HX-MS Data Analysis

4 Case Studies of the Application of HX-MS to Formulation Development of mAbs 

Although mAbs share high sequence similarity within an immunoglobulin subclass, they can have dramatically different physical and chemical stability properties unique to a specific mAb. Hence, individual formulation strategies are needed to address different types of physical and chemical instabilities observed for different mAbs formulated in various dosage forms [45]. Stabilization strategies 

Map significant differences on crystal structure or homology model 

Differential exchange mapped onto mAb homology model 

---

HX-MS data analysis in epitope mapping can be laborious due to the large number of peptides generated and associated kinetic measurements (i.e., different incubation/exchange time points). Recent development in user-friendly HX-MS software (see Chapter 3) relieves scientists of many hours of arduous data analysis [56-58]. The software advancements along with the addition of automation to the standard HX-MS platform have greatly facOitated the acceptance of HX-MS technology as a suitable technique for routine characterization of therapeutic proteins. 

The determination of the consensus measurement reproducibility for HX-MS is an integral part of method validation, as its derived uncertainty supports the estimation of precision under reproducibility conditions. Measurement reproducibility is determined through a statistical analysis of HX-MS measurements conducted on the same protein sample in many laboratories. The results from this analysis can help investigators detect measurement variance due to different realizations of the HX-MS technique. For studies of unknowns, measurement reproducibility defines where measurements of D uptake differ statistically. It is expected that rigorously evaluated measurement reproducibility can foster a broader understanding and acceptance of HX-MS data. To date, no study leading to the determination of the measurement reproducibility of HX-MS has been reported. 

Because these important drawbacks cannot be overlooked, it is critical to find alternative ways that are more capable of accurately displaying and making use of the entire HX-MS data set in a way that can be quickly, simply, and easily interpreted for comparabOity studies. In conducting an HX-MS comparison (or difference measurement), where one sample functions as a control or reference state that is then compared to some experimental state(s) of the same protein, it is important to consider the following in terms of data analysis and display ... 

Data Processing in Bottom-Up Hydrogen Exchange Mass Spectrometry 47 Table 3.1 List of available software utilities for HX-MS analysis... 

Once the filtered list is determined and tested against unlabeled data sets collected under HX-MS conditions, it can be applied for deuteration analysis of the full set of deuterated samples. The detection and quantification of deuterium levels rely upon the validity of this list. It proceeds to determine deuterium levels by budding theoretical isotopic expansions of a native peptide distribution across aU deuteration levels and calculating the best fit between the actual isotopic distribution and the... 

The methodologies described so far are suitable for extracting information from one peptide at a time. A typical HX-MS dataset, however, often contains hundreds of overlapping peptides. Taking advantage of all information contained in these peptides may increase the spatial resolution of the HDX result but requires major efforts. Therefore, a more comprehensive approach is required to include all the peptides in the data analysis. Several fairly successful attempts have been made in developing such a tool [16, 20, 34, 35]. 

---