| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Plant Pathology, University of Arkansas, 217 Plant Sciences, Fayetteville, Arkansas 72701
Peter R. Johnston
Herbarium PDD, Manaaki Whenua-Landcare Research, Private Bag 92170, Auckland, New Zealand
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
A diverse collection of isolates identified as Colletotrichum acutatum, including a range of fruit-rot and foliar pathogens, was examined for mtDNA RFLPs and RFLPs and sequence variation of a 900-bp intron of the glutamine synthetase (GS) gene and a 200-bp intron of the glyceraldehyde-3-phosphate dehydrogenase (GPDH) gene. RFLPs of mtDNA, RFLPs of the 900-bp GS intron and sequence analysis of each intron identified the same seven distinct molecular groups, or clades, within C. acutatum sensu lato. Sequence analysis produced highly concordant tree topologies with definitive phylogenetic relationships within and between the clades. The clades might represent phylogenetically distinct species within C. acutatum sensu lato. Mating tests also were conducted to assess sexual compatibility with tester isolates known to outcross to form the teleomorph Glomerella acutata. Mating compatibility was identified within one clade, C, and between two phylogenetically distinct clades, C and J4. The C clade represented isolates from a wide range of hosts and geographic origins. J4 clade contained isolates from Australia or New Zealand recovered from fruit rot and pine seedlings with terminal crook disease. That isolates in two phylogenetically distinct clades were capable of mating suggests that genetic isolation occurred before reproductive isolation. No other isolates were sexually compatible with the mating testers, which also were in groups C and J4. Certain clades identified by mtDNA and intron analysis (D1, J3 and J6) appeared to represent relatively host-limited populations. Other clades (C1, F1 and J4) contained isolates from a wide range of hosts. Isolates described as C. acutatum f. sp. pineum were clearly polyphyletic.
Key words: anthracnose, genealogy, gloeosporioides, lupini, miyabeana, phylogeny
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, Lardner et al 1999). Although the fungus first was described as a fruit-rot pathogen (Simmonds 1965), C. acutatum also has been reported to infect vegetative tissues of woody and herbaceous crops, ornamentals, conifers and forage plants (Britton and Redlin 1995, Brown and Soepena 1994, Chelemi et al 1993, Dingley and Gilmour 1972, Maas and Palm 1997, Reed et al 1996, Smith 1993, Strandberg 2001, Yang and Sweetingham 1998, Zulfiqar et al 1996, http://NT.ars-grin.gov/fungaldatabases). Based on morphological descriptions, many diseases reported before 1965 to be caused by C. gloeosporioides (or one of its synonyms) could have been caused by C. acutatum, (Baxter 1983, Halsted 1893, Saccardo 1884, Shear and Wood 1913, Walker et al 1991).
C. acutatum sensu lato (s. l.) represents a species that encompasses a wide range of morphological and genetic diversity. Characterization of C. acutatum s. l. has been enhanced by the use of molecular markers, which have identified genetically distinct and perhaps biologically discrete groups among morphologically similar isolates (Buddie et al 1999, Correll et al 1994, Forster and Adaskaveg 1999, Freeman et al 2001, Guerber and Correll 2001b, Johnston and Jones 1997, Lardner et al 1999, Sreenivasaprasad et al 1992). Although C. acutatum (sensu Simmonds) has been identified traditionally by predominantly ellipsoidal or fusiform conidia often described as "pointed" at both ends (Aa et al 1990, Arx 1970, Dyko and Mordue 1979, Gunnell and Gubler 1992, Simmonds 1965, Sutton 1980), isolates with more or less atypical conidia, with one or both ends rounded, also have been identified as C. acutatum based on molecular criteria (Brown et al 1996, Forster and Adaskaveg 1999, Lardner et al 1999, Sreenivasaprasad et al 1994). The concept of C. acutatum s. l. thus has been introduced to accommodate isolates that cluster with C. acutatum and diverge from other species of Colletotrichum based on molecular criteria (Johnston and Jones 1997). The diversity encountered within this broad species has been problematic, however, for plant pathologists who need to identify accurately specific pathogens for disease control or quarantine and regulatory purposes. This taxonomic confusion has prompted the need for molecular tools appropriate for the identification of intraspecific diversity within this broad species or species complex.
Sequence analyses of the intergenic transcribed spacers (ITS 1, ITS 2) of rDNA have been valuable for delineating species of Colletotrichum (Adaskaveg and Hartin 1997, Brown et al 1996, Freeman et al 2001, Sherriff et al 1994, Sreenivasaprasad et al 1992, Vinnere 2002) and analysis of sequences in the D2 domain of the large subunit rDNA distinguished C. acutatum s. l. from other species of Colletotrichum (Johnston 2000, Johnston and Jones 1997). Examination of conidial and cultural morphology, however, further divided C. acutatum s. l. into six subgroups (Johnston and Jones 1997, Lardner et al 1999). Among these morphological subgroups were C. acutatum Group A (C. acutatum sensu Simmonds 1965), Groups B and C (C. acutatum-like fruit-rot pathogens), and Group D (pathogens of Lupinus spp.). Also identified as belonging to distinct morphological subgroups of C. acutatum s. l. were isolates causing terminal crook disease of pine (C. acutatum f. sp. pineum Dingley & Gilmour, 1972) and isolates of Glomerella miyabeana Spiers and Hopcroft (1993). Lardner et al (1999) provided additional support for these morphological subgroups by RAPD analysis. RAPD profiles from random or minisatellite primers, RFLPs of mitochondrial and ribosomal DNA (mtDNA and rDNA) and analysis of ITS 1 and ITS 2 sequences have identified diversity within C. acutatum s. l. and among isolates in Group A (C. acutatum sensu Simmonds) that were not readily differentiated by morphological criteria (Buddie et al 1999; Correll et al 1994, 2000; Forster and Adaskaveg 1999; Freeman et al 2001; Guerber and Correll 2001a, b; Lardner et al 1999; Sreenivasaprasad et al 1992). Freeman et al (2001) found sequences of ITS 2 more informative than ITS 1 for examining variation in C. acutatum sensu Simmonds (Group A sensu Lardner et al 1999) and identified four groups among 14 isolates. Relatively low variation in ITS sequences, however, has hindered the resolution of intraspecific fungal taxa, and has resulted in short branch lengths in phylogenetic tree topologies that often have had low bootstrap values or consistency indices (Balardin et al 1999).
Colletotrichum acutatum frequently was isolated from apple fruit in a survey of apple bitter rot in the southeastern United States (Shi et al 1996). In an examination of genetic and molecular diversity in C. acutatum from apple, the vast majority of isolates had a single mtDNA haplotype designated C1 (Correll et al 2000). RFLP analysis indicated that additional isolates from a range of hosts and geographical origins, including some isolates placed in Group A by Johnston and Jones (1997), shared this common mtDNA haplotype (Guerber and Correll 2001a). Identification of multiple VCGs and nuclear DNA RFLPs demonstrated genetic and molecular diversity among isolates within the widely occurring C1 mtDNA haplotype (Correll et al 2000).
Subsequent studies to examine sexual fertility indicated that a number of archived and contemporary isolates with the C1 mtDNA haplotype were self-sterile but capable of outcrossing. Laboratory crosses of many isolates of C. acutatum with mtDNA haplotype C1 produced the newly described teleomorph Glomerella acutata (Guerber and Correll 1997, 2001a). Also capable of mating was a subculture of the type strain of C. acutatum, ATCC 56816, which had a distinct mtDNA haplotype, J4 (Guerber and Correll 2001a). The potential for sexual reproduction and gene flow within and between genetic subgroups of the broad species C. acutatum s. l. and their influence on population structure remains largely unexplored. Additional cultural and molecular data are needed to resolve C. acutatum s. l. into biologically relevant groupings and to characterize their genetic relationships.
Sequence analysis of conserved protein coding genes, such as beta-tubulin and translation elongation factor 1-alpha, which contain highly variable introns, have been particularly helpful for the phylogenetic examination of fungal species (Geiser et al 1998, O'Donnell et al 1998, 2000) and for developing a phylogenetic species concept for fungi (Taylor et al 2000). The objective of the present study was to examine phylogenetic relationships in a diverse worldwide collection of isolates of C. acutatum s. l., using sequencing and RFLP analysis of introns from two independent genes (Stephenson et al 1997, Templeton et al 1992, Weeds et al 2000) and RFLPs of mtDNA. Mating compatibility within and between genetically distinct subgroups was evaluated in this study in an effort to further delineate the mating population of Glomerella acutata.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
), and foliar pathogens of spinach (Spinacia oleraceae), ornamental hosts including Magnolia sp., Rhododendron sp., Morus sp., Vinca minor, and leatherleaf fern (Rumohra adiantiformis) and the parasitic plant dodder (Cuscuta sp.).
|
|
|
|
|
|
). These four isolates were self-sterile isolates of C. acutatum that had been shown to produce the G. acutata teleomorph when mated with one another. These four mating reference strains have been deposited in the American Type Culture Collection (ATCC, Manassas, Virginia) as isolates MYA-662, -663, -664 and -665, and in the International Collection of Microorganisms from Plants (ICMP, Landcare Research, Auckland, New Zealand) as ICMP 14065, 14066, 14067 and 14068. Several self-fertile isolates of Glomerella miyabeana that have been shown to have a molecular affinity to C. acutatum (Lardner et al 1999, Johnston and Jones 1997) also were examined. Three self-fertile and two self-sterile isolates of C. gloeosporioides (Guerber and Correll 2001a) were included as phylogenetic outgroups for comparative purposes.
Isolates were recovered from infected host tissue by the authors, supplied by other laboratories or purchased from ATCC and routinely were stored aseptically at 4 C or -20 C on desiccated colonized filter paper removed from the surface of potato-dextrose agar (Correll et al 1986).
Mitochondrial DNA (mtDNA) RFLPs –
Total DNA of all isolates was extracted and digested with the restriction enzyme MspI. A subset of 40 isolates also was examined with a second restriction enzyme, HhaI. The resultant restriction fragments were separated electrophoretically in 0.8% agarose, capillary transferred to charged nylon membranes and probed with two equimolar nonoverlapping mtDNA clones (4u40 and 2u18) from an isolate of C. orbiculare, as previously described (Correll et al 1993, Guerber and Correll 2001a). The RFLP data further were interpreted by cluster analysis using the UPGMA (unweighted pair-grouping method with arithmetic averages) algorithm of NTSYS-PC (F. James Rohlf, Department of Ecology and Evolution, State University of New York, Stony Brook, New York 11794-5245).
Intron RFLPs –
RFLPs were examined for a 900-bp intron of the glutamine synthetase (GS) gene (Stephenson et al 1997) and a 200-bp intron of the glyceraldehyde-3-phosphate dehydrogenase (GPDH) gene (Templeton et al 1992, Weeds et al 2000). These DNA regions have been phylogenetically informative in preliminary studies to examine inter- and intraspecific diversity in the genus Colletotrichum (Liu and Correll 2000, Liu et al 2001). Sequences of the 900-bp GS introns of two mating reference isolates of C. acutatum, ATCC 56816 and ATCC MYA-662, have been published as GenBank accessions AF285765 and AF285766, respectively (Guerber and Correll 2001a).
The forward primer GSF1 (5'-ATGGCCGAGTACATCTGG-3') and the reverse primer GSR1 (5'-GAACCGTCGAAGTTCCAC-3') were used to amplify the 900-bp intron region of the GS gene (Stephenson et al 1997). The forward primer GDF1 (5'-GCCGTCAACGACCCCTTCATTGA-3') and the reverse primer GDR1 (5'-GGGTGGAGTCGTACTTGAGCATGT-3') were used to amplify a 200-bp intron region of the GPDH gene (Templeton et al 1992). A Hybaid DNA thermocycler (Hybaid US, Franklin, Massachusetts) was used to perform PCR amplifications of the introns, using 35 cycles of denaturation at 94 C and annealing at 60 C for 1 min, with final extension at 72 C for 3 min.
Amplified DNA was digested with the restriction enzymes PstI, MspI, HaeIII, HhaI, HindIII and HinfI singly and in combination. The restriction fragments were separated electrophoretically in a 3.0% agarose gel in 0.5x TBE buffer for 4 h at 140 V. DNA fragments between 40 and 1000 bp were scored for their presence or absence, and the data were converted into a binary character matrix used to build a similarity matrix based on simple matching coefficients. Cluster analysis was performed with UPGMA to determine relative RFLP similarities.
DNA sequencing – The 900-bp and 200-bp double-stranded intron amplification products were purified with the Qiagen MinElute® system (Qiagen Inc., Valencia, California) and used as templates in dideoxy termination sequencing reactions using the ABI Prism Dye Terminator cycle sequencing system (Applied Biosystems Inc., Foster City, California) in an MJ Research thermocycler (MJ Research Inc., Waltham, Massachusetts) with the thermal profile suggested by ABI. Sequencing reactions were performed directly from both strands using primers GSF1 and GSR1 for the 900-bp GS intron, and GDF1 and GDR1 for the 200-bp GPDH intron. Sequencing reaction products were purified to remove unincorporated nucleotides and primers using the ethanol precipitation method described in the ABI manual and purified reaction products were vacuum dried and stored at -20 C until use. Before loading in the sequencer, the products were resuspended using formamide loading dye. Reaction products were run on either an ABI 377 automated sequencer in a 6% polyacrylamide gel, or an ABI 3100 capillary sequencer, in the University of Arkansas DNA Core Facility lab.
Sequence alignment and phylogenetic analysis –
The 900-bp GS and 200-bp GPDH introns from 118 isolates of C. acutatum were sequenced and phylogenetic analyses were run using three isolates of C. gloeosporioides as outgroup. Sequences of each intron were entered into the Seqpup DNA sequence editor (available from the Web page of the University of Illinois, Department of Biology), and the combined data were aligned using ClustalX (Thompson et al 1997). Phylogenetic analysis, as well as basic statistics, were performed using PAUP* 4.0 beta 10 (Swofford 2002). Three methods of tree building were used: maximum parsimony (MP), neighbor joining (NJ), and maximum likelihood (ML). In all three methods, alignment gaps were treated as missing data in the phylogenetic analysis and tree topologies were evaluated by statistical confidence in bootstrap values (Felsenstein 1985). One thousand replicates were performed to examine the relative bootstrap support for each group in the resultant topologies. The Hasegawa-Kishino-Yano model (HKY85, Hasegawa et al 1985) was used for NJ and ML tree-construction methods.
Several phylogenetic models were examined including JK, K2P, F84 and HKY85 with similar results. The HKY85 model was used for the analysis presented. In MP and ML analyses, a heuristic search was employed and starting trees always were obtained by random sequence addition. For MP analyses, the heuristic search had these parameters: substitution model set to a transition/transversion ratio of 2; the HKY two parameter model variant for unequal base frequencies; starting branch length obtained using the Rogers-Swofford approximation method; substitution rates set to conform to a gamma distribution; and molecular clock was not enforced. Tree visualization and drawing were carried out with TreeView (Win32) version 1.5.2 (http://taxonomy.zoology.gla.ac.uk/rod/rod.html). ML analyses were carried out with the heuristic algorithm TBR of PAUP because the dataset was too large to be used with the exhaustive or branch-and-bound algorithms. ML settings were: number of substitution types = 2 (HKY85 variant), transition/transversion ratio = 2, kappa = 4.027. Assumed nucleotide frequencies for both introns and the combined dataset were (empirical frequencies) A = 0.25788, C = 0.31081, G = 0.20218 and T = 0.22914. Assumed proportion of invariable sites = none, distribution of rates at variable sites = equal and settings corresponded to the HKY85 model. The sequences of the 900-bp and 200-bp introns obtained for each isolate were combined with the Seqpup DNA sequence editor, and MP analysis was performed as above on the combined dataset.
The tree length, consistency index (CI), CI excluding uninformative characters, homoplasy index (HI), HI excluding uninformative characters, retention index (RI) and rescaled consistency index (RC) were recorded for all MP trees. Kishino-Hasegawa tests were performed to assess significant differences among a subsample of 100 trees. NJ tree distance matrices were calculated on the HKY85 model.
Mating studies –
Mating capability was assessed within and between subgroups of C. acutatum s. l. Preliminary tests identified six isolates that were highly fertile when crossed with one another in all 30 outcrossing combinations (Table II). These six isolates, used as mating reference testers, were self-sterile and had predominantly fusiform conidia consistent with C. acutatum sensu Simmonds and morphological Group A (Johnston and Jones 1997, Lardner et al 1999). Five of these isolates had been reported to form the teleomorph Glomerella acutata in pairwise crosses (Guerber and Correll 2001a). These mating testers were crossed with each of the 118 isolates selected for detailed examination (Table I).
|
) on a modified Czapek-Dox agar media, pH 7.8 (2 g NaNO3, 1 g K2HPO4, 0.5 g MgSO4·H2O, 0.5 g KCl, 0.01 g FeSO4 and 20 g agar, per liter). Mycelial plugs of the two parental isolates were placed opposite each other and approximately 1 cm from the edge of 9 cm Petri plates. Autoclaved flat birch toothpicks (Diamond Brands Inc., Minneapolis, Minnesota, or Forster Inc., Wilton, Maine) were placed on the agar surface in an "N" configuration to provide a substrate for perithecia. Mating plates were incubated at 20 C under constant illumination provided by 4–8 40-watt, cool white, fluorescent tubes. The margins of parental colonies of C. acutatum merged after 9–10 d. After 26–32 d, the mating plates were examined under a 30–60x stereomicroscope for the presence of perithecia. To assess their developmental status, masses of perithecia were scraped off the center toothpick and crush mounts were prepared in water and examined with 400x phase-contrast microscopy. A combination of a qualitative and a quantitative rating system was used to score the sexual fertility of each cross. Crosses were scored on a scale of 0–7 with zero = no structures observed at the zone of colony interaction on the center toothpick where the two colonies merged; 1 = small sterile globose structures (possible protoperithecia) present; 2 = sterile perithecia with beaks but no asci; 3 = perithecia containing sterile asci with no ascospores; 4 = asci with very few ascospores; 5 = many asci and ascospores but with few asci containing eight spores; 6 = abundant ascospores with many eight-spored asci, and 7 = ascospores oozing from perithecial ostioles. Ascospore viability was assessed for crosses with a fertility score of >4.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) (Fig. 1). Based on UPGMA analysis, isolates clustered into six major groups (designated groups C, D, E, F, J and K) with >84% mtDNA RFLP band similarity within groups (Fig. 2, Table III). Specific band patterns within mtDNA groups were given numerical haplotype designations (e.g., C1, C2, etc.). Compared to MspI, HhaI differentiated the same general RFLP haplotype groups, with minor variations. HhaI failed to differentiate isolates with haplotypes F4 and F7 from F1, placed isolate LLB17 in haplotype D2 instead of D3, and identified minor polymorphisms among isolates with MspI haplotype J3 (Table III).
|
|
|
|
|
|
, Foster and Adaskaveg 1999, Guerber and Correll 2001a, Shi et al 1996) and fruit-rot pathogens from Australia and New Zealand from avocado, cherimoya, citrus, grape, kiwi, pear, puriri, quince, squash, strawberry and tomato, which have been assigned to morphological group A by Lardner et al (1999). MtDNA group C also included terminal crook pathogens of pine, foliar and shoot pathogens of spinach and several ornamentals, and a pathogen of dodder infesting cranberry bogs (Table III).
Of the isolates examined in detail, mtDNA group J contained a total of 33 isolates with six different haplotypes that were 85–96% similar (Figs. 1, 2; Table III). Mitochondrial DNA RFLPs clearly differentiated a large subgroup of isolates with mtDNA haplotype J4. These J4 isolates included anthracnose pathogens from New Zealand isolated from guava, nashi, sapote, tomato and tree lupine and a number of isolates from pine seedlings with terminal crook disease from New Zealand and South Africa (Table I). ATCC 56816, a subculture of the type strain of C. acutatum that was isolated from papaya in Australia (Simmonds 1968) also had the J4 haplotype (Guerber and Correll 2001a). Five isolates from Florida from Persian lime, orange and leatherleaf fern and two lupine pathogens from France and Canada belonged to mtDNA haplotype J2. Five Key lime anthracnose (KLA) pathogens from Florida belonged to haplotype J3 and four lupine pathogens from New Zealand and the United Kingdom belonged to haplotype J6. Single isolates from passion fruit from Florida and strawberry flowers from Brazil belonged to mtDNA haplotypes J1 and J5, respectively.
The isolates in mtDNA group D represented five haplotypes that were 86–98% similar (Figs. 1, 2; Tables I and III). MtDNA haplotype D1 included a number of isolates from strawberry from the United States, Venezuela and Israel. A single isolate from pecan and two from apple from the United States had haplotypes D2 and D4, respectively. Haplotype D3 was shared by a single isolate from guava from Brazil and 11 isolates from pepper fruit from Taiwan. Haplotype D5 was shared by two isolates recovered from terminal crook-diseased pine seedlings from Australia and a biocontrol isolate known as Lubao, identified as C. gloeosporioides, used in China since 1966 for the control of dodder (Cuscuta sp.) in soybean fields (Templeton 1992, Watson et al 2000).
The F mtDNA group included 19 isolates with eight individual haplotypes that were 84–96% similar, based on band sharing. These included isolates from almond from Israel (F1) and California (F3), strawberry from Norway (F1) and Florida (F7), Rhododendron sp. from Sweden and Latvia (F1), and a variety of fruit-rot pathogens from New Zealand placed in morphological groups B and C by Lardner et al (1999) (haplotypes F1, F2, F4, F5, F6 and F8).
Isolates of G. miyabeana from willow, strawberry, nashi and apple had mtDNA RFLP haplotype K1, and a single isolate of C. acutatum from persimmon had a unique haplotype (E1). Haplotypes K1 and E1 were <75% similar to the other haplotypes of C. acutatum s. l. (Figs. 1, 2) on the basis of MspI mtDNA fragments.
Intron RFLPs and sequence – The introns of the GS and GPDH genes were amplified successfully from each of the 118 representative isolates of C. acutatum s. l. and the three isolates of C. gloeosporioides used for comparison (Table III). Including the flanking regions, the GS amplicon was approximately 1000 bp, whereas the actual intron size was 885–904 bp for isolates of C. acutatum s. l. and 908–912 bp for C. gloeosporioides. The approximate size of the GPDH intron and flanking regions was 280 bp, whereas the actual intron size was 209–220 bp for C. acutatum s. l. and 210–216 bp for the C. gloeosporioides outgroup.
Although single restriction enzymes digested the 900-bp GS intron, combinations of enzymes, particularly HindIII+HinfI+HaeIII and HindIII+HinfI+MspI (HHH and HHM), produced more highly polymorphic profiles with 7–9 bands and were most effective for resolving subgroups within C. acutatum (Fig. 3, Table III). Restriction fragments from the enzyme combination HHH were used to draw the cluster dendrogram (Fig. 4), for which bands smaller than 40 bp were not considered. Comparison of GS intron RFLP profiles indicated distinct groups among the isolates of C. acutatum s. l. examined, which generally were congruent with the mtDNA RFLP groups (Figs. 2 and 4). However, UPGMA analysis based on intron RFLPs clustered the pepper isolates (mtDNA haplotype D3) with the isolates in mtDNA RFLP group J. The enzyme combination HHM resolved two haplotypes among isolates in mtDNA RFLP group C but did not differentiate isolates with mtDNA haplotypes J2 and J3 (Table III).
|
|
|
|
|
|
|
Sequence alignment of the 200-bp GPDH intron produced a total of 285 characters, of which 81 were phylogenetically informative in the MP analysis. Topologies for 100 MP trees sampled were not significantly different using the Kishino-Hasegawa test and six of the seven clades had 100% bootstrap support (Fig. 6). Clade F had 60% support. The tree length was 151 steps, CI 0.8543, HI 0.1457, CI excluding uninformative characters 0.8268, HI excluding uninformative characters 0.1732, retention index RI 0.9726 and RC was 0.8309. The size of the 200-bp GPDH intron for each of the molecular groups were 211–215 bp (C), 209–210 bp (D and J), 209 bp (J4 and E), 211–214 (F) and 219–220 bp (K).
Sequence alignment of the combined dataset (900-bp GS and 200-bp GPDH introns) produced a total of 1395 characters, of which 342 were phylogenetically informative in the MP analysis. Of these 342 phylogenetically informative characters, 261 (76%) were derived from the GS intron. Topologies for 100 MP trees sampled were not significantly different using the Kishino-Hasegawa test, and nodes for each of the seven clades had 100% bootstrap consensus (Fig. 7). The tree length was 699 steps, CI 0.8169, HI 0.1831, CI excluding uninformative characters 0.7690, HI excluding uninformative characters 0.2310, RI 0.9558 and RC was 0.7808.
Phylogenetic analysis of the GS intron and of the combined GS and GPDH intron sequence dataset, resolved two subgroups (a and b) within clade C (Figs. 5 and 7), which corresponded to two HHM intron RFLP haplotype subgroups (Table III). MP sequence analysis of the 200-bp GPDH intron (Fig. 6) placed subgroup b as derived from within subgroup a, although NJ analysis did separate them as reciprocally monophyletic (data not shown). Subgroup a was geographically diverse, whereas b contained only isolates from the United States.
Mating studies – The fertility of crosses between the six tester strains is summarized in Table II. All of the 30 outcrossing combinations were fertile, producing perithecia on the center toothpicks of the mating plates in the area where the parental colonies converged. No selfings occurred, however, which would have been evident by the formation of perithecia on the outer toothpicks in the mating plates.
The fertility of crosses between the six testers and 118 isolates are summarized in Table III. Ascospores recovered from crosses with scores 5–7 were viable. The viability of the rare ascospores in crosses with a score of 4 was not assessed. Sexual fertility was identified among 42 of the 43 isolates in mtDNA subgroup C1. The exception was isolate 1337 from tomato, which had atypical colony morphology and did not sporulate. The single isolate with mtDNA haplotype C3 (ATCC MYA-664) also was fertile. The three isolates with haplotype C2, all from dodder in Massachusetts, were not sexually compatible with the mating testers.
All isolates in clade J4 mated with at least some of the five tester strains having mtDNA haplotype C1, although the level of fertility generally was lower than the crosses within the C group described above (Table III). One isolate with mtDNA haplotype J4 (PJ57, from guava) produced a few ascospores when mated with PJ8, the single tester with haplotype J4. This cross was weakly fertile. No other J4 x J4 crosses were fertile. None of the isolates in the J clade, which had mtDNA haplotypes other than J4, mated with any of the tester isolates.
None of the isolates in molecular clades D, E, and F and K mated with the six mating tester strains. Certain test crosses with isolates in clades D, E and F did produce perithecia with beaks and a few contained sterile asci (Table III). However, no ascospores were observed in these crosses. Two isolates with mtDNA haplotype F1 and the four isolates of G. miyabeana were self-fertile (homothallic). These isolates produced fertile perithecia on the adjacent outside toothpicks and on the center toothpick up to the point of convergence with the mating testers, but no perithecia were observed at the contact zone with any of the mating tester strains.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, Forster and Adaskaveg 1999, Freeman et al 2001, Johnston and Jones 1997, Lardner et al 1999, Sreenivasaprasad 1992, Vinnere et al 2002). Independent analyses of mtDNA and intron RFLPs and sequences of the 900-bp GS and 200-bp GPDH introns each recognized the same seven groups or clades (C, D, E, F, J, J4 and K). Phylogenetic analysis of noncoding DNA intron sequences from two nuclear loci produced highly congruent tree topologies with strong bootstrap support for each of the seven clades (Figs. 5 and 6). Low indices of homoplasy indicated that, although phylogenetically distinct, the clades shared a common ancestry. Sequence analysis provided quantitative measures of divergence within and between the clades identified. Molecular variability within the clades ranged from low (clades C and J4) to relatively high (clades D, J and F) (Tables IV and V). The concordance of two separate intron gene genealogies furthermore indicated that the clades identified were fixed and, along with distinct mtDNA RFLP haplotypes, that the clades might represent genetically isolated independently evolving populations.
Clades C and J4 represented a large collection of isolates from diverse hosts (Table I) that were morphologically typical (sensu Simmonds) and were capable of mating in laboratory crosses to produce the teleomorph G. acutata (Tables II and III, Guerber and Correll 2001a). Although sexual reproduction in natural populations of C. acutatum remains undocumented, the impetus to examine mating capabilities in C. acutatum came from the discovery of a minimum of 9 VCGs among isolates with mtDNA RFLP haplotype C1 recovered from a single small orchard (Correll et al 2000; Guerber and Correll 1997, 2001a). In the laboratory, sexual interfertility among group C isolates generally was quite high (Tables II and III), consistent with the hypothesis that mating compatibility may be multiallelic (Correll et al 2000). Crosses between most J4 isolates and the C testers also were fertile but produced fewer ascospores (Table III). Only about a third of the C isolates and none of the J4 isolates were highly fertile when crossed with the single J4 tester. Inheritance of genetic markers, including the GS and GPDH introns (unpublished data), nitrate (nit) and sulfate (sul) metabolic mutations, VCG and colony color (Guerber and Correll 1998), has been tracked in a subset of crosses. The production of recombinant phenotypes in F1 ascospore progeny from certain C x C and C x J4 crosses has been documented (Guerber and Correll 1998) and has confirmed that the two introns were unlinked and segregated independently (data not shown).
Clades C and J4 possibly define a widely distributed mating population, or biological species. Conversely, they might represent two phylogenetically isolated clades that have retained the ability to mate, as reported for other phylogenetically distinct fungal species (Taylor et al 2000). These data suggest that genetic isolation occurred before reproductive isolation in C. acutatum. The C1 and J4 molecular profiles were identified among isolates collected as early as 1964 by J. H. Simmonds (ultimately becoming isolates ATCC 56813 and ATCC 56816) and other isolates collected as recently as 1998–2000 (Table III). The strict association of independent divergent molecular markers (mtDNA RFLPs and GS and GPDH intron sequences) in the isolates examined thus far indicates that natural sexual recombination has not occurred recently between these two populations, although they have retained the potential to cross.
Isolates in molecular clade J (mtDNA RFLP haplotypes J1, J2, J3, J5 and J6) represented an interesting group of phylogenetically related pathogens that, based on intron sequences, diverged from isolates in mtDNA subgroup J4. Subgroup J1 included a single isolate from passion fruit from Florida. Subgroup J2 included isolates from lupine from the United States and Canada and isolates from Florida known to cause citrus postbloom fruit-drop disease (Agostini et al 1992) and anthracnose of leatherleaf fern (Strandberg et al 1997). The genetic similarity of the fern and citrus pathogens from Florida might be significant in the epidemiology of these two diseases. Causal agents of Key lime anthracnose were in subgroup J3, whereas the nine isolates identified with mtDNA haplotype J6 originated from Lupinus spp. from New Zealand and the United Kingdom. A single isolate with mtDNA haplotype J5 came from strawberry from Brazil. Postbloom fruit drop, Key lime anthracnose and lupine pathogens originally were identified as C. gloeosporioides on the basis of conidial morphology (Dick 1994, Fagan 1979, Gondran et al 1986). However, some authors have recognized them as C. acutatum on the basis of ITS 1 sequences (Brown et al 1996, Reed et al 1996, Sreenivasaprasad et al 1994). Nirenberg et al (2002) recently have presented compelling morphological, physiological and molecular data in support of a new species designation, C. lupini, for certain anthracnose pathogens of lupine. Their phylogenetic analysis of rDNA sequences of several species of Colletotrichum infecting lupine and other hosts placed C. lupini and C. acutatum in sister clades. In the present study, intron sequences produced two congruent phylogenetic gene genealogies with 100% bootstrap support for distinguishing lupine pathogens in subgroups J2 and J6 from certain morphologically typical isolates of C. acutatum in clades C, D, E and J4. One lupine pathogen from France [PJ62 = Lars 163 (Gondran et al 1986)], had mtDNA haplotype J2 in the present study and an rDNA sequence (Sherriff et al 1994) that placed it in C. lupini, according to Nirenberg et al (2002).
Clades D and J clustered together based on intron sequence analysis (Figs. 5, 6 and 7). A total of 20 isolates originally examined from strawberry from the United States, Israel, and Venezuela had mtDNA haplotype D1 (Table I). These isolates corresponded, by HaeIII mtDNA RFLP agreement (data not shown), to mtDNA group MG1 of Buddie et al (1999), who recognized this group as the major pathogens of strawberry in Europe and North America, although, as corroborated by the current study, several additional molecular subgroups of C. acutatum s. l. were isolated from strawberry.
Lardner et al (1999) identified a unique RAPD band pattern for an isolate, PJ5 (=10.200=PRJ 1008.3), recovered from persimmon from New Zealand, that conformed morphologically to their Group A (C. acutatum sensu Simmonds). Likewise, in the current study PJ5 had a unique mtDNA haplotype, E1 (Figs. 1 and 2), and divergent intron sequences (Figs. 5, 6 and 7) that placed it in a unique molecular clade.
Mitochondrial DNA RFLPs and intron sequence data identified considerable diversity among isolates in clade F (Figs. 5, 6 and 7), several of which have been reported to have atypical conidial and colony morphologies (Forster and Adaskaveg 1999, Freeman et al 2000, Lardner et al 1999). Clade F contained several isolates that Lardner et al (1999) placed in morphological groups B and C that were comprised of "C. acutatum-like" fruit-rot pathogens that clustered in C. acutatum s. l. based on D2 rDNA sequences (Johnston and Jones 1997). Four perithecial isolates in morphological Group B sensu Lardner et al (1999) had mtDNA haplotype F1. Nine nonperithecial isolates placed in morphological Group C by Lardner et al (1999) had similar mtDNA haplotypes, F2, F4, F5, F6 and F8 (Figs. 1 and 2) and intron sequences that grouped them together (Figs. 5, 6 and 7). Similarly, a subgroup of morphologically atypical isolates that formed gray colonies in culture was identified from almond in California and Israel (Forster and Adaskaveg 1999, Freeman et al 1998). This subgroup was differentiated further from isolates forming pink colonies and ellipsoidal conidia by PCR profiles using random and simple-repeat primers (Forster and Adaskaveg 1999, Freeman et al 2000) and sequences of ITS 2 rDNA, which were found to be more informative than ITS 1 sequences at the subspecies level (Freeman et al 2001a). Our mtDNA and intron sequence data placed representative gray almond isolates from California and Israel in molecular subgroups F3 and F1, respectively, and pink almond isolates from California in clade C. Unique isolates from strawberry from Norway and Florida had mtDNA haplotypes F1 and F7, respectively.
Isolates of G. miyabeana had a unique mtDNA haplotype (K1), and their intron sequences clustered with clade F (Figs. 5, 6 and 7). Although most isolates in clade F were self-sterile, the group included some homothallic isolates with mtDNA RFLP haplotype F1, (e.g., PJ9) and GS intron sequences that were particularly homologous to those of G. miyabeana (Figs. 5 and 7). G. miyabeana is a primary pathogen of willow (Salix spp.) and considered to be a secondary opportunistic pathogen of fruit crops (Johnston et al 2000). Isolates of G. miyabeana were consistently self-fertile in culture and had distinct colony phenotypes and conidia with at least one rounded end (Lardner et al 1999). Buddie et al (1999) reported that, unlike the predominant molecular group from strawberry discussed above (D1), a second group with more variable rDNA and mtDNA RFLPs at least was partially reproducing sexually and included isolates of G. miyabeana. The present study supported the recognition of G. miyabeana as a species distinct from C. acutatum, and furthermore, intron sequences suggested a close relationship between G. miyabeana and fruit-rot pathogens in clade F, particularly homothallic strains with mtDNA haplotype F1 in morphological group B sensu Lardner et al (1999).
This study identified phylogenetically diverse clades within C. acutatum s. l. that cause similar anthracnose diseases on certain hosts, such as bitter rot of apple, strawberry anthracnose, and terminal crook of pine. Apple bitter-rot pathogens were identified in clades C, D, F and K, and strawberry pathogens were identified from clades C, D, F, J and K. Phylogenetically diverse isolates in subgroups D5 and J4 and the widely distributed C1 all were recovered from Pinus spp. with terminal crook disease, and it remains untested whether other isolates in C1 or other groups can infect pine seedlings. There is, therefore, a need for more intensive sampling, controlled pathogenicity studies and an ongoing analysis of informative molecular data, including intron sequences, to further characterize the range of pathogen diversity on many economically important hosts.
Conversely, to identify the host range of certain phylogenetic clades, more extensive sampling of a variety of host species over a wider geographic range is warranted. Our data demonstrate that biologically relevant populations, such as clades C and J4, infect a wide range of hosts. However, the limited sample of isolates from Asia and Africa leaves in question the biogeographic distribution of these clades. Examination of a small sample of isolates from hosts such as Key lime, Lupinus spp. and strawberry, suggested that a degree of host specialization might have occurred in certain phylogenetic groups (Table III).
The data from this study reflected the highly variable nature of mtDNA and sequences from two different nuclear gene introns and demonstrated their value for characterizing phylogenetic relationships within C. acutatum s. l. The two introns also have been useful for examining interspecific phylogenetic relationships in Colletotrichum (Liu and Correll 2000). These data were consistent with other studies demonstrating that variable sequences of introns of nuclear genes can be phylogenetically informative (Geiser et al 1998, O'Donnell et al 1998, 2000). Analysis of sequence data from a broader range of isolates ultimately might recognize the clades identified in the current study as phylogenetically distinct species. The recognition of species defined by phylogenetic analyses, perhaps in concert with subtle morphological or cultural characters, would aid communication among plant pathologists and could improve our understanding of evolutionary dynamics in this diverse group of plant pathogens.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Accepted for publication February 26, 2003.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Adaskeveg JE, Hartin RJ., 1997 Characterization of Colletotrichum acutatum isolates causing anthracnose of almond and peach in California. Phytopathology 87:979-987
Agostini JP, Timmer LW, Mitchell DJ., 1992 Morphological and pathological characteristics of strains of Colletotrichum gloeosporioides from citrus. Phytopathology 82:1377-1382
Arx JA von., 1970 A revision of the Fungi classified as Gloeosporium. Bibl Mycol 24:1-203
Balardin RS, Smith JJ, Kelly JD., 1999 Ribosomal DNA polymorphism in Colletotrichum lindemuthianum. Mycol Res 103:841-48
Baxter AP, van der Westhuizen GCA, Eicker A., 1983 Morphology and taxonomy of South African isolates of Colletotrichum. S African J Bot 2:259-289
Bernstein B, Zehr EI, Dean RA, Shabi E., 1995 Characteristics of Colletotrichum from peach, apple, pecan, and other hosts. Pl Dis 79:478-482
Britton KO, Redlin SC., 1995 Damping-off of flowering dogwood seedlings caused by Colletotrichum acutatum and Fusarium oxysporum. Pl Dis 79:1188
Brown AE, Sreenivasaprasad S, Timmer LW., 1996 Molecular characterization of slow-growing orange and Key lime anthracnose strains of Colletotrichum from citrus as C. acutatum. Phytopathology 86:523-527
Brown AE, Supoena H., 1994 Pathogenicity of Colletotrichum acutatum and C. gloeosporioides on leaves of Hevea spp. Mycol Res 98:264-266
Buddie AG, Martinez-Culebras P, Bridge PD, Garcia MD, Querol A, Cannon PF, Monte E., 1999 Molecular characterization of Colletotrichum strains derived from strawberry. Mycol Res 103:386-394
Chelemi DO, Knox G, Palm ME., 1993 Limb dieback of flowering dogwood caused by Colletotrichum acutatum. Pl Dis 77:100
Correll JC, Gordon TR., 1999 Population structure of ascomycetes and deuteromycetes. In: Worrall JJ, ed. Structure and dynamics of fungal populations. Dordrecht, The Netherlands: Kluwer Academic Publishers. p 225–250
Correll JC, Guerber JC, Wasilwa LA, Sherill JF, Morelock TE., 2000 Inter- and intrapecies variation in Colletotrichum and mechanisms which affect population structure. In: Prusky D, Dickman MB, Freeman S, eds. Colletotrichum: host specificity, pathogenicity, and host-pathogen interactions. St. Paul, Minnesota: APS Press. p 145–79
Correll JC, Puhalla JE, Schneider RW., 1986 Identification of Fusarium oxysporum f. sp. apii on the basis of colony size, virulence, and vegetative compatibility. Phytopathology 76:396-00
Correll JC, Rhoads DD, Guerber JC., 1993 Examination of mitochondrial DNA restriction fragment length polymorphisms, DNA fingerprints, and randomly amplified polymorphic DNA of Colletotrichum orbiculare. Phytopathology 83:1199-1204
Correll JC, Rhoads DD, Guerber JC., 1994 Genetic and molecular diversity of populations of Glomerella cingulata, Colletotrichum gloeosporioides, and Colletotrichum acutatum from apple fruit (Abstr). Int Mycol Cong 5:77
Dick MA., 1994 Blight of Lupinus arboreus in New Zealand. NZeal J Forest Sci 24:51-8
Dingley JM, Gilmour JW., 1972 Colletotrichum acutatum: Simmds. f. sp. pinea associated with "terminal crook" disease of Pinus spp. New Zealand J For Sci 2:192-201
Dyko BA, Mordue JEM., 1979 CMI descriptions of pathogenic fungi and bacteria. No. 630, Colletotrichum acutatum. Kew, Surrey, England: Commonwealth Mycological Institute. 2 p
Fagan HJ., 1979 Postbloom fruit drop, a new disease of citrus associated with a form of Colletotrichum gloeosporioides. Ann Appl Biol 91:13-0
Felsenstein J., 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791
Forster H, Adaskaveg JE., 1999 Identification of subpopulations of Colletotrichum acutatum and epidemiology of almond anthracnose in California. Phytopathology 89:1056-1065
Freeman S, Katan T, Shabi E., 1998 Characterization of Colletotrichum species responsible for anthracnose diseases of various fruits. Pl Dis 82:596-05
Freeman S, Minz D, Jurkevitch E, Maymon M, Shabi E., 2000 Molecular analyses of Colletotrichum species from almond and other fruits. Phytopathology 90:608-14
Freeman S, Minz D, Maymon M, Zveibil A., 2001 Genetic diversity within Colletotrichum acutatum sensu Simmonds. Phytopahology 91:586-592
Geiser DM, Pitt JI, Taylor JW., 1998 Cryptic speciation and recombination in the aflatoxin-producing fungus Aspergillus flavus. Proc Natl Acad Sci USA 95:388-393
TABLE I. Continued
