What is chain mispairing in bispecific antibody manufacturing?

Chain mispairing occurs when the four polypeptide chains of a bispecific antibody, two different heavy chains and two different light chains, associate incorrectly in the production cell, forming homodimers, half-antibodies, or assemblies where a light chain pairs with the wrong heavy chain. Only one of the possible pairings is the desired bispecific product. Without engineering interventions, mispairing products significantly reduce yield and create purification challenges because they share nearly identical physicochemical properties with the target.

How does knob-into-hole technology reduce heavy chain mispairing?

Knob-into-hole engineering introduces complementary steric mutations into the CH3 domains of the two heavy chains: a bulky amino acid substitution (the "knob") on one chain, and a complementary cavity-forming substitution (the "hole") on the other. The shape complementarity between the two modified domains makes heterodimer formation thermodynamically favorable over homodimer formation. KiH is used in mosunetuzumab and faricimab, among other approved bispecifics, but still produces low levels of hole-hole homodimer that require downstream removal.

Why is Protein A purification insufficient for bispecific antibodies?

Protein A captures the bispecific product and all Fc-containing impurities, including homodimers and half-antibodies, with equal affinity, because all of these species present the same Protein A binding site at the CH2-CH3 Fc junction. Standard Protein A elution cannot separate the product from co-eluting impurities. Optimized Protein A cycles with an intermediate low-pH wash step can achieve partial separation, and orthogonal polishing steps by mixed-mode or ion exchange chromatography remove remaining impurities.

What is controlled Fab arm exchange and how does it produce bispecific antibodies?

Controlled Fab arm exchange (cFAE) produces two half-antibodies with matched IgG4 Fc mutations separately, then combines them under controlled reducing conditions that drive strand exchange in the hinge region, forming a bispecific heterodimer. Because each half-antibody is produced and purified independently, the starting material is pure, and the exchange reaction produces a bispecific of substantially higher purity than direct CHO co-expression. cFAE is used in amivantamab, epcoritamab, and teclistamab.

How does bispecific antibody characterization differ from monoclonal antibody release testing?

Bispecific release testing adds identification and quantitation of the bispecific-specific impurities (homodimers, half-antibodies, and light chain mispairings) to the standard mAb quality attribute suite. SEC-HPLC must resolve the bispecific from homodimers at similar molecular weight. CEX charge variant analysis is more complex because two different Fab arms create a broader and more heterogeneous charge profile. Mass spectrometry is required to confirm correct assembly of all four chains and verify that both antigen-binding arms are structurally intact.

Articles

Overcoming production bottlenecks in next-generation biologics

Bispecific antibodies require a precision engineering approach that conventional Protein A purification wasn't designed to handle.

Written byTrevor J Henderson

| 8 min read

A researcher reviews a bispecific antibody molecular structure rendering on a monitor, with chromatography purification traces and protein analysis results visible on the laboratory bench.

The structural asymmetry of bispecific antibodies is their therapeutic advantage and their manufacturing challenge. Each distinct arm enables engagement of a different target, but that same asymmetry means two different heavy chains and two different light chains must assemble correctly in the same cell during production.

Flow (2026)

Listen with Speechify

0:00

8:00

Highlights in this article

Key takeaways
Why bispecific antibodies stress every part of the conventional mAb platform
What is chain mispairing and what engineering solutions have proven effective at scale?
Upstream expression: achieving correct stoichiometry in CHO cells
How are product-related impurities removed from bispecific purification trains?
Analytical characterization: the complexity of releasing a bispecific
What does the bispecific production challenge reveal about next-gen biologics?

Bispecific antibodies represent the most commercially mature class of next-generation biologic, with more than 35 approved products globally and hundreds more in clinical development. They are also the most rigorous test of whether conventional monoclonal antibody manufacturing platforms can handle formats that introduce structural asymmetry, chain mispairing, and downstream purification challenges that Protein A affinity capture alone cannot resolve.

Key takeaways

Bispecific antibodies require two different heavy chains and two different light chains to assemble correctly in the same CHO cell. Without engineering interventions, the four chains can form multiple mispairing combinations, with only one being the desired heterodimeric product. Chain mispairing engineering is a manufacturing prerequisite, not an optional optimization.

Knob-into-hole (KiH) technology addresses heavy chain heterodimerization through complementary steric mutations in the CH3 domains of the two heavy chains. It is used in several approved bispecific antibodies including mosunetuzumab and faricimab. However, KiH does not fully eliminate light chain mispairing and still produces "hole-hole" homodimer impurities that are physicochemically similar to the product.

Downstream purification of bispecific antibodies requires resolving the product from homodimers, half-antibodies, and mispaired species that share nearly identical properties with the target molecule. Optimized Protein A chromatography with an intermediate low-pH wash step can achieve approximately 60% monomeric purity increase from a single capture step for KiH bispecific formats.

Non-affinity purification platforms combining mixed-mode chromatography capture with anion exchange polishing have demonstrated greater than 98% purity by SEC-HPLC with approximately 60% overall recovery for KiH bispecific formats, offering a viable alternative to Protein A for programs where the format does not support standard affinity binding.

The bispecific antibody production challenge illustrates the broader principle governing all next-generation biologic formats: every structural departure from the standard IgG1 scaffold introduces manufacturing complexity that platform processes were not designed to handle. The commercial success of approved bispecifics reflects decades of engineering work to make their production viable, not an easy adaptation of mAb infrastructure.

For the strategic context of bispecific antibodies within the next-generation therapy manufacturing landscape, see Scaling the unscalable: manufacturing cell, gene, and mRNA therapies. For how ADCs face their own distinct set of manufacturing challenges at scale, see the DDN coverage of antibody-drug conjugates: mechanism, pipeline, and outlook.

Why bispecific antibodies stress every part of the conventional mAb platform

A standard IgG1 monoclonal antibody is a symmetric molecule: two identical heavy chains and two identical light chains, all encoded by the same gene sequences, express and assemble in the same stoichiometric ratios because there is nothing else for them to pair with. This symmetry is not just aesthetically convenient: it is the molecular basis for platform manufacturing, because every step in the process is designed around the physical and chemical properties of a molecule that is the same everywhere across a production run.

A bispecific antibody breaks this symmetry deliberately. By design, two different heavy chains with different variable domains must associate as a heterodimer rather than as homodimers, and two different light chains must each pair with their cognate heavy chain rather than interchanging. This structural asymmetry is what allows the bispecific to engage two different antigens simultaneously, which is its therapeutic mechanism. It is also what makes it difficult to manufacture at the purity and yield required for commercial supply.

The consequences propagate through every stage of the process. Upstream, CHO cell expression of two heavy chains and two light chains produces a mixture of correctly paired heterodimer and multiple mispairing products. Downstream, the correctly paired bispecific and its major impurities have nearly identical physicochemical properties: the same molecular weight, the same Fc region, and very similar charge. The analytical methods that detect charge heterogeneity and molecular size variants in monoclonal antibodies must be reconfigured for the more complex impurity profile of a bispecific, and the specifications that define acceptable product quality require additional product-specific development.

For the regulatory environment in which bispecific development programs must demonstrate manufacturing control over this complex impurity profile, see the related DDN coverage of why gene and cell therapies are stalling at the FDA. The characterization, comparability, and specification challenges described there for cell and gene therapies apply in a parallel form to next-generation biologics like bispecific antibodies.

What is chain mispairing and what engineering solutions have proven effective at scale?

Chain mispairing in bispecific antibody production refers to the incorrect association of the heavy and light chains expressed from two different antibody sequences in the same production cell. A 2025 study from Adimab on design of orthogonal constant domain interfaces for bispecific antibody heavy and light chain pairing surveyed the range of engineering approaches used in FDA-approved IgG-like bispecific antibodies and identified the specific technologies enabling each approval. Without engineering intervention, the four polypeptide chains can form multiple pairings, including homodimers of each heavy chain, half-antibodies, and light chains paired with incorrect heavy chains.

Knob-into-hole technology addresses heavy chain mispairing by introducing complementary steric mutations in the CH3 domains. One heavy chain receives a bulky "knob" substitution (typically T366W), and the other receives a "hole" substitution (T366S, L368A, Y407V) that creates a cavity that accommodates the knob. The complementary shape of the two modified CH3 domains thermodynamically favors heterodimer formation over homodimer formation. KiH is used in mosunetuzumab (Lunsumio, Roche) and faricimab (Vabysmo, Roche), among other approved bispecifics.

Controlled Fab arm exchange (cFAE) is used in several approved bispecific antibodies, including amivantamab, epcoritamab, and teclistamab. In cFAE, two different half-antibodies are produced separately with matched IgG4 Fc mutations that promote Fab arm exchange under reducing conditions. Mixing the two half-antibodies under controlled redox conditions drives the formation of a bispecific heterodimer. Because the exchange is driven by thermodynamics rather than relying on co-expression in a single cell, the purity of the bispecific product from cFAE is substantially higher than from direct co-expression, reducing the downstream purification burden.

For light chain mispairing, the engineering solutions include CrossMAb, which swaps the CH1 and CL domains of one Fab arm to prevent incorrect light chain pairing, and common light chain design, in which both binding arms are engineered to use the same light chain sequence. Common light chain design, used in emicizumab (Hemlibra, Roche), eliminates light chain mispairing entirely because there is no differential between the two light chains to mispair. Tandem scFv formats such as blinatumomab (Blincyto, Amgen) avoid both heavy and light chain mispairing entirely by fusing two scFv domains in a single polypeptide chain, at the cost of losing the Fc region and its associated half-life, effector function, and Protein A capture properties.

Technology	Mispairing addressed	Fc region present	Protein A capture	Downstream challenge	Approved examples
Knob-into-hole (KiH)	HC homodimerization; reduces but does not eliminate hole-hole homodimer	Yes; standard IgG Fc	Yes; standard conditions	Hole-hole homodimer impurity requires additional polishing; light chain mispairing must be addressed separately	Mosunetuzumab (Lunsumio); faricimab (Vabysmo)
Controlled Fab arm exchange (cFAE)	HC mispairing; high product purity by design of separate half-antibody production	Yes; IgG4-based Fc	Yes; each half-antibody captured separately before exchange	Requires reductant-controlled assembly step; both half-antibodies require GMP manufacture; light chain mispairing addressed by design	Amivantamab (Rybrevant); epcoritamab (Tepkinly); teclistamab (Tecvayli)
Common light chain	LC mispairing eliminated; HC mispairing requires KiH or pI engineering	Yes; standard IgG Fc	Yes	HC homodimers still require downstream removal; discovery constraints from requiring shared light chain across both binding arms	Emicizumab (Hemlibra); glofitamab (Columvi)
Tandem scFv (BiTE)	No HC or LC mispairing; single polypeptide chain format	No; typically no Fc domain	No; cannot use Protein A; alternative capture required	Short half-life requires continuous infusion; no standard affinity capture; alternative purification platforms required	Blinatumomab (Blincyto); many clinical-stage BiTE molecules

Upstream expression: achieving correct stoichiometry in CHO cells

For bispecific formats produced by co-expression of two heavy chains and two light chains in a single CHO cell, the ratio at which the four chains are expressed determines both the yield of correctly paired heterodimer and the composition of the mispairing impurity profile. If one heavy chain is expressed at twice the level of the other, the excess chain will preferentially form homodimers rather than contributing to the target heterodimer, reducing yield and increasing downstream purification burden.

Achieving controlled stoichiometric expression requires design of the expression construct with chain-specific promoters, selection of production clones that express all four chains at optimized ratios, and sometimes the use of individual plasmids or separate gene integration sites for each chain to allow independent regulation. The clone selection process for a bispecific program is substantially more complex than for a monoclonal antibody, where the only variable is expression level, because the process must identify clones with both high total expression and the correct inter-chain expression ratio.

Production process development must also account for the possibility that expression ratios shift as the cell culture progresses through inoculation, exponential growth, and stationary phase, since different metabolic conditions can affect the translation efficiency of each chain independently. Process characterization studies that monitor the mispairing profile across the production run, not just at harvest, are important for understanding whether upstream expression variability contributes to batch-to-batch differences in downstream purification performance.

How are product-related impurities removed from bispecific purification trains?

The proximity of bispecific product and homodimer impurities in physicochemical space is the central challenge of bispecific downstream processing. Research demonstrating excellent removal of knob-into-hole bispecific byproducts in a single Protein A step showed that an intermediate low-pH wash step incorporated into the standard Protein A cycle, before the standard low-pH elution, can differentially remove the hole-hole homodimer impurity while retaining the target bispecific, achieving approximately a 60% increase in monomeric purity from the clarified cell culture fluid in a single capture step. This result confirms that standard Protein A can be adapted for bispecific purification with optimized wash conditions, without requiring a fundamentally different capture platform.

For KiH bispecifics with kappa and lambda light chains on the two arms, a sequential affinity approach using two different VHH-derived affinity resins provides selective purification of the kappa-lambda heterodimeric product from both homodimers. A first resin targeting the CH1 domain (such as CaptureSelect CH1-LX) captures all antibody species, followed by a second resin targeting the kappa light chain (KappaSelect), which retains kappa-containing species and allows the lambda-only homodimer to flow through. The kappa-lambda bispecific, which binds both resins, is selectively enriched at each step.

The full analytical and strategic context for VHH-derived affinity ligands in bispecific antibody purification, including the mechanism of sequential CH1-LX and KappaSelect capture that separates kappa-lambda bispecifics from both kappa-kappa and lambda-lambda homodimers, is covered in the related Separation Science article on purifying next-generation modalities with nanobody binders.

Non-affinity purification platforms represent a complete departure from Protein A-dependent processing. A 2024 study demonstrating a non-Protein A platform for KiH bispecific antibody purification used mixed-mode Capto adhere resin at pH 7.9 as the capture step, followed by anion exchange chromatography in weak binding mode at pH 7.8, achieving greater than 98% purity by SEC-HPLC, host cell protein levels below 10 parts per million, and overall recovery of approximately 60%. This non-affinity platform offers full independence from Protein A, which is advantageous for bispecific formats that bind Protein A weakly or asymmetrically, and for commercial programs where Protein A resin cost and leachables management are significant considerations.

Analytical characterization: the complexity of releasing a bispecific

Release testing for a bispecific antibody must address an expanded impurity profile relative to monoclonal antibodies. Standard mAb size-based methods, such as SEC-HPLC, detect aggregates and fragments. For a bispecific, SEC must also detect and quantify the homodimer impurities, which run at a similar molecular weight to the desired heterodimer and may require method development to achieve adequate resolution. CEX charge variant analysis of a bispecific is inherently more complex than for a monoclonal antibody because the two different Fab arms of the bispecific have different isoelectric points, creating a broader and more heterogeneous charge variant profile.

The CEX method development challenge for bispecific charge variant analysis, including how pH gradient and salt gradient methods differ in their ability to resolve the more complex charge variant profile of a bispecific, is covered in the related Separation Science article on CEX chromatography for charge variant analysis in mAb development. The principles described for mAbs apply to bispecifics with additional complexity from the two distinct Fab charge distributions.

Bispecific identity and correct assembly must be confirmed by mass spectrometry to verify that the product contains the correct combination of heavy and light chains and that both antigen-binding arms are structurally intact. Intact mass analysis under non-denaturing conditions can confirm the expected molecular weight and distinguish the bispecific from the homodimers. Peptide mapping under denaturing conditions confirms the amino acid sequence of all four chains. Functional potency assays that depend on the simultaneous engagement of both targets provide direct evidence that both arms are present and active.

What does the bispecific production challenge reveal about next-gen biologics?

The engineering investments required to produce bispecific antibodies at commercial scale, and the purification and analytical innovation that has followed, represent a model for how the industry approaches each new next-generation biologic format. The same pattern is visible across other complex formats: ADCs require controlled chemical conjugation to precise drug-to-antibody ratios at commercial scale; multispecific antibodies beyond bispecific add further chain pairing complexity; and modified Fc variants with engineered effector functions or extended half-life require characterization methods that standard mAb analytics do not address.

The common thread is that platform manufacturing, the strategy of applying a single validated process to a broad class of molecules, becomes less applicable as formats diverge from the reference IgG structure. Programs that recognize this early and invest in format-specific process development from the outset, rather than attempting to retrofit standard mAb processes to non-standard molecules, consistently reach commercial manufacturing with fewer late-stage surprises.

The ADC manufacturing challenge, which involves controlled chemical conjugation of cytotoxic payloads to antibodies at precise stoichiometries while maintaining both antibody structure and payload stability, is a parallel example of next-generation biologic manufacturing complexity. For the DDN coverage of ADC mechanism and pipeline context, see What are antibody-drug conjugates (ADCs): mechanism, pipeline, and outlook.

This article was produced under Drug Discovery News' AI Editorial Guidelines.

Frequently Asked Questions (FAQs)

What is chain mispairing in bispecific antibody manufacturing?
Chain mispairing occurs when the four polypeptide chains of a bispecific antibody, two different heavy chains and two different light chains, associate incorrectly in the production cell, forming homodimers, half-antibodies, or assemblies where a light chain pairs with the wrong heavy chain. Only one of the possible pairings is the desired bispecific product. Without engineering interventions, mispairing products significantly reduce yield and create purification challenges because they share nearly identical physicochemical properties with the target.
How does knob-into-hole technology reduce heavy chain mispairing?
Knob-into-hole engineering introduces complementary steric mutations into the CH3 domains of the two heavy chains: a bulky amino acid substitution (the "knob") on one chain, and a complementary cavity-forming substitution (the "hole") on the other. The shape complementarity between the two modified domains makes heterodimer formation thermodynamically favorable over homodimer formation. KiH is used in mosunetuzumab and faricimab, among other approved bispecifics, but still produces low levels of hole-hole homodimer that require downstream removal.
Why is Protein A purification insufficient for bispecific antibodies?
Protein A captures the bispecific product and all Fc-containing impurities, including homodimers and half-antibodies, with equal affinity, because all of these species present the same Protein A binding site at the CH2-CH3 Fc junction. Standard Protein A elution cannot separate the product from co-eluting impurities. Optimized Protein A cycles with an intermediate low-pH wash step can achieve partial separation, and orthogonal polishing steps by mixed-mode or ion exchange chromatography remove remaining impurities.
What is controlled Fab arm exchange and how does it produce bispecific antibodies?
Controlled Fab arm exchange (cFAE) produces two half-antibodies with matched IgG4 Fc mutations separately, then combines them under controlled reducing conditions that drive strand exchange in the hinge region, forming a bispecific heterodimer. Because each half-antibody is produced and purified independently, the starting material is pure, and the exchange reaction produces a bispecific of substantially higher purity than direct CHO co-expression. cFAE is used in amivantamab, epcoritamab, and teclistamab.
How does bispecific antibody characterization differ from monoclonal antibody release testing?
Bispecific release testing adds identification and quantitation of the bispecific-specific impurities (homodimers, half-antibodies, and light chain mispairings) to the standard mAb quality attribute suite. SEC-HPLC must resolve the bispecific from homodimers at similar molecular weight. CEX charge variant analysis is more complex because two different Fab arms create a broader and more heterogeneous charge profile. Mass spectrometry is required to confirm correct assembly of all four chains and verify that both antigen-binding arms are structurally intact.

Add Drug Discovery News as a preferred Google source to see more of our trusted coverage.

About the Author

Trevor J Henderson
Trevor Henderson is the Creative Services Director for the Laboratory Products Group at LabX Media Group. With over two decades of experience, he specializes in scientific and technical writing, editing, and content creation. His academic background includes training in human biology, physical anthropology, and community health. Since 2013, he has been developing content to engage and inform scientists and laboratorians.
View Full Profile