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.

1 Introduction

The view that plastids have evolved from cyanobacteria by endosymbiosis and that this event took place originally only once is nowadays widely accepted (see Chap. 2). This primary endosymbiosis gave rise to the common ancestor of today’s three primary plastid-containing lineages (together known as Archaeplastida, Table 4.1): (1) green algae and plants, (2) red algae, and (3) glaucophyte algae (Reyes-Prieto and Bhattacharya 2007; Lane and Archibald 2008). From these lineages, plastids have spread laterally by secondary endosymbiosis to form further lineages (see Chap. 2). Thus, two unrelated lineages containing plastids of green algal ancestry (euglenoids and chlorarachniophytes) and several lineages containing plastids of red algal ancestry (haptophytes, cryptophytes, stramenopiles, dinoflagellates, and apicomplexa) have evolved (Cavalier-Smith 1999, 2002; Lane and Archibald 2008). In addition, the occurrence of serial secondary endosymbiosis and tertiary plastids that replaced the original secondary plastid (Yoon et al. 2005) is under discussion (Lane and Archibald 2008; Janouskovec et al. 2010). Most of the resulting species gain their energy through photoautotrophic carbon fixation. However, in probably every land plant and algal lineage, parasitic species have evolved that have abandoned photoautotrophic growth and rather live by parasitizing on other plants, fungi, or even animals.

Table 4.1 Plastid genomes sequenced from different lineages of the plant kingdom
Full size table

Species of colorless unicellular algae with reduced plastid genomes have been described among the green algae (de Koning and Keeling 2004; Tartar and Boucias 2004; Borza et al. 2005; de Koning and Keeling 2006), the euglenoid algae (Gockel and Hachtel 2000), and the dinoflagellates (Sanchez-Puerta et al. 2007; Matsuzaki et al. 2008). The apicomplexan group of unicellular parasitic organisms, last but not least, provides the most prominent example that plastid-containing parasites have been able to utilize a very wide host range (Lim and McFadden 2010). More recently, claims of a photosynthetic/red algal (and therefore, possibly, plastid) history of the oomycete Phytophthora (Tyler et al. 2006) and of ciliates (Reyes-Prieto et al. 2008) have been made (Janouskovec et al. 2010), taking the discussion around adaptations to a nonphotosynthetic life style another step ahead.

Parasitic land plants are – in contrast to their algal counterparts – quite eye catching (Fig. 4.1) and have received much attention not only due to their special lifestyle but also due to the damage they can inflict on agricultural land use. It is estimated that approximately 1% of all angiosperm species from at least 11 different lineages have resorted to a parasitic lifestyle (Barkman et al. 2007). Based on their attachment sites on their hosts, root parasites and shoot parasites are being distinguished (Fig. 4.1). The liverwort Aneura mirabilis (formerly known as Cryptothallus mirabilis) is one example – and as a matter of fact, to date, the only known example – of a nonvascular parasitic land plant that has evolved into a completely nonphotosynthetic lifestyle (Bidartondo et al. 2003).

Fig. 4.1
figure 1_4

Parasitic land plants. Parasitic plants can attach to the roots of their hosts, as in Monotropa uniflora (Indian pipe, a), or to the stems or leaves as in species of the genus Cuscuta (b, c). The attachment organs, or haustoria, are visible in (c)

Full size image

The following chapters will summarize our current knowledge on the structure and function of plastid genomes from algae to higher plants and will focus on the degeneration of plastid genomes in species that have resorted to a parasitic lifestyle.

2 The Plastid Genome

2.1 The Organization of Plastid DNA

Every plastid possesses its own genetic information that was inherited from the cyanobacterial ancestors of these organelles. Reflecting this ancestry, the plastid genome has retained many features of prokaryotic genomes, including the overall structure, physical properties, gene organization, and regulatory features necessary for gene expression (Bock 2007). Plastid DNA is present in multiple copies per plastid and is compacted by DNA-binding proteins (Kuroiwa 1991). Microscopic studies in combination with fluorescent staining of these high-molecular-weight DNA–protein complexes generally known as nucleoids, revealed that their number, shape, size, and distribution can vary significantly between species and depending on the developmental stage (Kuroiwa 1991). Plastid genome maps of all plant lineages traditionally show the organelle’s genetic information as circular molecules. The only striking exception to this seem to be the dinoflagellates where the normal single circle has been replaced by minicircles, which contain one or a few genes each (Zhang et al. 1999; Barbrook and Howe 2000). Electron microscopic pictures later revealed that the circular conformation is only one of many that the plastid DNA of higher plants can appear in. In addition to circular monomeric molecules, a variety of linear and branched monomers and oligomers (Deng et al. 1989; Lilly et al. 2001; Oldenburg and Bendich 2004; Scharff and Koop 2006) and possibly also shorter fragments (Kolodner and Tewari 1972) have been observed.

In contrast to the structural variability of the nucleoids, plastid coding capacity and gene organization have remained remarkably conserved across the plant kingdom, with higher plants, in particular, showing very little variation in their gene content and plastid genome sizes (Jansen et al. 2007). The selective pressure on the maintenance of photosynthesis-related genes and genes for subunits of the plastid gene-expression machinery seems to have been instrumental for maintaining the plastid genome and conserving a core set of plastid genes (Bock 2007). It is this selective pressure exerted by photosynthesis or, rather opposite, the obvious lack of it in parasitic plants that has drawn the attention to parasitic plant plastid genomes and their particular evolution.

2.2 Structure and Coding Capacity of Plastid Genomes

The first complete plastid genome sequences, namely that of the bryophyte Marchantia polymorpha and of the seed plant Nicotiana tabacum, were reported as early as 1986 (Ohyama et al. 1986; Shinozaki et al. 1986). Since then, more than 230 plastid genomes have been completely sequenced. The vast majority of these plastid genome sequences are from land plants and green algae, while other big groups, such as the red algae, dinoflagellates, or stramenopiles, are still strongly underrepresented (Table 4.1).

A common feature of plastid genomes are two inverted repeat regions (IRA and IRB) that split the remainder of the chromosome into a large and a small single-copy region (LSC and SSC, respectively). The inverted repeats can vary significantly in size, from 75 kbp in Pelargonium (Chumley et al. 2006) to 0.5 kbp in Pinus (Wakasugi et al. 1994). The functional role of this tetrapartite structure is unclear and examples of species without this organization are known from green (Hallick et al. 1993; Wakasugi et al. 1997; Jansen et al. 2008) and red plastid lineages (Glockner et al. 2000; Ohta et al. 2003; Hagopian et al. 2004), suggesting that it could be dispensable.

Compared to the ancestral prokaryotic genome, the plastome of all plants and algae is more or less drastically reduced. The genetic information for the majority of plastid-localized proteins either has been lost altogether in the course of evolution or has been transferred to the nucleus from where the gene products are imported back into the organelle (see Chap. 7). With a few exceptions, plastid genomes of the green lineage range between 120 and 160 kbp in size and code for 100 ± 20 proteins as well as about 40 genes encoding stable RNA species (rRNAs and tRNAs) (Palmer 1990). Even the plastome of the hypothetical ancestor to all green lineages was estimated to have contained a total of only 137 protein-coding genes (Turmel et al. 1999). Free-living descendents of the red plastid lineage have, on average, retained larger plastid genomes compared to the “standard plastome” of the green lineage. Nevertheless, gene numbers and sizes in these genomes are still in the same order of magnitude and feature around 200 protein-coding genes on approximately 160–180 kbp (Reith and Munholland 1993; Ohta et al. 2003). Extreme deviations from the average sizes, on the other hand, do exist in single cases and have been reported, for example, for Acetabularia species whose plastid genomes can be up to 1.5 Mbp in size (Simpson and Stern 2002), while probably the smallest plastid genome of a photosynthetic organism with only 72 kbp is that of the green alga Ostreococcus tauri (Robbens et al. 2007).

The information content of the plastid genome can be roughly divided into three large groups (see also Table 4.2): (1) genetic system genes comprising the RNA and protein components of the transcription and translation machineries as well as a few proteins involved in post-transcriptional and post-translational steps, (2) photosynthesis genes for subunits of the light and dark reactions and the ATPase, and (3) conserved open reading frames and genes with miscellaneous functions.

Table 4.2 Core set of genes encoded by the plastid genome of higher plants
Full size table

Genetic system genes required for transcription, translation, and processing steps represent a large portion of all plastid genomes and also represent the major fraction of strongly reduced plastid genomes, such as that of the apicomplexan parasite Toxoplasma gondii (Fig. 4.2). The number of photosynthesis and genetic system genes is slightly higher in red algae and most lineages derived from them, indicating specific losses in the green plastids, whereas a constant number of six genes (equaling 4–5% of the coding potential) on almost all plastid genomes is dedicated to the subunits of the plastid ATPase. In green plastid-derived lineages, the third group that comprises genes with other functions makes up around 20% of the genes, the majority of which are conserved reading frames of unknown function (YCFs) (Fig. 4.2). In most rhodophytes, glaucophytes, and chromalveolates, this gene group is considerably expanded and contains genes for amino acid, lipid, pigment, and cofactor metabolism (Fig. 4.2) that are absent from land plants and green algae. The only exception is the chromalveolate group of apicomplexans, where many typical gene groups of the red plastid lineage have been lost (Wilson et al. 1996). This reduction in genome-coding capacity is related partly to the parasitic life that members of this group are leading and partly to losses that must have occurred independent of the parasitic lifestyle.

Fig. 4.2
figure 2_4

Plastome coding capacity expressed in proportion of the various functional categories in the different lineages of the plant kingdom. Genomes used as examples were published as follows: N. tabacum (Shinozaki et al. 1986; Wakasugi et al. 1998); C. rheinhardtii (Maul et al. 2002); Cyanidium caldarium (Glockner et al. 2000); Cyanophora paradoxa (Genbank Acc # U30821); Euglena gracilis (Hallick et al. 1993); Bigelowiella natans (Rogers et al. 2007); Emiliania huxleyi (Sanchez Puerta et al. 2005); Toxoplasma gondii (Genbank Acc # U87145); Vaucheria litorea (Rumpho et al. 2008); and Guillardia theta (Douglas and Penny 1999). Background colors behind the pies indicate whether the plastids are of green algal origin (green), red algal origin (pink) or of glaucophyte origin (blue)

Full size image

3 Plastid Genomes from Parasitic Species

The angiosperm holoparasite Epifagus virginiana (Beechdrops) has been one of the first parasites to be thoroughly investigated with respect to its plastid genome sequence (Wolfe et al. 1992b); and these analyses have revealed a number of drastic changes that involve mainly gene losses and pseudogenizations. The significant reductions of the coding potential that embrace, among others, all photosynthesis genes in E. virginiana (Table 4.3 and Fig. 4.4), have been explained with the relaxation of the selective pressure exerted otherwise by photosynthesis. The discovery that, among the more than 150 species that are assembled within the holoparasitic genus Cuscuta, not all exhibit the same severe physiological reductions found in Epifagus but that some have, in fact, retained some basal photosynthetic activity (Hibberd et al. 1998; van der Kooij et al. 2000) has more recently enabled glimpses into the transition from photoautotrophy via intermediate mixotrophic states to complete heterotrophy (Krause 2008). A likewise gradient has, so far, not been found anywhere else.

Table 4.3 Plastid gene losses in parasitic versus nonparasitic higher plants
Full size table

3.1 Structural Changes

The typical organization of plastid chromosomes with a large single-copy region (LSC) and a small single-copy region (SSC) separated by two inverted repeat regions (IRA and IRB) has been retained by all parasitic angiosperms (Wolfe et al. 1992b; Funk et al. 2007; McNeal et al. 2007) (Fig. 4.3). For two Cuscuta species, C. reflexa and C. gronovii, overlapping PCR products have indicated the existence of a circular form of the plastid chromosomes (Funk et al. 2007), suggesting overall structural similarities between parasitic and nonparasitic plastid genomes. The same holds true for parasitic algae. Divergences from the standard pattern, for example, in Euglena longa are also present in the photosynthetic relative E. gracilis and are, therefore, unrelated to parasitism (Hallick et al. 1993; Gockel and Hachtel 2000).

Fig. 4.3
figure 3_4

Sizes of the large and small single-copy regions and the inverted repeats. Two photosynthetic angiosperms, Pelargonium and tobacco, and three parasitic angiosperms, Cuscuta reflexa, Cuscuta gronovii, and Epifagus virginiana, are shown. The size of each respective region is shown in basepairs

Full size image

The inverted repeats have been assigned a role as a stabilizing factor that limits genome rearrangements in chloroplasts. Nevertheless, the boundaries of inverted repeats were found to be hot spots for gene duplications or deletions (Yue et al. 2008). In line with this, the IRA–LSC junction (JLA) in C. reflexa and C. exaltata was found within the ycf2 gene, leaving one copy of this gene truncated. As a result of this reduction in IR size, there is only one copy each of rpl2 and trnI-CAU and one complete ycf2 gene (Funk et al. 2007; McNeal et al. 2007). Generally, the size of the inverted repeats in Cuscuta was reduced proportionally to the size of the plastid genome (Fig. 4.3). In E. virginiana, in contrast, the inverted repeats have suffered much less reductions so that their sizes relative to the single-copy regions are much larger. The main reason for this is that the ribosomal genes that are located on the IRs have been retained while many genes that are normally part of the single-copy regions (such as the photosynthesis genes) have been lost. However, differences between IR length and IR gene content are fairly common in higher plants, anyway, as exemplified by the comparison between tobacco and Pelargonium (Fig. 4.3), and even between species of the same genus (Goulding et al. 1996), so inverted repeat sizes cannot be correlated with a particular lifestyle.

Insertions and deletions (“indels”) of larger fragments and inversions that can affect the order of genes on the plastid genome are considered to be almost as important for the evolution of genomes as nucleotide substitutions (Yue et al. 2008). Compared to tobacco and other angiosperm plastid genomes, species of the Cuscuta subgenus Monogyna exhibit three typical sequence inversions within the plastid chromosome, two in the large single-copy region ~2 kb and ~13 kb in length and one of ~1.5 kb length in the small single-copy region (Stefanovic and Olmstead 2005; Funk et al. 2007). Otherwise, the gene order in parasitic plants was not found to be different. Whether the three inversions are of any functional significance is, however, unclear.

A high ratio of coding versus noncoding sequences was found for all Cuscuta species and some parasitic algae, such as Helicosporidium and Euglena longa, as well as for aplicoplast genomes. In contrast, Epifagus exhibits almost the same coding to noncoding sequence ratio as photosynthetic plants (Krause 2008). It has been speculated that an early reaction of the plastid genome to the parasitic lifestyle was the condensation of the genome by loss of many noncoding and possibly unimportant parts of the plastid DNA (Funk et al. 2007). As the adaptations to parasitism became more pronounced, pseudogenization and loss of functional reading frames occurred (Fig. 4.4), with the result that the relative amount of noncoding areas have increased again, as observed in Epifagus. The observation of highly compact genomes in parasitic algae and in Cuscuta might indicate, however, that the Epifagus plastid genome represents an exception.

Fig. 4.4
figure 4_4

Amount of functional genes and pseudogenes in three parasitic and one nonparasitic angiosperm species

Full size image

3.2 Gene Losses in Parasitic Plant Plastomes

A recent study of 64 nonparasitic higher plant plastid genomes has revealed that 66 individual losses of genes have occurred in different species during evolution (Jansen et al. 2007). As many as 62 of these losses were confined to the more derived monocot and eudicot clades, and only four genes (chlB, L, and N as well as trnP-GGG) have disappeared from the plastid genome at a very early stage (Jansen et al. 2007). Two genes were lost particularly often: infA for which 11 independent losses have been recorded in eudicots and that is missing, among others from all Solanaceae (Table 4.3), and accD with a total of six independent losses in monocots and some eudicots. The reported losses seem to follow no specific pattern.

The adaptation to parasitism, on the other hand, has resulted in a loss of genes that is the more pronounced, the lesser the photosynthetic capacity of the parasite has become. Here, functional correlations between gene losses and metabolic activities can be drawn more easily.

3.2.1 Ndh Genes

Eleven ndh genes on a “normal” plastid genome code for a chloroplast NAD(P)H dehydrogenase (Table 4.2). These genes were reported missing in several photosynthetic genera, such as Pinus, Phalaenopsis, and Chlamydomonas (Wakasugi et al. 1994; Maul et al. 2002; Chang et al. 2006), which was interpreted as evidence that these genes are nonessential. This question seems, however, far from being resolved, as a recent report defends the view that the Ndh complex plays an essential role in photosynthesis and discusses evidence for a possible transfer of ndh genes to the nucleus in conifers and Gnetales (Martin and Sabater 2010).

A loss of ndh genes was also reported for all sequenced plastid genomes of parasitic dicots [C. reflexa, C. exaltata, C. gronovii, C. obtusiflora, and E. virginiana (Wolfe et al. 1992b; Funk et al. 2007; McNeal et al. 2007)] (Table 4.3) and the recently published underground orchid Rhizantella gardneri (Delannoy et al. 2011) and besides also for the parasitic liverwort Aneura mirabilis (Wickett et al. 2008a, b). Like in all other cases where these genes were reported missing, the entire set consisting of 11 genes has been lost without exception, since these species do no longer rely on photosynthesis.

In E. longa, the ndh genes are also reported missing. However, in this case, the loss is shared with the photosynthetic species E. gracilis.

3.2.2 Photosystem Genes

Genes for photosystem I and photosystem II as well as the cytochrome b6f complex are encoded on all nonparasitic species. Most photosystem genes were retained in the four species of Cuscuta whose plastid genome sequences are available. C. reflexa exhibits no losses, which coincides with comparatively mild reductions in the rate of photosynthesis (van der Kooij et al. 2000), while photosynthesis rates are more severely compromised in C. gronovii. Since all genes were found to be transcribed (Berg et al. 2003) in both species, the finding that the psaI gene was lost in C. gronovii (Funk et al. 2007) could be of significance in that context. The retention of photosynthesis genes is the biggest difference to the plastid genome of E. virginiana (Table 4.3). All genes associated with the bioenergetic processes of photosynthetic electron transport and ATP synthesis were either lost or pseudogenized (Table 4.3 and Fig. 4.4). The condensation grade of the plastid genome of E. virginiana with only 70 kbp and a coding capacity of just 42 genes (Wolfe et al. 1992b) is in higher plants only outmatched by the underground orchid Rhizantella (Delannoy et al. 2011).

Analyses of nucleotide substitution rates showed that the psaI gene that is missing in, for example, C. gronovii, showed significantly increased KA/KS values in C. reflexa and C. exaltata (Krause 2010). PsaI is a small subunit of photosystem I that has only one transmembrane domain and is involved in the docking of the PsaL subunit to this photosystem (Yu et al. 2008; Vanselow et al. 2009). The high KA/KS values suggest that this protein is obviously not evolving under selective constraint, and that this lack of selective pressure is what presumably leads to its eventual loss. It has, however, not yet been determined whether a copy of this gene has been transferred to the nucleus and can functionally replace the lost plastid gene.

3.2.3 The rbcL Gene

The large subunit of Rubisco, rbcL, is encoded by the plastid genome. In accordance with the loss of photosynthetic activity, this gene was lost in some aphotosynthetic species, such as E. virginiana (Wolfe et al. 1992b), C. odorata (van der Kooij et al. 2000), and R. gardneri (Delannoy et al. 2011). However, many reports on other parasitic plant families, among them holoparasitic Scrophulariaceae, showed that rbcL was surprisingly conserved as a functional plastid gene independent of whether the photosynthetic capacity was retained or not. Open reading frames for rbcL were detected, for example, in Lathrea clandestina, a parasite of alder (Alnus glutinosa), where it seems to be expressed, despite the fact that this plant lacks chlorophyll (Lusson et al. 1998). Similar situations have been described for other holoparasites (Thalouarn et al. 1994; Wolfe and dePamphilis 1997, 1998; Delavault and Thalouarn 2002).

Likewise, the parasitic liverwort Aneura mirabilis and the euglenoid alga E. longa have retained seemingly functional rbcL genes, while all photosystem genes were deleted (Gockel and Hachtel 2000; Wickett et al. 2008a). The fact that rbcL has been retained in many but not all parasitic plant plastomes makes it difficult to associate a particular meaning to this, but it has been discussed that Rubisco could have a separate metabolic function independent of photosynthesis (Krause 2008).

3.2.4 Ribosomal Protein Genes

A total of 25 genes of the ribosomal small and large subunits are encoded by most higher plant plastomes (Table 4.2). Although no tendency toward enhanced nucleotide substitution rates in ribosomal protein genes was observed in Cuscuta species (Krause 2010), parasitic plant genomes exhibit several losses of rpl and rps genes (Morden et al. 1991; Funk et al. 2007). Like with the photosynthesis genes and other gene groups, the number of losses roughly follows the gradient of dependency on heterotrophic growth. In C. reflexa, only two genes, rpl23 and rps16, have nonfunctional reading frames and behave as pseudogenes. Along with a third gene, rpl32, both genes have been completely lost in C. gronovii. E. virginiana has even suffered four losses (rpl22, rpl32, rps15, and rps16) in addition to two pseudogenizations (rpl14 and rps23) (Table 4.3).

The rpl32 gene is not only missing from some parasitic plant plastid genomes but was also lost in a number of photosynthetic angiosperms (Jansen et al. 2007), among them Populus alba (Table 4.3). In P. alba, the corresponding chloroplast gene appears to have been transferred to the nucleus. The transfer of chloroplast genes to the nucleus is a process that requires many steps, including the removal of possible introns, the gain of suitable regulatory elements, as well as the acquisition of a transit peptide that can direct the nuclear gene product back to the plastids (Bungard 2004; Ravi et al. 2008). In case of the P. alba rpl32 gene, it could be shown that it acquired the transit peptide from another plastid targeted gene, cp sod-1 (Ueda et al. 2007), thereby paving the way for the Rpl32 protein’s return into the chloroplast.

Another example for a ribosomal gene whose loss has also been observed in P. alba is rps16 (Ueda et al. 2008). In this case, however, the gene was not simply transferred to the nucleus. Rather, the original plastid rps16 gene has been lost, and this loss has been compensated for by import of the mitochondrial rps16 gene. Apparently, the nuclear gene for the mitochondrial rps16 gene has acquired a dual-targeting signal that is able to direct it to the plastids in addition to mitochondria, rendering the plastid’s own gene dispensable.

This functional replacement of a ribosomal gene from one organelle by a dually targeted counterpart from the other organelle is not a unique case. A recent report has shown that, for example, the rpl10 gene in several plant mitochondrial genomes has been replaced by a dually targeted copy of the original “chloroplast-only” targeted rpl10 isoform (Kubo and Arimura 2010). It is possible that some gene losses from parasitic plant plastomes have been compensated for in a similar manner.

3.2.5 Rpo Genes

Higher plants possess two RNA polymerases, PEP and NEP, which share the responsibility of transcribing the plastid genetic information (Hess and Börner 1999). The PEP is a multi-subunit enzyme and is encoded by four plastid genes: rpoA, rpoB, rpoC1, and rpoC2 (Table 4.2). The encoded subunits display a similarity to the eubacterial multisubunit RNA polymerase, and PEP recognizes promoters with a structure similar to bacterial promoters (Hess and Börner 1999). PEP is predominantly responsible for the transcription of photosynthesis-related genes. The generation of homoplastomic rpo knock-out mutants in tobacco, consequently, leads to a loss of photosynthetic capacity that is accompanied by an off-white phenotype (DeSantis-Maciossek et al. 1999) and a reduction in the amount of transcripts of photosynthesis genes (Krause et al. 2000; Legen et al. 2002).

The entire set of rpo genes was reported missing in E. virginiana, which has lost any photosynthetic activity and with that also any of the PEP-dependent genes. Losses of the rpoA subunit were both previously and subsequently reported for photosynthetic species, such as Euglena gracilis (Hallick et al. 1993), Pelargonium (Chumley et al. 2006), and Passiflora (Jansen et al. 2007) (Table 4.3). The operon containing rpoB, C1, and C2, on the other hand, has generally been retained on plastid genomes. To date, the only exceptions are the plastid genomes of C. gronovii and C. obtusiflora, which were the first and only plastid genomes with a confirmed loss of the entire rpo gene set in a plastid genome where photosynthesis genes are not only present (Funk et al. 2007) but, moreover, actively expressed (Berg et al. 2003). It has been shown that transcription of these genes was taken over by the nuclear-encoded RNA polymerase, NEP, which is highly homologous to the mitochondrial phage-type RNA polymerase and might have evolved from it by gene duplication (see Chap. 12).

Unlike parasitic angiosperms, apicomplexans, Helicosporidium and E. longa have retained the rpo genes in their reduced plastid genomes (Wilson et al. 1996; Gockel and Hachtel 2000; Cai et al. 2003; de Koning and Keeling 2006; Janouskovec et al. 2010). In contrast to higher plants, algae appear to possess only a single nuclear gene for a phage-type RNA polymerase and its gene product seems to be exclusively localized to the mitochondria. A second nuclear-encoded but plastid-localized RNA polymerase, NEP, seems to be missing in algae (Smith and Purton 2002). A loss of the PEP subunits encoded by the plastid rpo genes would therefore deprive the plastids of the possibility to transcribe their genetic information, making these genes essential.

3.2.6 Hypothetical Conserved Reading Frames (Ycfs)

Plastid genomes contain a number of conserved open reading frames of unknown function. The conservation of some of these sequences within plastomes from higher plants down to algae or even cyanobacteria is interpreted as strong indication for their functional importance (Ravi et al. 2008). Attempts to uncover the function of the ycf gene products have been successful in some cases, and the use of aliases to the ycf designations is becoming more common. For example, ycf5 has been renamed ccsA [for c-type cytochrome synthesis, (Xie and Merchant 1996)], ycf9 is psbZ (Swiatek et al. 2001), and ycf10 is now called cemA [for chloroplast envelope membrane protein A, (Rolland et al. 1997)]. For ease of reference, the older ycf designations will be used here in this chapter.

Nicotiana species contain nine conserved reading frames in their plastid genome sequence (ycf1-5, ycf9,10, and 15 and orf350) (Wakasugi et al. 1998; Yukawa et al. 2006), six of which have been lost in E. virginiana (ycf3, 4, 5, 9, 10, and 15). Only the pseudogenization of ycf15 is shared with the Cuscuta species (Table 4.3). The retention of ycf1 and ycf2 on all of the parasitic plant plastid genomes indicates that a function of these genes in photosynthesis can be most likely excluded and that rather, a possible role in gene expression or a photosynthesis-unrelated metabolic function can be envisaged. This observation corroborates previous findings that knockout mutants of ycf1 and ycf2 never yielded viable homoplastomic lines and that these genes are therefore essential for the survival of higher plants (Drescher et al. 2000).

3.2.7 tRNA Genes

The set of tRNA genes encoded by plastomes of photosynthetic plants encompasses around 30 genes and it has been argued that a transfer of tRNA genes to the nucleus and re-import of the RNAs is impossible (Barbrook et al. 2006; Howe and Purton 2007). Nevertheless, some parasitic species do exhibit extensive losses of tRNA genes.

In C. reflexa, only a single tRNA, that for lysine (trnK-UUU) is missing. More extended losses of tRNA genes were observed in C. gronovii, where in addition to trnK-UUU the sequence for trnV-UAC was completely eliminated from the plastid DNA, and four tRNA genes (trnA-UGC, trnG-UCC, trnI-GAU, and trnR-AGC) have been reduced to pseudogenes. In E. virginiana, a total of five tRNAs (trnA-UGC, trnC-GCA, trnI-GAU, trnR-UCU, and trnS-GGA) were pseudogenized and eight tRNAs (trnG-GCC, trnG-UCC, trnK-UUU, trnL-UAA, trnT-GGU, trnT-UGU, trnV-GAC, and trnV-UAC) lost.

The extensive loss of tRNA genes from parasitic plastids has, therefore, raised the question whether the missing tRNAs have been replaced by nuclear-encoded ones or whether the codon usage was adapted to these losses. An analysis of codon usages in two Cuscuta species has shown that all 61 sense codons were found in the coding regions of plastid genes in a similar proportion as in nonparasitic plants that possess a “full” plastid tRNA set (Funk et al. 2007). A similar picture emerges from the Epifagus plastid genome. This was interpreted as circumstantial evidence for an import of cytosolic tRNAs into the chloroplasts (Wolfe et al. 1992a; Funk et al. 2007). It is, however, also possible that an “extended wobble” behavior of the remaining tRNAs partly compensates for the losses.

3.2.8 trnK-UUU and matK

Among the tRNA genes, trnK-UUU has a special role as this tRNA gene harbors in its intron the only RNA maturase gene found on the plastid genome, matK. The trnK-UUU gene was found missing on all parasitic angiosperm plastomes, but neither the gene nor its intron and the matK gene contained within have been reported missing in any nonparasitic plants. In C. reflexa, C. exaltata, and E. virginiana, surprisingly, matK has been retained as a free-standing gene, confirming the probably essential function in plastid intron maturation that has been attributed earlier to its gene product (Hess et al. 1994; Hubschmann et al. 1996; Jenkins et al. 1997; Vogel et al. 1997). Unprecedented in higher plants, however, was the complete loss of matK from C. gronovii and C. obtusiflora (Funk et al. 2007; McNeal et al. 2007).

The exceptional loss of matK from the plastid genomes of C. gronovii and C. obtusiflora is probably closely related to the loss of group IIa introns from a whole suite of genes in parasitic plant plastids. Group IIa introns have been discussed as targets of MatK (Liere and Link 1995; Hubschmann et al. 1996; Jenkins et al. 1997; Vogel et al. 1997). Out of originally eight group IIa introns, only a single group IIa intron, namely intron 2 of clpP, is found in C. gronovii and C. obtusiflora, where the matK gene was deleted. This intron was shown to be faithfully spliced and might therefore not be a target of MatK action (Funk et al. 2007). This conclusion was supported by recent biochemical and molecular evidence linking MatK to all group IIa introns, except the clpP intron 2 (Zoschke et al. 2010).

In Apicomplexa and in E. longa, an absence of the gene matK along with all group II introns was observed as well but this does not seem to be the result of a parasitic lifestyle here. The loss of matK in these species is shared with their photosynthetic relatives since matK was reported to be already missing in Chromera velia (Janouskovec et al. 2010) and E. gracilis (Gockel and Hachtel 2000).

3.3 Reduction of RNA-Editing Sites

The loss of introns was not the only posttranscriptional processing step that has experienced a reduction. A surprise was the reduction of RNA-editing sites that was not only the result of the loss of genes that are typically richer in editing sites. Of the 30–40 editing sites of photosynthetic seed plants (Tsudzuki et al. 2001), only 17 potential sites remain in C. reflexa, while 12 have been found in C. gronovii. Several sites are only partially edited or not edited at all. The loss of editing competence as well as the reduction in the number of introns has been discussed in a very recent review (Tillich and Krause 2010) and the reader is referred to this review for further details.

3.4 Loss of PEP Promoters

Each of the two enzymes sharing the responsibility for plastid transcription (PEP and NEP) differs with respect to the promoters they bind to (Hess and Börner 1999). PEP promoters resemble prokaryotic promoters and occur mainly upstream of the photosynthetic genes, whereas the phage-type NEP promoters can be found upstream of genes for the genetic system. Many genes and operons, such as the gene cluster for the ribosomal RNAs, even possess both promoter types.

The loss of the rpo genes from two Cuscuta plastid genomes where photosynthesis genes are still present (Krause et al. 2003; Berg et al. 2004) raised the question of how the corresponding promoters have developed. The analysis of promoter motifs upstream of photosynthesis genes revealed that the consensus −10 and −35 boxes of PEP promoters have been so severely changed that they must be considered to be nonfunctional (Funk et al. 2007). For the rrn operon and the rbcL gene it could be, moreover, demonstrated that the start sites of transcription have been shifted relative to those of tobacco and that the 5′ region of the novel transcription start site revealed striking similarities to the consensus sequence recognized by the NEP polymerase, indicating a shift from PEP- to NEP-driven transcription of these genes. This shift obviously enables the plastids to transcribe the previous PEP genes with high enough efficiency to allow for low photosynthetic activity.

4 Gene Retentions and the “Raison d’être” of Reduced Plastid Genomes

Overall, many of the changes that were seen in connection with a parasitic lifestyle seem to be shared between higher plant and algal parasites. The postulated forces that must exist, according to deKoning and Keeling, for algal parasites and that shape plastid genomes even after relaxation of photosynthetic selection pressure (de Koning and Keeling 2006) do seemingly also apply to higher plants. The most reduced plastid genomes in both groups (i.e., Epifagus for seed plants and Apicomplexa for algae) are characterized by a domination of genetic system genes and only two to four genes with functions outside of gene expression are present.

One question that has repeatedly been asked but so far never satisfactorily been answered is the mystery why plastid genomes were retained in nonphotosynthetic organisms. In this context, genes that encode for subunits of the gene-expression machinery, such as ribosomal RNAs and proteins, are hardly of much interest, since they are presumed to only secure the expression of the “key” plastid gene(s). Consequently, the answers why the plastid genome has not been lost altogether have been sought in the other retained genes.

In many plastid genomes of parasitic plant species, intact reading frames of the rbcL gene that codes for the photosynthesis protein Rubisco have been retained, despite the loss of photosynthetic activity (Krause 2008). It has been suggested that Rubisco could assume a second function in lipid biosynthesis (Schwender et al. 2004). An argument that strengthens the “essential lipid biosynthesis” hypothesis is the retention of the accD gene even in strongly reduced genomes, such as that from E. virginiana. Tobacco plastome mutants, where accD was interrupted, did never reach a homoplastomic state, underlining the essentiality of this gene for plants (Kode et al. 2005). However, in Apicomplexa there is no accD gene and rbcL is missing in apicomplexan parasites as well as in Epifagus and some Cuscuta species, just to name some (see Sect. 4.3.2).

Another set of genes that appear essential for plastid development independent of photosynthetic capacity are those that encode the subunits of the Clp Protease, clpP and clpC. Clp most likely performs chaperone functions and is engaged in protein import into the plastid. While in seed plants, the clpP gene was retained even in reduced plastomes, such as that of Epifagus, Apicomplexa have retained the gene clpC. However, exceptions are found also here (e.g., Helicosporidium), weakening any hypothesis that was tentatively built up around the clp genes.

Among the protein-coding genes, the two hypothetical reading frames ycf1 and ycf2 remain. Knockouts of each gene resulted persistently in heteroplasmy (Drescher et al. 2000; Shikanai et al. 2001; Kuroda and Maliga 2003) and both belong to the reduced gene set of extremely reduced plastid genomes, such as that of E. virginiana. However, their function might well be found to be associated with gene expression (Bock 2007), which would eliminate also these genes from being candidates for the “raison d’être” of plastid genomes.

Some recent alternative attempts at an explanation for the retention of plastid genomes circle around two transfer RNA genes. The first is the gene for the glutamyl-tRNA (trnE). This (trnE) gene fulfills three tasks in plastids: besides its role in protein biosynthesis, it plays a well-known role in the synthesis of δ-aminolaevulinic acid and, thereby, in heme biosynthesis (Jahn et al. 1992), and may regulate transcriptional activity of the NEP (Hanaoka et al. 2005), although this last function has been later challenged (Bohne et al. 2009). The essentiality of heme biosynthesis and the belief that a functional transfer of tRNA genes from the plastid to the nuclear genome is unlikely if not impossible (Barbrook et al. 2006) has been used as an argument for why a plastid genome has been retained, however much reduced it is. It has even been predicted some years ago that trnE may be the only gene that is found in all genomes, regardless of the degree of reductions (Barbrook et al. 2006). So far, all sequenced plastomes fulfill this prediction. However, also this hypothesis has a caveat, since in Plasmodium, at least, heme was found to be synthesized by an exclusively mitochondrial-located pathway, and it is therefore independent of plastid trnE. A second hypothesis was brought up recently, where the formylmethionyl-tRNA (tRNA fM) plays the main role (Howe and Purton 2007). tRNA fM is needed for translation initiation in plastids and mitochondria, but the only tRNA fM gene in Plasmodium is the one that is located in the apicoplast. Therefore, the formylmethionyl-tRNA pool of the plastids was proposed to be shared by the mitochondria, rendering this particular tRNA indispensable (Howe and Purton 2007). Whether this holds true for further parasitic species still awaits confirmation.

5 From Loss of Photosynthesis to Loss of Plastids?

As described in the previous chapters, all nonphotosynthetic land plants without exception have retained a more or less cryptic plastid with a plastome that exhibits a set of typical losses and also of typical gene retentions. A number of nonphotosynthetically living algae and descendents thereof have likewise retained a cryptic plastid that apparently fulfills some essential functions for the parasites.

For a long time, the debate has been going on why these plastids with their “crippled” plastid genomes have remained so steadfast in the nonphotosynthetic species. The discovery that the apicomplexan species Cryptosporidium has lost its plastid (Huang et al. 2004) has given the debate a new spin. Evidence for plastid-derived genes in the nonphotosynthetic oomycete Phytophthora (Tyler et al. 2006), in Ciliates (Reyes-Prieto et al. 2008), and in trypanosomatid parasites (Bodyl et al. 2010) that has recently been presented, has nourished the discussion of whether these lineages have evolved from a photosynthetic, plastid-bearing ancestor (Janouskovec et al. 2010). This scenario would imply that the loss of photosynthesis genes and the reduction of other plastome features could be just one intermediate step and that there are no evolutionary restrictions that would preclude the total loss of the plastid genome or the entire plastid. It is surprising, however, that so far no species has been found where the plastid DNA but not the plastid compartment as a whole were lost.

6 Conclusion and Perspectives

Analyses of nucleotide substitution rates have revealed that the mutation rates in plastid genomes are considerably lower than in their nuclear counterparts (Wolfe et al. 1987). The hypothesis that the selective pressure exerted by the photoautotrophic lifestyle has contributed significantly to this conservation of the plastid genome can best be tested by analyzing species that have evolved under a different type or level of selective constraint. Such species are present in the various groups of parasitic plants and algae. In the “omics” age, tools for high-throughput analysis and annotation of genomes are not only available, they are, more importantly, also affordable and require very small amounts of plant material. Consequently, not only the number of published chloroplast genomes from agriculturally important plants have increased considerably over the last 5 years, but also the number of genomes from “cryptic” plastids of parasitic species. While this information has been instrumental in getting an insight into the evolution of plastid genomes, a number of questions await answers in the future. Among those is the extent of a possible nuclear transfer of genes that are regarded as essential and that are as of now reported missing from the plastid genome. Another question concerns the coordination of cryptic plastid and nuclear gene expression in a nonphotosynthetic setting. To answer these questions, it will be necessary to concentrate on the nuclear genomes of some of these species in the future.