eDNA-stimulated cell dispersion from Caulobacter crescentus biofilms upon oxygen limitation is dependent on a toxin–antitoxin system

  1. Cecile Berne
  2. Sébastien Zappa
  3. Yves V Brun

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This study will be of interest to a broad audience of microbiologists by providing one of the few examples of a clear phenotype for a toxin-antitoxin system. The conclusion that an oxygen-regulated toxin-antitoxin system is required for an important step in biofilm development in the model organism Caulobacter crescentus is well supported by the data and experiments are well designed and controlled. Some possible limitations in interpretations from incompletely controlled phenotype reporters should be resolved by simple experiments.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

In their natural environment, most bacteria preferentially live as complex surface-attached multicellular colonies called biofilms. Biofilms begin with a few cells adhering to a surface, where they multiply to form a mature colony. When conditions deteriorate, cells can leave the biofilm. This dispersion is thought to be an important process that modifies the overall biofilm architecture and that promotes colonization of new environments. In Caulobacter crescentus biofilms, extracellular DNA (eDNA) is released upon cell death and prevents newborn cells from joining the established biofilm. Thus, eDNA promotes the dispersal of newborn cells and the subsequent colonization of new environments. These observations suggest that eDNA is a cue for sensing detrimental environmental conditions in the biofilm. Here, we show that the toxin–antitoxin system (TAS) ParDE 4 stimulates cell death in areas of a biofilm with decreased O 2 availability. In conditions where O 2 availability is low, eDNA concentration is correlated with cell death. Cell dispersal away from biofilms is decreased when parDE 4 is deleted, probably due to the lower local eDNA concentration. Expression of parDE 4 is positively regulated by O 2 and the expression of this operon is decreased in biofilms where O 2 availability is low. Thus, a programmed cell death mechanism using an O 2 -regulated TAS stimulates dispersal away from areas of a biofilm with decreased O 2 availability and favors colonization of a new, more hospitable environment.

Article activity feed

  1. Version published to 10.7554/elife.80808 on eLife
  2. eLife

    Author Response

    Reviewer #1 (Public Review):

    This study sets out to decipher whether the eDNA that promotes biofilm dispersal in Caulobacter crescentus biofilms is released when a random portion of cells lyse within biofilms, or whether eDNA release is a regulated process. They start by investigating whether any of the C. crescentus TA systems contribute to biofilm-associated cell death, and find that one of the systems, ParDE4 is responsible for cell death and eDNA release. They go on to show that this system is O2-regulated and thus contributes to cell death in particular in the oxygen limited interior regions of biofilms. These findings contribute significantly to our understanding of the biological functions of toxin-antitoxin systems, mechanisms of bacterial programmed cell death, and biofilm growth. The notion that TA systems function in cell death in particular has been controversial, and often based on overexpression of the toxin component, therefore the fact that this study uses a TA system in its native genomic context is notable. The authors also show clearly the somewhat counterintuitive result that the cell death (and presumably, toxin activity) is negatively correlated with transcription of the TA system. This is consistent with what is known about TA biology (but not with many past TA papers, which often correlated TA transcription with toxin activation). The study also provides a logical rationale for how ParDE4 mediated cell death ultimately contributes to bacterial fitness. The paper is well written and figures are clear and easy to follow.

    There are two relatively minor shortcomings of the paper, both acknowledged as caveats by the authors in their discussion. First, while the authors do include one experiment that addresses whether the toxin is responsible for the cell death (Fig 3), they do not show direct evidence of the activity of the toxin other than cell death/eDNA release. Second, the authors do not address whether the reduced TA transcription they observe is what causes the release of the toxin and thus the cell death phenotype. This seems likely to be the case based on previous studies of other TA systems (e.g. TA systems involved in plasmid segregation, most clearly shown for CcdAB, or more recently the ToxIN system during phage infection). Connecting this directly would be a very valuable addition to this study.

    We thank the reviewer for those positive comments. We agree that the TA system we describe in this study needs to be characterized in more detail. Understanding how this TA expression levels are linked to cell death is our next goal and will be the scope of a future publication.

    We now discuss the important missing point about possible TA expression being linked to cell death and refer to CcdAB, ToxIN and other relevant systems, as well characterized examples of such mechanisms. In the introduction, we now present the role of TAS in plasmid addiction and phage defense mechanisms. We also provide more information about those systems in the discussion and speculate the similarities with the TAS described here (see our reply to essential revisions above).

    Reviewer #2 (Public Review):

    In this work, the authors present compelling evidence that a toxin-antitoxin system contributes to biofilm dispersal under oxygen limited conditions. This work makes important contributions to two areas of microbial physiology; functional understanding of toxin-antitoxin systems, which have remained largely elusive, and mechanistic regulation or biofilm dispersal, is a critical, but less understood aspect of biofilm physiology.

    A major goal of the work described in this manuscript was to better understand the regulation of biofilm dispersal. These authors provide compelling evidence that the parDE4 toxin-antitoxin (TA) system in Caulobacter crescentus mediates enhanced cell death under conditions of oxygen limitation. This group previously reported that extracellular DNA (eDNA) inhibits attachment of new-born swarmer cells. Here they build on that observation by identifying a genetic module that contributes to cell death and DNA release under oxygen limitation, a sub-optimal condition present in a dense biofilm community, and demonstrate that parDE4 affects biofilm development. Together, this work makes important contributions toward understanding functional roles for toxin-antitoxin systems and regulation of mature stages of biofilm development. In addition, although eDNA is often depicted as having a structural role in strengthening and maintaining biofilms in some species, this work further establishes that eDNA can have multiple roles in biofilms including contributing to dispersal in Caulobacter.

    Strengths of this work include 1) comprehensive evaluation of multiple paralogous TAS and specific identification of the contribution of parDE4 to cell death, eDNA release and biofilm restriction, 2) genetic dissection of the TA pair to establish that the ParD4-antitoxin prevents eDNA release and promotes biofilm formation in a ParE4-toxin dependent manner, 3) provision of evidence that the parDE system affects cell death / eDNA release, but not responsiveness to eDNA, 4) demonstration of an anti-correlation between expression of parDE and ccoN, a hypoxic responsive gene, at both the population level under different growth conditions and at the single cell level within different growth conditions.

    We thank the reviewer for these positive comments.

    One weakness of this work is that the authors do not directly measure O2 concentrations in their growth conditions. However, they do monitor activity of an established hypoxic responsive promoter, which provides strong evidence that the various conditions tested do indeed affect oxygen concentrations in the culture medium. Nevertheless, it is difficult to assess oxygen availability in the flow cell experiments, which will be dependent on both dissolved oxygen in the media pumped through the flow cell and cell density within the flow cells. In the competition experiments, the ∆parDE4 mutant has an advantage before there seems to be an appreciable cell density, perhaps reflecting low oxygen in the growth medium or a monolayer of cells that is not obvious in the images as presented. It would be interesting to evaluate expression of ccoN in biofilms grown under these flow conditions.

    We agree with the reviewer that one limitation of our study is that we could not directly measure the O2 concentration in our different growth conditions. Unfortunately, we were unable to find a way to reliably and reproducibly assay the dissolved O2 concentration in our experimental set-ups (both static biofilms and flow-cells). We think that regulation of parDE4 expression is linked to the composition of the local environment surrounding each cell, and offering a proxy via ccoN expression is the best method we could provide to assess this. Results provided in Figures 7 and S3 (now S5) clearly show that cells that respond to limiting O2 levels (by activating ccoN expression) have low parDE4 expression. We also show in this set of experiments that, at the population level, there are cells highly expressing ccoN or parDE4 regardless of the culture conditions and the overall O2 levels.

    We now provide the expression of ccoN in different areas of biofilms, in addition to the already presented parDE4 expression, in Fig. 8A. We quantified ccoN transcription levels using the PccoN-lacZ construct (already used to generate data in Figure 5) and the fluorogenic ß-galactosidase substrate we used to quantify parDE4 expression in biofilms in the first version of this manuscript (Figure 8A). These new results now show that in biofilm areas where parDE4 is more expressed, ccoN expression is low and vice-versa and confirm other observations made throughout this work.

    The discussion regarding the observation that parDE expression drops under activating (oxygen limiting) conditions is contradictory to what I would expect based on the early findings about TA systems as genetic stabilization systems. The authors seem to expect that conditions that activate the toxin should correspond to increase expression of the TA operon. However, TA systems have frequently been characterized as DNA stabilization systems for plasmids or other mobile elements because the toxin proteins are more stable than the antitoxin proteins. In these cases, if the gene pair is lost (or in this case if expression is decreased) then the toxin protein persists longer than the antitoxin protein, effectively activating the toxin to arrest or kill cells that have lost (or in this case turned off) the gene pair. Thus I disagree with the statement that this is a "novel regulatory mechanism of PCD that remains to be understood" (line 436-7).

    The sentence preceding this one was "We are unaware of cases where reduced TAS expression is correlated with the condition that activates the PCD in biofilm regulation." and we suggested a "novel regulatory mechanism of PCD" in the context of biofilm formation. However, we realize now that our statements could be misleading and we entirely rewrote this section (Lines 510-519: " It is interesting to note that the "neutralized" steady state of the ParDE4 TAS, when the toxin is inactivated, seems to be when O2 is abundant, i.e, when parDE4 transcription is at its highest. In most studied TAS, stresses have been shown to induce transcription of TAS (LeRoux et al., 2020, Jurėnas et al., 2022), but here, the stress inflicted on the cells by O2 limitation is accompanied by a lower expression of parDE4. We are unaware of cases where reduced TAS expression is correlated with the condition that activates the PCD in biofilm regulation. This suggests a novel regulatory mechanism of PCD, in the context of biofilms, that remains to be understood.").

    Differential stability of toxin and antitoxin proteins provides a reasonable regulatory mechanism to explain the programed cell death observed. Testing of this, or other, mechanistic model(s) will be important in future studies of this system.

    We agree with the reviewer and testing protein stability is definitively on the list of experiments to do to dissect this TA killing mechanism in the near future. As mentioned above, we have been unable to obtain antibodies to these proteins so far, delaying these types of experiments.

  3. eLife

    Evaluation Summary:

    This study will be of interest to a broad audience of microbiologists by providing one of the few examples of a clear phenotype for a toxin-antitoxin system. The conclusion that an oxygen-regulated toxin-antitoxin system is required for an important step in biofilm development in the model organism Caulobacter crescentus is well supported by the data and experiments are well designed and controlled. Some possible limitations in interpretations from incompletely controlled phenotype reporters should be resolved by simple experiments.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  4. eLife

    Reviewer #1 (Public Review):

    This study sets out to decipher whether the eDNA that promotes biofilm dispersal in Caulobacter crescentus biofilms is released when a random portion of cells lyse within biofilms, or whether eDNA release is a regulated process. They start by investigating whether any of the C. crescentus TA systems contribute to biofilm-associated cell death, and find that one of the systems, ParDE4 is responsible for cell death and eDNA release. They go on to show that this system is O2-regulated and thus contributes to cell death in particular in the oxygen limited interior regions of biofilms. These findings contribute significantly to our understanding of the biological functions of toxin-antitoxin systems, mechanisms of bacterial programmed cell death, and biofilm growth. The notion that TA systems function in cell death in particular has been controversial, and often based on overexpression of the toxin component, therefore the fact that this study uses a TA system in its native genomic context is notable. The authors also show clearly the somewhat counterintuitive result that the cell death (and presumably, toxin activity) is negatively correlated with transcription of the TA system. This is consistent with what is known about TA biology (but not with many past TA papers, which often correlated TA transcription with toxin activation). The study also provides a logical rationale for how ParDE4 mediated cell death ultimately contributes to bacterial fitness. The paper is well written and figures are clear and easy to follow.

    There are two relatively minor shortcomings of the paper, both acknowledged as caveats by the authors in their discussion. First, while the authors do include one experiment that addresses whether the toxin is responsible for the cell death (Fig 3), they do not show direct evidence of the activity of the toxin other than cell death/eDNA release. Second, the authors do not address whether the reduced TA transcription they observe is what causes the release of the toxin and thus the cell death phenotype. This seems likely to be the case based on previous studies of other TA systems (e.g. TA systems involved in plasmid segregation, most clearly shown for CcdAB, or more recently the ToxIN system during phage infection). Connecting this directly would be a very valuable addition to this study.

  5. eLife

    Reviewer #2 (Public Review):

    In this work, the authors present compelling evidence that a toxin-antitoxin system contributes to biofilm dispersal under oxygen limited conditions. This work makes important contributions to two areas of microbial physiology; functional understanding of toxin-antitoxin systems, which have remained largely elusive, and mechanistic regulation or biofilm dispersal, is a critical, but less understood aspect of biofilm physiology.

    A major goal of the work described in this manuscript was to better understand the regulation of biofilm dispersal. These authors provide compelling evidence that the parDE4 toxin-antitoxin (TA) system in Caulobacter crescentus mediates enhanced cell death under conditions of oxygen limitation. This group previously reported that extracellular DNA (eDNA) inhibits attachment of new-born swarmer cells. Here they build on that observation by identifying a genetic module that contributes to cell death and DNA release under oxygen limitation, a sub-optimal condition present in a dense biofilm community, and demonstrate that parDE4 affects biofilm development. Together, this work makes important contributions toward understanding functional roles for toxin-antitoxin systems and regulation of mature stages of biofilm development. In addition, although eDNA is often depicted as having a structural role in strengthening and maintaining biofilms in some species, this work further establishes that eDNA can have multiple roles in biofilms including contributing to dispersal in Caulobacter.

    Strengths of this work include 1) comprehensive evaluation of multiple paralogous TAS and specific identification of the contribution of parDE4 to cell death, eDNA release and biofilm restriction, 2) genetic dissection of the TA pair to establish that the ParD4-antitoxin prevents eDNA release and promotes biofilm formation in a ParE4-toxin dependent manner, 3) provision of evidence that the parDE system affects cell death / eDNA release, but not responsiveness to eDNA, 4) demonstration of an anti-correlation between expression of parDE and ccoN, a hypoxic responsive gene, at both the population level under different growth conditions and at the single cell level within different growth conditions.

    One weakness of this work is that the authors do not directly measure O2 concentrations in their growth conditions. However, they do monitor activity of an established hypoxic responsive promoter, which provides strong evidence that the various conditions tested do indeed affect oxygen concentrations in the culture medium. Nevertheless, it is difficult to assess oxygen availability in the flow cell experiments, which will be dependent on both dissolved oxygen in the media pumped through the flow cell and cell density within the flow cells. In the competition experiments, the ∆parDE4 mutant has an advantage before there seems to be an appreciable cell density, perhaps reflecting low oxygen in the growth medium or a monolayer of cells that is not obvious in the images as presented. It would be interesting to evaluate expression of ccoN in biofilms grown under these flow conditions.

    The discussion regarding the observation that parDE expression drops under activating (oxygen limiting) conditions is contradictory to what I would expect based on the early findings about TA systems as genetic stabilization systems. The authors seem to expect that conditions that activate the toxin should correspond to increase expression of the TA operon. However, TA systems have frequently been characterized as DNA stabilization systems for plasmids or other mobile elements because the toxin proteins are more stable than the antitoxin proteins. In these cases, if the gene pair is lost (or in this case if expression is decreased) then the toxin protein persists longer than the antitoxin protein, effectively activating the toxin to arrest or kill cells that have lost (or in this case turned off) the gene pair. Thus I disagree with the statement that this is a "novel regulatory mechanism of PCD that remains to be understood" (line 436-7).

    Differential stability of toxin and antitoxin proteins provides a reasonable regulatory mechanism to explain the programed cell death observed. Testing of this, or other, mechanistic model(s) will be important in future studies of this system.

  6. Version published to 10.1101/2022.06.17.496608v1 on bioRxiv