Alex B. Burgin

Selection, design, and characterization of a new potentially therapeutic ribozyme

An in vitro selection was designed to identify RNA-cleaving ribozymes predisposed for function as a drug. The selection scheme required the catalyst to be trans-acting with phosphodiesterase activity targeting a fragment of the Kras mRNA under simulated physiological conditions. To increase stabilization against nucleases and to offer the potential for improved functionality, modified sequence space was sampled by transcribing with the following NTPs: 2-F-ATP, 2-F-UTP, or 2-F-5-[(N-imidazole-4-acetyl) propylamine]-UTP, 2-NH2-CTP, and GTP. Active motifs were identified and assessed for their modified NMP and divalent metal dependence. The minimization of the ribozymes size and the ability to substitute 2-OMe for 2-F and 2-NH2 moieties yielded the motif from these selections most suited for both nuclease stability and therapeutic development. This motif requires only two 2-NH2-Cs and functions as a 36-mer. Its substrate sequence requirements were determined to be 5-Y-G-H-3. Its half-life in human serum is >100 h. In physiologically relevant magnesium concentrations [approximately 1 mM] its kcat = 0.07 min(-1), Km = 70 nM. This report presents a novel nuclease stable ribozyme, designated Zinzyme, possessing optimal activity in simulated physiological conditions and ready for testing in a therapeutic setting.

In vitro selection of a novel nuclease-resistant RNA phosphodiesterase

BACKGROUNDnRibonucleotide-based enzymes (ribozymes) that cleave pathological RNAs are being developed as therapeutic agents. Chemical modification of the hammerhead ribozyme has produced nuclease-resistant catalysts that cleave targeted mRNAs in cell culture and exhibit antitumor activity in animals. Unfortunately, stabilizing modifications usually reduce the catalytic rate in vitro. An alternative to rationally designed chemical modifications of existing ribozymes is to identify novel motifs through in vitro selection of nuclease-stable sequence space. This approach is desirable because the catalysts can be optimized to function under simulated physiological conditions.nnnRESULTSnUtilizing in vitro selection, we have identified a nuclease-stable phosphodiesterase that demonstrated optimal activity at simulated physiological conditions. The initial library of 10(14) unique molecules contained 40 randomized nucleotides with all pyrimidines in a nuclease-stabilized 2-deoxy-2-amino format. The selection required trans-cleaving activity and base-pairing specificity towards a resin-bound RNA substrate. Initial selective pressure was permissive, with a 30 min reaction time and 25 mM Mg(2+). Stringency of selection pressure was gradually increased until final conditions of 1 mM Mg(2+) and less than 1 min reaction times were achieved. The resulting 61-mer catalyst required the 2-amino substitutions at selected pyrimidine positions and was stable in human serum (half-life of 16 h).nnnCONCLUSIONSnWe demonstrated that it is possible to identify completely novel, nuclease-resistant ribozymes capable of trans-cleaving target RNAs at physiologically relevant Mg(2+) concentrations. The new ribozyme motif has minimal substrate requirements, allowing for a wide range of potential RNA targets.

Can DNA Topoisomerases Be Ribonucleases

DNA topoisomerases catalyze changes in the linkage of DNA strands through a concerted mechanism of DNA strand cleavage and ligation reactions (for review, seeWang 1996xWang, J.C. Annu. Rev. Biochem. 1996; 65: 635–692Crossref | PubMedSee all ReferencesWang 1996). Eukaryotic type I topoisomerases are effective at relaxing supercoiled DNA, and this activity is intimately involved in transcription and DNA replication. In addition, this activity has been directly implicated in DNA recombination and DNA repair. These enzymes have been extensively studied because of their fundamental importance in many different biological functions and because topoisomerases are the target of therapeutic agents for human cancer (Chen and Liu, 1996). In the December issue of Molecular Cell, Sekiguchi and Shuman 1997xSekiguchi, J. and Shuman, S. Molec. Cell. 1997; 1: 89–97Abstract | Full Text | Full Text PDF | PubMedSee all ReferencesSekiguchi and Shuman 1997 present biochemical evidence that eukaryotic topoisomerase I can also act as a site-specific endoribonuclease. From this result, the authors postulate that an additional biological function of topoisomerases may be the identification, through cleavage, of ribonucleotides misincorporated in DNA and that these enzymes may even be involved in RNA processing reactions.At first glance, the modulation of DNA topology and RNA processing may appear to be very disparate activities. However, the data presented by Sekiguchi and Shuman 1997xSekiguchi, J. and Shuman, S. Molec. Cell. 1997; 1: 89–97Abstract | Full Text | Full Text PDF | PubMedSee all ReferencesSekiguchi and Shuman 1997 demonstrate that the mechanism of RNA strand scission is similar to that used to modulate DNA topology. Topoisomerase I from vaccinia virus was used for these studies. This enzyme is particularly useful because, unlike other eukaryotic topoisomerases, the vaccinia enzyme exhibits high sequence specificity. Topoisomerases relax supercoiled DNA by cleaving and rejoining one strand of the duplex (see Figure 1Figure 1). DNA cleavage is the result of a transesterification of an active site tyrosine creating a covalent 3′-phosphotyrosyl (DNA–enzyme) intermediate and liberating a 5′-hydroxyl (5′-OH). Although this covalent intermediate is chemically very stable, the steady-state concentration of this intermediate is usually low. This is because the ligation reaction is faster than the cleavage reaction. During ligation, the 5′-OH generated during cleavage displaces the covalent DNA–enzyme intermediate in another transesterification reaction and reforms the 5′–3′ phosphodiester linkage. A change in DNA linking number results when strand passage occurs between the cleavage and ligation steps. The cleavage and ligation reactions are typically viewed as a pseudoequilibrium because multiple cleavage/ligation cycles can be performed on the same substrate (Stivers et al. 1994xStivers, J.T., Shuman, S., and Mildvan, A.S. Biochemistry. 1994; 33: 327–339Crossref | PubMedSee all ReferencesStivers et al. 1994).Figure 1Comparison of Topoisomerase I–Mediated Relaxation of Supercoiled DNA and RNA Strand CleavageView Large Image | View Hi-Res Image | Download PowerPoint SlideAlthough the normal topoisomerase reaction uses the 5′-OH to attack the covalent DNA–enzyme intermediate, it has been demonstrated that a variety of other nucleophiles can attack this intermediate. For example, a 5′-OH from another DNA strand can displace the enzyme resulting in a recombinant DNA molecule. In addition, other hydroxyls from the solvent (e.g., glycerol) can attack the intermediate (Christiansen et al. 1994xChristiansen, K., Knudsen, B.R., and Westergaard, O. J. Biol. Chem. 1994; 269: 11367–11373PubMedSee all ReferencesChristiansen et al. 1994). It may therefore not be surprising that a 2′-hydroxyl (2′-OH) can be an active nucleophile at the ligation step of the topoisomerase reaction. In the present paper, the authors demonstrate that when topoisomerase I cleaves a phosphodiester at a ribonucleotide (see Figure 1Figure 1), the 2′-OH from the ribonucleotide can displace the enzyme forming a 2′–3′ cyclic phosphate in an intramolecular transesterification reaction (cyclization). This cyclic phosphodiester is apparently not a substrate for topoisomerase cleavage and does not act as acceptor for attack of the 5′-OH. As a result, the cyclization reaction traps the cleaved intermediate by displacing the enzyme. Cleaved RNA accumulates because the pseudoequilibrium normally associated with these reactions is destroyed.The authors provide thorough evidence that the vaccinia topoisomerase-mediated RNase reaction proceeds as detailed in Figure 1Figure 1. First, the RNase activity requires an active site tyrosine and a ribonucleotide at the site of cleavage. Second, indirect evidence strongly argues that the cleaved RNA contains a 2′, 3′ cyclic phosphate, and it is reasonable to presume that a 5′-OH is also generated. Third, kinetic data are consistent with the formation of a covalent RNA–enzyme intermediate that is chased (via cyclization) into cleaved product. And finally, vaccinia topoisomerase I acts catalytically since it is able to turn over during the RNase reaction.During vaccinia topoisomerase-mediated RNA cleavage, greater than 90% of the RNA substrate is typically converted to a 2′, 3′ cyclic product. Because the enzyme is capable of converting essentially all of the RNA-containing substrate to cleaved product, it is reasonable to propose that the observed in vitro activity is not an irrelevant side reaction. In addition, some ribonuclease activity is also demonstrated for human topoisomerase I, indicating that the reaction is not unique to vaccinia topoisomerase I. However, further characterization is necessary to determine how efficiently this reaction proceeds with human topoisomerase I. It is important to note that the extent of cleavage may not be the best indicator of efficiency. For example, even if the rate of cyclization is extremely slow, the cleaved product will still accumulate since the cyclization reaction irreversibly traps the cleaved intermediate. Preliminary kinetic data, using vaccinia topoisomerase I, does argue that the rate of cyclization (0.45 min–1) is approximately the same rate as ligation (0.6 min–1).The fact that the vicinal 2′-OH and 5′-OH appear to compete equally for attack on the RNA–enzyme intermediate provides important insight into structure/function relationships of topoisomerases. The enzyme must align the attacking nucleophile and the phosphodiester for in-line attack, and the enzyme can apparently accommodate two nucleophiles with different chemical reactivities (secondary versus primary alcohols) and different spatial constraints. It remains to be understood how the enzyme accommodates such a diverse array of acceptors in the ligation reaction.It will also be important to characterize how well topoisomerases recognize and cleave ribonucleotides during the first step (cleavage) of the reaction. Vaccinia topoisomerase-mediated RNA cleavage, like the standard cleavage/ligation reaction, is sequence specific. It is therefore tempting to speculate that vaccinia topoisomerase I may be involved in specific RNA processing reactions in vivo. A specific and efficient RNA substrate will be required before such a biological function can be assigned. Even though no substrate is available, the fact that another type I topoisomerase, E. coli DNA topoisomerase III, can act as an RNA topoisomerase (Wang et al. 1996xWang, H., Di Gate, R.J., and Seeman, N.C. Proc. Natl. Acad. Sci. USA. 1996; 93: 9477–9482Crossref | PubMed | Scopus (58)See all ReferencesWang et al. 1996) does provide some credibility to this proposal.Finally, the experiments described have been performed on DNA substrates containing ribonucleotides at the cleavage site or substrates containing RNA 3′ to the cleavage site. It has been previously demonstrated that substrates containing RNA 5′ to the cleavage site are poor substrates for vaccinia topoisomerase I. The RNA cleavage reaction is therefore more similar to RNase H cleavage than strict endoribonuclease activity. However, if a biological function for topoisomerase I is to cleave ribonucleotides that have been misincorporated into DNA, then the observed in vitro properties are expected.The ability of an abundant enzyme, like topoisomerase I, to act as a sensor for the misincorporation of ribonucleotides is a particularly attractive possibility for several reasons. First, the lack of sequence specificity in the DNA cleavage reaction of topoisomerase I would allow ribo-directed strand scission at a broad spectrum of sites in genomic DNA. The 2′, 3′ cyclic phosphate and 5′-OH end would presumably be recognized as a deleterious lesion and targeted by enzymes involved in excision repair. Second, topoisomerase I is up-regulated in some cancer cells, and one would expect the misincorporation of ribonucleotides to be particularly prevalent in such rapidly dividing cells. Clearly, for any reaction characterized in vitro, genetic experiments are ultimately required to address biological function.