Evidence for preferential Ig gene usage and differential TdT and exonuclease activities in human naïve and memory B cells

Abstract

Memory B cells and the antibodies they encode are important for protective immunity against infectious pathogens. Characterization of naïve and memory B cell antibody repertoires will elucidate the molecular basis for the generation of antibody diversity in human B cells and the optimization of antibody structures that bind microbial antigens. In this study we aimed to investigate the influence of antigenic selection on the antibody genes of the two CD27+ memory B cell subsets, comparing them with the naïve repertoire in CD27- cells. We analyzed and compared the Ig heavy chain gene transcripts in three recently defined circulating naïve and memory B cell subsets (CD19+IgD+CD27- [naïve], CD19+IgD+CD27+ [un-class-switched memory] or CD19+IgD- CD27+ [class-switched memory]) at the single cell level. We found similar biased patterns of variable, diversity and joining heavy chain gene usages in all three groups of cells. CD19+IgD+CD27+ memory B cells harbored as diverse an antibody gene repertoire as CD19+IgD-CD27+ memory B cells. Interestingly, CD19+IgD+CD27+ memory B cells possessed a lower frequency of somatic mutations, a higher incidence of exonuclease activity at the 3' end of D regions, and a lower frequency of N and P nucleotide additions at both VH-D and D-JH junctions of CDR3 regions compared to CD19+IgD-CD27+ memory B cells. These data suggest distinct functional mechanisms underlying selection of this unique subset of un-class-switched memory B cells.

FIGURE 1. A. Representative data from flow cytometric analyses demonstrating human naïve and memory B cells were isolated from adult blood samples with high purity

The numbers indicate the percentage of cells that were of naïve or memory phenotypes (IgD⁺CD27⁻, IgD⁺CD27⁺ or IgD⁻CD27⁺) as indicated by the oval gates. All cells plotted were gated to be CD19⁺. All post-sort samples contained >99% CD19⁺ B cells. B. Isotypes of antibody produced by single B cell clones were determined by Ig capture ELISA. Culture supernatants from 80 IgD⁺CD27⁻ naïve B cells, 69 IgD⁺CD27⁺ memory B cells and 80 IgD⁻CD27⁺ memory B cell clones were used for determining isotypes of antibody produced at 21 days post culture. Approximately 95% of IgD⁺CD27⁻ naïve B cells and 97% of IgD⁺CD27⁺ memory B cell clones produced IgM. The IgD⁻CD27⁺ memory B cell clones produced IgM, IgG or IgA (19%, 39%, or 36%, respectively).

FIGURE 2. Similar V_H, D and JH gene family utilization in circulating human naïve or memory B cells

Single circulating naïve or memory B cells from healthy blood donors were sorted, amplified in culture, and tested for antibody production by ELISA followed by sequence analysis of the expressed V_H gene segment. Frequencies of V_H, D and J_H gene family utilization in all three circulating naïve and memory B cells are presented in 2A, 2B and 2C, respectively. No difference was found in the distribution of V_H, D or J_H gene family utilization between the three types of B cells (p>0.5 for all comparisons, Chi-square tests).

FIGURE 3. Similar V_H, D and J_H gene segment utilization in circulating human naïve or memory B cells

Frequencies of V_H, D and J_H gene segments utilized in three circulating naïve and memory B cells are shown in 3A, 3B and 3C. No significant difference between cell types was found for any of the V_H, D or J_H gene segment utilizations (p>0.2). Numbers of V_H, D or J_H gene segments identified (used at least once within a group of cells) in all three circulating naïve and memory B cells are presented in 3D, 3E and 3F.

FIGURE 4. Shorter variable heavy chain CDR3 regions in IgD⁺CD27⁺ memory B cells

A. The CDR3 lengths for three circulating naïve and memory B cells are shown with medians and 25^th and 75^th percentiles (interquartile range) indicated as bars. The CDR3 lengths for IgD⁺CD27⁺ memory B cells were shorter than those of IgD⁺CD27⁻ naïve B cells, p=0.022. B. Total N and P nucleotides in both the V_H-D and D-J_H junctions of naïve and memory B cells are presented with medians and 25^th and 75^th percentiles indicated as bars. Total N and P nucleotides in IgD⁺CD27⁺ memory B cells were significantly less than those in IgD⁻CD27⁺ memory B cells, p = 0.0425.

FIGURE 5. A. Lower frequency of somatic mutations in IgD⁺CD27⁺ memory B cells

The number of mutations identified in variable heavy chain genes for circulating naïve or memory B cells is shown. Medians and 25^th and 75^th percentiles interquartile values are indicated. Naïve B cells possessed fewer mutations (median 0 bp) than did IgD⁺CD27⁺ (median 10 bp) or IgD⁻CD27⁺ memory B cells (median 18 bp) (p<0.0001). IgD⁺CD27⁺ memory B cells possessed fewer mutations than IgD⁻CD27⁺ memory B cells (p<0.0001). B. Distributions of total somatic mutations and replacement mutations were similar in IgD⁺CD27⁺ and IgD⁻CD27⁺ memory B cells. Mutations in both types of memory B cells were distributed across the variable region of heavy chain genes with an increased frequency of replacement mutations (R), especially in the CDR1, CDR2 and FR3 regions. Higher replacement mutation (R) /silent mutation (S) ratios (>3) are detected in the CDR regions of the two memory B cell groups than in the FR regions.

FIGURE 6. Frequency of mutations within hotspot motifs and the distribution of hotspot mutations along the variable regions were similar between IgD⁺CD27⁺ and IgD⁻CD27⁺ memory B cells

A. Numbers of total mutations and hotspot-targeted replacement or silent mutations within CDRs and FRs are shown. B. Frequency of hotspot-targeted mutations in the CDRs or FRs for the three subsets of circulating naïve and memory B cells are similar.