Proteomic analysis of tea plants (Camellia sinensis) with purple young shoots during leaf development

Tea products made from purple leaves are highly preferred by consumers due to the health benefits. This study developed a proteome reference map related to color changes during leaf growth in tea (Camellia sinensis) plant with purple young shoots using two-dimensional electrophoresis (2-DE). Forty-six differentially expressed proteins were detected in the gel and successfully identified by using MALDI-TOF/TOF-MS. The pronounced changes in the proteomic profile between tender purple leaves (TPL) and mature green leaves (MGL) included: 1) the lower activity of proteins associated with CO2 assimilation, energy metabolism and photo flux efficiency and higher content of anthocyanins in TPL than those in MGL may protect tender leaves against photo-damage; 2) the higher abundance of chalcone synthase (CHS), chalcone isomerase (CHI) and flavonol synthase (FLS) likely contributes to the synthesis of anthocyanins, catechins and flavonols in TPL tissues; 3) higher abundance of stress response proteins, such as glutathione S-transferases (GST) and phospholipid hydroperoxide glutathione peroxidase (PHGPx), could enhance the tolerance of TPL tissues to adverse condition in; and 4) the increased abundance of proteins related to protein synthesis, nucleic acids and cell wall proteins should be beneficial for the proliferation and expansion of leaf cell in TPL tissues. qPCR analysis showed that the expression of differentially abundant proteins was regulated at the transcriptional level. Therefore, the results indicated that higher abundance of CHI and CHS may account for the production of the purple-shoot phenotype in Wuyiqizhong 18 and thereby, enhancing the anthocyanin biosynthesis. The higher abundance of glutamine synthetase (GS) proteins related to the theanine biosynthesis may improve the flavor of tea products from TPL materials. Thus, this work should help to understand the molecular mechanisms underlying the changes in leaf color alteration.


Introduction
Tea (Camellia sinensis (L.) O. Kuntze) is an important commercial crop that is grown mostly in the tropical and subtropical countries of Asia, Africa and to some extent Latin America [1].In recent years, tremendous attention has been focused on tea plant due to its pleasant flavor and bioactive substances.Tea leaves are a valuable source of secondary metabolic products, including flavonoids, alkaloids, polysaccharides and theanine [2].Among these secondary metabolites, flavonoids (which comprise polyphenols, flavones, flavanonols and anthocyanins) are considered to contribute numerous pharmacological beneficial effects on human health [3][4].The health benefits of tea are thought to account for tea's protective role against cardiovascular disease [3], atherosclerosis [4], oxidant activity [5], and cancer [6].
The color of tea leaves has been diversified through long-time natural hybridization and artificial selection.Study on purple leaves of tea plant will positively promote the diversification of tea products and the fully utilization of different tea germplasm resources.Furthermore, tea products made from purple tea leaves are highly preferred by consumers.Compared with conventional tea, the anthocyanin-rich purple tea has multiple benefic functions including strong antioxidant activity [7], inhibition of colorectal carcinoma cell proliferation [8], and reinforcing brain antioxidant capacity [9].In addition, anthocyanins can be used as commercial food colorants and have been reported to exhibit antioxidants, chemopreventive, anti-bacterial, antiangiogenic, anti-inflammatory and anti-atherosclerotic properties [10].Therefore, to enhance the health potency of tea, purple-colored leaves have become one of the main qualitative attributes targeted in tea breeding programs.New tea cultivars with purple leaves have recently been developed and released in Kenya, China, Japan and India [11,[12][13][14].
The contents of total phenolic compounds and total anthocyanins, including catechins were higher in purple tea leaves than in green tea ones [15,16].Previous study has found that malvidin is the most abundant anthocyanidin in the tea products derived from the purple colored tea clones [11].Anthocyanins often accumulate in young leaves of plants (including tea plants), and make leaves present a purple color, then the purple color in leaves gradually turn into green as the leaves mature [17].Additionally, anthocyanin in young leaves correlates with tolerance to photoinhibition by the attenuation of visible radiation.Maayan [18] have reported that ribulose-1,5-bisphosphate carboxylase/oxygenase (rubisco), as well as chlorophyll, is not present in early development stages and only slowly accumulates throughout the developmental stages of the leaves.The lower abundance of rubisco is closely related to the rate of photosynthesis in young purple leaves [19].In previous study, we used cDNA-AFLP to investigate the gene expression profiles associated with changes in leaf color during the developmental stages of tea leaves, and 148 differentially expressed genes, which were involved in cell respiration, nucleic acid and protein metabolism, photosynthesis, cell expansion, biosynthesis of flavonoids and anthocyanins, were examined [19].However, the identification of differentially expressed genes at mRNA level is insufficient for the prediction of protein expression levels, and the available proteomics information on the changes in metabolites during shoot development in purple-leaf-tea cultivars is limited.Therefore, in this study, we used the newly developed tea cultivar Wuyiqizhong18 as source material, and we conducted 2-DE to separate differentially abundant proteins at the proteomic level and characterize the molecular mechanisms underlying changes in leaf color during the growth of tea plants with purple young shoots.

Plant materials
Tea plants (C.sinensis (L.) O. Kuntze cv.Wuyiqizhong18) were grown in the Germplasm Repository of Tea Research Institute in Wuyi Mountain, Fujian province, China.Only the healthy, tender shoots (one bud and two leaves) were sampled on April 13, 2015, and then the purple and green leaves were packaged separately in aluminum foil.The plant characteristics in 'Wuyiqizhong18' tea cultivar were as follows: the bud were green, the color of the second leaf changes gradually to purple, and the leaves finally turned green again (Fig 1).The samples were immediately frozen in liquid nitrogen and stored at −80˚C until use.

Measurement of polyphenolic compounds
Anthocyanins were extracted with 1% HCl-methanol solution and measured using a spectrophotometer at 530 nm, according to method described by Hatlestad et al. [20].The contents of total polyphenols were determined according to the International Organization for Standardization (ISO) 14502-1:2005 using Folin-Ciocalteu method [16].The identification and quantification of catechins and caffeine were measured according to the method described by Wei et al. [21].

Protein extraction
Total protein was extracted from frozen leaf samples using a phenol extraction procedure according to Saravanan et al. [22] with some modifications.Briefly, about 1 g leaf sample was ground into fine powder in liquid N 2 with a pestle and mortar.Approximately, 4 mL of icecooled extraction buffer [100 mM Tris-HCl (pH 7.8), 50 mM L-ascorbic acid, 1% (v/v) Triton X-100, 1% (v/v) β-mercaptoethanol, 100 mM KCl, and 1 mM phenylmethylsulfonyl fluoride] was added to the powder.The mixture was allowed to thaw slowly at room temperature and then gently pulverized for another two minutes.The homogenate was transferred to 10 ml centrifuge tube and centrifuged at 13,000 g for 15 min at 4˚C after equal volume of Tris-saturated phenol (pH 8.0) was added.The phenolic phase was transferred to a 50 mL corning centrifuge tube with five volumes of methanol (with 100 mM ammonium acetate).The protein pellets were collected and re-suspended in 25 mL of ice-cooled pure methanol for 2 h at -20˚C after centrifugation at 13,000 g for 10 min at 4˚C.Protein pellets were collected by centrifugation at 13,000 g for 15 min at 4˚C and washed twice with 25 mL of ice-cooled acetone, and then lyophilized at -20˚C until use.About 10 mg protein pellet was dissolved in lysis buffer [8 M urea, 1.5 M Thiourea, 4% (w/v) CHAPS, 2% (v/v) ampholyte (3-10), 15 mM DTT].The total protein concentration was measured by using Bradford protein assay [23].
Image analysis was carried out with PDQuest software version 8.0.1 (Bio-Rad, Hercules, CA, USA).Spot detection, matching and normalization were performed automatically, followed by manual verification.Only those candidate spots in all triplicate gels which had both a fold change of more than 1.5 and a p value of less than 0.05 by statistical test were considered differentially abundant proteins for mass spectrometry analysis.

Protein identification by MALDI-TOF/TOF-MS and bioinformatics analysis of proteins
Protein spots were excised from the stained gels and digested with trypsin as described by You et al. [24].Mass spectrometric analysis was conducted with a MALDI-TOF/TOF-MS 5800 (AB SCIEX, Shanghai, China).All acquired full-scan MS of samples were processed using TOF/ TOF Explorer™ Software (AB SCIEX) in-default mode.The data were analyzed through the GPS Explorer software (Applied Biosystem, Foster City, CA, USA) using MASCOT software (Matrix Science Inc., MA, USA) with the database viridiplantae from NCBI website (1850050 sequences; 642453415 residues).The search criteria specified trypsin, allowing 1 missed cleavage, with a peptide tolerance of 100 ppm of and an MS/MS tolerance of 0.6 Da.Proteins exhibiting a score greater than 75 and statistical significance (p < 0.05) were considered to be successfully identified proteins and were accepted.
Functional annotation of proteins was conducted using Gene Ontology (GO) annotation [25].The differentially accumulated proteins were assigned to the Cluster of Orthologous Groups of proteins (COG) database and the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [26].

qPCR analysis
Total RNA extraction and qPCR were performed as previously described [19].Specific primers were designed to investigate the gene expression at the transcript level of these representative differentially accumulated proteins.The sequences of the forward (F) and reverse (R) primers were listed in S1 Table .For the normalization of gene expression, Camellia sinensis 18S ribosomal RNA (AB120309.1)was used as a reference gene.Each treatment contained 3 biological replicates and each biological replicate had 3 technical replicates.The specificity of each primer pair was verified by melting curve analysis.The relative gene expression was calculated using the 2 -ΔΔCt algorithm below [27].

Statistical analysis
Significance tests were conducted using Student's t-test at p < 0.05 level.The results are represented by the mean ± standard errors (SE).
To elucidate the overall proteomic responses of tea leaves, we performed a heatmap approach by using mean centered and standard deviation scaled with The Unscrambler (Version 10.0.1, Camo Software Inc., Woodbridge, NJ, USA) for the normalization of the values to give all variables an equal chance.After normalization of all the data sets, normalized values were used in the color scales system from Microsoft Office Excel 2010 (Microsoft Corporation, Redmond, WA, USA) to generate the heatmap.
The normalized data set was used for the principal component analysis (PCA) by performing normalized volume of the differentially abundant protein spots using the Unscrambler (version 10.0.1, Camo Software Inc., Woodbridge, NJ, USA).The PCA loading plots were used to determine the separation of the differentially expressed proteins based on the presence and absence between TPL and MGL samples.The PCA loading plots were performed in triplicate (n = 3).

Identification of differentially expressed proteins
To explore the correlation between the proteomic and metabolite profiles of TPL and MGL tissues, 2-DE coupled with MS/MS technique was carried out with three biological replicates.More than 900 clear and reproducible spots were detected on each gel (Fig 3).After conducting analyses with PDQuest 8.0.1 software and MALDI-TOF/TOF-MS, we successfully identified 46 proteins (S1 Fig, Table 1).Of these identified proteins, 26 proteins had high abundance in TPL tissues, while 20 proteins had low abundance compared with those in MGL tissues.The results from the database are presented in Table 1.

PCA and heatmap responses
PCA was performed with all 46 of the identified proteins.The PC1 and PC2 values, which accounted for 89% and 5% of the total X and Y variance, revealed that the differential protein responses between TPL and MGL were clearly separated by the first principal component

Transcriptional expression analysis of differentially abundant proteins
To verify the proteomic results and evaluate the association between mRNA and protein levels, qPCR assays were conducted (

Discussion
Anthocyanins often accumulate in young leaves, fruit or flowers, and then degrade as leaves develop [28].This tendency has been reported on many plant species, including Chrysobalanus icaco [29], Capsicum spp [30] and Brunfelsia calycina [31].A previous study by our group also showed that anthocyanin contents decreased with leaf growth and development, while the contents of chlorophyll a and chlorophyll b were higher in MGL samples.Thus, we used the cDNA-AFLP technique to reveal the regulatory mechanism of the alterations of leaf color during leaf development in Wuyiqizhong18 [19].Nevertheless, due to post-transcriptional regulation, gene transcription data are insufficient to predict the expression levels of proteins.To further understanding the molecular mechanism underling the development of tea leaves, we used 2-DE technique to investigate the proteomic profile of different leaves and successfully identified 46 differentially abundant proteins in MGL and TPL samples in this study.

Proteins involved in carbohydrate and energy metabolism
We identified 6 increased and 17 decreased proteins that were involved in carbohydrate and energy metabolisms from TPL tissues (Table 1).In plants, the tricarboxylic acid cycle (TCA cycle) and the pentose phosphate pathway (PPP) provide not only energy and cofactors but also substrates for the synthesis of metabolites, or signals for feedback [32].Interestingly, in the present study, we found that some key enzymes involved in carbon assimilation had high abundance in TPL, such as phosphoglycerate kinase (PGK, spot No. 6410), fructose-bisphosphate aldolase (spot No. 8308), sedoheptulose-1,7-bisphosphatase (spot No. 2403) and phosphoribulokinase (PRK, spot No. 2409) (Fig 8).Moreover, TPL sample has a lower abundance of rubisco (spot No. 4401, 2402, 8204, 9502, 8506 and 7505), a bi-functional enzyme that catalyzes the initial step of the Calvin cycle and the initial oxygenation reaction of photorespiration.This observation was well corresponded with our previous study, reporting that the transcriptional levels of the genes encoding some photosynthetic enzymes are decreased in TPL tissues [19].The lower abundance of rubiscos would be responsible for the lower activity of CO 2 assimilation in TPL tissues than in MGL ones.It is also possible that the lower abundance of rubiscos in TPL tissues due to the immature chloroplasts of TPL tissues [18].
Four enzymes (malate dehydrogenase, spot No. 8304 and 806; fructokinase, spot No. 4304; pyruvate decarboxylase, spot No. 8702) involved in organic acid metabolism and glycolysis have significantly higher abundance in TPL tissues than in MGL ones.It means that energy and secondary metabolism would be enhanced in TPL tissues.Previous research has suggested that anthocyanins in young leaves are likely associated with tolerance against photoinhibition by attenuating visible radiation [33,34].Since TPL tissues are rich in anthocyanins, light attenuation by anthocyanins may display shade acclimation of the photosynthetic machinery in purple leaves, which is compatible with a photoprotective function of anthocyanins [35].The accumulation of anthocyanins in TPL tissues may serve as a filter of light and protect chloroplasts against photoinhibition [36].Thus, anthocyanins likely play a key role in the photoprotection of young leaves, which is a hypothesis that has been addressed in a previous study [37].

Proteins involved in secondary metabolism
Although the proportion of differentially abundance proteins related to secondary metabolism in our results was relatively low (9%) compared with the other categories, proteins related to anthocyanin and caffeine biosynthesis were found to be differentially abundances between TPL and MGL tissues (Table 1).Tea leaves are a valuable source of secondary metabolites, including polyphenols, alkaloids, volatile oils, and others [2,38].Based on the proteomic results in this study, several proteins related to polyphenol biosynthesis were differentially expressed between the TPL and MGL tissues.CHS is the first committed enzyme involved in anthocyanin and flavonoid biosynthesis [39].Gene silencing (via RNA interference or virusinduced gene silencing) of CHS in both of Torenia hybrida and petunia resulted in lower contents of total flavonoid and anthocyanins, and CHS enzyme activity of flower tissues [39,40].In the present study, the abundance of CHS (spot No. 8406) was increased in TPL tissues, indicating that the biosysnthesis of anthocyanin and flavonoid was enhanced.This outcome was consistent with our previous finding that CHS expression was a positive regulator of anthocyanin biosynthesis [19].Furthermore, qPCR was employed to verify the higher gene expression level of CHS in TPL tissues (Fig 7F).CHI and FLS are two key enzymes involved in the flavonoid biosynthesis pathway [41][42].In CHI-suppressed transgenic tobacco plants, they showed up to 25% reduction in anthocyanins content [43].In the present study, the abundance of CHI (spot No. 4106) and FLS (spot No. 412) were higher in TPL tissues, which was consistent with the qPCR results, than in MGL ones ( Fig 7).Thus, higher abundance of CHI, CHS and FLS may account for the production of the purple-shoot phenotype in Wuyiqizhong 18 and enhance the flavonoid biosynthesis.These results were also agreed with our metabolic data, which show that the contents of TA and TP were significantly higher in TPL tissues than in MGL ones (Fig 2).

Proteins involved in stress responses
The production of reactive oxidative species (ROS) is inevitable in plant grown in aerobic circumstance.As an adaptive strategy, plants have evolved several capabilities to scavenge ROS including endogenous antioxidant enzymes system and antioxidant products, such as superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), and glutathione reductase (GR).Due to different sub-cellular localization, the abundance of these anti-oxidative enzymes did not change in the same pattern in response to an environmental stimulus [44].In our study, four proteins involved in stress responses have higher abundance in TPL tissues, while one protein named copper/zinc superoxide dismutase (Cu/Zn SOD, spot No. 5116) has lower abundance in TPL tissues than in MGL ones (Table 1; Fig 7M).Cu/Zn SOD catalyzes the dismutation of ion superoxide into oxygen and hydrogen peroxide in plant chloroplast [45].The lower abundance of Cu/Zn SOD in TPL tissues might be due to the well protection of anthocyanins.PHGPx, one of the protein isoforms of GPx, protects plants from pathogen attack and tolerance to other abiotic stresses [46].A previous study suggested that only part of the absorbed light energy is utilized in photosynthesis, and the excess photon flux could lead to photo-damage and ROS production due to decreased CO 2 assimilation in TPL tissues [19,47].As hypothesis, the increased abundance of PHGPx (spot No. 5114) and GST (spot No. 5106) observed in TPL tissues could quench the generated ROS by excess light and ameliorate the oxidative damage caused by ROS.Moreover, GPx and GSTs with GPx activity also decompose other toxic compound like alkyl hydroperoxides in addition to H 2 O 2 generated by SODs activity [48].Based on these results, we concluded that the accumulation of stress-related proteins in TPL tissues might dissipate the excess excitation energy and protect tender leaves against photo-damage.

Proteins involved in protein metabolism
Most of the proteins involved in protein metabolism have higher abundance in TPL tissues, compared with those in MGL tissues.In this study, the higher abundance of protein disulfide isomerase-like 1-4-like isoform 2 (PDIL, spot No.801), glutamine synthetase (GS, spot No.4407) and chaperonin family proteins (spot No.8701) indicated that there was an increased requirement for protein synthesis, correct assembly and turnover due to the development of tender leaves.Especially, GS (EC 6.3.1.2) is a key enzyme involved in the assimilation of ammonium into glutamine (Gln) [49,50].GS plays an important role in the synthesis of theanine, which is a unique amino acid found in tea plants and is the main component responsible for the taste of tea [51].GS showed higher in tender leaves and lower abundance in older leaves, indicating that the synthesis of theanine in tender leaves was enhanced [19,52].

Proteins involved in nucleic acid metabolism, the cellular cytoskeleton and transport
Glycine-rich RNA-binding proteins (GRPs) harbor an RNA-recognition motif (RRMs) at the N-terminus and a glycine-rich region at the C-terminus [53].The expression patterns of GRPs are regulated by a number of external stimuli, such as cold stress conditions [54], high salinity and osmotic stress [55].The high abundances of GRPs (spot No.5109) in TPL tissues could enhance stress resistance of tender tissue.Microtubules are dynamic heteropolymers of α-and β-tubulin, which are structural component of the cytoskeleton in eukaryote [56].It plays key roles in many basic processes of eukaryotic cells, such as cell division, cell motility, intracellular transport and the control of cell shape [57].The high abundance of α-tubulin (spot No. 2504) and β-tubulin (spot No. 1607) could maintain the stability of the cytoskeleton system, which is very important for the development of plants.

Conclusions
Tea plants with purple coloration represent important tea germplasms due to their economic value and health benefits.The pronounced changes in the proteomic profile between TPL and MGL tissues were included: 1) the lower activity of proteins associated with CO 2 assimilation, energy metabolism and photo flux efficiency and higher content of anthocyanins in TPL samples may protect tender leaves against photo-damage; 2) the higher abundance of CHS, CHI and FLS likely contributes to the synthesis of anthocyanins, catechins and flavonols in TPL tissues; 3) higher abundance of stress response proteins, such as GST and PHGPx, could enhance the tolerance of TPL tissues to adverse condition; and 4) the increased abundance of proteins related to protein synthesis, nucleic acids and cell wall proteins should be beneficial for the proliferation and expansion of leaf cell in TPL tissues.In summary, the results revealed that the higher abundance of two key proteins including CHS and CHI were responsible for the higher content of anthocyanins in TPL tissues compared with MGL ones.The higher abundance of GS protein related to the theanine biosynthesis may improve the flavor of tea products from TPL materials.This work enhances our understanding of the molecular mechanisms behind the changes in leaf color alteration between TPL and MGL tissues.Therefore, future studies should focus on the regulatory role of CHS and CHI in anthocyanin biosynthesis pathway of purple leaves of tea plants.

b:
gi number is from NCBI database of matched protein.c : Ratio means the ratio of TPL to MGL; 0 means protein spots were only detected in MGL tissues; 100 means protein spots were only detected in the TPL tissues.https://doi.org/10.1371/journal.pone.0177816.t001(Fig 5A).Interestingly, the majority of differentially expressed proteins involved in carbohydrate and energy metabolism were highly clustered (Fig 5B).The heatmap of protein responses in TPL and MGL samples shows the correlation patterns of the differentially abundant proteins (Fig 6).

Fig 6 .
Fig 6.Heatmap of proteomic responses from mature green leaf (MGL) and tender purple leaf (TPL) tissues of tea plants.Proteomic levels correspond to the color scale.A color gradient from blue through to red represents a low level through to a high level of proteomic response.After calculating the mean of each metabolite, the normalized data were generated using the Unscrambler software.https://doi.org/10.1371/journal.pone.0177816.g006

Table 1 .
(Continued) a : Spot number corresponds to the 2-DE gel in Fig 3.