Back

Engineered Fab glycosylation of a blood-brain barrier transporter single-chain variable fragment preserved functionality but did not impact anti-drug antibody formation

Hederoth, G.; Liu, J.; Godec, A.; Metzendorf, N. G.; Rethi, B.; Hultqvist, G.

2025-12-02 bioengineering
10.1101/2025.11.28.691100 bioRxiv
Show abstract

The single-chain variable fragment (scFv) of the 8D3 antibody enables blood-brain barrier transport via transferrin receptor-mediated transcytosis. However, when fused to therapeutic antibodies, scFv8D3 increased immunogenicity and triggered anti-drug antibodies (ADA). Sialylation of N-linked glycans can critically influence protein immunogenicity, potentially conferring immunosuppressive properties and extended half-life. We investigated whether introducing sialic acid residues to scFv8D3 through engineered fragment antigen-binding (Fab) glycosylation could reduce ADA formation. We generated two novel variants with distinct N-linked glycosylation motifs (NXT/S), which both displayed sialic acid residues and retained functionality, as evidenced by the preservation of transferrin receptor binding and brain uptake. However, Fab glycosylation did not significantly alter pharmacokinetics or ADA production in mice. Our findings demonstrated that introducing Fab glycosylation is feasible, without compromising biological activity. Nonetheless, this strategy alone was insufficient to mitigate immunogenicity, underscoring the need for alternative approaches to improve the long-term safety of brain-targeting therapeutics. Statement of significanceSialylation of N-linked glycans can critically influence protein immunogenicity. We introduced N-linked Fab glycosylation sites into a scFv-based therapeutic construct with the aim of reducing ADA production upon repeated administration to mice. Although the sialylated variants retained functionality, pharmacokinetics and ADA levels remained unaffected, suggesting limited immunomodulation via engineered sialylation.

Matching journals

The top 12 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.