Front-End Electron Transfer Dissociation Coupled to a 21 Tesla FT-ICR Mass Spectrometer for Intact Protein Sequence Analysis
- 889 Downloads
High resolution mass spectrometry is a key technology for in-depth protein characterization. High-field Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) enables high-level interrogation of intact proteins in the most detail to date. However, an appropriate complement of fragmentation technologies must be paired with FTMS to provide comprehensive sequence coverage, as well as characterization of sequence variants, and post-translational modifications. Here we describe the integration of front-end electron transfer dissociation (FETD) with a custom-built 21 tesla FT-ICR mass spectrometer, which yields unprecedented sequence coverage for proteins ranging from 2.8 to 29 kDa, without the need for extensive spectral averaging (e.g., ~60% sequence coverage for apo-myoglobin with four averaged acquisitions). The system is equipped with a multipole storage device separate from the ETD reaction device, which allows accumulation of multiple ETD fragment ion fills. Consequently, an optimally large product ion population is accumulated prior to transfer to the ICR cell for mass analysis, which improves mass spectral signal-to-noise ratio, dynamic range, and scan rate. We find a linear relationship between protein molecular weight and minimum number of ETD reaction fills to achieve optimum sequence coverage, thereby enabling more efficient use of instrument data acquisition time. Finally, real-time scaling of the number of ETD reactions fills during method-based acquisition is shown, and the implications for LC-MS/MS top-down analysis are discussed.
KeywordsFTMS Ion-ion reaction Fourier transform mass spectrometry Electrospray ionization MS/MS MS2 Top-down ETD
Continued development of high-end technologies is needed to advance the field of top-down proteomics . Full characterization of proteoforms, not just transcribed gene products (RNA), is central to unraveling the most complex cellular processes. Intact protein analysis can provide structural information that cannot be obtained through traditional, bottom-up proteomics . For example, antibody-drug conjugates (ADCs), the most rapidly growing class of therapeutic compounds , require top-down analysis. Although these compounds are very effective in the treatment and tissue/cell specific targeting of various types of cancers , they can be highly heterogeneous  due to inherent problems in manufacturing. The heterogeneity arises from differences in antibody primary structure, the spatial and number distribution of drug conjugates , and the various linker species used for the antibody-drug tether. That heterogeneity cannot be captured appropriately with any combination of enzymatic approaches and necessitates that ADCs be analyzed intact to ensure that they are manufactured consistently and are safe for the public. Enzymatic approaches remove the inherent heterogeneity of the system and render extraction of the desired structural information problematic, even for the highest-performance commercial mass spectrometers. To achieve complete protein characterization, one must obtain primary sequence, site-localization of any post-translational modifications, and preferably higher-order structure characteristics. High-impact applications for top-down mass spectrometry, such as ADC characterization, should help drive the effort to close the analytical gap, in terms of proteins identified per experiment, quantification strategies, and diversity in protein vs. peptide chemistry, between bottom-up and top-down proteomics.
Closing the gap between top-down and bottom-up approaches will require advancement in sample preparation, front-end separations, and instrumentation. Thus, intact protein analysis continues to advance with the development of higher performance mass spectrometers  and new/versatile ion activation approaches [8, 9, 10, 11, 12, 13, 14]. With these advancements, routine implementation of top-down proteomics is an increasingly attainable goal. The inherently open-sourced technology and analytical merits of Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) makes it well suited for top-down protein characterization . State of the art ion introduction methods , cell design , operation , magnetic field strength , and homogeneity  provide contemporary FT-ICR MS instruments with excellent sensitivity and dynamic range. Former FT-ICR MS issues with cryogen supply and consumption have been obviated by the availability of superconducting magnets operating without liquid nitrogen and with essentially zero liquid helium consumption. Finally, FT-ICR MS is very easily coupled with orthogonal technologies such as laser-based [13, 20] or electron-based dissociation techniques as well as a host of other front-end hardware modifications.
Electron transfer dissociation (ETD) for top-down- or middle-down-size polypeptides has produced unprecedented sequence coverage [21, 22, 23]. Fragment ions produced by ETD retain labile post-translational modifications (PTMs) , and thus enable identification of their sequence location(s). However, fragmentation of intact proteins by ETD produces fragment ions spread over a relatively narrow m/z domain about the precursor. Product ion spectra exhibit very high spectral complexity surrounding the precursor m/z, often resulting in multiple overlapping product ion isotopic distributions. Consequently, these overlapping isotopic envelopes remain unresolved at even moderate resolving power. Ion–ion proton transfer (IIPT) reactions provide a route to simplify and better use the analytical m/z range; however, direct improvement in resolving power provides a similar effect. Also, the ETD process inherently produces a fairly uniform distribution of ions for each product ion channel; therefore, the precursor ion signal is distributed over dozens (or even hundreds in the case of proteins) of product ions. These aspects demand that top-down mass spectrometry instrumentation possess ultra-high mass resolution, broad dynamic range, and heightened sensitivity to take full advantage of data produced by ETD.
Technological advances in fragmentation efficiency and scan rate appropriate for LC time-scale have firmly cemented ETD as an indispensable tool for biomolecule analysis [25, 26, 27, 28, 29, 30]. A particularly valuable recent advance is the adaptation of ETD reagent ion source inside the front-end ionization source (FETD) . This manuscript details a multi-laboratory/industry collaboration to pair the complementary technology behind the high field FT-ICR MS instruments at the National High Magnetic Field Laboratory (NHMFL) with FETD. We are able to generate ETD product ion spectra with high sequence coverage for proteins up to 29 kDa with very little need for signal averaging. Further, we explore the relationship between acheivable sequence coverage and number of ETD fragment ion fills to an external ion accumulation device. Prediction of the minimum number of ETD fragment ion fills to achieve maximal sequence coverage for a given protein MW was optimized for on-line top-down LC-MS/MS performance.
All data were acquired with a 21 T FT-ICR mass spectrometer described in detail elsewhere . In brief, this hybrid system consists of a linear rf ion trap (Velos Pro; ThermoFisher Scientific) coupled with a custom-built FT-ICR mass analyzer. The Velos Pro is equipped with a commercial Orbitrap Fusion API inlet/FETD reagent  source (Thermo FisherScientific). The instrument control software is modified to control the ETD electronics. Q00 of the Velos is shortened to accommodate the reagent ion source (RIS) inside the existing source housing. The stacked ring ion guide (SRIG) DC offset voltage is used to repel ions produced by electrospray during reagent anion production within the RIS. A DC voltage was appropriated as the second of two DC potentials used to create a differential offset between adjacent rods of Q0 within the Velos to provide low-pass mass filtering of ions produced by the RIS. Owing to the flat-pole geometry of Q0, the mass resolution of this device is fairly poor and represents only half of the circuit required to constitute a mass-resolving quadrupole. Nevertheless, it provided adequate filtering of the higher m/z ions (>216 m/z) generated from the RIS to produce 90%–95% pure fluoranthene anions at the cost of about 50% drop in transmission efficiency (from ~1.5 ms to ~3.0 ms increase in reagent ion injection period). Alternatively, a broadband waveform can be applied to the linear ion trap prior to reaction, to remove the most common contaminant, the fluoranthene-nitrogen adduct species (m/z 216). In this implementation, we are able to use one or both approaches simultaneously to ensure gas-phase reagent ion purity. ETD reaction periods were normalized to roughly the time required for two half-lives of precursor decay (e.g., 45 ms for melittin and 8 ms for carbonic anhydrase).
Data Acquisition and Processing
Data were acquired with a dynamically harmonized ICR cell  by use of the standard LTQ-FT Ultra FTMS acquisition electronics (ThermoFisher Scientific), except for replacement of the preamplifier (Stahl Electronics, Mettenheim, Germany). Data acquisition for all direct infusion experiments were acquired with four microscans unless otherwise noted. Time-domain ICR transient duration was 0.76 s. Data was stored as the standard Thermo file .raw format, and transient data were stored as .dat files and processed with Predator  (8 Mword time-domain data, one zero-fill, magnitude mode). Externally calibrated FT-ICR mass spectra were deconvolved by use of Thrash or Xtract with a S/N threshold of 6. Neutral masses were submitted as input to Prosight Lite . Search tolerance for precursor mass and product ions was set to ±10 ppm. Fractional sequence coverage was determined directly from search results for each protein.
A set of standard proteins was used to test the relationship between number of fills and fractional sequence coverage: melittin, ubiquitin, cytochrome c, hemoglobin α subunit, apomyoglobin, histone H1, and carbonic anhydrase (molecular weight 3–30 kDa). All proteins were used as received (Sigma Aldrich, St. Louis, MO, USA) without further purification. Stock solutions were prepared for each protein in ultra-pure 18 MΩ water at a concentration of ~1 mM. Each stock solution was diluted directly into an electrospray solution (49:49:2; H2O:MeOH:HOAc) to a final protein concentration of 0.5 μM. We also prepared an equimolar mixture of three proteins at 0.5 μM each into electrospray solution (melittin, ubiquitin, and cytochrome c). Samples were directly infused (50 μm i.d., 360 μm o.d. polyimide-coated fused silica capillary, 500 nL/min flow rate) into the mass spectrometer by use of a custom nano-ESI source (University of Washington Proteomics Resource NSI source) at an ESI voltage of 1.8–2.5 kV. Capillary tips were produced by means of either an in-house developed grinding process or laser pulling (Sutter Instruments, Novato, CA, USA).
Results and Discussion
Front-End ETD Interfaced to a 21 T FT-ICR Mass Analyzer
The Velos Pro linear ion trap is equipped with electronics from a first generation ETD unit, such that the available DC outputs could be reassigned to the new FETD related devices, including a SRIG DC bias, a Q0 x-rod differential DC for low-pass mass filtering, and command current for the glow discharge. A DC bias applied to the SRIG allowed us to gate ions from the ESI source, such that ion injection from the RIS and the ESI source were decoupled. Application of a –50 V DC bias to the SRIG was capable of the rejection of all ions from the ESI source, whereas ion transmission from the RIS remained unaffected. During normal transmission from the ESI source, the SRIG bias was set to 0 V DC. Reagent ion purity remains a potential issue for implementing FETD with a Velos Pro system due to incomplete reagent ion isolation prior to ion–ion reaction. The main contributing contaminant species, the fluoranthene:nitrogen adduct (m/z 216), undergoes ion–ion reaction primarily through proton transfer. Although most good ETD reagents react with protein cations to a small degree via ion–ion proton transfer (or produces hydrogen loss), the mixing of two or more reagents with uncharacterized propensities for ETD/ion–ion proton transfer reaction pathways complicates product ion interpretation and lowers the yield of sequence-informative product ions. Therefore, two measures were taken to remove contaminant species from the reagent ion beam to achieve high quality ETD spectra. The first measure is to operate Q0 as a simple low-pass mass filter by application of a differential DC bias between adjacent rods. The planar Q0 electrode geometry limits performance but still achieves removal of 95%–99% of ions at or above the m/z of the fluoranthene:nitrogen adduct (m/z 216). Because the quadrupole geometry is not ideal for isolation, transmission efficiency was sacrificed for specificity (~50% loss in transmission efficiency). The second measure to ensure reagent ion purity is through the application of a broadband “clean-up” waveform. This waveform is constructed such that ions within a 20 Da window centered at the fluoranthene:nitrogen adduct are ejected from the linear ion trap prior to ion–ion reaction. ETD is a versatile technology that in principle may be coupled to any mass spectrometry platform, but benefits from hardware that places a true quadrupole mass filter  downstream from the RIS.
Molecular Weight and ETD Fragment Ion Fills
As discussed in depth by Riley et. al. , achieving maximum S/N is especially critical for top-down product ion mass spectra because the original precursor ion signal is distributed among many available fragmentation channels and fragment ion charge distributions. The two primary means for increasing S/N are spectral averaging and/or increasing the trapped ion population. Spectral averaging produces a gain in S/N that is proportional to n 1/2, in which n is the number of spectra averaged. However, for FTMS analyzers, spectral averaging is very costly in terms of time and may not always be appropriate, especially for on-line LC-MS/MS experiments. For this reason, it is preferable to improve S/N by increasing the ion population, which is possible with our hardware configuration. This can be done by means of a multiple ETD reaction fill scheme [22, 37] or by increasing ETD reaction capacity [36, 38]. Here, we demonstrate a multiple ETD reaction fill scheme (see Figure 1b). We repeat ion–ion reactions at relatively low precursor target ion number (typically 1E5 precursor and 2E5 reagent anion targets) and accumulate the reaction products in the MSD until maximum sequence coverage has been achieved. This device is estimated to have a storage capacity in excess of 50 million charges (see Supplemental Figure 1, 2) and has been designed to match or exceed the analytical charge capacity of the ICR cell such that there is no bottleneck in the number of ions/charges that can be supplied for mass analysis. Use of the MSD allows the ion trap assembly to be optimized as an ETD reaction vessel operated under conditions in which the pseudo-first order kinetic approximation holds true. Therefore, optimal and reproducible ETD fragmentation performance is attained. In this regard, the MSD effectively bridges the ion capacity gap between a single ETD reaction event in the linear rf ion trap and the ICR cell. It should be noted that the linear rf ion trap also has a very high charge capacity if operated as a storage-only device (tens of millions of charges), however, at those target values ETD performance is diminished.
From Figure 2a–g, we extracted the lowest number of fills at which sequence coverage is maximized (90–95 the maximum value observed) via curve-fit and plotted optimal fill number versus analyte MW, to reveal a linear relationship. Although this linear fit model represents the data well, we expect non-linearities based on protein tertiary structure, intact cysteine disulfide bonds, gas-phase intramolecular interactions, etc. Therefore, the linear model is a generalization that we shall apply to all proteins during LC-MS acquisition. It is worth noting that as we build up information from large-scale intact protein analysis, our model could be adapted to accommodate any observed non-linearity in the MW versus cumulative target based on sample type or other characteristics. Within this linear fit model, each data point represents the highest scan duty cycle without sacrificing sequence coverage for interrogated intact proteins. This model serves as a guide for more advanced data acquisition rules applied to LC-MS/MS ETD acquisition for top-down proteomics to more appropriately utilize instrument time as a function of the observed precursor MW.
Instrument-Controlled ETD Fragment Ion Fills
FETD Top-Down Analysis for Larger Proteins
This work was supported by NSF Division of Materials Research through DMR-11-57490, the State of Florida, and National Institutes of Health grants AI 033993 (DFH) and GM 037537 (DFH).
- 6.Behrens, C.R., Ha, E.H., Chinn, L.L., Bowers, S., Probst, G., Fitch-Bruhns, M., Monteon, J., Valdiosera, A., Bermudez, A., Liao-Chan, S., Wong, T., Melnick, J., Theunissen, J.W., Flory, M.R., Houser, D., Venstrom, K., Levashova, Z., Sauer, P., Migone, T.S., van der Horst, E.H., Halcomb, R.L., Jackson, D.Y.: Antibody-drug conjugates (ADCs) derived from interchain cysteine cross-linking demonstrate improved homogeneity and other pharmacological properties over conventional heterogeneous ADCs. Mol. Pharm. 12, 3986–3998 (2015)CrossRefGoogle Scholar
- 7.Hendrickson, C.L., Quinn, J.P., Kaiser, N.K., Smith, D.F., Blakney, G.T., Chen, T., Marshall, A.G., Weisbrod, C.R., Beu, S.C.: 21 Tesla Fourier transform ion cyclotron resonance mass spectrometer: a national resource for ultrahigh resolution mass analysis. J. Am. Soc. Mass Spectrom. 26, 1626–1632 (2015)CrossRefGoogle Scholar
- 12.Cannon, J.R., Cammarata, M.B., Robotham, S.A., Cotham, V.C., Shaw, J.B., Fellers, R.T., Early, B.P., Thomas, P.M., Kelleher, N.L., Brodbelt, J.S.: Ultraviolet photodissociation for characterization of whole proteins on a chromatographic time scale. Anal. Chem. 86, 2185–2192 (2014)CrossRefGoogle Scholar
- 20.Shaw, J.B., Li, W., Holden, D.D., Zhang, Y., Griep-Raming, J., Fellers, R.T., Early, B.P., Thomas, P.M., Kelleher, N.L., Brodbelt, J.S.: Complete protein characterization using top-down mass spectrometry and ultraviolet photodissociation. J. Am. Chem. Soc. 135, 12646–12651 (2013)CrossRefGoogle Scholar
- 21.Zhang, L., English, A.M., Bai, D.L., Ugrin, S.A., Shabanowitz, J., Ross, M.M., Hunt, D.F., Wang, W.H.: Analysis of monoclonal antibody sequence and post-translational modifications by time-controlled proteolysis and tandem mass spectrometry. Mol. Cell. Proteom. 15, 1479–1488 (2015)Google Scholar
- 23.Zheng, Y., Fornelli, L., Compton, P.D., Sharma. S., Canterbury, J., Mullen C., Zabrouskov, V., Fellers, R.T., Thomas, P.M., Licht, J.D., Senko, M.W., Kelleher, N.L.: Unabridged analysis of human histone H3 by differential top-down mass spectrometry reveals hypermethylated proteoforms from MMSET/NSD2 overexpression. Mol. Cell. Proteom. 15, 776–790 (2015)Google Scholar
- 24.Chi, A., Huttenhower, C., Geer, L.Y., Coon, J.J., Syka, J.E., Bai, D.L., Shabanowitz, J., Burke, D.J., Troyanskaya, O.G., Hunt, D.F.: Analysis of phosphorylation sites on proteins from Saccharomyces cerevisiae by electron transfer dissociation (ETD) mass spectrometry. Proc. Natl. Acad. Sci. U. S. A. 104, 2193–2198 (2007)CrossRefGoogle Scholar
- 35.Senko, M.W., Remes, P.M., Canterbury, J.D., Mathur, R., Song, Q., Eliuk, S.M., Mullen, C., Earley, L., Hardman, M., Blethrow, J.D., Bui, H., Specht, A., Lange, O., Denisov, E., Makarov, A., Horning, S., Zabrouskov, V.: Novel parallelized quadrupole/linear ion trap/Orbitrap tribrid mass spectrometer improving proteome coverage and peptide identification rates. Anal. Chem. 85, 11710–11714 (2013)CrossRefGoogle Scholar
- 36.Riley, N.M., Mullen, C., Weisbrod, C.R., Sharma, S., Senko, M.W., Zabrouskov, V., Westphall, M.S., Syka, J.E., Coon, J.J.: Enhanced dissociation of intact proteins with high capacity electron transfer dissociation. J. Am. Soc. Mass Spectrom. 3, 520–531 (2015)Google Scholar
- 39.Olsen, J.V., Schwartz, J.C., Griep-Raming, J., Nielsen, M.L., Damoc, E., Denisov, E., Lange, O., Remes, P., Taylor, D., Splendore, M., Wouters, E.R., Senko, M., Makarov, A., Mann, M., Horning, S.: A dual pressure linear ion trap Orbitrap instrument with very high sequencing speed. Mol. Cell. Proteom. 8, 2759–2769 (2009)CrossRefGoogle Scholar
- 40.Zhao, Y., Riley, N.M., Sun, L., Hebert, A.S., Yan, X., Westphall, M.S., Rush, M.J., Zhu, G., Champion, M.M., Mba Medie, F., Champion, P.A., Coon, J.J., Dovichi, N.J.: Coupling capillary zone electrophoresis with electron transfer dissociation and activated ion electron transfer dissociation for top-down proteomics. Anal. Chem. 87, 5422–5429 (2015)CrossRefGoogle Scholar