Design strategy for serine hydroxymethyltransferase probes based on retro-aldol-type reaction

Serine hydroxymethyltransferase (SHMT) is an enzyme that catalyzes the reaction that converts serine to glycine. It plays an important role in one-carbon metabolism. Recently, SHMT has been shown to be associated with various diseases. Therefore, SHMT has attracted attention as a biomarker and drug target. However, the development of molecular probes responsive to SHMT has not yet been realized. This is because SHMT catalyzes an essential yet simple reaction; thus, the substrates that can be accepted into the active site of SHMT are limited. Here, we focus on the SHMT-catalyzed retro-aldol reaction rather than the canonical serine–glycine conversion and succeed in developing fluorescent and 19F NMR molecular probes. Taking advantage of the facile and direct detection of SHMT, the developed fluorescent probe is used in the high-throughput screening for human SHMT inhibitors, and two hit compounds are obtained.


Response to the referees' comments and revisions that have been made
We thank all of the reviewers for their comments. These have been very helpful in further improving the manuscript. We revised our manuscript in lights of all the comments as follows. 2

Original comments from Reviewer #1
In this submission to Nature Communications, Sando and coworkers describe probes that allow for convenient fluorescence-based and 19-F NMR-based direct readout assays for SHMT (serine hydroxylmethyltransferase) activity. This is cleverly done; the authors have taken advantage of the ability of this enzyme to catalyze a retro-aldol reaction with erythrobeta-aryl substituted L-serine analogues independent of the tetrahydrofolate (THF) cofactor.
This allows for direct release of either a fluoresent or a 19-containing aromatic aldehyde, and its immediate detection by a plate-reading fluorimeter or by NMR, respectively. Such assays are useful chemical biological tools for the rapid screening of the effects of modulators or PTM (post-translational modification) on the activity of native SHMT, and, as such, will certainly draw the attention of many in the chemical biology community. The authors also demonstrate that this new assay serves as a useful platform for parallel screening in the search for new SHMT inhibitors.
Moreover, they demonstrate proof of principle here by utilizing the new platform to identifyi two new hit scaffolds with submicromolar IC(50)s. The authors use complementary assays (ITC, thermal denaturation, etc) to verify these hits, and they also demonstrate that both compounds inhibit the physiologically important L-serine to glycine plus N5,N10-CH2-THF reaction. Temperature dependence of inhibitor binding is also studied, allowing the authors to sort out entropic vs. enthalpic contributions.
Because the literature indicates that the L-erythro-stereoisomers are the best retro-aldolase substrates for SHMT, the authors are careful to target these isomers. Relatively stereochemistry is controlled by utilizing an N,N-dibenzylglycinate enolate that tends to give the E-enolate geometry, thereby providing predominantly the erythro (anti) aldol product. Subsequent chiral derivatization allows the authors to generate separable diastereomers and thereby gain access to desired L-erythro-stereoisomer as well as establish the absolute stereochemistry via extra crystallography.
All in all, this is a very nice study, and, in my view, with a few modifications and enhancemnts of the discussion, such a piece would appeal to the wide-ranging readership of 3 Nature Communications.
The authors should discuss clearly that the biologically relevant L-Ser to Gly + N5,N10-CH2-THF reaction can be assayed through a coupled reaction with methylene-THF dehydrogenase, as was done in the studies by Diederich and Chaiyen, for example (J. Med. Chem. -2017 -Reference 17 andChemMedChem 2018, 13, 931-934 (please add this reference!). This results in the conversion of NADP+ to NADPH which can be observed by UV absorbance at 340 nm or, in principle, by fluorescence emissiion at ~460 nm. This would be a two-enzyme coupled alternative to the assay presented in this paper.
The authors should also highlight the importance of developing inhibitors of the SHMT enzyme more clearly and comprehensively. Beyond being a validated target for malaria (Plasmodium falciparum enzyme), to my knowledge, the human enzyme remains the only enzyme of the three enzyme one carbon cycle for which established chemotherapeutic agents have not yet been developed. The authors should discuss the importance of inhibitors of the other two enzymes of the one carbon cycle, DHFR (dihydrofolate reductase) and TS (thymidylate synthase), for chemotherapy. As for the 19-F assay, the authors should add a broader discussion of previous uses of 19-F NMR to study PLP enzyme inhibition. Some references for these discussion points are suggested below:

Point-by-point response to the comments of Reviewer #1
We wish to thank the reviewer for the comments and for providing constructive suggestions.
We have addressed all the points raised by the reviewer through new text. ü Comment. The authors should discuss clearly that the biologically relevant L-Ser to Gly + N5,N10-CH2-THF reaction can be assayed through a coupled reaction with methylene-THF dehydrogenase, as was done in the studies by Diederich and at 340 nm or, in principle, by fluorescence emissiion at ~460 nm. This would be a two-enzyme coupled alternative to the assay presented in this paper.

>>>Response to Comment.
According to the reviewer's comment, we have added new sentences and references about an assay system of SHMT enzymatic reaction utilizing a coupled reaction with methylene-THF dehydrogenase.
[ Revised manuscript,p.7, For example, in the case of SHMT coupled assay, 11,15 SHMT firstly produces Gly and CH 2 -THF from Ser and THF. Then, the conversion of coenzyme NADP + to NADPH is caused by a coupled enzyme methylene-THF dehydrogenase using the CH 2 -THF as a substrate. By monitoring this NADP + -NADPH conversion with UV or fluorescence, SHMT activity can be indirectly detected. >>>Response to Comment.
We thank the reviewer for the constructive suggestion. Following the reviewer's comment, we have changed the main text so that the significance of development of SHMT inhibitors is clearly and comprehensively explained.
[Revised manuscript, p.4, line 13 -p.5, line 3] The development of SHMT inhibitors has been performed especially toward treatment of two types of diseases. The first one is antimalarial drug. [8][9][10][11][12][13][14][15] Malaria is a life-threatening disease that spreads to people through infected anopheles mosquitoes. It has a tremendous impact globally, 216 million people are infected in 2016, and 445,000 people have died. 13 In addition, the resistance of malaria parasites against existing antimalarial drugs has become a serious problem. Under such circumstances, researches and investigations for new inhibitors against malaria SHMT has been conducted. The second one is anticancer drug. 1,2 In chemotherapy, three enzymes of one-carbon metabolism, SHMT, dihydrofolate reductase (DHFR), 16 and thymidylate synthase (TS) [17][18][19] are potent target enzymes strongly related to cell proliferation (Fig. 1c).
Actually, inhibitors targeting DHFR and TS, such as methotrexate and fluorouracil respectively, have been used for a long time as anticancer agents. Among the three enzymes of the one-carbon metabolism, to our knowledge, human SHMT is the only enzyme to which an established chemotherapeutic agent has not yet been developed. Therefore, human SHMT has attracted attention as a target enzyme of anticancer drug.
ü Comment. The authors should discuss the importance of inhibitors of the other two enzymes of the one carbon cycle, DHFR (dihydrofolate reductase) and TS (thymidylate synthase), for chemotherapy.

>>>Response to Comment.
We agree with the reviewer's constructive comment, we have added a new sentence and figure (Fig. 1c) showing the importance of inhibitors of DHFR and TS in chemotherapy.
Actually, inhibitors targeting DHFR and TS, such as methotrexate and fluorouracil respectively, have been used for a long time as anticancer agents. Among the three enzymes of the one-carbon metabolism, to our knowledge, human SHMT is the only enzyme to which an established chemotherapeutic agent has not yet been developed. Therefore, human SHMT has attracted attention as a target enzyme of anticancer drug.  The authors claim their probes hold distinct advantages over existing methods such as coupled enzyme assays. The authors also noted that their probes contain two asymmetric centers and thus motived them to synthesize the various isomers to elucidate which form would be the optimal substrate for SHMT. The author complimented these studies by determining the rates for each isomer. The author then performed NMR / MRI studies using the 19F probe and showed a difference in signal when comparing the unactivated probe and the SHMT-treated probe. Finally, the authors employed their fluorescent probe in a screen to identify 2 inhibitors of SHMT activity. Overall this study was nice but as a reviewer I do not believe it is at the caliber of Nature Communications. The authors are urged to consider resubmitting to a different journal after addressing the points below.
1) One of the major advantages of having a direct SHMT probe over existing methods such as the coupled-enzyme assay is that it can be used on intact samples (e.g., live cells).
However, in each of the instances the authors selected to use the probe in vitro or in homogenates. I suspect the issue is that the probes are not cell permeable and thus, the authors were constrained to working around this limitation. If my suspicions are correct the authors should mask the carboxylate with a chemical group such as AM ester which can then be removed by intracellular esterases.
2) The authors mentioned there are two SHMT isoforms, why was this only tested against only one of the isoforms? How can you ensure that you are not getting pan-reactivity? And if you do have reactivity with both (which should be the case based on the proposed 12 mechanism) this decreases the utility of the probe because in each instance you will need to do much more elaborate control experiments with siRNA KO etc.
3) The authors were not careful in characterizing the stability of the probe or of the turned over product. This was a major sticking point for me. In particular, are there enzymes found in living systems can activate or change the properties of the probe? The authors should at least test this against CYPs. More concerning, however, is the stability of the turned over aldehyde products. The authors made no attempts to evaluate its stability after the product has been formed. Fluorogenic aldehyde probes can form Schiff-bases, iminiums, they can cyclize with Cys and homo-Cys (see aldehyde based probes for these aminio acids) and they can be oxidized to the acid product by a variety of ROSs. Since the authors claim their probe will likely find utility in cancer, oxidative stress in tumors is a major problem.
4) When I review papers for high profile journals such as Nat Commun, I consider whether the paper can spark new discovery in various fronts. From a probe perspective the current design is limited to only producing aldehyde products. The major problem with this is that aromatic aldehydes are typically non-fluorescent due to donor-PeT quenching. This means any other dye scaffold other than the naphthalene used by the authors would result in an on-off response. Turn-off probes are not desirable in the community because other factors can lead to a decrease in signal such as dye efflux, bleaching etc. In fact, this is likely why the authors decided to make a 19F probe instead of other fluorescent analogs because the MRI version is not impacted by quenching. I thought this was clever and resourceful. 5) While it may seem on first glance that the authors were being scientifically rigorous by determine which isomer is the best substrate. These experiments are simply not important, nor do add significantly to the paper. If this information was responsible for helping the authors design the current probe that is a different story. As is, it is just distracting. It is similar to how a lot of probe-based papers add DFT calculations that have no true value beyond bulking up the study.
6) Developing 19F MRI probes is an area that our group has strongly considered in the past. 7) I understand it was necessary to do something more with the probe to try make the study more comprehensive. I would have preferred the cell studies mentioned above but because it does not seem like the probes are compatible with live systems, the authors decided to perform a screen. You can do exactly the same screen using the coupled enzyme assay since it is in vitro. While the coupled assay may be operationally more complex, it can still be done. This brings us back to the question of whether the current probes have made a big enough of an advance to warrant publication in a high profile journal.

Point-by-point response to the comments of Reviewer #2
We wish to thank the reviewer for the comments and for providing constructive suggestions.
We have addressed all the points raised by the reviewer through discussions and/or new experiments. In fact, the present design strategy realized the development of the first fluorescent and 19 F chemical probes targeting SHMT (Fig. 2), which could be used for inhibitor screenings under crude/opaque biological conditions, and realized the successful finding of two new SHMT inhibitor candidates (Fig. 6). We believe that this is one of the most significant features of this paper, and in-cell applications are beyond this initial scope.
However, we agree with your comment that it is also important to show the possibility that the present design strategy can be used in designing fluorescent probes for various biological applications.
Therefore, based on the present design strategy, we additionally demonstrated the development of a new type of fluorescence turn-on probe S1 for SHMT, which produces fluorescent chromophore after tandem retro-aldol-β-elimination reaction (Supplementary Fig. S13).
This result clearly indicates that our probe design can be applied for not only aldehyde-carrying DMANA but also other bright fluorophores such as hydroxycoumarin and resorufin which have been used in the design of fluorescent molecular probes for biological applications including cellular assays (e.g. C. J. Chang et al., Chem. Commun., 44, 4647-4649 (2007)  [Revised manuscript, p. 23, line 18 -p. 24, line 2] In fact, based on the strategy, we could also develop a new type of fluorescence turn-on probe for SHMT, which produces fluorescent chromophore upon reaction with SHMT through tandem retro-aldol-β-elimination reaction (Supplementary Fig. S13, S14). This new turn-on mechanism could allow researchers to design SHMT probe with various bright fluorophores such as hydroxylcoumarin and resorufin, 41,42 demonstrating the versatility of this design strategy.
2) The authors mentioned there are two SHMT isoforms, why was this only tested against only one of the isoforms? How can you ensure that you are not getting pan-reactivity? And if you do have reactivity with both (which should be the case based on the proposed mechanism) this decreases the utility of the probe because in each instance you will need to do much more elaborate control experiments with siRNA KO etc. >>>Response to Comment. We thank for this suggestion. According to the reviewer's comment, we have checked the SHMT isoform selectivity of our fluorescence probe. It was found that fluorescence probe 1 reacts with both isoforms SHMT 1 and 2. This pan-reactivity would be useful for SHMT researches such as high throughput inhibitor screening against SHMT2. Since SHMT is present in various species and has several isoforms, the wide-reactivity against SHMTs could be a benefit for various applications. The information about the SHMT isoform selectivity has been added in the main text (p. 14, lines 9-12) and supplementary information

18
In addition, under our experimental conditions (units of SHMT were defined by the reactivity with DL-erythro-β-phenylserine), fluorescent probe 1 was found to be more responsitve to SHMT2 than SHMT1. This fact indicates that it would be possible to design SHMT molecular probes with higher isoform selectivity. In fact, we have already obtained several SHMT substrates that show high isomer selectivity depending on the structure of chromophore. We had not include this information in this manuscript because it had been a different topic from the main purpose of this research, reporting the general molecular design and the inhibitor screening for SHMT. However, if the editor or reviewer thinks this data is better to be included in this manuscript, we will consider adding it.
3) The authors were not careful in characterizing the stability of the probe or of the turned over product. This was a major sticking point for me. In particular, are there enzymes found in living systems can activate or change the properties of the probe? The authors should at least test this against CYPs. More concerning, however, is the stability of the turned over aldehyde products. The authors made no attempts to evaluate its stability after the product has been formed. Fluorogenic aldehyde probes can form Schiff-bases, iminiums, they can cyclize with Cys and homo-Cys (see aldehyde based probes for these aminio acids) and they can be oxidized to the acid product by a variety of ROSs. Since the authors claim their probe will likely find utility in cancer, oxidative stress in tumors is a major problem.
We thank the reviewer for this important advice. We agree with the reviewer's comment concerning the stability of the probe and the turned over product. The stability of the probe under biological conditions was partly shown in Fig. 5d ( 19 F NMR in rat liver homogenate condition). The 19 F NMR probe worked in rat liver homogenate including various metabolizing enzymes including CYPs. Such a crude enzyme mixture did not affect the 19 F NMR signal of the probe and product. >>>Response to Comment.
We thank the reviewer's comment. To rebut the reviewer's comment about the limitation of our fluorescent probe design, we have newly designed a new type of turn-on fluorescence probe S1. The probe S1 was synthesized by 4-steps as shown in Supplementary Fig. S14.  The new probe S1 is designed to produce fluorescent chromophore upon reaction with SHMT thorough tandem retro-aldol-β-elimination reaction (Supplementary Fig. S13a). In fact, the new probe S1 reacted with SHMT1 and produced fluorescent 7-hydroxycoumarin carrying no aldehyde. It was confirmed to be a turn-on type SHMT fluorescent probe ( Supplementary Fig. S13b). This result clearly indicates that our probe design is not limited to only aromatic aldehyde and that other bright fluorophores such as hydroxycoumarin and resorufin can be used as same 22 as previously reported turn-on fluorescent probes based on the β-elimination mechanism, e.g. C. J. Chang et al., Chem. Commun., 44, 4647-4649 (2007) andS. Q. Yao et al., Nat. Commun., 5, 3276 (2014).
Thanks to this revision based on the comment by reviewer, we are sure that our design strategy is further strengthened.
These points are discussed at the conclusion section in main text (p. 23, line 18 -p. 24, line 2 and Supplementary Fig. S13).

Reviewer #3
Original comments from Reviewer #3 Reviewer #3 (Remarks to the Author): Recommendation: Publish in Nature Comm. after minor revisions. Comments: The manuscript "Design Strategy for Serine Hydroxymethyltransferase Probes Based on Retro-Aldol-Type Reaction" from the Prof. Sando et al. describes the development of chemical probes for the enzyme SHMT as markers for the application in fluorescence and 19F NMR spectroscopy. Furthermore, they were able to demonstrate the convenience and benefit of their methods by e.g. applying the chemical probes to a high-throughput screening to identify two potential lead structures for drug development, which is a highlight of the paper.
This work nicely illustrates the use of a SHMT-induced retro-aldol reaction to access either fluorescent or 19F labelled chemical probes on a rather complex target. Especially, the 19F labelled chemical probe was shown be a potential tool for the SHMT detection in biological samples. Overall the work is of high quality and merits the publication in Nature Comm. It will find a broad audience that includes pharmacologists, medicinal chemists, and biochemists. This publication will fill a gap in the development of SHMT-related potential drugs and will contribute to a success in this field.
Major revisions: Despite the great science in this work, the major weakness of this manuscript is the writing, which does not appear the standard of Nature Comm. and needs some improvement.
Furthermore, the literature is not appropriately covered in the introduction part, which is already proven by "only" 19 citations in total.
1) The fields of fluorine MRS and MRI in biomedicine and fluorescence spectroscopy are 2) The research on the target SHMT does not only include the therapeutic area of cancer but also Malaria, which needs to be mentioned and introduced, because the pyrazolopyran-based inhibitors reported in the manuscript were also used against

Point-by-point response to the comments of Reviewer #3
We wish to thank the reviewer for the comments and for providing constructive comments and comprehensive English proofreading. We have addressed all the points raised by the reviewer through new experiments and/or new text.
ü Comment. Despite the great science in this work, the major weakness of this manuscript is the writing, which does not appear the standard of Nature Comm. and needs some improvement. Furthermore, the literature is not appropriately covered in the introduction part, which is already proven by "only" 19 citations in total.

>>>Response to Comment.
We thank the reviewer for the constructive comment to improve this manuscript. Following the reviewer's comment, we have revised the manuscript and added 25 references for covering related literatures appropriately. We hope that our revisions have adequately addressed the reviewer's concerns. ü Comment. 4) The authors could comment on possible side-effects and toxicity of the corresponding aldehydes especially Schiff-base formation in vivo.

>>>Response to Comment.
We agree with the reviewer's comment. The product aldehyde might react with biological relevant nucleophiles through forming Schiff's base.
Furthermore, we have evaluated the toxicity of DMANA (Supplementary Fig. S7). Significant toxicity was not observed from the cytotoxicity test. These results indicate that the significant side-effects and toxicity are not problematic under the concentration and time range of fluorescence experiments. We have added the discussion about the stability and toxicity of DMANA in the main text and Supplementary   Information (p. 14, lines 1-8, Supplementary Fig. S6, Supplementary   Fig. S7).
However, under the concentration range of 19 F NMR/MRI, 19 F product aldehyde might show the side-effects and toxicity, because the required concentration is high for detection of 19 F NMR/MRI signal. For practical in vivo MRI, it might be necessary to improve the sensitivity to lower the required concentration. These points have also been discussed in the main text (p. 24, lines 15-19).  Basically, we think that it is difficult to conduct in vivo experiments by using the current probes and machine setup. However, regarding fluorescence probe 1, depending on the two-photon efficiency of the turned-over product, it might be possible to detect the product near the surface of the mouse/rat by using a two-photon excitation microscope. We thank the reviewer for the comment. Following the reviewer's comment, we have added an explanation of manual modeling process as follows (caption in Supplementary Fig. S3). We hope that our revisions have adequately addressed the reviewer's concerns.

Regarding practical in vivo
[Revised Supplementary  We thank the reviewer for the constructive comment. Following the reviewer's comments, we have changed to "highlights". ü

>>>Response to Comment.
We thank the reviewer for the constructive comment. Following the reviewer's comments, we have unified to "CH 2 -THF".

>>>Response to Comment.
We thank the reviewer for the constructive comment. Following the reviewer's comments, we have changed positive charge to superscript "NADP + ".
ü Comment. Page 6, line 6: Figure 2a shows … >>>Response to Comment. We thank the reviewer for the constructive comment. Following the reviewer's comments, we have changed.
ü Comment. Page 6, line 10: Figure 2b shows … (redundant to the previous section and bumpy to read).

>>>Response to Comment.
We thank the reviewer for the constructive comment. In the revised manuscript, we have rewritten the sentences. We hope that our revisions have adequately addressed the reviewer's concerns. If the reviewer needs more rewriting, we are ready to rewrite again.

Figure 2a
shows the crystal structure of the (5-CHO-THF)-(Gly-PLP)-SHMT ternary complex of mouse SHMT1, which has high homology to human and rat SHMT (Fig. 2a). 21 Here, 5-CHO-THF plays a role as an analogue of THF in the intermediate state.
In the case of THF-dependent pathway (R = H; upper arrow), SHMT transfers one carbon to THF from Ser-PLP complex, to afford glycine ( Fig. 2b). 23 As shown in Fig. 2a, the vicinity of the serine recognition site is very limited. This limited space of the substrate binding site hampers the development of SHMT-responsive probes.

>>>Response to Comment.
We thank the reviewer for the constructive comment. In the revised manuscript, we have rewritten the sentence. We hope that our revisions have adequately addressed the reviewer's concerns.
We thank the reviewer for the constructive comment. Following the reviewer's comment, we have changed "crowded" to "limited". We thank the reviewer for the constructive comment. In the revised manuscript, we have rewritten the sentence. We hope that our revisions have adequately addressed the reviewer's concerns.
[ Revised manuscript,p.8, In other words, using THF-independent pathway, SHMT has substrate acceptance for β-substitution of serine.
ü Comment. Page 6, line 18: The sentences on this page require rewriting.

>>>Response to Comment.
We thank the reviewer for the constructive comment. In the revised manuscript, we have rewritten the sentences.
[ Revised manuscript,p.8, In other words, using THF-independent pathway, SHMT has substrate acceptance for β-substitution of serine.

>>>Response to Comment.
We thank the reviewer for the constructive comment. Following the reviewer's comment, we have changed. Following the reviewer's comment, we have changed.

>>>Response to Comment.
Following the reviewer's comment, we have changed.
ü Comment. Page 9, Figure 3 ; From a chemist point of view, the scheme should contain the temperatures, rxt times, and the deprotection conditions in detail, or should at least be mentioned in the caption of the Figure. >>>Response to Comment.
We thank the reviewer for the constructive comment. Following the reviewer's comment, we have revised the figure and caption. Following the reviewer's comment, we have changed.
We thank the reviewer for the constructive comment. Following the reviewer's comment, we have changed "naked eye" to "unaided human eye".
We thank the reviewer for the constructive comment. Following the reviewer's comment, we have changed as follows. We hope that our revisions have adequately addressed the reviewer's concerns.
[Revised manuscript, p.13, lines 17-19] Upon comparing the reaction rates toward SHMT1 among the DL-erythro form, the DL-threo form, and the L-erythro form, it emerged that the L-erythro enantiomer reacted faster.
We thank the reviewer for the constructive comment. Following the reviewer's comment, we have changed as follows.
[ Revised manuscript,p.13, I would like to thank the authors for the careful and thorough corrections of the manuscript and for taking the addressed recommendations into consideration to improve the quality of the manuscript. The manuscript "Design Strategy for Serine Hydroxymethyltransferase Probes Based on Retro-Aldol-Type Reaction" from the Prof. Sando et al. fulfills the publication requirements for Nature Comm. with minor revisions.
Point to Point Response >>>Author response. We thank the reviewer for the constructive comment to improve this manuscript. Following the reviewer's comment, we have revised the manuscript and added 25 references for covering related literatures appropriately. We hope that our revisions have adequately addressed the reviewer's concerns. >>>Referee response Thank you for adding additional literature to the manuscript to cover the field.
>>>Author response. We thank the reviewer for the constructive advice. Following the reviewer's comment, we have added new sentences about the advantage and precedents of 19F NMR/MRI, with appropriate references (Revised manuscript, p.10, lines 5-10). We hope that our revisions have adequately addressed the reviewer's concerns. >>>Referee response Thank you for the elaboration of 19F NMR techniques.
>>>Author response. Thank you for your valuable comment to improve this manuscript. Following the comments, we have added sentences to explain the importance of SHMT in the therapeutic area of Malaria and the previous efforts of anti-Malaria drug to the manuscript. [Revised manuscript,p.4, The development of SHMT inhibitors has been performed especially toward treatment of two types of diseases. The first one is antimalarial drug.8-15 Malaria is a life-threatening disease that spreads to people through infected anopheles mosquitoes. It has a tremendous impact globally, 216 million people are infected in 2016, and 445,000 people have died.13 In addition, the resistance of malaria parasites against existing antimalarial drugs has become a serious problem. Under such circumstances, researches and investigations for new inhibitors against malaria SHMT has been conducted. >>>Referee response I appreciate that malarial SHMT has been further introduced to the reader. But nevertheless, please rewrite this paragraph to underline the interesting and strong content with an adequate writing.
>>>Author response. We thank the reviewer for the constructive comment. Following the reviewer's comment, we have revised the manuscript. We hope that our revisions have adequately addressed the reviewer's concerns. If the reviewer needs more rewriting, we are ready to rewrite again. >>>Referee response Please rewrite the introduction. It is hard to read and not harmonized yet. No changes has been conducted compared to previous version.
>>>Author response. We agree with the reviewer's comment. The product aldehyde might react with biological relevant nucleophiles through forming Schiff's base. Following the reviewer's comment, first we have checked the stability of fluorescence product aldehyde (DMANA) against biological reactants, e.g. glutathione, L-cysteine, DL-homocysteine, H2O2, NaOCl, KO2, TBHP, and NOC7 ( Supplementary  Fig. S6). Under all conditions, no significant changes in fluorescence were observed. Furthermore, we have evaluated the toxicity of DMANA ( Supplementary Fig. S7). Significant toxicity was not observed from the cytotoxicity test. These results indicate that the significant side-effects and toxicity are not problematic under the concentration and time range of fluorescence experiments. We have added the discussion about the stability and toxicity of DMANA in the main text and Supplementary Information (p. 14, lines 1-8, Supplementary Fig. S6, Supplementary Fig. S7). However, under the concentration range of 19F NMR/MRI, 19F product aldehyde might show the side-effects and toxicity, because the required concentration is high for detection of 19F NMR/MRI signal. For practical in vivo MRI, it might be necessary to improve the sensitivity to lower the required concentration. These points have also been discussed in the main text (p. 24, lines 15-19).
Supplementary Figure S7 | Cytotoxicity evaluation of turned over fluorescence product DMANA. HeLa cells were plated in a 6-well plate at a density of 50,000 cells/mL in DMEM media. After incubation for 24 h, each samples were supplemented with either 1 μM or 5 μM DMANA, containing 1% DMSO. Equivalent samples were supplemented with 1% DMSO as a vehicle control. At 6 and 24 hours, cells were dissociated from wells by trypsin, a 10 μL sample was removed from each of the samples and mixed 1:1 with a 0.4% wt/volume trypan blue solution in PBS. Samples were incubated for 1 minute at room temperature before being loaded onto a hemocytometer where live and dead cells were counted. Each sample was made in triplicate for each time point. Error bars represent s.d., n = 3. >>>Referee response I appreciate the effort to answer my question about stability and toxicology. The question is very well addressed, which further improves the quality of the manuscript.
Human SHMT shares 91% sequence identity with mouse SHMT and 42% sequence identity with P. vivax SHMT (PvSHMT) (Chaiyen et al., FEBS J., 276, 4023-4036 (2009)). Because human and mouse have a high homology, mouse SHMT is considered to be appropriate for this modeling study. We have added this point about homology in the main text (P8, lines 5-8). >>>Referee response Thank you for further elucidation of the homology of the species. Nevertheless, a short comment about the homology of the active site would be appropriate.
>>>Author response. Basically, we think that it is difficult to conduct in vivo experiments by using the current probes and machine setup. However, regarding fluorescence probe 1, depending on the two-photon efficiency of the turned-over product, it might be possible to detect the product near the surface of the mouse/rat by using a two-photon excitation microscope. Regarding practical in vivo 19F MRI using probe 2, there is a problem of sensitivity, so it would be necessary to enhance the sensitivity, for example, by increasing the number of 19F atoms on molcular probes or by improving NMR/MRI instruments. In order to clarify this point, we have added this discussion about current limitation and future direction in an application for 19F MRI in main text. (P24, lines 15-19) >>>Referee response As this point was not only mentioned by me, it was very necessary to clarify the concerns. I appreciate the additional comment.
>>>Author response. We thank the reviewer for the comment. Following the reviewer's comment, we have added an explanation of manual modeling process as follows (caption in Supplementary Fig. S3). We hope that our revisions have adequately addressed the reviewer's concerns.

Point-by-point response to the comments of Reviewer #1
We wish to thank the reviewer for the expert comment and for providing constructive suggestions. We have addressed all the points raised by the reviewer through new text. therefore strongly related to cell proliferation (Fig. 1c). In fact, inhibitors targeting DHFR and TS, such as methotrexate and fluorouracil, respectively, have been used for a long time as anticancer agents. Among the three enzymes involved in one-carbon metabolism, to our knowledge, human SHMT (hSHMT) is the only enzyme for which an established chemotherapeutic agent has not yet been developed. Therefore, hSHMT has attracted attention as a potential target enzyme for inhibitor development.
[revMT_EnglishEditing.pdf, p6, Figure   We sincerely thank the reviewer for the positive comment for publication and constructive advice to improve our manuscript. We have conducted additional experiment to answer the question raised by the reviewer.
ü Comment. This reviewer thanks the authors for their response. I am almost satisfied. I would still like the authors to evaluate the stability of their probe(s) against CYPs. In particular the authors can use newly purchased or prepared rat liver microsomes for these experiments.
Once this key experiment has been completed I would be supportive of publication.

>>>Response to Comment.
We sincerely thank the reviewer for the expert comment. Following the reviewer's comment, we have newly purchased rat liver microsomes [revMT_EnglishEditing.pdf, p14, line 5-9] To evaluate the probe stability against cytochrome P450 (CYP450) enzymes, we incubated probe 1 and DMANA with NADPH-supplemented rat liver microsomes. Under our experimental conditions, majority of probe 1 and DMANA remained intact upon incubation with the microsomes (Supplementary Fig. S6).

Original comments from Reviewer #3
General response to the corrections I would like to thank the authors for the careful and thorough corrections of the manuscript and for taking the addressed recommendations into consideration to improve the quality of the manuscript. We agree with the reviewer's comment. The product aldehyde might react with biological relevant nucleophiles through forming Schiff's base. Following the reviewer's comment, first we have checked the stability of fluorescence product aldehyde (DMANA) against biological reactants, e.g. glutathione, L-cysteine, DL-homocysteine, H2O2, NaOCl, KO2, TBHP, and NOC7 ( Supplementary Fig. S6). Under all conditions, no significant changes in fluorescence were observed. Furthermore, we have evaluated the toxicity of DMANA ( Supplementary Fig. S7). Significant toxicity was not observed from the cytotoxicity test.
These results indicate that the significant side-effects and toxicity are not problematic under the concentration and time range of fluorescence experiments. We have added the discussion about the stability and toxicity of DMANA in the main text and Supplementary Information (p. 14, lines 1-8, Supplementary Fig. S6, Supplementary Fig. S7). However, under the concentration range of 19F NMR/MRI, 19F product aldehyde might show the 12 side-effects and toxicity, because the required concentration is high for detection of 19F NMR/MRI signal. For practical in vivo MRI, it might be necessary to improve the sensitivity to lower the required concentration. These points have also been discussed in the main text (p. 24, lines 15-19).
Because human and mouse have a high homology, mouse SHMT is considered to be 13 appropriate for this modeling study. We have added this point about homology in the main text (P8, lines 5-8).
>>>Referee response Thank you for further elucidation of the homology of the species. Nevertheless, a short comment about the homology of the active site would be appropriate.
Basically, we think that it is difficult to conduct in vivo experiments by using the current probes and machine setup. However, regarding fluorescence probe 1, depending on the two-photon efficiency of the turned-over product, it might be possible to detect the product near the surface of the mouse/rat by using a two-photon excitation microscope. Regarding practical in vivo 19F MRI using probe 2, there is a problem of sensitivity, so it would be necessary to enhance the sensitivity, for example, by increasing the number of 19F atoms on molcular probes or by improving NMR/MRI instruments. In order to clarify this point, We thank the reviewer for the comment. Following the reviewer's comment, we have added an explanation of manual modeling process as follows (caption in Supplementary   Fig. S3). We hope that our revisions have adequately addressed the reviewer's concerns.
[Revised Supplementary  Then, the co-crystal structure of PLP-Gly-5-CHO-THF-mSHMT1 (PDB ID: 1EJI) and probe 1 are displayed using the PyMOL software, and 5-CHO-THF is deleted. Finally, L-erythro probe 1 was manually placed to the active site of mSHMT1 with PLP-Gly as amino groups and carboxylic groups of Gly and probe 1 were each superimposed. This supports the hypothesis that L-erythro form can be accommodated in the substrate pocket. Color code of stick model: We thank the reviewer for this advice. According to reviewer's comment, we have added the corresponding citations.