The Immunoreactive Exo-1,3-β-Glucanase from the Pathogenic Oomycete Pythium insidiosum Is Temperature Regulated and Exhibits Glycoside Hydrolase Activity

The oomycete organism, Pythium insidiosum, is the etiologic agent of the life-threatening infectious disease called “pythiosis”. Diagnosis and treatment of pythiosis is difficult and challenging. Novel methods for early diagnosis and effective treatment are urgently needed. Recently, we reported a 74-kDa immunodominant protein of P. insidiosum, which could be a diagnostic target, vaccine candidate, and virulence factor. The protein was identified as a putative exo-1,3-ß-glucanase (Exo1). This study reports on genetic, immunological, and biochemical characteristics of Exo1. The full-length exo1 coding sequence (2,229 bases) was cloned. Phylogenetic analysis showed that exo1 is grouped with glucanase-encoding genes of other oomycetes, and is far different from glucanase-encoding genes of fungi. exo1 was up-regulated upon exposure to body temperature, and its gene product is predicted to contain BglC and X8 domains, which are involved in carbohydrate transport, binding, and metabolism. Based on its sequence, Exo1 belongs to the Glycoside Hydrolase family 5 (GH5). Exo1, expressed in E. coli, exhibited ß-glucanase and cellulase activities. Exo1 is a major intracellular immunoreactive protein that can trigger host immune responses during infection. Since GH5 enzyme-encoding genes are not present in human genomes, Exo1 could be a useful target for drug and vaccine development against this pathogen.


Introduction
The filamentous, aquatic, oomycetous microorganism, Pythium insidiosum [1][2][3], is unlike most other oomycetes, in that it infects humans and other animals, leading to a life-threatening disease, called pythiosis. The disease has been increasingly reported from tropical and subtropical areas around the world [2,3]. Clinicopathological features of pythiosis include arteritis of extremities (vascular pythiosis), keratitis (ocular pythiosis), and skin ulcer (cutaneous/subcutaneous pythiosis). Diagnosis of pythiosis is difficult. Antimicrobial chemotherapy is not effective against P. insidiosum. Immunotherapy using a P. insidiosum vaccine is available, but with limited efficacy [1,4,5]. Surgical removal of an infected organ (i.e., eye and leg) is the main treatment option [1]. Many patients die from uncontrolled and progressive infection. An effective treatment for patients with pythiosis is urgently needed. Better understanding the biology and pathogenesis of P. insidiosum could lead to better methods of infection control.
Many pathogens produce immunogenic proteins that are involved in important biological processes, host immunity, or pathogenesis. For example, Cryptococcus neoformans produces the immunogens MP84 and MP115 [6]. In the dimorphic fungus Blastomyces dermatitidis, the 120-kDa immunogen, BAD1, was identified as a surface protein that contributes to pathogenesis of blastomycosis [7]. The 65-kDa immunogen, Camp65p, from Candida albicans is a glucanase protein that functions as adhesin and virulence factor [8]. MP65 elicits protective immunity against C. albicans, and therefore, has been considered as a potential vaccine candidate [8,9].
A 74-kDa immunogen has been consistently identified in various strains of P. insidiosum [10]. The 74-kDa immunogen is recognized by sera from patients with pythiosis, but not sera from healthy individuals. By using proteomic and molecular genetic approaches, the 74-kDa immunogen was identified as a putative exo-1,3-β-glucanase (Exo1; formerly known as Pin-sEXO1 [11]), and a 924-bp partial Exo1-encoding sequence was cloned from genomic DNA (gDNA) of P. insidiosum, as reported by Krajaejun et al [12]. Details of the biological role of Exo1 in the life of P. insidiosum are unknown.
β-glucan is major cell wall component of fungi and oomycetes (including P. insidiosum) [13][14][15][16]. β-glucan-degrading enzymes, β-glucanases, can facilitate morphogenesis, cell wall remodeling, hyphal elongation, and growth, which are crucial for fungal physiology and virulence [8,13,14]. β-glucanase-encoding genes have been identified in oomycetes [16,17], but information on the structure and function of these genes is limited. Here, we aim to identify the full-length coding sequence of the P. insidiosum's exo-1,3-beta glucanase gene (exo1; formerly known as PinsEXO1 [11,18,19]), and immunologically and biochemically characterize its gene product. We found that exo1 was up-regulated upon exposure to body temperature, and its gene product is a major intracellular immunoreactive protein that exhibited glycoside hydrolase activities. Detailed characterization of exo1 could promote better understanding of P. insidiosum's biology and pathogenesis.

Ethics statement
This study was approved by the Committee on Human Rights Related to Research Involving Human Subjects, at the Faculty of Medicine, Ramathibodi Hospital, Mahidol University (approval number MURA2006/418/NS 1:2 ). An informed consent was not obtained from patients (from whom microorganisms and clinical specimens were obtained) because the data were analyzed anonymously, and the institutional ethics committee waived the need for written informed consent from the participants.

Primer
Sequence

RACE PCR
cDNAs were generated from P. insidiosum's RNA using a GeneRacer kit (Invitrogen). Briefly, 4 μg of total RNA underwent dephosphorylation using calf intestinal phosphase, removal of the mRNA cap using tobacco acid pyrophosphatase, ligation of the GeneRacer RNA oligo (5 0 -CGACUGGAGCACGAGGACACUGACAUGGACUGAAGGAGUAGAAA-3 0 ) using T4 RNA ligase, and cDNA synthesis using the SuperScript III reverse transcriptase and the GeneRacer Oligo dT primer (Table 1; Fig 1C and  . PCR products were assessed by 1% agarose gel electrophoresis. If more than one PCR product were observed, a second-round PCR amplifications (25-μL reaction) was then performed as mentioned above, using 1 μl of primary PCR product, 0.2 μM each of the primer GeneRacer 5' Nested and Pr62 (or the primer GeneRacer 3' Nested and Pr68). Amplification and sequencing of full-length exo1 Full-length exo1 gene sequence was PCR amplified from gDNA of P. insidiosum (strain P06, P16, P17, P24, P29, and Pi-S) using the primer Pr71 and Pr72, and from cDNA of P. insidiosum (strain P06) using the primer Pr71 and GeneRacer 3' Nested (Table 1; Fig 1D and Fig 1). Automated sequencing was performed using an ABI 3100 Genetic Analyzer (Applied Biosystems) and analyzed using the Applied Biosystems Sequencing software.
Measurement of exo1 transcript cDNA was synthesized, using a reverse transcription system kit (Promega), in a 20-μl reaction, comprising 1,000 ng total RNA, 1 μl random hexamer, 1 mM dNTPs, 0.5 μl recombinant RNasin ribonuclease inhibitor, 0.5 μl AMV reverse transcriptase, 1x reverse transcription buffer, and nuclease-free water. The reaction was performed in a Biorad MyCycler thermal cycler, using the following conditions: 25°C for 10 min, 42°C for 30 min, 95°C for 5 min. The primer Pr77 and Pr78 were used to amplify exo1, while the primer Pr79 and Pr80 were used to amplify the P. insidiosum's actin gene, act1 (accession number: HS975373). Real-time PCR was performed in a 20-μl reaction, comprising 100 ng cDNA template, 10 mM each primer, 1x SsoFast EvaGreen (Biorad), and RNase-free water, using a CFX96 Touch real-time PCR machine (BioRad) with the following conditions: the initial denaturation at 95°C for 3 min, 40 cycles of 95°C for 10 s and 57°C for 30 s, the final extension at 65°C for 5 s, and then 95°C for 5 s. A reaction without cDNA (no-template control) served as negative control. Each reaction was performed in triplicate. Expression of exo1, in relation to act1, was analyzed using the CFX 3.0 relative normalized expression program (Biorad). Table 2. List of oomycete and fungal microorganisms whose top BLAST-hit DNA sequences vs. the P. insidiosum's exo1 gene were used for phylogenetic analysis. All sequences were retrieved from the FungiDB database [27], and BLAST searched against the NCBI nucleotide database.

Microorganism
Group FungiDB gene ID BLAST search against NCBI database Synthetic peptides and rabbit anti-Exo1 peptide serum B-cell epitopes of the Exo1 protein were predicted using the PREDITOP program [34]. Three predicted peptides (Peptide-A, GLIGSQNGFDNSGKT; Peptide-B, GPYGTGTSGPSFGL; and Peptide-C, QQAAAIPCYNPIGRP) were synthesized (>95% purity) and conjugated with keyhole limpet hemocyanin (Mimotopes, Australia). Rabbit antisera, raised against a pool of these three peptides (anti-Exo1 peptide serum), was purchased from Mimotopes. ELISA titer of the rabbit antisera against each peptide was measured, using the Mimotopes's protocol.

ELISA
Serum samples from three pythiosis patients (who were diagnosed by culture identification or serological tests [35]) and three healthy blood donors (who came to the Blood Bank Division, Ramathibodi Hospital) were used for ELISA analyses. Wells of 96-well polystyrene plate were coated with 100 μl/well of 5 μg/ml of either Peptide-A, -B or -C, or with a pool of these peptides (1:

SDS-PAGE and Western blot
Proteins of SABH, CFA and cell lysate of Exo1-expressing E. coli (10 μg/lane) were separated by SDS-PAGE (4% stacking gel and 12% resolving gel), using a Biorad MiniProteon II apparatus (setting: 100 V, for 90 min) and blotted onto a polyvinylidene difluoride (PVDF) membrane (0.2-μm pore size; Millipore), using a Biorad Mini Trans-Blot cell apparatus (setting: 100 V for 60 min). The PVDF membrane was blocked with 5% nonfat-dried milk in TBS-T buffer (1.0 M Tris-base, 1.5 M NaCl, and 1.0% Tween 20) for 1 hr at room temperature, and washed once with TBS-T. The membrane was incubated with rabbit anti-Exo1 peptide serum (1:2,000 in 5% nonfat-dried milk in TBS-T) at 37°C for 1 hr. After the membrane was washed twice with TBS-T, goat anti-rabbit IgG conjugated with alkaline phosphatase (Southern Biotech; 1:4,000 in 5% nonfat-dried milk in TBS-T) was added to the membrane, and incubated at 37°C for 60 min. The membrane was washed three times with TBS-T. Western blot signals were developed using NBT and BCIP (Roche).
To block the anti-Exo1 antibodies in the rabbit immune serum, 3 ml of diluted rabbit serum (1:1,000 in 1% BSA in TBS-T) and a combination of Peptide-A, -B, or -C (5 μg each), were incubated at 4°C overnight, and then centrifuged at 440 x g. The resulting pre-absorbed serum was used in Western blot, as mentioned above. exo1 expression in E. coli exo1 coding sequence was amplified from the P. insidiosum strain P06 (CBS119452) in a 25-μl PCR reaction, containing 100 ng gDNA, 0.5 μl Elongase (Invitrogen) and its 1x buffer (buffer A: B = 1:4), 0.4 μM dNTPs, and 0.2 μM each of the primer PinsEXO1BamHI and PinsEXO1NcoI ( Table 1). The reaction was carried out in a MyCycler thermal cycler (Biorad), using the following conditions: 94°C for 1 min, 35 cycles of 94°C for 30 s, 55°C for 2 min, and 68°C for 4 min, and then 68°C for 15 min. The PCR product was purified using a PCR cleanup kit (GeneAid), double digested with BamHI and NcoI (New England Biolabs), and directionally cloned into pRSET-C (Invitrogen). The resulting plasmid, pPinsEXO1, was checked for correct in-frame translation, and transformed into the Escherichia coli strain BL21 (DE3) pLys for protein expression.
A bacterial clone, containing pPinsEXO1, was grown in LB broth supplemented with ampicillin (50 μg/ml). The bacteria were incubated with shaking (200 rpm) at 37°C for~3 hr to an optical density (600 nm) of 0.6-0.8. Protein expression was induced with 1 mM IPTG (Invitrogen) at 37°C for another 3 hr. Bacteria were harvested by centrifugation (6,000 x g) at 4°C for 15 min, and disrupted with BugBuster protein extraction reagent (Novagen) and sonication. The cell lysate was centrifuged and the induced protein was enriched from the supernatant using a Ni-NTA agarose column (Qiagen) and elution with immidazole. The eluted proteins underwent SDS-PAGE and Western blot analyses, using 1:5,000 mouse anti-6x histidine (anti-6xHis) antibody (Southern Biotech) and 1:4,000 goat anti-mouse IgG conjugated with alkaline phosphatase (Southern Biotech).

Nucleotide sequence accession numbers
All exo1-coding sequences from P. insidiosum strain P06, Pi-S, P16, P24, P29, and P17 have been submitted to the DNA Data Bank of Japan database, under accession numbers LC033486 to LC033491, respectively.
The exo1 open reading frame encodes a 742-amino-acid protein (Fig 1G). The SignalP program indicated that the first 23 N-terminus amino acids of the Exo1 protein is a signal peptide. The PREDITOP program [34] predicts Peptide-A and -B (Fig 1G) to be B-cell epitopes. Peptide-C is a chimeric peptide comprising two short amino acid regions from N-terminus end (QQAAAIP) and C-terminus end (YNPIGRP) of Exo1 (Fig 1G). BLAST search of these peptides against the P. insidiosum transcriptome [21] indicated that the peptides have homologies to several different transcripts representing different glucanase genes of P. insidiosum. All three peptides had matches to the transcript #UN05080; whereas only Peptide-A and Peptide-B had matches in the transcript #UN00475; Peptide-B alone was found in transcripts #UN03240, UN24957 and UN22794; and finally, only Peptide-C was found in transcript #UN01457 (Table 3). NCBI accession number, predicted protein length, calculated protein molecular weight, number of 454-derived transcript reads (when P. insidiosum grew at 37°C [21]), and sequence alignment analysis against Exo1 (as the query sequence), corresponding to each transcriptome-derived protein (as the subject sequence), are summarized in Table 3. Predicted protein structure (size and domain) of Exo1 and the transcriptome-derived proteins are shown in Fig 2. The exo1-coding sequences (accession number: LC033486 to LC033491) from six strains of P. insidiosum, and 35 top exo1-BLAST hit sequences from 9 oomycetes and 26 fungi [27] ( Table 2), were included for phylogenetic analysis. In the resulting tree, the sequences from P. insidiosum and other oomycetes form a clade that is separate from the sequences of fungi (Fig  3). Among the oomycetes, all exo1 sequences group together, and their phylogenetic positions placed them closer to the sequences from S. parasitica and P. ultimum than the sequences from other oomycetes.

Exo1 is an intracellular immunogen
Rabbit anti-Exo1 peptide serum was raised against the combination of Peptides-A, -B, and -C (Fig 1G). Based on ELISAs, the rabbit pre-immune serum (dilution 1:1,000) had only trace reactivity against the peptides, while the rabbit anti-Exo1 peptide serum (dilution 1:1,000) reacted strongly with each peptide (Fig 5). When the rabbit anti-Exo1 peptide serum was tested at a higher dilution (i.e., 1:64,000), Peptide-A showed the strongest immunoreactivity (Fig 5).
The rabbit anti-Exo1 peptide serum was used to detect the cellular location of Exo1. The proteins in SABH (representing intracellular proteins) and CFA (representing secreted proteins) were separated by SDS-PAGE, blotted on a Western blot membrane, and probed with the rabbit anti-Exo1 peptide serum (Fig 6). The 82-and 78-kDa bands (estimated by in-gel protein markers) in SABH reacted strongly to the rabbit anti-Exo1 peptide serum, whereas the corresponding bands were faint in CFA (Fig 6), indicating that Exo1 is an intracellular protein. The rabbit anti-Exo1 serum, pre-absorbed with a combination of all peptides or Peptide-A and -B, failed to detect 82-and 78-kDa bands in SABH (Fig 7). Peptide-A, -B, -C, or combination of these peptides was used to coat wells of an ELISA plate, and incubated with serum samples from pythiosis patients (n = 3; PS1-3) and healthy blood donors (n = 3; CS1-3). With an exception of the Peptide-C (Fig 8C), all pythiosis patient sera reacted strongly to the Peptides-A, B, or combination of all peptides (Fig 8A, 8B and 8D, respectively), as indicated by ELSIA signals well above the cutoff. In contrast, all control sera reacted poorly to all peptides (Fig 8A-8D), as indicated by ELISA signals below the cutoff. These results demonstrate that, during P. insidiosum infection, humans mount an antibody response to peptides of the Exo1 protein. The Immunoreactive Exo-1,3-β-Glucanase from Pythium insidiosum

Exo1 exhibits glycoside hydrolase activity
In order to demonstrate the expected hydrolytic activity of Exo1, we cloned the full-length coding region into an E. coli expression vector pRSET-C, to make pPinsEXO1, as a HIS-tagged fusion (Methods). The E. coli strain harboring pPinsEXO1 produced only trace amounts of protein (molecular weight:~60-90 kDa) recognized by the rabbit anti-Exo1 peptide serum (S1 Fig), and nickel-immobilized affinity chromatography failed to enrich the protein (data not shown). As an alternative for demonstrating Exo1 enzyme activity, we used an agar plate assay (Methods) to measure hydrolytic activity of Exo1, being expressed in the bacteria. The pPin-sEXO1-harboring E. coli (Fig 9A and 9B) and the positive controls [T. harzianum's lysing Western blot analysis of P. insidiosum's crude protein extracts using rabbit anti-Exo1 peptide serum. Crude proteins (SABH and CFA) extracted from P. insidiosum were separated in a SDS-PAGE gel, transferred to a Western blot membrane, and probed with the rabbit pre-immune or post-immune serum. The arrow and arrow head indicate the 82-and 78-kDa band, respectively. Protein molecular weight markers  are shown in kDa. (SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; CFA, culture filtrate antigen; SABH, soluble antigen from broken hyphae; Pre-immune, rabbit pre-immune serum; Post-immune, rabbit anti-Exo1 peptide serum).
doi:10.1371/journal.pone.0135239.g006 enzyme ( Fig 9C) and T. reesei's cellulase (Fig 9D)] provided a hydrolytic/clear zone in LB agar supplemented with laminarin, in a dose-dependent manner [i.e., with higher bacterial density or higher enzyme concentration, there is a larger clearance zone]. None of the negative controls [bacteria with pRSET-C empty plasmid (Fig 9E), and plain LB broth (Fig 9F)] produced a clear zone in the laminarin plates. In Fig 10 we show a linear relationship of either cellulase or Exo1-expressing E. coli concentrations on the clear zone diameters in the laminarin plates. Similar findings were observed in the LB plate supplemented with Avicel (a microcrystalline cellulose). Taken together, these findings indicate that the transgenically-expressed Exo1 has glycoside hydrolase activity. The Immunoreactive Exo-1,3-β-Glucanase from Pythium insidiosum
As expected from their evolutionary histories [15], the glucanase-encoding genes from the oomycetes (including P. insidiosum) were phylogenetically grouped together and separate from the glucanase-encoding genes from fungi ( Table 2; Fig 3). Among oomycetes, glucanasebased phylogenetic analysis (Fig 3) surprisingly showed that P. insidiosum was more closelyrelated to Saprolegnia parasitica (a member of the Saprolegnian lineage), than Pythium ultimum and also Phytophthora spp (Pythium and Phytophthora species are members of the Peronosporalean lineage) [41]. This paradoxical relationship may be tied to the observations that both P. insidiosum and S. parasitica are animal pathogens, while all other oomycetes included in this study are plant pathogens [42]. The Immunoreactive Exo-1,3-β-Glucanase from Pythium insidiosum Changing expression of exo1 in response to stresses, such as elevated temperature, aging, and limited carbon source, have been investigated here. The ability to grow at human body temperature (37°C) is a crucial factor for successful pathogens, including P. insidiosum. As demonstrated by quantitative real-time PCR (Fig 4), increasing the temperature, from 28°C to 37°C, significantly up-regulated exo1 expression, suggesting that Exo1 is required as a part of pathogen's adaptation to high temperature. In synthetic growth medium, dextrose is the main source of carbon for basic metabolisms of P. insidiosum. Aging P. insidiosum (the organism was grown in the medium for a relatively long period) up-regulated exo1 expression (Fig 4). One hypothesis to explain an increase in Exo1 with aging or senescence is that the release of dextrose/glucose from the cell wall or intracellular β-glucan stores could support continued basal metabolism as external supplies of glucose dwindle. In contrast, growth of the organism in medium lacking dextrose from the initiation of the culture resulted in significant down-regulation of exo1. We propose that under the condition of no added dextrose, the metabolism of the cell is dramatically slowed, and there is likely a global effect on the expression of many genes.
Western blot analysis, using rabbit anti-Exo1 peptide serum, detected two major immunogens (~82 and~78 kDa), and trace amounts of several lower molecular weight proteins (~20-60 kDa), in SABH (representing intracellular proteins) (Fig 6). All of these immunogens were also detected as secreted proteins of low abundance in CFA (Fig 6). These results suggests that, although small amount of the protein may be are found in the culture supernatant, Exo1 is predominantly a cytoplasmic protein. The rabbit anti-Exo1 serum, pre-absorbed with all three peptides (Peptide-A, -B, and -C), failed to detect 82-and 78-kDa proteins, whose sizes match the transcriptome-derived Exo1 homologous proteins that contains both Peptide-A and -B (UN05080 and UN00475; Table 3), in SABH (Fig 7), suggesting that the rabbit anti-Exo1 serum was specific to the proteins containing Peptide-A, -B or -C, including the 82-kDa protein, Exo1 (Table 3; Fig 2). In addition, when the rabbit anti-Exo1 serum was pre-absorbed with a combination of Peptide-A and -B (but not Peptide-A and -C, nor Peptide-B and -C), the 82-and 78-kDa proteins disappeared (Fig 7), indicating that these two proteins contain both Peptide-A and -B, which had strong antibody recognition. The Immunoreactive Exo-1,3-β-Glucanase from Pythium insidiosum BLAST searches of the Exo1 Peptide-A, -B, and -C (Fig 1G) against the 454-derived transcriptome of P. insidiosum [21] revealed that at least two of the Exo1 peptides matched two abundant transcripts (Table 3): UN05080 (number of transcript reads, 109; predicted protein size, 83 kDa) and UN00475 (17 reads; 76 kDa). Either Peptide-B or -C matched several other transcripts (Table 3): UN03240 (1 read; 33 kDa), UN01457 (1 read; 28 kDa), UN24957 (4 reads; 27 kDa), and UN22794 (1 read; 20 kDa). These transcriptome-derived proteins have high degree of sequence similarity to Exo1 (identity 88-100%; E-value < -87; Table 3). The two major bands on the Western blot (Fig 6) match well with the expected sizes predicted by the genes represented by UN05080 and UN00475 transcript clones (Table 3). Based on peptide mapping, sequence homology, and predicted protein size and domain, UN05080-derived protein is most similar to Exo1 (Table 3; Fig 2). According to BLAST searches against the NCBI database, the P. insidiosum transcriptome-derived protein sequences found matches to putative exo-1,3-β-glucanases from Phytophthora spp (identity 48-72%; E-value < -51; Table 2). Like other oomycetes, which have several glucanase genes in their genomes [16,17], the P. insidiosum transcriptome contains at least 6 glucanase-homologous genes that encode proteins with predicted BglC or X8 domains (Table 3; Fig 2). These homologues are expected to serve roles in carbohydrate metabolism, but their precise roles and how they may differ from each other remain to be investigated.
ELISA analysis revealed that Peptide-A and Peptide-B were strongly recognized by pythiosis patient sera (PS1-3), but not by control sera (CS1-3; Fig 8A-8B). Additionally, Peptide-A was recently reported as an efficient ELISA marker used for diagnosis of pythiosis [11]. This result confirmed that Exo1 is a major immunoreactive protein [10,12] with B-cell epitopes that trigger a host immune response during natural infection of P. insidiosum.
In conclusion, full-length exo-1,3-β-glucanase-encoding gene (exo1) was successfully cloned and expressed. Glucanase genes from oomycetes, including P. insidiosum, form a clade that is distantly related to the glucanase genes from fungi. Exo1 was predicted to contain the BglC and X8 domains found in proteins with a role in carbohydrate metabolism. Exo1 was characterized as a major intracellular immunoreactive protein that exhibits GH5 hydrolytic activity. exo1 is up-regulated upon exposure to body temperature and prolonged incubation, but down-regulated upon exposure to long term carbon-depleted conditions. Exo1 contains B-cell epitopes that trigger host immune responses during P. insidiosum infections in humans. Since GH5 enzyme-encoding genes are not present in the human genome, Exo1 could be a good candidate for development of drugs or vaccines against P. insidiosum.