An automated analyzer for methylated arginines in rat plasma by high-performance liquid chromatography with post-column fluorescence reaction

Abstract

A fully automated analyzer for methylated L-arginine metabolites [N,N-dimethyl-L-arginine (ADMA), N-methylarginine (NMMA) and N,N′-dimethyl-L-arginine (SDMA)] by high-performance liquid chromatography with post-column fluorescence derivatization was developed. This system consists of an on-line extraction, a separation on a reversed phase ion-pair chromatograph, a post-column derivatization by o-phthaladehyde (OPA) and thiol reaction, and fluorescence detection. NMMA, ADMA and SDMA were separated in 40 min with isocratic elution by a combination of octanoate and cyclohexane carboxylate as ion-pair reagents. The eluate was monitored at 450 nm with excitation at 337 nm. The calibration curves for NMMA, ADMA and SDMA showed linearity over the range from 0.05 μmol l⁻¹ (0.5 pmol on column) to 5.0 μmol l⁻¹ (50 pmol on column). This method does not require any time-consuming pre-treatment and requires only 10 μl of plasma sample for assay.

---

Introduction

Nitric oxide (NO), synthesized from L-arginine by NO synthase (NOS, EC 1.14.13.39), plays an important role in the regulation of cardiovascular tones and neurotransmission. In the cardiovascular system, NO synthesized in endothelial cells regulates blood flow and blood pressure. Recently, it was reported that a methylated L-arginine metabolite (Fig. 1), N,N-dimethyl-L-arginine (asymmetric DMA, ADMA) and N-methylarginine (NMMA), NOS inhibitors, decrease NO production and can cause hypertension.¹ ADMA and biologically inactive N,N′-dimethyl-L-arginine (symmetric DMA, SDMA) come to exist in several tissue proteins by post-translational methylation, and are released in body fluids after protein degradation. The alteration of ADMA concentration has recently been reported in plasma from patients with renal failure,² hypertension,³ microangiopathy,⁴ atherosclerosis,⁵ schizophrenia,⁶ and hypercholesteromia.⁷ Therefore, to investigate the relationship between methylated arginines and these diseases, a determination method for methylated arginines in biological samples is essential.

Fig. 1 Chemical structures of arginine, NMMA, ADMA and SDMA.

There have been several reports for determination of methylated arginines in biological samples using high-performance liquid chromatography (HPLC) and capillary electrophoresis (CE). However, these published methods have their drawbacks. The direct UV detection method of these compounds after separation by HPLC or CE gave low sensitivity.^8–10 To increase the sensitivity, the pre-column derivatization using o-phthalaldehyde (OPA) and thiol followed by the separation on a reversed phase column and fluorescence detection has been widely adopted.^11–16 Recently, fluorescent derivatization with fluorescein isothiocyanate (FITC) and CE-LIF (laser-induced fluorescence detection)¹⁷ was also reported. These methods, however, require tedious and complicated pre-treatment procedures for extraction and derivatization of methylated arginines in biological samples. Also, although the post-column derivatization method with OPA and thiol after separation on ion-exchange chromatography made it possible to avoid such tedious pre-treatment procedures, it requires complicated gradient profiles and a long separation time (>80 min).^18,19 Nevertheless, the methylated arginines were still not satisfactorily separated.

Thus, a simple, rapid, and sensitive determination method of methylated arginines needs to be developed. In this paper, we report an automated analyzer for methylated arginines. It includes an on-line extraction of methylated arginines by an ion-exchange pre-treatment column followed by the separation on a reversed phase ion-pair chromatography, post-column derivatization with OPA and thiol, and the fluorescence detection.

Experimental

Reagents

Arginines (arginine, Arg; NMMA; ADMA; SDMA) and other amino acids were all purchased from Sigma (St. Louis, MO, USA). Methanol of HPLC grade, sodium dihydrogen phosphate dihydrate (NaH₂PO₄·2H₂O), disodium hydrogen phosphate (Na₂HPO₄), sodium butane sulfonate, sodium hexane sulfonate and sodium octanoate were purchased from Wako Pure Chemicals (Tokyo, Japan). Boric acid, cyclohexanecarboxylic acid and o-phthalaldehyde (OPA) were purchased from Nacalai Tesque (Kyoto, Japan). Sodium hydroxide was purchased from Kanto Chemical Co. Inc. (Tokyo, Japan). 3-Mercaptopropionic acid was purchased from Aldrich Japan (Tokyo, Japan). Water was used after purification by a Milli-Q reagent system (Nihon Millipore, Tokyo, Japan). All other reagents were of analytical grade.

Instrumentation

The block diagram of the HPLC system is shown in Fig. 2. It consists of an RT730 system controller (GL Sciences, Tokyo, Japan), a model 33 autosampler, a plate cooler (System Instruments, Tokyo, Japan), three intelligent pumps (model 301) (Labo System, Tokyo, Japan), two valve units (model 401) (Labo System), a model 505 column oven (Labo System), Gastorr 154 (Labo System), an L-7480 fluorescence detector (Hitachi, Tokyo, Japan) and a D-7500 integrator (Hitachi). A pre-treatment column, CAPCELL PAK MF-SCX (10 × 4.0 mm, id, 5 μm, Shiseido, Tokyo, Japan), a pre-column, CAPCELL PAK MG-C18 (35 × 4.6 mm id, Shiseido) and an analytical column, CAPCELL PAK MG-C18 (250 × 4.6 mm id, 5 μm, Shiseido) were used. A reaction coil (5.0 m × 0.25 mm) was used for the post-column derivatization reaction.

Fig. 2 Block diagram of automated analyzer for methylated arginines: (solvent reservoir A) pretreatment buffer (1.0 ml min⁻¹) and eluent (1.0 ml min⁻¹), 70 mmol l⁻¹ sodium phosphate buffer (pH 6.67) containing 15 mmol l⁻¹ cyclohexane carboxylate and 1.5 mmol l⁻¹ octanoate; (solvent reservoir B) OPA solution (0.3 ml min⁻¹), 3.0 mmol l⁻¹ OPA in 0.2 mol l⁻¹ borate buffer (pH 9.8); (solvent reservoir C) thiol solution (0.3 ml min⁻¹), 3.7 mmol l⁻¹ 3-mercaptopropionic acid in 0.2 M borate buffer (pH 9.8). FC, CAPCELL PAK MF-SCX (10 mm × 40 mm id) × 2; PC, CAPCELL PAK MG-C18 (35 mm × 4.6 mm id); AC, CAPCELL PAK MG-C18 (250 mm × 4.6 mm id); column oven, 40 °C; C1, reaction coil, 0.25 mm id × 5 m; C2, heating coil, 0.13 mm id × 1 m; C3, heating coil, 1.0 mm × 1 m; C4, mixing coil, 1.0 mm id × 1 m.

Chromatographic conditions

The mobile phase was 70 mmol l⁻¹ sodium phosphate buffer (pH 6.67 ± 0.05) containing 15 mmol l⁻¹ cyclohexane carboxylate and 1.5 mmol l⁻¹ octanoate. The flow rate was 1.0 ml min⁻¹. The derivatization reagent solutions were 3.7 mmol l⁻¹ 3-mercaptopropionic acid in 0.2 mol l⁻¹ borate buffer (pH 9.8) and 3.0 mmol l⁻¹ OPA in 0.2 M borate buffer (pH 9.8) containing 1% methanol, respectively. The flow rates of these solutions were 0.3 ml min⁻¹. The eluate was monitored at 450 nm with excitation at 337 nm. The column oven and reaction coil temperature was maintained at 40 °C. The autosampler plate temperature was kept below 10 °C.

Mobile phase preparation

Commercially available cyclohexane carboxylic acid contains some impurities and must be purified before use. A solution of sodium hydroxide (1.0 mol l⁻¹, 80 ml) was added to sodium octanoate (4.5 mmol), cyclohexane carboxylic acid (45 mmol) and water (20 ml). Hexane (500 ml) was passed into the turbid solution through the needle by the HPLC pump (flow rate, 4.0 ml min⁻¹) and the solution turned clear gradually. The solution was added to 2.9 L of water containing Na₂HPO₄ (9.82 mmol), NaH₂PO₄·2H₂O (102.9 mmol) and NaN₃ (3.0 mmol).

Preparation of plasma samples

Male Sprague-Dawley (SD) rats (8 weeks old, 340–370 g) purchased from Charles River Japan Inc. (Kanagawa, Japan) were anaesthetized with diethyl ether. Blood was taken from the abdominal aorta, transferred to a heparinized polyethylene tube, and centrifuged at 3000g for 20 min at 4 °C. The plasma fraction was collected and stored at −70 °C until analyzed. The plasma samples (10 μl) were injected into the HPLC after the addition of four times the volume of mobile phase solution.

Calibration curve

A 20 μl aliquot of sample was diluted five times with mobile phase solution and a 50 μl aliquot of the solution was injected into the HPLC by autosampler. The calibration curve for peak heights versus concentrations was obtained using the solutions containing methylated arginines (NMMA, ADMA, SDMA) of 0.05, 0.1, 0.2, 0.4, 1.0, 2.0 and 5 μmol l⁻¹, respectively. Least-squares regression was used to calculate the slope and correlation coefficient.

Precision and accuracy

For the intra-day precision and accuracy study, to 20 μl of rat plasma, 20 μl of mobile phase solution containing 0.4, 1.0 and 2.0 μmol l⁻¹ of NMMA, ADMA and SDMA, respectively, was added. After an addition of 60 μL of mobile phase solution, a 50 μl aliquot of the obtained mixture was injected into the HPLC. The inter-day precision of this system was evaluated by analyzing the same rat plasma sample five consecutive times for five successive days.

Recovery

To determine the recovery of methylated arginines by on-line extraction with a cation-exchange column, the standard solutions of 0.05, 0.1, 0.4, 1.0 and 2.0 μmol l⁻¹ of methylated arginines was injected into the HPLC without the cation-exchange column. The slopes of the calibration curves for concentration versus peak area were compared with the calibration curves obtained with the cation-exchange column.

Results and discussion

On-line extraction of methylated arginines from plasma samples

At first, the extraction of basic amino acids from plasma samples using the cation-exchange column was examined. NMMA, ADMA and SDMA, bearing the guanidino group, are basic amino acids. To extract basic amino acids selectively, the pH region of the mobile phase, where only basic amino acids have a plus charge, was selected. Fig. 3 shows the ionization states and formal charge of the amino acids. The pK_a values for arginine are 2.18 (pK_a1), 9.09 (pK_a2) and 13.20 (pK_a3) while those values of valine, a neutral amino acid, are 2.29 (pK_a1) and 9.72 (pK_a2). Those values of aspartic acid, an acidic amino acid, are 1.99 (pK_a1), 3.96 (pK_a2) and 10.00 (pK_a3), respectively. Therefore, the neutral pH region seems to be suitable for the extraction of basic amino acids with a cation-exchange column, since the total charge of basic amino acids is +1, and those of neutral and acidic amino acids are 0 and −1, respectively. Fig. 4(a) shows the typical chromatogram of amino acids under these conditions. The guadininoamino acids (Arg, NMMA, ADMA and SDMA) were retained by the cation-exchange column. In comparison, acidic and neutral amino acids were not retained under these conditions. The retention times of Asp, Val, Leu, Phe and Tyr were 0.39, 0.46, 0.46, 0.40, and 0.54 min, respectively. His (0.62 min), Trp (0.65 min) and Lys (1.05 min) were also clearly separated from guanidinoamino acids. Thus, methylated arginines were clearly separated from other amino acids.

Fig. 3 Relationship between ionization and formal charge of amino acids.

Fig. 4 Chromatogram of a standard mixture of Arg, NMMA, ADMA and SDMA (30 nmol each on column) obtained with cation-exchange column (a) and cation-exchange column and C18 pre-column (b). The eluate was monitored by absorbance at 210 nm.

HPLC System

At first, the fraction of methylated arginines was directly injected into the separation column by valve switching. However, the base line was gradually elevated and considerably drifted. It was assumed that the impurities eluted from the cation-exchange column were responsible for the unstableness of the base line. Although we have tried several types of cation-exchange columns, no stable base line was obtained. Therefore, a C18 pre-column and a switching valve were used to avoid the inflow of impurities from the cation exchange column to the analytical column. Fig. 4(b) shows the chromatogram of guanidinoamino acids obtained on a C18 pre-column connected in series with the cation-exchange column. Methylated arginines were eluted in 3.5–6.5 min, thus, the fraction was injected into the analytical column by valve switching. In Fig. 2, valve 1 and valve 2 control the position of the cation exchange column and the C18 pre-treatment column, respectively. The valve switching protocol consists of the following 5 steps: (1) fractionation by the ion-exchange column, (2) injection of fractionated components into the C18 pre-column, (3) fractionation by the C18 pre-column, (4) injection of the required fraction into the analytical column and ion-exchange column washing, and finally, (5) separation on the analytical column and C18 pre-column washing (Fig. 5).

Fig. 5 Column switching protocol for the HPLC system.

Separation of methylated arginines

For the separation of guanidinoamino acids from the other amino acids by reversed phase chromatography using an ODS column, ion-pair chromatography was adopted, utilizing common ion-pair reagents such as straight-chain alkylsulfonates and octanoate. The standard NMMA, ADMA and SDMA were well separated. However, ADMA and SDMA were not clearly separated from the several interfering peaks on the chromatograms obtained from rat plasma sample using butane sulfonate as an ion-pair reagent [Fig. 6(a)], hexane sulfonate [Fig. 6(b)] and octanoate [Fig. 6(c)], respectively. Next, cyclohexane carboxylate, a cyclic compound, was tried as use as ion-pair reagent, since it was expected to give the unknown interfering compounds different chromatographic behavior as compared with the straight-chain ion-pair reagents. As expected, the interfering peaks were eluted differently, but still observed between ADMA and SDMA [Fig. 6(d)]. Then, the combined use of two ion-pair reagents was examined. Since with octanoate an interfering peak was eluted later than SDMA, the last analyte, octanoate was selected as a partner for cyclohexane carboxylate. After examination of the concentrations of the ion-pair reagents, finally the combined use of 15 mmol l⁻¹ cyclohexane carboxylate and 1.5 mmol l⁻¹ octanoate gave a good separation of methylated arginines from the interfering peaks within 40 min using isocratic elution as shown in Fig. 7. The post-column derivatization methods previously reported for methylated arginines require complicated gradient profile and long separation time of more than 80 min.^18,19 It should be noted that, as far as we know, cyclohexanecarboxylic acid has not been used as an ion-pair reagent and is not available in the market in a pure form, thus it must be entirely purified before use as described in the experimental section.

Fig. 6 Chromatograms obtained from a standard solution and a rat plasma sample. Mobile phase conditions were as follows; (a) 70 mmol l⁻¹ sodium phosphate buffer (pH 6.76) containing 25 mmol l⁻¹ butane sulfonate; (b) 70 mmol l⁻¹ sodium phosphate buffer (pH 6.76) containing 3.0 mmol l⁻¹ hexane sulfonate; (c) 70 mmol l⁻¹ sodium phosphate buffer (pH 6.76) containing 1.5 mmol l⁻¹ octanoate; (d) 70 mmol l⁻¹ sodium phosphate buffer (pH 6.97) containing 10 mmol l⁻¹ cyclohexane carboxylate. Other conditions are described in the text.

Fig. 7 Chromatogram obtained from a SD rat plasma sample and standard mixture of NMMA, ADMA and SDMA (0.8 μmol l⁻¹) by the present method. HPLC conditions are described in the text and Fig. 2.

Derivatization conditions

2-Mercaptoethanol is a most widely used thiol for OPA post column reactions. However, we used 3-mercaptopropionic acid to avoid the strong unpleasant smell.²⁰ The post column derivatization conditions such as the concentration of reagents, pH and the flow rate of the reagent solution, reaction temperature and the reaction coil length were determined according to previous papers.^20–22

Validation

Calibration curves, linearity and limits of detection. The calibration curves for NMMA, ADMA and SDMA showed linearity over the range from 0.05 μmol l⁻¹ (0.5 pmol on column) to 5.0 μmol l⁻¹ (50 pmol on column) with the correlation coefficient greater than 0.998. The limits of detection (LODs) for NMMA, ADMA and SDMA were 0.005 μmol l⁻¹ (0.05 pmol on column), 0.008 μmol l⁻¹ (0.08 pmol on column) and 0.01 μmol l⁻¹ (0.1 pmol on column), respectively (signal-to-noise ratio = 2). The LODs for ADMA using OPA derivatization and fluorescence detection are reported to be 0.18 μmol l⁻¹, 0.1 μmol l⁻¹, 0.015 μmol l⁻¹ and 1 pmol on column from human plasma.^7,11–13 The LOD for ADMA by LC-MS/MS determination was 1 ng ml⁻¹ (0.006 μmol l⁻¹).²³Therefore, the presented method is sensitive compared to the reported methods and is satisfactory for the determination of methylated arginines in 10 μl of plasma samples.

Recovery. The recovery of methylated arginines using a cation-exchange column from standard mixtures was determined by the comparison of the slopes of calibration curves for concentration versus peak area to those obtained without using the cation-exchange column. The recoveries of NMMA, ADMA and SDMA were 96, 94 and 94%, respectively. The slopes of the calibration curves of the standard solution with the cation-exchange column were identical with those obtained from spiked rat plasma. Thus, the recoveries of NMMA, ADMA and SDMA from rat plasma were determined as 96, 94 and 94%, respectively. These results were superior to those previously reported methods by manual extraction using a cation-exchange column, ranging from 65.3 to 75.2%,¹¹ 80 to 85%,¹² and 79.4 to 88.1%.¹³

Precision and accuracy. The precision and accuracy of methylated arginines determination in rat plasma with the present HPLC system were examined by adding three different known amount of arginines to rat plasma samples, and the results are summarized in Table 1. The values of intra-day precision in this method were less than 7.4% and the accuracy values were between 94.3 and 103.7% (n = 5). The RSD of the inter-day assay evaluated from the same rat plasma for NMMA, ADMA and SDMA were 6.1, 5.8 and 7.0%, respectively. These results suggest that the proposed method is appropriate for the routine assay of methylated arginines.

Table 1 Precision and accuracy in the determination of methylated arginines in rat plasma samples

	Concentration added/μM	Precision (RSD, %)	Accuracy (%)
NMMA	0	4.6	—
	0.4	5.1	103.3
	1.0	4.8	95.9
	2.0	5.1	95.5
ADMA	0	4.3	—
	0.4	3.8	96.8
	1.0	4.2	102.0
	2.0	5.1	101.0
SDMA	0	6.4	—
	0.4	6.2	94.3
	1.0	7.4	103.7
	2.0	5.8	97.1

---

Determination of methylated arginines in rat plasma

Using the presented method, the concentrations of methylated arginines in rat plasma were determined. The concentration of NMMA, ADMA and SDMA were 0.18 ± 0.05, 0.60 ± 0.19 and 0.30 ± 0.09 μmol l⁻¹, respectively. The values of the latter two are comparable to those previously reported (ADMA: 0.56–0.92 μmol l⁻¹, 0.80 μmol l⁻¹ and SDMA: 0.38 μmol l⁻¹).^24,25

Conclusion

An automated analyzer for methylated arginines was developed. The method includes an on-line extraction of methylated arginines, a separation by ion-pair chromatography, a post-column derivatization by an OPA and thiol reaction, and a fluorescence detection. This method does not require any time-consuming pre-treatment. The LOD for ADMA was 0.008 μmol l⁻¹ (0.08 pmol on column), enabling the determination of methylated arginines in 10 μl of plasma. The separation was completed within 40 min with isocratic elution, suggesting that this system is suitable for routine assay in biological samples. The method promises to be particularly useful in the investigation of the relationship of methylated arginines in a number of pathological conditions, including renal failure, hypertension, microangiopathy, atherosclerosis, schizophrenia, and hypercholesteromia.

Acknowledgement

We thank Dr. Chiho Lee for his valuable suggestions and discussion. This work was supported in part by a grant-in-aid 13672250 for scientific research from the Ministry of Education, Science and Culture of Japan.

References

1. J. Leiper and P. Vallance, Cardiovasc. Res., 1999, 43, 542.
2. R. J. MacAllister, G. St. J. Whitley and P. Vallance, Kidney Int., 1994, 45, 737.
3. H. Matsuoka, S. Itoh, M. Kimoto, K. Kohno, O. Tamai, Y. Wada, H. Yasukawa, G. Iwami, S. Okuda and T. Imaizumi, Hypertension, 1997, 29, 242.
4. H. Herlitz, A. Petersson, L. Sigstrom, A. Wennmalm and G. Westberg, Scand. J. Uronephrol., 1997, 31, 477.
5. H. Miyazaki, H. Matsuoka, J. P. Cooke, M. Usui, S. Ueda, S. Okuda and T. Imaizumi, Circulation, 1999, 99, 1141.
6. I. Das, N. S. Khan, B. K. Puri and S. R. Hirsch, Neurosci. Lett., 1996, 215, 209.
7. R. Böger, M. Bode-Böger, A. Szuba, P. S. Tsao, J. R. Chan, O. Tangphao, T. F. Blaschke and J. P. Cooke, Circulation, 1998, 98, 1842.
8. P. Vallance, A. Leone, A. Calver, J. Collier and S. Moncada, Lancet, 1992, 339, 572.
9. E. S. Kang, M. T. Tevlin, Y. B. Wang, T. M. Chaing, R. Cardenas, L. K. Myers and S. R. Acchiardo, Am. J. Med. Sci., 1999, 317, 9.
10. R. J. MacAllister, H. Parry, M. Kimoto, T. Ogawa, R. J. Russel, H. Hodson, G. St. L. Whitley and P. Vallance, Br. J. Pharmacol., 1996, 119, 1533.
11. J. Pi, Y. Kumagai, G. Sun and N. Shimojo, J. Chromatogr., B: Biomed. Appl., 2000, 742, 199.
12. A. Pettersson, L. Uggla and V. Backman, J. Chromatogr., B: Biomed. Appl., 1997, 692, 257.
13. B.-M. Chen, L.-W. Xia and R. -Q Zhao, J. Chromatogr., B: Biomed Appl., 1997, 692, 467.
14. S. M. Bode-Böger, R. H. Böger, S. Kienke, W. Junker and J. C. Frolich, Biochem. Biophys. Res. Commun., 1996, 219, 598.
15. R. Böger, S. M. Bode-Böger, W. Thiele, W. Junker, K. Alexander and J. C. Frolich, Circulation, 1997, 95, 2068.
16. J. Meyer, N. Richter and M. Hecker, Anal. Biochem., 1997, 247, 11.
17. E. Causśe, N. Siri, J. F. Arnal, C. Bayle, P. Malatray, P. Valdiguie, R. Salvayre and F. Couderc, J. Chromatogr., B: Biomed. Appl., 2000, 741, 77.
18. V. R. Villanueva, M. Mardon, T. L. Goff and F. Moncelon, J. Chromatogr., 1987, 393, 107.
19. H. Azuma, J. Sato, H. Hamasaki, A. Sugimoto, E. Isotani and S. Obayashi, Br. J. Pharmacol., 1995, 115, 1001.
20. A. Friorino, G. Frigo and E. Cucchetti, J. Chromatogr., 1989, 476, 83.
21. R. L. Cunico and T. Schlabach, J. Chromatogr., 1983, 266, 461.
22. N. Nimura and T. Kinoshita, J. Chromatogr., 1986, 352, 169.
23. K. Vishwanathan, R. L. Tackett, J. T. Stewart and M. G. Bartlett, J. Chromatogr., B: Biomed. Appl., 2000, 748, 157.
24. J. Sato, H. Masuda, S. Tamaoki, H. Hamasaki, K. Ishizaka, O. Matsuoka and H. Azuma, Br. J. Pharmacol., 1998, 125, 469.
25. Q. Feng, X. Lu, A. J. Fortin, A. Petterson, T. Hender, R. L. Kline and J. M. O. Arnold, Cardiovasc. Res., 1998, 37, 667.

---