An experimentally validated structure-based computational framework for humanisation of anti-orthopoxvirus antibodies

Area of Science:

• Immunology
• Virology
• Structural Biology

Background:

• Orthopoxviruses, including mpox virus (MPXV), represent a significant global health concern.
• Murine antibodies are limited by anti-mouse responses, and sequence-based humanization can reduce efficacy.

Purpose of the Study:

• To develop and validate a structure-guided computational framework for humanizing murine anti-orthopoxvirus antibodies.
• To assess the efficacy of humanized antibodies against orthopoxvirus infections.

Main Methods:

• A novel framework prioritizing 3D structural congruence over sequence identity was developed.
• The framework integrated Foldseek structural alignment and interface-residue constraints.
• Two murine antibodies (7D11, A27D7) were humanized (POX1.1, POX2.1) and extensively validated in vitro and in vivo.

Main Results:

• Humanized variants (POX1.1, POX2.1) maintained high conformational similarity to parental antibodies.
• Both variants demonstrated full epitope specificity and natural humanness.
• POX1.1 enhanced neutralization of vaccinia virus (VACV) and MPXV; POX2.1 retained broad cross-reactivity.
• Monotherapies provided significant protection in lethal VACV and MPXV models.
• Combination therapy achieved 100% survival in VACV models and suppressed MPXV dissemination.

Conclusions:

• The structure-centric workflow enables efficient humanization of murine anti-orthopoxvirus antibodies.
• This validated framework supports the development of countermeasures against orthopoxvirus pandemics.