Charge heterogeneity of therapeutic monoclonal antibodies by different cIEF systems: views on the current situation

ABSTRACT

Therapeutic mAbs show a specific “charge fingerprint” that may affect safety and efficacy, and, as such, it is often identified as a critical quality attribute (CQA). Capillary iso-electric focusing (cIEF), commonly used for the evaluation of such CQA, provides an analytical tool to investigate mAb purity and identity across the product lifecycle. Here, we discuss the results of an analysis of a panel of antibody products by conventional and whole-column imaging cIEF systems performed as part of European Pharmacopoeia activities related to development of “horizontal standards” for the quality control of monoclonal antibodies (mAbs). The study aimed at designing and verifying an independent and transversal cIEF procedure for the reliable analysis of mAbs charge variants. Despite the use of comparable experimental conditions, discrepancies in the charge profile and measured isoelectric points emerged between the two cIEF systems. These data suggest that the results are method-dependent rather than absolute, an aspect known to experts in the field and pharmaceutical industry, but not suitably documented in the literature. Critical implications from analytical and regulatory perspectives, are herein thoughtfully discussed, with a special focus on the context of market surveillance and identification of falsified medicines.

KEYWORDS:

• Charge heterogeneity profile
• cIEF
• identity
• isoelectric point
• purity
• quality control
• therapeutic mabs

MAb analysis by cIEF technique

In the past decade, the number and types of therapeutic monoclonal antibodies (mAbs) approved for marketing have grown substantially. Notably, with the rise of mAb biosimilars, these products have dominated the biopharmaceutical field. Currently, more than hundred mAbs have been approved by the European Medicines Agency (EMA) and the U.S. Food and Drug Administration for the treatment of a broad range of diseases, such as cancers, infections, and autoimmune disorders^1–3 (Figure 1(a)). In 2022, EMA approved 20 new mAbs, including biosimilars, bispecific mAbs, and antibody-drug conjugates (ADCs).⁴

Figure 1. (a) The number (cumulative frequency) of approved monoclonal antibodies and their biosimilars increased over the last decades. (b) The charge variant profile is a critical quality attribute (i.e., potentially affecting the safety and efficacy profile) evaluated for molecular characterization and monitored to evaluate identity and purity of mAb – based medicinal products along their lifecycle. A list of different analytical methods currently used to analyze charge heterogeneity is reported.

The inherent structural and functional complexity of mAbs has posed analytical challenges for manufacturers and regulators, both committed to ensuring the quality, efficacy, and safety of the medicinal products throughout their entire life cycle. Due to their protein nature, and depending on the manufacturing process, mAbs can undergo various post-translational modifications (PTMs), degradation, and conformational alterations with potentially critical effects on the product efficacy and safety.⁵

As a result of PTMs, each therapeutic mAb shows a highly specific charge heterogeneity profile, which can be described as its specific “fingerprint”. Charge heterogeneity is a critical quality attribute (CQA) of mAbs because charged variants can affect bioavailability, tissue distribution, and pharmacokinetics of this product class.^5–7 In the charge heterogeneity profile, each charge variant is positioned according to its specific isoelectric point (pI), defined as the pH value where the net charge of the molecule is zero. The pI of a molecule depends on its primary, secondary, and tertiary structure, but also on PTMs, and should therefore be an intrinsic biophysical property of a molecule.

One of the most sensitive techniques for the evaluation of mAb charge heterogeneity is capillary iso-electric focusing (cIEF),^8,9 which provides an important tool to investigate the identity, purity, and stability of a mAb and allows product consistency to be monitored during the manufacturing process, covering the whole lifecycle of the product¹⁰ (Figure 1(b)). mAbs contain ionizable acidic and basic groups that can result in a net positive or negative charge depending on the interactions with the surrounding environment. This enables separation of different charged variants during cIEF analysis under the influence of an electric field. In order to determine precise numerical values for the pI of different charged variants, the analysis is carried out by exploiting a pH gradient created from a mixture of carrier ampholytes and a set of pI markers with known pI values as calibrators, usually spiked into the sample. During the analysis, the pI markers and analytes migrate together with the ampholytes to the position in the pH gradient where the net charge and mobility of the molecules is zero, reflecting the pI.¹¹ Importantly, the pI value ‘measured’ by means of currently available techniques is defined as ‘apparent’ pI, because it is recognized that the experimental value does not always match the theoretical one, typically calculated on the basis of the primary sequence of the molecule. The main component in the resulting electropherogram is usually defined as the main peak, whereas charged variants migrating at lower or higher pI values are conventionally called ‘acidic’ or ‘basic’ species, respectively. Typically, the acidic variants contain various degrees of sialylation, Asn deamidation, and glycation, while the basic species are mainly characterized by non-cyclic N-terminal Gln, C-terminal lysine, and C-terminal amidation.^5,12

Challenges of establishing a cIEF procedure universally applicable to conventional and imaged-cIEF systems

Apart from the old and more laborious iso-electric focusing methods based on slab gels, the most widespread systems currently on the market are based on the separation of the analytes inside capillaries to ensure greater resolving power and can be essentially classified into two main categories, traditional cIEF and imaged-cIEF (icIEF).

In conventional cIEF, the formation of the pH gradient and separation of the analytes according to their pI (focusing) and detection (mobilization) are two distinct steps. In the mobilization step, the stable pH gradient is moved through the capillary from the anode to the cathode past the detection window (UV-detection at 280 nm), either by chemical or pressure mobilization. In this setup, basic species (high pI) pass through the detection window first, appearing in the left part of the electropherogram, when plotted against their migration time.^13,14 On the icIEF apparatus, the focusing step is monitored online by a charge-coupled device (CCD) camera extending along the entire length of the capillary, allowing direct detection without mobilization. In this case, the pI of the separated variants is plotted against their position (in pixels) along the capillary, from the cathode to the anode. Thus, in contrast to cIEF, the basic species appear in the right part of the electropherogram.¹⁵ Depending on the type of icIEF instrument, the capillary, which is typically pre-assembled in a ready-to-use cartridge, can have a vertical or horizontal orientation. It has been demonstrated that these differences do not affect the separation efficiency, so that the different icIEF instruments currently available provide comparable results.^16,17 In both systems, pI values of the sample are determined after calibration with pI-markers, assuming a linear dependence throughout the pH gradient.

cIEF and icIEF are not currently considered interchangeable, even though they are based on the same principle for the separation of the analytes. Recently, regulatory agencies and the Official Medicines Control Laboratory network have aimed to standardize analytical procedures, which would simplify control activities by making them less dependent on specific equipment. In the study discussed here, the aim was to rationalize the methodology for mAb charge heterogeneity analysis. An experimental study was carried out in order to establish a common, transversal analytical procedure, applicable to both conventional and icIEF systems. In this study, capillary electrophoresis systems from two manufacturers were used: a conventional cIEF apparatus from Sciex (PA800+) and a icIEF-instrument from BioTechne-ProteinSimple (Maurice). Both cIEF and icIEF systems use the same physico-chemical principle for the separation and identification of charged variants, while the detection of the analytes differs (according to the presence or absence of a mobilization phase), along with other specific technical parameters (e.g., separation gel or capillary length).¹⁸

The study consisted of testing a panel of representative and structurally different mAbs by means of both systems using protocols based on strictly similar experimental conditions. In brief, the procedures shared the same sample concentration and pre-treatment, the same pH gradient in terms of pharmalytes ratio and composition, as well as the same anodic and cathodic stabilizers. Infliximab was used as an internal control in the analysis for monitoring the method’s performance. Predefined system suitability criteria and further guidance were implemented to address the reliability and reproducibility of the results. In addition, an adaptation to specific requirements for the two platforms was made (see Supplementary information for detail on technical aspects and the experimental design).

However, despite the attempts to optimize the experimental conditions, the results differed, in some cases markedly, between the conventional cIEF and icIEF, in terms of pI values, electropherogram profile and resolution, as depicted in Figure 2. Collectively, these data reveal a significant analytical inconsistency when comparing results obtained from the two systems. Despite further efforts to establish a single common procedure (e.g., by exploiting different focusing times), we finally concluded that use of separate, system-specific procedure is required to capture the optimal approach with the best resolution and sensitivity (unpublished data).

Figure 2. Systematic inconsistencies for measured pI values and charge heterogeneity profiles have been observed in conventional vs imaged cIEF, for a panel of different mAbs, spanning a range between pH 6.8–10. The same panel of therapeutic mAbs was analyzed under comparable experimental conditions with two different devices: PA800±AB (cIEF, sciex) and Maurice™-(icIEF, Biotechne/Protein simple). The electropherograms of four representative mAbs (here numbered from 1 to 4) obtained on the two systems according to the respective calibration curves, were scaled on the same pH range reporting normalized absorbance signal versus pI, to make the results comparable, on the same Y-scale interval. In the table on the left, the averages of the main peak pI values from both systems together with the associated RDS% are also shown. The data presented were obtained from four European Official Medicines Control Laboratory sites, three of which were equipped with conventional cIEF and one with icIEF.

Do the cIEF systems affect mAbs charge analysis? Reflections and implications

The validation of a particular analytical procedure using a specific tool is the general practice in industry, and it is commonly accepted that cIEF and icIEF systems typically fail to demonstrate the same output. Despite this, to our knowledge there is no experimental evidence on this subject in the literature. We would also like to point out that, while inconsistent results are largely expected from analytical approaches based on fully different mechanisms (such as ion exchange chromatography and capillary zone electrophoresis), they are not expected from systems using the same chemical principle, once the conditions are made comparable. Therefore, we strongly believe the results deserve reflection shared by all stakeholders (industry and regulators) involved in the quality control of mAbs. Notably, the study highlights the importance of understanding factors affecting the results, prompting further investigations. In the context of the primary characterization of a molecule using cIEF-techniques, the question arises as to if the experimentally determined pI values of charged variants should be considered as a univocal biophysical property of a biomolecule, or, rather, a rough estimate, which, to a great extent, depends on the experimental conditions (Figure 1(b)). Furthermore, another question pertains as to whether the differences in the absolute pI values impair the suitability of the method to serve its intended use, such as to provide support for control activities.

It is widely recognized that the charge heterogeneity profiles of a mAb can differ with respect to the number of charged variants as well as the measured pI values, when determined with different orthogonal methods, such as gel-based IEF, ion exchange chromatography, or cIEF¹² (Figure 1(b)). This is mainly based on the fundamental differences in the mode of separation. Further, the differences may be caused by additional effects such as column interactions, which can lead to separation of variants sharing the same charge distribution (same pI).^19,20 Our findings suggest that similar phenomena can occur even when charged variants of mAbs are separated by two capillary electrophoretic methods based on the same principles, such as cIEF and icIEF. It is likely that the intrinsic differences, such as the use of different separation gels, capillaries and the presence or absence of the mobilization step, will result in the formation of a different microenvironment in the two systems, which may in turn have an impact on the overall charge distribution and the measured pI values of the different variants.

Again, the problem may be related to the accepted practice of using only two or three pI markers as internal calibrators, which makes the methods susceptible to divergences in pI estimations when small deviations from the pre-assumed linearity along the pH gradient occur. In this case, the approach currently used to calibrate cIEF systems should be critically revised, introducing more stringent system suitability criteria to exclude this possibility.

At present, it does not seem feasible to directly compare the theoretical pI values with those measured by cIEF or icIEF methods.

As a general note, cIEF methods for controlling charge heterogeneity of mAbs are usually developed early in the product lifecycle, and validation shows the suitability to ensure product consistency within defined limits. The connection between product consistency and the efficacy and safety of the product is a major point that is critically assessed by regulators during marketing authorization application. The profile of charged variants, established with a specific method, is a CQA of the product and needs to be carefully controlled. However, the relevance of absolute pI values has to be questioned in this context. Accordingly, the pI value is not typically presented in the release specification, even if it may sometimes appear as an acceptance criterion for the method’s system suitability (for example, in terms of differences between the test sample and the reference standard).

The observed differences between the cIEF and icIEF methods in determining the charge heterogeneity pattern of a mAb clearly show that it is not easy to substitute one method with another since identical experimental conditions do not lead to identical results in the two systems. The discrepancy could lead to false interpretation of the quality of the product, along with the associated risks.

The finding that apparent pI values might differ depending on the system used has notable implications, both at an industrial and regulatory level. For monitoring the quality of a product, it does not matter which tool is used, as long as it is used consistently throughout the mAb development and lifecycle, and it demonstrates a consistent charge distribution including the apparent pI of the isoforms. However, the manufacturer must then use the same method throughout the life of a product, unless the equivalence of a new application is convincingly demonstrated. Importantly, this interdependence also has repercussions at the regulatory level, as it is unlikely that an official control laboratory will always have the same cIEF method/equipment as the manufacturer, which bears the risk of misleading analyses.

The above implications are amplified in the context of market surveillance against falsified medicinal products, where standards or reference products are not always readily available. In this case, screening by means of a general ‘system-independent’ cIEF procedure could be greatly useful to unmask the true entity of a fraudulent product, making the assessment of product identity fast and unambiguous. One solution to this is the use of a reference standard, which allows the identification of the charge variant profile by a direct comparison with the respective reference electropherogram within the same IEF method, validated by the control laboratory. However, this path is not always feasible.

Importantly, the development of standardized analytical methods not constrained to a particular system would strengthen the independence of control laboratories in all related control activities.

We encourage a critical review of cIEF-based procedures, proposing studies and measures aimed at a better understanding of the phenomena underlying the discrepancies outlined above. Such findings might lead to further development and standardization of cIEF and icIEF methods, resulting in truly comparable procedures. Ultimately, this would enable the harmonization of results from different cIEF-based methods, an important step for the comparison of absolute pI values of charged variants and toward reaching the goal of making mAb charge profiles more comparable across different cIEF-methodologies.

Abbreviations

ADC, CCD, cIEF, CQA, EMA, FDA, icIEF, mAb, pI, PTMs, OMCL.

Acknowledgments

The authors are grateful to the EDQM for having organized the collection of materials for this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19420862.2024.2313737

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.