RNA structure-dependent uncoupling of substrate recognition and cleavage by Escherichia coli ribonuclease III

Abstract

Members of the ribonuclease III superfamily of double-strand-specific endoribonucleases participate in diverse RNA maturation and decay pathways. Ribonuclease III of the gram-negative bacterium Escherichia coli processes rRNA and mRNA precursors, and its catalytic action can regulate gene expression by controlling mRNA translation and stability. It has been proposed that E.coli RNase III can function in a non-catalytic manner, by binding RNA without cleaving phosphodiesters. However, there has been no direct evidence for this mode of action. We describe here an RNA, derived from the T7 phage R1.1 RNase III substrate, that is resistant to cleavage in vitro by E.coli RNase III but retains comparable binding affinity. R1.1[CL3B] RNA is recognized by RNase III in the same manner as R1.1 RNA, as revealed by the similar inhibitory effects of a specific mutation in both substrates. Structure-probing assays and Mfold analysis indicate that R1.1[CL3B] RNA possesses a bulge- helix-bulge motif in place of the R1.1 asymmetric internal loop. The presence of both bulges is required for uncoupling. The bulge-helix-bulge motif acts as a 'catalytic' antideterminant, which is distinct from recognition antideterminants, which inhibit RNase III binding.

Figure 1

(Previous page and above) In vitro selection strategy for isolating cleavage-resistant variants of R1.1 RNA. (A) Structure of R1.1[SxN] RNA. The nine sequence-randomized sites (N) in the internal loop are indicated. The inset figure shows the sequence of R1.1 RNA, with the primary (1°) and secondary cleavage sites indicated by the solid and dashed arrows, respectively. The positions of the SacI and XhoI restriction sites are indicated, as are the positions and sequences of the forward PCR primer and reverse transcriptase (RT) primer. (B) In vitro selection strategy. The oligodeoxynucleotide encoding R1.1[SxN] RNA is transcribed to provide an RNA pool, which then is incubated with RNase III. The products are reversed transcribed and the cDNAs subjected to PCR with the primers shown in (A). Only cDNAs corresponding to uncleaved RNA sequences can be amplified, providing a template pool for a new round of transcription and selection. The final amplified DNA sequences are cleaved by XhoI and SacI, and cloned into a pBluescript plasmid. See Materials and Methods for additional information.

Figure 1

(Previous page and above) In vitro selection strategy for isolating cleavage-resistant variants of R1.1 RNA. (A) Structure of R1.1[SxN] RNA. The nine sequence-randomized sites (N) in the internal loop are indicated. The inset figure shows the sequence of R1.1 RNA, with the primary (1°) and secondary cleavage sites indicated by the solid and dashed arrows, respectively. The positions of the SacI and XhoI restriction sites are indicated, as are the positions and sequences of the forward PCR primer and reverse transcriptase (RT) primer. (B) In vitro selection strategy. The oligodeoxynucleotide encoding R1.1[SxN] RNA is transcribed to provide an RNA pool, which then is incubated with RNase III. The products are reversed transcribed and the cDNAs subjected to PCR with the primers shown in (A). Only cDNAs corresponding to uncleaved RNA sequences can be amplified, providing a template pool for a new round of transcription and selection. The final amplified DNA sequences are cleaved by XhoI and SacI, and cloned into a pBluescript plasmid. See Materials and Methods for additional information.

Figure 2

(Previous page and above) RNase III cleavage reactivity of R1.1[SxN] RNA as a function of selection round. (A) Cleavage patterns of R1.1[Sx] RNA, R1.1[SxN] RNA (initial pool) and R1.1[SxN] RNA (round 6). Lanes 1, 3 and 5 show reactions incubated for 40 min in the presence of RNase III (200 nM), but without MgCl₂. Lanes 2, 4 and 6 show reactions incubated for 40 min with RNase III in the presence of MgCl₂. Products were electrophoresed in a 15% polyacrylamide gel containing 7 M urea and were visualized by phosphorimaging. To highlight the cleavage-resistance the amounts of RNA analyzed in lanes 5 and 6 were greater than in the other lanes. The arrow marked ‘1°’ indicates the position of two fragments of approximately equal lengths (53 and 56 nt), created by cleavage of the 109 nt RNA at the primary site (shown in Fig. 2A). ‘US’ and 5′ indicate the additional products of cleavage (both 28 nt in size) at the secondary site. The species indicated by arrowheads indicate products of cleavage at unidentified sites. (B) Resistance of R1.1[SxN] RNA to RNase III cleavage as a function of selection round. Aliquots of ³²P-labeled RNA transcribed from the amplified DNA after each round were incubated with RNase III, then electrophoresed as described above. Percent resistance to cleavage was determined by phosphorimaging (see Materials and Methods).

Figure 2

(Previous page and above) RNase III cleavage reactivity of R1.1[SxN] RNA as a function of selection round. (A) Cleavage patterns of R1.1[Sx] RNA, R1.1[SxN] RNA (initial pool) and R1.1[SxN] RNA (round 6). Lanes 1, 3 and 5 show reactions incubated for 40 min in the presence of RNase III (200 nM), but without MgCl₂. Lanes 2, 4 and 6 show reactions incubated for 40 min with RNase III in the presence of MgCl₂. Products were electrophoresed in a 15% polyacrylamide gel containing 7 M urea and were visualized by phosphorimaging. To highlight the cleavage-resistance the amounts of RNA analyzed in lanes 5 and 6 were greater than in the other lanes. The arrow marked ‘1°’ indicates the position of two fragments of approximately equal lengths (53 and 56 nt), created by cleavage of the 109 nt RNA at the primary site (shown in Fig. 2A). ‘US’ and 5′ indicate the additional products of cleavage (both 28 nt in size) at the secondary site. The species indicated by arrowheads indicate products of cleavage at unidentified sites. (B) Resistance of R1.1[SxN] RNA to RNase III cleavage as a function of selection round. Aliquots of ³²P-labeled RNA transcribed from the amplified DNA after each round were incubated with RNase III, then electrophoresed as described above. Percent resistance to cleavage was determined by phosphorimaging (see Materials and Methods).

Figure 3

Cleavage resistance of Class I R1.1 RNA variants, and involvement of the A9U mutation. (A) Analysis of R1.1[CL68B] RNA. The sequence of the RNA is shown on the left. The right side presents a cleavage assay showing the resistance of internally ³²P-labeled R1.1[CL68B] RNA to RNase III (lanes 1 and 2), and restoration of reactivity by the U9A reversion (lanes 3 and 4). The reactions involved incubation with 10 nM enzyme for 40 min. The four bands in lane 4 include the products of cleavage at the primary site (indicated by 1° and 3′), or at the primary and secondary sites (indicated by US and 5′; US, upper stem). (B) Analysis of R1.1[CL14A] RNA reactivity. The left side shows the structure of R1.1[CL14A] RNA. The assay on the right shows the cleavage resistance of internally ³²P-labeled substrate (compare lanes 1 and 2) and restoration of reactivity by the U9A reversion (lanes 3 and 4). The reaction conditions were the same as the experiment in (A).

Figure 4

Analysis of the Class II RNA, R1.1[CL3B] RNA. (A) The sequence of R1.1[CL3B] RNA. The nucleotide changes in the internal loop are indicated by bold font. The sequence of the R1.1 RNA internal loop is shown in the inset. Also shown is the distal box (db) and the UG→GU ‘db17’ mutation used in the experiment shown in Figure 5. (B) Cleavage assay of R1.1[CL3B] RNA (lanes 7–12) and R1.1 RNA (lanes 1–6) as a function of RNase III concentration. Lanes 1 and 7 represent incubation of RNA for 10 min in the absence of RNase III. Lanes 2–6 and lanes 8–12 show 10 min reactions using RNase III at a concentration of 2, 5, 8, 50 and 100 nM, respectively. The small amount of cleavage of R1.1[CL3B] RNA observed in lane 12 corresponds to 4.6% conversion to product, with cleavage occurring at the canonical cleavage site, based on similar gel electrophoretic mobilities in a denaturing gel (20). (C) Gel shift assay of R1.1[CL3B] RNA (lanes 1–6) and R1.1 RNA (lanes 7–12). Shift assays were performed using RNase III in the presence of Ca²⁺, which enhances substrate binding but does not support catalysis (see Materials and Methods). Lanes 1 and 7 show the mobility of the two RNAs in the absence of added RNase III. Lanes 2–6 and 7–12 represent RNase III concentrations of 2, 5, 8, 50 and 100 nM, respectively.

Figure 5

Common mechanism of RNase III recognition of R1.1[CL3B] RNA and R1.1 RNA. (A) Gel shift assays were performed using 5′ ³²P- labeled RNA, and RNase III in Ca²⁺-containing buffer (see Materials and Methods). Lanes 1–3, R1.1 RNA. Lanes 4–6, R1.1[CL3B] RNA. Lanes 7–9, R1.1[db17] RNA. Lanes 10–12, R1.1[CL3B;db17] RNA. See Figure 4A for the location of the db17 mutation. Lanes 1, 4, 7 and 10 show RNA mobility in the absence of RNase III. The paired lanes 2 and 3, 5 and 6, 8 and 9, and 11 and 12 show the mobilities of each RNA in the presence of 50 or 100 nM RNase III, respectively. The positions of the free and bound RNAs are indicated. (B) Graphic presentation of the effect of the db17 mutation on RNase III binding to R1.1[CL3B] RNA and R1.1 RNA. The graph shows the percent RNA bound, as determined by the ratio of the amount of bound RNA to the total (bound plus unbound) RNA. Two RNase III concentrations used were 50 and 100 nM. Shown to the right is the key to the RNAs examined. The binding of R1.1[CL3B; db17] RNA was undetectable.

Figure 5

Common mechanism of RNase III recognition of R1.1[CL3B] RNA and R1.1 RNA. (A) Gel shift assays were performed using 5′ ³²P- labeled RNA, and RNase III in Ca²⁺-containing buffer (see Materials and Methods). Lanes 1–3, R1.1 RNA. Lanes 4–6, R1.1[CL3B] RNA. Lanes 7–9, R1.1[db17] RNA. Lanes 10–12, R1.1[CL3B;db17] RNA. See Figure 4A for the location of the db17 mutation. Lanes 1, 4, 7 and 10 show RNA mobility in the absence of RNase III. The paired lanes 2 and 3, 5 and 6, 8 and 9, and 11 and 12 show the mobilities of each RNA in the presence of 50 or 100 nM RNase III, respectively. The positions of the free and bound RNAs are indicated. (B) Graphic presentation of the effect of the db17 mutation on RNase III binding to R1.1[CL3B] RNA and R1.1 RNA. The graph shows the percent RNA bound, as determined by the ratio of the amount of bound RNA to the total (bound plus unbound) RNA. Two RNase III concentrations used were 50 and 100 nM. Shown to the right is the key to the RNAs examined. The binding of R1.1[CL3B; db17] RNA was undetectable.

Figure 6

Structural analysis of R1.1[CL3B] RNA. (A) Tb³⁺ ion structure probing. Terbium ion (Tb³⁺)-dependent RNA cleavage was carried out essentially as described (38,39) with some modification. TbCl₃ was dissolved in 5 mM HEPES (pH 5.5) at a final concentration of 0.5 M and stored at –20°C prior to use. Briefly, 5′ ³²P-labeled RNA (∼15 000–20 000 c.p.m.) was incubated with 5 mM TbCl₃ in 20 mM NaCl, 50 mM HEPES (pH 7.5) in a 10 µl reaction for 15 min at 37°C. The reaction was stopped by EDTA (2 µl of a 250 mM solution). Samples were ethanol precipitated and resuspended in a small volume of TE buffer. An equal volume of deionized formamide (containing 0.025% bromophenol blue) was added, and the sample electrophoresed (1500 V, 2–4 h) in a 10% polyacrylamide gel containing 7M urea. The reactions were visualized by phosphorimaging. Lanes 1–3, R1.1[CL3B] RNA. Lanes 4–6, R1.1[WC-R] RNA. The latter RNA migrates faster in the gel due to a persistent secondary structure. Lanes 7–9, R1.1 RNA. Lanes 3, 6 and 9 represent the products of incubation with TbCl₃. Lanes 1, 4 and 7 show an alkaline ladder obtained by heating 5′ ³²P-labeled RNA at 90°C for 10 min in 1 mM sodium carbonate buffer (pH 9.2). Lanes 2, 5 and 8 represent partial RNase U2 (A>G-specific) reactions, obtained by incubating RNA at 55°C for 15 min with 10 U of RNase U2 in 25 mM sodium citrate (pH 3.5), 5 M urea, 0.75 mM EDTA and 0.5 mg/ml tRNA. Shown are the positions of the R1.1 internal loop (IL) sequences and the tetraloop sequence. (B) Mfold structures for R1.1 RNA and R1.1[CL3B] RNA, also showing the sites of significant cleavage by Tb³⁺.

Figure 6

Structural analysis of R1.1[CL3B] RNA. (A) Tb³⁺ ion structure probing. Terbium ion (Tb³⁺)-dependent RNA cleavage was carried out essentially as described (38,39) with some modification. TbCl₃ was dissolved in 5 mM HEPES (pH 5.5) at a final concentration of 0.5 M and stored at –20°C prior to use. Briefly, 5′ ³²P-labeled RNA (∼15 000–20 000 c.p.m.) was incubated with 5 mM TbCl₃ in 20 mM NaCl, 50 mM HEPES (pH 7.5) in a 10 µl reaction for 15 min at 37°C. The reaction was stopped by EDTA (2 µl of a 250 mM solution). Samples were ethanol precipitated and resuspended in a small volume of TE buffer. An equal volume of deionized formamide (containing 0.025% bromophenol blue) was added, and the sample electrophoresed (1500 V, 2–4 h) in a 10% polyacrylamide gel containing 7M urea. The reactions were visualized by phosphorimaging. Lanes 1–3, R1.1[CL3B] RNA. Lanes 4–6, R1.1[WC-R] RNA. The latter RNA migrates faster in the gel due to a persistent secondary structure. Lanes 7–9, R1.1 RNA. Lanes 3, 6 and 9 represent the products of incubation with TbCl₃. Lanes 1, 4 and 7 show an alkaline ladder obtained by heating 5′ ³²P-labeled RNA at 90°C for 10 min in 1 mM sodium carbonate buffer (pH 9.2). Lanes 2, 5 and 8 represent partial RNase U2 (A>G-specific) reactions, obtained by incubating RNA at 55°C for 15 min with 10 U of RNase U2 in 25 mM sodium citrate (pH 3.5), 5 M urea, 0.75 mM EDTA and 0.5 mg/ml tRNA. Shown are the positions of the R1.1 internal loop (IL) sequences and the tetraloop sequence. (B) Mfold structures for R1.1 RNA and R1.1[CL3B] RNA, also showing the sites of significant cleavage by Tb³⁺.

Figure 7

Both sets of bulged nucleotides in R1.1[CL3B] RNA are required to uncouple binding and cleavage. RNAs were synthesized in 5′ ³²P-labeled form (see Materials and Methods). RNAs were incubated in the presence or absence of RNase III (50 nM) for 10 min at 37°C, then electrophoresed in a denaturing polyacrylamide gel and analyzed by phosphorimaging (see Materials and Methods). Lanes 1 and 2, R1.1 RNA. Lanes 3 and 4, R1.1[CL3B] RNA. Lanes 5 and 6, R1.1[CL3B;ΔA₄₈] RNA. Lanes 7 and 8, R1.1[CL3B;ΔA₄₈ΔU₄₉] RNA. Lanes 9 and 10, R1.1[CL3B;ΔA₂₀;ΔA₄₈ΔU₄₉] RNA. Note the significantly greater electrophoretic mobility of the deletion variants, due to their shorter lengths and perhaps also to an altered secondary structure in the presence of 7 M urea.

Figure 8

Proposed alternative conformation of the Class I variant R1.1[CL68B] RNA, and the involvement of the A9U mutation. On the left is R1.1[CL68B] RNA lacking the A9U mutation. Shown are the positions of the proximal and distal boxes, and the RNase III cleavage site. In the presence of the A9U mutation, the bracketed nucleotides are proposed to pair, producing the structure shown on the right. Note the disruption of the proximal box (pb), as well as loss of the R1.1 internal loop and lower stem.