A Novel Data Processing Strategy for Detection of Low-Abundance HCPs: Increased Sensitivity and Accuracy with Fewer False-Positive Identifications
September 14, 2020
Because their presence introduces foreign proteins that may trigger an immune response, host cell proteins (HCPs) present a potential safety risk in biopharmaceutical products. Currently, the most commonly used techniques for characterizing HCPs are enzyme-linked immunosorbent assay (ELISA) and LC-MS/MS, which is an orthogonal method that provides information complementary to ELISA analyses. However, while being able to identify individual proteins, current MS-based methods for HCP detection typically generate large numbers of false-positive identifications which must be validated by expert MS users. We present a novel approach to identifying and quantifying HCPs in biopharmaceuticals that significantly reduces the number of reported false-positive identifications—and consequently analysis time–by leveraging data sets from control sample replicates containing high levels of HCPs.
In our two-stage method, qualitative information from upstream samples containing high concentrations of HCPs is leveraged to increase sensitivity and accuracy of protein identification in downstream samples. First, data sets from control samples containing high concentrations of HCPs are aligned to correct retention time shifts. MS signals and MS/MS spectra originating from the same species are consolidated across the samples to improve data quality. Finally, the consolidated spectral data are used to create a bioprocess-specific HCP library containing the three most abundant (“Hi3”) unmodified peptides. Second, target in-process samples are analyzed using the HCP library generated in the first stage in an automated data processing workflow.