Biochemical and genetic analyses of the oomycete Pythium insidiosum provide new insights into clinical identification and urease-based evolution of metabolism-related traits

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 ID 05-60-77).

Microorganisms and growths

Twenty-six strains of P. insidiosum isolated from humans (n = 14) or equines (n = 10) with pythiosis and from the environment (n = 2), were available for this study (Table 1). Identity and genotyping (i.e., Clade-I, II, and III) of P. insidiosum were confirmed through culture identification, single nucleotide polymorphism-based multiplex PCR, and rDNA sequence analysis (Chaiprasert et al., 1990, 2009; Badenoch et al., 2001; Rujirawat et al., 2017). Because P. insidiosum has been classified as a Biosafety Level 2 organism (https://www.atcc.org), Biosafety Level 2 precautions were followed throughout this study (https://www.cdc.gov/biosafety). All of the organisms were retrieved from stock cultures, and maintained on Sabouraud dextrose (SD) agar at 37 °C for at least three passages. SD agar plugs (5 mm in diameter) from one-week-old, actively growing cultures of P. insidiosum were then prepared (Krajaejun et al., 2010; Lerksuthirat et al., 2017) for biochemical assays. Radial growth rate (mm/day) of P. insidiosum was evaluated, using the previously described method (Krajaejun et al., 2010; Lerksuthirat et al., 2017). Strains with growth rates ≥5 mm/day were defined as fast-growing strains, while the rest were defined as slow-growing strains.

Biochemical assays

To set up biochemical assays, 10 different routinely available agars were each prepared in test tubes (except the DNase assay agar, which was prepared in a Petri dish), using ingredients purchased from BD Difco and BBL (if not stated otherwise), and the recommended protocols of the manufacturers. These agars included: urea agar (urease assay), Simmons’citrate agar (citrate hydrolysis assay), bile esculin agar (esculin hydrolysis assay), DNA agar (DNase assay), and purple agar base (sugar hydrolysis assay) with 2% (wt/v) dextrose, lactose, maltose, sucrose (Merck, Darmstadt, Germany), trehalose (Sigma-Aldrich, St. Louis, MO, USA), or xylose. A 5 mm diameter agar plug of an actively growing colony of each P. insidiosum strain was placed upon each type of agar and incubated at 37 °C for two days before biochemical reactions were read. Each biochemical assay was interpreted as “negative” if the agar color remained unchanged, and interpreted as “positive” when the agar color changed: (i) from yellow to pink (urease assay); (ii) from brown to black (esculin hydrolysis assay); (iii) from green to blue (citrate hydrolysis assay); (iv) from dark blue to yellow (all sugar hydrolysis assays); and (v) from blue to colorless (DNase assay). All biochemical assays were performed in duplicate.

Identification of urease- and urease accessory protein-encoding genes

The Oomycete Gene Table is an online comparative genomic analysis tool, derived from sequence similarity based gene grouping of the genome sequences of P. insidiosum, 19 related oomycetes, and two diatoms (Table S1) (Kittichotirat et al., 2011; Rujirawat et al., 2018). In the current study, the Oomycete Gene Table shows identification of putative urease- and urease accessory protein-encoding genes in the genomes of the oomycetes and diatoms (Fig. 1). Predicted urease protein sequences of the oomycetes and diatoms were aligned using MUSCLE (Edgar, 2004; Dereeper et al., 2008, 2010), and assessed for sequence identity and similarity using NCBI BLAST (https://blast.ncbi.nlm.nih.gov/).

The urease and urease accessory protein sequences of the plant Arabidopsis thaliana (accession numbers: NP_176922 (urease structure protein, URE); NP_850239 (urease accessory protein D, URED); NP_173602 (urease accessory protein F, UREF); and NP_180994 (urease accessory protein G, UREG)) (Witte, Rosso & Romeis, 2005) were retrieved from the NCBI database. To assess the presence of the orthologs in P. insidiosum, all of these Arabidopsis proteins were TBLASTN searched against the genome of the P. insidiosum strain Pi35 (also known as Pi-S), and two Illumina-derived genomes of the P. insidiosum strains Pi07 (also known as CBS 573.85) and Pi45 (Rujirawat et al., 2015; Kittichotirat et al., 2017; Patumcharoenpol et al., 2018), using the locally installed blast 2.2.28+ program (http://www.ncbi.nlm.nih.gov/) and the cut-off E-value ≤10⁻⁶.

Phylogenetic analysis

Phylogenetic analysis of 24 urease-encoding sequences from P. insidiosum (strains Pi07, Pi35, and Pi45), related oomycetes, and diatoms (outgroup) (Table S1) was executed online at www.phylogeny.fr (Dereeper et al., 2008). In brief, the sequence alignment was performed by MUSCLE (Edgar, 2004). Poorly aligned positions or gaps were eliminated by Gblocks (Castresana, 2000). Phylogenetic relationships were calculated by PhyML, using the maximum-likelihood algorithm and the branch-assessing aLRT test (Anisimova & Gascuel, 2006; Guindon et al., 2010). The phylogenetic tree was reconstructed using TreeDyn (Chevenet et al., 2006).

Sequence accession numbers

Sequences of the putative urease genes of P. insidiosum identified in the genomes of P. insidiosum strains Pi35 (accession number, LC317047 for Ure1), Pi07 (accession number LC325168 for Ure1), and Pi45 (LC325169 for Ure1A, and LC325170 for Ure1B) have been submitted to the DDBJ database.

Growth and biochemical characteristics of P. insidiosum

Twenty-six strains of P. insidiosum included in the current study were derived from different sources (humans, n = 14; animals, n = 10; and the environment, n = 2) and geographic origins (Asia, n = 15; Americas, n = 10; and Australia, n = 1). Based on the growth rates, P. insidiosum can be divided into two groups: (i) fast-growing strains (growth rate ≥5 mm/day; n = 17; 65% of all strains), and (ii) slow-growing strains (growth rate <5 mm/day; n = 9; 35% of all strains) (Table 1). Each group contained representatives from all phylogenetically distinct Clades (-I, -II, and -III), and from both humans and animals. Both environmental strains belonged to the fast-growing group.

As summarized in Table 1, all strains of P. insidiosum hydrolyzed esculin in the presence of bile but failed to breakdown citrate and two sugars (i.e., lactose and xylose). The majority of the strains can hydrolyze dextrose (n = 22; 85% of all strains), maltose (n = 22; 85%), sucrose (n = 20; 77%), trehalose (n = 22; 85%), and DNA (n = 23; 89%), while those that cannot utilize these substrates were almost all slow-growing. Unlike the other fast-growing strains, Pi03 did not hydrolyze sucrose. With regard to the urease assay, 71% (n = 12) of the fast-growing and 78% (n = 7) of the slow-growing strains could catabolize urea. Biochemical characteristics of some representative strains at day 0 (all agar colors remained unchanged) and day 2 post-inoculation (all biochemical reactions were read) were displayed in Fig. 2.

Ureases and urease accessory proteins of P. insidiosum and related oomycetes

Urease requires a number of urease accessory proteins to mediate enzymatic activity. Genes annotated as “urease” or “urease accessory protein” were searched using the Oomycete Gene Table (Rujirawat et al., 2018). All oomycetes and diatoms harbored a single copy of urease-encoding sequence (Gene cluster ID, #057948; average protein length: 849 amino acids; range: 761–1,345 amino acids), except the oomycete A. invadans, which contained three copies of this gene (Fig. 1; Table S1). Protein sequence alignment showed a high degree of identity (59–81%) and similarity (72–88%) between the ureases of oomycetes and diatoms (Fig. 3; Table S1).

A total of eight clusters of urease accessory protein-encoding genes were differentially presented in the genomes of 20 oomycetes (Fig. 1). These gene clusters included Cluster IDs: #051204 (found in 19 species), #291367 (17 species), #181024 (16 species), #152345 (15 species), #205644 (13 species), #213938 (12 species), #122775 (10 species), and #007410 (five species). Each oomycete genus possessed a different number of urease accessory gene clusters, for example: seven to eight clusters in Phytophthora, five to eight in Pythium, seven in Phytopythium, four in Saprolegnia, one to three in Aphanomyces, and one each in Albugo and Hyaloperonospora. None of these urease accessory gene clusters was identified in the diatom genomes.

TBLASTN search of the function-verified urease URE and urease accessory proteins URED, UREF, and UREG of the plant A. thaliana showed significant matches (E-value ≤−6) in the genomes of three representative P. insidiosum strains (Table 2): Pi07 (Clade-I strain), Pi35 (Clade-II strain), and Pi45 (Clade-III strain). One exception is UREF, which failed to find match in the genome of strain Pi07.

Urease-based phylogenetic relationships

A set of 24 urease-encoding sequences identified in the genomes of P. insidiosum, related oomycetes, and diatoms (Fig. 1; Table S1), were subjected to reconstruction of a maximum likelihood-based phylogenetic tree. As expected, phylogenetic locations of the ureases of the diatoms (serving as an outgroup) were separated from that of the oomycetes. The oomycete ureases were allocated into three phylogenetically distinct clades (Fig. 4): (i) the clade of Pythium, Phytophthora, Phytopythium, and Hyaloperonospora species; (ii) the clade of Aphanomyces and Saprolegnia species; and (iii) the clade of Albugo species. Most of the organisms contain one copy of the urease-encoding gene, except A. invadans (three copies) and P. insidiosum strain Pi45 (two copies). Four urease-encoding sequences from the P. insidiosum strains Pi07, Pi35, and Pi45 were grouped together, and placed more proximally to non-insidiosum Pythium, Phytophthora, Phytopythium, and Hyaloperonospora species than to other oomycete species.