Molecular Phylogeny of Acremonium and Its Taxonomic Implications Author(S): Anthony E

Mycological Society of America

Molecular Phylogeny of Acremonium and Its Taxonomic Implications Author(s): Anthony E. Glenn, Charles W. Bacon, Robert Price and Richard T. Hanlin Source: Mycologia, Vol. 88, No. 3 (May - Jun., 1996), pp. 369-383 Published by: Mycological Society of America Stable URL: http://www.jstor.org/stable/3760878 . Accessed: 04/09/2013 15:25

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Mycological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Mycologia.

http://www.jstor.org

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Mycologia,88(3), 1996, pp. 369-383. ? 1996 by The New YorkBotanical Garden, Bronx,NY 10458-5126

Molecular phylogeny of Acremonium and its taxonomic implications

Anthony E. Glenn1 phytes, we are proposing that the anamorphs of Epi? Department of Plant Pathology, Universityof Georgia, chloe and closely related asexual grass endophytes be 30602 Athens, Georgia reclassified into the new form genus Neotyphodium. and taxonomic considerations are also Charles W. Bacon Phylogenetic presented for other taxa. USDA-ARS, Toxicologycjf Mycotoxin Research Unit, Words: Russell Research Center,Athens, Georgia 30613 Key Clavicipitaceae, endophytes, Epichloe, Hypocreales, Neotyphodium, rDNA Robert Price Department of Botany, Universityof Georgia, Athens, Georgia 30602 INTRODUCTION Richard T. Hanlin Members of the family Clavicipitaceae (Hypocreales; Department of Plant Pathology, Universityof Georgia, are well known of a diverse Athens, Georgia 30602 Ascomycotina) pathogens assemblage of hosts including grasses, sedges, other ascomycetes, and insects. These fungi are found the and of the Abstract: Acremonium is generally considered to be throughout tropical temperate regions world 1950; White, a highly polyphyletic form genus containing distantly (Diehl, 1994b). Rogerson (1970) listed 31 in the but of these are related fungi. Sectional divisions within Acremonium genera family, eight considered to be The is divided into distinguish the clavicipitaceous grass endophytes of synonyms. family three subfamilies with the insect and sect. Albolanosa from the generally saprobic species fungal patho? in and of sections Acremonium, Chaetomioides, Gliomastix, gens variously placed the Oomycetoideae Cor- The are and Nectrioidea. In an effort to assess the possible dycipitoideae (Diehl, 1950). plant pathogens far the most studied and im? number of lineages currently placed within Acremon? by widely economically members of the and are classified as ium and to determine which groups of sexual asco? portant family the Diehl mycetes are phylogenetically affiliated with Acremon? subfamily Clavicipitoideae (Diehl, 1950). further divides the taxa of the ium species, maximum parsimony and neighbor-join- (1950) Clavicipito? ideae into the three tribes Balansieae, ing analyses were performed using partial sequences Clavicipiteae, of the nuclear small subunit ribosomal DNA (18S and Ustilaginoideae. The fungus-host interactions of these taxa can be as rDNA). Acremonium was shown to be a polyphyletic broadly categorized being epibiotic or or and heavi? taxon with affiliations to at least three ascomycetous endophytic, systemic localized, or mutualistic. orders: 1) most of the examined species from the ly pathogenic, moderately pathogenic, of are in the tribe Bal? sections Acremonium, Gliomastix, and Nectrioidea Most the grass pathogens ansieae and are either localized to showed a relationship to the Hypocreaceae even (Diehl, 1950) epibiotic stromata on leaves or inflores- though many of these species have never been asso? reproductive cences or form infections inter- ciated with any teleomorph; 2) the grass endophytes systemic endophytic within their hosts in addition to of sect. Albolanosa and other taxa from the Clavicip? cellularly forming external stromata. The Balansieae in? itaceae formed a monophyletic group derived from reproductive clude the taxa within the Hypocreales; 3) the thermophilic A. ala- teleomorphic Atkinsonella, Balansia, bamense of sect. Chaetomioides was derived from with? Balansiopsis, Epichloe, and Myriogenospora (Diehl, Luttrell and et in the Sordariales. Acremonium alternatum, the type 1950; Bacon, 1977; Rykard al, 1984). is without species of the genus, was one of the species showing Myriogenospora completely superficial any of the host and Ba? affiliation to the Hypocreaceae. In order to eliminate penetration epidermis (Luttrell White and but Atkinsonella some of the heterogeneity within Acremonium while con, 1977; Glenn, 1994), has some localized intercellular associated also emphasizing the unique biological, morpholog? hyphae with the stroma and Mor? ical, and ecological characteristics of the grass endo- (Leuchtmann Clay, 1988; gan-Jones and White, 1989). Balansia contains a few Accepted for publicationFebruary 8, 1996. epibiotic species (Leuchtmann and Clay, 1988; Clay 1 email: [email protected]. and Frentz, 1993), but most species have a well de-

369

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions 370 Mycologia veloped endophytic habit (White et al., 1995). All untenable as a generic character and merged the two species of the Epichloe typhina (Pers. : Fr.) Tul. group, under the older Acremonium. The form genus Gliom- including the apparent asexual derivatives of Epich? astix and monophialidic species of Paecilomyces were loe, are intercellular grass endophytes (White, also merged into Acremonium (Gams, 1971), thereby 1994b). Asexual grass endophytes were largely over- expanding its definition to comprise generally slow- looked as being related to Epichloe until Bacon et al. growing species with hyaline or pigmented conidia (1977) found intercellular hyphae growing within tall that are one-celled or exceptionally two-celled, in fescue. They referred to the fescue endophyte as a chains or slimy heads, and produced from ortho- biotype of E. typhina based on concurrent in vitro phialides or basitonously branched conidiophores. examinations and the earlier report by Sampson Acremonium is perceived to be a heterogeneous (1933). The obscurity of these asexual endophytes taxon because several morphologically distinct teleo? was a result of the host grasses never showing any morphic genera have Acremonium-like anamorphs. symptoms or signs of a fungal infection. For the ana? Most of the known teleomorphs of Acremonium are morph ofthe fungus, Bacon et al. (1977) applied the species of Nectria (Gams, 1971; Samuels, 1973, name Sphacelia typhina Sacc. which was the binomial 1976a,b; Lowen, 1995), but teleomorphic genera designated by Saccardo (1881) for the anamorph of such as Hypocrea, Hypomyces, Thielavia, Pronectria, E. typhina. However, S. typhina was later considered Nectriopsis, Epichloe, Emericellopsis, Mycoarachis, and a nomen dubium by Morgan-Jones and Gams (1982) Nigrosabulum also have anamorphs classified in Acre? because of insufficient type material. Typhodium monium (Malloch and Cain, 1970; Gams, 1971; Mor? (Link, 1826) is another genus of uncertain applica? gan-Jones and Gams, 1982; Samuels, 1988; Lowen, tion to the anamorph of E. typhina and is now con? 1995). Gams (1971, 1975) divided Acremonium into sidered a synonym of Epichloe (Hawksworth et al., three sections: Acremonium, Nectrioidea, and Gliom- 1983). However, as a "convention of convenience," astix. The majority of the species in these sections Diehl (1950) applied Typhodium as a form genus and are saprobic in a wide variety of habitats, but some used the term "typhodial" when referring to the ana? are plant parasites. Some species of sect. Nectrioidea morph of Epichloe. have teleomorphs in the genus Nectria, but most of Based on in vitro similarities, Morgan-Jones and the species in the three sections do not have any Gams (1982) placed the anamorphs of both the fes? known association to teleomorphs. Along with the es- cue endophyte and E. typhina in the form genus Acre? tablishment of sect. Albolanosa, Morgan-Jones and monium. The authors erected the new sect. Albolan? Gams (1982) also erected the new sect. Chaetomioides osa for these endophytic fungi with the in vitro char? for anamorphs of ascomycetes within the family acteristics of slow-growing white to yellow colonies, Chaetomiaceae that have, among other characteris? solitary phialides narrower at the base than the sub? tics, short-aculeate to lageniform phialides. Acremon? tending hyphae, hyaline and smooth-walled conidia, ium alabamense Morgan-Jones, the type species of and teleomorphs in the Clavicipitaceae. They stated sect. Chaetomioides, is a thermophilic fungus origi? that the exclusively solitary phialides of the endo? nally isolated from fallen pine needles (Morgan- phytes preclude them from classification in Verticil- Jones, 1974), and it is the anamorph of Thielavia ter- lium sect. Prostrata which was restricted to those taxa, restris (Apinis) Malloch & Cain (Morgan-Jones and such as Cordyceps spp., that possess verticillate as well Gams, 1982). Other species of this section are ana? as solitary phialides (see also Gams, 1971). Morgan- morphs of several Chaetomium species and cannot be Jones and Gams (1982) reported slight in vitro mor? distinguished without their teleomorphs (Morgan- phological differences between the anamorph of E. Jones and Gams, 1982). Most recently, sect. Licheno- typhina and those of Holcus mollis L., Dactylis glom- idea was erected for lichenicolous species (Lowen, erata L., and Sphenopholis obtusata (Michx.) Scribn., 1995), and some of these species have hypocreaceous which they treated as A. typhinum Morgan-Jones & teleomorphs while others are not associated with any W. Gams, and the tall fescue endophyte, which they teleomorph. described as the separate species A. coenophialum Because of the possible heterogeneity of Acremon? Morganjones 8c W. Gams. ium, the placement of the grass endophytes in sect. Acremonium is a cosmopolitan, morphologically Albolanosa was done with no claim of phylogenetic simple genus (see Gams, 1971). Acremonium alterna- relationship to the nonendophytic species in the oth? tum Link : Fr. is the lectotype and produces conidia er sections (Morgan-Jones and Gams, 1982; White in chains or heads. The genus Cephalosporium has and Morgan-Jones, 1987). This classification scheme often been distinguished from Acremonium because was based strictly on in vitro morphological similari- of the accumulation of conidia in slimy heads. How? ties. Some disagreement with this classification of the ever, Gams (1971) concluded heads versus chains was endophytes was initially expressed, and it continues

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Glenn et al.: Molecular Phylogeny of Acremonium 371

Table I. Acremoniumtaxa included in the analyses and the various sections in which they are classified3

a Teleomorphs are indicated if they are known, and precedence in the listing is given to the teleomorphs except for the ease of A. alabamense which was obtained from ATCC (#26796) as A. alabamense and not as T terrestris. bGams (1971). cGams (1975). d Morgan-Jones and Gams (1982). e See Samuels et al. (1991) and Gams and Van Zaayen (1982).

to be debated (Latch et al, 1984; Rykard et al., 1984; tomioides, Gliomastix, and Nectrioidea for inclusion in Morgan-Jones et al., 1992; Gams, 1995). The simple the rDNA comparisons (Tables I, II). Isolates from of re- morphology of the anamorphs may have been de? sect. Lichenoidea were not sampled because the rived multiple times in the evolution of fungi. How? cent establishment of this newest section. All isolates ever, the connection of so many Acremonium species of Acremonium were either type or authenticated cul? to Nectria and other genera of the Hypocreaceae sug? tures, and additional taxa were either personally col? researchers gests that while there may be some heterogeneity cre? lected or obtained from other (Table II). ated by sect. Albolanosa and sect. Chaetomioides, Acre? The fungi were maintained on cornmeal-malt agar monium may be more homogeneous than postulated. (CMM) (Rykard et al., 1982) at room temperature. The application of strict monophyly has been sug? Nucleic acid extraction.?For extraction of total ge? gested for fungal systematics and taxonomy so as to nomic DNA from mycelium, all isolates were grown reflect the phylogenetic history of a group of organ? in M102 liquid medium (Rykard et al., 1982) on a isms (Vilgalys et al, 1993). If a monophyletic classi? rotary shaker (200 rpm) at room temperature until fication scheme is to be the ultimate goal for any adequate growth occurred (usually 1-2 wk). The my? group of fungi, then the current delimitation of Acre? celium was collected by centrifugation, and the pel- monium is to Molecular open question. phylogenetic leted tissue was washed once with sterile distilled wa? data are in particularly helpful resolving relationships ter to remove excess medium. The tissue was then of morphologically simple organisms such as those in ? ground in liquid nitrogen and stored at 80 C until Acremonium. Therefore, using sequences of the nu? ready for nucleic acid extraction. Extractions were clear encoded small subunit ribosomal DNA (18S made from 0.2-0.5 g (wet weight) of ground myce? this was undertaken to 1) assess the rDNA), project lium. minimum number of separate lineages currently The nucleic acid extraction procedure was a mod- placed within the form genus Acremonium, 2) deter? ification of Lee and Taylor (1990) in which the chlo? mine which groups of sexual ascomycetes are closely roform :phenol step was repeated at least once so as affiliated with Acremonium species in the current clas? to remove as much cellular debris as possible. The sification, 3) determine if Acremonium sect. Albolan? extracted nucleic acid samples were diluted to pro- osa is appropriate for classification of the grass en? vide solutions with a DNA concentration range of 0.1 dophytes, and if not, 4) propose a new taxon for this to LO ng |xL l. group. Polymerase chain reaction and sequencing.?Fifty jjlL of each diluted was used as for poly? MATERIALS AND METHODS sample template merase chain reactions (PCR) (Mullis and Fallona, Fungal isolates.?A total of fifteen taxa having ana? 1987; Saiki et al., 1988). For each amplification, a morphs currently classified in Acremonium were se? total reaction volume of 100 pU was made containing lected from sections Acremonium, Albolanosa, Chae- diluted template, 10 mM Tris-HCl (pH 8.3), 50 mM

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions 372 Mycologia

Table II. Taxa included in the analyses, source of isolates, and accession numbers for 18S rDNA sequences

Taxon Isolate sourcea GenBank accession #

Acremonium alabamense ATCC 26796b U43969 Acremonium alternatum CBS 223.70 8c 406.66 U43970 Acremonium chrysogenum ATCC 14615b U43971 Acremonium coenophialum ATCC 52274b; Festuca arundinacea U45942 Acremoniumfurcatum CBS 122.42b & 508.65 U43972 Acremonium kiliense ATCC 347l6b U43973 Acremonium murorum CBS 214.69 U43966 Acremonium rutilum CBS 225.70 8c 394.66 U43967 Acremonium strictum ATCC 34717b U43968 Acremonium uncinatum J. F. White; Festuca pratensis U45943 Ascosphaera apis UCB 78-018 M83264 Atkinsonella hypoxylon A. E. Glenn; Danthonia spicata U44034 Balansia aristidae J. F. White; Aristida sp. U44035 Balansia henningsiana A. E. Glenn; Andropogon sp. U44036 Balansia obtecta J. F. White; Cenchrus echinatus U44037 Balansia sclerotica ATCC 16582 U32399 Balansia strangulans J. F. White; Panicum sp. U44038 Candida tropicalis MUCL 30002 M55527 Ceratocystisfimbriata T. C. Harrington C89 U32418 Chaetomium globosum A. E. Glenn U44039 Claviceps purpurea R. T. Hanlin; Dactylis glomerata U44040 Cordycepscapitata A. E. Glenn; Elaphomyces sp. U44041 Cryhonectriaparasitica R. T. Hanlin 5806 Denise Silva, unpubl. Daldinia concentrica ATCC 36659 U32402 Diaporthe phaseolorum F. A. Uecker 458 L36985 Echinodothis tuberiformis J. F. White; Arundinaria tecta U44042 Emericellopsisminima ATCC 16216 U44043 Emericellopsisterricola CBS 120.40b U44112 Epichloe amarillians J. F. White; Agrostis hiemalis (Ah3) U35034 Epichloe festucae J. F. WTiite; Festuca rubra rubra U44113 Eurotium rubrum UCB 88-016 U00970 Hypocrea pallida G. J. Samuels 89-83 U32408 Hypomycespolyporinus ATCC 46844 U32410 Microascus trigonosporus RSA 1942 L36987 Monascus purpureus ATCC 16365 M83260 Myriogenosporaatramentosa A. E. Glenn; Andropogon sp. U44114 Myriogenosporaatramentosa A. E. Glenn Erianthus brevibarbis U44115 Nectria cinnabarina G.J. Samuels 89-107 U32412 Nectria vilior ATCC 16217 U44116 Neocosmospora vasinfecta ATCC 28867 U44117 Saccharomycescerevisiae M27607 Taphrina deformans ATCC 34556 U00971 Xylaria hypoxylon ATCC 42768 U20378 = = aATCC American Type Culture Collection, USA; CBS Centraalbureau voor Schimmelcultures, Netherlands; MUCL = = = Mycotheque de l'Universite Catholique, Louvain-la-Neuve, Belgium; RSA Rancho Santo Ana, USA; UCB University of California, Berkeley Microgarden, USA. b Type culture.

KC1, 2.5 mM MgCl2, 0.5 jjlM of each primer, 200 jjlM gene (18S rDNA) was amplified using primers NS1 of each dNTP, and 2.5 units of AmpliTaq? DNA poly? and NS6 (White et al., 1990). Thermal cycling pa? merase (Roche Molecular Systems, Inc, Branchburg, rameters for amplification consisted of one initial NJ). Each reaction mixture was topped off with a thin cycle with denaturation at 95 C for 5 min, annealing layer of mineral oil and amplified using a Perkin-El- at 54 C for 30 s, and extension at 72 C for 45 s. This mer DNA Thermal Cycler 480 (Norwalk, CT). cycle was followed by 38 cycles with denaturation at A segment of the small subunit ribosomal RNA 95 C for 30 s, annealing at 54 C for 30 s, and exten-

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Glenn et al.: Molecular Phylogeny of Acremonium 373

sion at 72 C for 45 s. (plus 4 see addition to exten? (1992), Bruns et al. (1992), and Spatafora and sion segment per cycle). A final cycle was performed Blackwell (1993). with an extension segment of 72 C for 10 min. Phenetic analyses were also performed on the Amplified products were separated from unineor- aligned 18S rDNA sequences using MEGA v. 1.01 et Since MEGA isn't of porated nucleotides and primers using either mini- (Kumar al., 1993). capable columns (Wizard? PCR Preps, Promega Corp., Mad? evaluating binary code, the alignment gap was not ison, WI) or microconcentrators (Microcon? 100, included as an additional character in the data set. For and Amicon, Inc, Beverly, MA) following each manufac- neighborjoining analyses (Saitou Nei, 1987), the distance method turer's protocol. Purified samples were sequenced by gamma (Kimura 2-parameter) was used with deletion of all sites contain? the Molecular Genetics Facility of the University of complete or information. of Georgia (Athens, GA) using an Applied Biosystems ing gaps missing Bootstrapping the tree was with 500 automated sequencer (model 373A, version 1.2.1). neighbor-joining performed S. T. and C. Primers NS1, NS2, and NS3 (White et al, 1990) were replications. Again, cerevisiae, deformans, were used as taxa. used for sequencing the 18S rDNA. Primers NS1 and tropicalis outgroup NS2 provide complementary sequences, but the com? plementary sequence to NS3 was not determined. RESULTS The rDNA sequences were easily aligned by direct of the ssrRNA encom- examination. Segments gene (18S rDNA) passing 937 bp were analyzed. Alignment of sequenc? es was direct Data analysis.?Maximum parsimony analysis of the easily accomplished by examination due to the conserved nature. An of aligned sequences was conducted using PAUP v. 3.1.1 gene's alignment gap (Swofford, 1993) on a Macintosh Performa 6115CD. one base pair, common to the outgroup taxa and the was added to the data set as an ad? Alignment gaps were treated as missing data (GAP- Clavicipitaceae, ditional character. Of the 43 taxa included in this MODE=MISSING). However, one gap was included from 16 of these were as an additional character in the data matrix (ab? study (Table II), sequences obtained from either GenBank or other researchers. sence and presence of gap were coded as 0 and 1, The 27 were for this Our respectively). In total, 43 isolates were included in the remaining sequenced study. efforts focused on the form analysis (Table II). As a result of the large data set, sequencing primarily ge? nus Acremonium and the ascomycete only heuristic searches were performed with the fol? family Clavicip? itaceae. Aligned sequences are available from the au? lowing options in effect: tree-bisection-reconnection thors upon request. (TBR) swapping algorithm, collapsing zero length A maximum parsimony analysis with all taxa in? branches, and saving all minimal length trees (MUL- cluded yielded eight equally parsimonious trees of PARS). Ten replications with random addition of taxa 656 steps. For each of these trees, the consistency were performed for each heuristic search in order to index (Cl) was 0.585 with autapomorphies included, find any additional islands of minimum length trees the retention index (RI) was 0.712, and the rescaled (Maddison, 1991). To measure the relative support consistency index (RC) was 0.417. The strict consen? and stability of the resulting clades, bootstrap values sus of the eight trees is presented in Fig. 1. In order (Efron, 1982; Felsenstein, 1985) and decay indices to determine bootstrap and decay values for the (Bremer, 1988; Donoghue et al., 1992) were calcu? broad range of taxa included in this study, a smaller lated using PAUP v. 3.1.1. The computational inten? subset of selected taxa was chosen to represent the sity of these two values is directly dependent on the various clades, and the support values were comput- number of taxa and characters. these Executing ed on the smaller collection of taxa. The maximum computations on all ofthe taxa included was beyond parsimony strict consensus cladogram of this subset the capability of the computer. Therefore, selected is presented in Fig. 2 along with the bootstrap and taxa were excluded, and the values and bootstrap decay indices. This same subset of taxa was also an? indices were calculated on the smaller decay repre- alyzed using the neighbor-joining method, and the sentative selection. was Bootstrapping performed results are presented in FlG. 3. with 250 indices to 4 replications. Decay up steps Based on these results, Acremonium is a polyphy- than the most trees were de? longer parsimonious letic taxon having species associated with at least termined. cerevisiae E. C. Saccharomyces Hansen, three or four currently recognized ascomycete orders and Taphrina deformans (Berk.) Tul., Candida tro- (Fig. 1). First, A. furcatum F. 8c V. Moreau : W. Gams picalis (Castellani) Berkhout were used as outgroup is potentially associated with Microascus and Cerato? taxa based on the results of previous analyses at a cystis of the order Microascales. The bootstrap values broader taxonomic scale by Berbee and Taylor for this connection are 67% using maximum parsi-

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions 374 Mycologia

Candidatropicalis Saccharomycescerevisiae Acremoniumalabamense ? Chaetomiumglobosum Sordariales Cryphonectriaparasitica Diaporthephaseolorum Diaporthales Daldiniaconcentrica Xylariahypoxylon H Xylariales Acremoniumalternatum ? Emericellopsisminima # Emericellopsisterricola ? Acremoniumchrysogenum # Acremoniumkiliense ? Hypocreales Acremoniumstrictum ? L^-Hypocreales Acremoniummurorum ? Acremoniumrutilum ? Acremoniumcoenophialum ? Epichloeamarillians ? E Epichloefestucae ? Acremoniumuncinatum ? Atldnsonellahypoxylon # Balansiaaristidae Balansiasclerotica Myriogenosporaatramentosa A Clavicipitales Myriogenosporaatramentosa Eb Balansiahenningsiana Balansiastrangulans Balansiaobtecta Clavicepspurpurea Echinodothistuberiformis ? Cordycepscapitata Hypocreapatlida Hypomycespolyporinus Nectriacinnabarina Hypocreales Neocosmosporavasinfecta Nectriavilior ? Acremoniumfurcatum ? Ceratocystisfimbriata Microascales Microascustrigonosporus j Ascosphaeraapis Eurotiumrubrum 1Eurotiales Monascuspurpureus Taphrinaaeformans = = = Fig. 1. Strict consensus of eight cladograms (each of 656 steps, CI 0.585, RI 0.712, RC 0.417) resulting from maximum parsimony analysis of 18S rDNA sequences of all taxa included in the study. Orders of the Euascomycetes are indicated. Black dots indicate species having Acremonium-likeanamorphs.

mony (Fig. 2) and 83% with the neighbor joining Among the Acremonium species having affiliation analysis (Fig. 3). Second, A. alabamense is the known to the Hypocreales is the type species of the genus, anamorph of T. terrestrisof the Sordariales (see Ta? A. alternatum. It was weakly associated with other taxa ble I), and our results definitively show this ana? such as A. kiliense Griitz, A. strictum W. Gams, Emer? morph is strongly associated with Chaetomium from icellopsis minima Stolk, and E. terricola]. F. H. Beyma. the Sordariales. Both maximum parsimony and The parsimony clade of these five taxa had a boot? neighbor-joining analyses resulted in a 100% boot? strap of only 59% and a decay index of +1 (Fig. 2). strap value for the A. alabamense-Chaetomium clade The phylogenetic placement of the cleistothecial (Figs. 2 and 3, respectively). Third, Epichloe and re? Emericellopsis in the Hypocreales is supported here by lated asexual endophytes are well known members of the 18S rDNA sequences of both E. terricola and E. the order Clavicipitales (Fig. 1), and their anamorphs minima. are currently classified in Acremonium sect. Albolano? The monophyletic Clavicipitales appears to have sa (see Table I). Fourth, a large number of Acremon? been derived from within the Hypocreales (Figs. 1- ium species show affiliation to the Hypocreales (Fig. 3). Monophyly of the Clavicipitaceae, the only family 1). Six conidial fungi classified in Acremonium, two in the order Clavicipitales, was supported by neigh? Emericellopsis species which have Acremonium ana? bor-joining analysis with an 89% bootstrap value (Fig. morphs, and Nectria vilior Starback, which also has 3), and support from maximum parsimony analysis an Acremonium anamorph (see Table I), are all was only slightly weaker with a bootstrap of 85% and placed within the moderately well supported clade a decay index of +3 (Fig. 2). comprising the Hypocreales. Based on maximum As indicated in FiGS. 2 and 3, FJpichloeand its asex? parsimony analysis, the Hypocreales had a 76% boot? ual derivatives in Acremonium sect. Albolanosa formed strap value and a decay index of +2 (Fig. 2). MEGA's a well supported, monophyletic group with 84% and neighbor-joining analysis gave a bootstrap value of 96% bootstrap values, respectively. Atkinsonella, Bal? 87% for the comparable clade (Fig. 3). ansia, and Myriogenospora formed a sister group to

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Glenn et al.: Molecular Phylogeny of Acremonium 375

Acremonium alabamense 100 [?? >4 L? 86 Chaetomium globosum 5 10 +4 100 j" Cryphonectria parasitica 74 >4 I? Diaporthe phaseolorum +2 Daldinia concentrica

Xylaria hypoxylon

r- Acremonium aUernatum ?

59 100 r Acremonium kiliense +1 >4 i L- Acremonium strictum 52/+1- 1001 Emericellopsis minima ? >4 I Emericellopsis terricola ? 100 - Acremonium chrysogenum >4 rC Nectria cinnabarina

71/+3 84/+2r Acremonium coenophialum ? E M L Enir.hlne amnriWanx ? 55/+1 Epichloe amarilUans ? J] 76/+2 Atkinsonella hypoxylon ? e \ C ? Balansia henningsiana ? 85/+3 !? Claviceps purpurea e ? 99 Cordyceps capitata >4 97 100 f- Hypocrea pallida >4 >4 L Hypomycesj polyporinus

*? Nectria vilior ?

??????????? Acremoniumfurcatum ? 67 ? +1 98 Ceratocystis fimbriata >4 Microascus trigonosporus

100 Eurotium rubrum >4 Monascus purpureus

100 Candida tropicalis >4 ?? Saccharomyces cerevisiae

Taphrina deformans = = = Fig. 2. Strict consensus of two cladograms (each of 527 steps, Cl 0.639, RI 0.727, RC 0.465) resulting from is with maximum parsimony analysis of 18S rDNA sequences of a selected subset of taxa. Consensus cladogram depicted unit of measure. values proportional branch lengths. Scale bar indicates the number of nucleotide changes per Bootstrap below internodes. Black dots indicate greater than 50% are indicated above internodes. Decay indices are indicated species = = having AcremoniumAikeanamorphs. E grass endophyte; e grass epibiont.

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions 376 Mycologia

96 r Acremonium coenophialum ? E

rL Epichloe amarillians ? E 60 ? Atkinsonella hypoxylon ? e 58 Balansia henningsiana E 89 'Claviceps purpurea e L? 52 Cordyceps capitata

100 r~ Hypocrea pallida

Hypomyces polyporinus

Nectria cinnabarina

Nectria vilior ? 87 ?? Acremonium chrysogenum

Acremonium alternatum ?

f 100 r Acremonium kiliense i

99 ? Acremonium strictum ?

minima ? liooi1001 Emericellopsis Emericellopsis terricola ?

Acremoniumfurcatum ? 50 ri 83 97 Ceratocystis fimbriata ? Microascus trigonosporus

100 Acremonium alabamense

99 Chaetomium globosum

64 100 Cryphonectria parasitica t ? Diaporthe phaseolorum 100 ?Daldinia concentrica

'Xylaria hypoxylon 99 ? 100 Eurotium rubrum ?? Monascus purpureus

? Taphrina deformans

Candida tropicalis

-Saccharomyces cerevisiae

FiG. 3. Neighbor-joining phenogram resulting from analysis of 18S rDNA sequences of a selected subset of taxa. Bootstrap confidence measures greater than 50% are indicated at internodes. Symbols are the same as in Fig. 2.

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Glenn et al.: Molecular Phylogeny of Acremonium 377

the Epichloe clade (FiGS. 1), but little support was septo carente. Conidia oblonga, ellipsoidea vel cylindrica vel fusiformia vel lunata vel ame- available for resolving these relationships. uncinata, hyalina, levia, facientes In general, there was a high degree of congruence rospora. Fungi parasitae obligati graminum, aliqui ectostroma hymenio obtectum. Genus continens between the resulting phylogenies of the cladistic phialidum anamorphoses Clavicipitacearum. (Fig. 2) and phenetic (Fig. 3) analyses. Relationships TYPUS: Acremonium & W. within the differed in the coenophialum Morgan-Jones Hypocreales only place? Gams. ment of N. vilior and A. chrysogenum (Thirum. 8c Su- kapure) W. Gams. Placement of the Xylariales also Colonies white to yellowish, very to moderately differed between the two analyses. Bootstrap values slow growth rate, aerial hyphae often abundant, cot- for comparable clades in the two resulting phyloge? tony but usually not fasciculate; phialides solitary, nies were generally equivalent. However, three clades rarely verticillate, aculeate, orthotropic, arising from did show distinctly different values: the monophyletic aerial hyphae, bases frequently lacking a septum. Co? clade comprising the Sordariales and Diaporthales, nidia oblong, ellipsoidal, cylindrical, fusiform, lunate, the Hypocreales clade, and the clade containing A. uncinate, hyaline, smooth-walled, amerosporous. Ob? furcatum and the Microascales. These differences are ligate parasites of grasses, some forming exposed ec? probably inherent in the individual algorithms of the tostroma with palisade of phialides. Genus contain? synapomorphy-based cladistical analysis and the sim- ing anamorphs of the Clavicipitaceae. = ilarity-based phenetic analysis and how they each af? Etymology. Greek neos new + Typhodium, a ge? fect the areas of the trees where the degree of sup? nus erected by Link (1826) that was of uncertain ap- port for the topology is limited. plicability to the anamorph of Epichloe.

DISCUSSION Neotyphodium coenophialum (Morgan-Jones & W. Gams) Glenn, Bacon 8c Hanlin comb. nov. New taxon and combinations.?Based on the 18S rDNA sequence analyses presented here, the present Basionym. Acremonium coenophialum Morgan-Jones 8c W. classification of the anamorphs of Epichloe and relat? Gams, Mycotaxon 15: 311. 1982. ed mutualists in Acremonium is untenable. Acremon? No known teleomorph. ium, as typified by A. alternatum, appears to be re? stricted to the family Hypocreaceae. Epichloe and re? Neotyphodium typhinum (Morgan-Jones 8c W. Gams) lated genera of the Clavicipitaceae form a well cir- Glenn, Bacon 8c Hanlin comb. nov. cumscribed, monophyletic family that appears to be derived from within the Hypocreales. The monophy? Basionym. Acremonium typhinum Morgan-Jones 8c W. ly of this family is supported by the unique mor? Gams, Mycotaxon 15: 311. 1982. phology, ecology, and obligate parasitism of the Clav? Teleomorphs are species of Epichloe. icipitaceae. The anamorphs of this family are unusual in that they are associated with stromatic tissue Neotyphodium lolii (Latch, Christensen 8c Samuels) formed on a host (Diehl, 1950; Luttrell, 1980; Rykard Glenn, Bacon 8c Hanlin comb. nov. et al, 1984; White and Morgan-Jones, 1987; Leucht? mann and Clay, 1988; Morgan-Jones and White, Basionym. Acremonium lolii Latch, Christensen & Sa? 1989). Therefore, we are proposing the erection of muels [as Acremonium loliae], Mycotaxon 20: 535. 1984. a new genus to accommodate anamorphs of clavicip- No known teleomorph. itaceous fungi that form a palisade of simple phialidic conidiogenous cells over the surface of a stroma and Neotyphodium chisosum (J. F. White & Morgan- in culture produce simple, strictly aculeate phialides Jones) Glenn, Bacon 8c Hanlin comb. nov. often lacking a basal septum and usually arising as lateral branches from aerial As thus hyphae. defined, Basionym. Acremonium chisosumJ. F. White 8c Morgan- this genus should include the conidial fungi current- Jones, Mycotaxon 28: 179. 1987. ly classified in Acremonium sect. Albolanosa. No known teleomorph.

Neotyphodium Glenn, Bacon 8c Hanlin gen. nov. Neotyphodium starrii (J. F. White 8c Morgan-Jones) Glenn, Bacon 8c Hanlin comb. nov. Coloniae albae vel flavidae, lente vel modice crescentes, hyphae aeriae saepe abundantes, byssaceae, typice non fas- Basionym. Acremoniumstarrii]. F. White & Morgan-Jones, ciculatae; phialides solitariae, raro verticillatae, aculeatae, Mycotaxon 30: 87. 1987. orthotropicae, exoriundae ex hyphis aeriis, basi plerumque No known teleomorph.

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions 378 Mycologia

Neotyphodium huerfanum (J. F. White, G. T. Cole 8c The anamorph of Echinodothis tuberiformis (Berk. Morgan-Jones) Glenn, Bacon 8c Hanlin comb. nov. 8c Ravenel) G. F. Atk. presents a similar problem as found with Atkinsonella. The anamorph of this epi- Basionym. Acremonium huerfanumJ. F. White, G. T. Cole biont is similar to in that a of 8c Morgan-Jones, Mycologia 79: 148. 1987. Neotyphodium palisade cells are over the No known teleomorph. phialidic conidiogenous produced surface of a host-associated stroma, but the conidia are different in being didymosporous (White, uncinatum (W. Gams, Petrini & D. Neotyphodium 1993b). White (1993b) suggested that Echinodothis Schmidt) Glenn, Bacon & Hanlin comb. nov. and Epichloe were closely related and might deserve Basionym. Acremoniumuncinatum W. Gams, Petrini 8c D. classification together in a separate tribe. While lack? Schmidt, Mycotaxon 37: 67. 1990. ing statistical support, cladistical analysis suggested No known teleomorph. that Echinodothis may be intermediate between Cor- dyceps and Claviceps and is not closely related to Epi? chloe (FlG. 1). As with Atkinsonella, the holomorphic Neotyphodium chilense (Morgan-Jones, J. F. White 8c of E. to its and te? Piont.) Glenn, Bacon 8c Hanlin comb. nov. application tuberiformis anamorph leomorph means there is no need to establish an ana? Acremonium chilense E White Basionym. Morgan-Jones,J. morphic binomial. 8c Piont., Mycotaxon 39: 441. 1990. Clav? No known teleomorph. The Clavicipitales-Hypocreales relationship.?The icipitales is a morphologically well circumscribed or? While not included in our analyses, N. starrii and der of mostly parasitic ascomycetes possessing mor? N. lolii have been shown to be closely related to N. phologically distinctive characters such as deliques- coenophialum based upon maximum parsimony anal? cent lateral paraphyses, cylindrical asci with a thick? ysis of the two internal transcribed spacer (ITS) ened apical cap, and filiform ascospores that are regions of rDNA (Schardl et al., 1991). Unfortunate- often septate and may disarticulate into part-spores ly, N. chilense, N. chisosum, and N huerfanum were (Rogerson, 1970). However, the ordinal affinity of also unavailable to us for examination, and these the clavicipitaceous genera has been evaluated and three have yet to be characterized by DNA-based phy? interpreted differently by several researchers [see Ro? logenetic analyses. We do, however, propose the gerson (1970) for a thorough historical review of the transfer of these species into Neotyphodium based taxonomic literature]. upon the examinations and personal communica- In recent years, the ordinal relationships among tions of James F. White, Jr. and Walter Gams. hypocreaceous and clavicipitaceous genera has again The conidial stromata of species of Epichloe are eas? come into question as a result of phylogenetic eval- ily and logically addressed by the teleomorphic no- uations using molecular data. Maximum parsimony men even though anamorphic binomials and varie? analyses using the nuclear encoded 18S rDNA have ties have been erected for the various species (Mor? indicated a monophyletic relationship exists between gan-Jones and Gams, 1982; White, 1992). However, genera of the Clavicipitales and the Hypocreales, sug? the abundance of asexual endophytes that rarely or gesting that all taxa may be classifiable under the sin? never produce teleomorphic stromata [clonal type-III gle order Hypocreales with the monophyletic Clavi? endophytes of White (1988)] means that a dual sys? cipitaceae being a separate family from the paraphy- tem of nomenclature must be maintained whereby letic Hypocreaceae (Spatafora and Blackwell, 1993). the endophytes are classified in Neotyphodium type-III Our results, which also utilized 18S rDNA, show the and the sexually reproducing species are classified in same general relationship (Figs. 1-3). Additionally, Epichloe. gene phylogenies based on both the nuclear encoded As defined, Neotyphodium is the most appropriate large subunit ribosomal DNA (28S rDNA) and nucle? form genus for the typhodial stage of Atkinsonella ar encoded orotidine-5'-monophosphate decarboxyl- (Rykard et al., 1984). However, the applicability of ase also support the derivation of the monophyletic the nomen Atkinsonella to all stages of its develop? Clavicipitaceae from within the Hypocreales (Rehner ment is empirical and conveys a holomorphic view of and Samuels, 1994b, 1995). All these gene phyloge? this fungus. Exclusively asexual species are not nies support the treatment by Kreisel (1969) with a known. Therefore, since no binomial is currently de? single order incorporating the families Hypocreaceae fined for the typhodial stage of Atkinsonella, we do {sensu Rogerson) and Clavicipitaceae. not see the necessity of defining one under Neoty? phodium. The binomial Ephelis borealis Ellis 8c Everh. The expanding Hypocreales.?Just as molecular system? has been established for the sympodially proliferating atics has aided in clarifying some of the taxonomic synanamorph of Atkinsonella (Diehl, 1950). confusion concerning the ordinal relationships

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Glenn et al.: Molecular Phylogeny of Acremonium 379

among hypocreaceous and clavicipitaceous genera, in chains or heads from simple phialides, but A. al? the same approaches are allowing systematists to pro- abamense is different in that it grows much more rap? pose phylogenetically informative holomorphic con? idly than other species and is thermophilic (Morgan- nections between anamorphs and teleomorphs Jones, 1974; Morgan-Jones and Gams, 1982). Before (Schardl et al., 1991; Rehner and Samuels, 1994a). any formal generic reassignment of this anamorph is Such phylogenetic analyses are also generating pro- proposed, a detailed study is needed to determine if posals of novel intraordinal teleomorphic relation? other morphological characteristics exist which dis? ships which are expanding the delimitation of the tinguish it from other species of Acremonium. How? Hypocreales (Rehner and Samuels, 1994a, 1995; re? ever, because of the abundance of morphological sults reported herein). and biological information that is available concern- Our approach to the form genus Acremonium was ing the grass endophytes, the generic placement of similar to that used to investigate other polyphyletic the anamorphs of Epichloe and related grass symbi- form genera (Rehner and Samuels, 1994a, 1995). onts is now ready to be addressed. A discussion of Our results (Figs. 1-3) suggest Acremonium is indeed pertinent biological and taxonomic information is polyphyletic as currently circumscribed. Affiliation of presented below. the type species, A. alternatum, and other common Numerous systems of higher categories for the As? species such as A. kiliense, A. strictum, A. chrysogen- comycotina have been suggested and are summarized um, A. murorum (Corda) W. Gams, and A. rutilum in Hawksworth et al. (1983). Common to a few of W. Gams to the Hypocreaceae suggests that Acremon? these systems are the categorical divisions Plectomy? ium should be restricted to anamorphs of only this cetes and Pyrenomycetes at varying hierarchical lev? family. Since many Acremonium species, including A. els. These two divisions are based on characters of alternatum, do not produce ascomata in vitro, apply- the ascomata. Plectomycetes includes those fungi ing the taxonomic criterion of relationship to the Hy? having typically globose, nonostiolate cleistothecia pocreaceae is currently not feasible based on mor? that produce asci at varying levels throughout the phological characteristics. In order to strictly reserve centrum (Fennell, 1973; Benny and Kimbrough, Acremonium as anamorphs of the Hypocreaceae, 1980), and Pyrenomycetes includes those fungi hav? DNA sequence analyses would be needed for each of ing flask-shaped, mostly ostiolate perithecia that pro? the species of Acremonium that do not have a known duce asci from a single basal hymenium (Muller and teleomorph. While perhaps justified, such an im- von Arx, 1973). Suggestions have been made that the mense task is not yet practicable. However, it is pos? plectomycetes are a heterogeneous assemblage of sible that most of the orphaned Acremonium species morphologically similar fungi resulting from conver- may be phylogenetically affiliated with the Hypocre? gent evolution, and many plectomycetous genera are ales, as suggested by Fig. 1 which shows six orphaned suggested to be derived from within the Pyrenomy? species of Acremonium from the sections Acremonium, cetes (Malloch, 1981; Benny and Kimbrough, 1980). Gliomastix, and Nectrioidea derived from within the Recent molecular phylogenetic studies have con? Hypocreales. Of those sampled, only one orphaned firmed that such a heterogeneity indeed exists. An species, A. furcatum of sect. Nectrioidea, is not derived initial maximum parsimony analysis indicated that from within the Hypocreales. there was a natural sister-group relationship between At present the best we can do to eliminate some selected taxa corresponding well to the traditional of the heterogeneity within Acremonium is to remove ascomycete classes Plectomycetes and Pyrenomycetes those species which are known to be associated with (Berbee and Taylor, 1992). Such traditional plecto? teleomorphs that are not within the Hypocreaceae mycetous taxa as Ascosphaera, Ajellomyces, Monascus, (Gams, 1995). Though not examined, some, if not Talaromyces, and Thermoascus were found to form a all, species of sect. Lichenoidea would be expected to monophyletic sister group to a monophyletic group show affiliation to the Hypocreaceae since four of the of pyrenomycetes including Ophiostoma, Chaetom? nine species have hypocreaceous teleomorphs. ium, and Neurospora. Among the Acremonium species that were examined While a true higher level division appears to nat- by us, the grass endophytes (sect. Albolanosa) and A. urally exist between pyrenomycetes and some plec? alabamense (sect. Chaetomioides) have teleomorphs tomycetes, molecular analyses are identifying other that are in the Clavicipitaceae and Chaetomiaceae "plectomycetes" that appear to be derived from with? (Sordariales), respectively. Therefore, the generic in the pyrenomycetes. Roumegueriella and Heleococ- placement of these anamorphs is in need of reeval- cum are cleistothecioid genera that have been previ? uation. The morphological characteristics of A. ala? ously placed in the Hypocreales (Rogerson, 1970; bamense are very similar to those of other 'true' Acre? Malloch, 1981). This placement was supported by monium species in that it produces guttulate conidia analysis of 28S rDNA sequences (Rehner and Sa-

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions 380 Mycologia muels, 1994a, 1995). Mycoarachis (Rehner and Sa? considered, more detailed morphological studies are muels, 1994a) and Emericellopsis (Figs. 1-3), so far needed to determine if other tangible characters ex- classified in the cleistothecioid family Pseudeuroti- ist that could be used for its generic distinction from aceae, also show affinity to the Hypocreales. The Acremonium. A comparative analysis of conidium on? Pseudeurotiaceae was suggested by Malloch (1981) to togeny of A. alternatum and A. alabamense may pro- be related to the Diaporthales on the basis of cen? vide additional needed characters. trum development, but some members of this family The potential taxonomic problems posed by A. fur- now appear to be phylogenetically related to the Hy? catum are complex. Since its affiliation to an asco? pocreales. This relationship is directly relevant to this mycete order is currently vague, it must be retained study in that Mycoarachis, Emericellopsis, and other in the genus Acremonium. If further molecular or members of the Pseudeurotiaceae possess Acremon? morphological analyses should indicate a phyloge? ium anamorphs (Malloch and Cain, 1970). Heleococ- netic affiliation to the Hypocreaceae, then its current cum also produces an Acremonium anamorph (Uda- status would be maintained. If additional molecular gawa et al., 1995). analyses should confirm an affiliation to the Microas- While ascomatal characteristics appear to vary dra- cales as suggested by our results, or if its teleomorph matically between hypocreaceous genera, the posses- is discovered and classified within the Microascales, a sion of an Acremonium anamorph by many of these more phylogenetically appropriate classification for teleomorphs appears to be a phylogenetically infor? A. furcatum may then be justified. mative character supporting their link to the Hypo? The gene phylogenies presented here indicate that creaceae. It is interesting to note that Wu and Kim? the Clavicipitaceae is a distinctive, monophyletic fam? brough (1990) found much similarity in the ascogon- ily derived from within the Hypocreales. Other phy? ial and ascogenous systems of Emericellopsis, Asco- logenetic analyses have also indicated this same re? sphaera, and Monascus even though Emericellopsis is lationship (Spatafora and Blackwell, 1993; Rehner phylogenetically distinct from the more closely relat? and Samuels, 1995). The monophyletic nature of the ed Ascosphaera and Monascus (Fig. 1). Such similarity Clavicipitaceae is substantiated by its unique ecology in ascomatal development despite the phylogenetic and morphology. Mating compatibility studies have separation implies that these ascomatal characteris? shown that graminicolous species of Balansia, Atkin? are tics may be subject to convergent evolution. Also, the sonella, Echinodothis, and Epichloe heterothallic, the transfer of conidia from anamorphs of Ascosphaera (= Chrysosporium) and requiring (= spermatia) a stroma of one to a stroma of the Monascus (= Basipetospora) are distinctly different mating type op- and from the Acremonium anamorph of Emericellopsis. posite mating type (White Bultman, 1987; Leuchtmann and White Some of the other Acremonium-producing genera of Clay, 1989; White, 1993a, b; et of the Pseudeurotiaceae that may prove to be hypocrea? al., 1995). Mating compatibility Myriogenospora has not been examined. Conidia of are in- ceous once molecular analyses are performed in? Claviceps fective and do not function as clude Nigrosabulum, Hapsidospora, and Leucosphaeri- spermatia (Luttrell, Diehl the of na (Malloch and Cain, 1970; Malloch, 1989). The in? 1980). (1950) emphasized importance conidial fructification for taxonomic and clusion of cleistothecioid forms within the Hypocre? systematic evaluations of these the form ales is creating a need for reevaluation and fungi. Sphacelia, genus for the of is characterized a refinement of the distinguishing characters defining anamorph Claviceps, by this order. palisade of doliiform to lageniform phialides pro? duced along the convoluted surface of the develop? Taxonomic considerations.?The fact that the teleo? ing sclerotium, and this palisade bears a wet mass of morph of A. alabamense is T terrestrisof the family amerosporous conidia (Diehl, 1950; Luttrell, 1980). Chaetomiaceae is in opposition to the criterion of In contrast, the "typhodial" conidial fructifications of Acremonium being restricted to anamorphs of the Hy? Epichloe, herein classified in the form genus Neoty? pocreaceae. The thermophilic nature of A. alaba? phodium, produce dry masses of amerosporous co? mense, its rapid growth rate, and its truncated, dac- nidia from a continuous, nonconvoluted palisade of ryoid conidia were all considered by Morgan-Jones narrowly aculeate phialides over the surface of the and Gams (1982) to be characters distinctive enough host-associated stroma (Diehl, 1950; White and Mor? to warrant the creation of sect. Chaetomioides within gan-Jones, 1987). The anamorphs of Balansia and Acremonium. While A. alabamense is typically ortho- Myriogenospora are classified in the form genus Ephel? phialidic like most species of Acremonium, perhaps its is which is characterized by holoblastic, scolecospo- unique morphological and biological characters are rous conidia produced successively from sympodially more distinctive at the generic versus sectional level. proliferating conidiophores in apothecioid cavities in Before any formal reclassification of A. alabamense is host-associated stromata or pseudosclerotia (Diehl,

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Glenn et al.: Molecular Phylogeny of Acremonium 381

1950; Rykard et al., 1984; White et al., 1995). Atkin? the anamorph of Epichloe. As an informal "conven- sonella is intermediate in that it initially has a Neoty- tion of convenience," Diehl (1950) applied Typhod- phodium-like anamorph which is followed by an ium as the form genus for Epichloe conidial fructifi- Ephelis anamorph (Rykard et al., 1984). cations. In following Diehl's concept, we have pro? The close evolutionary connection between asex? posed the form genus Neotyphodium for the ana? ual grass endophytes and the sexually reproducing morphs of Epichloe and its related asexual grass species of Epichloe has been examined from an eco? endophytes. logical perspective (White, 1988), but a greater The establishment of Neotyphodium helps to alle- amount of attention has been given to molecular viate some of the heterogeneity of Acremonium while evaluations of relationships. The isozyme and molec? emphasizing the unique, monophyletic nature of the ular analyses previously performed on the asexual en? grass endophytes. The employment of molecular data dophytes have mainly dealt with the variation and re? to substantiate this reclassification indicates the utility lationships that exist among some of the species. of DNA-based phylogenetic techniques in evaluating Leuchtmann and Clay (1990) found that isozyme the taxonomic affiliations of morphologically simple profiles based on ten enzymes are virtually uniform taxa such as Acremonium. However, limitations do ex? throughout most isolates of N. coenophialum (= A. ist. For example, the basic morphology of A. alaba? coenophialum). They also found that no distinction mense is essentially the same as that of A. alternatum, could be made between the Epichloe endophytes and but the two are affiliated with different ascomycete the asexual endophytes, thereby suggesting that they orders. More detailed morphological studies are now jointly comprise a monophyletic group. Schardl et al. needed to determine if informative characters exist (1991) performed maximum parsimony analyses on which could be used for their generic distinction. Re? sequences of the internal transcribed spacer (ITS) sults of phylogenetic analyses are useful as frame- regions of rDNA and showed that the asexual endo? works for further, more focused morphological and phytes formed a monophyletic group with Epichloe molecular comparisons. Such comparisons may help and that asexual species apparently arose from Epi? to alleviate some of the remaining heterogeneity of chloe on multiple occasions. Additionally, they found Acremonium. that sequence comparisons do not support some of the based classifications of morphologically species ACKNOWLEDGMENTS asexual endophytes. Based on maximum parsimony analyses of ITS sequences, An et al. (1992) reported We are grateful to James F. White, Jr. for supplying us with several cultures and for his comments the that at least two distinct evolutionary origins of Neo- concerning pro? and to and Denise Silva typhodium-endophytes from Epichloe had occurred in ject manuscript, Joseph Spatafora for to us DNA to their the single host species Festuca arizonica Vasey. The providing sequences prior deposition into GenBank, and to Walter Gams and an unidentified studies by Leuchtmann and Clay (1990) and Schardl reviewer for offering suggestions to improve the manu? et al. (1991) are supportive ofthe existence of several script. different species of Epichloe (see also White, 1993a; White, 1994a; Leuchtmann et al., 1994). The monophyletic nature of the Clavicipitaceae, LITERATURE cited the separation of Epichloe from A. alternatum, the ob? An, Z.-q., J.-S. Liu, M. R. Siegel, G. Bunge, and C. L. Schardl. of and its asexual deriva- ligate parasitism Epichloe 1992. Diversity and origins of endophytic fungal sym- and the formation of a tives, sporodochium-like pal? bionts of the North American grass Festuca arizonica. isade of on an external stroma all indicate phialides Theor. Appl. Genet. 85: 366-371. the distinct nature of Epichloe and support the re- Bacon, C. W., J. K. Porter,J. D. Robbins, and E. S. Luttrell. classification of its anamorphs. The doubtful identity 1977. Epichloe typhina from toxic tall fescue grasses. of Sphacelia as a form genus for these anamorphs Appl. Environm. Microbiol. 34: 576-581. G. W. 1980. A ofthe (Diehl, 1950; Morgan-Jones and Gams, 1982) and the Benny, L., andj. Kimbrough. synopsis orders and families of with to lack of any other valid or appropriate form genus Plectomycetes keys gen? era. Mycotaxon 12: 1-91. means a new genus is needed to accommodate these Berbee, M. L., andj. W. Taylor. 1992. Two ascomycete class- fungi. Typhodium was erected by Link (1826), but his es based on fruiting-body characters and ribosomal meager description is vague and confusing so there DNA sequence. Molec. Biol. Evol. 9: 278-284. is as to whether he was the uncertainty describing Bremer, K. 1988. The limits of amino acid sequence data of or the Diehl anamorph Epichloe teleomorph. in angiosperm phylogenetic reconstruction. Evolution (1950) felt that the conidial fructifications of Epichloe 42: 795-803. were significantly different from those of Claviceps Bruns, T. D., R. Vilgalys, S. M. Barns, D. Gonzales, D. S. and chose not to apply the form genus Sphacelia to Hibbett, D. J. Lane, L. Simon, S. Stickel, T. M. Szaro,

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions 382 Mycologia

W. G. Weisburg, and M. L. Sogin. 1992. Evolutionary symbiotic with fine fescue grasses. Mycologia 86: 802- relationships within the fungi: Analyses of nuclear 812. small subunit rRNA sequences. Molec. Phylogenet.Evol. Link, H. F. 1826. Entwurf eines phytologischen Pflanzen- 1: 231-241. systems nebst einer Anordnung der Kryptophyten. Clay, K., and I. C. Frentz. 1993. Balansia pilulaeformis,an Akad. Wiss. Berlin, Phys. Kl. 1824: 145-194. epiphytic species. Mycologia 85: 527-534. Lowen, R. 1995. Acremonium section Lichenoidea section Diehl, W. W. 1950. Balansia and the Balansiae in America. nov. and Pronectria oligospora species nov. Mycotaxon Agriculture Monograph No. 4, USDA, Washington, 53: 81-95. D.C. 78 pp. Luttrell, E. S. 1980. Host-parasite relationships and devel? Donoghue, M. J., R. G. Olmstead, J. F. Smith, and J. D. opment of the ergot sclerotium in Claviceps purpurea. Palmer. 1992. Phylogenetic relationships of Dipsacales Canad. J. Bot. 58: 942-958. of based on rbcL sequences. Ann. Missouri Bot. Gard. 79: -, and C. W. Bacon. 1977. Classification Myrioge? 333-345. nospora in the Clavicipitaceae. Canad. f. Bot. 55: 2090- 2097. Efron, B. 1982. The Jackknife,the Bootstrap and Other Re- D. R. 1991. The and of sampling Plans. CBMS-NSF Regional Conference Series Maddison, discovery importance islands of tress. Zool. in Applied Mathematics, Monograph 38, SIAM, Phila- multiple most-parsimonious Syst. 40: 315-328. delphia. 92 pp. D. 1981. The centrum. 73-91. Felsenstein, J. 1985. Confidence limits on phylogenies: An Malloch, plectomycete Pp. In: the Luttrellian Ed., approach using the bootstrap. Evolution 39: 783-791. AscomyceteSystematics, Concept. D. R. New York. Fennell, D. I. 1973. Plectomycetes; Eurotiales. Pp. 45-68. Reynolds. Springer-Verlag, -. 1989. An undescribed of In: The Fungi, an Advanced Treatise. Vol. 4A. Eds., G. species Leucosphaerina. Stud. 31: 107-111. C. Ainsworth, F. K. Sparrow, and A. S. Sussman. Aca? Mycol. and R. F. Cain. 1970. Five new in the new demic Press, New York. -, genera familyPseudeurotiaceae. Canad. J. Bot. 48: 1815-1825. Gams, W. 1971. Cephalosporium-artzg^ Schimmelpilze(Hy? Morgan-Jones, G. 1974. Notes on Hyphomycetes. V. A new phomycetes).Gustav Fischer Verlag, Stuttgart. 262 pp. thermophilic species of Acremonium. Canad. J. Bot. 52: -. 1975. Cephalosporium-MkeHyphomycetes: Some 429-431. tropical species. Trans. Brit. Mycol. Soc. 64: 389-404. -, and W. Gams. 1982. Notes on Hyphomycetes. XLI. -. 1995. How natural should anamorph genera be? An endophyte of Festuca arundinacea and the ana? Canad. J. Bot. 73: S747-S753. morph of Epichloe typhina,new taxa in one of two new -, and A. Van Zaayan. 1982. Contribution to the tax? sections of Acremonium. Mycotaxon 15: 311-318. onomy and pathogenicity of fungicolous Verticillium -, R. A. Phelps, andj. F. White, Jr. 1992. Systematic species. I. Taxonomy Neth. J. Pl. Path. 88: 57-78. and biological studies in the Balansieae and related an? Hawksworth, D. L., B. C. Sutton, and G. C. Ainsworth. 1983. amorphs. I. Prologue. Mycotaxon 43: 401-415. Ainsworth and Bisby s Dictionary of the Fungi. 7th ed. -, and J. E White, jr. 1989. Concerning Atkinsonella CAB International, Kew. 445 pp. texensis,a pathogen of the grass Stipa leucotricha: De? Kreisel, H. 1969. Grundziige eines natiirlichen Systemsder velopmental morphology and mating system. Mycotax? Pilze. Lehre. 245 J. Cramer, pp. on 35: 455-467. Kumar, S., K. Tamura, and M. Nei. 1993. MEGA: Molecular Miiller, E., and J. A. von Arx. 1973. Pyrenomycetes: Me- version 1.01. The Penn? evolutionary genetics analysis, liolales, Coronophorales, Sphaeriales. Pp. 87-132. In: State Park, sylvania University,University Pennsylvania. The Fungi, an Advanced Treatise. Vol. 4A. Eds., G. C. Latch, G. C. M., M. Christensen, and G. Samuels. 1984. J. J. Ainsworth, F. K. Sparrow, and A. S. Sussman. Academic Five of Lolium and Festuca in New Zealand. endophytes Press, New York. 20: 535-550. Mycotaxon Mullis, K. B., and F. A. Fallona. 1987. Specific synthesis of 1990. Isolation of DNA from Lee, S., andj. Taylor. fungal DNA in vitrovia a polymerase-catalyzed chain reaction. and 282-287. In: PCR Pro- mycelia single spores. Pp. Methods Enzymol 155: 335-350. tocols: A Guide to Methods and M. A. Applications. Eds., Rehner, S. A., and G. J. Samuels. 1994a. Taxonomy and D. H. and T. White. Ac? Inns, Gelfand, J. J. Sninsky, J. phylogeny of Gliocladium analysed from nuclear large New York. ademic Press, subunit ribosomal DNA sequences. Mycol Res. 98: 625- Leuchtmann, A., and K. Clay. 1988. Atkinsonella hypoxylon 634. and Balansia cyperi,epiphytic members of the Balan? -, and -. 1994b. Molecular systematics of the siae. Mycologia 80: 192-199. Hypocreales: Teleomorph gene phylogenies and the -, and -. 1989. Morphological, cultural and status of orphan anamorphs. Fifth International My? mating studies on Atkinsonella, including A. texensis. cological Congress (Abstracts). P. 178. Mycologia 81: 692-701. -, and-. 1995. Molecular systematicsofthe Hy? -, and -. 1990. Isozyme variation in the Acre? pocreales: A teleomorph gene phylogeny and the status monium/Epichloe fungal endophyte complex. Phytopa? of their anamorphs. Canad. J. Bot. 73: S816-S823. thology80: 1133-1139. Rogerson, C. T. 1970. The hypocrealean fungi (Ascomy? -, C. L. Schardl, and M. R. Siegel. 1994. Sexual com? cetes, Hypocreales). Mycologia 62: 865-910. patibility and taxonomy of a new species of Epichloe Rykard, D. M., E. S. Luttrell, and C. W. Bacon. 1982. De-

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions Glenn et al.: Molecular Phylogeny of Acremonium 383

velopment of the conidial state of Myriogenosporaatra? netic implications of generic concepts in fungal tax? mentosa. Mycologia 74: 648-654. onomy: The impact of molecular systematic studies. -, and-. 1984. Conidiogenesis and co? Mycol. Helv. 6: 73-90. nidiomata in the Clavicipitoideae. Mycologia 76: 1095- White, J. F.,Jr. 1988. Endophyte-host associations in forage 1103. grasses. XI. A proposal concerning origin and evolu? Saccardo, R A. 1881. Fungi Veneti novi vel critici v. Mycol- tion. Mycologia 80: 442-446. ogiae Venetae addendi. Michelia 2: 241-301. -. 1992. Endophyte-host associations in grasses. XVII. Saiki, R. K, D. H. Gelfand, S. Stoffel, S. J. Scharf, R. Hig- Ecological and physiological features characterizing uchi, G. T. Horn, K. B. Mullis, and H. A. Erlich. 1988. Epichloe typhinaand some anamorphic varieties in En- 84: 431-441. Primer-directed enzymatic amplification of DNA with gland. Mycologia -. 1993a. associations in XIX. thermostable DNA polymerase. Science 239: 487-491. Endophyte-host grasses. A of some of Saitou, N., and M. Nei. 1987. The neighbor-joining meth? systematicstudy sympatricspecies Epichloe in 85: 444-455. od: A new method for reconstructing phylogenetic England. Mycologia trees. Molec. Biol. Evol. 4: 406-425. -. 1993b. Structure and mating systemof the graminicolous Echinodothis Sampson, K. 1933. The systemic infection of grasses by Epi? fungal epibiont tuberiformis(Clav? Bot. 80: 1465-1471. chloe typhina (Pers.) Tul. Trans. Brit. Mycol. Soc. 18: icipitales). Amer.f. 30-47. -. 1994a. Endophyte-host associations in grasses. XX. Structural and studies of amaril- Samuels, G. J. 1973. The myxomyceticolous species of Nec? reproductive Epichloe lians sp. nov. and comparisons to E. typhina. Mycologia tria. Mycologia 65: 401-420. 86: 571-580. -. 1976a. A revision of the fungi formerly classified -. 1994b. Taxonomic relationships among the mem? as Nectria subgenus Hyphonectria. Mem. New YorkBot. bers of the Balansieae 3-20. In: Gard. 26: 1-126. (Clavicipitales). Pp. Biotechnologyof EndophyticFungi of Grasses. Eds. C. W. -. 1976b. Perfect states of Acremonium: The genera Bacon andj. F. White, Jr. CRC Press, Boca Raton. Nectria, Actiniopsis, Ijuhya, Neohenningsia, Ophiodic- -, and T. L. Bultman. 1987. Endophyte-host associa? tyon,and Peristomialis.New Zealand J. Bot. 14: 231-260. tions in forage grasses. VIII. Heterothallism in Epichloe -. 1988. Fungicolous, lichenicolous, and myxomyce? typhina. Amer.J. Bot. 74: 1716-1721. ticolous species of Hypocreopsis,Nectriopsis, Nectria, Per? -, and A. E. Glenn. 1994. A study of two fungal epi- istomialis, and Trichonectria.Mem. New YorkBot. Gard. bionts of grasses: Structural features, host 48: 1-78. relationships, and classification in the genus Myriogenospora (Clavi? -, A. Y. Rossman, R. Lowen, and C. T. Rogerson. 1991. cipitales). Amer.]. Bot. 81: 216-223. A synopsis of Nectria subgen. Dialonectria. Mycol. Pap. -, and G. Morgan-Jones. 1987. Endophyte-host asso? 164: 1-47. ciations in forage grasses. IX. Concerning Acremonium Schardl, C. L., J.-S. Liu, J. F. White, Jr.,R. A. Finkel, Z. An, typhinum,the anamorph of Epichloe typhina.Mycotaxon 1991. rela? and M. R. Siegel. Molecular phylogenetic 29: 489-500. of and tionships nonpathogenic grass mycosymbionts -, L. T. Sharp, T. I. Martin, and A. E. Glenn. 1995. Pl. Evol. 178: 27- clavicipitaceous plant pathogens. Syst. Endophyte-host associations in grasses. XXI. Studies on 41. the structure and development of Balansia obtecta.My? W., and M. Blackwell. 1993. Molecular Spatafora, J. system? cologia 87: 172-181. atics of unitunicate The Clav- perithecial ascomycetes: White, T. J., T. Bruns, S. Lee, andj. W. Taylor. 1990. Am? connection. 85: 912- icipitales-Hypocreales Mycologia plification and direct sequencing of fungal ribosomal 922. RNA genes for phylogenetics. Pp. 315-322. In: PCR D. L. 1993. PAUP: Swrofford, Phylogeneticanalysis using par? Protocols: A Guide to Methods and Applications. Eds., M. version 3.1.1. Illinois Natural simony, History Survey, A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White. Champaign, Illinois. Academic Press, New York. Udagawa, S.-L, S. Uchiyama, and S. Kamiya. 1995. Two new Wu, C.-g., andj. W. Kimbrough. 1990. Ultrastructural stud? species of Heleococcum with Acremonium anamorphs. ies on the cleistothecial development of Emericellopsis Mycoscience36: 37-43. microspora(Eurotiales, Ascomycetes). Canad. J. Bot. 68: Vilgalys, R, J. Hopple, and D. S. Hibbett. 1993. Phyloge- 1877-1888.

This content downloaded from 142.51.1.212 on Wed, 4 Sep 2013 15:25:54 PM All use subject to JSTOR Terms and Conditions