Secretome Profiling Reveals Virulence-Associated Proteins of Fusarium proliferatum during Interaction with Banana Fruit

Secreted proteins are vital for the pathogenicity of many fungi through manipulating their hosts for efficient colonization. Fusarium proliferatum is a phytopathogenic fungus infecting many crops, vegetables, and fruit, including banana fruit. To access the proteins involved in pathogen–host interaction, we used label-free quantitative proteomics technology to comparatively analyze the secretomes of F. proliferatum cultured with and without banana peel in Czapek’s broth medium. By analyzing the secretomes of F. proliferatum, we have identified 105 proteins with 40 exclusively secreted and 65 increased in abundance in response to a banana peel. These proteins were involved in the promotion of invasion of banana fruit, and they were mainly categorized into virulence factors, cell wall degradation, metabolic process, response to stress, regulation, and another unknown biological process. The expressions of corresponding genes confirmed the existence of these secreted proteins in the banana peel. Furthermore, expression pattern suggested variable roles for these genes at different infection stages. This study expanded the current database of F. proliferatum secreted proteins which might be involved in the infection strategy of this fungus. Additionally, this study warranted the further attention of some secreted proteins that might initiate infection of F. proliferatum on banana fruit.


Introduction
The "secretome" is well-known as the collection of proteins that are delivered into the extracellular milieu via the secretory pathway [1]. Although fungal secretomes vary in composition and responsiveness between species [2], some secreted proteins, such as enzymes involved in carbohydrate degradation (the so-called cell wall-degrading enzymes or CWDEs), protein degradation, along with housekeeping enzymes, allergens, and proteins of unknown function are typically identified [3].

Strain and Culture Condition
The F. proliferatum strain ZYF was stored in 30% glycerol at −80 • C. Before use, the strain was dropped onto a potato dextrose agar (PDA, Oxoid, Basingstoke, Hampshire, England) plate and then incubated at 28 • C in the dark for 7 days. To collect fungal culture filtrate for MS, fungal agar plugs from 7-d-old plates were incubated in 200 mL Czapek's broth (CB) medium (3.0 g/L NaNO 3 , 1.0 g/L K 2 HPO 4 , 0.5 g/L MgSO 4 ·7H 2 O, 0.5 g/L KCl, 0.01 g/L FeSO4 and 30 g/L sucrose) sterilized previously at 121 • C for 20 min. Before incubation of F. proliferatum, the CB medium were cooled at room temperature and supplemented with and without approximately 5 g surface-sterilized banana peel (+BP/-BP). After incubation at 28 • C with shaking at 200 rpm for 3 days, the banana peel was removed, and the filtrate was collected for protein and fumonisin extraction. Three independent biological replications were carried out in this study.

Protein Extraction and Sample Preparation for a Label-Free Experiment
The filtrate was ultra-filtrated in 3KD ultrafiltration device, with 14,000 g for 10 min. The solution was filtered through a 0.22 µm Millipore membrane filter before being mixed with SDT solution (4% SDS, 100 mM DTT, 150 mM Tris-HCl, pH 8.0). After grounded with liquid nitrogen, five volumes of 10% TCA in acetone were added. Then the sample was vortexed and stored at −20 • C for at least 4 h. The pellets were collected after centrifugation at 6 000 g and 4 • C for 40 min and then washed with cold acetone for three times. The pellets were dried, and then the protein powder was re-suspended in 30 volumes of SDT lysis buffer (4% SDS, 100 mM DTT, 150 mM Tris-HCl, pH 8.0) and boiled for 5 min. The samples were further ultra-sonicated and boiled again for another 15 min. Undissolved cellular debris was removed by centrifugation at 14,000 g for 40 min. The supernatant was filtered through a 0.22 µm Millipore membrane filter and quantified with a Bicinchoninic acid (BCA) Protein Assay Kit (Bio-Rad, USA).

Protein Digestion
Digestion of protein (250 µg for each sample) was performed according to the filter aided sample preparation (FASP) procedure described by Wisniewski et al. [24]. Briefly, the detergent, DTT and other low-molecular-weight components were removed using 200 µL Urea (UA) buffer (8 M Urea, 150 mM Tris-HCl, pH 8.0) by repeated ultrafiltration (Microcon units, 30 kD) facilitated by centrifugation. Then 100 µL 0.05 M iodoacetamide in UA buffer was added to block reduced cysteine residues and the samples were incubated for 20 min in dark. The filter was washed with 100 µL UA buffer three times and then 100 µL 25 mM NH 4 HCO 3 twice. Finally, the protein suspension was digested with 3 µg trypsin (Promega) in 40 µL of 25 mM NH 4 HCO 3 overnight at 37 • C, and the resulting peptides were collected as a filtrate. The peptide content was estimated by UV light spectral density at 280 nm using an extinction coefficient of 1.1 of 0.1% (g/L) solution that was calculated based on the frequency of tryptophan and tyrosine in vertebrate proteins.

Liquid Chromatography (LC)-Electrospray Ionization (ESI) Tandem MS (MS/MS) Analysis by Q Exactive
The peptide of each sample was desalted on C18 Cartridges (Empore™ SPE Cartridges C18 (standard density), bed ID 7 mm, volume 3 mL, Sigma), then concentrated by vacuum centrifugation and reconstituted in 40 µL of 0.1% (v/v) trifluoroacetic acid. MS experiments were performed on a Q Exactive mass spectrometer that was coupled to Easy nLC (Proxeon Biosystems, now Thermo Fisher Scientific). 5 µg peptide was loaded onto a the C18-reversed phase column (Thermo Scientific Easy Column, 10 cm long, 75 µm inner diameter, 3 µm resin) in buffer A (2% acetonitrile and 0.1% Formic acid) and separated with a linear gradient of buffer B (80% acetonitrile and 0.1% Formic acid) at a flow rate of 250 nL/min controlled by IntelliFlow technology over 120 min. MS data were acquired using a data-dependent top10 method dynamically choosing the most abundant precursor ions from the survey scan (300-1800 m/z) for higher energy collision dissociation (HCD) fragmentation. Determination of the target value is based on predictive Automatic Gain Control (pAGC). Dynamic exclusion duration was 25 s. Survey scans were acquired at a resolution of 70,000 at m/z 200 and resolution for HCD spectra was set to 17,500 at m/z 200. The normalized collision energy was 30 eV and the underfill ratio, which specifies the minimum percentage of the target value likely to be reached at maximum fill time, was defined as 0.1%. The instrument was run with peptide recognition mode enabled. MS experiments were performed triply for each sample.

Sequence Database Searching and Data Analysis
The MS data were analyzed using MaxQuant software version 1.3.0.5. MS data were searched against the Fusarium proliferatum ET1 protein database (https://www.ncbi.nlm.nih.gov/genome/ proteins/2434?genome_assembly_id=295197) and banana genome database (http://banana-genomehub.southgreen.fr/home). Though there were several published genome databases of F. proliferatum (https://www.ncbi.nlm.nih.gov/genome/genomes/2434), only four strains had annotation information. Among them, ET1 strain had the best assembly results with fewest scaffolds. In addition, ET1 strain had more annotated genes and proteins. Hence, we chose ET1 strain for effector identification in this study. The detailed information for search parameters is shown in Table S2, Supplementary Materials. Label-free quantification was carried out in MaxQuant as previously described (Quantitative proteomics reveals subset-specific viral recognition in dendritic cells). Protein abundance was calculated based on the normalized spectral protein intensity (LFQ intensity). Proteins with LFQ intensity only in +BP sample were identified as exclusively secreted proteins in the +BP sample. Proteins with higher accumulation in +BP sample were identified as those with normalized total intensity ratios of (+BP: -BP) >2 combined with p-value < 0.05.

Fumonisin Production Analysis
After collection of 3-day-old culture (50 mL), Fumonisin B (FB) analysis was conducted according to previous research [19].

Fusarium Proliferatum Inoculations
The spore solution of F. proliferatum was diluted to 1 × 10 6 spores/mL for fruit inoculation. The green mature banana fruit were washed with sterile water, and then inoculated with 15 µL of aqueous conidia suspension described above. The inoculated fruit were stored at 22 • C and 85% relative humidity (RH). The peel tissues of banana fruit with and without F. proliferatum infection were collected at 4 days and 10 days after inoculation immediately ground into a powder with liquid nitrogen and stored at −80 • C for further analysis. Three biological replicates were conducted.

Quantitative Real-Time PCR (qRT-PCR) Analysis
RNA was extracted from 5 g of ground banana peel according to Shan et al. [30] followed by the removal of genomic DNA and cDNA synthesis. Specific primers for F. proliferatum genes were designed using Primer Premier 6 and shown in Table S3, Supplementary Materials. F. proliferatum Actin served as a reference gene for the quantification of gene expression. qRT-PCR was conducted according to Li et al. [19]. Three biological replicates were performed.

Statistics
All data were the average value of three replicate assays ± standard errors. Data for each sample were statistically analyzed using the Student's t-test (p < 0.05).

Identification of Secreted Proteins in Response to the Banana Peel
To investigate the possible effectors which are secreted extracellularly during the interaction with the host, F. proliferatum was cultured in CB medium supplemented with or without banana peel (+BP/-BP). Three days after culture, the culture filtrate was collected and concentrated for Liquid Chromatography-Electrospray Ionization Tandem MS (LC-ESI-MS/MS) analysis. We identified 1324 non-redundant proteins with 340696 total spectra and 7799 peptides. The details for the identification and quantitation are shown in Figure S1, Supplementary Materials. After the bioinformatics analysis, 108 proteins were identified as secreted proteins. Three banana proteins, mitochondrial phosphate transport protein, GTP-binding protein SAR1and 40S ribosomal protein S27-2, were found in the culture. All other 105 identified proteins originated from the fungus, and the detailed information as shown in Table S1, Supplementary Materials. Among them, 40 proteins were only identified in the +BP sample (Table 1) while 65 exhibited greater intensity in +BP sample (Table 2). In total, 70 fungal proteins had an N-terminal signal peptide sequence predicted by SignalP or TargetP, indicating that they are secreted proteins. Moreover, SecretomeP suggested that 35 proteins might be secreted in a nonclassical way (Tables 1 and 2). The Basic Local Alignment Search (BLAST) comparison with the database of fungal secreted proteins enabled further confirmation of these secreted proteins.   According to Blast2GO analysis, the secreted proteins only identified in the +BP sample were mainly categorized into the single-organism process, protein metabolic process, metabolic process, regulation of cellular process, and others ( Figure 1a). Meanwhile, the secreted proteins with higher accumulation in the +BP sample were mainly categorized into metabolic process, nitrogen compound metabolic process, cell wall organization or biogenesis, response to a stimulus, oxidation-reduction process, and others ( Figure 1b). The predicted secretome of F. proliferatum primarily contained proteins of unknown function and uncharacterized proteins (Figure 1, Tables 1 and 2).
Of particular interest are proteins that could be acting as virulence factors. Hence, we identified protein sequences against Pathogen-Host Interactions database (PHI-base) that is an invaluable resource in the discovery of virulence and effector genes (The Pathogen-Host Interactions database (PHI-base, http://www.phi-base.org/index.jsp): additions and future developments). With few exceptions, nearly all the secreted proteins got positive results in PHI-base, and the corresponding PHI ID is shown in Table S1, Supplementary Materials. These candidate virulence proteins merit further investigation. mainly categorized into the single-organism process, protein metabolic process, metabolic process, regulation of cellular process, and others ( Figure 1a). Meanwhile, the secreted proteins with higher accumulation in the +BP sample were mainly categorized into metabolic process, nitrogen compound metabolic process, cell wall organization or biogenesis, response to a stimulus, oxidation-reduction process, and others (Figure 1b). The predicted secretome of F. proliferatum primarily contained proteins of unknown function and uncharacterized proteins (Figure 1, Tables 1 and 2). Of particular interest are proteins that could be acting as virulence factors. Hence, we identified protein sequences against Pathogen-Host Interactions database (PHI-base) that is an invaluable resource in the discovery of virulence and effector genes (The Pathogen-Host Interactions database (PHI-base, http://www.phi-base.org/index.jsp): additions and future developments). With few exceptions, nearly all the secreted proteins got positive results in PHI-base, and the corresponding PHI ID is shown in Table S1, Supplementary Materials. These candidate virulence proteins merit further investigation.

Gene Expression in Planta
To determine whether genes encoding the secreted proteins identified in liquid media were also expressed in planta, banana fruit were inoculated with the fungus. At 4 dpi, F. proliferatum infected

Fumonisin Production
To further analyze whether banana peel amendment affected the fumonisin production, a 3-day culture was used for the analysis of the fumonisin production. As shown in Figure 3, the concentrations of FB1 and FB2 in +BP sample were much lower than those in -BP sample.

Fumonisin Production
To further analyze whether banana peel amendment affected the fumonisin production, a 3-day culture was used for the analysis of the fumonisin production. As shown in Figure 3, the concentrations of FB1 and FB2 in +BP sample were much lower than those in -BP sample.

Discussion
Many plant pathogens can modulate their secretomes in response to their plant hosts [7,[31][32][33]. In particular, fungal pathogens of plants adapt to the host environment through the secretion of proteins and other molecules to facilitate nutrient acquisition and overcome the immune response. Proteomics techniques have been widely used to identify many extracellular proteins [34,35]. F.

Discussion
Many plant pathogens can modulate their secretomes in response to their plant hosts [7,[31][32][33]. In particular, fungal pathogens of plants adapt to the host environment through the secretion of proteins and other molecules to facilitate nutrient acquisition and overcome the immune response. Proteomics techniques have been widely used to identify many extracellular proteins [34,35]. F. proliferatum causes great loss of crops, fruit, and vegetables worldwide. This study was aimed to investigate the infection mechanism of F. proliferatum on banana fruit based on the secretome change in response to banana using label-free quantitative proteomic technology.
The exploration of phytopathogen secretomes has been mainly achieved through functional proteomics analyses. Due to the challenge to identify fungal proteins in planta, Yang et al. identified the secreted proteome of F. graminearum cultured with barley or wheat flour to represent the natural hosts of the fungus [9]. In this study, to further mimic the natural infection environment of the fungus in planta, F. proliferatum was cultured in medium supplemented with or without banana peel (+BP/-BP). The banana peel closely represents the natural hosts of the fungus in contrast with previous studies of the F. proliferatum in vitro secretome [20,21]. This approach, combined with label-free quantitate proteomic technology, which is ideally suited for a gel-free shotgun analysis [36], can identify many fungal secreted proteins. In the present study, we identified 40 highly abundant proteins only observed in the secretome of +BP sample (Table 1). Meanwhile, we also identified 65 proteins observed with up-regulation in the secretome of +BP sample ( Table 2). The detection of all the selected transcripts in banana peel (Figure 2) has confirmed the relevance of the approach for the identification of fungal proteins involved in the interaction with the host. Therefore, this study increased the value of the information gained from the analysis of F. proliferatum secretome in vitro and provided an ideal model to understand plant-pathogen interactions.
Secretory pathways in fungi are complex and involved with the classical secretory pathway and alternate routes for protein trafficking. For the classical secretory pathway, signal peptides are added to target translated proteins outside the cell or to an organelle. In contrast, proteins with no signal peptides can be secreted via a nonclassical secretory pathway (on-conventional protein secretion in yeast). Proteins can also be secreted through other unconventional pathways, such as secretory lysosomes, exocytosis, microvesicles or ATP-binding cassette transporters [37]. Indeed, 35 of the proteins identified in the present study, including GAPDH, malate dehydrogenase and FBA1-fructose-bisphosphate aldolase, and 26S proteasome regulatory subunit RPN8, etc. were predicted to be secreted in a nonclassical way using SecretomeP (Tables 1 and 2). Previous research also showed that several housekeeping enzymes, including GAPDH, can be secreted extracellularly by pathogens and serve as virulence factors [38,39]. Usually, ribosomal proteins in the secretome would likely be related to contaminants of the secretome when cell breakage occurs [3]. However, in the present study, we identified one 60S ribosomal protein L8 and one ribosomal protein L7a. After a BLASTp search of these protein sequences against Fungal Secretome Database (http://fsd.riceblast.snu.ac.kr/index.php? a=view), we indeed got positive results with 60S ribosomal protein from Histoplasma capsulatum (87.4% identity) and 60S ribosomal protein L8 from Blastomyces dermatitidis (82.06% identity), respectively. Mulugeta et al. also found ribosomal proteins in the secretome of Staphylococcus carnosus, and they pointed that the appearance of these proteins referred to as "nonclassical protein excretion" [40], which has also been observed in other pathogens. Importantly, Lu et al. showed that ribosomal proteins play essential roles in C. albicans virulence [41]. Thus, the exact role of ribosomal proteins in the infection ability of F. proliferatum still need further investigation.
PHI-base is a web-accessible database that catalogues experimentally verified pathogenicity, virulence and effector genes from bacterial, fungal and protist pathogens [42]. Analysis of secreted proteins against PHI-base indicated that proteins belonging to pathogenicity-related class were abundant in secretome in response to banana peel (Table S1, Supplementary Materials). All of them were either exclusively secreted or increased to several folds in abundance in response to banana peel tissue (Tables 1 and 2). Highly represented proteins in this class included CAP20-virulence factor, lipase, SnodProt1, associated with virulence of fungus on plant host [43][44][45][46]. Other proteins previously associated with pathogenicity of fungus included CWDEs (beta-glucosidase, glucan 1,3-beta-glucosidase, endo-1,3-beta-glucanase and pectinesterase), proteases (aspartic proteinase OPSB, subtilisin-like serine protease), endochitinase 2, all of which have well documented roles in plant pathology as enzymes that either degrade the cell wall to permit access to the host or neutralize the host defenses [47][48][49]. Additionally, the results of the in planta expression profiling indicated that these genes had different expression patterns at different infection stages (Figure 2). For example, genes encoding glucan endo-1, 3-beta-glucanase, pectinesterase and aspartic proteinase OPSB showed higher expression level at early infection stage, suggesting they might mainly aid F. proliferatum at early time points in penetration and colonization of banana peel. Barnabas et al. also reported that fungus could secrete and regulate the expression of some secretory proteins at distinct stages of infection [6].
Proteins involved in nutrition absorption were also present and included SUC2-invertase, acid phosphatases and they were both involved the successful colonies of the fungus through deriving nutrients from their host [50]. Some proteins involved in the oxidation-reduction process, nitrogen compound metabolic process, carbon use process was also identified in the secretomes in the present study ( Figure 1, Tables 1 and 2). As reported by previous research, nitrogen metabolism, carbohydrate metabolism and oxidoreductase reactions are beneficial for the fungus to cope up and adapt to different nutritional environments during infection stages [6].
The analysis of the secretome also revealed 13 proteins with NLS (Tables 1 and 2). NLS was reported to predict the in planta localizations of fungal proteins [51]. Previous research has reported that some various bacterial, oomycete and fungal effectors targeted in host nucleus [52][53][54]. However, no F. proliferatum effectors have been found to target the host nucleus and trigger plant immunity. Zhang et al. also identified a novel secreted protein VdSCP7 that targets the plant nucleus and modulate plant immunity through analyzing the secretomes of V. dahlia [22]. Indeed, we identified two transcription regulation factor (transcription factor BTF3a and transcriptional repressor rco-1), one zuotin as well as one rAsp f 9 allergens, which were nuclear localized effector (Tables 1 and 2). Rampitsch et al. also identified allergen proteins in the secretomes of F. graminearum [3]. Furthermore, an allergen Asp F4-like protein was the causal agent of stem rot in corn [55]. So far, the role of a transcription factor in pathogenesis mainly focused on fungus itself [56,57]. No transcription factor has directly been secreted extracellular to regulate plant immunity. The secreted proteome in this study could anticipate their role in pathogen-host interactions. Our present results cannot shed light on the function of these proteins, but these proteins are indeed widely secreted by F. proliferatum in response to the plant host. Therefore, these are good candidates for further analysis of their exact role in pathogenicity.
Mycotoxin is also an important secreted molecule of fungus. Fumonisins are a group of mycotoxins mainly derived from Fusarium, Liseola section including F. proliferatum [58]. Whether the production of fumonisin facilitates the pathogenesis of fungus remains an open issue [59]. Some research indicated that fumonisins appeared to be involved in disease development [60], while other research indicated that deletion of fumonisin production has no effect on the aggressiveness of fungus [61]. In this study, lower fumonisins (FB1 and FB2) content were identified in +CB culture (Figure 3). Our results suggested that the initiation of infection might not be caused by FB production but other pathogenicity factors. Desjardins et al. also indicated that the role of fumonisins depends on complex environmental and genetic contexts in host-pathogen interaction [62].

Conclusions
Our current results provided a comprehensive, comparable proteomic analysis of the secreted proteins of F. proliferatum induced by the interaction with the banana host. Our results indicated that F. proliferatum could modulate itself with a set of extracellular proteins that prepares it for encountering and infection of the host plant. Our data suggested a subset of the secreted proteins whose presence might be required to initiate infection of F. proliferatum and provided a foundation for future investigation of virulence factors. Gene expression in planta confirmed their secretion and variable roles in different infection stages. In sum, this study provided valuable insight into secretory capacity and pathogenicity of F. proliferatum as well as the molecular interactions between fungi and the plant host.
Supplementary Materials: The following are available online at http://www.mdpi.com/2218-273X/9/6/246/s1, Figure S1: The evaluation results of the identification and quantitation of proteins in F. proliferatum cultured with and without banana peel; Figure S2: Visual picture of Fusarium proliferatum infection on banana fruit; Table S1: Detailed information for identified secreted proteins in this study; Table S2: Details for protein LFQ parameters; Table S3: Primers used for qRT-PCR in this study.