Prime time for minor subunits of the type II secretion and type IV pilus systems

Authors

  • Lori L. Burrows

    Corresponding author
    • Department of Biochemistry and Biomedical Sciences, the Michael G. DeGroote Institute for Infectious Diseases Research, McMaster University, Hamilton, Ontario, Canada
    Search for more papers by this author

For correspondence. E-mail burrowl@mcmaster.ca; Tel. (+1) 905 525 9140 ×22029; Fax (+1) 905 522 9033.

Summary

The type II secretion system (T2SS) exports folded proteins from the periplasms of Gram-negative bacteria. The type IV pilus system (T4PS) is a multifunctional machine used for adherence, motility and DNA transfer in bacteria and archaea. Partial sequence identity between the two systems suggests that they are related and might function via a similar mechanism, the dynamic assembly and disassembly of pseudopilus (T2SS) or pilus (T4PS) filaments. The major subunit in each system is thought to form the bulk of the (pseudo)pilus, while minor (low-abundance) subunits have proposed roles in assembly initiation, antagonism of disassembly, or modulation of (pseudo)pilus functional properties. In this issue, Cisneros et al. (2012) extend their previous finding that pseudopilus assembly is primed by the minor pseudopilins, showing that the same proteins can initiate assembly of Escherichia coli T4P. Similarly, they show that the E. coli minor pilins prime the polymerization of T2S pseudopili, although unlike genuine pseudopili, the chimeric filaments did not support secretion. This work reinforces the notion of a common assembly mechanism for the T2S and T4P systems.

The type IV pilus (T4P) system is broadly distributed among prokaryotes (Craig and Li, 2008; Pohlschroder et al., 2011) and is related to the type II secretion (T2S) system that secretes folded proteins from the periplasms of Gram-negative bacteria (Pugsley, 1993). Both systems are proposed to function via the dynamic assembly and disassembly of a pilus or pseudopilus respectively. For T4P, the repeated extension and retraction of adherent pili results in ‘twitching’ motility and contributes to pathogenicity, allowing bacteria to breach host barriers (Alarcon et al., 2009; Zahavi et al., 2011; Burrows, 2012; Coureuil et al., 2012). By analogy, the T2S system is proposed to form a short periplasmic pseudopilus that acts as a piston to push its substrates out of the cell (Korotkov et al., 2012). The relatedness of the two systems has been substantiated through structural studies that revealed common architectures even for components with limited sequence identity (Ayers et al., 2010; McLaughlin et al., 2012). Functional overlap is suggested by studies showing that the T2S system can assemble long surface-exposed ‘hyper-pseudopili’, and that some T4P systems secrete proteins (Sauvonnet et al., 2000; Kennan et al., 2001; Kirn et al., 2003; Vignon et al., 2003; Hager et al., 2006). Although assumed to be similar, the mechanistic details of pseudo (pilus) assembly remain poorly understood.

Each system has the same basic set of components: a predominant subunit called the major pilin (T4P) or major pseudopilin (T2S); a pre-pilin peptidase that removes the leader sequence from (pseudo)pilins to make them competent for assembly; a cytoplasmic membrane-embedded ‘platform’ protein; and a cytoplasmic polymerization ATPase, proposed to power assembly through ATP hydrolysis-driven conformational changes that are transmitted through the platform protein to the subunits (Strom et al., 1991; Py et al., 2001; Yamagata and Tainer, 2007; Craig and Li, 2008). Many T4P systems have a second, depolymerization ATPase. T2S and T4P systems in Gram-negative bacteria have large gated secretins that provide a conduit for substrate proteins or pili to transit the outer membrane (Wolfgang et al., 2000).

In addition to these basics, both systems have less conserved components, many of whose functions remain unclear. Among these are a set of ‘minor’ (pseudo)pilins, so designated not because of their insignificance, but rather their low abundance relative to the major subunit (Alm et al., 1996). Differences in the exact number, sequence and mutant phenotypes of minor subunits in T2S and T4P systems have hampered understanding of their specific roles. Proposed functions include initiation of assembly, contact-dependent opening of the secretin, antagonism of retraction, and/or modulation of fibre properties, including the selection of secretion substrates in the case of T2S (Winther-Larsen et al., 2005; Forest, 2008; Cisneros et al., 2011; Douzi et al., 2011; Korotkov et al., 2012).

In the T2S system, there are four conserved minor subunits designated H, I, J and K in the standardized Gsp (general secretory pathway) nomenclature (Pugsley, 1993). The four proteins form a complex in vitro, and GspIJK can be cross-linked in vivo (Douzi et al., 2009; Cisneros et al., 2011). The GspIJK equivalents from enterotoxigenic Escherichia coli were crystallized as a heterotrimer capped by GspK, whose arrowhead-like shape suggested that additional subunits were added below, with the complex forming the eventual tip of the pseudopilus (Forest, 2008; Korotkov and Hol, 2008). A related set of proteins are found in type IVa pili (T4aP) systems, one of two T4P subfamilies (Pelicic, 2008). Like GspHIJK, T4aP minor pilins are encoded together (Fig. 1) and are important for pilus assembly (Winther-Larsen et al., 2005; Giltner et al., 2010). However, their poor sequence identity with the T2S proteins makes it difficult to predict with any confidence whether their structures will be similar. Certain features of the conserved N-termini are shared however, including a conserved glutamate at +5 relative to the mature N-terminus in all but the GspK equivalents, and a conserved proline at +25 in GspH equivalents that might modulate interactions with the assembly machinery (Cisneros et al., 2011).

Figure 1.

Minor subunits prime assembly. The organization of minor subunit genes in model organisms Klebsiella oxytoca (T2S), Escherichia coli (T4aP), Pseudomonas aeruginosa (T4aP) and Neisseria spp. (T4aP) is shown at the top, with major subunit genes in white on the left, and minor subunit genes on the right. Equivalent conserved minor subunits (based on sequence identity and synteny) are shown in light pink, blue, tan and green. The gene order is not conserved in E. coli. Species-specific minor pilins are shown in orange (P. aeruginosa pilE), yellow (Neisseria pilX) and dark pink (Neisseria pilV). At the bottom are three models showing how the minor subunits could promote assembly. In model 1, spontaneous formation of a minor subunit complex (colour-coded to the genes above) could introduce a local membrane deformation (red arrow) that stimulates the activity of the polymerization ATPase–platform protein subcomplex, leading to polymerization of major subunit fibres. Model 2, spontaneous formation of a minor subunit complex that interacts with the platform could template the subsequent addition of major subunits. Model 3, the spontaneous formation of a minor subunit complex both activates the assembly platform and provides a scaffold for the addition of major subunits. CM, cytoplasmic membrane.

Cisneros et al. (2011) showed recently that the Klebsiella oxytoca GspIJK equivalents form an inner membrane subcomplex with quasihelical symmetry. Based on molecular dynamics simulations, they proposed that formation of the subcomplex led to local membrane deformations. They showed that the minor pseudopilin subcomplex primed pseudopilus assembly, either by activating the platform protein and assembly ATPase through membrane deformations to begin polymerization, or by forming a stable nucleation site for the subsequent addition of major subunits (Cisneros et al., 2011). In this issue of Molecular Microbiology, they asked whether the T4aP minor pilins have a similar function in pilus assembly.

They exploited the observation that heterologous expression of the K. oxytoca T2S Pul system in E. coli leads to assembly of the E. coli T4a pilin, PpdD (Sauvonnet et al., 2000). They show that PpdD assembly occurs only in the presence of a complete set of T2S genes and not with a subset lacking the gspHIJK equivalents, pulHIJK. The amount of PpdD assembly was similar regardless of whether the Pul minor pseudopilins or the native E. coli minor pilins were provided, suggesting that priming is independent of the nature of the minor subunits. In support of this idea, formation of pseudopili could be initiated by the T4a minor pilins, although not to the same extent as the cognate minor pseudopilins. Coexpression of the major pseudopilin PulG and major pilin PpdD reduced total T4aP levels compared to expression of PpdD alone, suggesting that the major subunits compete for a limiting number of priming complexes. Only homopolymers of PulG and PpdD were observed upon coexpression, suggesting that regardless of the priming mechanism, the major subunits do not form mixed fibres. Interestingly, expression of the Pul genes in E. coli allows secretion of the enzyme pullulanase. However, when the T4a minor pilins were used to prime pseudopilus assembly, the resulting fibres were unable to support pullulanase secretion. This finding may support an additional, non-assembly role for the T2S minor pseudopilins, corroborating a previous report of direct interaction between minor pseudopilins and secreted substrates (Douzi et al., 2011). However, it is possible that the amount of pseudopilus assembly supported by the T4a minor pilins is simply too low for efficient secretion.

This study has important implications for our understanding of assembly initiation in the T2S and T4P systems, but leaves many open questions. In particular, the issue of whether minor subunit complexation simply spurs the assembly platform and its associated ATPase into action without the complex becoming part of the resulting fibre (Fig. 1) needs further study. An ‘activation’ model in which various sets of minor subunits stimulate the assembly machinery without incorporation is consistent with the authors' inability to detect minor pseudopilins in assembled fibres, and with the low sequence identity between minor pilins and pseudopilins that could preclude direct interaction with heterologous major subunits. However, studies in the T4aP systems of Neisseria gonorrhoeae and Pseudomonas aeruginosa showed that minor subunits are present in sheared pili (Winther-Larsen et al., 2005; Giltner et al., 2010), suggesting that they are an integral part of the final product and supporting a ‘scaffolding’ model (Fig. 1). It should be noted that these two models are not mutually exclusive. If scaffolding occurs, the fact that major subunits preferentially form homopolymers – an indicator of specificity – needs to be reconciled with potential interactions of a particular major subunit with heterologous tip complexes.

The results clearly support a common functional role for at least some minor subunits, although whether priming of T4aP assembly requires all four minor subunits as tested here, or fewer, as shown previously for hyper-pseudopilus assembly (Sauvonnet et al., 2000; Cisneros et al., 2011), is not yet clear. Earlier studies suggested that although GspH (PulH) is required for optimal pullulanase secretion in K. oxytoca, it was dispensable for hyper-pseudopilus formation (Sauvonnet et al., 2000). The role of this particular subunit is currently unclear. Douzi et al. (2009) suggested that it might couple the GspIJK subcomplex to the major subunit GspG, but could not demonstrate a GspG–GspH interaction. The equivalent T4aP minor subunit (PpdB in E. coli, PilH in Neisseria, FimU in P. aeruginosa; Fig. 1) is important for pilus assembly, although it and all other minor pilins are dispensable in retraction-deficient backgrounds (Winther-Larsen et al., 2005; Carbonnelle et al., 2006; Giltner et al., 2010). As pili can assemble – albeit with reduced efficiency – in the absence of the minor subunits when retraction is blocked, those components have been proposed to antagonize fibre disassembly. However, this facet of their biology is poorly defined.

While many members of the T4aP subfamily have homologues of the GspHIJK proteins, the T4bP subfamily, including the monophyletic tight adherence (Tad) pili, has fewer minor subunits (Tomich et al., 2007; Pelicic, 2008). Among those that can be identified (with a type IV pilin-like N-terminal sequence), similarity to GspHIJK ranges from limited to non-existent. In some cases, only a single minor subunit has been identified, which functions as a tip adhesin (Yoshida et al., 1999; Ishiwa and Komano, 2004). Thus, it is not clear how far the model of assembly initiation described by Cisneros et al. can be extrapolated when considering more distant relatives. Based on the reported ability of GspIJ alone to initiate pseudopilus assembly (Cisneros et al., 2011), it is conceivable that fewer minor subunits are sufficient for priming in some systems.

In addition to the four conserved ‘core’ minor subunits (Fig. 1), many bacteria have ancillary species-specific minor pilins that make important contributions to T4P biology (Brown et al., 2010). Among the best characterized examples is the PilX protein from Neisseria meningitidis, which antagonizes pilus retraction and modulates fibre properties (Helaine et al., 2005; 2007; Brissac et al., 2012). Its structure revealed a hook-like protrusion in the C-terminus that was proposed to catch on PilX subunits in adjacent antiparallel pili, opposing their retraction (Helaine et al., 2007). The resulting force generation can cause the partial unravelling of pili (Brissac et al., 2012). PilX activity is important for the ability of meningococci to signal to epithelial cells (Coureuil et al., 2012). Another Neisseria-specific minor pilin, PilV, is essential for the T4P-dependent recruitment of the host polarity complex and cell junction proteins to cortical plaques beneath bacteria bound to endothelial cells (Lecuyer et al., 2012). The resulting depletion of junction proteins compromises the blood–brain barrier, allowing meningococci to cross.

Besides the functions listed above, minor subunits may have others that we have not yet imagined. Kuchma et al. (2012) recently found that the GspJ and GspK orthologues in the P. aeruginosa T4aP system repress surface-associated swarming motility by modulating levels of the intracellular secondary messenger, cyclic-di-GMP. Surprisingly, the proteins regulated these behaviours while in the inner membrane, as they continued to function when processing of their signal peptides was blocked. With the large number of potential type IV pilin-like proteins that were discovered in a limited search using the recently developed PilFind algorithm (Imam et al., 2011), we are likely seeing only the tip of the function iceberg.

The current study by Cisneros et al. is an important step forward in understanding the shared mechanisms of assembly in the T2S and T4P systems. With growing interest in the nature and extent of their functional contributions, it is now prime time for the minor subunits.

Acknowledgements

I thank Ylan Nguyen for helpful input. Work in my laboratory on type IV pilins is supported by Operating Grant MOP 86639 from the Canadian Institutes of Health Research.

Ancillary