Info

Functional Principles: Recognition, Switches and Catalysis

Although the range of activities of natural nucleic acids seems limited when compared with that of proteins, a remarkable range of nucleic acid functions (Fig. 10.5) have been unraveled by in vitro evolution techniques such as SELEX (Section 10.4) and by the recent developments in molecular and cellular biology [40].

Deco Texte

Fig. 10.5 Functional principles of nucleic acid foldamers. All single-stranded RNA and DNA foldamers folds in presence of salt ions. (a) Aptamers: RNA or DNA foldamers able to specifically recognize and bind a ligand. In the absence of ligand, the aptamer can be either flexible or rigid; (b) Riboswitch: RNA foldamer able to switch from one conformation to another upon binding of a ligand effector; (c) Ribozyme or DNAzyme:

Fig. 10.5 Functional principles of nucleic acid foldamers. All single-stranded RNA and DNA foldamers folds in presence of salt ions. (a) Aptamers: RNA or DNA foldamers able to specifically recognize and bind a ligand. In the absence of ligand, the aptamer can be either flexible or rigid; (b) Riboswitch: RNA foldamer able to switch from one conformation to another upon binding of a ligand effector; (c) Ribozyme or DNAzyme:

RNA or DNA foldamer can catalyze chemical reactions with multiple turnovers; (d) Aptazyme (allosteric ribozyme or DNAzyme): responsive catalytic foldamer that is activated by the presence (or absence) of a molecular effector. (e) Self-assembling DNA tiles or tectoRNAs: foldamers that assemble into stable nanostructures in the presence of salts and divalent ions (grey spheres) (Section 10.7).

10.2.2.1 Aptamers and Nucleic Acid Switches

RNA and DNA sequences called aptamers (Fig. 10.5), can act as receptors for a limitless number of ligands such as small bio-organic and synthetic compounds, ions, peptides, proteins, polysaccharides, lipid membranes and mineral surfaces (see reviews [7, 29, 40]). They bind their respective ligands with remarkable specificity and equilibrium constants of dissociation ranging from the micromolar range to the femtomolar range [42, 43]. The specific sequence signature of an ap-tamer can be surprisingly small and can vary from 10 to 100 nucleotides. In the absence of ligands, an aptamer sequence can readily fold into a rigid, stable 3C structure (Fig. 10.5a). However, it can also adopt a ''flexible'' metastable conformational state that is cooperatively stabilized into a unique ''rigid'' structure by induced fit mechanism upon ligand binding (Fig. 10.5a). Several artificial and natural RNA aptamers acting as switches or riboswitches have been found to exploit this property [44-46] (Figures 10.5(b) and 10.6(a)). Natural riboswitches are found to regulate the expression of genes by either activating or inhibiting the transcription or translation upon binding of a small molecular effector [46-49]. These RNA domains are generally able to adopt two distinct alternative conforma-

Fig. 10.6 Three-dimensional structures of a natural riboswitch and artificial ribozyme. (a) 3-D crystallographic structure of the thiamine pyrophosphate (TPP) sensing riboswitch: a natural riboswitch involved in gene regulation in bacteria [53]. The RNA bases and bound TPP (red) are shown as cylinders and the backbone is depicted with a ribbon. At the bottom left, detailed view of the TTP binding domain; (b) 3-D crystallographic structure of the Diels-Alder ribozyme: an artificial ribozyme that catalyzes carbon-carbon bond formation between anthracene and N-pentyl maleimide. The RNA bases and bound product (blue) are shown as cylinders. The backbone is depicted with a ribbon, whereas the hydrated Mg2+ are in a mesh representation [219]. On the right, detailed view of the catalytic site with the bound product. No Mg2+ seems to be directly involved in catalysis. Adapted with permission from Refs. [53, 219].

Fig. 10.6 Three-dimensional structures of a natural riboswitch and artificial ribozyme. (a) 3-D crystallographic structure of the thiamine pyrophosphate (TPP) sensing riboswitch: a natural riboswitch involved in gene regulation in bacteria [53]. The RNA bases and bound TPP (red) are shown as cylinders and the backbone is depicted with a ribbon. At the bottom left, detailed view of the TTP binding domain; (b) 3-D crystallographic structure of the Diels-Alder ribozyme: an artificial ribozyme that catalyzes carbon-carbon bond formation between anthracene and N-pentyl maleimide. The RNA bases and bound product (blue) are shown as cylinders. The backbone is depicted with a ribbon, whereas the hydrated Mg2+ are in a mesh representation [219]. On the right, detailed view of the catalytic site with the bound product. No Mg2+ seems to be directly involved in catalysis. Adapted with permission from Refs. [53, 219].

tions that are in equilibrium, one of the conformations being favored upon binding of a small target compound. The conformation that binds the target is generally metastable in its absence, allowing another thermodynamically more favored conformation to occur: the free energy change between the two conformational states is small and depends on few key 3° contacts directly involving ligand binding [50, 51]. The NMR and X-ray structures of several aptamers and riboswitches are presently available and shed light on the molecular recognition characteristic of these molecules [29, 51-56] (Fig. 10.6a). DNA switches have not yet been identified in Nature, but there is no conceptual reason why they should not be engineered.

10.2.2.2 Ribozymes and DNAzymes

Since the discovery of the first catalytic RNA molecules (ribozymes) in the early 1980s, RNA has been shown to exhibit a large repertoire of catalytic functions [57]. This has been extensively reviewed recently [58-60] and will only be described briefly. Ribozymes can efficiently achieve catalysis by bringing the reactive groups close to each other via specific binding, by precise orientation of the reactive groups and by structural complementarity to the substrate transition state (Figures 10.5(c) and 10.6(b)). Ribozymes are often known as metalloenzymes, although they may not directly involve divalent ions in RNA catalysis (Fig. 10.6b). They can also perform general acid/base and covalent catalysis. Besides reactions at phosphoryl centers, RNA is able to catalyze the formation of esters, amides, glycosidic and carbon-carbon bonds as well as alkylation, isomerization, metala-tion, peroxidation, oxido-reduction and aldol reactions [60, 61]. Recently, pyridyl-modified RNA sequences isolated by in vitro selection were found to catalyze the growth of palladium nanocrystals in short reaction times (@1 min) and with a high degree of shape specificity [62-64], suggesting that RNA can actively take part in the evolution of inorganic materials. The mechanism and exquisite detail of the 3° topology and catalytic site of several ribozymes has recently been revealed by X-ray crystallography [58, 59, 65-68].

The structural diversity of DNA aptamers suggests that DNA can form many of the same secondary structures that are exhibited by ribozymes. Despite the lack of 2'-OH groups, there is now compelling evidence that DNA can also efficiently catalyze various chemical reactions [69-71]. The present scope of reactions catalyzed by DNA enzymes (DNAzymes) is somewhat less impressive than that of ribozymes. Nevertheless, beside reactions of phosphorylation, adenylation, ligation or cleavage occurring on phosphoryl centers, some DNAzymes catalyze more exotic reactions such as peroxidation, porphyrin metalation, DNA depurination and thymine dimer photoreversion [71]. Interestingly, DNA higher order structures do not automatically require the presence of divalent ions, and some deoxyribo-zymes rely only on potassium ions, which are known to promote the formation of G-tetrads [72].

10.2.2.3 Multifunctional Nucleic Acid Foldamers

Strikingly, most of the new catalytic functions isolated by in vitro selection do not require large structural motifs. RNA pools containing fewer than 100 random po sitions (N100 random pool) in their sequences are sufficient for finding ribo-zymes with very different catalytic functions [57]. Longer random pools are generally required to select more complex RNA molecules, however. The degree of functional complexity reached by an RNA molecule apparently correlates with its structural complexity [73]. Thus, the longer the sequence of the random pool, the better is the probability of selecting complex and interesting ribozymes. Moreover, only zeptomoles of nucleic acid might be necessary to find small functional modules within large random pools [74]. Another interesting aspect is that RNA molecules composed of several separate and equally sized modules might have a strong selective advantage compared with larger unique structures [74].

In view of the modular organization of large RNA molecules [30], it is possible to take advantage of a known RNA domain that displays a specific function, such as substrate binding or catalysis, to select multifunctional ribozymes. The assumption is that some molecules selected from an RNA pool which consists of a known RNA module associated with a random domain, will combine the properties of the constant RNA module to a new function. This approach has been successful for the isolation of highly complex template-directed, sequence-independent RNA ligases from a pool of 1016 molecules consisting of a preexisting structural scaffold, appended to random RNA segments [75]. Similar modular approaches were applied to generate a ribozyme that is able to polymerize any RNA sequences up to 14 nucleotides by RNA-template primer extension [76], and a bifunctional ribozyme that can recognize an activated glutaminyl ester and subsequently amino-acylate a tRNA molecule [77]. Interestingly, structurally and functionally complex ribozymes can be isolated from libraries formed of stable structural modules associated with random regions of very limited size (N30 random positions) [78, 79]. Combination of functional motifs was also used to generate complex molecules with dual activities such as RNA cleavage and liga-tion [80], and allosteric ribozymes [81-84]. The later are also called aptazymes because they result from the combination of an aptamer joined to a catalytic domain by a communication module (Fig. 10.5d). Hammerhead self-cleaving aptazymes have been trained by in vitro evolution to switch their activity on or off with remarkable allosteric responses that are orders of magnitude greater than those typically seen for protein enzymes [85]. DNA aptazymes have also been engineered [86]. Recently, bifunctional RNA molecules combining binding and catalytic activities were identified from random pools by a new two-step selection method [87].

In the future, it is likely that novel ligand-responsive RNA self-assemblies will be generated likewise by taking advantages of artificial or natural riboswitches (Fig. 10.5e).

0 0

Post a comment