Caught in the Act: The Dialog Between Bacteriophage R17 and the Type IV Secretion Machine of Plasmid R1

Abstract

Bacteria communicate with each other through contact-independent and-dependent signaling mechanisms. Sensory perception of both types of signals is needed for conjugative transfer of mobile DNA elements via type IV secretion systems (T4SSs) to bacterial or eukaryotic target cells. While the regulatory circuitries coupling extracellular quorum and environmental signals to transcription of T4SS genes are increasingly understood, it remains fundamentally unknown how a potential recipient cell stimulates donor conjugative DNA transfer upon contact. In this issue, Zechner and colleagues report use of the male-specific bacteriophage R17, a phage that binds conjugative pili elaborated by IncF plasmid R1, to define requirements for phage-contact-mediated T4SS activation and phage penetration. They report that R17 penetrates only through T4SS channels engaged for delivery of their plasmid cargo to recipient cells. Engagement requires docking of catalytically-active relaxase TraI bound at oriT with the TraD substrate receptor (also termed the T4CP). The data, together with recent ultrastructural and biochemical findings, support an intriguing new model that the T4CP cumulatively senses an intracellular signal (substrate docking) and an extracellular signal (pilus bound by phage or a recipient cell) to coordinate a late stage morphogenetic or gating reaction that enables bidirectional transmission of nucleoprotein substrates through the T4SS.

Signal exchange among bacteria activates a wide array of physiological responses, including transmission of mobile genetic elements (Frost and Koraimann, 2010, Wozniak and Waldor, 2010). Conjugative transfer of mobile elements such as plasmids and integrative and conjugative elements (ICEs) is often stimulated by donor cell perception of signals in the environment. Some of these extracellular signals are quorum molecules released by potential recipient cells in the vicinity; for example, a homoserine lactone induces transfer of the Agrobacterium tumefaciens Ti plasmid (White and Winans, 2007) and a small peptide pheromone activates Enterococcus faecalis pCF10 (Dunny and Johnson, 2011). Other conjugation systems, exemplified by the Escherichia coli F transfer system, respond to other environmental (oxygen, temperature) and physiological signals (nutrient starvation, extracytoplasmic stress) signals (Frost and Koraimann, 2010). Studies of these and other well-characterized conjugation systems have defined features of complex regulatory circuitries coupling extracellular signals to activation of transfer gene expression, processing of the DNA substrate for transfer, and assembly of the translocation apparatus.

Conjugation systems are also activated in response to signals conveyed from recipient cells upon establishment of the donor - target cell contact (Lu and Frost, 2005). The nature of this ‘mating signal’, postulated more than 50 years through studies of the E. coli F transfer system, has eluded identification despite the accumulation of biochemical data in the interim of its existence (Brinton, 1965, Ou, 1975, Ou and Reim, 1978). The signal(s) eludes us still, but the elegant studies by Lang et al. reported in this issue of Molecular Microbiology advance our understanding of this phenomenon an important step further.

The conjugation systems are classified as members of the type IV secretion systems (T4SSs) (Lawley et al., 2003, Alvarez-Martinez and Christie, 2009). The T4SSs are a functionally diverse group of translocators that mediate the transfer of DNA and protein substrates across the cell envelopes of many species of Gram-negative and -positive bacteria and some species of the archaea. In Gram-negative bacteria, these systems elaborate two surface organelles, the envelope-spanning translocation channel and an extracellular filament termed the conjugative pilus (Christie et al., 2005). A dozen or more proteins called the mating pair formation (Mpf) proteins are needed to build these structures (Lessl et al., 1993). In addition, the type IV coupling protein (T4CP) associates at the cytoplasmic entrance of the translocation channel, functioning as the substrate receptor and linking secretion substrates to delivery through the channel (Gomis-Ruth et al., 2001, Schroder et al., 2002). In this capacity, the T4CP is required for substrate translocation, but it is dispensable for biogenesis of the conjugative pilus.

Some bacteriophages are ‘male-specific’ because they use conjugative pili as receptors for penetration into the cell (Roberts and Steitz, 1967). They attach to the tip or side of the pilus and gain access to the cell by mechanisms that are not defined but require the pilus and the rest of the Mpf proteins (Frost et al., 1985). Some ‘male-specific’ phages require only the pilus and Mpf channel subunits, but the group I RNA phages R17, f2, and MS2 additionally require the T4CP (Schoulaker and Engelberg-Kulka, 1978). Essentially, these phages have evolved to adsorb and utilize intact T4SSs - systems competent for substrate translocation - to gain access to the cell interior. Lang et al. postulated that group I phages might in effect mimic recipient cell contacts by conveying a signal equivalent to a recipient cell ‘mating signal’ to activate T4SS channel opening. Further, because these phages exhibit the unusual requirement of the T4CP for penetration, the signal could be conveyed to the T4CP, which then coordinates a response culminating in channel activation.

T4CPs are fascinating components of the T4SSs (Alvarez-Martinez and Christie, 2009). They are associated with all described conjugation systems and most other T4SSs dedicated to intercellular trafficking of proteins or other substrates. They typically possess an N-terminal transmembrane domain (TMD) and a soluble nucleotide-binding domain (NBD). The NBD of the TrwB T4CP from plasmid R388 was crystallized and shown to adopt a ring-shaped structure similar to RecA and DNA ring helicases. Six protomers of TrwB form the homohexameric ring with outer dimensions of 110 Å in diameter and 90 Å in height and a central channel of 20 Å in diameter constricting to 8 Å at the cytoplasmic pole (Gomis-Ruth, et al., 2001, Gomis-Ruth, et al., 2002). T4CPs bear sequence and structural similarities with the SpoIIIE and FtsK DNA translocases, and one model posits that these subunits function as motor proteins to drive translocation of secretion substrates across the inner membrane and into the Mpf channel for conveyance to the cell surface (Errington et al., 2001, Gomis-Ruth et al., 2004, Massey et al., 2006).

For the IncFII plasmid R1-16 under investigation by Lang et al., the TraD T4CP initiates the plasmid transfer process through establishment of contacts with the relaxosome composed of TraI relaxase/helicase and other auxiliary factors, e.g., TraM, bound at the R1-16 origin of transfer (oriT). The T4CP-relaxosome interaction stimulates TraI catalytic cleavage at the nic site within oriT of the DNA strand destined for translocation. To explore the requirements for R17 phage uptake through the R1-16 pilus/transfer channel, Lang et al. confirmed the importance of TraD and also established that mutation of the nucleotide-binding site abolished phage uptake. Armed with earlier biochemical evidence that the T4CP stimulates TraI relaxase/helicase activities in vitro (Mihajlovic et al., 2009, Sut et al., 2009, Lang et al., 2010), the investigators next tested and found, surprisingly, that the TraI relaxase/helicase and an intact oriT sequence are required for phage penetration. Mutations of genes for the auxiliary factors TraM or TraY also abolished phage adsorption; however, these mutations could not be complemented by trans expression of traM or traY, preventing conclusions about the importance of these proteins for phage uptake.

These initial results suggested the intriguing possibility that a TraI-oriT complex docked at the TraD T4CP is required for channel activation by phage R1. Through analyses of various TraI truncation derivatives, Lang et al. showed that an N-terminal fragment of TraI (residues 1-992) carrying the relaxase domain (residues 1-332) joined to an internal translocation signal (TSA) sufficed for phage uptake. The helicase domain and a C-terminal TraM-interaction domain were both dispensable, despite their importance for conjugative transfer of plasmid R1-16 to bacterial recipients. The investigators went on to confirm that TraD stimulates the nicking activity of the TraI(1-992) fragment in vitro, and conversely, TraI mutations in the relaxase catalytic pocket (Y16F/Y17F) and of residues predicted to disrupt ssDNA binding (E153D/Q193R/R201Q) abolished R17 phage uptake. As is typical of this group’s work, they carried out robust biochemical follow-up studies, confirming that the latter mutations exerted their effects specifically through attenuation of relaxase - nic site binding affinity as opposed to nonspecific disruption of global protein architecture.

Taken together, these studies defined the requirements for R17 phage uptake through the plasmid R1-16 T4SS: TraD bearing an intact nucleotide-binding site, a catalytically-active TraI relaxase domain, an adjoining TSA domain implicated in establishment of the TraI-TraD interaction, and the R1-16 oriT sequence. Lang et al. thus propose that the TraD T4CP essentially functions as a ‘switchboard operator’ by coupling perceived intra-and extracellular signals to type IV channel activation. The intracellular signal corresponds to relaxosome docking with the T4CP. One response output of signal perception is T4CP-mediated stimulation of relaxase activity, resulting in generation of the ssDNA transfer intermediate. The extracellular signals include phage adsorption to the pilus or a pilus - recipient cell contact. These signals are propagated through the pilus and translocation channel to the T4CP. Here, the response output is a stimulation of T4CP ATPase activity and channel activation. The T4CP therefore cumulatively perceives an intracellular signal, processed ssDNA substrate, and an extracellular signal conveyed from phage or recipient cell, to mount a response enabling passage of nucleoprotein complexes through the T4SS (Fig. 1).

Fig. 1.

Model for contact-dependent stimulation of the T4SS. Left: Signals converge on the T4CP from i) phage binding or contact with a recipient cell (yellow dashed arrow) and ii) a docked relaxosome (red dashed arrow). Right: The T4CP in turn stimulates channel gating through ATP hydrolysis and energy-induced conformational changes in the T4SS. The gated channel serves as a conduit for nucleoprotein substrates: i) phage uptake, e.g., R17 protein A bound to RNA (yellow); or ii) plasmid transfer, e.g., R1 TraI relaxase bound to ssDNA (red). The DNA strand of plasmid R1 translocated through the T4SS is depicted in red in the left panel.

Two questions surface from this work: i) What is the nature of the still-elusive phage or recipient cell signal transmitted across the T4SS to the T4CP, and ii) what is the underlying mechanism of T4CP-mediated channel activation? Answers to these questions await further study, although recent biochemical and ultrastructural studies of related T4SSs suggest that the T4CP regulates channel activity through structural changes that modulate channel gating. One of most significant advances in our understanding of the type IV channel architecture in recent years was achieved through structural resolution of a multisubunit ‘core’ complex of the pKM101 conjugation system. The complex, composed of 14 copies of each of the TraN lipoprotein, TraO, and TraF subunits, is a 1.05 MDa barrel-shaped structure, 185 Å wide and 185 Å long, with the potential of spanning the entire Gram-negative bacterial cell envelope. The N termini of the 14 TraF monomers interact to form a 55 Å diameter hole in the inner membrane. The rest of TraF extends through the periplasm to the outer membrane where, intriguingly, C-terminal α-helical domains of the 14 monomers interact to form a pore. The core complex has a central chamber of approximately 110 Å in diameter, sufficiently large to house other components of the translocation channel and/or a pilus or pseudopilus structure (Chandran et al., 2009, Fronzes et al., 2009).

Comparisons of core complex structures solved by cryoelectron microscopy (CryoEM) and X-ray crystallography identified structural variations consistent with two states, one in which the channel is open and the other closed (Fronzes et al., 2009, Chandran et al., 2009). These observations are of interest in view of work on the related Agrobacterium tumefaciens VirB/VirD4 system showing that the TraF homolog, VirB10, undergoes a conformational switch upon sensing of ATP binding/hydrolysis activities of the VirD4 T4CP and VirB11 ATPase (Cascales and Christie, 2004a). By use of a ChIP-based crosslinking assay, evidence was presented that this VirB10 structural transition is necessary for passage of DNA substrates through the distal portion of the translocation channel (Cascales and Christie, 2004b). Furthermore, a role for VirB10 in energy-dependent gating of the outer membrane pore was suggested by isolation of a mutation near the outer membrane pore that ‘locks’ VirB10 in the energized conformation and allows for release of secretion substrates to the cell surface independently of target cell contact (Banta, et al., 2011).

The T4CP thus might perceive extracellular signals (phage or recipient cell binding to the pilus) through structural transitions propagated along the core complex or pilus. Receipt of this signal together with that accompanying engagement of a catalytically active relaxosomal complex would stimulate T4CP ATP hydrolysis and further structural changes necessary for channel activation (Fig. 1). It will be exciting to test some of the predictions of this model, namely, that i) binding of male-specific phage or recipient cells induce structural transitions in the core complex and/or pilus, ii) phage or recipient-cell-contact mediated signals stimulate T4CP ATPase activity, and iii) a combination of substrate docking and phage or recipient cell binding induce channel gating. Finally, as noted above, the T4CP hexamer sits at the entrance of the Mpf channel and might function as the translocase for delivery of DNA substrates as well as phage nucleic acids across the inner membrane. The role of the T4CP in substrate translocation has been a subject of intense debate, however, and an alternative model proposes that the T4CP functions only to recruit and process substrates. The T4CP then delivers the transfer intermediate to a translocase composed of Mpf subunits for passage across the inner membrane and the rest of the T4SS channel (Cascales and Christie, 2004b, Atmakuri et al., 2004). By capitalizing on the observation that group I phages require T4CPs for penetration and by assaying for phage RNA contacts with native and mutant forms of T4CPs, it should now be feasible to discriminate between these models.

Finally, why have phages evolved to exploit T4SSs that are engaged for translocation? Lang et al. suggest that this stringency benefits the conjugative plasmid, as this window of opportunity is relatively compressed in the life of a plasmid. Another possibility derives from the recent ultrastructural and biochemical evidence that T4SS core complexes and VirB10 subunits undergo ATP-mediated structural changes consistent with a late stage morphogenetic or gating reaction. Some phages might have evolved the capacity to recognize these nearly completely assembled T4SSs and then, through a contact-dependent signal, activate final steps in channel morphogenesis or gating to generate an open conduit across the cell envelope for penetration. This strategy benefits the phage and likely suppresses concomitant transfer of the conjugative plasmid docked at the hijacked channel. Indeed, suppression by phage of conjugative transfer is observed in nature (Novotny et al., 1968, Ou, 1973) and is currently considered a viable therapy for inhibiting dissemination of plasmid-encoded antibiotic resistance and other virulence traits in nature (Lin et al., 2011, Jalasvuori et al., 2011). Male-specific phages bound to pili might simply sterically impede formation of mating junctions, but the present findings suggest phage hijacking of activated conjugation channels is another possible mechanism of suppression. As Lang et al. note, the dialog between a conjugative plasmid and infecting phage is a war won through mutation. The accumulation of mutations in pilin subunits and T4CPs offer a selective advantage to conjugative plasmids in suppressing recognition by male-specific phages, whereas these phages continually evolve to recognize functional T4SSs. The outcome of this dialog is diversification of T4SSs, as is clearly a prominent feature of this translocation superfamily (Alvarez-Martinez and Christie, 2009), as well as the invading phage. Finally, it is intriguing to speculate that some phages have acquired the capacity to target host cells carrying functional conjugative plasmids through recognition of activated mating channels. Phage penetration followed by integration into the resident conjugative plasmid and plasmid transfer offers an alternative means for phage dissemination among host populations, and might have contributed over evolutionary time to the extreme mosaicism evident among phages, plasmids, and ICEs.