TY - JOUR
AU - Guy, Bruno
AB - Abstract The ability to induce a protective response against Helicobacter pylori infection has been investigated by systemic immunization of mice with urease formulated with the cationic lipid DC Chol. This compound acts both as a formulating agent and as an adjuvant and induces a balanced Th1/Th2 response shown to be more effective for protection in our previous studies. Urease-DC Chol induced a significant protection in prophylaxis but not in therapeutic immunization. The protection level was between 1.5 and 2 log reduction of bacterial density measured by quantitative culture compared to unimmunized-infected mice. In parallel, the protective efficacy of other H. pylori antigens formulated in a similar way and administered with DC Chol was tested. These antigens were tested alone or in combination in prophylactic and therapeutic regimens. Some combinations of antigens induced a better prophylactic or therapeutic activity than urease alone (0.5–1.5 log further reduction in prophylaxis and therapy respectively, P<0.05). The combinations that induced the best protection were different in prophylaxis and therapy. In conclusion, DC Chol provides a convenient and efficient method to formulate different antigens even when they are present in non-compatible buffers initially. Moreover, the results obtained in protection against H. pylori with such formulations should lead the way to future clinical trials. 1 Introduction Since 1982 with the initial work of Marshall and Warren [1], the link between Helicobacter pylori infection and gastroduodenal disorders (gastritis, ulcer and cancer) has been clearly established [2,3] and thus opened the way to vaccine research. In this respect, different routes and formulations have been tested in animal models, and most experiments have been performed using mucosal routes (mainly intragastric) and potent mucosal adjuvants like cholera toxin (CT), the heat-labile toxin of Escherichia coli (LT) or their mutant non-toxic derivatives. Under these conditions, prophylactic and therapeutic immunizations using Helicobacter sonicates, killed whole cell preparations or different purified antigens induced significant levels of protection, especially in murine models [4–13] as determined by urease activity, bacterial culture or histology. However, conditions used in some of these studies, in particular the use of toxic adjuvants and the need for large doses in mucosal immunizations, prompted us to explore other directions in the search of a vaccine against H. pylori suitable for human usage. We have thus investigated other routes of immunization by studying different immunization schedules and, in parallel, the importance of the formulations and adjuvants used with the objective of increasing the protection obtained so far. We have observed that in outbred mice systemic immunization could elicit levels of protection equivalent or sometimes superior to those induced by the oral route in the presence of CT or LT when appropriate formulations were applied [14–17]. In agreement with our results, K. Eaton and coworkers recently obtained a significant reduction in bacterial density in piglets infected by H. pylori, using systemic or mucosal immunization. In their experiments, complete eradication was seen only in groups immunized parenterally [18]. Other authors also reported the efficacy of parenteral immunization in mice [19] or dogs [20]. In addition, in a therapeutic setting in mice, we have confirmed and extended our prophylactic studies showing the potential of systemic immunization in the presence of balanced Th1/Th2 adjuvants [16]. However, our results obtained with urease were shown to be variable among different experiments and there is a clear need to improve protection. Many different antigens have been tested so far, including urease [4–8,14–16], catalase [21], VacA [9], or whole cell lysate [10,11,14,18]. The potential antigens were first identified by conventional fractionation techniques, and more recently by genomic screening after the genome sequence of two different strains of Helicobacter pylori became available [22,23]. Direct access to the genome sequence has allowed a large selection of different antigens in our laboratories; 400 genes have been cloned, 200 tested in a murine animal model for protection using mucosal immunization [24]. Several of these tested antigens have shown good protection when tested individually in both prophylactic and therapeutic models and therefore, were selected to be tested in combinations in order to increase the protection offered by urease alone. However, some of these proteins were expressed as inclusion bodies in the heterologous host cell E. coli and their purification required the use of denaturing buffers containing guanidinium chloride or urea. These denaturing agents were sometimes still present at high concentrations in the final products, which made the combination of different antigens problematic. We therefore evaluated the use the cationic lipid adjuvant DC Chol [25], to formulate the different antigens present in incompatible buffers on the one hand and to try to induce an adequate immune response against H. pylori on the other hand. This adjuvant is actually a ‘balanced’ Th1/Th2 adjuvant [25], a profile likely to be associated with protective responses against H. pylori; indeed, our first results suggesting the critical role of a Th1 arm in addition to a Th2 response [15] have been then confirmed by many other authors [26–28]. In addition, although systemic immunization in the presence of QS21 gave promising results [15,16], this adjuvant may not be a proper one for human usage due to its relative systemic toxicity. This incited us to evaluate the use of DC Chol to prepare antigenic combinations for parenteral administration. In the present study we have evaluated the ability of DC Chol adjuvant to induce an adequate immune response against H. pylori urease. In parallel, the results were compared with those obtained from other antigens selected from the genomic screening [24]. Due to the size of the experiments, the selected antigens were restricted to AlpA [23,29] (TIGR 912), catalase [21,23] (TIGR 875), BabB [23,30,31] (TIGR 896), and a serine protease (Pr) [23] (TIGR 1012). These antigens were formulated the same way as urease with DC Chol and tested, individually or in combination, for their ability to induce higher protective responses in a mouse model. The evaluation was made in both prophylactic and therapeutic regimens. 2 Materials and methods 2.1 Antigens and adjuvants Recombinant H. pylori urease was expressed and purified as previously described [7]. Briefly, after cloning of the ureA and ureB genes under an inducible promoter and transformation in E. coli, inactive recombinant urease was expressed and purified from cell pellets. After several chromatographic steps including ion exchange and gel filtration, purified urease was lyophilized and stored at −20°C. After reconstitution, urease was stored at 4°C. The same urease preparation was used for all the experiments described in this study. The other antigens were cloned in and expressed from E. coli[24]. 2.2 Formulation of H. pylori antigens The liposomal formulations were prepared using the general detergent dialysis technique as described by Weder and Zumbuehl [45]. A detailed description of the preparation of H. pylori antigens will be given elsewhere (C. Geoffroy et al., in preparation). Briefly, chloroform solutions of lipids in the presence or absence of lipoidal adjuvants were mixed, evaporated, vacuum desiccated and resuspended in a buffer to yield a liposome suspension. The suspension was homogenized by either extrusion, microfluidization or sonication, and the resulting vesicles were transformed into lipid/detergent mixed micelles by the addition of excess detergent (e.g. alkylglycosides, bile salts, etc.). The antigens of interest were then added to the mixed micelles to form a homogeneous solution. Finally the detergent was removed by controlled dialysis to restore the liposomes with proper incorporation of the antigens. The procedure is applicable to formulate cocktails of proteins up to five different antigens as used in the present study, or to formulate any combination of H. pylori antigens. 2.3 Bacterial strain and culture H. pylori X43–2AN, a streptomycin-resistant mouse-adapted [32], was used in this study. This strain was stored at −70°C in Brucella broth (BB) (BioMérieux) supplemented with 20% v/v glycerol and 10% v/v fetal bovine serum (FBS) (Hyclone). The suspension of the challenge strain was prepared as follows: for pre-culture, H. pylori was grown on Mueller-Hinton Agar (MHA; Difco) containing 5% v/v sheep blood (bioMérieux) and antibiotics: 5 µg ml−1 thrimethoprim, 10 µg ml−1 vancomycin, 1.3 µg ml−1 polymyxin B sulfate, 5 µg ml−1 amphotericin, and 50 µg ml−1 streptomycin (selective marker of strain X43-2AN) (TVPAS). All antibiotics were purchased from Sigma. MHA-TVPAS plates were incubated for 3 days at 37°C under micro-aerobic conditions (Anaerocult C, Merck). The pre-culture was used to inoculate a 75-cm2 vented flask (Costar) containing 50 ml of BB supplemented with 5% v/v FBS and all antibiotics (TVPAS). The flask was kept under micro-aerobic conditions with gentle shaking for 24 h. The suspension was characterized by Gram staining, urease activity (Urea indole medium, Diagnostic Pasteur), catalase (H2O2, 3% v/v) and oxidase activities (bioMérieux disks). Viability and motility were checked by contrast phase microscopy. The suspension was diluted in BB to OD550=0.1 (which was equivalent to 1×107 CFU ml−1) and used for challenge. 2.4 Animal model Outbred OF1 female mice 6–8 weeks old were purchased from Iffa Credo (France). During the studies cages were covered (using Isocaps), mice were given filtered water and irradiated food and autoclaved material was used. For prophylactic immunization, mice were immunized on days 0, 21, and 42. Immunization was performed by the subcutaneous (s.c.) route (300 µl under the skin of the left part of the lumbar region). 5 µg of recombinant H. pylori urease and of each antigen (alone or within the cocktails) was administered by the s.c. route. Mice were challenged 4 weeks after the second boost by gastric gavage with 300 µl of the H. pylori suspension (3×106 CFU). For therapeutic immunization, mice were infected on day 0 with the challenge strain. Infection was assessed by analyzing gastric urease activity on 10% randomly selected mice 1 month after infection, and all mice were infected. The remaining animals were then immunized as previously described. Evaluation of protection was done 1 month after the last immunization. 2.5 Evaluation of the protection Four weeks after the challenge, mice were killed and stomachs (antrum+corpus) were sampled in a sterile hood. The protection level was assessed by three different methods: urease activity (Jatrox test, Procter and Gamble), bacterial culture, and histological analyses. Urease activity was assessed on a quarter of the stomach 4 and 24 h post-mortem by measuring pH change of phenol red at 550 nm, as previously described [15]. As for the quantitative culture, the mucosa from one half stomach was stored in the culture transport medium (Portagerm, BioMérieux) and transferred to the culture room within 2 h. The specimen was removed and homogenized with a sterile Dounce tissue grinder (Wheaton, Millville, USA) containing 1 ml of BB, and serially diluted to 10−3. 100 µl of each dilution was inoculated onto MHA+TVPAS plates and incubated under micro-aerobic conditions at 37°C for 4–5 days. Viable counts were recorded. H. pylori was identified by positive urease, catalase, and oxidase activities and by typical appearance on Gram stain. For histology, a quarter of the stomach was placed in 10% buffered formalin (Labo-Moderne) and then processed for tissue sectioning. Sections were stained with hematoxylin and eosin, and gastritis was scored based upon the infiltration of lymphocytes, plasma cells and neutrophils [7]. 2.6 Western blot analysis Inactivated H. pylori were sonicated and total extract was separated by electrophoresis on a SDS-PAGE gel. After transfer of proteins and saturation with milk, the membrane strips were incubated with sera from different immunizations, and the presence of specific IgG1 and IgG2a antibodies detected according to standard procedures. Visualization of bands was carried out with the ECL technique (Amersham). 2.7 Measurement of cytokines/ELISPOTs with spleen cells Nitrocellulose plates (Millipore) were coated with 5 µg ml−1 of anti-mouse interleukin (IL) 10 or interferon γ (IFNγ) (Pharmingen). The spleens were teased through a 70-µm filter (Falcon). After treatment with Gey's solution to eliminate red cells and three further washes, the cells were counted and loaded into the wells of the plates at a final concentration of 2×105 cells in 100 µl in each well. Three different concentrations (final concentration of 30, 10 and 3 µg ml−1) of filtered H. pylori extract (containing 25% urease) were added to the wells to stimulate the cells for 44 h at 37°C with 5% CO2. Each assay was done in triplicate in RPMI 1640 (Gibco) supplemented with 5% decomplemented fetal calf serum, sodium pyruvate, β-mercaptoethanol, glutamine and antibiotics. A positive control (concanavalin A, Sigma, at a 5 µg ml−1 final concentration) and a negative control (medium alone) were performed for each mouse. Secondary biotinylated anti-mouse IL5 or IFNγ antibodies (Pharmingen) were used at 1 µg ml−1. Spots were revealed with AEC substrate (Sigma) and once the plates dried, counted with an automated spot counter (Microvision, France). The number of spots for 106 cells induced by 10 µg ml−1H. pylori extract was determined and the background (spots induced by medium alone, negative control) was subtracted. 2.8 ELISA ELISAs were performed according to standard protocols (biotinylated conjugates, streptavidin peroxidase complex were from Amersham and OPD substrate from Sigma). Plates (Maxisorb, Nunc) were coated overnight at 4°C with H. pylori extracts (5 µg ml−1) in carbonate buffer. After saturation with bovine serum albumin (Sigma), plates were incubated with the sera (1.5 h), biotinylated conjugate (1.5 h), streptavidin peroxidase complex (1 h) and substrate (10 min). A polyclonal mouse serum directed against H. pylori extract served as a control in each experiment. The titers were expressed as the inverse of the dilution giving 50% of the maximal absorbance value at 492 nm. 2.9 Statistical analysis Protection was assessed by quantitative culture from infected stomachs and differences between groups was estimated by Newman-Keuls and Dunnett's tests. 3 Results In previous experiments, we had observed that urease injected with 200 µg of DC Chol was able to induce levels of protection in OF1 mice similar to those induced orally in the presence of LT (unpublished results). We then used the systemic route to compare the urease-induced protection to that induced by other antigens and/or combinations. 3.1 Immune responses against H. pylori urease Urease administered s.c. with DC Chol induced a balanced serum IgG1/IgG2a response, consistent with both the IFNγ response and the IL5 response observed in the spleen cells re-stimulated with urease in vitro (Fig. 1). In similar experiments, urease administered with alum did not induce significant IFNγ or IL5 production (not shown). Experiments carried out with six different preparations of urease-DC Chol induced consistently the same pattern of immune responses in mice (not shown). Figure 1 View largeDownload slide Immune responses against recombinant urease formulated with DC Chol. a: Antibody response in serum. b: IFNγ production by spleen cells. c: IL5 production by spleen cells. Figure 1 View largeDownload slide Immune responses against recombinant urease formulated with DC Chol. a: Antibody response in serum. b: IFNγ production by spleen cells. c: IL5 production by spleen cells. 3.2 Immune responses against antigen combinations The immune responses in terms of serum antibody isotypes against the different antigens were analyzed by Western blot. Fig. 2 shows an example of this analysis, and for each formulation reactivity was observed against the expected proteins. When formulated in multivalent combinations recombinant urease and catalase induced both IgG1 and IgG2a while AlpA, BabB and Pr induced a predominant IgG1 response, the signal against the latter protein being very weak. On the other hand, when formulated alone, all antigens induced a significant and balanced IgG1 and IgG2a response (BabB and Pr, see Fig. 2; AlpA, catalase, and urease, not shown). When multiple antigens were used, interference was then observed between them for the induction of antibody responses, and urease and catalase were immunodominant in this respect. The recombinant BabB induced reactivity against two different proteins or isoforms of the total extract, one migrating at the expected size (76 kDa) and the other at 55 kDa. Knowing that BabB belongs to a family of paralog sequences, it is interesting to note that only two of them showed cross-reactivity under the conditions analyzed. Figure 2 View largeDownload slide Western blot analysis of serum responses raised against some antigen combinations, as indicated at the top of the gels. Total extracts of H. pylori were loaded on 10–15% SDS gel gradients. The UreA subunit is not visible on the figure. Figure 2 View largeDownload slide Western blot analysis of serum responses raised against some antigen combinations, as indicated at the top of the gels. Total extracts of H. pylori were loaded on 10–15% SDS gel gradients. The UreA subunit is not visible on the figure. 3.3 Protection induced by prophylactic immunization Protection was assessed by measuring the level of urease activity in the stomachs of all mice, and by quantitative culture in the stomachs of all or half of the mice per group (Fig. 3). When analyzed by quantitative culture, the best known antigen urease induced about 2 log reduction of bacterial colonization (median CFU values), but the reduction was quite heterogeneous. When formulated with DC Chol, similar protection was also achieved by the other antigens, except for Pr (P<0.05). The combinations of antigens (including three to five antigens) also induced about 2 log reduction in bacterial charge, and the results were more homogeneous for most of the combinations. In the single-antigen groups, about 50% of mice had bacterial counts higher than 3000 CFU, particularly in the group of urease, while less than 20% of the multiple-antigen groups presented such a high value. Similarly, less than 25% of the mice presented low bacterial counts (below 1000 CFU) in the single-antigen groups, and more than 50% of the mice achieved this value in the multiple-antigen groups. A similar degree of homogeneity in protection was also observed in two other separate experiments using DC Chol or a combination of DC Chol and Bay adjuvants (unpublished data). Figure 3 View largeDownload slide Analysis of prophylactic efficacy of the different antigen combinations (monovalent vs. tri- to pentavalent formulations). A: Urease activity in stomach (bar=mean urease activity). B: Quantitative culture (bar=median CFU value). Mice were immunized three times with different monovalent or multivalent DC Chol formulations (5 µg of each antigen and 200 µg DC Chol final), and challenged 2 weeks after the second boost. Analysis was performed 1 month later. A quarter of the stomach was sampled for urease analysis and half for quantitative culture. Figure 3 View largeDownload slide Analysis of prophylactic efficacy of the different antigen combinations (monovalent vs. tri- to pentavalent formulations). A: Urease activity in stomach (bar=mean urease activity). B: Quantitative culture (bar=median CFU value). Mice were immunized three times with different monovalent or multivalent DC Chol formulations (5 µg of each antigen and 200 µg DC Chol final), and challenged 2 weeks after the second boost. Analysis was performed 1 month later. A quarter of the stomach was sampled for urease analysis and half for quantitative culture. In order to know if the number of antigens can be reduced to less than three, we then performed another experiment comparing monovalent and bivalent combinations. Fig. 4 presents the bacterial culture results obtained after challenge of immunized mice. As previously observed, AlpA, catalase and urease provided about 1.5 log reduction of bacterial load (median CFU value; P<0.05), but no better reduction of bacterial load was observed with the bivalent combinations among the three groups tested, and AlpA repeatedly induced the most homogeneous protection when administered with DC Chol adjuvant in prophylaxis. Figure 4 View largeDownload slide Analysis of prophylactic efficacy of some monovalent vs. bivalent formulations (quantitative culture). Mice were immunized three times with different monovalent or bivalent DC Chol formulations (5 µg of each antigen and 200 µg DC Chol final), and challenged 2 weeks after the second boost. Analysis was performed 1 month later. A quarter of the stomach was sampled for urease analysis and half for quantitative culture. Figure 4 View largeDownload slide Analysis of prophylactic efficacy of some monovalent vs. bivalent formulations (quantitative culture). Mice were immunized three times with different monovalent or bivalent DC Chol formulations (5 µg of each antigen and 200 µg DC Chol final), and challenged 2 weeks after the second boost. Analysis was performed 1 month later. A quarter of the stomach was sampled for urease analysis and half for quantitative culture. 3.4 Clearance induced by therapeutic immunization Therapeutic activity was then assessed using the same formulations. Western blot analysis performed in mice immunized after challenge showed similar profiles as in the prophylactic experiment, indicating that prior colonization did not influence the level and the quality of the antibody responses induced by the different formulations by the systemic route. The different cocktails were then compared to urease in their ability to reduce colonization. As shown in Fig. 5A,B, urease formulated with DC Chol did not induce a significant reduction in bacterial density, while some combinations did. The combinations containing catalase, BabB and Pr induced the best levels of reduction with almost 2 log in median CFU values (P<0.05). Figure 5 View largeDownload slide Analysis of therapeutic efficacy of some different antigen combinations (monovalent vs. tri- to pentavalent formulations). A: Urease activity in stomach (bar=mean urease activity). B: Quantitative culture (bar=median cfu value). Mice were first infected with H. pylori. One month later, 10% of mice were randomly analyzed. As all mice were infected, the remaining mice were then immunized three times with the different monovalent or multivalent DC Chol formulations tested in prophylactic vaccination. Analysis was performed 1 month after the second boost. A quarter of the stomach was sampled for urease analysis, and half for quantitative culture. Figure 5 View largeDownload slide Analysis of therapeutic efficacy of some different antigen combinations (monovalent vs. tri- to pentavalent formulations). A: Urease activity in stomach (bar=mean urease activity). B: Quantitative culture (bar=median cfu value). Mice were first infected with H. pylori. One month later, 10% of mice were randomly analyzed. As all mice were infected, the remaining mice were then immunized three times with the different monovalent or multivalent DC Chol formulations tested in prophylactic vaccination. Analysis was performed 1 month after the second boost. A quarter of the stomach was sampled for urease analysis, and half for quantitative culture. Similarly, in another therapeutic experiment we compared the ability of monovalent versus bivalent combinations. Fig. 6 shows that, as in the previous experiment, urease or other single antigens did not induce significant reduction in bacterial density, and confirmed that some combinations — AlpA/Cat and AlpA/BabB — could increase the protection level. Figure 6 View largeDownload slide Analysis of therapeutic efficacy of some monovalent vs. bivalent formulations (quantitative culture; bar=median CFU value). Mice were first infected with H. pylori. One month later, 10% of mice were randomly analyzed. As all mice were infected, the remaining mice were then immunized three times with the different monovalent or multivalent DC Chol formulations tested in prophylactic vaccination. Analysis was performed 1 month after the second boost. A quarter of the stomach was sampled for urease analysis, and half for quantitative culture. Figure 6 View largeDownload slide Analysis of therapeutic efficacy of some monovalent vs. bivalent formulations (quantitative culture; bar=median CFU value). Mice were first infected with H. pylori. One month later, 10% of mice were randomly analyzed. As all mice were infected, the remaining mice were then immunized three times with the different monovalent or multivalent DC Chol formulations tested in prophylactic vaccination. Analysis was performed 1 month after the second boost. A quarter of the stomach was sampled for urease analysis, and half for quantitative culture. In both therapeutic experiments, urease did not increase protection when present in the formulations, and the best combinations selected in these experiment were different from that of the prophylaxis (Fig. 3 compared to Fig. 5, same combinations). It is to be noted that, contrary to what we generally observed in prophylactic studies where some correlation exists between urease activity and quantitative culture (the latter test being more sensitive, [15]), such a correlation was not present among different therapeutic studies conducted in our laboratory, including that in the present study. Finally, the level of gastritis was analyzed in some groups that did or did not present a reduced colonization. Although moderate gastritis was observed in infected mice compared to uninfected mice (average score 1–2 in the former group vs. 0–1 in the latter), no differences were observed in immunized-infected mice compared to unimmunized-infected mice (not shown), in agreement with our previous results [15,16]. 4 Discussion Our results demonstrate that the use of some combinations of antigens formulated with DC Chol adjuvant reduces the variation of protection induced with individual antigens by the systemic route in mice. Moreover, some combinations induced a further reduction in colonization compared to individual antigens, in particular urease, which had been extensively studied in many laboratories including ours, and was generally agreed to be the gold standard. In our first study using DC Chol and urease [15], the results obtained were not as good as the present ones. This discrepancy can be explained by the quantity of adjuvant used, three times less in the previous case, and the way formulations were prepared; accordingly, the IgG1/IgG2a ratio was inverted compared to what obtained in this study [15]. When formulated with DC Chol, urease was effective in prophylaxis but not in therapy. We had observed that when mice were killed a long time after the challenge (after 3 months or more), urease activity was negative in some colonized stomachs although quantitative culture allowed recovery of bacteria with the expected level (unpublished observations). It was thought that these bacteria would be able to escape the immune response induced by urease unless the adjuvant used has a strong ‘non-specific’ activity that allows a reactivation of the bacteria or a broader stimulation of not yet identified effector mechanisms. This could be the case for LT by mucosal the route or for QS 21 by the systemic route for instance [15,16]. More generally, the positive effect of antigenic combination may be linked to the presence of different subset populations in the infected stomachs, which are heterogeneous with respect to antigen expression. Some subpopulations could constitute targets for protective immune responses if expressing the corresponding vaccinal antigen, while other subpopulations would not. Increasing the number of antigens in vaccine formulations would lower this risk. In vivo kinetic analysis of antigen expression by RT-PCR carried out after infection is in agreement with such downregulation of antigen expression (B. Rokbi et al., in preparation). In addition, two recent studies have indicated the existence of quorum sensing-like mechanisms in H. pylori, supporting the existence of gene expression regulation according to bacterial density [33,34]. Alternatively or in addition, the enhanced protection observed in some combinations may also have been due to simple additive effects. In an outbred mice population, and even more in humans, individuals may react differently to different antigens, and it is conceivable that for instance urease will induce the best level of protection in some individuals, while in others AlpA or catalase will do so. Therefore the probability of inducing a homogeneous level of protection can be increased by combining these different antigens. Nevertheless, some synergy/interference may have occurred in some of our conditions. Although only a cellular immune response has been linked with protection so far [27], formulation of membrane-associated proteins like AlpA and BabB in DC Chol liposomes mimicking a membrane-like environment may have allowed a better refolding of these antigens; and hence an antibody-mediated response against conformational epitopes, which was not induced by previous formulations, may have played a positive role in protection. Actually, a real synergy would, in theory, mean protection by two different mechanisms: cellular and humoral, and it could have been the case observed in this study especially with AlpA in prophylaxis and BabB in therapy. Interestingly, the cocktails that induced the best protection were different in prophylaxis and therapy. While AlpA and catalase were the most effective for prophylaxis, it seems that BabB and Pr to a lesser extent were involved in therapy. One may propose that different antigens are expressed during early invasion and chronic infection phases [35–37]. Since AlpA- and BabB-related proteins bind to different receptors [29–31], these proteins may also be expressed differently during different time periods of infection and this may explain their different reactivity in prophylaxis and therapy. The hypothesis that a differential expression of antigens may occur in vivo with time can also be addressed in the context of the expected correlates of protection. In fact, although an adjuvant-induced Th1 arm is required for protection [16,26,28], we have to keep in mind that such a response is harmful for the host when chronic infection occurs [38,39]. Then, if we need a balanced Th1/Th2 response, do we need these two arms simultaneously or successively? If we hypothesize that different antigens are expressed at different stages of the disease, as has been shown by various authors [35–37], an optimal immunization could induce the following events if the situation is comparable to Lyme disease [40]. First, induction of Th1 inflammatory response against early antigens [41], and, if not sufficient, then a Th2 response against late antigens would take place allowing, in theory, a better reduction of colonization (elimination?) or at least a reduction of inflammation [42]. This would be what we could call a ‘dynamic’ immunization in a sense that immunization would take place as usual, but that the immune response following challenge would be sequential and possibly qualitatively different with time according to the antigens expressed. Antigen cocktails like those used here with DC Chol, inducing balanced Th1/Th2 (urease, catalase) or more predominant Th2 responses (AlpA, BabB, Protease) could in theory be more effective under the conditions that these antigens are indeed expressed differently at different stages of infection. RT-PCR studies conducted on murine and human biopsies are currently being performed in our department, and should help us to answer these questions. In any case, we have shown that a single multivalent formulation could induce different qualitative responses against different antigens. This is in agreement with recent studies published by Ismail and Bretscher [43]. In conclusion, the use of the cationic lipid adjuvant DC Chol allowed, first of all, the induction of protective responses when used with individual antigens and, more importantly, a combination possibility to formulate different protein antigens, and lastly, an increased protection as compared with urease alone. These promising results suggest an alternative approach for vaccine formulation against H. pylori and provide a strong argument to evaluate this approach in future clinical trials [44]. Acknowledgements We acknowledge Marie José Quentin Millet for constant support, and Tom Ermak, Paul Giannasca, Richard Weltzin, and Emanuelle Trannoy for helpful discussions. References [1] Marshall B.J. Warren J.R. ( 1984) Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet  1, 1311– 1315. Google Scholar CrossRef Search ADS PubMed  [2] Correa P. ( 1995) Helicobacter pylori and gastric carcinogenesis. Am. J. Surg. Pathol.  19 ( Suppl. 1), S37– S43. Google Scholar PubMed  [3] Parsonnet J. ( 1995) The incidence of Helicobacter pylori infection. Aliment. Pharmacol. Ther.  9 ( Suppl. 2), 45– 51. Google Scholar PubMed  [4] Corthésy-Theulaz I. et al.   ( 1998) Mice are protected from Helicobacter pylori infection by nasal immunisation with attenuated Salmonella typhimurium phoPc expressing Urease A and B subunits. Infect. Immun.  66 ( 2), 581– 586. Google Scholar PubMed  [5] Ferrero R.L. Thiberge J.-M. Huerre M. Labigne A. ( 1994) Recombinant antigens prepared from the urease subunits of Helicobacter spp.: evidence of protection in a mouse model of gastric infection. Infect. Immun.  62, 4981. Google Scholar PubMed  [6] Gomez-Duarte O.G. Lucas B. Yan Z.X. Panthel K. Haas R. Meyer T. ( 1998) Protection of mice against gastric colonisation by Helicobacter pylori by single oral dose immunisation with attenuated Salmonella typhimurium producing urease subunits A and B. Vaccine  16 ( 5), 460– 471. Google Scholar CrossRef Search ADS PubMed  [7] Lee C.K. Weltzin R. Thomas W.D. Kleanthous H. Ermak T. Soman G. Hill J.E. Ackerman S. Monath T.P. ( 1995) Oral immunisation with recombinant Helicobacter pylori urease induces secretory IgA antibodies and protects mice from challenge with Helicobacter felis. J. Infect. Dis.  172, 161. Google Scholar CrossRef Search ADS PubMed  [8] Michetti P. Corthesy-Theulaz I. Davin C. Haas R. Vaney A.-C. Heitz M. Bille J. Krahaenbuhl J.-P. Saraga E. Blum A.L. ( 1994) Immunisation of BALB/c mice against H. felis infection with H. pylori Urease. Gastoenterology  107, 1002. Google Scholar CrossRef Search ADS   [9] Marchetti M. Arico B. Burroni D. Figura N. Rappuoli R. Ghiara P. ( 1995) Development of a mouse model of Helicobacter pylori infection that mimics human disease. Science  267, 1655. Google Scholar CrossRef Search ADS PubMed  [10] Chen M. Lee A. Hazell S. ( 1992) Immunisation against gastric Helicobacter infection in a mouse Helicobacter felis model. Lancet  339, 1120– 1121. Google Scholar CrossRef Search ADS PubMed  [11] Doidge C. Gust I. Lee A. Buck F. Hazell S. Manne U. ( 1994) Therapeutic immunisation against helicobacter infection. Lancet  343, 914. Google Scholar CrossRef Search ADS PubMed  [12] Czinn S.J. Cai A. Nedrud J. ( 1993) Protection of germ-free mice from infection by Helicobacter felis after oral or passive IgA immunisation. Vaccine  11, 637. Google Scholar CrossRef Search ADS PubMed  [13] Ghiara P. Rossi M. Marchetti M. Di Tommaso A. Vindigni C. Ciampolini F. Covacci A. Telford J. De Magistris M.T. Pizza M. Rappuoli R. Del Giudice G. ( 1997) Therapeutic intragastric vaccination against Helicobacter pylori in mice eradicates an otherwise chronic infection and confers protection against reinfection. Infect. Immun.  65, 4996. Google Scholar PubMed  [14] Guy B. Hessler C. Fourage S. Lecoindre P. Chevalier M. Peyrol S. Boude M. Haensler J. Rokbi B. Quentin-Millet M.J. ( 1998) Mucosal, systemic or combined therapeutic immunisations in cynomolgus monkeys naturally infected with Gastrospirillum hominis like organisms. Vaccine Res.  6, 141. [15] Guy B. Hessler C. Fourage S. Haensler J. Vialon-Lafay E. Rokbi B. QuentinMillet M.J. ( 1998) Systemic immunisation with Urease protects mice against Helicobacter pylori infection. Vaccine  16, 850. Google Scholar CrossRef Search ADS PubMed  [16] Guy B. Hessler C. Fourage S. Rokbi B. QuentinMillet M.J. ( 1999) Comparison between targeted and untargeted systemic immunisations with adjuvanted urease to cure Helicobacter pylori infection in mice. Vaccine  17, 1130– 1135. Google Scholar CrossRef Search ADS PubMed  [17] Weltzin R. Guy B. Thomas W.D. Giannasca P.J. Monath T.P. ( 2000) Parenteral adjuvant activities of Escherichia coli heat labile toxin (LT) and its B subunit for immunisation of mice against gastric Helicobacter pylori infection. Infect. Immun.  68, 2775– 2782. Google Scholar CrossRef Search ADS PubMed  [18] Eaton K.A. Ringler S.S. Krakowka S. ( 1998) Vaccination of gnotobiotic piglets against H. pylori. J. Infect. Dis.  178, 1399. Google Scholar CrossRef Search ADS PubMed  [19] Blanchard T.G. Czinn S.J. Nedrud J.G. ( 1998) Host response and vaccine develoment to H. pylori infection. Curr. Top. Microbiol. Immunol.  241, 181. [20] Del Giudice G. Rappuoli R. ( 1999) Immunisation against and Treatment for Infection by H. pylori. Patent WO 99/57278. [21] Radcliff F.J. Hazell S.L. Kolesnikow T. Doidge C. Lee A. ( 1997) Catalase, a novel antigen for Helicobacter pylori vaccination. Infect. Immun.  65, 4668– 4674. Google Scholar PubMed  [22] Alm R.A. et al.   . ( 1999) Genomic sequence comparison of two unrelated isolates of the human gastric pathogen H. pylori. Nature  397, 176 Google Scholar CrossRef Search ADS PubMed  [23] Tomb J.F. et al.   ( 1997) The complete genome sequence of the gastric pathogen Helicobacter pylori. Nature  388, 539– 547. Google Scholar CrossRef Search ADS PubMed  [24] Sodoyer R. Mazarin V. Guy B. Pion C. Dupuy M. Peubez I. Fourage S. Colombani S. Fourrichon L. Dubayle J. Hessler C. Berard Y. Seguin D. Deforges S. Al-Garawi A. Miller C. Oomen R. Quentin Millet M.J. Kleanthous H. Lissolo L. ( 1998) Post-genomic strategies for vaccine target identification against Helicobacter pylori. 98th General Meeting of ASM , Atlanta, GA, May 1998. [25] Brunel F. Darbouret A. Ronco J. ( 1999) Cationic lipid DC Chol induces an improved and balanced immunity able to overcome the unresponsiveness to the Hepatitis B vaccine. Vaccine  17, 2192– 2203. Google Scholar CrossRef Search ADS PubMed  [26] Akhiani A. Kjerrulf M. Redline R. Nedrud J. Czinn S. Lycke N. ( 1998) Th1 and Th2 cells contribute to protection but are also involved in the development of gastritis. Abstract C1. European Meeting on H. pylori Immunology and Pathology , Helsingor. [27] Ermak T.H. et al.   ( 1998) Immunisation of mice with urease vaccine affords protection against Helicobacter pylori infection in the absence of antibodies and is mediated by MHC class II-restricted responses. J. Exp. Med.  188, 1– 12. Google Scholar CrossRef Search ADS PubMed  [28] Sawai N. Kita M. Kodama T. Tanahashi T. Yamaoka Y. Tagawa Y.-I. Awakura Y. Imanishi J. ( 1999) Role of IFNγ in H. pylori induced gastric inflammatory responses in a mouse model. Infect. Immun.  67, 279– 285. Google Scholar PubMed  [29] Odenbreit S. Till M. Hofreuter D. Faller G. Haas R. ( 1999) Genetic and functional characterisation of the alpAB gene locus essential for the adhesion of Helicobacter pylori to human gastric tissue. Mol. Microbiol.  31, 1537– 1548. Google Scholar CrossRef Search ADS PubMed  [30] Ilver D. Arnqvist A. Ögren J. Frick I.M. Kersulyte D. Incecik E.T. Berg D.E. Covacci A. Engstrand L. Boren T. ( 1998) Helicobacter pylori adhesin binding fucosylated histo-blood group antigens revealed by retagging. Science  279, 373– 377. Google Scholar CrossRef Search ADS PubMed  [31] Su B. Hellström P.M. Rubio C. Celik J. Granström M. Normark S. ( 1998) Type I Helicobacter pylori shows Lewisb-independent adherence to gastric cells requiring de novo protein synthesis in both host and bacteria. J. Infect. Dis.  178, 1379– 1390. Google Scholar CrossRef Search ADS PubMed  [32] Kleanthous H. Tibbits T. Bahios T.J. Georgokopoulos K. Myers G. Ermak T.H. Fox J. Monath T. ( 1995) In vivo selection of a highly adapted H. pylori isolate and the development of an H. pylori mouse model for studying vaccine efficacy. In: VIIIth International Workshop on Gastroduodenal Pathology and Helicobacter pylori , 7–9 July 1995, Edinburgh. [33] Forsyth M.H. Cover T.L. ( 2000) Intercellular communication in Helicobacter pylori: luxS is essential for the production of an extracellular signaling molecule. Infect. Immun.  68, 3193– 3199. Google Scholar CrossRef Search ADS PubMed  [34] Joyce E.A. Bassler B.L. Wright A. ( 2000) Evidence for a signaling system in H. pylori: detection of a luxS-encoded autoinducer. J. Bacteriol.  182, 3638– 3643. Google Scholar CrossRef Search ADS PubMed  [35] Benaissa M. Babin P. Quellard N. Pezennec L. Cenatiempo Y. Fauchètre J.L. ( 1996) Changes in H. pylori ultrastructure and antigens during conversion from the bacillary to the coccoid form. Infect. Immun.  64, 2331– 2335. Google Scholar PubMed  [36] Cole S.P. Cirillo D. Kagnoff M.F. Guiney D.G. Eckmann L. ( 1997) Coccoid and spiral H. pylori differ in their abilities to adhere to gastric epithelial cells and induce IL8 secretion. Infect. Immun.  65, 843– 846. Google Scholar PubMed  [37] Mitchell H. Hazell S. Kolesnikow T. Mitchell J. Frommer D. ( 1996) Antigen recognition during progression from acute to chronic infection with a CagA positive strain of H. pylori. Infect. Immun.  64, 1166– 1172. Google Scholar PubMed  [38] D'Elios M.M. Manghetti M. de Carli M. Costa F. Baldari C.T. Burroni D. Telford J.L. Romagnani S. Del Prete G. ( 1997) Th1 effector cels specific for H. pylori in the gastric antrum of patients with peptic ulcer disease. J. Immunol.  158, 962– 967. Google Scholar PubMed  [39] Mohammadi M. Czinn S. Redline R. Nedrud J. ( 1996) Helicobacter specific cell-mediated immunity responses display a predominant Th1 phenotype and promote a delayed type hypersensitivity response in the stomachs of mice. J. Immunol.  156, 4729– 4738. Google Scholar PubMed  [40] Kang I. Barthold S.W. Persing D.H. Bockenstedt L.K. ( 1997) T helper cell cytokines in the early evolution of murine Lyme arthritis. Infect. Immun.  65, 3107– 3111. Google Scholar PubMed  [41] Dubois A. Berg D.E. Incecik E.T. Fiala N. Heman-Ackah L.M. PerezPerez G.I. Blaser M. ( 1996) Transient and persistent experimental infection of non-human primates with Helicobacter pylori: implications for human disease. Infect. Immun.  64, 2885– 2891. Google Scholar PubMed  [42] Saldinger P.F. Porta N. Launois P. Louis J.A. Waanders G.A. Bouzourène H. Michetti P. Blum A.L. Corthesy-Theulaz I. ( 1998) Immunisation of Balb/c mice with H. pylori B induces a T helper 2 response absent in Helicobacter infection. Gastroenterology  115, 891– 897. Google Scholar CrossRef Search ADS PubMed  [43] Ismail N. Bretscher P.A. ( 1999) The Th1/Th2 nature of concurrent immune responses to unrelated antigens can be independent. J. Immunol.  163, 4842– 4850. Google Scholar PubMed  [44] Graham D.Y. Opekun A.R. Osato M.S. El-Zimaity H.M. Cadoz M. Qureshi W.A. Monath T.P. ( 1999) H. pylori vaccine development in humans: challenge model: 3rd Annual Winter H. pylori Workshop , Orlando, FL, 26–27 Februart, Abstract. [45] Weder H.G. Zumbuehl O. ( 1984) The preparation of variably sized homogeneous liposomes for laboratory, clinical, and industrial use by controlled detergent dialysis. In: Liposome Technology  ( Gregoriadis G., Ed.), Vol. 1, pp. 79– 105, CRC Press, Boca Raton, FL. © 2001 Federation of European Microbiological Societies.
TI - Formulations of single or multiple H. pylori antigens with DC Chol adjuvant induce protection by the systemic route in mice Optimal prophylactic combinations are different from therapeutic ones
JF - Journal of the Endocrine Society
DO - 10.1111/j.1574-695X.2001.tb01565.x
DA - 2001-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/formulations-of-single-or-multiple-h-pylori-antigens-with-dc-chol-S2xUOaAhoj
SP - 157
EP - 165
VL - 30
IS - 2
DP - DeepDyve
ER -