Capturing Conformational Changes in Biotherapeutics by Hydrogen Deuterium Exchange and UHPLC–MS

August 2, 2010

BioPharm International

Volume 2010 Supplement, Issue 6

By providing information on the relative accessibility of locations within a protein, HDX by mass spectrometry opens new windows into the higher order structure of biomolecules.

ABSTRACT

Hydrogen deuterium exchange (HDX) studies on proteins, pioneered by Linderstrom-Lang in the 1940s, provides information that helps clarify tertiary structure. In the 1990s, linking HDX to mass spectrometry made it possible to characterize ever larger biomolecules and to identify the location of deuterium uptake. In the past three years, another advance has taken place with the development of cooled, fast separation chromatography combined with mass spectrometers that can separate ions in multiple orthogonal dimensions. Further, the technique has matured to the level where it is no longer an esoteric research topic, but a tool that can be applied over several stages in biotherapeutic development. This broad applicability facilitates faster development for promising biotherapeutics. This article examines some of the recent advances in HDX technology and how these advances have informed practical applications.

Hydrogen deuterium exchange (HDX) studies on labile hydrogens in proteins were pioneered by Hvidt and Linderstrom-Lang starting in the 1940s.1 They studied the dynamic exchange of labile hydrogens with deuterium in protein structures in deuterated solvent. This same principle is exploited today with analytical techniques that far outperform those available two generations ago.

(WATERS CORPORATION)

Recent technological improvements in ultra high pressure liquid chromatography (UHPLC) and mass spectrometry (MS) have provided separation and detection tools to delve into fine-grained detail unimagined as little as five years ago. The precision with which modern mass spectrometers can measure changes in molecular weight means that the location of deuterium uptake can be accurately determined. The extent of deuterium uptake depends on protein structure and conformation and thus provides insights that are vital to biotherapeutic development. Consequently, HDX is likely to become a routine technique in the near future.

HDX with mass spectrometry complements the information obtained by crystallography, X-rays, and other analytical methods. Furthermore, HDX by UHPLC–MS requires far less sample than those methods and sometimes can supply information when those techniques do not work.

Most recently, fast separation by UHPLC–MS at 0 °C has allowed HDX techniques to mature into a tool that is usable by biochemists and can be used routinely in biotherapeutic development. Such routine use facilitates a faster path for promising biotherapeutics. This article examines some of the recent advances in this technology and how those advances have informed practical applications.

Higher Order Structure (HOS) Characterization of Biopharmaceutical Products

As the industry has developed, the concept of well-characterized biological products (WCBP) emerged.2 Detailed information about the higher order structure (HOS) of biotherapeutics provides better characterization. As a result, regulatory agencies are now encouraging sponsors to provide more information on dynamic studies of a biopharmaceutical product, particularly at the submission stage.3 Methods in routine use for HOS studies include calorimetry, nuclear magnetic resonance, X-ray crystallography, and advanced fluorescence. With the advent of ultra-fast HDX with UHPLC–MS, a new level of detail is possible.4,5 HDX with mass spectrometry already is an important tool at the discovery stage for leading biotech companies, and is likely to rapidly progress into other departments.

Basic Principles of HDX

The basic principle of HDX is that amide hydrogens on the backbone of a biomolecule are more exposed in solution and therefore more prone to exchanging with deuterium in a solution of D2O. Over a time-course experiment, the number of hydrogens that are exchanged can be measured by MS and therefore the degree of activity at various sites can be inferred. Certain locations are more prone to exchange because they are less protected. These locations can be identified, sometimes to the residue level, and mapped to the sequence. Thereafter, the relationship to biological activity can be correlated to a three-dimensional map of the protein.

Modern HDX workflows tend to favor LC combined with ESI Q-Tof MS (Figure 1), because of the level of detail required. HDX experiments benefit from faster UHPLC separations with sub-2-µm particles: The chromatographic efficiency obtained means a shorter run time and a better resolution than with HPLC, with a smaller amount of protein loaded. Being able to perform separations in less than 10 min minimizes the loss of deuteration. This loss is termed "back-exchange," reflecting that the amide deuteriums in the protein backbone can re-exchange to hydrogen in solution. Lower pH, fast separations, and a cold environment all contribute to minimizing this effect.

Figure 1. The workflow of a typical hydrogen-deuterium exchange (HDX) experiment with mass spectrometry (MS) detection. The digestion and chromatographic portion of the experiment is kept at 0 °C to minimize back exchange.

In the workflow outlined in Figure 1, the protein in H2O is continuously labeled with deuterium and sampled at multiple time points (from seconds to hours). Typically a 15- to 20-fold excess of D2O is used in physiological buffer at neutral pH. The exchange is quenched at each time point by lowering the pH to 2.5 and the temperature to 0 °C for each sample drawn.

Two pathways are open to the user. For rapid monitoring of global conformational changes, the protein is injected directly into the MS detector. For local analysis, to determine the specific location of changes in deuteration, peptides are digested online using an immobilized pepsin column. Pepsin is a non-specific enzyme, and has the advantage of working at a lower pH. .

Figure 2. Fluidic schematic for online pepsin digestion using ultra high pressure liquid chromatography (UHPLC) with hydrogen-deuterium exchange (HDX). The peptide separation is performed inside the HDX manager at 0 °C.

The reproducibility of digestion and chromatographic separation is an important component of a successful experiment, allowing the comparison of the same peptide across different time points (Figure 3). The mass shifts for selected peptides can be compared. The faster exchange occurs for a peptide, the more accessible is that site in solution relative to other locations on the protein. Therefore, the relative folding and dynamics also can be inferred from the different rates of uptake at different locations. Differences in uptake for different conditions can be mapped onto tertiary structures for easier visualization.

Figure 3. HDX chromatographic separation of Interferon alpha 2b peptides by online digestion and their ESI–MS spectra. The chromatography is identical for all time points. In the illustration, the same peptide is extracted at each of the time points and varies in mass-to-charge ratio (m/z) because of increasing uptake of deuterium. The blue bar indicates the m/z of the 12C isotope with no deuteration.

Calmodulin is a calcium-binding protein, and the loop regions with an EF-hand motif are folded when the calcium is added in protein solution. The highlighted region in Figure 4A depicts the apo and holo versions of the protein. No calcium is bound in the apo version; the corresponding area in the holo version of the protein is folded differently. The color scheme in Figure 4B superimposes a visual representation of the differing proportions of deuterium incorporation on the sequence (a "heat map"). In the apo form, a yellow color represents >60% of relative deuterium uptake, whereas the same location in the holo form bound with calcium is <10% relative uptake by the end of the experiment (4 h).

Figure 4a. A model structure of calmodulin protein in apo and holo forms. Highlighted in yellow is the region responsible for calcium binding. The different deuterium incoporation is observed under two forms of the conformational state.

Peptide Mapping for HDX

Mass spectrometry has contributed to studying tertiary structure by increasing the information available. Biomolecules in an HDX experiment typically are digested with a non-specific enzyme. By using a data-independent acquisition MS technique (MSE ), the many overlapping peptides can be identified with high confidence, as shown in Figure 5.7 The linear sequence coverage can be up to 100% because of the many overlapping peptides, and therefore the location of uptake can be pinpointed more precisely.

Figure 4b. The color scheme for percent relative deuterium uptake of the calmodulin peptide, residues 46–65. This region showed a different amount and rate of exchange because of the conformation differences between the apo and holo forms. The information displayed in this heat map can be superimposed on the 3D model as in Figure 4a for easier visualization.

A peptide map generated by non-specific proteases can be highly complex and rapid chromatography may lead to many coeluting peptides. In the past, mass spectrometers were programmed to use the ion signal intensity to switch to MS–MS mode, known as data-dependant acquisition (DDA). Despite ever more sophisticated techniques to cope with switching back and forth, low abundance peptides still can be missed. Even worse, targeted DDA techniques are poor at quantitation and sometimes the analysis has to be repeated because of lack of reproducibility. So a robust strategy that simultaneously provides quantitation and identification has productivity advantages for a biotherapeutics developer.

Figure 5. A part of peptide map coverage of interferon alpha-2b by pepsin online digestion using data-independent acquisition MS. The high coverage with many overlapping peptides is helps locate regions of uptake more accurately on the sequence map.6

By using data-independent acquisition MS, a comprehensive dataset can be generated without prior knowledge of the sample. This dataset is achieved by acquiring two parallel channels, one at low collision energy (CE) and the other at high CE. The resulting raw data files contain parallel data channels with the molecular ions in one, and fragment ions in the other (Figure 6). The time link between these channels is used to automatically relate the molecular ions and their fragments by a number of criteria such as retention time and peak characteristics. From a single injection, LC–MS can provide comprehensive structural information. The data generated are quantitative, and the simplicity of the technique means that non-experts can use it. The comprehensive nature of the technique means that all of the peptides are measured, without bias being introduced because of peptide signal intensity.7 This is an important feature for HDX studies where key information might be found in low intensity peptides.

Figure 6. UHPLC–MS is a method of unbiased data acquisition that comprehensively catalogs complex samples in a single analysis.

More Information in Less Time

The incorporation of robotics is an aspect that helps make such complex experiments more routine and less prone to error. The labeling, quenching, and injection can be fully automated to better handle laborious HDX experiments. The variability involved in time-sensitive labeling experiments is minimized by robotics and leaves scientists free to do higher level tasks instead. Biotherapeutics developers thereby obtain greater levels of information in less time, and make better use of their capital, both human and hardware.

Possible Future Trends

Ion mobility mass spectrometry (IMS) has been commercialized for a few years, and early publications already showed that HDX studies benefitted by helping to reduce spectral complexity.8 This additional dimension of separation is likely to speed up adoption of HDX with MS, although the lack of automated data processing will remain a challenge in the short term. Only now that HDX is moving into the mainstream is there greater impetus to rely on software rather than specialized human interpretation.

In prototype work with a Synapt HDMS System (Waters Corporation), gas-phase reactions have been carried out to simplify the tasks and reduce the steps needed to provide useful information. One example is recent work on gas-phase HDX, where instead of performing the exchange in solution outside the instrument, it was instead performed inside the mass spectrometer,9 demonstrating that it is possible to perform rapid and efficient gas-phase HDX. The advances in instrument architecture have made this type of exploratory study possible, re-using commercially available instrumentation with a minimum of modification.

Because different parts of the protein are being labeled, greater understanding of the dynamic behavior of the proteins in the gas phase is possible. Naturally, much work remains to be done to relate this information to biological conditions, but it is already clear that the speed of this technique has positive implications for comparing different samples. Furthermore, the simplicity of the experiment makes it appealing for routine use in environments where a quick view of conformation would be beneficial.

Conclusion

HDX by mass spectrometry opens new windows into the dynamics of biomolecules by providing information on the relative accessibility of different conformations of a protein, or locations within a protein. As instruments have developed, a number of important advances have made HDX with mass spectrometry a more accessible tool to study the HOS of biotherapeutics. Industry and regulatory trends have very clearly moved toward an emphasis on conformation and the relationships between biomolecules. Therefore HDX has received a welcome impetus as one of the tools that is likely to be adopted rapidly in more areas of biotherapeutic analysis.

JOOMI AHN is senior research chemist and ST JOHN SKILTON, PHD, is a senior marketing manager, both at Waters Corporation, Milford, MA, 508.478.2000, stjohn_skilton@waters.com

References

1. Berger A, LinderstrØm-Lang K. Deuterium exchange of poly- alanine in aqueous solution. Arch Biochem Biophys.1957;69:106–18.

2. Brown F, Lubiniecki A, Murano G, editors. Characterization of Biotechnology Pharmaceutical Products. Proceedings of a symposium organized and sponsored by the US FDA Center for Biologics Evaluation and Research. 1995 Dec 11–13; Washington, DC. New York: Karger; 1998.

3. US Food and Drug Administration. Guidance for Industry. Comparability protocols. Chemistry, manufacturing, and controls information. Rockville, MD: 2003 Feb.

4. Wales TE, Fadgen KE, Gerhardt GC, Engen JR. High-speed and high-resolution UHPLC separation at zero degrees celsius. Anal Chem. 2008;80(17):6815–20.DOI:10.1021/ac8008862.

5. Zhang J, Adrián FJ, Jahnke W, Cowan-Jacob SW, Li AG, Iacob RE, et al. Targeting Bcr–Abl by combining allosteric with ATP-binding-site inhibitors. Nature. 2001 Jan 28;463:501–6. doi:10.1038/nature08675.

6. Laboratory of Molecular Structure Characterization [homepage on the Internet]. Draw map for peptic peptide coverage map. Prague, Czech Republic: Institute of Microbiology; [cited 2010 July 9]. Available from: http://ms.biomed.cas.cz/MSTools/.

7. Silva JC, Gorenstein MV, Li GZ, Vissers JP, Geromanos SJ. Absolute quantification of proteins by LCMSE: a virtue of parallel MS acquisition. Mol Cell Proteomics. 2006;5:144–56.

8. Iacob RE, Murphy III JP, Engen JR. Ion mobility adds an additional dimension to mass spectrometric analysis of solution-phase hydrogen/deuterium exchange. Rapid Commun Mass Spec. 2008;22:2898–904; DOI: 10.1002/rcm.3688.

9. Rand KD, Pringle SD, Murphy III JP, Fadgen KE, Brown J, Engen JR. Gas-phase hydrogen/deuterium exchange in a traveling wave ion guide for the examination of protein conformations. Anal Chem. 2009;81:10019–28.

10. Alverdi V, Mazon H, Versluis C, Hemrika W, Esposito G, Van den Heuvel R, et al. cGMP-binding prepares PKG for substrate binding by disclosing the c-terminal domain. J Mol Biol. 2008;375:1380–93. doi:10.1016/j.jmb.2007.11.053.