In the first blog of this series, we entered the microscopic world of the cell, where we discovered how incredibly varied, ubiquitous and dynamic proteins are. To further our understanding of biochemistry and also allow us to develop next generation medical treatments, effective robust analytical tools are required. One such tool employed by researchers is Hydrogen Deuterium exchange (HDX) interfaced with LC-MS detection.
As mentioned in the first blog, HDX allows us to not only image the 3D configuration of proteins but also to observe conformational changes . Moreover, HDX also offers us the opportunity to monitor associations with molecules such as, hormones, drugs, antibodies and other proteins (such as subunits of quaternary structures).
The HDX technique exploits the rate in which hydrogen atoms are replaced with deuterium isotopes on specific sites on a protein of interest. For HDX, the sites of interest, are the hydrogens found on the amide backbone1.
Note: hydrogens attached to O or S can also undergo exchange but not at the frequency that is useful for HDX experimentation.
The rate of exchange is dependent on the microenvironment of a protein and so, as a result, there is a difference in exchange frequencies at different locations on the molecule. Simply put, if the hydrogens are not hydrogen bonded to either a ligand or another part of the protein, then they are considered ‘exposed’. The result of this is that they have the opportunity to readily exchange with deuterium isotopes when exposed to a D2O enriched buffer. Furthermore, exchange is also observed, when certain amides are hydrogen bonded in tertiary conformations, quaternary associations or when they are involved in ligand binding. The key point is that the H-D exchange frequency for these hydrogen locations, is slower compared to fully exposed hydrogens. Therefore, as a result, the HDX technique is able to exploit this phenomenon to obtain 3D structural characteristics of proteins .
Note: no exchange is observed from hydrogen bonding in secondary protein conformations such as alpha helices, as the hydrogen bonding is too strong.
In summary, HDX works by measuring the rate of HD exchange at different sites along a peptide or protein sequence. Then, using sophisticated software like HD Examiner, these exchange rates are analysed to generate a 3D molecular map of the protein, revealing any ligand binding sites and conformational changes.
1. The HDX experiment – this where a protein sample is exposed to a D2O containing buffer for a fixed amount of time. The sample is then quenched to stop further exchange. This is repeated with multiple samples at increasing D2O exposure times. This allows us to measure HD exchange rates.
2. Proteomic analysis - each sample is fragmented (ether in-source or enzymatically), measured and sequenced using high resolution LC-MS.
3. HDX specific software – analyses the LC-MS data of the combined samples to identify HD exchange sites. The software then measures rates of exchange to allow for structural characteristics and dynamics to be elucidated.
How to deploy D2O for exchange
There are three primary ways to initiate deuteration.
1. Use of a D2O buffer diluent - this is the most popular but due to dilution, sensitivity sometimes can be an issue.H-D exchange can be either acid base or water catalysed, although exchange is at its slowest at a pH (pD) of 2.5-3. Indeed, as a result, a pH of 2.5-3 at low temperature (<0 °C) is used to quench the HD exchange reaction at the end of each exposure of deuterium.
In its most simplistic form, proteomic analysis, using LCMS, works by breaking a peptide up into smaller chunks and then, having analysed these into the LC-MS, the pieces of the ‘puzzle’ are reassembled (in silico) to ascertain the structure and possible function of the molecule. This is achieved primarily through bottom-up or top-down approaches.
In this approach, a protein is digested to into peptide fragments using a peptidase such as pepsin (which can be in-line or in-well). To optimise the digestion, often the cys-cys disulfide bridges (formed during the tertiary formation stage of protein synthesis) are reduced and derivatised through methylation (preventing back oxidation). The only difference in HDX is that the protein or peptide is deuterium labelled prior to the reduction, alkylation and digestion steps.
The advantage of the bottom-up approach, compared to top-down, is as follows: it is a more established methodology, easier to interpret the results and there is less H-D scrambling (see Top-down) in the source.
This approach avoids reduction, alkylation and digestion by instead conducting in-source fragmentation. The advantage of this approach is that it is the only MS method that generates true site-specific H-D exchange. This is because the technique maintains the protein’s native structure. This leads to a more precise mapping of protein dynamics and conformational changes. The disadvantage of the top-down approach is that the MS spectra is a lot more complicated to interpret as the produced fragments are often a lot larger than those generated by bottom-up approaches. Furthermore, as a result of the energetics of the in-source fragmentation process, unwanted H-D scrambling events are often observed, whereby both protons and deuterium ions migrate along the protein backbone. This leads to a loss of accuracy of where the H-D exchange sites were originally located.
As H is exchanged for D, the protein becomes heaver and so changes in the MS profiles are observed from sample to sample (with increasing longer exposures to deuteration) within an experiment. In other words, the heavier the protein becomes, the more the peak profiles on the mass spectrum are shifted towards the right.
There are two kinetics in play.
· The first is where the H-D exchange is much slower than the open-close rate of a specific region on the peptide. This is known as the EX2 regime and is characterised by a gradual shift in the isotopic distribution of the peptide peaks. This is dominant for ‘native-like’ conditions.Based on these two types of kinetics, rates of H-D exchange are calculated along the protein and the results are interpreted by the HDX analysis software.
HDX generates enormous amounts of data for every new protein interrogated. Firstly, conditions allowing for HDX to be measured need to be optimised. This includes optimising proteomic conditions and also scoping out the right exchange conditions for the HDX experiment. Once the conditions have been set, multiple timepoints are then collected to identify exchange sites of interest. Due to bottlenecks in the process such as digestion times and analytical runtimes, this is a time-consuming process. Furthermore, the huge amount of time dependent data that is produced, needs to be processed by highly trained scientists using specific bespoke data interpretation software packages. As a result, it can take many days to elucidate a full HDX experiment.
Note: To improve HDX workflows, Trajan has innovated fully automated HDX CHRONECT workstations. Moreover, Trajan has developed new HDX data processing software which cuts down analysis from days to hours.
Understanding the structure and dynamics of proteins ligand interactions is highly complicated. However, HDX coupled with LC-MS, gives us an unparalleled insight into the remarkable world of these molecular machines and the vital roles they play.
Over the coming blogs, we are going to explore the remarkable list of growing applications for HDX. For example, we will look at how HDX reveals, in greater resolution, the impact of genetic mutations on the perturbations in protein conformations leading to loss of function. Moreover, how this remarkable technique allows us to see more clearly the interplay of other molecules such as drugs and antibodies with proteins. In doing, so this gives us a progressively clearer insight enabling the future development of more targeted treatments and more efficacious vaccines.