TLR7- and TLR9-responsive human B cells share phenotypic and genetic characteristics

Abstract

B cells activated by nucleic acid-sensing TLR7 and TLR9 proliferate and secrete immune globulins. Memory B cells are presumably more responsive due to higher TLR expression levels, but selectivity and differential outcomes remain largely unknown. In this study, peripheral blood human B cells were stimulated by TLR7 or TLR9 ligands, with or without IFN-α, and compared with activators CD40L plus IL-21, to identify differentially responsive cell populations, defined phenotypically and by BCR characteristics. Whereas all activators induced differentiation and Ab secretion, TLR stimulation expanded IgM(+) memory and plasma cell lineage committed populations, and favored secretion of IgM, unlike CD40L/IL-21, which drove IgM and IgG more evenly. Patterns of proliferation similarly differed, with CD40L/IL-21 inducing proliferation of most memory and naive B cells, in contrast with TLRs that induced robust proliferation in a subset of these cells. On deep sequencing of the IgH locus, TLR-responsive B cells shared patterns of IgHV and IgHJ usage, clustering apart from CD40L/IL-21 and control conditions. TLR activators, but not CD40L/IL-21, similarly promoted increased sharing of CDR3 sequences. TLR-responsive B cells were characterized by more somatic hypermutation, shorter CDR3 segments, and less negative charges. TLR activation also induced long positively charged CDR3 segments, suggestive of autoreactive Abs. Testing this, we found culture supernatants from TLR-stimulated B cells to bind HEp-2 cells, whereas those from CD40L/IL-21-stimulated cells did not. Human B cells possess selective sensitivity to TLR stimulation, with distinctive phenotypic and genetic signatures.

Figure 1. TLR stimulation of human B cells expands IgM+ memory cells and plasma cell lineage committed cells

A: Representative day 7 cultured cells characterized by flow cytometry, gated on live singlet cells. B: Within the CD27⁺ memory gate, TLR stimulation expanded the proportion of IgM+ cells relative to the mean of 49% for unstimulated cells. C: Within the plasma cell lineage committed CD27^hi gate, TLR stimulation expanded IgM+ cells, unlike CD40L/IL-21. B–C: bars represent mean, error bars represent SEM. † = p < 0.1; *** = p < 0.001; CS = class-switched, defined as IgM-; PC = plasma cell. A: representative figures from one donor out of four, with experiments performed in duplicate; B–C: summary of four independent donors with two replicates each.

Figure 2. CD40L/IL-21 promotes proliferation of most B cells while TLR stimulation is selective

A: Representative day 7 cultured cells characterized by flow cytometry, gated on live singlet cells. B: The proportion of live cells did not differ among stimuli at day 7. B: bars represent mean, error bars represent SEM. * = p < 0.05; ** = p < 0.01; *** = p < 0.001; PC = plasma cell. A: representative figures from one donor out of two, with experiments performed in duplicate; B: summary of four independent donors with two replicates each.

Figure 3. TLRs skews antibody secretion towards IgM

A: TLR9 and TLR/IFN stimulated cells secreted high levels of IgM. B: Stimulated B cells also secreted IgG, with CD40L/IL-21 inducing more robust secretion. C: TLR stimulation skewed responding B cells towards secreting higher levels of IgM relative to IgG. A–B: bars represent mean, error bars represent SEM. C: bar represents median, error bars represent IQR. ns = p > 0.1; * = p < 0.05; ** = p < 0.01; *** = p < 0.001; A–C: summary of four independent donors with two replicates each.

Figure 4. Deep Sequencing of the IgH locus reveals clustering of TLR stimulated cells based on V and J usage

A: Unbiased clustering of IgHV and IgHJ pairing frequencies of the mutated sequence subset, showing grouping of TLR9 and TLR/IFN (in black) stimulated samples from four donors in contrast to other cultures (dendrogram shading added for ease of viewing). To control for inter-donor differences and nonspecific changes due to culture, data were normalized based on unstimulated samples prior to clustering. B: Changes in frequency of IgHV genes of specific interest (see results section for description). A: Algorithm: complete clustering based on Euclidean distances for IgHV-IgHJ pairings, filtered on pairings with at least 1% change in frequency between any two samples. B: bars represent mean, error bars represent SEM. Columns appear in order of legend. † = p < 0.1; * = p < 0.05; ** = p < 0.01; A–B: data represent four independent donors, split across two independent sequencing runs.

Figure 5. Deep sequencing of the IgH locus shows convergence of BCRs following TLR stimulation

TLR9 and TLR/IFN activated B cells of each donor demonstrated increased CDR3 amino acid sequence sharing following stimulation, while two similarly showed convergence following TLR7 stimulation (sequence sharing as indicated.) To control for differing sequencing depth, repertoire subsets of equal size were examined pair-wise for all samples from each donor, counting overlapping sequences. These counts were then clustered based on complete clustering of Euclidean distances, with dendrogram shading added for ease of viewing. The greatest extent of overlap represents 0.5%, 3.9%, 3.3%, and 1.75% sharing of sequences for donors 1 through 4. Data represent four independent donors, split across two independent sequencing runs, with subsampling performed 5 times.

Figure 6. TLR stimulation expands mutated B cells

The proportion of somatically hypermutated sequences (2+ IgHV mutations) increased with TLR stimulation relative to mean proportion of 25% for unstimulated cells. CD40L/IL-21 stimulation did not change the proportion of mutated sequences. Plot indicates data normalized intra-donor to unstimulated cultures. Bars represent mean and error bars represent SEM. † = p < 0.1, * = p < 0.05, ** = p < 0.01. Data represent four independent donors, split across two independent sequencing runs.

Figure 7. TLR stimulation expands cells bearing shorter CDR3 segments

A: TLR stimulation selectively expanded B cells with shorter CDR3 segments, particularly among mutated B cells, relative to mean length of 46 bases for unstimulated cells. B: TLR stimulation shifts the frequency distribution of CDR3 lengths towards shorter CDR3s. A: Bars represent mean, error bars represent SEM. B: bars represent sum of the proportion of unique sequences with a CDR3 length in the indicated length range. Columns appear in order of legend. * = p < 0.05. Data represent four independent donors, split across two independent sequencing runs.

Figure 8. TLR stimulation expands cells bearing less negatively charged CDR3 segments

TLR9 and TLR/IFN stimulation induced less negatively charged CDR3 segments in both non-mutated (A) and mutated (B) sequence subsets relative to mean charges of -0.55 and -0.47 for unstimulated cells. Bars represent mean, error bars represent SEM. † = p < 0.1, * = p < 0.05, ** = p < 0.01. Data represent four independent donors, split across two independent sequencing runs.

Figure 9. TLR stimulation induced autoreactive antibodies

A: TLR activation induced a mean charge increase that was positively correlated with CDR3 length. Plots indicate data normalized intra-donor to unstimulated cultures, and CDR3 length range represents modal 90% of lengths. Grey line represents 95% confidence interval tracing for linear model and dotted black line as reference for no change. r refers to Pearson correlation coefficient. B: Representative images obtained from probing culture supernatants for reactivity against HEp2 cells, showing induction of autoreactive antibodies following TLR stimulation. Summary of antinuclear staining scores for all donors using either anti-light chain (C) or anti-IgM isotype (D) detection reagents. Supernatant reactivity was graded on a scale of 0 (negative) or 1+ (weak staining) through 4+ (intense staining). * = p < 0.05; ** = p < 0.01; *** = p < 0.001. A: Data represent four independent donors, split across two independent sequencing runs. B: representative images from one donor. C–D: summary of four independent donors, each with two slides stained.