tRNA Creation by Hairpin Duplication Jeremy Widmann, 1 Massimo Di Giulio, 2 Michael Yarus, 3 Rob Knight 1 1 Department of Chemistry and Biochemistry, University of Colorado, Boulder, CO 80309, USA 2 International Institute of Genetics and Biophysics, CNR, Via G. Marconi 10, Naples 80125, Italy 3 Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, CO 80309, USA Received: 4 November 2004 / Accepted: 3 May 2005 [ReviewingEditor: Dr. Niles Lehman] Abstract. Many studies have suggested that the modern cloverleaf structure of tRNA may have arisen through duplication of a primordial hairpin, but the timing of this duplication event has been unclear. Here we measure the level of sequence identity between the two halves of each of a large sample of tRNAs and compare this level to that of chimeric tRNAs constructed either within or between groups defined by phylogeny and/or specificity. We find that actual tRNAs have significantly more mat- ches between the two halves than do random se- quences that can form the tRNA structure, but there is no difference in the average level of matching be- tween the two halves of an individual tRNA and the average level of matching between the two halves of the chimeric tRNAs in any of the sets we constructed. These results support the hypothesis that the modern tRNA cloverleaf arose from a single hairpin dupli- cation prior to the divergence of modern tRNA specificities and the three domains of life. Key words: tRNA — Hairpin duplication — Clo- verleaf Introduction Many lines of evidence suggest that the two halves of tRNA may be evolutionarily distinct. For example, the ‘‘operational code’’ that links amino acids to tRNAs depends only on the acceptor stem for certain amino acids (Schimmel and Henderson 1994), and aminoacyl tRNA synthetases can even charge mini- helices that resemble only one half of the tRNA molecule (Tamura and Schimmel 2001). These charged minihelix structures have been shown to function in peptide synthesis and may have been part of the primordial protein synthesis machinery (Dick and Shamel 1995). It has also been suggested that the top half of modern tRNAs has an ancient origin in replication and is recognized separately by RNaseP, the CCA-adding enzyme, telomerase, and aminoacyl- tRNA synthetases (Weiner and Maizels 1987; Maizels and Weiner 1994). The 3¢ half of modern tRNAs has been proposed to be older than the 5¢ half due to its base composition and repetitive sequence patterns (Eigen and Winkler-Oswatitsch 1981). More recently, it has been shown that the archaeon Nano- archaeum equitans can create functional tRNAs from the 3¢ and 5¢ tRNA halves, which are encoded by different loci and trans-spliced to form the final product (Randau et. al. 2005). Similarities between nucleotides at comparable positions within the two halves of the tRNA molecule have often been taken as evidence that the modern cloverleaf structure arose through direct duplication of a hairpin (Jukes 1995; Di Giulio 1995 and refer- ences cited therein). If this duplication theory is cor- rect, corresponding positions in the two halves of each modern tRNA molecule should match more than chance predicts (i.e., should have greater se- quence identity). Additionally, the halves of different tRNA molecules should match to greater or lesser extents depending on how many tRNA-creating Correspondence to: Rob Knight; email: rob@spot.colorado.edu J Mol Evol (2005) 61:524–530 DOI: 10.1007/s00239-004-0315-1