A Translational Perspective – Biacore Systems in Discovery and Early-Stage Development

Abstract Translational science has launched a new era of antibody therapeutics, driving demand for novel antibody formats designed to improve efficacy of therapies and reach new targets. The trend towards increasingly complex drug targets drives the need for increased sensitivity during development, screening, and lead optimization. Biacore™ surface plasmon resonance (SPR) systems are extensively used in biotherapeutic antibody discovery and development. Here, we discuss the utility of Biacore systems from selection of first candidates to clinical lead. We show that a combination of Biacore SPR systems, software, sensor chips, and kits support the setup of screening and characterization assays and reduce the difficulties that come with assay development. During screening, antibody capture followed by an antigen injection permits the selection of monophasic and stable binders with preferred kinetics and stoichiometry. Standardized Biacore epitope binning procedures ensure reliable determination of epitope specificity, while the effect of antibody engineering efforts can be investigated by analysis of antibody binding to antigen and Fcγ-receptors. Biacore SPR systems also address key developability aspects, including ensuring critical binding properties remain unchanged in forced degradation studies. While using Biacore SPR systems, antibody concentration and kinetics can be monitored in the presence of nonbinding unfolded fractions, host cell proteins, and other impurities. For translational scientists seeking to progress their drug discovery from benchtop to clinical testing and beyond, Biacore SPR systems can provide the reliability, sensitivity, and automation to reduce risk of failure and assist with a smooth journey to clinic. Introduction Over the last 30 years, recombinant proteins including hormones, cytokines, and therapeutic antibodies were developed by academics and advanced by translational scientists for the treatment of diseases including diabetes, cancer, and rheumatic disorders. Recombinant insulin in various forms is a commonly prescribed biotherapeutic medicine. While hormones and cytokines represent an important class of biotherapeutics, antibodies are now the focus and will eventually have a wider applicability. The number of FDA approved antibodies is steadily increasing (1), demonstrating that the translational science space is accelerating drug discovery, with approved antibodies now directed at over thirty different target molecules. In 2021, the FDA approved the 100th monoclonal antibody product. (Fig 1). Other emerging areas of translational medicine include; antisense oligonucleotides, mRNA- based drugs, and targeted protein degraders. In this white paper, we outline the value of Biacore SPR system technology to the translational scientist concerned with antibody discovery and development. From antibody generation to clinical lead: the value of developability assessments Modern techniques for antibody generation often yield numerous candidates that achieve desired functional properties. This can make it challenging for translational scientists to select a lead candidate for downstream resources and labor-intensive steps such as cell line development, process development, and formulation. Assessing a candidate’s developability — the likelihood of successfully developing a lead candidate into a stable, manufacturable, safe, and efficacious drug — is highly advantageous. By interrogating lead candidate developability aspects early, both the success rate and speed of preclinical and clinical development can be enhanced. Liabilities like product heterogeneity, stability, and unwanted in vivo effects are avoided. The antibody development workflow (Fig 1) has evolved, with early development no longer focused solely on potency and functional aspects such as specificity, affinity, and kinetics for molecular targets. The developability aspects play an increasingly important role for reducing risk of failure later in the development program (2). During antibody development (Fig 1), there are two stages where developability assessments can be performed. At each assessment there will likely be multiple promising candidates selected for further development. A thorough characterization of biophysical antibody properties is not always necessary and, to save time and costs, it can be beneficial to focus on the most prevalent liabilities to eliminate candidates. These include: •Examining complementary determining regions (CDRs) for potential degradation sites (2). •The impact of post-translational modifications (PTMs), for example, methylation, acetylation, phosphorylation, glycosylation on stability and conformation (3,4). •Aggregation and fragmentation tendencies (5). •Solubility and solution stability (6). •Biological factors such as immunogenicity (7) and pharmacokinetic properties (8). The core questions that translational scientists seek to answer during developability assessment are: Can the antibody be manufactured? While a candidate lead antibody may show high affinity, potency, and specificity — if the manufacturing process or physical properties are suboptimal, it may be unsuitable for production at therapeutic scale. Suboptimal characteristics can provide difficulty in streamlining and accelerating the manufacturing process because monetary investments may be required to implement additional downstream processes to move the molecule forward into development. Preferred routes of administration may not be achievable if the molecules cannot sustain therapeutic concentrations because of aggregation or viscosity (9). Key manufacturability assessment attributes include purity/heterogeneity, stability, solubility, upstream/ downstream processing, and formulation. Assessing these attributes enables translational scientists to achieve their key manufacturability endpoints, including long-term stability in formulation buffer and viscosity at high protein concentrations. Is it safe? Since antibodies directly interact with the immune system, many carry an inherent risk of adverse immune-related reactions. Some promising lead candidates fail because of toxicity to humans. Detecting these issues early in the selection process is highly favorable. This includes assessing immunogenicity/immunotoxicity, affinity for known toxicological targets, off-target binding, and half-life. Will it have acceptable bioavailability and efficacy? A lack of efficacy is one of the primary reasons why drugs fail at clinical trials. Ensuring that an antibody lead candidate displays a satisfactory level of bioavailability and efficacy is vital to success further down the line. Selecting lead candidates with optimal affinity, potency, specificity, and pharmacokinetics leads to better success in selecting a high-efficacy candidate. New antibody formats A large majority of FDA approved antibodies are full-length and of IgG1, IgG2, or IgG4 subclass. However, non-traditional antibody formats are slowly emerging. For example, blinatumomab (Blincyto, approved 2014) is a bispecific antibody while brentuximab vedotin (Adcetris, approved 2011) and ado-trastuzomab emtansine (Kadcyla, approved 2013) are based on conventional antibodies but conjugated with cytotoxic agents, and are so-called antibody-drug conjugates (ADCs). New formats are introduced to improve the efficacy (10) of therapies to reach new targets, for instance by designing antibodies that have the capability to cross the blood-brain barrier (11). Antibody formats that retain the basic structure of IgGs may inherit their pharmacokinetic properties (12) but novel constructs that lack the Fc-part of the antibody may have reduced half-life (13). This can be an advantage if the antibody is used for imaging purposes (14). For therapeutic purposes, there needs to be a balance between efficacy and half-life for smaller antibody formats such as single chain Fvs and nanobodies, or other scaffolds that have the potential to reach more hidden targets and even act as intrabodies to target intracellular antigens (15). Several pharmaceutical companies now have bispecific antibodies (16) and antibody-drug conjugates in their clinical pipeline (17) while intracellular antibodies may still be in research phase (Fig 2). (A) (B) (C) Canonical antibody Antibody–drug conjugates Bispecifics (D) Fragments Fab scFvNanobody Fig 2. Antibody formats. Antibody formats include canonical (A), antibodydrug conjugates (B), bispecifics (C) and fragments (D). Fragments include antigen- binding fragments (Fabs), single- chain variable region (scFv) constructs, and domain antibodies. Radiolabelled antibodies and antibodyimmunotoxins are not shown. These formats can be further subcategorized, and antibodies can span classifications. There are at least 30 different bispecific formats, for example, some of which include fragments. Modified from Nature Reviews Drug Discovery. Fig 1. The antibody development workflow. (A) First selection round with focus on antigen binding, and (B) Re-engineering and selection of clinical lead. Antibody generation Thousands Humanization Re-engineering Hundreds Developability assessment CDR focus A few Developability assessment Whole molecule focus A few Clinical need Screening and selection Functional aspects Tens Screening and selection Functional aspects Tens (A) (B) 2 CY37785-25Jul23-WH CY37785-25Jul23-WH 3 Efficiently select and optimize antibody candidates Biacore system technology was incorporated in the antibody development workflow almost immediately after launch in 1990, when kinetic analysis of antibody-antigen interactions (18)and epitope binning procedures (19) were described. Antibody D2E7, which later became Humira, the bestselling therapeutic antibody in 2021 (20), was selected from a Biacore screen in the mid-1990s (21). From screening of candidates to antibody engineering and final development, Biacore systems have been consistently used to determine specificity of binding to characterize antibody-antigen and antibody-Fc receptor interactions and to guide development towards a clinical lead. In developability studies, Biacore systems are typically used to monitor effects of forced degradation on antigen and Fc gamma receptor binding and for assessment of pharmacokinetic properties where binding to FcRn is related to antibody half-life. More recently, the use of binding mode-specific reagents (22) has been described for detection of changes in antibody topography as a consequence of forced degradation. Biacore systems in antibody development Depending on the design of the flow system, direct interaction analysis with several target molecules can be performed with a single injection of sample. Typically, one measuring spot is used for active analysis and one for referencing (Fig 3). Depending on the Biacore system, one to eight active/reference pairs can be used. Independently of the number of active reference pairs, double referencing is typically performed to arrive at high quality data by subtracting blank cycles. Sample throughput is linked to the number of microwell plates that can be handled in an automated run. Higher throughput systems are suited for antibody screening and large-scale epitope binning experiments while other systems have functions that make them ideal for detailed characterization studies. There is a considerable overlap in the systems and those intended for screening are also used for characterization and vice versa. Minimize time spent on assay development with dedicated sensor chips and reagents Antibody applications are supported by several sensor chips and reagents provided with ready-to-use protocols, enabling rapid assay development using well known and reversible capture formats. Fig 5. A stepwise approach to antibody selection. The clone library is analysed to allow selection of candidates for re-engineering. Construction of clone library Yes/No binding Ranking, specificity, kinetics Developability aspects Clones for re-engineering Biacore Sensor Chip Protein A, Sensor Chip Protein G, or Sensor Chip Protein L (Fig. 4A) can be used directly for antibody concentration measurements or for capture of antibodies and subsequent kinetic analysis of antibodyantigen interactions. Sensor Chip PrismA (Fig 4A) is suitable for antibody concentration measurements but not recommended for kinetic analysis of antibody–antigen interactions. Biacore Sensor Chip CM5 (Fig. 4B) can be used for direct immobilization and be combined with several antibody capture kits. With the Mouse Antibody Capture Kit, all IgG subclasses, IgM and IgA can be captured to immobilized polyclonal rabbit anti-mouse immunoglobulin. The Human Antibody Capture Kit includes a monoclonal mouse anti-human IgG (Fc) antibody capable of capturing all IgG subclasses. The Human Fab Capture Kit features a mix of monoclonal antibodies and capture Fab through kappa and lambda chains. ScFv antibodies may be captured using protein L or its variants (23). In cases where a small antibody fragment, antigen or an Fc-receptor is captured on the sensor surface the His Capture Kit can be used to capture histidinetagged molecules using a monoclonal anti-histidine antibody. Biotinylated reagents can be captured onto Biacore Sensor Chip SA, Sensor Chip NA, or for a reversible biotin-streptavidin interaction, Biotin CAPture Kit (Fig. 4C) can be used. (A) (B)-50 50 2500 2000 1500 1000 500 0-500 150250 Time (s) Response (RU) Response (RU) Time (s) 350 450 100200 250 200 150 100 50 0-50 300200400450 150250350350 Ab capture Ag binding Ag dissociation Ag binding Ag dissociation Very stable Stable Biphasic-50 50 2500 2000 1500 1000 500 0-500 150250 Time (s) Response (RU) Response (RU) Time (s) 350450 100200 250 200 150 100 50 0-50 300200400 450 150250350350 Ab capture Ag binding Ag dissociation Ag binding Ag dissociation Very stable Stable Biphasic Fig 6. Screening data. (A) Assay setup with capture of antibody from media and injection of antigen, and (B) Comparison of different antigen binding profiles for identification of stable, very stable, and biphasic binding properties. Fig 4. Biacore™ sensor chips and kits for reversible capture of biomolecules. Sensor Chip Protein A Sensor Chip Protein G Sensor Chip Protein L Sensor Chip PrismA Antigen binding to captured antibody Capture reagents Histidine capture Anti mouse Fc Anti human Fc Anti human Fab Direct immobilization Antibody binding to histidine captured FcgRI Reversible streptavidin Biotin CAPture kit Antibody binding to biotin captured FcRn Products Application example-20 0 20 40 60 80 100 120-100 0 100 200 300 400 500 600 700 800 900 Time (s) Response (RU)-20 30 80 130 180-100 0 100 200 300 400 500 600 700 800 900 Time (s) Response (RU)-20 0 20 40 60 80 100 120-100 0 100 200 300 400 500 600 700 800 900 Time (s) Response (RU) Fig 3. Active and reference measuring spots and their use in sample and blank cycles. ActiveReference Sample cycle ActiveReference Blank cycle Buffer Sample antibody Sample molecule Immobilized capture antibody Deeper insight into biotherapeutic characteristics with improved efficiency Translational scientists are looking to conduct earlier assessment of expression levels, target specificity, and binding stability for clone selection. The number of samples from hybridoma cells or recombinant expression systems varies greatly but may range up to thousands. A multi-step approach is often applied as s