CpG-ODN Shapes Alum Adjuvant Activity Signaling via MyD88 and IL-10

Abstract

Aluminum-containing adjuvants usually referred as Alum are considered as T helper type-2 (Th2) adjuvants, while agonists of toll-like receptors (TLRs) are viewed as adjuvants that favor Th1/Th17 immunity. Alum has been used in numerous vaccine formulations; however, its undesired pro-Th2 adjuvant activity constitutes a caveat for Alum-based vaccines. Combining Alum with TLR-dependent, pro-Th1/Th17 adjuvants might dampen the pro-Th2 activity and improve the effectiveness of vaccine formulations. Here, using the ovalbumin (OVA) model of allergic lung inflammation, we found that sensitization with the synthetic TLR9 agonist, which is composed of oligodeoxynucleotides containing CpG motifs adsorbed to Alum, inhibited the development of OVA-induced lung allergic Th2 responses without shifting toward a Th1 pattern. The conversion of T cell immunity from the polarized allergic Th2 response to a non-polarized form by sensitization with OVA/Alum/CpG was dependent on MyD88 signaling in myeloid cells. Notably, sensitization of IL-10-deficient mice with OVA/Alum/CpG resulted in the development of neutrophilic lung inflammation associated with IFNγ production. However, in IL-10/IL-12-deficient mice, it resulted in neutrophilic inflammation dominated by IL-17 production. We conclude that OVA/Alum/CpG sensitization signaling via MyD88 and IL-10 molecules results in non-polarized immunity. Conversely, OVA/Alum/CpG sensitization in presence of MyD88 but absence of IL-10 or IL-10/IL-12 molecules results, respectively, in neutrophilic inflammation associated with IFNγ or IL-17 production. Our work provides novel OVA models of lung inflammation and suggests that Alum/CpG-based formulations might be of potential use in anti-allergic or anti-infectious processes.

Keywords: Alum; CpG-ODN; OVA model; T helper cells; TLR agonists; adjuvants; lung inflammation.

Figure 1

Toll-like receptors 4 and 9 agonists adsorbed to Alum prevent T helper type-2 allergic responses. C57BL/6 wild-type mice were sensitized with ovalbumin (OVA)/Alum or with OVA/Alum/CpG or with OVA/Alum/LPS on days 0 and 7 and challenged with OVA on days 14 and 21. Samples were obtained on day 22. (A) Total cell and (B) differential cell counts in bronchoalveolar lavage (BAL); (C) IL-5 and (D) IFNγ concentrations in BAL. (E) Representative microphotographs of periodic acid-Schiff staining and lung mucus score (see Materials and Methods); (F) total IgE and (G) OVA-specific IgE serum levels. Control group (n = 4) consisted of non-manipulated animals. OVA/Alum/phosphate buffer saline (PBS) group (n = 10), OVA/Alum/LPS group (n = 5), and OVA/Alum/PBS group (n = 5). Values represent the mean ± SD and are representative of three independent experiments. One-way ANOVA: *p < 0.05; **p < 0.01; ****p < 0.0001, different from OVA/Alum/PBS group.

Figure 2

TLR9 agonist (CpG) attenuates airway allergic responses in a dose-dependent manner. C57BL/6 wild-type mice were sensitized with ovalbumin (OVA)/Alum or with CpG at different doses (0.1, 1, or 10 μg/animal) absorbed to OVA/Alum on days 0 and 7 and challenged with OVA on days 14 and 21. Samples were obtained on day 22. (A) Total cell and (B) differential cell counts in bronchoalveolar lavage (BAL); (C) lung inflammation score by hematoxylin/eosin staining and (D) lung mucus score by periodic acid-Schiff staining (see Materials and Methods); (E) total IgE, (F) OVA-specific IgE, and (G) OVA-specific IgG2c serum levels. The percentage of lymphocytes in BAL was less than 10% in all groups. Values represent the mean ± SD of one experiment. One-way ANOVA: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, different from OVA/Alum/phosphate buffer saline (PBS) group (n = 5). One-way ANOVA: ••p < 0.01; ••••p < 0.0001, different between groups (n = 5).

Figure 3

Optimal inhibition of allergic sensitization by CpG requires adsorption to Alum. C57BL/6 wild-type mice were sensitized with ovalbumin (OVA)/Alum [phosphate buffer saline (PBS)] or with OVA/Alum plus CpG absorbed to OVA/Alum (CpG) or given by s.c. (CpG s.c.) or by intraperitoneal (i.p.) route (CpG i.p.) on days 0 and 7 and challenged with OVA on days 14 and 21. Samples were obtained on day 22. (A) Total cell and (B) eosinophil cell counts in bronchoalveolar lavage (BAL); (C) total IgE and (D) OVA-specific IgE serum levels; (E) FACS analysis of cytokine-producing CD4⁺ T cells performed with pooled lung cells from three animals. Values represent the mean ± SD of one experiment. One-way ANOVA: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, different from OVA/Alum/PBS group (n = 5). One-way ANOVA: ••p < 0.01; ••••p < 0.0001, different between groups (n = 5).

Figure 4

Attenuation of allergic sensitization by CpG requires signaling through MyD88 pathway but not IL-12/IFNγ axis. MyD88-knockout (KO), IL-12-KO, or IL-12/IFNγ-KO mice were sensitized with ovalbumin (OVA)/Alum or with OVA/Alum/CpG on days 0 and 7 and challenged with OVA on days 14 and 21. Samples were obtained on day 22. (A) Total cell and (B) eosinophil cell counts in bronchoalveolar lavage (BAL); (C) lung mucus score by periodic acid-Schiff staining; (D) total IgE and (E) OVA-specific IgE serum levels. Values represent the mean ± SD and are representative of two experiments. One-way ANOVA: n.s. (non-significant), *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, difference between OVA/Alum and OVA/Alum/CpG (n = 5).

Figure 5

MyD88 expression on adaptive lymphoid cells is not required for inhibition of allergic sensitization by CpG. (A) RAG^−/− mice on C57BL/6 background were reconstituted or not with 20 × 10⁶ spleen cells of C57BL/6 wild-type (WT) or MyD88-knockout (KO) mice. Fourteen days later, mice were sensitized with ovalbumin (OVA)/Alum or OVA/Alum/CpG on days 0 and 7 and challenged with OVA on days 14 and 21. Samples were collected on day 22. (A,F) Schematic protocol of reconstituted RAG^−/− mice with WT or MyD88-KO spleen cells; (B,G) total cell and (C,H) eosinophil cell counts in bronchoalveolar lavage (BAL); (D,I) total IgE and (E,J) OVA-specific IgE serum levels; (K,L) lung mucus (periodic acid-Schiff) and inflammation (hematoxylin/eosin) scores (see Materials and Methods); RAG^−/− group represents non-manipulated RAG^−/− mice. Control group represents reconstituted RAG^−/− mice without further manipulation. Values represent the mean ± SD and a representative of three independent experiments. One-way ANOVA: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, compared to OVA/Alum group (n = 7).

Figure 6

Alum-based CpG sensitization induces, respectively, in IL-10-knockout (KO) or IL-10/IL-12-KO mice airway inflammation dominated by IFNγ or IL-17. C57BL/6 wild-type and IL-10-KO mice were sensitized with ovalbumin (OVA)/Alum or with OVA/Alum/CpG on days 0 and 7 and challenged with OVA on days 14 and 21. Samples were obtained on day 22. (A) Total cell counts and (B) eosinophil and neutrophil numbers in bronchoalveolar lavage (BAL); (C) IL-4, (D) IL-10, (E) IFNγ, and (F) IL-17 in BAL; (G) total IgE, (H) OVA-specific IgE, and (I) OVA-specific IgG2c serum levels. Values represent the mean ± SD and are representative of two independent experiments. One-way ANOVA: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, difference between OVA/Alum and OVA/Alum/CpG (n = 5).

Figure 7

Lung histology and lung inflammation score of wild-type (WT), IL-10-knockout (KO), or IL-10/IL-12-KO mice. C57BL/6 WT, IL-10-KO, or IL-10/IL-12 double-KO mice were sensitized with ovalbumin (OVA)/Alum or with OVA/Alum/CpG on days 0 and 7 and challenged with OVA on days 14 and 21. Samples were obtained on day 22. Control group consisted of non-manipulated animals. (A) Representative microphotographs of lung sections stained with hematoxylin/eosin and (B) lung inflammation score (see Materials and Methods); values represent the mean ± SD and are representative of two independent experiments. One-way ANOVA: ••••p < 0.0001, different between groups (n = 5).