A maturase-encoding group III twintron is conserved in deeply rooted euglenoid species: Are group III introns the chicken or the egg?

The fourth intron of the Euglena gracilis chloroplast photosystem II gene, psbCi4, is a 1,605-bp twintron composed of two group III introns and a coding locus for a 458-aa polypeptide, matI, located in the internal intron, psbCi4 homologs have been identified in seven euglenoids, including E. myxocylindracea, E. viridis, E. deses, E. pisciformis, Cryptoglena pigra, Eutreptia sp., and Lepocinclis beutschlii. All of the species examined contain both the group III twintron and the mart locus, revealing a more widespread occurrence of group III introns than previously known. The L. beutschlii mat1 locus is interrupted by two novel mini-group II introns of 224 and 258 nt, the smallest group II introns yet identified. Reverse transcriptase polymerase chain reaction analysis confirmed the splicing boundaries of the external and internal E. myxocylindracea, E. viridis, and E. deses introns as well as the novel L. beutschlii mat1 introns. As determined by comparative phylogenetic analysis, group III introns contain a structural homolog of group II intron domain VI. The mat1 loci encode peptide motifs characteristic of group II intron maturases. A group III intron-encoded protein whose predicted sequence is similar to group II intron-encoded maturases and a bona fide domain VI within group III introns are compelling evidence for a common ancestor of group II and group III introns.