Empowering Protein Characterization

Protein therapeutics are an important class of medicines that are used to treat a variety of diseases. According to market research, the protein therapeutics market is estimated to grow at a compound annual growth rate of 6.86% between 2020 and 2027 (1). This growth is expected to be driven by the rapid rise in chronic disorders, advancements in technologies, broadening awareness of protein therapeutics, increasing adoption of plasma-derived therapies to manage chronic disorders, and mounting government initiatives to develop healthcare sectors (1).

However, therapeutic proteins are notoriously less stable than conventional pharmaceuticals; and due to their complex molecular structures, they can be challenging to develop successfully (2). “Compared to small molecules and peptides, proteins are much more complex,” explains Khanh Courtney, PhD, senior director, Biologics, Element. “Therefore, the understanding of the biophysical and biochemical characteristics of a protein biomolecule is critical in the identification of critical quality attributes (CQAs) to ensure consistent safety and efficacy.”

As proteins are formed by long chains of amino acids that interact with each other, secondary and tertiary molecular structures are formed, notes Courtney. Beyond the tertiary structure, the protein can bind to itself, which results in even higher order structures, such as dimer, trimer, and oligomer, she states.

“Certain amino acids could be post-translationally modified with phosphate groups, methyl groups, simple sugars or complex sugars, and so on, leading to a highly complex protein therapeutic molecule,” Courtney says. “The folding of the protein, the higher order structure, [and] the composition of its post-translational modifications, particularly of the glycans, are all examples of protein characterizations that could impact the potency and function of the molecule.”

The highest importance

“Protein characterization is of [the] highest importance throughout all stages of the value chain from discovery to development, through to quality control,” emphasizes Martin Vollmer, strategic program office lead for Biopharma, Agilent Technologies. “Characterization of a protein provides information about its structure and composition, which is directly linked with its function, leading to a greater understanding of differences and ensuring that quality criteria of developed biopharmaceutical products are met according to regulatory guidelines, ensuring drug efficacy and drug safety.”

For Scott J. Berger, PhD, senior manager, Biopharmaceutical Markets, Waters Corporation, whether making a protein drug via a recombinant or synthetic process, development should start with protein characterization and having as much knowledge as possible about the product. “These processes are imperfect, so they are going to make products with low-level variations that are subject to degradation pathways that further raise product variability,” he specifies. “It’s this complexity that creates reliance on sophisticated analytical tools.”

Tools, such as liquid chromatography (LC) and mass spectrometry (MS), have provided process scientists with the ability to monitor important attributes of a molecule in response to changes in the process, Berger adds. “This [capability] gives scientists the flexibility to incrementally refine their process to improve drug quality and yield by establishing analytical comparability,” he says. “Characterization assays often translate to targeted methods for process monitoring and product release where you are verifying that you have control of your process and product quality.”

A number of challenges

A key challenge related to protein characterization is that methods usually require a lot of test material, Courtney stresses. “Ideally, protein characterization is done in the early stages of chemistry, manufacturing, and controls (CMC) program development so that critical attributes are identified early on. Unfortunately, the early stage of CMC, for instance during process development, consists of small-scale productions done at the lab bench and have low yield,” she states. “This challenge can be, and is being, dealt with by the advancement of analytical methodologies that can assess more than one characteristic at a time in a single sample preparation or method.”

An example of how this challenge is being overcome can be found in the use of high resolution, high sensitivity MS. “This approach can be used to decipher the peptide map of the protein, post-translational modifications, deamidation, disulfide bonds, aggregation, molecular mass verification, glycan structure elucidation, and even tertiary structure determination, all in one approach,” Courtney explains. “By packaging multiple attributes into one method, a small yield of protein produced at the lab bench could produce a wealth of data to assess protein characteristics to better inform manufacturing and quality decisions.”

There are several immediate challenges to protein characterization in Berger’s opinion. “First, is the increasing complexity of the molecules,” he says. “We’re seeing the appearance of bi- and tri-specific molecules, antibody-drug conjugates [ADCs], and complex fusion proteins, all of which have more attribute variation than your typical monoclonal antibody, requiring enhanced analytics to define and monitor them.”

Additionally, in light of the increasing number of accelerated reviews by regulatory agencies, a challenge has arisen in the form of being able to fully characterize molecules in a compressed timescale and at the same time as stability studies and formulation development are taking place, Berger points out. “So, there is an increasing need for efficient analytical platform methods that can utilize prior knowledge to accelerate these studies,” he states.

“Another challenge many organizations face is the lack of actionable knowledge during the clone selection stage of product discovery/development,” Berger continues. “Use of LC–MS at this stage can ensure that a biosimilar, for example, matches up to the target analytical profile of the innovator molecule when it comes to a set of key and critical quality attributes.” When used as a part of a quality-by-design (QbD) approach, LC–MS can help to ensure that the desired product attributes are built into the molecule, he notes. If there is insufficient knowledge about a molecule at this stage of development, it is possible that impurities or instabilities won’t be discovered until after the product is placed into clinical production, Berger warns.

For bench scientists, experimental bias can be a significant challenge, Berger highlights. “A top reason why assay-to-assay test results vary from one scientist to another has to do with complicated and multi-step sample preparation methods,” he says. “For multi-attribute monitoring of biomolecules by LC–MS, it’s important that you have a very high-quality and reproducible sample for meaningful results. To address this need automated sample prep technologies are being investigated by many groups as a tool to improve sample quality and reproducibility by eliminating the human factor from potentially biasing assay results.”

Undesirable secondary interactions can adversely impact analytical data when analyzing organic acids, organophosphates, oligonucleotides, phosphopeptides, acidic glycans, and phospholipids, Berger notes. “One example of a secondary interaction is the non-specific adsorption by certain biomolecules to the metal surfaces—or even to bio-compatible materials like titanium—of liquid chromatographs and chromatographic columns,” he states. “Another type of secondary interaction is caused by the highly active surfaces of proteins which have a propensity to interact with the hydrophobic and electrostatically-active sites on the silica and hybrid silica chromatographic particles in the UHPLC [ultra-high performance LC] or UPLC [Ultra-Performance LC] column.”

These secondary interactions cause difficulties when characterizing and monitoring drug products by size exclusion chromatography (SEC), which is used to determine protein sample size variants claims Bill Warren, Principal Bioseparation product manager, Waters Corporation. If these attributes are not carefully controlled then the safety and efficacy of the product can be adversely affected, Warren specifies. “It’s for this reason that regulatory agencies require drug firms to accurately quantify the protein size variants like mAb aggregates, monomers and fragments in protein drug products,” he says.

Vollmer emphasizes that protein characterization is highly complex and is only becoming more complex over time. “In the past, it was mainly monoclonal antibodies (mAbs), hormones, vaccines, modified human proteins, and similar therapeutic proteins. Now with the discovery of antibody [drug conjugates (ADCs) fusion-proteins, bi-specific Abs, and other modalities, the task of protein characterization now requires a much more flexible approach,” he asserts.

This increasing complexity is then also tied to the skill of the analytical scientist that needs to perform the characterization, Vollmer stresses. “There are a multitude of assays that need to be performed and the instrumentation required is more sophisticated, but costly,” he explains. “Big hurdles are, therefore, the cost of implementation, operation, and skilled personnel.”

“Another important challenge is the large amount of data that needs to be analyzed for meaningful information and further action,” Vollmer continues. “More broadly, regulatory hurdles and inertia in the industry are often roadblocks for fast innovation.”

Empowering technologies

“LC and MS have been the empowering technologies for the well-characterized biotherapeutic,” Warren says. “LC is essential for fully analyzing an intact protein or its sub-units for glycan profiling or peptide mapping or for looking at protein charge or size variants.”

A technique that is useful for characterizing protein size variants is SEC, Warren adds. There have been several new SEC columns that have been brought to the market recently, he states, that address “the problem of secondary interactions (both ionic and hydrophobic) when separating protein aggregates, monomers, and fragments over a molecular weight range of 10,000 to 6,500,000 Daltons.”

Berger also notes that through combining LC and MS, it is possible to establish the primary structure (product sequence), which is a regulatory filing requirement. “With LC–MS, you get to explore the product variation, its stability, and degradation pathways, which ‘hot spots’ to monitor to ensure that the product qualities and your processes are under your control,” he says. “LC–MS is also ideal for looking at a molecule’s higher order structure, which relates to the folding and stability of these biotherapeutics, and how they interact with their protein targets, or themselves when it comes to things like molecular aggregation.”

The structural information obtained by LC–MS can drive product development and can also form the basis of intellectual property, Berger points out. “Understanding higher order structure and stability dynamics may require the use of techniques like hydrogen deuterium exchange or ion mobility mass spectrometry (IMS-MS) and more recent techniques, such as collision-induced unfolding (CIU). CIU is an approach where the mass spectrometer is almost like a calorimeter, adding energy to the protein molecule and seeing how it unfolds within the IMS-MS,” he says. “These techniques allow you to ask questions that go beyond primary structure to understanding the dynamics of its folded structure and interactions with its protein target.”

Courtney highlights three techniques that are important for protein characterization: “High resolution, high sensitivity, orbitrap MS for protein and glycan identification, and structural elucidations; capillary electrophoresis to determine charge heterogeneity; and circular dichroism for secondary structures.”

Although there is no single standard technique to characterize proteins, Vollmer concurs that LC and high-resolution MS are regarded as the gold-standard thanks to the “rich” information such techniques offer on physiochemical properties of the protein. “Other alternative techniques are capillary electrophoresis (CE) and CE–MS because they can provide excellent resolution,” he adds.

“Spectroscopy, such as ultraviolet, Raman, near infrared, or fluorescence, are also widely applied to determine more specific requirements, such as concentration or purity. These techniques are fast and can even be applied inline,” Vollmer concludes. “Cell analysis technologies, such as metabolic analyzers, real-time cell analyzers, cell-imagers, and cell counters are applied to characterize suitable clones and host cells for protein manufacturing.”

References

1. Market Research Future, Protein Therapeutics Market Research Report: by Type, Application, Protein Function, End User—Forecast to 2027, Market Report (February 2021).
2. A. Bolje and S. Gobec, Pharmaceutics, 13 (4) 534 (2021).

About the Author

Felicity Thomas is the European editor for the Pharmaceutical Technology Group.

Article Details

Pharmaceutical Technology
Vol. 46, No. 3
March 2022
Pages: 36–39

Citation

When referring to this article, please cite it as F. Thomas, “Empowering Protein Characterization,” Pharmaceutical Technology 46 (3) (2022).