High-throughput analysis of sub-visible mAb aggregate particles using automated fluorescence microscopy imaging

Abstract

Aggregation of therapeutic proteins is a major concern as aggregates lower the yield and can impact the efficacy of the drug as well as the patient's safety. It can occur in all production stages; thus, it is essential to perform a detailed analysis for protein aggregates. Several methods such as size exclusion high-performance liquid chromatography (SE-HPLC), light scattering, turbidity, light obscuration, and microscopy-based approaches are used to analyze aggregates. None of these methods allows determination of all types of higher molecular weight (HMW) species due to a limited size range. Furthermore, quantification and specification of different HMW species are often not possible. Moreover, automation is a perspective challenge coming up with automated robotic laboratory systems. Hence, there is a need for a fast, high-throughput-compatible method, which can detect a broad size range and enable quantification and classification. We describe a novel approach for the detection of aggregates in the size range 1 to 1000 μm combining fluorescent dyes for protein aggregate labelling and automated fluorescence microscope imaging (aFMI). After appropriate selection of the dye and method optimization, our method enabled us to detect various types of HMW species of monoclonal antibodies (mAbs). Using 10 μmol L 4,4'-dianilino-1,1'-binaphthyl-5,5'-disulfonate (Bis-ANS) in combination with aFMI allowed the analysis of mAb aggregates induced by different stresses occurring during downstream processing, storage, and administration. Validation of our results was performed by SE-HPLC, UV-Vis spectroscopy, and dynamic light scattering. With this new approach, we could not only reliably detect different HMW species but also quantify and classify them in an automated approach. Our method achieves high-throughput requirements and the selection of various fluorescent dyes enables a broad range of applications.

Advanced Characterization of Silicone Oil Droplets in Protein Therapeutics Using Artificial Intelligence Analysis of Imaging Flow Cytometry Data

“…The Amnis® AI model developed herein classified protein and silicone oil droplets >2 mm with high accuracy in the absence of fluorescence label, however fluorescence labeling was required for identification of protein-adsorbed silicone. Fluorescence labeling has been used for detection by flow cytometry 41,42 and fluorescence microscopy, 43,44 and is routinely used in protein stability screening with the ThermoFluor assay. 45 Fluorescence labeling remains an important concern within the pharmaceutical formulation community as previous reports have indicated that fluorescence probes may affect the stability of the protein.…”

Identification of process conditions influencing protein aggregation in Chinese hamster ovary cell culture

“…Automated FMI (Paul et al, 2017) was performed on a NyONE fluorescence microscope (SynenTec, Elmshorn, Germany). Cell concentration and mAb particle formation were analyzed using image processing and automatic analysis with the NyONE customer software for suspension cell count and Nuclei counting (EMGU2 Nuclei Count), respectively.…”

“…Using our recently developed method for high-throughput analysis of soluble aggregates (Paul et al, 2015) and a method to quantify mAb aggregate particles (Paul et al, 2017), we investigated the impact of the process relevant factors temperature, osmolality, agitation, antifoam, and VPA on the cell concentration and formation of mAb aggregates in mAb-producing CHO cell cultures. Using our recently developed method for high-throughput analysis of soluble aggregates (Paul et al, 2015) and a method to quantify mAb aggregate particles (Paul et al, 2017), we investigated the impact of the process relevant factors temperature, osmolality, agitation, antifoam, and VPA on the cell concentration and formation of mAb aggregates in mAb-producing CHO cell cultures.…”

“…In this work we applied DoE methodology to screen for process parameters and cell culture additives influencing mAb aggregate formation in CHO cell culture, in order to find conditions to reduce mAb aggregation during cell culture. Using our recently developed method for high-throughput analysis of soluble aggregates (Paul et al, 2015) and a method to quantify mAb aggregate particles (Paul et al, 2017), we investigated the impact of the process relevant factors temperature, osmolality, agitation, antifoam, and VPA on the cell concentration and formation of mAb aggregates in mAb-producing CHO cell cultures. Host cell protein aggregation was investigated by analysis of a non-producing mock CHO cell line.…”

Stress Factors in mAb Drug Substance Production Processes: Critical Assessment of Impact on Product Quality and Control Strategy

“…Additional complexity arises due to difficulty of determining the conformation and aggregation state of the protein in culture broth. Formation of large aggregate particle was investigated by Paul et al 30,36 to screen for process parameters and cell culture additives that influence mAb aggregate formation in CHO cell culture. They 33 established an experimental protocol that allows analyzing mAb aggregate formation using SEC, directly in the supernatant of CHO cell culture without the influence of prepurification steps.…”