An alternative pathway for the formation of aromatic aroma compounds derived from L-phenylalanine via phenylpyruvic acid in tea (Camellia sinensis (L.) O. Kuntze) leaves

Article abstract of DOI:10.1016/j.foodchem.2018.07.056

Aromatic aroma compounds contribute to flavor of tea (Camellia sinensis (L.) O. Kuntze) and they are mostly derived from L-phenylalanine via trans-cinnamic acid or directly from L-phenylalanine. The objective of this study was to investigate whether an alternative pathway derived from L-phenylalanine via phenylpyruvic acid is involved in formation of aroma compounds in tea. Enzyme reaction with phenylpyruvic acid showed that benzaldehyde, benzyl alcohol, and methyl benzoate were derived from phenylpyruvic acid in tea leaves. Feeding experiments using [²H₈]L-phenylalanine indicated that phenylpyruvic acid was derived from L-phenylalanine in a reaction catalyzed by aromatic amino acid aminotransferases (AAATs). CsAAAT1 showed higher catalytic efficiency towards L-phenylalanine (p ≤ 0.001) while CsAAAT2 showed higher catalytic efficiency towards L-tyrosine (p ≤ 0.001). Both CsAAATs were localized in the cytoplasm of leaf cells. In conclusion, an alternative pathway for the formation of aromatic aroma compounds derived from L-phenylalanine via phenylpyruvic acid occurred in tea leaves.

Full text of DOI:10.1016/j.foodchem.2018.07.056

Accepted Manuscript
An alternative pathway for the formation of aromatic aroma compounds derived
from L-phenylalanine via phenylpyruvic acid in tea (Camellia sinensis (L.) O.
Kuntze) leaves
Xiaoqin Wang, Lanting Zeng, Yinyin Liao, Ying Zhou, Xinlan Xu, Fang Dong,
Ziyin Yang
PII:
S0308-8146(18)31195-6
DOI:
https://doi.org/10.1016/j.foodchem.2018.07.056
FOCH 23171
Reference:
To appear in:
Food Chemistry
Received Date:
Revised Date:
Accepted Date:
18 March 2018
5 July 2018
9 July 2018
Please cite this article as: Wang, X., Zeng, L., Liao, Y., Zhou, Y., Xu, X., Dong, F., Yang, Z., An alternative pathway
for the formation of aromatic aroma compounds derived from L-phenylalanine via phenylpyruvic acid in tea
(
Camellia sinensis (L.) O. Kuntze) leaves, Food Chemistry (2018), doi: https://doi.org/10.1016/j.foodchem.
2
018.07.056
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.

Title: An alternative pathway for the formation of aromatic aroma compounds derived from
L-phenylalanine via phenylpyruvic acid in tea (Camellia sinensis (L.) O. Kuntze) leaves
Running title: Aromatic aroma compounds in tea
a, b, #
a, b, #
a, b
a, b
a
Authors: Xiaoqin Wang
, Lanting Zeng
, Yinyin Liao , Ying Zhou , Xinlan Xu ,
c
a, b, *
Fang Dong , Ziyin Yang
Affiliation:
a
Key Laboratory of South China Agricultural Plant Molecular Analysis and Genetic
Improvement & Guangdong Provincial Key Laboratory of Applied Botany, South China
Botanical Garden, Chinese Academy of Sciences, Xingke Road 723, Tianhe District, Guangzhou
5
10650, China
b
University of Chinese Academy of Sciences, No.19A Yuquan Road, Beijing 100049, China
Guangdong Food and Drug Vocational College, Longdongbei Road 321, Tianhe District,
c
Guangzhou 510520, China
*
Corresponding author: Ziyin Yang, South China Botanical Garden, Chinese Academy of
Sciences, Xingke Road 723, Tianhe District, Guangzhou 510650, China; Tel: +86-20-38072989;
Email address: zyyang@scbg.ac.cn.
#
Co-first authors.
1

Abstract
Aromatic aroma compounds contribute to flavor of tea (Camellia sinensis (L.) O. Kuntze) and
they are mostly derived from L-phenylalanine via trans-cinnamic acid or directly from
L-phenylalanine. The objective of this study was to investigate whether an alternative pathway
derived from L-phenylalanine via phenylpyruvic acid is involved in formation of aroma
compounds in tea. Enzyme reaction with phenylpyruvic acid showed that benzaldehyde, benzyl
alcohol, and methyl benzoate were derived from phenylpyruvic acid in tea leaves. Feeding
2
experiments using [ H
8
]L-phenylalanine indicated that phenylpyruvic acid was derived from
L-phenylalanine in a reaction catalyzed by aromatic amino acid aminotransferases (AAATs).
CsAAAT1 showed higher catalytic efficiency towards L-phenylalanine (p ≤ 0.001) while
CsAAAT2 showed higher catalytic efficiency towards L-tyrosine (p ≤ 0.001). Both CsAAATs
were localized in the cytoplasm of leaf cells. In conclusion, an alternative pathway for the
formation of aromatic aroma compounds derived from L-phenylalanine via phenylpyruvic acid
occurred in tea leaves.
Keywords: Aroma; Camellia sinensis; Phenylalanine; Phenylpyruvic acid; Tea; Volatile
Abbreviations:
AAATs, aromatic amino acid aminotransferases; AIP, 2-aminoindan-2-phosphonic acid; DTT,
dithiothreitol; GC-MS, gas chromatography-mass spectrometry; MSTFA, N-Methyl-N-
(
trimethylsilyl)-trifluoroacetamide; PLP, pyridoxal phosphate; PVPP, polyvinylpolypyrrolidone;
TMEMD, N, N, N’, N’-tetramethylethylenediamine; SDS, sodium dodecyl sulfate; SPME,
solid-phase microextraction; UPLC-QTOF-MS, ultra performance liquidchromatography
2

coupled with quadrupole time-of-flight mass spectrometry; VPBs, volatile phenylpropanoids/
benzenoids.
1
. Introduction
Tea (Camellia sinensis (L.) O. Kuntze) volatile compounds contribute to the aroma property of
tea flavour, and determine the final quality of the tea product. Besides non-volatile compounds
including catechins, amino acids, and caffeine, tea volatile compounds are also representative
specialized metabolites in tea leaves (Wan, & Xia, 2015). Our current knowledge of the
biosynthesis of tea volatiles has been obtained substantially from studies on other plant species.
Similar to volatiles in other plants, tea volatiles can be classified into four major classes
according to their metabolic origin: volatile terpenoids, volatile phenylpropanoids/benzenoids
(
VPBs, also known as aromatic aroma compounds), volatile fatty acid derivatives, and volatiles
derived from carotenoids (Yang, Baldermann, & Watanabe, 2013). Some plant volatile formation
pathways are shared among different plant species, but some differ among plant species because
of the complex networks of plant volatile biosynthesis. Therefore, direct evidence of volatiles
biosynthesis in tea leaves is required. As a genetic transformation system has not yet been
established for tea, it is difficult to study the biosynthetic pathways of specialized metabolites in
tea leaves in detail. Consequently, few enzymes related to tea aroma have been functionally
characterized. -Glycosidases including -primeverosidases and -glucosidases, and glycosyl
transferases, which are involved in the transformation between glycoside-bound volatiles and
free volatiles in tea leaves, have been intensively studied and functionally characterized
(
Mizutani et al., 2002; Zhou et al., 2014; Ohgami et al., 2015). Recently, several enzymes and
genes involved in the final steps of the biosynthesis of tea aroma compounds including
3

(
S)-linalool, (E)-nerolidol, indole, and methyl salicylate, have been isolated, identified, and
functionally characterized (Fu et al., 2015; Zeng et al., 2016; Mei et al., 2017; Zhou et al., 2017;
Liu et al., 2018). In contrast, little attention has been paid to upstream pathways in the formation
of tea aroma compounds, especially VPBs.
The VPBs are the second most ubiquitous class of plant volatiles and have important
ecological functions and potential economic applications, for example, in sedation, improvement
of memory, and improvement of food storage and flavor (Pichersky, Noel, & Dudareva, 2006;
Schwab, Davidovich-Rikanati, & Lewinsohn, 2008). Most volatiles in this class contain an
aromatic ring and originate from shikimate via L-phenylalanine. The well-known upstream
pathway of VPBs formation is the deamination of L-phenylalanine to trans-cinnamic acid,
catalyzed by L-phenylalanine ammonia lyase (Pichersky, Noel, & Dudareva, 2006; Schwab,
Davidovich-Rikanati, & Lewinsohn, 2008) (Figure 1A). However, the two VPBs
phenylacetaldehyde and 2-phenylethanol are not produced from trans-cinnamic acid and are
directly derived from L-phenylalanine (Tieman et al., 2006) (Figure 1A). Recent studies have
indicated that phenylpyruvic acid may also be involved in the formation of VPBs, as validated in
melon, rose flowers, and tea flowers (Gonda et al., 2010; Dong et al., 2012; Hirata et al., 2012).
However, it is still unknown whether phenylpyruvic acid is involved in the formation of VPBs in
tea leaves. To answer this question, we supplied different substrates to enzyme extracts of tea
leaves, and found that benzaldehyde, benzyl alcohol, and methyl benzoate were derived from
2
phenylpyruvic acid in tea leaves. Then, stable isotope-labeled [ H
8
]L-phenylalanine was supplied
to tea leaves to investigate the phenylpyruvic acid pathway. Finally, the gene involved in
phenylpyruvic acid formation was isolated, cloned, sequenced, and functionally characterized.
The aim of this study was to discover the biosynthesis of phenylpyruvic acid in tea leaves and
4

the role of phenylpyruvic acid in the formation of tea aroma, especially VPBs.
2
. Materials and methods
2
.1. Chemicals and regents
2
-Aminoindan-2-phosphonic acid (AIP) was purchased from Absin Bioscience Inc., Shanghai,
China. Benzaldehyde, benzyl alcohol, acetophenone, 2-phenylethanol, methyl salicylate,
trans-cinnamic acid, and ethyl decanoate were purchased from Wako Pure Chemical Industries
Ltd., Japan. EasySee Western Marker (25-90 kDa) was purchased from TransGen Biotech,
Beijing. ECL kit, Precision Plus Protein standard (10-250 kDa), 2 × SYBR Green Universal PCR
Mastermix, N, N, N’, N’-tetramethylethylenediamine (TMEMD), 30% acrylamide/bis solution,
and sodium dodecyl sulfate (SDS) were purchased from Bio-Rad Laboratories, Hercules, CA,
2
USA. [,,,2,3,4,5,6- H
8
]L-Phenylalanine was purchased from Cambridge Isotope Laboratories
Inc., Cambridge, MA, USA. Methyl benzoate was purchased from Shanghai Macklin
Biochemical Co., Ltd., China. Ni-NTA resin was purchased from Qiagen Inc., USA.
N-Methyl-N-(trimethylsilyl)-trifluoroacetamide (MSTFA) was purchased from Regis Chemical
Co., Morton Grove, Illinois, USA. PD-10 desalting column was purchased from GE Healthcare
Life Sciences, Chandler, Arizona, USA. Phenylpyruvic acid and polyvinylpolypyrrolidone
(
PVPP) were purchased from Sigma-Aldrich Company Ltd., St. Louis, MO, USA. Quick RNA
isolation kit was purchased from Huayueyang Biotechnology Co., Ltd., China.
2
.2. Plant materials
Tea plants of C. sinensis cv. Jinxuan were widely cultivated in South China and the experiment
5

materials were acquired from Yingde Tea Experimental Station of Tea Research Institute,
Guangdong Academy of Agricultural Sciences (23°N, 113°E, Yingde, China).
2
.3. Extraction of enzymes from tea leaves and analyses of enzymatic products
The L-phenylalanine transaminase assay was conducted according to a literature method
(
Gonda et al., 2010) with slight modifications. The detailed enzyme extraction method is shown
in the ‘Supplementary Information’. A reaction mixture (400 μL) containing 5 mM
2
[
8
H ]L-phenylalanine and crude enzyme solution (350 μL) was incubated at 30 °C for 60 min,
then acidified with 1 M hydrochloric acid (80 μL), and then incubated for another 30 min at
0 °C. The samples were extracted with ethyl acetate (800 μL), and then centrifuged at 2,500 g
3
to separate the phases. The organic phase was collected, blown dry under a stream of nitrogen
gas and derivatized with MSTFA (80 μL) at 37 °C for 60 min. The MSTFA-derivated sample (1
μL) was analyzed by gas chromatography–mass spectrometry (GC-MS) (The GC-MS analysis
condition is described in the ‘Supplementary Information’).
The phenylpyruvic acid and trans-cinnamic acid conversion assays were conducted as
described previously (Hirata et al., 2012) with slight modifications. The detailed enzyme
extraction method is shown in the ‘Supplementary Information’. A reaction mixture (400 μL)
containing 5 mM phenylpyruvic acid or trans-cinnamic acid (dissolved in 5% (v/v) methanol),
and crude enzyme solution (350 μL) was incubated at 30 °C for 60 min, and then placed on ice
for 10 min to stop the reaction. To determine internal volatiles, the mixture was extracted with an
equal volume hexane/ethyl acetate (1:1) containing ethyl decanoate (5 nmol). Headspace
volatiles were collected using a solid-phase microextraction (SPME) fiber (2 cm-50/30 μm
TM
TM
DVB/Carboxen / PDMS Stable Flex , Supelco Inc., Bellefonte, PA, USA). The SPME fiber
6

or 1 μL of the extract was subjected to GC-MS analysis (The GC-MS analysis condition is
described in the ‘Supplementary Information’).
2
2
2
.4. Supplementation of [ H
8
]L-phenylalanine or cosupplementation of [ H ]L-phenylalanine
8
and AIP to tea leaves and identification of deuterium labeled products
Individual tea sample (one bud and two or three leaves) was injected with 4 μmol
2
2
[
H
8
]L-phenylalanine and 0.4 μmol AIP. Samples treated only with [ H
8
]L-phenylalanine served
as the control group. All samples were kept in an incubator under a 16-h light/8-h dark
photoperiod at 70% humidity and 25 °C for 3 days, and then immediately frozen in liquid
nitrogen and stored at -80 °C until further analyses. Four independent replicates were performed.
The labeled phenylpyruvic acid product was identified using a literature method (Dong et al.,
2
012) with slight modifications. Tea samples (0.8 g, finely powdered) were extracted with cold
methanol (4 mL) by vortexing for 2 min followed by ultrasonic extraction in ice cold water for
0 min. Each extract was mixed with cold chloroform (4 mL) and cold water (1.6 mL), and the
1
phases were allowed to separate. The upper layer was blown dry under a stream of nitrogen gas
to remove the organic solvent. The water extract was mixed with 0.25 M hydrochloric acid (50
μL) and extracted twice with ethyl acetate (1.0 mL). The ethyl acetate extract was blown dry
under a stream of nitrogen gas and derivatized with MSTFA (80 μL) at 37 °C for 60 min. The
MSTFA-derivated samples were then analyzed by GC-MS (The GC-MS analsis condition is
described in the ‘Supplementary Information’).
To analyze the labeled volatiles, tea samples (200 mg, finely powdered) were extracted by 1
mL dichloromethane containing of ethyl decanoate (5 nmol) as an internal standard in a shaker at
room temperature for 6 hours. The extraction solution was dried over anhydrous sodium sulfate
7

and then a 1-μL aliquot was subjected to GC-MS analysis (The GC-MS analysis condition is
described in the ‘Supplementary Information’).
2
.5. Phylogenetic analyses of CsAAATs and AAATs from various plant species
The protein sequence of Cucumis melo L. CmArAT1, a phenylalanine transaminase (GenBank:
FJ896816.1), was used for a BlastN searching in the Transcriptome Shotgun Assembly (TSA)
database of C. sinensis. Two aromatic amino acid transaminases members, CsAAAT1 (accession
no: MH544095) and CsAAAT2 (accession no: MH544096), were highly familiar and selected
from the TSA database online. Phylogenetic analysis was conducted and viewed using MEGA
7
.0 software on the basis of the neighbor–joining algorithms with a bootstrap test of 1000
replicates.
2
.6. CsAAATs recombinant expression and western blot analyses
The full length open reading frames (ORFs) of CsAAAT1 and CsAAAT2 were amplified by
PCR with the primers listed in Table S1 (Supplementary Information). For CsAAATs
recombinant expression, the detailed method is described in the ‘Supplementary Information’.
Western blot analyses were performed according to a literature method (Mahmood, & Yang,
2
012). Protein samples were electrophoresed on a SDS–polyacrylamide gel containing a
separation part (2700 μL of ddH O, 2400 μL of 30% acrylamide/bis (acrylamide), 1800 μL of
Tris-HCl (pH 8.8), 70 μL of 10% SDS, 70 μL of 10% ammonium persulfate, and 3 μL of
TMEMD) and a concentration part (1400 μL of ddH O, 340 μL of 30% acrylamide/bis
acrylamide), 250 μL of Tris-HCl (pH 6.8), 20 μL of 10% SDS, 20 μL of 10% ammonium
2
2
(
persulfate, and 2 μL of TMEMD). Precision Plus Protein standard (10-250 kDa) and EasySee
8

Western Marker (25-90 kDa) were used as molecular weight marker. After electrophoresis,
proteins were stained with Coomassie Brilliant Blue R-250 and destained with dye-destaining
solution. Following SDS-PAGE analysis, the proteins were transferred from a gel onto a
membrane (polyvinylidine fluoride (PVDF), 0.2 μm pore size; NuPAGE®) and then labeled with
a His-Tag mouse monoclonal antibody (Signalway antibody, Pearland, TX, USA) at a dilution of
1
:5000, continued with a HRP-conjugated goat anti-mouse IgG (Signalway antibody, Pearland,
TX, USA) diluted at 1:10000. Immunoreactive spots were detected using the ECL kit (Bio-Rad,
Hercules, CA, USA) according to the manufacturer's instructions.
2
.7. CsAAATs recombinant enzyme assay and substrate selectivity
CsAAATs recombinant enzyme assay was conducted according to a literature method (Gonda
et al., 2010) with slight modifications.The reaction mixture (1 mL) contained purified protein (25
μg of protein), and 50 mM BIS-TRIS propane buffer (pH 8.5) containing 10 mM α-ketoglutarate,
1
0 mM L-phenylalanine or L-tyrosine, 1 mM dithiothreitol (DTT), 225 μM pyridoxal phosphate
(
(
PLP), and 10% (w/v) D-sorbitol. The reactions were incubated at 30 °C for 2 h. The products
phenylpyruvic acid derived from L-phenylalanine and 4-hydroxyphenylpyruvic acid derived
from L-tyrosine) were extracted twice with ethyl acetate (2  1 mL). The ethyl acetate extract
was blown dry under a stream of nitrogen gas, and then re-dissolved in acetonitrile (200 μL).
After filtering through a 0.22-μm membrane, the ten-fold diluted extract was analyzed by an
ultra-performance liquid chromatography/quadrupole time-of-flight mass spectrometry
(
UPLC-QTOF-MS; ACQUITY UPLC I-Class/Xevo® G2-XS QTOF, Waters Corp., Milford,
MA, USA). The UPLC-QTOF-MS analysis conditions are described in the ‘Supplementary
Information’. The compounds were identified from their retention time and MS data
9

−
(
phenylpyruvic acid, t
R
= 5.13 min, m/z 163.0395 [M-H] ; 4-hydroxyphenylpyruvic acid, t
R
=
−
3
2
2
.12 min, m/z 179.0344 [M-H] ).
.8. Functional identification and subcellular localization of CsAAATs in vivo
.8.1. Agrobacterium-mediated transient expression of CsAAATs in Nicotiana benthamiana
The primers used to amplify the sequences to express the GFP-fusion protein are listed in
Table S1 (Supplementary Information). The reorganization vectors were transformed into
Agrobacterium GV3101 by electroporation according to the manufacturer’s instructions. The
detailed method is described in the ‘Supplementary Information’.
2
.8.2. Analyses of CsAAATs activities in overexpressed Nicotiana benthamiana
To analyze CsAAATs activities in overexpressed Nicotiana benthamiana, tobacco leaves (0.8 g)
of Nicotiana benthamiana (finely powdered) were used for phenylpyruvic acid extraction and
detection using the method described above.
2
.8.3. Subcellular localization of CsAAATs
2
Tobacco leaves were cut into 1-2 cm pieces, and confocal microscope images were taken
using a Zeiss LSM 510 META confocal laser microscope (Carl Zeiss, Jena, Germany) with a 40×
water objective under an excitation wavelength of (488 nm) for GFP fluorescence observation.
2
.9. Transcript expression analyses of CsAAATs in different tissues
Total RNA was extracted from roots, stems, young leaves, mature leaves, and flowers of C.
10

sinensis, respectively, using a Plant Total RNA Kit (Huayueyang Biotechnology CO., Ltd.,
China). The first-strand cDNA was synthesized using a PrimeScript RT Reagent Kit with gDNA
Eraser (Takara Industry CO., Ltd., Japan). Gene transcript expression was quantified using
quantitative real-time PCR (qRT-PCR). The reaction (20 μL) was consisted of iTaq Universal
SYBR Green Supermix (10 μL, 2×, Bio-Rad, Hercules, CA, USA), the specific forward and
reverse primers (0.2 μM each), and the 20-fold diluted template (2 μL). The encoding elongation
factor 1 (EF1) was used as an internal reference gene (Hao et al., 2014). The CsEF1, CsAAAT1
and CsAAAT2 gene specific primers of qRT-PCR are listed in Table S1 (Supplementary
Information). The qRT-PCR was conducted on a Roche LightCycle 480 system (Roche Applied
Science, Mannheim, Germany) under the following conditions: One cycle of 95 °C for 60 s
followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s. A melt curve was obtained for each
sample at the end of each run to ensure the purity of the amplified products. The 2C
method
T
was used to measure the relative expression level (Livak et al., 2001). Changes in mRNA levels
of CsAAAT1 and CsAAAT2 were normalized to that of CsEF1. Each tissue point represents an
average of three independent biological samples.
2
.10. Statistical analysis
Statistical analysis was performed using SPSS Statistics 23.0 software. The statistical
significances of the differences between two treatments were calculated using a two-tailed
Student’s t-test (* p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001). One-way ANOVA followed by Duncan’s
multiple comparison tests was used to determine the differences among three or more than three
groups. A probability level of 5% (p ≤ 0.05) was defined as significant.
11

3
. Results and discussion
3
.1. Benzaldehyde, benzyl alcohol, and methyl benzoate were derived from phenylpyruvic acid in
tea leaves
Based on previous reports (Dudareva, Negre, Nagegowda, & Orlova, 2006; Koeduka et al.,
006), the upstream pathway of VPBs formation is mainly via trans-cinnamic acid. However,
recent studies (Orlova et al., 2006; Yoo et al., 2013; de la Torre, El-Azaz, Ávila, & Cánovas,
014) have suggested that there might be an alternative pathway from another precursor,
2
2
possibly phenylpyruvic acid (Figure 1A). To investigate whether phenylpyruvic acid is involved
in the formation of VPB compounds in tea leaves, we prepared a crude enzyme extract from tea
leaves and monitored the reactions using trans-cinnamic acid and phenylpyruvic acid as
substrates. Then, we detected the changes in the contents of the related aroma compounds. The
contents of 2-phenylethanol, acetophenone, and methyl salicylate increased when using
phenylpyruvic acid as a substrate, but no statistically significant differences were detected
between the CK and treatment group (Figure S1). However, benzaldehyde, benzyl alcohol, and
methyl benzoate significantly accumulated in the reaction using phenylpyruvic acid as the
substrate (Figure 1B), whereas the contents of these three compounds were almost unchanged
when using trans-cinnamic acid as the substrate (Figure 1C). These results provided in vitro
evidence that three important VPBs (benzaldehyde, benzyl alcohol, and methyl benzoate) are
derived from phenylpyruvic acid in tea leaves. Because phenylpyruvic acid is unstable, an
isotopically labeled form of this compound is not commercially available. To further obtain in
vivo evidence, we extracted and analyzed the aroma compounds from tea leaves after
2
2
supplementation with [ H
8
]L-phenylalanine or co-supplementation with [ H ]L-phenylalanine
8
12

and AIP, a potent inhibitor of phenylalanine ammonia lyase. Labeled benzaldehyde was detected
2
in tea leaves after supplementation with [ H
8
]L-phenylalanine ( Figure S2). Howevere, there was
no obvious difference in the amounts of labeled benzaldehyde and benzyl alcohol between the
2
two treatment groups including supplementation with [ H
8
]L-phenylalanine and
2
co-supplementation with [ H
8
]L-phenylalanine and AIP (Figure 1D), and labeled methyl
benzoate was not detected. The in vivo test further confirmed that phenylpyruvic acid was the
precursor of benzaldehyde and benzyl alcohol in tea leaves. Labeled methyl benzoate was not
detected, possibly because it was beneath the limits of detection. In most plants, many important
VPBs are derived from trans-cinnamic acid (Dudareva, Negre, Nagegowda, & Orlova, 2006).
However, several studies have shown that phenylpyruvic acid is also likely to be an important
precursor of VPB compounds in some plants, such as melon (Cucumis melo) (Gonda et al., 2010),
rose flowers (Rosa × hybrida ‘Yves Piaget’) (Hirata et al., 2012), and tea flowers (Dong et al.,
2
012). Our study is the first confirmation that phenylpyruvic acid is an important precursor of
VPB compounds in tea leaves.
Benzaldehyde and benzyl alcohol are two prominent scent compounds that greatly contribute
to the fruity, floral smells of flowers, fruits, and plants. Benzaldehyde has a typical almond-like
odor and is one of the key volatile components in all types of tea, especially green tea, oolong tea,
and black tea (Yang, Baldermann, & Watanabe, 2013). Benzyl alcohol also largely contributes to
the floral aroma of oolong tea and black tea (Mizutani et al., 2002). Methyl benzoate is one of
essential floral compounds of many bee-pollinated flowers, for instance snapdragon flowers
(
Pichersky, & Gershenzon, 2002). However, the upstream pathway in the synthesis of
benzaldehyde, benzyl alcohol, and methyl benzoate remains unclear. In micro-organisms,
especially yeasts, phenylpyruvic acid is the precursor for benzaldehyde, benzyl alcohol, and
13

methyl benzoate, which contributes to the flavor of cheese and alcohol (Hazelwood, Daran, van
Maris, Pronk, & Dickinson, 2008; Etschmann, Bluemke, Sell, & Schrader, 2002). Some reports
have speculated that these compounds may be derived from phenylpyruvic acid in plants (Orlova
et al., 2006; Oliva et al., 2017) like in microorganisms, but there is no direct evidence for this.
Our study has provided the first evidence that phenylpyruvic acid was the precursor of
benzaldehyde, benzyl alcohol, and methyl benzoate in tea leaves.
3
.2. Phenylpyruvic acid was derived from L-phenylalanine in tea leaves
To investigate whether phenylpyruvic acid is derived from L-phenylalanine in tea leaves, we
carried out both in vitro and in vivo tests. First, we conducted the in vitro crude enzyme reaction
2
using [ H
8
]L-phenylalanine as the substrate and analyzed the products. Labeled phenylpyruvic
acid was detected by GC-MS, based on the detection of peaks of the characteristic ion (m/z 225,
99 and 314) compared with the retention time of authentic phenylpyruvic acid. The labeled
phenylpyruvic acid content was calculated from the peak area of the characteristic ion m/z 299
Figure 2A). The results revealed that L-phenylalanine was converted to phenylpyruvic acid by
2
(
enzymes in tea leaves. To confirm this result, we carried out an in vivo experiment by feeding the
2
stable isotope [ H
8
]L-phenylalanine to tea leaves. Labeled phenylpyruvic acid was also detected
2
after supplementation of [ H
8
]L-phenylalanine to tea leaves (Figure 2B & 2C), consistent with
the isotopic tracer results from rose flowers and tea flowers (Dong et al., 2012; Hirata et al.,
012). Since a genetic transformation system has not yet been established for tea, it is very
2
difficult to obtain direct in vivo evidence of secondary metabolite synthesis. Therefore, the use of
stable isotope-labeled secondary metabolites of precursors for in vivo tracing can provide
sensitive and direct evidence of metabolic networks in plants for which no genetic
14

transformation system is available (Chokkathukalam, Kim, Barrett, Breitling, & Creek, 2014).
The stable isotope labeling method has been successfully used to analyze tea secondary
metabolites pathways in other studies (Zeng et al., 2016; Cheng et al., 2017). In our study, we
found that phenylpyruvic acid was produced from L-phenylalanine in tea leaves via enzyme
assays and stable isotope labeling experiments.
3
.3. CsAAAT1 was involved in the pathway leading from L-phenylalanine to phenylpyruvic acid
in tea leaves
To further explore the key genes and enzymes involved in the formation of phenylpyruvic acid
from L-phenylalanine, we conducted BlastN searches to find putative CsAAATs sequences in C.
sinensis. Data mining of the TSA database revealed two unique aromatic amino acid
transaminases members, CsAAAT1 and CsAAAT2 (Figure 3). CsAAAT1 was more closely related
to CmArAT1, which encodes an L-phenylalanine transaminase (Figure 3). The two recombinant
proteins were overexpressed in BL21 E. coli competent cells (Figure 4A). The CsAAAT1
sequence was a full-length clone encoding a 421-amino-acid protein with a native molecular
mass of 46.75 kDa, and the CsAAAT2 sequence was a full-length clone encoding a
4
21-amino-acid protein with a native molecular mass of 46.2 kDa, as predicted by ExPASy
(
https://www.expasy.org/). Figure 4B shows the western blot analyses of CsAAAT1 and
CsAAAT2 expressed successfully in E. coli. The semi-purified CsAAAT1 and CsAAAT2 were
able to catalyze the L-phenylalanine transamination reaction to synthesize phenylpyruvic acid.
Furthermore, CsAAAT1 showed higher L-phenylalanine transamination activity while CsAAAT2
showed higher tyrosine transamination activity (Figure 5A).
To obtain in vivo evidences of CsAAATs functions, CsAAAT1 and CsAAAT2 were introduced
15

into the A. tumefaciens strain GV3101 and transiently overexpressed in leaves of N. benthamiana.
Both enzymes showed phenylpyruvic acid production activity in vivo (Figure 5B). We measured
the transcript levels of CsAAAT1 and CsAAAT2 in different tea tissues by qRT-PCR, and detected
higher levels of CsAAAT1 transcripts in mature leaves, young leaves, and flowers (Figure S3A),
and higher levels of CsAAAT2 transcripts in flowers (Figure S3B). In subcellular localization
analyses, both CsAAAT1 and CsAAAT2 were located in the cytoplasm of leaf cells (Figure 5C).
The AAATs can be divided into three families; phenylalanine aminotransferases, tryptophan
aminotransferases, and tyrosine aminotransferases (Wang, & Maeda, 2017). Several AAATs have
been reported in plants (Lopukhina, Dettenberg, Weiler, & Holländer-Czytko, 2001; Stepanova et
al., 2008; Gonda et al., 2010). The AAATs are PLP-dependent enzymes (Alexander, Sandmeier,
Mehta, & Christen, 1994) that catalyze the transamination reaction of aromatic amino acids to
their corresponding keto acids, and the products of AAATs in plants serve as key precusors of
secondary metabolites, such as the hormone auxin, phenylpropanoids, and alkaloids (Facchini,
2
001; Vogt, 2010). Phenylalanine aminotransferases are a subset of the tyrosine aminotransferase
family. They are localized outside of plastids, and interlink phenylalanine and tyrosine
metabolism. In our study, we sequenced and identified two AAATs that were localized in the
cytoplasm of leaf cells. A phylogenetic analysis and substrate selectivity assays showed that
CsAAAT1 was a phenylalanine aminotransferase while CsAAAT2 was a tyrosine
aminotransferase. These results showed that CsAAAT1 was the key enzyme catzlyzing the
conversion of L-phenylalanine into phenylpyruvic acid in tea leaves.
4
. Conclusions
Our results showed that benzaldehyde, benzyl alcohol, and methyl benzoate in tea leaves were
16

derived from L-phenylalanine via phenylpyruvic acid (Figure 6). Two unique CsAAATs were
sequenced and functionally characterized. Both in vitro and in vivo analyses verified that the two
CsAAATs had L-phenylalanine transaminase functions. Substrate selectivity assays showed that
CsAAAT1 was a L-phenylalanine aminotransferase and CsAAAT2 was a L-tyrosine
aminotransferase. Subcellular localization analyses of CsAAATs showed that both were located
in the cytoplasm of leaf cells. Our study clarifies the pathway of volatile
phenylpropanoids/benzenoids synthesis in plants, and is the first direct evidence that
benzaldehyde, benzyl alcohol, and methyl benzoate are derived from phenylpyruvic acid in tea
leaves.
Acknowledgments
This study was supported by grants from the National Natural Science Foundation of China
(
31601787), the Guangdong Natural Science Foundation (2016A030313652 and
016A030306039), the Guangdong Innovation Team of Modern Agricultural Industry
2
Technology System (2017LM1143), and the Guangdong Special Support Plan for Training
High-Level Talents (2016TQ03N617).
Competing financial interests
The authors declare no competing financial interests.
References
Alexander, F. W., Sandmeier, E., Mehta, P. K., & Christen, P. (1994). Evolutionary
relationships among pyridoxal-5’-phosphate-dependent enzymes. The FEBS
17

Journal, 219, 953-960.
Cheng, S. H., Fu, X. M., Wang, X. Q., Liao, Y. Y., Zeng, L. T., Dong, F., & Yang, Z. Y.
(
2017). Studies on the biochemical formation pathway of the amino acid L-
theanine in tea (Camellia sinensis) and other plants. Journal of Agricultural and
Food Chemistry, 65, 7210-7216.
Chokkathukalam, A., Kim, D. H., Barrett, M. P., Breitling, R., & Creek, D. J. (2014). Stable
isotope-labeling studies in metabolomics: new insights into structure and dynamics of
metabolic networks. Bioanalysis, 6, 511-524.
de la Torre, F., El-Azaz, J., Ávila, C., & Cánovas, F. M. (2014). Deciphering the role of aspartate
and prephenate aminotransferase activities in plastid nitrogen metabolism. Plant Physiology,
1
64, 92-104.
Dong, F., Yang, Z.Y., Baldermann, S., Kajitani, Y., Ota, S., Kasuga, H., ... Watanabe, N. (2012).
Characterization of L-phenylalanine metabolism to acetophenone and 1-phenylethanol in
the flowers of Camellia sinensis using stable isotope labeling. Journal of Plant Physiology,
1
69, 217-225.
Dudareva, N., Negre, F., Nagegowda, D. A., & Orlova, I. (2006). Plant volatiles: recent advances
and future perspectives. Critical Reviews in Plant Sciences, 25, 417-440.
Etschmann, M., Bluemke, W., Sell, D., & Schrader, J. (2002). Biotechnological production of
2
-phenylethanol. Applied Microbiology and Biotechnology, 59, 1-8.
Facchini, P. J. (2001). Alkaloid biosynthesis in plants: biochemistry, cell biology, molecular
regulation, and metabolic engineering applications. Annual Review of Plant Biology, 52,
2
9-66.
Fu, X. M., Chen, Y. Y., Mei, X., Katsuno, T., Kobayashi, E., Dong, F., ... Yang, Z. Y. (2015).
18

Regulation of formation of volatile compounds of tea (Camellia sinensis) leaves by single
light wavelength. Scientific Reports, 5, 16858.
Gonda, I., Bar, E., Portnoy, V., Lev, S., Burger, J., Schaffer, A.A., ... Lewinsohn, E. (2010).
Branched-chain and aromatic amino acid catabolism into aroma volatiles in Cucumis melo
L. fruit. Journal of Experimental. Botany, 61, 1111-1123.
Hao, X., Horvath, D. P., Chao, W. S., Yang, Y., Wang, X., & Xiao, B. (2014). Identification and
evaluation of reliable reference genes for quantitative real-time PCR analysis in tea plant
(
Camellia sinensis (L.) O. Kuntze). International Journal of Molecular Sciences, 15,
2155-22172.
2
Hazelwood, L. A., Daran, J. M., van Maris, A. J., Pronk, J. T., & Dickinson, J. R. (2008). The
Ehrlich pathway for fusel alcohol production: a century of research on Saccharomyces
cerevisiae metabolism. Applied and Environmental Microbiology, 74, 2259-2266.
Hirata, H., Ohnishi, T., Ishida, H., Tomida, K., Sakai, M., Hara, M., & Watanabe, N. (2012).
Functional characterization of aromatic amino acid aminotransferase involved in
2
-phenylethanol biosynthesis in isolated rose petal protoplasts. Journal of Plant Physiology,
69, 444-451.
1
Koeduka, T., Fridman, E., Gang, D. R., Vassão, D. G., Jackson, B. L., Kish, C. M., ... Pichersky,
E. (2006). Eugenol and isoeugenol, characteristic aromatic constituents of spices, are
biosynthesized via reduction of a coniferyl alcohol ester. Proceedings of the National
Academy of Sciences, 103, 10128-10133.
Liu, G.F., Liu, J.J., He, Z.R., Wang, F.M., Yang, H., Yan, Y.F., ... Wei, S. (2018). Implementation
of CsLIS/NES in linalool biosynthesis involves transcript splicing regulation in Camellia
sinensis. Plant, Cell & Environment, 41, 176-186.
19

Livak, K. J., & Schmittgen, T. D. (2001). Analysis of relative gene expression data using
real-time quantitative PCR and the 2C
T
method. Methods, 25, 402-408.
Lopukhina, A., Dettenberg, M., Weiler, E. W., & Holländer-Czytko, H. (2001). Cloning and
characterization of a coronatine-regulated tyrosine aminotransferase from Arabidopsis.
Plant Physiology, 126, 1678-1687.
Mahmood, T., & Yang, P. C. (2012). Western blot: technique, theory, and trouble shooting. North
American Journal of Medical Sciences, 4: 429.
Mei, X., Liu, X.Y., Zhou, Y., Wang, X.Q., Zeng, L.T., Fu, X.M., ... Yang, Z.Y. (2017). Formation
and emission of linalool in tea (Camellia sinensis) leaves infested by tea green leafhopper
(
Empoasca (Matsumurasca) onukii Matsuda. Food Chemistry, 237, 356-363.
Mizutani, M., Nakanishi, H., Ema, J. I., Ma, S. J., Noguchi, E., Inohara-Ochiai, M., ... Sakata, K.
2002). Cloning of -primeverosidase from tea leaves, a key enzyme in tea aroma formation.
(
Plant Physiology, 130, 2164-2176.
Ohgami, S., Ono, E., Horikawa, M., Murata, J., Totsuka, K., Toyonaga, H., ... Ohnishi, T. (2015).
Volatile glycosylation in tea plants: sequential glycosylations for the biosynthesis of aroma
β-primeverosides are catalyzed by two Camellia sinensis glycosyltransferases. Plant
Physiology, 168, 464-477.
Oliva, M., Bar, E., Ovadia, R., Perl, A., Galili, G., Lewinsohn, E., & Oren-Shamir, M. (2017).
Phenylpyruvate contributes to the synthesis of fragrant benzenoid-phenylpropanoids in
Petunia × hybrida flowers. Frontiers in Plant Science, 8, 769.
Orlova, I., Marshall-Colón, A., Schnepp, J., Wood, B., Varbanova, M., Fridman, E., ... Dudareva,
N. (2006). Reduction of benzenoid synthesis in petunia flowers reveals multiple pathways
to benzoic acid and enhancement in auxin transport. The Plant Cell, 18, 3458-3475.
20

Pichersky, E., & Gershenzon, J. (2002). The formation and function of plant volatiles: perfumes
for pollinator attraction and defense. Current Opinion in Plant Biology, 5, 237-243.
Pichersky, E., Noel, J.P., & Dudareva, N. (2006). Biosynthesis of plant volatiles: nature's
diversity and ingenuity. Science, 331, 808-811.
Schwab, W., Davidovich-Rikanati, R., & Lewinsohn, E. (2008). Biosynthesis of plant-derived
flavor compounds. The Plant Journal, 54, 712-732.
Stepanova, A. N., Robertson-Hoyt, J., Yun, J., Benavente, L. M., Xie, D. Y., Doležal, K., ...
Alonso, J. M. (2008). TAA1-mediated auxin biosynthesis is essential for hormone crosstalk
and plant development. Cell, 133, 177-191.
Tieman, D., Taylor, M., Schauer, N., Fernie, A.R., Hanson, A.D., & Klee, H.J. (2006). Tomato
aromatic amino acid decarboxylases participate in synthesis of the flavor volatiles
2
-phenylethanol and 2-phenylacetaldehyde. Proceedings of the National Academy of
Sciences, 103, 8287-8292.
Vogt, T. (2010). Phenylpropanoid biosynthesis. Molecular Plant, 3, 2-20.
Wang, M., & Maeda, H. A. (2017). Aromatic amino acid aminotransferases in plants.
Phytochemistry Reviews, 1-29.
Wan, X.C., & Xia, T. (2015). Secondary metabolism of tea plant (in Chinese) (1st ed.). Beijing:
Science Press.
Yang, Z. Y., Baldermann, S., & Watanabe, N. (2013). Recent studies of the volatile compounds in
tea. Food Research International, 53, 585-599.
Yoo, H., Widhalm, J. R., Qian, Y., Maeda, H., Cooper, B. R., Jannasch, A. S., ... Dudareva, N.
(
2013). An alternative pathway contributes to phenylalanine biosynthesis in plants via a
cytosolic tyrosine: phenylpyruvate aminotransferase. Nature Communications, 4, 2833.
21

Zeng, L. T., Zhou, Y., Gui, J. D., Fu, X. M., Mei, X., Zhen, Y. P., ... Yang, Z. Y. (2016).
Formation of volatile tea constituent indole during the oolong tea manufacturing process.
Journal of Agricultural and Food Chemistry, 64, 5011-5019.
Zhou, Y., Dong, F., Kunimasa, A., Zhang, Y. Q., Cheng, S. H., Lu, J. M., ... Yang, Z. Y. (2014).
Occurrence of glycosidically conjugated 1-phenylethanol and its hydrolase
β-primeverosidase in tea (Camellia sinensis) flowers. Journal of Agricultural and Food
Chemistry, 62, 8042-805.
Zhou, Y., Zeng, L. T., Liu, X. Y. , Gui, J. D., Mei, X., Fu, X. M., ... Yang, Z. Y. (2017). Formation
of (E)-nerolidol in tea (Camellia sinensis) leaves exposed to multiple stresses during tea
manufacturing. Food Chemistry, 231, 78-86.
22

Figure Legends
Figure 1. Identifications of precursor of some volatile phenylpropanoids and benzenoids in
tea leaves.
(
A) Potential pathways of volatile phenylpropanoids and benzenoids synthesis from
L-phenylalanine in plants. (B and C) Changes in amounts of benzaldehyde, benzyl alcohol, and
methyl benzoate using phenylpyruvic acid (PPA) or trans-cinnamic acid (CA) as substrate,
respectively. CK1, sum of contents volatiles produced from reaction without crude enzyme and
reaction without phenylpyruvic acid. CK2, sum of contents volatiles produced from reaction
a
without crude enzyme and reaction without trans-cinnamic acid. , contents of the products
b
extracted from 400 μL enzyme solution after 60 min enzyme reaction; , contents of the products
absorbed on SPME for 60 min collection after 60 min enzyme reaction. (D) Amounts of labeled
2
benzaldehyde and benzyl alcohol in tea leaves after supplementation with [ H
8
]L-phenylalanine
2
2
(
[ H
8
]Phe) and co-supplement of [ H ]Phe and 2-aminoindan-2-phosphonic acid (AIP). Data are
8
expressed as means ± S.D. (n = 3). *, p ≤ 0.05) and **, p ≤ 0.01.
Figure 2. Transformation from L-phenylalanine into phenylpyruvic acid in vitro and in vivo
in tea leaves.
(
A) Production of labeled phenylpyruvic acid (PPA) contents after crude enzyme reaction. CK,
2
crude enzyme reaction without [ H
8
]L-phenylalanine; Treatment, crude enzyme reaction using
2
[
H
8
]L-phenylalanine as substrate. N.D., not detected. Data are expressed as means ± S.D. (n = 3).
*
**, p ≤ 0.001. (B) Characteristic ions of products from phenylpyruvic acid/labeled
phenylpyruvic acid derivatized by MSTFA. (C) GC-MS chromatography of phenylpyruvic
23

acid/labeled phenylpyruvic acid.
Figure 3. Phylogenetic tree of CsAAATs and AAATs from other plants.
Multiple alignments were performed with MEGA 7.0. At, Arabidopsis thaliana; Cm, Cucumis
melo (melon); Cs, Camellia sinensis; Os, Oryza sativa; Pf, Perilla frutescens; Ry, Rosa 'Yves
Piaget'; Sm, Salvia miltiorrhiza; Ss, Solenostemon scutellarioides; Sp, Solanum pennellii
(
tomato); Zm, Zea mays. TyrATs, tyrosine aminotransferases; PheATs, phenylalanine
aminotransferases; AlaATs, alanine aminotransferases; TrpATs, tryptophan aminotransferases.
Accession numbers are as follows: AtTAT1, NP_200208.1; AtTAT2, NP_198465.3; AtTrpAT,
AEE35079.1; CmArAT1, NP_001284465.1; CmBCAT, ADC45390.1; CsAAAT1, MH544095;
CsAAAT2, MH544096; OsAlaAT, BAA77261.1; PfTAT, ADO17550.1; RyAAAT3, BAF64843.1;
SmTAT, ABC60050.1; SpTAT1, ADZ24702.1; SsTAT, CAD30341.1; ZmAlaAT, AAC62456.1;
ZmTrpAT, AMD16119.1.
Figure 4. Identification of CsAAAT1 and CsAAAT2 recombinant proteins expressed in
Escherichia coli.
(
A) SDS-PAGE analysis of E. coli-expressed CsAAAT1 and CsAAAT2. Arrows indicate target
proteins. (B) Western blot analyses of CsAAAT1 and CsAAAT2 expressed in E. coli.
Figure 5. Functional identification and subcellular localization of CsAAAT1 and CsAAAT2.
(
A) Enzyme activities of CsAAAT1 and CsAAAT2 expressed in Escherichia coli. Phe as sub.,
enzyme reaction using L-phenylalanine as substrate; Tyr as sub., enzyme reaction using
L-tyrosine as substrate; N.D., not detected. Data are expressed as means ± S.D. (n = 3). ***, p ≤
24

0
.001. (B)Analysis of phenylpyruvic acid content in per gram of fresh N. benthamiana leaves.
Vector, empty vector GV3101 overexpressed in N. benthamiana. CsAAAT1 and CsAAAT2
represent CsAAAT1 and CsAAAT2 overexpressed in N. benthamiana, respectively. Data are
expressed as means ± S.D. (n = 5). Columns with different letters above are significantly
different (Duncan’s test, p ≤ 0.05; one-way ANOVA). (C) Subcellular localization analyses of
CsAAAT1-GFP and CsAAAT2-GFP. Upper photos: CsAAAT1-GFP in cytoplasm of tobacco leaf
cell. Lower photos: CsAAAT2-GFP in cytoplasm of tobacco leaf cell. Green and red panels show
GFP fluorescence and chloroplast auto-fluorescence, respectively. Merged panel shows overlay
of green and red fluorescence images.
Figure 6. Proposed biosynthesis pathway of benzaldehyde, benzyl alcohol, and methyl
benzoate from L-phenylalanine via phenylpyruvic acid in tea leaves.
L-Phe, L-phenylalanine; PPA, phenylpyruvic acid; CsAAAT1, Camellia sinensis aromatic amino
acid transaminase 1; BAld, benzaldehyde; BAlc, benzyl alcohol; MeBA, methyl benzoate.
25

Figure 1
26

Figure 2
27

Figure 3
28

Figure 4
29