Integrated Phloem Sap mRNA and Protein Expression Analysis Reveals Phytoplasma-infection Responses in Mulberry*

To gain insight into the response of mulberry to phytoplasma-infection, the expression profiles of mRNAs and proteins in mulberry phloem sap were examined. A total of 955 unigenes and 136 proteins were found to be differentially expressed between the healthy and infected phloem sap. These differentially expressed mRNAs and proteins are involved in signaling, hormone metabolism, stress responses, etc. Interestingly, we found that both the mRNA and protein levels of the major latex protein-like 329 (MuMLPL329) gene were increased in the infected phloem saps. Expression of the MuMLPL329 gene was induced by pathogen inoculation and was responsive to jasmonic acid. Ectopic expression of MuMLPL329 in Arabidopsis enhances transgenic plant resistance to Botrytis cinerea, Pseudomonas syringae pv tomato DC3000 (Pst. DC3000) and phytoplasma. Further analysis revealed that MuMLPL329 can enhance the expression of some defense genes and might be involved in altering flavonoid content resulting in increased resistance of plants to pathogen infection. Finally, the roles of the differentially expressed mRNAs and proteins and the potential molecular mechanisms of their changes were discussed. It was likely that the phytoplasma-responsive mRNAs and proteins in the phloem saps were involved in multiple pathways of mulberry responses to phytoplasma-infection, and their changes may be partially responsible for some symptoms in the phytoplasma infected plants.

To gain insight into the response of mulberry to phytoplasma-infection, the expression profiles of mRNAs and proteins in mulberry phloem sap were examined. A total of 955 unigenes and 136 proteins were found to be differentially expressed between the healthy and infected phloem sap. These differentially expressed mRNAs and proteins are involved in signaling, hormone metabolism, stress responses, etc. Interestingly, we found that both the mRNA and protein levels of the major latex protein-like 329 (MuMLPL329) gene were increased in the infected phloem saps. Expression of the MuMLPL329 gene was induced by pathogen inoculation and was responsive to jasmonic acid. Ectopic expression of MuMLPL329 in Arabidopsis enhances transgenic plant resistance to Botrytis cinerea, Pseudomonas syringae pv tomato DC3000 (Pst. DC3000) and phytoplasma. Further analysis revealed that MuMLPL329 can enhance the expression of some defense genes and might be involved in altering flavonoid content resulting in increased resistance of plants to pathogen infection. Finally, the roles of the differentially expressed mRNAs and proteins and the potential molecular mechanisms of their changes were discussed. It was likely that the phytoplasma-responsive mRNAs and proteins in the phloem saps were involved in multiple pathways of mulberry responses to phytoplasma-infection, and their changes may be partially responsible for some symptoms in the phytoplasma infected plants. Molecular  Information exchange between cells and tissues to coordinate responses to environmental changes, development, and pathogen defense is required for plant viability (1), and higher plants have evolved different strategies to allow efficient intercellular exchange of information (2). The phloem network represents a well-developed long-distance translocation system for both nutrients and systemic signals (3). Despite the importance of phloem, the mechanisms involved in phloem transport and signaling remain unknown, and there is little information reported about the way in which its function is regulated (4 -5). It has been established that there are many mRNA transcripts in the phloem translocation stream, and plants can utilize the mobile RNAs to regulate development and responses to biotic and abiotic stresses at the wholeplant level (6 -11). Compared with the number of mRNA transcripts identified, a limited number of proteins have been identified because phloem sap sample is difficult to obtain in sufficient quantities in most plant species (12)(13). Interestingly, a large proportion of RNAs and proteins identified in plant phloem sap were predicted to be associated with stress and defense responses, although their exact physiological functions were unclear (14 -16). However, a comprehensive understanding of the mechanisms involved in the translocation of RNAs and proteins and the underlying cellular processes taking place within the phloem sap is still lacking. Therefore, it is important to identify and characterize a more comprehensive set of phloem sap RNAs and proteins that may be candidates that have significance regarding functional phloem and whole plant physiology.
Mulberry yellow dwarf disease is one of the most devastating diseases of mulberry (Morus spp.) caused by phytoplasma (17). Because phytoplasmas are difficult to culture in vitro, the underlying molecular mechanisms of their pathogenicity are still poorly understood (18). In the process of plant-pathogen interaction, plants not only can perceive the pathogen invasion and initiate local defense responses but also can transmit the signal of pathogen infection over long distances and activate a multicomponent response at the whole-plant level (19). Because phytoplasmas are strictly confined to the phloem compartment and contact with the sieve element (SE) contents directly (20), the influence of phytoplasma infection on phloem sap protein and RNA composition is easily conceivable. Identification and characterization of the responsive RNAs and proteins present in the phloem sap are a prerequisite to understanding the molecular mechanisms involved in the disease symptom development. In recent years, genomic and proteomic strategies have been successfully used to analyze plant-phytoplasma interactions, and many genes and proteins regulated following infection of the phytoplasma have been recognized in several plant species (21)(22)(23)(24)(25). However, as far as we know, very little information is available on phloem sap RNAs and proteins associated with phytoplasma infection.
In the present study, high-throughput transcriptomic and iTRAQ 1 proteomic approaches were combined to profile the phloem sap mRNAs and proteins involved in the response of mulberry to phytoplasma infection, and the differentially expressed genes and proteins were identified and their functions were discussed. Moreover, one of the differentially expressed gene, major latex protein-like 329 (MuMLPL329) was cloned and its functions were analyzed. Our results demonstrate that MuMLPL329 acts as a positive regulator participating in plant defense response. The information provided will help better understand the plant-phytoplasma interactions and shed light on the underlying molecular mechanisms of phytoplasma pathogenicity.

EXPERIMENTAL PROCEDURES
Plant Materials-Dormant hardwood stem branches of the same tree, Husang 32 (M. multicaulis Perr.), were cut into 17-cm cuttings and incubated in a growth chamber at 26°C, humidity 90% and under 12 h of light. In the next summer, the cutting seedlings were inoculated with phytoplasma by being grafted with the scions collected from phytoplasma-infected mulberry trees (Husang 32), and the seedlings grafted with the scions collected from healthy mulberry trees used as controls. Six weeks after inoculation, the plants showing typical symptoms as yellowing of the leaves, stunting, and witches'broom were used to detect phytoplasma by the PCR assay with an amplified fragment of the 16S rRNA gene of phytoplasma (GenBank Accession No. EF532410) as described previously (21). The plants showing positive symptoms and the controls were used to collect phloem saps.
Phloem Sap Sampling-Phloem sap was collected from infected and healthy mulberry plants using the shoot exudation method (26) with modifications. Briefly, the shoot was cut with a sterile razor blade between the fourth and sixth leaves from the top of the grafted shoots, and the first droplet was discarded, and the cut surface was blotted with sterile filter paper (3 mm; Whatman, Maidstone, UK) several times to avoid contamination. Exuding phloem saps thereafter were collected using sterile micropipette tips (200 l) and stored immediately at Ϫ80°C.
Library Construction and Sequencing-The gene expression libraries were prepared using the Illumina Gene Expression Sample Preparation Kit (Illumina San Diego, CA) according to the manufacturer's instructions. Briefly, total RNA was isolated from phloem sap samples using TRIzol ® reagent (Invitrogen, CA) following the manufacturer's instructions, and mRNA was isolated from the total RNA using magnetic oligo (dT) beads, and the ds-cDNAs were digested by the restriction enzymes Nla III and Mme I. After ligation with sequencing adaptors, PCR was performed, and the products were purified by 6% polyacrylamide Tris borate-EDTA gel. The tag libraries constructed were deep-sequenced on the Illumina sequencing platform (GAII) (Illumina).
Differential Expression Analyses of mRNAs-The raw sequence output data were processed using the Illumina pipeline, and "Clean Tags" were obtained by filtering and removing the 3Ј adaptor sequences, adaptor-only tags, and low-quality tags. The clean reads obtained were then aligned to our in-house mulberry transcriptome dataset, and the ambiguous tags with multiple hits were excluded. The number of annotated clean tags for each gene was then normalized by RESM based algorithm using perl scripts in the Trinity package (v2013-02-25) (27) to obtain the number of transcripts per million clean tags (TPM). The fold change and p value were calculated from the normalized expression, and the false discovery rate (FDR) was applied to determine the threshold of the p value in multiple tests and analyses. An "FDR Ͻ 0.01 and the absolute value of log2 Ratio Ն 1" was used as the threshold to judge the significance of gene expression difference.
Protein Preparation and Isobaric Labeling-Aliquots of 300 l of phloem sap were collected from diseased and healthy plants, and were expelled into 700 l of cold trichloroacetic acid/acetone and precipitated overnight at Ϫ20°C. The precipitated proteins were collected by centrifugation for 30 min at 40,000 ϫ g at 4°C and washed with acetone, after which they were air dried. The dried powder was resuspended in lysis buffer at 25°C for 1.5 h, and then was centrifuged at 25,000 ϫ g for 1 h at 15°C. Protein was purified using the Clean Up Kit (GE Healthcare, Uppsala, Sweden) following the manufacturer's instructions. The cleaned samples were dissolved in 0.1% sodium dodecyl sulfate containing 500 mmol/L triethylammonium bicarbonate and then reduced, alkylated and trypsin-digested before labeling with an 8-plex iTRAQ Reagents Kit (Applied Biosystems, Foster City, CA) according to manufacturer's instructions.
LC-MS/MS Analysis-After being labeled, the samples were mixed, dried, resuspended, and then loaded on a MacroSpin Vydac C18 reverse-phase minicolumn (Nestgroup Inc., Southborough, MA, USA). After washing and elution, the samples were dried down and fractionated using a strong cation-exchange (SCX) column. Mobile phase A (25% ACN, 10 mmol/L KH 2 PO 4 ) and mobile phase B (25% ACN, 2 M KCl, 10 mmol/L KH 2 PO 4 ) were selected. Peptides were eluted at a flow rate of 200 l/min with a linear gradient of 0 -5% solvent B for 30 min and then 30 -50% solvent B for 30 min, maintained for 10 min, and then ramped up to 100% solvent B in 5 min, after which they were held for 10 min. The absorbance at 280 nm was monitored: a total of 20 fractions were collected and pooled into 10 fractions. Each SCX fraction was lyophilized and redissolved in solvent [5% (v/v) acetonitrile, 0.1% (v/v) acetic acid] plus 0.01% trifluoroacetic acid. The peptides were loaded onto a C18 capillary trap cartridge (LC Packings) and then separated on a 15-cm nanoflow C18 column at a flow rate of 200 nL/min with a Proxeon EASY-nLC system (Odense, Denmark). Mobile phase A (0.1% formic acid in water) and mobile phase B [100% acetonitrile, 0.1% formic acid (v/v)] were selected. The peptides were eluted from the HPLC column by a linear gradient as follows: 5% solvent B for 10 min, 5-30% solvent B for 30 min, 30 -60% solvent B for 5 min, 60 -80% solvent B for 5 min and holding for 5 min, and 80 -5% solvent B for 5 min, after which they were maintained at 5% solvent B for 5 min. Flow from the column was directed to a Q-Exactive quadrupole Orbitrap mass spectrometer (Thermo Scientific, Bremen, Germany) operating in positive ion mode.
Precursors with a mass range of 300 -2000 m/z and a calculated charge of ϩ2 to ϩ4 were selected for fragmentation. For each MS spectrum, a maximum of three of the most abundant peptides above 5-count threshold were selected for MS/MS. Each selected precursor ion was dynamically excluded for 30 s with a mass tolerance of 0.03 Da. The fragment intensity multiplier was set to 20, and the maximum accumulation time was 2 s.
MS Spectra and Differential Expression Analysis of Proteins-The MS/MS data were analyzed by a thorough search considering biological modifications against the protein database deduced using the GENScan software from our mulberry transcriptome using the Paragon algorithm in ProteinPilot software version 2.0.1. Briefly, fixed modification of methylmethanethiosulphate-labeled cysteine, fixed iTRAQ modification of free amine at the N terminus and lysine, variable modification of oxidation on methionine, and two miscleavages of trypsin digestion were considered. The mass tolerance for both MS and MS/MS was 0.2 Da. A concatenated target-decoy database search strategy was also employed to estimate the FDR. FDR was calculated as the 2-fold percentage of decoy matches divided by the total matches, and FDR of the reported iTRAQ data set was Ͻ1%. The ProteinPilot software employed the peak area of iTRAQ reporters for quantification. For relative protein quantification, the mean, standard deviation, and p values to estimate statistical significance of the protein changes were calculated using Pro Group (Applied Biosystems). The mass spectrometry proteomics raw data and identifying information have been deposited to PeptideAtlas (http://www. peptideatlas.org/) with the data set identifier PASS00997.
Bioinformatics-BlastN searches against the reference Arabidopsis thaliana database downloaded from TAIR (http://www.arabidopsis.org/; release 10) were used to provide gene ontologies for the genes and proteins identified. GO analysis was performed for BLAST-matched Arabidopsis accession entries of the target genes based on their TAIR GO categories, and the assignment of functional terms was supported by Blast2GO. MapMan software (http://gabi.rzpd.de/projects/ MapMan) was used to provide a graphical overview of the metabolic and regulatory pathways for the detected genes as described by Gao et al. (28).
RNA Gel Blotting-Total RNA extracted was separated on 1.2% (v/v) formaldehyde denaturing agarose gel and then was blotted onto a nylon Hybond N membrane. The blots were hybridized with digoxigenin-labeled RNA probes which were complementary to the genes and prepared using the PCR DIG Probe Synthesis Kit (Roche, Mannheim, Germany). The probes used are given in the supplemental Table  S1. Prehybridization, hybridization, membrane washing, and detection were performed according to the procedure described by Umezawa et al. (29).
Protein Gel Blot Analysis-Prepared protein was mixed with 5 ϫ SDS-PAGE sample buffer. Samples were heated at 95°C for 3 min and loaded on 12% (v/v) SDS-polyacrylamide gels. After electrophoresis, proteins were electroblotted onto polyvinylidene difluoride (PVDF) membranes. Western blot analysis was performed according to a previously described method (30) using the horseradish peroxidase (HRP)-conjugated polyclonal anti-Rubisco or anti-MuMLPL329 protein polyclonal antibody which were generated by immunizing rabbits with the purified proteins.
qRT-PCR Analysis-RNA was extracted using the TRIzol ® reagent (Invitrogen) and digested with DNase I. qRT-PCR was performed using the CFX96TM Real-time System (Bio-Rad, CA) with the SYBR Premix Ex Taq TM kit from TaKaRa according to the manufacturer's protocol. The EF1-␣ or actin gene was amplified as a reference gene. The relative gene expression was evaluated using the comparative cycle threshold (Ct) method (31). All samples were assayed in triplicate. The primers used for qRT-PCR are given in the supplemental Table S2.
Gene Cloning-RNA was isolated from leaves of M. multicaulis using TRIzol R reagent (Invitrogen) and digested with DNase I and used to synthesize cDNA with 100 units of reverse transcriptase M-MLV (Promega, Madison, WI) in 20 ml reactions. The specific oligonucleotide primers were designed based on our available mulberry transcriptome data for PCR amplifications, and the DNA fragment obtained from RT-PCR was subcloned individually into the pMD18-T vector (Invitrogen). After transformation, positive clones were selected and further sequenced. The primers used for RT-PCR are given in the supplemental Table S3.
Sequence and Phylogenetic Analysis-The deduced amino acid sequences were aligned using DNAMAN software (version 6.0). The theoretical isoelectric point (pI) was calculated using pI/Mw tool within the ExPASy Proteomics Server (http://www.ca.expasy.org/tools/ pi_tool.html), and structural prediction was performed with SWISS-MODEL tools (http://www.swissmodel.expasy.org/). The neighborjoining method was used to produce the phylogenetic tree using the MEGA program. Bootstrapping was performed 1000 times to obtain support values for each branch.
Production of Transgenic Plant Lines-The coding region of the MuMLPL329 gene was amplified from pMD18-MuMLPL329 plasmid DNA with the sense primer containing an Xba I sequence at the 5Ј end, and the antisense primers containing a SacI sequence at the 3Ј end, and the amplified DNA was integrated into pMD18-T vector. The primers used for RT-PCR are given in the supplemental Table S3. The positive plasmid was digested with Xba I and Sac I, and the products were analyzed by agarose gel electrophoresis. The DNA fragment of ϳ450 bp was recovered and subcloned into binary plasmid vector pBI121 (digested with Xba I and Sac I) under the control of the 35 S promoter. Then, the construct was introduced into wild-type Arabidopsis plants through an Agrobacterium tumefaciens-mediated (strain GV3101) T-DNA transformation with the floral dip method. After transformation, the T1 seeds were sterilized and plated on kanamycin selection plates (MS media supplemented with 50 mg/ml kanamycin) to select transformed plants.
Determination of MuMLPL329 Subcellular Localization-The cDNA fragment of MuMLPL329 was cloned into the binary plasmid vector pROKII-EGFP under the control of 35 S to produce 35 S::MuMLPL329-EGFP expression vector. Mesophyll protoplasts from Arabidopsis thaliana were isolated according to the method described previously (32) and transformed with the 35 S::MuMLPL329-EGFP construct. After 12 h of incubation at 28°C, fluorescence images were acquired with a Bio-Rad MRC1024 confocal laser scanning microscope (Bio-Rad Microscience).
Plant Treatment-Mulberry and Nicotiana benthamiana seedlings used in the experiment were planted in a growth chamber at 26°C, 90% RH, and 12 h of light. The salicylic acid (SA) and jasmonate (JA) treatments were achieved by spraying the leaves with 5 mmol/L SA or 100 mmol/L JA solution. The control plants were sprayed with distilled water. Mulberry seedlings were inoculated with Pseudomonas syringae pv. mori by brushing the bacterial suspension (10 8 CFU/ml) onto the abaxial surfaces of young leaves. For Colletotrichum dematium inoculation, a filter paper disc (8 mm in diameter) soaked in conidial suspension (2.5 ϫ 10 6 conidia/ml) of C. dematium was placed on the adaxial surface of the young leaves. Inoculated mulberry plants were covered with polyethylene bags for 48 h after inoculation. N. benthamiana seedlings were inoculated with Pst. DC3000 by injecting 50 l of Pst. DC3000 (10 5 CFU/ml) bacterial suspensions into a leaf with a syringe, and inoculated with Botrytis cinerea by placing 5-l droplets of a spore suspension (2 ϫ 10 5 conidia/ml) in 24 g/L potato dextrose broth on the leaves. Inoculated plants were covered with a transparent plastic lid to maintain high humidity.
Detection of Resistance Against Pathogens-The MuMLPL329 gene was ligated to pBI121 vector and then introduced into GV3101 under the control of the 35 S promoter. The wild type Arabidopsis plants were transformed, and 4-week-old transgenic Arabidopsis plants were used for the resistance analysis. Inoculation with Pst. DC3000 was conducted by injecting 50 l of Pst. DC3000 (10 5 CFU/ml) bacterial suspensions or 10 mmol/L MgCl 2 (mock) into the rosette leaves with a syringe. To quantify the bacterial growth within the leaves, the inoculated leaves were ground in 1000 l of sterile water, serially diluted l/10 with sterile water and plated on King's B agar medium. Plates were placed at 28°C for 2 days, after which the colony-forming units were counted. Each treatment was conducted independently at least three times. Inoculation with B. cinerea was performed as described above. Inoculated plants were covered with a transparent plastic lid to maintain high humidity and incubated at 22°C with a 12-h photoperiod. The disease incidence and disease severity were examined daily after inoculation.
Inoculation of Arabidopsis with phytoplasma was performed following the method described before (20). Simply, to obtain phytoplasma-infected leafhoppers, leafhoppers (Hishmonus sellatus) were transferred to phytoplasma-infected mulberry for 2 weeks to allow oviposition and the hatched nymphs were kept on the infected plants until adulthood. Five adult leafhoppers were transferred to Arabidopsis seedlings which were at the stage of 4 to 5 rosette leaves and covered with clear tubes and then removed after 5 days. Plants infected with uninfected leafhoppers were used as controls. The presence of phytoplasma in the Arabidopsis samples was determined using qRT-PCR described previously (33). The primers and TaqMan probes (Forward primer 5Ј CGTACGCAAGTATGAAACTTAAAGGA 3Ј; Reverse primer 5Ј TCTTCGAATTAAACAACATGATCCA 3Ј; Probe 5Ј TGACGGGACTCCGCACAAGCG 3Ј) was used to amplify the 16S rRNA gene of phytoplasma. The primers and TaqMan probes (Forward primer 5Ј GACTACGTCCCTGCCCTTTG 3Ј; Reverse primer 5Ј AACACTTCACCGGACCATTCA 3Ј; Probe 5Ј ACACACCGCCCG-TCGCTCC 3Ј) was used to amplify the 18S rRNA gene of Arabidopsis. Normalization of phytoplasma quantities was performed by assays for plant 18S rDNA from each sample, and the number of phytoplasma cells per microgram of plant DNA was obtained by dividing the phytoplasma amount by the amount of DNA in each sample. In each qRT-PCR plate, water control was also included. All the samples were run in triplicate.
Promoter Analysis-To obtain the promoter sequence of MuM-LPL329, chromosome walking was performed using the TAIL-PCR method. Three specific primers SP1, SP2 and SP3 were designed based on the MuMLPL329 sequence, and four arbitrary primers LAD-1, LAD-2, LAD-3, and LAD-4 and the AC1 primer complementary to the adaptor sequence within the LAD primers were designed according to the method described before (34). All the primers used are given in the supplemental Table S3. Mulberry genomic DNA was isolated using CTAB (cetyltrimethyl-ammonium bromide) method (35) and used to preamplification using LAD and SP1 primers. In the primary TAIL-PCR, the amplification product was used as the template and the primer pairs AC1 and SP2 were used. In the secondary TAIL-PCR, the primary TAIL-PCR product was used as the template and the primer pairs AC1 and SP3 were used. The secondary TAIL-PCR amplified products were fractionated on an agarose gel, and the interest bands were isolated and sequenced. The potential cisregulatory elements within the sequence were analyzed with the PlantCARE software (http://bioinformatics.psb.ugent.be/webtools/ plantcare/html/). The promoter was cloned into the vector pBI121 to replace the 35 S promoter and fused to the ␤-glucuronidase (GUS) reporter gene to create the promoter expression vector pMLPL329::GUS which was then introduced into GV3101. For transient expression, tobacco leaves were infiltrated with transformed Agrobacterium according to the method described by Arpat et al. (36). For stable expression, Arabidopsis was transformated using the floral dipping method. Histochemical staining for GUS activity was performed as described by Jefferson et al. (37).
Determination of Flavonoid Content-Leaves sampled were dried to constant weight and then were grinded to a powder. Leaf powder was extracted with 70% (v/v) ethanol and the extract was filtered. Total flavonoid content was measured with a NaNO 2 -Al(NO 3 ) 3 method as described by Zhang et al. (38).
Experimental Design and Statistical Rationale-For the proteomic analysis, we obtained 3 biological replicates with healthy and infected phloem sap used for iTRAQ-based mass spectrometry analysis. For each of the 3 biological replicates, we performed three LC-MS/MS experiments as technical replicate. For gene expression and other biochemical, functional and quantitative assays, all the experiments were performed with at least three biological replicates, and the significance of the results was analyzed afterward via analysis of variance. Differences were considered significant when p values were less than 0.05.

RESULTS
Purity Assessing of Phloem Sap-To assess the purity of the sampled phloem sap, the Rubisco subunit which is found in other cell types but not in the SEs, and CmPP16, which is found in the SEs but not in other cell types, were used as specific cellular protein and mRNA markers, respectively, and their frequency in the samples was determined. The Western blotting results showed that the large subunit of Rubisco was not detected in the phloem sap sampled from either the healthy or the infected trees, but it was clearly present in leaf tissue samples free of major veins (Fig. 1A). The purity of the collected phloem sap samples was further confirmed by Northern blotting experiments using a Rubisco large subunit RNA specific probe and a phloem-specific CmPP16 RNA probe. The results showed that Rubisco large subunit RNA were present in the leaf tissue samples but absent in the phloem sap samples (Fig. 1B). In contrast, the phloem-specific CmPP16 RNA was detected in the phloem sap samples but barely detectable in the leaf tissue samples (Fig. 1C). Therefore, the contamination from surrounding tissues in the phloem sap samples was very low.
Gene Differential Expression Analysis Between Phytoplasma-infected and Healthy Mulberry Phloem Sap Libraries-After discarding low-quality and single-copy tags, totals of 3392 293 and 3421 516 clean tags remained were obtained in the infected phloem sap (IPS) and healthy phloem sap (HPS) libraries, respectively. Differences between the tag frequencies that appeared in the IPS and HPS libraries were used to estimate the unigene expression levels in response to phytoplasma infection. A total of 955 unigenes identified were found to be responsive to phytoplasma infection, among which 663 unigenes increased and 292 unigenes decreased significantly in the IPS library (p Ͻ 0.05, fold-change Ͼ 2.0). The top 20 up-and downregulated genes were given in Table  I, and the unigenes detected with at least 2-fold-change differences in the two libraries are shown in supplemental Table S4.
To validate the Solexa expression profiles, qRT-PCR analysis for 14 individual unigenes covering different expression patterns were performed (Fig. 2). Even though, the different scales of these unigenes detected by qRT-PCR analysis not in accordance with those detected by the Solexa sequencing, there was a very strong correlation between the PCR results and the Solexa-sequencing results, indicating that the profiles of these unigenes detected by Illumina sequencing are reliable.
To better understand the functions of these differentially expressed unigenes, the comprehensive tool MapMan was used to visualize the pathways affected by phytoplasmainfection in mulberry phloem saps. In the category "metabolism," primary metabolic pathways like TCA, lipid metabolism, carbohydrate (CHO) metabolism, as well as cell wall was changed. In addition secondary metabolisms in the terpene, flavonoid, phenylpropanoid, and amino acid metabolism pathways were also changed (Fig. 3A). This indicated that the metabolism in mulberry phloem saps was affected by phytoplasma-infection. Next to the category "regulation," differentially expressed unigenes are attributed to the bins for protein degradation and modification, redox, regulation of transcription, receptor kinases, G-proteins, MAP kinases, and phosphe inositides. Specifically it was showed that some unigenes are attributed to the bins for the metabolism of hormone such as abscisic acid (ABA), indole acetic acid (IAA), cytokinins (CK), gibberellic acid (GA), 6-benzyladenine (6-BA), JA, SA, etc. This suggested that the regulation pathways and hormone-related crosstalk in mulberry phloem saps were profoundly affected by phytoplasma-infection (Fig. 3B). When focusing on the category "biotic stress," it was revealed that the unigenes involved in hormone metabolism, and some transcription factors and defense stress-related genes were significant changes in the phytoplasma-infected phloem saps (Fig. 3C). Therefore, these differentially expressed unigenes were related to a variety of biologic processes, and the regulatory networks of the sap mRNAs involved in the response to phytoplasma-infection are intricate.
Identification and Quantification of Differentially Expressed Proteins Between Phytoplasma-infected and Healthy Mulberry Phloem Saps-The protein database deduced from our mulberry transcriptome was used to identify phloem sap proteins, and there were 739 proteins were identified and quantified, and 136 proteins were differentially expressed proteins (p Ͻ 0.05, fold-change Ͼ 2.0) between infected and healthy phloem saps. A total of 96 proteins increased, whereas 40 proteins decreased in the infected phloem sap compared with the healthy sap (supplemental Table S5). The top 20 up-and down-regulated proteins were given in Table II.
The GO analysis of the differentially expressed proteins was performed, and these proteins were classified into 14 functional categories (Fig. 4). The first category of proteins was involved in carbohydrate transport and metabolism (15%), and the second category included proteins associated with protein posttranslational modification and chaperones. The proteins associated with stress-related and defense mechanisms belong to the third category, and all the proteins, except for three in this category, were upregulated in the infected phloem sap. The proteins whose functions are unknown were numerically equal to the third-category proteins. Further analyses of these proteins will likely yield new scientific insights into the phytoplasma-mulberry interactions and reveal the new biological functions of phloem sap. The other differentially expressed proteins belong to categories such as energy production and conversion, transcription and regulation of transcription, secondary metabolite biosynthesis, transport and catabolism, and signal transduction mechanisms, among others. Therefore, the differentially expressed proteins in the phloem sap may play important roles in diverse biologic processes, and the regulatory networks of proteins involved in the response of mulberry to phytoplasma infection are also intricate.
Correlation of mRNA and Protein Profiles in Response to Phytoplasma Infection-In contrast to the proteome data, which showed that 18.4% of the proteins measured, were differentially expressed, the transcriptomic analysis showed that only 14.1% of transcripts measured were differentially expressed. Among all the differentially expressed genes, only 14 were regulated both at the mRNA and protein levels in the response to phytoplasma infection; moreover, 11 of the 14

Phytoplasma-responsive mRNAs and Proteins in Mulberry Saps
genes were changed with the same trend. However, there were three genes, CL6661.Contig1 (heat shock 70-kDa protein), Unigene20103 (ATP synthase subunit delta) and Uni-gene8723 (GDSL esterase/lipase) that were regulated with the opposite trend at the protein and mRNA levels (supplemental Table S4 and supplemental Table S5). Therefore, integrative transcriptomic and proteomic analysis of the phloem sap would provide more information for the identification of genes involved in the biological response of mulberry to phytoplasma infection.

Identification and Characterization of the Phytoplasmaresponsive MLP-like Protein 329 Gene-Integrated analysis
showed that the Unigene32585, which was annotated as MLP-like protein 329 gene, was increased at the protein and mRNA levels in the phytoplasma-infected phloem sap, and this result was confirmed by Northern blotting and Western blotting analyses (Fig. 5). This indicates that the Uni-gene32585 gene may have important roles in the response of mulberry to phytoplasma infection. However, MLP-like protein 329 gene was previously unrecognized related with phytoplasma infection. To examine the potential role of the gene in mulberry response to phytoplasma infection, the gene was cloned and a full length encoding cDNA of 456 bp was obtained, which encoded a protein 151 AA residues in length with a predicted size of 1.73 kDa and pI of 5.50 (GenBank accession MG871460). Putative conserved domain of the protein was detected and the results showed that it has a Bet v1-like domain which is primarily found in major latex proteinlike (MLP-like) proteins. However, its homology to other subfamily proteins from the Bet v 1 family is low (Fig. 6). SWISS-MODEL predictions showed that the structural properties of Unigene32585 gene were similar to other Bet v 1 family proteins, which contain a Y-shaped hydrophobic cavity formed by seven antiparallel ␤-sheets and three ␣-helices (Fig. 7A). Phylogenetic analysis of this gene and those from other plants was conducted, and the result showed that it was closest to the MLP-like protein 329 gene from M. notabilis (Fig.  7B). Therefore, the Unigene32585 gene was named as MuMLPL329. N-terminal extension prediction suggested that MuMLPL329 contained neither obvious signal peptide (http:// www.cbs.dtu.dk/services/SignalP/) nor obvious sublocalization sequence (http://www.cbs.dtu.dk/services/targetP). To clarify its subcellular localization, the coding region of the MuMLPL329 gene was fused in frame to an EGFP at its C terminus and introduced into Arabidopsis mesophyll protoplasts. Fig. 8 shows that MuMLPL329-EGFP fusion proteins were accumulated in both the nucleus and cytoplasm of mesophyll protoplasts.
Expression Profile of MuMLPL329 Gene-To examine the potential role of MuMLPL329 gene in plant defense, the ex- pression pattern of the gene in various organs of mulberry was examined by qRT-PCR analysis. As shown in Fig. 9A, the gene was ubiquitously expressed in all organs investigated. In stem bark where the phytoplasma are mainly restricted, the abundance of MuMLPL329 transcript was higher than in other organs of mulberry. At the same time, the mulberry seedlings

Phytoplasma-responsive mRNAs and Proteins in Mulberry Saps
were challenged with P. syringae pv. mori and C. dematium, and treated with JA and SA, respectively (Fig. 9B). The induced expression of MuMLPL329 was verified by qRT-PCR analysis, and the results showed that the expression of MuMLPL329 was induced after P. syringae pv. mori and C. dematium challenge. Moreover, the expression of MuM-LPL329 was also enhanced by exogenous application of JA.
Whereas, there was no remarkable change in the expression level of MuMLPL329 gene after SA treatment. To further investigate whether MuMLPL329 was involved in the disease responses, the putative promoter, 2000 bp DNA upstream of the MuMLPL329 coding region sequence was cloned (designed as pMuMLPL329; GenBank accession MH370165) and fused to the reporter gene encoding GUS and transient expression of the GUS gene in tobacco leaves was performed. Staining results showed that GUS driven by pMuM-LPL329 was induced after inoculation of B. cinerea or Pst. DC3000, meanwhile, GUS activity was enhanced by exogenous application of JA, but not by application of SA (Fig. 10). These results described above indicated that MuMLPL329 was involved in defense against pathogen infection and its expression may be modulated by different plant hormone signaling in mulberry. dopsis plants ectopic expression of MuMLPL329 were incubated with the bacterial pathogen Pst. DC3000, respectively. The results showed that there were severe disease symptoms showing gray-brown lesion with chlorosis in the wild-type plants. In contrast, the disease symptoms were not evident in the leaves of transgenic plants, although mild chlorosis or necrosis was occasionally observed (Fig. 11A). Bacterial growth in the leaves was monitored to ascertain whether the lack of symptom development reflected the restriction of bacterial growth and multiplication inside the leaves. The result of detection showed that the CFU of Pst. DC3000 in the leaves of wild-type plants was significantly higher than that in the leaves of transgenic plants (Fig. 11B). Similar experiments were performed with transgenic Arabidopsis plants following inoculation with B. cinerea to examine the role of MuMLPL329 in the defense response to fungal pathogens. Four days after inoculation (dai), dark necrotic lesions and fungal hyphae were observed at the inoculation sites on the leaf surface of wild-type plants, and the beginning of chlorosis was also observed around the inoculation site. However, no disease sign was observed on the leaves of MuMLPL329-overex-pressing plants at 4 dai (Fig. 11C). Therefore, the results presented above indicate that MuMLPL329-overexpressing in Arabidopsis enhances plant resistance to B. cinerea and Pst. DC3000.

Overexpression of MuMLPL329 Gene Enhances the Disease Tolerance of Transgenic Arabidopsis Plants-Because
Because phytoplasmas are difficult to culture in vitro, to explore whether plant resistance to phytoplasma can also be enhanced by MuMLPL329, wild-type and MuMLPL329overexpressing plants inoculated with phytoplasma via sapfeeding of phytoplasma-infected insect vectors, leafhoppers. Three weeks post infection, transgenic lines that expressed MuMLPL329 did not develop dwarfism though showed some symptoms of witches' broom showing improved resistance to phytoplasma (Fig. 12A). In contrast, wild-type plants inoculated with phytoplasma via sap-feeding of phytoplasma-infected leafhoppers exhibited severe developmental abnormalities, including symptoms of witches' broom and dwarfism (Fig. 12B). Therefore, the expression of MuMLPL329 gene alleviates the phytoplasma-associated disease symptoms and partially reduces the growth inhibition of phytoplasma in transgenic Arabidopsis. Phytoplasma growth in the leaves was also monitored using qRT-PCR to ascertain whether the lack of symptom development reflected the restriction of phytoplasma growth and multiplication inside the leaves. The results showed that many phytoplasma was detected in the wild type plants inoculated with phytoplasma via phytoplasma-infected leafhoppers. Though phytoplasma was also found in the MuMLPL329-overexpressing plants inoculated with phytoplasma, the number of phytoplasma was lower than that in the wild type plants. There was no phytoplasma detected in the wild type and transgenic Arabidopsis plants infected by healthy leafhoppers (Fig. 12C). Therefore, MuMLPL329-overexpressing in Arabidopsis restrains phytoplasma growth and multiplication in some extent and alleviates the symptoms of phytoplasma diseased.

Ectopic Expression of MuMLPL329 Affects Defense-related
Gene Expression-To explore whether the disease resistance observed in the MuMLPL329-overexpressing transgenic plants was resulted from the expression of defense genes, the expressions of some defense-related genes were monitored in the transgenic lines. Fig. 13 shows that the defense-related gene PR-1, PR-5, ␤-1,3-glucanase and PDF1.2 were expressed at very low levels in the wild type plants. However, the PR-5, ␤-1,3-glucanase and PDF1.2 genes were highly expressed in the MuMLPL329-overexpressing transgenic plants. These results suggest that the enhanced disease resistance in the transgenic plants may be because of the constitutively high-level expression of some defense-related genes.

DISCUSSION
Transcripts and Proteins Identified as Responding to Phytoplasma Infection-Thus far, the composition of phloem proteins and mRNAs has been characterized in many plant species and many transcripts and proteins present in the phloem of mulberry were identified in this study. When the transcripts identified were compared with previously published collections of mobile RNAs in Arabidopsis thaliana (39) through BLAST analyses, 1124 mobile RNAs were found in the present study. Based on this approach, we compared the phloem proteins identified in this study to the previously published sequences of the pumpkin phloem proteome (12), and 214 pumpkin phloem proteins were also found in our data. Many pumpkin proteins have also been found in rice, rape, and castor bean. Therefore, the phloem proteins and mRNAs in species may have a high degree of conservation in higher land plants (12,14). However, to our knowledge, this study is the first large-scale analysis of phloem proteins and mRNAs involved in the response to phytoplasma in mulberry. The data obtained from transcriptomic and proteomic analyses showed a poor correlation, suggesting there are selectively import mechanisms for maintaining proper levels of transcripts and proteins in the phloem during the response to phytoplasma infection.
Recognition is the initial event in the response of plants to pathogens, and it can occur through adhesins, fimbriae, flagella, and Type III and Type IV secretion systems (TTSS) (40). Though phytoplasma has no flagella and TTSS, it resides within the plant cell and can secrete proteins into plant hosts via the bacterial Sec translocation system, and the proteins secreted may function in the infected-plant cytoplasm like TTSS virulence factors (41). Mulberry may recognize these proteins and activate a signal transduction cascade and elicit defense responses. Our data showed that there were 40 mRNAs and 5 proteins involved in signaling were found to be differentially expressed between healthy and infected phloem sap (supplemental Table S4 and supplemental Table S5). These differentially expressed mRNAs and proteins probably serve as signaling molecules in response to phytoplasmainfection when they are released from the phloem, and they likely activate their signaling cascades in the non-infected tis- sues and play important roles in modulating the response to phytoplasma infection at the whole-plant level.
Phytohormone signaling is crucial for plant growth and development, as well as plant response to environment stresses (42). It was reported that multiple auxin responsive genes and auxin efflux carrier genes of the infected plants can be down regulated by the phytoplasma effector TENGU (2), and the effector SAP11 can reduced JA synthesis (20). Our transcriptomic analysis showed that 23 mRNAs associated with the signaling and metabolism of diverse phytohormones, including IAA, ABA, 6-BA, JA, CK, were differentially expressed in the infected phloem sap. These differentially expressed mRNAs partially accounted for the disturbance of hormonal signaling and metabolism in the infected plants, and this is in accordance with previous reports that phytoplasma infection disrupts the phytohormone balances in the host plants, resulting in reprogramming of growth and developmental patterns and in other symptoms in infected plants (18,(43)(44)(45)(46)(47).
As intracellular parasites, phytoplasmas lack many genes related to amino acid and fatty acid biosynthesis, the tricarboxylic acid cycle and oxidative phosphorylation, suggesting that they must obtain essential metabolites from their hosts and this will have a great impact on the metabolome of infected plants (48). Our results showed that phytoplasma infection alters the expression of several genes and proteins correlated with metabolic processes of TCA, lipid, carbohydrate metabolism, as well as secondary metabolism (supplemental Table S4 and supplemental Table S5). The differential expression of the genes and proteins might meet the requirements of phytoplasma for energy, growth and spread. From a different perspective, the changes of these genes or proteins may alter the activity of specific enzymes involved in some metabolic processes and disturb the normal metabolic processes in the infected plants, and this may be partly responsible for some of the mulberry yellow dwarf symptoms.
In addition, our data showed that 38 mRNAs and 16 proteins involved in various aspects of biotic and abiotic stress were differentially expressed (supplemental Table S4 and  supplemental Table S5). However, the expressions of these mRNAs and proteins do not all increased, the defense responses are induced but not completely or sufficiently to kill the phytoplasmas in the diseased plants. Another possible explanation is that phytoplasmas may have achieved a strategy to overcome plant defenses at this level. Interestingly, although the mulberry plants were planted in well-watered conditions and suffered no temperature stress, some mRNAs and proteins involved in the response to temperature and dehydration stress were differentially expressed in the infected phloem sap. For example, HSP70, known to be involved in response to a variety of abiotic stresses, was also found to be differentially expressed between the healthy and infected phloem sap. HSP70 is known to perform chaperone functions by stabilizing new proteins to ensure correct folding or by helping refold proteins that were damaged by cellular stress (49). Therefore, the increased expression of HSP70 may reflect an adaptive response in the infected plants to stabilize proteins or to help refold proteins that were damaged by phytoplasma infection.
Moreover, it was reported that the effector SAP11 of phytoplasma targets the nuclei of plant cells and interacts with transcription factors and changes plant gene transcription levels (20). Our results showed that some transcription factors (TFs), such as transcription factor VIP1, WRKY transcription factor 11 and squamosa promoter-binding-like protein 1, were also differentially expressed in the phloem sap under phytoplasma infection (supplemental Table S4 and supplemental Table S5). These TFs are known to be involved in responses to both biotic and abiotic stresses and play crucial roles in the regulation of defense genes, signaling, growth and development, and metabolism. Therefore, phytoplasma infection may be more than a biotic stress; the infection may also result in some abiotic stresses in the infected plants, and these differentially expressed genes may mediate the crosstalk between the abiotic and biotic stress responses. Furthermore, the differential expression of their TFs may result in up or downregulation some genes associated with plant growth and development and disturb the plant development process and be responsible for some mulberry yellow dwarf symptoms, which may aid phytoplasma colonization and increase plant attractiveness to insect vectors.
MuMLPL329 Is a Defense-related Protein in Mulberry-Though lots of genes involved in the response to phytoplasma infection have been reported in different plants, there are few reports to confirm the roles of these genes in resistance against phytoplasma diseases. As far as we know, this is the first report that MLP-like protein 329 gene was involved in the response to phytoplasma infection. Major latex protein (MLP) was first identified from the latex of opium poppy (Papaver somniferum) (50). Its orthologs, named MLP-Like proteins (MLPL), were later found in Arabidopsis as well as in other plants. Based on modest sequence similarity, they have been characterized as members of the Bet v 1 protein superfamily which contains proteins with low sequence similarity, but a similar three-dimensional (3D) structure (51). Although the MuMLPL329 protein has low similarity to other MLP proteins, it has a similar 3D structure to the Bet v 1 family proteins (Fig. 7A). In addition, as other MLP subfamily proteins, MuMLPL329 has no signal peptide, and subcellular localization analysis showed it was accumulated in both nucleus and cytoplasm (Fig. 8). In this study, the MuMLPL329 protein was detected in the phloem sap of mulberry, and this is consistent with the role of Bet v 1 family protein as a type of receptor taking part in binding or transporting molecules like plant hormones and secondary metabolites (52).
It has been reported that the function of MLPs is related to fruit and flower development in peach (53) and kiwifruit (54). Our data showed that MuMLPL329 was also expressed in various organs of mulberry plants indicating that the gene may also related to fruit development in mulberry. Meanwhile, several studies have reported that the expression of MLP genes is responsive to pathogen invasion, but the biological function of this protein family in defense responses is poorly understood (55), and there was little report about the function of MLPL329 proteins. In this study, MuMLPL329 was identified as a phytoplasma-responsive gene, and the induced expression of MuMLPL329 in mulberry in response to P. syringae pv. mori and C. dematium was also detected (Fig.  9B). This indicated that MuMLPL329 may play a role in the defense response in mulberry. Our results demonstrate that ectopic overexpression of the MuMLPL329 gene confers the transgenic Arabidopsis plants enhanced disease tolerance to B. cinerea and Pst. DC3000 (Fig. 11). Moreover, the transgenic Arabidopsis plants showed more tolerance to phytoplasma infection than wild type plants (Fig. 12). These results showed that the function of MuMLPL329 was tightly associated with plant defense response, and the expression of MuMLPL329 has been a strategy to overcome phytoplasmas in the diseased mulberry plant.
The role of Bet v 1 family protein in resistance to biotic and abiotic stresses is thought to be a type of receptor taking part in binding or transporting molecules like plant hormones and secondary metabolites (51). It was reported that Gh-MLP can binding or transport of ligand in transgenic Arabidopsis, thereby directly or indirectly affected flavonoid biosynthesis and change transgenic plant resistance to stress (52). Interestingly, it was found that some genes involved in flavonoid synthesis were changed in the grapevines (Vitis vinifera L.) plants infected by Flavescence doré e phytoplasma (54). In our data, the gene involved in flavonoid synthesis, CL773.Contig1 (flavonoid 3Ј-monooxygenase), was also found to be induced at both mRNA and protein levels in the phloem sap of phytoplasma infected mulberry plants (supplemental Table S4 and   supplemental Table S5). The total flavonoid content in the leaves of infected mulberry was significantly higher than that in the leaves of healthy mulberry (Fig. 14). In addition, the total flavonoid of the leaves from transgenic Arabidopsis plants were also determined, and the results showed that the average flavonoid content of transgenic Arabidopsis plants was significantly higher than that of wild type plants (Fig.  14). It was reported that many flavonoid compounds function as passive or inducible barriers against herbivores or microbial pathogens, and the flavonoid content can increase response to pathogen attack (57). These results suggest that MuMLPL329 is possible changing transgenic Arabidopsis plants resistance to pathogens infection by changing the flavonoid levels. However, the mechanism of MuMLPL329 involved in the variation of flavonoid content is not clear. Previous studies have shown that phytoplasma infections resulted in some phytohormones and metabolites being accumulated in the infected phloem saps and an inhibition of phloem transport, and this may be partly responsible for some of the mulberry yellow dwarf symptoms (18). Because MLPs were thought to take part in binding or transporting molecules like plant hormones and secondary metabolites, the increased expression of MuMLPL329 may make for transportation of metabolites and alleviating the symptoms.
It has been demonstrated that one of the MLPs of upland cotton, GhMLP28, can interact with ethylene response factor 6 (GhERF6) and enhance its transcription factor activity, which facilitate the expression of some GCC-box genes, such as PDF1.2 and PR-5, and led to an enhanced disease tolerance of the transgenic tobacco plants (55). In our data, the expression levels of PR-5, ␤-1,3-glucanase and PDF1.2 genes were also found to be increased in the MuMLPL329 transgenic Arabidopsis plants (Fig. 13). Because the Bet v 1 family proteins have similar 3D structure (Fig. 7A), MuM-FIG. 14. Total flavonoid content of mulberry and Arabidopsis leaves. Data represent the mean Ϯ S.D. of three independent biological samples. Different letters above the columns indicate significant differences (p Ͻ 0.05) according to Duncan's multiple range tests. Three biological repeats were performed. OE1, OE2 and OE3: Transgenic Arabidopsis lines overexpressing MulMLPL329.
LPL329 protein may as well interact with ethylene response factor 6 and facilitate the expression of PDF1.2 and PR-5 genes, and enhance disease tolerance of the transgenic Arabidopsis plants. In addition, our data showed that MuMLPL329 gene can be induced by JA treatment (Fig. 9B,  Fig. 10). So, one possibility is that MuMLPL329 acts as a receptor, binding plant hormone and activating the transduction signal, and then affecting the expressions of some defense genes. It is to be regretted that the PDF1.2 and PR-5 genes were not found to be induced in the infected mulberry phloem saps (supplemental Table S4), though the expression level of MuMLPL329 protein was increased in the infected phloem saps. Therefore, the regulatory networks of genes involved in the response of mulberry to phytoplasma-infection is complex, and further studies are required to clear the molecular basis of the function of MuMLPL329.
In conclusion, interpretation of the transcriptomic and proteomic data has uncovered several phytoplasma-responsive candidate genes/proteins and provides a global picture of the gene/protein expression changes in the phloem sap of mulberry under phytoplasma infection. Our results provide a critical line of evidence showing MuMLPL329 acts as a positive regulator participating in plant defense response. The information provided here is particularly useful to understand the function of genes/proteins in the phloem sap of mulberry and to help reveal the mechanisms underlying phytoplasma pathogenicity and better understand the plant-phytoplasma interactions.

DATA AVAILABILITY
The mass spectrometry proteomics data have been deposited to PeptideAtlas (http://www.peptideatlas.org/) with the dataset identifier PASS00997.