Mass Spectrometry Method for

Sep 1, 2001 - We have developed a new technique for quantifying methionine sulfoxide (MetSO) in protein to assess levels of oxidative stress in physio...
0 downloads 10 Views 78KB Size
Download PDF
Anal. Chem. 2001, 73, 4662-4667

Isotope Dilution Gas Chromatography/Mass Spectrometry Method for the Determination of Methionine Sulfoxide in Protein Mark A. Sochaski,† Alicia J. Jenkins,‡ Timothy J. Lyons,‡ Suzanne R. Thorpe,† and John W. Baynes*,†,§

Department of Chemistry and Biochemistry and School of Medicine, University of South Carolina, Columbia, South Carolina, 29208, and Division of Endocrinology, Diabetes and Medical Genetics, Medical University of South Carolina, Charleston, South Carolina, 29425

We have developed a new technique for quantifying methionine sulfoxide (MetSO) in protein to assess levels of oxidative stress in physiological systems. In this procedure, samples are hydrolyzed with methanesulfonic acid (MSA) in order to avoid the conversion of MetSO to methionine (Met) that occurs during hydrolysis of protein in HCl. The hydrolysate is fractionated on a cation exchange column to remove the nonvolatile MSA from amino acids, and the amino acids are then derivatized as their trimethylsilyl esters for analysis by selected ion monitoring-gas chromatography/mass spectrometry. The limit of detection of the assay is 200 pmol of MetSO per analysis, and the interassay coefficient of variation is 5.8%. Compared to current methods, the SIM-GC/MS assay avoids the potential for conversion of Met to MetSO during sample preparation, requires less sample preparation time, has lower variability, and uses mass spectrometry for sensitive and specific analyte detection. Oxidative stress arises in vivo when there is an imbalance between the production of pro- and antioxidants, in favor of the oxidants.1 This imbalance is characterized by increased concentrations of reactive oxygen species (ROS) such as hydrogen peroxide, superoxide, hydroxyl radicals, peroxynitrite, and singlet oxygen, all of which can oxidatively modify proteins, lipids, and/or DNA.2 The biological damage caused by ROS is believed to contribute, in part, to the aging process3,4 and to the development and progression of chronic, age-related pathologies such as rheumatoid arthritis, atherosclerosis, and diabetes.2,5 Despite the importance of oxidative stress in the pathogenesis and progression of disease, there is currently no broad spectrum in vivo “oxidometer” * Corresponding author: (tel) (803) 777-7272; (fax) (803) 777-7272; (e-mail) [email protected]. † Department of Chemistry and Biochemistry, University of South Carolina. ‡ Medical University of South Carolina. § School of Medicine, University of South Carolina. (1) Sies, H. In Oxidative stress: Oxidants and Antioxidants; Sies, H., Ed.; Academic Press: New York, 1991; pp xv-xxii. (2) Halliwell, B.; Gutteridge, J. M. C. Free Radicals in Biology and Medicine, 3rd ed.; Oxford University Press (Clarendon): Oxford, U.K., 1999; pp 1-66. (3) Ames, B. N.; Shigenaga, M. K.; Hagen, T. M. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 7915. (4) Stadtman, E. R.; Berlett, B. S. Chem. Res. Toxicol. 1997, 10, 485. (5) Baynes, J. W.; Thorpe, S. R. Diabetes 1999, 48, 1-9.

4662 Analytical Chemistry, Vol. 73, No. 19, October 1, 2001

to measure oxidative stress, analogous to the thermometer for measuring body temperature. As an alternative approach, relative oxidative stress is quantified by measuring changes in, or comparing the levels of, oxidatively modified biomolecules or biomarkers of oxidative damage in body fluids and tissues. Examples of these biomarkers include plasma malondialdehyde or urinary isoprostanes, derived from lipid oxidation, 8-hydroxyguanosine derived from DNA damage, and o-tyrosine, dityrosine, and methionine sulfoxide (MetSO) from protein oxidation. In this report, we describe an improved assay for quantification of MetSO, one of the most sensitive biomarkers of oxidative damage to protein. MetSO is formed through the oxidation of free or proteinbound methionine (Met) (Figure 1). Because Met is so easily oxidized by ROS, it is thought to play an important role in defending proteins against oxidative stress. Met residues located near the active sites of enzymes (e.g., glutamine synthase) protect the catalytic site of the protein during times of increased oxidative stress.6,7 The amount of MetSO in tissue protein may also reflect the severity or duration of oxidative stress in disease. Thus, significantly elevated levels of MetSO occur in lens cataract8 and as a result of inflammation and oxidative stress in the bronchoalveolar lavage fluid of idiopathic pulmonary fibrosis patients9 and in R-1-protease inhibitor in synovial fluid of rheumatoid arthritis patients.10 In addition, MetSO increases with age in long-lived tissue proteins such as human skin collagen11 and in the trabecular meshwork collagen of the eye.12 A major limitation in the routine use of MetSO as a biomarker of oxidative stress is the cumbersome nature of the current analytical technique, the cyanogen bromide method described by (6) Levine, R. L.; Mosoni, L.; Berlett, B. S.; Stadtman, E. R. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 15036. (7) Levine, R. L.; Berlett, B. S.; Moskovitz, J.; Mosoni, L.; Stadtman, E. R. Mech. Aging Dev. 1999, 107, 323. (8) Garner, M. H.; Spector, A. Proc. Natl. Acad. Sci. U.S.A. 1980, 77, 12741277. (9) Maier, K.; Leuschel, L.; Costabel, U. Am. Rev. Respir. Dis. 1991, 143, 271274. (10) Wong, P. S.; Travis J. Biochem. Biophys. Res. Commun. 1980, 96, 1449. (11) Wells-Knecht, M. C.; Huggins, T. G.; Dyer, D. G.; Thorpe, S. R.; Baynes, J. W. J. Biol. Chem. 1993, 268, 12348. (12) Horstmann, H. J.; Rohen, J. W.; Sames, K. Mech. Aging. Dev. 1983, 21, 121. 10.1021/ac010228k CCC: $20.00

© 2001 American Chemical Society Published on Web 09/01/2001

In this paper, we describe a new technique for quantifying MetSO in tissue protein to assess levels of oxidative stress in physiological systems. In this procedure, samples are hydrolyzed with MSA, avoiding the conversion MetSO to Met. The hydrolysate is fractionated batchwise on a cation exchange column to separate the nonvolatile MSA from amino acids, and the amino acids are then derivatized for analysis by selected ion monitoringgas chromatography mass spectrometry (SIM-GC/MS). The method was evaluated for its limit of detection, interassay variability, and linearity of response and was applied to analysis of the MetSO content of human skin collagen and lens proteins as a function of age. The results for collagen were compared to a previous analysis of similar samples by the CNBr method.11 Finally, the kinetics of formation of MetSO during metal-catalyzed oxidation of human low-density lipoprotein (LDL) was compared to the rate of formation of N-(carboxymethyl)lysine (CML), a lipoxidation product known to accumulate in LDL during oxidation in vitro.22

Figure 1. Scheme for formation of Met oxidation products. Met is readily oxidized by reactive oxygen species (ROS) to form MetSO (A). Conversion of the sulfoxide to the sulfone (B) requires stronger oxidation conditions, not commonly observed in vivo. The enzyme methionine sulfoxide reductase (MSR) provides an intracellular repair mechanism for converting MetSO back to Met (C), but there is no biological mechanism for repairing methionine sulfone.

Schechter.13 Incubation of proteins containing Met with CNBr results in cleavage of the protein on the carboxyl side of Met residues, followed by conversion of Met to homoserine and homoserine lactone during acid hydrolysis. The CNBr incubation step itself may need to be repeated 2-3 times in order to achieve complete cleavage of a protein, resulting in long workup times. Following CNBr treatment, the protein is hydrolyzed with HCl in the presence of a reducing agent (e.g., dithiothreitol), which facilitates the conversion of MetSO to Met.14 The hydrolyzed protein is then analyzed by cation exchange HPLC, with the homoserine and homoserine lactone representing the original Met, and the Met representing the original MetSO content of the protein. The CNBr procedure suffers from other procedural and analytical shortcomings. These include incomplete CNBr cleavage resulting from neighboring group interference15 and oxidation of Met to form MetSO during the CNBr cleavage reaction.16 Interassay variability is as high as 20%.17 Other methods to measure MetSO include protein hydrolysis in methanesulfonic acid (MSA)18,19or NaOH20,21 followed by HPLC analysis, but neither is widely used. (13) Shechter, Y.; Burstein, Y.; Patschornik, A. Biochemistry 1975, 14, 4497. (14) Scorrano, G. Acc. Chem. Res. 1972, 5, 132. (15) Doyen, N.; Lapresle, C. Biochem. J. 1979, 177, 251. (16) Joppich-Kuhn, R.; Corkill, J. A.; Giese, R. W. Anal. Biochem. 1982, 119, 73. (17) Maier, K. L.; Lenz, A.-G.; Beck-Spier, I.; Costabel, U. Methods Enzymol. 1995, 251, 455. (18) Chiou, S. H.; Wang, K. T. J. Chromatogr. 1988, 448, 404. (19) Puchala, R.; Pior, H.; von Keyserlingk, M.; Shelford, J. A.; Barej, W. Anim. Feed Sci. Technol. 1994, 48, 121. (20) Neumann, N. P. Methods Enzymol. 1967, 11, 487.

EXPERIMENTAL SECTION Reagents and Chemicals. Met, MetSO, type XII-A ribonuclease (RNase), Sephadex G-25, and Dowex 50W-X8-200 were purchased from Aldrich Chemical Co. (Milwaukee, WI). Deuterated Met (3,3,4,4-d4) was purchased from CDN isotopes (PointeClaire, Quebec, Canada); MetSO-d4 was prepared by oxidation of Met-d4 according to the method of Brot.23 MSA was obtained from Acros Chemicals (Atlanta, GA), and N,O-bis(trimethylsilyl)acetamide (BSA), Methyl-8, and amino acid standard H were from Pierce Chemical Co. (Rockford, IL). All other reagents were of the highest quality obtainable from Fisher Chemical Co. (Atlanta, GA) Preparation of Internal and External Standards. Standard solutions of Met (10nmol/µL) and MetSO (1 nmol/µL) as well as Met-d4 (2.5 nmol/µL) and MetSO-d4 (0.5 nmol/µL) were prepared gravimetrically in deionized water and stored at -20 °C until needed. Final standard concentrations were confirmed by HPLC amino acid analysis using postcolumn o-phthalaldehyde derivatization and fluorescence detection,24 based on a calibration curve prepared with a commercial amino acid standard. Preparation of Cation-Exchange Resin. Dowex 50W-X8-200H+ was cleaned by boiling 3 times for 15 min in 10 volumes of 1 N NaOH, followed each time by decanting fine particles. After washing with deionized water, this procedure was repeated 3 times with 1 M HCl. The resin was then washed with 3 volumes of 0.3 M MSA and equilibrated in 0.03 M MSA. The resin was stored in 0.03 M MSA at room temperature until needed. Preparation of Oxidized RNase. Native RNase was oxidized by the method of by Savige and Fontana25 to convert all Met residues to MetSO. Briefly, RNase (10 mg) was dissolved in a solution consisting of 15 µL of dimethyl sulfoxide, 70 µL of 12 M HCl, and 150 µL of glacial acetic acid. After incubation at room temperature for 15 min, the solution was diluted 1:20 with (21) Hayashi, R.; Suzuki, F. Anal. Biochem. 1985, 149, 521. (22) Fu, M. X.; Requena, J. R.; Jenkins, A. J.; Lyons, T. J.; Baynes, J. W.; Thorpe, S. R. J. Biol. Chem. 1996, 271, 9982. (23) Brot, N.; Weissbach, H. Arch. Biochem. Biophys. 1983, 223, 271. (24) Cunico, R. L.; Schlabach, T. J. Chromatogr. 1983, 266, 461. (25) Savige, W. E.; Fontana, A. Methods Enzymol. 1977, 25, 453.

Analytical Chemistry, Vol. 73, No. 19, October 1, 2001

4663

deionized water and dried in vacuo. The oxidized RNase was reconstituted in 1 mL of deionized water and desalted on a 2 × 30 cm Sephadex G-25 column in deionized water. The column was eluted with deionized water, and 1-mL fractions were collected and analyzed at 280 nm for protein. Protein fractions were pooled, dried in vacuo, reconstituted in deionized water, and stored at -20 °C until used. The concentration of oxidized RNase was determined using the Lowry protein assay,26 calibrated by comparison to a nonoxidized RNase standard curve. Isolation of Human LDL, Skin Collagen, and Lens Proteins. Studies on human collagen and lens samples were carried out in accordance with the guidelines of the Institutional Review Boards of the University of South Carolina, Columbia SC, and the Medical University of South Carolina, Charleston SC, and informed consent was obtained from all subjects. Subjects were normal, healthy volunteers. LDL was prepared from blood plasma by single vertical ultracentrifugation and oxidized using CuCl2 as described previously.22 Briefly, LDL in phosphate-buffered saline (∼0.1 mg/mL) was adjusted to 5 µM in CuCl2 and incubated at 37 °C. The progress of the oxidation reaction was monitored by measuring conjugated diene formation at 234 nm. At various times during the oxidation reaction, two aliquots, corresponding to 1 mg of protein each, were removed, and the oxidation was quenched by the addition of a concentrated diethylenetriaminepentaacetic acid solution to achieve a final 1 mM concentration. One of the aliquots was immediately reduced with NaBH4 overnight at 4 °C for analysis of CML.22 The second aliquot was adjusted to 0.1% in butylated hydroxytoluene (BHT) to inhibit propagation of lipid peroxidation reactions. Reduced and BHTtreated samples were dialyzed against several changes of deionized water, dried in vacuo, then delipidated,22 and stored at -70 °C until analyzed. Human skin collagen was prepared as described by Dunn et al.27 Briefly, skin samples were thawed and scraped vigorously to remove adherent fat and vascular tissue. The remaining tissue was extracted sequentially with 50 volumes (w/v) each of 1 M NaCl solution, 0.5 M acetic acid, and chloroform-methanol (2:1) for 24 h at 4 °C to remove soluble protein and lipids. Samples were lyophilized and stored at -70 °C until used. Human lens proteins were obtained from the Columbia Lions Eye Bank. The lenses were decapsulated, and total lens proteins were homogenized at 4 °C in 1 mL of deionized water using a Potter-Elvejhem Teflon-glass homogenizer, dialyzed against distilled water at 4 °C, and stored frozen at -70 °C. Sample Hydrolysis and Preparation for GC/MS Analysis. All solutions (i.e., MSA, pyridine, and deionized water) were thoroughly degassed and purged with nitrogen prior to use. Protein, 0.5-1 mg, was placed in a 1-mL screw cap Reacti-vial (Pierce, Rockford, IL) followed by addition of 350 µL of freshly prepared MSA (3 M) containing 0.2% phenol as antioxidant and deuterated Met (25 nmol) and MetSO (5 nmol) as internal standards. After purging the headspace with nitrogen for 10 s, the vials were sealed with Teflon-lined caps. Samples were hydrolyzed in a heating block at 150 °C for 2 h. After cooling to (26) Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. J. Biol. Chem. 1951, 193, 265. (27) Dunn, J. A.; McCance, D. R.; Thorpe, S. R.; Baynes, J. W. Biochemistry 1991, 30, 1205.

4664

Analytical Chemistry, Vol. 73, No. 19, October 1, 2001

room temperature, the hydrolysate was diluted with deionized water to 35 mL (1:100, final concentration 0.03 M MSA, pH ∼1) and applied to a Dowex cation exchange column containing 1 mL of resin prepared as described above. The column was washed with 3-mL aliquots of deionized water to remove residual MSA. Amino acids were then eluted into a 13 × 100 mm glass screw cap test tube with 3 mL of a 2 M aqueous pyridine solution and dried in vacuo. The dried amino acids were derivatized in a mixture of pyridine (10 µL) and BSA (150 µL) for 2 h at room temperature under a nitrogen atmosphere. Derivatized samples were transferred to 1.5-mL Eppendorf vials and centrifuged at 9000×g to remove any particulate matter that formed during sample workup and then transferred to autosampler vials. Instrumentation and SIM-GC/MS Analysis of MetSO. Analysis of derivatized amino acids was carried out using a Hewlett-Packard (Palo Alto, CA) series HP5890 gas chromatograph coupled to a HP5970 mass-selective detector. The GC column was a 30 m Rtx-5 capillary column (0.25-µm i.d.) (Restek, Bellefonte, PA). The temperature program was as follows: initial, 120 (hold 4.3 min) to 280 °C at 15 °C/min (hold 3 min.); total run time 18 min. One microliter of sample was injected in splitless mode with an inlet temperature of 250 °C. Helium was used as the carrier gas, maintained at a constant flow rate of 1 mL/min. Detector settings were as follows: temperature, 250 °C; scan time, 90 ms; solvent delay, 8.0 min. SIM was used to record the abundances of the TMS derivatives of Met, Met-d4, MetSO, and MetSO-d4 at m/z ) 250, 254, 264, and 268, respectively. Quantification of analytes was by isotope dilution MS, using standard curves constructed by adding increasing amounts of unlabeled analyte to a constant amount of the corresponding labeled compound. The peak areas of both labeled and unlabeled Met and MetSO were integrated, and the resulting values were used to calculate the amount of Met and MetSO present in the sample. The percent MetSO was expressed as MetSO/(Met + MetSO). RESULTS AND DISCUSSION GC and Full-Scan Analysis of TMS Derivatives of Met and MetSO. To determine which derivatization reagent would be best suited for this assay, a number of different methods were evaluated. These procedures included derivatization with acetic or trifluoroacetic anhydrides, methyl- or ethylchloroformates, diazomethane, Methyl-8, and methanolic HCl. Although only one Met product was observed for any given method, modification with BSA to form the trimethylsilyl (TMS) adduct was the only procedure yielding a single product peak for MetSO. Figure 2 shows a typical chromatogram for a mixture of the Met and MetSO standards derivatized with BSA, eluting at ∼10 and 13 min, respectively. In addition, the full-scan electron impact mass spectrum and fragmentation pattern for each analyte is also illustrated.28 The ion at m/z ) 250 (m/z ) 254 for Met-d4) was selected as the analytical ion for Met because the other major ions (m/z ) 73, 128, 176, and 219) were common to many amino acids; the ion at m/z ) 293 could not be resolved from a closely eluting peak in protein hydrolysates containing the same ion. For MetSO, m/z ) 264 (m/z ) 268 amu for MetSO-d4), was selected, again because m/z ) 73, 128, 176, and 218 occur in most amino acids. (28) Leimer, K. R.; Rice, R. H.; Gehrke, C. W. J. Chromatogr. 1977, 141, 355.

Figure 3. SIM-GC/MS chromatogram of skin collagen from an 18year-old donor. An acid hydrolysate of skin collagen (1 mg of protein) was analyzed for Met (A) and MetSO (B) by SIM-GC/MS. No interfering peaks were detected for either Met or MetSO.

Figure 2. GC/MS chromatogram and mass spectra of TMS derivatives of Met and MetSO standards. (A) Total ion current chromatogram (TIC) of the selected ions (SIM) used for Met and MetSO, showing elution times for Met (20 nmol) and MetSO (5 nmol). Panels B-D show mass spectra and proposed fragmentation patterns for the TMS derivatives of Met, MetSO, and MetSO-d4, respectively. The TMS derivative of MetSO is formed by a Pummerer rearrangement35 of the trimethylsilylated R-methylenic sulfoxide.

Limit of Detection and Interassay Variability of MetSO Assay. The limit of detection (LOD) of the MetSO assay was determined using a propagation of error formula for chromatography.29 A calibration curve was constructed in triplicate with increasing amounts of MetSO (0-0.5 nmol/sample) and a constant amount of MetSO-d4 (0.3 nmol/sample). The LOD for the SIM-GC/MS analysis was calculated to be 200 pmol of MetSO/ injection. Additional calculations were performed to determine the minimum quantity of 18-year-old human skin collagen that would be required to satisfy this LOD calculation. This sample was chosen because of its low MetSO concentration. Assuming that (1) Met accounts for 0.7% of the amino acids in type I collagen30 and (2) 18-year-old skin collagen contains ∼5% MetSO, a minimum (29) Foley, J. P.; Dorsey, J. G. Chromatographia 1984, 18, 503. (30) Rojkind, M. Biochem. Dis. 1973, 3, 1.

of 150 µg of protein would have to be analyzed to accommodate the LOD minimum. This amount of collagen is well below the actual 0.5-1 mg of protein used in the assay. Skin collagen from a pool of 18-year-old collagen was routinely included in sample analyses for quality control (QC) purposes. Typical SIM-GC/MS chromatograms for this sample are shown in Figure 3, illustrating both the natural-abundance Met and MetSO from the sample and the heavy labeled internal standards. The interassay variability of the method was calculated from a QC plot constructed over a 6-month period. The average value for the amount of MetSO in the QC sample was 5.8 ( 0.5% or 9.0% interassay coefficient of variation. This compares favorably to 10-20% reported for the CNBr assay.17 Assay Response. The ability to prepare sample mixtures with known percentages of oxidatively modified Met allowed us to determine the actual versus theoretical response of the assay. Figure 4 shows the results of the analysis of MetSO in mixtures of oxidized and nonoxidized RNase prepared in vitro. An important observation in this experiment is that the slope of the graph does not equal the predicted value of 1. This implies that the observed MetSO values were slightly lower than the predicted values present in the sample. One possible explanation for this discrepancy was that nonoxidized RNase was used as a standard in the Lowry protein assay26 for quantifying the protein content of the oxidized RNase solution. This may have contributed to an overestimation of the actual oxidized RNase concentration. The Analytical Chemistry, Vol. 73, No. 19, October 1, 2001

4665

Figure 4. Observed vs calculated percent MetSO in mixtures of oxidized and nonoxidized RNase. Mixtures of the two forms of RNase were prepared to simulate varying degrees of oxidized protein. Samples were analyzed in triplicate; data are mean ( standard deviation.

relative protein concentrations of the two standards were therefore compared by amino acid analysis. RNase and oxidized RNase solutions were each hydrolyzed in HCl overnight at 110 °C, effectively converting all MetSO to Met residues in both samples. The hydrolysate was dried in vacuo and analyzed by HPLC with postcolumn derivatization.24 Protein concentrations were determined by integrating the valine peak (selected because of its acid stability and baseline resolution) and back-calculating to determine the original protein concentration present in the oxidized RNase solution. Values for the protein concentration based on this procedure resulted in only a slight improvement of the slope, from 0.91 to 0.93, suggesting that oxidation of RNase may not have been complete. To determine the lowest statistically significant difference in protein MetSO that could be detected, we applied a Student t-test to the RNase experiment described in Figure 4. Using a 95% significance level (t ) 2.160), a standard deviation calculated from the pooling of the slope and y-intercept standard deviations (0.01 and 0.22, respectively), and 25 degrees of freedom, we calculated the lowest statistically significant difference to be 0.4% MetSO. In other words, there would need to be a difference in the MetSO values of 0.4% or greater between any two samples before they could be considered statistically significant. To the best of our knowledge, a comparable calculation has never been performed on the CNBr assay. One final observation in the RNase experiment was the presence of a 1% MetSO intercept at 0% theoretical MetSO. There are two possible explanations for this result. The first is that during the assay itself there was a 1% increase in baseline levels of MetSO (i.e., 1% of the Met on the protein is oxidized to MetSO during sample workup). A second possible explanation for this result is that the commercially available RNase contains 1% oxidized Met. This is not unreasonable since Met may be oxidized to MetSO during isolation and purification of the enzyme. Comparison between CNBr and GC/MS Methods. We have previously reported analyses of skin collagen from donors of different ages (1-75, n ) 17) using the CNBr method.11 For the present study, we analyzed another group of collagen samples from donors ages 18-72 (n ) 26) by GC/MS, shown in Figure 4666

Analytical Chemistry, Vol. 73, No. 19, October 1, 2001

Figure 5. MetSO increases with age in human skin collagen. Skin collagen from donors of various ages was analyzed by SIM-GC/MS.

Figure 6. MetSO vs age of lens. MetSO does not increase significantly with age in total lens proteins. Noncataractous lens proteins were analyzed by SIM-GC/MS.

5. Linear regression of each data set gave the following equations:

CNBr method:11

y ) 0.12 (age) +3.4; r2 ) 0.49

GC/MS method:

y ) 0.12 (age) +2.1; r2 ) 0.75

Thus, while the slopes of the lines for the two methods of analysis were identical, the correlation coefficient was much stronger with the GC/MS method, consistent with lower interassay variability in this method, and the y-intercept was closer to 0. Analysis of Lens Proteins. The MetSO content of human lens proteins was also measured as a function of age (Figure 6). We measured approximately 2-3% MetSO in lens crystallins but observed little change with age. These results are consistent with

measured using the GC/MS method. The progress of the oxidation reaction was monitored by the formation of conjugated dienes, measured at 234 nm (inset, Figure 7). The lag, propagation, and plateau phases of the lipid peroxidation reaction were observed as expected.33 Figure 7 also shows that the oxidation of LDL led to a concerted increase in both MetSO and CML. CML is an adduct formed on LDL from products of lipid peroxidation.22 Thus, peroxidation of lipids in LDL is accompanied by simultaneous protein modification by lipid oxidation products and oxidation of Met residues. The formation of MetSO on apolipoproteins A1 and AII during Cu2+ oxidation of HDL has been reported.34

Figure 7. MetSO vs age of lens. MetSO increases during metalcatalyzed oxidation of LDL. Progress of the oxidation reaction was monitored by measurement of conjugated dienes at 234 nm (inset). Aliquots were removed at indicated times, and LDL was analyzed for MetSO and CML content by SIM-GC/MS.

previous work of Garner and Spector,8 who reported negligible levels of MetSO in normal lens proteins but significant increases in MetSO in cataractous lenses. Using electrospray ionization MS/ MS analysis of peptides, Lund et al.31 reported 5-9% oxidation of Met residues in R-crystallins isolated from the water-insoluble fraction of human lens proteins, a fraction that increases with age and with the onset of cataracts. Application of GC/MS Method to Oxidized LDL. Oxidative modification of LDL is thought to enhance the atherogenicity of LDL in diabetes.32 We evaluated whether analysis of MetSO formation during metal-catalyzed oxidation of LDL could be (31) Lund, A. L.; Smith, J. B.; Smith, D. L. Exp. Eye Res. 1996, 63, 661. (32) Bierman, E. L. Arterioscler. Thromb. 1992, 12, 647. (33) Esterbauer, H.; Gebicki, J.; Puhl, H.; Jurgens, G. Free Radical Biol. Med. 1992, 13, 341. (34) Garner, B.; Witting, P. K.; Waldeck, A. R.; Christison, J. K.; Raftery, M.; Stocker, R. J. Biol. Chem. 1998, 273, 6080. (35) Horner, L.; Kaiser, P. Justus Liebigs Ann. Chem. 1959, 626, 19. Horner, L. Justus Liebigs Ann. Chem. 1960, 631, 198.

CONCLUSIONS In this paper, we describe a new method for the rapid determination of the protein content of Met and MetSO by SIMGC/MS. The method has a lower interassay variability (9%), has lower baseline values for percent MetSO, and requires substantially less workup time than the commonly used procedures. The LOD for the assay is 200 pmol/ injection. The new method takes advantage of the sensitivity and selectivity of SIM-GC/MS. Comparison of percentMetSO for an age range of human skin collagen between the SIM-GC/MS method and the CNBr method validates the new method as a substantial improvement for the analysis of MetSO in biological samples. We also applied the GC/ MS method to measure MetSO formation in human lens proteins and during copper-catalyzed oxidation of LDL. The latter experiment demonstrated a correlative increase in MetSO with another biomarker of oxidative stress, CML. The SIM-GC/MS method should provide an important tool for assessing the role of oxidative stress in the development and progression of pathology and for assessing the efficacy of antioxidant therapy for treatment of disease. ACKNOWLEDGMENT This work was supported by USPHS grants (DK19971, PO1HL-55782, EY10697), by the Juvenile Diabetes Research Foundation (JDF-4-1998-272, 996001), and by the Diabetes Research and Wellness Foundation. Received for review February 26, 2001. Accepted July 25, 2001. AC010228K

Analytical Chemistry, Vol. 73, No. 19, October 1, 2001

4667