• Nessuna parola chiave trovata

High-resolution solid-state Nuclear Magnetic Resonance (ssNMR) is rapidly emerging as a powerful structural tool in chemistry and biology and is applicable to a wide range of problems that cannot be addressed by solution state NMR or X-ray crystallography. Steady ongoing methodological developments combined with tremendous advances in probe and spectrometer hardware have led to a variety of strategies for resonance assignment, paving the way to the first 3D structure determinations of a wide range of samples at atomic resolution, ranging from inorganic frameworks and catalysts to membrane proteins and fibrils. Solid-state NMR does not suffer from molecular weight limitations (unlike its solution counterpart) and can be applied to non-crystalline samples. Therefore, this method is uniquely positioned to answer key questions about the chemistry and the structure of macromolecular assemblies which, because of their size and structural flexibility, are often difficult to characterize.
However several important problems remain to be solved before ssNMR is ready to cope with challenging solid biochemical assemblies, and many methodological developments are still expected in this fast evolving field.
The principal goal of this thesis has been to test a panel of experimental ssNMR methodologies for the structural investigation of large non-crystalline RNA-protein complexes such as determining the architecture and function of viral capsids.
In detail, we have focused on three highly relevant biological systems, the nucleocapsids of Measles virus (MeV), of Acinetobacter phage 205 (AP205), and of Rice Yellow Mottle Virus (RYMV). These molecules constitute a very challenging target whose molecular weight is 10 to 100 times larger than any structural study previously performed on globular proteins by ssNMR.
In these viral particles, genomic RNA is protected by multiple copies of a coat protein (the so-called N-protein, of about 400 amino acids for MeV), which are organized into complex superstructures (helical for MeV, and icosahedral for AP205 and RYMV, respectively). These systems are too large for solution state NMR, and the difficulty to obtain large single crystals prevents conventional X-rays analysis. However, recombinant (13C,15N)-labelled nucleocapsids can be prepared, and low resolution structures obtained by electron microscopy are available.
In the present work, solid-state NMR was used to study the nucleoproteins in the capsid superstructure. The repetitive positioning of the coat proteins provide precisely the degree of order necessary for high resolution NMR spectra. We have optimized sample conditions such that the proteins can be studied in situ, and have recorded the first (13C,13C), (15N,13C) and (1H, 31P)-correlation spectra (notably, with the aid of high magnetic fields and ultra-fast magic angle sample spinning), which constitute the first steps towards resonance assignment and structure determination, including determination of the quaternary interaction between neighboring proteins on the RNA.
The analysis of the NMR data was enriched by a series of circular dichroism (CD) spectra, which probe the relative arrangements of aromatic bases in the RNA, and of peptide groups along the capsid protein backbone.
The work is highly interdisciplinary, going from research in virology and molecular biology to quantum chemistry, via structural biology and bioinformatics.