Deoxyribozymes for peptide substrates: exploring the landscape of nucleophiles and electrophiles

While DNA is most familiar in its double-stranded form as a storehouse of genetic information, its chemical similarities to naturally occurring ribozymes suggest that it can act as a catalyst. Single-stranded DNA molecules that have the ability to catalyze various bioorganic reactions are called deoxyribozymes. Deoxyribozymes do not occur naturally and are identified using a combinatorial chemistry technique called in vitro selection. Since the discovery of the first artificial deoxyribozyme that catalyzes cleavage of phosphodiester bonds in RNA molecules, a number of deoxyribozymes have been identified that catalyze different bioorganic reactions, including RNA ligation, DNA phosphorylation, DNA deglycosylation, etc. Most of these reactions involve RNA or DNA substrates, where catalysis is facilitated by Watson-Crick base pairing between the substrate and the DNA catalyst. A major challenge in the field is to identify deoxyribozymes that can catalyze reactions involving non-nucleic acid substrates, such as proteins and sugars. Addressing this challenge, we investigated the ability of DNA to catalyze reactions between amino acids that have nucleophilic side chains, such as tyrosine, serine, and lysine with a 5???-triphosphate-RNA electrophile. Previous efforts from our lab revealed DNA-catalyzed tyrosine reactivity, whereas serine proved refractory to catalysis. In the research described in Chapter 3, we identified for the first time, novel DNA enzymes that catalyze chemical modification of serine side chains. These deoxyribozymes were identified in a structurally preorganized three-helix-junction (3HJ) architecture that places the peptide nucleophile close to the triphosphate-RNA electrophile. We showed that these deoxyribozymes can discriminate between Ser and Tyr when presented at the same amino acid position and can also distinguish between multiple Ser side chains at different positions in the peptide. In studies performed in parallel, our results indicated that obtaining DNA-catalyzed Lys side chain reactivity is relatively difficult. Selection experiments with lysine substrate led to deoxyribozymes that catalyze reaction of a phosphoramidite functional group instead of the Lys side chain. To address the difficulties in obtaining deoxyribozymes for Lys side chain reactions, we explored the ability of DNA to catalyze the reaction of amines with a more reactive electrophile, 2???,3???-cyclic phosphate as described in Chapter 4. However, multiple selection efforts involving the 2???,3???-cyclic phosphate electrophile, deoxyribozymes that catalyze a side reaction involving the ribose 2???-hydroxyl emerged. This study highlights some of the limitations of deoxyribozymes and the challenges associated with identifying deoxyribozymes that catalyze reactions involving amine nucleophiles. We also sought to identify deoxyribozymes that can covalently tag phosphopeptides with RNA. These studies are described in Chapter 5. Such deoxyribozymes can be employed as reagents to isolate phosphopeptides from a mixture of peptides. We identified DNA enzymes that catalyze the nucleophilic attack of the phosphate group present in phosphotyrosine- and phosphoserine-containing peptides on 5???-triphosphate-RNA. These deoxyribozymes showed greater than 200-fold selectivity for phosphorylated peptides over non-phosphorylated analogs. The catalytic efficiency of one of these deoxyribozyme was assayed with a wide range of peptide substrates that differ in the identity of amino acid flanking the phosphorylated tyrosine. This deoxyribozyme catalyzes the reaction of these peptide sequences with similar catalytic efficiency. Finally, as described in Chapter 6, we investigated the ability of DNA to catalyze phosphorylation of tyrosine and serine hydroxyls, that may be employed as artificial kinases. Artificial kinases could modulate numerous metabolic processes where naturally occurring kinases play an important role. In our first effort to identify kinase deoxyribozymes, we employed 5???-thiotriphosphate-RNA or GTP???S as a phosphate donor. However, these selection experiments were unsuccessful, and we later identified that this was due to the instability of the thiotriphosphate moiety under the selection conditions. We then developed a novel selection approach that eliminated the use of thiotriphosphates. In this approach, capture deoxyribozymes that can attach an RNA molecule to phosphorylated peptides were employed to identify kinase deoxyribozymes. In these experiments the observed activity was found to depend on the covalent connection between the substrate and the DNA pool