Accepted Manuscript

Characterization of white tea metabolome: Comparison against

green and black tea by a nontargeted metabolomics approach

Weidong Dai, Dongchao Xie, Meiling Lu, Pengliang Li, Haipeng

Lv, Chen Yang, Qunhua Peng, Yin Zhu, Li Guo, Yue Zhang,
Junfeng Tan, Zhi Lin

PII: S0963-9969(17)30102-3
DOI: doi: 10.1016/j.foodres.2017.03.028
Reference: FRIN 6635
To appear in: Food Research International
Received date: 2 August 2016
Revised date: 14 February 2017
Accepted date: 10 March 2017

Please cite this article as: Weidong Dai, Dongchao Xie, Meiling Lu, Pengliang Li, Haipeng
Lv, Chen Yang, Qunhua Peng, Yin Zhu, Li Guo, Yue Zhang, Junfeng Tan, Zhi Lin ,
Characterization of white tea metabolome: Comparison against green and black tea by
a nontargeted metabolomics approach. The address for the corresponding author was
captured as affiliation for all authors. Please check if appropriate. Frin(2017), doi:
10.1016/j.foodres.2017.03.028

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

Characterization of white tea metabolome: comparison against

green and black tea by a nontargeted metabolomics approach

Weidong Dai1, Dongchao Xie1, Meiling Lu2, Pengliang Li1, Haipeng Lv1, Chen Yang1,

Qunhua Peng1, Yin Zhu1, Li Guo1, Yue Zhang1, Junfeng Tan1, *, Zhi Lin1, **

PT
RI
SC
1
Key Laboratory of Tea Biology and Resources Utilization, Ministry of Agriculture,

Tea Research Institute, Chinese Academy of Agricultural Sciences, 9 Meiling South

NU
Road, Hangzhou, Zhejiang 310008, PR China
MA

2
Agilent Technologies (China) Limited, No. 3 Wangjing North Road, Chaoyang Distr.,

Beijing, 100102, P. R. China

E D

*
Corresponding author: Tel.: +86 571 86653154; E-mail addresses:
PT

tjfcallme@tricaas.com
CE

**
Corresponding author: Tel.: +86 571 86650617; E-mail addresses:
AC

linz@tricaas.com
 ACCEPTED MANUSCRIPT

Abstract

White tea is considered the least processed form of tea and is reported to have a

series of potent bioactivities, such as antioxidant, anti-inflammatory, anti-mutagenic,

and anti-cancer activities. However, the chemical composition of white tea and the

dynamic changes of the metabolites during the manufacturing process are far from

PT
clear. In this study, we applied a nontargeted metabolomics approach based on

RI
ultra-high performance liquid chromatography coupled with quadrupole time-of-flight

SC
mass spectrometry (UHPLC-QTOF/MS) to comprehensively profile the characteristic

metabolites of white tea. There were significant differences in the content of amino
NU
acids, catechins, dimeric catechins, flavonol and flavone glycosides, and aroma
MA

precursors in white tea compared with green and black teas that were manufactured

from the same fresh tea leaves. Furthermore, the dynamic changes of the metabolites
D

in the tea samples with various withering durations of 0, 4, 8, 12, 16, 20, 24, 28, and
E

36 h were also profiled. This study offers a comprehensive characterization of the

PT

metabolites and their changes in white tea.

CE

Keywords: white tea, withering, metabolomics, LC-MS, dynamic change

AC

Chemical compounds studied in this article

Theanine (PubChem, CID: 439378); Glutamic acid (PubChem, CID: 33032);

Quercetin (PubChem, CID: 5280343); Kaempferol (PubChem, CID: 5280863);

Myricetin (PubChem, CID: 5281672); Apigenin (PubChem, CID: 5280443);

Theaflavin (PubChem, CID: 114777); Epigallocatechin gallate (PubChem,

2
 ACCEPTED MANUSCRIPT

CID: 65064); Catechin (PubChem, CID: 73160); Caffeine (PubChem, CID: 2519)

PT
RI
SC
NU
MA
E D
PT
CE
AC

3
 ACCEPTED MANUSCRIPT

1 Introduction

Tea is the most consumed beverage after water in the world because of its health

benefits and satisfactory sensory experience (Ho, Zheng, & Li, 2015; Namita, Mukesh,

& Vijay, 2012; Sharangi, 2009). Compared with green tea and black tea, the two most

popular teas, white tea is a rare form of tea that undergoes the least amount of

PT
processing. Only a prolonged withering process and drying process are included in the

RI
manufacturing of white tea. In the prolonged withering process, a slight fermentation

SC
(oxidation) occurs, which is catalyzed by endogenous polyphenol oxidase (PPO) and

peroxidase (POD) in tea leaves, and these enzymes produce the unique aroma and
NU
taste of white tea (Hashimoto, Goto, Sakakibara, Oi, Okamoto, & Kanazawa, 2007; K.
MA

Wang, Liu, Liu, Huang, Xu, Li, et al., 2011; Yang, Baldermann, & Watanabe, 2013).

Recent studies have revealed that white teas have a series of potent bioactivities,
D

including antioxidant (T. Dias, Tomás, Teixeira, Alves, Oliveira, & Silva, 2013;
E

, Socorro, Silva, & Oliveira, 2014), anti-inflammatory (Thring,

PT

Hili, & Naughton, 2011; L. Zhao, La, & Grenier, 2013), anti-mutagenic (Santana-Rios,
CE

Orner, Amantana, Provost, Wu, & Dashwood, 2001), anti-cancer (Hajiaghaalipour,

AC

Kanthimathi, Sanusi, & Rajarajeswaran, 2015; Mao, Nie, Tsu, Jin, Rao, Lu, et al.,

2010; Shukla, 2007), and neuroprotective activities (Almajano, Vila, & Ginés, 2011;

López & Calvo, 2011).

These bioactivities depend on the unique metabolite phenotype of white tea. There

are some studies that have been carried out to investigate the major compositions of

white tea and have compared white tea with other teas. Alcazar et al. (Alcazar,

4
 ACCEPTED MANUSCRIPT

Ballesteros, Jurado, Pablos, Martín, Vilches, et al., 2007) reported that white teas have

the highest contents of alanine, arginine, asparagine, histidine, isoleucine, leucine,

phenylalanine, serine, and theanine compared with green, black, oolong, and pu-erh

teas, and the teas can be classified using the free amino acid profile combined with a

chemometric method. Santana-Rios et al. (Santana-Rios, Orner, Amantana, Provost,

PT
Wu, & Dashwood, 2001) reported that white tea has higher contents of gallic acid,

RI
theobromine, epigallocatechin (EGC), caffeine, and epicatechin gallate (ECG) and

SC
lower contents of theophylline, catechin (C), and epicatechin (EC) compared with

green tea. Zhao et al. (Y. Zhao, Chen, Lin, Harnly, Yu, & Li, 2011) tentatively
NU
identified 68 compounds in green pu-erh, green, and white teas using the UPLC
MA

combined with diode array detector (DAD) and MS detection, and found that gallic

acid, 1,2,6-trigalloylglucose, and caffeine concentrations were significantly higher in

D

white tea than in the other two teas.

E

However, the chemical constituent investigations in these studies were carried out
PT

using tea samples purchased from markets, which differ in variety, region,
CE

manufacturing procedure, and storage. The initial metabolite contents in fresh tea
AC

leaves are also different in these teas. Therefore, these tea samples might be not

appropriate for the investigations of the impact of the white tea manufacturing process

on tea metabolites (Unachukwu, Ahmed, Kavalier, Lyles, & Kennelly, 2010). In

addition, only a small number of major constituents, such as catechins, caffeine,

theobromine, and free amino acids were compared in these studies. A comprehensive

survey of the metabolome of white tea and its comparison to those of green tea and

5
 ACCEPTED MANUSCRIPT

black tea are urgently needed. Furthermore, the dynamic changes of the metabolome

during the manufacturing process have been less studied. The in-depth mapping of the

dynamic changes of the characteristic metabolites will be helpful for the

manufacturing of high-quality white teas.

Metabolomics, which is emerging as an important part of system biology, detects

PT
dozen, even hundreds, of endogenous metabolites simultaneously. It provides a global

RI
view of the metabolome and has been widely applied in food and tea studies (W. Dai,

SC
Qi, Yang, Lv, Guo, Zhang, et al., 2015; Lee, Lee, Chung, Shin, Lee, Lee, et al., 2011;

Tan, Dai, Lu, Lv, Guo, Zhang, et al., 2016; Xu, Hu, Wang, Wan, & Bao, 2015). In
NU
this work, we manufactured white, green, and black tea from the same fresh tea leaves
MA

and their metabolite profiles were compared by using an UHPLC-QTOF/MS-based

nontargeted metabolomics approach to discover the chemical characteristics of white

D

tea. In addition, the dynamic changes of characteristic metabolites during the

E

manufacturing process of white tea were also profiled.

PT
CE

2 Experimental
AC

2.1 Chemicals

Deionized water was produced by a Milli-Q water purification system (Millipore,

Billerica, Massachusetts). Methanol of LC–MS grade was purchased from Merck

(Darmstadt, Germany). Formic acid, EGC, C, epigallocatechin gallate (EGCG), ECG,

EC, kaempferol 3-O-glucoside, kaempferol 3-O-galactoside, kaempferol

3-O-rutinoside, vitexin, luteolin-8-C-glucoside, isovitexin, quercetin 3-O-glucoside

6
 ACCEPTED MANUSCRIPT

(isoquercitrin), tryptophan, glutamic acid, proline, glutamine, aspartic acid, asparagine,

leucine, isoleucine, threonine, lysine, histidine, valine, arginine, γ-aminobutyric acid

(GABA), myricitrin, kaempferol 3-O-arabinoside, guanosine, quinic acid, chlorogenic

acid, sulfacetamide, sulfafurazole, flumequine, and adenosine were from Sigma (St.

Louis, MO). Myricetin 3-galactoside, procyanidin B1, procyanidin B2, theogallin,

PT
apigenin-6,8-C-diglucoside, theaflavin, theaflavins-3-gallate, and

RI
theaflavins-3,3’-digallate were purchased from Chemfaces (Wuhan, China). Theanine,

SC
quercetin 3-rhamnoglucoside (rutin), quercetin 3-O-galactoside, theobromine,

gallocatechin (GC), tyrosine, and phenylalanine were obtained from J&K Scientific
NU
Ltd. (Beijing, China). Epiafzelechin was purchased from Yuanye Bio-Technology Co.,
MA

Ltd. (Shanghai, China). Caffeine was purchased from Enzo Life Sciences Inc.

(Farmingdale, NY).
D

2.2 Tea sample treatment

E

Clonal tea leaves of the “Fuding Dabaicha” r ty (one bud with three leaves)
PT

were collected, and divided into 3 portions for the manufacturing of white, green, and
CE

black teas (Figure 1). The manufacturing process of white tea was as follows: 40 kg
AC

fresh tea leaves were withered at 30 ºC and a relative humidity of 47% for 36 h, and

then dried at 120 ºC for 20 min and 80 ºC for 20-30 min to a moisture content of

approximately 5%. The manufacturing process of green tea was modified from a

previous work (W. Dai, et al., 2015). Briefly, after withering for 2 h, the tea leaves

were fixed using a rotary continuous fixation machine to terminate the activities of the

endogenous enzymes. Then, the leaves were rolled for 1 h before they were dried at

7
 ACCEPTED MANUSCRIPT

120 °C for 20 min and then at 80 °C for 20−30 min to a moisture content of

approximately 5%. The manufacturing process of black tea was modified from

another previous work (Tan, et al., 2016). After 12 h of withering at 30 ºC and a

relative humidity of 47%, a roller was used to roll the tea leaves for 80 min. Then, the

tea leaves were fermented during 4 h in an environment control cabinet at 30°C and

PT
90% of relative humidity, and then were immediately dried to a moisture content of

RI
approximately 5% to terminate the fermentation using a hot air dryer at 120 °C for 20

SC
min and then at 80 °C for 20−30 min. For the sampling, one part of the tea sample

was collected. The sampling was repeated six times for the manufactured teas (green,
NU
white, and black teas) and was repeated three times for the tea samples with various
MA

withering times (0, 4, 8, 12, 16, 20, 24, and 28 h). 50 milliliters of hot deionized water

(100 °C) was added to 0.3 g ground tea powder (< 0.15 mm) and held at 100 °C for 5
D

min to extract tea metabolites. Then, 1.6 mL of the solution was centrifuged at 10,000
E

g (Centrifuge 5810R, Eppendorf) for 10 min. The supernatants were filtered through a
PT

0.22 μm membrane and were then analyzed by LC–MS. Three internal standards of
CE

sulfacetamide, sulfafurazole, and flumequine (0.2 μg/mL) were spiked into the tea
AC

samples to evaluate the stability during the LC–MS process. In addition, quality

control (QC) samples were prepared by mixing equal amount of each tea sample and

were also used to evaluate the metabolomics process.

2.3 Nontargeted metabolomics analysis

The metabolomics measurements were carried out following procedures we

developed previously (Tan, et al., 2016). Briefly, an UHPLC system (UHPLC Infinity

8
 ACCEPTED MANUSCRIPT

1290, Agilent Tech., Santa Clara, CA) coupled to a Q-TOF mass spectrometer

(Q-TOF 6540, Agilent Tech., Santa Clara, CA) was used for the initial LC-MS data

acquisition. The chromatographic separation of the tea metabolome was conducted on

a Zorbax Eclipse Plus C18 column (100 × 2.1 mm, 1.8 μm, Agilent Tech., Littlefalls,

DE). The column was maintained at a constant temperature of 40 °C. Binary mobile

PT
phases were used for elution at a flow rate of 0.4 mL/min, where solvent A was an

RI
aqueous solution containing 0.1% (v/v) formic acid and solvent B was pure methanol.

SC
The linear gradient elution profile was as follows: 0 min, 10% B; 4 min, 15% B; 7

min, 25% B; 9 min, 32% B; 16 min, 40% B; 22 min, 55% B; 28 min, 95% B; and 30
NU
min, 95% B. The total analysis time for one sample injection was 30 min. 4 min was
MA

allowed for column equilibration between two consecutive injections. The injection

volume was 3 μL. The effluent from the column was detected using a Q-TOF mass
D

spectrometer with a dual jet stream electrospray ionization (ESI) source and operated
E

in the positive mode. The major MS parameters were the same as previously reported
PT

(Tan, et al., 2016). A mass scan range of 100–1100 m/z was applied for the full scan
CE

analysis. The Q-TOF/MS was calibrated daily following the manufacturer's procedure,
AC

and the reference ions with an m/z of 121.0509 (purine) and 922.0098 (hexakis

phosphazine) were continuously infused via the reference sprayer during data

acquisition for online calibration to ensure the MS accuracy. The metabolites were

identified according to authentic standards, accurate mass, MS2 spectra, metabolomics

databases (Metlin and Human Metabolome Database), and previous work (W. Dai, et

al., 2015; Lv, Zhu, Tan, Guo, Dai, & Lin, 2015; Tan, et al., 2016). Raw LC-MS files

9
 ACCEPTED MANUSCRIPT

are available at MetaboLights with accession number MTBLS403.

2.4 Data processing

The raw data files acquired by LC-MS analysis were first processed by the DA

Reprocessor software (Agilent Tech., Santa Clara, CA) to extract the metabolite

feature ions, and then, the data were imported into the Mass Profiler Professional

PT
software (Version 13.0, Agilent Tech., Santa Clara, CA) for peak alignment. The

RI
tolerances of retention time shift and mass shift for peak alignment were set as 0.15

SC
min and 2 mDa, respectively. Ions with a relative standard deviation (RSD) less than

30% in the QC sample analyses were used for further univariate and multivariate
NU
statistics (W. Dai, Yin, Zeng, Kong, Tong, Xu, et al., 2014; W. D. Dai, Wei, Kong, Jia,
MA

Han, Zhang, et al., 2011). Principal component analysis (PCA) was performed using

the Simca-P t r tr , Sweden) after weight normalization

D

and Pareto scaling to investigate the overall tea metabolome variation among the
E

white, green, and black teas and to determine the characteristic metabolites in white
PT

tea. Heat-map analysis and clustering analysis were performed using the
CE

MultiExperiment Viewer (version 4.8.1) to illustrate the dynamic changes of

AC

characteristic metabolites in white tea after the data transformation into the fold

change of the individual mass intensity to average mass intensity. The significance of

the difference of the metabolite between groups was tested using a non-parametric

Kruskal-Wallis H test or a Tukey s-b(K) test with the PASWstat software (Version

18.0, USA).

3 Results and discussion

10
 ACCEPTED MANUSCRIPT

3.1 Analysis of the tea metabolome by UHPLC Q-TOF/MS

As shown in Figure S1-A, the typical total ion current (TIC) chromatograms of

white tea, green tea, and black tea showed visually distinct differences. After peak

alignment, 2137, 1802, and 1719 metabolite ion features were detected in white, green,

and black tea, respectively. Among them, 101 metabolites were structurally identified

PT
(Table S2), and 85 out 101 metabolites were found in three teas simultaneously

RI
(Figure S1-B in Supplementary Material). To evaluate the performance of the LC-MS

SC
analysis, PCA analysis of QC samples and relative standard deviation (RSD)

calculation of three internal standards of sulfacetamide, sulfafurazole, and flumequine

NU
were applied. The QC samples were crowded in the center of the PCA score plot
MA

(Figure S2). The RSD of the retention times, m/z, and mass intensities of the three

internal standards were ranged from 0.06–0.18%, 4.7×10-5–8.1×10-5%, and

D

4.24–5.71%, respectively (Table S1). These results indicated good reproducibility of

E

the LC-MS analyses.

PT

3.2 Characteristics of white tea metabolome

CE

The differences among the metabolomes of white, green, and black tea that were
AC

manufactured from the same batch of tea leaves were investigated using PCA analysis

(Figure 2). The principal components 1 and 2 explained 56.2% and 26.2% of total

variances, respectively, and distinct differences were observed among the three groups

of teas in the PCA score plot (Figure 2-A). The PCA loading plot illustrated the main

differential metabolites in white tea (Figure 2-B).

3.2.1 Amino acids

11
 ACCEPTED MANUSCRIPT

Several previous studies have reported that white tea contains a higher amino acid

content than green and black teas (Alcazar, et al., 2007; Chen, Chen, Zhang, & Wan,

2009). In this study, the free amino acid contents showed drastic variations among

three teas (Table S3). Most of the amino acids, including valine, phenylalanine,

proline, leucine, isoleucine, tryptophan, threonine, lysine, histidine, arginine, and

PT
tyrosine exhibited the highest contents in white tea and the lowest contents in green

RI
tea. The high levels of these amino acids in white tea might be due to the protein

SC
breakdown during the prolonged withering process (Y Lu J g C ’ r y

Singanusong, et al., 2006). By contrast, theanine, glutamic acid, glutamine, and

NU
aspartic acid, which are recognized as main contributors for the umami taste of tea
MA

infusion, showed the lowest contents in white tea and the highest contents in green

tea.
D

3.2.2 Catechins and dimeric catechins

E

Catechins are considered the main compounds that account for the antioxidant
PT

activity of tea. They can be converted to dimeric, oligomeric, and polymeric

CE

compounds, such as theaflavins, theacitrins, theasinensins, theanaphthoquinones, and

AC

thearubigins in the presence of polyphenol oxidase (PPO) and peroxidase (POD)

during the tea fermentation process (Stodt, Blauth, Niemann, Stark, Pawar, Jayaraman,

et al., 2014). Catechins, including EGCG, EGC, EC, ECG, GC, C, and GCG, were at

lower levels in white tea and had the lowest levels in black tea when compared with

those in green tea (Table S3). This result is consistent with the fermentation degree in

that black, white, and green tea is fully fermented, slightly fermented, and

12
 ACCEPTED MANUSCRIPT

non-fermented, respectively. However, these results are different from Santana- ’s

findings, who reported that white tea had higher contents of EGC and ECG compared

with green tea (Santana-Rios, Orner, Amantana, Provost, Wu, & Dashwood, 2001).

This contrast might be due to the inclusion of different varieties of white tea and green

tea in Santana- ’s work, whereas the same batch of tea leaves was included in this

PT
study.

RI
Methylated catechins are naturally present in tea leaves and were recently reported

SC
to have a stronger antiallergic (Fujimura, Umeda, Yano, Maeda-Yamamoto, Yamada,

& Tachibana, 2007; Maeda-Yamamoto, Ema, Monobe, Tokuda, & Tachibana, 2012)
NU
and antihypertensive effects (Kurita, Maeda-Yamamoto, Tachibana, & Kamei, 2010)
MA

than catechins. In this study, interestingly, methylated catechins, such as EC

3-O-(3-O-methylgallate) and EGC 3-methylgallate, exhibited the highest contents in

D

white tea compared with those in green tea and black tea. Because methylated
E

catechins are consumed during the fermentation process (Tan, et al., 2016) (EC
PT

3-O-(3-O-methylgallate) and EGC 3-methylgallate were at a significantly low level in

CE

black tea), it is assumed that the methylation of catechins may occur during the
AC

prolonged withering process of white tea.

However, theaflavins, including theaflavin, theaflavins-3-gallate, and

theaflavins-3,3’-digallate, were at significantly high concentrations in white tea

because of the slight fermentation, but the levels were much lower than those in black

tea (Figure S3). Theasinesins and procyanidins are also important dimeric catechins in

teas, however, the variations in their levels after different manufacturing processes are

13
 ACCEPTED MANUSCRIPT

rarely surveyed. In a previous work on black tea fermentation, the levels of

theasinensins increased sharply within the early period of fermentation and then

slowly decreased, whereas the level of procyanidins reached the maximum within the

first hour of fermentation and then decreased rapidly within 1–6 h to low contents

(Tan, et al., 2016). In this study, theasinesins (theasinesin A and B) had significantly

PT
higher contents in white tea and the highest in black tea compared with those in green

RI
tea (Figure S3), indicating that the fermentation process could be responsible of

SC
theasinesins formation. By contrast, procyanidins (procyanidin B1, B2, B3, B5, and

C1) showed an opposite trend: they had the lowest contents in black tea and lower
NU
contents in white tea compared with those in green tea (Figure S3), indicating that
MA

procyanidins consumption occurs during the withering and fermentation processes.

3.2.3 Flavonol glycosides and flavone glycosides

D

Flavonol glycosides and flavone glycosides are also major phenolic constituents in
E

tea and have strong antioxidative bioactivity and potential benefits for the
PT

cardiovascular system (Wu, Xu, Héritier, & Andlauer, 2012). In this study, flavonol
CE

glycosides and flavone glycosides had different profiles according to their aglycones
AC

and conjugated sugars (Table S3). Kaempferol glycosides, including kaempferol

3-O-glucoside, kaempferol 3-O-galactoside, kaempferol 3-O-rutinoside, kaempferol

3-glucosylrutinoside, kaempferol 3-galactosyrutinoside, kaempferol 3,7-dirhamnoside,

kaempferol 3-O-arabinoside, and kaempferol 3-(6’’-galloylglucoside), showed the

highest levels in white tea compared with those in green tea and black tea. The

quercetin glycosides showed interesting tendencies. Glucosylated quercetin

14
 ACCEPTED MANUSCRIPT

compounds, including quercetin 3-O-glucoside (isoquercitrin), quercetin

3-rhamnoglucoside (rutin), quercetin 3-O-glucosylrutinoside, quercetin diglucoside,

and quercetin triglucoside, presented slightly lower contents in white tea than in green

tea, whereas galactosylated quercetin compounds (quercetin 3-O-galactoside and

quercetin 3-O-galactosylrutinoside) had relatively high contents in white tea.

PT
Myricetin glycosides also showed similar tendencies. In white tea, myricetin

RI
3-glucoside showed a lower content, while myricetin 3-galactoside showed a slightly

SC
higher content. Apigenin-C-monoglycosides, such as vitexin and isovitexin, presented

the highest contents in green tea, whereas apigenin-C-diglycosides, including

NU
apigenin-6-C-glucosyl-8-C-arabinoside, apigenin-6,8-C-diglucoside, and
MA

apigenin-6-C-arabinoside-8-C-glucoside, showed the highest contents in white tea.

3.2.4 Aroma precursors

D

Aroma primeverosides and glucosides are regarded as aroma precursors, which

E

release aroma compounds during the tea manufacturing and brewing processes. In this
PT

study, phenylethyl primeveroside, benzyl primeveroside, linalool primeveroside, and

CE

linalool oxide primeveroside were found in low contents in white tea and were nearly
AC

consumed completely in black tea (Figure S4). The extremely low contents of aroma

precursors in black tea were consistent with the results obtained by Wang et al., who

observed that phenylethyl primeveroside, benzyl primeveroside, and linalool oxide

primeveroside were almost disappeared after the fermentation stage of black tea (D.

M. Wang, Kurasawa, Yamaguchi, Kubota, & Kobayashi, 2001).

3.3 Dynamic changes of metabolites during the manufacturing process of white

15
 ACCEPTED MANUSCRIPT

tea

Long-time withering is the key process for the manufacturing of white tea. In this

study, the dynamic changes of the characteristic metabolites during the withering

process (0, 4, 8, 12, 16, 20, 24, 28, and 36 h) were mapped. The PCA analysis showed

obvious stepwise alterations of the tea metabolome during the withering process from

PT
0 to 36 h (Figure 3). The amino acid contents changed dramatically (Figure 4 and

RI
Table S4). The content of tryptophan, histidine, isoleucine, lysine, phenylalanine,

SC
proline, leucine, valine, and tyrosine increased significantly during the withering

process. The fold changes of the levels of tyrosine, valine, leucine, phenylalanine,
NU
lysine, proline, and isoleucine reached up to > 5. The level of glutamic acid and
MA

aspartic acid decreased significantly during the withering process, particularly during

the first 20 h. Nearly half of their contents were depleted after 36 h of withering.
D

Other characteristic metabolites, such as catechins, theasinensins, procyanidins,

E

theaflavins, flavonol glycosides, flavone glycosides, and aroma precursors, were

PT

clustered into 3 groups in the heat-map (Table S4). The levels of the metabolites in
CE

group I, including TF, TF 3-gallate, TF digallate, theasinensin A, and theasinensin B,

AC

increased significantly during the withering process. These metabolites are dimeric

catechins and their fold changes reached up to >15. Metabolites in group II were

mainly flavonol-O-glycosides and flavone-C-glycosides, and they did not change

drastically. The level of the methylated catechins of EGC 3-methylgallate and EC

3-O-(3-O-methylgallate) increased during the first 28 h and then slightly decreased

during the period of 28-36 h. The amounts of methylated catechins in white tea were

16
 ACCEPTED MANUSCRIPT

more than 1.7-fold than those in fresh leaves (0 h). This finding will be useful for

manufacturing methylated catechin-rich teas. The following compounds decreased

significantly during the withering process: group III metabolites, including catechins

of GC, GCG, C, EGC, and EC; the aroma precursors of benzyl primeveroside,

linalool primeveroside, and linalool oxide primeveroside; and procyanidins. Less than

PT
50% of the content of GC, GCG, and C were retained in the final white tea product

RI
compared with the initial content in fresh leaves. Approximately 59-90% of the

SC
contents of aroma primeverosides were retained in the tea leaves after the 36 h of the

withering process.
NU
4 Conclusion
MA

In this study, white, green, and black teas were manufactured from the same fresh

tea leaves, and their chemical constituents were compared using a

D

UHPLC-QTOF/MS-based non-targeted metabolomics approach. Multivariate and

E

univariate statisticcl analyses suggested great variations in the tea metabolome among
PT

white, green, and black teas. Amino acids, catechins, dimeric catechins, and aroma
CE

precursors were the metabolites with most variation among the three types of tea.
AC

Their dynamic changes during the white tea withering process (0, 4, 8, 12, 16, 20, 24,

28, and 36 h) were also mapped. The levels of the amino acids, tyrosine, valine,

leucine, phenylalanine, lysine, proline, and isoleucine and the dimeric catechins of TF,

TF 3-gallate, TF digallate, theasinensin A, and theasinensin B increased more than

5-fold, whereas the level of the catechins of GC, GCG, and C were reduced by > 50%.

17
 ACCEPTED MANUSCRIPT

Acknowledgements

The authors appreciate the funding support from the National Natural Science

Foundation of China (No. 31500561), the Earmarked Fund for China Agricultural

Research System (No. CARS-23), and the Science and Technology Innovation Project

of Chinese Academy of Agricultural Sciences (No. CAAS-ASTIP-2014-TRICAAS).

PT
RI
SC
NU
MA
E D
PT
CE
AC

18
 ACCEPTED MANUSCRIPT

References

Alcazar, A., Ballesteros, O., Jurado, J., Pablos, F., Martín, M., Vilches, J., &

Navalon, A. (2007). Differentiation of green, white, black, Oolong, and Pu-erh teas

according to their free amino acids content. Journal of Agricultural and Food

Chemistry, 55(15), 5960-5965.

PT
Almajano, M., Vila, I., & Ginés, S. (2011). Neuroprotective effects of white tea

RI
against oxidative stress-induced toxicity in striatal cells. Neurotoxicity Research,

SC
20(4), 372-378.

Chen, L., Chen, Q., Zhang, Z., & Wan, X. (2009). A novel colorimetric
NU
determination of free amino acids content in tea infusions with
MA

2,4-dinitrofluorobenzene. Journal of Food Composition and Analysis, 22(2), 137-141.

Dai, W., Qi, D., Yang, T., Lv, H., Guo, L., Zhang, Y., Zhu, Y., Peng, Q., Xie, D., Tan,
D

J., & Lin, Z. (2015). Nontargeted analysis using ultraperformance liquid

E

chromatography–quadrupole time-of-flight mass spectrometry uncovers the effects of

PT

harvest season on the metabolites and taste quality of tea (Camellia sinensis L.).
CE

Journal of Agricultural and Food Chemistry, 63(44), 9869-9878.

AC

Dai, W., Yin, P., Zeng, Z., Kong, H., Tong, H., Xu, Z., Lu, X., Lehmann, R., & Xu,

G. (2014). Nontargeted modification-specific metabolomics study based on liquid

chromatography–high-resolution mass spectrometry. Analytical Chemistry, 86,

9146-9153.

Dai, W. D., Wei, C., Kong, H. W., Jia, Z. H., Han, J. K., Zhang, F. X., Wu, Z. M.,

Gu, Y., Chen, S. L., Gu, Q., Lu, X., Wu, Y. L., & Xu, G. W. (2011). Effect of the

19
 ACCEPTED MANUSCRIPT

traditional Chinese medicine tongxinluo on endothelial dysfunction rats studied by

using urinary metabonomics based on liquid chromatography-mass spectrometry.

Journal of Pharmaceutical and Biomedical Analysis, 56(1), 86-92.

Dias, T., Tomás, G., Teixeira, N., Alves, M., Oliveira, P., & Silva, B. (2013). White

tea (Camellia sinensis (L.)): antioxidant properties and beneficial health effects.

PT
International Journal of Food Sciences and Nutrition, 2(2), 19-26.

RI
, G. a. D., Socorro, S., Silva, B. M., & Oliveira,

SC
P. F. (2014). White tea as a promising antioxidant medium additive for sperm storage

at room temperature: a comparative study with green tea. Journal of Agricultural and
NU
Food Chemistry, 62(3), 608-617.
MA

Fujimura, Y., Umeda, D., Yano, S., Maeda-Yamamoto, M., Yamada, K., &

Tachibana, H. (2007). The 67kDa laminin receptor as a primary determinant of

D

anti-allergic effects of O-methylated EGCG. Biochemical and Biophysical Research

E

Communications, 364(1), 79-85.

PT

Hajiaghaalipour, F., Kanthimathi, M., Sanusi, J., & Rajarajeswaran, J. (2015).

CE

White tea (Camellia sinensis) inhibits proliferation of the colon cancer cell line,
AC

HT-29, activates caspases and protects DNA of normal cells against oxidative damage.

Food Chemistry, 169, 401-410.

Hashimoto, T., Goto, M., Sakakibara, H., Oi, N., Okamoto, M., & Kanazawa, K.

(2007). Yellow tea is more potent than other types of tea in suppressing liver toxicity

induced by carbon tetrachloride in rats. Phytotherapy Research, 21(7), 668-670.

Ho, C.-T., Zheng, X., & Li, S. (2015). Tea aroma formation. Food Science and

20
 ACCEPTED MANUSCRIPT

Human Wellness, 4(1), 9-27.

Kurita, I., Maeda-Yamamoto, M., Tachibana, H., & Kamei, M. (2010).

Antihypertensive effect of Benifuuki tea containing O-methylated EGCG. Journal of

Agricultural and Food Chemistry, 58(3), 1903-1908.

López, V., & Calvo, M. I. (2011). White tea (Camellia sinensis Kuntze) exerts

PT
neuroprotection against hydrogen peroxide-induced toxicity in PC12 cells. Plant

RI
Foods for Human Nutrition, 66(1), 22-26.

SC
Lee, J.-E., Lee, B.-J., Chung, J.-O., Shin, H.-J., Lee, S.-J., Lee, C.-H., & Hong, Y.-S.

(2011). 1 H NMR-based metabolomic characterization during green tea (Camellia

NU
sinensis) fermentation. Food Research International, 44(2), 597-604.
MA

Lv, H.-P., Zhu, Y., Tan, J.-F., Guo, L., Dai, W.-D., & Lin, Z. (2015). Bioactive

compounds from Pu-erh tea with therapy for hyperlipidaemia. Journal of Functional
D

Foods, 19, 194-203.

E

Maeda-Yamamoto, M., Ema, K., Monobe, M., Tokuda, Y., & Tachibana, H. (2012).
PT

Epicatechin-3-O- 3″-O-methyl)-gallate content in various tea cultivars (Camellia

CE

sinensis L.) and its in vitro inhibitory effect on histamine release. Journal of
AC

Agricultural and Food Chemistry, 60(9), 2165-2170.

Mao, J. T., Nie, W.-X., Tsu, I.-H., Jin, Y.-S., Rao, J. Y., Lu, Q.-Y., Zhang, Z.-F., Go,

V. L. W., & Serio, K. J. (2010). White tea extract induces apoptosis in non–small cell

lung cancer cells: the role of peroxisome proliferator-activated receptor-γ and

15-lipoxygenases. Cancer Prevention Research, 3(9), 1132-1140.

Namita, P., Mukesh, R., & Vijay, K. J. (2012). Camellia Sinensis (green tea): a

21
 ACCEPTED MANUSCRIPT

review. Global Journal of Pharmacology, 6(2), 52-59.

Santana-Rios, G., Orner, G. A., Amantana, A., Provost, C., Wu, S.-Y., & Dashwood,

R. H. (2001). Potent antimutagenic activity of white tea in comparison with green tea

in the Salmonella assay. Mutation Research/Genetic Toxicology and Environmental

Mutagenesis, 495(1), 61-74.

PT
Sharangi, A. (2009). Medicinal and therapeutic potentialities of tea (Camellia

RI
sinensis L.)–A review. Food Research International, 42(5), 529-535.

SC
Shukla, Y. (2007). Tea and cancer chemoprevention: a comprehensive review. Asian

Pacific Journal of Cancer Prevention, 8(2), 155-166.

NU
Stodt, U. W., Blauth, N., Niemann, S., Stark, J., Pawar, V., Jayaraman, S., Koek, J.,
MA

& Engelhardt, U. H. (2014). Investigation of processes in black tea manufacture

through model fermentation (oxidation) experiments. Journal of Agricultural and

D

Food Chemistry, 62(31), 7854-7861.

E

Tan, J., Dai, W., Lu, M., Lv, H., Guo, L., Zhang, Y., Zhu, Y., Peng, Q., & Lin, Z.
PT

(2016). Study of the dynamic changes in the non-volatile chemical constituents of

CE

black tea during fermentation processing by a non-targeted metabolomics approach.

AC

Food Research International, 79, 106-113.

Thring, T. S., Hili, P., & Naughton, D. P. (2011). Antioxidant and potential

anti-inflammatory activity of extracts and formulations of white tea, rose, and witch

hazel on primary human dermal fibroblast cells. Journal of Inflammation, 8(1), 27.

Unachukwu, U. J., Ahmed, S., Kavalier, A., Lyles, J. T., & Kennelly, E. J. (2010).

White and green teas (Camellia sinensis var. sinensis): variation in phenolic,

22
 ACCEPTED MANUSCRIPT

methylxanthine, and antioxidant profiles. Journal of Food Science, 75(6), C541-C548.

Wang, D. M., Kurasawa, E., Yamaguchi, Y., Kubota, K., & Kobayashi, A. (2001).

Analysis of glycosidically bound aroma precursors in tea leaves. 2. Changes in

glycoside contents and glycosidase activities in tea leaves during the black tea

manufacturing process. Journal of Agricultural and Food Chemistry, 49(4),

PT
1900-1903.

RI
Wang, K., Liu, F., Liu, Z., Huang, J., Xu, Z., Li, Y., Chen, J., Gong, Y., & Yang, X.

SC
(2011). Comparison of catechins and volatile compounds among different types of tea

using high performance liquid chromatograph and gas chromatograph mass

NU
spectrometer. International Journal of Food Science & Technology, 46(7), 1406-1412.
MA

Wu, C., Xu, H., Héritier, J., & Andlauer, W. (2012). Determination of catechins and

flavonol glycosides in Chinese tea varieties. Food Chemistry, 132(1), 144-149.

D

Xu, J., Hu, F.-L., Wang, W., Wan, X.-C., & Bao, G.-H. (2015). Investigation on
E

biochemical compositional changes during the microbial fermentation process of Fu

PT

brick tea by LC–MS based metabolomics. Food Chemistry, 186, 176-184.

CE

Yang, Z., Baldermann, S., & Watanabe, N. (2013). Recent studies of the volatile
AC

compounds in tea. Food Research International, 53(2), 585-599.

Yao, L. L u X J g Y C N ’ r y S g u g tt N & Xu

Y. (2006). Compositional analysis of teas from Australian supermarkets. Food

Chemistry, 94(1), 115-122.

Zhao, L., La, V. D., & Grenier, D. (2013). Antibacterial, antiadherence, antiprotease,

and anti-inflammatory activities of various tea extracts: potential benefits for

23
 ACCEPTED MANUSCRIPT

periodontal diseases. Journal of Medicinal Food, 16(5), 428-436.

Zhao, Y., Chen, P., Lin, L., Harnly, J. M., Yu, L., & Li, Z. (2011). Tentative

identification, quantitation, and principal component analysis of green pu-erh, green,

and white teas using UPLC/DAD/MS. Food Chemistry, 126(3), 1269-1277.

PT
RI
SC
NU
MA
E D
PT
CE
AC

24
 ACCEPTED MANUSCRIPT

Figure captions

Figure 1. The manufacturing processes of white, green, and black teas.

Figure 2. Metabolomics analysis of the differences of the tea metabolome of white,

green, and black teas (n = 6): (A) a score plot of the PCA demonstrating distinct

metabolomes for three types of teas. R2X = 0.824, Q2 = 0.780; (B) a loading plot of

PT
the PCA (black triangles with red frames indicate characteristic metabolite features in

RI
white tea).

SC
Figure 3. PCA score plot of tea samples with various withering durations from 0 to 36

h. R2X = 0.821, Q2 = 0.686.

NU
Figure 4. Dynamic changes of amino acid contents during the manufacturing process
MA

of white tea; data are shown as mean ± SD (n = 3). The mass intensity was detected

with 0.3 g tea powder.

E D
PT
CE
AC

25
 ACCEPTED MANUSCRIPT

Figure 1

PT
RI
SC
NU
MA
ED
PT
CE
AC

26
 ACCEPTED MANUSCRIPT

Figure 2

PT
RI
SC
NU
MA
ED
PT
CE
AC

27
 ACCEPTED MANUSCRIPT

Figure 3

PT
RI
SC
NU
MA
ED
PT
CE
AC

28
 ACCEPTED MANUSCRIPT

Figure 4

PT
RI
SC
NU
MA
ED
PT
CE
AC

29
 ACCEPTED MANUSCRIPT

Graphical abstract

PT
RI
SC
NU
MA
DE
PT
CE
AC

30
 ACCEPTED MANUSCRIPT

Highlights
 White, green, and black teas were manufactured from the same fresh tea leaves.
 The three tea metabolomes were compared by a LC-MS based metabolomics
approach.
 Amino acids, catechins, dimeric catechins, and aroma precursors are the most
changeable metabolites.

PT
 Dynamic changes of metabolites during white tea withering process were
profiled.

RI
SC
NU
MA
E D
PT
CE
AC

31