Combinatorial Chemistry & High Throughput Screening - Volume 5, Issue 4, 2002

In vitro selection can be used to generate nucleic acid binding species (aptamers) and catalysts (ribozymes) that can recognize a variety of molecules. Because nucleic acid function is largely derived from readily tabulated secondary structures, it has proven possible to engineer aptamers and ribozymes to function as biosensors. Labeling nucleic acids with reporter molecules has yielded simple antibody substitutes, but by relying on ligand-dependent conformational changes it has also proven possible to generate biosensors that can recognize and specifically report the presence of ligands in homogenous solution. It may prove possible to generate signaling aptamers and allosteric ribozymes (aptazymes) that are responsive to a large fraction of an organismal proteome or metabolome using automated methods. Nucleic acid biosensor arrays for non-nucleic acid targets could likely be generated with the same facility as DNA chips.

Investigating ligand-DNA interactions using type IIS restriction enzymes (IISRE) as footprinting reagents is reviewed and contemplated. Ligand binding at a IISRE's cleavage but not sequence recognition site protects DNA from strand scission. This spatial arrangement has been exploited in the development of qualitative (combinatorial) and quantitative ligand-DNA investigative methods collectively termed Type IIS Restriction Enzyme Footprinting (cIISREF and qIISREF respectively). In cIISREF, the consensus binding sequence of a ligand is sought by using a IISRE to segregate combinatorial library members that are bound by ligand from those that are not. A PCR is performed following the segregation step to enrich the library in ligand binding (i.e. uncut) sequences. It might be possible that diversities approaching 10 30 unique sequences could be simultaneously searched using this homogeneous and biologically relevant method. For qIISREF, a ligand-DNA equilibrium constant is measured by quantifying the amounts of target and control DNA IISRE cleavage products as a function of ligand concentration. The control sequence is engineered to not bind ligand. Along with illustrating these methods by reviewing published works, current concerns and future prospects for IISREF are discussed.

While the in vitro selection of nucleic acid binding species (aptamers) requires numerous liquidhandling steps, these steps are relatively straightforward and the overall process is therefore amenable to automation. Here we demonstrate that automated selection techniques are capable of generating aptamers against a number of diverse protein targets. Automated selection techniques can be integrated with automated analytical methods, including sequencing, determination of binding constants, and structural analysis. The methods that have so far been developed can be further multiplexed, and it should soon be possible to attempt the selection of aptamers against organismal proteomes or metabolomes.

Hammerhead ribozymes that are subject to allosteric control by small molecule and oligonucleotide effectors have been reported recently. Rational design has been an effective strategy for the creation of these ribozymes, which incorporate structurally interdependent hammerhead motifs and effector-binding sequences. In this paper we report the rational design of the first protein-responsive allosteric ribozymes that are regulated by the HIV-1 Tat. The TAR-Tat interaction of HIV-1 has the interesting feature that both Tat and arginine are able to bind to and bring about comparable conformational changes in the TAR loop. Here we describe the construction of two classes of TAR-modified hammerhead ribozymes and their response to Tat protein and to its derivatives. Instances of both allosteric activation and inhibition were found. Interestingly, the activation response was stimulated by both Tat and argininamide while the inhibitory response was stimulated by Tat and by its derivative peptide, ADP1, but not by argininamide. Overall, the extent of allosteric response in our ribozymes was modest relative to those reported for ribozymes with small molecule effectors. Future work utilizing combinatorial approaches along with elements of rational design should reveal the means by which highly efficient, protein-mediated allostery of ribozymes may be achieved.

In vitro selection with either DNA or RNA libraries was performed against the TAR RNA element of HIV-1. The role of the selection conditions on the outcome of the selection was evaluated by varying the magnesium concentration and the temperature. The selection stringency was demonstrated to determine i) the affinity of the best identified aptamers for the TAR target, and ii) the type of interaction between the two partners. Selections performed with a DNA library under low (4 C, 10 mM magnesium) and high stringency (23 C, 3 mM magnesium) led to the emergence of “kissing aptamers” but even if the motif interacting directly with the TAR loop were identical in the two kinds of aptamers, the consensus was extended from eight to thirteen nucleotides when the Mg2+ concentration was decreased from 10 to 3 mM. Similar kissing aptamers were selected at 23 C and 37 C starting with two different RNA libraries under identical ionic conditions. In addition, selection performed at 37 C yielded a significant proportion of antisense sequences. Only antisense RNAs complementary to the TAR loop competitively inhibited the association of a Tat peptide with TAR.

Metal ions play important roles in both the structure and function of catalytic DNA and RNA. While most natural catalytic RNA molecules (ribozymes) are active in solutions containing Mg2+, in vitro selection makes it possible to search for new catalytic DNA / RNA that are specific for other metal ions. However, previous studies have indicated that the in vitro selection protocols often resulted in catalytic DNA / RNA that were equally active or sometimes even more active with metal ions other than the metal ion of choice. To improve the metal ion specificity during the in vitro selection process, we implemented a negative selection strategy where the nucleic acid pool was subjected to a solution containing competing metal ions. As a result, those nucleic acids that were active with those metal ions are discarded. To demonstrate the effectiveness of the negative selection strategy, we carried out two parallel in vitro selections of Co2+- dependent catalytic DNA. When no negative selection was used in the selection process, the resulting catalytic DNA molecules were more active in solutions of Zn2+ and Pb2+ than in Co2+. On the other hand, when the negative selection steps were inserted between the normal positive selection steps, the resulting catalytic DNA molecules were much more active with Co2+ than in other metal ions including Zn2+ and Pb2+. These results suggest strongly that in vitro selection can be used to obtain highly active and specific transition metal iondependent catalytic DNA / RNA, which hold great promise as versatile and efficient endonucleases as well as sensitive and selective metal ion sensors.