Label-free Quantification of Host-Cell Protein Impurity in a Recombinant Hemoglobin Reference Material

Abstract

AbstractQuantitative analysis depends on pure-substance primary calibrators with known mass fractions of impurity. Here, label-free quantification (LFQ) is being evaluated as a readily available, reliable method for determining the mass fraction of host-cell proteins (HCPs) in bioengineered proteins. For example, hemoglobin-A2 (HbA2) is being used as obtained through overexpression inE.coli.Two different materials had been produced: natural, and U-15N-labeled HbA2. For quantification of impurity, precursorion (MSl-) intensities were integrated over allE.coli-proteins identified, and divided by the intensities obtained for HbA2. This ratio was calibrated against the corresponding results forE.coli-cell lysate, which had been spiked at known mass-ratios to pure HbA2. To demonstrate the universal applicability of LFQ, further proteomes (yeast and human K562) were then alternatively used for calibration and found to produce comparable results. Valid results could also be obtained when the complexity of the calibrator is reduced to a mix of nine proteins, and a minimum of five proteins is estimated to be sufficient to keep the sampling error below l5%. For the studied materials, HbA2-mass fractions of 916±15 mg/g and 922±11 mg/g were found. Value assignment by LFQ thus contributes 1-2% to the overall uncertainty of HbA2-quantification when these materials are used as calibrators. Further purification of the natural HbA2yielded 999.1± 0.15 mg/g, corresponding to ≈ 0.2% of uncertainty contribution, though at a significant loss of material. If an overall-uncertainty of 5% is acceptable for protein-quantification, working with the original materials would definitely be viable, therefore.