LIBRA HuEasy Monoclonal antibody (mab) humanization service (fully humanized ab)
LIBRA HuEasy™ Technology for Monoclonal antibody (mab) humanization service
With extensive in-house data and know-how in protein engineering as well as antibody research and development, GeneMedi developed a unique big-data-driven mature antibody humanization technology, LIBRA HuEasy™. GM’s LIBRA HuEasy™ technology combines artificial intelligence (AI) and prior wet data to form a self-evolution antibody humanization Algorithm. We are committed to offering an Easier, Faster, higher efficient humanization service and solution to balance high affinity with low immunogenicity for your parental antibodies.
About LIBRA HuEasy™
LIBRA HuEasy™ is a Machine learning-based antibody humanization technology to balance high affinity with low immunogenicity. Easier, Faster, higher accuracy rate with less wet-validation procedure.
About LIBRA ™
LIBRA™-Antibody Engineering Platform is GeneMedi’s innovative antibody engineering platform. LIBRA is developed with the characteristics: Light, Intelligent, balanced, Rational, and Algorithm-driven. LIBRA,(♎) is also Latin, meaning “Balance” in astrology.
Definition: What is monoclonal antibody humanization (mab humanization)?
Monoclonal antibody humanization is an important part of the development of antibody drugs, it has been widely used for non-human (such as mouse, rat, rabbit, and chicken) antibody engineering. Compared with the parental antibody, humanized antibody is almost entirely composed of human sequences and can substantially reduce potentially immunogenicity when administered to human patients.
Humanization of antibody from species: mouse/rat/rabbit/chicken/camel/llama and shark antibody humanization
At present, for mice, rats, and rabbits, the main immune globulin isotype used for Humanization is IgG, while for chicken is IgY. Besides the whole IgG, nanobody, mainly engineered from camel, llama, and shark, which is also called variable domain of a heavy-chain antibody (VHH) or single domain antibody(sdAb), due to its high thermal stability, high affinity as well as high specificity with in vivo molecules, is getting more and more attention in Humanization too.
Species | Isotypes |
---|---|
mouse | IgG1, IgG2a/IgG2c, IgG2b, IgG3 |
rat | IgG1, IgG2a, IgG2b, IgG2c |
human | IgG1, IgG2, IgG3, IgG4 |
rabbit | IgG |
chicken | IgY |
camel | hcIgG |
shark | IgNAR |
Table 1. Different animal species have different sets of antibody isotypes for genetically engineering
How to get humanized monoclonal antibody?--Procedure & Protocol of antibody humanization
The main technical process is modifying antibody sequence from non-human species to increase the similarity to antibodies produced naturally in humans.
Specifically, first blast the parental antibody variable domain sequence with the human antibody Germline database to find a suitable human template, then graft the complementarity-determining region (CDR) segment of the parental antibody to the corresponding CDR area of a human antibody, antibodies obtained in this way are also called reshaping antibodies.
Fig 1. Antibody humanization workflow
Humanization of antibody formats: whole antibody/vhh(nanobody, sdAb)
The homology of nanobody and human antibody heavy chain variable domain (VH) can reach 80~90%. However, there is a small but significant difference between camel-derived VHH and human VH, the CDR1 and CDR3 of VHH are longer than human VH. In addition, in the framework (FR)-2, several conserved amino acid residues are hydrophobic in humans while the corresponding conserved residues in VHH are hydrophilic. Therefore, besides backmutation, in order to get better stability, solubility, and biological activity, humanized nanobodies also need to modify some FR residues.
humanized vs fully human antibody vs chimeric mouse/human antibody
Fig 2. Schematic overview of antibody humanization from the murine antibody (blue domains) to fully human antibody (green domains)
Chimeric antibody contains variable domains from its original species (like a mouse), and constant regions from another species (such as a human being) while retaining the specificity. Compared with the mouse monoclonal antibody, a human-mouse chimeric antibody can efficiently reduce the production of a human anti-murine antibody (HAMA) in clinical trials. However, a sizeable proportion of patients still have HAMA responses, so far chimeric antibody is not ideal for clinical treatment due to the immunological risk.
For a fully humanized antibody, its genetic source is 100% human-derived, the main ways include three technical routes, one is antibody library screening, creating a large library of variable domains of antibodies (in the form of so-called single-chain Fv fragment, scFv) on the surface of phage and therewith selecting and amplifying in E.coli. The 2nd route is the transgenic mouse, which refers to transferring all human antibody-encoding genes into its genome, which can express human antibodies. The 3rd route is available as an endogenous antibody from the human body. These agents are often harvested by identifying pathogen-specific B cells of patients who have recently recovered from an infection. For instance, Eli Lilly’s monoclonal antibody bamlanivimab (also known as LY-CoV555, aka LY3819253) was originally derived from the blood of one of the first U.S. patients who recovered from COVID-19.
Compared with chimeric antibodies, a humanized antibody has lower immunogenicity, wider clinical applications, and higher difficulty in preparation, while a fully humanized antibody has got to a marginal effect in reducing immunogenicity, and also a higher preparation cost.
Mouse antibody | Chimeric antibody | Humanized antibody | Fully Human antibody | |
---|---|---|---|---|
Human Immune Response | high, unavoidable | relatively high, unavoidable | low, avoidable | rare, avoidable |
clinical application | restricted, low application range | restricted | widely used | widely used |
production difficulty | low | relatively high | relatively high | high |
structure | fully mouse derived | fusing mouse derived variable regions to human constant regions | grafting mouse CDR into human variable framework regions | fully human derived |