Epitope-specific human influenza antibody repertoires diversify by B cell intraclonal sequence divergence and interclonal convergence

Abstract

We generated from a single blood sample five independent human mAbs that recognized the Sa antigenic site on the head of influenza hemagglutinin and exhibited inhibitory activity against a broad panel of H1N1 strains. All five Abs used the V(H)3-7 and J(H)6 gene segments, but at least four independent clones were identified by junctional analysis. High-throughput sequence analysis of circulating B cells revealed that each of the independent clones were members of complex phylogenetic lineages that had diversified widely using a pattern of progressive diversification through somatic mutation. Unexpectedly, B cells encoding multiple diverging lineages of these clones, including many containing very few mutations in the Ab genes, persisted in the circulation. Conversely, we noted frequent instances of amino acid sequence convergence in the Ag combining sites exhibited by members of independent clones, suggesting a strong selection for optimal binding sites. We suggest that maintenance in circulation of a wide diversity of somatic variants of dominant clones may facilitate recognition of drift variant virus epitopes that occur in rapidly mutating virus Ags, such as influenza hemagglutinin. In fact, these Ab clones recognize an epitope that acquired three glycosylation sites mediating escape from previously isolated human Abs.

FIGURE 1

Therapeutic efficacy of Ab 4K8 against disease caused by the 1918 A (H1N1) virus in mice. Mice were inoculated on day 0 and treated on day 1 with the indicated antibody and dose. In each group, six mice were monitored every other day for survival (A) and weight (B). At all dose levels, the differences in survival distribution between the 4K8 and the human IgG control groups were significant by log-rank test (p = 0.001 for the high-dose level, p < 0.001 for the medium-dose level, p < 0.01 for the low-dose level).

FIGURE 2

Comparison of antibody gene junctional sequences reveals four independent clones. The IgH gene segment junctions of the five V_H3-7/J_H6 antibodies 4A10, 2O10, 4K8, 6D9, and 2K11 are shown in amino acid and DNA sequence. Mutated amino acids and nucleotides are underlined. The amino acid residues are color-coded per standard IMGT color scheme based on chemical properties (43). Briefly, aliphatic (A, I, L, V) residues are dark blue, phenylalanine light blue, sulfur (C, M) residues cyan, glycine dark green, residues with hydroxyl groups (S, T) medium green, tryptophan pink, tyrosine light green, proline yellow, acidic (D, E) residues light orange, amide (N, Q) residues dark orange, and basic (H, K, R) residues red. Kabat numbering for amino acids is used instead of IMGT numbering, and is shown at the top level; the CDR H3 margins are denoted in red. The contributions of the V_H, D, and J_H genes are shown in light grey (V_H), medium grey (D), and dark grey (J_H).

FIGURE 3

Phylogram and sequence alignment to the V_H3-7*01 germline sequence of the heavy variable chain genes of Abs 4A10 (cyan), 2O10 (orange), 4K8/6D9 (medium blue), and 2K11 (green) from hybridoma technology and pyrosequencing. Five-letter alphanumeric labels denote sequences derived from pyrosequencing; hybridoma names are italicized. The location of the CDRs (based on IMGT analysis) is shown on top; Kabat numbering is shown at the bottom. Amino acids similar to the V_H3-7*01 germline sequence (for the V_H region) or to the consensus sequence (for the D/J regions) are in light gray, dissimilar amino acids in white. Variable gene segment (V-GENE) encoded sequences are separated from the diversity and joining gene segment (D/J-GENE) encoded sequences by a dashed line. Residues within the V_H region with evidence of convergence are identified with an asterisk at the bottom. The phylogram was generated based on the protein sequences with MacVector software version 12 using neighbor joining, best tree, symmetric tie breaking, uncorrected (“p”) distance settings, and rooted to the deduced V_H3-7*01 germ line protein sequence.