Journal of Medical and Veterinary Mycology (1994), 32, Supplement 1, 113-122

Changing taxonomic concepts and their impact on nomenclatural stability

G. S. DE HOOG 1, L. SIGLER 2, W. A. UNTEREINER 3, K. J. KWON-CHUNG 4, E. GUI~HO 5 AND J. M. J. UIJTHOF 1 1Centraalbureau voor Schimmelcultures, Baarn, The Netherlands," 2 University of Alberta Microfungus Collection, Edmonton, and 3Department of Botany, University of Toronto, Toronto, Canada," 4Clinical Mycology Section, National Institute of Health, Bethesda, USA; and s Unitd de Mycologie, Institut Pasteur, France

Experimental techniques, which are routinely applied in yeast systematics, are currently finding recognition in a growing number of filamentous taxa. Some other biochemical markers have recently been developed in hyphomycete . The spectrum of potential criteria now comprises characters of coenzyme-Q systems, secondary metabo- lites, protein electrophoresis, serology, nuclear (n) DNA/DNA reassociation, mole% G+C of DNA, protein electrophoresis, nutritional physiology and ultrastructural and karyological data. In addition, a wide range of molecular techniques is gaining rapid acceptance in evolutionary, systematic, ecological and epidemiological studies. Depend- ing on the aim of the study, partial DNA sequencing (mainly of SS and LS ribosomal genes and their spacers, but also of other genes), PCR-ribotyping and mitochondrial (mt) DNA restriction analyses are particularly powerful; for the establishment of the taxo- nomic position of taxa, 5.8S ribosomal (r) rRNA sequencing and Southern blotting with conserved genes are useful. Together with renewed in-depth morphological studies and the elucidation of teleomorph connections and (syn)anamorph life cycles, these tech- For personal use only. niques provide tools for an improved understanding of the phylogeny and ecological role of the distinguished taxa. Taxa are increasingly being classified along natural lines. The resulting changes in the taxonomic system, and thereby its impact on nomen- clature, may be considerable. This does not help in making taxonomy any more popular among medical mycologists. However, the significance to medical mycology of an improved taxonomic system should not be underestimated. Taxonomy is now becoming enabled to describe fungi as living entities in their natural ecological niche and thereby provides a better insight into their behaviour on the human body. We are now Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 beginning to realize that several fungi have a narrow ecological amplitude and hence their role as aetiological agents of human disease may be well defined and characteristic for the species. The clinical significance of found taxa can thus be evaluated with more precision. Below some examples of fungal groups, in which the taxonomic concepts are subject to rapid change, are presented.

Taxonomy of the : holomorphs of black yeasts The Herpotrichiellaceae is a well defined family of loculoascomycetes encompassing species characterized by their small, inconspicuous, setose ascomata, bitunicate asci with a thickened endotunica and hyaline to brown, septate ascospores. Although originally circumscribed to include a number of genera differentiated on the basis of Correspondence address: Dr G. S. de Hoog, Centraalbureau voor Schimmelcultures, PO Box 273, 3740AG, Baarn, The Netherlands. 113 114 DE HOOG ET AL.

65~ salmonis t ~ Exophia/a jeanselmei I pulcherrima 9~ Capronia moravica 89 I =Caproniamoravica ~- Capronia epimyces ICapronia species 1 i Capronia fungicola 97 t Tubeufia helicomyces r--- Pleospora rudis L~ ~ Leptosphaeria doliolum L~ [- Leptosphaeria microscopica Leptosphaeria bicolor - [- Chaetomium elaturn r~ [---- Hypornyces elatum I I ~ Ophiostoma ulmi / 96 I i Microascus cirrosus N ~ Neurospora crassa II - Leucostoma persoonii I I I t__ Histoplasma capsulatum | Eurotium rubrum IO-~DS [~ Talaromyces flavus P FIG. 1. Phylogenetic tree of partial 18S rDNA sequences of 23 fungal taxa. A heuristic search was performed using TBR (tree bisection-reconnection) branch swapping with the MULPARSand STEEPEST DESCENT options (PAUV version 3.1.1). A single, most parsimonious tree was found (169 steps, CI=0.539 excluding uninformative characters). E. rubrum Konig, Spiekermann & Bremer and T. flavus (K16cker) Stolk & Samson were employed as out-groups. Branch lengths are proportional to the number of inferred substitutions. The numbers above the branches are the frequencies with which

For personal use only. clades appeared in 100 boot-strap replicates.

stromal development, ascospore septation and the number of ascospores per ascus, only two genera are recognized in current taxonomic treatments of the Herpotrichiellaceae [15]. Capronica Sacc., the largest in the family, includes species possessing dark ascomata, pigmented ascospores and dematiaceous, conidial anamorphs, the majority of which belong to the Exophiala-Ramichloridium Rhinocladiella complex. This

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 complex constitutes a group of medically significant opportunists known as 'black yeasts'. Anamorphs belonging to the genus Exophiala Carmichael are unknown outside the Herpotrichiellaceae [6], and the predominance of black yeast anamorphs within the family suggests the close phylogenetic relationship of Exophiala and associated synanamorphs to the genus Capronia. Phylogenetic analysis of partial 18S rDNA sequences of 23 taxa (Fig. 1) demon- strates that species of Capronia possessing Exophiala or Ramichloridium anamorphs and species of Exophiala for which teleomorphs are unknown form a monophyletic group within the Ascomycotina. This clade is distinct from the Pleosporales (Tubeufiaceae and Pleosporaceae) and the unitunicate pyrenomycetes. Although this analysis supports the close phylogenetic relationship of species of Capronia and Exophiala, partial 18S sequence data do not permit the elucidation of inter-familial relationships. For example, Capronia fungicola (Samuels & Miiller) Untereiner, a species with a well developed stroma and an anamorph similar to Ramichloridium anceps (Sacc. & Ellis) de Hoog, and TAXONOMIC CONCEPTS AND NOMENCLATURAL STABILITY 115

Capronia species 1, with muriform ascospores, are not delimited from species of Capronia that lack a basal stroma and sympodial conidiogenous cells and possess phragmoseptate ascospores. The phylogram indicates, however, that Capronia moravica (Petrak) M/iller et al., the type species of the Herpotrichiellaceae, and Capronia epimyces Barr, which produces an Exophiala anamorph and Phialophora synanamorph, are distinct within the Herpotrichiellaceae. The type strains of Exophiala salmonis Carmichael (CBS 157.67) and (Langer.) McGinnis & Padhye var. jeanselmei (CBS 664.76) form a monophyletic group within the family (boot-strap support 65%), but the position of this clade relative to species of Capronia is unresolved. Sequence data from the more rapidly evolving rDNA internal transcribed spacers (ITS 1 and ITS 2) are presently being evaluated to address questions concerning the reliability of anamorphs and morphological characters in delimiting species of Capronia and the relationship between parasitic and saprotrophic species of Exophiala. Although teleomorphs are not known for the majority of black yeasts, the loculoascomycetous affinity of species of Exophiala has been confirmed by the produc- tion of pseudoparenchymatous ascomatal initials in isolates of E. jeanselmei var. jeanselmei and Exopthiala mansonii (Castell.) de Hoog [22]. Ascomatal initials and ascomata are not produced in vitro in the type strains of species of Exophiala [22] but immature, setose ascomata have been observed in additional isolates of (Kano) de Hoog and pallid, non-setose, pseudoparenchymatous initials are formed in one isolate of E. salmonis [W. A. Untereiner, unpublished data]. Morpho- logical and cultural characteristics are inadequate for the reliable separation of species of Exophiala and it is unlikely that these taxa can be clearly delimited using traditional taxonomic methods. The in vitro production of teleomorphs or their discovery in nature would furnish criteria necessary to determine unambiguously the systematic position of Exophiala and also provide information for elucidating the life cycles of these fungi.

For personal use only. Despite their occurrence on decaying plant material and on the decaying ascomata and basidiomata of other fungi, species of Capronia are unable to enzymatically degrade cellulose, lignin, starch, pectin and chitin [W. A. Untereiner, unpublished data]. Their inability to utilize these compounds supports the suggestion of Munk [16] that the members of the Herpotrichiellaceae are saprotrophs of secondary incidence (hyper- saprophytes) rather than primary degraders of plant material or mycoparasites. Species of Rhinocladiella Nannf., Ramichloridium Stahel ex de Hoog and Link resembling the anamorphs of some members of the genus Capronia are also found on

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 decaying wood and fungal fruit-bodies, and members of the genus Exophiala are commonly isolated from these substrata. The ecology and biological activity of black yeasts in natural systems are poorly understood, and until further investigated, questions concerning the significance of pleomorphism in the life cycles of these fungi and the evolution of parasitism within the genus Exophiala will remain unanswered. Nucleic acid sequence data, the near-exclusive anamorph-teleomorph connections and the production of pseudoparenchymatous ascomatal initials in isolates of Exophiala illustrate the close phylogenetic relationship of Exophiala and Capronia. It is questionable to assume, however, that all species of Exophiala represent the anamorphs of teleomorphic fungi or that the complex of synanamorphic genera associated with Exophiala and Capronia are related. Additional molecular, morphological, physiologi- cal and ecological data will be required to establish that the Herpotrichiellaceae represents the holomorphs of black yeasts and to accommodate Exophiala within the teleomorphic system of fungal classification. 116 DE HOOG ET AL.

Current research on taxonomy and ecology of black yeasts Most species of the anamorph genus Exophiala Carmichael, defined as filamentous fungi having annellidic conidiogenesis, also exhibit abundant yeast-like growth. E. dermatitidis is among the most hydrophilic black yeasts. The species is supposed to complete its entire anamorph cycle in submersion [7], while the less hydrophilic E. jeanselmei is more likely to inhabit solid substrates in moist environments [17]. Some strains of E. dermatitidis exhibit an additional, catenate form of growth. In one strain, originating from a systemic in Japan, and proven to be E. dermatitidis by nutritional physiology, polymerase chain reaction (PCR) ribotyping and partial 26S rRNA sequencing, this catenate synanamorph was remarkably similar to Cladosporium bantianum (Sacc.) Borelli. The synanamorph proved to be hydrophobic using MATH tests (microbial adhesion to hexadecane). It thus formed a link between Exophiala and the human pathogenic Cladosporium species. Several Exophiala species, among which is E. dermatitidis, are known to have Phialophora synanamorphs. The ascomycete genus Capronia (Herpotrichiellaceae) is known to have Exophiala, Phialophora and catenate, Cladosporium-like anamorphs, the latter differing markedly from the saprophytes around Cladosporium herbarum (Pers.: Fr.) Link (teleomorph: Mycosphaerella (de Not.) Johanson, Mycosphaerellaceae) (e.g. in partial 26S rRNA sequences and in conidial surface ultrastructure [K. Takeo, personal communication]). Consequently, the human- pathogenic Cladosporium species cannot be maintained in Cladosporium. Another unrelated group, comprising the black yeast genus Aureobasidium Viala & Boyer and the human associated species (Horta) Nishimura & Miyaji, are dothideaceous black yeasts. They have Q-10 (H2) ubiquinone systems and multikaryotic expanding hyphae. Most species are saprophytes on somewhat osmotic substrates such as phyllosphere [5, 8] or are opportunistic plant pathogens. The clinical spectrum is either widely variable, as in Aureobasidium pullulans (de Bary) Arn., or can be explained

For personal use only. by hydrophobic adhesion, as in H. werneckii. Herpotrichiellaceous black yeasts are identified with difficulty on the basis of morphology alone. Character sets applied more recently include serology, coenzyme-Q systems, thallus maturation, karyology, nutritional physiology, restriction fragment length polymorphism (RFLP) patterns of mitochondrial DNA and PCR-amplified partial ribosomal genes, rRNA sequencing, PCR-randomly amplified polymorphic DNA (RAPD) and Southern hybridization with internal transcribed spacer sequence (ITS) based probes. It is to be expected that on the basis of these new criteria the

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 taxonomy of black yeasts and their allies will change rather drastically within the next few years. For the routine distinction of herpotrichiellaceous anamorph species, assimilation of D-ribose, methyl-a-D-glucoside, lactose, soluble starch, DL-lactate, succinate, citrate, nitrate, nitrite, creatin and creatinin are particularly significant. PCR-ribotyping data show that Exophiala angulospora Iwatsu et al., E. jeanseImei var. heteromorpha (Nannf.) de Hoog, E. dermatitidis and E. salmonis Carmichael are different from the core of Exophiala around E. jeanselmei. Characters like gelatin liquefaction and tolerance of cycloheximide are applicable at higher taxonomic levels. The taxonomic position of Nadsoniella Issatchenko has to be reconsidered now that type strains of two described taxa from Russia have become available. Careful re-definition of species enables a more in-depth study of their ecology. Several species exhibit a characteristic clinical picture. E. dermatitidis is known from Asia as a rare neurotropic pathogen, but is commonly encountered in Europe in stools TAXONOMIC CONCEPTS AND NOMENCLATURAL STABILITY 117

and also asymptomatically in the mucus of lungs of patients with cystic fibrosis. Using PCR-RAPD it was shown that the two kinds of clinical behaviour can be found in a single population of identical strains. Research is being carried out, aiming for a circumscription of the in its natural ecological niche, for which the possibility of an endosymbiotic life cycle with bats is considered. The family Herpotrichiellaceae exhibits a remarkable tendency towards specializ- ation on the human host. It seems likely that several species have a natural ecological niche in association with vertebrates, either warm or cold-blooded. The future taxonomic system should reflect these preferences. It enables us to predict and to evaluate the clinical significance of these fungi.

Changing taxonomic concepts in medically important yeasts The basis for the definition of yeast taxa, as with other micro-organisms, has progressively shifted from the gross morphological/physiological level to the cellular biochemical and molecular level. The generic classification of the yeasts at the morphological/physiological level was based on the sexual or asexual state combined with nutritional profiles, mainly centered on the utilization of different carbon sources. For example, the genus Candida Berkhout used to be differentiated from the genus Cryptococcus Kfitzing by the presence of pseudo- or true hyphae; the genus Crypto- coccus was differentiated from Rhodotorula Harrison by the lack of carotenoid pigments and the ability to utilize inositol as the sole source of carbon. As the criteria shifted to the cellular/biochemical level, the presence or absence of hyphae and pseudohyphae was found to be insignificant. Rather, the presence or absence of xylose in cell hydrolysates and the diazonium blue B reaction (DBB), ultrastructures of cell wall and septal pore complexes, as well as the manner of budding (percurrent or holoblastic),

For personal use only. became more important criteria, since these characteristics were found to be more reliable for the phylogenetic relationship of the anamorphic yeasts to either Asco- mycetes or Basidiomycetes. By these criteria, several species previously classified in Candida have been transferred to the genus Cryptococcus or Rhodotorula and some species of Cryptococcus were transferred to Rhodotorula. At the molecular level, both DNA complementarity and nucleotide sequences of rRNAs have been used as the most reliable criteria. For example, rRNA sequences of Wingea van der Walt show no significant differences from those of Debariomyces

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 Lodder & Kreger-van Rij; hence, the genus Wingea was judged congeneric with Debariomyces. Such treatment is not surprising since the sexual state of Wingea is not significantly different from that of Debariornyces. The concept of species for teleomorphic yeasts has been different from that for asporogeneous yeasts. In heterothallic yeasts, the formation of viable spores by mating between two compatible strains has been, and still is, the most reliable criterion for species delimitation, while the morphology of the meiotic product has been the most important criterion for the speciation of homothallic yeasts. Molecular criteria such as DNA base composition (mole% G+C) or degree of DNA/DNA reassociation have clarified much confusion surrounding the conspecificity among the originally hetero- thallic yeasts that have lost their mating ability by mutation. Likewise, the molecular approaches resolved much taxonomic and nomenclatural confusion existing among homothallic or asporogeneous yeasts with minor morphological, physiological and biochemical differences. The analysis of isozymes, ubiquinones, karyotypes, RFLPs and 118 DE HOOG ET AL.

the content of fatty acid all contribute as supportive factors for classification when a species is heterogeneous. Among medically important yeasts, nomenclatural changes from new taxonomic concepts have been moderate in the genus Candida. On the basis of DNA/DNA homology data, Candida stellatoidea (Jones et al.) Langer & Guerra, Candida claussenii Lodder & Kreger-van Rij and Candida langeronii Dietrichson ex van Uden were judged conspecific. These species were originally separated from (Robin) Berkhout on the basis of a few morphological or physiological differences. For example, C. claussenii and C. Iangeronii were differentiated from C. albicans by the lack of germ tube and chlamydospores, respectively, while C. stellatoidea was differentiated from C. albicans by the lack of sucrose assimilation. Electrophoretic karyotyping and DNA fingerprinting using C. albicans-specific nucleotide sequences confirmed the synonymy of the four species [23]. On the other hand, the physical linkage analysis, karyotypes and the DNA fingerprinting of C. stellatoidea merited a varietal status. C. albicans var. stellatoidea (Jones et al.) Diddens & Lodder is found not to be merely a sucrose negative mutant of C. albicans, but is a variant resulting from a significant genetic rearrangement in the species [13]. DNA homology studies clarified that Candida pseudotropicalis (Castell.) Basgal is synonymous with Candida kefyr (Beijerinck) v. Uden & Buckley and that it is the anamorph of Kluyveromyces marxianus (Hansen) van der Walt var. marxianus. The conspecificity of Candida tropicalis (Castell.) Berkhout with Candida paratropicalis (Castell.) Basgal and the proposed conspecificity of (Castell.) Berkhout with Issatchenkia oriental& Kudryavtsev, based on DNA reassoci- ation values, were supported by karyotyping as well as Southern analysis of their DNA [23]. In the genus Cryptococcus, transient instability of nomenclature occurred in on the basis of teleomorphic morphology between the two

For personal use only. sub-groups of the species. DNA/DNA reassociation and genetic analysis of meiotic products from the crosses between the sub-groups clarified the confusion. The two- varietal status of C. neoformans is widely accepted [13] and further nomenclatural instability is not foreseen at this time of writing. In the genus Baillon, nomenclatural changes have been minimal. A new species of Malassezia sympodialis Simmons & Gu~ho was described on the basis of two isolates, obtained from a human source, which had a significantly lower G+C content of DNA compared to that of (Robin) Baillon, and had occasional sympodial budding. DNA reassociation values also indicated that M. sympodialis is Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 different from the other two Malassezia species [21]. The medical importance of M. sympodialis, however, is not known. Several nomenclatural changes in the genus Behrend have been proposed recently on the basis of new taxonomic concepts. The amount of confusion surrounding such a proposal is expected to be immense among medical mycologists. Within the genus Trichosporon, Trichosporon beigelli (Kiichenm. & Rabenh.) Vuill., a pathogenic species and also the generic type species, has been known as one of the most heterogeneous taxa on the basis of colonial and microscopic morphology, ecology and physiology. A recent comprehensive study [11] of the genus Trichosporon, employing physiological, ultrastructural and molecular approaches suggested that T. beigelii contains several distinct groups of organisms. As a result, seven species were proposed to replace the name T beigelii. The criteria used to reclassify the species of Trichosporon included the ultrastructure of septal pores, G+C content and reassociation values of TAXONOMIC CONCEPTS AND NOMENCLATURAL STABILITY 119

DNA, Co-Q enzymes, nutritional profiles and partial 26S rRNA nucleotide sequences. The sequences of 26S rRNA and DNA/DNA reassociation values, however, were used as the most important criteria for species demarcation. There will be considerable difficulties in adopting this new classification system in clinical laboratories until species-specific diagnostic probes are developed. It appears that further study is necessary to clarify the precise relationship between genetically close species, especially between those showing an intermediate range of DNA reassociation values. For example, Trichosporon inkin (Oho ex Ota) do Carmo-Sousa & v. Uden and Tricho- sporon ovoides Behrend both cause piedra, one mainly on pubic hair and the other mainly on capital hair. They both have Co-Q with 9 isoprene units and their G+C content falls between 58.3 and 59.5%. The DNAJDNA reassociation values between the two species was 12- 53% and such a difference was reflected in rRNA by five nucleotide sequence substitutions. The only physiological marker that can be used for clear separation of the two species is L-rhamnose assimilation by T. ovoides but not by T. inkin. The number of isolates tested of T. ovoides for L-rhamnose, however, was only three and is too small to draw a definitive conclusion. The most reliable characteristic for the separation of these two species appears to be morphology, namely the presence or absence of sarcinae. Such a marker, however, could not be considered a reliable criterion in other Trichosporon species since the sarcinae-producing species Fissuricella filamenta Pore et al. has nearly identical rRNA sequences and high DNA/DNA reassociation value with Trichosporon asteroides (Rischin) Ota which lacks the sarcinae. The newly proposed system revealed, however, that T. beigelii indeed is a heterogeneous species containing genetically diverse fungi.

Systemic pathogens and other Onygenales Historically, Histoplasma Darling, Blastomyces Gilchr. & Stokes, Stiles,

For personal use only. Paracoccidioides Almeida and Emmonsia cir. & Mont. have been treated in separate genera because of the taxonomic weight given to their parasitic forms. The first four pathogens were originally defined on the basis of histopathology with both Histoplasma and Coccidioides described as members of the protozoa. Blastomyces was considered to be a form of 'blastomycetes' or yeast. Mating experiments established connections to teleomorphs in the ascomycetes for Blastomyces dermatitidis Gilchrist & Stokes in Ajellomyces McDonough & Lewis in 1968, and for Histoplasrna capsulatum Darling in Emmonsiella Kwon-Chung in 1972.

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 Later, Ernmonsiella was reduced to synonymy with Ajellornyces on morphological grounds. The tiny (1-1.5/zm diam.), globose ascospores, evanescent asci and mesh-like ascocarp composed of swollen hyphal cells ending in coiled appendages placed Ajellomyces within the family then known as the Gymnoascaceae. Later revision by Currah [4] placed Ajellomyces in the family Onygenaceae, one of four families within the order Onygenales, which included the teleomorphs of many of the human pathogenic fungi such as the . Sexual reproduction by the other dimorphic pathogens is unknown or does not exist, but nucleic acid sequencing today allows for the testing of relatedness among strictly anamorphic fungi. Suggestions of a possible biological relationship between B. dermatitidis and Emmonsia parva (Emmons & Ashburn) Cif. & Mont. and between these fungi and H. capsulatum were made as early as 1942 but largely ignored. During studies for the natural reservoir of Rixford & Gilchrist, Emmons & Ashburn [10] observed a second fungus in the lungs of rodents, characterized by single, spherical, 120 DE HOOG ET AL.

non-budding cells varying between 10 and 14 pm in diameter and having dense staining contents. They described the isolated fungus as a 'phycomycete' under the name Haplosporangium parvum. These authors speculated on a possible biological relation- ship between C. immitis and Histoplasma parvum based on the occasional finding of mixed infections in animals, cross-reactivity with skin test antigens, and the resemblance in tissue forms, but they recognized that the morphological features of their new fungus in culture were closer to B. dermatitidis and H. capsulatum than to C. immitis. Later, Dowding [9] observed similar, but larger (up to 300 pm diam.) tissue forms in the lungs of rodents captured in Alberta, Canada. Established cultures were identified as H. parvum. She judged that B. dermatitidis, Blastomyces (Paracoccidioides) brasiliensis and H. capsulatum were closely related to H. parvum. In particular, B. dermatitidis appeared identical to H. parvum in macroscopic and microscopic features, even though separable in tissue forms. C. immitis appeared to be less closely related because of its formation of arthroconidia and development in tissue of cells containing endospores. The appropriate generic disposition of H. parvum has been controversial. In 1951, Carmichael suggested that the fungus was wrongly placed in the genus Haplosporangium and confirmed the close relationship between H. parvum and B. dermatitidis. Only the tuberculate nature of the conidia of H. capsulatum appeared to set it apart. Ciferri and Montemartini erected the genus Emmonsia for H. parvum in 1959, and Emmons and Jellison added a second species, Emmonsia crescens in 1960. In 1962, Carmichael [3] took up the genus for fungi forming solitary aleurioconidia and having teleomorphs, where known, in the Gymnoascaceae. The need for a new generic placement for B. dermatitidis has been recognized for some time, in part because of uncertainty about the intentions of Gilchrist and Stokes. They did not state clearly whether they were adding another species or proposing a new genus, in which case it would be a later homonym of Blastomyees Cost. & Roll. In either case

For personal use only. Blastomyees could not be used since Vuillemin, in 1911, had shown that the type species, Blastomyces luteus Cost. & Roll., was conspecific with a fungus later identified as Chrysosporium merdarium (Link) Carmichael. Previously, Dodge had proposed a new combination for B. dermatitidis in Zymonema, a genus of yeast-like fungi, but Carmichael transferred it, together with H. parvum, to Chrysosporium [3]. Neither transfer was widely accepted. Histoplasma was retained separately, but considered doubtfully distinct. New information derived from various sources, including nucleic acids, raises anew

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 the question of whether all genera should be retained. Phylogenetic analysis of the 18S rDNA sequences of four pathogenic species by Bowman et al. [2] grouped H. capsulatum with B. dermatitidis and determined that C. immitis was a near relative. H. capsulatum differed from B. dermatitidis at only 12 positions within the 1713 base sequences, whereas they differed from C. immitis by 35 and 33 substitutions, respec- tively. A later study [1], which included several pathogenic and non-pathogenic species having onygenalean affinities, showed strong support for the grouping of the E. parva with B. dermatitidis and H. capsulatum, but excluded the dimorphic C. immitis. Previously, Sigler [in 14] suggested that the teleomorph of E. parva could be predicted in Ajellomyces. Mating tests recently conducted among isolates of E. crescens have demonstrated the formation of a teleomorph classifiable in Ajellomyces [L. Sigler, unpublished data]. A description of this new species of Ajellomyces is in preparation. Evidence obtained from mating, nucleic acid sequence data and morphology support the observations of previous workers of a phylogenetic relationship between E. parva, TAXONOMIC CONCEPTS AND NOMENCLATURAL STABILITY 121

B. dermatitidis and H. capsulatum. While the maintenance of H. capsulatum in a separate genus can be argued on stability alone, serious consideration should be given to placing E. parva and B. dermatitidis into the same anamorphic genus. Such a move would demonstrate the close relationship between these species and be of value to researchers investigating improved diagnostic methods such as DNA-based detection systems. Indeed, it may be questioned whether the diagnostic tests developed thus far are reliable in distinguishing the two species; the in vivo and in vitro morphologies currently remain the most reliable criteria. Whether host adaptation has resulted in a distinct evolutionary lineage among the pathogens is a hypothesis which has been tested recently by molecular techniques [1], but the emphasis on dimorphism or pathogenicity as predictors of relationship may be misplaced. Lesser weight has been given to the morphological features of the saprophytic phase in predicting possible affinities for C immitis, but both Emmons and Dowding judged that the formation of arthroconidia set C. immitis apart from the other dimorphic fungi. The manner of arthroconidial formation, racket hyphae and somatic nuclear structures, have suggested close relationships to some species of Malbranchea Sacc. & Penz. and to some members of the Gymnoascaceae [12, 19, 20]. Although the guanine plus cytosine content (49.41~49.61) placed C. immitis at the low end of the range for ascomycetes and apart from several taxa of gymnoascaceous fungi known to produce arthroconidia [18], some of the compared taxa, such as Arachniotus Schroet. or Pseudoaraehniotus Kuehn would be unlikely to be close relatives. Phylogenetic analysis using 18S rDNA sequences [1, 2] provided support for the contention that C. immitis has affinities to the ascomycetes, and for the grouping of C. immitis with Uncinocarpus reesii Sigler & Orr, a heterothallic member of the Onygenaceae having a Malbranchea anamorph, and with two other anamorphic and teleomorphic species of Malbranchea.

For personal use only. Considerably compelling evidence is now available for a better understanding of the taxonomic affinities of the major human pathogens. The close relationship of two of them--B, dermatitidis and E. parva--argues for placement within the same genus. The merits of maintaining//, capsulatum separately also requires consideration. Finally, C. immitis has affinities to some (but not all) species of Malbranchea and will most likely prove to be a heterothallic ascomycete whose teleomorph might be expected in the Onygenaceae.

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 CONTRIBUTORS The contributors to this symposium were: W. A. Untereiner, Taxonomy of the Herpotrichiellaceae: holomorphs of black yeasts; G. S. de Hoog, Current research on taxonomy and ecology of black yeasts; K. J. Kwon-Chung, Changing taxonomic concepts in medically important yeasts; L. Sigler, Systemic pathogens and other Onygenales. The co-convenors were G. S. de Hoog and L. Sigler.

REFERENCES

1. BOWMAN, B. H. & TAYLOR, J. W. 1993. Molecular phylogeny of pathogenic and non-pathogenic Onygenales. In: D. R. REYNOLDS & J. W. TAYLOR (Eds) The Fungal Holomorph: Mitotic, Meiotic and Pleomorphic Speeiation in Fungal Systematics, pp. 169-178. CAB International, Wallingford. 2. BOWMAN,I. H., TAYLOR, J. W. & WHITE, T. J. 1992. Molecular evolution of the fungi: human pathogens. Molecular Biology and Evolution, 9, 893 904. t22 DE HOOG ET AL.

3. CARMICHAEL,J. W. 1962. Chrysosporium and some other aleuriosporic hyphomycetes. Canadian Journal of Botany, 40, 1137 1173. 4. CURRAH, R. S. 1985. Taxonomy of the Onygenales: Arthrodermataceae, Gymnoascaceae, Myxotrichaceae and Onygenaceae. Mycotaxon, 24, 1 216. 5. DE COCK, A. W. A. M. 1994. Population biology of Hortaea werneckii based on restriction patterns of mitochondrial DNA. Antonie van Leeuwenhoek International Journal of General and Molecular Microbiology, 65, 21 28. 6. DE HOOG, G. S. & McGTNNIS, M. R. 1987. Ascomycetous black yeasts. Studies in Mycology, 30, 187-199. 7. DE HOOG, (J. S., TAKEO, K., YOSHIDA, S., GOTTLICH, E., NISHIMURA, K. & MIYAJI, M. 1994. Pleoanamorphic life cycle of Exophiala (Wangiella) dermatitidis. Antonie van Leeuwenhoek International Journal of General and Molecular Microbiology, 65, 143-153. 8. DE HOOG, G. S. & VAN DEN ENDE, A. H. G. 1992. Nutritional pattern and eco-physiology of Hortaea werneckii, agent of human tinea nigra. Antonie van Leeuwenhoek International Journal of General and Molecular Microbiology 62, 321-329. 9. DOWDING, E. S. 1947. The pulmonary fungus, Haplosporangium parvum, and its relationship with some human pathogens. Canadian Journal of Research Section E, 25, 195-206. 10. EMMONS,C. W. & ASHBURN, L. L. 1942. The isolation ofHaplosporangiumparvum n.sp. and Coccidioides immitis from wild rodents. Public Health Reports, 57, 1715-1727. II. Gu~)io, E., SMITH, M. T., DE HOOG, G. S., BILLON-GRAND, G., CHRISTEN, R. & BATENBURG-VANDER VEGTE, W. H. 1992. Contributions to a revision of the genus Trichosporon. Antonie van Leeuwenhoek International Journal of General and Molecular Microbiology, 61, 289-316. 12. KWON-CHUNG, K. J. 1969. Coccidioides immitis: cytological study on the formation of the arthrospores. Canadian Journal of Genetics and Cytology, 11, 43 52. 13. KWON-CHuNG, K. J. & BENNETT, J. E. 1992. Medical Mycology. Lea & Febiger, Philadelphia, PA. 14. McGINNIS, M. R., SIGLER, L., BOWMAN, B. H., MASUDA, M. & WANG, C. J. K. 1991. Impact of conidiogenesis, teleomorph connections, pleomorphism and molecular genetics on evolving hyphomycete systematics. Journal of Medical and Veterinary Mycology, 29, (Suppl. 1), 261 269. 15. MOLLER, E., PETRINI, O., FISHER, P. J., SAMUELS, G. J. & ROSSMAN, A. Y. 1987. Taxonomy and anamorphs of the Herpotrichiellaceae with notes on generic synonymy. 16. MUNK, A. 1957. Danish Pyrenomycetes. Dansk Botanisk Arkiv, 17, 1-491. 17. NISHIMURA, K., MIYAJI, M., TAGUCHI, H. & TANAKA, R. 1987. Fungi in bathwater and sludge of bathroom drainpipes. 1. Frequent isolation of Exophiala species. Mycopathologia, 97, 17-23. For personal use only. 18. PAPPAGIANIS, D., ORNELAS, A. & HECTOR, R. 1985. Guanine plus cytosine content of the DNA of Coccidioides Onmitis. Sabouraudia: Journal of Medical and Veterinary Mycology, 23, 451~54. 19. S1GLER, L. 1993. Perspectives on Onygenales and their anamorphs by a traditional taxonomist. In: D. R. REYNOLDS & J. W. TAYLOR (Eds) The Fungal Holomorph: Mitotic, Meiotic and Pleomorphic Speciation in Fungal Systematics pp. 161-168. CAB International, Wallingford. 20. SIGLER, L. & CARM~CHAEL, J. W. 1976. Taxonomy of Malbranchea and some other hyphomycetes with arthroconidia. Mycotaxon, 4, 349-488. 21. SIMMONS, R. B. & GU~HO, E. 1990. A new species of Malassezia. Mycological Research, 94, 1146 1149. 22. UNTEREINER, W. A. 1994. A simple method for the in vitro production of pseudothecia in species of

Med Mycol Downloaded from informahealthcare.com by 192.87.37.12 on 03/26/12 Capronia. Mycologia, 86, 290-295. 23. WICKES, B. L., HICKS, J. B., MERZ, W. G. & KWON-CHUNG, K. J. 1992. The molecular analysis of synonymy among medically important yeasts within the genus Candida. Journal of General Microbiology, 138, 901-907.