Mimicking immune complexes for efficient antibody responses

Abstract

Efficient antibody responses are crucial for combating infectious diseases and vaccination remains a cornerstone of this effort. This study introduces a novel approach for enhancing immune responses in wild-type mice by utilizing pre-formed immune complexes, using the receptor-binding domain (RBD) of SARS-CoV-2 as a model antigen to illustrate the broader potential of the concept. Specifically, we found that pre-treating the antigen with bis-maleimide, a chemical linker that facilitates protein cross-linking, significantly enhances antibody production. Moreover, in vitro cross-linking of antigen to unrelated IgG using bis-maleimide generated immune complexes that markedly enhanced antigen-specific antibody responses, likely by mimicking natural memory-like mechanisms, suggesting that bis-maleimide pre-treated antigens may similarly engage IgG in vivo. In contrast, antigen crosslinking with IgA or IgM did not yield comparable effects, highlighting the unique capacity of IgG to boost immunogenicity. By leveraging the principles of immune memory, this study demonstrates the potential of pre-formed immune complexes to significantly enhance vaccine efficacy using an antigen-independent strategy broadly applicable to diverse pathogens.

Keywords: B cells; IgG; adjuvant; antibody responses; memory.

Figure 1

Antibody responses by native or complex RBD. (A) Schematic illustration of SARS-CoV-2 spike protein: Receptor-binding domain (RBD, orange), which interacts with human angiotensin converting enzyme 2 (ACE2) and thereby mediates entry of viral particles into the host cell was described as a target for neutralizing antibodies. Native RBD (nRBD) was produced in HEK293-6E cells, biotinylated and complexed with streptavidin (SAV). (B) nRBD (~27kDa) was biotinylated and complexed by addition of SAV, samples were separated under non-reducing (- β-mercaptoethanol (2-ME)) and reducing conditions (+ 2-ME) on a 10% SDS-PAGE and stained with Coomassie. RBD forms self-aggregates that can be dissolved by reducing disulfide bonds with 2-ME, highlighted by red rectangle. (C) Schematic overview of immunization procedure: WT mice were either control-immunized (CI), immunized intraperitoneally (i. p.) with 50 µg of native RBD (nRBD) or biotinylated RBD complexed with SAV (cRBD) in presence of CpG-ODN #1826 as adjuvant. Immunization was repeated on day 21 in CI, nRBD- and cRBD-immunized mice with the same vaccination composition used for primary immunization. Serum was collected and analyzed for RBD-specific antibodies on days 7, 14 and after booster vaccination on day 28. (D–F) Serum was harvested from immunized mice at the indicated time points after primary vaccination and RBD-specific IgM (D) and IgG (E) titers were determined by indirect ELISA. 7 days following booster immunization, serum was collected and the measurement of RBD-specific IgM and IgG levels was repeated for day 28 (F). CI, n = 29; nRBD, n = 4; cRBD, n = 8 or 10, respectively. Mean ± SD, statistical significance was calculated by using the Kruskal-Wallis test.

Figure 2

Robust antibody responses by RBD complexes generated by chemical cross-linking (A) Native (n)RBD (~27kDa) was produced in HEK293-6E cells and complexed by addition of 1,2-phenylen-bis-maleimide (1,2-PBM) for generation of RBD*. Samples were separated under non-reducing (- 2-ME) and reducing conditions (+ 2-ME) on a 10% SDS-gel and detected by Coomassie-staining. (B) Serum was collected from mice immunized with nRBD (n = 7) and RBD* (n = 11), as illustrated in Supplementary Figure S2C on day 28. RBD-specific IgG concentrations were measured by ELISA and compared to titers measured in CI mice. Mean ± SD, statistical significance was calculated by applying the Mann-Whitney-U test. (C) The neutralizing potential of generated antibodies was analyzed by a neutralization assay using pseudoviral particles in 5 serum samples from (B) as compared with neutralizing (blue) and non-neutralizing (gray) human serum as positive or negative control, respectively.

Figure 3

Activated RBD is required for enhanced immune responses. (A) Schematic illustration of the “quenching” reaction in which activated RBD* is inactivated by exposure to cysteine. (B) RBD was generated by addition of 1,2-PBM to nRBD. RBD* was subsequently incubated in presence of cysteine. Samples were separated under non-reducing (- 2-ME) and reducing conditions (+ 2-ME) on a 10% SDS-PAGE and detected by Coomassie-staining. (C) WT mice were immunized either with 50 µg of RBD* (n = 11) or “quenched” RBD*Cys, which was dialyzed before injection (n = 9). Serum was collected on day 28, one week after secondary immunization and RBD-specific IgG concentrations were measured by ELISA. Anti-RBD IgG concentrations elicited by RBD*Cys were compared to the respective concentrations, induced by RBD* immunization, which were already shown in Figure 2B . Mean ± SD, statistical significance was calculated by applying the Mann-Whitney-U test.

Figure 4

IgG is required for efficient immune responses by RBD* (A) Schematic illustration of cross-linking RBD* via 1,2-PBM with polyclonal murine antibodies of IgG isotype. (B) Dot blots of IgG immunoprecipitates against RBD. nRBD was activated with 1,2-PBM to generate RBD* and subsequently dialysed against PBS to remove free 1,2-PBM. Dialysed RBD* was either directly incubated with IgG (RBD* + IgG) or following inactivation with cysteine-solution (RBD*Cys + IgG). All samples were precipitated for IgG and developed against RBD. Representative data from 3 individual experiments are shown. (C) Dot blots of IgG immunoprecipitates against RBD. RBD*IgG generated by activating nRBD with 1,2-PBM in presence of IgG was used as control. Samples were precipitated for IgG and developed against RBD. Representative data from 3 individual experiments are shown. (D) Immunization was performed in WT mice using 50 μg of RBD* (n = 11), or RBD complexed with 1,2-PBM in presence of 25 μg polyclonal murine IgG (RBD*IgG, n = 11), IgM or IgA (RBD*IgM, n = 3 or RBD*IgA, n = 3; data already shown in Supplementary Figure S4D ). 50 μg CpG-ODN #1826 was used as adjuvant in all conditions. Serum was collected on day 28, one week after secondary immunization and RBD-specific IgG concentrations were measured by ELISA. Anti-RBD IgG concentrations elicited by RBD* complexed with immunoglobulins of different isotypes were compared to the respective concentrations, induced by RBD* immunization, which were already shown in Figure 2B and Figure 3C . Mean ± SD, statistical significance was calculated by applying the ordinary one-way ANOVA. (E) Immunization was performed in WT mice as described in Supplementary Figure S2C , using RBD* (n = 11), or RBD*IgG (n = 11) in presence or absence of 50 μg CpG-ODN #1826 as adjuvant. Serum was collected on day 28, one week after secondary immunization and RBD-specific IgG concentrations were measured by ELISA. Anti-RBD IgG concentrations elicited by RBD* and RBD*IgG in absence of CpG-ODN #1826 were compared to the respective concentrations, induced by RBD* and RBD*IgG in presence of adjuvant, which were already shown in Figures 2B , 3C and 4D . Mean ± SD, statistical significance was calculated by applying the Kruskal-Wallis test. (F) The neutralizing potential of generated antibodies was analyzed by a neutralization assay using pseudoviral particles in 5 serum samples from the RBD*IgG group, shown in (D) and (E). The 50% neutralization titer (NT50, right panel) was compared between serum samples used in the neutralization assay shown in Figures 2C and 4F (left panel). Mean ± SD, statistical significance was calculated by applying the Mann-Whitney-U test.

Figure 5

IgG immune complexes act with other adjuvants. (A) RBD*IgG was generated as described previously. RBD* was subsequently incubated in presence of either Alum or CpG. Samples were separated under reducing conditions (+ 2-ME) on a 10% SDS-PAGE and detected by Coomassie-staining. (B) RBD*IgG complexes were generated according to the previously described standard procedure. Prior to immunization, Alum was added to the previously prepared complexes and both nRBD or KLH, were used as controls. Serum was collected from immunized mice on day 28, one week after secondary immunization and RBD-specific IgG concentrations were measured by ELISA. n = 5, mean ± SD, statistical significance was calculated by applying the paired t test. (C) RBD*IgG complexes were generated as previously described and incubated in presence or absence cysteine (Cys). Immunization was performed in WT upon addition of either CpG or Alum as adjuvants. Serum was collected on day 28, one week after secondary immunization and RBD-specific IgG concentrations were measured by ELISA. n = 5, mean ± SD, statistical significance was calculated by applying the paired t test or Mann-Whitney-U test. (D) Immunization was performed in WT mice using 50 μg of RBD complexed by 1,2-PBM in presence of 25 μg polyclonal murine IgG, and subsequent addition of either CpG or Alum as adjuvant. Serum was collected on day 28, one week after secondary immunization and RBD-specific IgG concentrations were measured by ELISA. n = 5, mean ± SD, statistical significance was calculated by applying the paired t test. (E) KLH*IgG complexes were generated according to the described standard procedure with KLH at a concentration of 100 µg per mouse. Prior to immunization, Alum was added to the previously prepared complexes and untreated KLH was used as control. Serum was collected from immunized mice on day 28, one week after secondary immunization and RBD-specific IgG concentrations were measured by ELISA. n = 5, mean ± SD, statistical significance was calculated by applying the paired t test.

Figure 6

Autoreactive antibodies upon immunization. (A) Anti-RBD IgG concentrations were compared in serum from mice immunized twice with Comirnaty^® from BioNtech/Pfizer (BNT, n = 9) and RBD*/RBD*IgG (+CpG, n = 12), which reached the highest concentrations in previous experiments. Pre-immune (n = 22), mean ± SD, statistical significance was calculated by applying the Kruskal-Wallis test. (B) Paired comparison of anti-ProC IgG concentrations in mice prior to and after RBD*/RBD*IgG (+CpG) immunization (left, n = 6) or Comirnaty^® immunization (right, n = 5), respectively. Mean ± SD, statistical significance was calculated by applying the paired t test. (C) Grouped comparison of anti-ProC IgG concentrations from (C) Mean ± SD, statistical significance was calculated by applying the ordinary one-way ANOVA.