sign in sign up
<<
Home , Acanthamoeba, Archamoebae, Balamuthia mandrillaris, Clydonella, Collodictyonidae, Dactylopodida

Article

Phylogenetic position of marina and the evolution of

NIKOLAEV, Sergey Igorievich, et al.

Abstract

Recent molecular phylogenetic studies have led to the erection of the Amoebozoa, uniting naked and testate lobose amoebae, the mycetozoan slime moulds and amitochondriate amoeboid (). Molecular data together with ultrastructural evidence have suggested a close relationship between and Archamoebae, classified together in the Conosea, which was named after the cone of microtubules that, when present, is characteristic of their kinetids. However, the relationships of conoseans to other amoebozoans remain unclear. Here, we obtained the complete small-subunit (SSU) rRNA gene sequence (2746 bp) of the enigmatic, multiflagellated Multicilia marina, which has formerly been classified either in a distinct phylum, Multiflagellata, or among lobose amoebae. Our study clearly shows that Multicilia marina belongs to the Amoebozoa. Phylogenetic analyses including 60 amoebozoan SSU rRNA gene sequences revealed that Multicilia marina branches at the base of the Conosea, together with another flagellated amoebozoan, solitarium, as well as with sp., Filamoeba nolandi [...]

Reference

NIKOLAEV, Sergey Igorievich, et al. Phylogenetic position of Multicilia marina and the evolution of Amoebozoa. International Journal of Systematic and Evolutionary Microbiology, 2006, vol. 56, no. 6, p. 1449-1458

DOI : 10.1099/ijs.0.63763-0

Available at: http://archive-ouverte.unige.ch/unige:121594

Disclaimer: layout of this document may differ from the published version.

1 / 1 International Journal of Systematic and Evolutionary Microbiology (2006), 56, 1449–1458 DOI 10.1099/ijs.0.63763-0

Phylogenetic position of Multicilia marina and the evolution of Amoebozoa

Sergey I. Nikolaev,1,231 Ce´dric Berney,241 Nikolai B. Petrov,1 Alexandre P. Mylnikov,3 Jose´ F. Fahrni2 and Jan Pawlowski2

Correspondence 1Department of Evolutionary Biochemistry, A. N. Belozersky Institute of Physico-Chemical Sergey Nikolaev Biology, Moscow State University, Moscow, Russia Sergey.Nikolaev@medecine. 2Department of Zoology and Biology, University of Geneva, Geneva, Switzerland unige.ch 3Institute of Biology of Inland Waters, Russian Academy of Sciences, Yaroslavskaya obl., Borok, Ce´dric Berney Russia [email protected]

Recent molecular phylogenetic studies have led to the erection of the phylum Amoebozoa, uniting naked and testate lobose amoebae, the mycetozoan slime moulds and amitochondriate amoeboid protists (Archamoebae). Molecular data together with ultrastructural evidence have suggested a close relationship between Mycetozoa and Archamoebae, classified together in the Conosea, which was named after the cone of microtubules that, when present, is characteristic of their kinetids. However, the relationships of conoseans to other amoebozoans remain unclear. Here, we obtained the complete small-subunit (SSU) rRNA gene sequence (2746 bp) of the enigmatic, multiflagellated protist Multicilia marina, which has formerly been classified either in a distinct phylum, Multiflagellata, or among lobose amoebae. Our study clearly shows that Multicilia marina belongs to the Amoebozoa. Phylogenetic analyses including 60 amoebozoan SSU rRNA gene sequences revealed that Multicilia marina branches at the base of the Conosea, together with another flagellated amoebozoan, Phalansterium solitarium, as well as with Gephyramoeba sp., Filamoeba nolandi and two unidentified amoebae. This is the first report showing strong support for a containing all flagellated amoebozoans and we discuss the position of the root of the phylum Amoebozoa in the light of this result.

INTRODUCTION single kinetosome surrounded by a conical microtubular sheath. Submembrane interkinetosomal fibres connect Multicilia marina is a multiflagellated amoeboid protist that each kinetosome to a neighbouring one and coordinate was first isolated in 1880 by L. Cienkowski (Cienkowski, the movement of the flagella. No microtubular connectives 1881) at the Solovetskaya station (Onega Bay, White Sea). between the basal apparatus and the nucleus have been Its cells are usually spherical, about 30–40 mm in diameter, observed (Mikrjukov & Mylnikov, 1996, 1998). The mito- although some of them can have an oblong or irregular chondrial cristae are tubular. Multicilia marina normally form. Typically, each cell has approximately 20–30 flagella. feeds by , using lobopodia for the capture of The basal apparatus of each flagellum is represented by a large prey (gymnamoebae).

3Present address: Department of Genetic Medicine and Development, Cienkowski (1881) described Multicilia as an intermediate Centre Me´dical Universitaire, 1 rue Michel-Servet, 1211 Geneva 4, stage between flagellates and heliozoans. Kudo (1954) con- Switzerland. sidered it to be a member of the Rhizomastigina 4Present address: Department of Zoology, University of Oxford, The Butschli, together with the Mastigamoebidae and two Tinbergen Building, South Parks Road, Oxford OX1 3PS, UK. groups of heliozoans, the actinophryids and the dimorphids. 1These authors contributed equally to this work. Later ultrastructural studies revealed that the unusually Abbreviations: BV, bootstrap value; GTR, general time-reversible; ML, weak and oscillating movements of the flagella of Multicilia maximum likelihood; PP, posterior probability; SSU, small subunit. marina (resulting in their superficial similarity to heliozoan The GenBank/EMBL/DDBJ accession numbers for the SSU rRNA axopodia), together with the poor coordination in their gene sequences of Multicilia marina and the unidentified activity and a lack of anterior and posterior ends, make clone Borok are AY268037 and AY626163, respectively. Multicilia marina distinct from all other known multi- A figure showing the predicted secondary structure of the small-subunit flagellated/ciliated protistan taxa, such as Caryoblastea, rRNA gene of Multicilia marina is available as supplementary material in Pseudociliata, , Ciliophora, Opalinata IJSEM Online. and Parabasalida. A new phylum, Multiflagellata, of

63763 G 2006 IUMS Printed in Great Britain 1449 S. I. Nikolaev and others uncertain affinities was proposed to include the vector system (Promega), transformed in XL-2 ultracompetent cells Multicilia (Mikrjukov & Mylnikov, 1996, 1998). In contrast, (Stratagene), sequenced using an ABI PRISM BigDye Terminator cycle Cavalier-Smith (1998) proposed that Multicilia belongs sequencing kit and analysed using an ABI-377 DNA sequencer (Perkin Elmer), all according to the manufacturers’ instructions. in the phylum Amoebozoa, uniting lobose amoebae, the mycetozoan slime moulds and the amitochondriate enta- In addition, we obtained the SSU rRNA gene sequence (as described moebids and pelobionts (Archamoebae), and that the genus above) of an unidentified (unidentified eukaryote clone is either related to Archamoebae (Cavalier-Smith, 2000) or Borok), which was a contaminant in a culture of the to (Cavalier-Smith, 2003; Cavalier-Smith et al., Clathrulina elegans isolated from waste treatment in Borok. 2004). In the second edition of the Illustrated Guide to the SSU rRNA gene phylogenetic analyses and secondary- , Multicilia was placed among ‘residual hetero- structure modelling. Complete SSU rRNA gene sequences of trophic flagellates’ (Patterson et al., 2000). Multicilia marina and the unidentified eukaryote clone Borok were aligned manually with sequences from diverse using the In spite of the rapid accumulation of sequences from amoe- Genetic Data Environment software (Larsen et al., 1993), following boid organisms during the past few years (Amaral Zettler the secondary-structure model proposed by Wuyts et al. (2001). A et al., 2000; Bolivar et al., 2001; Fahrni et al., 2003; Peglar first dataset, including the two sequences obtained in this study and et al., 2003; Cavalier-Smith et al., 2004; Nikolaev et al., 2004, sequences of 23 lobose amoebae, 25 and 50 , 2005), to date no molecular data have been reported for was used to infer the position of Multicilia marina among eukary- Multicilia. Previous molecular phylogenetic studies, based otes (1421 unambiguously aligned positions). More-detailed phylo- genetic relationships within the Amoebozoa were then assessed on actin and small-subunit (SSU) rRNA gene sequences, using an alignment of 64 sequences, including the two sequences indicated the monophyly of Amoebozoa, although with obtained in this study and 58 sequences of other amoebozoans very weak support (Bolivar et al., 2001; Fahrni et al., 2003; (including the more divergent mycetozoans, pelobionts and enta- Cavalier-Smith et al., 2004). Furthermore, several higher- moebids) and four opisthokonts as an outgroup (1290 unambigu- level taxonomic lineages have been distinguished within ously aligned positions). GenBank/EMBL/DDBJ accession numbers Amoebozoa (e.g. Smirnov et al., 2005). However, the for all sequences used in our analyses are indicated in Figs 1 and 2. relationships between and within these lineages remain For both datasets, a maximum-likelihood (ML) tree was obtained with unclear. It has been proposed that the ancestral amoebozoan PAUP* (Swofford, 1998), using the general time-reversible (GTR) may have been a flagellated organism similar to Phalans- model of evolution (Lanave et al., 1984; Rodriguez et al., 1990), taking terium or and that lobose amoebae evolved into account a proportion of invariable sites and a gamma distribution by multiple flagella losses (Cavalier-Smith et al., 2004). of the rates of substitution for the variable positions, with eight rate categories. All parameters were estimated from the dataset. To ascertain the amoebozoan affiliation of Multicilia Bayesian analyses were performed on the two datasets with MRBAYES and to investigate the relationships between flagellated (Huelsenbeck & Ronquist, 2001), using the same model. Two runs and non-flagellated members of the phylum, we obtained starting from different random trees were performed for 1 200 000 generations and sampled every 100 generations, with four simulta- the complete SSU rRNA gene of Multicilia marina and neous chains. The trees sampled before the chains reached stationarity performed phylogenetic analyses, including an exhaustive were discarded as a burn-in (2000). In addition, a ML bootstrap sampling of eukaryotes and all available sequences of analysis was performed for the second dataset with PHYML (Guindon Amoebozoa. We also constructed a predicted secondary & Gascuel, 2003). The predicted secondary structure of the SSU structure of the SSU rRNA molecule for Multicilia marina ribosomal molecule was modelled with MFOLD (Zuker et al., 1999) and and compared it with those of other eukaryotes. Our study visualized with RnaViz (De Rijk & De Wachter, 1997). confirmed that Multicilia is an amoebozoan and allowed us to propose a new hypothesis about the phylogenetic RESULTS relationships within the phylum Amoebozoa. Observations on the culture of Multicilia marina METHODS The culture of Multicilia marina was maintained in artificial Cell cultures and DNA extraction, amplification, cloning and Shmaltz-Pratt medium (NaCl, 16?07 %; KCl, 0?38 %; sequencing. The culture of Multicilia marina used was obtained MgCl2.6H2O, 3?15 %;MgSO4.7H2O, 3?95 %; CaCl2.H2O, from the culture collection of the Institute of Biology of Inland 0?83 %; KNO3,0?06 %;K2HPO4.3H2O, 0?006 %;pH6?5– Waters, Russian Academy of Sciences (Borok, Russia). It was iso- 7?5) and was fed with the lobose amoeba sp. All lated from samples of brown collected from the Black Sea at a depth of 1 m and a salinity of 18 %, between the settlement Noviy attempts to cultivate Multicilia using as food the bacterium Svet and the village Vesyolye (Sudak district of Crimea). Pseudomonas fluorescens (which grew after addition of rice or wheat grains to the culture), the alga Chlorella, the DNA was extracted using a DNeasy Minikit (Qiagen). The com- bodonids Bodo designis, Bodo saltans or Procryptobia soro- plete SSU rRNA gene sequence of Multicilia marina was amplified using kini, or the chrysomonads Spumella or Paraphysomonas, the universal primers sA (59-ACCTGGTTGATCCTGCCAGT-39)andsB failed. During light microscopic observation, it was establ- (59-TGATCCTTCTGCAGGTTCACCTAC-39). PCR was performed in a total volume of 50 ml with an amplification profile consisting of 40 cycles ished that, at the time of feeding, cells of Multicilia marina of 30 s at 94 uC, 30 s at 50 uC and 2 min at 72 uC, followed by 5 min at used their ventral side to cover the amoeba Vannella, and 72 uC for the final extension. The PCR products were purified using a then stood motionless. The predator resumed its motion High Pure PCR Purification kit (Roche), then ligated into the pGEM-T several seconds after the capture of its prey.

1450 International Journal of Systematic and Evolutionary Microbiology 56 Small-subunit rRNA gene of Multicilia

Fig. 1. SSU rRNA gene-based phylogeny of eukaryotes, showing the positions of Multicilia marina and the unidentified eukaryote clone Borok. Both organisms clearly belong to the Amoebozoa (grey box). The topology was inferred using the ML method with the GTR model, with a gamma distribution and a proportion of invariable sites. Numbers at nodes represent posterior probabilities resulting from a Bayesian analysis using the same model. Values under 0?5 have been omitted. All branches are drawn to scale; the bar represents 0?05 substitutions (corrected) per site. The tree is presented in an unrooted format, with a basal trifurcation. http://ijs.sgmjournals.org 1451 S. I. Nikolaev and others

Fig. 2. SSU rRNA gene-based phylogeny of the Amoebozoa, showing the positions of Multicilia marina and the unidentified eukaryote clone Borok. Both organisms branch within a clade containing Phalansterium solitarium, Gephyramoeba sp., Filamoeba nolandi, the environmental clone RT5iin44, the Archamoebae and the Mycetozoa, which is named Conosea (grey box). The topology was inferred using the ML method with the GTR model, with a gamma distribution and a proportion of invariable sites. Numbers at nodes represent bootstrap values after 100 data resamplings (lower) and posterior probabilities resulting from a Bayesian analysis using the same model (upper). Solid circles show nodes supported by a BV of 100 % and a PP of 1?0. Values under 50 %/0?5 are indicated by dashes. All branches are drawn to scale, except the five branches marked with an asterisk, which were reduced by half; the bar represents 0?1 substitutions (corrected) per site. The tree was rooted with four opisthokonts.

Unusual length of the SSU rRNA gene sequence second-longest amoebozoan SSU rRNA gene sequence of Multicilia marina after that of palustris (3502 bp). The length of the SSU rRNA gene sequence of Multicilia marina is sim- The length of the complete, amplified SSU rRNA gene ilar to those of pelobionts and nana. In other sequence of Multicilia marina was 2746 bp, which is the Amoebozoa, the SSU rRNA gene is generally longer than

1452 International Journal of Systematic and Evolutionary Microbiology 56 Small-subunit rRNA gene of Multicilia that in most eukaryotes, but does not exceed 2200 bp, with performed on a second dataset including the two sequences the exception of spp. (Table 1). obtained in this study and those of 12 mycetozoans, 5 pelobionts, 6 entamoebids and 35 other amoebozoans, According to a predicted secondary structure (see Supple- with 4 opisthokonts as an outgroup (Fig. 2). As in Fig. 1, mentary Fig. S1 in IJSEM Online), nine insertions are Multicilia marina branches within a clade containing Pha- responsible for the extraordinary length of the SSU rRNA lansterium solitarium, Gephyramoeba sp., F. nolandi and the gene of Multicilia marina, which are located in the variable two unidentified clones Borok and RT5iin44. Analysis of regions V2, V4, V5, V7 and V8 (see Table 1). One of the the second dataset revealed that this clade, supported by a three insertions in region V2 corresponds to the additional PP of 0?91 and a bootstrap value (BV) of 84 %, also includes helix E8_1, which is absent from most eukaryotes. The Archamoebae and Mycetozoa. other two correspond to extensions in the terminal loops of helices 10 and 11. Two insertions are present in region V4 Resolution within this clade was relatively good. As in (extensions in the terminal loops of helices E23_12 and Fig. 1, Multicilia marina branches as the sister to E23_14). The insertion in region V5 corresponds to an Gephyramoeba sp. (PP=1?00; BV=91 %). F. nolandi and extension in the terminal loop of helix 29. The insertion the two unidentified clones Borok and RT5iin44 form a in region V8 corresponds to the additional helix E45_1, strongly supported clade (PP=1?00; BV=93 %). Archa- which is absent from most eukaryotes, but present in all moebae (supported by a PP of 1?00 and a BV of 99 %) and Archamoebae and some other amoebozoans. Remarkably, Mycetozoa (PP=1?00; BV=99 %) branch together (PP= at 426 bp, variable region V7 (helices 43 and E43_1 to 0?81; BV=68 %) as a sister group to the ‘Filamoeba clade’ E43_4) of the SSU rRNA gene of Multicilia marina is (PP=0?93; BV=71). Finally, Phalansterium solitarium among the longest of all eukaryotic sequences available occupies the most basal position in the clade, with a PP to date, with longer insertions present to our knowledge of 0?99 and a BV of 80 %. only in the strepsipteran insects. The secondary structure proposed for this region in Supplementary Fig. S1 is tenta- As in Fig. 1, other groupings within the Amoebozoa are con- tive and other structures are possible. gruent with a recent molecular phylogeny and classification of the phylum (Smirnov et al., 2005), and were named accordingly in Fig. 2. Three well-supported are Phylogenetic position of Multicilia marina recognized, the classes and Flabellinea and the Fig. 1 shows the result of an ML analysis of 100 SSU rRNA order Acanthopodida. As shown previously (Fahrni et al., gene sequences of eukaryotes, including Multicilia marina, 2003; Cavalier-Smith et al., 2004; Smirnov et al., 2005), the the unidentified eukaryote clone Borok, 23 lobose amoebae relationships between these clades and the three indepen- and 75 other eukaryotes representing the most sequenced dent lineages formed by Dermamoeba algensis, lineages of opisthokonts and bikonts. In the absence of a sp. and ‘’ stenopodia were very poorly resolved prokaryotic outgroup, the tree is presented in an unrooted (the generic status of the latter needs revision; see, for format, with a basal trifurcation. The analysis clearly demon- example, Smirnov et al., 2005). In the tree shown in Fig. 2, strates that Multicilia marina belongs to the Amoebozoa, the the Tubulinea occupy a basal position within Amoebozoa, monophyly of which was supported by a Bayesian posterior with a PP of 0?56 and a BV of 81 %. probability (PP) of 1?00. Opisthokonts also form a strongly supported clade (PP=1?00). In contrast, bikonts form two clusters: and apusomonads branch together DISCUSSION next to the opisthokonts, although without Bayesian sup- port, whereas the remaining lineages form a weakly Multicilia marina is an amoebozoan supported group (PP=0?54) and resolution among them is typically poor. Phylogenetic analysis of SSU rRNA gene sequences clearly showed that, despite its numerous flagella, Multicilia marina In the tree presented in Fig. 1, Multicilia marina branches is more closely related to lobose amoebae than to any other next to Gephyramoeba sp. (PP=0?84) and together they group of eukaryotes. Furthermore, the monophyly of the form a sister-group to a strongly supported clade (PP=0?97) Amoebozoa including Multicilia marina was strongly sup- comprising Filamoeba nolandi, the unidentified eukaryote ported in our analyses (Fig. 1), in spite of the inclusion of clone Borok and an environmental clone (RT5iin44) that several bikont lineages of unknown affinities, such as apuso- was described by Amaral Zettler et al. (2002). The Multicilia monads, , Ancyromonas, Diphylleia, Trimastix + Filamoeba + Gephyramoeba clade is supported by a PP of and anathema [after this paper had been accepted, 0?80. Phalansterium solitarium branches at the base of this ‘Mastigamoeba’ invertens was redescribed as Breviata clade, but with lower support (PP=0?65). anathema (Walker et al., 2006)], some of which have been shown to group near or with lobose amoebae in previous large-scale phylogenies of eukaryotes (e.g. Brugerolle et al., Phylogenetic relationships among Amoebozoa 2002; Cavalier-Smith & Chao, 2003; Cavalier-Smith, 2004). To further investigate the position of Multicilia marina This result contradicts the initial idea that Multicilia is an among Amoebozoa, ML and Bayesian analyses were intermediate stage between flagellates and heliozoans, as http://ijs.sgmjournals.org 1453 1454 others and Nikolaev I. S.

Table 1. Sequence length and unusual secondary-structure features of the SSU rRNA gene of Multicilia marina and other Amoebozoa

The size of helices 6, 10, 11, 16, E23_12, E23_14, 26, 29, 37, 44 and 46 is conserved in most eukaryotes; ‘h’ indicates that, in the taxon concerned, the terminal loop of the helix extends into a hairpin that is usually absent from other eukaryotes. Helices E8_1, E43_4 and E45_1 are absent from most eukaryotes; + and 2 indicate the presence or absence of these helices in the different amoebozoan taxa, respectively. In some Acanthamoeba species the helix E10_1 branches into two terminal hairpins, whereas in spp. two helices instead of one are found between helices 10 and 11. Parentheses indicate that the structure is present only in a subset of the members of the taxon.

Taxon Sequence 6 E8_1 10 E10_1 11 16 E23_1D E23_4D E23_12 E23_14 26 29 37 E43_1D E43_4 44 E45_1 46 length (bp)

Tubulinea 1766–1980 2 (h) 1 1 0–2 22 Flabellinea, 1960–2107 (+) 1 1–3 (h) 1 (+)(+) Flabellinea, Vannellida 1838–1962 (+) 1 1–3 h 0 2 (+) Mayorella sp. 2131 2 *d 12 122 ‘Platyamoeba’ stenopodia 2143 2 h23 122 Dermamoeba algensis 2037 2 22 122 Acanthopodida, Balamuthia 1972 2 12 h 122

nentoa ora fSseai n vltoayMicrobiology Evolutionary and Systematic of Journal International Acanthopodida, Acanthamoeba spp. 2303–2517 (+) (h) 2 2–3 2–3 (h) h 1 2 + Phalansterium solitarium 1876 2 11 122 Filamoeba nolandi 1846 2 11 122 Gephyramoeba sp. 1859 2 11 122 Multicilia marina 2746 + hh22hhh** + Archamoebae, Mastigamoeba spp. 2529–2741 (+) (h) * 1 2 * h 1 ++ Archamoebae, Pelomyxa 3502 h 2 h* h* * h *hhh* * + h Archamoebae, Endolimax 2589 2 *11*h12 + Archamoebae, Entamoeba spp. 1858–2104 2 211 h12 h + Mycetozoa, Dictyosteliida 1857–1918 2 1–2 1 1 22 Mycetozoa, 1839–2042 (+) (h) 1 2 1 2 (+)

DAccording to Wuyts et al. (2001), up to two helices arise from helix E23_1 (E23_2 and E23_3), up to three helices arise from helix E23_4 (E23_5 to E23_7) and up to two helices arise from helix E43_1 (E43_2 and E43_3); numbers indicate how many helices arise from these helices in the different amoebozoan taxa. dAn asterisk indicates the presence of an important insertion of as yet unknown secondary structure. 56 Small-subunit rRNA gene of Multicilia proposed by Cienkowski (1881) on the basis of the similarity It appears that the addition of the sequences of Multicilia of its body form, mode of locomotion and the presence of and one unidentified amoeba was important to recover multiple flagella. It is also in disagreement with the more the monophyly of all flagellated amoebae. The monophyly recent placement of Multicilia in the phylum Multiflagellata of Conosea is indeed sensitive to taxon sampling, as of uncertain affinities (Mikrjukov & Mylnikov, 1996, 1998) Phalansterium does not usually branch with Filamoeba, or among heterotrophic flagellates of uncertain affinities Gephyramoeba, Archamoebae or Mycetozoa in the absence (Patterson et al., 2000). of Multicilia and the two unidentified clones Borok and RT5iin44 (data not shown). Comparison of the present The results of our study confirm the hypothesis that results with those of previous analyses also suggests that Multicilia belongs in the phylum Amoebozoa, as first the use of a larger sampling of Archamoebae and Mycetozoa suggested by Cavalier-Smith (1998). However, the position was important to stabilize the relationships within Conosea of Multicilia among Amoebozoa in our SSU rRNA gene (data not shown). In particular, the addition of two recently trees differs from the classification of this genus in the published sequences of (Fiore-Donno et al., 2005) , together with Vannellidae, Vexilliferidae and allows the breaking of the very long stem branch leading (Cavalier-Smith, 2003; Cavalier-Smith et al., to myxogastrids. This apparently increases the bootstrap 2004). This classification was based on observations of cell support for the close relationship between Multicilia and surface glycostyles in Multicilia (Mikrjukov & Mylnikov, Gephyramoeba, as well as for the monophyly of Mycetozoa 1996), possibly similar to the glycostyles present in vannel- and their sister-group relationship with Archamoebae. lids (Bovee, 1965). However, it cannot be excluded that the glycostyles observed in Multicilia originate from its prey, Apart from Multicilia, the flagellated amoebae include Vannella. According to our analyses, the closest sequenced pelobionts, myxogastrids and Phalansterium, a genus that relative of Multicilia is Gephyramoeba sp. ATCC 50654, was previously considered as a zooflagellate (Ekelund, 2002) and both organisms branch far from Vannellidae and and was only recently transferred to the Amoebozoa, based Vexilliferidae. Although well-supported in our analyses, on an analysis of its SSU rRNA gene sequence (Cavalier- the relationship between Multicilia and Gephyramoeba is Smith et al., 2004). These flagellated Amoebozoa are all difficult to assess morphologically, because the ATCC characterized by a unikont flagellar apparatus, with a kinetid strain of Gephyramoeba sequenced by Amaral Zettler et al. composed of one centriole and one flagellum, with a (2000) was not described in detail and was possibly mis- microtubular cone (Cavalier-Smith, 1998). Ultrastructural identified (A. Smirnov, personal communication). studies of Multicilia marina (Mikrjukov & Mylnikov, 1996, 1998) revealed that the basal apparatus of each flagellum is Monophyly of all flagellated amoebozoans also represented by a single centriole and possesses a micro- Our SSU rRNA gene analyses, including the first sequence of tubular cone. However, in contrast with other flagellated Multicilia to be obtained and an exhaustive sampling of amoebozoans, which possess a kinetid with the broad end available complete sequences of Archamoebae and of the conus of microtubules orientated towards the nucleus Mycetozoa, allow us to recover for the first time a (Cavalier-Smith, 1998), in Multicilia the kinetid broad end is monophyly of all flagellated amoebozoans (pelobionts, orientated towards the cell surface (Mikrjukov & Mylnikov, myxogastrids, Multicilia and Phalansterium), which branch 1996). This might be an ancestral state or a consequence of together with a few non-flagellated lineages (entamoebids, the multiciliarity, where the flagellar apparatus does not lean , Filamoeba and Gephyramoeba) within a on the nucleus but on the surface skeletal elements. strongly supported clade (PP of 0?91 and BV of 84 %). As Some of the features observed in the primary and secondary this clade contains all flagellated amoebozoans examined to structures of the SSU rRNA gene of Multicilia suggest its date, we propose to name it Conosea (after Cavalier-Smith, possible close relationship to Archamoebae (Table 1). The 1998), in agreement with the new system of Amoebozoa extraordinary length of the SSU rRNA gene in Multicilia is published recently by Smirnov et al. (2005). shared with the majority of Archamoebae (Hinkle et al., A common origin for Archamoebae and Mycetozoa was 1994; Milyutina et al., 2001), although this is not strong proposed previously by Cavalier-Smith (1998). In a multi- evidence for a common origin, because even closely related gene analysis of expressed sequence tag data from Masti- amoebae species may differ widely in their SSU rRNA gene gamoeba, Entamoeba and Dictyostelium, Bapteste et al. length (in the Acanthopodida, for example, it varies from (2002) later claimed to have confirmed the monophyly of 1972 bp in to greater than 2500 bp an Archamoebae + Mycetozoa clade, but the lack of any in some Acanthamoeba spp.). Multicilia and all Archa- lobose amoebae in their study made this conclusion irrele- moebae, except Entamoeba, possess an insertion in variable vant. On the other hand, analyses of actin and SSU rRNA region V4, in helix E23_14 (data not shown). This insertion genes from a broad taxonomic sampling of Amoebozoa corresponds in Multicilia to an additional hairpin in the provided opposite results, showing Archamoebae and loop of helix E23_14. Unfortunately, because the secondary Mycetozoa as separate lineages (Fahrni et al., 2003). structure of region V4 has not yet been fully determined in Finally, these two taxa did not group together, even with Archamoebae, we do not know whether this region also the inclusion of Phalansterium (Cavalier-Smith et al., 2004). corresponds to an additional hairpin in the loop of helix http://ijs.sgmjournals.org 1455 S. I. Nikolaev and others

E23_14 in this taxon. Multicilia also shares with all Archa- moebae an insertion in region V8 (additional helix E45_1). The presence of helix E45_1 is a characteristic feature of Archamoebae, although it is present in some other Amoebozoa (Acanthamoeba spp., sp. and minutissima).

Finally, there are also some ultrastructural features of the flagellar root system that may suggest a relationship between Multicilia and Archamoebae (Cavalier-Smith, 2000). These include the presence of the microtubular cone and of a short flagellar band around the flagellum, as well as a single microtubular band (Brugerolle, 1991; Simpson et al., 1997; Mikrjukov & Mylnikov, 1998; Brugerolle & Mu¨ller, 2000; Walker et al., 2001). The loss of some components of the axoneme of the flagellum (Brugerolle & Patterson, 2000), Fig. 3. Schema indicating two possible positions of the root of such as the outer dynein arms, and the weak motion of the Amoebozoa. The first is that suggested by the phylogenetic flagella in some Archamoebae (Mastigina, Pelomyxa)are analysis shown in Fig. 2 (Tubulinea is the most basal clade also shared by Multicilia. Perhaps the ancestor of Archa- within Amoebozoa), whereas the second implies the minimum moebae lived, as Multicilia, in an aerobic environment, number of flagella losses during the evolution of Amoebozoa possessed tubulocristate mitochondria and was multi- (Conosea is the most basal clade within Amoebozoa). Flagella flagellated. If this hypothesis is correct, then Pelomyxa, losses given this most-parsimonious hypothesis are indicated with its multiple flagella (Griffin, 1988; Goodkov & Seravin, by solid bars or by curved bars when only some lineages are 1991), may have maintained the ancestral state of the group concerned within the corresponding taxon. and, because of a sedentary lifestyle, its flagella have been preserved in a reduced form only. the ), or twice if Entamoeba and Endolimax Where is the root of Amoebozoa? are not monophyletic, as suggested by SSU rRNA gene analyses (see Fig. 2). In this particular case, the loss of The monophyly of flagellated amoebae depends not only on the sequences included in the analyses, but also on the flagella may be related to the parasitic mode of life of position of the root of Amoebozoa. Because the radiation of members of these genera. In the whole Amoebozoa, up to Amoebozoa is not well resolved, its rooting varies depending ten flagella losses have been proposed to match the observed on the gene, type of analysis, taxonomic sampling and tree topology (Cavalier-Smith et al., 2004). However, the selected outgroup. In the present study, the root of Amoe- number of flagella losses depends on the phylogenetic bozoa is either placed between Tubulinea + Mayorella + position of the flagellated species within Amoebozoa and ‘Platyamoeba’ stenopodia + Acanthopodida and Flabellinea the position of the root in the group. According to the + Dermamoeba algensis + Conosea (Fig. 1) or between topologies shown in Figs 1 and 2, which place Multicilia Tubulinea and all other amoebozoans (Fig. 2). The latter and Phalansterium in the same clade, the number of flagella result was also recovered in actin phylogenies published by losses can be reduced to eight. If the rooting possibilities Fahrni et al. (2003) and Nikolaev et al. (2005) and in the SSU suggested by our analyses are not correct and the root rRNA gene phylogeny published by Smirnov et al. (2005). In actually lies between Conosea and all other amoebozoans, other SSU rRNA gene trees, the root was placed between as observed in some ML trees (data not shown), this number Tubulinea + Archamoebae and Mycetozoa + other lobose would be reduced to six (Fig. 3). In this most-parsimonious amoebae (Fahrni et al., 2003) or between Acanthopodida case, the flagellum would have been lost once in the and other Amoebozoa (Cavalier-Smith, 2002). Cavalier- common ancestor of Tubulinea and Flabellinea, once in Smith et al. (2004) used Breviata anathema as an outgroup, Gephyramoeba sp., once in the common ancestor of the but this species is not related to Amoebozoa (Fig. 1; see also Filamoeba clade, once in the common ancestor of dictyo- Edgcomb et al., 2002) and did not increase the stability of stelids and once or twice in the parasitic Archamoebae. the relationships within Amoebozoa. Finally, in other pro- tein trees, the taxonomic sampling is too small to make any The rooting would be much more complex if Amoebozoa assumption about the position of the root. are paraphyletic. This possibility cannot be excluded entirely because of (i) the as yet rather weak evidence for their Alternatively, the position of the root might be deduced monophyly, and (ii) the as yet not well-established posi- from such features as the presence of flagella. Assuming that tion of Amoebozoa between opisthokonts and bikonts. the ancestor of Ameobozoa was flagellated, as proposed by According to a recent hypothesis on the early divergence Cavalier-Smith (1998, 2000), the flagellum was lost several of eukaryotes (Stechmann & Cavalier-Smith, 2003; Richards times during the evolution of amoebae. For example, a & Cavalier-Smith, 2005), the Amoebozoa are close to the flagella loss occurred at least once within Archamoebae (in eukaryotic root. This root has been placed between the

1456 International Journal of Systematic and Evolutionary Microbiology 56 Small-subunit rRNA gene of Multicilia ancestrally uniflagellate opisthokonts and Amoebozoa on Cavalier-Smith, T. (2002). The phagotrophic origin of eukaryotes the one hand and the ancestrally biflagellate bikonts on the and phylogenetic classification of Protozoa. Int J Syst Evol Microbiol other. However, it cannot be excluded that the Amoebozoa 52, 297–354. might be paraphyletic. If this was the case, it would be Cavalier-Smith, T. (2003). Protist phylogeny and the high-level obvious to search for early eukaryotes among flagellated classification of Protozoa. Eur J Protistol 39, 338–348. amoebae. One of them, Phalansterium, has already been Cavalier-Smith, T. (2004). Only six kingdoms of life. Proc R Soc Lond proposed as a candidate for the closest relative to the B Biol Sci 271, 1251–1262. ancestral eukaryotes (Cavalier-Smith et al., 2004). Further Cavalier-Smith, T. & Chao, E. E.-Y. (2003). Molecular phylogeny of multigene studies of this and other flagellated amoebozoans, centrohelid , a novel lineage of bikont Eukaryotes that arose by ciliary loss. J Mol Evol 56, 387–396. including Multicilia, may possibly bring answers to some of the questions concerning early eukaryote evolution. Cavalier-Smith, T., Chao, E. E.-Y. & Oates, B. (2004). Molecular phylogeny of Amoebozoa and the evolutionary significance of the unikont Phalansterium. Eur J Protistol 40, 21–48. ACKNOWLEDGEMENTS Cienkowski, L. (1881). An account on the White Sea excursion in 1880. Proc St-Petersb Imp Soc Nat 12, 130–171 (in Russian). The authors thank Alexey Smirnov for useful comments on the De Rijk, P. & De Wachter, R. (1997). RnaViz, a program for the manuscript. This work was supported by the Russian Foundation visualisation of RNA secondary structure. Nucleic Acids Res 25, for Basic research 02-04-48265, 02-04-48958, 02-04-49987 and 00-15- 4679–4684. 97905, and Swiss National Science Foundation grants 3100-064073.00 Edgcomb, V. P., Simpson, A. G. B., Amaral Zettler, L., Nerad, T. A., and 7SUPJ062343. Patterson, D. J., Holder, M. E. & Sogin, M. L. (2002). Pelobionts are degenerate protists: insights from molecules and morphology. Mol Biol Evol 19, 978–982. REFERENCES Ekelund, F. (2002). A study of the soil flagellate Phalansterium solitarium Sandon 1924 with preliminary data on its ultrastructure. Amaral Zettler, L. A., Nerad, T. A., O’Kelly, C. J., Peglar, M. T., Protistology 2, 152–158. Gillevet, P. M., Silberman, J. D. & Sogin, M. L. (2000). A molecular Fahrni, J. F., Bolivar, I., Berney, C., Nassonova, E., Smirnov, A. & reassessment of the leptomyxid amoebae. Protistologica 151, 275–282. Pawlowski, J. (2003). Phylogeny of lobose amoebae based on Amaral Zettler, L. A., Go´ mez, F., Zettler, E., Keenan, B. G., Amils, R. actin and small-subunit ribosomal RNA genes. Mol Biol Evol 20, & Sogin, M. L. (2002). Microbiology: Eukaryotic diversity in Spain’s 1881–1886. River of Fire. Nature 417, 137. Fiore-Donno, A.-M., Berney, C., Pawlowski, J. & Baldauf, S. L. Bapteste, E., Brinkmann, H., Lee, J. A. & 8 other authors (2002). (2005). Higher-order phylogeny of plasmodial slime molds The analysis of 100 genes supports the grouping of three highly (Myxogastria) based on elongation factor 1-a and small subunit divergent amoebae: Dictyostelium, Entamoeba, and Mastigamoeba. rRNA gene sequences. J Eukaryot Microbiol 52, 201–210. Proc Natl Acad Sci U S A 99, 1414–1419. Goodkov, A. V. & Seravin, L. N. (1991). New ideas on the nature of Bolivar, I., Fahrni, J. F., Smirnov, A. & Pawlowski, J. (2001). SSU the giant amoeba Pelomyxa palustris; the position of this organism in rRNA-based phylogenetic position of the genera Amoeba and the system of lower eukaryotes (Peloflagellata classis n.). Zool Zh 70, (Lobosea, Gymnamoebia): the origin of Gymnamoebae revisited. 5–16 (in Russian). Mol Biol Evol 18, 2306–2314. Griffin, J. L. (1988). Fine structure and taxonomic position of the Bovee, E. C. (1965). An emendation of the amoeba genus giant amoeboid flagellate Pelomyxa palustris. J Protozool 35, 300–315. and a description of Vannella gen. nov. Trans Am Microsc Soc 84, Guindon, S. & Gascuel, O. (2003). 217–227. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Brugerolle, G. (1991). Flagellar and cytoskeletal systems in amito- Biol 52, 696–704. chondrial flagellates: Archamoeba, Metamonada, and Parabasalia. Hinkle, G., Leipe, D. D., Nerad, T. A. & Sogin, M. L. (1994). Protoplasma 164, 70–90. The unusually long small subunit ribosomal RNA of Phreatamoeba Brugerolle, G. & Mu¨ ller, M. (2000). Amitochondriate flagellates. In balamuthi. Nucleic Acids Res 22, 465–469. The : Unity, Diversity and Evolution, pp. 166–189. Edited by J. R. Green & B. S. C. Leadbeater. London: Taylor & Francis. Huelsenbeck, J. P. & Ronquist, F. (2001). MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17, 754–755. Brugerolle, G. & Patterson, D. J. (2000). Order Pelobiontida Page, 1976. In An Illustrated Guide to the Protozoa, 2nd edn, Kudo, R. (1954). Protistology, pp. 333–338. Springfield, IL: Charles C. pp. 1097–1103. Edited by J. J. Lee, G. F. Leedale & P. Bradbury. Thomas. Lawrence, KS: Society of Protozoologists. Lanave, C., Preparata, G., Saccone, C. & Serio, G. (1984). A new Brugerolle, G., Bricheux, G., Philippe, H. & Coffea, G. (2002). method for calculating evolutionary substitution rates. J Mol Evol triciliatum and (=Aulacomonas 20, 86–93. submarina) form a new family of flagellates () Larsen, N., Olsen, G. J., Maidak, B. L., McCaughey, M. J., Overbeek, with tubular mitochondrial cristae that is phylogenetically distant R., Macke, T. J., Marsh, T. L. & Woese, C. R. (1993). The ribosomal from other flagellate groups. Protist 153, 59–70. database project. Nucleic Acids Res 21, 3021–3023. Cavalier-Smith, T. (1998). A revised six- system of life. Biol Mikrjukov, K. A. & Mylnikov, A. P. (1996). Protist Multicilia marina Rev Camb Philos Soc 73, 203–266. Cienk. or a Heliozoon? Dokl Acad Nauk 346, 136–139 Cavalier-Smith, T. (2000). Flagellate megaevolution: the basis for (in Russian). eukaryote diversification. In The Flagellates: Unity, Diversity and Mikrjukov, K. A. & Mylnikov, A. P. (1998). The fine structure of a Evolution, pp. 361–390. Edited by J. R. Green & B. S. C. Leadbeater. carnivorous multiflagellar protist, Multicilia marina Cienkowski 1881 London: Taylor & Francis. (Flagellata ). Eur J Protistol 34, 391–401. http://ijs.sgmjournals.org 1457 S. I. Nikolaev and others

Milyutina, I. A., Aleshin, V. V., Mikrjukov, K. A., Kedrova, O. S. & Simpson, A. G. B., Bernard, C., Fenchel, T. & Patterson, D. J. (1997). Petrov, N. B. (2001). The unusually long small subunit ribosomal The organization of Mastigamoeba schizophrenia n. sp.: more RNA gene found in amitochondriate amoeboflagellate Pelomyxa evidence of ultrastructural idiosyncrasy and simplicity in pelobiont palustris: its rRNA predicted secondary structure and phylogenetic protists. Eur J Protistol 33, 87–98. implication. Gene 272, 131–139. Smirnov, A., Nassonova, E., Berney, C., Fahrni, J., Bolivar, I. & Nikolaev, S. I., Berney, C., Fahrni, J. F., Bolivar, I., Polet, S., Pawlowski, J. (2005). Molecular phylogeny and classification of the Mylnikov, A. P., Aleshin, V. V., Petrov, N. B. & Pawlowski, J. (2004). lobose amoebae. Protist 156, 129–142. The twilight of Heliozoa and rise of , an emerging Stechmann, A. & Cavalier-Smith, T. (2003). The root of the supergroup of amoeboid eukaryotes. Proc Natl Acad Sci U S A eukaryote tree pinpointed. Curr Biol 13, R665–R666. 101, 8066–8071. Swofford, D. L. (1998). PAUP*: Phylogenetic Analysis Using Nikolaev, S. I., Mitchell, E. A. D., Petrov, N. B., Berney, C., Fahrni, J. Parsimony (*and other methods). Sunderland, MA: Sinauer & Pawlowski, J. (2005). The testate lobose amoebae (order Associates. Kent, 1880) finally find their home within Amoebozoa. Walker, G., Simpson, A. G. B., Edgcomb, V., Sogin, M. L. & Protist 156, 191–202. Patterson, D. J. (2001). Ultrastructural identities of Mastigamoeba Patterson, D. J., Vørs, N., Simpson, A. G. B. & O’Kelly, C. (2000). punctachora, Mastigamoeba simplex and commutans and Residual free-living and predatory heterotrophic flagellates. In An assessment of hypotheses of relatedness of the pelobionts (Protista). Illustrated Guide to the Protozoa, 2nd edn, pp. 1302–1328. Edited by Eur J Protistol 37, 25–49. J. J. Lee, G. F. Leedale & P. Bradbury. Lawrence, KS: Society of Walker, G., Dacks, J. B. & Embley, T. M. (2006). Ultrastructural Protozoologists. description of Breviata anathema, n. gen., n. sp., the organism Peglar, M. T., Amaral Zettler, L. A., Anderson, O. R. & 7 other authors previously studied as Mastigamoeba invertens. J Eukaryot Microbiol (2003). Two new small-subunit ribosomal RNA gene lineages within 53, 65–78. the subclass gymnamoebia. J Eukaryot Microbiol 50, 224–232. Wuyts, J., Van de Peer, Y. & De Wachter, R. (2001). Distribution of Richards, T. A. & Cavalier-Smith, T. (2005). substitution rates and location of insertion sites in the tertiary evolution and the primary divergence of eukaryotes. Nature 436, structure of ribosomal RNA. Nucleic Acids Res 29, 5017–5028. 1113–1118. Zuker, M., Mathews, D. H. & Turner, D. H. (1999). Algorithms and Rodriguez, F., Oliver, J. F., Martin, A. & Medina, J. R. (1990). thermodynamics for RNA secondary structure prediction: a practical The general stochastic model of nucleotide substitution. J Theor Biol guide. In RNA Biochemistry and Biotechnology, pp. 11–43. Edited by 142, 485–501. J. Barciszewski & B. F. C. Clark. Dordtrecht: Kluwer.

1458 International Journal of Systematic and Evolutionary Microbiology 56