Euglenozoan kleptoplasty illuminates the early evolution of photoendosymbiosis

**Background:** The endosymbiotic origin of plastids is well established, but the early evolutionary steps leading to permanent plastids remain debated. Two main models exist: the integration model, where a heterotrophic cell ingests an alga and retains it as an endosymbiont, and the targeting-ratchet model, where protein-targeting systems evolve early and drive permanent integration. Rapaza viridis, described in 2012 as the only known mixotrophic euglenid and the closest sister taxon to the photosynthetic Euglenophyceae, was hypothesized to possess intermediate traits between phototrophic and phagotrophic euglenids. This study investigated whether R. viridis has permanent plastids or transient kleptoplasts and examined the molecular mechanisms underlying plastid retention.

**Methods:** The authors used time-course transmission electron microscopy (TEM) and light microscopy to observe the phagocytotic process and plastid fate in R. viridis after ingestion of Tetraselmis sp. Plastid 16S rRNA genes were sequenced using Sanger sequencing, and whole plastid genomes were assembled from Illumina reads from starved R. viridis cultures and Tetraselmis sp. cultures. Transcriptomes of R. viridis and Tetraselmis sp. were sequenced and assembled. Kleptoplast-targeted proteins were predicted by searching for bipartite targeting signals (signal peptide + transit peptide) using SignalP, PrediSI, and ChloroP, and by homology to the plastid proteome of Euglena gracilis. Phylogenetic analyses of 147 candidate proteins were performed using maximum likelihood with 1,000 bootstrap replicates in IQ-TREE. Polysaccharide accumulation was quantified using the phenol-sulfuric acid assay, and 13C isotope labeling experiments with NaH13CO3 were conducted to confirm photosynthetic carbon fixation.

**Key Results:** 1) R. viridis lacks canonical plastids. Plastid 16S rRNA sequences and whole plastid genomes from starved R. viridis cultures were identical to Tetraselmis sp., not to Pyramimonas or Euglenophyceae. 2) TEM time-course observations showed that within 30 minutes of ingestion, the Tetraselmis plastid is segregated from other cellular material, which is excreted. The plastid is subdivided into ≤10 small pieces within 12-18 hours, forming a three-membrane-bounded envelope typical of the Rapaza-type plastid. Starch grains from the original Tetraselmis plastid gradually disappear. 3) R. viridis undergoes 3-4 rounds of cell division after a single ingestion event, with a 10-fold increase in cell number over 3 weeks. Chlorophyll per cell decreases with division, indicating no net increase in photosynthetic capacity after ingestion. 4) Polysaccharide grains accumulate in the cytoplasm during the light period and decrease in the dark. 13C isotope labeling showed extensive incorporation of 13C into polysaccharide grains, confirming they are photosynthetic products from kleptoplasts. No polysaccharide accumulation occurred in continuous darkness. 5) R. viridis can survive no longer than 35 days without acquiring new kleptoplasts. 6) Transcriptome analysis identified 274 putative kleptoplast-targeted proteins: 166 class I (bipartite signal sequences with two transmembrane domains), 73 class II (signal peptide only with one transmembrane domain), and 35 transporters. In contrast, the heterotrophic euglenid Rhabdomonas costata yielded only 63 candidate proteins, of which only one had a hit to a plastid-targeted protein in E. gracilis, confirming specificity. 7) Components of the translocon complex were identified: Tic21 (shared with Euglenophyceae), Tic20 (related to red-lineage), and Tic110 (related to green-lineage). 8) Phylogenetic analyses revealed diverse origins of kleptoplast-targeted proteins: some from Tetraselmis (e.g., chaperonin 60 beta subunit), some shared with Euglenophyceae (e.g., RCA), and approximately 30% from red algal lineages. 9) The R. viridis transcriptome was 87% complete (BUSCO), and the Tetraselmis sp. transcriptome was 83.9% complete.

**Clinical Implications:** While this study does not have direct clinical applications, it provides fundamental insights into the evolutionary mechanisms of endosymbiosis and organellogenesis. Understanding how protein-targeting systems evolve before permanent organelle establishment has implications for synthetic biology and for understanding the evolutionary origins of eukaryotic cells. The targeting-ratchet model supported by this work suggests that horizontal gene transfer and the evolution of protein import machinery can precede the permanent integration of endosymbionts, which may inform efforts to engineer artificial endosymbiotic relationships.