Genes encoding putative effector proteins of the …...Genes encoding putative effector proteins of the type III secretion system of Salmonellapathogenicity island 2 are required for

Genes encoding putative effector proteins of the type IIIsecretion system of Salmonella pathogenicity island 2are required for bacterial virulence and proliferation inmacrophages

Michael Hensel, 1 Jacqueline E. Shea, 2 Scott R.Waterman, 2 Rosanna Mundy, 2 Thomas Nikolaus, 1

Geoff Banks, 2 Andre s Vazquez-Torres, 3 ColinGleeson, 2 Ferric C. Fang 3 and David W. Holden 2*1Lehrstuhl fur Bakteriologie, Max von Pettenkofer-Institutfur Hygiene und Medizinische Mikrobiologie, Munich,Germany.2Department of Infectious Diseases, Imperial CollegeSchool of Medicine, Du Cane Road, London W12 ONN,UK.3Department of Medicine, Division of Infectious Diseases,University of Colorado Health Sciences Center, 4200 EastNinth Avenue, Denver, CO 80262, USA.

Summary

The type III secretion system of Salmonella pathogeni-city island 2 (SPI-2) is required for systemic infection ofthis pathogen in mice. Cloning and sequencing of acentral region of SPI-2 revealed the presence of genesencoding putative chaperones and effector proteinsof the secretion system. The predicted products ofthe sseB , sseC and sseD genes display weak but sig-nificant similarity to amino acid sequences of EspA,EspD and EspB, which are secreted by the type IIIsecretion system encoded by the locus of enterocyteeffacement of enteropathogenic Escherichia coli. Thetranscriptional activity of an sseA :: luc fusion genewas shown to be dependent on ssrA , which is requiredfor the expression of genes encoding components ofthe secretion system apparatus. Strains carrying non-polar mutations in sseA , sseB or sseC were severelyattenuated in virulence, strains carrying mutations insseF or sseG were weakly attenuated, and a strainwith a mutation in sseE had no detectable virulencedefect. These phenotypes were reflected in the abilityof mutant strains to grow within a variety of macro-phage cell types: strains carrying mutations in sseA ,sseB or sseC failed to accumulate, whereas the growthrates of strains carrying mutations in sseE, sseF or

sseG were only modestly reduced. These data suggestthat, in vivo , one of the functions of the SPI-2 secretionsystem is to enable intracellular bacterial proliferation.

Introduction

Several Gram-negative bacterial pathogens secrete viru-lence proteins via specialized type III secretion systems(Mecsas and Strauss, 1996). These secretion systemscomprise a large number of proteins required to transferspecific effector proteins into eukaryotic host cells in acontact-dependent manner (Rosqvist et al., 1994; Zierlerand Galan, 1995; Collazo and Galan, 1997). Although sev-eral components of the secretion system apparatus showevolutionary and functional conservation across bacterialspecies (Salmond and Reeves, 1993), the effector pro-teins are less well conserved and have different functions.The Yersinia effectors YpkA and YopH have threonine/serine kinase and tyrosine phosphatase activities respec-tively (Galyov et al., 1993; Persson et al., 1995). The actionsof these and other Yops inhibit bacterial phagocytosis byhost cells, which is thought to enable extracellular bac-terial proliferation (for review see Cornelis and Wolf-Watz,1997). The Shigella Ipa proteins, secreted by the mxi/spatype III secretion system, promote entry of this bacteriuminto epithelial cells (Menard et al., 1996). The proteinsEspA, EspB and EspD, encoded by the locus of entero-cyte effacement (LEE) of enteropathogenic Escherichiacoli (EPEC) are secreted by a type III secretion systemand cause cytoskeletal rearrangements of host epithelialcells resulting in the formation of pedestal-like structureson the host cell surface (for a review see Donnenberg etal., 1997).

Salmonella typhimurium is unusual in that it containstwo type III secretion systems for virulence determinants.The first controls bacterial invasion of epithelial cells (Galanand Curtiss, 1989; Galan, 1996) and is encoded by geneswithin a 40 kb pathogenicity island (SPI-1) located at 63centisomes on the chromosome (Mills et al., 1995). Evi-dence for the second type III secretion system is basedon a set of corresponding genes within a second patho-genicity island (SPI-2) at 30 centisomes, which are requiredfor systemic growth of this pathogen in its host (Hensel et

Molecular Microbiology (1998) 30(1), 163–174

Q 1998 Blackwell Science Ltd

Received 7 May, 1998; revised 27 June, 1998; accepted 2 July, 1998.*For correspondence. E-mail [email protected]; Tel. (0181) 3833487; Fax (0181) 383 2078.

al., 1995; Ochman et al., 1996; Shea et al., 1996). We pro-posed that the SPI-2 secretion system genes should bedesignated as follows: ssa for genes encoding the secre-tion system apparatus, ssr for genes encoding secretionsystem regulators, ssc for genes encoding secretion sys-tem chaperones and sse for genes encoding secretionsystem effectors (Hensel et al., 1997a).

Many of the genes encoding components of the SPI-2secretion system are located in a 25 kb segment beginningat the 31 centisomes boundary of SPI-2 (Hensel et al.,1997b). On the basis of similarities with genes of otherbacterial pathogens, the first 13 genes from this boundary(the ssaK /U operon and ssaJ ) encode components of thesecretion system apparatus (Hensel et al., 1997a). A num-ber of additional genes including ssaC (orf 11 in Shea etal., 1996; spiA in Ochman et al., 1996) and ssrA (orf 12in Shea et al., 1996; spiR in Ochman et al., 1996), whichencode a secretion system apparatus protein and a twocomponent regulatory protein, respectively, are found ina region <8 kb upstream of ssaJ (Shea et al., 1996).

In this paper we describe an analysis of genes thatencode other components of the secretion system, locatedin the region between ssaJ and ssaC of SPI-2. DNA andprotein database searches with these genes identifiedthree whose products display weak but significant similar-ity to proteins secreted by the type III secretion systems of

EPEC and Yersinia, and two that are potential chaperonesfor these proteins. We show that transcriptional activity ofsseA is dependent on ssrAB, which encodes a two-com-ponent regulatory system required for the expression ofssa genes of SPI-2 (Valdivia and Falkow, 1997). Virulencetests and in vitro assays with strains containing non-polarmutations show that several sse genes are critical for Sal-monella virulence in mice and are required for bacterialaccumulation within macrophages.

Results

Organization of sse and ssc genes

As part of our effort to characterize SPI-2 genetically andfunctionally, we cloned and sequenced a central region ofthe pathogenicity island (Fig. 1A). DNA fragments cover-ing the region between ssaC and ssaJ were subclonedin plasmids p5-2 and p5-4 as indicated in Fig. 1C. Thearrangement and designation of genes in the 8 kb regionbetween ssaC and ssaK is shown in Fig. 1B. This sequ-ence is available from the EMBL database under acces-sion number AJ224892. The sequenced region extendsthe open reading frame (ORF) of a gene encoding a puta-tive subunit of the type III secretion apparatus referredto as spiB (Ochman et al., 1996). For consistency withthe universal nomenclature for type III secretion system

Q 1998 Blackwell Science Ltd, Molecular Microbiology, 30, 163–174

Fig. 1. Genetic organization of the effector gene region of SPI-2 and position of mutations.A. The position of this region in relation to other genes of the secretion system is shown.B. Genes encoding proteins with sequence similarity to effector proteins of other type III secretion systems (shaded arrows), putativechaperones of SPI-2 (hatched arrows), components of the type III secretion system (filled arrows) and without sequence similarity (openarrows) are indicated. Positions of a mTn5 insertion in mutant strain P10E11 and the aphT gene cassette in mutant strains HH100, HH102,HH104, HH107 and HH108 are marked.C. A restriction map is shown for BamHI (B), EcoRI (E), HindIII (H), EcoRV(V), SmaI (S) and ClaI (C). The positions of subclones p5-2,p5-4, p5-5, p5-7 and p5-8 are also indicated.

164 M. Hensel et al.

subunits (Bogdanove et al., 1996) and the nomenclature ofother SPI-2 genes (Hensel et al., 1997a), we propose thatthis gene be designated ssaD. The deduced amino acidsequence of ssaD is 24% identical to YscD of Y. entero-colitica. This is followed by an ORF with coding capacityfor a 9.3 kDa protein, 34% identical to YscE of Y. entero-colitica. Therefore, this gene is designated ssaE. A sequ-ence of 263 bp separates ssaE and a set of nine genes,several of which encode proteins with sequence similarityto secreted proteins or their chaperones from other patho-gens. These genes are separated by short intergenicregions or have overlapping reading frames and it is likelythat some are co-transcribed and translationally coupled.Therefore, the genes with similarity to those encodingchaperones were designated sscA and sscB, and theothers sseA-G. The amino acid sequence deduced fromsscA shows 26% identity/49% similarity over 158 aminoacid residues to SycD, the product of lcrH of Y. pseudo-tuberculosis that acts as a secretion-specific chaperonefor YopB and YopD (Wattiau et al., 1994). The amino acidsequence deduced from sscB shows 23% identity/36%similarity over 98 amino acid residues to IppI of Shigellaflexneri. IppI is a chaperone for S. flexneri invasion pro-teins (Ipas) (Baundry et al., 1988). As is the case for thesecretion chaperones SycD, IppI and SicA (Kaniga et al.,1995), SscB has an acidic pI (Table 1), whereas SscA hasan unusually high pI of 8.8.

SseA, SseE, SseF and SseG showed no significantsimilarities to DNA or protein database entries. SseB is25% identical/47% similar to EspA of EPEC over the entirelength of the 192-amino-acid-residue protein. SseD is 27%identical/51% similar to EspB of EPEC over 166 aminoacid residues. SseC has sequence similarity to a class ofeffector proteins involved in the translocation of othereffectors into the target host cell. These include YopB ofY. enterocolitica, EspD of EPEC and PepB of Pseudo-monas aerugunosa. SseC is <24% identical/48% similarto both EspD of EPEC and YopB of Y. enterocolitica(Fig. 2). EspD and YopB have two hydrophobic domainsthat are predicted to insert into target cell membranes (Pal-len et al., 1997). SseC contains three hydrophobic regionsthat could represent membrane-spanning domains. Otherfeatures of these predicted effector proteins are shown

in Table 1. Using the TMPREDICT program (Hofmann andStoffel, 1993), transmembrane helices are predicted forall the effector proteins apart from SseA, which is veryhydrophilic. Alignments of SseC to homologues in otherpathogens are shown in Fig. 2. Conserved amino acidsare mainly clustered in the central, more hydrophobic por-tion of the protein, but unlike YopB, there is no significantsimilarity to the RTX family of toxins. The conserved resi-dues in SseD are present mainly in the N-terminal halfof the protein. Comparison of the deduced amino acidsequences of sseABCDEF with entries in the PROSITEdatabase did not reveal the presence of any characteristicprotein motifs. We subjected the predicted amino acidsequences of the sse genes to searches using the pro-grams COIL and MULTICOIL (Lupas, 1997) as described byPallen et al. (1997). SseA and SseD are predicted tohave one trimeric coil each, and SseC is predicted to havetwo trimeric coils (Table 1). As EspB and EspD are pre-dicted to have one and two trimeric coils, respectively(Pallen et al., 1997), this provides further evidence thatthese proteins are functionally related.

Between sseG and ssaJ [a gene encoding a componentof the type III secretion apparatus (Hensel et al., 1997a)],three short ORFs were identified, designated ssaG, ssaHand ssaI, encoding proteins of 7.9 kDa, 8.1 kDa and 9.0 kDarespectively. SsaG has sequence similarity to YscF of Y.enterocolitica (25% identity over 71-amino-acid residues)and MxiH of S. flexneri (34% identity over 46 amino acidresidues). No significant sequence similarities were obtainedwith SsaH and SsaI.

Expression of sseA is dependent on ssrAB

To establish if the sse genes are part of the SPI-2 secre-tion system, we investigated the expression of an sseA::lucreporter gene fusion, integrated by homologous recombi-nation into the chromosome of different SPI-2 mutantstrains. In accordance with results from a previous study(Valdivia and Falkow, 1997), no significant transcriptionwas detected during growth in rich medium. However,transcriptional activity of sseA in a wild-type backgroundwas detected during growth in minimal medium (Fig. 3).Transposon insertions in ssrA and ssrB, encoding the


Table 1. Features of predicted proteins.Protein Mr ( kDa) pI Tm predictions Predicted coils

SseA 12.5 9.3 Hydrophilic At least one (trimer)SseB 21.5 4.7 One transmembrane helix NoneSseC 52.8 6.3 Three transmembrane helices At least two (trimers)SseD 20.6 4.8 Three transmembrane helices At least one (trimer)SseE 16.3 9.7 One transmembrane helix NoneSseF 26.9 4.4 Four transmembrane helices NoneSseG 24.4 9.2 Three transmembrane helices NoneSscA 18.1 8.8 Hydrophilic NoneSscB 16.4 4.7 Hydrophilic None

SPI-2 sse genes 165

sensor component and the transcriptional activator, res-pectively, resulted in 250- to 300-fold reduced expressionof sseA. Inactivation of hilA, the transcriptional activator ofSPI-1 (Bajaj et al., 1996), had no effect on sseA geneexpression. Transposon insertions in two genes encodingcomponents of the SPI-2 type III secretion apparatus(ssaJ::mTn5; ssaT::mTn5; Shea et al., 1996) also had nosignificant effect on the expression of sseA. These datashow that ssrAB is required for the expression of sseA,but that hilA is not.

Virulence tests with strains carrying non-polar mutations

DNA sequence analysis suggested that the sse genesmight encode effector proteins of the secretion system,but apart from a possible polar effect from a transposoninsertion in sscA (strain P10E11, Fig. 1) no strains carry-ing mutations in these genes were recovered in theoriginal STM screen for S. typhimurium virulence genesusing mTn5 mutagenesis (Hensel et al., 1995), and theirrole in virulence was unclear. To address this question,

we constructed strains carrying non-polar mutations insseA, sseB, sseC, sseE, sseF and sseG (Fig. 1) and sub-jected these strains to virulence tests by inoculating groupsof BALB/c mice intraperitoneally with each strain, as thiswas the method by which the original SPI-2 mutants wereidentified. Table 2 shows that all mice inoculated withstrains carrying mutations in sseA, sseB and sseC sur-vived a dose of 1 ×104 cfu, three orders of magnitudegreater than the LD50 of the wild-type strain, which isless than 10 cfu when the inoculum is administered by thei.p. route (Buchmeier et al., 1993; Shea et al., 1996). Thesame strains containing a plasmid carrying the corre-sponding wild-type allele were also inoculated into miceat a dose of 1 ×104 cfu. No mice survived these infections,which shows that each mutation can be complemented atleast partially by the presence of a functional copy of eachgene, and that each of these genes plays an important rolein Salmonella virulence. Strains carrying non-polar muta-tions in sseE, sseF and sseG caused lethal infectionswhen <1 ×104 cells of each strain were inoculated intomice by the i.p. route (Table 2) and were analysed in


Fig. 2. Alignment of the deduced SseC amino acid sequence to EspD of EPEC, YopB of Yersinia enterocolitica (Hakansson et al., 1993) andPepB of Pseudomonas aeroginosa (Hauser et al., 1998). The CLUSTALW algorithm of the MACVECTOR 6.0 program was used to construct thealignments. Similar amino acid residues are boxed, identical residues are boxed and shaded.


more detail by a competition assay with the wild-typestrain in mixed infections (three mice per test) to deter-mine whether they were attenuated in virulence. Thecompetitive index, defined as the output ratio of mutantto wild-type bacteria, divided by the input ratio of mutantto wild-type bacteria, shows that sseF and sseG do contri-bute to virulence, but their absence in mutant strains is notdetectable by the relatively insensitive LD50 test (Table 2).By comparison, very low competitive indices were obtainedusing strains carrying mutations in either ssrA or sseB.The competitive index for the sseE mutant was not signifi-cantly different to that of a fully virulent strain carrying anantibiotic resistance marker, which implies that this genedoes not play a significant role in systemic Salmonellainfection of the mouse.

Intramacrophage replication of mutant strains

We tested several mutant strains for their ability to growinside macrophages and macrophage-like cell lines, asmacrophage survival and replication are thought to repre-sent an important aspect of Salmonella pathogenesis invivo (Fields et al., 1986), and because Ochman et al.(1996) reported that an S. typhimurium strain carrying amutation in a SPI-2 gene was unable to survive in macro-phages. We have reported previously that a number ofSPI-2 mutant strains were not defective for survival or repli-cation within RAW macrophages (Hensel et al., 1997a), butsubsequent experiments have revealed that some SPI-2mutants can be shown to have a defect if aerated station-ary-phase bacterial cultures opsonized with normal mouseserum are used (see also accompanying paper: Cirillo etal., 1998). The increase in cfu for different strains in RAWmacrophages over a 16 h period is shown in Fig. 4. Growthdefects were observed for strains carrying mutations inssaV (encoding a component of the secretion apparatus),sseA, sseB, sseC and to a lesser extent for strains carryingmutations in sseE, sseF and sseG. Partial complementationof this defect was achieved with strains harbouring plasmidscarrying functional copies of sseC and sseB, and very slightcomplementation was observed for sseA. We also investi-gated the ability of SPI-2 mutant strains to accumulateinside the J774.1 macrophage cell line (Fig. 5A) and inperiodate-elicited peritoneal macrophages from C3H/HeNmice (Fig. 5B). Similar defects of S. typhimurium carryingtransposon or non-polar mutations in SPI-2 genes wereobserved, regardless of the phagocyte cell-type examined,although the peritoneal elicited cells had superior anti-microbial activity than either cell line.

Discussion

We have identified a group of genes encoding putativeeffector proteins and chaperones of the SPI-2 type III


Fig. 3. Expression of an sseA::luc fusion in wild-type and mTn5mutant strains of S. typhimurium. Expression in each strain wasdetermined in triplicate and the standard errors from the means areshown.

Table 2. Virulence of S. typhimurium strains inmice.

Strain Genotype

Mouse survivalafter inoculationa

with bacterialstrain

Mouse survival afterinoculationa withmutant þ complementingplasmid

Competitiveindex in vivo

12023 Wild-type 0/5 0.98b

HH100 sseAD: :aphT 5/5 0/5HH102 sseBD ::aphT 5/5 0/5 0.013HH104 sseC::aphT 5/5 0/5HH106 sseED 0/5 0.79HH107 sseFD ::aphT 0/5 0.222HH108 sseG ::aphT 0/5 0.158P3F4 ssrA ::mTn5 0.01

a. Mice were inoculated intraperitoneally with 1 ×104 cells of each strain and survival wasdetermined after 14 days.b. Result of competition between wild-type strain 12023 and a virulent mTn5 mutant identifiedin the STM screen.

SPI-2 sse genes 167

secretion system. These genes are located between twoclusters of genes encoding components of the secretionsystem apparatus. In terms of their overall arrangementand deduced amino acid sequences, the SPI-2 ssa andsse genes appear to share greatest similarity to the escand esp genes of the EPEC LEE (Elliot et al., 1998).Although sseBCD are in the same order as the genes towhich they show similarity (espADB), there is no genebetween espA and espD corresponding to sscA, whichencodes a putative chaperone. The corresponding geneproducts of sseCD and espDB are also predicted tohave transmembrane helices and to form trimeric coiledcoils which suggests that the two sets of genes have simi-lar functions.

EspA, EspB and EspD are necessary to activate epi-thelial signal transduction pathways leading to host cellcytoskeletal rearrangements and pedestal-like structureson intestinal epithelial cell surfaces, to which the bacteriaadhere (Donnenberg et al., 1997). Recent work has shownthat both EspB (Wolff et al., 1998) and EspA (Knutton etal., 1998) are involved in the translocation of EspB intoboth the plasma membrane and cytoplasm of epithelialcells. EspA is a component of a filamentous surface appen-dage that forms a contact between the bacterium and theeukaryotic cell surface (Knutton et al., 1998). These struc-tures resemble specific appendages (invasomes) that canbe detected on the surface of S. typhimurium during epi-thelial cell invasion, although their relationship to the SPI-1type III secretion system is uncertain (Ginocchio et al.,

1994; Reed et al., 1998). The SPI-1 secretion systemforms a needle-like structure that spans the bacterial cellenvelope (Kubori et al., 1998). Both SseC and EspD aresimilar to YopB, which is a pore-forming protein requiredfor translocation of effector proteins of the Yersinia typeIII secretion system across the eukaryotic cell plasma


Fig. 4. Intracellular survival and replication of SPI-2 mutant S.typhimurium in RAW 264.7 macrophages. After opsonization andinfection, macrophages were lysed and cultured for enumeration ofintracellular bacteria (gentamicin protected) at 2 h and 16 h afterinfection. The values shown represent the fold increase calculatedas a ratio of the intracellular bacteria between 2 h and 16 h afterinfection. Each strain was infected in triplicate and standard errorfrom the mean is shown.

Fig. 5. Intracellular survival and replication of SPI-2 mutant S.typhimurium in (A) J774.1 cells and (B) periodate elicited peritonealmacrophages from C3H/HeN mice. After opsonization andinternalization, phagocytes were lysed and cultured for enumerationof viable intracellular bacteria at time 0 h. The values shownrepresent the proportion of this intracellular inoculum viable at20 h 6 the standard error of the mean. Samples were processed intriplicate, and each experiment was performed at least twice.


membrane (Hakansson et al., 1996). These similaritiessuggest that at least some of the sse genes encode aSPI-2 translocon that integrates into a target cell mem-brane and mediates the translocation of other SPI-2 effec-tor proteins. Although intracellular Sse proteins can bedetected in bacterial lysates after growth in laboratorymedia using polyclonal antibodies and epitope tags (ourunpublished results), we have not been able to establishin vitro conditions that induce the secretion of these pro-teins. We are currently investigating their secretion invivo and within cultured host cells. Evidence that the ssegenes encode secreted proteins of the SPI-2 secretionsystem is therefore based on their chromosomal locationand order, similarities of predicted sequence and hydro-phobicity to Esps of EPEC and Yersinia YopB, and onthe fact that the expression of sseA requires SsrAB, atwo-component regulatory system required for the expres-sion of other genes of the secretion system (Valdivia andFalkow, 1997; and accompanying paper Cirillo et al., 1998).

The importance of the sse genes was originally unclearbecause all the STM-derived transposon insertions in SPI-2are in genes for structural and regulatory components (Sheaet al., 1996) and sscA, which is predicted to encode achaperone. The failure to recover transposon insertionsin the sse genes may have resulted from a low frequencyof mTn5 insertions in this region, because strains carryingtargeted mutations in sseA, sseB and sseC are stronglyattenuated in virulence. These strains also exhibited defi-cient replication or survival in two different macrophage-like cell lines and elicited mouse peritoneal macrophages.This phenotype was observed using opsonized bacteriabut does not require exogenous cellular activation by cyto-kines. This finding prompted us to re-examine the intra-macrophage accumulation of strains carrying transposoninsertions in ssrA and ssaV because in an earlier studyusing non-opsonized bacterial cells grown under condi-tions that make them highly invasive, these strains werenot found to have a replication defect (Hensel et al.,1997a). When ssrA and ssaV mutant strains were opsonizedand grown under non-invasive conditions, their intrama-crophage numbers were as low as the strains carryingmutations in sseA and sseB. These differences may reflectthe different intracellular fate of organisms entering macro-phages by pathogen-directed invasion processes (non-opsonized) or by phagocyte-directed uptake (opsonized).Furthermore, as high invasion leads to a greater degree ofhost cell cytotoxicity through SPI-1-mediated apoptosis(Monack et al., 1996; Chen et al., 1996), cytotoxic effectsof replicating wild-type cells may have released bacteriafrom the intracellular environment and resulted in killingby gentamicin, and therefore an under-estimation of thenumber of wild-type cells that had undergone intracellularreplication in our previous studies.

As the phenotypic behaviour of opsonized bacterial

strains carrying transposon or non-polar mutations in vari-ous SPI-2 genes inside cultured macrophages reflectstheir virulence phenotype in vivo, we conclude that theSPI-2 secretion system is required for bacterial prolifera-tion inside macrophages in vivo. However, this does notexclude the possibility that the SPI-2 genes have otherfunctions not apparent from the macrophage assays.These results confirm and extend the earlier findings ofOchman et al. (1996), who found that the secretion systemis required for bacterial survival in J774 macrophages.SPI-2 mutant strains were originally found to be attenu-ated in virulence by mixed infections of mice with poolsof signature-tagged mutant strains, the majority of whichare virulent (Hensel et al., 1995). This means the SPI-2mutant strains cannot be rescued in trans by the presenceof virulent strains. The demonstration of SPI-2 gene expres-sion within macrophages (Valdivia and Falkow, 1997), alongwith the reduced numbers of SPI-2 mutant cells in culturedmacrophages, suggests that the failure of virulent cells torescue the SPI-2 defect may be due to physical separationof bacteria within phagocytic cells. Their reduced accumu-lation could be due to either reduced survival (greater sus-ceptibility to macrophage killing) or reduced replication, ora combination of the two. Ochman et al. (1996) concludedthat the function of the secretion system might be to modifyhost factors required for phagosome–lysosome fusion orphagosome acidification. More recently Valdivia and Falkow(1997) showed that at least one of the SPI-2 secretion sys-tem genes is induced in a variety of host cells includingmacrophage-like RAW cells, and speculated that the func-tion of the secretion system might be to translocate bac-terial proteins across the vacuolar membrane. Therefore,the putative translocon encoded by the sse genes reportedhere might be inserted in the vacuolar membrane andinfluence the intracellular fate of the Salmonella-contain-ing vacuole. Further investigation of Salmonella–macro-phage interactions and the functions of the sse geneproducts are likely to provide important insights into themechanism of SPI-2-mediated virulence.

Experimental procedures

Bacterial strains, phages and plasmids

The bacterial strains, phages and plasmids used in this studyare listed in Table 3. Unless otherwise indicated, bacteriawere grown in LB broth or on LB agar plates with the addition,when appropriate, of ampicillin (50 mg ml¹1), kanamycin (50 mgml¹1) or chloramphenicol (50 mg ml¹1).

DNA cloning and sequencing

Clones harbouring fragments of SPI-2 were identified from alibrary of genomic DNA of S. typhimurium in l 1059, whichhas been described previously (Shea et al., 1996). The sse


SPI-2 sse genes 169

and ssc genes were subcloned from clone l 5 on a 5.7 kbEcoRI fragment and a 5.8 kb HindIII fragment in pKSþ asindicated in Fig. 1 and Table 3. DNA sequencing was per-formed using a primer walking strategy. The dideoxy method(Sanger et al., 1977) was applied using the Pharmacia T7sequencing system for manual sequencing and the dye termi-nator chemistry for automatic analysis on a ABI377 sequenc-ing instrument. Assembly of contigs from DNA sequenceswas performed by means of ASSEMBLYLIGN and MACVECTOR soft-ware (Oxford Molecular). For further sequence analyses, pro-grams of the GCG package version 8 (Devereux et al., 1984)were used on the HGMP network.

Construction of non-polar mutations

The construction of non-polar mutations in sseA, sseB, sseC,sseE and sseF are described below. All chromosomal modi-fications were confirmed by PCR and Southern hybridizationanalysis.

SseA

Plasmid p5-5 was cut at a unique HpaI site within sseA. Thelinear product was then used as template for a PCR withprimers sseA1 (58-GAAGGCCTTTTTCTTTATCATCATTC-CCC-38) and sseA2 (58-GAAGGCCTGAAACAACTTAATG-CTCAAGCC-38) to delete an internal region of 246 bp fromsseA, corresponding to amino acids 8–90. The linear productcontains Stu I sites at each terminus and was circularized byligation after StuI digestion and transferred into E. coli DH5a.A 900 bp HincII fragment of pSB315 containing an amino-glycoside 38-phosphotransferase gene (aphT ) from which thetranscriptional terminator had been removed (Galan et al.,1992) was ligated in the same orientation into the uniqueStu I site created in the sseA deletion plasmid. An NheI frag-ment containing the DsseA::aphT gene was inserted into theXbaI site of the plasmid pCVD442 (Donnenberg and Kaper,1991) and transferred by conjugation from E. coli S17-1 lpirto S. typhimurium 12023. Exconjugants were selected for kana-mycin resistance and screened for ampicillin sensitivity.


Table 3. Phages, plasmids and bacterial strains used in this study.

Phage, plasmid or strain Description Reference

Phagesl5 Clone from a library of S. typhimurium genomic DNA in l1059 Shea et al. (1996)

PlasmidspKSþ, pSKþ Ampr; high-copy-number cloning vectors Stratagene, HeidelbergpUC18 Ampr; high-copy-number cloning vector Gibco-BRL, EggensteinpACYC184 Cmr,Tet r; low-copy-number cloning vector Chang and Cohen (1978)pGP704 R6K ori, Ampr; lpir-dependent suicide vector Miller and Mekalanos (1988)pCVD442 R6K ori, Ampr; lpir-dependent sucrose selection suicide vector Donnenberg and Kaper (1991)pBL02 R6K ori, Ampr; luc fusion suicide vector Gunn and Miller (1996)pGPL01 R6K ori, Ampr; luc fusion suicide vector Gunn and Miller (1996)pSB315 Kanr, Ampr Galan et al. (1992)p5-2 Ampr; 5.7 kb EcoRI fragment of l5 in pKSþ This studyp5-4 Ampr; 5.8 kb HindIII fragment of l5 in pSKþ This studyp5-5 Ampr; Pst I digestion of p5-4 and religation of the larger fragment This studyp5-7 Ampr; 1.9 kb fragment of p5-4 in pKSþ

p5-8 Ampr; 2.2 kb Pst I /HindIII fragment of p5–2 in pSKþ This studypsseA Cmr; sseA in pACYC184 This studypsseB Cmr; sseB in pACYC184 This studypsseC Cmr; sseC in pACYC184 This study

E. coli strainsDH5a See reference Gibco-BRLS17-1 lpir lpir phage lysogen (see reference) Miller and Mekalanos (1988)CC118 lpir lpir phage lysogen (see reference) Herrero et al. (1990)

S. typhimurium strains12023 Wild-type NCTC, Colindale, UKCS015 phoP-102::Tn10d-Cm Miller et al. (1989)P3F4 ssrA ::mTn5 Shea et al. (1996)P2D6 ssaV ::mTn5 Shea et al. (1996)P4H2 hilA ::mTn5 Monack et al. (1996)P11D10 ssaJ ::mTn5 Shea et al. (1996)HH100 sseAD ::aphT, Kmr; non-polar mutation This studyHH101 HH100 containing psseA This studyHH102 sseBD ::aphT, Kmr; non-polar mutation This studyHH103 HH102 containing psseB This studyHH104 sseC::aphT, Kmr; non-polar mutation This studyHH105 HH104 containing psseC This studyHH106 sseED ; non-polar mutation This studyHH107 sseFD ::aphT, Kmr; non-polar mutation This studyHH108 sseG ::aphT, Kmr; non-polar mutation This study


SseB

Plasmid p5-5 was cut at the unique ClaI site within sseB.This linear DNA was used as the template for PCR usingprimers sseB1 (58-ATTGGATCCGGTGGAGATACCGTC-38)and sseB2 (58-TATGGATCCTGTTGTTAGGGTCGGG-38).The product, with terminal StuI sites, was digested with StuIand self-ligated to generate an sseB gene containing an internaldeletion of 333 bp corresponding to amino acids 55–166. Theblunt-ended aphT cassette (see above) was ligated into theStu I site in the sseB deletion. An NheI fragment containingthe sseB deletion was ligated into the unique XbaI site of plas-mid pCVD442 and transferred to the S. typhimurium chromo-some as described above.

SseC

A 2.6 kb fragment was recovered after BamHI and ClaI diges-tion of p5-2 and subcloned in BamHI/ClaI-digested pKSþ.The resulting construct was digested by HindIII, blunt endedusing the Klenow fragment of DNA polymerase and ligatedto the aphT cassette as indicated above. The resulting plas-mid was digested with Sal I and XbaI and the insert wasligated to Sal I/XbaI-digested pGP704. This plasmid was elec-troporated into E. coli S17-1 lpir and transferred into S. typhi-murium 12023 by conjugation. Exconjugants in which the sseCgene had been replaced by the cloned gene disrupted by inser-tion of the aphT cassette were selected by resistance to kana-mycin and screened for sensitivity to carbenicillin.

SseE

Plasmid p5-8 was used as template for a PCR with primerssseE1 (58-ATTATGCATGCATGGGAGCGACCTTTACACA-GCTT-38) and sseE2 (58-ATTTAGCATGCGGCGGTCTCCC-CTAAATATGCAGG-38). The PCR product, containing terminalSphI sites, was digested with SphI and self-ligated to createan internal deletion of 261 bp in sseE corresponding to aminoacids 26–112. The plasmid was digested with Sal I and Sst Iand a 2.0 kb fragment containing the deleted sseE genewas ligated into pCVD442. The resulting plasmid was usedto transfer the mutated gene onto the Salmonella chromo-some as described above, using sucrose selection to obtaincells from which the suicide vector had been lost (Donnenbergand Kaper, 1991).

SseF

The N-terminal and C-terminal regions of sseF were isolatedfrom plasmid p5–2 on a 2.3 kb Pst I/HindIII fragment and a2.7 kb EcoRI/Sst II fragment respectively. The Pst I/HindIIIfragment was subcloned in pBluescript SK, the resulting con-struct linearized at the polylinker XbaI site and the cohesiveends made flush by treatment with T4 DNA polymerase.The EcoRI/Sst II fragment was treated with T4 DNA polymer-ase and ligated to the linearized vector to form plasmidpsseF2, which contains sseF with a central deletion corre-sponding to amino acids 81–178. Plasmid psseF2 was linear-ized at a BamHI site adjacent to the destroyed XbaI site andligated with the aphT gene (see above). A plasmid was iden-tified carrying the aphT gene in the required orientation.

SseF::aphT was isolated on a EcoRI/Sst II fragment, blunt-ended as before, ligated into the SmaI site of pCVD422 andthe mutation transferred to the Salmonella chromosome asdescribed above.

SseG

A 1.0 kb SauIIIA fragment containing sseG was subclonedfrom p5-7 into pUC18. The resulting construct was digestedwith Asc I, blunt-ended using Klenow fragment of DNA poly-merase and ligated to the aphT cassette as indicated above.The resulting plasmid was digested with EcoRI and XbaI andthe insert was ligated into EcoRI/XbaI-digested pGP704.This plasmid was introduced into S. typhimurium 12023 andexconjugants carrying the aphT cassette in sseG were iso-lated as described above.

For complementation of non-polar mutations in sseA, sseBand sseC, the corresponding genes were amplified by PCRfrom genomic DNA using a series of primers correspondingto the region 58 of the putative start codons and to the 38

ends of the genes. These primers introduced BamHI restric-tion sites at the termini of the amplified genes. After digestionby BamHI, the genes were ligated to BamHI-digestedpACYC184 (Chang and Cohen, 1978) and transferred intoE. coli DH5a. The orientation of the inserts was determinedby PCR, and, in addition, DNA sequencing was performedto confirm the orientation and the correct DNA sequence ofthe inserts. Plasmids with inserts in the same transcriptionalorientation as the Tetr gene of pACYC184 were selected forcomplementation studies and electroporated into the S. typhi-murium strains harbouring corresponding non-polar mutations.

Virulence tests and macrophage survival assays

Groups of female BALB/c mice (20–25 g) were inoculatedintraperitoneally with either single or mixed S. typhimuriumstrains. The inoculum consisted of either 1 × 105 or 1 × 104 cfuin 0.2 ml of physiological saline. Bacterial strains were culturedas described by Shea et al. (1996).

For mixed infections, wild-type and mutant strains weregrown separately and mixed before injection. The cfu of bothstrains was checked by plating a dilution series of the inoculumonto LB and LB supplemented with kanamicin. For mixedinfections involving HH106, strain 12023 was first transformedwith pACYC184 (which does not affect its virulence, unpub-lished results) and cfu were checked by plating onto LB andLB supplemented with chloramphenicol. Mice were killed 48hafter inoculation and bacterial cfu were counted after platingdilution series of spleen homogenates onto LB and LB supple-mented with the appropriate antibiotic.

RAW 264.7 cells (ECACC 91062702), a murine macro-phage-like cell line, were grown in Dulbecco’s modified Eaglemedium (DMEM) containing 10% fetal calf serum (FCS) and2 mM glutamine at 378C in 5% CO2. S. typhimurium strainswere grown in LB to stationary phase and diluted to an OD600

of 0.1 and opsonized for 20 min in DMEM containing 10%normal mouse serum. Bacteria were then centrifuged ontomacrophages seeded in 24-well tissue culture plates at amultiplicity of infection of <1:10 and incubated for 30 min.After infection, the macrophages were washed twice withPBS to remove extracellular bacteria and incubated for 90 min


SPI-2 sse genes 171

(2 h after infection) or 16 h in medium containing gentamicin(12 mg ml¹1). Infected macrophages were washed twice withPBS and lysed with 1% Triton X-100 for 10 min and appropri-ate aliquots and dilutions were plated onto LB agar to enume-rate cfu.

Survival of opsonized S. typhimurium strains in J774.1cells (Ralph et al., 1975) or C3H/HeN murine peritoneal exu-date cells (from Charles River Laboratories) was determinedessentially as described by DeGroote et al. (1997), but with-out the addition of interferon g. Briefly, peritoneal cells har-vested in PBS with heat-inactivated 10% fetal calf serum 4days after intraperitoneal injection of 5 mM sodium periodate(Sigma) were plated in 96-well flat-bottomed microtitre plates(Becton-Dickinson) and allowed to adhere for 2 h. Non-adhe-rent cells were flushed out with prewarmed medium contain-ing 10% heat-inactivated fetal calf serum. In previous studies,we have established that >95% of the cells remaining afterthis procedure are macrophages. S. typhimurium from aeratedovernight cultures was opsonized with normal mouse serumand centrifuged onto adherent cells at an effector to targetratio of 1:10. The bacteria were allowed to internalize for15 min and washed with medium containing 6 mg ml¹1 gent-amicin to kill extracellular bacteria. At 0 h and 20 h, cells werelysed with PBS containing 0.5% deoxycholate (Sigma), withplating of serial dilutions to enumerate cfu.

Macrophages were examined microscopically over the courseof these experiments and did not show significant levels ofcytotoxicity.

Construction and analysis of sseA reporter gene fusion

A 1.1 kb SmaI/HincII fragment of p5-4 was subcloned intopGPL01, a suicide vector for the generation of luc fusions(Gunn and Miller, 1996). The resulting construct, in which1.0 kb upstream and 112 bp of sseA is translationally fusedto luc was used to transform E. coli S-17 lpir, and conjuga-tional transfer to S. typhimurium performed as described pre-viously (Gunn and Miller, 1996). Strains that had integratedthe reporter gene fusion into the chromosome by homologousrecombination were confirmed by PCR and Southern hybrid-ization analysis. Subsequently, the fusion was moved by P22transduction into the wild-type and various mutant strainbackgrounds with mTn5 insertions in SPI-1 or SPI-2 genes(Maloy et al., 1996). As a control, a strain was constructedharbouring a chromosomal integration of pLB02, a suicideplasmid without a promoter fusion to the luc gene (Gunnand Miller, 1996). For the analysis of gene expression, strainswere grown for 16 h in minimal medium with aeration. Aliquotsof the bacterial cultures were lysed and luciferase activity wasdetermined using a luciferase assay kit according to themanufacturer’s protocol (Boehringer Mannheim). Photondetection was performed on a Microplate scintillation/lumi-nescence counter (Wallac). All assay were carried out in tripli-cate and replicated on independent occasions.

Acknowledgements

This project was supported by grants from the MRC (UK) toDavid Holden, and DFG (Germany) (grant no. 1964/2–1) toMichael Hensel. A.V.T. and F.C.F. were supported in partby a grant from the National Institutes of Health (AI39557).

We gratefully acknowledge the use of the network serviceat HGMP Resource Centre, Hinxton, UK. We are grateful toDr J. S. Gunn for providing pGPL01 and pLB02 and to Dr Car-men Beuzon for help with the sequence analysis.

References

Bajaj, V., Lucas, R.L., Hwang, C., and Lee, C.A. (1996). Co-ordinate regulation of Salmonella typhimurium invasiongenes by environmental and regulatory factors is mediatedby control of hilA expression. Mol Microbiol 22: 703–714.

Baudry, B., Kaczorek, M., and Sansonetti, P.J. (1988) Nucleo-tide sequence of the invasion plasmid antigen B and C genes(ipaB and ipaC) of Shigella flexneri. Microb Pathog 4: 345–357.

Bogdanove, A.J., Beer, S.V., Bonas, U., Boucher, C.A., Coll-mer, A., Coplin, D.L., et al. (1996) Unified nomenclature forbroadly conserved hrp genes of phytopathogenic bacteria.Mol Microbiol 20: 681–683.

Buchmeier, N.A., Lipps, C.J., So, M.Y., and Heffron, F. (1993)Recombination-deficient mutants of Salmonella typhimuriumare avirulent and sensitive to the oxidative burst of macro-phages. Mol Microbiol 7: 933–936.

Chang, A.C., and Cohen, S.N. (1978). Construction andcharacterization of amplifiable multicopy DNA cloningvehicles derived from the P15A cryptic miniplasmid. J Bac-teriol 134: 1141–1156.

Chen, L.M., Kaniga, K., and Galan, J.E. (1996) Salmonellaspp. are cytotoxic for cultured macrophages. Mol Microbiol21: 1101–1115.

Cirillo, D.M., Valdivia, R.H., Monack, D. and Falkow, S. (1998)Macrophage-dependent induction of the Salmonella patho-genicity island 2 type III secretion system and its role inintracellular survival. Mol Microbiol 30: 175–188.

Collazo, C.M., and Galan, J.E. (1997). The invasion-associ-ated type III system of Salmonella typhimurium directs thetranslocation of Sip proteins into the host cell. Mol Microbiol24: 747–756.

Cornelis, G.R., and Wolf-Watz, H. (1997) The Yersinia Yopvirulon: a bacterial system for subverting eukaryotic cells.Mol Microbiol 23: 861–867.

DeGroote, M.A., Ochsner, U.A., Shiloh, M.U., Nathan, C.,McCord, J.M., Dinauer, M.C., et al. (1997) Periplasmicsuperoxide dismutase protects Salmonella from productsof phagocyte NADPH-oxidase and nitric oxide synthase.Proc Natl Acad Sci USA 94: 13997–14001.

Devereux, J., Haeberli, P., and Smithies, O. (1984) A com-prehensive set of sequence analysis programs for theVAX. Nucleic Acids Res 12: 387–395.

Donnenberg, M.S., and Kaper, J.B. (1991) Construction of aneae deletion mutant of enteropathogenic Escherichia coliby using a positive-selection suicide vector. Infect Immun59: 4310–4317.

Donnenberg, M.S., Kaper, J.B., and Finlay, B.B. (1997) Inter-actions between enteropathogenic Escherichia coli andhost epithelial cells. Trends Microbiol 5: 109–114.

Elliot, S.J., Wainwright, L.A., McDaniel, T.K., Jarvis, K.G.,Deng, Y.K., Lai, L.C., et al. (1998) The complete sequenceof the locus of enterocyte effacement (LEE) from entero-pathogenic Escherichia coli E2348/69. Mol Microbiol 28:1–4.



Fields, P.I., Swanson, R.V., Haidaris, C.G., and Heffron, F.(1986) Mutants of Salmonella typhimurium that cannot sur-vive within the macrophage are avirulent. Proc Natl AcadSci USA 83: 5189–5193.

Galan, J.E. (1996) Molecular genetic bases of Salmonellaentry into host cells. Mol Microbiol 20: 263–271.

Galan, J.E., and Curtiss, R. III (1989) Cloning and molecularcharacterization of genes whose products allow Salmon-ella typhimurium to penetrate tissue culture cells. ProcNatl Acad Sci USA 86: 6383–6387.

Galan, J.E., Ginocchio, C., and Costeas, P. (1992) Molecularand functional characterization of the Salmonella invasiongene invA: homology of InvA to members of a new proteinfamily. J Bacteriol 174: 4338–4349.

Galyov, E.E., Hakansson, S., Forsberg, A., and Wolf-Watz,H. (1993) A secreted protein kinase of Yersinia pseudo-tuberculosis is an indispensable virulence determinant.Nature 361: 730–732.

Ginocchio, C.C., Olmsted, S.B., Wells, C.L., and Galan, J.E.(1994) Contact with epithelial cells induces the formation ofsurface appendages on Salmonella typhimurium. Cell 76:717–724.

Gunn, J.S., and Miller, S.I. (1996). PhoP-PhoQ activatestranscription of pmrAB, encoding a two-component regula-tory system involved in Salmonella typhimurium antimicro-bial peptide resistance. J Bacteriol 178: 6857–6864.

Hakansson, S., Bergman, T., Vanooteghem, J.C., Cornelis,G., and Wolf-Watz, H. (1993) YopB and YopD constitutea novel class of Yersinia Yop proteins. Infect Immun 61:71–80.

Hakansson, S., Schesser, K., Persson, C., Galyov, E.E.,Rosqvist, R., Homble, F., and Wolf-Watz, H. (1996) TheYopB protein of Yersinia pseudotuberculosis is essentialfor the translocation of Yop effector proteins across thetarget cell plasma membrane and displays a contact-dependent membrane disrupting activity. EMBO J 15:5812–5823.

Hauser, A.R., Fleiszig, S., Kang, P.J., Mostov, K., and Engel,J.N. (1998) Defects in type III secretion correlate with inter-nalization of Pseudomonas aeruginosa by epithelial cells.Infect Immun 66: 1413–1420.

Hensel, M., Shea, J.E., Gleeson, C., Jones, M.D., Dalton, E.,and Holden, D.W. (1995) Simultaneous identification ofbacterial virulence genes by negative selection. Science269: 400–403.

Hensel, M., Shea, J.E., Raupach, B., Monack, D., Falkow, S.,Gleeson, C., and Holden, D.W. (1997a) Functional analy-sis of ssaJ and the ssaK/U operon, 13 genes encodingcomponents of the type III secretion apparatus of Salmon-ella pathogenicity island 2. Mol Microbiol 24: 155–167.

Hensel, M., Shea, J.E., Baumler, A.J., Gleeson, C., andHolden, D.W. (1997b) Analysis of the boundaries of Sal-monella pathogenicity island 2 and the correspondingchromosomal region of Escherichia coli K-12. J Bacteriol179: 1105–1111.

de Herrero, M., Lorenzo, V., and Timmis, K.N. (1990) Trans-poson vectors containing non-antibiotic resistance selec-tion markers for cloning and stable chromosomal insertionof foreign genes in gram-negative bacteria. J Bacteriol172: 6557–6567.

Hofmann, K., and Stoffel, W. (1993). TMbase – a database of

membrane spanning proteins segments. Biol Chem Hoppe-Seyler 347: 166.

Kaniga, K., Tucker, S., Trollinger, D., and Galan, J.E. (1995)Homologs of the Shigella IpaB and IpaC invasins arerequired for Salmonella typhimurium entry into culturedepithelial cells. J Bacteriol 177: 3965–3971.

Knutton, S., Rosenshine, I., Pallen, M.J., Nisan, I., Neves,B.C., Bain, C., et al. (1998) A novel EspA-associated sur-face organelle of enteropathogenic Escherichia coli involvedin protein translocation into epithelial cells. EMBO J 17:2166–2176.

Kubori, T., Matsushima, Y., Nakamura, D., Uralil, J., Lara-Tejero, M., Sukhan, A., et al. (1998). Supramolecular struc-ture of the Salmonella typhimurium type III protein secretionsystem. Science 280: 602–605.

Lupas, A. (1997) Predicting coiled-coil regions in proteins.Curr Opin Struct Biol 7: 388–393.

Maloy, S.R., Steward, V.L., and Taylor, R.K. (1996) Geneticanalysis of pathogenic bacteria, Cold Spring Harbor, NewYork: Cold Spring Harbor Laboratory Press.

Mecsas, J., and Strauss, E.J. (1996) Molecular mechanismsof bacterial virulence: type III secretion and pathogenicityislands. Emerg Infect Dis 2: 271–288.

Menard, R., Prevost, M.C., Gounon, P., Sansonetti, P., andDehio, C. (1996) The secreted Ipa complex of Shigella flex-neri promotes entry into mammalian cells. Proc Natl AcadSci USA 93: 1254–1258.

Miller, V.L., and Mekalanos, J.J. (1988) A novel suicide vec-tor and its use in construction of invertion mutations:osmoregulation of outer membrane proteins and virulencedeterminants in Vibrio cholerae requires toxR. J Bacteriol170: 2575–2583.

Miller, S.I., Kukral, A.M., and Mekalanos, J.J. (1989) A twocomponent regulatory system (phoP and phoQ) controlsSalmonella typhimurium virulence. Proc Natl Acad SciUSA 86: 5054–5058.

Mills, D.M., Bajaj, V., and Lee, C.A. (1995) A 40 kb chromo-somal fragment encoding Salmonella typhimurium inva-sion genes is absent from the corresponding region ofthe Escherichia coli K-12 chromosome. Mol Microbiol 15:749–759.

Monack, D.M., Raupach, B., Hromockyj, A.E., and Falkow,S. (1996) Salmonella typhimurium invasion induces apop-tosis in infected macrophages. Proc Natl Acad Sci USA 93:9833–9838.

Ochman, H., Soncini, F.C., Solomon, F., and Groisman, E.A.(1996) Identification of a pathogenicity island required forSalmonella survival in host cells. Proc Natl Acad Sci USA93: 7800–7804.

Pallen, M.J., Dougan, G., and Frankel, G. (1997) Coiled-coildomains in proteins secreted by type III secretion systems.Mol Microbiol 25: 423–425.

Persson, C., Nordfelth, R., Holmstrom, A., Hakansson, S.,Rosqvist, R., and Wolf-Watz, H. (1995) Cell-surface-bound Yersinia translocate the protein tyrosine phospha-tase YopH by a polarized mechanism into the target cell.Mol Microbiol 18: 135–150.

Ralph, P., Prichard, J., and Cohn, M. (1975) Reticulum cellsarcoma: and effector cell in antibody-dependent cell-mediated immunity. J Immunol 114: 898–905.

Reed, K.A., Clark, A., Booth, T.A., Hueck, C.J., Miller, S.I.,


SPI-2 sse genes 173

Hirst, B.H., and Jepson, M.A. (1998). Cell-contact-stimu-lated formation of filamentous appendages by Salmonellatyphimurium does not depend on the type III secretionsystem encoded by Salmonella Pathogenicity Island 1.Infect Immun 66: 2033–2039.

Rosqvist, R., Magnusson, K.E., and Wolf-Watz, H. (1994)Target cell contact triggers expression and polarizedtransfer of Yersinia YopE cytotoxin into mammalian cells.EMBO J 13: 964–972.

Salmond, G.P., and Reeves, P.J. (1993) Membrane trafficwardens and protein secretion in gram-negative bacteria.Trends Biochem Sci 18: 7–12.

Sanger, F., Nicklen, S., and Coulson, A.R. (1977) DNA sequ-encing with chain terminating inhibitors. Proc Natl Acad SciUSA 74: 5463–5467.

Shea, J.E., Hensel, M., Gleeson, C., and Holden, D.W. (1996)Identification of a virulence locus encoding a second type

III secretion system in Salmonella typhimurium. Proc NatlAcad Sci USA 93: 2593–2597.

Valdivia, R.H., and Falkow, S. (1997) Fluorescence-basedisolation of bacterial genes expressed within host cells.Science 277: 2007–2011.

Wattiau, P., Bernier, B., Deslee, P., Michiels, T., and Cor-nelis, G.R. (1994) Individual chaperones required for Yopsecretion by Yersinia. Proc Natl Acad Sci USA 91: 10493–10497.

Wolff, C., Nisan, I., Hanski, E., Frankel, G., and Rosenshine,I. (1998) Protein translocation into host epithelial cells byinfecting enteropathogenic Escherichia coli. Mol Microbiol28: 143–156.

Zierler, M.K., and Galan, J.E. (1995) Contact with culturedepithelial cells stimulates secretion of Salmonella typhi-murium invasion protein InvJ. Infect Immun 63: 4024–4028.



Documents

Genes encoding putative effector proteins of the …...Genes encoding putative effector proteins of the type III secretion system of Salmonellapathogenicity island 2 are required for