Identification of Differentially Expressed Proteins in Sugarcane in Response to Infection by Xanthomonas albilineans Using iTRAQ Quantitative Proteomics

Sugarcane can suffer severe yield losses when affected by leaf scald, a disease caused by Xanthomonas albilineans. This bacterial pathogen colonizes the vascular system of sugarcane, which can result in reduced plant growth and plant death. In order to better understand the molecular mechanisms involved in the resistance of sugarcane to leaf scald, a comparative proteomic study was performed with two sugarcane cultivars inoculated with X. albilineans: one resistant (LCP 85-384) and one susceptible (ROC20) to leaf scald. The iTRAQ (isobaric tags for relative and absolute quantification) approach at 0 and 48 h post-inoculation (hpi) was used to identify and annotate differentially expressed proteins (DEPs). A total of 4295 proteins were associated with 1099 gene ontology (GO) terms by GO analysis. Among those, 285 were DEPs during X. albilineans infection in cultivars LCP 85-384 and ROC20. One hundred seventy-two DEPs were identified in resistant cultivar LCP 85-384, and 113 of these proteins were upregulated and 59 were downregulated. One hundred ninety-two DEPs were found in susceptible cultivar ROC20 and half of these (92) were upregulated, whereas the other half corresponded to downregulated proteins. The significantly upregulated DEPs in LCP 85-384 were involved in metabolic pathways, the biosynthesis of secondary metabolites, and the phenylpropanoid biosynthesis pathway. Additionally, the expression of seven candidate genes related to photosynthesis and glycolytic pathways, plant innate immune system, glycosylation process, plant cytochrome P450, and non-specific lipid transfer protein was verified based on transcription levels in sugarcane during infection by X. albilineans. Our findings shed new light on the differential expression of proteins in sugarcane cultivars in response to infection by X. albilineans. The identification of these genes provides important information for sugarcane variety improvement programs using molecular breeding strategies.


Introduction
Sugarcane (Saccharum spp. hybrids) is an important food and bioenergy source and a significant component of the economy in more than 100 countries in the tropics and subtropics [1]. Commercial sugarcane cultivars (2n = 100-130) have a highly polyploid, aneuploid, heterozygous, and interspecific genome. This genome is composed of about 80% of S. officinarum (2n = 80) chromosomes, 10-15% of Population size of X. albilineans was determined in inoculated plants at 0 and 48 hpi using a quantitative PCR (qPCR) assay developed by Garces et al. [25]. Briefly, leaf sampling was identical to those for iTRAQ analysis and then total genomic DNA was extracted using the standard CTAB protocol [10]. One microliter (µL) of total leaf DNA (100 ng/µL) and serial 10-fold dilutions (10 7 -10 copies/µL) of pMD-albI plasmid were used as qPCR templates [24]. Total DNA of a disease-free sugarcane leaf and sterile distilled water were used as negative and blank controls, respectively. Three biological replicates and three technical replicates were used for all the samples.

Total Protein Extraction and Peptide Preparation
To extract total proteins, frozen samples were individually ground into powder in a pre-chilled mortar with liquid nitrogen. The powder was mixed with lysis buffer containing 50 mM Tris-HCl (pH 8), 8 M Urea and 0.2% SDS. The homogenate was incubated with an ultrasonic homogenizer (JY92-IIDN, Ningbo, China) (power 150 W) on ice for 5 min and then centrifuged at 12,000× g for 15 min at 4 • C The~700 µL supernatant was transferred to a new 1.5 mL centrifuge tube. Then, 7 µL 2 mM dithiothreitol was added and samples were incubated at 56 • C for 1 h, followed by addition of sufficient iodacetic acid to the sample and an additional incubation for 1 h at 25 • C in the dark. Cold acetone (four-fold volume of supernatant) was added to the sample and vortexed vigorously for~10 s before placing samples overnight at −20 • C. The samples were centrifuged at 12,000× g for 15 min at 4 • C and the pellets were washed twice with cold acetone. Finally, the pellets were dissolved using dissolution buffer containing 0.1 M triethylammonium bicarbonate (TEAB, pH 8.5) and 8 M urea. Protein concentration was determined with the Bradford assay [26]. The supernatant of each sample containing precisely 0.1 mg of protein was digested with Tripsin Gold (Promega, Madison, WI, USA) at 37 • C for 16 h. The proteins were dried by vacuum centrifugation at 1000 rpm for 2 h after removal of the urea using a C18 desalting cartridge (3M Corporation, Saint Paul, MI, USA).

iTRAQ Labeling of Peptides
The desalted peptides were labeled with iTRAQ reagents (iTRAQ ® Reagent-8PLEX Multiplex Kit, Sigma-Aldrich, Shanghai, China), following the manufacturer's instructions. For each 0.1 mg of peptide, 1 unit of labeling reagent was used. Peptides were dissolved in 20 µL of 0.5 M TEAB and the labeling reagent was added to 70 µL of isopropanol. After incubation for 1 h, the reaction was stopped with 50 mM Tris-HCl (pH 8.0). Differentially labeled peptides were mixed equally and then desalted using peptide desalting spin columns (89852; ThermoFisher Scientific, Waltham, MA, USA).
For DDA, the Q-Exactive HF-X mass spectrometer was operated in positive polarity mode with a spray voltage of 2.3 kV and a capillary temperature of 320 • C. Full MS scans from 350 to 1500 m/z were acquired at a resolution of 60,000 (at 200 m/z) with an automatic gain control (AGC) target value of 3 × 10 6 and a maximum ion injection time of 20 milliseconds (ms). From the full MS scan, a maximum number of 40 of the most abundant precursor ions were selected for higher-energy collisional dissociation (HCD) fragment analysis at a resolution of 15,000 (at 200 m/z) with an automatic gain control (AGC) target value of 1 × 10 5 , a maximum ions injection time of 45 ms, a normalized collision energy of 32%, an intensity threshold of 8.3 × 10 3 , and the dynamic exclusion parameter set at 60 s.

Data Quality Control
The raw data obtained from MS detection were uploaded directly into proteome discovery v. 2.2 (Thermo Fisher Scientific, Waltham, MA, USA) software for database retrieval, peptide mapping and protein quantification. The retrieved results were filtered using proteome Discoverer v. 2.2. Peptide Spectrum Matches (PSMs) with 95% confidence intervals. Proteins containing at least one unique peptide fragment were considered reliable. The reliable PSMs and proteins were verified with other reliable proteins. Peptide fragments and proteins with false discovery rates (FDR) of >5% were excluded. The protein sequences of sugarcane infected by X. albilineans were deposited into the United States National Center for Biotechnology Information (NCBI) SRA database under accession number PXD015930.
Proteome Discoverer v. 2.2 software was used to acquire the relative quantitative values of the PSMs of all samples. The values were based on the peak area of the plot generated by the original spectrograph. The relative quantitative values of the unique peptide fragments determined following calibration were obtained based on the quantitative data for all unique peptide fragments of each protein. For the differential analysis of proteins, the mean quantitative value of all biological repeats of each sample was used to calculate the ratio (fold change) between two samples. To determine the statistical significance of the difference, a T-test was conducted on the relative quantitative value of each protein for two samples to be compared. When the fold change (FC) was ≥ 1.5 and p ≤ 0.05, protein expression was considered significantly increased (upregulated). When FC ≤ 0.67 and p ≤ 0.05, protein expression was considered significantly decreased (downregulated). Among all identified DEPs, protein-protein interactions (PPIs) were predicted in silico using software STRING v11.0. (https://string-db.org/) [28]. The PPIs (more than two protein interactions) data that were highly similar to sorghum proteins were retrieved and their networks were drawn using Cytoscape V3.6.1 (http://www.cytoscape.org/).

Quantitative Real-Time PCR (qRT-PCR) Analysis
Total RNA was extracted with the TRIzol ® Kit (Invitrogen, Carlsbad, CA, USA) following the manufacturer's recommendations. Quantitative real-time PCR (qRT-PCR) analysis of the seven candidate genes was performed with the QuantStudio ® Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). The GAPDH (Glyceraldehyde 3-phosphate dehydrogenase) gene was used as the internal control. First-strand cDNA was synthesized from 1 µg of total RNA using the HiScript ® III RT SuperMix of the qPCR (+gDNA wiper) kit (Vazyme, Nanjing, China). The resulting cDNA was then used for the qPCR assay with the ChamQ TM Universal SYBR ® qPCR Master Mix (Vazyme, Nanjing, China) following the manufacturer's instructions. All the primer pairs were designed with Beacon Designer software v. 8.20 (Primer Biosoft International, Palo Alto, CA, USA) and the primer sequences for each gene were listed in Supplementary Table S1. To confirm the specificity of the product and to avoid the production of primer dimmers, a dissociation curve was designed after each qPCR run. The 2 −∆∆CT method was used to analyze the relative changes in expression of each gene [29]. Three independent biological replicates (aliquots) and three replicates were set for each leaf tissue sample.

Statistical Analyses
The paired comparison T-test was used to determine if the differences between protein (log 2 Fold Change) and transcriptional (2 -∆∆CT ) data of seven candidate genes at 48 hpi were significant. A general linear model was fitted to all relative expression levels (2 -∆∆CT ) of each gene using the one-way ANOVA procedure, and Student-Newman-Keuls (SNK) test was performed with the mean values. All the statistical analyses were conducted with SAS version 8.1 (SAS Institute, Cary, NC, USA).

Pathogen Population Size in the Resistant and the Susceptible Sugarcane Cultivar after Inoculation with X. albilineans
The population size of X. albilineans was six time higher in leaf scald susceptible cultivar ROC20 (mean of 611 copies of X. albilineans genome/µL) than in resistant cultivar LCP 85-384 (mean of 102 copies of X. albilineans genome/µL) at 48 hpi. No genome of X. albilineans was detected in the control plants inoculated with sterile liquid medium (Ct values greater than 35).

Identification of Differentially Expressed Proteins (DEPs) in Response to X. albilineans Infection
A total of 285 DEPs were identified for the two sugarcane cultivars after inoculation with X. albilineans and based on a p value ≤ 0.05 and an expression change ≥ 1.5 (up-regulation) or ≤ 0.67 (down-regulation) (Supplementary Figure S3). Among those, 164 DEPs were upregulated (Supplementary Table S2) and 123 DEPs were downregulated (Supplementary Table S3). For resistant cultivar LCP 85-384, 172 DEPs were found after inoculation with X. albilineans (R48_ vs. R0_CK) and among those, 113 were upregulated and 59 were downregulated. For susceptible cultivar ROC20, 192 DEPs were identified after inoculation with the pathogen (S48_Xa vs. S0_CK), and the same number of DEPs (96) was upregulated and downregulated ( Figure 1A). Seventy-nine DEPs were shared by both cultivars. Among those, 45 were upregulated and 32 were downregulated, whereas two proteins (Cluster-4871.159982 and Cluster-4871.183445) were upregulated in LCP 85-384 but downregulated in ROC20 ( Figure 1B).
A total of 285 DEPs were identified for the two sugarcane cultivars after inoculation with X. albilineans and based on a p value ≤ 0.05 and an expression change ≥ 1.5 (up-regulation) or ≤ 0.67 (down-regulation) (Supplementary Figure S3). Among those, 164 DEPs were upregulated (Supplementary Table S2) and 123 DEPs were downregulated (Supplementary Table S3). For resistant cultivar LCP 85-384, 172 DEPs were found after inoculation with X. albilineans (R48_ vs. R0_CK) and among those, 113 were upregulated and 59 were downregulated. For susceptible cultivar ROC20, 192 DEPs were identified after inoculation with the pathogen (S48_Xa vs. S0_CK), and the same number of DEPs (96) was upregulated and downregulated ( Figure 1A). Seventy-nine DEPs were shared by both cultivars. Among those, 45 were upregulated and 32 were downregulated, whereas two proteins (Cluster-4871.159982 and Cluster-4871.183445) were upregulated in LCP 85-384 but downregulated in ROC20 ( Figure 1B).

Gene Annotation of DEPs
Seventy-six of the 285 DEPs identified by GO enrichment analysis were attributed to 21 functional groups. The most significantly enriched GO terms (>5 DEPs) in the library of resistant cultivar LCP 85-384 (R48_Xa vs. R0_CK) included "response to stimulus (GO:0050896)" and "response to oxidative stress" (GO:0006979) of the biological process category, as well as "heme binding (GO:0020037)" and "peroxidase activity" (GO:0004601) of the molecular function category (Figure 2). For the library of susceptible cultivar ROC20 (S48_Xa vs. S0_CK), the most significantly enriched GO terms (>5 DEPs) were "hydrolase activity, acting on ester bonds (GO:0016788)" of the molecular function category and "cellular component organization or biogenesis (GO:0071840)" of the biological process category. The DEPs that corresponded to the most significantly enriched GO terms in LCP 85-384 were upregulated proteins, whereas those in ROC20 were downregulated proteins (Supplementary Figure S4). medium, S48_Xa = cultivar ROC20LCP85-384 inoculated with X. albilineans The red scale color in (B) is associated with upregulation whereas the blue scale color corresponds to downregulation.

Gene Annotation of DEPs
Seventy-six of the 285 DEPs identified by GO enrichment analysis were attributed to 21 functional groups. The most significantly enriched GO terms (>5 DEPs) in the library of resistant cultivar LCP 85-384 (R48_Xa vs. R0_CK) included "response to stimulus (GO:0050896)" and "response to oxidative stress" (GO:0006979) of the biological process category, as well as "heme binding (GO:0020037)" and "peroxidase activity" (GO:0004601) of the molecular function category (Figure 2). For the library of susceptible cultivar ROC20 (S48_Xa vs. S0_CK), the most significantly enriched GO terms (>5 DEPs) were "hydrolase activity, acting on ester bonds (GO:0016788)" of the molecular function category and "cellular component organization or biogenesis (GO:0071840)" of the biological process category. The DEPs that corresponded to the most significantly enriched GO terms in LCP 85-384 were upregulated proteins, whereas those in ROC20 were downregulated proteins (Supplementary Figure S4).  . LCP 85-384 is resistant to leaf scald. DEPs were annotated with a GO term belonging to one of three biological process categories: biological process (BP, in pink color), molecular function (MF, in blue color), and cellular component (CC, in green color). The X-axis represents the p-value and Y-axis shows the GO term names.

Functional Classification of DEPs by KEGG Analysis
Most proteins identified by KEGG analysis were involved in metabolic pathways (map01100) of the two sugarcane cultivars (Supplementary Table S4). Proteins involved in biosynthesis of secondary metabolites (map01110) were significantly upregulated in resistant cultivar LCP 85-384 (Supplementary  Table S4). Some amino acid metabolism pathways (map00250 and map00270), purine and pyrimidine metabolism pathways (map00230 and map00240), and amino sugar and nucleotide sugar metabolism (map00520) were significantly enriched in both cultivars ( Table 2). Proteins associated with defense response pathways were essentially expressed in the leaf scald resistant cultivar (LCP 85-384) and included DEPs associated with phenylpropanoid biosynthesis pathway (map00940), ubiquitin mediated proteolysis (map04120), and glutathione metabolism (map00480) (Supplementary Table S4). Two photosynthesis-related pathways were also involved in the response of sugarcane to X. albilineans infection: photosynthesis pathway (ko00195) and photosynthesis-antenna proteins pathway (ko00196) ( Table 2).

Protein-Protein Interactions (PPIs) Network Predicted in the STRING Database
The STRING database is a large repository of protein-protein interaction networks, including functional interactions, regulatory interactions, and stable complexes of proteins. The PPIs of the 285 sugarcane DEPs were identified by submitting a protein query sequence in the search box of the database. Ninety-two DEPs interacted with each other in the two varieties inoculated with X. albilineans. Among those, 51 DEPs (36 upregulated and 15 downregulated) were found in LCP 83-384 whereas 61 DEPs (29 upregulated and 32 downregulated) were found in ROC20 (Figure 3). A majority of these proteins were enriched in the metabolic (ko01100), the biological secondary metabolic (ko01110), glutathione (ko00480) and glycolysis/gluconeogenesis pathways (ko00010).

Protein-Protein Interactions (PPIs) Network Predicted in the STRING Database
The STRING database is a large repository of protein-protein interaction networks, including functional interactions, regulatory interactions, and stable complexes of proteins. The PPIs of the 285 sugarcane DEPs were identified by submitting a protein query sequence in the search box of the database. Ninety-two DEPs interacted with each other in the two varieties inoculated with X. albilineans. Among those, 51 DEPs (36 upregulated and 15 downregulated) were found in LCP 83-384 whereas 61 DEPs (29 upregulated and 32 downregulated) were found in ROC20 (Figure 3). A majority of these proteins were enriched in the metabolic (ko01100), the biological secondary metabolic (ko01110), glutathione (ko00480) and glycolysis/gluconeogenesis pathways (ko00010).

Transcript Profiling of Seven Selected Genes by qRT-PCR
Seven candidate genes were selected for further investigation of their transcript profiling because the DEPs encoded by these genes were involved in two main KEGG pathways and five important disease-resistance gene families (Supplementary Table S1). The transcription level of these seven genes was determined by qRT-PCR analysis at 0 and 48 hpi. The PCR amplification efficiency of the seven genes ranged from 98% to 101%. Relative expression between protein levels (log 2 Fold Change) and transcriptional levels (2 -∆∆CT ) of these genes at 48 hpi was not significantly different according to the paired comparison T-test (p-values ranged from 0.1678 to 0.2172) ( Figure 5).
DEPs encoded by these genes were involved in two main KEGG pathways and five important disease-resistance gene families (Supplementary Table S1). The transcription level of these seven genes was determined by qRT-PCR analysis at 0 and 48 hpi. The PCR amplification efficiency of the seven genes ranged from 98% to 101%. Relative expression between protein levels (log2 Fold Change) and transcriptional levels (2 -ΔΔCT ) of these genes at 48 hpi was not significantly different according to the paired comparison T-test (p-values ranged from 0.1678 to 0.2172) ( Figure 5). Four genes were highly expressed in the leaf scald resistant and susceptible cultivars as compared to the pathogen-free control plants: GAPC3 coding for the cytosolic glyceroldehyde-3-phosphate dehydrogenase, UGT coding for the UDP-glycosyltransferase, nsLTP coding for the non-specific lipid transfer protein, and UBA1 coding for the ubiquitin-activating enzyme E1 ( Figure 6). Three genes coding respectively for the photosystem I P700 apoprotein A1 (psaA, Cluster-4871.13787), a non-specific lipid-transfer protein (nsLTP, Cluster-4871.183445), and the ubiquitin-activating enzyme E1 (UBA1, Cluster-4871.278138) were all downregulated in ROC20 in comparison to pathogen-free control plants. Four genes were highly expressed in the leaf scald resistant and susceptible cultivars as compared to the pathogen-free control plants: GAPC3 coding for the cytosolic glyceroldehyde-3-phosphate dehydrogenase, UGT coding for the UDP-glycosyltransferase, nsLTP coding for the non-specific lipid transfer protein, and UBA1 coding for the ubiquitin-activating enzyme E1 ( Figure 6). Three genes coding respectively for the photosystem I P700 apoprotein A1 (psaA, Cluster-4871.13787), a non-specific lipid-transfer protein (nsLTP, Cluster-4871.183445), and the ubiquitin-activating enzyme E1 (UBA1, Cluster-4871.278138) were all downregulated in ROC20 in comparison to pathogen-free control plants. The UDP-glycosyltransferase (UGT, Cluster-4871.235701) and nsLTP gene (Cluster-4871.183445) were highly upregulated (p < 0.01) in resistant cultivar LCP 85-384 with 3.5-and 9.9-fold changes, respectively. The psaA gene was also significantly highly expressed (p < 0.05), with a fold change of 2.1 in LCP 85-384. The genes coding for cytochrome P450 (P450, Cluster-4871.249909), the cytosolic glyceroldehyde-3-phosphate dehydrogenase (GAPC3, Cluster-4871.143463), and the argonaute family protein (AGO, Cluster-4871.119964) were significantly upregulated with 5.3-, 1.8-, and 1.7-fold change, respectively, in ROC20. The UBA1 (Cluster-4871.278138) gene was significantly upregulated with a 1.7-fold change in LCP 85-384 and was almost not expressed in ROC20 after inoculation with X. albilineans ( Figure 6). respectively, in ROC20. The UBA1 (Cluster-4871.278138) gene was significantly upregulated with a 1.7-fold change in LCP 85-384 and was almost not expressed in ROC20 after inoculation with X. albilineans ( Figure 6).

Overall Assessment of DEPs Involved in Response to X. albilineans Infection
This comparative proteomic study using the iTRAQ technique resulted in identification of 4295 proteins involved in response of sugarcane to infection by X. albilineans. More than 500 of these proteins were attributed to the oxidation-reduction process (biological process category) and ATP binding (molecular function category), while most of the predicted proteins were annotated as proteins of carbohydrate metabolism, as well as translation and amino acid metabolism pathways. For the leaf scald resistant cultivar LCP 85-384, 27 DEPs were enriched in metabolic pathways and 16 DEPs were enriched in biosynthesis of secondary metabolites, indicating that metabolites play key roles in sugarcane in response to X. albilineans infection. Large numbers of primary and secondary metabolites play vital roles in plant defense mechanisms involving complex cascades [30][31][32]. Besides metabolic pathways and the biosynthesis of secondary metabolites, infection of sugarcane by X. albilineans also triggered plant-defense related pathways such as phenylpropanoid biosynthesis, ubiquitin mediated proteolysis, glutathione metabolism, and photosynthesis. PPI network analysis also indicated that these important pathways participate in sugarcane resistance and defense response during X. albilineans infection. Indeed, the above-mentioned pathways are commonly activated in sugarcane in response to various pathogens [1].

Regulation of Photosynthesis and Glycolytic Pathways of Sugarcane in Response to X. albilineans Infection
Generally, genes related to photosynthesis are downregulated as chlorotic and necrotic tissues develop during infection of plants by pathogens [33,34]. However, in our study, the psaA1 gene (Cluster-4871.13787) was significantly upregulated in the leaf scald resistant cultivar and only slightly downregulated in the susceptible cultivar at protein and transcription levels. This suggested that stimulation of the first step of photosynthesis occurs during the early stages of expression of sugarcane lead scald resistance to allow efficient light-driven electron transport. Notably, psaA1 is one of PSI P700 apoproteins that are the primary electron donors of photosystem I (PSI) [35,36]. At the early stage (48 hpi) of plant infection by X. albilineans, no white-pencil lines nor chlorotic symptoms appeared on the two sugarcane cultivars. Similarly, the final receptors of electrons in light-dependent reactions (encoded by the ferredoxin [2Fe-2S] and ferredoxin-DNADP+ reductase genes) were overproduced during infection of sugarcane by Acidovorax avenae subsp. avenae [37]. Genes related to the photosynthesis-antenna proteins were also upregulated in Cucumis sativus against Cucurbit chlorotic yellows virus infection [38].
GAPC3 is one of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene families which play crucial roles in cellular processes [39]. In addition to being involved in glycolysis, GAPC is a phosphorylating NAD-dependent GAPDH catalyzing the conversion of glyceraldehyde-3-P (Ga3P) to 1,3-bisphosphoglycerate in the cytoplasm [40]. In our study, GAPC3 (Cluster-4871.143463) was highly expressed in both sugarcane cultivars but was upregulated in susceptible ROC20 at both protein and transcription levels. This suggested that GAPC3 does not contribute to disease resistance but may favor progress of the pathogen during plant colonization. Overexpression of MeGAPCs in cassava resulted in decreased disease resistance against X. axonopodis pv manihotis, the causal agent of cassava bacterial blight. In contrast, MeGAPCs-silenced cassava plants by virus-induced gene silencing conferred improved disease resistance, as evidenced by physical interaction of MeGAPCs with autophagy-related protein 8b (MeATG8b) and MeATG8e, and inhibition of autophagic activity [41].

Activation of Plant Innate Immune Systems in Sugarcane after Inoculation with X. albilineans
Ubiquitination is one of the posttranslational protein modifications governing plant immune responses [42][43][44]. The process of ubiquitination involves covalent attachment of the highly conserved small protein ubiquitin to substrate proteins through a stepwise enzymatic cascade. The ubiquitin-activating enzyme (E1 or UBA) is one of three different catalytic enzymes that typically catalyze at the initial step of ubiquitination [45,46]. In tobacco, the expression of NtUBA1 and NtUBA2 was regulated in response to viral infection, wounding, and defense-related hormones [47]. In Arabidopsis, enzyme E1 was involved in R-protein-mediated resistance [42]. Genome-wide analysis of genes encoding core components of the ubiquitin system in soybean revealed that a large number of UBS-related genes (including E1 gene GmUBA1) played a role in immunity against soybean cyst nematode [48]. In our study, UBA1 was overexpressed in leaf scald resistant LCP 85-384 at protein and mRNA levels, but highly repressed in susceptible ROC20, especially at transcription level. These results indicated the UBA1 gene plays an active role in sugarcane in response to colonization by X. albilineans.
RNA silencing plays a major role in regulating plant developmental processes and environmental adaptation to diverse biotic and abiotic stresses through transcriptional gene silencing (TGS) and post-transcriptional gene silencing (PTGS) [49]. In addition to a ribonuclease III-type dicer-like (DCL) enzyme and RNA-dependent RNA polymerases (RDRs), an Argonaute (AGO) protein (catalytic core in RNA-induced silencing complex (RISC) is also a core component of the RNA silencing process [50]. Of the 10 Arabidopsis AGO protein families, AtAGO4 participates in the repeat-associated siRNA (ra-siRNA) pathway mediating methylation of DNA repeats [51,52]. This protein is also involved in the miRNA pathway mediating gene expression, as evidenced by a subset of the miRNAs preferential association with AtAGO4 [53,54]. AtAGO4 is required for the resistance of Arabidopsis to Pseudomonas syringae and AGOs (including AGO4 protein) of Brassica napus are involved in plant resistance to the necrotrophic fungal pathogen Sclerotinia sclerotiorum [55,56]. In our study, the sugarcane AGO gene (a homologue of Zea mays AGO4, LOC100381552) was highly upregulated in the leaf scald resistant and in the susceptible cultivar, but especially in ROC20. This indicated that the AGO4 protein was activated in sugarcane during infection by X. albilineans. However, the molecular mechanisms involving AGO4 during the interactions between sugarcane and X. albilineans remain to be deciphered.

High Level Expression of the UDP-Glycosyltransferase in the Sugarcane Cultivar Resistant to Leaf Scald
The UDP-glycosyltransferase (UGTs; EC 2.4.1.91) catalyzes the transfer of sugar molecules to a variety of acceptor molecules, such as hormones, lipids and other small molecules [57,58]. The glycoside molecule then regulates the biological activity, water solubility and stability of receptors [57,59,60]. Up to now, 106 GT families have been identified in the carbohydrate-active enzyme database (CAZy; http://www.cazy.org/). The largest of these families is GT Family 1 that comprises a large number of UGT members [61,62]. UGTs play an important role in the regulation of plant hormone balance, detoxification of endogenous and exdogenous substances, and modification of secondary metabolites [58,60]. Recently, several studies suggested that the UDP-glycosyltransferase was positively regulated in plant species in response to pathogens. For example, expression of TaUGT4 in wheat was higher in a Fusarium head blight (FHB) resistant cultivar than in a susceptible one after treatment with deoxynivalenol (DON) produced by Fusarium graminearum [63]. Wheat overexpressing TaUGT5 was also more resistant to F. graminearum as evidenced by reduced proliferation and destruction of plant tissue by the pathogen [64]. Gene HvUGT-10W1 from barley conferred resistance to FHB [65]. The overexpression of the BnUGT74B1 gene in B. napus increased the aliphatic and indolic glucosinolates levels by a factor 1.7 and 1.5 in leaves infected by Sclerotinia sclerotiorum and Botrytis cinerea, respectively [66]. BrUGT74B1 was also involved in phytoalexins biosynthesis which has an important role in plant disease resistance [67].
In our study, a sugarcane UDP-glycosyltransferase (a homolog of UGT73C2 of A. thaliana) was highly upregulated at 48 hpi in the resistant cultivar following inoculation by X. albilineans, suggesting that this gene plays an important role during the defense of sugarcane against this pathogen. The receptor molecules of sugarcane involved in the glycosylation process remain to be identified. A primary response of sugarcane to Ustilago scitaminea infection appears to be the production of glycoproteins inhibiting germination and inducing aggregation of fungal teliospores [68].

Different Expression Patterns of Plant Non-Specific Lipid Transfer Proteins (nsLTPs) in the Leaf Scald Resistant and the Leaf Scald Susceptible Sugarcane Cultivar
nsLTPs are a group of small, basic proteins that are abundantly expressed in plants, having the ability to bind or transfer various types of hydrophobic molecules in vitro, such as fatty acids, fatty acyl-CoA, phospholipids, glycolipids, and cutin monomers [69,70]. nsLTPs are involved in key cellular processes such as the stabilization of membranes, cell wall organization, and signal transduction, in addition to responses to stress and developmental processes [70]. Notably, nsLTPs exhibit strong antimicrobial activity in vitro and interfere with the membrane of target organisms, thus leading to the loss of membrane integrity [70]. In our study, a sugarcane nsLTP gene (homolog of Zea may nsLTP I) was strongly overexpressed in leaf scald resistant cultivar LCP 85-384 at both protein and transcription levels, suggesting that this gene plays a positive role in resistance to X. albilineans. Similar observations were reported in other plants in response to infection by various pathogens. For example, several nsLTP isoforms of Trichoderma harzianum T39-treated grapevines increased after Plasamopara viticola inoculation [71]; the early expression level of the nsLTP gene was significantly increased in wheat infected by the rust pathogen Puccinia triticina before visible haustoria formation [72]. The nsLTP gene is also involved in the priming acquisition at the early priming stage and memory in beta-aminobutyric acid (BABA)-primed mango fruit after Colletotrichum gleosporioides inoculation [73]. These findings provided additional proofs of participation of nsLTPs in the defense response of plants to pathogens.

Upregulation of the Plant Cytochrome P450 after Colonization of Sugarcane by X. albilineans
Plant P450s participate in a large number of primary and secondary metabolisms, including the phenylpropanoid, flavonoid, cyanogenic glucoside, essential sterols and steroid hormones, and other biosynthetic pathways which are thought to convey adaptive advantages in specific ecological niches [74][75][76][77]. P450s acting on fatty acids (FA)-involved oxygenation reactions in plants is enhanced by biotic and abiotic stress at the transcriptional level [78]. For example, CYP709C1 (a P450 protein) was the first sub-terminal hydroxylase of long-chain FAs characterized in plants and its induction by methyl jasmonate resulted in plant defense reaction [79]. The P450 protein CYP74 of A. thaliana catalyzed the generation of oxylipins (jasmonates, aldehydes, divinyl ether, and alcohols) that acted as not only signaling molecules, but also exhibited antimicrobial and antifungal properties [80]. Oxidation of JA-isoleucine conjugate (JA-Ile) by cytochrome P450 monooxygenase is the major mechanism for turning off JA signaling [81,82]. Jasmonates are lipid-derived compounds that act as signals in plant stress responses and developmental processes [82,83].
In sugarcane, cytochrome P450 sequences have been annotated in the genome of S. spontaneum AP85-441 [3,84]. A sugarcane P450 protein showed interaction activities with ScMat1, a putative sugarcane transcription factor TFIIH subunit that has kinase activity [85]. Additionally, sugarcane gene ScCPR450 was highly expressed at the mRNA level in plants under SA or PEG stresses, suggesting that this gene plays a role in the response of sugarcane to stresses [86]. In our study, expression of a sugarcane P450 gene (homolog of Setaria italica P450 72A15) was increased in the leaf scald resistant and susceptible cultivars, particularly in the susceptible cultivar. A similar observation showed that GbCYP86A1-1 positively regulated the defense of Gossypium barbadense against Verticillium dahliae by cell wall modification and the activation of immune pathways [87]. However, the precise role of P450 proteins in sugarcane in response to pathogen infection remains to be determined.

Conclusions
This study provides the first global proteomic dataset of sugarcane in response to infection by X. albilineans. The DEPs that were identified are predicted to be involved in metabolic pathways and biosynthesis of secondary metabolites as well as some plant-defense related pathways, such as phenylpropanoid biosynthesis and plant immune signal transduction. Seven candidate genes coding for a UDP-glycosyltransferase, a non-specific lipid-transfer protein, an acytochrome P450 protein, a cytosolic glyceroldehyde-3-phosphate dehydrogenase, an argonaute family protein, the ubiquitin-activating enzyme E1, and the photosystem I P700 apoprotein A1 (chloroplast) were activated or repressed at the transcriptional level after inoculation with the leaf scald pathogen. These genes are good candidates for further investigations of the pathways involved in the sugarcane response to infection by X. albilineans.