Hydrogen/Deuterium Exchange Mass Spectrometry: Fundamentals, Limitations, and Opportunities

Abstract

Hydrogen/deuterium exchange mass spectrometry (HDX-MS) probes dynamic motions of proteins by monitoring the kinetics of backbone amide deuteration. Dynamic regions exhibit rapid HDX, while rigid segments are more protected. Current data readouts focus on qualitative comparative observations (such as "residues X to Y become more protected after protein exposure to ligand Z"). At present, it is not possible to decode HDX protection patterns in an atomistic fashion. In other words, the exact range of protein motions under a given set of conditions cannot be uncovered, leaving space for speculative interpretations. Amide back exchange is an under-appreciated problem, as the widely used (m-m₀)/(m₁₀₀-m₀) correction method can distort HDX kinetic profiles. Future data analysis strategies require a better fundamental understanding of HDX events, going beyond the classical Linderstrøm-Lang model. Combined with experiments that offer enhanced spatial resolution and suppressed back exchange, it should become possible to uncover the exact range of motions exhibited by a protein under a given set of conditions. Such advances would provide a greatly improved understanding of protein behavior in health and disease.

Keywords: electrospray ionization; isotope effect; mass spectrometry; molecular dynamics simulation; protein aggregation; protein binding; protein digestion; protein dynamics; protein folding; protein modeling; protein mutation; protein stability.

Graphical abstract

Fig. 1

MD simulations of amide backbone fluctuations for several NH OC contacts in cytochrome c. Panels on the left show hydrogen-oxygen distances versus time. For each pair, the NH donor is listed first. Vertical dotted lines at 0.25 nm indicate the H-bond cut-off. The top panel illustrates an H-bond that remains permanently closed during the 1 μs simulation window. All others undergo closed ⇌ open transitions. Panels on the right illustrate MD snapshots as overlays of NH_closed (blue dashed) and NH_open (red dashed) conformers, for time points indicated by the blue and red arrows. Element coloring: N (blue), H (white), O (red), C (green). Reproduced with permission from ref. (80) Copyright 2021, American Chemical Society.

Fig. 2

Protection factor effects on HDX kinetics. The deuteration kinetics of an NH_j site were calculated using Equation 1 for different protection factors P_j in a 2 h time window. Parameters used: k_{ch_j} = 6.9 s⁻¹, corresponding to poly-Ala at 295 K in D₂O solution at pD 7.4.

Fig. 3

Schematic depiction of the standard bottom-up HDX-MS workflow. The red box refers to a D₂O environment. Blue boxes indicate steps that take place in H₂O. Blue spheres represent hydrogen (¹H) and red spheres represent deuterium (²H). The locations of these spheres represent NH_j sites in ubiquitin. The bottom-left panel schematically indicates the behavior of a peptide that exhibits enhanced protection in the presence of a ligand that binds to the protein.

Fig. 4

Kinetics of protein backbone deuteration and peptide back exchange during bottom-up HDX-MS.A, fictitious stretch of protein sequence, producing the peptide ACDEFGHI after digestion. B, deuterium content of each NH site j vs. time, with k_{ch_j} values from refs. (59, 78). (i) 10 min HDX in D₂O under native conditions at pD 7.4, 22 °C, with log P_j = 4 for all sites. (ii) 40 s pepsin digestion at 15 °C, pH 2.5 in H₂O. (iii) 15 min peptide LC at 0 °C, pH 2.5 in H₂O. (iv) Peptide ESI-MS with 5% droplet/gas phase back exchange. C, Peptide deuterium content D_pep(t) calculated from panel B for residues 2 to 8 and 3 to 8. ESI-MS analysis takes place much faster than indicated here.

Fig. 5

Performance of theEquation 9back exchange correction for the model peptide ACDEFGHI. Each panel shows three HDX profiles, calculated as discussed in the text: D_{pep_IDEAL,} ideal scenario without back exchange for residues 3 to 8; D_{pep_RAW}, experimentally measurable raw data; D_{pep_CORR}, raw data after back exchange correction (Equations 9 and 10). A–C, 10% back exchange, t_BX = 200 s; (D–F) 33% back exchange, t_BX = 1000 s; (G–I) 50% back exchange, t_BX = 2000 s. Each column shows data for one set of log P_j values, listed along the top.