The origin of plastids and their spread via secondary symbiosis

Abstract

The endosymbiotic, cyanobacterial nature of plastids is clearly established, but several fundamental issues concerning the origin and early evolution of plastids remain unresolved. One key question is whether plastids are monophyletic (derived from a single cyanobacterial ancestor) or polyphyletic (derived from more than one ancestor). This issue is complicated by the presence in many photosynthetic eukaryotes of secondary plastids, acquired by ingestion of a eukaryote, itself already equipped with plastids, rather than by direct ingestion of a free-living cyanobacterium. A review of the phylogenetic evidence from plastid genes indicates that the three major lineages of primary plastids (red, green, and glaucocystophyte) are probably monophyletic. Mitochondrial data further support this conclusion for red and green plastids (but are unavailable for glaucocystophytes), while nuclear data are largely unresolved. If plastids are monophyletic, then the pigment diversity of plastids must postdate their status as endosymbiotic organelles, but whether this diversity arose primarily by acquisition or loss is nuclear. Secondary endosymbiosis has greatly multiplied the variety of photosynthetic eukaryotes. A secondary origin of plastids is unequivocal for cryptomonads and chlorarachniophytes, is likely for heterokonts, haptophytes, and euglenophytes, and is suggested for the nonphotosynthetic parasites of phylum Apicomplexa. The remarkable plastid diversity of dinoflagellates appears to be the result of multiple secondary and tertiary endosymbiotic events. A consistent feature of all plastid genomes is extreme reduction relative to their cyanobacterial progenitors via outright gene loss, transfer of genes to the nuclear genome, and substitution by genes of nuclear ancestry. Most of this reduction seems to have occurred relatively soon after primary endosymbiosis, before the emergence of the major lineages of plastids, yet recent data also reveal surprising diversity of gene content among these lineages. The rubisco genes (rbcLS) of primary plastids on the red lineage are not related to those of cyanobacteria and seem to have been acquired via horizontal gene transfer.