UBC Faculty Research and Publications

The dinoflagellates Durinskia baltica and Kryptoperidinium foliaceum retain functionally overlapping… Imanian, Behzad; Keeling, Patrick J Sep 24, 2007

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
52383-12862_2007_Article_463.pdf [ 386.8kB ]
Metadata
JSON: 52383-1.0223423.json
JSON-LD: 52383-1.0223423-ld.json
RDF/XML (Pretty): 52383-1.0223423-rdf.xml
RDF/JSON: 52383-1.0223423-rdf.json
Turtle: 52383-1.0223423-turtle.txt
N-Triples: 52383-1.0223423-rdf-ntriples.txt
Original Record: 52383-1.0223423-source.json
Full Text
52383-1.0223423-fulltext.txt
Citation
52383-1.0223423.ris

Full Text

ralssBioMed CentBMC Evolutionary BiologyOpen AcceResearch articleThe dinoflagellates Durinskia baltica and Kryptoperidinium foliaceum retain functionally overlapping mitochondria from two evolutionarily distinct lineagesBehzad Imanian and Patrick J Keeling*Address: Canadian Institute for Advanced Research, Department of Botany, University of British Columbia, 3529-6270 University Boulevard, Vancouver, British Columbia V6T 1Z4, CanadaEmail: Behzad Imanian - ashkboos@can.rogers.com; Patrick J Keeling* - pkeeling@interchange.ubc.ca* Corresponding author    AbtractBackground: The dinoflagellates Durinskia baltica and Kryptoperidinium foliaceum are distinguishedby the presence of a tertiary plastid derived from a diatom endosymbiont. The diatom is fullyintegrated with the host cell cycle and is so altered in structure as to be difficult to recognize it asa diatom, and yet it retains a number of features normally lost in tertiary and secondaryendosymbionts, most notably mitochondria. The dinoflagellate host is also reported to retainmitochondrion-like structures, making these cells unique in retaining two evolutionarily distinctmitochondria. This redundancy raises the question of whether the organelles share any functionsin common or have distributed functions between them.Results: We show that both host and endosymbiont mitochondrial genomes encode genes forelectron transport proteins. We have characterized cytochrome c oxidase 1 (cox1), cytochromeoxidase 2 (cox2), cytochrome oxidase 3 (cox3), cytochrome b (cob), and large subunit of ribosomalRNA (LSUrRNA) of endosymbiont mitochondrial ancestry, and cox1 and cob of host mitochondrialancestry. We show that all genes are transcribed and that those ascribed to the host mitochondrialgenome are extensively edited at the RNA level, as expected for a dinoflagellate mitochondrion-encoded gene. We also found evidence for extensive recombination in the host mitochondrialgenes and that recombination products are also transcribed, as expected for a dinoflagellate.Conclusion: Durinskia baltica and K. foliaceum retain two mitochondria from evolutionarily distinctlineages, and the functions of these organelles are at least partially overlapping, since both expressgenes for proteins in electron transport.BackgroundThe endosymbiotic origins of plastids and mitochondriashare a number of characteristics in common, [1,2], butdiffer in the complexity of their evolutionary history fol-lost [3-5], plastids have spread between eukaryotic line-ages several times in events referred to as secondary andtertiary endosymbioses. Generally these secondary andtertiary endosymbionts have degenerated so far that allPublished: 24 September 2007BMC Evolutionary Biology 2007, 7:172 doi:10.1186/1471-2148-7-172Received: 30 April 2007Accepted: 24 September 2007This article is available from: http://www.biomedcentral.com/1471-2148/7/172© 2007 Imanian and Keeling; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 11(page number not for citation purposes)lowing their origin and initial integration. Whereas mito-chondria originated once and have apparently never beenthat remains is a plastid with extra membranes [6], but ina few exceptional cases intermediate stages of reductionBMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172are known, and these may provide interesting glimpsesinto how complexity is lost.One of the characters that is absent from nearly all knownexamples of secondary and tertiary endosymbionts is themitochondrion. This contrasts with the fact that mito-chondria have never been lost in any other eukaryotic lin-eage. Even in the most severely reduced, anaerobicparasites which lack oxidative phosphorylation, highlyreduced organelles called mitosomes and hydrogeno-somes are found [3-5]. Some of these have no direct rolein energy metabolism, but iron-sulfur cluster biosynthesisis a common function [5,7,8]. These relict organelles sug-gest mitochondria are resistant to outright loss, raisingquestions about why mitochondria appear to be one ofthe more dispensable features of algae taken up duringsecondary and tertiary endosymbiosis events.The single clear exception to this is found in a group ofrelated dinoflagellates that harbour a diatom tertiaryendosymbiont. This group contains several species (see[9] for a recent summary), and here we have examinedtwo: Durinskia baltica [10] and Kryptoperidinium foliaceum[11,12]. Several of these genera (including Durinskia andKryptoperidinium) have been shown to share a commonpennate diatom endosymbiont, arguing that the endo-symbiosis is stable through evolutionary time [13,14].Interestingly, this may not hold for the whole group, sincethe endosymbiont of Peridinium quinquecorne is a centricdiatom [15], suggesting that the integration may havespanned a long period of time and different transientendosymbionts were ultimately fixed in these two sub-groups. Nevertheless, the endosymbionts of D. baltica andK. foliaceum are no longer transient in the short term, theyhave lost motility and cell wall and, although some chro-matin condensation occurs during sexual reproduction inD. baltica, typical chromosomes are not found within theendosymbiont nucleus at any stage of its life cycle [16,17].During the endosymbiotic nuclear division, neither aspindle apparatus nor any microtubules have beenobserved [18], and the amitotic division of this nucleusresults in unequal daughter nuclei and significantly largeramount of DNA in the nucleus than that reported in otherdiatoms [19].The endosymbiont has clearly been reduced in manyways, but some of its most interesting characteristics arewhat it has retained. This includes plastids surrounded byendoplasmic reticulum (ER) that is continuous with theouter membrane of the nucleus, a plasma membrane thatseparates it from the host cytoplasm, a multi-lobed, prom-inent nucleus with a genome, ribosomes, dictyosomes,and mitochondria [16,20].It is the retention of mitochondria that makes this endo-symbiont stand out, in particular since D. baltica and K.foliaceum host cells have also been reported to retain mito-chondria. The loss of endosymbiont mitochondria in vir-tually all known examples of secondary and tertiaryendosymbiotic events suggests retaining two mitochon-dria is either unnecessary or even deleterious. The loss ofone of these organelles may be ongoing, but it is also pos-sible that both compartments require mitochondrialfunction or that they have distributed essential functionsbetween them. We have previously shown [9] that the K.foliaceum endosymbiont mitochondrion contains agenome and expresses genes for cytochrome c oxidasesubunit 1 (cox1), cytochrome c oxidase subunit 3 (cox3),and cytochrome b (cob). However, no data are availablefrom the host mitochondrion, and with the function ofthis organelle completely unknown we cannot address thepossibility that the two organelles have overlapping or dif-ferentiated function.In order to determine whether this unique mitochondrialredundancy extends to the functional level, we character-ized seven mitochondrial genes of D. baltica: five from theendosymbiont (cox1, cox2, cox3, cob, and LSUrRNA) andtwo from the host (cox1 and cob), and confirmed that cox2and LSUrRNA from the endosymbiont and cox1 from thehost are also present in K. foliaceum. Most significantly, inD. baltica, cox1 and cob are present and expressed in bothmitochondria, and those in the host are heavily edited, asexpected for a functional dinoflagellate mitochondrialgene [21-23]. All available data therefore suggest that boththe host and endosymbiont mitochondria are activelyexpressing genes functional in oxidative phosphorylationand energy production.Results and discussionCharacterization of endosymbiont mitochondrial genes and transcriptsPCR amplification using diatom-specific primers andtotal D. baltica DNA (or RNA, see below), resulted in frag-ments of the expected sizes for five genes: cox1, cox3, cob,cox2, and LSUrRNA. Sequencing multiple clones yielded asingle copy of each gene. Only a short fragment of cox1could be amplified from DNA, probably due to the pres-ence of long type II introns such as those found in diatomsPhaeodactylum tricornutum and Thalassiosira pseudonanaand the endosymbiont of K. foliaceum [9], so the remain-der was recovered from RNA by RT-PCR. Diatom-derivedgenes for cox1, cox3 and cob are already known from K.foliaceum [9], so to complement the D. baltica data we alsoamplified the K. foliaceum cox2 and LSUrRNA.The phylogenies of all five genes (Fig. 1, 2, 3, 4, 5) gener-Page 2 of 11(page number not for citation purposes)ally resembled trees based on nuclear genes, with rela-tively strong support for the monophyly of alveolates andBMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172Page 3 of 11(page number not for citation purposes)Protein maximum likelihood phylogeny of cytochrome c oxidase 1 (cox1)Figure 1Protein maximum likelihood phylogeny of cytochrome c oxidase 1 (cox1). Numbers at nodes indicate bootstrap support for major nodes over 50% from ML (top) and distance (bottom). A dash (-) indicates support less than 50%. Major groups are labeled to the right, with diatoms (red) and dinoflagellates (purple) indicated by a box and D. baltica and K. foliaceum genes in black.Cafeteria roenbergensisChrysodidymus synuroideuPylaiella littoralisLaminaria digitataUndaria pinnatifidaPseudochorda nagaiiChorda filumDesmarestia viridisFucus vesiculosusMischococcus sphaerocephalusTribonema aequaleBotrydium granulatumBotrydiopsis alpinaChattonella antiquaOphiocytium majusMonodus sp. CCMP505Rhizosolenia setigeraMelosira ambiquaFragilaria striatulaThalassionema nitzschioidesNitzschia frustulumCylindrotheca closteriumPhaeodactylum tricornutumThalassiosira pseudonanaDitylum brightwelliiCryptomonas ovataRhodomonas salinaPhaeocystis pouchetiiIsochrysis galbanaEmiliania huxleyiDiacronema vlkianumPavlova lutheriSaprolegnia feraxPhytophthora infestansPythium aphanidermatumHaptophyteCryptomonadsOomycetesEustigmatophyteYellow-Green AlgaRaphidophyteYellow-GreenAlgaeBrown AlgaeBicosoecidGolden Alga100  99100  85100100100100100  99100 99  - 62995188725992989199  -1001008195815652 -Paramecium aureliaTetrahymena thermophila0.2Ciliates9595100  -66  -Kryptoperidinium foliaceumDiatomsDurinskia balticaPlasmodium gonderiTheileria parvaTheileria annulataCrypthecodinium cohniiPeridinium willeiProrocentrum micansSymbiodinium sp. ZsKarenia brevisKarlodinium micrumGyrodinium galatheanumHeterocapsa circularisquamaApicomplexa56825591100 60100 9797 -Kryptoperidinium foliaceumDurinskia balticaDinoflagellatesBMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172sister relationship between dinoflagellates and apicompl-exans [24,25]. We also noted that haptophytes and cryp-tophytes are sister groups with strong support in cob trees(Fig. 4) and weakly so in cox1 and cox3 trees (Fig. 1 and 3),as has been suggested in other analyses [26-29]. Mostimportantly, however, in all five phylogenies the distinc-tion between the expected positions of host and endosym-biont-derived genes was unambiguous, and in all fivetrees the D. baltica gene amplified with diatom-specificprimers branched within the diatom clade with strongsupport (Fig. 1, 2, 3, 4, 5). Moreover, D. baltica consist-ently grouped with K. foliaceum with strong support (withthe exception of distance analysis of cox1, where the over-lap between the two sequences is relatively short). This isin disagreement with the proposal that these dinoflagel-lates are products of separate endosymbiotic events [30],but consistent with the analyses of nuclear small-subunitrRNA genes from the hosts in these two dinoflagellates[14] and the hypothesis that these two species, and mostlikely their other close relatives, resulted from a singletoms (i.e. Nitzschia or Cylindrotheca, Phaodactylum), whichis also consistent with evidence that the endosymbiont isa descendent of a pennate diatom [9,14,33,34]. Overall,these trees strongly support the conclusion that cox1, cox3,cob, cox2, and LSUrRNA are all present in the mitochon-drial genome of the endosymbiont.To confirm that all five genes are actively expressed, eachwas also amplified from RNA using RT-PCR. All diatom-derived cDNA sequences were identical to their corre-sponding genes, providing further evidence for thesegenes being from the endosymbiont mitochondrion,because the mitochondrial transcripts in dinoflagellatesare extensively edited [22,35,36].Characterization of host mitochondrial genes and transcriptsThe presence and expression of cox and cob genes in theendosymbiont mitochondria suggests this organelle isengaged in electron transport. The pressing question isProtein maximum likelihood phylogeny of cytochrome oxidase 2 (cox2)Figure 2Protein maximum likelihood phylogeny of cytochrome oxidase 2 (cox2). Numbers at nodes indicate bootstrap sup-port for major nodes over 50% from ML (top) and distance (bottom). A dash (-) indicates support less than 50%. Major groups are labeled to the right, with diatoms (red) and dinoflagellates (purple) indicated by a box and D. baltica and K. foliaceum genes in black.Saprolegnia feraxPythium heterothallicumPhytophthora lateralisEmiliania huxleyiThraustochytrium aureumOchromonas danicaC.synuroideusRhodomonas salinaDictyota dichotomaThalassiosira pseudonanaPhaeodactylum tricornutum0.2949199969910092938677Tetrahymena pyriformisParamecium aurelia9595CiliatesOomycetesHaptophyteLabyrinthulidGolden AlgaeBrown AlgaCryptophyte5652Toxoplasma gondiiPlasmodium falciparumTheileria annulataKarlodinium micrumOxyrrhis marina825062686251ApicomplexaDinoflagellatesDurinskia balticaKryptoperidinium foliaceumDiatomsPage 4 of 11(page number not for citation purposes)endosymbiotic event [13,15,31,32]. Together, D. balticaand K. foliaceum branched specifically with pennate dia-therefore the nature of the host organelle, but no datahave been gathered from it for comparison. DinoflagellateBMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172mitochondrial genomes only encode three protein-codinggenes: cox1, cox3, and cob [23]. The LSU rRNA has beenextensively fragmented and rearranged [37], and cox2 hasbeen split and moved to the nucleus [38], so only cox1,cox3, and cob were sought in the mitochondrial genome.We used dinoflagellate-specific primers for all three genesin RT-PCR with total D. baltica RNA, and identified frag-ments of cox1 and cob. Using 3' RACE, we also recoveredthe 3' end of the D. baltica cob gene. We also recovered 6additional copies of cox1 transcripts (Fig. 6), each ofwhich contained inserts that differed in position, size, andsequences, and disrupted the reading frame. Insertsranged from 81 to 453 bp. Two inserts at slightly differentpositions of RNA2 and RNA3 were similar in size andsequences, and both of these contained a 151 bp portionof the cob gene flanked by two small (about 20 bp) non-coding fragments (Fig. 6). Another insert contained a 75(Genbank accession: AB265207) Gonyaulax polyedra (Gen-bank accession: AF142472). We did not sequence a DNAclone lacking an insert, but the mitochondrial genome ofother dinoflagellates is known to contain many copies ofeach gene and many rearrangements [21,23], so the intactcopy of the gene was most likely simply not sampled.Given the highly fragmented and divergent nature ofdinoflagellate mitochondrial rRNAs [37], it is possiblethis 75 bp represents an as yet unidentified fragment ofeither the LSU or SSU.PCR using genomic DNA from D. baltica resulted in a sin-gle cob gene fragment (12 identical clones weresequenced), and six different cox1 fragments (seven differ-ent clones were sequenced). As with cDNAs, each cox1gene contained a unique insert (Fig. 6). Most inserts wereunique, but many contained small imperfect repeats inProtein maximum likelihood phylogeny of cytochrome oxidase 3 (cox3)Figure 3Protein maximum likelihood phylogeny of cytochrome oxidase 3 (cox3). Numbers at nodes indicate bootstrap sup-port for major nodes over 50% from ML (top) and distance (bottom). A dash (-) indicates support less than 50%. Major groups are labeled to the right, with diatoms (red) and dinoflagellates (purple) indicated by a box and D. baltica and K. foliaceum genes in black.Theileria parvaTheileria annulataPlasmodium falciparumKarenia brevisCafeteria roenbergensisC.synuroideusOchromonas danicaLaminaria digitataUndaria pinnatifidaPylaiella littoralisDesmarestia viridisFucus vesiculosusPhaeodactylum tricornutumDurinskia balticaKryptoperidinium foliaceumThalassiosira pseudonanaSaprolegnia feraxPhytophthora infestansRhodomonas salinaEmiliania huxleyi0.551 -10085100891008210072998410085657192  -99  -655010095DiatomsApicomplexansDinoflagellateBicosoecidGolden AlgaeBrown AlgaeOomycetesHaptophyteCryptomonadPage 5 of 11(page number not for citation purposes)bp fragment with over 90% identity to non-coding frag-ments from the dinoflagellates Alexandrium catenellacommon, and the positions of the inserts within cox1 werevariable and inevitably a portion of cox1 was missing atBMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172the point of insertion. One pair of DNA and cDNA clones(Fig. 6, DNA1 and RNA3) were found to be identical, withthe exception of edited sites (see below).A single transcript of the K. foliaceum cox1 was also recov-ered from sequencing 15 identical clones. This transcriptalso contained an insert, but without sequence similarityto anything known and not at a position in common withany D. baltica clone. We failed to identify cob, but consid-ering its presence in D. baltica we feel it is unlikely to beThe many variants of the D. baltica host cox1 gene and theinsert in the K. foliaceum cox1 gene are both consistentwith the nature of mitochondrial genomes in other dino-flagellates. This has been best described in Crypthecodin-ium cohni and Amphidinium carterae, where protein-codinggenes are flanked by non-coding, repeat-rich sequencesand that the context of a gene can vary in different copiesdue to homologous recombination [21,39], and in Oxyr-rhis marina where protein coding genes are found in differ-ent genomic contexts, and are often fragmented [23].Protein maximum likelihood phylogeny of cytochrome b (cob)Figure 4Protein maximum likelihood phylogeny of cytochrome b (cob). Numbers at nodes indicate bootstrap support for major nodes over 50% from ML (top) and distance (bottom). A dash (-) indicates support less than 50%. Major groups are labeled to the right, with diatoms (red) and dinoflagellates (purple) indicated by a box and D. baltica and K. foliaceum genes in black.Paramecium aureliaTetrahymena thermophilaHaemoproteus sylvaePlasmodium falciparumCrypthecodinium cohniiSymbiodinium microadriaticumProrocentrum minimumKarlodinium micrumPeridinium aciculiferumKarenia brevisOchromonas danicaChrysodidymus synuroideusLaminaria digitataPylaiella littoralisDesmarestia viridisFucus vesiculosusPhaeodactylum tricornutumThalassiosira pseudonanaSaprolegnia feraxPhytophthora infestansCafeteria roenbergensisRhodomonas salinaEmiliania huxleyi0.5DinoflagellatesApicomplexaCiliatesBrown AlgaeGolden AlgaeDiatomsOomycetesBicosoecidHaptophyteCryptomonad937480  -100100100 92100 829555100 9994888073  -6877985297  -100 56100 94100549594Durinskia balticaKryptoperidinium foliaceumDurinskia balticaPage 6 of 11(page number not for citation purposes)absent in K. foliaceum.BMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172Host mitochondrial transcripts are extensively editedDinoflagellate mitochondria possess a distinctive form ofRNA editing. Editing sites typically involve A to G, T to C,and C to U changes at first and second positions, affectingabout 2% of positions in cox1 and cob genes [22,35,40].The presence of such editing would provide further evi-dence for the dinoflagellate mitochondrial location of thecox1 and cob genes identified here, so all D. baltica genefragments were aligned to their respective transcripts andconserved editing sites examined. In total, 786 bp of cox1were comparable, and 352 bp of cob were comparable,from which 11 and 7 edited sites were identified, respec-tively. The nature of these edits was similar to that of otherdinoflagellates (Table 1) and, significantly, all but three ofthe editing sites were conserved in other dinoflagellatemitochondria [22,35,40]. Overall, the characteristics ofthis editing are consistent with these genes and cDNAsbeing located in the dinoflagellate host mitochondria.Reduction and functional redundancy of mitochondriaElectron microscopy has shown that mitocondria exist inboth the host and endosymbiont cytosolic compartmentsof D. baltica and K. foliaceum [11,12,16]. More recentlydiatom-derived genes for cox1, cox3, and cob have beenshown to be expressed in K. foliaceum [9]. However, with-out comparable data from the host mitochondria it isimpossible to determine whether the two organelles arefunctionally redundant, or have distributed functionsbetween them. Here, we have shown that both organellescontain at least two genes with central functions in elec-tron transport, cox1 and cob. Accordingly, these two spe-Schematic representation of cox1 gene and cDNA fragments characterised from Durinskia balticaFigure 6Schematic representation of cox1 gene and cDNA fragments characterised from Durinskia baltica. The black rec-tangles represent coding regions of the gene or the transcript. The white rectangles represent the inserts. The inserts that contain a fragment of another gene have been represented by gray rectangles. The scale is proportional to the number of cobcobrRNA?cobRNA1RNA0RNA6RNA5RNA4RNA3RNA2DNA6DNA5DNA4DNA3DNA2DNA1Maximum likelihood phylogeny of large subunit ribosomal RNA (LSUrRNA) un er Gen ral Time-Reversible (GTR) model of substitutionFigure 5Maximum likelihood phylogeny of large subunit ribosomal RNA (LSUrRNA) under General Time-Reversible (GTR) model of substitution. Numbers at nodes indicate bootstrap support for major nodes over 50% from ML (top) and distance (bottom). A dash (-) indicates support less than 50%. Major groups are labeled to the right, with diatoms (red) indicated by a box and D. baltica and K. foliaceum genes in black.Paramecium primaureliaTetrahymena pigmentosaChrysodidymus synuroideusOchromonas danicaCafeteria roenbergensisPhytophthora infestansSaprolegnia feraxThalassiosira pseudonanaPhaeodactylum tricornutumUndaria undarioidesLaminaria digitataPilayella littoralisDesmarestia viridisFucus vesiculosusDictyota dichotoma0.1OomycetesBicosoecidBrown AlgaeGolden AlgaeCiliates10010010010084  -7375100100957710010010010087915786100959595Durinskia balticaKryptoperidinium foliaceum DiatomsPage 7 of 11(page number not for citation purposes)nucleotides.BMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172cies are unique among eukaryotes in having retainedactive, functional mitochondrial genomes from two dis-tantly related eukaryotic lineages, the dinoflagellate hostand the pennate diatom endosymbiont. Moreover, theseorganelles now appear to be at least partially functionallyredundant, since both express genes for proteins in theelectron transport chain.Indeed, no characteristic of either host or endosymbiontmitochondrial genes or genomes has so far been shown tobe significantly different from those of their dinoflagellateor diatom relatives, which points to the conclusion thatneither organelle has been much affected by the presenceof the other. It seems unlikely that the two mitochondriaare retained due to functional differentiation, but theirgenetic redundancy may be related to spatial differentia-tion. If the membrane separating the host and the endo-symbiont, which is thought to be derived from the diatomplasma membrane [20], were deficient in transporters toefficiently shuttle either substrates or products of mito-chondrial reactions between the two compartments, thenneither compartment could eliminate those functionswithout consequences. In other endosymbiotic eventssuch difficulties would have been overcome or made irrel-evant by the continued reduction of the endosymbiontand integration with the host. Whether the mitochon-drion of D. baltica is a snapshot in the progression oflarly, although it is generally assumed that the endosym-biont will be reduced and the corresponding host featureretained, this does not need to be the case. With thealready highly reduced and unusual nature of dinoflagel-late mitochondrial genome [21,23,39], it is not unreason-able to hypothesize that the host organelle may be lost aseasily as the relatively normal mitochondrion of the endo-symbiont.ConclusionWe have shown that two related dinoflagellates, D. balticaand K. foliaceum, retain redundant genes in their host-derived and endosymbiont-derived mitochondrialgenomes, including several genes related to electron trans-port. Host-derived genes are edited at the RNA level andsubject to extensive recombination, as is expected fordinoflagellate mitochondria. All genes characterized havebeen shown to be expressed at the mRNA level, suggestingthe two organelles overlap in function, making theseunique among eukaryotes in retaining two partiallyredundant mitochondria with different evolutionary ori-gins.MethodsCulture conditions and nucleic acid extractionCultures of Durinskia baltica (Peridinium balticum) CS-38were obtained from CSIRO Microalgae Supply ServiceTable 1: Editing sites in the host mitochondrial mRNA of cox1 and cob in Durinskia balticaSite Relative to Cc or Pp DNA RNA Codon Position Change aa Conserved incox1 330 A G 1st I – V Pp and Cccox1 351 T C 1st F – L Pp and Cccox1 469 C U 2nd T – I Unique1cox1 481 C U 2nd S – F Pp and Cccox1 495 T C 1st F – L Pp and Cccox1 621 A G 1st I – V Pp and Cccox1 691 A G 2nd Y – C Ppcox1 924 A G 1st I – V Pp and Cccox1 952 T C 2nd L – S Pp and Cccox1 1174 A G 2nd K – R Pp and Cccox1 1180 A G 2nd N – S Pp and Cccob 782 T C 2nd V – A Pp, Cc, Pmic, Km, Pmin, Ps, Atcob 788 G C 2nd G – A Pp, Pmic, Km, Pmin, Ps, Atcob 861 C U 3rd Silent Pmic, Pmin, Kmcob 883 A G 1st I – V Pmic, Pmin, K.m.cob 904 A G 1st T – A Pmic and Pmincob 1064 G C 2nd G – A Unique2cob 1081 C U 1st H – V Unique2Column 1 is the site numbered according to genes from Crypthecodinium cohnii (cox1) or Pfiesteria piscicida (cob). Column 2 and 3 are pre-edited and post-edited states in D. baltica. Column 4 is the position within the codon. Column 5 is the amino acid change. Column 6 lists other dinoflagellates where the same editing is found. Abbreviations: Pp, Pfiesteria piscicida; Cc, Crypthecodinium cohnii; Pmic, Prorocentrum micans; Prorocentrum minimum; Km, Karlodinium micrum; Ps, Pfiesteria shumwyae; At, Alexandrium tamarense.1 This change is absent in Pp and Cc.2 These changes are absent in Pp and not sampled from Cc.Page 8 of 11(page number not for citation purposes)events that will ultimately lead to its loss, or whether theprocess has been 'stuck' in some way is unknown. Simi-(CSIRO Marine and Atmospheric Research Laboratories,Tasmania, Australia) and maintained in GSe medium atBMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/17222°C (12 : 12 light : dark cycle). Cultures of Kryptoperidin-ium foliaceum CCMP 1326 were obtained from the Prova-soli-Guillard National Center for Culture of MarinePhytoplankton (West Boothbay Harbor, ME, USA) andmaintained in F/2-Si medium under the above-men-tioned conditions. Cultures were grown both with andwithout antibiotics to reduce the number of bacteria: 500µg/ml penicillin G, 200 µg/ml ampicillin, 50 µg/ml strep-tomycin sulphate, and 50 µg/ml neomycin, modifiedfrom [19]. Cultures used in some molecular experimentsdid not contain antibiotics, while others did. Exponen-tially growing cells were harvested by centrifugation at3,220 g for 5 min at 8°C, and the pellet was frozen andground under liquid nitrogen. The total genomic DNAwas extracted from about 100 mg of the ground cells usingDNeasy Plant DNA isolation kit (Qiagen, Mississauga,ON). Total RNA was isolated using TRIzol Reagent (Invit-rogen, Burlington, ON) from the pelleted cells followingmanufacturer's instructions, and it was treated with Deox-yribonuclease I (Invitrogen). PCR was carried out usingPuReTaq (Amersham Biosciences, Baie d'Urfé, QC) andlong range PCR using Elongase Enzyme Mix (Invitrogen).RT-PCR was carried out using SuperScript III One-StepSystem with Platinum Taq DNA Polymerase (Invitrogen).Amplification and sequencing of mitochondrial genes and cDNAsFrom genomic DNA of D. baltica, for amplification of theendosymbiont genes we used the following primers: forLSUrRNA gene, 5'-TTCTGCGAAATCTATTKAAGTA-GAGCG-3' and 5'-CYGGCGTACCTTTTATCCRTTGMGC-3'; for cob gene, 5'-CCCTTACAGCAATTCCATTCGGAG-GTCAAA-3' and 5'-TTCGCCCTTCTGGAATACAATTAT-CAGGAT-3'; for cox3 gene, 5'-TTACAGGTGGTGTTCTTTATATGCACAAAA-3' and 5'-AGCCGAAGTGGTGGGGTATTTGTTGAGTGGT-3'; forcox2 gene, first we used the following two degenerateprimers, 5'-ATCGGGCATCAGTGGTAYTGGWSNTAYGA-3' and 5'-GTTTATCCCGCAGATYTCNSWRCAYTGNCC-3'and later the following specific primers, 5'-GTATTGGAG-GTACGAGATTTCGGACTTTGA-3' and 5'-CGGAGCACT-GACCAAAGAACATACCCACA-3'; for cox1 gene, 5'-GTTGTTACCCACCTTCTCTTTTACTACTGAT-3' and 5'-GCAACAACGTAATAAGTATCGTGAGGAGCA-3'. Foramplification of the host genes in D. baltica from genomicDNA, and all the cox1 products containing an insert, weused the following primers: for cox1, first we used the fol-lowing two primers previously described [35], 5'-AAAAATTGTAATCATAAACGCTTAGG-3' and 5'-TGTT-GAGCCACCTATAGTAAACATTA-3', and later the follow-ing two specific primers, 5'-GCACTTCTTTCATGAGTTTATCACCTTCAAG-3' and 5'-TTCTGAGCTGTAACAATGGCGGATTCCCA-3'; for cob,TGCCTAACAAAAATGCAGGATTCATAGTCT-3', and laterthe following primer was used to amplify the 3' end of thegene in 3' RACE using RLM-RACE kit (Ambion, Austin,TX, USA) following the manufacturer's instructions, 5'-GCATTAGAAGCTTGTGCATTACTTACTCCT-3'. Fromgenomic DNA of K. foliaceum, for amplification of theendosymbiont genes we used the following primers: forLSUrRNA gene, 5'-AACAGACAGTCCATGAGTGCTAA-GATTCAT-3', and 5'-CACACAGAATTACCGGATCAC-TATAACCGA-3'; for cox2 gene, we first used the followingtwo degenerate primers, 5'-GGGCATCAGTGGTATT-GGWSNTAYGARWW-3' and 5'-GTTTATCCCGCA-GATYTCNSWRCAYTGNCC-3', and later the followingtwo specific primers, 5'-GGGCATCAGTGGTATTGGTGG-TACGAAAT-3' and 5'-GTTTATCCCGCAGATTTCGCT-GCACTGGCC-3'. For amplification of the host cox1 in K.foliaceum from total RNA, we used the two previouslydescribed primers [35], 5'-AAAAATTGTAATCAT-AAACGCTTAGG-3' and 5'-TGTTGAGCCACCTATAG-TAAACATTA-3', using RT-PCR. Transcripts of all the geneswere characterized by RT-PCR using the same primers,and all these amplifications were carried out with controlslacking RT enzyme, from which no products wereacquired.We also used three pairs of dinoflagellate-specific primersto search for the host cob in K. foliaceum, which were basedon the most conserved regions of this gene found in dino-flagellate mitochondria. Two pairs of these primers weretested successfully to amplify this gene from D. baltica(data not shown). However, no product was obtainedwith any of these primers from the total DNA or RNAextracted from K. foliaceum used in PCR and RT-PCRrespectively.All PCR and RT-PCR products were gel purified andcloned using pCR 2.1 TOPO Cloning kit (Invitrogen). Ineach case, several clones were sequenced on both strandsusing BigDye terminator chemistry. New sequences havebeen deposited into GenBank as accessions EF434607–EF434629.Phylogenetic analysesThe conceptual translations of new cox1, cox2, cox3, andcob, and DNA sequences for LSUrRNA from D. baltica andK. foliaceum were aligned with homologues from publicdatabase using ClustalX 1.83.1 [41] under the default gapopening and gap extension penalties and the alignmentsedited manually. Phaeodactylum tricornutum homologueswere kindly provided by Marie-Pierre Oudot-Le Secq fromthe P. tricornutum genome sequencing project [42]. Phylo-genetic analyses were carried out including a diversity ofeukaryotes to determine the overall position of newPage 9 of 11(page number not for citation purposes)initially the following two primers were used, 5'-GGGGT-GCTACGGTTATTACGAACCTACTA-3' and 5'-sequences, and subsequently restricted to homologuesfrom chromalveolate taxa (dinoflagellates, apicomplex-BMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/172ans, ciliates, heterokonts, haptophytes, and cryptomon-ads), since both the host and endosymbiont are thoughtto be members of this supergroup [43]. No LSUrRNAsequences for dinoflagellates or apicomplexans wereincluded since these genes are fragmented, only partiallydescribed to date, and highly divergent, so ciliates alonerepresent alveolates. These alignments consisted of 52, 20,20, and 26 amino acid sequences, and 17 DNA sequenceswith 378, 109, 251, 372, and 1304 unambiguouslyaligned sites for cox1, cox2, cox3, cob, and LSUrRNA,respectively. For cox1 and cob sequences, several align-ment alternatives were attempted, which were independ-ently followed by phylogenetic analyses. These alternativealignments differed only in the inclusion or exclusion ofmissing sites. In order to make use of all the recoveredsequence data, the alignments of cox1 and cob that wereanalysed included some missing data and either the 5' or3' ends of certain D. baltica or K. foliaceum genes. Phylog-enies of cox1 and cob were also performed using excludingeither host or endosymbiont genes, resulting in no signif-icant differences (not shown). All alignments are availableupon request.Phylogenetic trees were inferred using maximum likeli-hood and distance methods. The proportion of invariablesites (i) and shape parameter alpha (α) with 8 variablerate categories were estimated from the data with PhyML2.4.4 [44] under the Whelan and Goldman (WAG) modelof substitution for cox1, cox2, cox3, and cob phylogenies,and under General Time-Reversible (GTR) model of sub-stitution for LSUrRNA phylogeny with the frequency ofamino acid or nucleotide usage calculated from the data.The i and α parameters estimated from the data were0.000, 0.057, 0.004, 0.010, and 0.113, and 1.002, 1.297,1.186, 1.402 and 1.002 for cox1, cox2, cox3, cob, and LSUr-RNA, respectively. For all five data sets 1,000 bootstrapreplicates were analyzed using PhyML. For distance trees,distances were calculated using TREE-PUZZLE 5.2 [45]with 8 variable rate categories and invariable sites. The iand α parameters were estimated by TREE-PUZZLE to be0.00, 0.05, 0.00, 0.00, and 0.10, and 0.94, 1.18, 1.03,1.12, and 0.91 for cox1, cox2, cox3, cob, and LSUrRNA,respectively. Trees were constructed by weighted neigh-bor-joining using WEIGHBOR 1.0.1a [46]. Distance boot-strapping of 1,000 replicates was carried out usingPUZZLEBOOT (shell script by A. Roger and M. Holder).Competing interestsThe author(s) declares that there are no competing inter-ests.Authors' contributionsBI characterized new molecular sequences, carried outpated in sequence analysis, and helped draft the manu-script. Both BI and PJK read and approved the finalmanuscript.AcknowledgementsThis work was supported by a grant from the Natural Sciences and Engi-neering Research Council of Canada to PJK. We would like to thank Marie-Pierre Oudot-Le Secq and Chris Bowler for providing mitochondrial sequences from P. tricornutum prior to publication, Ross Waller for sharing cox2 alignment, and Allen Larocque for technical assistance. PJK is a Fellow of the Canadian Institute for Advanced Research and a Senior Investigator of the Michael Smith Foundation for Medical Research.References1. Gray MW, Burger G, Lang BF: Mitochondrial evolution.  Science1999, 283:1476-1481.2. Palmer JD: The symbiotic birth and spread of plastids: Howmany times and whodunit?  J Phycol 2003, 39:4-11.3. Bui ETN, Bradley PJ, Johnson PJ: A common evolutionary originfor mitochondria and hydrogenosomes.  Proc Natl Acad Sci USA1996, 93:9651-9656.4. Roger AJ: Reconstructing early events in eukaryotic evolu-tion.  Am Nat 1999, 154:S146-S163.5. Williams BAP, Keeling PJ: Cryptic organelles in parasitic protistsand fungi.  Adv Parasitol 2003, 54:9-68.6. Archibald JM, Keeling PJ: Recycled plastids: A 'green movement'in eukaryotic evolution.  Trends Genet 2002, 18:577-584.7. Tachezy J, Sanchez LB, Muller M: Mitochondrial type iron-sulfurcluster assembly in the amitochondriate eukaryotes Tri-chomonas vaginalis and Giardia intestinalis, as indicated bythe phylogeny of IscS.  Mol Biol Evol 2001, 18:1919-1928.8. van der Giezen M, Tovar J: Degenerate mitochondria.  EMBO Rep2005, 6:525-530.9. Imanian B, Carpenter K J., Keeling PJ: The mitochondrial genomeof a tertiary endosymbiont retains genes for electron trans-port proteins.  J Eukaryot Microbiol 2007, 54:146-153.10. Carty S, Cox ER: Kansodinium new-genus and durinskia new-genus two genera of freshwater dinoflagellates pyrrophyta.Phycologia 1986, 25:197-204.11. Dodge JD: A dinoflagellate with both a mesokaryotic and aeukaryotic nucleus: Part 1 fine structure of the nuclei.  Proto-plasma 1971, 73:145-157.12. Jeffrey SW, Vesk M: Further evidence for a membrane boundendosymbiont within the dinoflagellate Peridiniumfoliaceum.  J Phycol 1976, 12:450-455.13. Tamura M, Shimada S, Horiguchi T: Galeidinium rugatum gen. etsp. nov. (Dinophyceae), a new coccoid dinoflagellate with adiatom endosymbiont.  J Phycol 2005, 41:658-671.14. Inagaki Y, Dacks JB, Doolittle WF, Watanabe KI, Ohama T: Evolu-tionary relationship between dinoflagellates bearing obligatediatom endosymbionts: Insight into tertiary endosymbiosis.Int J System Evol Microbiol 2000, 50:2075-2081.15. Horiguchi T, Takano Y: Serial replacement of a diatom endo-symbiont in the marine dinoflagellate Peridinium quin-quecorne (Peridinales, Dinophyceae).  Phycological Res 2006,54:193-200.16. Tomas RW, Cox ER: Observations on the symbiosis of Peridin-ium balticum and its intracellular alga.  J Phycol 1973, 9:304-323.17. Chesnick J, Cox ER: Fertilization and zygote development inthe binucleate dinoflagellate Peridinium balticum (Pyrrho-phyta).  Am J Bot 1989, 76:1060-1072.18. Tippit DH, Pickett-Heaps JD: Apparent amitosis in the binucle-ate dinoflagellate Peridinium balticum.  J Cell Sci 1976,21:273-289.19. Kite GC, Rothschild LJ, Dodge JD: Nuclear and plastid DNAsfrom the binucleate dinoflagellates Glenodinium foliaceumand Peridinium balticum.  Biosystems 1988, 21:151-164.20. Eschbach S, Speth V, Hansmann P, Sitte P: Freeze-fracture studyof the single membrane between host cell and endocytobi-ont in the dinoflagellates Glenodinium foliaceum and Perid-Page 10 of 11(page number not for citation purposes)phylogenetic analysis and drafted the manuscript. PJKconceived of the study, participated in its design, partici-inium balticum.  J Phycol 1990, 26:324-328.Publish with BioMed Central   and  every scientist can read your work free of charge"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."Sir Paul Nurse, Cancer Research UKYour research papers will be:available free of charge to the entire biomedical communitypeer reviewed and published immediately upon acceptancecited in PubMed and archived on PubMed Central BMC Evolutionary Biology 2007, 7:172 http://www.biomedcentral.com/1471-2148/7/17221. Norman JE, Gray MW: A complex organization of the geneencoding cytochrome oxidase subunit 1 in the mitochondrialgenome of the dinoflagellate Crypthecodinium cohnii:humologous recombination generates two different cox1open reading frames.  J Mol Evol 2001, 53:351-363.22. Zhang H, Lin S: Mitochondrial cytochrome b mRNA editing indinoflagellates: Possible ecological and evolutionary associa-tions?  Journal of Eukaryotic Microbiology 2005, 52:538-545.23. Slamovits CH, Saldarriaga JF, Larocque A, Keeling PJ: The highlyreduced and fragmented mitochondrial genome of the early-branching dinoflagellate Oxyrrhis marina shares characteris-tics with both apicomplexan and dinoflagellate mitochon-drial genomes.  J Mol Biol 2007, 372:356-268.24. Fast NM, Xue L, Bingham S, Keeling PJ: Re-examining alveolateevolution using multiple protein molecular phylogenies.  JEukaryot Microbiol 2002, 49:30-37.25. Van de Peer Y, De Wachter R: Evolutionary relationship amongthe eukaryotic crown taxa taking into account site-to-siterate variation in 18S rRNA.  J Mol Evol 1997, 45:615-630.26. Rice DW, Palmer JD: An exceptional horizontal gene transferin plastids: gene replacement by a distant bacterial paralogand evidence that haptophyte and cryptophyte plastids aresisters.  BMC Biol 2006, 4:31.27. Harper JT, Waanders E, Keeling PJ: On the monophyly of chro-malveolates using a six-protein phylogeny of eukaryotes.  IntJ System Evol Microbiol 2005, 55:487-496.28. Patron NJ, Inagaki Y, Keeling PJ: Multiple gene phylogenies sup-port the monophyly of cryptomonad and haptophyte hostlineages.  Curr Biol 2007, 17:887-891.29. Hackett JD, Yoon HS, Li S, Reyes-Prieto A, Rummele SE, BhattacharyaD: Phylogenomic analysis supports the monophyly of crypto-phytes and haptophytes and the association of 'Rhizaria' withchromalveolates.  Mol Biol Evol 2007, 24:1702-1713.30. Morris RL, Fuller CB, Rizzo PJ: Nuclear basic proteins from thebinucleate dinoflagellate Peridinium foliaceum (Pyrro-phyta).  J Phycol 1993, 29:342-347.31. Horiguchi T, Pienaar RN: Ultrastructure of a new marine sand-dwelling dinoflagellate, Gymnodinium quadrilobatum sp.nov. (Dinophyceae) with special reference to its endosymbi-otic alga.  Europ J Phycol 1994, 29:237-245.32. Horiguchi T, Pienaar RN: Ultrastructure of a marine dinoflagel-late, Peridinium quinquecorne Abe (Peridiniales) fromSouth Africa with special reference to its chrysophyte endo-symbiont.  Botanica Marina 1991, 34:123-131.33. Chesnick JM, Kooistra WHC, Wellbrock U, Medlin LK: RibosomalRNA analysis indicates a benthic pennate diatom ancestryfor the endosymbionts of the dinoflagellates Peridiniumfoliaceum and Peridinium balticum (Pyrrophyta).  J EukaryotMicrobiol 1997, 44:314-320.34. McEwan M, Keeling PJ: HSP90, tubulin and actin are retained inthe tertiary endosymbiont genome of Kryptoperidiniumfoliaceum.  J Eukaryot Microbiol 2004, 51:651-659.35. Lin S, Zhang H, Spencer DF, Norman JE, Gray MW: Widespreadand extensive editing of mitochondrial mRNAS in dinoflag-ellates.  J Mol Biol 2002, 320:727-739.36. Zhang H, Lin S: Mitochondrial cytochrome b mRNA editing indinoflagellates: possible ecological and evolutionary associa-tions?  J Eukaryot Microbiol 2005, 52:538-545.37. Gray MW: Diversity and evolution of mitochondrial RNAediting systems.  IUBMB Life 2003, 55:227-233.38. Kamikawa R, Inagaki Y, Sako Y: Fragmentation of mitochondriallarge subunit rRNA in the dinoflagellate Alexandrium cat-enella and the evolution of rRNA structure in alveolate mito-chondria.  Protist 2007, 158:239-245.39. Waller RF, Keeling PJ: Alveolate and chlorophycean mitochon-drial cox2 genes split twice independently.  Gene 2006,383:33-37.40. Nash EA, Barbrook AC, Edwards-Stewart RK, Bernhardt K, Howe CJ,Nisbet RE: Organisation of the mitochondrial genome in thedinoflagellate Amphidinium carterae.  Mol Biol Evol 2007,24:1528-1536.41. Gray MW, Lang BF, Burger G: Mitochondria of protists.  Annu RevGenet 2004, 38:477-524.42. Thompson JD, Higgins DG, Gibson TJ: Clustal W: improving thethrough sequence weighting, position specific gap penaltiesand weight matrix choice.  Nucleic Acids Res 1994, 22:.43. Phaeodactylum tricornutum genome sequencing project[http://www.jgi.doe.gov/index.html]44. Keeling PJ, Burger G, Durnford DG, Lang BF, Lee RW, Pearlman RE,Roger AJ, Gray MW: The tree of eukaryotes.  Trends Ecol Evol2005, 20:670-676.45. Guindon S, Gascuel O: A simple, fast and accurate method toestimate large phylogenies by maximum-likelihood.  Syst Biol2003, 53:696-704.46. Schaefer G, Anemueller S, Moll R: Archaeal complex II: 'Classical'and 'non-classical' succinate:quinone reductases with unu-sual features.  Biochim Biophys Acta 2002, 1553:57-73.47. Bruno WJ, Socci ND, Halpern AL: Weighted Neighbor Joining: ALikelihood-Based Approach to Distance-Based PhylogenyReconstruction.  Mol Biol Evol 2000, 17:187-197.yours — you keep the copyrightSubmit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.aspBioMedcentralPage 11 of 11(page number not for citation purposes)sensitivity of progressive multiple sequence alignment

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.52383.1-0223423/manifest

Comment

Related Items