Delivery determinants of an Acinetobacter baumannii type VI secretion system bifunctional peptidoglycan hydrolase

Abstract

Acinetobacter baumannii is a Gram-negative opportunistic pathogen and is a common cause of nosocomial infections. The increasing development of antibiotic resistance in this organism is a global health concern. The A. baumannii clinical isolate AB307-0294 produces a type VI secretion system (T6SS) that delivers three antibacterial effector proteins that give this strain a competitive advantage against other bacteria in polymicrobial environments. Each effector, Tse15, Tde16, and Tae17, is delivered via a non-covalent interaction with a specific T6SS VgrG protein (VgrG15, VgrG16, and VgrG17, respectively). Here we define the regions of interaction between Tae17 and its cognate delivery protein VgrG17 and identify that amino acids G1069 and W1075 in VgrG17 are essential for Tae17 delivery via the T6SS, the first time such specific delivery determinants of T6SS cargo effectors have been defined. Furthermore, we determine that the Tae17 effector is a multidomain, bifunctional, peptidoglycan-degrading enzyme that has both amidase activity, which targets the sugar-peptide bonds, and lytic transglycosylase activity, which targets the peptidoglycan sugar backbone. Moreover, we show that the Tae17 transglycosylase activity is more important than amidase activity for the killing of Escherichia coli. This study provides molecular insight into how the T6SS allows A. baumannii strains to gain dominance in polymicrobial communities and thus improve their chances of survival and transmission.

IMPORTANCE

We have shown that the Acinetobacter baumannii T6SS effector Tae17 is a modular, bifunctional, peptidoglycan-degrading enzyme that has both lytic transglycosylase and amidase activities. Both activities contribute to the ability to degrade peptidoglycan, but the transglycosylase activity was more important for the killing of Escherichia coli. We have defined the specific regions of Tae17 and its cognate delivery protein VgrG17 that are necessary for the non-covalent interactions and, for the first time, identified specific amino acids essential for T6SS cargo effector delivery. This work contributes to our molecular understanding of bacterial competition strategies in polymicrobial environments and may provide a window to design new therapeutic approaches for combating infection by A. baumannii.

INTRODUCTION

Acinetobacter baumannii is a Gram-negative bacterium that withstands hospital disinfection procedures and is a common cause of hospital-acquired infections. A. baumannii has increasing resistance to antibiotics and is listed as one of the three pathogenic species where the need for research and development of new antibiotics is “critical” (1–3). Many A. baumannii genomes encode an active type VI secretion system (T6SS), which plays an important role in interbacterial competition and enhanced survival in polymicrobial environments (4–6).

The T6SS is weaponry used by many Gram-negative bacterial species, especially those classified within the α-, β-, and γ-proteobacteria classes (7). This molecular “nanomachine,” which resembles the inverted tail of a bacteriophage (8), is anchored in the cell membrane and delivers effector proteins directly into target cells or the extracellular environment. Some T6SS effectors can target eukaryotic cells and therefore play roles in virulence, but many are antibacterial toxins that mediate interbacterial competition. The apparatus comprises a membrane complex, which anchors the T6SS in the bacterial cell (9), an outer sheath, composed of proteins TssB and TssC, and an inner needle composed of repeating Hcp hexamers (10) topped by a spike consisting of a trimer of VgrG proteins (11) and a single PAAR protein (12). Contraction of the outer sheath propels the inner needle proteins and attached effectors outside of the cell.

T6SS effectors can be translationally fused to T6SS needle proteins (specialized effectors) but most are cargo effectors that each require a specific non-covalent interaction with a T6SS protein for delivery (13). Often these interacting proteins are VgrG proteins that form the T6SS needle tip. The interactions required for the delivery of cargo effectors are poorly understood. VgrG proteins have a highly conserved N-terminal VgrG domain and a variable C-terminal region. The N-terminal VgrG domain is crucial for T6SS spike-formation (11, 14) while the C-terminal region is proposed to interact with the cognate effector (15–17). However, there is currently no information on the precise residues critical for the delivery of these cargo effectors.

Since T6SS-mediated killing is contact dependent (18), this form of interbacterial competition can become self-limiting if killing does not lead to cell lysis and there is the formation of “corpse barriers” that provide steric protection from T6SS-mediated attack (19). To ensure prey cell lysis occurs, many bacterial species encode at least one lytic T6SS effector that targets bacterial peptidoglycan, an essential component of the bacterial cell wall. These T6SS effectors usually cleave either the glycan linkages between N-acetylmuramic acid (MurNAc) and N-acetylglucosamine (GlcNAc) (20) or the cross-linking peptide stems of peptidoglycan (4). Currently, only one peptidoglycan hydrolase effector has been described that has two separate catalytic domains and targets both the glycan strands and the interpeptide bonds (21).

The A. baumannii strain AB307-0294 (22) encodes a single, constitutively active T6SS that delivers three cargo effectors, Tse15, Tde16, and Tae17. Each effector can act alone and kill competing bacteria, including Escherichia coli and Acinetobacter baylyi (6). The effector Tae17 is a predicted peptidoglycan hydrolase that is proposed to be delivered via non-covalent interactions with its cognate VgrG protein, VgrG17 (6). This study aimed to functionally characterize Tae17 and define the specific interactions required for its delivery by the T6SS. Here we show that the C-terminal region of VgrG17 interacts with the N-terminal region of the Tae17 effector and for the first time identify specific amino acids within the C-terminal region of VgrG17 that are essential for this interaction. Furthermore, we show that Tae17 is a multi-domain, bifunctional, peptidoglycan-degrading effector containing both lytic transglycosylase and amidase activities that are both required for maximal toxicity.

RESULTS

Tae17 is a modular, multi-domain effector with both lytic transglycosylase and amidase activity

Previous bioinformatic analysis of the wild-type Tae17 effector identified a predicted peptidoglycan binding domain, LysM, between residues 112 and 155, suggesting that Tae17 acts as a peptidoglycan hydrolase (6, 23). Tae17 showed limited amino acid sequence identity to proteins outside the Acinetobacter genus (best hit 35% identity with 62% coverage to a LysM domain-containing protein from Psychrobacter maritimus) and no significant shared amino acid sequence identity to any functionally characterized proteins. To predict other potential domains and active site residues, we produced a molecular model of wild-type Tae17 using the AI system AlphaFold2 (Deepmind). The model predicted with high confidence that Tae17 contained four domains (residues 1–88, 110–156, 234–446, and 469–582) each joined by a poorly modeled linker region (Fig. 1A and B). The N-terminal domain was predicted to form an Ig-like fold (Fig. 1C), which is found in a diverse set of proteins and often facilitates protein-protein interactions (12, 24). The second domain (residues 110–156), predicted to be a LysM domain, aligned well with characterized LysM structures (25). Using the DALI server for comparison to known structures (26), the third domain (residues 234–446) was shown to share structural homology with a domain in the lytic transglycosylase Cj0843c from Campylobacter jejuni (PDB ID: 7LAM, overall RMSD: 20.8 Å, active site RMSD: 11.0 Å) (27). While the overall alignment between the third domain of Tae17 and Cj0843c was poor, likely due to the difference in the length of the proteins, the alignment of the active site could be visualized. The key active site residues for Cj0843c are R388, E390, and K505, which aligned closely with Tae17 residues H276, E278, and K436 respectively (Fig. 1D). The fourth domain of Tae17 (residues 469–582) aligned with the amidase domain of the T6SS effector Tae3 from Ralstonia pickettii (PDB ID: 4HZ9, RMSD: 2.2 Å) (28). The catalytic dyad of Tae3 is known to include residues C23 and H81, closely aligning with Tae17 residues C488 and H550, respectively (Fig. 1E). These residues are consistent with the position of the active site residues in other proteins belonging to the CHAP amidase protein family (29). Together these data suggest that Tae17 contains a lytic transglycosylase and an amidase domain, both with distinct active sites.

FIG 1 The Tae17 AlphaFold2 model has a four-domain structure with two likely active sites. (A) Tae17 AlphaFold2 model colored according to confidence. (B) Tae17 sequence mapped by structural domains. Each domain is indicated in a different color and residue numbers are shown. (C) Tae17 AlphaFold2 structure mapped by domains, colors are as used in panel B. (D) Alignment of the C. jejuni lytic transglycosylase 7LAM (green) and the proposed Tae17 lytic transglycosylase domain (light blue, residues 234–446). Active site resides are indicated by pink boxes: E278 (E390 in 7LAM), H276 (R388 in 7LAM; labeled +), and K436 (K505 for 7LAM). (E) Alignment of the proposed Tae17 amidase domain (dark blue, residues 469–582) to the full-length Ralstonia pickettii T6SS effector Tae3 (orange, PDB ID: 4HZ9). Active site residues are indicated by a pink box: C23 and H81 for Tae3 and the equivalent C488 and H550 for Tae17. For panel E, 4HZ9 chains B and C are removed as these are from the immunity protein Tai3.

To experimentally confirm that Tae17 had both lytic transglycosylase and amidase activity, we utilized the glucose-repressible and arabinose-inducible vector pBAD30 (30) to separately express wild-type and mutated Tae17 proteins in E. coli to assess their toxicity. As Tae17 is only toxic to E. coli when targeted to the periplasm (6), a sequence encoding a PelB leader was added to the 5′ end of each recombinant tae17 gene. Expression plasmids encoding wild-type Tae17 (Tae17^WT), the lytic transglycosylase domain mutant Tae17^E278A, the amidase domain mutant Tae17^C488A, or a double mutant Tae17^E278A,C488A were separately expressed in E. coli from plasmids pAL1857, pAL1871, pAL1854, and pAL1870, respectively (Table 2).

To quantify the ability of each E. coli strain to survive heterologous expression of wild-type Tae17 or one of the Tae17 catalytic mutants, growth curves were performed, with or without induction of Tae17 expression. Both optical density (Fig. 2A) and cell viability (Fig. 2B) were assessed. All E. coli strains (expressing either recombinant Tae17^WT, Tae17^E278A, Tae17^C488A, or Tae17^E278A,C488A) grew normally when glucose was included in the medium to repress recombinant protein expression. E. coli cells expressing Tae17^E278A,C488A were as viable as the uninduced controls and grew at the same rate, showing that mutation of single active site residues in both predicted catalytic domains completely inactivated Tae17. Upon induction of protein expression, E. coli expressing Tae17^WT or the amidase mutant Tae17^C488A, which has the transglycosylase domain as the only remaining functional catalytic domain, generated similar, low viability counts 5 h after induction (approximately 10⁵-fold lower than the corresponding uninduced strains). Thus, Tae17 with only an active transglycosylase domain can kill E. coli at levels similar to the wild-type protein. An intermediate phenotype was observed when the expression of the lytic transglycosylase mutant Tae17^E278A,, which has the amidase domain as the only remaining functional catalytic domain, was induced in E. coli. This strain showed reduced growth and viability when grown in the presence of arabinose but was not as attenuated for growth and viability as the strains expressing Tae17^WT or Tae17^C488A (Fig. 2A and B). Thus, under the conditions tested, both domains have lethal activity against E. coli; however, the predicted lytic transglycosylase domain appears to be more important than the amidase domain for killing E. coli.

FIG 2 Activity of the A. baumannii AB307-0294 T6SS predicted bifunctional effector Tae17. (A) Growth curves were generated for recombinant E. coli strains harboring a gene encoding Tae17WT, Tae17E278A, Tae17C488A, or Tae17E278A,C488A (each with an added PelB leader sequence) on the vector pBAD30 (Table 2). Expression of the recombinant proteins was either induced with arabinose (solid markers and lines) or repressed with glucose (open markers and dotted lines) and optical density was measured at 0 h, 1 h, 3 h, and 5 h. (B) Viable counts were performed on the same cultures at 0 h, 1 h, 3 h, and 5 h. On both graphs, each data point represents the mean of three biological replicates, and the vertical bars show ±SD. The statistical significance of the difference in means was analyzed by ANOVA with Tukey’s multiple comparisons test. ***P < 0.001, ****P < 0.0001. (C) Degradation of FITC-labeled peptidoglycan following incubation for 2 h with purified lysozyme (control), Tae17WT, Tae17C488A, Tae17E278A, or Tae17E278A,C488A. For purification of the Tae17 proteins, see Fig. S1. (D) An image of a Tris-Tricine gel following electrophoresis of the following samples: Gram-negative dansylated LII (DAP LII), polymerized DAP LII (Pol. DAP LII), and polymerized DAP LII following incubation for 2 h with purified Tae17WT, Tae17C488A, Tae17E278A, or Tae17E278A,C488A. (E) An image of a Tris-Tricine gel following electrophoresis of Gram-positive dansylated LII (L-Lysine LII), polymerized L-Lysine LII (Pol. L-Lysine LII), and polymerized L-Lysine LII following incubation for 2 h with purified Tae17WT, Tae17C488A, Tae17E278A, or Tae17E278A,C488A.

As most characterized T6SS peptidoglycan hydrolase effectors have only a single catalytic domain (4, 20, 28, 31), we were interested in further defining the function of these two domains in Tae17. Therefore, we purified recombinant Tae17^WT, Tae17^E278A, Tae17^C488A, and Tae17^E278A,C488A from E. coli strains harboring pAL1789, pAL1919, pAL1920, and pAL1885, respectively (Table 2), using nickel affinity-coupled size exclusion chromatography. For all constructs, the purified products resolved one predominant band at ~67 kDa and a second band at ~57 kDa on SDS-PAGE (Fig. S1). To determine the nature of the bands, we conducted peptide fingerprinting of Tae17^WT. These data showed that the high MW species was indeed full-length Tae17 with 86% coverage, whereas the lower MW species was Tae17 with decreased peptide coverage prior to residue 91 (80% overall coverage) (Fig. S1). This suggests that for a proportion of the purified protein, the Ig-like domain (residues 1–88) had been cleaved. As this N-terminal degradation is unlikely to affect the catalytic activity of Tae17, we proceeded with functional analysis.

To assess the ability of each of the purified proteins to degrade peptidoglycan, proteins were incubated with purified, intact FITC-labeled E. coli peptidoglycan, and degradation activity was measured. Incubation with Tae17^WT resulted in significant peptidoglycan degradation, while the activity of the dual catalytic mutant Tae17^E278A,C488A was undetectable. Both Tae17^C488A and Tae17^E278A showed significantly reduced but measurable peptidoglycan hydrolase activity (Fig. 2C). We then assessed their ability to degrade in vitro synthesized dansylated lipid II (LII) polymers that resembled either Gram-positive LII (L-Lys LII) or Gram-negative LII (DAP LII). Moreover, the Gram-positive representative L-Lys LII was amidated at position 2 of the peptide stem to iso-D-glutamine and had a lysine (Lys) at position 3 in the amino acid chain of peptidoglycan, which is representative of the chemical variant seen in Staphylococcus aureus, Streptococcus pneumoniae, and other Gram-positive pathogens (32). The DAP LII, on the other hand, contained a diaminopimelic acid at position 3 (33). In both LII species, the presence of dansyl at DAP or Lys prevents transpeptidase crosslinking but allows for lipid II glycosyltransferase assembly of the glycan polymer. These polymers were separately incubated with the Tae17 recombinant proteins and the breakdown products generated were visualized on a Tris-Tricine gel (34). This assay is specific for lytic transglycosylase activity. Following the incubation of DAP LII with Tae17^WT or Tae17^C488A, the DAP LII polymer was completely degraded (Fig. 2D), indicative of lytic transglycosylase activity. Incubation of DAP LII with Tae17^E278A or Tae17^E278A,C488A did not result in the breakdown of the DAP LII polymer, confirming that E278 is essential for the lytic transglycosylase activity. Both Tae17^WT and Tae17^C488A were also able to hydrolyze the glycosidic bonds of the L-Lys LII polymer, while Tae17^E278A and Tae17^E278A,C488A were not (Fig. 2E). This indicates that the lytic transglycosylase domain of Tae17 can hydrolyze the glycosidic linkage between MurNAc and GlcNAc in either the L-Lys LII or DAP LII polymers. Taken together, these data show that E278 is essential for the lytic transglycosylase activity against both Gram-positive and Gram-negative representative peptidoglycan, and while C488 is not involved in the breakdown of the glycosidic linkages, it is essential for the amidase activity.

The N-terminus of Tae17 interacts with the C-terminus of the VgrG17 T6SS tip protein for effector delivery

We previously showed that the A. baumannii strain AB307_0294 lytic cargo effector Tae17 was crucial for interbacterial killing and dependent on VgrG17 for its delivery (6). However, the specifics of the interaction between the two proteins had not been defined. We therefore aimed to identify specific residues in VgrG17 that were required for interactions with Tae17. Previous studies have indicated that the C-terminal region of VgrG proteins confers specificity for their cognate effectors. Furthermore, Ig-like (or TTR-like) domains, ranging from 8 to 62 amino acids in length, have been identified in this region and are important for these interactions (15–17).

To determine whether the C-terminal region of VgrG17 was involved in the interactions with, and T6SS-mediated delivery of, Tae17, we constructed expression plasmids (pAL1565 and pAL1620) encoding VgrG17 variants truncated at the C-terminal end by 16 amino acids (VgrG17_1–1,085) and 27 amino acids (VgrG17_1–1,074) (Fig. 3A). Each of the recombinant VgrG17 proteins was then tested for their ability to deliver Tae17 into E. coli using interbacterial killing assays. For these assays, a triple mutant predator strain AB307-0294Δtse15Δtde16ΔvgrG17 was constructed. This mutant lacks the T6SS effectors Tse15 and Tde16 which are also toxic to E. coli (6). Moreover, the mutant lacks a genomic copy of vgrG17. This predator strain can therefore only kill E. coli if a recombinant VgrG17 protein provided in trans interacts with and delivers the Tae17 effector. Interbacterial killing assays showed, as expected, that the AB307-0294Δtse15Δtde16ΔvgrG17 strain containing empty vector (EV) could not kill E. coli but providing the strain with wild-type VgrG17 in trans resulted in significant killing with E. coli viable counts generally below the level of detection (Fig. 3B). The AB307-0294Δtse15Δtde16ΔvgrG17 strain expressing VgrG17_1–1,085, lacking the C-terminal 16 amino acids, was still able to kill E. coli at similar levels as when the wild-type VgrG17 was expressed. However, the AB307-0294Δtse15Δtde16ΔvgrG17 strain expressing VgrG17_1–1,074 could not kill E. coli. Therefore, these data indicate that one or more of the VgrG17 residues between amino acids 1,075 and 1,085 are crucial for the delivery of Tae17.

FIG 3 The effect of C-terminal truncations on the ability of VgrG17 to deliver Tae17 and kill E. coli prey. (A) A schematic representation of wild-type and C-terminal truncated VgrG17 proteins. The majority of A. baumannii VgrG sequence (white region) is conserved and predicted to participate in the formation of the stalk formed by the trimerization of VgrG proteins at the T6SS tip. This region is not shown in its entirety. The C-terminal end of the protein (green) is unique to VgrG17 and specific regions within it are predicted to interact with Tae17 for its delivery. (B) The ability of C-terminal truncated VgrG17 proteins VgrG171–1,085 and VgrG171–1,074 to deliver Tae17 via the T6SS was investigated using an interbacterial killing assay. The A. baumannii predator strain AB307-0294Δtse15Δtde16ΔvgrG17 (strain AL4060) containing empty vector (EV), VgrG17, VgrG171–1,085, or VgrG171–1,074 was co-incubated for 3 h with E. coli prey strain DH10B harboring an empty pWH1266 vector (AL3340). To check the growth of the prey in the absence of a predator, the E. coli prey was also grown alone (Media Only). Each data point represents the mean of two technical replicates, the horizontal bars show the mean of the biological replicates, and the vertical bars show ±SEM. The dotted line signifies the limit of detection of the assay. The statistical significance of the difference in means was analyzed by ANOVA with Tukey’s multiple comparisons test. n = 7–8. ****P < 0.0001.

To confirm there was a direct physical interaction between VgrG17 and Tae17 and identify the precise region involved in this interaction, we used bacterial adenylate cyclase two-hybrid (B2H) assays. A set of pUT18C-based plasmids, encoding the T18 bacterial adenylate cyclase fragment fused to fragments of VgrG17, and pKT25-based plasmids, encoding the T25 fragment fused to fragments of Tae17, were constructed (Table 2) and co-introduced into E. coli to assess the interaction between each recombinant protein (Fig. S2A and B). Tae17 amino acids 1–162 (fused to the C terminal end of T25) were observed to interact with three T18-VgrG17 proteins, containing VgrG17 amino acids 1,001–1,101, 1,001–1,085, and 1,051–1,085 (Fig. S2C). These data show that the first 162 amino acids of Tae17 (containing the Ig-like domain and the LysM domain) interact with VgrG17 residues 1,051–1,085, which included the same region identified using interbacterial killing assays as important for interaction and delivery of Tae17 into E. coli.

VgrG17 residues G1069 and W1075 are important for the delivery of Tae17

The interbacterial killing assays (above) indicated that one or more amino acids between amino acids 1,075 and 1,085 in VgrG17 were required to facilitate the Tae17-mediated killing of E. coli. The importance of this region was supported by the B2H data, which indicated that the region in VgrG17 required for interaction with Tae17 was within amino acids 1,051–1,085. To determine which of the specific residues in this region are most important for this interaction and subsequent delivery of the Tae17 effector, we used alanine scanning mutagenesis. We first made a series of VgrG17 expression plasmids using the A. baumannii pBASE expression vector (Table 2), each encoding full-length VgrG17 but with specific dual amino acid substitutions between residues 1,069 and 1,086 (Fig. 4A). In total, nine proteins were separately expressed in trans in the AB307-0294Δtse15Δtde16ΔvgrG17 predator strain and each strain assessed for the ability to kill E. coli DH10B prey cells in a Tae17-dependent manner. Killing activity was initially assessed qualitatively by plating all co-incubation growth onto LB agar with tetracycline that allowed only E. coli survivors to grow. After 3 h co-incubation, only two of the nine A. baumannii predator strains failed to kill E. coli (Fig. S3), indicating that the expressed recombinant proteins did not interact with, and deliver, Tae17. These strains expressed, in trans, VgrG17^{G1069A,Y1070A}, or VgrG17^{W1075A,H1076A}, suggesting that one or both of residues G1069 and Y1070, and W1075 and H1076 were essential for delivery of Tae17.

FIG 4 VgrG17 amino acids G1069 and W1075 are important for the delivery of the effector Tae17. (A) Schematic representation of VgrG17 with the amino acid sequence in the region of interest (1,051–1,101) shown. Amino acids targeted for dual alanine substitutions are shown in underlined pairs while amino acids targeted by single alanine substitutions (G1069, Y1070, W1075, and H1076) are shown in bold. The white region of VgrG17 indicates the portion predicted to participate in the formation of the stalk formed by the trimerization of VgrG proteins at the T6SS tip, while the region in green is predicted to be the general region available for interaction with the effector Tae17. (B) An Interbacterial killing assay using E. coli DH10B harboring empty pWH1266 vector as prey, co-cultured for 3 h with the A. baumannii predator strain AB307-0294Δtse15Δtde16ΔvgrG17 (AL4060) as predator, provided with empty pBASE (EV) or a pBASE-derivative encoding wild-type VgrG17, or VgrG17 with a single amino substitution as shown. As a prey growth control, E. coli DH10B harboring empty pWH1266 was cultured alone (Media Only) (C) Western immunoblot assessing Hcp secretion (top) or intracellular amounts of Hcp (whole-cell lysates [WCL]; bottom) in the A. baumannii wild-type AB307-0294 (WT), the tssM mutant (negative control for Hcp secretion), or AB307-0294Δtse15Δtde16ΔvgrG17 strain provided with plasmids encoding wild-type VgrG17 or one of the VgrG17 alanine mutants. The secretion of the T6SS needle protein, Hcp, is used as an indicator of a functional T6SS in A. baumannii. (D) Interbacterial killing assay using E. coli DH10B harboring a plasmid encoding immunity proteins Tdi16 and Tsi15 as prey (to neutralize T6SS effectors Tse15 and Tde16). The A. baumannii predator strain in this assay was AB307-0294ΔvgrG17 (AL4329) containing EV, or a pBASE-derivative encoding wild-type VgrG17, or VgrG17 with a single amino substitution as shown. (E) Western immunoblot assessing Hcp secretion (top) or intracellular amounts of Hcp (WCL; bottom) produced by the A. baumannii wild-type AB307-0294 (WT), the tssM mutant (negative control for Hcp secretion), or the AB307-0294ΔvgrG17 predator strain provided with plasmids encoding wild type VgrG17 or one of the VgrG17 alanine mutants. For panels B and D, each data point represents the mean of two technical replicates, the horizontal bars show the mean of the biological replicates, and the vertical bars show ±SEM. The dotted line signifies the limit of detection of the assay. The statistical significance of the difference in means was analyzed by ANOVA with Tukey’s multiple comparisons test. n = 4–12. ***P = 0.0001, ****P < 0.0001. For panels C and E, images were cropped to only include the ~10 kDa section of the gel containing the Hcp protein. Supernatant samples were concentrated 10× (10×sup).

To determine which of the four VgrG17 amino acids identified in the dual alanine mutagenesis qualitative assays were important for interaction with Tae17, we generated expression plasmids (again using the A. baumannii expression plasmid pBASE) encoding the following single mutants: VgrG17^G1069A, VgrG17^Y1070A, VgrG17^W1075A, and VgrG17^H1076A (Fig. 4A). We then used quantitative interbacterial killing assays to assess which of the alanine substituted VgrG17 proteins was still able to interact with and deliver Tae17 to kill E. coli. The separate expression of VgrG17^Y1070A and VgrG17^H1076A in the AB307-0294Δtse15Δtde16ΔvgrG17 predator strain fully restored the ability of A. baumannii to kill E. coli via Tae17 activity, indicating that these residues alone are not crucial for interaction with Tae17 (Fig. 4B). However, the predator cells expressing VgrG17^G1069A were unable to kill E. coli. Thus, amino acid G1069 in VgrG17 appears essential for the interaction with Tae17. The predator strain expressing VgrG17^W1075A also showed significantly reduced killing, suggesting that amino acid W1075 in VgrG17 is also important for the interaction with Tae17.

To assess whether the single alanine substitutions in VgrG17 were specifically affecting Tae17 binding and delivery, and not abrogating the formation or activity of the T6SS apparatus itself, we used the secretion of the needle protein Hcp as a marker for T6SS assembly and activity. We assessed the ability of the AB307-0294Δtse15Δtde16ΔvgrG17 strains, each provided with a pBASE plasmid encoding a different alanine-substituted VgrG17 protein, to secrete Hcp into the supernatant (Fig. 4C). Interestingly, the AB307-0294Δtse15Δtde16ΔvgrG17 strain harboring empty pBASE vector had an inactive T6SS with no observable Hcp secretion, although Hcp was still produced. This indicates that the AB307-0294 T6SS lacking VgrG17 does not function in the absence of other functional VgrG/effector pairs; similar results have been observed previously in other species (35). However, providing the AB307-0294Δtse15Δtde16ΔvgrG17 strain with wild-type VgrG17 or any of the single VgrG17 alanine mutants in trans restored Hcp secretion (Fig. 4C). This included the AB307-0294Δtse15Δtde16ΔvgrG17 strain expressing VgrG17^G1069A, which our interbacterial killing assay showed cannot deliver Tae17. Together, the data suggest that in the absence of Tse15 and Tde16 effector proteins, a full-length VgrG17 protein is essential for T6SS function, even if the VgrG17 protein supplied to the strain, for example, VgrG17^G1069A, cannot bind and deliver the Tae17 effector. These data also show that complementation with any of the mutant VgrG17 proteins can form a fully active system, which indicates that each of these mutated VgrG17 proteins can form a functional VgrG trimer and T6SS structure.

To confirm the interbacterial killing results obtained using the AB307-0294Δtse15Δtde16ΔvgrG17 predator strain, which showed no T6SS activity in the absence of a functional VgrG17, we performed additional interbacterial killing assays using a predator strain that produced all three T6SS effectors, Tse15, Tde16, and Tae17. The AB307-0294ΔvgrG17 mutant (AL4329) can deliver the effectors Tse15 and Tde16 (via VgrG15 and VgrG16, respectively) but can only deliver Tae17 when a functional VgrG17 protein is expressed in trans from the complementing plasmid. To allow the use of this predator strain for specifically assessing Tae17 delivery, the E. coli prey strain was provided with the appropriate cognate immunity proteins to protect against the T6SS effectors Tse15 and Tde16 (Tsi15 and Tdi16, respectively). In this case, any T6SS-mediated killing of E. coli by AB307-0294ΔvgrG17 can be attributed directly to successful VgrG17-mediated delivery of Tae17, as has been previously shown (6). As expected, in these interbacterial killing assays the in trans expression of wild-type VgrG17, VgrG17^Y1070A, or VgrG17^H1076A fully restored the ability of the AB307-0294ΔvgrG17 predator strain to kill E. coli via Tae17 delivery (Fig. 4D). However, when plasmids expressing either VgrG17^G1069A or VgrG17^W1075A were provided to AB307-0294ΔvgrG17, the strains were unable to kill E. coli (Fig. 4D). These data confirm that the VgrG17 amino acids G1069 and W1075 are crucial for interaction with, and delivery of, Tae17. To check the activity of the T6SS in these strains, we measured Hcp secretion as described earlier. All AB307-0294 ΔvgrG17 strains harboring empty vector or a VgrG17 expression plasmid produced similar levels of secreted Hcp, thereby confirming the T6SS was active in this mutant with or without provision of a functional VgrG17 (Fig. 4E).

Tae17 and VgrG17 are predicted to interact via edge-to-edge contacts across two domains

The VgrG17 experimental data obtained using truncated and alanine substituted proteins and B2H data indicated that the VgrG17 region from 1,051 to 1,085 is necessary for interaction with Tae17 and that the VgrG17 amino acids G1069 and W1075 are required for this interaction. Furthermore, for Tae17, the B2H data indicated that the first 162 amino acids are important for the interaction with this region of VgrG17. To understand the likely molecular basis of these interactions, we used the AlphaFold2 multimer to model the interaction of Tae17 with this region of VgrG17.

The T6SS needle tip is composed of a trimeric VgrG protein complex, where the N-terminal and central sections of each VgrG protein interact to form a stalk that is held together by strong intermolecular interactions (13). However, existing sequence and structural data indicate that the C-terminal regions of the VgrG proteins do not interact with each other and are free to interact with the effectors (13). Therefore, to identify potential contacts and align this to our experimental data, we modeled the C-terminal amino acids 1,010–1,101 of VgrG17, which fully encompassed the region shown to be important for VgrG17:Tae17 interaction, with full-length Tae17. Initial inspection identified two regions of interacting β-strands. One of these regions includes residues 1,013–1,070 of VgrG17, which fold into two small β-sheets that together form a sandwich-like structure that interacts with a β-strand formed by residues 90–93 of Tae17 (Fig. 5). We note the structural and functional similarity of this sub-structure to that identified by Flaugnatti et al. (36), and hence, we term this the transthyretin-like (TTR) domain. One sheet of this TTR comprises two β-strands from VgrG17, while the other sheet comprises three β-strands from VgrG17 and a single β-strand from Tae17. The second of the interacting β-strand regions included residues 1,081–1,083 of VgrG17 and a six-stranded Ig-like domain formed by two β-sheets of Tae17 from residues 4–93 (Fig. 5). Overall, the C-terminal region of VgrG17 (from residues ~1,071–1,084) appears to thread through the multi-domain structure of Tae17 toward the amidase domain (Fig. S4). The C-terminal ~17 amino acids of VgrG17 (residues 1,084–1,101) are modeled as an α-helix that does not interact with Tae17, which supports our VgrG17 C-terminal truncation data that this region is not involved in the interaction (Fig. S4).

FIG 5 AlphaFold2 model of VgrG171010-1085 interacting with Tae17 with predicted hydrogen bonds. While the AlphaFold2 interaction was modeled with the full-length Tae17, only the interacting region is shown here for clarity. (A) AlphaFold2 model of VgrG171,010–1,085 interacting with Tae171–156. VgrG17 is shown in green. For Tae17, the Ig-like domain is shown in pink, the LysM domain is shown in purple, and the linker region is colored gray. Edge-to-edge contact backbone hydrogen bonds are shown in red. Residues of interest mutated during the alanine scanning mutagenesis are shown. The amidase and lytic transglycosylase domains are removed for clarity but are included in Fig. S4. (B) The same AlphaFold2 model of VgrG171,010–1,085 interacting with Tae171–156 but rotated forward 90 degrees. The VgrG17 N- and C-termini and TTR domains are indicated.

To look more specifically at how VgrG17 and Tae17 interact in the model, the coordinates were visually inspected and input into PISA (37). No disulfide or covalent bonds were present, consistent with what is known about cargo effectors. However, approximately 17 backbone hydrogen bonds between the edge-to-edge contacts were predicted (Fig. 5 in red) as well as up to 12 salt bridges (Table S1; Fig. S4 and S5; Movie S1). The large number of salt bridges predicted across the interface is due to the highly negatively charged VgrG17 C-terminus and highly positively charged Tae17 N-terminus. This suggests that electrostatics are likely important for the interaction between VgrG17 and Tae17. Notably, a number of these charged residues were mutated during the dual alanine scanning (D1073, E1078, E1080, E1081, E1083, D1084, and D1086) yet these mutations showed no change in the delivery of Tae17. Hence, Tae17 can deliver VgrG17 even when some of the interacting residues have been mutated. As charged interactions are mediated by overall electrostatic potential, individual or dual mutation of these charged residues likely does not reduce the overall electrostatic potential enough to prevent the interaction.

DISCUSSION

Numerous Gram-negative bacteria employ a T6SS to deliver antibacterial effector proteins into neighboring bacterial cells as a means of gaining dominance in polymicrobial environments. The majority of bacteria that possess a T6SS, including A. baumannii, are human, animal, or plant pathogens (38). Consequently, manipulating the T6SS could be an important novel strategy for the control of pathogens such as A. baumannii, Pseudomonas aeruginosa, Vibrio cholerae, and Serratia marcescens (39–41). Recent bioinformatic analysis of T6SS effectors in A. baumannii strains identified >30 phylogenetically distinct families (23), with at least eight of these predicted to be involved in the degradation of peptidoglycan. The amino acid sequences of these effectors are very diverse, suggesting that these different effector families will have unique mechanisms of action. This study aimed to explore the structure, function, and delivery determinants of the peptidoglycan hydrolase effector Tae17 from the A. baumannii clinical isolate AB307-0294.

AlphaFold2 modeling of Tae17 revealed a modular protein with four domains separated by linkers and suggested that Tae17 contained two catalytic domains: one with lytic transglycosylase activity and one with amidase activity. Mutation of the proposed active site residues in these domains, alone and in combination, confirmed that Tae17 is indeed a bifunctional peptidoglycan hydrolase. Tse4 from A. baumannii ATCC 17978 has also recently been identified as a bifunctional peptidoglycan hydrolase (21). Tse4 has both lytic transglycosylase and endopeptidase activity but belongs to a different effector family and shares no readily identifiable amino acid sequence identity with Tae17 (23). Previous research on amidase effectors produced by other bacterial species, including Tae3 (28), Tse1/Tae1 (42), and Tae4 (43), has demonstrated that mutation of a single catalytic residue can completely abolish effector activity. However, mutation of the predicted amidase domain alone in Tae17 did not impede the ability of the effector to kill E. coli. This suggests that under these conditions, the lytic transglycosylase domain is the dominant domain of Tae17 and that E. coli may not be the primary target of the Tae17 amidase domain.

Gram-positive and Gram-negative bacteria produce a chemically different peptidoglycan. Both contain a similar backbone of repeating MurNAc and GlcNAc sugar units, but different stem peptide chemical components are observed. We tested Tae17 activity against in vitro synthesized Gram-positive LII-type or Gram-negative LII-type peptidoglycan backbone. The Tae17 proteins with an intact lytic transglycosylase domain were able to degrade both types of in vitro synthesized peptidoglycan. Therefore, this domain can recognize and cleave the glycosidic bonds in the peptidoglycan sugar chain that is representative of both Gram-positive and Gram-negative peptidoglycan (33).

For the Tae17 effector to be delivered, it must interact with the structural components of the T6SS. Our previous study showed that Tae17 specifically interacts with VgrG17 for delivery (6). In this current study, we have shown that the N-terminal 162 amino acids of Tae17 interact directly with the C-terminal region of VgrG17. Deletions from either end of this short Tae17 fragment resulted in the loss of interaction with VgrG17, strongly suggesting that interaction required both the Ig-like domain and LysM domain of Tae17. We also identified the region of VgrG17 between residues 1,051 and 1,085 as important for interaction with Tae17, and specifically residues G1069 and W1075, as measured by interbacterial killing assays. These data support and significantly extend a number of previous studies that show the effector N-terminal region interacts with the C-terminal region of VgrG for delivery by the T6SS (15–17). Our data present, for the first time, residues essential for the interaction required for T6SS cargo effector delivery.

During our analysis of Tae17 delivery, we found that a mutant strain (AB307-0294Δtse15Δtde16ΔvgrG17) lacking VgrG17 as well as the effectors Tse15 and Tde16 had no observable Hcp secretion. Thus, the AB307-0294 T6SS was non-functional in the absence of any complete VgrG/effector pairs. These data support previous work in Agrobacterium tumefaciens, where loading of at least one cargo effector on its cognate VgrG was necessary for efficient T6SS firing (35). However, complementation of the AB307-0294Δtse15Δtde16ΔvgrG17 mutant with the VgrG17 alanine mutant VgrG17^G1069A, which our interbacterial killing assays showed cannot deliver Tae17 and therefore is unlikely to form a cognate pair, restored Hcp secretion and therefore T6SS activity. This suggests that the AB307 T6SS is active in the absence of any complete VgrG/effector pairs, so long as VgrG17 is present. We are currently working to further define the exact VgrG and effector requirements for T6SS loading and firing.

Utilizing the experimental interaction data generated in this study, we produced an AlphaFold2 model of the Tae17:VgrG17 interaction. These data predicted two edge-to-edge contact interactions between two different domains, reminiscent of a related experimentally determined structure (36). One interaction involved the Ig-like domain of Tae17, and a single short β-strand of VgrG17, and the second interaction involved the TTR domain of VgrG17 and a single β-strand of Tae17; this Tae17 β-strand is between the Ig-like domain and the LysM domain. Hydrogen bonding would anchor the length of β-strands from the opposing molecules together, and side chains form further stabilizing salt bridges. These salt bridges are mediated by the negatively charged VgrG17 C-terminal region and the positively charged Tae17 N-terminal region. The presence of two interaction sites highlights the importance of a strong interaction between the two proteins, given the mechanical force of T6SS ejection. Further interactions may occur outside these domains, including with the VgrG17 stalk, when the Tae17 effector is loaded for delivery. When we map VgrG17 residues G1069 and W1075 onto the interaction structure, G1069 is found at the end of the TTR domain before a loop region is formed that threads through Tae17. Mutation of this residue to an alanine may disrupt the VgrG17:Tae17 interaction as alanine is more rigid than glycine and may not allow for the loop region to form as readily. By contrast, W1075 is found buried within the Tae17 Ig-like domain, forming hydrophobic interactions that anchor the loop region of VgrG17 to Tae17. Mutation of the tryptophan to alanine would reduce the size of the residue and disrupt hydrophobic contacts, destabilizing the interaction between VgrG17 and the Tae17 Ig-like domain. While it is possible that the loss of these interactions could be due to changes in the folding of VgrG17, we note that the predicted interaction between the VgrG17 TTR domain and Tae17 shows strong structural similarity to that of VgrG and Tle1 from E. coli (36). In the E. coli VgrG:Tle1 interactions, one side of a β-strand-rich sandwich comprising three anti-parallel β-strands interacts with a small β-strand of the effector, providing the necessary interactions for delivery. Given that Tae17 and Tle1 are different types of cargo effectors but appear to maintain the same interactions with VgrG for delivery, this suggests that this may be a strongly conserved mechanism of cargo effector tethering to the T6SS.

In this study, we have shown that Tae17 is a highly modular protein with four domains separated by flexible linkers. This T6SS effector is bifunctional as it possesses both lytic transglycosylase and amidase activity. Delivery of Tae17 is mediated by specific interactions between the Tae17 region encompassing the Ig-like and LysM domains and the C-terminal region of VgrG17. We have precisely defined two VgrG17 amino acids, VgrG17^G1069 and VgrG17^W1075, that are crucial for interaction with Tae17 and predict that these allow VgrG17 to thread through the Tae17 structure and allow for the correct alignment of the Ig-like and TTR domains. The AlphaFold2 and PISA interaction model predicted that these residues are located between two edge-to-edge contact interactions with numerous hydrogen bonds and are further stabilized by salt bridges, which together likely anchor the two molecules with sufficient strength to overcome the shear forces associated with T6SS-mediated effector delivery through target cell membranes.

MATERIALS AND METHODS

Bacterial strains and growth conditions

Bacterial strains used in this study are listed in Table 1 and all plasmids used in this study are listed in Table 2. All plasmids and bacterial strains generated in this study will be made available on request. All strains were cultured in 2YT broth or Lysogeny broth (LB) at 37°C with shaking (200 rpm). For growth on solid media, 1% wt/vol agar was added. Where appropriate, media was supplemented with ampicillin (100 µg/mL), tetracycline (10 µg/mL), kanamycin (50 µg/mL), or carbenicillin (100 µg/mL). For induction or repression of the P_araBAD promoter in pBAD30 (29), media were supplemented with 0.4% arabinose or 0.4% glucose, respectively.

TABLE 1 Strains used in this study

Name	Description	Source
E. coli DH5α	Strain for routine cloning purposes. F-, ∅80dlacZΔM15, Δ(lacZYA-argF)U169, recA1, endA1, hsdR17(rK-, mK+), phoA, supE44, λ-, thi-1, deoR, gyrA96, relA1	(44)
E. coli DH10B	Used as prey in T6SS interbacterial killings assays. F-, mrcA, Δ(mrr-hsdRMS-mrcBC), ∅80dlacZΔM15, ΔlacX74, recA1, endA1, araD139, Δ(ara leu), 7697, galU, galK, rpsL, npuG, λ-	(44)
E. coli C41 (DE3)	Used for large-scale production of recombinant Tae17. F-, ompT hsdSB (rB- mB-) gal dcm (DE3)	Sigma Aldrich/ Merck
E. coli BTH101	Used for bacterial two-hybrid assays. F-, cya-99, araD139, galE15, glaK16, rpsL1 (Str r), hsdR2, mcrA1, mcrB1	(45)
E. coli XL1-Blue	Used for QuikChange mutagenesis of Tae17. recA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F ́ proAB lacIqZΔM15 Tn10 (TetR)]	Stratagene
AL3340	E. coli DH10B harboring empty vector pWH1266, TetR	This study
AL3648	E. coli DH10B harboring pAL1263 (encoding Tsi15 and Tdi16), TetR	(6)
A. baumannii AB307-0294	A. baumannii AB307-0294 wild type, clinical isolate	(22)
AL3927	A. baumannii AB307-0294 Δtse15Δ tde16 markerless mutant	This study
AL4020	A. baumannii AB307-0294 ΔtssM, markerless mutant. Unable to produce a functional T6SS	This study
AL4060	A. baumannii AB307-0294 Δtse15Δtde16ΔvgrG17. KanR, aph (3′) in place of vgrG17	This study
AL4329	A. baumannii AB307-0294 ΔvgrG17 markerless mutant	This study

TABLE 2 Plasmids used in this study

Name	Description	Source
pAT03	pMMB67EH encoding FLP recombinase that interacts with tandem FRT sites flanking kanamycin cassette (aph 3′) to facilitate its removal. For generation of markerless A. baumannii mutants. AmpR	(46)
pBASE	A. baumannii/E. coli shuttle and recombinant protein expression vector (via the action of Ptac promoter upstream of MCS). AmpR	(6)
pBAD30	E. coli protein expression vector. Recombinant gene expression controlled ParaBAD promoter (induced by arabinose and repressed by glucose). AmpR	(30)
pCR™-Blunt II-TOPO™	A vector used as template for the PCR amplification of the kanamycin resistance gene aph (3′). Gene used in SOE-PCR mediated mutagenesis of A. baumannii AB307-0294	Invitrogen
pKT25	B2H vector. Encoded the T25 fragment (1–224) of CyaA under control of a lac promoter. KanR	(45)
pKT25-zip	B2H positive control plasmid. KanR	(45)
pET28a	E. coli recombinant protein expression vector. KanR	EMD Biosciences
pUT18C	B2H vector. Encodes the T18 fragment (amino acids 225–399) of CyaA under the control of the lac promoter. AmpR	(45)
pUT18C-zip	B2H positive control plasmid. AmpR	(45)
pWH1266	E. coli/A. baumannii shuttle vector. Used in T6SS interbacterial killing assays to provide a counter selection for E. coli DH10B. TetR.	(47)
pAL1263	pWH1266 encoding the T6SS immunity proteins Tsi15 and Tdi16, TetR. Used in E. coli DH10B for counter selection (TetR) and to provide immunity against the A. baumannii AB307_0294 T6SS effectors Tse15 and Tde16.	(6)
pAL1446	pBASE encoding a truncated VgrG17 protein: amino acids 1–546.	(6)
pAL1447	pBASE encoding the full-length VgrG17 protein (amino acids 1–1,101)	(6)
pAL1565	pBASE encoding truncated VgrG17: amino acids 1–1,085	This study
pAL1620	pBASE encoding truncated VgrG17: amino acids 1–1,074	This study
pAL1621	B2H plasmid pUT18C encoding VgrG17 amino acids 1,001–1,101 translationally fused to C terminal end of T18	This study
pAL1623	B2H plasmid pUT18C encoding VgrG17 amino acids 673–852 translationally fused to C terminal end of T18	This study
pAL1624	B2H plasmid pUT18C encoding VgrG17 amino acids 673–1,101 translationally fused to C terminal end of T18	This study
pAL1629	B2H plasmid pKT25 encoding Tae17 amino acids 1–162 translationally fused to C terminal end of T25	This study
pAL1630	B2H plasmid pKT25 encoding Tae17 amino acids 155–485 translationally fused to C terminal end of T25	This study
pAL1689	B2H plasmid pKT25 encoding Tae17 amino acids 1–32 translationally fused to C terminal end of T25	This study
pAL1699	B2H plasmid pKT25 encoding Tae17 amino acids 33–162 translationally fused to C terminal end of T25	This study
pAL1701	B2H plasmid pKT25 encoding Tae17 amino acids 1–51 translationally fused to C terminal end of T25	This study
pAL1702	B2H plasmid pKT25 encoding Tae17 amino acids 14–63 translationally fused to C terminal end of T25	This study
pAL1703	B2H plasmid pUT18C encoding VgrG17 amino acids 1,001–1,050 translationally fused to C terminal end of T18	This study
pAL1704	B2H plasmid pUT18C encoding VgrG17 amino acids 1,051–1,085 translationally fused to C terminal end of T18	This study
pAL1705	B2H plasmid pUT18C encoding VgrG17 amino acids 1,001–1,085 translationally fused to C terminal end of T18.	This study
pAL1715	B2H plasmid pKT25 encoding Tae17 amino acids 1–63 translationally fused to C terminal end of T25	This study
pAL1716	B2H plasmid pKT25 encoding Tae17 amino acids 1–116 translationally fused to C terminal end of T25	This study
pAL1779	pBASE containing vgrG17 with a silent SpeI restriction site introduced for cloning purposes. Generated using SOE-PCR using products made from primer pairs BAP9260/BAP9262 and BAP9263/BAP9261	This study
pAL1784	pBASE encoding alanine substituted VgrG17G1069A,Y1070A	This study
pAL1785	pBASE encoding alanine substituted VgrG17S1071A,A1072G	This study
pAL1786	pBASE encoding alanine substituted VgrG17D1073A,K1074A	This study
pAL1787	pBASE encoding alanine substituted VgrG17W1075A,H1076A	This study
pAL1789	pET28a encoding a C-terminal Hisx6-tagged Tae17, codon-optimized for expression in E. coli	GenScript Biotech
pAL1793	pBASE encoding alanine substituted VgrG17K1077A,E1078A	This study
pAL1794	pBASE encoding alanine substituted VgrG17T1079A,E1080A	This study
pAL1795	pBASE encoding alanine substituted VgrG17E1081A,F1082A	This study
pAL1796	pBASE encoding alanine substituted VgrG17E1083A,D1084A	This study
pAL1797	pBASE encoding alanine substituted VgrG17F1085A,D1086A	This study
pAL1809	pBASE encoding alanine substituted VgrG17G1069A	This study
pAL1810	pBASE encoding alanine substituted VgrG17Y1070A	This study
pAL1811	pBASE encoding alanine substituted VgrG17W1075A	This study
pAL1812	pBASE encoding alanine substituted VgrG17H1076A	This study
pAL1854	pBAD30 encoding Tae17C488A with a PelB leader, codon-optimized for expression in E. coli	This study
pAL1857	pBAD30 encoding Tae17WT with a PelB leader, codon-optimized for expression in E. coli	This study
pAL1870	pBAD30 encoding Tae17E278A,C488A with a PelB leader, codon-optimized for expression in E. coli	This study
pAL1871	pBAD30 encoding Tae17E278A with a PelB leader, codon-optimized for expression in E. coli	This study
pAL1885	pET28a encoding a C-terminal Hisx6-tagged Tae17E278A,C488A, codon-optimized for expression in E. coli	GenScript Biotech
pAL1919	pET28a encoding a C-terminal Hisx6-tagged Tae17E278A, codon-optimized for expression in E. coli	GenScript Biotech
pAL1920	pET28a encoding a C-terminal Hisx6-tagged Tae17C488A, codon-optimized for expression in E. coli	GenScript Biotech

DNA manipulations

Primers used for amplification of DNA for cloning/mutagenesis and Sanger sequencing are listed in Table 3. Isolation of genomic and plasmid DNA, PCR amplification using KOD high fidelity polymerase or Taq polymerase (colony PCR), restriction enzyme digestions, ligations, Sanger sequencing, transformation of chemically competent E. coli and transformation of electrocompetent A. baumannii via electroporation were performed as described previously (6, 48). For all cloning reactions, ligated products were used to transform E. coli chemically competent cells. Transformants containing plasmids with the correct sequence were identified by colony PCR and the sequence of isolated plasmids was confirmed using Sanger sequencing. For some experiments, the constructed plasmid was used to transform A. baumannii. To generate recombinant plasmids encoding truncated VgrG17 proteins, VgrG17_1–1,085 (pAL1565) and VgrG17_1–1,074 (pAL1620, Table 2), high-fidelity PCR products were generated using primer pairs BAP8874/BAP8842 and BAP8874/BAP9017, respectively, digested with SacI and BspEI, then ligated in-frame to similarly digested pAL1446, encoding amino acids 1–546 of VgrG17.

TABLE 3 Primers used in this study

Name	Sequence (5′–3′)a	Description
UP	GTAAAACGACGGCCAGT	Universal primer. Anneals downstream of multiple cloning site in pBASE. Used for sequencing cloned DNA
BAP1230	TAATACGACTCACTATAGG	pET28a forward primer. Anneals to T7 promoter region
BAP1231	CCGCTGAGCAATAACTAGC	pET28a reverse primer. Anneals to region following T7 terminator
BAP1430	TTAGCGGATCCTACCTGACG	pBAD30 forward primer. Anneals to the promoter region upstream of the multiple cloning site
BAP1431	TTTTATCAGACCGCTTCTGC	pBAD30 reverse primer. Anneals downstream of the multiple cloning site
BAP6009	ATGCTTCCGGCTCGTATG	pUT18C forward primer. Anneals upstream of the gene encoding T18 fragment
BAP6011	GAGCGGACGTTCGAAGTT	pUT18C forward primer. Anneals upstream of the multiple cloning site
BAP6012	TGTCTGTAAGCGGATGCC	pUT18C reverse primer, which anneals downstream of the multiple cloning site
BAP7545	AGCTCTGATCAAGAGACAGGATGA	For PCR confirmation of mutants made using SOE-PCR-mediated mutagenesis. Reverse primer that anneals within kanamycin resistance cassette, approximately 12 nucleotides upstream of aph (3′).
BAP7546	AGCGTTGGCTACCCGTGATATTG	For PCR confirmation of mutants made using SOE-PCR- mediated mutagenesis. Forward primer located at the end of kanamycin resistance gene aph (3′).
BAP7648	CCGGAATTGCCAGCTGGGGC	For SOE-PCR-mediated mutagenesis. Forward primer for the amplification aph (3′) from the plasmid pCRä-Blunt II-TOPOä.
BAP7649	GCCTTCTTGACGAGTTCTTCTGAA	For SOE-PCR-mediated mutagenesis. Reverse primer for the amplification aph (3′) from the plasmid pCRä-Blunt II-TOPOä.
BAP7689	AGTGGGAGCGAGCCCATCG	For SOE-PCR-mediated mutagenesis of A. baumannii to generate AL4060. Forward primer for amplification of region downstream of vgrG17.
BAP7690	CGTTTATATATCCTTATGATTATAAAAAAATGCTT	For SOE-PCR-mediated mutagenesis of A. baumannii to generate AL4060. Reverse primer for amplification of region downstream of vgrG17 for SOE-PCR-mediated mutagenesis. 5′ end is complementary to the KanR cassette from the plasmid pCRä-Blunt II-TOPOä.
BAP7691	TCTAATTAAGGGGAGAGAAAATGAATAAGAAA	For SOE-PCR-mediated mutagenesis of A. baumannii to generate AL4060. Forward primer for amplification of region upstream of vgrG17 for SOE-PCR-mediated mutagenesis. 5′ end is complementary to the KanR cassette from the plasmid pCRä-Blunt II-TOPOä.
BAP7692	CGACACTTCAGGACATCCTGTC	For SOE-PCR-mediated mutagenesis of A. baumannii to generate AL4060. Reverse primer for amplification of region upstream of vgrG17 for SOE-PCR-mediated mutagenesis.
BAP7767	GCCACCTGACGTCTAAGAA	Sequencing/colony PCR primer. Reverse primer which anneals downstream of multiple cloning site in pBASE.
BAP7835	GGTTTGCCGTTGTTTTCTTACG	For SOE-PCR-mediated mutagenesis of A. baumannii to generate AL4060. Forward primer for confirmation of vgrG17 deletion.
BAP7836	CATTAGAGGCAGGAGGAGCC	For SOE-PCR-mediated mutagenesis of A. baumannii to generate AL4060. Reverse primer for confirmation of vgrG17 deletion.
BAP8412	CAGGTCGTAAATCACTGCATAATTCG	Sequencing/colony PCR primer. Forward primer which binds upstream of the multiple cloning region in pBASE
BAP8527	GCTGGAGAAGGTTATGGTG	Sequencing/colony PCR primer. Forward primer which binds to nucleotides 1,498–1,513 of vgrG17.
BAP8528	TTTTTAGTATCAGGAAGC	Sequencing/colony PCR primer. Reverse primer which binds to nucleotides 1,661–1,644 of vgrG17.
BAP8842	AAAAGAGCTCTTAAAAATCCTCAAACTCTTCTG	For constructing pAL1565 encoding truncated vgrG17. Reverse primer which binds to nucleotides 3,236–3,254 of vgrG17. Includes SacI site and stop codon.
BAP8874	AAAAATCCGGAACAGCATCGACCAATTGCTTACC	For constructing pAL1565 and pAL1620 encoding truncated versions of vgrG17. Forward primer which binds to nucleotides 1,299–1,321 of vgrG17. Includes BspEI site.
BAP9017	AAAAGAGCTCTTATTTATCAGCACTATATCCCAC	For constructing pAL1620 encoding truncated vgrG17. Reverse primer which binds to nucleotides 3,203–3,222 of vgrG17. Includes SacI site and a stop codon.
BAP9021	AAAACTGCAGGGAATAAGAAAGGTTTAATTCAGTTTAG	For constructing two-hybrid plasmid pAL1629. Forward primer which binds to nucleotides 4–30 of tae17. Includes PstI site.
BAP9022	AAAACCCGGGTTAACTTGATCCACCTTTGTATGG	For constructing two-hybrid plasmid pAL1629. Reverse primer which binds to nucleotides 466–486 of tae17. Includes XmaI site and a stop codon.
BAP9023	AAAACTGCAGGGCCATACAAAGGTGGATCAAGTG	For constructing two-hybrid plasmid pAL1630. Forward primer, which anneals to nucleotides 466–487 of tae17. Includes PstI site.
BAP9024	AAAACCCGGGTTATTGACTTTTTTCTTTAGCTCTC	For constructing two-hybrid plasmid pAL1630. Reverse primer, which anneals to nucleotides 1,434–1,455 of tae17. Includes XmaI site and a stop codon.
BAP9025	AAAACTGCAGGACTGATGCTTTGAACGTGATT	For constructing two-hybrid plasmids pAL1623 and pAL1624. Forward primer, which anneals to nucleotides 2017–2037 of vgrG17. Includes PstI site.
BAP9026	AAAACCCGGGTTATTGTTCTTGCTGTTTAGTATC	For constructing two-hybrid plasmid pAL1623. Reverse primer, which anneals to nucleotides 2,536–2,556 of vgrG17. Includes XmaI site and a stop codon.
BAP9027	AAAACTGCAGGTCTTATGAGGTTCCTGAGTTG	For constructing two-hybrid plasmid pAL1621. Forward primer, which anneals to nucleotides 3,001–3,021 of vgrG17. Includes PstI site.
BAP9028	AAAACCCGGGTTAATTAGAATTATTATTTAAATTAGAATC	For constructing two-hybrid plasmids pAL1621 and pAL1624. Reverse primer, which anneals to nucleotides 3,277–3,306 of vgrG17. Includes XmaI site and a stop codon.
BAP9050	GTTCGCCATTATGCCGCATC	pKT25 forward primer for sequencing/colony PCR. Anneals upstream of the multiple cloning site
BAP9051	GGATGTGCTGCAAGGCGATT	pKT25 forward primer for sequencing/colony PCR. Reverse primer, which anneals downstream of the multiple cloning site
BAP9195	AAAACCCGGGTTAGCCCTTTTCATTCTTAATAACATG	For constructing two-hybrid plasmid pAL1689. Reverse primer, which anneals to nucleotides 73–96 of tae17. Includes XmaI site and a stop codon.
BAP9222	AAAACTGCAGGGACAACTATTGCTTCTGGAATGACAA ATAGTAATGGC	For constructing two-hybrid plasmid pAL1699. Forward primer, which anneals to nucleotides 97–132 of tae17. Includes PstI site.
BAP9234	AAAACCCGGGTTATCTACTTATCATTACTGTTAAGCC	For constructing two-hybrid plasmid pAL1701. Reverse primer, which anneals to nucleotides 130–153 of tae17. Includes XmaI site and a stop codon.
BAP9235	AAAACTGCAGGGTTATTTACTGGTAATAATATACC	For constructing two-hybrid plasmid pAL1702. Forward primer, which anneals to nucleotides 40–62 of tae17. Includes PstI site.
BAP9236	AAAACCCGGGTTAATTTTTTAAATATACATAAAGAGT ATCGCC	For constructing two-hybrid plasmid pAL1702. Reverse primer, which anneals to nucleotides 160–189 of tae17. Includes XmaI site and a stop codon.
BAP9237	AAAACCCGGGTTATAATTCATCTAAATTCCCTTGCG	For constructing two-hybrid plasmid pAL1703. Reverse primer, which anneals to nucleotides 3,128–3,150 of vgrG17. Includes XmaI site and a stop codon.
BAP9238	AAAACTGCAGGGGAAGAACAGAAAGATTCAAGAG	For constructing two-hybrid plasmid pAL1704. Forward primer, which anneals to nucleotides 3,151–3,173 of vgrG17. Includes PstI site.
BAP9239	AAAACCCGGGTTAAAAATCCTCAAACTCTTCTG	For constructing two-hybrid plasmid pAL1704. Reverse primer, which anneals to nucleotides 3,236–3,255 of vgrG17. Includes XmaI site and a stop codon.
BAP9246	AAAACCCGGGTTATTTTACTTTATGCGTTAAATATC	For constructing two-hybrid plasmid pAL1716. Reverse primer, which anneals to nucleotides 326–348 of tae17. Includes XmaI site and a stop codon.
BAP9260	TTCGAAAATATTAAATATAGTGC	For constructing pBASE plasmid pAL1779 for alanine scanning of VgrG17. Forward primer, which anneals to nucleotides 3,088–3,104 of vgrG17. Used for SOE-PCR to introduce a silent SpeI site into vgrG17 on plasmid pAL1447. Includes BstBI site
BAP9261	GAGCTCGGATCCTTAATTAGAATTATTATTTAAATTAG	For constructing pBASE plasmid pAL1779 for alanine scanning of VgrG17. Reverse primer, which anneals to 3′ end of vgrG17 and MCS of pBASE. For SOE-PCR to introduce a silent SpeI site into vgrG17 on plasmid pAL1447. Includes stop codon, BamHI and SacI sites
BAP9262	CTATATCCCTTTAGTACCTTCATTTTGTCAGGC	For constructing pBASE plasmid pAL1779 for alanine scanning of VgrG17. Reverse primer, which anneals to nucleotides 3,180–3,212 of vgrG17. Used SOE-PCR to introduce a silent SpeI site into vgrG17 on pAL1779. Includes SpeI site
BAP9263	GAAGGTACTAGTGGGATATAGTGCTGATAAATG	For constructing pBASE plasmid pAL1779 for alanine scanning of VgrG17. Forward primer, which anneals to nucleotides 3,192–3,224 of vgrG17. SOE-PCR to introduce a silent SpeI site into vgrG17 on pAL1779. Includes SpeI site
BAP9317	GTAAAACGACGGCCAGTGAATTCGAGCTCTTAATTAG	For constructing all plasmids encoding VgrG17 alanine mutations. Reverse primer, which anneals to the 3′ end of vgrG17 in pAL1779. Includes SacI site and stop codon
BAP9318	AAAAAAACTAGTGGCAGCTAGTGCTGATAAATGGCATAAAG	For construction of alanine scanning plasmid pAL1784 encoding VgrG17 G1069A, Y1070A. Forward primer, which anneals to nucleotides 3,196–3,228 of vgrG17 in pAL1779. Includes SpeI site
BAP9319	AAAAAAACTAGTGGGATATGCTGGTGATAAATGGCATAAAGAAAC	For construction of alanine scanning plasmid pAL1785 encoding VgrG17 S1071A, A1072G Forward primer, which anneals to nucleotides 3,196–3,234 of vgrG17. Includes SpeI site
BAP9320	AAAAAAACTAGTGGGATATAGTGCTGCTGCATGGCATAAAGAAACAGAAG	For construction of alanine scanning plasmid pAL1786 encoding VgrG17 D1073A, K1074A. Forward primer, which anneals to nucleotides 3,196–3,153 of vgrG17. Includes SpeI site
BAP9321	AAAAAAACTAGTGGGATATAGTGCTGATAAAGCGGCTAAAGAAACAGAAGAGTTTG	For construction of alanine scanning plasmid pAL1787 encoding VgrG17 W1075A, H1076A. Forward primer, which anneals to nucleotides 3,196–3,245 of vgrG17. Includes SpeI site
BAP9428	AAAAAAACTAGTGGGATATAGTGCTGATAAATGGCATGCAGCAACAGAAGAGTTTGAGGATTTTG	For construction of alanine scanning plasmid pAL1793 encoding VgrG17 K1077A, E1078A. Forward primer, which anneals to nucleotides 3,196–3,254 of vgrG17. Includes SpeI site
BAP9429	AAAAAAACTAGTGGGATATAGTGCTGATAAATGGCATAAAGAAGCAGCAGAGTTTGAGGATTTTGATG	For construction of alanine scanning plasmid pAL1794 encoding VgrG17 T1079A, E1080A. Forward primer, which anneals to nucleotides 3,196–3,257 of vgrG17. Includes SpeI site
BAP9430	AAAAAAACTAGTGGGATATAGTGCTGATAAATGGCATAAAGAAACAGAAGCTGCTGAGGATTTTGATGACTATG	For construction of alanine scanning plasmid pAL1795 encoding VgrG16 E1081A, F1082A. Forward primer, which anneals to nucleotides 3,196–3,263 of vgrG17. Includes SpeI site
BAP9431	AAAAAAACTAGTGGGATATAGTGCTGATAAATGGCATAAAGAAACAGAAGAGTTTGCGGCTTTTGATGACTATGA AACC	For construction of alanine scanning plasmid pAL1796 encoding VgrG17 E1083A, D1084A. Forward primer, which anneals to nucleotides 3196–3268 of vgrG17. Includes SpeI site
BAP9432	AAAAAAACTAGTGGGATATAGTGCTGATAAATGGCATAAAGAAACAGAAGAGTTTGAGGATGCTGCTGACTATGAAACCAACAATG	For construction of alanine scanning plasmid pAL1797 encoding VgrG17 F1085A, D1086A. Forward primer, which anneals to nucleotides 3,196–3,275 of vgrG17. Includes SpeI site
BAP9440	GCACGCAGAAAGTCAATTC	For sequencing pAL1789 Tae17 expression plasmid. Forward primer, which anneals to nucleotides 544–563 of the recombinant tae17 gene which has been codon optimized for expression in E. coli
BAP9465	ATGCTTTTTAATATATTTAGTGTCTTGGAGAAGTTCCTA TACTTTCTAGAGAATAGGAACTTCCCGGAATTGCCAGC TGGGG	For construction of A. baumannii AB307_0294 vgrG17 deletion mutant AL4329. Forward primer for the amplification of the KanR cassette. Contains 5′ overhang specific for vgrG17
BAP9466	TTAATTAGAATTATTATTTAAATTAGAATCGAAGTTCCT ATTCTCTAGAAAGTATAGGAACTTCTTCAGAAGAACTC GTCAAGAAGGCG	For construction of A. baumannii AB307_0294 vgrG17 deletion mutant AL4329. Reverse primer for the amplification of the KanR cassette. Contains 5′ overhang specific for vgrG17
BAP9467	CTTACTCGATCAGACCTATGC	For construction of A. baumannii AB307_0294 vgrG17 deletion mutant AL4329. Forward primer, which anneals ~ 1,000 bp upstream of vgrG17. For amplification of region upstream of vgrG17 for SOE-PCR-mediated mutagenesis
BAP9468	TCCAAGACACTAAATATATTAAAAAGC	For construction of A. baumannii AB307_0294 vgrG17 deletion mutant AL4329. Reverse primer complementary to the 5′ end of vgrG17. For amplification of region upstream of vgrG17 for SOE-PCR-mediated mutagenesis
BAP9469	GATTCTAATTTAAATAATAATTCTAATTAAG	For construction of A. baumannii AB307_0294 vgrG17 deletion mutant AL4329. Forward primer complementary to the 3′ end of vgrG17. For amplification of region downstream of vgrG17 for SOE-PCR-mediated mutagenesis
BAP9472	GAATTGACTTTCTGCGTGC	For sequencing pAL1789 Tae17 expression plasmid. Reverse primer, which anneals to nucleotides 544–563 of tae17, which has been codon optimized for expression in E. coli.
BAP9473	GCGTCTTTGGGTGCCGCATATAC	For construction of A. baumannii AB307_0294 vgrG17 deletion mutant AL4329. Reverse primer, which anneals ~ 1,000 bp downstream of vgrG17. For amplification of region upstream of vgrG17 for SOE-PCR-mediated mutagenesis
BAP9474	AAAATCGGTACTTTCATCAGC	For sequencing pAL1789 Tae17 expression plasmid. Reverse primer, which anneals to nucleotides 855–876 of tae17, which has been codon optimized for expression in E. coli
BAP9487	AAAAAAACTAGTGGCATATAGTGCTGATAAATGGCATA AAG	For construction of alanine scanning plasmid pAL1809 encoding VgrG17 G1069A. Forward primer, which anneals to nucleotides 3,198–3,233 of vgrG17. Includes SpeI site
BAP9488	AAAAAAACTAGTGGGAGCTAGTGCTGATAAATGGCATA AAG	For construction of alanine scanning plasmid pAL1810 encoding VgrG17 Y1070A. Forward primer, which anneals to nucleotides 3,198–3,233 of vgrG17. Includes SpeI site
BAP9489	AAAAAAACTAGTGGGATATAGTGCTGATAAAGCGCATA AAGAAACAGAAGAGTTTG	For construction of alanine scanning plasmid pAL1811 encoding VgrG17 W1075A. Forward primer, which anneals to nucleotides 3,198–3,248 of vgrG17. Includes SpeI site
BAP9490	AAAAAAACTAGTGGGATATAGTGCTGATAAATGGGCTA AAGAAACAGAAGAGTTTG	For construction of alanine scanning plasmid pAL1812 encoding VgrG17 H1076A. Forward primer, which anneals to nucleotides 3,198–3,253 of vgrG17. Includes SpeI site
BAP9592	CCCGGTTTCCAATGCGATTAAAC	For colony PCR/ sequencing of A. baumannii AB307_0294 vgrG17 deletion mutant AL4329. Forward primer anneals ~ 500 bp upstream of the region of recombination.
BAP9593	CCCGGGAAGGAGCTAGCTATGAAATACCTGCT GCCGACCGCTGCTGCTGGTCTGCTGCTCCTCG CTGCCCAGCCGGCGATGGCCAATAAAAAGGGA CTAATACAATTTAGGTTTGTTG	For constructing via PCR Tae17 pBAD30 expression plasmids pAL1854, pAL1856, and pAL1857 using the Genscript plasmid pAL1789 as template. Forward primer, which anneals to nucleotides 1–34 of E. coli codon optimized tae17. Includes an Xmal site, a ribosome-binding site, start codon, and PelB leader. Used for constructing pAL1854, pAL1856, and pAL1857
BAP9594	AAGCTTTTACATGCGGAAATAGCGG	For constructing Tae17 pBAD30 expression plasmids pAL1854, pAL1856, and pAL1857 using the Genscript plasmid pAL1789 as template. Reverse primer, which anneals to nucleotides 1,719–1,746 of the recombinant tae17 gene, which has been codon optimized for expression in E. coli. Includes a stop codon and HindIII site
BAP9595	CATCAGGCTGCGAAGTATGTTCGCATC	For constructing Tae17 pBAD30 expression plasmid pAL1854 via SOE-PCR. Forward primer, which anneals to nucleotides 1,452–1,479 the recombinant tae17 gene, which has been codon optimized for expression in E. coli. Encodes the C488A mutation
BAP9596	CTTCGCAGCCTGATGTTGACTTTTCTC	For constructing Tae17 pBAD30 expression plasmid pAL1854 via SOE-PCR. Reverse primer, which anneals to nucleotides 1,440–1,467 the recombinant tae17 gene which has been codon optimized for expression in E. coli. Encodes the C488A mutation
BAP9673	CAGTTATGCATTTAGCATCGGCGGGTACGTTC	For constructing Tae17 pBAD30 expression plasmid pAL1870 and pAL1871. Forward primer, which anneals to nucleotides 814–846 of the recombinant tae17 gene, which has been codon optimized for expression in E. coli. Encodes the E278A mutation
BAP9674	GAACGTACCCGCCGATGCTAAATGCATAACTG	For constructing Tae17 pBAD30 expression plasmid pAL1870 and pAL1871. Reverse primer, which anneals to nucleotides 814–846 of the recombinant tae17 gene, which has been codon optimized for expression in E. coli. Encodes the E278A mutation

^a

Introduced restriction sites in primers are underlined.

To construct each plasmid (Table 2) for the adenylate cyclase two-hybrid assay, a region of vgrG17 or tae17 was PCR-amplified using a forward primer containing the desired mutations together with the appropriate primer pair, both with suitable restriction sites (Table 3) then ligated to similarly digested pUT18C or pKT25 so that each encoded product was in-frame and at the C-terminal end of the encoded adenylate cyclase fragment T18 or T25, respectively. Pairs of recombinant pUT18C and pKT25 plasmids with the correct in-frame vgrG17 and tae17 sequence, respectively, were then introduced into chemically competent E. coli BTH101 cells and co-transformants recovered on ampicillin and kanamycin.

For single-site mutagenesis of tae17 in pBAD30 (pAL1854, pAL1871, Table 2), splicing by overlap-extension PCR (SOE-PCR) (49), using pAL1789 plasmid template and primers encoding the desired change and restriction enzyme sites (Table 3), was employed to generate each DNA fragment encoding a mutated tae17 gene. Each SOE-PCR product was then restriction enzyme digested and ligated to similarly digested pBAD30. For generation of tae17 encoding both the E278A and C488A mutations, QuikChange site-directed mutagenesis (Agilent Technologies, Inc., Santa Clara, CA) was performed with primer pair BAP9673/9674 to introduce the mutation encoding E278A into the gene using pAL1854 (encoding Tae17^C488A) as the template. Following digestion with DpnI to remove template DNA, products were introduced into E. coli XL-1 Blue to allow for nick repair.

For alanine scanning mutagenesis of recombinant vgrG17, the cloning strategy required modification of pAL1447 to introduce a silent SpeI restriction site at the 3′ end of vgrG17 as follows. An SOE-PCR product spanning vgrG17 nts 3082 (BstBI) to SacI (in the vector sequence of pAL1447) was made with first-round products generated with pAL1447 template and primer pairs BAP9260/BAP9262 and BAP9263/BAP9261. The product was digested with BstB1 and SacI and cloned into similarly digested pAL1447 (Table 2). To generate SOE-PCR products containing mutations for introducing single and double alanine substitutions into VgrG17, we used one of a series of long forward primers containing the desired mutations and an SpeI site (BAP9318-9321, 9428–9432, Table 3), together with reverse primer BAP9317 containing a SacI site (located downstream of vgrG17 in the vector). Each SOE-PCR product, representing nucleotides 3,196–3,306 of vgrG17 with the desired mutations, was then digested and ligated into similarly digested pAL1779.

Construction of A. baumannii site-directed mutants

A. baumannii strain AB307_0294 mutants AL4060 and AL4329 (Table 2) were constructed by allelic replacement of each gene using SOE-PCR products generated with the appropriate primers (Table 3) as previously described (6) with minor modifications as follows. The central kanamycin resistance gene aph (3′) was flanked by FRT sites to allow for later excision of the kanamycin cassette leaving an in-frame FRT scar. The SOE product was column-purified and used to transform A. baumannii as previously described (50). The kanamycin resistance cassette was removed from the mutation site via the introduction of the plasmid pAT03 (Table 2), which encodes the Flp recombinase (46). The second and third site-directed mutations in AL4060 were constructed following the same procedure as above using SOE-PCR products with the exception that the kanamycin cassette was retained following vgrG17 deletion in AL4060. All regions of the genome that underwent mutagenesis were subjected to diagnostic PCR using sets of primers flanking the deleted region, followed by DNA Sanger sequencing across the site of mutagenesis.

His-tagged Tae17 expression and purification

For in vitro experiments involving FITC-labeled peptidoglycan and in vitro synthesized dansylated lipid II polymers, recombinant Tae17 and mutated Tae17 proteins were produced as follows. Protein expression plasmids (pET28a+), each with an E. coli codon-optimized gene encoding C-terminal His_x6-tagged Tae17^WT (pAL1789) Tae17^E278A (pAL1919), Tae17^C488A (pAL1920), or Tae17^E278A,C488A (pAL1885) were constructed by Genscript. E. coli C41 cells, each containing a Tae17 expression plasmid, were separately grown in 400 mL of 2YT supplemented with kanamycin at 37°C with shaking at 200 rpm to an OD_{600 nm} of 0.6. Recombinant protein expression was induced via the addition of 1 mM IPTG and further incubation at 18°C with shaking (200 rpm) for ~18 h. Following centrifugation (5,000 × g, 20 min), cells were resuspended in 40 mL of a phosphate buffer consisting of 436 mM NaCl, 7 mM K₂HPO₄, 2.5 mM KH₂ PO₄ (pH8.0), 5% wt/vol glycerol (used in all subsequent steps), 20 mM imidazole and Pierce Protease Inhibitor (one mini-tablet per 10 mL buffer). Cells were sonicated on ice (Soniprep 150 Plus, 10 mAmp, 6 × 30 s) and then centrifuged (15,000 × g, 20 min) to remove insoluble material. His-tagged proteins were purified from each cell lysate by gently mixing lysate with 1 mL HisPur Ni-NTA resin (Thermoscientific), pre-equilibrated in the above buffer. After loading onto a column for immobilized metal-affinity chromatography (IMAC), the resin was washed several times in the above buffer, and then washed several times with phosphate buffer containing 50 mM imidazole. Proteins were eluted with phosphate buffer containing 250 mM imidazole. Elution fractions (1 mL) were concentrated using an Amicon Ultra-15 Centrifugal Filter Unit (Merck) and further purified by size exclusion chromatography using HiLoad 16/600 Superdex 75 column (Cytiva) pre-equilibrated in PBS containing 300 mM NaCl. Fractions with recombinant protein were pooled and concentrated, and purity were assessed using SDS-PAGE and Coomassie blue staining.

E. coli viability assays

Cultures of E. coli DH5α containing pBAD30 plasmids encoding Tae17^WT, Tae17^E278A, Tae17^C488A, or Tae17^E278A,C488A (all with PelB leader sequences) were grown in LB, supplemented with ampicillin and 0.4% glucose (to suppress recombinant protein expression), at 37°C with shaking (200 rpm) to an optical density (OD_{600 nm}) of 0.2. Cells were centrifuged (5,000× g for 5 min) and then resuspended in 30 mL phosphate-buffered saline (PBS). Cells were washed once more as above, resuspended in 30 mL of fresh LB supplemented with ampicillin, then divided equally (15 mL) into two sterile flasks; one supplemented with 0.4% glucose and the other with 0.4% arabinose to induce P_araBAD promoter and recombinant protein expression. Induced and uninduced cultures were incubated at 37°C as above and OD_600nm measurements were taken at 0 h, 1 h, 3 h, and 5 h as well as samples to determine CFU (10-fold dilutions, 10 µL aliquots of each dilution spotted onto LB agar supplemented with 0.4% glucose and ampicillin).

T6SS interbacterial killing assays

To assess the efficiency of Tae17 delivery by the T6SS, interbacterial killing assays were used. For each assay, an appropriate A. baumannii predator strain was used in co-culture on solid media with an E. coli DH10B prey strain, harboring either vector only (pWH1266) or pAL1263 (pWH1266 encoding the immunity proteins Tsi15 and Tdi16 to neutralize the activity of the two other T6SS antibacterial effectors). Assays were performed (≥3 biological replicates) as described previously (6), with minor modifications; co-cultures (10 µL) were spotted onto LB agar plates and incubated for 3 h at 37°C. For qualitative interbacterial killing assays (survival or non-survival of E. coli following co-culture), each co-culture growth was scraped off the plate and then all collected material added as a primary streak only onto LB agar supplemented with tetracycline to select for surviving E. coli prey, and LB agar supplemented with kanamycin (50 µg/mL) or carbenicillin (100 µg/mL) to select for the presence of the A. baumannii AB307-0294 predator strain. For the quantitative T6SS interbacterial killing assay, each co-culture spot was excised from the agar, resuspended in 1 mL of PBS by vigorous vortexing, and then serially diluted in PBS. Aliquots (10 µL) of each dilution were spotted onto LB agar supplemented with appropriate antibiotics (as above) to select for either predator or prey.

Bacterial two-hybrid adenylate cyclase assays

To identify interacting regions between VgrG17 and Tae17, we used bacterial adenylate cyclase two-hybrid assays as described previously (43, 51).

Detection of Hcp to assess T6SS activity

For the detection of the T6SS protein Hcp in A. baumannii whole-cell lysates and culture supernatants by western immunoblotting, samples were prepared as follows. One milliliter of the mid-log phase A. baumannii (OD_600nm~2.0) was centrifuged (5,000 × g, 5 min), supernatant collected, and cell pellet resuspended in 100 µL SDS-PAGE sample buffer. The supernatant was concentrated 10× using the Amicon Ultra Centrifugal Filter Units (Millipore), and an equal volume of 2× SDS-PAGE sample buffer was added. For loading, samples were heated to 99°C (10 min) and then centrifuged (11,000 × g for 5 min) to remove insoluble material. A 10 µL aliquot of each sample was subjected to electrophoresis (15% PAGE gel, 100 V for 50 min) and then transferred by western blot to a 0.2 µm nitrocellulose membrane (BioRad) (52). Hcp was detected using a polyclonal chicken anti-Hcp serum (1:500) (6), followed by a secondary HRP-conjugated donkey anti-chicken antibody (1:1000) (Merck Millipore). The western blots were analyzed with Clarity Western ECL substrate (BioRad) and visualized on an Amersham Imager 680.

Peptidoglycan degradation assays

DAP-containing (Gram-negative) peptidoglycan was purchased from Merck and labeled with fluorescein isothiocyanate (FITC) (53). Peptidoglycan degradation assays followed the method from (54). FITC-labeled peptidoglycan was mixed with 1 µM of protein and buffer A (1× PBS pH 8.0, 5% glycerol, 300 mM NaCl, 20 mM imidazole) was added with the protein up to a total volume of 100 µL with 5 µL of FITC-labeled peptidoglycan (10 mg/mL). The protein-peptidoglycan mix was incubated at 37°C for 2 h to allow for breakdown to occur. Following this, the samples were added to a filter plate (0.2 µM filter) to quench the reaction, and the flow through was analyzed in a plate reader (Ex. 495 nm and Em. 519 nm). The plate reader then measured the fluorescence against the positive control (Lysozyme) to adjust the gain and a single timepoint measurement was recorded.

Dansylated lipid II transglycosylase activity assay

Dansylated lipid II (LII) was polymerized with the highly processive monofunctional glycosyltransferase enzyme (MGT) from S. aureus. Polymerization occurred by incubating 10 µM of dansylated LII with 0.5 µM MGT for 2 h at 37°C in a reaction mixture consisting of Reaction Buffer (100 mM NaCl, 50 mM HEPES, 10 mM MgCl₂), 20% (vol/vol) dimethylsulfoxide (DMSO) and 600 µM lauryldimethylamine oxide (LDAO). This allowed for polymerization of the LII into a high molecular weight polymer. The reaction was stopped by denaturation of the enzyme at 80°C for 10 min. Tae17^WT, Tae17^E278A, Tae17^C488A, or Tae17^E278A,C488A were added separately to the reaction mixture at a concentration of 1 µM (final volume 15 µL) and the reactions were incubated at 37°C for a further 2 h. If the Tae17 species were active, this would allow for the high molecular weight polymer to be broken down. A 3 µL aliquot of 5× loading dye (10% SDS, 250 mM Tris-HCl pH 6.8, 5% β-mercaptoethanol, 0.02% bromophenol blue) was added to stop the reaction. The samples were loaded onto a Novex 16% Tricine gel (Invitrogen) with anode buffer (100 mM Tris-HCl, pH 8.8) in the main body of the gel tank and cathode buffer (100 mM Tris-HCl, 100 mM Tricine, 0.1% [wt/vol] SDS) in the gel cavity. The gel was electrophoresed at 100 V for 2 h and imaged with BioRad ChemiDoc MP imaging system using the ethidium bromide setting with a 30 s exposure.

Structural modeling and analysis

For de novo protein structure models, AlphaFold version 2.1.1 was used on the MASSIVE M3 computing cluster (55). For the Tae17 model, the full-length protein sequence was used as input and monomer mode was used. For the VgrG17:Tae17 interaction model, residues 1,010–1,101 of VgrG17 and full-length Tae17 were used as sequence input and multimer mode was used. For both runs, the full database was utilized, and the directory was set as /mnt/reference/alphafold/alphafold_20211129. Five models were produced, with an unrelaxed and relaxed output for each. For the monomer mode, the relaxed models were ranked on their predicted local distance difference test (pLDDT) scores, and the best pLDDT score was used for model building. The top score for Tae17 full-length modeling was an average pLDDT of 76.5. For the multimer mode, relaxed models were ranked using a predicted template modeling (ipTM) score. The top-ranked model (by ipTM score) was used for analysis. The AlphaFold2-generated Tae17 coordinates are now publicly available on the AlphaFold database under accession A0A5K6CL85 (55). All structures were visualized using UCSF Chimera (v1.15) or ChimeraX (v1.6.1). PDBePISA, an online tool for examining protein interfaces, was used to further assess the interactions between Tae17 and VgrG17 (37).

Statistical analyses

For E. coli survival in interbacterial killing assays, E. coli recombinant strain viability counts, and optical density of cultures, statistical significance between samples was determined by log₁₀ transforming the data and performing ordinary one-way ANOVA with Tukey’s multiple comparisons test (GraphPad Prism) with significance defined as P < 0.05. For the growth curve experiment, involving recombinant E. coli strains, statistical analysis was only performed on the final (5 h) timepoint samples. For the peptidoglycan degradation assays, ordinary one-way ANOVA was used with Tukey’s multiple comparisons test (GraphPad Prism) with significance defined as P < 0.05.

ACKNOWLEDGMENTS

This work was supported by NHMRC grants 1165036 (to J.D.B. and M.H.) and 1128981 (J.D.B., S.M., and M.H.). V.B. and B.K.H. were supported by RTP scholarships from the Australian Department of Education. This work was supported by the MASSIVE HPC facility () (56). We thank David Steer and the Monash Proteomics and Metabolomics Platform (MPMP) for Mass spectrometry protein identification. We thank Jordan Thompson for the construction and sharing of the A. baumannii ΔtssM mutant strain.

References

Tacconelli E, Carrara E, Savoldi A, Harbarth S, Mendelson M, Monnet DL, Pulcini C, Kahlmeter G, Kluytmans J, Carmeli Y, Ouellette M, Outterson K, Patel J, Cavaleri M, Cox EM, Houchens CR, Grayson ML, Hansen P, Singh N, Theuretzbacher U, Magrini N, WHO Pathogens Priority List Working Group. 2018. Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis 18:318–327.

Bock LJ. 2019. Bacterial biocide resistance: a new scourge of the infectious disease world? Arch Dis Child 104:1029–1033.

Dijkshoorn L, Nemec A, Seifert H. 2007. An increasing threat in hospitals: multidrug-resistant Acinetobacter baumannii. Nat Rev Microbiol 5:939–951.

Russell AB, Singh P, Brittnacher M, Bui NK, Hood RD, Carl MA, Agnello DM, Schwarz S, Goodlett DR, Vollmer W, Mougous JD. 2012. A widespread bacterial type VI secretion effector superfamily identified using A heuristic approach. Cell Host Microbe 11:538–549.

Carruthers MD, Nicholson PA, Tracy EN, Munson RS. 2013. Acinetobacter baumannii utilizes a type VI secretion system for bacterial competition. PLoS One 8:e59388.

Fitzsimons TC, Lewis JM, Wright A, Kleifeld O, Schittenhelm RB, Powell D, Harper M, Boyce JD. 2018. Identification of novel Acinetobacter baumannii Type VI secretion system antibacterial effector and immunity pairs. Infect Immun 86:e00297–00218.

Zoued A, Brunet YR, Durand E, Aschtgen M-S, Logger L, Douzi B, Journet L, Cambillau C, Cascales E. 2014. Architecture and assembly of the type VI secretion system. Biochim et Biophys Acta (BBA) - Mol Cell Res 1843:1664–1673.

Cianfanelli FR, Monlezun L, Coulthurst SJ. 2016. Aim, load, fire: the type VI secretion system, a bacterial nanoweapon. Trends Microbiol 24:51–62.

Brunet YR, Zoued A, Boyer F, Douzi B, Cascales E. 2015. The type VI secretion TssEFGK-VgrG phage-like baseplate is recruited to the TssJLM membrane complex via multiple contacts and serves as assembly platform for tail tube/sheath polymerization. PLoS Genet 11:e1005545.

Osipiuk J, Xu X, Cui H, Savchenko A, Edwards A, Joachimiak A. 2011. Crystal structure of secretory protein Hcp3 from Pseudomonas aeruginosa. J Struct Funct Genomics 12:21–26.

Renault MG, Zamarreno Beas J, Douzi B, Chabalier M, Zoued A, Brunet YR, Cambillau C, Journet L, Cascales E. 2018. The gp27-like hub of VgrG serves as adaptor to promote hcp tube assembly. J Mol Biol 430:3143–3156.

Shneider MM, Buth SA, Ho BT, Basler M, Mekalanos JJ, Leiman PG. 2013. PAAR-repeat proteins sharpen and diversify the type VI secretion system spike. Nature New Biol 500:350–353.

Günther P, Quentin D, Ahmad S, Sachar K, Gatsogiannis C, Whitney JC, Raunser S. 2022. Structure of a bacterial Rhs effector exported by the type VI secretion system. PLoS Pathog 18:e1010182.

Pukatzki S, Ma AT, Revel AT, Sturtevant D, Mekalanos JJ. 2007. Type VI secretion system translocates a phage tail spike-like protein into target cells where it cross-links actin. Proc Natl Acad Sci U S A 104:15508–15513.

Bondage DD, Lin JS, Ma LS, Kuo CH, Lai EM. 2016. VgrG C terminus confers the type VI effector transport specificity and is required for binding with PAAR and adaptor-effector complex. Proc Natl Acad Sci U S A 113:E3931–40.

Liang X, Pei T-T, Li H, Zheng H-Y, Luo H, Cui Y, Tang M-X, Zhao Y-J, Xu P, Dong T. 2021. VgrG-dependent effectors and chaperones modulate the assembly of the type VI secretion system. PLoS Pathog 17:e1010116.

Flaugnatti N, Le TTH, Canaan S, Aschtgen M-S, Nguyen VS, Blangy S, Kellenberger C, Roussel A, Cambillau C, Cascales E, Journet L. 2016. A phospholipase A1 antibacterial Type VI secretion effector interacts directly with the C-terminal domain of the VgrG spike protein for delivery. Mol Microbiol 99:1099–1118.

Pukatzki S, Ma AT, Sturtevant D, Krastins B, Sarracino D, Nelson WC, Heidelberg JF, Mekalanos JJ. 2006. Identification of a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system. Proc Natl Acad Sci U S A 103:1528–1533.

Smith WPJ, Vettiger A, Winter J, Ryser T, Comstock LE, Basler M, Foster KR. 2020. The evolution of the type VI secretion system as a disintegration weapon. PLoS Biol 18:e3000720.

Russell AB, Hood RD, Bui NK, LeRoux M, Vollmer W, Mougous JD. 2011. Type VI secretion delivers bacteriolytic effectors to target cells. Nature New Biol 475:343–347.

Le N-H, Pinedo V, Lopez J, Cava F, Feldman MF. 2021. Killing of Gram-negative and Gram-positive bacteria by a bifunctional cell wall-targeting T6SS effector. Proc Natl Acad Sci U S A 118:e2106555118.

Adams MD, Goglin K, Molyneaux N, Hujer KM, Lavender H, Jamison JJ, MacDonald IJ, Martin KM, Russo T, Campagnari AA, Hujer AM, Bonomo RA, Gill SR. 2008. Comparative genome sequence analysis of multidrug-resistant Acinetobacter baumannii. J Bacteriol 190:8053–8064.

Lewis JM, Deveson Lucas D, Harper M, Boyce JD. 2019. Systematic identification and analysis of Acinetobacter baumannii type VI secretion system effector and immunity components. Front Microbiol 10:2440.

Berisio R, Ciccarelli L, Squeglia F, De Simone A, Vitagliano L. 2012. Structural and dynamic properties of incomplete immunoglobulin-like fold domains. Protein Pept Lett 19:1045–1053.

Bateman A, Bycroft M. 2000. The structure of a LysM domain from E. coli membrane-bound lytic murein transglycosylase D (MltD). J Mol Biol 299:1113–1119.

Holm L, Laiho A, Törönen P, Salgado M. 2023. DALI shines a light on remote homologs: one hundred discoveries. Protein Sci 32:e4519.

Kumar V, Mathure SA, Lee M, Boorman J, Zeng X, Lin J, Hesek D, Lastochkin E, Mobashery S, van den Akker F. 2021. Turnover chemistry and structural characterization of the Cj0843c lytic transglycosylase of Campylobacter jejuni. Biochemistry 60:1133–1144.

Dong C, Zhang H, Gao Z-Q, Wang W-J, She Z, Liu G-F, Shen Y-Q, Su X-D, Dong Y-H. 2013. Structural insights into the inhibition of type VI effector Tae3 by its immunity protein Tai3. Biochem J 454:59–68.

Bateman A, Rawlings ND. 2003. The CHAP domain: a large family of amidases including GSP amidase and peptidoglycan hydrolases. Trends Biochem Sci 28:234–237.

Guzman LM, Belin D, Carson MJ, Beckwith J. 1995. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol 177:4121–4130.

Fukuhara S, Nakane T, Yamashita K, Ishii R, Ishitani R, Nureki O. 2018. Crystal structure of the Agrobacterium tumefaciens type VI effector-immunity complex. Acta Cryst F Struct Biol Commun 74:810–816.

Zapun A, Philippe J, Abrahams KA, Signor L, Roper DI, Breukink E, Vernet T. 2013. In vitro reconstitution of peptidoglycan assembly from the Gram-positive pathogen Streptococcus pneumoniae. ACS Chem Biol 8:2688–2696.

Vollmer W, Blanot D, de Pedro MA. 2008. Peptidoglycan structure and architecture. FEMS Microbiol Rev 32:149–167.

Schägger H, von Jagow G. 1987. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 166:368–379.

Wu C-F, Lien Y-W, Bondage D, Lin J-S, Pilhofer M, Shih Y-L, Chang JH, Lai E-M. 2020. Effector loading onto the VgrG carrier activates type VI secretion system assembly. EMBO Rep 21:e47961.

Flaugnatti N, Rapisarda C, Rey M, Beauvois SG, Nguyen VA, Canaan S, Durand E, Chamot-Rooke J, Cascales E, Fronzes R, Journet L. 2020. Structural basis for loading and inhibition of a bacterial T6SS phospholipase effector by the VgrG spike. EMBO J 39:e104129.

Krissinel E, Henrick K. 2007. Inference of macromolecular assemblies from crystalline state. J Mol Biol 372:774–797.

Pukatzki S, McAuley SB, Miyata ST. 2009. The type VI secretion system: translocation of effectors and effector-domains. Curr Opin Microbiol 12:11–17.

Murdoch SL, Trunk K, English G, Fritsch MJ, Pourkarimi E, Coulthurst SJ. 2011. The opportunistic pathogen Serratia marcescens utilizes type VI secretion to target bacterial competitors. J Bacteriol 193:6057–6069.

Chen L, Zou Y, She P, Wu Y. 2015. Composition, function, and regulation of T6SS in Pseudomonas aeruginosa. Microbiol Res 172:19–25.

Crisan CV, Hammer BK. 2020. The Vibrio cholerae type VI secretion system: toxins, regulators and consequences. Environ Microbiol 22:4112–4122.

Chou S, Bui NK, Russell AB, Lexa KW, Gardiner TE, LeRoux M, Vollmer W, Mougous JD. 2012. Structure of a peptidoglycan amidase effector targeted to Gram-negative bacteria by the type VI secretion system. Cell Rep 1:656–664.

Benz J, Reinstein J, Meinhart A. 2013. Structural insights into the effector - immunity system Tae4/Tai4 from Salmonella typhimurium. PLoS One 8:e67362.

Grant SGN, Jessee J, Bloom FR, Hanahan D. 1990. Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants. Proc Natl Acad Sci USA 87:4645–4649.

Karimova G, Pidoux J, Ullmann A, Ladant D. 1998. A bacterial two-hybrid system based on a reconstituted signal transduction pathway. Proc Natl Acad Sci U S A 95:5752–5756.

Tucker AT, Nowicki EM, Boll JM, Knauf GA, Burdis NC, Trent MS, Davies BW. 2014. Defining gene-phenotype relationships in Acinetobacter baumannii through one-step chromosomal gene inactivation. MBio 5:e01313–14.

Hunger M, Schmucker R, Kishan V, Hillen W. 1990. Analysis and nucleotide sequence of an origin of DNA replication in Acinetobacter calcoaceticus and its use for Escherichia coli shuttle plasmids. Gene 87:45–51.

Gulliver EL, Sy BM, Wong JL, Deveson Lucas DS, Powell DR, Harper M, Tree JJ, Boyce JD. 2022. The role and targets of the RNA-binding protein ProQ in the Gram-negative bacterial pathogen pasteurella multocida. J Bacteriol 204:e0059221.

Heckman KL, Pease LR. 2007. Gene splicing and mutagenesis by PCR-driven overlap extension. Nat Protoc 2:924–932.

Choi KH, Kumar A, Schweizer HP. 2006. A 10-min method for preparation of highly electrocompetent Pseudomonas aeruginosa cells: application for DNA fragment transfer between chromosomes and plasmid transformation. J Microbiol Methods 64:391–397.

Hayes BK, Harper M, Venugopal H, Lewis JM, Wright A, Lee H-C, Steele JR, Steer DL, Schittenhelm RB, Boyce JD, McGowan S. 2024. Structure of a Rhs effector clade domain provides mechanistic insights into type VI secretion system toxin delivery. Nat Commun 15:8709.

Mahmood T, Yang PC. 2012. Western blot: technique, theory, and trouble shooting. N Am J Med Sci 4:429–434.

Maeda H. 1980. A new lysozyme assay based on fluorescence polarization or fluorescence intensity utilizing A fluorescent peptidoglycan substrate. J Biochem 88:1185–1191.

Gurnani Serrano CK, Winkle M, Martorana AM, Biboy J, Morè N, Moynihan P, Banzhaf M, Vollmer W, Polissi A. 2021. ActS activates peptidoglycan amidases during outer membrane stress in Escherichia coli. Mol Microbiol 116:329–342.

Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, Tunyasuvunakool K, Bates R, Žídek A, Potapenko A, et al. 2021. Highly accurate protein structure prediction with AlphaFold. Nature New Biol 596:583–589.

Goscinski WJ, McIntosh P, Felzmann U, Maksimenko A, Hall CJ, Gureyev T, Thompson D, Janke A, Galloway G, Killeen NEB, Raniga P, Kaluza O, Ng A, Poudel G, Barnes DG, Nguyen T, Bonnington P, Egan GF. 2014. The multi-modal Australian ScienceS imaging and visualization environment (MASSIVE) high performance computing infrastructure: applications in neuroscience and neuroinformatics research. Front Neuroinform 8:30.