Prediction of RNA secondary structures: from theory to models and real molecules

RNA secondary structures are derived from RNA sequences, which are strings built form the natural four letter nucleotide alphabet, {AUGC}. These coarse-grained structures, in turn, are tantamount to constrained strings over a three letter alphabet. Hence, the secondary structures are discrete objects and the number of sequences always exceeds the number of structures. The sequences built from two letter alphabets form perfect structures when the nucleotides can form a base pair, as is the case with {GC} or {AU}, but the relation between the sequences and structures differs strongly from the four letter alphabet. A comprehensive theory of RNA structure is presented, which is based on the concepts of sequence space and shape space, being a space of structures. It sets the stage for modelling processes in ensembles of RNA molecules like evolutionary optimization or kinetic folding as dynamical phenomena guided by mappings between the two spaces. The number of minimum free energy (mfe) structures is always smaller than the number of sequences, even for two letter alphabets. Folding of RNA molecules into mfe energy structures constitutes a non-invertible mapping from sequence space onto shape space. The preimage of a structure in sequence space is defined as its neutral network. Similarly the set of suboptimal structures is the preimage of a sequence in shape space. This set represents the conformation space of a given sequence. The evolutionary optimization of structures in populations is a process taking place in sequence space, whereas kinetic folding occurs in molecular ensembles that optimize free energy in conformation space. Efficient folding algorithms based on dynamic programming are available for the prediction of secondary structures for given sequences. The inverse problem, the computation of sequences for predefined structures, is an important tool for the design of RNA molecules with tailored properties. Simultaneous folding or cofolding of two or more RNA molecules can be modelled readily at the secondary structure level and allows prediction of the most stable (mfe) conformations of complexes together with suboptimal states. Cofolding algorithms are important tools for efficient and highly specific primer design in the polymerase chain reaction (PCR) and help to explain the mechanisms of small interference RNA (si-RNA) molecules in gene regulation. The evolutionary optimization of RNA structures is illustrated by the search for a target structure and mimics aptamer selection in evolutionary biotechnology. It occurs typically in steps consisting of short adaptive phases interrupted by long epochs of little or no obvious progress in optimization. During these quasi-stationary epochs the populations are essentially confined to neutral networks where they search for sequences that allow a continuation of the adaptive process. Modelling RNA evolution as a simultaneous process in sequence and shape space provides answers to questions of the optimal population size and mutation rates. Kinetic folding is a stochastic process in conformation space. Exact solutions are derived by direct simulation in the form of trajectory sampling or by solving the master equation. The exact solutions can be approximated straightforwardly by Arrhenius kinetics on barrier trees, which represent simplified versions of conformational energy landscapes. The existence of at least one sequence forming any arbitrarily chosen pair of structures is granted by the intersection theorem. Folding kinetics is the key to understanding and designing multistable RNA molecules or RNA switches. These RNAs form two or more long lived conformations, and conformational changes occur either spontaneously or are induced through binding of small molecules or other biopolymers. RNA switches are found in nature where they act as elements in genetic and metabolic regulation. The reliability of RNA secondary structure prediction is limited by the accuracy with which the empirical parameters can be determined and by principal deficiencies, for example by the lack of energy contributions resulting from tertiary interactions. In addition, native structures may be determined by folding kinetics rather than by thermodynamics. We address the first problem by considering base pair probabilities or base pairing entropies, which are derived from the partition function of conformations. A high base pair probability corresponding to a low pairing entropy is taken as an indicator of a high reliability of prediction. Pseudoknots are discussed as an example of a tertiary interaction that is highly important for RNA function. Moreover, pseudoknot formation is readily incorporated into structure prediction algorithms. Some examples of experimental data on RNA secondary structures that are readily explained using the landscape concept are presented. They deal with (i) properties of RNA molecules with random sequences, (ii) RNA molecules from restricted alphabets, (iii) existence of neutral networks, (iv) shape space covering, (v) riboswitches and (vi) evolution of noncoding RNAs as an example of evolution restricted to neutral networks.