Functional and Structural Analysis of a Novel Acyltransferase from Pathogenic Phytophthora melonis.

This investigation characterizes an acyltransferase enzyme responsible for the pathogenicity of Phytophthora melonis. The protein was characterized in vitro for its physicochemical properties. The biochemical characterization, including thermal and pH stability, revealed the 35 °C temperature and 7.0 pH as the optimum conditions for the enzyme. Applying the Tween-80 solution enhanced the activity up to 124.9%. Comprehensive structural annotation revealed two domains, A (ranging from residues 260 to 620) and B (ranging from 141 to 219). Domain A had transglutaminase (T-Gase) elicitor properties, while B possessed antifreeze features. Rigorous sequence characterization of the acyltransferase tagged it as a low-temperature-resistant protein. Further, the taxonomic distribution analysis of the protein highlighted three genera in Oomycetes, i.e., Pythium, Phytophthora, and Plasmopara, bearing this protein. However, some taxonomic groups other than Oomycetes (i.e., archaea and bacteria) also contained the protein. Functional studies of structurally analogous proteins spanned 10 different taxonomic groups. These revealed TGase elicitors (10%), phytopathogen effector proteins RxLR (4%), transporter family proteins (3%), and endonucleases (1%). Other analogues having one percent of their individual share were HIV tat-specific factor 1, protocadherin fat 4, transcription factor 1, and 3-hydroxyisobutyrate dehydrogenase. Because the plant infection by P. melonis is a complex process regulated by a profusion of extracellular signals secreted by both host plants and the pathogen, this study will be of help in interpreting the cross-talk in the host-pathogen system.

Significance Statement

Acyltransferases are imperative for various cellular processes of the organisms. They are responsible for the precise interconversion of biochemicals and efficient cell secretory systems. We found a novel acyltransferase associated with pathogenic Phytophthora melonis. Because there is very little known about the cucumber–Phytophthora interactions, this study attracts tremendous attention from the experts working on microbial pathogenicity mechanisms. This research explained the functional and structural characteristics of the pathogenic acyltransferase and also elucidated its taxonomic distribution among living organisms. Overall, the study adds valuable information about the molecular kinetics and supramolecular characterization of the enzyme.

Introduction

Transferases are a class of enzymes that transfer specific functional groups from one donor molecule to an acceptor. This implicates them in hundreds of different biochemical pathways and integrates them with some of the most important processes in living cells. They take part in a myriad of cellular processes and responses, including translation.[1] Some members of the class adhere to an acylation event directed by a specialized acyltransferase structure and are sometimes also classified as the acyl carrier protein (ACP). They drive the transfer of fatty acyl moieties from one molecule to another to catalyze acylation events.[2,3] Acylation is essential to the entire family of pore-forming microbes producing or activating toxins.[4] It is also a critical step in the biosynthesis of triacylglycerol (an important component of biological membranes and a source of energy). Acyltransferase catalyzes the production of triacylglycerol and is a reason for pathogenicity in Colletotrichum gloeosporioides.[5]

Phytophthora melonis is a devastating plant pathogen and attacks a wide range of important plants (agricultural and ornamental), including cucumber, Cucumis sativus.[6] The disease is progressing each year in the damp and humid environment of Guangzhou and costs millions of dollars to commercial cucumber production.[7] The pathogen is responsible for a disease called cucumber blight, which not only affects the yield (up to 80%) but also the quality of the produce, severely declining its market value. Because the pathogen already has a wide host range along with the tendency to infect more crops, it poses a biosecurity threat to the local agricultural system.[7,8] An acyltransferase (H3GZF6) was detected as a pathogenicity protein in our research project funded by Dean Funds of the Guangdong Academy of Agricultural Sciences (201818B), aimed to discover the pathogenicity factors of P. melonis. The protein was also the cause of sudden oak death.[9] Severino et al.[10] reported the identical protein species as a pathogenicity tool of Phytophthora infestans, a causal agent of the Irish potato famine. The pathogenicity protein also persists in Phytophthora plurivora to cause diseases in plants.[10] Some reports have already been published about different transferase proteins in plant cells.[11,12] However, the involvement of an acyltransferase in the pathogenicity of P. melonis was a new observation, and its submolecular characterization could add some novel information to scientific knowledge. The protein has no previously reported information about its physicochemical properties/characterization. Therefore, we studied the structure, and physicochemical features of the enzyme. This study reports the first detailed characterization of the acyltransferase from P. melonis. Because the plant infection by P. melonis is a complex process coordinated by a plethora of extracellular signals secreted by both host plants and the pathogen, this study will be of help in understanding the cross-talk between the host–pathogen system.

Results and Discussion

Enzyme Activity

Temperature studies of the enzyme activity showed the maximum value at 40 °C. The enzyme retained a relative activity of 14.6% at 60 °C. However, the trend line predicted the seized enzyme activity at 65 °C (Figure A). The maximum stability in the enzyme activity was at pH 7.0, declining to a complete inactivation at pH 2 and 10 (Figure B). We recorded an acyltransferase activity of 3.2 NBD-TG μM·min–1·g–1 of the protein. The acyltransferase activity demonstrated the control value (100 units) to calculate the relative activity of the enzyme under different physical conditions. Tween-80 was the only chemical that enhanced the enzyme activity (124.9%) in comparison to the control treatment, while Na+, H2O2, and Ca2+ could maintain the activity up to 87.9, 81.9, and 81.2%, respectively. Among the representative biochemicals of metal salts, organic acids, and detergents, Zn+2 and Fe+2 suppressed the activity of the enzyme to zero (Figure C).

Figure 1

Evaluation of the enzyme activity against temperature gradient ranging from 0 to 80 °C. Temperature is plotted on the X-axis, relative activity is plotted on the Y-axis, and absolute numeric values of the enzymatic activity are labeled at respective data points on the activity curve (A). Activity retention of the enzyme acyltransferase at pH gradient 2–10. The X-axis shows pH gradient values, the Y-axis represents relative activity, and absolute numeric values of the enzymatic activity are labeled at each data point on the activity curve (B). Acyltransferase activity assay under the influence of various biochemicals representing metal salts, organic acids, and detergents. The enzyme activity was measured in terms of the formation of NBD-TG μM·min–1·g–1 of protein (C). Data labels show the numeric values of acyltransferase activity at each data point in terms of NBD-TG μM·min–1·g–1 protein.

Structural Analysis

The three-dimensional (3D) structure of the protein had two distinct lobes and a total of 775 residues (Figure A). Four residues, i.e., 538 (I), 432 (V), 558 (T), and 593 (F), took part in joining the two lobes of the acyltransferase (Figure B). There were two domains in the protein structure, designed on 3tw5.1.A and 4dt5.1.A. standards. The first domain had a ligand type 1 × CXS acidic (3-cyclohexyl-1-propylsulfonic acidic). However, detailed analysis revealed a total number of 132 active sites in the structure, 217 sequence repeats, and 76 binding sites (Supplementary Data Set 1, see Tables S1–S4).

Figure 2

Standard 3D structure of the protein H3GZF6 (A). Rotating structure of bilobed acetyltransferase with abutting residues I, V, T, and F. The amplified section is showing the adjoining residues (i.e., 538 (I), 432 (V), 558 (T), 593 (F)) of the two distinct lobes of the enzyme (B).

TGase elicitor is a family of oomycete proteins that elicit transglutaminase/acyltransferase activity and enable the organism to cause infections in plants. The enzyme accession number is E. C:2.3.2.13. Considering the enzyme presence in Vibrio spp., some studies have proposed a lateral gene transfer event that occurred between bacteria and oomycetes.[13] Transglutaminase elicitor is one of the members of a protein family containing 770 protein species. The protein belongs to the architectural domain along with the other 32 proteins and has a short name, TGase elicitor. Based on primary structure and kinetic properties, acyltransferases are classified into several groups, i.e., α, theta, mu, sigma, pi, zeta, kappa, and omega.[14] They may vary in their molecular weight, i.e., 23.0–28.0 kDa, but the presence of the two binding domains (a GSH binding site-G-site and a substrate-binding site-H-site) is a common feature for the enzyme.[11,15] In the current investigation, we detected two similar binding domains of the oomycete transferase, proving its nomenclature as “acyltransferase”.

Residues 260–620 structured “domain A”, and its local identity with the standard structure template was ≥0.8 for 97% of residues (Figure A). The energy distribution score was over 0.7 for domain A (Figure B). Residues 141–219 constructed the “domain B” in which the local similarity score of 86% residues was ≥0.4 (Figure C). The energy distribution Q mean value was 0.8 for domain B (Figure D).

Figure 3

Local similarity chart of “domain A (260–620)” with the standard template of the protein. Residues have been plotted on the X-axis, while the Y-axis shows local similarity with the target residue (A). Energy distribution chart for domain A. The red star highlights the position of domain A among the energy distribution matrix (B). Local similarity chart of “domain B (141–219)” with the standard template of the protein. Residues have been plotted on the X-axis, while the Y-axis shows local similarity with the target residue (C). Energy distribution chart for domain B. The red star highlights the position of domain B among the energy distribution matrix (D).

The acyltransferase activity of the protein species was more than the activities reported before,[16] and its catalytic efficiency was even increased by up to 124.92% in the Tween-80 solution. Tween-80 is a surfactant, and enzyme activities are significantly influenced by surfactants in the environment.[17] Most of the enzymes working in liquid/liquid interfaces are positively influenced by the addition of surfactants.[18] In addition, most of the plant enzymes, including cellulases, amylases, proteases, and carboxymethyl cellulases, are triggered in the presence of Tween-60.[19] A similar phenomenon could be involved in the enhanced activities of acetyltransferase under the Tween-80 application. Acyltransferases regulate the responses against external and cellular toxins, including those induced by oxidative stress.[12,20] They play important protection roles against oxidative stress induced by chilling environments.[16] In the current investigation, the sequence 141–219 was also found similar to the antifreeze protein, 4dt5.1.A, and may play a role during temperature stress. All of these factors show the P. melonis is tolerant against chilling stress and can cause pathogenicity at temperatures close to the freezing temperature, 2 °C.[21,22] However, the trend line showed that the optimum temperature for acyltransferase was 35 °C.

The protein sequence was carrying a transglutaminase (TGase) elicitor, ranging from residues 266–624, and had reference identities IPR032048 and PF16683. The particulars recorded for TGase were Uniprot entry (Q01928_PHYSO) and PDB ID: 3TW5 (Supplementary Data Set 1, see Table S5). The hidden Markov model logo of the TGase domain showed six tryptophan residues at a maximum deviation of 2.58% among all of the amino acids. However, there were four cysteine residues, which showed a variability of 2.15%. Five tyrosine residues were at 1.39% variation (Supplementary Data Set 1, see Figure S1). The overall view of the sequence depicted that the TGase elicitor started from residue 169 and ended at 527, with an E-value of 4.4e-182 and a domain range of 5.4e-182 (Table ).

Table 1

Particulars of Transglutaminase Elicitor in H3GZF6

			gathering threshold (bits)		score (bits)		E-value
domain	start	end	sequence	domain	sequence	domain	sequence	domain
TGase elicitor	169	527	25.00	25.00	617.30	617.00	4.4e-182	5.4e-182

This study describes the optimum physical conditions for the enzyme. A trend was noticed that a low temperature slowed down the enzyme activity.[23,24] Similarly, the optimum pH for the enzyme was found at 7.0. Agarwal and Choudhury[25] investigated two similar transferases of microbial origin and concluded their tentative use in the synthesis of organic acids (e.g., hydroxamic acid). Previously, it was a common notion that the fatty acids transferred by the CoA-independent system were restricted to C20 and C22 polyunsaturated fatty acids.[26] The reaction esterifies the 2-position of diradyl phospholipids, especially diacylglycerophosphocholine (diacyl-GPC). However, the CoA-independent transacylation system catalyzes the transfer of C20 and C22 polyunsaturated fatty acids either from exogenously added phospholipids or from endogenous membrane phospholipids.[27] Considering the of membrane phospholipids engenders a hypothesis that could be used to explain the pathogenicity of P. melonis. Together with the findings of this investigation and the results of the previous studies, acyltransferase emerged as a novel pathogenicity tool of P. melonis. In addition, its characterization provided valuable grounds to understand the pathogenicity mechanisms of Phytophthora.

A comprehensive evaluation of the complete residual sequence H3GZF6 explored two aliphatic regions in domain 1 and 24 aliphatic regions in domain 2. The aromatic parts were nonexistent in the domain, while 11 aromatic fields were present in domain 2. Domain 1 had balanced charges (positive and negative charges, each on two distinctive zones). However, the charged areas in domain 2 were a little more in number (19 positive and 19 negative). The number of hydrophobic regions (2) was lesser than the number of hydrophilic areas (5) in domain 1. We observed a similar situation in the case of domain 2, where 19 hydrophobic regions were present compared to 25 hydrophilic areas. Six polar regions were significantly lower than the 16 polar regions in domain 2. Domain 1 contained four proline domains; however, the similar regions in domain 2 were 11. We could find only one cysteine in domain 1 and six in domain 2. Serine–threonine regions were 13 in number in domain 1 and 43 in domain 2 (Figure ).

Figure 4

Detailed structural analysis of the two domains of the protein sequence H3GZF6. Supramolecular observations were made about the two domains (i.e., domain A and domain B), and characteristics of residues were highlighted.

We detected two domains of the protein H3GZF6, one of which (141–219) was the antifreeze protein.[28] The antifreezing property of the protein allows the pathogenic Phytophthora to cause infections at low temperatures and makes it a successful pathogen of subtropic regions. However, in a study conducted by Shi et al.,[16] the acyltransferase was declared as a typical cold-active enzyme. The activation of the enzyme at low temperatures and the presence of the antifreeze protein support the pathogenicity of Phytophthora at low temperatures.

Sequence Characteristics

We categorized the H3GZF6 molecule into different classes, and the results highlighted two zones as the regions having post-translational modifications. It included the first 26 residues (1–26), which acted as a signal peptide. The second region consisted of a total length of 749 residues (27–775) and was annotated as ChainPRO_5003587112. Two biologically distinct areas consisting of 140 residues (108–247) and 117 residues (659–775) were present besides three polar regions. The lengths of the polar regions were 113 (108–220), 41 (659–699), and 65 (711–775) residues (Table ).

Table 2

Characteristics of the Acyltransferase

	position(s)	length
post-translational modifications (processing events)	1–26	26	signal peptide
	27–775	749	ChainPRO_5003587112
domains (biologically distinguished region)	108–247	140	region 1
	659–775	117
polar regions	108–220	113	polar region 1
	659–699	41	polar region 2
	711–775	65	polar region 3

Interaction with Ligand Molecules

Two ligand molecules of 3-cyclohexyl-1-propylsulfonic acid (CXS) were attached with the protein H3GZF6 within 3Å (Figure A). Six residues of chain A (i.e., E.242, T.244, M.246, F.255, Y.272, and Y.323) were involved in the interaction with one CXS. However, the protein–ligand interaction profiler assay revealed that the interaction of each residue was not identical. There were three hydrophobic interactions of residues, i.e., A: T.244, A: F.255, and A: Y.272, of chain A. However, two hydrogen bonds were detected with residues A: T.244 and A: T.244 of chain A (Figure B). We found the second ligand molecule of CXS interacting with residues K.70, F.75, A.78, V.345, T.346, S.347, V.348, and G.349 of chain A within 3 Å. A hydrophobic interaction residue A: V.345 was observed in chain A. We also found a salt bridge with residue A in chain A of H3GZF6 (Figure C).

Figure 5

Two-dimensional structure of ligand 3-cyclohexyl-1-propylsulfonic acid (CXS) (A). Ligand molecules of CXS are attached to protein H3GZF6. A CXS ligand is interacting with six residues of chain A (i.e., E.242, T.244, M.246, F.255, Y.272, and Y.323 (B)). The second CXS ligand is hydrophobically interacting with residue V.345 (C).

Interactive docking of CXS ligand molecules argues for the synthesis of various glycerophospholipids, which constitute biological membranes.[29,30] The de novo synthesis of organic acids involves over one acyltransferase. The group of distinctive acyltransferases catalyzes the sequential acylation. The acidic ligand molecules play a prime role in the enzyme’s catalysis.[31,32] Previously, the molecular study by Coleman[33] revealed four acyltransferase motifs. However, the physicochemical properties of the enzymes had never been a focused issue of the researchers. This study has focused on this aspect and analyzed the physicochemical profiles of the amino acid sequences of the protein H3GZF6.

Protein domain mapping of H3GZF6 highlighted a cathgene3d consisting of residues 303–384 and documented with accession number G3DSA:1.10.220.130 belonging to a homologous superfamily. The IPR032048 transglutaminase elicitor of Interpro family PF16683 with reference number GO:0016755 was present in the residues 266 to 624. The same sequence 266–624 represented another transglutaminase elicitor PF16683 categorized in Pfam domain IPR032048. The residues 1–770 belonged to the Mansc domain-containing protein 1 of the panther family with accession number PTHR16021: SF13. Two sequence portions 108–247 and 659–775 were categorized as MobiDB Lite (Mobidblt). We featured three similar sequences, i.e., 108–220, 659–699, and 711–775, as Mobidblt Polar and MobiDB Lite and residues 27–775 belonging to the noncytoplasmic domain (Supplementary Data Set 1, see Figure S2).

Acyltransferase catalyzes the acylation of organic acids (e.g., lysophosphatidic acid). It takes part in de novo acid production and serves as a precursor for the synthesis of various membrane glycerophospholipids.[34] Because acyltransferases are integral membrane proteins, it is difficult to dissolve them without inactivation. The problem had been hampering the researchers for biochemical characterization.[35] The researchers have studied most of the ubiquitous acyltransferases of physiological importance using crude membrane preparations or intact cells.[36] However, it is not a reliable method to characterize a protein due to the presence of several contaminants of diverse biochemical nature.[37] The novelty of this study lies in the solubilization of acyltransferase in its active form. The biochemical studies of the enzyme defined the limit of catalytic reactions. The physicochemical characterization of the complete sequence of H3GZF6 protein is a first-time study in this direction, which enlightens the enzyme behavior in different physical environments.

Acyltransferase protein H3GZF6 had the maximum similarity with H3GZF3. Both protein species were present in Phytophthora ramorum. The second most similar protein was AN-156 (SF-13) encoded in a genome draft (from a eukaryotic cell of Thalassiosira pseudonana) EnsebmblGenome Thapsdr belonging to domain Thpasdraft-22002. Another SF13 protein species CHLRE (AN161) encoded by the gene EnsebmblGenome at LMJF_31_2670 domain of the organism Leishmania major was in close identity with H3GZF6. Some other organisms, i.e., Dictyostelium discoideum, Ustilago maydis, Monosiga brevicollis, Anopheles gambiae, and Drosophila melanogaster, were the second clade species close to H3GZF6 (Figure A). We found the overall distribution of the protein species interkingdom, e.g., archaea, bacteria, eukaryote, etc. The sequence was also part of some viruses and viroids. The protein distribution spectrum consisted of 72% of eukaryotes, and 90% of those were oomycetes. Peronosporales shared 95% species spectra of oomycetes and are further divided into three genera, i.e., Pythium, Phytophthora, and Plasmopara. Plasmopara halstedii was the only species representing genus Plasmopara. However, five Phytophthora species (Phytophthora nicotianae, Phytophthora megakarya, Phytophthora infestans, P. ramorum, and Phytophthora parasitica) contained the protein. Pythium ultimum was the only species found with protein H3GZF3 in the genus Pythium (Figure B). A detailed list of organisms with acyltransferase and the phylogenetic tree showing their relations have been provided in the Supporting Information (Supplementary Data Set 1, see Figure S3). One hundred structurally similar proteins were categorized into 10 different groups. Among them, 70% of the most identical protein species were uncharacterized. The percentage shares of other functional groups were TGase elicitors (10%), phytopathogen effector protein RxLR (4%), transporter family proteins (3%), and endonucleases (1%). HIV tat-specific factor 1, protocadherin fat 4, transcription factor 1, and 3-hydroxyisobutyrate dehydrogenase each shared one percent among structurally similar proteins (Figure C). The details of the protein species and their roles have been enlisted in Supplementary Data Set 1 (see Table S6).

Figure 6

Dendrogram showing the proteins of similar homology with H3GZF6 along with their respective genes, domains, organism names, and characterization status. The data were collected from publicly available databases, e.g., UniProt, NCBI, etc. (A). The global prevalence of the protein with the taxonomic positions of the carrying organisms. The data were collected and confirmed from the NCBI database (B). Functional details of the protein species structurally similar to H3GZF6 (ProBis algorithm) and their share in the protein pool, already reported and submitted to publicly available databases (C).

We have also investigated the global prevalence of the protein in the studied genomes of organisms. This part of the research estimates the global impact of acetyltransferase. Several members of our biome (e.g., eukaryotes, prokaryotes, and oomycetes) possessed the homologous sequences, revealing how much enzyme is important for life around the globe. Both the sequence and the function of acetyltransferase were conserved among the organisms.[38,39] It was also a surprise that 70% of structurally similar proteins were uncharacterized, which highlighted the need for extensive research efforts directed toward this field.

Conclusions

This study characterized a novel acyltransferase from highly pathogenic P. melonis. Low temperature, neutral pH, and Tween-80 solution supported the enzyme activity and stability. However, the enzyme consists of two domains with antifreezing and TGase elicitor properties. The protein also contained two acidic ligand molecules of CXS, docked at different positions of chain A. The study also determined the global prevalence of the enzyme in other organisms and performed the functional characterization of structurally resembling proteins.

Experimental Section

Composition of Henninger Synthetic Medium

One liter of Henninger synthetic medium[40] contained KH2PO4 (0.4 g), NaNO3 (0.4 g), CaCl2 (0.1 g), MgCO3 (0.1 g), (NH4)2SO4 (0.1 g), FeSO4 × 7 H2O (0.02 g), succinic acid (0.2 g), arginine (0.2 g), glycine (0.2 g), aspartic acid (0.4 g), glutamic acid (0.4 g), alanine (0.1 g), leucine (0.1 g), cysteine HCl (0.15 g), thiamine hydrochloride (1 mg), glucose (10 g), and sucrose (5 g, pH 5.0).

P. melonis Strain and Culture Conditions

A total of six strains of P. melonis with contrasting pathogenicity on cucumber were procured from the Vegetable Research Institute, Guangdong Academy of Agricultural Sciences, Guangzhou, China. The pathogen cultures were maintained on the V8 agar medium in the dark at 28 °C. For culture filtrate preparation, each P. melonis strain was grown in 1 L of autoclaved (121 °C, 30 min) Henninger liquid culture medium at 20 °C with shaking at 120 rpm for eight days.[40] The culture filtrates were then filtered on 0.2 μm filters and lyophilized before performing the protein profiling. A separate pathogenicity assay was performed for each strain by following the method of Ahmad and Ashraf.[41] The liquid cultures were then prepared as described for the untreated samples.

Protein Profile Analysis

The protein profile of P. melonis was studied by following the method of Khan et al.[2] The complete protein profiles were evaluated to screen differentially expressed proteins. The differentially expressed proteins were compared with the pathogenicity data. For this purpose, total protein contents were extracted by adopting the method of Ahmad et al.[42] Hence, the protein extraction solvent (phosphate-buffered saline) contained 140 mM NaCl, 10 mM Na2HPO4, 1.8 mM NaH2PO4, and 2.5 mM KCl. Protein samples were dissolved in 8 M Urea solution before their native 2D electrophoresis. Extracted protein samples were electrophoresed on 12% native polyacrylamide gel (Native-PAGE). Electrophoresis in the second dimension was performed under identical conditions with the addition of sodium dodecyl sulfate (SDS) to attain fine resolution. Then, Coomassie blue staining of proteins was performed to visualize and record results. Digital images of protein gels were captured for their detailed analyses. For image analysis and identification of the protein species, the method of Ahmad et al.[42] was adopted. Protein profiles of all highly pathogenic strains were compared with those of low pathogenic strains, and the profusion index for each protein species was calculated with the formula used by Khan et al.[2] Profusion behavior of different protein species was plotted in matrix plots to screen the proteins by the “profusion index”. It provided the proteins most actively playing pathogenic roles against cucumber plants.

Protein spots were compared using digital software SAMESPOTS (TotalLab Ltd., U.K.) and TOPSPOT (Kroger and Prehm, Berlin, Germany). Characteristic features of highly pathogenic protein species were checked from the online database UniProt. Solution-state NMR spectroscopy was carried out to precisely determine the structural properties of the most pathogenic protein species (H3GZF6). Data obtained through NMR were statistically analyzed and compared with the online protein database using the software PSVS (North East Structural Genomics, NESG).

Protein Sample Preparation

The protein sample was prepared by adopting the method of Shi et al.[16] Total protein extracts were purified using Ni-NTA resin affinity chromatography. Proteins were eluted in a stepwise manner with five resin-bed volumes of 40, 100, and 250 mM imidazole elution buffer at a flow rate of 1.0 mL/min. The purified protein was dialyzed by PBST containing urea of a concentration decreasing from 6 to 4 to 2 to 0 M. After thorough dialysis, the protein solution was collected and stored at 4 °C. Protein samples were collected at different stages of the purification process and run on 12.5% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) along with a protein marker and stained with 0.05% Coomassie blue G-250, followed by a standard destaining procedure. The protein concentrations were determined by the Coomassie Brilliant Blue G-250 standard method.

Enzymatic Activity Assay for Acyltransferase

The method of McFie and Stone[43] was strictly followed to elucidate the enzymatic activity of the acyltransferase. The stock solutions used for the assay were 1 M Tris-HCl (pH 7.6), 1 mM MgCl2, 4 mM DOG in acetone, 12.5 mg/mL BSA, 500 μM NBD-palmitoyl CoA in 20 mM Tris-HCl (pH 7.6), and 50 μg of the protein sample (cell lysate or total membranes). A solvent system of 50 μL of 50 mM Tris-HCl (pH 7.6)/250 mM sucrose diluted the stock solution. Assays were performed in 16 × 100 mm glass test tubes in a final reaction volume of 200 μL. A master mix was prepared to contain 20 μL of 1 M Tris-HCl (pH 7.6), 4 μL of 1 mM MgCl2, 10 μL of 4 mM DOG, 10 μL of 12.5 mg/mL BSA, 10 μL of 500 μM NBD-palmitoyl CoA, and 96 μL of water per reaction. The volumes were scaled up proportionally to accommodate the desired number of reactions. Aliquots of 150 μL master mix were prepared for individual sample analysis. The tubes were incubated in a 37 °C water bath for 2 min. The reaction was started by adding 50 μL of the protein sample. The reaction was incubated at 37 °C for 10 min with occasional shaking. An addition of CHCl3/methanol (2:1, v/v, 4 mL) terminated the reaction, and mixing of 800 μL of water diluted it before allowing reactions an incubation of 1 h at room temperature. The control buffer was also diluted to 0.1 mM bivalent cation concentration. A serial vortex and centrifuge at 3000 rpm for 5 min separated aqueous and organic phases. Nitrogen currents dried the organic phase after the aspiration of the upper aqueous phase. CHCl3/methanol (2:1, 50 μL) suspended lipids prior to channeling their spots on a 20 × 20 cm thin-layer chromatography (TLC) plate. A solvent system of hexane/ethyl ether/acetic acid (80:20:1, v/v/v) developed the TLC plate, which was allowed for 1 h air-drying before quantification of reaction products. The TLC plate was analyzed with a VersaDoc 4000 molecular imaging system (Bio-Rad Laboratories, Inc.), and fluorescence was quantified with the Quantity One software (Bio-Rad Laboratories, Inc.). The excitation and emission wavelengths of NBD were 465 and 535 nm, respectively. A blue LED laser light source and a 530BP emission filter were used. Data were presented as units (fluorescence intensity) of acyltransferase formed per minute per mg protein.

Biochemical Properties of Acyltransferase

The optimal temperature for acyltransferase activity was determined from 0 to 60 °C and in Tris-HCl (pH 7.0) over 10 min using the method.[16] A preincubation of the enzyme was carried out for 10, 20, 30, and 40 min at 40 °C and 50 °C in Tris-HCl (pH 7.0) to determine thermostability. Then, instant exposure to the ice water bath gave the residual activity as described in the standard assay method. The pH stability of acyltransferase activity was determined after preincubation in 50 mM of different buffers within the pH range of 5.0–9.0 for 30 min at 37 °C. The buffers used were sodium acetate/acetic acid (pH 5.0), NaH2PO4/Na2HPO4 (pH 6.0, 7.0), and Tris-HCl (pH 8.0, 9.0). The influence of metal ions and various other agents on the acyltransferase activity was determined, and the concentration of each reagent was maintained as ethylenediaminetetraacetic acid (EDTA) (5 mM), thiourea (10 mM), dithiothreitol (DTT) (10 mM), urea (10 mM), SDS (10 mM), Tween-80 (0.2%), Triton X-100 (0.2%), H2O2 (0.2%), Mg2+ (5 mM), Ca2+ (5 mM), Cu2+ (5 mM), Zn2+ (5 mM), Fe2+ (5 mM), Mn2+ (5 mM), Ni2+ (5 mM), K+ (5 mM), and Na+ (5 mM). For that, the enzyme was incubated with individual compounds in 50 mM Tris-HCl (pH 7.0) at 20 °C for 1 h. The residual activity in the samples and the control (without reagents) was determined by the standard assay.[1] The highest enzyme activity was used as control (100% of relative activity).

In-Solution Tryptic Digestion

Pure protein samples were resuspended in 100 μL of 0.1 M triethylammonium hydrogen carbonate (TEAB) buffer pH 8.0. An equal amount (1 μg) of bovine β-lactoglobulin (LACβ) was spiked in each sample to serve as an internal standard for experimental bias correction. Proteins were reduced by adding 1 μL of 1% SDS and 2 μL of 50 mM tris (2-carboxyethyl) phosphine (TCEP) and heating at 60 °C for 1 h. The addition of 1 μL of iodoacetamide (400 mM) with subsequent dark incubation at room temperature (30 min) alkylated the free thiol groups of cysteine residues. Proteins were then digested overnight at 37 °C with trypsin in 0.1 M TEAB (pH 8.0, protein/trypsin ratio 50:1 w/w) using the method of Ahmad et al.[44]

Labeling and Peptide Fractionation

The resulting peptides were labeled with isobaric tags to determine their relative and absolute quantities by iTRAQ reagents Multiplex Kit (AB Sciex, Foster City, CA). Each sample was labeled with one of three isobaric tags reconstituted with 50 μL of isopropanol. The reaction was left to stand at room temperature for 60 min and then blocked by incubating with 8 μL of hydroxylamine 5% for 15 min. The mixtures of labeled peptides were then pooled and dried under the vacuum. The lyophilized peptides were dissolved in 800 μL of 5% CH3CN and 0.1% formic acid (FA) and loaded (2 × 400 μL) onto C18 Macro SpinColumns (Harvard Apparatus). Elution was performed with 2 × 200 μL of 50% CH3CN/0.1% FA. The samples were then dried under vacuum and dissolved in 360 μL of deionized water. A solution containing 6% glycerol and 0.3% IPG buffer (pH 3–10, Agilent, Santa Clara, CA) was added to a final volume of 1.8 mL. Peptides were fractionated according to their pI on an Agilent 3100 OFFGEL fractionator using commercial 12 cm IPG pH 3–10 linear strips (GE Healthcare, Waukesha, WI). The strips were rehydrated with 20 μL of rehydration solution (4.8% glycerol, 0.24% IPG buffer, pH 3–10) per well. After a 30 min incubation, 150 μL of the sample solution was loaded per well. The isoelectric focalization was carried out at 20 °C until a total voltage of 20 kV/h with a maximum current of 50 μA and a maximum power of 200 mW. After the focalization, peptide fractions (12/for each group) were recovered in separate tubes, and pH values were measured to check for the efficiency of the pH gradient. Fractions were then dried under vacuum, dissolved in 300 μL of 5% CH3CN/0.1% FA, and loaded (2 × 150 μL) onto C18 Micro SpinColumns (Harvard Apparatus). Elution was performed with 2 × 100 μL of 50% CH3CN/0.1% FA, and eluted fractions were dried under vacuum and stored at −20 °C until MS analysis.[45]

Liquid Chromatography-Tandem Mass Spectrometry (LC–MS)

Lyophilized peptides obtained from OFFGEL fractionation were dissolved in 8 μL of 5% CH3CN/0.1% FA; 5 μL of the resulting sample was injected for LC–MS/MS analysis. MS analysis was performed on an LTQ Orbitrap Velos Pro from Thermo Electron (San Jose, CA) equipped with a NanoAcquity UPLC system from Waters (Milford, MA). Peptides were trapped on a home-made (5 μm 200 Å Magic C18 AQ 0.1 × 2 mm) precolumn (Michrom, Auburn, CA) and separated on a home-made (5 μm 100 Å Magic C18 AQ, 0.75 × 15 mm) column (Michrom). The analytical separation was run for 65 min using a gradient of 99.9% H2O/0.1% FA (solvent A) and 99.9% CH3CN/0.1% FA (solvent B). The gradient was run as follows: 0–1 min 95% A and 5% B, then to 65% A and 35% B at 55 min, and 20% A and 80% B at 65 min at a flow rate of 220 nL/min. For MS survey scans, the OT resolution was set to 60000, and the ion population was set to 5 × 105 with an m/z window from 400 to 2000. A maximum of three precursors was selected for both the collision-induced dissociation (CID) in LTQ and the high-energy C-trap dissociation (HCD) with analysis in the OT. For MS/MS in the LTQ, the ion population was set to 7 × 103 (isolation width of 2 m/z). In contrast, for MS/MS detection in the OT, it was set to 2 × 105 (isolation width of 2.5 m/z), with a resolution of 7500; the first mass was set at m/z  =  100, and the maximum injection time was 750 ms. The normalized collision energies were set to 35% for CID and 60% for HCD.[46]

Data Extraction, Relative Protein Quantification, and Database Interrogation

Peak lists were generated from raw data using the software MZMine version 2.30 (Pluskal, Okinawa, Japan). The conditions and methods were followed as used by Bashir et al.[47] After peak list generation, the CID and high-energy C-trap dissociation spectra were merged for simultaneous identification and quantification. The merged files were used for protein identification and quantification. For protein identification, parameters were specified as follows: databases  = uniprot_sprot/ uniprot_trembl; taxonomy  =  Phytophthora; precursor error tolerance  =  25 ppm; variable modification  =  oxidized methionine; fixed modifications  =  carbamidomethylated cysteine, iTRAQ-labeled amino terminus and lysine; enzyme  = trypsin; potential missed cleavage  =  2; cleavage mode  =  normal; search round  = 1. Protein and peptide scores were set up to maintain the false positive peptide ratio below 5%. For protein quantification, the isotopic correction was applied to reporter intensities according to the iTRAQ reagents certificate of analysis. iTRAQ reporter peak intensities were further normalized using the spiked LACβ standard. For each protein, the mean, the standard deviation, and the coefficient of variation of relative peptide intensities were obtained. The ratio of the protein was then computed as the geometric mean of all peptide ratios belonging to the protein. A Student’s t-test distribution was computed by the algorithm with a null hypothesis stating the log2 of the protein ratio equal to zero (confidence interval  =  95%).

The protein sequence was modeled by SIMULINK, an add-in of MATLAB. The package constructed the structure based on its homology (comparative modeling) by constructing an atomic-resolution model. For this purpose, the amino acid sequence was run a BLAST analysis, from which the most similar protein sequence (3tw5 belonging to Phytophthora sojae) was selected. Alignment analysis was carried out between H3GZF6 and 3tw5 before the development of the 3D structure. After the structure development, it was compared with the structure of the homologous protein (3tw5). After the successful production of alignment mapping residues, the template structure was used to produce a structural model. For 3D simulations, protein cleavage analysis was performed to determine the distance, and terminal selection boxes (N and C) were set from end to end as default. A built-in graphical user interface was used to visualize the constructed structure for detailed analysis and for the detection of the ligand attachment α-factor and the protein subunits participating in reactions. Ligands appeared on both sides of the reaction. Ordinary differential equations (ODEs) were used to rate the model for biochemical properties, and then binding for the formation of a heterotrimeric protein complex was performed. kGa and kGd were calculated on the basis of dose–response curves.[48]

The identifiers of protein-detected domains underwent a BLAST analysis with a bit score > 600 to find out peptides with similar sequences. The algorithm was used at a maximum threshold level of 0.05, a bit threshold of 25, and an initiating size of six. A matrix Blosum62 was selected with Gap Costs extension 1.0 and existence 11.0. The compositional adjustments were opted to be the conditional compositional score matrix adjustment in the absence of any filter and/or mask settings. The results were used to construct a dendrogram showing the taxonomic distribution of the protein in other species and higher taxa. Furthermore, a list of structurally similar proteins was developed using the ProBis algorithm based on similar interaction patterns of the binding sites[49] (Supplementary Data Set 1, see Figure S4).