| Nucleic acid replication via chemically catalyzed monomer polymerization has been shown to be extremely error-prone, an observation that has cast doubt on the plausibility of an ancient RNA world. Alternatively, it has been postulated that oligonucleotide ligation may have been able to overcome the problems inherent in the chemical polymerization of monomers. To assess the fidelity of replication via oligonucleotide ligation, we performed chemical and enzymatic ligation reactions with constant sequence DNA templates and random sequence DNA pools as substrates. Ligation products were isolated, amplified, cloned, and sequenced. For both chemical and enzymatic ligation, fidelity was found to increase with temperature and decrease with the length of the random pools. As was expected, mismatches were often common; however, conditions were found for both methods of ligation in which no mismatches were tolerated. At times alternative structures formed, leading to some minor products with unique sequences. Even so, under the proper conditions, oligonucleotides of defined sequence can be faithfully copied by a chemical catalyst or by a ligase in the absence of polymerases. These studies have important implications for scenarios for the origin of life. In particular, it is likely that self-replicating nucleic acids were originally short, but gained in length and catalytic prowess over time, while the substrates for replication conversely became shorter and simpler. Error-prone but simple oligonucleotide ligases eventually gave way to faithful but complex mononucleotide polymerases.;In order to study the replicability of short oligonucleotides, two sets of fully randomized pools were to be synthesized: one containing a 3... |
