Plastid translocon recycling in dinoflagellates demonstrates the portability of complex plastids between hosts

Abstract

The plastids of photosynthetic organisms on land are predominantly "primary plastids," derived from an ancient endosymbiosis of a cyanobacterium. Conversely, the plastids of marine photosynthetic organisms were mostly gained through subsequent endosymbioses of photosynthetic eukaryotes generating so-called "complex plastids." The plastids of the major eukaryotic lineages-cryptophytes, haptophytes, ochrophytes, dinoflagellates, and apicomplexans-were posited to derive from a single secondary endosymbiosis of a red alga in the "chromalveloate" hypothesis. Subsequent phylogenetic resolution of eukaryotes has shown that separate events of plastid acquisition must have occurred to account for this distribution of plastids. However, the number of such events and the donor organisms for the new plastid endosymbioses are still not resolved. A perceived bottleneck of endosymbiotic plastid gain is the development of protein targeting from the hosts into the new plastids, and this supposition has often driven hypotheses toward minimizing the number of plastid-gain events to explain plastid distribution in eukaryotes. But how plastid-protein-targeting is established for new endosymbionts is often unclear, which makes it difficult to assess the likelihood of plastid transfers between lineages. Here, we show that Kareniaceae dinoflagellates, which possess complex plastids known to be derived from haptophytes, acquired all the necessary protein import machinery from these haptophytes. Furthermore, cryo-electron tomography revealed that no additional membranes were added to the Kareniaceae complex plastid during serial endosymbiosis, suggesting that the haptophyte-derived import processes were sufficient. Our analyses suggest that complex red plastids are preadapted for horizontal transmission, potentially explaining their widespread distribution in algal diversity.

Keywords: algae; dinoflagellate; dinotom; endosymbiosis; evolution; haptophyte; photosynthesis; plastid; secondary plastid; tertiary plastid.

Figure 1. Inner membrane translocons are derived from haptophytes in Kareniaceae

(A) Schematic of four-membrane-bound plastids and the protein translocation machinery conserved in haptophytes, ochrophytes, cryptophytes, and api-complexans. Protein delivery to the ER or endomembrane system (endo) can occur either via vesicular transport or, in some taxa, cotranslational import via ribosomes docked on the outer membrane. The symbiont-specific ERAD-like machinery (SELMA) transfers proteins into the periplastidal space as a relic of the red algal cytoplasm, and TOC and TIC derived from the primary plastid transfer proteins into the plastid stroma. (B–E) Protein maximum likelihood phylogenies of Tic110, Tic20, Tic22, and PPP1, respectively. Sequences found in the Kareniaceae monophyletic clades are indicated with magenta triangles or text. Dinotom sequences of the Kryptoperidiniaceae are shown with yellow triangles. Bootstrap support values >70 are shown and the number of sequences per clade is given in parentheses. Scale bars indicate estimates of amino acid substitutions per site. Full phylogenies are given in Data S1. See also Table S1.

Figure 2. SELMA components Cdc48, Der1, and Uba1 are derived from haptophytes in Kareniaceae

Protein maximum likelihood phylogenies of (A) Cdc48, (B) Der1, and (C) Uba1 shown as for Figure 1. Full phylogenies are given in Data S1. See also Table S1.

Figure 3. SELMA components Ubc4, Ubi, and Hsp70 are derived from haptophytes in Kareniaceae

Protein maximum likelihood phylogenies of SELMA paralogs of (A) the chaperone Hsp70, (B) the ubiquitin conjugating enzyme Ubc4, and (C) ubiquitin (Ubi) are shown as in Figure 1. Full phylogenies are given in Data S1. See also Table S1.

Figure 4. Haptophyte-derived Kareniaceae translocon machinery possess bipartite plastid-targeting presequences

ER-directing signal peptide (SP) predictions by Phobius for 20 Kareniaceae translocation components. A logo plot made using WebLogo of protein pre-sequences aligned on the predicted cleavage site. Alignment of protein presequences with transit peptide-type features (40 residues post-SP cleavage site) is shown for each.

Figure 5. Karlodinium veneficum plastids are surrounded by four membranes

(A) Transmission electron micrograph of a high-pressure frozen, resin-embedded, thin section of the cell periphery showing a plastid (P) within the cytoplasm (Cy). (B) Cryo-electron tomogram showing the membranes separating the plastid thylakoids (Th, black arrowhead) and stroma from the cytoplasm. White dashed boxed region is magnified in (B) where white arrowheads indicate four bounding membranes of the plastid. (C) Z series of five successive virtual sections of the tomogram area shown by the dashed box in (B). The bounding membranes show the outer and inner membrane pairs maintaining an approximately fixed separation distance between each membrane, but variation in the spacing between these pairs is seen (asterisk). Nu, nucleus; M, mitochondrion; Cy, cytoplasm; P, plastid; PM, plasma membrane; Th and black arrowheads, thylakoids.

Figure 6. Presence and absence of plastid translocons in Kareniaceae

Phylogeny of Kareniaceae dinoflagellates and the detection of expressed plastid translocon proteins in each taxon indicated by colored circles. BUSCO scores, as an estimate of transcriptome coverage, are given for each. Bootstrap support values for phylogeny nodes are shown, and the scale bar indicates estimated number of nucleotide substitutions per site.

Figure 7. Model of ancestry and evolution of plastid-protein translocation machinery in Kareniaceae dinoflagellates

Ancestral peridinin plastids of dinoflagellates are surrounded by three membranes, and protein targeting occurs via the ER translocon Sec61 and vesicle delivery. Upon outer membrane fusion, protein cargo then passes through the TOC and TIC complexes. Most known Kareniaceae dinoflagellates inherited new plastids derived from haptophytes and likely adopted the dinoflagellate routing through the ER to the outermost plastid membrane. From here, haptophyte-derived translocons (green), including SELMA, complete protein import. Further plastid replacement in T. helix and K. armiger with alterative haptophyte plastids has maintained the preexisting TICs but eliminated all SELMA components and presumably the equivalent plastid membrane. Gray translocons remain unidentified.