Bispecific antibodies (BsAbs) are therapeutic proteins engineered to recognize and bind to two different targets simultaneously. Unlike natural antibodies, which are restricted to a single target, BsAbs possess two distinct antigen-binding sites on the same molecule. This dual-targeting capability allows them to act as a molecular bridge, often linking a disease-related cell, such as a tumor cell, and an immune effector cell, like a T-cell. By physically bringing the immune cell into close proximity with the target cell, the BsAb enhances the body’s immune response, facilitating the destruction of the unwanted cell. This mechanism increases the efficacy and precision of antibody-based treatments, particularly in oncology.
The Engineering Challenge of Dual Specificity
Manufacturing a bispecific antibody is complicated by the inherent structure of a conventional antibody. A standard immunoglobulin G (IgG) molecule is composed of two identical heavy chains and two identical light chains, which naturally assemble into a symmetrical “Y” shape. Producing a BsAb requires combining two different heavy chains and two different light chains, each forming a unique antigen-binding arm.
When all four unique chains are co-expressed, the cell machinery allows them to associate randomly, leading to a complex mixture of products. This combinatorial chaos results in up to ten different molecular combinations. Only one combination is the desired functional bispecific molecule; the others are non-functional homodimers or half-antibodies, which significantly reduce the yield.
This challenge involves two primary mispairing issues: heavy chain mispairing and light chain mispairing. The heavy chains must correctly form a heterodimer (two different chains), and the light chain must associate exclusively with its correct heavy chain partner to form the functional antigen-binding fragment (Fab) arm. Historically, this structural hurdle made the large-scale production of pure, functional BsAbs nearly impossible.
Chemical Conjugation: Early Approaches
Early attempts to create bispecific antibodies bypassed cellular assembly by linking two pre-formed antibodies using chemical methods. This involved creating and purifying two separate monoclonal antibodies, then connecting them in vitro using chemical cross-linkers. A common strategy used two F(ab’)2 fragments—the two-armed binding portions of an antibody lacking the Fc region.
To prepare the fragments for linkage, the inter-heavy chain disulfide bonds were chemically reduced, exposing reactive thiol groups. The two different fragments, each with a single antigen-binding site, were then mixed and chemically linked using agents like o-phenylenedimaleimide (o-PDM) or reagents that regenerate new disulfide bonds.
While these methods produced bispecific molecules, they had significant limitations. The reactions were often inefficient, resulting in low yields and a highly heterogeneous mixture of molecules. The non-specific chemical linkage sites compromised the stability and function of the final product. Furthermore, the process was difficult to scale up for therapeutic manufacturing. These issues spurred the field toward genetic engineering solutions.
Engineering Full IgG-like Bispecific Antibodies
Modern production of full-sized, IgG-like bispecific antibodies relies on protein engineering to genetically force the correct pairing of the four distinct polypeptide chains. The goal is to ensure the two different heavy chains pair exclusively with each other (forming the correct heterodimer), and that each heavy chain pairs only with its intended light chain.
The most widely adopted technology to solve heavy chain mispairing is the “Knob-into-Hole” (KiH) approach. This method introduces specific amino acid mutations into the constant region 3 (CH3) domains of the two heavy chains. On one chain, a small amino acid is replaced with a bulky one, creating a protruding “knob.” On the partner chain, a complementary “hole” is engineered by replacing a bulky amino acid with smaller ones. The steric hindrance created by the knob and the complementary space of the hole strongly favors the association of the two different chains, dramatically increasing the yield of the desired heterodimer and minimizing unwanted homodimers.
Once heavy chain pairing is controlled, the next challenge is light chain pairing. One genetic solution is the “common light chain” strategy, where both antigen-binding arms utilize an identical light chain sequence. This avoids mispairing entirely but limits the diversity and specificity of the antibodies, as both arms must accommodate the same light chain.
A more flexible solution is the “CrossMab” technology, which addresses mispairing by swapping domains between the heavy and light chains in one of the Fab arms. For instance, the constant domain of the light chain (CL) can be exchanged with the constant domain 1 of the heavy chain (CH1). This domain-swapping creates an orthogonal interface that forces the light chain to pair specifically with its intended heavy chain. Combining a heavy chain heterodimerization technology like KiH with a light chain pairing solution like CrossMab allows for highly efficient production of functional, full-length bispecific antibodies.
Non-Conventional Bispecific Antibody Formats
Many bispecific antibodies are created using smaller, non-conventional formats that bypass the heavy/light chain pairing problem by eliminating the heavy chain constant region (Fc domain). These molecules are often composed only of the variable domains responsible for antigen binding.
One prominent format is the tandem single-chain variable fragment (scFv), which forms the basis of the Bispecific T-cell Engager (BiTE) technology. An scFv links the variable domains of the heavy chain (VH) and light chain (VL) via a short, flexible peptide. The tandem scFv format connects two of these scFv units, each with a different specificity, using a longer linker peptide. This results in a single, continuous polypeptide chain that binds two different targets simultaneously.
Another fragment-based format is the Diabody, a bivalent molecule composed of two polypeptide chains. Each chain contains a VH domain from one antibody connected by a short linker to the VL domain of a second antibody. The short linker prevents intra-chain pairing, forcing the domains to pair with the complementary domains on the other chain to form the bispecific dimer.
These fragment-based BsAbs are significantly smaller than a full IgG molecule, offering unique pharmacokinetic properties. Their compact size allows for better tissue penetration, which is advantageous for reaching dense tumor microenvironments. However, lacking the Fc region, they are rapidly cleared by the kidneys, resulting in a much shorter half-life compared to full IgG antibodies. This often necessitates continuous or frequent intravenous infusions.
Ensuring Purity and Function
Even with optimized genetic engineering, the final harvested product is always a mixture of the desired molecule and various impurities. A robust downstream manufacturing process is necessary to separate the functional BsAb from mispaired species, aggregates, and other contaminants.
Purification typically starts with an initial capture step. For Fc-containing BsAbs, this is often Protein A affinity chromatography, which binds to the Fc region. For fragment-based BsAbs lacking an Fc region, alternative affinity methods are used, such as Protein L chromatography (which binds to the kappa light chain) or Immobilized Metal Affinity Chromatography (IMAC).
After capture, a series of polishing steps achieve the high purity required for a therapeutic product. These steps involve specialized chromatography techniques that separate molecules based on differences in charge, hydrophobicity, or size. Ion-Exchange Chromatography (IEX) is frequently employed to separate the correctly paired heterodimer from mispaired homodimers, which have slightly different surface charges. Hydrophobic Interaction Chromatography (HIC) or Mixed-Mode Chromatography (MMC) may also be used to remove aggregates and fragments.
Following purification, quality control assays confirm the final product’s identity and function. These include tests to confirm the correct molecular weight and structure, and functional assays to verify that the BsAb correctly binds both target antigens. Techniques like Surface Plasmon Resonance (SPR) measure binding kinetics, while flow cytometry confirms the ability of the BsAb to simultaneously engage two different cell populations, such as T-cells and tumor cells.