Analytical and Bioanalytical Chemistry

, Volume 392, Issue 5, pp 831–838

Fragmentation of intra-peptide and inter-peptide disulfide bonds of proteolytic peptides by nanoESI collision-induced dissociation

Authors

  • Michael Mormann
    • Institute for Medical Physics und BiophysicsUniversity of Münster
  • Johannes Eble
    • Institute for Physiological ChemistryUniversity of Münster
  • Christian Schwöppe
    • Department of Medicine/Haematology and OncologyUniversity Hospital of Münster
  • Rolf M. Mesters
    • Department of Medicine/Haematology and OncologyUniversity Hospital of Münster
  • Wolfgang E. Berdel
    • Department of Medicine/Haematology and OncologyUniversity Hospital of Münster
  • Jasna Peter-Katalinić
    • Institute for Medical Physics und BiophysicsUniversity of Münster
    • Institute for Medical Physics und BiophysicsUniversity of Münster
Original Paper

DOI: 10.1007/s00216-008-2258-7

Cite this article as:
Mormann, M., Eble, J., Schwöppe, C. et al. Anal Bioanal Chem (2008) 392: 831. doi:10.1007/s00216-008-2258-7

Abstract

Characterisation and identification of disulfide bridges is an important aspect of structural elucidation of proteins. Covalent cysteine-cysteine contacts within the protein give rise to stabilisation of the native tertiary structure of the molecules. Bottom-up identification and sequencing of proteins by mass spectrometry most frequently involves reductive cleavage and alkylation of disulfide links followed by enzymatic digestion. However, when using this approach, information on cysteine-cysteine contacts within the protein is lost. Mass spectrometric characterisation of peptides containing intra-chain disulfides is a challenging analytical task, because peptide bonds within the disulfide loop are believed to be resistant to fragmentation. In this contribution we show recent results on the fragmentation of intra and inter-peptide disulfide bonds of proteolytic peptides by nano electrospray ionisation collision-induced dissociation (nanoESI CID). Disulfide bridge-containing peptides obtained from proteolytic digests were submitted to low-energy nanoESI CID using a quadrupole time-of-flight (Q-TOF) instrument as a mass analyser. Fragmentation of the gaseous peptide ions gave rise to a set of b and y-type fragment ions which enabled derivation of the sequence of the amino acids located outside the disulfide loop. Surprisingly, careful examination of the fragment-ion spectra of peptide ions comprising an intramolecular disulfide bridge revealed the presence of low-abundance fragment ions formed by the cleavage of peptide bonds within the disulfide loop. These fragmentations are preceded by proton-induced asymmetric cleavage of the disulfide bridge giving rise to a modified cysteine containing a disulfohydryl substituent and a dehydroalanine residue on the C-S cleavage site.

Keywords

Inter-and intramolecular disulfide bridgeUnderivatised peptidesLow-energy collision-induced dissociationAsymmetric disulfide-bridge cleavage

Introduction

Oxidative coupling of the reactive SH groups of cysteines, yielding disulfide bonds, is one of the most important processes in the formation of the native structure of proteins. Often, disulfide bridges enhance the stabilisation of the tertiary structure of folded proteins [14].

Classical approaches for disulfide bond determination involve chemical or enzymatic cleavage of the intact protein followed by separation of the resulting peptides. Disulfide-linked peptides are then identified by rupture of the S-S bond, thereby modifying the separation properties. The latter are monitored by either electrophoretic or chromatographic techniques [57].

More recently, mass spectrometric techniques have been used for analysis of disulfide-linked proteins and peptides [8, 9]. Identification and sequencing of proteins is most frequently performed by use of proteolytic digestion followed by mass spectrometric analyses. Standard procedures for enzymatic degradation usually involve reductive cleavage of disulfide bonds present in the protein, to disrupt its tertiary structure. Reductive cleavage of the disulfide bonds is then followed by alkylation of the reactive free sulfhydryl groups preventing spontaneous oxidation by other sulfhydryl groups. However, when this approach is used, information on covalent cysteine-cysteine contacts within the protein is lost.

Various approaches have been shown to result in valuable information on analyte species in which two peptides are linked by a disulfide bridge. Cleavage of the disulfide bonds has been observed in both long-lived metastable ions and in ions fragmenting on a much shorter time-scale in the ion source when gas-phase ions are generated by matrix-assisted laser desorption/ionisation [1017]. Along with either symmetric or unsymmetric fragmentation of the S-S bridge, cleavage of amide bonds is often observed, giving rise to (partial) sequence information on the peptide ions. Electron-capture dissociation (ECD) is an alternative approach to the analysis of disulfide or sulfide-containing species, because it has been shown that C-S and S-S bonds are cleaved extremely rapidly and easily in hypervalent radical cations formed upon capture of an electron by a multiply charged precursor ion species [1823].

Defining the linkage pattern and obtaining sequence information from peptide ions containing an intramolecular disulfide loop by low energy collision-induced dissociation (CID) still remains a somewhat elusive goal. Peptide bonds within the disulfide loop are reported to be usually resistant to fragmentation under low-energy CID conditions which are typically applied to de-novo sequencing of peptides [8]. This finding is corroborated by recent results obtained by Lioe and O’Hair [24]. They examined the fragmentation of protonated model compounds containing intermolecular disulfide links both experimentally and theoretically. The results indicate that most cleavages of the disulfide bridge proceed usually via dissociation pathways that cannot compete with amide bond cleavages energetically. However, especially when peptide ions containing non-mobile protons are considered, disulfide bond cleavage reactions might be observed [24].

In some cases this problem can be overcome by use of high-energy collisional activation leading to both cleavage of the disulfide bond and fragmentation of the amide bonds of the peptide backbone [25]. However, the high-energy regime is not accessible with most instruments commonly used in proteomics laboratories. Complexation with gold(I) ions and other transition metals is an alternative approach to disulfide link cleavage, owing to the high sulfur affinity of heavier transition metals [2628]. Kim and Beauchamp have reported a novel method for identifying disulfide linkages in peptides by use of multiple-stage CID of analyte species cationised with alkali and alkaline earth metal ions [29]. Recently, Thakur and Balaram have reported on the fragmentation of positively charged gas-phase ions generated from contryphans, peptides containing a single disulfide bridge [30]. It has been reported that these analyte species mainly fragment by initial cleavage of an amide bond within the disulfide loop, giving rise to a linearised sequence. This process is promoted by the presence of at least one proline residue within the S-S loop, because amide bond cleavage adjacent to proline is a fast and facile process [31]. Ring-opening is typically succeeded by several dissociation processes giving rise to valuable information about the species under investigation. However, if no proline residues are present within the peptide sequence fragmentation by linearisation is rarely observed.

In this contribution we present recent results on the fragmentation of intra and inter-peptide disulfide bonds of proteolytic peptides by nanoESI collision-induced dissociation (nanoESI CID). Careful examination of the fragment ion spectra revealed the presence of low-abundance fragment ions formed by the cleavage of peptide bonds within the disulfide loop. These fragmentations are preceded by proton-induced asymmetric cleavage of the disulfide bridge, giving rise to a modified cysteine-containing a disulfohydryl substituent and a dehydroalanine residue on the C-S cleavage site.

Experimental

Reagents and materials

Rhodocetin alpha subunit from the venom of the Malayan pit viper (Calloselasma rhodostoma) was obtained by isolation and purification as described before [32]. A modified fusion protein comprising the extracellular domain of the tissue factor fused to the peptide GNGRAHA (tTF) was expressed in E. coli, isolated and purified according to the procedure reported by Kessler et al. [33]. Trypsin was purchased from Roche Diagnostics (Mannheim, Germany). All solvents were of HPLC-grade purity.

Omega glass capillaries used in nanoESI experiments were purchased from Hilgenberg (Malsfeld, Germany), and pulled by use of an in-house-built vertical pipette puller.

Sample preparation

Proteins were dissolved in 10 mmol L−1 ammonium hydrogen carbonate to final concentrations of 10 pmol μL−1 and first partially denatured by thermal defolding (95 °C, 5 min). Treatment of 10 μL protein solution with trypsin (20 μg mL−1) in solution (NH4HCO3, 10 mmol L−1, pH 7) for 12 h at 37 °C furnished the peptide mixtures analysed in this study. The mixtures were dried in a Speedvac (Savant, Farmingdale, NY, USA), and the dried residue was dissolved in 10 μL water to yield a concentration of 10 pmol μL−1 for the stock solution.

Mass spectrometric measurements

Nanoelectrospray Fourier transform ion cyclotron resonance mass spectrometry

Measurements were performed using a Bruker Apex II Fourier-transform ion-cyclotron resonance mass spectrometer (FT-ICR MS) equipped with a 9.4 T actively shielded magnet. Gas-phase ions were generated from solutions containing approx. 7.5 pmol μL−1 of the analyte material in water-methanol-formic acid 68:22:10 (v/v/v) by nanoESI in the positive-ion mode using an Apollo ion source. Typical source conditions were: capillary voltage −650 V and a capillary exit voltage of 60 V. The electrospray-generated ions were accumulated for 0.5 s in the hexapole located after the 2nd skimmer of the ion source and then transferred into the ICR cell. The ions were trapped inside the Infinity ICR cell by application of a “sidekick”, the trapping voltages were set to +0.9 V at both trapping electrodes. All mass spectra were acquired in the broadband mode with 512 kword data points. The time-domain signals were zerofilled once and apodized by a quadratic sine bell function prior to Fourier transformation. For all spectra 64 scans were accumulated.

Spectra were calibrated internally by use of the characteristic peptide ions generated by auto-digestion of trypsin.

Nanoelectrospray quadrupole time-of-flight tandem mass spectrometry (nanoESI-Q-TOF MS-MS)

NanoESI mass spectrometry experiments were carried out by use of a quadrupole time-of-flight (Q-TOF) mass spectrometer (Micromass, Manchester, UK) in the positive-ion mode. Gas-phase ions were generated from solutions containing approx. 7.5 pmol μL−1 of the analyte material in water-methanol-formic acid 68:22:10 (v/v/v) by nanoESI in the positive-ion mode using a Z-spray source. Typical source conditions were: source temperature 80 °C, desolvation gas (N2) flow rate 75 L h−1, capillary voltage potential 1.1 kV, and cone voltage 40 V. For low-energy CID experiments, the peptide precursor ions containing a disulfide loop or an inter-peptide disulfide bridge (PA–PD, cf. Table 1) were selected in the first quadrupole analyzer and fragmented in the collision cell using a collision gas (Ar) pressure of 3.0 × 10−3 Pa and collision energies of 20–40 eV (Elab).
Table 1

Peptide sequences, and theoretical and experimental (FT-ICR MS) m/z values for individual disulfide bridge-containing peptides examined in this study

https://static-content.springer.com/image/art%3A10.1007%2Fs00216-008-2258-7/MediaObjects/216_2008_2258_Tab1_HTML.gif

Results and discussion

Peptide mixtures obtained from proteolytic digests of partially denatured proteins, achieved by thermal defolding, where first investigated by high-resolution Fourier-transform ion-cyclotron resonance mass spectrometry (FT-ICR MS). Owing to the high mass accuracy of the instrument used oxidised species, i.e., those with a disulfide bridge, were readily identified by their exact mass (cf. Table 1).

The disulfide bridge-containing peptides were submitted to nanoESI low-energy CID using a quadrupole time-of-flight instrument (Q-TOF) as mass analyser. Figure 1 shows a nanoESI Q-TOF fragment-ion spectrum obtained from a CID experiment with the triply charged precursor ions [PA + 3H]3+ with m/z 911.05 derived from an in-solution tryptic digest of a protein isolated from the venom of the Malayan pit viper (Calloselasma rhodostoma) (manuscript in preparation).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-008-2258-7/MediaObjects/216_2008_2258_Fig1_HTML.gif
Fig. 1

(a) NanoESI Q-TOF fragment-ion spectrum obtained from a CID experiment on the triply-charged precursor peptide ions [PA + 3H]3+ of m/z 911.05 derived from an in-solution tryptic digest of a protein isolated from the venom of the Malayan pit viper (Calloselasma rhodostoma) (aa: 1–23). (y-type ions, black; b-type ions, red). (b) Fragmentation scheme of the peptide under inspection deduced from (a)

A complete set of y-type ions originating from amide-bond cleavages located outside the disulfide loop give rise to intense signals (y1–y8, y21, and y22) accompanied by some complementary b-type ions (b2, b3, b17–b19, and b22). Careful examination of the spectrum reveals the presence of further fragment ions also arising from peptide-bond cleavages. The corresponding fragment ions exhibit a mass shift of +32 u for the y-type ion series and a loss of −34 u for the N-terminal b-type ions as compared with the predicted m/z values for the fully reduced species. Fragment ions indicating the formation of a linearised sequence by initial cleavage of an amide bond within the disulfide loop, as reported previously [30], are not detected. This finding clearly points to an asymmetric ring opening of the intramolecular disulfide bridge prior to amide-bond fragmentation. Similar dissociation processes have already been reported for tandem MS spectra of intermolecularly disulfide-bonded peptides [8], bio-thiols [34] and β-lactam antibiotics [35, 36]. Asymmetric cleavage of the disulfide bridge gives rise to a modified cysteine containing a disulfohydryl substituent (thiocysteine) and a dehydroalanine (dhA) residue on the remote cleavage site. A possible mechanism for this type of fragmentation is illustrated in Scheme 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-008-2258-7/MediaObjects/216_2008_2258_Sch1_HTML.gif
Scheme 1

Fragmentation of peptide ions containing a disulfide loop—ring opening by β-elimination followed by proton transfer (salt-bridge mechanism, cf. text)

Charge-induced cleavage of the disulfide bridge, i.e., protonation of S-S as the initial fragmentation step, is improbable, because of the low proton affinity of the disulfide bridge [18, 37]. Therefore, a charge-remote intramolecular proton abstraction from the α-position of a cysteine residue by a nucleophilic and highly basic functional group present within the peptide leads to ring opening of the disulfide loop. Then, proton transfer to the disulfide moiety gives rise to a disulfohydryl substituent and a dehydroalanine residue at the C-S cleavage site of the former disulfide loop. The results obtained by Lioe and O’Hair indicate that the activation barrier for such a “salt bridge mechanism” is sufficiently low (145 kJ mol−1) to compete with the dissociation of the amide bonds outside and inside the loop (100–165 kJ mol−1). This process is followed by rupture of the amide bonds present in the peptide ions leading to b and y-type ions with the respective mass shift. For proper annotation of the resulting fragment ions those species containing the Dha amino acid residue have been denoted α-type ions and those containing the S-SH modification were denoted β-type ions.

To further investigate whether this type of cleavage can be regarded as a general fragmentation process of disulfide-bridge-containing peptides under low-energy CID conditions tryptic fragments obtained from a modified truncated tissue factor (tTF) have also been submitted to MS-MS experiments.

Figure 2 shows a nanoESI Q-TOF fragment-ion spectrum obtained from a CID experiment with the triply charged precursor ions with m/z 737.10 [PB + 3H]3+. Again, complete series of b and y-type fragment ions from both termini arising from cleavage of the amide bonds outside the disulfide ring can be observed. Closer inspection of the signals exhibiting minor intensity reveals the presence of fragment ions formed by asymmetric ring opening followed by amide bond cleavage. No evidence was found for the formation of linearised sequences by cleavage of peptide bonds inside the disulfide ring. Notably, only C-terminal fragment ions (y-type ions) contain the thiocysteine residue while b-type ions show incorporation of the dehydroalanine modification exclusively, which was also observed for the peptide ions [PA + 3H]3+ (vide supra).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-008-2258-7/MediaObjects/216_2008_2258_Fig2_HTML.gif
Fig. 2

(a) NanoESI Q-TOF fragment-ion spectrum obtained from a CID experiment on the triply-charged precursor peptide ions [PB + 3H]3+ of m/z 737.10 derived from an in-solution tryptic digest of a truncated tissue factor protein (aa: 93–111). (y-type ions, black; b-type ions, red). (b) Fragmentation scheme of the peptide under inspection deduced from (a)

Because peptide PB (tTF, aa 93–111) contains an additional cleavage site for trypsin, i.e., a lysine at position 94, peptide PC (tTF, aa 95–111), bearing a cysteine residue as N-terminal amino acid, is also formed by tryptic digestion. Therefore, peptides PC lack a basic residue at the N-terminal end of the disulfide loop enabling proton abstraction from the α-position of the N-terminal cysteine. Surprisingly, collisional activation of gas-phase ions generated from PC also leads to asymmetrical disulfide-bridge cleavage towards the C-terminal end of the peptide backbone (cf. Fig. 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-008-2258-7/MediaObjects/216_2008_2258_Fig3_HTML.gif
Fig. 3

(a) NanoESI Q-TOF fragment-ion spectrum obtained from a CID experiment on the doubly-charged precursor peptide ions [PC + 2H]2+ of m/z 997.49 derived from an in-solution tryptic digest of a truncated tissue factor protein (aa: 95–111). (y-type ions, black; b-type ions, red). (b) Fragmentation scheme of the peptide under inspection deduced from (a)

Figure 3 shows a nanoESI Q-TOF fragment ion spectrum obtained from a CID experiment with the doubly charged precursor ions with m/z 997.49 [PC + 2H]2+. A similar fragmentation pattern as observed for peptide ions [PA + 3H]3+ and [PB + 3H]3+ is detected. Cleavage of the intramolecular disulfide bridge giving rise to formation of a dhA residue at the C-terminal cysteine residue and a thiocysteine at the cysteine residue located closer to the N-terminus is followed by formation of b and y-type ions. Even if the resulting fragment ions give rise to rather weak signals these data can provide almost the complete peptide sequence assignment.

To further evaluate whether the dissociation pathway outlined above is also observed for peptides linked by an intermolecular S-S bridge, ions [PD + 3H]3+ were submitted to tandem mass spectrometry. The fragmentation of inter-peptide disulfide bonds is demonstrated by a CID experiment on a tryptic peptide, PD derived from an in-solution tryptic digest of the rhodocetin alpha subunit from the venom of the Malayan pit viper (Calloselasma rhodostoma) (cf. Fig. 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-008-2258-7/MediaObjects/216_2008_2258_Fig4_HTML.gif
Fig. 4

(a) NanoESI Q-TOF fragment-ion spectrum obtained from a CID experiment on the triply-charged precursor peptide ions [PD + 3H]3+ of m/z 719.26 derived from an in-solution tryptic digest of the rhodocetin alpha unit from the venom of the Malayan pit viper (Calloselasma rhodostoma) (aa: 1–18). (y-type ions, black; b-type ions, red). (b) Fragmentation scheme of the peptide under inspection deduced from (a)

The peptide PD represents the amino acids 1 to 18 and the peptide backbone is cleaved at K10. The peptides are referred to as P1 (aa 1–10) and P2 (aa 11–18) and b and y-type ions are prefixed with the corresponding numbers. Again, complete sets of y-type ions and a number of b-type ions can be detected and allow the determination of the peptide sequences and location of the disulfide bridge. Moreover, fairly abundant ions derived from α and β-fragmentations of the disulfide bond and from their combinations with b and y-type cleavages can be found. However, in contrast to intra-peptide disulfide bonds no preferred β-cleavage in combination with y-type ions is found. This observation is made also in the CID spectra of other peptides containing intermolecular disulfide bridges (e.g., chymotryptic peptide ions derived from the rhodocetin alpha subunit; data not shown). This finding might indicate that asymmetric cleavage of the intermolecular disulfide bridges does not necessarily precede amide bond fragmentation.

Conclusion

In this contribution we have presented a novel type of fragmentation process of peptides containing disulfide bridges. All gaseous ions obtained by positive-ion mode nanoESI of tryptic peptides with both intra and intermolecular disulfide links dissociate via similar dissociation pathways:
  1. 1.

    asymmetric cleavage of the disulfide bridge giving rise to a modified cysteine containing a disulfohydryl substituent and a dehydroalanine (dhA) residue on the C-S cleavage site; and

     
  2. 2.

    fragmentation of the amide bonds giving rise to modified b and y-type ions shifted by −34 u and +32 u, respectively.

     

This type of fragmentation can be used for characterisation of proteins containing disulfide bridges with respect to both de-novo sequencing and disulfide bridge location.

Acknowledgements

We gratefully acknowledge financial support by the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 492, project Z2 to JPK).

Copyright information

© Springer-Verlag 2008