Accurate Determination of Protein Methionine Oxidation by Stable

Nov 7, 2013 - Methionine (Met) oxidation is a major modification of proteins, which converts Met to Met sulfoxide as the common product. It is challen...
0 downloads 0 Views 676KB Size
Download PDF
Technical Note pubs.acs.org/ac

Accurate Determination of Protein Methionine Oxidation by Stable Isotope Labeling and LC-MS Analysis Hongcheng Liu,* Gomathinayagam Ponniah, Alyssa Neill, Rekha Patel, and Bruce Andrien Protein Characterization, Alexion Pharmaceuticals, Inc., 352 Knotter Drive, Cheshire, Connecticut 06410, United States S Supporting Information *

ABSTRACT: Methionine (Met) oxidation is a major modification of proteins, which converts Met to Met sulfoxide as the common product. It is challenging to determine the level of Met sulfoxide, because it can be generated during sample preparation and analysis as an artifact. To determine the level of Met sulfoxide in proteins accurately, an isotope labeling and LC-MS peptide mapping method was developed. Met residues in proteins were fully oxidized using hydrogen peroxide enriched with 18O atoms before sample preparation. Therefore, it was impossible to generate Met sulfoxide as an artifact during sample preparation. The molecular weight difference of 2 Da between Met sulfoxide with the 16O atom and Met sulfoxide with the 18O atom was used to differentiate and calculate the level of Met sulfoxide in the sample originally. Using a recombinant monoclonal antibody as a model protein, much lower levels of Met sulfoxide were detected for the two susceptible Met residues with this new method compared to a typical peptide mapping procedure. The results demonstrated efficient elimination of the analytical artifact during LC-MS peptide mapping for the measurement of Met sulfoxide. This method can thus be used when accurate determination of the level of Met sulfoxide is critical.

M

containing either Met of Met sulfoxide from other peptides is not necessary if monitored by mass spectrometry, because of the unique molecular weights of the peptides.10,10,20,30 One of the challenges for the accurate determination of the level of Met oxidation is analytical artifacts because Met oxidation can occur during sample preparation and analysis. It has been demonstrated that the level of Met oxidation increases during digestion and chromatography separation.30 The presence of residual metals in digestion buffer or sample contact with metal surfaces has also been shown to generate Met sulfoxide artifacts.30 It has also been demonstrated that the levels of oxidation detected by MALDI-TOF correlate with ozone levels in the air, because of the oxidation of Met and tryptophan during co-crystallization of the sample with the matrix on the plate.32 The oxidation of Met residue can occur during sample analysis by capillary zone electrophoresis (CZE) and electrospray mass spectrometry, as a result of the electrolysis of water.33 The level of oxidation increases with the increase of the potential applied to the electrospray needle when the potential is above the on-site of corona discharge.34 The oxidation of Met can also occur in the gas phase caused by electrical discharge from an eroded emitter when samples are analyzed using electrospray ionization mass spectrometry.35 In order to eliminate oxidation of Met during various steps of sample preparation and analysis, a procedure of forced oxidation was developed in the current study. The samples of

ethionine oxidation is a major protein degradation pathway both in vivo1−4 and in vitro.5,6 In vivo oxidation has been related to aging and several pathological conditions.1−3 In vitro oxidation can occur under many conditions, such as during purification,7−9 storage,10−13 light exposure,12,14 and exposure to oxidizing reagents8,10,15−21 or free radicals generated in the presence of metals.3,12,21−23 In vitro oxidation of proteins has enabled detailed studies of the impact of oxidation on protein structure, stability, and function. Oxidation of the susceptible Met residues of proteins has been shown to result in structural changes,1,8,15,19,24 decreased stability,19,24 increased propensity to aggregation and precipitation,15,22 loss of biological functions,1,5,8,8,9,15,16,25−27 and decreased half-life.28 Met oxidation may also result in an increase in immunogenicity.13,29 Therefore, it is critical to measure the accurate level of Met oxidation in recombinant therapeutic proteins to understand its impact on structure, stability, and function, and to control its level during storage to ensure appropriate product quality, safety, and efficacy. The common approach to detect and quantify site-specific Met sulfoxide typically involves digesting proteins into peptides and then analysis via reversed-phase high-performance liquid chromatography with either UV or mass spectrometry (MS) detection.1,8,10,11,16,17,20,22,30,31 Oxidation of Met residues results in an earlier elution of the oxidized peptides from reversed-phase columns and a molecular weight increase of 16 Da, compared to the respective unoxidized peptides. If complete separation of the oxidized and unoxidized peptides from other peptides were achieved, the level of oxidation can be determined using the peak areas monitored by UV.1,8,16,20,22,30 On the other hand, complete separation of the peptide © 2013 American Chemical Society

Received: September 25, 2013 Accepted: November 7, 2013 Published: November 7, 2013 11705

dx.doi.org/10.1021/ac403072w | Anal. Chem. 2013, 85, 11705−11709

Analytical Chemistry

■

Technical Note

RESULTS AND DISCUSSION Principle of the Method. The principle of the method is shown in Figure 1. In brief, each site of a Met residue in the

interest were oxidized using hydrogen peroxide enriched with 18 O atoms before sample preparation. Full conversion of Met to Met sulfoxide by 18O hydrogen peroxide prevents further oxidation during sample preparation and analysis, thus eliminating this artifact and allowing for the calculation of Met sulfoxide originally in the samples. The results demonstrated a much lower level of Met sulfoxide by this new method, compared to a typical LC-MS peptide mapping method.

■

MATERIALS AND METHODS The recombinant monoclonal antibody was expressed in Chinese hamster ovary (CHO) cell line and purified at Alexion (Cheshire, CT). Acetonitrile, ammonium bicarbonate, dithiothreitol (DTT), formic acid, hydrogen peroxide, iodoacetic acid, and trifluoroacetic acid (TFA) were purchased from Sigma (St, Louis, MO). Hydrogen peroxide (2−2.5%) with enriched 18 O atoms was purchased from Cambridge Isotope Laboratories, Inc. (Tewksbury, MA). Trypsin was purchased from Worthington Biochemical Corp. (Lakewood, NJ). Trypsin Digestion. Samples at final concentrations of 1 mg/mL were denatured and reduced using 6 M guanidine hydrochloride and 10 mM DTT in 20 mM Tris buffer, pH 8.0, at 37 °C for 30 min. The samples were then alkylated using iodoacetic acid at a final concentration of 30 mM at 37 °C for 30 min at pH ∼8.0. The samples were then buffer-exchanged into 50 mM ammonium bicarbonate using Zeba spin desalting columns (Thermoscientific, Rockford, IL). The buffer exchanged samples were digested using tryspin at a trypsin:antibody ratio of 1:10 (w:w) at 37 °C for 4, 8, 16, and 24 h. Oxidation of the Samples Using Hydrogen Peroxide with Enriched 18O Atom. The antibody was oxidized using hydrogen peroxide enriched with 18O atoms. Aliquots of the antibody at a concentration of 4 mg/mL in 0.1 M Tris were mixed with equal volume of 18O hydrogen peroxide and incubated at room temperature for 30 min. The samples were buffer exchanged into 0.1 M Tris, pH 8.0 using Zeba spin desalting columns. The samples were then prepared and digested using the same procedure, as described in the previous sections. A synthesized peptide was also oxidized to determine the presence of low levels of 16O hydrogen peroxide in the 18O hydrogen peroxide and its contribution to the analysis. The peptide at a concentration of 0.2 mg/mL in Tris buffer, pH 8.0 was oxidized by mixing with equal volume of the 18O hydrogen peroxide and incubated at room temperature for 30 min. LC-MS. An Agilent 1100 series HPLC system, an LC/MSD mass spectrometer, and a Zorbax C18 column (1.0 mm × 150 mm) (Agilent, Santa Clara, CA) were used to analyze the samples. Each sample, of ∼10 μg, was injected into the column at 95% mobile phase A (0.02% TFA, 0.08% formic acid) and 5% mobile phase B (0.02% TFA and 0.08% formic acid in acetonitrile). After 5 min, mobile phase B was increased to 35% within 100 min to elute and introduce the peptides into the mass spectrometer. The column was then washed using 95% mobile phase B and equilibrated using 5% mobile phase B before the next injection. Throughout the analysis, the flow rate was set at 50 μL/min and the column temperature was set at 60 °C. The mass spectrometer was run at positive mode with m/z from 200 to 3000.

Figure 1. Principle of the method.

proteins likely contains a mixture of Met and Met sulfoxide. Typically, protein samples are denatured, reduced, alkylated, and then digested using various proteases into peptides for LCMS analysis. The entire procedure lasts for several hours, up to 18 h (overnight) or even longer. The oxidation of Met residues during the sample preparation procedure will result in an overestimation of Met sulfoxide. To accurately determine the level of Met sulfoxide, the samples were oxidized using 18O hydrogen peroxide, which converts unoxidized Met to Met sulfoxide. Therefore, Met sulfoxide can no longer be generated as an artifact during sample preparation. The samples were then denatured, reduced, alkylated, and digested for LC-MS analysis. Because naturally occurring Met sulfoxide contains an 16O atom, while Met sulfoxide generated from 18O hydrogen peroxide oxidation contains an 18O atom, mass spectra of Met sulfoxide containing peptides will have two partially overlapped isotopic peak distributions with a molecular weight difference of 2 Da. The 2 Da molecular weight difference was used to differentiate and calculate the levels of Met sulfoxide in the samples and Met sulfoxide generated from 18O hydrogen peroxide oxidation. A recombinant monoclonal antibody was used as a model protein to demonstrate the utility of this procedure. Oxidation of Met in the Recombinant Monoclonal Antibody. The recombinant monoclonal antibody contains two susceptible Met residues located in the Fc region.10,24,27,36,37 Peptides that contain the two Met residues are shown in Table 1 and are referenced as “peptide 1” and “peptide 2”. Typical extracted ion chromatograms of peptide 1 Table 1. Amino Acid Sequences and Molecular Weights of Two Peptides Containing Susceptible Met Residues in Human IgG1

11706

peptide

sequence

MH+

Peptide 1 Peptide 2

DTLMISR WQQGNVFSCSVMHEALHNHYTQK

835.4 2802.2

dx.doi.org/10.1021/ac403072w | Anal. Chem. 2013, 85, 11705−11709

Analytical Chemistry

Technical Note

Figure 2. EIC of peptide 1 containing either Met (A) or Met sulfoxide (B). Mass spectra of the corresponding peptides are shown as insets as singly charged ions. Other peaks shown in the EIC containing peptides with m/z similar to peptide 1.

Figure 3. Mass spectra of (A) peptide 1 and (B) peptide 2 obtained from analysis of the sample oxidized for 30 min and digested for 4 h. A doubly charged ion is shown for peptide 1 and a triply charged ion is shown for peptide 2.

by searching m/z of the singly charged ions containing either Met or Met sulfoxide from the sample digested by trypsin for 4 h are shown in Figure 2. Mass spectra of the singly charged peptides are shown in the figure as insets. The observed molecular weight (MH+) of the peak eluted at ∼41 min is 835.5, which is in good agreement with the calculated molecular weight of 835.4. The observed molecular weight of the peak eluted at ∼36 min is ∼851.4 Da, which is 16 Da higher than the peak eluted at 41 min. This molecular weight is also in good agreement with the calculated molecular weight of this peptide containing Met sulfoxide instead of Met. The peptide containing Met sulfoxide eluted earlier than the peptide containing Met, because of increased polarity. The level of oxidation was determined by dividing the EIC peak areas of the Met sulfoxide containing peptides by the sum of the EIC peak areas of the peptide containing Met and the peptide containing Met sulfoxide. The levels of Met sulfoxide in peptide 1 were determined to be 11.5%, 21.0%, 28.6%, and 30.0% for the samples digested by trypsin for 4, 8, 16, and 24 h, respectively.

Typical extracted ion chromatograms of peptide 2 by searching m/z of the triply charged ions containing either Met or Met sulfoxide from the sample digested by trypsin for 4 h are shown in Supplemental Figure 1, in the Supporting Information. Using the EIC peak areas of the peptide containing either Met or Met sulfoxide, the levels of Met sulfoxide of peptide 2 were determined to be 7.3%, 7.6%, 11.1%, and 12.3% for the samples digested by trypsin for 4, 8, 16, and 24 h. Clearly, there is a trend of increased levels of Met sulfoxide with the increase of trypsin digestion time, as demonstrated by the analysis of peptide 1 and peptide 2. Therefore, it is challenging to determine the level of Met sulfoxide that was present originally in the samples accurately. A method that can accurately determine the level of Met sulfoxide is highly desirable. Accurate Determination of the Level of Met Sulfoxide. As shown in Figure 1, the antibody was first oxidized using 18O hydrogen peroxide to convert Met to Met sulfoxide. 11707

dx.doi.org/10.1021/ac403072w | Anal. Chem. 2013, 85, 11705−11709

Analytical Chemistry

Technical Note

Step 2: The contribution to the peak intensity at m/z 851.4, which is due to the presence of low level of 16O hydrogen peroxide in the 18O hydrogen peroxide, was calculated. The contribution of 16O hydrogen peroxide to the peak intensity at m/z of 851.4 was experimentally determined using a synthesized peptide with the same sequence as peptide 1. This synthesized peptide was oxidized using the 18O hydrogen peroxide reagent and then analyzed by LC-MS. A mass spectrum from analysis of the synthesized peptide after oxidation is shown in Supplemental Figure 4 in the Supporting Information. The ratio of the peak at m/z 851.4 over the peak at m/z 853.4 is 6.6%, suggesting the contribution of the low levels of 16O hydrogen peroxide in the 18O hydrogen peroxide is 6.6% of the peak intensity containing Met sulfoxide with 18O atoms (the peak intensity at m/z 853.4 after subtraction of the overlap). The same 6.6% was used for the calculation of peptide 2, since this contribution is independent of peptides, but rather on the purity of the reagent. Step 3: The percentage of Met sulfoxide originally in the samples was calculated by dividing the peak intensity at m/z 851.4 after subtraction of the contribution from 16O hydrogen peroxide in the reagent by the sum of the peak intensity at m/z 851.4 and m/z at m/z 853.4 after subtraction of the overlap from the third peak of the m/z 851.4 series. The percentage of Met sulfoxide was determined to be 7.8% for peptide 1. Calculation by following the same procedure resulted in 0.6% Met sulfoxide for peptide 2. Significantly lower levels of Met sulfoxide determined by this new method suggested that the levels of Met sulfoxide were overestimated by using the typical LC-MS peptide mapping procedure.

Then, the antibody was denatured, reduced, alkylated, and digested with trypsin for 4 h. Because it is critical to achieve complete oxidation, the data were first analyzed to determine whether or not oxidation was complete. By following the same procedure of searching m/z of the peptides as described in the previous section, it was found that 3.9% of peptide 1 and 0.9% of peptide 2 still contained the original Met after forced oxidation using18O hydrogen peroxide. As discussed in the previous section, ∼11.5% and ∼7.3% of Met sulfoxide were detected for peptide 1 and peptide 2, respectively, after 4 h of digestion. Assuming all of the Met sulfoxide was generated as an artifact, the remaining Met in peptide 1 and peptide 2 after forced oxidation could potentially produce 0.4% (11.5% × 3.9%) and 0.07% (7.3% × 0.9%) Met sulfoxide as artifacts for peptide 1 and peptide 2, respectively. The contribution of the remaining low percentage of Met as an analytical artifact is negligible and suggests that oxidation at room temperature for 30 min is sufficient. The mass spectra of peptide 1 and peptide 2 from analyzing the sample after oxidation with 18O hydrogen peroxide and 4 h of trypsin digestion are shown in Figure 3. For peptide 1, the peak with m/z 851.4 corresponds to the peptide containing Met sulfoxide with the 16O atom. The peak with m/z 853.4 corresponds to Met sulfoxide containing the 18O atom. Similarly, for peptide 2, the peak with m/z of 940.1 corresponds to the peptide containing Met sulfoxide with the 16O atom. The peak with m/z of 940.8 corresponds to Met sulfoxide containing the 18O atom. The presence of two isotope peak distributions for peptide 1 and peptide 2 indicated that there were low levels of Met sulfoxide present originally in the antibody for both peptides. The level of Met sulfoxide originally in the antibody can be calculated by dividing the peak intensity representing Met sulfoxide originally in the sample by the sum of the peak intensity representing both Met sulfoxide originally in the sample and Met sulfoxide generated during sample preparation. The mass spectrum of peptide 1 in Figure 3A was used to elucidate the calculation procedure. The calculation procedure is also detailed in the Supporting Information. It is worthwhile to mention that the peak at m/z of 851.4 corresponds to Met sulfoxide with the 16O atom. However, it represents not only Met sulfoxide originally in the sample but also a low amount of Met sulfoxide generated during oxidation with hydrogen peroxide, because of the presence of low levels of 16O hydrogen peroxide in the 18O hydrogen peroxide reagent. In addition, the peak at m/z of 853.4 represents not only Met sulfoxide with the 18 O atom, but also the overlap from the third peak of the m/z 851.4 series. Keeping in mind the presence of low levels of 16O hydrogen peroxide in the 18O hydrogen peroxide reagent and the overlap of the third peak of the m/z 851.4 series with the first peak of the m/z 853.4 series, the percentage of Met sulfoxide originally in the sample was calculated by the following steps. Step 1: The overlap from the third peak of the m/z 851.4 series to the peak at m/z 853.4 was calculated. This overlap was calculated by multiplying the peak intensity at m/z of 851.4 by the ratio of the third peak over the first peak using the isotope distribution of the peptide containing the original Met residue (such as the inset in Figure 2A). Subtraction of the overlap from the peak intensity at m/z 853.4 resulted in the peak intensity corresponding to peptide 1 containing Met sulfoxide with the 18O atom.

■

CONCLUSION A procedure for accurate determination of Met sulfoxide was established by eliminating oxidation of Met during sample preparations for LC-MS analysis. Each susceptible Met residue likely contains low level of Met sulfoxide. The remaining Met was oxidized to Met sulfoxide using 18O hydrogen peroxide before sample preparations. Therefore, Met sulfoxide cannot be further generated during sample preparations for peptide mapping and LC-MS analysis. The resulting mass spectra from the peptide with the fully oxidized Met residues contain two partially overlapped isotopic distributions, corresponding to either Met sulfoxide with 16O or Met sulfoxide with 18O. The molecular weight difference of 2 Da between the two isotopic distributions was used to differentiate and calculate the level of Met sulfoxide in the samples originally. Using a recombinant monoclonal antibody as a model protein, it was found that significantly lower levels of Met sulfoxide at two susceptible sites were determined using this new method compared to the typically employed peptide mapping procedure, indicating artifacts introduced during sample preparation. This procedure can be utilized where an accurate determination of the level of Met sulfoxide is critical.

■

ASSOCIATED CONTENT

S Supporting Information *

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

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: 203-271-8354. E-mail: [email protected]. 11708

dx.doi.org/10.1021/ac403072w | Anal. Chem. 2013, 85, 11705−11709

Analytical Chemistry

Technical Note

Notes

(32) Cohen, S. L. Anal. Chem. 2006, 78 (13), 4352−4362. (33) Bateman, K. P. J. Am. Soc. Mass Spectrom. 1999, 10, 309−317. (34) Morand, K.; Talbo, G.; Mann, M. Rapid Commun. Mass Spectrom. 1993, 7 (8), 738−743. (35) Chen, M.; Cook, K. D. Anal. Chem. 2007, 79 (5), 2031−2036. (36) Loew, C.; Knoblich, C.; Fichtl, J.; Alt, N.; Diepold, K.; Bulau, P.; Goldbach, P.; Adler, M.; Mahler, H. C.; Grauschopf, U. J. Pharm. Sci. 2012, 101 (11), 4248−4257. (37) Schlothauer, T.; Rueger, P.StrackeJ. O.; Hertenberger, H.; Fingas, F.; Kling, L.; Emrich, T.; Drabner, G.; Seeber, S.; Auer, J.KochS.Papadimitriou, A.MAbs.201354576586 (DOI: 10.4161/ mabs.24981).

The authors declare no competing financial interest.

■

REFERENCES

(1) Stadtman, E. R. Science 1992, 257 (5074), 1220−1224. (2) Brot, N.; Weissbach, H. Arch. Biochem. Biophys. 1983, 223 (1), 271−281. (3) Stadtman, E. R. Free Radic. Biol. Med. 1990, 9 (4), 315−325. (4) Beck, A.; Wagner-Rousset, E.; Ayoub, D.; Van, D. A.; SanglierCianferani, S. Anal. Chem. 2013, 85 (2), 715−736. (5) Manning, M. C.; Patel, K.; Borchardt, R. T. Pharm. Res. 1989, 6 (11), 903−918. (6) Manning, M. C.; Chou, D. K.; Murphy, B. M.; Payne, R. W.; Katayama, D. S. Pharm. Res. 2010, 27 (4), 544−575. (7) Linde, S.; Nielsen, J. H.; Hansen, B.; Welinder, B. S. J. Chromatogr. 2013, 530, 29−37. (8) Hsu, Y. R.; Narhi, L. O.; Spahr, C.; Langley, K. E.; Lu, H. S. Protein Sci. 1996, 5 (6), 1165−1173. (9) Vale, W.; Spiess, J.; Rivier, C.; Rivier, J. Science 2013, 213, 1394− 1397. (10) Chumsae, C.; Gaza-Bulseco, G.; Sun, J.; Liu, H. J. Chromatogr. B 2007, 850 (1−2), 285−294. (11) Kroon, D. J.; Baldwin-Ferro, A.; Lalan, P. Pharm. Res. 1992, 9 (11), 1386−1393. (12) Lam, X. M.; Yang, J. Y.; Cleland, J. L. J. Pharm. Sci. 1997, 86 (11), 1250−1255. (13) Hochuli, E. J. Interferon Cytokine Res. 1997, 17 (Suppl. 1), S15− S21. (14) Liu, H.; Gaza-Bulseco, G.; Zhou, L. J. Am. Soc. Mass Spectrom. 2009, 20 (3), 525−528. (15) Liu, J. L.; Lu, K. V.; Eris, T.; Katta, V.; Westcott, K. R.; Narhi, L. O.; Lu, H. S. Pharm. Res. 1998, 15 (4), 632−640. (16) Teh, L.-C.; Murphy, L. J.; Hug, N. L.; Surus, A. S.; Friesen, H. G.; Lazarus, L.; Chapman, G. E. J. Biol. Chem. 1987, 262 (14), 6472− 6477. (17) Frelinger, A. L., III; Zull, J. E. J. Biol. Chem. 1984, 259 (9), 5507−5513. (18) Nguyen, T. H.; Burnier, J.; Meng, W. Pharm. Res. 1993, 10 (11), 1563−1571. (19) Liu, D.; Ren, D.; Huang, H.; Dankberg, J.; Rosenfeld, R.; Cocco, M. J.; Li, L.; Brems, D. N.; Remmele, R. L., Jr. Biochemistry 2008, 47 (18), 5088−5100. (20) Houde, D.; Kauppinen, P.; Mhatre, R.; Lyubarskaya, Y. J. Chromatogr. A 2006, 1123 (2), 189−198. (21) Ji, J.; Zhang, B.; Cheng, W.; Wang, Y. J. J. Pharm. Sci. 2009, 4485−4500. (22) Li, S.; Nguyen, T. H.; Schoneich, C.; Borchardt, R. T. Biochemistry 1995, 34 (17), 5762−5772. (23) Li, S.; Schoneich, C.; Borchardt, R. T. Pharm. Res. 1995, 12 (3), 348−355. (24) Liu, H.; Gaza-Bulseco, G.; Xiang, T.; Chumsae, C. Mol. Immunol. 2008, 45 (3), 701−708. (25) Coltrera, M.; Rosenblatt, M.; Potts, J. T., Jr. Biochemistry 1980, 19 (18), 4380−4385. (26) Bertolotti-Ciarlet, A.; Wang, W.; Lownes, R.; Pristatsky, P.; Fang, Y.; McKelvey, T.; Li, Y.; Li, Y.; Drummond, J.; Prueksaritanont, T.; Vlasak, J. Mol. Immunol. 2009, 46 (8−9), 1878−1882. (27) Gaza-Bulseco, G.; Faldu, S.; Hurkmans, K.; Chumsae, C.; Liu, H. J. Chromatogr. B 2008, 870 (1), 55−62. (28) Wang, W.; Vlasak, J.; Li, Y.; Pristatsky, P.; Fang, Y.; Pittman, T.; Roman, J.; Wang, Y.; Prueksaritanont, T.; Ionescu, R. Mol. Immunol. 2011, 48 (6−7), 860−866. (29) Hermeling, S.; Crommelin, D. J.; Schellekens, H.; Jiskoot, W. Pharm. Res. 2004, 21 (6), 897−903. (30) Zang, L.; Carlage, T.; Murphy, D.; Frenkel, R.; Bryngelson, P.; Madsen, M.; Lyubarskaya, Y. J. Chromatogr. B 2012, 895−896, 71−76. (31) Ren, D.; Pipes, G.; Xiao, G.; Kleemann, G. R.; Bondarenko, P. V.; Treuheit, M. J.; Gadgil, H. S. J. Chromatogr. A 2008, 1179 (2), 198−204. 11709

dx.doi.org/10.1021/ac403072w | Anal. Chem. 2013, 85, 11705−11709