MAb Aggregation: A Problem Within the Biopharmaceutical Industry
.jpg)
K. H. Mok , Associate Professor, Trinity College Dublin, Ireland
Educated at Seoul National University, UC Berkeley, and Purdue University with postdoctoral research at the University of Oxford, K.H. Mok specializes in utilising biomolecular NMR spectroscopy to protein folding-misfolding-aggregation research. Co-author of publications in Nature, Nature Communications, Nature Immunology, PNAS and others, he currently serves as a Eureka Independent Evaluation Panel member, technical consultant to start-up companies LMNTIC Biotech and Meta-Flux, and editorial board member for three peer-reviewed journals.
The explosive growth of the global therapeutic antibody market has brought to the fore protein product aggregation as a problem which compromises efficacy, stability, and safety. Low pH, freezing, light exposure, lyophilization and ionic strength have all been identified to induce aggregation, and while excipients help in combating this, future inhibiting mechanisms will involve engineering of the molecular structures.
Introduction
Amongst the diverse biological macromolecules found in the cell, it would not be an exaggeration to say that proteins are the original "nano machines" of life. Approximately 20,000 in number, proteins are responsible for carrying out biochemical reactions (enzymes), transport (ion channels), intracellular signaling (receptors), protection (antibodies), storage (milk proteins), and cellular architecture (fibrous proteins), to name just a few.
Of the top ten blockbuster drugs in 2024, half of them were proteins, while the other half were peptides or small molecules designed to interact with proteins. All five proteins on this list (Keytruda®, Dupixent®, Skyrizi®, Darzalex®, and Stelara®; Combined sales of $77.4 billion) are monoclonal antibodies (MAbs), with pinpoint precision in binding other proteins or cells within our body.
We spoke with Professor Ken H. Mok, who has been characterising the structures and properties of proteins throughout his professional career.
1. What are your research interests and how have they evolved over time?
KHM: Beginning from my undergraduate years, I have always been intrigued with how proteins fold, also known as the ‘Protein Folding Problem’ (schematic figure below). Conceptually, the issue at hand was initially simple and straightforward: How is the three-dimensional, bioactive structure of a protein acquired from a nascent polypeptide chain released from the ribosome? Or metaphorically, how does information defined from a one-dimensional amino acid sequence attain three-dimensional biological context? From the 1960’s onto the 1980’s, the solution to this “Protein Folding Problem” was primarily thought to be of concern only to a small circle of biophysical chemists intellectually interested in this “esoteric game”. However, with the advent of genetic engineering, biotechnology and the birth of the biopharmaceutical industry - where overproduction of therapeutic proteins became a medical necessity and a commercial reality - it soon became, according to the journal Science in 1987, a “challenge that [could] no longer be ignored”.
Over a period of approximately three decades, scientists both experimental and theoretical have grappled with this complex subject, step-by-step, fragment-by-fragment, protein-by-protein. In industry, heterologously produced recombinant proteins often resulted in misfolded and hence biologically inactive inclusion bodies, and here, biochemical engineers stepped in to provide workable solutions. Throughout these endeavours, computational calculations had always served integral to deciphering the rules of folding. In the end, it turned out that the most efficient and holistic solution that encompassed all of the nuances and complexities was through the implementation of artificial intelligence (AI). Once considered to be one of the most intriguing and elusive scientific puzzles of life, this “win” for the computer was recognised in the 2024 Nobel Prize for Chemistry, with one half awarded to David Baker "for computational protein design", the other half jointly to Demis Hassabis and John Jumper "for protein structure prediction".
Figure: Protein Folding / Misfolding / Aggregation
Figure from Kitamura A, Nagata K, Kinjo M, Int J Mol Sci 16:6076-6092, 2015.
2. Why have the fundamentals in solving this problem been important in not only medicine but also for the biopharmaceutical industry?
KHM: Over the years, along with asking how a protein folds, it dawned upon scientists and clinicians that cellular proteins misfold and aggregate in vivo. This would not be of great consequence if the resulting clump of aggregated protein simply remained as harmless (but perhaps inconvenient) ballast. Rather, it turns out that these aggregates of proteins, many of which are identified in the brain such as amyloid beta and tau, α-synuclein, and huntingtin, display toxicity, induce inflammatory responses and showed strong correlations with neurodegenerative diseases such as Alzheimer’s, Parkinson’s, and Huntington’s disease, respectively.
To an individual, being afflicted with any of these diseases is tragic in itself, but imagine how detrimental it is when such diseases are transmissible. This is indeed the case for variant Creutzfeld-Jakob's disease (vCJD), where the misfolded prion protein originating from an animal suffering from bovine spongiform encephalopathy (BSE or mad cow disease) is unintentionally ingested by a human, transported to the individual's brain, and starts off a cascade of aggregation of the perfectly-functioning human prion protein.
While these neurodegenerative diseases were being studied in medicine and the neurosciences, the biopharmaceutical industry was experiencing unforeseen problems with its classic blockbuster protein therapeutics. FDA-approved drugs such as recombinant human erythropoietin (epoetin-alfa), insulin, interferon beta (IFNβ), and factor VIII (FVIII), previously administered to patients and generally well tolerated, began to show the formation of anti-drug antibodies (ADA). As a result, the importance of monitoring the immunogenicity of protein therapeutics post-approval warranted a closer look.
The common strand in all of these phenomena was the secondary event of misfolding and aggregation of previously-intact, functioning proteins.
3. With MAbs accounting for greater than 50% of all new approvals, and with over 750 FDA-approved biopharmaceutical agents, why is protein aggregation emerging as a problem in biotherapeutics?
KHM: The explosive growth of the global therapeutic antibody market has led to an interest in understanding the aggregation of protein products as it can compromise efficacy, concentration, and safety. Knowing that high concentration liquid formulations are necessary to achieve the therapeutic dose levels, various production and storage conditions such as low pH, freezing, light exposure, lyophilisation and increased ionic strength have all been identified as capable of inducing aggregation of polyclonal and monoclonal antibody therapeutics. The addition of stabilizing excipients has helped to combat the formation of aggregates, however future aggregation-inhibiting mechanisms involving engineering of the molecular structure are being explored as well.
For more than a decade, it has been known that different types of stress (such as oxidation, heat, or pH-shift) applied to MAb formulations result in increased ADA titers. One dramatic example where the clinical development itself was terminated involved the humanised anti-PCSK9 antibody bococizumab, which gave rise to ADAs in 48% of the patients. In this case, the propensity to aggregate with high poly-reactivity had been identified, and variants with improved biophysical properties and reduced immunogenicity are currently in development. Learning from this lesson, the introduction of point mutations and glycoengineering within aggregation prone regions (APRs) are being explored relatively early in the development process. We note that the amino acid sequence and glycosylation pattern modification themselves can potentially elicit immunogenicity, so further caution is required to take into account all outcomes.
4. How does your work translate to drug development?
KHM: Having focused on protein folding and protein misfolding problems during the past thirty-plus years, it is exciting to report that our work is further evolving to studying “Protein Alternative Folding”, which explores proteins and peptides that adopt multiple three-dimensional structures despite sharing identical amino acid sequences. Such proteins form a subset of Intrinsically Disordered Proteins (or Regions; IDP or IDR) which do not form any well-defined three-dimensional structure on its own, only to snap into a well-ordered structure upon binding with its biological molecular partner.
By using high-magnetic field biomolecular NMR spectroscopy, we have been able to provide insight to each area over the years, contributing to the body of knowledge both in breadth and depth. In addition, more specifically to alternative folding, I have been working with pioneering colleagues to develop a conformationally-fluctuating, protein-fatty acid complex (called Alpha1H) that has successfully concluded Phase I/II clinical trials against bladder cancer. It is even more gratifying to report that this candidate drug has received fast-track status by the US FDA.
