Adsorption of protein antigen to the cationic liposome adjuvant CAF®01 is required for induction of Th1 and Th17 responses but not for antibody induction

Abstract

The degree of antigen adsorption to adjuvants in subunit vaccines may significantly influence the immune responses they induce upon vaccination. Commonly used approaches for studying how the level of adsorption affects the induction of antigen-specific immune responses include (i) using adjuvants with different abilities to adsorb antigens, and (ii) comparing different antigens selected based on their ability to adsorb to the adjuvant. A weakness of these approaches is that not only the antigen adsorption level is varied, but also other important functional factors such as adjuvant composition and/or the B/T cell epitopes, which may affect immunogenicity. Hence, we investigated how changing the adsorption capabilities of a single antigen to an adjuvant influenced the vaccine-induced immune responses. The model antigen lysozyme, which displays a positive net charge at physiological pH due to an isoelectric point (pI) of 11, was succinylated to different extents, resulting in a reduction of the pI value to 4.4-5.9, depending on the degree of succinylation. A pronounced inverse correlation was found between the pI value of the succinylated lysozyme analogues and the degree of adsorption to a cationic liposomal adjuvant consisting of dimethyldioctadecylammonium bromide (DDA) and trehalose dibehenate (TDB) (CAF®01). Furthermore, increased adsorption to this adjuvant correlated directly with the magnitude of lysozyme-specific Th1/Th17 immune responses induced by the vaccine in mice, while there was an inverse correlation with antibody induction. However, high lysozyme-specific antibody titers were induced with an increased antigen dose, even upon vaccination with a strongly adsorbed succinylated lysozyme analogue. Hence, these data illustrate that the degree of lysozyme adsorption to CAF®01 strongly affects the quality of the resulting immune responses.

Keywords: Adjuvant; Adsorption; CAF01; Depot; Succinylation; T-cell response.

Graphical abstract

Fig. 1

Succinylation of lysozyme. (A) The amino acid sequence of lysozyme with possible succinylation i.e. lysine residues (K) highlighted in red and tyrosine residues (Y) highlighted in green (structure sourced from NCBI). (B) Succinylation of a primary amino group and a tyrosine residue on a protein (P) with succinic anhydride at pH 8.0, resulting in the formation of an amide bond and an ester bond, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Physicochemical characterization of succinylated LYS analogue mixtures evaluated by (A) native gel electrophoresis and (B) native gel electrophoresis with reversed current direction. Both native gels were run on a 4–15% Mini-PROTEAN® TGX™ Precast Gel with premixed electrophoresis buffer containing 25 mM Tris, 192 mM glycine (pH 8.3) at 300 V for 20 min with 2 µg protein loaded onto each well: (a) the reference Precision Plus ProteinTM Standards, (b) LYS-100, (c) LYS-10, (d) LYS-5, (e) LYS-0.5 and (f) unmodified lysozyme (LYS). (C) The average number-weighted hydrodynamic diameter measured by dynamic light scattering (DLS) with standard deviation (st.d) and the percent of the particle population at each size. (D) The secondary structure of succinylated LYS analogue mixtures determined by far-UV circular dichroism. All generated spectra are an average of three scans. The shown spectra have not been smoothed.

Fig. 3

Degree of succinylation of LYS analogues investigated by (A) isoelectric focusing (IEF) with (a) the reference IEF Marker 3–10, SERVA Liquid Mix, reference values are included, (b) LYS-100, (c) LYS-10, (d) LYS-5, (e) LYS-0.5 and (f) unmodified lysozyme (LYS). (B) MALDI-TOF spectra with data provided by Alphalyse (Odense, Denmark), molecular weight (Mw) [M] + showed in Dalton (Da) as a function of the intensity (%) and (C) table summing the mass of the LYS analogues from MALDI-TOF in Da with the most intense peak underlined together with the corresponding sites of succinylation with the Mw of one succinyl group is 100.07 Da. The theoretical isoelectric point (pI) is included in the table. The pI is calculated using ExPasy Bioinformatics Resource Portal (http://web.expasy.org/protparam/) and replacing lysine and tyrosine residues with glutamate residues. The primary amine group has not been replaced. The actual pI of succinylated lysozyme is included and read of the IEF gel. *The practical pI of unmodified LYS is found in literature .

Fig. 4

Adsorption of lysozyme to CAF®01. (A) Adsorption ratio of lysozyme to CAF®01 shown as the fraction of succinylated LYS analogue adsorbed per gram lipid as a function of the added antigen concentration (g/L). Data represent mean values of three technical replicates (n = 1). Values below zero were corrected to zero. (B) Percentage of LYS adsorbed to CAF®01 in each LYS analogue containing different concentrations (0.25–1 g/L) of LYS. Data represent mean values of three technical replicates due to shortage of succinylated LYS.

Fig. 5

Cell-mediated immune response. (A–D) Electrochemoluminiscence assay by MSD of supernatants from splenocytes from female BALB/C mice harvested three weeks post s.c. immunization with three times 200 µl 5 µg Ag/dose vaccine and restimulated with unmodified LYS for three days. (A) IFN-γ, (B) IL-17, (C) IL-5 and (D) IL-10 cytokine secretion. Data are shown as mean values ± max/min. Data are representative of three independent experiments. Statistical analysis was performed using one-way ANOVA with Dunnett’s multiple comparison test against the control group (LYS + CAF®01). P-values are stated in the figures.

Fig. 6

Antibody responses. Unmodified LYS-specific (A) IgG1 and (B) IgG2 antibodies were detected 3 weeks post immunization using an antibody enzyme-linked immunosorbent assay (ELISA). Female BALB/C mice were injected s.c. with 200 µl 5 µg Ag/dose of unmodified LYS alone (n = 5) or with CAF®01 or LYS-5 + CAF®01 or LYS-100 + CAF®01 (n = 8) three times, with 14 days between immunizations. Blood was harvested on day 49 (3 weeks p.i.) and the serum analyzed. Data are shown as mean ± SD. (C) and (D) Detection of unmodified LYS-specific (C) IgG1 and (D) IgG2a antibodies 2, 4, 6 and 8 weeks after the last immunization by ELISA. Data shown represent mean values ± SEM at a dilution factor of 10³. Female BALB/C mice were injected s.c. with 5 µg/dose LYS + CAF®01, 5 µg/dose LYS-100 + CAF®01, 20 µg/dose LYS-100 + CAF®01, or 40 µg/dose LYS-100 + CAF®01 (n = 4) three times, with 14 days between immunizations. Blood was harvested, and the serum was analyzed. Statistical analysis was performed using one-way ANOVA with Dunnett’s multiple comparison test for each time-point with LYS + CAF®01 (5 µg/dose) as reference. Significant differences [P-value < α (0.05)] are marked with #. The optical density (OD) was measured with an absorbance of 450 corrected at 650 nm.