Initiation of H1-T6SS dueling between Pseudomonas aeruginosa

ABSTRACT The Type VI secretion system (T6SS) is a multicomponent apparatus, present in many Gram-negative bacteria, which can inhibit bacterial prey in various ecological niches. Pseudomonas aeruginosa assembles one of its three T6SS (H1-T6SS) to respond to attacks from adjacent competing bacteria. Surprisingly, repeated assemblies of the H1-T6SS, termed dueling, were described in a monoculture in the absence of an attacker strain; however, the underlying mechanism was unknown. Here, we explored the role of H2-T6SS of P. aeruginosa in triggering H1-T6SS assembly. We show that H2-T6SS inactivation in P. aeruginosa causes a significant reduction in H1-T6SS dueling and that H2-T6SS activity directly triggers retaliation by the H1-T6SS. Intraspecific competition experiments revealed that elimination of H2-T6SS in non-immune prey cells conferred protection from H1-T6SS. Moreover, we show that the H1-T6SS response is triggered independently of the characterized lipase effectors of the H2-T6SS, as well as those of Acinetobacter baylyi and Vibrio cholerae. Our results suggest that H1-T6SS response to H2-T6SS in P. aeruginosa can impact intraspecific competition, particularly when the H1-T6SS effector-immunity pairs differ between strains, and could determine the outcome of multistrain colonization. IMPORTANCE The opportunistic pathogen Pseudomonas aeruginosa harbors three different Type VI secretion systems (H1, H2, and H3-T6SS), which can translocate toxins that can inhibit bacterial competitors or inflict damage to eukaryotic host cells. Unlike the unregulated T6SS assembly in other Gram-negative bacteria, the H1-T6SS in P. aeruginosa is precisely assembled as a response to various cell damaging attacks from neighboring bacterial cells. Surprisingly, it was observed that neighboring P. aeruginosa cells repeatedly assemble their H1-T6SS toward each other. Mechanisms triggering this “dueling” behavior between sister cells were unknown. In this report, we used a combination of microscopy, genetic and intraspecific competition experiments to show that H2-T6SS initiates H1-T6SS dueling. Our study highlights the interplay between different T6SS clusters in P. aeruginosa, which may influence the outcomes of multistrain competition in various ecological settings such as biofilm formation and colonization of cystic fibrosis lungs.

Genetic clusters of T6SS generally consist of a set of 13 core genes that are conserved in Gram-negative species such as Escherichia coli, Vibrio cholerae, and P. aeruginosa (11).These 13 genes encode cytoplasmic, periplasmic, and membrane proteins.The T6SS functionally resembles a contractile bacteriophage tail (12,13).It is composed of a cytoplasmic contractile sheath, which is tethered to the cell envelope by a base plate structure and a membrane complex (12,(14)(15)(16).The sheath contraction powers the translocation of effectors bound to the T6SS spike/tube components, namely, VgrG, PAAR, and Hcp (2,3,17,18).The contracted sheath is specifically disassembled by the AAA+ ATPase ClpV, which facilitates recycling of sheath subunits for new rounds of T6SS assembly (19)(20)(21).T6SS activity comprising sheath polymerization, contraction, and disassembly can be visualized by live-cell fluorescence microscopy (21).Cells secret ing anti-bacterial toxins are protected by cognate immunity proteins often encoded immediately downstream of the effector genes (22).
Several mechanisms regulate the T6SS expression and activity in P. aeruginosa.The posttranscriptional regulator RsmA negatively regulates the expression of all the three T6SS clusters (23).The RsmA-dependent posttranscriptional repression can be relieved by the removal of RetS, which is a sensor kinase (24).The deletion of retS activates the GacS/GacA two-component system, which leads to the expression of two small noncoding RNAs, RsmY and RsmZ.These noncoding RNAs further sequester the T6SS-inhibitory transcriptional regulator RsmA (2,25,26).In addition, previous studies showed that the H1-T6SS is posttranslationally controlled by the threonine phosphory lation pathway (TPP) encoded next to the H1-T6SS cluster (2,27,28).The TPP compo nents are the TagQRST sensor module, the PpkA kinase that can phosphorylate its substrate, Fha, and its cognate phosphatase, PppA (29,30).Activation of PpkA by a poorly characterized mechanism dependent on TagQRST results in Fha phosphorylation, which triggers H1-T6SS assembly precisely positioned to the site where P. aeruginosa is attacked by a neighboring bacterium (21,(29)(30)(31).Apart from T6SS-mediated attacks from competing bacteria, membrane perturbations by T4SS, reagents like EDTA or polymyxin B and extracellular DNA can also stimulate the TagQRST sensor module to trigger H1-T6SS assembly (32,33).Furthermore, it was shown that V. cholerae V52 T6SS effector TseL with phospholipase activity was required and sufficient for triggering a response for the H1-T6SS of P. aeruginosa (34).However, another report demonstrated that Acinetobacter baylyi ADP1 lacking all known effectors, including a lipase, can also trigger H1-T6SS (35), suggesting that physical puncture may be sufficient under some conditions to elicit an H1-T6SS response.Interestingly, it was observed that pairs of P. aeruginosa sister cells engage in spatially and temporally correlated rounds of H1-T6SS assemblies, termed dueling (21,31).Because H1-T6SS assembly is a result of an external stimulus, such dueling was attributed to accidental H1-T6SS attacks from neighboring sister cells (21).
Here, we reasoned that the assembly of H2-T6SS or H3-T6SS may be responsible for initiating H1-T6SS dueling and indeed show that inactivation of H2-T6SS reduced H1-T6SS dueling.In addition, we show by fluorescence microscopy and intraspecific competition experiments that the H2-T6SS, or the T6SS of V. cholerae 2740-80 and A. baylyi, can trigger an H1-T6SS response, even in the absence of lipase effectors, suggesting that such effectors are dispensable for triggering retaliation by the H1-T6SS.

H1-T6SS assembly rate is increased by cell-cell contact
To test the hypothesis that assembly of H2-or H3-T6SS could be triggering H1-T6SS response, we examined the activities of all the three T6SS clusters in a ΔretS mutant (21,36).We constructed chromosomal fusions between genes encoding the T6SS sheath components (tssB) and mNeonGreen (for the H1-T6SS or the H3-T6SS) or mCherry2 (for H2-T6SS) encoding genes.We measured single-cell expression levels by flow cytometry in cells incubated under conditions similar to the conditions used below for testing bacterial cell-cell interactions.We detected high expression levels of TssB1-mNG and TssB2-mCh2 in cells that were grown to mid-exponential phase, concentrated and incubated for 3 h on Luria-Bertani (LB) agar.However, no expression of the H3-T6SS could be detected under those conditions (Fig. S1A and B).Furthermore, we performed live-cell fluorescence microscopy imaging of the H1-and H2-T6SS sheath-labeled strains and we observed H1-T6SS sheath dynamics and dueling (defined here as spatially and temporally correlated sheath assembly, contraction, and disassembly in two neighbor ing cells) as well as multiple dynamic H2-T6SS sheath structures, both in a ΔretS and wild-type (WT) background (Fig. 1A and B).
In order to test if the H1-T6SS sheath dynamics (assembly, contraction, disassembly) depend on cell-cell contact, we also measured the TssB1 sheath assembly in cells incubated and imaged at low and high cell density (Fig. 1A and B).We show that about 17% of cells imaged at high density (cells in close contact) assembled H1-T6SS over a period of 3 min, whereas only about 6% of cells assembled H1-T6SS when incubated at low density (minimal cell-cell contact) (Fig. 1C).In contrast, the H2-T6SS sheath assembly was detected in a major subset (around 70%) of the ΔretS cells irrespective of cell-cell contact (Fig. 1A through C).Similarly, TssB1-mNG and TssB2-mCh2 assemblies can be identified in a wild-type (retS positive) genetic background, albeit with lower overall expression and fewer assembled structures (Fig. 1B).This indicates that retS deletion induces H1-T6SS and H2-T6SS expression and assembly, and that the H1-T6SS activity is enhanced by cell-cell contact.

Inactivation of H2-T6SS reduces H1-T6SS dueling
The observation that cell-cell contact increases H1-T6SS assembly but not H2-T6SS activity under those conditions led us to test the influence of H2-T6SS on H1-T6SS activity.We compared TssB1-mNG assembly in a wild-type or ΔretS parental strain and a mutant strain that lacked TssL2, an essential structural component of H2-T6SS (Fig. S1C; Fig. 1D).Remarkably, deletion of tssL2 significantly reduced the H1-T6SS assemblies both in a wild-type and a ΔretS genetic background.In a ΔretS strain, we found that 15% of cells display H1-T6SS sheath dynamics compared to only around 5% in the tssL2-negative strain.This decrease in H1-T6SS activity was reversed by addition of a copy of tssL2 in trans (Fig. 1E).To test if H2-T6SS-dependent increase in H1-T6SS assembly required cellcell contact, we also investigated the impact of H2-T6SS on H1-T6SS activity in a liquid culture.Because Hcp secretion is a hallmark of functional T6SS, we monitored the Hcp1 abundance in supernatants harvested from liquid cultures of the parental and tssL2negative strains.Interestingly, the Hcp1 secretion was comparable in the parental and ΔtssL2 strains (Fig. 2A).
Next, we tested if H2-T6SS had an effect on H1-T6SS ability to counter T6SS attacks from V. cholerae.Interbacterial competition experiments showed that the H2-T6SS plays no role in the ability of P. aeruginosa to retaliate to V. cholerae T6SS-mediated attacks.We recovered about 100-fold less T6SS-positive V. cholerae than T6SS-negative V. cholerae cells upon incubation with both the parental and tssL2 deletion P. aeruginosa strains (Fig. 2B and C; Fig. S2A and B).This specific inhibition of T6SS-positive V. cholerae was dependent on H1-T6SS (Fig. 2B; Fig. S2A and B) as shown previously (31).
Altogether, these observations suggest that H2-T6SS activity can trigger H1-T6SS assembly in a contact-dependent manner.However, the H2-T6SS activity has no influence on H1-T6SS activity in liquid culture or the ability of P. aeruginosa H1-T6SS to respond to attacks from other bacteria.
Together, our results indicate that P. aeruginosa retaliates with H1-T6SS in response to incoming H2-T6SS attacks from a neighboring sister cell or cells of another P. aeruginosa isolate.

Loss of H2-T6SS activity confers protection from H1-T6SS
To test if H2-T6SS-triggered H1-T6SS counterattack has any impact on P. aeruginosa cell interactions, we tested if the H1-T6SS could kill a non-immune P. aeruginosa cell, as a response to H2-T6SS attack.First, we performed in-frame single deletions of H1-T6SS effector-immunity pairs in PAO1 (tse1-tsi1 to tse7-tsi7) and concluded that, under our experimental conditions, most H1-T6SS dependent killing is mediated by Tse5 effector (Fig. S4).Non-immune prey strain lacking tse5-tsi5 was recovered about 10-to 15-fold less than the parental, immune prey after co-incubation with H1+/H2+ or H1+/H2attacker strains (Fig. 5A and C), and this inhibition was solely dependent on H1-T6SS activity in the attacker (Fig. 5B).Because P. aeruginosa sister cells duel with H1-T6SS The data are represented as mean ± standard deviation of five independent biological replicates.Number of H1-T6SS-active H2-T6SS-negative cells, n = 342.(C) Representative images of a retaliation event of TssB1-mNG in a wild-type genetic background against a PAO1 strain carrying a TssB2-mCh2 label in a wild-type genetic background.The event is highlighted with a white arrow in each frame.(21), we wondered if the elimination of H1-T6SS activity in the prey can minimize its killing.Interestingly, H1-T6SS-negative, non-immune prey was inhibited by H1+/H2+ or H1+/H2-attacker cells similarly to the H1-T6SS-positive non-immune prey (Fig. 5A and  C).

DISCUSSION
In this study, we aimed to identify the mechanism that triggers the spontaneous H1-T6SS assemblies in P. aeruginosa cells.Previous reports showed that treatment with polymyxin B, extracellular DNA, and EDTA chelator, as well as T4SS and T6SS attacks from sister cells or other heterologous bacteria, can activate H1-T6SS via the TagQRST signaling pathway (21,(31)(32)(33).Interestingly, sensing of kin cell lysis by the Gac/Rsm pathway was shown to lead to an elevated expression of genes that encode H1-T6SS components (45).However, spontaneous H1-T6SS assemblies that arise in P. aeruginosa ΔretS as well as in RetSpositive cells, termed dueling, were poorly understood.The fact that the H1-T6SS can sense external T6SS attacks motivated us to examine if H2-T6SS/H3-T6SS activity of a sister cell could trigger H1-T6SS response.However, the expression of only H1 and H2-T6SS was detected in P. aeruginosa PAO1 ΔretS by flow cytometry (Fig. S1A and B).
Our microscopic analysis of P. aeruginosa cells on a solid agarose pad indicated that absence of H2-T6SS led to a significant decrease in H1-T6SS assemblies (Fig. 1D and E).On the other hand, the ability to retaliate to T6SS-mediated attacks by V. cholerae was independent of H2-T6SS (Fig. 2B).Because Hcp1 was also present in the culture superna tant of H2-T6SS-negative mutant (Fig. 2A), we concluded that the decreased assembly of the H1-T6SS in an H2-T6SS-negative strain is not due to a structural defect in the assembly of the H1-T6SS.We hypothesized that H1-T6SS activity is triggered by H2-T6SS only when sister cells are in close contact on a solid surface.Indeed, time-lapse imaging experiments revealed that H1-T6SS is assembled in direct response to H2-T6SS attacks from a neighboring cell, or toward cells that had an active H2-T6SS (Fig. 3A through C).Importantly, H1-T6SS in PA14 also responded to H2-T6SS attack from PAO1 (Fig. 4A and  B), which indicated that H1-T6SS retaliation to H2-T6SS attacks happens in between P. aeruginosa strains.
Competition experiments showed that a lethal H1-T6SS counterattack from an attacker strain is dependent on H2-T6SS activity in the non-immune prey.The nonimmune prey without a functional H2-T6SS was spared from H1-T6SS-mediated inhibi tion (about three-or five-fold increase in its recovery) (Fig. 5A and C).Importantly, the observed H2-T6SS-mediated triggering of H1-T6SS can also be observed in strain PA14 (Fig. S6B).
Altogether, we propose a model that describes the events that lead to H1-T6SS dueling between sister cells (Fig. 8).H2-T6SS attack from an initiator cell (Fig. 8, green shade) can elicit an H1-T6SS response from the neighboring cell (Fig. 8, magenta shade).As a retaliation to this response, the H1-T6SS is assembled precisely toward the initiator.Hitting the initiator cell with H1-T6SS will trigger localized assembly of its H1-T6SS and thus result in rounds of H1-T6SS assemblies in those two cells, observed as dueling (Fig. 8).However, the exact mechanism on how H2-T6SS activates the TagQRST sensory module remains unknown.An earlier study described the role of V. cholerae V52 T6SS effector TseL in activating the TagQRST cascade (34).However, our results show that V. cholerae 2740-80 T6SS without any effector activity can effectively trigger an H1-T6SS response (Fig. 7C and D).We also found that H2-T6SS phospholipase effectors Tle1, Tle3, Tle4/TplE, Tle5a/PldA, and Tle5b/PldB, although essential for fully active H2-T6SS, are not required to elicit an H1-T6SS response (Fig. 6B and C).Finally, an A. baylyi mutant strain lacking all five known T6SS effectors, including a lipase, was shown to elicit H1-T6SS responses as previously described (35).Taken together, our study supports the model that H1-T6SS can be activated independently of effector activity.
Our findings shed light on P. aeruginosa intraspecific competitions that may occur in various habitats.It was shown earlier that P. aeruginosa strains, PAO1, and PAK had differences in H1-T6SS toxin-immunity pair tse7-tsi7.Intriguingly, the Tsi7 PAK was unable to block the toxic effects of Tse7 PAO1 and therefore PAO1 could outcompete PAK (46).Furthermore, the three T6SS clusters were reported to be upregulated in P. aeruginosa PAO1 biofilm cells (47), which suggests that the H1-T6SS might also respond to H3-T6SS puncture in conditions that are conducive to H3-T6SS assembly.P. aeruginosa biofilm is regulated by effective cell-cell communication mediated by quorum sensing (5,48), and H1-T6SS response to H2-T6SS may affect intercellular signaling in the course of biofilm formation.Our results also suggest that the interplay between the different T6SS clusters in P. aeruginosa can influence the outcome of interstrain competitions, particularly when effector-immunity pairs differ between strains, as is often the case in clinical isolates of P. aeruginosa (49), and could promote colonization by specific isolates in lung microenvironments of CF patients.Considering the role of P. aeruginosa biofilms in long-term persistence, infection of CF patients, and antimicrobial resistance, additional research is needed to understand T6SS-mediated cell interactions within biofilms and during infection (5,10,50,51).

Bacterial strains and growth conditions
P. aeruginosa PAO1/PA14 ∆retS-labeled strains, as well as knockout strains, were generated as described previously (31).A list of strains used can be found in Table S1.Recombinant clones were checked by colony PCR and were sequence verified.A list of plasmids used in this study can be found in Table S2.Bacteria were grown in LB broth at 37°C.Liquid cultures were grown aerobically.E. coli DH5α was used as a cloning strain; genetic manipulations in P. aeruginosa were carried out using E. coli SM10 λpir for conjugation.Antibiotics used were streptomycin (100 µg/mL for V. cholerae and A. baylyi), tetracycline (100 µg/mL for P. aeruginosa and 12.5 µg/mL for E. coli), and gentamicin (20 µg/mL).

Flow cytometry
Bacteria with different fluorescently labeled T6SS sheaths (TssB1-mNG, TssB2-mCh2, or TssB3-mNG) were grown to mid-exponential phase at 37°C from an overnight culture in LB and measured, or concentrated to an optical density (OD) of 10, spotted on an LB agar plate and incubated for 3 h at 37°C, then resuspended and measured.Prior to measurement, all samples were diluted 100× in phosphate-buffered saline (PBS) and were analyzed by a BD Fortessa flow cytometer, with 950 V green and 950 V red laser on a high-throughput system (HTS) in a high-throughput mode, with a sample flow rate of 1 µL/s for a total of 10 µL of sample.Data were analyzed using FlowJo v10, and the median fluorescence intensity (MFI) was plotted using Prism.

Bacterial killing assays
Quantitative killing assays were performed as described previously (31,52).Briefly, overnight cultures were diluted 1 to 100 into fresh LB.Bacterial pellets were harvested at OD ≈ 1, washed twice in 1 mL LB and concentrated 10 times (OD ≈ 10).Indicated strains were mixed at a ratio of 10:1 (P.aeruginosa to V. cholerae), and 5 µL of the mixtures was spotted on dry LB plates and incubated for 3 h at 37°C.Subsequently, bacterial spots were cut out, and cells were resuspended in 0.5 mL LB.Cell mixtures were spotted in serial dilutions (1:10) on selective recovery plates (streptomycin for V. cholerae, irgasan for P. aeruginosa).CFU was counted after ≈16 h of incubation at 37°C.Three independent biological replicates were analyzed.

Intraspecific competition assays
For intraspecific competition experiments using P. aeruginosa strains, plasmids with antibiotic resistance markers for prey and attacker strains were used.In prey strains, we electroporated pME6032 vector (53) (that encodes the gene for tetracycline resistance), and attacker strain was transformed with pPSV38 vector (54) (gentamicin resistance marker), by conjugation.Cultures were treated in the same manner as described in bacterial killing assays, but with some variations.Instead of concentrating the cultures to OD ≈ 10, we normalized to OD ≈ 1.Furthermore, the attacker and prey strains were mixed at a 20:1 ratio.Then, 5 µL of cell mixture was spotted on LB agar plates and incubated for 16 h at room temperature.Afterward, the bacterial spots were excised, resuspended in LB, and spotted in serial dilutions as described in bacterial killing assays.The attacker strains were selected on LB plates with gentamicin antibiotic, and prey strains were selected on LB plates with tetracycline.Three or four independent biological replicates were analyzed.

Electroporation in P. aeruginosa
Electroporation to deliver plasmid into P. aeruginosa was performed similarly as described earlier (55).Briefly, 2 mL overnight grown culture was spun down in a tabletop microcentrifuge for 1 min at room temperature.The supernatant was discarded, and the pellet was washed twice with 800 µL of ddH 2 O.The final pellet was dissolved in an appropriate volume of ddH 2 O, typically 100 to 500 µL.One-microliter amounts of plasmid were added to 100 µL of cells in 2-mm cuvettes, and electroporation was performed using a Bio-Rad GenePulser with the following settings: 25 µF, 400 Ω, 1.8 kV.All steps described were performed at room temperature.LB broth (1 mL) was added immediately after the pulse, and cells were incubated for 2 h at 37°C with shaking before plating on selective plates.

Western blots
Experiments were performed as described previously (56).Bacteria were cultivated as described for the bacterial killing assay.Proteins in 900 µL culture supernatant were precipitated by TCA/acetone.Primary antibodies were used at a final concentration of 1 µg/mL in 5% milk in Tris-buffered saline (pH 7.4) containing 0.1% Tween (TBST).Secondary antibodies were incubated for 1 h with horseradish peroxidase-labeled anti-rabbit antibody (Jackson Lab), washed with TBST, and peroxidase was detected by LumiGLO Chemiluminescent Substrate (Cell Signaling Technology, USA) on a gel imager (GE ImageQuant LAS 4000).To monitor protein abundance in cell pellets, 100 µL of culture was centrifuged, and the supernatant was discarded.For detection of Hcp1 in cell pellets, 100 µL cells was harvested, resuspended in 100 µL Laemmli buffer, and boiled for 10 min at 95°C.Proteins were separated on Novex 4%-12% Bis-Tris SDS-polyacrylamide gel electrophoresis gels (Thermo Fisher Scientific) and were transferred to nitrocellulose membrane for immuno-detection.

Fluorescence microscopy
Procedures similar to those described previously (21,31,36) were used to detect fluorescence signal in P. aeruginosa.Overnight cultures of P. aeruginosa were washed in LB, diluted 1:100 into fresh medium, and cultivated for 2.5-3.0 h to an OD at 600 nm of about 0.8-1.2.Cells from 1 mL of the culture were resuspended in 50-100 µL of fresh LB, concentrated to OD 10 to achieve high cell density.For low cell density imaging without cell-cell contact, cell suspension with OD ~ 1 was used.Cell suspensions were placed on a thin pad of 1% agarose in LB and were covered with a glass coverslip.Furthermore, the cells under coverslip were incubated at 37°C for 1 h and then imaged.Cells close to the periphery of the agarose pad were imaged.Nikon Ti-E inverted motorized microscope with Perfect Focus System and Plan Apo 100× Oil Ph3 DM (NA 1.4) objective lens were used.Spectra X light engine (Lumencor), ET-GFP (Chroma 49002), and ET-mCherry (Chroma 49008) filter sets were used to excite and filter fluorescence.The microscope was equipped with SPECTRA X light engine (Lumencor), and ET-EGFP (Chroma #49002) and ET-mCherry (Chroma #49008) filter sets were used to excite and filter fluorescence.The setup further contained a sCMOS camera pco.edge 4.2 (PCO, Germany) (pixel size 65 nm) and VisiView software (Visitron Systems, Germany) to record images.Temperature control was set at 30°C, and humidity was adjusted to 95% by an Okolab T-unit (Okolab).For time-lapse imaging of bacterial mixtures (Fig. 3A; Fig. S3A and S5A), the cultures were prepared and concentrated to OD 10 (as described above) and mixed at a 1:1 ratio.Then, 2 µL of this mixture was spotted on a thin pad of 1% agarose in LB, covered with a coverslip and incubated at 37°C for 1 h before imaging.

Image analysis
Fiji (57) was used for all image analysis and manipulations as described previously (31).For quantification of T6SS activity, total cell number was assessed using the "find maxima" options with noise tolerance adjusted manually for each case.All quantifications were verified manually and carried out using the "edge maxima exclusion" function.For quantification of T6SS activity from time-lapse movies, the "temporal color code" function was used.For high-density conditions, three different 15 × 15 µm fields were used; for low-density conditions, at least 10 different 50 × 50 µm fields were used for counting.The total number of cells analyzed (n) is indicated in the graphs (Fig. 1C).For quantification of H1-T6SS activity in ΔretS and ΔretS ΔtssL2 mutants, five different 130 × 130 µm fields were used, and the total number of analyzed cells is indicated in Figure 1E.The total number of cells was estimated using the Find Maxima function with a prominence of 3000 on the phase contrast channel, and the amount of TssB1-mNG assemblies was estimated with the Find Maxima function with a prominence of 700 on the GFP channel.For quantification of H1-T6SS response to H2-T6SS and spontaneous H1-T6SS activity in mixtures, "cell counter" plugin was used.Quantification of ClpV2-mNG foci in monoculture (Fig. 6A) was performed with TrackMate v7.11.1 (58) using the DoG detector mode with an estimated blob diameter of 0.25 µm and a threshold of 20, using median filtering and subpixel localization.Tracks were calculated with the Simple LAP tracker tool with a linking maximum distance of 0.2 µm, a gap-closing maximum distance of 0.4 µm, and a gap-closing maximum frame of 1. Tracks with a maximum duration of more than 12 frames (2 min) were excluded.The remaining tracks in each field of view were normalized by the amount of cells in the field, as estimated by the Find Maxima function with a light background (phase contrast channel) with a prominence of 3000.All imaging experiments were performed with at least three to five biological replicates.

Statistics
Ordinary one-way analysis of variance (ANOVA) with multiple comparisons and Tukey post hoc test was used to determine the significance between all groups using GraphPad Prism version 9.3.1.Two-tailed Student's t-test was performed when comparing two means.If not indicated differently, data are represented as mean ± standard deviation (SD).

FIG 1
FIG 1 Inactivation of H2-T6SS causes significant reduction in H1-T6SS assembly.Representative images of cells with TssB1-mNG and TssB2-mCherry2 fluorescence signals in (A) a low-density field and (B) high-density fields in a wild-type and ΔretS genetic background.Scale bar is 2 µm.(C) Quantification of cells showing dynamic TssB1-mNG or TssB2-cherry2 foci in high-density and low-density fields.The data are represented as mean ± standard deviation of three independent biological replicates.*P value < 0.1; ns, non-significant; two-tailed Student's t-test.(D) Representative images of cells with TssB1-mNG are depicted for each indicated strain.Scale bar is 2 µm.(E) Quantification of H1-T6SS activity in ΔretS and ΔretS ΔtssL2 mutants, complemented or not with a pME6032-TssL2 plasmid.The strains were spotted on 1% LB-agarose pads and were incubated for 1 h before imaging.The data are represented as mean ± standard deviation of 8-10 fields of view from two independent biological replicates.*P value < 0.05; ordinary one-way ANOVA with multiple comparisons and Tukey post hoc test.

FIG 2
FIG 2 Hcp1 secretion and H1-T6SS retaliation to V. cholerae T6SS is independent of H2-T6SS.(A) Presence of Hcp1 detected in culture supernatant and cell pellets.The molecular weight is indicated on the left in kilodaltons.(B) Summary of competition assays showing recovery of T6SS-positive V. cholerae 2740-80 prey strain after co-incubation with different P. aeruginosa attacker strains.(C) Summary of competition assays showing recovery of T6SS-negative V. cholerae 2740-80 Δhcp prey strain after co-incubation with different P. aeruginosa attacker strains.Data are presented as mean of log 10 (CFU/mL) of recovered V. cholerae 2740-80 strains.Error bars represent standard deviation of three independent replicates.*P value < 0.1; ns, non-significant; ordinary one-way ANOVA with multiple comparisons and Tukey post hoc test.

FIG 7
FIG 7 The T6SS of A. baylyi and V. cholerae induces H1-T6SS retaliation independently of known effectors.(A) Survival of A. baylyi ADP1 strains during competition with H1-T6SS-positive or negative P. aeruginosa.Error bars represent standard deviation of three independent replicates.****P value < 0.0001; ns, non-significant; ordinary one-way ANOVA with multiple comparisons and Tukey post hoc test.Data are presented as mean of log 10 (CFU/mL).(B) Microscopy images showing retaliation events triggered by WT or T6SS-effector-negative A. baylyi.(C) Survival of V. cholerae 2740-80 strains incubated with H1-T6SS-positive or negative P. aeruginosa.Error bars represent standard deviation of three independent replicates.**P value < 0.01, ****P value < 0.0001; ns, non-significant; ordinary one-way ANOVA with multiple comparisons and Tukey post hoc test.Data are presented as mean of log 10 (CFU/mL).(D) Microscopy images showing retaliation events triggered by WT or T6SS-effector inactivated V. cholerae.

FIG 8
FIG 8 Model of H2-T6SS-mediated triggering of H1-T6SS activity and dueling in P. aeruginosa.The H2-T6SS sheath contraction in initiator P. aeruginosa cell (green shade) powers the delivery of H2-Hcp and H2-VgrG spike into neighboring sister cell (magenta shade).Consequently, the H1-T6SS is triggered in the neighboring cell, which delivers the H1-Hcp and H1-VgrG into the initiator cell.Finally, the H1-T6SS dueling response is triggered in the initiator cell.The sheath contraction events coincide with the respective ClpV-sheath interactions.