sign in sign up
<<
Home , Acanthamoeba, Amaurochaete, Badhamia, Badhamia panicea, Badhamia utricularis, Comatricha

The Journal of Published by the International Society of Eukaryotic Microbiology Protistologists

J. Eukaryot. Microbiol., 57(2), 2010 pp. 189–196 r 2010 The Author(s) Journal compilation r 2010 by the International Society of Protistologists DOI: 10.1111/j.1550-7408.2009.00466.x Invalidation of Hyperamoeba by Transferring its to Other Genera of

ANNA MARIA FIORE-DONNO,a AKIKO KAMONO,b EMA E. CHAO,a MANABU FUKUIb and THOMAS CAVALIER-SMITHa aZoology Department, University of Oxford, South Parks Road, OX1 3PS Oxford, United , and bThe Institute of Low Temperature Science, Hokkaido University, Kita 19, Nishi 8, Kita-ku, Sapporo, Hokkaido 010-0819, Japan

ABSTRACT. The Hyperamoeba Alexeieff, 1923 was established to accommodate an aerobic exhibiting three life stages— amoeba, flagellate, and cyst. As more species/strains were isolated, it became increasingly evident from small subunit (SSU) gene phylo- genies and ultrastructure that Hyperamoeba is polyphyletic and its species occupy different positions within the Myxogastria. To pinpoint Hyperamoeba strains within other myxogastrid genera we aligned numerous myxogastrid sequences: whole small subunit ribo- somal (SSU or 18S rRNA) gene for 50 dark-spored (i.e. Stemonitida and Physarida) Myxogastria (including a new ‘‘Hyperamoeba’’/ sequence) and a 400-bp SSU fragment for 147 isolates assigned to 10 genera of the Physarida. Phylogenetic analyses show unambiguously that the type species Hyperamoeba flagellata is a (Physarum flagellatum comb. nov.) as it nests among other Physarum species as robust sister to Physarum didermoides. Our trees also allow the following allocations: five Hyperamoeba strains to the genus ; Hyperamoeba dachnaya, Pseudodidymium cryptomastigophorum, and three other Hyperamoeba strains to the genus Didymium; and two further Hyperamoeba strains to the Physaridae. We therefore abandon the polyphyletic and redundant genus Hyperamoeba. We discuss the implications for the ecology and evolution of Myxogastria, whose amoeboflagellates are more widespread than previous inventories supposed, being now found in freshwater and even marine environments. Key Words. , molecular phylogeny, , Myxomycetes, slime molds, SSU rRNA gene.

genus Hyperamoeba being polyphyletic, its ‘‘raison d’eˆtre’’ is INCE its description in 1923, 12 amoeboflagellates have been questionable. To avoid taxonomic confusion, a more precise iden- S allocated to the genus Hyperamoeba Alexeieff. They have tification of Hyperamoeba taxa and isolates is highly desirable. been found in very diverse terrestrial and freshwater Three strains are still available in culture (Table 1, and Hyper- (Karpov and Mylnikov 1997; Michel, Walochnik, and Aspo¨ck amoeba sp. ATCC 50247 deposited by T.K. Sawyer, not included 2003; Walker et al. 2003; Walochnik, Michel, and Aspo¨ck 2004; in this study) thus deserving a better determination for future use. Zaman and Adoutte 1999). An organism phylogenetically close to Also, Hyperamoeba SSU rRNA gene sequences have been de- Hyperamoeba dachnaya has been found in sea urchins of the Ad- posited in GenBank, and confuse the BLAST results when iden- riatic Sea (Dykova´ et al. 2007). Hyperamoeba is characterized by tifying environmental sequences (Kamono and Fukui 2006; the alternation of three life stages, amoeboid, flagellate, and Kamono et al. 2009a, b; Win Ko Ko et al. 2009). To avoid tax- cyst—a feature that is found in many unrelated protistan groups, onomic and phylogenetic confusion, clarifying the nomenclature such as Cercozoa, Heterolobosea, and Amoebozoa. of the two taxa for which EST projects are in progress—H. dach- The classification of Hyperamoeba has been controversial, es- naya, which has been often mis-spelled as ‘‘dachnya,’’ and pecially because a significant character, the second flagellum, Hyperamoeba sp. ATCC 50570—is especially important, since sporadic and adhering to the cell, has sometimes been overlooked the massive data that will be generated are and will be used to (see the excellent introduction by Walochnik et al. 2004 and ref- infer evolutionary hypotheses (Watkins and Gray 2008). We used erences therein; Dykova´ et al. 2007). Ultrastructural studies have our extensive database of dark-spored myxogastrian SSU rRNA revealed that a second basal body is always present, whether the gene sequences to find the closest relative to each isolate of second flagellum is developed or not, and striking ultrastructural Hyperamoeba, including a new one sequenced here, and renamed similarities with Myxogastria have been underlined (Dykova´ et al. all as precisely as possible. We abandon this redundant genus, 2007; Karpov and Mylnikov 1997; Walker et al. 2003; Walochnik which is a life form not a , and establish new combinations et al. 2004). The flagellar apparatus in myxogastrids (including within the genera Stemonitis, Didymium, and Physarum for all fruticulosa) is highly conservative (Karpov and My- named Hyperamoeba species. lnikov 1997; Nelson and Scheetz 1975) and is characterized by two centrioles mutually oriented at a roughly orthogonal angle varying from 601 to 1201, connected by a ‘‘short striated fibre’’ whose variations among taxa may have phylogenetic meaning MATERIALS AND METHODS (see Karpov, Novozhilov, and Chistiakova 2003, Walker et al. 2003) for detailed illustrations and discussion). This particular Culture, DNA extraction, and sequencing. We found an flagellar apparatus is probably an apomorphy of the clade My- Hyperamoeba-like flagellate as a contaminant in plates of Mi- xogastria1C. fruticulosa—Myxogastrea sensu (Cavalier-Smith, cronuclearia, whose origin could not be established. The culture, Chao, and Oates 2004), as confirmed by molecular phylogeny named Didymium sp. OX13PS, is maintained in Prof. Cavalier- (Fiore-Donno et al. 2010). Smith’s laboratory in Volvic natural mineral water (available in Phylogenies based on the small subunit (SSU) rRNA gene have most shops) together with bacteria carried over during subcultur- confirmed that all Hyperamoeba sequenced to date belong to ing. DNA was extracted with the MoBio UltraClean Soil DNA kit different groups of Myxogastria: some cluster in the order Physar- (MO BIO Laboratories Inc., Carlsbad, CA). The SSU rRNA gene ida in the families Physaridae and Didymiidae, and some in the ( 1,900 bp long) was amplified with primers G and F (G: 50-AA order Stemonitida (Dykova´ et al. 2007; Michel et al. 2003; Walker CCTGGTTGATCCTGCCAGTAGTCATATGC-30;F:50-GATC et al. 2003; Walochnik et al. 2004; Zaman and Adoutte 1999). The CTTCTGCAGGTTCACTAC-30). The sequence was obtained in three overlapping fragments, using the primer pairs G-9R (9R: 50-TT 0 0 Corresponding Author: A. M. Fiore-Donno, Zoology Department, AGAGCTGGAATTACCG-3 ), 4F-12R (4F: 5 -CCGCGGTAATT 0 0 0 University of Oxford, South Parks Road, OX1 3PS Oxford, United CCAGCTCC-3 ; 12R: 5 -ACCGGCCATGCACCACC-3 ) and 6F- Kingdom—Telephone number: 1441865281322; FAX number: F (6F: 50-GGTGGTGCATGGCCG-30). Although these primers 144165281310; e-mail: afi[email protected] have some mismatches with the sequence, this has not impaired 189 190 J. EUKARYOT. MICROBIOL., 57, NO. 2, MARCH–APRIL 2010

Table 1. New names given to Hyperamoeba species and strains. Asterisks indicate EST ongoing projects.

New name Previous name Reference Origin Substrate GenBank # Stemonitis sp. Hyperamoeba sp. strain A1K Walochnick et al. DE, Melsbach Maple tree bark AY321107 (2004) Stemonitis sp. Hyperamoeba sp. strain AH1 Walochnick et al. DE, Melsbach Maple tree bark AY321108 (2004) Stemonitis aff. flavogenita Hyperamoeba sp. strain B1/2 Walochnick et al. DE, Plauen Drinking water AY321109 (2004) treatment plant Stemonitis sp. Hyperamoeba sp. strain BuP Walochnick et al. DE, Melsbach tree bark AY321110 (2004) Stemonitis aff. flavogenita ÃHyperamoeba sp. ATCC50570 Zaman & Adoutte PK, Karachi Human fecal sample AF093247 (1999) Physarum flagellatum (Alexeieff) Hyperamoeba flagellata Karpov & Mylnikov RU, Yaroslavl Surface ice of a pond AF411289 Fiore-Donno, Kamono & Alexeieff ATCC 50637 (1997) Cavalier-Smith 2009, comb. nov. Undetermined Physaridae Hyperamoeba sp. G1a Walochnick et al. DE, Giessen Physiotherapy bath AY321113 (2004) Undetermined Physaridae Hyperamoeba sp. W2i Walochnick et al. DE, Wielbad Physiotherapy bath AY321115 (2004) Didymium sp. Hyperamoeba sp. E2/8 Walochnick et al. DE, Dresden Drinking water AY321111 (2004) treatment plant Didymium sp. Hyperamoeba sp. E3P Walochnick et al. DE, Dresden Drinking water AY321112 (2004) treatment plant Didymium sp. Hyperamoeba sp. Hpl Walochnick et al. DE, Andernach Sycamore tree bark AY321114 (2004) Didymium dachnayum (Walker, ÃHyperamoeba dachnaya Walker et al. RU, St-Petersburg Sediments under the AY062881 Karpov, Frolov & Patterson) Walker, Karpov, Frolov & (2003) ice, lake Osinovoye Fiore-Donno, Kamono & PattersonCCAP1535/1 Cavalier-Smith Didymium cryptomastigophorum Pseudodymium Michel et al. (2003) DE, Wildbad Physiotherapy bath AY207466 (Michel, Walochnik & Aspock) cryptomastigophorum Fiore-Donno, Kamono & Michel, Walochnik & Aspock Cavalier-Smith CCAP1573/1 their effectiveness. The sequence is deposited in GenBank under added where longer sequences had insertions, in order to keep as accession number GQ249857. many homologous positions as possible in the final alignment Sequence alignments. (Fig. 1). The length of Helix 44 was larger in Didymiidae (mean Large dark-spored clade alignment. A first dataset of length: 45.9, standard deviation: 3.2, mode: 47) than in Physaridae nearly complete SSU rRNA gene sequences, including the 12 Hyper- (mean length: 41.6, standard deviation: 2.0, mode: 41). The se- amoeba strains/species, Pseudodidymium cryptomastigophorum,and quence was also more variable in Didymiidae (Fig. 1). The two representatives of the Didymium-like marine myxogastrids, trimmed alignment comprised 147 taxa and 417 nucleotide posi- was assembled with nearly all available Stemonitida and Physar- tions, of which 203 were constant. The alignment was character- ida sequences plus one new sequence of Didymium sp. OX13PS. ized by high diversity in only 214 variable sites, for a Sequences that were identical (or nearly so) or partial sequences in proportionally high number of taxa, so it can provide only lim- clades already well represented were not included. ited phylogenetic information, but will add useful information to spp. and Semimorula liquescens, the sister group of Stemonitida our first analysis. The alignment was split into the two families, and Physarida in SSU rRNA gene analyses, were used as out- Physaridae (75 taxa) and Didymiidae (72 taxa) for separate ana- groups (Fiore-Donno et al. 2008, 2009, 2010). An alignment of lyses and display of the trees. The three alignments used for our 1,451 positions, comprising 54 taxa, was used for phylogenetic final analyses are deposited in TreeBASE (S2513 and M4797-99). analyses. Phylogenetic analyses. Short Physarida alignments. To better allocate the Physar- Large dark-spored clade alignment. The general time re- ida-related hyperamoebae, a second dataset was made for a SSU versible (GTR) model taking into account invariant characters and rRNA fragment obtained for 10 genera and 95 nominal species of a gamma-distributed rate heterogeneity among sites (GTR1I Physarida (Kamono and Fukui 2006; Kamono et al. 2009b), along 1gamma) (Lanave et al. 1984; Rodriguez et al. 1990) was chosen with the truncated Physarida sequences used in the previous align- using Modeltest 3.7 (Posada and Crandall 1998). Maximum like- ment, totalling 147 strains. From the 242 short fragments pub- lihood (ML) analyses were run using RAxML 7.0.4 (Stamatakis, lished by Kamono and Fukui (2006) and Kamono et al. (2009b), Ludwig, and Meier 2005) with the GTR model of substitution and only unique or nearly unique ( 98% similarity) sequences were a 25 rate category discrete gamma distribution. The alignment selected for each taxon. All sequences were manually aligned us- comprised 743 patterns where incompletely undescribed positions ing BioEdit software version 7.0.5.3 (Hall 1999). The 400 nu- were negligible (0.02%). The best-scoring ML tree was inferred, cleotide fragment comprises the highly variable region V7, and in under the GTRMIX model, from 200 randomized starting maxi- particular, the Helix 44, for which a putative secondary structure mum parsimony trees (average lkl: 15,980.3, best lkl: was inferred from its 39–53 nucleotides, taking as model that of 15,974.16). The best-scoring tree was used to report the con- (Johansen, Johansen, and Haugli 1988). fidence values as percentages (below called BML) obtained Paired bases were found in the shortest helices first and gaps were through 1,000 non-parametric bootstraps under the GTRCAT FIORE-DONNO ET AL.—INVALIDATION OF HYPERAMOEBA 191

GCAA T TCG 1,000 trees that were summarized in an extended majority-rule GCAA GCAA consensus tree by CONSENSE. GAAA TCAC CCC GGG CCC GGG RESULTS CCC GGG CCC GGG Large dark-spored clade phylogeny. CG To identify the species or strains described as Hyperamoeba, we included all relevant GC SSU rRNA gene sequences available. The resulting SSU rRNA CC GG gene tree shows a topology consistent with previously published CGCGTGresults (Fiore-Donno et al. 2008), with Stemonitida being para- TGA CGA phyletic to Physarida (Fig. 2). Stemonitida is divided into four, GGG CTC GT GCT C well supported clades: (1) atrosporum group— GGG CCT CTT TTC ‘‘Meriderma’’ ad interim (Poulain, Meyer, and Bozonnet 2002); CCTTGG TTT TCT (2) the group, including comata and CCC GGG CCC GGG papillatum; (3) the Stemonitis group, including GGG CCT GGG CCC oblonga and Symphitocarpus impexus; and (4) Lam- G ATCGGGTTCproderma species with shining peridium (Lamproderma ovoideum C C C C C C and Lamproderma sauteri), sister to Physarida. Five Hyperamoe- CCC AAA CCC AAA ba are included in the Stemonitis group (Hyperamoeba sp. B1/2, CCC GGG CCC GGG Hyperamoeba sp. ATCC50750, Hyperamoeba sp. BuP, Hyper- GGG CCC GGG CCC amoeba sp. A1K, Hyperamoeba sp. AH1), with strong support AAATAG AAA AAA (i.e. Bayesian inference [BI] posterior probability: 1.0; bootstrap CCT GGG CCC GGG replicates reported on the best ML tree: 98; bootstrap starting with GGGCCCCGGCCG a NJ tree: 100—the statistical supports are henceforth given in this GCG CGC GAG CTC order). There is more support for a monophyletic group of Lam- GCC GGC TCC GGG proderma1Physarida (BI/ML/NJ—1.0/99/100) than for Physar- GGG CCC GGG CCT ida alone (BI/ML/NJ—1.0/63/not recovered). GGG T TC A GG TCC The two families composing Physarida are recovered, with Didymiidae being paraphyletic to a holophyletic Physaridae A 123 B 456 (maximum support in all analyzes). In Didymiidae, two well-sup- Fig. 1. Diagram of two models (A, B) of secondary structure of Helix ported main groups emerge, leaving apart carestiae, 44 of the variable area V7 of the small subunit rRNA of Myxogastria. The Didymium dubium K15, and Protophysarum phloiogenum, which paired regions are in columns, the terminal loops in lines. Paired nucleo- occupy weakly supported basal positions: a group composed tides (including the pair G-T) are shown in bold. (A) Physaridae: 1, Phys- of ‘‘Didymium’’ 1 and 2 and a robust clade named ‘‘’’ arum hongkongense AB259511; 2, reticulatum AB259476 & (Fig. 2). In the first group (BI/ML/NJ—1.0/99/98), two groups can 77; 3, alpina AB259449. (B) Didymiidae: 4, Protophysarum be distinguished: a terminal, well-supported clade, labelled ‘‘Did- phloiogenum AY23018; 5, Lepidoderma carestianum AM231296; 6, Did- ymium 2’’ (BI/ML/NJ—1.0/99/100), and a paraphyletic clade la- erma deplanatum AB259367 & 68. belled ‘‘Didymium 1’’ (Fig. 2). The terminal clade ‘‘Didymium 2’’ is composed of the Didymium iridis-nigripes complex (excluding strain HA4-1), H. dachnaya, P. cryptomastigophorum, and the Didymium-like marine organisms (Fig. 2). ‘‘Didymium 1,’’ a pa- model. Additional ML analyses were conducted using Treefinder raphyletic group, includes Didymium iridis HA4-1, Hyperamoeba (version of June 2007) (Jobb 2007), with the same substitution sp. Hpl, and Didymium sp. OX13PS. Didymium squamulosum model and four rate categories. A thousand non-parametric boot- AM231293, two other strains of Hyperamoeba, E2/3 and E3P, and strap replicates were run, using a BIONJ distance-based starting crustacea are in poorly resolved positions. The second tree. Bayesian search of tree space used MrBayes, version 3.1.1 group, the ‘‘Diderma’’ clade, comprises Didymium anellus,two (Huelsenbeck and Ronquist 2003), with the GTR1I1gamma Diderma isolates, and Lepidoderma tigrinum (maximum support model of substitution, the gamma distribution being approximated in all analyses) and is paraphyletic to the Physaridae (Fig. 2). by eight categories. (a 5 0.307, proportion of invariable sites Physaridae is split into two clades, one comprising Badhamia pa- 5 0.067). Two runs starting from different random trees were nicea var. nivalis and Hyperamoeba sp. G1a and W2i (BI/ML/ performed and sampled every 10 generations, with eight simulta- NJ—1.0/88/89) and the second with Physarum spp., spp., neous chains, for 2 million generations. Convergence of the two and Hyperamoeba flagellata (BI/ML/NJ—0.71/-/-) (Fig. 2). runs, evaluated by a standard deviation of split frequencies Short Physarida trees. The unrooted NJ trees are composed o0.01, was reached after 253,000 generations, and the trees of often well-resolved terminal clades, with an overall topology obtained beforehand were discarded. The remaining 174,700 congruent with the complete SSU tree, and unresolved basal trees were assembled in a consensus tree (log likelihood 5 branches, as expected from the characteristics of these alignments. 15,951.15). In Didymiidae, Lepidoderma carestianum (with the morphologi- Short Physarida alignments. Distance methods were used cally similar Lepidoderma granuliferum) is in an outlying position to check the allocation of the physarid hyperamoebae. Using the (Fig. 3). Diderma darjeelingense is separate from other Diderma. PHYLIP package version 3.6 (Felsenstein 2004), a Neighbor Join- The remaining Diderma would be monophyletic if L. tigrinum ing bootstrap was conducted as follows: 1,000 replicates of the were included in it. The placement of D. anellus AM231292— alignment were generated using SEQBOOT. Distance matrices isolate provided by Jim Clark (Wikmark et al. 2007)—in the were calculated using DNADIST, using the F84 model, with tran- Diderma clade, in both trees, is intriguing and deserves better sitions/transversion rates set, and gamma distributed rates across investigation. The splitting of Didymium into two groups is retrie- sites, both parameters as calculated by Modeltest 3.7 (Posada and ved, and is even more evident: ‘‘Didymium 1,’’ paraphyletic in Crandall 1998). From these matrices, Neighbor Joining calculated Fig. 2, is here holophyletic. Most importantly, the species common 192 J. EUKARYOT. MICROBIOL., 57, NO. 2, MARCH–APRIL 2010

Echinostelium arboreum AY842030 Semimorula liquescens EU544544 Echinostelium coelocephalum AY842033 84; 75 Echinostelium minutum AY842034 Lamproderma fuscatum DQ903668 “Meriderma” Lamproderma atrosporum var. retisporum DQ903671 Amaurochaete comata AY842031 74; 72 Comatricha sinuatocolumellata DQ903684 0.79; -; Comatricha nigra DQ903683 Comatricha 54 group 61; 89 Comatricha nigricapillitia 1 DQ903685 Comatricha nigricapillitia 2 DQ903686 Enerthenema papillatum AY643823 Macbrideola oblonga DQ903682 Symphitocarpus impexus AY230188 93;98 AY145528 93;95 Hyperamoeba sp. B1/2 AY321109 Stemonitis flavogenita AF239229 Stemonitis 98; 100 87; 94 99; 83 Hyperamoeba sp. ATCC50750 AF093247 Hyperamoeba sp. A1K AY321107 Hyperamoeba sp. BuP AY321110 STEMONITIDA Hyperamoeba sp. AH1 AY321108 Lamproderma ovoideum s.str. DQ903675 94;97 Lamproderma Lamproderma sauteri DQ903674 Lepidoderma carestianum AM231296 99; 100 Didymium dubium K15 AM231295 Protophysarum phloiogenum 0.98; 78; 57 AY230189 63; - Mucilago crustacea DQ903679 PHYSARIDA Hyperamoeba sp. E2/8 AY321111 99; 98 Hyperamoeba sp. E3P AY321112 ‘‘Didymium’’ 1 81; 74 Didymium squamulosum AM231293 0.77; -; - Didymium sp. OX13PS GQ249857 97; 99 Hyperamoeba sp. Hpl AY321114 0.88; -; - Didymium iridis HA4-1 AJ938149 0.82; 90; 97 50; - Didymiidaey 98; 97 Didymium iridis CUR1-4CUR1 4 AJ938150 99; 100 Didymium “nigripes” AF239230 0.94; -; - Didymium iridis CR19-1 AJ938151 Pseudodidymium 0.56; -; - cryptomastigophorum AY207466 ‘‘Didymium’’ 2 69;71 Didymium iridis CR8-1 AJ938154 95; 93 Hyperamoeba dachnaya AY062881 0.98; 97; 100 Didymium sp. ECH1/I EF118758 Didymium spsp. ECH49/I EF118761 Didymium anellus AM231292 Diderma globosum var. europaeum DQ903677 98; 99 ‘‘Diderma’’ Lepidoderma tigrinum DQ903678 0.55; -; - Diderma niveum AM231291 0.79; 54; 52 Hyperamoeba sp. G1a AY321113 88; 89 Hyperamoeba sp. W2i AY321115 92; 95 var. nivalis DQ903680 Physarum album DQ903681 Physaridae 0.71; -; - Fuligo leviderma DQ903676 87; 89 AJ584697 0.58; -; - Physarum polycephalum X13160 0.1 0.92; -; - Hyperamoeba flagellata AF411289 Physarum didermoides AY183449 Fig. 2. Small subunit rRNA tree using 1,451 nucleotide positions for 54 taxa and derived using Bayesian inference. Dots correspond to a Bayesian posterior probability (BPP) of 1.0% and 100% bootstrap support for two separate maximum likelihood analyses. Thick branches correspond to 1.0 BPP and the specified values of bootstrap replicates reported on the best maximum likelihood tree and bootstrap starting with a NJ tree; hyphens indicate bootstrap values o50%. The scale bar shows the fraction of substitutions per site. GenBank accession numbers follow the species names. FIORE-DONNO ET AL.—INVALIDATION OF HYPERAMOEBA 193 to both trees are found in the same groups, showing that overall Lepidoderma granuliferum AB259444 Lepidoderma topology is not affected by the addition of numerous taxa. In Lepidoderma carestianum AM231296 100 Lepidoderma carestianum AB259440-3 pro parte ‘‘Didymium 1,’’ the position of the Hyperamoeba strains E2/8 and Diderma darjeelingense AB259366 E3P is not better resolved in this tree (cf. Fig. 2, 3), while Hyper- D. rugosumugosu AB259377-8593 8 amoeba sp. Hpl and Didymium sp. OX13PS are still associated Diderma aurantiacum AB435322 with Didymium iridis HA4-1, but also with three strains of D. floriforme var. subfloriforme AB435323 Diderma niveum AM231291 D. squamulosum (i.e. AB259430,35 and AB259438 from Japan, D. globosum var. europaeum DQ903677 and AM231293 from Costa Rica). In ‘‘Didymium 2,’’ the close Diderma umbilicatum AB259386 ‘‘D Lepidoderma tigrinum DQ903678 i relationship between Didymium sp. EC1/l and ECH49/1 and 100Lepidoderma tigrinum AB259445-6 92 d H. dachnaya is retrieved, with the addition of Didymium bahiense Diderma radiatum AB259375-6 (90% bootstrap). The clade including P. cryptomastigophorum Diderma alpinum AB435320-1, AB249364 e r with D. iridis CR19-1 is recovered (Fig. 3). 52 Didymium anellus AM231292 Diderma hemisphaericum AB259372-3 m In the Physaridae-only tree (Fig. 4), H. flagellata is sister to Diderma microsporum AB259374 a’’ three strains of Physarum didermoides (bootstrap support: 99%). Diderma deplanatum AB259367-8 Hyperamoeba G1a and W2i are in a non-supported clade includ- Diderma simplex var. applanatum AB259382 Diderma saundersii AB259379 ing Physarum reniforme, Physarum pusillum, Badhamia alpina, 84 76 Diderma saundersii AB259380-1 and B. panicea var. nivalis (Fig. 4). 52 Diderma effusum AB259369-71 68 Diderma testaceum AB259385 Didymium clavus AB259390, 92 56 Didymium clavus AB259389, 91 Didymium dictyopodium AB259397-8 DISCUSSION 51 Didymium squamulosum AB259436 77 99Didymium squamulosum AB259432,34 Our phylogenetic analyses show that there are at least two dis- Didymium squamulosumAB259431,33,37 tinct clades of Hyperamoeba, as previously shown but with fewer DidDidymium i crustaceum t AB259395- 6 taxa (Dykova´ et al. 2007; Walochnik et al. 2004). We can now 87 Mucilago crustacea DQ903679 ‘‘D 99 Muculago crustacea AB259447 i affirm that one belongs to the order Stemonitida and the other to Hyperamoeba sp. E2/8 AY321111 100 d the order Physarida. Hyperamoeba sp. E3P AY321112 y Didymium dubium AB259399 Renaming Hyperamoeba strains. Physarid hyperamoebae 100 m Didymium dubium K15 AM231295 comprise five distinct groups, two in the family Physaridae 50 D. comatum AB259393-4 i (H. flagellata, Hyperamoeba sp.G1a,andHyperamoeba sp. W2i), 97 Protophysarum u phloiogenum m and three within Didymiidae. Two of the Physaridae hyperamoe- AY230189 bae are close to B. panicea var. alpina (i.e. Hyperamoeba sp. G1a Didymium squamulosum AB259438 64 Didymium squamulosum AB259430,35 1’’ and W2i), while H. flagellata is most closely related to P. did- Didymium squamulosum AM231293 ermoides. In the Physaridae-only tree, H. flagellata is associated Hyperamoeba sp. Hpl AY321114 90 Didymium sp. OX13PS GQ249857 with two more strains of P. didermoides (bootstrap support: 99%). 92 This clade is not disrupted by adding numerous Physaridae. Given Didymium iridis HA4-1 AJ938149 61 Didymium flexuosum AB259400-1 the robust genetic closeness to P. didermoides we consider that the Didymium serpula AB259429 ‘‘absence’’ of fruiting bodies is insufficient reason for maintain- Didymium leoninum AB259411-2 Didymium panniforme AB259428 ing Hyperamoeba as a distinct genus. We therefore assign H. fla- Didymium floccoides AB259402-4 gellata to the genus Physarum as Physarum flagellatum 97Didymium marineri AB259413 (Alexeieff) Fiore-Donno, Kamono & Cavalier-Smith 2009 comb. Didymium megalosporum AB259414-8 Didymium iridis AB259409 nov. (Table 1). This act makes Hyperamoeba an unnecessary D. nigripes junior of Physarum, which means that the name Hyper- 93 AB259424,7 amoeba cannot now be validly applied to any of the Hyperamoeba Didymium nigripes 100 AB259425 strains that do not group within the genus Physarum on well- Didymium iridis resolved phylogenetic trees. This assumes that this strain was AB259407 correctly identified as H. flagellata, the type species of the genus 89 Didymium iridis 100 AB259408 Hyperamoeba (i.e. is more similar to Alexeieff’s strain). Because Didymium nigripes of past confusion over the of strains identified as AB259426 Didymium laccatipes AB259410 Hyperamoeba and the unavailability of any type specimen or type 90 Didymium floccosum AB259405-6 65 culture to resolve this confusion, we designate the strain ATCC Didymium melanospermum AB259419-20 68 50637, isolated by Mylnikov from a freshwater pond, at Yaro- Didymium minus AB259421-33 Didymium iridis CR8-1 AJ938154 ‘‘D slavl, Russia in 1993, as the neotype and hapantotype of P. fla- Didymium bahiense AB259388 90 i gellatum, in accordance with article 75.3 of the International Code Didymiumsp. ECH49/I EF118761 d for Zoological Nomenclature. Because H. flagellata was origi- 51 Didymium sp. ECH1/I EF118758 Hyperamoeba dachnaya AY062881 y nally described only crudely by light microscopy and no holotype Didymium iridis CR19-1 AJ938151 m ever existed (Alexeieff 1923), we cannot be sure that strain ATCC 92Pseudodidymium cryptomastigophorum i AY207466 u 50637 better represents the species. We chose this isolate as neo- Didymium bahiense AB259387 type primarily because it was the first to be studied by electron 70 m 88 Didymium iridis CUR1-4 AJ938150 microscopy (Karpov and Mylnikov 1997) and was among the 79 Didymium ‘‘nigripes’’ AF239230 Didymium verrucosporum AB259439 2’’ earliest by SSU sequencing (Cavalier-Smith et al. 2004), and be- 0.1 cause it has been longer in a major public culture collection than any other, and thus should reliably continue to be available to re- Fig. 3. Unrooted Neighbor Joining tree based on 417 positions of the searchers. Because of the close morphological similarities among small subunit rRNA gene of 72 taxa of the family Didymiidae (Myxoga- Hyperamoeba strains, this strain can be clearly distinguished from stria). The results of 1,000 bootstrap replicates are shown as percentages, if all others only by its unique SSU sequence (AF411289 Cavalier- 450%. The scale bar shows the fraction of substitutions per site. Taxa also present in Fig. 2 are in bold. Smith et al. 2004), but this is sufficient to recognize other strains 194 J. EUKARYOT. MICROBIOL., 57, NO. 2, MARCH–APRIL 2010

Diachea leucopodia AB259360-1 that might be isolated in future unambiguously as P.(5 Hyper- subsessilis AB259363 amoeba) flagellatum. Craterium dictyosporum AB259462 Craterium dictyosporum AB259463 Hyperamoeba sp. strains G1a and W2i are still associated with Diachea radiata AB259362 Badhamia (B. panicea var. nivalis and B. alpina), but are closer to Physarum crateriforme AB259503 P. reniforme (group not supported). Delimitation of the genera Physarum stellatum AB259545 Badhamia and Physarum has been considered to be artificial Physarum digitatum AB259506 60 76 (Neubert, Nowotny, and Baumann 1995): there are badhamioid Physarum confertum AB259500 60 Physarum lateritium AB259514 Physarum and physaroid Badhamia. Therefore, it is not surprising P. pulcherrimum AB259536 to find those two probably polyphyletic genera intermingled in our 59 Fuligo candida AB259480-1 tree. These tentative inferences need testing by complete SSU se- Fuligo candida AB435344 quencing of more taxa, since, if correct, generic distinctions 85 Fuligo septica f. flava AB259483-4 99 Fuligo aurea AB259478-9 within Physaridae would need extensive revision. Because of 97 Fuligo septica AB259482 these uncertainties Hyperamoeba sp. G1a and W2i cannot be as- Badhamia panicea AB259454AB259454-5 5 58 signed to either Physarum or Badhamia, so we designate them Physarum nigripodum AB259527 Badhamia affinis AB259448 simply ‘‘undetermined Physaridae’’ (Table 1). 74 Badhamia affinis AB435341-2 Didymiidae-related Hyperamoeba, including the Didymium-re- Physarum album AB259531 lated strains found in sea urchins, P. cryptomastigophorum, and 100 Physarum album AB259532 our new strain, all belong to a well-supported clade grouping Badhamia macrocarpa AB259452-3 100Badhamia macrocarpa AB259451 M. crustacea, D. squamulosum, D. nigripes, and all strains of Physarum viride AB259551 D. iridis. The fact that all these hyperamoebae are more closely Physarum viride AB435352-4 related to the genus Didymium than to Diderma and Lepidoderma 83 P. tenerum AB259550 is confirmed in the Didymiidae-only tree, but Didymium species Physarum flavicomum AB259507 P. flavicomum AB259508 are mixed with two other genera, Protophysarum (a monospecific Hyperamoeba flagellata AF411289 genus) and Mucilago. The latter is associated with Didymium 99 Physarum didermoides AB259505 crustaceum; in both species, sporophores are covered with a crust 99 Physarum didermoides AY183449 86 Physarum didermoides AB259504 of lime crystals, but Mucilago is a quite large (i.e. a few centi- Physarum psittacinum f. fulvum AB259535 meters) conglomerate of fused sporophores. Didymium exhibits Craterium aureum AB259459-61 intricate internal relationships. The Hyperamoeba sp. Hpl and 78 C. leucocephalum var. cylindricum AB259464-9 Didymium D. iridis 51 sp. OX13PS are closely related to HA4-1 from C. leucocephalum var. scyphoides AB259470-2 Haiti. Strain HA4-1 of D. iridis does not cluster with the other Craterium minutum AB259473-5 strains of the same species, but it is closer to D. squamulosum 68 Physarum roseum AB259543-4 AM231293 as shown elsewhere (Wikmark et al. 2007), and also to Craterium reticulatum AB259476-7 two Japanese strains of D. squamulosum. Because of the genetic Physarum florigerum AB259509-10 70 Physarum nucleatum AB259528-30 diversity shown by D. iridis and D. squamulosum (i.e. three other fragilis var. bisporus AB259485-6 Japanese D. squamulosum strains group with D. dictyopodium), it AB259498-9 88 is not possible to allocate Hyperamoeba Hpl and Didymium sp. Physarum luteolum AB259518-9 OX13PS either to D. squamulosum or to D. iridis. These two Physarum bivalve AB259488-9, 91,94 92 Physarum bivalve AB259490, 92, 93 nominal species are common, widespread, and very variable, and Physarum loratum AB259515-7 probably composed of distinct biological species. The study of the 61 Physarum mutabile AB259525-6 reproductive system of a considerable number of strains of D. Physarum pusillum AB259537 squamulosum and D. iridis has revealed an intricate pattern of Badhamia alpina AB259449 92 Badhamia panicea var. nivalis DQ90680 core sexual species, surrounded by a swarm of asexually repro- Hyperamoeba sp. G1a AY321113 ducing clones that vary from small localized patches to very ex- Hyperamoeba sp. W2i AY321115 tensive populations that differ both morphologically and 51Physarum reniforme AB259539 Physarum reniforme AB259538,40 genetically (see Clark 1995 and references therein; ElHage, Lit- Fuligo septica AJ584697 tle, and Clark 2000). The polyphyly of D. squamulosum and D. Fuligo leviderma DQ903676 iridis in our tree is consistent with each comprising several cryptic Physarum rigidum AB259541-2 83 biological species. To that may be added the existence of very AB259456-8 similar morphospecies, making the distinction between D. iridis, Physarum polycephalum X13160 52 Physarum albescens AB259487 D. melanospermum, D. megalosporum, D. bahiense, D. ve- Physarum bogoriense AB259495-7 rrucosporum, and D. nigripes not an easy task. Didymium nigripes Physarum hongkongense AB259511 AF239230 sequenced by Horton and Landweber (2000) may Physarum lakhanpalii AB259513 99 actually be ‘‘D. iridis’’ (Clark and Collins 1976). To resolve Physarumyp lakhanpalii AB259512 Physarum melleum AB259521-2 this confused taxonomy, a comprehensive monograph of 74 Physarum melleum AB435350 Didymium and allied genera is highly desirable. Meanwhile, we 99 Physarum melleum AB259520 consider it appropriate to allocate Hyperamoeba strains E2/8, Physarum melleum AB259523 58 Physarum mortonii AB259524 E3P, Hpl, H. dachnaya, and P. cryptomastigophorum to the ge- Physarum superbum AB259547-9 nus Didymium (Table 1). H. dachnaya Walker et al. thus becomes P. sulphureum AB259546 Didymium dachnayum (Walker et al.) Fiore-Donno, Kamono & Physarum plicatum AB259533-4 0.1 Cavalier-Smith 2009, comb. nov. and P. cryptomastigophorum 71 P. conglomeratum AB259501-2 Michel et al. becomes Didymium cryptomastigophorum (Michel Fig. 4. Unrooted Neighbor Joining tree based on 417 positions of the et al.) Fiore-Donno, Kamono & Cavalier-Smith 2009, comb. nov. small subunit rRNA gene of 75 taxa of the family Physaridae (Myxoga- (Table 1). stria). The results of 1,000 bootstrap replicates are shown as percentages, if Our trees suggest that there are probably at least four or five 450%. The scale bar shows the fraction of substitutions per site. Taxa also different species called D. iridis, and in the absence of sequences present in Fig. 2 are in bold. for a type specimen or culture there is no way of knowing which, FIORE-DONNO ET AL.—INVALIDATION OF HYPERAMOEBA 195 if any, of those on our trees deserve to retain that name. Therefore patible with all hyperamoebae nesting into the dark-spored clade although the sequence of P. cryptomastigophorum is very similar of Myxogastria. Nevertheless, it is still unclear whether the ca- to that of D. iridis CR8-1 (AJ938154), we cannot synonymize the pacity to sporulate has actually been lost or is simply not ob- two species as there is no reason to think that Strain CR8-1 in served. All reported attempts to make hyperamoebae strains fruit particular is the true D. iridis. have failed, although amoebae were transferred from liquid cul- Stemonitid Hyperamoeba all belong to the Stemonitis clade, tures to agar plates (Dykova´ et al. 2007; Karpov and Mylnikov Hyperamoeba sp. B1/2 and Hyperamoeba sp. ATCC50750 group- 1997; Michel et al. 2003). However, fruiting is a sexual process ing with maximum support in all analyses with Stemonitis flavoge- that involves fusion of two haploid, sexually compatible and thus, nita. We therefore name these two strains S. aff. flavogenita genetically diverse, amoeboflagellates in a diploid zygote which, (Table 1), not S. flavogenita itself in case further study reveals a under appropriate conditions, can differentiate into fruiting bodies closer relative. Strains A1K, BuP, and AH1 constitute a robust (Collins 1979). Many myxogastrid strains are heterothallic, and homogeneous clade distinct from S. flavogenita, but are still cannot form fruiting bodies in clonal cultures because of the ab- within the clade comprising S. flavogenita and S. axifera. With sence of the compatible type. Since all hyperamoeba reasonable confidence we can therefore allocate these three strains strains were single clones, originated from a single cell or cyst to the genus Stemonitis (Table 1), as Stemonitis sp., but not to any (Dykova´ et al. 2007; Karpov and Mylnikov 1997; Michel et al. of the two species represented here, as the genus Stemonitis com- 2003; Walochnik et al. 2004) or by serial dilution (Walker et al. prises 16 species (Lado 2001). 2003) or frequent subcultures (Zaman and Adoutte 1999), the Why are Hyperamoeba SSU rRNA gene sequences devoid of simplest explanation is that they are clones of heterothallic myxo- Group I introns? Group I introns are very common in Myxoga- gastrian species, many of which may have already been described stria (Fiore-Donno et al. 2008), being found in nine insertion sites from their fruiting bodies. However, the existence of several self- (Lundblad et al. 2004) always located in very conserved parts of mating myxogastrian strains, apparently derived from the sexual the gene (Haugen et al. 2003; Wikmark et al. 2007). They are isolates by suppression of during formation (Collins mobile or inferred to be mobile and introns at the same position et al. 1983), suggests that at least some hyperamoebae may have in diverse organisms are more related to each other than introns in lost the ability to mate and genuinely lack fruiting bodies. To dis- the same organism, suggesting lateral transfer (Lundblad et al. tinguish between these possibilities a large number of cultures of 2004)—but there are cases of vertically inherited introns (Lundb- hyperamoebae and their closest fruiting relatives must be obtained lad et al. 2004). Introns can represent up to 70% of the whole and experimental crosses carried out, probably revealing distinct sequence, as in the 7,257-nucleotide sequence of L. tigrinum patterns in different strains. Also, searching for meiosis-linked DQ903678 (Fiore-Donno et al. 2008). How is it that hyperamoe- genes, as defined by Ramesh, Malik, and Logsdon (2005) and two bae, P. cryptomastigophorum, the marine Didymium, and our of which are found in P. polycephalum (Watkins and Gray 2008), Didymium OX13PS are totally devoid of introns? As these taxa might help clarify whether they are still sexual. Interestingly, none are not phylogenetically related and come from diverse environ- of these genes has been found to date in the ESTs of Didymium ments, the hypothesis of a common factor inducing the loss of ( 5 Hyperamoeba) dachnayum and Stemonitis aff. flavogenita introns seems unlikely. Another hypothesis is that, as intron-con- ( 5 Hyperamoeba sp. ATCC50570). Should this fact be confirmed taining sequences are more difficult to obtain by PCR, only or- with data from complete , then it will be certain that these ganisms without introns get sequenced. For example, the SSU of Myxogastria have irreversibly lost the sexual part of their life-cycle. marine Didymium-like amoebae and of our Didymium OX31PS Ubiquity of the genus Didymium. Because of their peculiar were amplified in one step, with primers at the 50- and 30-ends, sporophores, Myxogastria are the only for which exten- which would not work for very large amplicons. Also, because sive, worldwide inventories have been made, proving that they are myxogastrian ribosomal sequences are so variable between taxa, common and widespread in nearly every terrestrial ecosystem only very conserved regions are selected to design primers and (Stephenson, Schnittler, and Novozhilov 2008). About 875 spe- these are often intron insertions sites. For example, all internal cies have been described (Lado 2001), based on characteristics of primers used by Walochnik et al. (2004) are designed across int- the sporophore. The most conspicuous group is the dark-spored ron insertions sites (e.g. P1fw/rev—across intron S516; P2fw/ clade comprising the orders Physarida and Stemonitida, and to- rev—S788; P3fw/rev—S1199). Consequently, any organism pos- talling 68% of the described species. In contrast, little is known of sessing these introns would not be sequenced. their distribution in soil, based mostly on a single set of studies Pseudodydymium cryptomastigophorum was described as a new conducted in United Kingdom in the 1980s (Feest and Madelin species and genus (Michel et al. 2003) mainly because it lacked a 1988a, b). Didymium, a genus of 76 species (Lado 2001), appeared 654-bp insert found in the only Didymium sequence available be- to be dominant in the studied soils (Feest and Madelin 1985). Re- fore 2003 (i.e. D.‘‘nigripes’’ AF239230). This insert corresponds cently, molecular environmental sampling focusing on the dark- to intron S1389, which has been investigated since then in the fam- spored Myxogastria also found Didymium in decaying wood and ily Didymiidae (Wikmark et al. 2007). This intron is unusual as it forest floor leaf litter in Thailand (Win Ko Ko et al. 2009), as 50% has lost the ability to self-splice, being removed by host enzymes of the species detected in different kinds of soils in Japan instead, making it vertically inherited in the family, though it is (Kamono et al. 2009a), and as in the air (Kamono et al. sometimes lost (e.g. in four stains of D. iridis). In light of these new 2009b). Didymium hyperamoebae have indeed also been found in findings, the genus Pseudodidymium is no longer justifiable, but the freshwater habitats where myxogastrid specialists would never species name should be retained until the probably polyphyletic look: under the ice of a frozen lake, in physiotherapy baths, and in D. iridis is properly revised and divided into separate biological drinking water treatment plants (Table 1 and references therein). species; D. cryptomastigophorum comb. nov. might even turn out Perhaps the most striking discovery is that of sea-inhabiting Did- to be the valid name for D. iridis CR8-1 AJ938154. ymium in the coelomic cavity of sea urchins (Dykova´ et al. 2007). Why do strains described as Hyperamoeba not form fruiting These discoveries suggest that Myxogastria may be much more bodies? The chief character separating Hyperamoeba strains widespread than previously thought, and that systematic searches from Myxogastria is the absence of fruiting bodies in cultures. for them in aquatic environments by amplifying and sequencing Earlier suggestions that Hyperamoeba could be the ancestral phe- environmental DNA might be illuminating. Past studies using notype directly derived from a non-fruiting ancestor (Cavalier- -wide primers would generally not have found them be- Smith and Chao 1998; Karpov and Mylnikov 1997) are incom- cause of the exceptional length and divergence of their rRNAs. 196 J. EUKARYOT. MICROBIOL., 57, NO. 2, MARCH–APRIL 2010

ACKNOWLEDGMENTS Kamono, A. & Fukui, M. 2006. Rapid PCR-based method for detection and differentiation of and (Myxomycetes) in This work was supported by the Leverhulme Trust research grant environmental samples. J. Microbiol. Methods, 67:496–506. R1008101 and by a Royal Society International Joint Project. For Kamono, A., Matsumoto, J., Kojima, H. & Fukui, M. 2009a. Character- critical reading of the manuscript, we thank David Mitchell and ization of myxomycete communities in soil by reverse transcription Alexander Kudryavtsev. We are grateful to the authors of the pre- polymerase chain reaction (RT-PCR)-based method. Soil Biol. Bioc- vious studies on Hyperamoeba for their accurate descriptions. hem., 41:1324–1330. Kamono, A., Kojima, H., Matsumoto, J., Kawamura, K. & Fukui, M. 2009b. Airborne myxomycete spores: detection using molecular tech- LITERATURE CITED niques. Naturwissenschaften, 96:147–151. Alexeieff, A. G. 1923. Hyperamoeba flagellata n. gen. n. sp. Arch. Russ. Karpov, S. A. & Mylnikov, A. P. 1997. Ultrastructure of the colourless Protistolog. Obs., 2:280–288. (in Russian). flagellate Hyperamoeba flagellata with specifical reference to the flag- Cavalier-Smith, T. & Chao, E. E.-Y. 1998. Hyperamoeba rRNA phylog- ellar apparatus. Eur. J. Protistol., 33:349–355. eny and the classification of the phylum Amoebozoa. The Society of Karpov, S. A., Novozhilov, Y. & Chistiakova, L. V. 2003. A comparative Protozoologists, 51st Annual Meeting, Flagstaff, AZ. Available at http: study of zoospore cytoskeleton in impexus, //www.uga.edu/protozoa/meetings/secabs/1998.html (accessed Novem- cinerea and epidendrum (Eumycetozoa). Protistology, 3:15– ber 15, 2009). 29. Cavalier-Smith, T., Chao, E. E.-Y. & Oates, B. 2004. Molecular phylog- Lado, C. 2001. Nomenmyx. A nomenclatural taxabase of Myxomycetes. eny of Amoebozoa and the evolutionary significance of the unikont Cuadernos de Trabajo Flora Micolo´gica Ibe´rica, 16:1–221. . Eur. J. Protistol., 40:21–48. Lanave, C., Preparata, G., Saccone, C. & Serio, G. 1984. A new method Clark, J. 1995. Myxomycete reproductive systems: additional information. for calculating evolutionary substitution rates. J. Mol. Evol., 20:86– Mycologia, 87:779–786. 93. Clark, J. & Collins, O. N. R. 1976. Studies on the mating systems of eleven Lundblad, E. W., Einvik, C., Ronning, S., Haugli, K. & Johansen, S. 2004. species of Myxomycetes. Am. J. Bot., 63:783–789. Twelve group I introns in the same pre-rRNA transcript of the my- Collins, O. N. R. 1979. Myxomycete biosystematics: some recent devel- xomycete Fuligo septica: RNA processing and evolution. Mol. Biol. opments and future research opportunities. Bot. Rev., 45:145–201. Evol., 21:1283–1293. Collins, O. N. R., Gong, T., Clark, J. & Tang, H.-C. 1983. Apomixis and Michel, R., Walochnik, J. & Aspo¨ck, H. 2003. Pseudodidymium crypt- heterothallism in Stemonitis flavogenita (Myxomycetes, Stemonitales). omastigophorum nov. gen., nov. sp., a Hyperamoeba or a slime mould? Mycologia, 75:614–622. A combined study on morphology and 18S rDNA sequence data. Acta Dykova´, I., Lom, J., Dvorˇa´kova´, H., Peckova´, H. & Fiala, I. 2007. Did- Protozool., 42:331–343. ymium-like myxogastrids (class Mycetozoa) as endocommensals of sea Nelson, R. K. & Scheetz, R. W. 1975. Swarm cell ultrastructure in urchins (Sphaerechinus granularis). Folia Parasitol., 54:1–12. Ceratiomyxa fruticulosa. Mycologia, 67:733–740. ElHage, N., Little, C. & Clark, J. 2000. Biosystematics of the Didymium Neubert, H., Nowotny, W. & Baumann, K. 1995. Die Myxomyceten squamulosum complex. Mycologia, 92:54–64. Deutschlands und des angrenzenden Alpenraumes. Band 2: . Feest, A. & Madelin, M. F. 1985. Numerical abundance of myxomycetes K. Baumann Verlag, Gomaringen (D), 368 pp. (myxogastrids) in soils in the West of England. FEMS Microbiol. Ecol., Posada, D. & Crandall, K. A. 1998. MODELTEST: testing the model of 31:353–360. DNA substitution. Bioinformatics, 14:817–818. Feest, A. & Madelin, M. F. 1988a. Seasonal population change of my- Poulain, M., Meyer, M. & Bozonnet, J. 2002. Lamproderma nivicoles du xomycetes and associated organisms in four woodland soils. FEMS groupe atrosporum Meylan. Scipta Botanica Belgica, 22:74–75. Microbiol. Ecol., 53:133–140. Ramesh, M. A., Malik, S.-B. & Logsdon, J. M. J. 2005. A phylogenomic Feest, A. & Madelin, M. F. 1988b. Seasonal population changes of my- inventory of meiotic genes: evidence for sex in Giardia and an early xomycetes and associated organisms in five non-woodland soils, and eukaryotic origin of meiosis. Curr. Biol., 15:185–191. correlations between their numbers and soil characteristics. FEMS Mi- Rodriguez, F., Oliver, J. L., Marin, A. & Medina, J. R. 1990. The general crobiol. Ecol., 53:141–152. stochastic model of nucleotide substitution. J. Theor. Biol., 142:485– Felsenstein, J. 2004. PHYLIP (Phylogeny Inference Package) version 3.6. 501. Distributed by the author. Department of Sciences, University Stamatakis, A., Ludwig, T. & Meier, H. 2005. RAxML-III: a fast program of Washington, Seattle. for maximum likelihood-based inference of large phylogenetic trees. Fiore-Donno, A. M., Haskins, E. F., Pawlowski, J. & Cavalier-Smith, T. Bioinformatics, 21:456–463. 2009. Semimorula liquescens is a modified echinostelid myxomycete Stephenson, S. L., Schnittler, M. & Novozhilov, Y. 2008. Myxomycete (Mycetozoa). Mycologia, 101:773–776. diversity and distribution from the record to the present. Biodiv. Fiore-Donno, A.-M., Meyer, M., Baldauf, S. L. & Pawlowski, J. 2008. Conserv., 17:285–301. Evolution of dark-spored Myxomycetes (slime-molds): molecules ver- Walker, G., Silberman, J. D., Karpov, S. A., Preisfeld, A., Foster, P., Fro- sus morphology. Mol. Phyl. Evol., 46:878–889. low, A. O., Novozhilov, Y. & Sogin, M. L. 2003. An ultrastructural and Fiore-Donno, A.-M., Nikolaev, S. I., Nelson, M., Pawlowski, J., Cavalier- molecular study of Hyperamoeba dachnaya, n. sp., and its relationship Smith, T. & Baldauf, S. L. 2010. Deep phylogeny and evolution of to the mycetozoan slime moulds. Eur. J. Protistol., 39:319–336. slime moulds (Mycetozoa). , 161:55–70. Walochnick, J., Michel, R. & Aspo¨ck, H. 2004. A molecular biological Hall, T. A. 1999. BioEdit: a user-friendly biological sequence alignment approach to the phylogenetic position of the genus Hyperamoeba. J. editor and analysis program for Windows 95/98/NT. Nucl. Acids Symp., Eukaryot. Microbiol., 51:433–440. 41:95–98. Watkins, R. F. & Gray, M. W. 2008. Sampling gene diversity across the Haugen, P., Coucheron, D. H., Ronning, S. B., Haugli, K. & Johansen, S. supergroup Amoebozoa: large EST data sets from caste- 2003. The molecular evolution and structural organization of self-splic- llanii, vermiformis, Physarum polycephalum, Hyper- ing group I introns at position 516 in nuclear SSU rDNA of My- amoeba dachnaya and Hyperamoeba sp. Protist, 159:269–281. xomycetes. J. Eukaryot. Microbiol., 50:283–292. Wikmark, O.-G., Haugen, P., Lundblad, E. W., Haugli, K. & Johansen, S. Horton, T. L. & Landweber, L. F. 2000. Evolution of four types of RNA 2007. The molecular evolution and structural organization of Group I editing in myxomycetes. RNA, 6:1339–1346. introns at position 1389 in nuclear small subunit rDNA of My- Huelsenbeck, J. P. & Ronquist, F. 2003. Mr Bayes v3.0b4. Available at xomycetes. J. Eukaryot. Microbiol., 54:49–56. http://morphbank.ebc.uu.se/mrbayes/info.php Win Ko Ko, T., Stephenson, S. E., Jeewon, R. & Hyde, K. D. 2009. Mo- Jobb, G. 2007. TREEFINDER version of June 2007. Distributed by the lecular diversity of myxomycetes associated with decaying wood and author at http://www.treefinder.de, Munich, Germany. forest floor leaf litter. Mycologia, 101:592–598. Johansen, T., Johansen, S. & Haugli, F. B. 1988. Nucleotide sequence of Zaman, V. & Adoutte, A. 1999. Hyperamoeba isolated from human feces: Physarum polycephalum small subunit ribosomal RNA as inferred from description and phylogenetic affinity. Eur. J. Protistol., 35:197–207. the gene sequence: secondary structure and evolutionary implications. Curr. Genet., 14:265–273. Received: 07/22/09, 11/30/09; accepted: 12/2/09