Antibody Affinity Shapes the Choice between Memory and Germinal Center B Cell Fates

Abstract

Immunological memory is required for protection against repeated infections and is the basis of all effective vaccines. Antibodies produced by memory B cells play an essential role in many of these responses. We have combined lineage tracing with antibody cloning from single B cells to examine the role of affinity in B cell selection into germinal centers (GCs) and the memory B cell compartment in mice immunized with an HIV-1 antigen. We find that contemporaneously developing memory and GC B cells differ in their affinity for antigen throughout the immune response. Whereas GC cells and their precursors are enriched in antigen binding, memory B cells are not. Thus, the polyclonal memory B cell compartment is composed of B cells that were activated during the immune response but whose antigen binding affinity failed to support further clonal expansion in the GC.

Keywords: antibody affinity; memory B cell.

Figure 1.. Antigen binding GC and memory B cells

(A) Schematic representation of the experiment: S1pr2^CreERT2/+ R26^ZSGreen/+ mice immunized with NP-OVA (footpad injection) at day 0, tamoxifen gavage on day 5, analysis of ZSGreen⁻ follicular B cells and ZSGreen⁺ GC cells on day 14 and ZSGreen⁺ memory B cells (MBC) on day 42. (B-C) Representative flow cytometry profiles and graph showing the percentage of Igλ⁺ cells among the ZSGreen⁻ follicular B cells on day 14 (left panel), ZSGreen⁺ GC cells on day 14 (middle panel) and ZSGreen⁺ memory B cells on day 42 (right panel) after NP-OVA immunization (3 independent experiments, n=7-8, 1way Anova, **** p ≤ 0.0001). (D) Schematic representation of the experiment: S1pr2^CreERT2/+ R26^tdT/+ mice immunized with TM4-Core (footpad injection) on day 0, tamoxifen gavage on day 5, analysis of tdT⁻ follicular B cells and tdT⁺ GC cells on day 14 and tdT⁺ memory B cells on day 42. (E-F) Representative flow cytometry profiles and graph showing the percentage of TM4-Core⁺ binding cells among the tdT⁻ follicular B cells on day 14 (left panels), the tdT⁺ GC B cells on day 14 (middle panel) and the tdT⁺ memory population on day 42 (right panels) after TM4-Core immunization (3 independent experiments, n=7-8, 1way Anova, **** p ≤ 0.0001).

Figure 2.. Antigen binding in contemporaneous GC and memory compartments

(A) Schematic representation of the experiment: S1pr2^CreERT2/+ R26^ZSGreen/+ mice immunized on day 0 with TM4-Core (footpad injection), tamoxifen gavage on days 9, or 15 or 21. Fate mapped ZSGreen⁺ GC and ZSGreen⁺ memory B cells analyzed 3 days after the tamoxifen injection. (B) Number of ZSGreen⁺ GC (left panel) and ZSGreen⁺ memory B cells (right panel) in control or TM4-Core immunized mice on day 12, which was 3 days after tamoxifen administration (3 independent experiments, n=6-10, Mann-Whitney, *** p ≤ 0.001). (C-D) Representative flow cytometry profiles and graph showing the percentage of TM4-Core⁺ cells among the ZSGreen⁻ Fo, and ZSGreen⁺ GC and memory B cell populations after TM4-Core immunization as in (A) (2 independent experiments, n=7, 2 way Anova, **** p ≤ 0.0001). (E) Schematic representation of the experiment: Vav^Tg col1a^mCherry/+ mice immunized on day 0 with TM4-Core or NP-OVA (footpad injection), doxycycline i.p injection on day 9. Fate mapped mCherry^low GC and mCherry^low switch memory B cells analyzed 3 days after doxycycline injection. Representative flow cytometry histogram: GC cells are mCherry^hi before doxycycline injection (day 9) and mCherry^low after (day 12); Fo B cells are mCherry^hi after doxycycline (day 12). (F) Representative flow cytometry histogram showing the percentage of mCherry^low switch memory B cells after doxycycline alone (left panel) immunization alone (middle panel) and both (right panel). (G-H) Representative flow cytometry profiles and graphs showing the percentage of TM4-Core⁺ and Igλ⁺ cells among the mCherry^hi Fo, and mCherry^low GC and memory B cell populations after TM4-Core or NP-OVA immunization as in (E) (3 independent experiments, n=6, 1 way Anova, **P ≤ 0.01, ***P ≤ 0.001, **** p ≤ 0.0001).

Figure 3.. Mutations in contemporaneous GC and memory B cells

(A) Graph shows number of somatic mutations (nucleotides, IgH + IgL) in the antibodies obtained from GC (gray) and memory B cells (black) purified after immunization and tamoxifen gavage as in Figure 2A (Groups of 15-20 mice, 3 independent experiments, n=112-221 antibody genes sequenced in each group, 2 way Anova, **** p ≤ 0.0001). (B-D) tdT⁺ GC and memory B cell were sorted from the same lymph node on day 12 after TM4-Core immunization and tamoxifen gavage on day 9. (B) Pie charts depicting the distribution of antibody sequences (IgH + IgK) for GC (left) and memory B cells (right) from 3 individual lymph nodes. The number in the inner circle indicates the number of sequences analyzed. White indicates sequences isolated only once, and grey or colored pie slices are proportional to the number of clonally related sequences. Colored slices indicated shared sequences between GC and memory B cells. (C) Graph shows the number of mutations (nucleotides: IgH + IgL) of the shared sequences between GC and memory B cells for each lymph node. (D) Trees shows phylogenetic relationships between IgH and IgK sequences from GC (grey) and memory B cells (colors) from the clones identified in (B).

Figure 4.. Affinity in contemporaneous GC and memory B cells

(A) Graphs show biolayer interferometry traces for biosensor chips loaded with TM4-Core immersed into solutions containing Fabs. 3BNC60 Fab was used as a positive control (red curve) and 38 Fabs cloned obtained from random naïve B cells as negative controls (gray curves). (B) Antibodies depicted in (Fig. 3A) were produced as Fabs and their affinity measured as in (A). Curves in gray represent Fabs with affinities similar to antibodies obtained from naive B cells. Curves are red for GC cells (left) and green for memory B cells (right) when affinities were higher than naive B cells. Numbers show the total red, green or gray antibodies in each graph. (C) Graph shows biolayer interferometry traces for GC Fabs measured as in (C) but grouped on the basis of somatic mutations. Top: 0 to 5 mutations; middle: 6 to 12 mutations; bottom: more than 13 mutations. (D) Bar graph shows the percentage of GC B cell derived Fabs with a measurable affinity grouped by number of mutations indicated on the X axis. (E) Scatterplot shows the correlation between mutation and K_D (r²=0.079 p=0.35). (F) Graphs show biolayer interferometry traces for biosensor chips loaded with TM4-Core immersed in solutions containing GC or memory B cells Fabs isolated on day 12 after immunization. Curves in gray represent Fabs with affinities similar to control naive B cells and ED38 Fabs. Curves in red for GC cells (top) and green for memory B cells (bottom) with binding higher than controls. Numbers show the total red, green or gray Fabs in each graph. (G) Graph shows the proportion of GC (red) and memory B cells (green) showing binding in (F) (H) Graphs show biolayer interferometry traces for biosensor chips loaded with individual Fabs that had no measurable affinity in monovalent binding experiments (F) immersed in solutions containing TM4-Core. Curves in gray represent Fabs with affinities similar to control naive B cells and ED38 Fabs. Curves in blue for GC cells (top) and orange for memory B cells (bottom) with binding higher than controls. Numbers show the total blue, orange and gray Fabs in each graph. (I) Graph show the proportion of Fabs showing monovalent (GC: red; memory B cells: green) or polyvalent binding (GC: blue; memory B cells: orange). (J) Graphs show biolayer interferometry traces for biosensor chips loaded with individual Fabs that had no measurable affinity in monovalent and polyvalent configurations (F and H) immersed in solutions containing tetrameric TM4-Core. Curves in gray represent Fabs with affinities similar to ED38 control Fab. Curves in light green for GC cells (top) and purple for memory B cells (bottom) with binding higher than ED38 control. Numbers show the total light green, purple and gray Fabs in each graph. (K) Graph show the proportion of Fabs binding in (F) GC: red; memory B cells: green; (H) GC: blue; memory B cells: orange or (J) GC: light green; memory B cells: purple.

Figure 5.. GC recall memory B cells

(A) Schematic representation of the experiment: S1pr2^CreERT2/+ R26^ZSGreen/+ mice were immunized on day 0 with HIV-1 TM4-Core. Tamoxifen gavage was administered on days 4 and 8. Left footpad immunization with NP-OVA or HIV-1 TM4-Core on day 30. Analysis of the left draining lymph node on day 36. (B-D) Representative flow cytometry profiles and graph showing the percentage of ZSGreen⁺ cells among the GC cells after a second immunization with NP-OVA or TM4-Core (2 independent experiments, n=8). (C) Graph shows the number of ZSGreen⁺ GC cells 6 days after left footpad immunization with NP-OVA or HIV-1 TM4-Core (2 independent experiments, n=7-8, Mann-Whitney, *p ≤ 0.05). (E-F) Representative flow cytometry profiles and graph showing the percentage of TM4-Core⁺ cells among the ZSGreen⁺ recall GC cells and the ZSGreen⁻ GC cells that were of naïve origin in GCs (1° GC) after TM4-Core immunization (3 independent experiments, n=11, paired t test, ** p ≤ 0.01). (G-J) recall GC cells analyzed from the contralateral lymph node 14 days after boosting with HIV-1 DMRS-Core. (G) Representative flow cytometry profiles showing the binding of HIV-1 DMRS-Core by recall GC cells from lymph nodes of 3 mice. (H) Pie charts depicting the distribution of antibody sequences from the lymph nodes in G. The number in the inner circle indicates the number of sequences (IgH+IgK) analyzed. White indicates sequences isolated only once, and colored pie slices are proportional to the number of clonally related sequences. (I-J) Graph shows the HIV-1 TM4-Core mean of fluorescence (I) and the number of somatic mutations (nucleotides, IgH + IgL) (J) of the recall GC cells from the clones describe in (H).

Figure 6.. Antigen binding after multiple immunizations

(A) Schematic representation of the experiment. S1pr2^CreERT2/+ R26^tdT/+ mice (n=10-15) received 1 immunization with HIV-1 TM4-Core (footpad injection) on day 0, tamoxifen gavage on day 5, and then weekly tamoxifen injection until day 42. Alternatively, S1pr2^CreERT2/+ R26^ZSGreen/+ mice (n=10-15) received 4 immunizations on days 0, 20, 40 and 60, tamoxifen gavage on day 65 and then weekly tamoxifen injection until day 102. (B-C) Representative flow cytometry profiles and graph showing the percentage of tdT⁺ or ZSGreen⁺ memory B cells after one (left panel) or four immunizations (right panel). Each point represents a pool of 10-15 mice (3 independent experiments, 2 way Anova, *p ≤ 0.05). (D) Graph shows number of somatic mutations (nucleotides, IgH + IgL) in the antibodies obtained from GC antibodies after 1 or 4 immunizations (3 independent experiments, n=59-221 antibodies, Mann-Whitney test, **** p ≤ 0.0001). (E) Graph shows number of somatic mutations (nucleotides, IgH + IgL) in the antibodies obtained from memory B cells after 1 or 4 immunizations (3 independent experiments, n=216-223 antibodies, Kruskal-Wallis, p=0.46). (F-G) Graphs show biolayer interferometry traces for 74 of the Fabs shown in (E). (F) Monovalent binding analysis. (G) Avidity at increased valency. Curves in gray represent Fabs with affinities similar to antibodies obtained from naive B cells. Curves in green or orange when affinities were higher than naive B cells. The numbers in green, orange and gray on the left corner of each graph represents the number Fabs tested.

Figure 7.. Affinity of GC and memory B cell precursors

(A-D) Graphs show biolayer interferometry traces for monovalent (A-B) or polyvalent (C-D) binding by mutated Fabs (left), and their corresponding UCAs monovalent binding (middle). The right panel shows binding under conditions of increased valency by the UCAs that didn’t show monovalent binding. (A) 17 GC Fabs that showed monovalent binding. (B) 1 memory B cell Fab that showed monovalent binding and 23 Fabs with no measurable monovalent and polyvalent affinity. (C) 11 GC cells Fabs that only showed binding under conditions of increased valency. (D) 7 memory B cell Fabs that only showed binding under conditions of increased valency.