Cellular Microbiology (2015) 17(2), 174–182

doi:10.1111/cmi.12400 First published online 8 January 2015

Microreview Cytoskeletal mechanics during Shigella invasion and dissemination in epithelial cells Cesar M. Valencia-Gallardo,1,2,3,4,5 Nathalie Carayol1,2,3,4 and Guy Tran Van Nhieu1,2,3,4* 1 Equipe Communication Intercellulaire et Infections Microbiennes, Centre de Recherche Interdisciplinaire en Biologie (CIRB), Collège de France, Paris, France. 2 Institut National de la Santé et de la Recherche Médicale (Inserm) U1050, Paris, France. 3 Centre National de la Recherche Scientifique (CNRS) UMR7241, Paris, France. 4 MEMOLIFE Laboratory of Excellence and Paris Science Lettre, Paris, France. 5 Université Paris Diderot – Paris 7, Paris, France. Summary The actin cytoskeleton is key to the barrier function of epithelial cells, by permitting the establishment and maintenance of cell–cell junctions and cell adhesion to the basal matrix. Actin exists under monomeric and polymerized filamentous form and its polymerization following activation of nucleation promoting factors generates pushing forces, required to propel intracellular microorganisms in the host cell cytosol or for the formation of cell extensions that engulf bacteria. Actin filaments can associate with adhesion receptors at the plasma membrane via cytoskeletal linkers. Membrane anchored to actin filaments are then subjected to the retrograde flow that may pull membrane-bound bacteria inside the cell. To induce its internalization by normally nonphagocytic cells, bacteria need to establish adhesive contacts and trick the cell into apply pulling forces, and/or to generate protrusive forces that deform the membrane surrounding its contact site. In this review, we will focus on recent findings on actin cytoskeleton reorganization within epithelial

Received 24 October, 2014; revised 26 November, 2014; accepted 1 December, 2014. *For correspondence. E-mail guy.tran-van-nhieu@ college-de-france.fr; Tel. (+33) 1 44 27 14 89; Fax (+33) 1 44 27 14 19.

cells during invasion and cell-to-cell spreading by the enteroinvasive pathogen Shigella, the causative agent of bacillary dysentery.

The Shigella invasion model Shigella are gram-negative entereoinvasive bacteria responsible for diarrhoeal diseases. Shigella dysenteriae and S. flexneri are responsible for bacillary dysentery, causing hundreds of thousands of casualties predominantly in developing countries. Following ingestion, Shigella multiplies in the colonic mucosa and induces an intense inflammatory reaction that leads to its destruction. Key to the virulence of this pathogen, Shigella has the capacity to invade epithelial cells, to replicate freely in the cell cytosol following lysis of the phagocytic vacuole, and to disseminate within the epithelium using actin-based motility. Shigella does not express any classical adhesin and hence does not show constitutive cell binding activity, which likely accounts for its moderate capacity to invade cultured cells in vitro. In vivo, however, it is estimated that between a few hundreds to a few thousands bacteria are sufficient to cause the disease following ingestion, pointing to Shigella as being a highly efficient invasive pathogen. It is possible that the absence of constitutive cell binding and exposure to host recognition systems favours in vivo intracellular bacterial replication at the onset of the infectious process.

Shigella invasion without constitutive cell binding Shigella can be considered as Escherichia coli species adapted to the intracellular lifestyle. Encoded by a large virulence plasmid, a type III secretion (T3S) system (T3SS) and the autotransporter IcsA are required for the invasion and dissemination process in epithelial cells (Marteyn et al., 2012). Type III secretion systems are flagellum-related structures found in many gram-negative bacterial pathogens that allow the injection of bacterial effectors into host cells (Mueller et al., 2008). T3SSs are composed of a basal body, spanning the bacterial inner

© 2014 John Wiley & Sons Ltd

cellular microbiology

Actin mechanics and Shigella and outer membranes, prolonged by a needle capped with a so-called ‘tip complex’ exposed at the bacterial surface. T3S is only activated upon cell contact, providing bacteria the means to hide virulence factors from host recognition systems under non-secreting conditions. The cell contact-dependent activation of the T3SS implicates the recognition of host membrane components by the ‘tip complex’ (Mattei et al., 2011). This recognition leads to the membrane insertion of the T3SS ‘translocon’ connected to the needle, required for the injection of bacterial T3S effectors. The Shigella T3SS permits the injection of approximately 30 bacterial effectors, but only about half of these are expressed constitutively, while the expression of the other half is only up-regulated when the T3SS is active. Constitutively expressed T3S effectors include those required for invasion, while many of the second wave of up-regulated effectors down-regulate inflammatory responses (Ashida et al., 2011). Although Shigella does not express constitutively active adhesins, the surface exposed autotransporter protein IcsA implicated in actin-based motility shows inducible adhesin function during Shigella invasion (Brotcke Zumsteg et al., 2014). The adhesin property of IcsA is only detected following T3S activation that occurs upon cell contact, or during bacterial exposure to bile salts, expected during the gastric journey in the intestinal lumen (Brotcke Zumsteg et al., 2014). There are precedents for a role in bile salts on the conformation of Shigella proteins exposed at the bacterial surface. The bile salt deoxycholate was proposed to bind to the tip complex protein IpaD (Dickenson et al., 2011). Binding of deoxycholate to IpaD may releases the conformation constraints that it imposes on IpaB, allowing IpaB oligomerization and formation of an IpaB homopentamer at the T3S apparatus (T3SA) tip (Dickenson et al., 2011). Although the adhesin property of IcsA is revealed upon bacterial exposure to deoxycholate, it is not observed in bacteria that lack a functional T3SS (Brotcke Zumsteg et al., 2014). This suggests that rather than being directly triggered by bile salts, the IcsA adhesin property requires interaction with one or several T3SS effectors, which still remain to be identified. The fact that mutations that alter IcsA-mediated adhesion but not actin-based motility impair the virulence of Shigella in the mouse keratoconjunctivitis assay, is consistent with the adhesin property of IcsA as being important for Shigella invasion (Brotcke Zumsteg et al., 2014). Bacterial capture by filopodia Filopodia are 100–200 nm diameter wide extensions present on virtually every cell types, containing a limited number of actin filaments and protruding up to several tens of microns from the cell surface. They are involved in © 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

175

a variety of cell processes that implicate adhesion, or cell–cell junctions in the case of epithelial cells, by acting as sensory organelles probing the environment for productive cell–cell contacts (Khurana and George, 2011). Filopodia have been implicated in primary steps of cell invasion by viral pathogens such as the human papilloma virus (Mothes et al., 2010). During this process, viral particles adhere through receptors on the side of filopodia and glide along the filopodial length towards the cell body in a process involving myosin II. However, in the case of Shigella, interaction occurs preferentially between a bacterial pole and the filopodal tip, via the Shigella tip complex components IpaB and IpaD (Romero et al., 2011). Following contact, filopodia retraction brings the bound bacterium in contact with the cell body. As opposed to the retrograde surfing viruses on filopodia, filopodial retraction induced by bacterial contact is not inhibited by the myosin II inhibitor blebbistatin, but depends on actin polymerization (Romero et al., 2011). Force measurements based on optical tweezers, however, suggested that although not essential, myosin II may also participate in filopodial retraction (Quatela, Romero et al., unpubl. results). The dynamics of filopodial elongation and retraction result from the balance between actin polymerization at the filopodial tip leading to filopodia elongation, and actin polymerization in the cell cortex powering the inward pulling of filopodial filaments (Fig. 1A). Bacterial contact leads to decreased polymerization at the filopodial tip and filopodial retraction (Romero et al., 2011). Although still unidentified, receptors may transduce signals through cytoskeletal linkers that reduce actin polymerization at the filopodial tip, upon association with the T3SS tip complex (Fig. 1A). By determining maximal forces exerted by retracting filopodia using optical tweezers, it was found that cytoskeletal–receptor link is likely the key determinant of the adhesion strength at the filopodial tip (Romero et al., 2012). Thus, if this adhesion strength is not sufficient to allow filopodial retraction, bacteria will detach. If the adhesion is productive at the filopodial tip, filopodial retraction brings bacteria in contact with the cell body where invasion occurs. The fact that filopodial capture occurs preferentially via one bacterial pole is consistent with evidence indicating that T3S is polar (Jaumouille et al., 2008). IcsA also localizes at one bacterial pole (Janakiraman and Goldberg, 2004), which also correspond to the pole where the translocon component IpaC is located prior to secretion, perhaps underlying a functional link between IcsA and T3S, as discussed above. Of note, filopodial capture has also been observed for other invasive bacteria, such as Yersinia, that constitutively express invasins at their surface and that invades cells in a ‘zippering mode’, but the role of capture during this internalization mode is unknown (Young et al., 1992).

176 C. M. Valencia-Gallardo, N. Carayol and G. Tran Van Nhieu

Fig. 1. Dynamics of actin filaments during Shigella capture by filopodia and invasion of epithelial cells. Blue arrows represent the direction of forces generated by actin polymerization or the retrograde flow. A. Yellow rectangles depict actin filaments bundling proteins. Actin assembly at the barbed ends of actin filaments is depicted by red circles. Filopodial dynamics of elongation and retraction is controlled by the rates of actin polymerization at the filopodial tip (Vf) and that of the retrograde flow, which includes cortical actin polymerization (Vc) and myosin motors, which may also pull on filopodial actin filaments (blue circles with tails). Non-productive bacterial interaction with the filopodial tip leads to detachment (left panel). Upon productive interaction with the Shigella T3SS tip complex with receptors at the filopodial tip (yellow circles), capping of actin filaments ensues (red bar) inhibiting actin polymerization at the filopodial tip (middle panel). Bacteria-bound filopodia retract through the action of the retrograde flow (right panel). B. Upon contact with the cell body, T3S is induced (purple arrow), thereby triggering the adhesin activity of IcsA. Bacterial adhesion, combined with the anchoring activity of IpaA mediated by its interaction with vinculin allows its pulling, perhaps with IpgB2, by the retrograde flow. The combined action of IpaC, IpgB1 and IpgD stimulates actin polymerization driving the formation of extensions surrounding invading bacteria.

Calcium (Ca2+) signalling and cytoskeletal reorganization during Shigella invasion Cytoskeletal reorganization induced by Shigella promotes pushing and pulling forces at the cell membranes, responsible for the generation of membrane leaflets that surround the bacteria, and bacterial anchoring into the actin foci. As opposed to filopodial capture, which can occur in the absence of T3S, actin foci induced by Shigella upon cell contact are critically dependent on T3S activation (Carayol and Tran Van Nhieu, 2013).

Actin foci linked to Shigella invasion require signalling dependent on inositol(1,4,5) triphosphate (InsP3; Tran Van Nhieu et al., 2013). However, while the effects of InsP3 are largely considered to be exclusively mediated through Ca2+ signalling, Shigella invasion can occur in the absence of global Ca2+ variations (Tran Van Nhieu et al., 2013). Elements of response to this seemingly paradoxical situation were brought about by the characterization of local Ca2+ responses during bacterial invasion. During agonist stimulation, the activation of phospholipases C (PLC) leads to the production of InsP3 following © 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

Actin mechanics and Shigella degradation of phosphatidylinositol (PI)(4,5) bisphosphate (PIP2). Increases in cytosolic Ca2+ then result from the release of Ca2+ from the endoplasmic reticulum following activation of InsP3 receptors. In addition to global Ca2+ increases, Ca2+ can also increase as a result of the stimulation of a discrete number of InsP3R, resulting in local short duration responses, typically lasting a few hundreds of milliseconds. Shigella invasion, however, is associated with local Ca2+ responses lasting up to several seconds (Tran Van Nhieu et al., 2013). The extremely long duration of these local responses challenge the classical model of Ca2+ diffusion, and is explained by the pronounced recruitment of InsP3 receptors, combined with a restriction in diffusion imposed by the polymerization of actin in bacterial invasion sites. The presence of a dense mesh of polymerized actin at invasion sites is probably sufficient to explain delays in the diffusion of InsP3 and therefore, the long duration and local nature of Ca2+ release at the bacterial foci. Ca2+ may directly regulate the function of cytoskeletal proteins involved in Shigella invasion. The initial events leading to PLC activation and InsP3 production during Shigella invasion have not been identified. The injected effectors IpgB1, IpaA and IpgD appear dispensable, but mutants defective for the translocon components IpaB or IpaC fail to recruit PLCs (Tran Van Nhieu et al., 2013). In agreement with the latter observation, there is evidence that destabilization of the plasma membrane by pore forming toxins, is sufficient to activate PLCs, and in some instances, to induce Ca2+ release and cytoskeletal reorganization (Garcia-Saez et al., 2011; Schwan et al., 2013). Thus, it is possible that the insertion of T3SS translocons in the plasma membrane triggers Ca2+ signalling during Shigella invasion. Shigella T3S effectors implicated in actin polymerization and bacterial invasion During the initial stages of cellular invasion, IpaC is critical for actin polymerization and membrane ruffling at bacterial invasion sites, but other T3S effectors also contribute to this process (Fig. 1B; Carayol and Tran Van Nhieu, 2013). The T3S-injected effector IpgB1 stimulates Racdependent actin polymerization through the Arp2/3 complex. While IpgB1 was reported to act as a RhoG mimic recruiting the ELMO/Dock180 complex a GEF for Rac, more recent evidence point that IpgB1 directly acts as a GEF for Rac (Handa et al., 2007; Orchard and Alto, 2012). IpgD, a T3S-injected effector, has a phosphatase activity towards PI(4,5)P2 and generates PI(5)P (Pendaries et al., 2006). Both PI(4,5)P2 hydrolysis and PI(5)P synthesis mediated by IpgD have been implicated in actin polymerization. PI(4,5)P2 allows the tethering of actin filaments to the plasma membrane. Its hydrolysis © 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

177

destabilizes cortical actin, thereby freeing more actin monomers and cytoskeletal components to fuel de novo actin polymerization at bacterial invasion sites (Niebuhr et al., 2000). Consistently, actin foci induced by a Shigella ipgD mutant appear smaller, although the strain does not show detectable defect in epithelial cell invasion. PI(5)P produced by IpgD can also bind to the DH-PH domains of Tiam-1, a GEF for Rac, to stimulate Rac activation at the plasma membrane as well as on early endosomes (Viaud et al., 2014). Tiam-1 is known to also bind to PI(3)P, but with a lower affinity than PI(5)P. These results suggest that in addition to PI(3)P signalling endosomes, PI(5)P may also contribute to the homeostatic balance of cargo degradation and vesicular trafficking, but that PI(5)P is the predominant PI involved in Tiam-1 recruitment at the plasma membrane, Rac activation and actin polymerization (Viaud et al., 2014). How PI(5)P and Tiam-1-dependent Rac activation cooperate with IpgB1 to promote ruffle formation and Shigella invasion remains to be characterized. Interestingly, the Salmonella IpgD orthologue, SopB, was shown to stimulate the Abi/WAVE complex acting in synergy with SopE, a T3S effector acting as a Rac GEF (Humphreys et al., 2012). SopB-mediated PI(4,5)P2 hydrolysis favours the synthesis of PI(3,4,5)P3 and stimulates the recruitment of Arf6 at bacterial invasion sites, and ARNO, a GEF for Arf1 (Humphreys et al., 2012). Similarly, accumulation of PI(3,4,5)P3 also occurs at Shigella invasion sites in an IpgD-dependent manner, as a result from IpgD-mediated dephosphorylation of PIP2, increase in PIs and activation of class I PI3-kinase (Niebuhr et al., 2002). Studies on Salmonella, however, showed that PI(3)P accumulation at bacterial invasion sites does not result from the hydrolysis of PI(3,4)P2 or PI(3,5)P2 by SopB, but from the SopB-dependent recruitment of Rab5 and Vps34 vesicles at invasion sites (Mallo et al., 2008). How PI(3,4,5)P3 accumulates at bacterial invasion sites is not fully understood, since its recruitment appears insensitive to PI3-kinase inhibitors (Mallo et al., 2008). As described for SopB, and through the accumulation of PI(3,4,5)P3, it is possible that IpgD also activates the Abi/WAVE complex and synergize with IpgB1 and IpaC to induce actin polymerization. Shigella effectors of invasion promoting anchoring to the cytoskeleton and Rho-dependent acto-myosin contraction Among the Shigella effectors of invasion, IpaA and IpgB2 do not participate in processes leading to actin polymerization (Fig. 1B). IpaA has been implicated in the anchoring of bacteria in foci of actin polymerization, as well as in the depolymerization of actin filaments during the completion of the Shigella invasion process (Carayol and Tran

178 C. M. Valencia-Gallardo, N. Carayol and G. Tran Van Nhieu Van Nhieu, 2013). The IpaA carboxyterminal domain contains three vinculin binding sites with different characteristics, allowing IpaA to to act as a super talin mimic (Izard et al., 2006; Tran Van Nhieu and Izard, 2007; Park et al., 2011). Vinculin is a focal adhesion protein that stabilizes the anchoring of integrin receptors to the actin cytoskeleton. Upon association of beta-1 integrin receptors with the extracellular matrix, vinculin binds to talin. This leads to the displacement of the inhibitory vinculin head-tail intramolecular interaction and association of the vinculin tail domain with F-actin. In addition to binding F-actin, the vinculin carboxyterminal domain also acts as an actin filament barbed-end capper, preventing the incorporation of actin monomers at the barbed end and filament elongation (Le Clainche et al., 2010). The reasons for the accumulation of VBSs on the IpaA carboxyterminal domain is not understood, but this feature is not uncommon in endogenous vinculin ligands. As suggested for talin, it is possible that IpaA allows the scaffolding of vinculin molecules to stabilize bacterial anchoring to the cytoskeleton during the invasion process (del Rio et al., 2009). The mechanism by which IpaA switches from cytoskeleton anchoring during the initial stages to actin depolymerization at later stages of invasion is not understood. These two activities may be related since the IpaA carboxyterminal domain in complex with vinculin also shows capping activity towards the barbed end of actin filaments (Ramarao et al., 2007). Inhibiting actin polymerization at the filaments’ barbed end is expected to result in filament depolymerization at the pointed end. In transfected cells, the aminoterminal domain of IpaA has also been implicated in the down-regulation of actin stress fibres, in a process involving the combined activation of the RhoA GTPase and competition with talin association with the cytoplasmic domain of the beta-1 subunit of integrins (Demali et al., 2006). The peculiar action of the IpaA aminoterminal domain represents an example of a function reverse to canonical induced by a bacterial effector through multiple actions. While RhoA activation is classically associated with increased stress fibres and focal adhesions, the inhibition of talin association with beta-1 integrins by the IpaA aminoterminal domain, combined with Rho-dependent myosin activation favour stress fibre destabilization, presumably by the pulling of actin filaments away from the cell cortex (Demali et al., 2006). In addition to the IpaA aminoterminal domain, the Shigella T3S effector IpgB2 also activates RhoA, by directly acting as a GEF for RhoA, thereby inducing acto-myosin contractions (Orchard and Alto, 2012). Nevertheless, the role of IpgB2 in invasion is not clear, since a single Shigella ipgB2 mutant does not show a significant impairment in bacterial uptake. In polarized cells, a combination

of ipgB1 and ipgB2 mutations is required to detect an effect on bacterial invasion (Hachani et al., 2008). These results suggest that in these cells, IpgB1 and IpgB2 play a redundant function during Shigella invasion. Rac and RhoA are required for the establishment of adherens junctions (Harris and Tepass, 2012). It is tempting to speculate that IpgB1 and IpgB2 act through the coordinated activation of the Rac and Rho GTPases to hijack processes involved in adherens junctions formation in polarized intestinal cells. Actin-based motility and cell-to-cell spreading Following invasion and escape from the vacuole, Shigella moves intracellularly by promoting actin polymerization and the formation of actin comet tail at one bacterial pole. This is permitted by IcsA, a protein exposed at one pole of the bacterial surface, which binds to the N-WASP protein and activates Arp2/3-dependent actin polymerization (Egile et al., 1999). N-WASP recruitment by IcsA is not sufficient to confer intracellular actin-based motility. Through the action of T3SS effectors that remain to be identified, Shigella also recruits the TOCA-1 protein, which stimulates the conversion of N-WASP from an inactive to an active opened conformation (Leung et al., 2008) (Fig. 2). In addition, tyrosine kinases are also required for Shigella actin-based motility (Burton et al., 2005; Dragoi et al., 2012). The Bruton’s tyrosine kinase (Btk) is up-regulated during Shigella infection and phosphorylates N-WASP. Bruton tyrosine kinase-dependent N-WASP phosphorylation favours its recruitment on the bacterial surface and the nucleation of actin comet tails (Fig. 2; Dragoi et al., 2012). How actin comet tails allow the propelling of bacteria in the host cell cytosol is not fully understood. In the case of actin tails induced by baculovirus, ultrastructural characterization using electron tomography enabled the visualization of a network of actin filaments with branched junctions, favouring a model where filaments continuously tethered to the viral particle, push this latter through the polymerization of actin (Mueller et al., 2014). While this model has been proposed for baculovirus comet tails that only contains a limited number of actin filaments, other models such as the ‘squeezing’ or ‘pushing’ bundles models have been proposed for the propelling by the Listeria actin comet tail, which similar to the Shigella actin comet tail, contains several hundreds of actin filaments organized tangentially to the bacterial surface and filaments branching from these tangential filaments forming the comet tail (Jasnin et al., 2013). Actin-based motility is required but not sufficient to allow Shigella to spread in neighbouring cells. Spreading from cell to cell requires the formation of protrusions at cell–cell junctions, the resolving of these protrusions into © 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

Actin mechanics and Shigella

TOCA-1

179

T3S

P

Btk Dia1

N-WASP

MyoX Dia1

Tricellulin

P

STK11

PI3K Eps1 Dynamin Fig. 2. Regulation of actin-based motility and Shigella cell-to-cell spreading. Following vacuolar lysis, Shigella replicates freely in the host cell cytoplasm. Actin-based motility is triggered by IcsA at one bacterial pole (blue) that recruits N-WASP (purple). TOCA-1 (yellow) is recruited through the activity of the T3SS (grey arrow), and relieves N-WASP’s auto-inhibition to trigger actin polymerization and the formation of comet tails. Btk phosphorylates N-WASP to stimulate its association with bacteria and actin-based motility. At cell–cell junctions, the formation of bacteria-containing protrusions require the formin Dia1 that may orientate the pushing of the IcsA-mediated actin comet at the onset of protrusion formation and help pushing the membrane by nucleating actin polymerization at the filopodial tip. Myosin X also contributes to protrusion elongation, possibly by allowing the transport of membranes or components at the filopodial tip. Protrusions form preferentially at tricellular junctions and their endocytosis by neighbouring cells requires the activation of PI3K, and a clathrin-dependent non-canonical pathway. Tyrosyl phosphorylation of junctional proteins by STK11 also stimulates resolving of protrusions into vacuole. Bacterial replication and dissemination resumes following lysis of the double membrane.

vacuoles and the lysis of these vacuoles leading to bacterial release in the cytosol of neighbouring cells. Junctional, cytoskeletal and signalling components not involved in actin-based motility are also critical for the formation of productive intercellular bacteria-containing protrusions. Activation of the diaphanous-related formin Dia1 is required for protrusions to form. Dia1 activation can allow the proper orientation of actin filaments into an array perpendicular to the cell–cell junction, in order to produce an optimal protrusive force (Fig. 2). Through its actin nucleating activity, Dia1 could also participate to protrusive force generation, which in addition to that produced by the actin comet tail may be critical to overcome constraints at cell–cell junctions (Heindl et al., 2010). The unconventional plus-end directed motor myosin X, also appear to contribute to Shigella-induced protrusions to favour bacterial dissemination (Bishai et al., 2013). Myosin X has been implicated in the transfer of membrane from the basis to the tip of filopodia at the cell periphery. MyoX is anchored through its head domain to actin filaments in the bacterial protrusion, while its tail domain associates with the protrusion’s membrane (Fig. 2; Bishai et al., 2013). Thus, through its plus-end motor activity, MyoX could transport membrane or © 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

membrane-associated components towards the protrusion’s tip to favour the elongation of protrusions containing Shigella (Bishai et al., 2013). Myosin II has also been implicated in Shigella dissemination in epithelial cells, but whether it is implicated in force generation required for protrusion engulfment, or indirectly involved in the maintenance of junctional integrity is not entirely clear (Rathman et al., 2000; Lum and Morona, 2014). The targeting of specific junctional structures is also required during Shigella cell-to-cell spread. This became clear from seminal work implicating the adherens junctions cadherin receptors in the Shigella spreading process (Sansonetti et al., 1994), and has been further documented in recent years. Indeed, recent evidence indicate that Shigella cell-to-cell spread preferentially occurs at the levels of tri-cellular junctions, corresponding to the intersection point of three of more cells of the epithelial layer (Fukumatsu et al., 2012). This preferential site for dissemination is linked to the involvement of tricellular junction proteins, such as tricellulin, in the spreading process. At tricellular junctions, Shigella-induced protrusion formation activates a PI3K-dependent clathrin pathway implicating epsin-1 and dynamin, but not the classical clathrin adaptor AP-2 or Eps15, which eventually leads to engulfment of the bacteria-containing protrusions by adjacent

180 C. M. Valencia-Gallardo, N. Carayol and G. Tran Van Nhieu cells (Fig. 2; Fukumatsu et al., 2012). The resolution of protrusions into bacteria-containing vacuoles in neighbouring cells was also shown to depend on the activity of the serine/threonine kinase STK11/LKB1 (Dragoi and Agaisse, 2014). STK11 activation enables the tyrosine phosphorylation of junctional components associated with the apposition of cell–cell contacts, which await identification (Dragoi and Agaisse, 2014).

Conclusions Here, we have reviewed how the hijacking of cytoskeletal processes enables the production of forces required for the formation of cell extensions and stabilization of bacterial anchorage during Shigella invasion, as well as the formation and engulfment of protrusions by neighbouring cells during bacterial dissemination. In addition to actin reorganization induced at close range by bacteria during invasion and dissemination, Shigella also manipulates the cytoskeleton through T3S effectors that globally act in infected cells. For example, the OspE T3S effector binds to the integrin-linked kinase to stimulate integrin-based adhesion with the extracellular matrix (Kim et al., 2009). Enhanced cell attachment mediated by OspE may serve to prevent the exfoliation of cells, thus preserving time for the bacteria to replicate and disseminate into neighbouring cells (Kim et al., 2009). While for individual effectors, significant progress has been achieved with regard to our understanding of their mode of action, how these effectors and downstream signalling events coordinate is still poorly understood. Among outstanding questions, the precise role of translocon components, how the presumably antagonistic action of the Rac and Rho are regulated downstream of IpgB1 and IpgB2, or how the T3SS converts the IcsA autotransporter into an adhesin during Shigella invasion or allows the recruitment of TOCA-1 to nucleate actin comet tails during dissemination are the most important to tackle. As for invasion, active cytoskeletal reorganization and signalling in recipient cells are required for the dissemination process (Fukumatsu et al., 2012; Dragoi and Agaisse, 2014; Lum and Morona, 2014). Whether this active signalling is induced by the mechanical action of bacterial protrusions invading neighbouring cells, or whether it implicates T3SS-mediated signalling by effectors injected from the donor cell to recipient cells will require further investigation.

Acknowledgements The authors thank F-X Campbell-Valois for critical reading of the article. C V-G is a recipient of PhD grants from the ‘Université Paris-Diderot’ and the Labex ‘Memolife’. NC is funded by a ‘Marie-Curie IRG’ grant and by the Collège de France (CDF). The

work has been supported by grants from the CDF, the ‘Institut National de la Santé et de la Recherche Médicale’, the ‘Centre National de la Recherche Scientifique’ and ‘l’Agence Nationale de la Recherche’.

References Ashida, H., Ogawa, M., Kim, M., Suzuki, S., Sanada, T., Punginelli, C., et al. (2011) Shigella deploys multiple countermeasures against host innate immune responses. Curr Opin Microbiol 14: 16–23. Bishai, E.A., Sidhu, G.S., Li, W., Dhillon, J., Bohil, A.B., Cheney, R.E., et al. (2013) Myosin-X facilitates Shigellainduced membrane protrusions and cell-to-cell spread. Cell Microbiol 15: 353–367. Brotcke Zumsteg, A., Goosmann, C., Brinkmann, V., Morona, R., and Zychlinsky, A. (2014) IcsA is a Shigella flexneri adhesin regulated by the type III secretion system and required for pathogenesis. Cell Host Microbe 15: 435– 445. Burton, E.A., Oliver, T.N., and Pendergast, A.M. (2005) Abl kinases regulate actin comet tail elongation via an N-WASP-dependent pathway. Mol Cell Biol 25: 8834– 8843. Carayol, N., and Tran Van Nhieu, G. (2013) Tips and tricks about Shigella invasion of epithelial cells. Curr Opin Microbiol 16: 32–37. Demali, K.A., Jue, A.L., and Burridge, K. (2006) IpaA targets beta1 integrins and rho to promote actin cytoskeleton rearrangements necessary for Shigella entry. J Biol Chem 281: 39534–39541. Dickenson, N.E., Zhang, L., Epler, C.R., Adam, P.R., Picking, W.L., and Picking, W.D. (2011) Conformational changes in IpaD from Shigella flexneri upon binding bile salts provide insight into the second step of type III secretion. Biochemistry 50: 172–180. Dragoi, A.M., and Agaisse, H. (2014) The serine/threonine kinase STK11 promotes Shigella flexneri dissemination through establishment of cell-cell contacts competent for tyrosine kinase signaling. Infect Immun 82: 4447– 4457. Dragoi, A.M., Talman, A.M., and Agaisse, H. (2012) Bruton’s tyrosine kinase regulates Shigella flexneri dissemination in HT-29 intestinal cells. Infect Immun 81: 598–607. Egile, C., Loisel, T.P., Laurent, V., Li, R., Pantaloni, D., Sansonetti, P.J., and Carlier, M.F. (1999) Activation of the CDC42 effector N-WASP by the Shigella flexneri IcsA protein promotes actin nucleation by Arp2/3 complex and bacterial actin-based motility. J Cell Biol 146: 1319– 1332. Fukumatsu, M., Ogawa, M., Arakawa, S., Suzuki, M., Nakayama, K., Shimizu, S., et al. (2012) Shigella targets epithelial tricellular junctions and uses a noncanonical clathrin-dependent endocytic pathway to spread between cells. Cell Host Microbe 11: 325–336. Garcia-Saez, A.J., Buschhorn, S.B., Keller, H., Anderluh, G., Simons, K., and Schwille, P. (2011) Oligomerization and pore formation by equinatoxin II inhibit endocytosis and lead to plasma membrane reorganization. J Biol Chem 286: 37768–37777. Hachani, A., Biskri, L., Rossi, G., Marty, A., Menard, R., Sansonetti, P., et al. (2008) IpgB1 and IpgB2, two homolo© 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

Actin mechanics and Shigella gous effectors secreted via the Mxi-Spa type III secretion apparatus, cooperate to mediate polarized cell invasion and inflammatory potential of Shigella flexenri. Microbes Infect 10: 260–268. Handa, Y., Suzuki, M., Ohya, K., Iwai, H., Ishijima, N., Koleske, A.J., et al. (2007) Shigella IpgB1 promotes bacterial entry through the ELMO-Dock180 machinery. Nat Cell Biol 9: 121–128. Harris, T.J., and Tepass, U. (2012) Adherens junctions: from molecules to morphogenesis. Nat Rev Mol Cell Biol 11: 502–514. Heindl, J.E., Saran, I., Yi, C.R., Lesser, C.F., and Goldberg, M.B. (2010) Requirement for formin-induced actin polymerization during spread of Shigella flexneri. Infect Immun 78: 193–203. Humphreys, D., Davidson, A., Hume, P.J., and Koronakis, V. (2012) Salmonella virulence effector SopE and Host GEF ARNO cooperate to recruit and activate WAVE to trigger bacterial invasion. Cell Host Microbe 11: 129– 139. Izard, T., Tran Van Nhieu, G., and Bois, P.R. (2006) Shigella applies molecular mimicry to subvert vinculin and invade host cells. J Cell Biol 175: 465–475. Janakiraman, A., and Goldberg, M.B. (2004) Recent advances on the development of bacterial poles. Trends Microbiol 12: 518–525. Jasnin, M., Asano, S., Gouin, E., Hegerl, R., Plitzko, J.M., Villa, E., et al. (2013) Three-dimensional architecture of actin filaments in Listeria monocytogenes comet tails. Proc Natl Acad Sci USA 110: 20521–20526. Jaumouille, V., Francetic, O., Sansonetti, P.J., and Tran Van Nhieu, G. (2008) Cytoplasmic targeting of IpaC to the bacterial pole directs polar type III secretion in Shigella. EMBO J 27: 447–457. Khurana, S., and George, S.P. (2011) The role of actin bundling proteins in the assembly of filopodia in epithelial cells. Cell Adh Migr 5: 409–420. Kim, M., Ogawa, M., Fujita, Y., Yoshikawa, Y., Nagai, T., Koyama, T., et al. (2009) Bacteria hijack integrin-linked kinase to stabilize focal adhesions and block cell detachment. Nature 459: 578–582. Le Clainche, C., Dwivedi, S.P., Didry, D., and Carlier, M.F. (2010) Vinculin is a dually regulated actin filament barbed end-capping and side-binding protein. J Biol Chem 285: 23420–23432. Leung, Y., Ally, S., and Goldberg, M.B. (2008) Bacterial actin assembly requires toca-1 to relieve N-wasp autoinhibition. Cell Host Microbe 3: 39–47. Lum, M., and Morona, R. (2014) Myosin IIA is essential for Shigella flexneri cell-to-cell spread. Pathog Dis 72: 174–187. Mallo, G.V., Espina, M., Smith, A.C., Terebiznik, M.R., Alem, A., Finlay, B.B., et al. (2008) SopB promotes phosphatidylinositol 3-phosphate formation on Salmonella vacuoles by recruiting Rab5 and Vps34. J Cell Biol 182: 741–752. Marteyn, B., Gazi, A., and Sansonetti, P. (2012) Shigella: a model of virulence regulation in vivo. Gut Microbes 3: 104– 120. Mattei, P.J., Faudry, E., Job, V., Izore, T., Attree, I., and Dessen, A. (2011) Membrane targeting and pore formation © 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

181

by the type III secretion system translocon. FEBS J 278: 414–426. Mothes, W., Sherer, N.M., Jin, J., and Zhong, P. (2010) Virus cell-to-cell transmission. J Virol 84: 8360–8368. Mueller, C.A., Broz, P., and Cornelis, G.R. (2008) The type III secretion system tip complex and translocon. Mol Microbiol 68: 1085–1095. Mueller, J., Pfanzelter, J., Winkler, C., Narita, A., Le Clainche, C., Nemethova, M., et al. (2014) Electron tomography and simulation of baculovirus actin comet tails support a tethered filament model of pathogen propulsion. PLoS Biol 12: e1001765. Niebuhr, K., Jouihri, N., Allaoui, A., Gounon, P., Sansonetti, P.J., and Parsot, C. (2000) IpgD, a protein secreted by the type III secretion machinery of Shigella flexneri, is chaperoned by IpgE and implicated in entry focus formation. Mol Microbiol 38: 8–19. Niebuhr, K., Giuriato, S., Pedron, T., Philpott, D.J., Gaits, F., Sable, J., et al. (2002) Conversion of PtdIns(4,5)P(2) into PtdIns(5)P by the S. flexneri effector IpgD reorganizes host cell morphology. EMBO J 21: 5069–5078. Orchard, R.C., and Alto, N.M. (2012) Mimicking GEFs: a common theme for bacterial pathogens. Cell Microbiol 14: 10–18. Park, H., Valencia-Gallardo, C., Sharff, A., Tran Van Nhieu, G., and Izard, T. (2011) Novel vinculin binding site of the IpaA invasin of Shigella. J Biol Chem 286: 23214– 23221. Pendaries, C., Tronchere, H., Arbibe, L., Mounier, J., Gozani, O., Cantley, L., et al. (2006) PtdIns5P activates the host cell PI3-kinase/Akt pathway during Shigella flexneri infection. EMBO J 25: 1024–1034. Ramarao, N., Le Clainche, C., Izard, T., Bourdet-Sicard, R., Ageron, E., Sansonetti, P.J., et al. (2007) Capping of actin filaments by vinculin activated by the Shigella IpaA carboxyl-terminal domain. FEBS Lett 581: 853– 857. Rathman, M., de Lanerolle, P., Ohayon, H., Gounon, P., and Sansonetti, P. (2000) Myosin light chain kinase plays an essential role in S. flexneri dissemination. J Cell Sci 113 (Part 19): 3375–3386. del Rio, A., Perez-Jimenez, R., Liu, R., Roca-Cusachs, P., Fernandez, J.M., and Sheetz, M.P. (2009) Stretching single talin rod molecules activates vinculin binding. Science 323: 638–641. Romero, S., Grompone, G., Carayol, N., Mounier, J., Guadagnini, S., Prevost, M.C., et al. (2011) ATP-mediated Erk1/2 activation stimulates bacterial capture by filopodia, which precedes Shigella invasion of epithelial cells. Cell Host Microbe 9: 508–519. Romero, S., Quatela, A., Bornschlog, T., Guadagnini, S., Bassereau, P., and Tran Van Nhieu, G. (2012) Filopodium retraction is controlled by adhesion to its tip. J Cell Sci 125: 4999–5004. Sansonetti, P.J., Mounier, J., Prevost, M.C., and Mege, R.M. (1994) Cadherin expression is required for the spread of Shigella flexneri between epithelial cells. Cell 76: 829– 839. Schwan, C., Kruppke, A.S., Nolke, T., Schumacher, L., Koch-Nolte, F., Kudryashev, M., et al. (2013) Clostridium difficile toxin CDT hijacks microtubule organization and

182 C. M. Valencia-Gallardo, N. Carayol and G. Tran Van Nhieu reroutes vesicle traffic to increase pathogen adherence. Proc Natl Acad Sci USA 111: 2313–2318. Tran Van Nhieu, G., and Izard, T. (2007) Vinculin binding in its closed conformation by a helix addition mechanism. EMBO J 26: 4588–4596. Tran Van Nhieu, G., Kai Liu, B., Zhang, J., Pierre, F., Prigent, S., Sansonetti, P., et al. (2013) Actin-based confinement of calcium responses during Shigella invasion. Nat Commun 4: 1567.

Viaud, J., Lagarrigue, F., Ramel, D., Allart, S., Chicanne, G., Ceccato, L., et al. (2014) Phosphatidylinositol 5-phosphate regulates invasion through binding and activation of Tiam1. Nat Commun 5: 4080. Young, V.B., Falkow, S., and Schoolnik, G.K. (1992) The invasin protein of Yersinia enterocolitica: internalization of invasin-bearing bacteria by eukaryotic cells is associated with reorganization of the cytoskeleton. J Cell Biol 116: 197–207.

© 2014 John Wiley & Sons Ltd, Cellular Microbiology, 17, 174–182

Copyright of Cellular Microbiology is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

Cytoskeletal mechanics during Shigella invasion and dissemination in epithelial cells.

The actin cytoskeleton is key to the barrier function of epithelial cells, by permitting the establishment and maintenance of cell-cell junctions and ...
393KB Sizes 0 Downloads 5 Views