Exploring the Genotype/Phenotype Landscape of Self-Assembling Modules in RNA

Abstract

GNRA tetraloops (N for base, R for purine) in tertiary interaction with distant helices or receptors are among the most widespread self-assembling modules identified in functional RNAs in nature. A remarkable example is the GAAA/11nt receptor motif, which displays a strong binding affinity and selectivity for its cognate tetraloop. However, in vitro evolution has yielded novel receptors with binding affinities and selectivity for GNRA tetraloops similar to that of the GAAA/11nt receptor, raising the question why only a limited subset of these interactions occurs in nature. It has been proposed that the natural evolution of GNRA receptors may incur a selection pressure to favor sequences that display a high degree of mutational robustness, meaning that single point mutations to their sequence do not significantly alter their self-assembly function. To this end, we have performed a comprehensive mutational analysis of the 11nt motif and two artificial receptors isolated by in vitro selection and shown to recognize their cognate tetraloops with binding affinity comparable to that of the GAAA-11nt interaction. Using a tectoRNA system, a comprehensive genetic/phenotypic landscape was derived for each of these interactions by screening their variants for the ability to bind GNRA tetraloops. The results indicate that both natural and artificial receptors maintain good binding affinity for their cognate tetraloop (GAAA-11nt; GUAA-R5.58; R1-GGAA) upon single point mutation. However, the artificial receptor R5.58 displays a capacity to change its phenotype in terms of its affinity and selectivity toward a particular GNRA tetraloop, which may indicate a greater degree of plasticity and evolvability through single point mutation. Artificially selected receptors may once have occurred in nature, as evolutionary intermediates on the way to optimal solutions, like the 11nt motif. Further experiments are planned to validate this hypothesis.