Antigen-driven Convergent Evolution of Polysaccharide-specific "DH-less" B Cells in Glycoconjugate Immunized Mice

Abstract

Glycoconjugate vaccines elicit robust anti-polysaccharide Ab response by recruiting T-cell help. Multiple doses of glycoconjugate vaccine are required to induce long-lasting immunity. The characteristics of anti-polysaccharide Ab response have been reported previously. However, the effect of glycoconjugate booster immunization on anti-polysaccharide and anti-carrier protein Ab repertoire remains poorly understood. In this study, we used clinically relevant pneumococcal capsular polysaccharide type 14 (PCP14) conjugated with cross-reactive material 197 (CRM197) as a model glycoconjugate Ag (PCP14-CRM197). We performed a comprehensive sequence analysis of mouse mAbs generated against PCP14 and CRM197 following immunization with one or three doses of PCP14-CRM197. Analysis of the paired Ig H and L chain transcripts revealed that anti-PCP14 Ab repertoire is extremely restricted. The reoccurrence of five replacement mutations at identical positions in anti-polysaccharide mAbs generated from different mice provided evidence for Ag-driven selection in PCP14-specific B cells. Convergent evolution was observed wherein distinct V(D)J rearrangements resulted in identical or nearly identical CDR3 in anti-PCP14 mAbs. Abs that lacked DH encoded amino acids dominated the anti-PCP14 Ab response. In contrast, anti-CRM197 Ab response was quite diverse, with fewer mutations compared with the anti-PCP14 mAbs, suggesting that conjugation of the polysaccharide to a carrier protein interferes with the development of carrier protein-specific Ab responses. Our findings provide molecular insights into the maturation of Ab responses driven by booster doses of glycoconjugate. This has fundamental implications for the design of glycoconjugate vaccines, especially where the development of Ab response against the carrier protein is also crucial.

FIGURE 1.

Serum anti-PCP14 and anti-CRM197 Ab response kinetics following immunization with PCP14-CRM197. Groups of female BALB/c mice (n = 10) were immunized with PCP14-CRM197 with alum, i.p. In the case of the single-dose regimen, the mice received a single dose at day 0 (shown as filled triangle below the y-axis), whereas in the case of the three-dose regimen, the mice were immunized on days 0, 14, and 28 (shown as open triangles). Control groups received only alum. The mice were bled at weekly intervals until week 14. The sera samples were pooled, and anti-PCP14 (A, C, and E) and anti-CRM197 (B, D, and F) endpoint Ab titers were determined by ELISA. The one- and three-dose regimens are indicated by filled and open circles, respectively. The curves were nudged in (D) for the ease of visualization.

FIGURE 2.

PCP14- and CRM197-specific mAbs exhibit distinct Ig gene segment usage. (A–D) Variable (VH and VL; A and C, respectively) and joining (JH and JL; B and D, respectively) gene segments used in the Ig H and L chain transcripts expressed in the PCP14-specific (blue) and CRM197-specific (red) hybridomas are shown. Only the gene segments that were present in our hybridoma set are shown. ^$The VH gene segment used could be either VH1-4 or VH1-7. ^#The VH gene segment used could be either VH1-4, VH1-7, or VH1-26. ^&The VH gene segment used could be either VH1-64 or VH1-S81. Dark and light shades represent the single- and triple-dose regimens, respectively. (E) Ig H and L chain pairing for the PCP14- and CRM197-specific mAbs is shown. The size of the circle is proportionate to the relative number of times a given VH and VL pair was observed in our set of mAbs. The Vλ and Jλ gene segment usage and VH–VL pairing for Igλ chain is not shown.

FIGURE 3.

CDR3 length distribution in anti-PCP14 and anti-CRM197 mAbs. HCDR3 (A) and LCDR3 (B) lengths (in codons) for the PCP14- and CRM197-specific mAbs are plotted. CDR3 lengths for the anti-PCP14 and anti-CRM197 mAbs are represented as black and gray bars, respectively.

FIGURE 4.

Clonal expansion of PCP14-specific B cells in mice immunized with PCP14-CRM197. Sequence alignment of the Ig H and L chain genes expressed in hybridomas derived from clonally related B cells with their germline counterpart are shown. Clone 1 comprising B07B02, F03A11, and F04F09 and clone 2 with members D03E02 and B12G10 are presented in (A) and (B), respectively. The nucleotide sequence of germline VH, DH, JH, VL, and JL gene segments used are shown on top. The suffixes “-H” and “-L” denote H and L chain genes expressed in the anti-PCP14 hybridomas, respectively. The FR, CDR, and codon positions are according to the IMGT numbering system. The amino acid translation (in single-letter code; in italics) is given below the codons. Nucleotide identity with the corresponding nucleotide in the germline gene is indicated by a dash (-). The figure shows only those codon positions where mutation(s) have occurred. Replacement and silent mutations are shown in upper and lowercase, respectively. CDR3 sequence is, however, presented in full. P- and N-nucleotide addition(s) are indicated, and the nucleotides lost due to exonuclease activity in the CDR3s during the V(D)J recombination are shown as missing. Deduced genealogical trees for clones 1 (C) and 2 (D) are shown as schematic diagrams. The number of nucleotide changes observed in the expressed H chain gene is written first, followed by the number of changes in the expressed L chain gene (italicized). HP denotes the hypothetical intermediate precursor. V(D)J rearrangements used in the hypothetical germline precursor and the isotypes of the expressed H and L chains are also indicated.

FIGURE 5.

Booster shots of PCP14-CRM197 drives mutational load in PCP14 specific B cells. Mutational load in the rearranged Ig H and L chain genes in the PCP14- and CRM197-specific hybridomas. Representative mAb from each clone was considered for the analysis. The number of mutations in PCP14-specific (A) and CRM197-specific (B) hybridomas obtained from the one-dose (filled circle) and three-dose (open circle) regimens are shown. The mutational loads in mAbs of IgM and IgG isotype are represented in red and blue, respectively. Error bars represent means ± SEM. The p values were calculated using two-tailed Mann–Whitney U test.

FIGURE 6.

Recurrent linked replacement mutations observed in PCP14-specific mAbs. The amino acid sequence of the Ig H chain expressed in PCP14-specific mAbs was aligned with the corresponding germline VH gene segment VH1-26. The codon position of recurrent amino acid replacement mutation is shown at the bottom of the multiple sequence alignment. The FRs and CDRs were assigned according to the IMGT numbering system. The time point following immunization at which the hybridomas were generated is indicated on the right. Amino acid identity with the corresponding residue in the germline gene is indicated by a dash (-).

FIGURE 7.

Convergent evolution observed in DH-less PCP14 reactive B cells. (A) The HCDR3 sequence ARWDY was identical in a subset of PCP14-specific hybridomas despite (i) using different VH, DH, and JH gene segments, (ii) difference in the N- and P-nucleotides, (iii) extent of exonuclease chewing at the V–D and V–J junctions in PCP14-specific hybridomas, and (iv) the fact they were generated from different mice. *The VH gene segment used could be either VH1-4 or VH1-7. ^#The VH gene segment used could be either VH1-4, VH1-7, or VH1-26. P- and N-nucleotides are shown in red and green, respectively. Additional DH-less anti-PCP14 mAbs are shown in (B). Two anti-CRM197 mAbs that lacked DH derived amino acids are presented in (C). The codons between the conserved cysteine (codon position 104 in VH) and conserved tryptophan (codon position 118 in JH) are shown. The nucleotide sequence of the HCDR3 portion of the IgH transcript expressed in the hybridoma and the corresponding germline VH, DH, and JH gene segments is shown in bold type. P- and N-nucleotide addition(s) are indicated on top of the alignment. Identity with the corresponding nucleotide in the germline gene segment is indicated with a dash (-). Silent mutations are shown in lowercase letters. The amino acid translation is shown in single-letter code below the codons (in italics). It may be noted that some nucleotides are missing at the V–D and D–J junctions due to exonuclease activity. The frequency (Fq.) of occurrence, dose regimen (Rg.), and the mouse ID (Mo. ID.) from which the hybridoma was derived are shown on the right.