Yield and fidelity of early steps of ribosome-free translation depend on the RNA anchoring

Before the translational machinery came into being, a simpler form of reading RNA sequences to instruct peptide synthesis must have existed. What this earliest form of translation was is unclear. Ribosome-free synthesis, relying solely on Watson-Crick base pairing and chemical reactivity, is a likely candidate. The established version of this process involves chain growth at the C-terminus of N-terminally anchored peptides. Here, we show that ribosome- and enzyme-free translation via chain growth at the N-terminus of C-terminally anchored peptides is also feasible. Yields for the formation of dipeptides via the latter process were between 6% and 46%, though, as compared to 60%-99% for dipeptides with opposite strand growth orientation. Translation fidelity to dipeptides was also lower, with just 55%-84% correct incorporation, versus 64%-92% for N-terminally linked dipeptides. On the tripeptide level, the preference for C-terminal chain growth was smaller, suggesting that only the very first step has a strong bias for N-terminal anchoring. Finally, mixed anhydride pre-activation was found to be an alternative to in situ activation for peptidyl RNAs. This data sheds light on putative early steps in the evolution of translation.