Clonal structure, stability and dynamics of human memory B cells and circulating plasmablasts

Abstract

Memory B cells persist for a lifetime and rapidly differentiate into antibody-producing plasmablasts and plasma cells upon antigen re-encounter. The clonal relationship and evolution of memory B cells and circulating plasmablasts is not well understood. Using single-cell sequencing combined with isolation of specific antibodies, we found that in two healthy donors, the memory B cell repertoire was dominated by large IgM, IgA and IgG2 clonal families, whereas IgG1 families, including those specific for recall antigens, were of small size. Analysis of multiyear samples demonstrated stability of memory B cell clonal families and revealed that a large fraction of recently generated plasmablasts was derived from long-term memory B cell families and was found recurrently. Collectively, this study provides a systematic description of the structure, stability and dynamics of the human memory B cell pool and suggests that memory B cells may be active at any time point in the generation of plasmablasts.

Functional antibodies exhibit light chain coherence

“…This finding implies that for memory cells, which bear functional antibodies and are typically the products of thymic and peripheral selection, heavy chain coherence implies light chain coherence. We further tested light chain coherence for memory B cells using independent datasets, finding 93% coherence on a heterogeneous compendium of older data (data generated by 10x Genomics for internal development and distributed with enclone 30 , as well as several other datasets related to multiple sclerosis 31 , COVID-19 32,33 , Kawasaki disease 34 and LIBRA-seq 35 ) and 79% coherence on a recent dataset 36 (Methods, 'Additional data'). We redistribute these data as part of our paper and provide details about each dataset in Supplementary Table 1.…”

“…Even so, large coherent groups can be formed. Figure 2b shows a transitive group of clonotypes computed using 90% percent identity, using both the data of this work and that of Phad et al 36 and sharing the heavy chain V gene IGHV3-9. The light chain coherence for this group is 99.7% (726 out of 728 cells), and all but four cells use the light chain genes IGKV2-30 and IGKJ2.…”

“…Namely, we identify and use for simulator training truly naive junction sequences (with no detectable SHM in the other CDR or FR regions), and we account for central and peripheral selection using the post-selection model 41 . 36 . These cells use the heavy chain V gene IGHV3-9.…”

“…We used several other datasets. These included single-cell data for 247,516 cells from Phad et al 36 , which were approximately evenly distributed between two donors. A second combined collection ('older data') included data generated by 10x Genomics for internal development, and distributed with enclone, as well as several other datasets related to multiple sclerosis 31 , COVID-19 32,33 , Kawasaki disease 34 and LIBRAseq 35 .…”

“…The graph shows the relationship between light chain coherence and the number of cells appearing in each group. b, Top, at 90% identity, the transitive group containing a cell with the CDRH3 sequence CIKDILPGGADSW is shown, using the data from this Article and from Phad et al (2022)36 . These cells use the heavy chain V gene IGHV3-9.…”