Genetic characterization of Pseudomonas‘NZI7’– a novel pathogen that results in a brown blotch disease of Agaricus bisporus

Authors


Godfrey New Zealand Institute of Crop & Food Research Ltd, Private Bag 4704, Christchurch, New Zealand (e-mail: godfreys@crop.cri.nz).

Abstract

Aims: To characterize a novel pseudomonad isolate capable of causing brown blotch disease of Agaricus bisporus.

Methods and Results: Using the white-line-in-agar (WLA) assay, fluorescent pseudomonads isolated from a New Zealand mushroom farm were screened for the lipodepsipeptide tolaasin, a characteristic marker of Pseudomonas tolaasii. One isolate, NZI7, produced a positive WLA assay and caused brown lesions of A. bisporus comparable with those produced by Ps. tolaasii. However, genetic analysis suggested that Ps. tolaasii and NZI7 were genetically dissimilar, and that NZI7 is closely related to Pseudomonas syringae. Nucleotide sequence analyses of a gene involved in tolaasin production indicated that similar genes are present in both NZI7 and Ps. tolaasii.

Conclusions: NZI7 represents a novel Pseudomonas species capable of causing brown blotch disease of A. bisporus.

Significance and Impact of the Study: Phenotypic identification of Ps. tolaasii based on A. bisporus browning and positive WLA may have limited specificity.

INTRODUCTION

The fluorescent bacterium Pseudomonas tolaasii Paine causes a brown blotch disease of commercial mushrooms (Tolaas 1915; Paine 1919). This disease results in significant crop losses worldwide, especially of Agaricus brunnescens Peck (A. bisporus (Lange) Imbach), the ‘button mushroom’ (Fermor et al. 1991). Although correlation has been established between environmental conditions and disease symptoms, there are still no effective strategies for blotch control.

Pseudomonas tolaasii produces an extracellular toxin, tolaasin, that is the primary bacterial agent responsible for disease symptoms (Brodey et al. 1991; Nutkins et al. 1991; Rainey et al. 1991). Tolaasin is a low molecular weight lipodepsipeptide (LDP), the production of which requires expression of at least three high MW polypeptides, TL1, TL2, and TL3 (Rainey et al. 1993). When Ps. tolaasii is cultured in close proximity to a second pseudomonad, Pseudomonasreactans’, a white precipitate forms between the colonies, defining the white-line-in-agar (WLA) assay (Wong and Preece 1979). Tolaasin forms a dense white precipitate with the ‘white line inducing principle’ (WLIP), an LDP produced by Ps. ‘reactans’ (Mortishire-Smith et al. 1991). Isolates of Ps. tolaasii that are defective for either TL1, TL2 or TL3 fail to form a white precipitate, suggesting that tolaasin is the sole agent that reacts with WLIP (Rainey et al. 1993).

Although Ps. tolaasii is included in the Approved Lists of Bacterial Names (Skerman et al. 1989), its validity as a distinct species and its place within the Pseudomonas genus have not been fully resolved. Bergey’s Manual of Systematic Bacteriology (Palleroni 1984) includes Ps. tolaasii in Pseudomonas (Section V), but hybridization experiments involving 23S rRNA—DNA place Ps. tolaasii in Section I of the pseudomonads (De Vos et al. 1985). Although conserved regions in the gene encoding the outer-membrane protein (OprF) in both Pseudomonas fluorescens and Ps. tolaasii support the close relatedness of these two species, Ps. tolaasii is distinguishable from Ps. fluorescens (De Mot et al. 1994). Pseudomonas tolaasii homogeneity and differentiation from both Ps. fluorescens and Ps.reactans’ was demonstrated by Goor et al. (1986) using substrate utilization tests, electrophoresis of soluble proteins and DNA:DNA hybridization experiments. Goor et al. (1986) also identified several Ps. tolaasii strains that were WLA negative and non-pathogenic to mushrooms. Thorn and Akihiko (1996) addressed the homogeneity of geographically-diverse isolates of Ps. tolaasii using DNA restriction fragment length polymorphisms (RFLP) and partial sequence analysis of PCR-amplified 16S rRNA genes. These experiments indicated that isolates of Ps. tolaasii were highly similar to each other, but readily distinguished from Ps. fluorescens isolates from different environments. Homogeneity of Ps. tolaasii was further demonstrated by Moore et al. (1996) based on comparison of the small subunit rRNA (16S rRNA) nucleotide sequences. This phylogenetic study included 24 validly-described species of the genus Pseudomonas (sensu stricto), and showed that the 16S rRNA hyper-variable region 2 was identical in Ps. tolaasii isolates but differed between Ps. tolaasii and all other species within the ‘Pseudomonas fluorescens lineage’ (Moore et al. 1996).

The physiological properties of Ps. tolaasii are similar to those of Ps. fluorescens and other fluorescent pseudomonads. However, Ps. tolaasii has routinely been distinguished from these organisms by its pathogenicity to mushrooms and by a positive WLA (Rhodes 1959; Wong and Preece 1979; Zarkower et al. 1983; Palleroni 1984; Janse et al. 1992). As far as is known, there have been no published reports of a non-Ps. tolaasii isolate causing a positive WLA.

In this paper, the characterization of a fluorescent pseudomonad, designated NZI7, that has been isolated from a New Zealand (NZ) mushroom farm, is described. NZI7 resembles Ps. tolaasii in producing brown lesions in mushroom bioassays and inducing a positive WLA. However, genetic examinations demonstrate that NZI7 and Ps. tolaasii are markedly different.

MATERIALS AND METHODS

Bacterial strains and culture conditions

Bacterial isolates used in this study are listed in Table 1. All strains used were maintained at −80°C in Kings B medium (KB) (King et al. 1954) with a final concentration of 20% (v/v) glycerol. Strains were transferred onto KB agar supplemented with 1·5% agar, incubated at 28°C for 16 h and then maintained at 4°C for short-term use. For cloning experiments involving Escherichia coli (DH5α), transformants were cultured on Luria Bertani medium (Difco) supplemented with 1·5% agar (LBA), and antibiotics (ampicillin, Ap (50 μg ml–1), kanamycin, Km (30 μg ml–1)) as necessary.

Table 1.   (a) Bacterial strains used in this study. (b) Reference sequences (with accession numbers) of 14 validly-described pseudomonads (Moore et al. 1996) used for phylogenetic analyses Thumbnail image of

WLA assay

The ability to produce tolaasin was determined using the WLA assay on KB agar as described by Wong and Preece (1979). Bacterial colonies were inoculated at a distance of 7 mm from a centre streak of the indicator bacterium, Ps.reactans’ (NCPPB 1311). Pseudomonas tolaasii NCPPB 2192T served as a positive control. Plates were inverted and incubated at 28°C. A white precipitation line was observable after 24–48 h of incubation.

Mushroom bioassay for pathogenicity

Bioassays were performed as described by Gandy (1968) using healthy, one-day-old A. bisporus. Cubes (1 cm3) of cap tissue were obtained with sterile scalpel blades and placed in a sterile Petri dish containing a 50 mm paper filter dampened with 800 μl of sterile ddH2O. Cubes were placed 2 cm apart to eliminate cross contamination by motile pseudomonads. Bacterial strains were cultured using KB medium to a density of approximately 1 × 109 cfu ml–1. A 50 ml aliquot of cells (or uninoculated control) was placed on each cube and incubated (undisrupted) at ambient temperature for 24 h. Mushroom caps incubated with bacterial isolates were scored for the degree of browning and decay relative to the pathogenic isolate Ps. tolaasii NCPPB 2192T.

Analytical Profile Index (API) 20 NE strip analysis

The API 20 NE micro-method for the identification of non-fastidious Gram-negative rods, using eight conventional biochemical and 12 carbohydrate assimilation tests, was performed as described by the manufacturer (BioMerieux). Numerical profiles obtained from the NZI7 and Ps. tolaasii strains tested (Table 1) were compared with the profiles stored in the 1999 Analytical Profile Index Software database (BioMerieux).

Genomic DNA isolation, PCR and DNA sequencing

DNA was isolated from pure cultures of bacteria using the Wizard Genomic DNA Isolation Kit (Promega). PCR amplifications were carried out in a Perkin Elmer 9700 thermocycler (Perkin Elmer, Auckland, NZ). Unless stated otherwise, a standard PCR reaction mixture (25 μl total) was used, consisting of 1 × buffer (10 mmol–1 Tris-HCl, pH 9·0, 50 mmol–1 KCl, 2·5 mmol–1 MgCl2, 0·01% gelatin and 0·1% Triton X-100), deoxyribonucleotide triphosphates (dATP, dCTP, dGTP, dTTP) at a final concentration of 200 μmol–1, 0·6 U Taq DNA polymerase (Roche Molecular Biochemicals), oligonucleotide primers (Table 2) at a final concentration of 2 mmol–1, and 100 ng of template DNA. PCR consisted of 30 cycles of 1 min 94°C, 1 min 55°C and 1 min 72°C. Prior to cycling, samples were heated at 94°C for 5 min, and as part of the terminal cycle, the extension step was increased to 5 min, 72°C. Contaminating primers and dNTPs were removed from PCR products using the High Pure™ PCR Product Purification Kit (Roche Molecular Biochemicals, MT Wellington, Auckland, NZ). PCR amplicons were directly sequenced with a variety of oligonucleotide primers (Table 2) using the Big Dye Terminator Kit and ABI Prism 3TIXLCPE (PE Biosystems). All 16S rRNA genes sequenced in this study were confirmed by determining contiguous overlapping sequences of PCR-DNA produced. The nearly complete 16S rRNA gene sequences determined in this study have been deposited with the GenBank Database under the accession numbers listed in Table 1.

Table 2.   Oligonucleotide PCR primers used in this study Thumbnail image of

Phylogenetic analysis

Nucleotide sequences were initially compared with those stored in GenBank using BLASTN (version 2,0.14) (Altschul et al. 1997) to determine species similarities. Clustal W (version 1.60) (Thompson et al. 1994) was used for sequence alignments of isolates sequenced in this study and selected validly-described species of the genus Pseudomonas (sensu stricto). Sequence dissimilarities were converted to evolutionary distances according to Jukes and Cantor (1969). The construction of neighbor-joining trees (Saitou and Nei 1987) and bootstrap analysis of 1000 re-samplings (Felsenstein 1985) were performed using the software package Treecon for Windows (Van de Peer and De Wachter 1994).

Cloning and primer design of DNA flanking Tn5 insertions in Ps. tolaasii

Transposon (Tn5)-mutagenized Ps. tolaasii strains, defective in tolaasin production (Rainey et al. 1993), were generously supplied by Paul Rainey, Oxford University, UK. To determine the nucleotide sequence neighbouring the Tn5 transposon in Ps. tolaasii PT182 (Rainey et al. 1993), chromosomal DNA was prepared using the Wizard Genomic DNA Preparation Kit (Promega), digested with restriction endonuclease SalI and ligated to pBluescript KS (Stratagene, Birkenhead, Auckland, NZ). Ligation mixes were used to transform E. coli DH5α and plated onto LB medium supplemented with Ap and Km. Plasmid DNA from resultant transformants was prepared and the nucleotide sequence of the SalI fragment flanking the Tn5 transposon was determined using T3 and T7 primers (Table 2).

PCR primers pvd1 and pvd2 (Table 2) were designed from the DNA sequence flanking the Tn5 insertion in Ps. tolaasii strain PT182 using the Primer Express software (Perkin Elmer).

REP PCR

The primers (REP1R-I and REP2-1 (Table 2)) and protocols used for REP PCR were those described by De Bruijn 1992). Template chromosomal DNA (1 ng) was used for PCR reactions.

MLEE

MLEE methods used were those described by Selander et al. (1986) for enzymes glucose-6-phosphate isomerase (GPI), glucose-6-phosphate dehydrogenase (G6PDH), malate dehydrogenase NADP (ME) and 6-phosphogluconate dehydrogenase (6PGDH). Scoring of electromorphs was performed as described by Selander et al. (1986), and cluster analysis was carried out using unweighted pair group algorithm averages with the S-Plus statistical analysis package (version 4·5; MathSoft Inc., Seattle, WA, USA).

RESULTS

Survey of pathogenic pseudomonads from an NZ mushroom farm

As part of a population study, 132 fluorescent pseudomonads were isolated from nine locations in and around a mushroom shed on a farm from the South Island of New Zealand (data not shown). One isolate discovered during this screen, designated NZI7, was WLA(+) (reacting with Ps. ‘reactans’) (Fig. 1) and caused brown blotch lesions in bioassay consistent with Ps. tolaasii (Fig. 2).

Figure 1.

 Screening Pseudomonas strains using the WLA (Wong and Preece 1979). Vertical inoculations are Ps.reactans’ NCPPB1311. Lipodepsipeptide precipitation between Ps. tolaasii strains and Ps. ‘reactans’ NCPPB1311 is observed as a positive ‘white line’. (1) Pseudomonas tolaasii NCPBB2192; (2) Ps. tolaasii NCPPB1116; (3) Ps. tolaasii NCPPB 741; (4) Ps. tolaasii NZ027; (5) Ps. tolaasii NZ032; (6) NZI7; (7) Ps. putida; (8) Ps. fluorescens; (9) Ps.reactans’ NCPPB1311; (10) Ps. gingeri NCPPB3147

Figure 2.

 Cube pathogenicity bioassay to determine pathogenic pseudomonads able to induce brown blotch of Agaricus bisporus tissue. A positive reaction is indicated by a sunken brown lesion with uniform brown discoloration. (1) Pseudomonas tolaasii NCPBB2192T; (2) Ps. tolaasii NCPPB1116; (3) Ps. tolaasii NCPPB741; (4) Ps. tolaasii NCPPB2325; (5) Ps. tolaasii NZ027; (6) Ps. tolaasii NZ032; (7) NZI7; (8) inoculated with KB; (9) inoculated with ddH2O

Comparison of NZI7 with Ps. tolaasii isolates

Dissimilarity between NZI7 and Ps. tolaasii was initially observed using API-20NE assays (Table 1). Examination and enumeration of polymorphic variations using MLEE of the enzymes GPI, G6PDH, ME and 6PGDH also revealed that NZI7 had different allelic profiles to Ps. tolaasii strains (Table 1).

REP-PCR produced indistinguishable banding patterns for Ps. tolaasii strains. However, NZI7 had a distinctly different profile (Fig. 3). Figure 4 shows the single shortest neighbor-joining tree of 16S rRNA gene sequences from 14 selected validly-described species of the genus Pseudomonas (sensu stricto) (Moore et al. 1996) with NZI7 and Ps. tolaasii isolates. When the 16S rRNA gene nucleotide sequence from NZI7 was compared with those retrieved from Genebank, highest nucleotide identity (97%) was observed to Ps. syringae pv. savastanoi (ATCC 13522T).

Figure 3.

 REP-PCR patterns of Pseudomonas strains. Lane 1: 1 kb size marker (Life Technologies); lane 2: Ps. tolaasii NCPPB2192T; lane 3: Ps. tolaasii NCPPB1116; lane 4: Ps. tolaasii NCPPB741; lane 5: Ps. tolaasii NZ027; lane 6: Ps. tolaasii NZ032; lane 7: NZI7; lane 8: Ps. gingeri NCPPB3147T; lane 9: Ps.reactans’ NCPPB1311

Figure 4.

 The inferred phylogenetic relationships among NZI7 and Pseudomonas tolaasii isolates with other species of Pseudomonas (sensu stricto). Evolutionary distances were determined with pairwise dissimilarities of the 16S rRNA gene sequences, and the dendrogram was generated using the neighbor-joining algorithm. Two major intrageneric clusters and five evolutionary lineages are defined, as described by Moore et al. (1996). Bootstrap proportions of confidence are represented as percentages for those branchings with values greater than 50%

Lipodepsipeptide gene similarity between NZI7 and Ps. tolaasii

Because NZI7 was observed to give a positive WLA, a comparison of the LDP from both NZI7 and Ps. tolaasii was made. From the nucleotide sequence flanking Tn5 in Ps. tolaasii PT182, primers were designed (pvd1, pvd2) to amplify a fragment within the TL cluster from Ps. tolaasii NCPPB2192T. PCR using pvd1 and pvd2 produced a fragment from NZI7 (NZI7 pvd) that, once translated, shared 97% amino acid identity to a similarly-translated amplicon from Ps. tolaasii NCPPB2192T (2192pvd) (Fig. 5a). The translated NZI7 pvd nucleotide sequence was compared with amino acid sequences within existing databases, and amino acid identity (72%) was observed between NZI7 pvd and a syringomycin synthetase (syrE) gene from Ps. syringae pv. syringae (Guenzi et al. 1998) (Fig. 5b).

Figure 5.

 Amino acid comparison of translated NZI7 pvd nucleotide sequence with (a) translated Pseudomonas tolaasii 2192 pvd and (b) translated Ps. syringae pv. syringae syringomycin synthetase (Genebank accession T14593). Shaded regions indicate amino acid identity

DISCUSSION

Pseudomonas tolaasii has traditionally been characterized as a fluorescent pseudomonad that results in brown blotch disease of A. bisporus and produces a white precipitation line when cultured next to Ps. ‘reactans’. The WLA assay was initially used in this study to establish the prevalence of Ps. tolaasii amongst fluorescent pseudomonads from a New Zealand mushroom farm. One isolate, designated NZI7, gave a positive WLA and formed brown lesions in mushroom bioassay consistent with those produced by Ps. tolaasii; it was therefore initially assumed to be a Ps. tolaasii isolate. However, differences in API NE 20 biochemical profiles suggested NZI7 to be unique. While the carbon assimilation tests included in the API 20 NE strip do not effectively discriminate between isolates of Ps. tolaasii, or between Ps. tolaasii isolates and other Ps. fluorescens isolates, there were clear profile differences between isolates of Ps. tolaasii and NZI7. This was further supported by genetic analysis (MLEE and REP-PCR), and although other geographically-unlinked environmental Ps. tolaasii isolates (NZ027, NZ032) were genetically similar to previously identified Ps. tolaasii isolates, NZI7 was consistently an outlier.

To determine the phylogenetic relationship of NZI7 to Ps. tolaasii, nucleotide sequence analysis of the 16S rRNA gene was performed. The general topology of the tree generated in this study (Fig. 4) is the same as that described by Moore et al. (1996), in which partial and nearly complete 16S rRNA genes were applied in a polyphasic study of natural intrageneric relationships of the genus Pseudomonanas (sensustricto). Moore et al. (1996) argued that there are at least two distinct intageneric divisions within the genus Pseudomonas designated (i) the ‘Ps. aeruginosa intrageneric cluster’ and (ii) the ‘Ps. fluorescens intrageneric cluster’. NZI7 and Ps. tolaasii isolates are observed within the ‘Ps. fluorescens intrageneric cluster’. However, within this cluster, NZI7 groups between the Ps. syringae lineage and the Ps. cichorii lineage, clearly external to the closely-clustered Ps. tolaasii isolates.

The observed relatedness of NZI7 to the Ps. syringae lineage is not surprising. Pseudomonas syringae has long been considered to be a genetically-diverse species, sub-classified into approximately 50 pathovars according to plant pathogenicity (Dye et al. 1980; Rudolph 1995) with many strains producing LDP toxins (Segre et al. 1989; Ballio et al. 1990, 1991, 1994; Fukuchi et al. 1990; Isogai et al. 1990a, b; Fukuchiet al. 1992; Scaloni et al. 1997). NZI7 is probably an LDP-producing Ps. syringae isolate that has entered the mushroom environment. If this is the case, it is, as far as is known, the first report of a Ps. syringae identified within a mushroom farm that is able to induce bacterial brown blotch of cultivated A. bisporus.

As NZI7 exhibited a positive WLA typically associated with a positive identification of Ps. tolaasii tolaasin, genes involved in LDP production were investigated and compared between NZI7 and Ps. tolaasii. PCR amplification of a coding region of DNA between the TL2 and TL1 of the Ps. tolaasii tolaasin gene cluster was targeted using PT182 (a Ps. tolaasii Tn5 generated mutant deficient in tolaasin production (Rainey et al. 1993)). This region contains a gene that does not affect the expression of TL1, TL2 or TL3, but is nevertheless required for tolaasin synthesis (Rainey et al. 1993). The resulting 188 bp PCR amplicons from NZI7 (NZI7 pvd) and Ps. tolaasii NCPPB2192T (2192 pvd) were translated, and comparison of amino acid composition resulted in strong identity (95·2%) over the entire region (Fig. 5). However, when NZI7 pvd was compared with the syringomycin synthetase (syrE) gene from Ps. syringae pv. syringae, identity was significantly lower (72%) (Fig. 5). Taken together, results from 16S rDNA analysis, MLEE and REP-PCR suggest that NZI7 is not a Ps. tolaasii isolate, due to its distinctly different genetic characteristics. Phylogenetic analysis based on the 16S rRNA gene shows NZI7 to cluster most closely to the Ps. syringae lineage. However, NZI7 pvd appears more similar to tolaasin than to syringomycin. There are at least two possible explanations for the apparently contradictory data. First, NZI7 may be a Ps. syringae isolate that has acquired the tolaasin gene locus responsible for A. bisporus pathogenicity by horizontal transfer amongst bacterial populations present in a mushroom farm. A second possibility is that NZI7 has acquired a mosaic lipodepsipeptide, again through horizontal gene acquisition. Studies have revealed that horizontal transfer and recombination of virulence genes plays a major role in generating genetic diversity amongst bacterial species (Kehoe et al. 1996), and a recent study has suggested that pathovars of Ps. syringae have the ability to acquire at least one gene via horizontal gene transfer (Sawada et al. 1999). Further examinations of the genes involved in LDP production in Ps. tolaasii and NZI7 will be required before conclusions can be drawn on the origins of the pathogenicity genes in NZI7. It should also be noted that a closer investigation of Ps. syringae isolates which produce LDP may identify other LDPs that form a precipitate with WLIP in the WLA.

From these findings, it is suggested that NZI7 represents a novel Pseudomonas species (which appears most closely related to Ps. syringae) capable of causing brown blotch disease of A. bisporus. It is not the intention of the authors to speciate NZI7, but rather to report that identification of Ps. tolaasii based on A. bisporus pathogenicity and positive WLA may have limited specificity, based on the present findings. Additionally, as PCR primers directed to Ps. tolaasii tolaasin also amplify a similar product in NZI7, phenotypic and genetic tests based on LDP production may not satisfactorily assign Ps. tolaasii identity. It is suggested that partial 16S rRNA sequencing of the hyper-variable region 2 (Moore et al. 1996) should be used for confirmation of Ps. tolaasii identity.

The finding of NZI7 that is pathogenic to A. bisporus sheds light on the epidemiological complexity of bacterial blotch, and highlights the need for continued characterization of bacteria causing disease of cultivated mushrooms. Such studies may yield insights into the genetic origin(s) of Ps. tolaasii pathogenicity factors; they may also identify other brown blotch-causing organisms similar to NZI7, as well as their mode of transmission, and the susceptibility of particular mushroom crops.

Acknowledgements

Funding was provided by the PGSF and support and samples were provided by the Commercial Mushroom Growers Federation (New Zealand) Ltd. The authors also thank Paul Rainey for discussion on NZI7 and for generously supplying Ps. tolaasii Tn5 strains used in this study, Hamish Reid for initial Pseudomonas isolation, Meredith Williams for initial investigation into PT182 and Simon Bulman for reading the manuscript.

Ancillary