Functional antibodies exhibit light chain coherence

Abstract

The vertebrate adaptive immune system modifies the genome of individual B cells to encode antibodies that bind particular antigens¹. In most mammals, antibodies are composed of heavy and light chains that are generated sequentially by recombination of V, D (for heavy chains), J and C gene segments. Each chain contains three complementarity-determining regions (CDR1-CDR3), which contribute to antigen specificity. Certain heavy and light chains are preferred for particular antigens^2-22. Here we consider pairs of B cells that share the same heavy chain V gene and CDRH3 amino acid sequence and were isolated from different donors, also known as public clonotypes^23,24. We show that for naive antibodies (those not yet adapted to antigens), the probability that they use the same light chain V gene is around 10%, whereas for memory (functional) antibodies, it is around 80%, even if only one cell per clonotype is used. This property of functional antibodies is a phenomenon that we call light chain coherence. We also observe this phenomenon when similar heavy chains recur within a donor. Thus, although naive antibodies seem to recur by chance, the recurrence of functional antibodies reveals surprising constraint and determinism in the processes of V(D)J recombination and immune selection. For most functional antibodies, the heavy chain determines the light chain.

Fig. 1. Functional public and private antibodies exhibit light chain coherence.

Pairs of B cells were examined if (1) they had the same heavy chain V gene name, (2) they had the same CDRH3 length, and (3) both cells were either memory (red) or naive (blue). The percentage of cell pairs using the same light chain V gene (or paralog) is shown as a function of CDRH3 amino acid identity, rounded down to the nearest 10%. a, Probability of having a common light chain V gene for public antibodies: the two cells in each pair originated from different donors. One curve is shown for each donor pair. Additional curves (grey and black (hidden below the grey in the graph)) show the light chain coherence when heavy and light chain correspondence is randomly permuted. Data are mean ± s.e.m. We tested the differences between regression curve slopes using a sum-of-squares F test (P < 0.0001, F = 20.89, d.f. numerator = 13, d.f. denominator = 140). b, Probability of having a common light chain V gene for private antibodies: the two cells in each pair are from the same donor, but from different computed clonotypes and exhibit additional evidence that they lie in different true clonotypes (Methods). One curve is shown for each donor.

Fig. 2. Transitive linking yields large coherent groups.

a, We transitively grouped clonotypes at given per cent CDRH3 amino acid identities (boxed) while requiring the same heavy chain V gene. 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). These cells use the heavy chain V gene IGHV3-9. Each dot represents a cell, and each cluster is a computed clonotype. With the exception of three cells, all computed clonotypes use the light chain gene IGKV2-30, and cells from all six donors (d1, d2, d3, d4, d5 and d6) are present. Bottom, logo plot for CDRH3 (top) and CDRL3 (bottom) amino acid sequences in this group.

Fig. 3. Public antibody properties are consistent with recombination biology.

Heavy chain junction sequences were aligned to concatenated reference sequences comprising VJ, VDJ or VDDJ, with up to two different D genes, and the most likely reference. We determined the number of bases inserted in the junction relative to this reference (counting deletions separately), as well as the number of substituted bases. a, The heavy chain junction region for a memory cell with the heavy chain junction CARDGGYGSGSYDAFDIW is shown. We found IGHD3-10 to be the most likely D gene. There are eight inserted bases in the junction and seven substitutions. The substitution rate is 7 out of 46, where the denominator (46) is the total number of matching and mismatching bases. b, For each of four types of antibodies in the data, we computed the number of inserted bases in the heavy chain junction region, relative to the concatenated VDJ (or in some cases VJ or VDDJ) reference sequence. Most of the inserted bases are insertions in non-templated region 1 or 2. The frequency is shown as a function of the number of inserted bases. See also Table 1.

Extended Data Fig. 1. Flow cytometry gating schemes for B cell subsets.

Gating strategy for B cell isolation. Panels a—d show naive cell gating, panels e—g show memory cell gating, and panels h—i show plasmablast gating. a, Hierarchical gating scheme for lymphocytes, single cells, live cells, and CD3-negative cells. b, We gated CD19+CD27± cells from CD3− cells for further analysis. Donor samples displayed noticeable differences in CD19 and CD27 expression. c, We analyzed CD19+CD27− cells for surface IgD expression and gated IgD+ cells for further analysis. d, We selected naive B cells by sorting CD19+CD27–IgD+CD24±CD38± B cells. e, For memory cell gating, we selected CD19+CD27+ cells from b for CD24+CD38+ positivity. f, We analyzed cells from e and isolated unswitched memory cells using IgD±IgM++ gating. g, We analyzed cells from e and isolated switched memory cells using IgD−CD95+ gating. h, We analyzed CD19+CD27+ cells from b and gated the IgD–CD27+ population. i, We sorted plasmablasts using CD24–CD38++ gating.

Extended Data Fig. 2. Light chain coherence is visible by sequence similarity.

Each point represents a pair of memory cells from different donors. Heavy and light chain edit distances are plotted, using the amino acids starting at the end of the leader and continuing through the last amino acid in the J segment. Points with identical coordinates are combined by showing a large point whose area is proportional to the number of such points. a, Cell pairs are displayed if the two cells in the pair have the same CDRH3 amino acid sequence. To increase readability, only one third of such pairs were selected at random for display. Of the pairs, 78% have light chain edit distance ≤ 20. This number (78%) is the fraction of cell pairs lying below the horizontal line at light chain edit distance 20, and was computed separately. It is proportional to the fraction of red below the line, if overlap is taken into account. b, [control] The same number of cell pairs were selected at random for display, without regard to CDRH3. Of the pairs, 9% have light chain edit distance ≤ 20.