Application of locked nucleic acids to improve aptamer in vivo stability and targeting function.

Aptamers are powerful candidates for molecular imaging applications due to a number of attractive features, including rapid blood clearance and tumor penetration. We carried out structure-activity relationship (SAR) studies with the Tenascin-C binding aptamer TTA1, which is a promising candidate for application in tumor imaging with radioisotopes. The aim was to improve its in vivo stability and target binding. We investigated the effect of thermal stabilization of the presumed non-binding double-stranded stem region on binding affinity and resistance against nucleolytic degradation. To achieve maximal thermal stem stabilization melting experiments with model hexanucleotide duplexes consisting of unmodified RNA, 2'-O-methyl RNA (2'-OMe), 2'-Fluoro RNA (2'-F) or Locked Nucleic Acids (LNAs) were initially carried out. Extremely high melting temperatures have been found for an LNA/LNA duplex. TTA1 derivatives with LNA and 2'-OMe modifications within the non-binding stem have subsequently been synthesized. Especially, the LNA-modified TTA1 derivative exhibited significant stem stabilization and markedly improved plasma stability while maintaining its binding affinity to the target. In addition, higher tumor uptake and longer blood retention was found in tumor-bearing nude mice. Thus, our strategy to introduce LNA modifications after the selection procedure is likely to be generally applicable to improve the in vivo stability of aptamers without compromising their binding properties.

Evaluation of bacterial RNase P RNA as a drug target.

RNA has gained increasing importance as a therapeutic target. However, so far mRNAs rather than stable cellular RNAs have been considered in such studies. In bacteria, the tRNA-processing enzyme RNase P has a catalytic RNA subunit. Fundamental differences in structure and function between bacterial and eukaryotic RNase P, and its indispensability for cell viability make the bacterial enzyme an attractive drug target candidate. Herein we describe two approaches utilized to evaluate whether the catalytic RNA subunit of bacterial RNase P is amenable to inactivation by antisense-based strategies. In the first approach, we rationally designed RNA hairpin oligonucleotides targeted at the tRNA 3'-CCA binding site (P15 loop region) of bacterial RNase P RNA by attempting to include principles derived from the natural CopA-CopT antisense system. Substantial inactivation of RNase P RNA was observed for Type A RNase P RNA (such as that in Escherichia coli) but not for Type B (as in Mycoplasma hyopneumoniae). Moreover, only an RNA oligonucleotide (Eco 3') complementary to the CCA binding site and its 3' flanking sequences was shown to be an efficient inhibitor. Mutation of Eco 3' and analysis of other natural RNase P RNAs with sequence deviations in the P15 loop region showed that inhibition is due to interaction of Eco 3' with this region and occurs in a highly sequence-specific manner. A DNA version of Eco 3' was a less potent inhibitor. The potential of Eco 3' to form an initial kissing complex with the P15 loop did not prove advantageous. In a second approach, we tested a set of oligonucleotides against E. coli RNase P RNA which were designed by algorithms developed for the selection of suitable mRNA targets. This approach identified the P10/11-J11/12 region of bacterial RNase P RNA as another accessible region. In conclusion, both the P15 loop and P10/11-J11/12 regions of Type A RNase P RNAs seem to be promising antisense target sites since they are easily accessible and sufficiently interspersed with nonhelical sequence elements, and oligonucleotide binding directly interferes with substrate docking to these two regions.

5S rRNA is a leadzyme. A molecular basis for lead toxicity.

This paper reports that the D-loop sequence of cellular mammalian ribosomal 5S RNAs is a natural leadzyme that specifically binds and cleaves in trans other RNA molecules in the presence of lead. The D-loops of these 5S rRNAs are similar in sequence to the active site of the leadzyme derived from tRNA(Phe), which cleaves a single bond in cis. We have devised a 12 nt model substrate based on the leadzyme sequence cleaved in trans by a 12 nt RNA molecule containing of the D-loop sequence. The model reaction occurs only at the appropriate concentration of lead and enzyme/substrate stoichiometry. The native 5S rRNA carries the same cleavage activity, although with different optimal lead concentration and stoichiometry. On the other hand, the isolated D-loop does not serve as a substrate when incubated with an RNA molecule with the potential to base pair with it and form the same internal loop (the bubble) present in the leadzyme-substrate complex. We show that the leadzyme cuts C-G, but not G-G or U-G linkages. The 5S rRNA leadzyme appears to have the shortest asymmetric pentanucleotide purine-rich loop flanked by two short double stranded RNAs. The leadzyme activity of native 5S rRNA may be an important aspect of lead toxicity in living cells. Because the leadzyme motif has been found in natural RNA species, its activity can be expressed in vivo even at a very low lead concentrations, of lead leading to the inactivation of other cellular RNAs. This might be one of the ways in which lead poisoning manifests itself at the molecular level. Lead toxicity is based not only on its binding to calcium and zinc binding proteins (such as Zn-fingers) and random hydrolysis of nucleic acids, but also, and most importantly, on the induction of the hydrolytic properties of RNA (RNA catalysis).