Systematic evaluation of the dependence of deoxyribozyme catalysis on random region length

Abstract

Functional nucleic acids are DNA and RNA aptamers that bind targets, or they are deoxyribozymes and ribozymes that have catalytic activity. These functional DNA and RNA sequences can be identified from random-sequence pools by in vitro selection, which requires choosing the length of the random region. Shorter random regions allow more complete coverage of sequence space but may not permit the structural complexity necessary for binding or catalysis. In contrast, longer random regions are sampled incompletely but may allow adoption of more complicated structures that enable function. In this study, we systematically examined random region length (N(20) through N(60)) for two particular deoxyribozyme catalytic activities, DNA cleavage and tyrosine-RNA nucleopeptide linkage formation. For both activities, we previously identified deoxyribozymes using only N(40) regions. In the case of DNA cleavage, here we found that shorter N(20) and N(30) regions allowed robust catalytic function, either by DNA hydrolysis or by DNA deglycosylation and strand scission via β-elimination, whereas longer N(50) and N(60) regions did not lead to catalytically active DNA sequences. Follow-up selections with N(20), N(30), and N(40) regions revealed an interesting interplay of metal ion cofactors and random region length. Separately, for Tyr-RNA linkage formation, N(30) and N(60) regions provided catalytically active sequences, whereas N(20) was unsuccessful, and the N(40) deoxyribozymes were functionally superior (in terms of rate and yield) to N(30) and N(60). Collectively, the results indicate that with future in vitro selection experiments for DNA and RNA catalysts, and by extension for aptamers, random region length should be an important experimental variable.

Figure 1

Two reactions systematically examined for the effect of random region length on DNA catalysis. In each case, the key selection step is illustrated. (A) DNA-catalyzed DNA cleavage. (B) DNA-catalyzed tyrosine-RNA nucleopeptide linkage formation.

Figure 2

Various substrate cleavage sites for deoxyribozymes from the initial (A) N₂₀ and (B) N₃₀ selections for DNA-catalyzed DNA cleavage. Individual deoxyribozymes are denoted 8VA followed by a clone number for N₂₀ and 8VB followed by a clone number for N₃₀. Each deoxyribozyme was allowed to cleave the 5′-³²P-radiolabeled DNA substrate (open arrowhead), resulting in a product band (filled arrowhead) whose PAGE migration rate relative to the standard suggests the cleavage site. The standard sample corresponds to hydrolysis of the substrate sequence at the position noted in the Experimental Procedures. Incubation conditions: 70 mM HEPES, pH 7.5, 1 mM ZnCl₂, 20 mM MnCl₂, 40 mM MgCl₂, and 150 mM NaCl at 37 °C (t = 20 h). Quantification of cleavage kinetics for each deoxyribozyme is provided in Figure 3. The reaction type (i.e., substrate hydrolysis, or substrate deglycosylation followed by two β-eliminations) and precise site of each cleavage reaction was assigned with confidence for each deoxyribozyme by MALDI mass spectrometry of the products (Figure 4; see Tables S1 and S2 for all data values and cleavage-site assignments). The type of reaction catalyzed by each deoxyribozyme is marked below its lane with “H” for hydrolysis or “D” for deglycosylation.

Figure 3

Kinetic plots for individual deoxyribozymes from the (A) N₂₀ and (B) N₃₀ selections for DNA cleavage. Kinetic plots for several additional deoxyribozymes from each selection experiment are shown in Figure S2. k_obs values are tabulated in Table S4. Incubation conditions as in Figure 2.

Figure 4

Representative MALDI mass spectra to assign cleavage reactions and sites for deoxyribozymes from the initial N₂₀ and N₃₀ selections for DNA cleavage. All mass spectra data and precise cleavage-site assignments are tabulated in Tables S1 and S2. The indicated C nucleotide of the substrate was not base-paired with either deoxyribozyme binding arm (see Figure 1). (A) 8VB4 deoxyribozyme, which hydrolyzes the DNA substrate at a specific phosphodiester linkage and forms 3′-phosphate and 5′-hydroxyl products. Some of the new DNA-hydrolyzing deoxyribozymes lead instead to 3′-hydroxyl and 5′-phosphate products. (B) 8VB5 deoxyribozyme, which deglycoyslates the DNA substrate at a specific guanosine. After two subsequent β-elimination reactions, the products are missing the entire G nucleoside and have 3′-phosphate and 5′-phosphate groups. Some of the new DNA-deglycosylating deoxyribozymes are not site-specific and deglycosylate the substrate at either of two adjacent nucleotide positions. In the spectrum of panel B, the asterisk denotes the peak for uncleaved substrate with z = 2.

Figure 5

Activities of deoxyribozymes from the follow-up N₂₀, N₃₀, and N₄₀ selections for DNA-catalyzed DNA hydrolysis. (A) PAGE image for each of the six new deoxyribozymes, showing representative timepoints (t = 30 s, 15 min, 2 h, and 20 h). Substrate (open arrowhead) and product (filled arrowhead) are marked. Incubation conditions: 70 mM HEPES, pH 7.5, 1 mM ZnCl₂, 20 mM MnCl₂ if indicated, and 150 mM NaCl at 37 °C. (B) Kinetic plots. k_obs values are tabulated in Table S4.

Figure 6

Activities of deoxyribozymes from the selections for DNA-catalyzed nucleopeptide linkage formation. (A) PAGE image showing representative timepoints (t = 30 s, 30 min, 2 h, and 20 h). Incubation conditions: 50 mM HEPES, pH 7.5, 20 mM MnCl₂, 40 mM MgCl₂, 150 mM NaCl, and 2 mM KCl at 37 °C. Substrate (open arrowhead) and product (filled arrowhead) are marked. For N₃₀, 8TM3 was the only new deoxyribozyme identified, along with two closely related sequence variants. For N₄₀, 11MN5 was the best deoxyribozyme identified previously by direct selection; data shown here were newly acquired for comparison. For N₆₀, 7TQ20 was one of two different deoxyribozymes found, along with six close sequence variants; the other deoxyribozyme, 7TQ46 (no variants found), had slightly lower activity. (B) Kinetic plots. k_obs values are tabulated in Table S4.