Physicochemical Characterization of the Catalytic Unit of Hammerhead Ribozyme and Its Relationship with the Catalytic Activity

Round 1

Reviewer 1 Report

Tanaka et al. used NMR spectroscopy to monitor the protonation state of the two functionally important guanosine residues, G12* and G10.1*, in the model RNA duplexes 1 and 2 of hammerhead ribozymes. They talked about the possible link between the newly discovered physicochemical properties and the catalytic activity of hammerhead ribozymes. They proposed that properties such as G10.1unusually *'s high pKa could explain, for example, its higher metal affinity.

The article is well written and contains useful information that will aid in understanding the catalytic activity of hammerhead ribozymes.

I only have a few comments:

1.       The authors should explain why they don't use more basic or acidic pH values to more accurately determine the pKa of the residues. In Figure 7, for example, they tried pH as low as 1.0 for one experiment, but the minimum pH for the other was 4.5. I'm also curious if they can increase the pH for the experiments in Figure 3. If there are compelling technical reasons to avoid such extreme pHs, they should be stated; otherwise, the experiments should be repeated at additional pH levels.

2.       Please, add legends to Figures 3 and 7 to make them easier to understand.

3.       Authors should think about combining some figures to improve readability. Figures 2, 3, and 4 could, for example, be panels in a single figure. The same is true for Figures 6 and 7.

Author Response

Response to Reviewer 1 Comments:

(General Comment)

Tanaka et al. used NMR spectroscopy to monitor the protonation

state of the two functionally important guanosine residues, G12* and G10.1*, in the model RNA duplexes 1 and 2 of hammerhead ribozymes. They talked about the possible link between the newly discovered physicochemical properties and the catalytic activity of hammerhead ribozymes. They proposed that properties such as G10.1*'s unusually high pKa could explain, for example, its higher metal affinity.

 The article is well written and contains useful information that will aid in understanding the catalytic activity of hammerhead ribozymes.

I only have a few comments:

(Point 1)

The authors should explain why they don't use more basic or acidic pH values to more accurately determine the pKa of the residues. In Figure 7, for example, they tried pH as low as 1.0 for one experiment, but the minimum pH for the other was 4.5. I'm also curious if they can increase the pH for the experiments in Figure 3. If there are compelling technical reasons to avoid such extreme pHs, they should be stated; otherwise, the experiments should be repeated at additional pH levels.

(Response 1)

Thank you for pointing out this important issue. As the referee pointed out, the pK_a analysis of N7(G12*) of duplex 1 doesn't contain data below pH 4.5. However, there is the following reason.

In the pH-titration experiment under acidic conditions, we performed pH-titration below pH 4.5, and NMR spectra under pH 3.70 - 4.53 are included in Figure 6. Unfortunately, below pH 4.5, another equilibrium between a duplex state and a denatured state is also included. In fact, the NMR spectrum at pH 3.57 showed a totally different spectral shape from that of the initial duplex state under pH 7.36. This indicates that the data below pH 4.5 does not fit to a simple two-state equilibrium system that was used in the current pK_a analysis. Consequently, the spectral data below pH 4.5 were excluded from pK_a analysis of N7(G10.1) in Figure 7.

As for the pH-titration experiment within the basic region, RNA duplexes are denatured above pH 10.5. Actually, we monitored a denatured state of duplex 2 at pH 10.9 (the next titration point to pH 10.4). The NMR spectral at pH 10.9 was totally different from that at pH 10.4.

Based on the reviewer’s comment, we added the relevant figure of NMR spectra at pH 10.9 with several reference spectra at pH 7.0 and 10.4 in the supplementary materials as Figure S4 with its legend. An explanatory text was added to the legend to Figure 5.

 Line 201-202 (the legend to Figure 5):

Addition: The 1D ¹H NMR spectrum of denatured duplex 2 at pH 10.9 in the absence of CdCl₂ is shown in Figure S4 in the Supplementary Materials.

(Point 2)

Please, add legends to Figures 3 and 7 to make them easier to understand.

(Response 2)

Thank you for advising of improper legends of the two figures. Below we present actual legends as it appears in our original manuscript. These were suppressed likely due to some transformation error(s) during the submission. We are sorry for the inconvenience that could affect clarity of our work.

The legend to Figure 3 (Line 138-142):

Figure 3. Plot of the chemical shift of H8(G12*) of the model RNA duplex 1 against pH (basic region). (red circle) H8(G12*) chemical shift in the absence of CdCl₂. (blue diamond) H8(G12*) chemical shift in the presence of CdCl₂. (open circle) H8(5’-GMP) chemical shift in the absence of CdCl₂. In the respective titration data, their corresponding theoretical curves are indicated. In the theoretical curves for model RNA duplex 1, extrapolated theoretical curves are presented as broken lines.

The legend to Figure 7 (Line 238-242):

Figure 7. Plot of the chemical shift of H8(G10.1*) of the model RNA duplex 1 against pH (acidic region) in the absence of CdCl₂. (red circle) H8(G10.1*) chemical shift. (open circle) H8(5’-GMP) chemical shift. In the respective titration data, their corresponding theoretical curves are indicated. In the theoretical curves for model RNA duplex 1, extrapolated theoretical curves are presented as broken lines.

(Point 3)

Authors should think about combining some figures to improve readability. Figures 2, 3, and 4 could, for example, be panels in a single figure. The same is true for Figures 6 and 7.

(Response 3)

We thank the reviewer for his/her useful suggestion that could enhance the presentation of our results. However, Figures 2 and 4 are pretty large and their merging may not be easy considering the page space. After careful examination, we think that merging Figures 2-4 into one as well as merging Figures 6 and 7 could improve our manuscript in line with the reviewer’s suggestion. I will consult this issue with the editorial office to find an optimal form of presentation of the graphical artwork. Currently, relevant figures in the revised manuscript are preserved, as in the original manuscript.

Typos and grammatical corrections:

The line 153:

Deletion of "the": Since the direction of the chemical shift perturbation above pH 9.22 was opposite to the that of N1-deprotonation for 5’-GMP (Figure 3, open circles)

The line 177:

Addition of comma ",": This highly basic property of G12* is notable, especially when considering the imino proton of G12* is

The reference 56:

Original: The hydrolysis of rna

 Revised: The hydrolysis of rnaRNA

Reviewer 2 Report

This manuscript reports NMR analysis of the basicity (pK_a) of the catalytic sites of hammerhead ribozyme (HHRz) using model RNAs and discusses possible reaction mechanisms of HHRz based on the NMR data. The physicochemical analysis of HHRz like this study is important. However, the environment of the catalytic site (H1 of G12) of the model RNA is considered to be different from that of HHRz. This is because H1 of G12 and 2’-OH of C17 are close to each other in the crystal structure of HHRz. On the other hand, if the environment of the metal binding site (N7 of G10.1) is similar in the model RNAs and HHRz, pK_a may be similar.  

In addition to the concerns mentioned above, there are some points that the authors should address:

 1. The authors estimate the pK_a of N1 and N7 of G by observing the signal change of H8 in pH titration. However, the chemical shift of the H8 signal is affected not only by the protonation/deprotonation of N1 and N7, but also by the conformational change of RNAs induced by pH change. In the case of monomer or dimer, there is no worry about the structural change, but the effects of structural changes are a concern in the case of long RNA. Furthermore, a large chemical shift changes due to ring current effect may occur depending on the degree of stacking with the adjacent bases.

 2. During pH titration experiments, H8 signals of G12 or G10.1 overlaps with other signals. Did you confirm the signal assignment when they overlapped?

3. In relation to the points 1 and 2 above, it is recommended to label G12 and G10.1 with 15N and observe the chemical shift change of 15N signals.

 4. The authors performed pH titration experiment between 3.57 and 10.48 and estimated pK_a of G12 >> 10.48. The authors also mentioned that the double-stranded structure is destroyed at the pH outside this range. What kind of spectra were observed above pH11?

 5. Line 142, the authors wrote “An effect on the pK_a of H8(G12*)”, but “An effect on the chemical shift of H8(G12*)” should be correct.

 6. Line 144, the authors should explain why they used Cd²⁺ as a divalent metal ion.

 7. Line 172, the authors described “This very basic property of G12* is notable especially considering that the imino proton of G12* is structurally exposed to bulk solvent due to the sheared-type G-A pairing.” This is strange, so the authors should discuss why N1 of G12* is highly basic, even though the imino proton is structurally exposed to bulk solvent.

 8. Did you analyze the chemical shift change of H8 of G12 at low pH and that of G10.1 at high pH? Do the pK_as of N1 and N7 affect each other?

Author Response

Response to Reviewer 1 Comments:

(General Point)

This manuscript reports NMR analysis of the basicity (pKa) of the catalytic sites of hammerhead ribozyme (HHRz) using model RNAs and discusses possible reaction mechanisms of HHRz based on the NMR data. The physicochemical analysis of HHRz like this study is important. However, the environment of the catalytic site (H1 of G12) of the model RNA is considered to be different from that of HHRz. This is because H1 of G12 and 2'- OH of C17 are close to each other in the crystal structure of HHRz. On the other hand, if the environment of the metal binding site (N7 of G10.1) is similar in the model RNAs and HHRz, pKa may be similar.

(Response to the General Point)

We thank the reviewer for important remark. We agree that the pK_a values determined in our model RNA duplexes and values in a whole HHRz may be different. Our work is aimed at obtaining of intrinsic pK_a values within the extracted A9-G10.1 motif from a whole HHRz. The pK_a values of relevant sites in a whole HHRz (pK_a[HHRz]) are likely modulated as compared to the intrinsic pK_a values of the motif (pK_a[intrinsic]) due to interactions that occur only within whole HHRz (pK_a[modulation]):

pK_a[HHRz] = pK_a[intrinsic] + pK_a[modulation]

We report for the first time the pK_a[intrinsic] value that can be regarded a baseline pK_a value. Once pK_a values in a whole HHRz system are determined, the pK_a modulation due to specific interactions within a whole HHRz can be clearly identified. Whether intrinsic pK_a of the basal catalytic motif is preserved in a whole HHRz system remains at the moment an interesting question. In any case, by determining the pK_a[intrinsic] value, we can set a plausible baseline for the pK_a value in the catalytic mechanism of HHRzs. We therefore anticipate that the intrinsic physicochemical properties determined in our work are fundamental in respect of future mechanistic studies of HHRzs.

However, the issue pointed out by the referee is important. Therefore, the following sentences in the original manuscript are revised to highlight the corresponding issue.

Line 281-284 in the revised manuscript (Line 260-262 in the original manuscript):

Original: Therefore, the derived property of G12* and G10.1* in RNA duplexes 1 and 2 might be an intrinsic nature of the catalytic unit although their pK_a values within the full-length HHRz could be further modulated.

Corrected: Therefore, the derived property of G12* and G10.1* in RNA duplexes 1 and 2 might be an intrinsic nature of the catalytic unit although their pK_a values within the full-length HHRz could be further modulated owing to additional interactions in the whole HHRz system.

(Point 1)

In addition to the concerns mentioned above, there are some points that the authors should address:

The authors estimate the pKa of N1 and N7 of G by observing the signal change of H8 in pH titration. However, the chemical shift of the H8 signal is affected not only by the protonation/deprotonation of N1 and N7, but also by the conformational change of RNAs induced by pH change. In the case of monomer or dimer, there is no worry about the structural change, but the effects of structural changes are a concern in the case of long RNA. Furthermore, a large chemical shift changes due to ring current effect may occur depending on the degree of stacking with the adjacent bases.

(Response 1)

We thank the reviewer for his/her remark. In order to explore this possibility, we performed the pH-titration experiments in two different RNA duplexes. As shown in the manuscript (Figures 2 and 5), the relevant pH-titration experiments both provided coherently the same result; the H8(G12*) resonance was not perturbed within the pH range. The available experimental data are consistent with the current interpretation, so far.

However, the comment mentioned above is an important remark. To highlight this issue and issue-3, we added the following sentence as indicated.

Line 272-275 in the revised manuscript

Addition: As a remark, current analyses rely on the ¹H NMR chemical shift perturbations which may be in principle affected by various structural effects. For more solid pK_a analysis, an ¹⁵N NMR study would be desirable in the future although there is no obvious inconsistency between experimental data and current interpretation.

(Point 2)

During pH titration experiments, H8 signals of G12 or G10.1 overlaps with other signals. Did you confirm the signal assignment when they overlapped?

(Response 2)

The assignment of H8(G12) and H8(G10.1) resonances under neutral pH was carried out following the assignments published in our previous work (Reference 33).

To assure the trace of the NMR signals of the two base protons within the titration experiment (basic region), a ¹H-¹H NOESY spectrum was recorded at pH 8.18. By using sequential NOE walks, base protons were assigned, and the H8(G12) resonance was unambiguously identified. Thus, the signal trace of the H8(G12) proton is reliable. In addition, a natural abundance ¹H-¹³C HSQC spectrum was also recorded. In a basic pH region, ¹H-¹H NOESY and ¹H-¹³C HSQC spectra were almost the same as those at the neutral pH (Reference 33). These spectra are shown in the Supplementary Materials, Figures S2 and S3.

To assure the trace of the NMR signals of base protons within the titration experiment under an acidic region, a natural abundance ¹H-¹³C HSQC spectrum was recorded at pH 4.40. The cross peaks were assigned with reference to the corresponding peaks in the ¹H-¹³C HSQC spectra under neutral and basic conditions. The ¹H-¹³C HSQC spectrum under acidic conditions was added to the Supplementary Materials, Figure S5.

The legends to Figures 2 and 6 were modified accordingly as follows.

Line 127-128 (The last sentence in the legend to Figure 2):

Addition: To assure the trace of the NMR signals of base protons, ¹H-¹H NOESY and natural abundance ¹H-¹³C HSQC spectra were recorded at pH 8.18 (Figures S2 and S3).

Line 232-233 (The last sentence in the legend to Figure 6):

Addition: To assure the trace of the NMR signals of base protons, a natural abundance ¹H-¹³C HSQC spectrum was recorded at pH 4.40 (Figure S5).

(Point 3)

In relation to the points 1 and 2 above, it is recommended to label G12 and G10.1 with 15N and observe the chemical shift change of 15N signals.

(Response 3)

As the referee proposed, such well-designated NMR experiment would eliminate the uncertainty. Unfortunately, we don't have an ¹⁵N-labeled sample currently. Considering the time for sample preparation and ¹⁵N measurement, we anticipate that performing of the proposed experiment would exceed the available period for revision (10 days). However, this remark is important in general as it may inspire other researchers and therefore, we added the following sentence in the revised manuscript as written in issue 1.

Line 272-275 in the revised manuscript

Addition: As a remark, current analyses rely on the ¹H NMR chemical shift perturbations which may be in principle affected by various structural effects. For more solid pK_a analysis, an ¹⁵N NMR study would be desirable in the future although there is no obvious inconsistency between experimental data and current interpretation.

(Point 4)

The authors performed pH titration experiment between 3.57 and 10.48 and estimated pKa of G12 >> 10.48. The authors also mentioned that the double-stranded structure is destroyed at the pH outside this range. What kind of spectra were observed above pH11?

(Response 4)

In duplex 2, we recorded 1D ¹H NMR spectrum at pH 10.9 in the absence of CdCl₂. The corresponding spectrum exhibited sharpened and well-separated signals and became a totally different spectrum from those below pH 10.4, which indicates the denaturation of duplex 2. This spectrum was added to the Supplementary Materials, Figure S4, as indicated in the legend to Figure 5 (the last sentence).

Line 201-202 (the legend to Figure 5):

Addition: The 1D ¹H NMR spectrum of denatured duplex 2 at pH 10.9 in the absence of CdCl₂ is shown in Figure S4 in the Supplementary Materials.

(Point 5)

Line 142, the authors wrote "An effect on the pKa of H8(G12*)", but "An effect on the chemical shift of H8(G12*)" should be correct.

(Response 5)

We thank the referee for pointing out an erroneous misleading sentence that was corrected as follows.

Line 142-144 in the original manuscript:

Original: An effect on the pK_a of H8(G12*) in the presence of a divalent metal cation was examined because the HHRz can be activated in the presence of divalent metal cation [15-17,50-54].

Line 143-146 in the revised manuscript:

Revised: An effect on the pK_a of H8(G12*) in the presence of a divalent metal cation was examined We further confirm if the pK_a of N1(G12*) is affected by the presence of a divalent metal cation because the HHRz can be activated in the presence of divalent metal cation [15-17,50-54].

(Point 6)

Line 144, the authors should explain why they used Cd2+ as a divalent metal ion.

(Response 6)

We thank the referee for pointing out the insufficient explanation that was corrected in the revised manuscript as follows.

Line 148-150 in the revised manuscript:

Addition: As a divalent metal cation, Cd²⁺ was selected since it is a strong binder for the A9-G10.1 motif with an NMR-compatible "diamagnetic" character [37-40] and has been used for several kinetic experiments [53, 61, 62].

(Point 7)

Line 172, the authors described "This very basic property of G12* is notable especially considering that the imino proton of G12* is structurally exposed to bulk solvent due to the sheared- type G-A pairing." This is strange, so the authors should discuss why N1 of G12* is highly basic, even though the imino proton is structurally exposed to bulk solvent.

(Response 7)

Indeed, this issue should be better described as was pointed out by the referee.

In general, pK_a alteration may occur due to two effects. One, is a structural factor such as masking of the protonation/deprotonation site by hydrogen bonding, stacking, shape fitting, and so on. The other is an electronic effect where the electronic state of the acid/base-moiety is modulated, e.g. by π-π stacking, hydrogen bonding, electrostatic interaction, and so on. Based on this situation, we postulated that the reason for the pK_a alteration would not be due to the first case (structural factor) but the second case (electronic effect). The sentence shown above was written in the manuscript to deny the structural factor (sequestration of the protonation/deprotonation site). It implies that the A9-G10.1 motif would have a unique electronic structure caused by its specific intra-motif interaction, such as non-standard sheared-type G12-A9 base pairing and/or the π-π stacking of the G12-A9 and G10.1-C11.1 base pairs. We described these possible effects in relevant part of the revised manuscript as follows.

Line 179-185 in the revised manuscript:

Addition: In general, pK_a alteration is brought by structural factors (sequestration of the protonation/deprotonation-site), or electronic effects (modulation of the electronic structure of acid/base-moiety), or by a combination of both effects. In the case of the A9-G10.1 motif, the structural factor doesn't explain the pK_a alteration as mentioned above. Here the distinguished pK_a value may be most likely attributed to the electronic effect caused by non-standard sheared-type G12-A9 base pairing and π-π stacking of the G12-A9 and G10.1-C11.1 base pairs.

Line 186-187 in the revised manuscript:

Addition: To generalize the above observation (basicity enhancement of G12*), the basicity of G12* was studied ~

(Point 8)

Did you analyze the chemical shift change of H8 of G12 at low pH and that of G10.1 at high pH? Do the pKas of N1 and N7 affect each other?

(Response 8)

It is an important question, indeed. In the pH-titration experiment toward basic conditions, the H8(G10.1*) resonance can be traced, and almost retained its initial chemical shift. Furthermore, other resonances were not perturbed, either. Thus, no significant modulation of all the signals was observed, and a mutual relationship between G12* and G10.1* was unclear. In the pH-titration experiment toward acidic conditions, the H8(G12*) resonance almost retained its original chemical shift. Thus, the modulation of the H8(G12*) resonance due to the protonation of N7(G10.1*) was not so evident. From these data, the pK_a modulation of G12* and G10.1* may be attributed to some effect, other than direct interaction between the G12* and G10.1* bases. This viewpoint was added to the revised manuscript as follows.

Line 268-272 in the revised manuscript:

Addition: Noteworthy, the mutual correlation between the chemical shifts of H8(G12*) and H8(G10.1*) was not observed. It implies that the direct interaction of G12* and G10.1* residues is very weak and the altered pK_a values of N1(G12*) and N7(G10.1*) are most likely due to some effect(s) other than the direct interaction of the two residues.

Typos and grammatical corrections:

The line 153:

Deletion of "the": Since the direction of the chemical shift perturbation above pH 9.22 was opposite to the that of N1-deprotonation for 5’-GMP (Figure 3, open circles)

The line 177:

Addition of comma ",": This highly basic property of G12* is notable, especially when considering the imino proton of G12* is

The reference 56:

Original: The hydrolysis of rna

 Revised: The hydrolysis of rnaRNA

Round 2

Reviewer 1 Report

The manuscript has improved a lot and it seems I think it is ready for publication.

I think I did not explain properly my concerns about legends in Figures 3 and 7. What I meant by 'legend' was the 'figure legend' or 'graphical legend', it is (for instance) adding in Fig. 3 a blue diamond, a white circle, and a red circle and a short text showing what they represent, then the reader does not need to read the 'text legend' to know what each symbol or line means.

Reviewer 2 Report

The authors have made substantial efforts to address my concerns, and they revised their manuscript accordingly. I think the manuscript has improved a lot.