Skip to main content

In vitro selection of RNA structural motifs with self-splicing activity

Objective

The profect relies on in vitro darwinian selection of RNA molecules to produce RNA structural motifs with designed characteristics (`aptamers'): self-splicing activity, potential for recognition of other RNA motifs, ability to bind specifically antibiotics or amino acids. The selected amino acid binding RNA aptamers can be used for a selection of novel ribozymes which use amino acid substrates for catalysis or which catalyse the formation of an amide bond. Self-splicing of group I introns are inghibited by a number of small ligands, among these are the aminoglycoside neomycin B and the peptide viomycin. From detailed characterization of small RNAs isolated via SELEX, information on how RNA folds and accommodates ligands were obtained. The structural identification of the motifs is performed using chemical footprinting techniques, nuclear magnetic resonance spectroscopy and three-dimension modelling. Simultaneously, the development of new tools for modelling and visualizing RNAs is pursued for coping with the newly derived sequences.
Several catalytic RNAs have been built (from small ribozymes to very large autocatalytic introns) and used for developping the data base. Together with this data base, the programming of automatic or semi-automatic tools for assembling RNA molecules is coming to a stage where the programs can be used. Molecular dynamics simulations of RNAs were successfully performed on the tRNA-Asp and its anticodon hairpin. This development has allowed the realization that C-H..O/N bonds are of importance for the stability of base pairs and RNA 3D structure in general. Molecular dynamics simulations have also permitted the assessment of the lifetimes of water molecules around those RNAs. These molecular dynamics studies are the most extensive and thorough ever performed on complex RNA systems.
In vitro selection was used to identify high-affinity RNA receptors for three ENA tetraloops, GUGA, GAAA and UUCG, which are among the commonest terminal loops in natural RNAs with a stable, phylogenetically conserved structure. Some of the receptors we isolated had already been identified by comparative sequence analysis of natural RNAs, whereas others are new. Individual receptors were characterized kinetically and found to display specific preferences for some members of the GNRA and UNCG loop families. These data, together with two recently published crystal structures of loop-receptor pairs, make it possible to propose general rules for the recognition of GNRA loops by RNA. One application of this work has been the discovery of three novel long-range contacts in group II self-splicing intron. This is the first time interactions that ensure mutual recognition of the variousstructural domains of group II introns have been identified and all three of them involve tetraloops of the GNRA family. One or these interactions brings together the small domain V, which is essential for catalysis, and the large domain I, which carries exon-binding sites. A second one mediates a conformational rearrangement that occurs in between the two reaction steps of the self-splicing process. These data have opened the way for the modelling of the higher-order structure of group II introns. The occurrence of those loops (belonging to the -GNRA- family) in self-assembly of large RNA molecules is only surpassed by loop-loop interactions between complementary regions of hairpins. The modelling of complete group I and group II introns shows that those two broad structural motifs are the most frequent ones in self-assembly of RNAs.
A pool of RNAs which specifically bind the aminoglycoside antibiotic neomycin has also been produced. The enriched-pool was dominated by RNA sequences containing a hairpin stem-loop structural motif, a fact sugesting that structure rather than primary sequence is required for neomycin recognition. The neomycin-binding site is being foot-printed and structurally characterized. Based on cleavages induced by ions which do not promote splicing and mutational analyses, the location for two metal ions surrounding the splice site was deduced. More than 90% of the molecules selected for viomycin binding, a peptifde antibiotic, shared one continuous highly conserved region of 14 nucleotides harbouring the antibiotic-binding site and forms a stem-loop structure with a pseudoknot. A comparison between the novel viomycin-binding motif and the natural RNA target sites for viomycin showed that all form a pseudoknot at the antibiotic-binding site. Mutational analyses and backbone footprinting experiments with the T4 phage derived thymidylate synthase group I intron mapped the binding site of neomycin B to the internal loop between the P4 and P5 stems and the joining region J6/7 Again, RNA structure rather than sequence is the major determinant for neomycin B recognition.
Chemical probing, damage selection and lead cleavage of the aptamers recognizing arginine and citrulline have been crrie dout. The three-dimensional structures of those complexes were carried out by multidimensional NMR spectroscopy, a first internationally. Pool synthesis for the peptide bond forming ribozyme has been completed. RNA Aptamers with affinity to amino acids and flavin cofactors were selected from random sequence libraries of 10{15} different molecules. Functional and structzural characterizations of these RNA molecules were carried out. Secondary structure refinements were performed using chemical modification and damage selection approaches. During the structural characterizations of these RNA molecules were carried out. Secondary structure refinements were performed using chemical modification and damage selection approaches. During the structural characterization of the FMN-aptamer it was discovered that isoalloxazine derivatives can cleave RNA molecules in a photooxidative mechanism, a reaction occurring essentially at GU pairs within RNA helices. The study represents the first example in which a structure motif made of an unusual base pair in an RNA is specifically recognized and affected by a low-molecular-weight molecule. An in vitro selection for an RNA-cleaving DNA-enzyme which utilizes an amino acid cofactor, histidine, for the celavage reaction was performed. A deoxyribozyme with RNA-cleaving activity was indeed selected, however, this deoxyribozyme did non require the histidine, but magnesium or calcium ions.

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE
Address
15,Rue René Descartes 15
67084 Strasbourg
France

Participants (4)

CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE
France
Address
Avenue De La Terrasse, Bftiment 26
91198 Gif Sur Yvette
Centre National de la Recherche Scientifique (CNRS)
France
Address
15 Avenue Charles Flahault
34060 Montpellier
Ludwig-Maximilians-Universität München
Germany
Address
Karlstraße 23
80333 München
UNIVERSITY OF VIENNA
Austria
Address
9,Dr. Bohrgasse 9
1030 Wien