Biophysical Methods Enhance mAb Characterization

Over the past 20 years monoclonal antibodies (mAbs) have been the most impactful drug category in terms of both sales and patient uptake. More than 150 mAbs have been approved globally, including 122 in the U.S. and 114 in Europe. Between 1985 and 2010 manufacturing and regulatory issues prevented mAbs from reaching their full potential, but several critical developments occurring in parallel transformed this industry.

Improvements in cell culture, mainly in media and feed development, were responsible for a rise in titers from tens or hundreds of milligrams per liter to as high as ten grams per liter, thus improving cost of goods. Single-use processing, once beset by issues but now an indispensable component of biologics processing, provided operational flexibility and paved the way to multi-product facilities. On the regulatory front, approval agencies, particularly the U.S. Food and Drug Administration, issued guidances on Quality by Design and Process Analytic Technology, which presented new frameworks and perspectives for quality best practices and process understanding.

Last but not least, the maturation and development of advanced analytical methodologies and instrumentation provided protein drug developers with the analytical tools for better understanding the critical aspects of both products and processes.

These developments, alongside advanced target discovery and disease modeling, were responsible for increasing demand for mAbs from about $300 million in 1999 to $186 billion by 2021; 2028 sales are expected to more than double, to $445 billion, a 13% annualized growth rate.

Analysis modes

mAb characterization workflows may be divided into three main groups: chromatography/affinity, spectroscopic, and biophysical methods. Some overlap and crosstalk exist among these methods, and any one category would take many pages to review. Several methods, however, have been instrumental in the success of biologic treatments and for achieving the goal of “process understanding.” These methods, although diverse, have found application in every stage of biopharmaceutical development, from discovery through post-marketing studies.

The first method, high-performance liquid chromatography (HPLC), has evolved from a specialty instrument to an all-purpose analytic platform; from a low-pressure system using large particle size media of limited resolving capability to ultra-performance LC (UPLC) using very high back pressures, sub-2-micron media particle sizes, and a range of selectivities including affinity, chirality, anionic and cationic exchange, size exclusion, etc.

UPLC/HPLC are now commonly interfaced to mass spectrometers (MS), which have undergone extraordinary evolution in their own right. In 1985, when the first mAb was approved, MS instruments took up an entire room and required doctoral-level maintenance and operation. Today MS instruments fit on desktops, with many MS modalities (e.g., ionization methods) enjoying unique biologic analysis niches. For example, soft ionization methods preserve critical structural features of mAbs during ionization while other methods have been applied, through proteomics, in the analysis of post-translational modifications on both intact and fragmented mAbs.

Biophysical methods

The third group of methods, biophysical analysis, report on the fundamental physical properties of a mAb. These approaches include analytical ultracentrifugation (AUC), differential scanning calorimetry (DSC), circular dichroism (CD), dynamic light scattering (DLS), the apolar fluorescent dye 8-Anilinonaphthalene-1-sulfonic acid (ANS) method, surface plasmon resonance (SPR), and two “overlap” methods: fourier-transform infrared spectrometry (FTIR) and size-exclusion chromatography (SEC).

Biophysical methods are typically applied to the analysis of higher-order protein structure, which is critical for assuring a mAb’s function, activity, stability, and ultimately quality. For example, characterizing a mAb preparation for particle size and aggregation typically involves SEC, AUC, DLS, plus separation via polyacrylamide gel electrophoresis. Workflows for characterizing conformation might include CD, FTIR, plus HPLC-MS and/or fluorescence labeling. For thermal stability researchers usually turn to CD, DSC, and one or two spectroscopic techniques (intrinsic tryptophan fluorescence and/or extrinsic fluorescence ANS dye binding).

Biophysical methods that quantify interactions between two or more molecules (protein-protein, protein-small molecule) are increasingly applied to mAb development, from screening through quality testing. By focusing on binding rather than on protein function, these methods differentiate between drug-like affinity and ligand-mAb binding artifacts.

Since mAb development is all about the affinity between mAb and various targets, including off-target species, considerable effort has gone into two biophysical methods: fluorescence polarization anisotropy (FPA) and SPR.

FPA measures mAb-target binding constants through the interaction between relatively small fluorescent ligands with larger receptor molecules. The output, anisotropy as a function of varying concentration of the fluorescent ligand at fixed target concentrations, can then be used to calculate affinity constants.

SPR quantifies interactions between a ligand immobilized on a sensor chip, with a mAb in real time and without any sort of labeling. Binding data are derived from small changes in refractive index (RI) at the surface of the chip: RI changes are proportional to the masses of species binding to the sensor surface. Changes over time are then used to calculate binding rates and kinetics, interaction affinity, and analyte concentration. SPR is especially suited to the analysis and characterization of a mAb’s antigen binding fragment (Fab).

Developers use SPR to select candidate molecules, to investigate a mAb’s mode of action, and to keep track of binding-related critical quality attributes throughout the product’s lifecycle.

Biophysical approaches to mAb characterization stand on their own—that is, they provide hard data related to a molecule’s physical properties. The real science, as it relates to drug development, is in the interpretation of what those data reflect in terms of a mAb’s safety, efficacy, and quality.

That level of interpretation requires input from other supporting analysis methods. That is why “mAb characterization” is always a composite of outputs from several orthogonal methods, each supplying its unique value, with sufficient overlap with other methods to complete the analytical puzzle with a minimum of “empty spaces” or unanswered questions.

SPR may provide an independent check on results from other analytical studies, particularly when mAbs exhibit identical behavior in one analytical dimension, but markedly different activity through another assay. For example, two proteins originating from a fermentation that “look” identical via MS, HPLC, etc. but with very different activities. SPR binding studies may suggest that the two species are different conformations of the same mAb.

SPR is particularly useful in probling such higher-order structures. In a recent paper, an interdisciplinary group investigated such structural changes resulting from the forced oxidative stress of the mAb abituzumab by hydrogen peroxide. Investigators took a “multianalytical approach” combining NMR, MS, DSC, bioassays, computational tools, and SPR to provide “qualitative and semiquantitative characterization” of higher-order structural changes in the mAb leading to the loss of biological activity.

As expected the mAb showed decreased affinity for its target by bioassay, which for a drug usually signifies reduced activity. The question was how?

The clues:

• The primary structure checked out according to LC-MS, however
• Before-after examination of methionine residues showed that some were oxidized to a great extent while some were barely affected
• NMR spectra showed small perturbations but no major conformational changes
• DSC indicated that oxidation affected thermal stability

Piecing this evidence together, researchers concluded that oxidation did not affect the biological activity of the Fab region: “Only when the biological activity of the Fc portion ... was investigated by SPR ... in terms of binding to the FcRn receptor, the effects induced by oxidation were observed.”

mAb characterization, particularly during early development, requires evidence from multiple assays and analyses. Some tests, including most biophysical readouts and NMR, are used primarily to guide the discovery stage. The relatively few methods that persist through manufacturing and quality control include LC, LC-MS, and to a lesser degree, SPR. Of the three SPR is the most rapid, requires the least sample preparation, and can be thought of as a sentinel assay.