The Unexpected Base‐Pairing Behavior of Cyanuric Acid in RNA and Ribose versus Cyanuric Acid Induced Helicene Assembly of Nucleic Acids: Implications for the Pre‐RNA Paradigm

Abstract

The cyanurica cid (CA) heterocycle forms supramolecular structures with adenine nucleobases/nucleosides and oligonucleotides,l eading to speculation that they can act as forerunners to RNA. Herein, the assembly behavior of RNA containing CA and CA-ribose nucleoside was studied. Contrary to previous reports, CA in RNA and the CA-ribonucleoside resulted in destabilization of supramolecular assemblies, which led to ar eevaluation of the CA-adenine hexameric rosette structure. An unprecedented noncovalent supramolecular helicene structure is proposed to account for the striking difference in behavior,which has implications for novel paradigms for reorganizing the structures of nucleic acids, the synthesis of long helicenes, and pre-RNA world paradigms. The results caution against extrapolating the self-assembly behavior of individual heterocycles from the level of monomers to oligomers because the baseparing properties of (non-)canonical nucleobases are impacted by the type of oligomeric backbone to which they are attached.

Noncovalent Helicene Structure between Nucleic Acids and Cyanuric Acid

“…Self-assembly between DNA and CARC requires ah igh minimum assembly concentrationo fC ARC, approximately seven times greatert han the required CA concentration for the DNA-CA assemblies, and the melting temperature of the DNA-CARC assemblies is less than 15 8C. [33] Our simulations showamore pronounced effect of CARC on the stabilityo ft he RNA assemblies (Simulations1 1-14) than on the DNA assemblies (Simulations 9-10).…”

“…[29] Following this work, Krishnamurthy and co-workers examined the base-pairing properties of oligonucleotides containing CA along with the ribose-substituted CA (CARC). [33] Their observations, however, suggested ah ydrogen-bonded helicene as an alternative model. Our MD simulations indicatet hat the proposed planar hexad modelf or the structure of the assembly is unlikely and that the assemblies adopt an unprecedented noncovalent helicene geometry for both the DNA andR NA systems.…”

“…Our helicene model can rationalize the observed instability of the assemblies in the presence of ribose-substituted CA as observed by Krishnamurthy and co-workers. [33] The ribose-phosphate backbones hows ac lose association with the exposed face of CA and thusr ibose substituentso nC As terically deform the nucleic acid backbone, leadingt os tructurali nstability.O ur simulations show that the A-form conformation is compatible with the presence of three (the "helicene-triad model") and two (the "helicene-dyad model")n ucleic acid strandsi nt he assembly.I nc ontrast to the triad model,t he dyad model leads to structures with minimal deviation from the reportedf iber diffraction geometry forc anonical nucleic acidd uplexes [39] and improves the structuralu niformity of the RNA strands. It is possible that both helicene models are accessible conformationsu nder varying experimental conditions.…”

“…Having established the greater likelihood of the helicene model over the hexad model for the structure of the DNA and RNA assemblies with CA, we were curious to explorew hether the originally proposed helicene model by Krishnamurthy and co-workersw ith two nucleic acid strands (i.e.,t he "helicenedyad model") [33] instead of three is plausible. Our previous modeling of the assemblies between three nucleic acid strands and CA showedt hat the strandsh aves ubstantial slides and display the characteristich ole in the center of the assembly, which are features of the A-form conformation of canonical RNA and DNA.…”

“…[29] Based on the work of Sleiman and co-workers [29] and others, [18,20,23,32] Krishnamurthy and co-workersi nvestigated the base-pairing properties of cyanuric acid ribose conjugate (CARC) nucleosidea nd CA-containing oligonucleotides, the results of which led them to proposeahydrogen-bonded helicene structure as an alternative to the hexamericr osette model. [33] In this article, we provides upport for this alternative model for the structure of the poly(A)-CA assemblies based on computational corroboration. We show that the assemblies adoptanoncovalent, twisted ribbon structure (the "helicene model")i nw hich adenine and CA form ac ontinuous hydrogen-bonding network with intrinsic helicity,w hich is driven by the sugar-phosphate backbone.…”