DNA Transport through the Dynamic Type IV Secretion System

ABSTRACT

The versatile type IV secretion system (T4SS) nanomachine plays a pivotal role in bacterial pathogenesis and the propagation of antibiotic resistance determinants throughout microbial populations. In addition to paradigmatic DNA conjugation machineries, diverse T4SSs enable the delivery of multifarious effector proteins to target prokaryotic and eukaryotic cells, mediate DNA export and uptake from the extracellular milieu, and in rare examples, facilitate transkingdom DNA translocation. Recent advances have identified new mechanisms underlying unilateral nucleic acid transport through the T4SS apparatus, highlighting both functional plasticity and evolutionary adaptations that enable novel capabilities. In this review, we describe the molecular mechanisms underscoring DNA translocation through diverse T4SS machineries, emphasizing the architectural features that implement DNA exchange across the bacterial membrane and license transverse DNA release across kingdom boundaries. We further detail how recent studies have addressed outstanding questions surrounding the mechanisms by which nanomachine architectures and substrate recruitment strategies contribute to T4SS functional diversity.

INTRODUCTION

Bacterial type IV secretion systems (T4SS) are incredibly versatile molecular devices harbored by phylogenetically diverse Gram-negative and Gram-positive organisms (1–5). Studies exploring the function and architecture of these nanomachines have largely focused on paradigmatic Gram-negative systems that are deployed to translocate protein, DNA, multisubunit toxin, and other macromolecular substrates to target cells (5). The fascinating T4SS superfamily can be broadly segregated into three functional subtypes that uniquely contribute to pathogenicity and genome plasticity (1–3). Bacterial conjugation machineries represent the largest and most wide-spread T4SS subfamily. These molecular devices enable the export of single-stranded DNA (ssDNA) and associated proteins across the donor bacterial membrane and into target prokaryotic and eukaryotic cells via mechanisms that require direct cell-cell contact (1–6). First described by Lederberg and Tatum nearly 80 years ago (7), T4SSs facilitate rapid microbial adaptation to challenging environments through the acquisition of fitness traits encoded within mobile genetic elements. Notably, conjugative DNA transfer is the predominant mechanism by which antibiotic resistance determinants and virulence factors are transferred within and between phylogenetically diverse species (2, 8–12).

Specialized modifications to “minimized” conjugation systems may enable novel secretion phenotypes or expanded target cell range (5). In contrast to conjugation machineries, the “effector translocator” T4SS subfamily is used to deliver assorted protein effectors into the cytoplasm of eukaryotic host cells in a contact-dependent manner (1, 2, 4). In some cases, effector translocators implement the interkingdom transfer of nucleoprotein and polysaccharide substrates (1, 4, 6, 13–16). Thus, effector translocator systems play a critical role in establishing host-pathogen interactions that promote host tissue colonization and foster bacterial persistence within distinct niches (1, 2). In unusual cases, T4SS-dependent mechanisms facilitate interbacterial killing via toxin delivery (17, 18) or the contact-independent secretion of multisubunit protein toxins into the extracellular environment (2, 19). The third T4SS subfamily comprised of “DNA uptake and release” systems function in a contact-independent manner to import exogenous DNA or to secrete DNA into the extracellular milieu (1, 2, 5, 14, 20–23). Thus, these dedicated genetic exchange systems expand the molecular arsenal employed by Gram-negative bacteria to acquire survival elements during infection (14, 20, 24).

Recent progress enabled by single particle cryo-electron microscopy (cryo-EM), in situ cryo-electron tomography (cryo-ET), and high-resolution correlative fluorescence microscopy has yielded new insights into T4SS assembly dynamics and spatial organization in the bacterial envelope under near-native cellular conditions (5, 25–31). In Gram-negative organisms, minimized bi-membrane spanning T4SSs are composed of approximately 12 conserved components designated VirB1 through VirB11 and VirD4 in accordance with a unified nomenclature based on the prototypical Agrobacterium tumefaciens vir T4SS (4, 5, 32). In these simplified systems, the apparatus inner membrane complex (IMC) subassembly forms part of the periplasm-spanning translocation channel and connects to the outer membrane-associated core complex (OMCC) assembly via a stalk-like structure (30, 31, 33, 34). In some systems, elongated VirB10 domains bridge the OMCC and IMC (5, 30, 31). A trio of ATPases, including the VirD4 coupling protein that tethers DNA and protein effectors to the secretion channel (5, 35–38), form the energetic apparatus at the cytoplasmic face of the inner membrane. Our structural understanding of model T4SS architectures has also revealed intriguing symmetry mismatch at the junction between the IMC and OMCC or between various rings comprising the OMCC that may afford conformational flexibility or secretion channel gating as a mechanism to orchestrate effector selection and directional translocation (5, 30, 31) (Fig. 1). Similar symmetry mismatch has been identified in several “expanded” effector translocator systems by in situ cryo-electron tomography and single particle reconstruction of isolated T4SS complexes (17, 39–44), highlighting remarkable structural conservation across the T4SS superfamily (Fig. 1). In this review, we summarize mechanisms underlying T4SS-dependent DNA transport across the bacterial envelope and highlight how recent studies have broadened our understanding of T4SS architectural and biological diversity.

FIG 1.

Symmetry mismatch in minimized and expanded T4SS architectures. High-resolution structures of isolated minimized (R388 [30, 31]) and expanded (pED208 [43, 44] and cag T4SS [39, 41]) core complex machineries reveal asymmetric regions within apparatus subassemblies. The minimized R388 core complex is characterized by a tetradecameric O-layer composed of VirB7 and the C-terminal domains of VirB10 (VirB10_CTD) and VirB9 (VirB9_CTD) displaying 14-fold symmetry, and a corresponding hexadecameric I-layer composed of VirB10_NTD and VirB9_NTD exhibiting 16-fold symmetry (30) (PDB: 7O3J and 7O3T). The iconic F plasmid (encoded by pED208) OMCC forms two radial, concentric rings characterized by a peripheral O-layer ring exhibiting 13-fold symmetry and an interior central cone/inner ring (CC/IR) with 17-fold symmetry (43, 44) (PDB: 7SPB and 7SPC). The expanded H. pylori cag T4SS outer membrane complex (OMC, 14-fold rotational symmetry) and corresponding periplasmic ring complex (PRC, 17-fold rotational symmetry) exhibit striking symmetry mismatch (39, 41) (PDB: 6X6S and 6X6J). The figure illustrates conserved architectural asymmetry among diverse T4SS machineries.

INTERBACTERIAL DNA DELIVERY SYSTEMS

Pioneering work performed over the past 75 years has uncovered mechanisms underscoring targeted DNA excision and recruitment of corresponding nucleoprotein substrates to conjugative T4SS machineries. Intricate biochemical details involved in DNA processing have been extensively characterized in vitro and in vivo (45). The process of conjugative DNA transfer initiates when accessory factors, which are typically members of ribbon-helix-helix DNA binding proteins (such as TraM from F-family plasmids [46], or TrwA from plasmid R388 [47]), bind the cognate origin of transfer (oriT) sequence located on either an extrachromosomal plasmid or integrative conjugative element (ICE). A combination of double-stranded DNA bending and direct protein-protein interactions mediated by oriT-bound accessory factors (46, 48, 49) recruits the relaxase (a specialized site- and strand-specific phosphodiesterase [11]) to form the catalytically active relaxosome complex (45, 50). Within the relaxosome, the relaxase initiates substrate processing by triggering an active site-mediated nucleophilic attack of the oriT scissile phosphate group, forming a high-energy bond which covalently links excised ssDNA product to the relaxase (1, 5, 45, 50, 51). This bond enables (i) directional nucleoprotein piloting through the T4SS conjugative apparatus, (ii) protects the phosphate group from subsequent nucleophilic attack in the recipient cell, and (iii) facilitates plasmid ends rejoining in the host cell cytoplasm (45, 50).

Relaxases are large, multidomain proteins that typically contain an N-terminal transesterase domain (the “relaxase”) which carries out the phosphodiesterase reaction (9, 11, 51). Domains localized to the C-terminal region may be involved in DNA helicase or primase activities or execute unknown functions (11, 52). While conjugative relaxases are categorized into eight mobility (“MOB”) groups (53), recent efforts have uncovered the intricate biochemical details underlying the function of MOB_F nuclease associated with F-family conjugation systems (e.g., TraI associated with F, pKM101, RP4, R1, pED208, and TrwC encoded by R388) (11, 54). In addition to the N-terminal transesterase domain that facilitates the nicking and covalent attachment of transfer DNA to the relaxase (55), TraI exhibits a C-terminal vestigial helicase domain that binds ssDNA (56), and an active helicase domain that directionally unwinds DNA 5′ to 3′ (57, 58). An additional TraI C-terminal domain of unknown function may be used to recruit relaxosome components to target DNA sequences (57, 58). Recent work determined the structure of TraI complexed to the F plasmid-encoded oriT at atomic resolution (59). Cryo-EM analysis revealed that ssDNA binds deep within the structure of TraI, with the 5′ region bound longitudinally to the transesterase and active helicase domains and the 3′ region extending through the active and vestigial helicase domains, thus facilitating remarkably high enzymatic processivity (11, 59). Structural analysis also mapped two translocation signals proximal to the TraI C terminus that likely contribute to relaxosome recruitment to the cognate VirD4 coupling protein for transport through the secretion apparatus (11, 59–61).

Coupling proteins are phylogenetically and functionally related to SpoIIIE and FtsK ATPases that translocate DNA across membranes during sporulation and cell division (9, 51, 62). Structural data and biochemical evidence indicate that VirD4 ATPases assemble into a homohexamer with an N-terminal transmembrane domain, a nucleotide binding/hydrolysis domain (63, 64), and a sequence-variable domain that confers substrate binding and specificity (9, 65–67). Additional variable C-terminal domains may provide additional coupling specificity (46) or control the selection and translocation of substrates through the T4SS (68, 69). Relaxosome docking to VirD4 coupling proteins and the T4SS apparatus is achieved through cooperative interactions among the relaxase and accessory factors (11, 50, 51, 70). For example, the F plasmid accessory factor TraM binds a short C-terminal motif within the VirD4-like coupling protein TraD (46, 49), whereas internal TraI translocation signals are accessible only when the relaxase adopts defined tertiary structures (59, 60). Other relaxases may harbor C-terminal translocation signals that facilitate relaxase-VirD4 receptor interactions (51, 61); however, the structural basis of such interactions remains undefined. Additional specialized accessory factors, including ParA-like DNA partitioning proteins, physically tether the relaxosome to VirD4 receptors via multiple protein-protein interactions (71, 72). Subsequent to VirD4 docking, the relaxase is unfolded via unknown mechanisms (73) and released from associated accessory factors for delivery through the translocation channel (1, 11, 51). In the recipient cell cytoplasm, the relaxase catalyzes the recircularization of transferred cargo through a reversal of the strand-breaking reaction, followed by second-strand synthesis and DNA replication (74, 75).

Our understanding of conjugative T4SS architectures primarily stems from structures resolved for the pKM101-encoded OMCC (34, 76, 77) and nearly intact R388 systems (30, 31). The OMCC assumes a barrel-shaped structure incorporating 14 copies of the OM-associated VirB7 lipoprotein, VirB9, and the VirB10 C-terminal region. Crystallographic analyses of the barrel outer (O) layer revealed VirB10 domains lining the interior and connected to an outer protective crown composed of VirB7 and VirB9 (77). The barrel inner (I) layer is formed by VirB9 N-terminal domains surrounding VirB10, which extends across the inner membrane (34, 78). The IMC of conjugative nanomachines exhibits remarkable asymmetry and features two side-by-side VirB4 ATPase hexamers extending from the IMC into the cytoplasm. The IMC comprises 12 copies of VirB3, VirB4, VirB5, VirB6, and VirB8, as well as 14 copies of the VirB10 N-terminal region (5, 11, 33, 51). Contacts between VirB10 and additional subunits within the IMC thus form the symmetry mismatch within the conjugative machinery that may enable functional plasticity (5, 51) (Fig. 1).

Conjugation systems harbored by Gram-negative bacteria elaborate flexible, dynamic conjugative pili that extend from the donor cell and retract to bring potential recipient cells into proximity (5, 79). Other conjugative systems assemble brittle pili that are shed from the cell surface to induce bacterial aggregation to enable DNA transfer (51). Our understanding of conjugative pilus biogenesis and assembly has been significantly advanced by recent structural analyses of purified pili and in situ cryo-ET studies of intact T4SS machineries (30, 80, 81). Analyses of assembled machineries within the E. coli cell envelope revealed multiple F-encoded pilus configurations docked onto alternative basal platforms (80), providing the first direct evidence that, in contrast to other cell surface pili and type III secretion system needle complexes which nucleate on an inner membrane platform, the T4SS pilus assembles at the outer membrane (80). Recent cryo-EM studies of the R388 system provide a structural model for pilus biogenesis whereby VirB2 pilin pentamers bound by VirB6 subunits in the inner membrane are leveraged into the pilus assembly site through the O-layer channel (30).

Despite major advances in our understanding of conjugative pilus assembly, the role pili in DNA transfer has not been firmly established. Evidence suggests that conjugative pili can serve as translocation conduits through which ssDNA and unfolded relaxases can be delivered into recipient cells at long range (82). In support of this hypothesis, the E. coli F pilus (81) and A. tumefaciens T pilus (83, 84) lumens are lined by inner membrane-derived phospholipids which impart a weak negative charge that may facilitate DNA passage (81). Conflicting observations support an indirect role of pili in substrate delivery to target cells. For example, multiple studies identified secretion “uncoupling” mutations that selectively block T4SS pilus biogenesis but do not obstruct substrate transfer, suggesting that elongated pili are not required for cargo translocation (32, 78, 85). Additionally, T4SSs harbored by Gram-positive bacteria do not assemble conjugative pili, yet these systems transfer DNA to recipient cells at high frequencies (1, 3). Although luminal phospholipids may facilitate low-efficiency conjugation, these lines of evidence suggest that conjugative pili serve primarily as adhesive organelles that establish tight donor-recipient mating junctions required for cell contact-dependent DNA transfer (5, 51).

Mating channels may have initially served as protein transport systems that were later adapted into conjugation machineries following the recruitment of coupling proteins associated with rolling circle replicases (52). Based on the inferred evolutionary relationships between coupling proteins and cognate VirB4 ATPases, conjugative T4SSs emerged in Gram-negative organisms and were later domesticated by archaea and monoderm Gram-positive species (3, 86). Studies of T4SSs encoded on the Enterococcus faecalis sex-pheromone responsive plasmid pCF10, C. perfringens plasmid pCW3, and the S. agalactiae broad host-range plasmid pIP501 revealed that in contrast to orthologous systems in Gram-negative bacteria, Gram-positive conjugation systems lack paradigmatic OMCC subunits and the VirB11 ATPase (3, 9, 87–89). In addition to employing surface adhesins to mediate donor-recipient cell contacts (as opposed to conjugative pili), Gram-positive machineries utilize VirB1-like lytic transglycosylases with multiple catalytic domains to enable assembly of the machine across the dense peptidoglycan layer (3, 90).

Mounting evidence suggests substrate transfer across the cytoplasmic membrane is mechanistically and structurally conserved among Gram-negative and Gram-positive bacteria. For example, early substrate processing and coupling protein docking reactions closely mirror corresponding mechanisms underscoring DNA delivery in Gram-negative systems. As in Gram-negative systems, Gram-positive plasmid- and ICE-encoded relaxases display strand-specific cleavage of cognate oriT regions and may coordinate with accessory factors to mediate docking to the coupling protein (1, 3, 9, 91). Current models of DNA coupling posit that Gram-positive relaxases form a transient covalent adduct to excised ssDNA substrates via conserved active site tyrosine residues; however, in contrast to Gram-negative relaxases, translocation signals that enable nucleoprotein transfer intermediate delivery to cognate coupling proteins have not been defined in Gram-positive orthologs (1, 9, 91). Likewise, comparison of multiple Gram-positive conjugative systems identified six subunits that exhibit homology to the A. tumefaciens VirD4, VirB1, VirB3, VirB4, VirB6, and VirB8 channel components (1, 3). Currently, limited structural and biochemical analyses support the assumption that orthologous Gram-positive T4SS subunits facilitate substrate transfer across the cytoplasmic membrane (1, 3). While high resolution structures of individual components encoded within the pIP501 and pCW3 systems have been determined, Gram-positive conjugation machinery architectures have not been visualized in situ. Thus, future work to resolve the structures of Gram-positive T4SS assemblies by cryo-EM will significantly advance our understanding of conjugative DNA transfer through evolutionarily diverse and minimized machineries.

DNA EXPORT INTO THE EXTRACELLULAR MILIEU

The obligate human pathogen Neisseria gonorrhoeae employs a contact-independent T4SS apparatus to secrete chromosomally derived ssDNA substrates into the extracellular milieu that facilitate the robust transformation of other highly competent Neisseria species (21, 92–94). In addition to contributing to exceptional genetic diversity, secreted ssDNA promotes host colonization by stimulating the initial stages of biofilm formation to establish persistent N. gonorrhoeae infection (93). Encoded on the 59 kb gonococcal genetic island (GGI), the specialized DNA release apparatus is harbored by approximately 80% of N. gonorrhoeae and 17.5% of N. meningitis clinical isolates (21, 92, 95, 96). The GGI harbors T4SS structural homologs that share sequence similarities to the E. coli F plasmid conjugation system and other T4SS machineries (including TraB, TraK, and TraV, which comprise the outer membrane core complex, the inner membrane proteins TraL and TraE, and the periplasmic components TraW, TraU, TraH, and TraN) as well as highly conserved components involved in DNA processing and recruitment, including partitioning proteins (ParA and ParB) and the relaxase TraI (21, 92, 97). Additional T4SS components, such as the coupling protein TraD (98) and TraC, a homolog of the ATPase VirB4, are also required for DNA secretion (21, 92, 99–102). Interestingly, the pilin homolog TraA and TbrI, which exhibits homology to a serine protease that plays a role in circularization of the pilin subunit, are not required for DNA secretion, indicating that DNA release into the extracellular space is not dependent on formation of an extracellular pilus (100). This suggests that gonococcal homologs of T4SS components involved in pilus assembly and extension (TraW, TraU, TraH, TraF, TrbC, and TraN) play unique roles within the Neisseria T4SS (101). Other GGI-encoded proteins required for ssDNA secretion include peptioglycanases and the disulfide isomerase DsbC (21, 92, 100, 103).

N. gonorrhoeae ssDNA processing and translocation parallels interbacterial DNA conjugation mechanisms. For example, chromosomal DNA processing relies on TraI relaxase-mediated DNA cleavage at oriT-like sequences and recruitment to the secretion apparatus by the cognate coupling protein TraD for transport across the inner and outer membranes (21, 97, 98). Experimental evidence supports a model whereby secreted ssDNA is delivered across the bacterial envelope as a nucleoprotein complex that is 5′ protected (presumably mediated by covalently bound TraI relaxase) and is susceptible to nucleases that directionally cleave ssDNA from 3′ free ends (97). Amino acid substitutions identified two tyrosine residues in TraI that are functionally important for DNA secretion — substitution of Tyr⁹³ resulted in the complete loss of DNA secretion while Tyr²⁰¹ mutagenesis resulted in partial loss of DNA secretion, demonstrating that both residues are necessary for TraI-dependent DNA processing (97). These data suggest that Tyr⁹³ is important for the initial cleavage of DNA while Tyr²⁰¹ may be involved in a second cleavage that terminates DNA processing (97). Because tyrosine residues contribute to nonspecific DNA interactions, Tyr²⁰¹ may enhance TraI affinity for DNA backbone binding (97). Additional in vitro evidence supports the role of Tyr²¹² in DNA cleavage (104). A TraI HD phosphohydrolase domain, which contains highly conserved histidine and aspartate residues (105), is also essential for Neisseria T4SS-dependent DNA secretion. Amino acid substitutions of the signature histidine and aspartate within this domain significantly reduce DNA secretion, suggesting that charged residues are important for TraI function; however, the exact role of the phosphohydrolase remains unknown (97). Additionally, N. gonorrhoeae TraI harbors a distinct N-terminal domain that mediates membrane interactions required for DNA secretion, representing a unique step in contact-independent DNA transport through T4SS machinery (97).

ParA and ParB, partitioning proteins encoded on the GGI, are also essential for DNA secretion by the N. gonorrhoeae T4SS (21, 100). Partitioning proteins play a role in localizing chromosomal and plasmid DNA during cell division (106); however, the role of DNA segregation machinery in T4SS effector translocation is incompletely defined. In the prototypical A tumefaciens vir T4SS, ParA (VirC1) and ParB (VirC2) homologs spatially coordinate the T-DNA/relaxasome complex to the T4SS machinery localized to the cellular poles to initiate DNA secretion (72). Gonococcal ParA and ParB have been shown to interact with the TraI relaxase, suggesting that gonococcal ParAB may assume a similar role in facilitating N. gonorrhoeae DNA secretion via relaxasome cellular positioning and docking to the T4SS apparatus (107). Further spatial organization of the secretion machinery is influenced by GGI-encoded peptioglycanases that generate a localized break within the cell wall to implement T4SS apparatus biogenesis (100, 108). The GGI also encodes additional peptidoglycan-associated proteins, including EppA (a protein that severs peptidoglycan cross-links) and Yag, a OmpA-like protein that binds peptidoglycan and may facilitate machinery assembly by tethering the T4SS apparatus to the cell wall (20, 21, 100).

Gonococcal TraG, a homolog of the E. coli F plasmid mating pair formation protein, is also essential for ssDNA secretion through the GGI-encoded T4SS (99, 109). The observation that gonococcal DNA secretion occurs in a contact-independent manner suggests that in contrast to the canonical role in conjugative DNA transfer systems, gonococcal TraG evolved novel features to enable DNA export through the N. gonorrhoeae T4SS. TraG is localized to the Neisseria inner membrane via five transmembrane domains which orient the N terminus to the periplasm and the C terminus to the cytosol (109). Inner membrane localization and prominent periplasmic and cytosolic domains indicate that TraG may interact with other T4SS structural components or the relaxasome complex. While mapping experiments to determine TraG topology within the T4SS apparatus represent a snapshot of machinery architecture, the possibility that the TraG C-terminal domain is cleaved and delivered to the extracellular milieu remains an intriguing possibility (109). Additionally, strains harboring truncated TraG variants are unable to secrete DNA substrates, suggesting that TraG may orchestrate effector specificity (20, 109). Alternatively, TraG may provide architectural scaffolding during T4SS apparatus biogenesis. In support of this hypothesis, TraG is produced from a strain-variable operon that encodes additional T4SS components required for DNA secretion, including TraH and the lytic transglycosylase AtlA (20, 92, 99, 109). TraH localization is dependent on the presence of TraG, suggesting that TraG may stabilize TraH or facilitate TraH transport to the outer membrane (101). In F plasmid-encoded T4SSs, TraH homologs are involved in conjugative pilus assembly; however, because Neisseria ssDNA secretion does not require a T4SS-associated pilus, gonococcal TraH appears to assume a unique role within the GGI-encoded T4SS (100, 101). While biochemical analyses have significantly advanced our understanding of contact-independent ssDNA transport, the architecture of GGI machinery remains unresolved.

DNA UPTAKE BY COMPETENCE SYSTEMS

Natural transformation (NT) is a widespread mode of horizontal gene transfer that contributes to the evolutionary adaptation of diverse bacteria. Approximately 80 bacterial species, including both Gram-positive and Gram-negative organisms, have been reported to be naturally transformable (110). To achieve NT, bacteria assemble membrane-spanning machinery complexes and develop a physiological state known as competence that allows the uptake of exogenous DNA and subsequent integration into the recipient chromosome. During DNA uptake, double-stranded DNA (dsDNA) is captured from the extracellular environment and accumulates at the outer membrane (Gram-positive bacteria) or the periplasm (Gram-negative bacteria) (111). In a process that is incompletely defined, exogenous dsDNA is converted into a single-stranded form that is channeled via the inner membrane protein ComEC into the cell cytoplasm (112, 113). Single-stranded DNA (ssDNA) is integrated into the recipient chromosome by cytoplasmic recombination machinery that incorporates DprA (114), RecA (115), and other competence-related proteins such as ComFC (116). While the ComEC and homologous recombination machinery is conserved in most naturally transformable bacteria (110, 117), the composition of DNA uptake machinery varies widely. Many bacteria employ a retractable type IV competence pilus (T4P) for dsDNA capture from the extracellular environment (118). In these cases, exogenous DNA is pulled to the outer membrane or periplasm by T4P retraction and subsequent binding to the competence proteins ComEA (111, 119, 120). Additionally, a recent study demonstrates that cell wall-associated teichoic acid mediates and enhances DNA binding during NT (121).

In contrast to most naturally transformable bacteria, H. pylori employs the comB T4SS for DNA uptake into the periplasm (122, 123). The comB apparatus exhibits homology to corresponding components within the prototypical A. tumefaciens vir T4SS and is encoded by genes clustered into two distinct operons (comB2-comB4 including 3 genes, and comB6-comB10 including 5 genes) (122, 123). Structurally, ComB7, ComB9, and ComB10 are proposed to form the outer membrane complex (14, 33) while ComB6-ComB8-ComB10 subassemblies are postulated to span the periplasm to bridge the inner membrane translocation apparatus and the outer membrane-associated complex (122, 123). Comparison to the A. tumefaciens vir T4SS reveals that while comB harbors a VirB4 ATPase homolog, the apparatus lacks multiple components essential for vir T4SS function, including VirB1, the minor pilin VirB5, the DNA coupling protein VirD4, and the ATPase VirB11 (14). Similar to A. tumefaciens VirB2, the major pilin ortholog ComB2 is proposed to be involved in cell surface DNA binding (6, 14).

Although the mechanistic details underscoring comB T4SS-mediated DNA import are unresolved, a two-step DNA uptake process has been described for H. pylori (124, 125). In the first step, the comB T4SS mediates exogenous DNA transport across the outer membrane followed by secondary DNA translocation into the bacterial cytoplasm via ComEC (125). ComH, a periplasmic protein that binds dsDNA, provides an essential link between outer and inner membrane transport (126). During DNA import across the cell envelope, the ComH C terminus interacts with incoming DNA and bridges substrate exchange with ComEC via the N-terminal domain to facilitate delivery across the inner membrane (126), suggesting that ComH is important for periplasmic DNA loading into the translocation channel. It is thus possible that ComB2 shuttles incoming DNA through a “relay” mechanism with ComH prior to substrate loading into the inner membrane translocation channel for import into the recipient cytoplasm (14, 126).

DNA uptake is a highly regulated process that does not lead to bacterial growth arrest (127, 128). For example, levels of ComB8 and ComB10 correlate with transformation rates and overexpression of the comB6-comB10 operon leads to increased competence (128). The kinetics of high velocity comB T4SS-mediated DNA import exceed corresponding rates and total DNA import volumes observed for analogous type II secretion system/T4P uptake devices (14, 125, 129). For example, while Neisseria T4P retraction systems have the capacity to import around 40 kbp (119), the comB T4SS translocates up to one chromosome equivalent into the periplasm (127). In contrast to Neisseria T4P systems that transport both ssDNA and dsDNA substrates into the bacterial cell (129), it is unknown whether the comB apparatus supports exogenous ssDNA translocation into the periplasm (14). Finally, in addition to significantly enhancing genetic diversity, competence may represent a defensive strategy in which increased reservoirs of periplasmic DNA protect the chromosome by mitigating oxidative stress encountered in the gastric niche (127). Alternatively, NT may protect the core genome and serve as a mechanism to combat undesirable outcomes associated with hypervariability or genome instability in microbes that undergo frequent intrachromosomal recombination (130). While competence is not required to establish infection, comB T4SS-mediated transformation may provide competitive advantages that promote chronic gastric colonization or H. pylori dissemination (14, 131, 132). Thus, future studies to elucidate comB apparatus architecture and the role of natural competence in bacterial persistence are warranted.

TRANSKINGDOM DNA TRANSFER VIA CONJUGATIVE MECHANISMS

Interdomain conjugation is a major driving force underlying genetic exchange among varied microbial species. In addition to horizontal gene transfer from donor to recipient bacteria, conjugative mechanisms can facilitate genetic element dissemination from bacteria to archaea or to diverse eukaryotic cells. For example, the broad host range proteobacterial IncP1 and associated derivative transfer systems have the capacity to translocate plasmid-derived DNA to bacteria (133, 134), archaea, yeast, and higher order eukaryotic cells (134–139). Accordingly, multiple studies demonstrate that E. coli harboring IncP1 systems deliver plasmid-borne DNA to the methanogenic archaeon Methanococcus maripaludis (135), and evolutionarily divergent eukaryotes, including yeast (e.g., Saccharomyces cerevisiae [136, 140, 141], S. kluyveri [142], Kluyveromyces lactis [143], and Pichia angusta [143]), unicellular algae diatoms (137), plant cells (138), and cultured human cells (139). Although there is a paucity of evidence supporting the role of IncP-mediated genetic transformation in nature, these studies suggest that transkingdom conjugation is not limited to the prototypical Agrobacterium tumefaciens system.

Recent evidence supports the hypothesis that interkingdom DNA transfer is a vestigial consequence of ancestral mechanisms underlying generalized conjugation through T4SS machinery (3, 9). Thus far, “dual function” systems that translocate diverse effector proteins and achieve DNA conjugation have been discovered in very few organisms. The most extensively characterized dual function machinery is the A. tumefaciens vir T4SS that translocates tumor-inducing DNA (T-DNA) and associated effector proteins into recipient plant cells (1–4, 6, 9). Multiple reports indicate that additional eukaryotic cell types are susceptible to transformation via A. tumefaciens transkingdom conjugation, including filamentous fungi (144, 145), arachnid cells (146), and cultured human cell lines (147). In both the A. tumefaciens vir T4SS and prototypical interbacterial conjugation systems, DNA is transmitted cell-to-cell as a nucleoprotein complex in which the 5′ end of the ssDNA substrate is covalently complexed to a cognate relaxase protein that is secreted into target cells (4, 5). Thus, the relaxase serves to directionally pilot tethered DNA through the translocation channel (1, 4, 5).

When in contact with target plant cells, A. tumefaciens T-DNA is excised from the cognate Ti plasmid at defined T-border sequences by the VirD1 DNA topoisomerase, VirD2 endonuclease, and other accessory proteins that form the catalytically active relaxosome (4, 6, 9). Within the relaxosome, target DNA is cleaved by the VirD2 relaxase which remains covalently tethered to the 5′ end of the excised ssDNA via a phosphotyrosine linkage (4, 9). The relaxase and associated accessory proteins promote relaxosome recruitment to the VirD4 coupling protein for docking to the vir T4SS apparatus (4, 6, 8, 9, 148, 149). Although VirD4 serves as the relaxosome receptor, substrate loading into the secretion channel is likely coordinated by two additional ATPases, VirB11 and VirB4 (6, 8, 35, 150). Upon relaxosome docking to VirD4, the relaxase is unfolded via unknown mechanisms (73), resulting in the release of accessory proteins and nucleoprotein translocation through the vir T4SS apparatus (4, 6, 9). Thus, ssDNA cargo is piloted into recipient plant cells as a “hitchhiker” molecule that is coincidentally delivered with relaxase effectors. In the recipient plant cell, VirD2 catalyzes T-DNA recircularization and facilitates nucleoprotein complex trafficking into the nucleus for chromosomal integration (1, 4, 9). In support of a generalized protein secretion model, the vir T4SS independently transports additional proteins into plant cells, including VirE2, VirE3, VirD5, and VirF (2, 4). The VirE2 ssDNA-binding protein coats translocated DNA cargo in the recipient cell cytoplasm and complexes with VirD2 as a mechanism by which to protect effector DNA cargo from cellular nucleases (151). VirE3, VirD5, VirF, and other vir T4SS effectors injected into plant cells promote infection and are implicated in crown gall tumorigenesis in planta (4, 151, 152).

In addition to A. tumefaciens, recent studies demonstrate that other plant and animal pathogens, including Rhizobium etli (153), Bartonella henselae (61, 154–156), Coxiella burnetii (155) Legionella pneumophila (155, 157, 158), and Helicobacter pylori (159) have the intrinsic capacity to translocate DNA into eukaryotic cells via “dual purpose” T4SS-dependent mechanisms. For example, L. pneumophila and C. burnetii harboring artificial dot/icm T4SS chimeras integrating the MobA relaxase associated with the mobilizable plasmid RSF1010 gain the capacity to translocate ssDNA into mammalian cells (155). In support of the hypothesis that conjugation is an inherent and ancient function of dual purpose machineries, early studies demonstrate that the L. pneumophila dot/icm T4SS mobilizes RSF1010 plasmids into recipient Legionella, as well as some E. coli strains, via MobA relaxase-dependent conjugative mechanisms (155, 158). Similarly, the B. henselae vir T4SS can coopt the heterologous TrwC relaxase associated with the R388 self-transmissible plasmid to achieve transkingdom ssDNA delivery into endothelial cells (154, 160). Chimeric relaxase fusions harboring a C-terminal translocation signal derived from the cognate effector protein BepD significantly enhanced B. henselae ssDNA transkingdom conjugation efficiencies (154), highlighting the potential to engineer T4SS machinery that delivers therapeutic designer DNA into target human cells in vivo. Importantly, TrwC has been shown to facilitate the random integration of plasmid-derived effector DNA into the recipient genome (156) presumably via VirD2-like nuclear import mechanisms. Alternatively, translocated B. henselae DNA integration and transgene expression may occur through a nuclear import-independent pathway that relies on breakdown of the nuclear envelope during cellular division (154, 156, 160).

In contrast to A. tumefaciens-like conjugation systems that deliver plasmid-borne ssDNA into mammalian cells, indirect evidence indicates that the H. pylori cag T4SS delivers chromosomal fragments into gastric epithelial cells in a contact-dependent mechanism (159) (Fig. 2). H. pylori cag T4SS-dependent DNA translocation is partially susceptible to exogenous nucleases and neutralizing anti-DNA monoclonal antibodies (159), suggesting a two-step translocation mechanism that does not appear to require an extracellular T4SS-associated pilus conduit. Translocated H. pylori DNA activates gastric TLR9 and additional immunostimulatory pathways that induce both anti- and proinflammatory responses (159, 161). Thus, cag T4SS-dependent DNA translocation provides additional mechanisms by which H. pylori manipulates the immune response to generate a hospitable environment that fosters chronic colonization of the gastric niche (159, 161). In support of the observation that cag T4SS activity stimulates multiple nucleic acid reconnaissance systems (159, 162), recent work demonstrates that H. pylori has evolved mechanisms to counterbalance or suppress nucleic acid signaling within the gastric mucosa (161, 163). For example, H. pylori-induced NF-κB activation induces the upregulation of TRIM30a to antagonize innate responses elicited by TLR9 and other DNA sensors in the murine stomach (163) (Fig. 2). While these studies provide compelling evidence that H. pylori provokes nucleic acid surveillance signaling via cag T4SS activity (Fig. 2), the exact mechanism by which chromosomal DNA is excised and transported through the secretion apparatus is unresolved.

FIG 2.

H. pylori transkingdom conjugation elicits diverse immune outcomes in gastric epithelial cells. Chromosomal fragments delivered to gastric epithelial cells via cag T4SS activity activate TLR9 (159, 161) and potentially other nucleic acid reconnaissance systems. To control excessive inflammatory responses stimulated by TLRs and STING, H. pylori induces negative feedback regulators (e.g., Trim30a in the murine gastric mucosa [163]) that suppress type I interferon production and dampen TLR9-driven NF-κB responses (161, 163).

Remarkable progress enabled by single particle cryo-EM and cryo-ET has illuminated the architectures of dual purpose T4SS machineries. Compared to minimized DNA conjugation systems, expanded H. pylori and L. pneumophila effector translocator systems exhibit symmetry mismatch between the outer and inner layers of the OMCC (17, 30, 39, 41, 81, 164). For example, the H. pylori cag T4SS OMCC is comprised of distinct structural features, including an outer membrane cap (OMC) consisting of an outer layer (O-layer) and inner layer (I-layer), a periplasmic ring (PR) complex, and a stalk region (25, 36, 39, 41) that assembles with a striking symmetry mismatch occurring between the OMC (14-fold symmetry) and the PR (17-fold symmetry) (39, 41) (Fig. 1). Similarly, the L. pneumophila dot/icm T4SS OMC (13-fold symmetry) and PR (18-fold symmetry) displays remarkable mismatch (5, 164). Complex and subassembly asymmetry is noteworthy considering that minimized conjugation systems effectively translocate DNA or proteins to other bacteria and eukaryotic cell targets despite displaying symmetry mismatch between the OMCC and IMC (5). In situ cryo-ET of the cag T4SS and dot/icm T4SS revealed several densities not identified by complex purification and single particle cryo-EM approaches, including central periplasmic cylinders and wing-like collars anchored to large inner membrane complexes (25–27, 36). The biological significance of symmetry mismatch between large T4SS subassemblies is unknown; however, asymmetry may afford dynamic structural transitions associated with channel gating or effector secretion that are induced in response to intra- or extracellular stimuli. Within the expanded dual purpose systems, nanomachine conformational flexibility may be required for binding to specific eukaryotic receptors, the recruitment of nonprotein substrates to the secretion channel, or orchestrating effector protein delivery across the bacterial envelope (5). Continued structural and mechanistic studies will deepen our understanding of how the fascinating T4SS nanomachine evolved extreme functional diversity and expanded recipient cell range.

CONCLUDING REMARKS AND FUTURE PROSPECTS

Remarkable advances in T4SS structural definition have provided exciting insight into nucleoprotein docking reactions and elegant mechanisms regulating substrate passage through the secretion channel. Recent work illuminating the incredible structural and architectural diversity among paradigmatic T4SS machineries has raised intriguing questions surrounding the observed symmetry mismatch in both minimized and expanded nanomachines as well as the functional significance of associated pilus structures. Despite the incredible progress toward the structural resolution of phylogenetically diverse systems, many fundamental questions remain unanswered. For example, the molecular mechanisms underscoring effector translocation through the T4SS apparatus have not been elucidated, and the intra- and extracellular signaling requirements for secretion channel activation/gating have not been identified. Furthermore, it is unknown whether the unfolded relaxase or the tethered ssDNA is transported first through the secretion channel, or whether these substrates are secreted simultaneously.

Although current models posit that the relaxase is piloted through the secretion apparatus to pull tethered DNA into the recipient cell, the mechanisms underlying relaxase processing and unfolding have not been described. Thus far, a putative chaperone that maintains the primed relaxase in an unfolded or semiunfolded state has not been characterized and whether a specific “unfoldase” facilitates relaxase processing is unknown. Whether the T4SS apparatus transitions from or oscillates between multiple architecture states that trigger either effector protein or ssDNA substrate secretion is an intriguing possibility supported by cryo-EM snapshots of various conjugation architectures assembled in the cell envelope (11, 51, 80). Finally, while the details of DNA processing and conjugative transport have been intensely studied, the architecture of donor-recipient mating junctions has not been resolved and the mechanistic details governing ssDNA transfer across the outer and inner recipient membranes have not been established. Continued structure-function studies to address these outstanding questions will undoubtedly foster innovative translational investigations aimed at developing synthetic chemical T4SS inhibitors (12, 165–167) or engineering optimized T4SS machineries to serve as designer DNA delivery devices (139, 154, 160). Unraveling the inner workings of these sophisticated nanomachines will pave the way for biotechnological breakthroughs and broaden our understanding of T4SS evolutionary adaptations that enable bacterial survival in challenging host environments.

Biographies

Mackenzie E. Ryan is a Ph.D. candidate in the Department of Microbiology, Immunology and Molecular Genetics at the University of Kentucky. She received her B.S. in Biology from the University of Dayton where she graduated summa cum laude in 2018. Her dissertation research in the Shaffer laboratory focuses on understanding how architectural symmetry mismatch orchestrates substrate selection and facilitates effector molecule translocation by the Helicobacter pylori cag type IV secretion system.

Prashant P. Damke, Ph.D. is a postdoctoral scholar in the Shaffer lab at the University of Kentucky. He obtained his doctorate in biochemistry from the Indian Institute of Science in Bangalore, India, and subsequently completed a postdoctoral fellowship under the direction of Dr. J. Pablo Radicella at the Institute of Cellular and Molecular Radiobiology in Fontenay-aux-Roses, France. His research interests focus on understanding the molecular mechanisms of DNA processing and transport during bacterial horizontal gene transfer. He is also interested in unraveling the biochemical and cellular functions of proteins and enzymes that play critical roles in DNA transport across membranes, DNA repair, homologous recombination, and host-pathogen interactions.

Carrie L. Shaffer, Ph.D. is an Assistant Professor at the University of Kentucky in the Department of Veterinary Science; the Department of Microbiology, Immunology, and Molecular Genetics; the Department of Pharmaceutical Sciences; and the Markey Cancer Center. She completed a B.Sc. in Agricultural Biotechnology at the University of Kentucky and received her Ph.D. in Microbiology and Immunology from Vanderbilt University under the mentorship of Dr. Tim Cover. Following postdoctoral fellowships at Vanderbilt University Medical Center (Dr. Maria Hadjifrangiskou) and Caltech (Dr. Grant Jensen), she established her independent research program focused on understanding how Helicobacter pylori cag T4SS activity stimulates the development of infection-associated malignancies.