Ó Springer-Verlag 1999
Med Microbiol Immunol (1999) 188: 151±159
ORIGINAL INVESTIGATION
Andreas Roggenkamp á Lorenz Leitritz á Andreas Sing Volkhard A. J. Kempf á Kerstin Baus JuÈrgen Heesemann
Anti-recombinant V antigen serum promotes uptake of Yersinia enterocolitica serotype O8 by macrophages Received: 29 July 1999
Abstract Phagocytosis resistance even in the presence of opsonizing antibodies is a key feature of pathogenic Yersinia spp. Nevertheless, antibodies against the secreted V antigen and the outer membrane protein YadA are known to mediate protection against Y. enterocolitica serotype O8 in a mouse model with intravenous infection. To investigate the impact of anti-V antigen serum on the interaction of Y. enterocolitica and phagocytic cells, gentamicin kill assays and immuno¯uorescence staining were performed. In contrast to anti-YadA, the presence of V antigen-speci®c antibodies resulted in an increased uptake of yersiniae by macrophages. The inhibition of phagocytosis by cytochalasin D suppressed the anti-V antigen-mediated uptake. The uptake-promoting eect of anti-V antigen was more distinct for macrophages than for polymorphonuclear leukocytes. The ®ndings of the passive immunization experiments using an orogastric infection model were in agreement with those of cell-culture experiments. In the ®rst 3 days of infection both antisera exhibit no protective eect on the multiplication of the bacteria in the Peyer's patches. Only mice passively immunized with anti-V antigen survived lethal oral infections with Y. enterocolitica serotype O8. Taken together, the results support the assumption that V antigen might be part of the translocation apparatus and that anti-V antigen inhibits the Yop translocation. In addition, antisera against in-frame-deleted recombinant V antigen were generated. Protection experiments using these antisera suggested that the type-speci®c region (amino Declaration: The experiments performed in this study comply with the current law of Germany. A. Roggenkamp (&) á L. Leitritz á A. Sing á V. A. J. Kempf á K. Baus á J. Heesemann Max von Pettenkofer-Institute for Hygiene and Microbiology, Ludwig Maximilians University Munich, Pettenkoferstrasse 9a, D-80336 Munich, Germany e-mail:
[email protected] Tel.: +49-89-51605200; Fax: +49-89-5380584
acids 225±232) of the V antigen might not be a protective epitope. Key words V antigen á Yersinia enterocolitica á Phagocytosis-resistance á Passive immunization
Introduction The ability of pathogenic Yersinia spp. (Y. enterocolitica, Y. pseudotuberculosis, Y. pestis) to infect non-immune hosts and to multiply extracellularly in the host tissue is dependent on a plasmid-encoded virulence apparatus called Yop virulon. This apparatus enables bacteria adhering at the surface of eukaryotic cells to inject eector proteins (yersinia outer proteins, Yops) into the cytosol of these cells [5]. It is composed of (i) a type III secretion system called Ysc responsible for secretion of Yop proteins to the bacterial surface [15, 23], (ii) a set of toxic Yop proteins (YopE, YopH, YopM, YopO/YpkA, YopP/J, YopT) able to derange cellular signaling necessary for phagocytosis and mobilizing of an eective immune response [5, 9, 16, 43], and (iii) a translocation system encoded in the lcrGVHYopBD operon which is required for the targeting process across the eukaryotic cell membrane [4, 11, 30, 38, 41]. The Yop virulon is controlled by positive and negative regulatory circuits. The induction of Yop virulon expression mediated by the transcription activator VirF at 37 °C [6, 20] is counteracted by the presence of negative regulator proteins, especially LcrQ (YscM1 and YscM2 in Y. enterocolitica) [31, 47]. Export of negative regulator proteins results in full expression of the Yop virulon and links Yop secretion/translocation and Yop virulon regulation. LcrQ requires the presence of YopD for the down-regulatory function [50]. The activation of the Yop-secretion/translocation event involves the VirA operon (at least YopN and TyeA) and the LcrG protein encoded in the lcrGVHYopBD operon [10, 17, 45]. LcrG exerts a secretion-blocking eect and, additionally, might be directly involved in the translocation process [41].
152
The V antigen (LcrV) is one of the major Yops, yet the exact role in the virulon and the pathogenicity of Yersinia spp. is not fully understood. The lcrV gene is located in the lcrGVHYopBD operon. Recent studies suggested that the V antigen is involved in the translocation process. A complete deletion of lcrV resulted in the inability to secrete the translocators YopB and YopD [40]. Moreover, Nilles et al. [29] showed that the V antigen itself is part of the translocation apparatus acting in the proper deployment of YopB. Non-polar inframe deletions in lcrV suggested that the V antigen plays a regulatory role in the expression of other Yop proteins [2, 33, 45]. For the regulatory function of the V antigen, a balanced interaction with LcrG is decisive [28, 29]. In contrast to other Yop proteins, the V antigen is secreted in a non-polarized fashion upon bacteria-cell contact [29]. This feature might be important for the immunomodulatory eects of the V antigen [25±27]. The V antigen has long been known to be a protective antigen against yersinia infections and has been studied in active and passive immunization experiments [21, 22, 24, 36, 48]. At least two dierent types of Yersinia V antigen exist [36]. Antibodies generated against one type are unable to protect against the other. The major difference between the two types occurs at amino acids (aa) residues 225 and 232 [36]. Experiments with genetically engineered truncated V antigen proteins suggested that the protective epitopes are located between aa 168 and 275 [24] or aa 135 and 275 [14]. The protective epitopes are thought to be conformational [34]. In the present study, we analyzed the interaction of Y. enterocolitica and phagocytic cells in the presence of yersinia-speci®c antibodies. We show that the V antigen is in fact part of the translocation apparatus. Antibodies against V antigen neutralize to some degree a main function of the Yop virulon: the inhibition of phagocytosis by macrophages. This provides the ®rst hint of how antibodies against the V antigen mediate protection. Moreover, we demonstrate that aa 225±32 of the V antigen are not a protective epitope.
Materials and methods Bacterial strains and culture conditions Y. enterocolitica WA-314 contains the virulence plasmid pYVO8 and belongs to the serotype O8 [13]. Escherichia coli DH5a was used as host for DNA cloning, and M15 for production of recombinant V antigen. Bacteria were grown in Luria-Bertani (LB) medium, WA-314 at 27 °C, E. coli at 37 °C. For intravenous infection of mice, aliquots of 14-day-old glycerol stock cultures stored at )80 °C were thawed, washed in sterile phosphate-buered saline (PBS; pH 7.4), and diluted to the appropriate bacteria concentrations [1]. Antibiotics were used at the following concentrations: ampicillin (Ap) 100 lg/ml; kanamycin (Km) 25 lg/ml. DNA manipulations and construction of deletions Plasmid DNA preparations were isolated with Nucleobond kits (Macherey-Nagel, DuÈren, Germany) as recommended by the manufacturers. Cloning methods, PCR performance, and DNA-
sequencing were essentially performed as described previously [36]. DNA fragments were isolated from agarose gels, and PCR fragments were puri®ed using the appropriate kits (Macherey-Nagel). Enzymes, deoxynucleoside triphosphates and Taq-polymerase were purchased from Boehringer (Mannheim, Germany). Oligonucleotides were synthesized by Roth (Karlsruhe, Germany). In-frame deletions in lcrV of strain WA-314 were constructed by ligating two appropriate PCR products in pKS (Stratagene, Heidelberg, Germany). The deletions were veri®ed by DNA sequencing. The deletion Vd1 (aa 226±248) was generated by ligating the PCR products Va (primer Vf1: bp 1±25 of lcrV with a 5¢ BamHI restriction site 5¢-CTGGGATCCATGATTAGAGCCTACGAACAA-3¢ and primer Vrev5: bp 675±654 of lcrV with a 5¢ HindIII restriction site: 5¢-CAACTCGTCAAGCTTAATGGTGGTTTGAGGCAT-3¢) and Vb (primerVf4: bp 747±765 of lcrV with a 5¢ HindIII restriction site 5¢-TAAAGAACAAGCTTGAAAGTGAGAATAAAAGAA-3¢ and Vrev1: downstream of the stop codon of lcrV with a 5¢ SacI restriction site 5¢-CTCGAGCTCCTGGTATTCTTGAGTGTCTGT-3¢). The deletion Vd2 (aa 190±248) was generated by ligating the PCR products Vc (primer Vf1 and primer Vrev6: bp 542±566 of lcrV with a 5¢ HindIII restriction site: 5¢-GTACTCAAGCTTCTCATGTATATTTATGGTGCCAC-3¢) and Vb. The deletion Vd3 (aa 190±278) was generated by ligating the PCR products Vc and Vd (primer Vf5: bp 833±852 of lcrV with a 5¢ HindIII restriction site 5¢-CCACCGAAGCTTGCCACCGCCTGCTCGGAT-3¢ and primer Vrev1). The deletion Vd4 (aa 226±278) was generated by ligating PCR products Va and Vd. Vd5 carries the ®rst 225 aa of lcrV and was constructed by cloning the PCR product Va. The mutated lcrV gene fragments were ligated into pQE30 (Qiagen, Hilden, Germany) and transformed in M15 containing the repressor plasmid pREP4. The resulting plasmids were denoted pQEVd1-Vd5. The coding genes of LcrG and YopD were ampli®ed by PCR using WA-314 as DNAsource and the primers LcrG1 (5¢-AGTAGGATCCATGAAATCTTCCCATTTTGATG-3¢) and LcrG2 (5¢-AATTGAGCTCTTAAATAATTTGCCCTCGCATCA-3¢) for lcrG and YopD1 (5¢-CTCGGATCCATGACAATAAATATCAAGACA3¢) and YopD2 (5¢-CTCGAGCTCTCATAAATGGTCAGACAACA-3¢) for yopD. The PCR products were digested with the appropriate enzymes and ligated into pQE30. Expression and puri®cation of recombinant proteins For the production of recombinant proteins the histidine-tagged protein expression and puri®cation system QIAexpress of Qiagen was used. Production and puri®cation of recombinant mutated V antigen and LcrG in native form was performed as described previously [36]. Recombinant YopD was puri®ed in denatured form as recommended by the manufacturer. The puri®ed recombinant proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) [19]. The protein concentration was measured by bicinchoninic acid protein assay (Pierce, Freiburg, Germany). Aliquots of the proteins were stored at )80 °C. Puri®ed recombinant mutated V antigens were denoted as rVd1 (deletion: aa 226±248), rVd2 (deletion: aa 190±248), rVd3 (deletion: aa 190±278), rVd4 (deletion: aa 226±278), rVd5 (deletion: aa 226± 325, end of the V antigen), rVO8 (full length, serotype O8 [34]) and rVO3 (full length, serotype O3 [36]), respectively. Preparation of antisera For production of polyclonal rabbit antisera against the dierent recombinant mutated V antigens, LcrG and YopD, 2-month-old white New-Zealand rabbits were used. Before immunization blood samples were taken from both rabbits and pooled [normal rabbit serum (NRS)]. For each immunization 150 lg recombinant protein in 1 ml PBS, respectively, were mixed with lyophilized adjuvant (ABM adjuvant, Sebak, Germany) and injected 0.5 ml intramuscularly and 0.5 ml subcutaneously. The ®rst immunization was done with complete adjuvant (ABM 2) followed by three booster immunizations with incomplete adjuvant (ABM 1) at intervals of
153 4 weeks. Finally, the rabbits were killed, blood was sampled by heart puncture and serum was collected after clotting. The speci®city of the antisera was tested by immunoblotting using Yops [36]. The antisera were heat-inactivated and the immunoglobulin fractions from the dierent antisera and the NRS were enriched by ammonium sulfate precipitation and extensive dialyzation against PBS [49]. Dialyzed immunoglobulin fractions were resuspended in the same volumes as the original antisera and used in cell culture experiments. To compare the dierent V antigen antisera, the puri®ed immunoglobulin fractions were tested by ELISA using rVO8 [36]. Based on these values the immunoglobulin fractions were adjusted to the same amount of V antigen-speci®c antibodies (dilution factors 1:1.25 to 1:3, protein concentration of the adjusted immunoglobulin fractions: 1.9±4.8 mg/ml). When retested in the rVO8-ELISA all adjusted immunoglobulin fractions resulted in extinction values of max/2 (half of the maximal absorbance) at a dilution of 2 ´ 10)3. ELISA testings were performed in three independent experiments. Isolation of proteose-peptone stimulated peritoneal macrophages, polymorphonuclear leukocytes and cell culture Peritoneal exudate cells were obtained from BALB/c mice that had received an intraperitoneal injection of 1 ml 10% proteose-peptone (Difco) 3 days previously. The peritoneal cells were washed twice with PBS, resuspended in RPMI 1640 medium (Gibco BRL, Karlsruhe, Germany) supplemented with 10% heat-inactivated fetal calf serum (FCS; Biochrom) and seeded in cell culture dishes. After 2 h, nonadherent cell were removed by repeated washing with PBS. Human polymorphonuclear leukocytes (PMN) were isolated from peripheral blood obtained from healthy volunteers by a onestep separation method using Mono-Poly Resolving Medium (Flow Laboratories, Irvine, UK) [7]. The murine macrophage-like cell line J774A.1 was routinely grown to con¯uent layers in RPMI 1640 medium with 5 mM L-glutamine (Gibco) supplemented with 10% FCS in a 5% CO2 atmosphere. Three hours prior to infection J774A.1, proteose-peptone stimulated peritoneal macrophages (PPPM) and PMN, respectively, were subcultured into 17.5-mmdiameter 24-well tissue culture plates (Nunc, Roskilde, Denmark) to a density of 5 ´ 105 (macrophages) or 2 ´ 106 (PMN) per well. Phagocytosis and gentamicin protection assay WA-314 grown overnight at 27 °C were diluted 1:40 in LB and incubated at 37 °C for 2 h. Prior to infection WA-314 was washed twice with PBS and resuspended in prewarmed cell culture medium without antibiotics. Each well was infected with 3 ´ 106 (for macrophages) or 1 ´ 107±4 ´ 107 (for PMN) WA-314 in 500 ll for 1 h at 37 °C. The actual dosis of infection was determined in each assay by plating serial dilutions [(3 1.02) ´ 106/ml]. Together with the bacteria, antisera were added as enriched immunoglobulin fractions to the infection at a ®nal concentration of 10% (vol/vol) if not stated otherwise. In some experiments cytochalasin D was added 30 min before infection at a ®nal concentration of 1 lg/ml. To visualize the bacteria-eukaryotic cell interaction, wells were washed twice with PBS after infection, ®xed with ice-cold methanol, and stained with Giemsa solution. To discriminate between intra- and extracellularly located bacteria, cells were stained by the doubleimmuno¯uorescence technique (d-IFT) described by Heesemann and Laufs [12]. The results of Giemsa staining and d-IFT represent the mean of at least three independent experiments, for which at least 100 eukaryotic cells were evaluated. To determine the amount of internalized bacteria the gentamicin protection assay described by Isberg and Falkow was used [18]. After infection wells were washed once with PBS and incubated for 2 h at 37 °C with gentamicin at a ®nal concentration of 40 lg/ml in cell culture medium without other antibiotics. After washing twice with PBS eukaryotic cells in each well were lysed with 500 ll 0.1% Triton X-100. The number of surviving bacteria were determined by plating serial dilutions (see above).
Chemiluminescence assays with PMN The activation of the neutrophil oxidative burst was measured as luminol-enhanced chemiluminescence (CL) and monitored with a microplate-chemiluminometer (Hamamatsu Photonics, Herrsching, Germany) as described previously [39]. Bacteria were tested preopsonized with 5% normal human serum at a multiplicity of infection (MOI) of 40:1. Speci®c antisera were added as adjusted puri®ed immunoglobulin fractions at a ®nal concentration of 10% (vol/vol). For restimulation of PMNs, opsonized zymosan (250 lg/ ml) was added and the secondary CL signal was measured for 75 min [39]. Each assay was performed in duplicate and repeated at least three times. A representative and reproducible CL graph is shown (see Fig. 2). Passive immunization and protection experiments For the intravenous infection route 6-week-old female BALB/c mice (Charles River WIGA, Sulzfeld, Germany) were injected intraperitoneally with 500 ll PBS containing 10 ll or 100 ll of adjusted immunoglobulin fractions of anti-rVd1-d5, anti-rVO8 or NRS, respectively, 1 day before infection. For infection 30 ´ minimal lethal doses (30 ´ LD50) of WA-314 (3 ´ 104 resuspended in 100 ll PBS [49]) were injected intravenously. After 4 days the mice were killed by cervical dislocation. Spleen and liver were prepared aseptically and homogenized. The bacterial load in the organs was determined by plating serial dilutions on MuellerHinton agar and counting colony forming units (CFU) after incubation at 27 °C of 48 h as described previously [35]. For oral infection BALB/c mice were injected intraperitoneally with 100 ll of puri®ed immunoglobulin fractions of anti-rVO8, anti-YadA [49], and NRS, respectively, in 500 ll PBS 1 day before infection. Mice were orogastrically infected the next day with 3 ´ 108 WA-314 grown at 27 °C. At days 1, 3, 5 and 7 the bacterial load in the organs was determined as described above (Animal licensing committee permission no. 211-2531-40/96). Statistics The signi®cance of the dierences among the control and experimental groups was determined by the Student t-test. P values of 10 bacteria/cell) were generated by rising the MOI (>25:1) or using young J774A.1 cells (2 days old, noncon¯uent in stock culture), bacteria tended to autoagglutinate on the cell surface and the stimulatory eect of anti-rVO8 on the uptake of the bacteria decreased. The gentamicin kill assays were repeated with PPPM instead of J774A.1. The results are shown in Table 1. Anti-rVO8 stimulated the uptake of bacteria in PPPM in the same fashion as in J774A.1 cells. The presence of anti-V antigen serum has low impact on the interaction between Y. enterocolitica and human PMN We wondered whether the stimulating eect of antirVO8 on the uptake of WA-314 was unique to macrophages or if it also occured in human PMN. Therefore, we infected 2 ´ 106 PMN/wells with WA-314 at ratios of 5:1±20:1 bacteria/cell in the presence of dierent anti-
Table 1 Phagocytosis of WA-314 by PPPM in the presence of dierent antisera (10%) determined by gentamicin kill assay: 5 ´ 105 PPPM/well were infected with 2 ´ 106 bacteria. At 1 h extracellular bacteria were killed with gentamicin and surviving bacteria were determined by plating serial dilutions. At 1 h after infection cell-associated bacteria were determined by Giemsastaining. Results are the mean of at least three independent experiments SD (PPPM proteose-peptone stimulated peritoneal macrophages, NRS normal rabbit serum) Antisera
Bacteria/cell
Surviving bacteria/well
NRS Anti-rVO8 Anti-YadA
2.15 (0.68) 1.66 (0.48) 1.89 (0.81)
3.1 ´ 104 (2.1) 2.6 ´ 105 (1.4)* 4.2 ´ 104 (1.6)
*P value < 0.05, determined by the Student's t-test
sera. After 1 h the number of intracellular surviving bacteria were determined by gentamicin kill assays. Despite the use of higher cell and bacteria concentrations the number of recultivated bacteria after gentamicin treatment was low when PMN were used instead of macrophages. Less than 10 bacteria/well (n 4) survived if 10% NRS were added to the infection. The number of surviving bacteria rose to 3.6 1.7 ´ 102 (n 4, MOI 20:1) if 10% anti-rVO8 was used. Between 10 and 30 bacteria/well (n 4) survived in the presence of 10% anti-YadA serum. Translocation of Yop proteins into the cytosol of PMN results in inhibition of the phagocytosis and down-regulation of the oxidative burst [39]. The inactivation of PMN upon contact with virulent Y. enterocolitica strains could be demonstrated by measuring a secondary oxidative burst signal (restimulation by opsonized zymosan). PMN were incubated with WA-314 in the presence of dierent antisera for 1 h and the secondary oxidative burst was measured (Fig. 2). Irrespectively of the antisera added to the infection, WA-314 was able to inactivate the PMN. Anti-V antigen serum protects BALB/c mice against lethal intestinal infection with Y. enterocolitica
Fig. 1 Gentamicin kill assay. J774A.1 cells, 5 ´ 105/well were infected with 2 ´ 106 WA-314 in the presence of 10% antisera, if not stated otherwise. After 1 h extracellularly located bacteria were killed with gentamicin and the level of intracellularly surviving bacteria were determined after cell lysis by plating serial dilutions. Cytochalasin D (1 lg/ml) was added (+cyt.D) 30 min before infection to suppress the phagocytosis of J774A.1 cells. P value for rVO8 < 0.05 as determined by the Student's t-test
Anti-YadA and anti-V antigen sera are known to be protective in the intravenous-infection mouse model [24, 35, 48, 49]. However, an eect in orogastric infections has not been shown so far. PMN are part of the innate immune response and one of their main function is thought to be the clearance of bacteria at the primary site of infection (i.e., Peyer's patches in the case of an orogastric infection with Y. enterocolitica). To investigate the protective eect of the respective antisera in intestinal infection BALB/c mice were infected orogastrically with WA-314 after passive immunization with anti-YadA or anti-rVO8. The course of infection was monitored by determining the bacterial load in the organs at days 1, 3, 5, and 7. The results are shown in Fig. 3. Mice passively immunized with anti-rVO8 were able to overcome a lethal orogastric infection with WA-
155 Fig. 2 Secondary CL response. PMN were preincubated with (black symbols) or without (white symbols) WA-314 (MOI 20:1) in the presence of dierent sera. (j, h) 10% NRS; (r, e) 10% anti-YadA serum; (m, n) 10% anti-rVO8 serum. After 1 h the PMN were restimulated with opsonized zymosan and the secondary CL was measured (CL chemiluminescence, PMN polymorphonuclear leukocytes, MOI multipicity of infection, NRS normal rabbit serum)
314. Anti-YadA was able to prevent the establishment of a systemic infection. However, anti-YadA was unable to prevent the multiplication of WA-314 in the Peyer's patches. On day 7 post infection these mice had developed large abscesses in the Peyer's patches and would have died by ileus or perforation. The type-speci®c region (aa 225±232) of the V antigen is not a protective epitope To analyze if the type-speci®c region (aa 225±232) of the V antigen is a protective epitope we constructed in-frame deletions removing this region in lcrV. The mutated lcrV genes were transferred into an expression system (pQE) and rabbit antisera were generated against the puri®ed recombinant proteins (anti-rVd1±5). To compare the dierent antisera, the content of V antigen-speci®c antibodies were determined by ELISA using rVO8 as antigen. The antisera were adjusted to the same amount of speci®c antibodies and used in cell culture and passive immunization experiments. The results of the immunization experiments are shown in Table 2. All antisera tested exhibited a protective eect in BALB/c mice against a lethal infection with WA-314 if high doses (100 ll) of antisera were used for immunization. Lower doses of antisera (10 ll) resulted in a decrease of the protective eect and, especially, anti-rVd3 failed to prevent the multiplication of WA-314. Anti-rVd1±5 were also tested in cell culture experiments for their ability to promote the uptake of WA-314 in J774A.1 cells. The results are shown in Fig. 4. Antisera anti-rVd1, 2, 4, 5 which were protective against lethal intravenous infection with WA-314 at low-dose application (10 ll) were also able to promote the uptake of WA-314 in J774A.1 cells. In agreement with the
passive immunization experiments, anti-rVd3 had no stimulating eect on the phagocytosis by J774A.1 cells. The low activity of anti-rVd3 in the two assays and the fact that the content of V antigen-speci®c antibodies in anti-rVd3 was comparable to anti-rVO8 suggested that a protective epitope might be deleted in rVd3. The deletion in rVd3 ranges from aa 190 to 278. However, the antisera generated against the other constructs: rVd1 (deletion: aa 226±248), rVd2 (deletion: aa 190±248), rVd4 (deletion: aa 226±278), rVd5 (deletion: aa 226±325, end of the V antigen) were shown to be as active as rVO8 in immunization and phagocytosis assays. Based on these data it is not possible to localize the protective epitope between the aa 190±278.
Discussion An important feature of pathogenic Yersinia spp. is the ability to prevent phagocytosis. Cell-adhering bacteria achieve the inactivation of phagocytic cells by translocating Yop proteins (i.e., YopH and YopE) into the cytosol of these cells via a type III secretion system [5]. Labeling of the bacterial surface with antibodies (antiYadA or anti-Inv) did not result in phagocytic uptake of yersiniae, indicating that Fc receptor-mediated phagocytosis is also inhibited by the Yop virulon [3, 8, 39]. The question of how antibodies exhibit a protective eect against Yersinia is unsolved. Apart from YadA the V antigen is known as a protective immunogen [24, 35, 48, 49]. Our results suggest that antibodies against the V antigen are able to interfere with the translocation of Yop eector proteins and thereby may counteract the inhibition of the phagocytosis. The uptake-supporting eect of anti-rVO8 was obviously not due to opsonization: (i) in conventional
156 b Fig. 3A±C Course of intestinal infection with WA-314 in passively immunized mice. BALB/c mice were immunized by intraperitoneal injection of 0.6 mg: A NRS, B anti-YadA, C anti-VO8 1 day prior to infection. After orogastric infection with 2 ´ 108 WA-314, the bacterial loads in Peyer's patches (PP), mesenteric lymph nodes (MLN ), spleen (S ), and small intestine (SI ) were determined by plating serial dilutions at the indicated days. The values given are the mean of eight animals SD
immuno¯uorescence microscopy, anti-rVO8 did not signi®cantly label the surface of un®xed WA-314 grown at 37 °C with or without Ca2+ (data not shown) [44], (ii) anti-rVO8 had no in¯uence on the cell adherence of the bacteria, and (iii) in agreement with previous reports anti-YadA labeling the bacterial surface was ineective in promoting uptake in phagocytic cells [37, 40]. The uptake-supporting eect of anti-rVO8 was more distinct if macrophages were used instead of PMN. The reason for the dierent behavior of these two cell types is not clear. Recently, it has been shown that Yop proteins are translocated in a dierential manner dependent on TyeA [17]. The production of target cellspeci®c translocation systems (for dierent Yops or translocation pili of dierent length) by Yersinia might be an explanation for the dierent behavior of macrophages and PMN in our experiments. Another possibility might be due to dierent susceptibilities of the two cell types against the translocated Yop proteins. A YopE null mutant of Y. pseudotuberculosis has been shown to be more attenuated in the orogastric- than in the intravenous-infection model, indicating that the toxic eect of YopE may be especially directed against PMN [37]. The phagocytosis assays with PPPM and PMN are in accordance with the passive immunization experiments. Anti-rVO8 and anti-YadA [35, 49] were able to protect
Fig. 4 Gentamicin kill assay. J774A.1 cells, 5 ´ 105/well were infected with 2 ´ 106 WA-314 in the presence of 10% antisera. After 1 h extracellularly located bacteria were killed using gentamicin and the level of intracellularly surviving bacteria were determined after cell lysis by plating serial dilutions
157 Table 2 Protection of BALB/c mice after passive immunization with dierent antisera against intravenous infection with WA314. At 1 day after intraperitoneal injection of 10 or 100 ll puri®ed and adjusted rabbit antisera WA-314 was injected intravenously at an infectious dose of 30 ´ LD50 (3 ´ 104). At day 4 after injection the numbers of bacteria in spleen and liver were determined by plating serial dilutions. Results are the means of six animals SD. P values were determined by Student's t-test (NS no signi®cant dierence)
Serum used for passive immunization
Mean no. of bacteria (log10 CFU SD, P value) Spleen
Liver
NRS (10 ll) NRS (100 ll) Anti-rVO8 (10 ll) Anti-rVO8 (100 ll) Anti-rVdl (10 ll) Anti-rVdl (100 ll) Anti-rVd2 (10 ll) Anti-rVd2 (100 ll) Anti-rVd3 (10 ll) Anti-rVd3 (100 ll) Anti-rVd4 (10 ll) Anti-rVd4 (100 ll) Anti-rVd5 (10 ll) Anti-rVd5 (100 ll)
7.61 0.36 7.43 0.46 4.62 0.55,
