and peak alignment, a total of 379 feature ions were obtained (Table S6), with all detected features retained for constructing the PLS-DA model without further filtering to ensure a comprehensive metabolic profile. As shown in Figure 3a, the PLS-DA score plot illustrates the distribution of samples across different processing stages. As processing progressed (from FL to FT), sample points migrated from the negative to the positive half-axis along the first latent variable, indicating gradual alterations in the primary metabolite profile during processing. Notably, the spatial span of sample points was greatest between the FL and WL stages and between WL and PL stages, suggesting that the spreading and high-temperature pan-fixation processes exerted the most significant influence on the composition of primary metabolites. Furthermore, QC samples clustered tightly around the model center, demonstrating the stability of the analytical system and the good reproducibility of the data. The model parameters (R2Y = 0.664, Q2 = 0.442) indicate that the model possesses good explanatory and predictive capabilities. The permutation test results (n = 200) in Figure 3b shows an R2 intercept of 0.145 and a Q2 intercept of −0.457, confirming the absence of overfitting and the reliability of the statistical results.

Figure 3.

Figure 3

Open in a new tab

Dynamic changes of primary metabolites during LJGT processing. (a) PLS-DA score plot illustrating the separation of samples across processing stages; (b) permutation test result (n = 200) for validating the PLS-DA model; (c) heatmap showing the relative abundance of key primary metabolites that varied during processing, a bold font indicates that the compound has been certified as a standard; (d) pathway analysis of the effect of processing on LJGT based on pathway-matched metabolites. Note: FL, fresh leaves; SL, spreading leaves; PL, pan-fixation leaves; FT, finished tea.

Through comparison with the NIST database and verification against reference standards, a total of 52 metabolites were identified, comprising 14 carbohydrates, 17 carbohydrate derivatives, 13 organic acids, 4 amino acids, and 4 other compounds (Table S6). Among these, 16 compounds exhibited VIP values greater than 1 with p < 0.05 (one-way ANOVA), thereby being selected as key differential compounds in the LJGT processing process. As shown in the heatmap clustering analysis in Figure 3c, 12 compounds exhibited significant decreases in level during processing. Among these, 7 compounds, including fructose, glycerol, and galactose, showed a sustained and substantial decline trend during both the spreading and pan-fixation stages, with average decreases of 65.1% and 67.3%, respectively. Lactose, gallic acid, xylose, and inositol increased during the spreading stage with an average increase of 16.3%, followed by a decrease during the pan-fixation stage with an average decrease of 72.4%. In addition, 2-ketoglutaric acid decreased by 61.1% during the spreading stage and increased by 90.1% during the pan-fixation stage. Conversely, four compounds exhibited significant accumulation during processing. Quinic acid showed an upward trend across all stages, increasing by 8.3% during the spreading stage and 7.8% during the pan-fixation stage. Sucrose and turanose decreased during the spreading stage, with an average reduction of 76.4%, followed by a significant increase during the pan-fixation, averaging a substantial rise of 721.0%. Galactinol showed a marked increase during the pan-drying stage, rising by approximately 27.3%. The results of pathway analysis showed that the above key differential compounds were closely related to carbohydrate metabolism, especially galactose metabolism, starch and sucrose metabolism, citrate cycle (TCA cycle), and glyoxylate and dicarboxylate metabolism (Figure 3d).

During the spreading process of fresh tea leaves, slow dehydration, respiration, and metabolic conversion occur simultaneously [25]. On one hand, compounds such as sucrose, glucose, fructose, and glycerol, which serve as respiratory substrates or biosynthetic precursors, are extensively consumed. As shown in Figure 4a, the levels of metabolites associated with the TCA cycle, such as succinic acid, malic acid, and 2-ketoglutaric acid, significantly decreased (p < 0.05). Similarly, the levels of metabolites related to galactose metabolism and sucrose metabolism, including glucose, sucrose, galactose, and trehalose, also significantly decreased (p < 0.05). This result contrasts with reports of upregulation in galactose metabolism during black tea withering [26], potentially due to more thorough moisture loss in black tea withering. On the other hand, cellular dehydration induces specific biochemical responses or secondary metabolic processes. For instance, the observed upward trend in the level of xylose (Figure 3c) may indicate hydrolysis of flavonol glycosides yielding soluble sugar [27].

Figure 4.

Figure 4

Open in a new tab

Changes in primary metabolites of LJGT during the spreading and drying processes. (a) Down-regulation of metabolites associated with starch and sucrose metabolism (green), galactose metabolism (purple), TCA cycle (red), and glyoxylate and dicarboxylate metabolism (blue) during spreading; (b) reduction of reducing sugars during pan-drying; (c) increased levels of sucrose and turanose during pan-drying. Note: different metabolic pathways are highlighted in distinct colors. All abbreviated enzyme names are listed in Table S7. Significance determined by one-way ANOVA with Tukey’s HSD test, * p < 0.05; ** p < 0.01; and *** p < 0.001.

Upon entering the fixation stage, rapid increases in leaf temperature cause enzyme denaturation and inactivation, while some heat-sensitive compounds undergo degradation or transformation. For instance, reducing sugars such as glucose, fructose, xylose, and ribose continuously decrease under thermal action (Figure 4b). These reducing sugars readily undergo Maillard reactions with amino acids during high-temperature processing [28], forming compounds like pyrazines and pyrroles that contribute to roasted aromas. This process plays a crucial role in developing the “bean aroma” characteristic of LJGT. Notably, the significant increase in sucrose level, with the PL level rising to 10.3 times that of SL (Figure 4c), suggests that thermal decomposition of macromolecular polysaccharides like pectin may occur at high temperatures. Concurrently, the level of its isomer, turanose, also increased synchronously, with a fold change of 6.2 (Figure 4c). This isomerization phenomenon is consistent with the transformation pattern observed in “Niangniang tea” during drying, which exhibits pronounced sweetness characteristics [29].

3.4. Identification of Key Differential Primary Metabolites in LJGT from Different Geographical Origins

The susceptibility of primary metabolites to environmental factors allows for a detailed characterization of the metabolic differences in LJGT from distinct geographical origins [30,31]. In this study, 49 LJGT samples were collected from three producing areas of Xihu (XH), Qiantang (QT), and Yuezhou (YZ), and a total of 353 features were obtained by derivatized GC-MS analysis (Table S8). PLS-DA was used to explore the differences in metabolic profiles of samples from different producing areas. As shown in Figure 5a, the PLS-DA score plot shows the distribution of metabolic characteristics of LJGT samples from the three producing areas. The distribution of samples from the XH was more concentrated and was significantly different from the QT and YZ regions in the direction of latent variable 1. This difference in distribution may be related to the more consistent geographical environment and planting conditions in the XH producing area, while the QT and YZ producing areas have large metabolic differences within the samples due to their wide geographical range. The QC samples are clustered tightly in the center of the score plot, indicating that the experimental data have good repeatability and stability. The PLS-DA model yielded a Q2 of 0.184, indicating it may have limited predictive capability. However, receiver operating characteristic (ROC) curve analysis revealed excellent discriminatory power, with the area under the curve (AUC) exceeding 0.95 for each of the three producing regions (Figure S1). Furthermore, the permutation test (n = 200) of Figure 5b shows that the R2 intercept is 0.344 and the Q2 intercept is -0.213, and that there is no overfitting of the model.

Figure 5.

Figure 5

Open in a new tab

Multivariate statistical analysis of LJGT samples from different geographical origins. (a) PLS-DA score plot showing the clustering of samples; (b) result of the permutation test (n = 200) for validating the PLS-DA model; (c) heatmap depicting the relative abundance of key differential compounds, a bold font indicates that the compound has been certified as a standard; (d) pathway analysis of the effect of geographical origin on LJGT based on pathway-matched metabolites.

To identify key differential metabolites distinguishing LJGT from different geographical origins, 149 features were initially screened based on VIP > 1. Subsequently, peaks with low reliability (QC sample RSD > 20%), duplicate signals (retention time difference < 1 s and characteristic ions from the same compound), and no significant changes during processing (Kruskal–Wallis test, p > 0.05) were excluded. Ultimately, 17 key characteristic peaks were retained. Through NIST database matching and standard sample validation, 15 key differential metabolites were identified. Heatmap visualization and HCA clustering of the distribution patterns of these 15 key differential compounds revealed distinct distribution patterns across tea samples. Except for shikimic acid and quinic acid, which were significantly enriched, and xylose, succinic acid, and fructose, which were notably lower in the XH region, most key differential metabolites showed considerable level variation even within the same geographical origin (Figure 5c). Pathway analysis further indicated that these metabolites are associated with several core metabolic pathways, including galactose metabolism, butanoate metabolism, TCA cycle, alanine, aspartate and glutamate metabolism, starch and sucrose metabolism, and glyoxylate and dicarboxylate metabolism (Figure 5d). These findings suggest that geographical origin plays a significant role in shaping the primary metabolite profile of LJGT, likely driven by tea cultivar, soil properties, and fertilization practices. It is worth noting that commercial samples may exhibit metabolic heterogeneity due to varying processing parameters and post-harvest storage conditions, even within the same region.

3.5. Quantitative Analysis of Organic Acids and Soluble Sugars in LJGT from Different Geographical Origins

Given the pivotal role of carbohydrates and organic acids in primary metabolism and flavor formation in tea leaves, quantitative analyses were conducted on six organic acids and four sugars to further validate the differences in primary metabolites among LJGT from various geographical origins (Figure 6 and Table S9). The results indicated that the total organic acid content in LJGT ranged from 3.6 to 13.6 mg/g, with specific concentrations as follows: citric acid (0.13–4.87 mg/g), malic acid (0.12–3.54 mg/g), and gallic acid (0.26–2.91 mg/g) were relatively high. No significant differences were observed in total organic acid content or most individual organic acid components among LJGT from different geographical origins. However, the average content of quinic acid and shikimic acid in tea samples from the XH region was significantly higher than in the other two regions (p < 0.05). Quinic acid and shikimic acid are key intermediates in the shikimate pathway, which occupies a central position in plant secondary metabolism [32,33]. This pathway is known to play an important role in the biosynthesis of secondary metabolites in tea, such as alkaloids, polyphenols, and amino acids.

Figure 6.

Figure 6

Open in a new tab

Comparison of organic acid and soluble sugar contents in LJGT from the Xihu (XH), Qiantang (QT), and Yuezhou (YZ) regions. Note: Significance determined by Kruskal–Wallis test with Dunn’s test, * p < 0.05; ** p < 0.01; and *** p < 0.001.

The total soluble sugar content in LJGT ranged from 0.1 to 79.5 mg/g, with relatively high content of sucrose (0.02–76.27 mg/g) and fructose (0.07–4.53 mg/g). Notably, some samples from the QT and YZ regions (QT13, QT15, YZ13, YZ14, YZ17, YZ18) exhibited elevated contents of sucrose, fructose, glucose, and galactose, with an average total soluble sugar content reaching 40.88 mg/g. This aligns with previous findings indicating that higher sucrose content correlates with increased contents of other sugars [34,35,36]. Soluble sugar compound contents in the XH region were relatively lower, with fructose, glucose, and galactose content significantly lower than in the YZ region (p < 0.05).

Considering that LJGT from all three regions primarily consist of ‘Longjing 43’, ‘Zhongcha 108’, and ‘Qunti’ cultivars, their metabolic differences are more likely associated with varying geographical conditions [37]. Specifically, the XH region features abundant vegetation, providing excellent shading conditions for tea plant growth. Research indicates that shading significantly reduces photosynthetic rates in tea plants, thereby decreasing the accumulation of sugar compounds [38]. Concurrently, shading has been reported to enhance the activity of the shikimic acid pathway and promote the synthesis of aromatic amino acids and phenolic compounds by regulating the carbon-nitrogen metabolic ratio [39,40,41], while inhibiting glycolysis, galactose metabolism, and other related carbohydrate metabolic pathways [42]. Although further enzymatic or transcriptomic evidence is needed, the metabolic patterns observed in this study are consistent with these shading-induced physiological changes. Crucially, these findings provide precise quantitative data on organic acids and soluble sugars, which have been largely overlooked in previous LJGT studies that predominantly focused on secondary metabolites (e.g., catechins, alkaloids, and phenolic acids). These findings provide a novel biochemical basis for deeply understanding LJGT from different geographical origins.

4. Conclusions

In this study, an optimized pre-column derivatization method for the tea matrix was established using an orthogonal design, with optimal conditions determined as 75 μL of MOXat 30 °C for 1.5 h, which significantly enhanced peak detection capacity. Based on this optimized method, 52 primary metabolites, including carbohydrates and organic acids, were identified in LJGT. Metabolic profiling revealed that spreading and pan-fixation are the most critical stages for metabolic remodeling. During spreading, metabolites associated with carbohydrate metabolism, such as sucrose, glucose, malic acid, and succinic acid, decreased significantly. During pan-drying, thermal treatment likely promoted the Maillard reaction, which was accompanied by a significant reduction in reducing sugars, including glucose and fructose. Levels of sucrose, turanose, and quinic acid increased may be attributed to precursor cleavage or isomerization. To investigate the differences in primary metabolites among LJGT from different geographical origins, analysis of 49 samples from the Xihu, Qiantang, and Yuezhou regions identified 15 differential compounds, including shikimic acid and quinic acid. Quantitative analysis was performed on six representative soluble sugars and four organic acids. The results showed that although total organic acid and soluble sugar contents showed no statistical difference across regions, samples from the Xihu region were characterized by significantly elevated contents of shikimic and quinic acids. Collectively, this study not only elucidates the regulatory mechanism of processing on the primary metabolic network but also reveals differences in the content of primary metabolites among LJGT from different geographical origins.

Abbreviations

The following abbreviations are used in this manuscript:

LJGT Longjing green tea
GC–MS Gas chromatography–mass spectrometry
MSTFA N-methyl-N-(trimethylsilyl) trifluoroacetamide
FL Fresh leaves
SL Spreading leaves
PL Pan-fixation leaves
FT Finished tea
QC Quality control
PLS-DA Partial least squares discriminant analysis
Open in a new tab

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/foods15050865/s1, Table S1: The grade, origin, and source of commercially available Longjing green tea from different geographical origins; Table S2: Factor levels of orthogonal experiments; Table S3: Orthogonal design table; Table S4: Results of ANOVA of variance for three factors; Table S5: Results of range analysis in orthogonal experiments; Table S6: All chromatographic retention time, peak areas, VIP values, p-values, and compound identifications of Longjing green tea processing; Table S7: Abbreviation results and full name in Figure 4; Table S8: All chromatographic retention time and peak areas of Longjing green tea from different regions; Table S9: Contents of organic acids and soluble sugar in Longjing green tea from different regions; Figure S1: Receiver operating characteristic (ROC) curve evaluating the performance of the PLS-DA model in differentiating tea samples from different geographical origins. The solid line depicts the model’s ROC curve, and the grey dashed line represents the line of no discrimination (AUC = 0.5). Note: TPR, true positive rate; FPR, false positive rate; AUC, area under the curve.

foods-15-00865-s001.zip (1MB, zip)

Author Contributions

Z.L. (Zhiyuan Lin): Investigation, Writing—original draft, Writing—Review and Editing, Visualization. M.C.: Methodology, Investigation, Formal analysis, Validation. B.Z.: Investigation. J.T. (Junfeng Tan): Investigation, Project administration. L.Z.: Supervision. Z.L. (Zhi Lin): Resources, Project administration. J.T. (Jinchi Tang): Funding acquisition, Writing—Review and Editing. W.D.: Conceptualization, Funding acquisition, Writing—Review and Editing. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding Statement

The work was supported by the funds of National Key Research and Development Program of China (2023YFD1601700), Zhejiang Provincial Outstanding Youth Science Foundation (LR23C160002), and Project of the Department of Agriculture and Rural Affairs of Guangdong Province (NYLJ2024009). We appreciate the analytical technique support from Instrumental Analysis Center of Tea Research Institute Chinese Academy of Agricultural Sciences.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

References

  • 1.Fernie A.R., Pichersky E. Focus issue on metabolism: Metabolites, metabolites everywhere. Plant Physiol. 2015;169:1421–1423. doi: 10.1104/pp.15.01499. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Cheynier V., Comte G., Davies K.M., Lattanzio V., Martens S. Plant phenolics: Recent advances on their biosynthesis, genetics, and ecophysiology. Plant Physiol. Biochem. 2013;72:1–20. doi: 10.1016/j.plaphy.2013.05.009. [DOI] [PubMed] [Google Scholar]
  • 3.Erb M., Kliebenstein D.J. Plant secondary metabolites as defenses, regulators, and primary metabolites: The blurred functional trichotomy. Plant Physiol. 2020;184:39–52. doi: 10.1104/pp.20.00433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Li W., Zhang Z., Chen R., Sun L., Lai X., Li Q., Hao M., Zhang S., Li Q., Sun S., et al. Metabolomics-based analysis of the effects of differences in soluble sugars on the sweetness quality of six major tea types in China. Food Funct. 2025;16:3707–3720. doi: 10.1039/D5FO00232J. [DOI] [PubMed] [Google Scholar]
  • 5.Ho C.-T., Zheng X., Li S. Tea aroma formation. Food Sci. Hum. Wellness. 2015;4:9–27. doi: 10.1016/j.fshw.2015.04.001. [DOI] [Google Scholar]
  • 6.Liu H., Zeng Y., Zhang K., Zhao X., Yang T. Improving the geographical traceability of tea in China based on stable isotope ratios. J. Food Sci. Technol. 2024;61:1943–1954. doi: 10.1007/s13197-024-05970-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Shuai M., Peng C., Niu H., Shao D., Hou R., Cai H. Recent techniques for the authentication of the geographical origin of tea leaves from Camellia sinensis: A review. Food Chem. 2022;374:131713. doi: 10.1016/j.foodchem.2021.131713. [DOI] [PubMed] [Google Scholar]
  • 8.Zhu Y., Lv H., Dai W., Guo L., Tan J., Zhang Y., Yu F., Shao C., Peng Q., Lin Z. Separation of aroma components in Xihu Longjing tea using simultaneous distillation extraction with comprehensive two-dimensional gas chromatography-time-of-flight mass spectrometry. Sep. Purif. Technol. 2016;164:146–154. doi: 10.1016/j.seppur.2016.03.028. [DOI] [Google Scholar]
  • 9.Chen Y., Liu X., Zhu W., Wang Z., Wu J., Ni Z., Pan H., Li D., Wang Y., Zhou J. Roasting-induced transformations in Longjing tea: Sensory, chemical, and molecular insights into roasted aroma profiles. Food Res. Int. 2026;229:118502. doi: 10.1016/j.foodres.2026.118502. [DOI] [PubMed] [Google Scholar]
  • 10.Wang M., Ma W., Shi J., Zhu Y., Lin Z., Lv H. Characterization of the key aroma compounds in Longjing tea using stir bar sorptive extraction (SBSE) combined with gas chromatography-mass spectrometry (GC-MS), gas chromatography-olfactometry (GC-O), odor activity value (OAV), and aroma recombination. Food Res. Int. 2020;130:108908. doi: 10.1016/j.foodres.2019.108908. [DOI] [PubMed] [Google Scholar]
  • 11.Zeng L., Fu Y., Gao Y., Wang F., Liang S., Yin J., Fauconnier M.-L., Ke L., Xu Y. Dynamic changes of key metabolites in Longjing green tea during processing revealed by widely targeted metabolomic profiling and sensory experiments. Food Chem. 2024;450:139373. doi: 10.1016/j.foodchem.2024.139373. [DOI] [PubMed] [Google Scholar]
  • 12.Yu X., Li J., Yang Y., Zhu J., Yuan H., Jiang Y. Comprehensive investigation on flavonoids metabolites of Longjing tea in different cultivars, geographical origins, and storage time. Heliyon. 2023;9:e17305. doi: 10.1016/j.heliyon.2023.e17305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cui H., Mao Y., Zhao Y., Huang H., Yin J., Yu J., Zhang J. Comparative metabolomics study of four kinds of Xihu Longjing tea based on machine fixing and manual fixing methods. Foods. 2023;12:4486. doi: 10.3390/foods12244486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Shan X., Yu Q., Chen L., Zhang S., Zhu J., Jiang Y., Yuan H., Zhou Q., Li J., Wang Y., et al. Analyzing the influence of withering degree on the dynamic changes in non-volatile metabolites and sensory quality of Longjing green tea by non-targeted metabolomics. Front. Nutr. 2023;10:1104926. doi: 10.3389/fnut.2023.1104926. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Bassiony A., Peng Q., Baldermann S., Feng S., Yang K., Zhang Y., Fu J., Lv H., Lin Z., Shi J. Differential accumulation patterns of flavor compounds in Longjing 43 and Qunti fresh leaves and during processing responding to altitude changes. Food Res. Int. 2024;187:114392. doi: 10.1016/j.foodres.2024.114392. [DOI] [PubMed] [Google Scholar]
  • 16.Park J., Lee J., Cho C., Lee D.K., Song W., Choi K., Choi S.Y., Yang H. Mass spectral data of primary and secondary metabolites changes in medicinal plants by solvent polarity. Sci. Data. 2025;12:893. doi: 10.1038/s41597-025-05262-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Engelhardt U.H. Tea chemistry—What do and what don’t we know?—A micro review. Food Res. Int. 2020;132:109120. doi: 10.1016/j.foodres.2020.109120. [DOI] [PubMed] [Google Scholar]
  • 18.Lingwan M., Masakapalli S.K. A robust method of extraction and GC-MS analysis of Monophenols exhibited UV-B mediated accumulation in Arabidopsis. Physiol. Mol. Biol. Plants. 2022;28:533–543. doi: 10.1007/s12298-022-01150-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Stupak M., Filatova M., Benes F., Binova Z., Hajslova J. GC–MS strategy for a comprehensive assessment of cannabis metabolome. Microchem. J. 2025;210:113062. doi: 10.1016/j.microc.2025.113062. [DOI] [Google Scholar]
  • 20.Product of Geographical Indication—Longjing Tea. National Standards of the People’s Republic of China; Beijing, China: 2008. [Google Scholar]
  • 21.Zhang L., Zeng Z., Ye G., Zhao C., Lu X., Xu G. Non-targeted metabolomics study for the analysis of chemical compositions in three types of tea by using gas chromatographymass spectrometry and liquid chromatography-mass spectrometry. Chromatography. 2014;32:804–816. doi: 10.3724/sp.j.1123.2014.04029. (In Chinese) [DOI] [PubMed] [Google Scholar]
  • 22.Wang Y., Zhou L., Zhou Y., Zhao C., Lu X., Xu G. A rapid GC method coupled with quadrupole or time of flight mass spectrometry for metabolomics analysis. J. Chromatogr. B. 2020;1160:122355. doi: 10.1016/j.jchromb.2020.122355. [DOI] [PubMed] [Google Scholar]
  • 23.Chong I.-G., Jun C.-H. Performance of some variable selection methods when multicollinearity is present. Chemom. Intell. Lab. Syst. 2005;78:103–112. doi: 10.1016/j.chemolab.2004.12.011. [DOI] [Google Scholar]
  • 24.Pang Z., Chong J., Zhou G., de Lima Morais D.A., Chang L., Barrette M., Gauthier C., Jacques P.-É., Li S., Xia J. MetaboAnalyst 5.0: Narrowing the gap between raw spectra and functional insights. Nucleic Acids Res. 2021;49:W388–W396. doi: 10.1093/nar/gkab382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Qi D., Shi Y., Lu M., Ma C., Dong C. Effect of withering/spreading on the physical and chemical properties of tea: A review. Compr. Rev. Food Sci. Food Saf. 2024;23:e70010. doi: 10.1111/1541-4337.70010. [DOI] [PubMed] [Google Scholar]
  • 26.Li Y., Chen Y., Dai X., Pan K., Mi X., Qiao D., Chen J., Chen Z., Shen Q., Zhang X. A comprehensive study of the physiology and chemistry of tea withering based on untargeted metabolomic, transcriptomic, and biochemical analyses. Food Chem. 2025;464:141713. doi: 10.1016/j.foodchem.2024.141713. [DOI] [PubMed] [Google Scholar]
  • 27.Prawira-Atmaja M.I., Puangpraphant S. Flavonol glycosides in tea: The role of mitigating oxidative stress during the withering process. Beverage Plant Res. 2025;5:e031. doi: 10.48130/bpr-0025-0017. [DOI] [Google Scholar]
  • 28.Liu J., Qu Y., Liu Y., Zhang H., Luo L., Zeng L. Dynamic changes and formation mechanisms of volatile compounds in the Maillard reaction between theanine and fructose in tea. Food Res. Int. 2026;227:118208. doi: 10.1016/j.foodres.2025.118208. [DOI] [PubMed] [Google Scholar]
  • 29.Wei Q., Liu X., Long S., Zhu L., Yang H., Song X., Peng Y., Chen T., Liu J., Wen B., et al. Exploration of the substances and key processing steps related to the sweetness of Niangniang tea. Food Chem. X. 2025;28:102556. doi: 10.1016/j.fochx.2025.102556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Salam U., Ullah S., Tang Z.H., Elateeq A.A., Khan Y., Khan J., Khan A., Ali S. Plant metabolomics: An overview of the role of primary and secondary metabolites against different environmental stress factors. Life. 2023;13:706. doi: 10.3390/life13030706. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Peng Y., Li M., Song F., Liu S., Qin Y., Hu B., Cui X. Identification of primary metabolite profiles reveals quality characteristics of citrus maxima ‘Shatian Yu’ from different origins. Curr. Issues Mol. Biol. 2024;46:12830–12846. doi: 10.3390/cimb46110764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Tzin V., Galili G. New insights into the shikimate and aromatic amino acids biosynthesis pathways in plants. Mol. Plant. 2010;3:956–972. doi: 10.1093/mp/ssq048. [DOI] [PubMed] [Google Scholar]
  • 33.Wang L., Wang Y., Wang Y., Wu L., He M., Mao Z., Liu G., Wei K., Wang L. Two shikimate dehydrogenases play an essential role in the biosynthesis of galloylated catechins in tea plants. Hortic. Res. 2024;12:uhae356. doi: 10.1093/hr/uhae356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Ding Y., Yu H., Mou S. Direct determination of free amino acids and sugars in green tea by anion-exchange chromatography with integrated pulsed amperometric detection. J. Chromatogr. A. 2002;982:237–244. doi: 10.1016/S0021-9673(02)01650-3. [DOI] [PubMed] [Google Scholar]
  • 35.Kausar T., Akram K., Kwon J.-H. Comparative effects of irradiation, fumigation, and storage on the free amino acids and sugar contents of green, black and oolong teas. Radiat. Phys. Chem. 2013;86:96–101. doi: 10.1016/j.radphyschem.2012.12.011. [DOI] [Google Scholar]
  • 36.Zhou J., Chang X., Wang H., Li Y., Cui X., Hu L., Zhang Y., Xu B. Content of fructose, glucose and sucrose in tea and source analysis of sucrose. J. Tea Commun. 2021;48:311–318. (In Chinese) [Google Scholar]
  • 37.Wang L., Wei K., Cheng H., He W., Li X., Gong W. Geographical tracing of Xihu Longjing tea using high performance liquid chromatography. Food Chem. 2014;146:98–103. doi: 10.1016/j.foodchem.2013.09.043. [DOI] [PubMed] [Google Scholar]
  • 38.Yang F., Fan Y., Wu X., Cheng Y., Liu Q., Feng L., Chen J., Wang Z., Wang X., Yong T., et al. Auxin-to-Gibberellin ratio as a signal for light intensity and quality in regulating soybean growth and matter partitioning. Front. Plant Sci. 2018;9:56. doi: 10.3389/fpls.2018.00056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Lee L.-S., Choi J.H., Son N., Kim S.-H., Park J.-D., Jang D.-J., Jeong Y., Kim H.-J. Metabolomic analysis of the effect of shade treatment on the nutritional and sensory qualities of green tea. J. Agric. Food. Chem. 2013;61:332–338. doi: 10.1021/jf304161y. [DOI] [PubMed] [Google Scholar]
  • 40.Zhang Q., Liu M., Mumm R., Vos R.C.H., Ruan J. Metabolomics reveals the within-plant spatial effects of shading on tea plants. Tree Physiol. 2020;41:317–330. doi: 10.1093/treephys/tpaa127. [DOI] [PubMed] [Google Scholar]
  • 41.Liu Y., Xu X., Chen Y., Yu C., Liu W., Zhang Y., Huang J., Wei H., Jiang X., Gao Y. Impact of ecological shading on the quality of Lushan Yunwu summer tea. Ind. Crops Prod. 2025;232:121197. [Google Scholar]
  • 42.Shao C., Jiao H., Chen J., Zhang C., Liu J., Chen J., Li Y., Huang J., Yang B., Liu Z., et al. Carbon and nitrogen metabolism are jointly regulated during shading in roots and leaves of Camellia sinensis. Front. Plant Sci. 2022;13:894840. doi: 10.3389/fpls.2022.894840. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

foods-15-00865-s001.zip (1MB, zip)

Data Availability Statement

The original contributions presented in this study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

Articles from Foods are provided here courtesy of Multidisciplinary Digital Publishing Institute (MDPI)

RESOURCES