Exploring the Cooperativity of the Fast Folding Reaction of a Small Protein Using Pulsed Thiol Labeling and Mass Spectrometry
2007; Elsevier BV; Volume: 282; Issue: 52 Linguagem: Inglês
10.1074/jbc.m706714200
ISSN1083-351X
AutoresSantosh Kumar Jha, Jayant B. Udgaonkar,
Tópico(s)Chemical Synthesis and Analysis
ResumoIt has been difficult to obtain directly residue-specific information on side chain packing during a fast (ms) protein folding reaction. Such information is necessary to determine the extent to which structural changes in different parts of the protein molecule are coupled together in defining the cooperativity of the overall folding transition. In this study, structural changes occurring during the major fast folding reaction of the small protein barstar have been characterized at the level of individual residue side chains. A pulsed cysteine labeling methodology has been employed in conjunction with mass spectrometry. This provides, with ms temporal resolution, direct information on structure formation at 10 different locations in barstar during its folding. Cysteine residues located on the surface of native barstar, at four different positions, remain fully solvent-accessible throughout the folding process, indicating the absence of any ephemeral nonnative structure in which these four cysteine residues get transiently buried. For buried cysteine residues, the rates of the change in cysteine-thiol accessibility to rapid chemical labeling by the thiol reagent methyl methanethiosulfonate appear to be dependent upon the location of the cysteine residue in the protein and are different from the rate measured by the change in tryptophan fluorescence. But the rates vary over only a 3-fold range. Nevertheless, a comparison of the kinetics of the change in accessibility of the cysteine 3 thiol with those of the change in the fluorescence of tryptophan 53, as well as of their denaturant dependences, indicates that the major folding reaction comprises more than one step. It has been difficult to obtain directly residue-specific information on side chain packing during a fast (ms) protein folding reaction. Such information is necessary to determine the extent to which structural changes in different parts of the protein molecule are coupled together in defining the cooperativity of the overall folding transition. In this study, structural changes occurring during the major fast folding reaction of the small protein barstar have been characterized at the level of individual residue side chains. A pulsed cysteine labeling methodology has been employed in conjunction with mass spectrometry. This provides, with ms temporal resolution, direct information on structure formation at 10 different locations in barstar during its folding. Cysteine residues located on the surface of native barstar, at four different positions, remain fully solvent-accessible throughout the folding process, indicating the absence of any ephemeral nonnative structure in which these four cysteine residues get transiently buried. For buried cysteine residues, the rates of the change in cysteine-thiol accessibility to rapid chemical labeling by the thiol reagent methyl methanethiosulfonate appear to be dependent upon the location of the cysteine residue in the protein and are different from the rate measured by the change in tryptophan fluorescence. But the rates vary over only a 3-fold range. Nevertheless, a comparison of the kinetics of the change in accessibility of the cysteine 3 thiol with those of the change in the fluorescence of tryptophan 53, as well as of their denaturant dependences, indicates that the major folding reaction comprises more than one step. To obtain an understanding of the cooperativity of the structural transitions accompanying the folding of unfolded protein to its unique native fold has been the central objective of many protein folding studies (1Dill K.A. Fiebig K.M. Chan H.S. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 1942-1946Crossref PubMed Scopus (436) Google Scholar, 2Klimov D.K. Thirumalai D. Fold. Des. 1998; 3: 127-139Abstract Full Text Full Text PDF PubMed Google Scholar, 3Chan H.S. Shimizu S. Kaya H. Methods Enzymol. 2004; 380: 350-379Crossref PubMed Scopus (163) Google Scholar, 4Reich L. Weikl T.R. Proteins. 2006; 63: 1052-1058Crossref PubMed Scopus (12) Google Scholar). Many protein folding reactions have been described as cooperative two-state U ⇌ N transitions, implying that native structure forms in a concerted all-or-none manner, with the formation of one contact facilitating the formation of many others (3Chan H.S. Shimizu S. Kaya H. Methods Enzymol. 2004; 380: 350-379Crossref PubMed Scopus (163) Google Scholar, 5Finkelstein A.V. Ptitsyn O.B. Protein Physics. Academic Press, London2002: 205-277Google Scholar). There is, however, a growing body of work suggesting that protein folding/unfolding transitions may be highly noncooperative and even be gradual structural transitions (6Swaminathan R. Nath U. Udgaonkar J.B. Periasamy N. Krishnamoorthy G. Biochemistry. 1996; 35: 9150-9157Crossref PubMed Scopus (49) Google Scholar, 7Holtzer M.E. Lovett E.G. D'Avignon D.A. Holtzer A. Biophys. J. 1997; 73: 1031-1041Abstract Full Text PDF PubMed Scopus (60) Google Scholar, 8Song J. Jamin N. Gilquin B. Vita C. Menez A. Nat. Struct. Biol. 1999; 6: 129-134Crossref PubMed Scopus (30) Google Scholar, 9Lakshmikanth G.S. Sridevi K. Krishnamoorthy G. Udgaonkar J.B. Nat. Struct. Biol. 2001; 8: 799-804Crossref PubMed Scopus (114) Google Scholar, 10Ahmed Z. Beta I.A. Mikhonin A.V. Asher S.A. J. Am. Chem. Soc. 2005; 127: 10943-10950Crossref PubMed Scopus (142) Google Scholar, 11Sadqi M. Fushman D. Munoz V. Nature. 2006; 442: 317-321Crossref PubMed Scopus (224) Google Scholar, 12Li H. Frieden C. Biochemistry. 2007; 46: 4337-4347Crossref PubMed Scopus (24) Google Scholar). For example, time-resolved anisotropy and fluorescence resonance energy transfer experiments, as well as NMR experiments, have shown that structure is lost incrementally during the denaturant-induced equilibrium unfolding of barstar (6Swaminathan R. Nath U. Udgaonkar J.B. Periasamy N. Krishnamoorthy G. Biochemistry. 1996; 35: 9150-9157Crossref PubMed Scopus (49) Google Scholar, 9Lakshmikanth G.S. Sridevi K. Krishnamoorthy G. Udgaonkar J.B. Nat. Struct. Biol. 2001; 8: 799-804Crossref PubMed Scopus (114) Google Scholar, 12Li H. Frieden C. Biochemistry. 2007; 46: 4337-4347Crossref PubMed Scopus (24) Google Scholar). High resolution probes, such as UV resonance Raman spectroscopy and NMR, indicate that the equilibrium unfolding of Trp-cage (10Ahmed Z. Beta I.A. Mikhonin A.V. Asher S.A. J. Am. Chem. Soc. 2005; 127: 10943-10950Crossref PubMed Scopus (142) Google Scholar), GCN4-like leucine zipper (7Holtzer M.E. Lovett E.G. D'Avignon D.A. Holtzer A. Biophys. J. 1997; 73: 1031-1041Abstract Full Text PDF PubMed Scopus (60) Google Scholar), CHABII (8Song J. Jamin N. Gilquin B. Vita C. Menez A. Nat. Struct. Biol. 1999; 6: 129-134Crossref PubMed Scopus (30) Google Scholar), and BBL (11Sadqi M. Fushman D. Munoz V. Nature. 2006; 442: 317-321Crossref PubMed Scopus (224) Google Scholar) is spatially decoupled and occurs in many steps. Nearly all evidence for gradual or multistep folding comes from equilibrium unfolding studies, where high resolution structural probes can be used easily. Kinetic evidence is needed (13Eaton W.A. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 5897-5899Crossref PubMed Scopus (96) Google Scholar, 14Huang F. Sato S. Sharpe T.D. Ying L. Fersht A.R. Proc. Natl. Acad. Sci. U. S. A. 2007; 104: 123-127Crossref PubMed Scopus (90) Google Scholar), especially evidence that provides structural information at the individual residue level, but this has been limited so far (15Sabelko J. Ervin J. Gruebele M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 6031-6036Crossref PubMed Scopus (284) Google Scholar, 16Osvath S. Sabelko J.J. Gruebele M. J. Mol. Biol. 2003; 333: 187-199Crossref PubMed Scopus (48) Google Scholar, 17Ma H. Gruebele M. Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 2283-2287Crossref PubMed Scopus (95) Google Scholar, 18Sinha K.K. Udgaonkar J.B. J. Mol. Biol. 2005; 353: 704-718Crossref PubMed Scopus (50) Google Scholar, 19Sinha K.K. Udgaonkar J.B. J. Mol. Biol. 2007; 370: 385-405Crossref PubMed Scopus (49) Google Scholar, 20Schanda P. Forge V. Brutscher B. Proc. Natl. Acad. Sci. U. S. A. 2007; 104: 11257-11262Crossref PubMed Scopus (135) Google Scholar). The major question that remains unaddressed is whether different regions of a protein form a structure in a synchronized or in an unsynchronized manner, during the major folding reaction of any protein. Pulsed cysteine labeling (SX) 3The abbreviations used are: SXcysteine labelingMMTSmethyl methanethiosulfonateESIelectrospray ionization provides direct structural information on the fate of individual residues during the folding process and has been shown to be an excellent probe for studying structure formation during the fast folding/unfolding reactions of several proteins at the level of individual side chains (21Ballery N. Desmadril M. Minard P. Yon J.M. Biochemistry. 1993; 32: 708-714Crossref PubMed Scopus (54) Google Scholar, 22Ha J.H. Loh S.N. Nat. Struct. Biol. 1998; 5: 730-737Crossref PubMed Scopus (50) Google Scholar, 23Ramachandran S. Rami B.R. Udgaonkar J.B. J. Mol. Biol. 2000; 297: 733-745Crossref PubMed Scopus (26) Google Scholar, 24Sridevi K. Udgaonkar J.B. Biochemistry. 2002; 41: 1568-1578Crossref PubMed Scopus (49) Google Scholar, 25Silverman J.A. Harbury P.B. J. Biol. Chem. 2002; 277: 30968-30975Abstract Full Text Full Text PDF PubMed Scopus (31) Google Scholar). In brief, side chains located in different parts of the protein structure are mutated to cysteine, one at a time, and the solvent accessibility of the individual cysteine thiol group to rapid chemical labeling is measured at different times of folding. The extent to which a particular cysteine residue is involved in structure formation at any time of refolding is reflected by the fraction of molecules in which the cysteine thiol gets labeled at that time. In this study, the pulsed SX methodology has been coupled with mass spectrometry for the first time to explore the cooperativity of the refolding reaction of barstar. cysteine labeling methyl methanethiosulfonate electrospray ionization The folding pathway of the small protein barstar has been characterized extensively (26Schreiber G. Fersht A.R. Biochemistry. 1993; 32: 11195-11203Crossref PubMed Scopus (135) Google Scholar, 27Shastry M.C. Agashe V.R. Udgaonkar J.B. Protein Sci. 1994; 3: 1409-1417Crossref PubMed Scopus (43) Google Scholar, 28Shastry M.C. Udgaonkar J.B. J. Mol. Biol. 1995; 247: 1013-1027Crossref PubMed Scopus (99) Google Scholar, 29Agashe V.R. Shastry M.C. Udgaonkar J.B. Nature. 1995; 377: 754-757Crossref PubMed Scopus (192) Google Scholar, 30Pradeep L. Udgaonkar J.B. J. Mol. Biol. 2002; 324: 331-347Crossref PubMed Scopus (42) Google Scholar, 31Pradeep L. Udgaonkar J.B. J. Biol. Chem. 2004; 279: 40303-40313Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar), and under strongly stabilizing conditions it can be represented as follows.U⇋1msIE→100msIL→100sNSCHEME 1 where IE represents a highly heterogeneous early intermediate, consisting of different structural forms, and is shown to be populated during the initial few milliseconds of refolding (28Shastry M.C. Udgaonkar J.B. J. Mol. Biol. 1995; 247: 1013-1027Crossref PubMed Scopus (99) Google Scholar, 30Pradeep L. Udgaonkar J.B. J. Mol. Biol. 2002; 324: 331-347Crossref PubMed Scopus (42) Google Scholar, 31Pradeep L. Udgaonkar J.B. J. Biol. Chem. 2004; 279: 40303-40313Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). A rapid equilibrium between U and IE is established before the major structural transition to the late intermediate IL occurs. IL has also been shown to be a heterogeneous ensemble of intermediates (32Bhuyan A.K. Udgaonkar J.B. Biochemistry. 1999; 38: 9158-9168Crossref PubMed Scopus (45) Google Scholar, 33Sridevi K. Juneja J. Bhuyan A.K. Krishnamoorthy G. Udgaonkar J.B. J. Mol. Biol. 2000; 302: 479-495Crossref PubMed Scopus (43) Google Scholar, 34Sridevi K. Lakshmikanth G.S. Krishnamoorthy G. Udgaonkar J.B. J. Mol. Biol. 2004; 337: 699-711Crossref PubMed Scopus (51) Google Scholar). Different members of the IE ensemble are populated in different solvent conditions; hence, the structural properties of IE appear different in different folding conditions (30Pradeep L. Udgaonkar J.B. J. Mol. Biol. 2002; 324: 331-347Crossref PubMed Scopus (42) Google Scholar, 31Pradeep L. Udgaonkar J.B. J. Biol. Chem. 2004; 279: 40303-40313Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). The extent of cooperativity present in the major structural transition (the IE to IL structural transition), which constitutes the fast phase of folding, is poorly understood at the individual residue level. Barstar offers itself as an interesting model system for this type of study, because the kinetics of the fast phase (i.e. of the IE to IL transition) are different when monitored using different probes, under some but not all folding conditions (29Agashe V.R. Shastry M.C. Udgaonkar J.B. Nature. 1995; 377: 754-757Crossref PubMed Scopus (192) Google Scholar, 31Pradeep L. Udgaonkar J.B. J. Biol. Chem. 2004; 279: 40303-40313Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). In this study, the pulsed SX methodology in conjunction with mass spectrometry has been applied to a library of 10 single Cys, single Trp-containing mutant forms of barstar, in which the single Cys residue is located at various buried and exposed locations of the protein. Fig. 1a shows the locations of the single tryptophan and the cysteine residues so introduced. For each mutant protein, the accessibility of the individual cysteine thiol group to a cysteine-specific labeling reagent, methyl methanethiosulfonate (MMTS), and the fluorescence of the sole tryptophan were measured at different times during the folding process. Only a 3-fold dispersion is seen in the site-specific rates of the fast change in cysteine thiol accessibility during folding, but a more detailed analysis of the dependence on urea concentration of the kinetics of the fast folding reaction measured at two different sites (Cys3 and Trp53) in one of the mutant proteins suggests that the fast folding reaction comprises more than one step. The method for the purification of barstar and its mutant variants has been described in detail previously (36Khurana R. Hate A.T. Nath U. Udgaonkar J.B. Protein Sci. 1995; 4: 1133-1144Crossref PubMed Scopus (62) Google Scholar). Ten different mutant variants, Cys3, Cys14, Cys25, Cys36, Cys40, Cys42, Cys62, Cys67, Cys82, and Cys89, each with a single Trp residue (Trp53) and a single Cys residue (at the residue position indicated in the name), were generated by site-directed mutagenesis (18Sinha K.K. Udgaonkar J.B. J. Mol. Biol. 2005; 353: 704-718Crossref PubMed Scopus (50) Google Scholar, 19Sinha K.K. Udgaonkar J.B. J. Mol. Biol. 2007; 370: 385-405Crossref PubMed Scopus (49) Google Scholar). Protein purity was checked by mass spectrometry using a Micromass Q-TOF Ultima mass spectrometer coupled with an ESI source. The masses determined for Cys3, Cys14, Cys25, Cys36, Cys40, Cys42, Cys62, Cys67, Cys82, and Cys89 were 10,232, 10,216, 10,232, 10,232, 10,232, 10,202, 10,190, 10,232, 10,246, and 10,216 Da, respectively. These masses indicate that the N-terminal methionine residue remained uncleaved during the expression of the mutant proteins. Protein concentrations were determined for all of the above mutant proteins by measuring the absorbance at 280 nm, using an extinction coefficient of 10,000 m–1 cm–1. Boric acid, EDTA disodium salt, MMTS, 5,5′-dithiobis(2-nitribenzoic acid), and cysteine·HCl was of ultrapure grade from Sigma. Urea (ultrapure grade) was from U. S. Biochemical Corp. Dithiothreitol (ultrapure grade) was obtained from Invitrogen, formic acid (GPR grade) was from BDH, and acetonitrile (HPLC grade) was from Qualigens. The native buffer used for all of the equilibrium and kinetic experiments was composed of 200 mm sodium borate and 1 mm EDTA at pH 9.2. The unfolding buffer was native buffer containing 6 m urea for all the kinetic experiments or 8 m urea for all of the equilibrium unfolding experiments at pH 9.2. Urea concentrations were determined from the measurement of the refractive index on an Abbe 3L refractometer from Milton Roy. All buffers and solutions were filtered through 0.22-μm filters and degassed before use. All of the experiments were carried out at 25 °C. MMTS-labeled protein was prepared by reaction of the protein in 8 m urea (unfolding buffer) at pH 9.2, with a 100-fold molar excess of MMTS for ∼5 min. The labeling reaction was quenched by the addition of a 1000-fold molar excess (to the protein) of cysteine·HCl (in 1% formic acid) to the reaction mixture. The addition of cysteine·HCl also decreased the pH of the solution to ∼2, which ensured that labeled protein did not lose any label. Following this, the labeled protein was separated from cysteine, urea, and other small molecules present in the reaction mixture by passing the protein through a Hi-Trap Sephadex G-25 desalting column on an Akta chromatography system. The extent of labeling was checked by mass spectrometry, and the protein was found to be >95% labeled as judged from the expected 46-Da increase in the mass of the protein. The stability of MMTS, the labeling reagent used in this study, depends upon the pH of the solution in which it is reconstituted; hence, its decomposition process imposes a time constraint on the use of a MMTS solution after its preparation. The rate of decomposition of MMTS was measured under various conditions (i.e. at pH 9.5 and 8, in water at pH 6 ± 0.5, and in aqueous solutions containing 0.6–2.0 m urea at pH 6.3 ± 0.5, as described earlier) (37Roberts D.D. Lewis S.D. Ballou D.P. Olson S.T. Shafer J.A. Biochemistry. 1986; 25: 5595-5601Crossref PubMed Scopus (203) Google Scholar). In brief, fresh solutions of dithiothreitol and 5,5′-dithiobis(2-nitribenzoic acid) were prepared in 20 mm Tris buffer at pH 8. Dithiothreitol was added to the solution containing excess 5,5′-dithiobis(2-nitribenzoic acid), and the concentration of the released TNB2– ions was determined by measuring the absorbance at 412 nm, using an extinction coefficient of 14,100 m–1 cm–1. The concentration of MMTS in solution was determined as a function of time by monitoring the loss of absorbance of the TNB2– ions at 412 nm after the addition of an aliquot of the solution to the reaction mixture containing dithiothreitol and excess 5,5′-dithiobis(2-nitribenzoic acid). MMTS was found to decompose with a time constant of ∼4 min at pH 9.5 and ∼60 min at pH 8 (data not shown). It is, however, fairly stable when reconstituted in water and aqueous solutions containing urea. It was found to decompose less than 5% in 3 h when reconstituted in aqueous solution containing 2.0 m urea at pH 6.5 ± 0.3 (data not shown). For all of the pulsed SX experiments, the MMTS solution was reconstituted in water containing 0.6–2.0 m urea at pH 6.3 ± 0.5 and was used within 2 h of its preparation. All pulsed SX experiments were carried out using a Biologic SFM-400 Q/S unit operating in the quenched flow mode. Both refolding and labeling reactions were performed at pH 9.2 and 25 °C. The protein was unfolded in 6 m urea (unfolding buffer) for at least 3 h prior to refolding experiments. After a variable time of refolding, a 4 ms pulse of MMTS label was applied. The labeling reaction was quenched by the addition of excess cysteine in 1% formic acid. Three different quenched flow programs were used to achieve the refolding times of 0, 5–23, and >23 ms. For the 0 ms refolding time point, 30 μl of 140.4 mm MMTS (in water) were mixed with 330 μl of refolding buffer inside the quenched flow machine for 5 ms, and the resulting solution was mixed with 40 μl of unfolded protein solution (150 μm stock) for 4 ms. The labeling reaction was quenched with the addition of 103 μlof 410 mm cysteine·HCl (in 1% formic acid) solution. For refolding times in the range of 5–23 ms, refolding was initiated by mixing 24 μl of unfolded protein solution (195 μm stock) with 208 μlof refolding buffer at pH 9.2 and 25 °C in various delay loops to achieve the desired refolding time points, and then the resulting solution was pulsed with 80 μl of 40.9 mm MMTS solution (in water containing 0.6 m urea) for 4 ms. The labeling reaction was quenched with the addition of 80 μl of 410 mm cysteine·HCl (in 1% formic acid) solution. To achieve refolding times greater than 23 ms, refolding was initiated by mixing 12 μl of unfolded protein solution (195 μm stock) with 104 μl of refolding buffer, at pH 9.2 and 25 °C in a 90-μl delay loop (total intermixer volume = 116 μl). After a variable delay time, a pulse consisting of 40 μl of 40.9 mm MMTS (in water containing 0.6 m urea) was applied for 4 ms. The labeling reaction was quenched with the addition of 40 μl of 410 mm cysteine·HCl (in 1% formic acid) solution. The concentration of protein at the time of labeling was 15 μm in all of the above pulsed SX experiments. The concentration of MMTS at the time of labeling was 10.5 mm, except in the experiments where the dependence on MMTS concentration of the cysteine accessibility-monitored refolding kinetics was studied. In those experiments, an identical protocol was followed except that the calculated amount of MMTS solution (9.75 m stock solution) was dissolved in water containing 0.6 m urea, so as to give the desired MMTS concentration at the time of labeling. The concentration of cysteine·HCl in all of the above experiments was 10-fold higher than that of MMTS at the time of quenching. In the experiments where the dependence on urea concentration of the cysteine accessibility monitored refolding kinetics was studied, a protocol identical to that described above was followed, except that the refolding buffer contained the calculated amount of urea, so as to give the desired urea concentration at the time of refolding. Also, the MMTS solution was reconstituted in water containing the desired urea concentration. Control experiments were done to ensure that the concentrations of MMTS and cysteine·HCl used in the above experiments were sufficient to fully label the unfolded protein and quench the labeling reaction, respectively. All of the pulsed SX experiments were completed within 2 h of the preparation of MMTS solution. All samples were processed in an identical manner. Each sample was desalted on an Akta chromatography system, using a Hi-Trap Sephadex G-25 desalting column. MilliQ water at pH 3 (pH adjusted with formic acid) was used for elution. A control experiment was performed to ensure that there was no cross-contamination between two samples during desalting. Five samples collected at the same time point of refolding, when desalted one after the other, gave the same composition of labeled and unlabeled protein (within ±3%), as determined by mass spectrometry (data not shown). A high concentration of free cysteine was present in the samples after the pulsed SX experiments. It was conceivable that the labeled protein might be reduced if the samples were not desalted for a long time. The following control experiment was done to determine the time frame within which the samples had to be desalted. Samples were collected in duplicate after the pulsed SX experiments. One set (Set A) was desalted right after the pulsed SX experiment (within 2 h), and another set (Set B) was desalted 15 h later. The amount of labeled protein in each sample of Set B is found to be only 1–2% less than the corresponding sample of Set A. Moreover, no peaks corresponding to the addition of the cysteine moiety to the unlabeled protein were observed in any of the mass spectra in any of the above experiments. To be safe, desalting was completed within 2 h after the pulsed SX experiments for all samples. The extents of labeling in samples from the pulsed SX experiments were determined using ESI-mass spectrometry. A Micromass Q-TOF Ultima mass spectrometer coupled with an ESI source, which was operated under Mass Lynx software control, was used. For acquisition of the mass spectra, the capillary and cone voltages were maintained at 3 kV and 80 V, respectively, the desolvation temperature was set to 150 °C, and the source temperature was set to 80 °C. Samples collected after desalting were mixed with acetonitrile (containing 0.2% formic acid) in a 1:1 ratio and were infused into the mass spectrometer using a syringe pump (Harvard Apparatus, Holliston, MA) at a flow rate of 10 μl/min. All of the spectra were collected in the positive ion mode. The concentration of protein in each sample was typically 2–3 μm, and typically an ion count of ∼150 was obtained in a 1 s data acquisition window. Instrument calibration was achieved with a separate injection of horse heart myoglobin. Typically, a mass spectrum consisting of a series of multiply charged peaks corresponding to the masses of the two protein species (unlabeled and MMTS-labeled protein) was observed in each 1-s scan. For each sample, the data acquired over 60 s were averaged. All of the resultant m/z spectra were processed in the following way using the Mass Lynx version 4.0 software. Background noise subtraction was done using a second order polynomial below 30% of the curve with a tolerance value of 0.01, followed by a two-point smoothening with a Savitzky-Golay algorithm (supplied with the Mass Lynx software) using a smoothening window (in channels) of ±23. The extent of labeling was determined from these smoothened m/z spectra by calculating the average relative ion intensity of the labeled protein from the ninth, tenth, and eleventh charged state peaks (these were the three most intense peaks in the mass spectra). Folding was monitored using the change in fluorescence of Trp53 as a probe. All equilibrium unfolding experiments were performed on a Fluoromax-3 fluorimeter (Jobin Yvon). The protein samples were incubated in different concentrations of urea for at least 3 h prior to the fluorescence measurements. Identical results were obtained if the time of incubation was 24 h. Excitation of tryptophan fluorescence was carried out at 295 nm, using a slit width of 0.5 nm. Emission was monitored at 320 nm using a slit width of 10 nm. The final protein concentrations in equilibrium unfolding experiments were 3–5 μm. All kinetic experiments were carried out using a Biologic SFM-4 stopped-flow machine. Proteins were unfolded in 6 m urea (unfolding buffer) for at least 3 h prior to the refolding experiments. Refolding was initiated by mixing 30 μl of unfolded protein with 270 μl of refolding buffer inside the stopped-flow mixing module. Sample excitation was carried out at 295 nm, and emission was monitored at 320 nm using an Oriel bandpass filter with a bandwidth of ±10 nm. In all experiments, a mixing dead time of 1.8 ms was achieved by using an FC-08 cuvette with a path length of 0.8 mm and a total flow rate of 5 ml/s. The final protein concentrations in fluorescence-monitored kinetic refolding experiments were 15–25 μm. In the experiments where the dependence of the fluorescence-monitored refolding kinetics on urea concentration was studied, an identical protocol as above was followed, except that refolding was initiated by appropriate dilution of refolding buffer, unfolding buffer, and unfolded protein inside the stopped-flow mixing module, so as to give the final desired urea concentration at the time of refolding. Determination of the Bimolecular Rate Constants for MMTS Labeling of a Cysteine Thiol in a Protein—The exchange reaction (SX) between a thiol labeling reagent and a protected thiol group of a protein can be modeled by a Linderstrom-Lang type of equation (22Ha J.H. Loh S.N. Nat. Struct. Biol. 1998; 5: 730-737Crossref PubMed Scopus (50) Google Scholar, 24Sridevi K. Udgaonkar J.B. Biochemistry. 2002; 41: 1568-1578Crossref PubMed Scopus (49) Google Scholar), which was conceptualized originally to explain amide-hydrogen exchange phenomena in proteins. The exchange reaction between a thiol labeling reagent and a protected thiol group of a protein can be modeled as follows. Closed(-S-H)⇋kclosedkopenOpen(-S-H)→[MMTS]kbExchanged(-S-S-CH3)SCHEME 2 A cysteine-thiol protected in the protein structure can get labeled with MMTS only when a structural opening reaction (i.e. local or global unfolding) exposes that thiol transiently to the solvent. In the above scheme, kopen and kclosed are the kinetic rate constants for opening and closing of a cysteine thiol residue in the closed-to-open reaction, and kb is the second order rate constant of the reaction of that thiol group with MMTS in the unfolded protein. Under steady state conditions, the observed rate constant of exchange of the thiol in the closed state is given by the following. kex=kopen×kb[MMTS]kclosed+kb[MMTS](Eq. 1) Two limiting cases of Equation 1 exist, depending upon the relative rates of the closing reaction (kclosed) and of chemical exchange from the open unfolded state (kb[MMTS]). If kclosed kb[MMTS], then the following is true. kex=kopen(Eq. 2) This is known as the SX1 limit, and under this condition, kex measures the rate of structural opening in the closed-to-open reaction. On the other hand, if kclosed kb[MMTS], then the following is true. kex=kopen×kb[MMTS]kclosed=Kopen×kb[MMTS](Eq. 2) This is known as the SX2 limit, and under this condition, kex measures the equilibrium constant between the closed and open states. The above two mechanisms can be distinguished, because the rate of labeling for the SX2 mechanism is dependent on the concentration of MMTS, whereas it is not for the SX1 mechanism. Analysis of the Equilibrium Unfolding Data—All equilibrium unfolding transition curves were analyzed using a two state N ⇌ U model (38Agashe V.R. Udgaonkar J.B. Biochemistry. 1995; 34: 3286-3299Crossref PubMed Scopus (211) Google Scholar). Raw data were converted into fraction unfolded versus [urea] plots, and t
Referência(s)