Journal of Proteomics - 首页 - 植物抗逆整合生物学实验室€¦ · Recently, proteomics...

Contents lists available at ScienceDirect

Journal of Proteomics

journal homepage: www.elsevier.com/locate/jprot

Subcellular proteome profiles of different latex fractions revealed washedsolutions from rubber particles contain crucial enzymes for natural rubberbiosynthesis

Dan Wanga,b,1, Yong Suna,d,1, Lili Changa, Zheng Tonga, Quanliang Xiea,c, Xiang Jina,b,Liping Zhua,c, Peng Hed, Hongbin Lic,⁎, Xuchu Wanga,b,c,⁎

a Institute of Tropical Biosciences and Biotechnology, Chinese Academy of Tropical Agricultural Sciences, Haikou, Hainan 571101, Chinab College of Life Sciences, Ministry of Education Key Laboratory for Ecology of Tropical Islands, Hainan Normal University, Haikou, Hainan 571158, Chinac College of Life Sciences, Key Laboratory of Agrobiotechnology, Shihezi University, Shihezi, Xinjiang 832003, Chinad Rubber Research Institute, Chinese Academy of Tropical Agricultural Sciences, Danzhou, Hainan 571737, China

A R T I C L E I N F O

Keywords:Hevea brasiliensisNatural rubber biosynthesisRubber latexRubber particlesSubcellular proteomicsWashing solutions

A B S T R A C T

Rubber particle (RP) is a specific organelle for natural rubber biosynthesis (NRB) and storage in rubber treeHevea brasiliensis. NRB is processed by RP membrane-localized proteins, which were traditionally purified byrepeated washing. However, we noticed many proteins in the discarded washing solutions (WS) from RP. Here,we compared the proteome profiles of WS, C-serum (CS) and RP by 2-DE, and identified 233 abundant proteinsfrom WS by mass spectrometry. Many spots on 2-DE gels were identified as different protein species. We furtherperformed shotgun analysis of CS, WS and RP and identified 1837, 1799 and 1020 unique proteins, respectively.Together with 2-DE, we finally identified 1825 proteins from WS, 246 were WS-specific. These WS-specificproteins were annotated in Gene Ontology, indicating most abundant pathways are organic substance metabolicprocess, protein degradation, primary metabolic process, and energy metabolism. Protein-protein interactionanalysis revealed these WS-specific proteins are mainly involved in ribosomal metabolism, proteasome system,vacuolar protein sorting and endocytosis. Label free and Western blotting revealed many WS-specific proteinsand protein complexes are crucial for NRB initiation. These findings not only deepen our understanding of WSproteome, but also provide new evidences on the roles of RP membrane proteins in NRB.Significance: Natural rubber is stored in rubber particle from the rubber tree. Rubber particles were traditionallypurified by repeated washing, but many proteins were identified from the washing solutions (WS). We obtainedthe first visualization proteome profiles with 1825 proteins from WS, including 246 WS-specific ones. These WSproteins contain almost all enzymes for polyisoprene initiation and may play important roles in rubber bio-synthesis.

1. Introduction

Natural rubber is a plant-derived long chain cis-1,4-polyisoprenepolymer produced by more than 2,500 plant species, but the pararubber tree (Hevea brasiliensis) is the only commercially cultivated one[1–3]. Natural rubber cannot be replaced by chemical synthetic alter-natives owing to its unique properties [4,5]. Natural rubber latex, as thecytoplasm of a specialized cell called laticifer, is usually collected byregular tapping the trunk bark of the rubber tree [1,6]. Tapping, as anon-destructive method for latex harvesting, can facilitate continualrubber production [3,7]. During the regular tapping, latex regeneration

is an important limiting factor for rubber yield, which relies on both thecomplex rubber biosynthesis process on rubber particle (RP) and var-ious organic sources supplied to laticifer cells [8,9].

After ultracentrifugation, latex can be divided into three differentfractions: the top part, RP; the middle part, C-serum (CS); and thebottom fraction, lutoid [10,11]. RP is a specific latex-producing orga-nelle, and it is recently called as a natural rubber biosynthetic ma-chinery [12,13]. It is encapsulated by a phospholipid monolayer andcontains a high molecular weight rubber core [14]. The fresh latexsystem contains 30–45% weight of RP [15,16]. RP diameter is rangedfrom 0.02 to 3.0 μm [8,17,18]. It is supposed to originate from rough

https://doi.org/10.1016/j.jprot.2018.05.002Received 5 November 2017; Received in revised form 30 April 2018; Accepted 2 May 2018

⁎ Corresponding authors at: College of Life Sciences, Key Laboratory of Agrobiotechnology, Shihezi University, Shihezi, Xinjiang 832003, China.

1 Equal contributor.E-mail addresses: [email protected] (H. Li), [email protected] (X. Wang).

Journal of Proteomics 182 (2018) 53–64

Available online 03 May 20181874-3919/ © 2018 Elsevier B.V. All rights reserved.

T

http://www.sciencedirect.com/science/journal/18743919https://www.elsevier.com/locate/jprothttps://doi.org/10.1016/j.jprot.2018.05.002https://doi.org/10.1016/j.jprot.2018.05.002mailto:[email protected]:[email protected]://doi.org/10.1016/j.jprot.2018.05.002http://crossmark.crossref.org/dialog/?doi=10.1016/j.jprot.2018.05.002&domain=pdf

endoplasmic reticulum [19]. Rubber molecule biosynthesis is per-formed by various enzymes and protein factors, which are bound to orembedded in RP membrane. Among them, cis-prenyltransferase (CPT,also named rubber transferase), rubber elongation factor (REF) andsmall rubber particle protein (SRPP) are crucial [5,13,14,20,21]. REF isanchored inside RP membrane by its auto-assembly ability, whereasSRPP largely covers RP surface in an oriented anisotropic manner[5,22–25].

Recently, proteomics has been performed to identify proteins fromrubber latex and RP. Comparative proteomics of large and small RPsresulted in 53 differential spots corresponding to 22 gene products, andthe protein abundance of SRPP and 3-hydroxy-3-methylglutaryl-CoAsynthase (HMGS) in small RPs are higher than in large RPs, but REFabundance is lower in small RPs [18]. Moreover, 186 proteins wereidentified from RPs by shotgun [25]. Recently, we developed a phenol-based method for protein extraction from different latex fractions [26],and a method for protein extraction (REP) from the washing solution(WS) of RPs [16]. By comparison of the proteomes of primary andsecondary lutoids, we also found that chitinase and glucanase playcrucial roles in RP aggregation [11]. Proteomics of latex revealedethylene inhibits the expression of enzymes for RP aggregation toprolong latex flowing, and finally improves latex production. We alsofound that specific isoforms of REF and SRPP are mainly phosphory-lated at serine residues, which might be important for ethylene-stimu-lated latex production [8]. A latex proteomic study indicated thatproteins associated with latex regeneration and latex flow can be af-fected by both ethylene and methyl jasmonate [27]. Recently, 1,839unique proteins were identified from rubber latex by de novo sequen-cing from mass spectral data [28].

Traditionally, total RPs in the up cream layer were collected andthen resuspended and thoroughly washed three times with a Tris-buf-fered sucrose washing solution. The purified RPs were finally obtainedby centrifugation, and membrane proteins were isolated from the wa-shed RP for further proteomic analysis [18,25]. However, by extractingand identifying proteins with a re-extraction method with phenol (REP)from WS, we observed large amounts of proteins in these discardedsolutions [16]. In the past studies, WS proteins had been neglected ordiscarded [18,25]. We speculated that WS may contain many proteinsfor regulating natural rubber biosynthesis (NRB). Combined 2-DE andshotgun mass spectrometry (MS), we performed comprehensive pro-teomics of WS, CS and RP, and found many WS-specific proteins. As faras our best knowledge, it is the first proteomic study of WS from RP,and our results may deepen the understanding of RP proteins in NRB.

2. Materials and methods

2.1. Plant material

Fifteen newly tapped mature (8-year-old) rubber tree (H. BrasiliensisMull. Arg. Clone RY 7-33-97) were selected and randomized into threegroups for latex collection. These plants were grown in an experimentfarm of the Chinese Academy of Tropical Agricultural Sciences inDanzhou City, Hainan Province, China. After tapping, the first 10 latexdrops were discarded, and the subsequent drops were collected in ice-chilled glass beakers and taken back for lab analysis.

2.2. Collection of rubber particles, C-serum and washed solutions from latex

The fresh rubber latex was centrifuged at 40, 000g for 60min at4 °C. The middle clear CS and upper fraction containing RP were col-lected as described [18,25,26]. After collecting the upper cream of RP,the remained samples were immediately put into liquid nitrogen, andthe freezing middle CS icicle was cut out. The collected crude fractionsof RP and CS were put into different new tubes, respectively. Then, thecollected top creamy RPs were resuspended in an ice-cold washingsolution (10mM EDTA Na2, 0.5 mM dithiothreitol, 250mM sucrose,

and 10mM Tris-HCl, pH 7.0) in ratio of 1:10 (w/v) and stirred for30min and then ultracentrifuged at 30,000g for 15min at 4 °C. Theupper phrase and washed solution of non-rubber fractions were col-lected, respectively. Repeated the above steps for three times, andcollected the washed solution in each time. In the following research,the upper floated RP phrase and clear non-rubber fractions in thesecond washing time were collected as the purified RPs and WS. Threebiological replicates were performed for each sample.

2.3. Protein extraction and two-dimensional gel electrophoresis

Proteins in CS and RP were extracted as described [8,26]. For WSfrom RP, we collected the washed solutions from the first (WS-1),second (WS-2) and third (WS-3) washing time, then extracted theirproteins by the REP method as described [16]. In brief, an equal volumeof the extraction buffer was added to WS, and the mixtures were di-vided into six parts. Equal volume of Tris-saturated phenol (pH > 7.8)was added to one part of the divided samples. The mixtures were vor-texed thoroughly for 5min and centrifuged at 15, 000g for 5min at 4 °C.Then, the phenol-based upper phase was transferred into the nextportion of sample, and the extraction procedure was repeated until allWS solutions had been extracted using the same phenol. This extractionstep was repeated once more to remove interfering compounds from themixture solution. The final upper phenol phase was transferred into anew centrifuge tube, and five volumes of ammonium sulfate-saturatedmethanol were added to precipitate proteins.

Protein concentration was determined by Bradford assay using BSAas standard. For 2-DE, 1300 μg of proteins from WS, CS and RP wereloaded onto the 24 cm, pH 4–7 linear gradient IPG strips (GEHealthcare, Uppsala, Sweden). The separation of proteins in the seconddimension was performed with SDS polyacrylamide gels (12.5%). Thegels were visualized by CBB-G250 staining. Gel image analysis wasperformed with Image Master 2D Platinum Software Version 5.0 (GEHealthcare, Uppsala, Sweden). Three biological replicates were per-formed for all the samples.

2.4. Protein identification via mass spectrometry

All high abundance protein spots (Vol > 0.01%) in the 2-DE gels ofWS were manually excised and in-gel digested with bovine trypsin asdescribed [8,11]. After trypsin digestion, the collected peptides werespotted on the MALDI plate, and cyano-4-hydroxycinnamic acid asmatrix was added onto the dried peptides. The samples were submittedinto a 5800 MALDI-TOF/TOF mass spectrometer (AB SCIEX, FosterCity, CA). Twenty strongest peaks of the TOF spectra per sample werechosen for MS/MS analysis. The measured tryptic peptide masses weretransferred to ProteinPilot Software (Version 5.0) and searched againsta self-constructed database derived from the original H. brasiliensisgenome scaffolds (BioProject ID: PRJNA80191, www.ncbi.nlm.nih.gov/nuccore/448814761) and the draft genome (GenBank:AJJZ01000000) as described [29] using an in-house MASCOT server(Version 2.3, Matrix Science, USA). Tryptic digestion with a maximumof 1 missed cleavage was considered. The cysteine was considered fixedmodified by carbamidomethylation and a variable modification ofmethionine oxidation (M). Spots were considered positively identifiedwith a Mascot score higher than 75 (p < 0.001, the threshold with95% confidence is 31 for rubber tree protein identification), at least 2matched peptides, and maximum peptide coverage higher than 5%(Table S1). In addition, an in-house BlastP searching was performed atNCBI (http://www.ncbi.nlm.) to confirm all identifications and findtheir homologous proteins.

2.5. Shotgun and label free quantitative proteomic analysis of WS, CS andRP

The 100 μg of proteins obtained from WS, CS and RP were

D. Wang et al. Journal of Proteomics 182 (2018) 53–64

54

http://www.ncbi.nlm.nih.gov/nuccore/448814761http://www.ncbi.nlm.nih.gov/nuccore/448814761http://www.ncbi.nlm

respectively reduced and alkylated by dithiothreitol and iodoaceta-mide, then digested by trypsin. The products of enzymatic hydrolysiswere separated by an UltiMate 3000 system (Thermo Fisher Scientific,Rockford, USA). Fifty washed fractions were collected in 1.5 mL tubesat one minute interval. Fractions were then dehydrated and mergedinto twenty fractions, and these fractions were subjected to the highresolution mass spectrometer. The enzymatic hydrolysis peptides wereanalyzed by using HPLC-ESI-MS/MS system. Trap and elute mode wasused to separate the samples using the Dionex Ultra 3000 nano HPLCsystem equipped with a trap column and a separation column. Massspectrometer recalibration was performed at the start of each sampleusing a β-galactosidase digest standard.

Protein identification was performed using ProteinPilot™ softwareV5.0 (AB SCIEX) as described [8,11]. For LC-MS/MS quantification,peptides were automatically selected by ProGroupTM algorithm andProteinPilot™ software to calculate the reporter peak area, error factor(EF) and p value. A reverse database search strategy was adopted toestimate false discovery rate (FDR) for peptide identification. A strictunused confidence score of 2.0, which corresponds to at least a peptideconfidence 99%, was used as qualification criterion. The identifiedproteins contained at least two matched peptides higher than 95%confidence, and a FDR value≤ 1% was used to perform protein

quantification. Furthermore, genome scaffolds from the rubber tree asdescribed [29] were used for validating protein identification.

Label free proteomic analysis was further performed to determinethe quantitative changes in protein abundance as described [30]. It isbased on measuring peptide ion peak intensity observed in low collisionenergy mode in a triplicate set. For protein quantification, data setswere normalized using the auto-normalization function in Maxquantsoftware (version 15.3.30) and searched against a self-constructed da-tabase as above described. Tryptic digestion with a maximum of 2missed cleavages was considered. The cysteine was considered to bemodified by carbamidomethylation and a variable modification ofmethionine oxidation (M). Mass tolerances for peptide and fragmentions used was 40 ppm, criteria for acceptance of peptide assignmentand protein identification was that unused score > 1.3 and at least twomatched peptides. FDR assessment is 0.01. Three biological replicatesof proteins from WS, CS and RP were normalized to the intensity ofmany qualitatively matched proteins (or peptides). Included limits wereall protein hits that were identified with a confidence of 99%. Allproteins whose abundances were significantly different between dif-ferent samples were manually assessed by checking the matched pep-tide and replication level across samples.

Fig. 1. Proteome profile of WS and MS identificationof main WS proteins.The total proteins from WS in the second (A, WS-2),first (B, WS-1) and third (C, WS-3) washing times ofRPs, as well as from CS (D) and RP (E), were sepa-rated by 2-DE, and the gels were visualized by CBBG-250. The positively identified protein spots aremarked with numbers from 1 to 233. The proteinidentities are presented in Table S1 and Fig. S2.


55

2.6. Western blotting analysis of the typical identified proteins in differentlatex fractions

Western blotting was performed as described [8,11]. For validationof protein expression levels, 3 biological replicates were performed forWestern blotting. The changed ratios of these proteins in the differentfractions were calculated on band abundance by ImageQuant TL soft-ware (GE Healthcare, Uppsala, Sweden). At the same time, the averagematched peptide numbers in proteins were calculated and the proteinaccumulation patterns in WS, CS and RP were determined based onshotgun and label free results.

2.7. Protein classification and STRING analysis

The identified proteins were searched against the UniProt databaseto confirm their functions. Proteins were then classified by usingFunctional Catalogue software to obtain their corresponding COGcodes. Gene Ontology (GO) pathway analysis was performed by KOBAS3.0 and WEGO software. Finally, an in-house BLAST searching atUniProtKB was performed for each protein to find its homology andconfirm its cellular component, biological process and molecularfunction. In addition, the identified proteins were analyzed usingSTRING V.9.1 database (http://string-db.org/) to examine protein-protein interactions. A high confidence view (score 0.9) was selected topresent protein-protein networks.

3. Results

3.1. MALDI TOF/TOF MS identification of the main proteins in 2-DE gel ofWS

To obtain the proteomic profiling of WS from RP, we collected thewashed solutions from the three round of washing. By comparing 2-DEgels, we noticed that protein profiles in WS-1, WS-2 and WS-3 weredifferent with each other. Although the highly abundant protein spotswere all present, many low abundant spots were disappeared with in-creased washing in these 2-DE gels (Fig. 1A–C). We think the proteinsfrom WS-1 may contain some contaminated proteins from CS, andmany RP membrane-combined proteins were lost in WS-3. Therefore,we used the proteins from WS-2 as WS proteins in the following study.

We calculated the protein spot abundance in the 2-DE gels of WS,and detected 1164 ± 27 spots (Fig. 1; Fig. S1). We finally identified233 protein spots, representing 142 unique proteins (Fig. 1; Fig. S2;Table S1). Among them, 28 spots marked with red color cannot bedetected in CS and RP gels, thus were termed as WS-specific spots(Fig. 1A). These WS-specific spots contain 26 unique proteins. Pathwayanalysis showed that they participate in multiple pathways, includingmetabolic pathways, biosynthesis of secondary metabolites, metabolismof amino acids and so on (Fig. 1A; Fig. S3; Table S1). In particular,malate dehydrogenase (Spot 226, referred as S226 hereafter) was in-volved into 8 pathways. The 5 proteins named beta-glucosidase (S175),aconitate hydratase (S220), phosphoglycerate kinase (S129 and S225),malate dehydrogenase (S226) and glucose-6-phosphate 1-epimerase(S22) were involved in biosynthesis of secondary metabolites. SeveralWS-specific proteins, such as peptidyl-prolyl cis-trans isomerase (S79and S222), ubiquitin carboxyl-terminal hydrolase (S13), REF (Spots 27,229–231) and elongation factor 1 (S224), are well known to take im-portant roles in NRB. The positions of all the positively identifiedproteins are indicated by arrows and marked with numbers on the WS2-DE gel (Fig. 1A), and their MS identities are listed in the supple-mentary data (Fig. S2; Table S1).

Among these WS proteins, REF and SRPP are abundant (Fig. 1;Table S1). Notably, many protein spots on 2-DE gel were identified asthe same proteins. They were termed as different protein isoforms orprotein species in the 2-DE gel-based proteomic study. These proteinspecies included REF that identified from 19 spots (S27, 42, 43, 45, 73,

86, 93-98, 189, 229, 230 and 231, with a total of abundance volume of31.81%), SRPP identified from 13 spots (S31, 62, 74, 76, 87, 88, 99-102, 195, 196 and 200, with total volume of 17.93%), proteasomesubunits from 10 spots (S33, 34, 69, 109, 110, 145, 166, 169, 187 and193), prohevein from 7 spots, actin-7 from 6 spots, phosphoglyceratekinase from 5 spots, etc. It is noteworthy that about half abundance(49.79%) was occupied by REF and SRPP (Fig. 1; Table S1). Theseprotein species had the same protein name, same amino sequence, andsimilar functions, but demonstrated different experiment isoelectricpoint and/or molecular weight on the 2-DE gels. These protein speciesmay have changed their molecular weight and isoelectric point by post-translational modifications (PTM) from the original ones.

3.2. Functional analysis of the high abundance WS proteins from 2-DE gels

We used WEGO software to determine the detailed pathways in the233 high abundance WS proteins, and found they are involved in 60pathways (Fig. 2A; Table S1). There are 7 main pathways which at leastcontain 11 WS proteins, including metabolic pathway (MP, 39 pro-teins), biosynthesis of secondary metabolites (BSM, 31 proteins),carbon metabolism (CM, 24 proteins), biosynthesis of amino acids(BAA, 18 proteins), metabolism of amino acids (MAA, 16 proteins),proteasome (PT, 13 proteins), and glycolysis/gluconeogenesis (GG, 11proteins). Eight important pathways contain at least 5 WS proteins;they are protein processing in endoplasmic reticulum (PPER, 9 pro-teins), carbon fixation in photosynthetic organisms (CFPO, 7 proteins),glyoxylate and dicarboxylate metabolism (GDM, 6 proteins), pyruvatemetabolism (PM, 6 proteins), amino sugar and nucleotide sugar meta-bolism (ASM, 6 proteins), endocytosis (EC, 5 proteins), inositol phos-phate metabolism (IPM, 5 proteins) and starch and sucrose metabolism(SSM, 5 proteins) (Fig. 2A; Table S1).

GO distribution analysis of all WS proteins was further performed toconfirm the cellular component, molecular function and biologicalprocess (Fig. 2B). Finally, 207 proteins show GO information. Theywere classified into 3 large groups, containing 21 subgroups based ontheir functional annotation. In cellular component, the largest propor-tion including 53 proteins is in the cell (GO: 0005623), 17 proteinsoccur in macromolecular complex (GO: 0032991), and 12 proteins arein organelle (GO: 0043226). For molecular function ontology, 7 sub-categories were assigned. The largest portion including 96 proteinshave catalytic activity (GO: 0003824), followed by the 58 proteins withbinding function, and then, followed by several proteins with anti-oxidant activity, enzyme regulator activity, translation regulator ac-tivity, transporter activity, and molecular transducer activity (Fig. 2B).In biological process, the largest part including 67 proteins are relatedto metabolic process (GO: 0008152), followed by cellular process (GO:0009987) with 60 proteins, and the other important biological pro-cesses, including response to stimulus, localization, biological regula-tion, pigmentation, cellular component organization, multi-organismprocess and death (Fig. 2B). In addition, the 142 unique proteins weresearched against the Uniprot protein database, and the results are si-milarly with WEGO (Fig. S3; Table S1).

Furthermore, these WS proteins were subjected to the STRING da-tabase to evaluate their protein-protein interactions. In STRING data-base, there has no H. brasiliensis or its closely related species data, so wetransformed the H. brasiliensis proteins to their correspondingArabidopsis thaliana protein sequences, then used A. thaliana databasefor further analysis. In this study, we selected a high confidence view(score 0.9) to present the protein-protein networks. Finally, among theidentified 142 unique proteins, 45 are involved in protein-protein in-teractions (Fig. 2C). These interacted proteins were classified into 4subclusters. Cluster 1 includes 36 protein species, representing 21 un-ique proteins. They are HSP 17.6 II, AT1G53540 (18.5 kDa class I heatshock protein, S57), HSP 70b (S7 and S78), Hop3 (Hsp70-90 organizingprotein, S81), Hsp81.4, HSC70-1(heat shock cognate 70 kDa protein,spots 6, 9, 116 and 136), ROF2 (peptidyl-prolyl cis-trans isomerase),


56

http://string-db.org

F3O9.27 (proteasome alpha-subunit), PBF1 (proteasome subunit betatype-1), etc. In cooperation with other chaperones, HSP70 stabilizespreexistent proteins against aggregation and mediates the folding ofnewly translated polypeptides in the cytosol as well as organelles. Thesechaperones participate in these processes by recognizing nonnativeconformations of other proteins. RPN12a acts as a regulatory subunit ofthe 26S proteasome, which involves in the ATP-dependent degradationof ubiquitinated proteins. Therefore, proteins in cluster 1 might playimportant roles in protein folding and degradation.

Six unique proteins were defined to cluster 2. They are PGK(phosphoglycerate kinase, spots 51, 129, 158 and 225), TPI (triose-phosphate isomerase, spots 146, 147, 183 and 204), GAPC2 (glycer-aldehyde 3-phosphate dehydrogenase, S213), LOS2 (enolase, spots 14,107, 132 and 133), iPGAM1 (bisphosphoglycerate-independent phos-phoglycerate mutase 1, S153), iPGAM2 (bisphosphoglycerate-in-dependent phosphoglycerate mutase 2, S179). These proteins areknown as enzymes that participate in carbohydrate metabolism. Incluster 3, three unique proteins, named OASA1 (cysteine synthase, S130and S203), HOG1 (adenosylhomocysteinase 1, spots 155, 157 and 180)and SAM2 (S-adenosylmethionine synthase, S210), were observed.Among them, cysteine synthase catalyzes O-3-acetyl-L-serine and hy-drogen sulfide to format L-cysteine and acetate; adenosylhomocysteineis a competitive inhibitor of S-adenosyl-L-methinine-dependent methyltransferase reaction. In cluster 4, there are 7 partners, and each partnercontains 2 proteins showing protein interaction with each other(Fig. 2C). Among them, MMZ3 (ubiquitin-conjugating enzyme E2 var-iant 1D, S35) interacts with UBC35, and ADF4 (actin-depolymerizingfactor, S59 and S144) interacts with ACT7 (actin family protein, spots 8,11, 121 and 138). A nuclear transport factor (NTF2A, S214) can in-teract with ran-binding protein 1 (AT2G30060, S141) (Fig. 2C).

3.3. Identification of proteins from WS, CS and RP by high-throughputshotgun analysis

We further performed high-throughput shotgun analysis of theproteins from WS, CS and RP. Finally, we identified about two thousandunique proteins from these fractions (Fig. 3; Table S2). For CS, weidentified 1824, 1966 and 1873 proteins from three independentshotgun experiments. Totally, 2559 proteins were identified at least onetime. Among them, 1837 proteins were determined in at least two re-peats, and 1268 proteins were identified from the three experiments(Fig. 3A). For RP, 1216, 1247 and 860 proteins were identified in thefirst, second and third shotgun experiment, respectively. Among them,1743 proteins were identified at least one time, 1020 proteins wereobtained from at least two experiments, and 560 proteins were detectedin all three experiments (Fig. 3B). For WS, 1941, 1923 and 1707 pro-teins were obtained in the three experiments, respectively. A total of2575 WS proteins were examined from at least one experiment, 1799proteins were from at least two times, and 1197 proteins were gener-ated from all experiments (Fig. 3D). In the following study, we selectedthe proteins positively identified from at least two independent shogunexperiments as the finally identified ones. Therefore, a total of 1837,1799 and 1020 unique proteins were identified from CS, WS and RP,respectively.

We combined the proteins identified by 2-DE and shotgun experi-ments, and obtained a total of 1825 proteins in WS. Among them, 116shared proteins were generated from the two kinds of methods, andcompared to shotgun results, only 26 specific proteins were obtained by2-DE-based method (Fig. 3C). Venn diagram of WS, CS and RP proteinsresulted in 740 shared proteins. For CS and WS, 1444 shared proteinswere determined. Only 785 proteins were identified from both CS andRP. It is noteworthy that 852 proteins were detected from both WS and

Fig. 2. Functional analyses of the identified high abundant WS proteins.The 233 identified high abundant proteins from 2-DE gels for WS were performed bioinformatics by KOBAS 3.0 (A), Gene Ontology by WEGO (B), and interactionanalysis by STRING software (C).


57

RP, indicating many proteins may not tightly bind to the surface of RPand can be washed away from RP. There were 348, 269 and 123 pro-teins that were identified only from CS, WS and RP, respectively(Fig. 3E; Table S2). Label free quantitative proteomics resulted in 2638proteins and most of them were determined as differentially expressedproteins (DEPs) with more than 1.5-fold change among CS, WS or RP(Table S3). Interestingly, we blasted the 26 WS-specific proteins from 2-DE gel, and found that almost all highly abundant proteins could beidentified from WS, CS and/or RP. Only three proteins, named 26Sprotease regulatory subunit 7 (S156), non-specific phospholipase C3(S223) and actin-7 (S227), were only identified from WS by 2-DE gel.These results revealed that many DEPs from 2-DE gel-based methodcannot be detected as specific proteins by shotgun analysis. The verylow overlap between the outputs of the two proteomic approaches isprobably owing to their different characteristics. The detected proteinson 2-DE gels may be the changes only in protein forms, rather than thechanges in the overall protein accumulation. These proteins may havepost-modifications that showed as different protein spots in 2-DE gels,and these differential spots can often be identified as protein forms/species in proteomic study. Our special interest will be focused onanalyzing the potential biological functions of the 246 WS-specificproteins.

3.4. Gene Ontology classification and protein-protein interaction of WS-specific proteins

We further performed WEGO analysis of the 246 WS-specific pro-teins (Fig. 3E) and the enriched outputs of their biological process(Fig. 4A), molecular function (Fig. 4B) and cellular component (Fig. 4C)were presented. At the biological process level, 13 processes such asbiosynthesis, single-organism metabolism and organic substance me-tabolic processes were detected. Among them, the most portion in-cluding 110 proteins are taken part in organic substance metabolicprocess, followed by 106 proteins in primary metabolic process andcellular metabolic process, respectively (Fig. 4A; Table S4). A large partincluding 69 proteins are related to the biosynthetic process. They arepeptidyl-prolyl cis-trans isomerase, hydroxyacyl dehydratase, gluta-mine synthetase, methyltransferase, fucokinase, aminotransferase, eu-karyotic translation initiation factor, ribosomal protein, etc. For mole-cular function classification, nine pathways were determined. Amongthem, the most portion including 66 proteins are taken part in organiccyclic compound binding, followed by 65 proteins heterocyclic com-pound binding and 56 proteins with ion binding activities (Fig. 4B;Table S4). It is noteworthy that 32 proteins show transferase activity(GO: 0016740); they involve in catalysis of the transfer of one com-pound to another compound. These processes often result in post-translational modifications. They are phosphate guanyltransferase(N641 and 1669; N stands for the protein No., hereafter), methyl-transferase (N956, 1203, 1682 and 1892), aminotransferase (N929 and1646), uridyltransferase (N1113), pyrophosphorylase (N762), gluco-syltransferase (N1489), protein kinase (N760, 1072, 1482, 1499 and1946), farnesyl pyrophosphate synthase (N1835), acetyltransferase(N493), glucanotransferase (N1290) and ubiquitin-protein ligase(N1011). At subcellular level, 12 components were observed. Amongthem, almost half of these proteins are localized to intracellular part orintracellular organelle. There are also many proteins in membrane-bounded organelle or as protein complex in endomembrane system(Fig. 4C; Table S4).

Protein-protein interactions of the 246 WS-specific proteins showed96 WS proteins with strong relationship (Fig. 4D). These interactedproteins were classified into 5 clusters, including plant cytokinesis andendocytosis (9 proteins), proteasome (11 proteins), ribosomal meta-bolism (46 proteins), protein folding and energy metabolism (18 pro-teins), and protein sorting (12 proteins). The largest cluster including46 proteins is related to ribosomal metabolism (Fig. 4D). Almost half ofthem are 40S and 60S ribosomal proteins (Table S4). It is known that

most proteins are synthesized on ribosome and processed in the en-doplasmic reticulum to the trans-Golgi network, and then to their finaldestination compartment. These vesicles have specific coat proteins likeclathrin. Seven elongation factors and six subunits of importin wereobserved in this cluster. Four members (importin subunits 9, 14, 68 and72) of signal recognition particle are WS-specific proteins; they involvein ribosomal metabolism. Several proteins like ubiquitin, WD40 do-main-containing protein, and actin were also detected in this cluster.

The second cluster with 18 WS proteins is involved in proteinfolding and energy metabolism. They are ATPase subunits (N723, 752and 1102), peptidylprolyl isomerase ROF2 (N290 and 1149), phos-phatase, pyrophosphorylase, and four HSP members (HSP90.1, HSP20-L, MTHSC70-2, and Cpn60 beta2). The third cluster contains 12 pro-teins that take crucial roles in protein sorting. They are seven isoformsof importin, four members of actin, and a vacuolar sorting-associatedprotein. The fourth cluster contains 11 proteins related to proteasome,and they are six proteasome components, two members of tubulin(N1712 and 1948), polyadenylate-binding protein (N1864), vesicle-fusing ATPase (N699), and prefoldin (N1372). Nine proteins in the fifthcluster for cytokinesis and endocytosis are subunits of adaptin protein(AP) complexes (N826, 979, 988, 1302, 1339 and 1459), a coat proteinnamed clathrin or coatomer (N925, CSB-2), ras-related protein RABD1(N786, ATFP8), and vesicle transport protein (Fig. 4D; Table S4).

3.5. Determination of protein accumulation patterns in WS, CS and RP

The accumulation patterns of 15 typical proteins in different latexfractions were further determined by Western blotting (Fig. 5A), andtheir relatively changed ratios were calculated (Fig. 5B). Among them,14 proteins except for beta-cyanoalanine synthase (CYAS) could bepositively detected in WS. For these WS proteins, two proteins namedgeranylgeranyl diphosphate synthase (GGPS) and acetyl-CoA carbox-ylase (ACOC) could not be positively identified by shotgun and labelfree analyses from WS (Fig. 5C; Tables S2, S3). These proteins areclassified into three typical groups based on their functions; they areinvolved in natural rubber biosynthesis, defense system, and signaltransduction. Among them, five proteins named farnesyl pyrophosphatesynthase (FPPS), SRPP, REF, GGPS and ACOC are involved in naturalrubber biosynthesis; ten proteins have relation to defense response; andtwo proteins named calcium-dependent protein kinase (CDPK) and 14-3-3 protein (1433) take part in signal transduction. By comparing thechanged ratios from Western blotting, shotgun and label free analyses,we noticed that most proteins showed similar changed patterns in thethree kinds of methods (Fig. 5B and C).

REF and SRPP could be detected in all the checked componentsincluding CS, WS, URP (unwashed RP) and RP in this study. For FPPS ornamed as farnesyl diphosphate synthase (FADS), the most abundantband was found in CS by Western blotting, whereas the most abundantone was detected in WS by shotgun and label free. The abundance ofGGPS in CS, WS and URP is very low, and it wasn't detected in RP.ACOC is a biotin-dependent enzyme that catalyzes the irreversiblecarboxylation of acetyl-CoA to produce malonyl-CoA. It provides themalonyl-CoA substrate for fatty acid biosynthesis, which is crucial forNRB. This enzyme was found to mainly accumulate in CS, and couldalso be detected in WS, but it could hardly be detected from URP andRP. The 14-3-3 protein showed similar accumulation pattern to ACOC,and it was mainly found in CS and WS. CDPK plays a crucial role insignal transduction pathway, in which calcium is as a second mes-senger. This enzyme was mainly detected from URP and CS, and itsaccumulation is very low in RP.

Western blotting also showed that the accumulation patterns indifferent latex fractions are similar. Among them, CYAS is mainly ac-cumulated in RP and CS, and even cannot be detected from WS.However, it was identified from WS, CS and RP by shotgun and labelfree, although the largest peptide number was obtained from RP. HSP70is mainly accumulated in CS and WS, but it is poor in RP and URP.


58

Superoxide dismutase (SOD) and ascorbate peroxidase (APX) are im-portant in nearly all living cells exposed to oxygen. Their accumulationpatterns in different latex fractions are similar, and they are rich in CSand WS, but poor in RP. Pyridoxal biosynthesis protein PDX1 is anethylene inducible protein, and it is rich in CS, but poor in RP. Thechanged patterns for glutathione reductase (GTRD), glutamate-cysteineligase (GTCL) and latex cyanogenic beta glucosidase (GLUD) are similarwith each other, and they accumulate mainly in CS, followed by WS andURP (Fig. 5; Tables S2, S3).

3.6. Functional analysis of all WS proteins and comparison of theaccumulation patterns of enzymes involved in NRB

Functional analysis of the WS proteins resulted in 17 categories (Fig.S4; Table S2). Among them, the largest part including 560 proteins arefunction unknown, followed by 280 proteins involved in translationand ribosomal structure, and 276 proteins participate in protein de-gradation. There are 186 proteins involved in carbohydrate transportand metabolism. There are also many sucrose biosynthesis and meta-bolism related enzymes, including sucrose synthase, sucrose phosphatesynthase, and sucrose phosphate phosphatase, etc. A large number ofproteins are involved in signal transduction mechanism (166 proteins,such as 14-3-3 protein, calcium ion binding protein, rab family protein,and GTP-binding protein), posttranslational modification, proteinturnover and chaperones (162 proteins), material transportation (161proteins, including importin, protein transport protein, vacuolar pro-tein sorting-associated protein, and ABC transporter, etc.), amino acidtransport and oxidation defense. Sixty proteins are involved in energy

Fig. 3. Comparison of the proteins identified from CS, WS and RP.The numbers of identified proteins from CS (A), RP (B) and WS (D) by shotgunin three biological duplicates are highlighted. The proteins identified by both 2-DE and shotgun from WS were incorporated and 116 shared proteins wereobserved in the two methods (C). Venn diagram of the identified proteins fromCS, WS and RP are also presented (E).

Fig. 4. Classification and protein-protein interactions of WS-specific proteins.GO analysis of the 246 WS-specific proteins is presented at their biological process (A), molecular function (B), and cellular component (C). The numbers in figure Aare: 1, biosynthetic process; 2, cellular component organization; 3, nitrogen compound metabolic process; 4, single-organism metabolic process; 5, cellular locali-zation; 6, cellular metabolic process; 7, cellular component biogenesis; 8, establishment of localization; 9, single-organism cellular process; 10, macromoleculelocalization; 11, primary metabolic process; 12, organic substance metabolic process; and 13, regulation of cellular process. The numbers in figure B are: 1,transferase activity; 2, protein binding; 3, small molecule binding; 4, heterocyclic compound binding; 5, carbohydrate derivative binding; 6, ion binding; 7, hydrolaseactivity; 8, organic cyclic compound binding; and 9, structural constituent of ribosome. The numbers in figure C are: 1, ribonucleoprotein complex; 2, proteincomplex; 3, intracellular organelle part; 4, intrinsic component of membrane; 5, non-membrane-bounded organelle; 6, plasma membrane; 7, membrane-boundedorganelle; 8, cell periphery; 9, endomembrane system; 10, intracellular part; 11, intracellular; and 12, intracellular organelle. These proteins were further performedprotein-protein interaction analysis by STRING software (D). The detailed GO information and abbreviations are listed in Table S4.


59

production, and they are malate dehydrogenase, vacuolar ATP syn-thase, ATP synthase, V-type proton ATPase, etc. Proteins involved incell division, cell structure, nucleotide transport and metabolism arealso enriched in WS. It is noteworthy that 26 proteins are involved inlipid metabolism, and they are long-chain-fatty-acid CoA ligase, enoyl-ACP reductase, 2-hydroxyphytanoyl-CoA lyase, etc. These proteins areimportant for biosynthesis of isoprenoids in many plants.

The functions of the shared proteins in different latex fractions alsodetermined (Fig. S4; Table S2). Our special interests were paid on theshared proteins in WS and RP fractions, and many proteins are involvedin signal transduction mechanisms, posttranslational modification,protein degradation, and ribosomal metabolism (Fig. S4; Tables S2, S4).In WS, CS and RP, 1274 proteins were shared within at least twofractions. And 268 proteins are functional unknown, followed by pro-teins involved in protein degradation (149 proteins), translation andribosomal structure (141 proteins), carbohydrate metabolism, etc. Inthe shared 837 proteins between CS and WS, 161 proteins are func-tional unknown, followed by proteins involved in ribosomal structure,protein degradation and basic metabolism.

Finally, we compared the accumulation patterns of WS proteins thatinvolved in NRB (Fig. 6; Fig. S5; Table S2). We noticed many WS pro-teins are directly or indirectly involved in the initiation process of NRB;this process is divided into 3 main stages. The first stage is sucrosebiosynthesis and metabolism (Fig. S5). In sucrose biosynthesis, 8 en-zymes are crucial. Among them, triosephosphate isomerase (TPI) cat-alyzes dihydroxyacetone phosphate (DHAP) to glyceraldehyde 3-phos-phate (G3P), and aldolase (AL) catalyzes DHAP and G3P to formfructose-1,6-biphosphate (FBP). Then, FBP can be catalyzed by a seriesof enzymes to synthesize sucrose, and these enzymes include fructose-1,6-bisphosphatase (FBPase), glucose-6-phosphate isomerase (GPI),phosphoglucomutase (PGM), UDPG pyrophosphorylase (UGP), sucrose-phosphate synthase (SPS), and sucrose phosphate phosphatase (SPP).On the other side, invertase (IT) catalyzes the metabolism of sucrosedegradation to form glucose and fructose, and sucrose synthase (SS)catalyzes sucrose to fructose. In the following steps, glucose and

fructose can be catalyzed by hexokinase (HK) to form glucose-6-phos-phate (G6P). It is known that G6P can directly take part in pentosephosphate pathway, or catalyzed by GPI to form fructose-6-phosphate(F6P), and then fructose-6-phosphate participates in glycolysis pathwayto biosynthesis of acetyl CoA (Fig. 6). Excess sucrose can be stored asstarch which is a carbon source under amylase (AA) degradation. In thesecond stage, enzymes are mainly involved in glycolysis (Fig. S5C) andpentose phosphate metabolism (Fig. S5D). The two pathways provideinitiation substrates for mevalonic acid (MVA) pathway, and can pro-duce large amounts of isopentenyl diphosphate (IPP). In glycolysispathway, after a series of enzymatic reaction, F6P is finally turned intopyruvate, and several enzymes including fructose-1,6-biphosphate ki-nase (FBK), fructose-1,6-biphosphate aldolase (FBA), TPI, glycer-aldehyde 3-phosphate dehydrogenase (G3PD), phosphoglycerate kinase(PGK), phosphoglycerate mutase (PGAM), enolase (EL), and pyruvatekinase (PK) are involved into this process. Then, pyruvate can be cat-alyzed by pyruvate dehydrogenase complex (PDC) to form acetyl CoA, akey initiation substrate for MVA pathway. In pentose phosphatepathway, G6P is finally turned to G3P.

In this process, several enzymes including glucose-6-phosphat-de-hydrogenase (GPD), 6-phosphogluconolactonase (PGL), 6-phosphoglu-conat dehydrogenase (PGD), ribulose-phosphate epimerase (RPE),transketolase (TK), and transaldolase (TD) are important. The thirdstage contains the MVA and 2C-methyl-D-erythritol-4-phosphate (MEP)pathways (Fig. S5). It is widely known that isoprenoid precursors forrubber biosynthesis are mainly provided by the cytosolic mevalonatedependent pathway to form IPP. The MVA pathway starts with acetyl-CoA acetyltransferase (ACAT) to catalyze the formation of acetyl acetyl-CoA, and then, through a series of enzymatic reactions, acetyl-CoA canbe catalyzed to form IPP. During this process, several enzymes in-cluding 3-hydroxy-3-methylglutaryl coenzyme A synthase (HMGS),HMG-CoA reductase (HMGR), mevalonate kinase (MEVK), phospho-mevalonate kinase (PMVK), mevalonate-5-pyrophosphate decarbox-ylase (MVD), and IPP isomerase (IPI) are important (Fig. 6). The plas-tidic MEP pathway is also suggested to contribute IPP for rubber

Fig. 5. Accumulation patterns of the 15 typical pro-teins in WS, CS, URP and RP.The 15 typical proteins were examined by Westernblotting and their accumulation patterns in WS, CS,unwashed rubber particles (URP), and RP are high-lighted (A). The relative changed ratios of theseproteins determined by Western blotting (B),Shotgun (C) and Label free (D) are presented. Theabbreviations are: FPPS, farnesyl pyrophosphatesynthase;SRPP, small rubber particle protein; REF,rubber elongation factor; CYAS, beta-cyanoalaninesynthase; CDPK, calcium-dependent protein kinase;GGPS, geranylgeranyl diphosphate synthase; ACOC,acetyl-CoA carboxylase; 1433, 14-3-3 protein;HSP70, heat-shock 70 kDa protein; SOD, superoxidedismutase; APX, ascorbate peroxidase; GTRD, glu-tathione reductase; PDX1, ethylene inducible proteinpyridoxal biosynthesis protein PDX1 like; GTCL,glutamate—cysteine ligase; GLUD, latex cyanogenicbeta glucosidase.


60

biosynthesis, and this pathway starts with 1-deoxy-D-xylulose 5-phos-phate synthase (DXS) to catalyze G3P and pyruvate to form 1-deoxy-D-xylulose 5-phosphate (DXP), and finally form IPP. But in this study, wedidn't identify any enzymes in MEP pathway (Tables S2 and S5),

indicating that MEP pathway is not as important as MVA pathway forNRB. Noteworthy is that almost all aforementioned enzymes have beenidentified from WS by shotgun analysis.

In rubber elongation process, geranyl pyrophosphate synthase (GPS)catalyzes IPP and DMAPP to form geranyl diphosphate (GPP), andFADS to produce farnesyl diphosphate (FPP) by adding IPP onto GPP,then GGPS and geranylgeranyl pyrophosphate synthase (GGPPS) ortermed as geranylgeranyltransferase can catalyze farnesyl diphosphateto form geranylgeranyl diphosphate (GGPP). Finally, GGPP can gen-erate natural rubber hydrocarbons contain different length of carbonwith the help of CPT. During the elongation process of rubber bio-synthesis, REF, SRPP and a Nogo-B receptor named as HRT1-REFbridging protein (HRBP) are known as RP membrane binding proteinsthat play crucial roles. Although in this proteomic study, we did notidentify several enzymes such as hydroxymethylglutaryl coenzyme Areductase (HMGR), IPP isomerase (IPI), GPS (only identified from CS),and GGPPS from WS, most of the key proteins including FADS, GGPS,GGTF, CPT, REF, SRPP and HRBP are positively identified from WS(Fig. 6; Table S5). Accumulation patterns of these proteins revealed thatmost NRB enzymes are high abundance in WS (Figs. 5 and 6; Table S2),indicating WS proteins might play more important roles in NRB.

4. Discussion

4.1. Combined 2-DE and shotgun analyses revealed the first visualization ofWS proteome

Rubber latex contains various macromolecules, including proteins,starches, sugars, tannins, resins, and alkaloids [8,31]. Rubber moleculeis synthesized on RP, and many proteins involved in this process havebeen recently identified by omics-based technologies [6,27,28]. Now,proteomics is becoming a powerful tool for discovery and character-ization of proteins from different latex fractions in H. brasiliensis, suchas total latex [8,26,27,32–34], RPs [8,13,18,25], C-serum [26,35, 36]and lutoids [11,28]. It is noteworthy that, by using a phenol-basedmethod for proteins extraction from RPs [26] and searching the com-prehensive proteomic data from the high quality of rubber tree genomedatabase [29], we have identified far more number of proteins from RPsand WS than the reported results [13,18,25]. Till now, no proteomics isfocused on the RP washed solutions. In a previous study, we found WSstill contain many NRB-related proteins [16]. Furthermore, to performWS proteomics, we improved a protein extraction method from theoriginal one [26], and proved that this new REP method is suitable forWS [16].

In this work, we generated the first high-resolution 2-DE profiles ofWS and identified many protein species, including 19 REF and 13 SRPPmembers (Fig. 1; Table S1), which are similar with our previous pro-teomic observation that at least 17.2% of proteins belonged to REF inthe rubber latex [8]. REF is localized on both large and small RPs andall laticifer layers, but SRPP is mainly localized in small RPs and lati-cifer layers in the conducting phloem [37]. Protein structure analysisrevealed that REF contains a β-sheet, whereas SRPP mainly folds as ahelical protein [22]. Both REF and SRPP are all highly hydrophobicproteins, and they exhibit differences in their affinity towards themembrane, REF penetrates deeply into the monolayer, and SRPP onlybinds to the lipid [24]. Our recent results also demonstrated that REFsubfamily member has a conserved short C-terminal and a changeablelength in the N-terminal beyond REF domain, while SRPP has a short N-terminal and a changeable length of the C-terminal beyond REF domain[34].

Here, we identified more than 2500 proteins from WS at least oneshotgun experiment (Table S2). To our best knowledge, it is the firstvisualization on WS protein profiles. By analyzing the proteomic data,we believe that, during washing, some proteins that loosely bound withRP might be washed away, and many CS proteins can also be includedin WS.

Fig. 6. Schematic diagrams of WS proteins involved in natural rubber bio-synthesis.The main WS proteins involved in NRB from the proteomics of WS, CS and RPare highlighted. Proteins marked with red color were positively identified fromWS. The changed ratios of proteins in different latex fractions (from left to right,WS/RP, CS/RP and CS/WS) were determined by Label free. The bar for proteinchanged ratios is as same as that in Fig. 5. The abbreviations are: ACAT, Acetyl-CoA C-acetyltransferase; ACOC, Acetyl-CoA carboxylase; APX, ascorbate per-oxidase; CDPK, calcium-dependent protein kinase; CPT, cis-prenyltransferase;CYAS, beta-cyanoalanine synthase; DMAPP, dimethylallyl diphosphate; DPMD,diphosphomevalonate decarboxylase; DXP, 1-deoxy-D-xylulose 5-phosphate;FADS, farnesyl pyrophosphate/diphosphate synthase; FPP, farnesyl dipho-sphate; GA3P, glyceraldehyde 3-phosphate; GGPP, geranylgeranyl diphosphate;GGPS, geranylgeranyl diphosphate synthase; GGPPS, geranylgeranyl pyropho-sphate synthase; GGTF, geranylgeranyl transferase; GLUD, latex cyanogenicbeta glucosidase; GPP, geranyl diphosphate; GPPS, geranyl pyrophosphatesynthase; GSGT, galactinol-sucrose galactosyltransferase; HMGR, hydro-xymethylglutaryl coenzyme A reductase; HMGS, hydroxymethylglutaryl coen-zyme A synthase; HRBP, HRT1-REF bridging protein, a Nogo-B receptor;HRTase, Hevea rubber transferase; IPP, isopentenyl pyrophosphate; MEP, me-thylerythritol 4-phosphate; MEVD, mevalonate disphosphate decarboxylase;MEVK, mevalonate kinase; MVA, menvalonate; PPi, inorganic diphosphate;REF, rubber elongation factor; SOD, superoxide dismutase; SRPP, small rubberparticle protein.


61

4.2. Most enzymes in WS are key proteins involved in NRB

NRB is a typical isoprenoid metabolism in laticifer [38,39]. In Hevearubber tree, cis-1,4-polyisoprene is generated by a living carbocationicpolymerization process that includes three key steps named initiation,elongation and termination [12]. The initiation process is through aMVA pathway [2,12,40,41], which begins with the synthesis of IPP. Inearly steps, HMGS and HMGR activate the supply of substrates. CPT (orHRT, Hevea rubber transferase) uses pyrophosphates to initiate rubbermolecule as well as IPP to form the polymer, thus is a key enzyme inrubber biosynthesis [20,21]. During elongation process, REF and SRPPare key enzymes. In isoprenoid biosynthesis, FADS or FPPS catalyzesthe last common substrate, and thus it is also a key enzyme for rubberbiosynthesis [2,8,14,40]. It was reported that MVA pathway is in cy-tosolic [42], while MEP pathway is mainly occurred in plastid [38,43].Recently, a 13C-labelled study on Hevea seedlings suggested that MEPpathway contributes IPP mainly for carotenoid biosynthesis, but not forrubber biosynthesis [40], indicating enzymes involved in MVA pathwayare more important for NRB. Our proteomic data in rubber latex [8] andWS (Table S2) also demonstrated that the identified proteins are mainlybelonged to MVA pathway. Among them, ACAT represents the firstenzyme in MVA pathway; it catalyzes the Claisen-type condensation oftwo units of acetyl-CoA to yield acetoacetyl-CoA [44]. Here, twomembers of acetyl-CoA acetyltransferase (ACAT1 and ACAT2) wereidentified from WS, and only four amino acids are different betweenthem (Table S2). Functional complementation showed that ACAT1 hada higher functional affinity in yeast than ACAT2 [40]. The second en-zyme is HMGS, which catalyzes aldol-type condensation of acetoacetyl-CoA and acetyl-CoA to produce HMG-CoA [21,45]. In our data, onlyone HMGS (scaffold0082_781745.mRNA1) was found in WS, and it islocated as a cytosolic protein. HMG-CoA can be converted to MVA byHMG-CoA reductase (HMGR). But, in this proteomic study, we didn'tdetect HMGR in WS, probably due to HMGR is low abundance and it ismainly in lutoids [46].

The first stage in the conversion of mevalonate into isopentyl pyr-ophosphate is the formation of 5-phosphomevalonate catalysed bymevalonate kinase [47]. In Hevea, this enzyme is coded by a singlegene, and has a higher accumulation in latex [40]. Here, we onlyidentified one mevalonate kinase in CS and WS. Phosphomevalonatekinase catalyzes mevalonate-5-phosphoate to form mevalonate-5-pyr-ophosphate (MVAPP), it is acid, and needs the presence of a thiol tomaintain its activity [47]. This enzyme was also coded by a single genein Hevea [40], but we identified two phosphomevalonate kinases in thisproteomic data, one from RP, CS and WS, the other only from WS(Table S2).

Mevalonate-5-pyrophosphate decarboxylase (MVD) catalyzesMVAPP to take off one molecule of CO2 and one molecule H2O to formIPP [41]. MVD has a good stability under acid and heat conditions, butit is easily inhibited by IPP and ADP [48]. In Hevea, this enzyme wasconsidered to be coded by a single gene [40], but in this study, weobtained two MVD members in WS, both containing 415 amino acids.Isopentenyl diphosphate isomerase (IPI) can convert IPP to formDMAPP, we didn't found this enzyme in WS, but only identified it fromCS (Table S2), indicating it is mainly located in cytoplasm. Farnesylpyrophosphate/diphosphate synthase (FADS) plays key roles in rubberbiosynthesis, and it catalyzes the geranyl pyrophosphate formation offarnesyl pyrophosphate, and then helps IPP incorporation into poly-isoprene [49]. In this study, we found two FADS members in WS.Western blotting showed the similarly changed patterns of FADS withshotgun and label free analyses. We didn't identify the other two rubberelongation related enzymes, named GGPS and GGPPS, from WS, whichis probably due to they are mainly located in plastid [50].

HRT is closely combined with RP, and it is important for forminglong chain rubber and determining the size of rubber molecule[20,21,48]. Two members of HRT family (HRT1 and HRT2) were re-ported in the rubber tree [51]. In vitro analysis showed that only HRT2

has rubber transferase activity for extending prenyl chain [51]. Here,we found two HRT2 members in WS (Table S2). Although REF andSRPP are not named as enzyme, they play a key role in the elongationprocess of rubber biosynthesis [23,52]. In Hevea, there are 18 REF/SRPP family members, and these genes exhibit distinct expressionpatterns in different Hevea fractions, most of them are located in asingle 205-kb genome site Scaffold1222 [29]. In this study, three REFs,named REF138, REF175 and REF222, as well as four SRPPs, namedSRPP117, SRPP155, SRPP204 and SRPP216, were identified from WS.They were all located in the Scaffold1222 in the rubber tree genome(Table S2). These results are consistent with the recently reports thatSRPP is more important than REF for NRB [22,24,37,53].

4.3. Comprehensive proteomics of WS revealed many specific proteins andprotein complexes in WS for NRB

In this proteomics study, we confirmed that WS really containsthousands of proteins. Among the 246 WS-specific proteins, 3-hydro-xyacyl-acyl-carrier-protein dehydratase, also named stearoyl-ACP (acyl-carrier-protein) desaturase, is involved in lipid metabolism (Fig. 3;Table S2). It was reported that acyl carrier proteins can bind to fattyacids through a thioester bond, and thus can generate the acyl-ACPlipoproteins [54]. This process is required for fatty acid chain elonga-tion in plants [55]. In most plants, the primary products of fatty acidsynthesis are palmitoyl-acyl-carrier protein (16,0-ACP) and stearoyl-ACP (18,0-ACP). The soluble plastid enzyme, stearoyl-ACP desaturase,introduces the first double bond into stearoyl-ACP between carbons 9and 10 to produce oleoyl-ACP [56].

Several protein complexes were also observed in WS-specific pro-teins. Among them, AP-4 complex is involved in recognition and sortingof cargo proteins to endosomal-lysosomal system [57]. Here, we iden-tified the AP-4 complex subunit epsilon as a specific protein in WS,whereas another subunit was simultaneously existed in WS, CS and RP(Table S2). Two conserved oligomeric Golgi complex subunits wereidentified from WS (Table S2), which are capable of intracellulartransport and glycoprotein modification [58]. COP9 signalosomecomplex possesses kinase activity that can phosphorylate regulators,and this COP complex regulates the largest family of E3 ubiquitin li-gases [59]. We observed two COP9 signalosome complex subunits asWS-specific proteins, although they were only identified by shotgunanalysis (Table S2).

The exocyst complex serves to direct vesicles after the Golgi com-plex to specific locations on the plasma membrane [60]. We identifiedthe trafficking protein particle complex subunits 9 and 10 as WS-spe-cific proteins; they are part of a large multi-subunit complex for tar-geting endoplasmic reticulum-to-Golgi transport vesicles [19]. We alsoidentified several exocyst complexs and transportation related proteins.They involve in exocytosis and transportation in rubber latex [61,62],indicating that RP might origin from endoplasmic reticulum and WSproteins might be important in exocytosis.

Many enzymes in ubiquitin-proteasome and protease systems werealso found in WS. They are 26S protease regulatory, E3 ubiquitin-pro-tein ligase, ubiquitin-40S ribosomal protein, ubiquitin carboxyl-term-inal hydrolase, cullin-associated protein, etc. It is known that cullins area family of hydrophobic proteins that can provide scaffolds for ubi-quitin ligase [63]. Ubiquitin-proteasome system is crucial for in-tracellular protein degradation, which participates in the degradationof more than 80% proteins in plant cells [64].

There are many WS-specific proteins for protein translation andbiosynthesis. Among them, four members were identified as eukaryotictranslation initiation factor, three members as serine/threonine-proteinphosphatase, and two members as serine/threonine-protein kinase(Table S2). During the washing procession of RP, some subunits of theseproteins may be fully washed out from RP. These results suggested thatWS from RP has an ubiquitin-proteasome mediated proteolysis systemfor protein degradation, and these proteins might play important roles


62

for the development of small RP to big RP [6,23], as well as the pro-cession of NRB [25,65].

5. Conclusions

In this study, we obtained the first visualization proteome profiles ofWS. Combined 2-DE and shotgun analyses, we identified thousands ofunique proteins from WS, CS and RP. Among them, 246 are WS-specificproteins; they are mainly involved in ribosomal metabolism, protea-some system, vacuolar protein sorting and endocytosis. Most key en-zymes involved in NRB were identified from WS. Our proteomic dataalso revealed that protein complexes may actually play important rolesfor NRB. Future works should pay more attentions on determination thedetailed roles of WS-specific proteins.

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jprot.2018.05.002.

Conflict of interest

The authors declare that they have no conflict of financial interest.

Acknowledgments

This research was supported by the National Natural ScienceFoundation Grants in China (No. 31570301, 31500557, 31400523), theCentral Public-interest Scientific InstitutionBasal Research Fund forChinese Academy of Tropical Agricultural Sciences (No.1630052017006, 1630052017007) and the Scientific ResearchInnovation Team Program of Hainan Province (2018CXTD341). We aregrateful to Prof. Weimin Tian in the Rubber Research Institute for hishelpful suggestions.

References

[1] J.M. Hagel, E.C. Yeung, P.J. Facchini, Got milk? The secret life of laticifers, TrendsPlant Sci. 13 (2008) 1360–1385.

[2] J.H. Zhu, Z.L. Zhang, Ethylene stimulation of latex production in Hevea brasiliensis,Plant Signal. Behav. 4 (2009) 1072–1074.

[3] Y. Zhang, J. Leclercq, P. Montoro, Reactive oxygen species in Hevea brasiliensis latexand relevance to Tapping Panel Dryness, Tree Physiol. 37 (2017) 261–269.

[4] F. Cataldo, Guayule rubber: a new possible world scenario for the production ofnatural rubber? Prog. Rubber Plast. Technol. 16 (2000) 31–59.

[5] K. Cornish, Similarities and differences in rubber biochemistry among plant species,Phytochemistry 57 (2001) 1123–1134.

[6] W.K. Cho, Y. Jo, H. Chu, S.H. Park, K.H. Kim, Integration of latex protein sequencedata provides comprehensive functional overview of latex proteins, Mol. Biol. Rep.41 (2014) 1469–1481.

[7] W.M. Tian, S.G. Yang, M.J. Shi, S.X. Zhang, J.L. Wu, Mechanical wounding-inducedlaticifer differentiation in rubber tree: an indicative role of dehydration, hydrogenperoxide, and jasmonates, J. Plant Physiol. 182 (2015) 95–103.

[8] X.C. Wang, D. Wang, Y. Sun, Q. Yang, L.L. Chang, L.M. Wang, X.R. Meng,Q.X. Huang, X. Jin, Z. Tong, Comprehensive proteomics analysis of laticifer latexreveals new insights into ethylene stimulation of natural rubber production, Sci.Rep. 5 (2015) 13778.

[9] K. Tungngoen, P. Kongsawadworakul, U. Viboonjun, M. Katsuhara, N. Brunel,S. Sakr, J. Narangajavana, H. Chrestin, Involvement of HbPIP2;1 and HbTIP1;1aquaporins in ethylene stimulation of latex yield through regulation of water ex-changes between inner liber and latex cells in Hevea brasiliensis, Plant Physiol. 151(2009) 843–856.

[10] K. Adiwilaga, A. Kush, Cloning and characterization of cDNA encoding farnesyldiphosphate synthase from rubber tree (Hevea brasiliensis), Plant Mol. Biol. 30(1996) 935–946.

[11] X.C. Wang, M. Shi, D. Wang, Y. Chen, F. Cai, S. Zhang, L. Wang, Z. Tong, W.M. Tian,Comparative proteomics of primary and secondary lutoids reveals that chitinaseand glucanase play a crucial combined role in rubber particle aggregation in Heveabrasiliensis, J. Proteome Res. 12 (2013) 5146–5159.

[12] J.E. Puskas, E. Gautriaud, A. Deffieux, J.P. Kennedy, Natural rubber biosynthesis—aliving carbocationic polymerization? Prog. Polym. Sci. 31 (2006) 533–548.

[13] S. Yamashita, H. Yamaguchi, T. Waki, Y. Aoki, M. Mizuno, F. Yanbe, T. Ishii,A. Funaki, Y. Tozawa, Y. Miyagi-Inoue, et al., Identification and reconstitution ofthe rubber biosynthetic machinery on rubber particles from Hevea brasiliensis, Elife5 (2016) 19022.

[14] K. Cornish, D.F. Wood, J.J. Windle, Rubber particles from four different species,examined by transmission electron microscopy and electron-paramagnetic-re-sonance spin labeling, are found to consist of a homogeneous rubber core enclosed

by a contiguous, monolayer biomembrane, Planta 210 (1999) 85–96.[15] R. Wititsuwannakul, P. Pasitkul, P. Jewtragoon, D. Wititsuwannakul, Hevea latex

lectin binding protein in C-serum as an anti-latex coagulating factor and its role in aproposed new model for latex coagulation, Phytochemistry 69 (2008) 656–662.

[16] D. Wang, Y. Sun, Z. Tong, Q. Yang, L.L. Chang, X.R. Meng, L.M. Wang, W.M. Tian,X.C. Wang, A protein extraction method for low protein concentration solutionscompatible with the proteomic analysis of rubber particles, Electrophoresis 37(2016) 2930–2939.

[17] A.P. Singh, S.G. Wi, G.C. Chung, Y.S. Kim, H. Kang, The micromorphology andprotein characterization of rubber particles in Ficus carica, Ficus benghalensis andHevea brasiliensis, J. Exp. Bot. 54 (2003) 985–992.

[18] Q.L. Xiang, K.C. Xia, L.J. Dai, G.J. Kang, Y. Li, Proteome analysis of the large andthe small rubber particles of Hevea brasiliensis using 2D-DIGE, Plant Physiol.Biochem. 60 (2012) 207–213.

[19] M.J. Chrispeels, E.T. Herman, Endoplasmic reticulum-derived compartments func-tion in storage and as mediators of vacuolar remodeling via a new type of organelle,precursor protease vesicles, Plant Physiol. 123 (2000) 1227–1234.

[20] J. Epping, N.V. Deenen, E. Niephaus, A. Stolze, J. Fricke, C. Huber, W. Eisenreich,R.M. Twyman, D. Prufer, C.S. Gronover, A rubber transferase activator is necessaryfor natural rubber biosynthesis in dandelion, Nature Plants 1 (2015) 15048.

[21] K.A. Grabinska, E.J. Park, W.C. Sessa, cis-Prenyltransferase: new insights into pro-tein glycosylation, rubber synthesis, and human diseases, J. Biol. Chem. 291 (2016)18582–18590.

[22] K. Berthelot, S. Lecomte, Y. Estevez, B. Coulary-Salin, A. Bentaleb, C. Cullin,A. Deffieux, F. Peruch, Rubber Elongation Factor (REF), a major allergen compo-nent in Hevea brasiliensis latex has amyloid properties, PLoS One 7 (2012) 48065.

[23] L. Dai, Z. Nie, G. Kang, Y. Li, R. Zeng, Identification and subcellular localizationanalysis of two rubber elongation factor isoforms on Hevea brasiliensis rubber par-ticles, Plant Physiol. Biochem. 111 (2017) 97–106.

[24] K. Berthelot, S. Lecomte, Y. Estevez, F. Peruch, Hevea brasiliensis REF (Hev b 1) andSRPP (Hev b 3): an overview on rubber particle proteins, Biochimie 106 (2014) 1–9.

[25] L. Dai, G. Kang, Y. Li, Z. Nie, C. Duan, R. Zeng, In-depth proteome analysis of therubber particle of Hevea brasiliensis (para rubber tree), Plant Mol. Biol. 82 (2013)155–168.

[26] X.C. Wang, M.J. Shi, X.L. Lu, R.F. Ma, C.G. Wu, A.P. Guo, M. Peng, W.M. Tian, Amethod for protein extraction from different subcellular fractions of laticifer latex inHevea brasiliensis compatible with 2-DE and MS, Proteome Sci. 8 (2010) 35.

[27] L. Dai, G. Kang, Z. Nie, Y. Li, R. Zeng, Comparative proteomic analysis of latex fromHevea brasiliensis treated with Ethrel and methyl jasmonate using iTRAQ-coupledtwo-dimensional LC-MS/MS, J. Proteome 132 (2016) 167–175.

[28] M.H. Habib, G.C. Yuen, F. Othman, N.N. Zainudin, A.A. Latiff, M.N. Ismail,Proteomics analysis of latex from Hevea brasiliensis (clone RRIM 600), Biochem. CellBiol. 95 (2017) 232–242.

[29] C. Tang, M. Yang, Y. Fang, Y. Luo, S. Gao, X. Xiao, Z. An, B. Zhou, B. Zhang, X. Tan,et al., The rubber tree genome reveals new insights into rubber production andspecies adaptation, Nature Plants 2 (2016) 16073.

[30] Z. Shen, P. Li, R.J. Ni, R. Mark, C.P. Yang, G.F. Liu, W. Ma, G.J. Liu, L. Ma, S.J. Li,et al., Label-free quantitative proteomics analysis of etiolated maize seedling leavesduring greening, Mol. Cell. Proteomics 8 (2009) 2443–2460.

[31] J.R. Hunter, Reconsidering the functions of latex, Trees 9 (1994) 1–5.[32] A. D'Amato, A. Bachi, E. Fasoli, E. Boschetti, G. Peltre, H. Senechal, J.P. Sutra,

A. Citterio, P.G. Righetti, In-depth exploration of Hevea brasiliensis latex proteomeand “hidden allergens” via combinatorial peptide ligand libraries, J. Proteome 73(2010) 1368–1380.

[33] P. Rojruthai, J.T. Sakdapipanich, S. Takahashi, L. Hyegin, M. Noike, T. Koyama,Y. Tanaka, In vitro synthesis of high molecular weight rubber by Hevea small rubberparticles, J. Biosci. Bioeng. 109 (2010) 107–114.

[34] Z. Tong, D. Wang, Y. Sun, Q. Yang, X.R. Meng, L.M. Wang, W.Q. Feng, L. Li,E.S. Wurtele, X.C. Wang, Comparative proteomics of rubber latex revealed multipleprotein species of REF/SRPP family respond diversely to ethylene stimulationamong different rubber tree clones, Int. J. Mol. Sci. 18 (2017) 958.

[35] H.L. Li, D. Guo, F.Y. Lan, W.M. Tian, S.Q. Peng, Protein differential expression inthe latex from Hevea brasiliensis between self-rooting juvenile clones and donorclones, Acta Physiol. Plant. 33 (2011) 1853–1859.

[36] P. Havanapan, P. Bourchookarn, A.J. Ketterman, C. Krittanai, Comparative pro-teome analysis of rubber latex serum from pathogenic fungi tolerant and susceptiblerubber tree (Hevea brasiliensis), J. Proteome 131 (2016) 82–92.

[37] T. Sando, T. Hayashi, T. Takeda, Y. Akiyama, Y. Nakazawa, E. Fukusaki,A. Kobayashi, Histochemical study of detailed laticifer structure and rubber bio-synthesis-related protein localization in Hevea brasiliensis using spectral confocallaser scanning microscopy, Planta 230 (2009) 215–225.

[38] E. Swiezewska, W. Danikiewicz, Polyisoprenoids: structure, biosynthesis andfunction, Prog. Lipid Res. 44 (2005) 235–258.

[39] A.Y. Rahman, A.O. Usharraj, B.B. Misra, G.P. Thottathil, K. Jayasekaran, Y. Feng,S. Hou, S.Y. Ong, F.L. Ng, L.S. Lee, et al., Draft genome sequence of the rubber treeHevea brasiliensis, BMC Genomics 14 (2013) 75.

[40] T. Sando, C. Takaoka, Y. Mukai, A. Yamashita, M. Hattori, N. Ogasawara,E. Fukusaki, A. Kobayashi, Cloning and characterization of mevalonate pathwaygenes in natural rubber porducing plant, Hevea brasiliensis, Biosci. Biotechnol.Biochem. 72 (2008) 2049–2060.

[41] C.T. Wu, Y. Li, Z.Y. Nie, L.J. Dai, G.J. Kang, R.Z. Zeng, Molecular cloning and ex-pression analysis of the mevalonate diphosphate decarboxylase gene from the latexof Hevea brasiliensis, Tree Genet. Genomes 13 (2017) 22.

[42] R. Kekwick, L. Chrestin, GO: The formation of isoprenoids in Hevea latex, in:J. d'Auzac, J.L. Jacob (Eds.), Physiology of Rubber Tree Latex, CRC Press, BocaRaton, 1989, pp. 145–164.


63

https://doi.org/10.1016/j.jprot.2018.05.002https://doi.org/10.1016/j.jprot.2018.05.002http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0005http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0005http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0010http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0010http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0015http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0015http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0020http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0020http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0025http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0025http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0030http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0030http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0030http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0035http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0035http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0035http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0040http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0040http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0040http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0040http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0045http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0045http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0045http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0045http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0045http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0050http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0050http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0050http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0055http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0055http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0055http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0055http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0060http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0060http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0065http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0065http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0065http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0065http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0070http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0070http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0070http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0070http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0075http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0075http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0075http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0080http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0080http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0080http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0080http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0085http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0085http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0085http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0090http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0090http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0090http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0095http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0095http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0095http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0100http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0100http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0100http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0105http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0105http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0105http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0110http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0110http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0110http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0115http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0115http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0115http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0120http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0120http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0125http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0125http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0125http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0130http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0130http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0130http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0135http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0135http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0135http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0140http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0140http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0140http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0145http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0145http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0145http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0150http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0150http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0150http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0155http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0160http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0160http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0160http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0160http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0165http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0165http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0165http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0170http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0170http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0170http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0170http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0175http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0175http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0175http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0180http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0180http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0180http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0185http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0185http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0185http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0185http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0190http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0190http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0195http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0195http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0195http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0200http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0200http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0200http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0200http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0205http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0205http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0205http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0210http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0210http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0210

[43] J.H. Ko, K.S. Chow, K.H. Han, Transcriptome analysis reveals novel features of themolecular events occurring in the laticifers of Hevea brasiliensis (para rubber tree),Plant Mol. Biol. 53 (2003) 479–492.

[44] P. Liao, H. Wang, A. Hemmerlin, D.A. Nagegowda, T.J. Bach, M.F. Wang,M.L. Chye, Past achievements, current status and future perspectives of studies on3-hydroxy-3-methylglutaryl-CoA synthase (HMGS) in the mevalonate (MVA)pathway, Plant Cell Rep. 33 (2014) 1005–1022.

[45] N. Sirinupong, P. Suwanmanee, R.F. Doolittle, W. Suvachitanont, Molecular cloningof a new cDNA and expression of 3-hydroxy-3-methylglutaryl-CoA synthase genefrom Hevea brasiliensis, Planta 221 (2005) 502–512.

[46] H. Schaller, B. Crausem, P. Benveniste, M.L. Chye, C.T. Tan, Y.H. Song, N.H. Chua,Expression of the Hevea brasiliensis (H.B.K.) Müll. Arg. 3-Hydroxy-3-Methylglutaryl-Coenzyme A Reductase 1 in tobacco results in sterol overproduction, Plant Physiol.109 (1995) 761–770.

[47] B.Y. Williamson, R.G. Kekwick, The enzymes forming isopentenyl pyrophosphatefrom 5-phosphomevalonate (mevalonate 5-phosphate) in the latex of Hevea brasi-liensis, Biochem. J. 124 (1971) 407–417.

[48] K. Cornish, The separate roles of plant cis and trans prenyltransferases in cis-1,4-polyisoprene biosynthesis, Eur. J. Biochem. 218 (1993) 267–271.

[49] T.K. Uthup, T. Saha, M. Ravindran, K. Bini, Impact of an intragenic retrotransposonon the structural integrity and evolution of a major isoprenoid biosynthesispathway gene in Hevea brasiliensis, Plant Physiol. Biochem. 73 (2013) 176–188.

[50] K. Okada, T. Sato, T. Nakagawa, M. Kawamukai, Y. Kamiya, Five geranylgeranyldiphosphate synthases expressed in different organs are localized into three sub-cellular compartments in Arabidopsis, Plant Physiol. 122 (2000) 1045–1056.

[51] K. Ogura, T. Koyama, Enzymatic aspects of isoprenoid chain elongation, Chem. Rev.98 (1998) 1263–1276.

[52] E. Goyvaerts, M.S. Dennis, D.R. Light, N.H. Chua, Cloning and sequencing of thecDNA encoding the rubber elongation factor of Hevea brasiliensis, Plant Physiol. 97(1991) 317–321.

[53] K. Wadeesirisak, S. Castano, K. Berthelot, L. Vaysse, F. Bonfils, F. Peruch,K. Rattanaporn, S. Liengprayoon, S. Lecomte, C. Bottier, Rubber particle proteinsREF1 and SRPP1 interact differently with native lipids extracted from Hevea bra-siliensis latex, Biochim. Biophys. Acta 1859 (2017) 201–210.

[54] A.C. Price, K.H. Choi, R.J. Health, Z. Li, S.W. White, C.O. Rock, Inhibition of b-

ketoacyl-acyl carrier protein synthases by thiolactomycin and cerulenin. Structureand mechanism, J. Biol. Chem. 276 (2001) 6551–6559.

[55] J.A. Aznar-Moreno, M. Venegas-Caleron, E. Martinez-Force, R. Garces, J.J. Salas,Acyl carrier proteins from sunflower (Helianthus annuus L.) seeds and their influenceon FatA and FatB acyl-ACP thioesterase activities, Planta 244 (2016) 479–490.

[56] J. Shanklin, C. Somervilie, Stearoyl-acyl-carrier-protein desaturase from higherplants is structurally unrelated to the animal and fungal homologs, Proc. Natl. Acad.Sci. U. S. A. 88 (1991) 2510–2514.

[57] K. Fuji, M. Shirakawa, Y. Shimono, T. Kunieda, Y. Fukao, Y. Koumoto,H. Takahashi, I. Hara-Nishimura, T. Shimada, The adaptor complex AP-4 regulatesvacuolar protein sorting at trans-Golgi network by interacting with VACUOLARSORTING RECEPTOR 1, Plant Physiol. 170 (2016) 211–219.

[58] X. Wu, R.A. Steet, O. Bohorov, J. Bakker, J. Newell, M. Krieger, L. Spaapen,S. Kornfeld, H.H. Freeze, Mutation of the COG complex subunit gene COG7 causes alethal congenital disorder, Nat. Med. 10 (2004) 518–523.

[59] A.K. Hoareau, V. Bochard, S. Réty, P. Jalinot, Association of the mammalian proto-oncoprotein Int-6 with the three protein complexes eIF3, COP9 signalosome and26S proteasome, FEBS Lett. 527 (2002) 15–21.

[60] M. Fendrych, L. Synek, T. Pecenkova, E.J. Drdova, J. Sekeres, R. de Rycke,M.K. Nowack, V. Zarsky, Visualization of the exocyst complex dynamics at theplasma membrane of Arabidopsis thaliana, Mol. Biol. Cell 24 (2013) 510–520.

[61] S.C. Loh, G.P. Thottathil, A.S. Othman, Identification of differentially expressedgenes and signaling pathways in bark of Hevea brasiliensis seedlings associated withsecondary laticifer differentiation using gene expression microarray, Plant Physiol.Biochem. 107 (2016) 45–55.

[62] Z.Y. Nie, G.J. Kang, Y. Li, L.J. Dai, R.Z. Zeng, Whole-transcriptome survey of theputative ATP-binding cassette (ABC) transporter family genes in the latex-produ-cing laticifers of Hevea brasiliensis, PLoS One 10 (2015) 116857.

[63] C. Schwechheimer, L.I. Calderon-Villalobos, Cullin-containing E3 ubiquitin ligasesin plant development, Curr. Opin. Plant Biol. 7 (2004) 677–686.

[64] J. Moon, G. Parry, M. Estelle, The ubiquitin-proteasome pathway and plant de-velopment, Plant Cell 16 (2004) 3181–3195.

[65] R. Nawrot, R. Lippmann, A. Matros, O. Musidlak, G. Nowicki, H.P. Mock, Proteomiccomparison of Chelidonium majus L. latex in different phases of plant development,Plant Physiol. Biochem. 112 (2017) 312–325.


64

http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0215http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0215http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0215http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0220http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0220http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0220http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0220http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0225http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0225http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0225http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0230http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0230http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0230http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0230http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0235http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0235http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0235http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0240http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0240http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0245http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0245http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0245http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0250http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0250http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0250http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0255http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0255http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0260http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0260http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0260http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0265http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0265http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0265http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0265http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0270http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0270http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0270http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0275http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0275http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0275http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0280http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0280http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0280http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0285http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0285http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0285http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0285http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0290http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0290http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0290http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0295http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0295http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0295http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0300http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0300http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0300http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0305http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0305http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0305http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0305http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0310http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0310http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0310http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0315http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0315http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0320http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0320http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0325http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0325http://refhub.elsevier.com/S1874-3919(18)30196-9/rf0325

Subcellular proteome profiles of different latex fractions revealed washed solutions from rubber particles contain crucial enzymes for natural rubber biosynthesisIntroductionMaterials and methodsPlant materialCollection of rubber particles, C-serum and washed solutions from latexProtein extraction and two-dimensional gel electrophoresisProtein identification via mass spectrometryShotgun and label free quantitative proteomic analysis of WS, CS and RPWestern blotting analysis of the typical identified proteins in different latex fractionsProtein classification and STRING analysis

ResultsMALDI TOF/TOF MS identification of the main proteins in 2-DE gel of WSFunctional analysis of the high abundance WS proteins from 2-DE gelsIdentification of proteins from WS, CS and RP by high-throughput shotgun analysisGene Ontology classification and protein-protein interaction of WS-specific proteinsDetermination of protein accumulation patterns in WS, CS and RPFunctional analysis of all WS proteins and comparison of the accumulation patterns o

Journal of Proteomics - 首页 - 植物抗逆整合生物学实验室€¦ · Recently, proteomics...

Documents

Transcript of Journal of Proteomics - 首页 - 植物抗逆整合生物学实验室€¦ · Recently, proteomics...