|
|
||||||||
Automation and Analytical Techniques |
Institute of Clinical Pharmacology, Otto-von-Guericke-University Magdeburg, University Hospital, Magdeburg, Germany.
aAddress correspondence to this author at: Otto-von-Guericke-Universität, Institut für Klinische Pharmakologie, Leipziger Strasse 44, D-39120 Magdeburg, Germany. Fax 49-391-67-13062; e-mail jens.martens-lobenhoffer{at}medizin.uni-magdeburg.de.
| Abstract |
|---|
|
|
|---|
Methods: We quantified Arg, ADMA, and SDMA in human plasma, human urine, and cell culture supernatant by HPLCelectrospray tandem mass spectrometry. Sample preparation required only protein precipitation. Separation was by liquid chromatography on a 150 x 3 mm silica column with an isocratic mobile phase consisting of wateracetonitriletrifluoroacetic acidpropionic acid (10:90:0.025:1 by volume). The chromatographic run time was 7 min.
Results: The chromatograms were interference-free in all matrices. In the low-concentration quality-control samples, the interassay CVs in plasma were 4.7% for Arg, 7.7% for ADMA, and 4.9% for SDMA. Similar values were obtained in urine and cell culture supernatants. The calibration functions were linear and covered the ranges of healthy and pathologic samples.
Conclusion: The new method requires neither derivatization nor complete chromatographic separation between ADMA and SDMA for quantification of the 3 metabolites, has calibration functions that are independent of the sample matrix, and provides measured concentrations that agree with those reported in the literature.
| Introduction |
|---|
|
|
|---|
In another metabolic pathway, Arg serves as a substrate for the enzyme endothelial nitric oxide synthase (eNOS). It is converted by this enzyme in endothelial cells to NO and citrulline. The production rate of NO by eNOS is a key factor for endothelial and cardiovascular functions (4). ADMA, on the other hand, competitively inhibits eNOS (5). Consequently, increased ADMA concentrations, particularly with an increased ADMA/Arg ratio, lead to deficiency in endothelial NO production and a subsequent decrease of endothelial function (6). This situation occurs in many diseases, including renal insufficiency, diabetes mellitus, essential hypertension, and hypercholesterolemia (7)(8)(9)(10). SDMA is inactive with respect to eNOS, but it shares the pathway for cell entry with Arg and ADMA and may therefore indirectly influence the NO production rate (11). In this context, a fast, easy, and precise assay for Arg, ADMA, and SDMA in biological fluids is highly desirable.
A broad range of methods for the determination of Arg, ADMA, and SDMA have been published in recent years. Many of these methods rely on derivatization of these compounds with o-phthalaldehyde, separation by reversed-phase HPLC, and fluorescence detection (12)(13)(14)(15). This approach requires laborious sample preparation and long HPLC separation times to remove endogenous interferences and to achieve separation of the structurally similar compounds ADMA and SDMA. The need for complicated sample clean-up can be avoided with the combination of HPLC separation with mass spectrometric detection. Nevertheless, to achieve separation of the very polar analytes on reversed-phase columns, a derivatization step must still be included, limiting the throughput of these methods (16)(17). Methods in which the analytes were separated in their underivatized forms have also been reported (18)(19). In both types of methods, however, ADMA and SDMA must be separated completely by chromatographic means, and neither method type uses an isotopic labeled analog as an internal standard (IS). They therefore are prone to incorrect quantification results caused by uncontrolled ion suppression in unknown samples.
Given these shortcomings of existing techniques, we sought to develop a new high-throughput method for measuring Arg, ADMA, and SDMA in biological fluids such as plasma, urine, or cell culture supernatants.
| Materials and Methods |
|---|
|
|
|---|
chemicals
Arg, ADMA (as hydrochloride salt), and SDMA [as di(p-hydroxyazobenzene-p'-sulfonate salt] were purchased from Sigma. The IS 13C6-arginine (as its hydrochloride salt) was obtained from Cambridge Isotope Laboratory. The second IS, 2D6-ADMA, was synthesized in our laboratory as described previously (16). All other chemicals were of analytical grade or better.
sample collection
Blood samples of
5 mL were drawn into sampling tubes (BD Vacutainer Systems) containing EDTA as anticoagulant. Blood cells were separated by centrifugation at 2400g for 10 min, and the resulting plasma was stored at 80 °C until analysis. Urine samples were stored in plastic containers without additives at 80 °C until analysis.
The healthy volunteers enrolled in this study [group 1 (n = 14); age range, 2232 years; 8 female, 6 male; group 2 (n = 7); age range, 2349 years; all male; weight range, 78100 kg] gave informed consent, and the study was in accordance with the current revision of the Helsinki Declaration.
Cell cultures were incubated in Clonetics EBM-2 basal cell medium (Cambrex Bio Science), after which cells were separated by centrifugation at 2400g for 10 min. Cell culture supernatants were stored in plastic containers at 80 °C until analysis.
calibration samples
For calibration in plasma and cell culture supernatants, we prepared a stock solution containing 150 µmol of Arg, 3 µmol of ADMA, and 4 µmol of SDMA in 100 mL of water. Water, plasma, and cell culture medium samples were enriched with this solution to yield calibration samples with added concentration ranges of 0150 µmol/L Arg, 03 µmol/L ADMA, and 04 µmol/L SDMA.
For urine calibration, we prepared a stock solution containing 50 µmol of Arg, 100 µmol of ADMA, and 100 µmol of SDMA in 100 mL of water. Water and urine samples were enriched with this solution to yield calibration samples with added concentration ranges of 0 50 µmol/L Arg, 0100 µmol ADMA, and 0100 µmol/L SDMA, respectively.
quality-control samples
We prepared quality-control samples from plasma and urine from healthy volunteers and from unincubated cell culture medium. The lowest concentrations of the quality-control samples were similar to the unenriched concentrations of Arg, ADMA, and SDMA in these matrices. We prepared 2 more concentrations by enriching the matrices with calibration solutions to yield samples in the concentration range of the medium and the highest calibration concentrations. The quality-control samples were portioned and stored at 80 °C until use.
sample preparation
The sample preparation procedure was almost identical for plasma, urine, and cell culture supernatants. To 100 µL of a plasma or cell culture supernatant sample, we added 20 µL of the IS solution (460 µmol/L 13C6-Arg and 45 µmol/L 2D6-ADMA in water). The respective sample volumes for urine were 50 µL plus 50 µL of IS solution. The samples were then mixed with 900 µL of the HPLC mobile phase B, which consisted of 1 L of acetonitrile mixed with 0.25 mL of trifluoroacetic acid (TFA), and 10 mL of propionic acid. Precipitated proteins were separated by centrifugation at 10 000g for 5 min.
chromatographic conditions and mass spectrometer settings
The injection volumes of the prepared samples into the HPLC system were 10 µL for plasma extracts and 5 µL for urine and cell culture supernatant extracts. The same chromatographic conditions were used for plasma, urine, and cell culture supernatants. Mobile phase A consisted of 1 L of water mixed with 0.25 mL of TFA and 10 mL of propionic acid; mobile phase B was as described above. Separation of the analytes was accomplished by isocratic elution with 10% mobile phase A and 90% mobile phase B at a flow rate of 0.5 mL/min and a column temperature of 30 °C. Under these conditions, the retention times were 4.3 min for Arg and its IS (13C6-Arg), 4.7 min for SDMA, and 4.9 min for ADMA and its IS (2D6-ADMA).
The effluent of the HPLC column was directed to the ESI source of the mass spectrometer for the 36.5 min portion of the run time; otherwise, it was directed to the waste container. The settings of the ESI source were as follows: sheath gas, 32 (arbitrary units); auxiliary gas, 20 (arbitrary units); needle voltage, +4.5 kV; and capillary temperature, 300 °C. Under these conditions, the ESI source produced single charged adduct ions in the form [M+H]+. No dimers or other adduct ions of the analytes were observed.
The [M+H]+ ions were analyzed in the multiple-reaction monitoring mode of the mass spectrometer. Fragmentation took place at a collision gas pressure of 1.5 mTorr, with argon used as collision gas. The observed multiple-reaction monitoring transitions were m/z 175.2
70.1 for Arg, m/z 181.2
74.1 for its IS (13C6-Arg), m/z 203.2
172.1 for SDMA, m/z 203.2
46.1 for ADMA, and m/z 209.2
70.1 for its IS (2D6-ADMA).
| Results and Discussion |
|---|
|
|
|---|
chromatography and detection
A typical chromatogram obtained from human plasma is depicted in Fig. 1
(chromatograms from urine and cell culture supernatant can be seen in Figs. 1 and 2 in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol52/issue3/). The peaks in the chromatogram were sharp and symmetric, and no interferences from endogenous substances were apparent. A whole chromatographic cycle took only 8 min, which makes it possible to analyze
180 samples per day. Use of hydrophilic-interaction liquid chromatography led to good and stable retention of the polar substances. Furthermore, the mobile phase used in the hydrophilic-interaction liquid chromatography was very favorable for an ESI source, giving strong and stable signals (20). To achieve these chromatographic attributes, some TFA had to be added to the mobile phase. However, the ability of TFA to form gas-phase ion pairs with basic analytes is known to suppress sensitivity in ESI sources. To overcome this problem, we added an excess of weak acids such as propionic acid to the mobile phase, a method reported by Shou and Naidong (21). With the application of this "TFA-fix", the assay was sensitive enough for use with all kinds of samples in the 3 investigated matrices.
|
ESI sources are known to be susceptible to matrix effects caused by coeluting substances from biological samples. We used the experimental procedure described by Souverain et al. (22) to investigate the matrix effects introduced by the injection of samples that had undergone the described sample preparation procedure. In short, a solution of the IS was infused continuously via a tee-union into the column effluent. While constantly monitoring the ion traces of the IS, we injected samples prepared without addition of the IS. Any ion suppression would be indicated by a decrease of the otherwise constant detector signal (see Fig. 3 in the online Data Supplement). Strong ion suppression occurred in all 3 matrices, but the effects ceased early in the chromatographic run and did not compromise quantification of the analytes. Furthermore, because we used isotopically labeled IS for Arg and ADMA, quantification of these substances was not susceptible to matrix effects. The retention time of SDMA was very close to that of ADMA; therefore, 2D6-ADMA could also serve as a reasonable IS for this compound.
Tandem mass spectrometric (MS/MS) detection of the analytes led to signals that showed no interferences from endogenous substances. The isotopically labeled IS were 6 unified atomic mass units heavier than their unlabeled counterparts, and we observed no cross-talk between the signals. Total chromatographic separation of the pair ADMA and SDMA was not required because the MS/MS spectra gave unique signals for each. The fragment ion m/z 46.1 was related to the loss of a dimethylammonium ion, a moiety that was found only in ADMA and not in SDMA, whereas the fragment ion m/z 172.1 was related to the neutral loss of methylamine, a moiety that was present only in SDMA and not in ADMA. In observing these fragment ions, we found no cross-talk between ADMA and SDMA and vice versa. The signals of these fragment ions were not the most intense ones observed in the MS/MS spectra, but their selectivity with regard to their parent compounds rendered them advantageous. In the case of the IS D6-ADMA, no such mass spectrometric separation was necessary; therefore, the most intense transition m/z 209.2
70.1 was observed.
validation of the assay
The limits of detection, defined as a peak-to-noise ratio >3, were 0.4 µmol/L for Arg, 0.02 µmol/L for ADMA, and 0.01 µmol/L for SDMA. The limits of quantification, defined as the lower limits of the calibration ranges, were 7.5 µmol/L for Arg, 0.15 µmol/L for ADMA, and 0.2 µmol/L for SDMA in plasma; 3.75 µmol/L for Arg, 0.075 µmol/L for ADMA, and 0.1 µmol/L for SDMA in cell culture supernatants; and 2.5 µmol/L for Arg and 5 µmol/L for ADMA and SDMA in urine.
The calibration function parameters for all substances in all matrices are summarized in Table 1
. All calibration functions were linear, and the correlation coefficients were always >0.99. For all matrices, analogous calibration from water samples was prepared for comparison. In all instances, except SDMA in cell culture medium, the calibration slopes were nearly identical in authentic matrix and in water samples. This result reflects the benefits of the use of isotopically labeled analogs as IS, making quantification independent of the matrix. We observed a matrix effect on the calibration in cell culture medium only in the case of SDMA, for which no isotopically labeled analog was available. For this substance, the calibration function obtained from the authentic matrix had to be used instead the one obtained from water. The calibration function intercepts were not significantly different from zero in calibrations in water. In contrast, the corresponding calibration functions in biological matrices showed intercepts representing the basal concentrations of the analytes already present in the matrix before enrichment.
|
For inter- and intraassay precision and accuracy, the CVs and the differences from the expected values (see Tables 1 and 2 in the online Data Supplement) were always <8%, except for the basal reading of SDMA in cell culture medium. This slightly worse performance of the SDMA measurements compared with Arg and ADMA again highlights the benefits of using isotopically labeled analogs as IS. Quantification of Arg and ADMA was very stable and rugged because of such IS, whereas for SDMA such an isotopically labeled analog is not available at this time.
application of the method
We successfully applied the method on several hundred plasma, urine, and cell culture samples. In samples from a group of healthy volunteers (group 1), the mean (SD) measured values in plasma were 60.6 (18.3) µmol/L for Arg, 0.370 (0.061) µmol/L for ADMA, and 0.449 (0.055) µmol/L for SDMA. These results correspond very well with recently published reports (12)(13)(14)(17)(18), in which mean values obtained from different methods were 65.696.1 µmol/L for Arg, 0.40.58 µmol/L for ADMA, and 0.470.69 µmol/L for SDMA. A more comprehensive discussion concerning the reference values of Arg, ADMA, and SDMA in human plasma was provided by Martens-Lobenhoffer and Bode-Böger (23). In urine samples from another group of healthy volunteers (group 2), the measured concentrations were 1.67 (0.62) µmol/mmol creatinine for Arg, 2.92 (0.42) µmol/mmol creatinine for ADMA, and 3.72 (0.35) µmol/mmol creatinine for SDMA. Again, these values are comparable to the results obtained with a different method (24), for which mean concentrations of 3.94 µmol/mmol creatinine for ADMA and 3.70 µmol/mmol creatinine for SDMA were reported (calculated from the data presented in the original manuscript). In human umbilical vein endothelial cell culture supernatants after incubation for 2 days, we found typical values of 0.397 (0.017) µmol/L for ADMA and 0.119 (0.006) µmol/L for SDMA.
In conclusion, our new method for the determination of Arg, ADMA, and SDMA in biological fluids is easy, fast, and rugged. Furthermore, it is less expensive than conventional assays using HPLC with fluorometric detection because of its minimal use of consumables and the short analysis time, which compensates for the expensive instrumentation. A newly developed ELISA measures only ADMA, is expensive, and shows poor correlation with results obtained by chromatographic methods (25); it therefore does not seem to be a viable alternative. In our method, hydrophilic-interaction liquid chromatography makes it possible for the first time to achieve substantial and stable retention for Arg, ADMA, and SDMA in their underivatized form by use of an HPLC mobile phase that is favorable for the ion-forming process in an ESI source. This feature makes possible intense and stable MS signals. In combination with the easy sample preparation without solid-phase extraction or derivatization, only minor adjustments are needed to apply this method to different matrices, such as human plasma, urine, or cell culture supernatants. The use of isotopically labeled analogs as IS for Arg and ADMA enables precise quantification. In the case of SDMA, an isotopically labeled IS is not available at this time, but the retention time, which is very close to that of ADMA, makes it possible to use the IS 2D6-ADMA for SDMA as well. The MS/MS detection provides interference-free chromatograms for all analytes in all matrices. A unique feature of our method is that neither derivatization nor complete chromatographic separation between ADMA and SDMA is required. As such, the proposed method could be a useful addition to the repertoire of techniques for the analysis of methylated arginines and Arg.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |