Effect of Strobilanthes tonkinensis Lindau Addition on Black Tea Flavor Quality and Volatile Metabolite Content

The characteristic aroma of Chinese black tea (BT) produced in summer usually lacks intensity and persistence, reducing consumer acceptance and market demand. Strobilanthes tonkinensis Lindau (STL) possesses excellent biological characteristics, making it a promising novel tea ingredient. We investigated the effects of different addition methods and concentrations for the novel additive STL on the sensory quality of BT. A 20:1500 g/g addition to rolled tea leaves was identified as the best BT with STL (BoS) treatment. We identified 141 volatile metabolites (VMs) for the first time in STL, with high alcohol, ester, ketone, enyne, alkyne, and alkane contents. Partial least-squares discriminant analysis distinguished the samples and revealed 28, 26, and 14 differential VMs in STL vs. BoS, BT vs. STL, and BT vs. BoS comparisons, respectively. Using a combination of Venn, multiple experiment viewer, and odor activity value analysis, 16 key differential VMs were identified. Compared to BT, the 1-octen-3-ol, 1-hexanol, 1-dodecanol, (E)-3-hexen-1-ol, phenylethyl alcohol, and 2-methoxy-3-(2-methylpropyl)-pyrazine contents in BoS were 394.7%, 53.6%, 34.1%, 24.4%, 9.9%, and 5.7% higher, respectively. BoS combined the aromatic properties of BT and STL into a sweet and glutinous rice flavor. The results provide theoretical support for future research and development of novel BT-based products.


Introduction
Black tea (BT) is the most widely consumed tea product worldwide. It contains many functional components, such as theaflavins, theasinensins, amino acids, and flavonoids, and has important health benefits, including antimicrobial, anticancer, antioxidant, antiviral, and caries-preventive effects [1][2][3]. BT produced in China, especially in summer, lacks intensity and persistence in its characteristic aroma [4], which reduces consumer acceptance and, consequently, market demand. To solve this problem, a series of BT products with new flavors have been developed by blending raw BT with flowers from other plant species, such as jasmine, rose, and Osmanthus [5,6], to create new BT products with characteristic flower or fruit flavors. However, the development of these new products has predominantly been based on the use of raw BT, and systematic research on the optimal blending processes is lacking. Consequently, the quality of the new BT products is neither stable nor superior. Therefore, novel additives during BT processing are required to improve the aroma and flavor of BT and to ensure optimal blending and stable quality.
Strobilanthes tonkinensis Lindau (STL) is a medicinal and aromatic plant native to Yunnan Province, China, and is now extensively distributed across Thailand and Vietnam [7,8]. STL has a unique aroma of glutinous rice when dried and is rich in biologically active compounds such as amino acids and flavonoids [9][10][11] and aromatic components such as carene, benzenamine, geraniol, and linalool [7,12,13]. Moreover, STL has many biological activities, including anti-aging, antiviral, anti-obesity, and anticoagulant properties [14][15][16]. Thus, it may be a promising candidate ingredient to help modulate the flavors and fragrances of tea.
In this study, STL was organically blended with BT during BT processing. By determining the optimum addition methods and concentrations, the influence of STL on the aroma of BT was elucidated, and the best technical parameters for the processing of BT with STL (BoS) were determined. Moreover, combined with sensory quality evaluation, infrared-assisted extraction coupled to headspace solid-phase microextraction (IRAE-HS-SPME) pretreatment technology and gas chromatography-tandem mass spectrometry (GC-MS/MS) were used to determine the influence of STL on the volatile compounds and aroma of BT. Multiple experiment viewer (MEV) analysis, partial least-squares discriminant analysis (PLS-DA), Venn, and odor activity value (OAV) analysis were used to determine the key volatile substances present in STL and BoS, the influence of STL on the aroma and flavor of BT, and the key differential volatile substances. We aimed to provide sound theoretical support and innovative ideas to contribute to the future research and development of new BT products.

Materials and Reagents
Fresh tea shoots of 'Fudingdabai' (Camellia sinensis L.) were collected in June 2020 from the Zhejiang Kaihua tea production area. The moisture content of fresh shoots was approximately 79%. Each tea shoot consisted of two leaves and one bud. Dried STL shoots with two leaves were purchased from the Yunwei Guming Tea Company (Honghe, China).
All chromatography-grade standards were purchased from the Yuanye Biological Technology Company (Shanghai, China). The purity of the chemical standards was more than 98%.

Tea Processing and Sample Treatment
Tea processing and sample treatment in this article were carried out in line with published studies [17], as detailed below.

Rolling
The withered leaves were rolled in a rolling machine (6CR-35 type, Zhejiang Chunjiang Tea Machinery Co., Ltd., Hangzhou, China) for 75 min, initially under no pressure for 25 min, and then under light pressure for 20 min, followed by heavy pressure for 10 min, light pressure for 15 min, and finally, no pressure for 5 min.

Fermenting
The rolled leaves were fermented for 3 h in an artificial climatic room at 28 • C and 95% RH.

Drying
The fermented leaves were dried at 110 • C in a hot-air drying machine (Zhejiang Chunjiang Tea Machinery Co., Ltd.) for approximately 15 min until 100 g of the leaves contained 20-25 g of water, and then they were spread out for 30 min. Finally, the BT samples were dried for approximately 30 min at 90 • C to reduce their moisture content to approximately 5%. Different addition treatments and concentrations of STL were evaluated using the same BT leaf-processing conditions as described above. Details of the methods used for the addition of STL during leaf processing are shown in Table 1. All tea samples were collected after the final drying period and stored in an ultra-low temperature freezer (991 type, Thermo Fisher Scientific Inc., Waltham, MA, USA) at −80 • C until analysis.

Tea Sensory Quality Evaluation
Tea sensory quality was assessed according to the National Standard of China (GB/T23776-2018) by five professional tea expert panelists (three women and two men, 25-48 years of age). The 100-score tea quality grading system was used, in which 30% was assigned for the tea taste, 25% for aroma, 25% for appearance, 10% for the liquor color, and 10% for the infused leaf. Tea samples (3.0 g) were added to each corresponding teacup containing pure boiling water (150 mL) and soaked for 5 min, followed by immediate separation. The sensory properties of tea samples were evaluated in the following order: aroma for hot scent, liquor color, aroma for warm scent, taste, aroma for cooled-down scent, and infused leaves [17,18].

Sample Preparation
The samples (0.5 g) were accurately weighed, and 0.0001 g was placed in headspace vials (20 mL). To this, 10 µL of 20 mg/L ethyl caprate in 1 mL of boiling pure water was added. The headspace vial cap was immediately tightened. The cap was protected by being impaled with a DVB/CAR/PDMS fiber head (Supelco, Bellefonte, PA, USA) with a manual handle (SPME, Supelco). The vial was then placed over a 100 W infrared device at a distance of 6 cm for 15 min, and the fiber head was inserted into a GC-MS inlet and desorbed at 250 • C for 2 min. Each sample was assessed three effective times for IRAE-HS-SPME [19][20][21].

GC-MS/MS Analysis
A gas chromatograph (7890B-7000C type, Agilent Corporation, Santa Clara, CA, USA) was used in the splitless injection mode. High-purity helium (99.99%) at a 1.0 mL/min flow rate was used as the carrier gas. The GC injector temperature was 250 • C, and the program of column temperature was as follows: initial temperature of 50 • C maintained for 5 min; an increase to 150 • C at 4 • C/min, which was maintained for 2 min, and an increase to 270 • C at 10 • C/min, which was maintained for 6 min [19,21].

Qualitative and Quantitative Analyses
Based on the NIST11 standard library, volatile metabolites (VMs) with over 80% similarity were screened using the unknown analysis program Agilent Mass Hunter [20]. Kovàts retention indices (RI) for each volatile metabolite were obtained by calculating the linear formula of n-alkanes (C7-C40) and by comparing them with theoretical RI having the same chromatographic column in the NIST Chemistry WebBook (http://webbook.nist. gov/chemistry/, accessed on 1 December 2020), based on the difference of RI within 30. Moreover, some key VMs, such as trans-β-ionone, α-ionone, (E)-4-heptenal, 1-hexanol, linalool, octanal, 1-decanol, naphthalene, indole, decanal, geraniol, cedrol, linoleic acid ethyl ester, and hexadecenoic acid ethyl ester, were identified with authentic standards. For absolute quantitative analysis, the mass concentrations of the VMs were calculated with the internal or authentic standard method using the following equation: where C i is the mass concentration of any component (µg/g), C is is the mass concentration of the internal or authentic standard (µg/g), A i is the chromatographic peak area of any component, and A is is the chromatographic peak area of the internal or authentic standard.

OAV Analysis
OAV is the ratio of the concentration of VMs to the odor threshold in a tea sample [20] and is calculated using the following equation: where c is the volatile compound concentration (µg/kg), and OT is the volatile compound odor threshold in pure water (µg/kg).

Statistical Analysis
The tests were repeated three times, and all experimental data were calculated as the average of three replicates and reported as means ± standard deviation. SAS 9.4 (SAS Institute Inc., Cary, NC, USA) was used to analyze the significant differences in concentrations and contents of the VMs in the different tea samples. The differential volatiles among tea samples were analyzed using SIMCA 13.0 (Umetrics, Umeå, Sweden). Multiple experiment viewer (MEV) (MEV 9.0; https://mev-tm4-org.caas.cn, accessed on 15 June 2021) analysis was used to compare the contributions of the differential VMs. The correlations of VMs were plotted using Venn graphs and R (http://www.r-project.org/, accessed on 15 September 2021) according to the previously described methods [21].

Effects of STL Additions on the Sensory Qualities of BT
Aroma and taste have an important impact on BT quality and its market value as they are the key attributes in tea sensory evaluation [22,23]. The different methods of STL addition and the different STL concentrations tested on the tea leaves were found to have a substantial impact on the taste and aroma of BT but little impact on the appearance, liquor color, and infused leaves ( Table 2). STL leaves were found to have a strong glutinous flavor and a sweet and thick taste, which is in accordance with previous studies [11,12], and the addition of STL significantly influenced BT flavor. The comparison among the different methods of STL addition during tea leaf processing showed that the sensory quality of the BoS samples obtained using treatment #2 (i.e., adding STL to rolled tea leaves) was the best, resulting in a glutinous sweet aroma and a sweet mellow taste. In contrast, adding STL to withered tea leaves resulted in an excessively heavy glutinous flavor that concealed the characteristic aroma of the BT. Adding STL to fermented tea leaves rendered the BT aroma a little grassy, and although the taste was sweet and mellow, STL addition did not improve the sensory qualities of the BT. Note: Data are presented as mean ± standard deviation (n = 5). Different lowercase letters in the same column indicate significant differences based on the least significant difference test (p < 0.05).
The comparisons among the different concentrations of STL tested revealed that the sensory quality of BoS samples prepared using treatment #2 (i.e., 20:1500 w/w) was the highest, with a strong and lasting glutinous sweet aroma and a sweet, mellow, and thick taste. In contrast, STL concentrations exceeding 20:1500 (w/w) concealed the characteristic flavor of BT, which did not benefit BT quality. When the concentration was too low, the aroma and taste of the STL had no positive effect on the BT flavor. To clarify the influence of STL on the properties of BT aroma, BT, BoS (as per treatment #2), and STL samples were selected for further study to detect and analyze their VMs.
The total VMs content and the content and proportion of the volatile compound types were significantly (p < 0.05) different among the BT, BoS, and STL samples (Table S2), with the highest total volatile compound content being in the STL samples (p < 0.05), followed by those in the BT and then the BoS samples, although the difference between the last two sample types was not significant. The ester, alcohol, ketone, enyne, alkyne, and alkane contents were the highest in the STL (p < 0.05), which is consistent with the findings of Zhang [24]. However, the aldehyde content was the highest in the BT samples (p < 0.05). In turn, the proportions of the esters, ketones, enynes, alkynes, and alkanes were the highest in the STL samples (p < 0.05), whereas those of the aromatic hydrocarbons and alcohols were the highest in the BT and STL samples, respectively (p < 0.05 for each). Through appropriate addition technology, STL could confer BT samples with a variety of volatile metabolites, thereby enhancing the aroma quality of BT. The VMs present in the BT and STL samples were combined in the BoS samples. Among them, alcohol and alkane contents were significantly higher in the latter than in the BT samples, which was consistent with the sensory quality evaluation results.

MEV Analysis
MEV analysis was performed for the three tea sample types based on the 141 VMs detected (Table S1) to identify the key compounds contributing to the STL aroma and the influence of STL on BT aroma improvement. The BT and BoS samples were found to be clustered together into one class, whereas the STL samples formed a separate category (Figure 1).

PLS-DA Analysis
PLS-DA is widely used in food and agriculture research as a discriminant analytical method [18,[21][22][23]. Therefore, we applied PLS-DA to explore the influence of STL addition on VMs and aroma improvements of BT and to clarify the differences in volatile compound contents in the BT, BoS, and STL samples according to the 141 compounds found in dry tea. The three sample types were effectively distinguished by PLS-DA (Figure 2a). The STL samples clustered together in both quadrants on the left side, whereas the BoS and BT samples clustered in the upper-right and lower-right quadrants, respectively. These results were consistent with the sensory quality evaluation (Table 2) and the grouping pattern obtained using the heatmap cluster analysis (Figure 1). The model was verified by cross-validation after selecting three principal components. The related variables were as follows: R 2 X (cum) was 0.955, R 2 Y (cum) was 0.988, and Q 2 (cum) was 0.949, which indicated that this model was reliable and had strong prediction abilities. The loading plot (Figure 2b) clearly illustrates the differences in the volatile compound contents among the BT, STL, and BoS samples. For example, the concentrations of VMs such as hexanal, 1-methyl-1H-1,2,4-triazole, ethylbenzene, 3-octanone, o-cymene, benzeneacetaldehyde, linalool, isobutyl pentyl ester, and neophytadiene were significantly higher in BT (p < 0.05) than in the other two sample types, in which hexanal has a fruity and sweaty odor [2], and benzeneacetaldehyde and linalool impart a sweet and woody flavor [25]. Meanwhile, the concentrations of VMs such as 1-octen-3-ol, naphthalene, L-α-terpineol, 2-methoxy-3-(2-methylpropyl)-pyrazine, 5-methyl-tridecane, isophorone, n-pentadecanol, 1-dodecanol, and trans-β-ionone were the highest in the STL samples (p < 0.05). These accounted for glutinous rice, woody, floral, or sweet flavors [6,10,27,28] and played important roles in enhancing the aroma of BT. In contrast, the concentrations of VMs such as 2-methoxy-furan, cis-2-methyl-cyclopentanol, 1-hexanol, 9-methylheptadecane, 5-ethyl-5methyl-decane, 3-methyl-tetradecane, and O-(1-naphthoyl)-1,2-benzenediol were significantly higher in the BoS (p < 0.05) than in the BT or STL samples.

Analysis of Differential VMs
To further analyze the differential VMs among the BT, STL, and BoS samples, we constructed a thermography chart (Figure 4) based on 33 specific differential VMs (Table S3). The BT and BoS samples with a sweet and glutinous rice aroma were clustered into one category, whereas the STL samples formed their own category exhibiting a strong glutinous rice aroma (Figure 4). In addition, the 33 VMs were grouped into three major classes. The first class included eight VMs, namely, (E)-2-hexenal, 1-methyl-1H-1,2,4triazole, benzeneacetaldehyde, linalool, phenylethyl alcohol, and nerol, which had the highest contents in BT and distinctly contributed to the sweet aroma of the BT [5,23]. The second class included (E)-3-hexen-1-ol, cis-2-methyl-cyclopentanol, 1-hexanol, decane, and (Z)-hexanoic acid, 3-hexenyl ester, which were more abundant in the BoS samples than in either the BT or the STL samples; therefore, they were determined to be responsible for the sweet and glutinous rice aroma of BoS, following the combination of BT and STL. The third class included 20 VMs that were more abundant in the STL samples than in the other two samples; this category included compounds such as benzaldehyde, 1-octen-3-ol, 3-carene, 2-ethyl-1-hexanol, isophorone, linalool oxide (pyranoid), 2-methoxy-3-(2-methylpropyl)pyrazine, 1-dodecanol, n-valeric acid cis-3-hexenyl ester, formic acid dodecyl ester, α-ionone, and trans-β-ionone, which contributed greatly to the strong glutinous rice aroma of the STL samples [7,12]. When combined with the Venn diagram, these findings clearly showed that the concentrations and proportions of the volatile substances have strong effects on the odor of each type of tea blend.

Conclusions
The addition of Strobilanthes tonkinensis Lindau (STL) during black tea (BT) processing had a substantial effect on the volatile metabolites (VMs) composition and aroma improvement of BT. We identified 141 VMs in STL firstly with high alcohol, ester, ketone, enyne, alkyne, and alkane contents. STL has a unique aroma of glutinous rice owing to high contents of benzaldehyde, 1-octen-3-ol, isophorone, 2-methoxy-3-(2-methylpropyl)pyrazine, linalool oxide (pyranoid), 1-dodecanol, α-ionone, and trans-β-ionone in the STL leaves. Different addition methods and different addition concentrations of STL influence BT aroma and flavor considerably, and the addition of 20:1500 w/w STL to rolled tea leaves was the best method for processing BT with STL (i.e., BoS), with a strong and lasting glutinous sweet aroma and a sweet, mellow, and thick taste. PLS-DA clearly distinguished BT, STL, and BoS, and revealed 28, 26, and 14 differential volatile compounds when comparing STL vs. BoS, BT vs. STL, and BT vs. BoS, respectively. Combining S-plot, Venn, MEV, and OAV analysis, 16 key differential VMs were identified among BT, STL, and BoS. BoS combined the aroma properties of BT and STL and showed the highest contents of hexanal, (E)-2-hexenal, (E)-3-hexen-1-ol, 1-hexanol, 1-octen-3-ol, benzeneacetaldehyde, linalool, phenylethyl alcohol, 2-methoxy-3-(2-methylpropyl)-pyrazine, 1-dodecanol, and nerol. Our systematic study showed the effects of the new exogenous additive STL on the improvement of the BT aroma and value, thus providing sound theoretical support and insights for the development of new BT products as well as enriching our theoretical understanding of the mechanisms underlying aroma formation in BT. However, this study only shows the effects of STL addition on the VMs and aroma improvement of BT. Thus, the effect of STL addition on non-VMs and taste improvement requires further exploration.