Insights into the Cross-Immunity Mechanism within Effector Families of Bacteria Type VI Secretion System from the Structure of StTae4-EcTai4 Complex

The Gram-negative bacteria type VI secretion system (T6SS) has been found to play an important role in interbacterial competition, biofilm formation and many other virulence-related processes. The bacteria harboring T6SS inject the effectors into their recipient’s cytoplasm or periplasm to kill them and meanwhile, to avoid inhibiting itself, the cognate immunity proteins were produced to acts as the effector inhibitor. Tae4 (type VI amidase effector 4) and Tai4 (type VI amidase immunity 4) are newly identified T6SS effector-immunity (EI) pairs. We have recently solved the structures of StTae4-Tai4 and EcTae4-Tai4 complexes from the human pathogens Salmonella typhimurium and Enterobacter cloacae, respectively. It is very interesting and important to discover whether there is cross-neutralization between St- and EcTai4 and whether their effector inhibition mechanism is conserved. Here, we determined the crystal structure of StTae4 in complex with EcTai4. The solution conformation study revealed it is a compact heterotetramer that consists of an EcTai4 homodimer binding two StTae4 molecules in solution, different from that in crystal. A remarkable shift can be observed in both the flexible winding loop of StTae4 and protruding loop of EcTai4 and disulfide bonds are formed to stabilize their overall conformations. The in vitro and in vivo interactions studies showed EcTai4 can efficiently rescue the cells from the toxicity of its cognate effectors StTae4, but can not neutralize the toxic activities of the effectors from other families. These findings provide clear structural evidence to support the previous observation of cross-immunity within T6SS families and provide a basis for understanding their important roles in polymicrobial environments.


Introduction
The type VI secretion system (T6SS) is a novel multi-subunit needle-like apparatus and plays an important role in many processes of bacterial life cycles, such as interspecies competition, biofilm formation and virulence-related processes [1]. The Gramnegative bacteria harboring T6SS inject the effectors into their recipient's cytoplasm or periplasm to kill them. Meanwhile, to protect itself from accidental injury, the cognate immunity proteins were produced to protect the donor cells against the toxic effectors [2], [3]. Therefore, they can inhibit the growth of competitor cells without causing accidental injury to themselves and provide fitness advantages in the niche competition. Four broadly distributed and phylogenetically distinct families of T6SS peptidoglycan (PG) amidase effectors-immunity (EI) pairs have been recently identified based on overall primary sequence homology and different substrate specificities [4]. Tae4 (type VI amidase effector 4) and Tai4 (type VI amidase immunity 4) are T6SS effector-immunity pairs from the fourth family.
Our group has recently solved the crystal structures of Stand EcTae4-Tai4 complexes from the human pathogens Salmonella typhimurium and Enterobacter cloacae, respectively [5]. Structure-based mutational analysis of the EcTae4-Tai4 interface shows that a helix of one subunit in dimeric Tai4 plays a major role in binding Tae4, while a protruding loop in the other subunit is mainly responsible for inhibiting Tae4 activity. The inhibition process requires collaboration between the Tai4 dimer, distinctly different from that of Tse1 inhibiting by Tsi1 from the pathogen Pseudomonas aeruginosa (Tse1 and Tsi1 were recently renamed as Tae1 and Tai1, respectively) [4], [6], [7], [8]. Since Stand EcTae4-Tai4 complexes have similar structures, it is very interesting to discover whether there is cross-neutralization between the two immunity proteins within this family. Moreover, it is very important to find whether the mechanism of the effector inhibition and the key residues are conserved in the crossneutralization process. To this end, we determined the highresolution crystal structure of the effector StTae4 from S. typhimurium in complex with the immunity protein EcTai4 from E. cloacae and studied the cross-neutralization of EcTai4 toward various effectors from the four families by in vitro and in vivo interactions. Our study has provided clear structural evidence to support the previous observation that there is cross-immunity within effector families of bacteria T6SS and provided a structural basis for understanding the essential roles of cross-immunity in polymicrobial environments.

Overall Structure of StTae4-EcTai4 Complex
Our initial attempts at solving the crystal structure of StTae4-EcTai4 complex using the molecular replacement method with the known StTae4 or EcTai4 structures as the searching model have not been successful. Then we solved the structure by the singlewavelength anomalous dispersion (SAD) method using Se-Metlabeled protein and refined it to a final R/R free factor of 0.21/ 0.26 at 2.50 Å . The complex belonged to the P2 1 2 1 2 1 space group while the Ecand StTae4-Tai4 belonged to the C121 and P6 1 22 space group, respectively. There are sixteen molecules in the asymmetric unit of StTae4-EcTai4 complex ( Figures 1A and S1), while there are four and two molecules in that of Ecand StTae4-Tai4 (PDB code 4HFF and 4HFK), respectively. However, the retention volume of purified StTae4-EcTai4 complex eluted from analytical size exclusion chromatography (Superdex 200) corresponded to a molecular mass of ,57 kDa ( Figure S2), which is much smaller than that of the total sixteen molecules above. To solve this contradiction, the small-angle X-ray scattering (SAXS) study, was applied to study the solution structure of the complex. As shown in Figure 1B, the fit of the theoretical curve of one tetramer crystal structure to the experimental data is very good fit with a discrepancy value of 1.625. The results indicate that the active complex is a heterotetramer in solution, consisting of a Tai4 homodimer [named subunit I (cyan) and subunit II (orange)] binding two Tae4 molecules ( Figure 1A), consistent with the status of the Ecand StTae4-Tai4 complexes in solution. As previously demonstrated, molecular dimerization of EcTai4 is required in recognizing and binding EcTae4 [5], and the dimerization can also be observed in the present complex. The four heterotetramers in the asymmetric unit are likely the result of crystal packing ( Figure S1), and represent no significant biological relevance. Since the four heterotetramers are essentially identical, we hereafter confine our analyses and discussions to one StTae4-EcTai4 heterotetramer (composed of chains A and C for StTae4 and chains B and D for EcTai4).
Structural Comparisons of StTae4/EcTai4 with their Respective Status in St-or EcTae4-Tai4 Complexes The overall structure of StTae4, especially the N-terminal (Nt) subdomain in the complex, is basically identical to that in StTae4-Tai4 (PDB code 4HFF) and the catalytic triad Cys44-His126-Asp137 adopts very similar conformations in both structures (Figures 2A and S3). Structural alignments of StTae4 with the former structure only give an RMSD of 0.415. However, there are clear differences in the C-terminal (Ct) subdomain. The most striking and divergent region is the disordered winding loop, which can be characterized as two clips (Figure 2A). Clip I, composed of the residues from Gly132 to Leu142, adopts similar structure to that in the former StTae4, but a remarkable shift (,5.4 Å ) in the terminus occurs. It is worth to note that the conserved residues Cys135 and Cys139 form a disulfide bond (DSB) to stabilize the flexible loop, while there is no DSB between the two residues in the loop of the former one, where the two sulfydryl groups are in opposite directions ( Figure 2A). As previously reported, the DSB is also observed in EcTae4 and Tae1, which provides their structural stability for substrate recognition and is closely associated with their PG amidase activities [4], [5]. The residues from Asn143 to Val151 form clip II, folding over the catalytic region. The clip II portion in StTae4 is also remarkably different (a ,4.8 Å shift) from that in the former one. The conformational flexibility of this clip has been proved to significantly affect the enzyme activity of EcTae4 [4]. The residues Leu142-Gln148 in Clip II of the winding loop observed in the present structure were missed in the former Tae4 of StTae4-Tai4. Moreover, the overall conformation of the Clip II in the present structure is different from that in the former one from StTae4-Tai4, which may be adaptive to the binding and inhibition of EcTai4. This indicates remarkable changes in the winding loop of StTae4 will occur when recognized and inhibited by EcTai4 or other immunity proteins during the cross-immunity process. On the other hand, there are seldom changes in the closed lid loop covering the active pocket of Tae4 (Figure 2A), indicating the role of the loop in the inhibition process of EcTai4 against the active site of various effectors is similar.
The structure of EcTai4 in the present complex is also basically identical to that in EcTae4-Tai4 (PDB code 4HFK, Figure 2B). Structural alignments of EcTai4 with the former one give an RMSD of 0.468 Å . However, the dominant feature of the protruding loop in Tai4 is a shift by ,3.0 Å compared with the former EcTae4 ( Figure 2B), which may be adaptive for inserting into the active site of different effectors within this family. It is worth to note that the conserved residues Cys41 and Cys101, located in a2 and a5 respectively, form a DSB to stabilize the super helical conformation during the inhibition process, where there is no DSB formed between the two residues in the former EcTai4.

Interaction of StTae4 with Dimeric EcTai4
There is intimate association between StTae4 and EcTai4 homodimer in the interface ( Figure 3 and Table S1). The total buried surface area in the interface of EcTai4 dimer with one StTae4 monomer is 895 Å 2 , with 641 Å 2 contributed by two helices (a3 and a5) from StTae4 with two helices (a3 and a4) from EcTai4 subunit I, and 254 Å 2 contributed by the lid loop from StTae4 interacting with the protruding loop from the neighboring EcTai4 subunit II. A closer inspection of electrostatic potential mapped onto the molecular surfaces of StTae4 and EcTai4 dimer, reveals a perfect surfaces complementary in both shape and electric charge ( Figure 3, Middle), also suggesting there are extensive interactions between them. Notably, the highly conserved residues from Tyr78 to Asn81 in the loop b2-a5 and the helix a5 of StTae4 directly interact with Ala70, Leu63, Glu64 and Leu68 of subunit I from EcTai4 through a series of hydrogen bonds; the residues Lys33 in the helix a3 of StTae4 directly interacts with Glu74 of subunit I from EcTai4 through a salt bond (Figure 3, Right). On the other hand, the conserved residues Gly89, Thr91 and Tyr96 in the protruding loop of subunit II from EcTai4 forms direct interactions with the Ser121 and Asn122 in the lid loop of StTae4 through hydrogen bonds (Figure 3, Left).

Interaction Study between EcTai4 and different Effectors
The in vitro interactions of EcTai4 with the effectors PaTae1 (from P. aeruginosa), TyTae2 (from Salmonella Typhi), RpTae3 (from Ralstonia pickettii) or StTae4 from different families of T6SS were studied by pull-down assays ( Figure 4A). Our results showed that His-EcTai4 can pull down StTae4,while His-StTai4 can also pull down EcTae4 although we were unable to get the structure of the complex. However, His-EcTai4 can not pull down PaTae1, TyTae2 or RpTae3. The results indicate the immunity proteins may directly interact with their cognate effectors with the members of the fourth family, while there is no interaction between Tai4 and the effectors from other three families.
Co-expression of EcTai4 with the effectors above in periplasmic space of E. coli was further applied to test its neutralization capacity by observing the growth of E. coli ( Figure 4B). The results showed Ecand St-Tai4 can provide protection for the viability of E. coli harboring Stand Ec-Tae4, respectively, although their detoxification capacity is less lower compared with the native Tae4-Tai4 pairs. Meanwhile, Ec-Tai4 can not protect against the effectors from the other three families. This indicated both Ecand St-Tai4 can efficiently rescue the cells from the toxicity of their cognate effectors within this family, but can not neutralize the toxic activity of the other family effectors.

Conserved Inhibition Mechanism between Ec-and StTai4
Although there are remarkable changes in the key regions of both StTae4 and EcTai4, comparisons of the key residues in EcTai4 responsible for recognition and inhibition of StTae4 with those in EcTai4-Tai4 showed they are conserved and very similar. These indicate the neutralization process of Tae4 among different species within the effector families of T6SS is similar, but the conformations of the flexible loops, which are associated with their enzyme activities, may vary as a result of the inhibition of different Tai4 from different species of this family. Moreover, more residues in StTae4 are directly involved in binding and inhibiting StTai4 compared with those in EcTai4 of the present complex. For example, the conserved residue Ser121 (or Arg124) in the lid loop of StTae4 not only directly interacts Ala29 and Thr31 (or Tyr72) in the subunit I from StTai4, but also with Asn96 in the protruding loop of subunit II in StTae4-Tai4 complex. In the former EcTae4-Tai4 complex, the conserved residues E63A and E64A variants of EcTai4 cause a ,140-and ,10-fold reduction in affinity, respectively, indicating that these two residues are important for EcTae4 binding [5]. In the present complex, Leu63 directly  Cross-Immunity within Bacterial T6SS EI Families PLOS ONE | www.plosone.org interacts with Arg79, while Leu64 with Val80 and Asn81 of StTae4, respectively (Figure 3, Right). We can reasonably speculate these two residues play similar role in the interaction with StTae4. Besides, there is a direct interaction between Gly90 (in the protruding loop of EcTai4) and Ser151 (in the winding loop of EcTae4) in the former EcTae4-Tai4, while no direct interaction can be observed between these two loops in the present complex.
In the former StTae4-Tai4 complex, there is a direct interaction by a hydrogen bond formed by Ser98 in the protruding loop of Tai4 and the catalytically important His126 in Tae4 via a water molecule ( Figure S4). In the present structure and EcTae4-Tai4 complex, the tip of the protruding loop is also located near the catalytic triad. Moreover, although there is no direct interaction between the protruding loop of EcTai4 and the catalytic triad of Ecor StTae4, the variant D86-91 (the deletion from Gln86 to Thr91) in the protruding loop of EcTai4 has proved to be not capable of inhibiting amidase activity, but still can bind to EcTae4 [5]. In this complex, the residues Gly89, Thr91 and Tyr96 in the protruding loop of EcTai4 directly interact with Ser121 and Asn122 of StTae4 through a series of hydrogen bonds (Figure 3, Left, and Table S1), in order to tether EcTai4 to the active site of StTae4, in a similar mode to that in EcTai4-Tae4 complex. These results suggest the role of the protruding loop in inhibiting Tae4 is very similar among the family members.

Structural Implications for the Cross-immunity of T6SS Effector Families
Not surprisingly, both Ecand St-Tai4 can provide efficient protection for their cognate effectors Stand Ec-Tae4, but can not for the effectors from the other families. Similar observation has been reported that PaTai1 can rescue cells from the toxicity of both PaTae1 and the cognate BpTae1 (from Burkholderia phytofirmans), but can not neutralize the toxic activity of the effectors from other families [4]. These results showed all the species may avoid being killed from the toxicity of different effectors in the same family as a result of the intra cross-immunity. These interesting findings suggest a new strategy to develop anti-pathogen drugs. We can rationally design some small molecule inhibitors or small peptides that bind to the immunities, which could serve as a treatment against multiple pathogens within one effector families. However, an exception is that there no cross-immunity between TyTai2 and BtTai2 (from Burkholderia thailandensis) [4].
In this study, both the amino acid sequences (with 53.4% identity, Figure 5A) and the structures (with a RMSD value of 0.574 Å ) of Ecand StTae4 are very similar. More importantly, the structures of Ecand StTai4 are also similar with a RMSD value of 1.475 Å , although they share only 18.1% amino acid identity ( Figure 5B). This structure-determined inhibition mechanism within this effectors family may be very conserved. Meanwhile, there is very low amino acid identity between Tae4 (or Tai4) and the other three families (Figure S5), and their structural arrangements are distinct. Tai1 and Tai4 are all-strand and all- helical protein [5], [6], [7], respectively, and Tai2 and Tai3 display a+b folds with different structural arrangements based on secondary structure prediction [9].The varied structures of the immunity proteins from different effector families are most likely to cause distinct inhibition mechanism of the effectors. Therefore, Tai4 can not provide efficient protection for the effectors from the other families. The structure of RpTai3 revealed its dimerization manner is different from its homolog BtUCP from Bacteroides thetaiotaomicron, and the residues in the dimer interface are not conserved [9]. This indicated there may be different inhibition mechanisms within this effector family. Moreover, TyTai3 can neutralize the toxic activities of both TyTae3 and BtTae2, while BtTai2 can only provide protection for BtTae2 [4]. This indicates the cross-immunity is selective and the inhibition mechanisms of some members in these two families may be similar, as both of them have a+b structures. The variation of neutralization patterns shows the functional diversity of EI pairs.

Cross-immunity of T6SS Effector Families in the Environment
Most of T6SS EI pairs are discovered in pathogens that colonize polymicrobial sites in the host and natural environment, such as the gastrointestinal tract (GI tract), and the soil [10], [11]. This suggests they are closely associated with interbacterial interactions in the formation of environmentally and clinically relevant microbial communities. Under these conditions, the cross-immunity against multiple effectors may promote the cooperation of some species and play an important role in interbacterial competition [10]. Moreover, Russell et al found 27% of the immunity proteins they identified were not encoded adjacent to intact effector genes, while the effector genes always co-occur with immunity genes [4]. This indicates there is a selective pressure to retain immunity even in the absence of cognate effectors for the antagonistic interspecies competition.
In this study, both S. Typhimurium and E. cloacae harboring the Tae4-Tai4 pair are common pathogens causing gastroenteropathy and they inhabit the polymicrobial environments in GI tract during most of their life cycle [12], [13]. Therefore, for competitions for survival in these environments by T6SS, their immunity proteins may be required to be retained as a result of a selective pressure. In the view of structural aspect combined with the interaction studies, our study proved Ecand StTai4 can provide the interactive immunity for both Ecand StTae4 and they can cooperate in the interbacterial competition, supporting the idea the necessity of retaining the immunity.

Conclusion
The molecular insights into the StTae4-EcTai4 complex structure and function garnered from this study shed light on the mechanisms of cross-immunity within effector families of T6SS. This work provides novel structural insights into the EI interaction, which is critical for accurately understanding the interplay between effector and immunity sequence variation, and assisting in our comprehension of bacterial interaction networks and community structure. Moreover, these findings provide valuable information for development of novel antibacterial agent that can control multiple pathogens.

Cloning, Expression, Purification and Crystallization
The genes encoding full-length StTae4 and truncated EcTai4 (residues 19-117 without the N-terminal 18-residue signal peptide) were amplified from the S. Typhimurium and E. cloacae genomic DNA, respectively. The digested PCR products of StTae4 was cloned into the NcoI and XhoI sites of pET28at-plus (introducing an N-terminal TEV cleavage site, constructed by our lab), while the digested PCR products of EcTai4 was cloned into the NdeI and XhoI sites of pET21a (Novagen, USA) with a C-terminal His tag. The two recombinant plasmids were co-transformed into BL21 (DE3) cells for co-expression. Recombinant proteins were purified as previously described [14].
The StTae4-EcTai4 complex was concentrated to ,15 mg/ml using Millipore Amicon Ultra 10 KD. Crystallization screens were performed with Hampton Research and QIAGEN kits using sitting-drop vapour-diffusion method at 293K. The SeMet complex crystal was obtained in the mixture solution containing 20% (w/v) PEG 3350 and 0.2 M Magnesium formate after 3 weeks.

Data Collection, Structure Determination and Refinement
The diffraction data from a single crystal were collected on the beamline station BL17U1 of SSRF (Shanghai Synchrotron Radiation Facility) using an ADSC Q315r detector at a wavelength of 0.9792 Å . The total oscillation was 360u with 1u per image and the exposure time was 1 s per image. Before data collection, crystals were soaked for 5 s in a cryoprotectant consisting of 20% (v/v) glycerol in the crystal mother liquid and then flash-cooled in liquid nitrogen. The temperature was held at 100 K in cold nitrogen gas stream during data collection. The data were processed by HKL2000 [15]. At the first time, the space group was set to P422 with unit cell parameters a = b = 89.36Å , c = 272.72 Å , however, the phases were too poor to get an interpretable electron density, so the space group was changed to the present one. The Se atoms were located by the program Shelxd [16], and then used to calculate the initial phases in Shelxe. The phases from Shelxe were improved in Resolve [17], and then used in Buccaneer for model building [18]. Coot and Phenix.refine were used for manually building and refinement, respectively [19], [20]. All the structures were validated by Molprobity [21]. Refinement statistics and model parameters were given in Table 1. The program PyMOL (http://www.pymol.sourceforge. net/) was used to prepare structural figures.

Small-angle X-ray Scattering and Low Resolution Model Building
SAXS data were collected on the beamline station 1W2A in BSRF using a MARCCD165 detector. The scattering was recorded in the range of the momentum transfer 0.023,s,0.22 Å 21 , in which s = (4psinh)/l, 2h represents the scattering angle, and the X-ray wavelength l is 1.54 Å . The measurements were performed in a cuvette (100 ml) with exposure time of 100 seconds to diminish the parasitic scattering.
The PRIMUS program was used to process the scattering curves [18]. The sample was measured at the concentrations of 1, 3 and 5 mg/ml to exclude concentration dependence. The distance distribution functions p (r) was computed with experimental data by the program GNOM [23]. The theoretical curves were calculated by the program CRYSOL [24]. The program GASBOR was used to build the ab initio low-resolution shapes of the complex in solution [25]. The protein structure is represented by an ensemble of dummy residues.

Protein Pull-down Assay
The genes encoding PaTae1, TyTae2, RpTae3 and StTae4 were amplified from the genomic DNA of P. aeruginosa, S. Typhi, R.
pickettii and S. Typhimurium, respectively. The digested PCR products were cloned into the NcoI and XhoI sites of pET28atplus. These recombinant plasmids were transformed in to E. coli strain for expression, respectively. His-tag EcTai4 and different effector proteins were treated with Ni beads at 277 K for 15 min. Subsequently, the native effector proteins were loaded into the beads, respectively. After extensive washing with 20 mM imidazole, the proteins were eluted with 250 mM imidazole and analyzed by SDS-PAGE and Coomassie blue staining.

Cell Viability Assay
PaTae1, TyTae2, RpTae3 and StTae4 were subcloned into the vector pET22b, while EcTai4 was subcloned into the vector pET26b. The EcTai4-pET26b was co-transformed into BL21 (DE3) cells with plasmids containing the effectors for co-expression in the periplasmic space. A single colony harboring the expressing plasmid was grown in LB media at 310K. After overnight culture, the cells were serially diluted in 10-fold steps and plated onto the LB agar supplemented with antibiotic and IPTG. The plates were prepared for pictures after an additional 20 h growth at 310 K. The strains harboring StTae4-Tai4 or EcTae4-Tai4 were prepared as control.

Protein Data Bank Accession Code
The atomic coordinates and structure factor files of StTae4-EcTai4 complex have been deposited into the RCSB PDB with the code 4JUR. Figure S1 Four heterotetramers in the asymmetric unit of StTae4-EcTai4 complex crystal. The colors are shown as in Figure 1A.