Importance of residue 248 in Escherichia coli RNase P RNA mediated cleavage

tRNA genes are transcribed as precursors and RNase P generates the matured 5' end of tRNAs. It has been suggested that residue − 1 (the residue immediately 5ʹ of the scissile bond) in the pre-tRNA interacts with the well-conserved bacterial RNase P RNA (RPR) residue A248 (Escherichia coli numbering). The way A248 interacts with residue − 1 is not clear. To gain insight into the role of A248, we analyzed cleavage as a function of A248 substitutions and N−1 nucleobase identity by using pre-tRNA and three model substrates. Our findings are consistent with a model where the structural topology of the active site varies and depends on the identity of the nucleobases at, and in proximity to, the cleavage site and their potential to interact. This leads to positioning of Mg2+ that activates the water that acts as the nucleophile resulting in efficient and correct cleavage. We propose that in addition to be involved in anchoring the substrate the role of A248 is to exclude bulk water from access to the amino acid acceptor stem, thereby preventing non-specific hydrolysis of the pre-tRNA. Finally, base stacking is discussed as a way to protect functionally important base-pairing interactions from non-specific hydrolysis, thereby ensuring high fidelity during RNA processing and the decoding of mRNA.

.The heavy dashed demarcation line separates the S-and C-domains.The large gray box highlights the A 248 -region, and show the substitutions that were introduced at 248 (red arrows).The gray box in L15 marks residues that pair with the substrate 3ʹ end-the RCC A-RNase P RNA interaction (interacting residues underlined) 12 -in the RPR-substrate complex.The blue arrows and Roman numerals mark the Pb 2+induced cleavage sites as shown in Fig. 2 (black circles).The vertical line marked in blue marks the "332-region", which is also cleaved in the in presence of Pb 2+ (see also 85,91 ).Residues highlighted with gray circles correspond to RNase T1 cleavage sites (see also Fig. 2, bands marked with red dots) 92 .The green dashed line and arrows mark the area in P18, which becomes accessible to RNase T1 cleavage upon on substitution of A 248 with U (see Fig. 2, Eco RPR U248 ).(B) Sequence of alignment of the region which includes the conserved E. coli (Ec) A 248 , T. maritima (Tm 93 ) A 213 , M. tuberculosis (Mtb 28 ) A 248 and the Archaea P. furiosus (Pfu 9,12 ) A 218 , and neighboring sequences as indicated.
with residue -1-remains unclear.The crystal and cryo-EM structures of the RNase P-tRNA complexes (bacteria and archaea) do not provide guidance because these structures represent the post cleavage stage of the RNase P catalyzed reaction 18,20 .We therefore decided to revisit and investigate the interrelationship between residue −1 and A 248 .To achieve this, we studied cleavage of all ribo pre-tRNA and model hairpin loop substrates, carrying different nucleobases at position −1 with Eco RPR 248-variants.
Here we provide data that the identities of both residue N −1 in the substrate and residue 248 in the RPR influence cleavage site selection and rate of cleavage.However, our data do not support the model where the wellconserved residue A 248(wt) forms a cis WC/WC pair with N −1 .This was particularly apparent studying different substrates carrying 3-methyl U at the N −1 position.Our combined data support a model where the structural topology of the active site varies and depends on the identity of the nucleobases at, and in proximity to, the cleavage site and their potential to interact.As a consequence, this affects the positioning of Mg 2+ that activates the water that acts as the nucleophile resulting in efficient and correct cleavage.In this scenario we suggest that, besides participating in the anchoring of the substrate, the role of A 248 in wild type bacterial RPR, which stacks on the tRNA G +1 /C +72 base pair, is to exclude bulk water from accessing the amino acid acceptor stem and thereby prevent non-specific hydrolysis/cleavage of the pre-tRNA.

Results
Substituting residue 248 has minor effects on the overall structure of Eco RPR.To investigate the role and contribution of the well-conserved A 248 to Eco RPR mediated cleavage we used wild type Eco RPR A248(wt) and three 248 variants: Eco RPR C248 , Eco RPR G248 and Eco RPR U248 (Fig. 1).The generation and catalytic performance of Eco RPR G248 has been reported elsewhere 30,31 , while the other two RPRs were generated as outlined in "Materials and methods".As predicted on the basis of previous studies, the C 248 and U 248 variants were catalytically active 24 (see below).
First, we inquired whether substitution of A 248(wt) with any of the other nucleobases affected the structure of Eco RPR.On the basis of structural probing with Pb 2+ and RNase T1 (which cleaves 3ʹ of single stranded G residues) we reported that the overall structures of Eco RPR A248(wt) and Eco RPR G248 are very similar 31 .This was also the case for the C 248 variant [Fig.2; cf.lanes 2 and 3 (Eco RPR A248(wt) ) and lanes 11 and 12 (Eco RPR C248 )].By contrast, a U at 248 affected the structure such that G-residues between P15 and the P18-loop became accessible to RNase T1 [Fig.1; cf.lanes 3 (Eco RPR A248(wt) ) and 6 (Eco RPR U248 )].This suggested that a U at 248 influences the structural integrity of P18, which plays a role in connecting the C-and S-domain via the P8/P18-interaction (Fig. 1).Moreover, compared to Eco RPR A248(wt) , exposure to RNase T1 resulted in the appearance of an additional weak cleavage product located between residues 276 and 292, in particular in the case of G 248 (Fig. 2; bands marked with *).This might indicate a change in the structure in this region in response to mutating A 248(wt) , see also Ref. 31 .With respect to the Pb 2+ -induced cleavage patterns, we did not detect any apparent difference comparing the 248 variants [Fig.2; cf.lanes 2 (A 248(wt) ), 5 (U 248 ), 8 (C 248 ) and 11 (G 248 )].We conclude that substitution of A 248 in wild type Eco RPR resulted in a small (if any) overall structural effect with the exception of U 248 where a notable structural change was detected in P18.

Catalytic performance as a function of replacing the well-conserved A 248 in Eco RPR.
Choice of substrates and experimental outline.The role of residues A 248 in Eco RPR A248(wt) and −1 in the substrate has previously been analyzed using variants of a Bacillus subtilis tRNA Asp precursor 24,25 .From these studies, the authors proposed a model where the −1 residue in the substrate forms a cis Watson-Crick (WC) base pair with A 248 .The model predicts that (i) breakage of this interaction shifts cleavage from the correct to an alternative site (see below), and (ii) introduction of a compensatory change that restores the N −1 /N 248 pairing should increase (rescue) cleavage at the correct site.To test this model and to investigate the role of A 248 , we used N −1 derivatives of the E. coli tRNA Ser Su1 precursor, pSu1 13,33 , and two well-characterized model hairpin loop substrates, pATSer and pMini3bp, both derived from pSu1 (Fig. 3) 15,31,[34][35][36] .The pATSer substrates have the amino acceptor-stem and T-stem intact while pMini3bp lacks the T-stem, T-loop and part of the acceptor-stem.Two pATSer variants were used, the first has the original T-loop (e.g.pATSerUG where U and G correspond to the residues at −1 and " +73", respectively; numbering refers to the position in tRNA; Fig. 3).In the other, the T-loop is substituted with a GAAA-tetra loop (e.g.pATSerUG GAAA ).The latter interacts differently with Eco RPR A248(wt) ; it increases cleavage at the alternative site between −2 and −1 (Fig. 3C; see below) 15,30,37 .The short model substrates pMini3bp all have three-base-pair short stems, capped with GAAA-tetra loops (e.g.pMini3bpUG).Importantly, pSu1 and pATSer can interact with the TBS-region (see above) upon Eco RPR substrate complex formation, while pATSer variants with GAAA-tetra loops and pMini3bp cannot (or interact differently) due to their sizes and/ or the presence of the GAAA-tetra loop 15,30,31,38 .We introduced the natural ribonucleobases (A, C, G and U) at position −1 (N −1 ) in all four substrate variants.For pATSer and pMini3bp, we also used variants carrying chemically modified ribonucleobases at −1 and +73.Varying both residue −1 and +73 allowed us to investigate the importance of having nucleobases at −1 that can pair with residue +73 with different numbers of hydrogen bonds.To further investigate whether U −1 in the model substrates pairs with A 248(wt) in Eco RPR we introduced a methyl group (3mU) at − 1 (Fig. 3E), which interferes with cis WC/WC pairing with A 248(wt) .Finally, we replaced the 2ʹOH with 2ʹNH 2 (or 2ʹH) and varied the +1/+72 base pair in pATSerUG to probe the de-protonation of the 2ʹNH 2 (charge distribution; see below) at the canonical cleavage site in the RPR-substrate complex as a function of N 248 identity (Fig. 3) [38][39][40] .
The Mg 2+ concentration for optimal cleavage rates of pMini3bp substrates using Eco RPR A248(wt) and Eco RPR G248 is 800 mM; this is higher than for the other substrates 15,30,31 .Moreover, on the basis of our published data where we studied cleavage of pATSer and pMini3bp variants using Eco RPR A248(wt) and Eco RPR G248 , we assumed that optimal cleavage rates are reached at 800 mM Mg 2+ also for the other 248 variants 15,30,31,37 .To be    The seven-base loop (B, marked in gray) in pATSerNN was replaced with a GAAA-tetra loop (C, marked in gray) to generate pATSerNN GAAA , see 14,15  able to directly compare the cleavage rates, we decided to perform all the experiments discussed below at 800 mM Mg 2+ .Also, at this Mg 2+ concentration the likelihood of detecting cleavage increases, see e.g., 31 .We emphasize that the C5 protein interacts with residues N −4 -N −8 in the 5ʹ leader but not N −1 41,42 and that we were primarily interested in the catalytic performance of the RPR in the absence of C5.Hence, these studies were performed without the C5 protein.
Cleavage of the different substrates was studied with respect to (i) cleavage site recognition and (ii) rate of cleavage (single turnover; see "Materials and methods").The canonical (also referred to as correct cleavage or the +1 position) site corresponds to cleavage between residues −1 and +1 (Fig. 3), while cleavage at other positions are referred to as alternative sites or miscleavage; e.g., cleavage at −1 relates to cleavage between −2 and −1 in the 5ʹ leader.The frequencies of cleavage at +1 are presented in Figs. 4,    To calculate the frequencies of cleavage at +1 we used the 5ʹ cleavage fragments and mean and experimental errors were calculated from at least three independent experiments.and 3.For clarity and guidance, the experiments using different substrate/RPR (N 248 ) combinations are referred to as "Experiment Series (ExpS)" in the figures and tables where 1.1 corresponds to substrate 1, pSu1, having A at −1 while 1.2 has C, i.e. pSu1A −1 and pSu1C −1 , respectively, and substrate 2.1 pATSer having A −1 and G +73 (pATSerAG and A −1 /G +73 substrate/variant) etc.In the first set of experiments we studied cleavage of the full-size pre-tRNA Ser Su1 (pSu1; ExpS 1.1-1.4)and pATSer N −1 /N +73 (ExpS 2.1-2.8;Fig. 4A,B, and Table 1) variants that can interact productively with the TBS in the RPR S-domain (see above).These results are discussed below in "Substrates that can interact productively with the TBS-region-influence of changes of the nucleobase at −1 in substrates and residue 248 in the RPR".Following this we analyzed the impact of N −1 /N +73 variants in substrates that cannot form a productive interaction with the TBS, pATSer GAAA (ExpS 3.1-3.6;Fig. 4C and Table 1) and pMini3bp (ExpS 4.1-4.12;Fig. 4D and Table 2) variants.These data are discussed in "Substrates that cannot interact productively with the TBS-region-influence of changes of the nucleobase at −1 in substrates and residue 248 in the RPR".In "Altering the WC-surface of a U at position −1 and influence of the N248 identity", we discuss substrates carrying CH 3 at position 3 on the nucleobase that alter the WC-surface of U −1 in model substrates www.nature.com/scientificreports/(ExpS 5.1-5.3;Fig. 6 and Table 3).In each sections (A-C), we summarize the data with respect to the possible interaction between residues A 248 and −1.Following this, in "Kinetic constants k obs and k obs /K sto and activation energy as a function of N 248 identity" we present single turnover kinetic data for cleavage of the model substrate pATSerUG with the different 248 variants.These experiments were performed at different temperatures with the objective to determine the activation energy as a function of the N 248 identity.In the final "Differential effects due to replacement of the 2ʹOH at −1 with 2ʹH or 2ʹNH 2 in pATSerUG and influence of the N 248 identity on the charge distribution at the cleavage site", we probe the influence of the N 248 identity on the charge distribution at the cleavage site.
Substrates that can interact productively with the TBS-region-influence of changes of the nucleobase at −1 in substrates and residue 248 in the RPR.pSu1 variants (Fig. 4A and Table 1; ExpS 1.1-1.4):All the −1 variants were cleaved mainly at the +1 site and at the alternative position −1 (Fig. 4A), irrespective of the identity of residue 248.Except for U 248 , the U −1 variant (ExpS 1.4) was cleaved with the highest rate at +1 where k app(+1) was highest for A 248(wt) .With respect to the wild-type substrate, pSu1C −1 (ExpS 1.2),A 248(wt) was the most efficient catalyst; the k app values for the other three variants were lower.
For pSu1A −1 (ExpS 1.1), changing Eco RPR A248(wt) to any of the other nucleobases resulted in decreased cleavage frequency at +1 relative to −1, while comparing k app values for cleavage at +1 and −1 with the different RPRs differed ≤ two-fold.Cleavage of pSu1C −1 (ExpS 1.2) at +1 was reduced and increased at −1 using G 248 , but there  ) while no apparent change in k app(−1) was detected irrespective of RPR.We also noted that the 248 variants cleaved the different pSu1 N −1 substrates with low frequencies at other positions in the 5ʹ leader upstream of site −1 (not shown).Noteworthy, in E. coli the wild-type pSu1 has a C at position −1, which pairs with the discriminator base G +73 (Fig. 3A) and the C −1 /G +73 pairing influences cleavage efficiency and site selection, see e.g. 27.
pATSer variants (Fig. 4B and Table 1; ExpS 2.1-2.8):Cleavage of the N −1 variants carrying A, C, G and U more or less mirrored the results with pSu1 (cf.ExpS 1.1-1.4 vs. 2.1-2.4).Overall G 248 cleaved with the highest rates both at +1 (pATSerAG; ExpS 2.1) and at −1 (pATSerGG; ExpS 2.3).Moreover, having a purine at 248 gives a more efficient catalyst compared to when a pyrimidine is present at this position with the exception of the "pATSerCG/G 248 " combination (ExpS 2.2).
Specifically, first pATSerAG (ExpS 2.1) was cleaved with roughly the same frequencies at +1 by all 248 variants with the possible exception for C 248 , which cleaved this substrate both at +1 and −1 with a lower rate compared to the other RPR variants.Second, compared to the other RPRs G 248 cleaved pATSerCG more frequently at −1.This is also reflected in the k app values for cleavage at +1 and −1, while the G 248 and C 248 RPRs cleaved pATSerCG at +1 with the same rates (ExpS 2.2).These results are contradictory to the formation of cis WC/WC pairing between C −1 and G 248 in the RPR-substrate complex.Third, in contrast there appeared to be suppression/rescue of cleavage of pATSerGG at −1 using C 248 (comparing frequencies of cleavage at −1 and k app(−1) ; Fig. 4B 3F).However, U 248 cleaved pATSerUG with a significantly lower rate at +1 than the other RPR variants (Table 1).
For the pATSer variants with "unnatural" nucleobases at −1 and +73, reducing the number of potential hydrogen bonds between −1 and +73 from three to two restored cleavage at +1 for G 248 to the level observed for the other 248 variants (Fig. 4B; cf.ExpS 2.2 and 2.5, i.e. pATSerCG vs. pATSerCIno).This appeared to be the result of an increase in the rate of cleavage at +1 while hardly any effect on the rate was detected for cleavage at −1 (Table 1; cf.ExpS 2.2 and 2.5).By contrast, the potential formation of three hydrogen bonds between U −1 and DAP +73 resulted in increased miscleavage for all four 248 variants (Fig. 4B; cf.ExpS 2.4 and 2.8, i.e. pATSerUG vs. pATSerUDAP).This was accompanied with noticeable rates of cleavage at −1 (Table 1; cf.ExpS 2.4 and 2.8).In keeping with this, the k app values for cleavage at +1 were lower for pATSerUDAP relative to pATSerUG for all RPR variants.
In summary (see Fig. 5A), (i) when the T-loop can form a productive interaction with TBS in the S-domain we did not detect any conclusive evidence for cis WC/WC pairing between N −1 and N 248 .However, for some of the combinations, cis WC/WC pairing cannot be excluded (see also the "Discussion").(ii) The potential to form three H-bonds between N −1 and N +73 affected both cleavage site selection and dependent on substrate-RPR combination the rate of cleavage.(iii) For Eco RPR A248(wt) , cleavage of substrates with natural nucleobases at N −1 , the U −1 substrates are preferred.
pMini3bp variants (Fig. 4D and Table 2; ExpS 4.1-4.12):For the variants with natural nucleobases at −1 we did observe significant reduction in cleavage of the G −1 variant at +1 using A 248(wt) , G 248 and U 248 while C 248 cleaved the G −1 substrate preferentially at +1 (Fig. 4D; ExpS 4.3).Moreover, irrespective of 248 variants pMini3bpCG and pMini3bpCA were cleaved mainly at +1 with some cleavage at −1 and cleavage of pMini3bpUA was detected only With respect to the rate of cleavage we compared k app(+1) (except for two substrates, see below) because the rates were significantly lower than for the other substrates, in particular at −1 (Table 2).Overall, k app(+1) for A 248(wt) and G 248 were higher compared to C 248 and U 248 ; the highest k app(+1) was for pMini3bpUG (ExpS 4.4) with A 248(wt) .This is in keeping with the trend seen with the other substrate variants, i.e.U −1 variants were in general cleaved with the highest rates at +1 (but cf.e.g., the "pATSerAG/G 248 " combination above).For the "pMini3bpCG/G 248 " combination (ExpS 4.2), k app(+1) and k app(−1) were both higher than when the other 248 variants, including A 248(wt) , were used.Comparing cleavage of pMini3bpCG vs. pMini3bpGG (ExpS 4.2 and 4.3) with G 248 , k app(+1) was ≈1300fold higher for pMini3bpCG, while k app(−1) was ≈40-fold higher.No difference in k app(+1) was detected for C 248 cleaving these two substrates, while k app(−1) was 3000-fold lower in cleaving pMini3bpCG than when G 248 was used (Table 2; ExpS 4.2 and 4.3, cf.0.0001 vs. 0.3).In fact, C 248 was found to be a very poor catalyst with all pMini3bp substrates.Analyzing the (pMini3bpUG and pMini3bpAG)/A 248(wt) and (pMini3bpUG and pMini3bpAG)/U 248 combinations (ExpS 4.1 and 4.4) revealed a significant drop in k app(+1) for A 248(wt) by replacing U −1 with A −1 , but   only a two-fold rescue (cf.pMini3bpUG vs. pMini3bpAG) using U 248 .Interestingly, G 248 cleaved the pMini3bpAG substrate at +1 with a markedly higher rate (k app(+1) ) compared to using the other 248 variants (Table 2; ExpS 4.1).
The pMini3bp variants cannot interact with the TBS-region (see above) and comparison of the pATSer (with T-loop) and the pMini3bp data sets (cf.Fig. 4B,D) revealed that some pATSer variants such as pATSerAG (ExpS 2.1), pATSerInoG (ExpS 2.7) and pATSerUDAP (ExpS 2.8; except U 248 ) were cleaved at −1 with higher frequencies by all four 248 variants.Relative to cleavage of the pATSer derivatives with GAAA-tetra loops, the frequencies of cleavage at +1 were, in general, higher with the pMini3bp variants.
Considering rates of cleavage, introduction of 2NH 2 [Table 2; cf.pMini3bpAG (ExpS 4.1) vs. pMini3bpDAPG (ExpS 4.8)] and removal of the 6NH 2 [Table 2; cf.pMini3bpAG (ExpS 4.1) vs. pMini3bp2APG (ExpS 4.12)] on the −1 nucleobase resulted in a ≈four-and ≈ 160-fold decrease in k app(+1) for G 248 , while for A 248(wt) corresponding values were ≈ three-and ≈ ten-fold lower.These data suggested that in particular the exocyclic amine at position 6 on A −1 plays a more important role for cleavage with G 248 than for A 248(wt) .Cleavage of pMini3bpUG and pMini3bpUA with G 248 resulted in a 20-and ten-fold lower k app(+1) , respectively, compared to A 248(wt) (Table 2; ExpS 4.4 and 4.6) while only a small difference in k app(+1) was detected for cleavage of pMini3bpUDAP using these two RPRs (Table 2; ExpS 4.11).Moreover, the k app(+1) values for these three pMini3bp U −1 substrates using G 248 were similar, within a factor of two.This might indicate that the catalytic performance of A 248(wt) is influenced by pairing between N −1 and N +73 and/or the pairing between N +73 and U 294 in the RPR-substrate complex.
In summary (see Fig. 5B), (i) the cleavage site distribution data did not provide any conclusive evidence for cis WC/WC pairing between residues N −1 and 248 in the Eco RPR substrate complex when we interfered with/ or removed the interaction between the T-loop and TBS.However, there were a few possible exceptions, e.g., the combinations "pATSerCG GAAA /G 248 ", "pATSerGG GAAA /C 248 ", and "pMini3bpGG/C 248 ". (ii) As in cleavage of pSu1 and pATSer variants, the potential pairing between N −1 and N +73 influence the efficiency of cleavage and site selection also in the absence of a productive interaction between the T-loop and TBS in the RPR.(iii) Interfering with the interaction between TSL and TBS affect choice of cleavage site and rate of cleavage, see also 15,37 .
Together the combined data with the four different substrate series suggested that the influence of N −1 and N +73 on cleavage site recognition and rate of cleavage at +1 and −1 depend on substrate and/or "N −1 /N +73 -N 248 " combination.Moreover, in general we do detect larger variations in k app for cleavage at +1 than at −1.It therefore appears that the impact of the various changes either in the substrate or in the RPR is larger for cleavage at the correct position +1 than at −1.We also emphasize that the choice of cleavage site did not change during the course of the reactions as revealed from the time course experiments used to determine k app values.
Altering the WC-surface of a U at position −1 and influence of the N 248 identity.To further understand the importance of the Watson-Crick surface of the N −1 residue in the substrate we used substrates carrying substitutions of U with 3-methyl U (3mU) at −1.This modification would be expected to disturb the interaction with the Watson-Crick surface of U −1 (Fig. 3B-E).The data are shown in Fig. 6 and Table 3 (ExpS 5.1-5.3 vs. 2.4, 3.4 and 4.4).
A comparison of cleavage of pATSerUG vs. pATSer3mUG revealed no (or very minor) change in choice of cleavage site (Figs.4B and 6; cf.ExpS 2.4 vs. ExpS 5.1) for any of the four 248 variants.However, k app(+1) dropped three-to four-fold for all four RPRs, with U 248 being the least efficient catalyst (Table 3).www.nature.com/scientificreports/With pATSerUG GAAA , introduction of 3mU at −1 did not result in any apparent change in cleavage site preference with the notable exception for U 248 .Here we did detect an increase of cleavage at +1 compared to cleavage of pATSerUG GAAA (Figs. 4C and 6; cf.ExpS 3.6 vs. ExpS 5.2).The other three 248 variants cleaved both pATSerUG GAAA and pATSer3mUG GAAA at +1.As for pATSerUG, the presence of 3mU −1 influenced cleavage rates; k app(+1) values were down four-to six-fold using A 248(wt) and G 248 , respectively, while for C 248 the decrease was very modest, ≈1.5-fold.No apparent change was detected for U 248 .Interestingly, 3mU −1 influenced the rate of cleavage at +1 for A 248(wt) and C 248 while cleavage by G 248 resulted in a four-fold decrease (Table 3; cf.ExpS 3.4 vs. ExpS 5.2).
Comparing cleavage of pMini3bpUG vs. pMini3bp3mUG, we detected just a small increase in cleavage at −1 for all 248 variants (Figs.4D and 6; cf.ExpS 4.4 vs. ExpS 5.3).Moreover, k app(+1) for A 248(wt) was down 16-fold in response to the introduction of 3mU −1 .For G 248 and C 248 , the change was more modest, 2.7-fold lower for G 248 while C 248 cleaved 3mU −1 with a 2.5-fold higher rate than it cleaved the corresponding substrate lacking the methyl modification.No change was detected for U 248 .
In summary (see Fig. 5C), the presence of 3mU −1 that blocks the Watson-Crick surface has an impact on the rate of cleavage.The impact on the rate at +1 (k app(+1) ) appears to be dependent on RPR-substrate combination, as exemplified by cleavage of pMini3bpUG and pMini3bp3mUG with A 248(wt) vs. C 248 .Remarkably, introduction of 3mU −1 in the "pATSer-GAAA-tetra-loop" substrate rescued cleavage at +1 using the U 248 RPR variant.Hence, these findings do not support cis WC/WC pairing between N −1 and 248 for these substrates, see also 29 .Kinetic constants k obs and k obs /K sto and activation energy as a function of N 248 identity.The data presented above clearly suggested that the identity of residue 248 affect both cleavage site recognition and rate of cleavage.We therefore decided to determine the kinetic constants, k obs and k obs /K sto (for cleavage at +1), for the different 248 variants using pATSerUG.To gain insight into why a purine at 248 (in particular A at 248) is preferable over a pyrimidine, we also determined k obs and k obs /K sto at different temperatures.This would allow us to estimate the activation energy for the reaction catalyzed by the various 248 RPRs.These series of experiments were done under single turnover conditions at 800 mM Mg 2+ (see above) and the results are shown in Fig. 7 and Table 4. www.nature.com/scientificreports/ The k obs and k obs /K sto values for A 248(wt) at 37 °C agreed with our previous data (Table 4) 37 .A comparison of k obs and k obs /K sto for the four 248 variants revealed that having A or G at 248 resulted in the most efficient catalysts, in agreement with data discussed above.For A 248(wt) , lower temperature resulted in a modest but reproducible decrease in k obs .This trend was also detected for the other 248 variants.The k obs at different temperatures were highest for A 248(wt) and G 248 , and lowest for C 248 and U 248 .Irrespective of 248 variant and temperature, the K sto values were similar within a ≈ two-to three-fold range.We have argued that under these reaction conditions K sto ≈ K d (see "Materials and methods") 31 and references therein.On the basis of this, our data suggested that substituting A 248(wt) resulted in a modest change in binding affinity for pATSerUG.
Notwithstanding that the variation in k obs in response to temperature was modest (but reproducible) we plotted k obs as a function of temperature (Arrhenius plot).This would give an indication about the activation energy (E a ) for cleavage of pATSerUG by the different 248 variants.The E a values varied from 12 to 57 kJ/mole, with A 248(wt) having the lowest value followed by G 248 < C 248 and < U 248 (Fig. 7; Table 4).
Taken together, in keeping with the data discussed above, a purine at 248 is preferred over a pyrimidine, with U 248 being the weakest catalyst.From these data it also appears that this is, at least in part, due to the activation energy barrier being lower with a purine at 248, in particular with an adenosine as in Eco RPR A248(wt) .This provides one rational why A at position 248 in bacterial RPR (Eco numbering) is conserved (see also the "Discussion").
Differential effects due to replacement of the 2ʹOH at −1 with 2ʹH or 2ʹNH 2 in pATSerUG and influence of the N 248 identity on the charge distribution at the cleavage site.The 2ʹOH of residue −1 is important for both cleavage rates and site selection in bacterial RPR-mediated catalysis 43 .Hence, we decided to investigate whether replacement of the U −1 2ʹOH with 2ʹH or 2ʹNH 2 in pATSerUG (pATSerdUG and pATSeramUG, respectively; Fig. 3B,E) influenced the choice of cleavage site.
Introduction of a 2ʹH (pATSerdUG) resulted in reduced cleavage at +1 for all 248 variants irrespective of pH (5.2, 6.1 and 7.2) consistent with previous data using pre-tRNA 24,25 .Importantly, cleavage at −1 did not increase with pH (Fig. 8A).Cleavage of the 2ʹNH 2 substituted substrate (pATSeramUG) on the other hand resulted in increased cleavage at +1 at higher pH.In contrast to cleavage with A 248(wt) and G 248 higher pH was required to reach 50% cleavage at +1 using C 248 (Fig. 8B).The most dramatic effect however, was observed using the U 248 variant.Here we did not detect any significant change in the frequency of cleavage at +1 with increasing pH.
The pH dependent cleavage of pATSeramUG at +1 by Eco RPR A248(wt) is also influenced by the identity of N +1 / N +72 (cf.Fig. 5 in 40 ; see also 44 ; Fig. 8B,C; cf.G +1 /C +72 , A +1 /U +72 , 2AP +1 /U +72 , DAP +1 /U +72 and Ino +1 /C +72 substrate variants).This was also the case for the C 248 and G 248 variants.Of those substrate variants having an exocyclic amine at position 2 on the nucleobases (2NH 2 ) at +1 (Fig. 3B; cf.substrates with G +1 , 2AP +1 and DAP +1 ) C 248 showed a similar response to pH as A 248(wt) , while higher pH was needed to reach 50% cleavage at +1 for G 248 except using pATSeramUG(G +1 /C +72 ) (cf.Fig. 8A-C).For the substrates lacking a 2NH 2 on the nucleobase at +1 [pATSeramUG(A +1 /U +72 ) and pATSeramUG(Ino +1 /C +72 )], we detected only a small increase in cleavage at +1 with increasing pH for A 248(wt) , C 248 and G 248 while for U 248 no cleavage at +1 was observed.In fact, for U 248 we observed no or only a small increase in cleavage at +1 using all pATSeramUG(N +1 /N +72 ) variants with increasing Table 4.The kinetic constants for cleavage of pATSerUG at as a function of temperature and 248 variant.The experiments were performed under single-turnover conditions at 800 mM Mg 2+ concentrations at pH 6.1 as outlined in "Materials and methods".For details regarding the calculation of K d , see the main text, Wu et al. 31 and references therein.The activation energies were calculated using the k obs values as described in Tallsjö and Kirsebom 68 (see also Fig. 7).The data represent mean ± experimental errors calculated from at least three independent experiments.

variant
Temp (°C) k obs (min -1 ) k obs /K sto (min pH.For all the RPR substrate combinations we also detected cleavage at other positions both downstream of the +1 site and in the 5ʹ leader with increasing pH (not shown).Also, irrespective of residue at 248 no significant change in the frequencies of cleavage at +1 with changing pH using the all ribo substrate variants was detected (not shown).Taken together, these data suggest that the protonation (the pKa value) of the 2ʹNH 2 at −1 is affected by the nucleobase identity at position 248 in Eco RPR and at +1 (and +72) in pATSerUG (see "Discussion").

Discussion
Residues in the RNase P substrate interact with several regions of the RNA subunit (RPR) of bacterial RNase P (see introduction).Among these the N −1 residue in the substrate 5ʹ leader is close to the active center where cleavage occurs, and it has been proposed that the well conserved A 248(wt) forms a cis WC/WC base pair when U is present at −1 24,25 .These studies were primarily based on using pre-tRNAs carrying different deoxyribonucleobases at position N −1 .In E. coli ≈40% of the pre-tRNAs do not carry a U at −1 24,27,28 .Also, cross-linking studies suggest that N −1 and N +1 in the substrate are positioned close to A 248 -C 253 and G 332 -A 333 (E. coli numbering, see Fig. 1A) 26,45,46 .Hence, we have argued that A 248(wt) is a key nucleobase of a N −1 binding surface/pocket 16,27,29 .Here we provide data where we analyzed cleavage as a function of A 248(wt) substitutions and N −1 nucleobase identity using all ribo pre-tRNA and three all ribo model substrates to investigate whether N −1 and N 248 forms a cis WC/ WC base pair.If cis WC/WC base pair forms between N −1 and N 248 this means that the phenotypic change due to disruption of the N −1 /N 248 pairing can be rescued by a compensatory change that restores pairing between N −1 /N 248 .For the pre-tRNA substrate pSu1 and the model substrate pATSer, which both can form a productive TSL/TBS-interaction (see "Introduction", induced fit mechanism) 15,30,37,46 , the data supported cis WC/WC pairing for substrates carrying G at −1, while we did not find any conclusive evidence for cis WC/WC pairing using the other combinations (except the U −1 /A 248 vs.A −1 /U 248 combinations in the pATSer context; see summary, Fig. 5A).When we interfered with the TSL/TBS-interaction by using "pATSer-GAAA-tetra-loop" substrates our findings are consistent with cis WC/WC pairing using the C −1 , G −1 and U −1 substrate variants but not for A −1 (see summary, Fig. 5B).The impact of the N −1 /N 248 interaction was also detected using pre-tRNA substrates carrying a 2ʹH at −1 or substrates that could not form the RCC A-RNase P RNA interaction 24,25 , i.e. when additional RPR substrate interactions were disrupted.Moreover, our findings with the pMini3bp variants, which cannot interact with TBS in the S domain, lend less support for cis WC/WC pairing than when the "pATSer-GAAA-tetra-loop" series was used.But, support comes from using pMini3bpGG, pMini3bpDAPG and pMini3bpUDAP, where the latter can form three hydrogen bonds between N −1 and N +73 in the substrate (see summary, Fig. 5B).In summary, detection of possible cis WC/WC pairing between N −1 and N 248 depends on substrate and disruption of more than one RPR-substrate contact such as the TSL/TBS-interaction.
Residue A 248 is well conserved among bacterial RPRs and if the U −1 WC surface are involved in pairing with residue A 248(wt) blocking the N3 position on the nucleobase-by adding a methyl group (3mU)-would interfere with choice of cleavage site and rate of cleavage.As in pSu1, the model substrates carry an A at −2.Hence, following Zahler et al. 24,25 , who used pre-tRNA Asp that also carries A −2 , we argued that interfering with the formation of the "U −1 /A 248(wt) " potential pairing would result in a shift of cleavage from the correct site to the alternative site −1 due to the presence of the 3-methyl group at the N3 position of U −1 in the substrate.All three 3mU −1 model substrate variants were, however, preferentially cleaved at +1 irrespective of 248-variant.This is inconsistent with cis WC/WC pairing (see summary, Fig. 5C).Importantly, the introduction of 3mU −1 in the three all ribo model hairpin loop substrates did not shift choice of cleavage site for wild type Eco RPR A248(wt) , which would be expected if there was cis WC/WC pairing between U −1 and A 248(wt) , see also 29 .It is also noteworthy that the presence of 3mU −1 in the "pATSer-GAAA-tetra-loop" substrate rescued cleavage at +1 using the U 248 RPR variant.Together these data do not support cis WC/WC pairing between U −1 and A 248(wt) in wild type Eco RPR.In this context we emphasize that substituting A 248(wt) with U influenced the structure of the RPR, in particular in the P18 region, which has a role in connecting the S-and the C-domains.The P18 loop interacts with P8 and disruption of this interaction affects cleavage efficiency of both pre-tRNAs and model hairpin loop substrates [47][48][49][50][51] .Hence, this structural change in the RPR might therefore have an impact on the catalytic performance of the U 248 variant, both with respect to site selection and rate of cleavage; however, again this would be substrate dependent.This would be in keeping with a perturbed coupling (i.e.induced fit, see e.g.Ref. 15 ) between a productive TSL-TBS interaction and events at the cleavage.
Furthermore, in E. coli as well as in other bacteria a U is the most frequently (≈60%) occurring nucleobase at −1 in pre-tRNA 5ʹ leaders 24,25,27,28 .This also applies to the archaea Pyrococcus furiosus (65% U −1 ), which as E. coli possess a type A RPR and an A at the corresponding position to A 248 9,12,22 (Fig. 1B).As discussed above, there is limited support for cis WC/WC pairing between U −1 and A 248(wt) in wild type Eco RPR.High GC-content bacteria such as Mycobacterium tuberculosis (and other mycobacteria; see Fig. 1B) and Neisseria meningitides carry type A RPRs with A 248(wt) (E. coli numbering).In these bacteria, C at −1 is the most frequently occurring nucleobase, while U −1 is present in ≈13% and ≈32% of the pre-tRNAs, respectively 27,28 .This argues against formation of cis WC/WC pairing between N −1 and A 248(wt) for the majority of pre-tRNAs in these bacteria.
In conclusion for the majority of pre-tRNAs (and model substrates), A 248 does not interact with N −1 via cis WC/WC pairing.However, given that RNase P processes other RNA transcripts, including mRNAs 2 , we cannot completely exclude the possibility that A 248(wt) is engaged in cis WC/WC pairing with these substrates.In this context we also have to consider that our experiments were performed without the C5 protein and hence the presence of C5 might have an impact given that C5 interact with residues upstream of N −1 (see above 41,42 ).We propose that the structural architecture of the "active site" is flexible and varies dependent on the identity of the nucleobases at and near the cleavage site and their potential to interact with chemical groups in the RPR.This flexibility is also predicted to depend on the interaction between the pre-tRNA TSL-region and its binding site (TBS) in the RPR S-domain (see above) as well as the RCC A-RPR interaction 15,24,25,30,37,44,46 .
Structural architecture and Me(II)-binding near the cleavage site.RNase P mediated cleavage depends on Me(II)-ions, which are involved in activating the water molecule that acts as the nucleophile, substrate interaction and folding of the RPR 43,52 .On the basis of correctness and rate of cleavage available data suggest that Mg 2+ is the preferred ion.Perreault and Altman 53,54 suggested that binding of Mg 2+ at the junction between the single stranded 5ʹ leader and the amino acid acceptor stem involves the two 2ʹ hydroxyls at positions −1 and −2 forming a productive complex that acts as the true RNase P substrate, see also 25,38,39,46,55,56 .In RNA the structural topology of Me(II)-binding sites affects both binding affinity and positioning of the Me(II)-ion.This is evident from lead(II)-induced cleavage studies of yeast tRNA Phe and Eco RPR 15,[57][58][59] .For model substrates, introduction of U +1 (or C +1 ) in pATSerUG (or pATSerCG) affects lead(II)-induced cleavage at the cleavage site such that the frequency of cleavage 5ʹ of N +1 increases more than when a purine is present at +1 60 .Similarly, substituting the 2ʹOH at -2, −1 and "C +74 " in a model hairpin loop model substrate influences Mg 2+ -induced cleavage between −3 and −2 53 .In keeping with this, substituting the N −1 2ʹOH with 2ʹNH 2 in pATSerUG prevent Pb 2+ -induced cleavage between residue −1 and +1 (not shown).Also, the presence of a 2ʹNH 2 at N −1 in pATSe-rUG (and pATSerCG) result in a shift of cleavage from −1 to +1 with increasing pH [38][39][40]44 ; this report. ThepKa for 2ʹNH 2 is 6.0-6.2 (determined by NMR-spectroscopy using a dinucleotide) 61,62 .Therefore the 2ʹNH 2 at −1 in pATSeramUG is most likely protonated at lower pH.As a consequence, this results in a positive charge at the +1 cleavage site, which interferes with cleavage at +1, causing the cleavage to shift to −1 38,39 .With increasing pH, the 2ʹNH 3 + becomes deprotonated, resulting in cleavage at +1.The pH dependent shift of cleavage from −1 to +1 (i.e., de-protonation of the 2ʹNH 3 + at −1) is also dependent on the structure of the N +1 /N +72 base pair 40 ; this report.The data presented here using the 2ʹNH 2 substituted substrates suggest that the identity of residue 248 in the RPR also influences the pH dependent shift from −1 to +1, in particular with respect to U 248 .However, we also observed a shift in the pH dependence for G 248 when the structure of the N +1 /N +72 base pair was altered. Giventhat A 248(wt) is in close proximity to the cleavage site 18 these data are consistent with a model where changes of the structural architecture at and near the cleavage site in the RPR-substrate complex (see above) affect the charge distribution.As a consequence, this influences the positioning of the Mg 2+ that activates the water that acts as the nucleophile resulting in a shift of the phosphorus to be attacked 31,43 ; for an alternative rational see 25 .
Proposed function of the well-conserved residue A 248(wt) in wild type RPR and base stacking to prevent unspecific hydrolysis.In the RNase P tRNA crystal structure, which represents the post-cleavage stage, A 248(wt) stacks on top of the tRNA G +1 /C +72 base pair and presents the Hoogsteen surface facing the G +1 and the tRNA 5ʹ end (Fig. 9A) 17 ; see also Refs 20,63 .The importance of the A 248(wt) Hoogsteen surface for substrate interaction has been implicated on the basis of nucleotide analogue-modification interferences studies 32 .However, we provided data suggesting that the Hoogsteen surface of A 248(wt) is not engaged in pairing with N −1 , at least not in the case of pMini3bp substrates 31 .This raises the question about the role and function of A 248(wt) .The structure of yeast tRNA Phe reveals that the discriminator base at position +73 stacks on top of the G +1 /C +72 pair (Fig. 9B) 64 .As such, the discriminator base acts as a hydrophobic cap that restricts access of bulk H 2 O to the terminal base pair 65,66 .Binding of pre-tRNA to the RPR results in formation of the RCC A-RNase P RNA interaction where the discriminator base pairs with residue U 294 18,27,34   .In the RNase P-tRNA complex A 248(wt) stacks on the G +1 /C +72 base pair by occupying the position that the discriminator base has in free tRNA (Fig. 9A,B).This contributes to anchor the substrate to the RPR 18,20,63 .In addition, we propose that the A 248(wt) stacking on G +1 /C +72 prevents water from accessing the hydrophobic amino acid acceptor stem and potential unspecific hydrolysis of the tRNA after cleavage.We foresee that this also occurs prior to cleavage of the pre-tRNA and the recent cryoEM structures of Eco RNase P in complex with pre-tRNA support that this is indeed the case 63 .In this context the stacking free energy for A would be more favorable, followed by G, C and U 67 .Moreover, considering the activation energy (E a ), our findings indicated that the trend is A 248(wt) < G 248 < C 248 < U 248 with A 248(wt) having the lowest activation energy barrier (Table 4).These data provide reasons to why A 248(wt) in bacterial RPR is well conserved.
We also note that the E a value for cleavage of pATSerUG with A 248(wt) was determined to be 12 kJ/mole (Table 4), which is two-to three-fold lower than for cleavage of pre-tRNA Tyr Su3, both with and without the RNase P protein C5 68 .This difference could depend on substrate and/or reaction conditions.In pre-tRNA Tyr Su3 both the discriminator base (A +73 ) and the first 3ʹ C (C +74 ) pair with U −1 and G −2 in the 5ʹ leader, respectively, rendering A +73 and C +74 less accessible for interacting with RPR, i.e. formation of the "RCC A-RPR interaction" (see above 13 ), compared to pATSerUG (Fig. 3).Also, here the experiments were performed at high Mg 2+ and at a lower pH than in our previous study 68 , which are also factors to consider.
To conclude, in addition to its contribution to anchor the substrate 18,20,63 we suggest that the function of A 248(wt) is to replace the tRNA discriminator base and prevent access of water that would lead to unspecific hydrolysis/cleavage of the pre-tRNA in the RNase P-substrate complex.Saccharomyces cerevisiae RPR lacks an A at the position corresponding to Eco RPR A 248(wt) .Interestingly, in the cryo-EM structure of S. cerevisiae RNase P in complex with pre-tRNA the 5ʹ leader residues A −1 and A −2 stack on top of the tRNA G +1 /C +72 pair forming a hydrophobic cap 19 .According to our proposal this would also prevent unspecific hydrolysis/cleavage of the pre-tRNA.Given that POP5 amino acid residues are also positioned close to the G +1 /C +72 pair these might also contribute to prevent access of H 2 O and unspecific hydrolysis/cleavage (see also below).

Prevention of unspecific hydrolysis in PRORP.
Like RNase P, proteinaceous PRORPs cleave the 5ʹ leader of pre-tRNAs and recent data show that the N −1 identity also influences cleavage by PRORPs both with respect to cleavage site recognition and rate of cleavage [69][70][71] .The crystal structures of PRORP1 and PRORP2 are available 72,73 ; for a cryo structure see 74 , whereas the structure of PRORP in complex with its pre-tRNA substrate is not.Structural and mechanistic studies suggest that D474 and D475 coordinate Me(II) in the PRORP1 active site.Given the similarities between RNA and protein-based RNase P activities, i.e., the need to cleave pre-tRNAs correctly and prevent unspecific hydrolysis, it is likely that stacking on top of the N +1 /N +72 base pair is also present in the PRORP-pre-tRNA complex.Candidates to act as a hydrophobic cap during the PRORP catalyzed reaction might be aromatic amino acids such as W478 and F500, which both are positioned close to the Me(II)ion in the active site.Another possibility is that the pre-tRNA discriminator base keeps its position and stacks on  1) on the tRNA Phe G +1 /C +72 base pair (in green) in the crystal structure of the RNase P-tRNA Phe complex (PDB code 3Q1R) 18 .Grey spheres represent Me(II)-ions.(C) Stacking and the RCC A-RPR interaction (in green) in the crystal structure of the RNase P-tRNA Phe complex (PDB code 3Q1R) 18 .Stacking residues in magenta.D 73 corresponds to the discriminator base at position +73 in tRNA 94 while the RPR numbering refers to E. coli numbering (Fig. 1).Note that A 295 in E. coli corresponds to U 266 in T. maritima RPR 18 .Stacking residues, the tRNA 3ʹ terminal A 76 and the RPR residue, are marked in magenta.(D) Codon-anticodon interaction in the ribosomal A-site where residues in magenta stack as shown in the figure.p34-p37 correspond to positions in the tRNA anticodon loop.Gray residues represent the codon and residues marked in orange residues correspond to A1492 and A1493 in 16S rRNA (PDB code 2J02) 80 .(E,F) Stacking interactions in the ribosomal peptidyl transfer center, panel E (A-site) and panel F (P-site) as indicated.Orange residues correspond to rRNA residues interacting with the tRNA, green residues refer to tRNA and the tRNA discriminator base is highlighted in magenta (PDB code 5IBB) 79 .The images were created using PyMOL (Schrödinger, LLC).www.nature.com/scientificreports/top of the N +1 /N +72 pair (and/or residues in the pre-tRNA 5ʹ leader, see above) in the PRORP-substrate complex as observed in other protein-tRNA complexes (see below).It will be interesting to determine whether this is the case and, if so, how access of water to the "inside" of the hydrophobic amino acid acceptor stem is prevented in the PRORP-substrate complex.
Base stacking and prevention of unspecific hydrolysis of RNA.Crystal structures of amino-acyl-tRNA synthetase-tRNA complexes (such as ArgRS-tRNA Arg and MetRS-tRNA Met ), EF-Tu-tRNA Phe , the CCA adding enzyme in complex with a tRNA mimic and tRNA bound to the ribosome show that the discriminator base at +73 stacks on the G +1 /C +72 pair in a similar way as shown in Fig. 9A (see also E,F) [75][76][77][78][79] .In all these examples the discriminator is a purine.Moreover, inspection of the RCC A-RNase P RNA interaction in the RNase P-tRNA crystal structure reveals that U 266 stacks on the A +73 /U 265 base pair, while the 3ʹ terminal A +76 stacks on the C +75 /G 263 base pair (Fig. 9C; note that the T. maritima residues G 264 , U 265 and U 266 correspond to G 292 , U 294 and A 295 in wild type Eco RPR, see Fig. 1) 18 .A similar type of stacking can also be observed in the ribosomal A-and P-sites both in the case of tRNA and mRNA interaction as well as with respect to the pairing between C 74 and C 75 and rRNA (Fig. 9D-F) [79][80][81] .Together this further emphasizes the importance of stacking.It is conceivable that a function of this "type" of base stacking is to prevent the access of water to functionally important base pairing interactions, and thereby ensuring high fidelity during RNA processing and decoding of mRNA.

Materials and methods
Preparation of substrates and RPR.The tRNA Ser Su1 precursor (pSu1) N −1 variants were generated as run-off transcripts using T7 DNA-dependent RNA polymerase and PCR-amplified templates as described elsewhere 33,82 .The model hairpin loop substrate N −1 series (pATSer, pATSer-GAAA-tetra loop and pMini3bp) were purchased from Thermo Scientific Dharmacon, USA.The substrates were [γ-32 P]-ATP 5ʹ end-labeled and gel-purified followed by overnight Bio-Trap extraction (Schleicher and Schuell, GmbH, Germany; Elutrap in USA and Canada) and phenol-chloroform extraction as described elsewhere 15,31 .
The construction of the gene encoding Eco RPR G248 was recently reported 31 , while the C 248 and U 248 variants behind the T7 promoter were generated following the same procedure as outlined elsewhere using the wild type Eco RPR A248(wt) gene as template and appropriate oligonucleotides 12,31,83,84 .The RPRs were generated as run-off transcripts using T7 DNA-dependent RNA polymerase and PCR-amplified templates 31,82 .
Structural probing of the Eco RPR variants.The Eco RPR variants were 3ʹ-end labeled with [ 32 P]pCp and structurally probed using Pb 2+ and RNase T1 under native conditions as described elsewhere 31,34,35,45,85 .Briefly, approximately 2 pmols of labeled RPR in 10 µl was pre-incubated for 10 min at 37 °C in 50 mM Tris-HCl (pH 7.5), 100 mM NH 4 Cl and 10 mM MgCl 2 together with 4 µM of the unlabeled corresponding RPR.Cleavage was initiated by adding freshly prepared Pb(OAc) 2 to a final concentration of 0.5 mM and the reaction was stopped after 10 min.In the digestion with RNase T1, the RPR was pre-incubated as described above.One unit of RNase T1 was added followed by incubation on ice for 10 min.The reactions were stopped by adding two volumes of stop solution (10 M urea, 100 mM EDTA).The products were analyzed on 8% (w/v) denaturing polyacrylamide/7 M urea gels.
Cleavage assays and determination of k app .The cleavage reactions were conducted in buffer C [50 mM 4-morpholineethanesulfonic acid (MES) and 0.8 M NH 4 Cl (pH 6.1)] at 37 °C and 800 mM Mg(OAc) 2 .The RPRs were pre-incubated at 37 °C in buffer C and 800 mM Mg(OAc) 2 for at least 10 min to allow proper folding before mixing with pre-heated (37 °C) substrate.In all the experiments the concentrations of substrates were ≤ 0.02 µM, while the concentrations of the RPR variants were as indicated in Table and Figure legends.The reactions were terminated by adding two volumes of stop solution (see above).The products were separated on 25% (w/v) polyacrylamide/7 M urea gels.
Cleavage of pATSerU am G derivatives at 37 °C was performed in buffer C and 800 mM Mg(OAc) 2 at pH 5.2, pH 6.1 and pH 7.2 39,40 .
The rate constant k app was determined under single-turnover condition at 800 mM Mg 2+ in buffer C. The concentrations of Eco RPR variants used to generate the data are specified in the respective Table legends.The concentrations of pSu1 (precursor-tRNA Ser Su1 34 ) and model substrates 15,31,34 were ≤ 0.02 μM (see also the main text).For rate calculations, we used the 5ʹ cleavage fragment as a measure of product formed.In each assay, the time of incubation was adjusted to ensure that the velocity measurements were in the linear range (typically ≤ 10%, but never exceeding that 40% of the substrate had been consumed).Each k app value is reported as a mean ± deviation of this value, which was calculated using data (six time points) from at least three independent experiments.Determination of the kinetic constants k obs , k obs /K sto and K sto .The rate constants k obs and k obs /K sto were determined under saturating single-turnover conditions at pH 6.1 (where cleavage is suggested to be rate limiting) and 800 mM Mg 2+ using pATSerUG, as described elsewhere, e.g. 37.Under these conditions we have argued elsewhere that K sto ≈ K d in the Eco RPR-alone reaction 12,30,31,86,87 .The final concentrations of the different RPR variants were between 0.8 and 6.4 µM; the concentration of the pATSerUG substrate was ≤ 0.02 μM.To ensure that the experiments were done under single-turnover conditions, the lowest concentration of RPR was > 10 times higher than the concentration of the substrate.For the calculations we used the 5ʹ cleavage fragment, and the time of cleavage was adjusted to ensure that the velocity measurements were in the linear range (see above).To be able to compare with our previously published data, k obs and k obs /K sto were obtained by linear regression from Eadie-Hofstee plots as described elsewhere 12,30,31,37,88,89 .Each value is an average of at least three independent experiments and is given as a mean ± the deviation of this value.

Figure 1 .
Figure 1.Illustration of the Eco RPR secondary structures.(A) Eco RPR secondary structure according to Massire et al.90 .The heavy dashed demarcation line separates the S-and C-domains.The large gray box highlights the A 248 -region, and show the substitutions that were introduced at 248 (red arrows).The gray box in L15 marks residues that pair with the substrate 3ʹ end-the RCC A-RNase P RNA interaction (interacting residues underlined)12 -in the RPR-substrate complex.The blue arrows and Roman numerals mark the Pb 2+induced cleavage sites as shown in Fig.2(black circles).The vertical line marked in blue marks the "332-region", which is also cleaved in the in presence of Pb 2+ (see also85,91 ).Residues highlighted with gray circles correspond to RNase T1 cleavage sites (see also Fig.2, bands marked with red dots)92 .The green dashed line and arrows mark the area in P18, which becomes accessible to RNase T1 cleavage upon on substitution of A 248 with U (see Fig.2, Eco RPR U248 ).(B) Sequence of alignment of the region which includes the conserved E. coli (Ec) A 248 , T. maritima (Tm 93 ) A 213 , M. tuberculosis (Mtb 28 ) A 248 and the Archaea P. furiosus (Pfu 9,12 ) A 218 , and neighboring sequences as indicated.

Figure 2 .
Figure 2. Structural probing of Eco RPR.Probing the structures of the Eco RPR 248-variants with Pb 2+ and RNase T1.Roman numerals and black circles refer to Pb 2+ -induced cleavage sites in Eco RPR (Fig. 1) 31,85,91 .Numbers and red circles correspond to the RNase T1 cleavage sites according to Guerrier-Takada and Altman 92 , see Fig. 1A.The vertical black lines mark the P18-and 332-region.The vertical black line "P18" marks the extra RNase T1 cleavage sites between 292 and 314 in the U 248 variant.The reactions were conducted using 0.5 mM Pb(OAc) 2 and RNase T1 as described in "Materials and methods".

2 HFigure 3 .
Figure 3. Secondary structure of substrates used in the present study.(A) pSu1, (B) pATSerNN, (C) pATSerNN GAAA , (D) pMini3bp, (E) structures of nucleobases, and (F) cleavage of pATSerUG by the different Eco RPR 248 variants.Residues highlighted in gray were introduced to generate the different variants carrying alternative nucleobases at positions −1 and +73.The black boxes illustrate the changes that generated substrates carrying 2ʹNH 2 and 2ʹH as well as substitutions of residues at positions +1 and +72.The canonical (correct) cleavage sites between residues N −1 and N +1 in the different substrates are marked with black arrows.The gray arrows mark the alternative cleavage sites between N −2 and N −1 (referred to as position −1, see text).The seven-base loop (B, marked in gray) in pATSerNN was replaced with a GAAA-tetra loop (C, marked in gray) to generate pATSerNN GAAA , see14,15 .Panel (F): lane (L) 1, pATSerUG no RPR added; lane 2, cleavage of pATSerCG GAAA with Eco RPR A248(wt) ; lane 3, cleavage of pATSerUG with Eco RPR A248(wt) ; lane 4, cleavage of pATSerUG with Eco RPR C248 ; lane 5, cleavage of pATSerUG with Eco RPR G248 ; lane 6, cleavage of pATSerUG with Eco RPR U248 .Sub, substrate and 5ʹCL Frags marks the migration of the 5ʹ cleavage products as a result of cleavage at +1 and −1.The reaction was performed in buffer C at 800 mM Mg 2+ with 0.8 μM Eco RPR (irrespective of variant) and ≤ 0.02 μM substrate for 10 s as described in "Materials and methods".
6 and 8 while the rate constants (k app ), determined under single turnover conditions for the combinations discussed above, are shown in Tables1

Figure 5 .
Figure5.Summary of data for N −1 /N 248 cis WC/WC base paring.Boxes marked in gray are consistent with cis WC/WC base-pairing; light gray marks those combinations where one combination (or weak agreement/ non-WC/WC pairing e.g.GU-pairing) are consistent with cis WC/WC base-pairing, e.g.cf.pSu1U −1 /A 248vs pSu1A −1 /U 248 -combinations.Boxes marked in red highlight the combinations that are not in agreement with cis WC/WC base pairing, while no color indicates other combinations.The grey ExpS boxes refer to the Experimental Series, e.g.1.1-1.4 and 2.1-2.8etc., as shown in Figs.4 and 6, and Tables 1, 2 and 3. (A) Experiment series using pSu1 (ExpS 1.1-1.4)and pATSer (ExpS 2.1-2.8)variants, which can establish a productive interaction with the TBS region in the S-domain (see main text for details).(B) Experiment series using pATSerGAAA (ExpS 3.1-3.6)and pMini3bp (ExpS 4.1-4.12)variants, which cannot form a productive interaction with the TBS region in the S-domain (see main text for details).(C) Experiment series for model substrates with a 3-methyl group at U −1 (ExpS 5.1-5.3).

Figure 6 .
Figure 6.Frequencies of cleavage-site selection for 3-methylated substrates by Eco RPR 248 variants.Histograms summarizing frequencies of cleavage at +1 in % during Eco RPR-mediated cleavage of pATSer3mUG (ExpS 5.1), pATSer3mUG GAAA (ExpS 5.2) and pMini3bp3mUG (ExpS 5.3) as indicated.We used the 5ʹ cleavage fragments to calculate the frequencies of cleavage at +1; mean and experimental errors were calculated from at least three independent experiments.

Figure 7 .
Figure 7. Kinetics of cleavage of pATSer with the Eco RPR 248 variants and Arrhenius plots.(A) Rate of cleavage of pATSerUG as a function of increasing concentration of the Eco RPR 248 variants.The experiments were performed at 37 °C in buffer C containing 800 mM Mg 2+ as described in "Materials and methods".The data represent mean and experimental errors from at least three independent experiments.Insets correspond to Eadie-Hofstee plots using the primary data and the k obs and k obs /K sto values presented in Table4.(B) Arrhenius plots of temperature dependence of k obs for the Eco RPR 248 variants as indicated.The data are summarized in Table4and the temperatures are in Kelvin.The values given in the inset correspond to the calculated E a (activation energy) values.

Figure 9 .
Figure 9. Illustration of base stacking.(A) Stacking of the discriminator base, D +73 (in magneta), on the G +1 /C +72 base pair in the crystal structure of tRNA Phe (PDB code 1EVV) 64 .(B) Stacking of residue A 248 (in magenta and E. coli numbering, Fig.1) on the tRNA Phe G +1 /C +72 base pair (in green) in the crystal structure of the RNase P-tRNA Phe complex (PDB code 3Q1R)18 .Grey spheres represent Me(II)-ions.(C) Stacking and the RCC A-RPR interaction (in green) in the crystal structure of the RNase P-tRNA Phe complex (PDB code 3Q1R)18 .Stacking residues in magenta.D 73 corresponds to the discriminator base at position +73 in tRNA94 while the RPR numbering refers to E. coli numbering (Fig.1).Note that A 295 in E. coli corresponds to U 266 in T. maritima RPR18 .Stacking residues, the tRNA 3ʹ terminal A 76 and the RPR residue, are marked in magenta.(D) Codon-anticodon interaction in the ribosomal A-site where residues in magenta stack as shown in the figure.p34-p37 correspond to positions in the tRNA anticodon loop.Gray residues represent the codon and residues marked in orange residues correspond to A1492 and A1493 in 16S rRNA (PDB code 2J02)80 .(E,F) Stacking interactions in the ribosomal peptidyl transfer center, panel E (A-site) and panel F (P-site) as indicated.Orange residues correspond to rRNA residues interacting with the tRNA, green residues refer to tRNA and the tRNA discriminator base is highlighted in magenta (PDB code 5IBB)79 .The images were created using PyMOL (Schrödinger, LLC). https://doi.org/10.1038/s41598-023-41203-4

Table 1 . Rate of cleavage (k app ) for pSu1, pATSer and pATSer GAAA derivatives using different RPR variants without the C5 protein. The data represent mean ± experimental errors calculated from at least three
independent experiments and are expressed as cleavage per min per pmol of RPR.Dependent on RPR substrate combination, between 0.4 and 0.8 μM RPR was used, and 2 nM of substrate in all cases.The reactions were performed at 37 °C in buffer C at 800 mM Mg 2+ (see "Materials and methods") and the "substrate-N 248 " combinations showing the highest rates are highlighted in bold.

Table 2 .
Rate of cleavage (k app ) of pMini3bp for RPR variants without the C5 protein.The data represent mean ± experimental errors calculated from at least three independent experiments and are expressed as cleavage per min per pmol of RPR.Dependent on RPR substrate combination, between 0.4 and 0.8 μM RPR was used, and 2 nM of substrate in all cases.The reactions were performed at 37 °C in buffer C at 800 mM Mg 2+ (see "Materials and methods") and the "pMini3bp-N 248 " combination showing the highest rate is highlighted in bold.

Table 3 .
Ratewas no apparent difference for the other 248 variants.The increased cleavage at −1 for G 248 is also noticeable by comparing k app(+1) and k app(−1) values (Table1).The wild-type A 248 and G 248 variant cleaved pSu1G −1 at +1 with slightly lower frequencies than C 248 and U 248 (ExpS 1.3).Moreover, k app(+1) values for all 248 variants were similar (ExpS 1.3) while k app for cleavage at −1 were higher for A 248(wt) and G 248 compared to C 248 and U 248 .Finally, with pSu1U −1 (ExpS 1.4) we detected modest differences in cleavage frequency at +1 and k app(+1) (at most 2.6-fold change comparing A 248(wt) and U 248 of cleavage (k app ) of as a function of having 3-methyl U at −1.The data represent mean ± experimental errors calculated from at least three independent experiments and are expressed as cleavage per min per pmol of RPR.Dependent on RPR substrate combination, between 0.4 and 0.8 μM RPR was used, and 2 nM of substrate in all cases.The reactions were performed at 37 °C in buffer C at 800 mM Mg 2+ (see "Materials and methods").The bold ExpS numbers highlight the substrates with 3mU.ExpS CL site A 248 (wt) Vol.:(0123456789) Scientific Reports | (2023) 13:14140 | https://doi.org/10.1038/s41598-023-41203-4www.nature.com/scientificreports/ and Table 1; cf.ExpS 2.3 C 248 vs. G 248 ).Fourth, pATSerUG (cf.ExpS 2.4) was almost exclusively cleaved at +1 by all 248 variants (see also Fig. . pSu1U −1 /A 248vs pSu1A −1 /U 248 -combinations.Boxes marked in red highlight the combinations that are not in agreement with cis WC/WC base pairing, while no color indicates other combinations.The grey ExpS boxes refer to the Experimental Series, e.g.1.1-1.4 and 2.1-2.8etc., as shown in Figs.4 and 6, and Tables 1, 2 and 3. (A) Experiment series using pSu1 (ExpS 1.1-1.4)and pATSer (ExpS 2.1-2.8)variants, which can establish a productive interaction with the TBS region in the S-domain (see main text for details).(B) Experiment series using pATSerGAAA (ExpS 3.1-3.6)and pMini3bp (ExpS 4.1-4.12)variants, which cannot form a productive interaction with the TBS region in the S-domain (see main text for details).(C) Experiment series for model substrates with a 3-methyl group at U −1 (ExpS 5.1-5.3).