Skip to main content
NCBI home page
EMBO Reports logoLink to EMBO Reports
. 2023 Apr 17;24(6):e56849. doi: 10.15252/embr.202356849

A glycine zipper motif is required for the translocation of a T6SS toxic effector into target cells

Jemal Ali 1,2,3, Manda Yu 1,5, Li‐Kang Sung 1, Yee‐Wai Cheung 1, Erh‐Min Lai 1,2,4,✉
PMCID: PMC10240207  PMID: 37066763

Abstract

Type VI secretion systems (T6SSs) can deliver diverse toxic effectors into eukaryotic and bacterial cells. Although much is known about the regulation and assembly of T6SS, the translocation mechanism of effectors into the periplasm and/or cytoplasm of target cells remains elusive. Here, we use the Agrobacterium tumefaciens DNase effector Tde1 to unravel the mechanism of translocation from attacker to prey. We demonstrate that Tde1 binds to its adaptor Tap1 through the N‐terminus, which harbors continuous copies of GxxxG motifs resembling the glycine zipper structure found in proteins involved in the membrane channel formation. Amino acid substitutions on G39xxxG43 motif do not affect Tde1–Tap1 interaction and secretion but abolish its membrane permeability and translocation of its fluorescent fusion protein into prey cells. The data suggest that G39xxxG43 governs the delivery of Tde1 into target cells by permeabilizing the cytoplasmic membrane. Considering the widespread presence of GxxxG motifs in bacterial effectors and pore‐forming toxins, we propose that glycine zipper‐mediated permeabilization is a conserved mechanism used by bacterial effectors for translocation across target cell membranes.

Keywords: Agrobacterium tumefaciens, DNase effector, glycine zipper, translocation, Type VI secretion system

Subject Categories: Microbiology, Virology & Host Pathogen Interaction; Signal Transduction; Structural Biology

This study describes how the type VI secretion DNase effector Tde1 is translocated from the bacterial attacker into the competitor cell, demonstrating a new and maybe a conserved role of glycine zipper motif(s) in effector delivery.

graphic file with name EMBR-24-e56849-g001.jpg

Introduction

In a complex microbial community, bacteria have evolved versatile secretion systems for the export or import of substrates across their membranes in response to different environmental cues. Each specialized protein secretion system (type I to X secretion system [TISS to TXSS]; reviewed in Costa et al, 2015; Christie, 2019; Palmer et al, 2020) can recognize specific substrates for secretion and translocation across one or multiple membranes. The type VI secretion system (T6SS) is a molecular weapon deployed by many Proteobacteria for pathogenesis, antagonism, or nutrient acquisition (Coulthurst, 2019). The T6SS effectors discovered so far exert functions in antibacterial, anti‐eukaryotic, and metal acquisition (Russell et al, 2014; Hachani et al, 2016; Lien & Lai, 2017; Jurenas & Journet, 2021). The most established T6SS effectors are bacterial toxins, in which bacteria also produce cognate immunity proteins to prevent self‐intoxication and toxicity in the sibling cells.

T6SS is a multiprotein complex, composed of at least 13 conserved core proteins resembling a phage tail structure, that extends from the cytoplasm to the outer membrane of the attacker cell (Cherrak et al, 2019; Wang et al, 2019). The T6SS machine consists of the Tss(J)LM membrane complex (MC), TssEFGK base plate (BP), TssBC contractile sheath, and Hcp‐VgrG‐PAAR puncturing device. The MC interacts with the BP (Durand et al, 2015; Cherrak et al, 2018), which serves as a docking site of VgrG–PAAR–effector complex to initiate the polymerization of the tail (Zoued et al, 2016). The tail is composed of the Hcp inner tube and TssBC outer sheath, whose biogenesis is regulated by TssA cap protein, and when triggered, the sheath contracts and ejects out the effector decorated puncturing device into extracellular milieu or target cells (Basler et al, 2012; Vettiger & Basler, 2016; Ali & Lai, 2022).

The T6SS has multiple strategies for delivering diverse effectors. On the basis of the known effectors and their transport mechanisms, effectors can be classified as “specialized” or “cargo” effectors (Cianfanelli et al, 2016; Cherrak et al, 2019). Specialized effectors are fused to either of the C‐termini of three core structural proteins (Hcp, VgrG, or PAAR) while cargo effectors interact directly or require a specific chaperone/adaptor to be loaded into the lumen of the Hcp tube or onto the VgrG spike prior to secretion. Though diverse T6SS antibacterial effectors that act in the cytoplasm, membrane, or periplasm of the target cells have been reported (Russell et al, 2014; Lien & Lai, 2017; Jurenas & Journet, 2021), their mechanisms to breach outer and inner membranes for targeting cytoplasm of their targets still yet to be clarified.

A glycine zipper structure consisting of repetitive GxxxG motifs is commonly found in membrane‐associated proteins (Kim et al, 2005) and bacterial toxins (Kim et al, 2004; Fonte et al, 2011). Glycine zipper motifs are known to be involved in the toxicity of some bacterial effectors for membrane channel formation. For example, the transmembrane domain (TMD) of a vacuolating toxin, VacA of Helicobacter pylori encodes three GxxxG motifs forming helix–helix packing interactions (Kim et al, 2004), which are required for the vacuolation and membrane channeling contributing to VacA toxicity (McClain et al, 2003). Type I secretion effectors CdzC and CdzD of Caulobacter crescentus and T6SS effector Tse4 of Pseudomonas aeruginosa, also possess glycine zipper motifs involved in the antibacterial activity (Garcia‐Bayona et al, 2017; LaCourse et al, 2018). Expression of Tse4 disrupted the proton motive force of the inner membrane while CdzC and CdzD form surface aggregation for the contact‐dependent killing of target cells. However, how glycine zipper motifs of Tse4 and CdzCD involved in toxicity remains unknown.

A T6SS‐encoding locus is highly conserved in the genome of plant pathogenic bacterium Agrobacterium tumefaciens and the apparatus functions as an antibacterial weapon (Ma et al, 2014; Yu et al, 2020; Wu et al, 2021; Chou et al, 2022). We previously revealed that A. tumefaciens strain C58 deploys two Type VI DNase effectors (Tde1 and Tde2) as the major antibacterial weapons, in which the cognate immunity proteins (namely Tdi1 and Tdi2) prevent autointoxication (Ma et al, 2014). Both Tde1 and Tde2 harbor a C‐terminal Novel toxin 15 (Ntox15) domain (Zhang et al, 2012) containing an HxxD catalytic motif required for its DNase activity (Ma et al, 2014). Tde1 requires its cognate chaperone/adaptor Tap1 for loading onto VgrG1 for secretion (Bondage et al, 2016).

By obtaining the uncoupling Tde1 variants that remain capable of binding to Tap1 for export but are deficient in membrane permeability, translocation, and interbacterial competition, we reveal the secretion and translocation mechanism of Tde1 from the attacker cell to the target cell. We show that the N‐terminal region of Tde1 harboring repetitive glycine zipper motifs is sufficient for interacting with Tap1 for secretion. Once secreted, a conserved glycine zipper motif is necessary for translocation across target cell membranes. This finding demonstrates a new role of glycine zipper motif(s) in effector delivery into target cells.

Results

Tde1 can cause DNase‐independent growth inhibition in Escherichia coli

Our previous study showed that overexpression of Tde1 in A. tumefaciens C58 caused growth inhibition, and the immunity protein Tdi1 only partially protected against this cytotoxicity (Ma et al, 2014). We hypothesized that Tde1 has domains apart from the DNase domain that contributes to its toxicity. In addition to the C‐terminal Ntox15 DNase domain (amino acid 99–247; Ma et al, 2014), Tde1 has a predicted transmembrane domain (TMD, 22–42; Fig 1A). Thus, three fragments of Tde1, that being the N‐terminal, N‐Tde1(1–97), and two C‐terminal regions, C1‐Tde1(50–278), and C2‐Tde1(98–278) were tested for toxicity. To avoid confounding effects by the DNase activity, substitutions of catalytic residues (H190A, D193A) were introduced in the C1‐Tde1 and the full‐length wild‐type (WT) Tde1 to become C1‐Tde1(M) and Tde1(M), respectively (Fig 1A). Ectopic expression in E. coli (DH10B) under an IPTG‐inducible promoter showed that N‐Tde1 was sufficient to inhibit growth (Fig 1B). Tde1(M), but not the C1‐Tde1(M), is growth inhibitory. Although C2‐Tde1 retains the wild‐type DNase catalytic residues, it was not able to inhibit growth, suggesting the N‐terminus is required for the DNase activity. Both C1‐Tde1(M) and C2‐Tde1 are expressed at levels similar to or higher than N‐Tde1 or Tde1(M), indicating that their loss of growth inhibition is not due to the nonexpression of the proteins (Fig EV1A). This evidence suggests the N‐terminal region of Tde1 is sufficient to confer toxicity under the conditions tested and that the C‐terminal DNase domain requires the entire or part of the N‐terminus for it to cause toxicity.

Figure 1. Schematic domain organization, sequence alignment, growth inhibition assay of Tde1.

Figure 1

Open in a new tab
  • A
    Schematic domain organization of Tde1 protein and its variants. The N‐terminal repeated glycine zipper motifs (12–51) overlapping a predicted transmembrane domain (22–42) and Ntox15 DNase domain (99–247) are indicated. Tde1 and its variants with truncation or amino acid substitutions were illustrated.
  • B, C
    (B) Growth inhibition assay of E. coli DH10B cells harboring pTrc200 vector or each of its derivatives expressing Tde1 variants with IPTG induction. (C) Growth inhibition assay of E. coli DH10B cells co‐expressing the Tde1 variants expressed from pTrc200 plasmid and Tdi1 immunity gene expressed from pRL662 plasmid. Growth curve was determined at OD600. Graphs of panels B and C show mean ± SD of three biological replicates (n = 3), each averaged with 3 technical repeats. One‐way ANOVA was used for the analysis of statistical significance followed by the Tukey's multiple comparison. Different letters indicate statistically different groups of strains (P value, 4.6 × 10−5 and 5.19 × 10−8 for panels B and C, respectively).
  • D
    Multiple sequence alignment of N‐Tde1 homologs were presented with highly conserved amino acid residues highlighted in yellow. The bacterial species, strain name, and locus number of Tde1 orthologs (Agrobacterium/Rhizobium) or tape measure proteins (Paraburkholderia/Burkholderia) are indicated on the left and right of aligned sequences. Two conserved glycine residues (G39, G43) subjected to mutagenesis were indicated by the arrows above the sequences.

Source data are available online for this figure.

Figure EV1. Western blot analysis of Tde1 variants and Tdi1 in Escherichia coli growth inhibition assay.

Figure EV1

Open in a new tab
  1. Western blot for the detection of the expression of HA‐tagged Tde1 variants expressed in E. coli. *Other HA‐tagged truncated bands, related to Fig 1B.
  2. Western blot for the detection of Tdi1 from E. coli cells co‐expressing HA‐tagged Tde1 variants and strep‐tagged Tdi1, related to Fig 1C. The loading control is a nonspecific band from the western blot of anti‐strep.

To test whether Tdi1, the immunity protein for the DNase toxicity of Tde1 (Ma et al, 2014), can also neutralize the N‐Tde1 toxicity, the Tde1 variants were co‐expressed with the Tdi1. The result shows that Tdi1 could not rescue the growth inhibition caused by the N‐Tde1 and Tde1(M) (Figs 1C and EV1B). This indicates that Tdi1 cannot neutralize the N‐terminus‐mediated toxicity.

A glycine zipper motif in N‐terminus of Tde1 is required for toxicity and enhanced membrane permeability

To get an insight into the cause of growth inhibition by N‐terminus of Tde1, we used N‐Tde1 region as a query to search against the NCBI nonredundant (nr) database and identified Tde1 homologs encoded in the T6SS gene clusters of Agrobacterium/Rhizobium as well as tape measure proteins (TMP) encoded in genomes of Paraburkholderia/Burkholderia (Figs 1D and EV2A). We noticed the conservation of continuous copies of GxxxG motifs (12–51) in the N‐terminus of Tde1, which resembles the glycine zipper motifs overrepresented in membrane proteins and reported to be involved in the membrane channel formation (Kim et al, 2005). Thus, we hypothesized that these repetitive glycine zipper motifs are involved in membrane permeability and N‐Tde1 toxicity.

Figure EV2. Genetic organizations and domain architecture of Tde homologs.

Figure EV2

Open in a new tab
  1. Genetic organizations of genes encoding representative Tde1 orthologues and Tape Measure Proteins (TMPs) with sequence similarity to the N‐terminus of Tde1. The proteins encoded from the upstream and downstream of tde1 and tmp genes are shown with their identified domain organizations.
  2. Domain architecture of the Ntox15‐containing proteins. Top 10 classes of the Ntox15‐containing proteins are shown with the identifiable domains (not to scale). The number of proteins in each class was indicated on the right based on the information on June 29, 2022. The Agrobacterium tumefaciens Tde1 belonged to the first class where the N‐terminal region lacks an identifiable domain.

To verify the hypothesis, two highly conserved glycine residues at positions 39 and 43 of a glycine zipper motif were substituted with leucine (G39L and G43L), and the resulting N‐Tde1 and Tde1(M) variants were named as N‐Tde1GLGL and Tde1(M)GLGL, respectively. The growth analysis of E. coli DH10B cells by counting viable cells and OD600 measurement showed that both N‐Tde1GLGL and Tde1(M)GLGL lost the ability to cause growth inhibition (Figs 2A and EV3A). Similar results were also observed when they were overexpressed in A. tumefaciens ∆tde1 mutant (Fig EV3B), indicating that the G39xxxG43 glycine zipper motif of Tde1 is required for the observed toxicity.

Figure 2. Growth inhibition and membrane permeabilization assays of glycine zipper mutants.

Figure 2

Open in a new tab
  • A
    Growth inhibition assay of E. coli DH10B cells harboring pTrc200 vector or each of its derivatives expressing Tde1 variants with IPTG‐inducible expression. The growth of E. coli was monitored by CFU counting every 1 h.
  • B, C
    For membrane permeabilization assays, BW25113 WT alone or ∆lacY(pYTA‐lacZ) cells harboring pTrc200 vector or each of its derivatives expressing Tde1 variants were carried out for (B) β‐galactosidase activity assay to determine ONPG uptake, (C) propidium iodide permeability with cells treated with propidium iodide and Hoechst for detection by fluorescence microscope (Scale bar = 5 μm). For the quantification of cells with PI signals, a total of 6 randomly selected images obtained from two biological repeats were used to quantify the number of PI‐stained cells/number of Hoechst‐stained cells as indicated.
  • D
    Bacteriostatic activity assay. E. coli DH10B cells harboring pTrc200 vector or each of its derivatives expressing Tde1 variants were cultured with or without IPTG induction for 1 h. The IPTG‐induced cells were further centrifuged and resuspended in the fresh medium with or without IPTG. Cell density was measured again before continuous growth for additional 1 h.

Data information: Graphs of panels A, B, and D show mean ± SD of three biological replicates (n = 3), each averaged with 3 technical repeats. One‐way ANOVA was used for the analysis of statistical significance followed by the Tukey's multiple comparison. Different letters indicate statistically different groups of strains (P value, 1 × 10−16 and 2 × 10−16 for panels B and D, respectively).

Source data are available online for this figure.

Figure EV3. Growth inhibition assays of Tde1 glycine zipper variants in Agrobacterium tumefaciens and Escherichia coli .

Figure EV3

Open in a new tab
  1. Growth inhibition assay of E. coli DH10B cells harboring pTrc200 vector or each of its derivatives expressing Tde1 variants with IPTG‐inducible expression, monitored by OD600, related to Fig 2A. Graphs show mean ± SD of three biological replicates (n = 3), each averaged with 3 technical repeats.
  2. Growth curve and western blot analyses of A. tumefaciens C58 ∆tde1 carrying pTrc200 or its derivatives expressing HA‐tagged Tde1 variants. The growth curve was detected every 2 h in 523 media supplemented with 1 mM IPTG. Graphs show mean ± SD of three biological replicates (n = 3), each averaged with 3 technical repeats. One‐way ANOVA was used for the analysis of statistical significance followed by the Tukey's multiple comparison. Different letters indicate statistically different groups of strains (P value = 6.47 × 10−11). The proteins collected at the end point (6 h) were analyzed for western blotting with antibodies against HA. Representative results of three biological repeats were shown. Protein markers are indicated in kDa.
  3. Viability assay for E. coli cells derived from the ONPG uptake assay after 1 h IPTG induction, related to Fig 2B.
  4. The growth curve analysis of E. coli cells used for in vivo plasmid DNA degradation assay. The turbidity of E. coli BW25113 expressing Tde1 and its variants carried out for the in vivo plasmid DNA degradation assay was measured. The E. coli cells were supplemented with 0.5% glucose (glu) or 0.2% L‐arabinose (ara) for the repression or induction of Tde1 and its variants. The OD600 values were measured by DEN‐600 photometer (Biosan, Latvia) every hr.

Next, we tested whether N‐Tde1 can increase E. coli inner membrane permeability. To do so, we used the β‐galactosidase activity assay to measure the entry of ortho nitrophenyl galactopyranoside (ONPG; 301 Da) into the cytosol. ONPG normally requires a functional permease LacY to enter into the cytosol but can enter if the inner membrane is permeabilized/compromised (Casteels et al, 1993; Epand et al, 2009). N‐Tde1 and Tde1(M) as well as their glycine zipper substitution variants were expressed in E. coli BW25113∆lacY (Baba et al, 2006) carrying β‐galactosidase (pYTA‐lacZ). The BW25113∆lacY(pYTA‐lacZ) complemented with lacY was used as a positive control. The E. coli cells were induced with IPTG to express Tde1 variants for 1 h and collected for ONPG uptake assay. This time point was chosen because there is no obvious difference in the number of viable cells among the strains tested (Figs 2A and EV3C). The results showed that cells expressing either N‐Tde1 or Tde1(M) had similar β‐galactosidase activity as LacY‐expressing cells. By contrast, cells expressing N‐Tde1GLGL and Tde1(M)GLGL only exhibited background‐level activity as the negative controls (Fig 2B). These results indicate that the N‐Tde1 and Tde1(M) are able to increase membrane permeability depending on the G39xxxG43 motif. The data also suggest that the N‐terminus‐mediated growth inhibition is caused by its ability to enhance inner membrane permeability through glycine zipper motifs.

To further analyze the extent of enhanced membrane permeabilization, cells from the same experiment were normalized to the same OD600 and stained with Hoechst and propidium iodide (PI). Hoechst (616 Da) is a nucleic acid staining dye that is permeable to live Gram‐negative bacterial cells while PI (668.4 Da) can only enter through a compromised inner membrane or dead cells. The PI/Hoechst staining showed strong PI signals in cells expressing N‐Tde1 and Tde1 (M) but no or few signals were detected in cells expressing N‐Tde1GLGL, Tde1(M)GLGL, or vector control, indicating that N‐Tde1 is able to enhance membrane permeability to allow molecules with size 668.4 Da to pass (Fig 2C).

We next determined whether the N‐terminus of Tde1 is bacteriostatic or bactericidal by growth recovery assay (Mariano et al, 2019). E. coli cells were induced with IPTG to express N‐Tde1 or Tde1(M) and after 1 h, washed with fresh media without IPTG for continuous cultivation. We found that growth was recovered when cells were washed of the IPTG inducer, in contrast to the growth inhibition of cells with continuous IPTG induction (Fig 2D). Collectively, the data suggest that the N‐terminus of Tde1 is sufficient to facilitate membrane permeability for bacteriostatic toxicity, and such activity requires the conserved G39xxxG43 glycine zipper motif.

The N‐terminus of Tde1 is necessary and sufficient for Tap1 interaction

Tap1 is the adaptor for loading Tde1 onto VgrG1 (Ma et al, 2014; Bondage et al, 2016). However, the region that Tde1 and Tap1 interact remains undefined. Thus, we performed a co‐immunoprecipitation (co‐IP) assay to identify the specific region of Tde1 that can interact with Tap1 in A. tumefaciens. The HA‐tagged Tde1 variants were expressed in ∆tde1 and anti‐HA agarose bead was used to co‐precipitate the interacting proteins followed by western blotting to detect Tde1 variants and Tap1. The results showed that the N‐Tde1 and Tde1(M) interact with Tap1 but not the C‐terminal variants, C1‐Tde1(M) and C2‐Tde1 (Fig 3A). N‐Tde1GLGL and Tde1(M)GLGL remain capable of interacting with Tap1 (Fig 3A). The results suggest that Tap1 interacts with Tde1 through the N‐terminus and that the G39L and G43L substitution does not affect Tde1–Tap1 interaction.

Figure 3. The N‐terminus of Tde1 is sufficient for interaction with Tap1 and secretion.

Figure 3

Open in a new tab
  1. Co‐immunoprecipitation (Co‐IP) in Agrobacterium tumefaciens. A. tumefaciens C58 ∆tde1 harboring pTrc200 vector or its derivatives expressing HA‐tagged Tde1 variants. Anti‐HA resin was used to co‐precipitate the Tde1 variants and Tap1.
  2. Secretion assay for HA‐tagged Tde1 variants. Western blot for the cellular and extracellular fractions of A. tumefaciens C58 ∆tdei and ∆tdei∆tssK expressing the HA‐tagged Tde1 variants. Hcp secretion was detected as a positive control for active T6SS secretion. Representative western blot results of three biological repeats were shown with antibody against HA, Hcp, or EF‐Tu where EF‐Tu serves as a loading and nonsecreted protein control. Protein markers are indicated in kDa.

Source data are available online for this figure.

The N‐terminus of Tde1 is necessary and sufficient for secretion

Because N‐Tde1 interacts with Tap1, we hypothesized that this region is required for Tde1 secretion. Thus, we performed a secretion assay by expressing the various HA‐tagged Tde1 variants in ∆tdei, a deletion mutant lacking both tde1‐tdi1 and tde2‐tdi2 toxin immunity pairs. Both cellular and extracellular fractions were collected to determine their expression and secretion, respectively. The results showed that all Tde1 variants containing N‐terminus are secreted but not the C‐terminus, C1‐Tde1(M). The secretion is in a T6SS‐dependent manner as secretion was essentially abrogated in ∆tdei∆tssK, which lacks both tdei and tssk encoding the baseplate component. N‐Tde1GLGL and Tde1(M)GLGL are also stably expressed and secreted (Fig 3B). The data suggest that N‐terminus of Tde1 is necessary and sufficient for secretion and that the G39L and G43L substitution does not interfere with the secretion capacity of Tde1. Accordingly, Hcp secretion levels are highly correlated with Tap1–Tde1 interaction and secretion of Tde1 variants (Fig 3B). The data also confirmed the requirement of the Tap1–Tde1 interaction for Tde1 secretion and supported our previous finding that Tde loading onto VgrG is critical for active T6SS secretion (Wu et al, 2020).

G39xxxG43 motif of Tde1 is required for target cell delivery

Because the G39xxxG43 glycine zipper motif located in N‐Tde1 increased the membrane permeability but was not required for secretion, we hypothesized that G39xxxG43 is responsible for inserting Tde1 into the inner membrane and delivering it into the cytoplasm of target cells. Here, we engineered each of Tde1 variants fused to super‐folder green fluorescence protein (sfGFP) with a flexible (GGGS) linker between Tde1 and sfGFP to avoid the Tde1 functional/structural interference by GFP. The sfGFP‐fused Tde1 variants were expressed in A. tumefaciens ∆tdei and ∆tdei∆tssK mutants. The secretion assay results showed that both WT and G39L and G43L substitution of N‐Tde1‐sfGFP and Tde1(M)‐sfGFP are secreted (Fig EV4A). No or trace amounts of proteins were observed in the extracellular fractions of ∆tdei∆tssk mutants, demonstrating that the secretion was T6SS dependent. C1‐Tde1(M)‐sfGFP protein signal could not be unambiguously determined in the cellular fraction due to the overlapping of its predicted protein band with cross‐reacted proteins, and no corresponding C1‐Tde1(M)‐sfGFP band was detected in the extracellular fraction. The secretion assay of Tde1 variants fused with either HA or sfGFP concluded that N‐Tde1 is necessary and sufficient for secretion and the G39L and G43L substitution does not affect Tde1 being secreted, which is correlated with the ability to interact with Tap1.

Figure EV4. Secretion assay for sfGFP‐fused Tde1 variants and fluorescence microscopy for negative controls of translocation assay.

Figure EV4

Open in a new tab
  1. Secretion assay for Tde1 variants fused with sfGFP. Western blot for the cellular and extracellular fractions of Agrobacterium tumefaciens C58 ∆tdei and ∆tdei∆tssK expressing the Tde1 variants fused with sfGFP were detected by anti‐GFP antibody. Representative western blot results of three biological repeats were shown with antibody against GFP, Hcp, or EF‐Tu where EF‐Tu served as a loading and nonsecreted protein control. Hcp secretion served as a positive control for active T6SS secretion. Protein markers are indicated in kDa.
  2. A. tumefaciens C58 ∆tdei∆tssK expressing N‐Tde1‐sfGFP or Tde1(M)‐sfGFP (in green) and E. coli DH10B carrying mCherry (false colored in blue) were co‐cultured for 20 h. No cyan fluorescence with merged blue and green signals could be detected when attacker cells are T6SS‐inactive, which served as negative controls for the translocation assay (Scale bar = 5 μm).

We next investigated the translocation of Tde1 variants by mixing A. tumefaciens ∆tdei, expressing sfGFP‐fused Tde1 variants, with E. coli cells expressing mCherry. A. tumefaciens expressing sfGFP only (Vector‐sfGFP) was used as a negative control. After co‐culture, we imaged populations for mCherry (false colored in blue) and GFP (green) to detect E. coli and A. tumefaciens, respectively. We merged images to identify cyan‐colored cells (overlayed blue and green signals), which represent E. coli cells with translocated Tde1 variants carrying sfGFP (Fig 4A). We were able to observe ~ 50% of cells with cyan fluorescence when A. tumefaciens expressing N‐Tde1‐sfGFP and Tde1(M)‐sfGFP was co‐cultured with E. coli mCherry whereas the GFP and mCherry signals were not overlapped in the E. coli cells co‐cultured with A. tumefaciens strains expressing GFP only or sfGFP‐fused C1‐Tde1(M), N‐Tde1GLGL, Tde1(M)GLGL, respectively (Fig 4A and B). No cyan fluorescence was observed when N‐Tde1‐sfGFP and Tde1(M)‐sfGFP were expressed in the ∆tdei∆tssk mutant as the attacker (Fig EV4B).

Figure 4. Translocation of Tde1 variants fused with sfGFP by Agrobacterium tumefaciens‐Escherichia coli co‐culture.

Figure 4

Open in a new tab
  1. Fluorescence microscopy for Tde1 translocation. A. tumefaciens C58 ∆tdei expressing Tde1 variants fused with sfGFP (in green) and E. coli DH10B carrying mCherry (false colored in blue) were co‐cultured for 20 h. A cyan fluorescence with merged blue and green signals represented the translocation of Tde1 variants from A. tumefaciens to E. coli (Scale bar = 5 μm).
  2. The number of cells with overlayed GFP and mCherry fluorescence was quantified from a total of 6 randomly selected images obtained from three biological repeats (number of cells with cyan fluorescence/total E. coli cells counted).

Source data are available online for this figure.

The data suggest that Tde1 is translocated into target cells in a T6SS‐ and G39xxxG43‐dependent manner. Because N‐Tde1GLGL‐sfGFP and Tde1(M)GLGL‐sfGFP could be secreted but not translocated into target cells, G39xxxG43 motif is necessary for delivering Tde1 into the target cell.

G39xxxG43 is critical for interbacterial competition but not for DNase activity

To assess the role of the G39xxxG43 motif for target cell intoxication in the context of interbacterial competition, A. tumefaciens C58 ∆tdei expressing either Tde1‐Tdi1, Tde1(M)‐Tdi1, or single/double G39L and G43L substitution variants, was competed with target E. coli (DH10B) cells. By counting the survival rate of E. coli prey cells, the data showed that A. tumefaciens ∆tdei (Tde1‐Tdi1) exhibits an antibacterial activity but not in the negative controls, the secretion deficient mutants ∆tssL and ∆tdei∆tssK (Tde1‐Tdi1; Fig 5A). No antibacterial activity could be observed for A. tumefaciens ∆tdei expressing Tde1(M)‐Tdi1, indicating the DNase‐mediated killing of E. coli. The antibacterial activity of ∆tdei(Tde1GLGL‐Tdi1, Tde1G39L‐Tdi1, Tde1G43L‐Tdi1) was not detectable, similar to that of negative controls. We also performed interbacterial competition assays using A. tumefaciens strain 1D1609, which is susceptible to T6SS killing by C58 (Wu et al, 2019). The interbacterial competition between two A. tumefaciens strains was calculated by competitive index, which revealed the higher competitiveness of ∆tdei (Tde1–Tdi1) and C58 against 1D1609 but no competitive advantage could be detectable for any of glycine zipper variants or Tde1(M) (Fig 5B). The observed antibacterial activity is T6SS‐dependent because the killing activity of Tde1 was not observed when expressed in ∆tdei∆tssK. The results indicate that G39xxxG43 motif is required for interbacterial competition at both inter‐ or intra‐species levels. We also performed a secretion assay of these A. tumefaciens attacker strains and all glycine zipper variants were secreted (Fig 5C). It is notable that Tde1GLGL proteins accumulated at slightly lower levels while Tde1G39L and Tde1G43L had similar or even higher protein levels to that of Tde1 and Tde1(M). Accordingly, Tde1GLGL was secreted at lower levels.

Figure 5. G39xxxG43 glycine zipper motif of Tde1 is required for DNase‐mediated killing of target cells during interbacterial competition.

Figure 5

Open in a new tab
  1. Interbacterial competition of Agrobacterium tumefaciens C58 ∆tdei and ∆tdei∆tssK expressing the Tde1 variants against E. coli cells was carried out on LB medium and E. coli survival rate was quantified by CFU counting.
  2. Interbacterial competition between various A. tumefaciens C58 strains and A. tumefaciens 1D1609 on AK medium and the competition outcome was shown by competitive index.
  3. Secretion assay for Tde1 and its variants co‐expressed with its immunity protein Tdi1 in A. tumefaciens C58 ∆tdei and ∆tdei∆tssK.
  4. In vivo plasmid DNA degradation assay. E. coli BW25113 carrying pJN105 empty vector or the derivatives expressing different variants of Tde1 was supplemented with 0.5% glucose (“−”) or 0.2% L‐arabinose (“+”) for 3 h to either repress or induce Tde1 production. The plasmids were then extracted to observe the DNA degradation, and the bottom panel showed western blots of specific Tde1 protein bands.
  5. Growth inhibition assay of Tde1 and its variants. E. coli BW25113 cells were induced by adding 0.2% L‐arabinose for Tde1 production. The OD600 values were measured every 15 min. The OD600 values of the 4 h post‐L‐arabinose induction were analyzed for statistical analysis. Graphs show mean ± SD of three biological repeats.

Data information: Western blots were detected with a specific antibody against Tde1, Hcp, or EF‐Tu serving as a loading and nonsecreted protein control. Protein markers are indicated in kDa. Data in panel A are mean ± SD of four biological repeats of two independent experiments (n = 4). Panels B and E show mean ± SD of three biological repeats (n = 3). One‐way ANOVA was used for the analysis of statistical significance followed by the Fisher's least significant difference (LSD) test for panels A and B while the Tukey's test was done for panel E. Different letters indicate statistically different groups of strains (P value, 3.63 × 10−4, 2.70 × 10−3, 2.3 × 10−15 for panels A, B, and E, respectively). Results in panels C and D are representative of three biological repeats.

Source data are available online for this figure.

To exclude the possibility that G39L and G43L substitution may influence its DNase activity, we performed in vivo plasmid DNA degradation assay. Tde1 and the variants were each expressed by the tightly controlled arabinose‐inducible promoter for in vivo plasmid DNA degradation assay in E. coli BW25113 as described (Ma et al, 2014). It was observed that plasmid DNA was completely degraded in cells expressing Tde1, but not in the negative controls, the cells without arabinose induction nor cells expressing Tde1(M). Plasmid DNA was also degraded by Tde1GLGL but not as complete as Tde1 while both Tde1G39L and Tde1G43L exhibit wild‐type level DNase activity. (Fig 5D). The lower DNA degradation efficiency of Tde1GLGL could be correlated with the barely detected Tde1GLGL proteins (Fig 5D). We also found that the degree of plasmid DNA degradation is also correlated with the growth inhibition effect (Figs 5E and EV3D). The slight recovery of Tde1GLGL from growth inhibition is consistent with the instability of Tde1GLGL. The evidence that G39L and G43L substitutions abolished interbacterial competition but did not affect DNase activity and secretion of Tde1 suggest the G39xxxG43 motif is necessary for delivering Tde1 across the inner membrane into the cytoplasm of target cells.

Discussion

Through the dissection of Tde1 DNase effector, we provide strong evidence for a role of the N‐terminal glycine zipper motif(s) of Tde1 in delivering the T6SS effector into target cells. Here, we propose a model explaining the loading, firing, and translocation of Tde1 (Fig 6). In A. tumefaciens, Tde1 DNase activity is neutralized by Tdi1 by binding to C‐terminal DNase domain while its N‐terminal domain interacts with Tap1 for loading onto VgrG1 (Step 1). The VgrG1–Tap1–Tde1–Tdi1 complex is then recruited onto the membrane‐associated baseplate, which serves as a docking site for polymerization of Hcp tube and TssBC sheath (Step 2). Upon TssBC sheath contraction (Step 3), Tap1 and Tdi1 may fall off and Hcp‐VgrG‐Tde1 puncturing device is then ejected for secretion. In contact with a target cell, Tde1 may be delivered to the periplasm of the target cell where Tde1 permeabilizes the inner membrane in a G39xxxG43 motif‐dependent manner (Step 5). Once delivered, Tde1 exerts its toxicity by attacking DNA for degradation (Step 6).

Figure 6. Proposed model of the loading, firing, and translocation of Tde1.

Figure 6

Open in a new tab

The Tde1 translocation is proposed through six steps. Step 1: Tde1 forms a complex with Tdi1 and Tap1 in the attacker cell. Step 2: Tap1–Tde1–Tdi1 complex binds to the VgrG and the Hcp‐VgrG‐PAAR puncturing device carrying Tde1–effector complex is loaded onto the membrane‐associated baseplate. Step 3: Hcp tube and TssB/C sheath polymerize on the Tde1‐loaded VgrG/baseplate while Tdi1 and Tap1 fall off with unknown mechanisms before or upon firing. Step 4: TssBC sheath contracts and ejects Tde1 into the target cell periplasm or cytoplasm. Step 5: The glycine zipper(s) on the N‐terminus of Tde1 permeabilize the target cell membrane. Step 6: Intact or truncated Tde1 proteins attack DNA for degradation in the target cell.

T6SS cargo effectors often require the specific chaperone/adaptor for loading onto the puncturing device for secretion. Our previous findings demonstrated that Tap1, a DUF4123‐containing protein, specifically interacts with Tde1 for loading onto VgrG1 for secretion (Ma et al, 2014; Bondage et al, 2016). We now show that the N‐terminal region of Tde1 is necessary and sufficient for interaction with Tap1 for secretion and delivery into target cells. The evidence that Tde1GLGL variant remains capable of binding to Tap1 for export but is deficient in membrane permeability and translocation demonstrates a distinct role of this G39xxxG43 motif in target cell delivery. Among the 10 classes of the Ntox15‐containing proteins, the majority of them including Tde1 belong to class I without detectable N‐terminal domains (Fig EV2B). We identified the presence of glycine zipper motifs overlapping with the transmembrane domain (TMD) not only in N‐terminal region of all Tde1 orthologs encoded in Rhizobiaceae but also in C‐terminal region of tape measure proteins (TMP) encoded in genomes of Paraburkholderia/Burkholderia, likely as a prophage. TMP is a phage protein suggested to have a channel‐forming activity (Roessner & Ihler, 1984, 1986) and as a determinant in connecting host inner membrane proteins for injecting phage genome into bacterial host cells (Cumby et al, 2015). Such conservation in Tde1 orthologs suggests that this glycine zipper‐mediated delivery could be a common strategy deployed by these bacterial effectors for translocation across target cell membranes. It would be also interesting to investigate whether TMP also employs its C‐terminal glycine zipper to mediate phage genome entry into host cells.

A role of N‐terminal domain involved in the translocation of polymorphic toxins has been well documented in those contact‐dependent growth inhibition (CDI) system and bacteriocins (Ruhe et al, 2020). However, little is known about the translocation mode of bacterial toxins delivered by other systems. Previous study in P. aeruginosa showed that VgrG‐loaded Tse6–EagT6 complex is sufficient to translocate across a lipid bilayer in vitro (Quentin et al, 2018), suggesting a role of VgrG–effector complex itself in inserting across the inner membrane of target cells. A recent study further uncovered a widespread prePAAR motif in N‐terminal TMDs of T6SS effectors involved in interaction with Eag family chaperone for export (Ahmad et al, 2020). The findings from the Tap1 and Eag chaperone‐mediated T6SS toxins led us to propose that the bacterial toxins harboring a N‐terminal TMD may be protected by its cognate chaperone/adaptor from insertion into membranes in the attacker cell. However, once the effector is injected into the periplasm of the target cell, specific motifs (such as glycine zippers or perhaps prePAAR) may insert into the inner membrane for the delivery into the cytoplasm. By an elegant in vitro translocation assay, a recent study discovered a N‐terminal domain of a bacteriocin pyocin G is required for the import of its C‐terminal nuclease toxin into the cytoplasm cross inner membrane (Atanaskovic et al, 2022). This inner membrane translocation domain (IMD) is distinct from the glycine zipper repeats identified in this study but also found conserved in other bacterial toxins including some of T6SS. Thus, a bacterial toxin directing its own translocation into target cells could be a general strategy used by bacteria for interbacterial competition.

A few membrane‐permeabilizing T6SS toxins have been reported. The Vibrio cholerae VasX causes dissipation of membrane potential, leading to membrane permeabilization of target bacterial cells similar to the Tme effectors of V. parahaemolyticus, which represents a widespread family of T6SS effectors harboring C‐terminal TMD for membrane disruption (Miyata et al, 2013; Fridman et al, 2020). On the other hand, Tse4 disrupts the membrane potential and forms a cation‐selective pore without membrane permeabilization where the pore cannot even allow the permeability of a relatively smaller molecular weight (ONPG, 300 Da; LaCourse et al, 2018). Distinct from these toxins in which they confer pore‐forming activity for toxicity, the role of glycine zipper(s) of Tde1 appears to enhance membrane permeability for bringing the toxin domain into target cell cytoplasm because Tde1(M) with complete glycine zipper motifs but the loss of DNase activity did not show interbacterial competition activity against E. coli or A. tumefaciens under conditions tested (Fig 5A and B; Ma et al, 2014).

To date, no structural information is available for Ntox15 superfamily proteins where Tde1 belongs. While N‐terminus of Tde1 lacks sequence similarity to any of those known pore‐forming toxins, structural similarity to pyocin S5 and colicin Ia could be predicted by Phyre2 (Fig EV5A). Further structural modeling showed the structural similarity of two helixes containing consecutive glycine zipper motifs of N‐Tde1 (10–62) to the pore‐forming domain of pyocin S5 (Behrens et al, 2020; Fig EV5B–D). Pyocin S5 can cause ATP leakage and PI permeability (Ling et al, 2010) potentially to the inner membrane after translocation through FptA and TonB1 (Behrens et al, 2020). Tde1 allows the passage of a relatively larger molecule, PI (668 Da), suggesting that its N‐terminal glycine zipper(s) may form larger pores similar to pyocin S5. G39xxxG43 motif plays no role in DNase activity of Tde1 but is crucial for its protein stability. Tde1 with the substitution of one of the two glycine residues to leucine retains the stability of intact proteins, but Tde1 is prone to truncations and degradation when both glycine residues are substituted to leucine. The instability is most evident when ectopically expressed in E. coli and when retaining DNase activity (Figs 5 and EV1A). Single glycine substitution (Tde1G39L and Tde1G43L variants) does not influence protein stability may suggest that the adjacent glycine residues (G35 or G47) are sufficient to compensate the loss of one glycine of G39xxxG43 motif for structural integrity in both variants. The importance of G39xxxG43 motif in Tde1 protein stability is consistent with the role of glycine zippers in structural impact (Kim et al, 2005). However, the integrity of G39xxxG43 motif is critical for interbacterial competition because both Tde1G39L and Tde1G43L variants do not exhibit detectable antibacterial activity to either E. coli or A. tumefaciens (Fig 5A and B). These results suggest the role of G39xxxG43 motif in delivering Tde1 across the inner membrane into the cytoplasm of target cells.

Figure EV5. Structural prediction of the Tde1 N‐terminus with similarity to pyocin S5 and colicins.

Figure EV5

Open in a new tab
  1. Predicted results of N‐terminal Tde1 (1–97) as a query reveal structural similarity to pyocin S5 and colicin Ia with high confidence.
  2. N‐terminal Tde1 with structural similarity to pore‐forming domain of the pyocin S5, colicines, and other membrane perturbing proteins based on Phyre2 prediction.
  3. Cartoon model of the Tde1 (residue 10–62) by using on the basis of the crystal structure of Pyocin S5 (PDB 6THK) with 77.3% of confidence level. All glycine residues of the predicted glycine zipper motif of Tde1 were indicated.
  4. Superimposition of N‐Tde1 and pore‐forming domain of pyocin S5. Tde1 N‐terminus is in red, and the partially pore‐forming domain of pyocin S5 is in teal; G39 and G43 in the putative glycine zipper motif are highlighted in green.

Data information: All data were analyzed by Phyre2 server.

It is striking to observe such a high percentage of cells (~ 50%) representing N‐Tde1‐sfGFP and Tde1(M)‐sfGFP translocation from A. tumefaciens into E. coli cells expressing mCherry (Fig 4). Adding the flexible GGGS linker between sfGFP and Tde1 that retain both Tde1 secretion activity and GFP fluorescence may be the key to the success of this translocation experiment. Interestingly, we also observed many GFP foci from A. tumefaciens cells expressing translocation‐competent N‐Tde1‐sfGFP or Tde1(M)‐sfGFP while others including E. coli cells with GFP signals were found to be uniformly distributed throughout the cells. Based on the role of glycine zippers in interacting with membrane, we propose that Tde1 proteins may preferentially bind to the microdomain of the cytoplasmic membrane, which was recently found in A. tumefaciens (Czolkoss et al, 2021). We also found that Tde1 proteins (either tagged with HA or GFP, Figs EV1, 3, and EV4A) are prone to truncation especially when they are ectopically expressed in E. coli or when Tdi1 is absent or not equivalent. Thus, it is possible that Tde1‐GFP proteins are truncated after translocation into E. coli cells, in which most GFP signals are emitted from free GFP instead of Tde1‐GFP. The stability of free GFP derived from translocated Tde1‐GFP may also explain the high percentage of E. coli cells exhibiting overlayed GFP/mCherry signals. There is evidence that the truncation of T6SS effectors is critical for toxicity (Pei et al, 2020). Future work to investigate how Tde1 interacts with membrane and dissects the region required for DNase activity shall shed light to understand the biological significance and mechanisms underlying this interesting observation.

With the knowledge of effector translocation mechanisms, the bacterial protein secretion apparatus also offers a strategy for delivering heterologous proteins to specific cells. T6SS is a promising vehicle for such purpose because effectors or secreted proteins appear to be delivered with their folded or partially folded form, unlike those to be transported as unfolded forms in most of the other specialized secretion systems (Costa et al, 2015). Engineering T6SS carriers such as VgrG spikes to carry exogenous effector proteins into target cells are feasible but challenging (Ho et al, 2017; Wettstadt et al, 2020; Wettstadt & Filloux, 2020). By using a truncated variant of PAAR, a recent study showed delivering exogenous T6SS effectors and Cre recombinase for genetic modification in the target cells (Hersch et al, 2021). Our success in using N‐Tde1 in the delivery of sfGFP proteins into target E. coli cells where they exert fluorescence also suggests potential applications of N‐Tde1 for the delivery of proteins of interest such as genetic modifiers. This strategy provides advantages over transforming foreign DNA for expressing a protein of interest from creating undesired genome manipulation.

Materials and Methods

Strains and growth conditions

The strains and plasmids used in this study are listed in Appendix Tables S1 and S2. The E. coli strains used in this study are BW25113 and DH10B. All the A. tumefaciens strains were cultured on 523 medium (Kado & Heskett, 1970) at 28°C unless stated. The E. coli strains were cultured on Luria Bertani (LB) medium (10 g L−1 NaCl, 10 g L−1 tryptone, and 5 g L−1 yeast extract) at 37°C unless stated. Where appropriate, the media were supplemented with 100 μg ml−1 spectinomycin (Sp), gentamycin (Gm) 25 μg ml−1 (for E. coli) and 50 μg ml−1 (for A. tumefaciens), 50 μg ml−1 ampicillin (Amp), 50 μg ml−1 kanamycin (Km), 1 mM Isopropyl β‐d‐1‐thiogalactopyranoside (IPTG).

Growth inhibition assay

For growth inhibition assay of IPTG‐inducible expression of Tde1 and its variants, E. coli (DH10B) harboring pTrc200 vector or the derivatives expressing Tde1 variants were grown overnight in LB medium supplemented with Sp prior to 1:30 dilution in a fresh medium and incubated for 2 h at 37°C with 250 rpm. After 2 h, the cultures were normalized to OD600 0.1 in LB with 1 mM IPTG for continuous culture in the same growth condition. The growth of E. coli was monitored for OD600 every 1 h using ULTROSPEC® 10 cell density meter (Biochrom, UK) or viable cell by counting colony forming units (CFUs) on LB agar containing Sp. For growth inhibition assay of arabinose‐inducible expression, E. coli BW25113 harboring pJN105 vector or the derivatives expressing Tde1 variants were used. Overnight cultures of E. coli cells were adjusted to an OD600 of 0.1 in 200 μl LB with 0.2% L‐arabinose in a 96‐well plate. The OD600 values were measured by the Synergy H1 microplate reader (Agilent Technologies, USA) with agitation at 37°C. The OD600 values or CFUs of indicated time points were used to calculate mean ± SD of three biological repeats. One‐way analysis of variance (ANOVA) was used for the analysis of statistical significance followed by the Tukey's multiple comparison.

In vivo plasmid DNA degradation assay

The in vivo plasmid DNA degradation assay was performed as described (Ma et al, 2014) with minor modifications. Briefly, overnight cultures of E. coli BW25113 carrying pJN105 vector or the derivatives expressing Tde1 variants were adjusted to an OD600 of 0.3 in 4 ml LB with 0.5% D‐glucose or 0.2% L‐arabinose. After induction for 3 h, bacterial cells normalized by OD600 were collected for plasmid DNA extraction and western blot analysis. The plasmids were then extracted and applied to 0.6% agarose gel electrophoresis to detect DNA degradation. The OD600 values were measured by DEN‐600 photometer (Biosan, Latvia) every hour.

β‐Galactosidase and viability assays for ONPG update

β‐galactosidase assay was performed as described (Saint Jean et al, 2018) with minor modifications. BW25113 wild‐type, BW25113∆lacY(pYTA‐lacZ), or BW25113∆lacY harboring pTrc200 vector or the derivatives expressing Tde1 variants were grown overnight and refreshed to a fresh medium as stated for growth inhibition assay. After subculture for 2 h, the cells were induced with 1 mM IPTG, and incubated for one more hr. Part of the culture was adjusted to OD600 = 0.3 in Z‐buffer and the Intracellular β‐galactosidase activity was measured by mixing 100 μl of 4 mg ml−1 ONPG with 900 μl of the cells and incubation at room temperature for 10 min prior to measurement at OD420. The remaining cells were normalized to OD600 0.3 in 0.9% sterile saline and after serial dilution, 10 μl was spotted on the LB plate without antibiotics to recover the viable cells. Data from OD420 were used to calculate mean ± SD of three independent experiments. One‐way ANOVA was used for the analysis of statistical significance followed by the Tukey's multiple comparison.

Co‐immunoprecipitation (Co‐IP)

The co‐IP was performed according to the manufacturer's recommendations of EZview red Anti‐HA agarose (Sigma‐E6779) with minor modifications. To identify Tap1‐interacting domain of Tde1, the HA‐tagged Tde1 variants were expressed from pTrc200 plasmid. For co‐IP in A. tumefaciens, C58 ∆tde1 cells expressing the Tde1 variants grown in 523 medium overnight were resuspended in a 1:30 ratio to a fresh medium and incubated at 25°C for 3 h followed by 1 mM IPTG induction for additional 3 h. After 6 h postincubation, the cells were normalized to OD600 of 5 per ml in ice‐cold PBS buffer (pH 7.4). After cell lysis by lysozyme treatment and sonication, the lysate was centrifuged and a 100 μl aliquot of the lysate was saved for the input fraction. The remaining 900 μl lysate was mixed with 25 μl of pre‐equilibrated Ezview red Anti‐HA agarose and incubated at 4°C for 1 h. The beads were then washed 3 times with ice‐cold PBS buffer and the proteins bound to the beads were eluted with 100 μl of 2× SDS sample loading buffer. Similarly, the aliquoted input fraction was mixed with an equal volume of 2× SDS sample loading buffer for analysis by western blotting.

Secretion assay

Type VI secretion assay was performed in 523 medium as described previously (Wu et al, 2020). Briefly, A. tumefaciens strain was cultured overnight in 523 medium and normalized to OD600 of 0.2 in a fresh medium. After 6 h of culturing, the secreted proteins were collected by centrifuging at 10,000 g for 5 min. The resulting pellet was adjusted to OD600 of 10 as a cellular fraction. The culture supernatant was filtered with 0.22 μm Millipore filter membrane, and the resulting filtrate was subjected to TCA precipitation (Wu et al, 2008) and referred to as an extracellular fraction.

Western blotting

Western blot analyses were done as previously described (Lin et al, 2013). The following primary antibody titres used were: HA epitope (1:4,000), Tap1 (1:3,000; Lin et al, 2013), Strep (1:4,000), EF‐Tu (1:6,000), and C‐terminal Tde1 (1:4,000; Ma et al, 2014; Bondage et al, 2016), Hcp (1:2,500; Wu et al, 2008).

Interbacterial competition assays

For interbacterial competition with E. coli (target), A. tumefaciens strain (attacker) was grown overnight at 28°C in 523 broth with appropriate antibiotics if needed. E. coli DH10B harboring pRL662 plasmid was grown at 37°C in LB broth with Gm. After harvesting and washing the cells in 0.9% saline, the attacker to target cell density was adjusted to 30:1 (OD600 = 3: 0.1) and the mix was spotted on LB medium containing 1.5% (wt/vol) agar. After incubation of the mixed strains for 16 h at 28°C, the spot was resuspended in 0.9% saline, serial diluted, and spotted on a gentamycin‐containing LB agar square plate at 37°C to calculate E. coli survival rate by CFU counts. Similar procedure was used when using A. tumefaciens strain 1D1609 as a target, which was grown at 28°C in 523 broth prior to competition. The competition was carried out on AK medium for 16 h at 28°C with CFU counting at both initial and final time points by the selection of C58 strains with Sp resistance and 1D1609 with Gm resistance. To calculate the competitive index, CFUs of A. tumefaciens attacker C58 strain were divided by the CFUs of target 1D1609 strain at both 0 h and 16 h, and the ratio of C58/1D1609 at 16 h was divided by the ratio of C58/1D1609 at 0 h to obtain competitive index. One‐way ANOVA was used for the analysis of statistical significance followed by the Fisher's least significant difference (LSD) test.

Fluorescence microscopy

For propidium iodide and Hoechst staining, E. coli cells (BW25113) harboring pTrc200 vector or derivatives expressing Tde1 variants were grown overnight and refreshed to a fresh medium as stated for growth inhibition assay. After subculture for 2 h, the cells were induced with 1 mM IPTG for 1 h and OD600 equivalent to 0.3 was collected in 1 ml PBS and stained with Hoechst 33342 (H3570) to a final concentration of 12.3 and 1 μg ml−1 propidium iodide (2208511) and incubated for 2 min in dark. Finally, 2 μl was spotted on 2.5% agarose pad.

For the translocation experiment, the sfGFP‐fused Tde1 variants were expressed in A. tumefaciens ∆tdei cells (attacker). E. coli (target) cells were labeled with mCherry (false color blue) expressed from pBBRMCS2. A. tumefaciens attacker cells were cultured in 523 broth overnight, and E. coli target cells were separately cultured on LB broth. Overnight cultured attacker and target cells were mixed at a 5:1 ratio (OD600 = 1.0:0.2), and 10 μl of the mix was cultured on an LB agar plate without IPTG. After 20 h of co‐culture, the cells were washed with 100 μl PBS and 2 μl of suspension was spotted on the 2.5% agarose pad on a microscopic slide. The translocation signal was detected as the merge of GFP and mCherry (false colored in blue), which is of cyan fluorescence.

Fluorescence microscopy was performed using Axio Observer 7 (Zeiss, Germany) microscope equipped with an Axiochem 702 digital camera and a Plan‐Apochromat 100×/1.4 Oil DIC H objective. Exposure times were adjusted to 20 ms for Phase, 50 ms for Hoechst, 150 ms for PI, 200 ms for GFP, and 5,000 ms for mCherry. Multiple images were taken from different fields and all the experiments were performed at least in triplicate and a representative image is shown. Images were analyzed by using ZEN 2.3 (blue edition) software.

Domain prediction and analysis

Full‐length Tde1 (1–278) was used as a query for conserved domain search on the conserved domains database (CDD; Lu et al, 2020) of the National Center for Biotechnology Information (NCBI). Prediction of the transmembrane domain was done using the PRED‐TMR2 (Pasquier et al, 1999). The Tde1 homologs and tape measure proteins (TMPs) for the multiple sequence alignment were obtained by BLAST search of N‐Tde1 (1–97) against the NCBI nonredundant database (nr) with representative sequences selected for multiple sequence alignment. The domain architectures of the Ntox15 domain‐containing proteins were obtained using the full‐length Tde1 against the Conserved Domain Architectural retrieval tool (CDART) of NCBI. The information of gene clusters encoding Tde1 homologs and TMPs including upstream and downstream three genes was retrieved from their respective genomes. N‐Tde1(1–97) was used as a query for structural prediction on a Phyre2 (Kelley et al, 2015). Three‐dimensional structure modeling was done using Phyre2 in intensive modeling mode. Crystal structure served as the best template for the N‐terminus, and the percentage of confidence for three‐dimensional structure modeling is indicated in the legends of corresponding figures. The structural graphics were generated by using ChimeraX 1.1 (Goddard et al, 2018).

Author contributions

Jemal Ali: Conceptualization; investigation; writing – original draft. Manda Yu: Conceptualization; investigation; writing – review and editing. Li‐Kang Sung: Investigation; writing – review and editing. Yee‐Wai Cheung: Investigation; writing – review and editing. Erh‐Min Lai: Conceptualization; resources; supervision; funding acquisition; writing – original draft; project administration.

Disclosure and competing interest statement

The authors declare that they have no conflict of interest.

Supporting information

Appendix

Click here for additional data file. (173.3KB, pdf)

Expanded View Figures PDF

Click here for additional data file. (3.3MB, pdf)

PDF+

Click here for additional data file. (10.2MB, pdf)

Source Data for Figure 1

Click here for additional data file. (15.1KB, zip)

Source Data for Figure 2

Click here for additional data file. (160.4MB, zip)

Source Data for Figure 3

Click here for additional data file. (15.2MB, zip)

Source Data for Figure 4

Click here for additional data file. (40.3MB, zip)

Source Data for Figure 5

Click here for additional data file. (42.1MB, zip)

Acknowledgements

We would like to thank the former and current Lai lab members for their help and fruitful discussion throughout this study, Yun‐Wei Lien for generating the ∆tdei∆tssK strain, and Chih‐Horng Kuo for discussion and suggestion on BLAST analysis. The authors highly appreciate Jeff Chang, See‐Yeun Ting, Lay‐Sun Ma, and Chih‐Feng Wu for critically reading the manuscript and their valuable comments. We also acknowledge the technical assistance of the fluorescence microscope provided by Mei‐Jane Fang from Live Cell Imaging Division of Cell Biology Core and Sanger DNA sequencing service provided by Genomic Technology Core, both located at the Institute of Plant and Microbial Biology, Academia Sinica. The authors also thank Ying Wang for the illustration of the proposed model. The funding was supported by the National Science and Technology Council (NSTC) of Taiwan (NSTC 110‐2311‐B‐001‐032‐MY3) and the Academia Sinica Investigator Award (AS‐IA‐107‐L01) to EML. YWC was supported by the postdoctoral fellowship (NSCT 110‐2811‐B‐001‐524). The funders had no role in study design, data collection, and interpretation, or the decision to submit the work for publication.

EMBO reports (2023) 24: e56849

Data availability

No large primary datasets have been generated and deposited.

References

  1. Ahmad S, Tsang KK, Sachar K, Quentin D, Tashin TM, Bullen NP, Raunser S, McArthur AG, Prehna G, Whitney JC (2020) Structural basis for effector transmembrane domain recognition by type VI secretion system chaperones. Elife 9: e62816 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ali J, Lai EM (2022) Distinct TssA proteins converge in coordinating tail biogenesis of the type VI secretion systems. Bioessays 45: e2200219 [DOI] [PubMed] [Google Scholar]
  3. Atanaskovic I, Sharp C, Press C, Kaminska R, Kleanthous C (2022) Bacterial competition systems share a domain required for inner membrane transport of the Bacteriocin Pyocin G from Pseudomonas aeruginosa . mBio 13: e0339621 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M, Wanner BL, Mori H (2006) Construction of Escherichia coli K‐12 in‐frame, single‐gene knockout mutants: the Keio collection. Mol Syst Biol 2: 2006.0008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Basler M, Pilhofer M, Henderson GP, Jensen GJ, Mekalanos JJ (2012) Type VI secretion requires a dynamic contractile phage tail‐like structure. Nature 483: 182–186 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Behrens HM, Lowe ED, Gault J, Housden NG, Kaminska R, Weber TM, Thompson CMA, Mislin GLA, Schalk IJ, Walker D et al (2020) Pyocin S5 import into Pseudomonas aeruginosa reveals a generic mode of Bacteriocin transport. mBio 11: e03230‐19 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. 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–E3940 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Casteels P, Ampe C, Jacobs F, Tempst P (1993) Functional and chemical characterization of Hymenoptaecin, an antibacterial polypeptide that is infection‐inducible in the honeybee (Apis mellifera). J Biol Chem 268: 7044–7054 [PubMed] [Google Scholar]
  9. Cherrak Y, Rapisarda C, Pellarin R, Bouvier G, Bardiaux B, Allain F, Malosse C, Rey M, Chamot‐Rooke J, Cascales E et al (2018) Biogenesis and structure of a type VI secretion baseplate. Nat Microbiol 3: 1404–1416 [DOI] [PubMed] [Google Scholar]
  10. Cherrak Y, Flaugnatti N, Durand E, Journet L, Cascales E (2019) Structure and activity of the type VI secretion system. Microbiol Spectr 7 10.1128/microbiolspec.PSIB-0031-2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Chou L, Lin YC, Haryono M, Santos MNM, Cho ST, Weisberg AJ, Wu CF, Chang JH, Lai EM, Kuo CH (2022) Modular evolution of secretion systems and virulence plasmids in a bacterial species complex. BMC Biol 20: 16 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Christie PJ (2019) The rich tapestry of bacterial protein translocation systems. Protein J 38: 389–408 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cianfanelli FR, Monlezun L, Coulthurst SJ (2016) Aim, load, fire: the type VI secretion system, a bacterial Nanoweapon. Trends Microbiol 24: 51–62 [DOI] [PubMed] [Google Scholar]
  14. Costa TR, Felisberto‐Rodrigues C, Meir A, Prevost MS, Redzej A, Trokter M, Waksman G (2015) Secretion systems in gram‐negative bacteria: structural and mechanistic insights. Nat Rev Microbiol 13: 343–359 [DOI] [PubMed] [Google Scholar]
  15. Coulthurst S (2019) The type VI secretion system: a versatile bacterial weapon. Microbiology 165: 503–515 [DOI] [PubMed] [Google Scholar]
  16. Cumby N, Reimer K, Mengin‐Lecreulx D, Davidson AR, Maxwell KL (2015) The phage tail tape measure protein, an inner membrane protein and a periplasmic chaperone play connected roles in the genome injection process of E. coli phage HK97. Mol Microbiol 96: 437–447 [DOI] [PubMed] [Google Scholar]
  17. Czolkoss S, Safronov X, Rexroth S, Knoke LR, Aktas M, Narberhaus F (2021) Agrobacterium tumefaciens type IV and type VI secretion systems reside in detergent‐resistant membranes. Front Microbiol 12: 754486 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Durand E, Nguyen VS, Zoued A, Logger L, Pehau‐Arnaudet G, Aschtgen MS, Spinelli S, Desmyter A, Bardiaux B, Dujeancourt A et al (2015) Biogenesis and structure of a type VI secretion membrane core complex. Nature 523: 555–560 [DOI] [PubMed] [Google Scholar]
  19. Epand RF, Sarig H, Mor A, Epand RM (2009) Cell‐wall interactions and the selective bacteriostatic activity of a miniature oligo‐acyl‐lysyl. Biophys J 97: 2250–2257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Fonte V, Dostal V, Roberts CM, Gonzales P, Lacor PN, Velasco PT, Magrane J, Dingwell N, Fan EY, Silverman MA et al (2011) A glycine zipper motif mediates the formation of toxic beta‐amyloid oligomers in vitro and in vivo . Mol Neurodegener 6: 61 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Fridman CM, Keppel K, Gerlic M, Bosis E, Salomon D (2020) A comparative genomics methodology reveals a widespread family of membrane‐disrupting T6SS effectors. Nat Commun 11: 1085 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Garcia‐Bayona L, Guo MS, Laub MT (2017) Contact‐dependent killing by Caulobacter crescentus via cell surface‐associated, glycine zipper proteins. Elife 6: e24869 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Goddard TD, Huang CC, Meng EC, Pettersen EF, Couch GS, Morris JH, Ferrin TE (2018) UCSF ChimeraX: meeting modern challenges in visualization and analysis. Protein Sci 27: 14–25 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hachani A, Wood TE, Filloux A (2016) Type VI secretion and anti‐host effectors. Curr Opin Microbiol 29: 81–93 [DOI] [PubMed] [Google Scholar]
  25. Hersch SJ, Lam L, Dong TG (2021) Engineered type six secretion systems deliver active exogenous effectors and Cre recombinase. mBio 12: e0111521 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Ho BT, Fu Y, Dong TG, Mekalanos JJ (2017) Vibrio cholerae type 6 secretion system effector trafficking in target bacterial cells. Proc Natl Acad Sci U S A 114: 9427–9432 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Jurenas D, Journet L (2021) Activity, delivery, and diversity of type VI secretion effectors. Mol Microbiol 115: 383–394 [DOI] [PubMed] [Google Scholar]
  28. Kado CI, Heskett MG (1970) Selective media for isolation of Agrobacterium, Carynebacterium, Erwinia, Pseudomonas, and Xanthomonas . Phytopathology 60: 969–976 [DOI] [PubMed] [Google Scholar]
  29. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJ (2015) The Phyre2 web portal for protein modeling, prediction and analysis. Nat Protoc 10: 845–858 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kim S, Chamberlain AK, Bowie JU (2004) Membrane channel structure of helicobacter pylori vacuolating toxin: role of multiple GXXXG motifs in cylindrical channels. Proc Natl Acad Sci U S A 101: 5988–5991 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kim S, Jeon TJ, Oberai A, Yang D, Schmidt JJ, Bowie JU (2005) Transmembrane glycine zippers: physiological and pathological roles in membrane proteins. Proc Natl Acad Sci U S A 102: 14278–14283 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. LaCourse KD, Peterson SB, Kulasekara HD, Radey MC, Kim J, Mougous JD (2018) Conditional toxicity and synergy drive diversity among antibacterial effectors. Nat Microbiol 3: 440–446 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lien YW, Lai EM (2017) Type VI secretion effectors: methodologies and biology. Front Cell Infect Microbiol 7: 254 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lin JS, Ma LS, Lai EM (2013) Systematic dissection of the Agrobacterium type VI secretion system reveals machinery and secreted components for subcomplex formation. PLoS One 8: e67647 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ling H, Saeidi N, Rasouliha BH, Chang MW (2010) A predicted S‐type pyocin shows a bactericidal activity against clinical Pseudomonas aeruginosa isolates through membrane damage. FEBS Lett 584: 3354–3358 [DOI] [PubMed] [Google Scholar]
  36. Lu S, Wang J, Chitsaz F, Derbyshire MK, Geer RC, Gonzales NR, Gwadz M, Hurwitz DI, Marchler GH, Song JS et al (2020) CDD/SPARCLE: the conserved domain database in 2020. Nucleic Acids Res 48: D265–D268 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ma LS, Hachani A, Lin JS, Filloux A, Lai EM (2014) Agrobacterium tumefaciens deploys a superfamily of type VI secretion DNase effectors as weapons for interbacterial competition in planta. Cell Host Microbe 16: 94–104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mariano G, Trunk K, Williams DJ, Monlezun L, Strahl H, Pitt SJ, Coulthurst SJ (2019) A family of type VI secretion system effector proteins that form ion‐selective pores. Nat Commun 10: 5484 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. McClain MS, Iwamoto H, Cao P, Vinion‐Dubiel AD, Li Y, Szabo G, Shao Z, Cover TL (2003) Essential role of a GXXXG motif for membrane channel formation by Helicobacter pylori vacuolating toxin. J Biol Chem 278: 12101–12108 [DOI] [PubMed] [Google Scholar]
  40. Miyata ST, Unterweger D, Rudko SP, Pukatzki S (2013) Dual expression profile of type VI secretion system immunity genes protects pandemic Vibrio cholerae . PLoS Pathog 9: e1003752 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Palmer T, Finney AJ, Saha CK, Atkinson GC, Sargent F (2020) A holin/peptidoglycan hydrolase‐dependent protein secretion system. Mol Microbiol 115: 345–355 [DOI] [PubMed] [Google Scholar]
  42. Pasquier C, Promponas VJ, Palaios GA, Hamodrakas JS, Hamodrakas SJ (1999) A novel method for predicting transmembrane segments in proteins based on a statistical analysis of the SwissProt database: the PRED‐TMR algorithm. Protein Eng 12: 381–385 [DOI] [PubMed] [Google Scholar]
  43. Pei TT, Li H, Liang X, Wang ZH, Liu G, Wu LL, Kim H, Xie Z, Yu M, Lin S et al (2020) Intramolecular chaperone‐mediated secretion of an Rhs effector toxin by a type VI secretion system. Nat Commun 11: 1865 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Quentin D, Ahmad S, Shanthamoorthy P, Mougous JD, Whitney JC, Raunser S (2018) Mechanism of loading and translocation of type VI secretion system effector Tse6. Nat Microbiol 3: 1142–1152 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Roessner CA, Ihler GM (1984) Proteinase sensitivity of bacteriophage lambda tail proteins gpJ and pH in complexes with the lambda receptor. J Bacteriol 157: 165–170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Roessner CA, Ihler GM (1986) Formation of transmembrane channels in liposomes during injection of lambda DNA. J Biol Chem 261: 386–390 [PubMed] [Google Scholar]
  47. Ruhe ZC, Low DA, Hayes CS (2020) Polymorphic toxins and their immunity proteins: diversity, evolution, and mechanisms of delivery. Annu Rev Microbiol 74: 497–520 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Russell AB, Peterson SB, Mougous JD (2014) Type VI secretion system effectors: poisons with a purpose. Nat Rev Microbiol 12: 137–148 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Saint Jean KD, Henderson KD, Chrom CL, Abiuso LE, Renn LM, Caputo GA (2018) Effects of hydrophobic amino acid substitutions on antimicrobial peptide behavior. Probiotics Antimicrob Proteins 10: 408–419 [DOI] [PubMed] [Google Scholar]
  50. Vettiger A, Basler M (2016) Type VI secretion system substrates are transferred and reused among sister cells. Cell 167: 99–110.e12 [DOI] [PubMed] [Google Scholar]
  51. Wang J, Brodmann M, Basler M (2019) Assembly and subcellular localization of bacterial type VI secretion systems. Annu Rev Microbiol 73: 621–638 [DOI] [PubMed] [Google Scholar]
  52. Wettstadt S, Filloux A (2020) Manipulating the type VI secretion system spike to shuttle passenger proteins. PLoS One 15: e0228941 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wettstadt S, Lai EM, Filloux A (2020) Solving the puzzle: connecting a heterologous Agrobacterium tumefaciens T6SS effector to a Pseudomonas aeruginosa spike complex. Front Cell Infect Microbiol 10: 291 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Wu HY, Chung PC, Shih HW, Wen SR, Lai EM (2008) Secretome analysis uncovers an hcp‐family protein secreted via a type VI secretion system in Agrobacterium tumefaciens . J Bacteriol 190: 2841–2850 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wu CF, Santos MNM, Cho ST, Chang HH, Tsai YM, Smith DA, Kuo CH, Chang JH, Lai EM (2019) Plant‐pathogenic Agrobacterium tumefaciens strains have diverse type VI effector‐immunity pairs and vary in in‐planta competitiveness. Mol Plant Microbe Interact 32: 961–971 [DOI] [PubMed] [Google Scholar]
  56. Wu CF, Lien YW, Bondage D, Lin JS, Pilhofer M, Shih YL, Chang JH, Lai EM (2020) Effector loading onto the VgrG carrier activates type VI secretion system assembly. Embo Rep 21: e47961 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wu CF, Weisberg AJ, Davis EW 2nd, Chou L, Khan S, Lai EM, Kuo CH, Chang JH (2021) Diversification of the type VI secretion system in agrobacteria. mBio 12: e0192721 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Yu M, Wang YC, Huang CJ, Ma LS, Lai EM (2020) Agrobacterium tumefaciens deploys a versatile antibacterial strategy to increase its competitiveness. J Bacteriol 203: e00490‐20 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zhang D, de Souza RF, Anantharaman V, Iyer LM, Aravind L (2012) Polymorphic toxin systems: comprehensive characterization of trafficking modes, processing, mechanisms of action, immunity and ecology using comparative genomics. Biol Direct 7: 18 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Zoued A, Durand E, Brunet YR, Spinelli S, Douzi B, Guzzo M, Flaugnatti N, Legrand P, Journet L, Fronzes R et al (2016) Priming and polymerization of a bacterial contractile tail structure. Nature 531: 59–63 [DOI] [PubMed] [Google Scholar]

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    Appendix

    Click here for additional data file. (173.3KB, pdf)

    Expanded View Figures PDF

    Click here for additional data file. (3.3MB, pdf)

    PDF+

    Click here for additional data file. (10.2MB, pdf)

    Source Data for Figure 1

    Click here for additional data file. (15.1KB, zip)

    Source Data for Figure 2

    Click here for additional data file. (160.4MB, zip)

    Source Data for Figure 3

    Click here for additional data file. (15.2MB, zip)

    Source Data for Figure 4

    Click here for additional data file. (40.3MB, zip)

    Source Data for Figure 5

    Click here for additional data file. (42.1MB, zip)

    Data Availability Statement

    No large primary datasets have been generated and deposited.

    Articles from EMBO Reports are provided here courtesy of Nature Publishing Group

    RESOURCES