Competition between the Yops of Yersinia enterocolitica for Delivery into Eukaryotic Cells: Role of the SycE Chaperone Binding Domain of YopE

ABSTRACT A type III secretion-translocation system allowsYersinia adhering at the surface of animal cells to deliver a cocktail of effector Yops (YopH, -O, -P, -E, -M, and -T) into the cytosol of these cells. Residues or codons 1 to 77 contain all the information required for the complete delivery of YopE into the target cell (release from the bacterium and translocation across the eukaryotic cell membrane). Residues or codons 1 to 15 are sufficient for release from the wild-type bacterium under Ca2+-chelating conditions but not for delivery into target cells. Residues 15 to 50 comprise the binding domain for SycE, a chaperone specific for YopE that is necessary for release and translocation of full-length YopE. To understand the role of this chaperone, we studied the delivery of YopE-Cya reporter proteins and YopE deletants by polymutant Yersinia devoid of most of the Yop effectors (ΔHOPEM and ΔTHE strains). We first tested YopE-Cya hybrid proteins and YopE proteins deleted of the SycE-binding site. In contrast to wild-type strains, these mutants delivered YopE15-Cya as efficiently as YopE130-Cya. They were also able to deliver YopEΔ17–77. SycE was dispensable for these deliveries. These results show that residues or codons 1 to 15 are sufficient for delivery into eukaryotic cells and that there is no specific translocation signal in Yops. However, the fact that the SycE-binding site and SycE were necessary for delivery of YopE by wild-type Yersinia suggests that they could introduce hierarchy among the effectors to be delivered. We then tested a YopE-Cya hybrid and YopE proteins deleted of amino acids 2 to 15 but containing the SycE-binding domain. These constructs were neither released in vitro upon Ca2+ chelation nor delivered into cells by wild-type or polymutant bacteria, casting doubts on the hypothesis that SycE could be a secretion pilot. Finally, it appeared that residues 50 to 77 are inhibitory to YopE release and that binding of SycE overcomes this inhibitory effect. Removal of this domain allowed in vitro release and delivery in cells in the absence as well as in the presence of SycE.

the mRNA, so that the Yops are cotranslationally secreted from the bacteria (1,2). In the case of YopE, the first 15 codons or amino acids constitute this 5Ј secretion signal. In addition, efficient secretion of some Yops requires the assistance of individual cytosolic chaperones, called Sycs (32,33). These chaperones are small acidic proteins that possess a leucine repeat in their C-terminal moiety. SycE, the chaperone of YopE, binds amino acids 15 to 50 (the chaperone binding domain) of YopE (27,35), and it prevents the intrabacterial degradation of this Yop (5,10). The chaperone binding domain is not required for secretion of YopE fusion proteins by the 5Ј secretion signal (5,27,28). Moreover, in the absence of this chaperone binding domain, SycE becomes dispensable for secretion of YopE, suggesting that it is the presence of the chaperone binding domain that creates the need for the chaperone (35). However, data have been presented to show that hybrid YopE-neomycin phosphotransferase (designated YopE-Npt) proteins lacking the first 5Ј secretion signal are still secreted by the Ysc apparatus, suggesting that YopE could contain a second secretion signal (5). This proposed second secretion signal is localized to the site of the chaperone binding domain and, correspondingly, it is only operational in the presence of SycE (5).
Translocation of the effector Yops across the eukaryotic cell membrane was shown by several laboratories (4,12,20,23) to be dependent on YopB and YopD, two other proteins exported by the bacterium, but this view has recently been questioned (15). Translocation of effector Yops can be demonstrated by several methods. A classical approach makes use of a calmodulin-dependent adenylate cyclase (Cya) reporter strategy (29). Translocation of Yop effectors can also be demonstrated by fractionation of the infected cell culture or by indirect immunofluorescence and confocal scanning laser microscopy (23). Demonstration of the translocation turned out to be more difficult with some Yops than with others, and it has been observed that translocation can be improved if expression of the other Yop effectors is abolished (4,11). This could be due to a decrease in competition between the different Yop effectors for the secretion and translocation machineries. Strains of Y. enterocolitica that carry multiple yop mutations are thus sensitive tools for studying the translocation of Yop effectors.
It has been shown previously that a Cya reporter protein fused to just the first 15 amino acids of YopE (YopE 15 -Cya) can be released by wild-type (wt) bacteria upon Ca 2ϩ chelation; however, this fusion protein is not delivered into eukaryotic cells. Indeed, at least the first 50 amino acids are required for the reporter protein to be translocated into eukaryotic cells by wt bacteria (27,28). Therefore, amino acids 15 to 50, which are the residues that bind the SycE chaperone and which constitute the proposed second secretion signal, were thought to be a translocation domain (27,28), although they are not sufficient, in the absence of the 5Ј secretion signal, to direct delivery of YopE by wt bacteria into eukaryotic cells (14).
In this study, we investigated the requirement for the two proposed secretion signals for delivery of YopE into eukaryotic cells. We confirmed that SycE and residues 15 to 50 of YopE are required for delivery of YopE by wt bacteria, but we observed that they are dispensable for delivery by a multimutant strain. This suggests that SycE could be a hierarchy factor for YopE delivery. Moreover, we identified a secretion-inhibitory domain between residues 50 and 77.

MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions. Parental wt strain Y. enterocolitica MRS40(pYV40) is an ampicillin-sensitive derivative of serotype O:9 clinical isolate E40(pYV40) (25,29). Escherichia coli LK111, XL-1 Blue, and BL21(DE3) were used for plasmid construction and protein expression. E. coli SM10pir ϩ was used to conjugate plasmids into Y. enterocolitica. The full list of plasmids used in this study is given in Table 1. Bacterial strains were routinely grown in tryptic soy broth and plated on tryptic soy agar. For in vitro induction of the yop genes, Y. enterocolitica was grown in brain heart infusion (BHI), supplemented with 20 mM sodium oxalate, 20 mM MgCl 2 , and 0.4% (wt/vol) glucose (BHI-Ox). Yop induction under minimal-medium conditions was performed as described previously (5). Selective agents were used at the following concentrations: ampicillin, 200 g ⅐ ml Ϫ1 ; chloramphenicol, 10 g ⅐ ml Ϫ1 ; nalidixic acid, 35 g ⅐ ml Ϫ1 ; streptomycin, 100 g ⅐ ml Ϫ1 ; sucrose, 5% (wt/vol); and arsenite, 0.4 mM.
Molecular biology techniques. Molecular biology techniques were essentially performed as previously described (24). All chemicals were obtained from Sigma unless stated otherwise. Yops were precipitated from culture supernatants by ammonium sulfate (0.5 g ⅐ ml Ϫ1 ) (9), analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and, where appropriate, transferred to nitrocellulose membranes. Immunoblots were developed with secondary antibody conjugated to horseradish peroxidase (HRP) and Supersignal (NEN) as a chemiluminescent substrate. For detection of Cya fusion proteins on nitrocellulose membranes, biotinylated calmodulin (Calbiochem) was used according to the supplier's instructions, except that streptavidin-biotinylated HRP complex (Amersham Life Science) was substituted for streptavidin-alkaline phosphatase, and Supersignal was used as the chemiluminescent substrate.
Mutator plasmids were introduced into Y. enterocolitica strains by conjugation from E. coli SM10pir ϩ . After allelic exchange, the mutations were confirmed by PCR analysis or sequencing and, where appropriate, by Yop induction and SDS-PAGE analysis of the secreted proteins or by Western blot analysis of the bacterial proteins.
Plasmid constructions. (i) pAPBD18. The 2.4-kb BamHI-ClaI fragments of pMSLE15 (containing yopE 15 -cya and sycE) were cloned into the corresponding sites of pBluescript KS(Ϫ) to create pAPBD16. pAPBD16 was digested with EcoRI, the ends were filled in with Klenow, and the DNA was religated to create pAPBD17. The 2.4-kb BamHI-ClaI fragments of pAPBD17 (containing yopE 15 cya and sycE 54 ) were cloned into the corresponding sites of pTM100 to create pAPBD18.
(iii) pYOB2. The 2.2-kb yopO gene was amplified by PCR with oligonucleotides MIPA 471 (GCATGAACATATGGGAACTA) and MIPA 473 (TATAT CAAATGCATGGCTTAGGG) using pBC5 as a template. After digestion with NdeI (underlined) and NsiI (underlined), the DNA fragment was cloned into the NdeI and PstI sites of pCNR26 to give plasmid pYOB2. This construction places the second start codon of yopO at the NdeI site.
(v) pAPB24. SycE was amplified by PCR with oligonucleotides MIPA 599 (CGGGATCCTATTCATTTGAACAAGCTA), which is complementary to nucleotides 4 to 22 of sycE, and MIPA 521 (CTCAAGCTTCTACTCAACTAAA TGACCG), which is identical to nucleotides 379 to 393 of sycE and four additional bases with pAPBD16 as a template (introduced restriction sites are underlined). The 400-bp product was digested with BamHI and HindIII and cloned into the corresponding sites of pQE-30 to create pAPB24.
Yop translocation assay. The PU5-1.8 mouse monocyte/macrophage cell line (ATCC TIB-61) used in these studies was grown in RPMI 1640 medium (Gibco BRL) supplemented with 10% (wt/vol) fetal bovine serum, 2 mM L-glutamine, and streptomycin, 100 g ⅐ ml Ϫ1 . Translocation assays were carried out essentially as described by Sory and Cornelis (29). Cells were seeded into 24-well tissue culture plates at a density of 5 ϫ 10 5 cells per ml of medium per well and allowed to adhere for 20 h. Before infection with Y. enterocolitica, cells were washed and covered with RPMI 1640 supplemented only with 2 mM L-glutamine. Cytochalasin D was added 30 min before infection, at a final concentration of 5 g ml Ϫ1 (stock solution, 2 mg ml Ϫ1 in dimethyl sulfoxide). Cytochalasin D is not toxic to Y. enterocolitica at this concentration (29). A freshly isolated transconjugant colony of Y. enterocolitica was cultured overnight at 22°C and diluted the next day to an optical density at 600 nm of 0.2 in 5 ml of BHI medium. After being grown with shaking at 22°C for 2 h, bacteria were washed and suspended in saline. Samples of 100 l, containing about 10 7 bacteria (multiplicity of infection, 20:1), were added to the monolayer, and the infected cultures were incubated at 37°C for 2 h in a 6% CO 2 atmosphere. Cells were washed and then lysed under denaturing conditions (100°C for 5 min in 50 mM HCl and 0.1% [wt/vol] Triton X-100). The lysate was neutralized by NaOH, and cyclic AMP (cAMP) was extracted with ethanol. After centrifugation, the supernatant was dried, and cAMP was assayed by an enzyme immunoassay (Biotrak Amersham). All experiments were performed three times.
Cytotoxicity assay. The HeLa human epithelial cell line (ATCC CCL-2) used in these studies was grown in RPMI 1640 supplemented with 10% (wt/vol) fetal bovine serum, 2 mM L-glutamine, and streptomycin (100 g ⅐ ml Ϫ1 ) and prepared similarly to the PU5-1.8 macrophages as described above except that the HeLa cells were seeded at a density of 7 ϫ 10 4 cells ⅐ ml Ϫ1 . Bacteria were pregrown as described above, and cells were infected with Y. enterocolitica at a multiplicity of infection of 70. Two to three hours after infection, the morphology of the cells was observed by phase-contrast microscopy. The cells became rounded as a result of cytotoxicity.
Staining of actin filaments with phalloidin. Rat I fibroblasts grown on coverslips were infected with the different strains. After 2.5 h of infection, the cells were fixed in 2% (wt/vol) paraformaldehyde for 20 min. After being washed with phosphate-buffered saline (PBS) (136 mM NaCl, 2.7 mM KCl, 10.1 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 [pH 7.4]), membranes were permeabilized with 0.5% (wt/vol) Triton X-100 in PBS for 10 min. Cells were then incubated for 40 min at 37°C with fluorescein isothiocyanate-conjugated phalloidin. Samples were mounted on Mowiol and examined by fluorescence microscopy.
SycE-binding assays. Native SycE was produced and purified as described in reference 33. His 6 -SycE was produced in E. coli XL-1 Blue(pAPB24) and purified on a His-Trap column by elution with 300 mM imidazole according to the manufacturer's instructions (Pharmacia Biotech). Total cell proteins of Y. enterocolitica were separated by SDS-PAGE and transferred to nitrocellulose membranes. After blockage in PBST plus BSA (PBS plus 0.1% Tween 20 plus 0.5% bovine serum albumin [BSA]), the membrane was incubated with His 6 -SycE (0.5 g ⅐ ml Ϫ1 ) in PBST plus BSA for 2 h at room temperature. Bound SycE or His 6 -SycE was revealed with anti-SycE or anti-His antibody (Pharmacia Biotech), respectively, followed by HRP-conjugated secondary antibody and chemiluminescence detection.

Translocation of YopE 15 -Cya into eukaryotic cells by Yop effector polymutant Y. enterocolitica.
It has been previously demonstrated that wt bacteria deliver YopE 130 -Cya, but not YopE 15 -Cya, into eukaryotic cells, suggesting that the 5Ј se-  (28) were delivered into eukaryotic cells by ⌬HOPEM bacteria (data not shown). To confirm that delivery of YopE 15 -Cya into eukaryotic cells by ⌬HOPEM bacteria was due to the type III secretion-translocation system, the delivery of YopE 15 -Cya by ⌬HOPEMYscN bacteria (secretion deficient) and ⌬HOPEMYopB and ⌬HOPEMYopD bacteria (both translocation deficient) was tested ( Table 2). YopE 15 -Cya was not translocated by these strains, confirming that YopE 15 -Cya was indeed delivered into eukaryotic cells by the type III injectisome (Table 2). To assess the necessity for the 5Ј secretion signal, delivery of Cya fused to the first two amino acids of  These results show that the translocation system of ⌬HOPEM bacteria is still specific for the Yops. In conclusion, translocation of YopE-Cya hybrids is possible without the previously described translocation domain, which comprises amino acids 15 to 50, but not without the 5Ј secretion signal (residues or codons 1 to 15 and upstream RNA). Production and secretion levels of YopE 15 -Cya by wt and ⌬HOPEM bacteria. It was next investigated whether the delivery of YopE 15 -Cya by ⌬HOPEM, but not wt Y. enterocolitica, could result from differences between strains in the production and secretion of this hybrid protein. Protein levels were analyzed following growth of the bacteria in Ca 2ϩ -chelating conditions, which induce Yop production and release. No differences were seen between the two strains in the levels of YopE 15 -Cya associated with the bacteria or released into the extracellular medium ( Fig. 1A and Fig. 1B, lanes 5 and 6). Secretion of YopE 15 -Cya by both strains was strictly dependent on the Ysc system, since no secretion was observed in a yscN background (Fig. 1C). Although Yop release upon Ca 2ϩ chelation may not necessarily reflect exactly what occurs upon contact of Y. enterocolitica with eukaryotic cells, these data do show that synthesis of the two fusion proteins and their passage through the bacterial membranes was equally efficient in the two strains and equally dependent on Ysc. The only difference between the two strains with regard to YopE 15 -Cya was thus the level of translocation of this protein into eukaryotic cells (Table 2). This suggests that the presence of additional Yops in the wild type directly reduces translocation of YopE 15 -Cya and that in order to enter into eukaryotic cells the Yops must thus compete with one another for passage through the secretiontranslocation machinery.

Influence of SycE on secretion and translocation of YopE 15 -Cya.
In order to investigate the requirement for SycE for translocation of YopE 15 -Cya and YopE 130 -Cya into eukaryotic cells, an sycE mutation was introduced into the wt and ⌬HOPEM strains ( Table 1). The experiment was carried out with plasmids pMS111 and pMSLE15, which encode SycE along with YopE 130 -Cya and YopE 15 -Cya, respectively, and with plasmids pMSL30 and pAPBD18, which encode only the YopE-Cya fusion proteins ( Table 1). The presence or absence of SycE did not affect the steady-state levels of YopE 130 -Cya or YopE 15 -Cya associated with the wt or ⌬HOPEM bacteria when grown under BHI-Ox conditions (Fig. 1A). However, SycE was required for efficient secretion and translocation of YopE 130 -Cya into eukaryotic cells not only by wild-type but also by ⌬HOPEM bacteria (Fig. 1B, compare lanes 1 and 3 and lanes 2 and 4; Table 2). In contrast, the presence or absence of SycE did not influence secretion or translocation of YopE 15 -Cya into eukaryotic cells by ⌬HOPEM bacteria (Fig. 1B, compare lanes 5 and 7 and lanes 6 and 8;  (Fig. 2), and we checked this removal by a SycE overlay experiment (33). Purified SycE or His 6 -SycE bound YopE but failed to bind YopE ⌬17-77 , verifying that the chaperone binding domain had been deleted from the latter protein ( Fig. 3B and D). ⌬HOPEM bacteria could not be used for cytotoxicity experiments because they still produce the YopT cytotoxin (8). We thus turned to ⌬THE bacteria (Table 1), which do not induce any morphological changes in eukaryotic cells (Fig. 4). YopE ⌬17-77 was produced and released by ⌬THE bacteria (Fig. 3A and 3C), and release of YopE ⌬17-77 did not occur in an yscN background (Fig. 5), confirming that this release was type III dependent. Delivery was then assayed by monitoring the rounding up of HeLa epithelial cells and by staining the actin of Rat-I cells. ⌬THE Y. enterocolitica strains producing YopE or YopE ⌬17-77 were cytotoxic for HeLa epithelial cells (results not shown) and Rat I fibroblasts (Fig. 4), while ⌬THEB bacteria producing YopE ⌬17-77 were not cytotoxic, indicating that translocation of YopE ⌬17-77 was YopB dependent. This result confirmed that the first 16 amino acids of YopE are sufficient for delivery into Competition could play an important role in determining the level of translocation. YopE 15 -Cya was delivered into eukaryotic cells by ⌬HOPEM bacteria, but not by wt bacteria, suggesting that competition between the Yops is an important determinant for secretion and translocation and that the chaperone binding domain plays a significant role with regards to this competition. To investigate this theory, the ability of ⌬HOPEM bacteria to deliver YopE 15 -Cya (encoded by pMSLE15) into eukaryotic cells when overproducing another Yop effector in trans was tested. Therefore, the translocation of YopE 15 -Cya into eukaryotic cells by ⌬HOPEM bacteria overproducing YopH, YopO, YopP, YopE, or YopM was measured. ⌬HOPEM(pMSLE15)(pBC18R) served as a vector control. In each case, the rate of translocation of YopE 15 -Cya into eukaryotic cells was lower than that of ⌬HOPEM bacteria not overexpressing one of these Yop effectors in trans (Table 3). In contrast, translocation into eukaryotic cells of YopE 130 -Cya by ⌬HOPEM was unaffected by overproducing another Yop in trans. The strongest effect on delivery of YopE 15 -Cya was observed with YopE and YopH ( Table 3). As a control, we checked the profile of proteins released by these strains upon Ca 2ϩ chelation. This control (data not shown) confirmed the overproduction of the Yops encoded in trans. Unfortunately, it also showed a concomitant reduction in the release of the Cya reporter and of the translocators LcrV, YopB, and YopD, indicating that the previous results must be interpreted with caution. To circumvent these difficulties, presumably linked to titration, we tested whether YopE 15  lack of YopE alone significantly increased delivery of YopE 15 -Cya into eukaryotic cells. These results are consistent with the idea that amino acids 15 to 50 promote translocation of YopE by wt bacteria by assisting YopE to compete with other Yops for the secretion-translocation apparatus. If this was so, one would expect that YopE deprived of its chaperone binding domain (YopE ⌬17-77 ) would not compete with YopE 15 -Cya for delivery into eukaryotic cells. We thus overproduced YopE ⌬17-77 in trans, and we monitored translocation of YopE 15 -Cya. As expected, overproduction of YopE ⌬17-77 did not inhibit translocation of YopE 15 -Cya (Table 3). Thus, amino acids 15 to 50 of YopE, in conjunction with SycE, seem to give YopE a competitive advantage over the other Yops for the secretiontranslocation process.
To check this hypothesis, we removed residues 50 to 77 from YopE, and we monitored in vitro release of YopE in the presence and in the absence of SycE. As expected, it was released equally as well as YopE ⌬17-77 , and this release was independent of SycE. This contrasted with wt YopE, which was only released in the presence of SycE (Fig. 5). Thus, amino acids 50 to 77 of YopE inhibit secretion of YopE in the absence of SycE. Interestingly, although this construct does not need SycE  for secretion, it still binds SycE. Thus the secretion-inhibitory domain is distinct from the minimal SycE-binding domain, although this secretion-inhibitory domain must be somehow covered by SycE.

DISCUSSION
In this paper, we have analyzed the N-terminal domain of Y. enterocolitica YopE in order to clarify its roles in the in vitro release of YopE and its delivery into eukaryotic cells.
The results, summarized in Fig. 7, confirm previous data in showing that residues 1 to 50 of YopE are required for delivery of YopE into eukaryotic cells by wt Y. enterocolitica (27,28). However, the current results also show that delivery of YopE by Yop effector multimutant bacteria does not require amino acids 15 to 50 but rather that the secretion signal encompassing amino acids or codons 1 to 15 is sufficient. This implies that the chaperone binding domain does not need to interact with the Yop translocators for Yop effector translocation. In addition, this suggests that any protein that can be released by the Ysc secretion machinery also has the capacity to be delivered into eukaryotic cells. This conclusion hence implies a continuity between the secretion and translocation apparatuses, so that a Yop can pass through the secretion channel, syringe and needle, and then directly through the translocation apparatus into the target cell.
The requirement for amino acids 15 to 50 for translocation of YopE into eukaryotic cells by wt Y. enterocolitica, but not by Yop effector multimutant bacteria, implies that these amino acids give YopE a competitive advantage over the other Yops for the Ysc secretion-translocation apparatus. Due to the competition, only the Yops that are avidly recognized by the Ysc apparatus would be successfully delivered inside eukaryotic cells. Competition between the Yops could determine the order of precedence of Yop entry into eukaryotic cells and/or the relative quantities of each Yop delivered inside a cell.
However, if there is continuity between secretion and translocation, how could one explain that domains 15 to 50 of YopE are required for translocation by wt bacteria but not for release of YopE under Ca 2ϩ -chelating conditions? This could be due to differences in the structure of the Ysc apparatus when opening is caused by Ca 2ϩ chelation and when opening is triggered by contact with eukaryotic cells. It is possible that Ca 2ϩ chelation shears the external part of the Ysc apparatus, resulting . YopE 15 -X, containing the N-terminal 5Ј secretion signal but lacking the chaperone binding site, is prevented from entering eukaryotic cells (panel 1) but is nevertheless released under low-Ca 2ϩ conditions (4). We hypothesize that competition is stronger for delivery into cells (small channel) than for release under low-Ca 2ϩ conditions (large channel). In sycE mutant bacteria, the lack of SycE does not affect the pathway followed by YopE 15 -X (panels 2 and 5). However, full-length YopE is neither delivered into cells nor released under low-Ca 2ϩ conditions. Removal of the domain encompassing amino acids 50 to 77 (not shown in this figure) allows YopE to be released independently of SycE. We conclude that this domain is inhibitory for release and that this inhibition is prevented by SycE. In ⌬HOPEM sycE strains (panels 3 and 6), YopE 15 -X is not only released under low-Ca 2ϩ conditions but also delivered into cells. This indicates that the N-terminal 5Ј secretion signal is sufficient for delivery into cells. YopE and YopE 15 -X are partially degraded when blocked inside bacteria. This representation is based on the results presented in this paper and on previous results which are cited in the text. in a secretion channel (syringe) on the surface of the bacteria with an inner diameter that is much wider than that of the channel (needle) bridging the bacteria and the eukaryotic cell (Fig. 7). In support of this hypothesis, Ca 2ϩ chelation leads to the release of some external parts of the Ysc apparatus, such as YscP (19,30). Thus, passage through the secretion channel under Ca 2ϩ -chelating conditions would be far more abundant and far more permissive than upon bacteria-eukaryotic cell interaction.
In agreement with the observations of Lee et al. (14), domain encompassing amino acids 15 to 50 was not sufficient to direct YopE to the eukaryotic cytosol (14). However, unlike previous data (5), release of YopE to the extracellular milieu by this domain could not be detected, despite the use of various gene constructions, protein systems, and growth conditions. Although the same ϩ1 frame-shift mutation of codons 2 to 15 was used here as that employed by Cheng et al. (5), in the present work the mutation was inserted in yopE and yopE 130cya, while Cheng et al. (5,14) tested yopE-npt hybrids. This difference in protein backbone may explain the disparity of our results. In conclusion, the domain encompassing amino acids 15 to 50 is a secretion-translocation enhancer signal that is required for efficient delivery of YopE into eukaryotic cells by wt Y. enterocolitica, but it can not be considered as a physiological secretion signal.
Our results indicate that SycE plays a role as a factor introducing a hierarchical order in effector delivery, by abetting YopE to compete with the other Yops. This role should not be considered as exclusive, as SycE is required in addition when YopE contains amino acids 50 to 77. Indeed, the presence of this domain creates a need for the chaperone. This fits with older observations that bacteria missing SycE are unable to efficiently release or deliver full-length YopE or YopE 130 -Cya but are able to secrete YopE 40 -Cya (35). According to our previous observations, we suggested that it was the Syc-binding domain (residues 15 to 50) that created the need for the chaperone. The more refined present observations indicate that the secretion-inhibitory domain is localized immediately downstream of the minimal domain needed for Syc binding. Although residues 50 to 77 are neither sufficient nor necessary for SycE binding, they are likely to be covered by SycE. The determination of the three-dimensional structure of the YopE-SycE complex will clarify this.
The reason why residues 50 to 77 of YopE interfere with secretion of YopE is not clear. These amino acids could interfere with secretion through the Ysc machinery and/or they could affect the stability or solubility of YopE. Recently, Cheng et al. (6) have shown that SycE fused to glutathione S-transferase was unable to complement ⌬SycE bacteria for delivery of YopE into eukaryotic cells, even though the SycE hybrid protein bound YopE in the bacterial cytosol and stabilized this Yop (6). These experiments support the results presented here, as they show that in addition to stabilizing YopE in the bacterial cytosol, SycE is also required for efficient Yop translocation by wt bacteria. It seems that glutathione S-transferase-SycE fusion proteins do not have this secondary function. In conclusion, the data presented in this paper present a more-complete picture of the functions of the N-terminal domains of YopE for secretion and translocation of this protein.
Amino acids or codons 1 to 15 (secretion domain) are sufficient and absolutely necessary to direct translocation of YopE into eukaryotic cells by Yop effector multimutant Y. enterocolitica. Amino acids 15 to 50 bind the SycE chaperone and aid YopE to compete with the other Yops for entry into eukaryotic cells via the secretion-translocation machinery. Finally, amino acids 49 to 77 are inhibitory to YopE secretion, and this inhibition is overcome by binding of SycE to amino acids 15 to 50. Future crystallography studies of YopE alone and in complex with SycE will be very beneficial to the further studies of these domains, as would detailed studies on the other Yop-Syc interactions. It will be of great interest to investigate whether these other combinations have properties similar to those of YopE and SycE described here.