Fine-tuning of outer membrane–peptidoglycan tethering by the redox-active lipoprotein LppB from Salmonella enterica

Reviewer #1 (Public review):

Summary:

Pierre Despas et al. studied the role of Salmonella typhimurium LppB in outer membrane tethering. Using E. coli {delta}lpp mutant the authors showed that Salmonella LppB is covalently attached to PG through K58 and that these crosslinks are formed by the L,D-transpeptidase LdtB, primarily. Additionally, authors demonstrate that LppB forms homodimers via a disulfide bond through C57, but when Lpp is present it can also form heterotrimers with it. Thus, suggesting a regulatory role in Lpp-PG crosslinking.

Strengths:

In my view, this is a nice piece of work that expands our understanding of the role of lpp homologs. The experiments were well-designed and executed, the manuscript is well-written and the figures are well-presented.

Weaknesses:

I have some suggestions to give a clearer message, because I think a few images don't reflect much of what the authors wrote.

It'd be helpful for readers to see the phylogenetic tree of the rest of the organisms that harbor LppB homologs and Lpp.

Increased expression of LppB under low pH is subtle. This result would benefit from quantifying the blots (Fig. S1) and performing statistical analysis.

Similarly, the SDS-EDTA sensitivity result (Fig. S2) is not convincing; the image doesn't seem to show isolated colonies at low pH (Fig. S2B). Please measure CFU/mL and report endpoint growth graphs instead. Statistical analysis should also be presented.

The reduction to PG crosslinking of the C57R mutant is unclear (Fig 4B lane 22). The authors state: "suggesting that additional features of the LppB C-terminal region underlie its reduced efficiency." Does this mean additional amino acids play a role? Did the authors try to substitute Cys with other amino acid residues like Ala or Ser and quantify protein levels to find a mutant with similar expression levels? Do these have less crosslinking too?

Reviewer #2 (Public review):

Summary:

The manuscript by Pierre Despas and co-workers, reports the biochemical characterization of LppB a peculiar Lpp (Braun's lipoprotein) homolog found in Salmonella enterica. S. enterica encodes two Lpp homologs LppA and LppB: while LppA and Lpp function similarly, the role of LppB is less clear. LppB shares with Lpp the C-terminal Lys needed for covalent attachment to peptidoglycan (PG) but diverges in residues that precede the terminal Lys featuring a Cys residue at the penultimate position. By using E. coli as a surrogate model, the authors show that LppB can be covalently linked to PG via the terminal Lys residues and that the penultimate Cys residue can be used to form homodimer species when expressed alone and heterotrimeric complexes when co-expressed with Lpp. Interestingly, LppB expressed in E. coli seems to be stabilized at acidic pH a condition Salmonella encounters in macrophage phagosomes. Finally, based on decreased intensity of LppB-PG crosslinked bands as LppB expression increases the authors suggest that LppB is able to negatively modulate the outer membrane-peptidoglycan connectivity.

Strengths:

The manuscript is interesting, describes a novel strategy employed by bacteria to fine tuning outer membrane-PG attachment and provides new insights into how envelope remodeling processes can contribute to bacterial fitness and pathogenicity.

Weaknesses:

The analysis and quantification of muropeptides formed in E. coli strains overexpressing LppB would strengthen the main conclusion of the manuscript.

Reviewer #3 (Public review):

Summary:

The manuscript is interesting, and it is clearly written. While the experiments are well executed, a general flaw is that the LppA/B analyses are done in the E. coli K12 host as surrogate for Salmonella enterica. For the mechanistic and molecular analyses of LppB a surrogate host is certainly adequate, yet it limits extrapolation of the physiological implications of LppB in the natural context.

Strengths:

The work convincingly demonstrates that LppB forms disulfide-based dimers and that it is crosslinked to PG via LdtB in E. coli. Moreover, dimerisation is required for LppB abundance in E. coli and LppB can inhibit crosslinking of Lpp/A to PG in E. coli.

Weaknesses:

Regarding the key conclusion of the work: while it is shown that LppB is oxidized in E. coli, whether envelope integrity (or OMV production) changes arise from switches in oxidation of the LppB cysteines remains to be shown, for E. coli let alone in the native host Salmonella. Does expression of LppB influence Lpp/A activity or OM tethering in E. coli? Since the inhibition of the Lpp/A linking to PG is not affected by the oxidation state of LppB, the abstract/title implies redox-control of envelope integrity which is a bit misleading and an overstatement. Both are features of LppB: i.e. it dimerizes through disulfide bond formation and it reduces PG binding of Lpp/A through trimerisation. However, no link between the two is shown.

Author response:

Public Reviews:

Reviewer #1 (Public review):

Summary:

Pierre Despas et al. studied the role of Salmonella typhimurium LppB in outer membrane tethering. Using E. coli ∆lpp mutant the authors showed that Salmonella LppB is covalently attached to PG throug K58 and that these crosslinks are formed by the L,Dtranspeptidase LdtB, primarily. Additionally, authors demonstrate that LppB forms homodimers via a disulfide bond through C57, but when Lpp is present it can also form heterotrimers with it. Thus, suggesting a regulatory role in Lpp-PG crosslinking.

Strengths:

In my view, this is a nice piece of work that expands our understanding of the role of lpp homologs. The experiments were well-designed and executed, the manuscript is wellwritten and the figures are well-presented.

Weaknesses:

I have some suggestions to give a clearer message, because I think a few images don't reflect much of what the authors wrote.

We thank Reviewer #1 for this important comment. We agree that several figures could more directly illustrate the points made in the text. In a revised version, we intend to revise the relevant figure panels and legends to better align the visual message with the conclusions, and we will adjust the corresponding text to explicitly state what each figure demonstrates and how the data support our interpretation. We anticipate that these changes will improve clarity and strengthen the alignment between figures and text.

It'd be helpful for readers to see the phylogenetic tree of the rest of the organisms that harbor LppB homologs and Lpp.

We thank Reviewer #1 for this suggestion. We examined the distribution of Lpp-family proteins across closely related Enterobacteriaceae. While species such as Escherichia fergusonii, Shigella flexneri and Shigella dysenteriae encode Lpp and as well as a paralogous small lipoprotein (YqhH, see Fig.S7), we find that LppB-like orthologs (equivalent to lppB from Salmonella) appear to be restricted to Salmonella species to our knowledge. Because LppB shows this lineage-specific distribution, inclusion of a broader phylogenetic tree would primarily highlight its restricted presence rather that provide additional evolutionary insight. We will clarify this point in the revised manuscript.

Increased expression of LppB under low pH is subtle. This result would benefit from quantifying the blots (Fig. S1) and performing statistical analysis.

We thank Reviewer #1 for this observation. We agree that the increase in LppB levels at acidic pH appears modest. We will carefully reassess this result across independent experiments and, where technically appropriate, provide quantitative information to better document the magnitude of the effect. Additionally, we will revise the text to more accurately described the observed difference.

Similarly, the SDS-EDTA sensitivity result (Fig. S2) is not convincing; the image doesn't seem to show isolated colonies at low pH (Fig. S2B). Please measure CFU/mL and report endpoint growth graphs instead. Statistical analysis should also be presented.

We thank Reviewer #1 for this suggestion. We agree that the SDS-EDTA sensitivity assay presented in Fig. S2 could benefit from a more quantitative assessment. We will perform CFU/mL measurements from independent biological replicates to better quantify the observed differences and include statistical analysis when appropriate. In addition, we will revise the corresponding text to more accurately reflect the magnitude of the phenotype.

The reduction to PG crosslinking of the C57R mutant is unclear (Fig 4B lane 22). The authors state: "suggesting that additional features of the LppB C-terminal region underlie its reduced efficiency." Does this mean additional amino acids play a role? Did the authors try to substitute Cys with other amino acid residues like Ala or Ser and quantify protein levels to find a mutant with similar expression levels? Do these have less crosslinking too?

We thank Reviewer #1 for this important comment. As correctly noted, the reduced abundance of the LppB_C57R variant likely contributes to its reduced level of peptidoglycancrosslinked species. Therefore, we cannot formally distinguish whether the reduced peptidoglycan crosslinking reflects decreased intrinsic crosslinking efficiency or simply reduced protein abundance and stability. We will revise the text to clarify this point and explicitly acknowledge this limitation. The C57R substitution was chosen because arginine is present at the equivalent position in the Salmonella LppA homolog, allowing us to assess the functional consequences of a naturally occurring sequence variation between Lpp-family members. While substitutions such as C57A or C57S could further dissect the specific contribution of the cysteine residue, our use of the C57R substitution provides direct insight into the functional implications of this naturally occurring difference between Lpp homologs.

Reviewer #2 (Public review):

Summary:

The manuscript by Pierre Despas and co-workers, reports the biochemical characterization of LppB a peculiar Lpp (Braun's lipoprotein) homolog found in Salmonella enterica. S. enterica encodes two Lpp homologs LppA and LppB: while LppA and Lpp function similarly, the role of LppB is less clear. LppB shares with Lpp the Cterminal Lys needed for covalent attachment to peptidoglycan (PG) but diverges in residues that precede the terminal Lys featuring a Cys residue at the penultimate position. By using E. coli as a surrogate model, the authors show that LppB can be covalently linked to PG via the terminal Lys residues and that the penultimate Cys residue can be used to form homodimer species when expressed alone and heterotrimeric complexes when co-expressed with Lpp. Interestingly, LppB expressed in E. coli seems to be stabilized at acidic pH a condition Salmonella encounters in macrophage phagosomes. Finally, based on decreased intensity of LppB-PG crosslinked bands as LppB expression increases the authors suggest that LppB is able to negatively modulate the outer membrane-peptidoglycan connectivity.

Strengths:

The manuscript is interesting, describes a novel strategy employed by bacteria to fine tuning outer membrane-PG attachment and provides new insights into how envelope remodeling processes can contribute to bacterial fitness and pathogenicity.

Weaknesses:

The analysis and quantification of muropeptides formed in E. coli strains overexpressing LppB would strengthen the main conclusion of the manuscript.

We thank Reviewer #2 for this insightful comment. We agree that quantitative analysis of muropeptides in E. coli strains expressing LppB would strengthen the main conclusion. This point was also raised in the editorial assessment and by Reviewer #3, underscoring its importance. In a revised version, we plan to perform muropeptide profiling by HPLC, coupled where appropriate to mass spectrometry, to quantitatively assess peptidoglycan composition in the relevant strains.

Reviewer #3 (Public review):

Summary:

The manuscript is interesting, and it is clearly written. While the experiments are well executed, a general flaw is that the LppA/B analyses are done in the E. coli K12 host as surrogate for Salmonella enterica. For the mechanistic and molecular analyses of LppB a surrogate host is certainly adequate, yet it limits extrapolation of the physiological implications of LppB in the natural context. 

Strengths:

The work convincingly demonstrates that LppB forms disulfide-based dimers and that it is crosslinked to PG via LdtB in E. coli. Moreover, dimerization is required for LppB abundance in E. coli and LppB can inhibit crosslinking of Lpp/A to PG in E. coli. 

Weaknesses:

Regarding the key conclusion of the work: while it is shown that LppB is oxidized in E. coli, whether envelope integrity (or OMV production) changes arise from switches in oxidation of the LppB cysteines remains to be shown, for E. coli let alone in the native host Salmonella. Does expression of LppB influence Lpp/A activity or OM tethering in E. coli? Since the inhibition of the Lpp/A linking to PG is not affected by the oxidation state of LppB, the abstract/title implies redox-control of envelope integrity which is a bit misleading and an overstatement. Both are features of LppB: i.e. it dimerizes through disulfide bond formation and it reduces PG binding of Lpp/A through trimerization. However, no link between the two is shown.

We thank Reviewer #3 for this important comment and for highlighting the need to clarify the relationship between LppB oxidation, oligomerization, and its effect on peptidoglycan crosslinking. We agree that while our data demonstrate that LppB forms disulfide-linked oligomers and that LppB expression reduces Lpp/A attachment to peptidoglycan, our current results do not establish a direct causal link between the oxidation state of LppB and its ability to modulate outer membrane–peptidoglycan tethering. Therefore, we will revise the manuscript to avoid implying redox-dependent control of envelope integrity and to more clearly present these as distinct but potentially related properties of LppB.