The Injectisome, a Complex Nanomachine for Protein Injection into Mammalian Cells

Abstract

Type III protein secretion systems (T3SSs) or injectisomes are multi-protein nanomachines present in many gram-negative bacteria that have sustained long-standing close relationship with a eukaryotic host. These secretion systems have evolved to modulate host-cellular functions through the activity of the effector proteins they deliver. To reach their destination, T3SS effectors must cross the multi-barrier bacterial envelope and the eukaryotic cell membrane. Passage through the bacterial envelope is mediated by the needle complex, a central component of T3SSs that expands both the inner and outer membranes of gram-negative bacteria. A set of T3SS secreted proteins, known as translocators, form a channel in the eukaryotic plasma membrane through which the effector proteins are delivered to reach the host cell cytosol. While the effector proteins are tailored to the specific life style of the bacterium that encodes them, the injectisome is conserved among the different T3SSs. The central role of T3SSs in pathogenesis and their high degree of conservation make them a desirable target for the development of antimicrobial therapies against several important bacterial pathogens.

INTRODUCTION

Type III protein secretion systems (T3SSs) are multi-protein nanomachines present in many gram-negative bacteria with a close relationship with a eukaryotic host. The primary function of these machines is the delivery of bacterially encoded effector proteins into target eukaryotic cells(1–4), to modulate a myriad of cell biological processes for the benefit of the bacteria that encode them(5, 6). T3SSs are widespread in nature playing a central role in the pathogenic and symbiotic interactions between many bacteria and their hosts. Among the bacteria that encode T3SSs are many important human and plant pathogens. As the field has progressed, so has the amount of information available therefore precluding a comprehensive review of the literature. Therefore, in this chapter we will focus on the structural and architectural aspects of the type III system. To reflect current knowledge, and to help the reader better understand the structural organization of this machine, throughout this chapter, we will refer to the complete type III secretion machine as the injectisome, and we will describe in detail the different substructures that integrate it (i.e. the needle complex, the export apparatus, and the sorting platform). Readers are referred to other reviews for more specific aspects of the structure and function of these secretion machines(1–4).

The injectisome

The main structural element of the type III secretion system is the injectiome, a complex multi-protein structure composed of extracellular, envelope-associated and cytoplasmic elements or substructures. To facilitate their description, each of these elements will be discussed separately. However, it should be emphasized that the type III secretion machine is and operates as one single functional structural unit and that its separation into different sub-components, while useful, is somewhat arbitrary. Also, we note that the increasing knowledge on the structural organization of this system has rendered some previous descriptions of this system more confusing. Furthermore, the plethora of gene names has made comparison of systems across species somewhat challenging, thus, when referring to specific components of the injectisome, we will utilize a previously proposed universal nomenclature. Table 1 lists the specific names of proteins of the most studied T3SSs.

Table 1|.

Principal Components of Most Studied Type III Secretion Systems

Universal Nomenclature	Function	Yersinia	Salmonella SPI-1	Salmonella SPI-2	E. coli (EPEC/EHEC)	Shigella	Chlamydia	P. aeruginose	P. syringae	Rhizobium	Flagellum
Needle Complex (Flagellar Basal Body)
SctC	Secretin (OM ring)	YscC	InvG	SsaC	EscC	MxiD	CdsC	PscC	HrcC	RhcC1-RhcC2	N/A
—	Secretin pilotin	YscW	InvH	—	—	MxiM	—	ExsB	HrpT	—	N/A
SctD	IM ring	YscD	PrgH	SsaD	EscD	MxiG	CdsD	PscD	HrpQ	Y4yQ	FliG
SctJ	IM ring	YscJ	PrgK	SsaJ	EscJ	MxiJ	CdsJ	PscJ	HrcJ	NolT	FliF
SctF	Needle subunit	YscF	PrgI	SsaG	EscF	MxiH	CdsF	PscF	HrpA	NopA-NopB	FlgE (Hook protein)
SctI	Inner rod subunit	YscI	PrgJ	SsaI	EscI	MxiI	—	PscI	HrpB	NolU	—
Export Apparatus (Inner Membrane Proteins)
SctU	Autoprotease/Su bstrate switching	YscU	SpaS	SsaU	EscU	Spa40	CdsU	PscU	HrcU	RhcU	FlhB
SctV	Inner membrane channel	YscV	InvA	SsaV	EscV	MxiA	CdsV	PcrD	HrcV	Y4yR	FlhA
SctR	Inner membrane channel	YscR	SpaP	SsaR	EscR	Spa24	CdsR	PscR	HrcR	RhcR	FliP
SctS	Inner membrane channel	YscS	SpaQ	SsaS	EscS	Spa9	CdsD	PscS	HrcS	RhcS	FliQ
SctT	Inner membrane channel	YscT	SpaR	SsaT	EscT	Spa29	CdsT	PscT	HrcT	RhcT	FliR
Sorting Platform (Flagellar C-ring)
SctQ	Core scafold	YscQ	SpaO	SsaQ	SepQ	Spa33	CdsQ	PscQ	HrcQ	RhcQ	FliM-FliN
SctN	ATPase	YscN	InvC	SsaN	EscN	Spa47	CdsN	PscN	HrcN	RhcN	FliI
SctL	Core scafold	YscL	OrgB	SsaK	EscL	MxiN	CdsL	PscL	HrpE	NolV	FliH
SctO	SctN/SctV linker	YscO	InvI	SsaO	EscO	Spa13	CdsO	PscO	HrpO	Y4yJ	FliJ
SctK	Core scafold	YscK	OrgA	—	EscK	MxiK	—	PscK	HrpD	—	—
Regulatory proteins
SctP	Needle assembly regulator	YscP	InvJ	SsaP	EscP	Spa32	CdsP	PscP	HrpP	—	FliK
SctW	Regulator of translocase secretion	YopN-TyeA	InvE	SsaL	SepL	MxiC	CopN	PopN	HrpJ	—	—
—	Initiation of needle assembly	?	OrgC	?	?	MxiL	?	?	?	?	—
Translocators and tip complex protein
SctB	Translocon	YopD	SipC	SseD	EspB	IpaC	CopD	PopD	—	—	—
SctE	Translocon	YopB	SipB	SseC	EspD	IpaB	CopB	PopB	HrpK	NopX	—
SctA	Tip protein	LcrV	SipD	SseB	EspA	IpaD	CT584	PcrV	—	—	—

1). The needle complex

The needle complex (NC) is a major core element of the T3SS injectisome and its discovery in 1998 constituted a major breakthrough in the understanding of type III secretion machines(7). Up until then, T3SSs were simply a collection of genes required for protein secretion without a clear framework to explain how they may constitute a protein secretion machine. Since the first visualization and isolation from S. Typhimurium(7), it has been visualized in other bacteria showing a conserved architecture(8–10). It consists of a multi-ring base substructure embedded in the bacterial envelope and a needle-like extension protruding several nanometers from the bacterial surface (Figure 1). The needle is linked to the base through a substructure known as the inner rod, which docks into a socket-like structure within the NC composed of the export apparatus, which is thought to form a conduit to facilitate the passage of effector proteins through the bacterial inner membrane. At the distal side of the needle lies the needle tip complex, which senses the target host cell.

Figure 1.

Salmonella Typhimurium SPI-1 encoded type III secretion system. (A) Surface view of the 3D reconstruction of the single particle cryo-EM map of the needle complex substructure with the atomic structures of the different needle complex components docked. OR1: outer ring 1; OR2: outer ring 2; IR1: inner ring 1; IR2: inner ring 2. (B) Central section of an overall cryo-ET structure of the complete injectisome in situ. Of note is the location of IR2 in the cytosolic side of the bacterial envelope. (C) Molecular model of the organization of the injectisome in situ with available atomic structures fitted into the model (figure adapted from DOI: 10.1016/j.cell.2018.01.034)

The base

The NC base is a multi-ring structure that spans the inner and outer membranes of the bacterial envelope. Despite its architectural complexity, the base of the NC is composed of a relatively small number of proteins. The inner rings (IR1 and IR2), located in the bacterial inner membrane and the cytoplasm, are composed of two proteins (SctJ and SctD form the IR1 while SctD alone forms the IR2), whereas the outer ring (OR) and the neck of the base are composed of a single protein (SctC), a member of the secretin family of outer membrane proteins. Single particle cryo-electron microscopy (cryo-EM) of isolated complexes has revealed a detailed view of the NC of the S. Typhimurium Pathogenicity Island 1 (SPI-1) encoded T3SS(11–13). This structure revealed that 15 SctC molecules form a double-walled stranded β-barrel complex in the outer membrane, that connects directly to 24 molecules of the inner ring components, amounting to a symmetry mismatch(12, 13). This mistmatch confers flexibility to the neck, which may be important during NC assembly. The IRs are composed of two concentric rings. A larger peripheral ring formed by SctD encircling a smaller internal ring formed by SctJ. This smaller ring is shielded on the sides by the SctD ring and on top by the SctC neck. Both SctD and SctJ share a similar although inverted topology, with a periplasmic and a cytoplasmic domain separated by a single transmembrane segment. The N-terminal domain of SctJ is lipidated and located in the periplasm and the C-terminal cytoplasmic domain is very short or absent in some homologues which also lack the transmembrane domain(14). In contrast, SctD has a longer N-terminal cytoplasmic domain that forms the IR2 and links the NC base to the sorting platform (see below). The soluble domains of some base components have been crystalized [i. e. the periplasmic domains of the outer membrane secretin EscC (SctC)(15), PrgH (SctD)(16) and EscJ (SctJ)(17)]. In all cases, the atomic structures showed a modular architecture of three topologically similar α/β domains. A comparison of the structures revealed a strong similarity between them despite the lack of detectable sequence identity. The fact that these three proteins arrange in a ring led to the hypothesis that this modular fold may represent a ring-building motif(15). However, this domain is also present in proteins that do not organize in rings and has been shown to be dispensable for ring formation(18). Therefore, the relationship between the presence of this domain and the ability to organize in rings remains unclear. The structure of the NC in situ determined by cryo electron tomography (cryo-ET) has shown that although the outer ring, neck, and inner rings of the in situ and isolated structures are virtually identical, there are some unique features revealed by the cryo-ET analysis(19) (Figure 1). The outer ring of the NC is inserted into the inner leaflet of the outer membrane resulting in an “inward pinch”. Furthermore, although the IR1, which is predicted to be located in the periplasmic space, completely overlap in the isolated and in in situ structures, the IR2, which is located in the cytoplasm, does not. In the in situ structure this ring is pushed further away from IR1 to accommodate the inner membrane, which separates both rings. Upon assembly of the sorting platform, the IR2 undergoes a significant conformation change to adopt a “6 patch” organization to accommodate the 6 pods of the sorting platform (see below)(19).

The needle

The NC base has a several nanometer-long needle extension that confers the NC its “syringe-like” appearance. The needle substructure is composed of a single, small protein subunit, SctF, polymerized in a helical fashion(20–23). In its native arrangement the length of the needle ranges between 30 and 70 nm, and its width from 10 to 13 nm. The atomic structure of the needle protein shows a α-helical hairpin arrangement with two α-helixes of similar size separated by a short loop most often containing two proline residues separated by two amino acids (the PXXP motif)(24–26). More recent studies involving cryo-EM(27) and solid state NMR(23) of recombinant and native needle polymers indicate that the needle has a right-handed helical organization consisting of ∼5.7 subunits per turn and a helical pitch of ∼24 Å. The entire assemble is traversed by a channel ∼25 Å in diameter. Although initially there were two incompatible atomic models for the needle polymer, several studies have now conclusively shown that the subunit orientation within the needle polymer places the extended N-terminus of the protomer facing the outside of the polymer and the C-terminus of the subunit facing the lumen. Residues that face the lumen of the needle are highly conserved and mostly polar, and analysis of their electrostatic potential reveals alternating positively- and negatively-charged regions. The implications of that observation are unclear but it is possible that such organization could play a role in the mechanisms of secretion through the needle channel. Within the filament, the individual proteins are stabilized by multiple inter- and intra-subunit contacts resulting in a rather rigid structure. Attention has been placed on a small kink (residues Val-20 to Asn-22 in the case of the S. Typhimurium SPI-1 needle protein) that interrupts the N-terminal α-helix and that it is not observed in the crystal structure of the soluble protomer. The implications of this observation has not been determined but it is conceivable that it could play a role in signal transduction during activation of the injectisome upon contact with target cells.

The inner rod

The inner rod is a substructure that links the needle to the NC base by docking to the export apparatus(11). Although by analogy with the flagellar system this substructure was originally referred to as a “rod”, cross-linking as well as stoichiometry studies suggest that this structure is most likely akin to a “washer” rather than a rod since it is predicted to have ~6 subunits(28, 29). Like the needle substructure, it is built from a single small protein subunit, SctI. The atomic structure of the inner rod of S. Typhimurium has been solved by CD and NMR spectroscopy(30). In its soluble monomeric form this protein is largely unfolded lacking tertiary structure. However, computational methods have suggested that the inner rod subunit shares a similar structure to that of the needle protein, an α-helical hairpin shape flanked by flexible regions. in silico modeling has also determined that the domains critical for filament assembly are well conserved between the needle and inner rod protein. However, the two subunits differ significantly at their N-terminus, and based on the needle filament structure, these differences would not allow the inner rod to elongate beyond one turn of the helix (~6 subunits)(28). This substructure has been implicated in substrate switching and needle length control(31, 32).

The needle tip complex and needle extension

The T3SS is inactive prior to contact with the eukaryotic host. In this inactive state, the needle filament is capped by a single protein that organizes in a tip complex(33, 34), or it is extended by another filament longer than the needle itself(35). The tip complex and the needle extension, play a role in sensing the environment preventing the premature unproductive secretion of effectors.

Based on their structure and biochemical properties the proteins that make up the tip complex can be divided in two groups: the SipD/IpaD (from Salmonella and Shigella respectively) and the LcrV/PcrV (from Yersinia and Pseudomonas). SipD/IpaD-like tip proteins are organized in an N-terminal α-helical hairpin, a long central coiled-coil domain, and a C-terminal region containing a mixture α/β domains(36, 37). The central coiled-coil domain is characteristic of all tip proteins although in LcrV/PcrV it is flanked by globular domains on the N- and C-termini, that give them a dumb-bell appearance(38). The central coiled-coil domain is important for the interaction of the tip protein with the needle filament and the needle protomer is expected to bind at multiple sites on this domain. The N-terminal α-helical hairpin folds independently and has been shown to act as a self-chaperone preventing the untimely oligomerization of the SipD/IpaD subunits in the bacterial cytoplasm. This self-chaperoning domain is absent in the LcrV/PcrV family of tip proteins, where a small cytoplasmic protein (LcrG/PcrG) functions as a chaperone instead(39).

Several crystal structures of different tip proteins have been solved, either alone(37, 38, 40), in combination with the needle protein (PrgI-SipD)(41), or in complex with bile salts(42, 43). What is needed however, is an atomic resolution view of the tip protein complex assembled at the needle tip. Presently there are low-resolution 3D reconstructions from electron micrographs of the Yersinia LcrV tip(44) and the Shigella IpaD tip(33). The Yersinia LcrV tip complex showed a well-defined structure characterized for the presence of a base (formed by the N-terminal globular domain), a neck (formed by the coiled-coil region), and a head (comprising the C-terminal globular domain). This precise organization could be accounted by a pentameric LcrV ring. The low-resolution structure of the IpaD tip complex is also compatible with a pentameric organization of the tip protein. This complex however does not show the characteristic morphology of the LcrV tip complex. It has been proposed that the Shigella tip complex may exhibit two different compositions, a homo-pentameric IpaD tip complex and a hetero-pentameric complex consisting on four IpaD molecules along with one IpaB molecule(45). Quantification of the relative abundance of the two different complexes was not feasible, therefore it is possible that the IpaD-IpaB complexes may represent a low proportion of injectisomes that have been activated prematurely while the pentameric IpaD structure represents the resting tip complex.

A significant modification in the tip complex occurs in several bacterial species including pathogenic E. coli strains (EPEC/EHEC). These injectisomes are characterized by the presence of a filamentous extension to the needle substructure, which in E. coli is formed by a single protein, EspA(35). Similar to the mechanism by which flagellin assembles into the flagellum, the EspA filament assembles by coiled-coil interactions between EspA subunits. However, electron micrographs of negative stained EspA filaments showed that they are distinct from flagella(46). The 3D reconstruction of the EspA filaments shows that they consist of a helical tube ~120 Å wide, containing a hollow central channel of ~25 Å in diameter with a continuous channel through which effector proteins are translocated. The EspA filament shows helical symmetry having 28 subunits in 5 turns for a 1 start helix (5.6 subunits/turn). A later study using cryo-EM of frozen-hydrated filaments has shown that the EspA filament displays heterogeneity in the structure(47) due to a fixed rotation between subunits, but a variable axial rise between adjacent subunits. How this variability in the structure relates to the function of the filaments has not yet been addressed.

The export apparatus

All T3SSs contain five conserved inner membrane proteins that are essential for their function, SctV, SctR, SctS, SctT, and SctU(48–51). Cryo-EM studies have correlated the presence of a defined density inside the inner membrane rings with the presence of the inner membrane proteins, indicating that at least a subset of these inner membrane proteins are located inside the NC structure creating a channel through which the secreted proteins can traverse the inner membrane(52). More recently, a cryo-EM structure of an in vitro assembled complex of 3 export apparatus components, SctR, SctS and SctT, from the homologous flagellar export apparatus was solved providing major insight into the organization of this substructure(53). The complex adopts a helical configuration and is organized with a stoichiometry 5:4:1 for FliP(SctR), FliQ(SctS), and FliR(SctT), respectively. Remarkably, despite the presence of several predicted transmembrane domains, none of these proteins adopt a canonical membrane protein configuration. Rather, the complex is arranged in a helical configuration in which a FliP/FliQ pair and an additional FliP molecule combine with one copy of FliR. FliR is structurally equivalent to the FliP/FliQ pair so the structure consists of six copies of a FliR-like element forming a single helical turn. The helical arrangement of these export apparatus components provides an optimal platform onto which the inner rod/needle filament can be effectively assembled. Two additional components of the export apparatus, SctV and SctU, whose location within the needle complex have not been precisely determined, are likely to play a more specialized role in type III secretion. SctV has a large cytoplasmic domain that crystalizes as a circular nonamer(18) and by cryo-ET can be seen as toroidal shape density immediately bellow the cytoplasmic side of the IR2 ring of the NC(19). It is possible that SctV may play a role in preparing the substrates for translocation through the export apparatus. It has also been suggested that the SctV family of proteins may work as a proton channel to energize the secretion process, although this activity has not been formally demonstrated(54). Another member of these inner membrane proteins, SctU, plays a role in substrate switching of the machine specificity form early substrates (i.e. needle complex components and regulators) to translocators and effectors. SctU, also known as the “switch protein”, is a protease that undergoes autocatalytic cleavage(55–62). This auto-cleaving event was proposed to be the signal that triggers substrate switching from early substrates (needle and inner rod proteins and other accessory proteins) to middle and late substrates (translocators and effectors). However, more recent experiments have demonstrated that the cleavage event per se does not provide a signal for substrate switching and that the cleavage may simply be required to provide SctU with the appropriate conformation for its secretion function(63).

2). The sorting platform

There are several conserved cytosolic proteins, SctQ, SctK, SctL, and SctN, that are required for type III secretion. They arrange in a complex that operates as a sorting platform to organize the secretion process through the T3SS establishing a hierarchy in the order in which substrates are secreted(64). The proteins that constitute the sorting platform are highly conserved among all T3SSs and have been shown to interact with one another(64–66). Recently, the structural organization of these proteins in situ was revealed by cryo-ET(19, 67)(Figure 1). The sorting platform exhibits a cage-like architecture, enclosed by 6 pod-like structures that emerge from the NC and converge into a 6-spoke wheel-like structure at its cytoplasmic side. This structure serves as scaffold to place the associated ATPase SctN and SctO in close apposition to the export apparatus. These studies have also revealed that SctQ accounts for most of the protein density associated with the pods and is linked to the wheel-like capping structure formed by SctL on one side and SctK, which links the sorting platform to the NC, on the other. The presence of the 6-pod structure stands in contrast with the appearance of a related structure found in the flagellar apparatus known as the C-ring, which forms a closed ring stably linked to the flagellar basal body(68). These structural differences may reflect the fundamentally different roles they play. There is evidence indicating that the sorting platform exhibits a dynamic behavior with cycles of assembly and disassembly although the specific role of this behavior in the function of T3SS or the mechanisms by which this behavior may be controlled are not currently known(69–71). It is possible that an alternatively translated product of the open reading frame encoding the core sorting platform components SctQ may be involved in this process(72–75). However, more research will be required to clarify these poorly understood aspects of type III secretion. The specific mechanisms by which the sorting platform may engage substrates is also poorly understood although it is expected that the associated ATPase, SctN, with the help of SctO may play a central role in the process of bringing the substrates to the ring formed by the cytoplasmic domain of SctV on the cytosolic side of the NC base.

Assembly of the injectisome

The assembly of the injectisome occurs in a step-wise manner (Figure 2). First, the export apparatus components SctR, SctS, and SctT, are inserted in the membrane(52) and subsequently form a pseudo hexameric assembly(53). Later, two additional membrane proteins, SctU snd SctV, are added to the export apparatus. The assembled export apparatus then serves as a template for the assembly of the inner rings of the NC(52). Nucleation of the inner rings around the export apparatus may also result in the “extraction” of the export apparatus components from the plasma membrane(53).

Figure 2.

Model of the step-wise assembly of the injectisome. SctRST form a stable complex in the inner membrane (IM) to which SctU is recruited. This complex nucleates the assembly of the inner rings integrated by SctJ and SctD, which results in the extraction or “pulling” of the IM components from the bacterial plasma membrane. At the same time, the secretin is independently assembled into the outer ring and the two structures come together to form the needle complex base substructure to which SctV is subsequently recruited. Once the NC base is formed, the cytoplasmic sorting platform is recruited to the cytoplasmic side of the NC base and the system starts to function as a type III secretion machine dedicated to the delivery of early substrates such as the inner rod (SctI) and needle (SctF) subunits to complete the assembly of the entire injectisome.

The ORs and neck of the base are made of a single protein, SctC, belonging to the secretin family of outer membrane bacterial proteins(76). With the assistance of a “pilotin” SctC assembles into a pore in the outer membrane(77–80). Once the ORs and the IRs in association with export apparatus are independently assembled, they come together to form a complete NC base substructure. The cytoplasmic accessory proteins that form the sorting platform are recruited to the NC base that starts to function as a T3SS dedicated to the secretion of early substrates (22, 81). These early substrates include proteins that will create the needle, SctF, and the inner rod, SctI, as well as accessory regulatory proteins SctP, and the S. Typhimurium protein OrgC, which regulate the assembly process(82). Completion of the needle triggers substrate switching from early substrates (needle complex component and regulators of needle complex assembly) to translocators and effectors, and the injectisome is now ready to be activated by the contact with the host cell. An alternative outside-in assembly model has been proposed for Yersinia spp. although current evidence is not consistent with such model(83).

Activation of the type III secretion machine

A distinctive feature of T3SSs is that they require an activating signal to secrete and translocate effectors. Although the nature of the activating signal is poorly understood, in most cases it derives from the bacterial contact with target cells(84–86), although other agonists have also been described(87, 88). The activation step is presumably necessary to ensure that the effectors are not unproductively secreted into the extracellular environment prior to host cell contact. How cell contact triggers the activation of the T3SS is not known but the tip complex is predicted to be involved in the process. In support of this hypothesis it is possible to activate secretion in vitro by the addition of compounds such as bile salts or congo red that bind the tip complex and presumably induce conformational changes similar to those that may occur upon cell contact(42, 43, 89). These conformational changes are thought to be transduced to the secretion machine through the needle and inner rod structures. Consistent with this hypothesis, mutations in the needle and inner rod proteins have been identified that result in constitutive or altered secretion phenotypes(28, 90–92). Activation of the secretion machine leads to secretion of the translocators (SctB and SctE), which are deployed on the eukaryotic plasma membrane to form the translocation pore or translocon through which effector proteins directly reach the cytosol of the host cell (Figure 3). Recent cryo ET studies have provided insight into the organization of the translocon(93). These studies have observed a well-defined ‘bend’ on the target cell membrane in areas where the needle substructure makes contact with the host cell interface, reflecting the intimate association that is known to be required for optimal T3SS-mediated effector translocation(94). Notably, these studies showed the presence a distinct density within the region of the target host cell membrane in close apposition to the needle tip of the T3SS injectisome, which was correlated to the presence of the translocon. Although the structure was not detailed enough to provide insight into the stoichiometry and/or molecular organization of the translocon, it showed a structure ~13.5 nm in diameter and 8 nm in thickness, which was smaller than a structure (~60 nm in diameter) of the enteropathogenic E. coli translocons assembled from purified components on red blood cells(95). The reason for the different dimensions are unclear but they may reflect the differences between the in vivo and in vitro deployment of the translocases. For the translocation to be productive, the translocases must be engaged by the secretion machinery preceding the effectors, through mechanisms involving the sorting platform. Consistent with this notion, only the translocases can be detected in complex with the sorting platform prior to host cell contact, and it is only in the absence of the translocases that the effectors can be detected at this location(64).

Figure 3.

Model of the injectisome’s interaction with an eukaryotic host cell. Activation of the injectisome leads to secretion of the translocators that are deployed on the eukaryotic plasma membrane to form the translocon, which remains in contact with the needle to form a direct conduit between the bacterial and host cell cytosol that serves a passageway for the effector proteins (figure adapted from DOI: 10.7554/eLife.39514).

Concluding remarks

Type III secretion protein machines have sparked the interest of scientists for more than two decades. Although there has been remarkable advancements in the understating of the structure and function of these machines many knowledge gaps still remain. For example atomic information on some essential elements of the injectisome are still missing, the precise mechanism of substrate engagement by the secretion machine are poorly understood, and the mechanisms of eukaryotic cell sensing and signal transduction remain obscure. Addressing some of these fundamental questions most likely will require the development of novel experimental approaches to be able to study the function of these machine in live bacteria. The presence of these machines in many important bacterial pathogens has made them attractive targets for the development of next generation antimicrobials that can be deployed to prevent and combat many important infectious diseases.