Nonenzymatic assembly of active chimeric ribozymes from aminoacylated RNA oligonucleotides

Aminoacylated transfer RNAs, which harbor a covalent linkage between amino acids and RNA, are a universally conserved feature of life. Because they are essential substrates for ribosomal translation, aminoacylated oligonucleotides must have been present in the RNA world prior to the evolution of the ribosome. One possibility we are exploring is that the aminoacyl ester linkage served another function before being recruited for ribosomal protein synthesis. The nonenzymatic assembly of ribozymes from short RNA oligomers under realistic conditions remains a key challenge in demonstrating a plausible pathway from prebiotic chemistry to the RNA world. Here, we show that aminoacylated RNAs can undergo template-directed assembly into chimeric amino acid–RNA polymers that are active ribozymes. We demonstrate that such chimeric polymers can retain the enzymatic function of their all-RNA counterparts by generating chimeric hammerhead, RNA ligase, and aminoacyl transferase ribozymes. Amino acids with diverse side chains form linkages that are well tolerated within the RNA backbone and, in the case of an aminoacyl transferase, even in its catalytic center, potentially bringing novel functionalities to ribozyme catalysis. Our work suggests that aminoacylation chemistry may have played a role in primordial ribozyme assembly. Increasing the efficiency of this process provides an evolutionary rationale for the emergence of sequence and amino acid–specific aminoacyl-RNA synthetase ribozymes, which could then have generated the substrates for ribosomal protein synthesis.

The evolution of ribosomal protein synthesis would have required the presence of 2′,3′-aminoacylated RNAs, which are the universal substrates for protein synthesis. Without specialized aminoacyl-RNA synthetase enzymes, the initial aminoacylation of the 2′,3′-diol of RNA must have been mediated by spontaneous chemical processes. However, current chemical pathways leading to RNA aminoacylation are inefficient or rely on unstable, activated amino acids. For example, nonenzymatic aminoacylation of the 2′,3′-diol of RNA oligonucleotides with imidazole-activated amino acids affords low levels of aminoacylation (1, 2), primarily due to the pronounced hydrolytic instability of protonated aminoacyl esters (half-life of ca. 50 to 180 min at 22 °C, pH 8, depending on the amino acid residue) (3–5). Alternatively, interstrand transfer of an amino acid from a 5′-phosphate mixed anhydride to the 2′,3′-diol generates aminoacylated oligonucleotides but depends on the efficient synthesis of phospho-carboxy anhydrides, which are prone to hydrolysis and CO₂-mediated degradation (6, 7). Given these limitations, spontaneous synthesis alone may not have generated sufficient aminoacylated RNA substrates to facilitate the emergence of a primitive ribosome. Furthermore, spontaneous synthesis of aminoacylated RNAs is unlikely to have been highly sequence and amino acid specific; this lack of specificity would have made the evolution of coded peptide synthesis problematic. Here, we consider a scenario in which the initial inefficient and minimally specific aminoacylation of RNA played an alternative role that was nonetheless advantageous to primordial protocells. Such a role might have provided a selective pressure favoring the evolution of ribozymes that generated and maintained high levels of RNA aminoacylation. Accordingly, we have sought to identify a role for aminoacylated RNAs that could have been beneficial to primitive protocells, independent of ribosomal protein synthesis.Early life is thought to have used ribozymes to catalyze primordial metabolic reactions as well as RNA replication prior to the evolution of protein enzymes (i.e., the RNA world) (8, 9). In this scenario, the first ribozymes were generated through the nonenzymatic ligation of short RNA oligonucleotides and/or the template-directed polymerization of activated nucleotide monomers. Imidazole and its derivatives have traditionally been used to activate RNA building blocks (10, 11), and recent work has expanded on the prebiotic plausibility of imidazole-activated RNAs (12–15). While the activation of ribonucleotides with imidazoles has facilitated the assembly and templated copying of short RNAs (10, 16–19), oligomers long enough to exhibit enzymatic properties have remained out of reach. A major obstacle to efficient ribozyme assembly is the relatively poor nucleophilicity of the terminal 2′,3′-diol of RNA, mandating high Mg²⁺ concentrations that are not compatible with fatty acid-based vesicle models of protocells (20). Carbodiimide activation of phosphate (21, 22) or replacement of the 3′-hydroxyl with an amino group allow for efficient ligation of RNA oligonucleotides (23–28), but the prebiotic relevance of these approaches remains unclear.In the 1970s, Shim and Orgel demonstrated that glycylated nucleotides can take part in template-directed polymerization with imidazole-activated nucleotides, forming phosphoramidate linkages between the Gly amine and the 5′-phosphate of an adjacent nucleotide (29). Building on this discovery, recent investigations have focused on the possibility of an RNA–amino acid copolymer world (30) and the conditions that permit conjugation of peptido RNAs (31). The Richert group has recently expanded our understanding of peptidoyl (peptide-bridged) RNAs and how they may have played a role in template-directed peptide condensation (31, 32). However, to our knowledge, chimeric, amino acid–bridged ribozymes created with the building blocks of both proteins and RNA have not yet been reported. In our previous work, short aminoacylated RNAs were shown to form amino acid bridges in nonenzymatic ligation reactions with imidazole-activated oligonucleotides at rates much higher than observed with unmodified RNA (5). Here, we extend that approach to show that multiple ligation reactions can generate chimeric amino acid–bridged RNAs long enough to constitute catalytic RNAs. Importantly, we show that these amino acid–bridged RNAs function as catalytically proficient ribozymes. We demonstrate the assembly of three chimeric ribozymes that perform RNA cleavage, RNA ligation, and RNA aminoacylation, all thought to have been important enzymatic functions in the RNA world. We generate the chimeric ribozymes at 2.5 mM Mg²⁺, conditions that are compatible with fatty acid–based vesicles (20) and limit RNA and activated RNA degradation (33) but generally yield no RNA ligation. We further show that aminoacylated RNA assembly on splint templates can generate a chimeric ribozyme that functions in the same pot without purification, thereby highlighting the potential prebiotic plausibility of our chimeric ribozyme assembly method. Our work reveals a link between aminoacylation chemistry and the nonenzymatic assembly of RNA oligonucleotides into functional ribozymes. This functional link provides a potential rationale for the evolution of ribozyme-catalyzed RNA aminoacylation prior to the evolution of ribosomal protein synthesis.