Zn2+-dependent deoxyribozymes that form natural and unnatural RNA linkages

Abstract

We report Zn(2+)-dependent deoxyribozymes that ligate RNA. The DNA enzymes were identified by in vitro selection and ligate RNA with k(obs) up to 0.5 min(-)(1) at 1 mM Zn(2+) and 23 degrees C, pH 7.9, which is substantially faster than our previously reported Mg(2+)-dependent deoxyribozymes. Each new Zn(2+)-dependent deoxyribozyme mediates the reaction of a specific nucleophile on one RNA substrate with a 2',3'-cyclic phosphate on a second RNA substrate. Some of the Zn(2+)-dependent deoxyribozymes create native 3'-5' RNA linkages (with k(obs) up to 0.02 min(-)(1)), whereas all of our previous Mg(2+)-dependent deoxyribozymes that use a 2',3'-cyclic phosphate create non-native 2'-5' RNA linkages. On this basis, Zn(2+)-dependent deoxyribozymes have promise for synthesis of native 3'-5'-linked RNA using 2',3'-cyclic phosphate RNA substrates, although these particular Zn(2+)-dependent deoxyribozymes are likely not useful for this practical application. Some of the new Zn(2+)-dependent deoxyribozymes instead create non-native 2'-5' linkages, just like their Mg(2+) counterparts. Unexpectedly, other Zn(2+)-dependent deoxyribozymes synthesize one of three unnatural linkages that are formed upon the reaction of an RNA nucleophile other than a 5'-hydroxyl group. Two of these unnatural linkages are the 3'-2' and 2'-2' linear junctions created when the 2'-hydroxyl of the 5'-terminal guanosine of one RNA substrate attacks the 2',3'-cyclic phosphate of the second RNA substrate. The third unnatural linkage is a branched RNA that results from attack of a specific internal 2'-hydroxyl of one RNA substrate at the 2',3'-cyclic phosphate. When compared with the consistent creation of 2'-5' linkages by Mg(2+)-dependent ligation, formation of this variety of RNA ligation products by Zn(2+)-dependent deoxyribozymes highlights the versatility of transition metals such as Zn(2+) for mediating nucleic acid catalysis.

Figure 1

Two possible products from the ligation reaction of a 2′,3′-cyclic phosphate RNA substrate. Attack of the 5′-hydroxyl group of another RNA substrate leads to either a native 3′–5′ linkage or a non-native 2′–5′ linkage. When a nucleophile other than the 5′-hydroxyl group attacks the 2′,3′-cyclic phosphate, this forms an unnatural linkage that is neither 3′–5′ nor 2′–5′ (see text).

Figure 2

RNA ligation mediated by new Zn²⁺-dependent deoxyribozymes. (A) The key Zn²⁺-dependent step of each selection round. In the first round, the N₄₀ deoxyribozyme region is fully random, whereas in subsequent rounds, this region is enriched in those 40-nt DNA sequences that are competent for Zn²⁺-dependent RNA ligation. During the selection procedure, the covalent connection (*) between the right-hand RNA substrate and the deoxyribozyme strand (as created by T4 RNA ligase) means that catalytically active DNA sequences attach the left-hand RNA substrate to themselves via the right-hand RNA substrate. Thus the active DNA sequences are readily separable by PAGE; see our previous reports for details of the selection procedure (, ). However, for application of the new deoxyribozymes, the covalent RNA-DNA connection is neither required nor present, and ligation occurs in a trimolecular format (see Figure 3A). (B) Nucleotide sequences of the new Zn²⁺-dependent deoxyribozymes. The sequences correspond to the N₄₀ enzyme region of the deoxyribozyme strand of panel A; the DNA binding arm sequences are complementary to the RNA substrate sequences as listed in the Experimental Section. The 12BB8 enzyme region is only 39 nt long, reflecting a deletion during the selection process.

Figure 3

RNA ligation activity of the Zn²⁺-dependent 6J deoxyribozymes. (A) Trimolecular format of the ligation assay. (B) Representative 20% PAGE image for RNA ligation as mediated by the 6J12 deoxyribozyme (70 mM Tris, pH 7.9, 150 mM NaCl, 2 mM KCl, 1 mM ZnCl₂, 23 °C). (C) and (D) Kinetic plots for 6J12, 6J1, 6J18, and 6J2 under the conditions of panel B, except 37 °C for 6J2. See Table 1 for k_obs values. For 6J2, an additional data point at 240 min falls directly on the fit line but was omitted for clarity. See Supporting Information for full data sets for each of these deoxyribozymes.

Figure 4

Dependence of 6J12 ligation activity on pH, Zn²⁺ concentration, and metal ion identity. (A) The pH dependence of 6J12 at 23 °C. The pH was maintained with 70 mM Tris at the indicated values; the other solution components were 150 mM NaCl, 2 mM KCl, and 1 mM ZnCl₂. At pH 7.1 or 8.5 there was no detectable ligation product (<0.5% yield at 20 min). k_obs values (min⁻¹): pH 7.5, 0.13; pH 7.7, 0.39; pH 7.9, 0.54; pH 8.1, 0.48. (B) The [Zn²⁺] dependence of 6J12 activity (70 mM Tris, pH 7.9, 150 mM NaCl, 2 mM KCl, 23 °C). Yields are the average ± standard deviation of three determinations at each Zn²⁺ concentration. (C) Metal ion dependence of 6J12 activity. The RNA ligation reaction was performed under standard conditions (70 mM Tris, pH 7.9, 150 mM NaCl, 2 mM KCl, 23 °C) in the presence of either 1 mM ZnCl₂, NiCl₂, CoCl₂, or CdCl₂. For the Zn2+ assay, timepoints were taken at 0.5, 2, and 10 min; for the other assays, timepoints were taken at 0.5 min, 4 h, and 24 h. For a more comprehensive version of the experiment showing data for all of these ions and also for CuCl₂ from 10 μM to 10 mM, see Supporting Information.

Figure 5

Ligation activities for individual 12BB deoxyribozymes. Incubation conditions were 70 mM Tris, pH 7.5, 150 mM NaCl, 2 mM KCl, 1 mM ZnCl₂, 23 °C. The 5′-terminal nucleotide of the right-hand RNA substrate was either G (filled symbols) or 2′-deoxy-G (dG; open symbols). See Table 1 for k_obs values.

Figure 6

Analysis of the background ligation products formed with 1 mM Zn²⁺ (or 40 mM Mg²⁺ or Mn²⁺) and an exactly complementary DNA splint instead of a deoxyribozyme. (A) Cleavage assay with the 8–17 deoxyribozyme, which selectively cleaves 3′–5′ linkages (t = 0 to 7 h). (B) Cleavage assay at 100 mM Mg²⁺ and pH 9.0 in the presence of the exactly complementary DNA oligonucleotide, which selectively cleaves 2′–5′ linkages (t = 0 to 7 h). See ref. for the assays of panels B and C performed on standard 3′–5′-linked and 2′–5′-linked RNAs as control experiments that validate this approach. (C) Partial alkaline hydrolysis (AH; 50 mM NaHCO₃, pH 9.2, 90 °C, 10 min) and RNase T1 digestion assays. The brackets mark key bands that distinguish 3′–5′ from 2′–5′ linkages in RNase T1 digestion. The lower of the two bracketed bands observed for the 3′–5′ RNA is present only in greatly reduced intensity for the 2′–5′ RNA. This is because RNase T1 has difficulty cleaving after the G of A↓GG when the indicated ligation junction is 2′–5′, although cleavage after the next G to the 3′-side is unaffected (upper bracketed band).

Figure 7

Cleavage of the 6J ligation products with the 6J deoxyribozymes. Each of the four 6J deoxyribozymes cleaves the 6J2 product (open triangles) with different rate and yield than the products from the other three deoxyribozymes. These kinetic data separate the products into two distinct types, termed here the 6J12 product and the 6J2 product.

Figure 8

Partial alkaline hydrolysis and RNase T1 digestion assays of the ligated RNA products from the 6J12 and 6J2 deoxyribozymes. The ligation products were subjected either to no incubation (–), partial alkaline hydrolysis (AH), or treatment with RNase T1. L = left-hand RNA substrate standard with a 2′,3′-cyclic phosphate. The far right set of lanes are the assays performed on a 3′–5′-linked RNA standard. The sequence on the right side of the gel corresponds to the ligated RNAs with a 3′–5′ linkage. See Figure 6C for these assays performed on a 2′–5′-linked RNA standard. The brackets denote the key difference in the RNase T1 digestion pattern between the 6J products and the 3′–5′ standard, similar to the bands marked in Figure 6C.

Figure 9

Linkage assays for 12BB ligation products. (A) Partial alkaline hydrolysis. For preparation of these products, the 5′-terminal nucleotide of the right-hand RNA substrate was 2′-deoxy-G (dG), as was the case during selection itself. Therefore, in each ladder an expected 1-nt gap is observed at the ligation site, which is denoted by the small arrowhead on the right side. As expected, analogous assays on the products prepared with a 5′-G substrate showed an unbroken ladder (data not shown). However, instead of a regular ladder, the 12BB6 and 12BB8 products show rapid reversion to the initial substrates, indicating that they are neither 3′–5′-linked nor 2′–5′-linked (see text). (B) Cleavage assays with the 8–17 deoxyribozyme (3′–5′-selective; t_1/2 = 12–14 min), performed on the four 12BB products that showed a regular alkaline hydrolysis ladder and are therefore linear. Only the 12BB5 and 12BB12 products are 3′–5′-linked. (C) Cleavage assays at 100 mM Mg²⁺ and pH 9.0 in the presence of the exactly complementary DNA oligonucleotide (2′–5′-selective; t_1/2 = 9–12 h). Only the 12BB1 and 12BB2 products are 2′–5′-linked. (D) RNase T1 digestion of the 12BB6 and 12BB8 products. The brackets enclose the positions of two missing cleavage bands relative to the 3′–5′ standard (compare bracketed bands in Figure 6C and Figure 8). For preparation of all ligation products for panels B–D, the 5′-terminal nucleotide of the right-hand RNA substrate was G.

Figure 10

Cleavage assays to correlate the three 6J12, 6J2, and 12BB8 ligation products. Compare Figure 7, which shows analogous cleavage assays among the four 6J deoxyribozymes and their products.

Figure 11

Possibilities for the unnatural RNA linkages formed by the 6J12, 6J2, and 12BB8 deoxyribozymes. (A) The 3′–2′ and 2′–2′ linear linkages that would be formed by attack of the terminal guanosine 2′-hydroxyl of the right-hand RNA substrate (R) on the 2′,3′-cyclic phosphate of the left-hand RNA substrate (L), with the 2′- or 3′-oxygen of the cyclic phosphate as the leaving group, respectively. (B) The 2′,3′-branched and 2′,2′-branched linkages that would be formed by attack of an internal 2′-hydroxyl of R on the 2′,3′-cyclic phosphate of L, shown here using the second nucleotide’s 2′-hydroxyl group. Analogous branched products are possible using an internal 2′-hydroxyl group from further down the R substrate strand. All unnatural linkages are named systematically following the conventions for naturally occurring 3′–5′ and 2′–5′ linkages (i.e., native and non-native linear linkages) and for naturally occurring 2′,5′-branched linkages, which are distinct from the branches shown here (see Discussion). In all structures, the 3′-terminal adenosine nucleotide of the L substrate (formerly bearing the 2′,3′-cyclic phosphate) and the first three nucleotides of the R substrate (5′-GGA...) are drawn explicitly. The newly formed bond is marked with an arrow; the oxygen of the guanosine 2′-hydroxyl nucleophile is adjacent to the arrowhead.

Figure 12

Experimentally testing the removal of individual 2′-hydroxyl group nucleophiles from the R substrate (one 2′-hydroxyl at a time) with the 6J12, 6J2, and 12BB8 deoxyribozymes. The four 5′-terminal nucleotides of the R substrate used in each kinetic assay are shown (t = 0.5, 2, and 10 min for 6J12; 0.5, 30, and 120 min for 6J2; and 0.5, 15, and 90 min for 12BB8), where (dG) or (dA) denotes the single 2′-deoxy-G or 2′-deoxy-A in each substrate as appropriate. No activity (<0.5%) is detected with the (dG)GAA substrate for 6J12 and 6J2, whereas no activity is detected with the G(dG)AA substrate for 12BB8.