Many tRNA UAALeu genes from plastids contain a group I intron. An intron is also inserted in the same gene at the same position in cyanobacteria, the bacterial progenitors of plastids, suggesting an ancient bacterial origin for this intron. A group I intron has also been found in the tRNA fMet gene of some cyanobacteria but not in plastids, suggesting a more recent origin for this intron. In this study, we investigate the phylogenetic distributions of the two introns among cyanobacteria, from the earliest branching to the more derived species. The phylogenetic distribution of the tRNA UAA Leu intron follows the clustering of rRNA sequences, being either absent or present in clades of closely related species, with only one exception in the Pseudanabaena group. Our data support the notion that the tRNA UAA Leu intron was inherited by cyanobacteria and plastids through a common ancestor. Conversely, the tRNA fMet intron has a sporadic distribution, implying that many gains and losses occurred during cyanobacterial evolution. Interestingly, a phylogenetic tree inferred from intronic sequences clearly separates the different tRNA introns, suggesting that each family has its own evolutionary history.Ever since their discovery, the origin of introns has been a subject of controversy. One view, the introns-late hypothesis, proposes that introns are recent invaders and that split genes arose by late insertion of introns into originally uninterrupted genes (28). In that scenario, horizontal transfer and transposition of introns are frequent events, accounting for the scattered phylogenetic distribution of introns. Although the debate has focused on spliceosomal introns, such a scenario could apply as well to other types of introns, some of which are known to be mobile (22). In contrast, the introns-early view implies that introns are very ancient, being present in the progenote (universal ancestor) (7). The demonstration that some members of group I and group II introns are capable of in vitro autocatalytic activity (19,29,39) lends further support to the presence of these introns at an early stage of evolution, maybe as early as the putative precellular RNA world (13). In such a scenario, the observed phylogenetic distribution of introns could be explained by multiple losses in different lineages during evolution (7) and by their mobility, which is assumed to be a derived feature (2). A major obstacle for the introns-early hypothesis was the apparent absence of introns in eubacteria, although this was tentatively rationalized by pressure to streamline the genome in rapidly dividing bacteria (7). Discovery of group I introns in bacteriophages of gram-positive and gram-negative bacteria did not help to resolve the issue, due to uncertainties concerning the origin of the bacteriophages themselves (see discussion in reference 35). The recent discovery of both group I and group II introns in divergent eubacteria (4,11,12,20,31,44) was acclaimed as a breakthrough by introns-early proponents. In most cases, however, the relatio...