Nonhydrolyzable Ribonucleotide Phosphoroimidazolide Analogues for Mechanistic Studies of Nonenzymatic RNA Replication

Abstract

Efficient and faithful replication of genomic information is perhaps one of the most important hallmarks of cellular life. At some early point, such processes must have occurred in the absence of elaborate nucleic acid or protein-based enzymatic catalysts. Template-directed nonenzymatic RNA primer extension represents an important experimental model which aims to illustrate how functional ribonucleotide sequences could have transferred their genetic contents to their offspring. Prototypical nonenzymatic RNA primer extension experiments involve imidazole-activated monoribonucleotides; these activated monomers bind reversibly to an RNA template and polymerize to form the complementary strand, and such processes are catalyzed by a high concentration of divalent metal cations. In Ch. 1, I review recent chemical strategies that enable nonenzymatic RNA copying reactions to proceed efficiently. Extensive efforts have been invested in determining the thermodynamic parameters and the optimal reaction conditions for regioselectivity and fidelity of nonenzymatic RNA polymerizations, yet some mechanistic aspects of this reaction remain elusive. One of the most perplexing features of nonenzymatic RNA replication is that parent strand copying typically does not proceed to completion. Specifically, the copying of the last template nucleotide is dramatically more sluggish than prior nucleotide incorporations. Orgel and coworkers proposed that the leaving group of the activated monomer downstream of the extending monomer plays important catalytic role(s), but the mechanistic origin(s) of this effect remained elusive until recently. In Chapter 2, I detail my synthetic chemistry efforts to develop nonhydrolyzable analogues of 2-methylimidazole-activated monoribonucleotides, and the ensuing RNA crystallographic efforts to ascertain the mechanistic nature of the leaving group-leaving group interactions between consecutively-bound activated monomers. Notably, in these RNA-analogue complexes, noncovalent interactions are not observed, challenging the long-held assumption that noncovalent leaving group-leaving group interactions – presumably pi-pi,cation-pi, or steric in nature – are likely to be responsible for the catalytic effect. An alternate mechanistic proposal, wherein primer extension is driven by highly activated 5′-5′-linked imidazolium-bridged dinucleotides formed by two imidazoleactivated monomers, began to emerge. In Chapter 3, I describe the RNA crystallography efforts, using P1,P3-diguanosine-5′-triphosphate (GpppG) as a stable analogue of the activated imidazolium-bridged dinucleotides, to reveal the possible binding modes of the activated imidazolium-bridged dinucleotides with cognate RNA templates. In chapter 4, I detail the development of guanosine 5ʹ-(4-methylimidazolyl)-phosphonate (ICG) as the closest nonhydrolyzable isosteric analogue of 2-methylimidazole-activated guanosine mononucleotide (2-MeImpG). Using NMR and mass spectrometric techniques, I demonstrate that ICG and activated monomers react to form 5ʹ-5ʹ-linked imidazolium bridged dinucleotides, with the imidazolium bridge flanked by both labile nitrogenphosphorus and stable carbon-phosphorus linkages. Purified mixed-linkage dinucleotides react with cognate RNA primer-template complexes to afford extended primers, but only when the labile nitrogen-phosphorus linkage, and not the nonhydrolyzable carbon-phosphorus linkage, is in close proximity with the upstream primer 3′-hydroxyl group. These analogue-based model systems strongly bolster the hypothesis that the 5ʹ-5ʹ-linked imidazolium-bridged dinucleotides are important intermediates in primer extensions driven by imidazole-activated monoribonucleotides. The presence of activated “helper oligonucleotides” that bind complementarily downstream of the extending monomer encourages facile RNA copying with high fidelity, but the thermodynamic and/or kinetic origins of this phenomenon have not been well elucidated. In Chapter 5, I quantitatively demonstrate that complementary helper oligonucleotides that bind downstream of the extending dramatically enhance monomer binding: if the monomer is GMP and the downstream helper has a 5¢-G terminus, GMP binding is enhanced 100-fold. Nonetheless, unactivated helper oligomers only modestly accelerate upstream primer extensions, by ~10 fold, thus suggesting that the primerextension rate enhancement conferred by activated helper oligonucleotides must be a combined effect of thermodynamic (enhanced monomer binding affinity) and kinetic (leaving group-leaving group interaction) effects. In Ch.6, using the approach of X-ray crystallography, I investigate the potential structural and conformational effect(s) that activated (and unactivated) helper oligonucleotides could exert on the incoming activated monomer or the overall duplex structure. The helper oligonucleotides utilized in Ch. 6 have either a 5′-ICG cap, a 5′-phosphate or a 5′-hydroxyl group, and the ICG monomer (Ch. 4) is sandwiched between the primer and the helper oligonucleotides. Surprisingly, all three downstream oligonucleotides rigidify the conformation of the bound ICG monomer, forces that monomer to adopt the 3′-endo conformation and to bind to the template exclusively with Watson-Crick geometry, all while aligning the phosphonate group of the sandwiched ICG closer to the primer 3′-hydroxyl group for the subsequent nucleophilic attack. These results suggest that a downstream helper oligonucleotide can catalyze upstream primer extension by placing additional steric and conformational restraints on the incoming monomer relative to the primer. These results afford new insights into the mechanistic underpinnings of nonenzymatic RNA primer extensions, which will in turn aid in the optimization of this reaction to afford longer and more accurately copied RNA sequences in a reasonable timescale.