Does Electron Capture Dissociation (ECD) Provide Quantitative

Jul 9, 2014 - Electron capture dissociation (ECD) as a method of quantitative and qualitative study of glycated ubiquitin was investigated. ECD has be...
8 downloads 0 Views 1MB Size
Technical Note pubs.acs.org/ac

Does Electron Capture Dissociation (ECD) Provide Quantitative Information on the Chemical Modification of Lysine Side Chains in Proteins? The Glycation of Ubiquitin Piotr Stefanowicz,* Monika Kijewska, and Zbigniew Szewczuk Faculty of Chemistry, University of Wrocław, F. Joliot-Curie 14, 50-383 Wrocław, Poland S Supporting Information *

ABSTRACT: Electron capture dissociation (ECD) as a method of quantitative and qualitative study of glycated ubiquitin was investigated. ECD has been successfully applied for sequencing of modified peptides and assigning glycated Lys residues. By using a hybrid Fourier transform mass spectrometry (FT-MS) system equipped for ECD, a series of multiply glycated ubiquitin ions was observed. Ions of the glycated ubiquitin with a defined number of glucose moieties attached to the protein were isolated by quadrupol and fragmented in the ICR cell by the ECD method. The fragmentation spectrum was dominated by cn and (z+1)n ions. The ECD technique was tested for the quantitative analysis of the modified ubiquitin and isomeric glycated peptides (fragments of bovine serum albumin (BSA)). Obtained results indicate that the ECD fragmentation cannot be applied for the quantitative determination of the relative reactivities of respective Lys residues in the ubiquitin.

T

of more-complex proteins often required additional confirmation via the MS/MS technique. There are examples of application of the CID MS/MS method for sequencing of glycated peptides;9 however, a disadvantage of this approach is an abundant fragmentation of the hexose moiety attached to a peptide, resulting in multiple neutral losses of water molecules, formaldehyde, and a whole hexose, determining a poor sequence coverage.10 These difficulties may be overcome by applying ECD11 or electron transfer dissociation (ETD)12 methods, which have proven to be efficient in the analysis of peptide-derived Amadori products.11,12 In both approaches, the intense series of (z+1) and c ions covering the majority of sequence were observed. Recently, we analyzed the ECD fragmentation of synthetic peptide-derived Amadori products, as well as peptide fragments of glycated ubiquitin.10 Glycation of the protein results in a mixture of isomeric compounds that differ by number and location of glycated lysine moieties. Successful analysis of these nonhomogeneously glycated peptides encouraged us to apply the ECD method for the characterization of the intact molecule of glycated ubiquitin.

he high resolution of a Fourier transform ion cyclotron resonance (FT-ICR) instrument enables direct analysis of post-translational modifications of large molecules including intact proteins. Nowadays, there is a growing interest in application of the electron capture dissociation (ECD) fragmentation technique for “top down” identification of post-translational modifications sites, such as phosphorylation1 or enzymatic glycosylation.2 The idea of fragmenting a whole, nonspecifically modified protein, to quantitatively analyze the modifications occupancies at the specific sites was published by Pesavento et al.3 The authors studied the mixture of post-translationally acetylated histones and their fragments. They proved that the ratios of intensities of ions generated in ECD experiment are highly correlated with the ratios of isomeric peptides producing corresponding ions. This procedure was also successfully extended to analyze the naturally occurring mixture of histone H4 acetylation isomers,3 as well as the autoacetylated p300 HAT protein.4 However, it is not clear if this approach is general and can be applied to other modifications of proteins. Nonenzymatic glycation of proteins is considered to be an important cause of complications associated with diabetes.5−7 The application of mass-spectrometric methods to the glycated proteins analysis was reviewed recently.8 The early procedures of characterization of glycated proteins based on immunological, chromatographic, and chemical methods were not able to provide detailed information on glycation sites. Peptide mapping methods combining the enzymatic hydrolysis and mass-spectrometric analysis (LC-ESI-MS or MALDI-TOF-MS) in many cases identified modified amino acid residues. Analysis © 2014 American Chemical Society



EXPERIMENTAL SECTION

Chemicals and Reagents. Ubiquitin (from bovine blood cells) and [12C6]D-glucose were purchased from Sigma. Formic

Received: January 30, 2014 Accepted: July 9, 2014 Published: July 9, 2014 7247

dx.doi.org/10.1021/ac501329g | Anal. Chem. 2014, 86, 7247−7251

Analytical Chemistry

Technical Note

Figure 1. ESI-MS spectrum of modified ubiquitin (panels located above the signals show the theoretical ECD tandem mass spectra of unmodified protein and ubiquitin containing 1, 2, or 3 hexose moieties present in a series of expected cn and zn ions).

acid (conc. 99%) was purchased from Merck. Acetonitrile of the LC-MS grade was purchased from Fluka. Glycation of Ubiquitin. The high-temperature glycation of ubiquitin (bovine blood cells; Sigma−Aldrich, St. Louis, MO) was performed according to Boratyński’s procedure.13,14 Mass spectrometry. The ESI-MS experiments were performed using an Apex-Qe 7T instrument (Bruker) equipped with an ESI source and a heated hollow cathode dispenser. Spectra were recorded using aqueous solutions of acetonitrile (50%) and formic acid (0.1%) at the protein concentration 5 μM. The potential between the spray needle and the orifice was set to 4.5 kV. The cathode dispenser was heated gradually to 1.7 A. The ECD pulse length was set at 7 ms, and the ECD bias was 0.7 V. Analysis and deconvolution of the obtained spectra were carried out with Biotools (Bruker) software. A sophisticated numerical annotation procedure (SNAP) algorithm was used to find peaks.15 All the obtained signals had a mass accuracy error in a range of 5 ppm.

where n is the number of hexose moieties attached to the ion and An is the abundance of the ion corresponding to ubiquitin with n attached hexose moieties. The sequence of ubiquitin contains seven lysine residues and a N-terminal amino group. The glycation is relatively nonspecific.10 The ECD technique was used to indicate the modification of lysine residues in thermally glycated ubiquitin. Even though there are differences in the reactivity of the individual lysine moieties, each of the amino groups may be glycated to some extent. Therefore, the sample of glycated protein is a very complex mixture. It may be expected that the conjugate containing only one hexose moiety in the protein molecule (M = 8721.9) is a mixture of eight possible isomers. The number of possible isomers for doubly glycated ubiquitin is given as

RESULTS AND DISCUSSION Ubiquitin used in these experiments was glycated according to the procedure described previously.13,14 The deconvoluted spectrum of glycated protein presented in Figure 1 shows that the number of the hexose (fructosamine) moieties attached to the ubiquitin molecule ranges from 0 to 4. The average ratio of hexose moieties per 1 molecule of ubiquitin (n̅) is 1.82, calculated according to the formula

while for ubiquitin containing three hexose moieties, respectively, the number of possible isomers is given as

⎛8⎞ ⎜ ⎟ = 56 ⎝2⎠



n̅ =

⎛8⎞ ⎜ ⎟ = 336 ⎝ 3⎠ To analyze the sample of ubiquitin glycation products, we performed fragmentation on the mixtures of isomers characterized by a defined number of hexose moieties. For the ECD experiment, the ions with charge +11 were selected. The fragmentation was performed for the conjugates containing 0, 1, 2, or 3 fructosamine (Fru) moieties at m/z

∑n nA n ∑n A n

(1) 7248

dx.doi.org/10.1021/ac501329g | Anal. Chem. 2014, 86, 7247−7251

Analytical Chemistry

Technical Note

Figure 2. Stepped curve defining the degree of reactivity of individual lysine residues in glycated ubiquitin obtained from ECD fragmentation ions m/z 808.63 ([M+2 Fru+11H]11+).

779.17 ([M+11H]11+, m/z 793.90 ([M+Fru+11H]11+), m/z 808.63 ([M+2Fru+11H]11+) and m/z 823.36 ([M+3Fru +11H]11+), respectively (see Figure 1). Resulting spectra were analyzed using a SNAP algorithm to obtain isotopic envelopes for particular product ions. The processed spectrum of doubly glycated ubiquitin presented in Figure S1 in the Supporting Information reveals that mainly cn and (z+1)n ions are formed. The sequence of ubiquitin was covered on the basis of ions obtained by fragmentation of parent ions m/z 808.63 ([M+2Fru+11H]11+. The number of identified signals for the individual undergoing ECD fragmentation indicates all modifications in glycated ubiquitin (see Figure S2 in the Supporting Information). Identified ions, classified according to the ion type (cn or (z+1)n), n value, and the number of hexose moieties, are presented in Table S1 in the Supporting Information. The highest number of fragments was observed for ubiquitin glycated with three hexose moieties. Comparison of the ions differing by the number of hexose moieties only allows one to calculate the average level of glycation (n̅) for all the obtained fragments (characterized by a defined number of hexose moieties and corresponding to a specific ion). According to the previously described formula (eq 1), n̅ = 1.82. Plotting the glycation level as a function of the fragment length provides information on the distribution of the sugar moieties in the protein sequence.16,17 Data presented in Figure 2 represent separate plots for the protein containing two hexose moieties. A graph for the protein containing three hexose moieties is presented in Figure S3 in the Supporting Information. The graph presented in Figure 2 consists of several plateaus, corresponding to fragments resulting from the cleavage of peptide bonds located between the Lys moieties. The average glycation level of the corresponding lysine residues was different. Their values depended on the reactivity of each

lysine side chain in the glycation reaction. According to the data presented in Figure 2, the average glycation level for the sequence fragments 0−5 is close to 0, suggesting that the Nterminal amino group was not glycated. The ions between Lys11 and Lys27 formed a plateau with the average glycation level being 0.5. Between Lys27 and Lys33, the glycation level increased to a value of ∼1.5. The increase of glycation level for Lys48 is low, but there is only one point in the plot in this range. For Lys63, the level of glycation reaches the highest value (2). A similar graph was obtained for the mixture of isomers containing three hexose residues in the protein molecule. The analysis of ECD spectra suggests that the lessreactive moiety was the N-terminal amino group, which, according to our data, was not glycated at all. The ε-amino groups of the lysine residues showed a comparable reactivity, with the exception of Lys63, which turned out to be more prone to react with sugar than the remaining lysines. The results obtained for the ions corresponding to modified proteins containing 2 and 3 hexose moieties were qualitatively similar but not identical. To find the origin of this discrepancy, we performed additional experiments on the ECD fragmentation of mixtures of glycated peptides obtained by regioselective synthesis. The glycated isomers (H-Asp-Thr-Glu-Lys(Fru)Gln-Ile-Lys-Lys-Gln-Thr-OH (P4), H-Asp-Thr-Glu-Lys-GlnIle-Lys-Lys(Fru)-Gln-Thr-OH (P6), H-Asp-Thr-Ile-Ser-SerLys(Fru)-Leu-Lys-Glu-OH (P9), and H-Asp-Thr-Ile-Ser-SerLys-Leu-Lys(Fru)-Glu-OH (P10)) were synthesized according to the Fmoc strategy using a fully protected building block: Fmoc-Lys(i,i-Fru,Boc)−OH.18 The composition of the mixture was checked by high-performance liquid chromatography (HPLC). The ratio of isomers in the mixture was correlated with the ratios of abundances of diagnostic fragments in the ECD spectra of investigated peptides. The conducted experiments showed that the ratio of the intensity of diagnostic ions in the ECD experiment (Y-axis) correlates linearly with the 7249

dx.doi.org/10.1021/ac501329g | Anal. Chem. 2014, 86, 7247−7251

Analytical Chemistry

Technical Note

Figure 3. Quantitative analysis of the diagnostic ion pairs cn and zn differentiation of modifications in the studied isomers P4 and P6 (identified diagnostic ions in the ECD spectra for isomer P4: c7 1022.536 and for isomer P6: c7* 860.483; z6* 891.490. (In this figure, I represents the intensity of the diagnostic ion, and P represents the area under the signal in the chromatogram.)

glycation is preserved in fragment ions, the results do not allow the unambiguous quantitative determination of the relative reactivities of respective Lys residues in the ubiquitin sequence. Therefore, our results are not fully consistent with the previous research concerning the ECD analysis of a mixture of acetylated histone H4.3 This may suggest that the quantitative correlation of abundances of fragment ions and a content of particular forms of modified proteins is not a general phenomenon for all post-translational modifications. The ECD fragmentation-based method seems to be useful for the quantitative analysis of isomeric mixtures of modified peptides; however, for modifications different from acetylation, a calibration with synthetic peptide mixtures of known composition is required.

composition of the tested mixture of isomers (X-axis). The dependence of abundance on concentration was linear, although the slope differed in particular cases (see Figure 3 and Figure S4 in the Supporting Information). When we considered the pairs of c7/c7,* c7/z6* and c6/c6,* c6/c7* ions, the dependence of P4/P6 and P9/P10 was linear with a slope in the range of 1.0−1.15 and a correlation coefficient in the range of 0.93−0.98; in the same measurements, the same dependence for ions c7/c7* and c7/c6* (P9/P10) was characterized by a slope in the range of 0.35−0.36 and a correlation coefficient in the range of 0.93−0.94. The obtained results suggest that the slope value is very much dependent on the pairs of ions selected for analysis. It indicates that the ECD method can be applied to the analysis of mixtures of isomers of post-translationally modified peptides to unambiguously indicate the modification sites but quantitative analysis using ECD in each individual case requires a calibration using an appropriate mixture of regioselectivity modified peptides.



ASSOCIATED CONTENT

S Supporting Information *

This material is available free of charge via the Internet at http://pubs.acs.org/.





CONCLUSIONS We have shown that electron capture dissociation (ECD) is a powerful technique for localizing sites of glycation of proteins. Interpretation of spectra revealed significant number of fragments, mainly c type and (z+1) type. On the basis of abundances of ions with the glycation level 0, 1, 2, and 3, the average glycation level was calculated for fragments of various lengths. Plotting this value against length produced stepped curves consisting of several plateaus (see Figure 2 and Figure S3 in the Supporting Information). However, the differences between plots obtained for the ions at m/z 808.63 ([M+2Fru +11H]11+) and m/z 823.36 ([M+3Fru+11H]11+) (two and three hexose moieties, respectively) indicated that these data are not suggestive of a quantitative measurement of the individual lysine glycation. In addition, if (z+1) ions were taken into consideration, the obtained correlation is even less profound and no regularity was observed in the data. This result was confirmed by experiments on model glycated peptides. The dependence of concentration ratios of isomeric glycated peptide on ratios of abundances of diagnostic ions is linear, but the slope depends on the ion pair that is selected. It seems that even though the ECD fragmentation of glycated ubiquitin provides good sequence coverage and

AUTHOR INFORMATION

Corresponding Author

*Tel.: +48-71-3757213. Fax: +48 71 3282348. E-mail: piotr. [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by a grant from the Ministry of Science and Higher Education of Poland (No. N N204 180340).



REFERENCES

(1) Sweet, S. M. M.; Mardakheh, F. K.; Ryan, K. J. P.; Langton, A. J.; Heath, J. K.; Cooper, H. J. Anal. Chem. 2008, 80, 6650−6657. (2) Reid, G. E.; McLuckey, S. A. J. Mass Spectrom. 2002, 37, 663− 675. (3) Pesavento, J. J.; Mizzen, C. A.; Kelleher, N. L. Anal. Chem. 2006, 78, 4271−4280. (4) Karanam, B.; Jiang, L.; Wang, L.; Kelleher, N. L.; Cole, P. A. J. Biol. Chem. 2006, 281, 40292−40301. (5) Cohen, M. P.; Ziyadeh, F. N.; Chen, S. J. Lab. Clin. Med. 2006, 147, 211−219. 7250

dx.doi.org/10.1021/ac501329g | Anal. Chem. 2014, 86, 7247−7251

Analytical Chemistry

Technical Note

(6) Capote, F. P.; Sanchez, J.-C. Mass Spectrom. Rev. 2009, 28, 135− 146. (7) Zhang, Q.; Tang, N.; Schepmoes, A. A.; Phillips, L. S.; Smith, R. D.; Metz, T. O. J. Proteome Res. 2008, 7, 2025−2032. (8) Arena, S.; Salzano, A. M.; Renzone, G.; D’Ambrosio, C.; Scaloni, A. Mass Spectrom Rev. 2014, 33, 49−77. (9) Frolov, A.; Hoffmann, P.; Hoffmann, R. J. Mass Spectrom. 2006, 41, 1459−1469. (10) Lapolla, A.; Fedele, D.; Reitano, R.; Arico, N. C.; Seraglia, R.; Traldi, P.; Marotta, E.; Tonani, R. J. Am. Soc. Mass. Spectrom. 2004, 15, 496−509. (11) Stefanowicz, P.; Kijewska, M.; Szewczuk, Z. J. Mass Spectrom. 2009, 44, 1047−1052. (12) Zhang, Q.; Forlov, A.; Tang, N.; Hoffmann, R.; van de Goor, T.; Metz, T. O.; Smith, R. D. Rapid Commun. Mass Spectrom. 2007, 21, 661−666. (13) Boratyński, J. Biotechnol. Techniq. 1998, 12, 707−710. (14) Stefanowicz, P.; Boratyński, J.; Kańska, U.; Petry, I.; Szewczuk, Z. Acta Biochim. Pol. 2001, 48, 1137−1141. (15) Available via the Internet at: http://bruker.poznan.pl/images/ stories/Daltonics/noty/mt75.pdf (accessed Dec. 12, 2013). (16) Stefanowicz, P.; Petry-Podgórska, I.; Kowalewska, K.; Jaremko, Ł.; Jaremko, M.; Szewczuk, Z. Biosci. Rep. 2010, 29, 91−99. (17) Pan, J.; Han, J.; Borchers, C. H.; Konermann, L. J. Am. Chem. Soc. 2009, 131, 12801−12808. (18) Stefanowicz, P.; Kijewska, M.; Kapczynska, K.; Szewczuk, Z. Amino Acids 2010, 38, 881−889.

7251

dx.doi.org/10.1021/ac501329g | Anal. Chem. 2014, 86, 7247−7251