To establish an infection, Yersinia pseudotuberculosis utilizes a plasmid-encoded type III translocon to microinject several anti-host Yop effectors into the cytosol of target eukaryotic cells . YopD has been implicated in several key steps during Yop effector translocation, including maintenance of yop regulatory control and pore formation in the target cell membrane through which effectors traverse . These functions are mediated, in part, by an interaction with the cognate chaperone, LcrH . To gain insight into the complex molecular mechanisms of YopD function, we performed a systematic mutagenesis study to search for discrete functional domains . We highlighted amino acids beyond the first three N-terminal residues that are dispensable for YopD secretion and confirmed that an interaction between YopD and LcrH is essential for maintenance of yop regulatory control . In addition, discrete domains within YopD that are essential for both pore formation and translocation of Yop effectors were identified . Significantly, other domains were found to be important for effector microinjection but not for pore formation . Therefore, YopD is clearly essential for several discrete steps during efficient Yop effector translocation . Recognition of this modular YopD domain structure provides important insights into the function of YopD .

 
The ability to infect an animal or plant host is a feature common to many pathogens . Several of these pathogens utilize functionally homologous type III secretion systems (TTSSs) to translocate antihost effector proteins into target host cells (15, 36, 53) . Pathogenic Yersinia spp . contain a 70-kb virulence plasmid sufficient for establishing a TTSS (23, 30, 46, 66) . The type III needle complex is comprised of numerous Ysc (Yersinia secretion) components, which upon target cell contact secrete two classes of Yop proteins (Yersinia outer proteins), antihost effector proteins and proteins required for efficient translocation into target cells (14) . The translocated proteins aid in bacterial colonization by subverting host cell signaling, which enables bacteria to resist phagocytosis and compromise immune surveillance (1, 17) .

Induction of Yops by target cell contact (45, 49) can be mimicked in vitro by growing bacteria at 37°C in the absence of calcium (low-calcium response) (14) . Regulatory control of the TTSS is established by a positive and negative control loop that involves the AraC-like activator LcrF (also termed VirF) (12, 68) and the negative regulatory element LcrQ (also termed YscM) (45, 47, 57), respectively . Type III-dependent secretion of LcrQ appears to derepress yop transcription (45, 47) through a mechanism that requires another negative regulator, YopD (65) . While a yopD null mutant is growth restricted at 37°C and Yop synthesis is constitutively induced in vitro (22, 65), the presence of LcrQ in the bacterial cytoplasm of a yopD null mutant does not repress this constitutive Yop synthesis (65) . As YopD stability is dependent on the cytosolic cognate chaperone LcrH (also termed SycD) (20, 64), it follows that a YopD-LcrH complex, which may also include LcrQ (11), is essential for maintaining yop regulatory control (2, 21) . These data are consistent with the recent notion that YopD, not LcrQ, is the molecular switch that controls feedback inhibition of the TTSS (67) .

In addition, secreted YopD is essential for effector translocation (22, 28, 48) . Together with YopB and LcrV, YopD apparently participates in forming a pore complex in the plasma membrane through which effector translocation occurs (10, 26, 32, 41, 61) . However, YopD may associate with pores only transiently, since a portion of YopD localizes to the cytosol of infected cell monolayers (22) .

The observations that support the hypothesis that YopD plays a role at several steps of the translocation process are intriguing . However, it is difficult to conceptualize how YopD actually integrates these multiple activities . As there are homologues of this protein in other TTSSs (6, 8, 36), understanding the mechanism of YopD function would significantly increase our understanding of type III secretion . Until now, this has been impeded by the inability to investigate the functions of YopD in isolation from each other . In this study, we systematically mutagenized YopD . Analyses of the mutants verified that YopD is a multifunctional protein involved in several key steps that are necessary for efficient Yop effector translocation . In particular, discrete functional domains were isolated that enabled the role of YopD in translocation to be separated from its role in pore formation and regulation . The mosaic-like structure of YopD is likely to facilitate coordination of the multiple functions .

 
Bacterial strains, plasmids, and growth conditions. Bacterial strains and plasmids used in this study are listed in Table 1 . Unless indicated otherwise, bacteria were routinely cultivated in Luria-Bertani agar or broth at either 26°C (Yersinia pseudotuberculosis) or 37°C (Escherichia coli) with aeration . When required, appropriate antibiotics were added at the following final concentrations: carbenicillin, 100 µg per ml; kanamycin, 50 µg per ml; gentamicin, 20 µg per ml; spectinomycin, 20 µg per ml; and chloramphenicol, 25 µg per ml .

 
Recombinant DNA techniques, enzymes, and reagents. Chemical transformation in E . coli was performed by the method of Hanahan (27) . Plasmid DNA was purified from E . coli with a Quantum Prep miniprep kit (Bio-Rad Laboratories AB, Sundbyberg, Sweden) or with a Jetstar 2.0 plasmid midi kit (Genomed, Bad Oeynhausen, Germany) as described by the manufacturer . Standard DNA manipulation techniques were performed essentially as described elsewhere (51) . DNA fragments were recovered from agarose by spin column purification performed as described by the manufacturer (Amicon, Beverly, Mass.) . All modifying enzymes were purchased from either Roche Diagnostics Scandinavia (Bromma, Sweden) or New England Biolabs (Beverly, Mass.) . The DyNAzyme EXT DNA polymerase used for PCR amplification was purchased from Finnzymes (Espoo, Finland), oligonucleotides were purchased from DNA Technology A/S (Aarhus, Denmark), and deoxynucleoside triphosphates were purchased from Amersham Biosciences Europe GmbH (Uppsala, Sweden) .

Construction of sequential in-frame yopD deletion mutants. Amplified DNA fragments used for constructing the in-frame deletion mutations were generated by overlapping PCR (35) . The primer combinations used to create the in-frame yopD deletion mutants are listed in Table 2 . Each fragment containing a sequence flanking the specific yopD deletion was confirmed by sequence analysis by using a DYEnamic ET terminator cycle sequencing kit (Amersham Biosciences); this was facilitated by initial cloning into the pCR4-TOPO TA cloning vector (Invitrogen AB, Stockholm, Sweden) . The fragments were then cloned into XhoI- (or SpeI-) and XbaI-digested suicide mutagenesis vector pDM4 (40) . E . coli S17-1pir was used as the donor strain in conjugal mating experiments with Y . pseudotuberculosis . For selection for the appropriate allelic exchange events we used established methods (40) .

 
Protein stability. Protein stability in the presence of endogenous proteases was analyzed after growth in defined TMH medium as described previously (20) . In parallel, intrabacterial protein stability was also assessed by the method of Feldman and colleagues (18) for translocation-deficient yopD mutant alleles only . Protein fractions were analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and were transferred to a nitrocellulose membrane (Scheicher and Schuell, Dassel, Germany) by using a Hoefer SemiPhor semidry transfer unit (Amersham Biosciences) . Membrane-bound YopD was detected by using rabbit anti-YopD polyclonal antiserum, followed by horseradish peroxidase-conjugated anti-rabbit antibodies (Amersham Biosciences) prior to detection with the enhanced chemiluminescence system (ECL) used as directed by the manufacturer (Amersham Biosciences) .

Limited chymotrypsin digestion. Overnight cultures of translocation-deficient YopD mutants were diluted to an optical density at 600 nm of 0.2 in 10 ml of fresh culture medium (Luria-Bertani broth supplemented with 5 mM EGTA and 20 mM MgCl₂) . Bacteria were grown for 1 h at 26°C and then incubated at 37°C for 3 h . Cultures were centrifuged at 2,500 x g, and the supernatants containing secreted proteins were collected and passed through a 0.45-µm-pore-size sterile filter . CaCl₂ was added to a final concentration of 20 mM, and the tubes were placed on ice for 20 min . Each sample was divided into two 5-ml portions by transfer into new tubes . To one tube, chymotrypsin (Roche Diagnostics) was added to a final concentration of 10 µg per ml, while the other tube was treated as a control for spontaneous proteolysis . The tubes were incubated on ice for 30 min, and then proteinase activity was quenched and proteins were precipitated on ice for 1 h by addition of 0.1 volume of trichloroacetic acid . After centrifugation for 10 min at 2,500 x g, the supernatants were carefully removed by aspiration . Each remaining precipitate was dissolved in 50 µl of 2x protein sample buffer (100 mM Tris-HCl [pH 6.8], 4% SDS, 0.001% bromophenol blue, 10% glycerol, 2% ß-mercaptoethanol) and fractionated by SDS-PAGE by using 15% Tris-Tricine gels . Standard immunoblotting procedures were used to detect YopD peptides with polyclonal rabbit anti-YopD antiserum .

Growth phenotypes and the MOX test. Yersinia plating frequencies and the subsequent growth phenotypes when the organisms were grown under high- and low-Ca²⁺ conditions at 37°C were determined by using the MOX test (5, 23) . Briefly, wild-type Yersinia (YPIII/pIB102) was unable to grow on LA (Luria-Bertani broth supplemented with 2% agar) supplemented with magnesium oxalate (in the absence of Ca²⁺) at 37°C, and normal growth occurred on LA with 2.5 mM CaCl₂; this phenotype was termed calcium dependent (CD) . The yopD null mutant (YPIII/pIB621) was unable to grow at 37°C irrespective of the Ca²⁺ level (temperature sensitive [TS] phenotype) . For confirmation, we performed parallel experiments in which the growth phenotypes were determined at 37°C during logarithmic growth of bacteria in liquid TMH medium (lacking Ca²⁺) or medium supplemented with 2.5 mM CaCl₂ (21) .

Analysis of Yop synthesis and secretion. Yop synthesis and secretion were induced as previously described (20-22) . Briefly, total Yop levels were assessed by obtaining samples directly from the bacterial culture suspension, which contained a mixture of Yops secreted into the culture medium and Yops in intact bacteria . Samples of the cleared supernatant that contained only Yops secreted into the culture medium were obtained to assess protein secretion levels . All protein fractions were separated by SDS-PAGE and then subjected to immunoblotting . Specific proteins were detected on the membrane support by using rabbit polyclonal antisera raised against YopH, YopB, LcrV, YopD, and YopE .

Cultivation and infection of HeLa cells. Cultivation and infection of HeLa cells for cytotoxicity assays were performed by using standard methods (22) . The infection period was 5 h long, and at 1- h intervals the extent of morphological change was visualized by light microscopy . The degrees of HeLa cell cytotoxicity induced by Yersinia strains expressing mutant yopD derivatives were recorded by using a sliding scale . The cytotoxicity induced by wild-type Y . pseudotuberculosis (YPIII/pIB102) defined the upper limit, while the cytotoxicity induced by the yopD null mutant (YPIII/pIB621) defined the lower limit .

Ras modification assay. Y . pseudotuberculosis strains expressing ExoS from the high-copy-number vector pTS103-Gm were tested for the ability to induce ExoS-mediated modification of eukaryotic Ras in infected HeLa cells (59) . Modification of Ras was visualized by Western blotting with Immobilon-P membranes (Millipore, Bedford, Mass.) coupled to the use of an ECL+plus Western blot detection system (Amersham Biosciences) and an anti-Ras monoclonal antibody (Transduction Laboratories, Lexington, Ky.) .

Contact hemolysis of sheep erythrocytes. The contact hemolysis assay in the presence or absence of the carbohydrates raffinose, dextrin 15, and dextran 4 was performed as previously described (10, 26, 33) .

 
Generation and elemental characterization of sequential yopD mutants of Y . pseudotuberculosis. By using sequential deletional mutagenesis it has recently been demonstrated that YopD has two separate LcrH-binding sites, one at positions 53 to 149 and one at positions 278 to 292 (Fig . 1) (20) . To further investigate the presence of discrete YopD functional domains, these internal deletions were introduced in cis by allelic exchange onto the virulence plasmid of Y . pseudotuberculosis wild-type strain YPIII/pIB102 (Fig . 1) . We first examined the stability of the new mutants using an assay that detected their susceptibility to endogenous proteases (20) . Significantly, the wild-type and mutant proteins remained resistant to proteolysis (data not shown) . We interpreted this to indicate that the native structure of the YopD variants was preserved .

 
A yopD null mutant lacked yop regulatory control that permitted constitutive Yop synthesis at 37°C, even during growth in noninductive media, and it was concomitantly growth restricted at elevated temperatures irrespective of the Ca²⁺ level (TS phenotype) (5, 22, 65) . This phenotype is distinct from the growth phenotype of wild-type Y . pseudotuberculosis, which requires Ca²⁺ for growth at 37°C (CD phenotype) and exhibits a normal pattern of Yop synthesis (5, 22, 65) . We used a MOX test (5, 23) to analyze the regulatory status of each YopD deletion strain . As expected, a CD phenotype was obtained with Y . pseudotuberculosis wild-type strain YPIII/pIB102 (data not shown) . Furthermore, YopD deletion mutants 1, 2, 7 to 11, and 13 had an equivalent CD phenotype (Fig . 1) . On the other hand, YopD deletion mutants 3 to 6 and 12, which lacked essential domains required for LcrH binding (20), had a TS phenotype, being unable to form colonies at 37°C; this is comparable to the growth restriction observed for a yopD null mutant (Fig . 1) . In parallel, we confirmed by immunoblotting that the TS mutants were also derepressed for Yop synthesis (data not shown) . Taken together, these findings show that YopD-dependent yop regulatory control requires a YopD-LcrH complex, which is consistent with previous reports (2, 9, 21) .

We next investigated the secretion efficiency of each YopD variant and the effect on secretion of other essential type III substrates . Secreted proteins from cleared culture supernatants of bacteria grown both in the absence of Ca²⁺ (brain heart infusion [BHI] broth supplemented with 5 mM EGTA and 20 mM MgCl₂ [inductive conditions]) and in the presence of Ca²⁺ (BHI broth supplemented with 2.5 mM CaCl₂ [repressive conditions]) were analyzed . We observed efficient YopD secretion from mutants 5 to 13 and impaired YopD secretion from mutants 3 and 4 during growth in secretion-competent medium (without Ca²⁺) (Fig . 2) . On the other hand, secretion by mutants 1 and 2, although adequately expressed (data not shown), was not detected under these assay conditions (Fig . 2) . This observation showed that the N-terminal 47 residues are important for efficient YopD secretion . Significantly, other type III substrates (translocators [YopB and LcrV] and effectors [YopE and YopH]) were generally secreted at normal wild-type levels (Fig . 2), although more secretion was typically observed for YopD mutants lacking yop regulatory control (null mutant and mutants 3 to 6 and 12) . In addition, the same mutants specifically secreted LcrV in noninducing media (with Ca²⁺) (Fig . 2), a phenomenon reported previously for mutants with mutations in both YopD (65) and its cognate chaperone, LcrH (21, 56) .

 
Identification of discrete YopD domains essential for Yop effector translocation. Having defined the secretion effects imposed by the various yopD mutations, we tried to identify discrete translocation domains . A yopD null mutant does not translocate the YopE Rho-GAP cytotoxin or the YopH tyrosine phosphatase into infected HeLa cell monolayers (22) . To investigate if a particular region of YopD was involved in the translocation, we initially used the YopE-dependent HeLa cell cytotoxicity assay (48) to determine translocation competency . As expected, nonsecreted YopD mutants 1 and 2 and the yopD null mutant did not translocate YopE, as the cellular morphology of infected monolayers was indistinguishable from that of uninfected cell monolayers (Fig . 3, compare panels C and D with panel A) . Similarly, YopD mutants 3 to 6 and 10 to 13 also did not induce YopE-dependent cytotoxicity towards HeLa cells (Fig . 3E to H and L to O) . In fact, only YopD mutants 7 to 9 were indistinguishable from the wild type in terms of being able to efficiently translocate YopE into target cells (Fig . 3, compare panels I to K with panel B) .

 
In parallel, we introduced the Pseudomonas aeruginosa ExoS effector molecule (pTS103-Gm) (10) into the Yersinia yopE null mutant (YPIII/pIB522) (19) that also contained a variant of YopD (1 to 13) . Because there is an ADP-ribosylating domain that relies on the eukaryotic 14-3-3 protein family (4, 31), ExoS-dependent ADP ribosylation of cytoplasmic proteins, such as Ras, provides a sensitive method for detecting direct translocation of ExoS into target cells (10, 31, 59, 60) . A YopE defective background was chosen to avoid an inhibitory effect on pore formation (63) that may have had an impact on the ExoS translocation efficiency . We confirmed that ExoS encoded on plasmid pTS103-Gm was efficiently secreted from all strains, except when it was combined with the 1, 2, and 13 in-frame deletions of YopD (Fig . 4A) . At least for the 1 and 2 backgrounds, this phenomenon was also observed for Yop secretion (Fig . 2) . The ExoS-producing strains were then used to infect HeLa cell monolayers . Cytoplasmic Ras in infected HeLa cell lysates was fractionated by SDS-PAGE and immunoblotted with monoclonal anti-Ras antibodies . ExoS-dependent ADP ribosylation of Ras is easily observed as a more slowly migrating band after fractionation (10, 31, 59, 60) . Similar to their inability to translocate YopE, YopD deletion mutants 1 to 6 and 10 to 13 did not translocate ExoS, as Ras remained unmodified even after 135 min of infection, similar to the results obtained for the uninfected control (Fig . 4B and data not shown) . Significantly, strains harboring the wild-type YopD (yopE) and the 7 to 9 variants were all able to translocate ExoS, yet distinct differences in translocation efficiency were detected . Interestingly, the 8 mutant translocated ExoS at a higher rate than the parental strain, as judged by the more rapid disappearance of unmodified Ras . On the other hand, the rate of translocation by the 7 mutant was notably lower, while the rate of translocation by the 9 mutant was intermediate, despite the fact that both organisms were able to rapidly induce YopE-dependent cytotoxicity (Fig . 3 and 4) . This further established the utility of the ExoS reporter system as an important tool in establishing the efficiency of effector translocation .

 
The data described above highlight regions of YopD (regions in 7 to 9, encompassing residues 150 to 227) that are dispensable for translocation . Furthermore, in the secreted YopD mutants, residues 53 to 149 (in 3 to 6) and the C terminus extending from residue 234 (in 10 to 13) are required for Yop effector translocation . Despite this translocation defect, the mutants with deletions encompassing residues 234 to 275 (10 and 11) and the extreme C terminus from residue 293 (13) all exhibit normal yop regulatory control . This finding revealed novel YopD domains that are involved solely in effector translocation .

Examination of the intrabacterial stability of YopD. Identification of YopD mutants 10, 11, and 13, which are defective only in translocation, was an important finding . Therefore, we examined the stability of these three mutants to ensure that the phenotypic readout was not due to the in-frame deletions imparting a conformation biased toward one experimental assay over another . We used an intrabacterial stability assay (18) as a means to quantitate the native folding propensity of these mutants . We did not detect any significant differences in the susceptibilities of the wild-type and mutant proteins to endogenous proteases over a 60-min period (Fig . 5A) . This was true for either the 8 mutant, which was phenotypically indistinguishable from the wild type, or mutants 10, 11, and 13, which exhibited a defect in Yop effector translocation but not in regulation . This was reflected by the similar half-lives (range, 24 to 33 min) of all the variants tested (Fig . 5B) .

 
We also used chymotrypsin digestion of secreted YopD to further probe for structural differences between the wild-type and mutant proteins . Notable differences in protein digestion patterns would have been indicative of altered protein folds . Chymotrypsin was used because it was predicted to target about five recognition sites in YopD with a cleavage probability of 80%, as determined with the PeptideCutter web server (http://us.expasy.org/tools/peptidecutter) (data not shown), which resulted in an even distribution of peptide fragments in the size range from approximately 8 to 32 kDa . Significantly, YopD variants secreted from Yersinia all exhibited very similar peptide fragmentation patterns in the suggested size range (Fig . 5C and D) . Small deviations likely occurred because each in-frame deletion was predicted to remove at least one chymotrypsin recognition site .

Taken together, these results supported the conclusion that the YopD variants encoded by the 10, 11, and 13 mutant alleles had comparable tertiary structures . This conclusion was also supported by the ability of the mutants to still bind LcrH (20), a prerequisite for yop regulatory control (21) . We interpreted this to indicate that the loss of Yop effector translocation, but not the loss of yop regulatory control, in these YopD mutant backgrounds is a bona fide phenotype .

Contact hemolysis as a prelude to Yop effector translocation. A number of studies have implicated YopD in a tripartite complex with the other translocator proteins LcrV and YopB, which forms pores in the eukaryotic membrane through which Yop effectors are translocated (10, 32, 41, 61) . Therefore, we wanted to investigate whether the participation of YopD in pore formation always ensures functional Yop translocation . Workers in our laboratories routinely use hemoglobin release from erythrocytes infected with Yersinia as a measure of pore formation in eukaryotic membranes (10, 26, 32, 33) . All the yopD deletion alleles were introduced into the yopK null mutant (YPIII/pIB155) (34), since this background enhances the assay sensitivity (10, 26, 32) . As anticipated, YopD mutants 7, 8, and 9 were able to lyse erythrocytes (Fig . 6A), which paralleled the capacity of the mutants to translocate antihost effectors into target cells (Fig . 3 and 4) . Indeed, lysis by the 8 mutant was comparable to lysis by the parental YPIII/pIB155 strain producing wild-type YopD . On the other hand, not only were mutants 4, 5, and 10 to 13 unable to translocate Yop effectors, they were also not capable of lysing erythrocyte membranes, which is comparable to the results obtained for the control strain harboring a full-length deletion of yopD (Fig . 6A) . Surprisingly, however, lysis was still observed for mutants 1 to 3 and 6 (Fig . 6A), even though these mutants displayed no evidence of effector translocation (Fig . 3 and 4) . It is noteworthy that the degree of lysis was no different than the degree of lysis that we observed for mutants 7 and 9, which could translocate YopE and ExoS .

 
To confirm that the observed lytic activities of wild-type Y . pseudotuberculosis and isogenic YopD mutants 1 to 3 and 6 to 9 were actually due to the ordered membrane insertion of YopD-dependent pores having similar dimensions, we performed an osmoprotection assay . Osmoprotectants of suitable sizes can prevent osmotic lysis of erythrocytes (10, 32) . After infection of sheep erythrocytes, the largest carbohydrate, dextran 4 (diameter, 3 to 3.5 nm), and the medium-size carbohydrate, dextrin 15 (diameter, 2.2 nm), dramatically inhibited this activity in a manner that was typically size dependent (Fig . 7) . Moreover, the smallest carbohydrate, raffinose (diameter, 1.2 to 1.4 nm), generally poorly prevented hemoglobin release except with mutant 9, in which there was a >50% reduction in hemolysis . Nevertheless, the overall pattern of osmoprotection observed for the specific YopD mutants was comparable to the wild-type pattern, which clearly demonstrated that the lytic activity was due to the specific formation of YopD-dependent pores that were of similar sizes in erythrocyte membranes and not due to nonspecific membrane disruption . Significantly, this assay also confirmed that the 3 and 6 mutant proteins were able to form pores that resembled those formed by wild-type YopD yet were not sufficient for effector microinjection . Hence, YopD-dependent pore formation does not guarantee functional translocation . This strongly supports the hypothesis that there is a multistep process during translocation and that pore formation is only one discrete part of this process . Thus, YopD appears to play an important role in more than one step of the translocation process .

 
Even though the mutant 1 and 2 proteins were apparently not secreted at detectable levels (Fig. 2), it is intriguing that they could form pores with sizes similar to the size of wild-type pores in erythrocyte membranes (Fig . 7) . Therefore, to further examine this finding, into these YopD variants we introduced a second-site mutation (I288K) in the C-terminal amphipathic -helix that is known to eliminate pore formation (Olsson and Francis, unpublished data) . Significantly, the double YopD mutants were no longer able to form pores in erythrocyte membranes (Fig . 6B), which reinforces the idea that the pore-forming ability of the 1 and 2 mutant proteins is dependent on secreted forms of YopD_4-20 and YopD_23-47, respectively . Since this suggests that these variants are actually secreted, albeit at very low levels, we sought to clarify this issue . The YscU inner membrane protein is essential for type III secretion (37) . Therefore, we introduced a yscU deletion into 1 and 2 mutants that also contained a yopK deletion . However, as a loss-of-function mutation in the TTSS represses Yop synthesis (14), we restored Yop synthesis by introducing a deletion into the LcrQ repressor element (45, 47, 57) . The new mutants were compared to their isogenic parents for the ability to secrete YopD and the ability to form pores in erythrocyte membranes . Clearly, the yopK 1 and yopK 2 mutants could indeed secrete minor amounts of YopD_4-20 and YopD_23-47, respectively, but the secretion was visualized only after fivefold more protein was fractionated by SDS-PAGE (compared to the amounts in experiments whose results are shown in Fig . 2) and after overexposure of the ECL-detected membrane to X-ray film (Fig . 8) . This secretion was type III dependent because the quadruple mutants (having deletions in yopK, yscU, lcrQ, and yopD) did not secrete detectable levels of either YopD variant (Fig . 8) . Moreover, these nonsecreting mutants were not able to induce pore formation in our hemoglobin release assay (Fig . 6C) . Thus, like mutants 3 and 6, mutants 1 and 2 could generate pores, but they were unable to establish effector microinjection . The failure of these two mutants to translocate effectors may not have been due simply to low secretion levels, since we observed several LcrH mutants that exhibited comparably poor levels of YopD secretion yet still translocated effectors into infected cell monolayers with the efficiency of the wild-type strain (Edqvist and Francis, unpublished) . Hence, the collective data give credence to the hypothesis that there is multistep involvement of YopD in this complex virulence strategy and that there are discrete functional domains in YopD .

 
Several studies have pointed to involvement of YopD at multiple levels during type III secretion by Yersinia . Until now, it has been difficult to determine whether the collective findings are a true reflection of the role of YopD during an infection . Therefore, we performed comprehensive phenotypic mapping of sequential in-frame YopD deletion mutants of the plasmid-encoded TTSS of Y . pseudotuberculosis . Discrete functional domains were identified that were necessary to facilitate YopD secretion, LcrH chaperone binding, yop regulation, effector translocation, and pore formation (Table 3) . In particular, YopD mutants that could separate pore formation in eukaryotic membranes from microinjection of antihost effectors into target host cells were isolated . Importantly, this suggested that these two events could be separated . It is noteworthy that the different functions observed for YopD encompass nearly every step of the process leading to type III-mediated translocation of Yop effectors into target cells . In view of this, it is not surprising that independent studies may result in different conclusions with respect to the function of YopD . Presumably, this would also be the case for other type III proteins from different bacterial pathogens . Despite this, our results clearly show that YopD is truly a multifunctional protein that has a mosaic of discrete domains which have individual functions .

 
In bacterial pathogens of animals, members of the TTSS translocon family often have multiple functions involving regulatory, structural, and effector mechanisms . This is particularly true for IpaB and IpaC of Shigella spp . (16, 43), SipB and SipC of Salmonella enterica serovar Typhimurium (16), EspB and EspD of enteropathogenic and enterohemorrhagic E . coli (16), and YopD and LcrV of Yersinia spp . (3, 13, 14; this study) . This level of complexity presents obstacles to studying the individual functions of these important proteins . Therefore, the fact that we isolated three YopD mutants, 10 (YopD_234-254), 11 (YopD_256-275), and 13 (YopD_293-305),which are specifically defective in pore formation and translocation of antihost effectors into target cells but not yop regulatory control and chaperone binding, is important .

Not surprisingly, deletion of the putative coiled coil domain encompassing residues 248 to 277 (44) eliminated effector translocation . As this domain is intimately involved in protein-protein interactions (38) and is widespread in proteins of TTSSs (16, 44), this mutant phenotype supports the notion that the Yersinia translocon is a multiprotein complex that presumably includes YopD, YopB, and LcrV . However, the observed interactions of YopD with either LcrV (52) or YopB (29, 42) are not known to specifically involve the YopD coiled coil domain . Another consideration is that the coiled coil domain may initiate YopD multimerization (39, 62), which might be a requirement for biological function, as was recently proposed for PopD, a YopD homologue encoded by the TTSS of P . aeruginosa (54) . To investigate the translocon complex, one approach which we have used is a comparative analysis of the functional complementation of translocon components from Y . pseudotuberculosis and P . aeruginosa (8-10) . This analysis revealed that YopD and PopD, despite their similarities, were not functionally interchangeable, indicating that only interactions with native translocon members support effector translocation (8) . We are currently pursuing this line of investigation with the view that the coiled coil domain may indeed confer specificity within YopD toward its binding partners . This is particularly significant since a corresponding coiled coil domain in PopD was not predicted (8) .

The translocon proteins are thought to generate a pore complex in the eukaryotic membrane through which antihost effectors pass en route to the inside of infected cells (15, 36) . A paradoxical finding, however, is that several of the pore components may partially localize to the target cell cytosol, resulting in an intracellular effector function (6, 16) . In this regard, YopD was observed inside infected HeLa cells, but no obvious effector function was identified (22) . Perhaps translocating YopD might be necessary to maintain effector molecules in a translocation-competent state during microinjection . Indeed, the fact that YopD interacts with the YopE cytotoxin in vitro (29) may well be a consequence of this putative chaperone role in vivo . In this study, we identified four YopD mutants, 1 (YopD_4-20), 2 (YopD_23-47), 3 (YopD_53-68), and 6 (YopD_128-149), that did not translocate effector proteins (YopE and ExoS), even though the pore formation in erythrocyte membranes was equivalent to that induced by mutants 7 (YopD_150-170) and 9 (YopD_207-227), which could efficiently translocate effectors . This was a surprising finding and supported the hypothesis that pore formation is not sufficient for effector microinjection . The data for these mutants indicate that there is an additional function of YopD in the translocation process that extends beyond pore formation . Interestingly, deletion of the region encompassing residues 128 to 149, which formed pores without mediating effector translocation (6), corresponded to a putative hydrophobic domain (25) that is required for YopE binding (29) . Taken together, these data indirectly support the hypothesis that YopD has a guidance role during translocation, and they may also have implications for the intracellular role of translocon components from other animal pathogens . In this context, the observation that mutant 8 (YopD_174-198) has an enhanced rate of ExoS translocation compared to that of wild-type Yersinia, while it exhibits an equivalent degree of lytic activity, is also very significant . Indeed, this supports our notion that pore formation in target cell plasma membranes and effector microinjection are separate events and that discrete domains of YopD are dedicated to each activity .

It is also interesting that the poorly secreting mutants 1 (YopD_4-20) and 2 (YopD_23-47) could still induce pore formation but not effector translocation . This was not due to low levels of secreted YopD, since very low levels of secreted YopD and YopB translocator proteins were all that was necessary for a subset of LcrH mutants to efficiently translocate Yop effectors (Edqvist and Francis, unpublished) . Importantly, it appears that as little as the first three N-terminal amino acids are sufficient for low-level YopD secretion . To our knowledge, this is the first study that has defined the secretion domain of a translocator protein . However, a precedent for a requirement for so few N-terminal amino acids in type III-dependent protein secretion has been established . An analysis of the type III secretion signal of the S . enterica InvJ protein, which is required to control the needle length of the type III needle complex, revealed that residues 4 to 7 were predominantly required to ensure efficient secretion of InvJ (50) . Whether these miniature secretion signals are universal for the type III substrates presumably secreted early during an infection, such as determinants of the outer needle and the translocators, remains a fascinating challenge for further studies .

The regulatory loops that control expression of the Yersinia plasmid-borne TTSS are innately complex (14, 58) . One essential negative regulatory network was illustrated by the finding that mutants with mutations in the LcrH chaperone that are unable to bind YopD do not have yop regulatory control (21) . We confirmed this observation by showing that YopD mutants that are unable to bind LcrH also are not able to control yop regulation . In this context, the finding that an LcrH-YopD complex, in association with LcrQ, is necessary for repressing translation of mRNA derived from type III genes is a key observation (2, 11) . However, since LcrQ does not interact with YopD or LcrH (67), the exact mechanism of this regulatory cross talk remains unresolved .

In summary, we report the results of a thorough phenotypic analysis of in-frame sequential deletions of YopD . Our results confirmed that a YopD-LcrH interaction is necessary to maintain yop regulatory control . In addition, we established that pore formation is only one step in the process of effector translocation into target cells . Furthermore, we isolated YopD mutants that are indistinguishable from the wild type with respect to yop regulatory control but are unable to induce pores in eukaryotic membranes and to microinject antihost effectors . We concluded that YopD does function in several steps leading to efficient effector translocation by virtue of mutant phenotypes identified in (i) regulation, (ii) secretion, (iii) pore formation, and (iv) translocation . Our results should facilitate analysis of the molecular mechanisms that govern the multistep process of effector translocation by TTSSs in numerous bacterial pathogens . This may even preempt the discovery of a common antibacterial therapy targeted toward this highly refined and effective virulence strategy .

 
This work was supported by grants from the Swedish Research Council (to Å.F., H.W.W., and M.S.F.), the Foundation for Medical Research at Umeå University (to M.S.F.), the Swedish Foundation for Strategic Research (to Å.F . and H.W.W.), and the J . C . Kempes Memorial Fund (to J.O., P.J.E., and J.E.B.) .

We are indebted to Maria Vedin and Stefan Lindström for skillful technical assistance, and we thank Victoria Shingler for advice concerning the chymotrypsin digestion assay .

 
* Corresponding author . Mailing address: Department of Molecular Biology, Umeå University, SE-901 87 Umeå, Sweden . Phone: 46-(0)90-7852536 . Fax: 46-(0)90-771420 . E-mail: matthew.francis@molbiol.umu.se.

J.O., P.J.E., and J.E.B . contributed equally to this work .

1. Aepfelbacher, M., and J . Heesemann. 2001 . Modulation of Rho GTPases and the actin cytoskeleton by Yersinia outer proteins (Yops) . Int . J . Med . Microbiol . 291:269-276.
2. Anderson, D . M., K . S . Ramamurthi, C . Tam, and O . Schneewind. 2002 . YopD and LcrH regulate expression of Yersinia enterocolitica YopQ by a posttranscriptional mechanism and bind to yopQ RNA . J . Bacteriol. 184:1287-1295 .
3. Anderson, D . M., and O . Schneewind. 1999 . Type III machines of Gram-negative pathogens: injecting virulence factors into host cells and more . Curr . Opin . Microbiol . 2:18-24.
4. Barbieri, J . T. 2000 . Pseudomonas aeruginosa exoenzyme S, a bifunctional type-III secreted cytotoxin . Int . J . Med . Microbiol . 290:381-387.
5. Bergman, T., S . Håkansson, Å . Forsberg, L . Norlander, A . Macellaro, A . Bäckman, I . Bölin, and H . Wolf-Watz. 1991 . Analysis of the V antigen lcrGVH-yopBD operon of Yersinia pseudotuberculosis: evidence for a regulatory role of LcrH and LcrV . J . Bacteriol . 173:1607-1616.
6. Blocker, A., D . Holden, and G . Cornelis. 2000 . Type III secretion systems: what is the translocator and what is translocated? Cell Microbiol . 2:387-390.
7. Bölin, I., and H . Wolf-Watz. 1984 . Molecular cloning of the temperature-inducible outer membrane protein 1 of Yersinia pseudotuberculosis. Infect . Immun . 43:72-78.
8. Bröms, J . E., A.-L . Forslund, Å . Forsberg, and M . S . Francis. 2003 . Dissection of homologous translocon operons reveals a distinct role for YopD in type III secretion by Yersinia pseudotuberculosis. Microbiology 149:2615-2626 .
9. Bröms, J . E., A.-L . Forslund, Å . Forsberg, and M . S . Francis. 2003 . PcrH of Pseudomonas aeruginosa is essential for secretion and assembly of the type III translocon . J . Infect . Dis . 188:1910-1922.
10. Bröms, J . E., C . Sundin, M . S . Francis, and Å . Forsberg. 2003 . Comparative analysis of type III effector translocation by Yersinia pseudotuberculosis expressing native LcrV or PcrV from Pseudomonas aeruginosa. J . Infect . Dis . 188:239-249.
11. Cambronne, E . D., and O . Schneewind. 2002 . Yersinia enterocolitica type III secretion: yscM1 and yscM2 regulate yop gene expression by a posttranscriptional mechanism that targets the 5' untranslated region of yop mRNA . J . Bacteriol . 184:5880-5893 .
12. Cornelis, G., C . Sluiters, C . L . de Rouvroit, and T . Michiels. 1989 . Homology between virF, the transcriptional activator of the Yersinia virulence regulon, and AraC, the Escherichia coli arabinose operon regulator . J . Bacteriol . 171:254-262.
13. Cornelis, G . R. 2002 . The Yersinia Ysc-Yop 'type III' weaponry . Nat . Rev . Mol . Cell Biol . 3:742-752.
14. Cornelis, G . R., A . Boland, A . P . Boyd, C . Geuijen, M . Iriarte, C . Neyt, M . P . Sory, and I . Stainier. 1998 . The virulence plasmid of Yersinia, an antihost genome . Microbiol . Mol . Biol . Rev . 62:1315-1352 .
15. Cornelis, G . R., and F . Van Gijsegem. 2000 . Assembly and function of type III secretory systems . Annu . Rev . Microbiol . 54:735-774.
16. Delahay, R . M., and G . Frankel. 2002 . Coiled-coil proteins associated with type III secretion systems: a versatile domain revisited . Mol . Microbiol . 45:905-916.
17. Fällman, M., F . Deleuil, and K . McGee. 2002 . Resistance to phagocytosis by Yersinia. Int . J . Med . Microbiol . 291:501-509.
18. Feldman, M . F., S . Muller, E . Wuest, and G . R . Cornelis. 2002 . SycE allows secretion of YopE-DHFR hybrids by the Yersinia enterocolitica type III Ysc system . Mol . Microbiol . 46:1183-1197.
19. Forsberg, Å., and H . Wolf-Watz. 1990 . Genetic analysis of the yopE region of Yersinia spp.: identification of a novel conserved locus, yerA, regulating yopE expression . J . Bacteriol. 172:1547-1555.
20. Francis, M . S., M . Aili, M . L . Wiklund, and H . Wolf-Watz. 2000 . A study of the YopD-LcrH interaction from Yersinia pseudotuberculosis reveals a role for hydrophobic residues within the amphipathic domain of YopD . Mol . Microbiol . 38:85-102.
21. Francis, M . S., S . A . Lloyd, and H . Wolf-Watz. 2001 . The type III secretion chaperone LcrH co-operates with YopD to establish a negative, regulatory loop for control of Yop synthesis in Yersinia pseudotuberculosis. Mol . Microbiol . 42:1075-1093.
22. Francis, M . S., and H . Wolf-Watz. 1998 . YopD of Yersinia pseudotuberculosis is translocated into the cytosol of HeLa epithelial cells: evidence of a structural domain necessary for translocation . Mol . Microbiol . 29:799-813.
23. Gemski, P., J . R . Lazere, and T . Casey. 1980 . Plasmid associated with pathogenicity and calcium dependency of Yersinia enterocolitica. Infect . Immun . 27:682-685.
24. Håkansson, S. 1995 . YopB-mediated translocation of Yops and the YpkA Ser/Thr kinase of Yersinia. Ph.D . thesis . Umeå University, Umeå, Sweden.
25. Håkansson, S., T . Bergman, J . C . Vanooteghem, G . Cornelis, and H . Wolf-Watz. 1993 . YopB and YopD constitute a novel class of Yersinia Yop proteins . Infect . Immun . 61:71-80.
26. Håkansson, S., K . Schesser, C . Persson, E . E . Galyov, R . Rosqvist, F . Homble, and H . Wolf-Watz. 1996 . The YopB protein of Yersinia pseudotuberculosis is essential for the translocation of Yop effector proteins across the target cell plasma membrane and displays a contact-dependent membrane disrupting activity . EMBO J . 15:5812-5823.
27. Hanahan, D. 1983 . Studies on transformation of Escherichia coli with plasmids . J . Mol . Biol . 166:557-580.
28. Hartland, E . L., S . P . Green, W . A . Phillips, and R . M . Robins-Browne. 1994 . Essential role of YopD in inhibition of the respiratory burst of macrophages by Yersinia enterocolitica. Infect . Immun . 62:4445-4453.
29. Hartland, E . L., and R . M . Robins-Browne. 1998 . In vitro association between the virulence proteins, YopD and YopE, of Yersinia enterocolitica. FEMS Microbiol . Lett . 162:207-213.
30. Heesemann, J., B . Algermissen, and R . Laufs. 1984 . Genetically manipulated virulence of Yersinia enterocolitica. Infect . Immun . 46:105-110.
31. Henriksson, M . L., M . S . Francis, A . Peden, M . Aili, K . Stefansson, R . Palmer, A . Aitken, and B . Hallberg. 2002 . A nonphosphorylated 14-3-3 binding motif on exoenzyme S that is functional in vivo . Eur . J . Biochem . 269:4921-4929 .
32. Holmström, A., J . Olsson, P . Cherepanov, E . Maier, R . Nordfelth, J . Pettersson, R . Benz, H . Wolf-Watz, and Å . Forsberg. 2001 . LcrV is a channel size-determining component of the Yop effector translocon of Yersinia. Mol . Microbiol 39:620-632.
33. Holmström, A., J . Pettersson, R . Rosqvist, S . Håkansson, F . Tafazoli, M . Fällman, K . E . Magnusson, H . Wolf-Watz, and Å . Forsberg. 1997 . YopK of Yersinia pseudotuberculosis controls translocation of Yop effectors across the eukaryotic cell membrane . Mol . Microbiol . 24:73-91.
34. Holmström, A., R . Rosqvist, H . Wolf-Watz, and Å . Forsberg. 1995 . Virulence plasmid-encoded YopK is essential for Yersinia pseudotuberculosis to cause systemic infection in mice . Infect . Immun . 63:2269-2276.
35. Horton, R . M., and L . R . Pease. 1991 . Recombination and mutagenesis of DNA sequences using PCR, p . 217-247 . In M . J . McPherson (ed.), Directed mutagenesis: a practical approach . Oxford University Press, New York, N.Y.
36. Hueck, C . J. 1998 . Type III protein secretion systems in bacterial pathogens of animals and plants . Microbiol . Mol . Biol . Rev . 62:379-433 .
37. Lavander, M., L . Sundberg, P . J . Edqvist, S . A . Lloyd, H . Wolf-Watz, and Å . Forsberg. 2002 . Proteolytic cleavage of the FlhB homologue YscU of Yersinia pseudotuberculosis is essential for bacterial survival but not for type III secretion . J . Bacteriol 184:4500-4509 .
38. Lupas, A. 1996 . Coiled coils: new structures and new functions . Trends Biochem . Sci . 21:375-382.
39. Michiels, T., P . Wattiau, R . Brasseur, J . M . Ruysschaert, and G . Cornelis. 1990 . Secretion of Yop proteins by yersiniae . Infect . Immun . 58:2840-2849.
40. Milton, D . L., R . O'Toole, P . Horstedt, and H . Wolf-Watz. 1996 . Flagellin A is essential for the virulence of Vibrio anguillarum. J . Bacteriol . 178:1310-1319.
41. Neyt, C., and G . R . Cornelis. 1999 . Insertion of a Yop translocation pore into the macrophage plasma membrane by Yersinia enterocolitica: requirement for translocators YopB and YopD, but not LcrG . Mol . Microbiol . 33:971-981.
42. Neyt, C., and G . R . Cornelis. 1999 . Role of SycD, the chaperone of the Yersinia Yop translocators YopB and YopD . Mol . Microbiol . 31:143-156.
43. Nhieu, G . T., and P . J . Sansonetti. 1999 . Mechanism of Shigella entry into epithelial cells . Curr . Opin . Microbiol . 2:51-55.
44. Pallen, M . J., G . Dougan, and G . Frankel. 1997 . Coiled-coil domains in proteins secreted by type III secretion systems . Mol . Microbiol . 25:423-425.
45. Pettersson, J., R . Nordfelth, E . Dubinina, T . Bergman, M . Gustafsson, K . E . Magnusson, and H . Wolf-Watz. 1996 . Modulation of virulence factor expression by pathogen target cell contact . Science 273:1231-1233.
46. Portnoy, D . A., H . F . Blank, D . T . Kingsbury, and S . Falkow. 1983 . Genetic analysis of essential plasmid determinants of pathogenicity in Yersinia pestis. J . Infect . Dis . 148:297-304.
47. Rimpiläinen, M., Å . Forsberg, and H . Wolf-Watz. 1992 . A novel protein, LcrQ, involved in the low-calcium response of Yersinia pseudotuberculosis shows extensive homology to YopH . J . Bacteriol . 174:3355-3363.
48. Rosqvist, R., Å . Forsberg, and H . Wolf-Watz. 1991 . Intracellular targeting of the Yersinia YopE cytotoxin in mammalian cells induces actin microfilament disruption . Infect . Immun . 59:4562-4569.
49. Rosqvist, R., K . E . Magnusson, and H . Wolf-Watz. 1994 . Target cell contact triggers expression and polarized transfer of Yersinia YopE cytotoxin into mammalian cells . EMBO J . 13:964-972.
50. Russmann, H., T