Anal. Chem. 2005, 77, 4549-4555
Radiolytic Modification and Reactivity of Amino Acid Residues Serving as Structural Probes for Protein Footprinting Guozhong Xu†,‡ and Mark R. Chance*,†,‡,§
Departments of Physiology & Biophysics and Biochemistry, and Center for Synchrotron Biosciences, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461-1602
Hydroxyl radical-mediated protein footprinting is a convenient and sensitive technique for mapping solventaccessible surfaces of proteins and examining the structure and dynamics of biological assemblies. In this study, the reactivities and tendencies to form easily detectible products for all 20 (common) amino acid side chains along with cystine are directly compared using various standards. Although we have previously reported on the oxidation of many of these residues, this study includes a detailed examination of the less reactive residues and better defines their usefulness in hydroxyl radical-mediated footprinting experiments. All 20 amino amides along with cystine and a few tripeptides were irradiated by γ-rays, the products were analyzed by electrospray mass spectrometry, and rate constants of modification were measured. The reactivities of amino acid side chains were compared based on their loss of mass spectral signal normalized to the rate of loss for Phe or Pro that were radiolyzed simultaneously to serve as internal standards. In this way, accurate quantitation of relative rates could be assured. A reactivity order of amino acid side chains was obtained as Cys > Met > Trp > Tyr > Phe > cystine > His > Leu, Ile > Arg, Lys, Val > Ser, Thr, Pro > Gln, Glu > Asp, Asn > Ala > Gly. Ala and Gly are far too unreactive to be useful probes in typical experiments and Asp and Asn are unlikely to be useful as well. Although Ser and Thr are more reactive than Pro, which is known to be a useful probe, their oxidation products are not easily detectible. Thus, it appears that 14 of the 20 side chains (plus cystine) are most likely to be useful in typical experiments. Since these residues comprise ∼65% of the sequence of a typical protein, the footprinting approach provides excellent coverage of the side-chain reactivity for proteins. Structural proteomics includes a multitude of methods to examine the structure of protein complexes.1-3 Structural mass * To whom correspondence should be addressed. Tel.: (718) 430 4136. Fax: (718) 430 8587. E-mail:
[email protected]. † Department of Physiology & Biophysics. ‡ Center for Synchrotron Biosciences. § Department of Biochemistry. (1) Sali, A.; Glaeser, R.; Earnest, T.; Baumeister, W. Nature 2003, 422, 216225. 10.1021/ac050299+ CCC: $30.25 Published on Web 06/08/2005
© 2005 American Chemical Society
spectrometry approaches using an array of chemical probes are proving to be a promising avenue to partial structural characterization, particularly with respect to the spatial and topological organization of protein complexes.4-24 Mass spectrometry has advantages of high sensitivity and throughput as well as capacity to capture the transient intermediates of dynamic processes.4,8,14,24 Typical chemical probing methods include hydrogen-deuterium (H/D) exchange,4,5 chemical cross-linking,9 selective chemical modification at amino acid residues such as Lys, Arg7 and Cys8, and nonselective covalent modification with high reactive carbine generated by photolysis of diazirine10,11 and the hydroxyl radical (2) Guan, J. Q.; Almo, S.; Chance, M. R. Acc. Chem. Res. 2004, 37, 221-229. (3) Russell, R. B.; Alber, F.; Aloy, P.; Davis, F. P.; Korkin, D.; Pichaud, M.; Topf, M.; Sali, A. Curr. Opin. Struct. Biol. 2004, 14, 313-324. (4) Hoofnagle, A. N.; Resing, K. A.; Ahn, N. G. Annu. Rev. Biophys. Biomol. Struct. 2003, 32, 1-25. (5) Garcia, R. A.; Pantazatos, D.; Villarreal, F. J. Assay Drug Dev. Technol. 2004, 2, 81-91. (6) Zappacosta, F.; Ingallinella, P.; Scaloni, A.; Pessi, A.; Bianchi, E.; Sollazzo, M.; Tramontano, A.; Marino, G.; Pucci, P. Protein Sci. 1997, 6, 1901-1909. (7) Suckau, D.; Mak, M.; Przybylski, M. Proc. Natl. Acad. Sci. U. S. A. 1992, 89, 5630-5634. (8) Apuy, J. L.; Park, Z. Y.; Swartz, P. D.; Dangott, L. J.; Russell, D. H.; Baldwin, T. O. Biochemistry 2001, 40, 15153-15163. (9) Sinz, A. J. Mass Spectrom. 2003, 38, 1225-1237. (10) Craig, P. O.; Ureta, D. B.; Delfino, J. M. Protein Sci. 2002, 11, 1353-1366. (11) Richards, F. M.; Lamed, R.; Wynn, R.; Patel, D.; Olack, G. Protein Sci. 2000, 9, 2506-2517. (12) Sharp, J. S.; Becker, J. M.; Hettich, R. L. Anal. Biochem. 2003, 313, 216225. (13) Sharp, J. S.; Becker, J. M.; Hettich, R. L. Anal. Chem. 2004, 76, 672-683. (14) Maleknia, S. D.; Ralston, C. Y.; Brenowitz, M. D.; Downard, K. M.; Chance, M. R. Anal. Biochem. 2001, 289, 103-115. (15) Guan, J. Q.; Vorobiev, S.; Almo, S. C.; Chance, M. R. Biochemistry 2002, 41, 5765-5775. (16) Kiselar, J. G.; Maleknia, S. D.; Sullivan, M.; Downard, K. M.; Chance, M. R. Int. J. Radiat. Biol. 2002, 78, 101-114. (17) Guan, J. Q.; Almo, S. C.; Reisler, E.; Chance, M. R. Biochemistry 2003, 42, 11992-12000. (18) Liu, R.; Guan, J. Q.; Zak, O.; Aisen, P.; Chance, M. R. Biochemistry 2003, 42, 12447-12454. (19) Kiselar, J. G.; Janmey, P. A.; Almo, S. C.; Chance, M. R. Mol. Cell. Proteomics 2003, 2, 1120-1132. (20) Kiselar, J. G.; Janmey, P. A.; Almo, S. C.; Chance, M. R. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 3942-3947. (21) Maleknia, S. D.; Chance, M. R.; Downard, K. M. Rapid Commun. Mass Spectrom. 1999, 13, 2352-2358. (22) Gupta, S.; Mangel, W. F.; McGrath, W. J.; Perek, J. L.; Lee, D. W.; Takamoto, K.; Chance, M. R. Mol. Cell. Proteomics 2004, 3, 950-959. (23) Guan, J. Q.; Takamoto, K.; Almo, S. C.; Reisler, E.; Chance, M. R. Biochemistry 2005, 44, 3166-3175. (24) Chance, M. R. Biochem. Biophys. Res. Commun. 2001, 287, 614-621.
Analytical Chemistry, Vol. 77, No. 14, July 15, 2005 4549
(HO•).12-24 H/D exchange takes place at backbone amides and probes the solvent accessibility of the protein main chain. Covalent modification overcomes many shortcomings of H/D exchange (in particular back exchange) and facilitates sample analysis. Selective chemical modification requires multiple reagents and experiments to probe a subset of amino acids, while nonselective modification by high reactive carbine and HO• radicals is able to provide excellent protein sequence coverage in a single experiment. HO• is proving to be a convenient and powerful species for probing the solvent accessibility of proteins with resolution at the singleresidue level; such radicals can be generated by Fenton’s reagent,12 photolysis of hydrogen peroxide,13 radiolysis of water using ionizing radiation such as X-rays, γ-rays, β-particles, or fast neutrons, and electrical discharge.21 Radiolysis of water provides the cleanest source of HO• without need for addition of chemicals and supports a wide variety of experimental conditions. Based on oxidative modification of proteins and subsequent mass spectroscopic analysis, protein footprinting has become a powerful technique for mapping solvent-accessible surfaces allowing both protein structure and the direct and allosteric interactions of proteins with ligands in a variety of aqueous conditions to be explored in details.14-24 Experimentally, aqueous solutions of protein samples at micromolar concentration are exposed to X-rays or γ-rays for various intervals of time. The solvent-accessible amino acid side chains undergo stable oxidation mediated by HO• generated by the radiolysis of water, while the direct impact of protein molecules by radiation is negligible at such low concentrations. Oxidized protein samples are then typically digested by proteases, and quantitative LC-MS and MS/ MS are used to determine the extents and sites of modification.15,16 The solvent accessibility of local sites is evaluated based on the rate constants of individual peptides,16,19 while ligand-binding sites and sites of conformational change upon binding are mapped through examining the changes in oxidation rates of target peptides.18,22-25 Understanding the fundamental chemistry of radiolytic modifications of amino acid residues is a prerequisite for developing protein footprinting to its fullest potential. Our previous studies have indicated that many aromatic, aliphatic, sulfur-containing, and charged residues are useful footprinting probes.26-29 The potential of an amino acid residue to serve as footprinting probe in typical aerobic conditions is determined first by its reactivity to HO• and second by the ability to detect stable oxidation products by mass spectroscopy. Radiolytic oxidation of amino acids has been extensively studied using pulse radiolysis combined with UV/ visible and electron paramagnetic resonance spectroscopic detection of free radical intermediates.30 The bimolecular rate constants of reaction of the 20 common amino acids and cystine with hydroxyl radical and hydrated electron at neutral (25) Rashidzadeh, H.; Khrapunov, S.; Chance, M. R.; Brenowitz, M. Biochemistry 2003, 42, 3655-3665. (26) Maleknia, S. D.; Brenowitz, M.; Chance, M. R. Anal. Chem. 1999, 71, 39653973. (27) Xu, G.; Takamoto, K.; Chance, M. R. Anal. Chem. 2003, 75, 6995-7007. (28) Xu, G.; Chance, M. R. Anal. Chem. 2004, 76, 1213-1221. (29) Xu, G.; Chance, M. R. Anal. Chem. 2005, 77, 2437-2449. (30) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref. Data 1988, 17, 513-886.
4550 Analytical Chemistry, Vol. 77, No. 14, July 15, 2005
Table 1. Rate Constants for Reaction of Amino Acids with Hydroxyl Radical and Hydrated Electron at Or near Neutral pHa,b HO•
eaq-1
substrate
rate (M-1 s-1)
pH
rate (M-1 s-1)
pH
Cys Trp Tyr Met Phe His Arg cystine Ile Leu Val Pro Gln Thr Lys Ser Glu Ala Asp Asn Gly
3.5 × 1010 1.3 × 1010 1.3 × 1010 8.5 × 109 6.9 × 109 4.8 × 109 3.5 × 109 2.1 × 109 1.8 × 109 1.7 × 109 8.5 × 108 6.5 × 108 5.4 × 108 5.1 × 108 3.5 × 108 3.2 × 108 2.3 × 108 7.7 × 107 7.5 × 107 4.9 × 107 1.7 × 107
7.0 6.5-8.5 7.0 6-7 7-8 7.5 6.5-7.5 6.5 6.6 ∼6 6.9 6.8 6.0 6.6 6.6 ∼6 6.5 5.8 6.9 6.6 5.9
1.0 × 1010 3.0 × 108 2.8 × 108 4.5 × 107 1.6 × 108 6.0 × 107 1.5 × 108 1.5 × 1010 n/ac His > Leu, Ile > Arg, Lys, Val > Thr, Pro > Glu, Gln. Reactivity of Side Chains Compared to a Proline Standard. The reactivities of amino acid residues were also investigated using Pro-NH2 as internal standard. Met and Tyr were not evaluated using Phe-NH2 as internal standard because of 16-Da mass differences with Phe, and the amino acid residues of very low reactivity were hardly modified when they were mixed with highly reactive Phe. Evaluating the reactivity based on two different internal standards also provides cross-checks for the data. For the nine amino acids examined using both Phe-NH2 and ProNH2 as internal standard, the relative values agree quite well (e.g., average 8% difference). The rate constants of 15 amino amides compared with ProNH2 and the ratio of rate constants (k/kF) are shown in Table 4A. Leu-NH2 and Ile-NH2 were not studied with Pro as internal standards, because their molecular mass is 16-Da higher than ProNH2 and cannot be distinguished from the +16-Da oxidation
Table 4. Rate Constants (External-Based) of (A) Amino Amides Relative to Pro-NH2 and (B) Peptides GMG and GCG Relative GWG A side chain
k
kp
k/kp
Trp Tyr Phe cystine His Arg Lys Val Thr Ser Pro Glu Gln Asn Asp Ala Gly
0.54 ( 0.06 0.084 ( 0.008 0.157 ( 0.006 0.19 ( 0.01 0.093 ( 0.007 0.088 ( 0.008 0.082 ( 0.005 0.032 ( 0.001 0.082 ( 0.008 0.049 ( 0.007
0.031 ( 0.004 0.007 ( 0.003 0.014 ( 0.003 0.019 ( 0.004 0.010 ( 0.002 0.030 ( 0.004 0.037 ( 0.003 0.017 ( 0.001 0.050 ( 0.005 0.035 ( 0.004
0.038 ( 0.006 0.037 ( 0.005 0.019 ( 0.004 0.027 ( 0.008 0.014 ( 0.002 0.002 ( 0.004
0.055 ( 0.006 0.056 ( 0.007 0.043 ( 0.004 0.064 ( 0.009 0.10 ( 0.01 0.052 ( 0.005
17 12 11 10 9.3 3.0 2.2 1.9 1.5 1.4 1.0 0.69 0.66 0.44 0.43 0.14 0.03
B peptide
k
k/kW
GWG GMG GCG
0.028 ( 0.003 0.033 ( 0.003 0.047 ( 0.004
1.0 1.2 1.7
product of Pro-NH2. The relative reactivity order is the same as that obtained using Phe as internal standard in Table 3. Trp is the most reactive side chain except for the sulfur-containing residues. The reactivity of Trp, Tyr, Phe, cystine, and His is about 17-, 12-, 11-, 10-, and 9.3-fold that of Pro. The reactivity of Arg, Lys, and Val is about 3.0-, 2.2-, and 1.9-fold that of Pro. The hydroxyl-containing side chains, Ser and Thr, also seem to be slightly more susceptible to radiolytic modification than Pro. The reactivity generally decreases for short aliphatic side chains. The reactivities of Gln, Glu, Asn, Asp, and Ala are about 69, 66, 44, 43, and 14% that of Pro. Reactivity of Sulfur-Containing Side Chains. Met and Cys are the two most reactive amino acid residues29 and are susceptible to significant secondary oxidation reactions from mild oxidizing reagents such as hydrogen peroxide generated during radiolysis.32 The tripeptides GCG, GMG, and GWG were analyzed to minimize steric effects on the reactivity measurement. The three peptides (44) Hawkins, C. L.; Davies, M. J. Biochim. Biophys. Acta 2001, 1504, 196219. (45) Hawkins, C. L.; Davies, M. J. J. Chem. Soc., Perkin Trans. 2 1998, 26172622. (46) Garrison, W. M. Chem. Rev. 1987, 87, 381-398. (47) Stadtman, E. R. Methods Enzymol 1995, 258, 379-393.
along with GAG were mixed together and exposed to γ-rays. Because the reactivities of Cys, Met, and Trp are much higher than Ala, no oxidation of GAG was found in the irradiated peptide mixture, and GAG was used as internal signal standard for the peptide mixture. The peptide mixture was analyzed right after exposure to minimize secondary reactions. The ratios of the mass spectral signal intensities of the three unmodified peptides to that of GAG were used to represent the relative amount of GCG, GMG, and GWG, while the fraction of unmodified peptide was calculated as the ratio of the relative amount of irradiated peptide to that of control. The rate constants for the three peptides were obtained by nonlinear fitting the fraction of unmodified peptide and corresponding exposure time to first-order kinetics; the results are shown in Table 4B. In situations with a minimum of secondary oxidation, the rate constants of GCG and GMG are 1.7- and 1.2fold that of GWG, indicating Cys is the most reactive amino acid residue followed by Met, which is 20% more reactive than Trp, the most reactive non-sulfur-containing amino acid residue. CONCLUSION The results indicate the reactivity and detection efficiency of 20 common amino acid side chains and cystine in order of Cys > Met > Trp > Tyr > Phe > cystine > His > Leu, Ile > Arg, Lys, Val > Thr, Ser, Pro > Glu, Gln > Asn, Asp > Ala > Gly. The reactivity order is consistent with the expectation that hydroxyl oxidation is preferred at sites where the radical can be stabilized by neighboring functional groups such as unsaturated bonds or electron-rich heteroatoms through electron delocalization and to a lesser degree by electron-releasing alkyl groups through electron donation to the electron-deficient radical centers.31,44-47 Ala and Gly are far too unreactive to be useful probes in typical experiments, and Asp and Asn are unlikely to be useful as well. Although Ser and Thr are more reactive than Pro, which is known to be a useful probe, their oxidation products are not easily detectable. Thus, it appears that 14 of the 20 side chains (plus cystine) are most likely to be useful in typical experiments. Since these residues comprise ∼65% of the sequence of a typical protein, the footprinting approach provides excellent coverage of the side chain reactivity of proteins. ACKNOWLEDGMENT This research is supported in part by The Biomedical Technology Centers Program of the National Institute for Biomedical Imaging and Bioengineering under P41-EB-01979. Received for review February 17, 2005. Accepted May 4, 2005. AC050299+
Analytical Chemistry, Vol. 77, No. 14, July 15, 2005
4555
