Rotavirus 2/6 Virus-Like Particles Administered Intranasally in Mice, with or without the Mucosal Adjuvants Cholera Toxin and Escherichia coli Heat-Labile Toxin, Induce a Th1/Th2-Like Immune Response

Abstract

We investigated the rotavirus-specific lymphocyte responses induced by intranasal immunization of adult BALB/c mice with rotavirus 2/6 virus-like particles (2/6-VLPs) of the bovine RF strain, by assessing the profile of cytokines produced after in vitro restimulation and serum and fecal antibody responses. The cytokines produced by splenic cells were first evaluated. Intranasal immunization with 50 μg of 2/6-VLPs induced a high serum antibody response, including immunoglobulin G1 (IgG1) and IgG2a, a weak fecal antibody response, and a mixed Th1/Th2-like profile of cytokines characterized by gamma interferon and interleukin 10 (IL-10) production and very low levels of IL-2, IL-4, and IL-5. Intranasal immunization with 10 μg of 2/6-VLPs coadministered with the mucosal adjuvants cholera toxin and Escherichia coli heat-labile toxin (LT) considerably enhanced the Th1/Th2-like response; notably, significant levels of IL-2, IL-4, and IL-5 were observed. Since rotavirus is an enteric pathogen, we next investigated the production of IL-2 and IL-5, as being representative of Th1 and Th2 responses, by Peyer's patch and mesenteric lymph node cells from mice immunized intranasally with 2/6-VLPs and LT. The results were compared to those obtained from splenic and cervical lymph node cells. We found that both cytokines were produced by cells from each of these lymphoid tissues. These results confirm the Th1/Th2-like response observed at the systemic level and show, on the assumption that T cells are the primary cells producing the cytokines after in vitro restimulation, that rotavirus-specific T lymphocytes are present in the intestine after intranasal immunization with 2/6-VLPs and LT.

Rotaviruses are the most common etiologic agents of severe acute gastroenteritis in infants and young children throughout the world (8). The development of safe effective rotavirus vaccines is an international priority, and significant efforts have focused on live rotavirus vaccines (10). New strategies including nonreplicating rotavirus subunits, such as virus-like particles (VLPs), which are stable noninfectious particles that are morphologically and antigenically similar to the native virus, are also being investigated (1–4, 7, 11, 16, 17, 28). Rotavirus 2/6 VLPs (2/6-VLPs) are made by coinfecting insect cells with recombinant baculoviruses expressing VP2 and VP6, the major core and intermediate layer proteins, which self assemble spontaneously. Such VLPs have been shown to induce antibody responses and protection from rotavirus challenge, notably in rabbits after parenteral immunization (1) and in mice treated orally and intranasally (16, 17). Until now, the profile of cytokines induced by VLPs has not been investigated. In a previous study, we showed that spleen cells from suckling mice orally inoculated with highly replicative homologous or weakly replicative heterologous strains of rotavirus produced a mixed Th1/Th2 pattern of cytokines, including gamma interferon (IFN-γ), interleukin 5 (IL-5), and IL-10, after in vitro restimulation (5). In the present study, we first investigated the rotavirus-specific lymphocyte responses induced by intranasal immunization of adult BALB/c mice with 2/6-VLPs of the bovine RF strain, by assessing antibody responses as well as the profile of cytokines produced by splenocytes after in vitro restimulation with purified virus. We also investigated the effects of cholera toxin (CT) and Escherichia coli heat-labile-toxin (LT) adjuvants on these responses. Indeed, both adjuvants have been shown to significantly increase antibody responses and protection from rotavirus challenge in mice immunized with VLPs (16, 17). We report that 2/6-VLPs administered intranasally induced a mixed Th1/Th2 profile of cytokines very similar to that induced by live rotavirus administered orally to suckling mice and that both CT and LT considerably enhanced the mixed Th1/Th2-like immune response induced by 2/6-VLPs. Since rotavirus is an enteric pathogen and since there is a substantial degree of compartmentalization within the mucosal immune system (26), we next investigated the production of IL-2 and IL-5, as being representative of Th1 and Th2 responses, by Peyer's patch (PP) and mesenteric lymph node (MLN) cells from mice immunized with 2/6-VLPs in the presence of LT. This immunization schedule was selected because of the high cytokine production observed in the first experiments and of enhanced protection when VLPs are administered with LT (16). The results were compared to those obtained from splenic and cervical lymph node (CLN) cells. Both cytokines were found to be produced by rotavirus-specific lymphocytes from each of these lymphoid tissues after in vitro restimulation.

MATERIALS AND METHODS

Cells and viruses.

The MA-104 cell line, derived from African green monkey kidney, was maintained in Eagle's modified minimal essential medium supplemented with 2 mM l-glutamine, nonessential amino acids, 100 U of penicillin per ml, 100 μg of streptomycin per ml, and 5% heat-inactivated fetal calf serum (FCS).

The SA11 strain of simian rotavirus (serotype G3) and the RF strain of bovine rotavirus (serotype G6) were propagated in MA-104 cells in the presence of 0.5 μg of trypsin per ml. Partially purified virus for in vitro restimulation and for enzyme-linked immunosorbent assay (ELISA) coating was prepared by freeze-thawing culture flasks and subjecting tissue culture homogenates to ultracentrifugation through a 40% (wt/vol) sucrose cushion in 50 mM Tris-HCl (pH 8.0)–10 mM NaCl–2 mM CaCl₂. Protein concentrations were determined using a Bradford protein assay (Bio-Rad, Munich, Germany) with bovine serum albumin (BSA) as the standard, and the purity of the preparation was confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis.

The baculovirus recombinants used in this study were BacRF2A (9) and BacRF6 (22), which code, respectively, for the proteins VP2 and VP6 of the bovine rotavirus RF. Spodoptera frugiperda 9 (Sf9) insect cells were grown and maintained in TNM-FH medium (PharMingen) containing 10% FCS.

VLP preparation and characterization.

2/6-VLPs were prepared in Sf9 cells coinfected at a high multiplicity of infection with BacRF2A and BacRF6 and purified as previously described (9). Briefly, infected cells were collected 5 days postinfection and then extracted with Freon 113, and the aqueous phase that contained VLPs was subjected to an isopycnic CsCl gradient in a 20 mM PIPES (piperazine-N,N′-bis[2-ethanesulfonic acid]) buffer with 10 μM CaCl₂ (pH 6.6) (18 h at 35,000 rpm in a Beckman SW55 Ti rotor). Purified 2/6-VLPs were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis to confirm their protein composition and examined by negative-staining electron microscopy. Protein concentration was determined using a Bradford protein assay with BSA as the standard.

Immunization and sample collection.

Adult female BALB/c mice (4 to 5 weeks old) were from our animal facilities for the first set of experiments and from Iffa-Credo (L'Arbresle, France) for the second one. No mice had evidence of previous rotavirus infection, as determined by serum antibody titers. Mice were immunized intranasally with 2/6-VLPs on days 0 and 14. Prior to intranasal inoculation, mice were anesthetized by intraperitoneal administration of a mixture of ketamine (80 mg/kg) and xylazine (16 mg/kg). The immunogen was given by gradual inoculation of the nostrils of the mice. The different groups of mice received 50 μg of VLPs alone or 10 μg of VLPs mixed with 5 μg of CT or LT (both from Sigma, St Louis, Mo.) per dose in a volume of 20 μl. Control mice received phosphate-buffered saline (PBS) with or without CT or LT. To determine if a part of the immune response observed could be attributed to the fact that a portion of the inoculum may be swallowed, a group of five mice was immunized by the oral route with 10 μg of 2/6-VLPs in the presence of 5 μg of LT. Blood and fecal samples were collected from each mouse on day 35 (21 days after the second immunization) and stored at −40°C prior to use. Mice were then sacrificed, and spleens as well as, in one set of experiments, superficial and posterior CLNs, PPs, and MLNs were removed.

Measurement of rotavirus-specific antibodies in serum and fecal samples by ELISA.

Antibody titers in serum and fecal samples were determined by ELISA. Microtiter plates were coated overnight at room temperature with 100 μl of sucrose-purified RF virus (0.5 μg/ml in 0.1 M carbonate buffer, pH 9.6). Wells were blocked with PBS containing 5% nonfat dry milk. Fecal samples were made 10% (wt/vol) by suspension in PBS, pH 7.4. Serial twofold dilutions in PBS–5% nonfat dry milk of serum (starting at 1/100) or fecal extracts (starting at 1/40) were added to wells and incubated for 40 min at 37°C. After three washes, the plates were incubated for 30 min at room temperature with a 1:5,000 dilution of biotin-labeled goat anti-mouse α, γ, γ1, or γ2a heavy chain-specific antisera (Southern Biotechnology Associates, Inc., Birmingham, Ala.). The plates were washed, and peroxidase-labeled avidin (Southern Biotechnology Associates) was added. The color reaction was developed at room temperature in the dark with the chromogenic substrate orthophenylenediamine with 0.03% H₂O₂, and A₄₉₂ was determined. Endpoint titers were expressed as the reciprocal log₁₀ of the last dilution that gave an optical density at least twofold greater than the mean value obtained with samples from uninfected mice for the same dilution (cutoff value of 0.1). Negative serum samples (titer < 100) and fecal samples (titer < 40) were arbitrarily assigned titers of 50 and 20 (twofold below 100 and 40), respectively, for statistical calculations.

Preparation of spleen cells and superficial and posterior CLN, PP, and MLN cells and in vitro restimulation.

Spleen and superficial and posterior CLNs, PPs, and MLNs from each mouse were aseptically removed, and single-cell suspensions were prepared by mechanical dissociation. Cells were washed with incomplete medium consisting of RPMI 1640 supplemented with 2 mM l-glutamine, 1 mM sodium pyruvate, 0.3% glucose, 100 U of penicillin per ml, and 100 μg of streptomycin per ml. After splenic erythrocytes had been removed by lysis with sterile water, cells (4 × 10⁶/ml) were resuspended in complete medium (incomplete medium plus 10% heat-inactivated FCS). Cells from either immunized or nonimmunized mice (4 × 10⁵ cells/well) were cultured in the presence of 5 μg of sucrose-purified virus (RF or SA11 strain) per ml or RPMI medium only in 96-well plates. In the second set of experiments, cells from spleens and superficial CLNs were also cultured in the presence of 5 μg of VLPs per ml. The T-cell mitogen concanavalin A (5 μg/ml) was added to positive control wells. The cells were incubated at 37°C with 5% CO₂. The culture supernatants were harvested at various intervals (days 2, 4, and 5) after restimulation, on the basis of preliminary experiments, and frozen at −70°C until cytokine analysis.

Cytokine ELISA.

Cytokine levels in culture supernatants were determined by ELISA. IFN-γ, IL-2, and IL-4 levels were determined using monoclonal antibodies (MAbs) obtained from PharMingen (San Diego, Calif.). MAbs used for coating were clones R4-6A2 (2 μg/ml), JES6-1A12 (2 μg/ml), and 11B11 (2 μg/ml), and MAbs used for detection were clones XMG1.2 (0.5 μg/ml), JES6-5H4 (0.5 μg/ml), and BVD6-24G2 (1 μg/ml) for IFN-γ, IL-2, and IL-4, respectively. Microtiter plates were coated with 50 μl of anticytokine MAb and incubated overnight at 4°C. The wells were blocked with PBS containing 3% BSA at room temperature for 2 h. After four washes with PBS containing 0.05% Tween 20, 100 μl of supernatants was added to duplicate wells and incubated overnight at 4°C. The plates were then washed and incubated with the appropriate biotinylated anticytokine MAb diluted in PBS with 3% BSA for 1 h. The plates were washed, and peroxidase-labeled avidin was added and incubated for 30 min. The color reaction was developed with ABTS (2,2′-azino-bis[3-ethylbenzthiazoline-6-sulfonic acid]) and 0.03% H₂O₂. Standard curves were generated using recombinant murine IFN-γ (PeproTech, Rocky Hill, N.J.), IL-4 (PharMingen), and IL-2 (PeproTech). The sensitivities of these ELISAs were 15 pg/ml for IL-4, 5 pg/ml for IL-2, and 0.2 ng/ml for IFN-γ. Low levels of IL-4 were assessed using an ultrasensitive kit (R&D Systems, Minneapolis, Minn.), and IL-5 and IL-10 levels were measured using licensed kits (Endogen, Cambridge, Mass.) following the manufacturer's recommendations. The sensitivities of these ELISAs were 2 pg/ml for IL-4, 5 pg/ml for IL-5, and 12 pg/ml for IL-10. For statistical analysis, levels of cytokine below the detection limit were recorded as one-half the detection limit.

Statistics.

Antibody responses and cytokine productions were both analyzed for comparison between the different groups with one-way analysis of variance, after verification of variance's homogeneity by using Hartley's table. For each antibody or cytokine analyzed, in the case of a significant one-way analysis of variance test, post hoc analysis comparing results between the different experimental arms was conducted by using a Newman-Keuls test. In the few cases where the measurements were performed for two groups only, a direct comparison was done using the Mann-Whitney nonparametric U test. Comparison of cytokine production under different conditions of restimulation within a given group were performed either by using a Newman-Keuls test after one-way analysis of variance or, when only two conditions were compared, by using the Mann-Whitney nonparametric U test. Similarly, within a given group, comparisons between serologic or fecal IgA and IgG responses were done by using the Mann-Whitney nonparametric U test. For all the tests, a P value of <0.05 was considered significant.

RESULTS

Fecal and serum antibody responses in mice inoculated intranasally with 2/6-VLPs of the bovine RF strain of rotavirus with and without CT or LT.

We examined the serum and fecal antibody responses in mice inoculated twice intranasally either with 50 μg of 2/6-VLPs alone or with 10 μg of 2/6-VLPs in the presence of 5 μg of CT or LT as well as in control mice that received PBS with or without CT or LT. No significant antibody response was observed in control mice. All mice immunized with VLPs alone developed significant IgA and IgG serologic responses (Fig. 1A). We further characterized the systemic IgG response by assessing IgG1 and IgG2a subclass responses. Elevated IgG1 and IgG2a levels were observed (Fig. 1A). In fecal extracts, we could not detect any IgA response, whereas low levels of IgG were observed in all mice (Fig. 1B). The coadministration of 10 μg of 2/6-VLPs with 5 μg of CT or LT induced significantly higher titers of serum and fecal IgA and IgG antibody than administration of 50 μg of 2/6-VLPs alone (P < 0.0002) (Fig. 1). Notably, a significant fecal IgA response was observed (Fig. 1B). No statistically significant differences in fecal and serum antibody responses were observed between mice that received CT and LT as adjuvants (P > 0.05). 2/6-VLPs administered with CT or LT induced higher titers of serum IgG than serum IgA, whereas similar levels of IgG and IgA were observed in fecal extracts. We further characterized the effects of CT and LT on the IgG subclass response. Intranasal inoculation of 2/6-VLPs with CT or LT induced significantly higher titers of both IgG1 and IgG2a than intranasal administration of 2/6-VLPs alone (P < 0.0002) (Fig. 1A). No statistically significant differences in IgG1 response or in IgG2a response were observed between mice that received CT and LT as adjuvants (P > 0.05).

FIG. 1.

Serum (A) and fecal (B) rotavirus-specific antibody response in mice inoculated twice intranasally with 50 μg of 2/6-VLPs alone or 10 μg of 2/6-VLPs with CT (5 μg) or LT (5 μg) as an adjuvant. Serum antirotavirus IgA, IgG, and IgG subclass (IgG1 and IgG2a) titers were determined 21 days after the second immunization. The results are plotted as the geometric mean titers for the 3 groups, and error bars represent 1 standard error of the mean. Rotavirus-specific antibody responses were not detected in unimmunized mice. ∗, significant difference (P < 0.05) compared to the group given 50 μg of 2/6-VLPs alone for the same isotype.

Cytokine production by spleen cells from mice intranasally inoculated with 2/6-VLPs of the bovine RF strain of rotavirus with and without CT or LT.

We next examined the rotavirus-specific systemic cytokine response induced by intranasal inoculation of 2/6-VLPs alone or with CT or LT 3 weeks after the second immunization. IL-2, IFN-γ, IL-4, IL-5, and IL-10 were assayed in culture supernatants from spleen cells at different times after in vitro restimulation with purified virus (RF or SA11 strain), on the basis of preliminary experiments (day 2 for IL-2, day 4 for IL-4 and IL-10, and day 5 for IFN-γ and IL-5). Residual production from in vivo priming was evaluated in wells incubated without antigen in the presence of RPMI medium only. Spleen cells from control mice that received PBS only or PBS with CT or LT did not produce any cytokines after in vitro restimulation, and the results for these groups were combined (Fig. 2).

FIG. 2.

Th1 (IFN-γ and IL-2) and Th2 (IL-4, IL-5, and IL-10) cytokine production by spleen cells from groups of five mice inoculated twice intranasally with 50 μg of 2/6-VLPs alone or 10 μg of 2/6-VLPs with CT or LT and from unimmunized mice (n = 15). Spleen cells were removed 3 weeks after the second immunization and restimulated in vitro with purified virus RF (▪) or SA11 (▨) or cultured in the presence of RPMI medium only (□). The capacity of spleen cells to be stimulated in vitro and to produce cytokines was confirmed by production of cytokines in the presence of concanavalin A. Results are means for individual mice plus standard deviations. ∗, significant difference (P < 0.05) compared to the nonimmunized group after in vitro restimulation under the same conditions; ∗∗, when, in addition, significant difference (P <0.05) compared to restimulation with RPMI medium only was observed within the same group; @, significant difference (P < 0.05) in cytokine production compared to the group given 50 μg of VLPs alone after in vitro restimulation under the same conditions; #, significant difference (P < 0.05) in cytokine production between the groups immunized with CT and LT after in vitro restimulation under the same conditions.

After in vitro restimulation with purified virus RF, spleen cells from mice intranasally inoculated with 50 μg of 2/6-VLPs alone produced a mixed Th1/Th2-like cytokine profile characterized by IFN-γ and IL-10 production and very low levels of IL-2, IL-4, and IL-5, which were higher than those observed for controls, although not significantly different (Fig. 2). When 10 μg of 2/6-VLPs was administered intranasally with CT or LT, a mixed Th1/Th2-like cytokine profile was also induced and was characterized by IFN-γ and IL-10 production as well as significant levels of IL-2, IL-4, and IL-5 (Fig. 2). However, some differences were observed between the two adjuvants. Mice immunized with 10 μg of 2/6-VLPs and CT produced higher levels of IL-2, IL-4, and IL-5 than mice immunized with 50 μg of 2/6-VLPs alone (P = 0.0026, 0.043, and <0.0001, respectively), whereas IL-10 levels were higher but not statistically different and IFN-γ levels were lower (P = 0.205 and 0.0218, respectively). Mice immunized with 10 μg of 2/6-VLPs and LT produced higher levels of IL-2, IL-4, IL-5, and IL-10 than mice immunized with 50 μg of 2/6-VLPs alone but equivalent levels of IFN-γ (P = <0.0001, 0.0002, 0.0016, 0.0015, and 0.172, respectively). Compared to mice immunized with 10 μg of 2/6-VLPs and CT, mice immunized with 10 μg of 2/6-VLPs and LT produced higher levels of IFN-γ, IL-2, IL-4, and IL-10 but lower levels of IL-5 (P = 0.0021, <0.0001, 0.0069, 0.0126, and 0.0075, respectively).

For some mice in the three groups, a low residual production of some cytokines resulting from in vivo priming was observed in wells incubated without antigen in the presence of RPMI medium only.

Finally, levels of Th1 cytokines produced by spleen cells restimulated with purified virus of the simian SA11 strain (serotype G3) were not statistically different from those produced when purified virus of the bovine RF strain (serotype G6) was used for restimulation (P > 0.05) (Fig. 2). Although levels of Th2 cytokines produced by spleen cells restimulated with RF were higher than those obtained after SA11 restimulation, we observed a statistically significant difference only for IL-4 levels in mice immunized with 2/6-VLPs and CT (P = 0.04).

IL-2 and IL-5 production by spleen cells and superficial and posterior CLN, PP, and MLN cells from mice immunized intranasally or orally with 2/6-VLPs and LT.

Since rotavirus is an enteric pathogen and since mucosal immune responses may be stronger at sites adjacent to the site of induction than at more distant sites, we investigated IL-2 and IL-5 production, as being representative of Th1 and Th2 cells, respectively, by PP and MLN cells from mice intranasally immunized twice with 10 μg of 2/6-VLPs and LT. This immunization schedule was selected because it induced a strong production of both cytokines in the first set of experiments and of enhanced protection when VLPs are administered with LT (16). We also investigated IL-2 and IL-5 production from splenic and CLN (superficial and posterior lymph nodes) cells, where antigen-sensitized lymphocytes migrate after intranasal immunization (21, 25). Controls were inoculated with PBS and LT. Finally, to determine if a portion of the intranasal inoculum which might have been swallowed contributed to the immune response at the intestinal level, we also investigated IL-2 and IL-5 production by the different lymphoid tissues after oral immunization with 10 μg of 2/6-VLPs and 5 μg of LT. Cells from spleens, CLNs (superficial and posterior), PPs, and MLNs from control mice that received PBS with LT did not produce significant levels of IL-2 or IL-5 (Fig. 3).

FIG. 3.

IL-2 and IL-5 production by spleen cells and superficial and posterior CLN, MLN, and PP cells from mice inoculated twice intranasally (n = 5; for MLN cells, n = 3) or orally (n = 5) with 10 μg of 2/6-VLPs and LT (5 μg) and from unimmunized mice (n = 4). Cells were removed 3 weeks after the second immunization and restimulated in vitro with purified virus RF (▪) or VLP (▨) or cultured in the presence of RPMI medium only (□). The capacity of spleen cells to be stimulated in vitro and to produce cytokines was confirmed by production of both cytokines in the presence of concanavalin A. Results are means for individual mice plus standard deviations. ∗, significant difference (P < 0.05) compared to the nonimmunized group after in vitro restimulation under the same conditions; ∗∗, within the group inoculated by the intranasal route, significant difference compared to restimulation with RPMI medium only; #, significant difference (P < 0.05) between the groups immunized intranasally and orally after in vitro restimulation under the same conditions.

Cells from spleens and superficial and posterior CLNs, PPs, and MLNs from mice inoculated by the intranasal route produced high levels of IL-2 and IL-5 after in vitro restimulation with purified RF virus. A residual production of IL-2 and/or IL-5 resulting from in vivo priming was observed in wells cultured with RPMI medium only. Compared to restimulation with antigen, lower levels were observed. VLPs were used for in vitro restimulation of spleen and superficial CLN cells, and results were compared to those observed with the whole virus at the same concentration. IL-2 was produced at a similar level by cells from both spleens and CLNs (P > 0.05), whereas IL-5 production was higher in the presence of VLPs; however, the difference was significant only for spleen cells (P = 0.009).

Finally, IL-2 and IL-5 production was observed for spleens and superficial CLNs and MLNs from only two of the five mice given 2/6-VLPs by the oral route and for PP cells from only one. Moreover, it differed significantly from IL-2 and IL-5 production observed for all the lymphoid tissues from mice immunized by the intranasal route.

DISCUSSION

Recent studies have shown that the intranasal route of immunization with rotavirus VLPs or inactivated rotavirus particles induced higher levels of antibody responses and protection in mice than oral immunization (15–17). Moreover, CT and LT have been shown to significantly increase antibody responses and protection against rotavirus infection in adult CD-1 mice intranasally immunized with VLPs (16, 17). In the present study, we investigated the rotavirus-specific lymphocyte responses induced by intranasal administration of 2/6-VLPs of the bovine RF strain of rotavirus to adult BALB/c mice with and without the mucosal adjuvants CT and LT. 2/6-VLPs intranasally administered alone induced a rotavirus-specific systemic Th1/Th2-like cytokine response very similar to that produced by spleen cells from suckling mice orally inoculated with highly replicative homologous or weakly replicative heterologous strains of rotavirus (5). Indeed, spleen cells from mice immunized with 2/6-VLPs produced, after in vitro restimulation with purified rotavirus, IFN-γ, IL-10, and very low levels of IL-2, IL-4, and IL-5. This mixed Th1/Th2-like cytokine profile was in accordance with the fact that both IgG1 and IgG2a were produced. These results indicate that rotavirus antigens induce similar systemic cytokine responses under very different experimental conditions.

We also examined the lymphocyte responses after coadministration by the intranasal route of 2/6-VLPs with CT or LT. As previously reported for CD-1 mice, we found that rotavirus-specific antibody titers were greatly enhanced; notably, high fecal IgA and IgG titers were induced by administration of 10 μg of 2/6-VLPs with adjuvants, compared to the very weak fecal antibody response induced by 50 μg of 2/6-VLPs alone. Whether these fecal antibody responses are mucosa associated or serum derived remains to be determined. Concerning the cytokine response, we found that CT and LT significantly increased both Th1- and Th2-like cytokine production by spleen cells compared to immunization with 2/6-VLPs alone. In support of this, a strong upregulation of both serum IgG1 and IgG2a isotypes was observed. However, enhancement was not observed for all the cytokines assessed, and some differences were observed between CT and LT. Whereas IL-2, IL-4, and IL-5 production was greatly enhanced when a fivefold-lower dose of 2/6-VLPs was used with both adjuvants, IFN-γ production was either lower or equivalent, according to the adjuvant, and IL-10 was not upregulated with CT. This observation suggests that upregulation of cytokine production by CT and LT may be dissociated, the major effect concerning cytokines which are not produced when 2/6-VLPs were administered alone. Although it seemed that the immune response with CT was more Th2-like than that with LT, due to higher levels of IL-5 (Th2 type), its effects should be considered not purely Th2 but rather mixed. Indeed, we observed a strong increase in both IgG1 and IgG2a levels as well as an increase in Th1 (IL-2) and Th2 (IL-4 and IL-5) cytokine levels. Several studies have reported that oral or intranasal immunization of mice with tetanus toxoid and CT induced a strict Th2 response (12, 13, 27). However, a mixed Th1/Th2 immune response with CT has also been shown, in particular after intranasal immunization of mice with a human immunodeficiency virus type 1 peptide and CT (19) and after oral or intranasal immunization of macaques with p55 of simian immunodeficiency virus and CT (6, 14). These conflicting results remain to be explained; in particular, differences in antigens should be considered. With LT as adjuvant, both IgG1 and IgG2a titers were enhanced as well as Th1 and Th2 cytokine levels. Our results support other studies which showed that oral immunization with tetanus toxoid and LT or intranasal immunization with Helicobacter urease and LT induced a mixed Th1 and Th2 response (20, 24). Moreover, McNeal et al. (15) found that inclusion of LT-R192G, an attenuated E. coli heat-labile enterotoxin, during intranasal immunization with inactivated rotavirus particles caused increases in both serum IgG1 and IgG2a titers without significant change in IgG1/IgG2a ratios.

Since rotavirus is an enteric pathogen and because it has become evident that mucosal immune responses are stronger at sites adjacent to the site of induction than at more distant sites, we next investigated cytokine production at the intestinal level. IL-2 and IL-5 production, representative of Th1 and Th2 cells, respectively, was assessed in PPs and MLNs from mice immunized twice intranasally with 10 μg of 2/6-VLPs and LT. This immunization schedule was selected because of the high immune response induced in the first set of experiments and of enhanced protection when VLPs are administered with LT (16). Both cytokines were also investigated in spleens and CLNs (superficial and posterior), where antigen-sensitized lymphocytes migrate after intranasal immunization (21, 25). We found that, after intranasal immunization, rotavirus-specific cells producing high levels of IL-2 and IL-5 could be detected not only in spleens and CLNs but also in PPs and MLNs. Since after in vitro restimulation, cytokines are most likely produced by T cells, these results suggest that rotavirus-specific T lymphocytes are present in the intestine after intranasal immunization. This is in accordance with recent results reporting the presence of memory T cells in intestinal lymphoid tissue after intranasal inoculation of mice with the SAG1 protein of Toxoplasma gondii with CT (23) and Cryptosporidium parvum DNA (18) and of memory B cells in pigs inoculated with rotavirus 2/6-VLPs in the presence of LT (28). In our study, residual production of cytokines by VLP-induced effector cells still present 3 weeks after the second immunization was observed in some cases in wells incubated without antigen and should be deducted from the total response to assess the memory responses.

Although mice were anesthetized during immunization, we could not rule out the possibility that a portion of the intranasal inoculum was swallowed and contributed to the immune response induced by 2/6-VLPs at the intestinal level. To discard this hypothesis, IL-2 and IL-5 were also investigated after oral inoculation of mice under the same conditions. IL-2 and IL-5 production was significantly lower than in mice inoculated by the intranasal route, with detectable production being observed in only two of the five mice orally given 2/6-VLPs and LT. This result eliminates the possibility that a part of the immune response observed at the intestinal level was due to antigen swallowed during intranasal inoculation. Moreover, it is in accordance with previous results showing increased immunogenicity and protection after intranasal administration of VLPs compared to oral immunization (16, 17). Of interest, IL-2 and IL-5 production was detected in superficial CLNs from the two mice for which a response was observed after oral immunization, as was previously reported after immunization with bacterial protein antigens in the presence of CT (25).

In conclusion, we have shown that 2/6-VLPs of the bovine RF strain of rotavirus administered intranasally to adult BALB/c mice generated a rotavirus-specific cytokine response characterized by a mixed Th1/Th2 profile similar to that induced by oral inoculation of live rotavirus in suckling mice. Coadministration of VLPs with CT or LT considerably enhanced this mixed Th1/Th2 response, and this effect correlated with a large increase in fecal and serum antibody responses. Finally, rotavirus-specific production of both IL-2 and IL-5 was observed in PPs and MLNs after intranasal immunization with 2/6-VLPs in the presence of LT, suggesting the presence of rotavirus-specific T lymphocytes in these lymphoid tissues. Further testing will be required to determine if these results might be influenced by the strain of mouse.