Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

I. Introduction

Algae comprise a large number of most diverse unicellular and multi-cellular taxa that populate virtually all ecosystems on Earth. The best-known among the 20 or so algal groups are the green, red, brown, and golden algae, dinoflagellates, foraminifera, and euglenids. Algae are defined by exclusion – as plastid-containing eukaryotes outside plants or, what comes out to the same result, as all plastid-containing protists (protists being eukaryotes outside animals, fungi, and plants).

This chapter is subdivided into sections based on phylogeny. We will refer to generally accepted eukaryotic clades (Fig. 6.1) and leave proposed supergroups (Keeling et al. 2005) for critical review in other chapters of this book (see Chaps. 2, 4). One of these undisputed clades is Viridiplantae (green algae and land plants; aka ‘green plants’), yet a matter of debate remains the exact relationship of ‘greens’ to red and glaucophyte algae (for references see legend of Fig. 6.1). Another solid monophyletic assemblage unites alveolates (including dinoflagellates), stramenopiles (including golden and brown algae), and chlorarachniophytes. Haptophytes and cryptophytes, however, earlier thought to affiliate with alveolates and stramenopiles (to form the ‘chromalveolates’), cannot be placed reliably in the eukaryotic tree. A third well established eukaryotic clade is Euglenozoa plus ­heteroloboseans and jakobids. Opisthokonts ­(animals, fungi, and related protists) form a fourth coherent group, but as it lacks photosynthetic members, it will not be further discussed in this chapter.

Fig. 6.1.
figure 1

Schematic tree of the major eukaryotic groups. Shading circumscribes groups belonging to the taxon indicated in large font size. Black contiguous lines connect monophyletic groups, notably Viridiplantae; stramenopiles + alveolates + chlorarachniophytes; opisthokonts (e.g. Burki et al. 2007; Hackett et al. 2007; Rodriguez-Ezpeleta et al. 2007a), and jakobids + Euglenozoa + Heterolobosea (Rodriguez-Ezpeleta et al. 2007a). Other relevant references in this context are (Cavalier-Smith 1999; Rodriguez-Ezpeleta et al. 2005; Baurain et al. 2010). Dotted lines represent uncertain topologies. Boxed taxon names highlight photosynthetic groups.

Full size image

Eukaryotes acquired plastids long after the establishment of mitochondria. The first round of acquisition, by enslavement of a cyanobacterium, took place in the common ancestor of green, red, and glaucocystophyte algae (see Chap. 1). From here, plastids spread to the other eukaryotic lineages via secondary and higher-order symbiotic events, involving the engulfment of a photosynthetic eukaryote by a non-photosynthetic one. The origin of plastids in the various lineages can be quite convoluted as discussed in detail in Chap. 2. Mitochondria, in contrast, have apparently remained faithful to their host, and mitochondrial DNA (mtDNA) accurately reflects the evolutionary history of the nuclear genome as a whole. This view is corroborated by numerous independent phylogenetic reconstructions using either nuclear or mitochondrial genes (e.g. Baurain et al. 2010) yielding trees that are congruent in topology.

Complete mtDNAs have been determined at a large scale since the 1990s, and the body of data is growing at an accelerated pace with ever faster sequencing technologies. Over the years, reviews on mtDNAs have attempted to keep up with the growing genome data, some eukaryotic-wide (Gray et al. 1999), others with special attention to protists (Gray et al. 1998, 2004), plants (Kubo and Newton 2008), or animals (Boore 1999). Here, we will review currently available mitochondrial genome data from all plastid-containing eukaryotes except plants (embryophytes), whose mtDNAs will be discussed in Chaps. 8 and 9 of this book.

II. Mitochondrial Genome Structure and Gene Complement

We start with a short introduction on the structure of the mitochondrial genome and the type of genes contained in it, across all eukaryotes.

A. Structure of the Mitochondrial Genome

Generally, mitochondria contain a single type of chromosome (in multiple copies), as did most likely their alpha-proteobacterial ancestor. However, several exceptions to this rule are found in disparate eukaryotic lineages, with up to hundreds mitochondrial chromosomes (see discussion in (Burger et al. 2003a) and a recent review (Burger et al. 2011)). The shape of most mtDNAs is ‘circular-mapping’, that is a linear DNA molecule composed of head-to-tail concatenates that most likely arises by rolling-­circle DNA replication (Oldenburg and Bendich 2001; Ling and Shibata 2004). Less frequent are monomeric linear molecules and truly circular mitochondrial chromosomes that occur sporadically across the eukaryotic tree (Gray et al. 2004).

The size of mtDNAs varies drastically. The smallest ones are found in Plasmodium and relatives with 6 kbp only (Feagin et al. 1991) and the largest ones in the cucumber family measuring up to ∼3,000 kbp (Ward et al. 1981). At the time of writing, the largest fully sequenced mtDNA is that of Cucurbita pepo with nearly 1,000 kbp (Alverson et al. 2010).

B. Mitochondrion-Encoded Functions

Mitochondria perform numerous biological functions, most of which rely entirely on nucleus-encoded genes (for a review on the origin of imported mitochondrial proteins, see (Gray et al. 2001) and Chap. 1, Sect. II. D). Among the pathways and biological processes that involve at least some mtDNA-encoded genes, two are universal, notably electron transport plus oxidative phosphorylation, and mitochondrial translation (Table 6.1). MtDNA always encodes some or even all structural RNAs that are involved in translation, i.e. ribosomal RNAs (rRNAs) and transfer RNAs (tRNAs), but may not include ribosomal protein-coding genes at all. Additional processes specified by some mtDNAs include protein transport across the mitochondrial inner membrane via the twin-arginine translocase, and more rarely, the SecY-type transport system. Cytochrome c maturation, in particular heme import into the inner-membrane space and covalent heme linkage to apocytochrome c, is also often controlled by mitochondrial genes. Further, mtDNA-encoded components for tRNA processing and cytochrome oxidase assembly are found in several lineages. Mitochondrial genes for transcription occur in only a single group, the jakobids (for a review, see Gray et al. 2004). In general, mtDNAs adhere to ‘genetic conservatism’ in that they encode only a relatively narrow set of functions that are essential for biogenesis and activity of the organelle (Gray et al. 1999).

Table 6.1 Mitochondrion-encoded genes and their functionsa
Full size table

All of the mitochondrial processes listed above are directly derived from the bacterial ancestor of mitochondria (genes acquired by horizontal transfer are discussed below; see also Chap. 10). The range of functions encoded by mitochondrial genomes generally does not correlate well with the phylogenetic affinity of the organisms. Gene migration to the nucleus or loss for good has occurred numerous times and independently across the eukaryotic tree.

C. Gene Sets

Table 6.1 lists all mitochondrial genes and cognate products recognized to date. The large and small subunit rRNA genes (rnl, rns) are encoded on all known mtDNAs. In contrast, the mitochondrial 5S rRNA gene (rrn5) occurs rather sporadically. In several instances, it was overlooked in the original sequence annotation (see Tables 6.3, 6.5), and it might remain undetected in other cases, because it is small and often poorly conserved. For example, to discover this gene in amoebozoans mtDNAs, it required biochemical data combined with secondary structure analysis and comparative sequence data from related species (Bullerwell et al. 2010). Mitochondrion-encoded tRNAs are variable in number ranging from zero to nearly 30. Mitochondrial genes for the respiratory chain are found universally, notably cob specifying apo-cytochrome b of Complex III (ubiquinone-cytochrome c-oxidoreductase) and cox1 encoding subunit 1 of Complex IV (cytochrome c oxidoreductase, or briefly cytochrome oxidase). Most mtDNAs carry additional genes for Complex IV (cox2 and cox3). Subunits for Complex I (NADH-ubiquinone oxidoreductase) are mtDNA-encoded in most eukaryotes, the basic gene set consisting of nad1, 4, 5 and generally also nad2, 3, 6, while the nad7–11 genes occur less frequently. In some lineages, Complex I has been entirely lost and functionally substituted by a nucleus-encoded single-subunit enzyme (van Dooren et al. 2006). Complex V (ATP synthase) has typically several subunits specified by mitochondrial genes, notably atp6, 8, 9, and more rarely atp1 and atp4. Only a few lineages possess mitochondrial genes for Complex II (succinate-ubiquinone-oxidoreductase) such as sdh2–4, and it was only in the mid 1990s when mtDNAs carrying these genes were discovered for the first time (Burger et al. 1996).

The other major class of mitochondrial genes encodes proteins of the mito-ribosome. Up to 27 such genes can occur (jakobids), whereas several major lineages harbour only few or no such genes at all on their mtDNAs. Most frequent are genes for the small ribosomal subunit (rps3 etc.) with typically fewer genes for the large subunit (rpl2, etc.).

Genes that are rarely found in mtDNAs involve transmembrane transport and cytochrome c maturation (tatA, tatC, ccmA-C, F) and 5′ tRNA processing (rnpB, encoding the RNA subunit of RNase P). The most uncommon mitochondrial genes are rpoA-D, encoding RNA polymerase components, tufA, a translation elongation factor gene, secY that is involved in secY-dependent protein import, and ssrA specifying tmRNA that releases stalled ribosomes from non-stop mRNAs (Jacob et al. 2004). These latter seven genes are not found in mtDNAs of algae.

A small set of mitochondrial genes were probably gained through horizontal transfer. These include genes for phage-like DNA and RNA polymerases and reverse transcriptase (dpo, rpo, and rtl), as seen for instance in mtDNAs of red and brown algae but also in other eukaryotes. These genes likely spread via mobile genetic elements, sometimes integrating into mtDNA and sometimes establishing themselves as mitochondrial plasmids (Weber et al. 1995). Integration may trigger genome rearrangements or linearization. Plasmid versions of these genes have been studied particularly well in fungi (Griffiths 1995). Free-standing rtl genes may also be derived from group II intronic reading frames (see below). Despite contrary claims (Delaroque et al. 1996; Rousvoal et al. 1998), there is no evidence that these genes are involved in replication and transcription of mtDNA. Instead, dpo, rpo, and rtl sequences often accumulate frameshift and stop mutations (e.g. in Porphyra Burger et al. 1999) and are frequently fragmented, indicating that they are ‘on the way out’. For mitochondrial trans­cription, most eukaryotes have substituted the function of ancestral rpoA-D by a nuclear gene encoding a homolog of the single-polypeptide RNA polymerase of bacteriophages T3 and T7 (Cermakian et al. 1997). Another most likely horizontally transferred gene is dam (methyl transferase) that has been detected first in haptophytes (Sanchez Puerta et al. 2004). It is unknown whether it has a biological function.

ORFs (or unidentified open reading frames) are potential protein-coding genes and occur in most of the larger mtDNAs. Several ORFs turned out to be unrecognized, highly divergent versions of known genes (Table 6.1). For example, the former orfB is in fact atp8 (Gray et al. 1998), ymf39 is now known as atp4 (Burger et al. 2003c), murf1 is a highly derived nad2 in trypanosome mtDNA (Kannan and Burger 2008), and the ORFs in fungal mitochondria initially designated ‘5S’, var or ORF227, are extremely deviant rps3 homologs (Bullerwell et al. 2000). Function assignment of hypothetical (mitochondrial) proteins is indeed challenging, and progress in bioinformatics approaches (e.g. Kannan et al. 2008) will eventually uncover further incognito versions of typical mitochondrial protein genes. Alternatively, ORFs could be fortuitous, representing genome regions that happen to be conceptually translatable into proteins. This seems to apply to most of the hundred or so ORFs in the inflated mtDNAs of plants (see Chaps. 8, 9).

Introns in mitochondrial genes belong to group I or group II and occur in virtually all major eukaryotic lineages. These introns often harbour ORFs that participate in intron propagation and/or intron splicing (maturases). In the case of group I introns, ORFs encode homing endonucleases of the LAGLIDADG or GIY-YIG type, whereas group II intron ORFs usually encode phage-like reverse transcriptases (for a recent review, see Lang et al. 2007). Yet, there are a few exceptions where group II intron ORFs specify homing endonucleases (Toor and Zimmerly 2002).

The mtDNAs with the largest gene count are found among jakobids. Reclinomonas americana, for example, has 67 protein-encoding genes and an additional 30 genes for structural RNAs (Lang et al. 1997). This mtDNA contains virtually all genes represented on any other mtDNA, plus unique genes that also descend from the proteobacterial ancestor, such as tufA, rpoA-D, atp3 (Complex V subunit), cox11 (Complex IV assembly), secY and ssrA (Jacob et al. 2004). The recently sequenced mtDNA of the jakobid Andalucia contains in addition cox15 (Complex IV assembly) and one more ribosomal protein gene, rpl35 (BF Lang, pers. comm.), summing up to 31 mitochondrion-encoded ribo-protein genes and 69 assigned protein genes in total. These are 23 times more protein genes than in mtDNAs of certain chlorophycean algae. A comprehensive compilation of mtDNA gene content for sequences published prior to 2010 is available in the GOBASE database (O’Brien et al. 2009).

In the subsequent sections, we will review mtDNAs from the various algal groups in seven sections and the following order: (1) green algae (chlorophytes, streptophytes), (2) glaucophytes, (3) red algae, (4) dinoflagellates, (5) stramenopiles (golden, brown, and raphidophyte algae), (6) chlorarachniophytes, and (7) euglenids. Each section starts with a brief description of general characteristics, life cycle, ecology and taxonomic subdivisions of the algal group. Organismal information is mostly taken from (Margulis et al. 1989), unless indicated otherwise. The second part of each section reviews current mtDNA-related information comparing and contrasting genomic features within and between algal lineages.

III. Algal mtDNAs

A. Viridiplantae

Viridiplantae consist of two major divisions, Chlorophyta (chlorophytes) and Streptophyta (streptophytes). Within chlorophytes, three monophyletic groups are currently accepted: Chloro­phyceae, Ulvophyceae, and Treboux­iophyceae. Chlorophytes also include most of the so-called prasinophyte algae – a paraphyletic assemblage of unicellular species that are thought to be descendants of the ancestral lineages from which the main green algal groups evolved (Mattox and Stewart 1984). Streptophytes have separated early from the chlorophytes and include charophycean algae as well as land plants (Embryophyta; Friedl 1997), the latter group being subject of Chaps. 8 and 9. Plastids of Viridiplantae are characterized by containing the photosynthetic pigments chlorophyll a and b.

The first mitochondrial genome sequenced from Viridiplantae was that of the green alga Chlamy­domonas reinhardtii (Michaelis et al. 1990). The unusual, animal-like mtDNA features triggered a flurry of interest in sequencing the mitochondrial genome from other algal lineages. To date, about 20 complete green algal mitochondrial genome sequences are available, revealing an extraordinary diversity in size, shape, and gene content.

Prasinophyta (Chlorophyta)

The paraphyletic prasinophytes comprise unicellular algae with a variable number of flagella (from 0 to 16), and whose cell membranes are often covered by scales (Graham and Wilcox 2000). This group includes primarily marine species and represent a large and important part of the phytoplankton. Notably, the smallest free-living eukaryote known to date, the spherical, flagellum-less Ostreococcus, is a member of the prasinophytes. Seven clades are currently recognized, but the exact relationships within the group and to the other green algal taxa are not fully understood (Marin and Melkonian 2010).

Currently, five prasinophyte mtDNAs have been sequenced completely and one partially, and they show a wide range of genome sizes and gene repertoires (Table 6.2). Among these mitochondrial genomes, three exhibit the so-called “ancestral” (or minimally diverged) pattern of organization (Turmel et al. 1999), notably mtDNA of Nephroselmis olivacea (Pseudos­courfieldiales) (Turmel et al. 1999), Ostreococcus tauri (Robbens et al. 2007), and Micromonas sp. RCC299 (Mamiellales) (Worden et al. 2009). Ancestral genomes have a rather large gene complement, eubacteria-like ribo-protein gene clusters, a few introns, and slowly diverging gene sequences. Indeed, at 44–47 kbp, these rather compact genomes are the most gene-rich chlorophyte mtDNAs sequenced to date, coding for up to 19 respiratory and ATP synthase proteins, 15 ribosomal proteins, three rRNAs and 26 tRNAs (Table 6.3). All three mtDNAs contain almost the same gene set, display a circular map, use the standard genetic code, and vary little in A+T-content. Also, the gene order is quite similar in these three genomes. Introns have only been detected in Nephroselmis mtDNA, which possesses four group I introns. Interestingly, in Ostreococcus and Micromonas, the rns gene is fragmented in two pieces, a derived feature that is more common among chlorophycean mtDNAs (discussed later). The Ostreococcus mtDNA is also distinct in displaying very short intergenic regions. In fact, with only 7% intergenic sequences, this mitochondrial genome is the most gene-dense among Chlorophyta. Further, a segmental duplication covers 44% of the mitochondrial genome. A large duplicated region is also present in the close relative Micromonas sp., and a 1.5 kbp inverted repeat has been detected in the partially sequenced mtDNA of Tetraselmis (Platymonas) subcordiformis (Kessler and Zetsche 1995).

Table 6.2. General characteristics of mtDNAs from green algae
Full size table
Table 6.3. Mitochondrial gene content in green algae
Full size table

The other two fully sequenced prasinophyte mitochondrial genomes are from Pedinomonas minor (Pedinomonadales) (Turmel et al. 1999) and Pycnococcus provasolli (Pseudoscourfieldiales, Pycnococcaceae) (Turmel et al. 2010). These are both considerably smaller genomes (24–25 kbp) encoding a much smaller set of proteins and RNAs – a feature shared with the “reduced-derived” mtDNAs of chlorophycean algae (Tables 6.2, 6.3). Notable is the absence of several nad genes and most (in Pycnococcus) or all (in Pedinomonas) ribosomal protein genes. However, although of similar small size, the gene contents of these two mtDNAs differ substantially, with Pycnococcus coding for almost twice the number of tRNAs and proteins. The larger number of genes in Pycnococcus results in a higher overall gene density in this mtDNA relative to that of Pedinomonas. Nevertheless, in Pedinomonas all 22 genes are also found tightly packed on a 16-kbp segment, while the remaining 9-kbp region is composed of a complex array of repeated sequences (Turmel et al. 1999). As in several other reduced mtDNAs (e.g. those of the chlorophycean Chlamydomonas eugametos and Chlorogonium elongatum see below), all genes in Pycnococcus and Pedinomonas are encoded on the same strand. A deviant genetic code is employed in both algal mitochondria, with the unprecedented use in Pycnococcus of UUA and UUG ­(normally leucine codons) as stop codons (Table 6.2). The rnl gene is fragmented in two pieces in Pedinomonas but not in Pycnococcus, and a group II intron (at precisely the same site as a group II intron in the mitochondrial rnl gene of the brown alga Pylaiella) is found exclusively in Pedinomonas. Both Pycnococcus and Pedinomonas mitochondrial gene sequences appear to evolve more rapidly when compared to those of other prasinophyte lineages and chlorophytes in general.

A recent plastid-based phylogeny suggests an affiliation of Pedinomonas with the Trebouxio­phyceae (Turmel et al. 2009b), which is however not confirmed in a phylo­geny with substantially more plastid sequence data (see Chap. 3). Accumulating genomics data will eventually bring ‘order’ to the paraphyletic prasinophytes.

Chlorophyceae (Chlorophyta)

Chlorophyceae is a diverse assemblage of primarily freshwater algae, but secondarily marine species are also known. This group includes unicellular (flagellated or non-­flagellated) as well as multicellular species. Chlorophyceans are subdivided in two phylogenetically distinct clades, whose initial circumscription was based on the configuration of the flagellar basal bodies. Therefore, clades are commonly referred to as the CW (“clockwise”) and the DO (“directly opposed”) clades (Booton et al. 1998; Graham and Wilcox 2000). Note that several chlorophycean taxa, particularly Chlamydomonas (also termed the Chlamydomonas ‘complex’), are polyphyletic. Mitochondrial genomes from species of both clades have been sequenced. To date, mtDNA sequences are available from several CW lineages, including Chlamy­domonas, Polytomella, Chlorogonium, Dunaliella and Volvox. The only representative of the DO clade with completely sequenced mtDNA is Scenedesmus (for references, see below).

a. CW-Chlorophyceae

Mitochondrial genomes in the CW clade are of variable conformation (either a single circular-mapping, or 1–2 linear chromosomes) and rather small in size (Table 6.2), and possess extremely reduced gene complements limited to seven respiratory protein genes, 1–3 tRNA genes, rnl and rns (Table 6.3). The two latter genes are fragmented and scrambled around the genome. CW-chlorophyceans are subdivided in several subclades, and most of the completely sequenced mtDNAs are from the Reinhardtinia (as defined by Nakada et al. 2008). Among these are Polytomella capuana, Polytomella parva, Chlamydomonas reinhardtii, and Volvox carteri. Mitochondrial genomes from the Reinhardtinia are linear, small and relatively G+C-rich with the notable exception of Volvox.

At 13 kbp and 43% A+T, the mtDNA of Polytomella capuana (Smith and Lee 2008a) – a non-photosynthetic and wall-less relative of C. reinhardtii – is the smallest and most G+C-rich among Viridiplantae mtDNAs. With its nine genes common to mtDNAs of all CW-chlorophyceans plus one tRNA, it is also the most gene-poor mitochondrial genome among Viridiplantae. The single mitochondrial chromosome is linear with telomeres on both termini. A single-stranded loop or a ‘broken loop’ forms the very ends of the molecule (Mallet and Lee 2006; Smith and Lee 2008a).

In contrast to P. capuana, the mtDNAs of two other Polytomella species (P. parva and P. piriformis) occur as two chromosomes (13 and 3 kbp), with the smaller containing only one gene (nad6). Both molecules display at their termini identical 1.3-kbp inverted repeats with either closed or open telomeric structures, similar to the situation described in P. capuana (Fan and Lee 2002; Mallet and Lee 2006). Furthermore, there is evidence for variation in both size and number of mitochondrial chromosomes between various Polytomella isolates (Mallet and Lee 2006). Presumably, illegitimate recombination between short inverted repeats has transformed the single linear chromosome such as that in P. capuana into multiple chromosomes as found in P. parva and P. piriformis (Smith and Lee 2008). Finally, the three Polytomella mitochondrial genomes differ in their intergenic regions. Unlike the P. capuana mtDNA, where inverted repeats punctuate all but two of the genes, repeats are absent from P. piriformis and P. parva mtDNAs.

Both C. reinhardtii (Gray and Boer 1988; Vahrenholz et al. 1993) and its close relative, Chlamydomonas incerta for which a partial mtDNA sequence is available (Popescu and Lee 2007), possess linear mitochondrial genomes with terminal inverted repeats. Yet, the terminal structures in C. reinhardtii differ from those in Polytomella, as they consist of terminal inverted repeats including ∼40 nt-long non-complementary 3’ single-stranded extensions. Furthermore, compared to Polytomella spp., two more tRNAs are encoded in the mtDNAs of C. reinhardtii (Michaelis et al. 1990) and C. incerta, and a reverse transcriptase-like coding region (possibly a remnant of a group II intron; Nedelcu and Lee 1998) is present in the C. reinhardtii mtDNA. Another difference relative to Polytomella spp. is that several group I introns interrupt mitochondrial genes in C. reinhardtii and C. incertae. Note that although C. reinhardtii mtDNA is usually considered intronless, up to three optional group I introns have been found in certain isolates (Smith and Lee 2008b). A small (ca. 20 kbp) linear mtDNA with terminal inverted repeats of about 1.8–3.3 kbp have also been described for a colonial relative of C. reinhardtii, namely Pandorina morum (Moore and Coleman 1989).

The multicellular CW-alga Volvox carteri has a considerably larger (35 kbp) and more A+T-rich mtDNA (Smith and Lee 2009) than its unicellular and colonial Reinhardtinia relatives, and its genome maps as a circular molecule (Smith and Lee 2009, 2010) (see Table 6.2). Also, two group I introns (at least one of which is absent in some Volvox strains) and one group II intron have been found in this genome. The larger size of the Volvox mtDNA (i.e. twice the size compared to C. reinhardtii) is due to introns and long intergenic regions, most of which carry short palindromic repeats. Apparently, two mitochondrial genome isomers (A and B) are present in Volvox that differ from one another in gene arrangement. These isomers likely result from illegitimate recombination between repetitive elements, a phenomenon commonly seen in plastid DNAs (see Chap. 3) and mtDNAs of plants (see Chaps. 8, 9), but rather rarely observed in non-plant mtDNA (but see Sect. III, C of this chapter). Intergenic plus intronic regions of Volvox mtDNA taken together amount to approximately 60% (Smith and Lee 2009, 2010). Based on this calculation, the Volvox genome has been considered the most bloated chlorophyte mtDNA. However, when only intergenic regions are taken into account (see Table 6.2), it is the ulvophyte Pseudendo­clonium (discussed later) that has the most loosely packed mtDNA among chlorophytes.

From CW-algae outside the Reinhardtinia, three mtDNA sequences are known. These are from Chlamydomonas eugametos (Denovan-Wright et al. 1998), Chlorogonium elongatum (Kroymann and Zetsche 1998) and Dunaliella salina (Smith et al. 2010) (Table 6.2). The rns and rnl genes in these three mtDNAs are less fragmented than the ones in Reinhardtinia (i.e. three vs. six gene pieces for rns and six vs. at least eight gene pieces for rnl). In these mtDNAs, all genes are encoded on the same strand (i.e. have the same transcriptional polarity), which is a trait shared by all currently sequenced small (<35 kbp) and circular-mapping mtDNAs of CW-chlorophycean algae, but also seen in fungi and other eukaryotic taxa (e.g. Schizosaccharomyces pombe (Bullerwell et al. 2003) and Thraustochytrium aureum (Gray et al. 2004)). A single coding strand apparently emerged several times independently in mitochondrial genome evolution.

b. DO-Chlorophyceae

The only available mtDNA sequence from the DO-chlorophyceans belongs to Scenedesmus obliquus (Kück et al. 2000; Nedelcu et al. 2000). At 43 kbp, this genome is in the medium size range among green algal mtDNAs. Identified genes account for only 60% of the genome, and both group I and group II introns are present (Table 6.2). The larger gene content and lower degree of rRNA fragmentation (two rns and four rnl fragments (Table 6.3); note that the breakpoints are shared with Chlamydomonas) classify this mitochondrial genome as less derived relative to those of all other chlorophyceans. A particularity of Scenedesmus mitochondria is a deviant genetic code characterized by the use of UAG (normally a stop codon) to specify leucine, and the unprecedented recoding of UCA (normally a serine codon) as a stop codon. Most puzzling was the finding of a C-terminally truncated mitochondrial cox2 gene whose second half was suggested to be nucleus-encoded ((Funes et al. 2002); (reviewed in Burger et al. 2003b)). Consistent with this suggestion, Chlamydomonas and Polytomella species possess a split cox2 gene with both parts encoded in their nucleus (Pérez-Martínez et al. 2001). Other examples of split genes with one half in mtDNA and the other in nuclear DNA are rpl2 in Arabidopsis (Adams et al. 2001) and cox1 in several eukaryotes including algae (Gawryluk and Gray 2010). These latter cases likely represent an intermediate phase of gene migration toward the nucleus.

Trebouxiophyceae (Chlorophyta)

The third chlorophyte group, the Trebouxiophyceae (sensu (Friedl 1995)), consists of several widespread and biotically significant algae that inhabit mostly soil and freshwater, and include most green algal phycobionts of fungi (lichens), ciliates and cnidarian animals (corals). Trebouxiophytes are either unicellular nonflagellated or filamentous algae (Booton et al. 1998) with basal bodies displaced in a counterclockwise (CCW) configuration, a trait shared with the Ulvophyceae ((Kreimer and Melkonian 1990); see below).

The two trebouxiophyte mtDNAs sequenced thus far are from non-photosynthetic relatives of Chlorella – Prototheca wickerhamii (Wolff et al. 1993) and Helicosporidium sp. (Pombert and Keeling 2010) – which are common parasites of vertebrates and invertebrates, respectively. The two genomes are circular-mapping and very similar in size, A+T-content and gene repertoire (differing by a single tRNA), and their level of synteny is higher than between any other two sequenced chlorophyte mtDNAs. At 55 and 49 kbp, these mtDNAs are in the medium size range among chlorophytes, and their nucleotide composition places them among the most A+T-rich genomes in this group (see Table 6.2). The Helicosporidium mitochondrial genome is more densely packed than that of Prototheca and has a trans-spliced group I intron in its cox1 gene, otherwise described for only two other unrelated species (Burger et al. 2009; Grewe et al. 2009). In terms of gene complement, these mtDNAs resemble closely the “ancestral” type represented by the prasinophyte Nephroselmis, coding for a large number of respiratory and ribosomal proteins (see Table 6.3). Also, in contrast to chlorophycean mtDNAs, the two trebouxiophytes have an rrn5 gene, continuous rns and rnl coding regions, as well as a complete set of tRNA genes for translation of all mitochondrial codons. This latter capacity is shared by only one other chlorophyte, the ulvophyte Pseudendoclonium (discussed next).

Ulvophyceae (Chlorophyta)

Ulvophyceae, the fourth chlorophyte group, comprise both unicellular and multicellular species, including some of the largest and most conspicuous green algae (seaweeds). Organisms of certain macroscopic species consist of a single, large multinucleate cell (coenocyte) (Graham and Wilcox 2000). Ulvophyceae are common on rocky intertidal sea coasts in temperate regions, but freshwater species are also known. The phylogenetic position of Ulvophyceae within Chlorophyta, and especially their relationship to Chloro­phyceae, is unclear. Some studies propose that Ulvophyceae emerged before the divergence of Trebouxiophy­ceae and Chlorophyceae, while others suggest a possible sister-group relationship between the Ulvophyceae and Chlorophyceae, with the Trebouxiophyceae rather occupying a basal position (see (Pombert et al. 2004) for discussion and references; see also the plastid-gene-based phylogeny in Chap. 3).

The ulvophyceans whose mitochondrial genome has been sequenced to date are Oltmannsiellopsis viridis (Pombert et al. 2006) and Pseudendoclonium akinetum (Pombert et al. 2004). The two algae belong to distinct, deeply-diverging lineages in the Ulvophyceae (Friedl and O’Kelly 2002; Pombert et al. 2005). Their mtDNAs differ greatly in size, with that of Pseudendoclonium being almost 40 kbp larger than the Oltmannsiellopsis mtDNA (57 kbp). In fact, at 96 kbp, Pseudendo­clonium has the largest chlorophyte mtDNA known so far (see Table 6.2). Most of the mitochondrial genome in Pseudendoclonium is occupied by intergenic dispersed repeats. The mitochondrial gene complement of the two genomes is rather similar, but gene arrangement differs markedly. Further, mtDNA of Oltmannsiellopsis is special in showing strong evidence for intracellular, inter-organellar transfer of a group I intron, because both the plastid and the mitochondrial rnl contain a group I intron inserted at the same position, and these introns are more similar to each other than either of them is to any other known introns at that position (Pombert et al. 2006). Another particularity of this alga is that the mitochondrial genome appears to have recently captured via horizontal transfer a group II intron from a cryptophyte and an integrase from a bacterium (Pombert et al. 2006).

Pseudendoclonium mtDNA exhibits certain features typical of the “expanded” pattern of embryophytes (also seen in some charophytes discussed below). This challenges the previous notion that only land plants allowed a substantial gain of intergenic sequences in mtDNA (Turmel et al. 2007). Pseudendoclo­nium mtDNA also challenges the categorization of green algal evolutionary patterns into ancestral/reduced-derived/expanded types, calling for more fine-grained notions as detailed in Sect. IV of this chapter.

Charophyceae (Streptophyta)

Charophyceae belong to the second division of Viridiplantae – the streptophytes – and form a monophyletic lineage together with embryophytes ((e.g. Karol et al. 2001; Rodriguez-Ezpeleta et al. 2005); see Fig. 6.1). Charophyceae contain six lineages of mainly freshwater algae. These are (1) the Mesostig­matales including a single species, the scaly, unicelled biflagellate Mesostigma viride (previously regarded as a member of the Prasinophyceae, but now placed confidently at the base of streptophytes (Rodriguez-Ezpeleta et al. 2007b)); (2) the Chlorokybales, also constituted by a single unicellular species, the sarcinoid Chlorokybus atmophyticus; and further the multicellular (3) Klebsormidiales; (4) Zygnematales; (5) Coleochaetales; and (6) Charales. Mesostig­matales and Chlorokybales are likely the earliest diverging charophyceans, forming a distinct clade (Turmel et al. 2007), while the branching order of the other taxa remains uncertain. Some studies indicate that Charales are the closest relatives to land plants, but others suggest that this group diverged prior to Zygnematales and Coleochaetales (for references and discussion, see Turmel et al. 2006). Mitochondrial DNAs of four charophyte lineages have been sequenced to date. These are from Mesostigma and Chloroky­bus, Chaetosphae­ridium globosum (Coleochaetales), and Chara vulgaris (Charales).

Charophycean mtDNAs have similarly large gene contents, but vary considerably in genome size (Tables 6.2, 6.3). The medium-sized mitochondrial genome of Mesostigma (Turmel et al. 2002b) is the only one among green algae possessing trans-splicing group II introns that otherwise occur frequently in land plants (Bonen 2008). Chlorokybus mtDNA (Turmel et al. 2007) is not only the most gene-rich one among all Viridiplantae, but also the largest. At 202 kbp, this genome is about twice as large as the largest chlorophyte mtDNA (i.e. that of Pseudendoclonium (Ulvophyceae)) and exceeds even that of the bryophyte land plants, Marchantia polymorpha (Oda et al. 1992) and Physcomitrella patens (Terasawa et al. 2007). Contributory to the huge mtDNA size are the numerous group I and group II introns (also present in tRNA genes) and intergenic regions that account for about half of the genome. Chlorokybus mtDNA is the least densely packed green algal mitochondrial genome currently known.

The two multicellular charophyceans Chaeto­sphaeridium globosum (Turmel et al. 2002a) and Chara vulgaris (Turmel et al. 2003) both possess medium-sized mtDNAs. The Chara mtDNA is unique among green algae in encoding components involved in cytochrome c biogenesis (see Tables 6.1, 6.3). While common in land plants, these genes are found in only a few algal mtDNAs including that of the rhodophyte Cyanidioschyzon merolae (see below).

B. Glaucophyta

Glaucophytes (aka glaucocystophytes) are freshwater microalgae with planktonic or sessile life styles. Some taxa are flagellated, others are not, and unicellular and colonial species exist. At least nine genera have been recognized. Glaucophytes are characterized by the presence of a cyanelle in their cells. This photosynthetic organelle has retained from its cyanobacterial ancestor a thin peptidoglycan wall and the phycobiliproteins phycocyanin and allophycocyanin organized in typical phycobilisomes on the surface of unstacked thylakoids. Phycobilisomes and unstacked thylakoids are ancestral (cyanobacterial) features shared only by red algae (Graham and Wilcox 2000). In contrast to Viridiplantae, glaucophytes (and rhodophytes, see below) have chlorophylls a and c as photosynthetic pigments. The phylogenetic position of glaucophytes has been controversial. Recent phylogenomics analyses show their common ancestry with red algae and green algae plus plants, but the precise branching order remains elusive (Rodriguez-Ezpeleta et al. 2005), as this topology is not obtained in phylogenies that use different species and different sequence sampling.

Mitochondrial genome sequences are available for two glaucophytes, Cyanophora paradoxa and Glaucocystis nostochinearum (Price et al. 2012; GenBank acc. nos. HQ849544, HQ908425). The two genomes are similar in terms of A+T-composition and gene content (Tables 6.4, 6.5). The gene complement is large (including genes for 5S rRNA, 15 ribosomal proteins, 10 Complex I subunits, and two Complex II subunits), the gene order displays vestiges of the eubacterial str, S10, spc and alpha operons, and the gene sequences are little derived. Despite these similarities, the two mtDNAs exhibit a completely different gene order outside operon-like clusters and differ markedly in size, with 52 kbp for Cyanophora and 34 kbp for Glaucocystis. The larger size of the Cyanophora genome is due to a 5-kbp long duplicated region containing nad9, cox3, and rns, and several 0.3-kbp long direct repeats occurring 2–3 times and overlapping genes partially or completely.

Table 6.4. General characteristics of mtDNAs from non-green algae
Full size table
Table 6.5. Mitochondrial gene content in non-green algaea
Full size table

C. Rhodophyta

Rhodophyta is a morphologically diverse group with up to 6,000 unicellular and multicellular species in at least 12 orders. Red algal cells are unique among algae in lacking centrioles and flagella from all life stages, and in having plastids with unstacked thylakoids and containing phycoerythrin as an accessory photosynthetic pigment. Rhodophytes inhabit tropical and temperate near-shore marine waters; many of these species are of economic and ecological significance (Graham and Wilcox 2000). Several distinct red algal lineages are known to date but their phylogenetic ­relationships are not well understood (Saunders and Hommersand 2004).

To date, complete mitochondrial genomes have been sequenced from six species belonging to four distinct rhodophyte lineages, and partial information is available from one additional species. Species include two unicellular Cyanidales – Cyanidioschyzon merolae (Ohta et al. 1998) and Cyanidium caldarium (partial sequence, Viehmann et al. 1996) – and five multicellular species from three distinct groups – Chondrus crispus (Gigartinales) (Boyen et al. 1994; Leblanc et al. 1995), Porphyra purpurea (Bangiales) (Burger et al. 1999), and Gracilariopsis andersenii, Gracilariophila oryzoides, and Plocamiocolax pulvinata (Hancock et al. 2010) (Floridophyceae). Red algal mtDNAs are relatively similar in size (between 25 and 36 kbp), conformation (circular-mapping), and gene order (for details, see discussion in Burger et al. 1999) (Table 6.4). They are rather compact, having as little as 4% non-coding sequences and displaying several cases of overlapping genes (in Chondrus and Cyanidioschyzon). The presence of three mitochondrion-encoded succinate dehydrogenase genes is unique among algae (Table 6.5). In terms of coding capacity, the mtDNAs of five of the six red algae are essentially identical, while that of Cyanidioschyzon encodes several additional genes specifying ribosomal proteins and components involved in the biogenesis of cytochrome c (ccmA, B, C, F; see Table 6.1). Notable is the occurrence of a mitochondrion-encoded rpl20 that is otherwise only known from jakobids. Some confusion arose about whether or not red algae possess a mitochondrion-encoded 5S rRNA. The initial report that Chondrus mtDNA includes an rrn5 situated between cox2 and cox3 (Leblanc et al. 1995) was discounted (Lang et al. 1996). Yet, the claim of rrn5 being absent from red algal mtDNAs had to be rectified, when this gene was later detected in another genomic location (between nad3 and rps11) in Chondrus (Gray et al. 1998), Cyanidioschyzon and Cyanidium (Gray et al. 2004); a mtDNA-encoded 5S rRNA is lacking only in Porphyra (Burger et al. 1999) (Table 6.5). Another difference between the studied red algae is that mitochondrial translation uses the standard genetic code in Cyanidioschyzon and Cyanidium, but UGA specifies tryptophan in the other species (see Table 6.4).

The mitochondrial genome of Porphyra assumes two isomeric conformations that differ from each other in the orientation of a 15-kbp region between two ∼300 bp-long repeats (Burger et al. 1999). The same phenomenon has been described recently in Volvox mtDNA (see Sect. III.A.2), but is known for a long time from angiosperm mitochondria where repeats promote major genome rearrangements via a flip-flop mechanism of illegitimate recombination (Hanson and Folkerts 1992). Further, Porphyra mtDNA displays numerous sequence polymorphisms, including mostly substitutions (transitions and transversions), but also insertions and deletions. At the time of the initial report (Burger et al. 1999), it was not clear whether the observed sequence polymorphism and the two isomer conformations are due to true heteroplasmy (more than one type of mtDNA in the same organism) or rather to diversity within the population of thalli that were collected in the wild, propagated in the laboratory and then used to construct the clone library for mtDNA sequencing. When we later examined distinct blades of Porphyra including female, male and bisexual thalli, we could demonstrate that each individual contains two mtDNA conformations, and often but not always, sequence polymorphisms. The sequence deposited in GenBank (acc. no. NC_002007) is that of the predominant isomer (excess factor of about 5–10; G. Burger, D. Tremblay, unpublished). The observed heteroplasmy could be a consequence of mtDNA being inherited biparentally. In fact, a certain ­percentage of biparental mtDNA inheritance has been observed in the relative P. yezoensis (Choi et al. 2008).

A few group II introns (but no group I introns) are present in red algal mtDNAs (Table 6.4). Although the large majority of mitochondrial introns are found in protein-coding genes, four of the six sequenced rhodophyte mtDNAs have introns residing in a tRNA gene (trnI of Chondrus (Boyen et al. 1994) and Gracilariopsis, Gracilariophila and Plocamiocolax (G. Burger, this report)). The two introns in the Porphyra mito­chondrial rnl gene are remarkable, because they are most similar to introns in the cyanobacterium Calathrix sp., suggesting recent horizontal transfer between bacteria and mitochondria.

D. Stramenopiles

Stramenopiles comprise nearly 40 taxa (see Patterson 1989), whose monophyly is well supported by molecular phylogeny. About half of the stramenopile taxa are nonphotosynthetic such as oomycetes (Phytophthora) and bicosoecids (Cafeteria), while plastid-carrying groups include phaeophytes (brown algae), chrysophytes (golden algae), bacillariophytes (diatoms), raphidophytes, and xanthophytes. Complete mtDNA sequences are available for all these algal groups except xanthophytes.

Stramenopile mtDNAs have a large set of tRNA genes, but all lack trnT (the gene for tRNA-Thr), which apparently was lost from this genome in the common ancestor of stramenopiles (Gray et al. 2004). To sustain mitochondrial translation, tRNA-Thr must be either imported into mitochondria, or generated from another mitochondrial tRNA by post-transcriptional RNA modification or editing. Note that loss of mitochondrial trnT is not unique to stramenopiles, but also occurred in other groups independently (e.g. jakobids; Burger and Lang, unpublished).

Chrysophytes

Most golden-brown algae are unicellular, naked flagellates, but some species cover their cells with silica scales and others form complex colonies or filaments. Chrysophytes typically inhabit freshwater, where they are part of the phytoplankton. The golden-brown color of their plastids is due to chlorophyll a, chlorophylls of the c-group and typically beta-carotene. More than 120 chrysophyte genera are recognized today.

Complete mtDNA sequences are available from two different genera, represented by Chrysodidymus synuroides (Chesnick et al. 2000) and Ochromonas danica (Burger et al. unpublished; GenBank acc. no. AF287134; see Tables 6.4, 6.5). The mtDNA size is 34 kbp for Chrysodidymus and 41 kbp for Ochromonas, the former being circular-mapping and the latter linear with a 2.2-kbp terminal repeat at both ends that includes several ORFs and tRNA genes. The size difference between these two mtDNAs is due to these repeats and a larger number of ORFs in Ochromonas. Both taxa display a large mitochondrial gene complement including nad11 and 14 ribo-protein genes. Chrysodidymus has in addition tatC, whereas Ochromonas mtDNA contains a dpo gene; the latter has likely been acquired via a plasmid as also seen in other mtDNAs (for a review, see Gray et al. 2004). The two chrysophyte mtDNAs share several clusters of identical gene order. Some clusters exhibit the ancestral eubacteria order of ribosomal protein genes; other clusters such as nad2-nad9-nad7 have a common derived arrangement that must have be present already in the common ancestor of both golden algae, with a trend to group genes of similar functions (here genes for subunits of Complex I).

Diatoms

Diatoms (Bacillariophyta) are unicellular, silica-walled algae found in freshwater and marine habitats. They are considered the most abundant aquatic organisms after viruses and bacteria. Only a few diatoms have flagella; most are free-floating (planktonic) or attached to plants, sand or rocks (benthic). Plastids vary in color from yellowish-brown to deep brown, containing chlorophylls a and c, and xanthins. Diatoms are very diverse, and taxa are numerous with about 250 recorded genera and more than 100,000 species.

Complete mtDNA sequences are available for two species from different classes, the fragilariophycean Synedra acus (Ravin et al. 2010), and the coscinodiscophycean Thalassiosira pseudonana (Armbrust et al. 2004), while the mtDNA sequence from the bacillariophycean Phaeodac­tylum tricornutum is nearly completed (Bowler et al. 2008; Ravin et al. 2010; see Tables 6.4, 6.5). At 77 kbp, the mitochondrial genome of Phaeodactylum is nearly twice as large as that of the other two diatom species. This size difference is due to the length of their repeat regions, which stretch over 36 kbp in Phaeodactylum, but only 5 and 4 kbp in Synedra and Thalassiosira. The mitochondrial gene content of diatoms is quite similar to that of the golden algae discussed above. Remarkably, most diatoms possess a mitochondrion-encoded rpl10, a gene that is not seen in mtDNAs from other stramenopiles and that is generally infrequent. Not only the gene set but also the gene arrangement on mtDNA is well conserved among diatoms, which is unexpected since the three diatoms belong to distinct lineages with quite diverse morphologies.

Phaeophytes

Brown algae are multicellular organisms whose sizes vary from microscopic filaments to large blades sometimes more than 10 m long (kelp). Photosynthetic pigments include chlorophylls a and c, and fucoxanthins. The habitat of phaeophytes is mostly marine costal regions, where they grow attached to rocks etc. or sometimes invertebrate animals. The phylum comprises about 250 genera and nearly 1,000 species.

Complete mtDNA sequences are available for 12 species from five different phaeophyte lineages, Desmarestia viridis (Desmarestiales), Dictyota dichotoma (Dictyotales), Fucus vesiculosus (Fucurales) (Oudot-Le Secq et al. 2006), Laminaria digitata (Laminariales) (Oudot-Le Secq et al. 2002), Pylaiella littoralis (Ectocarpales) (Oudot-Le Secq et al. 2001) and Saccharina angustata (Laminariales; Yotsulura et al. unpublished, NC_013473) plus six other species of the Saccharina genus (see Tables 6.4, 6.5). The sizes of brown algal mtDNAs vary between 32 and 59 kbp with Dictyota having the smallest and Pylaiella the largest genome. Although the investigated species are quite distinct in morphology and habitat, their mtDNAs are most similar. The mitochondrial gene complement is large – including rpl31 that is seen only sporadically in other groups – and identical across phaeophytes. Only the number of introns (all group II), tRNAs, and ORFs varies. Even the mitochondrial gene order is essentially identical between brown algae (but completely different compared to diatom mtDNAs), with sporadic ORFs or introns inserted/deleted in one or the other mtDNA. Remarkable is nad11 that only encodes the N-terminal FeS-binding domain (ca. 230 residues), whereas the rest of the protein is missing (Oudot et al. 1999). A similarly short nad11 has also been observed in the non-photosynthetic stramenopile Cafeteria roenbergensis (Burger et al. unpublished; GenBank acc. no. NC_000946). It was speculated that nad11 functions as a short version (Oudot et al. 1999), but it is more probable that the second half of the protein is encoded in the nucleus and imported into mitochondria, as is the case for cox2 in Scenedesmus (see above).

Raphidophytes

Raphidophyte algae are flagellated unicellular organisms with usually bright green plastids that contain chlorophylls a and c, and large amounts of carotenoid pigments. In contrast to most other stramenopile algae, raphidophytes lack a rigid cell wall. The group includes planktonic freshwater and marine species that occur where vegetation is abundant, and several taxa are responsible for toxic algal blooms harmful to fish (for references, see Masuda et al. 2011). Raphidophytes is a small group of organisms with four genera and a total of nine species. The complete mtDNA sequence is available for two species, Chattonella marina var. marina and Heterosigma akashiwo ((Masuda et al. 2011); see Tables 6.4, 6.5).

The two raphidophyte mtDNAs are of moderate size (39 and 45 kbp). The size difference is due to ∼4 kbp more intergenic sequence and the presence of two group II introns in cox1 in Chattonella mtDNA, while the gene content is the same. Gene order is quasi identical with the exception of a cluster of seven genes that has changed orientation, and several open reading frames that are present/absent in various genomic locations. The mitochondrial gene content of raphidophytes is essentially the same as in brown algae, only that rpl31 is missing in the two raphidophytes.

E. Alveolates

Alveolates are subdivided into about eight taxa, one of which, the dinoflagellates, is mostly photosynthetic, while another, the apicomplexans, is predominantly non-­photosynthetic, but possesses plastid relicts (e.g. Plasmodium and Eimeria). The few photosynthetic apicomplexan species include Alveolata and Chromera (Janouskovec et al. 2010) (see Chap. 2 for the origin of their plastids). Apicomplexans are generally not considered algae, and therefore, their mt­DNAs will not be discussed in this chapter. More information on this subject is available in specialized reviews (e.g. McFadden and Waller 1997; Williams and Keeling 2003).

Dinoflagellates

Dinoflagellates inhabit marine and fresh­water ecosystems. Their great diversity of cell shapes comes from the flattened membrane sacs (alveoli) beneath the plasma membrane that form armours of most baroque shapes. All dinoflagellates have two flagella that are inserted at the same point, with one wrapping around the cell and the other oriented perpendicularly to the first. Photosynthetic taxa play a major role in ocean carbon fixation, and are equally notorious for toxic red tides as they are for symbiotic partnerships with reef-building corals. Plastid pigments include chlorophylls a and c2 together with the unique xanthophyll peridinin, further beta-carotene and xanthins. Some dinoflagellates have chlorophyll c1 and fucoxanthin instead, testifying to multiple independent acquisitions of plastids from diverse sources through higher-order endosymbioses.

The dinoflagellate species with most mtDNA sequence available today is Amphidium carterae (33 kbp), but the genome sequence is far from being completed. Data from numerous taxa combined indicate a set of only five mitochondrial genes, cob, cox1, cox3, rnl, and rns, implying that most of the traditional mitochondrial genes must have migrated to the nucleus (Tables 6.4, 6.5; (reviewed in Nash et al. 2008; Waller and Jackson 2009)). Interestingly, cox2 exists as a split nuclear gene similar to the situation in chlorophyceans (Waller and Keeling 2006). Mitochondrial genes exist not only as contiguous forms, but also in pieces and associated within different genomic contexts, as first described for Crypthecodinium (Norman and Gray 2001). Further, the dinoflagellate mitochondrial genomes appear to undergo considerable recombination (Jackson et al. 2007). On top, RNA editing (Lin et al. 2002, 2008) and trans-splicing of certain genes (Waller and Keeling 2006) make these mitochondrial genomes some of the most bizarre and most difficult to analyze. For a recent review, see (Burger et al. 2011).

F. Chlorarachniophytes (Cercozoa)

Chlorarachniophytes are a small group of photosynthetic flagellates that live in marine habitats. Plastids are bright green and contain chlorophylls a and b, but neither chlorophyll c nor biliprotein pigments. As in the unrelated cryptophytes (see below), a nucleomorph reveals that photosynthesis was acquired via secondary endosymbiosis, here with a green alga. Two chlorarachniophyte genera are recognized, Chlorarachnion and Bigelowiella.

A nearly complete mtDNA sequence is available for Bigelowiella natans (Burger et al. unpublished; GenBank acc. no. HQ840955; Table 6.4). The mitochondrial genome consists of a linear chromosome of ∼38 kbp, with about 400 bp and 700 bp remaining unsequenced at the two extremities. The mitochondrial gene set is moderately large (Table 6.5). All genes are encoded on the same strand with only a few nucleotides in intergenic regions, and some protein-coding genes overlap by a few nucleotides. Bacterial operon organisation has vanished, and the inferred protein sequences are divergent. Given the small genome size, it is surprising to find five ORFs longer than 100 residues (185–605 amino acids). Some of these ORFs are likely fortuitous as they include small stretches of regular genes; others might be ribo-proteins that are too derived to be recognized.

G. Cryptomonads

Cryptomonads are single-celled flagellates recognizable by their flattened asymmetric cells and distinctive swimming motion. They can be found in diverse aquatic habitats, from drainage ditches to tundra ponds and the open ocean. Most of the 20 or so genera are photosynthetic. The heterotrophic Goniomonas is believed to be primarily without plastids, i.e. closely related to the cryptomonad ancestor that engulfed an alga with red plastids (for a discussion on plastid origins see Chap. 4). Photosynthetic cryptomonads are important planktonic primary producers at the base of aquatic food webs, feeding rotifers, mussels and many other animals. Some species form toxic blooms. Cryptomonad plastids are characterized by chlorophyll c and phycobilins as photosynthetic pigments. Cells carry in addition to their indigenous nucleus a remnant ‘nucleomorph’ wrapped in a membrane sack together with the plastid, which are the leftovers of the red-algal endosymbiont (for a review, see Maier 1992).

Complete mtDNA sequences are available for two species, Hemiselmis andersenii (Kim et al. 2008) and Rhodomonas salina (Hauth et al. 2005) (see Tables 6.4, 6.5). The genomes are 61 kbp and 48 kbp in size, respectively, and minimally diverged as to gene complement, genetic code, and gene order. First, the mitochondrial gene set is among the richest outside jakobids including, in addition to the more common genes discussed above, also atp1 and nad8. Ribo-protein genes are arranged in eubacteria-like clusters in both cryptophyte mtDNAs, but the order of other genes is different. Overall, sequences are minimally divergent. Both genomes are compact with only a single sizeable intergenic region, which in Rhodomonas mtDNAs is 4.7 kbp long and contains an elaborate 4.5 kbp repeat region with two large blocks of inverted orientation. The blocks are composed of numerous tandem repeats and hairpin structures with unit length of ∼35–700 nt. Within this region map the predicted promoters and replication origins. Hemiselmis has a five times larger mitochondrial repeat region that also consists of non-palindromic and palindromic motifs, which, compared to those in Rhodomonas mtDNA, are shorter (∼20–350 nt long), differ in sequence, and are all arranged in the same orientation.

H. Haptophytes

Haptophytes (prymnesiophytes) are planktonic, biflagellated algae occurring in marine habitats. All members of this group are photosynthetic. Most species are unicellular, but colonial taxa are known as well. Haptophyte cells are covered with scales of often complex ornamentation. Species with calcite scales such as Emiliania are among the most productive lime producers on Earth. Haptophyte plastids appear yellow-brown due to high concentrations of carotenoids and xanthins, in addition to chlorophylls a, c1 and c2. We recognize about 50 genera with some 500 species.

Complete mtDNA sequences are available for two species, Emiliania huxleyi (Sanchez Puerta et al. 2004) and Pavlova lutheri (Burger et al. unpublished; GenBank acc. no. HQ908424; see Tables 6.4, 5). These mtDNAs resemble each other consi­derably in gene content, genome size (29–34 kbp), and a large intergenic repeat region of 1.3 kbp in Emiliania and 1.6 kbp in Pavlova. Mitochondria of Pavlova and other Pavlovales use the standard genetic code, whereas Emiliania as well as species from five other haptophyte orders read UGA ‘stop’ codons as tryptophan (Hayashi-Ishimaru et al. 1997). Remarkably, E. huxleyi mtDNA carries the dam gene encoding DNA adenine methylase, which was likely acquired by horizontal gene transfer. The only other mtDNA where this gene has been discovered is that of the charophyte Klebsormidium (BF. Lang, G. Burger, MW. Gray, unpublished). Haptophyte mitochondrial gene sequences are relatively derived accounting for long branches in mitochondrial-protein-based phylogenies and uncertain branching position (Baurain et al. 2007) (see Fig. 6.1).

I. Euglenozoa

Euglenozoa, one of the deepest-diverging eukaryotic lineages, include three taxa, kinetoplastids, diplonemids, and euglenids (Simpson et al. 2002). Euglenids branch basally to the other two clades, and they are the only euglenozoan group with photosynthetic members (see Fig. 6.1).

Euglenids

Euglenids are unicellular flagellates of ovoid to lanceolate body shape. Most are naked with a plasma membrane that is stiffened by distinctive pellicle strips of helical arrangement. Euglenids are cosmopolitans living in most different aqueous habitats that are typically rich in organic matter. They are not found in plankton but rather at the water-mud or water-air interface, since bacteria are their food source. Nearly 50 genera and 1,000 species are recognized. About one third of euglenid species possess plastids. Photosyn­thesis has apparently been acquired rather late in euglenid evolution, because basally branching taxa such as Peranema trichophorum and Petalomonas cantuscygni are heterotrophic (Moreira et al. 2001). Euglenid plastids are grass-green (chlorophyll a and b, beta-­carotene and xanthophylls); according to phylogenetic analyses, the plastid donor was a relative of the chlorophyte alga Pyramimo­nas ((Turmel et al. 2009a); see also Fig. 2.1 in Chap. 2).

Data on photosynthetic euglenid mtDNA are available for only a single species, Euglena gracilis, yet the genome sequence is far from being complete (Tables 6.4, 6.5). Defined chromosomes are not detectable, but rather numerous linear molecules of a broad range of different sizes, mostly around 4 kbp, with a smaller fraction around 7.5 kbp (Spencer and Gray 2011). In total, about 13 kbp mtDNA sequence is available in public databases. The A+T content is extremely high (75–80%), and gene sequences are evolving at accelerated rates yielding an extremely long branch for Euglena in mitochondrial-gene-based phylogenies (e.g. Yasuhira and Simpson 1997). Genes identified so far on mtDNA are rnl and rns (Spencer and Gray 2011), cox1 (Tessier et al. 1997; Yasuhira and Simpson 1997), cox2 and cox3 (Tessier and Paulus, unpublished; GenBank acc. no. AF156178.1). The cox3 sequence was initially annotated as nad6, but the corresponding reading frame rather represents the N-terminal portion (residues 1−∼160) of cox3 (Spencer and Gray 2011), which is most poorly conserved not only in Euglena but also in the other Euglenozoa, including trypanosomes and diplonemids (Vlcek et al. 2011), and therefore difficult to recognize. It remains to be seen whether a complete copy of cox3 exists in Euglena mtDNA, or whether the missing half is in the nuclear DNA. The gene may also be split into two mitochondrion-encoded pieces, just like the rRNA genes in this genome. Ribosomal RNA gene halves occur in different genomic contexts, together with numerous short dispersed fragments of authentic genes, as well as direct repeats arranged in multiple combinations. This indicates that the highly complex structure of Euglena mtDNA is due to frequent recombination.

A Euglena-like genome structure is also observed in the heterotrophic euglenid Peranema, whereas in Petalomonas (the second non-­photosynthetic euglenid for which some mtDNA data are available) appears to possess a rather traditional mtDNA architecture (Roy et al. 2007). The ‘chaotic’ mitochondrial genome organisation seen in Euglena has probably emerged in the common ancestor of euglenids, prior to the acquisition of plastids.

IV. Recurring Patterns of Mitochondrial Genome Evolution

The mtDNAs that resemble most the genome of the alpha-proteobacterial predecessor of mitochondria are found among the (plastid-less) jakobid flagellates, which typically have a ca. 70 kbp circular mtDNA that carries genes for ∼70 proteins and ∼30 structural RNAs. These mtDNAs appear to be ‘frozen in time’ (Lang et al. 1997) as to gene set, gene order and bacteria-like transcription. Algal mtDNAs are not only very different from the minimally derived mtDNAs of jakobids; they are also highly diverse amongst each other. This illustrates that most disparate evolutionary forces must have been at work to shape mitochondrial genomes in the various algal lineages. Still, several prominent evolutionary patterns emerge in algal mtDNAs discussed above. But these patterns occur sporadically and independent of taxonomic groupings, with sometimes astoundingly similar mitochondrial genomes in phylogenetically most distant lineages (convergent evolution).

The probably most conspicuous trend in mtDNA evolution is what is often referred to as ‘genome reduction’ including the decrease of physical size and gene set. This phenomenon occurred (independently), for instance, in the lineages leading to C. reinhardtii and Pedinomonas (Turmel et al. 2010). However, the two genome features do not always contract in parallel as exemplified by Volvox (Smith and Lee 2009), where a minimal mitochondrial gene complement is accompanied by an ordinary-sized mitochondrial genome. Alternatively, genome size reduction can occur by maintaining a substantial gene repertoire (yet for obvious reasons to only a certain degree). This is achieved, as exemplified by Bigelowiella mtDNA, via densely packing the genes on mtDNA combined with shortening the coding regions. Densely packed genomes with reduced intergenic portions also evolved in red algae (Plocamiocolax and Chondrus), chrysophytes and phaeophytes (Chrysodidymus and Dictyota). In addition, adjacent protein-coding genes may be even fused to economize the stop and start codons as in the rhizarian Acanthamoeba castellanii (Burger et al. 1995). Alternatively, genes can overlap partially so that the same stretch of DNA encodes different information. Overlapping protein-coding genes are specified in different reading frames, as seen in clusters of ribosomal protein genes, e.g. in the mtDNAs of the red alga Chondrus (Leblanc et al. 1995) and the cryptomonad Rhodomonas (Hauth et al. 2005). In other cases, protein-coding genes overlap with tRNA genes, such as in the mtDNA of the chrysophyte Chrysodidymus (Chesnick et al. 2000). Because of all the above, one should distinguish the two different reductive evolutionary trends: one that affects genome size and the other that affects gene number in mtDNAs, the latter mostly via accelerated gene transfer to the nucleus.

Genome expansion is another recurrent evolutionary pattern observed in mitochondria of various lineages. Size inflation is typically due to growing intergenic regions that accumulate dispersed repeats across the genome as, for example, in the charophyte Chlorokybus and the ulvophyte Pseudendoclonium. Intergenic regions may as well expand through accumulation of inactive mitochondrial gene fragments piling up between the complete, functional genes, as seen in dinoflagellates. Further, foreign DNA such as chloroplast sequences, may be hoarded in mtDNA, which to our knowledge has been observed exclusively in embryophyte, but not in algal mtDNAs. In contrast to an expansion of intergenic regions across the entire mitochondrial genome, there are also select cases of a single, several kbp-long ‘surplus’ region. Such mtDNA segments may be composed of numerous tandemly-arranged repeat motifs as in the green alga Pedinomonas (Turmel et al. 1999) or the cryptophyte Hemiselmis (Kim et al. 2008). Tandem repeats likely expand through slipped mispairing, and repeat inversions probably arise via strand switching during replication (for a review, see Bzymek and Lovett 2001). Alternatively, ‘surplus’ regions may arise via segmental duplication of mtDNA and contain supernumerary gene copies as seen in the prasinophyte Ostreo­coccus (Robbens et al. 2007).

Changes in genome conformation have also occurred independently in several algal lineages. This includes linearization (such as in the Reinhardtinia clade and in certain chrysophytes and chlorarachniophytes), fragmentation (in the non-photosynthetic chlorophycean algae Polytomella, in euglenids and dinoflagellates), and the presence of multiple isomeric genomic forms (in the red alga Porphyra and the green alga Volvox). Such conformational changes are thought to be consequences of illegitimate recombination involving short repeated sequences (Burger et al. 1999; Smith and Lee 2009).

Another pattern encountered in mtDNAs of various algal lineages is the gradual loss of the ancestral gene order of ribosomal protein genes. Among algae, the bacterial str, S10, spc, and alpha operon arrangement is most conserved in the prasinophyte Nephroselmis (Turmel et al. 1999) and in the cryptophyte Rhodomonas (Hauth et al. 2005). On the other hand, the degree of gene shuffling can be so extreme that not even a single pair of ribosomal protein gene neighbours is retained as for example in the mtDNA of the cercomonad alga Bigelowiella (Burger et al. unpublished; GenBank acc. no. HQ840955).

Further evolutionary trends are observed in mtDNAs of phylogenetically unrelated algae and eukaryotes in general. These include sometimes drastic deviations from the universal genetic code (in the chlorophycean Scenedesmus and the prasinophyte Pycnococcus (see also Hayashi-Ishimaru et al. 1996)), fragmentation of rRNA genes (in both the prasinophyte Pedinomonas and the ancestor of chlorophycean green algae as well as in Euglena), and pronounced nucleotide composition bias with an A+T content that is either extreme high (in Euglena and the trebuxiophyte Prototheca) or extremely low (in the chlorophycean Polytomella capuana).

In sum, the evolutionary history of algal mitochondrial genomes appears like an exploration of all possible avenues of tolerable deviations, which has brought about the astounding diversity of mtDNAs in algae and eukaryotes as a whole.