Rhizochaete, a new genus of phanerochaetoid fungi

A new basidiomycete genus, Rhizochaete (Phanerochaetaceae, Polyporales), is described. Rhizochaete is characterized by a smooth to tuberculate, pellicular hymenophore and hyphal cords that turn red or violet in potassium hydroxide, monomitic hyphal system of simple or nodose septate hyphae, cystidia, and small, cylindrical to subglobose basidiospores. It morphologically is most similar to Phanerochaete. Analyses of nuclear ribosomal and internal-transcribed spacer region sequence data support a close relationship between Rhizochaete and Phanerochaete. The new taxon R. brunnea, from southern Argentina, is described and illustrated. In addition, the new combinations R. americana, R. borneensis, R. filamentosa, R. fouquieriae and R. radicata are proposed. A key to the species of Rhizochaete is provided.


INTRODUCTION
During a survey of Corticiaceae sensu lato growing on Nothofagus pumilio (Poepp. & Endl.) Krass. (Greslebin and Rajchenberg 1997a, b, 1998, Greslebin 2001 in the southern forests of Argentina (Cabrera Accepted for publication August 27, 3003. 1 E-mail: alina@ciefap.cyt.edu.ar 2 Corresponding author. E-mail: knakasone@fs. fed.us and Willink 1973), an undescribed taxon whose hymenial surface turned violet with drops of 2-5% KOH was found. The generic placement of this taxon could not be determined readily from its morphological features because it possessed characters assignable to several genera. The basidiocarp and the hyphal system had a phanerochaetoid appearance, but the hyphae were clamped regularly. In addition, the tubular cystidia with thickened walls were similar to those developed in some species of Crustoderma but differed in being encrusted with crystals and granular material. The taxon was associated with white rot, but the test for extracellular oxidases resulted in a negative or a very weakly positive reaction. The affiliation of this taxon to Phanerochaete P. Karst., Phlebia Fr., Hyphoderma Wallr., Crustoderma Parmasto and Ceraceomyces Jülich was evaluated, but in all cases the new species did not conform to important features of these genera. Several species in the above-mentioned genera that had the hymenial surfaces turning red-violet with KOH solution showed similarities in hyphal morphology, type of encrustation and cystidia. Because morphological features were insufficient to establish a proper generic disposition, the large and small subunits of the nuclear ribosomal DNA and internal-transcribed spacer (ITS) region were sequenced and analyzed. The analyses showed a close relationship between the Argentinean taxon and Ceraceomyces americanus Nakasone, C. R. Bergman & Burds., Ceraceomyces fouquieriae (Nakasone & Gilb.) Nakasone, C.R. Bergman & Burds., Phanerochaete filamentosa (Berk. & M.A. Curtis) Burds. and Phanerochaete radicata (Henn.) Nakasone, C. R. Bergman & Burds. In this paper we describe the new genus Rhizochaete to accommodate these taxa, based on morphological and molecular studies.

MATERIALS AND METHODS
Morphological and cultural studies.-Freehand sections of fresh and dried basidiocarps were examined microscopically, mounted in 2-5% potassium hydroxide (KOH) and 1% aqueous phloxine, Melzer's reagent (reactions amyloid, dextrinoid or none [ϭIKI-]; Kirk et al 2001), 0.1% cotton blue in lactophenol and 1% aqueous cresyl-blue (reaction metachromatic when walls turn purple). Color descriptions were taken from Munsell (1954) and herbarium designations from Holmgren et al (1990). Cultures were obtained from context tissue of fresh basidiocarps or isolated from the associated wood rot and are kept at the culture collection at CIEFAP. Cultural features were studied and described according to Nobles (1965). The species code describing the cultures follows the system of Nobles (1965), with several modifications summarized by Nakasone (1990). Line drawings of microscopic features were made with a drawing tube on the microscope. Unless otherwise indicated, all specimens are at CIEFAP. Phylogenetic analyses.- Taxa used in the phylogenetic analyses are listed in TABLE I. Three datasets, representing three different gene regions of the nuclear ribosomal gene, were analyzed. Based on previous phylogenetic studies of Homobasidiomycetes using the nuclear small-subunit ribosomal RNA gene (SSU rRNA) (Hibbett andDonoghue 2001, Lim 2001), 28 taxa were included in the first dataset, and Gloeophyllum sepiarium (Wulf. : Fr.) P. Karst. was chosen as outgroup taxon. The nuclear large-subunit ribosomal RNA (LSU rRNA) gene region includes 36 taxa of which 19 also were represented in the SSU rRNA dataset. Results from the SSU rRNA analysis and Parmasto's and Hallenberg's (2000) study of the phylogenetic relationships of phlebioid fungi based on the LSU rRNA were used to determine the taxa included in the LSU rRNA dataset. Gloeophyllum sepiarium was designated outgroup taxon in the LSU rRNA dataset as well. The third dataset consists of sequences of the internal transcribed spacer region, including the 5.8S rRNA gene (ITS). Forty taxa were included in the ITS dataset. The taxa in this dataset were based on previous studies (Boidin et al 1998;de Koker et al 2003) and included a number of Phanerochaete species. Representative strains of five taxa of Rhizochaete were included in all three datasets.
Cultures and voucher specimens of strains sequenced in this study are deposited at CFMR (TABLE I). Cultures were grown in 50 ml of sterile 2% malt extract supplemented with 1% glucose and 0.1% yeast extract at 25 C for 1 wk. The cultures were harvested by filtration onto Miracloth (Chicopee Mills Inc., La Jolla, California), lyophilized and stored at Ϫ20 C. Total DNA was extracted following the protocol outline in Cenis (1992), with minor modifications, and further purified with a Geneclean kit (Qbiogene, Carlsbad, California). The ITS region was amplified with primers ITS5 and ITS4, the SSU RNA gene with primers NS1 and NS8, and the 5Ј end of the LSU RNA gene with primers LR0R and LR7 (White et al 1990) using a Taq PCR Core Kit (Qiagen, Hilden, Germany) in a PTC 200 DNA Engine thermal cycler (MJ Research, Watertown, Massachusetts). Primers were prepared by Operon Technologies Inc. (Alameda, California). Cycling parameters were: 1 cycle at 93 C for 2 min, followed by 35 cycles at 53 C for 2 min, 72 C for 2 min, and 93 C for 1 min. Amplified DNA products were cleaned with a QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) then sequenced with an ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems, Foster City, California), following the manufacturer's protocol. Primers used for sequencing were ITS1, ITS3 and ITS2 or ITS6 for the ITS region, NS1, NS2, NS3, NS5, NS7, NS6, NS8 and sometimes SR4 for the SSU rRNA gene, and LR0R, LR3R, LR17R, LR3, LR5 and LR7 for the LSU rRNA gene, (http://www.biology.duke.edu/fungi/mycolab/ primers.htm). Sequences were obtained from an ABI Prism 377-18 DNA Sequencer (PE Biosystems, Foster City, California). This overlapping sequencing strategy resulted in the DNA regions being sequenced in both directions, except in a few areas. Sequences obtained from this study were submitted to GenBank (AY219389-AY219404).
Sequences obtained from this study were aligned manually in PAUP* 4.0b10 (Swofford 2002) and McClade (Maddison and Maddison 1992) with those obtained from GenBank (TABLE I). The SSU rRNA, LSU rRNA and ITS regions were analyzed separately. The aligned sequences are available from TreeBASE (S972). In each region, ambiguous sites were excluded before analyses. Phylogenetic analyses of the sequence data were performed with maximumparsimony (MP) and maximum-likelihood (ML) methods, as implemented in PAUP, and with Bayesian inference, using MrBayes version 2.01 (Hulsenbeck and Ronquist 2000). For MP analyses, an initial heuristic search of 100 random taxon addition replicate searches was conducted with TBR branch-swapping, MAXTREES set to autoincrease, without constraints, unordered and equally weighted nucleotides, and retention of two shortest trees. The shortest trees were used as starting trees in a second heuristic search, with TBR branch swapping and MAXTREES set to 5000 to find the most-parsimonious trees. Bootstrap support for clades (Felsenstein 1985) was estimated from 1000 replicate heuristic searches with simple taxon addition sequence, retention of one tree per replicate, TBR branch swapping, and MAX-TREES set to 5000. Consistency indices (CI, Kluge and Farris 1969) and retention indices (RI, Farris 1989) exclude uninformative characters. Decay indices (DI, Bremer 1988) were determined with AUTODECAY 4.0 (Eriksson 1998). The program MODELTEST 3.06 (Posada and Crandall 1998) performed nested likelihood ratio tests to determine the best model of sequence evolution for the three datasets. The values obtained from MODELTEST then were used in ML and Bayesian analysis. ML heuristic searches were performed in PAUP with TBR branch swapping. Bayesian analysis was implemented in MrBayes with four Markov chain Monte Carlo chains with no molecular clock enforced. One million or 1 500 000 generations were performed, with every 100 trees sampled. The first 1000 or 1500 trees were excluded from construction of the consensus tree. Bayesian analyses were performed three times to confirm the consistency of the consensus tree and posterior clade probabilities.   Etymology. Rhizo (Gr. Rhiza ϭ root) referring to the rhizomorphs ϩ chaete (Gr. Chaite ϭ hair, setae, spine, bristle) referring to the presence of protruding cystidia.

Rhizochaete
Basidiocarp pellicular to membranaceous, subceraceous when fresh, coriaceous but friable to firm papyraceous upon drying; readily detachable from substrate and/or subiculum. Hymenial surface continuous, smooth to slightly tuberculate, velutinous, yellowish, orange or brownish colored, turning red to violet with drops of KOH solution. Context packedhypochnoid to densely wooly. Margin distinct, fimbriate to fibrillose; hyphal cords usually abundant, turning red to violet in KOH.
Hyphal system monomitic; generative hyphae with variable septation (some species simple septate, some species mostly simple septate but a few septa with clamps, and other species regularly clamped), thin to thick-walled, encrusted with dark yellow to yellowish brown, resinous-like granules that dissolve in KOH turning the solution pale violet; some hyphae also encrusted with hyaline crystals that do not dissolve in KOH. Crystalline encrustation usually restricted to subicular hyphae, and crystals aggregated in clusters or rosette-like structures larger and coarser than the colored granules. Subhymenium well developed, composed of tightly packed hyphae. Subiculum an open and loose textura intricata; a basal textura porrecta stratum usually present next to the substrate. Hyphal system and hymenial elements metachromatic. Cystidia present in all but one species, cylindrical to subfusiform, thin to thick-walled, always encrusted with dark yellow to yellowish-brown resinous-like granules that dissolve in KOH, hyaline crystals usually present. Basidia clavate to subcylindrical and sinuous, thinwalled or thickening towards the base. Basidiospores small, up to 6.5(-7) m long, short cylindrical to ellipsoid, widely ellipsoid to subglobose in one species, thin-walled, sometimes appearing slightly thickened, smooth, inamyloid. Associated rot white, but extracellular oxidase test of cultures may produce a weak or negative reaction.
Comments. This genus is characterized by the combination of detachable, subpellicular to membranaceous basidiocarps, hymenial surface and hyphal cords that turn red or violet in KOH, hyphae and cystidia with two types of encrustation (dark yellow to yellowish-brown granules and hyaline crystals), and small, cylindrical to subglobose basidiospores with thin to slightly thickened walls. The color change of the hymenial surface and hyphal cords is produced by the dark yellow to yellowish-brown granules that coat the hyphae as they dissolve and turn violet in KOH. It is an acid-base reaction as the application of an acid solution recovers the original hymenial color.
Rhizochaete easily is distinguished from morphologically similar corticioid genera. For example, Hyphoderma has cystidia that are similar to those in Rhizo- chaete but its basidiospores are significantly larger. Crustoderma is associated with a brown rot, whereas Rhizochaete is associated with a white rot. Phlebia sensu stricto (Hjortstam 1997) and Phlebiopsis Jülich can be distinguished by their ceraceous to subgelatinous basidiocarps and tightly packed, agglutinated, subicular hyphae.
Rhizochaete is easily distinguished from the closely related genus Phanerochaete by the hyphal septation in the case of regularly clamped species. Simple-septate species of Rhizochaete develop rare single clamps but never multiple clamps that can be found in Phanerochaete species. Furthermore, all species of Rhizochaete develop a red or violet reaction of both the hymenium and hyphal cords to KOH. Although in some species of Phanerochaete, the hymenium turns red in KOH, the hyphal cords do not, e.g., P. laevis, P. salmoneolutea and P. subceracea. In other species such as P. burtii, the hyphal cords but not the hymenium turn red in KOH. The hymenia of P. carnosa and P. sanguinea produce a dark green or olive green reaction in KOH. A comparison of the distinguishing basidiocarp traits mentioned above for Rhizochaete and morphologically similar taxa in Ceraceomyces, Phanerochaete and Phlebiopsis is presented in TABLE II.

Rhizochaete brunnea
Basidiocarp resupinate, membranous, thick (0.3-2 mm) when fresh, with a coriaceous aspect but breaking readily upon drying, detachable from the substrate. Hymenial surface even to slightly tuberculate, velutinous to pilose under a 10ϫ lens by the protruding cystidia, when fresh dark yellow, brownishyellow or brown (10YR 6/8, 5/6; 7/5YR 5/6), slight vinaceous when dried, turning violet in KOH solution (the original color being recovered upon the application of an acid solution), the coloration is due to the encrusted cystidia and its intensity varies according to cystidia abundance. Context up to 1.5 mm thick, with a compact hypochnoid texture, brownish yellow (10YR 6/8). Margin generally fibrillose, white or yellow, paler than the hymenial surface, with hyphal cords. Hyphal cords dark yellow, 100-1200 m diam, firm, branched, abundant in the margin, developed under the basidiocarp and throughout the substrate.
Hyphal system monomitic; generative hyphae clamped, thick-walled, metachromatic, heavily encrusted with small, granular, dark melleous to chestnut-colored material that readily dissolves in KOH solution and turns the solution lilaceous; some hyphae, especially subicular ones, encrusted with polyhedral, hyaline crystals that do not dissolve in KOH. Subhymenial hyphae tightly intertwined and arranged perpendicular to the substrate, a compact textura intricata or intricata-porrecta, 5-6 m diam, with walls thickened up to 1 m. Subiculum a loose and open textura intricata, subicular hyphae up to 10 m diam, with walls up to 2 m thick, sometimes with secondary simple or ampullate septa, clamps sometimes difficult to discern; toward the base of subiculum hyphae arranged more or less parallel to the substrate; a basal stratum next to substrate usually present, a textura porrecta arranged parallel to the substrate. Hyphal cords composed of an inner core of parallel, tightly packed hyphae, 4-6 m diam, clamped, with walls slightly thick to thickened, hyaline, and smooth, and wider hyphae 10-28 m diam, sparsely septate, with walls thin to thick, containing refringent material that strongly stain with phloxine (appearing gloeopleurous-like) or, if lacking staining material then with walls up to 4 m thick; outer layer composed of closely or loosely intertwined, yellowish hyphae 4-6 m diam, with walls slightly thick to thick, heavily encrusted with granular, chestnut-colored granules and scattered, hyaline, polyhedral crystals. Cystidia cylindrical, (80-)100-250 ϫ 8-15 m, with thickened walls up to 4 m except in the apex, metachromatic, some with adventitious septa, heavily encrusted with both chestnut-colored material and hyaline crystals. Basidia narrowly clavate, 40-55(-60) ϫ 5-6 m, with 4 sterigmata and a basal clamp, thickwalled toward the base, walls metachromatic. Basidiospores ellipsoid, 5-6.5(-7) ϫ 3-3.5 m, thin-walled, smooth, IKI-, guttulate.
Cultural description.-F IGS. 5-10 Cultures studied. No. 229, from basidiocarp M. Rajchenberg 11455;No. 230, from associated decayed wood and mycelia of basidiocarp A. Greslebin 1577. Macroscopic characters. Growth very slow, 6-6.5 cm radius by 6 wk. Margin regular, hyaline, submerged in the agar. Behind the margin a woolly mat is formed, first as isolated punctual flakes that develop into a heterogeneous, felty to woolly, dark yellow to brownish yellow mat, with scattered denser areas, often with incipient hyphal cord formation. Drops of KOH solution turn the mat lilac or violet, but the color vanishes rapidly. Reverse bleached. Odor slightly sweet, fruity.
Microscopic characters. Marginal hyphae clamped, 3-5 m diam, thin-walled, hyaline, branched, with long hyphal segments. Aerial mat with thin-and thick-walled generative hyphae covered with minute, dark yellow to brownish granules that readily dissolve in KOH solution. The size of these granules obscures their shape. At wk 6 some hyphae develop gelatinous, rough walls. Hyaline, polyhedral crystals formed on the hyphae and in the agar.
b Proportion of invariant sites. c Gamma distribution shape parameter. FIG. 11. Strict consensus of 385 most-parsimonious trees (tree length ϭ 209 steps, CI ϭ 0.496, RI ϭ 0.712) of the small-subunit ribosomal RNA gene region. Numbers above the branches are average posterior clade probabilities from three Bayesian analyses. Decay indices are indicated above the branches preceded by a ϩ sign. Bootstrap confidence levels are shown below the branches; values Ͻ50% are not shown. Asterisks (*) and triangles (᭡) indicate nodes that collapse in the Bayesian and maximum-likelihood trees, respectively. Diamonds (ࡗ) with the same superscript letter indicate branches that are joined in maximum-likelihood and Bayesian trees.
Sequence alignments.-The SSU rDNA region sequence alignment totaled 1817 base pairs (bp), of which 119 characters (6.5%) were excluded because of ambiguity in alignment; 139 remaining characters were variable, of which only 56 (3%) were parsimony informative. In contrast, the ITS region was the shortest at 866 bp and was the most difficult to align. More than half of the ITS characters, 455 bp (53%), were excluded; 161 characters were variable, and of these 115 (29%) were parsimony informative. The LSU rDNA region was 932 bp long; 86 (9%) ambiguous characters were excluded, 222 characters were variable and 155 (17%) characters were parsimony informative.  FIG. 11 is congruent with, but slightly less resolved than, the ML and Bayesian trees. Rhizochaete is included in the unresolved Phanerochaete clade, nested within the larger Phlebia clade of Hibbett and Donoghue (2001).
Phylogenetic analyses of the LSU rDNA produced trees that generally are congruent with the trees of FIG. 12. Strict consensus of 14 most-parsimonious trees (tree length ϭ 739 steps, CI ϭ 0.323, RI ϭ 0.519) of the large-subunit ribosomal RNA gene region. Symbols and numbers are described in FIG. 11. FIG. 13. Strict consensus of 2658 most-parsimonious trees (tree length ϭ 453 steps, CI ϭ 0.433, RI ϭ 0.613) from maximum-parsimony analysis of the internal transcribed spacer region. Symbols and numbers are described for FIGS. 11 and 12. the SSU region, although less than half of the taxa are shared between the two datasets. There were 14 MP trees of 739 steps with CI ϭ 0.323 and RI ϭ 0.519. The strict MP consensus tree of the LSU region is shown in FIG. 12. In this tree, the Rhizochaete species form a monophyletic clade that is sister to a heterogeneous clade containing Phanerochaete chrysosporium, Phan. sordida, Bjerkandera adusta, Pulcherricium caeruleum, Phlebia deflectens and Phl. lilascens. These two clades and Phlebiopsis constitute the Phanerochaete clade. The ML and Bayesian trees generally are congruent but more resolved than the MP consensus tree. Some branches paired in the ML and Bayesian trees but not in the MP trees are indicated on the figure. Note that in the ML and Bayesian trees Phlebiopsis gigantea is included in a clade with R. fouquieriae and R. brunnea

DISCUSSION
The most striking and consistent character of the new genus Rhizochaete is the red to violet reaction of the basidiocarp and hyphal cords to KOH that is related to the dark yellow to yellowish-brown granules that coat the hyphae and cystidia. This feature, though, is present in some species of Phanerochaete, Ceraceomyces, Phlebia and unrelated taxa such as Hy-phodontia australis (Berk.) Hjortstam (Greslebin et al 2000). Other significant characters are the ''phanerochaetoid'' appearance of the hyphal system, the structure of the subiculum and subhymenium, the hyphal cords, and the shape and size of the basidiospores that are similar to some Phanerochaete species. The sum of these morphological characters indicates a close relationship of Rhizochaete to the genus Phanerochaete, and this relationship also is supported by molecular data.
Rhizochaete consists of species that have a monomitic hyphal system with either regularly nodose septate, regularly simple septate, or simple septate with scattered single clamps. In most genera in the Aphyllophorales, the species have one type of hyphal septation. It is not unusual, however, for one or more species to have simple septate hyphae in a genus of primarily nodose septate species. Examples of these genera include Hyphoderma Wallr., Hyphodontia J. Erikss., Phlebia Fr., Radulodon Ryvarden, Resinicium Parmasto and Veluticeps (Cooke) Pat. A few corticioid genera, namely Botryobasidium Donk and Peniophora Cooke, include a significant number of nodose septate and simple septate species.
Ribosomal DNA analyses support the formation of the new genus Rhizochaete. In general, analyses of the SSU, LSU and ITS-sequence data by MP, ML and Bayesian methods produced trees that support the monophyly of the Rhizochaete species. However, with ML and Bayesian analyses of the LSU, Phlebiopsis gigantea was included also in the Rhizochaete clade. Rhizochaete is closely related to the Phanerochaete sensu stricto clade. Of the taxa included in the datasets, Phlebiopsis gigantea, Phan. crassa and Phan. hiulca, appear to be the most closely related to Rhizochaete. These results are congruent with a phylogenetic study of the genus Phanerochaete that employed the ITS region (de Koker et al 2003).
The Rhizochaete clade is relatively consistent in the analyses of the three datasets, although the positions of other taxa are not. Most of the conflicting results involve taxa in the LSU trees. For example, Bjerkandera adusta is embedded in the Phanerochaete clade in the LSU trees but not in the SSU and ITS trees. Boidin et al (1998) found that three species of Bjerkandera clustered together in a distinct clade basal to the Phanerochaete clade in an analysis of the ITS region. Phlebia subserialis similarly clusters with the Phanerochaete sensu stricto group in the ITS trees, also reported by Boidin et al (1998), but joins other Phlebia species in LSU (Parmasto and Hallenberg 2000) and SSU rDNA sequence analyses (Suhara et al 2002). In another example, Phan. sanguinea clusters with Phan. burtii and Phan. carnosa in the Phanerochaete sensu stricto clade in ITS trees. However, in the LSU trees, Phan. sanguinea is in a clade with Meripilus giganteus and Albatrellus syringae, basal representative taxa of the polyporoid clade (Hibbett and Donoghue 2001) and far removed from the Phanerochaete clade. Perhaps some of these inconsistencies and other minor ones not mentioned could be resolved with better taxon sampling and the inclusion of protein coding sequences. Sequences from all three DNA regions unfortunately were available only for eight taxa, so a combined sequence analysis was not attempted.
In a study of the mitochondrial SSU rRNA gene, Ko and others (2001) reported that Phan. filamentosa clustered with Antrodia carbonica and Oligoporus fragilis instead of B. adusta and Phan. chrysosporium. This is not consistent with results presented here and might reflect a misidentified specimen or a different mode of evolution of the mitochondrial SSU rRNA gene from that of the nuclear rRNA genes.
In conclusion, Rhizochaete is a polythetic genus that is defined by the combination of basidiocarp macromorphology, including hyphal cords, and its reaction with KOH, hyphal septation, hyphal arrangement, two types of encrustation, cystidia, and basidiospore shape and size. The recognition of this new genus also is supported by molecular data.