The study of protein chemical modification using polarized Excitation Emission Matrix (pEEM)

Abstract

The advances in drug development and the increased use of biologics as therapeutics drives the need for more effective Process Analytical Technology (PAT) tools to measure biophysical changes during manufacturing and formulation processes. This is particularly true for proteins undergoing chemical modification for the production of new entities such as Antibody drug conjugates (ADCs), PEGylated proteins or radioimmunoconjugates. The measurement of structural quality in solution can be challenging and time consuming, but intrinsic protein fluorescence (IPF) because of its high sensitivity, ease of use, and when implemented in via multi-dimensional techniques like polarized Excitation Emission (pEEM) spectroscopy, its high information content, might offer a solution. Conjugation reactions potentially exposes the protein to chemical and physical stressors such as the modification of specific amino acids and agitation, which can affect protein stability triggering unfolding, aggregation, and/or fragmentation, and potentially lead to reduction in process yields, loss of activity and/or immunogenicity issues. This, combined with the increased structural complexity of proteins and its conjugates, requires the use of more sensitive quality control (QC) tools to better identify and assess biophysical and structural changes during manufacturing. Here we demonstrate how pEEM measurements can be used in combination with simple chemometric tools like Principal Component Analysis (PCA) and partial least squares (PLS) to discriminate the raw product solutions according to the degree of PEGylation and also to predict conjugation degree (PEG to protein ratio, PPR) with good accuracy (root mean square error, RMSE for calibration ∼10%, relative error of prediction, REP< 20%), in comparison to the reference technique Size Exclusion Chromatography (SEC) (SEC error: ∼7.2%). The use of variable selection tools allows for similar predictions obtained with faster (from ~7min to less than 60s) and simpler two-dimensional spectra, which makes the method a more viable PAT tool for in or online measurements. Furthermore, pEEM offers a reproducible and fast alternative to simultaneously measure protein concentration (RMSE<0.01 g/L), asses structural variance, and particle/aggregate content. It allows one to generate quantitative vii prediction models for non-reversible aggregation content as described by SEC, and obtain qualitative information about reversible aggregate content, which cannot be obtained from SEC Finally, we show the feasibility of using pEEM for assessing and monitoring structural changes during the reaction. By using IgG PEGylation as a model system, we show that the three levels of information extracted from pEEM can be used for in-situ assessment of conjugation reactions: Rayleigh scatter (RS) to probe aggregate/particle formation, both transient and permanent; intrinsic fluorescence emission to identify subtle chemical and structural variations in the parent protein, and aniso-EEM maps that corroborated protein structural changes.