1

Electrophoresis 2014, 00, 1–8

Mariana Zuccherato Bocato1 Marcela Armelim Bortoleto2 ˆ Monica Tallarico Pupo1 Anderson Rodrigo Moraes de Oliveira2 1 Departamento

ˆ de Ciencias ˆ Farmaceuticas, Faculdade de ˆ ˆ Ciencias Farmaceuticas de ˜ Preto, Universidade de Ribeirao ˜ Paulo, Ribeirao ˜ Preto, SP, Sao Brasil 2 Departamento de Qu´ımica, ˆ Faculdade de Filosofia, Ciencias ˜ Preto, e Letras de Ribeirao ˜ Paulo, Universidade de Sao ˜ Preto, SP, Brasil Ribeirao

Received March 14, 2014 Revised June 9, 2014 Accepted June 17, 2014

Research Article

A new enantioselective CE method for determination of oxcarbazepine and licarbazepine after fungal biotransformation The present work describes, for the first time, the simultaneous separation of oxcarbazepine (OXC) and its active metabolite 10-hydroxy-10,11-dihydrocarbamazepine (licarbazepine, Lic) by chiral CE. The developed method was employed to monitor the enantioselective biotransformation of OXC into its active metabolite by fungi. The electrophoretic separations were performed using 10 mmol/L of a Tris-phosphate buffer solution (pH 2.5) containing 1% w/v of ␤-CD phosphate sodium salt (P-␤-CD) as running electrolyte, −20 kV of applied voltage and a 15°C capillary temperature. The method was linear over the concentration range of 1000–30 000 ng/mL for OXC and 75–900 ng/mL for each Lic enantiomer (r  0.9952). Within-day precision and accuracy evaluated by RSD and relative errors, respectively, were lower than 15% for all analytes. The validated method was used to evaluate the enantioselective biotransformation of OXC, mediated by fungi, into its active metabolite Lic. This study showed that the fungi Glomerella cingulata (VA1) and Beuveria bassiana were able to enantioselectively metabolize the OXC into Lic after 360 h of incubation. Biotransformation by the fungus Beuveria bassiana showed 79% enantiomeric excess for (S)-(+)-Lic, while VA1 gave an enantiomeric excess of 100% for (S)-(+)-Lic. This study opens a new route to the drug (S)-(+)-licarbazepine. Keywords: Chiral CE / Licarbazepine / Oxcarbazepine / Stereoselective fungal biotransformation DOI 10.1002/elps.201400137

1 Introduction Oxcarbazepine (OXC) (Fig. 1a) is a nonchiral antiepileptic drug used for the treatment of partial seizures [1]. This drug is an analogue of carbamazepine (CBZ), differing only in the keto group located at position-10 [2]. Both OXC and CBZ are enzyme-inducing antiepileptic drugs with the ability to increase or decrease serum concentrations of lipids in humans [3–5]. OXC has been reported to be preferable to CBZ, with respect to its effect on serum lipids in adults. OXC likewise presents better tolerability, more favorable pharmacokinetics and a better metabolic profile than CBZ [6]. A key difference between these two drugs is the fact that OXC is not metabolized to a toxic epoxide derivative metabolite, un-

Correspondence: Dr. Anderson Rodrigo Moraes de Oliveira, ˆ Departamento de Qu´ımica, Faculdade de Filosofia, Ciencias e Le˜ Preto, USP, Av. dos Bandeirantes, 3900, Ribeirao ˜ tras de Ribeirao Preto, SP 14040-901, Brasil E-mail: [email protected] Fax: +55-16-36024838

Abbreviations: CBZ, carbamazepine; CM-␤-CD, carboxymethyl-␤-CD; DS, degree of substitution; ee, enantiomeric excess; Lic, licarbazepine; MIC, minimal inhibitory concentration; OXC, oxcarbazepine; P-␤-CD, ␤-CD phosphate salt sodium; S-␤-CD, sulfated-␤-CD  C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

like CBZ [2]. The metabolism of OXC includes rapid presystemic reduction mediated by the hepatic cytosolic enzyme aryl ketone reductase, resulting in the formation of the chiral metabolite 10-hydroxy-10,11-dihydrocarbamazepine. This monohydroxy-derivative (also known as licarbazepine (Lic)) has a chiral center at position-10, existing as both the (S)(+)- and (R)-(−)-Lic enantiomers (Fig. 1B and C). While both enantiomers show similar antiepileptic effects [7–11], (S)-Lic is already marketed as a drug in the acetate form under the generic name eslicarbazepine acetate (or BIA 2–093). This drug was developed and patented in the mid-1990s by Bial (Portela e Co., S.A) [12]. As such, the ability to obtain enantiomerically pure Lic metabolites through alternative routes would be advantageous. The simultaneous stereoselective determination of OXC and its metabolite Lic by CE has not been reported. One prior report describes the use of capillary EKC to perform the chiral separation of Lic enantiomers in the absence of OXC or other analytes [13]. In this study, octakis-6-sulfo-␥ -CD was used to promote the enantioseparation, resulting in migration times of the Lic enantiomers between 10 and 13 min. Other studies have reported the analyses of OXC and its metabolites in biological matrices. Chiral separation of OXC and Lic in human plasma was achieved by LC-MS/MS [7, 14–16], and LC-UV was successfully used for the analyses of OXC and its metabolites in mouse tissue samples [17] and in mouse plasma and brain [18].

www.electrophoresis-journal.com

2

M. Z. Bocato et al.

Figure 1. Chemical structure of OXC (A), (S)-(+)-licarbazepine (B), (3) (R)-(−)-licarbazepine (C) and the internal standard, risperidone (D).

Microbial biotransformation is an alternative tool that enables molecular reactions mimicking mammalian metabolism [19]. Microbial biotransformation was initially used to obtain more active or less toxic metabolites, which could then be correlated with those obtained with in vivo or in vitro animal models [20]. The most common reactions observed in these processes with microorganisms are oxidations, reductions (similar to hepatic phase I in animal metabolism), and conjugation (phase II in animal metabolism) [19]. This methodology has the following advantages: (i) reduced animal demand, (ii) low cost (as it is unnecessary to isolate, purify, and stabilize the enzyme), (iii) mild incubation conditions, (iv) the possibility of obtaining higher concentrations of metabolites, facilitating structural elucidation and finally, (v) stereoselectivity [20–22]. Successful studies have demonstrated fungal biotransformation to be a useful pathway that is capable of drug metabolism mimicking mammalian systems [23–25]. Because Lic is a biologically active metabolite and (S)-Lic is already marketed as a drug, the use of fungi in a biotransformation process was evaluated to study the bioconversion of OXC into Lic enantiomers. Based on this, an enantioselective CE method for the simultaneous analysis of OXC and Lic in a liquid culture medium was developed and validated to reliably evaluate the enantioselective fungal biotransformation of OXC into Lic.

2 Materials and methods 2.1 Chemicals and reagents The solvent (HPLC grade) used in the sample preparation process was methyl-tert-butyl-ether from J.T. Baker (Philipsburg, PA, USA). Analytical grade sodium chloride, potassium chloride, potassium phosphate (KH2 PO4 ), magnesium sulfate (MgSO4 ·7H2 O), and iron sulfate (FeSO4 ·7H2 O) were all obtained from Merck (Darmstadt, Germany). The Tris(hydroximetyl)aminomethane and monosodium phosphate used to prepare the buffer solutions were purchased  C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Electrophoresis 2014, 00, 1–8

from J.T. Baker (Philipsburg, PA, USA). Sodium hydroxide and sodium nitrate were obtained from Nuclear (Diadema, Brazil). (Ortho)phosphoric acid was purchased from Vetec (Rio de Janeiro, Brazil) and acetic acid was obtained from Zilqu´ımica (Ribeir˜ao Preto, Brazil). Potato dextrose agar, sucrose, malt extract, dextrose, triptone, and yeast extract were obtained from Acumedia (Lansing, MI, USA). ␤-CD phosphate sodium salt, degree of substitution (DS)2–6, (P-␤-CD) was acquired from CycloLab (Budapest, Hungary). Sulfated␤-CD, DS5, (S-␤-CD) and carboxymethyl-␤-CD, DS3, (CM-␤-CD) were acquired from Sigma-Aldrich (Steinheim, Germany). All solutions used before electrophoretic analysis were filtered through a Millex-HV 0.45 ␮m disk filter from Millipore (Belford, MA, USA) and degassed by ultrasound for 5 min. Water used to prepare the solutions was purified using a Milli-Q plus system (Millipore, Bedford, USA). Standard stock solutions of OXC and rac-Lic (all purchased from Toronto Research Chem., Toronto, Canada) were prepared in methanol with concentrations of 5000 and 1000 ␮g/mL, respectively. Calibration curve solutions of OXC at concentrations ranging from 100 to 3000 ␮g/mL and rac-Lic ranging from 15 to 180 ␮g/mL were obtained by dilutions in the same solvent. A solution of risperidone (Toronto Research Chem., Toronto, Canada), prepared in methanol at a concentration of 100 ␮g/mL, was used as the internal standard. All solutions were stored at −20°C in amber glass tubes and protected from direct light. 2.2 Capillary electrophoresis Analyses were performed using CE instrumentation from Beckman Coulter (model P/ACE MDQ, Fullerton, CA, USA). The 32 KaratTM software package was used to control the instrument and to acquire data. An uncoated fused-silica capillary from Beckman Coulter with a 75 ␮m id, 30 cm total length, and 20 cm effective length was used. Before the first use, the capillary was conditioned by rinsing with aqueous 1.0 mol/L NaOH for 30 min, followed by water for 30 min. At the beginning of each working day, the capillary was rinsed with aqueous 0.1 mol/L NaOH for 10 min, water for 10 min, and the running buffer for 10 min. After each analysis, the capillary was rinsed with aqueous 0.1 mol/L NaOH for 2 min, water for 2 min, and running buffer for 3 min. After using the capillary, it was washed with aqueous 0.1 mol/L NaOH for 15 min followed by water for 15 min. The chiral electrophoretic separations were carried out in a 10 mmol/L phosphate buffer-Tris solution (pH adjusted to 2.5 with (ortho)phosphoric acid) containing 1% P-␤-CD (w/v). The capillary temperature was set at 15°C. The injections were performed hydrodynamically at a pressure of 0.5 psi for 8 s. A constant voltage of −20 kV was applied during the analyses. 2.3 Migration order of Licarbazepine enantiomers To establish the migration order of the Lic enantiomers, enantiomerically pure samples were isolated by HPLC. These www.electrophoresis-journal.com

Electrophoresis 2014, 00, 1–8

analytes were separated according to conditions previously described in the literature [11]. Briefly, the Lic enantiomers were separated on a Chiralcel OD column (250 mm × 4.6 mm, 10 ␮m particle size, acquired from Chiral Technologies, Exton, PA, USA) under isocratic conditions using a n-hexane-ethanol-2-propanol (18:2:1, v/v/v) plus glacial acetic acid (0.1%) as the mobile phase. After separation, each Lic enantiomer peak was collected, and the solvent was evaporated. The remaining residue was solubilized in water and analyzed by CE under the conditions described in Section 2.2. The migration order was established by comparing the migration and retention times of both studies.

2.4 Extraction procedure OXC and Lic were extracted from the liquid culture medium by a liquid–liquid extraction procedure. Aliquots of 2 mL of the liquid culture medium spiked with 20 ␮L of standard solutions of OXC, rac-Lic, and the internal standard (I.S.) were transferred to 10 mL glass tubes, alkalized with 0.5 mL of a 0.1 mol/L NaOH solution, and further treated with NaCl to achieve a 20% w/v ratio. The samples were then subjected to vortex agitation for 15 s, after which 5 mL of methyl-tert-butyl-ether were added. The tubes were capped R VRX agitator (IKA, Staufen, and agitated using a Vibrax Germany) at 1000 rpm for 30 min, followed by centrifugation at 1800 × g for 10 min at 10°C in a CF-15 centrifuge (Hitachi Koki, Kyoto, Japan). The organic layer (4 mL) was transferred to a 10 mL conical glass tube and the solvent was evaporated to dryness under a gentle stream of compressed air at room temperature. The remaining residues were dissolved in 150 ␮L of water, vortex-mixed for 10 s and analyzed by CE.

2.5 Method validation Because there is no specific guide specifying a standard procedure for analysis of drugs and metabolites in liquid culture medium, the EMA guidelines for analysis of drugs in biological matrices [26] were followed as closely as possible. Linearity was evaluated in quintuplicate. OXC and racLic were added in 2 mL of Czapek liquid culture medium in the concentration range of 1000–30 000 ng/mL and 75– 900 ng/mL, respectively. The correlation coefficient (r) was determined by least squares. The results were weighted by 1/x2 because the residual analysis of the analytical curve exhibited heteroscedastic behavior [27]. The F test for lack-of-fit (FLOF ) and a p value of 0.05 was employed for statistical analysis (MINITAB Release version of 14.1, State College, PA, EUA). The absolute recovery of each analyte was determined using calibration curves compiled from the data obtained from pure analytes. Samples of liquid culture medium were spiked with OXC concentrations of 3000, 15 000, and 22 500 ng/mL, and Lic enantiomer concentrations of 150, 450, and 675 ng/mL (n = 3, for each concentration). The LOQ was  C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

CE and CEC

3

defined as the lowest concentration of analyte that could be determined with accuracy and precision below 20% [26] over five analytical runs (n = 5). This value was obtained from liquid culture medium spiked with 1000 ng/mL of OXC and 75 ng/mL of each Lic enantiomer. The LOQ established for each Lic enantiomer was based on an OXC biotransformation of 1%. Within-day precision and accuracy were achieved by replicate analysis (n = 5) of 2.0 mL liquid culture medium samples spiked with standard solutions of the drug and metabolite at LOQ, low, medium, and high concentrations (1000, 3000, 15 000, and 22 500 ng/mL for OXC; 75, 150, 450, and 675 ng/mL for each enantiomer of Lic). Betweenday precision and accuracy were determined during routine operation of the system over a period of three consecutive working days. The overall precision of the method was expressed as RSD,% and accuracy was expressed as a percentage of relative error (RE,%) [26]. Freeze-thaw cycle stability and short-term room temperature stability, as well as stability under the biotransformation conditions for each analyte, were ensured. Stability tests involved five replicates (n = 5) of 2.0 mL of spiked liquid culture medium at low (1000 ng/mL for OXC and 150 ng/mL for each Lic enantiomer) and high (22 500 ng/mL for OXC and 675 ng/mL for each Lic enantiomer) concentration. To ensure stability under the biotransformation conditions, 3 mg aliquots of OXC were dissolved in the liquid culture medium and added to Erlenmeyer flasks containing 100 mL of liquid culture medium before being submitted to the conditions used in the biotransformation procedure (see Section 2.7). Each day during the period of biotransformation (15 days), 2 mL aliquots (n = 3) were analyzed. The results of the stability-study samples were compared with data obtained from fresh samples at the same concentration and were considered stable if the deviation (expressed as relative error, RE%) was within ± 15%. The selectivity of the method was evaluated by analyzing the liquid culture medium in the absence of OXC, but with the fungus, under the conditions previously established.

2.6 Fungus isolation and maintenance The endophytic fungus Glomerella cingulata (VA1) was isolated from the plant Viguiera arenaria. VA1 was maintained in agar plugs in a sterile glycerol:water (8:2, v/v) solution at −20°C. This strain was deposited in the Academic Laboratory of Chemistry of Microorganisms, Faculty of Pharmaceutical Sciences of Ribeir˜ao Preto (University de S˜ao Paulo, Ribeir˜ao Preto, Brazil). The fungus Beuveria bassiana ATCC 7159 was purchased from ATCC (University Boulevard, Manassas, VA, USA).

2.7 Oxcarbazepine biotransformation procedure The biotransformation procedure was carried out as previously described by our group [28–31]. Three disks with www.electrophoresis-journal.com

4

M. Z. Bocato et al.

0.5 cm diameters containing the fungal mycelia were aseptically transferred to separate 9.0 cm diameter Petri dishes containing potato dextrose agar and allowed to grow for 6 days at 22 ± 2°C. Three uniform disks (0.5 cm diameter) of the fungus mycelia were cut with a transfer tube (Fischer, Scientific, Pittsburgh, PA, USA) and inoculated in 50 mL Falcon tubes containing 10 mL of prefermentative medium (10.0 g malt extract, 10.0 g dextrose, 5.0 g triptone, 3.0 g yeast extract, and deionized water to 1 L, pH adjusted to 6.2 with a solution of 0.5 mol/L HCl) used for the appropriate growth of microorganism for 96 h, at 120 rpm at 30°C. After growth, each mycelium was completely transferred to a 250 mL Erlenmeyer flask containing 100 mL of modified Czapek medium (25.0 g sucrose, 2.0 g NaNO3 , 1.0 g KH2 PO4 , 0.5 g MgSO4 ·7H2 O, 0.5 g KCl, 0.01 g FeSO4 · 7H2 O, and deionized water to 1 L, pH adjusted to 5.0 with a solution of 1.0 mol/L HCl). At this point, 3.0 mg of OXC was dissolved in 500 ␮L of the same modified Czapek medium and added to the Erlenmeyer flask. The final concentration of OXC in the liquid culture medium was 30 ␮g/mL. All the experiments were performed in duplicate. The cultures were incubated for 360 h at 30°C, with shaking at 120 rpm. Control samples consisted of (i) Czapek liquid culture medium in the absence of OXC, but with the fungus, and (ii) Czapek liquid culture medium without the fungus, but added the OXC. The results obtained in the biotransformation process were expressed as enantiomeric excess (ee), determined by the equation ee = (A − B/A + B) × 100, where A is the enantiomer with higher concentration and B is the enantiomer with lower concentration.

2.8 Minimal inhibitory concentration The effect of OXC on fungal growth was evaluated by monitoring the minimal inhibitory concentration (MIC). Testing was performed in a 96-well plate, where each well was loaded with the culture medium (200 ␮L), the evaluated fungus and OXC. The OXC concentration was varied from 200 ␮g/mL to 97.7 ng/mL. The plate was incubated for 7 days at 30°C and the evaluation of the results was performed visually by comparing the fungal growth in the presence of an inhibitory agent and in the presence of OXC. The MIC is the lowest concentration capable of inhibiting fungal growth.

3 Results and discussion 3.1 Electrophoretic separation The optimization procedure for the simultaneous separation of OXC and Lic enantiomers employed uncoated fused-silica capillaries of 75 ␮m id and 40 cm effective length. The pKa of OXC is 10.7 [13], while the Lic pKa is 13.75 [32]. As these two compounds are weakly basic, an available analytical strategy for the resolution of these analytes by CE is screening with different CDs at acidic pH. Based on this, the following pa C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Electrophoresis 2014, 00, 1–8 Table 1. Comparison of migration times and resolution values obtained for the analytes employing 50 mmol/L phosphate buffer solution containing 1% of CM-␤-CD or P-␤-CD using an uncoated silica capillary of 40 cm effective length, an applied voltage of −20 kV and an analysis temperature of 20°C

Analytes

(E1)-Lica) (E2)-Lica) OXC

CM-␤-CD

P-␤-CD

Time (min)

Resolution

Time (min)

Resolution

35 26 28

1.9 2.4

23 29 35

5.8 5.1

a) The enantiomers were named as (E1)- and (E2)-Lic because the elution order has not been established for the CM-␤-CD.

rameters were evaluated in the resolution of Lic enantiomers: type of CD, electrolyte type, pH and concentration of electrolyte, concentration of the chiral selector, applied voltage, capillary temperature, and capillary length. To evaluate the effectiveness of individual CDs on the separation of the analytes, a monosodium phosphate buffer solution (50 mmol/L) was used at pH 2.5 (pH adjusted with (ortho)phosphoric acid) containing 1% w/v of the following CDs: CM-␤-CD, S-␤-CD, and P-␤-CD. The applied voltage was 10 kV in both normal (when CM-␤-CD was used) and reverse mode (when S-␤-CD or P-␤-CD was used). Among CDs tested, both CM-␤-CD and P-␤-CD successfully resolved all analytes. Next, the pH was varied from 2.0 to 5.0 with both CDs; however, the separation of all analytes was only achieved employing a pH value of 2.5. Table 1 shows the migration times and resolution values of the analytes. P-␤-CD was chosen for further use because it enabled lower analyte migration times and gave peaks with better resolution. The P-␤-CD is a negatively charged CD [33]; therefore, when using P-␤-CD, its electrophoretic mobility must be considered. At pH 2.5 the analytes will be in their cationic forms, so in reverse polarity they will migrate toward the cathode (injection side). On the other hand, anionic CDs will migrate toward the anode (detection side). When the anionic CD migrates toward anode it electrostatically interacts with the cationic analytes, carrying them to the detection side and promoting the analyte detection. Next, the type of BGE was evaluated. A Trisphosphate buffer (Tris(hydroximetyl)aminomethane solution) and monosodium phosphate buffer 50 mmol/L at were tested at pH 2.5. The best resolution for all analytes and shorter migration times were found for the Trisphosphate buffer. When phosphate buffer was used, the current level remained over 100 ␮A, which caused band enlargement due to Joule heating [34]. However, when the Trisphosphate buffer was used, the current level remained around of 90 ␮A, avoiding this phenomenon. As such, the Trisphosphate buffer was chosen as the background for further optimization. After choosing the buffer, its concentration was evaluated. The Tris-phosphate buffer (pH 2.5) was evaluated in the www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2014, 00, 1–8

Figure 2. Standard electropherogram for the separation of OXC and its metabolites after chiral separation optimization. Electrophoretic conditions: 10 mmol/L Tris-phosphate buffer solution (pH 2.5) containing P-␤-CD 1% (w/v). Capillary effective length: 20 cm. The applied voltage and capillary temperature were maintained constant at −20 kV and 15 °C, respectively. (1) OXC; (2) (S)-(+)-licarbazepine, (3) (R)-(−)-licarbazepine.

range of 10–100 mmol/L. Increasing the ionic strength of the buffer may offer better resolution and better theoretical plate values [35]; however, higher electrolyte concentration corresponds to longer migration times of the analytes and higher current levels. The optimal BGE concentration was set at 10 mmol/L Tris-phosphate buffer. After BGE optimization, the CD concentration was optimized. The CD concentration affects both the mobility of the enantiomers and their resolution [35]. The effect of P-␤-CD concentration on the enantioseparation was evaluated over a concentration range of 0.5–2.0% (w/v) P-␤-CD. Increasing the CD concentration increased the migration times and decreased the resolution values of all the analytes [35] as a result of high current levels, possibly leading to Joule heating. Therefore, 1.0% w/v P-␤-CD was chosen as it presented the best resolution values for all analytes with an acceptable current value [36]. Finally, the applied voltage (5–30 kV, in the reverse mode) and capillary temperature (10–25°C) were assessed. The best resolution values, with suitable analysis times and current levels, were accomplished employing −20 kV of applied voltage and capillary temperature of 15°C. To reduce the analysis time, the capillary effective length was

5

reduced to 20 cm. No loss in the resolution of the analytes was observed after the reduction of capillary length. The optimal CE conditions for the analysis of OXC and its chiral metabolite were achieved using a 10 mmol/L Trisphosphate buffer solution (pH 2.5) containing P-␤-CD 1% w/v employing a capillary of 20 cm effective length. The applied voltage and capillary temperature were maintained at −20 kV and 15°C, respectively. Under these conditions suitable resolutions were achieved in less than 8 min (Fig. 2) with a current level around 80 ␮A. The migration order was established by comparing the migration time of the pure enantiomers injected separately with the retention time described by Volosov et al. [11] In this previous HPLC method, the first enantiomer to elute was the (R)-(−)-Lic and the second was (S)-(+)-Lic. In our work, employing the optimized CE conditions, the first enantiomer to migrate was the (S)-(+)-Lic and the second was (R)-(−)-Lic.

3.2 Method validation To extract the OXC and Lic enantiomers from liquid culture medium, a liquid–liquid extraction was employed using methyl-tert-butyl ether as the extractor solvent. To improve the extraction, the medium was alkalinized with aqueous NaOH (0.1 mol/L) followed by the addition of 20% NaCl (w/v). Method validation was performed by internal standardization employing the drug risperidone as internal standard (Fig. 1D). Although risperidone presents a different chemical structure from the analytes separated in the present work, its chemical properties are similar to OXC and Lic [32, 38]. The method proved to be linear over the concentration range of 1000–30 000 ng/mL for OXC and 75–900 ng/mL for each Lic enantiomer with r ⬎ 0.9952 and relative error for each point of the analytical curve below 15% (Table 2). The LOQ was 75 ng/mL for each Lic enantiomer and 1000 ng/mL for OXC with relative error and RSD below 10% (Table 2). The lack of fit test presents suitable results for p- and F values. Therefore, the null hypothesis was not rejected, proving there is no lack of linear fit (Table 2). The recoveries of OXC and its metabolite were close to 55 and 80%, respectively, with RSDs lower than 9% for all analytes. The results for within-day and between-day precision and accuracy for each Lic enantiomer and OXC are summarized in Table 3. All results are in agreement with literature recommendations [26]. The freeze-thaw cycles and short-term room

Table 2. Limit of quantification and linearity of the developed method

Analytes

(S)-(+)-Lic (R)-(−)-Lic OXC

Range (ng/mL)

75–900 75–900 1000–30 000

Linear equationa)

y = 0.0084x + 0.0522 y = 0.0086x + 0.0130 y = 0.0022x + 0.2515

rb)

0.9952 0.9957 0.9955

ANOVA Lack of Fit F-value

p value

1.37 0.98 2.88

0.13 0.97 0.05

LOQ (ng/mL)

RE (%)

RSD (%)

75 75 1000

−5 −5 3

3 2 1

a) Five replicates (n = 5) for each concentration. b) r, coefficient of correlation.

 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.electrophoresis-journal.com

6

M. Z. Bocato et al.

Electrophoresis 2014, 00, 1–8

Table 3. Within-day and between-day precision and accuracy of the developed method

Analytes

(S)-(+)-Lic

(R)-(−)-Lic

OXC

Nominal concentrations (ng/mL)

Within-day Obtained concentrations (ng/mL)

RSDa) (%)

REb) (%)

Obtained concentrations (ng/mL)

RSDa) (%)

REb) (%)

75 150 450 675 75 150 450 675 1000 3000 15000 22500

68.7 157.9 434.2 575.1 68.5 160.3 427.5 582.2 963.9 3261.4 15191.1 22816.1

5 3 1 11 10 2 3 9 6 8 5 9

−9 5 −4 −14 −9 7 −5 −14 −4 9 1 2

68.5 164.6 411.4 694.0 68.7 167.3 412.2 698.6 940.1 3194.7 13854.9 21594.5

11 7 5 8 10 4 13 8 2 11 6 8

−9 10 −9 3 −8 12 −8 4 −6 7 −9 −4

Between-day

a) RSD expressed as a percentage. b) Relative error expressed as a percentage.

Table 4. Analyte stability test of the developed method

Analytes

(S)-(+)-Lic

Stability

CL

CH

CL

CH

CL

CH

153.0 5 2

586.2 1 −13

158.3 2 6

587.7 2 −13

2932.3 8 −2

19635.6 1 −13

153.1 8 2

604.5 1 −11

158.3 2 6

584.1 2 −14

3228.3 2 8

25202.2 2 12

Freeze-thaw cycles (n = 5) Concentration (ng mL−1 ) Precision (RSD)a) Accuracy (RE,%)b) Short-term (n = 5) Concentration (ng/mL) Precision (RSD)a) Accuracy (RE,%)b)

(R)-(−)-Lic

OXC

n = number of determinations. CL and CH = low concentration and high concentration. See Section 2.4. a) RSD expressed as a percentage. b) Relative error expressed as a percentage.

temperature stability tests showed no statistically significant degradation with RSD and relative errors lower than 15% (Table 4). The selectivity of the method was determined by analyzing the blank sample (liquid culture medium in the absence of OXC, but with the fungus). This study showed that the fungi did not produce any secondary metabolites with the same migration time of OXC or Lic enantiomers (Figs. 3 and 4).

3.3 Method application and enantioselective biotransformation studies The developed and validated method was employed for stereoselective fungal biotransformation studies described in Section 2.7. MIC tests indicated no toxicity of OXC to the evaluated fungi. The biotransformation reactions of OXC using fungi were monitored for 360 h. This biotransformation involves the reduction of a carbonyl group, yielding the Lic metabolites. Some enantioselective biotransformation  C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

studies employing fungi as catalytic agents have shown promising results with different type of substrates [28–31, 37, 38]. Based on these reported successes, we evaluated the fungi Glomerella cingulata (VA1) and Beuveria bassiana ATCC 7159 in the biotransformation of the drug OXC, producing the active and chiral metabolite licarbazepine. Both species of fungi were able to enantioselectively biotransform OXC into its metabolite. For both fungi, the (S)-(+)-Lic enantiomer was the primary product. Sample analyses were performed every 120 h during the biotransformation studies. Lic enantiomers were observed after 120 h of incubation and remained present until the end of the study. The results obtained in the biotransformation process (n = 2, for each fungus) were expressed as ee. As can be seen in Fig. 3, VA1 was able to biotransform OXC with 100% enantiomeric excess of the (S)-(+)-Lic enantiomer. This result was observed for all of the collected samples. This fungus has already demonstrated the ability to enantioselectively biotransform other drugs. Borges et al. [39] www.electrophoresis-journal.com

CE and CEC

Electrophoresis 2014, 00, 1–8

7

Table 5. Biotransformation study employing the fungi Glomerella cingulata and Beuveria bassiana

Sampling hour

120 240 360

Glomerella cingulata (VA1)

Beuveria bassiana ATCC 7159

(S)-(+)-Lic (ng/mL)

ee (%)

(S)-(+)-Lic (ng/mL)

(R)-(−)-Lic (ng/mL)

ee (%)

212.0 666.6 885.4

100 100 100

163.3 499.0 707.9

83.4

100 100 79

ee = enantiomeric excess.

reported that VA1 was able to biotransform the drug propranolol, achieving 50% enantiomeric excess of the metabolite (−)-(S)-4-hydroxypropranol. The Beuveria bassiana ATCC 7159 fungus was also able to stereoselectively biotransform OXC into both Lic enantiomers. This study showed preferential formation of the (S)(+)-Lic enantiomer (Fig. 4), resulting in 79% ee. Another report showed the Beuveria bassiana ATCC 7159 fungus was able to produce enantiomerically pure diethyl 1-hydroxy-1phenylmethanephosphonate [40]. Table 5 shows the enantiomeric excess for each fungus species and the concentration obtained in each period. These results show that VA1 can be used to produce the (S)-(+)Lic, the most active enantiomer of the Lic metabolite, in its enantiomerically pure form. Figure 3. Representative electropherogram after OXC biotransformation by the fungus Glomerella cingulata (VA1) after 360 h of incubation (A). Representative electropherogram of Czapek liquid culture medium (control) incubated with the Glomerella cingulata (VA1), showing that this fungus did not produce any secondary metabolites in the migration time of the analytes (B). Representative electropherogram at t = 0 (C). Electrophoretic and extraction conditions described in Sections 2.2 and 2.4, respectively. (1) Internal standard; (2) OXC; (3) (S)-(+)-licarbazepine.

4 Concluding remarks The present work describes the first simultaneous analysis of OXC and its chiral and active metabolite licarbazepine by CE. The advantages of the developed method over the described methods [7, 11, 13] are as follows: faster analysis, no consumption of organic solvents, and simplicity. The biotransformation studies described here showed the possibility of obtaining enantiomerically pure (S)-(+)-licarbazepine by employing fungi as a catalytic agent. This work was supported by grants# 2013/17658-9, the S˜ao Paulo Research Foundation (FAPESP), Conselho Nacional de Desenvolvimento Cient´ıfico e Tecnol´ogico (CNPq) and Coordenac¸a˜o de Aperfeic¸oamento de Pessoal de N´ıvel Superior (CAPES). The authors have declared no conflict of interest.

Figure 4. Representative electropherogram after OXC biotransformation by the fungus Beuveria bassiana ATCC 7159 after 360 h of incubation (A). Representative electropherogram of Czapek liquid culture medium (control) incubated with the Beuveria bassiana ATCC 7159 showing that this fungus did not produce any secondary metabolites in the migration time of the analytes (B). Representative electropherogram at t = 0 (C). Electrophoretic and extraction conditions are described in Sections 2.2 and 2.4, respectively. (1) Internal standard; (2) OXC; (3) (S)-(+)-licarbazepine, (4) (R)-(−)-licarbazepine.

 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

5 References [1] Flesch, G., Czendlik, C., Renard, D., Lloyd, P., Drug Metab. Disp. 2011, 39, 1103–1110. [2] Shorvon, S., Seizure 2000, 9, 75–79 [3] Sozuer, D. T., Atakil, D., Dogu, O., Baybas, S., Arpaci, B., Eur. J. Pediatr. 1997, 156, 565–567. [4] Bramswig, S. Sudhop, T., Luers, C., Von Bergmann, K., Berthold, H. K., Epilep. 2003, 44, 457–460.

www.electrophoresis-journal.com

8

M. Z. Bocato et al.

[5] Schwaninger, M., Ringleb, P., Annecke, A., Winter, R., Kohl, B., Werle, E., Filhn, W., Riesen, P. A., Walter-Sack, I., J. Neurol. 2000, 247, 687–690. [6] Schmidt, D., Elger, C. E., Epilep. & Behav. 2004, 5, 627–635. [7] Antunes, N. de J., Wichert-Ana, L., Coelho, E. B., Pasqua, O. D., Jr, V. A., Takayanagui, O. M., Tozatto, E., Lanchote, V. L., Chirality 2013, 25, 897–903.

Electrophoresis 2014, 00, 1–8

[25] Borges, W. S., Borges, K. B., Bonato, P. S., Said, S., Pupo, M. T., Curr. Org. Chem. 2009, 13, 1137–1163. [26] European Medicines Agency (EMA), Guideline on Bioanalytical Method 554 Validation, 2011 http://www. ema.europa.eu/docs/enGB/documentlibrary/555Scientifi cguideline/2011/08/WC500109686.pdf (accessed 12.13). ˜ A. C., J. [27] Almeida, A. M., Castel-Branco, M. M., Falcao, Chromatogr. B 2002, 774, 215–222.

[9] Schmutz, M., Brugger, F., Gentsch, C., McLean, M. J., Olpe, H. R., Epilep. 1994, 35, 47–50.

˜ [28] Jesus, L. I., Albuquerque, N. C. P., Borges, K. B., Simoes, R. A., Calixto, L. A. Furtado, N. A. J. C., de Gaitani, C. M., Pupo, M. T., de Oliveira, A. R. M., Electrophoresis 2011, 32, 2765–2775.

[10] Volosov, A., Xiaodongo, S., Perucca, E, Yagen, B., Sintov, A. Bialer, M., Clin. Pharmacol. Therap. 1999, 66, 547–553.

˜ D. B., Borges, K. B., Barth, T., Pupo, M. T., Bon[29] Carrao, ato, P. S., de Oliveira, A. R. M., Electrophoresis 2011, 32, 2746–2756.

[11] Volosov, A., Bialer, M., Xiaodong, S., Perucca, E., Sintov, A., Yagen, B., J. Chromatogr. B 2000, 738, 419–425.

[30] Borges, K. B., de Oliveira, A. R. M., Barth, T., Jabor, V. A. P., Pupo, M. T., Anal. Bioanal. Chem. 2010, 399, 915–925.

[8] French, J. A., Baroldi, P., Brittain, S. T., Johnson, J. K., Acta Neurol. Scand. 2014, 129, 143–153.

´ [12] Benes, J., Soares-da-Silva, P., USP TO patent no. 5753646, 19 May 1998. [13] Marziali, E., Raggi, M. A., Komarova, N., Kenndler, E., Electrophoresis 2002, 23, 3020–3026.

´ ˜ D. B., Barth, T., Borges, K. B., Fur[31] Hilario, V. C., Carrao, tado, N. A. J. C., Pupo, M. T., de Oliveira, A. R. M., J. Pharm. Biom. Anal. 2012, 61, 100–107.

[14] Mazzucchelli, J., Franco, V., Fattore, C., Marchiselli, R., Perucca, E., Gatti, G., Ther. Drug. Monit. 2007, 29, 319–324

[32] Bialer M., in: Levy, R. H., Mattson, R. H., Meldrum, B. S., Perucca, E., (Eds.), (Antiepileptic Drugs, 5th edition, Lippincott Williams & Wilkins, Philadelphia 2002, pp. 459–465.

[15] Loureiro, A. I., Fernandes-Lopes, C., Wright, L. C., Soares-da-Silva, P., J. Chromatogr. B 2011, 879, 2611–2618.

[33] Lehnert, P., Pribylka, A., Maier, V., Znaleziona, J., Sevcik, J. Dousa, M., J. Sep. Sci. 2013, 36, 1561–1567.

[16] Flesch, G., Francote, E., Hell, F., Degen, P. H., J. Chromatogr. 1992, 581, 147–151. [17] Alves, G., Figueiredo, I., Castel-Branco, M., Loureiro, A., ˜ A., Caramona, M., Anal. Chim. Acta. 2007, 596, Falcao, 132–140. ˜ [18] Fortuna, A., Alves, G., Almeida, A., Lopes, B., Falcao, A., Soares-da-Silva, P., Biomed. Chromatogr. 2012, 26, 384–392.

[34] Nelson, R. J., Burgi, D. S., in: Landers, J. P. (Ed.), Handbook of Capillary Electrophoresis, CRC Press, Boca Raton 1994, 549–562. [35] Gubitz, G., Schmid, M. G., in: Van Eeckhaut, A., Michote, Y. (Ed.), Chiral Separations by Capillary Electrophoresis, CRC Press, Boca Raton 2010, pp. 47–86.

[19] Azerad, R., in: Scheper, T. (Ed.), Advances in Biochemical Engineering Biotechnology. Springer-Verlag, Berlin, Germany 1999, p. 169.

[36] de Gaitani, C. M., de Oliveira, A. R. M., Bonato, P. S., in: Garc´ıa, C. D., Chumbimuni-Torres, K. Y., Carrilho, E. (Eds.), Capillary Electrophoresis and Microchip Capillary Electrophoresis, Principles, Applications and Limitations, John Wiley & Sons, Inc., Hoboken, New Jersey 2013, pp. 236–237

[20] Pervaiz, J., Ahmad, S., Madni, M. A., Ahmad, H., Khaliq, F. H., Appl. Biochem. And Microb. 2013, 49, 437–450.

[37] Fortes, S. S., Barth, T., Furtado, N. A. J. C., Pupo, M. T., de Gaitani, C. M., de Oliveira, A. R. M., Talanta 2013, 16, 743–752.

[21] van den Brink, H. J. M., van Gorcon, R. F. M., van den Hondel, C. A. M. J. J., Punt, P. J., Fungal Genet. Biol. 1998, 23, 1–17.

˜ [38] Bocato, M. Z., Simoes, R. A., Calixto, L. A., de Gaitani, C. M., Pupo, M. T., de Oliveira, A. R. M., Anal. Chim. Acta 2012, 742, 80–89.

[22] Das, S., Rosazza, J. P. N., J. Nat. Prod. 2006, 69, 499–508.

[39] Borges, K. B., Pupo, M. T., Bonato, P. S., Electrophoresis 2009, 30, 3910–3917.

[23] Clark, A. M., Hefford, C. D., Med. Res. Rev. 1991, 11, 473–501. [24] Abourashed, E. A. Clark, A. M., Hulford, C. D., Curr. Med. Chem. 1999, 6, 359–374

 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

´ [40] Zymancz-Duda, E., Brzezinska-Rodak, M., Klimek-Ochab, M., Lejczak, B., Curr Microbiol. Apr. 2011, 62, 1168–1172.

www.electrophoresis-journal.com