| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Mycology and Plant Pathology |
2Department of Biology, Duke University, Durham, North Carolina 27708 USA; 3CSIRO Forestry and Forest Products, P.O. Box 5, Wembley, Western Australia 6913; 4USDA Forest Service, Pacific Northwest Research Station, 3200 Jefferson Way, Corvallis, Oregon 97331 USA; 5Institute of Microbiology, Leopold Franzens-University Innsbruck, Technikerstr. 25, 6020 Innsbruck, Austria; and 6Department of Forest Science, Oregon State University, Corvallis, Oregon 97331-7501 USA
Received for publication February 2, 2001. Accepted for publication June 26, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Cortinarius • Descomyces • Descolea • Hymenogaster • phylogeny • Protoglossum • Thaxterogaster • Quadrispora
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) (Figs. 1–3). Sequestrate ascomycetes have commonly been called true-truffles, but for sequestrate basidiomycetes many different terms describing a particular fruit body habit have been used: trufflelike, false truffles, gastroid (a peridium encloses a loculate or highly convoluted gleba lacking a stipitate columella), and secotioid (the margin of the pileus remains appressed to the stipe and the lamellae are convoluted and anastomosed). The term sequestrate was proposed by Kendrick (1992
) as a neutral term as regards phylogenetic relationship. Hence, sequestrate fungi encompass a polyphyletic group of diverse taxa often resembling particular fungi that forcibly eject their spores (Thiers, 1984
).
|
; Heim, 1971
). Such relationships were proposed also for Cortinarius (Pers.) S.F. Gray, perhaps the largest genus of agaricoid basidiomycetes. This genus of gilled fungi is recognized by the rusty brown basidiospores, which have roughened walls, the presence of a typically thin, cobweb-like partial veil (cortina), and a mutualistic (ectomycorrhizal) lifestyle. The tendency toward establishing a trufflelike habit from lamellate ancestors within the Cortinarius series was evidenced by gradually divergent morphologies. Three different sequestrate forms were thought to have been derived from Cortinarius: (1) strongly velate and partially hypogeous (emergent) Cortinarius taxa (Thiers and Smith, 1969
); (2) secotioid Thaxterogaster spp. (Singer and Smith, 1963; Singer, 1951), and (3) gastroid Hymenogaster spp. (Thiers, 1984
) (Figs. 1–3).
However, morphological studies alone have been inadequate to resolve the question of the origin and evolution of sequestrate taxa related to Cortinarius. Even in cases where affinities have been generally agreed upon, e.g., the relationship between Cortinarius and Thaxterogaster, the direction of evolution has long been debated (Smith, 1971
; Singer, 1975
; Thiers, 1984
). Meanwhile, molecular studies have provided compelling evidence that taxa formerly classified in the Gasteromycetes have evolved repeatedly from nonsequestrate ancestors within the Homobasidiomycetes (Hibbett et al., 1997
; Hibbett and Thorne, 2000
). Most molecular studies have focused on Russulales or Boletales and their sequestrate counterparts: within Russulales, Macowanites, Gymnomyces, Cystangium, Zelleromyces, and Arcangeliella have arisen several times from Russula and Lactarius (Miller et al., 2001)
; within Boletales, the secotioid genus Gastrosuillus is derived several times from Suillus and is thus polyphyletic (Kretzer and Bruns, 1997
), while the trufflelike Rhizopogon is monophyletic, sharing a common ancestor with Suillus (Johannesson and Martin, 1999
; Grubisha et al., 2001
). Surprisingly little molecular research has been focused on the linkages between the Agaricales (true gilled mushrooms) and their sequestrate relatives: only the relationship between the gilled mushroom Coprinus and its sequestrate relatives Podaxis and Montagnea (Hopple and Vilgalys, 1994, 1999
), the relationship of the sequestrate Weraroa with Strophariaceae (Binder, Besl, and Bresinsky, 1997
), and the relationship between the secotioid Setchelliogaster and the agaric Descolea have been demonstrated (Martin, Hogberg, and Llistosella, 1999
; Martin and Roccabruna, 1999
).
In this study we used ITS sequences to investigate the phylogeny of sequestrate taxa related to Cortinarius. Is there a link between Cortinarius, Thaxterogaster, and Hymenogaster? Do these sequestrate genera represent monophyletic groups? Is it appropriate to split Hymenogaster into morphologically more homogenous units (Hymenogaster sensu stricto, Descomyces, Protoglossum, Quadrispora, and Timgrovea) (Bougher and Castellano, 1993
)? If monophyletic entities exist among these sequestrate forms, are morphological characters such as spore and peridium morphology useful for their delimitation?
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
DNA extraction, amplification, and sequencing
DNA was isolated from dried herbarium material following standard protocols using miniprep procedures employing 2x CTAB lysis buffer (2% CTAB = cetyltrimetylammoniumbromide, 100 mmol/L Tris pH 8, 20 mmol/L Na2EDTA, 1.4 mol/L NaCl) (Zolan and Pukkila, 1986
). Samples of dried basidiome fragments were homogenized with a plastic pestle in a microcentrifuge tube (1.5 mL) with 300–500 µL 2x CTAB lysis buffer. After soaking for several hours, tubes were incubated in 70°C for 30 min. An equal volume of chloroform was added, and the tubes were vortexed and centrifuged at 10 000 g for 15 min. The aqueous phase was transferred to a new tube and purified either by ethanol/isopropanol precipitation or Geneclean no. III (Bio 101, La Jolla, California, USA) and resuspended in 55 µL ddH2O. The DNA extracts were diluted to 0.1–1.0 ng/µL for use in the polymerase chain reactions (PCR).
Polymerase chain reaction amplifications (PCR) followed the protocol of Vilgalys and Hester (1990)
in a final volume of 25 µL. Primers ITS1 and ITS4 (White et al., 1990
) were used for PCR amplification and sequencing of the internal transcribed spacers from the ribosomal genes. Reactions were performed in a Perkin Elmer 480 thermocycler (Perkin Elmer, Foster City, California, USA).
Amplified PCR products were quantified by gel electrophoresis on a 0.8% agarose gel stained with ethidium bromide and purified by microfiltration using Ultrafree-MC centrifugal columns (Millipore, Bedford, Massachusetts, USA). Sequencing was performed by fluorescent dye terminator chemistries following the manufacturer's instructions (Perkin Elmer) with automated sequencers (ABI 3700 or ABI 377, Perkin-Elmer, Norwalk, Connecticut, USA). Sequence chromatograms were compiled with Sequencher software, version 2.0 (Gene Codes, Ann Arbor, Michigan, USA). Sequences from this study have been submitted to GenBank.
Sequence alignments
Sequences were aligned by hand in the text editor of PAUP* version 4.0D64 (Swofford, 1998
). Four sequence alignments were generated. The first alignment was primarily used to define the major clades, which were then analyzed separately. It included 151 sequences representing the 114 taxa under investigation. After the introduction of gaps, the alignment included 859 nucleotide positions. Several areas of ambiguous alignment were excluded. From the analysis of the 507 unambiguously aligned characters, 207 characters were constant, 49 were parsimony uninformative, and 251 characters were parsimony informative. To correspond to the dominant species in the three major clades, these three clades will be referred to as the Cortinarius clade, the Hebeloma/Hymenogaster/Naucoria clade, and Descolea clade. To distinguish between clade names and names of taxa, the former are not italicized.
The second alignment was made for 81 ingroup taxa belonging to the largest clade, the Cortinarius clade, using three Hebeloma spp. as outgroup. The alignment of those 106 sequences was 859 nucleotide positions long; 555 characters remained after excluding gapped and ambiguous areas; 266 characters were constant, 58 were parsimony uninformative, and 231 included characters that were parsimony informative.
The third alignment included the Hebeloma/Hymenogaster/Naucoria clade, consisting of 29 sequences for 11 species of Hymenogaster, 7 species of Hebeloma, and 3 species of Naucoria. Four Cortinarius spp. and three Gymnopilus spp. were used as outgroup. The alignment was 859 nucleotide positions long: 556 characters were included, of which 318 characters were constant, 68 parsimony uninformative, and 170 characters parsimony informative.
The fourth alignment included 31 sequences belonging to the Descolea clade including eight taxa of Descolea, two taxa of Descomyces, two taxa of Setchelliogaster, four Cortinarius spp., and three Hebeloma spp. This alignment was 859 nucleotide positions long, of which 534 remained after excluding gaps/areas of ambiguity and of which 345 characters were constant, 73 were parsimony uninformative, and 116 characters were parsimony informative. Alignments are available from TreeBASE (accession number S636: M988–990).
Phylogenetic analyses
Phylogenetic analyses were performed in PAUP* version 4.0D64 (Swofford, 1998
). The first alignment was analyzed by maximum parsimony (MP) with the following settings: MULPARS on, steepest descent not in effect, MAXTREES set to 100, equal character weights, gaps treated as missing. The most parsimonious trees were searched with tree-bisection-reconnection (TBR) branch swapping. Starting trees were obtained by random-sequence addition; 100 heuristic searches were performed, and the shortest trees over all replicates were kept and assumed to be the most parsimonious reconstructions. Support for branches was evaluated using the bootstrap method (heuristic search; bootstrap replicates = 100; addition sequence = random; TBR on, MAXTREES = 100 in each replicate). Three major clades were well supported and were used to generate the alignments two, three, and four. Each of those latter alignments was analyzed separately, which allowed the inclusion of more unambiguous positions.
Maximum parsimony (MP) was employed for analysis of alignment two with the PAUP* settings as above, but MAXTREES was set to 25 000. Topological constraints were employed to investigate alternative tree topologies (monophyly of Protoglossum and Thaxterogaster). To test whether constrained topologies were significantly worse, these were compared to the unconstrained trees using the nonparametric comparison test of Templeton (Templeton, 1983
) and the Kishino-Hasegawa test (Kishino and Hasegawa, 1989
) as implemented by PAUP*. Maximum-likelihood ratio tests (LRT) (Goldman, 1993
; Huelsenbeck and Rannala, 1997
) were employed in order to identify a simple and robust substitution model for each data set. The LRT were conducted from trees obtained in the MP analyses. After model estimation, one MP tree was used as a starting tree for TBR searches by use of maximum likelihood (ML) with parameters derived from the model fitting. The ML search was allowed to proceed for 85 h. Bootstrap support for branches obtained in ML searches were estimated with 2000 replicates of the "fast bootstrap" option in PAUP* using parsimony (MULPARS off and no branch swapping).
For alignment three, the Hebeloma/Hymenogaster/Naucoria clade, MP was employed as described above. Two hundred heuristic searches were performed, and the shortest trees were kept and assumed to be the most parsimonious reconstructions. These 35 trees were used as starting trees for an additional heuristic search with TBR and a reconnection limit of 27, MAXTREES = 5000. Support for each branch was estimated by 200 bootstrap replicates (MULPARS off and no branch swapping) (Felsenstein, 1985
). Topological constraints were employed to test the monophyly of Hymenogaster, Hebeloma, and Naucoria. After model estimation for ML analysis, one MP tree was used as starting tree for TBR searches using parameters derived from the LRT. Bootstrap supports for branches obtained in ML searches were estimated with 100 random bootstrap replicates under parsimony.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
) with rate heterogeneity, transition/transversion ratio = 1.9548. The model accounted for among-site rate variation with a gamma-shape parameter estimation of 1.302271, and four rate categories represented by mean. Assumed nucleotide frequencies were A = 0.24660, C = 0.19754, G = 0.22165, T = 0.33421. Number of substitution types = 2. The overall topology of the strict consensus tree of the MP analysis (not shown) corresponds with the ML tree (–ln(likelihood) = 6306.30371) (Fig. 5).
|
Bootstrap support for larger monophyletic groups (>3 taxa) within the Cortinarius clade was obtained for Dermocybe and for parts of the subgenera Telamonia, Phlegmacium, Myxacium. Three lineages of Telamonia spp. were found: Telamonia I, Telamonia II, and Myxotelamonia. Phlegmacium spp. group into four subclades, and Myxacium spp. into three. Several Sericeocybe taxa of section Anomali also form a clade, with remaining species grouped into Telamonia II. Species currently classified in the subgenus Leprocybe are found in different clades.
Placement of sequestrate taxa related to Cortinarius
Twenty-two sequestrate taxa were placed with bootstrap support >50% into nine clades together with agaric-like forms (Fig. 5). However, two Hymenogaster spp., Thaxterogaster violaceum, T. porphyreum, Protoglossum sp. Beever 996, and two sequestrate Cortinarius spp. (Tr18741, Tr8178) could not be clearly related to any of the >400 sequences in the current ITS database of Cortinarius and related genera (U. Peintner, unpublished data).
Searches with topological constraints for the monophyly of Thaxterogaster all yielded longer (i.e., less parsimonious) trees than searches without constraints. Constrained trees were all clearly rejected by the Templeton and the Kishino-Hasegawa tests (Table 1) indicating that Thaxterogaster is not monophyletic. The species currently classified in the genus Thaxterogaster have multiple origins: (1) the relationships of clades Thaxterogaster I and II are not known; and (2) the other Thaxterogaster spp. are related to agaricoid taxa of the three clades Phlegmacium I, Myxacium I, and Myxotelamonia (Fig. 5).
|
Sequestrate Cortinarius spp. are found within the Phlegmacium III, IV, and Dermocybe clades. Cortinarius Tr8178 is nested within Phlegmacium III and is a sistergroup to C. variecolor and C. citriolens. Three partly hypogeous cortinarii, C. magnivelatus, C. bigelowii, and C. verrucisporus, are nested within Phlegmacium IV. C. globuliformis is closely related to Dermocybe splendida.
Two taxa of Quadrispora tested form a clade closely related to T. redactus and T. piriformis and thus are a purely sequestrate subclade of Myxacium I. Furthermore, Hymenogaster remyi and H. sublilacinus form their own small Hymenogaster clade within the Cortinarius clade.
Analysis of individual data sets: the Hebeloma/Hymenogaster/Naucoria clade
The heuristic searches yielded 18 best trees with a score of 491, CI = 0.6680, HI = 0.3320, RI = 0.7504. The best fitting ML model for this alignment was as follows: assumed nucleotide frequencies: A = 0.25515, C = 0.20651, G = 0.20895, T = 0.32939; number of substitution types = 6; proportion of invariable sites: 0.49987; rates follow a gamma distribution with shape parameter 2.94413, four rate categories. This corresponds to the general time-reversible model (Yang, 1994
) with rate heterogeneity. The tree resulting from the ML analysis (–ln (likelihood) = 3262.13834) (Fig. 6) was completely congruent with the MP trees (not shown). The Hebeloma/Hymenogaster/Naucoria clade is a monophyletic group consisting of five well-supported clades: (1) the Hebeloma clade; (2) two clades of Naucoria representing the two currently accepted subgenera of the genus, subgenus Submelinoideae and subgenus Naucoria; and (3) two clades of Hymenogaster, with the clade Hymenogaster I including the type species H. bulliardii and clade Hymenogaster II. The deeper branches of each clade were not resolved. All trees enforcing the monophyly of Hebeloma and/or Hymenogaster had the same length than all unconstrained trees, indicating that monophyly for Hymenogaster cannot be rejected from our data. All trees with constraints for monophyly of Naucoria were five steps longer than unconstrained trees; nevertheless, a monophyletic Naucoria could not be rejected by the Kishino-Hasegawa test or by the Templeton test (Table 2).
|
|
) with rate heterogeneity. The tree resulting from the ML analysis (–ln (likelihood) = 3191.96939) (Fig. 7) was congruent with the MP trees (not shown). The Kishino-Hasegawa and the Templeton tests clearly reject monophyly of Setchelliogaster, while the monophyly of Descomyces is congruent with all unconstrained trees.
|
. However, for practical reasons, in this paper we will continue to use this name for the agaricoid form of S. tenuipes. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Moser, 1983
; Hansen and Knudsen, 1992
), eight subgenera of Cortinarius are recognized: Cortinarius, Cystogenes, Dermocybe, Leprocybe, Myxacium, Paramyxacium, Phlegmacium, Sericeocybe, and Telamonia. This study and others (Liu, Rogers, and Ammirati, 1997
; Hoiland and Holst Jensen, 2000
; Seidl, 2000
; Seidl et al., 2000
) indicate that these subgenera do not correspond to monophyletic groups. Therefore, a reclassification of Cortinarius into natural units would be appropriate, but is beyond the scope of this paper.
Unsettled family affinities: the genera around Cortinarius
The question of "family affinities" of Cortinarius and other brown-spored genera of Agaricales is still unsettled (Hibbett and Thorne, 2000
; Moncalvo et al., 2000)
. Based on alignability of ITS sequences, a close relationship between Cortinarius, Hebeloma, Naucoria, and Descolea can be assumed. Based on the same criterion, the genus Gymnopilus Karst. was chosen as the outgroup. This genus is suitable as an outgroup as the sampled species form a monophyletic clade, are saprobic in contrast to the ectomycorrhizal ingroup taxa, and no sequestrate relatives to Gymnopilus have been reported at present.
Internal transcribed spacer sequences of Inocybe, Callistosporium, Laccaria, and Leucopaxillus were also examined to serve as outgroup to root the Cortinarius phylogeny, as suggested by classical taxonomy (Singer, 1975, 1986), chemotaxonomy, or previous molecular studies (Moncalvo et al., 2000)
. (These sequences have also been deposited in GenBank.) However, none of these sequences were alignable with our ingroup data set; therefore, a close relationship of the above genera with Cortinarius, Hebeloma, Naucoria, Descolea, and Gymnopilus can be excluded. These preliminary results are particularly interesting when higher taxonomic categories, e.g., the delimitation of the family Cortinariaceae, are addressed, but more extensive molecular studies are needed to resolve these questions.
The puzzling phylogeny of the Cortinarius clade
The Cortinarius clade is a large monophyletic group, which includes agaricoid taxa of the genera Cortinarius, Dermocybe, and Rozites. No premature conclusions should be drawn about the phylogeny of Cuphocybe based on one single taxon, as neither the type species (Cuphocybe olivacea) nor other species of Cuphocybe have been included in this study.
The deeper relationships within Cortinarius are not well resolved by ITS sequences (Fig. 5). However, our results indicate that except for the subgenera Dermocybe and Cortinarius, the currently recognized subgenera are polyphyletic. This result agrees with previous studies (Liu, Rogers, and Ammirati, 1997
; Hoiland and Holst Jensen, 2000
; Seidl, 2000
).
The monophyletic Dermocybe clade nests within the Cortinarius clade, which consists of a well-supported core clade including most of the known boreal species of Dermocybe (Liu, Rogers, and Ammirati, 1997
) and 2–3 smaller clades. The monophyletic subgenus Cortinarius consists of only few species, which are characterized by large violaceous basidiomes with cheilo- and pleurocystidia.
Many taxa currently classified in the subgenus Telamonia also form a monophyletic group. Although we cannot confirm a consistent subdivision of Cortinarius into the two lineages Telamonia II and Cortinarius sensu stricto (s.s.) as proposed by Hoiland and Holst Jensen (2000)
, the apparent lack of sequestrate forms may support the isolated position of Telamonia II within the Cortinarius clade. However, a few species currently assigned to this subgenus are outside the core clade and show relationships to sequestrate taxa: C. acutus and C. obtusus (Telamonia I), as well as the type species of section Myxotelamonia, C. cinereobrunneus.
Subgenus Myxacium is polyphyletic, comprising at least three distinct lineages: Myxacium I, including the monophyletic sections Defibulati and Myxacium; Myxacium II (Cortinarius vibratilis = section Ochroleuci); and Myxacium III (C. salor and C. delibutus = section Delibuti). Interestingly, sequestrate taxa seem to be related only to Myxacium I.
The morphology-based classification of sequestrate fungi related to Cortinarius
Many sequestrate genera have been placed in the Cortinariaceae based on shared characters such as brown, verrucose spores, a nongelatinized trama, an array of specific pigments, and the ectomycorrhizal habit. The sequestrate forms with the highest morphological similarity with agarics are the emergent species of Cortinarius (Thiers and Smith, 1969
; Watling, 1980
; Bougher and Malajczuk, 1986
; Fogel, 1994
). The gills of these fungi are radially arranged and produce a spore deposit. Unlike typical agaricoid species of Cortinarius, however, the stipe is short and the gills often disintegrate with age while the partial veil commonly remains unbroken at maturity.
The next step of morphological reduction occurs within the genus Thaxterogaster (Singer, 1951
); taxa classified in this genus are characterized by basidiomes with a reduced to well-developed stipe and a basally exposed hymenium with a loculate to lamellate gleba (Fig. 2).
Hymenogaster Vittadini was widely considered as "end-point phylogenetic position in the Cortinariaceae presenting a truly hypogeous member" (Fig. 3) (Smith, 1966
; Beaton, Pegler, and Young, 1985
; Fogel, 1985
). However, as Bougher and Castellano (1993)
discussed at length, the generic concept of Hymenogaster sensu lato (s.l.) has been inflated beyond the original circumscription intended by Vittadini. Therefore, the authors recombined taxa from Hymenogaster s.l. into several genera, e.g., Destuntzia (Fogel, 1985
), Descomyces, Quadrispora, Gautieria, Timgrovea, and Protoglossum (formerly Cortinomyces; May, 1995
).
As redefined by Bougher and Castellano (1993)
, Hymenogaster s.s. is restricted to taxa with thick-walled, ellipsoid to fusiform spores bearing a large, cup-like hilar appendix, as in the type species H. bulliardii. Because such spores are very different from those of the genus Cortinarius, the relationship of Hymenogaster s.s. to other established taxa of the Agaricales remained unclear (Bougher and Castellano, 1993
).
Hymenogaster spp. with warty spores and closely adhering episporium, which closely resemble spores of Cortinarius spp., were recombined into the genus Protoglossum (type: P. luteum) (Bougher and Castellano, 1993
). As distinguished from Thaxterogaster spp., Protoglossum spp. have gastroid basidiomes without stipes or with very reduced stipes (columella). However, in species with highly variable development of columella (e.g, T. luteirufescens), boundaries between Thaxterogaster and Protoglossum are indistinct (Bougher, 1997
; Bougher and Syme, 1998
).
Quadrispora spp. are characterized by hypogeous basidiomes with a viscid peridium and irregularly warted statismospores, which adhere to each other in persistent tetrads after release from the basidium. The type of the genus is Q. oblongispora. The relationship of Quadrispora to other taxa of the Agaricales was not previously known (Bougher and Castellano, 1993
).
Descomyces was segregated from Hymenogaster because of the distinct spore morphology and peridium structure resembling Setchelliogaster and Descolea. In contrast to the secotioid Setchelliogaster, the basidiomes of Descomyces are gastroid. The type of the genus is Descomyces albus (Bougher and Castellano, 1993
).
Setchelliogaster accommodates taxa with a very reduced to absent stipe and basally exposed hymenophore. The key characters for this genus are a peridial surface consisting of inflated, isodiametric elements, and pale brown basidiospores with finely verruculose ornamentation overlaid by a hyaline myxosporium (type: S. tenuipes). Significant evidence for the close relationship between Setchelliogaster and Descolea was based on the observation that both genera can produce a variety of basidome morphologies, which transgress the formal taxonomic definitions of the genus (Martin and Roccabruna, 1999
; Lago, Bougher, and Castro, 2000
).
Multiple origins of sequestrate fungi from ectomycorrhizal ancestors
Sequestrate forms occur in all three major ectomycorrhizal lineages of Cortinariaceae (the Cortinarius, Hebeloma/Hymenogaster/Naucoria, and Descolea clades) but not within the saprobic Gymnopilus (Figs. 4–7). These results indicate multiple origins of sequestrate forms from ectomycorrhizal ancestors. Further evidence for this hypothesis arises from the fact that animals play an important role in dispersing spores of the generally obligately ectomycorrhizal, sequestrate fungi (Claridge, Castellano, and Trappe, 1996
). The production of spores at the root zone and the maintenance of spores in the soil for long periods are generally regarded as evolutionary advantages for mycorrhizal fungi (Fogel, 1992
; Miller, Torres, and McClean, 1994
). Long-distance dispersal is effected by mycovores, and spores in the faeces are likely to be deposited near suitable plant roots (Bougher and Lebel, 2001
).
Molecular data support the segregation of Hymenogaster s.l. in distinct, independently derived lineages as proposed by Bougher and Castellano (1993)
. Furthermore, ITS phylogeny also places these lineages in relationship to their agaricoid counterparts. Hence, only sequestrate Cortinarius spp., Thaxterogaster, Quadrispora, Protoglossum, and two Hymenogaster spp. (H. remyi, H. sublilacinus) share common ancestors with Cortinarius taxa. The other Hymenogaster taxa under investigation share a common ancestor with Hebeloma and Naucoria. Setchelliogaster and Descomyces are derived from Descolea, the former being polyphyletic and the latter monophyletic.
Sequestrate fungi within the Cortinarius clade: polyphyletic origin vs. radiation into monophyletic entities
Three different sequestrate basidiome morphologies (emergent, secotioid, and gastroid) occur within the Cortinarius clade (Figs. 1–3 and 5): emergent Cortinarius spp. are strikingly morphologically similar to their agaric-like relatives. The sequestrate Cortinarius spp. related to Phlegmacium IV have a strongly developed white veil like their agaric relatives, but in the most derived taxon C. magnivelatus, the pileus is dry. In addition C. globuliformis is related to Dermocybe splendida and has bright yellow colors on the cap, gills, stipe, and the surrounding matlike masses of mycelium (Bougher and Malajczuk, 1986
), a typical character for many species of Dermocybe. A unique combination of anthraquinone derivates has been found in D. splendida (Horak, 1987
; Gill, 1995
). Anthraquinonic pigments have been proven to be a very useful character in Dermocybe taxonomy, but they have not yet been investigated in sequestrate Cortinarius spp., though this may be a simple tool to test relationships in the future. As further indicated by other relationships found in this study, it might also be useful to apply other macrochemical tests currently used for Cortinarii (e.g., potassium hydroxide [KOH] 30% reaction) on sequestrate forms as well. Similarly, testing the fluorescence behavior of basidiomes under UV light might be worthwhile (Moser, 1960, 1983
).
Our results clearly demonstrate that emergent Cortinarius taxa evolved from agaric-like precursors, but also indicate that once successful sequestrate forms are established, they tend to radiate independently from their agaric relatives (e.g., the clade C. magnivelatus, C. bigelowii, and C. verrucisporus). This confirms the hypothesis proposed by Fogel (1994)
of a recent parallel evolution of emergent Cortinarius taxa based on distribution data.
The sequestrate Thaxterogaster spp. have multiple origins within the Cortinarius clade: T. albocanus is closely related to Cortinarius cinereobrunneus, a species belonging to section Myxotelamonia resembling typical Telamonia spp., but having a gelatinized epicutis or veil. Thaxterogaster campbellae and T. fragile are derived from taxa of clade Phlegmacium I, whereas T. violaceum has unknown relationships. However, our results indicate that most Thaxterogaster taxa could be related to clade Myxacium I: T. pingue and Thaxterogaster sp. Tr17864 form a well-supported clade as sister group of Myxacium I, sect. Myxacium (C. trivialis, etc.) (Fig. 5). Key characters of this section are presence of clamps and spores >10 µm. Thaxterogaster redactus and T. piriformis share a common ancestor with all these above-mentioned taxa.
Thaxterogaster paveleckii, with thick, slimy peridium and no clamp connections, is closely related to Myxacium I, section Defibulati (C. mucifluus, etc.). This secotioid species exhibits the main synapomorphy supporting this section Defibulati, namely the complete lack of clamp connections. The lack of clamp connections in T. porphyreum can be interpreted as a further indication for its affinities to Myxacium I, as clamp connections usually occur in Cortinarius. Hence, Myxacium I seems to be a "hot spot" of evolution of Thaxterogaster-like forms, including also the development of the gastroid Quadrispora taxa (Fig. 5).
Two other gastroid genera are related to Cortinarius, namely Protoglossum and two Hymenogaster spp., which are primarily delimited from each other based on spore morphology. Hymenogaster sublilacinus and H. remyi have no natural relationships to Hymenogaster s.s. (Figs. 5–6). They belong to a distinct group of species with small, nonbeaked spores. Such species have been placed in the subgenus Dendrogaster (formerly genus Dendrogaster Bucholtz). Unfortunately, the type species of Dendrogaster, D. connectens Bucholtz, has beaked spores and thus does not belong to this species complex, but is a synonym of Hymenogaster (Fogel, 1985
). The necessary nomenclatural changes to accommodate these species will be published in a forthcoming paper.
Overall, each of the different sequestrate basidiome forms related to Cortinarius appears in various subclades, clearly indicating multiple origins. However, certain sequestrate forms seem to be more frequent in certain groups, e.g., Thaxterogaster in Myxacium I and emergent Cortinarius spp. in clades of Phlegmacium. This observation might indicate an increased susceptibility of certain groups to develop a particular sequestrate form. Furthermore, this might also be interpreted as an indication for relationships between clades sharing similar sequestrate forms, e.g., between clades Phlegmacium II, III, and IV.
Although several types of sequestrate forms occur within the Cortinarius clade, these forms are phylogenetically not closely related with each other: only Thaxterogaster and Quadrispora can be found in the same clade, indicating that gastroid taxa could possibly derive from secotioid forms. However, Quadrispora could also share a common ancestor with Thaxterogaster. This implies that the different sequestrate basidiome forms can evolve independently from each other without intermediate forms.
When several sequestrate taxa occur in a clade, they tend to form a sister group to the agaric-like forms rather than being nested within them. Thus, although sharing a common ancestor, they evolve and speciate separately. This pattern of evolution is especially striking in Myxacium I, in which a sampling bias for agaric-like forms can be excluded. We conclude that sequestrate forms are of polyphyletic origin within the Cortinarius clade, but some of them are developing into monophyletic, purely sequestrate groups.
Overall, ITS phylogeny indicates a derived, recent radiation of sequestrate forms in Cortinarius. The low level of sequence divergence within the Cortinarius clade and Hebeloma (Aanen et al., 2000)
suggests that these fungi have a relatively recent origin, with one to several periods of rapid speciation.
Is Hymenogaster polyphyletic?
Hebeloma, Naucoria, and Hymenogaster form a well-supported monophyletic clade (Fig. 6). Basal nodes within the Hebeloma/Hymenogaster/Naucoria clade are not well resolved. In any case, Hymenogaster has a common ancestor with both Naucoria and Hebeloma.
Hebeloma is monophyletic. Small Hebeloma spp. could be confused with Naucoria spp. but the latter never have a gelatinized epicutis and are exclusively ectomycorrhizal with Alnus (or more rarely, Salix). Because of the somewhat difficult delimitation, Kühner (1980)
has proposed fusing Hebeloma and Naucoria. Our results confirm earlier data (Aanen et al., 2000)
, indicating that Naucoria is phylogenetically distinct from Hebeloma, which implies that the strict host preference of Naucoria has acted as a strong selection pressure. The two clades of Naucoria correspond to the two currently recognized sections: section Naucoria (N. escharoides, N. scolecina) with narrow cheilocystidia tapering to a subacute or acute apex, and section Submelinoideae (N. bohemica) with broad, nonattenuate cheilocystidia. The monophyly of Naucoria could not be rejected.
Also the monophyly of Hymenogaster could not be rejected. Hymenogaster I includes the type species H. bulliardii and others (Fig. 6), while Hymenogaster II consists of H. alnicola, H. glacialis, and H. tener. However, more extensive studies are needed to assess whether Hymenogaster is monophyletic, having achieved its special morphology only once.
The morphological plasticity of sequestrate taxa related to Descolea
Internal transcribed spacer data clearly show that Descolea, Setchelliogaster, and Descomyces form a distinct clade (Fig. 7) in which three types of basidiomes are present: Descolea, with agaric-like basidiomes, Setchelliogaster, with secotioid, sublamellate basidiomes, and Descomyces, with gastroid basidiomes and a loculate hymenium.
A broad variation in basidiome morphology has been observed for many species of Descolea and Setchelliogaster (Martin and Roccabruna, 1999
; Lago, Bougher, and Castro, 2000
). For instance, all transitional stages between agaricoid and secotioid fruit bodies have been observed for D. rheophylla, which therefore was synonymized with S. tenuipes (Lago, Bougher, and Castro, 2000)
. Based on morphological and molecular data, Martin and Roccabruna (1999)
argue that S. tenuipes is recently derived from D. rheophylla. However, our molecular data indicate that the agaricoid morphology was rather reestablished from a secotioid form, in which the basidiome development of present-day individuals remains very plastic.
Setchelliogaster australiensis exhibits less morphological plasticity (variation) than S. tenuipes (Lago, Bougher, and Castro, 2000)
. Basidiomes of this taxon are morphologically highly variable, but no agaricoid forms, as in S. tenuipes, have been observed. Furthermore, secotioid forms have also been reported to occur in Descolea maculata and D. gunnii (Lago, Bougher, and Castro, 2000)
. Hydnangium sublamellatum is another ectomycorrhizal fungus with Eucalyptus, in which different developmental patterns from agaricoid to gastroid can be expressed amongst and within single collections. Such highly morphologically plastic taxa are probably very recently derived from agaricoid precursors (Bougher, Tommerup, and Malajczuk, 1993
).
High levels of morphological plasticity pose problems for circumscribing taxa on a morphological basis. Taxa defined by basidiomal form are useful in a broad practical sense, but may not be biologically distinct, especially when their boundaries are transgressed by a high morphological variability. In such cases, the application of a phylogenetic species concept as defined by Harrington and Rizzo (1999)
seems most appropriate: they diagnose species as the smallest aggregation of populations with a common lineage that share unique, diagnosable phenotypic characters. Thus, collections belonging to the clade around Setchelliogaster tenuipes should be regarded as one morphologically very plastic species.
The lack of morphological plasticity of Descomyces is surprising, as Descomyces was regarded as the most reduced form in the phylogenetic series including Setchelliogaster and Descolea. According to our molecular data, one plausible scenario for the development of sequestrate forms related to Descolea is a two-step mutation: the first evolutionary event occurred at the basal node of the Descolea clade, which would explain the morphological plasticity of most of the investigated Descolea spp. Second, in Descomyces, an apparently stable morphology has evolved, forming a small, but distinct gastroid lineage. The stability of such gastroid morphologies implies that here the information for certain morphological pathways is not readily reexpressed or was definitely lost.
The extreme morphological changes between Descolea and Descomyces are analogous to the case of Rhizopogon and Suillus, in which striking macromorphological changes combined with low levels of genetic sequence divergence and extreme similarity in micromorphology have also been observed (Bruns et al., 1989
). Hence, this is another case in which morphological change is not related to molecular change in the ITS region of rDNA.
Evolution of sequestrate forms
The gastroid lineages without intermediate secotioid forms, as well as the high diversity of apparently stable (i.e., persistent over time) secotioid forms, brings into question the severity of selection against intermediate forms hypothesized by previous molecular studies (Bruns et al., 1989
; Baura, Szaro, and Bruns, 1992
). Our results suggest instead several possible evolutionary scenarios for the development of sequestrate forms: (1) evolution of morphologically very plastic forms (Setchelliogaster) from gilled precursors; (2) evolution of stable emergent taxa (e.g., Cortinarius magnivelatus group) from gilled precursors; (3) evolution of stable secotioid taxa from gilled precursors (Thaxterogaster); (4) fast evolution of stable gastroid taxa from gilled precursors, with or without unstable intermediate forms (Hymenogaster, Descomyces); and (5) gradual evolution of gastroid forms from stable secotioid ancestors (Quadrispora?).
For morphologically very plastic taxa, it has been suggested that the different basidiome forms are due to differences in the expression of, or relative timing and duration of, genetic events controlling various key processes during development (Bougher and Lebel, 2001
). Descolea rheophylla is a well-documented example for agaric-like forms, which can derive secondarily from morphologically very plastic sequestrate forms. Here the information for ballistospore development has not been lost, but has been suppressed.
This could indicate that the loss of important morphogenetic information probably occurs in a later stage in the evolution of sequestrate forms, although this is a key step in the development of stable, sequestrate species. Mutations conferring the loss of functional development may occur in only a few genes, therefore morphological reduction appears to be rapid relative to molecular changes, e.g., recessive alleles at a single locus can cause the formation of sequestrate forms (Hibbett, Tsuneda, and Murakami, 1994
). Such mutations and others may easily occur independently from other mutations, thus explaining why sequestrate forms within Cortinarius have repeatedly evolved from agaric-like precursors. Transformations resulting from a loss of genetic information usually cannot be reversed. Hence, we hypothesize that successfully adapted, stable sequestrate forms tend to radiate, but they can radiate only into species with the same form of sequestrate basidiomes or into further reduced basidiome types.
The utility of morphological characters to establish natural relationships
The hypothesis that gasteroid adaptions have arisen at successive levels of basidiomycete evolution was already proposed by Ingold (1965)
based on spore morphology and spore dispersal. Molecular phylogenies support this hypothesis: at higher taxonomic levels within basidiomycetes, Hibbett et al. (1997)
showed that puffballs and other forms of Gasteromycetes have been repeatedly derived across the Hymenomycetes. On a lower taxonomic level, our study clearly demonstrates that sequestrate taxa are also related to distinct clades of agaric-like Cortinarius, Hebeloma/Hymenogaster/Naucoria, and Descolea. This demonstrates that changes in morphology from agaricoid to gastroid with all transitional forms have been achieved independently much more frequently than previously suspected across all taxonomic levels.
Interestingly, as early as 1892 Rehsteiner already suggested a possible relationship of Hymenogaster with agarics based on their fruit body development (Rehsteiner, 1892, cited in Reijnders, 2000)
. Later, Hymenogaster was redefined and a close relationship of Protoglossum and Thaxterogaster to Cortinarius as well as of Descomyces and Setchelliogaster to Descolea were proposed based on the spore morphology and other morphological characters (Bougher and Castellano, 1993
). Our molecular data confirm these results, thus supporting the utility of spore morphology for delimiting groups. However, morphological methods have their limits: sometimes the morphology of "Gasteromycetes" is reminiscent of one of the groups from which they have descended, and in other cases, they have been subject to a unique evolution in response to challenge by particular ecological requirements (Reijnders, 2000)
. When the latter occurs, evolutionary relationships can probably only be elucidated by supplementing the available morphological and ecological data with molecular data.
Conclusions
Within the Cortinariaceae, ITS data appear well suited for elucidating relationships around the species level, as sequences of the same taxon from different geographical origins match (Fig. 5). Furthermore, the phylogenetic relationships of sequestrate taxa like Thaxterogaster, Protoglossum, Quadrispora, Setchelliogaster, Descomyces, and Hymenogaster can be determined in relation to their gilled relatives. These results clearly demonstrate that sequestrate taxa evolved independently many times. In most of the cases we studied, emergent, secotioid, and gastroid forms evolved independently from each other, resulting in what appear to be morphologically stable taxa with a tendency to radiate.
However, considerable work is needed not only to determine the phylogenetic relationships of sequestrate fungi, but also to illuminate the molecular mechanisms of the evolution of such striking morphological modifications as represented by sequestrate fungi. This study and others (Hibbett et al., 1997
; Miller et al., 2001
) further stress the necessity to reunify the form-group "Gasteromycetes" with the "Hymenomycetes." Neither group can be regarded as isolated from the other, but acknowledgment of natural, morphologically diverse lineages should be the new paradigm for fungal systematics.
|
FOOTNOTES |
|---|
7 Author for reprint requests, current address: Institute of Microbiology, Leopold Franzens-University Innsbruck, Technikerstr. 25, 6020 Innsbruck, Austria (ursula.peintner{at}uibk.ac.at
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Baura G. T. M. Szaro T. D. Bruns 1992 Gastrosuillus laricinus is a recent derivative of Suillus grevillei: molecular evidence. Mycologia 84: 592-597[CrossRef][ISI]
Beaton G. D. N. Pegler T. W. K. Young 1985 Gasteroid Basidiomycota of Victoria State, Australia 3. Cortinariales. Kew Bulletin 40: 167-204
Binder M. H. Besl A. Bresinsky 1997 Agaricales or Boletales? Molecular evidence towards the classification of some controversial taxa. Zeitschrift für Mykologie 63: 189-196
Bougher N. L. 1997 Three new sequestrate basidiomycetes from Western Australia. Mycotaxon 63: 37-48
Bougher N. L. M. A. Castellano 1993 Delimitation of Hymenogaster sensu stricto and four new segregate genera. Mycologia 85: 1034-1034[ISI]
Bougher N. L. A. Lebel 2001 Sequestrate (truffle-like) fungi in Australia and New-Zealand. Australian Systematic Botany 14: XX-XX
Bougher N. L. N. Malajczuk 1986 An undescribed species of hypogeous Cortinarius associated with Eucalyptus in Western Australia. Transactions of the British Mycological Society 86: 301-304[ISI]
Bougher N. L. K. Syme 1998 Fungi of Southern Australia. University of Western Australia Press, Melbourne, Australia
Bougher N. L. I. C. Tommerup N. Malajczuk 1993 Broad variation in developmental and mature basidiome morphology of the ectomycorrhizal fungus Hydnangium sublamellatum sp. nov. bridges morphological based generic concepts of Hydnangium, Podohydnangium and Laccaria. Mycological Research 97: 613-619[ISI]
Bruns T. D. R. Fogel T. J. White J. D. Palmer 1989 Accelerated evolution of a false-truffle from a mushroom ancestor. Nature 339: 140-142[CrossRef][Medline]
Claridge M. F. M. A. Castellano J. M. Trappe 1996 Fungi as a food resource for mammals in Australia. In A. E. Orchard [ed.], Fungi of Australia 239–267. CSIRO Publishing, Canberra, Australia
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][ISI]
Fogel R. 1985 Studies in Hymenogaster (Basidiomycotina): a re-evalution of the subgenus Dendrogaster. Mycologia 77: 72-82[CrossRef][ISI]
Fogel R. 1992 Evolutionary processes in truffles and false truffles: evidence from distribution of hypogeous fungi in Great Basin, USA. Micologia e Vegetazione Mediterranea 7: 13-30
Fogel R. 1994 Materials for a hypogeous mycoflora of the Great Basin and adjacent cordilleras of Western United States. II. Two submerged species Cortinarius saxamontanus sp. nov., and C. magnivelatus, plus comments on their evolution. Mycologia 86: 795-801[CrossRef][ISI]
Gill M. 1995 New pigments of Cortinarius Fr. and Dermocybe (Fr.) Wünsche (Agaricales) from Australia and New Zealand. Beihefte zur Sydowia 10: 73-87
Goldman N. 1993 Simple diagnostic statistical tests of models for DNA substitution. Journal of Molecular Evolution 37: 650-661[ISI][Medline]
Grubisha L. J. M. Trappe R. Molina J. W. Spatafora 2001 Biology of the ectomycorrhizal genus Rhizopogon. V. Phylogenetic relationships in the Boletales inferred from LSU rDNA sequences. Mycologia 93: 82-89[CrossRef][ISI]
Hansen L. H. Knudsen 1992 Nordic Macromycetes, vol. 2, Polyporales, Boletales, Agaricales, Russulales. Nordsvamp, Kopenhagen, Denmark
Harrington T. C. D. M. Rizzo 1999 Defining species in the fungi. In J. J. Worrall [ed.], Structure and dynamics of fungal populations, 43–70. Kluwer Academic Press, Dordrecht, The Netherlands
Hasegawa M. H. Kishino T.-A. Yano 1985 Dating of the human-ape splitting by molecular clock of mitochondrial DNA. Journal of Molecular Evolution 21: 160-174
Heim R. 1971 The interrelationships between the Agaricales and gasteromycetes. In R. H. Petersen [ed.], Evolution in the higher basidiomycetes, 505–534. University of Tennessee, Knoxville, Tennessee, USA
Hibbett D. S. E. M. Pine E. Langer G. Langer M. J. Donoghue 1997 Evolution of gilled mushrooms and puffballs inferred from ribosomal DNA sequences. Proceedings of the National Academy of Sciences, USA 94: 12002-12006
Hibbett D. S. R. G. Thorne 2000 Basidiomycota: Homobasidiomycetes. In D. S. McLaughlin, E. G. McLaughlin, and P. A. Lemke [eds.], The Mycota VII, Part B: Systematics and evolution, 121–168. Springer Verlag, Berlin, Germany
Hibbett D. S. A. Tsuneda S. Murakami 1994 The secotioid form of Lentinus tigrinus: genetics and development of a fungal morphological innovation. American Journal of Botany 81: 466-478[CrossRef][ISI]
Hoiland K. A. Holst Jensen 2000 Cortinarius phylogeny and possible taxonomic implications of ITS rDNA sequences. Mycologia 92: 694-710[CrossRef][ISI]
Hopple J. S., Jr. R. Vilgalys 1994 Phylogenetic relationship among coprinoid taxa and allies based on data from restriction site mapping of nuclear rDNA. Mycologia 86: 96-107[CrossRef][ISI]
Hopple J. S., Jr. R. Vilgalys 1999 Phylogenetic relationships in the mushroom genus Coprinus and dark-spored allies based on sequence data from the nuclear gene coding for the large ribosomal subunit RNA: divergent domains, outgroups, and monophyly. Molecular Phylogenetics and Evolution 13: 1-19[CrossRef][ISI][Medline]
Horak E. 1987 New species of Dermocybe (Agaricales) from New Zealand. Sydowia Annales Mycologici 40: 81-112
Huelsenbeck J. P. B. Rannala 1997 Phylogenetic methods come of age: testing hypotheses in an evolutionary context. Science 276: 227-232
Ingold A. 1965 Spore liberation. Clarendon Press, Oxford, UK
Johannesson H. M. P. Martin 1999 Cladistic analysis of European species of Rhizopogon (Basidiomycota) based on morphological and molecular characters. Mycotaxon 71: 267-284[ISI]
Kendrick B. 1992 The fifth kingdom. Focus Information Group, Newburyport, Massachusetts, USA
Kishino H. M. Hasegawa 1989 Evaluation of the maximum likelihood estimates of the evolutionary tree topologies from sequence data, and the branching order in Hominoidea. Journal of Molecular Evolution 29: 170-179[CrossRef][ISI][Medline]
Kretzer A. T. D. Bruns 1997 Molecular revisitation of the genus Gastrosuillus. Mycologia 89: 586-589[CrossRef][ISI]
Kühner R. 1980 Les Hyménomycètes agaricoïdes: étude générale et classification. Numéro spécial du Bulletin de la Société Linnéenne de Lyon, France.
Lago M. N. L. Bougher M. L. Castro 2000 Morphological variability and implications for definition of taxa in the Descolea-Setchelliogaster-Descomyces complex. Mycotaxon 18: 37-57
Liu Y. J. S. O. Rogers J. F. Ammirati 1997 Phylogenetic relationships in Dermocybe and related Cortinarius taxa based on nuclear ribosomal DNA internal transcribed spacers. Canadian Journal of Botany (Revue Canadienne De Botanique) 75: 519-532
Martin M. P. N. Hogberg J. Llistosella 1999 Macowanites messacipoides, a hypogeous relative to Russula messapica. Mycological Research 103: 203-208[CrossRef][ISI]
Martin M. P. A. Roccabruna 1999 The taxonomic boundaries between Naucoria rheophylla and Setchelliogaster tenuipes based on morphology and molecular data. Mycotaxon 71: 141-148[ISI]
May T. W. 1995 Notes on Protoglossum (Fungi, Cortinariales). Muelleria 8: 287-289
Miller S. L. P. Torres T. M. McClean 1994 Persistence of basidiospores and sclerotia of ectomycorrhizal fungi and Morchella in soil. Mycologia 86: 89-95[CrossRef][ISI]
Miller S. L. T. M. McClean J. F. Walker B. Buyck 2001 A molecular phylogeny of Russulales including agaricoid, gasteroid and pleurotoid taxa. Mycologia 93: 344-354[CrossRef][ISI]
