The phospholipase effector Tle1^Vc promotes Vibrio cholerae virulence by killing competitors and impacting gene expression

ABSTRACT

Vibrio cholerae utilizes the Type VI secretion system (T6SS) to gain an advantage in interbacterial competition by delivering anti-prokaryotic effectors in a contact-dependent manner. However, the impact of T6SS and its secreted effectors on physiological behavior remains poorly understood. In this study, we present Tle1^Vc, a phospholipase effector in atypical pathogenic V. cholerae E1 that is secreted by T6SS via its interaction with VgrG1^E1. Tle1^Vc contains a DUF2235 domain and belongs to the Tle1 (type VI lipase effector) family. Bacterial toxicity assays, lipase activity assays and site-directed mutagenesis revealed that Tle1^Vc possessed phospholipase A₁ activity and phospholipase A₂ activity, and that Tle1^Vc-induced toxicity required a serine residue (S356) and two aspartic acid residues (D417 and D496). Cells intoxication with Tle1^Vc lead to membrane depolarization and alter membrane permeability. Tli1^tox-, a cognate immunity protein, directly interacts with Tle1^Vc to neutralize its toxicity. Moreover, Tle1^Vc can kill multiple microorganisms by T6SS and promote in vivo fitness of V. cholerae through mediating antibacterial activity. Tle1^Vc induces bacterial motility by increasing the expression of flagellar-related genes independently of functional T6SS and the tit-for-tat (TFT) response, where Pseudomonas aeruginosa uses its T6SS-H1 cluster to counterattack other offensive attackers. Our study also demonstrated that the physical puncture of E1 T6SS can induce a moderate TFT response, which is essential to the Tle1^Vc-mediated strong TFT response, maximizing effector functions. Overall, our study characterized the antibacterial mechanism of phospholipase effector Tle1^Vc and its multiple physiological significance.

KEYWORDS:

• Vibrio cholerae
• T6SS
• phospholipase effector
• tit-for-tat
• motility

Introduction

Vibrio cholerae, a gram-negative bacterium, is the causative pathogen of cholera, a severe infectious disease.¹ They can be divided into toxigenic and non-toxigenic strains based on whether they carry the cholera toxin ctxAB genes.^2,3 The main mode of cholera transmission is through the consumption of water and food contaminated with toxigenic V. cholerae, with the typical symptom being watery diarrhea that can lead to death in severe cases. Secretion systems, such as the type III secretion system and the type VI secretion system (T6SS),^4,5 are widely used by bacteria to overcome host and competitor challenges. T6SS is a bacterial nanoweapon that can kill or inhibit adjacent cells by injecting toxic effectors through the telescopic structure of a needle tube⁶. T6SS enables microorganisms to obtain a competitive advantage over competitors in the same ecological niche, enhancing their environmental adaptability and pathogenicity to the host.⁷

T6SS was first discovered in Pseudomonas aeruginosa and V. cholerae in 2006^8,9 and is present in approximately 25% of all sequenced gram-negative bacteria.¹⁰ The T6SS nanomachine can be divided into three main parts: the membrane complex, baseplate, and contractile sheath, which perform injection functions.⁶ These structures are conserved among different species.^11,12 The tubular structure can be divided into an Hcp inner tube and outer tube composed of VipA/B.¹³ The inner Hcp tube can be secreted by the contraction of VipA/B.^14,15 Fast injection of the Hcp inner tube delivers the tip structure, including VgrG, PAAR protein, and accompanying effectors, into the target cell.^16,17 The killing effect of the T6SS is mainly attributed to the effectors loaded on the tip structure. Although effectors are extremely diverse, catalytic activity can be generally classified into lipase, peptidoglycan-degrading enzymes, proteases, nucleases, and pore-forming effectors, etc., which can exert their corresponding cytotoxic effect after being injected into neighboring cells.¹⁸ These effectors are secreted by T6SS in diverse ways, including covalently or non-covalently binding to VgrG, Hcp, and PAAR to achieve transportation. Occasionally, these interactions are aided by specific adaptors or chaperones.

Type VI lipase effectors (Tle) are highly abundant and can be found in a broad spectrum of bacterial species.^17,19–21 Tle can be divided into five divergent families (Tle1–5).¹⁹ Tle1 and Tle2 are enriched in aquatic metagenomes, especially in Vibrio species, and have been shown to be important for interbacterial competition in the complex microbial community.²⁰ In addition to their antibacterial activity, lipase effectors also regulate multiple physiological roles. For example, Tle1^AH and Tle2^Vc have been shown to affect biofilm formation and trigger tit-for-tat (TFT) responses in which Pseudomonas aeruginosa uses its T6SS encoded by T6SS-H1 cluster to counterattack other offensive attackers, respectively.^21,22

T6SS effector proteins belonging to the Tle1 family have been extensively studied in Gram-negative bacteria.¹⁹ Previous investigations have primarily focused on the antibacterial activity of Tle1, along with exploring other biological functions such as their impact on biofilm formation and colonization in host.^19,21,23,24 It is worth noting that the enzymatic activity and antibacterial mechanisms of Tle1 family proteins encoded by different bacteria can vary significantly. For instance, Tle1^BT of Burkholderia thailandensis BTH_I2698 and Tle1^PA of Pseudomonas aeruginosa have been shown to possess phospholipase A₂ (PLA₂) activity,^19,25 while Tle1^EAEC of entero-aggregative Escherichia coli exhibits both phospholipase A₁ (PLA₁) activity and a lower level of PLA₂ activity.²⁴ However, limited information is available regarding the molecular mechanism underlying Tle1^Vc-mediated killing activity in non-toxigenic V. cholerae strains, as well as its biological roles in virulence and in vivo fitness within host organisms. In this study, we report a lipase effector Tle1^Vc in an atypical pathogenic O1 serogroup V. cholerae E1 lacking the typical virulence factors (cholera toxin and toxin co-regulated pilus), but encoding a type III secretion system responsible for intestinal colonization of non-O1, non-O139 V. cholerae.^26,27 Tle1^Vc is a phospholipase effector with phospholipase A₁ activity and A₂ activity secreted by T6SS and is predicted to contain two conserved motifs in addition to the known GXSXG motif, which are essential for its catalytic toxicity. We also found that an immunity protein, Tli1^tox- can specifically abolish Tle1^Vc-mediated toxicity by direct interaction. It is possible that Tle1^Vc is secreted through noncovalent binding to the structural protein VgrG1^E1. T6SS-mediated delivery of Tle1^Vc demonstrated broad-spectrum antibacterial activity, providing V. cholerae E1 with a competitive advantage against multiple microorganisms. Colonization analysis revealed that V. cholerae E1 can colonize the adult mouse cecum and utilize its T6SS to enhance in vivo fitness. Moreover, our study found that Tle1^Vc regulates bacterial motility and elicits a strong tit-for-tat response from P. aeruginosa PAO1. Notably, physical puncturing of the T6SS can induce a moderate TFT response, which is critical to the Tle1^Vc-induced robust TFT response. Overall, our investigation has unveiled the vital physiological roles played by the T6SS of V. cholerae E1 and its substrate, Tle1^Vc, which is a member of the phospholipase Tle1 family.

Materials and methods

Bacterial strains and growth conditions

All the strains and plasmids used in this study are listed in Table S1. Strains were grown at 37°C in Luria – Bertani medium or with 1.2% agar. The final concentrations of antibiotics in the growth media were 10 μg/mL chloramphenicol (Cm) for E. coli, 2 μg/mL Cm for V. cholerae, 100 μg/mL ampicillin (Amp), 50 μg/mL kanamycin (Km), and 10 μg/mL nalidixic acid (Nal).

Plasmid construction

Primers used in this study are listed in Table S2. Gene deletion and chromosomal mutation of V. cholerae were performed using the suicide plasmid pWM91 or pDS132 containing a sacB counter-selectable mark.^28,29 Plasmid pK18 was used for gene deletion in Pseudomonas aeruginosa PAO1.³⁰

The tle1^Vc (Gene accession number: OQ980252) and vgrG1^E1 genes were directly cloned into the pBAD24 plasmid for complementation (cytoplasmic expression).³¹ The effector and its variants were also fused with an N-terminal periplasmic Tat (twin-arginine translocation pathway) signal (periplasmic expression) into the pBAD24 plasmid for V. cholerae and E. coli or the pPSV37 plasmid³² for P. aeruginosa PAO1. Immunity proteins were cloned into the pSRKKm vector for the complementation assays.³³ When required, these genes were introduced into the pETM3C plasmid with a 6× His purification tag for protein purification.

Bacterial competition assays

Overnight cultures of bacterial predators and prey cells were diluted to an OD₆₀₀ of 1.0 and mixed in a 1:1 (predator: prey) ratio. Mixtures (5 μL) were spotted onto 0.22 μm polyvinylidene fluoride (PVDF) membranes overlaid on an LB agar plate and incubated at 37°C. After 4 h, cells were resuspended in 1 mL of 0.8% NaCl, and 5 μL of serially diluted cells was spotted onto selective LB agar plates to determine the number of surviving prey.

Protein secretion and Western blots analysis

Overnight bacterial cultures were subcultured in fresh LB and grown to OD₆₀₀ 0.9 at 37°C. The culture was centrifuged, and the supernatant was filtered through a 0.22 μm PVDF filter. For each 1.8 mL of culture supernatant, 200 μL of 100% trichloroacetic acid was added, and the mixture was incubated on ice for 30 min before centrifugation at 20,000 g for 15 min. The resulting precipitate was then washed twice with ice-cold acetone. Protein samples were separated by 12% SDS-PAGE gel and transferred onto PVDF membrane, which was then blocked with 5% nonfat milk. The proteins were probed with anti-rabbit Tle1^Vc, anti-rabbit Hcp, and anti-mouse RpoB (Biolegend) primary antibodies overnight at 4°C, and further incubated with fluorescent secondary antibodies (LI-COR, Odyssey). The signal was screened using a two-color infrared laser imager. Polyclonal antibodies against Hcp and Tle1^Vc were purchased from Shanghai Youlong Biotech.

Bacterial toxicity assays

E. coli BL21(DE3) cells carrying the effectors and their variants were cultured overnight in LB medium. The cultures were then adjusted to an optical density at 600 nm (OD₆₀₀) of 1.0 and serially diluted in 0.8% NaCl at a 10-fold dilution. Subsequently, 5 μL of the resulting dilution was spotted onto LB agar plates containing either 0.2% arabinose or 0.2% glucose, followed by overnight incubation.

For the growth curve assays, bacterial cells expressing both the effector and immunity proteins were grown in LB medium overnight. The cultures were then inoculated at a 1% ratio into LB medium supplemented with 0.2% arabinose and 0.03 mM IPTG. The optical density was measured at the indicated time points.

Expression and purification of proteins

The effector gene tle1^Vc, the immunity genes tli1^tox- and tli1^tox+ were cloned into a pETM3C plasmid with a 6× His-tag and transformed into E. coli BL21(DE3). Overnight cultures were inoculated into the LB medium and grown to an OD₆₀₀ 0.6. After adding IPTG to a final concentration of 0.1 mM, the culture was incubated at 16°C for 16 h. Cells were collected by centrifugation, resuspended in lysis buffer (50 mM tris HCl, 100 mM NaCl, 30 mM imidazole, 1 mM PMSF, pH 7.8) and finally lysed by ultrasonication. Recombinant proteins with a 6× His tag were trapped on a Ni²⁺-NTA column (Sangon Biotech). The column containing the target protein was washed twice in wash buffer (50 mM tris HCl, 100 mM NaCl, 30 mM imidazole, pH 7.8) and eluted with elution buffer (50 mM tris HCl, 100 mM NaCl, 300 mM imidazole, pH 7.8). The eluted proteins were further purified using the AKTA FPLC system.

Enzyme activity assays

The lipase activity of purified Tle1^Vc and its inactivated mutants was detected using Tween 20 as degrading substrate.^34,35 The enzyme activity using protein (50 ng) and Tween 20 as the substrate was assayed at 37°C in a 200 µL reaction buffer containing 25 mM Tris-HCl, pH 7.8), 1.8% (v/v) Tween 20, and 3 mM CaCl₂. The optical density at 500 nm was measured over a 60-min period.

For phospholipase A activity assays were performed using PED-A1 (sn-1-labeled) and PED6 (sn-2-labeled) fluorescent substrates according to manufacturer’s directions (Invitrogen). Activity of Tle1^Vc and its variants on these substrates was measured at an enzyme concentration of 3 nM at 37°C.

Analytical gel filtration chromatography coupled with static light scattering

Molar mass measurements were performed using an AKTA FPLC system (GE Healthcare) coupled with a static light scattering detector (mini-DAWN, Wyatt) and a differential refractive index detector (Optilab, Wyatt). Protein samples (50 μM) were loaded onto a Superdex 200 Increase 10/300 GL column (GE Healthcare) pre-equilibrated with a buffer (50 mM Tris-HCl (pH 7.5), 100 mM NaCl, 1 mM EDTA, and 1 mM DTT). The effector and immunity proteins were mixed at a 1:1 ratio.

Isothermal titration calorimetry (ITC) assay

ITC experiments were performed using a MicroCal ITC200 calorimeter (Malvern, UK) at 25°C. Proteins at high concentrations (100 μM) were loaded into the syringe and titrated into cells containing the corresponding interactors at low concentrations (10 μM). The sample in the syringe was subsequently injected into the cell with a time interval of 120 s (0.5 μL for the first injection and 2 μL each for the following 18 injections). Titration data were analyzed using Origin 7.0 software and fitted with the one-site binding model.

Surface plasmon resonance (SPR) experiments

Surface plasmon resonance (SPR) experiments were conducted to measure the binding kinetic parameters between proteins. The surface of a CM5 sensor chip (GE Healthcare) was activated using NHS and EDC (1:1, v/v), followed by immobilization of the protein through standard amine-coupling at 25°C. The ligand was diluted to a concentration of 10 μg/μL in sodium acetate buffer (pH 4.0), and the protein immobilization level was set to approximately 1000 RU (response unit; 1 RU is roughly equivalent to a 1 pg/mm² change in the concentration of the bound substance on the chip surface). Subsequently, ethanolamine was used to block any unreacted NHS ester groups. To assess the activity of the immobilized protein on the chip surface, a potential interactive protein was utilized, and regeneration was performed using glycine hydrochloride at pH 1.5.

Different concentrations of analytes were sequentially injected into the channel to evaluate the binding kinetic parameters. A reference channel was activated and blocked to account for nonspecific binding of analytes to the chip surface. To remove any remaining sample from the different analytes, the chip was washed twice with PBS. The on-rates and off-rates of the compounds were determined by fitting the data to a 1:1 Langmuir binding model. Kinetic and affinity measurements were calculated based on the data after subtracting the reference channel signal using the Biacore T200 Evaluation Software (GE Healthcare).

Motility assays

Overnight cultures were adjusted to OD₆₀₀ ~1.0 and the 1 µL suspension was spotted on the center of the 0.3% agar LB plates containing appropriate antibiotics and 0.2% arabinose, then incubated at room temperature. The diameter of the migration areas was measured at the indicated time points.

RT-qPCR analysis

Bacterial cells grown on 0.3% agar LB plate were collected and total RNA was extracted using the RZ solution (TIANGEN). Single-stranded cDNA was synthesized from 1 μg of total RNA using HiScript® III RT SuperMix (Vazyme, R323) with random hexamers as primers in a 20 μL reaction mixture. The resulting cDNA mixture was diluted to 10 ng/μL as a template for the subsequent quantification assays. Reverse transcription-quantitative PCR (RT-qPCR) was performed using a LightCycler® 96 SW (Roche) with specific primers, as described in Table S2. To standardize the results, the relative abundance of V. cholerae 16S gene was used as the internal standard. The fold change in gene transcription was determined using the comparative threshold cycle (CT) method.³⁶

Electron microscopy

Overnight bacteria were inoculated into 0.3% agar LB plate containing appropriate antibiotics and 0.2% arabinose and incubated at room temperature overnight. Strains were collected, adsorbed on a carbon type B 300-mesh copper grid (Ted Pella, Inc.), and stained with 3% uranyl acetate. Samples were imaged using a 120 kV transmission electron microscope (Talos L120C G2).

Fluorescence microscopy imaging assays

BL21(DE3) cells periplasmically expressing Tle1^Vc and its catalytic mutant were grown to OD₆₀₀ 1.0, resuspended in 0.5× PBS and spotted on 1.5% agarose-LB pads containing 0.2% arabinose and fluorescence dyes. DiBAC₄(3) and SYTOX blue, two commonly used cell membrane integrity dyes, were used to evaluate the dynamic process of bacterial death³⁷. DiBAC₄(3) (Invitrogen) is sensitive to membrane potential and will aggregate and show a bright green fluorescent signal when the cell membrane loses its potential. SYTOX blue (Invitrogen) can enter cells after the cell membrane loses selective permeability and stain nucleic acids in bright blue. All microscopy images were obtained using a ZEISS LSM900 (Carl Zeiss) inverted microscope with a Definite Focus 2.0 system to acquire stable long-period images and 100× oil immersion objective lens (Carl Zeiss). Each sample was observed continuously for 2 h, and images were captured every 20 s.

Bioinformatics analysis

All nucleotide and protein sequences of the microorganisms were acquired from the National Center for Biotechnology Information (NCBI) database. The genome of V. cholerae E1 can be found in the NCBI database (GenBank accession numbers: chromosome 1: CP110188.1; chromosome 2: CP110189.1). A total of 1338 V. cholerae genomes were downloaded from the NCBI database. BLAST analysis was performed using Tle1^Vc as the query, and the cutoff value for Tle1^Vc-positive was an HMMER bit score of 20 over the overall sequence comparison. Phylogenetic tree reconstruction was performed using MEGA 7.0, and sequence logos were generated using the Weblogo tool.

In vivo colonization assays

The animal experiments were approved by the Ethics Committee on Animal Experimentation of the Southern University of Science and Technology (project code: SUSTech-JY202203087).

The adult mouse model was used as previously described with the following modifications³⁸. Five-week-old CD1 mice were fed water containing streptomycin (0.5%) and aspartame (0.4%) for 12 h before inoculation. For single colonization assays, V. cholerae wild-type and mutant V. cholerae were orally inoculated at 10⁸ CFU into mice pretreated with streptomycin. The feces were collected daily. After 3-day post-infection, mice were sacrificed, and the small intestines, cecum, and large intestines were harvested and homogenized, and colonizing bacteria were counted by plating on LB agar plates. For competitive colonization assays, a 1:1 mixture of the wild type (LacZ⁻) and mutant (LacZ⁺) was orally inoculated with 10⁸ CFU into mice, which were enumerated by plating on LB agar plates containing 40 μg/mL X-Gal (5-bromo-4-chloro-3-indolyl-β-Dgalactopyranoside). The competitive index (i.e., the output ratio of mutant to wild-type over the input ratio of mutant to wild-type) was calculated.

Results

Tle1^Vc is a phospholipase antibacterial effector belonging to the Tle1 family

In a previous study, we isolated an environmental V. cholerae strain E1 with constitutive T6SS expression.²⁷ Using comparative genomic analysis, we predicted a new effector-immunity pair in the Aux1 cluster of E1 (no contain TseL effector) by comparing it to the N16961 reference genome (Figure 1a). Through Pfam database analysis, we predicted that this potential effector contained a DUF2235 domain, which belongs to the phospholipase effector Tle1 subfamily (Figure 1b), and named it Tle1^Vc. We found that V. cholerae Tle1^Vc homologs were mainly distributed in non-toxigenic V. cholerae and rarely in toxigenic V. cholerae (Figure 1c), indicating that Tle1^Vc may be more important in non-toxigenic V. cholerae in complex environments. Phylogenetic analysis revealed that Tle1^Vc is closely related to the known Tle1^AH effector from Aeromonas hydrophila NJ-35 (Figure 1d, left) and that Tle1^AH-mediated antibacterial activity depends on its delivery into the periplasm of the target bacteria.²¹ To determine whether Tle1^Vc functions in a similar way, we expressed Tle1^Vc with or without a twin-arginine periplasmic translocation signal (Tat) in E. coli BL21 cells using an arabinose-inducible pBAD vector. The results showed that Tat-Tle1^Vc, but not Tle1^Vc, was toxic to E. coli BL21 (Figure 1e).

Figure 1. Tle1^Vc is a lipase effector with antibacterial activity.

(a) Comparison of genetic organization of T6SS Aux1 cluster between V. cholerae E1 and N16961. Genes were annotated in the genome database. (b) Functional domain of novel effector was predicted using pfam database. (c) The distribution of Tle1^Vc homologs in V. cholerae. The percentage of Tle1^Vc-positive strain among non-toxigenic (CT-) and toxigenic (CT+) V. cholerae strains was 60.89% and 2.89%, respectively. Related data are available in Data Set S1. (d) Phylogenetic analyses of Tle1^Vc in V. cholerae E1 and representative members of the Tle1 family from different bacteria using MEGA6.0. Maximum-likelihood phylogeny constructed with a bootstrap of 1000. Sequence logo of conservative motifs of Tle1^Vc were shown by WebLogo. (e) Toxicity of expressing Tle1^Vc and its catalytic mutants in E. coli. BL21 cells expressing Tle1^Vc and its variants in cytoplasm (Cyto-Tle1^Vc) and periplasmic space (Tat-Tle1^Vc, Tat-Tle1^VcS356A, Tat-Tle1^VcD417A and Tat-Tle1^VcD496A) were serially diluted onto LB agar plates containing either 0.2% glucose or 0.2% arabinose. The expression of Tle1^Vc and its variants in BL21 cells were detected through western blot (below). RpoB served as an internal reference. The representative image from three independent biological replicates was displayed. (f) The lipase activity assay of Tle1^Vc proteins. Purified Tle1^Vc and its inactivated mutants were incubated with Tween 20 catalytic substrate at 37°C, the OD₅₀₀ was monitored at indicated time points. Error bars represent the mean ±SD of three independent biological replicates. (g and h) The measurement of phospholipase activity of Tle1^Vc. Enzymatic activity of the Tle1^Vc and its protein variants against vesicles containing phospholipid derivatives with fluorescent signal at the sn1 or sn2 positions. (i) Schematic of a phospholipid indicating the activities defined in the effector protein Tle1^Vc.

Our analysis of multiple sequence homology comparisons revealed two conserved motifs, FXGXXD (X is for any amino acid) and EXXXPGXXXDIG, in addition to the known lipase catalytic center motif GXSXG (Figure 1d, right). To determine whether these motifs are essential for Tle1^Vc-induced toxicity, we mutated two aspartic acid residues (D417 and D496) and a serine residue (S356) to alanine within these motifs and tested the toxicity of Tle1^Vc and its variants in the periplasm of E. coli BL21. Heterologous expression of Tat-Tle1^Vc resulted in reduced survival, whereas mutant variants Tat-Tle1^VcS356A, Tat-Tle1^VcD417A, and Tat-Tle1^VcD496A grew as well as the empty control (Figure 1e). Western blot analysis indicated that the expression level of these mutant proteins in E. coli BL21 cells were comparable to wild type (Figure 1e, below), suggesting that loss of catalytic activity was caused by active site mutations rather than instability of proteins expressed. To determine the enzymatic activity of Tle1^Vc, we purified the effector and performed a lipase activity assay using Tween 20 as the catalytic substrate. We also purified the Tle1^Vc mutant alleles for the enzyme activity assay. Tle1^Vc showed significant catalytic activity, whereas the catalytic mutants Tle1^VcS356A, Tle1^VcD417A, and Tle1^VcD496A were unable to hydrolyze the substrate as buffer control (Figure 1f and S1a). Next we asked whether Tle1^Vc possesses phospholipase activity. Using vesicle substrates doped with fluorescent phospholipid derivatives, our results showed that Tle1^Vc had both PLA₂ activity and PLA₁ activity (Figure 1g–l). In summary, these data demonstrated that Tle1^Vc is a phospholipase effector belonging to the Tle1 family, and its toxicity requires two conserved motifs in addition to the known GXSXG motif.

Tle1^Vc alters membrane potential and permeability of target cell

TseL, a lipase effector previously reported in V. cholerae V52, belongs to the Tle2 family and has been shown to impairs cell membrane integrity by degrading phospholipids.^19,22 To investigate whether Tle1^Vc also induces membrane damage, we assessed its effect on membrane potential and permeability using voltage-dependent DiBAC₄(3) dye and SYTOX blue nucleic acid dye, respectively, which can bind to the depolarized membrane and visualize cell death via time-lapse microscopy.³⁷ E. coli BL21 cells carrying Tat-Tle1^Vc, Tat-Tle1^VcS356A, Tat-Tle1^VcD417A, or Tat-Tle1^VcD496A were spotted individually on a thin 1.5% agarose LB pad containing the two dyes and 0.2% arabinose as an inducer of pBAD promoter. Under toxin-inducing conditions, E. coli BL21 cells expressing Tat-Tle1^Vc showed simultaneous green and blue fluorescence, stained by DiBAC₄(3) and SYTOX blue dye, respectively, compared with the empty vector (Figure 2a,b). In contrast, no fluorescence signal was observed in the cells expressing the three variants (Figure 2c–e). Quantitative analysis showed that the percentage of fluorescence signal of cells (a 30 × 30 µm field of at least 30 cells) periplasmically expressed Tle1^Vc protein was notably higher than cells expressed its mutants or vector control (Figure 2f). These data demonstrated that Tle1^Vc triggers membrane depolarization and permeabilization, resulting in cell death.

Figure 2. Tle1^Vc effector impairs the integrity of bacterial membrane.

(a-e) Time-lapse microscopy of E. coli BL21 cells expressing pBAD24 empty plasmid (a), Tat-Tle1^Vc (b), Tat-Tle1^VcS356A (c), Tat-Tle1^VcD417A (d) or Tat-Tle1^VcD496A (e) grown on LB agar pads containing DiBAC₄(3) and SYTOX blue dyes and 0.2% arabinose (induction). Scale bar, 5 μm; The representative image from two independent biological replicates was shown. (f) Quantitative analysis of percentage of green fluorescence of cells expressing Tle1^Vc and its catalytic variants. Statistical significance is indicated (unpaired, Student’s t-test: ***P < .001).

Tli1^tox- specifically neutralizes Tle1^Vc-induced toxicity

Genes encoding T6SS effectors and their corresponding immunity proteins are often found in pairs, with immunity genes located downstream of the effector genes. In the case of tle1^Vc, an unannotated gene was identified downstream of the effector gene, which was designated Tli1^tox- due to its homology to the Tli1^tox+ protein encoded by toxigenic V. cholerae (Figure 1a and S2). To investigate whether Tle1^Vc-induced antibacterial activity could be neutralized by Tli1^tox-, we generated a Δtle1^Vc/tli1^tox- mutant as prey in a competitive killing assay. Co-culturing the prey cells with E1 wild-type, Δtle1^Vc, Δtle1^Vc+pTle1^Vc, and ΔvasK (inactive T6SS) as predators showed that cells lacking the tli1^tox- immunity gene were susceptible to Tle1^Vc positive predators, as the Δtle1^Vc mutant and ΔvasK did not kill the Δtle1^Vc/tli1^tox- prey cells, whereas the wild type and Δtle1^Vc+pTle1^Vc markedly reduced their survival (Figure 3a). However, introducing tli1^tox- into the Δtle1^Vc/tli1^tox- prey cells using an IPTG-inducible pSRKKm vector provided effective protection after induction (Figure 3a). These results demonstrate that Tle1^Vc toxicity can be neutralized by Tli1^tox-.

Figure 3. Tli1^tox- specifically neutralizes the toxicity of Tle1^Vc.

(a) Survival of the Δtle1^Vc/tli1^tox- mutant in competition against a panel of predator strains. The Δtle1^Vc/tli1^tox- mutant containing pSRKKm vector complemented with or without Tli1^tox- as prey competed with predators indicated at 1:1 ratio. Both 50 μM IPTG and 0.1% arabinose were included in LB agar plate to induce corresponding genes expression. Surviving prey cells were serially diluted and determined on the LB plate. Error bars represent the mean ±SD of at least three independent biological replicates. Statistical significance is indicated (unpaired, Student’s t-test: ns, not significant; **P < .01). (b) The effect of homologous immunity proteins on Tle1^Vc toxicity. BL21 cells co-expressed an arabinose-inducible pBAD24 vector or its variants of pBAD24 with Tat-Tle1^Vc and either an IPTG-inducible pSRKKm vector or its derivatives of pSRKKm with Tli1^tox+ or Tli1^tox- were grown in LB medium in presence of 0.1% arabinose and 30 μM IPTG. The OD₆₀₀ was recorded in time points indicated. Error bars represent the mean ±SD of three independent biological replicates. Statistical significance is indicated (unpaired, Student’s t-test: ns, not significant; ****P < .0001). (c) SEC analysis of complex formation between the proteins indicated. Protein sample (50 μM) were mixed at a 1:1 ratio and were separated on a Superdex 200 Increase 10/300 GL column and Molar mass was obtained using an AKTA FPLC system coupled with a static light scattering detector. (d) ITC analysis of the interaction between Tle1^Vc and Tli1^tox-. Proteins with high concentrations (100 μM) were loaded into the syringe and titrated into the cells containing corresponding interactors with low concentrations (10 μM).

Bioinformatics analysis identified a hypothetical protein (Tli1^tox+) encoded by the second gene downstream of tle2^Vc (tseL) within the Aux1 cluster of in toxigenic V. cholerae N16961, which shares ~ 80% identity with Tli1^tox- at the amino acid level (Figure 1a and S2). To determine whether Tli1^tox+ plays a defensive role against Tle1^Vc toxicity, we co-expressed Tli1^tox- and Tli1^tox+ using an IPTG-inducible Ptac promoter and Tat-Tle1^Vc using an arabinose-inducible PBAD promoter in E. coli BL21 cells. The results showed that cells expressing Tli1^tox- grew well under Tle1^Vc-induced toxic conditions as compared with the empty control, whereas cells expressing Tli1^tox+ failed to do so (Figure 3b), indicating that Tli1^tox- specifically neutralizes the toxicity of Tle1^Vc.

Furthermore, we purified the effector and the two immunity proteins (Figure S1) to investigate their interaction with Tle1^Vc using analytical gel filtration chromatography coupled with static light scattering and the isothermal calorimetric titration. When Tle1^Vc, Tli1^tox+, and Tli1^tox- were added individually to the column, their individual unique peaks and molecular weights could be detected by laser scatter. However, when we mixed Tle1^Vc with Tli1^tox- and loaded the mixture into the column, the peaks representing individual proteins disappeared, and a new peak with a higher molecular weight appeared. The new complex peak corresponded to a molecular weight of 113.3 kDa, which was approximately equal to the combined molecular weight of Tle1^Vc (84.5 kDa) and Tli1^tox- (30.7 kDa) (Figure 3c). In contrast, Tli1^tox+ failed to form a stable complex with Tle1^Vc (Figure S3a). Moreover, the intensity of interaction was measured by the Isothermal Titration Calorimetry (ITC) assays. Tle1^Vc and Tli1^tox- binding was clearly endothermic and tight, at least in the low nanomolar K_D range (~0.35 ± 0.20 nM) (Figure 3d), while no significant interaction was observed between the Tli1^tox+ and Tle1^Vc (Fig. S3B). These data suggested that Tli1^tox- can specifically protect cells from Tle1^Vc-induced intoxication by direct interaction with Tle1^Vc.

Tle1^Vc delivery requires T6SS and the carrier VgrG1^E1

To confirm whether functional T6SS is necessary for Tle1^Vc secretion, western blotting was performed to detect Tle1^Vc secretion in both the wild-type strain and the ΔvasK (T6SS-inactive) strain. The results showed that Tle1^Vc was present in the pellets of both strains, but only in the supernatant of the wild type, indicating that Tle1^Vc secretion strictly depends on functional T6SS (Figure 4a).

Figure 4. Tle1^Vc delivery requires functional T6SS and VgrG1^E1.

(a) The role of T6SS on Tle1^Vc secretion. Wild-type and ΔvasK mutant were grown to OD₆₀₀ 1.2; the pellets and supernatant were collected for western blot analysis. The Tle1^Vc polyclonal antibody was used to detect the Tle1^Vc secretion and RpoB served as a loading control. (b and c) The effect of VgrG1^E1 on Tle1^Vc-mediated killing activity. The E. coli MG1655 (b) and Δtle1^Vc/tli1^tox- mutant (c) as prey was co-incubated with predator indicated at 1:1 ratio. When needed, 0.1% arabinose were included in LB medium to induce vgrG1^E1 expression. Error bars represent the mean ±SD of three independent biological replicates. Statistical significance is indicated (unpaired, Student’s t-test: ns, not significant; *P < 0.05; **P < .01). (d) The interaction analysis of Tle1^Vc and VgrG1^E1. The protein Tle1^Vc was immobilized on a CM5 sensor chip. After coupling, ethanolamine was used to block unreacted NHS ester groups. Indicated gradient concentrations of VgrG1^E1 and T6SS effector TseV (unpublished data) as negative control were serially injected into the channel to evaluate the binding kinetic parameters. A reference channel was only activated and blocked to normalize results.

Since T6SS effectors require interaction with adjacent structural proteins such as PAAR, Hcp, and VgrG⁶, we assessed the effect of the upstream protein VgrG1^E1 on Tle1^Vc delivery via bacterial competition assays. As expected, the ΔvgrG1^E1 mutant, which retained T6SS activity, abolished the killing effect of the effector-immunity mutant Δtle1^Vc/tli1^tox- compared to the wild type. However, complementation of the ΔvgrG1^E1 mutant with vgrG1^E1 restored the killing ability of Δtle1^Vc/tli1^tox- prey (Figure 4c), indicating that VgrG1^E1 acts as a carrier for Tle1^Vc secretion. To further confirm this conclusion, VgrG1^E1 and Tle1^Vc were purified (Figure S1a,c), and their interaction intensity was analyzed using the Surface Plasmon Resonance (SPR) assays. Results showed that Tle1^Vc can bind to VgrG1^E1 with a K_D of 2.03 μM, but can not interact with negative control TseV (an unpublished T6SS effector) (Figure 4d). The results indicated that Tle1^Vc could directly bind to VgrG1^E1 for delivery.

T6SS-delivered Tle1^Vc contributes to Vibrio cholerae intra-/inter- species antagonism

Previous studies have demonstrated that the Type VI secretion system (T6SS) plays a crucial role in intra- and inter-species competition in Vibrio cholerae.^39,40 In V. cholerae, T6SS function was abolished by completely deleting multiple effectors but not by point mutagenesis, which inactivates multiple effectors.⁴¹ Among the four effectors encoded by V. cholerae E1, Tle1^Vc is a lipase effector, VasX is a pore-forming effector, VgrG3^cp targets the cell wall, and TseV is an uncharacterized effector.^27,42 To assess the role of Tle1^Vc in bacterial competition, we created a mutant strain (4eff) that inactivated the catalytic mutants of vgrG3^cpD867A27 and tle1^VcD417A, deleted vasX^ΔC1642, and deleted tseV. This mutant strain retained T6SS secretion but abolished the antibacterial killing toward effector-cognate immunity mutants (Figure S4a,b). In competitive killing assays, we co-incubated the 3eff predator strain (only Tle1^Vc remained active) and the 4eff predator strain with prey cells, such as V. cholerae C6706, V. vulnificus, V. parahaemolyticus 1360, E. coli MG1655, A. hydrophila, and E. aerogenes. The results demonstrated that the 3eff predator strain effectively killed all the prey cells tested, whereas the 4eff strain exhibited killing activity similar to that of the ΔvasK mutant in all competition assays (Figure 5a–f). Therefore, the T6SS-secreted Tle1^Vc effector can promote intra- and inter-species competition in Vibrio cholerae E1.

Figure 5. Tle1^Vc contributes to Vibrio cholerae intra- and interspecies antagonism.

(a – f) V. cholerae C6706 (a), V. vulnificus (b), V. parahaemolyticus 1360 (c), E. coli MG1655 (d), A. hydrophila (e), E. aerogenes (f) as prey competed with V. cholerae E1 predator indicated at the ratio of 5:1 (predator: prey). Surviving prey cells were serially diluted and determined on the LB plate containing antibiotics. Error bars represent the mean ±SEM of at least three independent biological replicates. Statistical significance is indicated (unpaired, Student’s t-test: *P < .05; **P < .01; ***P < .001; ****P < .0001).

Physical puncture of T6SS and catalytic activity of effector are essential to Tle1^Vc-induced strong tit-for-tat response of P. aeruginosa

To test whether Tle1^Vc could induce a tit-for-tat (TFT) response, we performed experiments with Pseudomonas aeruginosa PAO1, which uses its T6SS to counterattack other attacking bacteria.⁴³ The results showed that a tle1^Vc-inactivated strain of V. cholerae led to a higher number of surviving CFUs in PAO1 than in the wild type, whereas complementation with tle1^Vc in the inactivated strain reduced the survival number to that of the wild-type group (Figure 6a). When the T6SS-H1 cluster of PAO1 was inactivated, we failed to observe counterattack from PAO1 ΔtssB1 after a T6SS attack by V. cholerae E1 wild-type and Δtle1^Vc+pTle1^Vc strains (Figure 6a). This suggests that Tle1^Vc induces a strong TFT response in PAO1 via its T6SS-H1 cluster counterattack. We further evaluated the toxicity of Tle1^Vc on PAO1 and found that cells expressing Tat-Tle1^Vc grew poorly compared to the other Tat-Tle1^Vc variants and the Tat-GFP control (Figure 6b). However, the surviving CFU of cells expressing Tat-Tle1^Vc in LB-agar plates were comparable to those of the three Tat-Tle1^Vc variants and Tat-sfGFP control (Figure 6c). We also tested whether endogenous Tle1^Vc was sufficient to stimulate the TFT response of PAO1. The results showed that Tle1^Vc could not directly induce a TFT response (Figure 6d), which differs from the lipase effector TseL (or Tle2^Vc).⁴⁴ Thus, we speculate that the Tle1^Vc-induced strong TFT response may require the presence of physical puncture or the synergy of other T6SS effectors. We used the multiple-effector-inactivation mutant 4eff (retained T6SS physical puncture and lost the killing assay; Figure S4a,b) and ΔvasK (no physical puncture) to test the first hypothesis. V. cholerae E1 4eff and ΔvasK were co-incubated with PAO1 expressing pPSV37 empty vector or its derivatives of Tle1^Vc. As expected, the CFU of surviving 4eff after a T6SS attack by P. aeruginosa expressing Tat-Tle1^Vc obviously decreased compared to the empty vector control, whereas the CFU of surviving ΔvasK after the same T6SS attack did not change (Figure 6e), suggesting that T6SS physical puncture is essential for Tle1^Vc-induced strong TFT response. Notably, a single physical puncture can directly stimulate the TFT response of PAO1, but the effect was moderate compared to the presence of the effector Tat-Tle1^Vc (Figure 6e). We further assessed the role of the catalytic activity of Tle1^Vc on TFT response. The survival of V. cholerae E1 4eff and ΔvasK after a T6SS attack by PAO1 expressing three Tle1^Vc mutant variants were examined, and the results indicated that the CFU of surviving 4eff was obviously increased in the three Tle1^Vc variant groups compared to the Tle1^Vc group and was comparable to that of the empty plasmid group (Figure 6e), suggesting that the catalytic activity of Tle1^Vc was also found to be vital to TFT response. Collectively, these data demonstrate that Tle1^Vc triggers TFT response, which is dependent on its catalytic activity, the T6SS-H1 cluster of PAO1, and physical puncture of V. cholerae E1 T6SS.

Figure 6. Tle1^vc-induced strong tit-for-tat response requires the T6SS-H1 of PAO1, catalytic activity of effector and the physical puncture of T6SS.

(a) Surviving CFU of V. cholerae strains indicated after a T6SS attack by PAO1. V. cholerae strains wild type, Δtle1^Vc, Δtle1^Vc+pTle1^Vc (complemented with tle1^Vc), and ΔvasK were co-incubated either with P. aeruginosa PAO1 WT or ΔtssB1 (inactivated T6SS-H1 cluster) mutant at the ratio of 5:1 (P. a: V. c) on LB agar plate for 4 h. Surviving V. cholerae cells were serially diluted and determined on the LB plate containing antibiotics. Error bars represent the mean ±SD of three independent biological replicates. Statistical significance is indicated (unpaired, Student’s t-test: ns, not significant; *P < .05). (b) Growth of P. aeruginosa cells expressing the V. cholerae Tle1^Vc or its inactivated mutant in LB media supplemented with a final concentration of 100 μM IPTG. Cells expressing Tat-sfGFP were used as control. (c) Survival of P. aeruginosa periplasmically expressing the indicated effector. Cells of serial-dilution were used to compare. Representative image of plates is shown. (d) Quantification of surviving V. cholerae E1 ΔvasK after a T6SS attack by P. aeruginosa periplasmically expressing pPSV37 empty vector or its variants with effectors indicated. Statistical significance is indicated (unpaired, Student’s t-test: ns, not significant). (e) Surviving V. cholerae E1 (4eff or ΔvasK) after a T6SS attack by P. aeruginosa expressing the indicated effector. 4eff (maintained physical puncture) and ΔvasK (lost physical puncture) as prey. Statistical significance is indicated (unpaired, Mann-Whitney test: **P < .01; ***P < .001).

Tle1^Vc mediates motility of Vibrio cholerae E1 by controlling flagellar synthesis

In addition to its antibacterial and anti-eukaryotic effects, T6SS also plays a role in the function of multiple virulence factors.^45–47 The effect of Tle1^Vc on bacterial motility and biofilm formation was tested, and the results showed no significant difference in biofilm formation between the wild-type and Δtle1^Vc strains. However, the motility of the Δtle1^Vc mutant was noticeably decreased compared to that of the wild-type, and this phenotype could be restored after the Δtle1^Vc mutant was complemented with tle1^Vc (Δtle1^Vc +pTle1^Vc) (Figure 7a). Notably, the ΔvasK mutant maintained motility similar to that of the wild-type, indicating that Tle1^Vc-dependent motility does not require T6SS function. Since the catalytic activity of Tle1^Vc is vital to the tit-for-tat response and antibacterial activity, the effect of catalytic activity on motility was further evaluated. The Tle1^Vc-inactivated chromosomal mutant tle1^VcD417A showed attenuated motility compared to the wild type, indicating that Tle1^Vc-mediated motility requires its catalytic activity.

Figure 7. Phospholipase effector Tle1^Vc mediates bacterial motility through modulating flagellar gene expression.

(a) The effect of Tle1^Vc on V. cholerae motility. Overnight cultures of V. cholerae E1 strains wild type, tle1^VcD417A (Tle1^Vc catalytic mutation on genome), Δtle1^Vc, Δtle1^Vc+pTle1^Vc (complemented with tle1^Vc) and ΔvasK were blotted on 0.3% agar LB plate at room temperature. When needed, 0.2% arabinose was added to induce Tle1^Vc expression. Diameter was measured after 36 h. Error bars represent the mean ±SD of at least three independent biological replicates. Representative image of plates was shown in Figure 7a below. Statistical significance is indicated (unpaired, Student’s t-test: ns, not significant; **P < .01; ***P < .001; ****P < .0001). (b) Schematic regulation of the V. cholerae flagellar transcription hierarchy. (c) Quantitative analysis of flagellar gene expression. Cells grown in Figure 7a were collected after 16 h and total RNA was extracted and quantified using indicated primers (Table S2). Error bars represent the mean ±SD of three independent biological replicates. (d) Transmission electron microscopy of V. cholerae E1 strains indicated with 3% uranyl acetate staining. (e) The impact of flagellar genes on V. cholerae Δtle1^Vc motility was investigated. Cell cultures of V. cholerae E1 strains, including the wild type, Δtle1^Vc, Δtle1^Vc+pFlaA (complemented with flaA), and Δtle1^Vc+pFlaE (complemented with flaE), were spotted onto 0.3% agar LB plates at room temperature. In cases where flagellar gene expression was required, 0.2% arabinose was added as an inducer. The diameter of the motility zone was measured after 48 hours. A representative image of the plates can be seen in Figure 7d below. Statistical significance is indicated (unpaired, Student’s t-test: ***P < .001; ****P < .0001). (F) Transmission electron microscopy of V. cholerae E1 WT and its mutants complementing with flagellar genes indicated.

Vibrio cholerae is motile through a single polar flagellum⁴⁸, which is typically composed of three major structural components encoded by multiple genes: the basal body, hook, and filament.^49,50 The expression of V. cholerae flagellar genes is regulated by a four-tiered transcriptional hierarchy⁴⁸ (Figure 7b). Therefore, the effect of Tle1^Vc on flagellar gene expression was examined by qPCR. As expected, the expression of several flagellar genes was downregulated at the transcriptional level in the Δtle1^Vc and tle1^VcD417A mutants compared to that in the wild type (Figure 7c). However, the expression of downregulated flagellar genes was restored to the expression level of the wild-type in the Δtle1^Vc +pTle1^Vc strain (Figure 7c). Electron microscopy imaging assays showed that a single bacterium of the Δtle1^Vc and tle1^VcD417A mutants lost the singular flagellum, whereas the Δtle1^Vc +pTle1^Vc strain recovered the singular flagellum, as observed in the wild type (Figure 7d). To further determine whether Tle1^Vc-mediated motility defect could be restored by complementing downregulated flagellar genes in Δtle1^Vc mutant, we evaluated two representative flagellar genes, flaA and flaE, whose expressions are important for flagellar synthesis of Vibrio cholerae⁵¹. Results of complementation assays showed that expression of flaA or flaE could partially restore motility of Δtle1^Vc (Figure 7e), which was consistent with the result of electron microscopy imaging (Figure 7f), suggesting that the flagellar synthesis is responsible for motility defect mediated by the Tle1^Vc.

Antimicrobial activity mediated by Tle1^Vc contribute to in vivo fitness of V. cholerae E1

Toxigenic V. cholerae is a noninvasive pathogen that colonizes the small intestine through the expression of the toxin-co-regulated pilus (TCP), an essential colonization factor of Vibrio cholerae.⁴⁴ Although environmental V. cholerae E1 lacks known virulence factors such as TCP and CT, it was found to colonize the cecum rather than the small intestine of adult mice (Figure S5a,b). T6SS has been reported to regulate bacterial colonization in the host.^52,53 To investigate whether the T6SS of E1 plays a role in in vivo fitness, we orally infected mice pretreated with streptomycin (to suppress gut resident microbes) with a 1:1 mixture of E1 wild-type (T6SS⁺; LacZ⁻) strain and ΔvasK (T6SS⁻; LacZ⁺) mutant. We observed a significant colonization defect of the T6SS⁻ strain compared to the T6SS⁺ strain in the cecum as well as in the large intestine (Figure S5c,d). As the growth rate of T6SS⁻ mutant is similar to wild type (Figure S5e), suggesting that E1 can utilize its T6SS to increase in vivo competitiveness.

Next, we assessed the role of Tle1^Vc in the gut colonization of E1 by comparing the number of bacteria colonized by the wild-type and Δtle1^Vc mutant strains. We observed a significant colonization defect of the Δtle1^Vc mutant in the cecum compared to the wild-type strain (Figure 8a). As Tle1^Vc mediates both motility and antibacterial abilities in E1, motility plays a significant role in V. cholerae colonization. To investigate the importance of motility, we generated a motility-deficient mutant of E1, designated as ΔflaA, and compared its motility ability to that of the wild type. The results demonstrated that ΔflaA lost its motility ability (Figure 8b). Subsequently, we evaluated the impact of motility on E1 colonization. In comparison to the wild type, the colonization level of ΔflaA in the cecum was similar (Figure 8c), indicating that motility of E1 was not essential for cecum colonization. By performing competitive colonization assays with the wild-type (LacZ⁻) and sensitive Δtle1^Vc/tli1^tox- mutant (LacZ⁺). We found that the number of colonies of the Δtle1^Vc/tli1^tox- mutant gradually decreased over infection time compared to that of the wild-type strain in feces (Figure 8d), and the Δtle1^Vc/tli1^tox- mutant showed a huge colonization defect in the cecum (Figure 8e,f), showing that Tle1^Vc is active in vivo. These results suggest that Tle1^Vc-mediated antibacterial activity contributes to in vivo fitness rather than motility.

Figure 8. The effect of Tle1^Vc on E1’s in vivo fitness and survival of immunity mutant.

(a) The mice pretreated with streptomycin were infected by intragastric inoculation with the WT strain (10⁸ CFU) or the Δtle1^Vc mutant (10⁸ CFU). After 3 days post-infection, colonized bacteria in the cecum were enumerated by serial plating on selective media. Error bars represent the mean ± SD of four independent replicates. Significance was calculated using unpaired, Student’s t-test, * P < .05. (b) Motility assays of V. cholerae E1 were conducted by spotting cultures of the wild type (WT) and ΔflaA mutant onto 0.3% agar LB plates at room temperature. The diameter of the motility zone was measured after 48 hours. Error bars indicate the mean ± SD of at least three independent biological replicates. A representative image of the plates is displayed in Figure 8b below. Statistical significance is denoted (unpaired Student’s t-test: ****P < .0001). (c) The effect of motility on V. cholerae E1 colonization was assessed using an orally infected mouse model following the procedure described in Figure 8a. The mice were pretreated accordingly and subsequently infected with the WT strain and the ΔflaA mutant. Error bars represent the mean ± SD of eight independent replicates. Statistical significance was determined using an unpaired Student’s t-test (ns). (d and e) The competitive colonization of V. cholerae. The mice treated with continuous streptomycin were infected with the 1:1 mixture of the WT strain (lacZ⁻) and the sensitive Δtle1^Vc/tli1^tox- mutant (lacZ⁺). Output cells in feces (d) were detected at day 1 to 3 post-infection; after 3 days post-infection, colonized bacteria in the cecum (e) were counted by serial plating on selective media containing X-Gal. Error bars represent the mean ± SD of six independent replicates. Significance was calculated using unpaired, Student’s t-test, * P < .05; *** P < .001; **** P < .0001. (f) The competitive index (CI) of the Δtle1^Vc/tli1^tox- mutant versus the WT strain in the cecum from Figure 8c. Horizontal line represents mean of the CI.

Discussion

The type VI secretion system (T6SS) delivers effectors into neighboring cells to enhance competitiveness, and the cell membrane is a common target for these effectors. For example, Tle1 of Burkholderia thailandensis and Aeromonas hydrophila, Tle2 and VasX of V. cholerae, Tle1, Tle4 and Tse4 of P. aeruginosa, and Tle1 and Tle5 of K. pneumoniae all have been shown to hydrolyze the membrane phospholipids or act on membrane by pore-forming.^{19,21,25,42,54–56} In this study, we identified a putative T6SS lipase effector Tle1^Vc in Aux1 cluster of the atypical pathogenic V. cholerae E1 based on bioinformatics analysis. By analyzing its amino acid sequence, we identified a conserved DUF2235 domain and lipase catalytic core motif GXSXG, which are widely distributed in Tle1 lipase family. Tle1 proteins have been established as a subfamily of T6SS phospholipase effectors, exhibiting either phospholipase A₂ activity or a combination of phospholipase A₁ and A₂ activities.^19,24 In this study, we demonstrated that Tle1^Vc possesses both phospholipase A₂ activity and a similar phospholipase A₁ activity. Conversely, another Tle1 effector, Tle1^EAEC, has been shown to possess phospholipase A₁ activity and a lower level of phospholipase A₂ activity.²⁴ These findings underscore the importance of characterizing the phospholipase activity profile of the Tle1 family in elucidating the antibacterial mechanisms and biological roles of other Tle1 family effectors.

According to multiple sequence alignment, two Tle1 family conserved motifs FXGXXD and EXXXPGXXXDIG in addition to catalytic center motif GXSXG were identified in Tle1^Vc. By heterologous expression assays, we determined that the S356, D417 and D496 residues of Tle1^Vc were indispensable for Tle1^Vc-catalyzed toxicity, which was consistent with the result of enzyme activity assays. The structure and active domain of Tle1^PA of P. aeruginosa have been shown that the catalytic triad (S235-D279-H377) is vital for its toxicity.²⁵ Although Tle1^PA and Tle1^Vc only share ~ 16% identity in their amino acid sequences, the D417 and S356 residues of FXGXXD and GXSXG motifs corresponds to D279-S235 of Tle1^PA, and the D496 residue of Tle1^Vc and the H377 residue of Tle1^PA are adjacent within EXXXPGXHXDIG motif, showing that Tle1^Vc may use the same catalytic mode to degrade phospholipids as Tle1^PA. After being injected into neighboring cells by the T6SS, Tle1^BT and Tle2^Vc are able to increase cellular permeability by hydrolyzing the membrane phospholipids of target cells.¹⁹ Using time-lapse microscopy, we also demonstrate that Tle1^Vc increased the permeability of cell membrane and caused cell death. While P. aeruginosa PAO1 cells demonstrated differential resistance to Tle1^Vc under solid and liquid LB culturing conditions, the toxicity mediated by Tle1^Vc was notably milder in liquid medium compared to solid medium. These discrepancies may be attributed to the presence of metal ions in agar, as previous studies have reported their involvement in mediating the killing activity of T6SS effectors.⁵⁷ PAO1 has evolved unknown pathways to counteract T6SS effectors (VgrG3 and VasX) of V. cholerae, in contrast to E. coli MG1655.²² Furthermore, PAO1 encodes over 1000 more proteins than MG1655, which could contribute to its enhanced protective capacity against phospholipase-mediated membrane damage inflicted by Tle1^Vc. Further investigations are required to fully understand the underlying mechanisms.

In the present study, we identified two putative immunity proteins, Tli1^tox- and Tli1^tox+, encoded by genes downstream of tle1^Vc and tle2^Vc, respectively, in environmental V. cholerae E1 and toxigenic N16961 strains. Bioinformatics analysis revealed that these cognate proteins belong to the DUF2931 family and are predicted to possess signal peptide sequences at their N-termini, consistent with the localization of Tle1^Vc. However, functional analysis demonstrated that Tli1^tox- was essential for neutralizing the toxicity of Tle1^Vc through direct interaction, whereas Tli1^tox+ did not exhibit this protective role, despite sharing approximately 80.24% sequence identity. It has been reported that mutations in immunity proteins can lead to a decrease in V. cholerae‘s intraspecies competition capacity.⁵⁸ Further investigation is needed to elucidate whether variations in amino acid residues in Tli1^tox+ resulted in the loss of interaction with Tle1^Vc.

In addition to conferring interbacterial competition, the T6SS may also play multiple physiological roles, including bacterial motility, protease activity, biofilm formation and can provoke tit-for-tat response from Pseudomonas aeruginosa.^22,59 In this study, we demonstrated that inactivation or deletion of Tle1^Vc impacts motility of V. cholerae E1 by downregulating the expression of flagellar relative genes such as flgB, flgO, flaABE and motXY, whose expression is essential for flagellar synthesis.^48,50,51 And the motility of Δtle1^Vc could be partially restored by complementing with flagellar gene flaA or flaE. These data were consistent with the result of electron microscopy imaging assays. Furthermore, motility induced by Tle1^Vc is independent of the functional T6SS but relies on Tle1^Vc-catalytic activity. As T6 effectors, VgrG3 and TseL, despite lacking a classical Sec or TAT secretion signal, have been observed in the periplasm through heterologous expression in E. coli, even in the absence of T6SS. This suggests that T6 effectors can reach the periplasm independently of T6SS,⁶⁰ potentially leading to Tle1^Vc-mediated motility independent of an active T6SS. The regulation of flagellum biosynthesis is a complex process. Cell envelope stress response systems and defense-related genes have been identified as factors influencing flagellum-driven motility.⁶¹ For instance, the Rcs phosphorelay system, which responds to outer membrane damage, as well as two-component regulatory systems BaeSR and WigKR, which participate in the cellular response to T6SS effectors and provide protection against foreign toxins,^62–64 suggest that cell envelope stress response systems may integrate effector damage as a signal to initiate a cascade response, potentially associated with Tle1^Vc-driven motility. Additionally, T6SS-dependent Tle1^Vc also triggers tit-for-tat (TFT) response and this process depends on T6SS-H1 counterattack of P. aeruginosa and its catalytic activity, as observed in Tle2^Vc22. Interestingly, we found that physical puncture of V. cholerae T6SS can moderately induce T6SS-H1-mediated counterattack of P. aeruginosa, showing that single physical penetration is sufficient for inducing TFT response. Furthermore, the expression of Tle1^Vc in P. aeruginosa PAO1 could not stimulate TFT response and Tle1^Vc-mediated strong TFT response was dependent on the physical puncture of V. cholerae E1, which differs from Tle2^Vc-mediated TFT response that the expression of Tle2^Vc alone in PAO1 was sufficient for inducing T6SS-H1-mediated counterattack.²² As Tle1^Vc and Tle2^Vc have various phospholipase activities responsible for different bactericidal capacity toward P. aeruginosa and E. coli,^19,22,32 which may result in distinct TFT responses stimulated by them. Taken together, these findings indicate that the effects of T6SS and Tle1^Vc proteins on microorganism physiology are multidimensional and complicated.

In addition to being an anti-bacterial weapon, T6SS allows pathogens to compete with commensal microbiota during host colonization.^52,53,65 To better understand the role of E1 T6SS, we carried out mouse colonization assay and verified that T6SS could promote E1’s in vivo fitness. T6SS-delivered Tle1^Vc is shown to present bactericidal activity in vivo and confer V. cholerae E1 a competitive advantage in the cecum of adult mice, which was also further confirmed by competitive colonization experiments in which the E1 wild type strain showed a significant competitive advantage (>10⁴ fold, Figure 8e,f) over Δtle1^Vc/tli1^tox- prey strain in the cecum, similar to in vitro killing efficiency (Figure 3a). E1 T6SS had stronger dynamic in vivo than toxigenic strain which has a much lower in vivo killing rate (<10 fold).^57,65 These results imply that the atypical pathogenic V. cholerae E1 has more higher in vivo killing capacity than toxigenic V. cholerae strain. In conclusion, the data presented here provide a strong evidence that Tle1^Vc of V. cholerae E1 serves as a T6SS effector belonging to Tle1 lipase family and its homologs are widely distributed among different bacterial species. Our finding suggested that Tle1^Vc not only acts as a antimicrobial toxin that plays important roles in antagonistic bacterium-bacterium interactions, also is involved in a number of cellular processes.

Acknowledgments

We thank Prof. Zhaoqing Luo for helpful discussion and Prof. Liang Yang for providing us with the P. aeruginosa PAO1 strain.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19490976.2023.2241204

Additional information

Funding

This work was supported by the National Natural Science Foundation of China under Grants [32270061], [32100019] and [32070131], Shenzhen Science and Technology Program under Grant [KQTD20200909113758004], and the China Postdoctoral Science Foundation under Grant [2020M682773].