Timing of Toll-Like Receptor 9 Agonist Administration in Pneumococcal Vaccination Impacts Both Humoral and Cellular Immune Responses as Well as Nasopharyngeal Colonization in Mice

Abstract

Synthetic oligodeoxynucleotides (ODN) containing unmethylated CpG motifs, CpG ODN, are Toll-like receptor 9 agonists (TLR9a), which have been used as adjuvants in pneumococcal vaccines to improve antibody responses in immunodeficient patients. Here, we examined whether the coadministration of TLR9a with pneumococcal CRM₁₉₇-conjugate vaccine enhances protection against pneumococcal colonization, the levels of antipolysaccharide antibodies, and the CD4⁺ T-cell responses. Wild-type BALB/c mice and B-cell-deficient BALB/c Igh-J^tm1Dhu mice were immunized twice with the following: (i) PCV alone; (ii) simultaneous PCV and TLR9a; (iii) PCV and then TLR9a, after a 48-h delay; (iv) TLR9a alone; and (v) phosphate-buffered saline. Nasopharyngeal protection, serum antibodies, CD4⁺ T-cell responses, and clearance of bacteremia after intraperitoneal challenge with Streptococcus pneumoniae 6B were evaluated. We found decreased nasopharyngeal protection against S. pneumoniae 6B colonization after simultaneous immunization with PCV and TLR9a compared to immunization with PCV alone in wild-type BALB/c mice (P = 0.037). A similar trend was observed in B-cell-deficient BALB/c Igh-J^tm1Dhu mice. Simultaneous administration did not enhance antibody levels and lowered the CRM₁₉₇-specific cytokine release of gamma interferon, interleukin-2 (IL-2), IL-5 and IL-13. Immunization with PCV and then TLR9a, after a 48-h delay, significantly improved nasopharyngeal protection compared to simultaneous administration (P = 0.011). Furthermore, delaying TLR9a delivery increased antibody titers compared to both simultaneous administration (P = 0.001) and PCV immunization alone (P = 0.026). In conclusion, the immunological and clinical impact of adjuvanting a pneumococcal conjugate vaccine (Prevnar; Pfizer) with a TLR9a is highly depended on timing of the adjuvant administration. Thus, careful timing of adjuvant administration may improve novel vaccine formulations.

INTRODUCTION

The purpose of vaccination is to induce long-lasting protection against subsequent infections and disease. The generation of sufficient humoral and cellular adaptive immunity can be difficult to achieve by traditional vaccine design and delivery. Inducing a potent innate immune response as part of an enhanced vaccine regimen has been suggested (18). Innate immune pathways involving specific receptor recognition, such as the Toll-like receptors (TLR), are important for antigen-specific adaptive immunity (8). Short oligodeoxynucleotides (ODN) containing unmethylated CpG motifs are TLR9a agonists and are thought to directly activate B cells and plasmacytoid dendritic cells in humans (2, 13). Vaccine adjuvants such as TLR9a can enhance antibody responses, activate CD4⁺ T cells, and induce inflammatory cytokine production, which among other effects, stimulates the maturation of NK cells and macrophages (9). For these reasons, CpG ODN have been suggested as a potent adjuvant for use in vaccination (11).

Immunocompromised individuals are highly susceptible to infections by Streptococcus pneumoniae, which causes pneumonia, sepsis, and meningitis. The World Health Organization estimates that, among HIV-infected patients, pneumococci are the leading nonmycobacterial etiology of pneumonia worldwide and are responsible for ca. 30% of community-acquired pneumonia cases admitted to hospitals among adults in industrialized countries.

At present, two different types of pneumococcal vaccines are licensed: a thymus-dependent polysaccharide conjugate vaccine (PCV), e.g., the 10-valent Synflorix vaccine (GSK) and the 13-valent Prevnar (Pfizer), and a thymus-independent polysaccharide vaccine (PPSV), e.g., the 23-valent PneumoNovum, Sanofi, and PneumoVax vaccines (Merck). Both types induce serotype-specific anticapsular antibodies. The pneumococcal conjugate vaccine is effective against infections in infants (6, 7), but it has limited serotype coverage, and its protective capacities may diminish over time as a consequence of serotype replacement. On the other hand, the PPSV vaccine covers ca. 90% of the most frequently isolated serotypes, but it is not effective in children younger than 2 years of age. Vaccination of HIV-infected patients with PPSV induce specific antibodies, but the antipolysaccharide (α-ps) antibody titers at 1 year postvaccination may not be different from prevaccination levels, suggesting a short-lived effect from immunization (17).

It has been shown that the administration of a pneumococcal conjugate vaccine together with large doses of CpG ODN increases IgG α-ps levels in mice compared to immunization with a conjugate vaccine alone (4, 14). Furthermore, protein-conjugated polysaccharides restore diminished IgG α-ps responses in aged mice to young adult levels (19). In a clinical trial, we have previously shown that adding a TLR9a (CPG 7909) to a pneumococcal conjugate vaccine doubled the proportion of HIV-infected individuals achieving high vaccine-specific IgG antibody levels at 9 months (20). Recently, it was suggested that the timing of TLR9 adjuvant administration is crucial for the adjuvant effect in the context of pneumococcal polysaccharide immunity (23). Taillardet et al. suggest that an optimal adjuvant effect in mice is obtained when TLR9a administration is delayed 48 h [referred to hereafter as TLR9a(48 h)] relative to polysaccharide vaccination.

Hence, we sought here to analyze the immunological effects of TLR9a administrated in physiologic relevant doses with PCV. Further, we explored the importance of the timing of the adjuvant administration relative to the PCV vaccination.

MATERIALS AND METHODS

Bacterial methods.

S. pneumoniae serotype 6B, strain 0603, was cultured on 5% blood agar plates (SSI, Copenhagen, Denmark) and grown in Todd-Hewitt broth supplemented with 1% fetal calf serum. Frozen stocks were stored at −80 °C in concentrations of 1.2 × 10⁸ CFU/ml. The type 6B strain was chosen because it usually causes bacteremia without killing the animals (unpublished observations).

Immunization.

Female BALB/c AnNTac mice (Taconic, Ry, Denmark), aged 4 to 6 weeks, were used for the nasopharyngeal colonization study and analysis of cellular immunogenicity. Mice aged 8 to 10 weeks were used for the intraperitoneal challenge study. All mice were immunized subcutaneously with either (i) 0.4 mg of the pneumococcal conjugate vaccine (Prevnar, Pfizer, CRM₁₉₇-conjugated)/kg alone, (ii) 0.4 mg of PCV and 4 μg of TLR9a (ODN 1826; Invivogen, Denmark)/kg simultaneously, (iii) 0.4 mg of PCV/kg and then 4 μg of TLR9a/kg after a 48-h delay [TLR9a(49 h)], (iv) TLR9a alone, or (v) phosphate-buffered saline (PBS) control. After 3 weeks, all animal groups were reimmunized. Using the same procedure, both BALB/c Igh-J^tm1Dhu mice (Taconic) deficient in B cells and BALB/c NCr-Foxn1^nu mice deficient in T cells were immunized. During the course of the study, the Prevnar formulation was changed from a 7-valent to a 13-valent vaccine; however, the dose of individual antigens used for immunization of each mouse remained the same over the course of the study. A recombinant protein derived from diphtheria toxin (CRM₁₉₇) is used as a conjugation backbone in the Prevnar vaccine.

Nasopharyngeal colonization.

At 6 weeks after the first immunization, mice were inoculated intranasally with 10⁶ CFU of S. pneumoniae 6B in 25 μl of PBS. At 1 week after inoculation, the mice were euthanized, and the upper respiratory tracts were washed with 100 μl of saline through the transected trachea (16). The first six drops from the nostrils were collected, serial diluted, and cultured on blood agar plates containing 5 mg of gentamicin/ml. Pneumococcal colonies in the tracheal wash were counted, and the CFU/ml were calculated. The identity of a number of colonies was verified by the capsular reaction test by using diagnostic type 6B antisera (SSI). This was done on 10 randomly selected colonies from two different mice in each group.

Enzyme-linked immunosorbent assay antibody measurements.

Orbital blood was retrieved 5 weeks after the first immunization and sera were assayed for α-ps 14 and α-ps 6B IgG. To determine α-ps, sera were diluted with PBS containing 0.1% Tween 20 (PBS-T), phosphorylcholine (50 μl/ml), and cell wall polysaccharide (CWPS; SSI) (100 μg/ml) in order to block antibodies specific for contaminating CWPS (24). Samples were incubated 2 h at 37°C and overnight at 5°C. Wells were coated with either polysaccharide 14, polysaccharide 6B (both from the American Type Culture Collection [ATCC]) or CWPS overnight. The following day, the plates were washed once with PBS, blocked for 1.5 h using assay buffer (Invitrogen, Denmark), and washed three times with PBS-T. Serum samples were serially diluted in PBS-T and then incubated on plates for 1.5 h at room temperature. Subsequently, peroxidase-conjugated rabbit anti-mouse IgG (Dako, Denmark) was added at a 1:2,500 dilution, and the plates were incubated 1.5 h at room temperature. Finally, the plates were washed four times, and color was developed using TMB substrate (Kem-En-Tec, Denmark). The reaction was stopped using 0.5 M H₂SO₄, and the absorption was read at 450 nm with background subtraction at 650 nm. Each serum titration was assayed in duplicate, and an internal monoclonal 6B antibody positive reference (HASP4 [21]) was used to calibrate the assays.

Intraperitoneal challenge.

At 6 weeks after the first vaccination, the mice were intraperitoneally challenged with 10⁷ S. pneumoniae 6B. After 48 h, the mice were euthanized, and 200-μl portions of undiluted blood were cultured on blood agar plates. The next day, the mice were either scored as bacteremic or to have cleared the infection by the presence or absence of pneumococcal colonies. Colonies were verified as S. pneumoniae 6B by capsular reaction tests.

Splenocyte proliferation.

Harvested splenocytes were washed twice with PBS and subsequently stained with CFSE (carboxyfluorescein diacetate succinimidyl ester; Invitrogen). Briefly, cells were adjusted to 50 × 10⁶ cells per ml in PBS and mixed 1:1 with a 10 μM CFSE solution in PBS. After 5 min of incubation at room temperature, the staining was stopped by adding 1 volume of heat-inactivated fetal calf serum. The cells were washed twice with medium before plating the cells at 100,000 cells per round-bottom well in a 96-well plate. After 2 h, the cells were stimulated with CRM₁₉₇ (Pfizer) at 20 μg/ml, polysaccharide 6B (ATCC) at 10 μg/ml, or the combination of anti-CD28 (0.25 μg/ml) and anti-CD3 (1 μg/ml) (both from eBioscience), or the cells were left unstimulated (media) and incubated at 37°C for 4 days. At day 4, the cells were harvested to flow tubes and kept on ice. The cells were washed with PBS and stained with a live/dead fixable dead cell stain kit (Invitrogen) as recommended by the manufacturer. Subsequently, Fc receptors were blocked using anti-CD16/anti-CD33 (BD Bioscience, United Kingdom) for 10 min on ice before being stained with anti-CD4/APC, anti-CD3/PECy7, anti-CD138/PE, and CD19/PerCp-Cy5.5 (BD Bioscience). The cells were washed and fixed in 2% formaldehyde and analyzed on a FACSCanto flow cytometer (BD Bioscience). Single stain controls were used for compensations and results were analyzed by using FlowJo (TreeStar).

Lymphocyte activation.

For an assessment of splenocyte activation, the cells were plated at 200,000 cells per well in 96-well plates and stimulated with CRM₁₉₇ (20 μg/ml), concanavalin A (5 μg/ml), or medium alone as a control. After 8 h, the cells were harvested, placed on ice, and stained with the live/dead fixable dead cell stain kit. Subsequently, Fc receptors were blocked using anti-CD16/antiCD33 (BD Bioscience), and the cells were stained for the surface markers CD3/FITC, CD4/APC-Cy7, CD19/PerCpCy5.5, and CD69/PECy7 (all from BD Bioscience). After a washing step, the cells were fixed with stabilizing fixative (BD Bioscience) and kept at 4°C until analysis on a FACSCanto the following day. Single stain controls were used for compensation, and results were obtained using FlowJo.

Antigen-presenting cells.

To assess antigen-presenting capabilities, splenocytes were plated at 200,000 cells per well in 96-well plates and stimulated with CRM₁₉₇ (20 μg/ml), pokeweed mitogen (5 μg/ml), or medium alone as a control. After 48 h, the cells were harvested, placed on ice, and stained using the live/dead fixable dead cell stain kit. Subsequently, Fc receptors were blocked using anti-CD16/antiCD33 (BD Bioscience), and the cells were stained for the surface markers CD19/PerCpCy5.5, CD11c/PECy7, CD80/PE, and CD86/APC. After a washing step, the cells were fixed with stabilizing fixative (BD Bioscience) and kept at 4°C until analysis on a FACSCanto the following day. Single stain controls were used for compensation, and results were obtained using FlowJo.

Intracellular cytokine production.

Splenocytes were plated at 150,000 cells per 96-well plate and left to rest for 10 h, after which they were stimulated with phorbol myristate acetate (PMA) and ionomycin (50 and 200 ng/ml, respectively), CRM₁₉₇ (20 μg/ml), or medium alone as a control. After half an hour, GolgiPlug (BD Bioscience) containing brefeldin A was added as recommended by the manufacturer. After 6 h of incubation at 37°C, the cells were harvested and stained using the live/dead stain (far red), followed by Fc block with anti-CD16/anti-CD33 (10 min on ice). Subsequently, the cells were surface stained for CD4/APC-Cy7 and CD8/PerCpCy5.5 (all from BD Bioscience) for 30 min on ice. The cells were washed and subsequently fixed using BD Cytofix/Perm for 20 min on ice. After being washed with BD Bioscience Perm/Wash, the cells were stained for intracellular IL-2/APC and IFN-γ/PE-Cy7 (BD Bioscience) for 30 min at room temperature. After being washed, the cells were fixed using stabilizing fixative (BD Bioscience) and stored at 4°C until analysis using FACSCanto. Single-stain controls were included for compensation. Fluorescence-minus-one samples for both cytokines using cells stimulated with PMA and ionomycin (50 and 200 ng/ml, respectively) were used to determine unlabeled levels of gamma interferon (IFN-γ) and interleukin-2 (IL-2). Data analysis was performed using FlowJo.

Splenocyte cytokine secretion.

Splenocytes were plated at 200,000 cells per 96-wells. After six h of rest the cells were stimulated with CRM₁₉₇ (20 μg/ml), CpG DNA (ODN 1826, 10 μg/ml; Invivogen), or medium alone as a control. At day 2 after stimulation, the cell medium was harvested and immediately frozen at −80°C. The levels of interleukin-5 (IL-5), IL-13, and monocyte chemoattractant protein 1 (MCP-1) were assessed using a multiplex bead assays (Invitrogen) on the Luminex 100 platform.

Statistical analysis.

The Mann-Whitney test was used to compare the different vaccination groups with regard to nasopharyngeal colonization and α-ps antibody levels. Cellular immunogenicity assays were evaluated using Kruskal-Wallis with Dunn's multiple-comparison test. Correlations between antibody levels and nasopharyngeal colonization were analyzed by using the Pearson correlation test. The clearance of the intraperitoneal pneumococcal challenge was compared by using the Fisher exact test. A difference was considered significant with a P value of <0.05. Proliferation, activation, antigen presentation costimulation, together with the cytokine levels, were expressed as geometric means in box plots with 5 to 95% percentile whiskers. All graphing and statistical analysis were done in GraphPad Prism. All animal experiments were approved by the Danish National Experimental Animal Inspectorate (document 2010/561-1794).

RESULTS

Simultaneous immunization with TLR9a and PCV reduces protection against nasopharyngeal colonization.

Simultaneous coadministration of TLR9a with PCV significantly reduced protection against pneumococcal colonization compared to mice vaccinated with either PCV alone (P = 0.011) or TLR9a alone (P = 0.0006) (Fig. 1A). No differences in colonization between groups receiving PCV alone or TLR9a alone were observed. However, immunization with PCV and TLR9a(48 h) restored protection compared to the group immunized with PCV and TLR9a simultaneously (P = 0.023). A similar nasopharyngeal colonization study in B-cell-deficient BALB/c Igh-J^tm1Dhu mice showed the same tendency between immunization groups as did BALB/c mice (Fig. 1B). This suggests that the simultaneous administration of TLR9a with PCV aggravates protection against nasopharyngeal S. pneumoniae 6B colonization compared to the administration of either component alone.

Fig 1.

Pneumococcal nasopharyngeal colonization. (A) BALB/c mice were vaccinated in a prime boost regimen 3 weeks apart. At 3 weeks after the last vaccination, the mice were inoculated via the nasal cavity with S. pneumoniae serotype 6B (10⁶ CFU). One week later, the mice were sacrificed, and the nasopharynges were washed and plated on blood agar plates. Colonization titers (i.e., the median CFU calculated from three dilutions) were determined from mice from two separate experiments. Among vaccinated mice, the group receiving PCV and TLR9a (n = 26) had the highest level of colonization compared to the groups receiving either PCV alone (n = 24; P = 0.011), adjuvant alone TLR9a (n = 26; P = 0.0006), or PCV with delayed TLR9a (n = 14; P = 0.023). Group comparisons were made using the Mann-Whitney test (*, P < 0.05; **, P < 0.01). (B) BALB/c Igh-J^tm1Dhu mice devoid of B cells were vaccinated in a prime-boost regimen 3 weeks apart with either PCV alone (n = 12) or PCV+TLR9a (n = 12), with subsequent nasal colonization by S. pneumoniae serotype 6B (10⁶ CFU). Group comparison were made by using the Mann-Whitney test. The difference was not significant (P = 0.11).

Administration of TLR9a 48 h after PCV immunization increased α-ps antibody titers significantly.

Mice immunized with PCV and TLR9a(48 h) had a significantly higher α-ps 14 IgG response compared to mice immunized with PCV alone (P = 0.026) or PCV and TLR9a simultaneously (P = 0.001) (Fig. 2A). The antibody α-ps 6B levels yielded similar results (data not shown). There was no difference between antibody titers in the PCV-alone group and the simultaneous PCV and TLR9a group. No antibodies were detected in control groups immunized with either PBS or TLR9a alone. Furthermore, there was no correlation between α-ps 6B IgG antibody levels and nasopharyngeal colonization. This was true for all immunization strategies (Fig. 2B). These observations support that the differences in protection against nasopharyngeal colonization with S. pneumoniae 6B between groups are not dependent on antibodies and, consequently, that antibodies are not the primary mediator of protection against pneumococcal nasopharyngeal colonization in mice. Lastly, vaccination of T-cell-deficient mice (BALB/c NCr-Foxn1^nu) with PCV resulted in no antibody responses to polysaccharide types 6B and 14, highlighting the T-cell dependence of this vaccine (Fig. 2C).

Fig 2.

Antibodies. (A) Mice vaccinated with PCV+TLR9a(48 h) (n = 27) had significantly higher antibody titers to polysaccharide 14 in sera than mice receiving PCV alone (n = 27; P = 0.026) and mice vaccinated with PCV and TLR9a simultaneously (n = 27; P = 0.001). Similar observations were made for serotype 6B. Group comparisons were determined by using the Mann-Whitney test. (B) Plot of colonization in the nasal cavity and antibody titers. A Pearson correlation test was used to analyze the correlation between nasal colonization with S. pneumoniae serotype 6B and the antibody titers. No correlations were established (P = 0.25). The graph depicts data for pooled mice from the PCV-alone, PCV+TLR9a, and PCV+TLR9a(48 h) vaccination groups. No correlations were observed when analysis were performed stratified by immunization group (data not shown). (C) Athymic BALB/c NCr-Foxn1^nu nude mice vaccinated as described previously (n = 5 animals per group). No antibodies to polysaccharide 14 were detected in any vaccination group.

Lymphocyte proliferation potential is severely impaired by simultaneous PCV and TLR9a immunization.

The proliferation of CD4⁺, CD19⁺, and CD138⁺ cells in whole splenocyte preparations was evaluated. Figure 3 displays the collected data as box plots and representative flow plots of proliferating cells and CD138 positivity identifying differences between the PCV-alone and PCV+TLR9a immunizations. Mice receiving PCV and TLR9a simultaneously had a significantly reduced proliferative potential compared to those receiving PCV alone in all of the subsets analyzed (Fig. 3A, B, and D). Strikingly, the fraction of plasma cells (i.e., CD138⁺ lymphocytes) was far greater in both the PCV+TLR9a and the TLR9a-alone immunization groups than in the other two groups (P < 0.05 and P < 0.01). In turn, this increased plasma cell differentiation lead to a decrease in proliferative potential of the CD138⁺ population, as clearly seen in Fig. 3D and E. Hence, immunization with PCV alone relative to PCV+TLR9a ensured potent T- and B-cell proliferation. The reduction in proliferation was less profound when TLR9a was delayed relative to PCV vaccination. However, the PCV+TLR9a(48 h) results were not significantly different from the PCV+TLR9a results for any subset, but all displayed the same tendency.

Fig 3.

Proliferation potential of lymphocytes after pneumococcal vaccination. (A) Proliferation of live CD4⁺ lymphocytes among CFSE-labeled splenocytes from immunized mice (n = 5 in each group) after 4 days of culture in the presence of CRM₁₉₇. The proliferation of cells obtained from mice immunized with PCV alone was significantly higher than in mice immunized with PCV+TLR9a (P < 0.05). The differences between the PCV+TLR9a and PCV+TLR9a(48 h) groups were not significant. (B) Proliferation of live CD19⁺ lymphocytes among CFSE-labeled splenocytes from vaccinated mice (n = 5 in each group) after 4 days of culture in the presence of CRM₁₉₇. The proliferation of CD19⁺ lymphocytes in animals vaccinated with PCV alone was significantly higher than in PCV+TLR9a-treated mice (P < 0.05). The differences between the PCV+TLR9a- and PCV+TLR9a(48 h)-treated mice were not significant. (C) Proportions of live-gated lymphocytes that are CD138⁺. The difference between the PCV-alone and PCV+TLR9a animals, as well as between the PCV+TLR9a and PCV+TLR9a(48 h) animals, were significant (P < 0.05 and P < 0.01, respectively). (D) Proliferation of live CD138⁺ lymphocytes among CFSE-labeled splenocytes from vaccinated mice (n = 5 in each group). The CD138⁺ cell number is limited, which does not support proliferation analysis; hence, gate selection allows an estimation of cells with division potential. The differences between the PCV-alone and PCV+TLR9a groups were significant. (E) The graph depicts the relationship in individual mice vaccinated with PCV, PCV+TLR9a, and PCV+TLR9a(48 h) between the proportions of CD138⁺ cells and the proliferation potential of the CD138⁺ subset.

Simultaneous PCV and TLR9a immunization reduces B- and T-cell activation.

Next, lymphocyte activation was evaluated. Purified splenocytes were stimulated with CRM₁₉₇ for 8 h. No differences in activated CD4⁺ cells were observed between the TLR9a group and the PCV group, but simultaneous immunization with PCV and TLR9a reduced the activation (upregulation of CD69) of CD4⁺ T cells compared to the PCV-alone group (P < 0.05) (Fig. 4A). Identical results, were obtained when analyzing the B-cell compartment. Stimulation with CRM₁₉₇ significantly upregulated CD69 on CD19⁺ lymphocytes (P < 0.001) (Fig. 4B). These results show that the simultaneous administration of TLR9a and PCV significantly reduced the activation potential of both CD4⁺ T and CD19⁺ B cells.

Fig 4.

Cellular activation. (A) After 8 h stimulation with CRM₁₉₇, live/CD4⁺ gated cells were scored for CD69 positivity, as indicated in the flow diagram. Box plots with 5 to 95% percentile whiskers from five animals are shown. All group comparisons were performed using the Kruskal-Wallis with Dunn's multiple-comparison test. Mice vaccinated with PCV alone showed significantly higher mobilization of the activation marker compared to PCV+TLR9a-treated animals (P < 0.05). (B) MFI of CD69 on live/CD19⁺ lymphocytes, as indicated in the flow diagram. The histogram compares PCV (red) and PCV+TLR9a (blue) results. Box plots with 5 to 95% percentile whiskers from five animals are shown. All group comparisons were performed using the Kruskal-Wallis with Dunn's multiple-comparison test. Mice receiving PCV had significantly higher levels of CD69 compared to all other animals (P < 0.001).

Antigen presentation.

Antigen presentation capacity was evaluated by measuring the surface markers CD86 and CD80 on antigen-presenting cells. The top panel of Fig. 5 displays the data obtained with B cells (CD19⁺) and clearly shows vast differences in the ability to regulate CD86 upon CRM₁₉₇ stimulation. The mean fluorescence intensities (MFI) of PCV- and PCV+TLR9a(48 h)-vaccinated animals were significantly higher than those of PCV+TLR9a-vaccinated animals (P < 0.01 and P < 0.05, respectively; Fig. 5A). In addition, the impact of antigen presentation measured by regulating CD86 and CD80 on dendritic cells again defines the simultaneous delivery of adjuvant as less effective [P < 0.05 compared to both the PCV-alone and PCV+TLR9a(48 h) animals] (Fig. 5B). Hence, the ability for B cells and dendritic cells to present antigen via major histocompatibility complex molecules and deliver a potent costimulus appears to be hampered by vaccination of a thymus-dependent polysaccharide vaccine simultaneously with TLR9a.

Fig 5.

Antigen-presenting cells. (A) Representative diagram highlighting the difference in CD86 upregulation on live/CD19⁺ gated lymphocytes upon 48 h of CRM₁₉₇ stimulation. Data for the sample from the PCV group is red, and that for the PCV+TLR9a group is blue. An identical color code is used in the histogram. The MFI of CD86 is displayed in box plots with 5 to 95% percentile whiskers from five animals. Group comparison was performed using the Kruskal-Wallis with Dunn's multiple-comparison test. Mice receiving PCV and PCV+TLR9a(48 h) displayed significantly higher levels of CD86 compared to mice receiving PCV+TLR9a (P < 0.01and P < 0.05, respectively). (B) Representative flow diagrams highlighting the small but significant differences in CD80 and CD86 upregulation on live/CD11c⁺ gated dendritic cells upon 48 h of CRM₁₉₇ stimulation. Box plots with 5 to 95% percentile whiskers from five animals display the percentages of live/CD11c⁺ cells positive for either CD80 or CD86. Mice receiving PCV and PCV+TLR9a(48 h) expressed CD80/CD86 on CD11⁺ cells compared to mice receiving PCV+TLR9a (P < 0.05 for both).

Cytokine production.

The simultaneous administration of PCV and TLR9a significantly reduced cytokine production. Splenocytes from vaccinated mice were stimulated with CRM₁₉₇. PCV vaccination induced significantly higher proportions of IFN-γ-producing CD4⁺ T cells compared to all other groups (P < 0.01) (Fig. 6A), whereas immunization with TLR9a alone was associated with significantly higher proportions of IL-2-producing CD4⁺ T cells compared to PCV+TLR9a-treated mice (P < 0.05) but was not significantly different from animals treated with PCV alone or PCV+TLR9a(48 h) (Fig. 6B). Analysis of the supernatants from stimulated splenocytes identified highly statistical significant differences in responses from PCV-vaccinated versus PCV+TLR9a-vaccinated mice with regard to IL-5 (P < 0.001), IL-13 (P < 0.01), and MCP-1 (P < 0.01) (Fig. 6C). Thus, mice immunized simultaneously with PCV and TLR9a showed no detectable production of IFN-γ, IL-2, IL-5, and IL-13 and limited amounts of MCP-1 production, although immunization with PCV and TLR9a(48 h) restored the ability to produce cytokines (significant responses for IL-5 [P < 0.01] and IL-13 [P < 0.05]) (Fig. 6). These results suggest that antigen-specific cytokine induction after immunization with either PCV or TLR9a is reduced when the two components are delivered simultaneously.

Fig 6.

Cytokine production. (A) IFN-γ⁺ cells after gate selection on CD3⁺ CD4⁺ T cells after 8 h stimulation with CRM₁₉₇. Box plots with 5 to 95% percentile whiskers from five animals are shown. All group comparisons were performed using the Kruskal-Wallis with Dunn's multiple-comparison test. Mice vaccinated with PCV alone had significantly greater IFN-γ responses compared to PCV (P < 0.01). (B) IL-2⁺ cells after gate selection on CD3⁺ CD4⁺ T cells after 8 h of stimulation with CRM₁₉₇. Box plots with 5 to 95% percentile whiskers from five animals are shown. Group comparison was performed using Kruskal-Wallis with Dunn's multiple-comparison test. Mice receivingTLR9a produced more IL-2 than mice treated with either PCV+TLR9a or PBS (P < 0.05). (C) Splenocytes (2 × 10⁵) from vaccinated animals (five in each group) were stimulated for 48 h with CRM₁₉₇. The supernatant was harvested and analyzed on the Luminex 100 platform. Box plots with 5 to 95% percentile whiskers are shown. All group comparisons were performed using the Kruskal-Wallis with Dunn's multiple-comparison test. Mice vaccinated with PCV alone were significantly different from mice treated with PCV+TLR9a for IL-5 (P < 0.001), IL-13 (P < 0.01), and MCP-1 (P < 0.05). Mice treated with PCV+TLR9(48 h) were significantly different from mice treated with PCV+TLR9a for both IL-5 (P < 0.01) and IL-13 (P < 0.05).

Intraperitoneal pneumococcal challenge.

Immunization of mice with PCV and TLR9a(48 h) enhanced the clearance of a pneumococcal infection. At 6 weeks after the first immunization, mice were challenged intraperitoneally with 10⁷ S. pneumoniae 6B. The majority of mice in the PCV group (75%) and in the group immunized with PCV and TLR9a(48 h) (87%) cleared the bacteria (Fig. 7). In contrast, most control mice, i.e., those treated with PBS, or mice treated with TLR9a alone were unable to clear the infection (37 and 12% clearance, respectively). The group immunized simultaneously with PCV and TLR9a tended to have a lower pneumococcal clearance (50%) compared to the groups immunized with either PCV alone (75%) or PCV and TLR9a(48 h) (87%). Comparing the three antigen experienced groups to the combined antigen naive controls highlight the suppression observed after combining PCV and simultaneous TLR9a delivery. The mice receiving PCV alone and PCV, followed by TLR9a(48 h), were significantly better protected compared to controls (P = 0.0324 and P = 0.0078, respectively [Fisher exact test]). This protection relative to controls was not observed with simultaneous PCV and TLR9a administration (P = 0.365 [Fisher exact test]).

Fig 7.

Bacteremia upon intraperitoneal challenge. Eight-week-old BALB/c mice were vaccinated in a prime boost regimen 3 weeks apart (n = 8 in each group). At 3 weeks after the last vaccination, the mice were challenged via intraperitoneal inoculation of S. pneumoniae serotype 6B (10⁷ CFU). After 48 h, the mice were sacrificed, and blood samples were serially diluted on blood agar plates. Bacteremia was defined by the presence of detectable bacteria in the blood, and each mouse was scored as either clearing the infection or having bacteremia. Depicted are the percentages of mice clearing the infection. Group comparisons were performed using the Fisher exact test. Both PCV-vaccinated and PCV+TLR9(48 h)-vaccinated mice had significantly better bacterial clearance than did antigen-naive mice (P = 0.034 and P = 0.0078, respectively). PCV+TLR9a-treated mice were not different from either antigen-naive control mice or the other PCV-vaccinated groups.

DISCUSSION

TLR9 agonists are potent immunostimulatory molecules with great potential as novel vaccine adjuvants. However, we demonstrated here that the simultaneous administration of PCV and TLR9a in both wild-type BALB/c mice and B-cell-deficient BALB/c mice (BALB/c Igh-J^tm1Dhu mice) decrease protection against pneumococcal nasopharyngeal colonization. Furthermore, the simultaneous administration of PCV and TLR9a blunted postvaccination CRM₁₉₇-specific IFN-γ, IL-5, IL-13, and MCP-1 secretion. In contrast, delaying the administration of TLR9a for 48 h relative to the PCV injection significantly improved protection against nasopharyngeal colonization compared to simultaneous vaccine and adjuvant administration. The timing of adjuvant delivery affected IL-5 and IL-13 production, plasma cell differentiation, and antigen presentation capacity. Importantly, delaying TLR9a administration doubled antibody levels compared to both simultaneous PCV and TLR9a administration and immunization with PCV alone.

All previous vaccination studies injected very high doses of TLR9a (50 to 100 μg/mouse), which greatly exceed the more physiological relevant dose used in the present study (4 μg/mouse). Still, our results are consistent with those of a recent study reporting low α-ps3 IgM responses following the simultaneous administration of polysaccharide 3 and TLR9a but high α-ps3 IgM responses when TLR9a administration was delayed 1, 2, or 3 days (23). In contrast, others have reported increased α-ps antibody levels when coadministering TLR9a and protein-conjugated polysaccharides (4, 10).

Taillardet et al. (23) suggest that the TLR9a effect on the α-ps antibody response is independent of T cells when pure polysaccharides are used. Both the T- and B-cell-deficient mice described in the present study, together with the observed cytokine production and cellular activation analysis, indicate a T-cell-dependent effect of TLR9a when coadministered with a thymus-dependent vaccine. Hence, the observation that both T- and B-cell recall responses are blunted by TLR9a administration warrants further studies on selective antigen-adjuvant combinations. Importantly, we found no correlation between serotype-specific IgG levels and the number of pneumococci in pharyngeal washes. Thus, the reduced protection from colonization that we observed in simultaneously PCV- and TLR9a-immunized mice can thus be ascribed to an immune exhaustive effect independent of antibodies, a view supported by other studies advocating the importance of CD4⁺ T cells in protection against nasopharyngeal pneumococcal colonization (1, 16, 22).

Chu et al. reported that TLR9a administrated together with hen egg lysozyme induces IFN-γ production and decreases IL-5 production (5). IFN-γ and IL-2 are both Th-1 proinflammatory cytokines, whereas IL-5 and IL-13 are Th-2 cytokines that stimulate eosinophils, cytotoxic T cells, and B-cell proliferation and differentiation in mice (15). We observed that the induction of both Th-1 cytokines, IFN-γ and IL-2, was abrogated by the simultaneous administration of PCV and TLR9a. In addition, both IL-5 and IL-13 production were absent in splenocytes from mice vaccinated simultaneously with PCV and TLR9a but almost completely restored by delaying the administration of TLR9a by 48 h. Chu et al. suggested that TLR9a switches a Th-2-dominated response to a Th-1 response (5). Our analysis of cytokine production indicates that the simultaneous administration of PCV and TLR9a blocks both Th-1 and Th-2 responses, which are partly restored when TLR9a administration is delayed by 48 h. We argue that these findings support a model wherein effector cells (both B and T cells) are unable to respond to recall antigen (i.e., they display an exhausted phenotype). Clinically, the restriction of immune responses by simultaneous PCV and TLR9a administration is supported by both a significant decrease in protection against nasopharyngeal colonization and a reduced clearance of bacteria upon intraperitoneal challenge. A clear demonstration of the effect of adjuvant timing on antigen-presenting cells is evident in Fig. 3E. Interestingly, delaying TLR9a administration relative to antigen interfered with the formation of CD138⁺ plasma cells. Moreover, these CD138⁺ plasma cells have lowered the proliferation potential and, as such, appear blunted. However, mice receiving TLR9a 48 h after receiving PCV had significantly higher levels of circulating antibodies than did mice treated with PCV alone. This finding indicates that committing lymphocytes to an antigen prior to receiving a secondary stimulatory signal from the adjuvant could be beneficial in the combination tested here.

TLR9 receptors are known to be expressed differently in humans and mice. Human B cells and plasmacytoid dendritic cells express TLR9, whereas B cells, plasmacytoid dendritic cells, and myeloid dendritic cells, as well as monocyte/macrophages, express TLR9 in mice (3). Peak cytokine secretion after subcutaneous TLR9 immunization in mice and humans has previously been reported to occur within 2 to 12 h postinjection and at 24 h postinjection, respectively (9, 12, 25). However, in our study the biological effect was maintained 3 weeks after the last injection; mice receiving only TLR9a displayed significantly greater background IL-2 production by CD4⁺ T cells. These prolonged immune responses after TLR9a administration make it highly interesting to investigate TLR9a mechanisms and immune potentials further.

In conclusion, we demonstrated here that, compared to the immunization of PCV alone, immunization of mice with PCV and then TLR9a after a 48-h delay significantly improved immune responses to both the polysaccharide and the protein carrier of PCV (CRM₁₉₇). Our findings emphasize that the adjuvant mechanisms of TLR9a are complex and encompass antigen-presenting cell, B-cell, and T-cell components and, as such, have different immunological effects when used in conjunction with a T-cell-dependent or T-cell-independent vaccine. The applicability of delaying adjuvant administration relative to the administration of vaccine antigen in humans needs to be investigated carefully.