MBE Advance Access originally published online on April 18, 2007
Molecular Biology and Evolution 2007 24(8):1611-1621; doi:10.1093/molbev/msm075
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Articles |
Evolution of the Glucose-6-Phosphate Isomerase: The Plasticity of Primary Metabolism in Photosynthetic Eukaryotes
* Institut für Genetik, Technische Universität Braunschweig, Braunschweig, Germany
Georg-Speyer-Haus, Johann Wolfgang Goethe Universität Frankfurt, Frankfurt am Main, Germany
Département de Biochimie, Université de Montréal, Montréal, Quebec, Canada
E-mail: j.petersen{at}tu-bs.de
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Glucose-6-phosphate isomerase (GPI) has an essential function in both catabolic glycolysis and anabolic gluconeogenesis and is universally distributed among Eukaryotes, Bacteria, and some Archaea. In addition to the cytosolic GPI, land plant chloroplasts harbor a nuclear encoded isoenzyme of cyanobacterial origin that is indispensable for the oxidative pentose phosphate pathway (OPPP) and plastid starch accumulation. We established 12 new GPI sequences from rhodophytes, the glaucophyte Cyanophora paradoxa, a ciliate, and all orders of complex algae with red plastids (haptophytes, diatoms, cryptophytes, and dinoflagellates). Our comprehensive phylogenies do not support previous GPI-based speculations about a eukaryote-to-prokaryote horizontal gene transfer from metazoa to
-proteobacteria. The evolution of cytosolic GPI is largely in agreement with small subunit analyses, which indicates that it is a specific marker of the host cell. A distinct subtree comprising alveolates (ciliates, apicomplexa, Perkinsus, and dinoflagellates), stramenopiles (diatoms and Phytophthora [oomycete]), and Plantae (green plants, rhodophytes, and Cyanophora) might suggest a common origin of these superensembles. Finally, in contrast to land plants where the plastid GPI is of cyanobacterial origin, chlorophytes and rhodophytes independently recruited a duplicate of the cytosolic GPI that subsequently acquired a transit peptide for plastid import. A secondary loss of the cytosolic isoenzyme and the plastid localization of the single GPI in chlorophycean green algae is compatible with physiological studies. Our findings reveal the fundamental importance of the plastid OPPP for Plantae and document the plasticity of primary metabolism.
Key Words: algal evolution • endosymbioses • gene transfer • glycolysis • oxidative pentose phosphate pathway • plastid metabolism
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
With the discovery of the 3 domains of life by Carl Woese et al. (1990
), small-subunit (SSU) rDNA became the golden standard for the determination of evolutionary relationships. Phylogenetic reconstruction allowed to test the significance of morphological, biochemical, and ultrastructural traits and gave new insights into eukaryotic evolution. Together with other molecular markers, a more comprehensive systematic framework toward a natural classification was created by combining distinct lineages of protists into superensembles, such as stramenopiles, Plantae (also designated as Archaeplastida), or alveolates (Adl et al. 2005). The superensemble stramenopiles (Heterokontophyta) is characterized by 2 flagella of different length (van den Hoek et al. 1996) and comprises diverse algae including unicellular diatoms or huge brown algae (kelps), as well as nonphotosynthetic parasitic oomycetes, for example, the plant pathogen Phytophthora infestans causing the potato blight (Van de Peer and De Wachter 1997). It is widely accepted that the second superensemble Plantae (Cavalier-Smith 1981), including glauco-, rhodo-, chlorophytes, and land plants, obtained their plastid with 2 surrounding membranes through a single primary endosymbiosis with a cyanobacterium (Delwiche and Palmer 1997; Douglas 1998; Stoebe and Kowallik 1999; Rodriguez-Ezpeleta et al. 2005). All other photosynthetic eukaryotes contain complex plastids with 3 or 4 membranes, which originated from the engulfment of a photosynthetic eukaryote by a phagotrophic host cell. Two independent secondary endosymbioses with green algae are documented, which gave rise to eugleno- and chlorarachniophytes (Delwiche 1999; Rogers et al. 2007). Haptophytes, cryptophytes, diatoms, and dinoflagellates contain complex plastids of red algal origin. With respect to the host cell, the affiliation of hapto- and cryptophytes is not yet resolved, but together with photosynthetic stramenopiles (see above), they are designated by the operational term "chromists" due to the presence of chlorophyll c that causes a red pigmentation (Cavalier-Smith 1981). Dinoflagellates are the only photosynthetic lineage of the distinct superensemble alveolata, which moreover comprises heterotrophic ciliates, Perkinseae, and apicomplexa. The latter lineage, which includes the malaria parasite Plasmodium falciparum, harbors a reduced plastid (apicoplast) that also originated via secondary endosymbiosis (McFadden et al. 1996; Wilson et al. 1996), though the putative red algal ancestry (Foth and McFadden 2003) is not clearly established (Funes et al. 2002, 2003; Waller et al. 2003). Although host cell–related markers unequivocally support the monophyly of alveolates (Rodriguez-Ezpeleta et al. 2005), recent hints of a specific relationship between dinoflagellates and chromists (Harper and Keeling 2003; Petersen, Teich, Brinkmann, Cerff 2006) are restricted to plastid related genes of the Calvin cycle and may be related to tertiary endosymbioses (Teich et al. 2007).
In this study, we present comprehensive analyses of eukaryotic glucose-6-phosphate isomerases (GPI) and their prokaryotic homologs. GPI is an essential enzyme of catabolic glycolysis and anabolic gluconeogenesis that catalyzes the reversible isomerization of glucose-6-phosphate and fructose-6-phosphate. Although this cytosolic enzyme is universally present in eukaryotes (Henze et al. 2001; this study), land plants harbor an additional nuclear-encoded plastid-targeted isoenzyme, which was recruited from the cyanobacterial endosymbiont (Nowitzki et al. 1998). The plastid GPI is essential for both the oxidative pentose phosphate pathway (OPPP; Martin and Herrmann 1998) and starch synthesis (Yu et al. 2000). Here, we analyze the general distribution and evolution of GPI genes in photosynthetic eukaryotes. This study presents new data from chloro-, rhodo-, and glaucophytes as well as from complex algae, originating from secondary or tertiary endosymbioses, and indicate that the cytosolic GPI represents a genuine marker gene of the host cell.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Algal Material
The haptophytes Pavlova lutheri (strain 926-1) and Emiliania huxleyi (strain 33.90) and the glaucophyte Cyanophora paradoxa (strain 29.80 [Pringsheim isolate]) were obtained from the "Sammlung von Algenkulturen" at the University of Göttingen (SAG). Culturing was performed as previously described (Petersen et al. 2003
, Petersen, Teich, Brinkmann, Cerff 2006). The plant material of the rhodophyte Chondrus crispus was field collected on the North Sea island Helgoland.
Isolation of Nucleic Acids and Construction of Libraries
The isolation of nucleic acids and the preparation of 
ZAPII libraries from P. lutheri and C. crispus as well as the construction of a DASHII library from the glaucophyte C. paradoxa (strain 29.80 [Pringsheim isolate]) have been previously described (Petersen et al. 2003, Petersen, Teich, Brinkmann, Cerff 2006; Teich, Grauvogel, Petersen 2007). The ZAPII libraries from Lingulodinium polyedrum and Pyrocystis lunula were donated from Woodland Hastings (Harvard University, Cambridge, MA), the ZAP Express library from Paramecium tetraurelia was obtained from Jürgen Linder (Tübingen, Germany), and the NM1149 libraries from Hanusia phi and C. paradoxa (strain 45.84 [Kies isolate]) as well as the EMBL3 library from Phaeodactylum tricornutum (UTEX strain 642) were given by Marie-Fran
oise Liaud (Braunschweig, Germany).
Reverse Transcriptase–Polymerase Chain Reaction Amplification
Homologous probes were amplified via reverse transcriptase–polymerase chain reaction (RT-PCR) using the Thermoscript RT-PCR System (Invitrogen, Carlsbad, CA). Degenerated primers for PCR amplification were designed based on the universally conserved GPI motifs AFWDWVGG (5'-GCSTTCTGGGAYTGGGTNGGNGG-3') and FDQWGVEL (5'-AGCTCSACNCCCCAYTGRTCRAA-3'). Reverse transcription, PCR amplification, and cloning were performed as previously described (Petersen, Teich, Brinkmann, Cerff 2006). The different GPI clones were identified by sequencing using radioactive and fluorescence techniques.
Isolation and Sequencing of cDNA and Genomic Clones
Pavlova lutheri, P. tricornutum, L. polyedrum, P. lunula, C. crispus, C. paradoxa, and P. tetraurelia libraries were screened using 32P-labeled homologous RT-PCR probes for GPI genes (Petersen et al. 2003, Petersen, Teich, Brinkmann, Cerff 2006). The GPI screening of the H. phi library was performed with previously identified probes under less stringent conditions. The cDNA clones from ZAPII cDNA libraries were subcloned in pBluescript II SK(+) by single-clone excision (ZAP Express: pBK-CMV). The genomic clones from P. tricornutum and C. paradoxa were subcloned into the PstI and XhoI sites of pBluescript II SK(+), respectively. Specific GPI primers for RT-PCR amplification from E. huxleyi were constructed on the basis of a genomic contig merged from National Center for Biotechnology Information (NCBI) trace files. The Kazusa Institute (Japan) established a GPI expressed sequence tag (EST) of the rhodophyte Porphyra yezoensis [GenBank: AV429872] and kindly donated the corresponding cDNA clone. All clones were sequenced on both strands using pBluescript or gene-specific primers.
Identification of Clones by Database Analyses
GPI sequences from the red algae Galdieria sulphuraria and Cyanidioschyzon merolae were identified from the Galdieria database (GDB; Michigan State University) and the "Cyanidioschyzon merolae Genome Project" (http://merolae.biol.s.u-tokyo.ac.jp/), respectively, and those of the diatom Thalassiosira pseudonana were established at the "Joint Genome Institute" (JGI, Walnut Creek, CA) and preliminary sequence data of the alveolate Perkinsus marinus were obtained from "The Institute for Genomic Research" (TIGR; http://www.tigr.org). Nucleotide and deduced amino acid sequences of GPI were used as query sequences for Blast searches in the NCBI database (TBlastN; BlastP) and sequences from E. huxleyi, P. tricornutum and Tetrahymena thermophila were obtained using MEGABLAST. The respective clones were retrieved from GenBank and assembled into contigs if necessary (supplementary table S1, Supplementary Material online). The authenticity of the GPI sequences from Tetrahymena and Paramecium was confirmed by the ciliate-specific codon usage.
Sequence Handling and Phylogenetic Analyses
The initial alignments obtained with ClustalX (Thompson et al. 1997) were manually refined using the ED option of the MUST program package (Philippe 1993). All data sets were analyzed by 3 different likelihood methods with a model based on the Whelan and Goldman (WAG) matrix of amino acid replacements assuming a proportion of invariant positions and gamma distributed rates (WAG + I + 
4). The maximum likelihood (ML) analyses were carried out with PhyML version 2.4 (Guindon and Gascuel 2003) and Treefinder (Jobb et al. 2004), whereas the Bayesian inference was performed with MrBayes version 3.0B4 (Huelsenbeck and Ronquist 2001), where 200,000 generations were completed with trees collected each tenth generation. The number of generations needed until the likelihood values converged (burn-in) was typically less than 5% of the total. Bootstrap analyses with 100 replicates were performed with both Treefinder and PhyML using the same model as described above to estimate the support for internal nodes. The CONSENSE option of the PHYLIP package (http://evolution.genetics.washington.edu/phylip.html) was used to generate the bootstrap consensus trees.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Isolation and Characterization of GPI Sequences
Here, we present 12 new GPI sequences from 10 different species, including 7 complex algae with red plastids, 2 Plantae, and 1 ciliate. The cDNA clones were isolated from the cryptophyte H. phi (formerly designated as Cryptomonas phi), 2 haptophytes (P. lutheri and E. huxleyi), 2 dinoflagellates (L. polyedrum and P. lunula), 2 rhodophytes (P. yezoensis and C. crispus—two sequences), and the ciliate P. tetraurelia. From 2 strains of the glaucophyte C. paradoxa, we identified a partial cDNA (Kies; SAG 45.84), as well as an independent genomic GPI sequence (Pringsheim, SAG 29.80). Finally, we isolated a genomic GPI from the diatom P. tricornutum. The precise positions of 8 Cyanophora and 2 Phaeodactylum introns were determined by comparison with cDNA sequences obtained via RT-PCR. All newly established sequences were deposited at GenBank (accession numbers DQ812889–DQ812900).
Many GPI reference sequences were retrieved from public databases by extensive data mining via Blast searches. Raw sequence data (e.g., NCBI trace files) were merged into contigs and the deduced amino acid sequences subsequently used for phylogenetic analyses (see below). Although some Archaea are known to contain nonhomologous GPI genes belonging to the cupin superfamily (Verhees et al. 2001), we could identify 5 archaeal GPI sequences that are clearly homologous to their eukaryotic counterparts. Most bacteria contain homologous GPIs, and we used genes from various lineages including those of cyano- and proteobacteria for our phylogenetic analyses. We moreover assembled a broad data set of eukaryotic sequences including those of metazoa, fungi, kinetoplastids, Plantae (green and red algae, land plants), alveolates (ciliates, apicomplexa), anaerobic (amitochondriate) eukaryotes, and several other protists including P. infestans and Dictyostelium discoideum. The GPI sequence of the alveolate parasite P. marinus was established from a preliminary TIGR contig, and 8 intron positions were determined in comparison with our alignments. The distribution of GPI genes in complex algae with red plastids, such as the haptophyte E. huxleyi and the completely sequenced diatoms P. tricornutum and T. pseudonana, was thoroughly investigated. Whereas 4 genes are present in both diatoms, our analyses indicate the existence of only one GPI gene in the genome of E. huxleyi (
220 Mb) in spite of a 4-fold coverage at the time of our study (NCBI trace files).
Distribution of Plastid GPI Sequences among Plantae
The plastid GPI of land plants has a cyanobacterial origin (Nowitzki et al. 1998) but orthologs are neither present in the completely sequenced genomes of chlorophytes (Chlamydomonas and Volvox) nor in those of rhodophytes (Cyanidioschyzon and Galdieria). In contrast, we could identify orthologs of the cytosolic GPI genes of land plants from all 3 lineages of Plantae. In comparison with the start codon of Arabidopsis (P34795
[GenBank]
), translation generally starts in a range of about 5 amino acids throughout glycolytic GPI sequences, irrespective of their eukaryotic or eubacterial origin (alignment not shown). However, the deduced GPI sequences of 2 chlorophytes (Chlamydomonas, Volvox) and 3 rhodophytes (Porphyra, Galdieria, and Cyanidioschyzon) contain N-terminal extensions of more than 60 amino acids (supplementary fig. S1, Supplementary Material online). Their existence is confirmed by EST-data, excluding an artifactual assignment due to hidden introns. Analyses with the prediction programs TargetP (Emanuelsson et al. 2000) and ChloroP (Emanuelsson et al. 1999) clearly indicate that these chlorophycean as well as rhodophycean extensions are transit peptides for plastid import (supplementary fig. S1, Supplementary Material online). Accordingly, we designate these sequences with the suffix "pla" in the phylogenetic analyses below.
Phylogenetic Analyses of GPI Sequences
Figure 1 shows in a schematic form a comprehensive phylogenetic tree inferred by Treefinder using a ML approach based on 179 GPI sequences and 256 amino acid positions. The complete analysis and the accession numbers of the 85 eukaryotic, 89 bacterial, and 5 archaeal sequences are presented as supplementary material online (supplementary fig. S2 and table S1). The tree is divided into 3 distinct assemblages designated as I, II, and III following Nowitzki et al. (1998). Assemblage III, which consists of archaeal and bacterial sequences of Bacilli and a highly divergent gene of Mycoplasma, was used as an outgroup to root the tree (fig. 1). Assemblage II contains essentially nuclear-encoded plastidial genes of land plants and their cyanobacterial relatives, several 
-proteobacteria and the low GC Desulfitobacterium, as well as a basally branching group of three closely related GPIs from anaerobic protists (see Henze et al. 2001). The distinct assemblage I comprises the majority of eubacterial as well as eukaryotic sequences and the latter are located in 3 subtrees (I-A, I-B, and I-C; fig. 1). A conspicuous finding in the basal branch I-C is a specific association between the cryptophyte H. phi and chlamydiae (100% bootstrap proportion [BP]). We can exclude that the newly established cDNA sequence of H. phi is a bacterial contaminant because a very similar EST clone was identified from the closely related cryptophyte Guillardia theta (155 of 158 deduced amino acids are identical [Zauner S, unpublished data]). Thus, the location of this GPI sequence argues for a specific acquisition from eubacteria by cryptophytes via horizontal gene transfer (HGT). Another striking feature is the very long common branch of subtree I-B, which emerges from the broad diversity of unresolved eubacterial sequences. This highly supported subtree harbors GPI sequences from Plantae (green plants, rhodophytes, and the glaucophyte Cyanophora), alveolates (ciliates, apicomplexa, Perkinsus, and dinoflagellates), and stramenopiles (represented by diatoms and the oomycete Phytophthora). A separate analysis of this subtree without divergent and partial sequences is described below (see fig. 3). Finally, all remaining eukaryotic GPI sequences emerge from eubacteria (I-A), but this part of the tree is poorly resolved.
|
|
50% are shown.To increase phylogenetic resolution, we reanalyzed the sequences of subtree I-A separately after exclusion of the divergent GPI of Encephalitozoon and the basal branching bacterial sequences. The resulting ML tree based on 53 sequences and 460 amino acid positions is shown in figure 2. The monophyly of metazoa that was not recovered in the exhaustive "overview" tree (fig. 1) is now moderately supported (63/76% BP). Moreover, the new analysis shows a common branching of all fungi, though with a weak bootstrap support (51/60%), which probably results from the very fast evolutionary rates of the ascomycete sequences (Felsenstein 1978
; Brinkmann et al. 2005). Accordingly, a separate analysis without the divergent ascomycetes (supplementary fig. S3, Supplementary Material online) strengthens the bootstrap support for a common branching of the zygomycete Rhizopus, the ascomycete Schizosaccharomyces, and the 3 basidiomycetes (51/60% to 79/81%). Although there is no doubt about the monophyly of the closely related kinetoplastids, their affinity to other sequences as well as the affiliation of Dictyostelium and Entamoeba remains inconclusive (fig. 1). Diatoms are the sole eukaryotic lineage that is found in both subtrees (I-A: 1 gene; I-B: 3 genes) and the GPI-4 sequence of subtree I-A clusters unsupported with haptophyte GPI sequences (fig. 2). However, the rising statistical support upon successive elimination of divergent sequences (37% [fig. 1], 44/48% [fig. 2], 74/77% [supplementary fig. S3, Supplementary Material online]) suggests that this association actually reflects a common origin of the corresponding sequences.
|
A MrBayes analysis of subtree I-B based on 36 sequences and 450 amino acid positions, which represents the topology with the best likelihood value (ML; not recovered by PhyML and Treefinder due to problems with the heuristic search), is shown in figure 3. The GPI sequences of Plantae are exclusively located in this subtree, with the exception of plastid GPIs from land plants (assemblage II; fig. 1), whose cyanobacterial affiliation demonstrates their endosymbiotic origin. The cytosolic genes of land plants share a strongly supported common origin together with those of Chlamydomonas and Volvox (100% BP). This is of special interest because the presence of a transit peptide in the chlorophycean sequences (see above; supplementary fig. S1-A, Supplementary Material online) suggests a plastid localization of the corresponding proteins, which is compatible with physiological studies (Schnarrenberger et al. 1990
). Therefore, our analyses propose that the cytosolic GPI enzyme of chlorophycean green algae secondarily recruited a transit peptide for chloroplast import. The same explanation probably applies for a distinct group of closely related rhodophycean sequences (pla; fig. 3), which also contain transit peptides, but most likely obtained them independently from the chlorophycean pendants (see above; supplementary fig. S1-B, Supplementary Material online). Rhodophytes accordingly harbor a cytosolic as well as a plastid-targeted enzyme both located in subtree I-B (fig. 1). The presence of closely related plastidial genes in all red algae included, which represent 2 distant lineages (Yoon et al. 2006), suggests a unique ancient origin via a gene duplication. Finally, we cannot exclude that the 2 partial GPI sequences of Cyanophora (fig. 1) also represent cytosolic and plastid-targeted isoenzymes, and thus a third most likely independent case where a previously cytosolic GPI enzyme is relocated into the plastid.
In addition to the GPI sequences of Plantae, all alveolate genes are exclusively located in subtree I-B (fig. 1), and our separate phylogenetic analysis further indicates that the latter superensemble is monophyletic (72/81% BP; fig. 3). In agreement with the assumption that GPI sequences represent the host cell, both ciliates and dinoflagellates moreover form well-supported distinct branches (99/99 and 100/100% BP, respectively). We excluded the divergent GPI sequences of Theileria from the analysis because they provoke an in-group long branch attraction artifact within the branch of apicomplexan parasites (data not shown; Brinkmann et al. 2005; Philippe et al. 2005). The accelerated evolutionary rates of Theileria and Plasmodium seem to be a general characteristic of these species (data not shown) and may explain the low statistic support for the whole assembly. However, the quite slow evolutionary rate of cytosolic GPI sequences from dinoflagellates is noteworthy because several of their plastid-targeted proteins exhibit extraordinarily long branches that strongly disturb phylogenetic reconstruction (Bachvaroff et al. 2006; Petersen, Teich, Brinkmann, Cerff 2006). Our GPI analyses display a close and well-supported (fig. 3; 93/96% BP) affiliation between dinoflagellates and the oyster parasite Perkinsus (Goggin and Barker 1993). It contradicts previous speculations about a placement of Perkinsidae with apicomplexa (Levine 1978; Perkins 1978) and is in agreement with recent studies based on SSU rRNA, heat shock protein 90, and structural genes (Reece et al. 1997; Saldarriaga et al. 2003; Leander and Keeling 2004). Finally, subtree I-B (fig. 1) contains stramenopile GPI sequences of the oomycete Phytophthora and 3 of 4 diatom specific branches (GPI-1,-2, and -3). The close affiliation of diatom GPI-3 with the plastid subtree of rhodophytes strongly suggests a common origin of the respective sequences. In an analysis of a reduced data set, these diatom sequences are found within the rhodophycean clade together with the sequences of Chondrus and Porphyra (data not shown), which indicates that diatoms recruited this plastidial gene in the course of a secondary plastid endosymbiosis with a red alga. EST data are missing, but we could amplify by RT-PCR the putative bipartite signal-transit peptide including a characteristic consensus motif (A
FVP) for protease cleavage from Phaeodactylum, indicating that GPI-3 of diatoms may be still targeted into the complex plastid (Kilian and Kroth 2004). Diatom GPI-2 genes probably originated from a GPI-1/2 gene duplication early in diatom evolution. Though representing the most divergent clade of eukaryotic GPIs (fig. 1; supplementary fig. S2, Supplementary Material online), they are still expressed, as documented by a Phaeodactylum EST (CD377263). According to TargetP and ChloroP (Emanuelsson et al. 1999, 2000), the GPI-1 of diatoms is likely a cytosolic enzyme, and the exclusive presence of a Plantae-related GPI in oomycetes (various Phytophthora species; Ospina-Giraldo and Jones 2003) suggests that these sequences represent the ancient host cell enzyme. The weak statistic support for the stramenopile subtree (GPI-1 and Phytophthora; fig. 3) may be the result of an about 2.5-fold accelerated evolutionary rate of diatom GPI-1 genes (compared with GPI-3).
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Distribution of Plastid GPI in Photosynthetic Eukaryotes and Implications for Plant Physiology
GPI is an essential enzyme of glycolysis/gluconeogenesis and is present in heterotrophic and photosynthetic eukaryotes, most Bacteria, and some Archaea (fig. 1). A plastid-targeted homolog of land plants is required for starch production and degradation, and it is crucial for the plastid OPPP (Martin and Herrmann 1998
) that assures reduced nicotinamide adenine dinucleotide phosphate generation at night. Land plants acquired the respective gene from the engulfed cyanobacterium after primary endosymbiosis (fig. 1; Nowitzki et al. 1998), but cyanobacterial orthologs are missing in the completely sequenced genomes of 2 chlorophycean green algae (Chlamydomonas and Volvox) and of 2 rhodophytes (Cyanidioschyzon and Galdieria). However, our in silico analyses strongly suggest that both chlorophytes and rhodophytes harbor plastid GPIs, which independently evolved from cytosolic pendants by recruitment of transit peptides (fig. 3; supplementary fig. S1, Supplementary Material online). Moreover, the absence of cytosolic GPI is coherent with the loss of cytosolic fructose-1,6-bisphosphate aldolase (FBA) and fructose-1,6-bisphosphatase (FBP) genes in Chlamydomonas (Schnarrenberger et al. 1994; Teich et al. 2007) and with biochemical studies indicating that at least the upper half of glycolysis is exclusively present in the plastid of unicellular chlorophytes (Schnarrenberger et al. 1990; Plaxton 1996). An Arabidopsis mutant of the plastid-targeted GPI documents the essential function of plastid GPI for starch accumulation within the chloroplast (Yu et al. 2000). Thus, the plastid localization of the single GPI of Chlamydomonas and Volvox is consistent with the observation that starch is stored in plastids of both land plants and chlorophytes (van den Hoek et al. 1996). However, the presence of a plastid GPI in rhodophytes, which accumulate floridean starch outside the plastid in the cytoplasm (Viola et al. 2001), suggests that the raison d'être of this enzyme is probably its essential role in the OPPP. At least 2 scenarios can explain the patchy distribution of plastid-targeted GPI sequences in Plantae. First, this gene may have still been plastid encoded when the 3 primary lineages originated and the endosymbiotic gene transfer (EGT) occurred later in a common ancestor of land plants. Accordingly, chlorophytes, rhodophytes, and perhaps also glaucophytes would have replaced the genuine plastid-encoded gene by preexisting cytosolic duplicates. Second, the EGT may have occurred early in Plantae evolution and the nuclear encoded cyanobacterial GPI was secondarily replaced several times. Irrespective of which scenario is correct, the probably universal presence of plastid-targeted GPI genes in Plantae documents the importance of its function for the metabolism of primary plastids.
The presumable absence of a plastid-targeted GPI in E. huxleyi indicates that the plastid metabolism of haptophytes underwent fundamental changes after secondary endosymbiosis including the loss of the OPPP in the red algal endosymbiont. A similar observation was previously made for diatoms (Michels et al. 2005) and our group has proposed (Petersen, Teich, Becker 2006) that this loss might have been caused by the replacement of the red algal (plastid) glyceraldehyde-3-phosphate dehydrogenase (GapA) by a cytosolic isoenzyme (GapC-I; Liaud et al. 1997). In contrast to GapA of Plantae, the novel GapC-I is unable to mediate the inactivation of the Calvin cycle at night. Therefore, the presence of OPPP would result in futile cycling associated with a waste of ATP. Because GapC-I is present in all orders of complex algae with red plastids, such as stramenopiles, haptophytes, cryptophytes, and dinoflagellates (Harper and Keeling 2003), we hypothesize that the plastid OPPP is also absent from the latter 2 lineages. If true, this would imply that the formerly crucial plastid GPI is actually dispensable and may have been lost.
Finally, the presence of 4 different GPI genes in diatoms is noteworthy and a similar redundancy is observed for multiple isoenzymes of the primary metabolism such as FBA (Kroth et al. 2005) or FBP (Teich et al. 2007). However, it is unclear whether the additional GPI genes only ensure a more sophisticated regulation or have a different metabolic function, for example, in mitochondrial glycolysis. The presence of this pathway previously discovered in stramenopiles (oomycetes and diatoms; Liaud et al. 2000) is validated at least for the lower glycolytic part from triose-phosphate isomerase to pyruvate kinase (Río Bártulos C, personal communication), but we cannot exclude that it is complete in diatom mitochondria.
Cytosolic GPI Is a Marker of the Host Cell—Implications for Algal Evolution
In addition to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and FBA that represent the best investigated gene systems of primary metabolism, GPI is another marker that has been used for the reconstruction of gene and organismal evolution. In the past, peculiar relationships based on GPI phylogenies were proposed, such as, for example, a close affinity between the apicomplexan parasite Toxoplasma and green plants (Dzierszinski et al. 1999). Moreover, a sister-group relationship between -proteobacteria and metazoa was taken as a bona fide example of a nonendosymbiotic transkingdom gene transfer from eukaryotes to prokaryotes (Katz 1996). However, we could not recover this specific affiliation (see figs. 1 and 2) and follow the interpretation of Nowitzki et al. (1998). Their criticism especially pointed toward the limited availability of eubacterial reference sequences a decade ago, which may have led to invalid conclusions. In the meantime, the wealth of whole genome sequences from various prokaryotes as well as many additional eukaryotic sequences allow a much better taxon sampling. In that respect, the present study displays the most comprehensive phylogenetic analysis of GPI sequences so far (fig. 1; supplementary fig. S2, Supplementary Material online). We also established a representative data set of eukaryotic GPIs including sequences of the 2 superensembles Alveolata and Plantae, as well as of all lineages of complex algae with red plastids (haptophytes, diatoms, cryptophytes, and dinoflagellates). The distinct subtree of alveolate GPI sequences including well-supported branches of dinoflagellates and ciliates (fig. 3) is compatible with 18S rDNA analyses (Van de Peer and De Wachter 1997). Our comprehensive phylogenies document the alveolate affiliation of apicomplexan GPI sequences including those of Toxoplasma (see Dzierszinski et al. 1999) and suggest a specific relationship between Plantae, stramenopiles, and alveolates. Furthermore, the specific relationship of the oyster parasite Perkinsus together with dinoflagellates (fig. 3) is in agreement with analyses of structural genes such as actin and tubulin (Saldarriaga et al. 2003). The validation of the current classification of Perkinseae (Goggin and Barker 1993; Reece et al. 1997; Leander and Keeling 2004) is of special relevance because very recent studies, for example, of our group deliver strong evidence for the presence of a so far hidden plastid in these protists (Grauvogel C, Reece KS, Brinkmann H, Petersen J, in preparation; Stelter et al. 2007). Taken together, the cytosolic GPI of alveolates is a marker of the host cell and its distribution was not affected by plastid acquisition of dinoflagellates, Perkinseae, and apicomplexa via eukaryote-to-eukaryote endosymbiosis.
We have established GPI sequences from all lineages with complex red plastids (haptophytes, cryptophytes, diatoms, and dinoflagellates), and our phylogenetic analyses revealed that this host cell related gene was occasionally replaced in eukaryotic evolution. In particular, the GPI of cryptophytes was probably acquired via HGT as indicated by the unexpected placement of Hanusia in a bacterial subtree together with chlamydiales (I-C; fig. 1). Although diatoms harbor altogether 4 genes, it is likely that GPI-1 and GPI-2, which originate from a gene duplication, represent their genuine cytosolic isogene because they are found in the same subtree I-B as the sole GPI of the nonphotosynthetic stramenopile Phytophthora (figs. 1 and 3). On the other hand, the third gene of this clade (GPI-3) was probably recruited in an endosymbiotic context from a rhodophyte (fig. 3). With respect to our comprehensive GPI phylogeny (fig. 1), assemblage I-A generally obtains weak statistic support, but a separate analysis provides solid bootstrap values for a common branching of diatom GPI-4 and the exclusive homolog of haptophytes (74/77%; supplementary fig. S3, Supplementary Material online). Therefore, it is possible that diatoms obtained this gene by a lateral recruitment from a haptophyte, and we cannot exclude that this acquisition occurred via a eukaryote-to-eukaryote endosymbiosis (Teich et al. 2007).
The well-supported subtree of Plantae, alveolates, and stramenopiles (I-B, fig. 1; supplementary fig. S1, Supplementary Material online), which is separated by a long common branch from the rest of GPI sequences, is the most striking feature of our GPI phylogeny. A separate analysis of this clade including several bacterial sequences exhibits a moderately supported relationship (70–74%) with the GPI of the unnamed 
-proteobacterium HTCC2255 (EAU52689
[GenBank]
), whose draft genome sequence was recently released (supplementary fig. S4, Supplementary Material online). If the phylogeny is reliable and the gene was not recruited via HGT, this finding is compatible with an -proteobacterial/mitochondrial origin that was earlier proposed for the complete glycolytic pathway (Martin and Herrmann 1998). The remaining eukaryotic GPI sequences of assemblage I-A, including those of haptophytes, metazoa, and fungi (Opisthokonta) as well as kinetoplastids may have an independent proteobacterial origin (fig. 2). Although their specific relationship to -proteobacterial sequences cannot be excluded, it is not resolved due to the slow evolutionary rate of subtree I-A (fig. 1). The alternative explanation of a common origin of all eukaryotic sequences from subtrees I-A and I-B is quite speculative and not supported by our phylogenies. Therefore, we suppose that a gene duplication early in eukaryotic evolution followed by the differential loss of 1 of the 2 copies is unlikely. However, irrespective of its precise evolutionary origin, the accelerated mutation rate at the base of the eukaryotic GPI subtree I-B may be a fortunate coincidence that reveals an important aspect of the ancient eukaryotic evolution with a single marker of the host cell. Indeed, a sister-group relationship between the superensembles stramenopiles and Alveolata was previously observed for 18S rDNA (Van de Peer and De Wachter 1997). Furthermore, a specific association of these 2 superensembles with Plantae to the exclusion of metazoa and fungi (opisthokonts) was also recovered by a large concatenated data set of the host cell, which, however, lacks representatives of 2 currently proposed superensembles, that is, Excavates and Rhizaria (Rodriguez-Ezpeleta et al. 2005). Because the observed GPI phylogeny, that is the assemblage of Plantae, alveolates, and stramenopiles, is in agreement with these observations, it may actually reflect the organismal relationships. The reconstruction of eukaryotic evolution is one of the biggest challenges for present day biologists, and further analyses of single "lucky genes" (Bapteste et al. 2002) as well as large concatenated data sets (Rodriguez-Ezpeleta et al. 2005) will help to enlighten the puzzle of algal biodiversity and their genomic mosaics originating from eukaryote-to-eukaryote endosymbioses.
|
Supplementary Material |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Supplementary table S1 and figures S1–S4 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/). Sequences established in this study have been deposited in the DDBJ/EMBL/GenBank International Nucleotide Sequence Database under the following accession numbers: DQ812889–DQ812900.
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
We are grateful to Ulrike Brandt and Sarina Scharbatke for excellent technical assistance and Denis Baurain (Liége, Belgium) for very helpful comments on the manuscript. We thank Woody Hastings (Harvard University, USA) for the cDNA libraries from L. polyedrum and P. lunula, Marie-Fran
oise Liaud (Braunschweig, Germany) for cDNA libraries from H. phi and C. paradoxa, and a genomic library from P. tricornutum, Jürgen Linder (Tübingen, Germany) for a cDNA library from P. tetraurelia, the Kazusa DNA Research Institute (Japan) for providing a GPI cDNA clone from P. yezoensis and René Teich (Marburg, Germany) for establishment of a cDNA clone from P. lutheri. Preliminary sequence data of P. marinus were obtained from TIGR through the Web site at http://www.tigr.org and sequencing was accomplished with support from the National Science Foundation. The sequence data from G. sulphuraria were provided by the Michigan State University GDB (http://genomics.msu.edu/galdieria). Major financial support, including a PhD stipend for C.G., was received from the Deutsche Forschungsgemeinschaft (CE 1/27-2). The authors also want to thank 2 anonymous reviewers for careful reading and constructive criticism of the manuscript. |
Footnotes |
|---|
Martin Embley, Associate Editor
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionSupplementary MaterialAcknowledgementsReferences |
|---|
Adl SM, Simpson AG, Farmer MA. (28 co-authors). The new higher level classification of eukaryotes with emphasis on the taxonomy of protists. J Eukaryot Microbiol (2005) 52:399–451.[CrossRef][ISI][Medline]
Bachvaroff TR, Sanchez-Puerta MV, Delwiche F. Rate variation as a function of gene origin in plastid-derived genes of peridinin-containing dinoflagellates. J Mol Evol (2006) 62:42–52.[CrossRef][ISI][Medline]
Bapteste E, Brinkmann H, Lee JA. (11 co-authors). The analysis of 100 genes supports the grouping of three highly divergent amoebae: Dictyostelium, Entamoeba, and Mastigamoeba. Proc Natl Acad Sci USA (2002) 99:1414–1419.
Brinkmann H, van der Giezen M, Zhou Y, Poncelin de Raucourt G, Philippe H. An empirical assessment of long-branch attraction artefacts in deep eukaryotic phylogenomics. Syst Biol (2005) 54:743–757.[ISI][Medline]
Cavalier-Smith T. Eukaryote kingdoms: seven or nine? Biosystems (1981) 14:461–481.[CrossRef][ISI][Medline]
Delwiche C. Tracing the thread of plastid diversity through the tapestry of life. Am Nat (1999) 154:S164–S177.[CrossRef][ISI][Medline]
Delwiche C, Palmer J. The origin of plastids and their spread via secondary symbiosis. In: Origins of algae and their plastids—Bhattacharya D, ed. (1997) Wien (Germany): Springer Verlag. 53–86.
Douglas S. Plastid evolution: origins, diversity, trends. Curr Opin Genet Dev (1998) 8:655–661.[CrossRef][ISI][Medline]
Dzierszinski F, Popescu O, Toursel C, Slomianny C, Yahiaoui B, Tomavo S. The protozoan parasite Toxoplasma gondii expresses two functional plant-like glycolytic enzymes. Implications for evolutionary origin of apicomplexans. J Biol Chem (1999) 274:24888–24895.
Emanuelsson O, Nielsen H, Brunak S, von Heijne G. Predicting subcellular localization of proteins based on their N-terminal amino acid sequence. J Mol Biol (2000) 300:1005–1016.[CrossRef][ISI][Medline]
Emanuelsson O, Nielsen H, von Heijne G. ChloroP, a neural network-based method for predicting chloroplast transit peptides and their cleavage sites. Protein Sci (1999) 8:978–984.[Abstract]
Felsenstein J. Cases in which parsimony and compatibility methods will be positively misleading. Syst Zool (1978) 27:401–410.
Foth BJ, McFadden GI. The apicoplast: a plastid in Plasmodium falciparum and other apicomplexan parasites. Int Rev Cytol (2003) 224:57–110.[ISI][Medline]
Funes S, Davidson E, Reyes-Prieto A, Magallon S, Herion P, King M, Gonzalez-Halphen D. A green algal apicoplast ancestor. Science (2002) 298:2155.
Funes S, Davidson E, Reyes-Prieto A, Magallon S, Herion P, King M, Gonzalez-Halphen D. Response to comment on "A green algal apicoplast ancestor". Science (2003) 301:49b.
Goggin CL, Barker SC. Phylogenetic position of the genus Perkinsus (Protista, Apicomplexa) based on small subunit ribosomal RNA. Mol Biochem Parasitol (1993) 60:65–70.[CrossRef][ISI][Medline]
Guindon S, Gascuel O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol (2003) 52:696–704.[CrossRef][ISI][Medline]
Harper J, Keeling P. Nucleus-encoded, plastid-targeted glyceraldehyde-3-phosphate dehydrogenase (GAPDH) indicates a single origin for chromalveolate plastids. Mol Biol Evol (2003) 20:1730–1735.
Henze K, Horner DS, Suguri S, Moore DV, Sanchez LB, Muller M, Embley TM. Unique phylogenetic relationships of glucokinase and glucosephosphate isomerase of the amitochondriate eukaryotes Giardia intestinalis, Spironucleus barkhanus and Trichomonas vaginalis. Gene (2001) 281:123–131.[CrossRef][ISI][Medline]
Huelsenbeck JP, Ronquist F. MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics (2001) 17:754–755.
Jobb G, von Haeseler A, Strimmer K. TREEFINDER: a powerful graphical analysis environment for molecular phylogenetics. BMC Evol Biol (2004) 4:18.[CrossRef][Medline]
Katz LA. Transkingdom transfer of the phosphoglucose isomerase gene. J Mol Evol (1996) 43:453–459.[ISI][Medline]
Kilian O, Kroth PG. Presequence acquisition during secondary endocytobiosis and the possible role of introns. J Mol Evol (2004) 58:712–721.[CrossRef][ISI][Medline]
Kroth PG, Schroers Y, Kilian O. The peculiar distribution of class I and class II aldolases in diatoms and in red algae. Curr Genet (2005) 48:389–400.[CrossRef][ISI][Medline]
Leander BS, Keeling PJ. Early evolutionary history of dinoflagellates and apicomplexans (alveolata) as inferred from hsp90 and actin phylogenies. J Phycol (2004) 40:341–350.[CrossRef][ISI]
Levine ND. Perkinsus gen. n. and other new taxa in the protozoan phylum Apicomplexa. J Parasitol (1978) 64:549.[Medline]
Liaud MF, Brandt U, Scherzinger M, Cerff R. Evolutionary origin of cryptomonad microalgae: two novel chloroplast/cytosol-specific GAPDH genes as potential markers of ancestral endosymbiont and host cell components. J Mol Evol (1997) 44(Suppl 1):S28–S37.[CrossRef][ISI][Medline]
Liaud MF, Lichtle C, Apt K, Martin W, Cerff R. Compartment-specific isoforms of TPI and GAPDH are imported into diatom mitochondria as a fusion protein: evidence in favor of a mitochondrial origin of the eukaryotic glycolytic pathway. Mol Biol Evol (2000) 17:213–223.
Martin W, Herrmann RG. Gene transfer from organelles to the nucleus: how much, what happens, and why? Plant Physiol (1998) 118:9–17.
McFadden GI, Reith ME, Munholland J, Lang-Unnasch N. Plastid in human parasites. Nature (1996) 381:482.[CrossRef][Medline]
Michels AK, Wedel N, Kroth PG. Diatom plastids possess a phosphoribulokinase with an altered regulation and no oxidative pentose phosphate pathway. Plant Physiol (2005) 137:911–920.
Nowitzki U, Flechner A, Kellermann J, Hasegawa M, Schnarrenberger C, Martin W. Eubacterial origin of nuclear genes for chloroplast and cytosolic glucose-6-phosphate isomerase from spinach: sampling eubacterial gene diversity in eukaryotic chromosomes through symbiosis. Gene (1998) 214:205–213.[CrossRef][ISI][Medline]
Ospina-Giraldo MD, Jones RW. Characterization of the glucose-6-phosphate isomerase gene in Phytophthora infestans reveals the presence of multiple alleles. Fungal Genet Biol (2003) 40:197–206.[CrossRef][ISI][Medline]
Perkins FO. The structure of Perkinsus marinus (Mackin, Owen and Collier, 1950) Levine, 1978 with comments on taxonomy and phylogeny of Perkinsus ssp. J Shellfish Res (1978) 15:67–87.
Petersen J, Brinkmann H, Cerff R. Origin, evolution, and metabolic role of a novel glycolytic GAPDH enzyme recruited by land plant plastids. J Mol Evol (2003) 57:16–26.[CrossRef][ISI][Medline]
Petersen J, Teich R, Becker B, Cerff R, Brinkmann H. The GapA/B gene duplication marks the origin of Streptophyta (charophytes and land plants). Mol Biol Evol (2006) 23:1109–1118.
Petersen J, Teich R, Brinkmann H, Cerff R. A "Green" phosphoribulokinase in complex algae with red plastids: evidence for a single secondary endosymbiosis leading to haptophytes, cryptophytes, heterokonts and dinoflagellates. J Mol Evol (2006) 62:143–157.[CrossRef][ISI][Medline]
Philippe H. MUST, a computer package of management utilities for sequences and trees. Nucleic Acids Res (1993) 21:5264–5272.
Philippe H, Zhou Y, Brinkmann H, Rodriguez-Ezpeleta N, Delsuc F. Heterotachy and long-branch attraction in phylogenetics. BMC Evol Biol (2005) 5:50.[CrossRef][Medline]
Plaxton WC. The organization and regulation of plant glycolysis. Annu Rev Plant Physiol Plant Mol Biol (1996) 47:185–214.[CrossRef][ISI][Medline]
Reece KS, Siddall ME, Burreson EM, Graves JE. Phylogenetic analysis of Perkinsus based on actin gene sequences. J Parasitol (1997) 83:417–423.[CrossRef][Medline]
Rodriguez-Ezpeleta N, Brinkmann H, Burey SC, Roure B, Burger G, Löffelhardt W, Bohnert HJ, Philippe H, Lang BF. Monophyly of primary photosynthetic eukaryotes: green plants, red algae, and glaucophytes. Curr Biol (2005) 15:1325–1330.[CrossRef][ISI][Medline]
Rogers MB, Gilson PR, Su V, McFadden GI, Keeling PJ. The complete chloroplast genome of the chlorarachniophyte Bigelowiella natans: evidence for independent origins of chlorarachniophyte and euglenid secondary endosymbionts. Mol Biol Evol (2007) 24:54–62.
Saldarriaga JF, McEwan ML, Fast NM, Taylor FJ, Keeling PJ. Multiple protein phylogenies show that Oxyrrhis marina and Perkinsus marinus are early branches of the dinoflagellate lineage. Int J Syst Evol Microbiol (2003) 53:355–365.
Schnarrenberger C, Jacobshagen S, Müller B, Krüger I. Evolution of isoenzymes of sugar phosphate metabolism in green algae. Prog Clin Biol Res (1990) 344:743–764.[Medline]
Schnarrenberger C, Pelzer-Reith B, Yatsuki H, Freund S, Jacobshagen S, Hori K. Expression and sequence of the only detectable aldolase in Chlamydomonas reinhardtii. Arch Biochem Biophys (1994) 313:173–178.[CrossRef][ISI][Medline]
Stelter K, El-Sayed NM, Seeber F. The expression of a plant-type ferredoxin redox system provides molecular evidence for a plastid in the early dinoflagellate Perkinsus marinus. Protist (2007) 158:119–130.[Medline]
Stoebe B, Kowallik K. Gene-cluster analysis in chloroplast genomics. Trends Genet (1999) 15:344–347.[CrossRef][ISI][Medline]
Teich R, Grauvogel C, Petersen J. Intron distribution in Plantae: 500 million years of stasis during land plant evolution. Gene (2007) 394:96–104.[CrossRef][Medline]
Teich R, Zauner S, Brinkmann H, Petersen J, Forthcoming. Origin and distribution of Calvin cycle fructose and sedoheptulose bisphosphatases in plantae and complex algae. A single secondary origin of complex red plastids and subsequent propagation via tertiary endosymbioses. Protist (2007).
Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res (1997) 25:4876–4882.
van den Hoek C, Mann D, Jahns HM. Algae: an introduction to phycology (1996) Cambridge: Cambridge University Press.
Van de Peer Y, De Wachter R. Evolutionary relationships among the eukaryotic crown taxa taking into account site-to-site rate variation in 18S rRNA. J Mol Evol (1997) 45:619–630.[CrossRef][ISI][Medline]
Verhees CH, Huynen MA, Ward DE, Schiltz E, de Vos WM, van der Oost J. The phosphoglucose isomerase from the hyperthermophilic archaeon Pyrococcus furiosus is a unique glycolytic enzyme that belongs to the cupin superfamily. J Biol Chem (2001) 276:40926–40932.
Viola R, Nyvall P, Pedersen M. The unique features of starch metabolism in red algae. Proc R Soc B (2001) 268:1417–1422.[Medline]
Waller RF, Keeling PJ, van Dooren GG, McFadden GI. Comment on "A green algal apicoplast ancestor". Science (2003) 301:49a.
Wilson RJ, Denny PW, Preiser PR. (11 co-authors). Complete gene map of the plastid-like DNA of the malaria parasite Plasmodium falciparum. J Mol Biol (1996) 261:155–172.[CrossRef][ISI][Medline]
Woese CR, Kandler O, Wheelis ML. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria. and Eucarya. Proc Natl Acad Sci USA (1990) 87:4576–4579.
Yoon HS, Muller KM, Sheath RG, Ott FD, Bhattacharya D. Defining the major lineages of red algae (Rhodophyta). J Phycol (2006) 42:482–492.[CrossRef][ISI]
Yu T-S, Lue W-L, Wang S-M, Chen J. Mutation of Arabidopsis plastid phosphoglucose isomerase affects leaf starch synthesis and floral initiation. Plant Physiol (2000) 123:319–325.
CiteULike 
Connotea 
Del.icio.us What's this?
| |||||||||||