2759
Electrophoresis 2014, 35, 2759–2764
Margarita V. Lebedeva Aleksandra F. Prokhorova Elena N. Shapovalova Oleg A. Shpigun Chemistry Department, Lomonosov Moscow State University, Moscow, Russian Federation
Received March 14, 2014 Revised July 8, 2014 Accepted July 10, 2014
Research Article
Clarithromycin as a chiral selector for enantioseparation of basic compounds in nonaqueous capillary electrophoresis The first use of macrolide antibiotic clarithromycin (CLM) in nonaqueous media for enantioseparation (partial or baseline) of the following compounds: alprenolol, atenolol, metoprolol, clenbuterol, methoxyphenamine, pindolol, propranolol, sotalol, synephrine, labetalol, and fenoterol is reported. Each analysis took less than 15 min. To find optimal separation conditions, some properties of CLM (adsorption, solubility), as well as the effect of experimental parameters on the enantioseparation of analytes (background electrolyte composition, chiral selector concentration, temperature, and applied voltage) were studied. The best chiral resolution was achieved in methanolic solution of 100 mM citric acid, 10 mM NaOH, 240–300 mM H3 BO3 , and 60–75 mM CLM. Using the proposed procedure, adsorption of CLM on the capillary wall was negligible and the repeatability of the migration times (RSD) was as good as 1.6%. For the analysis of propranolol, the linearity was achieved in the concentration range 2.5 × 10−2 – 3.0 × 10−1 mg/mL with the LODs (3 × S/N) being equal 2.6 × 10−3 mg/mL and 2.8 × 10−3 mg/mL for the first and the second enantiomers, respectively. Linear range for metoprolol enantiomers was 1.0 × 10−2 –1.6 × 10−1 mg/mL. The LODs (3 × S/N) were determined as 2.8 × 10−3 and 3.0 × 10−3 mg/mL for the first and the second enantiomers, respectively. Keywords: -Blockers / Chiral separation / Clarithromycin / Macrolides / Nonaqueous capillary electrophoresis DOI 10.1002/elps.201400135
1 Introduction Analysis of chiral compounds is performed by HPLC, CE, supercritical fluid chromatography, TLC, and GC [1]. CE is often the method of choice for chiral analyses due to its improved efficiency and resolution. The fact that chiral selector is dissolved in a BGE enables a rapid method development and reduced cost of both the analysis and the method development. Nonaqueous CE (NACE) is a powerful tool for chiral analysis due to the higher solubility and different selectivity of some chiral selectors in a nonaqueous environment, and also flexibility of the technique [1–4]. Lower conductivity of electrolytes in NACE allows higher voltage to apply and higher buffer concentration to use. The enantioseparation mechanism in nonaqueous media differs from that in aqueous solutions resulting in the enantiomer migration order reversal as it was shown for heptakis (2,3-dimethyl-6-sulfo)Correspondence: Dr. Aleksandra F. Prokhorova, Chemistry Department, Lomonosov Moscow State University, Leninskie Gory, d.1, str.3, Moscow, 119991, Russian Federation E-mail: [email protected] Fax: +7-495-939-46-75
Abbreviations: AA, acetic acid; CA, citric acid; CLM, clarithromycin; MA, macrocyclic antibiotic; MeOH, methanol; NACE, nonaqueous CE; TBA, tributylamine; TEA, triethylamine C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
-CD [5, 6]. In addition, formation of external complexes between propranolol and heptakis (2,3-dimethyl-6-sulfo)-CD instead of inclusion complexes is observed in methanolic BGEs [6]. However, when compared NACE with aqueous CE, the reported chiral selectors and selectands are fewer. To date, different types of cyclodextrins [7], macrocyclic antibiotics, ion-pairing agents [8], and some other are used for chiral separations in CE. Macrocyclic glycopeptide antibiotics such as vancomycin, balhimycin, eremomycin provide high enantioselectivity in aqueous buffers but, unfortunately, they tend to adsorb on the capillary wall [9]. They also strongly absorb in the UV-region, therefore, significantly decreasing the detection sensitivity. From this point of view, macrolides, another type of macrocyclic antibiotics, are promising selectors for enantioseparation in nonaqueous BGEs because they are insoluble in water; they do not have significant absorbance in UV-region, while possessing many stereogenic centers [10–18]. The application spectrum of macrolide is still not as large as it could be. Only five papers on their use in aqueous CE have been published, including the first report of macrolide antibiotic, erythromycin, as a chiral selector in 2002 [10, 11] for enantioseparation of antihepatitis drugs (biphenyldimethylester derivatives). It was then used for baseline enantioseparation of amino alcohols (propranolol, duloxetine) and N,N-dimethyl-3-(2methoxyphenoxy)-3-propyl amine [14]. The attempt to use erythromycin and its different derivatives to enantioseparate
www.electrophoresis-journal.com
2760
M. V. Lebedeva et al.
aromatic carboxylic acids, some neutral compounds (warfarin), and amines with low concentration of antibiotic in BGE were not successful [13]. Clarithromycin (CLM) was once [12] used as lactobionate in the mixture of 12.5 mM phosphate buffer (pH 7.3) and methanol (50/50 v/v%) and showed enantioselectivity toward basic compounds such as amines and amino alcohols. In the above paper, the role of the lactobionate anion was not discussed. Also were published four papers about use of macrolides in NACE. Azithromycin dissolved in the mixture of ACN and methanol was beneficial for enantioseparation of carvedilol, darifenacine, cetirizine, citalopram, sertraline as well as tryptophan [15]. According to another paper [16], this antibiotic could resolve enantiomers of tetrahydrozoline and methoxyphenamine. Eryrthromycin lactobionate could resolve enantiomers of propranolol and duloxetine [17], and boromycin was successfully used for enantioseparation of tryptophanol, norepinephrine, octopamine, p-hydroxynorephedrine, 2-amino-1-phenylethanol, and ␣-benzylamine [18]. The possibilities of this type of antibiotics as chiral selectors are not very well established and, therefore, the intensive study of their enantioseparation capability is required. Many amines and amino alcohols have significant biological activity. Some of them are -blockers (atenolol, metoprolol, propranolol, etc.), salbutamol is sympathomimetic, methoxyphenamine and clenbuterol are bronchodilators, while synephrine is vasoconstrictor. Enantioseparation of these compounds have been performed by different separations techniques including both aqueous and NACE. Different chiral selectors were used (cyclodextrins, other polysaccharides, including combination of two selectors) [5,6,19–21]. The aim of this work was to evaluate enantioselective ability of the macrolide antibiotic CLM as a base in methanol media. Basic compounds such as alprenolol, synephrine, fenoterol, metoxyphenamine, propranolol, etc. were used as tested compounds. The influence of the BGE composition and concentration of chiral selector, as well as other experimental parameters (temperature, applied voltage), on the enantioseparation of analytes was investigated.
2 Materials and methods 2.1 Chemicals and reagents Boric (H3 BO3 ), citric (CA) and acetic acids (AA), sodium hydroxide, and all chiral compounds (alprenolol, atenolol, clenbuterol, metoprolol, pindolol, propranolol hydrochloride, sotalol, synephrine, fenoterol, methoxyphenamine, labetalol, tetrahydrozoline, epinephrine, salmeterol, octopamine, chlorpheniramine, isoproterenol, doxylamine) were provided by Sigma-Aldrich (St. Louis, MO, USA). Triethylamine (TEA) and tributylamine (TBA) were provided by Acros Organics (NJ, USA). Chiral selector with purity 98% was purchased from (Xi’an Lijun Pharmaceutical Co., Shanxi, China). Methanol (MeOH) of HPLC grade was purchased from LabScan (Gliwice, Poland). C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Electrophoresis 2014, 35, 2759–2764
2.2 Instrumentation Separations were performed using a Capel-105M CE system (Lumex, St. Petersburg, Russia) equipped with a UV-detector (190–380 nm) and 7100 CE equipped with a DAD (Agilent Technologies, Germany). Uncoated fused-silica capillaries (50 m id × 45 cm (35.5 cm to the detector) and 50 m id × 48.5 cm (40 cm to the detector), 363 m od) were purchased from Polymicro Technologies (Phoenix, AZ, USA). All chiral compounds were detected at 225 nm and temperature was set at 20°C. All samples were introduced using pressure injection (2.5 kPa, 15 s). Positive polarity was used. All measurements were repeated three times.
2.3 Procedure All the chiral analytes were dissolved in methanol (1 mg/mL) and diluted with methanol to the lower concentration (0.1 mg/mL). The BGEs were prepared daily from stock solution of NaOH, TEA, TBA, boric acid, CA, or AA in methanol by mixing the required volume of each component and diluting the mixture with MeOH. The BGEs were prepared by dissolving CLM in a BGE followed by filtration through a 0.22 m membrane filter. To provide the initial information about possible CLM concentration, its solubility in MeOH was determined by visual observation after 3 min of sonication. Acetone was used as an EOF marker. One sample studied was Anaprilin tablets (JSC Synthesis, Kurgan, Russian Federation) which contained propranolol hydrochloride. The powdered mass of ten tablets (100 mg) were dissolved in MeOH (10 mL), sonicated and filtered. The obtained solution was diluted 20 times with MeOH and then analyzed. Solutions for calibration were prepared by dissolving stock solution of propranolol hydrochloride in MeOH. The sample preparation of Vasocardin tablets containing metoprolol tartrate (Zentiva, Slovak Republic) was generally the same. New capillary was flushed with MeOH (30 min), 50 mM NaOH in MeOH (30 min), MeOH (30 min), and BGE (30 min). Before work, the capillary was flushed daily with methanol and BGE for 10 min and 15 min, respectively, and after work, it was washed with methanol (10 min). To improve the reproducibility, between runs the capillary was washed with 0.2 M CA in methanol (3 min), methanol (3 min), and BGE (5 min). It is preferable to use the freshly prepared solution because of the solvent evaporation.
3 Results and discussion CLM is a semisynthetic macrolide antibiotic, erythromycin derivative. The structure of CLM (Fig. 1) includes a macrocyclic 14 carbon atom lactone ring with several functional groups (two sugar moieties, several hydroxyl groups, two methoxy group, and one tertiary amino group). Methoxy group in position 6 of the lactone ring increased resistance to acids as compared with erythromycin [22]. Numerous functional groups of the antibiotic can undergo multiple www.electrophoresis-journal.com
Electrophoresis 2014, 35, 2759–2764
CE and CEC
2761
gration time is due to the protonation of metoprolol molecule. This fact could be explained by possible CLM reacting with boric acid to form a CLM–boric acid complex, which participated in enantiorecognition. This is unlikely under such acidic conditions. However, study of this phenomenon in more detail is essential for better understanding.
3.1.2 Effect of the nature of base additive and its concentration
Figure 1. Structure of clarithromycin.
interactions (including hydrogen bonding, charge-to-charge, hydrophobic, steric interactions) with analyte, enabling separation of enantiomers. CLM has no conjugated double bonds in the lactone ring; hence, significant UV absorbance is only observed at wavelengths below 210 nm. CLM in a base form practically insoluble in water, ACN, soluble in acetone and in methylene chloride, slightly soluble (up to about 90 mM) in methanol, but as a salt (citrate or acetate) it can be easily dissolved in methanol. Therefore, methanol was used as a solvent.
3.1 Effect of the BGE 3.1.1 Effect of the acid additive and its concentration In order to select the BGE composition, the experiments were started with a BGE made up of 75 mM TBA and 60 mM H3 BO3 . This BGE was successfully used previously with azithromycin [16]. Insufficient solubility of CLM (less than 65 mM) in the initial BGE dictated the necessity of addition of equimolar amount of an acid: acetic or citric. In case of AA, no enantioseparation was observed despite of high concentration (90 mM) of the chiral selector. The use of CA (60 mM) instead of AA enabled getting two peaks of enantiomers of metoprolol (Rs 0.4), clenbuterol (Rs 1.1), and methoxyphenamine (Rs 1.0). Due to the good solubility of CLM in MeOH in the presence of CA, further experiments were carried out using CA additive in BGE. As in the case of azithromycin, boric acid has a great influence on enantioseparation [16]. The effect of boric acid on the separation of metoprolol enantiomers was studied in BGE which consisted of 100 mM CA, 20 mM NaOH, 60 mM CLM and addition of H3 BO3 (0–360 mM) in MeOH. Decreasing boric acid concentration lead to peak resolution decrease up to its complete disappearance at 0 mM H3 BO3 . With increasing of concentrations up to 240 mM, resolution between the peaks of the enantiomers increased and further remained almost constant. The migration time of metoprolol decreased with increasing boric acid concentrations up to 120 mM and then remained constant (Fig. 2). A significant decrease in mi C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
To investigate the influence the buffer cation, NaOH, TEA, or TBA was used as a base additive, while concentrations of boric acid and CA, and CLM were kept constant (240, 100, and 60 mM, respectively). Unlike enantioseparation with azithromycin as a chiral selector [16], no significant difference was found between the bases. For example, the resolution factors for atenolol enantiomers were 1.4, 1.2, and 1.4 (for 10 mM NaOH, 20 mM TEA, and 20 mM TBA, respectively). All of them also provided sharp and efficient peaks of enantiomers. Then, the effect of NaOH concentration was investigated in the range between 2 and 20 mM. The higher concentration of sodium hydroxide was limited by poor solubility of CLM. In the case of metoprolol, pindolol, and alprenolol concentration of sodium hydroxide did not much affect the resolution, while increasing the NaOH concentration from 10 to 20 mM, the decrease of resolution factor was observed (for atenolol, propranolol, sotalol, synephrine enantiomers, and fenoterol diastereomers). Therefore, the optimal concentration of NaOH was selected 10 mM. As other MAs, CLM can be adsorbed on the walls of the fused-silica capillaries causing peak tailing and poor repeatability of the separation [9]. The adsorption of CLM citrate was evaluated as it was suggested in [23]. Two BGEs were used: 30 mM H3 BO3 and 75 mM TBA and 10 mМ NaOH, 100 mM CA, and 240 mM H3 BO3 (both were in MeOH). In the first BGE, antibiotic migrated very close to EOF and its electrophoretic mobility ( = (7.0 ± 0.3) × 10−10 m2 V−1 s−1 ) seemed to be independent of its concentration in the studied range (0.5–60 mM). The peak shapes and their symmetry did not changed drastically. This showed low, if any, adsorption of CLM on the fused-capillary wall under the above conditions. However, tertiary amino group of CLM can be protonated in acidic solution and electrostatical interaction between silanol groups of the capillary. This was probably the case in the second BGE because constant shift of migration times was observed (RSD = 9.3%). Antibiotic migrated as a cation. The adsorption study revealed that its electrophoretic mobility decreased with the increase of the injected CLM quantity. To improve the repeatability of the migration times, it is recommended to include washing with CA (200 mM) between runs before washing with MeOH and BGE. This easy step significantly improved the obtained results. CLM electrophoretic mobility was found to be equal (1.1 ± 0.4) × 10−6 m2 V−1 s−1 and the repeatability of the migration times of analytes was about 1.6%. www.electrophoresis-journal.com
2762
M. V. Lebedeva et al.
Electrophoresis 2014, 35, 2759–2764
Figure 2. Electropherograms showing effect of boric acid on enantioseparation of metoprolol. Conditions: BGE, A—0, B—60, C—120, D—240, E—360 mM H3 BO3 ; 100 mM citric acid, 20 mМ NaOH, 60 mM CLM.
Figure 3. Effect of BGE concentration on resolution of racemic drugs. Conditions: BGE, c (citric acid)/c (NaOH) = 5/1, 240 mM H3 BO3 , 60 mM CLM in MeOH.
3.2 Effect of the BGE concentration The influence of the BGE concentration (total concentration of CA and sodium hydroxide, at a constant ratio c (CA)/c (NaOH) = 5/1 and c (H3 BO3 ) = 240 mM) on the electromigration parameters was studied. Figure 3 displays the obtained dependence of resolution factor for some compounds in different BGEs. The best enantioseparation occurred when 120 mM citrate solution was used. Partial enantioseparation of sotalol (Rs 0.5) was achieved only at this buffer concentration. This BGE also provided the fastest analysis time. In addition to the data presented in Fig. 3, atenolol, pindolol, labetalol, fenoterol, and sotalol were enantioseparated too with the maximum peak resolution at 120 mM.
3.3 Effect of the chiral selector concentration The concentration of CLM is a parameter that is easy to vary; meanwhile it has a profound effect on both migration times and peak resolution of an analyte. In general, the resolution improved with the increase of chiral selector concentration. In our work, the CLM concentration was varied in the range of 40–90 mM. The migration times were generally increasing as the CLM concentration increased. At 40 mM of CLM, no enantioseparation for sotalol, synephrine, labetalol, and C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 4. Effect of CLM concentration on peak resolution of racemic drugs. Conditions: BGE, 100 mM citric acid, 10 mM NaOH, 240 mM H3 BO3 with addition of CLM in MeOH.
fenoterol was observed and for all other compounds resolution factor lower than 0.5. As shown in Fig. 4, the resolution of analytes increased with increasing of CLM concentration up to the maximum at 60–75 mM, then decreased or remained the same. For example, the maximum resolution of sotalol enantiomers, fenoterol, and labetalol diastereomers was observed at 75, 75, and 60 mM CLM, indicating low enantioselectivity of the antibiotic toward them under these conditions. For metoprolol and pindolol, the optimum concentration seemed to be higher than 90 mM, however peak broadening did not allowed further increase of CLM. Enantiomers of octopamine, epinephrine, salmeterol, chlorpheniramine, isoproterenol, and doxylamine were not enantioseprated during this work. Labetalol which possesses four stereo centers was only partially resolved into two peaks. Finally, BGE consisted of 100 mM CA, 10 mM NaOH, 240 mM H3 BO3 in MeOH with addition of 60–75 mM CLM was chosen to be optimal and used in further experiments. When compared to the aqueous-organic BGE containing CLM [14], although higher chiral selector concentration was required for some compounds, higher separation efficiency was achieved (⬎8 × 104 ) and more analytes were enantioseparated. Nonaqueous media enabled enantioseparating of alprenolol, clenbuterol, methoxyphenamine, synephrine, and pindolol (Fig. 5). It is interesting to note that under studied conditions of certain molecules (e.g. methoxyphenamine) are less separated than that of molecules having large substituents in the www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2014, 35, 2759–2764
2763
Figure 5. Electropherograms showing enantioseparation of tested compounds: A—atenolol, B*—metoprolol, C—pindolol, D—alprenolol, E—propranolol, F—sotalol, G—synephrine, H—fenoterol, I**—labetalol. Conditions: 75 mM claritrotomycin (*—90 mM, **—60 mM), 240 mM H3 BO3 , 100 mM CA, 10 mM NaOH in MeOH; +20 kV, 225 nm. Compounds: J—metoprolol, K—clenbuterol, L—methoxyphenamine. Conditions: 60 mM clarithromycin, 75 mM TBA, 60 mM H3 BO3 , 60 mM CA in methanol
aromatic ring like alprenolol or pindolol. Some of analytes (alprenolol, pindolol, sotalol, synephrine, clenbuterol, fenoterol, methoxyphenamine) were not previously enantioseparated using macrolides. 3.4 Effect of the applied voltage and temperature Effect of the applied voltage and temperature on the separation parameters was studied for alprenolol and propranolol. The increase in applied voltage from +10 kV to +25 kV brought shorter migration time and higher efficiency. Peak resolution was improved up to +20 kV and then remained constant. A voltage of 20 kV was chosen as optimal. To evaluate possible voltage to be applied, an Ohm’s graph was plotted (data not shown). Current did not exceed 12 A at 25 kV. Based on that, it can be concluded that a single BGE can be used for 12–15 analyses and evaporation of methanol seems to affect stronger on the repeatability. Temperature effect was not very noticeable. An increase in temperature from 20 to 30°C (5°C increment) brought about a slight decrease (about 1 min) in migration time of all the analytes due to slight viscosity decrease of the BGE. Resolution factor hardly depended on temperature. The increase in temperature led to the very slight decrease in resolution by a factor of 0.1–0.2 only. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.5 Quantification Finally, as an example for illustration the potential application, propranolol and metoprolol enantiomers were quantified (both in different tablets). The investigated samples were analyzed in BGE which consisted of 240 mM H3 BO3 , 10 mM NaOH, 100 mM CA in MeOH with 60 mM CLM. Calibration curves of the corrected peak area (area/migration time) of propranolol were found to be linear in the range 2.5 × 10−2 –3.0 × 10−1 mg/mL. The correlation coefficients, r2 , were determined as 0.9991 and 0.9990 for the first and the second enantiomer, respectively. The LODs (3 × S/N) were determined as 2.6 × 10−3 and 2.8 × 10−3 mg/mL for the first and the second enantiomers, respectively. The intraday precision (RSD) for the corrected peak areas (S/tm ) was 2.0%. Calibration curves of the corrected peak area (area/migration time) of metoprolol were found to be linear in the range 1.0 × 10−2 –1.6 × 10−1 mg/mL. The correlation coefficients, r2 , were determined as 0.9981 and 0.9979 for the first and the second enantiomer, respectively. The LODs (3 × S/N) were determined as 2.8 × 10−3 and 3.0 × 10−3 mg/mL for the first and the second enantiomers, respectively. The intraday precision (RSD) for the corrected peak areas (S/tm ) was 1.1 and 1.2%. Good agreement of the found and declared values www.electrophoresis-journal.com
2764
M. V. Lebedeva et al.
was demonstrated. Both tablets contained the racemate of -blockers. The content of metoprolol was found to be (47.7 ± 0.9) milligrams per tablet (declared 50 mg). The content of propranolol (the sum of both enantiomers) was (40.3 ± 0.9) milligrams per tablet (declared 40 mg).
4 Concluding remarks CLM was used as a chiral selector in NACE for enantioseparation (partial or baseline) of amines and especially amino alcohols (alprenolol, atenolol, metoprolol, pindolol, propranolol, sotalol, synephrine, labetalol, clenbuterol, fenoterol, methoxyphenamine). Among all the studied BGEs, the combination of boric acid and CA with NaOH in MeOH gave rise to the best enantioresolution. To better understand the separation mechanism, complexation between CLM and boric should be thoroughly investigated as well as the enantiomer migration order. These studies are in progress in our group. The BGE composition, chiral selector concentration, temperature, and applied voltage were optimized and these optimized conditions were used for the determination of propranolol and metoprolol in tablets. The work demonstrated simple and reproducible procedure for enantioseparation of amines and amino alcohols. This work was supported by Russian Foundation for Basic Research (grant nos. 12-03-31255 and 12-03-00405). The authors acknowledge Agilent Technologies for providing 7100 CE instrument. The authors have declared no conflicts of interest.
5 References
Electrophoresis 2014, 35, 2759–2764
[5] Servais, A.-C., Rousseau, A., Fillet, M., Lomsadze, K., Salgado, A., Crommen, J., Chankvetadze, B., Electrophoresis 2010, 31, 1467–1474. [6] Chankvetadze, L., Servais, A.-C., Fillet, M., Salgado, A., Crommen, J., Chankvetadze, B., J. Chrom. A 2012, 1267, 1206–1216. [7] Servais, A.-C., Fillet, M., Methods Mol. Biol. 2013, 970, 297–305. [8] Wang, L.-J., Liu, X.-F., Lu, Q.-N., Yang, G.-L., Chen, X.-G., J. Chromatogr. A 2013, 1284, 188–193. [9] Dixit, S., Park, J. H., Biomed. Chromatogr. 2014, 28, 10–26. [10] Hou, J. G., He, T. X., Han, X. A., Du, X. Z., Mao, X. F., Deng, H. L., Gao, J. Z., Chin. Chem. Lett. 2002, 3, 280–282. [11] Hou, J. G., He, T. X., Mao, X. F., Du, X. Z., Deng, H. L., Gao, J. Z., Anal. Lett. 2003, 36, 1437–1449. [12] Xu, G., Du, Y., Chen, B., Chen, J., Chromatographia 2010, 72, 289–295. [13] Ha, P. T., Hoogmartens, J., Van Schepdael, A., J. Pharm. Biomed. Anal. 2006, 41, 1–11. [14] Yu, T., Du, Y., Chen, B., Electrophoresis 2011, 32, 1898–1905. [15] Kumar, A. P., Park, J. H., J. Chromatogr. A 2011, 1218, 1314–1317. [16] Lebedeva, M. V., Bulgakova, G. A., Prokhorova, A. F., Shapovalova, E. N., Chernobrovkin, M. G., Shpigun, O. A., Chromatographia 2013, 76, 375–379. [17] Chen, B., Du, Y., Electrophoresis 2010, 1217, 1806–1812. ˇ [18] Maier, V., Ranc, V., Svidrnoch, ˇ ık, J., M., Petr, J., Sevc´ ´ E., Armstrong, D. W., J. Chrom. A 2012, 1237, Tesaˇrova, 128–132. ´ ´ F., J. Biochem. Biophys. [19] Gagyi, L., Gyeresi, A., Kilar, Methods. 2008, 70, 1268–75.
[1] Ward, T., Baker, B. A., Anal. Chem. 2008, 80, 4363–4372.
[20] Servais, A.-C., Fillet, M., Chiap, P., Dewe, W., Hubert, P., Crommen, J., Electrophoresis 2004, 25, 2701–2710.
[2] Ali, I., Sanagi, M. M., Aboul-Enein, H. Y., Electrophoresis 2014, 35, 926–936.
[21] Chen J., Du Y., Zhu, F., Chen B., J. Chromatogr. A 2010, 1217, 7158–7163.
[3] Teng, H., Yuan, B.-Q., You, T.-Y., Chin. J. Anal. Chem. 2010, 38, 1670–1677.
[22] Zuckerman, J. M., Infect. Dis. Clin. N. Am. 2004, 18, 621–649.
[4] Chankvetadze, B., Blaschke, G., Electrophoresis 2000, 21, 4159–4178.
[23] Fang, N., Zhang, H., Li, J., Li, H. W., Yeung, E. S. Anal. Chem. 2007, 79, 6047–6054.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.electrophoresis-journal.com
Electrophoresis 2014, 35, 2759–2764
Margarita V. Lebedeva Aleksandra F. Prokhorova Elena N. Shapovalova Oleg A. Shpigun Chemistry Department, Lomonosov Moscow State University, Moscow, Russian Federation
Received March 14, 2014 Revised July 8, 2014 Accepted July 10, 2014
Research Article
Clarithromycin as a chiral selector for enantioseparation of basic compounds in nonaqueous capillary electrophoresis The first use of macrolide antibiotic clarithromycin (CLM) in nonaqueous media for enantioseparation (partial or baseline) of the following compounds: alprenolol, atenolol, metoprolol, clenbuterol, methoxyphenamine, pindolol, propranolol, sotalol, synephrine, labetalol, and fenoterol is reported. Each analysis took less than 15 min. To find optimal separation conditions, some properties of CLM (adsorption, solubility), as well as the effect of experimental parameters on the enantioseparation of analytes (background electrolyte composition, chiral selector concentration, temperature, and applied voltage) were studied. The best chiral resolution was achieved in methanolic solution of 100 mM citric acid, 10 mM NaOH, 240–300 mM H3 BO3 , and 60–75 mM CLM. Using the proposed procedure, adsorption of CLM on the capillary wall was negligible and the repeatability of the migration times (RSD) was as good as 1.6%. For the analysis of propranolol, the linearity was achieved in the concentration range 2.5 × 10−2 – 3.0 × 10−1 mg/mL with the LODs (3 × S/N) being equal 2.6 × 10−3 mg/mL and 2.8 × 10−3 mg/mL for the first and the second enantiomers, respectively. Linear range for metoprolol enantiomers was 1.0 × 10−2 –1.6 × 10−1 mg/mL. The LODs (3 × S/N) were determined as 2.8 × 10−3 and 3.0 × 10−3 mg/mL for the first and the second enantiomers, respectively. Keywords: -Blockers / Chiral separation / Clarithromycin / Macrolides / Nonaqueous capillary electrophoresis DOI 10.1002/elps.201400135
1 Introduction Analysis of chiral compounds is performed by HPLC, CE, supercritical fluid chromatography, TLC, and GC [1]. CE is often the method of choice for chiral analyses due to its improved efficiency and resolution. The fact that chiral selector is dissolved in a BGE enables a rapid method development and reduced cost of both the analysis and the method development. Nonaqueous CE (NACE) is a powerful tool for chiral analysis due to the higher solubility and different selectivity of some chiral selectors in a nonaqueous environment, and also flexibility of the technique [1–4]. Lower conductivity of electrolytes in NACE allows higher voltage to apply and higher buffer concentration to use. The enantioseparation mechanism in nonaqueous media differs from that in aqueous solutions resulting in the enantiomer migration order reversal as it was shown for heptakis (2,3-dimethyl-6-sulfo)Correspondence: Dr. Aleksandra F. Prokhorova, Chemistry Department, Lomonosov Moscow State University, Leninskie Gory, d.1, str.3, Moscow, 119991, Russian Federation E-mail: [email protected] Fax: +7-495-939-46-75
Abbreviations: AA, acetic acid; CA, citric acid; CLM, clarithromycin; MA, macrocyclic antibiotic; MeOH, methanol; NACE, nonaqueous CE; TBA, tributylamine; TEA, triethylamine C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
-CD [5, 6]. In addition, formation of external complexes between propranolol and heptakis (2,3-dimethyl-6-sulfo)-CD instead of inclusion complexes is observed in methanolic BGEs [6]. However, when compared NACE with aqueous CE, the reported chiral selectors and selectands are fewer. To date, different types of cyclodextrins [7], macrocyclic antibiotics, ion-pairing agents [8], and some other are used for chiral separations in CE. Macrocyclic glycopeptide antibiotics such as vancomycin, balhimycin, eremomycin provide high enantioselectivity in aqueous buffers but, unfortunately, they tend to adsorb on the capillary wall [9]. They also strongly absorb in the UV-region, therefore, significantly decreasing the detection sensitivity. From this point of view, macrolides, another type of macrocyclic antibiotics, are promising selectors for enantioseparation in nonaqueous BGEs because they are insoluble in water; they do not have significant absorbance in UV-region, while possessing many stereogenic centers [10–18]. The application spectrum of macrolide is still not as large as it could be. Only five papers on their use in aqueous CE have been published, including the first report of macrolide antibiotic, erythromycin, as a chiral selector in 2002 [10, 11] for enantioseparation of antihepatitis drugs (biphenyldimethylester derivatives). It was then used for baseline enantioseparation of amino alcohols (propranolol, duloxetine) and N,N-dimethyl-3-(2methoxyphenoxy)-3-propyl amine [14]. The attempt to use erythromycin and its different derivatives to enantioseparate
www.electrophoresis-journal.com
2760
M. V. Lebedeva et al.
aromatic carboxylic acids, some neutral compounds (warfarin), and amines with low concentration of antibiotic in BGE were not successful [13]. Clarithromycin (CLM) was once [12] used as lactobionate in the mixture of 12.5 mM phosphate buffer (pH 7.3) and methanol (50/50 v/v%) and showed enantioselectivity toward basic compounds such as amines and amino alcohols. In the above paper, the role of the lactobionate anion was not discussed. Also were published four papers about use of macrolides in NACE. Azithromycin dissolved in the mixture of ACN and methanol was beneficial for enantioseparation of carvedilol, darifenacine, cetirizine, citalopram, sertraline as well as tryptophan [15]. According to another paper [16], this antibiotic could resolve enantiomers of tetrahydrozoline and methoxyphenamine. Eryrthromycin lactobionate could resolve enantiomers of propranolol and duloxetine [17], and boromycin was successfully used for enantioseparation of tryptophanol, norepinephrine, octopamine, p-hydroxynorephedrine, 2-amino-1-phenylethanol, and ␣-benzylamine [18]. The possibilities of this type of antibiotics as chiral selectors are not very well established and, therefore, the intensive study of their enantioseparation capability is required. Many amines and amino alcohols have significant biological activity. Some of them are -blockers (atenolol, metoprolol, propranolol, etc.), salbutamol is sympathomimetic, methoxyphenamine and clenbuterol are bronchodilators, while synephrine is vasoconstrictor. Enantioseparation of these compounds have been performed by different separations techniques including both aqueous and NACE. Different chiral selectors were used (cyclodextrins, other polysaccharides, including combination of two selectors) [5,6,19–21]. The aim of this work was to evaluate enantioselective ability of the macrolide antibiotic CLM as a base in methanol media. Basic compounds such as alprenolol, synephrine, fenoterol, metoxyphenamine, propranolol, etc. were used as tested compounds. The influence of the BGE composition and concentration of chiral selector, as well as other experimental parameters (temperature, applied voltage), on the enantioseparation of analytes was investigated.
2 Materials and methods 2.1 Chemicals and reagents Boric (H3 BO3 ), citric (CA) and acetic acids (AA), sodium hydroxide, and all chiral compounds (alprenolol, atenolol, clenbuterol, metoprolol, pindolol, propranolol hydrochloride, sotalol, synephrine, fenoterol, methoxyphenamine, labetalol, tetrahydrozoline, epinephrine, salmeterol, octopamine, chlorpheniramine, isoproterenol, doxylamine) were provided by Sigma-Aldrich (St. Louis, MO, USA). Triethylamine (TEA) and tributylamine (TBA) were provided by Acros Organics (NJ, USA). Chiral selector with purity 98% was purchased from (Xi’an Lijun Pharmaceutical Co., Shanxi, China). Methanol (MeOH) of HPLC grade was purchased from LabScan (Gliwice, Poland). C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Electrophoresis 2014, 35, 2759–2764
2.2 Instrumentation Separations were performed using a Capel-105M CE system (Lumex, St. Petersburg, Russia) equipped with a UV-detector (190–380 nm) and 7100 CE equipped with a DAD (Agilent Technologies, Germany). Uncoated fused-silica capillaries (50 m id × 45 cm (35.5 cm to the detector) and 50 m id × 48.5 cm (40 cm to the detector), 363 m od) were purchased from Polymicro Technologies (Phoenix, AZ, USA). All chiral compounds were detected at 225 nm and temperature was set at 20°C. All samples were introduced using pressure injection (2.5 kPa, 15 s). Positive polarity was used. All measurements were repeated three times.
2.3 Procedure All the chiral analytes were dissolved in methanol (1 mg/mL) and diluted with methanol to the lower concentration (0.1 mg/mL). The BGEs were prepared daily from stock solution of NaOH, TEA, TBA, boric acid, CA, or AA in methanol by mixing the required volume of each component and diluting the mixture with MeOH. The BGEs were prepared by dissolving CLM in a BGE followed by filtration through a 0.22 m membrane filter. To provide the initial information about possible CLM concentration, its solubility in MeOH was determined by visual observation after 3 min of sonication. Acetone was used as an EOF marker. One sample studied was Anaprilin tablets (JSC Synthesis, Kurgan, Russian Federation) which contained propranolol hydrochloride. The powdered mass of ten tablets (100 mg) were dissolved in MeOH (10 mL), sonicated and filtered. The obtained solution was diluted 20 times with MeOH and then analyzed. Solutions for calibration were prepared by dissolving stock solution of propranolol hydrochloride in MeOH. The sample preparation of Vasocardin tablets containing metoprolol tartrate (Zentiva, Slovak Republic) was generally the same. New capillary was flushed with MeOH (30 min), 50 mM NaOH in MeOH (30 min), MeOH (30 min), and BGE (30 min). Before work, the capillary was flushed daily with methanol and BGE for 10 min and 15 min, respectively, and after work, it was washed with methanol (10 min). To improve the reproducibility, between runs the capillary was washed with 0.2 M CA in methanol (3 min), methanol (3 min), and BGE (5 min). It is preferable to use the freshly prepared solution because of the solvent evaporation.
3 Results and discussion CLM is a semisynthetic macrolide antibiotic, erythromycin derivative. The structure of CLM (Fig. 1) includes a macrocyclic 14 carbon atom lactone ring with several functional groups (two sugar moieties, several hydroxyl groups, two methoxy group, and one tertiary amino group). Methoxy group in position 6 of the lactone ring increased resistance to acids as compared with erythromycin [22]. Numerous functional groups of the antibiotic can undergo multiple www.electrophoresis-journal.com
Electrophoresis 2014, 35, 2759–2764
CE and CEC
2761
gration time is due to the protonation of metoprolol molecule. This fact could be explained by possible CLM reacting with boric acid to form a CLM–boric acid complex, which participated in enantiorecognition. This is unlikely under such acidic conditions. However, study of this phenomenon in more detail is essential for better understanding.
3.1.2 Effect of the nature of base additive and its concentration
Figure 1. Structure of clarithromycin.
interactions (including hydrogen bonding, charge-to-charge, hydrophobic, steric interactions) with analyte, enabling separation of enantiomers. CLM has no conjugated double bonds in the lactone ring; hence, significant UV absorbance is only observed at wavelengths below 210 nm. CLM in a base form practically insoluble in water, ACN, soluble in acetone and in methylene chloride, slightly soluble (up to about 90 mM) in methanol, but as a salt (citrate or acetate) it can be easily dissolved in methanol. Therefore, methanol was used as a solvent.
3.1 Effect of the BGE 3.1.1 Effect of the acid additive and its concentration In order to select the BGE composition, the experiments were started with a BGE made up of 75 mM TBA and 60 mM H3 BO3 . This BGE was successfully used previously with azithromycin [16]. Insufficient solubility of CLM (less than 65 mM) in the initial BGE dictated the necessity of addition of equimolar amount of an acid: acetic or citric. In case of AA, no enantioseparation was observed despite of high concentration (90 mM) of the chiral selector. The use of CA (60 mM) instead of AA enabled getting two peaks of enantiomers of metoprolol (Rs 0.4), clenbuterol (Rs 1.1), and methoxyphenamine (Rs 1.0). Due to the good solubility of CLM in MeOH in the presence of CA, further experiments were carried out using CA additive in BGE. As in the case of azithromycin, boric acid has a great influence on enantioseparation [16]. The effect of boric acid on the separation of metoprolol enantiomers was studied in BGE which consisted of 100 mM CA, 20 mM NaOH, 60 mM CLM and addition of H3 BO3 (0–360 mM) in MeOH. Decreasing boric acid concentration lead to peak resolution decrease up to its complete disappearance at 0 mM H3 BO3 . With increasing of concentrations up to 240 mM, resolution between the peaks of the enantiomers increased and further remained almost constant. The migration time of metoprolol decreased with increasing boric acid concentrations up to 120 mM and then remained constant (Fig. 2). A significant decrease in mi C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
To investigate the influence the buffer cation, NaOH, TEA, or TBA was used as a base additive, while concentrations of boric acid and CA, and CLM were kept constant (240, 100, and 60 mM, respectively). Unlike enantioseparation with azithromycin as a chiral selector [16], no significant difference was found between the bases. For example, the resolution factors for atenolol enantiomers were 1.4, 1.2, and 1.4 (for 10 mM NaOH, 20 mM TEA, and 20 mM TBA, respectively). All of them also provided sharp and efficient peaks of enantiomers. Then, the effect of NaOH concentration was investigated in the range between 2 and 20 mM. The higher concentration of sodium hydroxide was limited by poor solubility of CLM. In the case of metoprolol, pindolol, and alprenolol concentration of sodium hydroxide did not much affect the resolution, while increasing the NaOH concentration from 10 to 20 mM, the decrease of resolution factor was observed (for atenolol, propranolol, sotalol, synephrine enantiomers, and fenoterol diastereomers). Therefore, the optimal concentration of NaOH was selected 10 mM. As other MAs, CLM can be adsorbed on the walls of the fused-silica capillaries causing peak tailing and poor repeatability of the separation [9]. The adsorption of CLM citrate was evaluated as it was suggested in [23]. Two BGEs were used: 30 mM H3 BO3 and 75 mM TBA and 10 mМ NaOH, 100 mM CA, and 240 mM H3 BO3 (both were in MeOH). In the first BGE, antibiotic migrated very close to EOF and its electrophoretic mobility ( = (7.0 ± 0.3) × 10−10 m2 V−1 s−1 ) seemed to be independent of its concentration in the studied range (0.5–60 mM). The peak shapes and their symmetry did not changed drastically. This showed low, if any, adsorption of CLM on the fused-capillary wall under the above conditions. However, tertiary amino group of CLM can be protonated in acidic solution and electrostatical interaction between silanol groups of the capillary. This was probably the case in the second BGE because constant shift of migration times was observed (RSD = 9.3%). Antibiotic migrated as a cation. The adsorption study revealed that its electrophoretic mobility decreased with the increase of the injected CLM quantity. To improve the repeatability of the migration times, it is recommended to include washing with CA (200 mM) between runs before washing with MeOH and BGE. This easy step significantly improved the obtained results. CLM electrophoretic mobility was found to be equal (1.1 ± 0.4) × 10−6 m2 V−1 s−1 and the repeatability of the migration times of analytes was about 1.6%. www.electrophoresis-journal.com
2762
M. V. Lebedeva et al.
Electrophoresis 2014, 35, 2759–2764
Figure 2. Electropherograms showing effect of boric acid on enantioseparation of metoprolol. Conditions: BGE, A—0, B—60, C—120, D—240, E—360 mM H3 BO3 ; 100 mM citric acid, 20 mМ NaOH, 60 mM CLM.
Figure 3. Effect of BGE concentration on resolution of racemic drugs. Conditions: BGE, c (citric acid)/c (NaOH) = 5/1, 240 mM H3 BO3 , 60 mM CLM in MeOH.
3.2 Effect of the BGE concentration The influence of the BGE concentration (total concentration of CA and sodium hydroxide, at a constant ratio c (CA)/c (NaOH) = 5/1 and c (H3 BO3 ) = 240 mM) on the electromigration parameters was studied. Figure 3 displays the obtained dependence of resolution factor for some compounds in different BGEs. The best enantioseparation occurred when 120 mM citrate solution was used. Partial enantioseparation of sotalol (Rs 0.5) was achieved only at this buffer concentration. This BGE also provided the fastest analysis time. In addition to the data presented in Fig. 3, atenolol, pindolol, labetalol, fenoterol, and sotalol were enantioseparated too with the maximum peak resolution at 120 mM.
3.3 Effect of the chiral selector concentration The concentration of CLM is a parameter that is easy to vary; meanwhile it has a profound effect on both migration times and peak resolution of an analyte. In general, the resolution improved with the increase of chiral selector concentration. In our work, the CLM concentration was varied in the range of 40–90 mM. The migration times were generally increasing as the CLM concentration increased. At 40 mM of CLM, no enantioseparation for sotalol, synephrine, labetalol, and C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 4. Effect of CLM concentration on peak resolution of racemic drugs. Conditions: BGE, 100 mM citric acid, 10 mM NaOH, 240 mM H3 BO3 with addition of CLM in MeOH.
fenoterol was observed and for all other compounds resolution factor lower than 0.5. As shown in Fig. 4, the resolution of analytes increased with increasing of CLM concentration up to the maximum at 60–75 mM, then decreased or remained the same. For example, the maximum resolution of sotalol enantiomers, fenoterol, and labetalol diastereomers was observed at 75, 75, and 60 mM CLM, indicating low enantioselectivity of the antibiotic toward them under these conditions. For metoprolol and pindolol, the optimum concentration seemed to be higher than 90 mM, however peak broadening did not allowed further increase of CLM. Enantiomers of octopamine, epinephrine, salmeterol, chlorpheniramine, isoproterenol, and doxylamine were not enantioseprated during this work. Labetalol which possesses four stereo centers was only partially resolved into two peaks. Finally, BGE consisted of 100 mM CA, 10 mM NaOH, 240 mM H3 BO3 in MeOH with addition of 60–75 mM CLM was chosen to be optimal and used in further experiments. When compared to the aqueous-organic BGE containing CLM [14], although higher chiral selector concentration was required for some compounds, higher separation efficiency was achieved (⬎8 × 104 ) and more analytes were enantioseparated. Nonaqueous media enabled enantioseparating of alprenolol, clenbuterol, methoxyphenamine, synephrine, and pindolol (Fig. 5). It is interesting to note that under studied conditions of certain molecules (e.g. methoxyphenamine) are less separated than that of molecules having large substituents in the www.electrophoresis-journal.com
CE and CEC
Electrophoresis 2014, 35, 2759–2764
2763
Figure 5. Electropherograms showing enantioseparation of tested compounds: A—atenolol, B*—metoprolol, C—pindolol, D—alprenolol, E—propranolol, F—sotalol, G—synephrine, H—fenoterol, I**—labetalol. Conditions: 75 mM claritrotomycin (*—90 mM, **—60 mM), 240 mM H3 BO3 , 100 mM CA, 10 mM NaOH in MeOH; +20 kV, 225 nm. Compounds: J—metoprolol, K—clenbuterol, L—methoxyphenamine. Conditions: 60 mM clarithromycin, 75 mM TBA, 60 mM H3 BO3 , 60 mM CA in methanol
aromatic ring like alprenolol or pindolol. Some of analytes (alprenolol, pindolol, sotalol, synephrine, clenbuterol, fenoterol, methoxyphenamine) were not previously enantioseparated using macrolides. 3.4 Effect of the applied voltage and temperature Effect of the applied voltage and temperature on the separation parameters was studied for alprenolol and propranolol. The increase in applied voltage from +10 kV to +25 kV brought shorter migration time and higher efficiency. Peak resolution was improved up to +20 kV and then remained constant. A voltage of 20 kV was chosen as optimal. To evaluate possible voltage to be applied, an Ohm’s graph was plotted (data not shown). Current did not exceed 12 A at 25 kV. Based on that, it can be concluded that a single BGE can be used for 12–15 analyses and evaporation of methanol seems to affect stronger on the repeatability. Temperature effect was not very noticeable. An increase in temperature from 20 to 30°C (5°C increment) brought about a slight decrease (about 1 min) in migration time of all the analytes due to slight viscosity decrease of the BGE. Resolution factor hardly depended on temperature. The increase in temperature led to the very slight decrease in resolution by a factor of 0.1–0.2 only. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.5 Quantification Finally, as an example for illustration the potential application, propranolol and metoprolol enantiomers were quantified (both in different tablets). The investigated samples were analyzed in BGE which consisted of 240 mM H3 BO3 , 10 mM NaOH, 100 mM CA in MeOH with 60 mM CLM. Calibration curves of the corrected peak area (area/migration time) of propranolol were found to be linear in the range 2.5 × 10−2 –3.0 × 10−1 mg/mL. The correlation coefficients, r2 , were determined as 0.9991 and 0.9990 for the first and the second enantiomer, respectively. The LODs (3 × S/N) were determined as 2.6 × 10−3 and 2.8 × 10−3 mg/mL for the first and the second enantiomers, respectively. The intraday precision (RSD) for the corrected peak areas (S/tm ) was 2.0%. Calibration curves of the corrected peak area (area/migration time) of metoprolol were found to be linear in the range 1.0 × 10−2 –1.6 × 10−1 mg/mL. The correlation coefficients, r2 , were determined as 0.9981 and 0.9979 for the first and the second enantiomer, respectively. The LODs (3 × S/N) were determined as 2.8 × 10−3 and 3.0 × 10−3 mg/mL for the first and the second enantiomers, respectively. The intraday precision (RSD) for the corrected peak areas (S/tm ) was 1.1 and 1.2%. Good agreement of the found and declared values www.electrophoresis-journal.com
2764
M. V. Lebedeva et al.
was demonstrated. Both tablets contained the racemate of -blockers. The content of metoprolol was found to be (47.7 ± 0.9) milligrams per tablet (declared 50 mg). The content of propranolol (the sum of both enantiomers) was (40.3 ± 0.9) milligrams per tablet (declared 40 mg).
4 Concluding remarks CLM was used as a chiral selector in NACE for enantioseparation (partial or baseline) of amines and especially amino alcohols (alprenolol, atenolol, metoprolol, pindolol, propranolol, sotalol, synephrine, labetalol, clenbuterol, fenoterol, methoxyphenamine). Among all the studied BGEs, the combination of boric acid and CA with NaOH in MeOH gave rise to the best enantioresolution. To better understand the separation mechanism, complexation between CLM and boric should be thoroughly investigated as well as the enantiomer migration order. These studies are in progress in our group. The BGE composition, chiral selector concentration, temperature, and applied voltage were optimized and these optimized conditions were used for the determination of propranolol and metoprolol in tablets. The work demonstrated simple and reproducible procedure for enantioseparation of amines and amino alcohols. This work was supported by Russian Foundation for Basic Research (grant nos. 12-03-31255 and 12-03-00405). The authors acknowledge Agilent Technologies for providing 7100 CE instrument. The authors have declared no conflicts of interest.
5 References
Electrophoresis 2014, 35, 2759–2764
[5] Servais, A.-C., Rousseau, A., Fillet, M., Lomsadze, K., Salgado, A., Crommen, J., Chankvetadze, B., Electrophoresis 2010, 31, 1467–1474. [6] Chankvetadze, L., Servais, A.-C., Fillet, M., Salgado, A., Crommen, J., Chankvetadze, B., J. Chrom. A 2012, 1267, 1206–1216. [7] Servais, A.-C., Fillet, M., Methods Mol. Biol. 2013, 970, 297–305. [8] Wang, L.-J., Liu, X.-F., Lu, Q.-N., Yang, G.-L., Chen, X.-G., J. Chromatogr. A 2013, 1284, 188–193. [9] Dixit, S., Park, J. H., Biomed. Chromatogr. 2014, 28, 10–26. [10] Hou, J. G., He, T. X., Han, X. A., Du, X. Z., Mao, X. F., Deng, H. L., Gao, J. Z., Chin. Chem. Lett. 2002, 3, 280–282. [11] Hou, J. G., He, T. X., Mao, X. F., Du, X. Z., Deng, H. L., Gao, J. Z., Anal. Lett. 2003, 36, 1437–1449. [12] Xu, G., Du, Y., Chen, B., Chen, J., Chromatographia 2010, 72, 289–295. [13] Ha, P. T., Hoogmartens, J., Van Schepdael, A., J. Pharm. Biomed. Anal. 2006, 41, 1–11. [14] Yu, T., Du, Y., Chen, B., Electrophoresis 2011, 32, 1898–1905. [15] Kumar, A. P., Park, J. H., J. Chromatogr. A 2011, 1218, 1314–1317. [16] Lebedeva, M. V., Bulgakova, G. A., Prokhorova, A. F., Shapovalova, E. N., Chernobrovkin, M. G., Shpigun, O. A., Chromatographia 2013, 76, 375–379. [17] Chen, B., Du, Y., Electrophoresis 2010, 1217, 1806–1812. ˇ [18] Maier, V., Ranc, V., Svidrnoch, ˇ ık, J., M., Petr, J., Sevc´ ´ E., Armstrong, D. W., J. Chrom. A 2012, 1237, Tesaˇrova, 128–132. ´ ´ F., J. Biochem. Biophys. [19] Gagyi, L., Gyeresi, A., Kilar, Methods. 2008, 70, 1268–75.
[1] Ward, T., Baker, B. A., Anal. Chem. 2008, 80, 4363–4372.
[20] Servais, A.-C., Fillet, M., Chiap, P., Dewe, W., Hubert, P., Crommen, J., Electrophoresis 2004, 25, 2701–2710.
[2] Ali, I., Sanagi, M. M., Aboul-Enein, H. Y., Electrophoresis 2014, 35, 926–936.
[21] Chen J., Du Y., Zhu, F., Chen B., J. Chromatogr. A 2010, 1217, 7158–7163.
[3] Teng, H., Yuan, B.-Q., You, T.-Y., Chin. J. Anal. Chem. 2010, 38, 1670–1677.
[22] Zuckerman, J. M., Infect. Dis. Clin. N. Am. 2004, 18, 621–649.
[4] Chankvetadze, B., Blaschke, G., Electrophoresis 2000, 21, 4159–4178.
[23] Fang, N., Zhang, H., Li, J., Li, H. W., Yeung, E. S. Anal. Chem. 2007, 79, 6047–6054.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.electrophoresis-journal.com
