The relationship between enzyme architecture and substrate specificity among archaeal pre-tRNA splicing endonucleases has been investigated more deeply, by using biochemical assays and model building. The enzyme from Archeoglobus fulgidus (AF) is particularly interesting: it cleaves the bulge-helix-bulge target without requiring the mature tRNA domain, but, when the target is a bulge-helix-loop, the mature domain is required. A model of AF based on its electrostatic potential shows three polar patches interacting with the pre-tRNA substrate. A simple deletion mutant of the AF endonuclease lacking two of the three polar patches no longer cleaves the bulge-helix-loop substrate with or without the mature domain. This single deletion shows a possible path for the evolution of eukaryal splicing endonucleases from the archaeal enzyme. molecular evolution ͉ RNA-protein interactions ͉ tRNA endonucleases A ccuracy in tRNA splicing is essential for the formation of functional tRNAs and therefore for gene expression. In Bacteria, pre-tRNA introns are self-splicing group I introns, and the splicing mechanism is autocatalytic. In Eukarya, tRNA introns are small and invariably interrupt the anticodon loop one base 3Ј to the anticodon. In Archaea, the introns are also small and often reside in the same location as eukaryal tRNA introns, but not always (1, 2). In both Eukarya and Archaea, the specificity for recognition of the pre-tRNA resides in the endonucleases (3-5). These enzymes remove the intron by making two independent endonucleolytic cleavages. It is generally accepted that the archaeal enzymes act without any reference to the mature domain but instead recognize specific structures that define the intron-exon boundaries (6-8). By contrast, the eukaryal enzyme normally acts in a mature-domain-dependent mode (3, 4).The exact way in which the eukaryal endonucleases recognize the precursors has yet to be determined, but many RNA-protein interactions are presumably required. Previous studies have shown the role of the three-dimensional structure and of specific invariant bases of the mature domain in recognition and binding by the enzyme (3, 4). The importance of structural regularities in the precursors suggested an anchor-and-measure mechanism in which the endonuclease binds to one or more reference sites in the mature domain, which are common to all pre-tRNAs, and measures the distance to the equivalently positioned intron-exon junctions. This hypothesis was supported by experiments that involved the engineering of changes in the distance that separate the mature domain from the splice sites. These alterations changed the size of the intron in a predictable way: the insertion of 1 bp in the anticodon stem increased the size of the intron by two bases, one at each end, and the insertion of 2 bp in the anticodon stem increased the size of the intron by 4 nt (3, 4). These striking results, which had been observed in specific instances, were generalized and taken to mean that there are no important recognition elements at t...