Article pubs.acs.org/ac
Simple Setup for Gas-Phase H/D Exchange Mass Spectrometry Coupled to Electron Transfer Dissociation and Ion Mobility for Analysis of Polypeptide Structure on a Liquid Chromatographic Time Scale Ulrik H. Mistarz,† Jeffery M. Brown,§ Kim F. Haselmann,‡ and Kasper D. Rand*,† †
Department of Pharmacy, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark Diabetes Protein Engineering, Novo Nordisk A/S, Novo Nordisk Park 1, 2760 Måløv, Denmark § Waters MS Technologies Centre, Stamford Avenue, Wilmslow SK94AX, U.K. ‡
S Supporting Information *
ABSTRACT: Gas-phase hydrogen/deuterium exchange (HDX) is a fast and sensitive, yet unharnessed analytical approach for providing information on the structural properties of biomolecules, in a complementary manner to mass analysis. Here, we describe a simple setup for ND3-mediated millisecond gas-phase HDX inside a mass spectrometer immediately after ESI (gas-phase HDX-MS) and show utility for studying the primary and higher-order structure of peptides and proteins. HDX was achieved by passing N2-gas through a container filled with aqueous deuterated ammonia reagent (ND3/ D2O) and admitting the saturated gas immediately upstream or downstream of the primary skimmer cone. The approach was implemented on three commercially available mass spectrometers and required no or minor fully reversible reconfiguration of gas-inlets of the ion source. Results from gas-phase HDX-MS of peptides using the aqueous ND3/D2O as HDX reagent indicate that labeling is facilitated exclusively through gaseous ND3, yielding similar results to the infusion of purified ND3-gas, while circumventing the complications associated with the use of hazardous purified gases. Comparison of the solution-phase- and gasphase deuterium uptake of Leu-Enkephalin and Glu-Fibrinopeptide B, confirmed that this gas-phase HDX-MS approach allows for labeling of sites (heteroatom-bound non-amide hydrogens located on side-chains, N-terminus and C-terminus) not accessed by classical solution-phase HDX-MS. The simple setup is compatible with liquid chromatography and a chip-based automated nanoESI interface, allowing for online gas-phase HDX-MS analysis of peptides and proteins separated on a liquid chromatographic time scale at increased throughput. Furthermore, online gas-phase HDX-MS could be performed in tandem with ion mobility separation or electron transfer dissociation, thus enabling multiple orthogonal analyses of the structural properties of peptides and proteins in a single automated LC-MS workflow.
E
performed by infusion of a deuterated basic gas such as D2O, deuterated alcohols or ND3, into the mass spectrometer.16−21 Studies have indicated that gas-phase deuterium exchange by ND3 informs on the conformation of proteins more directly than exchange by weaker reagent bases, such as D2O and deuterated alcohols,14,22 that require the reagent base to simultaneously hydrogen bond between a charged site and a carbonyl oxygen nearby for an exchange-competent ion-neutral complex to occur (the relay mechanism).23 This means that HDX requires a specific polypeptide orientation and does not necessarily correlate only to surface accessibility, making interpretation of data in terms of structure more complicated.
lectrospray ionization (ESI) has enabled the direct transfer of intact proteins from solution to the gas-phase.1 A number of studies suggest that proteins adopt stable non-native conformations upon equilibration in the gas-phase but can retain solution-like structures for tens to several hundreds of ms after desolvation.2−4 This time regime affords access to structures of proteins pertinent to the solution-phase prior to equilibration in a desolvated environment. Several gas-phase MS techniques have been applied for characterizing biomolecules in this time regime, including ion mobility spectrometry (IMS),5−8 that measures the collision cross section of ions by their drift time in an inert gas, and techniques that rely on chemical reactions in the gas-phase like proton-transfer reactions,9,10 radical-based reactions,11−13 and hydrogen/deuterium exchange (HDX).14,15 Gas-phase HDX provides a potentially powerful approach to study both simple and complex polypeptide structure, and has traditionally been © 2014 American Chemical Society
Received: September 16, 2014 Accepted: November 6, 2014 Published: November 6, 2014 11868
dx.doi.org/10.1021/ac5035456 | Anal. Chem. 2014, 86, 11868−11876
Analytical Chemistry
Article
Figure 1. Schematic overview of the two configurations of the gas inlets (cone entry and cone exit) of a Q-TOF mass spectrometer (Synapt G1/G2) to enable gas-phase HDX in the ion source. In cone entry labeling mode, modification of the cone gas inlet allows saturation of the cone gas by leading it through a liquid deuterated reagent (b). In cone exit labeling mode, a liquid deuterated reagent is added to the existing ETD reagent vial and N2-gas (“make-up gas”) is passed across the liquid deuterated reagent into the cone exit region via a corona discharge needle (c). In ion guide labeling, purified ND3-gas is infused directly to the source T-Wave.
B, Arg8-Vasopressin, Bradykinin, Tryptophan-Cage, substance P, and the proteins ubiquitin and cytochrome c.
In contrast, exchange by the stronger reagent base, ND3, does not require simultaneous hydrogen bond formation to both the charged site and the neighboring carbonyl for the exchangecompetent ion-neutral complex to occur (the onium mechanism).23 ND3 exchange appears to be dependent on both (a) the primary structure and the location of charged sites (i.e., covalent structure) and (b) the intramolecular hydrogen bonding and surface-accessibility of labile hydrogens (i.e., noncovalent structure).14,15,24 Structural transitions and dynamics of folded polypeptides in vacuum occur at much slower rates than in solution because of the absence of solvation that significantly lowers the energy required to transition between two conformational states. 25 Gas-phase HDX performed within the millisecond time frame after ESI using ND3-gas (referred to as gas-phase HDX-MS)14,15 has proven capable of probing the near-native or unfolded protein states present in the gas-phase following ESI from denaturing or native solution conditions, respectively.14,15,26 Several options for the infusion of purified ND3-gas to an ESI-MS setup have been reported including the nebulizer or curtain gas,18,27 the ion guide,14,15 or the mass analyzer.28 These techniques either require relatively complex modification to low pressure regions of the mass spectrometer that could potentially affect instrument performance and the purchase and handling of expensive and hazardous gas. At least in part for these reasons, gas-phase HDX-MS have yet to see more widespread use. Here, we describe a simple setup for routine gas-phase HDX-MS on a commercially available Q-TOF mass spectrometer without the need for modifications to the lowpressure regions and without the use of a hazardous deuteration gas as labeling reagent. The experimental setup is based on the use of a highly basic liquid reagent, aqueous deuterated ammonia (ND3/D2O), to achieve gas-phase HDX by enriching the N2-gas admitted to the ion source with ND3. The ND3/ D2O reagent enables labeling of peptides and proteins similar to the use of purified ND3-gas, in contrast to other liquid reagents tested including D2O, EtOD, and MeOD. We have coupled the simple gas-phase HDX-MS setup to liquid chromatography, an automated nanoESI interface and advanced mass spectrometry techniques including ion mobility spectrometry (IMS) and electron transfer dissociation (ETD) performed in tandem. Gas-phase HDX-MS analysis of biomolecules separated on a liquid chromatographic time scale (with or without tandem ETD or IMS) can thus be performed in an automated routine manner. We have demonstrated these combined capabilities by analysis of model peptides, such as Leu-enkephalin, Glu-fibrinopeptide
■
MATERIALS AND METHODS Materials. Deuterium oxide 99.9%, methanol-D4 99.8%, and ethanol-D6 99% were purchased from Sigma-Aldrich (Schnelldorf, Germany) and aqueous deuterated ammonia (26% Ammonia-D3 in D2O 99.5%) was purchased from Merck (Darmstadt, Germany). 99% ND3-gas (contained in a 25 L lecture bottle) was purchased from Cambridge Isotope Laboratories (Andover, MA) and fitted with a regulator (3222702-CGA180, CONCOA, Utrecht, Netherlands). Tryptophan-Cage was purchased from Anaspec (Fremont, CA), and all other peptides and reagents were purchased from SigmaAldrich (Schnelldorf, Germany), and were used without further purification. Peptides and proteins were, unless stated otherwise, dissolved in distilled water (1 mg/mL) and diluted to 10 μM in either 50% acetonitrile/0.1% formic acid or in 10 mM ammonium acetate. H3PO4 was diluted to 1% in 50% methanol (V/V). Mass Spectrometry. Samples were, unless stated otherwise, loaded in a pulled metal-coated glass capillary (Thermo, Odense, Denmark) and ionized using a nanoflow Z-spray ESI source (Waters, Wilmslow, UK) mounted on either a Synapt G1, Synapt G2 or a Synapt G2-Si HDMS mass spectrometer (Waters, Wilmslow, UK). Alternatively, samples were introduced by direct infusion using an external automated nanoESI system (NanoMate, Advion BioSciences, Ithaca, NY) mounted on a Synapt G2. The source and spray settings were tuned to obtain the best possible signal and HDX efficiency (Supporting Information Figure S-1 and Figure S-2); For the Z-spray source: Capillary voltage 1−1.3 kV, sampling cone voltage 2−20 V, extraction cone voltage 3−5 V, source T-Wave velocity 200−500 m/s and source block temperature 70−110 °C. Other settings were kept in default conditions and acquisition was performed in “resolution mode”. For the NanoMate, the capillary voltage was increased to 1.3−1.6 kV, with other settings unchanged. Mass spectra were accumulated for 1 min at 0.5 s per scan, from m/z 100 to m/z 4000, unless stated otherwise. Information on the settings used for ion mobility, electron transfer dissociation and LC-MS are provided in the Supporting Information. Gas-Phase HDX-MS. For cone entry labeling (via cone gas), HDX was performed by introducing the labeling reagent immediately upstream of the primary cone entry region (“sample cone”) of a Synapt G1 by modifying the N2 “cone gas” PTFE tubing so that the nitrogen passed through a closed 11869
dx.doi.org/10.1021/ac5035456 | Anal. Chem. 2014, 86, 11868−11876
Analytical Chemistry
Article
25 mL glass vial containing 15 mL liquid reagent; D2O, MeOD, EtOD, or ND3/D2O (Figure 1b). See Supporting Information for additional details. For cone exit labeling (via ETD makeup gas), HDX was performed in the lower pressure region downstream of the primary cone exit (sample cone) region on both a Synapt G2 and Synapt G2-Si. This configuration required no modifications to these commercially available instruments, as gas-phase HDXMS was facilitated simply by filling the ETD-reagent vial on each instrument with 0.1−5 mL aqueous ND3/D2O reagent (Figure 1c). Controlled gas-phase HDX-MS was performed simply by varying the N2 gas-flow leading from the ETD-vial to the cone exit region (makeup gas-flow, 0−50 mL/min). See Supporting Information for additional details. For ion source labeling, HDX was performed in the source Twave ion guide by the controlled infusion of purified ND3-gas, as described previously.15 Combined solution-phase- and gas-phase HDX-MS experiments are detailed in the Supporting Information. Data Analysis. Processing of all mass spectra were carried out by MassLynx V4.1 software (Waters, Wilmslow, UK) with a (1,5) Savitzky-Golay filter smoothing and subsequent centering of the peaks. Deuterium contents of peptide, protein, and ETD fragment ions were determined with Excel 2013 (Microsoft, Redmond, WA) by calculating the difference in the intensity weighted average masses of the deuterated sample and that of a nondeuterated control sample in the absence of deuteration.
Figure 2. Gas-phase HDX-MS of phosphoric acid ionic clusters using different liquid HDX reagents in cone entry labeling mode. (a) HDX plot of the deuterium uptake of phosphoric acid ionic clusters of four and eight molecules as a function of gas-flow across ND3/D2O, EtOD, MeOD, and D2O. (b) Mass spectra recorded of clusters of eight phosphoric acid molecules at increasing flow of N2-gas enriched by an aqueous ND3/D2O solution. The red lines indicate the measured centroid average mass of the ion cluster at each gas-flow setting and the blue line indicates the theoretical average mass value in the case of deuteration of all labile hydrogens in the ion cluster.
■
RESULTS AND DISCUSSION Setup for Gas-Phase HDX-MS. Three generations of QTOF instruments (Synapt G1, Synapt G2 or Synapt G2-Si) were modified to allow fast and efficient gas-phase HDX to be performed in either the cone entry region, cone exit region, or the source Traveling Wave Ion Guide (T-Wave) (Figure 1, see Materials and Methods for further details). Advantageously, the cone entry labeling- and the cone exit labeling configuration required no or minor hardware modifications and allowed the use of liquid HDX reagents to saturate existing N2 gas-flows to facilitate gas-phase HDX in the ion source. Evaluating Liquid Reagents for Gas-Phase HDX-MS. The use of different liquid HDX reagents to facilitate gas-phase labeling reactions in the ion source were tested using phosphoric acid, which forms ionic clusters in the gas-phase. These clusters cover a large mass range and contain an increasing number of labile heteroatom bound hydrogens, making it a useful and simple model system for gas-phase HDX-MS method development. The use of several common liquid HDX reagents was explored for gas-phase HDX including D2O, MeOD, and EtOD. Additionally, we speculated that aqueous deuterated ammonia (ND3/D2O) could represent a useful liquid reagent to facilitate highly efficient gas-phase HDX-MS. The gas-phase HDX of phosphoric acid ionic clusters was measured following enrichment of the N2-cone gas (cone entry labeling) by use of either of the three commonly used liquid HDX reagents; D2O, MeOD, and EtOD, as well as aqueous deuterated ammonia (ND3/D2O). The average deuterium uptake as a function of gas-flow for all four liquid reagents is shown in Figure 2a. An increase in gas-flow results in increased deuterium uptake. All four reagents showed full exchange of all labile hydrogens in ionic clusters of four as well as eight phosphoric acid molecules, at high gas-flow settings, with 11 and 23 deuteriums, respectively. A similar degree of HDX was
observed for other cluster sizes (data not shown). It was observed that the ND3/D2O reagent and MeOD, gave the highest deuterium uptake of the phosphoric acid cluster at low cone gas-flow settings, followed by EtOD and finally D2O. This correlates with the higher gas-phase basicity of ND3, as well as the volatility and solubility of the respective liquid deuterated reagents in N2-gas. Upon extended use, a decrease in deuteration degree (Supporting Information Figure S-3) and pD was observed for the ND3/D2O reagent, due to the higher volatility of ND3 than D2O. The reduced deuterium uptake observed for the lower gas-flow settings (10−20 L/h, Figure 2a and 2b) could indicate a partial back-exchange with atmospheric H2O present in the cone entry region at low flow settings.18,29 However, with increasing cone gas-flows, the pressure of deuterated labeling gas is increased in the cone region resulting in maximal deuterium labeling of phosphoric acid clusters (Figure 2b). Gas-Phase HDX-MS of Model Peptides. The applicability of the gas-phase HDX-MS setup for labeling of peptides was examined using Leu-Enkephalin (Leu-Enk, YGGFL). For all four liquid deuterated reagents tested in both cone entry and cone exit labeling, an increased N2 gas-flow resulted in a higher deuterium uptake of Leu-Enk (Figure 3a). The presence of D2O, MeOD, and EtOD reagent in the cone region (cone entry labeling) resulted in a 3-fold reduction in S/N ratio, corresponding to a broadening of the isotopic envelope of Leu-Enk upon deuteration. The ND3/D2O reagent showed a somewhat higher decrease in S/N (approximately 10-fold). This observation coincided with the presence of peptide-ND3 adduct ions, providing a possible explanation for this additional 11870
dx.doi.org/10.1021/ac5035456 | Anal. Chem. 2014, 86, 11868−11876
Analytical Chemistry
Article
mediated gas-phase HDX-MS almost exclusively labels heteroatom bound non-amide hydrogens in peptides and proteins.14,15 The apparent useful property of the aqueous ND3/D2O reagent for gas-phase HDX-MS was further explored in cone exit labeling mode on the Synapt G2 and the Synapt G2-Si, Figure 3a. Cone exit labeling on the Synapt G2 and Synapt G2-Si showed similar maximal deuteration of Leu-Enk to cone entry labeling on the Synapt G1. The small HDX differences as a function of gas-flow for the two labeling configurations are likely the result of differences in flow control, dimensions and construction of the cone region of each QTOF instrument. In particular, the cone region on the G2-Si is different from the G1/G2, that is, having a different cone diameter, no extraction cone and the absence of a pump immediately downstream of the sample cone. As cone exit labeling required no modification of instrumental hardware while allowing similar gas-phase deuterium labeling at reduced reagent consumption and improved S/N, the utility of this setup for gas-phase HDXMS was studied further on a Synapt G2 instrument. A series of control experiments demonstrated that gas-phase deuterium uptake in the cone exit labeling configuration was (a) largely unaffected by changes in cone voltage and source T-Wave settings (Supporting Information Figure S-1) and (b) highly responsive (
