Salmonella Typhi Vi capsule prime-boost vaccination induces convergent and functional antibody responses

Abstract

Vaccine development to prevent Salmonella Typhi infections has accelerated over the past decade, resulting in licensure of new vaccines, which use the Vi polysaccharide (Vi PS) of the bacterium conjugated to an unrelated carrier protein as the active component. Antibodies elicited by these vaccines are important for mediating protection against typhoid fever. However, the characteristics of protective and functional Vi antibodies are unknown. In this study, we investigated the human antibody repertoire, avidity maturation, epitope specificity, and function after immunization with a single dose of Vi-tetanus toxoid conjugate vaccine (Vi-TT) and after a booster with plain Vi PS (Vi-PS). The Vi-TT prime induced an IgG1-dominant response, whereas the Vi-TT prime followed by the Vi-PS boost induced IgG1 and IgG2 antibody production. B cells from recipients who received both prime and boost showed evidence of convergence, with shared V gene usage and CDR3 characteristics. The detected Vi antibodies showed heterogeneous avidity ranging from 10 μM to 500 pM, with no evidence of affinity maturation after the boost. Vi-specific antibodies mediated Fc effector functions, which correlated with antibody dissociation kinetics but not with association kinetics. We identified antibodies induced by prime and boost vaccines that recognized subdominant epitopes, indicated by binding to the de–O-acetylated Vi backbone. These antibodies also mediated Fc-dependent functions, such as complement deposition and monocyte phagocytosis. Defining strategies on how to broaden epitope targeting for S. Typhi Vi and enriching for antibody Fc functions that protect against typhoid fever will advance the design of high-efficacy Vi vaccines for protection across diverse populations.

Fig. 1.. Sequence characteristics of BCRs from IgG-positive plasmablasts isolated after Vi-TT primary or Vi-PS boost immunization.

(A) Subclass distribution for n = 3 participants, in response to prime (Vi-TT) and boost (Vi-PS) in a paired fashion. (B) Heavy-chain CDR3 length after Vi-TT prime (n = 3) or Vi-PS-boost (n = 3), showing paired samples only. (C to E) Individual heavy-chain CDR3 length after Vi-TT prime or Vi-PS boost. Gray lines are based on a normal distribution with the mean and SD of the CDR3 length data for sample shown. (F) Ig heavy-chain variable region usage in paired samples, comparing Vi-TT prime and Vi-PS boost. Four heavy-chain regions of interest are shown. Further regions are shown in fig. S4.

Fig. 2.. Selection strategy for the identification of Vi-specific mAbs.

(A) Family size of sequences identified in participant 8624 after Vi-PS boost. Families with fewer than three sequences were excluded for the purpose of visualization. (B) Example of a family from 8624, selected on the basis of a combination of expansion and persistence. (C) Example of a selected convergent antibody pair. (D) Strategy used to select each of the 96 mAbs, split by each of the different donors. Antibodies categorized on the basis of persistence were also expanded. (E) Vi binding of n = 96 isolated mAbs by BLI, where blue dots indicate mAbs isolated from prime time point. (F) Frequency of Vi binding mAbs for each of the selection strategies used, as measured by BLI. (G) Vi status of the n = 96 selected monoclonals, stratified by CD62L status.

Fig. 3.. No evidence for affinity maturation of Vi polysaccharide antibodies after repeat vaccination.

(A) Isoaffinity plot indicating avidity measurements for mAbs (n = 53) to polymeric Vi polysaccharide by BLI. Diagonal lines indicate the K_D values shown. (B) Spearman correlation of Vi antibody avidity and binding response to Vi antigen. Negative correlation represents decrease in K_D, which corresponds to an increase in avidity. (C) Avidity measurements (K_D) of Vi binding mAbs stratified by Vi-TT prime/Vi-PS boost. (D) Avidity measurements (K_D) of Vi binding mAbs stratified by participant, with blue indicating mAbs from prime time points. (E) Total serum (T1, n = 10; T2, n = 8; T3, n = 10) was used to measure residual Vi antibody binding after NaSCN exposure for 15 min in ELISA, as a proxy for avidity measurements. The avidity index was calculated as absorbance at 1 M NaSCN/(absorbance at 0 M NaSCN − absorbance at 5 M NaSCN). T1, 28 days after prime; T2, before Vi-PS boost; T3, 28 days after Vi-PS boost.

Fig. 4.. Convergent development of Vi-specific antibody sequences among Vi vaccine recipients.

(A) mAbs (n = 12) from vaccinees with the same variable heavy-chain gene segment were grouped and tested for similarity on the amino acid level by calculating the Levenshtein edit distance. Sequences from vaccinees with ≥75% identity were defined as a convergent group (G1 to G5). Convergent groups based on variable heavy- and light-chain gene segment usage with participant, time point, H- and L-CDR3 sequence, and antibody avidity for Vi antigen. Bold black letters indicate identical amino acids, teal letters indicate amino acids with similar chemical characteristics, and pink letters indicate different amino acids. Chemical characteristics were defined by grouping amino acids based on electrically charged, polar uncharged, and hydrophobicity. (B) Quantification of H-CDR3 sequence similarity by BLOSUM62 calculation of similarity index. (C) Cross-competition of convergent antibodies within convergent groups by BLI, where the first mAb was bound to Vi to complete saturation before association with the second mAb. Red indicates self-blocking, pink indicates cross-blocking, and green indicates no blocking.

Fig. 5.. Antibodies derived from Vi vaccinees recognize de–O-acetylated Vi.

(A) Molecular structure of Vi polysaccharide monomer with C3 O-acetyl conversion to ¹H after NH₄OH treatment. (B) Increasing signal intensity of two-dimensional NMR spectra centered around 1.9 ppm (¹H) and 26.5 ppm (¹³C), where new C3 ¹H-¹³C bond appears as a function of increasing amounts of NH₄OH treatment (n = 7) of polysaccharide for de–O-acetylation. (C) Binding sensorgram of AB-008053 IgG1 to Vi PS and De-O-Vi PS, where Vi is immobilized on biosensor and dipped into wells containing antibody. (D) Heatmap of binding response of all Vi binding mAbs (n = 53) to Vi PS and De-O-Vi PS antigens. Asterisk indicates that mAb was selected after Vi-TT prime. Left: mAbs with equal or greater binding to De-O-Vi PS compared with Vi PS (n = 6). Right: mAbs with diminished binding to De-O-Vi PS compared with Vi PS (n = 47). (E and F) Example binding sensorgrams of mAbs interacting similarly (E) to Vi PS (left) and De-O-Vi PS (right) or better (F) to De-O-Vi PS with corresponding avidity measurements indicated. Ф indicates lower antigen density used for antibody titration to reliably measure rate constants.

Fig. 6.. Vi PS mAbs recognize exposed and occluded epitopes.

(A) Competition sensorgram of O-acetyl targeting mAb AB-008053 as antibody 1 followed by antibodies (n = 3) targeting a subdominant but exposed epitope (SEE) as antibody 2. (B) Competition sensorgram of O-acetyl targeting mAb AB-008053 as antibody 1 followed by antibodies (n = 3) targeting a subdominant and occluded epitope (SEO) as antibody 2. (C) Cross-competition matrix of SEE (n = 3) and SEO (n = 3) targeting mAbs against each other with colored gradient, with blue indicating no blocking and pink indicating cross-blocking.

Fig. 7.. Vi PS antibodies mediate non-neutralizing Fc effector functions that are highly associated with slow antibody dissociation rates.

(A) ADMP scores and (B) ADCD scores for Vi binding mAbs (n = 53), where blue indicates antibodies that bind to subdominant exposed epitopes (SEE, n = 3) and pink indicates antibodies that bind to subdominant occluded epitopes (SEO, n = 3). (C) Spearman correlation of ADMP and ADCD scores for Vi mAbs (n = 53). (D) Spearman correlations of ADMP with antibody off-rate (left) and avidity (right). (E) Spearman correlations of ADCD with antibody off-rate (left) and affinity (right). Negative correlations represent decreases in off-rate and K_D, which corresponds to increases in avidity.