A new lineage of non-photosynthetic green algae with extreme organellar genomes

Background: The plastid genomes of the green algal order Chlamydomonadales tend to expand their non-coding regions, but this phenomenon is poorly understood. Here we shed new light on organellar genome evolution in Chlamydomonadales by studying a previously unknown non-photosynthetic lineage. We established cultures of two new Polytoma-like flagellates, defined their basic characteristics and phylogenetic position, and obtained complete organellar genome sequences and a transcriptome assembly for one of them. Results: We discovered a novel deeply diverged chlamydomonadalean lineage that has no close photosynthetic relatives and represents an independent case of photosynthesis loss. To accommodate these organisms, we establish a new genus, Leontynka, with two species L. pallida and L. elongata distinguished by both morphological and molecular characteristics. Notable features of the colourless plastid of L. pallida deduced from the plastid genome (plastome) sequence and transcriptome assembly include the retention of ATP synthase, thylakoid-associated proteins, carotenoid biosynthesis pathway, and plastoquinone-based electron transport chain, the latter two modules having an obvious functional link to the eyespot present in Leontynka. Most strikingly, the L. pallida plastome with its ~362 kbp is by far the largest among non-photosynthetic eukaryotes investigated to date. Instead of a high gene content, its size reflects extreme proliferation of sequence repeats. These are present also in coding sequences, with one repeat type found in exons of 11 out of 34 protein-coding genes and up to 36 copies per gene, affecting thus the encoded proteins. The mitochondrial genome of L. pallida is likewise exceptionally large, with its >104 kbp surpassed only by the mitogenome of Haematococcus lacustris among all members of Chlamydomonadales studied so far. It is also bloated with repeats, yet completely different from those in the L. pallida plastome, which contrasts with the situation in H. lacustris where both organellar genomes have accumulated related repeats. Furthermore, the L. pallida mitogenome exhibits an extremely high GC content in both coding and non-coding regions and, strikingly, a high number of predicted G-quadruplexes. Conclusions: With the unprecedented combination of plastid and mitochondrial genome characteristics, Leontynka pushes the frontiers of organellar genome diversity and becomes an interesting model for studying organellar genome evolution.