Analyzing Protein Aggregation in Biopharmaceuticals

LAGUNA DESIGN/Getty Images

Biopharmaceutical manufacturers are under increasing pressure from regulators to ensure the safety and quality of their products. Protein aggregation presents a key challenge in the development of biologic formulations as it can have an impact on product quality in terms of efficacy and immunogenicity. Matthew Brown, PhD, product technical specialist, Life Sciences, Malvern Instruments spoke to BioPharm International about the causes and risks of protein aggregation and discussed the analytical capabilities available to measure and characterize protein aggregates.

Causes of protein aggregation

BioPharm: What are the causes of protein aggregation in biologic formulations?

Brown: Proteins have a natural propensity to aggregate due to the dynamic nature of their structure, which is held together by a combination of Van der Waals forces, hydrogen bonds, disulfide linkages, and hydrophobic interactions. Disruption of this delicate balance can expose internal hydrophobic regions of the polypeptide chain, which may then interact with areas on other proteins to form larger complexes of misfolded proteins. This aggregation can be ‘native,’ in which the protein structure is maintained and the aggregation is largely reversible, or ‘non-native,’ where denaturation and structural changes mean this effect is largely irreversible. Aggregates may continue to grow and form over a wide size range, including up to and beyond the formation of visible particles, and ultimately this leads to precipitation.

Protein aggregation is a common consequence of many sample treatments and is a major problem facing the biopharmaceutical industry. These treatments include the addition of chemicals, incorrect reconstitution of lyophilized materials, the effect of mechanical stresses encountered during manufacturing, freeze/thaw cycles, and prolonged storage. The presence of contaminating particles is also known to promote aggregation, with materials such as silicone oil, a commonly used lubricant in pre-filled syringes, acting as a nucleation site for aggregation. Other examples of contaminating materials include silicone rubber from container-closure systems, glass particles from vials, and oxidized metal particles from sources such as filling lines. There is the potential for aggregation to occur at almost every stage of a biopharmaceutical process, such as in development, formulation, manufacturing, storage, and at the point of use, and it may have a number of deleterious consequences.

What is fundamentally important is to understand, at an early stage, the pathway of aggregate formation in order to put in place processes that will help to minimize it. The aggregation behavior of two proteins in a similar process may be quite different.

Risk of protein aggregation

BioPharm: In terms of safety and efficacy, what are the risks of protein aggregation?

Brown: While biological molecules can degrade in a number of ways that make their development as therapeutic agents challenging, the presence of aggregates in biopharmaceutical formulations remains one of the major quality and safety concerns. Not all aggregates will lose the functionality of the original protein, but as aggregation proceeds, the activity of the constituent protein molecules may well diminish or be lost from the therapeutic, thereby reducing its efficacy. Of even greater concern is that numerous studies have shown that the presence of protein aggregates in a therapeutic formulation destined for parenteral administration may trigger an unwanted and/or potentially dangerous immune response in the recipient.

Efficacy can be affected in a number of different ways, ranging from no impact through to rendering a drug completely ineffective. For example, large aggregates developing during the formulation process may well be filtered out at the later stages of production, which will also reduce the concentration of active molecule in the final product. While the most serious safety issue is the potential to trigger a life-threatening immune response that leads to anaphylactic shock, less devastating immune responses also have serious consequences. For example, the efficacy of a biopharmaceutical may be compromised if an immune response in the patient leads to elimination of the therapeutic protein, resulting in ineffective treatment perhaps where no other alternative treatments exist. Equally important is the administration of proteins designed to supplement levels of a naturally occurring endogenous protein. The triggering of an immune response here can lead not only to destruction of the therapeutic itself, but may also induce an immune response against the intrinsic protein, potentially leaving the patient with additional clinical complications. The case of Eprex-associated Pure Red Cell Aplasia (PRCA) is a well-documented example of such an effect.

The FDA Guidance for Industry: Immunogenicity Assessment for Therapeutic Protein Products, issued in August 2014 (1), provides useful summary descriptions of the clinical consequences of immune responses to therapeutic proteins. Reduced efficacy and increased immunogenicity of a drug product are both highly undesirable, so understanding and monitoring protein aggregation in drug formulations is crucial.

Measuring and characterizing protein aggregates

BioPharm: Can you describe the challenges in measuring and characterizing protein aggregates? What guidelines have FDA or EMA provided?

Brown: The FDA Guidance for Industry: Immunogenicity Assessment for Therapeutic Protein Products (1) is ‘intended to assist manufacturers and clinical investigators involved in the development of therapeutic protein products for human use.’ It is wide-ranging in its scope and sets out FDA’s current thinking and recommendations, but is not a statutory requirement.

In this guidance, aggregates are defined as any self-associated protein species, with a monomer defined as the smallest naturally occurring and/or functional subunit. The text points to the criticality of minimizing protein aggregation in therapeutic products and the need to develop minimization strategies as early as is feasible in product development. It states that, ‘methods that individually or in combination enhance detection of protein aggregates should be employed to characterize distinct species of aggregates in a product.’ Recommendation is made that the range and levels of subvisible particles (2–10 microns) present in therapeutic protein products, initially and over the course of the shelflife, should be assessed. The guidance goes on to indicate that as more analytical methods become available, there should be a move to characterize particles in smaller (0.1–2 microns) size ranges. Furthermore, there should be risk assessment of the impact of these particles on the clinical performance of the therapeutic protein product, with the development of control and mitigation strategies based on that assessment.

Subvisible particles are usually defined as particles that are not visible to the naked eye and have a size of <100 µm. They can further be defined into micron (1–100 µm) and sub-micron (1 µm) size ranges. The United States Pharmacopeia (USP) chapter <788> relating to particulate matter in injections requires the quantification of subvisible particles that are ≥ 10 µm and ≥ 25 µm in size, usually using light obscuration and flow imaging techniques (2). Meanwhile smaller aggregates (<0.1µm), typically caused by oligomerization, are characterized using size exclusion chromatography (SEC). However, for particles in the 0.1 to 10 µm size range cited in the FDA guidance, very few characterization techniques can provide quantitative sizing details, and there is no single instrument that can cover the full measurement range required for sub-visible particles. Consequently, different analytical approaches must be applied, each with their own distinct sizing range. This poses the challenge of relating results from quite different measurement technologies and raising the need to use truly orthogonal systems that deliver independently achieved data on the same parameter. The use of orthogonal approaches to the characterization of biotherapeutics is strongly encouraged by regulatory authorities. This is to ensure a more detailed and rounded view of the product is obtained, and to prevent reliance on single technologies. Each and every technique used to characterize the complex nature of proteins will have its own limitations; therefore, combining multiple technologies improves understanding of products.

BioPharm: What methods do you use to measure and characterize protein aggregates and how do they compare to each other?

Brown: A variety of techniques is available to characterize protein aggregates, which while helpful individually, collectively can offer even more valuable insight into the behavior of biotherapeutics. A number of widely used technologies are described in the following (see Figure 1).

Figure 1: Techniques to characterize protein aggregates. DLS is dynamic light scattering. AUC is analytical ultracentrifugation. NTA is nanoparticle tracking analysis. RMM is resonant mass measurement. SEC/GPC is size exclusion chromatography/gel permeation chromatography. LO is light obscuration.

Figure 1 is courtesy of Malvern Instruments.

SEC separates proteins on the basis of size as they pass through chromatography columns. Consequently, SEC allows characterization of a protein’s oligomeric state and is normally employed as a quality control (QC) release assay. The addition of multiple detectors (such as in the Viscotek TDAmax triple detection system [Malvern Instruments]) enables more extensive characterization, including the determination of molecular weight. However, large aggregates, typically above 200 nm, will block the column, and hence, will not be detected. Consequently, SEC will only provide a picture of the smallest aggregates present.

Dynamic light scattering (DLS) systems (such as the Zetasizer Nano [Malvern Instruments]) are widely used throughout the lifecycle of biopharmaceuticals to measure the size and size distribution of proteins in solution. DLS is an easy-to-use, rapid and non-invasive method with incredibly high sensitivity to the presence of large particles, allowing the detection of aggregates at the earliest stages of onset. Consequently, DLS is often used as a screening tool and in comparability studies to monitor the oligomeric state of the protein of interest. However, DLS provides qualitative data only, and will not provide particle quantification or concentration.

Nanoparticle tracking analysis (NTA) uses a high-resolution digital camera and specially designed software to track the movement of particles under a microscope, thereby tracking Brownian motion and determining hydrodynamic size. It generates a high-resolution particle size distribution by sizing each particle individually and also measures the concentration of particles present in the sample. Due to the low refractive index of protein, the lower limit of detection in NTA measurement is approximately 30 nm diameter. So while protein monomer units are not measurable by NTA, aggregates comprised of just a few tens of monomers through to many thousands of units can be sized and counted to determine particle concentration.

A recent addition to the toolkit is resonant mass measurement (RMM) in the form of the Archimedes system (Malvern Instruments). It uses RMM to detect and accurately count particles in the critically important size range 50 nm–5 µm, and to reliably measure their buoyant mass, dry mass, and size. Furthermore, it can also distinguish between proteinaceous material and contaminants, such as silicone oil, by means of comparing their relative resonant frequencies and buoyant masses.

Recent advances in analytical technologies

BioPharm: What recent advances have you seen in analytical technologies for measuring and characterizing protein aggregates? And what area is still lacking?

Brown: There is a growing need to be able to quantify aggregates within the sub-micron size range. Some of the latest technology identified in this article now provides the industry with this capability. This ability is becoming more important as companies are required to assess immunogenic risk for parenterals. In addition to detecting and characterizing aggregates, huge importance is now being placed on the identification of particulates and contaminants. Particles detected in a product may not be protein aggregates, but rather contaminants from manufacturing processes and product contact surfaces. New technologies, such as RMM and morphologically-directed Raman spectroscopy (MDRS), are now providing a means by which particulate matter can be distinguished from protein aggregates, thereby greatly facilitating troubleshooting and deviation resolution.

One of the newest analytical systems for protein characterization (Zetasizer Helix [Malvern Instruments]) combines industry-leading DLS technology for high sensitivity aggregate sizing with Raman spectroscopy, which allows monitoring of changes in secondary and tertiary protein structure. The combination of DLS and Raman spectroscopy enables measurement of protein size and structure from a single small-volume sample, providing unique insights into protein folding, unfolding, aggregation, agglomeration, and oligomerization. Ultimately, this can lead to identification of the degradation pathways that result in the formation of aggregates and identification of high-risk processes and parameters. Such detailed information supports both the effective application of quality by design (QbD) and the efficient development of biosimilars.

In conclusion, protein aggregation is a consequence, rather than a cause, of degradation. As discussed at the outset, there is a pressing need to understand aggregation pathways and the risk factors that can induce aggregation right from the start of the drug development process in order to devise mitigation strategies and to align manufacturing processes with QbD principles. Given the importance of aggregates and particles to immunogenicity, it is likely that in the future particle quantification and characterization methods will need to be implemented more readily in GMP-compliant environments and in support of QC activities. Indeed, particle counting is often far more sensitive than methods that measure loss of protein monomer, as the protein mass in sub-visible particles is usually very low, relative to the total protein content. However, this understanding must also extend to the manufacturing process and beyond. Not only is it necessary to identify those processing steps most likely to introduce particles and the types of particles involved, but consideration also has to be given to product handling post release, which includes storage and handling in a clinical setting.

References

1. FDA, Guidance for Industry: Immunogenicity Assessment for Therapeutic Protein Products (Rockville, MD, August 2014).

2. USP General Chapter <788>, “Particulate Matter in Injections” (US Pharmacopeial Convention, Rockville, MD, 2012). 

Matthew Brown, PhD, product technical specialist, Life Sciences, Malvern Instruments, Tel: +44 (0)1684 892456, matthew.brown@malvern.com.

Article DetailsBioPharm International
Vol. 28, Issue 1
Pages: 40–43
Citation: When referring to this article, please cite as A. Siew, “Analyzing Protein Aggregation in Biopharmaceuticals,” BioPharm International 28 (1) 2015.