Multispecific biotherapeutics offer significant advantages in the treatment of complex diseases by enabling enhanced efficacy and broader therapeutic options compared with traditional monospecific therapies. Designed to bind two or more targets, these advanced modalities are increasingly prominent in the clinical pipeline. Since 2019, the number of multispecific molecules in development has grown steadily, with nearly a 20% increase over the past five years, a trend that is expected to continue.
As molecular complexity increases, so do the challenges associated with developability, manufacturability, and clinical safety. Embedding quality by design (QbD) principles early in development is therefore essential to de-risk these programs, support efficient progression to the clinic, and enable robust downstream manufacturing. These challenges, and the practical strategies to address them, are explored in greater depth in a dedicated technical webinar - Building Quality by Design for Multi-Specific Therapeutic Proteins.
How to build QbD into multispecific therapeutics?
Developing multispecific biotherapeutics requires careful consideration of several key factors:
- Post-Translational Modifications (PTMs): Unwanted liabilities can impact in vitro activity, kinetic assays, pharmacology, toxicology, clinical safety and both drug substance and drug product stability.
- Immunogenicity: Immunogenicity risks may affect pharmacology, toxicology, and clinical safety, potentially resulting in anti-drug antibodies (ADAs) that reduce therapeutic effectiveness.
- Aggregation Propensity: Increased aggregation can negatively impact process performance, yield, cost, development timelines and clinical safety.
- Expression System: The choice of expression system affects future manufacturability, scalability, process optimization, speed to clinic and regulatory risk.
Our comprehensive toolbox approach
Our has extensive experience expressing multi-chain therapeutics. In our Early Development Services alone, over 2,000 protein-based therapeutics have been expressed in-house using the GS Gene Expression System® technology, with a growing shift away from standard monoclonal antibodies (mAbs). In 2023, more than 60% of expressed molecules were based on new molecular formats. Today, we support more than 20 launched products that are non-standard monoclonal antibodies or advanced molecular formats.
Based on this experience, we have developed an integrated toolbox approach with four key components, supported by comprehensive in silico and in vitro assessments. This is the key approach we take in building QbD, which supports early-stage screening of titer, product assembly, potency, and immunosafety, strengthening the data package for better candidate selection and accelerated clinical development.
- Molecular reformatting: upfront formatting of molecules, including humanization or dehumanization of whole molecules or individual domains.
- In silico assessments and protein engineering: predictive tools that support candidate screening and lead selection, followed by expert protein engineering to de-risk and optimize.
- Non-GMP protein expression: use of our in-house GS Gene Expression System® technology provides early insight into protein expression in a commercially relevant platform, generating material for in vitro and in vivo preclinical testing.
- Immunosafety testing services: panels of in vitro immunogenicity assays enable early detection of potential immune responses before clinical development.
What are our early development in silico tools?
Our in silico tools are designed to address two critical developability questions:
- Is it safe? – Is the molecule and its individual domains safe to administer to patients?
- Can I make it? – Do they contain liabilities for manufacturing?
Currently, our offers manufacturability assessment, immunogenicity risk assessment and an integrated screening platform.
Manufacturability assessment
This assessment uses primary amino acid sequence and structural homology modeling to evaluate:
- Sequence liabilities (glycosylation sites, deamidation, isomerization, fragmentation)
- Protein aggregation propensity (considering hydrophobicity, solvent accessibility, charge)
- Protein fitness parameters (molecular weight, extinction coefficient, absorbance, PI)
For multi-chain molecules, this assessment can also predict likely misassemblies, supporting the selection of appropriate purification and analytical strategies for downstream development.
In silico immunogenicity prediction
A rapid and cost-effective immunogenicity assessment identifies potential T cell epitopes by:
- Scanning amino acid sequences using sliding windows to identify T-cell epitopes
- Assessing promiscuous binding to multiple HLA allotypes
- Predicting binding strength
- Calculating overall risk scores based on population frequency and importance of affected allotypes
- Providing rankings comparable to marketed antibodies
Lower risk scores usually correlate with human antibody sequences, while higher scores are associated with non-human sequences that pose greater immunogenicity risk.
Integrated screening platform
Our AI-driven platform combines manufacturability and immunogenicity prediction to screen hundreds of amino acid sequences.
- The screening process involves:
- Sequence annotation
- AI-driven structural model generation
- In silico prediction of immunogenicity, PTM, and aggregation
- Data benchmarking against similar marketed molecular formats
- Application of filters and risk weights to rank molecules
Weightings can be adjusted depending on molecular requirements and function, enabling an efficient ranking of formats and lead candidates based on predicted developability.
How to build quality into molecular designs with our in silico platform?
Our in silico tools can be applied to both individual building blocks and fully assembled molecular formats. Key considerations include:
- Input Selection: use of de-risked molecular domains (e.g. scFVs, Fabs, VHH fragments). For non-human starting domains, humanization and dehumanization tools can be applied.
- Molecular Architecture: selection depends on multiple factors:
- Number of chains or domains
- Need for specific chain or domain pairing technologies
- Format requirements (stoichiometry and distance between binding domains)
- Valency and binding strength
- Functional requirements, including half-life extension
- Modifications required for conjugation
Once a panel of formats meeting design requirements is established, in silico tools can be applied again to rank molecular formats and select top leads for in vitro expression based on predicted developability attributes.
What are our early development in vitro tools?
Our early expression toolbox rapidly expresses and purifies representative material for screening multiple molecule formats and lead candidates. Key capabilities include:
- Rapid construction of multiple vector designs
- Use of the same cells, vectors and medium feeds as process development and manufacturing (critical for generating representative data)
- Early screening for titer and product assembly post-transfection
- Material production ranging from few milligrams to several grams, depending on requirements
- Flexible purification strategies aligned with downstream processes
- Comprehensive analytical capabilities
In vitro immunogenicity assessment
Our immunogenicity assessment using human peripheral blood mononuclear cells (PBMCs) evaluates three potential immune response outcomes:
- Wanted immune response, where immune response is the intended outcome (e.g., vaccines)
- Unwanted immune response, resulting in anti-drug antibodies that can reduce drug effectiveness
- Immunotoxicity, such as cytokine release or off-target effects
Both innate and adaptive immune responses are evaluated, along with immediate immunotoxicity driven by product mode of action.
- PBMC bank and assay strengths
- On-site processing of whole blood and leukapaks
- Access to over 500 healthy donors and patient samples
- Assessment of responses to API, formulation, and impurities
- Human Tissue Authority license and ethics approval for storage and use of human samples
How to build quality into molecular designs with our in vitro platform?
In vitro expression in a commercially relevant platform demonstrates expression of functional and manufacturable lead candidates while providing representative material for testing. Our GS Gene Expression System® is a fully integrated system with four key components:
- Host Cell Lines: GS Knockout® and GS Xceed® cell lines (GS Xceed® cell line is a FucT8 knockout for A-fucosylated antibody expression)
- Vector System: GS Xceed® Pro vectors containing the high-strength gene promoter LHP-1
- Expression Technology: Including media systems
- Process Know-How: Complete package for commercial development
Our GS System® technology has been used to develop over 100 commercial products to date. The GS Xceed® Pro vector system enables construction of expression vectors for one, two, three, or four product genes.
What are strategies for vector design for multispecifics?
We use multi-gene vector approaches rather than co-transfecting single gene vectors. Co-transfection can result in heterogeneous pools with limited control over gene copy number and integration sites.
By contrast, our approach optimizes product quality and assembly upfront. Gene order within a single expression cassette balances chain expression, with all genes integrated using piggyBac® technology. Screening then focuses on consistency in titer and product quality rather than extensive clone hunting.
Learn more about our early development services or get in touch with us.
