Heterologous Assembly of the Type VI Secretion System Empowers Laboratory Escherichia coli with Antimicrobial and Cell Penetration Capabilities

ABSTRACT

The synthetic biology toolbox has amassed a vast number of diverse functional modules, but protein translocation modules for cell penetration and cytosol-to-cytosol delivery remain relatively scarce. The type VI secretion system (T6SS), commonly found in many Gram-negative pathogens, functions as a contractile device to translocate protein toxins to prokaryotic and eukaryotic cells. Here, we have assembled the T6SS of Aeromonas dhakensis, an opportunistic waterborne pathogen, in the common laboratory strain Escherichia coli BL21(DE3). We constructed a series of plasmids (pT6S) carrying the T6SS structural and effector genes under native or tetracycline-inducible promoters, the latter for controlled expression. Using fluorescence microscopy and biochemical analyses, we demonstrate a functional T6SS in E. coli capable of secreting proteins directly into the cytosol of neighboring bacteria and outcompeting a number of drug-resistant pathogens. The heterologous assembly of T6SS not only confers the lab workhorse E. coli with the cytosol-to-cytosol protein delivery capability but also demonstrates the potential for harnessing the T6SS of various pathogens for general protein delivery and antibacterial applications.

IMPORTANCE The T6SS is a powerful and versatile protein delivery system. However, the complexity of its macromolecular structure and gene regulation makes it not a trivial task to reconstitute the T6SSs of pathogens in a nonpathogenic host. In this study, we have assembled an inducible T6SS in E. coli BL21(DE3) and demonstrated its functions in protein delivery and antimicrobial activities. The engineered T6SS empowers E. coli to deliver protein cargos into a wide range of prokaryotic and eukaryotic cells.

INTRODUCTION

The large number of specialized tools that microbial pathogens employ to interact with the host and the environment may be considered a valuable reservoir for developing unique synthetic biology modules. Indeed, both bacterial toxins and protein secretion systems have been engineered for diverse applications (1–6). However, tools for direct penetration of target cells and general protein delivery to both eukaryotic and prokaryotic cells remain scarce. The type VI secretion system (T6SS) is a double-tubular contractile molecular device that can deliver effectors to eukaryotic and prokaryotic cells in a contact-dependent manner, thereby playing a crucial role in interspecies interaction and survival (7–10). About 25% of Gram-negative bacteria encode the T6SS, including many important pathogens, such as Vibrio cholerae, enteroaggregative Escherichia coli, Pseudomonas aeruginosa, Burkholderia mallei, and Agrobacterium tumefaciens (8, 9, 11–13). These T6SSs can be classified into different subtypes, but they share a conserved core set of structural genes encoding a baseplate, an Hcp inner tube, a VipA/B contractile outer sheath, a VgrG trimeric spike, and a cone-shaped PAAR tip (14–17). Sheath contraction can eject the Hcp, VgrG, and PAAR proteins and their associated effectors outside the cell (7, 18). After contraction, an ATPase ClpV is recruited to disassemble the contracted sheath for recycling the sheath subunits while the secreted Hcp tube and effectors require replenishment (19, 20).

In contrast to the conserved structural proteins, T6SS effector proteins are highly divergent in structures and functions. Their secretion is mainly mediated via two routes, binding to the Hcp tube or to the VgrG-PAAR complex, the latter often requiring the presence of chaperone proteins (10, 15, 21–25). Some “evolved” Hcp, VgrG, and PAAR proteins with extended functional domains also act as effectors (15, 23, 26, 27). Known effectors have diverse functions, including actin cross-linking in eukaryotic cells (26, 28), lysing bacterial cell wall (29–32), damaging membranes (31, 33, 34), degrading nucleic acids (22, 35, 36), disrupting nutrient uptake, and inducing autophagy (37). These diverse effector functions enable the T6SS to be effective against not only eukaryotic species but also both Gram-negative and Gram-positive bacteria (8, 29, 37–39). In addition to those native toxins, the T6SS can also be engineered to secrete heterologous proteins, including β-lactamase (28), nuclease (40, 41), and Cre recombinase (3).

Aeromonas dhakensis is an opportunistic waterborne pathogen that can cause sepsis and gastroenteritis (42). The A. dhakensis model strain SSU is a diarrheal isolate with a constitutively active T6SS exhibiting strong antibacterial and antieukaryotic activities (24, 43, 44). The genome encodes three known T6SS effector proteins: a pore-forming toxin TseC (24, 41), a self-cleavable Rhs-family nuclease TseI (36), and a lysozyme, TseP (43). Each effector has its specific immunity protein for self-protection. These effectors enable SSU to efficiently outcompete E. coli, and T6SS-active pathogens V. cholerae and P. aeruginosa (3, 24, 45). Besides the T6SS, A. dhakensis also exhibits a number of additional virulence traits, including the secretion of aerolysin and the type III protein secretion system (46, 47). These additional toxins in A. dhakensis and similar toxins in other T6SS organisms hinder the application of directly using the T6SS for protein delivery in their native hosts.

Laboratory E. coli strains have been arguably the most widely used protein expression platforms not only for structural and biochemical research but also for producing valuable metabolites and engineering synthetic biological circuits (48–51). Here, we have expanded their capabilities by developing constitutive and inducible T6SSs that enable common E. coli to deliver proteins deep into the cytosol of a bacterial cell as well as to kill a number of clinically relevant bacterial pathogens. Using Red/ET recombineering, a useful tool for cloning large biosynthetic gene clusters (52–54), we constructed several plasmid vectors carrying the 38-kb T6SS gene cluster of A. dhakensis SSU under the control of a constitutively native promoter or a tetracycline-inducible promoter in E. coli strains BL21(DE3) and DH5α. We demonstrate that the resultant E. coli strains carrying pT6S plasmids (T6SS-encoding plasmids) can actively assemble tubular contractile structures, deliver toxic nuclease into a neighboring bacterial cell, and kill a number of important pathogens that infect plants, insects, and humans. The plasmid-borne pT6Ss may be used to transform a variety of existing E. coli chassis strains from a conventional protein production tool to a powerful protein secretion and antibacterial tool for broad applications.

RESULTS

Cloning of the T6SS cluster by Red/ET recombineering.

Almost all T6SS structural proteins of A. dhakensis SSU are encoded on a large gene cluster from the hcp1 gene (HMPREF1171_00923) to the tsiP gene downstream of HMPREF1171_00947, except for PAAR2, Hcp2, and VgrG3, which are encoded on an accessory cluster (Fig. 1A) (43, 44). To visualize T6SS assembly, we constructed a chromosomal VipA-sfGFP (superfolder green fluorescent protein) fusion within the large cluster in SSU. Two NheI restriction enzyme sites were first introduced flanking the T6SS cluster in the genome for a large gene cluster clone. The upstream NheI site was inserted either directly in front of the ribosome binding site (RBS) of the first gene, hcp1, to be added to a tetracycline-inducible promoter, P_tetO, or 662 bp upstream of the start codon of Hcp1 containing the native promoter. Using the p15A-cm-tetR-tetO-hyg-ccdB plasmid as the template (55), we amplified the linear p15A vectors carrying two 80-bp flanking arms homologous to the T6SS cluster. Genomic DNA, digested with NheI, and linear vector DNA were mixed and electroporated into the E. coli GB05RedTrfA strain carrying the RecET recombination system (55), in which the digested T6SS cluster was ligated with the linearized vector through homologous recombination (Fig. 1B). To confirm this, we isolated plasmids from independent colonies after the transformation and tested the restriction enzyme-digested fragments by electrophoresis. Using in silico digestion analysis, we first identified eight KpnI sites in the pT6S plasmids and their corresponding theoretical sizes of digested products. Electrophoresis results show that seven of the eight expected bands were found with the correct size, while the smallest product of 182 bp was not detected, which may be attributed to insufficient staining under the test conditions (Fig. 1C). Using Sanger sequencing analysis, we confirmed the ligation position of the 80-bp homologous arms and the existence of the 182 bp that was not detected by gel electrophoresis. To construct T6SS-inactive controls, we employed two separate strategies by deleting the vasK gene and introducing a noncontractile 3-amino-acid (Ala-Glu-Val) insertion in the VipA N-terminal region (VipA-N3), respectively (24, 56, 57). Altogether, we obtained two sets of T6SS plasmids: pT6S_NP for constitutive expression with the native promoter and pT6S_Tet for controlled expression under the tetracycline-inducible promoter (Fig. 1C and see Fig. S1A and B in the supplemental material).

FIG 1.

Construction of pT6S plasmids by Red/ET recombineering. (A) Organization of the T6SS cluster (from HMPREF1171_00923 to the tsiP gene downstream of HMPREF1171_00947) in A. dhakensis SSU. There are two effector-immunity protein pairs, TseI-TsiI and TseP-TsiP. (B) Schematic for RecET direct cloning of the T6SS gene cluster. The left panel shows linearization of the vector, while the right panel shows digestion of genomic DNA. (C) Restriction digestion spectra of four pT6S plasmids. Plasmids were digested with KpnI and resolved on the agarose gel by electrophoresis. The molecular weights of digested DNA bands are indicated. The molecular weights of the DNA standards (M) are shown in Fig. S1A.

BL21(DE3)/pT6S can translocate proteins into the cytosol of recipient cells.

To test whether these pT6S plasmids are functional, we transformed each of them into E. coli BL21(DE3), a common lab workhorse for protein production. For simplicity, we refer to these E. coli strains by the transformed plasmids here. Using fluorescence microscopy analysis, we found that pT6S_NP and pT6S_Tet strains both assembled dynamically contractile tubular structures, a hallmark for T6SS activities (Fig. 2A and Movies S1 and S2) (18). As expected, noncontractile tubular structures and bright loci were formed in the T6SS-inactive VipA-N3 and the ΔvasK mutants, respectively (Fig. 2A and Fig. S2). Competition analysis between an E. coli prey and the pT6S strains confirms that the T6SS is functional for both pT6S_NP and the pT6S_Tet (Fig. 2B). Notably, the killing ability of pT6S_Tet was dependent on the inducer concentration and was higher than that of pT6S_NP when anhydrotetracycline (aTc) was added at 100 ng/mL (Fig. 2B and Fig. S3).

FIG 2.

The pT6S constructs in E. coli BL21(DE3) are functional. (A) Fluorescence microscopy analysis of E. coli BL21(DE3) carrying different pT6S plasmids. The sheath subunit VipA or VipA-N3 was labeled with C-terminal sfGFP. See also Movies S1 and S2 in the supplemental material for time-lapse images. Image size, 5 by 5 μm. Scale bars, 1 μm. (B) Competition analysis of E. coli BL21(DE3) containing pT6S plasmids with the prey strain E. coli MG1655. The SSU wild-type (WT) strain and the ΔvasK mutant were included as controls. An empty vector, pBAD18, was used to provide kanamycin resistance to MG1655. (C) Competition assay of pT6S constructs against the T6SS immunity-defective ΔtseI^c tsiI ΔvasK mutant as prey. The prey was transformed with an empty vector, pPSV37 (p), or pPSV37 carrying the immunity gene tsiI (ptsiI) induced with 0.1 mM IPTG. (D) Competition analysis of effector-inactivated pT6S with the E. coli MG1655 prey. Catalytic residue mutations of the two effectors TseI (HFH-AAA) and TseP (E663A) were introduced to the pT6S constructs, resulting in corresponding pT6S_{_eff} plasmids. The SSU WT and the ΔvasK mutant were included as controls. The prey MG1655 carries a pBAD18 vector with kanamycin resistance. (E) Western blot of Hcp secretion. The RNA polymerase subunit RpoB serves as a control for equal loading and cell lysis. For panels B, C, and D, error bars indicate mean ± SD of values from three biological replicates. Two-tailed Student's t test was used to determine P values. ***, P < 0.001; ns, not significant. When there was no prey survival at the most concentrated dilution, the arbitrary number 1 was used for calculation to indicate the detection limit (DL).

To test whether the pT6S strains can deliver proteins into the cytosol of recipient bacteria, we employed a competition assay based on the delivery of the nuclease effector TseI (36). Using a TseI immunity-defective mutant of A. dhakensis SSU ΔtseI^c tsiI ΔvasK as prey, we found that both pT6S_NP and pT6S_Tet outcompeted the prey carrying an empty vector but not the ones ectopically expressing the immunity protein TsiI (Fig. 2C). As a control, the noncontractile mutants did not affect prey survival (Fig. 2C). These results suggest that the T6SS could deliver the nuclease TseI to the cytoplasm of recipient cells to achieve cytosol-to-cytosol delivery.

To confirm that the bacterial killing of pT6S is not due to the physical force of T6SS puncture or an unknown protein from the host E. coli BL21(DE3) strain, we constructed pT6S_{NP_eff} and pT6S_{Tet_eff}, in which the two pT6S-encoded known effectors TseI and TseP were catalytically inactivated (Fig. S1C) (36, 43). Competition and secretion analyses show that both pT6S_{NP_eff} and pT6S_{Tet_eff} strains lost the antibacterial activities but were able to secrete Hcp, indicative of an active T6SS (Fig. 2D and E). Secretion of Hcp was detected in the noncontractile pT6S_{N3-Tet_eff} sample, but at a much lower level than that of its corresponding T6SS-active sample, pT6S_{Tet_eff}, which we attributed to cell lysis (Fig. 2E and Fig. S4). These results confirm our previous findings that effector activities are not required for the T6SS secretion (41, 58). Collectively, these results show that the plasmid-borne engineered T6SS is functional in BL21(DE3). We also transformed these plasmids into another commonly used laboratory strain, E. coli DH5α, and the T6SS tubular structures were similarly detected in pT6S_Tet and pT6S_N3-Tet strains (Fig. S5).

pT6S plasmids can be stably maintained.

The plasmid-borne pT6S offers flexibility for convenient use in different strain backgrounds but may suffer from instability due to its large size. Therefore, we examined whether the pT6S plasmids can be stably maintained in E. coli BL21(DE3). BL21(DE3) strains carrying either pT6S or the p15A vector grew similarly in LB medium, suggesting T6SS expression has a rather minor effect on fitness (Fig. S6). Using flow cytometry analysis of the fluorescently labeled T6SS VipA-sfGFP sheath, we monitored GFP signals of different pT6S strains grown in LB over a time course of 96 h. The percentage of cells with GFP signals (GFP⁺) of the pT6S_NP strain was reduced gradually over time, but there were still 39.3% GFP⁺ cells after 96 h of incubation in the absence of antibiotic selection (Fig. 3A). When the appropriate antibiotic was added, over 60% of cells were GFP⁺ in the 96-h samples. The other pT6S plasmids exhibited similar stability (Fig. S7). To explore whether the pT6S plasmids are still functional, we also tested the ability of the E. coli/pT6S strains to outcompete a prey strain after 96 h of growth in the absence of antibiotic selection. The results show that the prey E. coli strain, MG1655, was efficiently killed by both pT6S_NP and pT6S_Tet killer strains at similar levels to those by killer strains prior to the continuous culture treatment (Fig. 2B and Fig. 3B), suggesting the pT6S plasmids are still functional.

FIG 3.

Stability of pT6S plasmids in E. coli BL21(DE3). (A) Dot plots of flow cytometry analysis for the portion of GFP⁺ cells in E. coli BL21(DE3) carrying pT6S_NP over time with antibiotics (Cm+) or without antibiotics (Cm−). The percentage of GFP⁺ cells is indicated at the bottom. (B) Competition analysis of E. coli BL21(DE3) carrying pT6S plasmids, isolated after 96 h of continuous culture in the absence of antibiotic selection, against the prey MG1655. The SSU wild-type (WT) strain and the ΔvasK mutant were included as controls, and MG1655 carries a pBAD18 vector with kanamycin resistance. Survival of prey was compared by serial dilutions (10×) on selective medium. The experiment was repeated twice. (C) Schematic models of TseC complementation plasmids. Shown are pTseCTsiC (effector TseC and its immunity protein TsiC), p2404-2401(VgrG3, chaperone, effector TseC, and its immunity protein TsiC), or p2404-2401(TseC^Δ14) (14 amino acids [M814 to F827] in the middle of the C-terminal colicin domain deletion). (D) Competition assay of effector-inactivated pT6S_{_eff} with TseC complementation plasmids against E. coli MG1655 prey. pT6S_{_eff} E. coli strains were transformed with pTseCTsiC, p2404-2401, or p2404-2401(TseC^Δ14). Error bars indicate the mean ± SD of values from three biological replicates. Two-tailed Student's t test was used to determine P values. ***, P < 0.001.

Next, we tested whether the BL21(DE3)/pT6S can be equipped with additional plasmid-borne effectors: we constructed three plasmids expressing an additional SSU T6SS effector TseC alone, with its cognate VgrG and chaperone proteins, or an inactive mutant TseC^Δ14, respectively (24, 43). Each plasmid also encodes both TseC and the cognate immunity TsiC protein for self-protection (Fig. 3C). To focus on TseC functions, we transformed the TseC-expressing plasmids to the effector-inactivated pT6S_eff strains. Competition analysis against the E. coli MG1655 prey reveals that TseC could be effectively delivered to kill the E. coli prey only when VgrG and the chaperone were coexpressed in the T6SS-dependent manner (Fig. 3D). Collectively, these results indicate that the pT6S plasmids are stable in E. coli and also amenable for extended functions with additional effector-encoding plasmids.

pT6S confers E. coli broad antimicrobial activities.

Considering the serious threat of antimicrobial resistance to public health, we next tested whether the E. coli/pT6S strains could be used to inhibit bacterial pathogens. We arbitrarily selected a panel of diverse microbial pathogens of plants, insects, humans, and animals. V. cholerae C6706 is a 7th cholera pandemic isolate, and Vibrio parahaemolyticus is a leading cause of foodborne gastroenteritis (59, 60). Diffusely adherent E. coli (DAEC) and enterotoxigenic E. coli (ETEC) can cause acute diarrhea in infants and adults (61), while Citrobacter rodentium is diarrheagenic to mice and an important model to study the pathogenesis of enteropathogenic E. coli (EPEC) and enterohemorrhagic E. coli (EHEC) (62). Klebsiella pneumoniae can cause severe infections in the urinary tract and the respiratory tract (63). Photorhabdus asymbiotica is an insect pathogen and opportunistic pathogen of human (64). Pseudomonas syringae pv. syringae is an important plant pathogen and also one of the best-studied plant pathogens (65). When these pathogens were competed with the E. coli/pT6S strains, all exhibited various degrees of sensitivity to the engineered pT6S strain, while K. pneumoniae was the least sensitive (Fig. 4). The induced pT6S_Tet strain exhibited greater killing efficiency than the pT6S_NP strain with the native promoter (Fig. 4).

FIG 4.

Broad antimicrobial activities of E. coli BL21(DE3)/pT6S strains. Survival levels of different preys between pT6S_NP (+) and pT6S_N3-NP (−) strains (A) and between pT6S_Tet (+) and pT6S_N3-Tet (−) strains (B) were calculated. Error bars indicate the mean ± SD of values from three biological replicates. Two-tailed Student's t test was used to determine P values. **, P < 0.01; ***, P < 0.001; ns, not significant. When there was no prey survival at the most concentrated dilution, the arbitrary number 1 was used for calculation to indicate the detection limit (DL).

We also test whether the pT6S strains can kill P. aeruginosa, another important human pathogen that can detect T6SS attacks and elicit strong tit-for-tat responses using its own H1-T6SS (66, 67). Results show that the pT6S strains and their T6SS-inactive mutants did not exhibit any difference in survival when competing with wild-type P. aeruginosa, suggesting that the physical puncture and the delivered effectors of the pT6S are insufficient to trigger the retaliatory response of P. aeruginosa (Fig. S8).

DISCUSSION

Although secretion systems are often employed by pathogens to interact with the host, their functions can be harnessed and engineered to deliver specific cargos for various applications, including competing with other pathogens and immune modulation (4, 40, 41, 68–73). In comparison with other secretion systems, the T6SS is capable of injecting proteins directly into diverse species, ranging from Gram-negative and Gram-positive bacteria to fungi and eukaryotic cells, in a contact-dependent manner and without the need for surface recognition (8, 28, 29, 37–39). These features make it a very powerful tool to modulate both bacteria and the host, depending on the cargo proteins. In this study, we have engineered a set of plasmid-borne T6SSs that can confer on the commonly used laboratory E. coli strain additional capabilities for cell penetration and cytosol-to-cytosol protein delivery into recipient cells. We demonstrate that the T6SS-armed E. coli can deliver antibacterial effectors and eliminate a number of bacterial pathogens. Considering the vast number of T6SSs and their effectors in Gram-negative bacteria, our study demonstrates the methodological feasibility and the potential for T6SS-mediated applications that are not only limited to controlling infectious diseases but could also include delivering a broad range of proteins and peptides for therapeutic purposes (Fig. 5).

FIG 5.

Schematic model of E. coli/pT6S applications. There are multiple clades of T6SSs in Gram-negative bacteria, with the majority of their functions uncharacterized. Of the four T6SS subtypes recently described, T6SSⁱ is widely found in Proteobacteria, T6SSⁱⁱ in Francisella species, T6SSⁱⁱⁱ in Bacteroidetes, and the T6SS^iv in Amoebophilus asiaticus (16, 17). Our approach demonstrates the feasibility of expressing these T6SSs in chassis strains, as exemplified in BL21(DE3) here. These T6SS-based protein delivery tools may have broad applications in various sectors, including biotechnology, health, and environment. The image was generated using BioRender.

In addition to the application potential, heterologous assembly of the A. dhakensis T6SS in E. coli has two important implications for understanding the T6SS biogenesis. First, it suggests that genes for the T6SS assembly are orthogonal and modular in that no other A. dhakensis genes but the pT6S ones are required for T6SS assembly in E. coli. This is supported by a recent report that the T6SS of V. parahaemolyticus can be functionally assembled in Vibrio natriegens, another member of the Vibrio genus but lacking a native T6SS (74). These findings pave the way for studying the divergent functions and biogenesis of the T6SSs from unculturable, genetically intractable, or high-risk pathogenic bacteria. Second, cell-envelope-spanning macromolecular complexes often require specific mechanisms to traverse the peptidoglycan (PG) layer while preserving the essential barrier function of PG (75). Our data suggest that the A. dhakensis T6SS does not require a cognate PG-degrading enzyme for the transenvelope complex TssJ/L/M to cross the PG layer. This is in contrast to the assembly of Acinetobacter T6SS, for which a T6SS-associated inner membrane-bound PG hydrolase, TagX, has been proposed to remodel the cell wall and permit assembly (76). Notably, a large number of T6SS genes do not encode such T6SS-associated enzymes. The T6SS in enteroaggregative E. coli employs a housekeeping PG-lytic enzyme, MltE, that directly interacts with the periplasmic domain of TssM (77). It is possible that the pT6S strain may hijack the E. coli MltE protein for the same purpose, but the sequence identity of the two TssM proteins is only 21.1% (see the supplemental material). Therefore, whether MltE is a promiscuous TssM-binding partner or other PG-modifying enzymes are also involved requires further investigation. Nonetheless, the pT6S plasmids provide a valuable toolkit to study the PG-crossing mechanism of the T6SS macromolecular complex by taking advantage of the powerful E. coli genetics.

In summary, the functional assembly of T6SS in laboratory E. coli overcomes the potential safety concern of using the native attenuated or avirulent pathogens. The constructed pT6S can be constitutively active or inducible, equipped with additional plasmid-borne effectors, and also flexibly transferred to other chassis strains. Depending on cargo activities and recipient cell types, the pT6S functions can be expanded way beyond the demonstrated antibacterial activities. These functions may have a transient and noninheritable effect or some lasting or permanent effect—e.g., through Cre-mediated genome editing and epigenetic modifying factors (3). Therefore, the adaptability of the engineered pT6S greatly expands the application of commonly used laboratory E. coli strains for diverse applications.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

All the strains and plasmids used in this study are listed in Table 1 and Table 2. Growth of the strains was routinely in lysogeny broth (LB) (1% [wt/vol] tryptone, 0.5% [wt/vol] yeast extract, 0.5% [wt/vol] NaCl). Antibiotics and chemicals were added when needed: 100 μg/mL streptomycin, 25 μg/mL chloramphenicol (15 μg/mL used in Red/ET recombineering), 4 μg/mL tetracycline, 50 μg/mL kanamycin, 50 μg/mL carbenicillin, 20 μg/mL gentamicin, 25 μg/mL irgasan, 0.01% (wt/vol) l-arabinose for the induction of pBAD24 vectors, 0.2% (wt/vol) l-arabinose for the RecET direct cloning, 0.2% (wt/vol) l-rhamnose, 0.1 mM IPTG (isopropyl-β-d-thiogalactopyranoside), and 100 ng/mL anhydrotetracycline (aTc).

TABLE 1.

Strains used in this study

Strain	Name or type used in this study	Genotype	Description	Reference or source
E. coli	GB05RedTrfA	DH10B fhuA::IS2 ΔybcC ΔrecET PRhaSR-γβαA pBAD-trfA	Strain used for Red/ET recombineering	55
	GBdir-gyrA462	DH10B fhuA::IS2, ΔybcC ΔrecET PBAD-ETγA gyrAArg462Cys	Arg462Cys mutation in GyrA subunit of DNA gyrase conferring CcdB resistance	55
	BL21(DE3)	F– ompT gal dcm lon hsdSB(rB–mB–) λ(DE3 [lacI lacUV5-T7 p07 ind1 sam7 nin5]) [malB+]K-12(λS)	Strain used as heterologous host of T6SS	Lab stock
	MG1655	K-12 F– λ– ilvG mutant rfb-50 rph-1	Strain used for competition assay	Lab stock
	WM6026	lacIq rrnB3 ΔlacZ4787 hsdR514 ΔaraBAD567 ΔrhaBAD568 rph-1 attl::pAE12(ΔoriR6K-cat::Frt5) ΔendA::Frt uidA(ΔmluI)::pir attHK::pJK1006Δ(oriR6K-cat::Frt5 trfA::Frt)	Strain used for conjugation, diaminopimelic acid auxotroph	Mekalanos lab
	PIR1	F− Δlac169 rpoS(Am) robA1 creC510 hsdR514 endA recA1 uidA(ΔmluI)::pir-116	Strain used for cloning	Invitrogen
	DH5α	F− ϕ80lacZΔM15 Δ(lacZYA-argF)U169 recA1 endA1 hsdR17 phoA supE44 thi-1 gyrA96 relA1 λ−	Strain used for cloning and gene expression	Invitrogen
	DAEC 2787		Strain used for competition assay	Lab stock
	ETEC H-10407		Strain used for competition assay	Lab stock
A. dhakensis SSU	WT	Parental strain	Parental strain	24
	ΔvasK	ΔvasK	T6SS null, in-frame deletion of vasK	24
	ΔtseIc tsiI	ΔtseIc tsiI	In-frame deletion of toxin-coding sequence of tseI and tsiI	36
	ΔtseIc tsiI ΔvasK	ΔtseIc tsiI ΔvasK	In-frame deletion of vasK in ΔtseIc tsiI background	This study
	VipA-sfGFP	vipA-sfgfp	Chromosomal VipA-sfGFP fusion	This study
	VipA-N3-sfGFP	vipA-N3-sfgfp	Chromosomal VipA-N3-sfGFP fusion, noncontractile sheath	This study
	VipA-sfGFP-NP	vipA-sfgfp	NheI sites inserted flanking T6SS cluster, upstream NheI site inserted in front of native promoter of hcp1	This study
	VipA-sfGFP-Tet	vipA-sfgfp	NheI sites inserted flanking T6SS cluster, upstream NheI site inserted in front of RBS of hcp1	This study
	VipA-N3-sfGFP-NP	vipA-N3-sfgfp	Chromosomal VipA-N3-sfGFP fusion, noncontractile sheath, NheI sites inserted flanking T6SS cluster, upstream NheI site inserted in front of native promoter of hcp1	This study
	VipA-N3-sfGFP-Tet	vipA-N3-sfgfp	Chromosomal VipA-N3-sfGFP fusion, noncontractile sheath, NheI sites inserted flanking T6SS cluster, upstream NheI site inserted in front of RBS of hcp1	This study
	VipA-sfGFP-NP_eff	vipA-sfgfp tseIHFH-AAA tsePE663A	Chromosomal mutations of TseI catalytic residues H1507, F1508, and H1509 and TseP catalytic residue E663; NheI sites inserted flanking T6SS cluster; upstream NheI site inserted in front of native promoter of hcp1	This study
	VipA-N3-sfGFP-NP_eff	vipA-N3-sfgfp tseIHFH-AAA tsePE663A	Chromosomal VipA-N3-sfGFP fusion, noncontractile sheath; chromosomal mutations of TseI catalytic residues H1507, F1508, and H1509 and TseP catalytic residue E663; NheI sites inserted flanking T6SS cluster; upstream NheI site inserted in front of native promoter of hcp1	This study
	VipA-sfGFP-Tet_eff	vipA-sfgfp tseIHFH-AAA tsePE663A	Chromosomal mutations of TseI catalytic residues H1507, F1508, and H1509 and TseP catalytic residue E663; NheI sites inserted flanking T6SS cluster; upstream NheI site inserted in front of RBS of hcp1	This study
	VipA-N3-sfGFP-Tet_eff	vipA-N3-sfgfp tseIHFH-AAA tsePE663A	Chromosomal VipA-N3-sfGFP fusion, noncontractile sheath; chromosomal mutations of TseI catalytic residues H1507, F1508, and H1509 and TseP catalytic residue E663; NheI sites inserted flanking T6SS cluster; upstream NheI site inserted in front of RBS of hcp1	This study
V. cholerae	C6706		Strain used for competition assay	Lab stock
V. parahaemolyticus	B5		Strain used for competition assay	Lab stock
C. rodentium			Strain used for competition assay	Lab stock
K. pneumoniae			Strain used for competition assay	Lab stock
P. aeruginosa	PAO1		Strain used for competition assay	Lab stock
P. asymbiotica			Strain used for competition assay	Lab stock
P. syringae pv. syringae			Strain used for competition assay	Lab stock

TABLE 2.

Plasmids used in this study

Plasmid	Description	Reference
pSC101-BAD-ETgA-tet	Plasmid used for Red/ET recombineering, tetracycline resistance	55
p15A-cm-tetR-tetO-hyg-ccdB	Plasmid used to amplify linear vector in Red/ET recombineering, chloramphenicol resistance	55
P15A-cm-tetR-tetO	Plasmid used as p15A empty vector without hyg and ccdB	This study
pBAD24Amp	Plasmid used to amplify bla gene in Red recombineering	Lab stock
pDS132kan	Suicide plasmid to construct in-frame deletions, kanamycin resistance	Lab stock
pDS132-ΔvasK	Suicide vector to construct in-frame deletion of vasK	24
pDS132-tseIHFH-AAA	Suicide vector to construct chromosomal tseIHFH-AAA	36
pDS132-tsePE663A	Suicide vector to construct chromosomal tsePE663A	43
pDS132-VipA-KI-sfGFP	Suicide vector to construct insertion of sfGFP	This study
pDS132-VipA-KI-N3-sfGFP	Suicide vector to construct insertion of N3 (noncontractile sheath mutation) and sfGFP	This study
pDS132-KI-NheI-F-NP	Suicide vector to construct insertion of NheI site in front of native promoter of hcp1	This study
pDS132-KI-NheI-F-Tet	Suicide vector to construct insertion of NheI site in front of RBS of hcp1	This study
pDS132-KI-NheI-B	Suicide vector to construct insertion of NheI site behind T6SS cluster	This study
pPSV37	IPTG-inducible expression vector, gentamicin resistance	81
pPSV37-tsiI-3V5	IPTG-inducible expression of TsiI with C-terminal 3V5 tag	This study
pBAD18	Arabinose-inducible expression vector, kanamycin resistance	82
pBAD24-tseCtsiC	Arabinose-inducible expression of TseC and TsiC	This study
pBAD24-2404-2401	Arabinose-inducible expression of genes from 2404 to 2401	This study
pBAD24-2404-2401(TseCΔ14)	Arabinose-inducible expression of genes from 2404 to 2401 with TseCΔ14 mutant	This study
pT6SNP	Plasmid for expression of T6SS cluster under native promoter of hcp1, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study
pT6SN3-NP	Plasmid for expression of T6SS cluster (noncontractile sheath mutation) under native promoter of hcp1, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study
pΔvasKNP	Deletion of vasK (ΔvasK::ampr) in pT6SNP background	This study
pT6STet	Plasmid for expression of T6SS cluster under tetracycline-inducible promoter PtetO, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study
pT6SN3-Tet	Plasmid for expression of T6SS cluster (noncontractile sheath mutation) under tetracycline-inducible promoter PtetO, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study
pΔvasKTet	Deletion of vasK (ΔvasK::ampr) in pT6STet background	This study
pT6SNP_eff	Plasmid for expression of T6SS cluster (mutations of TseI catalytic residues H1507, F1508, and H1509 and TseP catalytic residue E663) under native promoter of hcp1, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study
pT6SN3-NP_eff	Plasmid for expression of T6SS cluster (mutations of TseI catalytic residues H1507, F1508, H1509, and TseP catalytic residue E663; noncontractile sheath mutation) under native promoter of hcp1, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study
pT6STet_eff	Plasmid for expression of T6SS cluster (mutations of TseI catalytic residues H1507, F1508, and H1509 and TseP catalytic residue E663) under tetracycline-inducible promoter PtetO, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study
pT6SN3-Tet_eff	Plasmid for expression of T6SS cluster (mutations of TseI catalytic residues H1507, F1508, and H1509 and TseP catalytic residue E663; noncontractile sheath mutation) under tetracycline-inducible promoter PtetO, p15A-cm-tetR-tetO-hyg-ccdB backbone	This study

Construction of SSU derivative strains.

The methods of crossover PCR and homologous recombination were used to construct mutants of SSU as previously described (78, 79). The plasmids and suicide vectors for chromosomal recombination were generated using standard molecular biology techniques. Briefly, fragments of homologous arms and the target fragments were cloned into the suicide plasmid pDS132. Plasmids were transferred to the recipient cells by conjugation using a donor cell of E. coli WM6026. The plasmid was integrated into the chromosome by homologous recombination, and the two-step allelic exchange was completed by sacB-mediated sucrose counterselection. All the plasmids were verified by sequencing, and the primers used in this study are listed in Table 3.

TABLE 3.

Primers used in this study

Primer	Sequence (5′→3′)	Description
pDS132-hifi-f	CGATCCTTTTTAACCCATCAC	Forward primer to amplify pDS132 vector
pDS132-hifi-r	CTTCTAGAGGTACCGCATGC	Reverse primer to amplify pDS132 vector
pDS132-f	TGTTGCATGGGCATAAAGTTGC	Forward confirmation primer of pDS132 vector
pDS132-r	ACGGCTGACATGGGAATTCC	Reverse confirmation primer of pDS132 vector
pDS132-VipA-sfGFP KI-1	AAAAGGATCGATCCTCTAGAGAGTGCGTGGAGGATATGGG	Forward primer to amplify upstream to construct insertion of sfGFP
pDS132-VipA-sfGFP KI-2	TCCTCCTCCTGCGGCCGCGGCCTCGGCAGGCTTAATCAAA	Reverse primer to amplify upstream to construct insertion of sfGFP
pDS132-VipA-sfGFP KI-3	TTTGATTAAGCCTGCCGAGGCCGCGGCCGCAGGAGGAGGA	Forward primer to amplify gene of sfGFP
pDS132-VipA-sfGFP KI-4	AGTCAATCCAACCCGGCGATTTATTTGTAGAGCTCATCCA	Reverse primer to amplify gene of sfGFP
pDS132-VipA-sfGFP KI-5	TGGATGAGCTCTACAAATAAATCGCCGGGTTGGATTGACT	Forward primer to amplify downstream to construct insertion of sfGFP
pDS132-VipA-sfGFP KI-6	TCGCATGCGGTACCTCTAGAACCAGCGATACTTGGCGAAA	Reverse primer to amplify downstream to construct insertion of sfGFP
pDS132-VipA-N3-sfGFP KI-1	GTGATGGGTTAAAAAGGATCGGCATGGTATCTGCTTGCTGC	Forward primer to amplify upstream to construct insertion of N3
pDS132-VipA-N3-sfGFP KI-2	CTCTATTTCCGCAACCTCAGCCTGCTGACCACCGGTTGCCGGTA	Reverse primer to amplify upstream to construct insertion of N3
pDS132-VipA-N3-sfGFP KI-3	GCTGAGGTTGCGGAAATAGAGCTTCCGTTGA	Forward primer to amplify downstream to construct insertion of N3
pDS132-VipA-N3-sfGFP KI-4	GCATGCGGTACCTCTAGAAGGATAGTGGTCGTGATCCGAGCT	Reverse primer to amplify downstream to construct insertion of N3
pDS132-VipA-N3-sfGFP KI-5	GGTATGATGCGAGAGCAGGG	Forward confirmation primer of N3 mutation
pDS132-VipA-N3-sfGFP KI-6	ACCAGCGATACTTGGCGAAA	Reverse confirmation primer of N3 mutation
pDS132-VipA-N3-sfGFP KI-confirm	ACGCATCGTCAGCTGATGAACTA	Confirmation primer for sequencing of N3 insertion
pDS132-B-NheI-KI-1	GTGATGGGTTAAAAAGGATCGGGGAAGTAAAAACGGCGATG	Forward primer to amplify upstream of back NheI site to construct insertion of NheI
pDS132-B-NheI-KI-2	CAAACTCGCTAGCTTAACCATATCTTTCTTTCATATACTTCTTTACC	Reverse primer to amplify upstream of back NheI site to construct insertion of NheI
pDS132-B-NheI-KI-3	TGGTTAAGCTAGCGAGTTTGGCAAGGTGTGATAGAC	Forward primer to amplify downstream of back NheI site to construct insertion of NheI
pDS132-B-NheI-KI-4	GCATGCGGTACCTCTAGAAGGCCCTTAGCTCAGTCGGATAG	Reverse primer to amplify downstream of back NheI site to construct insertion of NheI
pDS132-B-NheI-KI-5	ATTATGCATTCAATCCTGTGGC	Forward confirmation primer of insertion of back NheI site
pDS132-B-NheI-KI-6	GCGCAACCTGTTTTCCATCA	Reverse confirmation primer of insertion of back NheI site
pDS132-B-NheI-KI-confirm	GCAGAAGGGATTGATGAGACA	Confirmation primer for sequencing of insertion of back NheI site
PDS132-F-KI-NheI-NP-KI-1	GTGATGGGTTAAAAAGGATCGCAAAAAGAAGCGAGGGCC	Forward primer to amplify upstream of front NP-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-NP-KI-2	CCACTCGGCTAGCGACACCTCACCAAATTTTTACTTCTTG	Reverse primer to amplify upstream of front NP-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-NP-KI-3	AGGTGTCGCTAGCCGAGTGGCTGAAGGAGCAC	Forward primer to amplify downstream of front NP-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-NP-KI-4	GCATGCGGTACCTCTAGAAGCTTGCACCAGCATCTCGTCT	Reverse primer to amplify downstream of front NP-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-NP-KI-5	TGGATGGTCATTTAGCCCCC	Forward confirmation primer of insertion of front NP-NheI site
PDS132-F-KI-NheI-NP-KI-6	AGCAGGCTGACCAGATTGC	Reverse confirmation primer of insertion of front NP-NheI site
PDS132-F-KI-NheI-NP-KI-confirm	TAAATTACGCCCCGTTCCAGC	Confirmation primer for sequencing of insertion of front NP-NheI site
PDS132-F-KI-NheI-RBS-KI-1	GTGATGGGTTAAAAAGGATCGGCGTTAAATGTGCGAACGGT	Forward primer to amplify upstream of front RBS-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-RBS-KI-2	TGCTCCTGCTAGCGATTGGTTGAACGGTAATGACACT	Reverse primer to amplify upstream of front RBS-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-RBS-KI-3	ACCAATCGCTAGCAGGAGCAATTCCATGCCAAC	Forward primer to amplify downstream of front RBS-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-RBS-KI-4	GCATGCGGTACCTCTAGAAGGAAGGTGTTTTCGGGCAGAG	Reverse primer to amplify downstream of front RBS-NheI site to construct insertion of NheI
PDS132-F-KI-NheI-RBS-KI-5	CGACCCATGAAGTGGTCTCC	Forward confirmation primer of insertion of front RBS-NheI site
PDS132-F-KI-NheI-RBS-KI-6	GCGCGAAGTCACTCACCACC	Reverse confirmation primer of insertion of front RBS-NheI site
PDS132-F-KI-NheI-RBS-KI-confirm	TTTCCTATTGTCAACGCATAGCTCA	Confirmation primer for sequencing of insertion of front RBS-NheI site
pDS132-tsePE663A-hifi-f	GTGATGGGTTAAAAAGGATCGCGGTGAGGTGGTTGGACATTG	Forward primer to amplify upstream and downstream of tsepE663 to construct mutation of tsepE663A
pDS132-tsePE663A-hifi-r	GCATGCGGTACCTCTAGAAGCGCTCCTGAGATTCGGGTAATC	Reverse primer to amplify upstream and downstream of tsepE663 to construct mutation of tsepE663A
tsePE663A-f	GCCAAAGATATGTTGGGTCAACC	Forward confirmation primer of tsePE663A mutation
tsePE663A-r	CAGTATATTCTCCACCAGCGTTCTTTTT	Reverse confirmation primer of tsePE663A mutation
tsePE663A-confirm	CTTATCACTCTAGAAATGGTTTTTGCAGC	Confirmation primer for sequencing of tsePE663A mutation
pDS132-tseIHFH-AAA-hifi-f	GTGATGGGTTAAAAAGGATCGCGAGCGCTTCTTGCATGATCC	Forward primer to amplify upstream and downstream of tseIHFH to construct mutation of tseIHFH-AAA
pDS132-tseIHFH-AAA-hifi-r	GCATGCGGTACCTCTAGAAGACGATATGCCTCTTCAAAAAACATCCT	Reverse primer to amplify upstream and downstream of tseIHFH to construct mutation of tseIHFH-AAA
tseIHFH-AAA-f	CTATGGCTATGACAGAACGGGCA	Forward confirmation primer of tseIHFH-AAA mutation
tseIHFH-AAA-r	CACTACCTTGTAATAAGACCAACGACCA	Reverse confirmation primer of tseIHFH-AAA mutation
tseIHFH-AAA-confirm	GCGCAATATAAAACACATAAATATGTTTATG	Confirmation primer for sequencing of tseIHFH-AAA mutation
pDS132-vasK-ko1	GTGATGGGTTAAAAAGGATCGTCTAGACCATGGTACTCAGCGGTGG	Forward primer to amplify upstream and downstream of vasK to construct deletion of vasK
pDS132-vasK-ko4	GCATGCGGTACCTCTAGAAGTCTAGATCTCCTGCTGATGTTGCTGC	Reverse primer to amplify upstream and downstream of vasK to construct deletion of vasK
pDS132-vasK-ko5	GTTGGTGATGGGGCTGAAA	Forward confirmation primer of deletion of vask
pDS132-vasK-ko6	GCGAATGTCGTCGTAGCTGA	Reverse confirmation primer of deletion of vask
p15A-Tet-r	GTGAAGGCACCGGCGGTGATGTTGCCCTGAGTTTTACCTTCGATGCTGATATAACATGGAGTTGGCATGGAATTGCTCCTCTTTAGATCTTTTGAATTCTTTTCTCTATC	Reverse primer to amplify p15A-cm-tetR-tetO-hyg-ccdB linear vector with 80-bp homologous arm target to T6SS cluster (without native promoter of hcp1)
p15A-f	AGGATGGGGCCGGTGATGGAACATATAAGTTTAACAATGCAATATCGGTAAAGAAGTATATGAAAGAAAGATATGGTTAAAGATCCGAAAACCCCAAGT	Forward primer to amplify p15A-cm-tetR-tetO-hyg-ccdB linear vector with 80-bp homologous arm target to T6SS cluster
p15A-NP-r	ATTTGGCGGTGAGGGAGGGATTCGAACCCTCGATACGTTGCCGTATACACACTTTCCAGGCGTGCTCCTTCAGCCACTCGCTTTAGATCTTTTGAATTCTTTTCTCTATC	Reverse primer to amplify p15A-cm-tetR-tetO-hyg-ccdB linear vector with 80-bp homologous arm target to T6SS cluster (with native promoter of hcp1)
bla-f	TCGAAATGAGTGCCGAAATGAGTCATTGAGTCGGAGTTGGATCACATTTTATGAGTATTCAACATTTCCGTGTCG	Forward primer to amplify bla with 50-bp homologous arm target to vasK
bla-r	ACATATGAACTTCCAACACAGGAAGGGGGCACAACGTGCCCCCGACGCGATTACCAATGCTTAATCAGTGAGGC	Reverse primer to amplify bla with 50-bp homologous arm target to vasK
p15A-cm-tetR-tetO-f	AGATCTAAAGAGATCCGAAAACCCCAAGT	Forward primer to amplify p15A empty vector without hyg and ccdB
p15A-cm-tetR-tetO-r	TTTCGGATCTCTTTAGATCTTTTGAATTCTTTTCTCTATC	Reverse primer to amplify p15A empty vector without hyg and ccdB
ΔvasK-confirm-f	CAAGAGGCCAGTCGCCTC	Forward confirmation primer of deletion of vasK and the insertion of bla
ΔvasK-confirm-r	GTACTCCGGCAACATCCTGG	Reverse confirmation primer of deletion of vasK and insertion of bla
p15A-confirm-f	GTTAAACCTTCGATTCCGACCT	Forward confirmation primer of sequence of front 80-bp homologous arm in recombinant plasmid
p15A-confirm-r	TATAAACGCAGAAAGGCCCAC	Reverse confirmation primer of sequence of back 80-bp homologous arm in recombinant plasmid
tsiI-3V5-KpnI-f	CCGGTACCAGGAGGAAACGATGAATTCT	Forward primer to amplify tsiI with 3V5 tag
tsiI-3V5-XbaI-r	GCTCTAGATTATTATGTTGAATCAAGTCCGAGAAGTGG	Reverse primer to amplify tsiI with 3V5 tag
pPSV37-f	AAGGGTGGAAACGCAAAA	Forward confirmation primer of pPSV37
pPSV37-r	CCCCTCAAGACCCGTTTAGA	Reverse confirmation primer of pPSV37
tseCtsiC-f	TTGGGCTAGCAGGAGGTACCATGAGTACGCCCAATCAAGCC	Forward primer to amplify tseC tsiC
tseCtsiC-r	GGTTTACCGCATGCTCTAGATTACTTATTGATCTCATCCAGCCTTCTTG	Reverse primer to amplify tseC tsiC
2404-2401-f	TTGGGCTAGCAGGAGGTACCATGGCAGACAGCACAGGATTACA	Forward primer to amplify genes from 2404 to 2401
pBAD24-KpnI-hifi-R	GGTACCTCCTGCTAGCCCAAAA	Reverse primer to amplify pBAD24 vector
pBAD24-KpnI-hifi-F	TCTAGAGCATGCGGTAAACCTATTCC	Forward primer to amplify pBAD24 vector

Preparation of genomic DNA and linear vector.

The isolation of genomic DNA was prepared by phenol-chloroform extraction and ethanol precipitation as described by Wang et al. (55). Briefly, overnight cultures were diluted 1:100 in fresh LB with appropriate antibiotics and grown at 37°C with shaking at 200 rpm until they reached an optical density at 600 nm (OD₆₀₀) of 1. Fifty milliliters of bacterial cultures was harvested by centrifugation at 8,300 × g for 5 min at room temperature. The pellets were washed with double-distilled H₂O (ddH₂O) and centrifuged again, and the pellets were then resuspended in 8 mL Tris-HCl (10 mM [pH 8.0]) with 500 μL proteinase K (20 mg/mL) and 1 mL SDS (10% [wt/vol]) and incubated at 50°C for 2 h. Ten milliliters of phenol-chloroform-isoamyl alcohol mixture (25:24:1) was added to make the mixture an emulsion and centrifuged at 8,300 × g for 30 min at room temperature. The 500-μL aqueous phase was transferred into a new microcentrifuge tube with 35 μL sodium acetate (3 M [pH 7.5]) and 1.2 mL absolute ethanol. The genomic DNA was pelleted by centrifugation at 9,400 × g for 1 min at room temperature. The air-dried DNA was resuspended in 500 μL Tris-HCl (10 mM [pH 8.0]) and digested by NheI restriction enzyme. Linear cloning vectors with 80-bp flanking arms homologous to the target cluster were amplified by the primer pairs p15A-NP-r and p15A-f and p15A-Tet-r and p15A-f.

Cloning of the T6SS cluster.

RecET direct cloning was performed as described previously (55). Briefly, 40-μL overnight cultures of E. coli GB05RedTrfA were subcultured in 1.4 mL fresh LB and incubated at 30°C for 2 h with shaking at 950 rpm. The ETγA operon in the pSC101-BAD-ETgA-tet plasmid was induced by adding 35 μL l-arabinose at 37°C for 40 min. Ten micrograms of digested genomic DNA and 1 μg of the linear vector were coelectroporated to E. coli cells at 2.5 kV using a 2-mm-gap cuvette. Cells were resuspended in LB, recovered at 37°C for 1 h, and plated on 15 μg/mL chloramphenicol. Colonies were screened for the presence of correct plasmids by restriction digestion using KpnI, and the regions of homology arms were confirmed by sequencing.

Gene deletion in pT6S plasmids.

To delete genes in the pT6S plasmids, we employed the rhamnose-inducible γβαAA system encoded in the chromosome of GB05RedTrfA (55). Briefly, target genes could be replaced with PCR products of the ampicillin-resistant gene bla with flanking 50-bp arms homologous to target genes through the γβαAA system-mediated recombination. PCR fragments were electroporated into GB05RedTrfA cells with pT6S, and the successful clones were selected on LB agar plates with 50 μg/mL carbenicillin and 25 μg/mL chloramphenicol. Positive colonies were further verified by restriction digestion of KpnI, and the region of homology arms was confirmed by sequencing.

Bacterial killing assay.

Overnight cultures of killer and prey cells were transferred into fresh LB medium with the appropriate antibiotics at a ratio of 1:100 and grown to an OD₆₀₀ of 1, except the strains with the pT6S_Tet, pT6S_N3-Tet, and pΔvasK_Tet plasmids were induced with aTc to a final concentration of 100 ng/mL at an OD₆₀₀ of 0.6 and incubated until they reached an OD₆₀₀ of 1. Cells were then harvested by centrifugation at 4,500 × g for 3 min, and the pellets were resuspended in fresh LB. Killer cells were mixed with prey cells at a ratio of 10:1 (1:10 and 10:10 for killing with P. aeruginosa) and spotted on LB agar plates or LB agar plates containing 100 ng/mL aTc for induction of the expression of T6SS genes. In the TsiI and TseC complementation assay, 0.1 mM IPTG and 0.01% arabinose were also added for induction of the pPSV37 and pBAD plasmids, respectively. After coincubation for 3 h at 37°C (with V. parahaemolyticus incubated at 30°C, P. syringae pv. syringae at 28°C, and P. asymbiotica at 28°C), cells were resuspended in 500 μL fresh LB, and 10-fold serial dilutions were performed on LB agar plates with antibiotics to assess the survival of prey cells. The mean log₁₀ CFU per milliliter of recovered cells was calculated, and error bars indicate mean ± standard deviation (SD) of values from three biological replicates. The two-tailed Student's t test was used to determine P values.

Protein secretion assay.

Overnight cultures were transferred into fresh LB with appropriate antibiotics at a ratio of 1:100 and incubated at 37°C with shaking at 200 rpm to an exponential-phase OD₆₀₀ of 1, except the strains with the pT6S_{Tet_eff} and pT6S_{N3-Tet_eff} plasmids were induced with aTc to a final concentration of 100 ng/mL at an OD₆₀₀ of 0.6 and incubated until they reached an OD₆₀₀ of 1. Two milliliters of cells at an OD₆₀₀ of 1 was collected by centrifugation at 10,000 × g for 2 min at room temperature. The pellets were resuspended in 100 μL 2× SDS-loading buffer and used as whole-cell samples. Supernatants were collected by centrifugation two times at 10,000 × g for 2 min at room temperature and precipitated in 20% (vol/vol) trichloroacetic acid (TCA) at −20°C for 1 h and then centrifuged at 15,000 × g for 30 min at 4°C. Acetone was used to wash the protein pellets twice. The air-dried pellets were mixed with 100 μL 2× SDS-loading buffer. Whole cells and secretion samples were heated at 98°C for 10 min.

Western blot analysis.

Samples of whole-cell and secretion proteins were resolved on SDS-PAGE gels and transferred to the polyvinylidene difluoride (PVDF) membrane (Bio-Rad). The membrane was blocked with 5% (wt/vol) nonfat milk in TBST buffer (50 mM Tris, 150 mM NaCl, 0.1% [vol/vol] Tween 20 [pH 7.6]) for 1 h at room temperature. Then the membrane was incubated with primary antibody in TBST with 1% (wt/vol) nonfat milk. The membrane was washed three times with TBST buffer and incubated with horseradish peroxidase (HRP)-conjugated secondary antibody in TBST with 1% (wt/vol) nonfat milk for 1 h followed by detection using ECL enhanced chemiluminescence solution (Bio-Rad). The beta subunit of RNA polymerase (RpoB) antibody (Biolegend; product 663905) was used as a control for cell lysis. The polyclonal antibody to Hcp was custom-made by Shanghai Youlong Biotech. The secondary antibodies were purchased from ZSGB-Bio (product no. ZB-2305 for mouse and ZB-2301 for rabbit).

Fluorescence microscopy.

Strains were grown in LB to an exponential-phase OD₆₀₀ of 1 prior to imaging, except the strains with the pT6S_Tet, pT6S_N3-Tet, and pΔvasK_Tet plasmids were induced with aTc (100 ng/mL) at an OD₆₀₀ of 0.6 and then grown to an OD₆₀₀ of 1. Cells were harvested by centrifugation at 2,500 × g for 8 min and resuspended to an OD₆₀₀ of 10 with 0.5× phosphate-buffered saline (PBS). One microliter of cells was spotted onto a 1% (wt/vol) agarose–0.5× PBS pad. Fluorescence images were acquired with a Nikon Ti-2E inverted microscope using NIS-Elements AR 5.20.00 software, and Fiji was used for image analysis (80).

Plasmid stability detection.

Overnight cultures were transferred into fresh LB with or without 25 μg/mL chloramphenicol at a ratio of 1:100 and incubated at 37°C with shaking at 200 rpm for continuous culture. Samples were collected every 24 h and analyzed by flow cytometry. For the pT6S_Tet and pΔvasK_Tet plasmids containing strains, the aTc inducer (100 ng/mL) was added every 24 h. Bacteria were collected by centrifugation at 2,500 × g for 8 min and resuspended in PBS. Samples were analyzed with a flow cytometer (Beckman). Data acquisition and analysis were performed on CytExpert 2.3 software. The 488-nm laser was used for excitation of GFP, and the GFP signal was measured in the fluorescein isothiocyanate (FITC) channel. The parameters were set as follows: forward scatter (FSC) threshold, 4,000; FSC gain, 200; side scatter (SSC) gain: 100; FITC gain, 400; events to record, 10,000; events to display, 1,000.

Growth curve.

Overnight cultures were transferred into fresh LB with appropriate antibiotics at a ratio of 1:100 and incubated at 37°C with shaking at 200 rpm for 12 h. Cells were collected by centrifugation, resuspended to OD₆₀₀ of 3, and diluted 1:100 in fresh LB with antibiotics and aTc (100 ng/mL), the latter for inducing the expression of pT6S_Tet and pT6S_N3-Tet plasmids. Bacteria were cultured in 96-well plates at 37°C with shaking for 10 h. The absorbance was measured at 600 nm using a microtiter plate reader (Biotek Synergy H1). LB was used as the blank control for normalization. Error bars indicate the mean ± standard deviation of values from three biological replicates.