Non-coding RNAs (ncRNAs) have emerged as abundant critical elements of the cellular machinery that are increasingly being targeted in drug discovery efforts. High-resolution 3D structure determination of ncRNAs provides the basis for understanding their functional mechanisms at the atomic level and for implementing structure-based approaches for drug discovery. However, RNA's unique characteristics continue to pose significant challenges to high-resolution structure determination by X-ray crystallography and NMR spectroscopy. The global structures of many ncRNAs are highly flexible and can therefore resist crystallization. Many ncRNAs have molecular weights that exceed the NMR limit of application (<100 nucleotides). This has made it necessary to excise individual RNA domains, suitable for characterization by X-ray and NMR, from their much larger context. This 'divide and conquer strategy'is often called into question by experiments showing that domains have overlapping functions and potentially associate to form higher order structures. By contrast, advances in computational and experimental methods are allowing determination of secondary structures for increasingly large and complex RNAs. Recently, we showed that topological constraints provide the missing link between RNA secondary structure and 3D global and dynamic adaptation. In this proposal, we propose to develop new computational methods that exploit these newly founded topological constraints to define global features of RNA structure based on secondary structure alone. By combining these global constraints with footprinting data and a new NMR chemical shift fingerprinting strategy for identifying tertiary motifs, we propose to develop a new paradigm for determining the 3D structural organization of large flexible RNAs. The methodology will be validated by determining the 3D conformation of the guanine sensing riboswitch aptamer domain in its stable ligand bound form and subsequently used to characterize the more flexible conformation of the domain in its ligand free form. Results will be used to test the hypothesis that topological constraints encoded in the unique three-way junction, and not long range loop-loop tertiary interactions, define the global structure of the aptamer domain giving rise to spatially tuned dynamics that are optimized for adaptive ligand binding.
The proposed research will test the feasibility of developing a novel computational NMR method for determining structures of very large RNAs under solution conditions. The method obviates the need for crystallization or sharp NMR spectra, can be applied to RNA structures as large as 1000 nucleotides, and is expected to increase throughput over conventional methods by one-to-two orders of magnitude. This new paradigm for RNA structure determination will bring into the realm of application several modes of investigation that are currently impossible due to limitations in X-ray and NMR.
