sign in sign up
<<
Home , Algae, Plastid, Satellite (biology)

J. Phycol. 39, 259–267 (2003)

MINIREVIEW

THE MESOZOIC RADIATION OF EUKARYOTIC : THE PORTABLE HYPOTHESIS1

Daniel Grzebyk,2 Oscar Schofield Environmental Biophysics and Molecular Program, Institute of Marine and Coastal Sciences, Rutgers, The State University of New Jersey, 71 Dudley Road, New Brunswick, New Jersey 08901, USA Costantino Vetriani Department of Biochemistry and , and Institute of Marine and Coastal Sciences, Rutgers, The State University of New Jersey, 71 Dudley Road, New Brunswick, New Jersey 08901, USA and Paul G. Falkowski Environmental Biophysics and Molecular Ecology Program, Institute of Marine and Coastal Sciences and Department of Geology, Rutgers, The State University of New Jersey, 71 Dudley Road, New Brunswick, New Jersey 08901, USA

Although all appear to have been de- In eukaryotic algae, all are derived from a rived from a common ancestor, a major schism oc- common photoautotrophic prokaryotic ancestor and curred early in the of eukaryotic algae that were sequentially appropriated by various nonphoto- gave rise to red and green photoautotrophic lin- synthetic host cells through primary, secondary, and eages. In and earlier times, the fossil even tertiary endosymbiotic events (Wolfe et al. 1994, record suggests that oceanic eukaryotic phytoplank- Delwiche 1999, Tomitani et al. 1999). Based on bio- ton were dominated by the green (chl b-containing) chemical and ultrastructural features of their plastids algal line. However, following the end-Permian ex- (van den Hoek et al. 1995), eukaryotic algae can be tinction, a diverse group of eukaryotic phytoplank- clustered into two major groups: a “green” and a “red” ton evolved from secondary symbiotic associations lineage. The green lineage comprises the Kingdom in the red (chl c-containing) line and subsequently , which includes all (Chloro- rose to ecological prominence. In the contemporary phyta) and higher (Streptophyta), all of which , red eukaryotic taxa continue have primary . Two green algal phyla to dominate marine pelagic webs, whereas the have secondary endosymbiotic plastids (euglenophytes green line is relegated to comparatively minor eco- and ), but both contain relatively logical and biogeochemical roles. To help elucidate few species. In the red lineage, the per se why the oceans are not dominated by green taxa, we (Rhodophyta), which have plastids derived from a pri- analyzed and compared whole plastid in mary endosymbiotic event, contain relatively few ex- both the red and green lineages. Our results suggest tant taxa and very few unicellular members. In con- that whereas all algal plastids retain a core set of trast to the green lineage, most eukaryotic algal taxa , red plastids retain a complementary set of that contain plastids from the red line are products of genes that potentially confer more capacity to auton- secondary or tertiary endosymbioses. These include omously express regulating oxygenic photo- the (Dinophyta) and the Kingdom synthetic and energy transduction pathways. We hy- comprised, in part, of , crypto- pothesize that specific losses in the primary phytes, and (including , brown endosymbiotic green plastid reduced its portability algae, and ). for subsequent symbiotic associations. This corollary In Paleozoic and earlier eras, the fossil record of of the plastid “enslavement” hypothesis may have acritarchs is interpreted to contain taxa associated limited subsequent evolutionary advances in the with the extant prasinophyte green algae, suggesting green lineage while simultaneously providing a com- that members of the green lineage dominated eukary- petitive advantage to the red lineage. otic phytoplankton communities in the oceans (Tap- Key index words: endosymbiosis; evolution; phytoplank- pan 1980, Lipps 1993, Mendelson 1993). After the ton; plastid ; RUBISCO end-Permian extinction, however, several unicellular algal phyla with secondary red plastids rose to ecologi- cal prominence. Although the possibility that second- ary symbiotic associations occurred in the red lineage 1Received 6 June 2002. Accepted 30 October 2002. before the end-Permian extinction cannot be ex- 2Author for correspondence: email [email protected]. cluded, the fossil evidence suggests that thecate di-

259

260 DANIEL GRZEBYK ET AL. noflagellates appeared in the early to mid-Triassic and core set of plastid genes intensively radiated in the Jurassic period (Bujak and At the time of this writing, we located 24 complete Williams 1979, Fensome et al. 1996), the coccolitho- genome sequences in public databases; 10 phorids (Haptophyta) evolved in the late Triassic and are from algae (Table 1). We do not include in our rose to ecological prominence in the analysis the plastid genome of dinoflagellates, for (Roth 1987, Bown et al. 1992), and the diatoms which the current knowledge remains fragmentary at emerged sometime in the Jurassic and rose to ecologi- best (Boczar et al. 1991, Zhang et al. 1999, Barbrook cal prominence from the late Cretaceous (Lipps 1993, and Howe 2000). Gene distributions in the nine pho- Harwood and Nikolaev 1995). These phyla continue tosynthetic algal plastid genomes (excluding the non- to dominate the eukaryotic phytoplankton assem- photosynthetic Astasia longa) were established with the blages in the contemporary . Although green 190 identified and 66 hypothetical (ycf ) -coding eukaryotic phytoplankton have not been competi- genes from the last update of plastid gene nomencla- tively excluded from the ocean (indeed, quantitative ture and distribution (Stoebe et al. 1998) (Table 2). molecular analyses suggest that prasinophytes remain The distribution reveals a set of approximately 50 relatively abundant in picoplankton; Moon-Van Der core protein-coding genes retained in all taxa (Fig. 1). Staay et al. 2001, Vaulot et al. 2002), green eukaryotic These genes are broadly clustered into three major phytoplankton seldom dominate in contemporary pe- functional domains: 1) genes encoding for the proton lagic marine ecosystems. For example, an analysis of coupled ATP synthase (atp genes); 2) genes encoding HPLC-based composition from the Joint for photosynthetic processes, including reaction cen- Coastal Ocean Flux Study reveals that chl c is 3.4 times ter proteins (psa, psb genes), most of the electron more abundant in the ocean than chl b on a molar ba- transport components that connect the two photosys- sis (Falkowski, unpublished data). Why have red plas- tems (pet genes), and the large subunit of RUBISCO tid-containing phyla become so ecologically successful (rbcL genes); and 3) housekeeping genes that include over the past 250 million years? the plastid ribosomal proteins (rpl, rps genes), RNA Why plastids (or any other ) retain any polymerase (rpo genes), and elongation factor (tufA genetic information at all is unclear (Allen and Raven gene). Only one biosynthetic gene, involved in chl 1996). It is clear, however, that no plastid is com- synthesis (chlI ), is retained in all photosynthetic al- pletely autonomous, that is, capable of growth and gae; this gene was transferred to the nucleus in higher replication outside of the host . The reason is that plants. In the heterotrophic A. longa, the plastid ge- after all symbiotic events, most protoplastid genes nome retained mostly housekeeping genes, losing all were transferred to the host (McFadden but one (rbcLg) gene involved in . 1999). However, all plastids retain a set of genes that Among , the third primary endosymbi- are essential for the physiological function of the or- otic with a few, rare, extant species, Cyano- ganelle in oxygenic photosynthesis. Here we present an phora paradoxa retains about 30 genes in common with analysis of whole plastid genomes from various eukary- all red plastids (rhodophytes, cryptophytes, and dia- otic algal taxa in an attempt to understand what ge- toms) and about 10 additional genes found only in netic features distinguish the two major plastid lineages rhodophytes but only 3 to 4 genes associated with from each other. From our analysis, we suggest that dif- green algal plastids (including the rbcLg gene coding ferences in gene composition between plastid geno- for the type IB large subunit of RUBISCO). types could account, in part, for the ecological success of phytoplankton containing secondary red plastids. In considering the relative ecological and evolu- tionary success of the red plastid phyla, we postulated Table 1. The algal plastid genomes available in databases. that genetic information retained by plastids is critical Number of in defining the potential fitness of the plastid in new protein-coding endosymbiotic associations. This hypothesis suggests Algal division, species, and reference genes that after a major selection event, such as a change in Rhodophyta (primary red plastids) redox chemistry of the ocean, red and green plastids Porphyra purpurea (Reith and Munholland 1995) 203 effectively compete with each other for accommoda- Cyanidium caldarium (Glöckner et al. 2000) 199 tion within new host cells. Assuming that, before the Cryptophyta (secondary red plastids) symbiotic association, the host cell did not contain Guillardia theta (Douglas and Penny 1999) 147 photosynthetic or other genes retained within what Bacillariophyta (secondary red plastids) Odontella sinensis (Kowallik et al. 1995) 139 would become a plastid, the selection and subsequent Glaucophyta (primary plastids) fitness of the plastid within the host must depend on Cyanophora paradoxa (Stirewalt et al. 1995) 136 the extent to which the host cell can support the ge- (primary green plastids) netically depleted organelle. Hence, the potential for Mesotigma viride (Lemieux et al. 2000) 99 Nephroselmis olivacea (Turmel et al. 1999) 91 “portability” of the plastid among host cells should in- vulgaris (Wakasugi et al. 1997) 78 crease with the amount of genetic information re- Euglenophyta (secondary green plastids) tained by the plastid. We call this the “portable plastid Euglena gracilis (Hallick et al. 1993) 52 hypothesis.” Astasia longa (Gockel and Hachtel 2000) 28

THE PORTABLE PLASTID HYPOTHESIS 261

Fig. 1. Distribution of the 190 first identified and 66 hypothetical (ycf ) protein-coding genes in the nine photosynthetic algal plastid genomes, using the last update of nomenclature and distribution of Stoebe et al. (1998). (See Table 2 for the summary of the gene nomenclature.) Primary endosymbiotic genomes (Glaucophyta, Rhodophyta, Chlorophyta, square boxes) are widely overlap- ping, in accordance with the monophyletic origin of all plastids. The secondary endosymbiotic genomes (Euglenophyta, Cryptophyta, Bacillariophyta, round boxes) are included in the primary lineages from which they emerged. In the secondary plastid genomes, minD-minE genes in cryptophytes and ycf66 in diatoms are assumed to originate from their ancestral rhodophyte and euglenophyte ycf13 and ycf67 from their ancestral chlorophyte endosymbiont. On the whole, distribution of ycfs does not differ signif- icantly from those of other protein-coding genes. Parentheses indicate genes not found in all rhodophyte plastid genomes; square brackets indicate genes not found in all chlorophyte plastid genomes. Genes involved in the three main plastid functions represented in the core set are highlighted: ATP synthase genes (), photosynthetic processes (green), and housekeeping genes (). Genes involved in protein regulatory pathways (transcriptional and post-transcriptional regulation) in the complementary gene set of red plastids are shown in pink.

The strong similarity among numerous functional logenetic analyses, using data from single plastid genes in all plastids, including the reaction centers genes or nuclear genes coding for plastid proteins, and ATPase, and the corresponding concordance suggest that the plastids of heterokonts, cryptophytes, with extant cyanobacterial gene sequences strongly haptophytes, and dinoflagellates were derived from support a monophyletic origin of chloroplasts from a rhodophytes (Chesnick et al. 1996, Bhattacharya and cyanobacterium-like ancestor (Douglas and Murphy Medlin 1998, Durnford et al. 1999, Takishita et al. 1994, Bhattacharya and Medlin 1995, 1998, Hess et al. 1999, 2000, Takishita and Uchida 1999, Zhang et al. 1995, Palenik and Swift 1996, Köhler et al. 1997). Phy- 1999). In the green lineage, euglenophyte and chlo-

262 DANIEL GRZEBYK ET AL.

Table 2. The protein-coding gene nomenclature in algal rarachniophyte plastids appear to have been acquired chloroplast genomes (adapted from Stoebe et al. 1998). from secondary endosymbiosis of chlorophytes (van de Peer et al. 1996, Bhattacharya and Medlin 1998, Gene code Protein function Palmer and Delwiche 1998, Turmel et al. 1999). Ac- accA,B,D Acetyl-CoA carboxylase cordingly, the set of genes retained in secondary en- acpP Acyl carrier protein apcA,B,D,E,F Allophycocyanin dosymbiotic plastids is almost totally included in the argB Acetylglutamate kinase primary lineage from which they descended (Fig. 1). atpA,B,D,E,F,G,H,I ATP synthase The only exceptions are a small number of genes re- bas1 Thiol-specific protein tained in secondary plastids that may not be present bioY Biotin synthase carA Carbamoyl phosphate synthetase in their related extant primary plastids: minD and cbbX Red type Calvin cycle minE in cryptophytes (Fig. 1), ycf66 in diatoms, and ccsA Heme attachment to plastid cytochrome c ycf13 and ycf67 in euglenophytes. These genes are rel- cemA Envelope membrane protein chlB,I,L,N reductase ics from ancestral primary plastids that were later lost. clpC/P Caseinolytic-like protease (Clp) cpcA,B,G phycobilisome gene retention and losses in chloroplast cpeA,B Phycoerythrin genomes reflect evolutionary patterns crtE Geranylgeranyl pyrophosphate synthetase cysA,T Probable transport proteins Whereas the presence of specific genes provides clues dfr Drug sensory protein about the origin of plastids, plastid gene losses, which dnaB DNA-replication helicase are effectively irreversible, also provide evolutionary dnaK Hsp 70-type chaperone dsbD Thiol:disulfide interchange protein information. There are 200 or fewer protein-coding fabH -Ketoacyl-acyl carrier protein synthase III genes in primary plastids. Assuming that the number fdx 2[4Fe-4S] ferredoxin of genes in the original primary ur-plastid was similar ftrB Ferredoxin-thioredoxin reductase ftsH,W Division proteins to that found in the extant cyanobacterium Synechocys- glnB Nitrogen regulatory protein tis PCC6803 (3168 protein-coding genes; Kaneko et gltB Glutamate synthase (GOGAT) al. 1996), more than 93% of the ancestral endosym- groEL,ES Chaperonins 60 and 10 kDa biont genome was lost or transferred to the host nu- hemA 5-Aminolevulinic acid synthase hisH Histidinol-phosphate aminotransferase cleus in the primary plastid lineages (Fig. 2). Green I-CvuI DNA endonuclease plastids exhibit the most numerous gene losses, either ilvB,H Acetohydroxyacid synthase specific losses or those in common with glaucophytes. infA,B,C Translational initiation factors minD,E Homologues of bacterial Subsequently, patterns of gene losses occurred differ- regulators ently at phylum level radiations within primary lin- mntA,B Manganese transport system proteins eages and at radiations accompanying secondary sym- moeB Molybdopterin protein biotic events. Between 1% and 10% of the remaining nadA Quinolinate synthase nblA Phycobilisome degradation protein plastid genes were lost at phylum level radiations ndhA-J NADH- oxidoreductase within rhodophytes and chlorophytes. A larger num- ndhK NADH-ubiquinone oxidoreductase ber of the remaining primary plastid genes were lost ntcA Global nitrogen transcriptional regulator odpA,B Pyruvate dehydrogenase E1 component at radiations linked to secondary symbiotic events. For pbsA Heme oxygenase example, with the divergence of cryptophyte and ba- petA,B,D,F,G,J,L,M Photosystem electron transport proteins cillariophyte plastids from the rhodoplasts, between pgmA Phosphoglycerate mutase preA Prenyl transferase 15% and 20% of the plastid genes were lost. Approxi- psaA-M PSI proteins mately 30% of the genes were lost at the divergence of psbA-X PSII proteins euglenoid plastids from chloroplasts. This analysis sug- rbcLg/r RUBISCO large subunit, green and red forms gests that endosymbiotic events resulted in relatively rbcR RUBISCO operon transcriptional regulator rapid and massive gene losses in plastids, whereas the rbcSg/r RUBISCO small subunit, green and red forms radiations within phyla were accompanied by slower rdpO Probable reverse transcriptase and more gradual genomic erosion. rne RNAseE rpl1-36 Large subunit ribosomal proteins We used the patterns of gene loss, inferred from a rpoA,B,C1,C2 RNA polymerase simple presence/absence analysis, to construct an rps1-20 Small subunit ribosomal proteins evolutionary (Fig. 3). In this analysis, we implic- secA,Y Preprotein-translocase itly assume that each gene retained in a plastid re- syfB Phenylalanine tRNA synthetase syh Histidine tRNA synthetase flects an ancestral status, and hence, at least within a thiG biosynthesis lineage, the more evolved a plastid, the more genes trpA Tryptophane synthase were lost. The patterns of gene loss clearly separate trpG Anthranilate synthase, glutamine amidotransferase the red plastid lineage, with rhodophytes at the base trxA Thioredoxin of the cluster, from the green lineage. Further out of tsf Translational elongation factor Ts the red lineage are the secondary endosymbiotic al- tufA Translational elongation factor Tu gae: the cryptophytes and bacillariophytes. In the upp Uracil phosphoribosyltransferase ycf Hypothetical proteins green lineage, the plastid appears ancestral within the chlorophytes, and the pattern of gene loss suggests that land plants diverged early from chloro- THE PORTABLE PLASTID HYPOTHESIS 263

Fig. 2. Hypothetical patterns of gene losses at radiations of phytoplankton and plastids inferred from 190 identified pro- tein-coding genes (Stoebe et al. 1998) in extant species plastid genomes. For clarity, the ycf losses, whose pattern is similar to identified protein-coding genes (from the gene distribution in Fig. 1), are not shown. From the ancestral photosynthetic genome, ra- diations leading to primary phytoplankton lineages after the primary endosymbiosis are indicated by a single circle. Radiations due to secondary endosymbiotic events are indicated by a double circle. After primary endosymbiosis, gene losses are shown from a hypo- thetical ancestral plastid genome containing the whole set of genes found in extant plastids: genes specifically retained in the red pri- mary plastids (i.e. lost in both green and plastids) are in red, and genes specifically retained by green and glaucophyte plastids are in green and gray, respectively. Other genes losses that occurred at a single radiation are in black, and those that occurred at several independent radiations are in blue.

phytes (Figs. 2 and 3). Our phylogenetic analysis (Fig. as intermediate between and the red 3) points out the high degree of evolution in eugleno- plastid lineage (van den Hoek et al. 1995, Ohta et al. phyte plastids, reflected by their remarkably reduced 1997, Martin et al. 1998, Stoebe and Kowallik 1999). genomes (Table 1). The retention of rpl22 in the cen- On the whole, the evolutionary relationships in- ter of the ribosomal protein cluster L2 in Euglena and ferred from gene composition of plastids, at least Astasia suggests, however, that euglenophyte plastids within the red and green plastid lineages, are surpris- radiated relatively early from chlorophytes (Fig. 2). ingly similar to those derived from concatenated plas- Our analysis places the glaucophyte Cyanophora at the tid protein-coding gene sequences and corroborates base of the green plastid cluster (Fig. 3). However, the that view of plastid phylogeny (Martin et al. 1998, 2002, plastid in this retains more genes (up to 53, Turmel et al. 1999, Lemieux et al. 2000). that is, approximately 39% of its genome) in common with rhodophyte plastids (Fig. 1). Glaucophytes are number of secondary and tertiary generally considered a primitive primary endosymbi- endosymbiotic events otic lineage; their plastids (cyanelles) have retained Before we elaborate further on our portable plastid some of the structural, biochemical, and genetic fea- hypothesis, we need to summarize current knowledge tures of the cyanobacterium-like ancestor and appear about the number of secondary and tertiary endosym- 264 DANIEL GRZEBYK ET AL.

Fig. 3. Phylogenetic analysis of plastids inferred from gene presence and gene loss in plastid ge- nomes, from the 190 identified and 66 hypothetical protein-coding genes of the list of Stoebe et al. (1998). The cyanobacterium Synechocystis sp. was used as an outgroup. Bootstrap values over 50% are shown at nodes. The consensus tree was inferred using the Camin-Sokal parsimony method in the PHYLIP 3.57 package (Felsenstein 1993) (program MIX, using the Jumble option to run the analysis 10 times for each of the 1000 bootstrapped data sets). Gene presence (re- flecting an ancestral status) was scored 0, and gene absence (i.e. loss, reflecting an evolved status) was scored 1.

biotic events assumed to have occurred in the green at least four independent secondary endosymbiotic and red lineages. Over the last decade, molecular phy- events involving red algae (as illustrated in Delwiche logenetic analyses have been used in an attempt to re- 1999). An alternative hypothesis suggests that all solve the evolutionary relationships between the main phyla containing secondary red plastids arose from a algal phyla. These analyses have been based largely on single secondary endosymbiotic event that led to the phylogenetic relationships between specific nuclear chromalveolate lineage (Cavalier-Smith 1999, 2000, and plastid-encoded genes. A major concern is whether, 2002a,b, McFadden 2001). This hypothesis is sup- within a lineage, a single host cell acquired its plastids ported by phylogenetic analyses of nuclear genes, before the divergence of the endosymbiotic associations which distinguish several , including into distinct clades or if various apoplastidic hetero- ( and , which include dinoflagel- trophic host cells, which were phylogenetically distinct, lates) and Chromista (heterokonts and cryptophytes). acquired plastids independently. Neither analysis included haptophytes (Baldauf et al. In the green plastid lineage, analyses of 18S rRNA 2000, Fast et al. 2001). genes (rDNA) clearly distinguished euglenophytes Altogether, there appears to be twice as many sec- from chlorarachniophytes, and according to specific ondary endosymbiotic events in the red line compared features of their plastids, the hypothesis of their emer- with the green line. Curiously, one phylum appears to gence from two independent secondary events is well be promiscuous in appropriating plastids. Although it accepted (Delwiche 1999, McFadden 2001, Cavalier- is questioned whether the ecologically predominant Smith 2002b). In the red plastid lineage, analyses of peridinin-containing plastid dinoflagellates arose from 18S rDNA and GAPDH genes suggest no close phylo- a secondary or tertiary endosymbiotic event (Morden genetic relationship between the various secondary and Sherwood 2002, Yoon et al. 2002), the great variety host cells that comprise the red line (Bhattacharya of chloroplasts found within dinoflagellates clearly re- and Medlin 1998, Fast et al. 2001). Moreover, analyses veals that several independent endosymbiotic events of plastid-encoded genes (mostly 16S rRNA and rbcL occurred in this phylum. These events involved at least genes) suggest an independent origin of several sec- three types of secondary red plastid algal phyla (dia- ondary red plastids within the ancestral rhodophytes toms, haptophytes, and cryptophytes) and one from (Bhattacharya and Medlin 1995, Takishita et al. 1999, primary green plastid algae (prasinophytes) (Schnepf 2000, Oliveira and Bhattacharya 2000, Müller et al. and Elbrächter 1999). It is unclear whether the green 2001, Martin et al. 2002). Consequently, one hypothe- plastids of dinoflagellates are monophyletic or if they sis suggests that in the red line arose from are secondary or tertiary endosymbionts. THE PORTABLE PLASTID HYPOTHESIS 265

”portable plastid” hypothesis erotrophic host cells that were under strong selection Why then have the disparate taxa within the red to become autotrophic. A corollary of this hypothesis line become ecologically successful in the contempo- is that, in contrast, the transfer of some of key core rary ocean, and what processes might have contrib- plastid genes to the host nuclear genome decreased uted to their increased competitive advantage in the the portability of primary green plastids. Once these Mesozoic ocean? It is clear from our whole plastid ge- genes were lost from the plastid, a potential secondary nome analysis that whatever the host cell, all red plas- host would have needed to retrieve them from the pri- tids retain a common set of genes that distinguishes this mary host nucleus. The probability that such an event plastid type from all other plastids in eukaryotic algae occurred is difficult to quantify, but the fact that most (Fig. 1). Specifically, all algae with red plastid genomes extant species within the green algal lineages are pri- exclusively retained a set of 14 genes. In addition, they mary endosymbionts suggests that it was rare; to wit, have about 30 genes retained in glaucophytes (Fig. 1). there are no known heterokonts, haptophytes, and These two groups of genes constitute a complementary cryptophytes with green plastids. The enigma, how- core set in red plastids, 75% of which are related to the ever, is how retrieving these few genes from the nu- core set described earlier: 20 ribosomal proteins (rpl cleus of potential green primary endosymbionts might and rps genes), 2 ATPase genes, 9 genes related to have presented a problem to the secondary hosts that photosynthesis (2 pet, 4 psa including psaD encoding had successfully appropriated or discarded the other for the ferredoxin docking protein in PSI, 3 psb), and approximately 2000 nuclear-encoded plastid protein the small subunit of RUBISCO (rbcS). Among the other genes from red primary endosymbionts. Alternatively, genes, several encode for proteins involved in various to replace the gene lost in the plastid, the secondary processes of post-transcriptional regulation of plastid host would have to obtain analogous nonplastid genes proteins, especially those encoded in the nucleus. from elsewhere by lateral gene transfer: a documented For example, the translocase complex secA/Y, located example of lateral gene transfer is the -proteobacte- in the membrane, imports such proteins as rial origin of the form L2 RUBISCO found in the peri- or PSI protein Psa-N from the stroma dinin-containing dinoflagellates (Morse et al. 1995, into the thylakoid lumen (Kloesgen 1997, Dalbey and Whitney et al. 1995). Robinson 1999). Several genes encode for stromal Why then are there green secondary endosymbi- chaperones (dnaK and groEL) and chaperone-pro- otic plastids in other algal taxa, the euglenophytes, teases (clpC) involved in shaping/folding, assembly, chlorarachniophytes, and the green dinoflagellates? and activation of oligomeric protein complexes (Chen For the euglenophytes and chlorarachniophytes, the and Schnell 1999, Agarraberes and Dice 2001). Inter- plastid phylogeny inferred from 16S rDNA suggests estingly, the red plastids retained the Clp regulatory their plastids are derived early within the green plas- ATP-binding subunit gene, clpC, whereas green plastids tid lineage (Bhattacharya and Medlin 1998). Indeed, retained the proteolytic subunit gene, clpP. Other genes to reconcile the plastid portability hypothesis with the encode for a DNA helicase (dnaB), a protein involved occurrence of secondary symbionts in the green line in organelle division (fstH), and an acyl carrier pro- requires that the secondary endosymbiotic events in tein (acpP gene) involved in an early step of syn- the green line occurred before the transfer of some thesis. The rbcR and cbbX genes are transcriptional core and complementary genes to the primary host regulators. The latter is specifically involved in tran- nucleus. Subsequent gene transfers from secondary scriptional regulation of the red-type RUBISCO gene green plastids paralleled those from primary green operon, rbcL-rbcS, retained in all plastids in the red line. plastids (i.e. the transfer of the rbcS gene to the nu- In contrast, in the green line, the transfer to the nucleus cleus of euglenophytes). of the gene encoding the small subunit of RUBISCO (which is essential for the catalytic activity) ensures potential role of host cells in the ecological nuclear (i.e. host cells) control of RUBISCO expression success of the endosymbiotic assemblages (Rodermel 1999). The retention of both RUBISCO Although the modern dominant eukaryotic marine genes in the red plastid genomes ensured much greater phytoplankton (diatoms, peridinin-containing dinoflagel- autonomy in RUBISCO expression. Altogether, the com- lates, haptophytes, cryptophytes) have rhodophyte- plementary core set of genes in the red lineage sug- derived plastids, the rhodophytes themselves are rare in gests that compared with green plastids, red plastids contemporary oceanic phytoplankton assemblages. This have retained significantly more capacity to autono- apparent paradox suggests that the ecological success of mously express proteins critical to maintaining and the secondary red symbiotic taxa is not solely a conse- regulating oxygenic photosynthetic and energy trans- quence of the composition of their plastids’ genetic duction pathways. and functional potential but also involves the eukary- We hypothesize that after the long and dramatic otic host cells, which may have provided new geneti- changes in ocean chemistry and circulation associated cally determined fitness. with the end-Permian extinction, gene retention in What genetic potential of the various eukaryotic the red line increased their potential to be appropri- hosts determined the ecological success of the endo- ated by a variety of new, phylogenetically diverse, het- symbiotic assemblages? One major attribute that sec- 266 DANIEL GRZEBYK ET AL. ondary hosts have brought in to new phytoplankton Agarraberes, F. A. & Dice, J. F. 2001. Protein translocation across phyla of the red line is armor, in the form of siliceous, membranes. Biochim. Biophys. Acta 1513:1–24. Allen, J. F. & Raven, J. A. 1996. Free radical-induced versus calcareous, or based cell walls, or formation redox regulation: costs and benefits of genes in . J. of exopolymers that permit aggrega- Mol. Evol. 42:482–92. tion of cells to form massive colonies or dramatically Baldauf, S. L., Roger, A. J., Wenk-Siefert, I. & Doolittle, W. F. 2000. increase the seawater viscosity. These host-dependent A kingdom-level phylogeny of eukaryotes based on combined properties afford protection against pressure protein data. Science 290:972–7. Barbrook, A. C. & Howe, C. J. 2000. Minicircular plastid DNA in the and may have conferred a critical ecological advan- Amphidinium operculatum. Mol. Gen. Genet. 263: tage to the secondary red plastid phytoplankters (Ver- 152–8. ity and Smetacek 1996). In contrast, the marine green Bhattacharya, D. & Medlin, L. 1995. The phylogeny of plastids: a re- plastid host cells are subjected to relatively heavy graz- view based on comparisons of small-subunit ribosomal RNA coding regions. J. Phycol. 31:489–98. ing pressure. Indeed, environments where green uni- Bhattacharya, D. & Medlin, L. 1998. Algal phylogeny and the origin cellular algae form blooms are lacustrine, including of land plants. Plant Physiol. 116:9–15. the Great Lake. In such environments the diver- Boczar, B. A., Liston, J. & Cattolico, R. A. 1991. Characterization of sity of grazers is low compared with the contemporary DNA from three marine dinoflagellates (Dinophyceae): ocean. Most extant red and green primary endosymbi- Glenodinium sp. and two members of the toxic , Protogon- yaulax. Plant Physiol. 97:613–8. otic taxa (green and red macroalgae and land plants) Bown, P. R., Burnett, J. A. & Gallagher, L. T. 1992. Calcareous nano- greatly diverged with the emergence of multicellular- evolution. Mem. Sci. Geol. 43:1–17. ity with parenchymatous development. A simi- Bujak, J. P. & Williams, G. L. 1979. Dinoflagellate diversity through lar evolutionary pathway only occurred one time in time. Mar. Micropaleontol. 4:1–12. Cavalier-Smith, T. 1999. Principles of protein and lipid targeting in the various secondary endosymbiotic phyla to give rise secondary endosymbiosis: euglenoid, dinoflagellate, and sporo- to the brown macroalgae (Phaeophyceae). Curiously, zoan plastid origins and the eukaryote family tree. J. Eukaryot. the red plastid line appears never to have colonized Microbiol. 46:347–66. terrestrial ecosystems. Cavalier-Smith, T. 2000. Membrane heredity and early chloroplast evolution. Trends Plant Sci. 15:174–82. Cavalier-Smith, T. 2002a. Chloroplast evolution: secondary symbio- genesis and multiple losses. Curr. Biol. 12:R62–4. conclusions Cavalier-Smith, T. 2002b. The phagotrophic origin of eukaryotes Although all major classes of plastids are found in and phylogenetic classification of . Int. J. Syst. Evol. Mi- the modern ocean, all the dominant phytoplankton crobiol. 52:297–354. Chen, X. & Schnell, D. J. 1999. Protein import into chloroplasts. taxa are secondary red endosymbionts. Our whole plas- Trends Cell Biol. 9:222–7. tid genome analysis suggests that plastid portability Chesnick, J. M., Morden, C. W. & Schmieg, A. M. 1996. Identity of was a major factor that contributed to the ecological the endosymbiont of Peridinium foliaceum (Pyrrophyta): analysis success of the red line. Plastid portability is based on of the rbcLS operon. J. Phycol. 32:850–7. the retention of functional photosynthetic and energy Dalbey, R. E. & Robinson, C. 1999. Protein translocation into and across the bacterial plasma membrane and the plant thylakoid transducing genes in the organelle. One key gene ap- membrane. Trends Biochem. Sci. 24:17–22. pears to be the small subunit of RUBISCO. Although Delwiche, C. F. 1999. Tracing the thread of plastid diversity the transfer of these plastid genes to the host nucleus through the tapestry of . Am. Nat. 154:S164–77. did not predestine the green lineage to extinction in Douglas, S. E. & Murphy, C. A. 1994. Structural, transcriptional, and phylogenetic analyses of the atpB gene cluster from the the marine phytoplankton, we propose that it re- plastid of (). J. Phycol. 30:329–40. duced the evolutionary plasticity of this lineage to Douglas, S. E. & Penny, S. L. 1999. The plastid genome of the cryp- evolve to rapidly changing ocean chemistry and redox tophyte alga, Guillardia theta: complete sequence and con- conditions in the Mesozoic period. Our plastid porta- served synteny groups confirm its common ancestry with red bility hypothesis suggests two major factors were criti- algae. J. Mol. Evol. 48:236–44. Durnford, D. G., Deane, J. A., Tan, S., McFadden, G. I., Gantt, E. & cal in determining the evolutionary trajectory of eu- Green, B. R. 1999. A phylogenetic assessment of the eukaryotic karyotic phytoplankton after the end-Permian extinction: light-harvesting antenn proteins, with implications for plastid 1) the loss of critical genes from plastid genomes during evolution. J. Mol. Evol. 48:59–68. endosymbiotic events, which differentially affected the Fast, N. M., Kissinger, J. C., Roos, D. S. & Keeling, P. J. 2001. Nu- clear-encoded, plastid-targeted genes suggest a single common genetic and functional potential of plastids, and 2) the origin for apicomplexan and dinoflagellates plastids. Mol. Biol. concurrent ecological fitness of the specific host cells, Evol. 18:418–26. which afforded selective advantages under changing en- Felsenstein, J. 1993. PHYLIP (Phylogeny Inference Package). Version vironmental conditions. The selection pressures brought 3.57c. Distributed by the author, Department of Genetics, Uni- about by the end-Permian extinction that led to the sub- versity of , Seattle. Fensome, R. A., MacRae, R. A., Moldowan, J. M., Taylor, F. J. R. & sequent ecological success of the red lineage remains Williams, G. L. 1996. The early Mesozoic radiation of di- to be elucidated. noflagellates. 22:329–38. Glöckner, G., Rosenthal, A. & Valentin, K. 2000. The structure and gene repertoire of an ancient red algal plastid genome. J. Mol. This research was supported by the National Science Founda- Evol. 51:382–90. tion (NSF) through the Biocomplexity program (NSF Grant Gockel, G. & Hachtel, W. 2000. Complete gene map of the plastid OCE-0084032). We thank Andrew Knoll, Andrew Irwin, Lee genome of the nonphotosynthetic euglenoid Astasia Kerkhof, Dion Durnford, Linda Medlin, John Raven, and two longa. 151:347–51. anonymous reviewers for constructive comments. Hallick, R. B., Hong, L., Drager, R. G., Favreau, M., Monfort, A., Or- THE PORTABLE PLASTID HYPOTHESIS 267

sat, B., Spielmann, A. & Stutz, E. 1993. Complete sequence of Doyle, J. J. [Eds.] Molecular of Plants. II. DNA Sequenc- Euglena gracilis chloroplast DNA. Nucleic Acids Res. 21:3537–44. ing. Kluwer Academic, Boston, pp. 375–409. Harwood, D. M. & Nikolaev, V. A. 1995. Cretaceous diatoms: mor- Reith, M. & Munholland, J. 1995. Complete nucleotide sequence of phology, , biostratigraphy. In Blome, C. D., Whalen, the Porphyra purpurea chloroplast genome. Plant Mol. Biol. P. M. & Reed, K. [Eds.] Siliceous . Paleontological So- Reptr. 13:333–5. ciety, Pittsburgh, PA, pp. 81–106. Rodermel, S. 1999. Subunit control of Rubisco biosynthesis—a relic Hess, W. R., Weihne, A., Loiseaux-de Goer, S., Partensky, F. & Vau- of an endosymbiotic past? Photosynth. Res. 59:105–23. lot, D. 1995. Characterization of the single psbA gene of Prochlo- Roth, P. H. 1987. Mesozoic calcareous nannofossil evolution: rela- rococcus marinus CCMP 1375 (Prochlorophyta). Plant Mol. Biol. tion to paleoceanographic events. Paleoceanography 2:601–11. 27:1189–96. Schnepf, E. & Elbrächter, M. 1999. Dinophyte chloroplasts and Kaneko, T., Sato, S., Kotani, H., Tanaka, A., Asamizu, E., Naka- phylogeny—a review. Grana 38:81–97. mura, Y., Miyajima, N., Hirosawa, M., Sugiura, M., Sasamoto, Stirewalt, V. L., Michalowski, C. B., Löffelhardt, W., Bohnert, H. J. & S., Kimura, T., Hosouchi, T., Matsuno, A., Muraki, A., Na- Bryant, D. B. 1995. Nucleotide sequence of the cyanelle genome kasaki, N., Naruo, K., Okumura, S., Shimpo, S., Takeuchi, C., from Cyanophora paradoxa. Plant Mol. Biol. Rep. 13:327–32. Wada, T., Watanabe, A., Yamada, M., Yasuda, M. & Tabata, S. Stoebe, B., Martin, W. & Kowallik, K. V. 1998. Distribution and no- 1996. Sequence analysis of the genome of the unicellular cy- menclature of protein-coding genes in 12 sequenced chloro- anobacterium Synechocystis sp. strain PCC6803. II. Sequence de- plast genomes. Plant Mol. Biol. Rep. 16:243–55. termination of the entire genome and assignment of potential Stoebe, B. & Kowallik, K. V. 1999. Gene-cluster analysis in chloro- protein-coding regions. DNA Res. 3:109–36. plast genomics. Trends Genet. 15:344–7. Kloesgen, R. B. 1997. Protein transport into and across the thyla- Takishita, K., Nakano, K. & Uchida, A. 1999. Preliminary phyloge- koid membrane. J. Photochem. Photobiol. B Biol. 38:1–9. netic analysis of plastid-encoded genes from an anomalously Köhler, S., Delwiche, C. F., Denny, P. W., Tilney, L. G., Webster, P., pigmented dinoflagellate Gymnodinium mikimotoi (Gymnodini- Wilson, R. J. M., Palmer, J. D. & Roos, D. S. 1997. A plastid of ales, Dinophyta). Phycol. Res. 47:257–62. probable green algal origin in apicomplexan parasites. Science Takishita, K. & Uchida, A. 1999. Molecular cloning and nucleotide 275:1485–9. sequence analysis of psbA from the dinoflagellates: origin of Kowallik, K. V., Stoebe, B., Schaffran, I., Kroth-Pancic, P. & Freier, the dinoflagellate plastid. Phycol. Res. 47:207–16. U. 1995. The chloroplast genome of the chlorophylle ac-con- Takishita, K., Nakano, K. & Uchida, A. 2000. Origin of the plastid in taining alga, Odontella sinensis. Plant Mol. Biol. Reptr. 13:336–42. the anomalously pigmented dinoflagellate Gymnodinium miki- Lemieux, C., Otis, C. & Turmel, M. 2000. Ancestral chloroplast ge- motoi (, Dinophyta) as inferred from phyloge- nome in Mesostigma viride reveals an early branch of green netic analysis based on the gene encoding the large subunit of . 403:649–52. form I-type RuBisCo. Phycol. Res. 48:85–9. Lipps, J. H. 1993. Fossil and . Blackwell, Oxford, Tappan, H. N. 1980. The Paleobiology of Plant Protists. Freeman W.H., 342 pp. San Francisco, 1028 pp. Martin, W., Stoebe, B., Goremykin, V., Hansmann, S., Hasegawa, M. Tomitani, A., Okada, K., Miyashita, H., Matthijs, H. C. P., Ohno, T. & & Kowallik, K. V. 1998. Gene transfer to the nucleus and the Tanaka, A. 1999. b and in the common evolution of chloroplasts. Nature 393:162–5. ancestor of cyanobacteria and chloroplasts. Nature 400:159–62. Martin, W., Rujan, T., Richly, E., Hansen, A., Cornelsen, S., Lins, T., Turmel, M., Otis, C. & Lemieux, C. 1999. The complete chloroplast Leister, D., Stoebe, B., Hasegawa, M. & Penny, D. 2002. Evolu- DNA sequence of the green alga Nephroselmis olivacea: insights tionary analysis of Arabidopsis, cyanobacterial, and chloroplast ge- into the architecture of ancestral chloroplast genomes. Proc. nomes reveals plastid phylogeny and thousands of cyanobacterial Natl. Acad. Sci. USA 96:10248–53. genes in the nucleus. Proc. Natl. Acad. Sci. USA 99:12246–51. van de Peer, Y., Rensing, S. A., Maier, U.-G. & De Wachter, R. 1996. McFadden, G. I. 1999. Plastids and protein targeting. J. Eukaryot. Substitution rate calibration of small subunit ribosomal RNA Microbiol. 46:339–46. identifies chlorarachniophytes endosymbionts as remnants of McFadden, G. I. 2001. Primary and secondary endosymbiosis and green algae. Proc. Natl. Acad. Sci. USA 93:7732–6. the origin of plastids. J. Phycol. 37:951–9. van den Hoek, C., Mann, D. G. & Jahns, H. M. 1995. Algae: An Intro- Mendelson, C. V. 1993. Acritarchs and prasinophytes. In Lipps, duction to . Cambridge University Press, Cambridge, J. H. [Ed.] Fossil Prokaryotes and Protists. Blackwell, Oxford, pp. UK, 627 pp. 77–104. Vaulot, D., Romari, K. & Not, F. 2002. Are less diverse than Moon-Van Der Staay, S. Y., De Wachter, R. & Vaulot, D. 2001. Oce- in marine picoplankton? Trends Microbiol. 10:266–7. anic 18S rDNA sequences from picoplankton reveal unsus- Verity, P. G. & Smetacek, V. 1996. Organism life cycles, predation, pected eukaryotic diversity. Nature 409:607–10. and the structure of marine pelagic ecosystems. Mar. Ecol. Prog. Morden, C. W. & Sherwood, A. R. 2002. Continued evolutionary Ser. 130:277–93. surprises among dinoflagellates. Proc. Natl. Acad. Sci. USA 99: Wakasugi, T., Nagai, T., Kapoor, M., Sugita, M., Ito, M., Ito, S., 11558–60. Tsudzuki, J., Nakashima, K., Tsudzuki, T., Suzuki, Y., Hamada, Morse, D., Salois, P., Markowic, P. & Hastings, J. W. 1995. A nuclear- A., Ohta, T., Inamura, A., Yoshinaga, K. & Sugiura, M. 1997. encoded form II RuBisCO in dinoflagellates. Science 268:1622–4. Complete nucleotide sequence of the chloroplast genome Müller, K. M., Oliveira, M. C., Sheath, R. G. & Bhattacharya, D. from the green alga Chlorella vulgaris: the existence of genes 2001. Ribosomal DNA phylogeny of the Bangiophycidae possibly involved in chloroplast division. Proc. Natl. Acad. Sci. (Rhodophyta) and the origin of secondary plastids. Am. J. Bot. USA 94:5967–72. 88:1390–400. Whitney, S. M., Shaw, D. C. & Yellowlees, D. 1995. Evidence that Ohta, N., Sato, N., Nozaki, H. & Kuroiwa, T. 1997. Analysis of the some dinoflagellates contain a ribulose-1,5-bisphosphate car- cluster of ribosomal protein genes in the plastid genome of a boxylase/oxygenase related to that of the -proteobacteria. unicellular red alga Cyanidioschyzon merolae: translocation of Proc. R. Soc. Lond. B. Biol. Sci 259:271–5. the str cluster as an early event in the rhodophyte-chro- Yoon, H. S., Hackett, J. D. & Bhattacharya, D. 2002. A single origin mophyte lineage of plastid evolution. J. Mol. Evol. 45:688–95. of the peridinin- and fucoxanthin-containing plastids in di- Oliveira, M. C. & Bhattacharya, D. 2000. Phylogeny of the Bangio- noflagellates through tertiary endosymbiosis. Proc. Natl. Acad. phycidae (Rhodophyta) and the secondary endosymbiotic ori- Sci. USA 99:11724–9. gin of algal plastids. Am. J. Bot. 87:482–92. Wolfe, G. R., Cunningham, F. X., Durnford, D., Green, B. R. & Palenik, B. & Swift, H. 1996. Cyanobacterial evolution and prochlo- Gantt, E. 1994. Evidence for a common origin of chloroplasts rophyte diversity as seen in DNA-dependent RNA polymerase with light-harvesting complexes of different pigmentation. Na- gene sequences. J. Phycol. 32:638–46. ture 367:566–8. Palmer, J. D. & Delwiche, C. F. 1998. The origin and evolution of Zhang, Z., Green, B. R. & Cavalier-Smith, T. 1999. Single gene cir- plastids and their genomes. In Soltis, P. S., Soltis, D. E. & cles in dinoflagellate chloroplast genomes. Nature 400:155–9.