Loading...
Development of a Novel Algorithm for Rapid HX-MS Data Analysis and Investigation of Antigen Induced Allosteric Effects on Monoclonal Antibodies by HX-MS
SHI, YUQI
SHI, YUQI
Citations
Altmetric:
Abstract
Hydrogen exchange-mass spectrometry (HX-MS) has been widely adopted for probing higher order structure and dynamic of proteins in various fields ranging from academia research to biopharmaceuticals. HX-MS is considered as a medium resolution approach due to its ability to localize structural information of a protein at peptide level. As a consequence, HX-MS plays an important role in a variety of applications, such as assessing the comparability and stability of drug molecules, conducting equivalence analysis for biosimilars, performing epitope mapping, etc. Compared to rapid experimental and instrument development, informatics of HX-MS is still lagging behind. Conventional "feature extraction" approach impede wider adoption of this application. Described herein, are projects designed to, 1. develop a new algorithm for hydrogen exchange-mass spectrometry data analysis with less manual interpretation; 2. investigate allosteric Fc response in neutralizing antibodies upon antigen binding; 3. develop a strategy to evaluate the accuracy of peptide assignment for established peptide mapping data interpretation criteria.We developed a novel algorithm for rapid and simplified HX-MS data analysis. HX-MS raw data was compressed by inducing sparsity using LASSO regression algorithm. The algorithm was validated in terms of accuracy, precision and reproducibility using experimental data collect on maltose-binding protein. The algorithm accomplishes two goals: firstly, it is able to extract individual feature with less manual interpretation; secondly, it can generate structural fingerprint of proteins under different conditions for global comparability analysis. We established an in-solution bottom-up HX-MS experiment to investigate allosteric alterations on Fc region of the antibody upon antigen binding. Using one antigen to target three neutralizing antibodies, structural changes on the Fc region of three were investigated separately. Apart from changes in the CDR binding regions, statistical analysis of HX-MS results showed structural alternation in both Fab constant regions and Fc regions. Changes identified in three antibodies share similarities and differences. More interestingly, a portion of the structural changes overlapped with Fc receptor binding sites and glycosylation site. Our work demonstrates not only allosteric signaling to effector and Fc receptor binding but also that even within neutralizing antibodies recognizing different epitopes within the same antigen, that there are interesting differences in the allosteric response which must depend on how antigen engages with the paratope. Bottom-up HX-MS experiment interprets data at peptide level, therefore, it is important to establish a peptide mapping data interpretation criteria for peptide assignment. We developed a strategy that can determine the score cut-off for peptide assignment. By defining and matching hits with a library containing multiple decoys, false assignment rate was determined at different score cut-off. With NIST mAb as the study model, we demonstrated a benchmarking approach for peptide assignment criteria assessment.
Description
Date
2012-12-31
Journal Title
Journal ISSN
Volume Title
Publisher
University of Kansas
Collections
Research Projects
Organizational Units
Journal Issue
Keywords
Chemistry, Biochemistry, allosteric change, HX-MS, LASSO, peptide mapping, Ricin toxin