Extended Data Fig. 1. Further analyses of TseH killing, delivery, and structure.

a. Relative survival (killing by V. cholerae V52 TseH^WT / TseH^H64A) of E. coli, A. hydrophila, Edwardsiella sp., A. baylyi ADP1 (ΔT6SS), or V. parahaemolyticus RIMD2210633 strains. Welch’s one-way ANOVA with Dunnett’s multiple comparisons test comparing each sample to E. coli MG1655; *, p < 0.05; **, p < 0.01; ns, not significant. b. Relative survival (killing by TseH^WT strain relative to killing by TseH^H64A strain) of A. dhakensis expressing TsiH or vector only control. Unpaired 2-tailed t-test; ***, p < 0.001. c. Survival of A. dhakensis after killing by TseH^H64A strain expressing TseH or vector only. Vector control data resembles no-plasmid data shown in Fig. 1a and was only repeated in duplicate. Unpaired 2-tailed t-test with Welch’s correction; **, p < 0.01. d. Extension of Figure 1b in the main text. Survival of A. dhakensis prey after killing by ΔPAAR2 strain expressing plasmid-borne TseH and full-length or C-terminal truncations of PAAR2. For PAAR2: V, vector; numbers, PAAR2 truncated to indicated amino acid length; Δin, codons 146-156 removed leaving downstream residues intact. One-way ANOVA with Dunnett’s multiple comparisons test comparing to the sample expressing TseH and Vector control of PAAR2; ***, p < 0.001; ns, not significant. For all graphs, the mean and standard deviation are shown. Dots show individual replicates. e. Western blot showing whole cell lysate or eluted fraction after His-tag pull-down. BL21 DE3 E. coli expressed Flag-tagged TseH in combination with either His-tagged PAAR2 or Spy as a non-binding control. Number at left indicated ladder band size in kDa. Data is representative of two independent replicates. f. Circular permutation of catalytic residues causes swapping of N- and C-terminal lobes in the structure of TseH compared to Tse1 (PDB: 4EOB¹⁸), the other T6SS effector that belongs to NlpC/P60 family has reversed N- and C-lobe structure in comparison to TseH. g. Restricted access to the catalytic cysteine in certain peptidoglycan endopeptidases. The surface representations of SaCwlT and TseH show that the conformational changes in residues surrounding the active site are necessary for the substrate binding^21,22. For all graphs, the mean and standard deviation are shown. Dots show individual replicates. TseH^WT, V. cholerae V52 with all anti-bacterial effectors inactivated except TseH; TseH^H64A, V52 with all anti-bacterial effectors inactivated including TseH.