Hydrogen/Deuterium Exchange-MS for Biologics: Higher-Order Structure Analysis and Epitope Mapping

Biologics are the fastest-growing sector in the pharmaceutical industry, with global sales approaching $200 billion and making up about 20% of all pharmaceuticals.1 More interestingly, seven out of the top 10 best-selling drugs in 2015 were biologics, indicating that they can command higher prices than small molecules in general.2 The higher price for biologics may be justified in two ways: their ability to fulfill previously unmet needs and the high cost for development and production. Certain hormone and enzyme replacement therapies, such as insulin for diabetes and β-glucocerebrosidase for Gaucher’s disease, are only possible with biologics. Because of their high specificity to the targets, therapeutic antibodies can achieve high efficacy and safety against some of the most confounding diseases, such as cancer and autoimmune disorders.3 The inherent large size and complex nature of biologics pose a more structural instability and challenging production process than that experienced with small molecule-based drugs.

Therapeutic antibodies are the most economically successful class of biologics, representing approximately half the total sales of biologics, and include five of the aforementioned seven biologics in the top 10.4 Antibody molecules are even larger and more complicated than most other therapeutic proteins, and their market is growing fastest among biologics. Most therapeutic antibodies entered the market in the late 1990s, about 15 years after the first recombinant therapeutic protein, insulin. This late start is in part due to technical challenges arising from chimerization and humanization. Combined with emerging technologies such as antibody–drug conjugates and bispecific antibodies, therapeutic antibodies remain the most cutting-edge area in the pharmaceutical industry.5

Biosimilars (generic versions of biologics), especially those for therapeutic antibodies, are attracting a lot of attention as the industry finds itself in the midst of a patent cliff of biologics. By the end of this period, which spans 2012–2019,6 patents for all of the top 10 biologics on the market will have expired, and will therefore create an opportunity for innovator companies and biosimilar companies to compete. To date, the European Medicines Agency (EMA) has approved 31 biosimilars since its first approval of a biosimilar in 2006,7 and the U.S. FDA is following the same trend.8 One challenge in biosimilar development is the biophysical and biochemical characterization of the drugs, because the majority of the top-selling biologics coming off patent are antibodies, with more complex structures than most other therapeutic proteins.

Higher-order structure analysis for biologics

In contrast to conventional small-molecule drugs, biologics are large and complex compounds. They contain a higher-order structure (consisting of secondary, tertiary, and quaternary structures), which has a significant impact on efficacy and safety.9 The complex nature of biologics requires a more stringent biophysical characterization to assure safety and efficacy.10 The FDA encourages drug makers to analyze the higher-order structure of a biologic for internal and external comparability studies.11,12

No single biophysical method can characterize the higher-order structure in solution in high resolution. All currently available analytical tools for higher-order structure analysis, such as circular dichroism (CD) and Fourier transform infrared spectroscopy (FTIR), can provide only the sum of higher-order structural information for the entire protein molecule. While X-ray crystallography can provide a high-resolution structure of a protein, the structure obtained is in a solid state, and this technique cannot monitor the higher-order structure of a formulated biologic. Although nuclear magnetic resonance (NMR) can determine high resolution of a protein structure in solution, this technique has a size limitation. This eliminates many biologics, including therapeutic antibodies, from being viable candidates for NMR analysis. Therefore, a widely applicable biophysical method capable of providing structural information in solution at high resolution is desirable.

Hydrogen/deuterium exchange mass spectrometry (HDX-MS) is an ideal method to probe a protein’s higher-order structure. It is widely applicable and is capable of monitoring the dynamic properties of a protein at the submolecular level in solution.13 The dynamic properties can, in turn, be correlated to hydrogen bonding and higher-order structure. Applications for HDX-MS range from quality control and formulation optimization to observing the effects of mutations in a biologic molecule. This analysis can also help establish “equivalence” of a biosimilar with the innovator’s biologic. A comprehensive biophysical characterization, in line with current regulatory expectations, will minimize the amount of clinical studies, and therefore the costs of the biosimilar development.12

Epitope mapping for therapeutic antibodies

Epitope mapping is an essential aspect of the discovery and development of therapeutic antibodies, and can assist in the selection of lead candidate molecules. Intellectual property considerations of patentability, and freedom to operate consequences, can be dependent on the epitope itself. Further, regulatory agencies recommend that prior to use in humans and, whenever possible, the protein bearing the reactive epitope should be biochemically defined and the antigenic epitope itself determined.14

Commonly employed epitope mapping strategies have some limitations.15 While an X-ray co-crystal of the antigen–antibody complex remains the gold standard of epitope determination, it is not always feasible due to the difficulty of obtaining high-quality, well-diffracting crystals. Conformational changes induced by mutagenesis may lead to loss of binding and can result in a falsely identified epitope. Discontinuous epitopes, nonlinear in origin and dependent on the structural conformation of the protein, may preclude the use or complicate the interpretation of limited proteolysis and overlapping peptide analysis.

HDX-MS epitope mapping technology is widely applicable and compatible with discontinuous conformational epitope in the solution state. The success of HDX-MS epitope mapping is reliant on various parameters, primarily concerned with the coverage of the antigen sequence by peptic fragments. Two major disadvantages of the technology are heavy disulfide bonds and heavy glycosylations. The success rate is expected to be around 95% (the number obtained at ExSAR Corp., South Brunswick Township, NJ).

Method for higher-order structure analysis by HDX-MS

HDX-MS can probe the dynamic properties of a protein in solution at a submolecular level. Upon transfer from water to a deuterated buffer, a protein gradually increases its mass as the protein’s hydrogen atoms are replaced with deuterium (Figure 1). Each hydrogen-to-deuterium exchange (HDX) rate depends on the protein’s dynamic properties around the atom. The dynamic regions of a protein exchange from hydrogen to deuterium faster than the rigid regions (top left in figure). After the HDX reaction, the HDX-MS platform digests the protein (top right) and determines the mass of each peptic fragment generated (middle right). Repetition of the experiments with various exchange time generates a deuterium buildup curve against time for each peptide (middle left). Deuterium buildup curves may be color-coded and placed in the protein sequence (bottom left). HDX-MS technology determines the deuterium incorporation of a target protein at the submolecular level, and the degree of deuterium incorporation of each peptic fragment infers the dynamic properties of the region of the protein.

Figure 1 – HDX-MS method overview.

Examples of higher-order structure analysis by HDX-MS

HDX-MS can compare higher-order structures of a biologic product between batches, formulations, manufacturers, and/or mutants. HDX-MS analysis of human growth hormone (hGH) at two different pH levels is shown in Figure 2.13 In this analysis, HDX-MS demonstrated that hGH has different deuterium incorporating profiles and thus different dynamic properties at various parts of the protein. In the case of hGH, helical regions of the protein are more rigid than loop regions. HDX-MS can also provide differences in dynamic properties with various experimental parameters in solution at submolecular levels. Interestingly, even at pH 2.6, the protein shows slow exchange regions around four helix bundles, indicating that hGH sustains molten globular-type structure even at the low pH.

Figure 2 – HDX-MS of hGH at pH 7.0 and 2.6. The level of deuterium incorporation is colorcoded, as shown in the top right. Light-blue cylinders on top of the sequence indicate the locations of helices in crystal structure.

Examples of epitope mapping by HDX-MS

HDX-MS can identify the epitope of antigen–antibody interaction in solution. Three epitopes identified by HDX-MS agreed well with the contact residues of the corresponding X-ray co-crystal structures.16 The agreement is more evident when the HDX-MS epitopes were overlaid on the corresponding X-ray co-crystal structures (Figure 3). It is worth pointing out that all three epitopes described here have discontinuous conformational epitopes.

 Figure 3 – Antigen–antibody co-crystal structures and epitopes identified by HDX-MS: a) cytochrome c, E8 antibody; b) IL-13, CNTO607 antibody; and c) IL-17A, CAT2200 antibody. White CPK models are antigens. Blue and light blue are epitopes identified by HDX-MS. Yellow are light chains and light green are heavy chains of antibodies.

Conclusion

Biologics, particularly therapeutic antibodies and biosimilars, promise to be an increasing therapeutic offering for years to come. Higher-order structure analysis of biologics and epitope mapping of therapeutic antibodies are key components of biophysical characterization. Both analyses are encouraged by the FDA, as these analyses are a means with which to ensure safety and efficacy. HDX-MS can play a major role in protein characterization by filling unmet needs in these areas with automation13 and analysis that is well-suited for regulatory filings.17

References

  1. http://www.mckinsey.com/industries/pharmaceuticals-and-medical-products/our-insights/rapid-growth-in-biopharma
  2. http://www.pharmacompass.com/radio-compass-blog/top-drugs-by-sales-revenue-in-2015-who-sold-the-biggest-blockbuster-drugs
  3. https://www.americanpharmaceuticalreview.com/Featured-Articles/183086-Advancing-Biologics-Development-and-Manufacturing/
  4. Ecker, D.M. and Jones, S.D. Mabs 2015, 7, 9–14.
  5. Carter, P. Experimental Cell Research 2011, 317, 1261–9.
  6. Calo-Fernández, B. and Martínez-Hurtado, J.L. Pharmaceuticals 2012, 5, 1393–1408.
  7. http://www.gabionline.net/Biosimilars/General/Biosimilars-approved-in-Europe
  8. https://en.wikipedia.org/wiki/Biosimilar
  9. https://www.gpo.gov/fdsys/pkg/CHRG-111hhrg52285/pdf/CHRG-111hhrg52285.pdf
  10. Berkowitz, S.A.; Engen, J.R. et al. Nat. Rev. Drug Discov. 2013, 11, 527–40.
  11. https://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/Q6B/Step4/Q6B_Guideline.pdf
  12. https://www.fda.gov/OHRMS/DOCKETS/98fr/2004d-0118-gdl0001.pdf
  13. Hamuro, Y; Coales, S.J. et al. J. Biomol. Tech. 2003, 14, 171–82
  14. https://www.fda.gov/downloads/BiologicsBloodVaccines/GuidanceComplianceRegulatoryInformation/OtherRecommendationsforManufacturers/UCM153182.pdf
  15. Gershoni, J.M.; Roitburd-Berman, A. BioDrugs 2007, 21, 145–56.
  16. Pandit,D.; Tuske, S.L. et al. J. Mol. Recognit. 2012, 25, 114–24.
  17. Hamuro, Y. J. Am. Soc. Mass Spectrom. 2017, 28, 486–97.

Yoshitomo Hamuro, Ph.D., is senior scientist, SGS Life Sciences, 606 Brandywine Pkwy., West Chester, PA 19380, U.S.A.; tel.: 610-696-8210; e-mail: [email protected]www.sgs.com/lifescience