Glycine, serine and threonine metabolism confounds efficacy of complement-mediated killing

Serum resistance is a poorly understood but common trait of some difficult-to-treat pathogenic strains of bacteria. Here, we report that glycine, serine and threonine catabolic pathway is down-regulated in serum-resistant Escherichia coli, whereas exogenous glycine reverts the serum resistance and effectively potentiates serum to eliminate clinically-relevant bacterial pathogens in vitro and in vivo. We find that exogenous glycine increases the formation of membrane attack complex on bacterial membrane through two previously unrecognized regulations: 1) glycine negatively and positively regulates metabolic flux to purine biosynthesis and Krebs cycle, respectively. 2) α-Ketoglutarate inhibits adenosine triphosphate synthase, which in together promote the formation of cAMP/CRP regulon to increase the expression of complement-binding proteins HtrE, NfrA, and YhcD. The results could lead to effective strategies for managing the infection with serum-resistant bacteria, an especially valuable approach for treating individuals with weak acquired immunity but a normal complement system.

absence of the cell surface proteins that bind complement in E. coli? What are some alternative hypotheses as to why glycine is potentiating killing, but other metabolites that alter membrane potential do not? 4. Lines 315-316: I believe this statement is worded wrong. Doesn't a decrease in glycine lead to an increase in killing? 5. Figure 3, panel f: Please define the green dotted line 6. Line 336: Suggest replacing "genes" with "the presence of genes" or "gene expression" 7. Figure  11. Line 576: change "explore" to "explored" 12. Lines 638-642 seem out of place in this data section. consider moving 13. Figure 8, panel B: MRSA is spelled wrong 14. Discussion: Please discuss how you think glycine is potentiating serum killing of the non-E. coli strains. I understand that this may be speculative.
15. Line 1076: Stating that bacterial strains "were from collections in our laboratory" does not provide enough information to the reader to enable reproducibility. Did these strains come from published work? Were they obtained from patients? Please add a strain table and appropriate references.

Jim Cassat
Reviewer #2 (Remarks to the Author): In this paper by Cheng et. al, the influence and mechanisms of glycine, serine and threonine on serum-mediated killing of bacteria is investigated. The paper is thorough and contains novel research of great interest that could have real world consequences for treatment of bacteremia. The authors were rigorous in their experimental design, with mostly good controls used throughout. The extension into clinically relevant bacterial strains and mouse model was also excellent. The paper is overall convincing however have a few major concerns as outlined below: Major comments: • As stated in the introduction, LPS is one of the main drivers of serum resistance-primarily due to the production of O-antigen that is thought to block access of complement to the surface/ outer membrane proteins of the bacteria. K-12 has a mutation in waaL which renders it O-antigen negative and thus its role in any glycine mediated sensitivity is unexplored. It is likely that many of the other strains used in this study are O-antigen positive however this is not mentioned. Thus, 1) As the second most used strain-Is the serum-resistant Y17 O-antigen positive?
2) Is it known if any of the other E. coli, Pseudomonas etc. used were O-antigen positive?
3) Was the effect of glycine on O-antigen expression investigated at all? 4) As we know from the serum resistome paper that O-antigen expression in EC958 is critical could you test the effect of glycine on killing of this strain?
• Gram negative bacteria are sensitive to complement-mediated killing as the MAC complex can insert into the outer membrane. Gram positive bacteria however are resistant to complement mediated killing as the peptidoglycan of their membrane is too thick for MAC insertion. Thus, 5) In figure 8A All Staph aureus strain (Gram positive) are killing by the addition of glycine to serum. By what mechanism is this killing happening? Was a HI-serum (or C3 depleted) + glycine control used in these assays? In factwas this control done for any of the strains in 8A?
Minor comments: 1) In the introduction there is no mention of the role of capsule in serum resistance (although reports differ). Also no mention of antibody mediated serum resistance found in Gram negatives such as Salmonella In the manuscript titled "Glycine, serine, and threonine metabolism confounds efficacy of complement-mediated killing", Cheng et al explore the metabolic mechanisms underlying bacterial serum sensitivity or resistance. More specifically, the authors determine that exposure of E. coli to serum (and by extension, complement) leads to global metabolic changes, including significant down regulation of glycine, serine, and threonine metabolism. In a series of comprehensive and wellcontrolled experiments, the authors go on to demonstrate that mechanistically, this alteration of amino acid catabolism is linked to changes in TCA cycle flux, ATP production, membrane potential, cAMP/CRP regulation, and finally, cell surface protein expression. It is the alterations in cell surface protein expression that at least partially underlie the altered susceptibility to serum killing, as the authors clearly demonstrate decreased binding of complement factors. The authors also show that exogenous glycine, serine, and threonine can restore serum-sensitivity to serumresistant E. coli, and that glycine treatment sensitizes many classes of human bacterial pathogens to serum. Finally, the authors show that mice treated with glycine have significantly lower bacterial burdens during disseminated infection with 4 2 different pathogens. Taken together, these experiments outline a new mechanistic understanding of bacterial serum sensitivity and suggest a potential avenue for modulating immune killing of common bacterial pathogens. The work is broad in scope, comprehensively executed, and will be of significant interest to the field of investigators studying innate immunology, bacterial pathogenesis, and host-pathogen interactions. The conclusions are soundly supported by the data, and the authors very comprehensively address potential confounding variables. Statistical analyses are rigorously applied throughout the manuscript. I do not believe that additional experiments are necessary to fortify the main conclusions. However, some items should be addressed in the text prior to publication.

Response. Thank you for your comment!
Minor comments: 1. The are grammatical errors throughout. The manuscript may benefit from a professional editor in this regard Response, We appreciate this comment! The manuscript has been carefully checked and revised by professional editor.
2. Lines 184-186: This statement is likely worded too strongly. Treatment with ampicillin and gentamicin did not potentiate glycine killing, but this certainly doesn't completely exclude changes in membrane integrity. To make this statement conclusively would require additional compounds that more directly test membrane 3 integrity, or perhaps EM.
Response, We appreciate this comment! According to Reviewer's comment, it has been revised in lines 192 -193 following as "This result indicates that glycine-mediated serum killing may not be related to impaired bacterial membrane integrity.".
3. Lines 284 and 285: The authors note that alanine, fructose, glucose, and glutamine stimulate the TCA cycle and increase membrane potential, but do not decrease survival in the presence of serum. This (and other comments below) point to a multi-factorial mechanism that may include more than just changes in membrane potential or altered surface protein expression. The authors should elaborate on this in the discussion. For example, why is MRSA killing potentiated by glycine in the absence of the cell surface proteins that bind complement in E. coli? What are some alternative hypotheses as to why glycine is potentiating killing, but other metabolites that alter membrane potential do not?
Response, According to Reviewer's comment, it has been revised in lines 727 -740 as following "Notably, other metabolites like alanine, fructose, glucose, and glutamine promote the membrane potential, but they fail to reverse serum resistance. This discrepancy is attributed to the dual roles of glycine that not only promotes membrane potential but also up-regulates the expression of complement-binding outer membrane proteins. The proposed mechanism can be applicable in Gram-negative bacteria with outer membrane proteins in addition to E. coli, but why this effect is also 4 observed in Gram-positive bacteria requires further investigation as they do not possess the outer membrane proteins for MAC insertion. The increased deposition of complement components on MRSA surface suggests other mechanisms may exist.
Nevertheless, that the triple mutations of htrE, nfrA and yhcD did not completely abrogate glycine-mediated serum killing and loss of waal affected the killing potential in several mutants indicates that serum resistance is a multi-factorial event that may include more than membrane potential and outer membrane proteins. 6. Line 336: Suggest replacing "genes" with "the presence of genes" or "gene expression" Response, We appreciate this comment! According Reviewer's comment, it has been 5 revised as "the presence of genes". 7. Figure 4, panel D: Ratio is misspelled as ration Response, We appreciate this comment! It has been corrected.
8. Figure 5: Please define ATPgammaS in the legend Response, We appreciate this comment! According Reviewer's comment, it has been defined in lines 440 as following "ATPγS, a nonhydrolyzable ATP analog.". 9. Figure  Response, We appreciate this comment! According Reviewer's comment, we have discussed this in lines 736 -740 as following "Nevertheless, that the triple mutations of htrE, nfrA and yhcD did not completely abrogate glycine-mediated serum killing and loss of waal affected the killing potential in several mutants indicates that serum resistance is a multi-factorial event that may include more than membrane potential and outer membrane proteins.". 6 11. Line 576: change "explore" to "explored" Response, We appreciate this comment! It has been revised.
12. Lines 638-642 seem out of place in this data section. consider moving Response, We appreciate this comment! They have been removed.

Figure 8, panel B: MRSA is spelled wrong
Response, We appreciate this comment! It has been corrected.
14. Discussion: Please discuss how you think glycine is potentiating serum killing of the non-E. coli strains. I understand that this may be speculative.
Response, We appreciate this comment! We have discussed this issue in the revised manuscript in lines 729 -733 as following "This discrepancy is attributed to the dual roles of glycine that not only promotes membrane potential but also up-regulates the expression of complement-binding outer membrane proteins. The proposed mechanism can be applicable in Gram-negative bacteria with outer membrane proteins in addition to E. coli,".
15. Line 1076: Stating that bacterial strains "were from collections in our laboratory" does not provide enough information to the reader to enable reproducibility. Did these strains come from published work? Were they obtained from patients? Please add a strain

Vibro. alginolyticus
The collections of our laboratory 2

Vibro. parahaemolyticus
The collections of our laboratory E.coli O157:H7 The collections of our laboratory 3

Pseudomonas aeruginosa
The collections of our laboratory, which were isolated from patients pneumonia strains (Supplementary Fig. 14). E. coli serotype O157:H7, with adhesive fimbriae and a cell wall that consists of an outer membrane containing LPS, was sensitive to the glycine-triggering killing (Fig. 6p) while exogenous glycine did not affect waal transcription (Fig. 6q). When waal was deleted, the glycine-mediated killing was either unaffected or reduced by 4.81-24.1% (Fig. 6r).
These results indicate that LPS is not the key target in the glycine-triggering serum killing.".   coli K12, Y5, Y10 and Y17 in the presence or absence of the indicated volume of serum or HI serum plus 100 mM glycine.
All critiques have been satisfactorily addressed. Congratulations on this exciting study.

Jim Cassat
Reviewer #2 (Remarks to the Author): In this re-submission by Cheng et al, they have responded to all the comments I raised in my initial review in a satisfactory manner. A brief summary below: -The addition of the O157:H7 result, as well as the O-antigen characterization are an excellent addition to the paper and strengthen the findings -The C3 and C5 results on MRSA strains are interesting but agree that further study on this is outside scope of this paper.
-The HI serum + glycine controls has also improved the veracity of the findings.
-All minor comments were responded to as requested 1

REVIEWERS' COMMENTS:
Reviewer #1 (Remarks to the Author): All critiques have been satisfactorily addressed. Congratulations on this exciting study.
Response, We appreciate it. Thanks.
Reviewer #2 (Remarks to the Author): In this re-submission by Cheng et al, they have responded to all the comments I raised in my initial review in a satisfactory manner. A brief summary below: -The addition of the O157:H7 result, as well as the O-antigen characterization are an excellent addition to the paper and strengthen the findings -The C3 and C5 results on MRSA strains are interesting but agree that further study on this is outside scope of this paper.
-The HI serum + glycine controls has also improved the veracity of the findings.
-All minor comments were responded to as requested