2785
J. Sep. Sci. 2014, 37, 2785–2790
Marta Pelcova´ 1,2 Monika Langmajerova´ 1,2 ˇ ´ Eliska Cvingrafov a´ 1,2 2,3 Jan Juˇrica ˇ Glatz1,2 Zdenek
Research Article
Nonaqueous capillary electrophoresis of dextromethorphan and its metabolites
1 Masaryk
University, Faculty of Science, Department of Biochemistry, Kamenice, Czech Republic 2 Masaryk University, CEITEC Central European Institute of Technology, Kamenice, Czech Republic 3 Masaryk University, Faculty of Medicine, Department of Pharmacology, Kamenice, Czech Republic Received June 2, 2014 Revised July 7, 2014 Accepted July 9, 2014
This study deals with the nonaqueous capillary electrophoretic separation of dextromethorphan and its metabolites using a methanolic background electrolyte. The optimization of separation conditions was performed in terms of the resolution of dextromethorphan and dextrorphan and the effect of separation temperature, voltage, and the characteristics of the background electrolyte were studied. Complete separation of all analytes was achieved in 40 mM ammonium acetate dissolved in methanol. Hydrodynamic injection was performed at 3 kPa for 4 s. The separation voltage was 20 kV accompanied by a low electric current. The ultraviolet detection was performed at 214 nm, the temperature of the capillary was 25⬚C. These conditions enabled the separation of four analytes plus the internal standard within 9 min. Further, the developed method was validated in terms of linearity, sensitivity, and repeatability. Rat liver perfusate samples were subjected to the nonaqueous capillary electrophoretic method to illustrate its applicability. Keywords: Dextromethorphan / Dextrorphan / Methanol / Nonaqueous media DOI 10.1002/jssc.201400582
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Organic solvents as a BGE component in CE have been documented since the introduction of CE into analytical chemistry [1–4]. In the traditional approach, an organic solvent forms part of an aqueous BGE and is thus considered a modifier. In contrast in nonaqueous capillary electrophoresis (NACE), the organic solvents are employed as the main dissolving agent of the BGE and thus some solvent-induced effects on ionic migration are more pronounced than in mixed hydroorganic solutions. Therefore NACE offers an interesting potential application of CE for systems that cannot be resolved in aqueous media. However, the same basic laws govern electromigration irrespective of CE mode. Their effect on the selectivity of the separation system and all relevant parameters affecting electromigration is the main reason for utilizing NACE. The first report on utilizing nonaqueous media as the BGE appeared in 1984 [3]. The BGE was based on the system tetraammonium perchlorate/hydrochloric acid in ace´ Department of Biochemistry, Correspondence: Dr. Marta Pelcova, Faculty of Science, Masaryk University, Czech Republic E-mail: [email protected] Fax: +42054949-3427
Abbreviations: ACN, acetonitrile; NH4 Ac, ammonium acetate; CYP, cytochrome P450; DXM, dextromethorphan; DTR, dextrorphan; FLD, fluorescence detection; HYM, (+)-3hydroxymorphinan; IS, internal standard; LEV, levallorphan; MeOH, methanol; MEM, (+)-3-methoxymorphinan; NACE, nonaqueous capillary electrophoresis C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
tonitrile (ACN), and was used for the separation of chinoline derivatives. Recently, NACE has become popular, especially for the separation of pharmaceuticals, highly hydrophobic species, and compounds unstable in water. The feasibility of NACE has been documented in several reviews of its theoretical characteristics and applications [5–9]. Critical evaluation of NACE and its theoretical characteristics have been covered in depth by the studies of Kenndler et al. [10–13], which comment on the aspects of organic solvent usage in CE and their benefits and drawbacks. Further, numerous publications have been devoted to the separation of pharmaceuticals by NACE, mostly concerning compounds that are basic in nature, such as local anesthetics, beta-blockers, alkaloids, or herbal drugs [14–19]. They confirm that organic solvents are beneficial for the CE analyses of drugs; they positively affect selectivity, exhibit good compatibility with the sample, and various detectors at low levels of electric current. This paper reports the separation of dextromethorphan (DXM) hydrochloride, as a parent drug, and its three metabolites: (+)-3-methoxymorphinan (MEM), dextrorphan (DTR), and (+)-3-hydroxymorphinan (HYM) (Fig. 1). DXM is an agonist at delta opioid receptors, which is clinically used as an antitussive. Despite other opioids exerting varying intensities of analgesic and sedative effects, DXM is believed to lack significant analgesic and sedative potency. It is also an antagonist of N-methyl-D-aspartate receptors. DXM is well established as a probe drug to determine cytochrome P450 (CYP) 2D6 isoform activity [20, 21]. DXM is primarily transformed to DTR via O-demethylation by CYP2D6, whose activity exhibits a high range of interindividual and interethnic variability,
www.jss-journal.com
2786
J. Sep. Sci. 2014, 37, 2785–2790
M. Pelcova´ et al.
2.2 Electrophoretic apparatus Analyses were performed with an Agilent 3D CE System instrument (Agilent, Waldbronn, Germany) equipped with a UV detector. Data were collected and evaluated with ChemStation (Agilent). The bare fused-silica capillary of 75 m id and 365 m od was purchased from Polymicro Technologies (Phoenix, AZ, USA). The pH values were measured using a pH meter (Orion Research EA 940, Orion Research Scottsdale, AZ, USA) and calibrated with standard aqueous buffer solutions. Distilled water was obtained from the DirectR 3 UV water purification system (Millipore, Darmstadt, Q Germany). Figure 1. Structure of studied compounds.
2.3 Sample preparation mainly caused by genetic polymorphism [22]. DXM is also N-demethylated to MEM, which is primarily mediated by CYP3A4 isoform in human liver microsomes [23, 24]. DTR and MEM are then demethylated to HYM via CYP3A4 and CYP2D6, respectively. Therefore, the simultaneous analysis of DXM and its metabolites could be useful for determining the metabolic activity of both CYP2D6 and CYP3A4. The separation of DXM and its metabolites by a variety of techniques has already been reported. There are several HPLC methods with different detection techniques focused on separating DXM and its metabolites in various matrices. An LC–MS/MS assay was developed and validated for the determination of DXM and all its metabolites in human urine [25], in human saliva and urine [21], and in everted gut sacs [26]. A combination of HPLC with fluorescence detection (FLD) was reported for the separation of DXM and its metabolites in human plasma and urine [27–29]. The DXM/DTR metabolic ratio in serum samples has been assessed by HPLC–FLD for CYP2D6 phenotyping [30]. Even GC–MS has been used for DXM and DTR determination in human hair [31]. CE methods for DXM analyses have also been published. The simultaneous determination of all the above-mentioned compounds in human plasma was reported ˇ aslavsk´a [33]. by Kristensen [32] and in human urine by C´ However, separation by NACE has only been accomplished for DXM and three other basic nitrogenous compounds [34]. In this study, a new NACE method for determining DXM and its metabolites was optimized and the method was employed for DXM metabolite separation in rat liver perfusates.
2 Materials and methods 2.1 Chemicals and reagents DXM, MEM, HYM, DTR, levallorphan (LEV), ACN, methanol (MeOH), and ammonium acetate (NH4 Ac) were purchased from Sigma (Sigma–Aldrich, Germany). All the solvents were of HPLC grade and filtered through a 0.45 m nylon filter before use. No other procedures were carried out. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Stock solutions of standards of 2 mM were prepared in MeOH and kept at –20⬚C. Their analytical solutions were prepared in specified solvents and diluted to the desired concentrations. The internal standard (IS) LEV was added to a final concentration of 100 M. The standard solutions were stored at 4⬚C. The real samples were perfusates drawn after 60 and 120 min of perfusion of isolated rat liver. The perfusion was accomplished according to the experiment as described in the paper by Jurica et al. [35]. The samples underwent the extraction procedure. Briefly, 0.5 mL of liver perfusate sample was transferred into a screw-capped glass tube and 0.25 mL of sodium carbonate and 0.4 mL of organic solvent mixture (n-hexane/butanol 9:1, v/v) was added followed by extensive shaking in a vortex device for 10 min at 2000 rpm. Afterwards the organic layer was transferred into another glass tube (3.5 mL) and 0.3 mL of 10 mM HCl was added. Again the mixture was extensively shaken for 10 min, frozen, and the organic phase was discarded. The aqueous phase was thawed and evaporated for 3 h in a vacuum concentrator (Eppendorf, Hamburg, Germany). The samples were then dissolved in MeOH and the IS was added.
2.4 BGE preparation The BGE was prepared from 0.2 M stock solutions of NH4 Ac in a particular solvent with the exception of ACN (the explanation is given below in Section 3.1) and diluted to the desired concentration. The pH* value was adjusted with an aqueous solution of ammonium hydroxide and measured with a standard pH meter calibrated with aqueous standard buffers. All BGEs were freshly prepared, filtered through a 0.45 m membrane filter, and degassed for 10 min before running.
2.5 CE experiment The separation was carried out in a 48.5 cm (40 cm effective length) × 75 m id bare fused-silica capillary. Before the www.jss-journal.com
J. Sep. Sci. 2014, 37, 2785–2790
Electrodriven Separations
2787
first usage, the capillary was activated with 1 M NaOH, 0.1 M NaOH, water, 1 M HCl, and water again, each for 10 min at 50⬚C. The capillary treated in this way was then rinsed with the tested BGE for 30 min. Samples were introduced by a pressure of 3 kPa for 4 s. The capillary temperature was set to 25⬚C. The separation voltage was kept at 20 kV. The detection wavelength was set to 214 nm. The mobility of the EOF was calculated from the mobility of thiourea as an EOF marker (n = 6). Preanalysis washing was conducted for 3 min at 100 kPa with the given BGE, and after each run the capillary was flushed with the corresponding pure organic solvent for 2 min and BGE for another 2 min, both steps at 100 kPa. The electrolyte was renewed after every two runs to obtain good repeatability. After each set of experiments, the capillary was washed with the pure organic solvent for 15 min and finally dried with air for 5 min. The capillary was stored dry. A new capillary was used for each solvent. To avoid sample evaporation, the samples were kept in a refrigerator when not in use.
3 Results and discussion 3.1 Optimization of NACE separation media To achieve satisfactory separations and sensitivity, the optimization of separation conditions was of primary importance. First, an appropriate solvent was determined. Viscosity and relative permittivity are among the most important parameters, which affect the mobility of EOF. The preliminary choice of solvents was set up according to the literature data [32, 34]. In the study presented by Kristensen [32], DXM was separated in a pure aqua-based BGE. The DXM separation from compounds with similar physicochemical properties was achieved in a MeOH/ACN system [34]. For the determination of optimal solvent, ACN, MeOH, and water were chosen. The solubility of electrolytes in ACN is low and therefore an NH4 Ac concentration >15 mM could not be achieved. Despite that, the highest value of EOF mobility was obtained for the BGE prepared from ACN (95 × 10−9 m2 ·V−1 ·s−1 for 10 mM NH4 Ac, pH* 8.3) accompanied by a very low electric current generated in the CE system (25 kV, a breakdown of the electric current was sometimes observed. The concentration of NH4 Ac produced the most pronounced effect on the RS of the targeted pair and the overall separation. The lowest RS value was determined to be 0.2 for the 10 mM electrolyte and an RS value of 2.2 was the best result, for 80 mM (see Fig. 2C). An increasing concentration of NH4 Ac and related BGE viscosity change resulted in a
www.jss-journal.com
Electrodriven Separations
J. Sep. Sci. 2014, 37, 2785–2790
2789
Table 2. Analyte concentrations in perfusate samples acquired at optimal separation conditions. Det. C, detected concentration, calculated c, concentration recalculated according to sample pretreatment
Sample 1 Analyte
Det. c (M)
MEM HYM DXM DTR
N.D.a) 3.7 Under LOQ 13.5
Sample 2 Calculated c (M)
Det. c (M)
Calculated c (M)
0.7
4.9
0.9
2.6
21.3
4.1
a) Not detected.
decrease in EOF. At the same time the electric current rose from 13 to 80 A. The total analysis time was evaluated in accordance with the migration time of LEV, the IS. Further, the separation efficiency was evaluated from the peak of the IS, since it was the last migrating analyte, so the effect of longitudinal diffusion would be the most pronounced and it should reflect the objective state of the separation system. A previously published paper showed that even a mixture of solvents can improve separation selectivity [34]. Therefore the effect of adding ACN to the methanolic BGE was tested. However, the addition of ACN resulted in a decrease in resolution and efficiency in the examined range from 0 to 50%. The optimization results can be found in Supporting Information Table S1. The amount of ACN affects the BGE viscosity and its overall relative permittivity [5] and hence EOF is also influenced. Accordingly, the analysis time was almost three times shorter with increased ACN content at a constant separation voltage. The main aim of this optimization was to reach the highest RS of the DXM–DTR pair and so it was decided to leave ACN out of the BGE. Finally, it can be concluded that 40 mM NH4 Ac in MeOH, pH* 8.0 is a suitable BGE for the given analysis performed at 25⬚C and a separation voltage of 20 kV. The typical separation of standards is depicted in Fig. 3A. After the optimal separation conditions were established, partial validation tests for linearity and LOD and LOQ were performed. The linearity data were obtained with standard samples containing 100 M LEV as the IS. The calibration curves were obtained as the dependence of the normalized peak area (a ratio of peak area of analyte and IS peak area) to analyte concentration. Table 1 shows the regression equations and coefficients of determination for the four analytes. Intraday repeatability was determined from ten consecutive runs. The LOD and LOQ was calculated on the basis of a detector S/N of 3:1 and 10:1, respectively.
3.3 Applications The applicability of the method was assessed by the determination of DXM and its metabolite concentration in rat liver perfusates. The samples underwent the extraction procedure as described in Section 2.3 and were finally dissolved C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
in MeOH containing the IS. The efficiency of the extraction procedure was quite high for all analytes, specifically 86% for MEM, 81% for HYM, 92% for DXM and 76% for DTR. A typical electropherogram is shown in Fig. 3B. The peaks were identified by comparing the migration times of separated compounds and standards and by spiking the sample with the standards. Each sample was analyzed three times by the optimized method. The detected amount of compound was recalculated to the ultimate compound content in perfusate samples as dilution and concentration took place during the sample preparation procedure. The recalculated concentrations of the analytes are summarized in Table 2.
4 Concluding remarks This paper describes the NACE method for the simultaneous detection and quantitation of DXM and its metabolites. From the optimization, we can conclude that MeOH was the most appropriate solvent for the BGE in which the studied compounds were separated from each other. Further, the resolution of all peaks was most affected by NH4 Ac concentration; the other operating parameters only exhibited minor influences. Among the advantages of the presented NACE method are its speed, high peak efficiency, and compatibility with other detection techniques, e.g. MS, if more precise identity confirmation is needed. The utilization of the optimized NACE method was documented by DXM metabolite determination and their quantitation in rat liver perfusate samples. Financial support was provided by the Czech Science Foundation, No. GBP206/12/G014. The authors have declared no conflict of interest.
5 References [1] Beckers, J. L., Everaerts, F. M., J. Chromatogr. 1972, 68, 207–230. [2] Reijenga, J. C., Staats, H. J. L. A., Everaerts, F. M., J. Chromatogr. 1983, 267, 85–89.
www.jss-journal.com
2790
M. Pelcova´ et al.
J. Sep. Sci. 2014, 37, 2785–2790
[3] Walbroehl, Y., Jorgenson, J. W., J. Chromatogr. 1984, 315, 135–143.
¨ [22] Schmid, B., Bircher, J., Preisig, J., Kupfer, A., Clin. Pharmacol. Ther. 1985, 38, 618–624.
ˇ [4] Bocek, P., Deml, M., Gebauer, P., Dolnik, V., Analytical Isotachophoresis, Wiley VCH, Weinheim 1988.
[23] Gorski, J. C., Jones, D. R., Wrigton, S. A., Hall, S. D., Biochem. Pharmacol. 1994, 48, 173–182.
[5] Steiner, F., Hassel, M., Electrophoresis 2000, 21, 3994– 4016.
[24] Yu, A., Haining, R. L., Drug. Metab. Dispos. 2001, 29, 1514–1520.
[6] Riekkola, M.-L., Electrophoresis 2002, 23, 3865–3883.
[25] Vengurlekar, S. S., Heitkamp, J., McCush, F., Velagaleti, P. R., Brisson, J. H., Bramer, S. L., J. Pharm. Biomed. Anal. 2002, 30, 113–124.
[7] Geiser, L., Veuthey, J.-L., Electrophoresis 2009, 30, 36–49. [8] Scriba, G. K. E., J. Chromatogr. A 2007, 1159, 28–41. [9] Geiser, L., Veuthey, J.-L., Electrophoresis 2007, 28, 45–57. [10] Kenndler, E., J. Chromatogr. A 2014, 1335, 16–30. [11] Kenndler, E., J. Chromatogr. A 2014, 1335, 31–41.
[26] Arrellano, C., Philibert, C., Dane a` Yakan, E. N., Vachoux, C., Lacombe, O., Houin, G., J. Chromatogr. B 2005, 819, 105–113. [27] Afshar, M., Rouini, M.-R., Amini, M., J. Chromatogr. B 2004, 802, 317–322.
[12] Jouyban, A., Kenndler, E., Curr. Anal. Chem. 2014, 10, 248–266.
[28] Daali, Y., Cherkaoui, S., Doffey-Lazeyras, F., Dayer, P., Desmeules, J. A., J. Chromatogr. B 2008, 861, 56–63.
[13] Tellez, A., Kenndler, E., Electrophoresis 2009, 30, 3978– 3985.
´ G., Chladek, ´ ´ J., Beranek, ´ [29] Zimova, J., Mart´ınkova, M., Chem. Listy 2000, 94, 132–135.
´ ´ [14] Morales-Cid, G., Cardenas, S., Simonet, B. M., Valcarcel, M., Electrophoresis 2009, 30, 1684–1691.
ˇ ´ A., Pindurova, ´ E., [30] Juˇrica, J., Bartecek, R., Zourkova, ˇ ´ A., Kasp ˇ arek, ´ Sulcov a, T., Zendulka, O., J. Clin. Pharm. Ther. 2012, 37, 486–490.
[15] Servais, A.-C., Fillet, M., Chiap, P., Abushoffa, A. M., Hubert, P., Crommen, J., J. Sep. Sci. 2002, 25, 1087–1095. [16] Roscher, J., Posch, T. N., Putz, M., Huhn, C., Electrophoresis 2012, 33, 1567–1570.
[31] Kim, J. Y., Suh, S. I., Paeng, K.-J., In, M. K., Chromatographia 2004, 60, 703–707.
[18] Chen, Q., Lil, P., Zhang, Z., Lil, K., Liu, J., Li, O., J. Sep. Sci. 2008, 31, 2211–2218.
[32] Kristensen, H., J. Pharm. Biomed. Anal. 1998, 18, 827– 838. ´ [33] Cˇ aslavska, J., Hufschmid, E., Theurillat, R., Desiderio, C., Wolfisberg, H., Thormann, W., J. Chromatogr. B 1994, 656, 219–231.
[19] Gao, W. H., Lin, S. Y., Jia, L., Gu, X. K., Chen, X. G., De Hu, Z., J. Sep. Sci. 2005, 28, 92–97.
[34] Dong, Y., Chen, X., Chen, Y., Chen, X., Hu, Z., J. Pharm. Biomed. Anal. 2005, 39, 285–289.
[20] Dayer, P., Leeman, T., Striberni, R., Clin. Pharmacol. Ther. 1989, 45, 34–40.
ˇ y, ´ J., Zahradn´ıkova, ´ L. Z., Tomandl, J., [35] Juˇrica, J., Konecn J. Pharm. Biomed. Anal. 2010, 52, 557–564.
[21] Lutz, U., Volkel, W., Lutz, R. W., Lutz, W. K., J. Chromatogr. B 2004, 813, 217–225.
[36] Wright, P. B., Lister, A. S., Dorsey, J. G., Anal. Chem. 1997, 69, 3251–3259.
¨ [17] Ganzera, M., Kruger, A., Wink, M., J. Pharm. Biomed. Anal. 2010, 53, 1231–1235.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
J. Sep. Sci. 2014, 37, 2785–2790
Marta Pelcova´ 1,2 Monika Langmajerova´ 1,2 ˇ ´ Eliska Cvingrafov a´ 1,2 2,3 Jan Juˇrica ˇ Glatz1,2 Zdenek
Research Article
Nonaqueous capillary electrophoresis of dextromethorphan and its metabolites
1 Masaryk
University, Faculty of Science, Department of Biochemistry, Kamenice, Czech Republic 2 Masaryk University, CEITEC Central European Institute of Technology, Kamenice, Czech Republic 3 Masaryk University, Faculty of Medicine, Department of Pharmacology, Kamenice, Czech Republic Received June 2, 2014 Revised July 7, 2014 Accepted July 9, 2014
This study deals with the nonaqueous capillary electrophoretic separation of dextromethorphan and its metabolites using a methanolic background electrolyte. The optimization of separation conditions was performed in terms of the resolution of dextromethorphan and dextrorphan and the effect of separation temperature, voltage, and the characteristics of the background electrolyte were studied. Complete separation of all analytes was achieved in 40 mM ammonium acetate dissolved in methanol. Hydrodynamic injection was performed at 3 kPa for 4 s. The separation voltage was 20 kV accompanied by a low electric current. The ultraviolet detection was performed at 214 nm, the temperature of the capillary was 25⬚C. These conditions enabled the separation of four analytes plus the internal standard within 9 min. Further, the developed method was validated in terms of linearity, sensitivity, and repeatability. Rat liver perfusate samples were subjected to the nonaqueous capillary electrophoretic method to illustrate its applicability. Keywords: Dextromethorphan / Dextrorphan / Methanol / Nonaqueous media DOI 10.1002/jssc.201400582
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Organic solvents as a BGE component in CE have been documented since the introduction of CE into analytical chemistry [1–4]. In the traditional approach, an organic solvent forms part of an aqueous BGE and is thus considered a modifier. In contrast in nonaqueous capillary electrophoresis (NACE), the organic solvents are employed as the main dissolving agent of the BGE and thus some solvent-induced effects on ionic migration are more pronounced than in mixed hydroorganic solutions. Therefore NACE offers an interesting potential application of CE for systems that cannot be resolved in aqueous media. However, the same basic laws govern electromigration irrespective of CE mode. Their effect on the selectivity of the separation system and all relevant parameters affecting electromigration is the main reason for utilizing NACE. The first report on utilizing nonaqueous media as the BGE appeared in 1984 [3]. The BGE was based on the system tetraammonium perchlorate/hydrochloric acid in ace´ Department of Biochemistry, Correspondence: Dr. Marta Pelcova, Faculty of Science, Masaryk University, Czech Republic E-mail: [email protected] Fax: +42054949-3427
Abbreviations: ACN, acetonitrile; NH4 Ac, ammonium acetate; CYP, cytochrome P450; DXM, dextromethorphan; DTR, dextrorphan; FLD, fluorescence detection; HYM, (+)-3hydroxymorphinan; IS, internal standard; LEV, levallorphan; MeOH, methanol; MEM, (+)-3-methoxymorphinan; NACE, nonaqueous capillary electrophoresis C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
tonitrile (ACN), and was used for the separation of chinoline derivatives. Recently, NACE has become popular, especially for the separation of pharmaceuticals, highly hydrophobic species, and compounds unstable in water. The feasibility of NACE has been documented in several reviews of its theoretical characteristics and applications [5–9]. Critical evaluation of NACE and its theoretical characteristics have been covered in depth by the studies of Kenndler et al. [10–13], which comment on the aspects of organic solvent usage in CE and their benefits and drawbacks. Further, numerous publications have been devoted to the separation of pharmaceuticals by NACE, mostly concerning compounds that are basic in nature, such as local anesthetics, beta-blockers, alkaloids, or herbal drugs [14–19]. They confirm that organic solvents are beneficial for the CE analyses of drugs; they positively affect selectivity, exhibit good compatibility with the sample, and various detectors at low levels of electric current. This paper reports the separation of dextromethorphan (DXM) hydrochloride, as a parent drug, and its three metabolites: (+)-3-methoxymorphinan (MEM), dextrorphan (DTR), and (+)-3-hydroxymorphinan (HYM) (Fig. 1). DXM is an agonist at delta opioid receptors, which is clinically used as an antitussive. Despite other opioids exerting varying intensities of analgesic and sedative effects, DXM is believed to lack significant analgesic and sedative potency. It is also an antagonist of N-methyl-D-aspartate receptors. DXM is well established as a probe drug to determine cytochrome P450 (CYP) 2D6 isoform activity [20, 21]. DXM is primarily transformed to DTR via O-demethylation by CYP2D6, whose activity exhibits a high range of interindividual and interethnic variability,
www.jss-journal.com
2786
J. Sep. Sci. 2014, 37, 2785–2790
M. Pelcova´ et al.
2.2 Electrophoretic apparatus Analyses were performed with an Agilent 3D CE System instrument (Agilent, Waldbronn, Germany) equipped with a UV detector. Data were collected and evaluated with ChemStation (Agilent). The bare fused-silica capillary of 75 m id and 365 m od was purchased from Polymicro Technologies (Phoenix, AZ, USA). The pH values were measured using a pH meter (Orion Research EA 940, Orion Research Scottsdale, AZ, USA) and calibrated with standard aqueous buffer solutions. Distilled water was obtained from the DirectR 3 UV water purification system (Millipore, Darmstadt, Q Germany). Figure 1. Structure of studied compounds.
2.3 Sample preparation mainly caused by genetic polymorphism [22]. DXM is also N-demethylated to MEM, which is primarily mediated by CYP3A4 isoform in human liver microsomes [23, 24]. DTR and MEM are then demethylated to HYM via CYP3A4 and CYP2D6, respectively. Therefore, the simultaneous analysis of DXM and its metabolites could be useful for determining the metabolic activity of both CYP2D6 and CYP3A4. The separation of DXM and its metabolites by a variety of techniques has already been reported. There are several HPLC methods with different detection techniques focused on separating DXM and its metabolites in various matrices. An LC–MS/MS assay was developed and validated for the determination of DXM and all its metabolites in human urine [25], in human saliva and urine [21], and in everted gut sacs [26]. A combination of HPLC with fluorescence detection (FLD) was reported for the separation of DXM and its metabolites in human plasma and urine [27–29]. The DXM/DTR metabolic ratio in serum samples has been assessed by HPLC–FLD for CYP2D6 phenotyping [30]. Even GC–MS has been used for DXM and DTR determination in human hair [31]. CE methods for DXM analyses have also been published. The simultaneous determination of all the above-mentioned compounds in human plasma was reported ˇ aslavsk´a [33]. by Kristensen [32] and in human urine by C´ However, separation by NACE has only been accomplished for DXM and three other basic nitrogenous compounds [34]. In this study, a new NACE method for determining DXM and its metabolites was optimized and the method was employed for DXM metabolite separation in rat liver perfusates.
2 Materials and methods 2.1 Chemicals and reagents DXM, MEM, HYM, DTR, levallorphan (LEV), ACN, methanol (MeOH), and ammonium acetate (NH4 Ac) were purchased from Sigma (Sigma–Aldrich, Germany). All the solvents were of HPLC grade and filtered through a 0.45 m nylon filter before use. No other procedures were carried out. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Stock solutions of standards of 2 mM were prepared in MeOH and kept at –20⬚C. Their analytical solutions were prepared in specified solvents and diluted to the desired concentrations. The internal standard (IS) LEV was added to a final concentration of 100 M. The standard solutions were stored at 4⬚C. The real samples were perfusates drawn after 60 and 120 min of perfusion of isolated rat liver. The perfusion was accomplished according to the experiment as described in the paper by Jurica et al. [35]. The samples underwent the extraction procedure. Briefly, 0.5 mL of liver perfusate sample was transferred into a screw-capped glass tube and 0.25 mL of sodium carbonate and 0.4 mL of organic solvent mixture (n-hexane/butanol 9:1, v/v) was added followed by extensive shaking in a vortex device for 10 min at 2000 rpm. Afterwards the organic layer was transferred into another glass tube (3.5 mL) and 0.3 mL of 10 mM HCl was added. Again the mixture was extensively shaken for 10 min, frozen, and the organic phase was discarded. The aqueous phase was thawed and evaporated for 3 h in a vacuum concentrator (Eppendorf, Hamburg, Germany). The samples were then dissolved in MeOH and the IS was added.
2.4 BGE preparation The BGE was prepared from 0.2 M stock solutions of NH4 Ac in a particular solvent with the exception of ACN (the explanation is given below in Section 3.1) and diluted to the desired concentration. The pH* value was adjusted with an aqueous solution of ammonium hydroxide and measured with a standard pH meter calibrated with aqueous standard buffers. All BGEs were freshly prepared, filtered through a 0.45 m membrane filter, and degassed for 10 min before running.
2.5 CE experiment The separation was carried out in a 48.5 cm (40 cm effective length) × 75 m id bare fused-silica capillary. Before the www.jss-journal.com
J. Sep. Sci. 2014, 37, 2785–2790
Electrodriven Separations
2787
first usage, the capillary was activated with 1 M NaOH, 0.1 M NaOH, water, 1 M HCl, and water again, each for 10 min at 50⬚C. The capillary treated in this way was then rinsed with the tested BGE for 30 min. Samples were introduced by a pressure of 3 kPa for 4 s. The capillary temperature was set to 25⬚C. The separation voltage was kept at 20 kV. The detection wavelength was set to 214 nm. The mobility of the EOF was calculated from the mobility of thiourea as an EOF marker (n = 6). Preanalysis washing was conducted for 3 min at 100 kPa with the given BGE, and after each run the capillary was flushed with the corresponding pure organic solvent for 2 min and BGE for another 2 min, both steps at 100 kPa. The electrolyte was renewed after every two runs to obtain good repeatability. After each set of experiments, the capillary was washed with the pure organic solvent for 15 min and finally dried with air for 5 min. The capillary was stored dry. A new capillary was used for each solvent. To avoid sample evaporation, the samples were kept in a refrigerator when not in use.
3 Results and discussion 3.1 Optimization of NACE separation media To achieve satisfactory separations and sensitivity, the optimization of separation conditions was of primary importance. First, an appropriate solvent was determined. Viscosity and relative permittivity are among the most important parameters, which affect the mobility of EOF. The preliminary choice of solvents was set up according to the literature data [32, 34]. In the study presented by Kristensen [32], DXM was separated in a pure aqua-based BGE. The DXM separation from compounds with similar physicochemical properties was achieved in a MeOH/ACN system [34]. For the determination of optimal solvent, ACN, MeOH, and water were chosen. The solubility of electrolytes in ACN is low and therefore an NH4 Ac concentration >15 mM could not be achieved. Despite that, the highest value of EOF mobility was obtained for the BGE prepared from ACN (95 × 10−9 m2 ·V−1 ·s−1 for 10 mM NH4 Ac, pH* 8.3) accompanied by a very low electric current generated in the CE system (25 kV, a breakdown of the electric current was sometimes observed. The concentration of NH4 Ac produced the most pronounced effect on the RS of the targeted pair and the overall separation. The lowest RS value was determined to be 0.2 for the 10 mM electrolyte and an RS value of 2.2 was the best result, for 80 mM (see Fig. 2C). An increasing concentration of NH4 Ac and related BGE viscosity change resulted in a
www.jss-journal.com
Electrodriven Separations
J. Sep. Sci. 2014, 37, 2785–2790
2789
Table 2. Analyte concentrations in perfusate samples acquired at optimal separation conditions. Det. C, detected concentration, calculated c, concentration recalculated according to sample pretreatment
Sample 1 Analyte
Det. c (M)
MEM HYM DXM DTR
N.D.a) 3.7 Under LOQ 13.5
Sample 2 Calculated c (M)
Det. c (M)
Calculated c (M)
0.7
4.9
0.9
2.6
21.3
4.1
a) Not detected.
decrease in EOF. At the same time the electric current rose from 13 to 80 A. The total analysis time was evaluated in accordance with the migration time of LEV, the IS. Further, the separation efficiency was evaluated from the peak of the IS, since it was the last migrating analyte, so the effect of longitudinal diffusion would be the most pronounced and it should reflect the objective state of the separation system. A previously published paper showed that even a mixture of solvents can improve separation selectivity [34]. Therefore the effect of adding ACN to the methanolic BGE was tested. However, the addition of ACN resulted in a decrease in resolution and efficiency in the examined range from 0 to 50%. The optimization results can be found in Supporting Information Table S1. The amount of ACN affects the BGE viscosity and its overall relative permittivity [5] and hence EOF is also influenced. Accordingly, the analysis time was almost three times shorter with increased ACN content at a constant separation voltage. The main aim of this optimization was to reach the highest RS of the DXM–DTR pair and so it was decided to leave ACN out of the BGE. Finally, it can be concluded that 40 mM NH4 Ac in MeOH, pH* 8.0 is a suitable BGE for the given analysis performed at 25⬚C and a separation voltage of 20 kV. The typical separation of standards is depicted in Fig. 3A. After the optimal separation conditions were established, partial validation tests for linearity and LOD and LOQ were performed. The linearity data were obtained with standard samples containing 100 M LEV as the IS. The calibration curves were obtained as the dependence of the normalized peak area (a ratio of peak area of analyte and IS peak area) to analyte concentration. Table 1 shows the regression equations and coefficients of determination for the four analytes. Intraday repeatability was determined from ten consecutive runs. The LOD and LOQ was calculated on the basis of a detector S/N of 3:1 and 10:1, respectively.
3.3 Applications The applicability of the method was assessed by the determination of DXM and its metabolite concentration in rat liver perfusates. The samples underwent the extraction procedure as described in Section 2.3 and were finally dissolved C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
in MeOH containing the IS. The efficiency of the extraction procedure was quite high for all analytes, specifically 86% for MEM, 81% for HYM, 92% for DXM and 76% for DTR. A typical electropherogram is shown in Fig. 3B. The peaks were identified by comparing the migration times of separated compounds and standards and by spiking the sample with the standards. Each sample was analyzed three times by the optimized method. The detected amount of compound was recalculated to the ultimate compound content in perfusate samples as dilution and concentration took place during the sample preparation procedure. The recalculated concentrations of the analytes are summarized in Table 2.
4 Concluding remarks This paper describes the NACE method for the simultaneous detection and quantitation of DXM and its metabolites. From the optimization, we can conclude that MeOH was the most appropriate solvent for the BGE in which the studied compounds were separated from each other. Further, the resolution of all peaks was most affected by NH4 Ac concentration; the other operating parameters only exhibited minor influences. Among the advantages of the presented NACE method are its speed, high peak efficiency, and compatibility with other detection techniques, e.g. MS, if more precise identity confirmation is needed. The utilization of the optimized NACE method was documented by DXM metabolite determination and their quantitation in rat liver perfusate samples. Financial support was provided by the Czech Science Foundation, No. GBP206/12/G014. The authors have declared no conflict of interest.
5 References [1] Beckers, J. L., Everaerts, F. M., J. Chromatogr. 1972, 68, 207–230. [2] Reijenga, J. C., Staats, H. J. L. A., Everaerts, F. M., J. Chromatogr. 1983, 267, 85–89.
www.jss-journal.com
2790
M. Pelcova´ et al.
J. Sep. Sci. 2014, 37, 2785–2790
[3] Walbroehl, Y., Jorgenson, J. W., J. Chromatogr. 1984, 315, 135–143.
¨ [22] Schmid, B., Bircher, J., Preisig, J., Kupfer, A., Clin. Pharmacol. Ther. 1985, 38, 618–624.
ˇ [4] Bocek, P., Deml, M., Gebauer, P., Dolnik, V., Analytical Isotachophoresis, Wiley VCH, Weinheim 1988.
[23] Gorski, J. C., Jones, D. R., Wrigton, S. A., Hall, S. D., Biochem. Pharmacol. 1994, 48, 173–182.
[5] Steiner, F., Hassel, M., Electrophoresis 2000, 21, 3994– 4016.
[24] Yu, A., Haining, R. L., Drug. Metab. Dispos. 2001, 29, 1514–1520.
[6] Riekkola, M.-L., Electrophoresis 2002, 23, 3865–3883.
[25] Vengurlekar, S. S., Heitkamp, J., McCush, F., Velagaleti, P. R., Brisson, J. H., Bramer, S. L., J. Pharm. Biomed. Anal. 2002, 30, 113–124.
[7] Geiser, L., Veuthey, J.-L., Electrophoresis 2009, 30, 36–49. [8] Scriba, G. K. E., J. Chromatogr. A 2007, 1159, 28–41. [9] Geiser, L., Veuthey, J.-L., Electrophoresis 2007, 28, 45–57. [10] Kenndler, E., J. Chromatogr. A 2014, 1335, 16–30. [11] Kenndler, E., J. Chromatogr. A 2014, 1335, 31–41.
[26] Arrellano, C., Philibert, C., Dane a` Yakan, E. N., Vachoux, C., Lacombe, O., Houin, G., J. Chromatogr. B 2005, 819, 105–113. [27] Afshar, M., Rouini, M.-R., Amini, M., J. Chromatogr. B 2004, 802, 317–322.
[12] Jouyban, A., Kenndler, E., Curr. Anal. Chem. 2014, 10, 248–266.
[28] Daali, Y., Cherkaoui, S., Doffey-Lazeyras, F., Dayer, P., Desmeules, J. A., J. Chromatogr. B 2008, 861, 56–63.
[13] Tellez, A., Kenndler, E., Electrophoresis 2009, 30, 3978– 3985.
´ G., Chladek, ´ ´ J., Beranek, ´ [29] Zimova, J., Mart´ınkova, M., Chem. Listy 2000, 94, 132–135.
´ ´ [14] Morales-Cid, G., Cardenas, S., Simonet, B. M., Valcarcel, M., Electrophoresis 2009, 30, 1684–1691.
ˇ ´ A., Pindurova, ´ E., [30] Juˇrica, J., Bartecek, R., Zourkova, ˇ ´ A., Kasp ˇ arek, ´ Sulcov a, T., Zendulka, O., J. Clin. Pharm. Ther. 2012, 37, 486–490.
[15] Servais, A.-C., Fillet, M., Chiap, P., Abushoffa, A. M., Hubert, P., Crommen, J., J. Sep. Sci. 2002, 25, 1087–1095. [16] Roscher, J., Posch, T. N., Putz, M., Huhn, C., Electrophoresis 2012, 33, 1567–1570.
[31] Kim, J. Y., Suh, S. I., Paeng, K.-J., In, M. K., Chromatographia 2004, 60, 703–707.
[18] Chen, Q., Lil, P., Zhang, Z., Lil, K., Liu, J., Li, O., J. Sep. Sci. 2008, 31, 2211–2218.
[32] Kristensen, H., J. Pharm. Biomed. Anal. 1998, 18, 827– 838. ´ [33] Cˇ aslavska, J., Hufschmid, E., Theurillat, R., Desiderio, C., Wolfisberg, H., Thormann, W., J. Chromatogr. B 1994, 656, 219–231.
[19] Gao, W. H., Lin, S. Y., Jia, L., Gu, X. K., Chen, X. G., De Hu, Z., J. Sep. Sci. 2005, 28, 92–97.
[34] Dong, Y., Chen, X., Chen, Y., Chen, X., Hu, Z., J. Pharm. Biomed. Anal. 2005, 39, 285–289.
[20] Dayer, P., Leeman, T., Striberni, R., Clin. Pharmacol. Ther. 1989, 45, 34–40.
ˇ y, ´ J., Zahradn´ıkova, ´ L. Z., Tomandl, J., [35] Juˇrica, J., Konecn J. Pharm. Biomed. Anal. 2010, 52, 557–564.
[21] Lutz, U., Volkel, W., Lutz, R. W., Lutz, W. K., J. Chromatogr. B 2004, 813, 217–225.
[36] Wright, P. B., Lister, A. S., Dorsey, J. G., Anal. Chem. 1997, 69, 3251–3259.
¨ [17] Ganzera, M., Kruger, A., Wink, M., J. Pharm. Biomed. Anal. 2010, 53, 1231–1235.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
