T cell independent antibody responses with class switch and memory using peptides anchored on liposomes

Abstract

Vaccines generally require T lymphocytes for B-cell activation and immunoglobulin class switching in response to peptide or protein antigens. In the absence of T cells, limited IgG class switch takes place, germinal centers are short-lived, and the B cells lack memory. Here, immunization of mice with liposomes containing 15mer peptides and monophosphoryl lipid A (MPLA) as adjuvant, induced T-cell independent (TI) IgG class switch within three days, as well as germinal center formation. The antibody responses were long-lived, strictly dependent on Toll-like receptor 4 (TLR4) signaling, partly dependent on Bruton's tyrosine kinase (BTK) signal transmission, and independent of signaling through T-cell receptors, MHC class II and inflammasome. The antibody response showed characteristics of both TI type 1 and TI type 2. All IgG subclasses could be boosted months after primary immunization, and the biological function of the secreted antibodies was demonstrated in murine models of allergic anaphylaxis and of bacterial infection. Moreover, antibody responses after immunization with peptide- and MPLA-loaded liposomes could be triggered in neonatal mice and in mice receiving immune-suppressants. This study demonstrates T-cell independent endogenous B-cell memory and recall responses in vivo using a peptide antigen. The stimulation of these antibody responses required a correct and dense assembly and administration of peptide and adjuvant on the surface of liposomes. In the future, TI vaccines may prove beneficial in pathological conditions in which T-cell immunity is compromised through disease or medicines or when rapid, antibody-mediated immune protection is needed.

Fig. 1. Peptide- and MPLA-loaded liposomes stimulate IgG class switch in vivo and independent on T cells, but dependent on TLR and BTK signaling.

a Schematic representation of ovalbumin (OVA) peptide- and MPLA-loaded liposomes. The peptide is anchored in the liposomal membrane by palmitoyl chains coupled on both ends of the peptide. The adjuvant monophosphoryl lipid A (MPLA) is integrated in the liposomal membrane through six acyl chains. b OVA-specific antibodies determined in BALB/c mice (n = 6) six weeks after two immunizations (days 0 and 7) with the indicated OVA peptide-liposomes (OVAaa) or with OVA protein on aluminum (OVA-Alu). Data were analyzed using a Kruskal Wallis test comparing any group to OVA-Alum. c Rapid OVA-specific IgG class switch in BALB/c mice after injection single immunization with 10 µg OVA_58-72 liposomes (n = 3). d–g OVA-specific antibody responses in d BALB/c wild-type (WT) and athymic (Nude) mice, e C57BL/6 WT, TCR-deficient (TCR^−/−) and MHC class II-deficient (MHC II^−/−) mice, f C57BL/6 WT and TLR4-deficient (TLR4^−/−) mice or g CBA WT and syngeneic BTK-deficient mice (BTK^−/−) six weeks after two immunizations with Lip-OVA_58-72 as indicated above (n = 3-5). Antibody levels are expressed as reciprocal geometric mean (GM) ± GSD of titers b, d–g or as OD ± SD c. Data were representative of two to four independent experiments with comparable results. *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 2. T-independent and long-lived germinal center formation.

BALB/c WT and nude mice were immunized with 10 µg Lip-OVA_58-72 or OVA-Alu. On day 5, mouse spleens from WT mice were analyzed histologically by staining with PNA a or for B220 b. Spleen from WT or nude were analyzed by immune fluorescence staining with PNA (red), B220 (green) and DAPI for determination of germinal center formation c. Splenocytes from immunized (n = 5) or untreated (n = 3) BALB/c mice were stained with B220, GL-7 and CD38 antibodies for flow cytometry, the germinal center being defined as B220-positive, GL-7-positive and CD38-negative as shown in zebra plots. The B-cell analysis was not antigen-specific. Lip-OVA_58-72–treated (n = 5) and untreated mice (n = 3) were compared statistically using Mann-Whitney U test.

Fig. 3. Antibodies are long-lived with affinity maturation and B-cell memory responses taking place independent on T cells.

WT BALB/c and athymic nude mice were immunized with 10 µg Lip-OVA_58-72 and boosted 14 weeks later. a Specific IgG1, IgG2a, IgG2b and IgG3 antibody responses to the whole OVA were measured at the indicated time points and the results are expressed as mean ± SD (n = 5). b Avidity of the measured antibodies on week 23 after immunization as analyzed by competitive ELISA. The number above the bars indicate the avidity fold increase between weeks 1 and 23.

Fig. 4. IgM and cytokine secretion from B cells are dependent on inflammasome activation.

a–f Naive B cells from spleens mice were stimulated in vitro with 10 µg/ml Lip-OVA_58-72 for two days. a Cells and supernatants were harvested to measure the IgM expression and total IgM by PCR and ELISA, respectively. Untreated B cells were used as negative controls. b IgM secretion of B cells of C57BL/6, TRL4-, MyD88- or TRIF-deficient mice after in vitro stimulation with Lip-OVA_58-72. c TNF-α and IL-10 secretion of B cells of C57BL/6 mice after stimulation with Lip-OVA_58-72 or LPS (5 µg/ml) as compared to untreated cells (n = 5). d, e IgM secretion from Lip-OVA_58-72-stimulated B cells of WT or caspase 1-deficient mice d, or from Lip-OVA_58-72-stimulated WT B cells in the presence or absence of pan caspase-inhibitor z-VAD e. f OVA-specific IgG2b and IgG3 responses (OD_1:150) in WT C57BL/6 or caspase 1-deficient mice (n = 5) as measured on day 12 (top panel) and day 35 (bottom panel) after immunization with Lip-OVA_58-72. g OVA- (gray bars) specific IgG2b and IgG3 responses (OD_1:150) in WT or ASC-deficient (ASC) mice (n = 5) after immunized with Lip-OVA_58-72. Means and SD are shown. Kruskal Wallis test or Mann-Whitney U test were applied for statistical analysis. *p < 0.05; **p < 0.01; n.s.: not significant.

Fig. 5. Overcoming immune suppression with T-cell independent vaccines.

Cultures of purified B cells a or whole splenocytes b from C57BL/6 mice were stimulated two days in vitro with Lip-OVA_58-72 a or concanavaline A b in the presence of immunosuppressive dexamethasone (blue), tacrolimus (pink) or cyclosporine A (ochre). IgM from B cells a and IL-2 from T cells b secretion were measured in the culture supernatants by ELISA. The indicated p values were calculated by Kruskal Wallis and testing if treatment affects readout. c OVA-specific IgG2b and IgG3 responses in C57BL/6 mice immunized with Lip-OVA_58-72 in the absence or presence of indicated immunosuppressive drugs (n = 4–6). Means and SD are shown. The area under the curve for each group was calculated and compared to untreated (Untr) mice by Kruskal Wallis with Dunn’s multiple comparison tests.

Fig. 6. Neonatal immunization with TI liposomal OVA_58-72 or TD aluminum OVA vaccines.

BALB/c pups (n = 5-12) from naive mothers were immunized with either Lip-OVA_58-72 (red) or OVA-Alu (black) at 1, 4, 7, 14 or 35 days of age. OVA-specific IgG subclass responses were determined two weeks later. Data are presented as geometric mean with 95% Cl. Mixed-effect model with Geisser-Greenhouse correction was used to compare statistically antibody responses between the two groups. ns: not significant.

Fig. 7. Prophylactic immunization with Lip-OVA_58-72 reduced anaphylactic responses in OVA-allergic mice.

BALB/c mice (n = 5) were immunized twice (weeks 0 and 4) with Lip-OVA_58-72 or OVA protein on alum (Alu-OVA) or left untreated (No immunization). On weeks 8 and 11, all mice were subject to sensitization with OVA protein on alum. One group of non-immunized mice were left non-sensitized (No sensitization). On week 14, all mice were challenged intraperitoneally with soluble OVA and the body temperature monitored as a measure for allergic anaphylaxis; the integrated AUC for the body-temperature curves was calculated using pre-challenge temperature as baseline. Results are expressed as mean + SEM and analyzed using Kruskal-Wallis tests comparing the two immunized groups to the non-immunized group. Statistical p values are indicated.

Fig. 8. Immunization with liposomal TI vaccines reduced bacterial infection in mice.

BALB/c mice (n = 5) were immunized with a combination of Lip-OVA_58-72 and Lip-OVA_179-195 (Lip), with OVA protein on alum (Alu) or left untreated. On day 7, all mice were infected with OVA-expressing L monocytogenes. On day 9, the mice were euthanized and the bacterial burden in liver was determined by plating tissue homogenates on agar and calculating CFU. Geometric means are shown and the results were analyzed using Kruskal Wallis tests comparing the two immunized groups to the non-immunized group. Statistical p values are indicated.