Optimize your protein therapeutics across the pipeline with hydrogen exchange mass spectrometry

Over the past three decades, hydrogen exchange mass spectrometry (HX-MS) has advanced from an esoteric research tool, to a highly repeatable analytical method sensitive to subtle changes in the higher-order structure of proteins. The method is now widely employed in drug discovery and vaccine development for the mapping of conformational epitopes. More recent work suggests that HX-MS data also has the potential to support process and analytical development, as well as provide information for regulatory considerations.

In this SelectScience^® webinar, now available to watch on demand, Dr. David Weis, Professor of Chemistry at the University of Kansas, provides a brief overview of what is measured in HX-MS and how results are interpreted. Plus, Weis discusses broad characterization of the epitopic surface of ricin toxin, development of an analytical framework to evaluate structural similarity of therapeutic proteins, and much more.

Read on to explore highlights from the live Q&A session and register now to watch the webinar on demand.

Can HX-MS be used to map the binding sites for small molecules binding to protein targets?

DW: The interaction of small molecules with protein targets is a little bit more complicated than a protein-protein interaction. That comes about because the small molecule has such a small contact surface with the protein target. When we look at small molecule-protein binding we do often see some evidence of slowed hydrogen exchange or protection in the vicinity of the binding site, but it's subtle. We'll also have various subtle allosteric effects that will be of about the same magnitude as direct contact with the small molecule. So, it's often difficult to distinguish the allosteric effects from the direct contact sites.

Nevertheless, a lot of those allosteric effects are useful and there's been a lot of very nice work done, in particular in establishing structural activity relationships across libraries of small molecules against a single protein target.

Can HX-MS be used to study the effects of changes in pH?

DW: In biopharmaceuticals, there's a lot of interest in how proteins change in response to changes in pH.

As a lot of proteins are found to be more stable at a particular pH than at other pHs, we'd like to be able to probe how that change in pH manifests itself in terms of altered patterns of hydrogen exchange. One of the challenges in hydrogen exchange is that the hydrogen exchange chemistry is extremely sensitive to changes in pH. As a rough guideline, a one-unit change in pH will cause roughly a tenfold change in the rate of hydrogen exchange. And that arises because the hydrogen exchange process is catalyzed by hydroxide in solution. For example, from pH 6 up to pH 7, there's an inherent tenfold increase in the rate of hydrogen exchange.

So, although one can explore pH effects, one needs to be very cautious about this effect. There are several ways that we can compensate for that inherent hydrogen exchange rate effect. One way is to use a mathematical correction. If we know ahead of time that 10 seconds of hydrogen exchange at pH 7 is equivalent to 100 seconds of hydrogen exchange at pH 6, we can simply use a different timescale for making those comparisons. This does introduce a bit of a challenge, though, because we're limited in terms of the dynamic range of the hydrogen exchange labeling time. So, we can't go much faster than about 10 seconds with an automated liquid handler, and we can't, for practical reasons, run for much longer than 24 hours.

We can measure hydrogen exchange in the middle where we can overlap times at different pHs. That's the most reliable way at this point to correct for these differences in pH.

How long does HX-MS data analysis take and what are some of the limitations associated with the technique?

DW: One of the limitations is that standard mass spec software is not too well designed for analyzing hydrogen exchange mass spec data. The reason for that is although we're looking at specific peptides with well-defined retention times, the mass itself is not well defined because the mass is changing as a function of how long the deuterium exposure is. This means that most traditional mass spec software is very focused on a molecule or a peptide, in this case, with a well-defined mass. We don't have a well-defined mass, so we need purpose-designed software. There is now quite a suite of both commercial and free software out there for doing hydrogen exchange data analysis.

One of the challenges that arise is the issue of back exchange. This is the loss of the deuterium label during the chromatographic step because we go from a deuterium-rich environment during labeling to a protium or H2O-rich environment during LC separation. So, we need to complete separations of these labeled peptides quickly. One of the downsides of this is we often wind up with overlaps of peptide pictures. We'll have co-eluting peptides that are at least partially overlapping in m/z space as well. These collisions in the data mean that we have to review data to eliminate these kinds of artifacts.

For a sort of a platformed molecule of roughly a few hundred peptides, 100 to 200 peptides, data analysis will take a few hours. There is some expert review required to ensure that these kinds of overlaps are minimized. But on a platform molecule, often we have that knowledge built-in, and it eliminates undesirable peptides from the dataset.

Can HX-MS be used to determine the structure of aggregated proteins?

DW: If there are specific contacts between the monomers in an aggregated structure, it may be possible to map those by hydrogen exchange. If the aggregates are nonspecific, we're going to see that hydrogen exchange is slowed in many locations. It's going to be much more challenging then to determine the nature of those aggregated species.

What do you see for the future of this technique? And is there anything exciting to look out for with this?

DW: I'm particularly interested in seeing a couple of things. One is adopting more of the principles of data science to these rather complicated four-dimensional datasets that we generate in hydrogen exchange. We have chromatographic retention time. We have the m/z axis of the spectrum. We have the ion abundance. The fourth dimension is really the HX labeling time. It'd be interesting to see if there are more ways that we can treat the data less invasively in a more automated fashion.

I'm also interested to see how hydrogen exchange will work at the back end of the development pipeline where we begin to think about these questions of equivalence and comparability. Comparability related to post-approval process changes, and then the potential use of hydrogen exchange in the biosimilar space.

To learn more about hydrogen exchange mass spectrometry, watch the webinar on demand>>

SelectScience runs 10+ webinars a month, discover more of our upcoming webinars>>