Hydrogen
exchange-mass spectrometry (HX-MS) is widely promoted
for its ability to detect subtle perturbations in protein structure,
but such perturbations will result in small differences in HX. However,
the detection limit of HX-MS has not been widely investigated, nor
is there a useful approach for defining the detection limit of HX-MS
measurements. In this work, we designed a well-characterized structural
variant spiking model to investigate the detection limit of conventional
peptide-based HX-MS. The detection limit was challenged by spiking
small fractions of a structural variant (modeled using maltose binding
protein W169G mutant) into a reference protein (wild-type maltose
binding protein). As little as 5% of the structural variant could
be detected. The small structural perturbation was not resolvable
by far UV circular dichroism, differential scanning calorimetry, or
size exclusion chromatography. Furthermore, we validated the ability
of the hybrid statistical analysis approach, presented in a companion
paper (10.1021/acs.analchem.9b01325), to reliably identify small,
significant differences in HX-MS measurements. With our structural
variant spiking model, we demonstrate a benchmarking approach for
determining a detection limit of HX-MS for detection of changes in
higher-order structure that might be encountered in protein structural
comparability and similarity assessment applications.