Mixed adjuvant formulations reveal a new combination that elicit antibody response comparable to Freund's adjuvants

Abstract

Adjuvant formulations capable of inducing high titer and high affinity antibody responses would provide a major advance in the development of vaccines to viral infections such as HIV-1. Although oil-in-water emulsions, such as Freund's adjuvant (FCA/FIA), are known to be potent, their toxicity and reactogenicity make them unacceptable for human use. Here, we explored different adjuvants and compared their ability to elicit antibody responses to FCA/FIA. Recombinant soluble trimeric HIV-1 gp140 antigen was formulated in different adjuvants, including FCA/FIA, Carbopol-971P, Carbopol-974P and the licensed adjuvant MF59, or combinations of MF59 and Carbopol. The antigen-adjuvant formulation was administered in a prime-boost regimen into rabbits, and elicitation of antigen binding and neutralizing antibodies (nAbs) was evaluated. When used individually, only FCA/FIA elicited significantly higher titer of nAbs than the control group (gp140 in PBS (p<0.05)). Sequential prime-boost immunizations with different adjuvants did not offer improvements over the use of FCA/FIA or MF59. Remarkably however, the concurrent use of the combination of Carbopol-971P and MF59 induced potent adjuvant activity with significantly higher titer nAbs than FCA/FIA (p<0.05). This combination was not associated with any obvious local or systemic adverse effects. Antibody competition indicated that the majority of the neutralizing activities were directed to the CD4 binding site (CD4bs). Increased antibody titers to the gp41 membrane proximal external region (MPER) and gp120 V3 were detected when the more potent adjuvants were used. These data reveal that the combination of Carbopol-971P and MF59 is unusually potent for eliciting nAbs to a variety of HIV-1 nAb epitopes.

Figure 1. gp140_SF162 conformation upon formulation with adjuvant.

The gp140_SF162 glycoprotein was formulated with or without adjuvants and assayed on PAGE gel for conformation. (a) Regardless of presence or absence of adjuvants, a single band at 140 kDa was observed in the reduced gel. (b) Similar observation was made in the non-reduced gel, where all the samples showed two bands at 140 kDa and 280 kDa, representing the non-crosslinked and disulfide crosslinked gp140 trimers, respectively. (c) A BN-PAGE gel was used to examine the native conformation of gp140 in the presence of adjuvants. None of the adjuvants appear to affect the structure of gp140.

Figure 2. Monoclonal antibody binding to gp140_SF162 in the presence of adjuvant.

The antigenicity of gp140_SF162 in the presence of adjuvants was measured by mAb binding. A panel of 12 different mAbs, representing different epitope domains of the Env, was used. The signal of mAb binding to the adjuvant-formulated gp140_SF162 was calculated relative to the signal measured in the unadjuvanted sample, which is assigned to have the value of 1. The data is color coded: Yellow represents no difference in mAb binding, or no change of antigenicity as compared to the unadjuvanted sample and light blue represents up to 3-fold decrease of mAb binding (or antigenicity). Both FIA and MF59 adjuvants appeared to have little effect on antigenicity, as mAb binding was very similar to the unadjuvanted sample. On the other hand, decreased mAb bindings were detected in the presence of both of the Carbopols.

Figure 3. Endpoint IgG binding titer.

The antigen-specific IgG titer was measured by ELISA. Antiserum (terminal time point) was serially diluted (2-fold) and the endpoint titer was defined as the last dilution that gave a positive signal (>3-fold signal of the prebleeds). Animals that received the FCA/FIA (groups 1 and 2) or combination of Carbopol-971P and MF59 have significantly higher endpoint IgG titers than others (p<0.05).

Figure 4. Relative avidity index.

Avidity of the antiserum was measured by its ability to remain in binding with gp140_SF162 in the presence of 8 M Urea. All samples were found to have relative high avidity, defined as >50% IgG binding in the presence of 8 M Urea, as compared to the no Urea control. Compared to the unadjuvanted group (no. 12), antisera from animals vaccinated with adjuvants have significantly higher RAI (p<0.05).

Figure 5. Epitope specificity across all antisera by mAb competition.

Across all samples, the polyclonal antibody responses appeared be mostly CD4bs and CD4i driven, as the antisera could efficiently out-competed mAbs against these two regions (b12 and VRC01 for CD4bs and 17b and E51 for CD4i). Modest to little activities against the V3 loop or the MPER were detected.

Figure 6. Antisera competition to different mAbs.

Antiserum was used to compete with Eu-labeled mAbs specific for different antigenic domains of Env. The fluorescent signal of Europium is proportional to residual mAb binding, relative to the prebleed controls. A decrease in residual mAb binding indicated an increase in specificity of the antiserum for the epitope site. Both the FCA/FIA (groups 1&2) and the combination of Carbopol-971P and MF59 (group 10) displayed potent activities towards both CD4bs and CD4i and modest activities towards the V3 loop and MPER. By contrast, antisera from animals vaccinated with other adjuvants displayed only modest activities to CD4bs and CD4i and little, if any, towards V3 and MPER.

Figure 7. 50% Neutralization titer of individual pseudovirus between groups.

For single adjuvant, animals vaccinated with FCA/FIA (groups 1 and 2) have the highest titers against the tier 1A viruses, and low titers against tier 1B. Sequential immunization with different prime-boost combination of Carbopol-971P/974P and MF59 (groups 6–9) performed no better than using FCA/FIA or MF59 alone. By contrast, the combination of Carbopol-971P and MF59 (group 10) conferred significantly higher neutralization titers over all others (p<0.05), particularly against the tier 1B viruses DJ263.8 and SS1196.1.