8430 Mycologia. 92(6), 2000, pp, 1133-1137. © 2000 by The Mycological Society of America, Lawrence, KS 66044-8897 Amoebidium parasiticum is a protozoan, not a Trichomycete Gerald L. Benny cete order Amoebidiales (Zygomycota, Fungi) (Licht­ Department ofPlant Pathology, 1453 Fifield Hall, wardt 1986). With 4 and 7 species, respectively, these University ofFlorida, Gainesville, Florida, USA arthropodophilous symbionts are typically found in 32611-0680 fresh water in many parts of the world on Crustacea Kerry O'DonnelP and Insecta (Lichtwardt 1986, 1997, Lichtwardt et al J.'vlicrobial Properties Research Unit, National Center 1999), bloodworms (Licht:wardt and Williams 1992) for Agricultural Utilization Research, United States and Daphnia spp. (Lichtwardt 1986), or in the hind­ Department ofAgriculture, Agricultural Research gut or larval gills of invertebrates (Lichtwardt 1986, Service, 1815 North University Street, Peoria, Illinois, USA 61604-3999 Lichtwardt and Arenas 1996, Lichtwardt and Williams 1992, Williams and Lichtwardt 1990). Species of Amoebidiwll produce unicellular thalli that are typi­ Abstract: Classification ofthe Amoebidiales (Tricho­ cally attached externally to a host by an acellular mycetes, Zygomycota) 'within the Fungi is problem­ holdfast, whereas thalli of Paramoebidiwll are at­ atical because their cell walls apparently lack chitin tached to the cuticle of the hindgut, or to other Tri­ and they produce amoeboid cells during their life chomycetes inhabiting the host hindgut. cycle, A nearly full length fragment of the nuclear Study of the Amoebidiales has been advanced sig­ small subunit (SSU) rRt'\JA ofAmoebidiwll parasiticum nificantly by the axenic culture of Amoebidium par­ was amplified by the polymerase chain reaction asiticum (Whisler 1962), which enabled vVhistler (PCR) and sequenced to examine its phylogenetic (1963) and Trotter and vVhisler (1965) to determine relationships. Results of a Blasu'\J search of GenBank that the cell walls were not composed of cellulose or revealed that the A. parasiticum SSU rRNA sequence chitin. Culturing also has facilitated sequencing the was most closely related to that of Ichthyophonus hof mitochondrial genome, which at ca 300 kbp, is sig­ eri, an ichthyosporean in the Protozoa near the ani­ nificantly larger than that reported for any fun­ mal-fungal divergence. Maximum parsimony analysis gus (Lang BF, Burger G. http://megasum.bch. of ichthyosporean and fungal SSU sequences, using umontreal.ca/ogmp/projects/apara/gen.html). In sequences of choanoflagellates to root the 18S rDNA addition, V\71J.isler (1968) was able to induce the gene trees, resolved A. parasiticum as a strongly sup­ amoeba-cyst phase of the life cycle. These findings ported sister of 1. hofeli within the Ichthyophonida indicate the Amoebidiales are unique among organ­ clade of the protozoan class Ichthyosporea. In con­ isms classified as Fungi in the autapomorphic pro­ trast to other members of this class, which are mostly duction of amoeboid cells. Amoebae encyst to form obligate or facultative parasites ofvarious animals, A. (cysto)spores which develop into new thalli. In con­ parasiticum and other members of the Amoebidiales trast to Paramoebidium, species of Amoebidium also are only known to be arthropodophilous symbionts. form sporangiospores. The results also provide the first evidence that mito­ The rationale for classifying Amoebidiales as Tri­ chondrial cristae types exhibit homoplastic distribu­ chomycetes has been based on putative similarity in tions within the Ichthyosporea. thallus morphology and a shared symbiotic associa­ Key Words: Amoebidiales, 18S rRt'\JA, Ichthyo­ tion 'with arthropods. V\71J.istler (1963) and Lichtwardt sporea, mitochondrial cristae, phylogeny, Zygomycota (1986), however, have theorized that this order may represent a nonfungal evolutionary lineage derived from a protozoan ancestor. This conclusion was INTRODUCTION based on the premise that similarities between the Amoebidiales and other Trichomycetes may be due Amoebidium Cienkowski and Paramoebidium Leger & to convergent evolution. Efforts to examine evolu­ Duboscq are genera of unusual fungal-like microor­ tionary relationships of the Amoebidiales have in­ ganisms traditionally classified 'within the Trichomy- cluded phylogenetic analysis of 5S ribosomal RNA Accepted for publication June 13, 2000, (Walker 1984), serological analysis (Sanger et al I Corresponding author; email: kodonnell@sunca,ncaur,usda,gov 1972) and comparison of ribosomal Rt'\JA molecular 1133 1134 MVCOLOGIA masses (Porter and Smiley 1979). Collectively, results authentic Perkinsus spp. are nested \vithin the alvoelates), of these studies, together with those of Trotter and Psorospermllm haeckelii Hilgendorf U33180, Rhinosporidium "Vhistler (1965), suggest that the Amoebidiales are seeberi Wernicke AF118851, rosette agent L29455, Smittium not closely related to other Zygomycota; however, culisetae Lichtlv. AF007540.1, Sphaerosomo arcticum Yl6260.2 (cited in GenBank asJoestensenJP,Johansen S, Sperstad S, they do not resolve the phyletic affinities ofthe order. Landfald B. Sphaerosoma arcticmn, a new member of a clade Because phylogenetic analyses of nuclear small sub­ of protists near the animal-fungal divergence: systematic po­ unit (SSU) 18S rRNAs have helped resolve evolution­ sition, in vitro growth characteristics, and gross biochemical ary relationships among fungi (Bruns et al 1992) and composition), Spizellomyces ocuminatus (DJ.S. Barr) DJ.S. pseudofungi (Gunderson et al 1987), we generated Barr M59759.1, Unknown ichthyosporean AJ130859, Usti­ a nearly full length SSU sequence for A. pamsiticum lago moydis (DC.) Corda X62396.1. of 1720 bp to which we added sequences from 19 Sequences were aligned using ClustalX (Thompson et al taxa obtained from GenBank based on the results of 1997) and then by eye using SemWare Editor Professionall a BlastJ.~ search (Benson et al 1999). 32 vel's. 2.80b (SemWare Corporation, Marietta, Georgia). The dataset was analyzed by equally weighted maximum parsimony using PAUP*4.0b2 (Swofford 1999). Phylogenet­ ivLUERIALS A>m METHODS ic analysis employed a heuristic search, with gaps treated as missing data, 1000 random addition sequences with MUL­ jV1aterial examined.-Amoebidium parasiticum (FRA-I-14 = PARS on and TBR branch swapping. Clade stability was as­ NRRL 20524 = ATCC 32708) was grown in a shallow layer sessed by 1000 parsimony bootstrap replications, using 10 of distilled water covering one-tenth strength brain heart random addition sequences per replicate, and Bremer sup­ infusion agar in 60 X 15 mm plastic petri dishes at room port (= BS, Bremer 1988) using TreeRot (Sorenson 1996). temperature (ca 25 C) as recommended by Lichtwardt The PAUP* file has been deposited in TreeBASE as S491. (1986). Mter 1 wk cells were harvested and lyophilized over­ night. RESULTS jV10lecular biology.-Total genomic DNA was isolated from the lyophilized cellular material by the phenollchloroform The dataset consisted of 20 aligned sequences 2081 procedure described by O'Donnell et al (1997) for herbar­ bp in length, but 732 ambiguously aligned characters ium specimens. Polymerase chain reaction (PCR) and se­ were excluded from the analyses. Equally weighted quencing protocols were done according to the procedure maximum parsimony analysis of the 1349 included described by O'Donnell et al (1998), using primers de­ characters, 150 of which were phylogenetically infor­ scribed by "Vhite et al (1990), and NS21d 5'-TTGATA­ mative, lielded 3 most-parsimonious trees 527 steps GGGCAGAAATTTG and NS41g 5'-CCAACTGTCCCTAT­ in length (FIG. 1) (consistency index = 0.712, reten­ TAATCAT. tion index = 0.709, rescaled consistency index = Dataset construction and phylogenetic analysis.-The SSU 0.505). With this dataset, Amoebidiwn pamsiticumwas rDNA sequence of Amoebidium parasiticum was submitted deeply nested witllin the Ichthyophonida (bootstrap to a Blasu'\f search (vers. 2.0.10) of GenBank (Benson et al = 90%, BS = 7), a monophyletic sister-group of the 1999). The SSU sequence of lchthyophonus hoferi (Ichthyo­ Dermocystida (bootstrap = 99%, BS = 7). The latter sporea, Protozoa) yielded the highest score (Spanggaard et clade contained Rllinosporidium seeberi, the etiologi­ al 1996). Based on the results of this search, we downloaded cal agent of rhinosporidiosis of humans and other 19 SSU 18S rRNA from GenBank or the TreeCon website animals (Herr et al 1999). As indicated by the BlastJ.~ (http://rrna.uia.ac.ve/ssu/index.html). These included ichthyosporean sequences (Baker et al 1999, Cavalier-Smith search of GenBank, A. parasiticum was strongly sup­ 1998a, Herr et al 1999, Ragan et al 1996) and representa­ ported as a sister to Ichthyophonus hoferi (FIG. 1) tives of the major clades of fungi including the Harpellales (bootstrap = 94%, BS = 5). However, relationships (Trichomycetes). In addition, sequences of tlvo choanofla­ of major lineages within the Ichthyophonidia and gellates, Acanthocoepsis unguiculata and Diaphanoeca gran­ fungi were incompletely resolved by the SSU 18S dis (see FIG. 1), were selected for rooting the tree by the rDNA data as evidenced by one node within each of outgroup method. GenBank accession numbers for the 20 these clades with a bootstrap score of < 50%. terminals are as follows: Acanthocoepsis unguicu lata Ll0823, Amoebidium parasiticum AF274051, Anumfeca richardsi Wong & Beebee AF070445.1, Aspergillus fumigatus Fres. DISCUSSION M60300.1, Capnomyces stellatus S. W. Peterson & Lichtlv. Results