Structure of Mycobacterium tuberculosis 1-Deoxy-D-Xylulose 5-Phosphate Synthase in Complex with Butylacetylphosphonate
, Rick Oerlemans 1, Robin M. Gierse 2,3, Riya Jotwani 1, Anna K. H. Hirsch 2,3 and Matthew R. Groves 1,*Round 1
Reviewer 1 Report
The manuscript "Structure of M. tuberculosis 1-Deoxy-D-Xylulose 5-Phosphate Synthase in Complex with Butylacetylphosphonate" by Gawriljuk et al. focuses on analyzing the active site of 1-deoxy-D-xylulose 5-phosphate synthase (DXPS) in complex with an inhibitor, butyacetylphosphonate (BAP), in the crystal. The manuscript is well written but needs some minor corrections, which I detail below:
1. Figure 1 should be supplemented by introducing the chemical formula for substituents R1 and R2.
2. Correct and uniform notations should be used in the whole manuscript, corrections are needed:
a. "D-xylulose" instead of "d-xylulose";
b. for time notation "h" instead of "hr";
c. Italic font should be used for bacteria strains, e.g.: E. coli and M. tuberculosis.
Author Response
Dear Sir/Madam,
Many thanks for the referee’s comments which have helped us to strengthen the manuscript. Please find attached our itemized response to both reviewers.
Yours,
Matthew
Reviewer 1:
The manuscript "Structure of M. tuberculosis 1-Deoxy-D-Xylulose 5-Phosphate Synthase in Complex with Butylacetylphosphonate" by Gawriljuk et al. focuses on analyzing the active site of 1-deoxy-D-xylulose 5-phosphate synthase (DXPS) in complex with an inhibitor, butyacetylphosphonate (BAP), in the crystal. The manuscript is well written but needs some minor corrections, which I detail below:
- Figure 1 should be supplemented by introducing the chemical formula for substituents R1 and R2.
Answer: We have added the chemical formulas for R1 and R2.
- Correct and uniform notations should be used in the whole manuscript, corrections are needed:
- "D-xylulose" instead of "d-xylulose";
- for time notation "h" instead of "hr";
- Italic font should be used for bacteria strains, e.g.: E. coli and M. tuberculosis
Answer: Many thanks for these corrections, which have been integrated.
Reviewer 2 Report
The manuscript of Groves et al. describes the crystal structure of 1-deoxy-D-xylulose 5-phosphate synthase (DXPS) in complex with the inhibitory ligand butylacetylphosphonate (BAP) obtained following soaking of a crystallized truncated enzyme with BAP. The thus obtained structure of this complex should guide scientists to the synthesis of improved inhibitors using in silico modelling to optimize the mutual interactions.
Selective inhibition of DXPS herein is very important as it is the rate limiting step in the methylerythritol phosphate pathway, which is used in bacteria for synthesis of isoprenoid precursors. Hence, the basic idea for obtaining structural information in order to facilitate the design of optimized inhibitory ligands is very interesting, and the obtained findings are very valuable. Overall, the manuscript however needs to be elaborated a bit more to increase its impact.
As the above comments require the inclusion of additional results, the description “major revision” is used despite the interesting findings which are presented.
Major comments:
- Although the various structures and the overall reaction mechanism is well known within the field, a scheme with the different reaction steps carrying all detailed various chemical structures is warranted for the general reader, to be able to follow the discussion in this paper. The (only) structure shown in figure 1 is also incomplete with undefined “Rs”.
- Line 71-78: Why was BAP used for soaking, as second-generation structures have higher affinity and selectivity? Can the authors comment on this choice in the manuscript?
- Line 156: “side chain of Ser112” – in contrast, according to the Fig 2A drawing it looks like it is the main chain carbonyl of Ser112 that is interacting.
- Line 157: no H-bonded water molecules are shown in Fig 2A in contrast with the text.
- Paragraph 215-226: the D-PheTrAP docked structure shows no interaction with Lys473, as the latter is facing out of the active site. This contrasts with the GAP substrate containing structure (soaked) where Lysine and the GAP phosphate are interacting as discussed in the text (previous findings). Yet mutating Lys473 to alanine leads to lower inhibitory activity of D-PheTrAP. How would the authors explain this and how reliable therefore is thus the crucial modeling which was carried out?
- Can modeled structures for respectively D-PheTrAP, BAP and GAP be made and compared (via superposed structures)?
- How is the orientation of Arg415 within the DXPS-GAP co-structure?
- Could the authors calculate the interaction energies for some ligands upon modeling and compare these results with the (published) affinity data? This likewise would give information on the quality of the modeling efforts.
- Figure S1: what is the distance between the phenyl ring of the modeled adduct and lysine473 as they seem to clash in this figure? No hydrogen-bonding with the carboxylate is seen (or shown) in this structure. The orientation of the amino acids and the ligand seems to be very different from Figure 4?
Additional remarks:
Line 10: While 3 affiliations are provided, there seems to be no author carrying the third affiliation?
Line 14: provide the full name for MEP (first use)
Line 21: provide the full name for GAP
Line 35: upon referring to the need for new scaffolds, an outdated reference of 2004 is provided, while many more could be given. Overall, some old references are used at specific places in the manuscript and some more recent reviews could / should be added.
Line 70: both the quoted reference [19] and this manuscript are too optimistic claiming low nanomolar affinity, while the description “submicromolar” seems more appropriate.
Line 74: in complex with “an” BAP. Remove “an” as there exists only one BAP structure.
Line 82: comply with the general referencing method and adjust to “Gierse et al. [20]”
At several places, blanks were omitted when indicating various units, e.g. at line 88 (correct to “1 mM”), lines 90, 95, 101, 108 and maybe additional ones.
Line 218: correct to “links”
Line 220: I presume the authors mean “with bound GAP” instead of “phosphate”
Line 281: “AlkylAPs”, use capital “A”
Line 282: insert “with peptidic enamide prodrug molecules already developed”
small mistakes - see report
Author Response
Dear Sir/Madam,
Many thanks for the referee’s comments which have helped us to strengthen the manuscript. Please find attached our itemized response to both reviewers.
Yours,
Matthew
Reviewer 2:
The manuscript of Groves et al. describes the crystal structure of 1-deoxy-D-xylulose 5-phosphate synthase (DXPS) in complex with the inhibitory ligand butylacetylphosphonate (BAP) obtained following soaking of a crystallized truncated enzyme with BAP. The thus obtained structure of this complex should guide scientists to the synthesis of improved inhibitors using in silico modelling to optimize the mutual interactions.
Selective inhibition of DXPS herein is very important as it is the rate limiting step in the methylerythritol phosphate pathway, which is used in bacteria for synthesis of isoprenoid precursors. Hence, the basic idea for obtaining structural information in order to facilitate the design of optimized inhibitory ligands is very interesting, and the obtained findings are very valuable. Overall, the manuscript however needs to be elaborated a bit more to increase its impact.
As the above comments require the inclusion of additional results, the description “major revision” is used despite the interesting findings which are presented.
Major comments:
- Although the various structures and the overall reaction mechanism is well known within the field, a scheme with the different reaction steps carrying all detailed various chemical structures is warranted for the general reader, to be able to follow the discussion in this paper. The (only) structure shown in figure 1 is also incomplete with undefined “Rs”.
Answer: We thank the reviewer for pointing out the missing points in the introduction. We have added a paragraph in Introduction detailing the reaction mechanism as well as a Scheme with all reaction steps. (4th paragraph - Section 1).
- Line 71-78: Why was BAP used for soaking, as second-generation structures have higher affinity and selectivity? Can the authors comment on this choice in the manuscript?
Answer: At the time of experiments, we only had direct access to BAP. Furthermore, the second-generation of alkylAPs were not designed through a structure-based approach, whereas we used BAP as an exemplar molecule as it is unelaborated and thus allowing for more directed structure-based optimization. A paragraph has been added to the Introduction to explain the rationale behind this approach. (8th Paragraph Section 1)
- Line 156: “side chain of Ser112” – in contrast, according to the Fig 2A drawing it looks like it is the main chain carbonyl of Ser112 that is interacting.
Answer: We thank the reviewer for the correction. This should indeed be the main chain carbonyl; it has been corrected in the text.
- Line 157: no H-bonded water molecules are shown in Fig 2A in contrast with the text.
Answer: We re-wrote the text and have changed the figure reference to clarify the hydrogen bonds made by the phosphonates, which are now more clearly represented in Figure 2 B.
- Paragraph 215-226: the D-PheTrAP docked structure shows no interaction with Lys473, as the latter is facing out of the active site. This contrasts with the GAP substrate containing structure (soaked) where Lysine and the GAP phosphate are interacting as discussed in the text (previous findings). Yet mutating Lys473 to alanine leads to lower inhibitory activity of D-PheTrAP. How would the authors explain this and how reliable therefore is thus the crucial modeling which was carried out?
Answer: We thank the reviewer for raising this observation. This is an inherent limitation of the rigid-body docking methodology used. Our ∆MtDXPS structure in complex with BAP, which was used as receptor for the docking, has the Lys473 in an outward facing conformation. The See-SAR docking algorithm used in this paper does not allow for conformational flexibility of the side-chains in the docking process. However, as seen from the phosphate bound structure (PDBID: 7A9G) the Lys473 can adopt different conformations. To further illustrate this point we performed flexible docking to highlight the lysine flexibility and its ability to potentially interact with the compound. (2nd Paragraph of Section 3.5) In addition, a discussion about the similarities between Arg480 in EcDXPS and Lys473 in MtDXPS was added to discussion (5th Paragraph of Section 4). Overall, from the docking alone, we cannot say with confidence that Lys473 will or will not interact with D-PheTrAP.
- Can modeled structures for respectively D-PheTrAP, BAP and GAP be made and compared (via superposed structures)?
Answer: We would like to thank the reviewer for this suggestion and we feel that this would aid in the interpretation of the binding modes and docking results. It has been added as figure 5 in Results Section 3.5.
- How is the orientation of Arg415 within the DXPS-GAP co-structure?
Answer: We would like to thank the reviewer for this query. The Arg415 in the DXPS-phosphate structure is in the same position as in the DXPS-BAP structure, see (new) figure 5.
- Could the authors calculate the interaction energies for some ligands upon modeling and compare these results with the (published) affinity data? This likewise would give information on the quality of the modeling efforts.
Answer: We calculated the binding energies of the BAP bound in our structure and the docked D-PheTrap using SMINA. They were -7.4 kcal/mol and -8.2 kcal/mol, respectively. These values indicate that the D-PheTrap indeed has a higher affinity for mtDXPS than BAP. However, calculated binding energies from docking approaches should be treated with caution as they can be unreliable due to the relatively simplistic scoring functions implemented, making comparisons between different molecules based purely on energies somewhat unreliable. Additionally, most of the published inhibition data available is against EcDXPS rather than MtDXPS, and the slight differences in the active site residues may mean that the measured affinities might not necessarily apply to MtDXPS. The text has been clarified to reflect this observation (Section 3.5).
- Figure S1: what is the distance between the phenyl ring of the modeled adduct and lysine473 as they seem to clash in this figure? No hydrogen-bonding with the carboxylate is seen (or shown) in this structure. The orientation of the amino acids and the ligand seems to be very different from Figure 4?
Answer: Even though the distance looks close in the figure, there is no clash between the phenyl and the Lys473 as the distance is 7 angstroms. Figure S1 was replaced by figure 5, which has a better view than the previous one. (Results Section 3.5.)
Additional remarks:
Line 10: While 3 affiliations are provided, there seems to be no author carrying the third affiliation?
Our apologies for the oversight. The third affiliation came after the editor corrections, we have added them to the corresponding authors.
Line 14: provide the full name for MEP (first use)
Line 21: provide the full name for GAP
Answer: Many thanks for these corrections, which have been integrated.
Line 35: upon referring to the need for new scaffolds, an outdated reference of 2004 is provided, while many more could be given. Overall, some old references are used at specific places in the manuscript and some more recent reviews could / should be added.
We have replaced the outdated reference for the following one:
Spellberg, B. The Future of Antibiotics. Crit. Care 2014, 18, 228, doi:10.1186/cc13948
As well as the latest publications in the development of DXPS inhibitors:
Zhu, D.; Johannsen, S.; Masini, T.; Simonin, C.; Haupenthal, J.; Illarionov, B.; Andreas, A.; Awale, M.; Gierse, R.M.; van der Laan, T.; et al. Discovery of Novel Drug-like Antitubercular Hits Targeting the MEP Pathway Enzyme DXPS by Strategic Application of Ligand-Based Virtual Screening. Chem. Sci. 2022, 13, 10686–10698, doi:10.1039/d2sc02371g.
Johannsen, S.; Gierse, R.M.; Olshanova, A.; Smerznak, E.; Laggner, C.; Eschweiler, L.; Adeli, Z.; Hamid, R.; Alhayek, A.; Reiling, N.; et al. Not Every Hit-Identification Technique Works on 1-Deoxy-d-Xylulose 5-Phosphate Synthase (DXPS): Making the Most of a Virtual Screening Campaign. ChemMedChem 2023, e202200590, doi:10.1002/cmdc.202200590.
Marcozzi, A.; Masini, T.; Zhu, D.; Pesce, D.; Illarionov, B.; Fischer, M.; Herrmann, A.; Hirsch, A.K.H. Phage Display on the Anti-Infective Target 1-Deoxy-d-Xylulose-5-Phosphate Synthase Leads to an Acceptor–Substrate Competitive Peptidic Inhibitor. ChemBioChem 2018, 19, 58–65, doi:10.1002/cbic.201700402.
Jumde, R.P.; Guardigni, M.; Gierse, R.M.; Alhayek, A.; Zhu, D.; Hamid, Z.; Johannsen, S.; Elgaher, W.A.M.; Neusens, P.J.; Nehls, C.; et al. Hit-Optimization Using Target-Directed Dynamic Combinatorial Chemistry: Development of Inhibitors of the Anti-Infective Target 1-Deoxy-d-Xylulose-5-Phosphate Synthase. Chem. Sci. 2021, 12, 7775–7785, doi:10.1039/d1sc00330e.
Line 70: both the quoted reference [19] and this manuscript are too optimistic claiming low nanomolar affinity, while the description “submicromolar” seems more appropriate.
Answer: Many thanks for this suggestion, which has been integrated.
Line 74: in complex with “an” BAP. Remove “an” as there exists only one BAP structure.
Line 82: comply with the general referencing method and adjust to “Gierse et al. [20]”
Answer: Many thanks for these corrections, which have been integrated.
At several places, blanks were omitted when indicating various units, e.g. at line 88 (correct to “1 mM”), lines 90, 95, 101, 108 and maybe additional ones.
Apologies for the oversight. We have changed all omitted blanks found in the Material and Methods section.
Line 218: correct to “links”
Answer: Many thanks for these corrections, which have been integrated.
Line 220: I presume the authors mean “with bound GAP” instead of “phosphate”
Unfortunately, in our previous structure (PDBID: 7A9G) the only direct evidence that could be found was electron density for the phosphate moiety of GAP in one of the chains, rather than unambiguous density for the entire GAP. Even though the complex was soaked with D-GAP and there are no other sources of phosphate, we cannot say with 100% confidence that what was observed in that structure was indeed D-GAP rather than a contaminating phosphate and we decided to label it as phosphate only. The text has been modified to clarify this point.
Line 281: “AlkylAPs”, use capital “A”
Line 282: insert “with peptidic enamide prodrug molecules already developed”
Answer: Many thanks for these corrections, which have been integrated.
Round 2
Reviewer 2 Report
The manuscript of Groves et al. on the studies of inhibition of 1-deoxy-D-xylulose 5-phosphate synthase (DXPS) has been considerably improved with a lot of supplementary background information and discussion as requested. Very nice (supplementary) pictures nicely illustrate now the findings of the authors and their data evaluation. The manuscript now will have broader appeal and is ready for publication.
I thank the authors for following up all referee requests and for careful answering all questions.
