De Novo Creation of Ribozymes

The general concept of SELEX and its use for the isolation of aptamer sequences is described elsewhere in this volume. The crucial step in this technique is selection - an effective distinction between desired and undesired species. It is in the selection step that the most decisive advances in ribozyme discovery are made. The standard selection event in SELEX is a binding event, and consequently the desired molecules can be distinguished from the rest by physical separation, namely by affinity chromatography.

Conceptually starting from the basic SELEX protocol, several strategies to create RNA constructs with catalytic properties have been developed. As already mentioned, the TSA approach has been emulated from catalytic antibodies. Aptamers against TSAs of different reactions have been raised and tested for catalytic activity. Prudent et al. (1994) reported an aptamer with catalytic properties in biphenyl isomerization. An aptamer with porphyrin metallation properties was reported by the Schultz group (Conn et al., 1996). Cholesterol esterase activity of another RNA isolated by the TSA approach was found by Chun et al. (1999). Interestingly, the generation of catalytic antibodies failed in the latter case possibly as a result of poor immunogenicity of the hapten. While all these RNAs act as catalysts, the rate acceleration was limited, at about 88-fold (Prudent et al., 1994), 110-fold (Chun et al., 1999), and 460-fold (Conn et al., 1996), as compared with 103-105 for TSA antibodies (Lolis and Petsko, 1990; Stewart and Benkovic, 1995) and up to 1017 for natural enzymes (Griffiths and Tawfik, 2000; Hilvert, 2000). Other attempts to isolate catalytic RNAs via TSA, for example for the Diels-Alder reaction, have thus far remained at the stage of aptamer binding (Morris et al., 1994; Arora et al., 1998).

Better acceleration rates and catalysis of more complex reactions were obtained by a different strategy, termed direct selection. In this strategy, selection is not based on a binding event; rather it is based directly on the desired catalytic event. For example, if the reaction to be catalyzed is phosphodiester hydrolysis, the RNA library is immobilized on solid support. The selection event is thus directly tailored to hydrolysis: active molecules cleave their covalent linkage to the solid support and thus dissociate from it. The catalytic event thus enables spatial separation of active and inactive molecules. The use of reactants other than RNA in direct selection allows targeting of more complex reactions than hydrolysis or RNA ligation. As stated before, a reactant X can, for example, be conjugated to the RNA library as initiator nucleotide. Assuming a desired RNA catalyst of the chemical reaction XpZ, which may include simple addition as in X + YpZ, the selection has to be designed in such a way as to isolate all RNA library members carrying the target reaction product Z. This can be through chemical trapping or affinity purification of Z directly, or by designing Y in such a way that the final product contains a functional moiety of Y, which allows for easy affinity purification.

An exciting example for the former case is the ribocatalytic oxidation of benzyl alcohol to benzaldehyde using NAD+ as a cofactor: the desired aldehyde product was trapped chemically by a biotin-conjugated hydrazine derivative (Tsukiji et al., 2003). This case will be more explicitly discussed below. The latter case has been successfully demonstrated by several labs employing a biotin-conjugated reactant Y thus allowing purification of the reactive species by streptavidin-agarose (Ekland and Bartel, 1996; Williams and Bartel, 1996; McGinness and Joyce, 2002). Our lab has developed yet another, further refined strategy, named direct selection with linker-coupled reactants. This design, tailored to find catalysts for the chemical addition of compound Y to compound X to yield the addition product Z, is depicted in Fig. 9.1. Reactant X is coupled to the RNA pool by transcription initiation, or by subsequent ligation to the 3' end. However, as opposed to direct selection, the reactant X is now separated from the RNA by a polyethylene glycol (PEG) linker, allowing it to interact with potential catalytic pockets in the RNA, which are spatially remote from either extremity. In our lab, using linkercoupled reactants, a ribozyme was selected that catalyzes a Diels-Alder reaction. Anthracene tethered to RNA via a PEG linker corresponds to reactant X, in this case a diene, and biotin-maleimide was employed as reactant Y, the free dieno-phile (Seelig and Jaschke, 1999). This design allowed easy isolation of the reaction product Z by streptavidin-agarose.

Depending on the reactivity of reactant Y towards the functional groups contained in unmodified RNA, its conjugation to species of the RNA library will occur with more or less site specificity to the conjugated X-reactant. More reactive Ys may add to RNA at positions other than the desired reactant X, thus also fulfilling the practical requirements for selection and amplification. Results of such selections are occasionally reported in the literature (e.g. Wilson and Szostak, 1995). A strategy to circumvent non-specific addition involves the incorporation of a chemically orthogonal cleavage site between reactant X and the RNA library. After immobilization as a primary selection event, a secondary selection event is created by a treatment targeted specifically to the cleavage site. If immobilization of a given RNA sequence is based on addition of reactant Y to reactant X with correct site specificity, the cleavage will release the desired sequence into solution for amplification. However, addition of Y to sites other than X anywhere in the

Fig. 9.1 Direct selection with linker-coupled reactants for the reaction of X with Y to yield product Z. The DNA pool is transcribed and linked with reactant X by conjugation with a functionalized initiator X or by ligation of the transcript to a functionalized dinucleotide. After the reaction the reactive RNA is isolated by immobilization, on a solid phase through an

Fig. 9.1 Direct selection with linker-coupled reactants for the reaction of X with Y to yield product Z. The DNA pool is transcribed and linked with reactant X by conjugation with a functionalized initiator X or by ligation of the transcript to a functionalized dinucleotide. After the reaction the reactive RNA is isolated by immobilization, on a solid phase through an sequencing anchor group at the former reactant Y and further purified by washing the solid phase. Subsequent reverse transcription and polymerase chain reaction (PCR) result in an enriched DNA library, which is used as input for the next round of selection (Seelig and Jaschke, 1999).

RNA will result in persistent immobilization during the secondary selection event, thus effectively removing unwanted species from the amplification cycle. The chemistry used for the cleavage event in the secondary selection step should not, of course, interact in any way with the rest of the conjugated library.

To our knowledge, two such orthogonal approaches have been described in literature. In one case, reductive cleavage of a disulfide is used (Sengle et al., 2000). In our lab, a dinucleotide was designed containing a 5'-pCC ligation site, a PEG linker with an embedded photocleavable o-nitrophenyl group, and a terminal attachment site which can be derivatized for coupling to a desired reactant X (Hausch and Jäschke, 1998). Such photocleavable linker has been successfully applied by the Famulok lab in the selection of a ribozyme catalyzing a Michael addition (Sengle et al., 2001).

0 0

Post a comment