| Nucleic Acids Research | Pages |
©1999 Oxford University Press |
Design and isolation of ribozyme-substrate pairs using RNase P-based ribozymes containing altered substrate binding sites
Introduction
Materials And Methods
Cloning of the modified ribozymes
The RNA substrate libraries
In vitro selection
Characterization of the selected substrates
Results And Discussion
Design of the ribozymes and the substrate libraries
The progress of selection
Characterization of six individual variants
Probing ribozyme-substrate interactions by structural mapping methods
Design and isolation of P RNA-based ribozyme-substrate pairs
Acknowledgements
References
Design and isolation of ribozyme-substrate pairs using RNase P-based ribozymes containing altered substrate binding sites
Received April 12, 1999; Revised July 30, 1999; Accepted September 7, 1999
ABSTRACT Substrate recognition and cleavage by the bacterial RNase P RNA requires two domains, a specificity domain, or S-domain, and a catalytic domain, or C-domain. The S-domain binds the T stem-loop region in a pre-tRNA substrate to confer specificity for tRNA substrates. In this work, the entire S-domain of the Bacillus subtilis RNase P RNA is replaced with an artificial substrate binding module. New RNA substrates are isolated by in vitro selection using two libraries containing random regions of 60 nt. At the end of the selection, the cleavage rates of the substrate library are ~0.7 min-1 in 10 mM MgCl2 at 37°C, ~4-fold better than the cleavage of a pre-tRNA substrate by the wild-type RNase P RNA under the same conditions. The contribution of the S-domain replacement to the catalytic efficiency is from 6- to 22 000-fold. Chemical and nuclease mapping of two ribozyme-product complexes shows that this contribution correlates with direct interactions between the S-domain replacement and the selected substrate. These results demonstrate the feasibility of design and isolation of RNase P-based, matching ribozyme-substrate pairs without prior knowledge of the sequence or structure of the interactive modules in the ribozyme or substrate.
INTRODUCTION
The ribozyme from bacterial RNase P is unique among natural ribozymes in its mode of substrate recognition (1-5). RNase P performs an endonucleolytic cleavage reaction to generate the mature 5[prime]-end of all tRNAs in vivo. The RNase P RNA (denoted P RNA) binds two structured regions in a natural pre-tRNA substrate, the T stem-loop and the acceptor stem/5[prime]-leader. In addition to the natural substrates, several classes of different RNA substrates have been isolated by in vitro selection using the wild-type Bacillus subtilis P RNA (6,7) or the Escherichia coli P RNA (8). The secondary structure of these substrates can be divided into three structural modules similar to the acceptor stem/5[prime]-leader, the T stem-loop, and the coaxially stacked linker of a natural tRNA substrate (5).
Recognition of the selected substrates by P RNA can be summarized on the basis of its domain structure (Fig. 1A and B; 9). P RNA is composed of two independently folding domains. The C-domain contains the entire active site and binds the acceptor stem/5[prime]-leader region of a tRNA substrate. The S-domain binds the T stem-loop region of a tRNA substrate and confers the substrate specificity. Chemical modification and ribozyme mutagenesis revealed that the T stem-loop-like module in three of the four classes of selected substrates is recognized by the S-domain (Fig. 1B). These three substrate classes contain a similar linker structure. The significantly different linker region in the fourth substrate class directs one module to interact with a hairpin loop region (designated the L1 module) in the C-domain (Fig. 1B; 7). Deletion of the S-domain has little effect on binding and cleavage of substrates in this fourth class, whereas deletion of the L1 module has little effect on binding and cleavage of a pre-tRNA or substrates in the three other classes.
Figure 1. Substrate binding by the B.subtilis P RNA-based ribozymes. The shaded components of the ribozyme include the C-domain and the two substrate binding regions, the S-domain or S-replacement and the L1 module. The boxed components of the substrate include the short helix around the cleavage site and the region interacting with the binding sites in the ribozyme. The linker region in the substrate is shown as thick, dashed lines. The cleavage site is indicated by an arrow and the 5[prime]-leader regions are shown as a thin line. (A) Binding of a pre-tRNA substrate by the wild-type P RNA (3,4,23). (B) Binding of four previously selected substrates by the wild-type P RNA (5). (C) Expected binding modes of the selected substrates isolated in this work by XS1 and XS2. (i) Both the S-replacement and the L1 module in the ribozyme are involved in binding (as shown for variant 1). (ii) The L1 module in the ribozyme is primarily involved in binding (as shown for variant 2). (iii) The S-replacement in the ribozyme is primarily involved in binding. No substrate has been identified so far that only binds to the S-replacement. Presumably, more stringent selection conditions including deletion of the L1 module in XS1 and XS2 are needed to obtain this kind of substrate.
These previous selection results suggest that the substrate binding properties of P RNA may be exploited to isolate new ribozyme-substrate pairs by in vitro selection. In this work, the entire S-domain in the B.subtilis P RNA is replaced with an artificial substrate binding module (designated S-replacement; Fig. 1C). Two RNA libraries are designed to allow the randomized regions to interact with S-replacement. Since the S-domain in the wild-type P RNA is proposed to be in proximity with the L1 module in the three-dimensional structure (4,10), however, S-replacement may act together with the L1 module in recognition of the selected substrates. It is therefore expected that the selected substrate may interact with both S-replacement and the L1 module [Fig. 1C, scenario (i)], or only with S-replacement or just the L1 module [Fig. 1C, scenarios (ii) and (iii)]. Many new substrates are isolated in this work; several variants are more efficiently cleaved by as much as 30-fold compared to cleavage of a pre-tRNA substrate by the wild-type P RNA under the same conditions. These selected substrates interact with S-replacement and with the L1 module to different extents.
MATERIALS AND METHODS
Cloning of the modified ribozymes
The entire S-domain of the B.subtilis P RNA was replaced by an artificial substrate binding module containing the circularly permuted yeast tRNAPhe sequence minus the D stem-loop (nt 27-76 + 1-9 of tRNAPhe; Fig. 2A). To generate ribozymes with S-replacement, two complementary DNA oligonucleotides, 5[prime]-TAAATCCGCTGGTGCGAATTCTGTGGATCGAACACA-GGACCTCCAGATCTTCAGTCTGG and 5[prime]-CCAGACTGAA-GATCTGGAGGTCCTGTGTTCGATCCACAGAATTCGCAC-CAGCGGATTTA, were annealed and ligated to the 3[prime]-end of two previously cloned C-domain constructs. These previously described C-domain constructs (designated XS1[Delta]S and XS2[Delta]S; 9) were also used to determine the contribution of S-replacement to the catalytic efficiency of the selected substrates.
Figure 2. The sequence and secondary structure representation (24) of the P RNA-based ribozymes and RNA libraries. (A) Both ribozymes contain the C-domain of the B.subtilis P RNA (residues 240-409 + 1-85). The S-domain (residues 86-239) is replaced by an RNA structure originating from the circularly permuted yeast tRNAPhe minus the D stem-loop. Compared to XS1, the insertion of two helices from the S-domain (P7 and P8) in XS2 changes the relative geometry between the active site in the C-domain and S-replacement. The other known substrate binding site in the C-domain (L1 module; 5,7) is not changed. According to the structural models of P RNA (3,4), S-replacement and the L1 module are supposed to be in close proximity in the tertiary structure of XS1 and XS2. (B) The RNA libraries contain 60 randomized nucleotides and have different linker regions (shaded).
The RNA substrate libraries
Two libraries were designed to allow interactions between the randomized regions with S-replacement and/or the unchanged L1 module in the C-domain (Fig. 2B). Both libraries had 60 randomized nucleotides, but different linker regions. These libraries were constructed by five cycles of PCR using the DNA oligonucleotides 5[prime]-TGGTGGATCTTTAGTCTGGA[N60]-CCCAGACTTGAGCGAATC and 5[prime]-TGGTGCGAATTC[N60]-GTGAATCCGC as templates and 5[prime]-TAATACGACTCAC-TATAGCGAACGCTTCACGGATTCGCTCAAGTCTGG/5[prime] TGGTGGATCTTTAGTCTGGA and 5[prime]-TAATACGACTCA-CTATAGCGAACGCTTCACGGCGGATTCAC/5[prime]-TGGTG-CGAATTC as primers, respectively (6). The PCR generated templates were used directly for RNA synthesis by in vitro transcription using T7 RNA polymerase (11).
In vitro selection
The libraries were designed to contain fixed cleavage sites, so that active substrates could be isolated by separating the 3[prime] cleavage product from the unreacted RNA on a denaturing gel (8). The selections were carried out at a ribozyme:substrate ratio of 1:5 with increasingly stringent conditions. Two parameters were controlled in the selection: the Mg2+ concentration and the concentration of the ribozyme and the substrate library (the detailed selection conditions are described in Fig. 3). In both selections, the Mg2+ concentration was 25 mM in early rounds and decreased to 10 mM in later rounds; the ribozyme/substrate concentration was 1 µM/5 µM in early rounds and decreased to 0.04 µM/0.2 µM in later rounds. Renaturation of the ribozyme and substrate libraries was identical to our previous selections (6,9). The ribozyme and substrate in buffer only were heated at 85-90°C for 2 min, followed by incubation at ambient temperature for 3 min. MgCl2 was added and the ribozyme or substrate was incubated at 50 or 37°C for 5 min, respectively. The cleavage reaction was initiated by mixing the renatured ribozyme and the substrate. Aliquots were taken at different time points and the cleavage product separated from unreacted substrate on polyacrylamide gels containing 7 M urea. For every round, the percentage of 32P-labeled cleavage product was determined by phosphorimaging and the cleavage rate was calculated from the slope of the initial 20% product. The cleavage product was eluted from the gel by the crush-and-soak method. The procedures of cDNA synthesis and PCR amplification were identical to those described in the early work (6) and according to the manufacturer's protocols for using AMV reverse transcriptase and Taq DNA polymerase (Amersham Life Science).
Figure 3. (A) The progress of selection of the XS1 library with the XS1 ribozyme. Cycles 1-7 were carried out in 50 mM Tris-HCl, pH 8, 25 mM MgCl2 at 1 µM ribozyme and 5 µM substrate. Cycles 8-18 were performed in 10 mM MgCl2 at different ribozyme:substrate concentrations (1 µM E:5 µM S, 0.2 µM E:1 µM S and 0.04 µM E:0.2 µM S). kobs was not determined for cycles 1, 2, 13 and 17. The conditions for the final selection cycle were 50 mM Tris-HCl, pH 8, 10 mM MgCl2, 37°C, 0.04 µM ribozyme and 0.2 µM substrate. (B) The progress of selection of the XS2 library with the XS2 ribozyme. Cycles 1-10 were carried out in 50 mM Tris-HCl, pH 8, 25 mM MgCl2 at different ribozyme:substrate concentrations (1 µM E:5 µM S, 0.2 µM E:1 µM S and 0.04 µM E:0.2 µM S). Cycles 11-13 were performed in 10 mM MgCl2, 0.04 µM ribozyme and 0.2 µM substrate. kobs was not determined for cycle 10.
Characterization of the selected substrates
The enriched substrate library was cloned and individual variants sequenced. The location of the cleavage site was determined by comparing the 5[prime]-32P-labeled cleavage product to an alkaline hydrolysis and partial nuclease T1 digestion ladder of the same substrate. Terminal truncation experiments were performed to identify residues from the 5[prime]- and 3[prime]-ends that could be deleted without affecting cleavage efficiency. Briefly, trace amounts of 5[prime]- or 3[prime]-32P-labeled substrates were hydrolyzed by boiling for 45 s in 1 mM glycine, 0.4 mM MgSO4, pH 9.5. The hydrolyzed RNA was neutralized by the addition of Tris-HCl, pH 8, to a final concentration of 50 mM, renatured as described above and cleaved by 0.3 µM XS1 ribozyme for up to 15 min at 37°C. The reaction mixture was then analyzed by 8% PAGE containing 7 M urea.
Structural mapping using nucleases V1 and T1 was performed as previously described with minor modifications (7,12). Briefly, reactions were carried out in 50 mM Tris-HCl, pH 8, 10 mM MgCl2 using end-labeled RNA substrates. The RNA was renatured as described above at a final concentration of 0.2 µM. Nuclease was then added to 0.2-0.4 mU/µl V1 and 0.4 U/µl T1 and the reaction proceeded for 5 min at 37°C. An equal volume of denaturing gel loading buffer (9 M urea, 100 mM EDTA) was added and the mixture quickly cooled on ice. The mixture was immediately loaded on 8 and 20% denaturing PAGE gels.
The ribozyme-substrate and the ribozyme-3[prime]-product complexes were probed by partial nuclease V1 and T1 digestion to identify residues in the substrate that are protected upon ribozyme binding. The ribozyme-substrate/product complexes were formed in 50 mM Tris-HCl, pH 8, 10 mM MgCl2, and 1 mM spermine using end-labeled substrates or 3[prime] cleavage products. The final ribozyme concentrations were 1.5 µM for variant 1 and 5 µM for variant 2, corresponding to five times the Km of the cleavage reaction under these conditions. In the absence of the ribozyme, unlabeled substrate was added to maintain the equivalent phosphate concentration. The ribozyme-substrate mixture was preincubated at 37°C for 5 min prior to the addition of the nuclease to 2-4 mU/µl V1 and 0.1 U/µl T1. The reactions were carried out at 37°C for 1 min, then treated as described under structural mapping.
To identify regions in S-replacement and the L1 module of the ribozyme that interact with the substrate, Fe(II)-EDTA footprinting of the ribozyme was carried out in the presence and absence of a molar excess of variants 1 and 2. The ribozyme and substrates were renatured separately as described above. The ribozyme-substrate/product complexes were formed in 50 mM Tris-HCl, pH 8, 10 mM MgCl2, and 1 mM spermine by preincubation of 0.1 µM ribozyme with 1.5 µM variant 1 or 6 µM variant 2 at 37°C overnight. Hydroxyl radical protection was performed using end-labeled ribozyme as described previously (13,14). The reactions were carried out in 1 mM Fe(II), 1.2 mM EDTA at 37°C for 30 min, then quenched by the addition of thiourea at a final concentration of 10 mM. An equal volume of denaturing gel loading buffer (9 M urea, 100 mM EDTA) was added and the mixture was analyzed on 15 or 6% PAGE gels.
The kcat/Km and the cleavage rate at saturating ribozyme concentrations (kc) under single turnover conditions (i.e. [E] >> [S]) were measured under the solution conditions of the final selection cycle (50 mM Tris-HCl, pH 8, 10 mM MgCl2, 37°C). These characterizations employed standard protocols in nucleic acid biochemistry with minor modifications as previously described (5,7). kcat/Km was measured using 10-80 nM XS1 or XS2 ribozyme or 2 µM XS1[Delta]S or XS2[Delta]S ribozyme and trace amounts of 5[prime]-32P-labeled substrates. kc was determined using >1 µM XS1 or XS2 ribozyme.
RESULTS AND DISCUSSION
Design of the ribozymes and the substrate libraries
The two ribozymes used in this work contained the entire C-domain of the B.subtilis P RNA originating from circularly permuted constructs with the 5[prime]-end at nt 240 or 235 (9,15). The entire S-domain of the B.subtilis P RNA was deleted and an artificial substrate binding module was inserted in its place. To increase the success of selection, two different ribozymes were constructed (Fig. 2A). S-replacement contains the portion of the yeast tRNAPhe in its circularly permuted form minus the D stem-loop. The truncated tRNA was chosen for the possibility that the conserved D stem-loop sequence may be selected in the substrate to restore the tertiary structure of tRNA. In principle, however, any RNA structure may be used to replace the S-domain. Similarly, any RNA structure that can interact with S-replacement in any way may be selected. The difference between these two ribozymes is the insertion of the P7 and P8 stem-loops of B.subtilis P RNA in XS2. The presence of P7/P8 changes the distance and/or orientation of the active site in the C-domain relative to S-replacement and therefore generates two distinct binding modules.
The two substrate libraries used in the selection contained a randomized region of 60 nt designed to interact with S-replacement in the ribozyme (Fig. 2B). Two different linker regions were used to increase our chance of success. The linker region in the XS1 library was modeled after the linker of a tRNA substrate. The linker region in the XS2 library was modeled after the linker of a previously selected substrate (substrate 8 in Fig. 1B). As derived from the P RNA models, S-replacement and the L1 module are likely to be proximal in the three-dimensional structure (3,4). Therefore, substrates from both libraries are expected to interact with either S-replacement, the L1 module or both (Fig. 1C).
The progress of selection
Matching ribozyme-substrate pairs were selected at increasing stringency with regard to the ionic conditions and ribozyme/substrate concentrations (Fig. 3). After 13-18 rounds of selection, both libraries had cleavage rates of 0.6-0.7 min-1, which is ~4-fold better than the cleavage of a pre-tRNAPhe substrate by the wild-type P RNA under the same conditions. The selected substrate mixture was then cloned and 63 clones were sequenced and tested for activity. Consistent with the final selection cycle, all individual substrates were cleaved at an equivalent rate of 0.1-2 min-1 under the final selection conditions. When assayed with the XS1 and XS2 ribozymes with or without S-replacement, cleavage of individual variants was decreased between 2- and 20 000-fold (kcat/Km) upon deletion of S-replacement. The wide dispersion of the deletion effect suggests that S-replacement and the L1 module are involved in the binding of selected substrates to different extents. Some variants have complementary sequences to the `anticodon loop' and the `T loop' in S-replacement and/or complementary sequences to the `L1 loop' in the L1 module. Without further investigation, it is uncertain whether these complementary sequences are directly involved in ribozyme-substrate interactions. Some other variants had no complementary sequences to either S-replacement or the L1 module. A `D stem-loop'-like sequence was not found in any variant, suggesting that the tRNA-like interaction is not efficient for high affinity binding in generating ribozyme-substrate complexes.
Characterization of six individual variants
Six variants were chosen for further characterization on the basis of their fast cleavage rates and the differential effects upon deletion of S-replacement (Table 1). We chose to determine the kcat/Km and the cleavage rate at saturating ribozyme concentrations (kc) in single turnover reactions (i.e. [E] >> [S]) to compare the catalytic efficiency of these substrates (16,17). The kcat/Km term describes the combined effects from binding and cleavage chemistry of a ribozyme-substrate pair, whereas the kc term describes the chemical step after the substrate is bound to the ribozyme. At 10 mM MgCl2, pH 8 and 37°C, the kcat/Km of these substrates is between 0.7 and 5.4 µM-1 min-1 under single turnover conditions, 4- to 27-fold better than cleavage of a pre-tRNAPhe substrate by the wild-type P RNA. The cleavage rate at saturating ribozyme concentrations (kc) was between 0.5 and 9 min-1, indicating that these selected variants are very efficient substrates in a RNA-catalyzed reaction.
Table 1. Ribozyme cleavage of six selected substrates under single turnover conditions
| Ribozyme/substrate | Substrate sequence (randomized region)a | kcat/Km (µM-1 min-1)b | kc (min-1)b | Relative reactivityc |
| XS1/1 | aaaac aaacu gaucg aacgu cacgg uccgc caccc agcuc uucac ugccc ccccc | 5.0 | 3.4 | 22 000 |
| XS1/2 | agcuc uaucg ccuga cacaa cggua ugacu gcccc cgugc ccacc ccccc c | 1.0 | 1.2 | 6 |
| XS2/3 | auauc gagac uccca aaugu uucug gugac ggguc ucuuc aacuc agucc accuc cucug | 3.3 | 3.1 | 17 |
| XS2/4 | gacau accac agaca cauug uagug gcuag agugg caaau gacuu cagca ugcag guccc | 0.7 | 5.2 | 1 600 |
| XS2/5 | acaca cccuc ugggu uggag cucuc uagcc acugc gaacu cuuca cucgc uuuuc gcucc | 5.4 | 8.8 | 640 |
| XS2/6 | agucg caucc uggac uuggg ccguc uugga cgugc gacca gacca ucgcu gacgu ugaug | 1.0 | 0.5 | 80 |
bConditions: 50 mM Tris-HCl, pH 8, 10 mM MgCl2, 37°C. The cleavage of a pre-tRNAPhe substrate by the wild-type P RNA under these conditions has a kcat/Km of 0.2 µM-1 min-1.
ckcat/Km (XS1 and XS2) divided by kcat/Km (XS1[Delta]S and XS2[Delta]S) under identical conditions.
Terminal truncation experiments (18) were performed to determine how many nucleotides can be deleted from the 5[prime]- and 3[prime]-ends without affecting cleavage efficiency (Fig. 4). Between 20 and 23 nt can be deleted from the 3[prime]-end of variants 1 and 2. These deleted nucleotides include a part of the designed linker in the original XS1 library. The dispensability of the designed linker suggests that variants 1 and 2 had acquired a new linker involving nucleotides from the randomized region. In contrast to variants 1 and 2, no 3[prime] nucleotides can be deleted from variants 3-6 without affecting cleavage efficiency (data not shown), suggesting that the designed linker in variants 3-6 is retained. Three nucleotides in the 5[prime]-leader region are sufficient to confer full reactivity for variants 1-6 (data not shown). This 5[prime]-leader requirement is identical to the cleavage of a pre-tRNA substrate by the wild-type P RNA (19,20).
Figure 4. Nuclease V1 and T1 mapping of variants 1 and 2 is applied to probe the secondary structure of these variants. Terminal truncation experiments of these variants were carried out to determine the number of nucleotides that can be deleted from the 3[prime]-end without affecting the catalytic efficiency. The results are summarized in Figure 5B.
The cleavage site of these selected substrates can be mapped by comparing the 5[prime]-32P-labeled cleavage product to partial alkaline hydrolysis and nuclease T1 digestion of the same RNA (data not shown). The cleavage sites of variants 1 and 2 do not coincide with the designated site in the XS1 library (Fig. 5). This result is consistent with terminal truncation showing that the linker in variants 1 and 2 has changed. The cleavage sites of variants 3-6 are located at the predicted positions (Fig. 2B), consistent with retention of the linker in these variants.
A
 |
B
 |
Figure 5. (A) The binding of variants 1 and 2 by the XS1 ribozyme was probed by their protection against nuclease digestion. The protected regions are marked by a thick line, whereas a region with enhanced V1 cuts are marked by a double thin line. (B) Summary of the nuclease mapping and protection data for variants 1 and 2 and their proposed secondary structure. The cleavage site is between the hatched nucleotides shown in bold. The 5[prime]-leader region is shown in lower case. The boxed nucleotides at the 5[prime]- and 3[prime]-ends can be deleted without affecting cleavage efficiency. Nucleotides that are protected upon ribozyme binding are shaded. Nucleotides that are cut more strongly by nuclease V1 are circled. Regions that are proposed to interact with S-replacement or the L1 module are indicated by oval lines.
Probing ribozyme-substrate interactions by structural mapping methods
Due to their extreme difference in catalytic efficiency upon deletion of S-replacement (Table 1), variants 1 and 2 were probed further for their interactions with the ribozyme. The secondary structures of these variants were analyzed by nuclease mapping (Fig. 4). Consistent with terminal truncation and relocation of the cleavage sites, the proposed secondary structure for variants 1 and 2 was different from the design (Fig. 5B).
Regions in variants 1 and 2 that interact with the ribozyme were identified by changes in the nuclease digestion pattern upon ribozyme binding (Fig. 5). Two regions in variant 1 can be proposed to interact with the ribozyme. One region involves approximately nt 21-29. Several nucleotides in this region are protected against nuclease digestion, while other nucleotides show enhanced nuclease V1 cuts upon ribozyme binding (Fig. 5A). The increased V1 cuts suggest that these nucleotides are involved in forming an intermolecular structure with the ribozyme. These nucleotides are complementary in sequence to the `T stem-loop' region in S-replacement. Therefore, it is likely that this region interacts with S-replacement. The other protected region involves approximately nt 38-48. This region probably interacts with a different region in the ribozyme due to their distal location to nt 21-29. A single region of ~8 nt in variant 2 was protected against nuclease digestion. Since S-replacement has only a small effect on the cleavage efficiency of variant 2 (Table 1), this region in the substrate is likely to interact with a region in the ribozyme other than S-replacement.
The interaction of variants 1 and 2 with S-replacement and the L1 module of the XS1 ribozyme was also demonstrated by ribozyme protection from hydroxyl radical attack upon substrate binding (Fig. 6). As indicated by the catalytic activity (Table 1) and by nuclease digestion (Fig. 5), variant 1 protects two regions in S-replacement of the ribozyme, including the `T stem-loop' implicated as interacting with nt 21-29 in this variant. These protections in the XS1 ribozyme were not seen upon binding of variant 2 (data not shown). A portion of the L1 module is protected by both substrates, suggesting that the L1 module is directly involved in the binding of both variants.
Figure 6. Protection of S-replacement (3[prime]-32P-label) and the L1 module (5[prime]-32P-label) in the XS1 ribozyme upon binding of variants 1 and 2 as probed by hydroxyl radical protection. The protected regions are indicated by thick lines.
In summary, variant 1 appears to interact with S-replacement and the L1 module, whereas variant 2 appears to interact primarily with the L1 module. Since variants 1 and 2 exhibited maximal and minimal effects on cleavage efficiency upon deletion of S-replacement, we propose that all other selected substrates obtained in this work can be described by the two proposed modes of interaction shown in Figure 1C [scenarios (i) and (ii)]. The failure to obtain any substrate that does not interact with the L1 module suggests that the L1 module may play a dominant role in substrate binding upon deletion of the natural S-domain. Deletion of the L1 module from the XS1 and XS2 ribozymes may be necessary to isolate substrates that exclusively interact with S-replacement.
Design and isolation of P RNA-based ribozyme-substrate pairs
The selection results described here are a logical extension from our previous work using only the wild-type P RNA. Based on the characteristics of the previously selected substrates, we have proposed a strategy to isolate new P RNA-based ribozyme-substrate pairs (5). These new ribozyme-substrate pairs would involve compensatory substitution of one or both binding modules in the P RNA with the corresponding binding module in the substrate. In the P RNA system, any RNA structure can in principle be used as S-replacement, provided that another RNA structure can interact with this S-replacement with sufficient affinity. As demonstrated in this work, it is relatively easy to obtain substrates that interact in some fashion with modified P RNA-based ribozymes. The selectivity for such substrates with or without S-replacement can be as high as 20 000-fold. A higher selectivity may be accomplished by deleting the L1 module in these modified ribozymes. The L1 module can constitute a part of the binding site in the ribozyme and may be proximal to S-replacement in the three-dimensional structure (3,4). As indicated by its protection upon substrate binding, it is likely that the L1 module is involved in binding of our selected substrates, albeit to different extents.
The isolation of variable RNA substrates containing complementary structure to defined binding sites in the ribozyme has become a routine exercise. The challenge now is to isolate variable ribozymes that cleave defined RNA structures. Two methods may be used to obtain such `custom designed' ribozymes. One method involves replacing the S-domain or the L1 module with the RNA structure of interest to select a complementary binding module in the substrate. These interactive RNA modules will then be switched between the ribozyme and the substrate. Another method involves covalently linking the RNA structure of interest to the ribozyme library to allow cleavage to occur in cis (21,22). Once the new ribozyme with the complementary binding module is selected, the RNA structure may be detached from the ribozyme to allow cleavage in trans. Either way, the substrate binding properties of P RNA should be useful in the design and isolation of ribozyme-substrate pairs without prior knowledge of the sequence/structure of the binding regions in the ribozyme and the substrate.
ACKNOWLEDGEMENTS
We thank V. Shelton for helpful discussions. This work was supported by a grant from the NIH (GM52993). E.M. was the recipient of NIH training grant 5T32GM07183.
REFERENCES
*To whom correspondence should be addressed. Tel: +1 773 702 4179; Fax: +1 773 702 0439; Email: taopan{at}midway.uchicago.edu
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Busch and R. Backofen INFO-RNA--a fast approach to inverse RNA folding Bioinformatics, August 1, 2006; 22(15): 1823 - 1831. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Piccinelli, M. A. Rosenblad, and T. Samuelsson Identification and analysis of ribonuclease P and MRP RNA in a broad range of eukaryotes Nucleic Acids Res., August 8, 2005; 33(14): 4485 - 4495. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J. Ohuchi, Y. Ikawa, H. Shiraishi, and T. Inoue Modular engineering of a Group I intron ribozyme Nucleic Acids Res., August 1, 2002; 30(15): 3473 - 3480. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Loria and T. Pan Modular construction for function of a ribonucleoprotein enzyme: the catalytic domain of Bacillus subtilis RNase P complexed with B.subtilis RNase P protein Nucleic Acids Res., May 1, 2001; 29(9): 1892 - 1897. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||