Abstract
Type IV secretion systems (T4SSs) are central to bacterial pathogenesis. Traditionally known for facilitating DNA transfer via conjugation, T4SSs also mediate biofilm formation. These biofilms are critical for the fitness of adherent-invasive Escherichia coli (AIEC), which are commonly isolated from Crohn’s disease patients and are known for propelling gut inflammation. Many AIEC strains carry F-like plasmids encoding the IncF subgroup of T4SSs. Unlike minimized systems with 12 core components, the IncF family is an expanded T4SS with additional genes that enhance conjugation. Here, we show that a biofilm-forming AIEC strain harbors an unusual IncF plasmid that lacks two conserved components essential for T4SS functionality. This strain forms a natural hybrid T4SS where the missing components are supplied by a co-residing chromosomal T4SS on an integrative and conjugative element (ICE). Biochemical assays reveal that this hybrid T4SS drives pilin polymerization and biofilm formation on epithelial cells. Furthermore, we show that a bacterial subpopulation expresses the IncF and ICE-encoded genes in response to host cells, leading to the assembly of biofilms and enhanced fitness in the gut. These findings uncover crosstalk between two evolutionary distant mobile genetic elements to form a hybrid T4SS that mediates biofilm biogenesis by a Crohn’s disease-associated pathogen.
Similar content being viewed by others
Introduction
Crohn’s disease (CD) is marked by transmural inflammation and the disruption of the gut barrier1,2. These changes facilitate the expansion of opportunistic pathogens, such as adherent invasive strains of Escherichia coli (AIEC)3. Although the pathogenesis of CD is complex and incompletely understood, members of this pathogroup have been frequently isolated from CD patients, with the prevalence of AIEC among this population reaching ~30% relative to healthy individuals4,5. This positive correlation between AIEC and inflammation was demonstrated in several preclinical models, where these strains triggered disease in susceptible hosts6,7,8,9. Several groups showed that the ability of these bacteria to adhere to and invade host cells is superior to commensals9,10,11,12,13. Furthermore, AIEC can survive eradication by macrophages through hijacking host defense responses such as autophagy14,15. However, the virulence factors that drive AIEC pathogenesis remain unclear. This information is cardinal to discriminate these strains from commensals and to design the next generation of selective antimicrobials.
Previous work identified a role for the type IV secretion system (T4SS) in promoting AIEC fitness in the gut16. In addition to mediating DNA transfer via conjugation, this secretion system facilitates diverse functions such as interbacterial killing, effector protein secretion, and biofilm formation17,18,19. T4SS-mediated biofilms are formed via promoting interbacterial adhesion as a ‘by-product’ of bacterial conjugation which also relies on forming cell-to-cell contacts, resulting in tightly aggregated bacterial communities. Given their limited impermeability towards antibiotics and immune cells, these structures promote bacterial persistence in the gut20. In this regard, ~30% of AIEC strains carry F-like plasmids, which encode T4SSs that belong to the IncF group. Additionally, IncF genes were enriched in the metagenomes of CD-associated microbiomes relative to healthy individuals16. We previously showed that AIEC strains employ these nanomachines to form biofilms on the surface of epithelial cells, promoting their persistence in vivo. Mutant bacteria incapable of elaborating the F pili show diminished ability to assemble biofilms on epithelial cells and are rapidly cleared from the murine gut16. These biofilms are facilitated by the ability of the long flexible pili of the IncF system to mediate aggregation16,21,22. Indeed, the interactions between F pili and the membrane phospholipids were shown to significantly enhance their biophysical and thermochemical stability, making these pili efficient in mediating bacterial aggregation and biofilm biogenesis under harsh conditions21,23.
Structurally, T4SSs are classified into minimized and expanded systems18. Minimized systems comprise at least 12 main components of which 11 are essential for functionality. These are the VirB2-11 and D4 proteins carried by the Ti plasmid of the plant pathogen, Agrobacterium tumefaciens, which has become the paradigm of all T4SSs. Throughout evolution, these minimized systems acquired additional components, leading to the emergence of expanded T4SSs18. These systems are exemplified by the T4SS carried on F and F-like plasmids (e.g., IncF). Canonical IncF plasmids usually carry an average of 36 genes organized in an operon, making this cluster one of the longest polycistronic transcripts in E. coli24,25,26. Both minimized and expanded systems share common structural features, and their topology can be divided into discrete subassemblies such as the inner membrane complex (IMC) and outer membrane core complex (OMCC), which together with the periplasmic stalk, constitute the platform for pilin assembly through the polymerization of the TraA (VirB2 homologue) subunit, also called the F-pilin. The OMCC is the largest and the most stable T4SS subassembly, hence it is considered the foundation for the nucleation of the remainder of the T4SS apparatus. It comprises two flexible concentric rings of mismatched symmetry formed by the assembly of TraV, K, B (VirB7, 9, 10)27. The inner membrane complex (IMC) is mainly composed of TraC (VirB4), an AAA+ ATPase that associates with the membrane as a hexameric collar of dimers28,29. Through its N-terminus domain, TraC interacts with inner membrane proteins such as TraL (VirB3), forming the IMC that connects with the OMCC through the periplasmic stalk. In the R388 system, this is a cone-shaped structure that comprises a pentamer of the inner membrane protein, TraG (VirB6), forming the base upon which a pentamer of TraE (VirB5) mounts. The stalk is further docked to the inner membrane through the interactions between VirB6 and VirB3 of the IMC29. Together, this complex serves as a platform for polymerizing pilin monomers, TraA. This process is energised by TraC through its ATPase activity, making this enzyme essential for the elaboration of the F-pilus.
Most of our knowledge on how T4SSs are regulated was obtained from laboratory-domesticated strains that constitutively express these genes24. While these studies provided instrumental structural insights into T4SSs, how the assembly of these machines is regulated in clinical isolates remains unknown. In this work, we investigated the mechanisms of T4SS assembly in AIEC str. NRG857c, previously isolated from the ileum of a CD patient30 and shown to use this secretion system for making biofilms in the gut16. We show that the IncF system carried by this AIEC strain is missing two of the core components that are required for pilin assembly. These are the VirB4 homologue, TraC, and the VirB7 homologue, TraV. We show that functional TraC and TraV carried by an integrative and conjugative chromosomal element (ICE) complement the function of IncF. ICEs are mobile genetic elements that can integrate into bacterial genomes. Furthermore, they encode for a T4SS that can facilitate their transmission to a new recipient via conjugation upon excision from the original host31,32,33. We show that both IncF and the ICE contribute to the formation of a chimeric T4SS by AIEC. We demonstrate that this mosaic system is upregulated in the presence of host cells, triggering biofilm formation. Importantly, we show that ICE-encoded TraC and TraV are required for in vivo fitness. This work uncovers new elements of AIEC pathogenesis and provides important insights into the evolution of T4SSs. This information will guide the development of novel therapeutics in preventing and managing Crohn’s disease.
Results
IncF T4SS in AIEC lacks two core components required for pilin assembly
Previous work showed that AIEC str. NRG857c carries a ~ 147 kbp F-like plasmid, pO83 (accession: NC_017659.1). This plasmid encodes for multiple antibiotic resistance cassettes, virulence factors such as iron uptake proteins, and an IncF T4SS, which was found to be used by these bacteria to form biofilms on epithelial cells. This process is key to bacterial fitness, as a mutant lacking the conjugative pili (∆traAIncF) was rapidly cleared from the host environment16. To identify other IncF genes in the pO83 F-like plasmid and examine their role in biofilm biogenesis, we compared the encoding cluster from AIEC to that of the reduced F plasmid derivative, pOX38. F plasmids encode for expanded T4SSs that comprise homologues of the 12 core components of the Vir system from A. tumefaciens (VirB1-11 and D4) in addition to other proteins that are required for efficient conjugation18,34. While the two clusters displayed high homology at the primary sequence level, we observed that several genes are present in pOX38, yet absent in AIEC (Fig. 1A). This includes traN, traW, and trbC, which play a role in mating pair stabilization during conjugation35,36. We then investigated the IncF cluster of AIEC to validate the presence of the minimal components required for functionality. We based this analysis on the archetypal Vir system of A. tumefaciens as a representative of minimized systems. Interestingly, we observed that the IncF cluster of AIEC lacks traC and traV homologues, which are required for T4SS assembly and function. TraC is an ATPase that forms a hexamer of dimers and dominates the IMC as two concentric rings28,29. Through ATP hydrolysis, TraC fuels the oligomerization of pilin subunits leading to the elaboration of the F-pilus. Similarly, TraV is required for the assembly of the OMCC. This lipoprotein anchors the outer layer of the TraK and TraB multimeric complex to the outer membrane18,27,29. Given that AIEC was shown to form T4SS-dependent biofilms16, we hypothesized that homologues of traC and traV are potentially encoded on the chromosome. Therefore, we performed a BLASTP37 search using TraC from the IncF plasmid, pOX38, as the query against the genome of AIEC str. NRG857c. Our analysis identified NRG857_20910 as the closest homologue to TraC from F. This gene is part of a cluster predicted to be an integrative and conjugative element (ICE) by ICEFinder, a webtool used for the detection of ICE-encoded T4SSs38. ICEs are mobile genetic elements that can integrate into bacterial chromosomes, which can be reversed using excisionases31,32,33. Because they encode their own T4SS, they can self-mobilize to a new host after excision. This ICE is homologous to Hin1056 from Haemophilus influenzae (Fig. 1A), which was previously shown to encode for a functional T4SS39. We used AlphaFold340 to predict the structure of the ICE-encoded TraC homologue (TraCICE) and queried this model against the Protein Data Bank using the DALI webserver41. In line with the primary sequence analysis, this search identified VirB4 of pR388 as the closest homologue to TraCICE29. The predicted structure of TraCICE shows the conserved domains of VirB4. This includes the four-helix bundle and the ɑ/ β RecA domain responsible for binding ATP42(Fig. 1B). To identify a TraV homologue in the ICE, we extended our bioinformatics approach to predict the structure of the proteins encoded by the genes nearby traCICE (Table S1). In this regard, the predicted structure of the orf upstream of traCICE exhibited the structural signature characteristic of TraV. This includes four β-strands that form two continuous β-sheets extending across adjacent TraV subunits, stabilizing the OMCC27. This domain also aligned with the solved structure of TraV from pED20827 (Fig. 1C). Additionally, the putative TraVICE has a lipobox with a serine residue at +2 position, which is characteristic of outer membrane lipoproteins43. Together, these data suggest the unique presence of two T4SS-encoding clusters with TraC and TraV homologues uniquely present in the chromosomal ICE.
A The archetypal T4SS present on Ti plasmid from Agrobacterium tumefaciens shows the 12 core components of minimized T4SSs (top). Primary sequence alignment of uncharacterized AIEC ICE demonstrates high homology to ICEHin1056 that encodes for T4SS in Haemophilus influenzae (middle). Primary sequence alignment of tra cluster from AIEC str. NRG857c to that of the canonical Escherichia coli F-plasmid pOX38 shows the absence of traV and traC (bottom). Scale bar is 2 kb. Predicted structures of TraCICE and TraVICE aligned to the homologues from pR388 (B) and pED208 (C) respectively.
TraVICE and TraCICE interact with the IncF-encoded T4SS components
To confirm the activity of TraCICE, we cloned and expressed the protein with a histidine tag at the C-terminus. Using affinity chromatography followed by size exclusion chromatography, we purified TraCICE and assessed its homogeneity and its purity using SDS polyacrylamide gel electrophoresis (PAGE) (Fig. 2A). We then confirmed the ATPase activity of the purified protein using a standard biochemical assay (Fig. 2B). TraCICE showed a specific activity of ~30 nmole.min-1.mg-1, which is close to that reported for VirB4 of pKM10142. TraC is known to form dimers that arrange into a hexameric complex that associates with the inner membrane. This serves as a platform for the assembly of other IMC components such as TraL, which interacts with the central subunits of the TraC hexamer29,44,45. IncF of AIEC encodes a TraL homologue (NRG857_30193) that is absent in the ICE. We then used AlphaFold340 to predict its structure and found that it displayed striking similarity to the solved structures of TraL homologues in other IncF systems (Fig. S1). In this regard, we hypothesized that TraCICE interacts with IncF-encoded TraL (TraLIncF). To test this, we co-expressed TraCICE with a Strep-tag at the C-terminus together with untagged TraLIncF. We then purified TraCICE using a combination of affinity and size-exclusion chromatography to eliminate co-impurities. The eluted fractions were separated using SDS-PAGE. In addition to a prominent band that corresponds to TraCICE (~100 kDa), we observed a ~ 11 kDa protein that co-eluted with it (Fig. 2C). We identified this protein as TraLIncF using mass spectrometry, confirming the interaction between TraCICE and TraLIncF. Similarly, we sought to test whether TraVICE can interact with IncF-encoded TraB and TraK. Together, these 3 proteins are known to form the concentric rings of the OMCC. To this end, we co-expressed TraBIncF carrying a Strep-tag at the C-terminus with TraKIncF and TraVICE. Moreover, we engineered TraVICE to have an HA tag at its C-terminus to monitor it in the complex using immunoblotting. The complex was purified using affinity chromatography followed by size exclusion chromatography. We then separated the eluates using SDS-PAGE to identify potential binding partners. As expected, we were able to detect TraKIncF in all the TraBIncF-containing fractions. Interestingly, TraVICE co-eluted with both proteins, which we validated using immunoblotting (Fig. 2D), confirming that these 3 proteins assemble into a hybrid OMCC. Together, these results suggest a crosstalk between the two systems in AIEC str. NRG857c to mediate the assembly of a T4SS that is a hybrid of ICE and IncF-encoded components.
A Purification of putative TraCICE using affinity followed by size exclusion chromatography. B Purified TraCICE shows ATPase activity in vitro. Data represent three independent replicates and are presented as mean values +/− SEM. C Co-purification of TraCICE and TraLIncF using affinity followed by size-exclusion chromatography. D Co-purification of the hybrid OMCC from E. coli co-expressing TraBHIncF, TraKIncF, and TraVICE (~17 kDa). An HA-tag was added to the C-terminus of TraVICE to confirm its identity via immunoblotting. Three independent experiments were performed, and similar results were obtained every time. Source data are provided as a Source Data file.
ICE complementation of IncF T4SS is required for biofilm formation and conjugation
We previously showed that AIEC uses T4SS to form biofilms on epithelial cells16. We compared human epithelial cells from different origins to evaluate their ability to induce biofilm formation. Unlike kidney (HEK 293 T) and cervical (HeLa) epithelial cells, AIEC wild-type robustly formed biofilms on Caco-2 intestinal epithelial cells, validating the use of these cells to study T4SS-dependent biofilm formation. Importantly, these differences are not driven by variation in cell adhesion, as AIEC attached equally to all epithelial cell lines tested (Fig. S2). Thus, we sought to determine whether the ICE-encoded traC and traV are required for biofilm formation. To test this, we used two-step allelic exchange to generate clean deletions of traCICE and traVICE. We then infected Caco-2 epithelial cells with either the wild-type or mutant strains. As a control, we infected a group of Caco-2 cells with a ∆traAIncF AIEC strain. This mutant lacks the conjugative pili, which were previously shown to be required for efficient biofilm formation16. Epithelial cells were infected with the different AIEC strains for 5.5 h, which is sufficient for biofilm formation16. Using immunofluorescence, we monitored microcolony formation across the different strains. As shown in Fig. 3A, wild-type AIEC formed robust biofilms on epithelial cells, which were dependent on the IncF-encoded traA. Interestingly, knocking out either traCICE or traVICE diminished biofilm formation, suggesting that these ICE-encoded elements complement the IncF T4SS cluster. To quantify these structures, we tallied the number of bacterial aggregates larger than 100 µm2, a threshold previously shown to correspond to the sizes of biofilms made by AIEC16. Compared to the wild-type strain, we observed a 6 to 10-fold reduction in biofilm formation when either traCICE or traVICE were deleted, which was rescued when these genes were expressed in trans (Fig. 3B). This analysis confirmed that ICE-encoded traC and traV are both required for efficient biofilm biogenesis by AIEC on epithelial cells. While our data show that ICE complements the missing tra genes of IncF, it is unclear whether the homologues common between the two systems are functionally redundant. Our AlphaFold3 analysis identified a TraK homologue on the ICE in addition to the one encoded by the IncF plasmid of AIEC (Table S1). Multimers of TraK interact with TraB and TraV to form the OMCC, which is predicted to be flexible enough to facilitate the extension of the polymerizing pili through the pore29. Thus, a functional TraK is required for T4SS-dependent biofilm assembly. To test whether these homologues are functionally redundant, we deleted traKICE and traKIncF independently and tested the ability of these strains to form biofilms on epithelial cells. Our assumption was that ∆traKICE and ∆traKIncF would be equally efficient in biofilm formation to wild-type if these homologues are functionally redundant. To this end, we infected epithelial cells with wild-type, ∆traKICE and ∆traKIncF AIEC and used immunofluorescence to monitor biofilm formation. Interestingly, we observed a ~ 5-fold reduction in biofilm formation with ∆traKIncF relative to wild-type, which was rescued upon expressing traKIncF in trans. On the contrary, ∆traKICE did not display any significant reduction in biofilm formation (Fig. 3A, B). To ensure that the different levels of biofilm formation by the different AIEC strains are not driven by a defect in cell adhesion, we quantified their ability to attach to the epithelial cells using standard assays. As shown in figure S3, all mutants adhered to Caco-2 cells at comparable levels to wild-type, confirming the role of IncF and ICE genes in biofilm biogenesis. While this hybrid system is able to drive robust biofilm formation on epithelial cells, DNA transfer via conjugation is one of the major functions of T4SS. Thus, we sought to assess the role of the hybrid T4SS in AIEC in bacterial mating using the commensal E. coli strain HS as a recipient. We engineered this strain to become gentamicin resistant by inserting a resistance cassette into a chromosomal intergenic region. The F-like plasmid of AIEC, pO83, carries several antibiotic resistance markers such as CAT (chloramphenicol acetyl transferase), facilitating the quantification of transconjugants upon its transfer. We then mated E. coli str. HS with AIEC strains lacking the different ICE and IncF genes. As a positive control, we used E. coli carrying the conjugative pOX38 as a donor. All mutants, including ∆traKICE, showed a modest decrease in conjugation (2 to 5 fold) relative to wild-type, which was rescued to varying levels when the genes were expressed in trans (Fig. 3C). Importantly, we observed a significant increase (~4 orders of magnitude) in mating when the pOX38-carrying E. coli was used as a donor compared to AIEC. These results suggest a co-operative relationship between IncF and ICE in mediating T4SS-dependent biofilm formation and conjugation by AIEC.
A Biofilm formation by the different AIEC strains on epithelial cells. Immunofluorescence staining shows the cytoskeleton of Caco-2 epithelial cells in red with DAPI-stained nuclei in blue, infected with wild-type (wt) or isogenic mutant AIEC strains in green. Isogenic strains carrying vector controls were used when applicable. Each panel is representative of 45 images per group, n = 3 independent biological replicates. Scale bar is 100 µm. B Quantification of biofilm formation by each strain shown as a percentage relative to wt. Biofilms were quantified across 15 fields of view per trial over three independent experiments. Data is presented as mean values +/− SEM. Statistical significance was tested by Kruskal–Wallis test with Dunn’s multiple comparisons test. C Wt AIEC donor displayed higher conjugation relative to the isogenic mutants lacking different T4SS components. E. coli /pOX38 was used as a positive control. Data represent 21 biological replicates from 6 independent experiments. Outliers were removed using Grubb’s test. Data is presented as mean values +/− SEM. Statistical significance was tested by repeated measures one-way ANOVA with Dunnett’s multiple comparisons test comparing the different AIEC strains to wt, ****P < 0.0001. Source data are provided as a Source Data file.
Co-expressing ICE-encoded TraC and TraV with IncF genes triggers pilin assembly and biofilm formation
The presence of traVC in the ICE suggests that these genes are regulated independently from the IncF cluster. To investigate this, we sought to derepress IncF as this cluster was shown to be expressed at low levels under standard laboratory conditions16. The expression of IncF-encoded tra genes is under the control of the pY promoter46. This is positively regulated by the transcriptional activator, TraJ, and the response regulator, ArcA47. However, TraJ is negatively regulated post-transcriptionally with the small RNA, finP, which can bind to the 5’ UTR of traJ to silence its expression. Since finP is susceptible to degradation by ribonuclease E, the expression of its cognate chaperone, FinO, is essential for its integrity48,49,50,51. In this regard, the F plasmid is naturally derepressed due to the disruption of finO by an insertion element52. Thus, we reasoned that knocking out finO will result in finP degradation, derepressing IncF-encoded tra genes. To test this, we deleted finO and monitored the expression of tra genes using a pY-dependent luminescent reporter. As shown in Fig. 4A, the ∆finO strain displayed significant luminescence compared to wild-type when grown under standard laboratory conditions, suggesting that the deletion of this chaperone relieves the finP-dependent repression of IncF. To test if traVCICE are regulated independently of IncF, we monitored their expression using RT-qPCR in the wild-type and ∆finO strains. As expected, traKIncF and traGIncF displayed ~100 fold increase in expression in the ∆finO strain relative to the isogenic wild-type. Interestingly, the deletion of finO did not affect the expression of traVCICE, confirming that these ICE genes are regulated independently of IncF (Fig. 4B). We then employed our biofilm assays to confirm our transcriptional data. We infected epithelial cells with wild-type and ∆finO AIEC and observed no increase in biofilm formation by the derepressed strain (Fig. 4C, D). We speculated that this is due to the overexpression of IncF genes by the ∆finO strain without a simultaneous increase in traVCICE levels. To address this, we expressed TraCICE in trans in the ∆finO background. We observed a modest increase in biofilm formation compared to wild-type (~1.5 fold increase). When we co-expressed TraVICE and TraCICE in the ∆finO strain, we observed a more robust increase (~2.4 fold) in biofilm formation compared to the wild-type group. As a control, we generated a TraCICE ATPase inactive variant by mutating the lysine 512 located in the Walker A motif to glutamine. This mutation was previously reported to abolish ATPase activity without affecting the oligomerization of TraC nor its association with the inner membrane complex53. We confirmed that this variant is expressed by AIEC at similar levels to that of the active enzyme using immunoblotting. Similarly, we confirmed the expression of traVICE across the different strains (Fig. S4). By co-expressing TraCICE K512Q with TraVICE, the levels of biofilm formation by the ∆finO strain dropped to those of wild-type, confirming that the ATPase activity of TraCICE is required for T4SS-mediated aggregation (Fig.4C, D). This is likely because the ATPase is essential for pilin oligomerization54, which was previously shown to be required for biofilm formation by AIEC16. Similarly, overexpressing both TraVICE and TraCICE in the ∆finO strain increased conjugation by ~4-fold relative to the isogenic strain (Fig. S5).
A Luminescent IncF reporter shows increased pY expression upon the deletion of finO. B Deleting finO results in increased expression of IncF genes but not traVCICE. Data represent 4 independent biological replicates and are presented as mean values +/− SEM. Groups were compared using unpaired two-tailed Mann-Whitney test. C Immunofluorescence staining shows cytoskeleton of Caco-2 epithelial cells in red with DAPI-stained nuclei in blue, infected with wild-type (wt) or isogenic AIEC strains in green. A significant increase in biofilm structures is observed when traVCICE are overexpressed. Isogenic strains carrying vector controls were used when applicable. Each panel is representative of 45 images per group, n = 3 independent biological replicates. Scale bar is 100 µm. D Quantification of biofilm formation by each strain shown as a percentage relative to wt per trial. Biofilms were quantified across 15 fields of view per trial over three independent trials. Data are presented as mean values +/− SEM. Groups were compared using Kruskal–Wallis test with Dunn’s multiple comparisons test, * P < 0.05, ***P < 0.001, ****P < 0.0001. Source data are provided as a Source Data file.
To confirm that the observed biofilms and conjugation levels are driven by the assembly of the T4SS pili, we employed a molecular approach to visualize these structures. Previous work showed that the pili can be engineered to incorporate surface-exposed cysteine residues, which can be labelled covalently with maleimide-fluorophore conjugates55. Thus, we engineered both wild-type and ∆finO strains by replacing the native TraAIncF with the TraAIncF S54C variant to allow maleimide bioconjugation. We observed a slight increase in piliation in ∆finO compared to wild-type, which can be driven by basal expression of chromosomal TraVICE and TraCICE. Thus, we deleted the chromosomal TraVICE and TraCICE in ∆finO to avoid any stochastic expression of these genes and expressed them in trans individually or together to properly quantify their contribution to piliation. After growing the different strains in vitro, we induced the expression of TraVICE or/and TraCICE, then labelled the pili using maleimide-fluorophore conjugates before examining the different strains using fluorescence microscopy. We then quantified the percentage of piliated bacteria relative to the entire population across the different strains. As expected, we did not detect any pili in the ∆finO ∆traVCICE strain. Only 5% of the population displayed piliation when TraCICE was overexpressed (Fig. 5A, B). On the contrary, we detected higher levels of piliation when TraVICE and TraCICE were co-expressed (~13%). This was diminished upon the expression of the inactive TraCICE K512Q, confirming that ATPase activity is required to fuel pilin polymerization (Fig. 5A, B). In agreement with our functional assays, these results show that ICE-encoded genes complement IncF to assemble a hybrid T4SS that can drive piliation, required for biofilm formation and conjugation. Nevertheless, both elements are potentially governed by independent regulatory mechanisms.
A T4SS pili formed by the different AIEC strains were labelled using maleimide-AlexaFluor488. Each panel is representative of 45 images per group, n = 3 independent biological replicates. Isogenic strains carrying vector controls were used when applicable. Arrows mark the pili. Scale bar is 20 µm. B Quantification of piliated bacteria, shown as a percentage relative to the total bacterial count per field of view. Samples were quantified across 15 fields of view per trial over four independent trials. Data is presented as mean values +/− SEM. Groups were compared using Kruskal–Wallis test with Dunn’s multiple comparisons test comparing the different strains to ∆finO∆traVC empty vector (EV) control, ****P < 0.0001, ns non-significant. Source data are provided as a Source Data file.
The host environment promotes the expression of T4SS genes encoded by ICE and F-like plasmid of AIEC
Previous work showed that host epithelial cells induce the expression of IncF genes, leading to biofilm formation16. Since TraVCICE are required for efficient biofilm formation by AIEC on epithelial cells (Fig. 3), we reasoned that these host cells upregulate the expression of ICE- and IncF-encoded genes, resulting in the assembly of a functional T4SS. Using BPROM (http://www.softberry.com/), we queried the regions upstream the operon encoding traVCICE for potential promoters. This analysis revealed the presence of a putative promoter carrying the canonical -10 and -35 elements, potentially driving the expression of ICE-encoded genes. We cloned the putative promoter upstream of a fluorescent reporter gene, generating a tractable tool to monitor the expression of ICE genes. While the expression of IncF genes was previously shown to be induced by Caco-2 epithelial cells, this analysis was performed using approaches lacking subpopulation resolution16. Thus, we also generated a fluorescent IncF reporter to monitor the expression of these genes. Two groups of epithelial cells were infected with either the ICE or IncF transcriptional reporters. Moreover, we incubated the reporter strains in the cell-free supernatants of Caco-2 cells to investigate whether a secreted factor is driving T4SS induction. As negative controls, we infected fixed epithelial cells or cell-free coverslips with the reporter strains. The high-binding nature of the coverslips allows bacteria to adhere to them in the absence of host cells. After 6.5 h, we stained the entire bacterial population using a specific antibody, while fluorescence from the IncF and ICE reporters was used to track the bacteria expressing these genes. Although the bacteria adhered to the cell-free coverslips, we could not detect significant expression of either the IncF or ICE genes as expected. However, when investigating the bacterial population adherent to epithelial cells, we observed a significant signal from the fluorescent reporters of both systems (Fig. 6A). We then quantified the percentage of IncF- and ICE-expressing bacteria relative to the entire population in the absence and presence of host cells. Both reporters revealed that 1-2% of AIEC expresses IncF and ICE in the absence of epithelial cells or when fixed cells were used. Upon adhesion to live epithelial cells, these subpopulations expanded to ~30% of IncF-expressing bacteria and ~10% of bacteria expressing ICE genes (Fig. 6B). These results reveal the presence of host cues that upregulate the genes required to assemble a functional T4SS in AIEC. To extrapolate our findings to the gut environment, we tracked the expression of ICE genes in vivo. We previously showed that the pY promoter of IncF is expressed in the murine gut within 24 h post-infection16. Therefore, we sought to track the expression of ICE using in vivo imaging, allowing us to monitor gene expression by the AIEC population in the gut in real time. Luminescent reporters are powerful tools that facilitate tracking gene expression with high sensitivity and minimal background signal56,57. In this regard, we cloned the ICE promoter upstream of the lux operon, which will result in luminescence upon expression. As expected, this reporter showed no expression under standard laboratory conditions (Fig. 7A). We then infected a cohort of mice with the luminescent ICE reporter for 16 h before imaging the animals. As a control for stochastic expression, a group of mice was infected with a promoterless reporter. As shown in Fig. 7B, we detected a significant upregulation of the ICE genes in the murine gut shortly after infection compared to the promoterless control. Together, these data demonstrate that AIEC expresses IncF and ICE genes in the host environment, resulting in the assembly of a functional T4SS.
A Immunofluorescence staining of AIEC strains (red) carrying fluorescent reporters (green) under the control of the pY or ICE promoter in the presence or absence of viable or fixed Caco-2 cells (blue) or cell-free supernatants. Arrows mark the bacteria expressing IncF or ICE genes. Each panel is representative of 45 images per group, n = 3 independent biological replicates. Scale bar is 20 µm. B Quantification of IncF-expressing and ICE-expressing bacteria, shown as a percentage relative to the total bacterial count per field of view. Samples were quantified across 15 fields of view per trial over three independent trials. Data are presented as mean values +/− SEM. Groups were compared using Kruskal–Wallis test with Dunn’s multiple comparisons test, ****P < 0.0001. Source data are provided as a Source Data file.
A ICE genes are not expressed in vitro under standard laboratory conditions. B ICE genes are expressed in the murine gut. Representative whole gut organs from female C57BL/6 J mice infected with AIEC carrying a luminescent reporter under the control of the putative ICE promoter shows significant expression compared to isogenic strains carrying empty vector, n = 3 biological replicates. C ∆traVICE, ∆traCICE, traCICE K512Q (traC*ICE), and ∆traAIncF AIEC strains are outcompeted by isogenic wild-type strain across three gut regions. Each datapoint represents a log10 competitive index from an independent animal, n = 5 biological replicates. Data are presented as mean values +/− SEM. Zero values in the ceca were replaced with pseudocounts representing the lowest value obtained prior to log10 transformation to enable calculation of competitive indices. D Mice infected with ∆traVICE, ∆traCICE, and ∆traAIncF show less lipocalin-2 and TNF-α in their ceca and colons compared to animals infected with wild-type, n = 5 biological replicates. Data are presented as mean values +/− SEM. Groups were compared using repeated measures one-way ANOVA with Dunnett’s multiple comparisons test, *P < 0.05, **P < 0.005, ***P < 0.001,****P < 0.0001. Source data are provided as a Source Data file.
ICE-encoded TraC and TraV are required for in vivo-fitness
We previously showed that T4SS-dependent biofilm formation by AIEC in vivo promotes their persistence in the gut. This is evident by the rapid clearance of a mutant lacking T4SS pili from the murine gut relative to the wild-type isogenic strain16. Given that our data revealed a role for TraVCICE in pilin assembly and biofilm formation on host cells, we hypothesized that these genes promote the fitness of AIEC in the gut. To test this, we competed ∆traVICE, ∆traCICE, and an isogenic strain carrying traCICE K512Q instead of the native gene against wild-type AIEC in the murine gut. As a control, we included ∆traAIncF strain in this analysis as we previously showed that it is less fit than wild-type16. Importantly, we engineered the wild-type strain to carry a unique antibiotic resistance marker compared to the mutants, enabling us to mix competing strains and enumerate each individually after plating the gut tissues on selective media. As a control for stochastic differences, we competed wild-type AIEC against an isogenic strain that carries traVCICE together with a unique selection marker. In each case, the competing strains were grown in vitro, then mixed 1:1 before orally infecting different cohorts of C57BL/6 J female mice. Total bacterial burdens were the same across all animals (Fig. S6). The animals were euthanized 8 days post-infection, and the gut tissues were plated on selective media to quantify each strain. Ratios between the two competing strains were calculated and normalized by their proportions in the infecting inoculum for each animal (competitive index). As shown in Fig. 7C, all mutants were outcompeted by the wild-type strain in all tissues (~0.5 fold reduction). On the contrary, the traVCICE+ strain did not show a significant difference from the isogenic wild-type, ruling out the possibility that any fitness defects were due to stochastic variability. Given that AIEC strains are known to display pro-inflammatory behaviour in the gut2,8, we sought to evaluate the role of TraVCICE in promoting inflammation. We infected several cohorts of C57BL/6 mice with wild-type and T4SS mutants. As a control, a group of animals remained uninfected. Acute colitis was triggered in all groups by administering 3% w/v dextran sulfate sodium (DSS) for 5 days58. Several groups have shown that AIEC strains exacerbate the DSS-mediated colitis in mice8,9,59. After euthanasia, we collected the gut tissues and monitored the levels of the pro-inflammatory cytokines lipocalin-2 and TNF-α. Compared to uninfected mice, wild-type AIEC triggered significant inflammation, evident by the high levels of lipocalin-2 and TNF-α in the ceca and colons of all infected mice. Interestingly, this pro-inflammatory behaviour requires functional T4SS as ∆traVICE, ∆traCICE and ∆traAIncF strains showed similar levels of lipocalin-2 and TNF-α relative to uninfected mice (Fig. 7D). Together, these data highlight the role of ICE-encoded TraVC in promoting the persistence and pro-inflammatory behaviour of AIEC in the gut.
ICEs are common among CD-associated E. coli
Since our data suggest that AIEC str. NRG857c assembles T4SS using IncF and ICE-encoded components, we sought to investigate the prevalence of these mobile genetic elements among CD-associated E. coli. We previously queried a collection of 40 published AIEC genomes for the presence of IncF systems and found that ~30% of them carry a system homologous to that of str. NRG857c16. We then used ICEFinder38 to mine these genomes for the presence of ICE. In addition, we included the prototype AIEC str. LF82 in this analysis. Interestingly, we found that ~58% of these strains carry at least one ICE (Table S2). Of those ICE-carrying strains, ~46% also have an IncF cluster. While the archetypal strain LF82 does not carry an IncF T4SS, we detected in its genome an ICE that is identical to that of str. NRG857c (Fig. S7). Together, these data show that most AIEC strains have either ICE or IncF T4SS with some strains carrying both elements.
Discussion
Although T4SSs are well-known for their role in DNA transfer via conjugation, the contribution of these secretion systems to bacterial pathogenesis extends to mediating biofilm formation16,21,22,60. Indeed, T4SS-mediated biofilms were shown to be crucial for the in vivo-fitness of prominent pathogens such as AIEC, known for their contribution to gut inflammation during CD3,4,16. This is potentially due to the ability of biofilms to shield bacteria from the immune cells61,62. In this regard, the conjugative pili can potentially initiate the aggregation required for biofilm formation due to their ability to connect neighbouring bacteria21.
Traditionally, T4SSs are classified as minimized or expanded systems18. Minimized systems are exemplified by the archetypal T4SS present in A. tumefaciens. These systems are minimally composed of the VirB1-B11 and D4 homologues. During their evolution, expanded systems, such as IncF, acquired additional proteins with specialized functions. This includes the ability of F pili to extend and retract, which is lacking in the pili elaborated by minimized systems63. In this work, we identified a functional T4SS in AIEC str. NRG857c that forms a functional hybrid of the IncF and ICE-encoded T4SSs. Unlike the canonical IncF plasmids, which feature expanded T4SSs, this AIEC strain carries a IncF plasmid with a reduced number of genes. Notably, it does not encode two of its essential components. These are the VirB4-like ATPase, TraC, and the VirB7-like OMCC component, TraV. Using bioinformatics, we identified some ICE-encoded homologues of these genes in the chromosome of AIEC. ICEs are mobile genetic elements that encode various beneficial genes such as antibiotic resistance and virulence factors, as well as their own T4SSs to facilitate host-to-host transmission31,32,33. We show that both TraVCICE can functionally complement the lack of their homologues in the IncF of AIEC, resulting in the assembly of a functional hybrid T4SS. This is supported by diminished biofilm formation when traVICE or traCICE are absent. Similarly, both genes are required to promote the persistence of AIEC in the gut. Additionally, deleting finO, the chaperone of the T4SS repressor finP, resulted in increased expression of IncF genes without a simultaneous increase in T4SS-dependent biofilms. This can be rescued by overexpressing TraVCICE, suggesting that ICE-encoded genes are regulated independently of IncF. In accord with this, the ∆finO AIEC strain does not assemble T4SS pili without the overexpression of TraVCICE. Importantly, our biochemical assays documenting the cross-interaction between TraLIncF and TraCICE suggest that TraCICE can associate with the IMC of IncF. Similarly, TraVICE can associate with TraBIncF and TraKIncF in AIEC, aligning with the structural diversity of the OMCC that was previously shown to have the ability to form hybrid structures under laboratory conditions18. Moreover, the lack of clear homologues of TraQFHW on the IncF plasmid suggests that the crosstalk between the two co-residing systems may extend beyond TraCICE and TraVICE, perhaps additionally recruiting homologues which functionally replace these components. Indeed, one of the identified ICE components, Tfc4, contains a TrbI-like domain (Table S1), suggesting it can functionally complement the lacking TrbI which is otherwise conserved in IncF systems. Future work will investigate the potential role of the other ICE genes in driving the assembly of hybrid T4SSs.
Interestingly, we observed the presence of certain T4SS components that are common between ICE and IncF. This includes TraK, which is essential for the formation of a stable OMCC18,27,29. We showed that TraKICE is dispensable for T4SS-dependent biofilm formation, unlike the IncF homologue. While this suggests that IncF-encoded components dominate the T4SS mediating biofilms, we cannot rule out the possibility that the redundant OMCC components in ICE may be involved in other processes. This is supported by the structural flexibility observed among OMCCs from the different T4SSs, which is likely an adaptation to their functions. For example, OMCCs from expanded systems are larger than those in minimized systems18,27. This is potentially to accommodate the dynamics of pilin extension and retraction associated with expanded systems such as IncF. Thus, one could envisage that AIEC str. NRG857c assembles more than one OMCC, each suited for different functions relevant to virulence. In accord with this, the host environment was shown to drive the adaptation of the VirB10 homologue in Helicobacter pylori, fine-tuning the functionality of its T4SS64. Thus, future work will involve identifying and solving the structure of the OMCCs assembled by AIEC str. NRG857c. In addition to str. NRG857c, we investigated the prevalence of ICEs among 41 AIEC strains. We found that 58% of these strains carry at least one ICE element. Of the ICE+ strains, 46% carry an IncF cluster. Thus, future work will investigate whether these strains can assemble hybrid T4SSs that can mediate different functions. Interestingly, we found that the prototypical AIEC str. LF82 does not carry an IncF cluster; yet, have an ICE that is identical to that of str. NRG857c. This finding raises questions about the ability of ICE+ IncF- strains to complement incomplete conjugative plasmids that they may receive from other bacteria in a complex polymicrobial environment such as the human gut. Indeed, more work will be required to study the prevalence of hybrid T4SSs among CD-associated E. coli.
While our data support the assembly of at least one chimeric T4SS in AIEC str. NRG857c, the results pose important outstanding questions. First, what are the regulatory mechanisms that govern the assembly of this system from two separate genetic elements? Using single-cell reporters, we show that epithelial cells induce the expression of both ICE and IncF-encoded T4SS genes. This is further supported by tracking the expression of these genes in the gut environment. Nevertheless, it is unclear whether a common signal triggers the expression of both the ICE and IncF in AIEC. Second, what is the evolutionary selection for chimeric T4SSs? Recent data show that ~25% of classical IncF systems lack one or more genes required for functionality, which can be provided by a co-residing conjugative plasmid65. Complementarity across conjugative plasmids that are not self-transmissible ensures that they are maintained in bacterial populations. Our data suggest that this modularity extends to different T4SS subtypes, such as those encoded by the unrelated IncF and ICE systems described here. However, we observed lower levels of piliation per cell in AIEC compared to other F-carrying strains, which can assemble up to 20 pili per cell66,67. Only ~13% of the AIEC population assembles pili with the overexpression of all required IncF and ICE-encoded components. This suggests that chimeric T4SSs are potentially less efficient in pilin assembly than the canonical IncF. In support of this, swapping components between two distantly related IncF systems yielded functional T4SSs; yet, they were less efficient in conjugation relative to the native systems65. Similarly, we show that the IncF and ICE components of the hybrid T4SS are required for conjugation. Nevertheless, the conjugation levels displayed by AIEC were significantly lower than the E. coli strain carrying the F plasmid derivative, pOX38. While making few pili, AIEC is able to assemble biofilms on epithelial cells, potentially at a lower metabolic cost relative to other F-carrying E. coli. In this regard, the production of a few pili by a subpopulation of bacteria is sufficient to facilitate interactions among kins, leading to microcolony formation and subsequent biofilm development. This is further facilitated by the high stability of F pili, which can support biofilm formation under harsh conditions such as the gut environment21.
A larger number of biofilms were detected in the gut biopsies of CD patients compared to healthy controls68. While the prevalence of T4SS-mediated biofilms in the CD environment has not been investigated, tra genes were detected at higher levels in the CD-associated metagenomes compared to that of healthy individuals16. However, the proportion of these genes that are carried by ICEs has not been determined. Our study reveals a role for ICE-encoded T4SS genes in biofilm formation by CD-associated pathogens. Thus, future research should focus on investigating the abundance of ICEs in the CD gut and their potential contribution to the formation of chimeric T4SSs. Elucidating the regulatory mechanisms governing T4SS and biofilm formation is pivotal in developing novel therapeutic strategies to prevent and manage CD.
Methods
Mice
All animal work was approved by the Canadian Council on Animal Care Committee under the University of Alberta Animal Use Protocol #3860. Female C57BL/6 J mice were purchased from Jackson Laboratory and bred in-house under specific-pathogen-free biocontainment level 2 conditions with 12 h light and dark cycles from 7 am to 7 pm at 21-25 °C and 45-55% humidity. Animals were provided a 4% fat (w/w) diet ad libitum.
Growth of bacterial strains
AIEC str. NRG857c strains, E. coli str. HS, and E. coli K12 carrying pOX38 were grown at 37 °C in Luria broth (LB, Miller) with shaking (220 rpm). When needed, antibiotics were added at the following concentrations: ampicillin 100 µg/ml, chloramphenicol 34 µg/ml, kanamycin 50 µg/ml, gentamicin 30 µg/ml.
Eukaryotic cell lines
Caco-2, HEK 293 T, and HeLa cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Caco-2 and HEK 293 T cell lines used in this study are not commonly misidentified cell lines. The HeLa cell line is a commonly misidentified cell line according to the ICLAC register. Regular morphological examination of Caco-2, HEK 293 T, and HeLa cell lines were performed for authentication, and all cell lines were tested negative for mycoplasma using PCR.
T4SS operon alignments
DNA sequences of the respective operons: A. tumefaciens Ti-plasmid encoded T4SS (GenBank: AF242881), H. influenzae 1056 ICE encoded T4SS (NCBI: NC_011409), the E.coli NRG857c ICE encoded T4SS (NCBI: NC_017634.1), the pO83-encoded (NCBI: NC_017659.1), the E. coli LF82 ICE encoded T4SS (NCBI: NC_011993.1) and pOX38-encoded T4SS (NCBI: NZ_OQ683454.1) were annotated based on SecReT4 database. Locus alignments were done using Clustal Omega and visualised using Clinker69 and Adobe Illustrator.
Structural predictions
Based on the primary sequence alignments, homologues between the different T4SSs were used to generate structural predictions in AlphaFold340, which were subsequently aligned and visualised in ChimeraX and annotated in Adobe Illustrator.
Bacterial strain construction
The gene sequences encoding for TraKIncF and TraBHIncF fused with a C-terminal Strep II-tag were cloned into a pASK-IBA3C plasmid and TraVICE fused with a C-terminal HA-tag was cloned into a pBAD24 plasmid using the In-Fusion HD enzyme Premix kit (TakaraBio). TraCICE was cloned with 10 histidine residues at the C-terminus into pExt22 plasmid70 to generate pTA7. traCICE was amplified with primers TA8/TA9 and ligated to pExt22, amplified with primers TA10/TA7, using Gibson assembly. Similarly, pJW8 and pJW9 were generated to express traVICE and traKIncF respectively. traVICE was amplified with primers JW69/wel1599 and ligated to pExt22 amplified using JW70/JW71 to generate pJW8, while traKIncF was amplified with primers JW72/JW75 and ligated to pExt22 after amplifying it with JW73/JW74 to generate pJW9. pWel64 was generated to co-express TraCICE and TraVICE. traVICE was amplified using wel1597 and wel1598, while traCICE was amplified using TA9 and wel1569 followed by a 3-way ligation into pExt22. To generate pWel65 that co-expresses TraVICE and TraCICE K512Q, primers wel1563/1568 were used to introduce the mutation in traCICE. To co-express TraCICE-TraLIncF, we cloned both encoding genes into pExt20_Km to generate pWel75. pExt20_Km was generated by amplifying the aminoglycoside-3’-O-phosphotransferase cassette from pExt22 using primers wel1522/1523 and ligated to pExt2070 that was amplified with primers wel1524/1525. Untagged traLIncF was amplified with primers wel1645/1646 and ligated upstream of traCICE, which was amplified with wel1647/1648 to include a Strep-tag at its C-terminus. Both amplicons were ligated into the pExt20_km, which was amplified with SM200/wel1572, to generate pWel75.
Two-step allelic exchange was used to knock out traCICE, traVICE, traKICE, traKIncF, finO, in wild-type AIEC and traVICE in the ∆traCICE mutant. In brief, upstream and downstream regions flanking the genes of interest were amplified and ligated to the suicide vector pRE1129,71 to generate pWel28 (∆traCICE), pWel72 (∆traVICE), pWel59 (∆traKICE), pWel60 (∆traKIncF), pWel29 (∆finO), and pWel71 (∆traVICE in ∆traCICE). Knock-out constructs were mobilized to AIEC str. NRG857c by conjugation with E. coli SM10. Co-integrants were selected for by plating conjugation lawns on LB agar supplemented with ampicillin (100 µg/ml), chloramphenicol (34 µg/ml), and gentamicin (20 µg/ml). Integrated plasmids were curated by plating co-integrants on 10% sucrose. Deletions were confirmed by Sanger sequencing. Similarly, pWel51 was constructed to swap TraAIncF with TraAIncF S54C. traAIncF and its flanking regions were amplified using primers wel1469/wel1539 and wel1534/wel1472. Both fragments were ligated to pRE112, which was amplified with wel1473/wel1474. To swap chromosomal TraCICE with TraCICE K512Q, pJW15 was constructed by amplifying traCICE and its flanking regions with JW90/JW91 and JW92/JW93, which also introduced K512Q mutation in the frame. Both fragments were ligated to pRE112, which was amplified with wel1577/wel1578. To generate the fluorescent reporter of IncF, pY was amplified using wel317/wel1029, while mClover3 was amplified from pNCS_mClover372 with primers wel1028/wel1027. Both were ligated to pGen_Lux56, which was amplified with primers wel1030/wel318, to generate pTraY_mClover3. ICE fluorescent reporter was generated by amplifying the predicted promoter upstream traVCICE with primers JW21/JW22. mClover3 was amplified with primers JW24 and wel1027 using pTraY_mClover3 as template. Both fragments were ligated into pGen_Lux amplified with JW23/wel1030 to generate pJW4. The luminescent ICE reporter, pWel78, was generated by ligating the ICE promoter amplified with wel1657/1658 to pGen_Lux, which was amplified with primers wel1659/wel1660. All primers are listed in Table S3. To generate a gentamicin-resistant E. coli str. HS, we used Lambda Red recombination to introduce the resistance cassette in the intergenic region upstream a putative intimin (EcHS_A0351) as described before9,73. Briefly, the resistant cassette was amplified from pCDF_GmFrt using wel1189 and wel1190. The PCR product was used to transform E. coli str. HS expressing the Red recombinase from pKD46_Km. The location of the chromosomal gentamicin resistance cassette was confirmed by PCR.
TraCICE purification
TraCICE with 10 histidine residues at the C-terminus was expressed and purified from E. coli Top10 carrying pTA7. Briefly, cells were grown at 37 °C and expression was induced with 0.5 mM IPTG. After overnight incubation, cells were pelleted down at 8000 rpm, resuspended in 20 mM Tris-HCL (pH 6.5), 150 mM NaCl, 1 mM β-mercaptoethanol, 10 mM imidazole, 0.01% w/v triton X−100, protease inhibitor cocktail (Roche) before lysis using Emulsiflex. Lysate was spun down at 18,000 rpm and filtered (0.45 µm) before loading on the Ni-NTA column. After washing with lysis buffer supplemented with 40 mM imidazole, TraCICE was eluted with a buffer containing 20 mM Tris-HCL (pH 6.5), 75 mM potassium acetate, 1 mM β-mercaptoethanol, 10% w/v glycerol and 0.5 M imidazole. Eluates were combined and further separated using HiLoad Superdex 16/60 200 pg. Purified TraCICE was eluted with 50 mM HEPES (pH 6.5), 75 mM potassium acetate,10% w/v glycerol, 2 mM Magnesium acetate and 0.1 mM EDTA. Protein homogeneity was verified using SDS-PAGE followed by Coomassie staining to visualize the separated proteins.
ATPase assay
Enzymatic activity of TraCICE was tested by adding 1.5 µM purified protein to a buffer containing 50 mM PIPES-NaOH (pH 7), 75 mM K acetate, 5% glycerol, 1 mM KCl, 0.1 mM EDTA, 1 mM DTT, 5 mM ATP and 2.5 mM Mg acetate in a final volume of 20 µl. For heat-inactivated control, an equivalent amount of protein was heated for 5 min at 80 °C. Reactions were incubated at 37 °C for increments of 15 min to a total of 1.5 h to monitor activity over time. Product formation was monitored by adding 3 µl of the reaction to 200 µl malachite green. Absorbance was measured at OD 620 nm using a standard spectrophotometer. Product was quantified using a standard curve developed using increasing concentrations of potassium phosphate monobasic.
Co-purification of TraCICE and TraLIncF
TraCICE with a strep-tag at the C-terminus and untagged TraLIncF were co-expressed and purified from E. coli Top10 carrying pWel75. Briefly, cells were grown at 37 °C and expression was induced with 0.5 mM IPTG. After overnight incubation, cells were pelleted down at 8000 rpm and resuspended in 100 mM Tris-HCL (pH 7.5), 150 mM NaCl, 5 mM β-mercaptoethanol, 1 mM EDTA, 0.1% w/v DDM, and protease inhibitor cocktail (Roche) before lysis using Emulsiflex. Lysate was spun down at 18,000 rpm and filtered (0.45 µm) before loading on the Strep-tactin column (Qiagen). After washing with lysis buffer, TraCICE was eluted with the same buffer supplemented with 2.5 mM desthiobiotin. Eluates were combined and further separated using S75 increase 10/300 GL. Purified TraCICE was eluted with desthiobiotin-free buffer. Eluates were separated using SDS-PAGE followed by Coomassie staining to visualize the co-eluted proteins. Bands corresponding to TraLIncF (~11.9 kDa) were excised and analyzed by mass spectrometry.
Mass spectroscopy-based identification of proteins
In-gel samples were reduced (10 mm BME in 50 mm ammonium bicarbonate), alkylated (55 mM iodoacetamide in 50 mm ammonium bicarbonate), and digested overnight at 37 °C (Promega sequencing-grade modified trypsin). Tryptic peptides were extracted twice from the gel with 30% acetonitrile and 1% formic acid in water. Peptides were separated and analysed using a Vanquish Neo UHPLC system (Thermo Scientific) and an EASY-Spray capillary HPLC column (ES75150, Thermo Scientific) coupled to an Orbitrap Exploris 480 mass spectrometer (Thermo Scientific). The mass spectrometer was operated in data-dependent acquisition mode with a resolution of 60,000 and m/z range of 350–1700. Multiply charged ions were fragmented via HCD dissociation with an NCE of 28, and spectra of their fragments were recorded in the orbitrap at a resolution of 15,000. Data were processed using Proteome Discoverer (Thermo Scientific, version 3.0) and the database (UniProt UP000008614) was searched using SEQUEST (Thermo Scientific). Search parameters included full trypsin cleavage, an allowance for 2 missed cleavages, a minimum peptide length of 6, a strict false discovery rate (FDR) of 0.01, a relaxed FDR of 0.05, a precursor mass tolerance of 10 ppm, and a fragment mass tolerance of 0.02 Da. Peptides were searched with carbamidomethyl cysteine as a static modification and oxidized methionine, and deamidated glutamine and asparagine as dynamic modifications.
Outer membrane core complex purification
The pBAD24_traKIncF, traBHIncF-strep and pASK-IBA3C_traVICE-HA plasmids were co-transformed into BL21(DE3) cells and incubated overnight in LB at 37 °C. The following day, cells were subcultured into 8 L of fresh LB medium, with the OD600 adjusted to 0.1, and grown at 37 °C until the OD600 reached 0.6. Protein overexpression was induced by adding 0.2% w/v arabinose and 200 µg/L anhydrotetracycline. The culture was incubated for an additional 18 h at 18 °C. Cells were harvested by centrifugation at 7,000 x g for 15 min at 4 °C, and the pellet was resuspended in 50 mM Tris-HCl (pH 7), 300 mM NaCl, 0.2 mg/ml lysozyme, and EDTA-free protease inhibitor (Roche). Cells were lysed by passing through a cell disruptor twice at 26,000 psi, and unbroken cells were removed by centrifugation at 30,000 x g for 20 min at 4 °C. The clarified lysate was subjected to ultracentrifugation at 120,000 x g for 2 h at 4 °C to isolate bacterial membranes. Membranes were resuspended in solubilisation buffer (50 mM Tris-HCl, pH 7, 300 mM NaCl, 1% w/v DDM, 0.8% LDAO) and homogenized using a Dounce homogenizer, then allowed to solubilise for 1 h with gentle stirring. The insoluble material was removed by ultracentrifugation at 120,000 x g for 1 h, and the supernatant was applied to a pre-equilibrated StrepTrapXT column (Cytiva) with equilibration buffer (50 mM Tris-HCl, pH 7, 300 mM NaCl, 0.23% LDAO). Proteins were eluted using the equilibration buffer supplemented with 50 mM D-Biotin, and the eluted proteins were further purified using size exclusion chromatography on a Superose 6 10/300 GL column (Cytiva) equilibrated with 50 mM Tris-HCl, pH 7, 300 mM NaCl, 0.23% LDAO. The protein content was analysed by coomassie blue-stained SDS-PAGE and western blotting.
Immunostaining of biofilms and T4SS-promoter expressing AIEC
Caco-2, HEK 293 T, and HeLa cells were grown in Dulbecco’s Modified Eagle Medium (DMEM, Gibco) supplemented with 20% fetal bovine serum (Gibco), 1x minimum non-essential amino acids (Gibco), and 1mM N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid (Gibco). Cells were seeded on coverslips pre-coated with poly-L-lysine and grown to 90% confluency. Caco-2 cells were infected with AIEC str. NRG857c at a multiplicity of infection (MOI) of 1:25 for 1.5 h. Non-adherent bacteria were washed off with 1X PBS twice, then the media was replaced with Iscove’s Modified Dulbecco’s Medium (Gibco) and cells were incubated for an additional 4 h. For T4SS-promoter expressing AIEC strains, cells were incubated for 5 h with Iscove’s Modified Dulbecco’s Medium. At the end of the infection period, cells were washed twice with 1X PBS, then fixed with 3% paraformaldehyde (Sigma-Aldrich) for 30 min at room temperature. HEK 293 T and HeLa cells were grown and infected similarly to assess their ability to trigger biofilm formation by AIEC. To investigate T4SS-promoter expression with non-viable cells, Caco-2 cells were fixed with 3% paraformaldehyde for 30 min at room temperature, then washed three times with 1x PBS prior to infection with AIEC as mentioned previously. To observe T4SS-promoter expression in the absence of epithelial cells, cell-free supernatants were harvested from Caco-2 cells after reaching 100% confluency. Coverslips were blocked overnight with 3% goat serum (Invitrogen) and 1% bovine serum albumin (BioShop) at 4 °C. To stain the cells and bacteria, samples were incubated with rabbit anti-O83 antibody (Statens Serum Institut) for 1 h at room temperature. Coverslips were washed three times in 1x PBST, then stained for 40 min at room temperature using goat anti-rabbit antibody conjugated to Alexa Fluor 488 (Invitrogen) or goat-anti-rabbit antibody conjugated to Alexa Fluor 568 (Invitrogen) and phalloidin Alexa Fluor 568 (Invitrogen), or phalloidin Alexa Fluor 350 (Invitrogen). To stain the nuclei, samples were incubated with DAPI (Invitrogen) for 5 min at room temperature and washed three times with 1x PBST. Biofilm images were captured with BioTek Cytation 5 Imager using a 20X objective. Samples were imaged blindly and processed with BioTek Gen5 software for biofilm quantification. Acquired images were preprocessed then analyzed to quantify aggregates with surface area of 100 µm2 or more, shown previously to represent biofilms formed by pathogenic E. coli16. Images of T4SS-promoter expression under the control of a fluorescent reporter were captured with a Leica Stellaris 8 confocal microscope using a 20X objective (Leica Microsystems). All images were obtained over three independent experiments.
AIEC adhesion to eukaryotic cells
Caco-2, HEK 293 T, or HeLa cells were infected with wild-type or isogenic AIEC mutants at a multiplicity of infection (MOI) of 10:1 for 1.5 h as described previously. Non-adherent bacteria were washed off with 1x PBS three times, then eukaryotic cells were lysed with 1 ml 1% triton X-100 in PBS. Bacteria were plated on selective media before and after infection to quantify cell adhesion by each strain.
Bacterial conjugation
Different AIEC strains, E. coli carrying pOX38, and E. coli str. HS were subcultured in LB supplemented with ampicillin (100 µg/ml) for AIEC and E. coli str. HS and chloramphenicol (34 µg/ml) for E. coli /pOX38 till log phase. To induce protein expression in AIEC strains, 1 mM IPTG was added for 2 h at 37 °C. Strains were washed three times with 1x PBS and normalized to OD600 of 1 before mixing donor and recipients (E. coli str. HS) at a ratio of 5:1. Bacterial mixtures were spotted on nitrocellulose membranes layered over LB agar plates supplemented with 1 mM IPTG and conjugated for 6 h at 37 °C. All conjugation lawns were resuspended in 1 ml sterile PBS and plated on selective media to recover transconjugants and donors separately. As a negative control, donor and recipients were plated separately on the selection media for transconjugants to monitor the emergence of spontaneous resistance.
Western blotting
Protein expression was induced with 1 mM IPTG for 2 h at 37 °C in AIEC. Total protein was standardized and loaded onto 4-20% SDS-PAGE gel in 1X Laemmli buffer. A semi-dry blotting system was employed to transfer the samples onto a nitrocellulose membrane, which was blocked overnight at 4 °C with Intercept blocking buffer (LI-COR). To detect TraCICE, the membrane was incubated for 1 h at room temperature with a primary mouse anti-His antibody (1:4000, GenScript), followed by incubation with a secondary IRDye® 800CW goat anti-mouse antibody (1:16000, LI-COR). Blots were visualized using a LI-COR Imaging System (LI-COR). The TraVICE-HA protein was detected by incubating the membranes with a monoclonal anti-HA antibody (1:2000 dilution, BioLegend) overnight at 4 °C. The next day, the membrane was washed with 0.1% PBST and incubated with a secondary anti-mouse antibody (1:5000 dilution, Sigma-Aldrich) for 1 h at room temperature. Signal detection was carried out using an enhanced chemiluminescence reagent kit (Thermo Fisher), and images were acquired using a ChemiDoc system (Bio-Rad).
Maleimide-based fluorescent labeling of pili
AIEC wild-type, ∆finO, and ∆finO ∆traVCICE strains engineered with chromosomal traAS54C and carrying the different plasmids were grown overnight at 37 °C for 14–16 h. Strains were normalized to OD 1 CFU/ml and subcultured in LB supplemented with ampicillin (100 µg/ml) and kanamycin (50 µg/ml) till log phase. 1 mM IPTG was added for 2 h at 37 °C to induce protein expression. Cells were washed once with 1x PBS, then incubated with 25 µg/ml Maleimide-Alexa Fluor488 (Thermo Fisher) for 30 min in the dark at 4 °C. Samples were washed twice with 1x PBS, then spotted onto 1% agarose pads with 40% glycerol. All slides were captured with a Leica Stellaris 8 confocal microscope using a 63X objective (Leica Microsystems). Images were obtained over three independent experiments.
RT-qPCR
AIEC strains were grown in LB overnight at 37 °C. When needed, strains were subcultured in LB to log phase and protein expression was induced with 1 mM IPTG for 2 h at 37 °C shaking. Nucleic acid was harvested from bacterial pellets using a phenol ethanol-based method, then digested with RNase-free DNase (Thermo Fisher) and added to LunaScript RT SuperMix (New England Biolabs) to obtain cDNA. qPCR was performed using a CFX96™ Real-Time PCR Detection System (Bio-Rad) with SYBR green SuperMix (Quanta Biosciences) and select primers (Table S3). Gene expression was normalized to the bacterial housekeeping gene gapA.
Murine infections
C57BL/6 J mice were orally gavaged with 20 mg/ml streptomycin 20 h prior to infection with 109 cfu/ml of the pre-mixed competing AIEC strains. The bacterial inoculum was plated to obtain an input ratio of the two strains. Mice were euthanized eight days post-infection, and the infected tissues were homogenized and plated on selective media to enumerate each strain. Output ratios were normalized to the input. To monitor luminescence from AIEC carrying ICE luminescent reporter, C57BL/6 J mice were pre-treated with streptomycin as mentioned previously and infected with AIEC carrying pWel78. Two days post-infection, mice were anesthetized (2% isoflurane carried in 2% oxygen) and imaged for 60 s in an IVIS Spectrum in vivo imaging system (PerkinElmer). To investigate the effects of AIEC strains on inflammation, mice were pre-treated with streptomycin and infected with wt or isogenic AIEC mutant strains as mentioned previously. 3% dextran sulfate sodium (Fisher Scientific) was administered in water to mice for 5 days to induce colitis. Mice were sacrificed 2 days post-dextran sulfate sodium, and cecum and colon tissues were harvested. Inflammation was monitored by body weight.
Enzyme-linked immunosorbent assays
Lipocalin-2/NGAL and TNF-α levels were determined in inflamed gut compartments using R&D Systems DuoSet ELISA. Briefly, a high-binding 96-well plate was coated with capture antibody overnight at room temperature. Wells were blocked with 1% BSA in PBS, incubated with samples and standard, then detection antibody, streptavidin-HRP, and a 1:1 mixture of Color Reagent A (H2O2) and Color Reagent B (tetramethylbenzidine) sequentially. Reaction was stopped with 2 N H2SO4. Plates were washed with 0.05% Tween 20 in PBS three times between each step. Absorbance was read at 450 nm and corrected with 540 nm.
Statistics and reproducibility
For analyses comparing two independent groups, unpaired two-tailed Mann Whitney U-tests were used at a 95% confidence interval. Shapiro-Wilk normality test was conducted on datasets with 30 or more datapoints. In such cases, Kruskal-Wallis test followed by Dunn’s multiple comparisons test was used to compare groups when the data did not follow the normal distribution. When datapoints were not sufficient to test normality, data distribution was assumed to be normal, and one-way ANOVA with Dunnett’s multiple comparisons test was used to compare groups. A p value < 0.05 indicates significance for all results. Data was analyzed and generated using GraphPad Prism 10 (GraphPad software). Number of replicates or sample size used are listed in figure legends. Where not stated, all results were obtained from at least three independent experiments.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The mass spectrometry data generated in this study have been deposited in the PRIDE database under accession code PXD066462. The microscopy image data generated in this study have been deposited in the Figshare repository at (https://doi.org/10.6084/m9.figshare.29634554.v1). All datasets are open access and publicly available. Source data are provided with this paper.
References
Wright, E. K. et al. Recent advances in characterizing the gastrointestinal microbiome in Crohn’s disease: a systematic review. Inflamm. Bowel Dis. 21, 1219–1228 (2015).
Mansour, S., Asrar, T. & Elhenawy, W. The multifaceted virulence of adherent-invasive Escherichia coli. Gut Microbes https://doi.org/10.1080/19490976.2023.2172669 (2023).
Darfeuille-Michaud, A. et al. Presence of adherent Escherichia coli strains in ileal mucosa of patients with Crohn’s disease. Gastroenterology 115, 1405–1413 (1998).
Darfeuille-Michaud, A. et al. High prevalence of adherent-invasive Escherichia coli associated with ileal mucosa in Crohn’s disease. Gastroenterology 127, 412–421 (2004).
Nadalian, B. et al. Prevalence of the pathobiont adherent-invasive Escherichia coli and inflammatory bowel disease: a systematic review and meta-analysis. J. Gastroenterol. Hepatol. 36, 852–863 (2021).
Small, C.-L. N., Reid-Yu, S. A., McPhee, J. B. & Coombes, B. K. Persistent infection with Crohn’s disease-associated adherent-invasive Escherichia coli leads to chronic inflammation and intestinal fibrosis. Nat. Commun. 4, 1957 (2013).
Chassaing, B., Koren, O., Carvalho, F. A., Ley, R. E. & Gewirtz, A. T. AIEC pathobiont instigates chronic colitis in susceptible hosts by altering microbiota composition. Gut 63, 1069–1080 (2014).
Kittana, H. et al. Evidence for a causal role for Escherichia coli strains identified as adherent-invasive (AIEC) in intestinal inflammation. mSphere 8, e00478-22 (2023).
Elhenawy, W., Tsai, C. N. & Coombes, B. K. Host-specific adaptive diversification of Crohn’s disease-associated adherent-invasive Escherichia coli. Cell Host Microbe 25, 301–312.e5 (2019).
Glasser, A. L. et al. Adherent invasive Escherichia coli strains from patients with Crohn’s disease survive and replicate within macrophages without inducing host cell death. Infect. Immun. 69, 5529–5537 (2001).
Cieza, R. J., Hu, J., Ross, B. N., Sbrana, E. & Torres, A. G. The IbeA invasin of adherent-invasive Escherichia coli mediates interaction with intestinal epithelia and macrophages. Infect. Immun. 83, 1904–1918 (2015).
Bringer, M.-A., Barnich, N., Glasser, A.-L., Bardot, O. & Darfeuille-Michaud, A. HtrA stress protein is involved in intramacrophagic replication of adherent and invasive Escherichia coli Strain LF82 isolated from a patient with crohn’s disease. Infect. Immun. 73, 712–721 (2005).
De la Fuente, M. et al. Escherichia coli isolates from inflammatory bowel diseases patients survive in macrophages and activate NLRP3 inflammasome. Int. J. Med. Microbiol. 304, 384–392 (2014).
Bretin, A. et al. Activation of the EIF2AK4-EIF2A/eIF2α-ATF4 pathway triggers autophagy response to Crohn’s disease-associated adherent-invasive Escherichia coli infection. Autophagy 12, 770–783 (2016).
Lapaquette, P., Glasser, A.-L., Huett, A., Xavier, R. J. & Darfeuille-Michaud, A. Crohn’s disease-associated adherent-invasive E. coli are selectively favoured by impaired autophagy to replicate intracellularly. Cell. Microbiol. 12, 99–113 (2010).
Elhenawy, W. et al. High-throughput fitness screening and transcriptomics identify a role for a type IV secretion system in the pathogenesis of Crohn’s disease-associated Escherichia coli. Nat. Commun. 12, 2032 (2021).
Christie, P. J., Gomez Valero, L. & Buchrieser, C. Biological Diversity and Evolution of Type IV Secretion Systems. in Type IV Secretion in Gram-Negative and Gram-Positive Bacteria (eds. Backert, S. & Grohmann, E.) 413, 1–30 (Springer International Publishing, Cham, 2017).
Costa, T. R. D., Patkowski, J. B., Macé, K., Christie, P. J. & Waksman, G. Structural and functional diversity of type IV secretion systems. Nat. Rev. Microbiol. https://doi.org/10.1038/s41579-023-00974-3 (2023).
Shen, X. et al. Identification of atypical T4SS effector proteins mediating bacterial defense. mLife 2, 295–307 (2023).
Flemming, H.-C. et al. Biofilms: an emergent form of bacterial life. Nat. Rev. Microbiol 14, 563–575 (2016).
Patkowski, J. B. et al. The F-pilus biomechanical adaptability accelerates conjugative dissemination of antimicrobial resistance and biofilm formation. Nat. Commun. 14, 1879 (2023).
Ghigo, J. M. Natural conjugative plasmids induce bacterial biofilm development. Nature 412, 442–445 (2001).
Costa, T. R. D. et al. Structure of the bacterial sex F pilus reveals an assembly of a stoichiometric protein-phospholipid complex. Cell 166, 1436–1444.e10 (2016).
Frost, L. S. & Koraimann, G. Regulation of bacterial conjugation: balancing opportunity with adversity. Future Microbiol 5, 1057–1071 (2010).
Silverman, P. M., Wickersham, E. & Harris, R. Regulation of the F plasmid tra Y promoter in Escherichia coli by host and plasmid factors. J. Mol. Biol. 218, 119–128 (1991).
Will, W. R. & Frost, L. S. Characterization of the Opposing Roles of H-NS and TraJ in Transcriptional Regulation of the F-Plasmid tra Operon. J. Bacteriol. 188, 507–514 (2006).
Amin, H., Ilangovan, A. & Costa, T. R. D. Architecture of the outer-membrane core complex from a conjugative type IV secretion system. Nat. Commun. 12, 6834 (2021).
Khara, P., Song, L., Christie, P. J. & Hu, B. In situ visualization of the pKM101-encoded type IV Secretion System Reveals a Highly Symmetric ATPase Energy Center. mBio 12, e02465–21 (2021).
Macé, K. et al. Cryo-EM structure of a type IV secretion system. Nature 607, 191–196 (2022).
Eaves-Pyles, T. et al. Escherichia coli isolated from a Crohn’s disease patient adheres, invades, and induces inflammatory responses in polarized intestinal epithelial cells. Int. J. Med. Microbiol. 298, 397–409 (2008).
Delavat, F., Miyazaki, R., Carraro, N., Pradervand, N. & Van Der Meer, J. R. The hidden life of integrative and conjugative elements. FEMS Microbiol. Rev. 41, 512–537 (2017).
Botelho, J. & Schulenburg, H. The role of integrative and conjugative elements in antibiotic resistance evolution. Trends Microbiol. 29, 8–18 (2021).
Wozniak, R. A. F. & Waldor, M. K. Integrative and conjugative elements: mosaic mobile genetic elements enabling dynamic lateral gene flow. Nat. Rev. Microbiol 8, 552–563 (2010).
Costa, T. R. D. et al. Type IV secretion systems: advances in structure, function, and activation. Mol. Microbiol 115, 436–452 (2021).
Shala-Lawrence, A., Bragagnolo, N., Nowroozi-Dayeni, R., Kheyson, S. & Audette, G. F. The interaction of TraW and TrbC is required to facilitate conjugation in F-like plasmids. Biochemical Biophysical Res. Commun. 503, 2386–2392 (2018).
Klimke, W. A. et al. The mating pair stabilization protein, TraN, of the F plasmid is an outer-membrane protein with two regions that are important for its function in conjugation. Microbiology 151, 3527–3540 (2005).
Altschul, S. F. et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389–402 (1997).
Liu, M. et al. ICEberg 2.0: an updated database of bacterial integrative and conjugative elements. Nucleic Acids Res. 47, D660–D665 (2019).
Juhas, M. et al. Novel type IV secretion system involved in propagation of genomic islands. J. Bacteriol. 189, 761–771 (2007).
Abramson, J. et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630, 493–500 (2024).
Holm, L., Laiho, A., Törönen, P. & Salgado, M. DALI shines a light on remote homologs: one hundred discoveries. Protein Sci. 32, e4519 (2023).
Walldén, K. et al. Structure of the VirB4 ATPase, alone and bound to the core complex of a type IV secretion system. Proc. Natl. Acad. Sci. Usa. 109, 11348–11353 (2012).
Narita, S. & Tokuda, H. Bacterial lipoproteins; biogenesis, sorting and quality control. Biochimica et. Biophysica Acta (BBA) - Mol. Cell Biol. Lipids 1862, 1414–1423 (2017).
Low, H. H. et al. Structure of a type IV secretion system. Nature 508, 550–553 (2014).
Mossey, P., Hudacek, A. & Das, A. Agrobacterium tumefaciens type IV secretion protein VirB3 is an inner membrane protein and requires VirB4, VirB7, and VirB8 for stabilization. J. Bacteriol. 192, 2830–2838 (2010).
Fowler, T., Taylor, L. & Thompson, R. The control region of the F plasmid transfer operon: DNA sequence of the traJ and traY genes and characterisation of the traY → Z promoter. Gene 26, 79–89 (1983).
Lu, J. et al. Cooperative function of TraJ and ArcA in regulating the F plasmid tra operon. J. Bacteriol. 201, e00448-18 (2019).
Frost, L., Lee, S., Yanchar, N. & Paranchych, W. finP and fisO mutations in FinP anti-sense RNA suggest a model for FinOP action in the repression of bacterial conjugation by the Flac plasmid JCFLO. Mol. Gen. Genet 218, 152–160 (1989).
Gubbins, M. J., Will, W. R. & Frost, L. S. The F-plasmid, a paradigm for bacterial conjugation. In The Dynamic Bacterial Genome (ed. Mullany, P.) 151–206 (Cambridge University Press, 2005).
Finnegan, D. J. & Willetts, N. S. Two classes of Flac mutants insensitive to transfer inhibition by an F-like R factor. Molec. Gen. Genet. 111, 256–264 (1971).
Wong, J. J. W., Lu, J. & Glover, J. N. M. Relaxosome function and conjugation regulation in F-like plasmids – a structural biology perspective. Mol. Microbiol. 85, 602–617 (2012).
Cheah, K.-C. & Skurray, R. The F plasmid carries an IS3 Insertion Within finO. Microbiology 132, 3269–3275 (1986).
Hu, B., Khara, P. & Christie, P. J. Structural bases for F plasmid conjugation and F pilus biogenesis in Escherichia coli. Proc. Natl. Acad. Sci. Usa. 116, 14222–14227 (2019).
Schandel, K. A., Maneewannakul, S., Ippen-Ihler, K. & Webster, R. E. A traC mutant that retains sensitivity to f1 bacteriophage but lacks F pili. J. Bacteriol. 169, 3151–3159 (1987).
Goldlust, K. et al. The F pilus serves as a conduit for the DNA during conjugation between physically distant bacteria. Proc. Natl. Acad. Sci. Usa. 120, e2310842120 (2023).
Lane, M. C., Alteri, C. J., Smith, S. N. & Mobley, H. L. T. Expression of flagella is coincident with uropathogenic Escherichia coli ascension to the upper urinary tract. Proc. Natl. Acad. Sci. 104, 16669–16674 (2007).
Chan, K., Kim, C. C. & Falkow, S. Microarray-based detection of Salmonella enterica serovar typhimurium transposon mutants that cannot survive in macrophages and mice. Infect. Immun. 73, 5438–5449 (2005).
Chassaing, B., Aitken, J. D., Malleshappa, M. & Vijay-Kumar, M. Dextran sulfate sodium (DSS)-Induced Colitis in Mice. Current Protocols in Immunology 104, 15.25.1-15.25.14 (2014).
Kitamoto, S. et al. Dietary l-serine confers a competitive fitness advantage to Enterobacteriaceae in the inflamed gut. Nat. Microbiol 5, 116–125 (2020).
Alvarez-Martinez, C. E. & Christie, P. J. Biological diversity of prokaryotic type IV Secretion Systems. Microbiol Mol. Biol. Rev. 73, 775–808 (2009).
Thurlow, L. R. et al. Staphylococcus aureus biofilms prevent macrophage phagocytosis and attenuate inflammation in vivo. J. I. 186, 6585–6596 (2011).
Domenech, M., Ramos-Sevillano, E., García, E., Moscoso, M. & Yuste, J. Biofilm formation avoids complement immunity and phagocytosis of Streptococcus pneumoniae. Infect. Immun. 81, 2606–2615 (2013).
Koraimann, G. et al. Spread and persistence of virulence and antibiotic resistance genes: a ride on the F plasmid conjugation module. EcoSal Plus 8, https://doi.org/10.1128/ecosalplus.ESP-0003-2018 (2018).
Barrozo, R. M. et al. Functional plasticity in the type IV secretion system of helicobacter pylori. PLoS Pathog. 9, e1003189 (2013).
Kishida, K. et al. Chimeric systems composed of swapped Tra subunits between distantly-related F plasmids reveal striking plasticity among type IV secretion machines. PLoS Genet 20, e1011088 (2024).
Armstrong, G. D., Frost, L. S., Sastry, P. A. & Paranchych, W. Comparative biochemical studies on F and EDP208 conjugative pili. J. Bacteriol. 141, 333–341 (1980).
Arutyunov, D. & Frost, L. S. F conjugation: back to the beginning. Plasmid 70, 18–32 (2013).
Swidsinski, A., Weber, J., Loening-Baucke, V., Hale, L. P. & Lochs, H. Spatial organization and composition of the mucosal flora in patients with inflammatory bowel disease. J. Clin. Microbiol 43, 3380–3389 (2005).
Gilchrist, C. L. M. & Chooi, Y.-H. clinker & clustermap.js: automatic generation of gene cluster comparison figures. Bioinformatics 37, 2473–2475 (2021).
Dykxhoorn, D. M., St. Pierre, R. & Linn, T. A set of compatible tac promoter expression vectors. Gene 177, 133–136 (1996).
Edwards, R. A., Keller, L. H. & Schifferli, D. M. Improved allelic exchange vectors and their use to analyze 987P fimbria gene expression. Gene 207, 149–157 (1998).
Bajar, B. T. et al. Improving brightness and photostability of green and red fluorescent proteins for live cell imaging and FRET reporting. Sci. Rep. 6, 20889 (2016).
Datsenko, K. A. & Wanner, B. L. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. Usa. 97, 6640–6645 (2000).
Acknowledgements
W.E. holds a Canada Research Chair in Microbiome Research (CRC-00249). This work is funded by Crohn’s & Colitis Canada (GAR 1034447), Natural Sciences and Engineering Research Council of Canada (DGECR-2023-00140), Major Innovation Fund-Antimicrobial Resistance - One Health Consortium (RES0059404), Canadian Institutes of Health Research (RES0069043), LKSIoV Research Support and Innovation Grant (RES0057474), and Striving for Pandemic Preparedness - The Alberta Research Consortium (SPP-ARC, RES0066221). W.E. is funded by Canada Foundation for Innovation - JELF (CFI 43658). J.W. is supported by a federal CGS-M. T.R.D.C. is supported by the Welcome Trust Award 215164/Z/18/Z. Experiments were performed at the University of Alberta Faculty of Medicine & Dentistry Cell Imaging Core, RRID:SCR_019200, which receives financial support from the Faculty of Medicine & Dentistry, the University Hospital Foundation, Striving for Pandemic Preparedness – The Alberta Research Consortium, and Canada Foundation for Innovation (CFI) awards to contributing investigators.
Author information
Authors and Affiliations
Contributions
J.W. and W.E. conceptualized the study and designed the experiments. J.W., A.R., P.V., N.L., and J.O. performed the experiments. J.W. and W.E. wrote the manuscript. A.R., P.V., J.P., N.L., J.O., K.D., B.R., S.T., M.C., A.T., D.K., B.W., X.C., and T.R.D.C. contributed to revising the manuscript.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Carrie Shaffer, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Wong, J., Refaat, A., Villacampa-Teixeira, P. et al. The assembly of a hybrid type IV secretion system by a Crohn’s disease-associated Escherichia coli strain. Nat Commun 16, 8797 (2025). https://doi.org/10.1038/s41467-025-63859-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-025-63859-4
