GABA and Its Shunt’s Contributions to Flavonoid Biosynthesis and Metabolism in the Tea Plat (Camellia Sinensis)

Background: γ-Aminobutyric acid (GABA), a signal molecule, is regarded as the intersection of carbon and nitrogen metabolism, but its contributions to avonoid metabolism in tea plants during the whole growth cycle remain unclear, and the correlation between the GABA shunt and avonoid metabolism in tea plants is worth exploring. Secondary metabolites and their correlations with the taste qualities of tea plants (Camellia sinensis) during different seasons have been investigated. Results: Related secondary metabolites and transcript proles of genes encoding enzymes in the GABA shunt, avonoid pathway and polyamine biosynthesis were measured throughout the whole tea plant growth season and after exogenous GABA applications. In addition, levels of differentially expressed proteins were measured after treatment with or without exogenous GABA. The tea leaves showed the highest metabolite concentrations in spring. CsGAD, CsGABAT, CsSPMS, CsODC, CsF3H and CsCHS were found to be important genes in the GABA and anthocyanin network. Conclusion: GABA and anthocyanin concentrations showed a positive correlation, to some extent, and CsF3H and CsCHS played important roles in the GABA and anthocyanin network. Further studies should focus on exploring GABA and avonoid metabolism through the transgenic engineering of tea plants.


Background
Tea is a popular beverage worldwide. The tea plant (Camellia sinensis) is an important commercial crop in China because of its importance in Chinese traditional culture and the special sensory qualities of tea produced from speci c regions. However, tea quality is easily affected by the environment, especially the climate in which the tea plants are grown, This is part of tea's appeal-special taste characteristics that can only be discerned by connoisseurs. [1] In different seasons, the metabolites and taste qualities of tea have been studied preliminarily. [2,3] Flavonoid compounds, like avonols and avan-3-ols, contribute greatly to tea quality and tea plant growth. Anthocyanins are tea avonoid compounds that play important roles in plant coloring and in preventing damage caused by abiotic and biotic stresses, like cold, drought, UV irradiation and fungal infection. [4] Anthocyanins in plants have been studied in depth because they are not only bioactive components but their consumption is also bene cial for human health, including reducing the risk of developing cancer and diabetes. Additionally, amino acid levels play signi cant roles in changes in the qualities of tea harvested from plants during different seasons. γ-Aminobutyric acid (GABA), a non-protein amino acid, has been a hot research subject for years owing to its roles in a variety of biological functions, such as improving the medicinal value of plants and their resistance to environmental stress. [5][6][7] In addition, exogenous GABA in uences the contents of free amino acid components in tea plants under both suitable environmental conditions and cold stress. [6] In plant cells, the GABA shunt, a part of the tricarboxylic acid cycle, is responsible for GABA synthesis through an irreversible reaction involving α-decarboxylation and glutamate, as well as polyamine (PA) degradation by glutamate decarboxylase (GAD), GABA transaminase (GABA-T) and succinic semialdehyde (SSA). [8] Ornithine/arginine (Arg) and PA biosynthetic pathways are associated with seed growth, while GABA has been predicted to play important roles in seed development, but its detailed modes of action remain poorly understood. [9] In this study, tea leaves were collected during regular stages over a whole year to investigate the contribution of the GABA shunt to seasonal differences in tea growth and quality. Although numerous studies regarding anthocyanins and GABA biosynthesis have been reported [4] , the correlations between GABA and anthocyanins have not been investigated.
In this work, isobaric tags for relative and absolute quanti cation (iTRAQ)-based proteomic analyses were performed to select the differentially expressed proteins (DEPs) involved in avonoid metabolism, the GABA shunt and amino acid transport. The tea plants were treated with exogenous GABA and compared to normal untreated controls. DEPs involved in the GABA shunt and avonoid metabolism were identi ed.
Furthermore, we discuss the morphological, chemical and transcript pro les related to the GABA shunt and avonoid metabolism at different developmental stages over the whole year. This study will increase our understanding of the effects of the GABA shunt on the synthesis of anthocyanins and possible correlations between the GABA shunt and avonoid metabolism in tea plants during different seasons.

Plant materials and treatments
Tea plants (C. sinensis cv. Longjingchangye) formal identi ed by Feng Li, were collected from the tea garden of the Zhongshan Palace in Nanjing, China. A voucher specimen of this material has been deposited in Department of Tea Science, Nanjing Agricultural University. In this study, single buds having two leaves were obtained seasonally, including 7 th Jan. Extraction and measurement of PAs and free amino acid components Tea leaves were subjected to extraction and analyzed using previously reported methods. [6,8] Determination of catechin and anthocyanin concentrations Methods of measuring the catechin and anthocyanin contents were described previously by Wang. [10] A Waters e2695 system (Waters, USA) was employed for catechin determination: ow rate, 1 ml/min; temperature, 35°C; column, Phenomenex 00G-4337-E0, 250 × 4.6 mm; injection volume, 10 μl; mobile phases, A = 1% formic acid (v/v) and B = 100% acetonitrile (v/v), 0-42 min; detection, λ = 280.

Determination of chlorophyll concentration
The determination of the chlorophyll concentration was performed in accordance with a previous report. [11] Gene expression level analysis A plant RNA extraction kit (Aidlab, China) was used to isolate total RNA from 100 mg fresh leaves. A cDNA Synthesis SuperMix (TransGen, China) was used for cDNA synthesis. Quantitative real-time PCR (qRT-PCR) was performed using a Bio-Rad iQ5 uorescence quantitative PCR platform. The primer pairs are listed in Table S1. CsGAPDH was used as an internal control gene in this study. The relative gene expression levels were calculated using the 2 −ΔΔCt method.
Protein extraction, digestion, iTRAQ labeling, mass spectrometry analysis and data analysis Protein extraction, digestion, iTRAQ labeling, mass spectrometry analysis and data analysis were followed as previously reported. [6] Brie y, 1 ml phenol extraction buffer was mixed with 0.1 g frozen tea leaves and then shaken for 40 min at 4°C with 1 ml saturated phenol (Tris-HCl, pH 8.0). Then, 0.1 M ammonium acetate: methanol and lysate solutions were used to extract proteins. An iTRAQ Reagent 4plex Multiplex Kit (AB Sciex) was used to label samples after being reduced, alkylated and trypsindigested.
A Triple TOF 5600 mass spectrometer (SCIEX, USA), a capillary C18 trap column (3 cm × 100 μm) and a C18 column (15 cm × 75 μm) were used to perform the analyses, load samples and separate samples, respectively. Additionally, a 2.4-kV ion spray voltage, 35-psi curtain gas, 5-psi nebulizer gas and an interface heater were used to acquire data.

DEP annotations and a protein-chemical interaction analysis
All the DEPs in the GABA shunt, avonoid metabolism pathway and amino acid transport were identified using iTRAQ-based quantitative proteomics [6] . In addition, the predicted interaction networks of proteins and chemicals in the GABA shunt, amino acid transport and avonoid metabolic pathway were constructed using STITCH (http://stitch.embl.de/).

Statistical analysis
The average value of replicates was calculated. SPSS17.0 (SPSS Inc., Chicago, IL, USA) was used to determine expression level differences. A heatmap was constructed using Tbtool vo.66583 (China).

Morphological changes in tea leaves during seasons
The coloring of the tea leaves occurred gradually during different seasons under different temperatures and light conditions (Fig. 1A-K). In summer, the bud and rst leaf were purplish red; they were a bit yellow and red in autumn, more yellow in winter and green in spring. In spring (D-K), there were less sprouting buds in progressive stages, such as H, I and J.
Changes in the chemical contents and gene expression levels in the GABA shunt A heatmap was constructed to identify correlations between transcripts and metabolite pro les in the GABA shunt at different times over the whole year. The spermine, spermidine, GABA, Arg and proline (Pro) contents showed obvious increasing trends when spring began. The PA content was highest in stage "G", while that of GABA was highest in stage "J". Additionally, the stages having the lowest GABA contents occurred in summer and winter, while those of spermine and spermidine occurred in summer and winter, respectively. Additionally, the GABA and glutamate contents showed opposite trends over the whole study period. Except for stages "J" and "K", the GABA and spermidine contents showed similar changes. Moreover, the spermine and spermidine contents changes in a consistent manner (Fig. 2B).
Changes in amino acids, chlorophyll and metabolite contents, as well as gene expression pro les in the avonoid pathway In winter, the anthocyanin content was highest, while it was lowest in stage "F". The chlorophyll a and b concentrations peaked at stage "D", while they were lowest at stages "C" and "J" (Fig. 3A). The epigallocatechin gallate and epicatechin gallate contents experienced steady downward trends from stages "A" to "K". The epicatechin content was highest in stage "D" and lowest in stage "B". The catechin content was lowest at stage "D". (Fig. 3B). The expression level of CsANR was highest in stage "D", while that of CsUFGF was highest in stage "C". The expression levels of CsCHS and CsFLS were highest in stage "K", while that of CsFLS was lowest in stages "C" and "D". Additionally, the expression level of CsCHS was at stage "D", and that of CsCHI was at stage "C" (Fig. 3C).
The Val, Ile, Leu and Try contents showed similar trends in the study, experiencing uctuations throughout the whole period. The contents of other amino acid components showed an increasing trend from the beginning of spring (Fig. 3D).
The GABA concentration was signi cantly negatively correlated with the gallocatechin gallate (GCG), epigallocatechin (EGC) and chlorophyll contents, while the glutamate content was signi cantly negatively correlated with the GCG and chlorophyll contents and extremely signi cantly negatively correlated with the EGC content. However, over the season, the contents of chemicals related to the GABA shunt were negatively correlated with the anthocyanin content (Table S2). In avonoid metabolism, the expression levels of CsPAL, Cs4CL, CsF3H, CsCHS, CsFLS, CsLAR, CsANS and CsANR were strongly correlated with the catechin content, while the anthocyanin content was only signi cantly positively correlated with the CsANR expression level. Additionally, the GABA content was also positively correlated with the CsF3H expression level (Table S3).

Interaction networks of proteins and chemicals
STITCH was used to analyze the interaction networks of chemicals and proteins in the GABA shunt and avonoid metabolic pathway. In the avonoid metabolic pathway, p-coumaric acid, catechin, EGC and epicatechin had no interactions with chemicals related to the GABA shunt. In addition, ketoglutarate and phenylalanine were the most important bridges to chemicals related to the GABA shunt and avonoid metabolism. Moreover, the proteins P5CS, CHS1, CHI and F3H played important roles in the correlations of these two pathways.
DEPs and chemicals related to GABA and anthocyanin metabolism in tea leaves after exogenous GABA applications compared with the control The expression levels of proteins CHS, FLSI, F3H, ANR and ALDH12A1 were down-regulated when an exogenous GABA treatment was applied compared with the control, while those of proteins SDH-2, LPD1, GLN1-1, MS1 and AGT1 were up-regulated. (Fig. 5A) The PA contents decreased when compared with the control treatment. The putrescine and spermine concentrations underwent signi cant downward trends compared with the control on day 4. Additionally, the GABA and anthocyanin concentrations in the GABAtreated samples increased dramatically compared with those of the control during the whole study period (Fig. 5B). The contents of all the amino acid components, except cysteine, methionine and tyrosine, increased after both treatments. In the GABA shunt, the CsGAD and CsSPMS2 expression levels were signi cantly positively correlated with the anthocyanin content, while the CsGAD1, CsGABAT1, CsSPMS2 and CsODC levels were heavily correlated with chemicals related to the GABA shunt (Table S4).

Discussion
Effects of climate on tea leaf morphology and chemical compound contents During the tea growth cycle, tea quality and morphology are affected by water levels, temperature and sunlight radiation intensity and duration. The anthocyanin contents were higher in summer than in winter. [12] However, in our results, the anthocyanin contents and the UFGF expression were highest in winter (stage "C"), while the purple color of tea leaves was the deepest in summer. The sharp enhancement in the anthocyanin concentration in winter may result from the increasing level of water present in the environment (snow was present in stage "C"). In stage "C", yellow-colored tea leaves had higher anthocyanin concentrations and lower chlorophyll a and b concentrations. During stage "A", purple-redcolored tea leaves showed the second highest anthocyanin and both chlorophyll a and b concentrations. Unfortunately, we did not measure the correlations among tea leaf colors and pigment levels. In the avonoid synthetic pathway, UFGT and CHI were key anthocyanin biosynthetic genes, because their expression levels were highly positively correlated with the anthocyanin content.
In spring, because of warming, bud are hastening to mature and open. [13] In plucking seasons, chemical compounds in tea leaves are in uenced by temperature, sun exposure time and water levels. The amino acid content is highest in spring, and amino acid and catechin levels are affected by temperature, water levels and sun exposure time. [3] In our results, uctuations in the catechins contents were moderate in spring, and epigallocatechin gallate and epicatechin gallate contents were highest in summer. Green tea leaves had higher amino acid contents compared with leaves harvested in later seasons. [14] In our results, the amino acid contents, especially those of Ala, Pro and Arg, in tea leaves were higher in summer and spring than in other seasons.
Chemicalaccumulations and transcript pro les of genes encoding enzymes involved in the GABA shunt and PA biosynthesis during the whole tea growth cycle The metabolites present and taste qualities of tea leaves harvested in different seasons have been investigated. [15,16] GABA may be regarded as a signal molecule, and it accumulates to adjust to stress reactions, especially those related to temperature. [6,17,18] However, GABA concentrations in tea leaves were higher in spring than those in winter and summer. Additionally, the expression CsGAD1 level increased signi cantly in summer and winter, while theCsGABAT1 and CsGABAT2 levels decreased. The CsCUAO expression level was fairly low in winter. Perhaps, CsCUAO could be used to explore the roles of the GABA shunt in the plant growth cycle. However, the contents of all the chemical compounds in the GABA shunt and PA biosynthesis were lower in summer and winter compared with in spring. In previous studies, the concentrations of most chemical compounds in plants were found to be highest in spring. [12,14] At the same time, PAs, Met and Pro concentrations increase to improve plant resistance to temperature stress. [19][20][21][22][23] In winter and summer, there are many other factors that in uence plant growth, like the sunlight and water levels. It is worthwhile exploring how the genes CsGAD, CsGABAT, CsSPMS and CsODC function during the tea growth cycle, because their expression levels were strongly correlated with the chemical compounds in the GABA shunt and PA biosynthetic pathway.
The possible contributions of avonoid metabolism to tea plants during the tea growth cycle Flavonoid accumulations in grapes change from summer to winter in response to many factors, including temperature, sunshine and water levels. [12] Unfortunately, the avonoid regulatory mechanisms that act in response to changes in climatic conditions remain elusive. Because of global warming, climatic in uences on tea quality must be taken into account. Flavonoids are produced by plant secondary metabolism. [24] Correlations between GABA and environmental conditions have been investigated. [6,17,25] We explored this point by investigating correlations between GABA and avonoid metabolism. The GABA shunt and avonoid metabolism may affect each other because they both contribute to the regulation of physiological processes of plants in response to biotic and abiotic stresses. [16] Here, the contents of compounds in the GABA shunt were correlated with catechin levels (Table S2). In particular, Glu and GABA concentrations were strongly negatively correlated with GCG and EGC levels. CsPAL1, CsC4H, CsCHS, C'F''5'H, CsDFR, CsDFR, CsLAR and CsANR play important roles in catechin synthesis. [17] Here, we were unable to de ne clear correlations between the GABA contents and transcript pro les of genes involved in avonoid metabolism. In the future, the overexpression or gene-silencing of CsODC, CsSPMS2 or CsSPDS1 will be performed (Fig. S1).

Effects of exogenous GABA on avonoid metabolism
The correlations of GABA and anthocyanins with avonoid metabolism have been studied, and anthocyanin biosynthesis in plants is in uenced by many factors, like hormones, temperature and light. [26][27][28] However, there is no research regarding correlations between GABA and anthocyanins, which are important compounds in avonoid metabolism. We found that the addition of exogenous GABA increased the anthocyanin concentrations in tea leaves (Fig. 5B). In the anthocyanin biosynthetic pathway, there are many functioning transcription factors . [29,30] The CHS, FLS1, F3H and ANR expression levels were down-regulated after the application of exogenous GABA compared with under normal conditions, and the CsF3H and CsCHS expression levels also decreased (Fig. 5). Tea cultivars with purple leaves are used to produce unique teas with speci c avors. The use of GABA treatments might represent a practical method to regulate avonoid metabolism and produce tea with optimal desired qualities.

Conclusion
In this study, the secondary metabolite contents and transcripts pro les of genes encoding enzymes in the GABA shunt, PA biosynthesis and avonoid pathway were measured during the whole tea plant growth period. The results suggested that CsGAD, CsGABAT, CsSPMS and CsODC may be effective and helpful in exploring the complicated metabolism of the GABA shunt and PA biosynthesis in tea plants during the whole year. To determine more about the correlations between GABA and avonoid metabolism, we detected the expression levels of proteins, some chemical compounds and transcripts pro les related to tea plants treated with and without exogenous GABA. GABA and anthocyanin concentrations showed a positive correlation, to some extent, and CsF3H and CsCHS played important roles in the GABA and anthocyanin network. Further studies should focus on exploring GABA and avonoid metabolism through the transgenic engineering of tea plants. Availability of data and materials The data sets are included within the article and its Additional les.
Competing interests