Nucleic-acid-templated synthesis as a model system for ancient translation Christopher T Calderone and David R Liu  The translation of nucleic acids into synthetic structures with expanded functional potential has been the subject of considerable research, with applications including small-molecule and polymer evolution, reaction discovery and sensing. Here, we review properties of nucleic-acid-templated synthesis in the context of requirements for prebiotic translation. This analysis highlights the chemical possibilities of ancient translation systems, as well as challenges that these systems may have faced. Addresses Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, Cambridge, Massachusetts 02138, USA  e-mail: drliu@fas.harvard.edu Current Opinion in Chemical Biology 2004, 8:645–653 This review comes from a themed section on Model systems Edited by David G Lynn and Nicholas V Hud Available online 18th October 2004 1367-5931/$ – see front matter # 2004 Elsevier Ltd. All rights reserved. DOI 10.1016/j.cbpa.2004.09.003 Introduction For half a century, researchers have appreciated that proteins are translated from information-bearing nucleic acids, which, in turn, are amplified by replication [1]. In the modern world, a network of highly complex macro- molecular assemblies mediates these processes of transla- tion and replication. We assume that before the advent of modern translation and replication machinery, informa- tion still flowed from replicable carriers to functional molecules to allow the evolution of living systems with diverse chemical capabilities. To better understand the requirements for the prebiotic replication of nucleic acids, researchers began to study the non-enzymatic, template- directed replication of nucleic acids as early as 1966 [2]. More recently, attention has turned to the non-enzymatic translation of nucleic acids into unrelated structures of greater functional potential. Although these recent efforts typically do not operate under reaction conditions chosen to mimic the prebiotic world, they nevertheless provide insights into the potential of nucleic acids to mediate translation without the aid of the modern translational apparatus. Regardless of its context, translation in the most abstract form involves four events: first, readout (in the biotic world, base pairing between an aminoacyl-tRNA anti- codon and a strand of mRNA); second, chemical reaction (acyl transfer from a peptidyl-tRNA to an aminoacyl- tRNA); third, elongation (iteration of template-directed reactions enabled by ribosomal translocation); and fourth, termination (hydrolytic release of the product polypep- tide with subsequent dissociation of the mRNA-ribosome complex). Analogs for each of these events have been achieved using nucleic-acid-templated synthesis in the absence of enzymes, and examples of each are presented in this article. Readout In the modern world, protein-encoding information is read by the adjacent annealing of three-base tRNA anti- codons with mRNA codons, facilitating an efficient and sequence-specific coupling reaction [3]. Although the contiguous annealing of codons made of three consecu- tive nucleotides is the outcome of hundreds of millions of years of evolution, it is unclear a priori whether this strategy is uniquely effective for transferring information from nucleic acids to corresponding functional molecules. Early research on alternative nucleic acid readout systems used information-carrying units (codons) as small as one nucleotide. When used to direct templated synthesis, however, mononucleotides often resulted in modest yields and poor sequence selectivities, possibly due to the unstable and transient nature of mononucleotide base pairing [4]. More recent work has focused on codons of four to ten bases or more. As a result of this change, the efficiencies of nucleic-acid-templated syntheses in- creased dramatically. For example, increasing codon length from three to eight nucleotides allowed Mattes and Seitz to reduce substrate concentrations by a factor of 320 in a DNA-templated amine acylation reaction, while maintaining acceptable reaction efficiencies [5]. Gartner and Liu observed that the ability of a given arrangement of codons to support DNA-templated synth- esis depends on the nature of the templated chemical reaction [6]. In some cases, reactive groups could be separated by dozens of nucleotides yet still react effi- ciently, while in other cases, precise codon alignment was required. These observations led to a kinetic model for DNA-templated synthesis in which the relative rates of annealing and reaction determine the ability of a given arrangement of codons to support reaction (Figure 1) [7  ]. www.sciencedirect.com Current Opinion in Chemical Biology 2004, 8:645–653