Mass Defect-Based Pseudo-Isobaric Dimethyl Labeling for

Mass Spectrometry-Based Proteomics for Pre-Eclampsia and Preterm Birth. Kai Law , Ting-Li Han , Chao Tong , Philip Baker. International Journal of Mol...
0 downloads 0 Views 802KB Size
Download PDF
Letter pubs.acs.org/ac

Mass Defect-Based Pseudo-Isobaric Dimethyl Labeling for Proteome Quantification Yuan Zhou,†,‡ Yichu Shan,† Qi Wu,†,‡ Shen Zhang,†,‡ Lihua Zhang,*,† and Yukui Zhang† †

National Chromatographic Research and Analysis Center, Key Lab of Separation Sciences for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian, China ‡ University of Chinese Academy of Sciences, Beijing, China S Supporting Information *

ABSTRACT: Discovering differentially expressed proteins in various biological samples requires proteome quantification methods with accuracy, precision, and wide dynamic range. This study describes a mass defect-based pseudo-isobaric dimethyl labeling (pIDL) method based on the subtle mass defect differences between 12C/13C and 1H/2H. Lys-C protein digests were labeled with CD2O/13CD2O and reduced with NaCNBD3/NaCNBH3 as heavy and light isotopologues, respectively. The fragment ion pairs with mass differences of 5.84 mDa were resolved by high-resolution tandem mass spectrometry (MS/MS) and used for quantification. The pIDL method described here resulted in highly accurate and precise quantification results with approximately 100-fold dynamic range. Furthermore, the pIDL method was extended to 4-plex proteome quantification and applied to the quantitative analysis of proteomes from Hca-P and Hca-F, two mouse hepatocarcinoma ascites syngeneic cell lines with low and high lymph node metastasis rates.

M

interference is common for MS/MS level-based quantification methods.9,10 To solve these problems, several innovative methods have been recently developed. Isobaric peptide termini labeling (IPTL)11−13 is an elegant solution in which the amino groups of the N-termini and C-termini of Lys-C protein digests were crosswise labeled with heavy/light isotope reagents according to the slightly different chemical properties of α- and ε-NH2. Fragment ion pairs specific to the labeled peptides were used for peptide/protein quantification. Although IPTL improved the quantification accuracy compared to other reported MS/ MS level methods, the side reactions that are inherent in the multistep labeling can adversely affect quantification accuracy and dynamic range.12 NeuCode SILAC14 is another strategy based on subtle mass differences resulting from the mass defects of 12C/13C, 14 N/15N, and 1H/2H. The small differences present when

ethods of stable-isotope incorporation with mass spectrometry (MS)-based proteome quantification have advanced rapidly in the past decade. Peptide samples can be differentially tagged with heavy or light isotopes by metabolic labeling1,2 or chemical labeling.3,4 The mass differences can be distinguished at either the MS or tandem mass spectrometry (MS/MS) level. Dimethyl labeling,3 a chemical labeling method, is widely used for proteome quantification at the MS level. Several advantages of this method include quick reaction, high labeling efficiency, low cost, and applicability to different types of samples including tissues, cells, and body fluids.5 Isobaric tags for relative and absolute quantitation (iTRAQ),4 an MS/MS-based method, allows the proteome quantification of up to eight samples simultaneously and provides more precise quantification results than the MS level quantification method.6,7 Although these strategies have been widely used for proteome quantification, their accuracy and dynamic range are limited by the signal-tonoise ratio and the increased MS spectral complexity leads to fewer quantified proteins for MS level-based quantification approaches.8 In addition, ratio distortion caused by precursor © 2013 American Chemical Society

Received: September 6, 2013 Accepted: November 1, 2013 Published: November 1, 2013 10658

dx.doi.org/10.1021/ac402834w | Anal. Chem. 2013, 85, 10658−10663

Analytical Chemistry

Letter

fractions were pooled as equal-interval fractions, such as fraction 1 mixed with fraction 21 and fraction 20 mixed with fraction 40.17 All of the fractions were then lyophilized in a SpeedVac, and the samples were stored at −20 °C until use. LC−MS/MS Analysis. All of the samples were resuspended in water containing 0.1% formic acid (FA). The samples were analyzed using a low-pH RPLC−ESI-MS/MS system consisting of a quaternary surveyor MS pump (Thermo, CA) and a LTQOrbitrap Velos mass spectrometer (Thermo, CA) with a nanospray source. Buffer C (98% H2O + 2% ACN + 0.1% FA) and Buffer D (2% H2O + 98% ACN + 0.1% FA) were used for gradient separation. Peptides were separated on a homemade C18 column (150 mm × 75 μm i.d., Agela, Tianjin, China). Gradient elution was performed using a gradient of 5% to 35% acetonitrile in 0.1% FA with run times of 49 min for the BSA samples, 94 min for the Hca-P samples, and 50 min for the Hca-P and Hca-F mixture at a flow rate of 300 nL/min. Peptides were detected using a LTQ-Orbitrap Velos mass spectrometer in a data-dependent acquisition mode. For each cycle, survey scan MS was acquired from m/z 350 to 1 800 at a resolution of 60 000 at m/z = 400. For optimization of MS/MS resolution, the three most intense ions were selected for MS/ MS scan by HCD at resolutions of 15k, 30k, and 60k, respectively. For the quantification of the peptides, the six most intense ions were selected for MS/MS scan at a resolution of 30k in centroid mode. The dynamic exclusion function was set to a repeat count of 1, and an exclusion duration of 40 s was used. The normalized collision energy was set to 40%. The temperature of the ion transfer capillary was set to 200 °C. The spray voltage was set to 2.0 kV. One microscan was used for each MS and MS/MS scan. Data Analysis. The acquired raw files were analyzed by MaxQuant18 (version 1.2.2.5). The Andromeda19 program, which is embedded in MaxQuant, was used to search the peak lists against BSA.fasta or the target-decoy IPI-mouse database (version 3.68, 56 729 entries). Common contaminants were added to this database. Cysteine carbamidomethylation was searched as a fixed modification, whereas N-terminal acetylation and methionine oxidation were searched as variable modifications. Heavy- and light-labeled samples were searched independently using dimethyl (+34.063 12 Da) and dimethyl (+34.063 12) N-termini and K as variable modifications. For the four-plex sample, dimethyl (+30.038 01)/dimethyl (+34.063 12 Da) or dimethyl (+30.043 85)/dimethyl (+34.068 96 Da) of N-termini and K were set as light/heavy labels for quantification in two independent search. The two results were then combined. Peptide identification was based on a search with an initial mass deviation of up to 6 ppm for the precursor ions and an allowed fragment mass deviation of 20 ppm. Enzyme specificity was set to trypsin for BSA samples or Lys-C for the complex samples. Two mis-cleavages were allowed and a minimum of six amino acids per identified peptide was required. A FDR of 0.01 for proteins and peptides was required. All of the raw files were converted to *.mgf files using pXtract (part of the pFind program20). The intensity values of the peptide a, b, and y fragment ion pairs were extracted from the mgf files using java scripts built in-house to calculate the H/ L ratio of the peptide. The ratio of each PSM was obtained by calculating the ratio of the total intensities of the heavy and light paired fragment ions.8 Peptide and protein ratios were calculated as the median of all spectra matching the same peptide and the median of all the quantified unique peptides

lysine (K) is composed of these different isotopes enable proteome quantification with good accuracy and dynamic range. However, it suffers from long cycle time to acquire the MS1 spectra since the resolution of MS1 was as high as 480k. We hypothesized that the incorporation of 12C/13C and 1 H/2H into Lys-C protein digests by dimethyl labeling would generate a new fragment ion-based quantification method, which was termed as pseudo-isobaric dimethyl labeling (pIDL). Compared to other reported fragment ion-based quantification methods, such as IPTL12 and QITL,15 by our proposed method, the labeling of the amino groups of N termini and K were performed simultaneously, beneficial to improve quantification accuracy and precision and decrease the side reaction.

■

METHODS Sample Preparation. Approximately 2 × 106 Hca-F or Hca-P cells were inoculated subcutaneously and grown in the abdominal cavity of inbred Chinese 615 mice for 7 days.16 The cells in ascites were collected. The cells from both cell lines were then washed three times with cold 1× PBS buffer, further homogenized in 2 mL of lysis buffer (1% (v/v) protease inhibitor cocktail in 8 M urea,) using a Tissue Tearor from Biospec Products (Bartlesville, OK) at approximately 10 000 rpm for 1 min, sonicated at 100 W for 100 s, and centrifuged at 25 000g for 40 min at 4 °C. The supernatant was collected and proteins were precipitated by the addition of cold acetone. After centrifugation, the pellets were lyophilized in a SpeedVac (Thermo Fisher Scientific, San Jose, CA) and subsequently resuspended in 8 M urea. Protein concentration was determined by the Bradford assay (Beyotime, Nantong, China), and the samples were stored at −20 °C until use. Disulfide bonds were reduced by incubating the protein with DTT at 56 °C for 1 h, followed by alkylation of the proteins by IAA in the dark at room temperature for 40 min. The solution was then diluted to 1 M urea with 50 mM phosphate buffer (pH 8.0). Finally, trypsin was added to the BSA samples at a ratio of 1:50 (enzyme/protein, w/w) and incubated at 37 °C for 24 h. For the Hca-F or Hca-P protein samples, Lys-C was added at a weight ratio of 1:50 (enzyme/protein) and incubated at 37 °C for 24 h. Dimethyl Labeling. Both BSA and Lys-C Hca-P digests were labeled with 13CD2O and NaCNBH3 (light labeling, L) and with CD 2O and NaCNBD3 (heavy labeling, H), respectively, as previously described.5 Next, the two differently labeled BSA digests were mixed at ratios of H/L=1:100, 1:50, 1:10, 1:5, 1:1, 5:1, 10:1, 50:1 and 100:1; Hca-P digests were mixed at ratios of H/L = 1:1, 5:1, 10:1, and 40:1. All the mixed samples were then desalted by C18-trap column and lyophilized in a SpeedVac. All the samples were then stored at −20 °C until use. Lys-C Hca-P protein digests were labeled by 13CH2O, NaCNBH3 (30L) and 13CD2O, NaCNBH3 (34L), whereas Hca-F protein digests were labeled by CH2O, NaCNBD3 (30H), and CD2O, NaCNBD3 (34H). After the samples were quenched with a 1% NH3 solution, they were mixed at a ratio of 1:1:1:1 (w/w). The mixture of labeled peptides was separated by high-pH RPLC using an Agilent 1290 Infinity LC system (Santa Clara, CA). Buffer A (100% H2O + 25 mM NH4FA, pH = 9.5) and Buffer B (10% H2O + 90% ACN + 25 mM NH4FA) were used for gradient separation. The gradient elution was performed using 2% B (0−2 min), 2−5% B (2−3 min), 5−35% B (3−78 min), 35−80% B (78−78.1 min), and 80% B (78.1− 82 min), with fractions collected every 2 min. The resulting 10659

dx.doi.org/10.1021/ac402834w | Anal. Chem. 2013, 85, 10658−10663

Analytical Chemistry

Letter

Figure 1. The pIDL method and the effect of MS/MS resolution on quantification. (a) Mass increments for one amino group labeled with different types of commercially available formaldehyde and sodium cyanoborohydride. (b) Schematic representation of the mass defect-based pseudo-isobaric dimethyl labeling method. Lys-C protein digests were labeled with 13CD2O and NaCNBH3, in addition to CD2O and NaCNBD3, as light and heavy isotopologues, respectively. The mass difference was 11.68 mDa, which is indistinguishable at the MS level. However, high-resolution MS/MS is capable of resolving fragment ions with a mass difference of 5.84 mDa, which allows proteome quantification. (c) MS/MS spectrum of TYFPHFDVSHGSAQVK. (d) Resolved fragment ions with the MS/MS resolution at 15k, 30k, and 60k (from top to bottom).

from a protein.7,21 Proteins quantified at least twice and with at least a 2-fold change were considered to be differentially expressed proteins.

lymph node metastasis rates16 as model samples to investigate the performance of the pIDL method. Effects of MS/MS Resolution on Proteome Quantification. Collision-induced dissociation (CID) in the ion trap is widely used for peptide identification. However, it suffers from low resolution and 1/3 cutoff. However, in the pIDL method, only a 5.84 mDa mass difference between light and heavy isotopologues, which can be resolved by high MS/MS resolution, was used for quantification. Therefore, we selected higher-energy collisional dissociation (HCD) as the collision mode and Orbitrap as the mass analyzer. The resolving power of Orbitrap-type mass analyzer decreases with increasing m/z of the measured ion.22 Since the resolution equals the ratio between M and ΔM for larger ions with larger m/z, the ΔM is larger than 5.84 mDa, so they cannot be resolved. Thus, only fragment ions in the low mass range can be resolved by a current LTQ-Orbitrap Velos mass spectrometer. Higher resolution enables more fragment ion pairs to be resolved, but longer cycle time is needed to collect the spectra. Therefore, we investigated the effect of MS/MS resolution (15k to 60k at m/z = 400) on proteome quantification. When peptide quantification was performed, to ensure the fragment ions were correctly selected, the mass tolerance of fragment ions was set as 2 mDa. Taking TYFPHFDVSHGSAQVK from hemoglobin as an example (Figure 1c,d), we plotted the distinguished fragment ion pairs at the resolution described

■

RESULTS AND DISCUSSION IDL Method for Proteome Quantification. Currently, four types of formaldehyde (CH2O, 13CH2O, CD2O, and 13 CD2O) and two types of sodium cyanoborohydride (NaCNBD3 and NaCNBH3) are commercially available (Figure 1a). The labeling reaction introduced two carbon atoms and four hydrogen atoms to the peptides. If 13CD2O + NaCNBH3 was applied, the mass increment is 34.063 12, which is the mass of (13CD2)2. If CD2O + NaCNBD3 was applied, the mass increment is 34.068 96, which is the mass of (CD3)2-H2. Other combination of different labeling reagents can also be used to generate pseudo-isobaric labeling. We used Lys-C to digest proteins because its cleavage site is lysine (K). N- and C-terminal amino groups of peptides can be pseudo-isobaric dimethyl labeled simultaneously. Therefore, all of the a-, b-, and y-type fragment ion counterparts would have mass shifts of 5.84 mDa (Figure 1b). The fragment ions with such subtle mass differences can be resolved by high-resolution MS/MS, which were used for proteome quantification. In this study, we selected two mouse hepatocarcinoma ascites syngeneic cell lines with low (Hca-P) and high (Hca-F) 10660

dx.doi.org/10.1021/ac402834w | Anal. Chem. 2013, 85, 10658−10663

Analytical Chemistry

Letter

Figure 2. Quantification results using the pIDL method. (a) Box plots showing the PSMs, peptides, and proteins measured at a mixing ratio of 1:1. (b) Box plots showing the ratios measured (box and whiskers) and the expected ratios (red dashed line) at mixing ratios of 5, 10, and 40. (c) Comparison of the expected ratios versus the measured ratios. Differentially labeled BSA digests were mixed at ratios of 100:1, 50:1, 10:1, 5:1, 1:1, 1:5, 1:10, 50:1, and 100:1. The average of three measured ratios (red cross) and SD values (error bar) are displayed. (d) The two quantification results obtained by four-plex pIDL method demonstrated a strong correlation for analyzing the differentially expressed proteins of Hca-F and the Hca-P cell lines (R2 = 0.887).

above. Only one fragment ion pair was used for quantification at the resolution of 15k, whereas 5 and 10 fragment ion pairs were used for quantification at resolutions of 30k and 60k, respectively. At the 30k resolution, 99.5% of all spectra can be used for quantification, and less cycle time was required than at the 60k resolution (Supporting Information S-Table 1). Therefore, the 30k MS/MS resolution was utilized in the following study. Accuracy and Dynamic Range. We next investigated the accuracy and precision of the pIDL method with heavily and lightly labeled Hca-P protein digests at mixing ratios of 1, 5, 10, and 40. The 1:1 ratio yielded 5374 peptide-spectrum matches (PSMs) using the pIDL method, of which 99.1% were quantified in three replicate runs. Additionally, over 99.9% of all the ratios ranged from 0.5 to 2 and the 25 and 75 percentiles were 0.959 and 1.100, respectively (Figure 2a). All identified proteins can be quantified with ratios ranging from 0.700 to 1.44, and the RSD values were less than 10% for 87.2% of the proteins that were quantified at least twice in the three replicate runs. For heavily and lightly labeled Hca-P protein digests at mixing ratios of 5, 10, and 40, the measured ratios were 5.48, 11.15, and 45.65, respectively, with relative errors of 9.6%, 11.5%, and 14.1%, respectively (Figure 2b). Thus, no significant

ratio compression was observed, which demonstrates the quantification accuracy of pIDL in a wide dynamic range. We further evaluated the dynamic range of pIDL method according to refs 12, 15, and 23 by analyzing the mixtures of tryptic BSA digests with H/L ratios of 100:1, 50:1, 10:1, 5:1, 1:1, 1:5, 1:10, 1:50, and 1:100 (w/w). According to the pIDL method shown in Figure 1b, only peptides with K as C terminal were used for BSA quantification. The quantification results were listed in S-Table 2 in the Supporting Information. We plotted the average values of measured ratios acquired from the three replicate runs against their expected ratios. Good linearity (y = 0.988x − 0.110,) across a 100-fold dynamic range with R2 = 0.999 (Figure 2c). All of the mixtures were precisely quantified with RSD values lower than 9%. Complete labeling of all peptides is imperative in accurate protein quantification. Both identified labeled and unlabeled peptides were used to calculate the labeling efficiency with variable modification of the N-terminal amine or lysine residue set as dimethyl (+34.068 96 Da or +34.063 12 Da, respectively) during the database search. Using the heavily and lightly labeled Hca-P protein digest ratio of 1:1 as an example, the values obtained were ≥99.9% for lysine and ≥99.4% for N-terminal amines (Supporting Information S-Table 3). 10661

dx.doi.org/10.1021/ac402834w | Anal. Chem. 2013, 85, 10658−10663

Analytical Chemistry

Letter

to provide a novel proteome quantification technique. Because of the high labeling efficiency via a one-pot reaction, specific fragment ion pairs for quantification, and enhanced purity of fragment ions by high-resolution MS/MS, such a strategy showed advantages of wide dynamic range, high accuracy, and good precision. Furthermore, pIDL was also successfully used as a 4-plex to improve the reproducibility of proteome quantification. All these results demonstrated that pIDL might become a promising technique to achieve the large scale proteome quantification.

Because fragment ions have higher specificity for peptides than reporter ions using the pIDL method, the ratio compression caused by coelution was reduced. Besides, fragment ions were purified since contaminating ions and fragment ions can be resolved at the MS/MS resolution of 30k (m/z=400). To ensure the fragment ions were correctly selected, the mass tolerance of fragment ions was set as 2 mDa. Using ETYGDMADCCEK from BSA with H/L=1 as an example (Supporting Information S-Figure 1), many contaminating ions around the lightly and heavily labeled a1 ions were resolved, thus the a1 ion pair used for quantification was purified. The 2 mDa mass tolerance of a1 ion ensured the correct a1 ions were correctly selected for further analysis. Several fragment ion-based methods were developed, such as IPTL,12 QITL15 and IVTAL,24 by which only the PSM with at least 4 pairs of fragment ions were used for peptide quantification to obtain accurate quantification results. In our method, we compared the obtained H/L ratios obtained from 1 to 10 fragment ion pairs, as shown in Supporting Information Figure 2. The median values of different pairs ranged from 1.01 to 1.05, which showed no significant difference among the results by t test. Additionally, the SD values obtained from the assay ranged from 0.0618 to 0.225 (Supporting Information SFigure 2). Therefore, by our proposed method, even the fragment-ion-pair number as low as 1 is capable of providing accurate quantification results. Four-Plex pIDL Method and Its Application. The pIDL method is based on the subtle mass defects between 12C/13C and 1H/2H and results in three isotopic peaks by calculating the incorporation of these six reagents according to Figure 1a. Each peak is capable of quantifying two samples. Therefore, quantification data of up to six samples (six-plex) can be acquired simultaneously using the pIDL method. We analyzed the proteomes from the Hca-P and Hca-F cell lines using the four-plex pIDL method. Hca-P protein digests were labeled by 13CH2O, NaCNBH3 (30L) and 13CD2O, NaCNBH3 (34L), whereas Hca-F protein digests were labeled by CH2O, NaCNBD3 (30H), and CD2O, NaCNBD3 (34H). In this method, two quantification results can be achieved in just one LC−MS experiment. In total 2403 proteins were quantified by both light (30H/30L) and heavy (34H/34L) isotopologues. The ratio of 30H/30L and the ratio of 34H/34L showed excellent correlation (R 2 = 0.887, Figure 2d), which demonstrates the high reproducibility of protein quantification obtained using the 4-plex pIDL method. For the four-plex pIDL method, the proteins quantified at least twice were further analyzed. In total, 2517 unique proteins were reliably quantified, of which 80 proteins were up-regulated in Hca-F and 114 proteins were down-regulated in Hca-F (Supporting Information S-Table 4). The average RSD value of all quantified proteins was 11.76%, demonstrating good quantification precision. However, in the four-plex pIDL method, about 55% of all the peptides can be fragmented by both lightly and heavily labeled isotopologues due to the limited scan speed of the mass spectrometer. Meanwhile, about 75% of all the quantified proteins can be quantified by both +30 and +34 isotope clusters. We believe that more peptides labeled with both light and heavy reagents can be fragmented with the development of the mass spectrometer.

■

ASSOCIATED CONTENT

S Supporting Information *

(S-Figure 1) Spectrum of a1 ion of ETYGDMADCCEK (H/L = 1) from BSA; (S-Figure 2) effect of fragment ion pair numbers of PSMs on quantification results; (S-Table 1) effect of resolving power on protein quantification; (S-Table 2) quantification results of heavily and lightly labeled BSA digests with different ratios; (S-Table 3) dimethyl labeling efficiency at N-terminal α-amine group and ε-amine group of lysine in the peptides; and (S-Table 4) differentially expressed proteins in Hca-P and Hca-F cell lines. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone and fax: +86-41184379720. Author Contributions

Yuan Zhou and Yichu Shan contributed equally to this work. Y.Z., L.Z., and Y.Z. designed the experiments, analyzed the data, and wrote the paper. Y.S. developed software for data analysis. Q.W. and S.Z. participated in isotope labeling of protein samples and data analysis. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The work was supported by the China State Key Basic Research Program (Grant 2012CB910604), National Natural Science Foundation (Grants 20935004 and 21005079), and The Creative Research Group Project by NSFC (Grant No. 21021004). We also thank Prof. Shujuan Shao from Dalian Medical University for providing Hca-P and Hca-F samples.

■

REFERENCES

(1) Oda, Y.; Huang, K.; Cross, F. R.; Cowburn, D.; Chait, B. T. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 6591−6596. (2) Ong, S. E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Mol. Cell. Proteomics 2002, 1, 376− 386. (3) Hsu, J. L.; Huang, S. Y.; Chow, N. H.; Chen, S. H. Anal. Chem. 2003, 75, 6843−6852. (4) Ross, P. L.; Huang, Y. N.; Marchese, J. N.; Williamson, B.; Parker, K.; Hattan, S.; Khainovski, N.; Pillai, S.; Dey, S.; Daniels, S.; Purkayastha, S.; Juhasz, P.; Martin, S.; Bartlet-Jones, M.; He, F.; Jacobson, A.; Pappin, D. J. Mol. Cell. Proteomics 2004, 3, 1154−1169. (5) Boersema, P. J.; Raijmakers, R.; Lemeer, S.; Mohammed, S.; Heck, A. J. Nat. Protoc. 2009, 4, 484−494. (6) Choe, L.; D’Ascenzo, M.; Relkin, N. R.; Pappin, D.; Ross, P.; Williamson, B.; Guertin, S.; Pribil, P.; Lee, K. H. Proteomics 2007, 7, 3651−3660.

■

CONCLUSIONS In summary, on the basis of the subtle mass defect differences between 12C/13C and 1H/2H, the pIDL method was developed 10662

dx.doi.org/10.1021/ac402834w | Anal. Chem. 2013, 85, 10658−10663

Analytical Chemistry

Letter

(7) Mertins, P.; Udeshi, N. D.; Clauser, K. R.; Mani, D. R.; Patel, J.; Ong, S. E.; Jaffe, J. D.; Carr, S. A. Mol. Cell. Proteomics 2012, 11, M111 014423. (8) Zhang, G.; Neubert, T. A. Mol. Cell. Proteomics 2006, 5, 401−411. (9) Ting, L.; Rad, R.; Gygi, S. P.; Haas, W. Nat. Methods 2011, 8, 937−940. (10) Wenger, C. D.; Lee, M. V.; Hebert, A. S.; McAlister, G. C.; Phanstiel, D. H.; Westphall, M. S.; Coon, J. J. Nat. Methods 2011, 8, 933−935. (11) Koehler, C. J.; Strozynski, M.; Kozielski, F.; Treumann, A.; Thiede, B. J. Proteome Res. 2009, 8, 4333−4341. (12) Koehler, C. J.; Arntzen, M. O.; Strozynski, M.; Treumann, A.; Thiede, B. Anal. Chem. 2011, 83, 4775−4781. (13) Koehler, C. J.; Arntzen, M. O.; de Souza, G. A.; Thiede, B. Anal. Chem. 2013, 85, 2478−2485. (14) Hebert, A. S.; Merrill, A. E.; Bailey, D. J.; Still, A. J.; Westphall, M. S.; Strieter, E. R.; Pagliarini, D. J.; Coon, J. J. Nat. Methods 2013, 10, 332−334. (15) Yang, S. J.; Nie, A. Y.; Zhang, L.; Yan, G. Q.; Yao, J.; Xie, L. Q.; Lu, H. J.; Yang, P. Y. J. Proteomics 2012, 75, 5797−5806. (16) Song, B.; Tang, J. W.; Wang, B.; Cui, X. N.; Hou, L.; Sun, L.; Mao, L. M.; Zhou, C. H.; Du, Y.; Wang, L. H.; Wang, H. X.; Zheng, R. S.; Sun, L. World J. Gastroenterol. 2005, 11, 1463−1472. (17) Song, C.; Ye, M.; Han, G.; Jiang, X.; Wang, F.; Yu, Z.; Chen, R.; Zou, H. Anal. Chem. 2010, 82, 53−56. (18) Cox, J.; Mann, M. Nat. Biotechnol. 2008, 26, 1367−1372. (19) Cox, J.; Neuhauser, N.; Michalski, A.; Scheltema, R. A.; Olsen, J. V.; Mann, M. J. Proteome Res. 2011, 10, 1794−1805. (20) Wang, L. H.; Li, D. Q.; Fu, Y.; Wang, H. P.; Zhang, J. F.; Yuan, Z. F.; Sun, R. X.; Zeng, R.; He, S. M.; Gao, W. Rapid Commun. Mass Spectrom. 2007, 21, 2985−2991. (21) Bantscheff, M.; Lemeer, S.; Savitski, M. M.; Kuster, B. Anal. Bioanal. Chem. 2012, 404, 939−965. (22) Scigelova, M.; Makarov, A. Proteomics 2006, 6 (Suppl2), 16−21. (23) Sohn, C. H.; Lee, J. E.; Sweredoski, M. J.; Graham, R. L.; Smith, G. T.; Hess, S.; Czerwieniec, G.; Loo, J. A.; Deshaies, R. J.; Beauchamp, J. L. J. Am. Chem. Soc. 2012, 134, 2672−2680. (24) Nie, A. Y.; Zhang, L.; Yan, G. Q.; Yao, J.; Zhang, Y.; Lu, H. J.; Yang, P. Y.; He, F. C. Anal. Chem. 2011, 83, 6026−6033.

10663

dx.doi.org/10.1021/ac402834w | Anal. Chem. 2013, 85, 10658−10663