Most of the available mitochondrial and plastid genome sequences are biased towards adenine and thymine (AT) over guanine and cytosine (GC). Examples of GC-rich organelle DNAs are limited to a small but eclectic list of species, including certain green algae. Here, to gain insight in the evolution of organelle nucleotide landscape, we present the GC-rich mitochondrial and plastid DNAs from the trebouxiophyte green alga Coccomyxa sp. C-169. We compare these sequences with other GC-rich organelle DNAs and argue that the forces biasing them towards G and C are nonadaptive and linked to the metabolic and/or life history features of this species. The Coccomyxa organelle genomes are also used for phylogenetic analyses, which highlight the complexities in trying to resolve the interrelationships among the core chlorophyte green algae, but ultimately favour a sister relationship between the Ulvophyceae and Chlorophyceae, with the Trebouxiophyceae branching at the base of the chlorophyte crown.
Citation: Smith DR, Burki F, Yamada T, Grimwood J, Grigoriev IV, Van Etten JL, et al. (2011) The GC-Rich Mitochondrial and Plastid Genomes of the Green Alga Coccomyxa Give Insight into the Evolution of Organelle DNA Nucleotide Landscape. PLoS ONE 6(8): e23624. doi:10.1371/journal.pone.0023624
Editor: Juergen Kroymann, French National Centre for Scientific Research, Université Paris-Sud, France
Received: May 13, 2011; Accepted: July 21, 2011; Published: August 26, 2011
Copyright: © 2011 Smith et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: DRS is supported by a postdoctoral fellowship from the Natural Sciences and Engineering Research Council (NSERC) of Canada. FB is supported by a prospective researcher postdoctoral fellowship from the Swiss National Science Foundation, and by a grant to the Centre for Microbial Diversity and Evolution from the Tula Foundation. JLVE was partially supported by grant P20-RR15635 from the COBRE program of the National Center for Research Resources (no additional external funding received for this study). PJK is funded by a grant from NSERC (227301). PJK is a Fellow of the Canadian Institute for Advanced Research and a Senior Scholar of the Michael Smith Foundation for Health Research. The work conducted by the DOE JGI is supported by the Office of Science of the U.S. Department of Energy under Contract Number DE-AC02-05CH11231. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Some of the most diverse and unusual mitochondrial and plastid DNAs (mtDNAs and ptDNAs) from all eukaryotes come from the Chlorophyta — a lineage comprising the majority of known green algal species . Both among and within the four main chlorophyte classes — Prasinophyceae, Trebouxiophyceae, Ulvophyceae, and Chlorophyceae — there is an impressive range of mitochondrial and plastid genome sizes, gene complements, and noncoding DNA compositions (see  for a compilation). Moreover, the Chlorophyceae boasts some of the few documented cases of species with guanine- and cytosine-rich (GC-rich) mtDNA, which is a significant feat when considering that most mitochondrial (and plastid) genomes are highly biased in adenine and thymine (AT). The nonphotosynthetic chlorophycean Polytomella capuana has one of the highest recorded mtDNA GC contents (57%) , and analyses of the mtDNA-encoded cox1 and cob from the chlorophyceans Lobochlamys segnis and Lobochlamys culleus revealed average GC compositions of 54 and 62%, respectively . The only other published examples of GC-rich mitochondrial genomes (as of March 2011) come from the parasitic fungus Candida subhashii , a few teleost species (, and references therein), and the lycophyte Selaginella moellendorffii , . The latter and other members of the Selaginella genus are also exceptional in that they are the only species known to contain GC-rich plastid genomes , .
The forces biasing organelle genomes towards G and C may differ among lineages. For instance, within the P. capuana mtDNA, the GC content is highest at silent sites, such as noncoding and third codon position, suggesting a neutral underpinning to the GC richness . Conversely, for the S. moellendorffii mitochondrial and plastid genomes, the levels of G and C are highest at the more functionally constrained sites, like rRNA-coding regions and first and second codon positions, implying that the GC bias is driven by natural selection; complicating this interpretation, however, is the fact that many of the cytosine residues within the Selaginella mtDNA and ptDNA are post-transcriptionally edited to uracil –. Unraveling the mechanism responsible for the GC enrichment of organelle genomes might in turn help explain the near-ubiquity of AT-rich mitochondrial and plastid DNAs throughout the eukaryotic domain, a feature that is poorly understood but probably linked to AT mutation pressure , .
Coccomyxa sp. C-169 (formerly referred to as Chlorella sp. C-169; hereafter Coccomyxa) is a unicellular, free-living trebouxiophyte  whose nuclear genome is being sequenced by the United States Department of Energy Joint Genome Institute (DOE JGI). Our cursory scan of the Coccomyxa sequencing reads revealed elevated levels of G and C in what appeared to be mtDNA and ptDNA sequences, suggesting that complete assemblies of the Coccomyxa mitochondrial and plastid genomes may reveal a novel set of GC-rich organelle DNAs.
Moreover, there is currently a lack of trebouxiophyte organelle genome data, which limits the utility of the available green algal organelle genomes for phylogenetic analyses. This is significant because, although there is strong support for the Prasinophyceae forming a paraphyletic assemblage at the base of the Chlorophyta, the branching order of the Ulvophyceae, Trebouxiophyceae, and Chlorophyceae (UTC) remains unresolved and controversial , . This is likely due to poor sampling and a lack of molecular sequence data from these groups, but it also reflects their antiquity, morphological diversity, and the potentially short timeframe over which they diverged from each other –. Depending on the ultrastructural characteristics and molecular data being used, all three possible branching orders of the UTC classes have been hypothesized –. Plastid genome multi-gene phylogenies have generally supported a sister relationship between the Ulvophyceae and Trebouxiophyceae , , whereas mtDNA phylogenies have tended to group the ulvophytes close to the chlorophyceans , . These analyses, however, have been hindered because the only complete sets of trebouxiophyte organelle genomes come from Pedinomonas minor ,  — a species that some have argued is not a trebouxiophyte at all, and whose mitochondrial genome shows high rates of evolution , ,  — and the nonphotosynthetic parasite Helicosporidium sp. ATCC 50920 ,  (hereafter Helicosporidium), whose plastid genome contains no genes for proteins that function in photosynthesis and is therefore poorly suited for multi-gene ptDNA phylogenies.
Here, we present the mitochondrial and plastid genome sequences from Coccomyxa. We investigate the architectures and nucleotide landscapes of these genomes as well as their potential for resolving the relationships among the crown chlorophytes.
Results and Discussion
Architecture of the Coccomyxa organelle genomes
The Coccomyxa mitochondrial and plastid genome sequences (GenBank under accession numbers HQ874522 and HQ693844, respectively) are both circular-mapping ,  (Figure 1) and, as predicted, are rich in G and C. These are the third set of trebouxiophyte mtDNA and ptDNA sequences described thus far; the other two, as mentioned above, come from Helicosporidium and P. minor , , , , and there is also complete sequence data for the ptDNAs of the trebouxiophytes Chlorella vulgaris, Parachlorella kessleri, and Leptosira terrestris ,  and for the mtDNA of the nonphotosynthetic parasite Prototheca wickerhamii . General features of the sequenced trebouxiophyte organelle genomes, including those from Coccomyxa, are shown in Table 1.
The ptDNA and mtDNA sequences have respective lengths of 175.7 and 65.4 kb, and are deposited in GenBank under accession numbers HQ693844 and HQ874522. Arrows within the coding regions denote transcriptional polarities. Transfer RNA-coding regions are designated by the single-letter abbreviation of the amino acid they specify; their anticodon sequences are shown in brackets.
With respective sizes of 65 and 175 kilobases (kb) and noncoding DNA contents of 52 and 56%, the Coccomyxa mitochondrial and plastid genomes are among the most inflated organelle DNAs observed from the Trebouxiophyceae, and are more akin to the prodigious organelles genomes often found within the Chlorophyceae and Ulvophyceae. Apart from the ptDNA of L. terrestris, which is 195 kb and 55% noncoding, most of the trebouxiophyte organelle genomes studied heretofore are relatively condensed (Table 1), as exemplified by the P. minor mtDNA, which is 25 kb and contains only 22 genes, and the Helicosporidium ptDNA, which is 37.4 kb and 95% coding, making it the most streamlined plastid genome observed from the Viridiplantae.
Annotation of the Coccomyxa organelle genomes revealed 59 (mtDNA) and 115 (ptDNA) putative genes, which are listed, along with the genes from the other sequenced trebouxiophyte organelle DNAs, in Supplementary Tables S1 and S2. Like other available trebouxiophyte organelle DNAs, save for the P. minor mtDNA, the Coccomyxa mitochondrial and plastid genomes use the standard genetic code. The mitochondrial gene content of Coccomyxa differs from those of P. wickerhamii and Helicosporidium in that it does not have rpl6, rps11, and trnTugu. We identified intact copies of rpl6 and rps11 within the Coccomyxa nuclear genome sequence (v2) on scaffolds 12 and 36, respectively, but did not find an mtDNA-like copy of trnTugu, leaving open the possibility that a nuclear-encoded cytosolic trnTugu is getting imported into the mitochondrion and compensating for the lost mitochondrial version of the gene. The ptDNA from Coccomyxa contains the same number of genes as that of C. vulgaris, which is about 3 to 10 more genes (mostly tRNAs) than those from other photosynthetic trebouxiophytes and around 50 more genes than the Helicosporidium plastid genome, which does not encode any photosynthetic proteins (Table 1; Supplementary Table S2).
Although replete with noncoding sequence, the Coccomyxa organelle genomes harbour few introns. We identified one group I intron in the ptDNA, found in psbB, and five introns in the mtDNA: a group I intron in rrnL and four group II introns, distributed among four tRNA-coding genes (trnHgug, trnSgcu, trnSuga, and trnWcca). Group II introns within tRNA-coding genes have been found in mtDNAs of other green algae, including those of Chlorokybus atmophyticus, Chaetosphaeridium globosum, and C. vulgaris. Three of tRNA group II introns in the Coccomyxa mtDNA (trnHgug, trnSgcu, trnSuga) are orthologous with those of C. atmophyticus. Low organelle intron contents are a common theme in trebouxiophytes and prasinophytes, whereas the organelle DNAs of chlorophyceans and ulvophytes are often intron dense.
Many of the intergenic regions in the Coccomyxa organelle genomes contain repetitive elements. Approximately 7% of both the mtDNA and ptDNA comprises repeats, most of which have forward (i.e., direct) or inverted (i.e., palindromic) orientations, are GC rich (60–70%), and are 20–250 nt in length (average 100 nt). Similar (even identical) repeats were found in both the mitochondrial and plastid compartments. In total, the Coccomyxa organelle genomes share ∼500 nt of repeat sequence with one another, distributed over approximately five different sites in the mtDNA and eleven in the ptDNA. To the best of our knowledge, this is one of only a few examples from all eukaryotes of a species harbouring the same repetitive element in both its mtDNA and ptDNA. GC-rich repeat elements have been identified in other organelle genomes, including the mtDNAs of the green algae P. capuana and Chlorogonium elongatum , . The repeats within the Coccomyxa organelle DNAs, however, do not show any obvious sequence similarity to those from other organelle DNAs.
The plastid genome of Coccomyxa, like those of C. vulgaris, L. terrestris, and Helicosporidium, does not have a “quadripartite structure”, meaning that it lacks inverted repeats. The only sequenced trebouxiophyte ptDNAs that have inverted repeats are those of Parachlorella kessleri and Pedinomonas minor (Table 1), implying that loss of the inverted repeat has occurred multiple times throughout the evolution of trebouxiophytes .
Organelle DNA nucleotide landscape
The Coccomyxa organelle genomes have overall GC contents of 53.2% (mtDNA) and 50.7% (ptDNA). Although only modestly biased in G and C, the nucleotide compositions of these two genomes are made more impressive by the AT bias found in nearly all other available mtDNA and ptDNA sequences (Figure 2). Indeed, Coccomyxa belongs to a small but eclectic list of species that are known to have organelle DNAs that exceed 50% GC (Table 2).
Species with high organelle GC contents are labeled (see Table 2 for full genus names). Organelle genome sequences were downloaded from the GenBank Reference Sequence collection on 1 March 2011.
The individual strands of the Coccomyxa organelle genomes contain almost equal proportions of G versus C (the GC skew of the mtDNA is 0.015 and that of the ptDNA is 0.008), a trait that is shared by other GC-rich organelle DNAs, except for the mtDNA of the wolf herring Chirocentrus dorab, which contains an abundance of cytosine on the main sense strand. Analyses of Coccomyxa EST sequence data revealed no evidence of post-transcriptional editing in either the mitochondrial or plastid compartments (∼10 organelle transcripts were explored), contrasting the massive levels of C-to-U editing found in the GC-biased mtDNA and ptDNA of the spikemoss Selaginella moellendorffii –. Selaginella organelle genomes are currently the only examples where GC richness is allied with post-transcriptional editing (Table 2).
The GC composition differs among the various regions of the Coccomyxa organelle genomes (Table 2). For the mtDNA, the GC contents of intergenic regions (55.7%) and third codon positions (59.8%) are higher than those of the more functionally constrained regions, such as first and second codon positions (51.4 and 40.8%) and sites coding for structural RNAs (52.5%). In the plastid genome the GC content at first codon positions (56.1%) exceeds that of intergenic (51%) and third codon positions (50.1%), and is lowest at second codon positions (43.6%). An inflated GC content at first codon positions relative to the other genomic regions is a general trend among plastid genomes (reflecting, in part, an amino acid composition rich in leucine) and is not unique to Coccomyxa (see  for a compilation of the codon position GC contents for various completely sequenced ptDNAs). For both the mtDNA and ptDNA, the highest levels of G and C are observed for the repetitive elements, most of which are between 60–75% GC.
The deduced amino acid sequences of the Coccomyxa mtDNA and ptDNA protein-coding genes, although slightly enriched in alanine and glycine, are not markedly different from their counterparts in AT-rich trebouxiophyte organelle DNAs (Supplementary Tables S3 and S4). It is therefore unlikely that the GC contents of the Coccomyxa organelle genomes are driven by selection for a particular amino acid composition. Moreover, the tRNA repertoire and anticodon suite encoded within the organelle DNAs of Coccomyxa are almost identical to those from other AT-rich trebouxiophyte organelle genomes (Supplementary Tables S1 and S2). Adaptation for DNA thermo stability (resulting from the extra hydrogen bond in G•C pairs) can also be ruled out as Coccomyxa sp. 169 originates from Marble Point Antarctica . So then what is the nature of the forces biasing the Coccomyxa organelle DNAs in G and C? When addressing this question, it may help to reflect on why most organelle genomes are enriched in A and T.
The forces shaping organelle genome nucleotide landscape are poorly understood and probably differ both among and within lineages. Some have invoked selection for translational efficiency to explain the high levels of A and T in organelle genomes (reviewed in ), whereas others contend that it stems from AT mutation pressure coupled with inefficient mitochondrial and plastid DNA repair processes –. Indeed, mitochondrial and plastid genomes, because they are housed in energy-producing organelles, are often exposed to reactive oxygen species (ROS), which, by causing the deamination of cytosine to uracil and the oxidative conversion of guanine to 8-oxo-guanine, promote C∶G→T∶A transitions and C∶G→A∶T transversions . That being said, the plastid genomes from nonphotosynthetic species, like Helicosporidium and Plasmodium falciparum, which are not currently exposed to high levels of ROS, are also AT rich, suggesting that there are a diversity of forces biasing organelle genomes towards A and T. Whatever these forces may be, Coccomyxa appears to be buffering or counteracting them. The following observations give insight into how it may be achieving this.
Within the Coccomyxa mtDNA and ptDNA there are, with some exceptions, an excess of guanines and cytosines at silent sites relative to the more functionally constrained positions (Table 2). This distribution of G and C is best explained by the negative selection principle of the neutral theory of molecular evolution . In other words, nonadaptive forces may be driving the nucleotide composition towards G and C. The two neutral processes that are thought to influence nucleotide landscape are biased mutation pressure, which, as discussed above, seems to be skewed towards A and T in mitochondria and plastids, and biased gene conversion, which favours G and C in nuclear genomes ,  but appears to be AT-biased in the sole organelle system in which it has been studied (the ptDNA of tobacco) . Although speculative, one possibility is that the inflated GC content of the Coccomyxa organelle DNA repeat elements — sequences that presumably undergo high levels of recombination and therefore experience a lot of gene conversion  — may be a sign of a GC-biased conversion process within the mitochondrion and plastid of this species; however, this would contrast the AT biased conversion process found in the tobacco plastid genome . Also, the presence of identical GC-rich repeats in both the mtDNA and ptDNA of Coccomyxa implies that there is a mobile nature to at least some of these elements, meaning that they may be able to propagate their GC richness by spreading their sequence throughout the various noncoding regions.
Another possibility is that a mutation conferring a GC-biased mutational pattern has been fixed in Coccomyxa. Given that both the mtDNA and ptDNA are GC rich, one would expect that this new mutation bias be derived from a system that affected both organelles. The mitochondrial and plastid compartments share comparatively few proteins with one another, but nucleus-encoded dual-targeted organelle proteins have been characterized in a number of lineages, and are the rule rather than the exception in some systems , . The similar GC bias of the mtDNA and ptDNA may imply that cell-wide features, like life history characteristics, environment, and/or metabolic processes, have influenced the mitochondrial and plastid nucleotide compositions. In this context, it is noteworthy that the overall GC content of Coccomyxa nuclear genome (∼53%) is comparable to that of the organelle DNAs.
Organelle DNA sequence data from other Coccomyxa species are scarce, but we were able to find in GenBank partial rbcL ptDNA sequences from 49 different Coccomyxa strains. Forty-seven of these strains have relatively GC-poor rbcL sequences (average = 40% GC), suggesting that their plastid genomes are AT rich. However, the rbcL GC contents from two strains — Coccomyxa chodatii (49%) and Coccomyxa rayssiae (47.3%) — are similar to that of Coccomyxa sp. C-169 (49.3%), the focus of this study. All three of these taxa are non-lichenized, free living species, whereas the other 47 species, which had typical AT-rich rbcL GC contents, are symbionts.
Sparse taxon sampling and a lack of DNA sequence data have made it difficult to resolve the phylogenetic relationships among the Ulvophyceae, Trebouxiophyceae, and Chlorophyceae. The completion of the Coccomyxa mitochondrial and plastid genomes allows for a more thorough phylogenetic evaluation of these groups. Using two amino acid sequence datasets — one with 29 mtDNA-encoded proteins and another containing 67 ptDNA-encoded proteins — we performed Maximum Likelihood (LG+Γ model) and Bayesian phylogenetic (CAT+GTR+Γ model) analyses on Coccomyxa and various other green algae for which complete or almost complete organelle DNA sequences are available. Helicosporidium was not included in the phylogenetic analyses because it lacked several important organelle genes for phylogenetic reconstruction; when we included Helicosporidium (which groups with species from the Chlorellaceae) in our analyses it did not alter the tree topologies discussed below.
Both the mitochondrial and plastid phylogenies (Figures 3 and 4, respectively) consistently grouped Coccomyxa with other trebouxiophytes. On the mitochondrial trees, Coccomyxa was sister to a clade containing Helicosporidium and P. wickerhamii, but the taxon sampling remains relatively poor, and elevated rates of evolution for P. minor and Scenedesmus obliquus prevented resolution of the UTC classes. The grouping of P. minor with Scenedesmus obliquus received high bootstrap support (95%) in the ML tree (Figure 3A), as it did in previous mtDNA-based green algal phylogenies , , but given that both of these species show high rates of mtDNA substitution, we argue that this arrangement is caused by long-branch attraction (LBA), which also appears to be attracting the relatively fast evolving ulvophyte P. akinetum, resulting in a paraphyletic Ulvophyceae (Figure 3A). Our concerns regarding LBA were confirmed by Bayesian inference with the site-heterogeneous CAT model [combined with a general time reversible (GTR) exchange rate matrix and gamma correction], which has been shown to handle saturated positions more efficiently than site-homogeneous models such as the LG matrix . In the Bayesian tree (Figure 3B), a monophyletic Ulvophyceae clade was recovered, with maximal statistical support, and the position of P. minor and S. obliquus were unresolved.
Phylogenetic tree of green algae, with 6 streptophyte species used as outgroup. Thick branches represent maximal statistical support (100% bootstrap support in A; 1.0 Bayesian posterior probabilities in B). When not maximal, values are indicated. The scale bar represents the estimated number of amino acid substitutions per site.
Phylogenetic tree of green algae, with 5 streptophyte species used as outgroup. Thick branches represent maximal statistical support (1.0 Bayesian posterior probabilities). When not maximal, values are indicated. The scale bar represents the estimated number of amino acid substitutions per site.
The plastid dataset contained both a larger diversity of chlorophyte species and more than twice the number of genes than the mitochondrial dataset. Figure 4 corresponds to the Bayesian tree inferred from the site-heterogeneous CAT+GTR+Γ model. On this tree the trebouxiophytes formed a monophyletic group comprising two subclades: one containing P. minor and members of the Chlorellaceae and another harbouring Coccomyxa, O. solitaria, and L. terrestris. Previous plastid multi-gene phylogenies positioned L. terrestris on a long branch next to the Chlorophyceae , , but our analyses place it with the trebouxiophytes and indeed sister to Coccomyxa with high support, which is consistent with its traditional classification . The plastid tree also recovered a weakly supported sister relationship between the Ulvophyceae and Chlorophyceae, with the Trebouxiophyceae branching at the base of the chlorophyte crown. Although the relationships among the UTC groups were not burdened by dramatically long branches, it is noticeable that the Chlorophyceae in general and the ulvophyte Bryopsis hypnoides evolve faster than the rest of the UTC members. In an attempt to test for an undetected misleading effect resulting from this heterogeneous rate of sequence evolution that would lower the support for the overall UTC branching pattern, we reconstructed a tree where six fast-evolving lineages were discarded: P. provasolii, B. hypnoides and all Chlorophyceae except S. obliquus (Supplementary Figure S1). Using this reduced dataset, the statistical support for the Ulvophyceae and Chlorophyceae grouping increased significantly (0.73 PP to 0.99 PP); the monophyly of Trebouxiophyceae, however, remained poor. This branching order is consistent with previous phylogenetic analyses using 18S rRNA ,  and plastid genome architectural features  and reinforces the belief that the ancestor of the UTC clade had a phycoplast and a non-persistent mitotic spindle  — features that were presumably lost in the Ulvophyceae. An early diverging Trebouxiophyceae also implies that a counter-clockwise flagellar orientation was an ancestral characteristic of the UTC, which evolved to directly opposed and clockwise orientations within the Chlorophyceae. This hypothesis is consistent with the counter clockwise flagellar root system in the prasinophyte class Chlorodendrophyceae .
Materials and Methods
Coccomyxa sp. C-169 (formerly referred to as Chlorella sp. C-169 and then redefined after phylogenetic analysis of the draft nuclear genome sequence [DOE JGI, unpublished data]) was maintained at the culture collection of the Institute of Molecular and Cellular Biosciences (IAM), University of Tokyo, until 2007. It is now available from the Microbial Culture Collection at the National Institute for Environmental Studies (NIES) under strain number 2166. Coccomyxa C-169/NIES-2166 originates from Marble Point Antarctica.
Assembly and annotation of the Coccomyxa organelle genomes
The Coccomyxa organelle genomes were sequenced as a part of the DOE JGI Coccomyxa genome project using whole-genome shotgun sequencing on a Sanger platform. Reads were data-mined from the National Center for Biotechnology Information (NCBI) Coccomyxa Trace Archive using the following sequences as BlastN (v2.2.25+) queries: the Helicosporidium and P. wickerhamii mitochondrial genomes and the C. vulgaris ptDNA. BlastN parameters were as follows: an expectation value (E-value) of 0.0001; a word size of 11; match and mismatch scores of 2 and −3, respectively; and gap-cost values of 5 (existence) and 2 (extension). Trace files corresponding to Blast hits were assembled with CodonCode Aligner v184.108.40.206 (CodonCode Corporation, Dedham, MA, USA). Gaps in the assemblies were filled by blasting the generated Coccomyxa mtDNA and ptDNA contigs against the Coccomyxa Trace Archive and then using the hits to extend the contigs. The entire Coccomyxa organelle genome sequences were blasted against the Coccomyxa draft nuclear genome sequence (v2) to verify that they were not generated from either nuclear mitochondrial DNAs or nuclear plastid DNAs. These analyses indicate that the mitochondrial and plastid genome assemblies were generated from organelle DNA and are consistent with the view that Coccomyxa harbours very little organelle-like sequences in its nuclear genome .
Introns and repeat elements were respectively detected with RNAweasel  and REPuter . Additional scans for repeats were performed by building custom BLAST databanks of the Coccomyxa mtDNA and ptDNA sequences and then blasting (BlastN v2.2.25+) these with specific regions from the mitochondrial and plastid genomes. Coccomyxa EST and nuclear genome sequence data (v2) were downloaded from the DOE JGI Coccomyxa genome portal.
The phylogenetic datasets included the concatenated alignments of 29 mtDNA-encoded genes (5,344 amino acids) from 15 taxa (mitochondrial dataset) and 67 ptDNA-encoded genes (1,1218 amino acids) from 25 taxa (plastid dataset); see Supplementary Table S5 for a list of species and genes that were employed. Sequences were aligned using the L-INS-I method of the MAFFT package . Poorly aligned positions were removed with Gblocks  using the following settings: no gaps allowed, minimum number of sequences for conserved and flank positions equal to 50% of the number of taxa plus one, a maximum number of contiguous non-conserved positions of 10, and a minimum block length of 8. Chlamydomonas reinhardtii was discarded from the mitochondrial dataset because it showed high rates of mtDNA evolution and has a gene-poor mitochondrial genome — a more slowly evolving representative for the Chlorophyceae was available (i.e., S. obliquus). The Helicosporidium ptDNA was not included in the plastid dataset because it is highly reduced and thus lacks several genes that were important for phylogenetic reconstruction. The final gene datasets were based on the presence of at least 70% of the species that were to be included in the final concatenations (Supplementary Table S5).
Maximum Likelihood (ML) phylogenetic analyses were run with RAxML 7.2.8  in combination with the rapid hill-climbing algorithm and the site-homogeneous LG+Γ+F model of evolution (−m PROTGAMMALGF, 4 discrete rate categories). The best-scoring ML tree was based on multiple searches of 20 randomized stepwise addition parsimony starting trees. Statistical support was evaluated with non-parametric bootstrapping using 100 replicates. Bayesian phylogenetic analyses were performed with PhyloBayes 3.2f  using the site-heterogeneous mixture CAT model with a general time reversible (GTR) exchange rate matrix and gamma correction (4 categories). Two independent Markov chains with a total length of 10,000 cycles were performed, discarding the first 2000 points as burnin, and calculating the posterior consensus on the remaining trees. Convergence between the two chains was ascertained by examining the difference in frequency for all of their bipartitions (<0.05 in all analyses).
For the fast-evolving species removal analysis on the plastid dataset, root-to-tip distances were calculated with the “ape” and “geiger” packages available in R, with a root position defined on the branch leading to the monophyletic clade of Marchantia polymorpha and Physcomitrella patens (within the outgroup Streptophyta). Taxa were sorted according to their distance to the root, and the fastest evolving taxa removed (Pycnococcus provasolii, Bryopsis hypnoides, and all of the Chlorophyceae except the relatively slowly evolving S. obliquus).
Bayesian phylogenetic tree of the ptDNA data with fast-evolving species removed
Gene contents of available trebouxiophyte mitochondrial genomes
Gene contents of available trebouxiophyte plastid genomes
Amino acid composition of proteins encoded in complete mitochondrial genome sequences from trebouxiophytes
Amino acid composition of proteins encoded in complete plastid genome sequences from trebouxiophytes
Gene and species lists for the mitochondrial and chloroplast datasets
Conceived and designed the experiments: DRS PJK. Performed the experiments: JG IVG TY JLVE. Analyzed the data: DRS FB JG. Contributed reagents/materials/analysis tools: JG IVG TY JVE. Wrote the paper: DRS. Helped interpret the data and revise the manuscript: DRS FB PJK JLVE.
- 1. Lewis LA, McCourt RM (2004) Green algae and the origin of land plants. Am J Bot 91: 1535–1556.
- 2. Smith DR, Lee RW (2010) Low nucleotide diversity for the expanded organelle and nuclear genomes of Volvox carteri supports the mutational-hazard hypothesis. Mol Biol Evol 27: 2244–2256.
- 3. Smith DR, Lee RW (2008) Mitochondrial genome of the colorless green alga Polytomella capuana: a linear molecule with an unprecedented GC content. Mol Biol Evol 25: 487–496.
- 4. Borza T, Redmond EK, Laflamme M, Lee RW (2009) Mitochondrial DNA in the Oogamochlamys clade (Chlorophyceae): high GC content and unique genome architecture for green algae. J Phycol 45: 1323–1334.
- 5. Fricova D, Valach M, Farkas Z, Pfeiffer I, Kucsera J, et al. (2010) The mitochondrial genome of the pathogenic yeast Candida subhashii: GC-rich linear DNA with a protein covalently attached to the 5′ termini. Microbiology 156: 2153–2163.
- 6. Clare EL, Kerr KCR, von Königslöw TE, Wilson JJ, Hebert PDN (2008) Diagnosing mitochondrial DNA diversity: applications of a sentinel gene approach. J Mol Evol 66: 362–367.
- 7. Smith DR (2009) Unparalleled GC content in the plastid DNA of Selaginella. Plant Mol Biol 71: 627–639.
- 8. Hecht J, Grewe F, Knoop V (2011) Extreme RNA editing in coding islands and abundant microsatellites in repeat sequences of Selaginella moellendorffii mitochondria: the root of frequent plant mtDNA recombination in early tracheophytes. Genome Biol Evol 3: 344–358.
- 9. Tsuji S, Ueda K, Nishiyama T, Hasebe M, Yoshikawa S, et al. (2007) The chloroplast genome from a lycophyte (microphyllophyte), Selaginella uncinata, has a unique inversion, transpositions and many gene losses. J Plant Res 120: 281–290.
- 10. Burger G, Lang BF (2003) Parallels in genome evolution in mitochondria and bacterial symbionts. IUBMB Life 55: 205–212.
- 11. Kusumi J, Tachida H (2005) Compositional properties of green-plant plastid genomes. J Mol Evol 60: 417–425.
- 12. Karsten U, Friedl T, Schumann R, Hoyer K, Lembcke S (2005) Mycosporine-like amino acids and phylogenies in green algae: Prasiola and its relatives from the Trebouxiophyceae (Chlorophyta). J Phycol 41: 557–566.
- 13. Pröschold T, Leliaert F (2007) Systematics of the green algae: conflict of classic and modern approaches. In: Brodie J, Lewis J, editors. Unravelling the algae: the past, present, and future of algal systematics. London (UK): Taylor and Francis. pp. 123–53.
- 14. O'Kelly CJ (2007) The origin and early evolution of green plants. In: Falkowski PG, Knoll AH, editors. Evolution of primary producers in the sea. Burlington (MA): Elsevier Academic Press. pp. 287–309.
- 15. Cocquyt E, Verbruggen H, Leliaert F, De Clerck O (2010) Evolution and cytological diversification of the green seaweeds (Ulvophyceae). Mol Biol Evol 27: 2052–2061.
- 16. Pombert JF, Otis C, Lemieux C, Turmel M (2005) The chloroplast genome sequence of the green alga Pseudendoclonium akinetum (Ulvophyceae) reveals unusual structural features and new insights into the branching order of chlorophyte lineages. Mol Biol Evol 22: 1903–1918.
- 17. De Wever A, Leliaert F, Verleyen E, Vanormelingen P, Van der Gucht K, et al. (2009) Hidden levels of phylodiversity in Antarctic green algae: further evidence for the existence of glacial refugia. Proc R Soc B Biol Sci 276: 3591–3599.
- 18. Turmel M, Otis C, Lemieux C (2009) The chloroplast genomes of the green algae Pedinomonas minor, Parachlorella kessleri, and Oocystis solitatia reveal a shared ancestry between the Pedinomonadales and Chlorellales. Mol Biol Evol 26: 2317–2331.
- 19. Pombert JF, Otis C, Lemieux C, Turmel M (2004) The complete mitochondrial DNA sequence of the green alga Pseudendoclonium akinetum (Ulvophyceae) highlights distinctive evolutionary trends in the Chlorophyta and suggests a sister-group relationship between the Ulvophyceae and Chlorophyceae. Mol Biol Evol 21: 922–935.
- 20. Pombert JF, Keeling PJ (2010) The mitochondrial genome of the entomoparasitic green alga Helicosporidium. PLoS ONE 5: e8954.
- 21. Turmel M, Lemieux C, Burger G, Lang BF, Otis C, et al. (1999) The complete mitochondrial DNA sequences of Nephroselmis olivacea and Pedinomonas minor. Two radically different evolutionary patterns within green algae. Plant Cell 11: 1717–1730.
- 22. Melkonian M (1990) Chlorophyte orders of uncertain affinities. Order Pedinomonadales. In: Margulis L, Corliss JO, Melkonian M, Chapman DJ, editors. Handbook of protoctista. The structure, cultivation, habitats and life histories of the Eukaryotic microorganisms and their descendants exclusive of animals, plants and fungi. Boston (MA): Jones and Bartlett Publishers. pp. 649–651.
- 23. Moestrup Ø (1991) Further studies of presumedly primitive green algae, including the description of Pedinophyceae class. nov. and Resultor gen. nov. J Phycol 27: 119–133.
- 24. de Koning AP, Keeling PJ (2006) The complete plastid genome sequence of the parasitic green alga Helicosporidium sp. is highly reduced and structured. BMC Biol 4: 12.
- 25. Bendich AJ (2004) Circular chloroplast chromosomes: the grand illusion. Plant Cell 16: 1661–1666.
- 26. Bendich AJ (2007) The size and form of chromosomes are constant in the nucleus, but highly variable in bacteria, mitochondria and chloroplasts. Bioessays 29: 474–483.
- 27. de Cambiaire JC, Otis C, Turmel M, Lemieux C (2007) The chloroplast genome sequence of the green alga Leptosira terrestris: multiple losses of the inverted repeat and extensive genome rearrangements within the Trebouxiophyceae. BMC Genomics 8: 213.
- 28. Wolff G, Plante I, Lang BF, Kück U, Burger G (1994) Complete sequence of the mitochondrial DNA of the chlorophyte alga Prototheca wickerhamii. Gene content and genome organization. J Mol Biol 237: 75–86.
- 29. Kroymann J, Zetsche K (1998) The mitochondrial genome of Chlorogonium elongatum inferred from the complete sequence. J Mol Evol 47: 431–440.
- 30. Holm-Hasen O (1964) Isolation and culture of terrestrial and fresh-water algae of Antarctica. Phycologia 4: 43–51.
- 31. Lynch (2007) The origins of genome architecture. Sunderland (MA): Sinauer Associates.
- 32. Kang D, Hamasaki N (2002) Maintenance of mitochondrial DNA integrity: repair and degradation. Curr Genet 41: 311–322.
- 33. Mason PA, Matheson EC, Hall AG, Lightowlers RN (2003) Mismatch repair activity in mammalian mitochondria. Nucleic Acids Res 31: 1052–1058.
- 34. Kimura M (1983) The neutral theory of molecular evolution. Cambridge (UK): Cambridge University Press.
- 35. Marais G (2003) Biased gene conversion: implications for genome and sex evolution. Trends Genet 19: 330–338.
- 36. Duret L, Galtier N (2009) Biased gene conversion and the evolution of mammalian genomic landscapes. Annu Rev Genomics Hum Genet 10: 285–311.
- 37. Khakhlova O, Bock R (2006) Elimination of deleterious mutations in plastid genomes by gene conversion. Plant J 46: 85–94.
- 38. Spencer CC (2006) Human polymorphism around recombination hotspots. Biochem Soc Trans 34: 535–536.
- 39. Millar A, Whelan J, Small I (2006) Recent surprises in protein targeting to mitochondria and plastids. Curr Opin Plant Biol 9: 610–615.
- 40. Carrie C, Giraud E, Whelan J (2009) Protein transport in organelles: Dual targeting of proteins to mitochondria and chloroplasts. FEBS J 276: 1187–1195.
- 41. Lartillot N, Philippe H (2008) Improvement of molecular phylogenetic inference and the phylogeny of Bilateria. Philos T Roy Soc B 363: 1463–1472.
- 42. Lü F, Xü W, Tian C, Wang G, Niu J, et al. (2011) The Bryopsis hypnoides Plastid Genome: Multimeric Forms and Complete Nucleotide Sequence. PLoS ONE 6: e14663.
- 43. Lockhorst GM, Rongen GPJ (1994) Comparative ultrastructural studies of division processes in terrestrial green alga Leptosira erumpens (Deason & Bold) Lukesova confirm the ordinal status of the Pleurastrales. Cryptogamic Botany 4: 394–409.
- 44. Watanabe S, Nakayama T (2007) Ultrastructure and phylogenetic relationships of the unicellular green algae Ignatius tetrasporus and Pseudocharacium americanum (Chlorophyta). Phycol Res 55: 1–16.
- 45. Mattox KR, Stewart KD (1984) Classification of the green algae: a concept based on comparative cytology. In: Irvine DEG, John DM, editors. Systematics of green algae. London (UK): Academic Press. pp. 29–72.
- 46. Smith DR, Crosby K, Lee RW (2011) Correlation between nuclear plastid DNA abundance and plastid number supports the limited transfer window hypothesis. Genome Biol Evol 3: 365–371.
- 47. Lang BF, Laforest MJ, Burger G (2007) Mitochondrial introns: a critical view. Trends Genet 23: 119–125.
- 48. Kurtz S, Schleiermacher C (1999) REPuter: fast computation of maximal repeats in complete genomes. Bioinformatics 15: 426–427.
- 49. Katoh K, Kuma K, Toh H, Miyata T (2005) MAFFT version 5: improvement in accuracy of multiple sequence alignment. Nucleic Acids Res 33: 511–518.
- 50. Castresana J (2000) Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol Biol Evol 17: 540–552.
- 51. Stamatakis A (2006) RAxML-VI-HPC: Maximum Likelihood-based Phylogenetic Analyses with Thousands of Taxa and Mixed Models. Bioinformatics 22: 2688–2690.
- 52. Lartillot N, Lepage T, Blanquart S (2009) PhyloBayes 3: a Bayesian software package for phylogenetic reconstruction and molecular dating. Bioinformatics 25: 2286–2288.