Nature Cell Biology homepage
Advanced search
 Journal home > Archive > Table of Contents > Review > Full text
Journal home
Advance online publication
Current issue
Archive
Press releases
Supplements and Focuses
Image gallery
Guide to authors
Online submissionOnline submission
Permissions
For referees
Free online issue
Contact the journal
Subscribe
Advertising
work@npg
naturereprints
About this site
For librarians
 
NPG Resources
Nature
Nature Reviews Molecular Cell Biology
UCSD-Nature Signaling Gateway
The Cell Migration Gateway
Nature Reports Stem Cells
Nature Reports Avian Flu
NPG Subject areas
Biotechnology
Cancer
Chemistry
Clinical Medicine
Dentistry
Development
Drug Discovery
Earth Sciences
Evolution & Ecology
Genetics
Immunology
Materials Science
Medical Research
Microbiology
Molecular Cell Biology
Neuroscience
Pharmacology
Physics
Browse all publications
Review
Cell Biology of Disease
November 2004 Contents Archive


Nature Cell Biology  6, 1026 - 1033 (2004)
doi:10.1038/ncb1104-1026

Subversion of phosphoinositide metabolism by intracellular bacterial pathogens

Javier Pizarro-Cerdá & Pascale Cossart

Javier Pizarro-Cerdá & Pascale Cossart are at the Unité des Interactions Bactéries-Cellules/Unité INSERM 604, Institut Pasteur, 75724 Paris Cedex 15, France. pizarroj@pasteur.fr pcossart@pasteur.fr

Phosphoinositides are short-lived lipids, whose production at specific membrane locations in the cell enables the tightly controlled recruitment or activation of diverse cellular effectors involved in processes such as cell motility or phagocytosis. Bacterial pathogens have evolved molecular mechanisms to subvert phosphoinositide metabolism in host cells, promoting (or blocking) their internalization into target tissues, and/or modifying the maturation fate of their proliferating compartments within the intracellular environment.
Microbial pathogens have developed different strategies to adapt and survive in their specific hosts. Many grow and replicate in the extracellular space in close association with eukaryotic cells, whereas intracellular microorganisms, after uptake in phagocytic or non-phagocytic cells, replicate in different organelles or intracellular compartments where they are protected from antibodies or complement1, 2. In all cases, the infection process requires a very precise subversion of the cellular machinery for the parasite to establish an appropriate niche for replication. Exploitation of the cellular cytoskeleton and associated proteins that modulate its plasticity has been reported for many bacterial pathogens1, 2. Lipids and lipid metabolism have only recently emerged as additional targets for bacterial virulence factors3. Cellular lipids, and in particular phosphoinositides, not only modulate the function of the actin cytoskeleton or its architecture, but also function as scaffolds recruiting specific effectors to membranes, and/or controlling maturation of intracellular compartments. Modification of phosphoinositides by kinases and phosphatases controls in a very precise way their temporal and spatial distribution (Fig. 1), making them ideal modulators of local and transient cellular mechanisms4, 5. In this review, we will address recent data concerning the subversion of phosphoinositide metabolism during bacterial infection.

Figure 1. Phophoinositide metabolism flowchart (modified from ref. 9).
Figure 1 thumbnail

Main steps in phosphoinositide synthesis and degradation, showing some of the enzymes that catalyse the reactions and the steps subverted by bacterial pathogens. Yersinia species induce activation of a PIP(5)K to promote PtdIns(4,5)P2 formation; Listeria monocytogenes and UPEC promote the activity of the class I PI(3)K to generate PtdIns(3,4,5)P3, whereas EPEC blocks its activity; Mycobacterium tuberculosis, blocks a class III PI(3)K involved in the formation of PtdIns(3)P; Shigella flexneri translocates into host cells the bacterial enzyme IpgD that dephosphorylates PtdIns(4,5)P2 into PtdIns(5)P; Salmonella enterica also translocates a bacterial effector, SigD/SopB, that has phosphatidylinositol polyphosphate phosphatase activity and it has been proposed to dephosphorylate the 5'-phosphate from PtdIns(3,5)P2 and PtdIns(3,4,5)P3; SigD/SopB is also implicated in the turnover of PtdIns(4,5)P2, but the precise PtdIns(4,5)P2 hydrolysis site remains to be determined. SigD/SopB additionally has inositol phosphate phosphatase activity (not shown in the diagram).



Full FigureFull Figure and legend (50K)
Phosphoinositides and phagocytosis
Actin rearrangements are required for phagocytosis6, and actin remodelling is linked to changes in phosphoinositide metabolism7, 8. In particular, phosphatidylinositol 4,5-biphosphate (PtdIns(4,5)P2) controls the function and distribution of several actin-binding proteins, through recognition of its headgroup by lysine- or arginine-rich hydrophobic motifs or pleckstrin homology (PH) domains9. Other phosphoinositides can be also recognized by PH domains of different specificities or by other domains such as PX or FYVE (Table 1). Their fusion to fluorescent proteins such as green fluorescent protein (GFP) has yielded instrumental probes for the analysis of the distribution of phosphoinositides during phagocytosis and phagosome maturation.

Table 1. Phosphoinositides and phosphoinositides binding domains
Table 1 thumbnail

Full TableFull Table
In macrophages, PtdIns(4,5)P2 accumulates in pseudopods during extension of the phagocytic cup, and this accumulation coincides with actin enrichment10, confirming the idea that PtdIns(4,5)P2 recruits effectors involved in regulating the actin cytoskeleton. One isoform of the type I phosphatidylinositol 4-phosphate 5-kinase (PIP(5)KIalpha) that synthesizes PtdIns(4,5)P2 from PtdIns(4)P is also recruited at sites of phagocytosis and is required for completion of the phagocytic process11, suggesting that PtdIns(4,5)P2 is directly produced at sites of particle internalization (Fig. 2). As the phagosome seals, PtdIns(4,5)P2 and actin disappear from the phagocytic cup. Catabolism of PtdIns(4,5)P2 could be explained, in part, by its hydrolysis to diacylglycerol (DAG) by phospholipase C-gamma1 (PLC-gamma1), which is activated during phagocytosis12. Another fraction of PtdIns(4,5)P2 is phosphorylated to phosphatidylinositol 3,4,5-triphosphate (PtdIns(3,4,5)P3), an event that class I PtdIns 3-kinase (PI(3)K) is responsible for; silencing of the alpha or beta isoforms of p85 — the regulatory subunit of the heterodimeric class I PI(3)K — reduces accumulation of PtdIns(3,4,5)P3 and blocks phagocytosis13. PtdIns(3,4,5)P3 accumulates at the phagocytic cup14 and seems to be required for its closure15 (Fig. 2). Several guanosine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs) of the Arf and Rho families of small GTP-binding proteins possess PH domains that interact with PtdIns(3,4,5)P3 and could be involved in completion of phagocytosis16. PtdIns(3,4,5)P3 disappears shortly after sealing of the phagosomal vacuole, and the 5' phosphatase SHIP or the 3' phosphatase PTEN are probably required for dephosphorylating PtdIns(3,4,5)P3 into PtdIns(3,4)P2 or PtdIns(4,5)P2, respectively17, 18.

Figure 2. Phosphoinositide species involved in classical phagocytosis.
Figure 2 thumbnail

PtdIns(4,5)P2 accumulates in pseudopods during extension of the phagocytic cup and probably recruits effectors involved in actin cytoskeleton regulation. As the phagosome seals, PtdIns(4,5)P2 disappears. This could be explained in part by its catabolism by phospholipases but also by its conversion into PtdIns(3,4,5)P3, and indeed, PtdIns(3,4,5)P3 appearance coincides with PtdIns(4,5)P2 clearance. PtdIns(3,4,5)P3 accumulates transiently in the phagocytic cup and is required for its closure. Once the phagosome is formed, PtdIns(3)P is produced on its surface and recruits proteins that control phagosome fusion and maturation. Other phosphoinositide species are present in the trans-Golgi complex (PtdIns(4)P) or in the nucleus (PtdIns(5)P), leading to the proposal that membrane identity can be mediated by compartmentalization of specific phosphoinositides79.



Full FigureFull Figure and legend (29K)
After closure of the phagocytic cup and through the acquisition of various components derived from endosomes and lysosomes, the phagosome matures in a compartment that has the capacity to kill microorganisms or to destroy apoptotic bodies. Maturation of the formed phagosome requires the recruitment/generation of a rare phosphoinositide, phosphatidylinositol 3-phosphate (PtdIns(3)P)19 (Fig. 2). Class III PI(3)K is involved in the production of PtdIns(3)P in phagosomal membranes, as its inhibition by specific inhibitory antibodies or by wortmannin causes the arrest of phagosomal maturation at an early stage13. Two main types of domains that specifically recognize PtdIns(3)P, the FYVE and PX domains, have been discovered in proteins that regulate endocytosis and endosomal trafficking9. One ligand of PtdIns(3)P is the early endosomal antigen 1 (EEA1), a FYVE-domain protein that also interacts with the small GTPase Rab5 and tethers early endosomes to endocytic vesicles, favouring their fusion20. Another ligand of PtdIns(3)P is the hepatocyte growth factor (HGF)-regulated tyrosine kinase substrate (Hrs), a protein that also possess a FYVE domain and is involved in the endocytic sorting of ubiquitinated membrane proteins and multivesicular body formation21. Subunits of the phagocytic NADPH oxidase complex present a PX domain that recognizes PtdIns(3)P, and probably contribute to phagosomal maturation22. Nexins, a diverse group of proteins involved in membrane trafficking and protein sorting, also interact with PtdIns(3)P through their PX domain23; some nexins also have a BAR domain that can sense membrane curvature, targeting them to microdomains defined by high curvature and PtdIns(3)P enrichment24. Although PtdIns(3)P disappears from the membrane of certain phagosomes, it can remain associated with other maturing phagosomes25; however, the functional relevance for this divergence remains unknown.

Bacterial invasion strategies
Bacterial pathogens use different strategies to colonize their specific hosts1, 2. Some bacteria remain extracellular (for example, enteropathogenic Escherichia coli or EPEC). Others are internalized by phagocytic cells and proliferate in vacuolar compartments (for example, Mycobacterium tuberculosis and Legionella pneumophila). The so-called 'invasive' pathogens have evolved mechanisms to induce their own uptake by cells that are normally non-phagocytic, and although some are only transiently present within a vacuole at one stage of their infectious process (for example, uropathogenic E. coli (UPEC), or Yersinia species), others establish residence in a membrane-bound vacuole in which they replicate (for example, Salmonella enterica). Finally, a few invasive bacteria escape from the internalization vacuole and multiply in the cytosol (for example, Listeria monocytogenes, Shigella flexneri and Mycobacterium marinum). The study of phosphoinositide metabolism during baterial infection reveals that pathogens control these lipids in different ways, exploiting and fostering their normal metabolism, or modifying their cellular signalling cascades to establish their replicative niches (Fig. 3).

Figure 3. Signalling pathways exploited by bacterial pathogens.
Figure 3 thumbnail

(a) Internalization of Yersinia species. Interaction of invasin with beta1-integrins induces Rac1 recruitment and PIP(5)K(1)alpha translocation to the bacterial entry site; Arf6 activates PIP(5)K(1)alpha leading to local PtdIns(4,5)P2 generation. (b) Cellular invasion by L. monocytogenes. (1) InlB binds to Met, inducing Gab/ Cbl/Shc and class I PI(3)K recruitment; Rac1 is then activated, controlling actin dynamics during entry; Akt is also activated, probably leading to cell survival after infection; PLC-gamma1 activation induces Ins(1,4,5)P3 increase and Ca2+ release from intracellular stores, but the role of this signalling cascade is unknown. (2) Intracellularly, L. monocytogenes resides transiently in an EEA1- and Rab5-enriched vacuole, favouring vacuolar fusion with early endosomes to delay phagosomal maturation. (c) UPEC entry into bladder cells. The FimH receptor is uroplakin Ia, which has a short cytoplasmic domain and interacts with unknown molecules to trigger entry; complexes required for invasion include Rac1/caveolin1 and class I PI(3)K/FAK, the latter probably modulating vinculin and alpha-actinin association to stabilize actin. (d) Internalization of S. flexneri. The TTSS effector IpgD dephosphorylates PtdIns(4,5)P2 into PtdIns(5)P leading to membrane detachment from the underlying cytoskeleton, facilitating Rac1 and Cdc42 ruffling activity; PtdIns(3,5)P2 and PtdIns(3,4,5)P3 also accumulate at the bacterial entry site, but their possible role in entry is unknown. (e) S. enterica invasion of target cells. (1) the Rac1/Cdc42 exchange factor SopE and the phosphatidylinositol phosphatase SigD/SopB promote Ins(1,4,5,6)P4 generation and indirect Cdc42 activation; SigD/SopB can also generate PtdIns(3,4)P2, activating Akt. (2) After invasion, S. enterica-containing vacuoles (SCV) are enriched in PtdIns(3)P, favouring SCV fusion with PtdIns(3)P−enriched empty macropinosomes, probably taking advantage of the EEA1 and Rab5 fusogenic machinery. (f) Phagocytosis inhibition by EPEC. Bacteria multiply extracellularly on top of pedestals; translocation to the host cell cytoplasm of an unknown TTSS effector leads to inactivation of class I PI(3)K, PtdIns(4,5)P2 accumulation and blockage of bacterial internalization. (g) Internalization of M. tuberculosis. The bacterial phosphatidylinositol analogue LAM inhibits intracellular Ca2+ rise, blocking a calmodulin−calmodulin kinase II pathway that activates class III PI(3)K; simultaneously, the mycobacterial phosphoinositide PIM activates a GTPase (probably Rab5) inducing fusion of the M. tuberculosis-containing compartment with early endosomes.



Full FigureFull Figure and legend (136K)
Phosphoinositide signalling through receptor engagement
Enteropathogenic Yersinia species translocate across the intestinal epithelium in mammalian hosts1, multiply in local lymph nodes or deep organs and induce systemic diseases26. Interaction between the bacterial outer membrane protein invasin and heterodimeric beta1 integrin receptors results in uptake27, a process requiring activation of the small GTP-binding protein Rac128. The Rac1 pathway involves recruitment of PIP(5)KIalpha and transient local production of PtdIns(4,5)P2 at sites of bacterial entry29. Indeed, transfection of Rac1 increases the level of PIP(5)KIalpha around internalized bacteria. Reduction of PtdIns(4,5)P2 produced by PIP(5)KIalpha using a membrane-targeted PtdIns(4,5)P2 phosphatase lowers bacterial uptake, highlighting the key role of this phosphoinositide in infection. PtdIns(4,5)P2 production is normally regulated by the GTP-binding protein Arf6 (ref. 30), and expression of a defective Arf6 variant in target cells inhibits PtdIns(4,5)P2 production and bacterial phagocytosis, but not PIP(5)KIalpha recruitment29. This observation reveals that wild-type Arf6 is involved in PIP(5)KIalpha activation to induce PtdIns(4,5)P2 generation. Together, these results suggest that engagement of beta1 integrins by invasin induces the activation of Rac1, which in turn favours the recruitment of PIP(5)KIalpha at the bacterial entry site; Arf6 then activates PIP(5)KIalpha, leading to PtdIns(4,5)P2 formation and translocation to the plasma membrane of PtdIns(4,5)P2-binding proteins involved in control of actin dynamics required for bacterial invasion. Interestingly, Arf6 exchange factors bind to PtdIns(4,5)P2 (ref. 31), suggesting that a positive feedback loop could be involved in the activation of Arf6 at sites of high PtdIns(4,5)P2 concentration. Arf6 has been implicated in the delivery of membranes at sites of phagocytosis32 and this function could also be involved in the formation of the phagocytic cup in which Yersinia is internalized. Alternatively, PtdIns(4,5)P2 could also be recruited to generate PtdIns(3,4,5)P3, a phosphoinositide species able to recruit other cellular effectors that could regulate actin dynamics during the invasion processs (Fig. 3a).

L. monocytogenes is a bacterial pathogen responsible for severe foodborne infections in humans leading to gastroenteritis, meningitis and abortions33. L. monocytogenes induces its internalization in non-phagocytic cells through the interaction of two bacterial invasion proteins, InlA and InlB, with two cellular receptors, the adhesion molecule E-cadherin (which interacts with InlA34) and the hepatocyte growth factor receptor Met (which interacts with InlB35). It is almost ten years since the observation that L. monocytogenes entry into target cells requires PI(3)K activity, as bacterial invasion is blocked by the PI(3)K inhibitors wortmannin and LY294002 (ref. 36). In Vero cells, which lack E-cadherin and in which entry is dependent only on the InlB pathway, the use of a dominant- negative form of the p85alpha regulatory subunit blocked infection, revealing a role for the class I PI(3)K in this process. L. monocytogenes takes advantage of the cellular signalling pathway associated with the tyrosine kinase receptor Met to recruit adaptors such as Gab1, Shc and Cbl, which translocate the class I PI(3)K from the cytosol to the plasma membrane to induce cytoskeletal rearrangements37. Interestingly, the purified InlB protein is capable of inducing membrane ruffles in target cells as HGF, the natural ligand of Met37, 38. The activation kinetics of Met by HGF and InlB are different35, but the signalling cascades triggered by both molecules seem similar39, 40 (Fig. 3b). The direct downstream effectors that could be recruited by PtdIns(3,4,5)P3 are, however, unknown. Invasion of Vero cells by L. monocytogenes requires the polymerization and subsequent depolymerization of actin from the phagocytic cup, a process that is dependent on the actin-depolymerizing factor cofilin, a downstream effector of Rac40. The Rac exchange factor Vav2, a PH-domain protein that binds to PtdIns(3,4,5)P3, was considered as a potential downstream PI(3)K effector; however, dominant-negative Vav2 does not block L. monocytognes entry (H. Bierne, P. Mandin and P.C., unpublished observations).

In the Caco-2 cell line, which is dependent on the InlA-invasion pathway, wortmannin also impairs bacterial entry36, but the mechanism of class I PI(3)K recruitment by E-cadherin is unknown. It has recently been found that E-cadherin localizes in the plasma membrane within lipid rafts, which are lipidic platforms that select molecular effectors into functional units required for efficient signalling and sorting processes41. L. monocytogenes could take advantage of this platform to recruit signalling molecules involved in translocation of the PI(3)K into the plasma membrane.

In the human epithelial cell line Hep-2, InlB activates a different isoform of class I PI(3)K (p85beta-p110) and subsequently activates PLC-gamma1, leading to a transient increase in intracellular inositol 1,4,5-triphosphate (Ins(1,4,5)P3) and release of intracellular calcium42. However, pharmacological or genetic inactivation of PLC-gamma1 does not affect the InlB-mediated bacterial uptake, indicating that PLC-gamma1 activation could be a side effect during the entry process; alternatively, PLC-gamma1 activation could function in post-internalization steps, but this possibility has not been explored. After invasion, L. monocytogenes resides transiently in a vacuole that is enriched in EEA1, suggesting the presence of PtdIns(3)P in this compartment (Fig. 3b). Interestingly, L. monocytogenes-containing phagosomes are also enriched in the small GTPase Rab5, which increases the fusion of early endosomal compartments with the bacterial vacuole, delaying its maturation43. As a consequence, L. monocytogenes resides as long as possible in a safe environment before lysing its phagosome and escaping into the cellular cytoplasm where the bacteria multiply. Regulation of PtdIns(3)P levels by L. monocytogenes is an attractive hypothesis that could explain the retention of EEA1 and Rab5 in the bacterial phagosome. Activation of the anti-apoptotic Akt kinase downstream of the class I PI(3)K has been detected in the murine cell line J774 (ref. 44), suggesting that L. monocytogenes could also modulate cell survival after infection (Fig. 3b).

More than 80% of urinary tract infections are produced by uropathogenic E. coli (UPEC). Although UPEC was long considered non invasive, recent results indicate that the type 1 pilus adhesin FimH mediates not only bacterial adherence but also invasion of human bladder epithelial cells45. The only known cellular receptor for FimH is uroplakin Ia46, a molecule that localizes to lipid rafts; caveolin-1 and Rac1 interact with each other within lipid rafts and are also required for UPEC entry47. Wortmannin inhibits invasion, implying that PI(3)K is required for uptake. Phosphorylation of the focal adhesion kinase (FAK) and formation of FAK−PI(3)K complexes are observed only with wild-type UPEC, and not with a FimH- mutant45. In the current scenario, during bladder cell invasion, UPEC probably take advantage of the physiological endocytosis of uroplakin Ia, which interacts with caveolin-1, Rac1 and possibly other uroplakins within the context of a lipid raft to trigger internalization signalling cascades requiring PI(3)K activity; PI(3)K and FAK could potentially modulate, directly or indirectly, a number of other molecules including vinculin, tensin, talin or alpha-actinin, which can function in stabilizing the host-cell cytoskeleton. Indeed, the infection of bladder cells with wild-type UPEC induces the formation of vinculin−alpha-actinin complexes, and the formation of these complexes is inhibited in the presence of PI(3)K inhibitors45 (Fig. 3c).

Translocation of bacterial phosphoinositide phosphatases
S. flexneri is the causative agent of bacillary dysentery in humans48. This pathogen induces its entry into non-phagocytic cells by injecting bacterial effectors directly into the host cell through a type III secretion system (TTSS). The TTSS is a needle-like structure that spans the inner and outer membranes of the bacterial envelope and after contact with mammalian cells forms a pore in the host-cell membrane; through this channel bacterial effector proteins are translocated into the host-cell cytoplasm where they subvert cellular functions49.

Some of the S. flexneri TTSS effectors, such as IpaC, directly interact with cytoskeletal proteins1, 50, 51. S. flexneri also modifies the phosphatidylinositol metabolism of host cells to favour its internalization through the activity of another TTSS effector, the virulence factor IpgD. During infection of non-phagocytic HeLa cells, S. flexneri translocates IpgD into the cytoplasm of target cells, which behaves as a phosphatidylinositol 4-phosphatase that specifically dephosphorylates PtdIns(4,5)P2 into PtdIns(5)P, leading to the accumulation of PtdIns(5)P in the eukaryotic cell52. PtdIns(5)P is a phosphoinositide species that was thought not to exist in vivo53. Whether PtdIns(5)P has a role in regulating specific cytoskeletal rearrangements is still unknown. However, dephosphorylation of PtdIns(4,5)P2 has an important role during invasion. PtdIns(4,5)P2 regulates membrane adhesion by controlling local binding interactions between the plasma membrane and the cytoskeleton54; accordingly, in cells transfected with IpgD, the force that tethers the plasma membrane and the cytoskeleton was reduced52. Moreover, IpgD transfection led to the formation of membrane blebbing and the disappearance of actin stress fibres. IpgD probably allows the local detachment of the plasma membrane from the cytoskeleton to facilitate extension of membrane filopodia and ruffles by changing the interaction between PtdIns(4,5)P2 and cytoskeletal anchoring proteins and/or by regulating actin dynamics through enzymes such as gelsolin or cofilin that increase or decrease actin polymerization (Fig. 3d). Cdc42 and Rac favour the extension of membrane filopodia and ruffles during S. flexneri invasion, and IpgD seems to facilitate the cytoskeletal modifications induced by these small Rho family GTPases52. The specificity of the IpgD inositol 4-phosphatase activity is revealed by the fact that, during infection, only PtdIns(4,5)P2 levels decrease whereas other phosphoinositide species such as PtdIns(3,4)P2, PtdIns(3,5)P2 and PtdIns(3,4,5)P3 increase, due to concomitant activation of PI(3)K. PI(3)K activity is linked to infection, as PtdIns(3,4)P2, PtdIns(3,5)P2 and PtdIns(3,4,5)P3 concentrations do not increase with non-invasive mutants52; wortmannin does not block entry, suggesting either that PI(3)K kinase activity is not necessary for invasion, or that wortmannin-insensitive class II enzymes or a bacterially derived kinase are involved in this process.

S. enterica is a main cause of food-borne gastroenteritis, as well as the etiological agent of typhoid fever55, inducing its entry into non-phagocytic cells and its survival in macrophages, through the action of two different TTSS1. The S. enterica TTSS effector SigD/SopB is an IpgD homologue that functions in the invasion of target cells as well as in the maturation of S. enterica-containing vacuoles (SCVs). It presents phosphatidylinositol polyphosphate and inositol polyphosphate phosphatase activities; however, there are conflicting reports about its enzymatic specificity56, 57, 58 (Fig. 1).

SigD/SopB stimulates the formation of membrane ruffles and Cdc42- dependent actin cytoskeletal rearrangements required for S. enterica entry in Henle-407 cells57. This function overlaps with the activity of SopE and SopE2, two other TTSS effectors that function as exchange factors for Cdc42 and Rac. SigD/SopB dephosphorylates inositol- 1,3,4,5,6-phosphate (Ins(1,3,4,5,6)P5) into inositol-1,4,5,6-phosphate (Ins(1,4,5,6)P4) in infected cells as well as in in vitro assays57. SopE does not have inherent inositol phosphate phosphatase activity but it does contribute to the hydrolysis of Ins(1,3,4,5,6)P5 and inositol-1,2,3,4,5,6- phosphate (Ins(1,2,3,4,5,6)P6) probably by activation of a PLC as well as of an endogenous phosphatase. The molecular pathway that leads to the indirect activation of Cdc42 by inositol phosphates is unknown; these molecules probably compete with phosphoinositides for binding to specific PH domains, displacing the subcellular localization of PH-domain-containing molecules required for actin cytoskeleton rearrangements during invasion57 (Fig. 3e).

As stated above, SigD/SopB also has phosphatidylinositol phosphatase activity, and has been implicated in PtdIns(4,5)P2 turnover during S. enterica invasion58 (the precise site of PtdIns(4,5)P2 hydrolysis remains to be determined). SigD/SopB is responsible for the rapid disappearance of PtdIns(4,5)P2 from the plasma membrane before fission and formation of the SCV58. Thus, it has been proposed that SigD/SopB weakens membrane−cytoskeleton interactions, probably by displacing molecules that recognize PtdIns(4,5)P2, promoting the separation and sealing of the SCV after invasion.

Phagosomal maturation is also controlled by the phosphatidylinositol phosphatase activity of SigD/SopB. In MDCK cells, PtdIns(3)P is produced periodically on the membranes of nascent SCVs in a wortmannin-insensitive manner, and it was suspected that SigD/SopB participated in this process by dephosphorylating PtdIns(3,5)P2 to PtdIns(3)P59. In HeLa cells, PtdIns(3)P also transiently accumulates on SCVs, but this phenomenon is wortmannin-sensitive, implying that a PI(3)K is involved in maturation of the SCV60. More recently, it has been observed that PtdIns(3)P cycles on latex-bead phagosomal membranes are dependent on the host cell and are cell type-specific61. Nevertheless, it has been confirmed in Henle-407 cells that SigD/SopB contributes to SCV maturation, inducing the formation of large PtdIns(3)-enriched macropinosomes during S. enterica invasion62. The appearance of PtdIns(3)P-enriched empty vesicles that fuse frequently with the SCV suggests that S. enterica induces the homotypic fusion of PtdIns(3)P compartments; the PtdIns(3)P-binding effectors EEA1 and Rab5 have been localized to the early SCV63 indicating that S. enterica could boost the early endosomal fusion machinery to divert SCV maturation from the normal phagolysosomal pathway (Fig. 3e).

Finally, SigD/SopB is also required for Akt activation in HeLa cells64. Cell infection with wild-type S. enterica induces the translocation of Akt to the plasma membrane and its phosphorylation on residues Thr 308 and Ser 473. Infection with a SigD/SopB deletion mutant does not induce Akt phosphorylation, although it is translocated to the plasma membrane64. Complementation with a SigD/SopB mutant that lacks inositol phosphatase activity does not restore Akt activation, highlighting the relevance of SigD/SopB enzymatic activity during Akt activation. It has been proposed that PtdIns(3,4)P2, produced from PtdIns(3,4,5)P3 by SigD/SopB (Fig. 1) could induce a conformational change in Akt favouring its phosphorylation and full activation leading to increased survival of infected epithelial cells65.

Bacterial interference with phosphoinositide kinases
Enteropathogenic E. coli (EPEC), a major cause of severe infantile diarrhoea, colonizes the human small intestine mucosa and exerts its pathological effects by inducing the degeneration of the enterocyte brush border66. Through a TTSS, adhering EPEC directly translocate bacterial effectors into the host-cell cytoplasm. These effectors induce the localized loss of microvilli and the formation of actin-rich protrusions called 'pedestals', on which bacteria reside67. EPEC multiply in the extracellular space and pedestal formation has been considered as a mechanism to block bacterial phagocytosis. Along this line, it has been shown that EPEC inhibit their internalization into host cells through the inhibition of PI(3)K-dependent pathways68. In macrophages, an EPEC TTSS mutant can be efficiently phagocytosed whereas a wild-type EPEC cannot. Moreover, chemical inhibition of PI(3)K by wortmannin, or functional inhibition of class I PI(3)K with a dominant-negative p85alpha subunit, blocked the internalization of the EPEC mutant. Wild-type EPEC do not inhibit recruitment but do inhibit the activity of PI(3)K, causing the dissociation of phosphotyrosine proteins associated with PI(3)K and inducing interruption of phagocytic signals (Fig. 3f). The accumulation of PtdIns(4,5)P2 at sites of bacterial adhesion in macrophages suggests that wild-type EPEC also mediate the dephosphorylation of PtdIns(3,4,5)P3 into PtdIns(4,5)P2 by an as yet uncharacterized phosphatase.

M. tuberculosis is the etiological agent of tuberculosis, a severe pulmonary disease characterized by the infection and chronic survival of the tubercle bacillus in host macrophages69. M. tuberculosis resides in a vacuole that remains in an early phagosomal stage characterized by fewer or less active vacuolar H+-ATPases70 and absence of mature lysosomal hydrolases71. As shown recently, the bacterial phosphatidylinositol analogue lipoarabinomannan (LAM) inhibits cytosolic calcium rise in infected cells, blocking a calmodulin−calmodulin kinase II cascade that is required for activation of class III PI(3)K, PtdIns(3)P generation and EEA1 recruitment to phagosomes72. EEA1, in cooperation with the trans-Golgi network (TGN) molecule syntaxin 6, is needed for the delivery of hydrolases from the TGN to the phagosome, and M. tuberculosis interferes with this pathway73. It has been recently reported that the PtdIns(3)P ligand Hrs is not present in M. tuberculosis phagosomes74; its absence is possibly related to the maturation arrest of this compartment. In parallel, another secreted mycobacterial phosphoinositide (phosphatidylinositol mannoside, PIM) stimulates early endosomal fusion (probably influencing Rab5 activity), functioning in concert with LAM to block phagosomal maturation75 (Fig. 3g).

It was recently appreciated that, during phagocytosis of latex beads, there are two phases of PtdIns(3)P acquisition: phase I consists of a strong burst of PtdIns(3)P after closure of the phagocytic cup and is calmodulin−calmodulin kinase II-independent, whereas phase II consists of multiple waves of lower PtdIns(3)P peaks and is calmodulin−calmodulin kinase II-dependent61. These oscillations are proposed to dictate orderly interactions with other compartments involved in phagosomal maturation61. During invasion of target cells by M. tuberculosis, phase I peak (dependent on the presence of cholesterol at the bacterial entry site) is shifted to earlier time points when bacteria are still entering the host cell and the phagosomal cup is not closed, whereas phase II peaks are not observed in mycobacterial phagosomes61. The absence of phase II peaks is readily explained by the inhibition of class III PI(3)K recruitment by mycobacterial LAM72. The shift in peak I could be explained by an early activation of Rab5 by PIM75, rendering the M. tuberculosis phagosome unique from the early stages of its formation, favouring its divergence from the normal phagolysosomal maturation cascade.

The idea that phosphoinositides could be used as targets for control of infectious diseases is illustrated in an interesting report76 that suggests that treating cells infected with M. tuberculosis with different lipids derived from phosphoinositide metabolism (such as arachidonic acid, ceramide or sphingosine) could alter the maturation fate of M. tuberculosis-containing phagosomes increasing their fusion with lysosomes, lowering their pH and resulting in a marked decrease in pathogen viability. These results may lead to the potential dietary use of phosphoinositide derivatives to control pathogen replication.

Conclusions
Phosphoinositides have an essential function in the coordination of actin cytoskeleton rearrangements and in phagosomal maturation steps during classical phagocytosis. Bacterial pathogens have taken advantage of the scaffolding role of phosphoinositides to recruit at the bacterial entry site effectors required for promoting (or inhibiting) invasion, or to modify the maturation of their intracellular compartments. Many of the specific effectors controlled by each bacterial pathogen through the modulation of phosphoinositide metabolism remain uncharacterized. We believe that this field will expand rapidly and that the study of bacterial intracellular parasitism will probably shed light onto the mechanisms used by the cell to proceed during classical phagocytosis.

 Top
REFERENCES
  1. Cossart, P. & Sansonetti, P. J. Bacterial invasion: the paradigms of enteroinvasive pathogens. Science 304, 242−248 (2004). | Article | PubMed  | ISI | ChemPort |
  2. Meresse, S. et al. Controlling the maturation of pathogen-containing vacuoles: a matter of life and death. Nature Cell Biol. 1, 183−188 (1999). | Article | PubMed  | ISI | ChemPort |
  3. Brumell, J. H. & Grinstein, S. Role of lipid-mediated signal transduction in bacterial internalization. Cell. Microbiol. 5, 287−297 (2003). | PubMed  | ISI | ChemPort |
  4. Botelho, R. J., Scott, C. C. & Grinstein, S. Phosphoinositide involvement in phagocytosis and phagosome maturation. Curr. Top. Microbiol. Immunol. 282, 1−30 (2004). | PubMed  | ISI | ChemPort |
  5. Payrastre, B. et al. Phosphoinositides: key players in cell signalling, in time and space. Cell. Signal. 13, 377−387 (2001). | Article | PubMed  | ISI | ChemPort |
  6. Greenberg, S., el Khoury, J., di Virgilio, F., Kaplan, E. M. & Silverstein, S. C. Ca2+-independent F-actin assembly and disassembly during Fc receptor-mediated phagocytosis in mouse macrophages. J. Cell Biol. 113, 757−767 (1991). | Article | PubMed  | ISI | ChemPort |
  7. Yin, H. L. & Janmey, P. A. Phosphoinositide regulation of the actin cytoskeleton. Annu. Rev. Physiol. 65, 761−789 (2003). | Article | PubMed  | ISI | ChemPort |
  8. Defacque, H. et al. Phosphoinositides regulate membrane-dependent actin assembly by latex bead phagosomes. Mol. Biol. Cell 13, 1190−1202 (2002). | Article | PubMed  | ISI | ChemPort |
  9. Lemmon, M. A. Phosphoinositide recognition domains. Traffic 4, 201−213 (2003). | PubMed  | ISI | ChemPort |
  10. Botelho, R. J. et al. Localized biphasic changes in phosphatidylinositol-4,5-bisphosphate at sites of phagocytosis. J. Cell Biol. 151, 1353−1368 (2000). | Article | PubMed  | ISI | ChemPort |
  11. Coppolino, M. G. et al. Inhibition of phosphatidylinositol-4-phosphate 5-kinase Ialpha impairs localized actin remodeling and suppresses phagocytosis. J. Biol. Chem. 277, 43849−43857 (2002). | Article | PubMed  | ISI | ChemPort |
  12. Cox, D. & Greenberg, S. Phagocytic signaling strategies: Fcã-receptor-mediated phagocytosis as a model system. Semin. Immunol. 13, 339−345 (2001). | Article | PubMed  | ISI | ChemPort |
  13. Vieira, O. V. et al. Distinct roles of class I and class III phosphatidylinositol 3-kinases in phagosome formation and maturation. J. Cell Biol. 155, 19−25 (2001). | Article | PubMed  | ISI | ChemPort |
  14. Marshall, J. G. et al. Restricted accumulation of phosphatidylinositol 3-kinase products in a plasmalemal subdomain during Fc gamma receptor-mediated phagocytosis. J. Cell Biol. 153, 1369−1380 (2001). | Article | PubMed  | ISI | ChemPort |
  15. Cox, D., Tseng, C. C., Bjekic, G. & Greenberg, S. A requirement for phosphatidylinositol 3-kinase in pseudopod extension. J. Biol. Chem. 274, 1240−1247 (1999). | Article | PubMed  | ISI | ChemPort |
  16. Hilpela, P., Vartiainen, M. K. & Lappalainen, P. Regulation of the actin cytoskeleton by PI(4,5)P2 and PI(3,4,5)P3. Curr. Top. Microbiol. Immunol. 282, 117−163 (2004). | PubMed  | ISI | ChemPort |
  17. Cox, D., Dale, B. M., Kashiwada, M., Helgason, C. D. & Greenberg, S. A regulatory role for Src homology 2 domain-containing inositol 5'-phosphatase (SHIP) in phagocytosis mediated by Fc gamma receptors and complement receptor 3 (alphaMbeta2; CD11b/CD18). J. Exp. Med. 193, 61−71 (2001). | Article | PubMed  | ISI | ChemPort |
  18. Cao, X. et al. The inositol 3-phosphatase PTEN negatively regulates Fc gamma receptor signaling, but supports Toll-like receptor 4 signaling in murine peritoneal macrophages. J. Immunol. 172, 4851−4857 (2004). | PubMed  | ISI | ChemPort |
  19. Ellson, C. D. et al. Phosphatidylinositol 3-phosphate is generated in phagosomal membranes. Curr. Biol. 11, 1631−1635 (2001). | Article | PubMed  | ISI | ChemPort |
  20. Simonsen, A. et al. EEA1 links PI(3)K function to Rab5 regulation of endosome fusion. Nature 394, 494−498 (1998). | Article | PubMed  | ISI | ChemPort |
  21. Raiborg, C. et al. Hrs sorts ubiquitinated proteins into clathrin-coated microdomains of early endosomes. Nature Cell Biol. 4, 394−398 (2002). | Article | PubMed  | ISI | ChemPort |
  22. Zhan, Y., Virbasius, J. V., Song, X., Pomerleau, D. P. & Zhou, GW. The p40phox and p47phox PX domains of NADPH oxidase target cell membranes via direct and indirect recruitment by phosphoinositides. J. Biol. Chem. 8, 4512−4518 (2002). | Article | ChemPort |
  23. Worby, C. A. & Dixon, J. E. Sorting out the cellular functions of sorting nexins. Nature Rev. Mol. Cell Biol. 3, 919−931 (2002). | Article | PubMed  | ISI | ChemPort |
  24. Cozier, G. E. et al. The phox homology (PX) domain-dependent, 3-phosphoinositidemediated association of sorting nexin-1 with an early sorting endosomal compartment is required for its ability to regulate epidermal growth factor receptor degradation. J. Biol. Chem. 277, 48730−48736 (2002). | Article | PubMed  | ISI | ChemPort |
  25. Henry, R. M., Hoppe, A. D., Joshi, N. & Swanson, J. A. The uniformity of phagosome maturation in macrophages. J. Cell Biol. 164, 185−189 (2004). | Article | PubMed  | ISI | ChemPort |
  26. Putzker, M., Sauer, H. & Sobe, D. Plague and other human infections caused by Yersinia species. Clin. Lab. 47, 453−466 (2001). | PubMed  | ChemPort |
  27. Isberg, R. R. & Leong, J. M. Multiple beta1 chain integrins are receptors for invasin, a protein that promotes bacterial penetration into mammalian cells. Cell 60, 861−871 (1990). | Article | PubMed  | ISI | ChemPort |
  28. Alrutz, M. A. et al. Efficient uptake of Yersinia pseudotuberculosis via integrin receptors involves a Rac1−Arp 2/3 pathway that bypasses N-WASP function. Mol. Microbiol. 42, 689−703 (2001). | Article | PubMed  | ISI | ChemPort |
  29. Wong, K. W. & Isberg, R. R. Arf6 and phosphoinositol-4-phosphate-5-kinase activities permit bypass of the Rac1 requirement for beta1 integrin-mediated bacterial uptake. J. Exp. Med. 198, 603−614 (2003). | Article | PubMed  | ISI | ChemPort |
  30. Brown, F. D., Rozelle, A. L., Yin, H. L., Balla, T. & Donaldson, J. G. Phosphatidylinositol 4,5-bisphosphate and Arf6-regulated membrane traffic. J. Cell Biol. 154, 1007−1017 (2001). | Article | PubMed  | ISI | ChemPort |
  31. Derrien, V. et al. A conserved C-terminal domain of EFA6-family ARF6-guanine nucleotide exchange factors induces lengthening of microvilli-like membrane protrusions. J. Cell Sci. 115, 2867−2879 (2002). | PubMed  | ISI | ChemPort |
  32. Niedergang, F., Colucci-Guyon, E., Dubois, T., Raposo, G. & Chavrier, P. ADP ribosylation factor 6 is activated and controls membrane delivery during phagocytosis in macrophages. J. Cell Biol. 161, 1143−1150 (2003). | Article | PubMed  | ISI | ChemPort |
  33. Dussurget, O., Pizarro-Cerdá, J. & Cossart, P. Molecular determinants of Listeria monocytogenes virulence. Annu. Rev. Microbiol. 58, 587−610 (2004). | Article | PubMed  |
  34. Mengaud, J., Ohayon, H., Gounon, P., Mege, R. -M. & Cossart, P. E-cadherin is the receptor for internalin, a surface protein required for entry of L. monocytogenes into epithelial cells. Cell 84, 923−932 (1996). | Article | PubMed  | ISI | ChemPort |
  35. Shen, Y., Naujokas, M., Park, M. & Ireton, K. InIB-dependent internalization of Listeria is mediated by the Met receptor tyrosine kinase. Cell 103, 501−510 (2000). | Article | PubMed  | ISI | ChemPort |
  36. Ireton, K. et al. A role for phosphoinositide 3-kinase in bacterial invasion. Science 274, 780−782 (1996). | Article | PubMed  | ISI | ChemPort |
  37. Ireton, K., Payrastre, B. & Cossart, P. The Listeria monocytogenes protein InlB is an agonist of mammalian phosphoinositide 3-kinase. J. Biol. Chem. 274, 17025−17032 (1999). | Article | PubMed  | ISI | ChemPort |
  38. Maulik, G. et al. Activated c-Met signals through PI3K with dramatic effects on cytoskeletal functions in small cell lung cancer. J. Cell. Mol. Med. 6, 539−553 (2002). | PubMed  | ISI | ChemPort |
  39. Banerjee, M. et al. GW domains of the Listeria monocytogenes invasion protein InlB are required for potentiation of Met activation. Mol. Microbiol. 52, 257−2571 (2004). | PubMed  | ISI | ChemPort |
  40. Bierne, H. et al. A role for cofilin and LIM kinase in Listeria-induced phagocytosis. J. Cell Biol. 155, 101−112 (2001). | Article | PubMed  | ISI | ChemPort |
  41. Seveau, S., Bierne, H., Giroux, S., Prevost, M. C. & Cossart, P. Role of lipid rafts in E cadherin- and HGF-R/Met-mediated entry of Listeria monocytogenes into host cells. J. Cell Biol. 166, 743−753 (2004). | Article | PubMed  | ISI |