Bacteria frequently colonize environmental habitats and infection sites by forming surface-attached multicellular communities called biofilms. Participating in the biofilm lifestyle allows bacteria to collectively acquire nutrients and resist threats (Flemming et al., 2016). By contrast, the individual free-swimming state allows bacteria to roam. The global pathogen Vibrio cholerae undergoes repeated rounds of clonal biofilm formation and disassembly, and both biofilm formation and biofilm exit are central to disease transmission as V. cholerae alternates between the marine niche and the human host (Conner et al., 2016; Gallego-Hernandez et al., 2020; Tamayo et al., 2010). The V. cholerae biofilm lifecycle occurs in three stages. First, a founder cell attaches to a substate; second, through cycles of growth, division, and extracellular matrix secretion, the biofilm matures; and finally, when the appropriate environmental conditions are detected, cells leave the biofilm through a process termed dispersal (Bridges et al., 2020). Bacterial cells liberated through dispersal can depart their current environment and repeat the lifecycle in a new niche, such as in a new host (Guilhen et al., 2017). In general, the initial stages of the biofilm lifecycle, from attachment through maturation, have been well studied in V. cholerae and other bacterial species, whereas the dispersal phase remains underexplored. Components underlying biofilm dispersal are beginning to be revealed. For example, in many bacterial species, the Lap system that is responsible for cell attachments during biofilm formation is also involved in dispersal (Boyd et al., 2012; Christensen et al., 2020; Gjermansen et al., 2005; Kitts et al., 2019; Newell et al., 2011).

To uncover mechanisms controlling V. cholerae biofilm dispersal, we recently developed a brightfield microscopy assay that allows us to follow the entire biofilm lifecycle (Bridges and Bassler, 2019). We combined this assay with mutagenesis and high-content imaging to identify mutants that failed to properly disperse (Bridges et al., 2020). Our screen revealed genes encoding proteins that fall into three functional groups: signal transduction, matrix degradation, and cell motility. Our hypothesis is that these classes of components act in sequence to drive biofilm dispersal. First, the cues that induce dispersal are detected by the signal transduction proteins. Second, activation of matrix digestion components occurs. Finally, cell motility engages and permits cells to swim away from the disassembling biofilm. The majority of the genes identified in the screen encoded signal transduction proteins. Elsewhere, we characterized one of the signaling systems identified in the screen, a new two-component phospho-relay that we named DbfS-DbfR (for Dispersal of Biofilm Sensor and Dispersal of Biofilm Regulator) that controls biofilm dispersal (Bridges et al., 2020). Of the remaining signaling proteins identified in the screen, four are proteins involved in regulating production/degradation of, or responding to, the second messenger molecule cyclic diguanylate (c-di-GMP).

c-di-GMP regulates biofilm formation in many bacteria including V. cholerae in which low c-di-GMP levels correlate with motility and high c-di-GMP levels promote surface attachment and matrix production, and thus, biofilm formation (Conner et al., 2017). c-di-GMP is synthesized by enzymes that contain catalytic GGDEF domains that have diguanylate cyclase activity (Conner et al., 2017). c-di-GMP is degraded by proteins with EAL or HD-GYP domains that possess phosphodiesterase activity. V. cholerae encodes >50 proteins harboring one or both of these domains, underscoring the global nature of c-di-GMP signaling in this pathogen. The activities of these enzymes are often regulated by environmental stimuli through ligand binding sensory domains; however, in most cases, the identities of ligands for GGDEF- and EAL-containing enzymes are unknown (Römling et al., 2013). Of the c-di-GMP regulatory proteins identified in our dispersal screen, only one, the hybrid GGDEF/EAL protein, MbaA, which has previously been shown to regulate V. cholerae biofilm formation (Bomchil et al., 2003), has a defined environmental stimulus: MbaA responds to the ubiquitous family of small molecules called polyamines, aliphatic cations containing two or more amine groups (Miller-Fleming et al., 2015). Our dispersal screen also identified PotD1, a periplasmic binding protein that mediates polyamine import and that has also been shown to control V. cholerae biofilm formation (Cockerell et al., 2014; McGinnis et al., 2009). Thus, identification of these two genes in our screen suggested that beyond roles in biofilm formation, polyamine detection could be central to biofilm dispersal. In the present work, we explore this new possibility.

Polyamines are essential small molecules derived from amino acids and were likely produced by the last universal common ancestor (Michael, 2018; Miller-Fleming et al., 2015). Different organisms make distinct blends of polyamines (Michael, 2018), and they are involved in biological processes ranging from stress responses to signal transduction to protein synthesis to siderophore production (Karatan et al., 2005; McGinnis et al., 2009; Michael, 2018). In some cases, mechanisms underlying polyamine function are known. For example, the eukaryotic translation initiation factor 5A is modified with hypusine that is derived from the polyamine spermidine (Puleston et al., 2019). In most cases, however, the mechanisms by which polyamines control physiology remain mysterious (Miller-Fleming et al., 2015).

Polyamines are known to be involved in biofilm development in many bacterial species (Hobley et al., 2014; Karatan and Michael, 2013; Michael, 2018; Nesse et al., 2015; Patel et al., 2006). Intriguingly, the biofilm alterations that occur in response to a given polyamine vary among species. In V. cholerae, norspermidine, a rare polyamine produced by Vibrionaceae and select other organisms, promotes biofilm formation (Karatan et al., 2005; Lee et al., 2009). In contrast, the nearly ubiquitous polyamine spermidine, which differs from norspermidine only by a single methylene group, is not produced at substantial levels by V. cholerae and it represses V. cholerae biofilm formation (Figure 1; Karatan et al., 2005; Lee et al., 2009; McGinnis et al., 2009). Likewise, the polyamine spermine, typically made by eukaryotes, also represses V. cholerae biofilm formation (Sobe et al., 2017). Both spermidine and spermine are abundant in the human intestine (Benamouzig et al., 1997; McEvoy and Hartley, 1975; Osborne and Seidel, 1990). These results have led to the hypothesis that V. cholerae detects norspermidine as a measure of ‘self’ and spermidine and spermine as measures of ‘other’ to assess the species composition of the vicinal community (Sobe et al., 2017; Wotanis et al., 2017). V. cholerae funnels that information internally to determine whether or not to make a biofilm (Sobe et al., 2017; Wotanis et al., 2017). V. cholerae produces intracellular norspermidine; however, norspermidine has not been detected in cell-free culture fluids of laboratory-grown strains (Parker et al., 2012). Thus, if norspermidine does indeed enable V. cholerae to take a census of ‘self,’ it is apparently not via a canonical quorum-sensing-type mechanism.

Figure 1

Download asset Open asset

As mentioned above, information contained in extracellular polyamines is transduced internally by MbaA, which is embedded in the inner membrane. MbaA contains a periplasmic domain that interacts with the periplasmic binding protein NspS. nspS is located immediately upstream of mbaA in the chromosome (Cockerell et al., 2014; Karatan et al., 2005; Young et al., 2021). MbaA also possesses cytoplasmic GGDEF (SGDEF in MbaA) and EAL (EVL in MbaA) domains (Figure 1). Genetic evidence suggests that when NspS is bound to norspermidine, the complex associates with MbaA and biofilm formation is promoted (Cockerell et al., 2014). Specifically, NspS interaction with MbaA inhibits MbaA phosphodiesterase activity, leading to increased c-di-GMP levels, and in turn, elevated biofilm formation (Figure 1, left panel) (Cockerell et al., 2014). It is proposed that Apo-NspS and NspS bound to spermidine or spermine do not bind to MbaA, and thus, under this condition, biofilm formation is reduced (Cockerell et al., 2014; Sobe et al., 2017). In this case, MbaA phosphodiesterase activity is not inhibited, which leads to decreased c-di-GMP levels and repression of biofilm formation (Figure 1, right panel). Currently, it is not known whether or not MbaA has diguanylate cyclase activity. The purified MbaA cytoplasmic domain functions as a phosphodiesterase in vitro (Cockerell et al., 2014).

In a proposed second regulatory mechanism, norspermidine and spermidine are thought to control biofilm formation via import through the inner membrane ABC transporter, PotABCD1 (McGinnis et al., 2009). Following internalization, via an undefined cytoplasmic mechanism, polyamines are proposed to modulate biofilm formation (Figure 1; McGinnis et al., 2009). This hypothesis was based on the finding that elimination of PotD1, which is required for import of norspermidine and spermidine, resulted in elevated biofilm formation. Importantly, the studies reporting the MbaA and PotD1 findings used quorum-sensing-deficient V. cholerae strains (Joelsson et al., 2006) that likely cannot disperse from biofilms. Furthermore, the end-point biofilm assays used could not differentiate between enhanced biofilm formation and the failure to disperse.

Here, we combine real-time biofilm lifecycle measurements, mutagenesis, a reporter of cytoplasmic c-di-GMP levels, and measurements of intra- and extracellular polyamine concentrations to define the roles that norspermidine and spermidine play in controlling the V. cholerae biofilm lifecycle. In wildtype V. cholerae, exogenous norspermidine and spermidine inversely alter cytoplasmic c-di-GMP levels and they exert their effects through the NspS-MbaA circuit. Norspermidine promotes biofilm formation and suppresses biofilm dispersal. Spermidine represses biofilm formation and promotes biofilm dispersal. Both the MbaA SGDEF and EVL domains are required for polyamine control of the biofilm lifecycle. We provide evidence that MbaA synthesizes c-di-GMP in the presence of norspermidine and degrades c-di-GMP in the presence of spermidine and that is what drives V. cholerae to form and disperse from biofilms, respectively. When MbaA is absent, V. cholerae is unable to alter the biofilm lifecycle in response to extracellular polyamines. We demonstrate that polyamine internalization via PotD1 is not required to promote V. cholerae entrance or exit from biofilms, but rather, periplasmic detection of polyamines by MbaA is the key regulatory step. Specifically, our results suggest that the ΔpotD1 mutant fails to disperse because it is unable to reimport self-secreted norspermidine. The consequence is that excess periplasmic norspermidine accumulates, is detected by MbaA, and leads to production of c-di-GMP and suppression of biofilm dispersal. Collectively, our work reveals the mechanisms by which V. cholerae detects and transduces the information contained in polyamine signals into modulation of its biofilm lifecycle. We propose that the polyamine sensing mechanisms revealed in this study allow V. cholerae to distinguish relatives from competitors and potentially the presence of predators in the vicinity and, in response, modify its biofilm lifecycle to appropriately colonize territory or disperse from an existing community.

In this study, we investigated the effects of the polyamines norspermidine and spermidine on V. cholerae biofilm dispersal. Norspermidine and spermidine, which differ in structure only by one methylene group, mediate starkly opposing effects on the V. cholerae biofilm lifecycle: norspermidine inhibits and spermidine promotes biofilm dispersal. Both function through the NspS-MbaA circuit. Thus, the polyamine binding protein NspS must harbor the exquisite capability to assess the presence or absence of a single chemical moiety as norspermidine binding drives NspS interaction with MbaA, whereas binding to spermidine prevents this interaction. Here, we speculate on the possible biological significance of our findings. We suspect that norspermidine and spermidine act as ‘self’ and ‘other’ cues, respectively. Specifically, norspermidine is a rare polyamine in the biosphere, produced only by select organisms, namely V. cholerae and closely related marine vibrios (Hamana, 1997; Michael, 2018; Yamamoto et al., 1991). The observation that laboratory-grown WT V. cholerae releases little norspermidine, at least in part due to PotD1-mediated reuptake (Figure 5C), suggests that this system does not behave like a canonical quorum-sensing pathway. However, it is possible that norspermidine is secreted by V. cholerae under some environmental conditions, or by other vibrios, and V. cholerae detects the released norspermidine via NspS-MbaA, and its dispersal from biofilms is prevented. Thus, when close relatives are nearby, as judged by detection of the presence of norspermidine, V. cholerae elects to remain in its current biofilm niche. More speculative is the notion that norspermidine functions as a cue for phage-, predator-, or toxin-induced cell lysis. Specifically, if V. cholerae can detect norspermidine released by neighboring lysed V. cholerae cells, in response to the perceived danger, V. cholerae would remain in the protective biofilm state. Conversely, detection of spermidine, a nearly ubiquitously produced polyamine (Michael, 2018), could alert V. cholerae to the presence of competing or unrelated organisms. In this case, V. cholerae would respond by dispersing from biofilms and fleeing that locale. In a seemingly parallel scenario, we previously discovered that autoinducer AI-2, a universally produced quorum-sensing signal, also drove V. cholerae biofilm repression and premature dispersal, while CAI-1, the V. cholerae ‘kin’ quorum-sensing autoinducer, did not (Bridges and Bassler, 2019). Together, our findings suggest that V. cholerae possesses two mechanisms to monitor its own numbers (norspermidine and CAI-1) and two mechanisms to take a census of unrelated organisms in the environment (spermidine and AI-2), and it responds by remaining in the biofilm state when closely related species are in the majority and by exiting the biofilm state when the density of non-related organisms is high, presumably to escape competition/danger.

We found that periplasmic polyamine detection by NspS-MbaA, but not polyamine import, controls the biofilm dispersal program and both the MbaA EVL and SGDEF domains are required for the V. cholerae response to polyamines. Despite previous work suggesting that MbaA functions exclusively as a phosphodiesterase and that its SGDEF domain, which possesses a serine substitution in the active site relative to the canonical GGDEF active site, is inert with respect to c-di-GMP synthesis, our data suggest that when norspermidine is present, MbaA does synthesize c-di-GMP (Figure 3C, Figure 4C), as has also been shown for other SGDEF proteins (Pérez-Mendoza et al., 2011). Therefore, MbaA is likely a bifunctional protein capable of both c-di-GMP synthesis and degradation. We wonder how such an arrangement benefits V. cholerae in its response to polyamines. We speculate that the ability of MbaA to inversely alter c-di-GMP levels in response to two discrete cues coordinates polyamine signaling and enables more rapid and greater magnitude changes in c-di-GMP levels than would be achievable if two separate receptors existed, each harboring only a single catalytic activity and each responsive to only one polyamine.

An analogous bifunctional arrangement exists for the quorum-sensing two-component receptor LuxN from Vibrio harveyi. LuxN is an inner membrane receptor that harbors a periplasmic sensory domain responsible for detection of a self-made small molecule cue, a homoserine lactone agonist with a four-carbon acyl tail (Bassler et al., 1994). The ligand encodes species-specific information about cell population density. Competing bacterial species that co-occupy the niche in which V. harveyi resides produce similar acyl homoserine lactone signal molecules, and notably, those molecules differ only in the acyl tail length or decoration (Papenfort and Bassler, 2016). The tails of the molecules produced by the competitors usually possess more than four carbons, and they act as potent LuxN antagonists. With respect to mechanism, the self-made agonist drives LuxN into phosphatase mode while the antagonists induce LuxN to act as a kinase (Ke et al., 2015). Thus, as in the MbaA circuit, binding of the different ligands to LuxN propels opposing enzymatic activities in the cytoplasm. We speculate that ligand-driven alternations between opposing enzymatic activities could be an underappreciated mechanism that sensors employ to transmit species-specific and species-non-specific information into discrete changes in downstream behaviors.

In addition to conveying information about the species composition of the vicinal community, the MbaA SGDEF and EVL domains could also participate in dimerization and/or allosteric binding to c-di-GMP or other metabolites to mediate feedback regulation of the opposing MbaA activities. Parallel examples exist, here we present one: Caulobacter crescentus contains a GGDEF/EAL protein called CC3396, in which the GGDEF domain is non-catalytic but it nonetheless binds to GTP, which allosterically activates the phosphodiesterase activity in the neighboring EAL domain (Christen et al., 2005). A feedback mechanism could provide MbaA with the ability to integrate the information encoded in extracellular polyamine blends with cytoplasmic cues, such as the metabolic state of the cell or the current cytoplasmic c-di-GMP level. Taken together, the double ligand detection capability linked to the dual catalytic activities of the NspS-MbaA circuit allows V. cholerae to distinguish between remarkably similar ligands and, in response, have the versatility to convey distinct information into the cell to alter the biofilm lifecycle.

In summary, four signaling pathways have now been defined that feed into the regulation of V. cholerae biofilm dispersal: starvation (RpoS) (Singh et al., 2017), quorum sensing (via CAI-1 and AI-2) (Bridges and Bassler, 2019; Singh et al., 2017), the recently identified DbfS/DbfR cascade (ligand unknown) (Bridges et al., 2020), and through the current work, polyamine signaling via NspS-MbaA. We propose that integrating the information contained in these different stimuli into the control of biofilm dispersal endows V. cholerae with the ability to successfully evaluate multiple features of its fluctuating environment prior to committing to the launch of this key lifestyle transition that, ultimately, impinges on its overall survival. Because V. cholerae biofilm formation and biofilm dispersal are intimately connected to cholera disease and its transmission, we suggest that deliberately controlling the biofilm lifecycle, possibly via synthetic strategies that target polyamine signal transduction via NspS-MbaA, could be a viable therapeutic strategy to ameliorate disease.

All data generated and analyzed in this study are included in the manuscript and supporting files. Source data files have been provided in Zenodo (https://doi.org/10.5281/zenodo.4651348).

1. 1. Bridges AA
   2. Bassler BL

   (2021) Zenodo

   Inverse regulation of Vibrio cholerae biofilm dispersal by polyamine signals.

   https://doi.org/10.5281/zenodo.4651348

1. 1. Bassler BL
   2. Wright M
   3. Silverman MR

   (1994) Multiple signalling systems controlling expression of luminescence in Vibrio harveyi: sequence and function of genes encoding a second sensory pathway

   Molecular Microbiology 13:273–286.

   https://doi.org/10.1111/j.1365-2958.1994.tb00422.x

   • PubMed
   • Google Scholar
2. 1. Benamouzig R
   2. Mahé S
   3. Luengo C
   4. Rautureau J
   5. Tomé D

   (1997) Fasting and postprandial polyamine concentrations in the human digestive lumen

   The American Journal of Clinical Nutrition 65:766–770.

   https://doi.org/10.1093/ajcn/65.3.766

   • PubMed
   • Google Scholar
3. 1. Bomchil N
   2. Watnick P
   3. Kolter R

   (2003) Identification and characterization of a Vibrio cholerae gene, mbaA, involved in maintenance of biofilm architecture

   Journal of Bacteriology 185:1384–1390.

   https://doi.org/10.1128/JB.185.4.1384-1390.2003

   • PubMed
   • Google Scholar
4. 1. Boyd CD
   2. Chatterjee D
   3. Sondermann H
   4. O'Toole GA

   (2012) LapG, required for modulating biofilm formation by Pseudomonas fluorescens Pf0-1, is a calcium-dependent protease

   Journal of Bacteriology 194:4406–4414.

   https://doi.org/10.1128/JB.00642-12

   • PubMed
   • Google Scholar
5. 1. Bridges AA
   2. Fei C
   3. Bassler BL

   (2020) Identification of signaling pathways, matrix-digestion enzymes, and motility components controlling Vibrio cholerae biofilm dispersal

   PNAS 117:32639–32647.

   https://doi.org/10.1073/pnas.2021166117

   • PubMed
   • Google Scholar
6. 1. Bridges AA
   2. Bassler BL

   (2019) The intragenus and interspecies quorum-sensing autoinducers exert distinct control over Vibrio cholerae biofilm formation and dispersal

   PLOS Biology 17:e3000429.

   https://doi.org/10.1371/journal.pbio.3000429

   • PubMed
   • Google Scholar
7. 1. Christen M
   2. Christen B
   3. Folcher M
   4. Schauerte A
   5. Jenal U

   (2005) Identification and characterization of a cyclic di-GMP-specific phosphodiesterase and its allosteric control by GTP

   Journal of Biological Chemistry 280:30829–30837.

   https://doi.org/10.1074/jbc.M504429200

   • Google Scholar
8. 1. Christensen DG
   2. Marsden AE
   3. Hodge‐Hanson K
   4. Essock‐Burns T
   5. Visick KL

   (2020) LapG mediates biofilm dispersal in Vibrio fischeri by controlling maintenance of the VCBS‐containing adhesin LapV

   Molecular Microbiology 114:742–761.

   https://doi.org/10.1111/mmi.14573

   • Google Scholar
9. 1. Cockerell SR
   2. Rutkovsky AC
   3. Zayner JP
   4. Cooper RE
   5. Porter LR
   6. Pendergraft SS
   7. Parker ZM
   8. McGinnis MW
   9. Karatan E

   (2014) Vibrio cholerae NspS, a homologue of ABC-type periplasmic solute binding proteins, facilitates transduction of polyamine signals independent of their transport

   Microbiology 160:832–843.

   https://doi.org/10.1099/mic.0.075903-0

   • Google Scholar
10. 1. Conner JG
    2. Teschler JK
    3. Jones CJ
    4. Yildiz FH

    (2016) Staying alive: Vibrio cholerae’s Cycle of Environmental Survival, Transmission, and Dissemination

    Microbiology Spectrum 4:2015.

    https://doi.org/10.1128/microbiolspec.VMBF-0015-2015

    • Google Scholar
11. 1. Conner JG
    2. Zamorano-Sánchez D
    3. Park JH
    4. Sondermann H
    5. Yildiz FH

    (2017) The ins and outs of cyclic di-GMP signaling in Vibrio cholerae

    Current Opinion in Microbiology 36:20–29.

    https://doi.org/10.1016/j.mib.2017.01.002

    • PubMed
    • Google Scholar
12. 1. Flemming HC
    2. Wingender J
    3. Szewzyk U
    4. Steinberg P
    5. Rice SA
    6. Kjelleberg S

    (2016) Biofilms: an emergent form of bacterial life

    Nature Reviews Microbiology 14:563–575.

    https://doi.org/10.1038/nrmicro.2016.94

    • PubMed
    • Google Scholar
13. 1. Gallego-Hernandez AL
    2. DePas WH
    3. Park JH
    4. Teschler JK
    5. Hartmann R
    6. Jeckel H
    7. Drescher K
    8. Beyhan S
    9. Newman DK
    10. Yildiz FH

    (2020) Upregulation of virulence genes promotes Vibrio cholerae biofilm hyperinfectivity

    PNAS 117:11010–11017.

    https://doi.org/10.1073/pnas.1916571117

    • PubMed
    • Google Scholar
14. 1. Gjermansen M
    2. Ragas P
    3. Sternberg C
    4. Molin S
    5. Tolker-Nielsen T

    (2005) Characterization of starvation-induced dispersion in Pseudomonas putida biofilms

    Environmental Microbiology 7:894–904.

    https://doi.org/10.1111/j.1462-2920.2005.00775.x

    • PubMed
    • Google Scholar
15. 1. Guilhen C
    2. Forestier C
    3. Balestrino D

    (2017) Biofilm dispersal: multiple elaborate strategies for dissemination of bacteria with unique properties

    Molecular Microbiology 105:188–210.

    https://doi.org/10.1111/mmi.13698

    • Google Scholar
16. 1. Hamana K

    (1997) Polyamine distribution patterns within the families Aeromonadaceae, vibrionaceae, Pasteurellaceae, and Halomonadaceae, and related genera of the gamma subclass of the proteobacteria

    The Journal of General and Applied Microbiology 43:49–59.

    https://doi.org/10.2323/jgam.43.49

    • PubMed
    • Google Scholar
17. 1. Hobley L
    2. Kim SH
    3. Maezato Y
    4. Wyllie S
    5. Fairlamb AH
    6. Stanley-Wall NR
    7. Michael AJ

    (2014) Norspermidine is not a self-produced trigger for biofilm disassembly

    Cell 156:844–854.

    https://doi.org/10.1016/j.cell.2014.01.012

    • PubMed
    • Google Scholar
18. 1. Joelsson A
    2. Liu Z
    3. Zhu J

    (2006) Genetic and phenotypic diversity of quorum-sensing systems in clinical and environmental isolates of Vibrio cholerae

    Infection and Immunity 74:1141–1147.

    https://doi.org/10.1128/IAI.74.2.1141-1147.2006

    • PubMed
    • Google Scholar
19. 1. Karatan E
    2. Duncan TR
    3. Watnick PI

    (2005) NspS, a predicted polyamine sensor, mediates activation of Vibrio cholerae biofilm formation by norspermidine

    Journal of Bacteriology 187:7434–7443.

    https://doi.org/10.1128/JB.187.21.7434-7443.2005

    • PubMed
    • Google Scholar
20. 1. Karatan E
    2. Michael AJ

    (2013) A wider role for polyamines in biofilm formation

    Biotechnology Letters 35:1715–1717.

    https://doi.org/10.1007/s10529-013-1286-3

    • PubMed
    • Google Scholar
21. 1. Kashiwagi K
    2. Innami A
    3. Zenda R
    4. Tomitori H
    5. Igarashi K

    (2002) The ATPase activity and the functional domain of PotA, a component of the Spermidine-preferential uptake system inEscherichia coli

    Journal of Biological Chemistry 277:24212–24219.

    https://doi.org/10.1074/jbc.M202849200

    • Google Scholar
22. 1. Ke X
    2. Miller LC
    3. Bassler BL

    (2015) Determinants governing ligand specificity of the Vibrio harveyi LuxN quorum-sensing receptor

    Molecular Microbiology 95:127–142.

    https://doi.org/10.1111/mmi.12852

    • PubMed
    • Google Scholar
23. 1. Kitts G
    2. Giglio KM
    3. Zamorano-Sánchez D
    4. Park JH
    5. Townsley L
    6. Cooley RB
    7. Wucher BR
    8. Klose KE
    9. Nadell CD
    10. Yildiz FH
    11. Sondermann H

    (2019) A conserved regulatory circuit controls large adhesins in Vibrio cholerae

    mBio 10:e02822-19.

    https://doi.org/10.1128/mBio.02822-19

    • PubMed
    • Google Scholar
24. 1. Lee J
    2. Sperandio V
    3. Frantz DE
    4. Longgood J
    5. Camilli A
    6. Phillips MA
    7. Michael AJ

    (2009) An alternative polyamine biosynthetic pathway is widespread in Bacteria and essential for biofilm formation in Vibrio cholerae

    Journal of Biological Chemistry 284:9899–9907.

    https://doi.org/10.1074/jbc.M900110200

    • Google Scholar
25. 1. McEvoy FA
    2. Hartley CB

    (1975) Polyamines in cystic fibrosis

    Pediatric Research 9:721–724.

    https://doi.org/10.1203/00006450-197509000-00007

    • PubMed
    • Google Scholar
26. 1. McGinnis MW
    2. Parker ZM
    3. Walter NE
    4. Rutkovsky AC
    5. Cartaya-Marin C
    6. Karatan E

    (2009) Spermidine regulates Vibrio cholerae biofilm formation via transport and signaling pathways

    FEMS Microbiology Letters 299:166–174.

    https://doi.org/10.1111/j.1574-6968.2009.01744.x

    • PubMed
    • Google Scholar
27. 1. Michael AJ

    (2018) Polyamine function in archaea and Bacteria

    Journal of Biological Chemistry 293:18693–18701.

    https://doi.org/10.1074/jbc.TM118.005670

    • Google Scholar
28. 1. Miller-Fleming L
    2. Olin-Sandoval V
    3. Campbell K
    4. Ralser M

    (2015) Remaining mysteries of molecular biology: the role of polyamines in the cell

    Journal of Molecular Biology 427:3389–3406.

    https://doi.org/10.1016/j.jmb.2015.06.020

    • PubMed
    • Google Scholar
29. Book

    1. Morgan DML

    (1998)

    Determination of polyamines as their benzoylated derivatives by HPLC

    In: Morgan D. M. L, editors. Polyamine Protocols, Methods in Molecular Biology. Humana Press. pp. 111–118.

    • Google Scholar
30. 1. Nesse LL
    2. Berg K
    3. Vestby LK

    (2015) Effects of norspermidine and spermidine on biofilm formation by potentially pathogenic Escherichia coli and Salmonella enterica wild-type strains

    Applied and Environmental Microbiology 81:2226–2232.

    https://doi.org/10.1128/AEM.03518-14

    • PubMed
    • Google Scholar
31. 1. Newell PD
    2. Boyd CD
    3. Sondermann H
    4. O'Toole GA

    (2011) A c-di-GMP Effector System Controls Cell Adhesion by Inside-Out Signaling and Surface Protein Cleavage

    PLOS Biology 9:e1000587.

    https://doi.org/10.1371/journal.pbio.1000587

    • Google Scholar
32. 1. Osborne DL
    2. Seidel ER

    (1990) Gastrointestinal luminal polyamines: cellular accumulation and enterohepatic circulation

    American Journal of Physiology-Gastrointestinal and Liver Physiology 258:G576–G584.

    https://doi.org/10.1152/ajpgi.1990.258.4.G576

    • Google Scholar
33. 1. Papenfort K
    2. Bassler BL

    (2016) Quorum sensing signal–response systems in Gram-negative bacteria

    Nature Reviews Microbiology 14:576–588.

    https://doi.org/10.1038/nrmicro.2016.89

    • Google Scholar
34. 1. Parker ZM
    2. Pendergraft SS
    3. Sobieraj J
    4. McGinnis MM
    5. Karatan E

    (2012) Elevated levels of the norspermidine synthesis enzyme NspC enhance Vibrio cholerae biofilm formation without affecting intracellular norspermidine concentrations

    FEMS Microbiology Letters 329:18–27.

    https://doi.org/10.1111/j.1574-6968.2012.02498.x

    • PubMed
    • Google Scholar
35. 1. Patel CN
    2. Wortham BW
    3. Lines JL
    4. Fetherston JD
    5. Perry RD
    6. Oliveira MA

    (2006) Polyamines are essential for the formation of plague biofilm

    Journal of Bacteriology 188:2355–2363.

    https://doi.org/10.1128/JB.188.7.2355-2363.2006

    • PubMed
    • Google Scholar
36. 1. Pérez-Mendoza D
    2. Coulthurst SJ
    3. Humphris S
    4. Campbell E
    5. Welch M
    6. Toth IK
    7. Salmond GP

    (2011) A multi-repeat adhesin of the phytopathogen, Pectobacterium atrosepticum, is secreted by a type I pathway and is subject to complex regulation involving a non-canonical diguanylate cyclase

    Molecular Microbiology 82:719–733.

    https://doi.org/10.1111/j.1365-2958.2011.07849.x

    • PubMed
    • Google Scholar
37. 1. Puleston DJ
    2. Buck MD
    3. Klein Geltink RI
    4. Kyle RL
    5. Caputa G
    6. O'Sullivan D
    7. Cameron AM
    8. Castoldi A
    9. Musa Y
    10. Kabat AM
    11. Zhang Y
    12. Flachsmann LJ
    13. Field CS
    14. Patterson AE
    15. Scherer S
    16. Alfei F
    17. Baixauli F
    18. Austin SK
    19. Kelly B
    20. Matsushita M
    21. Curtis JD
    22. Grzes KM
    23. Villa M
    24. Corrado M
    25. Sanin DE
    26. Qiu J
    27. Pällman N
    28. Paz K
    29. Maccari ME
    30. Blazar BR
    31. Mittler G
    32. Buescher JM
    33. Zehn D
    34. Rospert S
    35. Pearce EJ
    36. Balabanov S
    37. Pearce EL

    (2019) Polyamines and eIF5A hypusination modulate mitochondrial respiration and macrophage activation

    Cell Metabolism 30:352–363.

    https://doi.org/10.1016/j.cmet.2019.05.003

    • PubMed
    • Google Scholar
38. 1. Römling U
    2. Galperin MY
    3. Gomelsky M

    (2013) Cyclic di-GMP: the first 25 years of a universal bacterial second messenger

    Microbiology and Molecular Biology Reviews 77:1–52.

    https://doi.org/10.1128/MMBR.00043-12

    • PubMed
    • Google Scholar
39. 1. Singh PK
    2. Bartalomej S
    3. Hartmann R
    4. Jeckel H
    5. Vidakovic L
    6. Nadell CD
    7. Drescher K

    (2017) Vibrio cholerae combines individual and collective sensing to trigger biofilm dispersal

    Current Biology 27:3359–3366.

    https://doi.org/10.1016/j.cub.2017.09.041

    • PubMed
    • Google Scholar
40. 1. Sobe RC
    2. Bond WG
    3. Wotanis CK
    4. Zayner JP
    5. Burriss MA
    6. Fernandez N
    7. Bruger EL
    8. Waters CM
    9. Neufeld HS
    10. Karatan E

    (2017) Spermine inhibits Vibrio cholerae biofilm formation through the NspS–MbaA polyamine signaling system

    Journal of Biological Chemistry 292:17025–17036.

    https://doi.org/10.1074/jbc.M117.801068

    • Google Scholar
41. 1. Tamayo R
    2. Patimalla B
    3. Camilli A

    (2010) Growth in a biofilm induces a hyperinfectious phenotype in Vibrio cholerae

    Infection and Immunity 78:3560–3569.

    https://doi.org/10.1128/IAI.00048-10

    • PubMed
    • Google Scholar
42. 1. Wotanis CK
    2. Brennan WP
    3. Angotti AD
    4. Villa EA
    5. Zayner JP
    6. Mozina AN
    7. Rutkovsky AC
    8. Sobe RC
    9. Bond WG
    10. Karatan E

    (2017) Relative contributions of norspermidine synthesis and signaling pathways to the regulation of Vibrio cholerae biofilm formation

    PLOS ONE 12:e0186291.

    https://doi.org/10.1371/journal.pone.0186291

    • PubMed
    • Google Scholar
43. 1. Yamamoto S
    2. Chowdhury MA
    3. Kuroda M
    4. Nakano T
    5. Koumoto Y
    6. Shinoda S

    (1991) Further study on polyamine compositions in Vibrionaceae

    Canadian Journal of Microbiology 37:148–153.

    https://doi.org/10.1139/m91-022

    • PubMed
    • Google Scholar
44. 1. Young EC
    2. Baumgartner JT
    3. Karatan E
    4. Kuhn ML

    (2021) A mutagenic screen reveals NspS residues important for regulation of Vibrio cholerae biofilm formation

    Microbiology 167:1023.

    https://doi.org/10.1099/mic.0.001023

    • Google Scholar
45. 1. Zamorano-Sánchez D
    2. Xian W
    3. Lee CK
    4. Salinas M
    5. Thongsomboon W
    6. Cegelski L
    7. Wong GCL
    8. Yildiz FH

    (2019) Functional specialization in Vibrio cholerae Diguanylate Cyclases: Distinct Modes of Motility Suppression and c-di-GMP Production

    mBio 10:e00670-19.

    https://doi.org/10.1128/mBio.00670-19

    • PubMed
    • Google Scholar
46. 1. Zhou H
    2. Zheng C
    3. Su J
    4. Chen B
    5. Fu Y
    6. Xie Y
    7. Tang Q
    8. Chou SH
    9. He J

    (2016) Characterization of a natural triple-tandem c-di-GMP riboswitch and application of the riboswitch-based dual-fluorescence reporter

    Scientific Reports 6:20871.

    https://doi.org/10.1038/srep20871

    • PubMed
    • Google Scholar

Acceptance summary:

This paper presents evidence that the polyamines spermidine and norspermidine can inversely impact the dispersal of Vibrio cholerae biofilms through the protein MbaA, which contains a periplasmic domain that may interface with a possibly direct polyamine sensor, NspS. The paper presents compelling data to support the notion that the cyclic-di-GMP synthetase and phosphodiesterase domains of MbaA are ultimately responsible for transducing polyamine signals in the periplasm into changes in biofilm formation. The work indicates that external polyamines may function to provide self/non-self information, similar to quorum sensing systems.

Decision letter after peer review:

Thank you for submitting your article "Inverse regulation of Vibrio cholerae biofilm dispersal by polyamine signals" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by Gisela Storz as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Anthony Michael (Reviewer #2); Mark Mandel (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

As the editors have judged that your manuscript is of interest, but as described below that additional experiments are required before it is published, we would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). First, because many researchers have temporarily lost access to the labs, we will give authors as much time as they need to submit revised manuscripts. We are also offering, if you choose, to post the manuscript to bioRxiv (if it is not already there) along with this decision letter and a formal designation that the manuscript is "in revision at eLife". Please let us know if you would like to pursue this option. (If your work is more suitable for medRxiv, you will need to post the preprint yourself, as the mechanisms for us to do so are still in development.)

Summary:

This paper presents evidence that the polyamines spermidine and norspermidine can inversely impact the dispersal of Vibrio cholerae biofilms through the protein MbaA, which contains a periplasmic domain that may interface with a possibly direct polyamine sensor, NspS. The paper is easy to follow and presents some compelling data to support the notion that the cyclic-di-GMP synthetase and phosphodiesterase domains of MbaA are ultimately responsible for transducing polyamine signals in the periplasm into changes in biofilm formation. The work indicates that external polyamines may function to provide self/non-self information, similar to quorum sensing systems. There were several major concerns raised, as summarized below. In particular, there was concern from all three reviewers about how much polyamine-based signaling contributes to the formation of biofilms by wild-type cells, especially the self-produced norspermidine. In short, there's currently confusion about whether polyamines only impact biofilms when added exogenously or whether the endogenously produced/secreted levels can and do impact biofilms. Measuring these polyamines in biofilm cultures will be essential to a revision.

Essential revisions:

Experimental issues:

The deletion mutants (e.g. in Figure 2) need to be complemented.

The evidence for the periplasmic sensing model in the ΔpotD1 mutant is strong, but could be further strengthened by an intervention that blocks the proposed norspermidine export. If there is a way to interrupt endogenous norspermidine synthesis and/or export to the periplasm, this would add further support for the model in Figure 5CD to explain the role of PotD1. Similarly, a PotD1 allele that is still present in the periplasm but is nonfunctional for import would provide evidence that the key issue is flux/import, and not sequestration for PotD1.

The fact that the mbaA(SGDEF)* mutant still accumulates biomass at the same rate up to 8 hr as the wild-type with the same max value suggests that this pathway doesn't become relevant until later in biofilm formation. But, spermidine at high levels can inhibit early on (Figure 3A). To me, these two findings aren't reconciled in the authors' model (Figure 4F) which would lead one to believe that CDG synthesis by MbaA is critical for biofilm formation, but it's not (see Figure 4A) – to be clear, I recognize that the paper's title and abstract focus on dispersal, but the model leads one to think that the pathway being studied here is important throughout the biofilm development process. Maybe more generally, the issue is about how spermidine and norspermidine impact the normal development of a wild-type biofilm, if at all. It's not revealed until the Discussion that norspermidine isn't likely secreted at appreciable levels (something that should definitely be noted in the Introduction), so while exogenous addition can impact biofilms, the lack of secreted norspermidine makes it sound like it's actually not so critical to our understanding of WT biofilms and the behavior of WT biofilms documented in Figure 2. Following on this point above, I have two questions that seem critical to assessing how polyamines impact biofilm development: (i) What's the level of spermidine produced and secreted? If the dispersal is normally driven by spermidine, its levels should accumulate extracellularly or periplasmically at the time biofilms being to disperse. (ii) What's the biofilm phenotype of a mutant that cannot synthesize spermidine or norspermidine or both? It could be that mutants that can't synthesize polyamines behave like the WT and that neither is detectable in the periplasm/secreted fraction – such results wouldn't negate or diminish the conclusion that these polyamines can impact biofilm formation when added exogenously (which the Discussion speculates on), but it would mean that the usual pattern of biofilm mass accumulation and then drop off noted in Figure 2B isn't driven by the polyamines. At the very least, the initial accumulation seems very unlikely to be driven by norspermidine if it can't be detected.

Following on the point above, the group of Karatan, using HPLC, has repeatedly failed to detect extracellular norspermidine in the spent growth medium of V. cholerae planktonic and biofilm cultures (refs. Wotanis et al., 2017 and uncited Parker et al., (2012) FEMS Microbiol Lett 329, 19-27). The current authors have used LC-MS/MS to determine that spent growth medium from planktonic potD1 gene deletion cultures contain 15 times more norspermidine than the parental wildtype strain. The authors need to provide a quantitative analysis of the absolute levels of norspermidine produced by biofilm cultures and assess whether it exceeds the minimum amount required to see effects when exogenously adding norspermidine.

The hypothesis of the authors – that self-produced norspermidine in the Δ-potD1 mutant accumulates in the periplasm and elicits biofilm effects due to its periplasmic concentration, is elegant but requires two suppositions: firstly, norspermidine is secreted by the cell into the periplasm, and that the periplasmic concentration of norspermidine must be above a threshold that results in elicitation of biofilm and dispersal effects. Norspermidine is presumably able to freely diffuse out of the periplasm into the external medium via porins. Given the time scales involved, the concentration of norspermidine would be likely to equilibrate either side of the outer membrane. Therefore, either the concentration of norspermidine in the external cell-free medium must be above a threshold level that elicits biofilm effects, as discussed above. Or, norspermidine must be sequestered in the periplasm. How might sequestration of the norspermidine in the periplasm occur? The Karatan group showed that the cell-free spent medium from V. cholerae biofilm cultures contained 2 mM cadaverine, whereas the sterile cell-free growth medium contained only 3 μM cadaverine. This indicates that the biofilm cells were undergoing acid stress, which induces the cadaverine-producing and exporting cadAB system (lysine decarboxylase and lysine/cadaverine antiporter) but more relevantly, endogenously-produced cadaverine has been shown to block outer membrane porin channels, even if produced under pH neutral conditions (Samartzidou and Delcour (1999) "Excretion of endogenous cadaverine leads to a decrease in porin-mediated outer membrane permeability", J. Bacteriol., 181, 791-798). Production and excretion of cadaverine by CadAB during biofilm formation could prevent self-produced norspermidine diffusing out of the periplasm. In the WT cells, any periplasmic norspermidine could be reacquired by the cell through the potABCD transporter, whereas in the Δ-potD1 mutant, a cadaverine block of porins might result in accumulation of periplasmic norspermidine. Questions to the authors: (1) What was the concentration of cadaverine in the WT and Δ-potD1 cell cultures and cell-free spent medium? (2) What was the concentration of putrescine, spermidine and N-acetylnorspermidine? Presumably this data is part of the original LC-MS/MS output. (3) The authors' polyamine data was obtained with planktonic cultures containing an additional deletion of the vps1 gene that eliminates the biofilm matrix. Did the authors measure norspermidine levels in cell-free spent medium from biofilm cultures? If not, it would be very relevant to the authors' hypothesis to measure norspermidine and cadaverine in the cell-free spent medium of WT and Δ-potD1 biofilm cultures (that are +vps1) and compare with sterile, cell-free medium. The blockage of porin permeability in the Samartzidou and Delcours study occurred when external cadaverine concentration had reached only 0.2 mM, whereas the Karatan group found 2 mM cadaverine in the external medium, suggesting that in the V. cholerae biofilm cells, cadaverine-blocked porins would prevent free diffusion of norspermidine out of the periplasm. Thus, cadaverine measurements would be very relevant to the authors' conclusions.

Scholarship issues:

A more thorough scholarly analysis of what is known about NspS (versus proposed here) would bolster the role for NspS in the authors' model. The current manuscript does not describe the key literature on NspS. In particular, the sensing of polyamines by MbaA through NspS-polyamine binding and direct binding of NspS to MbaA seems well-supported by literature that could be described.

The paper generally lacks context and does not discuss in adequate detail the general biological roles for polyamines (including cytoplasmically), what's already known about their roles in biofilm formation (in vibrios or other organisms like B. subtilis), or how they are synthesized. Without this sort of context, I think the paper will be difficult to access or appreciate for the non-expert/non-biofilm researcher.

It is surprising that the authors did not discuss and cite the Sobe et al., paper, "Spermine inhibits Vibrio cholerae biofilm formation through the NspS-Mba polyamine signaling system" (2017) J. Biol. Chem. 292, 17025-17036. This paper discusses the concept of self/non-self mediated by polyamines and makes a persuasive argument that spermine (a primarily eukaryotic polyamine) is a more likely signal than spermidine, especially for a human pathogen.

The authors did not measure cyclic-di-GMP directly and did not measure polyamine transport directly. It would therefore be prudent to make less categorically emphatic statements about eg., diguanylate cyclase activity line 339 "our data show that MbaA does synthesize c-di-GMP when norspermidine is present (Figure 3C, Figure 4C). In fact, c-di-GMP was not measured and the diguanylate cyclase activity of MbaA was not biochemically assayed. Similarly, for norspermidine uptake – in the abstract ("Biofilm dispersal fails in the absence of PotD1 because reuptake of endogenously produced norspermidine does not occur..") and line 109. At this stage, more tentative language would be appropriate.

Regarding self/non-self, I had difficulty integrating the discussion proposing that "self" molecules promote biofilm with a role for biofilm dispersal in cholera transmission (when, presumably, the "self" molecules would be at high abundance). I think a more sophisticated discussion of the issue here would be helpful, given that this is the lab that has defined the relevant QS signaling, has distinguished surface biofilms vs. liquid aggregates, etc.

Essential revisions:

Experimental issues:

The deletion mutants (e.g. in Figure 2) need to be complemented.

We thank the reviewers for pointing out this issue. As requested, we expressed mbaA and potD1 from an ectopic locus on the V. cholerae chromosome. Introduction of each gene in the corresponding deletion mutant restored the WT biofilm lifecycle behavior. These results are provided in new Figure 2—figure supplement 1.

The evidence for the periplasmic sensing model in the ΔpotD1 mutant is strong, but could be further strengthened by an intervention that blocks the proposed norspermidine export. If there is a way to interrupt endogenous norspermidine synthesis and/or export to the periplasm, this would add further support for the model in Figure 5CD to explain the role of PotD1. Similarly, a PotD1 allele that is still present in the periplasm but is nonfunctional for import would provide evidence that the key issue is flux/import, and not sequestration for PotD1.

We agree that the suggested experiments would confirm our model. Unfortunately, we do not know the identity of the exporter that transports norspermidine from the cytoplasm to the periplasm, so we cannot perform that test. However, we can disrupt norspermidine biosynthesis via deletion of nspC, which has the added benefit of eliminating norspermidine export to the periplasm. Consistent with our model (and as has been shown previously by other groups), the ΔnspC strain failed to form biofilms (shown in new Figure 5F; discussed in text lines 319-327). Furthermore, we made the ΔnspC ΔpotD1 double mutant and it also failed to form biofilms, demonstrating that norspermidine production is epistatic to norspermidine import (shown in new Figure 5F).

To distinguish between polyamine import and PotD1-mediated sequestration of norspermidine, we deleted potA. PotABCD is the transport apparatus responsible for spermidine/norspermidine import. The role of the PotA ATPase is to supply the energy required for transport. Thus, in the absence of PotA, PotD1 remains capable of binding polyamines, yet transport does not occur. The ΔpotA mutant exhibited the identical dispersal defect as the ΔpotD1 mutant validating that transport of norspermidine, not sequestration by PotD1, is the key activity required for biofilm dispersal. The results are reported in new Figure 5—figure supplement 1, along with text in lines 333-340.

The fact that the mbaA(SGDEF)* mutant still accumulates biomass at the same rate up to 8 hr as the wild-type with the same max value suggests that this pathway doesn't become relevant until later in biofilm formation. But, spermidine at high levels can inhibit early on (Figure 3A). To me, these two findings aren't reconciled in the authors' model (Figure 4F) which would lead one to believe that CDG synthesis by MbaA is critical for biofilm formation, but it's not (see Figure 4A) – to be clear, I recognize that the paper's title and abstract focus on dispersal, but the model leads one to think that the pathway being studied here is important throughout the biofilm development process. Maybe more generally, the issue is about how spermidine and norspermidine impact the normal development of a wild-type biofilm, if at all.

We thank the referees for this set of comments/suggestions. Here, we respond to them one by one for clarity:

We take the referee’s first question to be: what level of control does the MbaA signaling pathway exert over the biofilm lifecycle in WT V. cholerae in the absence of exogenously supplied polyamines? As the reviewer notes, in the case of the WT strain under conditions in which exogenous polyamines are not supplied, it appears, based on our results, that MbaA plays a minor role in driving progression of the biofilm lifecycle because the ∆mbaA mutant forms nearly WT biofilms and only exhibits a modest dispersal defect (Figure 2B). However, we have now analyzed and included results for the ΔnspC strain (discussed above) which is unable to synthesize and secrete norspermidine (Figure 5F). Due to the absence of endogenously-produced norspermidine, the ΔnspC strain does not form biofilms, as has also been demonstrated by other groups. The ΔnspC mutant phenotype shows that, in WT, when MbaA is present, periplasmic norspermidine is required to suppress the MbaA phosphodiesterase activity, which in turn, allows biofilms to form. Thus, in WT, the MbaA pathway has a major role as it is required for the normal biofilm lifecycle to take place.

It's not revealed until the Discussion that norspermidine isn't likely secreted at appreciable levels (something that should definitely be noted in the Introduction), so while exogenous addition can impact biofilms, the lack of secreted norspermidine makes it sound like it's actually not so critical to our understanding of WT biofilms and the behavior of WT biofilms documented in Figure 2.

As requested, we have clarified this point. We have included the following text in the Introduction (lines 92-96): “V. cholerae produces intracellular norspermidine, however, norspermidine has not been detected in cell-free culture fluids of laboratory grown strains (Parker et al., 2012). Thus, if norspermidine does indeed enable V. cholerae to take a census of ‘self’, it is apparently not via a canonical quorum-sensing type mechanism.”

In the discussion, we now write (lines 395-400), “The observation that laboratory grown WT V. cholerae releases little norspermidine, at least in part due to PotD1-mediated reuptake (Figure 5C), suggests that this system does not behave like a canonical quorum sensing pathway. However, it is possible that norspermidine is secreted by V. cholerae under some environmental conditions, or by other vibrios, and V. cholerae detects the released norspermidine via NspS-MbaA, and its dispersal from biofilms is prevented.”

Following on this point above, I have two questions that seem critical to assessing how polyamines impact biofilm development: (i) What's the level of spermidine produced and secreted? If the dispersal is normally driven by spermidine, its levels should accumulate extracellularly or periplasmically at the time biofilms being to disperse.

We appreciate this insight and suggestion. We performed the requested analyses. Please see lines 367-383 and the new Figure 6B. In brief, in the revised manuscript, we now provide quantitation of intracellular and extracellular spermidine and norspermidine concentrations at two timepoints – one prior to (5 h) and one after (10 h) biofilm dispersal. We find that, consistent with previous reports, V. cholerae makes almost no spermidine under all conditions tested. Furthermore, no substantial fluctuations in norspermidine levels (intracellular or extracellular) occurred between the timepoints that could be responsible for driving biofilm dispersal. We also considered the possibility that MbaA levels could change over the biofilm lifecycle, which could alter MbaA output activity. We therefore quantified MbaA-3xFLAG levels prior to (5 h) and after (10 h) biofilm dispersal (new Figure 6A). We found that MbaA levels remained constant. Our conclusion is that MbaA is poised to regulate the biofilm lifecycle in response to exogenous polyamines. Moreover, in the WT strain in the absence of environmentally-supplied polyamines, MbaA activity is constant and biased toward phosphodiesterase function throughout the biofilm lifecycle.

(ii) What's the biofilm phenotype of a mutant that cannot synthesize spermidine or norspermidine or both? It could be that mutants that can't synthesize polyamines behave like the WT and that neither is detectable in the periplasm/secreted fraction – such results wouldn't negate or diminish the conclusion that these polyamines can impact biofilm formation when added exogenously (which the Discussion speculates on), but it would mean that the usual pattern of biofilm mass accumulation and then drop off noted in Figure 2B isn't driven by the polyamines. At the very least, the initial accumulation seems very unlikely to be driven by norspermidine if it can't be detected.

We thank the referee for this suggestion. We have now assayed the biofilm lifecycle of the ∆nspC mutant that cannot synthesize norspermidine (new Figure 5F; discussed in lines 319-327). No biofilm formation occurs in this mutant demonstrating that, for normal biofilm development to occur, periplasmic norspermidine is required to suppress the MbaA phosphodiesterase activity (as discussed above in response to point 3). As noted above, we now also included a Results section demonstrating that changes in MbaA activity do not drive biofilm dispersal in the absence of exogenously supplied polyamines (new Figure 6; discussed in lines 367-383).

Following on the point above, the group of Karatan, using HPLC, has repeatedly failed to detect extracellular norspermidine in the spent growth medium of V. cholerae planktonic and biofilm cultures (refs. Wotanis et al., 2017 and uncited Parker et al., (2012) FEMS Microbiol Lett 329, 19-27). The current authors have used LC-MS/MS to determine that spent growth medium from planktonic potD1 gene deletion cultures contain 15 times more norspermidine than the parental wildtype strain. The authors need to provide a quantitative analysis of the absolute levels of norspermidine produced by biofilm cultures and assess whether it exceeds the minimum amount required to see effects when exogenously adding norspermidine.

We thank the reviewer for this suggestion. We have now repeated our measurements and we provide molar concentrations for all of the polyamines quantified by LC-MS/MS (see new Figure 5I, J and text lines 341-364). We write, “At high cell density, approximately 25-fold more norspermidine was present in cell-free culture fluids collected from the ΔpotD1 mutant (average 2.3 µM) than in those prepared from the WT (average 90 nM) (Figure 5I). […] There was no difference in norspermidine levels in whole cell extracts prepared from WT (average 0.6 µmol/g) and the ΔpotD1 mutant (average 0.5 µmol/g) (Figure 5J).”

Regarding our measurements from WT biofilm cultures, we now write, lines 376-379 “The concentration of extracellular norspermidine did increase between 5 h and 10 h, from <10 nM to ~75 nM (Figure 6B), however this range is far below the NspS-MbaA detection threshold (Figure 3C). Extracellular spermidine was nearly undetectable at both timepoints (Figure 6B).”

The hypothesis of the authors – that self-produced norspermidine in the Δ-potD1 mutant accumulates in the periplasm and elicits biofilm effects due to its periplasmic concentration, is elegant but requires two suppositions: firstly, norspermidine is secreted by the cell into the periplasm, and that the periplasmic concentration of norspermidine must be above a threshold that results in elicitation of biofilm and dispersal effects. Norspermidine is presumably able to freely diffuse out of the periplasm into the external medium via porins. Given the time scales involved, the concentration of norspermidine would be likely to equilibrate either side of the outer membrane. Therefore, either the concentration of norspermidine in the external cell-free medium must be above a threshold level that elicits biofilm effects, as discussed above. Or, norspermidine must be sequestered in the periplasm. How might sequestration of the norspermidine in the periplasm occur? The Karatan group showed that the cell-free spent medium from V. cholerae biofilm cultures contained 2 mM cadaverine, whereas the sterile cell-free growth medium contained only 3 μM cadaverine. This indicates that the biofilm cells were undergoing acid stress, which induces the cadaverine-producing and exporting cadAB system (lysine decarboxylase and lysine/cadaverine antiporter) but more relevantly, endogenously-produced cadaverine has been shown to block outer membrane porin channels, even if produced under pH neutral conditions (Samartzidou and Delcour (1999) "Excretion of endogenous cadaverine leads to a decrease in porin-mediated outer membrane permeability", J. Bacteriol., 181, 791-798). Production and excretion of cadaverine by CadAB during biofilm formation could prevent self-produced norspermidine diffusing out of the periplasm. In the WT cells, any periplasmic norspermidine could be reacquired by the cell through the potABCD transporter, whereas in the Δ-potD1 mutant, a cadaverine block of porins might result in accumulation of periplasmic norspermidine. Questions to the authors: (1) What was the concentration of cadaverine in the WT and Δ-potD1 cell cultures and cell-free spent medium? (2) What was the concentration of putrescine, spermidine and N-acetylnorspermidine? Presumably this data is part of the original LC-MS/MS output. (3) The authors' polyamine data was obtained with planktonic cultures containing an additional deletion of the vps1 gene that eliminates the biofilm matrix. Did the authors measure norspermidine levels in cell-free spent medium from biofilm cultures? If not, it would be very relevant to the authors' hypothesis to measure norspermidine and cadaverine in the cell-free spent medium of WT and Δ-potD1 biofilm cultures (that are +vps1) and compare with sterile, cell-free medium. The blockage of porin permeability in the Samartzidou and Delcours study occurred when external cadaverine concentration had reached only 0.2 mM, whereas the Karatan group found 2 mM cadaverine in the external medium, suggesting that in the V. cholerae biofilm cells, cadaverine-blocked porins would prevent free diffusion of norspermidine out of the periplasm. Thus, cadaverine measurements would be very relevant to the authors' conclusions.

We thank the referee for these insightful comments. To test the referee’s hypothesis that cadaverine blockage of porins causes norspermidine trapping in the periplasm, we performed several experiments. Our data are provided in Author response image 1. We find no evidence for cadaverine export in controlling extracellular norspermidine levels or in driving the biofilm lifecycle.

In the revised manuscript, we now provide quantitation of cadaverine, putrescine, spermidine, and norspermidine levels in all of our polyamine measurements. Those data are provided in new Figure 5 I, J; new Figure 5—figure supplement 2; new Figure 6B; and new Figure 6—figure supplement 1.

Regarding cadaverine: First, in WT planktonic, shaken cultures at high cell density, cadaverine was present at 0.5 µM in cell-free culture fluids (new Figure 5—figure supplement 2B; middle panel). In biofilm cells grown under static conditions, concentrations of both intracellular and extracellular cadaverine were low at the 5 h timepoint and dramatically increased at 10 h (new Figure 6—figure supplement 1). These results are consistent with those in the Karatan manuscript that demonstrated higher cadaverine levels for statically grown biofilm cells than for cells grown in shaken culture. Both sets of results suggest that, at late times, acid stress occurs in the static cultures, cadAB is induced, and cadaverine accumulates. Our temporal results, however, are inconsistent with the referee’s hypothesis that cadaverine blockage of porin channels leads to elevated norspermidine and subsequent changes in NspS-MbaA signaling. Specifically, cadaverine is in low abundance at the 5 h timepoint, when biofilm formation is favored. Cadaverine concentration increases by 10 h, after dispersal has occurred. This timing suggests that, if cadaverine had blocked porins, norspermidine would have accumulated in the periplasm by the 10 h timepoint, promoting biofilm formation, however, that does not occur. Rather, we observe biofilm dispersal.

We further examined the relationship between cadaverine and the biofilm lifecycle by constructing a ΔcadB mutant that lacks the lysine/cadaverine antiporter responsible for cadaverine secretion. The biofilm lifecycle of the ΔcadB mutant was identical to that of WT V. cholerae (see Author response image 1), showing that periplasmic/extracellular cadaverine does not alter norspermidine signaling in WT V. cholerae. Moreover, we made the ΔcadB ΔpotD1 double mutant and it behaved identically to the single ΔpotD1 mutant. Thus, the presence or absence of cadaverine does not alter the biofilm dispersal failure phenotype of the ΔpotD1 mutant, so the ΔpotD1 mutant phenotype cannot be due to blocking of porins by cadaverine. Together, our results show that secreted cadaverine does not impinge on the V. cholerae biofilm lifecycle, and therefore, cadaverine has no effect on periplasmic norspermidine levels nor on MbaA activity.

Because our results do not support any role for cadaverine in the V. cholerae biofilm lifecycle, we want to keep the message of our manuscript streamlined. Therefore we have not included an extended analysis of cadaverine in the current manuscript. As noted above, we do include our cadaverine measurements in the Results, and we review information on polyamines in general in the revised Introduction.

Author response image 1

Download asset Open asset

Scholarship issues:

A more thorough scholarly analysis of what is known about NspS (versus proposed here) would bolster the role for NspS in the authors' model. The current manuscript does not describe the key literature on NspS. In particular, the sensing of polyamines by MbaA through NspS-polyamine binding and direct binding of NspS to MbaA seems well-supported by literature that could be described.

We thank the reviewers for this suggestion. We have written a more in-depth passage in the revised Introduction concerning what is known about the MbaA-NspS signaling circuit. Please see lines 78-111.

The paper generally lacks context and does not discuss in adequate detail the general biological roles for polyamines (including cytoplasmically), what's already known about their roles in biofilm formation (in vibrios or other organisms like B. subtilis), or how they are synthesized. Without this sort of context, I think the paper will be difficult to access or appreciate for the non-expert/non-biofilm researcher.

We regret overlooking a basic introduction to polyamines and we agree including such text will broaden the scope of the manuscript. In response, we now describe general polyamine biology in the revised Introduction (lines 69-96).

It is surprising that the authors did not discuss and cite the Sobe et al., paper, "Spermine inhibits Vibrio cholerae biofilm formation through the NspS-Mba polyamine signaling system" (2017) J. Biol. Chem. 292, 17025-17036. This paper discusses the concept of self/non-self mediated by polyamines and makes a persuasive argument that spermine (a primarily eukaryotic polyamine) is a more likely signal than spermidine, especially for a human pathogen.

We thank the reviewer for catching this oversight. We have included a discussion the effect of spermine on V. cholerae biofilm formation in our revised Introduction (lines 85-91).

The authors did not measure cyclic-di-GMP directly and did not measure polyamine transport directly. It would therefore be prudent to make less categorically emphatic statements about eg., diguanylate cyclase activity line 339 "our data show that MbaA does synthesize c-di-GMP when norspermidine is present (Figure 3C, Figure 4C). In fact, c-di-GMP was not measured and the diguanylate cyclase activity of MbaA was not biochemically assayed. Similarly, for norspermidine uptake – in the abstract ("Biofilm dispersal fails in the absence of PotD1 because reuptake of endogenously produced norspermidine does not occur..") and line 109. At this stage, more tentative language would be appropriate.

We thank the reviewer for this comment. We have now toned down our language in these and other cases.

Regarding self/non-self, I had difficulty integrating the discussion proposing that "self" molecules promote biofilm with a role for biofilm dispersal in cholera transmission (when, presumably, the "self" molecules would be at high abundance). I think a more sophisticated discussion of the issue here would be helpful, given that this is the lab that has defined the relevant QS signaling, has distinguished surface biofilms vs. liquid aggregates, etc.

As requested, we have expanded the text on the topic of polyamines as self vs non-self-signals in the Introduction, Results, and Discussion sections, with a focus on comparing the polyamine system to traditional quorum-sensing circuits. For example, see lines 391-408 for our overarching ideas concerning how the NspS-MbaA system could function to distinguish “self” vs “other.” Additionally, see lines 430-444 where we compare and contrast the opposing MbaA enzymatic activities to those of the established quorum-sensing receptor, LuxN. Beyond those new passages, at present, we hesitate to speculate on the role of MbaA signaling in disease transmission absent any data on which to hang our ideas. To our knowledge, NspS-MbaA-mediated signaling has not yet been investigated in animal models of infection.

Author details

1. Andrew A Bridges

   1. Department of Molecular Biology, Princeton University, Princeton, United States
   2. The Howard Hughes Medical Institute, Chevy Chase, United States

   Contribution

   Conceptualization, Resources, Data curation, Formal analysis, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - reviewing and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8132-751X

2. Bonnie L Bassler

   1. Department of Molecular Biology, Princeton University, Princeton, United States
   2. The Howard Hughes Medical Institute, Chevy Chase, United States

   Contribution

   Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Writing - original draft, Project administration, Writing - review and editing

   For correspondence

   bbassler@princeton.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0043-746X

• 2,344

  Page views

• 365

  Downloads

• 12

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.