Created: 1st June 1999, last updated: 12th July 1999, © 1999 ABRF

RESEARCH COMMUNICATIONS

 

Trapping and Identification of Folding Intermediates of Disulfide Bond-Forming Proteins Based on Cyanylation, Cleavage, and Analysis by Mass Spectrometry

Jack Throck Watson, Ying Yang, and Jiang Wu

Biochemistry Department, Michigan State University, East Lansing, MI

Reprinted from the electronic version of JBT available at http://www.abrf.org/JBT/JBT.html accession #0014.

Address correspondence and reprint requests to Jack Throck Watson, Biochemistry Building, Michigan State University, East Lansing, MI 48824-1319 (email: watsonj@pilot.msu.edu).

 

A new method for trapping folding intermediates is based on cyanylation of free sulfhydryl groups. Cyanylation of sulfhydryl groups by cyanodiaminopyridinium tetrafluoroborate (CDAP) in acidic media prevents disulfide bond exchange and terminates any folding that relies on disulfide bond formation. Exposing the cyanylated protein to basic conditions causes cleavage of the peptide bond on the amino terminal side of the cyanylated cysteines, producing an iminothiazolidine (itz)-blocked peptide for each free cysteine plus one truncated peptide. Knowledge about the sequence of the original protein allows deduction of the location of free cysteines in the original molecule based on analysis of the peptide fragments by mass spectrometry using matrix-assisted laser desorption ionization (or electrospray). For intermediates containing more than one disulfide bond, partial reduction of the trapped intermediate followed by additional cyanylation, cleavage, and mass-mapping of the fragments allows deduction of the pairing of the disulfide bonds in the intermediate in most cases. (J Biomol Tech 1999;10:75-81)

Key words: protein folding, cyanylation, mass spectrometry, specific cleavage, disulfide bond pairing.

A good trapping agent for collection of folding intermediates of proteins that rely on disulfide bond formation should block the free sulfhydryl groups quickly, completely, and without modifying the protein at sites other than at the sulfhydryl functionality.1 The pioneering work in this area was carried out by Creighton et al.,1 who studied the bovine pancreatic trypsin inhibitor (BPTI) by trapping with iodoacetate. They found eight well-populated 1- and 2-disulfide intermediates, of which five contained exclusively native disulfide bonds, and those that adopted nonnative disulfides were thought to be kinetically important intermediates. The BPTI model was reexamined using modern separation methodologies, and it was concluded that there existed only five species of well-populated intermediates (two 1-disulfide and three 2-disulfide species), all of which contained only native disulfide bonds.3,4 The inconsistency between the results of the original and later studies is thought to arise largely from differences in the methods of trapping the intermediates.

In general, conventional approaches for recognition of disulfide bond structures of folding intermediates is tedious and cumbersome, especially for large proteins for which proteolytic digestion produces a large number of fragments that are irrelevant to the disulfide bond linkage, and impractical or unsuitable for proteins containing closely spaced or adjacent cysteines, for which proteolytic or chemical degradation cannot achieve cleavage between such residues.5,6

We have developed a new method for assigning disulfide bond pairings in proteins by using the procedure of partial reduction, cyanylation, cleavage, and mass mapping.7 Briefly, our procedure subjects a denatured protein to limited chemical reduction to produce a mixture of singly reduced protein isomers. The nascent sulfhydryls are then cyanylated, and the resulting isomers are separated by high-performance liquid chromatography (HPLC). Under alkaline conditions, cleavage of the peptide bond occurs on the N-terminal side of cyanylated cysteine to form truncated peptides that, after reduction of the remaining disulfide bonds, can be mass mapped by desorption ionization mass spectrometry.8 The masses of the fragments can be related to the location of the paired cysteines that had undergone reduction, cyanylation, and cleavage. The methodology minimizes disulfide bond scrambling and is applicable to proteins containing closely spaced or adjacent cysteine residues.

We assessed the suitability of our cyanylation methodology for trapping disulfide-bond forming intermediates during refolding experiments with recombinant human epidermal growth factor (hEGF). Our preliminary results9 agree well with those reported by Chang et al.10

An important aspect of our methodology is that the cyanylation reaction that traps the intermediates is also the first step of our analytic procedure leading to structural characterization of the protein (Fig. 1).7 The key steps of our trapping and identification procedure are represented in Figure 2 and are described in the context of characterizing one of the intermediates in the refolding of hEGF.

FIGURE 1. The structure of human epidermal growth factor.

 

FIGURE 2. Flow chart of trapping and identification procedures.

 

MATERIALS AND METHODS

Materials

Recombinant hEGF was obtained from the Protein Institute (Broomall, PA). The purity of the protein was greater than 97%. Tris(2-carboxyethyl)phosphine (TCEP) hydrochloride was purchased from Pierce Chemical Co. (Rockford, IL). Guanidine hydrochloride was obtained from Boehringer-Mannheim Biochemicals (Indianapolis, IN); citric acid, sodium citrate, hydrochloric acid, acetic acid, and 1-cyano-4-dimethylamino-pyridinium tetrafluoroborate (CDAP) were purchased from Sigma and used without further purification. Acetonitrile and trifluoroacetic acid (TFA) were of HPLC grade. The TCEP solution in 0.1 M citrate buffer at pH 3.0 was prepared as 0.10 M stock solution and stored under nitrogen at -20°C for weeks with little deterioration. The 0.10 M CDAP solution in 1 M citrate buffer at pH 3.0 was freshly prepared before use.

Protocol for Refolding of the Protein

Recombinant hEGF (1.0 mg) was dissolved in 0.2 mL of 0.1 M citrate buffer (pH 3.0) containing 6 M guanidine HCl and 0.1 M TCEP reducing agent. The denaturation and reduction of hEGF were carried out at 37°C for 2 hours. The reaction mixture was separated by HPLC, and the fraction corresponding to the reduced hEGF was collected, dried, and stored at -80°C.

The refolding of reduced and denatured protein was initiated by diluting the protein sample with 0.05 M Tris-HCl buffer (pH 8.5) to a final protein concentration of 1 mg/mL. The protein was subjected to folding in open air to provide a valid basis for comparing our results with those of Chang et al.10

Trapping of Folding Intermediates

Folding intermediates were trapped in a time course manner by removing aliquots (0.1 mL) of the protein solution and mixing them with 10 µL of 0.5 M HCl solution containing freshly prepared 0.2 M CDAP. The pH of the solution was adjusted to 3, if necessary, and cyanylation of free sulfhydryl groups by CDAP proceeded at room temperature for 15 minutes. The trapped intermediates were immediately separated by HPLC. The fractions were collected manually and analyzed by MALDI-TOF MS. Those with 0-, 52-, 104-, and 156-dalton increases over the mass of the intact protein correspond to the 3-disulfide (nonnative [III] or native [N]), 2-disulfide (II-), 1-disulfide (I-), and 0-disulfide (R) species, respectively. The collected intermediates were dried in a speedvac and stored in a -80°C freezer.

Mass Spectrometry

MALDI mass spectra were obtained on a Voyager Elite time-of-flight (TOF) mass spectrometer (PerSeptive Biosystems Inc., Framingham, MA) equipped with delayed extraction and a model VSL-337ND nitrogen laser (Laser Science, Newton, MA). The accelerating voltage in the ion source was set to 20 kV. Grid voltage and guide wire voltages were 93.6% and 0.2% of the accelerating voltage, respectively. Data were acquired in the positive linear DE mode of operation. Time-to-mass conversion was achieved by external or internal calibration using standards of bradykinin (m/z 1061.2), bovine pancreatic insulin (m/z 5734.5), and horse skeletal myoglobin (m/z 16,952) obtained from Sigma Chemical Company (St. Louis, MO). All experiments were performed using alpha-cyano-4-hydroxycinnamic acid (Aldrich Chemical Co., Milwaukee, WI) as the matrix. Saturated matrix solutions were prepared in a 50% (v/v) solution of acetonitrile and aqueous 0.1% TFA, mixed in equal volumes with peptide or protein samples, and applied to a stainless-steel sample plate. The mixture was allowed to air dry before being introduced into the mass spectrometer. Any problem with guanidine HCl during analysis by MALDI was avoided by dilution or washing of the sample spot on the MALDI plate.7 Experimentally observed m/z values were within 0.1% of the calculated values listed in Tables 1 and 2; rather than display observed m/z values in Figure 4, we indicated the residues in the fragments represented by the peaks to facilitate interpretation by the reader.

 

RESULTS

A profile of the abundances of the various intermediates is indicated by the HPLC chromatogram in Figure 3, which results from analysis of an aliquot of the folding solution as trapped (and therefore cyanylated) after 30 minutes under refolding conditions.

FIGURE 3. Profile of intermediates in the form of an HPLC chromatogram of cyanylated intermediates trapped after 30 minutes of refolding.

 

The molecular mass of each of the fractionated intermediates was obtained by MALDI-TOF MS. Peak 1 represents 3-disulfide species; they each have the same molecular mass as the native protein because there were no free sulfhydryl groups to be cyanylated and cause a mass shift. Peaks 2, 5, 6, 7, 10, and 12 represent 2-disulfide bond species, because the molecular mass of each was determined to be 6268 daltons, a 52-dalton shift from that of the native protein because there were two free sulfhydryls available to react with CDAP. Peaks 13, 14, 16, and 17 represent 1-disulfide bond species, because the molecular mass of each was determined to be 6320 daltons, a 104-dalton shift from that of the native form (a 26-dalton shift for each derivatizable sulfhydryl group). Peak 18 represents the reduced or unfolded hEGF. Peaks 8, 9, 11, and 15 represent mixtures of 2- and 1-disulfide species. Most 1-disulfide species eluted after the 2-disulfide species, but there are some cases of overlap or reversed order.

Aliquots of the compound represented by peak 2 in Figure 3, now known to contain two disulfide bonds, was subjected to cleavage and mass mapping before and after partial reduction and additional cyanylation. Exposing an aliquot of the 2-disulfide bond intermediate to cleavage directly allows the recognition of which two cysteines were free at the moment of trapping. Exposing a second aliquot of the 2-disulfide bond intermediate to partial reduction, further cyanylation, and then cleavage allows recognition of the connectivity of the disulfide bonds present in the intermediate.

The MALDI mass spectrum of the cleavage products of the 2-disulfide bonded intermediate (Fig. 3, peak 2) is shown in Figure 4A (see Table 1 for calculated m/z values of possible cleavage fragments); labeled peaks correspond to fragments consisting of residues itz-6-19, itz-20-53, and a beta-elimination product consisting of residues beta(itz-6-53), indicating that cleavage occurred at cysteines 6 and 20. Cysteines 6 and 20 are free residues in II-A. The peak labeled as 1-53 represents some cyanylated II-A that did not undergo any cleavage under our experimental conditions. The peaks representing itz-6-53 and 1-53 appear as doublets, because each of the fragments consist of two species; in this case, the more abundant corresponds to a cyanylated but uncleaved species, and the other (59 daltons lower) corresponds to a beta-elimination product.

FIGURE 4. (A) The MALDI mass spectrum of peptide mixtures resulting from cleavage of cyanylated II-A. (B and C) The mass spectra of cleavage products of cyanylated singly reduced isoforms of II-A, as represented by peaks 1 and 2 in Figure 5. Itz, iminothiazolidine derivative.

 

TABLE 1
Expected m/z Values for Fragments Resulting From Cleavage of Each of the Possible 15 Singly Reduced or Cyanylated hEGF Isomers
Reduction of
disulfide pair		     Fragment     m/z     Reduction of
disulfide pair		     Fragment     m/z
6-14 1-5 551.5 14-42 1-13 1423.5
itz-6-13 915.9 itz-14-41 3160.6
itz-14-53 4837.6 itz-42-53 1721.0
6-20 1-5 551.1 20-31 1-19 2047.1
itz-6-19 1539.6 itz-20-30 1344.6
itz-20-53 4214.0 itz-31-53 2913.3
6-31 1-5 551.5 20-33 1-19 2047.1
itz-6-30 2840.2 itz-20-32 1560.8
itz-31-53 2913.3 itz-33-53 2697.2
6-33 1-5 551.5 20-42 1-19 2047.1
itz-6-32 3056.4 itz-20-41 2536.9
itz-33-53 2697.2 itz-42-53 1721.0
6-42 1-5 551.5 31-33 1-30 3347.7
itz-6-41 4031.5 itz-31-32 261.3
itz-42-53 1721.0 itz-33-53 2697.2
14-20 1-13 1423.4 31-42 1-30 3347.7
itz-14-19 667.7 itz-31-41 1236.3
itz-20-53 2913.3 itz-42-53 1721.0
14-31 1-13 1423.4 33-42 1-32 3563.9
itz-14-30 1968.3 itz-33-41 1020.1
itz-31-53 2913.3 itz-42-53 1721.0
14-33 1-13 1423.4
  itz-14-32 2184.5
itz-33-53 2697.2

itz, iminothiazolidine derivative.

 

A different aliquot of the fraction labeled as peak 2 in Figure 3 was subjected to partial reduction and further cyanylation of the nascent sulfhydryls; an aliquot of this reaction mixture was subjected to HPLC to give the separation shown in Figure 5. Peak II-A in Figure 5 represents the original 2-disulfide intermediate that survived the partial reduction reaction, whereas peaks 1 and 2 represent the two isomeric 1-disulfide species, and peak R represents the species consisting of no disulfide bonds (it was completely reduced during the partial reduction procedure). The material represented by peak 1 in Figure 5 was exposed to cleavage conditions at high pH, and the resulting reaction mixture was subjected to analysis by MALDI-MS to give the mass spectrum shown in Figure 4B. Table 2 lists the calculated m/z values for cleavage fragments of the 15 possible 1-disulfide-bonded species. Peaks in Figure 4B correspond to fragments consisting of residues itz-6-19, itz-33-41, itz-42-53 and a beta-elimination product consisting of residues (itz-33-53), indicating that a disulfide pair, Cys 33 and Cys 42, must have been reduced, cyanylated, and cleaved. Likewise, the MALDI mass spectrum in Figure 4C was obtained from the mixture of cleavage products from the species represented by HPLC peak 3 in Figure 5. The labeled peaks in Figure 4C correspond to fragments consisting of residues itz-6-13, itz-20-30, itz-14-30, and itz-31-53, indicating cleavage at cysteine residues 6, 20, 14, and 31, respectively.

FIGURE 5. HPLC chromatogram of the reaction mixture after partial reduction and recyanylation of an aliquot of the fraction corresponding to II-A (peak 2 in Figure 3). Peaks 1 and 2 represent the two isoforms of singly reduced II-A; and R is the fully reduced, cyanylated protein.

 

TABLE 2
Fragments of Cleavage Products Corresponding to Each of the 15 Possible Cyanylated 1-Disulfide Intermediates of hEGF
Disulfide
linkage		     Fragment     m/z     Disulfide
linkage		     Fragment     m/z
6-14 1-19 2047.1 14-42 1-5 551.5
itz-20-30 1344.6 itz-6-19 1539.6
itz-31-32 262.3 itz-20-30 1344.6
itz-33-41 1020.1 itz-31-32 262.3
itz-42-53 1721.0 itz-33-53 2697.2
6-20 1-13 1423.4 20-31 1-5 551.5
itz-14-30 1968.3 itz-6-13 915.9
itz-31-32 262.3 itz-14-32 2184.5
itz-33-41 1020.1 itz-33-41 1020.1
itz-42-53 1721.0 itz-42-53 1721.0
6-31 1-13 1423.4 20-33 1-5 551.5
itz-14-19 667.7 itz-6-13 915.9
itz-20-32 1560.8 itz-14-30 1344.6
itz-33-41 1020.1 itz-31-41 1236.3
itz-42-53 1721.0 itz-42-53 1721.0
6-33 1-13 1423.4 20-42 1-5 551.5
itz-14-19 667.7 itz-6-13 915.9
itz-20-30 1344.6 itz-14-30 1968.3
itz-31-41 1236.3 itz-31-32 262.3
itz-42-53 1721.0 itz-33-53 2697.3
6-42 1-13 1423.4 31-33 1-5 551.5
itz-14-19 667.7 itz-6-13 915.9
itz-20-30 1344.6 itz-14-19 667.7
itz-31-32 262.3 itz-20-41 2536.9
itz-33-53 2697.2 itz-42-53 1721.0
14-20 1-5 551.5 31-42 1-5 551.5
itz-6-30 2840.2 itz-6-13 915.9
itz-31-32 262.3 itz-14-19 667.7
itz-33-41 1020.1 itz-20-32 1560.8
itz-42-53 1721.0 itz-33-53 2697.2
14-31 1-5 551.5 33-42 1-5 551.5
itz-6-19 1539.6 itz-6-13 915.9
itz-20-32 1560.8 itz-14-19 667.7
itz-33-41 1020.1 itz-20-30 1344.6
itz-42-53 1721.0 itz-31-53 2913.3
14-33 1-5 551.5
itz-6-19 1539.6
itz-20-30 1344.6
itz-31-41 1236.3
tz-42-53i 1721.0

itz, iminothiazolidine derivative.

 

Because we know from the data in Figure 4A that cysteines 6 and 20 were free in II-A, we deduce that Cys 14 is linked to Cys 31. There is a minor degree of unexpected cleavage at Cys 42, resulting in the formation of fragments consisting of residues itz-31-41 and itz-42-53. As we described in a more complete report of the folding intermediates of hEGF,11 the species itz-42-53 is exceptionally sensitive to MALDI. A possible explanation for the appearance of a peak for itz-42-53 in the MALDI spectra in Figures 4A and 4C is a minor cross reaction between cysteine residues in the high pH conditions during cleavage giving rise to a trace of itz-42-53. Such interference is plausible because we have found that the sensitivity of itz-42-53 to MALDI is more than 100 times greater than other peptides in this study. Fortunately, we have rarely observed this phenomenon, and although the presence of itz-42-53 in Figure 4 makes this a less than perfect didactic example, disulfide bond assignment is still possible.11

 

DISCUSSION

Conventional methodology to quench the folding process uses the alkylation of sulfhydryl groups by iodoacetate (iodoacetic acid or iodoacetamide) under alkaline conditions, although the fidelity of iodoacetate trapping has been questioned. Torella et al.12 found that alkylation by iodoacetic acid results in a side reaction at histidine residues. Acid quenching is another common method used to stop folding. However, the kinetics and the distribution patterns of intermediates trapped by iodoacetate and by acidic solution are frequently different, implying possible artifact formation during either of the two trapping methods. The kinetic analysis of acid quenching by Weissman and Kim3 indicated that acid quenching can effectively slow down folding and further diminish thiol/disulfide exchange by reducing the concentration of the thiolate anion. On the other hand, the results from iodoacetate trapping should be interpreted with caution because the thiol/disulfide exchange cannot be eliminated completely under the reaction conditions.

Cyanylation of sulfhydryl groups has rarely been used as an analytic reaction, even though the thiolate anion is one of the most reactive functional groups in proteins. Typically, cyanylation has been carried out by reaction with 2-nitro-5-thiocyanobenzoic acid (NTCB), which modifies SH groups under alkaline conditions; however, the kinetics of cyanylation by NTCB is slower than alkylation by iodoacetate. An alternative reagent, CDAP, proved to be suitable for the cyanylation of SH groups under acidic conditions. Our experiments with the CDAP reagent indicated that cyanylation is complete at pH 3 within 15 minutes at room temperature, under which conditions thiol/disulfide exchange is minimized. CDAP can effectively trap intermediates in low pH solution to minimize the exchange reaction, and the cyanylated products can be subjected directly to disulfide mapping after isolation by our simple methodology.7

The pH change from folding conditions (pH 8.5) to cyanylation conditions (pH 3.0) is accomplished by acidifying the solution with a small volume of HCl solution containing CDAP, as described in the experimental section. Control of the pH is important for successful cyanylation by CDAP, because the lower pH (<2) slows the reaction significantly, whereas the higher pH (>4.5) catalyzes hydrolysis of the CDAP reagent. The pH of the reaction can be monitored with a pH meter.

In summary, trapping of free cysteinyl intermediates (with a half-life of seconds or longer) can be achieved efficiently under conditions that minimize or eliminate disulfide exchange. The trapped intermediates are in a form (cyanylated) that is already part of a newly developed procedure for disulfide mass mapping based on specific chemical cleavage.7,13

 

REFERENCES

1. Rothwarf DM, Scheraga HA. Regeneration of bovine pancreatic ribonuclease A. 1. Steady-state distribution. Biochemistry 1993;32:2671-2679.

2. Creighton TE. The disulfide folding pathway of BPTI. Science 1992;256:111-114.

3. Weissman JS, Kim PS. Reexamination of the folding of BPTI: Predominance of native intermediates. Science 1991;253:1386-1393.

4. Weissman JS, Kim PS. The pro region of BPTI facilitates folding. Cell 1992;71:841-851.

5. Hirayama K, Akashi S. Assignment of disulfide bonds in proteins. In Matsuo T, Caprioli RM, Ross ML, Seyama Y (eds): Biological Mass Spectrometry: Present and Future. New York: Wiley, 1994:299-312.

6. Smith DL, Zhou Z. Strategies for locating disulfide bonds in proteins. Methods Enzymol 1990;193:374-389.

7. Wu J, Watson JT. A novel methodology for assignment of disulfide bond pairings in proteins. Protein Sci 1997; 6:391-398.

8. Wu J, Gage DA, Watson JT. A strategy to locate cysteine residues in proteins by specific chemical cleavage followed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Anal Biochem 1996; 235:161-174.

9. Wu J, Yang Y, Watson JT. Disulfide mapping of folding intermediates of recombinant human epidermal growth factor (hEGF) after trapping by cyanylation, cleavage, and mass spectrometry. Presented at the 43rd annual meeting of the American Society for Mass Spectrometry (p. TPG221 of the proceedings); Palm Springs, CA, June 1997.

10. Chang J-Y, Schindler P, Ramseier U, Lai P-H. The disulfide folding pathway of human epidermal growth factor. J Biol Chem 1995;270:9207-9216.

11. Wu J, Yang Y, Watson JT. Trapping of intermediates during the refolding of recombinant human epidermal growth factor (hEGF) by cyanylation, and subsequent structural elucidation by mass spectrometry. Protein Sci 1998;7:1017-1028.

12. Torella C, Ruoppolo M, Marino G, Pucci P. Analysis of RNase A refolding intermediates by electrospray/mass spectrometry. FEBS Lett 1994;352:301-306.

13. Schutte CG, Lemm T, Glombitza GJ, Sandhoff K. Complete localization of disulfide bonds in GM2 activator protein. Protein Sci 1998;7:1039-1045.

 

Return to table of contents