Matrix-trapped viruses can prevent invasion of bacterial biofilms by colonizing cells
- Department of Biological Sciences, Dartmouth College, United States
- Max Planck Institute for Terrestrial Microbiology, Germany
- Department of Physics, Philipps University Marburg, Germany
- Biozentrum, University of Basel, Switzerland
Abstract
Bacteriophages can be trapped in the matrix of bacterial biofilms, such that the cells inside them are protected. It is not known whether these phages are still infectious and whether they pose a threat to newly arriving bacteria. Here, we address these questions using Escherichia coli and its lytic phage T7. Prior work has demonstrated that T7 phages are bound in the outermost curli polymer layers of the E. coli biofilm matrix. We show that these phages do remain viable and can kill colonizing cells that are T7-susceptible. If cells colonize a resident biofilm before phages do, we find that they can still be killed by phage exposure if it occurs soon thereafter. However, if colonizing cells are present on the biofilm long enough before phage exposure, they gain phage protection via envelopment within curli-producing clusters of the resident biofilm cells.
Introduction
Bacteria and their bacteriophage predators, or phages, are found alongside each other in nearly every environment examined (Rodriguez-Brito et al., 2010; Zablocki et al., 2016; Clokie et al., 2011; Manrique et al., 2016; Correa et al., 2021). Phages inject their genomes into the cytoplasm of their hosts, and in the case of obligate lytic phages, immediately begin co-opting host resources to replicate. Eventually, host cells are lysed to release a new cohort of phage virions. Predatory pressure from phage attack drives bacterial evolution, diversification, and ultimately the community structure of many microbiomes (Blazanin and Turner, 2021; Harrison et al., 2013; Lenski and Levin, 1985; Abedon, 2008; Koskella and Brockhurst, 2014; Gómez and Buckling, 2011; Keen and Dantas, 2018). The mechanistic, ecological, and evolutionary features of phage-bacteria interactions have a deep history of study, including many seminal theoretical and experimental papers that have characterized the population and evolutionary dynamics of phage-bacteria interactions (Koskella and Brockhurst, 2014; Abedon, 2009; Chao et al., 1977; Susskind and Botstein, 1978). The traditional literature in this area mostly considers well-mixed culture conditions such as those in shaken liquids, which can reveal fundamental aspects of phage-bacteria interactions without spatial structure. However, in nature bacteria often reside in spatially constrained, surface-bound communities, or biofilms (Nadell et al., 2016; Hellweger et al., 2016; Flemming and Wingender, 2010; Flemming et al., 2016). The limited work that has focused on phage infection in biofilm environments has often found outcomes that differ substantially from those observed in mixed liquid environments (Hansen et al., 2019; Levin and Bull, 2004; Simmons et al., 2020; Davies et al., 2016; Abedon, 2016; Schrag and Mittler, 1996; Eriksen et al., 2018; Simmons et al., 2018; Chaudhry et al., 2018; Pires et al., 2021; Vidakovic et al., 2018).
A defining feature of biofilm populations is the presence of a self-secreted adhesive polymer matrix that modulates cell-cell and cell-surface interactions, in addition to influencing collective biofilm architecture (Nadell et al., 2016; Flemming and Wingender, 2010; Erskine et al., 2018; Teschler et al., 2015; Hartmann et al., 2019; Colvin et al., 2012). Several recent papers have demonstrated that the biofilm matrix can be central to phage-host coevolution. Pseudomonas fluorescens and Escherichia coli rapidly evolve mucoid colony phenotypes – which reflect increased and/or altered matrix secretion – when they are under phage attack (Scanlan and Buckling, 2012; Chaudhry et al., 2020). Curli fibers, a proteinaceous component of the E. coli matrix, can block biofilm-dwelling cells from T7 and T5 phage exposure (Vidakovic et al., 2018). In this case, phages can be seen directly enmeshed in the curli mesh without infecting biofilm cells unless the integrity of the curli layer is compromised (Simmons et al., 2020; Vidakovic et al., 2018). Recent papers by Darch et al., 2017, Díaz-Pascual et al., 2019, and Dunsing et al., 2019, respectively, suggest a similar pattern of matrix-dependent phage protection in Pseudomonas aeruginosa, Vibrio cholerae, and Pantoea stewartii. Earlier work from Kay et al., 2011 used plaque-forming unit (PFU) count assays to show that if phage-exposed biofilms are disassociated by a matrix-degrading agent, the phage titer in the surrounding media increases, suggesting that phages released from the matrix are potentially still infectious. It is not known, however, whether phages remain threatening to newly arriving bacteria while the phages are still embedded in the intact biofilm matrix; it could be, for example, that matrix-embedded phages are mostly degraded or trapped in configurations that render them unable to infect host cells. The answer to this question may have a significant impact on the processes of population assembly as cells encounter and attempt to colonize pre-existing biofilms.
Inspired by the findings above, we investigated the consequences of phage entrapment in the matrix for biofilm population assembly. If they remain infectious, these phages could pose a threat to new cells that attempt to colonize the biofilm surface. Here, we explored this possibility by studying how matrix-embedded phages influence the invasion of bacteria into pre-existing biofilm populations, and whether biofilm-invading cells can integrate into the existing matrix and gain phage protection. To address these questions, we used a combination of microfluidic culture, phage infection reporter techniques, high resolution confocal microscopy, and detailed spatial analysis of the resulting image data. We find that matrix-trapped phages do indeed remain infectious and reduce the ability of newly arriving cells to colonize existing biofilms, and that this effect is dependent on the relative timing of the arrival of phages and colonizing cells.
Results and discussion
Influence of matrix-trapped phages on invading planktonic cells
E. coli produces a variety of matrix components at different times during biofilm formation, including flagellar filaments, curli amyloid fibers, and polysaccharides such as colanic acid and cellulose (Vidakovic et al., 2018; Beloin et al., 2008; Barnhart and Chapman, 2006; Evans and Chapman, 2014; Hammar et al., 1996; Pesavento et al., 2008; Serra et al., 2013a; Serra et al., 2013b). Curli fibers are the single matrix component known to be essential for T7 phage protection in E. coli biofilms (Vidakovic et al., 2018) and are assembled by extracellular polymerization of CsgA monomers on an outer membrane baseplate comprised of CsgB (Evans and Chapman, 2014; Barnhart and Chapman, 2006). Curli proteins secreted by E. coli localize primarily in the upper half of the biofilm-dwelling population, and introduced phages become trapped in the outer part of this curli matrix layer (Vidakovic et al., 2018; Serra et al., 2013a; Serra et al., 2013b; Figure 1A). The biofilm-dwelling cells amongst and below the curli fiber matrix layer thus become protected against T7 exposure in the absence of the evolution of physiological resistance to T7 phages (Simmons et al., 2020; Vidakovic et al., 2018). Although these phages are blocked from diffusing into the biofilm interior, we hypothesized that trapped phage particles may remain capable of infecting cells that reach them from the liquid environment that surrounds the biofilm. To test this hypothesis, we first grew biofilms of E. coli AR3110 for 62 hr (Figure 1B; Figure 1—figure supplement 1); prior work has established this growth period as optimal for full curli expression along the biofilm front (Vidakovic et al., 2018), and we confirmed this result in our microfluidic growth conditions (Figure 1—figure supplement 2). The resident cells harbored a chromosomal construct for constitutive expression of the far-red fluorescent protein mKate2 (Shcherbo et al., 2007) to make them visible with fluorescence time-lapse microscopy. The biofilms were then exposed to a 1 hr pulse of media containing lytic T7 phages at 2x108 PFU/mL, such that phages could accumulate in the outer matrix curli layer (Figure 1C). The phages contained a sfGFP (Pédelacq et al., 2006) expression construct engineered into their genome, allowing us to visualize any cells that became phage-infected (Vidakovic et al., 2018). For a control comparison, we performed the same procedure but exposed the resident biofilms to sterile media containing no phages. Following phage exposure or control treatment, we performed a population invasion step in which high-density planktonic cultures of phage-susceptible wild type E. coli AR3110 (OD600 = 6.0) were added to the chambers for 3 hr to colonize the resident biofilms (Figure 1D). The full experimental procedure for these experiments is summarized in Figure 1—figure supplement 1. Colonizing cells were isogenic to resident wild-type cells, except they constitutively expressed mKO-κ (Tsutsui et al., 2008) so that they could be distinguished from resident cells during imaging.
Figure 1 with 2 supplements see all
Visualization and quantification of biofilm invasion with or without phage exposure.
(A) Visualization of E. coli biofilm (red) with stained curli matrix (white), including one x-y optical section (main image) and z-projection (inset). (B–D) Illustration of invasion assay procedure. (B) Resident biofilm-producing cells (red) were allowed to grow for 62 hr prior to phage exposure. (C) Inlet tubing was then swapped for 1 hr to new tubing and syringe containing a concentrated phage suspension (2x108 PFU/mL). (D) Resident biofilms were then challenged with isogenic E. coli expressing a different fluorescent protein (yellow); this was performed by swapping for 3 hr to new tubing and syringe containing high density E. coli culture (OD600 = 6.0). Biofilms were imaged 10 hr following this step. (E) Invading cells (yellow) can successfully attach to the periphery of resident biofilm (red) in the absence of phages. (F) Invading cells fail to colonize when biofilms are pretreated with phages, which become trapped in the biofilm matrix. The phage-encoded reporter (cyan) indicates invading cells that have become phage-infected. (G,H) An E. coli mutant that does not permit phage amplification (ΔtrxA, magenta) invades equally well in control and phage pre-treatment conditions. (I) Quantification of image data shown in E-H; average invading biovolume per field of view (150 µm x 150 µm x 15 μm; length x width x height). Error bars represent SEM. Pairwise comparisons were performed with Mann-Whitney Signed Ranks tests with Bonferroni correction (n = 4-6 biological replicates; ** denotes p<0.05). (J–K) Invading cell cluster size distributions for phage-susceptible cells invading biofilms without (J) or with (K) phage pre-treatment (n = 4-6 biological replicates).
We found that in the absence of phage exposure, resident biofilms could be colonized by planktonic cells, albeit not at high efficiency (Figure 1E,I). The colonizing cells were restricted to the outer surface of the resident biofilms and could not enter the biofilm interior, similar to what we have seen previously for V. cholerae (Nadell et al., 2015). However, when resident biofilms were pre-exposed to phages, colonization by phage-susceptible cells was almost completely eliminated (Figure 1F,I). Further, most invading cells that were detected on phage-exposed biofilms were fluorescent in the sfGFP channel, indicating that they had been phage-infected but not yet lysed.
Our interpretation of this result is that susceptible invading cells encounter phages in the curli matrix, become infected, and lyse to release new phages, or fail to divide further. We assessed this idea quantitatively by measuring the cell cluster size distributions of invading cells on biofilms with or without phage-pre-exposure. Without pre-exposure, we found numerous instances of cell clusters with 20 cells or more, indicating several rounds of division in the 10 hr between invasion and imaging (Figure 1J). On biofilms exposed to phages prior to invasion of new E. coli cells, there were almost no groups larger than one or a few cells (Figure 1K), indicating few or no divisions in the period between invasion and imaging. This is consistent with the observation noted above that on phage pre-exposed biofilms, any remaining invading cells were expressing the phage infection reporter, which would preclude further growth and division.
Another (though not mutually exclusive) explanation for our results is that phages, by occupying potential sites of attachment, block the physical interaction of invading cells with the biofilm outer surface. We tested this possibility by repeating the experiment above with invading cultures of an E. coli mutant harboring a clean deletion of trxA; this strain does not allow for phage amplification (Qimron et al., 2006). The ΔtrxA deletion mutant was found to invade resident biofilms at equal rates whether or not the resident was pre-exposed to T7 phages, which suggests that phages do not block attachment sites for invading cells (Figure 1G,H,I). Noting that the ΔtrxA deletion mutant undergoes abortive infection upon exposure to phage T7, the fact that ΔtrxA can colonize resident biofilms, while WT E. coli cannot, indicates that phage amplification from initial sites of infection may be important for elimination of phage-susceptible invading cells.
Phages that are trapped in the curli matrix of E. coli biofilms are thus persistent as a threat to incoming susceptible bacteria. This result also implies that in the event that curli-protected cells within resident biofilms disperse individually or en masse, they may too be susceptible to phages that are released from the biofilm exterior. This remains an important question for future work.
Invading cells gain phage protection, after a delay
Once we determined that matrix-embedded phages could infect recently attached susceptible cells, we asked whether invading cells fare better if they arrive prior to the phage pulse. To explore this question, we again grew resident biofilms of E. coli AR3110 for 62 hr, followed by a planktonic population invasion step as described above. Chambers were separated into two groups: In the first group, phages were pulsed into the chamber immediately after colonization by the invading cell population. In the second group, biofilms were incubated post-invasion for 10 hr – the time at which we observed the colonized biofilms to reach a population steady-state – and then subjected to a phage pulse. This experimental procedure is summarized in Figure 2—figure supplement 1. Images were taken of each chamber approximately 10 hr after the phage pulse, allowing for multiple infection and lysis cycles to occur and the system to reach its new equilibrium. When phage pulses occurred immediately after biofilms were colonized by an invading strain, the invading cells were mostly killed by T7 phage exposure (Figure 2A,B). Interestingly, however, invading cells were not significantly affected by phage pulses that arrived 10 hr after the colonizing strain attached to the resident biofilm outer periphery (Figure 2A,C).
Figure 2 with 1 supplement see all
Visualization and quantification of colonization success with phage exposure post-colonization.
(A) Average invading biovolume per field of view (150 µm x 150 µm x 15 μm; length x width x height). Error bars represent SEM. Pairwise comparisons were performed with a Mann-Whitney Signed Ranks test (n = 7-11 biological replicates; **** denotes p<0.00005). (B) Invading cells (yellow, though absent in B) are killed when phages are introduced immediately after their arrival. Resident biofilm cells are shown in red. (C) Invading cells are not killed when phages are introduced 10 hr after their arrival.
Invading cells indirectly co-opt matrix of resident biofilms
Having observed that invading cells gain phage protection, but only after a delay between attaching to a resident biofilm and phage exposure, we next investigated how invading cells gain phage protection over this delay period. Given that phage protection is dependent on being embedded in curli polymers of the biofilm matrix (Vidakovic et al., 2018), it is possible that colonizing cells produce their own curli after an initial transition period following attachment to the resident biofilm. Alternatively, invading cells could co-opt matrix produced by the resident population. We tested these possibilities using a combination of experiments in which either the invading strain or the resident strain produced a 6xHis-labeled variant of CsgA, allowing us to localize and quantify curli production as a function of time and space inside biofilms via immunostaining. His-labeling of CsgA has been shown previously not to influence curli function in live biofilms (Vidakovic et al., 2018; Serra et al., 2013b). We cultivated resident biofilms and performed the invasion assay as above, but in this case we included an anti-His, AlexaFluor-647-conjugated antibody in the inflowing medium such that any curli produced in the biofilm became fluorescent and detectable by confocal microscopy (Figure 3A).
Figure 3 with 2 supplements see all
Spatial and temporal dynamics of curli fiber localization around resident and invading cells.
(A) Illustration of matrix quantification method. Localized curli matrix was quantified by measuring CsgA-His immunofluorescence in 3 µm shells surrounding individual segmented cell volumes. (B) Quantification of matrix localization surrounding the resident and invading strain populations during an 11 hr time course after invading cells arrive (n = 3-4 biological replicates per time point, errors bars denote SEM). (C–F) Representative images displaying resident cells (red) producing labeled curli matrix (white); invading cells are shown in yellow. Images in (C-F) were acquired respectively at 1 , 5, 7, and 10 hr following the arrival of invading cells.
When invading cells harbored the His-labeled variant of csgA, we detected negligible anti-His fluorescence in the 10 hr following invasion of the biofilm exterior (Figure 3—figure supplement 1). It was not clear why colonizing cells do not more rapidly produce their own curli, despite occupying conditions presumably identical to those experienced by resident cells on the biofilm surface; we speculate that the progression of the curli expression program must be quite slow relative to the time scale of surface occupation of the invading strain in our experiments. This could be the case, for example, if curli production were dependent on cell-cell packing that emerges gradually over the course of biofilm growth. By contrast, when cells in the resident biofilm harbored the His-labeled variant of csgA, abundant anti-His staining was detectable in the upper surface of the resident biofilm (Figure 3B–F), including the surroundings of the colonizing cells (Vidakovic et al., 2018; Serra et al., 2013a; Serra et al., 2013b). These two observations indicate that invading cells do not gain phage protection in the 10 hr following colonization because they produce their own curli fibers, but rather because they coopt the curli fibers being produced by cells in the resident biofilm (Figure 3B). To support this interpretation, we used image analysis to quantify curli accumulation in the immediate neighborhood of resident and colonizing cell biomass following invasion; this analysis showed a steadily increasing amount of curli in the immediate neighborhood of colonizing cells over the course of 10 hr following their arrival to the outer surface of the resident biofilm. At the 10 hr mark, invading cells were surrounded by the same amount of curli as the resident strain at the beginning of the invasion experiment.
We next asked if the invading strain directly or indirectly co-opts curli matrix material from the resident biofilm. Invading cells producing CsgB baseplates could potentially collect freely diffusing CsgA monomers being released by the resident biofilm population, accumulating curli matrix material and thus phage protection. This might not necessarily be the case, as our experiments above showed that the invading cells were not producing CsgA, whose corresponding gene is in the same operon as csgB; one would typically not expect the production of one without the other. Alternatively, without directly sequestering resident-produced CsgA to their cell surface, invading cells could become enveloped within curli-producing clusters of the resident biofilm population with enough curli in their surroundings to block phage diffusion. This would be an indirect way of exploiting the phage protection of the resident biofilm’s curli layer.
To differentiate between these two possibilities, we repeated the post-invasion phage pulse experiments using a ΔcsgB mutant that cannot produce its own CsgB base plate, and which therefore cannot nucleate CsgA polymerization on its outer surface. Here, as above (see Figure 2), few cells survived if phages arrived immediately after invading cells colonized the resident biofilm surface (Figure 3—figure supplement 2). However, despite lacking CsgB curli baseplates, invaders were still protected after a 10-hr delay prior to phage exposure in the system (Figure 3—figure supplement 2). This prompted the interpretation that invading cells do not have to be able to directly polymerize CsgA on their exterior to gain phage protection over the 10 hr after attaching to the biofilm surface. Rather this result suggests that the invading cells – despite remaining distinct physiologically from the resident cells by virtue of not producing CsgA or CsgB themselves – indirectly coopt curli produced by the resident biofilm by becoming sufficiently enveloped amidst clusters of resident cells that their exposure to incoming phages is greatly reduced or eliminated.
Conclusions
We have shown that matrix-embedded T7 phages can remain infectious on the curli-protected E. coli biofilm surface and kill newly arriving susceptible bacteria. In this sense, biofilm-dwelling microbes, by trapping phages on the biofilm periphery, can incidentally weaponize them against incoming phage-susceptible cells. On the other hand, we found that if invading E. coli cells attach to a resident biofilm and have sufficient time to become entangled in the curli matrix, they too gain protection from subsequent phage exposure. This protection is obtained by indirect exploitation of the resident biofilm’s curli matrix: invading cells did not significantly use resident-produced curli monomers to polymerize curli on their own exterior, but rather became sufficiently enveloped in curli-producing groups of resident biofilm cells that they were no longer exposed to an incoming phage attack.
Our observations bear an interesting analogy to those of Barr et al., 2013, who found that phages trapped in host mucosal linings can kill incoming bacteria (see also Barr et al., 2015). We speculate on the basis of our results here that phage entrapment and their blocking effect against bacterial colonization is important not just in host associated mucosal environments but even more broadly to many biofilm contexts in which phage-trapping matrix material could potentially influence the pattern of community succession. We would not say that obligate lytic phages like T7 evolved to serve this function per se; their evolutionary interest is to infect host cells, replicate, and spread to new hosts. But the biofilm matrix almost certainly did evolve in part to protect the cells within from external threats, including exposure to phages; when phages are trapped in the matrix and remain viable, they can subsequently influence whether other incoming bacteria (depending on their phage susceptibility) colonize the resident biofilm surface.
As a demonstration of principle, we studied here the effect of matrix-embedded T7 phages on the colonization ability of cells isogenic to those in the resident E. coli biofilm, showing that biofilm colonization by susceptible cells was effectively eliminated by the presence of phages on the biofilm surface. This does not constitute a proof of generality that matrix-trapped phages always block colonization by susceptible bacteria, but we would think that this phenomenon is not unique to E. coli. We hope our report will prompt further tests for other bacterial and phage species; recent reports from other groups have suggested matrix-dependent protection against and potentially sequestration of incoming phages, but it remains to be seen whether phages remain active against other incoming cells in these cases (Darch et al., 2017; Díaz-Pascual et al., 2019; Dunsing et al., 2019). The centrally important and open question for future work prompted by our results is whether multi-species biofilm consortia trap multiple phage types of different strain and species specificities, and whether some or all of these trapped phages have the potential to kill off invading cells of their target host range. In this regard, the spatial ecology of biofilm-phage interaction may play a key role in the successional dynamics of polymicrobial biofilm communities.
Materials and methods
Strains
E. coli strains used in these experiments were all derived from E. coli AR3110 and are listed in Table 1 below. Each strain used for imaging contained a codon-optimized fluorescent protein construct (mKO-κ, mKate2 or mTFP1) under the control of a constitutive Ptaq promoter. Fluorescent protein constructs were inserted using the traditional lambda-red recombineering method by amplifying constructs with primer overhangs corresponding to the attB site (Datsenko and Wanner, 2000). iProof High-Fidelity DNA Polymerase (Bio-Rad, Hercules, CA, USA) was used to amplify insertion sequences for fluorescent protein expression constructs. The E. coli ΔtrxA deletion strain was also constructed using lambda-red recombineering. The deletion construct was made by amplifying a kanamycin resistance marker flanked by FRT sites. Primer overhangs added upstream and downstream regions flanking start and stop codons of the trxA locus for replacement of the full reading frame with the Kan cassette. FRT recombinase was subsequently expressed in trans to remove the kanamycin resistance marker.
Table 1
Strains, plasmids, and oligos used in this study.
| Strain | Relevant markers/Genotype | Source | ||
|---|---|---|---|---|
| E. coli | ||||
| CNE 689 | AR3110 wild type Ptac-mKo-κ inserted at attB site | This study | ||
| CNE 762 | AR3110 wild type Ptac-mKate2 inserted at attB site | This study | ||
| CNE 760 | AR3110 wild type Ptac-mTFP inserted at attB site | This study | ||
| CNE 284 | AR3110 ΔcsgB::scar, Ptac-mruby inserted at attB site | (19) | ||
| CNE 773 | AR3110 CsgA with C-terminal 6x His Tag, Ptac-mKo-κ inserted at attB site | This study, (19) | ||
| CNE 691 | AR3110 ΔtrxA::scar, Ptac-mKo-κ inserted at attB site | This study | ||
| CNE 198 | AR3110 wild type | (19) | ||
| T7 Phage | ||||
| CNX 06 | T7 with sfgfp under phi 10 promotor control | (19) | ||
| Plasmid | Origin, marker | Comments | Templates, primers | Source |
| pCN754 | pR6Kγ, Kan | Housing plasmid/template for Ptac-mTFP insert | CNO 198, CNO 199, pCN 752, pCN 664 | This study |
| pCN755 | pR6Kγ, Kan | Housing plasmid/template for Ptac-mKate2 insert | CNO 198, CNO 199, pCN 753, pCN 664 | This study |
| pCN664 | pR6Kγ, Kan | Housing plasmid/template for Ptac-mKo-κ insert | CNO 198 CNO 199 | This study |
| pNUT1336 | pR6Kγ,Kan | pUC housing IDT-synthesized double mKO optimized for E. coli AR3110 | This study | |
| pCN753 | pR6Kγ, Kan | pUC housing IDT-synthesized double mTFP optimized for E. coli AR3110 | This study | |
| pCN752 | pR6Kγ, Kan | pUC housing IDT-synthesized double mKate2 optimized for E. coli AR3110 | This study | |
| Primer name | Sequence (designed using Snapgene) | Description | ||
| CNO 198 | ACAACTTTTTGTCTTTTTACCTTCCCGTTTCGCTCAAGT TAGTATttgacaattaatcatcggctcg | Universal primer to amplify new E. coli Ptac_FP construct, with attB integration tail | ||
| CNO 199 | TCCGGGCTATGAAATAGAAAAATGAATCCGTTGAAGCC TGCTTTTcatgggaattagccatggtcc | Universal primer to amplify new E. coli Ptac_FP construct, with attB integration tail | ||
| CNO 138 | ACAACGAAACCAACACGCCAGGCTTA TTCCTGTGGAGTTATATgtgtaggctggagctgcttc | Forward primer to amplify pKD3 FRT-Cm-FRT with homology to trxA up flank | ||
| CNO 139 | GCGTCCAGTTTTTAGCGACGGGGCACCCGAACATG AAATTCCCCcatatgaatatcctccttagt | Reverse primer to amplify pKD3 FRT-Cm-FRT with homology to trxA down flank | ||
| CNO 146 | GAATGGGCGTACAGTTATGAAAC | Forward primer to check sequence of trxA deletion in E. coli, 200 bp upstream | ||
| CNO 147 | TGCCTGGTCACAGGAGAGT | Reverse primer to check sequence of trxA deletion in E. coli, 200 bp downstrm | ||
| CNO 179 | GTGGATTGGGAACCGAGCA | Sequencing primer for pCN 664 | ||
| CNO 180 | GGAGATCCCAGACTACTTCAAAC | Sequencing primer for pCN 664 | ||
| CNO 181 | gtcaagaccgacctgtcc | Sequencing primer for pCN 664 | ||
| CNO 182 | ggacatagcgttggctacc | Sequencing primer for pCN 664 | ||
| CNO 183 | caccaatttcatattgctgtaagtg | Sequencing primer for pCN 664 | ||
| CNO 223 | TCTTTCACTTCCAGGTTAATGGTG | Sequencing primer for pCN 754 | ||
| CNO 224 | CGTTTGTGATTGAAGGCGAAG | Sequencing primer for pCN 754 | ||
| CNO 225 | GTGTATGAAAGCGCGGTGG | Sequencing primer for pCN 754 | ||
| CNO 226 | GGAACGTATGTACGTTCGTGAC | Sequencing primer for pCN 754 | ||
| CNO 229 | CATAATCCACCTCCTTTACTGGTC | Sequencing primer for pCN 755 | ||
| CNO 230 | AAACCGTATGAAGGCACCC | Sequencing primer for pCN 755 | ||
| CNO 231 | AAAGAAACCTATGTGGAACAGCAT | Sequencing primer for pCN 755 | ||
| CNO 232 | CGAATGGTCCGGTTATGC | Sequencing primer for pCN 755 | ||
| CNO 173 | TTTGGATCCTCTAAGCTTCATcctag | Forward primer for amplification of IDT 2x-mKO E. coli optimized (kde1336 template) | ||
| CNO 174 | ctccagcctacactttGAATTCtttTCTAGAAAGGAGCTCatg | Reverse primer for kde1336 template with overlap to FRT-Kan from pKD4 | ||
| CNO 175 | GAATTCaaagtgtaggctggagctgc | Forward primer for FRT-Kan amplification from pKD4 template with overlap to IDT 2x-mKO from pNUT1336 | ||
| CNO 176 | ggaagaaatagcgcatgggaattagccatggtcc | Reverse primer for FRT-Kan from pKD4, with overlap to pSC101 from kde970 | ||
| CNO 177 | ctaattcccatgcgctatttcttccagaattgc | Forward primer for pSC101 from kde970 with overlap to FRT_Kan from pKD4 | ||
| CNO 178 | aaaGGATCCattggtgagaatccaagcactag | Reverse primer for amplifying pSC01 from kde970, with BamHI site for ligating to upstream fragment from pNUT1336 | ||
| CNO 189 | aaaGGATCCttgacaattaatcatcggctcgtataatgcctaggc CTAAGCTTCATcctaggGACAc | introduction of Ptac (no lacO) into new FP construct vector for E. coli | ||
| CNO 190 | GCTCGCGGTAATTTTTTCGG | for sequencing of Ptac inserted with CNO 189 |
Phage propagation and titer
Request a detailed protocolT7 lytic phages were propagated and lysates collected in a manner adapted from Bonilla et al., 2016. Briefly, AR3110 wild type E. coli cultures were grown overnight in 5 mL lysogeny broth at 37°C at 250 rpm in a New Brunswick orbital shaking incubator. The host strain was then back diluted 1:20 into 100 mL lysogeny broth and allowed to grow to mid exponential phase (0.3–0.5 OD600). At this time, T7 phages were spiked in from a frozen stock and MgSO4 was added to a final concentration of 5 mM. The culture was placed back into the incubator for 3–4 hr, until the culture clarified. The entire volume was then vacuum filtered (0.22 µm filter Millipore Sigma). Phage titer was determined by traditional plaque assay (Adams, 1959). Briefly, host E. coli was grown overnight and sub-cultured as described above to achieve mid exponential phase (0.3–0.5 OD600). Phage preparation was serially diluted by passing 10 µL into 990 µL for 100-fold dilutions. Top agar (0.5% agar, lysogeny broth) was melted and aliquoted into 3 mL volumes. Subsequently, 50 µL of a dilution was added to each sample along with MgSO4 (5 mM). Molten top agar was then poured evenly onto lysogeny broth plates and placed at 37°C for 3 hr. Plates were removed and plaques were counted in order to calculate plaque forming units (PFU) per milliliter.
Biofilm phage pretreatment invasion assay
Request a detailed protocolWe measured attachment and growth of exogenously added planktonic cells to curli protected biofilms with and without the prior addition of T7 phage. E. coli AR3110 expressing mKate2 was cultured in 5 mL lysogeny broth overnight at 37°C at 250 rpm in a New Brunswick orbital shaking incubator. E. coli AR3110 was used due to its strong biofilm formation ability relative to other K12 domesticated lab strains of E. coli, and the literature history of establishing the timing and components of matrix expression in this strain background (Simmons et al., 2020; Vidakovic et al., 2018; Serra et al., 2013a; Serra et al., 2013b; Serra and Hengge, 2014). Cultures were then pelleted and washed twice with 0.9% NaCl and standardized to OD600=0.2 prior to inoculation into microfluidic devices. The inoculum was incubated for 1 hr at room temperature (approximately 22°C). Media syringes (1 mL BD plastic) were prepared by loading 1 mL of 1% tryptone broth (W/V) and attaching a 25-gauge needle. Tubing (#30 Cole palmer PTFE ID 0.3 mm) was then carefully attached to the needle and syringes were subsequently placed in Harvard Apparatus syringe pumps. After affixing inlet and outlet tubing to the microfluidic devices, a 40 s pulse at 40 µL/min was conducted to remove unattached cells, before standard flow regime (0.1 µL/min) was started. Biofilms were grown at room temperature for 62 hr at which time, tubing was swapped from clean media to either purified phages (2x108 PFU/mL) or clean media control. Flow was continued at 0.1 µL/min for 1 hr. After phage pretreatment, tubing was again removed and switched to syringes containing invading cells expressing mKO-κ for three hours at 0.1 µL/min. Invading cells were prepared prior to this step. Cells were grown overnight as previously described before in lysogeny broth. Invading cells were then sub-cultured 1:20 into 100 mL of 1% tryptone broth for 3 hr at 37°C at 250 rpm. Cells were then pelleted, concentrated, and standardized to OD600=6.0 (~5x109 CFU/mL). Following the conclusion of the invasion, microfluidic chambers were allotted 10 hr of incubation at room temperature under standard flow conditions for growth and phage infection to occur. Three to five image fields (150 µm x 150 µm x 25 µm; length x width x height, slice interval 0.5 µm) were then acquired on a Zeiss 880 LSCM and image analysis performed using BiofilmQ software tool (Hartmann et al., 2021).
Biofilm phage posttreatment invasion assay
Request a detailed protocolAttachment and growth of invading E. coli were measured under two different regimes. Invasions of resident biofilms were conducted prior to phage application in this assay; however, timing of the phage application was varied. Resident biofilms were cultured in the same manner described above in the pretreatment invasion assay. At 62 hr of growth, clean media tubing was exchanged for invading cells (prepared in an identical manner as above) and allowed to flow for 3 hr at 0.1 µL/min. Following the conclusion of the invasion, chambers were separated into two groups corresponding to phage treatment regime. Half of the chambers were exposed to phage treatment (2x108 PFU/mL at 0.1 µL/min for 1 hr) immediately following the invasion, while the other half was incubated for 10 hr at room temperature prior to the phage treatment. In both regimes, images were acquired in the manner described above. Imaging took place 10 hr following their respective phage treatments.
Curli matrix localization assay
Request a detailed protocolCurli matrix monomers, CsgA, were labeled with a 6X-His tag as previously published (Vidakovic et al., 2018). Curli fibers were detected via direct fluorescent immunostaining with an α−6X-His antibody (Invitrogen) conjugated to a fluorescent dye, Alexafluor647. Antibody was added to clean media at 0.1 mg/mL and flowed continuously throughout the course of the experiment. Biofilms were grown as described above, using E. coli expressing labeled curli.
Confocal microscopy and image analysis
Request a detailed protocolAll imaging was performed using a Zeiss 880 line-scanning confocal microscope with either a 10x/0.4NA or a 40x/1.2NA water objective to minimize axial aberration effects. Representative images for each experiment (with n independent replicates as indicated in each figure legend) were taken at random locations throughout the corresponding microfluidic devices. The sfGFP fluorescent protein was excited using a 488 nm laser line, the mKO-κ fluorescent protein was excited using a 543 nm laser line, the mKate2 fluorescent protein was excited using a 594 nm laser line, and Alexafluor647 was excited using a 633 nm laser line. All image stacks were trimmed if necessary (e.g. if area outside of the microfluidic devices had been acquired in addition to the biofilm itself) using the native Zeiss Zen Blue software. All subsequent quantifications were performed using the BiofilmQ image analysis framework (Hartmann et al., 2021).
Data availability
Raw data for the entire study has been provided in the source data files.
References
-
BookBacteriophage Ecology: Population Growth, Evolution, and Impact of Bacterial VirusesCambridge University Press.https://doi.org/10.1017/CBO9780511541483
-
Phage evolution and ecologyAdvances in Applied Microbiology 67:1–45.https://doi.org/10.1016/S0065-2164(08)01001-0
-
Bacteriophage exploitation of bacterial biofilms: phage preference for less mature targets?FEMS Microbiology Letters 363:fnv246.https://doi.org/10.1093/femsle/fnv246
-
Curli biogenesis and functionAnnual Review of Microbiology 60:131–147.https://doi.org/10.1146/annurev.micro.60.080805.142106
-
Escherichia coli biofilmsCurrent Topics in Microbiology and Immunology 322:249–289.https://doi.org/10.1007/978-3-540-75418-3_12
-
Mucoidy, a general mechanism for maintaining lytic phage in populations of BacteriaFEMS Microbiology Ecology 96:fiaa162.https://doi.org/10.1093/femsec/fiaa162
-
Revisiting the rules of life for viruses of microorganismsNature Reviews Microbiology 19:501–513.https://doi.org/10.1038/s41579-021-00530-x
-
Functional amyloid and other protein fibers in the biofilm matrixJournal of Molecular Biology 430:3642–3656.https://doi.org/10.1016/j.jmb.2018.07.026
-
Curli biogenesis: order out of disorderBiochimica Et Biophysica Acta (BBA) - Molecular Cell Research 1843:1551–1558.https://doi.org/10.1016/j.bbamcr.2013.09.010
-
Biofilms: an emergent form of bacterial lifeNature Reviews Microbiology 14:563–575.https://doi.org/10.1038/nrmicro.2016.94
-
Big impact of the tiny: bacteriophage-bacteria interactions in biofilmsTrends in Microbiology 27:739–752.https://doi.org/10.1016/j.tim.2019.04.006
-
Rapidly fluctuating environments constrain coevolutionary arms races by impeding selective sweepsProceedings of the Royal Society B: Biological Sciences 280:20130937.https://doi.org/10.1098/rspb.2013.0937
-
Quantitative image analysis of microbial communities with BiofilmQNature Microbiology 6:151–156.https://doi.org/10.1038/s41564-020-00817-4
-
Advancing microbial sciences by individual-based modellingNature Reviews Microbiology 14:461–471.https://doi.org/10.1038/nrmicro.2016.62
-
Bacteriophage ecology in Escherichia coli and Pseudomonas aeruginosa mixed-biofilm communitiesApplied and Environmental Microbiology 77:821–829.https://doi.org/10.1128/AEM.01797-10
-
Close encounters of three kinds: bacteriophages, commensal Bacteria, and host immunityTrends in Microbiology 26:943–954.https://doi.org/10.1016/j.tim.2018.05.009
-
Bacteria-phage coevolution as a driver of ecological and evolutionary processes in microbial communitiesFEMS Microbiology Reviews 38:916–931.https://doi.org/10.1111/1574-6976.12072
-
Population and evolutionary dynamics of phage therapyNature Reviews Microbiology 2:166–173.https://doi.org/10.1038/nrmicro822
-
Spatial structure, cooperation and competition in biofilmsNature Reviews Microbiology 14:589–600.https://doi.org/10.1038/nrmicro.2016.84
-
Engineering and characterization of a superfolder green fluorescent proteinNature Biotechnology 24:79–88.https://doi.org/10.1038/nbt1172
-
Inverse regulatory coordination of motility and curli-mediated adhesion in Escherichia coliGenes & Development 22:2434–2446.https://doi.org/10.1101/gad.475808
-
Viral and microbial community dynamics in four aquatic environmentsThe ISME Journal 4:739–751.https://doi.org/10.1038/ismej.2010.1
-
Host-Parasite coexistence: the role of spatial refuges in stabilizing Bacteria-Phage interactionsThe American Naturalist 148:348–377.https://doi.org/10.1086/285929
-
Cellulose as an architectural element in spatially structured Escherichia coli biofilmsJournal of Bacteriology 195:5540–5554.https://doi.org/10.1128/JB.00946-13
-
Bright far-red fluorescent protein for whole-body imagingNature Methods 4:741–746.https://doi.org/10.1038/nmeth1083
-
Molecular genetics of bacteriophage P22Microbiological Reviews 42:385–413.https://doi.org/10.1128/mr.42.2.385-413.1978
-
Living in the matrix: assembly and control of Vibrio cholerae biofilmsNature Reviews Microbiology 13:255–268.https://doi.org/10.1038/nrmicro3433
-
Diversity and ecology of viruses in hyperarid desert soilsApplied and Environmental Microbiology 82:770–777.https://doi.org/10.1128/AEM.02651-15
Decision letter
-
Wenying ShouReviewing Editor; University College London, United Kingdom
-
Gisela StorzSenior Editor; National Institute of Child Health and Human Development, United States
-
Wenying ShouReviewer; University College London, United Kingdom
In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.
Acceptance summary:
Biofilm matrix-dependent phage protection has been observed in a variety of bacteria. Using techniques including microfluidic culture, single-cell resolution confocal microscopy, and phage infection reporter, authors show here that phages are trapped in curli fibres of outer E. coli biofilm layer, and that these phages can attack newly-arriving susceptible cells.
Decision letter after peer review:
Thank you for submitting your article "Matrix-trapped viruses can protect bacterial biofilms from invasion by colonizing cells" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including Wenying Shou as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen Gisela Storz as the Senior Editor.
The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.
Summary:
Biofilm matrix-dependent phage protection has been observed in a variety of bacteria. Using techniques including microfluidic culture, single-cell resolution confocal microscopy, and phage infection reporter, authors show here that phages are trapped in curli fibres of outer E. coli biofilm layer, and that these phages can attack newly-arriving susceptible cells.
Essential Revisions:
The reviewers are concerned about the generality of your findings. On the other hand, reviewers realize that perhaps, the strength of this paper is in quantitative methodology. Thus, you may choose to make it a Method/Resource paper (but you will need to state the novelty and the applicability of the Methods), or you can do more experiments to argue that the phenomenon is observed for other type(s) of phages.Reviewer #1:
Biofilm matrix-dependent phage protection has been observed in a variety of bacteria. Using techniques including microfluidic culture, single-cell resolution confocal microscopy, and phage infection reporter, authors show here that phages are trapped in curli fibres of outer E. coli biofilm layer, and that these phages can attack newly-arriving susceptible cells. Invading cells that arrived at the biofilm 10 hrs prior to phage arrival are protected by using the existing curli fibre as protection.
The experiments are nicely illustrated and well-done. The article is also well-written. I just have one question: What happens to the long-term fate of biofilm? Presumably, a fraction of cells can still grow in the biofilm and a fraction of cells will detach from the biofilm. How might the presence of phage affect these processes? Speculation is fine.Reviewer #2:
The motivation for this study is right on and important. In the real world, bacteria and bacteriophage (phage) live in physically structured habitats, where the bacteria exist as colonies or microcolonies are often embedded and stuck together in polymeric matrices known as biofilms. As the authors point out, most theoretical and experimental studies neglect this inconvenient reality. They use mass action models and their empirical analog, bacteria, and phage in well-agitated liquid culture. A full understanding of the population biology, ecology, and evolution of bacteria and phages require an understanding of how these populations intact in physically structured habitats.
This state-of-the-art study presents compelling evidence that the phage T7, which the authors have previously shown to be embedded in the matrices of biofilms are viable and capable of replicating on sensitive bacteria. It is not clear why these phages would not be viable and capable of replicating under these conditions. Is there evidence that suggests that phages are unable to replicate in biofilm populations of their host bacteria? If so, they should present this evidence.
This study considers only one form of physically structured bacterial populations, biofilms of E. coli on the surfaces of plastic tubes, and a single phage, T7. Does this result not obtain when the E. coli exist as colonies? Is this result unique to this rapid lysis phage? Will the same results obtain with other lytic phages, like T4? On surfaces and soft agar, the phage infection dynamics in these physically structured populations are different for these phages. T4 will be an excellent phage to examine whether rapid lysis is needed for this result; rapid lysis mutants of T4 can be obtained.
This study seems to have an agenda to provide evidence for, rather than broadly test, the hypothesis that lytic phages embedded in biofilms reduce the likelihood of those biofilms being colonized by other bacteria. The demonstration that matrix embedded phage T7 can reduce the likelihood of E. coli colonizing an established biofilm under the limited conditions of this experiment is of little general interest. Biofilms are commonly composed of multiple species of bacteria. The host range of phages is usually limited within a species and among strains of those species. The embedded phages of one species or strain of bacteria will not prevent colonization by other bacterial species.
More consideration should be given to the population and evolutionary dynamics of the phage in biofilm populations of bacteria. Do the phages in these E. coli biofilms continue to replicate and maintain their populations? What about resistance? As is commonly seen in liquid cultures of bacteria and lytic phages, resistance evolves, and the dominant population of bacteria is not susceptible to the phage. It would be of some interest and add generality to this study to know if the E. coli in the biofilms are sensitive or resistant to T7, and how that affects their premise that these embedded phages prevent colonization biofilms.
Biofilms are often associated with the pathogenesis of bacteria and resistance to antibiotics. It would be of some interest to know what the results presented in this study relate to these practical elements of biofilms' biology.
Although well-designed and well-executed, the inferences drawn from his study of E. coli and T7 in biofilms are too limited to support the publication of this manuscript in eLife.
Specific Comments and Suggestions:
– The abstract does a fine job of summarizing the main points of the paper.
– Line 44: How do they differ? Expand!
– Line 52: The articles cited do not provide compelling support for the statement that biofilms shape bacteria and phage evolution. This statement, which seems fundamental to this study's motivation, requires expansion and support from the literature.
– In general, the introduction provides the necessary background to understand the paper for microbiologists. On the other hand, the questions addressed in this study are not presented in a manner that would be of interest to the broad audience eLife aspires to.
– Line 80: The authors do not present the rationale behind the 62 hour-growth or biofilm confluent growth.
– Line 89: The densities of bacteria and phage in these experiments seem high. They should justify the choice of densities and multiplicities of infection. It will be of some interest if the same qualitative results obtain with seemingly more relativistic lower densities of bacteria and phage and different starting conditions. Of particular interest in this regard is Figure 1
– Line 99: There is no consideration of the dynamics of phage infection, particularly the latent period. Could the infecting population of bacteria replicate? Are they all killed by the embedded phage?
– Curiously, they emphasize their elegant and cool technology but don't the rigorous experiments and controls necessary to test their hypothesis that embedded phages prevent the colonization of biofilm populations of bacteria.
– Line 116: The authors need to provide the rationale behind selecting the pulse times (10 – hours). Is this period supported by previous findings that are not presented?
–
– Line 168: The authors present a conclusion at the end of this paragraph that could be considered an example of wishful thinking. The result presented seems to be inconsistent with their hypothesis. Nevertheless, they present these results as indirect evidence in support of their hypothesis.
– Line 173: From an evolutionary perspective, it would be difficult to develop a mechanism by which bacteria domesticate phage for their protection. They should consider the selective pressures responsible for the phage protecting the biofilm populations of bacteria from invasion by other bacteria. Is it a coincidental or evolved phenomenon?
– The authors implicitly assume the phage-bacteria interaction is that of a predator (parasitoid, really) and its host, negatively impacting the bacterial population. They may want to point out; their results suggest that in biofilms, the association between phage and bacteria is more that of a symbiont and its host. If this is correct, they should point this out.
– The authors should clearly state that their observations, albeit well-supported, may be specific for the model system they used and for strictly virulent phage infections.
– To a large extent, this report can be described as a methods paper, a proof in principle that the elegant methods they present can be used to address some of the interactions between phage and bacteria in biofilms. We suggest the authors focus the paper on the utility and the novelty of the technique used, rather than the role of virulent phage infection for protecting bacteria in biofilms from colonization by other bacteria.
– Figure 1 would be clearer if it was divided into three figures. The diagram (cartoon) could be presented as supplemental material since it is not required to understand the results presented in the manuscript's body. Initially, we found it difficult to understand the results presented in I, J and K are about.
– Figure 2 provides the information needed to understand this control experiment. But as suggested earlier, the logic behind the 10 hours of the pulse has to be provided.
– Figure 3 is hard to understand. The cartoon in Figure 3A does not provide a clear explanation of the experiment.
– The three main figures of the paper illustrate the method rather than provide evidence in support of their hypothesis. As elegant as the methods presented are, we don't consider this technical contribution appropriate for publication in eLife.
– In Supplemental Table S1. The authors should explain why some nucleotides are presented in upper case and others in lower case.
– The software used to design the primers should be made available.
– Figure S4. They can improve the resolution of this figure.Reviewer #3:
This short report builds on work by Vidakovic et al. (1) that had demonstrated protection of E. coli cells growing in biofilms against killing by bacteriophage T7 through sequestration of T7 phage particles by curli, a protein component of the E. coli biofilm matrix. In the work presented here, Bond et al. provide evidence that these entrapped T7 phage particles are still infectious and can kill invading E. coli cells that would otherwise be able to integrate into the biofilm. They further show that cells that invade the biofilm before T7 is introduced can integrate into the biofilm and acquire protection from killing by the phage. Intriguingly, this protection appears to be granted by curli produced by the resident cell population rather than curli produced by the invading cells themselves.
The data are convincing and clearly presented, and the authors' arguments are well supported. The findings are a significant experimental advance in the largely theoretical field of phage-host dynamics in structured communities.
There are a few opportunities for improvement in clarity or content:
A) Although the localization of matrix-entrapped T7 particles can be inferred by the patterns and cell death and curli staining shown here, it is never explicitly shown. The foundational paper for this report (1) did show by microscopy that fluorescently labeled T7 accumulates at the biofilm periphery in a curli- and flagellin-dependent manner and that fluorescently labeled T7 colocalizes with fluorescently labeled curli. Therefore, while not strictly necessary to show colocalization of T7 and curli in these experiments, it might strengthen the work to explicitly refer readers to the previous work.
B) What are the precedents for His-tagged curli monomers (with and without fluorescently antibody labeling) polymerizing and performing their normal function in E. coli biofilms? It's possible that matrix containing a significant amount of tagged protein and/or antibody might have altered functional properties compared to wild-type matrix.
C) What is the interpretation of the difference between the distributions of invading cell cluster sizes shown in Figures 1J and 1K? It's not necessarily an intuitive result. One might imagine that the phage burst resulting from lysis of an invading cell would lead to a locally high phage density and a high chance of the entire invading cell cluster being eliminated rather than some small number of cells surviving-something like a microscopic analog of clear plaques on a plate.
D) It's curious that invading cells, being in locally near-identical environments to their neighboring curli-producing resident cells, do not begin producing curli within 10 h of colonization. Speculation on this phenomenon would be welcome. If invading cells do eventually begin producing curli, what toggles the switch? If not, how do resident and invader cells "know" that they belong to different populations despite being congenic?
E) Image acquisition and analysis could be described in more detail. As is, it's unclear how many fields of view are examined per experimental replicate and how the fields of view were selected (for instance, in invasion experiments are the fields selected randomly for biovolume calculation or are fields showing invasion events specifically selected?).
Reference:
1. Vidakovic L, Singh PK, Hartmann R, Nadell CD, Drescher K. 2018. Dynamic biofilm architecture confers individual and collective mechanisms of viral protection. Nat Microbiol 3:26-31. doi:10.1038/s41564-017-0050-1
https://doi.org/10.7554/eLife.65355.sa1Author response
Reviewer #1:
Biofilm matrix-dependent phage protection has been observed in a variety of bacteria. Using techniques including microfluidic culture, single-cell resolution confocal microscopy, and phage infection reporter, authors show here that phages are trapped in curli fibres of outer E. coli biofilm layer, and that these phages can attack newly-arriving susceptible cells. Invading cells that arrived at the biofilm 10 hrs prior to phage arrival are protected by using the existing curli fibre as protection.
The experiments are nicely illustrated and well-done. The article is also well-written. I just have one question: What happens to the long-term fate of biofilm? Presumably, a fraction of cells can still grow in the biofilm and a fraction of cells will detach from the biofilm. How might the presence of phage affect these processes? Speculation is fine.
We would like to thank the referee for their positive evaluation of our work, and their inquiry regarding the downstream consequences of phage trapping for biofilm-dwelling cells. This is an excellent question and one which we hope to pursue in the future experimentally. For the time being, we can say that with sufficient time, biofilms of E. coli grow large enough that they eventually slough off of the surface to which they are attached. As our work shows that matrix-trapped phages remain infectious, the phage population presumably disperses with them and may infect newly planktonic cells. A key question is whether the resulting phage propagation is sufficient to halt further surface colonization by surviving bacteria and another round of biofilm growth. We have added a note in the text to this end, and we look forward to reporting on this question in future work (Lines 132-135).
Reviewer #2:
The motivation for this study is right on and important. In the real world, bacteria and bacteriophage (phage) live in physically structured habitats, where the bacteria exist as colonies or microcolonies are often embedded and stuck together in polymeric matrices known as biofilms. As the authors point out, most theoretical and experimental studies neglect this inconvenient reality. They use mass action models and their empirical analog, bacteria, and phage in well-agitated liquid culture. A full understanding of the population biology, ecology, and evolution of bacteria and phages require an understanding of how these populations intact in physically structured habitats.
We very much appreciate the referee’s agreement on the timeliness of the research questions pursued in our paper. We are very much of the same mind that phage-host interactions in nature can be completely different in the biofilm context. The referee brings up a number of important and constructive criticisms below, and we have endeavored at all junctures to address them. Some would require more experimental work than we currently have the ability to perform, so we are hopeful that a combination of new experiments and expanded discussion will be sufficient.
This state-of-the-art study presents compelling evidence that the phage T7, which the authors have previously shown to be embedded in the matrices of biofilms are viable and capable of replicating on sensitive bacteria. It is not clear why these phages would not be viable and capable of replicating under these conditions. Is there evidence that suggests that phages are unable to replicate in biofilm populations of their host bacteria? If so, they should present this evidence.
We appreciate the referee’s comments here. The rationale for questioning whether curli-trapped phages might not be active against cells is that they either degrade after being embedded in the matrix, or that they become embedded in orientations that prevent them from being able to readily infect cells that are arriving on the biofilm surface. This possibility had not been tested previously to our knowledge. For biofilms that are not protected by matrix (including E. coli biofilms before they begin producing curli fibers) phages are able to diffuse among and infect biofilm-dwelling bacteria. This was shown previously by the Vidakovic et al. paper to which the referee refers in the comment above. To clarify this point, we have added new text to the introduction in support of the rationale for the paper (Lines 59-62).
This study considers only one form of physically structured bacterial populations, biofilms of E. coli on the surfaces of plastic tubes, and a single phage, T7. Does this result not obtain when the E. coli exist as colonies? Is this result unique to this rapid lysis phage? Will the same results obtain with other lytic phages, like T4? On surfaces and soft agar, the phage infection dynamics in these physically structured populations are different for these phages. T4 will be an excellent phage to examine whether rapid lysis is needed for this result; rapid lysis mutants of T4 can be obtained.
These are all excellent suggestions, and we appreciate the insight on the part of the referee. The spatiotemporal matrix composition of colonies grown on agar and biofilms grown in flow chambers may differ substantially, although Regine Hengge’s work and the Vidakovic et al. paper also show some surprising similarities in matrix localization despite the different directions from which nutrients enter the cell groups in the different assays. While it would certainly be interesting to determine if our results hold in agar colony models, the level of single-cell level resolution fluorescence imaging required for our work is generally not possible for agar colonies. A main goal of the present paper, which demonstrates that matrix-trapped phages kill off invading cells, is to motivate further work on this question by others in the biofilm and phage fields using different growth conditions, host bacteria, and phages. We would respond similarly for the question of whether the same results hold for phage T4. We are not able to make this assessment at present due to lab restrictions, but testing the generality of these results is an important avenue for further work. We added new emphasis in the conclusion section of the paper to make it clear that our work established a proof of principle, but not a proof of generality across all environments, all bacteria, and all phages, and that we hope other researchers in the field will join us to address this question in the future (Lines 231-236).
This study seems to have an agenda to provide evidence for, rather than broadly test, the hypothesis that lytic phages embedded in biofilms reduce the likelihood of those biofilms being colonized by other bacteria. The demonstration that matrix embedded phage T7 can reduce the likelihood of E. coli colonizing an established biofilm under the limited conditions of this experiment is of little general interest. Biofilms are commonly composed of multiple species of bacteria. The host range of phages is usually limited within a species and among strains of those species. The embedded phages of one species or strain of bacteria will not prevent colonization by other bacterial species.
While we understand the skepticism on the part of this referee, we do not agree that our results will be of little general interest. First, the microfluidic flow conditions used in our experiments are an excellent model system that reflects many natural settings in which biofilms grow on a solid-liquid interface. Though we only examine a single-species and single-phage condition, this is the first result of its kind and meant as a demonstration of principle, not proof of universality, to spark interest in the question of how trapped phages influence bacterial colonization of resident. We agree that since phages are species and strain-specific, they will not kill all invading bacteria; in fact, we demonstrate this directly using a T7-resistant mutant of invading E. coli in the paper. Our results prompt the question of whether natural biofilms might trap phages of many different origins and then become resistant to colonization by the species/strains to which those trapped phages are virulent. An important future step is to determine whether this is the case, as we originally highlighted in the conclusion section of the paper. With this referee comment in mind, though, we understand we did not emphasize the caveats of our paper in sufficient depth, and we have expanded this section of the conclusion (to the extent possible given space constraints for eLife short reports) to make it clear what questions are left open by our results. We hope this modification to the conclusion will be more effective in prompting the readership to include similar assays in their work (Lines 236-244).
More consideration should be given to the population and evolutionary dynamics of the phage in biofilm populations of bacteria. Do the phages in these E. coli biofilms continue to replicate and maintain their populations? What about resistance? As is commonly seen in liquid cultures of bacteria and lytic phages, resistance evolves, and the dominant population of bacteria is not susceptible to the phage. It would be of some interest and add generality to this study to know if the E. coli in the biofilms are sensitive or resistant to T7, and how that affects their premise that these embedded phages prevent colonization biofilms.
Biofilms are often associated with the pathogenesis of bacteria and resistance to antibiotics. It would be of some interest to know what the results presented in this study relate to these practical elements of biofilms' biology.
These are also appreciated criticisms from the referee, and we can address the majority based on our previously published work. The phages trapped in the curli matrix might be able to replicate to a degree, but not to the extent that they any major impact on net positive growth of the biofilm. This was shown by Vidakovic et al. (2018). But overall, the phages are trapped in the curli matrix of mature E. coli biofilms, and neither proliferate nor degrade. Our central question in this paper was whether these phages can in principle infect and replicate and newly arriving, non-curli-protected cells arrive, and we find that the answer is yes.
We have shown in a previous publication that in a naïve resident biofilm typical of those used in our experiments, there are about 4x106 resident E. coli, of which ~15 cells are de novo resistant mutants. In all likelihood these mutants are buried in the biofilm in random locations and remain unexposed to phages blocked on the biofilm periphery. Among cells in the invading population, clearly few or none are resistant mutants because they are all killed by phages trapped on the resident biofilm exterior. The population sizes we are working with are orders of magnitude smaller than those in well-mixed experiments in test tubes – because the number of de novo resistant E. coli cells is small relative to the whole population, and phage-host interactions are spatially constrained, phage resistance evolution is very unlikely to make any difference to the outcome of our experiments. We have added new text stating this point and appreciate the note from the reviewer to make sure there is no confusion among readers (Lines 85-87).
Although well-designed and well-executed, the inferences drawn from his study of E. coli and T7 in biofilms are too limited to support the publication of this manuscript in eLife.
Specific Comments and Suggestions:
– The abstract does a fine job of summarizing the main points of the paper.
– Line 44: How do they differ? Expand!
We expand on this point in the following paragraph of the introduction.
– Line 52: The articles cited do not provide compelling support for the statement that biofilms shape bacteria and phage evolution. This statement, which seems fundamental to this study's motivation, requires expansion and support from the literature.
We have cited all the papers that we could find directly examining the impact of phage on host evolution in the biofilm context, including our own recently published work to this effect. We’re afraid that without further clarification from the referee on what was missing here, we are not able to expand further. We are also working with the limited word space of the short report format of eLife, and so not every introductory point can be expanded upon in full.
– In general, the introduction provides the necessary background to understand the paper for microbiologists. On the other hand, the questions addressed in this study are not presented in a manner that would be of interest to the broad audience eLife aspires to.
While we appreciate the reviewer’s concern about general interest, we do not agree that this is the case. This has been evident in numerous departmental talks and a biofilm conference at which this material has been recently presented and exceptionally well received by broad audiences. We are quite confident in the general interest of this work and therefore feel it is an excellent fit to eLife. We hope that after the clarifications noted above and below, that the referee will have change of heart about this.
– Line 80: The authors do not present the rationale behind the 62 hour-growth or biofilm confluent growth.
This point from the referee is much appreciated, and to support our addition to the manuscript we have performed additional experiments showing that it is at the ~62 h mark that resident biofilms of E. coli develop enough of a curli matrix layer to effectively block phage diffusion and trap phages in place on the biofilm exterior. We have added text and a new supplemental figure in light of this comment (Lines 91-93).
– Line 89: The densities of bacteria and phage in these experiments seem high. They should justify the choice of densities and multiplicities of infection. It will be of some interest if the same qualitative results obtain with seemingly more relativistic lower densities of bacteria and phage and different starting conditions. Of particular interest in this regard is Figure 1.
We are afraid that this is not possible to do. In order to obtain replicable E. coli biofilm conditions, we must start experiments with the surface density of cells that is shown in the first submission. As noted above, the subsequent growth period for 62 hours is needed for sufficient curli production to trap phages after exposure. Presumably, the effect of phages killing incoming susceptible cells might be weakened with very few phages trapped in the matrix, i.e. lower MOI. In principle this could be tested in the future, but given the difficulty of these experiments and current limitations on lab work, this is not feasible at the present time.
– Line 99: There is no consideration of the dynamics of phage infection, particularly the latent period. Could the infecting population of bacteria replicate? Are they all killed by the embedded phage?
To our knowledge, when invading bacteria were added to biofilms with phages embedded in the curli matrix, the invading bacteria were nearly all killed. The only invading cells we could find were expressing the phage infection reporter, indicating that they were in the middle of the phage latent period. This point is highlighted in the main text.
– Curiously, they emphasize their elegant and cool technology but don't the rigorous experiments and controls necessary to test their hypothesis that embedded phages prevent the colonization of biofilm populations of bacteria.
This criticism is not possible to address because it does not specify which experiments are missing. We feel that all of the needed controls are present in Figure 1, to establish the core point that matrix-embedded phages can prevent the colonization of resident biofilms by phage-susceptible (but not phage-resistant) bacteria. The clarity of this result seems to be supported by the comments from the other referees as well.
– Line 116: The authors need to provide the rationale behind selecting the pulse times (10 – hours). Is this period supported by previous findings that are not presented?
This was the duration of time over which we observed the biofilm population dynamics to come to equilibrium after phage pulse. This is now noted in the text (Lines 142-143).
–
– Line 168: The authors present a conclusion at the end of this paragraph that could be considered an example of wishful thinking. The result presented seems to be inconsistent with their hypothesis. Nevertheless, they present these results as indirect evidence in support of their hypothesis.
We appreciate the skepticism on the part of the referee, but here the referee does not specify exactly what they consider to be wishful thinking in the interpretation of these results. Furthermore, the hypothesis at the end of this paragraph is explicitly presented as interpretation.
– Line 173: From an evolutionary perspective, it would be difficult to develop a mechanism by which bacteria domesticate phage for their protection. They should consider the selective pressures responsible for the phage protecting the biofilm populations of bacteria from invasion by other bacteria. Is it a coincidental or evolved phenomenon?
This point is much appreciated and one which we did not communicate clearly enough in the original text. The matrix evolved without questions to address many environmental challenges, including resisting fluid shear, controlling cell spatial localization, and protection from external threats including phages and antibiotics. The fact that trapped phages can act as a defense against invasion by competitors may well be incidental, but nevertheless could contribute to the adaptive value of matrix production. We have changed our wording in the conclusion to reflect this point (Lines 226-230). We have also changed our title to be more descriptive of the main result to help avoid confusion.
– The authors implicitly assume the phage-bacteria interaction is that of a predator (parasitoid, really) and its host, negatively impacting the bacterial population. They may want to point out; their results suggest that in biofilms, the association between phage and bacteria is more that of a symbiont and its host. If this is correct, they should point this out.
We have adjusted the text in light of this comment from the referee (Line 226).
– The authors should clearly state that their observations, albeit well-supported, may be specific for the model system they used and for strictly virulent phage infections.
We have added text to the conclusion section to better emphasize this point (Lines 231-236).
– To a large extent, this report can be described as a methods paper, a proof in principle that the elegant methods they present can be used to address some of the interactions between phage and bacteria in biofilms. We suggest the authors focus the paper on the utility and the novelty of the technique used, rather than the role of virulent phage infection for protecting bacteria in biofilms from colonization by other bacteria.
We respectfully do not agree with this point from the referee. Though we are grateful for the referee’s appreciation of our technical approach, we do still feel that the highlight of this work is the ecological/biological result that matrix-trapped phages can in principle block the colonization of the biofilm by other bacteria that are susceptible to the trapped phages. Though this observation needs to be tested for other species and environments, this core result has not been reported previously and will, we are confident, prompt others in the field to think in a new way about biofilm-phage interactions. This hope has been borne out in numerous presentations to biofilm-specific and broad-audience lecture settings, so we do not think we are being unrealistic in this regard.
– Figure 1 would be clearer if it was divided into three figures. The diagram (cartoon) could be presented as supplemental material since it is not required to understand the results presented in the manuscript's body. Initially, we found it difficult to understand the results presented in I, J and K are about.
We’re afraid this is not possible given the figure constraints of this publication format for eLife. We are also confident that all of the information in Figure 1 is best presented in one package for the sake of rigor and completeness in the experiments. We have added a new paragraph to clarify the interpretation of Figure 1J,K (Lines 115-124).
– Figure 2 provides the information needed to understand this control experiment. But as suggested earlier, the logic behind the 10 hours of the pulse has to be provided.
This has been addressed in response to the referee’s comment above.
– Figure 3 is hard to understand. The cartoon in Figure 3A does not provide a clear explanation of the experiment.
We’re afraid that without further specification as to what is unclear, we are not able to address this comment from the referee.
– The three main figures of the paper illustrate the method rather than provide evidence in support of their hypothesis. As elegant as the methods presented are, we don't consider this technical contribution appropriate for publication in eLife.
We respectfully disagree with this assessment, which is also not consistent with the feedback from the other two referees. The referee states here that the evidence presented does not justify the conclusions, but she/he does not specify in detail why they feel this is the case. Without more specific feedback, this criticism is not possible to address, nor is it consistent with the sentiments of the other referees on this submission.
– In Supplemental Table S1. The authors should explain why some nucleotides are presented in upper case and others in lower case.
This has been corrected.
– The software used to design the primers should be made available.
This has been corrected.
– Figure S4. They can improve the resolution of this figure.
This figure is at the limit of the resolution permitted by the confocal microscopy equipment we have access to.
Reviewer #3:
This short report builds on work by Vidakovic et al. (1) that had demonstrated protection of E. coli cells growing in biofilms against killing by bacteriophage T7 through sequestration of T7 phage particles by curli, a protein component of the E. coli biofilm matrix. In the work presented here, Bond et al. provide evidence that these entrapped T7 phage particles are still infectious and can kill invading E. coli cells that would otherwise be able to integrate into the biofilm. They further show that cells that invade the biofilm before T7 is introduced can integrate into the biofilm and acquire protection from killing by the phage. Intriguingly, this protection appears to be granted by curli produced by the resident cell population rather than curli produced by the invading cells themselves.
The data are convincing and clearly presented, and the authors' arguments are well supported. The findings are a significant experimental advance in the largely theoretical field of phage-host dynamics in structured communities.
We are delighted for the referee’s overall positive impression of our work, and we appreciate both the close reading of the paper and suggestions for improvement below.
There are a few opportunities for improvement in clarity or content:
A) Although the localization of matrix-entrapped T7 particles can be inferred by the patterns and cell death and curli staining shown here, it is never explicitly shown. The foundational paper for this report (1) did show by microscopy that fluorescently labeled T7 accumulates at the biofilm periphery in a curli- and flagellin-dependent manner and that fluorescently labeled T7 colocalizes with fluorescently labeled curli. Therefore, while not strictly necessary to show colocalization of T7 and curli in these experiments, it might strengthen the work to explicitly refer readers to the previous work.
This comment is much appreciated, and we now make a clearer reference to the Vidakovic et al. (2018) paper with respect to the observation of matrix-trapped phages on the periphery of curli-expressing biofilms (Lines 52-54).
B) What are the precedents for His-tagged curli monomers (with and without fluorescently antibody labeling) polymerizing and performing their normal function in E. coli biofilms? It's possible that matrix containing a significant amount of tagged protein and/or antibody might have altered functional properties compared to wild-type matrix.
This is a helpful note as we did not mention the precedent for localizing curli via immunostaining of His-tagged csgA. This method has been used in several previous publications studying E. coli biofilm architecture by the Hengge and Drescher groups, with no indication of interference with normal matrix function (i.e. no major differences from WT biofilm morphology, shear resistance, etc.). We now make a more careful note of this in the main text (Lines 162-163).
C) What is the interpretation of the difference between the distributions of invading cell cluster sizes shown in Figures 1J and 1K? It's not necessarily an intuitive result. One might imagine that the phage burst resulting from lysis of an invading cell would lead to a locally high phage density and a high chance of the entire invading cell cluster being eliminated rather than some small number of cells surviving-something like a microscopic analog of clear plaques on a plate.
Thank you for pointing out that this portion of the figure needs some additional explanation in the text. Figure 1J and K depict the change in cluster size distribution of invading cells in the absence (J) and presence (K) of curli-trapped phages on the resident biofilm’s surface. They show that when phages are absent when invading cells arrive, the invading strain is able to establish cell clusters over time that most likely result from several rounds of division by one or a few cells that attach to a given location on the resident biofilm surface. When phages are present, on the other hand, the majority of invading cells are killed, and so there are almost no changes for larger cell clusters to develop. We have added new text to the Results section to make sure this interpretation is explained clearly (Lines 115-124).
D) It's curious that invading cells, being in locally near-identical environments to their neighboring curli-producing resident cells, do not begin producing curli within 10 h of colonization. Speculation on this phenomenon would be welcome. If invading cells do eventually begin producing curli, what toggles the switch? If not, how do resident and invader cells "know" that they belong to different populations despite being congenic?
On the basis of our His-labeling experiments, we are confident that the invading E. coli strain does not begin producing their own curli on the 10 h time scale before phages are introduced, but from a mechanistic perspective we are not entirely sure as yet why this is the case. From our experience, cells colonizing surface from the planktonic phase take at least 2 days before they begin to produce curli. Though the curli regulon has been explored in detail by other groups, it is not fully clear what mechanisms control this timing. We now add direction discussion of this point (Lines 170-175). We can speculate that invading cells, which do not appear to be expressing the csg operon and therefore most likely do not have a CsgB baseplate on their exterior, cannot directly coopt the CsgA produced by curli-producing resident biofilm cells, and that this controls the separation of the two populations despite the fact that they are congenic. As we show, though, the invading strain does appear to be folded into the growing front of the resident biofilm, despite the fact that the invading strain does not appear to be activating the csg operon. Additional commentary on this point is now provided at the end of this results subsection (Lines 206-210).
E) Image acquisition and analysis could be described in more detail. As is, it's unclear how many fields of view are examined per experimental replicate and how the fields of view were selected (for instance, in invasion experiments are the fields selected randomly for biovolume calculation or are fields showing invasion events specifically selected?).
We appreciate this input from the referee and have updated the methods section to make it clear that image locations were selected randomly, and to highlight sample sizes as directly as possible for each quantification.
Reference:
1. Vidakovic L, Singh PK, Hartmann R, Nadell CD, Drescher K. 2018. Dynamic biofilm architecture confers individual and collective mechanisms of viral protection. Nat Microbiol 3:26-31. doi:10.1038/s41564-017-0050-1
The basic story reminds me of the study (now 10 years old, maybe) out of Forest Rohwer's group showing that phage bind the mucous surrounding corals. The authors suggested that the phage provided a kind of defense system for the coral against bad bacteria. My thought was that phages sticking in the mucous probably included those that attacked good bacteria; it would be wishful thinking to suppose that phages were evolving to benefit the coral. The authors never mentioned this obvious point.
In this case (from the abstract), as with Rowher's group, there is an implied purpose to the defense rather than more benign explanations. I don't like that, as it's almost as if they are trying to create a story. (The text may be more reasonable.)
We appreciate this feedback from the 4th informal referee. We are also familiar with the Rohwer study, and in fact this was one of the main motivations for this paper. The Rohwer study and our results’ connection to it was a central part of the original submission’s Discussion section, and remains so. As to the interpretation that phages evolved to help the biofilms in which they become trapped, we never held this interpretation and agree with the referee that the fact that phages may help biofilms repel invading cells would be a coincidental benefit to the fact that the matrix is trapping the phages and preventing them from harming the biofilm-resident cells in the first place; this latter function is what we would presume to be the primary evolutionary explanation for the biofilm matrix- i.e. protecting the biofilm from threats, including phages. The fact that the trapped phages may help the biofilm-resident cells, even if a coincidental benefit, is nevertheless a potentially important ecological phenomenon. This is the main message of the paper. In response to this note and that of referee #2 above, we have modified the title of the paper to be more descriptive, and we have modified the text of the conclusion so as not to give the impression that we are arguing phages have evolved to help biofilms (Lines 228-233).
https://doi.org/10.7554/eLife.65355.sa2Article and author information
Author details
Funding
Dartmouth College (GANN Fellowship)
- Matthew C Bond
National Institutes of Health (P30-DK117469)
- Matthew C Bond
- Carey D Nadell
European Research Council (StG-716734)
- Knut Drescher
Deutsche Forschungsgemeinschaft (DR 982/5–1)
- Knut Drescher
Behrens-Weise-Foundation
- Knut Drescher
Simons Foundation (826672)
- Carey D Nadell
National Science Foundation (MCB 1817342)
- Carey D Nadell
National Science Foundation (IOS 2017879)
- Carey D Nadell
Dartmouth College (Burke Award)
- Carey D Nadell
Cystic Fibrosis Foundation (STANTO15RO)
- Carey D Nadell
National Institutes of Health (2R01AI081838)
- Carey D Nadell
National Institutes of Health (P20-GM113132)
- Carey D Nadell
Human Frontier Science Program (RGY0077/2020)
- Carey D Nadell
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Acknowledgements
The authors are grateful to members of the Nadell lab at Dartmouth for feedback on the project. MCB was supported by a GANN Fellowship from Dartmouth College and NIH grant P30-DK117469 to the Dartmouth Cystic Fibrosis Research Center. KD is supported by the European Research Council (StG-716734), the Deutsche Forschungsgemeinschaft (DR 982/5–1), and the Behrens-Weise-Foundation. CDN is supported by the Simons Foundation Award Number 826672, NSF grant MCB 1817342, NSF grant IOS 2017879, a Burke Award from Dartmouth College, a pilot award from the Cystic Fibrosis Foundation (STANTO15RO), NIH grant P30-DK117469, NIH grant 2R01AI081838 to PI Robert Cramer, NIH grant P20-GM113132 to the Dartmouth BioMT COBRE, and grant RGY0077/2020 from the Human Frontier Science Foundation with co-PI Alexandre Persat.
Senior Editor
- Gisela Storz, National Institute of Child Health and Human Development, United States
Reviewing Editor
- Wenying Shou, University College London, United Kingdom
Reviewer
- Wenying Shou, University College London, United Kingdom
Publication history
- Received: December 1, 2020
- Accepted: July 8, 2021
- Accepted Manuscript published: July 9, 2021 (version 1)
- Version of Record published: August 6, 2021 (version 2)
Copyright
© 2021, Bond et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 2,197
- Page views
-
- 318
- Downloads
-
- 12
- Citations
Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Matrix-trapped viruses can prevent invasion of bacterial biofilms by colonizing cells
eLife 10:e65355.
https://doi.org/10.7554/eLife.65355
Further reading
-
- Ecology
- Evolutionary Biology
Circadian clocks infer time of day by integrating information from cyclic environmental factors called zeitgebers, including light and temperature. Single zeitgebers entrain circadian rhythms, but few studies have addressed how multiple, simultaneous zeitgeber cycles interact to affect clock behavior. Misalignment between zeitgebers (‘sensory conflict’) can disrupt circadian rhythms, or alternatively clocks may privilege information from one zeitgeber over another. Here, we show that temperature cycles modulate circadian locomotor rhythms in Nematostella vectensis, a model system for cnidarian circadian biology. We conduct behavioral experiments across a comprehensive range of light and temperature cycles and find that Nematostella’s circadian behavior is disrupted by chronic misalignment between light and temperature, which involves disruption of the endogenous clock itself rather than a simple masking effect. Sensory conflict also disrupts the rhythmic transcriptome, with numerous genes losing rhythmic expression. However, many metabolic genes remained rhythmic and in-phase with temperature, and other genes even gained rhythmicity, implying that some rhythmic metabolic processes persist even when behavior is disrupted. Our results show that a cnidarian clock relies on information from light and temperature, rather than prioritizing one signal over the other. Although we identify limits to the clock’s ability to integrate conflicting sensory information, there is also a surprising robustness of behavioral and transcriptional rhythmicity.
-
- Ecology
- Evolutionary Biology
Division of labor, where subpopulations perform complementary tasks simultaneously within an assembly, characterizes major evolutionary transitions of cooperation in certain cases. Currently, the mechanism and significance of mediating the interaction between different cell types during the division of labor, remain largely unknown. Here, we investigated the molecular mechanism and ecological function of a policing system for optimizing the division of labor in Bacillus velezensis SQR9. During biofilm formation, cells differentiated into the extracellular matrix (ECM)-producers and cheater-like nonproducers. ECM-producers were also active in the biosynthesis of genomic island-governed toxic bacillunoic acids (BAs) and self-resistance; while the nonproducers were sensitive to this antibiotic and could be partially eliminated. Spo0A was identified to be the co-regulator for triggering both ECM production and BAs synthesis/immunity. Besides its well-known regulation of ECM secretion, Spo0A activates acetyl-CoA carboxylase to produce malonyl-CoA, which is essential for BAs biosynthesis, thereby stimulating BAs production and self-immunity. Finally, the policing system not only excluded ECM-nonproducing cheater-like individuals but also improved the production of other public goods such as protease and siderophore, consequently, enhancing the population stability and ecological fitness under stress conditions and in the rhizosphere. This study provides insights into our understanding of the maintenance and evolution of microbial cooperation.
