Introduction

Phytophthora capsici was first reported by Leon H. Leonian in 19221, and is currently regarded as one of the 10 most important oomycete pathogens in molecular plant pathology2. This devastating pathogen has a global distribution and can infect more than 45 species of plants including both crops and weed species3,4,5,6, causing crown, root, and fruit rot7,8,9, which lead to significant economic losses every year2,5,10. Although, crop rotation and other management tools contribute to the control of diseases caused by P. capsici, in practice there is a heavy reliance on fungicides11,12. The carboxylic acid amide (CAA) fungicide flumorph, 4-[3-(3,4-dimethoxyphenyl)-3-(4-fluorophenyl)-1-oxo-2-propenyl] morpholine, which was developed by the Shenyang Research Institute of Chemical Industry of China in 199413, has been patented in China (ZL.96115551.5), the United States (US6020332), and Europe (0860438B1). It is currently registered for the control P. capsici, Phytophthora infestans, Pseudoperonospora cubensis, and Plasmopara viticola in China, and remains an effective fungicide to control diseases caused by P. capsici14.

The sexual reproduction of P. capsici15 plays an important role in its disease cycle initiating infection in host plants5,16, while the resulting genetic recombination can contribute to the development of isolates that exhibit complete insensitivity to certain fungicides17. P. capsici is a heterothallic pathogen that produces two mating types, A1 and A2, and it has been shown that the co-occurrence of both mating types in regions of the United States, South Africa, and the northern provinces of China, can facilitate frequent outcrossing and increase the risk of resistance developing18,19,20,21. Previous studies have shown that flumorph resistance in P. capsici is controlled by two dominant genes, which implies that once resistance has developed it could rapidly spread through a population via both sexual and asexual reproduction22.

Proteomics has become a useful tool for studying the biological effects of fungicides. For example, 2-DE has been used to investigate the global response of Saccharomyces cerevisiae in the early stages of exposure to mancozeb23, while MALDI-TOF-MS/MS has been used to study the mode of action of the fungicide JS399-19 in Fusarium graminearum24, and iTRAQ (isobaric tags for relative and absolute quantitation) technology to study the effect of pyrimorph in P. capsici25. The current study adopted a similar approach, using iTRAQ to compare the response of a wild-type parental P. capsici isolate (PCAS1) and its flumorph resistant sexual progeny (S2-838). The proteomics data produced would hopefully provide a greater understanding of the adaptive mechanisms associated with flumorph resistance in P. capsici, as well as highlighting target proteins for not only the early monitoring of flumorph resistance, especially that associated with sexual reproduction, but also for the design of novel fungicides.

Results

Overview of quantitative proteomics analysis

A total of 2396 individual proteins with at least one unique peptide and protein scores >20 were identified from the wild-type (PCAS1) and flumorph-resistant (S2-838) isolates of P. capsici cultured in the presence or absence of flumorph (1.5 μg/ml or 100 μg/ml, respectively) using iTRAQ-LC-MS/MS analysis (identified protein and peptide information, Supplementary Table S1, S2).

Effect of flumorph on protein levels

In total, 189 and 26 proteins were found to be significantly altered in PCAS1 and S2-838, respectively (Supplementary Table S3, S4). Of the 189 proteins detected in the wild-type isolate PCAS1, a total of 80 were up-regulated, and the other 109 down-regulated. In contrast, the flumorph-resistant isolate S2-838 was much less affected with only 21 up-regulated proteins and 5 down-regulated ones.

Identification of candidate proteins for the adaptive response of P. capsici to flumorph

It was found that 181 proteins were associated with the adaptive response of P. capsici to flumorph, with altered levels of abundance in the wild-type isolate PCAS1, but not in the flumorph-resistant isolate S2-838, when comparing the control cultures to those treated with flumorph (Table 1). The subsequent GO analysis categorized these proteins into 14 functional groups according to their biological activity (Fig. 1). The majority of the proteins fell into just two categories metabolic process (83) and cellular process (54). The other proteins fell into 12 categories including developmental process, cellular component biogenesis, cellular component organization, death, pigmentation, localization, response to stimulus, multicellular organismal process, growth, multi/-organism process, establishment of localization, and biological regulation. However, it should be noted that a single protein can be assigned to more than one category. Metabolic pathway enrichment analysis was then performed by matching the proteins with altered abundance to annotated proteins in the KEGG Pathway database. Although it was not possible to classify a large number of the proteins (51), the majority were assigned to a diverse range of metabolic pathways, including amino acid metabolism, carbohydrate metabolism, energy metabolism, lipid metabolism, nucleobase-containing compound metabolism, response to stimulus, transport, and other metabolic pathway (Fig. 2).

Table 1 Candidate proteins identified by iTRAQ analysis for the adaptive response of P. capsici to flumorph.
Full size table
Figure 1: GO annotation of candidate proteins associated with the adaptive response of P. capsici to flumorph.
figure 1

Numbers indicate the number of proteins categorized into each functional group.

Full size image
Figure 2: Distribution of candidate proteins associated with the adaptive response of P. capsici to flumorph as categorized by KEGG pathway analysis.
figure 2

Numbers indicate the number of proteins in each category.

Full size image

Discussion

An iTRAQ-LC-MS/MS approach was used to investigate the effect of the CAA fungicide flumorph on wild-type and resistant isolates of P. capsici. Altogether, 2406 individual proteins were identified, a number of 189 and 26 proteins were found to have altered levels of abundance in response to flumorph stress in PCAS1 and S2-838, respectively. One reason for the big difference in the number of differentially expressed proteins between the isolates PCAS1 and S2-838 can be contributed to the different genetic background. Compared to the wild-type isolate, the point mutations in cellulose synthase 3 caused the resistance to CAA fungicides in the mutant25,26. The 181 proteins related with genetic background of flumorph resistance were identified as candidates for the adaptive response of P. capsici to flumorph. The subsequent GO analysis categorized the proteins into 14 biological processes. However, KEGG pathway analysis indicated that 51 of the proteins have yet to be assigned metabolic pathways and therefore provide little insight into the effect of flumorph, although they could be utilized as candidate proteins for future study. The roles of the remaining 130 proteins were discussed below.

Carbohydrate metabolism was the pathway most affected by flumorph and was associated with the altered abundance of 46 proteins, of which 10 and 36 were up-regulated and down-regulated in response to flumorph, respectively. Two of the up-regulated proteins, Glucan-1,3-beta-glucosidase (Accession number: 262097763, and 262098611) and exo-1,3-beta-glucanase are involved in the break down glucan to release glucose27,28. Previous investigations into the mode of action of CAA fungicides have revealed that mandipropamid and pyrimorph can inhibit cell wall biosynthesis in P. infestans and P. capsici, respectively25,26. Given that the cell walls of oomycetes mainly consist of cellulose and 1,3-β-glucans29, the increased abundance of glucan-1,3-beta-glucosidase and exo-1,3-beta-glucanase suggested that P. capsici utilized carbohydrate from the host environment to compensate for the cell wall stress induced by flumorph. Similarly, the increased abundance of PHYSODRAFT_261542, PHYSODRAFT_302104, and PHYSODRAFT_565653, which are involved in glucan biosynthesis30, could also represent an adaptation to the cell wall stress induced by flumorph. The most down-regulated proteins were associated with glycolysis and the citric acid (TCA) cycle, which are involved in the utilization of glucose and other carbohydrates to generate ATP. However, interestingly, all the proteins involved in lipid metabolism, which can also result in the production of large amounts of energy, were up-regulated in response to flumorph. The altered level of these flumorph-responsive proteins suggested that flumorph might induce a redistribution of the metabolic processes associated with energy production. This hypothesized shift of energy generation from glycolysis and the citrate cycle to lipid metabolism could allow for the redistribution of glucose or carbohydrate in response to the inhibition of cell wall biosynthesis caused by flumorph.

Although the majority of the proteins with altered levels of abundance were associated with energy metabolism, a significant number of proteins (9 up-regulated, 20 down-regulated) were associated with amino acid metabolism. Several of the down-regulated proteins were found to play a role in the biosynthesis of amino acids and proteins including argininosuccinate lyase31 and the hypothetical protein PHYSODRAFT_48539930, which are involved in arginine biosynthesis; hypothetical protein PHYSODRAFT_47171330 and 5-methlytetrahydropteroyltriglutamate-homocysteine methyltransferease32, which play a role in methionine biosynthesis; and cysteine synthase, which participates in cysteine biosynthesis33; as well as the eukaryotic translation initiation factor 334, all of which might play important roles in protein synthesis. It is therefore possible that a reduced rate of global protein synthesis could be another adaptive response of P. capsici to flumorph stress as the pathogen attempts to maintain the fidelity of its protein biosynthesis. Similar results have also been observed in the response of P. capsici to another CAA fungicide, pyrimorph25.

It was also interesting that some proteins with altered levels of abundance were associated with response to stimulus and signal transduction. For example, alkaline phosphatase35 was up-regulated in response to flumorph. This important hydrolase enzyme catalyzes dephosphorylation during the post-translational modification of proteins36. Dephosphorylation and phosphorylation of S, T, Y and H residues are the best characterized modifications associated with the reversible, activation and inactivation of enzyme activity and the modulation of molecular interactions in signaling pathways37. It was also found that elicitin protein was up-regulated. Elicitin superfamily of proteins are structurally related to extracellular proteins that induce hypersensitive cell death and other biochemical changes associated with the defense response38,39,40. The up-regulation of alkaline phosphatase and elicitin in P. capsici therefore suggested that signal transduction was an important factor in responding to flumorph stress.

It was also found that a protein ABCA1 (from lipid exporter family) associated with transmembrane transport, an ATP binding cassette A (ABCA) superfamily protein was up-regulated in P. capsici in response to flumorph. Members of the ABCA family proteins have also been implicated in the adaptation to environmental changes in the free-living state of Phytophthora sojae41. The activation of this transporter could confer significant selective advantage to P. capsici isolates responding to flumorph stress in their environment.

Taken together, these results indicated that the adaptive response of P. capsici to flumorph was complex and regulated by multiple pathways, including utilising carbohydrate from the host environment to compensate for the cell wall stress induced by flumorph, a shift in energy generation from glycolysis and citrate cycle to lipid metabolism, decreased amino acids biosynthesis, and elevated levels of proteins associated with the pathogen’s response to stimulus and transmembrane transport. The proteomic data produced in the current study could provide important insight into the adaptive response of P. capsici to flumorph that would be useful for monitoring the emergence of resistance in field populations.

Material and Methods

Strains, medium, and growth conditions

The wild-type P. capsici isolate PCAS1 (P1314, mating type A1), which was originally collected from diseased green pepper (Capsicum annuum L.), was kindly provided by Professor Michael Coffey (University of California, Riverside, USA), while the sexual progeny S2-838 was generated in a previous study22. The EC50 values (the effective concentration for 50% inhibition of mycelial growth) for flumorph in the two isolates was approximately 1.5 μg/ml and 100 μg/ml (the maximal soluble concentration of flumorph), respectively (Supplementary Fig. S1). Potato dextrose agar (PDA) or potato dextrose broth (PDB) was used for routine maintenance of the cultures, which were dark incubated at 25 °C.

Sample preparation for proteomic analysis

Mycelium collected from 4-day cultures growing on PDA medium with a cellophane sheet were harvested and used to inoculate PDB medium in the presence or absence of flumorph (1.5 μg/ml or 100 μg/ml for the wild-type and resistant isolates, respectively). Each treatment has 10 biological replicates. After 24 hours dark-incubation at 25 °C with shaking, the mycelia were collected by filtration, washed profusely with sterile distilled water, dried, and ground thoroughly in liquid nitrogen. The resulting samples were stored at −80 °C until required. The experiment was repeated three times. The protein was extracted from approximately 100 μg of each sample, which were resuspended in 1 ml lysis buffer [8 M urea, 30 mM HEPES, 5 mM TCEP, and 2 mM EDTA] with the aid of a sonicator (Branson® Sonifier 250, BRANSON Ultrasonics Corporation, Danbury, U. S. A.). Any undisrupted cells were removed by centrifugation with the supernatant being transferred to fresh tubes. The samples were then incubated at 60 °C for one hour before the addition of 1% iodoethanol for another one hour in dark. The proteins were precipitated overnight in a freezer using 4 volumes of cold acetone, before being collected by centrifugation at 20000 rpm for 20 min. Finally, the proteins were dissolved in 50 mM triethylammonium bicarbonate (TEAB) containing 1% SDC.

Isobaric Tags for Relative and Absolute Quantitation (iTRAQ) and Liquid Chromatography Coupled with Tandem Mass Spectrometry (LC−MS/MS) Analysis

Since technical variation of iTRAQ measurements was demonstrated to be on the order of 20%42, pooling samples were used to produce such biases43,44. In our study, ten biological replicates for each sample were pooled together to produce one sample, and 100 μg aliquots digested with 1 μg/μl trypsin overnight at 37 °C. After being lyophilized the samples belonging to individual treatments were labeled with different iTRAQ reagents (Applied Biosystems, Foster City, CA.) according to the protocol of the manufacturer. The untreated PCAS1 control was labeled with 114, while the PCAS1 treated with flumorph, S2-838 control and S2-838 treated with flumorph were labeled with 115, 116 and 117, respectively. The labeled peptides were then combined and dried in a vacuum concentrator. The first dimension of the 2D-LC consisted of extensive fractionation of the peptide mixtures by strong cation exchange (SCX) chromatography to improve proteome coverage. Briefly, the dried samples were reconstituted in buffer A [25% (v/v) acetonitrile (ACN), 10 mM potassium phosphate; pH adjusted to 3.0] and loaded onto a Lumn A column (4.6-mm i.d. × 100-mm length, 5 μm, 100 Å; Phenomenex, USA). The column was equilibrated for 10 min in buffer A before the peptides were eluted at a flow rate of 1 mL/min using buffer B [25% (v/v) ACN, 2 M potassium chloride, 10 mM potassium phosphate; pH adjusted to 3] at a succession of increasing gradients 0–30% for 15 min, followed by 30–100% for 15 min, and finally 100% buffer B for 10 min. A total of fifteen peptide fractions were collected, which were then dried using a SpeedVac centrifugal vacuum concentrator and purified on a strata-X C18 column (Phenomenex, USA) prior to mass spectrometry (MS) analysis.

The LC-MS/MS experiments were performed using an integrated system consisting of a Q Exactive™ Mass Spectrometer (Thermo Fisher Scientific, USA) coupled with a nanoflow HPLC system (Easy nLC, Proxeon Biosystems, now Thermo Fisher Scientific, USA). Each fraction was reconstituted in 0.1% formic acid before being injected into the LC-MS/MS system. The samples eluted from the trap column were separated on a PepMap C18 column (100 mm × 75 mm, 300-Å pore size, 5 μm particle size, Thermo scientific, USA) at a rate of 400 nL/min, using 0.1% formic acid as solvent A and 0.1% formic acid in acetonitrile as solvent B, in increasing gradients: 0.1–5% B (0–10 min), 5–30% B (10–40 min), 30–60% B (40–45 min), 60–80% B (45–48 min), 80% B (48–55 min), 80–0.1% B (55–65 min). The eluting peptides were sprayed into the mass spectrometer at an ion spray voltage of 1800 eV, and their MS/MS spectra acquired using automated data-directed switching between the MS and MS/MS modes. The five most abundant signals from each survey scan (350–2000 m/z range) were selected by charge state, and the collision energy applied accordingly for the sequential MS/MS fragmentation scanning as described previously45. The entire experiment was conducted three times.

Data Processing and Analysis

The raw MS/MS data were merged and transformed using the Proteome Discoverer software package (version 1.3; Thermo Fisher Scientific, USA)46 before Mascot version 2.3.01 (Matrix Sciences, Ltd., London, UK) was used to identify and quantify the individual proteins according to sequences contained in the NCBI Oomycetes database using the following settings: trypsin specific digestion with one missed cleavage allowed, peptide tolerance of 15 ppm, MS/MS tolerance of 20 mmu, iTRAQ 4-plex for peptide N-t and Lys as fixed modifications, and in variable mode, iTRAQ 4-plex on Tyr, oxidized Met and methylthio on Cys. The false positive rate, which was checked using a concatenated target-decoy database search strategy, was set to be less than 1%. Only proteins with at least one unique peptide and having protein scores of more than 20 were initially recorded. Only proteins with two or more peptides were used for the quantitative analysis. The LIBRA tool from the TPP software47 was used for protein quantification using the default parameters. The relative abundance of proteins in the different treatments were calculated from three replicates using the log2 of the iTRAQ ratios, which were normalized before the standard deviations from the corresponding normal distributions of ratios were used to determine the cutoff point of the experiment48. Proteins whose average ratios fell outside a standard deviation of ±1 from the global mean were considered to have differential abundance. Gene Ontology (GO) annotation was conducted using information retrieved from the UniProt and BGI WEGO (http://wego.genomics.org.cn) databases49, while the pathway enrichment analysis was conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway database50.

Additional Information

How to cite this article: Pang, Z. et al. Insights into the adaptive response of the plant-pathogenic oomycete Phytophthora capsici to the fungicide flumorph. Sci. Rep. 6, 24103; doi: 10.1038/srep24103 (2016).