IcmR-regulated Membrane Insertion and Efflux by the Legionella pneumophila IcmQ Protein

Abstract

Legionella pneumophila proliferates within alveolar macrophages as a central property of Legionnaires' disease. Intracellular growth involves formation of a replicative phagosome, which requires the bacterial Dot/Icm system, a multiprotein secretion apparatus that translocates proteins from the bacterium across the macrophage plasma membrane. Two components of this system, IcmR and IcmQ, are proposed to exhibit a chaperone/substrate relationship similar to that observed in other protein translocation systems. We report here that IcmQ inserts into lipid membranes and forms pores that allow the efflux of the dye calcein but not Dextran 3000. Both membrane insertion and pore formation were inhibited by IcmR. Trypsin digestion mapping demonstrated that IcmQ is subdivided into two functional domains. The N-terminal region of IcmQ was necessary and sufficient for insertion into lipid membranes and calcein efflux. The C-terminal domain was necessary for efficient association of the protein with lipid bilayers. IcmR was found to bind to the N-terminal portion of the protein thus providing a mechanism for its ability to inhibit IcmQ pore-forming activity. Localization of IcmQ on the surface of the L. pneumophila shortly after infection as well as its pore-forming capacities suggest a role for IcmQ in forming a channel that leads translocated effectors out of the bacterium. Legionella pneumophila proliferates within alveolar macrophages as a central property of Legionnaires' disease. Intracellular growth involves formation of a replicative phagosome, which requires the bacterial Dot/Icm system, a multiprotein secretion apparatus that translocates proteins from the bacterium across the macrophage plasma membrane. Two components of this system, IcmR and IcmQ, are proposed to exhibit a chaperone/substrate relationship similar to that observed in other protein translocation systems. We report here that IcmQ inserts into lipid membranes and forms pores that allow the efflux of the dye calcein but not Dextran 3000. Both membrane insertion and pore formation were inhibited by IcmR. Trypsin digestion mapping demonstrated that IcmQ is subdivided into two functional domains. The N-terminal region of IcmQ was necessary and sufficient for insertion into lipid membranes and calcein efflux. The C-terminal domain was necessary for efficient association of the protein with lipid bilayers. IcmR was found to bind to the N-terminal portion of the protein thus providing a mechanism for its ability to inhibit IcmQ pore-forming activity. Localization of IcmQ on the surface of the L. pneumophila shortly after infection as well as its pore-forming capacities suggest a role for IcmQ in forming a channel that leads translocated effectors out of the bacterium. Legionella pneumophila is a Gram-negative bacterium that multiplies inside amoebae found in fresh water ecosystems (1Fields B.S. Trends Microbiol. 1996; 4: 286-290Abstract Full Text PDF PubMed Scopus (444) Google Scholar). It is believed that inhalation of aerosols containing either free-living L. pneumophila or amoeba laden with the bacteria leads to colonization of the lungs. Phagocytosed L. pneumophila then grow within alveolar macrophages in a fashion similar to that observed with amoeba (2Glavin F.L. Winn Jr., W.C. Craighead J.E. Ann. Intern. Med. 1979; 90: 555-559Crossref PubMed Scopus (44) Google Scholar, 3Katz S.M. Hashemi S. Lab. Invest. 1982; 46: 24-32PubMed Google Scholar), eventually causing a severe form of pneumonia known as Legionnaires' disease (4Fraser D.W. Tsai T.R. Orenstein W. Parkin W.E. Beecham H.J. Sharrar R.G. Harris J. Mallison G.F. Martin S.M. McDade J.E. Shepard C.C. Brachman P.S. N. Engl. J. Med. 1977; 297: 1189-1197Crossref PubMed Scopus (1307) Google Scholar). The ability of L. pneumophila to grow intracellularly can be recapitulated within cultured macrophages or cell lines (1Fields B.S. Trends Microbiol. 1996; 4: 286-290Abstract Full Text PDF PubMed Scopus (444) Google Scholar). Phagosomes harboring L. pneumophila avoid fusion with the endocytic network as early as 5-min postinfection (5Swanson M.S. Hammer B.K. Ann. Rev. Microbiol. 2000; 54: 567-613Crossref PubMed Scopus (303) Google Scholar, 17Roy C.R. Berger K.H. Isberg R.R. Mol. Microbiol. 1998; 28: 663-674Crossref PubMed Scopus (295) Google Scholar), as endosomal markers such as Rab5 or the transferrin receptor are excluded from the phagosome (6Horwitz M.A. J. Exp. Med. 1983; 158: 2108-2126Crossref PubMed Scopus (510) Google Scholar, 7Clemens D.L. Lee B.Y. Horwitz M.A. Infect. Immun. 2000; 68: 2671-2684Crossref PubMed Scopus (136) Google Scholar). Instead, the compartment appears to mature via a unique pathway that involves recruitment of components found in the early secretory pathway (8Tilney L.G. Harb O.S. Connelly P.S. Robinson C.G. Roy C.R. J. Cell Sci. 2001; 114: 4637-4650PubMed Google Scholar, 9Kagan J.C. Roy C.R. Nat. Cell Biol. 2002; 4: 945-954Crossref PubMed Scopus (363) Google Scholar). For example, the secretory pathway marker Arf1 has been recently localized on the L. pneumophila phagosome shortly after infection (10Nagai H. Kagan J.C. Zhu X. Kahn R.A. Roy C.R. Science. 2002; 295: 679-682Crossref PubMed Scopus (453) Google Scholar), and markers of early secretory vesicles are found surrounding the phagosome within 15 min after bacterial contact with macrophages (9Kagan J.C. Roy C.R. Nat. Cell Biol. 2002; 4: 945-954Crossref PubMed Scopus (363) Google Scholar). Within a few hours, rough endoplasmic reticulum accumulates and potentially docks with the phagosome (11Horwitz M.A. J. Exp. Med. 1983; 158: 1319-1331Crossref PubMed Scopus (496) Google Scholar, 12Swanson M.S. Isberg R.R. Infect. Immun. 1995; 63: 3609-3620Crossref PubMed Google Scholar). Inside this intracellular niche, the bacteria replicate, yielding dozens of bacteria before lysing out of the macrophage to initiate a new replication cycle. Several independent genetic screens have led to the identification of bacterial factors necessary for intracellular proliferation of L. pneumophila (13Berger K.H. Isberg R.R. Mol. Microbiol. 1993; 7: 7-19Crossref PubMed Scopus (496) Google Scholar, 14Brand B.C. Sadosky A.B. Shuman H.A. Mol. Microbiol. 1994; 14: 797-808Crossref PubMed Scopus (141) Google Scholar, 15Segal G. Shuman H.A. Infect. Immun. 1997; 65: 5057-5066Crossref PubMed Google Scholar, 16Vogel J.P. Andrews H.L. Wong S.K. Isberg R.R. Science. 1998; 279: 873-876Crossref PubMed Scopus (592) Google Scholar). Single mutations in the majority of the 25 dot/icm genes lead to complete inhibition of intracellular growth. Most dot/icm mutants fail to form a replicative phagosome and cannot avoid the endocytic pathway (17Roy C.R. Berger K.H. Isberg R.R. Mol. Microbiol. 1998; 28: 663-674Crossref PubMed Scopus (295) Google Scholar, 18Wiater L.A. Dunn K. Maxfield F.R. Shuman H.A. Infect. Immun. 1998; 66: 4450-4460Crossref PubMed Google Scholar). The majority of these genes show similarity to components of conjugative DNA transfer systems found on IncI plasmids R64 and Col1b-P9 G. Shuman H.A. Infect. Immun. 1997; 65: 5057-5066Crossref PubMed Google Scholar, 16Vogel J.P. Andrews H.L. Wong S.K. Isberg R.R. Science. 1998; 279: 873-876Crossref PubMed Scopus (592) Google Scholar, K. N. Mol. Microbiol. 2000; PubMed Scopus Google Scholar). with this L. pneumophila can the bacterial in a J.P. Andrews H.L. Wong S.K. Isberg R.R. Science. 1998; 279: 873-876Crossref PubMed Scopus (592) Google Scholar, G. Shuman H.A. Sci. S. 1998; PubMed Scopus Google Scholar). to other was that the of the dot/icm genes are the of a secretion that translocation of bacterial factors into the with this which the protein is a translocation of the Dot/Icm (10Nagai H. Kagan J.C. Zhu X. Kahn R.A. Roy C.R. Science. 2002; 295: 679-682Crossref PubMed Scopus (453) Google Scholar), as is a protein that with the membrane J.P. Isberg R.R. Mol. Microbiol. PubMed Scopus Google Scholar). L. pneumophila but not dot/icm mutants cell of the macrophages a of infection J.E. J.P. Andrews H.L. Isberg R.R. Mol. Microbiol. 1998; PubMed Scopus Google Scholar). It has been that the Dot/Icm proteins allow the secretion of a pore-forming protein that inserts into the cell membrane as of the secretion and the pore is for this with this a of in of infection but not intracellular proliferation in that the pore formation by L. pneumophila not be in translocation of into Mol. Microbiol. 2002; PubMed Scopus Google Scholar, S. Infect. Immun. 2002; PubMed Scopus Google Scholar). Several of the Dot/Icm proteins have been in have been demonstrated for components of this for the ability of IcmR to inhibit of IcmQ J. Kagan J.C. H. Roy C.R. Mol. Microbiol. 2000; PubMed Scopus Google Scholar, G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar). of and that these genes proteins necessary for intracellular intracellular as well as of infection J. Kagan J.C. H. Roy C.R. Mol. Microbiol. 2000; PubMed Scopus Google Scholar, G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar). these suggest in the secretion channel either within the bacterial or on the plasma that IcmR and IcmQ are found in J. Kagan J.C. H. Roy C.R. Mol. Microbiol. 2000; PubMed Scopus Google Scholar). The of these two proteins is found in a to of of IcmQ and two or of IcmR G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar). the of IcmQ forms a of a property of the of secretion in the of S. Mol. Microbiol. 1996; PubMed Scopus Google Scholar, Microbiol. Mol. Biol. Rev. 1998; PubMed Google Scholar). chaperone/substrate has been observed in the secretion of the that conjugative systems similar to Dot/Icm such W. L. X. S. L. Mol. Microbiol. PubMed Scopus Google Scholar, Sci. S. 2001; PubMed Scopus Google Scholar). We report here that the of IcmQ that is a IcmQ inserts into lipid membranes efflux of a of is of pore-forming such as S. J. Microbiol. Rev. PubMed Google Scholar). L. pneumophila and pneumophila in this were from the K.H. Isberg R.R. Mol. Microbiol. 1994; 14: PubMed Scopus Google Scholar). is a to K.H. Isberg R.R. Mol. Microbiol. 1994; 14: PubMed Scopus Google Scholar). is to and a were on and as (13Berger K.H. Isberg R.R. Mol. Microbiol. 1993; 7: 7-19Crossref PubMed Scopus (496) Google Scholar). For L. was For L. pneumophila were to of bacterial as a The PubMed Scopus Google was to IcmQ the of the in L. pneumophila and digestion of the G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google the was into the and macrophages were from of and in as J. Exp. Med. PubMed Scopus Google Scholar). of the was to the fusion protein in G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar). was to IcmQ in the of allow in formation of the and were into Scholar), of were to of and with for were by in of containing and on for min on with a were then by for 15 of was with the bacterial for a of of the with the was from by of to the and the The was then to the and the was was then by the with of and for 15 The was then and the was was the and the was by by of of IcmQ was into the to allow of IcmQ to a C-terminal and is were to a DNA the IcmQ from was by and into by and was with The C-terminal portion of IcmQ was into the to allow of of IcmQ to a C-terminal The DNA to of IcmQ was by and is from was by and into by and was with The was in the containing the of was into to allow of of IcmQ to a C-terminal The DNA to of was by and is from was by and into by and was with The was in the containing of the or C-terminal portion of IcmQ were to of and with for were in of and The was on for min on with a for 15 and were a a of the with proteins were of a containing IcmQ or were and IcmQ in this were from from this a from bacteria with the and in were a in a and then of were vesicles were by two either by G. PubMed Scopus Google or by G. PubMed Scopus Google Scholar). For the were in of and in the or of The were then for For the the lipid was for min in of and of and were and a water either or on the of the was or The was then by a pore membrane a lipid such as the a membrane to of the association of proteins with lipid of protein was with of lipid vesicles in and for min to in was to the to a of and of the was the in was on the and of was on The were then to in a for 15 The of the were by by of the or by IcmQ, the the with the was to with as in in was was observed The bacteria were the was with and were from the bacteria C.R. Isberg R.R. Infect. Immun. 1997; 65: PubMed Google Scholar). was then the by J. Kagan J.C. H. Roy C.R. Mol. Microbiol. 2000; PubMed Scopus Google that was as a of were in and on a were the with a containing and and then with of the were on and by L. pneumophila vesicles were by as that were in containing calcein N. PubMed Scopus Google Scholar). calcein was then a with The lipid vesicles were in to a of in and to was after to Dextran was in to the and lipid vesicles were as was after vesicles containing IcmQ were the a vesicles were in the of protein but in the of the lipid vesicles were a the of the vesicles were in and was as were found such that was of the of the vesicles of the been to with were then to lipid vesicles with and were as H. G. PubMed Scopus Google that the For were for min in on and membranes was then as G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar). were with the and were either to or the was on a to allow of the IcmQ was to of in before for the were then either in J. PubMed Scopus Google containing for min or to were on macrophages of infection of by for min and then for were for min in containing of the by with were with for bacteria in the of bacteria were with with macrophages or bacteria were then in were and with pneumophila as M.S. Isberg R.R. Infect. Immun. 1996; PubMed Google Scholar). were then with G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar). was to the were then with IcmQ with found in and secretion systems with lipid membranes from Rev. Microbiol. 2000; 54: PubMed Scopus Google Scholar). IcmQ such IcmQ and with into by IcmQ with in the of and was to the as a The was by to allow the formation of vesicles and the was on a to the of of the IcmQ in the with to the of the the of with the to the of the protein within the of the were to the of IcmQ was found to be the of IcmQ was found in the of the not IcmQ is thus to with and the of association on the of the We IcmQ with lipid vesicles vesicles were by with by to the IcmQ was to the vesicles in the of to a and that these of IcmQ with vesicles and of IcmQ was with vesicles or was found with the vesicles IcmQ appears to with vesicles in the of to of IcmQ with vesicles be the of association or into the the of the association of IcmQ with IcmQ was to containing or and to of a to IcmQ and a that to of IcmQ, not the association of IcmQ with lipid vesicles that IcmQ is not with the lipid IcmQ be by of that was on the of lipid that IcmQ was not within the of IcmQ that been into vesicles by the was to to the IcmQ was after in which of the were and that was or protein in the of the vesicles these that lipid IcmQ was into the lipid of IcmQ in the lipid be by in the of IcmQ or IcmQ in the of was to for min and by the of IcmQ was to to IcmQ was into a by and to the of IcmQ IcmQ into vesicles was to of IcmQ by to were with not insertion of the N-terminal portion of the IcmR of IcmQ into IcmQ has been to as a protein in L. pneumophila in J. Kagan J.C. H. Roy C.R. Mol. Microbiol. 2000; PubMed Scopus Google Scholar). IcmQ is found with IcmR in a that IcmR IcmQ from into this the of IcmQ was in a of L. from bacteria in were by and proteins were from proteins by a and The membranes were then with and with IcmQ was localized in the in a of IcmQ was found in the and the a of IcmQ is with the membrane and was for the as was not observed in a the of and proteins were was found in the in was in the membrane in The ability of IcmR to insertion of IcmQ in was then was to vesicles and the ability of IcmQ to with on a was of IcmQ was from 25 to of the IcmQ IcmR was IcmR is thus to IcmQ insertion into lipid vesicles by to of IcmQ of the on the of L. pneumophila with of IcmQ association with membranes by the of IcmR that bacterial with IcmQ is from IcmR to with a membrane that to be a were not to IcmQ by on L. macrophages G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar). the IcmQ was in L. pneumophila with and this was into contact with macrophages by a min of were IcmQ and and L. pneumophila and IcmQ be on the surface of the bacteria proteins such as IcmR or not be these not bacteria IcmQ in not show on surface and after contact with the of the bacteria with the macrophages for IcmQ was a in the of bacteria to the IcmQ on the surface of L. pneumophila on other Dot/Icm IcmQ was in dot/icm of the mutations in the dot/icm genes was found to the ability of IcmQ to on the surface of the bacteria shortly after the of IcmQ on the surface of Legionella that contact with macrophages a of IcmQ is to the membrane of the bacteria in a Most of this with the bacterial surface translocated across the membrane. IcmQ of of of IcmQ to in a form with a and to into lipid in the of its is of proteins such as or that have pore-forming Sci. S. 2001; PubMed Scopus Google Scholar, J. 2000; PubMed Scopus Google Scholar). the dye calcein to IcmQ was to form pores in a membrane N. PubMed Scopus Google Scholar). of this dye was into in the of and from the dye by and IcmQ is to form pores as inserts into these the dye be in to and of of the with was to the of after of IcmQ to vesicles containing calcein led to in in a protein as as 25 calcein of that observed with IcmQ of the of calcein to be the of the were IcmQ was not to the of calcein a of was to calcein efflux. of IcmR to IcmQ was to inhibit calcein efflux. IcmQ calcein in fashion The of calcein by IcmQ be either by the formation of a pore or by the by activity. these two of a calcein K. S. K. J. 1997; Full Text PDF PubMed Scopus Google was IcmQ the of the a dye be by Dextran in L. J. 1993; 65: Full Text PDF PubMed Scopus Google was to was a to the of by IcmQ insertion into vesicles of IcmQ, of of be calcein these be by was into the of the lipid that pores are in and by for to IcmQ for lipid the pore-forming of IcmQ were on lipid from and out of or calcein were efficient calcein efflux from lipid vesicles protein to association were in with the calcein efflux with vesicles in and The of association of IcmQ was in or association membrane association is not sufficient to pore formation and the lipid for the two to be of calcein efflux observed with the lipid was not to in the IcmQ was to calcein efflux from of the not a of and which calcein with of IcmQ sufficient to of the efflux observed with The of calcein observed with the be to the of of the or to the of this IcmR to the N-terminal of IcmQ ability of IcmR to association of IcmQ with lipid vesicles to that of IcmR with IcmQ a region of IcmQ in membrane the region of IcmQ in IcmR and in membrane the was to and the region of IcmQ that was from digestion by IcmR was the of as as IcmR was IcmQ was into C-terminal as the was were for the of IcmQ that been with IcmR as the to IcmQ and the to as by N-terminal and and of proteins were into protein The two to and of IcmQ and as with IcmQ the two and were to the The to of IcmR and the to of IcmQ and The of IcmR in the thus of the of IcmQ a central portion of IcmR It is that a of IcmR to is the that a portion of IcmR be in contact with IcmQ the region of the protein The N-terminal of IcmQ with and the N-terminal of IcmQ was by and IcmR IcmQ the ability of the N-terminal region of IcmQ to with and calcein efflux was to of IcmQ and was found to be and was with and the was to to membrane of protein that were on a was for IcmQ association of with the vesicles be the other the protein were for the association of with the be 25 these protein the were in the with The protein was found in the of the to the in these association of with the a association of with lipid is in the of the in efficient association of the N-terminal with vesicles observed for The was in the calcein was to calcein similar to that observed for IcmQ of protein The of was by either the of IcmR or as been observed for the association of with membranes of protein observed with IcmQ the ability to the vesicles on calcein The ability to calcein efflux that were to of insertion the that a few of in a be sufficient to allow calcein efflux from that the N-terminal portion of IcmQ was for membrane this region was and the protein was and the from was with this protein was to the was necessary for this activity. with this was to calcein efflux with vesicles be into lipid vesicles by the protein with in the of by of the by the of the on of the not vesicles were by and a portion of the protein was that the association of with was not to of the protein within the of the The association of with the vesicles not to be to of a that is not by this is as the protein is in and that the of the protein was in the these that the N-terminal portion of IcmQ region in IcmR is necessary for the insertion of the protein into can protein and is to form The C-terminal portion of IcmQ, on the other cannot into but appears to protein insertion in lipid can be into vesicles by the this portion of the protein have a with the in the of the a that IcmR has similar to secretion found in bacterial and secretion systems G. Isberg R.R. Mol. Microbiol. 2001; PubMed Scopus Google Scholar, J.P. Microbiol. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar, J.E. Science. PubMed Scopus Google Scholar). to the of IcmQ, the of IcmR. We found that a IcmQ was to into lipid membrane insertion of IcmQ efflux of but not Dextran that was a pore and in K. S. K. J. 1997; Full Text PDF PubMed Scopus Google Scholar, L. J. 1993; 65: Full Text PDF PubMed Scopus Google Scholar). of the of IcmQ in this the that the N-terminal of the protein is in with the C-terminal this in the N-terminal is in is with with a led to of the N-terminal portion of IcmQ from that this portion of the protein is in the a protein the of IcmQ was to into lipid vesicles and calcein that this portion of the protein is necessary and sufficient for these of the N-terminal portion of the protein led to a of the ability to into bilayers. The of IcmQ are thus to in into a lipid and calcein efflux of of IcmQ in this by protein protein on on in a new The C-terminal region of IcmQ is necessary for efficient association with lipid bilayers. a this be into lipid vesicles a this region of the protein is water membrane of the C-terminal is the that the insertion of the N-terminal of this the insertion of IcmQ into membranes as is a two with the N-terminal portion by the of the which this by with the lipid that membrane insertion and pore formation are to For with of such as the lipid IcmQ is to calcein efflux but with the the that into the membrane a protein or a a two was demonstrated for Mol. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar). this a membrane formation of the region of into the membrane and is for membrane but not into the and is not of the membrane the formation of a pore via insertion of a domain into the pore formation in lipid IcmQ in the of IcmR was not to the with its role as a of IcmQ IcmR the insertion of IcmQ into membranes in and in bilayers. a mechanism for this The of IcmR to IcmQ led to the of in the N-terminal portion of IcmQ, that IcmR to the region for membrane insertion into vesicles IcmR thus IcmQ insertion into lipid membranes by the region of IcmQ in this the in of IcmQ, its infection of IcmQ be by to L. pneumophila shortly after infection of that of IcmQ was localized on the bacterial surface after contact with IcmQ not be on the bacterial cell surface growth in that a after contact with the to of the It be that this was observed of IcmQ, and that of IcmQ in to be the mechanism of IcmQ to the membrane is as was independent of other Dot/Icm these a role for IcmQ either in the portion of the translocation channel or in of this portion of the known of IcmQ of this are a property to be with a dot/icm translocation apparatus J.P. Isberg R.R. Microbiol. PubMed Scopus Google Scholar). mutations in the allow of L. pneumophila the in are not but mutations in genes to components of the secretion the IcmQ is of the portion of the translocation are with its to the of the the of bacterial contact with the pore of IcmQ and its are to IcmR and IcmQ be in the of the and of the channel in contact with be out that for IcmQ in the bacterial membrane as well as on the cell of the translocation channel are not the of mutations in and the of the IcmQ protein are of translocated secretion translocated proteins such as and have been observed on the bacterial cell and has been to the of other effectors J. 1994; PubMed Scopus Google Scholar, S. K. H. J. 1996; PubMed Scopus Google Scholar, G. Mol. Microbiol. PubMed Scopus Google Scholar, Mol. Microbiol. 2000; PubMed Scopus Google Scholar). The observed of IcmQ the bacterial surface to translocation into the and pore formation are by the of the of IcmQ in the membrane from the of a that IcmR to be from to such a IcmQ has the property of and to into membranes forms similar to such as the S. S. J. Microbiol. Rev. PubMed Google or 1994; PubMed Scopus Google Scholar). similar pore-forming is observed with the protein J. 2000; PubMed Scopus Google and the protein Sci. S. 2001; PubMed Scopus Google Scholar), which are translocated to on the role of the pore-forming of IcmQ infection and to the other Dot/Icm We for and of the for the and for the with