B cell imprinting in children impairs antibodies to the haemagglutinin stalk
Summary
This Nature paper links B cell receptor sequence to antigen specificity (using LIBRA‑seq, single‑cell RNA/V(D)J sequencing and monoclonal antibody characterisation) to ask how childhood influenza exposures shape long‑term B cell memory. The team studied young children with sequential primary exposures to H3N2 and H1N1 (in both orders), compared them with children after single H1 infections and with adults. Children mount predominantly de novo primary responses with lower class‑switching and fewer somatic hypermutations than adults, but 4–6% of memory B cells after consecutive heterosubtypic infections are H1/H3 cross‑reactive and target the conserved central haemagglutinin (HA) stalk.
Crucially, an H3→H1 imprinting pattern produced cross‑group anti‑stalk antibodies in children that were substantially narrower and less potent than those from adults or from H1→H3 children. Structural and biophysical analyses showed that a single residue at HA2 position 46 (Asp versus Asn; D46N) explains most of the altered binding: the change of a single atomic group (carboxylate vs amide) shifts electrostatic contacts and reduces affinity to many pre‑2009 H1 strains. Cryo‑EM, mutagenesis and MD simulations support Asp46 as a key contact for imprinted stalk antibodies. Finally, limited data from infant vaccination suggest simultaneous first exposure to H1 and H3 (vaccine) may avoid the harmful heterosubtypic imprinting seen after sequential infection.
Key Points
- LIBRA‑seq + 10x single‑cell sequencing mapped >3,000 HA‑specific B cells and validated binding with recombinant monoclonal antibodies.
- Children’s primary HA responses are de novo, show less class‑switching and fewer somatic hypermutations than adult responses.
- After sequential heterosubtypic infection, ~4–6% of children’s memory B cells are cross‑group (H1/H3) and concentrate on the central HA stalk epitope.
- H3‑first imprinting (H3→H1) yields cross‑group stalk antibodies with markedly reduced neutralisation breadth and potency compared with adult or H1‑first responses.
- A single amino‑acid difference at HA2 position 46 (D46 vs N46) accounts for most of the loss of binding to older (pre‑pdm) H1 strains — a tiny chemical change with large functional impact.
- Cryo‑EM structures and MD simulations show Asp46 makes key electrostatic and hydrogen‑bond contacts that are weakened by the D46N substitution.
- Preliminary infant vaccine data suggest simultaneous H1+H3 priming may avoid the heterosubtypic imprinting that narrows stalk antibody breadth.
Context and relevance
This study is highly relevant to anyone working on influenza immunity or universal‑vaccine strategies. It shows how early antigenic history — even a single residue difference between strains — can shape which B cell clones are recalled later and whether those clones produce broadly protective stalk antibodies. The finding that heterosubtypic imprinting can actively reduce breadth and potency of anti‑stalk antibodies is important because the HA stalk is a major target for broadly protective vaccines. The work combines cohort immunology, single‑cell genomics, monoclonal antibody functional assays and high‑resolution structural biology, so it connects clinical observations to precise molecular mechanisms.
Author style
Punchy: the paper isn’t just another imprinting study — it drills down to the atomic level and shows a single‑atom difference (carboxylate vs amide) can flip the protective capacity of imprinted stalk antibodies. If you care about vaccine design or childhood priming policies, the details matter here.
Why should I read this?
Quick and blunt: if you want to understand why some kids (and by extension some birth cohorts) may be left with narrower anti‑stalk protection — and what that means for universal‑vaccine approaches — read this. It’s the kind of paper that explains a real problem (imprinting that weakens broadly neutralising stalk responses) and gives clear molecular evidence you can act on.