Systematic Identification of Characteristic Metabolites and Analysis of Quality and Metabolomic Profiles of Yunnan Kucha White Tea

Abstract

Kucha, a unique tea germplasm rich in theacrine, imparts its fresh leaves with a particularly bitter taste and multiple bioactivities. However, systematic studies on processed Kucha—especially white tea—remain limited. In this study, white teas were produced from two Yunnan Kucha accessions (YLKC1, YLKC2) and two conventional cultivars. Their quality characteristics and non-volatile metabolic profiles were systematically compared using sensory evaluation, targeted quantification and widely targeted metabolomics. Results indicated that Kucha white teas displayed pronounced bitterness, with YLKC1 presenting a richer, well-layered flavor, indicating promising quality potential. Targeted quantification demonstrated a remarkably high theacrine content (~30 mg/g) in Kucha white teas, whereas caffeine and several catechin monomers were significantly lower than those in conventional cultivars. Widely targeted metabolomic analysis identified 3376 non-volatile metabolites. PCA and OPLS-DA demonstrated a clear separation in metabolic profiles between Kucha and control groups. In total, 601 significantly differential metabolites were identified. Taste-driven annotation against ChemTastesDB revealed 17 known bitter compounds, 10 of which were significantly accumulated in Kucha white tea—including theacrine, theophylline, theobromine, L-arginine, neohesperidin, pinocembrin, kaempferol-3-O-(6”-malonyl)glucoside, fraxin, adenosine, and xanthine. Among these compounds, theacrine showed the highest upregulation (9.30-fold). In addition, several galloylated flavonoid glycosides also exhibited significant accumulation. KEGG enrichment analysis further indicated that flavonoid biosynthesis and caffeine metabolism were crucial pathways contributing to these metabolic differences. Collectively, these findings demonstrate that the characteristic bitterness of Kucha white tea arises from the coordinated accumulation of a specific set of bitter phytochemicals rather than a single compound and provide a prioritized panel of candidate compounds for flavor-oriented breeding and processing.

1. Introduction

Kucha (Camellia sinensis) is a unique and valuable tea resource primarily distributed in the Yunnan, Guangdong, Guangxi, Hunan, and Fujian provinces of China, with the most extensive distribution occurring in Yunnan and both sides of the Nanling Mountains [1,2]. Its defining characteristic is the significant bitterness present in both fresh leaves and processed tea, which clearly distinguishes Kucha from other cultivated tea plant varieties in terms of sensory quality [3].

The characteristic bitterness and physiological activities of Kucha are closely correlated with its unique chemical composition. Previous studies have indicated that Kucha is particularly rich in theacrine (1,3,7,9-tetramethyluric acid), a purine alkaloid structurally analogous to caffeine [4]. Theacrine exhibits a strong bitter taste, substantially higher than that of caffeine and theobromine [5,6]. From a biosynthetic perspective, theacrine is obtained from caffeine through oxidation at the C8 position to form the intermediate 1,3,7-trimethyluric acid, followed by methylation at the N9 position [7,8,9]. This metabolic conversion from caffeine to theacrine not only reduces direct stimulatory effects on the central nervous system but also confers a range of distinct physiological functions [10,11]. Studies have reported that theacrine can enhance locomotor activity, reduce fatigue, and improve cognitive performance [12,13]. Additionally, theacrine exhibits several biological activities, including hypnotic [14,15], antidepressant [10], antioxidant [16], and anti-inflammatory effects [17]. It has also been reported to inhibit breast cancer cell metastasis [9] and to prevent non-alcoholic fatty liver disorder [18].

Owing to its distinctive flavor and diverse pharmacological activities, Kucha has attracted increasing research attention, making the elucidation of its chemical constituents a primary focus of investigation [19]. Previous studies have systematically characterized the composition of alkaloids and catechins in Kucha, identifying theacrine and epigallocatechin gallate (EGCG) as its predominant constituents, respectively [20,21,22]. Furthermore, several studies employing diverse metabolomic platforms have conducted in-depth investigations from multiple perspectives, including the identification of characteristic metabolites [3], tissue-specific metabolic profiling [23], and the screening of bitter compounds [19,24]. However, these investigations have largely been confined to fresh tea leaves, whereas systematic investigations on processed commercial tea remain notably limited. Given that it is the final product directly consumed by consumers, a comprehensive understanding of Kucha tea’s quality characteristics is essential for its production and for the scientific development of consumer-oriented products.

In this study, fresh leaves from two Yunnan Kucha accessions and two conventional tea cultivars were processed into their samples employing a standardized protocol. The sensory quality of the resulting samples was assessed through sensory evaluation. Additionally, their non-volatile metabolite profiles were systematically analyzed employing targeted quantification by high-performance liquid chromatography (HPLC) and widely targeted metabolomics based on ultra-performance liquid chromatography–tandem mass spectrometry (UPLC-MS/MS). This study provides a comprehensive analysis of the quality characteristics and non-volatile metabolic profiles of Kucha white tea, highlights its differences from conventional tea cultivars, and offers a scientific basis for the development and utilization of Kucha-derived tea products, while laying a foundation for future research into their nutritional properties and physiological activities.

2. Materials and Methods

2.1. Tea Sample Preparation

Sample collection and processing were performed in August 2023 in Menghai County, Yunnan Province, China. The experimental materials included two Kucha accessions (‘Yunling Kucha 1’ and ‘Yunling Kucha 2’) and two control cultivars (‘Yunkang 10’ and ‘Yuncha 1’). Fresh leaves were harvested following the standard of one bud with two leaves. The freshly harvested leaves were naturally withering indoors at room temperature for 72 h. Subsequently, the withered leaves were dried at 80 °C for 3 h to produce white tea samples. Three biological replicates were prepared for each sample, with all replicates derived from different plants. All processed tea samples were then immediately sealed in airtight, light-proof aluminum foil bags and stored at −80 °C until the sensory evaluation and metabolite analysis.

2.2. Sensory Evaluation of Tea

Sensory evaluation was performed in accordance with the China National Standard “Methodology for Sensory Evaluation of Tea” (GB/T 23776-2018) [25]. Briefly, 3 g of white tea sample was accurately weighed and transferred into a 150 mL cylindrical cup. Samples were then brewed with freshly boiled water for 5 min. After brewing, the tea infusion was carefully filtered prior to evaluation. A panel of seven experienced panelists, each holding the national senior tea assessor certification, scored each of the five core quality attributes on a hundred-point scale. The scoring was performed according to the descriptive lexicons and scoring references for white tea specified in Appendix B.7 of the Chinese national standard GB/T 23776-2018 [25]. All sensory evaluations were conducted in a blinded manner. A weighted composite score was then calculated based on the following predetermined weighting coefficients: dry leaf appearance (25%), infusion color (10%), aroma (25%), taste (30%), and infused leaf (10%). All sensory evaluations were completed in October 2023.

2.3. Absolute Quantification of Catechins and Purine Alkaloids

The contents of catechin monomers and purine alkaloids in white tea samples were investigated by HPLC, after the analytical method explained by Jin et al. [26]. The reference standards included gallocatechin (GC), catechin (C), epicatechin (EC), epicatechin gallate (ECG), epigallocatechin (EGC), EGCG, theobromine, caffeine, and theacrine. The purities of all standards were no less than 98% and were obtained from Sigma–Aldrich Shanghai Trading Co., Ltd. (Shanghai, China).

2.4. Determination of Non-Volatile Metabolites Employing UPLC-MS/MS

The samples were ground into a homogeneous powder employing a ball mill (MM 400, Retsch (Haan, Germany)) at 30 Hz for 90 s. Precisely 30 mg of the powdered sample was weighed and extracted with 1.5 mL of a 70% (v/v) methanol aqueous solution pre-cooled to −20 °C, which was spiked with 2-chlorophenylalanine (98% purity) as an internal standard at a final concentration of 1 mg/L. This internal standard was used only for quality control and not involved in metabolite quantification. The samples were vortex-mixed for 30 s at 30 min intervals, with the cycle repeated six times. The mixture was then centrifuged at 12,000 rpm for 3 min, and the supernatant was collected and filtered through a 0.22 μm microporous membrane, before being transferred to a sample vial for UPLC-MS/MS analysis.

UPLC-MS/MS analysis was performed on an ExionLC^TM AD UPLC system coupled with an ESI-Q TRAP-MS/MS (SCIEX, Framingham, MA, USA). Chromatographic separation was achieved on an Agilent SB-C18 column (1.8 µm, 2.1 mm × 100 mm) at 40 °C. The mobile phase consisted of (A) 0.1% formic acid in water and (B) 0.1% formic acid in acetonitrile. The gradient program (0.35 mL/min) was: 0–9 min, 5–95% B; 9–10 min, 95% B; 10–11.1 min, 95–5% B, held at 5% B for 2.9 min. The injection volume was 2 µL.

ESI source parameters: temperature 500 °C; ion spray voltage +5500 V (positive) and −4500 V (negative); ion source gas I and II at 50 and 60 psi; curtain gas 25 psi; collision-activated dissociation (CAD) gas high. Metabolite quantification was performed in multiple reaction monitoring (MRM) mode with nitrogen as collision gas (medium). Declustering potential (DP) and collision energy (CE) were optimized for each MRM transition.

Metabolite identification was based on matching the acquired MS/MS spectra against the in-house Metware Database (MWDB). Relative quantification was performed by integrating the peak areas of characteristic ion pairs across samples. Peak area normalization against the internal standard (2-chlorophenylalanine) was applied to correct for instrument response variations. Rigorous quality control procedures were applied to ensure data reliability, which included the analysis of quality control (QC) samples, imputation of missing values, filtration of features with high coefficients of variation (CV > 0.5), and peak alignment with retention time correction. Representative total ion chromatograms (TIC) of the detected metabolites in positive and negative ion modes are provided in Figures S1 and S4 to demonstrate the stability and reproducibility of the UPLC-MS/MS runs. Additionally, representative MS/MS spectra of six annotated metabolites are shown in Figure S5 to support their structural identification based on database matching or fragmentation patterns. All metabolite detection and instrumental analysis were completed in October 2024 at Wuhan Metware Biotechnology Co., Ltd. (Wuhan, China).

2.5. Orthogonal Partial Least Squares-Discriminant Analysis (OPLS-DA)

Orthogonal partial least squares discriminant analysis (OPLS-DA) was performed to identify differential metabolites between the YLKC and CT groups. Both groups consisted of six replicates each: the YLKC group comprised samples YLKC1 and YLKC2 (three biological replicates each), and the CT group comprised samples YK10 and YC1 (three biological replicates each). Data were preprocessed by log₂ transformation and mean-centering. The analysis was conducted using the MetaboAnalystR (version 1.0.1) package in R (version 4.4.2), with 7-fold K-fold cross-validation (implemented natively in the OPLSR.Anal function) for model validation. The model was evaluated based on R²X, R²Y, and Q² (the cross-validated predictive ability). A permutation test with 200 randomizations was applied to validate the model against overfitting; the model was considered robust if the original Q² value was significantly higher than that of the permuted models (p < 0.05). Based on the validated model, differential metabolites were screened using the following criteria: Variable Importance Projection (VIP) > 1, log₂(Fold Change) > 1 or <−1, and an unadjusted p < 0.05.

2.6. Statistical Analysis

Data collation was performed employing WPS Office software (version 12.1.0.15712). One-way analysis of variance was conducted using the Statistical Package for the Social Sciences software (version 26.0), followed by the least significant difference test for post hoc analysis. Bar graphs and heatmaps were generated employing GraphPad Prism software (version 8.2.1) and TBtools (version 2.441) [27], respectively.

3. Results

3.1. Sensory Evaluation

A systematic sensory evaluation was first performed on the white teas produced from two new Kucha accessions (YLKC1 and YLKC2) and two control cultivars (YK10 and YC1) (Figure 1 and Table S1). Regarding dry leaf appearance, all tea samples displayed intact and well-formed strips, with fat buds and significant pekoe. YLKC1 achieved the highest score (92 points), indicating superior strip morphology compared to controls YK10 (90 points), YC1 (90 points), and YLKC2 (89 points). In terms of infusion color, all samples displayed apricot-yellow and bright appearances, indicative of high quality, with YC1 scoring highest for this attribute (93 points). Aroma and taste were crucial factors distinguishing Kucha from conventional cultivars. Both Kucha accessions (YLKC1 and YLKC2) exhibited a sweet aroma, scoring 90 and 91 points, respectively. Although slightly lower than the herbaceous aroma of YC1 (92 points), their aromas were preferable to the slightly greenish-sweet notes of YK10 (87 points). Taste profile analysis demonstrated that YLKC1 and YLKC2 displayed distinctive and significant bitterness, leading to lower scores than the two control cultivars. Notably, YLKC1 demonstrated a mellow, slightly bitter taste with a sweet aftertaste, suggesting a richer and well-layered flavor profile. For infused leaves, YLKC1 and YLKC2 exhibited greenish-auburn, tender, and evenly spread leaves, each scoring 89 points, outperforming YK10 and YC1. Overall sensory scores indicated that YLKC1 (90.35 points) surpassed YK10 (88.85 points) and YLKC2 (88.55 points) in overall quality and was only slightly lower compared to YC1 (91.30 points). These findings demonstrate that the two new Kucha accessions, YLKC1 and YLKC2, possess a competitive overall sensory quality. In particular, YLKC1 displayed a distinctive flavor profile, indicating its strong potential for application in white tea processing.

3.2. Analysis of Purine Alkaloids and Catechins in Kucha White Tea

Purine alkaloids and catechins are two core classes of secondary metabolites in tea, whose content and composition substantially influence its quality [28]. In this study, these two groups of compounds were quantitatively analyzed employing HPLC. The findings for purine alkaloid are illustrated in Figure 2. Specific accumulation of theacrine was the most essential chemical marker in the YLKC (YLKC1 and YLKC2), with a content reaching approximately 30 mg/g, whereas it was not detected in the conventional cultivars YK10 and YC1. In contrast, caffeine levels in the YLKC group were remarkably low (<15 mg/g) and substantially lower than those in the control group (CT), especially in YLKC1, which contained <10 mg/g. Additionally, the theobromine content in YLKC2 was significantly higher compared to the other three samples. Non-significant difference in theobromine content was observed between YLKC1 and YK10, but both were significantly higher compared to YC1. In terms of catechin composition, compared to YK10 and YC1, YLKC1 and YLKC2 had significantly lower levels of simple catechins (C, EC, GC) and ECG, but significantly higher EGCG content (Figure 2). Additionally, YLKC2 demonstrated the highest EGC content among all samples, whereas a non-significant difference in EGC content was maintained among the other three samples. In summary, the chemical fingerprint of the YLKC is characterized by high theacrine, high EGCG, and low caffeine.

3.3. Distinct Non-Volatile Metabolic Profiles Between YLKC and Control Samples

To comprehensively characterize the non-volatile metabolome of Kucha resources and to explore the material foundation underlying the significant bitterness of Kucha white tea, a comprehensive analysis was performed on four samples (YLKC1, YLKC2, YK10, and YC1) employing UPLC-MS/MS-based widely targeted metabolomics. A total of 3376 metabolites were measured across the four samples (Figure 3A, Table S2). According to their chemical structures, these metabolites were classified into the following categories: flavonoids (691), terpenoids (441), phenolic acids (437), lipids (289), alkaloids (287), amino acids and derivatives (280), lignans and coumarins (147), organic acids (110), nucleotides and derivatives (93), tannins (88), quinones (27), steroids (19), and others (467). The classification findings suggested that flavonoids indicated the most abundant category, accounting for 20.47% of the total detected metabolites (Figure 3A).

To evaluate the overall metabolic differences among sample groups, principal component analysis (PCA) was performed (Figure 3B). The findings demonstrated that the variance contributions of the first (PC1) and second (PC2) principal components were 35.25% and 23.67%, respectively, with a cumulative contribution rate of 58.92%, suggesting that these two components adequately captured the primary structure of the inter-sample metabolic variation. The PCA score plot demonstrated minimal variation within groups but a clear separation trend between groups. Specifically, YLKC1 and YLKC2 were clearly separated from YC1 and YK10, whereas YLKC1 and YLKC2 clustered closely together (Figure 3B), indicating that Kucha possesses highly conserved and unique metabolic characteristics. Furthermore, a heatmap clustering analysis based on the relative abundance of all identified metabolites was performed (Figure 3C). The results indicated that metabolites grouped into several clusters according to their abundance patterns. Hierarchical clustering of samples indicated that YLKC1 and YLKC2 formed a single branch, which was clearly separated from the CTs (YC1 and YK10). This finding is consistent with the PCA results and collectively demonstrates systematic differences in the overall metabolic profiles between Kucha (YLKC1 and YLKC2) and the control samples. Additionally, PCA and heatmap analyses revealed that although the two control samples (YC1 and YK10) are both conventional tea cultivars, their metabolite compositions still displayed certain differences, reflecting distinct sensory characteristics, such as taste. The inclusion of control samples with varying metabolic backgrounds also facilitated more accurate identification of metabolites specific to Kucha white tea.

3.4. Validation of Distinct Metabolic Differences Between YLKC and Control Groups Through OPLS-DA

Given the unique bitter taste characteristics and similar metabolite abundance patterns observed in the YLKC group (Figure 3B,C), the samples were divided into two groups for differential metabolite analysis to systematically identify the potential bitter compounds in white tea processed from YLKC germplasms. YLKC1 and YLKC2 were classified as the YLKC group, while YC1 and YK10 were designated as the CT group.

To visualize the separation of metabolic differences between YLKC and CT groups, OPLS-DA was employed. The OPLS-DA score plot indicated a clear separation between the two groups: the CT group (orange) clustered on the left, and the YLKC group (green) clustered on the right (Figure 4A), with tight clustering of replicates within each group suggesting good reproducibility. In terms of variance explanation, T score[1] and Orthogonal T score[1] accounted for 33.7% and 26.2% of the variance, respectively, supporting the statistical significance of the inter-group metabolic differences. The reliability of the model was further verified by a permutation test with 200 permutations (Figure 4B). Model parameters were R²Y = 1, Q² = 0.985 (p < 0.005), and R²X = 0.59. As an exploratory multivariate analysis, these metrics supported that the model exhibited reliable explanatory and predictive ability without overfitting and effectively distinguished between the YLKC and CT groups. This provided a solid statistical foundation for the subsequent screening of bitterness-associated differential metabolites.

3.5. Differential Analysis of Non-Volatile Compounds Between YLKC and Control White Tea

To identify the core bitter compounds in white tea processed from YLKC germplasms, differential metabolites between YLKC and CT groups were screened employing the following criteria: VIP > 1 (from an OPLS-DA model with biological replicates ≥3), |log₂FC| > 1, and p < 0.05. A total of 601 differential metabolites were verified (Figure 5A, Table S3). Compared to the CT group, 303 metabolites were upregulated and 298 were downregulated in the YLKC group (Figure 5A). These differential metabolites were primarily classified into the following categories: flavonoids (177), phenolic acids (95), terpenoids (86), amino acids and derivatives (39), alkaloids (37), tannins (26), lignans and coumarins (24), lipids (17), organic acids (11), nucleotides and derivatives (4), quinones (4), steroids (3), and others (78). Most differential metabolites were concentrated in flavonoids, phenolic acids, and terpenoids (Figure 5B).

Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was conducted on the 601 differential metabolites. The findings demonstrated that 56 metabolites were enriched in 48 metabolic pathways (Table S4). Figure 5C exhibits the top 20 pathways with the smallest p-values. Among these, flavonoid biosynthesis, caffeine metabolism, and stilbenoid, diarylheptanoid and gingerol biosynthesis were pronouncedly enriched.

Furthermore, the 601 differential metabolites were queried against ChemTastesDB (https://doi.org/10.1016/j.fochms.2022.100090 accessed on 13 February 2026). Among them, 19 metabolites were annotated with defined taste attributes (Figure 5D). With the exception of two umami compounds, Glu-Glu and L-glutamic acid, the remaining 17 metabolites were characterized as bitter substances. Notably, 10 of these bitter metabolites exhibited higher accumulation in the YLKC group than in the CT group, including theacrine, theophylline, theobromine, L-arginine, neohesperidin, pinocembrin, kaempferol-3-O-(6”-malonyl)glucoside, fraxin, adenosine, and xanthine.

Additional analysis of the top 20 metabolites with the largest absolute fold-change values between YLKC and CT groups demonstrated that theacrine was the most significantly upregulated compound in the YLKC group (9.30-fold increase), which is highly aligned with the HPLC quantification findings (Figure 5E). Conversely, the most significantly down-regulated metabolite was isovitexin 2”-O-β-D-glucoside, with a fold change of −9.90. Notably, 16 of the 20 metabolites were flavonoids, and all were present in glycosylated forms. Among these, more than half (9 compounds) were upregulated in YLKC group, including kaempferol-3-O-(6”-galloyl)galactoside (5.76-fold), kaempferol-3-O-(2”-galloyl)galactoside (5.28-fold), kaempferol 7-(6”-galloylglucoside) (5.23-fold), delphinidin 3’-O-(2”-O-galloyl-beta-galactopyranoside) (4.87-fold), and quercetin-3-O-(6”-O-galloyl)galactoside (4.81-fold). Given that galloyl moieties are often correlated with astringency and bitterness, these specifically highly expressed flavonoid glycosides are likely essential potential taste-active compounds contributing to the characteristic flavor profile of YLKC white tea, alongside alkaloids.

4. Discussion

As a rare tea germplasm, Kucha holds significant potential for the development of functional tea beverages due to its specific enrichment of theacrine, an alkaloid that is non-toxic to humans and possesses several health-promoting properties [29,30]. The two Kucha white tea samples examined in this study not only exhibited high theacrine contents but also low caffeine levels (<1.5%), with YLKC1 containing less than 1% caffeine. This metabolic profile, marked by “low caffeine, high theacrine”, suggests that these accessions are suitable for developing novel functional tea products, including low-stimulant beverages for caffeine-sensitive populations. Additionally, screening superior low-caffeine germplasm from Kucha resources offers a novel pathway for breeding and product innovation of low-caffeine tea cultivars. Given that theacrine has been reported to exert antidepressant, hypnotic, anti-inflammatory, and antioxidant effects [10,14,15,16,17], the high theacrine accumulation in YLKC white tea may confer similar neuroregulatory and anti-fatigue benefits, while its low caffeine content minimizes stimulatory side effects, making it a potentially health-promoting beverage for broader consumers.

Previous studies have indicated that green, white, and black teas processed from the Guangdong Kucha accession ‘Kucha 6’ all display different degrees of bitterness [31]. The current study further confirms that even after processing into white tea, Kucha samples retain a significant and recognizable characteristic bitterness, distinctly varying from the taste profile of conventional white tea. In product development, balancing bitterness remains a crucial challenge. Although YLKC1 presents bitterness, its “slightly bitter, with a sweet aftertaste” provides valuable insights for flavor optimization (Table S1). This offers insights for flavor optimization in product development: through the selection of suitable cultivars and strategic blending of raw materials, it is possible to achieve a harmonious taste profile without compromising its distinctive traits, thereby improving consumer acceptability.

Current research on Kucha tea has primarily focused on the compositional analysis of fresh leaves, whereas systematic investigations into the quality characteristics of its processed products remain limited. Although fresh leaves offer the foundational material basis for the internal composition of finished teas, processing methods (natural withering of white tea) can induce significant alterations in the composition and content of biochemical components, which are crucial processes for forming the unique quality attributes of different tea types [32,33,34]. Consequently, developing and using Kucha resources requires systematic research on the comprehensive quality of processed products, including the associated flavor and functional metabolites (volatile and non-volatile compounds), to provide scientific support for product innovation and market expansion. The identification and quantitative analysis of tea metabolites are crucial, as the health-promoting effects of tea largely depend on its metabolic profile [35]. This study systematically analyzed the quality characteristics and non-volatile metabolite composition of Kucha white tea, identifying 3376 metabolites (Table S2). Additionally, 601 differential metabolites were screened between the Kucha and conventional white tea samples (Table S3). These differences likely constitute the material basis for their sensory quality distinctions—especially the formation of the unique bitter taste profile of Kucha white tea—while also potentially differentiating it from conventional white tea in terms of health benefits. Consequently, follow-up studies should further explore the specific differences in health efficacy between Kucha and conventional teas when processed into white tea or other tea types.

The characteristic bitterness of Kucha tea is a crucial quality attribute, and identifying its chemical basis has been a central focus of research. Using a taste-driven metabolomics strategy, we queried the 601 differential metabolites against ChemTastesDB. This analysis revealed 17 known bitter compounds, 10 of which were significantly accumulated in YLKC white tea (Figure 5D), including theacrine, theophylline, theobromine, L-arginine, neohesperidin, pinocembrin, kaempferol-3-O-(6”-malonyl)glucoside, fraxin, adenosine, and xanthine. These results provide direct chemical evidence that the intense bitterness of YLKC white tea arises from the coordinated accumulation of a specific set of upregulated bitter phytochemicals, rather than from a single compound. Among these, theacrine showed the most dramatic upregulation (9.30-fold), consistent with its established role as a key bitter metabolite in Kucha germplasms [24,36]. Taste recombination experiments by Xing et al. further demonstrated that removal of theacrine significantly reduced bitterness intensity (p < 0.05) [36]. Additionally, several galloylated flavonoid glycosides (e.g., kaempferol-3-O-(6”-galloyl)galactoside, 5.76-fold) also accumulated to high levels (Figure 5E). We therefore propose that the unique taste profile of YLKC white tea results from the combinatorial effect of theacrine, specific galloylated flavonoids, and the other nine upregulated bitter compounds identified via ChemTastesDB. Future taste recombination studies are warranted to elucidate their individual and synergistic contributions.

5. Conclusions

In summary, this study unraveled the metabolic basis of the characteristic bitterness of Kucha white tea. The bitterness arises from the coordinated upregulation of a specific set of bitter phytochemicals, rather than from any single compound. Among these, theacrine (9.30-fold) and several galloylated flavonoid glycosides (up to 5.76-fold) were identified as major contributors, alongside nine other ChemTastesDB-annotated bitter compounds that also accumulated significantly in YLKC. The YLKC germplasm is further characterized by a “high-theacrine, high-EGCG, low-caffeine” metabolic signature, distinguishing it from conventional white tea cultivars. These findings offer a chemical framework for flavor-oriented breeding and processing of Kucha resources and provide a prioritized panel of candidate compounds for subsequent taste recombination and bioactivity validation.