Cyclic di-GMP Modulates a Metabolic Flux for Carbon Utilization in Salmonella enterica Serovar Typhimurium

ABSTRACT Salmonella enterica serovar Typhimurium is an enteric pathogen spreading via the fecal-oral route. Transmission across humans, animals, and environmental reservoirs has forced this pathogen to rapidly respond to changing environments and adapt to new environmental conditions. Cyclic di-GMP (c-di-GMP) is a second messenger that controls the transition between planktonic and sessile lifestyles, in response to environmental cues. Our study reveals the potential of c-di-GMP to alter the carbon metabolic pathways in S. Typhimurium. Cyclic di-GMP overproduction decreased the transcription of genes that encode components of three phosphoenolpyruvate (PEP):carbohydrate phosphotransferase systems (PTSs) allocated for the uptake of glucose (PTSGlc), mannose (PTSMan), and fructose (PTSFru). PTS gene downregulation by c-di-GMP was alleviated in the absence of the three regulators, SgrS, Mlc, and Cra, suggesting their intermediary roles between c-di-GMP and PTS regulation. Moreover, Cra was found to bind to the promoters of ptsG, manX, and fruB. In contrast, c-di-GMP increased the transcription of genes important for gluconeogenesis. However, this effect of c-di-GMP in gluconeogenesis disappeared in the absence of Cra, indicating that Cra is a pivotal regulator that coordinates the carbon flux between PTS-mediated sugar uptake and gluconeogenesis, in response to cellular c-di-GMP concentrations. Since gluconeogenesis supplies precursor sugars required for extracellular polysaccharide production, Salmonella may exploit c-di-GMP as a dual-purpose signal that rewires carbon flux from glycolysis to gluconeogenesis and promotes biofilm formation using the end products of gluconeogenesis. This study sheds light on a new role for c-di-GMP in modulating carbon flux, to coordinate bacterial behavior in response to hostile environments. IMPORTANCE Cyclic di-GMP is a central signaling molecule that determines the transition between motile and nonmotile lifestyles in many bacteria. It stimulates biofilm formation at high concentrations but leads to biofilm dispersal and planktonic status at low concentrations. This study provides new insights into the role of c-di-GMP in programming carbon metabolic pathways. An increase in c-di-GMP downregulated the expression of PTS genes important for sugar uptake, while simultaneously upregulating the transcription of genes important for bacterial gluconeogenesis. The directly opposing effects of c-di-GMP on sugar metabolism were mediated by Cra (catabolite repressor/activator), a dual transcriptional regulator that modulates the direction of carbon flow. Salmonella may potentially harness c-di-GMP to promote its survival and fitness in hostile environments via the coordination of carbon metabolic pathways and the induction of biofilm formation.

most direct effect. However, how these regulators mediate this response is not determined. Showing an effect of cdG on central metabolism is interesting, and I find their model that cdG helps to switch S. Typhimurium to a metabolic state more favorable for extracellular polysaccharide production something that would be of interest to the field. However, the manuscript does not provide any evidence that the measured gene changes have any functional consequences, which would greatly strengthen its conclusions. Furthermore, the authors need to explore if mutations in the central regulators impact basal cdG or the activity of the diguanylate cyclase, AdrA, which they use to alter cdG levels.
1. Line 74-make this a new paragraph 2. Line 102, 294-change "cholera" to "cholerae" 3. In some species, overproduction of cdG can slow growth. The authors should perform a growth curve of their WT and AdrA overexpressing strain to test if this is a factor in their experiments. 4. Line 159-It would be prudent to mention that SgrS also encodes the SgrT peptide, which inhibits PstG transport of glucose. 5. Fig. 4-Lines 175-177. It is not possible to determine if the reduction in ptsB and manXYZ is due to cdG or SrgS overexpression because this experiment is missing one key strain, the pSrgS + pBbS2k (vector control) to show what effect overproduction of SgrS has on expression of these genes in the absence of cdG overproduction from AdrA. Without this strain, it is not possible to appreciate how much of the reduced expression in the SgrS/AdrA strain is due to cdG or SgrS (NOTE: I realize they explored SgrS alone in Fig. S1, but this did not have the other vector control and was not done side by side with AdrA expression). 6. Although I do not find this likely, it is possible that mutations in the metabolic regulators SgrS, Mlc, and Cra could influence the diguanylate cyclase activity of AdrA. The reason this is a possibility is that most DGCs have N-terminal sensory domains that respond to either environmental or host derived cues, and AdrA could be sensing metabolic intermediates in the cell. Such regulation of AdrA activity could explain many of their results. The authors should therefore quantify cdG upon overexpression of AdrA in these transcription factor mutants to test this possibility. In addition, mutations to these key regulators could impact the basal concentrations of c-di-GMP even in the absence of AdrA which may impact their expression studies. 7. Lines 193-195-This statement is speculation without evidence. There are many possible mechanisms. It should be removed. 8. Fig. 6C-This EMSA experiment with Cra should be tested with the addition of cdG to determine if cdG directly impacts binding of Cra to its target promoters. Given that the EMSA has already been developed, that is a simple experiment to try which greatly informs the model in Fig. 8. 9. One weakness of the paper is that all the conclusions are based on gene expression results. It would strengthen the paper to demonstrate a functional consequence to the demonstrated cdG gene regulation. Are there specific growth conditions or substrates that can be used to demonstrate that cdG is decreasing PTS activity while increasing gluconeogenesis? 10. The authors use standard T-tests to analyze statistical differences in large datasets, but they should really be using ANOVAs.
Reviewer #2 (Comments for the Author): This manuscript shows a new role for c-di-GMP, which is the modulation of carbon flux. The authors analyzed the effect of the overproduction of c-di-GMP in S. Typhimurium, by overexpressing the diguanylate cyclase AdrA, and found that it decreased the expression of genes encoding components of phosphoenolpyruvate (PEP):carbohydrate phosphotransferase systems (PTSs) for the uptake of glucose, mannose and fructose. Furthermore, c-di-GMP was shown to act through the regulators SrgS, M1c and Cra to control the three PTSs, being Cra the major factor for this c-di-GMP-mediated regulation. The authors demonstrate that CrA binds to genes of the three PTSs. Additionally, they found that c-di-GMP-Cra also regulates the expression of genes for gluconeogenesis but positively. Thus, this manuscript supports that c-di-GMP rewires carbon flux from glycolysis to gluconeogenesis to favor biofilm formation, which requires the end products of gluconeogenesis. The manuscript is well written, and conclusions are completely supported by data.

Comments.
What about deletion of adrA? Does it affect the expression of genes for carbon flow?
It would be good to mention in Discussion that results from EMSAs (number or DNA/protein complexes) agree with the number of predicted Cra-binding sites on fruB, pstG and manX.
Do the genes for gluconeogenesis also harbor the CsrA-binding consensus sequence?
Minor comments.
Line 218. pCra instead pFruR Line 226. The bands from EMSAs were seen by staining with ethidium bromide, not by using an anti-His antibody.

Preparing Revision Guidelines
To submit your modified manuscript, log onto the eJP submission site at https://spectrum.msubmit.net/cgi-bin/main.plex. Go to Author Tasks and click the appropriate manuscript title to begin the revision process. The information that you entered when you first submitted the paper will be displayed. Please update the information as necessary. Here are a few examples of required updates that authors must address: • Point-by-point responses to the issues raised by the reviewers in a file named "Response to Reviewers," NOT IN YOUR COVER LETTER. • Upload a compare copy of the manuscript (without figures) as a "Marked-Up Manuscript" file. • Each figure must be uploaded as a separate file, and any multipanel figures must be assembled into one file. For complete guidelines on revision requirements, please see the journal Submission and Review Process requirements at https://journals.asm.org/journal/Spectrum/submission-review-process. Submissions of a paper that does not conform to Microbiology Spectrum guidelines will delay acceptance of your manuscript. " Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me. If you do not wish to modify the manuscript and prefer to submit it to another journal, please notify me of your decision immediately so that the manuscript may be formally withdrawn from consideration by Microbiology Spectrum.
If your manuscript is accepted for publication, you will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail. Arrangements for payment must be made before your article is published. For a complete list of Publication Fees, including supplemental material costs, please visit our website.
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees. Need to upgrade your membership level? Please contact Customer Service at Service@asmusa.org.
Thank you for submitting your paper to Microbiology Spectrum.

Manuscript Number: Spectrum03685-22
Dear Editor, We appreciate the faithful and valuable review. Our detailed responses to the review comments are provided below and the changes are highlighted in blue in the revised manuscript. We have addressed all of the concerns of the reviewers and hope that the revised manuscript is now suitable for considering publication in Microbiology Spectrum.

Reviewer #1 (Comments for the Author):
This manuscript by Baek and Yoon explores the role of cyclic di-GMP (cdG) in the central metabolism of S. Typhimurium. The authors find using gene expression studies that cdG downregulates PTS components while upregulating gluconeogenesis pathways genes. They suggest the regulators SgrS (sRNA), Mlc, and Cra are necessary for this response, with Cra having the most direct effect. However, how these regulators mediate this response is not determined. Showing an effect of cdG on central metabolism is interesting, and I find their model that cdG helps to switch S. Typhimurium to a metabolic state more favorable for extracellular polysaccharide production something that would be of interest to the field. However, the manuscript does not provide any evidence that the measured gene changes have any functional consequences, which would greatly strengthen its conclusions. Furthermore, the authors need to explore if mutations in the central regulators impact basal cdG or the activity of the diguanylate cyclase, AdrA, which they use to alter cdG levels.
Response: We value the reviewer's perceptive feedback. We acknowledge that the experimental evidence based just on gene regulation is insufficient to predict the functional outcome. Therefore, in the revised manuscript, we further explored whether the downregulation of three PTSs by c-di-GMP might slow the pace of extracellular glucose import. We also looked into the possibility that the absence of three regulatory mediators-SgrS, Mlc, and Cra-could influence the production of c-di-GMP and its cellular activity. The detailed results are discussed in the relevant comments below.

Line 74-make this a new paragraph
Response: According to the reviewer's suggestion, the sentences describing the roles of c-di-GMP, CsgD and σ S in biofilm formation and stress adaptation have been separated, making a new paragraph (line 75 -85 in the revised).
2. Line 102, 294-change "cholera" to "cholerae" Response: We appreciate the considerate correction. Bacterial name has been corrected according to the binomial nomenclature (lines 103 and 301).
3. In some species, overproduction of cdG can slow growth. The authors should perform a growth curve of their WT and AdrA overexpressing strain to test if this is a factor in their experiments.
Response: We were also aware of the growth deficiency triggered by the overexpression of c-di-GMP, as the reviewer mentioned. To minimize the growth inhibitory effect of cdi-GMP, we optimized the induction condition by employing multiple strategies, including a tight regulation promoter, a low copy number plasmid, and low concentrations of inducer. Comparison of bacterial growth curves is provided in Fig. S1 in the revised manuscript. Despite several attempts, Salmonella producing c-di-GMP showed attenuated growth, reaching an OD 600 of 2.3 at its maximum, which was around 30% lower than the control with an empty plasmid. Instead, we used bacterial cells that were in the same stage of growth. The PTS genes were tested at the log phase, or approximately 3 h in both strains, while the gluconeogenesis-associated genes were examined at the stationary phase, which was roughly 8 h in both cases. The poor growth caused by c-di-GMP overexpression, which downregulates the expression of three PTSs, may be an inevitable result given that the three PTSs are primarily in charge of acquiring carbon sources.
4. Line 159-It would be prudent to mention that SgrS also encodes the SgrT peptide, which inhibits PstG transport of glucose.
Response: We appreciate the thoughtful suggestion. The role of SgrS encoding SgrT has been added in the revised text with a relevant reference (line 163 -165). 5. Fig. 4-Lines 175-177. It is not possible to determine if the reduction in ptsB and manXYZ is due to cdG or SrgS overexpression because this experiment is missing one key strain, the pSrgS + pBbS2k (vector control) to show what effect overproduction of SgrS has on expression of these genes in the absence of cdG overproduction from AdrA. Without this strain, it is not possible to appreciate how much of the reduced expression in the SgrS/AdrA strain is due to cdG or SgrS (NOTE: I realize they explored SgrS alone in Fig. S1, but this did not have the other vector control and was not done side by side with AdrA expression).
Response: We concur with the reviewer that the previous version of Fig. 4  controls the production of c-di-GMP, the levels of c-di-GMP were compared between a wild-type and three mutant strains lacking SgrS, Mlc, or Cra. To ensure the production of AdrA similar amongst these strains, pAdrA was used to transform all strains. Under the tested condition (LB broth, 37℃, 48 h), the absence of either regulator had no effect on the cellular concentrations of c-di-GMP (Fig. S5A in the revised manuscript), indicating that the absence of these regulators did not influence the activity of the diguanylate cyclase, AdrA. The primary role of c-di-GMP is to stimulate biofilm formation by upregulating the production of curli and cellulose, both of which may be easily recognized by rdar and rugose morphotypes. When the morphotypes were compared between bacterial strains, the mutants lacking SgrS, Mlc, or Cra promoted the development of the rdar and rugose morphotypes, which was unexpected because the c-di-GMP levels were comparable between the wild-type and these mutant strains. This result suggests that these regulators may obstruct c-di-GMP action or independent of c-di-GMP, they may also regulate the production of curli and cellulose. This result has been added as Fig. S5B in the revision. In addition, we also examined the possibility that these three regulators could impact the basal concentrations of c-di-GMP even in the absence of AdrA. Unfortunately, the basal c-di-GMP levels were too low to compare between the wild-type and three mutant strains. Instead, we compared the rdar morphotypes between bacterial strains without AdrA overexpression (see bacterial strains harboring pBbS2k-RFP, an empty plasmid in Response: The statement has been removed according to the reviewer's suggestion. 8. Fig. 6C-This EMSA experiment with Cra should be tested with the addition of cdG to determine if cdG directly impacts binding of Cra to its target promoters. Given that the EMSA has already been developed, that is a simple experiment to try which greatly informs the model in Fig. 8. Response: According to the suggestion, the binding affinity of Cra to the promoters of fruBKA, ptsG, and manXYZ were examined in the presence of c-di-GMP. Even with the inhibitory FBP, the addition of c-di-GMP tended to facilitate Cra to bind to P fruB and P manX . However, c-di-GMP failed to restrain FBP which attenuated the binding affinity of Cra toward P ptsG . The experimental condition might be optimized still more to clear the interaction between c-di-GMP and Cra. But, for now, we speculate that the binding affinities of Cra differ between promoters and that, as a result, the effects of c-di-GMP on the binding affinity of Cra vary between promoters. The result has been added as Fig. S4 with the interpretation (line 322 -327). 9. One weakness of the paper is that all the conclusions are based on gene expression results.
It would strengthen the paper to demonstrate a functional consequence to the demonstrated cdG gene regulation. Are there specific growth conditions or substrates that can be used to demonstrate that cdG is decreasing PTS activity while increasing gluconeogenesis?
Response: We totally agree with the reviewer's suggestion. In order to verify that c-di-GMP downregulates the expression of three PTSs, PTS Glc , PTS Man , and PTS Fru , and subsequently slows sugar transport by these PTSs, the amount of extracellular glucose in the culture broth was measured. A wild-type Salmonella consumed 50% glucose at 3 h, when the genes of PTS Glc and PTS Man were highly activated.
However, Salmonella producing c-di-GMP consumed only 24% glucose for the same time, which was presumably due to the downregulation of PTS Glc and PTS Man . The result has been added in the revised manuscript (line 137 -139) as 10. The authors use standard T-tests to analyze statistical differences in large datasets, but they should really be using ANOVAs.
Response: According to the suggestion, we used one-way analysis of variance (ANOVA) with Tukey's post hoc test in the revised figures, including Fig. 4, Fig. 5, Fig. 6,   Fig. 7, and Fig. S2. The method describing statistical analysis has been corrected accordingly (line 461 -464).

Reviewer #2 (Comments for the Author):
This manuscript shows a new role for c-di-GMP, which is the modulation of carbon flux. The authors analyzed the effect of the overproduction of c-di-GMP in S. Typhimurium, by overexpressing the diguanylate cyclase AdrA, and found that it decreased the expression of genes encoding components of phosphoenolpyruvate (PEP):carbohydrate phosphotransferase systems (PTSs) for the uptake of glucose, mannose and fructose. Furthermore, c-di-GMP was shown to act through the regulators SrgS, M1c and Cra to control the three PTSs, being Cra the major factor for this c-di-GMP-mediated regulation. The authors demonstrate that CrA binds to genes of the three PTSs. Additionally, they found that c-di-GMP-Cra also regulates the expression of genes for gluconeogenesis but positively. Thus, this manuscript supports that c-di-GMP rewires carbon flux from glycolysis to gluconeogenesis to favor biofilm formation, which requires the end products of gluconeogenesis.
The manuscript is well written, and conclusions are completely supported by data.
Response: We appreciate the reviewer's favorable evaluation. The detailed responses to each comment are listed below. Comments.
What about deletion of adrA? Does it affect the expression of genes for carbon flow?
Response: We also considered to delete adrA to investigate the roles of c-di-GMP. However, the deletion of adrA did not influence the transcription of genes responsible for the production of curli and cellulose, which are primarily regulated by c-di-GMP.
We speculated that other diguanylate cyclases except AdrA might have supplemented the absence of adrA. Due to multiple alternatives to AdrA, we decided to overexpress c-di-GMP, instead, to explore its role in Salmonella physiology. Although we have not examined the transcription of genes associated with PTSs and gluconeogenesis, it's unlikely that the deletion of adrA will change their expression.
It would be good to mention in Discussion that results from EMSAs (number or DNA/protein complexes) agree with the number of predicted Cra-binding sites on fruB, pstG and manX.
Response: As suggested by the reviewer, we asserted in Discussion that "We found that Cra could directly bind to the promoters of fruBKA, ptsG, and manXYZ containing the consensus Cra boxes and repress their transcription", indicating that the EMAS results were consistent with the predicted Cra-binding sites (line 310 -312).
Do the genes for gluconeogenesis also harbor the CsrA-binding consensus sequence?
Response: The Csr system composed of CsrA (regulator), CsrB/CsrC (noncoding RNAs), and CsrD (regulator for CsrB/CsrD) is known to positively regulate glycolysis while inhibiting gluconeogenesis. The consensus binding motif of CsrA (ARGGAN, where GGA is mandatory) has been predicted in multiple genes associated with gluconeogenesis as well as glycolysis (Liu B. et al. 2021. Microorganisms 9(11): 2383). However, we don't think that the reviewer questioned the CsrA-binding sites in gluconeogenesis-relevant genes, because we've not tested and mentioned Csr system in the manuscript. If the reviewer intended Cra instead of CsrA, lots of studies have suggested the binding sites of Cra in the genes of gluconeogenesis and glycolysis (Kim D. et al. 2018. Nucleic Acids Res. 46(6): 2901-2917).