A phylogenetic mosaic plastid proteome and unusual plastid-targeting signals in the green-colored dinoflagellate Lepidodinium chlorophorum

Abstract

Background: Plastid replacements through secondary endosymbioses include massive transfer of genes from the endosymbiont to the host nucleus and require a new targeting system to enable transport of the plastid-targeted proteins across 3-4 plastid membranes. The dinoflagellates are the only eukaryotic lineage that has been shown to have undergone several plastid replacement events, and this group is thus highly relevant for studying the processes involved in plastid evolution. In this study, we analyzed the phylogenetic origin and N-terminal extensions of plastid-targeted proteins from Lepidodinium chlorophorum, a member of the only dinoflagellate genus that harbors a green secondary plastid rather than the red algal-derived, peridinin-containing plastid usually found in photosynthetic dinoflagellates.

Results: We sequenced 4,746 randomly picked clones from a L. chlorophorum cDNA library. 22 of the assembled genes were identified as genes encoding proteins functioning in plastids. Some of these were of green algal origin. This confirms that genes have been transferred from the plastid to the host nucleus of L. chlorophorum and indicates that the plastid is fully integrated as an organelle in the host. Other nuclear-encoded plastid-targeted protein genes, however, are clearly not of green algal origin, but have been derived from a number of different algal groups, including dinoflagellates, streptophytes, heterokonts, and red algae. The characteristics of N-terminal plastid-targeting peptides of all of these genes are substantially different from those found in peridinin-containing dinoflagellates and green algae.

Conclusions: L. chlorophorum expresses plastid-targeted proteins with a range of different origins, which probably arose through endosymbiotic gene transfer (EGT) and horizontal gene transfer (HGT). The N-terminal extension of the genes is different from the extensions found in green alga and other dinoflagellates (peridinin- and haptophyte plastids). These modifications have likely enabled the mosaic proteome of L. chlorophorum.

Figure 1

Genes of green algal origin. Maximum likelihood trees inferring a green-algal origin of L. chlorophorum of PsbO, RuBisCO activase, PsbP, PsbR and PAP protein kinase. The phylogenies were inferred using RAxML. Bootstrap values >50%are indicated on the branches. Green and red lineages are indicated by color. Secondary green algae are in bold. Filled dots indicate 100% bootstrap support. Bayesian posterior probability values are indicated for some of the most important splits.

Figure 2

Genes of peridinin- and heterokont origin. Maximum likelihood trees of 1-deoxy-D-xylolose-5 phosphate reductoisomerase (Dvr1), ferredoxin NADP oxireductase, 4-disphophocytidyl-2C-methyl-D-erythriol kinase demonstrating that these genes are of peridinin dinoflagellate origin. Csp41 clusters with heterokonts. All trees were inferred using RAxML. Bootstrap values >50%are indicated on the branches. Green and red lineages are indicated by color and secondary green algae are in bold. Filled dots indicate 100% bootstrap support. Bayesian posterior probability values are indicated for some of the most important splits.

Figure 3

Phylogenies supported by shared sequence characteristics. A: Maximum likelihood trees inferring a heterokont origin of RuBisCO small subunit and phosphoribulokinase in L. chlorophorum. B: Alignment of the genes demonstrating shared sequence characteristics between L. chlorophorum and streptophytes and heterokonts, respectively. Filled dots indicate 100% bootstrap support. Bayesian posterior probability values are indicated for some of the most important splits.