Characterizing Short-Lived Protein Folding Intermediates by Top

Sep 17, 2010 - ... Heidi Vitrac , Daniel Kavan , Sylvain Pichard , Daniel Gillet , Eric Forest , Vincent Forge ... Weidong Cui , Henry W. Rohrs , Mich...
0 downloads 0 Views 1MB Size
Download PDF
Anal. Chem. 2010, 82, 8591–8597

Characterizing Short-Lived Protein Folding Intermediates by Top-Down Hydrogen Exchange Mass Spectrometry Jingxi Pan,† Jun Han,‡ Christoph H. Borchers,‡ and Lars Konermann*,† Department of Chemistry, The University of Western Ontario, London, Ontario, N6A 5B7, Canada, and University of Victoria-Genome BC Proteomics Centre, Victoria, British Columbia, V8Z 7X8, Canada This work combines pulsed hydrogen/deuterium exchange (HDX) and top-down mass spectrometry for the structural characterization of short-lived protein folding intermediates. A custom-built flow device with three sequential mixing steps is used for (i) triggering protein folding, (ii) pulsed D2O labeling, and (iii) acid quenching. The earliest folding time point that can be studied with this system is 10 ms. The mixing device was coupled online to the electrospray source of a Fourier transform mass spectrometer, where intact protein ions are fragmented by electron capture dissociation (ECD). The viability of this experimental strategy is demonstrated by applying it to the refolding of horse apo-myoglobin (aMb), a reaction known to involve a transient intermediate. Cooling of the mixing device to 0 °C reduces the reaction rate such that the folding process occurs within the experimentally accessible time window. Top-down ECD provides an average spatial resolution of ca. 2 residues, surpassing the resolution typically achieved in traditional proteolytic digestion/HDX studies. Amide back exchange is virtually eliminated by the short (∼1 s) duration of the acid quenching step. The aMb folding intermediate exhibits HDX protection in helices G and H, whereas the remainder of the protein is largely unfolded. Marginal protection is seen for helix A. Overall, the top-down ECD approach used here offers insights into the sequence of events leading from the unfolded state to the native conformation, with envisioned future applications in the areas of protein misfolding and aggregation. The time-resolved experiments reported herein represent an extension of our previous work, where HDX/MS with top-down ECD was employed for monitoring “static” protein structures under equilibrium conditions (Pan et al. J. Am. Chem. Soc. 2009, 131, 12801). Understanding the mechanisms by which disordered proteins fold into their native structures continues to be one of the grand * To whom correspondence should be addressed. Phone: (519) 661-2111 ext. 86313. Fax: (519) 661-3022. E-mail: [email protected]. † The University of Western Ontario. ‡ University of Victoria. 10.1021/ac101679j  2010 American Chemical Society Published on Web 09/17/2010

challenges of biology.1 Fortunately, experimental and theoretical/ computational investigations have resulted in considerable progress over the past few years. The perplexing speed of folding has been attributed to funneled energy landscapes that direct the conformational motions of polypeptide chains toward the native state.2 Nonetheless, simplified pathway descriptions (e.g., the foldon model3) remain useful for interpreting these stepwise structural changes, thereby reconciling “old” and “new” views of the folding problem.4 Intriguing insights continue to emerge about the linkages between neurodegenerative disorders and protein misfolding.5 Kinetic folding experiments commence with the denaturation of isolated proteins in vitro, for example by exposure to acidic pH. Refolding is then triggered by a suitable change in solvent conditions. The ensuing structural transitions can be monitored by spectroscopic tools in a time-resolved fashion.6 When employed in conjunction with protein engineering methods, these experiments may be used for Φ-value analyses that provide information on transition states.2 One of the most direct approaches for tracking time-dependent structural changes during folding is the characterization of transient intermediates that are formed en route to the native state.7-9 Yet, the fact that intermediates typically only exist for fractions of a second makes it challenging to obtain detailed insights into their conformational properties. An additional complication is the fact that many intermediates do not become strongly populated.10 A careful choice of experimental conditions (pH, temperature, solvent additives, etc.) may therefore be required for detection of these semifolded species.11 (1) Service, R. F. Science 2008, 321, 784–786. (2) Fersht, A. R. Nat. Rev. Mol. Cell Biol. 2008, 9, 650–654. (3) Maity, H.; Maity, M.; Krishna, M. M. G.; Mayne, L.; Englander, S. W. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 4741–4746. (4) Lazaridis, T.; Karplus, M. Science 1997, 278, 1928–1931. (5) Bernstein, S. L.; Dupuis, N. F.; Lazo, N. D.; Wyttenbach, T.; Condron, M. M.; Bitan, G.; Teplow, D. B.; Shea, J.-E.; Ruotolo, B. T.; Robinson, C. V.; Bowers, M. T. Nat. Chem. 2009, 1, 326–331. (6) Bartlett, A. I.; Radford, S. E. Nat. Struct. Mol. Biol. 2009, 16, 582–588. (7) Hartl, F. U.; Hayer-Hartl, M. Nat. Struct. Mol. Biol. 2009, 16, 574–581. (8) Brockwell, D. J.; Radford, S. E. Curr. Opin. Struct. Biol. 2007, 17, 30–37. (9) Gianni, S.; Ivarsson, Y.; Jemth, P.; Brunori, M.; Travaglini-Allocatelli, C. Biophys. Chem. 2007, 128, 105–113. (10) Kato, H.; Vu, N.-D.; Feng, H.; Zhou, Z.; Bai, Y. J. Mol. Biol. 2007, 365, 881–891. (11) Krishna, M. M. G.; Lin, Y.; Mayne, L.; Englander, S. W. J. Mol. Biol. 2003, 334, 501–513.

Analytical Chemistry, Vol. 82, No. 20, October 15, 2010

8591

Pulsed amide hydrogen/deuterium exchange (HDX) is one of the premier strategies for studying short-lived folding intermediates.12-16 These experiments rely on the fact that D2O exposure can induce isotope exchange at backbone amide hydrogens. Deuteration occurs readily at sites that lack hydrogen bonding and that are accessible to the solvent. In contrast, tightly folded elements are protected against HDX. Exposure of the protein to a brief labeling pulse at well-defined time points during folding therefore reveals structural snapshots, akin to individual frames of a motion picture. Two-dimensional NMR spectroscopy has traditionally been the method of choice for monitoring HDX protection patterns.17,18 In recent years, however, most HDX studies have been conducted with electrospray ionization (ESI) mass spectrometry (MS) detection.19-21 Attractive features of ESI-MS include its greater sensitivity, conceptual simplicity, virtually unlimited size range, and the capability of distinguishing parallel from sequential folding pathways.15,22-24 In addition, recent work has demonstrated the applicability of HDX-MS to membrane proteins which are notoriously difficult to study by many other analytical techniques.25-29 HDX-MS usually involves a low temperature quenching step at acidic pH after isotope exchange. In most experiments the protein is then digested with a suitable protease, and information on the deuteration pattern is uncovered from the mass shifts of proteolytic peptides by liquid chromatography (LC)/MS.19 Limitations of the conventional HDX-LC/MS approach include the occurrence of isotope back exchange during digestion and LC.30 Moreover, the spatial resolution is often quite low, around 5-10 residues, as determined by the size of the proteolytic fragments.31 Recent work suggests that these shortfalls can be addressed through the use of top-down MS.32 Rather than relying on enzymatic digestion, top-down strategies involve the transfer of intact protein ions into the vacuum of the mass spectrometer, (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32)

Udgaonkar, J. B.; Baldwin, R. L. Nature 1988, 335, 694–699. Roder, H.; Elo ¨ve, G. A.; Englander, S. W. Nature 1988, 335, 700–704. Raschke, T. M.; Marqusee, S. Nat. Struct. Biol. 1997, 4, 298–304. Yang, H.; Smith, D. L. Biochemistry 1997, 36, 14992–14999. Konermann, L.; Simmons, D. A. Mass Spectrom. Rev. 2003, 22, 1–26. Uzawa, T.; Nishimura, C.; Akiyama, S.; Ishimori, K.; Takahashi, S.; Dyson, H. J.; Wright, P. E. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 13859–13864. Krishna, M. M. G.; Hoang, L.; Lin, Y.; Englander, S. W. Methods 2004, 34, 51–64. Engen, J. R. Anal. Chem. 2009, 81, 7870–7875. Eyles, S. J.; Kaltashov, I. A. Methods 2004, 34, 88–99. Englander, S. W. J. Am. Soc. Mass Spectrom. 2006, 17, 1481–1489. Tsui, V.; Garcia, C.; Cavagnero, S.; Siuzdak, G.; Dyson, H. J.; Wright, P. E. Protein Sci. 1999, 8, 45–49. Miranker, A.; Robinson, C. V.; Radford, S. E.; Aplin, R.; Dobson, C. M. Science 1993, 262, 896–900. Heidary, D. K.; Gross, L. A.; Roy, M.; Jennings, P. A. Nat. Struct. Biol. 1997, 4, 725–731. Hebling, C. M.; Morgan, C. R.; Stafford, D. W.; Jorgenson, J. W.; Rand, K. D.; Engen, J. R. Anal. Chem. 2010, 82, 5415–5419. Zhang, X.; Chien, E. Y. T.; Chalmers, M. J.; Pascal, B. D.; Gatchalian, J.; Stevens, R. C.; Griffin, P. R. Anal. Chem. 2010, 82, 1100–1108. Rey, M.; Mrzek, H.; Pompach, P.; Novk, P.; Pelosi, L.; Brandolin, G.; Forest, E.; Havlek, V.; Man, P. Anal. Chem. 2010, 82, 5107–5116. Busenlehner, L. S.; Salomonsson, L.; Brzezinski, P.; Armstrong, R. N. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 15398–15403. Joh, N. H.; Min, A.; Faham, S.; Whitelegge, J. P.; Yang, D.; Woods, V. L.; Bowie, J. U. Nature 2008, 453, 1266–1270. Wu, Y.; Kaveti, S.; Engen, J. R. Anal. Chem. 2006, 78, 1719–1723. Kaltashov, I. A.; Bobst, C. E.; Abzalimov, R. R. Anal. Chem. 2009, 81, 7892– 7899. Pan, J.; Han, J.; Borchers, C. H.; Konermann, L. J. Am. Chem. Soc. 2009, 131, 12801–12808.

8592

Analytical Chemistry, Vol. 82, No. 20, October 15, 2010

followed by gas phase fragmentation.33,34 Electron capture dissociation (ECD)35,36 represents a highly promising approach for top-down HDX-MS.37,38 This technique provides extensive backbone cleavage through formation of numerous c and z• fragment ions.39 Even more importantly, ECD preserves the solutionphase H/D pattern along the amide backbone during the fragmentation process.32,40,41 Similar experiments may also be conducted by electron transfer dissociation (ETD).42,43 A number of earlier studies have attempted to employ collision-induced dissociation (CID) for spatially resolved HDX-MS experiments.44-47 However, CID tends to induce intramolecular H/D migration (“scrambling”), which can make a proper interpretation of experimental data problematic.48-52 ECD and ETD, on the other hand, occur without any apparent scrambling as long as excessive collisional heating in the ion sampling interface is avoided.32,40,41 Apo-myoglobin (aMb, 153 residues) represents a widely used system for protein folding studies.53 NMR analyses54,55 have revealed that native aMb adopts an overall fold resembling that of the heme-containing holo-protein.56 Accordingly, native aMb forms seven R-helices referred to as A-E, G, and H. Notably, the F region encompassing residues 83-96 remains disordered in aMb.32,54 Refolding of acid-denatured sperm whale aMb has previously been shown to proceed through an intermediate that becomes populated within the millisecond dead-time of typical quench-flow experiments.57 Pulsed HDX/NMR studies revealed that this intermediate exhibits strongly developed hydrogen(33) Siuti, N.; Kelleher, N. L. Nat. Methods 2007, 4, 817–821. (34) Han, X.; Jin, M.; Breuker, K.; McLafferty, F. W. Science 2006, 314, 109– 112. (35) Zubarev, R. A.; Zubarev, A. R.; Savitski, M. M. J. Am. Soc. Mass Spectrom. 2008, 19, 753–761. (36) Cooper, H. J.; Hakansson, K.; Marshall, A. G. Mass Spectrom. Rev. 2005, 24, 201–222. (37) Charlebois, J. P.; Patrie, S. M.; Kelleher, N. L. Anal. Chem. 2003, 75, 3263– 3266. (38) Kweon, H. K.; Hakansson, K. Analyst 2006, 131, 275–280. (39) Kruger, N. A.; Zubarev, R. A.; Horn, D. M.; McLafferty, F. W. Int. J. Mass Spectrom. 1999, 185-187, 787–793. (40) Pan, J.; Jun, H.; Borchers, C. H.; Konermann, L. J. Am. Chem. Soc. 2008, 130, 11574–11575. (41) Rand, K. D.; Adams, C. M.; Zubarev, R. A.; Jorgensen, T. J. D. J. Am. Chem. Soc. 2008, 130, 1341–1349. (42) Abzalimov, R. R.; Kaplan, D. A.; Easterling, M. L.; Kaltashov, I. A. J. Am. Soc. Mass Spectrom. 2009, 20, 1514–1517. (43) Rand, K. D.; Zehl, M.; Jensen, O. N.; Jørgensen, T. J. D. Anal. Chem. 2009, 81, 5577–5584. (44) Kim, M.-Y.; Maier, C. S.; Reed, D. J.; Deinzer, M. L. J. Am. Chem. Soc. 2001, 123, 9860–9866. (45) Deng, Y.; Pan, H.; Smith, D. L. J. Am. Chem. Soc. 1999, 121, 1966–1967. (46) Abzalimov, R. R.; Kaltashov, I. A. Anal. Chem. 2010, 82, 942–950. (47) Anderegg, R. J.; Wagner, D. S.; Stevenson, C. L.; Borchardt, R. T. J. Am. Soc. Mass Spectrom. 1994, 5, 425–433. (48) Jørgensen, T. J. D.; Gårdsvoll, H.; Ploug, M.; Roepstorff, P. J. Am. Chem. Soc. 2005, 127, 2785–2793. (49) Demmers, J. A. A.; Rijkers, D. T. S.; Haverkamp, J.; Killian, J. A.; Heck, A. J. R. J. Am. Chem. Soc. 2002, 124, 11191–11198. (50) Ferguson, P. L.; Pan, J.; Wilson, D. J.; Dempsey, B.; Lajoie, G.; Shilton, B.; Konermann, L. Anal. Chem. 2007, 79, 153–160. (51) Percy, A. J.; Slysz, G. W.; Schriemer, D. C. Anal. Chem. 2009, 81, 7900– 7907. (52) Johnson, R. S.; Krylov, D.; Walsh, K. A. J. Mass. Spectrom. 1995, 30, 386– 387. (53) Eliezer, D.; Wright, P. E. J. Mol. Biol. 1996, 263, 531–538. (54) Eliezer, D.; Yao, J.; Dyson, H. J.; Wright, P. E. Nat. Struct. Biol. 1998, 5, 148–155. (55) Hughson, F. M.; Wright, P. E.; Baldwin, R. L. Science 1990, 249, 1544– 1548. (56) Evans, S. V.; Brayer, G. D. J. Mol. Biol. 1990, 213, 885–897. (57) Jennings, P. A.; Wright, P. E. Science 1993, 262, 892–896.

Figure 1. Continuous-flow mixing device for pulsed HDX/MS. Exposure of acid-unfolded protein to ammonium hydroxide solution triggers refolding at mixer M1. After a variable amount of folding time (tfold, ranging between 10 ms and 1 s), addition of D2O initiates the 20 ms labeling pulse at mixer M2. HDX is quenched by formic acid solution after M3 for 1 s, before the mixture is infused into the electrospray source of the mass spectrometer. Uncorrected pH-meter readings for D2O-containing solutions are referred to as pH*.

bonding in helices A, G, and H, as well as in portions of the B helix.17,22 The current work focuses on horse aMb (16952 Da), which differs from the sperm whale protein in 19 residues. The folding mechanism of horse aMb resembles that of the sperm whale variant.58 However, much less is known about the exact structural properties of the horse aMb folding intermediate. Our laboratory has previously demonstrated the use of rapid mixing devices in conjunction with ESI-MS for HDX-based protein folding studies.16,59 Here we develop this approach further, by reporting the first application of online pulsed HDX with top-down ECD-MS for the structural characterization of a short-lived folding intermediate, focusing on horse aMb as model system. The approach used here should be applicable to a wide range of protein systems. EXPERIMENTAL SECTION Materials. Holo-myoglobin from horse skeletal muscle and bovine ubiquitin were obtained from Sigma (St. Louis, MO). Heme-free protein (aMb) was prepared by butanone extraction under acidic conditions,60 followed by dialysis. Protein stock solutions were flash frozen in liquid nitrogen and stored at -80 °C. Bradykinin (Bachem, King of Prussia, PA) served as rapidly exchanging internal HDX standard.32 D2O was from Cambridge Isotope Laboratories (Andover, MA). Pulsed HDX. Protein refolding and pulsed HDX were carried out using a custom-made four-syringe continuous-flow setup with three sequential mixing steps (Figure 1). The basic design of this device is similar to systems described previously,61 adapted from quench-flow strategies originally developed for NMR spectroscopy.12,13 Syringe 1 was filled with an aqueous solution of 75 µM acid-unfolded aMb and 30 µM bradykinin in 10 mM HCl. Syringe 2 contained water with dilute ammonium hydroxide. Both syringes were connected to mixer M1 via fused silica capillaries (i.d. 100 (58) Uzawa, T.; Akiyama, S.; Kimura, T.; Takahashi, S.; Ishimori, K.; Morishima, I.; Fujisawa, T. Proc. Nat. Acad. Sci. U.S.A. 2004, 101, 1171–1176. (59) Pan, J. X.; Rintala-Dempsey, A.; Li, Y.; Shaw, G. S.; Konermann, L. Biochemistry 2006, 45, 3005–3013. (60) Teale, F. W. J. Biochim. Biophys. Acta 1959, 35, 543. (61) Konermann, L.; Collings, B. A.; Douglas, D. J. Biochemistry 1997, 36, 5554– 5559.

µm, Polymicro Technologies, Phoenix, AZ). Refolding was initiated by mixing the contents of syringes 1 and 2 at M1 in a 1:1 volume ratio at a flow rate of 5 µL min-1 each, resulting in a pH jump from 2 to 10. The outlet of M1 was connected to a “folding capillary”. Depending on the folding time t different capillary i.d. values and lengths l were used (10 ms: i.d. 20 µm, l 5 mm; 100 ms: i.d. 50 µm, l 8.5 mm; 1 s: i.d. 75 µm, l 38 mm). Data for t ≈ 1 h were obtained using the same setup as for 10 ms, however, instead of online mixing, both syringes were filled with the same refolded aMb solution at pH 10.0 prepared by manual mixing. For all time points, the outflow of the refolding capillary was mixed at M2 with ammonium hydroxide in D2O from syringe 3 (pH meter reading 10.0, 40 µL min-1). Downstream of M2 was a 8.5 mm “pulsed HDX capillary” (i.d. 50 µm), which provided an average labeling time of 20 ms in 80% (v/v) D2O. HDX was quenched at mixer M3 by exposure to acidic solution from syringe 4 (0.4% formic acid in acetonitrile/H2O/D2O in a 1: 1.8: 7.2 volume ratio, 25 µL min-1). The measured pH after this final mixing step was 2.0. This solvent composition ensures that the side chain deuteration downstream of M2 remains constant, thereby facilitating the subsequent data analysis. The outlet of M3 was connected to the mass spectrometer via a 20 cm fused silica capillary (i.d. 75 µm), followed by the stainless steel electrospray capillary (i.d. 100 µm, 9 cm for FT-MS and 15 cm for Q-TOF-MS, see below), resulting in a quenching time on the order of 1 s. Amide back exchange under these conditions is less than 1%.21 The total flow rate of the final mixture was 75 µL min-1 at a protein concentration of 5 µM. Acid-induced protein unfolding during quenching results in the formation of high electrospray charge states which enhance the electron capture cross section in ECD experiments.62 Note that under these conditions all ESI charge states of the acid-unfolded protein exhibit the same HDX characteristics.32 All four syringes were advanced continuously by syringe pumps (Harvard Apparatus, Holliston, MA). Prior to data acquisition the system was operated for 30 min in order to achieve equilibration. Some of the experiments were carried out at room temperature (22 ± 1 °C). Low temperature (0 °C) studies were conducted by immersing the entire mixing system, including all solvent delivery lines in an ice-water bath. The mixers M1-M3 were custom built as described elsewhere, with a dead-volume of approximately 3 nL.61 Mixer performance was tested by bromocresol purple discoloration.17 A number of continuous-flow devices that were assembled for this work did not pass this quality control and had to be discarded. Microfluidic mixers are available from various vendors. However, the suitability of those commercial devices was not tested in this work, and we relied on in-house-built mixers instead. The dependence of the intrinsic HDX rate constant on pH63 dictates that millisecond pulse-labeling has to be conducted under basic conditions.12,13 At pH ≈ 10 the labeling of unprotected amides occurs in the submillisecond range.63 It is possible to conduct refolding in near-neutral solution, followed by a change in pH concomitant with the labeling pulse.12,13,17 However, such a pH change can lead to complications as it may cause the protein (62) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998, 120, 3265–3266. (63) Bai, Y.; Milne, J. S.; Mayne, L.; Englander, S. W. Proteins: Struct. Funct. Genet. 1993, 17, 75–86.

Analytical Chemistry, Vol. 82, No. 20, October 15, 2010

8593

to deviate from its original folding pathway during the HDX step.11,64 For this reason the entire mixing sequence of the current work up until the quenching step was conducted at a constant pH of 10. Prerequisite for this strategy is knowledge about the behavior of the native state under basic conditions. In the case of aMb it has been demonstrated that the native state is stable at pH 10.65,66 For proteins that are more sensitive to a basic environment it is advisible to employ mixing sequences where the pH during the refolding step is closer to neutral. A less basic labeling pulse that results in somewhat lower intrinsic exchange63 will be sufficient for many applications.13 Mass Spectrometry. Global HDX experiments on intact aMb were conducted on a quadrupole-time-of-flight instrument (Q-TOF Ultima API, Waters, Manchester, UK), utilizing a standard Z-spray electrospray source at +3 kV. Top-down ECD data were acquired on a 12 T Apex-Qe hybrid Fourier transform (FT) mass spectrometer (Bruker Daltonics, Billerica, MA) equipped with an Apollo II electrospray source. The ion optics and parameter settings used for this instrument have been described previously.32 Approximately 2600 scans were accumulated over the m/z range 250-2800, corresponding to an acquisition time of ca. 60 min for each ECD spectrum. FT-MS calibration was performed with intact ions and ECD fragments of ubiquitin. Data Analysis. Of the 261 exchangeable hydrogens in neutral aMb 148 are backbone NH sites, 110 are on amino acid side chains and 3 on the termini. Charge states of the ECD fragments used for HDX analyses were as follows: 1+ for c3-c23 and z3-z8, 2+ for c17-c38 and z9-z18, 3+ for c40-c44 and z21-z29, 4+ for c45-c65 and z32, 5+ for z36-z44, z49, z57, 6+ for c86 and z55-56, 7+ for c69 - c78 and z61, 8+ for c83 and z70, 9+ for z67, and 10+ for c98 and z76. HX-Express67 was used for determining centroid m/z values (R) for all partially deuterated ECD fragments. Centroids of the corresponding unlabeled ions (R0) were obtained from ProteinProspector (http://prospector.ucsf.edu) based on the known protein sequence (pdb file 1wla).56 The number of protected hydrogens Nprot in each ECD fragment was calculated as

Nprot ) Ntot -

(

)

n(R - R0) -n P(mD - mH)

(1)

where Ntot is the total number of exchangeable hydrogens in each of the fragment ions, including amide backbone, side chain, and terminal sites, but excluding charge carriers. The charge state of the ion is denoted as n, and the atomic masses of deuterium and hydrogen are mD ) 2.0141 Da and mH ) 1.0078 Da, respectively. The deuteration level of the bradykinin internal standard (P) was determined by isotope modeling50 to be 77.5%, slightly below the expected value of 80%. This small deviation is compensated by including the normalization factor P in eq 1. For our analysis it is assumed that exchangeable side chain sites and termini are uniformly labeled to the same level P (64) Bieri, O.; Kiefhaber, T. J. Mol. Biol. 2001, 310, 919–935. (65) Kirby, E. P.; Steiner, R. F. J. Biol. Chem. 1970, 245, 6300–6306. (66) Weisbuch, S.; Ge´rard, F.; Pasdeloup, M.; Cappadoro, J.; Dupont, Y.; Jamin, M. Biochemistry 2005, 44, 7013–7023. (67) Weis, D. D.; Engen, J. R.; Kass, I. J. J. Am. Soc. Mass Spectrom. 2006, 17, 1700–1703.

8594

Analytical Chemistry, Vol. 82, No. 20, October 15, 2010

Figure 2. Partial ESI mass spectra of aMb, depicting the 17+ charge state. The data were obtained after exposing the protein to an HDX pulse at different time points during folding (10 ms, 100 ms, etc., as indicated). Spectra on the left were recorded at 22 ( 1 °C. Panels on the right represent data obtained at 0 °C. Dotted vertical lines indicate the approximate m/z positions of the intermediate and the native state after pulsed HDX. The experiments were conducted by coupling the flow-mixing device of Figure 1 to a Q-TOF mass spectrometer. Data for the t ) 1 h native control were obtained using manual mixing. Intact protein data (not shown) qualitatively similar to those depicted here were recorded by FT-MS.

as the internal standard.68 Transformation of the Nprot progressions using a previously developed strategy32 provides the backbone amide deuteration level D as a function of residue position. D ) 1 corresponds to sites that are completely deuterated (lack of protection), whereas D ) 0 for fully protected sites that exclusively carry hydrogen. Fractional values of D reflect partial exchange. The error of the D values determined in this way is on the order of ±0.1.32 RESULTS AND DISCUSSION Intact Protein Mass Distributions. Pulsed HDX/MS on horse aMb was carried out using a custom-built flow-mixing device that was directly coupled to the electrospray source of a mass spectrometer. This system employs three sequential mixing steps for (i) triggering refolding by a pH jump, (ii) pulsed HDX, and (iii) acid quenching (Figure 1). Importantly, the quenching step lasts for merely one second, thereby virtually eliminating amide back exchange. Mass distributions obtained for the intact protein reveal that folding at room temperature goes to completion on a time scale of ∼100 ms (Figure 2). Even for the earliest accessible point (10 ms) the spectrum is already dominated by proteins exhibiting HDX properties similar to the t ) 1 h native control. Yet, the presence of an intermediate is evident from a distinct (68) Wang, L.; Smith, D. L. Anal. Biochem. 2003, 314, 46–53.

Figure 3. Representative ECD fragments of aMb obtained after pulsed HDX for a folding time of 10 ms (top row, folding intermediate), and for a folding time of 1 h (bottom row, native control). The data were recorded on a 12 T FT mass spectrometer.

high mass shoulder in the 10 ms data (Figure 2, top left). Unfortunately, the predominance of native-like aMb at all time points precludes an ECD analysis of the intermediate at room temperature. The reaction rate can be dramatically reduced by embedding the flow-mixing system in a 0 °C ice bath. For a folding time of 10 ms at this temperature almost the entire aMb population resides in the intermediate conformation (Figure 2, top right). As discussed elsewhere,22 this behavior implies that the intermediate represents an obligatory step during folding. In other words, every single protein in the sample has to pass through this intermediate on its trajectory toward the native state. The resolution of the isotope exchange profiles in Figure 2 greatly exceeds that of earlier HDX/MS studies on sperm whale aMb where a commercial quench-flow system was used.22 The bimodal mass distribution at 100 ms/0 °C, reflects the coexistence of the intermediate with more tightly folded conformers. Notably, the HDX level of the low mass species at this time point differs from that of the native state. This observation reveals that the transition from the 10 ms intermediate to the fully folded protein involves more than a single step. Top-Down HDX. Electrosprayed aMb ions were subjected to top-down ECD on a Fourier transform (FT) mass spectrometer following pulsed HDX at various folding times. Fragmentation was conducted without precursor selection in order to enhance the signal-to-noise ratio. Our primary focus is on the 10 ms time point at 0 °C, where the intermediate is maximally populated. For comparison, we will also consider data for native aMb, obtained after t ) 1 h at 0 °C. HDX mass shifts could be determined for 36 c and 34 z• ions. Unprocessed data for these fragment ions show larger mass shifts for the intermediate than for the refolded control. Examples are depicted in Figure 3. A number of additional low intensity c and z• signals were detectable, but those fragments could not be considered for HDX analyses due to their limited S/N ratio. This behavior is due to the fact that partial deuteration reduces peak intensities because the signals are spread over a greater number of isotope peaks.69 The use of C-13/N-15 depleted protein samples has been suggested as a partial remedy for this problem.37,70 Nonetheless, even without (69) Slysz, G. W.; Percy, A. J.; Schriemer, D. C. Anal. Chem. 2008, 80, 7004– 7011.

isotope depletion the “useful” fragments in the case of our aMb experiments cover the entire protein sequence with an average spatial resolution of 2.2 residues. Top-down ECD therefore allows pinpointing the location of deuteration sites with a resolution which surpasses that of typical proteolysis-based HDX/MS experiments. From the mass shifts measured for the individual c and z• ions the number of protected hydrogens Nprot in each of the fragment ions can be determined on the basis of eq 1. Regions in the resulting plots (Figure 4) with a slope close to unity correspond to segments that are protected from isotope exchange due to H-bonding and/or solvent exclusion. Regions with a slope around zero represent unfolded segments that are not protected. Thus, from Figure 4 it is immediately apparent that the t ) 10 ms intermediate is mostly unstructured throughout the first ∼100 residues. In contrast, two strongly protected regions are apparent closer to the C-terminus (Figure 4B, red trace). For comparison, data for the t ) 1 h refolded control were included in Figure 4 (black symbols). Not surprisingly, Nprot is generally much higher for the refolded protein. Spatially-Resolved Deuteration Patterns. The progressions of amide deuteration values for folding times of t ) 10 ms and t ) 1 h are depicted in Figure 5. Note that the central region encompassing residues 79-99 is covered by both c and z• fragments for the aMb data sets analyzed here. Deuteration levels determined from the two ion types agree closely with each other, attesting to the internal consistency of the method used. The deuteration pattern measured for the refolded protein at t ) 1 h is consistent with the known secondary structure of native aMb (Figure 5).54-56 Helical regions are largely protected from pulsed HDX, whereas extensive deuteration takes place at the intervening loops and in the F-region. Partial protection in the C-D loop reflects hydrogen bonding seen in the X-ray structure of holo-myoglobin.56 Overall, the deuteration pattern in Figure 5 for the refolded protein is similar to that reported in our earlier top-down HDX study on native aMb.32 Slight differences are attributed to the labeling conditions used (20 ms of HDX at a pH meter reading of 10 here, vs 5 s of D2O exposure at near-neutral pH in ref 32). A dramatically different HDX pattern is seen for the folding intermediate (Figure 5, t ) 10 ms). Native-like protection is present for helix G and parts of helix H. The A helix is only marginally developed with deuteration levels around 0.8. The remainder of the protein sequence does not show any discernible protection at 10 ms. This behavior is quite different from that of sperm whale aMb where the intermediate is more structured, with protection in the A(B)GH regions.17 Interestingly, completion of the folding process occurs 1 order of magnitude more slowly for sperm whale aMb than for the horse variant under comparable temperature conditions.22 Nonnative interactions within the highly developed core of the sperm whale intermediate, requiring thermally activated “error repairs”, may be responsible for these different kinetics.71,72 The results of the current work are compatible with the idea that formation of a less structured intermediate (70) Bou-Assaf, G. M.; Chamoun, J. E.; Emmett, M. R.; Fajer, P. G.; Marshall, A. G. Anal. Chem. 2010, 82, 3293–3299.

Analytical Chemistry, Vol. 82, No. 20, October 15, 2010

8595

Figure 4. Number of protected hydrogen atoms (Nprot) in aMb c ions (a) and z• ions (b) determined by top-down ECD (eq 1). Data for the folding intermediate (t ) 10 ms) are in red, those for the native control (t ) 1 h) are in black. Blue lines represent the hypothetical scenario where all backbone amide hydrogens are protected, gaps are due to prolines in positions 37, 88, 100, and 120. Amide deuteration levels of Figure 5 were determined by transforming the data of this Figure as described.32

Figure 5. Deuteration level D of backbone amides for the aMb folding intermediate (top, refolding time t ) 10 ms), and for the refolded native control (bottom, t ) 1 h). The data were recorded at 0 °C. Open and closed symbols represent HDX data obtained from c and z• ions, respectively. Also shown is the R-helical secondary structure of native aMb, based on X-ray data for holo-myoglobin (pdb code 1wla).56 Helices are denoted as A-E, G, and H. The F-region in native aMb is unfolded. D values around zero correspond to protected backbone amides (not deuterated), whereas unprotected sites are characterized by D ≈ 1. Shown on the right is a visualization of the deuteration levels, mapped to the structure of holo-myoglobin.56 Color coding is as follows: blue, D < 0.33; green, 0.33 < D < 0.66; red D > 0.66.

in the case of horse aMb is an effective mechanism for circumventing kinetic traps during folding. CONCLUSIONS The results discussed above demonstrate how the combination of online pulsed HDX with top-down MS provides insights into the temporal sequence of events during protein folding. By characterizing the structure of a short-lived aMb intermediate it is possible to uncover which segments fold first, and which ones (71) Nishimura, C.; Dyson, H. J.; Wright, P. E. J. Mol. Biol. 2010, 396, 1319– 1328. (72) Bedard, S.; Krishna, M. M. G.; Mayne, L.; Englander, S. W. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 7182–7187.

8596

Analytical Chemistry, Vol. 82, No. 20, October 15, 2010

fold later as the protein navigates toward the native state on its multidimensional energy landscape. This time-resolved approach represents a significant extension of earlier work from our group32,40 as well as from other laboratories42,73 where top-down HDX/MS was used for studying protein structures under “static” (equilibrium) conditions. The ∼2 residue spatial resolution achieved here with top-down ECD exceeds that of typical LC/MS HDX studies. At the same time, back exchange is virtually eliminated by using a very short (∼1 s) quenching step immediately prior to ESI. An advantage of the classical LC/MS HDX strategy compared to the approach used (73) Stefanowicz, P.; Petry-Podgorska, I.; Kowalewska, K.; Jaremko, L.; Jaremko, M.; Szewczuk, Z. Biosci. Rep. 2009, 30, 91–99.

here is its virtually unlimited size range. However, ongoing improvements in FT-MS technology74 are paving the way toward top-down HDX studies on much larger biomolecular systems, including pharmaceutically interesting drug targets, complexes and membrane proteins. We envision that the pulsed HDX/topdown MS approach of this work will also be suitable for gaining a better understanding of the mechanisms by which misfolded proteins assemble into cytotoxic aggregates. Although this work employed nonselective ECD of entire ion populations, future experiments could be conducted with quadrupole selection for limiting the fragmentation to specific precursor ions. Such a (74) Schaub, T. M.; Hendrickson, C. L.; Horning, S.; Quinn, J. P.; Senko, M. W.; Marshall, A. G. Anal. Chem. 2008, 80, 3985–3990.

modification would allow the interrogation of individual conformers in a heterogeneous mix. ACKNOWLEDGMENT This work was supported by the Natural Sciences and Engineering Council of Canada, the Canada Foundation for Innovation, Genome Canada, Genome BC, and the Canada Research Chairs Program.

Received for review June 25, 2010. Accepted September 7, 2010. AC101679J

Analytical Chemistry, Vol. 82, No. 20, October 15, 2010

8597