Delivery of an Rhs‐family nuclease effector reveals direct penetration of the gram‐positive cell envelope by a type VI secretion system in Acidovorax citrulli

Abstract The type VI secretion system (T6SS) is a double‐tubular nanomachine widely found in gram‐negative bacteria. Its spear‐like Hcp tube is capable of penetrating a neighboring cell for cytosol‐to‐cytosol protein delivery. However, gram‐positive bacteria have been considered impenetrable to such T6SS action. Here we report that the T6SS of a plant pathogen, Acidovorax citrulli (AC), could deliver an Rhs‐family nuclease effector RhsB to kill not only gram‐negative but also gram‐positive bacteria. Using bioinformatic, biochemical, and genetic assays, we systematically identified T6SS‐secreted effectors and determined that RhsB is a crucial antibacterial effector. RhsB contains an N‐terminal PAAR domain, a middle Rhs domain, and an unknown C‐terminal domain. RhsB is subject to self‐cleavage at both its N‐ and C‐terminal domains and its secretion requires the upstream‐encoded chaperone EagT2 and VgrG3. The toxic C‐terminus of RhsB exhibits DNase activities and such toxicity is neutralized by either of the two downstream immunity proteins, RimB1 and RimB2. Deletion of rhsB significantly impairs the ability of killing Bacillus subtilis while ectopic expression of immunity proteins RimB1 or RimB2 confers protection. We demonstrate that the AC T6SS not only can effectively outcompete Escherichia coli and B. subtilis in planta but also is highly potent in killing other bacterial and fungal species. Collectively, these findings highlight the greatly expanded capabilities of T6SS in modulating microbiome compositions in complex environments.


INTRODUCTION
To survive in natural and host environments, microbes have developed a number of strategies to outcompete neighboring species, including the production of diffusible molecules, releasing phage or phage-like particles, and utilization of contactdependent antibacterial weapons [1][2][3][4][5][6][7][8] .One such lethal weapon is the type VI secretion system (T6SS) that gram-negative bacteria commonly employ to deliver toxins into neighboring cells through a phage-tail-like double tubular contractile structure 7,9,10 .The T6SS ejects a spear-like tube with toxic effectors outward in milliseconds, with enough power to drill through the cell wall and two cellular membranes of gram-negative cells [10][11][12] .T6SSs show great diversity in the arsenal of toxic effectors and, as a result, potent killing capabilities against susceptible bacteria, amoeba, yeast, and other eukaryotic cells 7,[13][14][15][16][17] .
To defend against the T6SS attack, bacteria have developed both specific and nonspecific protective mechanisms 15,18,19 .The effector-specific protection is mediated by effector-cognate immunity proteins that can neutralize effectors through interaction 15,20,21 .However, such specific protection can be insufficient in the absence of bacterial innate-immunity-like nonspecific mechanisms 18,19 , including the production of exopolysaccharide (EPS) capsules, oxidative stress response, and envelope stress responses 19,[22][23][24] .Notably, most T6SS-susceptible bacteria are gram-negative, while gram-positive bacteria are not generally believed to be targets of T6SS [25][26][27] .The T6SS resistance of gram-positive cells has been largely attributed to a stronger envelope consisting of thicker peptidoglycan layers, polymers of teichoic acids, and surface proteins, which may present an impenetrable barrier to the T6SS spear 28 .Although T6SSsecreted cell-wall-lysing effectors have been shown effective against gram-positive Bacillus subtilis cells 26,29 , and the T6SS of Pseudomonas chlororaphis appears to induce spore formation of B. subtilis 30 , it remains elusive whether the T6SS is capable of penetrating into gram-positive cells to achieve direct cytosol-to-cytosol delivery.
Acidovorax citrulli (AC) is a gram-negative seed-borne plant pathogen that causes bacterial fruit blotch of cucurbit crops (including melons and watermelons) and is extremely difficult to eradicate 31 .Similar to many human pathogens, A. citrulli has rapidly spread globally in the last few decades and is the most economically important pathogen to cucurbits 31,32 .A. citrulli possesses a number of virulence traits, of which the type III secretion system is required for infection 33 and the T6SS appears to contribute to seed-to-seedling transmission 34 .Here, we report that the T6SS of A. citrulli type strain AAC00-1 possesses potent antibacterial activities against gram-negative and gram-positive bacteria.Using secretome, bioinformatic, and genetic-screening analyses, we characterized T6SS-dependent effectors in AC and determined two Rhs (rearrangement hot spot) family effectors, RhsB and RhsE, that are crucial for outcompeting B. subtilis.We report that RhsB is a nuclease effector whose secretion is mediated by the upstream-encoded chaperone EagT2 and VgrG3.We reveal that RhsB is subject to self-cleavage that is crucial for its toxicity.In summary, our results demonstrate that the T6SS of AC could breach the thick cell envelope of gram-positive B. subtilis and deliver the RhsB nuclease to B. subtilis in a cytosol-to-cytosol manner.The diverse antibacterial and antifungal T6SS functions in the plant pathogen A. citrulli highlight an expanded ecological impact of the T6SS in multispecies communities far beyond the pathogenesis of human diseases.

AC T6SS displays potent antibacterial and antifungal activities
To test whether the A. citrulli AAC00-1 strain (AC) possesses a functional T6SS, we first examined the expression and secretion of Hcp that forms the shaft of the T6SS spear-like secreted structure 9,10 .We found that Hcp was well expressed during aerobic growth in LB but its secretion was increased substantially during the transition from exponential to stationary phase growth (Figures 1A and S1A).These results indicate that the T6SS of AC is active and functional under lab conditions.
Next, we tested whether the T6SS of AC exhibits antibacterial and antifungal activities as previously reported for other T6SS systems 13,15,25 .First, we constructed a T6SS-null mutant lacking the tssM gene that encodes an essential T6SS inner membrane protein 7 .Using interspecies competition assays, we compared the survival of different prey species against wild type or the ΔtssM mutant of AC. Results show that the T6SS could effectively kill bacterial competitors Escherichia coli, Enterobacter cloacae, P. aeruginosa, B. subtilis, and Mycobacterium smegmatis, as well as fungal competitors Pichia pastoris, Saccharomyces cerevisiae, and Candida albicans (Figures 1B,C and S1,B-I).These prey strains exhibited variable susceptibilities to the T6SS ranging from several-fold reduction for C. albicans to near one-million-fold reduction in survival for E. coli between wild-type and the ΔtssM samples (Figure 1B,C).Collectively, the T6SS of AC targets a wide range of microbial competitors.
To validate the predicted effectors, we first employed a secretome approach using LC-MS/MS analysis to compare the secreted proteins between wild type and the ΔtssM mutant.We successfully detected unique peptides corresponding to Hcp, all the 12 VgrG proteins, and 7 of the predicted T6SS effectors (Figure 2B and Table S1).These results suggest that the small T6SS operons are also actively expressed, similar to the large cluster.
Because VgrG proteins often directly interact with their cognate effectors for secretion, we also used pull-down assays with a select group of VgrG proteins as bait to determine their interacting effectors.We expressed plasmidborne N-terminal 6His-tagged VgrG1, VgrG3, VgrG5, and VgrG10, respectively, in AC and analyzed the eluted proteins from His-affinity columns using LC-MS/MS analysis.Using this approach, we detected nine Rhs-family effectors, suggesting it is an effective approach to identify VgrG-dependent effectors (Figure 2C and Table S2).

RhsB and RhsE are critical for T6SS antibacterial activities
To determine whether these effectors contribute to interspecies competition, we constructed a panel of effector mutants by deletion or insertion.Using bacterial competition assays against E. coli and B. subtilis preys, we screened these effector mutants and found that, while most mutants exhibited wild-type level killing abilities against prey cells, two mutants (Aave_0499, named RhsB and Aave_2838, named RhsE) showed impaired killing activities (Figures 2D,E and  S2A,B).To test whether deletion of rhsB or rhsE affects T6SS secretion, we compared Hcp secretion of these two mutants with the wild type.Results show that mutants of rhsB and rhsE had wild-type level of Hcp secretion (Figure S2C).Collectively, these results suggest RhsB and RhsE are key antibacterial effectors.

RhsB is neutralized by its downstream immunity proteins RimB1&2
Both rhsB and rhsE encode Rhs-family proteins with an Nterminal PAAR domain, a middle YD-repeat/Rhs domain, and a C-terminal domain of unknown function (Figure 3A).Using Phyre2 sequence analysis, we found no significant hit for RhsE but the C-terminal domain of RhsB is distantly related to a virus-type replication-repair nuclease (PDB: 4qbn) with 24% identity 40 .Downstream of rhsB reside two small predicted genes of unknown functions sharing 69% identity and equal length in the protein sequence.We name the two downstream genes rimB1 and rimB2 (Rhs-immunity B), respectively.
To test if the C-terminus of RhsB (RhsB CT ) is toxic, we expressed it using an arabinose inducible vector (pBAD) in E. coli and compared cell survival in the presence of arabinose (induced) or glucose (repressed) 41 .Results show that RhsB CT was highly toxic, reducing survival by 100 folds when induced (Figure 3B).Although RhsB CT belongs to the PD-(D/E) XK superfamily 42 , there are no obvious catalytic sites in 4qbn or RhsB CT .We constructed several point mutations around the PD site and tested their toxicity in E. coli.Both the KE-AA  mutant and the KD-AA mutant displayed severely attenuated toxicity compared with the wild-type RhsB CT (Figure 3B).Interestingly, the K1561A mutant exhibited stronger toxicity than its parental protein (Figure 3B).And the results of Western blot analysis confirmed that these mutant proteins were expressed (Figure S3A).
To test whether RimB proteins confer protection, we first expressed RhsB with and without the downstream genes in E. coli using the arabinose-inducible pBAD vector.Results show that RhsB toxicity can be neutralized by co-expression of RimB1 or RimB2 similarly (Figure 3C).We also constructed the ΔrhsB-rimB1&2 mutant lacking the immunity genes and transformed it with an empty vector or vectors encoding one or both immunity proteins.Competition analyses against wild-type AC or the ΔtssM mutant reveal that survival was significantly increased when immunity genes were ectopically expressed in the ΔrhsB-rimB1&2 mutant (Figures 3D  and S3B).Using bacterial two-hybrid assays 16,43 , we found that both RimB1 and RimB2 could bind to the nontoxic RhsB KE-AA construct, while the Aave_0500 protein, encoded downstream of rimB2, did not interact with any of the proteins (Figure 3E).As control, Pal and TolB proteins showed positive interaction 44,45 .These results collectively suggest both immunity proteins could confer protection.To examine if RhsB exhibits nuclease activity, we purified the C-terminal domain of RhsB and its inactive mutants RhsB KE-AA and RhsB KD-AA , all with an N-terminal 6His-SUMO tag.Results show that only the wild-type RhsB CT but not the two catalytically inactive mutants could efficiently degrade DNA (Figures 3F and S3C).
Collectively, these results indicate that RhsB is a nuclease effector and has two cognate immunity proteins RimB1 and RimB2.

RhsB is delivered into the cytosol of B. subtilis
Next, we tested whether RhsB is delivered to B. subtilis using a competition assay.The prey B. subtilis was transformed with the empty vector pHT01 or vectors expressing one or both immunity proteins, separately.Results show that all B. subtilis strains expressing the immunity genes survived significantly better than the one expressing the vector alone when they competed with wild-type AC (Figures 3G  and S3D).When the ΔrhsB mutant was used as the killer, survival of B. subtilis with vector alone was increased to a similar level to that of B. subtilis expressing immunity genes.In addition, to test whether RhsB could be translocated into B. subtilis through diffusion, we co-incubated B. subtilis with wild-type AC or the ΔtssM mutant in liquid culture, respectively.There was no significant difference in B. subtilis survival (Figure S3E,F).These results collectively indicate that RhsB is delivered by the T6SS to B. subtilis and its toxicity is neutralized by the RimB immunity proteins expressed in B. subtilis.We also constructed the chromosomal nucleaseinactivated mutant RhsB KE-AA and found this mutant also exhibited attenuated killing ability of B. subtilis to the same level as deletion of rhsB did (Figure 3H), suggesting RhsB intoxicated B. subtilis due to its nuclease activity.

RhsB secretion is mediated by the chaperone EagT2 and VgrG3
RhsB is encoded in a gene cluster with its immediate upstream genes encoding VgrG3 and a DUF1795-domain chaperone EagT2 46,47 .To determine whether these two proteins are important for RhsB secretion, we first performed a pull-down analysis using the nontoxic RhsB ΔC mutant that lacks the C-terminus toxin domain.The results confirmed that RhsB could directly interact with EagT2 and VgrG3 (Figures 4A and S4A).We next constructed the deletion mutants of eagT2 and vgrG3 and tested their killing abilities against the ΔrhsB-rimB1&2 prey cells.The ΔeagT2 displayed a killing defect similar to the ΔrhsB mutant, suggesting that EagT2 is crucial for RhsB secretion (Figures 4B and S4B).In contrast, the ΔvgrG3 mutant showed impaired killing ability relative to wild type but still significantly stronger killing ability than the ΔrhsB mutant (Figures 4B and S4B).Considering the ΔrhsB-rimB1&2 cells possess active T6SS and could donate VgrG3 but not EagT2 into sister cells, we postulated that the partial defect in the ΔvgrG3 mutant may result from substrate exchange between T6SS neighboring cells, a known phenotype previously described in Vibrio cholerae 11 .Indeed, when we performed competition analysis against the B. subtilis prey, results show that the ΔeagT2 and the ΔvgrG3 mutants exhibited impaired killing activities similar to the ΔrhsB mutant (Figures 4C and S4C).

RhsB is self-cleaved at both N-and C-terminus
We have previously described a self-cleavable Rhs effector TseI in Aeromonas dhakensis 36 .Although RhsB possesses an N-terminal PAAR domain that is absent in TseI, sequence alignment of these two proteins shows that RhsB contains the conserved aspartic cleavage residues around the  S1.Different colors of dots indicate the corresponding proteins pulled down with VgrG1, VgrG3, VgrG5, and VgrG10 shown in (C) and Table S2.(B) Secretome analysis summarizing the detected VgrG and effector proteins between the WT and the T6SS-null mutant (ΔtssM).(C) Pull-down analysis summarizing VgrG and effector proteins pulled down with VgrG1, VgrG3, VgrG5, and VgrG10.For (B) and (C), all represented proteins were detected with unique peptides.The Y-axis shows the number of peptides detected of the corresponding protein shown in X-axis, while the specific number of peptides, unique peptides, and coverage are shown in Tables S1 and S2, respectively.Survival of E. coli MG1655 (D) and B. subtilis PY79 (E) attacked by AAC00-1 WT, ΔtssM, and effector inactivation mutants for 1 and 3 h, respectively.Survival of prey cells was determined by serial dilutions on selective media.Error bars indicate the standard deviation of at least three biological replicates and statistical significance was calculated using one-way ANOVA test for each group, **p < 0.01, ****p < 0.0001.DL, detection limit.To test whether RhsB is cleaved, we expressed the nontoxic RhsB KE-AA mutant protein in E. coli since the wild type is highly toxic.Western blot analysis shows that the N-terminus and the C-terminus of RhsB KE-AA were also cleaved (Figure 4E).Mutation of the conserved residue D1484 to alanine abolished C-terminal cleavage while the D280 to alanine mutation impaired both N-and C-terminal cleavage (Figure 4E).We also performed in vitro expression of RhsB using the In Vitro Protein Synthesis kit consisting of defined transcription-translation protein components.Western blot analysis of in vitro expressed RhsB confirms that RhsB is self-cleaved at the C-terminus while the N-terminal cleavage is indiscernible due to high background noise signals under the in vitro expression condition (Figures 4F and S4D).In addition, we detected RhsB cleavage in the cytosol of A. citrulli by expressing its nontoxic mutant RhsB KE-AA in AC wild type, T6SS-null mutant ΔtssM, and ΔrhsB, suggesting that self-cleavage of RhsB could occur before secretion and be independent of T6SS functionality (Figure S4E).
In addition, we purified the His-SUMO-tagged RhsB KE-AA with His-tagged EagT2 together, while expressing RhsB KE-AA alone had much less yield.The purified N-terminal His-SUMOtagged RhsB KE-AA protein exhibited clear cleavage products (Figure S4F).Using N-terminal protein sequencing, we determined the C-terminal cleavage site after the residue tryptophan 1488, which is consistent with the predicted site based on the sequence alignment (Figure 4D).However, the Nterminal cleavage site was unsolved due to poor detection by N-terminal sequencing.Nonetheless, Western blot analysis of the predicted N-terminus (RhsB NT ) expressed in E. coli shows a nearly-identical size to the cleaved N-terminal product from full-length RhsB (Figure S4G).The predicted cleavage site of RhsB is also consistent with the recently reported N-terminal cleavage site of Rhs1 in Photorhabdus laumondii 48 .

Cleavage is crucial for RhsB-mediated toxicity
To test whether cleavage is important for RhsB functions, we first examined RhsB-dependent intra-and inter-species competition.Results show that the cleavage-defective mutants rhsB D280A and rhsB D1484A displayed impaired killing abilities, similar to the ΔrhsB mutant, against the ΔrhsB-rimB1&2 and B. subtilis prey strains (Figures 4G,H  and S4H,I).To test whether cleavage also affects the intracellular toxicity of RhsB, we expressed the full-length RhsB and its cleavage-defective mutants using arabinose-inducible pBAD vectors in E. coli and compared their effects on survival after arabinose induction.The results show that the survival of cells expressing wild-type RhsB was reduced by 10 folds compared with that of the cells expressing the cleavage-defective mutants (Figure 4I).Collectively, these data show that cleavage is crucial for RhsB-mediated intracellular toxicity and competition.

AC employs its T6SS to outcompete other bacteria in planta
To explore whether the T6SS confers competitive fitness in planta, we performed competition assays between AC and E. coli MG1655 or B. subtilis PY79 using the Nicotiana benthamiana leaf infection model 49 .The results show that wildtype AC could effectively outcompete prey cells in comparison with the ΔtssM mutant (Figures 5A,B and S5A,B).Because E. coli and B. subtilis strains are not natural plant pathogens, we next tested whether the T6SS of AC can kill P. syringae pv.syringae, which is one of the most common plant pathogens that infect the phyllosphere 50 .Through performing competition assay in vitro and in planta (Figures 5C  and S5C), we found that AC significantly inhibited the growth of P. syringae in both conditions.We also validated the T6SSmediated competition between AC and E. coli using a natural host of AC infection, Citrullus lanatus (Figure S5D,E).Collectively, these results indicate that the T6SS of AC is functional and promotes its fitness during plant infection.

DISCUSSION
In conclusion, we report the potent activities of A. citrulli T6SS against a panel of bacterial and fungal species.We have systematically identified the T6SS effectors in A. citrulli and determined two Rhs-family effectors, RhsB and RhsE, which contribute the most to killing both gram-negative and grampositive bacteria.We have determined key residues and interacting proteins that are required for the activity, cleavage, and secretion of RhsB.We further show that the AC T6SS is active against competing species in planta.These data underscore the broad ecological impact of the T6SS on diverse natural microbial communities, beyond the previously known susceptible microbes and pathogens of human infections (Figure 5D).
These findings have several important implications.First, in comparison with previous efforts focused on human pathogens, T6SS functions are substantially less studied in plant pathogens despite the vast number of effectors, distinct ecological environments, and enormous impact on world economy and food security.Previous studies have reported key antibacterial functions of the T6SS in several plant-associated bacteria, including Agrobacterium tumefaciens, Burkholderia glumae, and P. protegens [51][52][53] , but all are limited to inhibiting gram-negative bacteria.Our findings will likely stimulate further research in novel T6SS functions of plant pathogens.
Second, our findings address a long-standing critical question on whether contraction of the T6SS sheath is powerful enough to drill through the thick gram-positive cell envelope 10,28 .A recent report shows that the T6SS of Acinetobacter baumannii could secrete a cell-wall-lysing effector Tse4 to kill gram-positive bacteria 29 .However, it remains unclear whether the secreted Tse4 destroys the grampositive cell wall from the outside or from inside, or whether Tse4 can be delivered into the cytosol of gram-positive cells.Since expressing immunity proteins in B. subtilis confers protection against the delivered RhsB nuclease, our study provides the first evidence that the T6SS of AC could penetrate through the cell envelope and achieve cytosol-to-cytosol delivery into gram-positive cells.Notably, other effectors provide redundant functions to RhsB in killing B. subtilis, as evidenced by the attenuated but not abolished killing activities of the ΔrhsB.Further research is required to examine this penetration process in detail to elucidate the cellular death pathways using combinatorial effector mutations.
In addition, our results reveal that the AC T6SS is capable of killing a variety of target species that differ greatly in cell envelope composition and defense mechanisms.These functions are likely dependent on the T6SS-dependent effectors, including many belonging to the Rhs family.The findings on the self-cleavage of RhsB are consistent with previous reports that Rhs-family effectors in A. dhakensis (TseI), P. syringae (PSPTO_5438), P. aeruginosa (PA2684), V. parahaemolyticus (VP1517), and P. laumondii (Rhs1) are subject to self-cleavage 36,48 .Although critical cleavage residues have been determined, the molecular mechanism remains elusive and warrants future research.In addition, it is likely that different effectors are involved in targeting different species.The effector mutant library constructed in this study will help elucidate effector specificity against different bacterial and fungal competitors in future research.
Lastly, the demonstration that the AC T6SS can kill both gram-negative and gram-positive bacterial and fungal species highlights the potential of developing T6SS-based treatment strategies as green alternatives to chemical agents in mitigating infectious diseases in agricultural and medical applications.In short, there seems to be no barrier too thick for the T6SS to break in the microbial world.

Bacterial strains and plasmids
Plasmids and strains are listed in Table S3.All strains were routinely grown in LB, 7H9, 7H10, or YPD media following standard culturing conditions for each species.Antibiotics were used as follows: kanamycin (25 µg/ml for bacterial strains, and 100 µg/ml for fungal strains), chloramphenicol (10 µg/ml), irgasan (25 µg/ml), gentamicin (20 µg/ml), and ampicillin (50 µg/ml).Gene expression vectors were constructed as previously described 54 .Full-length RhsB, RhsB CT , or its mutants were cloned into the pBADexpression vector, respectively.Cells harboring different constructs were grown in LB supplemented with 0.2% (w/v) glucose to repress gene expression.All constructs were confirmed by sequencing.All primers and plasmids are available upon request.

Bacterial competition assay
For AC intraspecies competition, exponential-phase killer cells and stationary-phase prey cells were mixed at a ratio of 20:1 (killer:prey) and spotted on LB agar plates for 24 h at 28°C.For the interspecies competition with M. smegmatis, exponential-phase killer cells and stationary-phase prey cells were mixed at a ratio of 2:1 (killer:prey), spotted on 7H10 plates, and co-incubated for 8 h at 37°C.For the interspecies competition with other strains, exponential-phase killer cells and stationary-phase prey cells were mixed at a ratio of 10:1 (killer:prey), spotted on LB agar plates, and co-incubated for 10 h at 28°C (for S. cerevisiae), or 3 h at 28°C (for P. syringae and all the fungal prey cells), or 3 h at 37°C (for all other bacterial prey cells) unless stated otherwise.For the competition assay in liquid media, killer and prey cells were mixed with the same ratio used in the competition assay on solid media and incubated in liquid LB for 3 h at 37°C with shaking.Survival of prey cells was quantified by serial dilution and plating on selective media.For the relative survival in Figure 3D, it was calculated by using the final survival of the specific prey cells attacked by wild type divided by the survival of the same prey cells competed with the ΔtssM mutant.For the relative survival in Figure 4B,G, it was calculated by using the final survival of the prey cells competed with different T6SS-active killer strains (T6SS + ) divided by the survival of prey cells competed with the ΔtssM mutant.Error bars show standard deviation of at least three biological replicates.

Competition assay in planta
For competition assay using N. benthamiana, 4-5 weeks old plants were inoculated on the second and third leaves.For competition assay using C. lanatus "Jingxin No. 2", 2-weekold plants were inoculated on the cotyledons.Selected regions were injected with bacterial mixtures, AC (OD 600 ~5) mixed with prey cells (OD 600 ~0.5) as indicated, into the abaxial side of the leaves.After 6 h infection at room temperature, discs of infiltrated areas were collected and quickly frozen by liquid nitrogen, followed by grinding.The samples were then resuspended in liquid LB.Prey cells were quantified by ten-fold serial dilution and plating on selective media.Error bars show standard deviation of at least three biological replicates.

Protein secretion assay
Cultures were grown aerobically in liquid LB at 28°C to OD 600 ~2 and harvested by centrifugation at 2500g for 8 min.Pellets were resuspended in fresh LB and incubated at 28°C for 1 h without shaking.Cells were spinned down twice at room temperature at 10,000g for 2 min.Pellets were resuspended in SDS-loading buffer (Epizyme Biotech) and used as whole-cell samples.Supernatants were precipitated in TCA (trichloroacetic acid, 20% [v/v]) at −20°C for 30 min and centrifuged at 15,000g for 30 min at 4°C.The pellets were then washed with acetone and the air-dried pellets were resuspended in SDSloading buffer.Whole-cell and secretion samples were boiled for 10 min prior to SDS-PAGE and Western blot analysis.For secretome analysis, the supernatants of overnight cultures were collected and subject to SDS-PAGE analysis.Gel slices containing secreted proteins were cut out and sent for LC-MS/ MS analysis performed by the Instrumental Analysis Center of Shanghai Jiao Tong University.

Protein purification and enzymatic assay
RhsB and its variants were expressed using the pETSUMO vectors in E. coli BL21(DE3).Cells were grown in liquid LB medium to exponential phase (OD 600 ~0.6) at 37°C.For the expression of His-SUMO-RhsB CT or its non-toxic mutants, cells were induced with 1 mM IPTG at 37°C for 5 h.For the expression of His-SUMO-RhsB KE-AA , cells were induced with 1 mM IPTG at 20°C for 18 h.The cells were harvested by centrifugation at 4500g for 10 min.The pellets were resuspended in lysis buffer (20 mM Tris-HCl, 150 mM NaCl, 10 mM imidazole, pH 8.0) and lysed by sonication.Lysates were centrifuged at 15,000g for 20 min and the supernatants were incubated with Ni-NTA resin (Smart-lifesciences). Proteins were eluted in elution buffer (20 mM Tris-HCl, 150 mM NaCl, and variable concentrations of imidazole, pH 8.0).Eluted samples were analyzed by SDS-PAGE analysis.
Protein activity in vitro was detected by incubating with 100 ng plasmid at 37°C for 1 h.NEB CutSmart buffer (50 mM potassium acetate, 20 mM tris-acetate, 10 mM magnesium acetate, 100 µg/ml bovine serum albumin, pH 7.9) was chosen as the reaction buffer.Purified proteins (0.1 μg) and 0.5 units DNase I (positive control) were used separately in each reaction.

Protein pull-down assay
Genes of interest were cloned into pET and pBBRT vectors for expression.Cells were grown in liquid LB medium with appropriate antibiotics to exponential phase (OD 600 ~0.6) at 37°C and induced with 1 mM IPTG overnight at 20°C for pET vectors or 100 ng/ml anhydrotetracycline (aTc) for 3 h at 37°C for pBBRT vectors.Cells were harvested by centrifugation at 4500g for 10 min, resuspended in lysis buffer (20 mM Tris, pH 8.0, 500 mM NaCl, 50 mM imidazole with protease inhibitor [Thermo Scientific]), and lysed by sonication.After cell debris was removed by centrifugation at 15,000g for 5 min, supernatants were mixed and incubated with Ni-NTA resin (Smartlifesciences) at 4°C for 1 h.The samples were then washed five times with wash buffer (20 mM Tris, pH 8.0, 500 mM NaCl, 50 mM imidazole), and eluted in elution buffer (20 mM Tris, pH 8.0, 500 mM NaCl, 500 mM imidazole).Input and eluted samples were boiled for 10 min prior to SDS-PAGE and Western blot analysis.

Protein toxicity assay
Cells harboring different plasmids were grown on LB agar plates with appropriate antibiotics and 0.2% (w/v) glucose at 30°C overnight.Cells were then harvested and resuspended in fresh liquid LB medium and grown to OD 600 = 1.A series of ten-fold dilutions were plated on LB agar plates containing 0.1% (w/v) L-arabinose or 0.2% (w/v) glucose for induction and repression, respectively.For Figure 4I, the cells were induced with 0.1% (w/v) L-arabinose for 2 h.The survival of E. coli before and after induction was enumerated by ten-fold dilutions on LB agar plates containing 0.2% (w/v) glucose and appropriate antibiotics.Each experiment was performed at least two times, with one representative experiment shown.

Bacterial two-hybrid assay
Proteins were fused to the T25 and the T18 domains of the Bordetella adenylate cyclase as previously described 54 .Plasmids encoding fusion proteins were co-expressed in the reporter strain BTH101.Single colonies for each transformation were inoculated into 300 μl of LB medium, respectively, and grown for 4 h at 30°C with shaking.Each culture (3 μl) was spotted onto LB agar plates supplemented with kanamycin, ampicillin, IPTG (0.02 mM), and X-Gal (40 µg/ml).Plates were incubated for 6 h at 30°C and then 10 h at room temperature.The experiments were performed in triplicate and a representative result is shown.

Bioinformatic and statistical analysis
Gene sequences of A. citrulli AAC00-1 were downloaded from GenBank NC_008752.1.DNA sequences were managed and analyzed by Benchling.RhsB sequence was analyzed with Phyre2 55 .The model figure was generated using BioRender (https://biorender.com).Statistical analysis was performed using the GraphPad Prism software (9.3.0).The statistical significance was evaluated using a two-tailed Student's t-test or one-way ANOVA test as indicated.

N-terminal Edman sequencing
Purified proteins were subjected to an SDS-PAGE gel.Protein bands corresponding to RhsB M-fragment and C-terminus were excised individually and sent for Edman sequencing that was performed at the BiotechPack Scientific Company.

Figure 1 .
Figure 1.The T6SS of A. citrulli displays antibacterial and antifungal activities.(A) Growth curve of AC at 28°C.The secretion levels of Hcp were examined during growth.Competition assay of wild type (WT) and the T6SS-null ΔtssM mutant against a panel of bacterial (B) and fungal (C) strains.Survival of prey cells was determined by serial dilutions on selective media.Error bars indicate standard deviation of three biological replicates and statistical significance was calculated using a two-tailed Student's t-test for each group, ***p < 0.001, ****p < 0.0001.DL, detection limit.

Figure 2 .
Figure 2. Identification of T6SS effectors.(A) T6SS gene cluster and operons.Predicted effector and chaperone genes are shown in red and blue, respectively.Asterisks indicate the corresponding proteins detected in the secretome results of (B) and TableS1.Different colors of dots indicate the corresponding proteins pulled down with VgrG1, VgrG3, VgrG5, and VgrG10 shown in (C) and TableS2.(B) Secretome analysis summarizing the detected VgrG and effector proteins between the WT and the T6SS-null mutant (ΔtssM).(C) Pull-down analysis summarizing VgrG and effector proteins pulled down with VgrG1, VgrG3, VgrG5, and VgrG10.For (B) and (C), all represented proteins were detected with unique peptides.The Y-axis shows the number of peptides detected of the corresponding protein shown in X-axis, while the specific number of peptides, unique peptides, and coverage are shown in TablesS1 and S2, respectively.Survival of E. coli MG1655 (D) and B. subtilis PY79 (E) attacked by AAC00-1 WT, ΔtssM, and effector inactivation mutants for 1 and 3 h, respectively.Survival of prey cells was determined by serial dilutions on selective media.Error bars indicate the standard deviation of at least three biological replicates and statistical significance was calculated using one-way ANOVA test for each group, **p < 0.01, ****p < 0.0001.DL, detection limit.

Figure 3 .
Figure 3. Characterization of RhsB function.(A) Operon structure of rhsB.The sequence of the RhsB C-terminus is aligned with the sequence of a virus-type replication-repair nuclease domain (VRR-Nuc, PDB: 4qbn).(B) Toxicity of expressing RhsB C-terminus (Parental) and its mutants in E. coli.(C) Toxicity of expressing RhsB with an empty vector (pBAD) or a vector carrying the immunity gene rimB1, rimB2, or rimB1 and rimB2 together as indicated in E. coli.For (B) and (C), all constructs were cloned on pBAD vectors and survival of E. coli was enumerated by serial plating on arabinose (induction) and glucose (repression) plates.(D) Competition assay of WT and the T6SS-null ΔtssM mutant against the effector-immunity deletion mutant ΔrhsB-rimB1&2 complemented with an empty vector (pBBR) or a vector expressing the immunity protein RimB1, RimB2, or RimB1 and RimB2 together as indicated.The data point indicates the relative survival of prey cells attacked by WT compared with that by T6SS mutant.(E) Bacterial two-hybrid analysis of RhsB-RimB1&2 interaction.Fusion proteins with T18 and T25 domains were co-expressed in the reporter strain BTH101 as indicated.(F) DNA degradation by RhsB C-terminus (RhsB CT ) and its mutants.The qualities of purified proteins are shown by SDS-PAGE analysis in Figure S3C.(G) Competition assay of WT, ΔtssM and ΔrhsB mutant against the B. subtilis PY79 carrying an empty vector (pHT01) or a vector expressing the immunity protein RimB1, RimB2 or RimB1 and RimB2 together as indicated.Cells of killer and prey were mixed at a ratio of 20:1 (killer:prey) and spotted on LB agar plates with 1 mM IPTG for 3 h at 37°C.(H) Competition assay of WT, ΔtssM, ΔrhsB, and rhsB KE-AA mutant against the B. subtilis PY79.For (D), (G), and (H), error bars indicate the standard deviation of three biological replicates and statistical significance was calculated using a one-way ANOVA test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 4 .
Figure 4. Secretion and cleavage of RhsB.(A) Interaction of RhsB ΔC with VgrG3 and chaperone EagT2.Pull-down analysis was performed using His-tagged sfGFP (control), VgrG3 or EagT2 and V5-tagged RhsB ΔC .Competition assay of WT, ΔtssM, ΔrhsB, ΔeagT2, and ΔvgrG3 mutants against the ΔrhsB-rimB1&2 (B) and B. subtilis PY79 (C).(D) Alignment of the predicted sequences of N-terminus and C-terminus cleavage regions between RhsB and TseI.(E) Western blot analysis of RhsB KE-AA and its cleavage-defective mutants D1484A, D280A, and D1484A D280A.Proteins were induced in E. coli with IPTG (0.01 mM).The nontoxic KE-AA mutant is used as the parental construct.The same constructs were also used for in vitro expression shown in (F).(F) In vitro expression of RhsB KE-AA and its cleavage-defective mutants D1484A, D280A, and D1484A D280A.The proteins were expressed using a PURExpress ® In Vitro Protein Synthesis Kit.Samples were subject to SDS-PAGE analysis, followed by Western blot analysis.Competition assay of WT, ΔtssM, ΔrhsB, rhsB D1484A , and rhsB D280A mutant against the ΔrhsB-rimB1&2 (G) and B. subtilis PY79 (H).(I) Toxicity of expressing RhsB and its cleavage-defective mutants in E. coli.All constructs were cloned on pBAD vectors and the survival of E. coli after arabinose induction was examined by serial plating on plates.For the relative survival in (B) and (G), it was calculated by using the final survival of the prey cells competed with different T6SS-active killer strains (T6SS + ) divided by the survival of prey cells competed with the ΔtssM mutant.For (B), (C), (G), and (H), error bars indicate the standard deviation of at least three biological replicates and statistical significance was calculated using a one-way ANOVA test for each group, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant.

Figure 5 .
Figure 5. T6SS-mediated competition in planta.Competition assay of WT and ΔtssM against the E. coli MG1655 (A) and B. subtilis PY79 (B).(C) Competition assay of WT and ΔtssM against P. syringae pv.syringae in vitro and in planta, respectively.Error bars indicate the standard deviation of three biological replicates and statistical significance was calculated using a two-tailed Student's t-test for each group, **p < 0.01, ***p < 0.001, ****p < 0.0001.(D) Schematic of the T6SS-dependent killing of AC against gram-negative and gram-positive competitors in planta.CW, cell wall; IM, inner membrane; OM, outer membrane.