Liquid Chromatography–Tandem Mass Spectrometry Method for the Analysis of Asymmetric Dimethylarginine in Human Plasma

Nitric oxide (NO) is essential in numerous physiologic processes and may be involved in related pathologic processes. Asymmetric dimethylarginine (ADMA) is an endogenous inhibitor of isoforms of NO synthase in humans (1). ADMA originates from protein arginine methylation by protein arginine methyltransferases after protein hydrolysis (2). Enzymatic hydrolysis by dimethylarginine dimethylaminohydrolases to dimethylamine and citrulline is the major pathway for elimination of ADMA (3). Circulating ADMA is altered in patients with cardiovascular and neurologic diseases, erectile dysfunction, and many other disorders (4)(5)(6), and increased circulating ADMA independently predicts future cardiovascular events and mortality (7)(8). Short-time infusion of ADMA affects hemodynamics and cardiac function in humans (3)(9). Analytical methods for the measurement of ADMA include HPLC, capillary electrophoresis, ELISA, and mass spectrometry (MS). The commonly used HPLC methods with fluorescence detection (10) measure o-phthaldialdehyde derivatives of ADMA. These derivatives are not stable and must be analyzed on-line. Moreover, ADMA must be separated by chromatographic means from its biologically inactive isomer, symmetric dimethylamine (SDMA). Thus, these HPLC methods have long analysis times of up to 45 min. Alternatively, ADMA and SDMA can be analyzed by gas chromatography–mass spectrometry (11)(12), which shows different fragmentation patterns for the respective ions. These methods include several extraction and derivatization steps and require considerable analysis times. Recently, Kirchherr and Kühn-Velten (13) developed a mass spectrometric method that omits derivatization procedures; however, chromatographic separation of ADMA from SDMA is still required for this method. Here we report a simple and rapid liquid chromatography–tandem MS (LC-MS/MS)-based method with 1 derivatization step and sample pretreatment consisting of protein precipitation. This is the first LC-MS–based method that separates methylated arginine derivatives by specific fragmentation patterns instead of by chromatographic separation. Analyses were performed on a Varian 1200L Triple Quadrupole MS equipped with 2 Varian ProStar model 210 HPLC pumps and a Varian analytical column [50 × 2.0 mm (i.d.)] packed with Polaris C18-Ether (3 μm bead size). The mobile phases consisted of methanol containing 1 mL/L formic acid (mobile phase A) and water containing 1 mL/L formic acid (mobile phase B). Chromatography was performed at 25 °C with a flow rate of 0.4 mL/min. The gradient started with 2% A for 0.5 min and increased linearly to 50% A over 1.5 min, with subsequent reequilibration with 2% A for 2 min. Nitrogen was used as the nebulizing and drying gas (380 °C) at 90 and 180 L/h, respectively. For ionization in the positive electrospray ionization mode (ESI+), the needle and shield voltages were set at 5850 and 400 V, respectively. The following transitions were observed after fragmentation with argon (1.5 mTorr): m/z 231→70 [collision energy (CE), −22 eV] for l-arginine (USP Reference Standard); m/z 238→77 (CE, −24 eV) for l-[2H7]-arginine (98 atom% 2H isotopic purity; 2,3,3,4,4,5,5-l-[2H7]-arginine; Euriso-top); m/z 259.3→214 (CE, −16 eV) for ADMA (>99% purity; Sigma-Aldrich); m/z 259.3→228 (CE, −14 eV) for SDMA (>99% purity; Sigma-Aldrich); and m/z 265.3→220 (CE, −16 eV) for [2H6]-ADMA (98 atom% 2H isotopic purity; 3,3,4,4,5,5-[2H6]-ADMA). The synthesis and characterization of [2H6]-ADMA were performed as described previously (12). All compounds were analyzed as their butyl ester derivatives (see below). All parent ions (MS) represented the molecular protonated cations, i.e., [M+H]+. The major daughter ions formed, i.e., m/z 70 and m/z 77, for l-arginine and l-[2H7]-arginine, respectively, correspond to a pyrrolinium ion as a cyclization product (14). Major daughter ions of m/z 259.3 were m/z 214 [M+H − NHCH3CH3]+ and m/z 228 [M+H − NH2CH3]+ for authentic ADMA and SDMA, respectively. The m/z 228 and m/z 214 ions were absent in the product ion mass spectra of ADMA and SDMA, respectively. We checked ion suppression by comparing peak areas of aqueous calibrator solutions without matrix and calibrator added to plasma matrix (15). Ion suppression was <10% for 5 different specimens. We used 3 different calibrators, l-arginine, ADMA, and SDMA, at 6 different concentrations (n = 5 each): 0, 25, 50, 100, 250, and 500 μmol/L for l-arginine and 0, 0.25, 0.5, 1, 2, and 4 μmol/L for ADMA and SDMA, respectively. Aqueous stock solutions of l-arginine, ADMA, and SDMA were made by weighing authentic material supplied by the manufacturer. The calibrators were treated exactly the same as patient plasma samples. l-[2H7]-Arginine and [2H6]-ADMA were added as internal standards at concentrations of 50 and 2 μmol/L, respectively. [2H6]-ADMA was used as internal standard for ADMA and SDMA. Quantitative determinations in biological samples (50 μL) were performed by dilution of 5 μL of aqueous internal standards (500 μmol/L l-[2H7]-arginine and 20 μmol/L [2H6]-ADMA) corresponding to concentrations of 50 and 2 μmol/L, respectively, in 50 μL of biological sample. Subsequently, proteins were precipitated with 100 μL of acetone, and dried supernatants were derivatized. Compounds were derivatized with 100 μL of 1 mol/L HCl in 1-butanol for 17 min at 65 °C. After evaporation, samples were reconstituted in 1 mL of water; 20 μL of the water extract was injected on the column. The mean (SD) retention times of the butyl ester derivatives of l-arginine, ADMA, and SDMA, respectively, were 1.03 (0.04) min (CV = 3.9%; n = 6), 1.84 (0.02) min (CV = 1.3%), and 2.03 (0.02) min (CV = 0.8%). Differences in retention times for unlabeled and labeled compounds were not statistically significant. The linear regression equations for peak-area ratio (LC-MS/MS; y) and ratio of injected calibrators (x) were as follows: y = 0.94x − 0.001 (r2 = 0.999) for l-arginine; y = 0.98x + 0.02 (r2 = 0.999) for ADMA; and y = 2.05x + 0.01 (r2 = 0.999) for SDMA. Ideally, regression analysis of peak-area ratios should give a slope of 1; however, the intensities of the daughter ions obtained for labeled and unlabeled compounds were not the same. Thus, only under MS conditions was a slope of 1 obtained: The linear regression equations for peak-area ratio (LC-MS; y) and ratio of injected calibrators (x) were as follows: y = 1.003x − 0.02 (r2 = 0.999) for l-arginine; y = 1.001x + 0.01 (r2 = 0.999) for ADMA; and y = 1.002x + 0.03 (r2 = 0.998) for SDMA. Nevertheless, calibration curves were calculated to account for these differences, and the slopes of the calibration curves were included in all plasma concentration calculations. The lower limits of detection, defined as a signal-to-noise ratio of 3, were 45 nmol/L for l-arginine, 3 nmol/L for ADMA, and 2 nmol/L for SDMA. We validated our LC-MS/MS method by adding different concentrations of l-arginine, ADMA, and SDMA in quintuplicate to samples. l-Arginine was added at 0, 0.5, 1, 25, 50, 100, and 250 μmol/L, and ADMA and SDMA were added at 0, 0.05, 0.1, 0.5, 1, 2, and 4 μmol/L. Linear regression analysis between measured (y) and added (x) concentrations yielded the following slopes and y-intercepts: 1.01 and 70.3 μmol/L (r2 = 0.999) for l-arginine; 1.00 and 0.45 μmol/L (r2 = 0.999) for ADMA; and 1.01 and 0.42 μmol/L (r2 = 0.999) for SDMA. Data from these validation experiments are listed in Table 11 . The observed recoveries of the added analytes were 97.7%–111% for l-arginine, 96.9%–105% for ADMA, and 90.8%–109% for SDMA. All CVs were <6% for the imprecision and <20% for recovery except for the addition of 0.5 μmol/L l-arginine. Thus, the analytical range of the method was 1–250 μmol/L for l-arginine and 50 nmol/L to 4 μmol/L for ADMA and SDMA, with 1 μmol/L and 50 nmol/L representing the lower limits of quantification for l-arginine and ADMA and SDMA, respectively. The addition of l-arginine, ADMA, or SDMA did not influence the measurement of the other analytes. We also determined within- and between-run imprecision (CVs; n = 10). The within-run CVs were 3.3% for l-arginine, 2.6% for ADMA, and 2.5% for SDMA. Between-run CVs were 4.7% for l-arginine, 4.1% for ADMA, and 3.9% for SDMA We also measured ADMA in plasma samples from healthy human volunteers (n = 22) by this LC-MS/MS method. Shown in Fig. 11 is a typical chromatogram of a human plasma sample. Samples were aliquoted (50 μL) and frozen at −20 °C before analysis. The mean (SD) concentrations of l-arginine, ADMA, and SDMA were 65.6 (23.4), 0.55 (0.14), and 0.69 (0.23) μmol/L, respectively. The mean (SD) molar ratio of l-arginine to ADMA was 132 (55). Our LC-MS/MS–based method allows for the simultaneous determination of ADMA, SDMA, and l-arginine and has been validated for the measurement of these 3 analytes in human plasma. Values obtained for l-arginine, ADMA, and SDMA in human plasma are within previously reported ranges (11)(12)(13)(16). Major advantages over previous reported methods for the measurement of ADMA include a minimized sample volume and reduction in laboratory work. In particular, in comparison with other LC-MS–based methods (13)(14)(16)(17), sample run time has been reduced to 4 min including reequilibration. Chromatographic separation required only slightly more than 2 min. This considerable reduction in analysis time was achieved by use of the specific fragmentation patterns of the ester derivatives of ADMA and SDMA. Identical MS spectra were obtained from the ester derivatives of ADMA and SDMA, but the major fragments in the daughter ion spectra for ADMA and SDMA were at m/z 214 and m/z 228, respectively. These ions were specific for ADMA and SDMA: they were absent in the corresponding spectra of SDMA and ADMA, respectively. In contrast to previous reports, the proposed method allows, for the first time, spectrometric instead of chromatographic separation of these methylated arginines (see the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol51/issue7/). Precision of the LC-MS/MS method for the measurement of l-arginine, ADMA, and SDMA in human plasma. NA, not applicable. Precision of the LC-MS/MS method for the measurement of l-arginine, ADMA, and SDMA in human plasma. NA, not applicable. Representative LC-MS/MS chromatograms for l-arginine, l-[2H7]-arginine (internal standard), SDMA, ADMA, and [2H6]-ADMA (internal standard). Chromatograms from the analysis of a human plasma sample (50 μL) to which 50 μmol/L l-[2H7]-arginine and 2 μmol/L [2H6]-ADMA had been added as internal standards. Trace 1 (top), l-arginine; trace 2, l-[2H7]-arginine; trace 3, SDMA; trace 4, ADMA; trace 5 (bottom), [2H6]-ADMA. All of the above analytes could be separated based on their specific daughter ions in LC-MS/MS analysis: m/z 70 for l-arginine, m/z 77 for l-[2H7]-arginine, m/z 228 for SDMA, m/z 214 for ADMA, and m/z 220 for [2H6]-ADMA. Scan rate was 1 s. (+) ESI SRM, electrospray ionization single reaction monitoring.