Embed Size (px)
Citation preview
ARTICLE IN PRESS
Protist, Vol. 156, 149—161, August 2005
1Correspondinfax +44 1865 2e-mail david.b
& 2005 Elsevdoi:10.1016/j
elsevier.de/protis
http://www.Published online date 11 July 2005ORIGINAL PAPER
Polyubiquitin Insertions and the Phylogeny ofCercozoa and Rhizaria
David Bassa,1, David Moreirab, Purificacion Lopez-Garcıab, Stephane Poletc, Ema E. Chaoa,Sophie von der Heydena, Jan Pawlowskic, and Thomas Cavalier-Smitha
aDepartment of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, UKbUnite d’Ecologie, Systematique et Evolution, UMR CNRS 8079, Universite Paris-Sud,Batiment 360, 91405 Orsay CEDEX, FrancecDepartment of Zoology and Animal Biology, University of Geneva, CH-1211 Geneva 4, Switzerland
Submitted July 8, 2004; Accepted March 26, 2005Monitoring Editor: Michael Melkonian
A single or double amino acid insertion at the monomer—monomer junction of the universaleukaryotic protein polyubiquitin is unique to Cercozoa and Foraminifera, closely related ‘core’ phyla inthe protozoan infrakingdom Rhizaria. We screened 11 other candidate rhizarians for this insertion:Radiozoa (polycystine and acantharean radiolaria), a ‘microheliozoan’, and Apusozoa; all lack it,supporting suggestions that Foraminifera are more closely related to Cercozoa than either is to othereukaryotes. The insertion’s size was ascertained for 12 additional Cercozoa to help resolve their basalbranching order. The earliest branching Cercozoa generally have a single amino acid insertion, like allForaminifera, but a large derived clade consisting of all Monadofilosa except Metopion, Helk-esimastix, and Cercobodo agilis has two amino acids, suggesting one doubling event and noreversions to a single amino acid. Metromonas and Sainouron, cercozoans of uncertain position, havea double insertion, suggesting that they belong in Monadofilosa. An alternative interpretation,suggested by the higher positions for Metopion and Cercobodo on Bayesian trees compared withmost distance trees, cannot be ruled out, i.e. that the second insertion took place earlier, in theancestral filosan, and was followed by three independent reversions to a single amino acid inChlorarachnea, Metopion and Cercobodo.& 2005 Elsevier GmbH. All rights reserved.
Key words: Cercozoa; Rhizaria; polyubiquitin insertion; Radiozoa; Foraminifera.
Introduction
Recent advances in eukaryote phylogenetic re-construction stem from progress in four areas:novel organisms being cultured, thereby providingraw materials for genetic and morphological
g author;[email protected] (D. Bass)
ier GmbH. All rights reserved..protis.2005.03.001
analysis; increased choice and sophistication ofanalytical techniques; basing phylogenetic ana-lyses on several or many genes (e.g. Baldauf et al.2000; Bapteste et al. 2002; Martin et al. 1998); andrefining sequence-based trees with data frommolecular cladistic characters, which can behighly resolving (Archibald et al. 2003; Stechmannand Cavalier Smith 2002, 2003), although such
ARTICLE IN PRESS150 D. Bass et al.
data should be treated cautiously as indelcharacters may be homoplasic, for example byconvergence or reversion (Bapteste and Philippe2002). Increasingly many previously unclassifiedtaxa are being placed into known groups, andmisleading classifications and taxon designationsarising from single-gene trees are being corrected(e.g. Cavalier-Smith, 2003a, 2004; Philippe andAdoutte 1998; Philippe et al. 2000). Consequently,the eukaryote tree is resolving into a topology witha rather small number of internally diversetaxonomic ‘supergroups’ (Berney et al. 2004;Cavalier-Smith 2003a, 2004). One of these, thebiciliate infrakingdom Rhizaria (Cavalier-Smith2002), has recently been the subject of severalstudies (Burki et al. 2002; Cavalier-Smith 2004;Cavalier-Smith and Chao 2003a—c; Nikolaev et al.2003, 2004), which have sought to determine itsmember phyla, their interrelationships, and itsclosest relatives.
The two ‘core rhizarian’ phyla are the recentlyestablished Cercozoa (Bass and Cavalier-Smith2004; Cavalier-Smith 1998; Cavalier-Smith andChao 2003c) and the long-known Foraminifera(reviewed in Pawlowski et al. 2002). A commonorigin of Cercozoa and Foraminifera was initiallysuggested by actin trees (Keeling 2001) and by themarked propensity of both groups to formreticulopodia (Cavalier-Smith 2002). Support forthis relationship was strengthened by the discov-ery that both phyla carry an insertion of one or twoamino acid residues at the monomer—monomerjunctions of the polyubiquitin gene (Archibald et al.2003; Archibald and Keeling 2004), and by RNApolymerase II phylogenies (Longet et al. 2003).
Ubiquitin is a 76 aminoacid protein found in alleukaryotes, rivalling histone 4 in its degree ofconservatism (Sharp and Li 1987). Ubiquitin genesoccur in three main forms: single open readingframes, fused to ribosomal protein genes, or ashead-to-tail multimers of ubiquitin coding regions(polyubiquitin). The protein is essential to manybasic processes in eukaryotic cells, for exampleapoptosis, signal transduction, DNA repair, stressresponse, transcriptional regulation, endocytosis,and cell cycle regulation (Finley and Chau 1991;Hershko and Ciechanover 1998; Sun and Chen2004).
Members of Haeckel’s Radiolaria also belong inRhizaria, even though Radiolaria itself is nowshown to be polyphyletic (Polet et al. 2004).Phaeodarea (e.g. Coelodendrum, Aulosphaera,Aulacantha) are actually monadofilosan cercozo-ans (Polet et al. 2004), probably most closelyrelated to the order Tectofilosida (Bass and
Cavalier-Smith 2004), while Acantharea and Poly-cystinea often group together (Lopez-Garcıa et al.2002). The latter grouping constitutes phylumRadiozoa (Cavalier-Smith 1993, despite the re-moval of Phaeodarea: Cavalier-Smith 2004). Re-cent molecular studies (Nikolaev et al. 2004) alsosupport the grouping of Sticholonchea withAcantharea as the subphylum Spasmaria withinthe Radiozoa (Cavalier-Smith 1993), confirmingthat the previous placement of Sticholonche inHeliozoa (Febvre-Chevalier 1990) was erroneous.Radiozoa are considered to be rhizarians on thebasis of a shared propensity to make reticulopodiaand their usual grouping with Foraminifera andCercozoa on rRNA trees (Cavalier-Smith and Chao2003b; Lopez-Garcıa et al. 2002; Nikolaev et al.2004; Polet et al. 2004).
However, the branching order of Cercozoa,Foraminifera, and Radiozoa is not well resolved.On 18S rRNA trees that include all three groups,Foraminifera often branch within or as sister toRadiozoa (Cavalier-Smith and Chao 2003b,c), orsometimes as sister to the cercozoan haplospor-idia (Nikolaev et al. 2004). When Radiozoa are notincluded, Foraminifera branch as sister to thecercozoan Gromia oviformis (Berney and Paw-lowski 2003; Longet et al. 2004; Nikolaev et al.2003). In the protein (actin, RPB1) trees, in theabsence of Radiozoa, Foraminifera may branchwithin core Cercozoa (Filosa) (Keeling 2001), or assister to plasmodiophorids (Archibald and Keeling2004) or to Gromia (Longet et al. 2003; 2004).However, when radiolarian actin was included,one of the foraminiferan actin paraloguesbranched with Radiozoa (Nikolaev et al. 2004).Here we sequence polyubiquitin genes fromseveral acantharea and polycystines and anadditional foraminiferan in order to decide whetherForaminifera are closer to Radiozoa or Cercozoa.
Two other groups originally but more doubtfullyincluded in Rhizaria are centrohelid heliozoa andApusozoa (Cavalier-Smith 2002). Recent 18SrDNA phylogenetic analyses, in combination withultrastructural data, have led to these beingomitted from the current circumscription of Rhi-zaria (Cavalier-Smith 2003a, 2004; Cavalier-Smithand Chao 2003a,b). However, in view of theirhistorical association with Rhizaria, it was impor-tant to check whether polyubiquitin structure isconsistent with this exclusion. We have thereforealso screened two members of Apusozoa and amarine microheliozoan (a probable sister tocentrohelids: Cavalier-Smith and Chao 2003a)in order to see if they have polyubiquitin insertionsor not.
ARTICLE IN PRESSPhylogeny of Cercozoa and Rhizaria 151
As some Cercozoa have one and others twoamino acids in the polyubiquitin insertion, thischaracter could in principle be used to partitionCercozoa cleanly into two groups if the number ofamino acids was doubled only once and has neverundergone reversion. This would be particularlyvaluable phylogenetically and taxonomically asthe basal branches of existing cercozoan rDNAtrees are already known to be poorly resolved,especially within the subphylum Filosa (Cavalier-Smith and Chao 2003c). Present evidence sug-gested that the cercozoan subphylum Endomyxahas an insertion of a single amino acid likeForaminifera (which may branch within it onsparsely sampled trees) and that all studied Filosaexcept for chlorarachnean algae have an insertionof two amino acids (Archibald et al. 2003;Archibald and Keeling 2004). Therefore, we alsosequenced polyubiquitins from a much largervariety of filosan Cercozoa, to see which lineageshave insertions of one or two amino acids andwhether the evolutionary transition between thetwo types is indeed sufficiently rare to be avaluable phylogenetic marker within Filosa.
Results
Figure 1 shows the polyubiquitin sequences of 59taxa, including numerous cercozoans. It can beseen that all Cercozoa have an insertion of eitherone or two amino acids compared with mosteukaryotes. By contrast all Foraminifera have asingle amino acid insertion, except for one species,where two different genes were found, two copies ofone with a single insertion, which we consider theauthentic foraminiferan sequence, and a single copywith no insertion that is likely to be a contaminant.None of the new sequences from Radiozoa,Apusozoa, and Heliozoa contained inserts.
To enable us to interpret the insertion patternwithin Cercozoa, especially of Cercobodo agilisfor which no previous sequences were available,we constructed more comprehensive 18S rDNAtrees for Cercozoa (Figs 2,3).
Discussion
We have now shown that the unnamed marinemicroheliozoan, two Apusozoa, and five Radiozoa(two acanthareans and three polycystineans) alllack amino acid insertions between the ubiquitinmonomers of the polyubiquitin gene. On thesimplest interpretation, which assumes that this
insertion once gained has never been lost, theabsence of the insertion in these two groupssuggests that Foraminifera and Cercozoa aremore closely related to each other than they areto any taxa lacking an insertion (Fig. 4). Becausewe cannot rule out the possibility that the insertionmight sometimes be lost, other evidence isneeded to confirm this probable closer relation-ship of Foraminifera and Cercozoa.
The Evolutionary Position of Radiozoa
If it is true that the insertion has never been lost inany group, then the frequent grouping of For-aminifera with Polycystinea on 18S rDNA trees(Cavalier-Smith and Chao 2003b,c) is most likely along-branch attraction artifact (Embley and Hirt1998; Philippe and Adoutte 1998; Philippe et al.2000). Initially, that grouping of Radiozoa andForaminifera seemed to support the classificationof both groups as the supergroup Retaria (Cava-lier-Smith 1999). The discovery that Phaeodareaare cercozoans (Polet et al. 2004) has shown thatthe taxon Retaria as originally conceived (Cavalier-Smith 1999) was polyphyletic. Therefore, Forami-nifera and Radiozoa (modified by excludingPhaeodarea: Cavalier-Smith 2003a) are againtreated as distinct phyla and Retaria not used asa formal taxon (Cavalier-Smith 1993, 2004). Un-less the polyubiquitin insertion was lost in Radio-zoa, our results indicate that even omittingPhaeodarea, Retaria would not be holophyletic.
The term ‘core rhizaria’ was coined for Cerco-zoa, Foraminifera, and Radiolaria, all character-ized by a propensity to form reticulopodia and/orfilopodia and all grouping together on rDNA trees(Cavalier-Smith and Chao 2003b). However, withthe exclusion of Apusozoa from Rhizaria this termwould become synonymous with Rhizaria, whichcurrently comprises three phyla: Cercozoa, For-aminifera, and Radiozoa (Cavalier-Smith 2004),and is very probably a clade (Fig. 4). We suggestthat it is now useful to restrict the term ‘coreRhizaria’ to Cercozoa plus Foraminifera, the twogroups with the shared synapomorphy of thepolyubiquitin insertion. Thus Rhizaria comprisecore Rhizaria plus Radiozoa.
Exclusion of Apusozoa from Rhizaria
When Apusozoa were originally tentatively placedin Rhizaria it was recognised that their inclusionmade the group paraphyletic (Cavalier-Smith
ARTICLE IN PRESS
Euglypha rotunda* Thaumatomonas sp.
* 'Spongomonas minima UT1' Cercomonas sp. E
Cercomonas sp. 22 (ATCC 50318)* Cercomonas 'longicauda'-1* Cercomonas 'longicauda'-2
Cercomonas sp. 18-1 (ATCC 50316)Cercomonas sp. 18-2 (ATCC 50316)
* Heteromita sp.* Sainouron mikroteron* Metromonas simplex
* Massisteria marina* Dimorpha-like
* Gymnophrys sp.* Metopion fluens
* Helkesimastix sp.* Cercobodo agilisBigelowiella natans
Lotharella amoebiformis 1Lotharella amoebiformis 2
Lotharella globosaSpongospora subterraneaPlasmodiophora brassicae
* Gromia oviformisReticulomyxa filosa
Haynesina germanica* Bathysiphon spp. 1&2
* Bathysiphon sp. 3* Collozoum sp.
* Collozoum inerme 1* Sphaerozoum italicum
* Stauracon pallidus* Xiphacantha alata
* marine microheliozoan* Ancyromonas sp.
* Amastigomonas sp.
TLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSAMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSAMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGAGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGNGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGSGMQIFVKTLTGKTLHLVLRLRGGS-MQIFVKTLTGKTLHLVLRLRGGA-MQIFVKTLTGKTLHLVLRLRGGS-MQIFVKTLTGKTLHLVLRLRGGS-MQIFVKTLTGKTLHLVLRLRGGA-MQIFVKTLTGKTLHLVLRLRGGS-MQIFVKTLTGKTLHLVLRLRGGS-MQIFVKTLTGKTLHLVLRLRGGT-MQIFVKTLTGKTLHLVLRLRGGT-MQIFVKTLTGKTLHLVLRLRGGS-MQIFVKTLTGKTLHLVLRLRGGA-MQIFVKTLTGKTLHLVLRLRGGA-MQIFVKTLTGKTLHLVLRLRGGT-MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGK
Trichomonas vaginalisNaegleria fowleri
Tetrahymena pyriformisEuplotes eurystomus
Plasmodium falciparumTrypanosoma cruzi
Phytophthora infestansAcanthamoeba castellaniiDictyostelium discoideumPhysarum polycephalum
Volvox carteriGracilaria gracilis
Arabidopsis thalianaOryza sativa
Pinus sylvestrisSaccharomyces cerevisiae
Neurospora crassaCandida albicans
Homo sapiensDrosophila melanogaster
Caenorhabditis elegans
TLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGKTLHLVLRLRGG--MQIFVKTLTGK
Filosa
Proteomyxidea
Endomyxa
CERCOZOA
FORAMINIFERA
RADIOZOA
APUSOZOAHELIOZOA
Monadofilosa
Chlorarachnea
basal filosan flagellates
Polycystinea
Acantharea
OTHER EUKARYOTES
Monomer N Monomer N+1
Plasmodiophorida
Clade A
Clade B
Figure 1. Junction between two ubiquitin monomers in the polyubiquitin genes of rhizarians, apusozoans, amarine microheliozoan (possibly sister to the centrohelid heliozoa: Cavalier-Smith and Chao 2003a), and arange of other eukaryotes. The insertions of one and two amino acids between the monomers are marked inbold. The 23 results generated by this study are marked with an asterisk.
152 D. Bass et al.
ARTICLE IN PRESS
12-4.3
0.1 substitutions/site
Euglyphida
NOVEL CLADE 1
Spongomonadida
NOVEL CLADE 2
NOVEL CLADE 3
BASAL 'GROUP T'
Thaumatomonadida
NOVEL CLADE 4
Cryomonadida
Tectofilosida
Phaeodarea
Tectofilosida
Heteromitidae
Cercomonadidae
Clade A
Clade B
Proteomyxidea
NOVEL CLADE 5
Metopiida
NOVEL CLADE 6
NOVEL CLADE 7
Metromonadida
Chlorarachnea
Phytomyxea
AscetosporeaGromiidea
Imbricatea
Imbricatea
Monadofilosa
Filosa
Sarcomonadea
Thecofilosea
Endomyxa
19
90
39
75
99
85
88
39
90
94
63
72
9091
97
7-4.5NOVEL CLADES 8 & 9
Plasmodiophorida
Phagomyxida
99
85
80
83
50
Cyphoderia ampulla AJ4187934-2.5: AY620259
A12: AY620315Paulinella chromatophora X81811
13-1.8: AJ620296Euglypha filifera AJ418786
Euglypha acanthophora AJ418788Trinema enchelys AJ4187927-6.6: AJ620336
9-2.2: AJ62030812-4.1: AJ620337Pseudopirsonia mucosa AJ561116
'Spongomonas minima' UT1 AF411281Spongomonas sp. KV/DB AJ620252
C1: AJ620329C5: AJ620331
7-1.3: AJ6203337-1.4: AJ620334Thaumatomonas seravini AF411259
Allas diplophysa AF411262A15: AY620317
9-2.4: AY620311D6: AY620316
13-2.8: AY62030412-4.5: AY620341
Cryothecomonas aestivalis 1 AF290539Cryothecomonas longipes AF290540
7-4.3: AY6203567-2.1: AY620354PP2-3D: AY620277
Lecythium sp. AJ5148677-6.1: AY620355A14: AY620321
F9: AY6202817-5.5: AY620347
7-6.2: AY6203489-1.4: AY620314
4-2.3: AY620274Pseudodifflugia cf. gracilis AJ418794
Aulosphaera trigonopa AY266292Aulacantha scolymantha AY266294
Coelodendrum ramosissimum AY266293Proleptomonas faecicola AF411275
Heteromita globosa U42447Allantion sp. AF411265
Bodomorpha minima AF411276Cercomonas sp. U42449Cercomonas sp. ÔLargeÕ AF411266
Cercomonas sp. AZ6 AF411268Cercomonas sp. 'Tempisque' AF411271
Cercomonas sp. (anaerobic) AF411272Cercomonas sp. 'Small' AF534712
D5: AY6203184-1.8: AY6202679-2.3: AY620313
4-2.7: AY62026812-1.3: AY620364
D8: AY620319N-Por filose amoeba AF289081
13-2.7: AY620300Dimorpha-like AF411283
4-6.2: AY620270Gymnophrys sp. TC-S AF411284Massisteria marina AF411286
13-4.8: AY6203014-6.1: AY620271
12-3.3: AY620358A17: AY620324
Cercobodo agilis AY748806 4-5.2: AY620285
6-2.7: AY6202867-3.4: AY620361F5: AY620287
F6: AY620288PP1-14: AY620290
Metopion sp. AF411278Metromonas simplex DB AY620255
Lotharella globosa AF076169Chlorarachnion reptans U03477
10-3.2: AY62029113-1.5: AY620306
Phagomyxa bellerocheae AF310903Phagomyxa odontellae AF310904
Polymyxa betae AF310902Spongospora nasturtii AF310901
Urosporidium crescens UCU47852Haplosporidium costale U20858
Gromia oviformis AJ457814Gromia oviformis AJ457812
Figure 2. Distance tree of 91 cercozoan 18S rRNA sequences (including some from environmental librariesand novel clades first described in Bass and Cavalier-Smith 2004) using 1163 positions (GTR+G+ i model). Allbootstrap percentages X75% are shown; others lower than this are shown for groups of particular interest tothis study. Black blobs indicate 100% support. Black squares to the right of the tree represent the number ofamino acid residues in the polyubiquitin insertion in the representative member(s) of the taxa indicated.
Phylogeny of Cercozoa and Rhizaria 153
ARTICLE IN PRESS
NOVEL CLADE 4
Cryomonadida
Tectofilosida
Phaeodarea
Tectofilosida
Heteromitidae
Cercomonadidae
Clade A
Clade B
Proteomyxidea
NOVEL CLADE 5Metopiida
NOVEL CLADE 6
NOVEL CLADE 7
MetromonadidaChlorarachnea
Phytomyxea
AscetosporeaGromiidea
Monadofilosa
Filosa
Sarcomonadea
Thecofilosea
EndomyxaNOVEL CLADES 8 & 9
PlasmodiophoridaPhagomyxida
NOVEL CLADE 1
Euglyphida
Spongomonadida
BASAL 'GROUP T'
Thaumatomonadida
Imbricatea
Imbricatea
NOVEL CLADE 3
NOVEL CLADE 2
Thecofilosea
Desmothoracida
Figure 3. Bayesian tree of 100 cercozoan sequences using 1118 positions. The figures and blobs at thenodes are bootstrap percentages (upper; blobs ¼ 100% support) and Bayesian posterior probabilities (lower)for this dataset. See Methods for details. The criteria for inclusion of bootstrap support values are the sameas for Figure 2 and also apply to the PP values.
154 D. Bass et al.
2002). Subsequently a possible relationship ofApusozoa to Excavata was suggested, enablingApusozoa to be excluded from Rhizaria (Cavalier-
Smith 2003a, b). Neither Amastigomonas norAncyromonas polyubiquitin genes carry the inser-tion, from which we infer that Apusozoa do not
ARTICLE IN PRESS
SticholonchePolycystinea Acantharea
RADIOZOA
Reticulopodia
Filosa
Monadofilosa
Proteomyxidea
Metopion
Cercobodo agilis
Helkesimastix
Endomyxa
Phytomyxea AscetosporeaFORAMINIFERA
OTHER EUKARYOTES, includingHeliozoa & Apusozoa
CERCOZOA
Core Rhizaria RHIZARIA
GA-AG 18S rRNA deletion
Chlorarachnea
Metromonas
Sainouron
Polyubiquitin insertion = 2 amino acids
Polyubiquitin insertion = 1 amino acid
Figure 4. Schematic tree showing relative phylogenetic positions of the taxa under study as jointly informedby 18S rRNA gene phylogenies and the partitioning suggested by the distribution of presence/absence, andthe size, of the polyubiquitin insertion. Solid branches summarize information from 18S rDNA analyses,morphological, and polyubiquitin data; dashed lines indicate those taxa (Sainouron and Helkesimastix) forwhich 18S rDNA sequences are unavailable. The dotted line to Metromonas signifies concordance ofmorphological and polyubiquitin insertion data, but discordance with 18S rDNA sequence data; that toCercobodo indicates its position according to some analyses and as suggested by the nature of the insertionin its polyubiquitin gene, but acknowledges the instability of its branching position (cf. Figs 2, 3).
Phylogeny of Cercozoa and Rhizaria 155
branch within core rhizaria. Apusozoa was acandidate for inclusion in Rhizaria on the basisof its mode of locomotion (active anterior andtrailing posterior flagellum used for gliding onsurfaces characterise Cercozoa) and sharing thecharacter of kinetocyst extrusomes with someCercozoa. The phylogenetic position (or evenholophyly) of Apusozoa cannot be resolved by18S rDNA analyses (Atkins et al. 2000; Cavalier-Smith 2000; Cavalier-Smith and Chao 2003b;Cavalier-Smith 2004). It is now considered incer-tae sedis in subkingdom Biciliata; whether it is themost divergent bikont group or is related toexcavates remains unclear (Cavalier-Smith 2004;Cavalier-Smith and Chao 2003b).
Heliozoa
Centrohelid heliozoa, like Apusozoa, fail to groupconsistently with any other phylum on rDNA trees
(Cavalier-Smith and Chao 2003b) and are nowalso provisionally excluded from Rhizaria (Cava-lier-Smith 2004). As all our stocks of genomic DNAfrom cultures of centrohelid Heliozoa are probablycontaminated with other (prey) eukaryotes, we didnot consider it worthwhile to sequence polyubi-quitin from them, as its provenance would havebeen uncertain. Heliozoa are therefore repre-sented in our polyubiquitin data only by theunnamed marine microheliozoan (which feeds onbacteria) that usually groups weakly as sister tocentrohelids in 18S rDNA trees (Cavalier-Smithand Chao 2003a). However, it is ultrastructurallyuncharacterised and probably not a centrohelid.The absence of the insertion from this microhe-liozoan suggests that it is not part of thecercozoan-foraminiferan clade, but this shouldnot be generalised to centrohelids until there isdirect evidence from them or firmer evidence thatthe two are related, though it is likely thatcentrohelids will turn out also to lack the insert.
ARTICLE IN PRESS156 D. Bass et al.
Centrohelids have also been considered putativeRhizaria: their hexagonally arranged axopodialaxonemes superficially resemble those of radi-olaria, and they also have kinetocyst extrusomes,but both characters may be convergent. They donot group within or as sister to Rhizaria on 18SrDNA and actin trees (Cavalier-Smith and Chao2003a; Nikolaev et al. 2004). As desmothoracids(e.g. Clathrulina, Hedriocystis) are not related tocentrohelids but group within Cercozoa, as sug-gested by Cavalier-Smith and Chao (2003a,c) andshown in Nikolaev et al. (2004), they should carrythe insert. This is currently being investigated.
The Relationship between Foraminiferaand Cercozoa
There is now evidence of single amino acidinsertions in the polyubiquitin gene of threephylogenetically divergent foraminiferan genera—Haynesina, Reticulomyxa (both have an alanineinsert) (Archibald et al. 2003), and Bathysiphon(threonine) (this study). This information, in com-bination with several other lines of evidence (actin,RPB1, and 18S rDNA phylogenies; see referencesabove), confirms that Foraminifera and Cercozoaare very closely related, but whether they aresisters or whether Foraminifera occupy a derivedposition within Cercozoa remains uncertain.
The facts that Foraminifera do not share theunique single nucleotide deletion in the 18S rRNAgene shown by all Cercozoa (Cavalier-Smith andChao 2003b), and that more highly sampled 18SrDNA distance and ML trees do not showForaminifera nested within Cercozoa, suggest thatForaminifera and Cercozoa may be sisters. How-ever, it is possible, given the rapid evolution of theforaminiferan rRNA genes (Pawlowski and Berney2003) that the thymine-containing nucleotide inForaminifera at the site of the cercozoan deletionmight be secondarily inserted, and that Foramini-fera are actually derived from Cercozoa. Thiswould be congruent with some actin trees, whichshow Foraminifera branching within Cercozoa, assister to Cercomonas (Keeling 2001) or as sister toPlasmodiophorida (Archibald and Keeling 2004).However, establishing the phylogenetic position ofForaminifera using actin sequences is madedifficult by the presence in Foraminifera of twoparalogous actin genes, which often do not grouptogether. For example, in some recent actinphylogenies, one paralogue branches with Gromiaoviformis, while the other appears as sister toPolycystinea (Longet et al. 2004; Nikolaev et al.
2004); another recent very well sampled actin treedid not even group Foraminifera and Cercozoa(Stechmann and Cavalier-Smith unpublished).This problem was resolved recently by increasingtaxon sampling of foraminiferan actin and byremoving fast-evolving actin sequences from theanalyses (Flakowski et al. 2005). However, actin istoo conserved to resolve the exact relationshipbetween Cercozoa and Foraminifera and otherproteins should be investigated.
Phylogeny of Cercozoa
As Figure 1 indicated, the following cercozoanshave a single amino acid insertion: Bigelowiellaand Lotharella (both chlorarachneans), Cercobodoagilis, Metopion, and Gromia. Helkesimastix alsohas a single amino acid insertion and is therefore acore rhizarian, but we cannot confirm that it is acercozoan as data from other genes are incon-clusive, even though its gliding zooflagellatemorphology strongly suggests that it is; thepresence of only one amino insertion, in contrastto the two in Cercomonas, provides additionalevidence that it is not a cercomonad. The taxawith a single amino acid insertion either belong tothe subphylum Endomyxa (Gromia), or are basallineages in subphylum Filosa in many 18S rDNAtrees (e.g. Fig. 2). The possible exceptions to thisare Metopion and Cercobodo, which in someanalyses group above proteomyxids and as sisterto heteromitids respectively, but with negligiblesupport (Fig. 3). At other times they branch belowproteomyxids, but equally unreliably.
In many cases the evidence from 18S rDNAtrees and the polyubiquitin insertion data areconcordant. The few cases where there areweakly supported discrepancies between treesand the distribution of the polyubiquitin insertionare priorities for further analysis. If there are noproven cases of loss of either amino acid residuesin the insertion, our data would provide strongevidence that taxa with a single amino acidinsertion are indeed the most basal filosans, andthose with insertions of two amino acids arederived. However, as Figure 3 suggests that thiscondition may have been violated at least threetimes, albeit with weak support (Cercobodo agilis,Metopion, and Metromonas), evidence from addi-tional sources is required to resolve the branchingorder within Cercozoa, and thus verify that somereversions from a double to a single amino acidinsertion have occurred.
ARTICLE IN PRESSPhylogeny of Cercozoa and Rhizaria 157
On the basis of limited morphological evidence(light microscope and ultrastructural) Metopion,Helkesimastix, and Cercobodo agilis might bethought to belong to Sarcomonadea. However,weak indications from 18S rDNA trees are nowstrengthened by our new polyubiquitin data, whichsuggest that they represent at least two different,divergent filosan lineages (Figs 2,3). As Cercobo-do agilis was originally described as Cercomonasagilis Morott, but had a single amino acid insertunlike all other Cercomonas strains, we se-quenced the 18S rRNA gene to see whether ornot it has been misclassified. As Figure 3 shows,Cercobodo agilis does not group with Cercomo-nas Clade A or B, but instead with Novel Clade 7(Bass and Cavalier-Smith 2004), in some analysesone of the most deeply diverging filosan clades.The suggestion, on the basis of morphologicalobservations, that Sainouron is a sarcomonad,and not a deep-branching filosan (Karpov 1990) issupported by the fact it has a two-amino acidinsertion. Both Sainouron and Helkesimastix arevery difficult to work with in terms of rDNAsequencing. As we have not been able to obtaincomplete 18S rDNA sequences from either ournew data provide the first molecular evidence thatthese genera are cercozoan and core rhizarianrespectively.
Almost all other cercozoans screened are morederived filosans (including Monadofilosa), accord-ing to 18S rDNA phylogenies, and have aninsertion comprising two amino acids. The onlyexception to this is Metromonas, which has thesame insertion of two amino acids as severalmonadofilosans but is the only taxon with such aninsertion to group basally—as sister to, or near toChlorarachnea in ML, distance, and Bayesiananalyses. However, this relationship has very littlestatistical support (see also Bass and Cavalier-Smith 2004) and is not corroborated by morpho-logical evidence. We suggest that Metromonasmay really be related to monadofilosans, but moredata are required to confirm this, and we do notrule out the possibility that Metromonas is a basalfilosan and that either the insertion of two aminoacids is a character convergent with that in thetrue monadofilosans or there have been second-ary reversions to a single amino acid. Indeed, if thehigher positions shown for Metopion and Cerco-bodo on the Bayesian tree (Fig. 3) turn out to becorrect, rather than their deeper branching seenon distance trees such as Figure 2, a reasonablyparsimonious alternative interpretation would bethat the second insertion took place somewhatearlier than indicated in Figure 4, in the ancestral
filosan, and was later followed by three indepen-dent reversions to a single amino acid in Chlorar-achnea, Metopion and Cercobodo.
Identification of Novel Cercozoan Clades
Of the nine novel clades revealed by screeningenvironmental 18S rRNA gene libraries in Bassand Cavalier-Smith (2004) (our knowledge of themwas previously limited to these sequence data),two have now been partially characterised by theknowledge that they contain Pseudopirsoniamucosa (Kuhn et al. 2004) (a modified NovelClade 1; 80% bootstrap support) and Cercobodoagilis (Novel Clade 7; 88% bootstrap support) (Fig.3). Figures 2 and 3 are the first published treesshowing convincingly that Pseudopirsonia muco-sa is a monadofilosan cercozoan, not a heterokontas stated in GenBank (Bass and Cavalier-Smith2004); although Kuhn et al. (2004) also showPseudopirsonia to be a cercozoan, parts of theirtree in Figure 3 are discordant with previousstudies of both heterokonts and Cercozoa; forexample alveolates appear polyphyletic; withinheterokonts Placidiales, Pseudofungi and hypo-gyrists are not monophyletic; and within CercozoaFilosa are not holophyletic and the genus Cerco-monas is polyphyletic, being split into threeclades, one of which misleadingly appears to besister to Pseudopirsonia. Furthermore, the Baye-sian tree (Fig. 3) strongly suggests for the first timethat Novel Clade 6 is closely related to Metopionand that Novel Clade 3 belongs in Thecofilosea.
DNA from Uncultivated Samples
It might be argued that as neither the polycystinesnor the acantharea were cultivated, the DNAsamples from them could have been contami-nated by non-rhizarian protists and that it may betheir DNA that was amplified by the polyubiquitinprimers, and that Radiozoa themselves mightactually have the insert. To ensure that this is notthe case, our sequences are based on severalindependent samples processed in separatelaboratories. Each sample was carefully washedin an attempt to exclude such contamination andall of them yielded only authentic radiozoan rRNAgenes when tested using universal 18S rDNAprimers. Although the possibility of contaminationof all of them seems remote, because we cannotaltogether exclude it (and also because there is apossibility that the ancestral radiozoa might have
ARTICLE IN PRESS158 D. Bass et al.
lost the insertion), our conclusion that Foramini-fera are more probably more closely related toCercozoa than to the Radiozoa requires corro-boration from other molecules.
In the foraminiferan Bathysiphon, even thoughtwo clones carried a single amino acid insertion asin all other studied foraminifera, one clone showedno insertion. This latter sequence is very unlikelyto originate from a foraminiferan as it lacks threesignature sequences from elsewhere in the poly-ubiquitin gene that are shared by all othersequenced foraminiferans, shown here in bolditalics with the non-foraminiferan state as sub-scripts: GGA-MQIFVKTLTGKTITLDVEPNSDTIQE/
D/SNVKAKIQDKEGIPPED.
Conclusions
The possible partitions in the rhizarian tree in thelight of the taxonomic distribution of the two typesof polyubiquitin insertion in combination withinformation from 18S rDNA trees were sum-marised in Figure 4. Because they are the onlytwo groups to share the insert, Cercozoa andForaminifera appear to be more closely related toeach other than any of the other groups screened,but the distribution of the polyubiquitin insertiondoes not allow us to decide whether Cercozoa aresisters of or ancestral to Foraminifera. If there turnout to be no cases where the insertion has beendeleted, then its absence from Apusozoa, Radio-zoa and the microheliozoan is strong evidencethat these groups are excluded from core Rhizaria,although Radiozoa probably belong to Rhizaria,branching just outside the Cercozoa—Foramini-fera clade.
For cercozoan phylogeny, Figure 3 is largelycongruent with that in Bass and Cavalier-Smith(2004). The major novelties are that Cercobodoagilis is strongly supported as a member of NovelClade 7 (Bass and Cavalier-Smith 2004) anddesmothoracids branch well within the proteo-myxids. Assuming that insertions have not arisenconvergently, we propose that Sainouron is amonadofilosan and Helkesimastix is a moredivergent cercozoan (though it might group out-side Cercozoa, it is nonetheless definitely a corerhizarian). Because of the conflict between thedistance and Bayesian trees in the positions ofMetopion and Cercobodo, we cannot determinewhether they are basal filosans that never had thesecond amino acid insertion (Fig. 4) or are part ofthe later sarcomonad radiation and have secon-darily lost it. Thus the second amino acid insertion
occurred either in the ancestral filosan or some-what later (Fig. 4), after the divergence of Helk-esimastix, Novel Clade 7, Chlorarachnea, and therest of Filosa. It also remains unclear whetherMetromonas is sister to Chlorarachnea, as weaklyindicated by the rDNA trees, or closer to Mon-adofilosa as morphology and the doubleness of itspolyubiquitin insertion suggest.
Post Scriptum
Since this paper was written, Vickerman andMoreira (pers. comm.) have shown that a newlydiscovered cercozoan, ‘Aurigamonas solis’, isrelated to Cercobodo agilis, belonging to NovelClade 7 according to 18S rDNA phylogenies, andlike C. agilis, has a single amino acid polyubiquitininsertion. Therefore it is possible that there was areversion to a single amino acid insertion in thecommon ancestor of all NC7 lineages.
Methods
Detection of polyubiquitin insertion: GenomicDNA from the following taxa was amplified withprimers UBIQ1: 5’-GGCCATGCARATHTTYGT-NAARAC-3’ and IUB2: 5’-GATGCCYTCYTTRT-CYTGDATYTT-3’: ‘Spongomonas minima UT1’(ATCC 50405), Cercomonas ’longicauda’(ATCC50344), Heteromita sp. (culture now dead),Sainouron mikroteron (ATCC 50340), Metromonassimplex (culture now dead), Massisteria marina(ATCC 50266), Dimorpha-like sp. (ATCC 50522),Gymnophrys sp. (ATCC 50638), Metopion fluens(AP Mylnikov, Borok, Russia), Helkesimastix faeci-cola (ATCC 50328), Cercobodo agilis (CCAP 1910/1), Gromia oviformis*, Bathysiphon sp.*, Collo-zoum sp.**, Collozoum inerme**, Sphaerozoumitalicum**, Stauracon pallidus**, Xiphacantha ala-ta**, Ancyromonas sigmoides (ATCC 50267),Amastigomonas sp. (ATCC 50062), and marinemicroheliozoan (TC-S lab). The species markedwith single asterisks are uncultivatable. Those withtwo asterisks were collected directly from plank-ton samples taken from Villefranche sur Mer(Mediterranean coast of France) at 25 m depthusing a plankton net. The different species wereidentified using a stereomicroscope, isolated andwashed many times with sterile sea water beforeDNA extraction. Bathysiphon sp. was collectedfrom the coast of Scotland. G. oviformis originatedfrom the Mediterranean coast. As for the plank-tonic species, the foraminifera were thoroughlycleaned prior to extraction, which was performed
ARTICLE IN PRESSPhylogeny of Cercozoa and Rhizaria 159
with 10—50 specimens. Gromia’s extraction wasdone using single specimens undergoing game-togenesis, as described in Arnold (1972). Toascertain the authenticity of ubiquitin sequences,amplifications were obtained from at least twoDNA extracts for each uncultured species and1—3 clones were sequenced for each PCRproduct.
PCR reactions were performed as follows: 3 mindenaturation at 94 1C, followed by 35 or 45 cyclesof 45 s at 92 1C, 1 min at 50 1C, and 1.5 min at72 1C, followed by a final extension of 5 min at72 1C. The products were run on a gel. This primerpair produces a ladder, the shortest fragment isless than a complete ubiquitin monomer; thebands increase in size by one whole monomerlength. Bands corresponding to 1.5 and 2.5monomer lengths were excised from the gel andpurified. rDNA clone libraries were constructedusing the TOPO TA Cloning kit (Invitrogen). Whiteor light blue colonies were picked and grown in2 ml LB medium:1 ml ampicillin overnight. Theplasmid DNA was extracted and cleaned usingthe QIAquick PCR Purification kit (Qiagen). Thecleaned products were run on a gel with knownnegative transformants; positive transformantswere sequenced using dye terminators andseparated on an automated ABI-377 sequencer.The sequencing primers used were M13 forwardand M13 reverse. The sequences were edited andtranslated in the trace editor TED, then aligned byeye using the Genetic Data Environment software.
18S rDNA distance and Bayesian trees: arepresentative sample of 89 partial (about 1.26 kbonly to allow inclusion of the incomplete novelclade sequences) cercozoan sequences from ourdatabase was aligned by eye using the GeneticData Environment software. New 18S rDNAsequences for Cercobodo agilis (this study:culture from K. Vickerman, Glasgow) and Pseu-dopirsonia mucosa (GenBank) were aligned to thisdataset. Gaps and ambiguously aligned positionswere also removed, leaving an alignment of 1163positions. ModelTest v.3.06 (Posada and Crandall1998) selected the GTR model with gammacorrection for intersite rate variation (G) andallowance for invariant sites (i) for the dataset(a ¼ 0:631; i ¼ 0:258). These parameters wereused for distance trees constructed using PAUP4.0b10 (Swofford, 2003). A heuristic search (500replicates; random addition) using minimum evo-lution was made (with TBR branch swapping). Thetree was rooted between Filosa and Endomyxa asshown by maximum likelihood trees with closeoutgroups in Cavalier-Smith and Chao (2003c).
Bootstrap support values were estimated using500 bootstrap replicates within one additionsequence for a distance analysis using a MLmodel of substitution with the same values of aand i.
The Bayesian tree (Fig. 3) was calculated usingMrBayes v3.0b4 (Huelsenbeck and Ronquist2001) based on 100 taxa, including all those inFigure 2 with the addition of desmothoracids(Clathrulina & Hedriocystis) and some moreenvironmental cercozoan sequences); 1118 posi-tions were used for the analysis. Two separateMCMCMC runs with randomly generated startingtrees were carried out for 4 million generationseach with one cold and three heated chains. Theevolutionary model applied included GTR sub-stitution matrix, gamma correction with shapeparameter estimated from the data, the covarionmodel, and autocorrelation. Trees were sampledevery 100 generations. Six thousand and fivehundred trees were discarded as burn-in (treessampled before the likelihood plots reached aplateau). A consensus of the remaining trees wasgenerated to reveal the posterior probabilities(PPs) of the branching pattern. Both independentruns resulted in basically identical tree topologiesand posterior probabilities, indicating that the runslasted long enough to converge. The tree shown isnot the consensus tree produced by the Bayesiananalysis but the tree with the highest likelihoodselected from the posterior distribution, annotatedwith the consensus PPs. The bootstrap supportvalues on Figure. 3 were estimated in PAUP4.0b10 using 500 bootstrap replicates for eachof 50 addition sequence replicates for a distanceanalysis using a GTR+G+I model (a ¼ 0:611;i ¼ 0:237).
Sequences have been deposited in GenBankunder accession numbers AY748806-AY748813,AJ937550-AJ937554, DQ098270-DQ098278, andDQ100436-DQ100438. Sequence alignments areavailable on request from the correspondingauthor.
Acknowledgements
The work was supported by NERC ResearchStudentships to DB and SvdH. TCS thanks NERCUK for research grants and NERC and theCanadian Institute for Advanced Research Evolu-tionary Biology Program for fellowship support.DM, PL-G, SP, and JP would like to thank Jeanand Colette Febvre (Station Biologique de Ville-franche sur Mer) for hospitality and help during the
ARTICLE IN PRESS160 D. Bass et al.
isolation and identification of Polycystinea andAcantharea. Their part of the work was supportedby an ATIP grant of the CNRS to PL-G. Researchof JP and SP was supported by Swiss NSF grant3100A0-100415. We thank John M Archibald forthe polyubiquitin primers.
Reference
Arnold ZM (1972) Observations on the biology of theprotozoan Gromia oviformis. Dujardin. Univ CalifPubl Zool 100: 76—107
Atkins MS, McArthur AG, Teske AP (2000) Ancyr-omonadida: a new phylogenetic lineage among theprotozoa closely related to the common ancestor ofmetazoans, fungi, and choanoflagellates (Opistho-konta). J Mol Evol 51: 278—285
Archibald JM, Keeling PJ (2004) Actin and ubiquitinprotein sequences support a Cercozoa/Foraminiferaancestry for the plasmodiophorid plant pathogens. JEukaryot Microbiol 51: 113—118
Archibald JM, Longet D, Pawlowski J, Keeling PJ(2003) A novel polyubiquitin structure in Cercozoaand Foraminifera: evidence for a new eukaryoticsupergroup. Mol Biol Evol 20: 62—66
Baldauf SL, Roger AJ, Wenk-Siefert I, DoolittleWF (2000) A kingdom-level phylogeny of eukaryotesbased on combined protein data. Science 290:972—977
Bapteste E, Philippe H (2002) The potential value ofindels as phylogenetic markers: position of tricho-monads as a case study. Mol Biol Evol 19: 972—977
Bapteste E, Brinkmann H, Lee JA, et al (2002) Theanalysis of 100 genes supports the grouping of threehighly divergent amoebae: Dictyostelium, Entamoe-ba, and Mastigamoeba. Proc Natl Acad Sci USA 99:1414—1419
Bass D, Cavalier-Smith T (2004) Phylum-specificenvironmental DNA analysis reveals remarkably highglobal biodiversity of Cercozoa (Protozoa). Int J SystEvol Microbiol 54: 2393—2404
Berney C, Pawlowski J (2003) Revised smallsubunit rRNA analysis provides further evidence thatForaminifera are related to Cercozoa. J Mol Evol 57:S120—S127
Berney C, Fahrni J, Pawlowski J (2004) How manynovel eukaryotic ‘‘kingdoms’’ Pitfalls and limitationsof environmental DNA surveys. BMC Biol 2: 13
Burki F, Berney C, Pawlowski J (2002) Phyloge-netic position of Gromia oviformis Dujardin inferred
from nuclear-encoded small subunit ribosomal DNA.Protist 153: 251—260
Cavalier-Smith T (1993) Kingdom Protozoa and its18 phyla. Microbiol Rev 57: 953—994
Cavalier-Smith T (1998) A revised six-kingdomsystem of life. Biol Rev Camb Philos Soc 73:203—266
Cavalier-Smith T (1999) Principles of protein andlipid targeting in secondary symbiogenesis: eugle-noid, dinoflagellate, and sporozoan plastid originsand the eukaryotic family tree. J Eukaryot Microbiol46: 347—366
Cavalier-Smith T (2000) Flagellate Megaevolution:the Basis for Eukaryote Diversification. In Green JR,Leadbeater BSC (eds) The Flagellates. Taylor andFrancis, London, pp 361—390
Cavalier-Smith T (2002) The phagotrophic origin ofeukaryotes and phylogenetic classification of Proto-zoa. Int J Syst Evol Microbiol 52: 297—354
Cavalier-Smith T (2003a) Protist phylogeny and thehigh-level classification of Protozoa. Europ J Protis-tol 39: 338—348
Cavalier-Smith T (2003b) The excavate protozoanphyla Metamonada Grasse emend. (Anaeromona-dea, Parabasalia, Carpediemonas, Eopharyngia) andLoukozoa emend. (Jakobea, Malawimonas): theirevolutionary affinities and new higher taxa. Int J SystEvol Microbiol 53: 1741—1758
Cavalier-Smith T (2004) Only six kingdoms of life.Proc R Soc Ser B 271: 1251—1262
Cavalier-Smith T, Chao E-Y (2003a) Molecularphylogeny of centrohelid heliozoa, a novel lineageof bikont eukaryotes that arose by ciliary loss. J MolEvol 56: 387—396
Cavalier-Smith T, Chao E-Y (2003b) Phylogeny ofChoanozoa, Apusozoa, and other protozoa and earlyeukaryote megaevolution. J Mol Evol 56: 540—563
Cavalier-Smith T, Chao E-Y (2003c) Phylogeny andclassification of phylum Cercozoa (Protozoa). Protist154: 341—358
Embley TM, Hirt RP (1998) Early branching eukar-yotes? Curr Opin Genet Dev 8: 624—629
Febvre-Chevalier C (1990) Phylum Actinopoda:Class Heliozoa. In Margulis L, Corliss JO, MelkonianM, Chapman D (eds) Handbook of Protoctista.Jones and Bartlett, Boston, MA, pp 399—416
Finley D, Chau V (1991) Ubiquitination. Annu RevCell Biol 7: 25—69
ARTICLE IN PRESSPhylogeny of Cercozoa and Rhizaria 161
Flakowski J, Bolivar I, Fahrni J, Pawlowski J(2005) Actin phylogeny of Foraminifera. J Foram Res35: 93—102
Hershko A, Ciechanover A (1998) The ubiquitinsystem. Annu Rev Biochem 67: 425—479
Huelsenbeck JP, Ronquist F (2001) MRBAYES:Bayesian inference of phylogenetic trees. Bioinfor-matics 17: 754—755
Karpov SA (1990) System of Protists. Omsk,Mezhvuzoskaya
Keeling PJ (2001) Foraminifera and Cercozoa arerelated in actin phylogeny: two orphans find a home?Mol Biol Evol 18: 1551—1557
Kuhn S, Medlin LK, Eller G (2004) Phylogeneticposition of the parasitoid nanoflagellate Pirsoniainferred from nuclear-encoded small subunit riboso-mal DNA and a description of Pseudopirsonia n. gen.and Pseudopirsonia mucosa (Drebes) comb. nov.Protist 155: 143—156
Longet D, Archibald JM, Keeling PJ, Pawlowski J(2003) Foraminifera and Cercozoa share a commonorigin according to RNA polymerase II phylogenies.Int J Syst Evol Microbiol 53: 1735—1739
Longet D, Burki F, Flakowski J, Berney C, Polet S,Fahrni J, Pawlowski J (2004) Multigene evidencefor close evolutionary relationships between Gromiaand Foraminifera. Acta Protozoologica 43: 303—311
Lopez-Garcıa P, Rodrıguez-Valera, Moreira D(2002) Toward the monophyly of Haeckel’s Radiolar-ia: 18S rRNA environmental data support thesisterhood of Polycystinea and Acantharea. Mol BiolEvol 19: 118—121
Martin W, Stoebe B, Goremykin V, Hansmann S,Hasegawa M, Kowallik KV (1998) Gene transfer tothe nucleus and the evolution of chloroplasts. Nature393: 162—165
Nikolaev SI, Berney C, Fahrni J, Mylnikov AP,Aleshin VV, Petrov N, Pawlowski J (2003) Gym-nophrys cometa and Lecythium sp. are core Cerco-zoa: evolutionary implications. Acta Protozool 42:183—190
Nikolaev SI, Berney C, Fahrni J, Bolivar I, Polet S,Mylnikov AP, Aleshin VV, Petrov NB, Pawlowski J(2004) The twilight of Heliozoa and rise of Rhizaria: a
molecular revolution in the phylogeny of amoeboidprotists. Proc Natl Acad Sci USA 101: 8066—8071
Pawlowski J, Berney C (2003) Episodic Evolution ofNuclear Small Subunit Ribosomal RNA Gene in theStem Lineage of Foraminifera. In Donoghue PC,Smith MP (eds) Telling the Evolutionary Time:Molecular Clocks and the Fossil Record. Systema-tics Association Special Volume.No.66.. Taylor andFrancis, London, pp 107—118
Pawlowski J, Holzmann M (2002) Molecular phy-logeny of Foraminifera—a review. Europ J Protistol38: 1—10
Philippe H, Adoutte A (1998) The MolecularPhylogeny of Eukaryota: Solid Facts and Uncertain-ties. In Coombs G, Vickerman K, Sleigh M, Warren A(eds) Evolutionary Relationships among Protozoa.Chapman and Hall, London, pp 25—56
Philippe H, Lopez P, Brinkmann H, Budin K,Germot A, Laurent J, Moreira D, Muller M, LeGuyader H (2000) Early branching or fast evolvingeukaryotes? An answer based on slowly evolvingpositions. Proc R Soc Ser, B 267: 1213—1222
Polet S, Berney C, Fahrni J, Pawlowski J (2004)Small-subunit ribosomal RNA gene sequences ofPhaeodarea challenge the monophyly of Haeckel’sRadiolaria. Protist 155: 53—63
Posada D, Crandall KA (1998) MODELTEST: testingthe model of DNA substitution. Bioinformatics 14:817—818
Sharp PM, Li W-H (1987) Molecular evolution ofubiquitin genes. Trends Ecol Evol 2: 328—332
Stechmann A, Cavalier-Smith T (2002) Rooting theeukaryote tree by using a derived gene fusion.Science 297: 89—91
Stechmann A, Cavalier Smith T (2003) The root ofthe eukaryote tree pinpointed. Curr Biol 13:R665—R666
Sun L, Chen ZJ (2004) The novel functions ofubiquitination in signalling. Curr Opin Cell Biol 16:119—126
Swofford DL (2003) PAUP*: Phylogenetic Analysisusing Parsimony (* and other methods). Version 4.Sinauer Associates, Sunderland, MS
