Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jan 25.
Published in final edited form as: Anal Chim Acta. 2020 Nov 21;1143:65–72. doi: 10.1016/j.aca.2020.11.022

High-throughput hydrogen deuterium exchange mass spectrometry (HDX-MS) coupled with subzero-temperature ultrahigh pressure liquid chromatography (UPLC) separation for complex sample analysis

Mulin Fang 1, Zhe Wang 1, Kellye A Cupp-Sutton 1, Thomas Welborn 1, Kenneth Smith 2, Si Wu 1
PMCID: PMC8265693  NIHMSID: NIHMS1652486  PMID: 33384131

Abstract

Hydrogen deuterium exchange coupled with mass spectrometry (HDX-MS) is a powerful technique for the characterization of protein dynamics and protein interactions. Recent technological developments in the HDX-MS field, such as sub-zero LC separations, large-scale data analysis tools, and efficient protein digestion methods, have allowed for the application of HDX-MS to the analysis of multi protein systems in addition to pure protein analysis. Still, high-throughput HDX-MS analysis of complex samples is not widespread because the co-elution of peptides combined with increased peak complexity after labeling makes peak de-convolution extremely difficult. Here, for the first time, we evaluated and optimized long gradient subzero-temperature ultra-high-pressure liquid chromatography (UPLC) separation conditions for the HDX-MS analysis of complex protein samples such as E. coli cell lysate digest. Under the optimized conditions, we identified 1419 deuterated peptides from 320 proteins at –10 °C, which is about 3-fold more when compared with a 15-minute gradient separation under the same conditions. Interestingly, our results suggested that the peptides eluted late in the gradient are well-protected by peptide-column interactions at −10 °C so that peptides eluted even at the end of the gradient maintain high levels of deuteration. Overall, our study suggests that the optimized, sub-zero, long-gradient UPLC separation is capable of characterizing thousands of peptides in a single HDX-MS analysis with low back-exchange rates. As a result, this technique holds great potential for characterizing complex samples such as cell lysates using HDX-MS.

Graphical Abstract

graphic file with name nihms-1652486-f0007.jpg

INTRODUCTION

Hydrogen/deuterium exchange coupled with mass spectrometry (HDX-MS) is a powerful protein foot-printing method for characterization of protein dynamics and protein/protein interactions.[1] Over the past years, HDX-MS has become widely used in structural and functional biology due to the advancement in HDX sample handling, separation, MS detection, and the availability of open source data processing software.[1–3] Currently, HDX-MS is expanding to investigate complex protein samples such as large/multi-protein complexes and holds great potential for characterization of protein-protein interactions in more native-like environments like cell lysates without extensive purification.[4, 5]

HDX-MS monitors the exchange of protein backbone amide hydrogens to deuterium upon exposure of proteins to deuterium oxide (D2O).[6] Typical HDX-MS workflow includes deuterium labeling, quenching, protein digestion, LC separation, and MS detection.[7] In proteomics studies, including HDX-MS analysis, reversed-phase LC (RPLC) separation is the most commonly used LC separation format. Typically, RPLC is performed at room temperature (20 °C) or higher for more efficient mass transfer of analytes to achieve better separation.[8] However, normal RPLC conditions are not adequate for HDX-MS experiments as back-exchange of deuterium upon injection of the labeled peptides onto the LC column, due to the high H2O content of the LC buffers, can result in loss of site-specific information and can bias the identification of sites of interest. Thus, optimized LC separation conditions are critical to minimize back-exchange rates and perform more robust HDX-MS analysis.[4]

In recent years, many efforts have been devoted to improving LC separation performance and reducing the back-exchange rates for HDX-MS to perform more robust and complicated HDX experiments.[4] Subzero-temperature LC separation functions to improve HDX-MS analysis by decreasing back-exchange during separation.[5, 9–11] In fact, the back-exchange rate during LC separation is dependent on the separation temperature with a rate decrease of approximately 3 fold per 10-degree decrease in temperature.[12–14] For example, Venable et al. demonstrated a negligible loss of deuterium for fully deuterated fibrinopeptide A after 100 minutes at −30 °C.[9] Wales et al. recorded an average 81.7% deuteration of the tryptic peptides of phosphorylase B after a one-hour gradient separation at −20 °C.[5] This study also indicates that LC separation at −10 °C maintains a similarly high degree of deuterium incorporation of peptides when compared with the LC separation at −20 °C (74.9% and 81.7%, respectively, using a one-hour gradient). In order to perform LC under subzero-temperature conditions, mobile phase modifiers can be used to reduce the freezing point of mobile phase A and avoid freezing of the mobile phase in the LC system.

Short-gradient UPLC separations (e.g. <10 minute gradient) with small UPLC packing material sizes (e.g. sub-2 μm particle size) have previously been used to boost separation resolution when compared with traditional HPLC separation for the HDX-MS analysis of protease digests of large proteins or multi-protein complexes.[15, 16] To further increase the separation efficiency, increased flow rates are often used in UPLC separations compared to traditional HPLC flow rates. As a consequence of small particle size, the operational pressure increases significantly with the increased flow rates in the UPLC system; this effect is particularly prominent under subzero-temperature conditions which result in an increase in mobile phase viscosity [8]. The collective effects of small particle size, high pressure, and low temperature necessitated the optimization of the high-throughput HDX-MS system for the analysis of complex samples.

In this study, we optimized a long-gradient sub-zero-temperature UPLC platform (e.g. > 1-hour separation gradient) for the separation of deuterium labeled complex cell lysate digest samples. We examined the effects of mobile phase modifiers, separation temperatures, flow rates and pressure, and gradient lengths on separation efficiency and deuterium retention. Our results demonstrated that a 90-minute UPLC separation at −10 °C and ~10,000 psi with 10% acetonitrile as the mobile phase modifier could be implemented to maintain high levels of deuterium retention in HDX-MS analysis. Additionally, our results showed that the average fractional deuterium incorporation was actually increased slightly for peptides that eluted late in the gradient (after 50 minutes in a 90-minute gradient), making long-gradient UPLC separation a promising approach for the HDX-MS analysis of highly complex samples such as cell lysates.

MATERIAL AND METHODS

Materials and reagents

All chemicals, including Deuterium oxide were purchased from Sigma-Aldrich (Milwaukee, WI) unless noted otherwise. Phenylmethylsulphonyl fluoride (PMSF) was purchased from VWR (Radnor, PA) and trypsin (TPCK treated) was obtained from ThermoFisher (Rockford, IL). The desalting column, Strata C18-U (55 μm, 70 Å, 100 mg/mL) was purchased from Phenomenex (Torrance, CA). The ACE® Excel® SuperC18™ column (100 mm × 2.1 mm, 1.7 μm, 90 Å) was purchased from Advanced Chromatography Technologies Ltd (Aberdeen, Scotland).

Sample preparation

Escherichia coli (E. coli) K12 cells were grown in 2% LB media and cultured at 37 °C with gentle shaking at 250 rpm overnight. The cultured E. coli cells were collected and centrifuged at 13,000 rpm for 60 minutes at 4 °C. The supernatant was discarded, and the E. coli pellets were resuspended in 25 mM ammonium bicarbonate at a ratio of 1 gram of cell pellets to 5 mL of buffer. 0.1% (v/v) PMSF was added to inhibit proteases from degrading proteins. Then, E. coli pellets were lysed using an Avestin EmulsiFlex-C3 homogenizer. The cell lysate was then centrifuged at 4 °C and 13,000 rpm for 60 minutes to remove cell debris. 6 M urea, 200 mM dithiothreitol and 200 mM iodoacetamide were prepared in 25 mM ammonium bicarbonate. 1 mg (350 μL) of the proteins in the supernatant were incubated with 100 μL of 6 M urea and 25 μL of 200 mM dithiothreitol for denaturation and reduction at 37 °C for 1 hour, followed by incubating with 100 μL of 200 mM iodoacetamide for alkylation at room temperature in dark for 30 minutes. The alkylated proteins then were incubated with 100 μL of 200 mM dithiothreitol at room temperature immediately, followed by adding 25 mM ammonium bicarbonate to make the volume of protein solution 1 mL. Proteins were digested with trypsin (prepared in 25 mM ammonium bicarbonate) at 37 °C under pH 7 condition overnight with a protein to enzyme ratio of 50:1 (w/w). The digested peptides were desalted using SuperC18™ columns before vacuum concentration.

Deuterated E. coli peptides were prepared by reconstituting the concentrated peptides into D2O at a volume ratio of 1:9 to a final peptide concentration of 0.68 μg/μL. The labeling mixture was then incubated at room temperature for three days. The final labeling D2O fraction was 0.9, which is used for the calculation of fractional deuterium incorporation. In order to avoid freezing of peptide samples during the sample injection step at −10 °C, acetonitrile was added to the sample before sample injection so the final acetonitrile fraction is 10%. Non-deuterated E. coli peptides were prepared using the same procedure as the deuterated samples, but H2O rather than D2O was used to reconstitute the concentrated peptides.

Subzero-temperature liquid chromatography

Subzero-temperature liquid chromatography was performed by placing a six-port two-way valve and a C18 UPLC column (100 mm × 2.1 mm, 1.7 μm, 90 Å) into a portable freezer (Figure S1). The temperature of the portable freezer could be controlled from −20 °C to +20 °C. The UPLC system utilized in this study is a Thermo Accela pump with a maximum pressure of 14,000 psi. The mobile phase modifier, operation temperature, and flow rate were evaluated for their effect on separation efficiency. For all LC runs, 25 μg of non-deuterated E. coli lysate peptides were manually injected onto the column at −10 °C. The LC gradient was controlled using the Thermo Accela UPLC system. Two mobile phase systems (mobile phase systems 1 and 2) were evaluated to avoid freezing of solvent in the LC system. Mobile phase system 1 was made up of 0.1% formic acid and 10% acetonitrile in water for mobile phase A and 0.1% formic acid in acetonitrile for mobile phase B. Mobile phase system 2 was made up of 0.1% formic acid and 35% methanol in water for mobile phase A and 0.1% formic acid in acetonitrile for mobile phase B.

The gradient for this set of evaluations started with 0% mobile phase B for sample loading over 10 minutes followed by an increase from 0% to 33% mobile phase B with varying gradient lengths (15 minutes to 90 minutes in 15-minute increments) for peptide separation. Then the column was washed with 90% mobile phase B for 5 minutes followed by a decrease to 0% mobile phase B for 3 minutes. The column was re-equilibrated by flushing with 100% mobile phase A for 10 minutes. For each gradient, non-deuterated E. coli peptides were analyzed for peptide identification and the deuterated E. coli lysate peptides were analyzed to evaluate deuterium incorporation after separation. All experiments were performed in triplicate. Eluted peptides were detected online using a photodiode array (PDA) detector and/or a Thermo Velos Elite mass spectrometer (Thermo Fisher Scientific, Hanover Park, IL, USA).

Bottom-up MS analysis

An LTQ Orbitrap Elite mass spectrometer (Thermo Fisher Scientific, Hanover Park, IL, USA) with a custom nano-ESI interface was used for the MS experiments.[17, 18] The flow was split to result in approximately 1.5 μL/min flow to the nano-ESI interface. The heated capillary temperature was set to 275 °C with a spray voltage of 3.5 kV. High-resolution MS scans were obtained with a MS resolution setting of 100,000 and the m/z range was 350 to 1350. MS/MS scans were acquired using the ITMS with collisional induced dissociation (CID) at a normalized collision energy setting of 35%. The ten most abundant precursor ions were selected for MS/MS. The AGC for MS/MS was 3E4 and the max ion injection time was 50 ms with 3 microscans.

Data analysis

Peptide identification.

Non-deuterated E. coli lysate peptides were identified using MSGF+ [19] to search the mass spectra from the LC-MS/MS analysis against the annotated E. coli database (E. coli K12) that was downloaded from www.uniprot.org (proteome ID is UP000000625). The decoy database is automatically generated in MSGF+ software. The peptide identifications were filtered using a SpecE value cut-off of 1E-10 (i.e. the calculated FDR <1% at the unique peptide level).

Fractional deuterium incorporation calculation.

Peak detection and fractional deuterium incorporation of peptides were calculated as previously described in the literature [12, 20] using the in-house developed software. Furthermore, this software fits peptide mass distributions to a Gaussian model and calculates the R2. Peptide isotope distributions with an R2 value above 0.9 were selected for deuterium incorporation calculation.

Fractional deuterium incorporation simulation.

The theoretical deuterium level of each peptide was simulated as previously described [21] for comparison to experimental values. Briefly, for each identified peptide, each amide deuterium back-exchange rate was calculated using algorithms provided by Englander (http://hx2.med.upenn.edu/) [13], parameterized using T = 263 K and pH = 2.5. These rates were used to calculate overall expected deuterium remaining on the peptide at the time it eluted.

RESULTS AND DISCUSSION

Optimization of subzero-temperature UPLC conditions for separation of E. coli tryptic peptides

It has been reported that UPLC [7] and low temperatures separation conditions [4] improve LC-HDX-MS analysis by enhancing the separation performance and decreasing back-exchange rates. However, until now, the evaluation of these parameters has only been performed on simple protein or peptide samples [5, 9, 11]. Here we optimized a subzero-temperature UPLC system for complex sample HDX-MS analysis using tryptic peptides from E. coli cell lysate as a model system.

We first compared two mobile phase systems (details in the method section) under subzero-temperature conditions to determine the effect of the mobile phase modifier on separation efficiency. We evaluated 35% methanol as suggested by Wales et al. [5] and acetonitrile as suggested by Zhang et al. [22]. In the study conducted by Zhang et al., 4.5% acetonitrile was added to mobile phase A to perform LC-HDX-MS at −9 °C for characterization of the epitope of birch pollen allergen.[22] Here, we increased the percentage of acetonitrile in mobile phase A to 10% to reduce the column pressure at −10 °C and maintain high operational flow rates (Figure S2). Triplicate LC-HDX-MS runs were performed at −10 °C and demonstrated reproducible separation without column or sample freezing when 35% methanol or 10% acetonitrile were utilized as the mobile phase modifier. We observed that many peptides did not bind to the RPLC column and eluted out during the sample loading time when methanol was used as the modifier, which may result in low peptide recovery in the analysis (Figure S3). Therefore, 10% acetonitrile was utilized as the modifier in mobile phase A at −10 °C for the following evaluation and optimization.

We then evaluated the UPLC separation efficiency of E. coli lysate peptides under different flow rates at −10 °C using 10% acetonitrile as the modifier (Figure 1). The full peak width at half maximum (FWHM) under each flow rate condition was calculated using 7 peaks selected from various points in the entire UPLC elution window, as previously suggested[23]. The FWHM decreased as the flow rate increased, suggesting higher separation efficiency at a higher flow rate. Additionally, the backpressure of the LC system at equilibrium (100% mobile phase A) was measured at each corresponding flow rate. We found that a flow rate of 150 μL/min offered the optimal operational pressure (9,824 psi ± 404) and FWHM for our system. Replicate runs were performed to evaluate the robustness of the ultra-high-pressure operating conditions (Figure S4).

Figure 1. Flow-rate optimization of the UPLC separation at −10 °C.

Figure 1.

Open in a new tab

Base-peak chromatogram of UPLC separation under different flow rates and corresponding pressures when the UPLC column was at equilibrium with 100% buffer A at −10 °C. 7 peaks selected from various point in the gradient (1–7 in the figure) were evaluated. The FWHM (in seconds) of each peak as well as their average FWHMs were reported in the table. The corresponding pressure under each condition was calculating as the average pressure from triplicate runs.

Finally, we compared our optimized subzero-temperature separation results with an analogous separation conducted at room temperature (Figure S5). At a flow rate of 150 μL/min, the backpressure was increased from 5,729 psi ± 157 at room temperature to 9,824 psi ± 404 at −10 °C. The backpressures were significantly increased at the same flow rate when the temperature was decreased from room temperature to −10 °C because the viscosity of mobile phase increases as temperature decreases [8]. The average FWHM decreased slightly at −10 °C (e.g. from 8.28s ± 1.19 at RT to 6.50s ± 0.56 at −10 °C), suggesting that the UPLC separation at subzero temperatures shows similar or slightly better separation performance compared to room temperature separation using the same flow rate.

Deuterium incorporation of E. coli lysate peptides using the UPLC separation at −10 °C

To determine the deuterium incorporation of the peptides after LC separation, UPLC-HDX-MS analysis was performed using a flow rate of 150 μL/min with a 30-minute gradient at both room temperature and −10 °C. In total, the fractional deuterium incorporation was calculated for 840 identified peptides from 222 proteins at room temperature and 879 identified peptides from 222 proteins at −10 °C. Figure 2A shows the relative percentage of peptides with different fractional deuterium incorporation at room temperature and −10 °C. In total, 772 peptides have greater than 50% deuterium incorporation at −10 °C, and only 39 peptides have greater than 50% deuterium incorporation at room temperature. The average deuterium incorporation was approximately 40% higher at −10 °C (62.2±0.81%) when compared with equivalent room temperature separation (22.6±0.88%) (Figure 2B). We also evaluated the fractional deuterium incorporation at 4 °C and found that the majority of the characterized peptides showed the fractional deuterium incorporation in the range from 30% to 60% (Figure S6). We found that the optimized UPLC separation at −10 °C greatly reduces the back-exchange rate in complex samples such as E. coli lysate peptides in a 30-minute separation gradient. We further evaluated the run-to-run reproducibility of our UPLC-HDX-MS platform. Figure 2C demonstrates the correlation between two replicate runs at −10 °C with a 30-minute gradient. The fractional deuterium incorporation was calculated for the same 520 identified peptides in both runs and we observed good reproducibility between these two runs (R2=0.9326). Our results suggested that the optimized subzero-temperature UPLC-HDX-MS platform can efficiently and reproducibly characterize complex deuterated samples such as cell lysate digest and maintain high deuteration levels.

Figure 2. Evaluation of fractional deuterium incorporation in the 30-minute UPLC-HDX-MS analysis.

Figure 2.

Figure 2.

Open in a new tab

(A) Relative percentage of peptides in each fractional deuterium incorporation range at RT and −10 °C (840 peptides from RT, 879 peptides from −10 °C, flow rate: 150 μL/min); (B) Average fractional deuterium incorporation of the peptides at RT and −10 °C. The error bars are the 95% confidence interval from the mean of the identified peptides based on the mean, the standard deviation, and sample size; (C) Linear regression shows the reproducibility of the two replicates performed at −10 °C.

Sub-zero temperature UPLC-HDX-MS of the E. coli lysate peptides with respect to gradient length

Good quality isotopic envelopes with few or no overlapped peaks are desirable for HDX-MS analysis, but a major hurdle for achieving this in complex samples is the limited peak capacity associated with LC separation. The incorporation of UPLC has significantly improved the peak capacity of HDX-MS, allowing for the characterization of larger proteins and protein complexes [16]. Still, many of these studies are performed with very short gradients at 0 °C. We have demonstrated that our optimized UPLC method can be efficiently operated at −10 °C with high deuterium retention after a 30-minute gradient, so we evaluated the separation performance of the UPLC and the back-exchange with longer gradients (from 15 min to 90 min) through the analysis of E. coli lysate peptides.

The separation performance of the UPLC system at −10 °C was evaluated using non-deuterated E. coli lysate peptides with different separation gradients (Figure S7). In a 90-minute gradient, 27,246 deconvoluted mass features were detected, which is ~3-fold higher than the number of deconvoluted mass features detected using a 15-minute gradient (8,530). Only 1,200 peptides were identified in the 15-minute gradient run which was significantly lower than the 4,520 peptides identified for the 90-minute gradient run, suggesting better separation efficiency with longer UPLC gradients.

We further evaluated the back-exchange observed for each gradient by comparing the average fractional deuterium incorporation of deuterium labeled E. coli peptides using the optimized UPLC-HDX-MS analysis. In total, we characterized 1419 deuterated peptides from 320 proteins in the 90-minute gradient run, which is about 4-fold more than were identified in the 15-minute run (382 peptides from 141 proteins). As shown in Figure 3, detected deuterated peptides in from different gradient lengths have similar fractional deuterium incorporation ranges. In all the gradients, we observed that a majority of detected deuterated peptides have greater than 50% deuterium incorporation. The average deuterium incorporation slightly decreased as the length of the gradient increased (64.5±14.9% for the 15-minute gradient and 56.9±14.8% for the 90-minute gradient). However, the number of deuterated peptides that were detected and characterized in the UPLC-HDX-MS analysis increased drastically with increasing lengths of gradient (382 deuterated peptides detected for the 15-minute gradient and 1419 deuterated peptides detected for the 90-minute gradient). Overall, our results suggested that longer UPLC gradients at −10 °C provide better separation capacity for increased peptide detection while maintaining similar deuterium incorporation.

Figure 3. Evaluation of fractional deuterium incorporation in the UPLC-HDX-MS analysis at varying gradient lengths (15 minutes to 90 minutes, −10 °C, flow rate: 150 μL/min).

Figure 3.

Open in a new tab

(A) Relative percent of peptides in each fractional deuterium incorporation range. (B) Box and Whisker plots of the fractional deuterium incorporation of the identified peptides separated.

It is generally accepted that deuterium incorporation should decrease over time when the concentration of D2O is lowered if the primary structure of the peptide is the only contributor to the rate of back-exchange [13]. Using this model, we calculated the theoretical fractional deuterium incorporation of the detected E. coli peptides after separation using a 90-minute gradient and plotted them against their elution times (Figure S8). We then compared the simulated results to our experimental results for the 90-minute gradient (Figure 4). The deuterium incorporation for both the simulated and experimental peptides decreased at the beginning of the gradient. However, the simulated deuterium incorporation decreased linearly throughout the entire gradient, while the deuterium incorporation for the experimental peptides gradually increased at the end of the gradient. Interestingly, the same trend was observed under all gradient length conditions (15 minutes to 90 minutes) (Figure S9).

Figure 4. Comparison of simulated fractional deuterium incorporation and experimental fractional deuterium incorporation in the 90-minute UPLC-HDX-MS analysis at −10 °C.

Figure 4.

Open in a new tab

A total of 1419 identified deuterated peptides with R2 above 0.9 were used here.

Previous literature indicates that the observed deuteration trend may be the result of interactions between the peptides and the RPLC column altering the deuteration back-exchange rate depending on peptide sequence and length [21]. We first plotted the length of the identified deuterated peptides versus elution time and found that the average peptide length increased with increased elution time (Figure 5A). Longer peptides may have more interaction with the RPLC column which may block sites and decrease the overall rate of back exchange for some longer peptides. Another factor that may affect the back-exchange rates of peptides is the formation of secondary structure caused by hydrophobic interactions within the peptide. Formation of secondary structure is more common for longer peptides that generally eluted later in the gradient [21, 24, 25]. These longer peptides may undergo induced formation of stabilized secondary structures [26, 27] and the deuterium embedded in the secondary structure will have less solvent accessibility and a reduced back-exchange rate. It has also been suggested that column interactions favor retention of the secondary helical structures [21]; this may further decrease the back exchange rate of longer peptides eluted later gradient due to column-induced deceleration of back-exchange.

Figure 5. Investigation of the effects of peptide properties on deuterium incorporation for 1419 identified deuterated peptides in the 90-minute UPLC-HDX-MS analysis at −10 °C.

Figure 5.

Open in a new tab

(A) Calculated peptide lengths in different elution windows. (B) GRAVY values of peptides in different elution windows.

Other peptide properties may also contribute to the back-exchange of individual peptides. We calculated the relative hydrophobicity for each identified peptide using the grand average of hydropathy (GRAVY) calculator (http://www.gravy-calculator.de/) [21] and plotted these values against the elution time for each peptide (Figure 5B). Interestingly, even though RPLC separates based on hydrophobicity, the relative hydrophobicity of the peptides only increased in the first 30 minutes of the separation, and the average hydrophobicity does not change significantly from 30 minutes to 100 minutes. Thus, the overall increase in hydrophobicity may be another factor contributing to the reduced back-exchange rate for peptides eluted late in the gradient. The combined effects of column interaction, secondary structure, and hydrophobicity may explain the nonlinear deuteration trend observed in this study.

Overall, the ability to maintain high fractional deuteration throughout a long gradient HPLC separation is valuable to HDX-MS analysis because higher deuterium incorporation allows more information regarding protein binding to be retained and reduces the bias of characterization. Furthermore, longer separation allows for more peptides identification to obtain higher proteome sequence coverage. While accurate identification of altered deuterium labeling is generally required for typical HDX-MS experiments on single proteins or simple protein complexes, it is especially important for analysis of complex protein mixtures as the likelihood of errant identification of interaction sites increases with sample complexity and could lead to false conclusions regarding protein binding or altered folding. The long gradient HPLC separation at subzero temperature introduced here exhibits the potential for the high proteome coverage and deuterium retention required to perform HDX-MS analysis on complex samples such as cell lysates for the study of protein binding and dynamics in more native-like environments which could allow the more in depth study of protein interactions and pathways.

CONCLUSION

We have performed, for the first time, the systematic optimization of the UPLC-HDX-MS analysis of complex protein samples such as E. coli lysate digest at −10 °C. Under the optimized conditions, we characterized 1419 deuterated peptides from 320 proteins using a 90-minute separation gradient at −10 °C. Our results suggested that peptides eluted across the entire gradient (including peptides eluted at the end of the gradient) all maintain relatively high levels of deuteration. Overall, the optimized, sub-zero temperature, long-gradient UPLC separation is capable of characterizing thousands of peptides with low back exchange, which greatly increased the throughput in a single HDX-MS analysis.

Improvement and optimization of the sub-zero temperature UPLC-HDX-MS system may still be made; for example, the separation efficiency of highly complex samples can be further improved for high throughput HDX-MS by optimizing the parameters of the UPLC column including particle size, column length, ion pairing reagents, and packing materials. Additionally, we have to note that our average fractional deuterium incorporation is slightly lower than some reported results [5]. This is likely due to our sample preparation and manual sample injection process. Sample preparation can be improved to further reduce the back exchange by flash-freezing samples using dry ice, thawing the sample right before sample injection, controlling solution conditions such as pH and ionic strength, etc. Other measures such as controlled desolvation in the MS ionization step can also help improve the overall deuterium retention as previously reported [28]. The potential application of this method may suffer from typical limitations involved in LC-MS based proteomics studies such as run-to-run variations in LC-MS analysis [29]. Additionally, as is common in complex sample analysis, the possibility of false positive results is increased; in this case, erroneous detection of shifts in deuteration could falsely indicate protein interactions, so further confirmation of potential interactions screened with this method using other analytical techniques may be necessary to corroborate results. Furthermore, analysis of complex HDX-MS data sets is complicated and there are not currently any software packages available designed to analyze HDX-MS datasets of complex protein mixtures such as cell lysates. This type of data analysis also requires statistical analysis to interpret significant shifts in deuteration and methods by which to achieve this analysis have not been deeply studied [3]. However, all potential drawbacks considered, the application of long gradient low temperature separation to HDX-MS analysis holds great promise for broad screening of protein dynamics in samples with high complexity and in more native-like environments such as human cell lysates. As such, we intend to apply our optimized subzero temperature long-gradient HDX-MS system to the analysis of functional biological systems to further assess its ability to screen and characterize protein interactions and dynamics. Future studies will incorporate the recommendations provided by the International Conference on Hydrogen-Exchange Mass Spectrometry (IC-HDX) such as multiple labeling time points and incorporation of biological replicates[1]. Optimization of different buffer systems capable of lower temperature (−20 °C) separation can also be explored for analysis of complex samples using HDX-MS in future studies.

Supplementary Material

1
2
3

Highlights:

  • Long-gradient UPLC hydrogen deuterium exchange mass spectrometry of complex cell lysate digest at −10 °C

  • Deuterium incorporation remains high after 90-minute separation gradient

  • Characterization of thousands of deuterated peptides from hundreds of proteins in a single separation

ACKNOWLEDGMENTS

This work was partly supported by grants from NIH NIAID R01 AI41625 and NIH NIH/NIAID-2U19AI062629. MLF would like to acknowledge Jiwon Kang for the helps in the web lab.

S.W. and K.S. provided funding; S.W. provides supervision of the project; M.F. and Z.W. prepared samples and performed MS and LC experiments; M.F, K.A.C, and T.W. performed data analysis; all authors contributed to the manuscript preparation.

Footnotes

COMPETING INTERESTS

The authors declare no competing interests (financial and non-financial).

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCE

  • 1.Masson GR, et al. , Recommendations for performing, interpreting and reporting hydrogen deuterium exchange mass spectrometry (HDX-MS) experiments. Nat Methods, 2019. 16(7): p. 595–602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Harrison RA and Engen JR, Conformational insight into multi-protein signaling assemblies by hydrogen-deuterium exchange mass spectrometry. Curr Opin Struct Biol, 2016. 41: p. 187–193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Claesen J. and Burzykowski T, Computational methods and challenges in hydrogen/deuterium exchange mass spectrometry. Mass Spectrom Rev, 2017. 36(5): p. 649–667. [DOI] [PubMed] [Google Scholar]
  • 4.Trabjerg E, Nazari ZE, and Rand KD, Conformational analysis of complex protein states by hydrogen/deuterium exchange mass spectrometry (HDX-MS): Challenges and emerging solutions. Trac-Trends in Analytical Chemistry, 2018. 106: p. 125–138. [Google Scholar]
  • 5.Wales TE, et al. , Subzero Celsius separations in three-zone temperature controlled hydrogen deuterium exchange mass spectrometry. J Chromatogr A, 2017. 1523: p. 275–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Konermann L, Pan J, and Liu YH, Hydrogen exchange mass spectrometry for studying protein structure and dynamics. Chem Soc Rev, 2011. 40(3): p. 1224–34. [DOI] [PubMed] [Google Scholar]
  • 7.Engen JR and Wales TE, Analytical Aspects of Hydrogen Exchange Mass Spectrometry. Annu Rev Anal Chem (Palo Alto Calif), 2015. 8: p. 127–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Wu N. and Clausen AM, Fundamental and practical aspects of ultrahigh pressure liquid chromatography for fast separations. J Sep Sci, 2007. 30(8): p. 1167–82. [DOI] [PubMed] [Google Scholar]
  • 9.Venable JD, et al. , Subzero temperature chromatography for reduced back-exchange and improved dynamic range in amide hydrogen/deuterium exchange mass spectrometry. Anal Chem, 2012. 84(21): p. 9601–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Amon S, et al. , Spatially resolved protein hydrogen exchange measured by subzero-cooled chip-based nanoelectrospray ionization tandem mass spectrometry. Anal Chem, 2012. 84(10): p. 4467–73. [DOI] [PubMed] [Google Scholar]
  • 11.Pan J, et al. , Subzero temperature chromatography and top-down mass spectrometry for protein higher-order structure characterization: method validation and application to therapeutic antibodies. J Am Chem Soc, 2014. 136(37): p. 13065–71. [DOI] [PubMed] [Google Scholar]
  • 12.Walters BT, et al. , Minimizing Back Exchange in the Hydrogen Exchange-Mass Spectrometry Experiment. Journal of the American Society for Mass Spectrometry, 2012. 23(12): p. 2132–2139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Bai Y, et al. , Primary structure effects on peptide group hydrogen exchange. Proteins: Structure, Function, and Bioinformatics, 1993. 17(1): p. 75–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Connelly GP, et al. , Isotope effects in peptide group hydrogen exchange. Proteins, 1993. 17(1): p. 87–92. [DOI] [PubMed] [Google Scholar]
  • 15.Wu Y, Engen JR, and Hobbins WB, Ultra performance liquid chromatography (UPLC) further improves hydrogen/deuterium exchange mass spectrometry. J Am Soc Mass Spectrom, 2006. 17(2): p. 163–7. [DOI] [PubMed] [Google Scholar]
  • 16.Wales TE, et al. , High-speed and high-resolution UPLC separation at zero degrees Celsius. Anal Chem, 2008. 80(17): p. 6815–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Wang Z, et al. , Top-down Mass Spectrometry Analysis of Human Serum Autoantibody Antigen-Binding Fragments. Sci Rep, 2019. 9(1): p. 2345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Wang Z, et al. , Two-Dimensional Separation Using High-pH and Low-pH Reversed Phase Liquid Chromatography for Top-down Proteomics. Int J Mass Spectrom, 2018. 427: p. 43–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Kim S. and Pevzner PA, MS-GF+ makes progress towards a universal database search tool for proteomics. Nat Commun, 2014. 5: p. 5277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Mayne L, Hydrogen Exchange Mass Spectrometry. Methods Enzymol, 2016. 566: p. 335–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sheff JG, Rey M, and Schriemer DC, Peptide-Column Interactions and Their Influence on Back Exchange Rates in Hydrogen/Deuterium Exchange-MS. Journal of the American Society for Mass Spectrometry, 2013. 24(7): p. 1006–1015. [DOI] [PubMed] [Google Scholar]
  • 22.Zhang Q, et al. , Epitope Mapping by HDX-MS Elucidates the Surface Coverage of Antigens Associated with High Blocking Efficiency of Antibodies to Birch Pollen Allergen. Anal Chem, 2018. 90(19): p. 11315–11323. [DOI] [PubMed] [Google Scholar]
  • 23.Jorgenson JW, Capillary liquid chromatography at ultrahigh pressures. Annu Rev Anal Chem (Palo Alto Calif), 2010. 3: p. 129–50. [DOI] [PubMed] [Google Scholar]
  • 24.Krokhin OV, Sequence-specific retention calculator. Algorithm for peptide retention prediction in ion-pair RP-HPLC: application to 300- and 100-A pore size C18 sorbents. Anal Chem, 2006. 78(22): p. 7785–95. [DOI] [PubMed] [Google Scholar]
  • 25.Zhou NE, Mant CT, and Hodges RS, Effect of preferred binding domains on peptide retention behavior in reversed-phase chromatography: amphipathic alpha-helices. Pept Res, 1990. 3(1): p. 8–20. [PubMed] [Google Scholar]
  • 26.Mant CT, et al. , EFFECT OF PEPTIDE-CHAIN LENGTH ON PEPTIDE RETENTION BEHAVIOR IN REVERSED-PHASE CHROMATOGRAPHY. Journal of Chromatography, 1988. 458: p. 193–205. [DOI] [PubMed] [Google Scholar]
  • 27.Babushok VI and Zenkevich IG, Retention Characteristics of Peptides in RP-LC: Peptide Retention Prediction. Chromatographia, 2010. 72(9–10): p. 781–797. [Google Scholar]
  • 28.Coales SJ, Tomasso JC, and Hamuro Y, Effects of electrospray capillary temperature on amide hydrogen exchange. Rapid Commun Mass Spectrom, 2008. 22(9): p. 1367–71. [DOI] [PubMed] [Google Scholar]
  • 29.Bantscheff M, et al. , Quantitative mass spectrometry in proteomics: a critical review. Anal Bioanal Chem, 2007. 389(4): p. 1017–31. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1652486-supplement-1.pdf (458.8KB, pdf)
2
NIHMS1652486-supplement-2.xlsx (262KB, xlsx)
3
NIHMS1652486-supplement-3.xlsx (42.8KB, xlsx)

RESOURCES