2′-Fluoro-c-di-GMP as an oral vaccine adjuvant

Bis-(3′–5′)-cyclic dimeric 2′-deoxy-2′-fluoroguanosine monophosphate (2′-F-c-di-GMP) was synthesized through the modified H-phosphonate chemistry. Oral immunization of C57BL/6 mice with Helicobacter pylori cell-free sonicate extract adjuvanted with 2′-F-c-di-GMP led to the production of antigen-specific antibodies in feces and sera, and lowered bacterial counts in the stomach upon post-vaccination infections in immunized mice. Similarly, oral vaccination of BALB/c mice with flagillin proteins from Clostridium difficile and Listeria monocytogenes adjuvanted with 2′-F-c-di-GMP led to production of antigen-specific antibodies both systemically and mucosally. The adjuvanticity of 2′-F-c-di-GMP is associated with the enhanced induction of interferon γ. These results demonstrated the excellent oral adjuvanticity of 2′-F-c-di-GMP.


Introduction
In the last decade or so, 3 0 ,5 0 -cyclic diguanylic acid (c-di-GMP) has been recognized as a potent immunostimulator and a useful mucosal adjuvant in a number of models. 1,2 It was previously demonstrated by us that intranasal administration of c-di-GMP prior to bacterial challenges provides mice with protection against Acinetobacter baumannii infection by chemokine induction and enhanced neutrophil recruitment. 3 Furthermore, we showed that intranasal immunization of mice with pneumococcal surface adhesion A (PsaA) adjuvanted with c-di-GMP invoked strong antigen-specic serum immunoglobulin G (IgG) and secretory IgA antibody responses, and the nasopharyngeal Streptococcus pneumoniae colonization in immunized mice was signicantly reduced. 4 In the present study, we wish to demonstrate the adjuvanticity of c-di-GMP and its 2 0 -uoro-analog (2 0 -F-c-di-GMP) in oral immunization of mice against Helicobacter pylori. In this respect, uorine atoms are small and electronegative. Incorporation of uorine at the 2 0 -position of nucleosides is an effective approach to modulate sugar puckers. 5 Furthermore, introduction of uorine into therapeutic agents has been well recognized as a useful modication to modulate pharmacological properties. [6][7][8][9] We report herein that oral immunization of C57BL/6 mice with H. pylori cell-free sonicate extract (HPCE) adjuvanted with 2 0 -F-c-di-GMP led to the production of antigen-specic antibodies, and provide excellent protective immunity of immunized mice against H. pylori challenges. In a similar manner, productions of antigen-specic antibodies were also demonstrated in mice immunized with agillin proteins from Gram-positive bacterium Clostridium difficille and intracellular pathogen Listeria monocytogenes.
In the modied H-phosphonate approach the fully protected linear dimer phosphorothioate triester 5 was prepared by reacting H-phosphonate triethylammonium salt 3 and 3 0 -O-Lev nucleoside 4 in the presence of pivaloyl chloride followed by oxidation with 1-phenylsulfanyl-pyrrolidine-2,5-dione 11. Aer removal of 3 0 -O-levulinyl group from the dimer 5 by the treatment with hydrazine hydrate in pyridine-acetic acid solution, the product was transformed into the corresponding linear dimer H-phosphonate triethylammonium salt, followed by removal of 5 0 -O-dimethoxytrityl group. The resulting linear dimer H-phosphonate 7 was then subjected to cyclization under high dilution conditions to give the fully protected 2 0 -deoxy-2 0 -uoro cyclic diguanylic acid 8 in 59% yield. study, we rst determined if i.n. immunization of mice with 2 0 -F-c-di-GMP can elicit antigen specic mucosal immune responses at a comparable magnitude to those induced by the parental c-di-GMP. As shown in Fig. 1, co-administration of pneumococcal protein PsaA with 2 0 -F-c-di-GMP induced higher levels of PsaA-specic IgA in feces and vaginal wash, and serum IgG2a than the co-administration with c-di-GMP whereas the serum PsaA-specic IgA and IgG1 levels were comparable between the two adjuvants. As expected, sham-immunized mice showed no specic antibody responses in the serum or mucosal samples.
More importantly, we found that the mucosal immune responses induced by the i.n. immunization with 2 0 -F-c-di-GMP adjuvanted vaccine were protective against mucosal infections in the well-established mouse S. pneumoniae colonization model (Fig. 2) in that mice i.n. immunized with PsaA + 2 0 -F-c-di-GMP showed signicantly reduced colonization of S. pneumoniae when compared to sham-immunized mice (P < 0.05). The magnitude of this reduction was comparable to that attained in mice immunized with PsaA adjuvanted with cholera toxin (CT), 4 the golden standard of mucosal adjuvant which has undesirable toxicity for human applications. We have previously shown that immunization with PsaA alone at this dose showed no effect on the bacterial colonization. 4 These results demonstrated that 2 0 -Fc-di-GMP is a potent mucosal adjuvant when administered by intranasal route, and that 2 0 -F-c-di-GMP induces a potent, protective immunity against i.n. challenge with S. pneumoniae when co-administered with the PsaA antigen via i.n. route. Therefore, further exploration of this molecule as a potential mucosal adjuvant is warranted.
Oral immunization with 2 0 -F-c-di-GMP-adjuvanted vaccine induces strong antigen-specic antibody responses in the serum and at multiple mucosal sites Despite the well-recognized socioeconomic and safety advantages of oral immunization over the parenteral or i.n. immunization, only a limited number of oral vaccines are currently approved for human use. 15 Oral vaccination is the most challenging vaccination method due to the administration route. Indeed, we found Fig. 1 Induction of antigen-specific mucosal IgA responses by intranasal administration of 2 0 -F-c-di-GMP. Groups of 5 BALB/c mice were intranasally immunized with 2 mg PsaA admixed with 2.5 mg 2 0 -F-c-di-GMP, 2.5 mg c-di-GMP or 10 mg c-di-GMP at day 0, 14 and 21. Additional group of mice were immunized with phosphate-buffered saline (PBS) and served as sham-immunized group. The feces, vaginal washing and blood samples were collected at day 28 and assayed by ELISA for PsaA-specific IgA and IgG isotypes (IgG1 and IgG2a) responses. *P < 0.05 vs. PsaA + c-di-GMP groups.
that oral administration of the parental c-di-GMP as a mucosal adjuvant failed to induce reliable mucosal or systemic immune responses (unpublished data). In this study, we therefore assessed if oral administration of 2 0 -F-c-di-GMP induces antigen-specic mucosal immune responses. As shown in Fig. 3, oral coadministration of both high and low doses of HPCE with 2 0 -F-cdi-GMP induced substantial amount of antigen-specic fecal IgA and serum IgG2a responses, which were similar in the magnitude to those induced by CT (Fig. 3A). As expected, shamimmunized mice showed no specic antibody responses in the serum or fecal samples. Similarly, oral co-administration of 2 0 -Fc-di-GMP with agellin antigens puried from L. monocytogenes (50 mg) or C. difficile (30 mg) induced substantial amount of C. difficile agellin-specic IgA and small amount of Listeria agellin-specic IgA in feces as well as serum IgG1 and IgG2a responses, as compared with sham-immunized mice (Fig. 3B). These results demonstrated that 2 0 -F-c-di-GMP enhances mucosal immune responses to microbial antigens when administered via the oral route, and indicate that 2 0 -F-c-di-GMP can be used in oral vaccines as a potent mucosal adjuvant.
Oral immunization of mice with 2 0 -F-c-di-GMP-adjuvanted HPCE signicantly reduces gastric colonization by H. pylori We next determined if the mucosal immune responses induced by the 2 0 -F-c-di-GMP adjuvanted H. pylori oral vaccine protect against H. pylori challenge in a mouse model of H. pylori infection. 16 As shown in Fig. 4, quantitative bacteriology showed that oral immunization of mice with both low (250 mg) and high (500 mg) doses of H. pylori HPCE + 2 0 -F-c-di-GMP vaccines signicantly reduced the bacterial burdens in the gastric mucosa at 4 weeks post-challenge when compared to shamimmunized mice (P < 0.001). Moreover, the magnitude of this reduction was comparable to that attained in mice immunized with HPCE adjuvanted with CT. As anticipated, immunization of mice with HPCE alone failed to reduce the bacterial colonization. These results suggest that 2 0 -F-c-di-GMP is capable of inducing protective mucosal immunity against mucosal pathogens upon oral immunization with specic antigen.
Oral immunization with 2 0 -F-c-di-GMP-adjuvanted vaccine induces antigen-specic Th1/Th17 cytokine responses Previous studies by others have implied that Th1/Th17 immune responses may play an important role in host defence against H. pylori infection. 17 We next examined if oral administration of 2 0 -  F-c-di-GMP + HPCE induced antigen-specic Th1/Th17 cytokine responses. Compared to sham-immunized mice, the splenocytes from 2 0 -F-c-di-GMP immunized mice produced substantially higher amount of IFN-g, IL-2 and IL-17 in response to HPCE stimulation (Fig. 5). The amount of IFN-g produced by the splenocytes from 2 0 -F-c-di-GMP immunized mice was even higher than that produced by the cells from CT-immunized mice although the latter mice produced higher amount of IL-2 and IL-17 than the former mice. These results indicate that the protection against H. pylori infection induced by oral immunization with 2 0 -F-c-di-GMP in mice is associated with the production of antigen-specic IFN-g.

Experimental
Synthesis of compounds 3.51 mmol) was dried in vacuo at 60 C for 5 h followed by addition of dry pyridine (10.0 ml). Aer the solution was cooled (ice-water bath), chlorotrimethylsilane (2.23 ml, 17.6 mmol) was added and the reaction mixture was stirred for 30 min. The mixture was then evaporated to ca. half of the original volume. To the residue, dry pyridine (5.0 ml) was added and mixture was cooled (ice-water bath) followed by dropwise addition of isobutyric anhydride (3.02 ml, 18.2 mmol). Aer 3 h, the reaction was quenched by addition of water (3.0 ml) followed aer 15 min aqueous ammonium hydroxide (30-33%, 10.0 ml). The products were rst evaporated under reduced pressure while the temperature was kept below 10 C. Aer the bulk ammonia was removed, the products were evaporated to dryness under reduced pressure while the temperature was kept below 35 C. The residue was co-evaporated with dry pyridine (2 Â 10 ml) and then dissolved in dry pyridine (15.0 ml) followed by addition of 4,4 0 -dimethoxytrityl chloride (1.15 g, 3.39 mmol). Water (2.0 ml) was added aer 30 min to quench the reaction. The products were concentrated under reduced pressure and the residue was dissolved in dichloromethane (25 ml) and extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated, and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were dried (MgSO 4 ) and concentrated with a rotary evaporator. The residue was puried by column chromatography on silica gel. The appropriate fractions, which were eluted with dichloromethanemethanol (98 : 2 v/v) containing 0.5% triethylamine, were concentrated under reduced pressure to give the title compound as a colourless glass (1.83 g, 79%).
1.00 g, 1.52 mmol) was co-evaporated with dry toluene (2 Â 5 ml) and then dissolved in dry pyridine (10.0 ml) followed by addition of levulinic anhydride (0.65 g, 3.04 mmol). Aer 16 h, Additional group of mice were immunized with PBS and served as sham-immunized group. The mice were orally challenged 3Â between day 35 and 42 with 10 8 CFU H. pylori SS1 and the bacterial numbers in the gastric mucosa of challenged mice were determine 4 weeks later. ***P < 0.001 vs. immunized groups. Additional group of mice were immunized with PBS and served as shamimmunized group. The spleens were collected at day 28 and single cell suspension was prepared for the determination of antigen-specific cytokine responses. The cells were stimulated by 10 mg ml À1 HPCE and cultured at 37 C in a 5% CO 2 atmosphere for 72 hours. At the end of the culture, the supernatant was collected and assayed for the levels of IFN-g, IL-2 and IL-17 by Luminex. *P < 0.05 and ***P < 0.001 vs. PBS group.
This journal is © The Royal Society of Chemistry 2019 RSC Adv., 2019, 9, 41481-41489 | 41485 water (3.0 ml) was added and the solution was concentrated under reduced pressure. The residue was dissolved in dichloromethane (25 ml) and extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were dried (MgSO 4 ) and evaporated under reduced pressure. The residue was coevaporated with dry toluene (3 Â 5 ml) and dissolved in dry dichloromethane (10.0 ml). Distilled pyrrole (1.05 ml, 15.1 mmol) followed by dichloroacetic acid (0.56 ml, 6.79 mmol) were added. Aer 10 min, the products were extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were dried (MgSO 4 ) and concentrated under reduced pressure. The residue was puried by column chromatography on silica gel. The appropriate fractions, which were eluted with dichloromethane-methanol (97 : 3 v/v), were combined and concentrated under reduced pressure to give the title compound as a colourless glass (0.51 g, 74%   10 g, 1.67 mmol), and the mixture was co-evaporated with dry pyridine (2 Â 5 ml) and then dissolved in dry pyridine (25 ml). To the cooled (ice-water bath) mixture was added dropwise pivaloyl chloride (0.68 ml, 5.55 mmol) over 5 min. Aer 1 h water (5.0 ml) was added. Aer a further period of 1 h at room temperature, the products were concentrated under reduced pressure and the residue was dissolved in dichloromethane (25 ml) and extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were extracted with triethylammonium phosphate buffer (20 ml, 0.5 M, pH 7.0) and the organic layer was back extracted by dichloromethane (15 ml). The combined organic layers were dried (MgSO 4 ) and concentrated under reduced pressure. The residue was puried by short column chromatography on silica gel. The appropriate fractions, which were eluted with dichloromethane-methanol (90 : 10 v/v), were pooled and concentrated under reduced pressure to give triethylammonium salt of the title compound as a colourless glass (1.18 g, 86%). ESI-MS found M À ¼ 720. DMTr-G 0 p(s 0 )G 0 -Lev 5. 2 0 -Fluoro-3 0 -O-levulinyl-2-N-isobutyryl-2 0deoxyguanosine 4 (0.44 g, 0.97 mmol) and 2 0 -uoro-5 0 -O-dimethoxytrityl-2-N-isobutyryl-2 0 -deoxyguanosine 3 0 -H-phosphonate triethylammonium salt 3 (1.00 g, 1.22 mmol) were co-evaporated with dry pyridine (2 Â 5 ml) and then dissolved in dry pyridine (5.0 ml) and cooled (ice-water bath). Pivaloyl chloride (0.31 ml, 2.53 mmol) followed by 1-phenylsulfanyl-pyrrolidine-2,5-dione 11 (0.56 g, 2.20 mmol) were added aer 5 min. The reaction mixture was stirred for 30 min at room temperature, and then water (0.2 ml) was added. Aer 10 min the products were concentrated under reduced pressure. The residue was dissolved in dichloromethane (25 ml) and extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were dried (MgSO 4 ) and concentrated under reduced pressure. The residue was puried by column chromatography on silica gel. The appropriate fractions, which were eluted with dichloromethane-methanol (98 : 2 v/v) containing 0.5% of triethylamine, were pooled and concentrated under reduced pressure to give the title compound as a colourless glass (1.35 g, 109%. The yield is over 100%, due to inseparable impurities, likely phthalimide DMTr-G 0 p(s 0 )G 0 -OH 6. DMTr-G 0 p(s 0 )G 0 -Lev 5 (1.20 g, 0.95 mmol) was dissolved in pyridine (10.0 ml) followed by addition of a mixture of hydrazine monohydrate (0.40 ml, 8.24 mmol), acetic acid (4.8 ml), water (0.8 ml) and pyridine (18.0 ml). The reaction mixture was stirred at room temperature for 15 min followed by addition of pentane-2,4-dione (1.20 ml). Aer 10 min the products were concentrated under reduced pressure. The residue was dissolved in dichloromethane (25 ml) and extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were dried (MgSO 4 ) and concentrated under reduced pressure. The residue was puried by column chromatography on silica gel. The appropriate fractions, which were eluted with dichloromethanemethanol (95 : 5 v/v), were concentrated under reduced pressure to give the title compound as a colourless solid (1.05 g, 95% HO-G 0 p(s 0 )G 0 p(H) 7. Ammonium p-tolyl H-phosphonate 10 (0.57 g, 3.01 mmol) was co-evaporated with triethylamine (0.84 ml, 6.03 mmol) and methanol (5.3 ml). To the residue was added DMTr-G 0 p(s 0 )G 0 -OH 6 (1.03 g, 0.88 mmol) and the mixture was co-evaporated with dry pyridine (2 Â 5 ml) and then dissolved in more dry pyridine (10.0 ml). To this cooled (ice-water bath) reaction mixture was added pivaloyl chloride (0.37 ml, 3.0 mmol) over a period of 5 min. Aer 30 min water (1.0 ml) was added. The products were allowed to warm up to room temperature and stirring was continued for another 1 h. The products were concentrated under reduced pressure and the residue was dissolved in dichloromethane (25 ml) and extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were dried (MgSO 4 ) and concentrated under reduced pressure. The residue was co-evaporated with dry toluene (3 Â 10 ml) and then dissolved in dry dichloromethane (20 ml). Freshly distilled pyrrole (1.04 ml, 15.0 mmol) and dichloroacetic acid (0.86 ml, 10.4 mmol) were added. The reaction mixture was stirred for 10 min and then extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were extracted with triethylammonium phosphate buffer (20 ml, pH 7.0, 0.5 M) and the aqueous layer was back extracted by dichloromethane (15 ml Preparation of fully protected 2 0 -uoro-c-di-GMP 8. HO-G 0 p(s 0 )G 0 p(H) triethylammonium salt 7 (100 mg, 0.097 mmol) was co-evaporated with dry pyridine (2 Â 3 ml) and then dissolved in dry dichloromethane (5.0 ml). This solution was added to a solution of diphenyl chlorophosphate (0.41 ml, 1.98 mmol) in dry pyridine (5.0 ml) dropwise over 20 min while the temperature was kept at À40 C (dry ice-acetone bath). Aer 20 min 1-phenylsulfanyl-pyrrolidine-2,5-dione 11 (0.45 g, 1.76 mmol) was added and the reaction mixture was allowed to warm up to room temperature. Stirring was continued for another 30 min and then water (0.5 ml) was added. Aer 5 min the products were concentrated under reduced pressure. The residue was dissolved in dichloromethane (25 ml) and extracted with saturated aqueous sodium hydrogen carbonate (15 ml). The layers were separated and the aqueous layer was back extracted with dichloromethane (2 Â 5 ml). The combined organic layers were dried (MgSO 4 ) and concentrated under reduced pressure. The residue was puried by short column chromatography on silica gel. The appropriate fractions, which were eluted with dichloromethane-methanol (97 : 3 v/v), were combined and concentrated under reduced pressure to give the fully protected 2 0 -uoro-c-di-GMP 8 as a colourless glass (58 mg, 59% 2 0 -Fluoro-c-di-GMP 9. Fully protected 2 0 -uoro-c-di-GMP 8 (50 mg, 0.049 mmol) was co-evaporated with dry toluene (2 Â 2 ml) and then dissolved in dry acetonitrile (3.0 ml), followed by addition of 2-nitrobenzaldoxime 13 (0.12 g, 0.72 mmol) and N,N,N 0 ,N 0 -tetramethylguanidine (86 ml, 0.69 mmol). Aer 16 h the products were evaporated to dryness under reduced pressure and the residue was taken up in concentrated aqueous ammonium hydroxide (33%, 2.0 ml). The reaction mixture was sealed and heated at 55 C for 15 h. Aer the mixture was cooled, the products were concentrated to dryness under reduced pressure and then co-evaporated with ethanol (2 Â 2 ml). The residue was dissolved in methanol (1.0 ml) and precipitated from diethyl ether (30.0 ml). The product was obtained by centrifugation at 5000 rpm. This precipitationcentrifugation process was repeated 3 more times. The residue was dissolved in HPLC grade water (0.2 ml) followed by addition of 1-butanol (10 Â volume). The mixture was vortexed and then frozen. Aer centrifugation at 10 000 rpm for 10 min, the supernatant was discarded. This 1-butanol extraction step was repeated two more times and the nal residue was dissolved in HPLC water (1.0 ml) and passed through an Amberlite (Na + form) cation exchange column (1 Â 20 cm). The fractions containing nucleic acid were freeze-dried to give the fully unprotected disodium salt of 2 0 -uoro-c-di-GMP 9 as a colourless sponge solid ( Antigens and adjuvants. c-di-GMP was synthesized according to procedures reported previously. 10 Cholera toxin (CT) was purchased from Sigma-Aldrich Canada Ltd (Oakville, Ontario, Canada). The recombinant pneumococcal surface adhesion A (PsaA) of S. pneumoniae and H. pylori cell-free sonicate extract (HPCE) were prepared as described in details previously. 4,18 For the preparation of agellins from L. monocytogenes and C. difficile, bacterial strains were grown overnight at 37 C on agar plates. Cells were scraped from surface and resuspended in 10 mM Tris pH 7.4. Flagella were sheared from surface of bacterial cells using a warring blender (4 Â 1 min). Bacterial cells were removed by low speed centrifugation and then agellar laments were collected from supernatant by ultracentrifugation (100 000 Â g, 1 hour). Flagellar samples were analysed by SDS-PAGE.
Mice. Six to eight-week-old female BALB/c or C57BL/6 mice were purchased from Charles Rivers Laboratories (St. Constant, Quebec). The animals were housed under specicpathogen-free conditions in a federally licensed small animal containment level 2 facility, and given free access to sterile water and certied mouse chow. The animals were maintained and used in accordance with the recommendations of the Canadian Council on Animal Care Guide to the Care and Use of Experimental Animals. All experimental procedures were approved by the institutional animal care committee (Institute for Biological Sciences, National Research Council Canada, Ottawa, Ontario).

Intranasal or oral immunization in mice
For intranasal immunization, mice were lightly anesthetized under isouorane. On day 0, 14 and 21, groups of mice (n ¼ 5) were inoculated with 50 ml of various vaccine preparations or controls as detailed in Results section. For oral immunization, various vaccine preparations or controls in 0.5 ml volume were administered by gavage via an 18-gauge feeding needle.
At day 28, blood and mucosal samples (fecal pellets or/and vaginal wash) were collected as described elsewhere. 19 In some experiments, nasal wash samples were also collected at the end of experiment. For both the vaginal and nasal wash, the volume of the lavage uid recovered from each mouse was recorded. Since the recovery of lavage volumes among different mice in our study were very similar (about 90%), the small variation among individual samples was not adjusted. All samples were stored at À20 C until assay.
Enzyme-linked immunosorbent assay (ELISA) for measurement of antigen-specic antibodies in the serum and mucosal samples Levels of antigen-specic antibodies in serum and mucosal samples were measured by an enzyme-linked immunosorbent assay (ELISA) modied from a previously described procedure. 16 Briey, 96-well microplates (Thermo Electron Corporation, Milford, MA) were coated with puried PsaA (5 mg ml À1 , 100 ml per well), puried agellins (5 mg ml À1 , 50 ml per well), or HPCE (20 mg ml À1 ; 50 ml per well) in 0.015 M sodium carbonate and 0.035 M sodium bicarbonate buffer (pH 9.6) at 4 C overnight. All the subsequent incubations were carried out at room temperature. The wells were blocked by incubation with 5% bovine serum in PBS for 1 h, and then rinsed three times with PBS-0.05% Tween 20. Duplicates of 100 ml pre-diluted samples were added to the wells. Sample dilutions were as follows: 1 : 2 for fecal extracts and nasal washes, 1 : 2000 for serum IgG1, 1 : 500 for serum IgG2a and 1 : 50 for serum IgA. Aer the plates were incubated for 3 h, alkaline phosphatase-conjugated goat antibodies specic for mouse IgA, IgG1 and IgG2a (all from Caltag Laboratories, Burlingame, CA) were added and incubated for 1 h. Colour reactions were developed by the addition of p-nitrophenyl phosphate (pNPP) substrate (KPL, Inc., Gaithersburg, MD), and optical density was measured at 405 nm with an automated ELISA plate reader (Multiskan Ascent, Thermo Labsystems, Vantaa, Finland). Pooled samples collected from mice that had been intranasally immunized with the relevant vaccines or from the naïve mice were used as positive or negative controls respectively.

Intranasal challenge of mice with S. pneumoniae
Fresh inocula were prepared for each experiment from a frozen stock of S. pneumoniae (type 14). Stock vials of S. pneumoniae were thawed and the culture revived on chocolate agar, which was then used to inoculate Todd-Hewitt and Columbia broth supplemented with 1% glucose and 0.1% sodium bicarbonate. The broth culture was incubated in a candle jar at 37 C for approximately 6 h. The broth culture in mid-logarithmic growth phase was centrifuged at 12 000 Â g for 15 min, cells were resuspended in PBS and used immediately. Fourteen days aer the last immunization, mice were anesthetized and intranasally inoculated with approximately 10 7 colony-forming units (CFU) S. pneumoniae in 50 ml saline. The actual inoculum concentrations were determined by plating 10-fold serial dilutions on chocolate agar. Inoculated mice were sacriced 3 days later and the nasal cavity was lavaged with 0.5 ml lavage uid and aliquots (100 ml) of 10-fold serial dilutions of the lavage uids were cultured, in duplicates, on chocolate agar plates to quantify the number of viable organisms.

H. pylori and oral H. pylori infection
The mouse-adapted H. pylori SS1 that was established by Lee et al. 20 was used as the challenge strain. Bacteria were grown on brain-heart infusion (BHI) broth supplemented with 5% horse serum under a microaerophilic atmosphere created by a Cam-pyGen generator (Oxoid Ltd, Hampshire, England) at 37 C for 48 h. The bacterial cells were harvested by centrifugation at 12 000 Â g for 10 min at 4 C, washed with PBS three times and then suspended in PBS.
The mice were inoculated orally with 1 Â 10 8 CFU of freshly harvested H. pylori in 500 ml BHI broth by using a 18-gauge feeding needle as described previously. 16 The inoculations were repeated twice (a total of three inoculations) over a period of 5 days. Mice were killed by an overdose of carbon dioxide 4 weeks aer the last inoculation, and the serum and stomachs of the animals were collected for analysis. In some experiments, spleens were removed for splenocyte culture and determination of cytokine responses.

Assessment of H. pylori infection
The presence of H. pylori infection in individual mice was determined by quantitative bacterial culture as described previously. 16 Briey, the mouse stomach was opened longitudinally along the greater curvature and gently washed three times in PBS to remove the stomach contents. The stomach tissue was then homogenized in 5 ml saline with Teon pestles and glass tubes. The homogenates were serially diluted in saline, and 100 ml of the dilutions were plated onto GSSA (Glaxo Selective Supplement Antibiotics)-supplemented agar plates. The plates were incubated for 96 h at 37 C under microaerophilic conditions, and the number of colonies were counted and expressed as CFU per stomach. With this method, log 10 2.7 CFU per stomach represented the limit of detection.

Splenocyte culture and cytokine assay
Spleens were aseptically removed and placed in 5 ml Dulbecco's modied Eagle's medium (DMEM). Single cell suspensions were prepared by pressing the spleen through a 100 mm Nylon cell strainer into a sterile Petri dish, using the rubber end of a 1 ml syringe plunger. Erythrocytes from spleens were lysed using an ACK lysis buffer. Cells were washed one time with DMEM and resuspended at a concentration of 3 Â 10 7 cells per ml in DMEM supplemented with 10% fetal bovine serum (FBS). Cells were seeded in duplicate on a 96-well cell culture plate and stimulated with the mixed H. pylori antigens (10 mg ml À1 each) for 48 h at 37 C and 5% CO 2 . Culture supernatants were collected and assayed for the concentrations of IFN-g, IL-2 and IL-17 using Luminex kits. 21