MOLECULAR GENETIC VARIATION IN POPULATIONS OF VERTICILLIUM FUNGICOLA ON THE CULTIVATED BUTTON MUSHROOM AGARICUS BISPORUS
P. D. Collopy1, M. L. Largeteau-Mamoun2, C. P. Romaine1*, and D. J. Royse1*
1Department of Plant Pathology, The Pennsylvania State University, University Park, PA, 16802, USA; 2Mushroom Research Station, INRA-Bordeaux, Villenave d'Ornon Cedex, France
*D. J. Royse: < and C. P. Romaine: <
ABSTRACT
Molecular genetic analyses were performed on 40 isolates of Verticillium fungicola collected from various Pennsylvania mushroom farms in 1999 and 28 isolates of Verticillium spp collected during the last 50 years from various geographic locations. Two molecular approaches (sequence analysis of internal transcribed spacers 1 and 2 [ITS-1 and 2] and 5.8 S regions of the nuclear ribosomal DNA [rDNA] transcriptional unit and random amplified polymorphic DNA [RAPD]) were performed for the 68 isolates of Verticillium spp. Sequence data indicated the existence of a geographic isolation between North American and European isolates. Overall, isolates with unique rDNA sequences produced unique RAPD patterns. Results from six decamer RAPD further supported the presence of a clonal population of V. fungicola among Pennsylvanian isolates. All isolates with rDNA sequence and RAPD grouping similarity to ex-type strains V. fungicola var. aleophilum and V. fungicola var. fungicola displayed the highest level of virulence in assays. Based on rDNA sequence and RAPD analyses, isolates displaying reduced or no virulence were distantly related to these two varieties. All results obtained for the analyses of ex-type strain V. fungicola var. flavidum indicate that this fungal isolate should form a distinct species rather than a variety of V. fungicola.
INTRODUCTION
Verticillium fungicola (Pruess) Hassebrauk is the causal agent of dry bubble disease of the cultivated button mushroom Agaricus bisporus (Lange) Imbach and is worldwide in distribution. Taxonomic revisions of this species placed V. fungicola into Section Prostrata and divided it into three different varieties based on morphological taxonomy and pathogenicity studies (Gams and Zaayen 1982, Zaayen and Gams 1982). These three varieties were named fungicola, aleophilum, and flavidum. Conidia of V. fungicola var. aleophilum are variable in size and shape much like V. fungicola var. fungicola. The only consistent difference between the two varieties is the temperature maximum for optimal growth. The variety flavidum differs mainly from the other two varieties in its lower optimum temperature, presence of sclerotia, yellowish colonies, pungent odor, and non-pathogenic abilities on A. bisporus(Gams 1995, Gams and Zaayen 1982).
Bidochka et al. (1999) investigated the production of hydrolytic enzymes and performed pathogenicity assays against a plant and an insect host as a means for gaining insight into the evolution of pathogenicity of Verticillium spp. In addition, the authors performed a sequence analysis of the genus Verticillium using the internal transcribed spacer region 1 (ITS-1) and a portion of the relatively more conserved small nuclear subunit (NS) of ribosomal RNA (rDNA). These data were used to provide phylogenetic information among 13 different species of Verticillium. Two isolates of V. fungicola from two hosts (mushroom and Lepidoptera) were analyzed for their nuclear rDNA phylogeny and extracellular enzyme production. Despite the physiological similarities of these two isolatesin their protease, pectinase, and chitinase profiles, rDNA sequence analyses of the ITS-1 revealed a 49.3% divergence between isolates. These results did not produce monophyly of V. fungicola for any of the parsimony trees generated; thus, the possibility of high levels of diversity exist within populations of V. fungicola. Therefore, Verticillium may be of polyphyletic origin and, consequently, a form genus.
Genetic variability within isolates of V. fungicola was previously investigated by Bonnen and Hopkins (1997). The authors used a random amplified polymorphic DNA (RAPD) analysis to examine the within-species variation of this organism. Sixty-six isolates collected over a 45-year period (1950-1995) were compared with respect to colony morphology, response to fungicides, virulence, geographic origin, and RAPD grouping. When all 66 isolates were analyzed together, no correlation was found between RAPD grouping, colony morphology and virulence. However, all recently collected isolates (1993-1995) were similar in their fungicide responses, virulence, colony morphologies, geographic origins, and RAPD groupings. This high level of homogeneity within recent isolates of V. fungicola greatly contrasts the results obtained by Bidochka et al. (1999) andsuggests the possibility of a clonal population of V. fungicola is responsible for recent dry bubble outbreaks on commercial mushroom farms.
In the present study, our objective was to examine the genetic variability within isolates of V. fungicola responsible for past and recent dry bubble outbreaks on mushroom farms. Sequence analyses of a portion of the nuclear rDNA, specifically the internal transcribed spacers-1 and -2 (ITS-1 and -2) and 5.8 S rRNA gene regions were performed to examine the genetic diversity within this species. Isolates collected from various geographic locations during the past 50 years were included to understand how the genetic diversity of this pathogen has changed over time. A clearer understanding of the genetic structure of temporal V. fungicola populations is useful to provide clues on speciation and biogeography. In addition to comparing rDNA sequences of various isolates collected over the past 50 years, a random amplified polymorphic DNA (RAPD) analysis was performed. This allowed for the genetic assessment of isolates collected during 1999 from 27 southeastern Pennsylvania mushroom farms.
MATERIALS AND METHODS
Cultures
Forty isolates of V. fungicola were collected from infected basidiomes on 27 mushroom farms in Pennsylvania in 1999. An additional 28 isolates from various culture collections were included in this study as representative isolates collected during the past fifty years from various geographic locations. All isolates were regenerated from single conidia to ensure genetic uniformity and pure culture before molecular analyses. Ex-type strains for V. tenerum and all varieties of V. fungicola were obtained from Centraalbureau voor Schimmelcultures (Baarn, The Netherlands).
DNA extraction and PCR amplification of ITS-1 and -2 regions and 5.8 S rDNA
Isolates of Verticillium spp. were grown at room temperature in 50 ml of 2% malt extract broth contained in 250-ml Erlenmeyer flasks. Fungal mycelial mats were collected by vacuum filtration, washed with sterile ddH2O, and allowed to air-dry. Genomic DNA was extracted using the Puregene DNA isolation kit® (Gentra Systems, Inc., Minneapolis, MN). DNA (10 ng) was amplified in reaction mixtures (23.5 µl) containing 50 mM Tris (pH 8.3), 0.25 mg/ml BSA, 2% sucrose, 0.1 mM Cresol Red, 0.2 mM dNTPs, 2 mM MgCl2, 0.1% triton, 1.0 U Taq DNA polymerase (Promega, Madison, WI), and 0.5 µM for each primer ITS-1 (5' TCC GTA GGT GAA
CCT GCG G 3') (White et al. 1990) and ALR0 (5' CAT ATG CTT AAG TTC AGC GGG 3') (Figure 1). Reactions were performed in 0.2 ml thinwall PCR tubes using a PTC-100TM (MJ Research, Inc., Watertown, MA) with the following program: 1 cycle at 94C for 2 min.; 35 cycles of 94C for 15 sec., 57C for 30 sec., and 72C for 45 sec.; and 1 cycle of 72C for 4 min.
Figure 1. Map of ITS and 5.8 S rDNA. Filled boxes indicate exons; introns are unfilled. Labeled arrows indicate primers used for PCR amplification and sequencing.
PCR product purification and sequencing reaction
Amplified PCR products were directly purified using the Wizard® PCR prep system (Promega, Madison, WI). Purified amplification products (10 ng) were sequenced in 5 µl reactions containing 1µl Applied Biosystems (ABI) BigDye terminator (Perkin-Elmer, Foster City, CA), 1 µl HBD100 halfBD (Genpak Inc., Stony Brook, NY), and 0.5 mM PCR primer. Sequencing reactions were performed in 0.2 ml thinwall PCR tubes using a PTC-100TM (MJ Research, Inc., Watertown, MA) with the following program: 1 cycle of 92C for 90 sec., 50C for 15 sec., and 60C for 4 min.; and 25 cycles of 92C for 30 sec., 50C for 15 sec., 60C for 4 min.
Phylogenetic sequence analysis
Sequence data was analyzed using the Wisconsin Package (Genetics Computer Group, Madison, WI). DNA sequences of the ITS-1, ITS-2, and 5.8 S rDNA were aligned and compared independently for the entire set of V. fungicola isolates using the software for multiple sequence alignment Clustal W (Higgins et al. 1991, Thompson et al. 1994). Phylogenetic analyses were performed using PAUP, version 4.0 (Swofford 1999). Nucleotide distances were calculated by the Jukes Cantor (Jukes and Cantor 1969) and Kimura 2-parameter models (Kimura 1980) with gaps treated as missing data, and phylogenetic inferences were performed by the Neighbor-joining (NJ) method (Saitou and Nei 1987).
Representative sequence contigs from each clade resulting from the NJ analysis of isolates (Table 1) were used for comparison to the nucleotide database at the National Center for Biotechnology Information (NCBI, Bethesda, MD). Basic Local Alignment Search Tool for nucleotides (BLASTnr) was used to query all nucleotide databases at the NCBI (Altschul et al. 1990). Organisms with sequence matches for ITS-1, 5.8 S and ITS-2 rRNA gene sequences were used in a maximum parsimony (MP) analysis using the heuristic search option with gaps treated as missing and starting tree(s) obtained through stepwise addition. Duplicate matches for searches were not included in the analysis. Bootstrap tests with 1,000 replications (Felsenstein 1985) were conducted for phylogenetic analyses to examine reliability of interior branches and validity of trees obtained. Verticillium tenerum (Ni) was designated as the outgroup for all phylogenetic analyses.
RAPD analysis
DNA (10 ng) was amplified in reaction mixtures (15.0 µl) containing 50 mM Tris (pH 8.3), 0.25 mg/ml BSA, 2% sucrose, 0.1 mM Cresol Red, 0.2 mM dNTPs, 2 mM MgCl2, 1.0 U Taq DNA polymerase (Promega, Madison, WI), and 0.5 µM decamer primer (University of British Columbia, Vancouver, Canada). Reactions were performed in 0.2 ml thinwall PCR tubes using a PTC-100TM (MJ Research, Inc., Watertown, MA) with the following program: 1 cycle at 94C for 2 min. and 35 cycles of 94C for 1 min., 37C for 2 min., and 72C for 2 min. Products were visualized on 1.5% NuSieve® 3:1 agarose gels (BioWhittaker Molecular Applications, Rockland, ME) containing 0.05 µg ml-1 ethidium bromide. Primers that produced polymorphisms were chosen based on previous work by Bonnen and Hopkins (1997).
Pathogenicity assay
Mature A. bisporus basidiomes were harvested from The Pennsylvania State University Mushroom Test Demonstration Facility and inoculated the same day with a V. fungicola spore suspension. Eighteen representative isolates of 7 different RAPD groupings were used for this assay. Virulence tests were performed and rated for each isolate as previously described by Bonnen and Hopkins(1997). Stipes were removed from all mushrooms prior to inoculation to retard cap opening and basidiospore release. Caps were inoculated with a spore suspension containing 1 106 conidia/ml and rated as follows: 0 = no symptoms; 1 = faint rings around inoculation site; 2 = light brown rings around inoculation site; 3 = dark brown rings at inoculation site; 4 = dark brown rings, some sporulation and some pitting of the cap tissue at inoculation site; 5 = symptoms extend beyond the inoculation site, severe pitting of the tissue and profuse sporulation.
RESULTS
Amplified PCR products obtained from the ITS-1 and -2 gene regions and 5.8 S rDNA yielded fragments approximately 500 bp in length. Purification and sequencing of these products resulted in 433-476 bp of readable sequence for 68 samples after sequences were assembled and manually trimmed. An alignment of DNA sequences generated a total of 504 sites, including gaps, for phylogenetic analyses. Variation within the entire alignment was 36.7%, with internal variation of the ITS-1, 5.8 S, and ITS-2 gene regions of 55.8%, 11.0%, and 44.6%, respectively. Most variation observed within these regions could be attributed to sequences obtained from V. fungicola var. flavidum and V. tenerum. When these samples were removed from the alignment, variation within the total number of sites dropped to 18.2% and variation within the ITS-1, 5.8 S, and ITS-2 gene regions dropped to 35.8%, 0.0% and 19.9%, respectively. Transitions accounted for most of the observed substitutions with transition/transversion ratios of 1.2, 5.8, and 1.4 for the ITS-1, 5.8 S, and ITS-2 gene regions when all sequence contigs were compared. A transition/transversion ratio could not be calculated for the 5.8 S rDNA region when V. fungicola var. flavidum and V. tenerum were removed from the analysis due to a lack of variable sites within this region.
Independent nucleotide sequence alignments between ex-type strain varieties of V. fungicola showed variation of 16.5%. Most of the variation was accounted for by nucleotide differences between sequences of V. fungicola var. fungicola and V. fungicola var. flavidum (16.4%), followed by differences between sequences of V. fungicola var. aleophilum and V. fungicola var. flavidum (15.5%). Very low nucleotide sequence divergence (1.5%) was observed between ex-type strains V. fungicola var. fungicola and V. fungicola var. aleophilum (Table 1).
A Neighbor-Joining distance analysis for the entire set of sequence contigs (rDNA) placed isolates into four different groups. All isolates collected from Pennsylvanian mushroom farms in 1999 grouped into a clade with V. fungicola var. aleophilum and 14 isolates from The Pennsylvania State University culture collection (PSUCC). This clade, referred to as clade aleophilum, possessed no sequence variation except for a single transition substitution within the ITS-2 gene region of isolate DC130 from Korea. Similarly, no sequence variation was observed within all other clades derived from this analysis.
Results obtained from sequence analyses were further supported by RAPD analysis. All isolates collected from Pennsylvanian mushroom farms during 1999 gave identical RAPD patterns for all primers screened. RAPD analysis of isolates from all clades produced from NJ analysis revealed distinct banding patterns for each clade and further delineated isolates DC130 and the ex-type strain of V. fungicola var. aleophilum from all other isolates of the clade aleophilum. A total of seven different RAPD groupings were observed from all isolates identified as V. fungicola. Samples that produced unique RAPD patterns were consistent with those samples that yielded unique sequences of the ITS-1, 5.8 S, and ITS-2 rDNA. The unique RAPD pattern observed from V. tenerum was not counted as a RAPD grouping and was simply included as a positive control. Similarly, negative control reactions with water rather than template DNA did not produce DNA bands (data not shown). Previous work by Bonnen and Hopkins (1997) identified only 4 RAPD groupings. The additions of European isolates and the ex-type strains of V. fungicola to these analyses were responsible for the observed increase in the number of RAPD groupings.
Eighteen representative isolates from each of the observed 7 different RAPD groupings, were assessed for virulence levels on A. bisporus. All isolates of clades V. fungicola var. aleophilum and V. fungicola var. fungicola induced the most severe disease symptoms on A. bisporus pilei. Neither sterile H2O, used as a control, produced symptoms, nor did some isolates of the PSU disease collection (DC124 and DC131) and V. tenerum and V. fungicola var. flavidum.
Table 1: Nucleotide variation of the ITS-1, 5.8 S, and ITS-2 of type specimens ofVerticillium fungicola and V. tenerum.
Ex-type strains compared‡ / No. of variable sites / Total sites compared / Percent variationVfvf - AL - FL - Ni / 154 / 491 / 31.4%
Vfvf - AL - FL / 80 / 485 / 16.5%
Vfvf vs FL / 79 / 483 / 16.4%
AL vs FL / 75 / 485 / 15.5%
Vfvf vs AL / 7 / 466 / 1.5%
‡Vfvf=V. fungicola var. fungicola; AL=V. fungicola var. aleophilum; FL=V. fungicola var. flavidum; Ni=V. tenerum.
Results of queried sequences in the nucleotide databases at the NCBI yielded sequence matches exclusively for ITS-1, 5.8 S, and ITS-2 nuclear rDNA. The top three matches for each queried sequence were downloaded and used for phylogenetic analysis. A total of twelve sequences representing eight different fungal genera produced sequence matches with E values ranging from 0 - 1.0E-124 (Table 2). Isolates V-17 and Vfvf both produced matches for Cordyceps nutans and Paecilomyces spKACC40219. Isolate DC120 matched two identical rDNA sequences of Beauveria parasitica (only one included for phylogenetic analysis).
Alignment of sequences listed in Table 2 produced 494 sites with 47.2% variability within the entire alignment. A Maximum Parsimony analysis produced a single most parsimonious tree with a length of 471 steps, and consistency and retention indices of 0.73 and 0.88 respectively (Figure 1). Verticillium tenerum was chosen as the outgroup for this and all subsequent phylogenetic analyses.
Table 2: BLASTnr results of ITS-1, 5.8 S, and ITS-2 rDNA sequence data.
Organism / Isolate(s) match‡‡ / GenBank Accession(s) / E Value(s) / % identity (bp)Acremonium strictum ATCC 34717‡ / FL / ASU57671 / 1.0E-157 / 89.8
Beauveria parasitica 38002 / DC 120 / BPU35305, BPU19039 / 1.0E-135 / 94.5
Cordyceps nutans KACC500169 / V17,
Vfvf / AF224274 / 0.0E+00,
1.0E-164 / 91.4,
92.9
Cordyceps sp97009 / V17 / AB027378 / 0.0E+00 / 91.5
Gaeumannomyces sp 159 / DC 114 / GSP10040 / 1.0E-128 / 98.4
Helicoon sessile 129 / FL / HSU72605 / 1.0E-124 / 93.0
Orbilia luteorubella 125 / FL / OLU72604 / 1.0E-124 / 93.0
Paecilomyces spKACC40219 / V17,
Vfvf / AF224690 / 1.0E-168,
0.0E+00 / 92.2
92.1
Paecilomyces spKACC40221 / Vfvf / AF208525 / 0.0E+00 / 91.7
Tolypocladium parasiticum
(B. parasitica) ATCC52203 / DC 120 / TP58SRR / 1.0E-135 / 94.5
Verticillium lamellicola UAMH2055 / DC 114 / AF108471 / 1.0E-128 / 98.4
Verticillium lamellicola UAMH4785 / DC 114 / AF108480 / 1.0E-132 / 98.8
‡ Type specimen of A. strictum
‡‡ Vfvf=V. fungicola var. fungicola; AL=V. fungicola var. aleophilum; FL=V. fungicola var. flavidum; DC114=Pennsylvanian isolate collected in 1951; DC120=Pennsylvanian isolate collected in 1958
The top three matches for each queried sequence in NCBI nucleotide databases were downloaded and used for phylogenetic analysis. Duplicate matches for searches were not included in analyses.
A phylogenetic tree generated from a NJ analysis with 1000 bootstrap replicates was identical in tree topology had identical tree topology (data not shown).
DISCUSSION
A total of four clades were inferred from a NJ analysis of the ITS-1, 5.8 S, and ITS-2 sequences. Two of these clades, V. fungicola var. aleophilum and V. fungicola var. fungicola, were comprised entirely of isolates responsible for recent dry bubble outbreaks in Europe and North America with the exception of one Korean isolate (DC130). European and North American isolates were consistently divergent with regards to geographic location of origin. All recent European isolates had identical rDNA sequences to V. fungicola var. fungicola and, hence, were grouped into clade V. fungicola var. fungicola. Currently, no isolates of this cladehave been detected on North American mushroom farms. Similarly, isolates responsible for recent dry bubble outbreaks in North America have identical ITS-1, 5.8 S, and ITS-2 rDNA sequence to V. fungicola var. aleophilum and, therefore, are the primary constituents of that clade. Although V. fungicola var. aleophilum has been collected in Europe (Zaayen and Gams 1982), V. fungicola var. fungicola appears to be the predominant causal agent of dry bubble on A. bisporus on that continent. Phylogenetic analyses of the green mold mushroom pathogen, Trichoderma harzianum, have revealed a similar phenomenon of geographic isolation. Two biotypes (Th2 and Th4) of T. harzianum were responsible for recent green mold epidemics in Europe and North America, respectively (Castle et al. 1998, Chen et al. 1999, Muthumeenakshi et al. 1994, Ospina-Giraldo et al. 1998, Rinker 1994).