Research article Received: 13 October 2014,
Revised: 4 February 2015,
Accepted: 4 March 2015
Published online in Wiley Online Library: 9 April 2015
(wileyonlinelibrary.com) DOI 10.1002/bmc.3472
Evaluation of the enantioselective binding of imazalil to human serum albumin by capillary electrophoresis Lucía Asensi-Bernardia, Yolanda Martín-Bioscaa, Laura Escuder-Gilaberta, Salvador Sagradoa,b and María José Medina-Hernándeza* ABSTRACT: In this work, a methodology for the evaluation of enantioselective binding of imazalil (IMA) enantiomers to human serum albumin (HSA) that does not require the separation of free and bound to HSA fractions is developed. This methodology comprises the incubation of IMA–HSA designed mixtures for 30 min directly in the capillary electrophoresis system and the subsequent direct injection and chiral separation of IMA employing highly sulfated β-cyclodextrin as chiral selector and the complete filling technique. Two mathematical approaches were used to estimate apparent affinity constants (K1), protein binding and enantioselectivity (ES) for both enantiomers of IMA. Moderate enantioselective binding of IMA enantiomers to HSA (ES = 2.0) was shown by the 1:1 stoichiometry and log K1 values of 3.4 ± 0.4 and 3.1 ± 0.3 for the first and second eluted enantiomers, respectively. Copyright © 2015 John Wiley & Sons, Ltd. Keywords: chiral capillary electrophoresis; enantioselective binding; human serum albumin; imazalil enantiomers; toxicokinetics
Introduction
Biomed. Chromatogr. 2015; 29: 1637–1642
* Correspondence to: M. J. Medina-Hernández, Departamento de Química Analítica, Facultat de Farmacia, Universitat de Valencia, C/Vicent Andrés Estellés s/n, E-46100 Burjassot, Valencia, Spain. Email: [email protected] a
Departamento de Química Analítica, Facultad de Farmacia, Universitat de València, Burjassot, Spain
b
Centro Interuniversitario de Reconocimiento Molecular y Desarrollo Tecnológico, Unidad Mixta Universidad Politécnica de Valencia-Universitat de València, Valencia, Spain Abbreviations used: BGE, background electrolyte; CD, cyclodextrin; CE, capillary electrophoresis; CFT, complete filling technique; ES, enantioselectivity; HS-β-CD, highly sulfated β-cyclodextrin; HSA, human serum albumin; IMA, imazalil; K1, apparent affinity constant; PB, protein binding; Rs, enantiomers resolution.
Copyright © 2015 John Wiley & Sons, Ltd.
1637
Pesticides, which are widely used worldwide to protect crops from pests, constitute a major source of potential environmental hazard to humans and animals since they are present and concentrated in the food chain (Sisman and Türkez, 2010). Approximately 25% of the agrochemicals employed are chiral, and several differences in the efficacy and safety profiles of their enantiomers have been found, suggesting that sometimes, when only one of the enantiomers is active or toxic, the use of the racemate (which is the usual procedure) leads to unnecessary contamination that could be avoided and/or reduced with the use of a single enantiomer. For this reason, the main effort of research involving this kind of products concerns the development of new single-enantiomer pesticides with the same effectiveness over pests but showing higher selectivity and lower side effects (Pérez-Fernández et al., 2011). Xenobiotic–protein interactions are determining factors in the toxicokinetics and toxicity properties of xenobiotics. Human serum albumin (HSA) is the most abundant protein in human plasma, and it presents a high degree of enantioselective interaction with drugs (Ascoli et al., 2006), so it may be also enantioselective in its interaction with other chiral xenobiotics such as agrochemicals. Some studies have shown differential binding of pesticides to serum albumin (Carter et al., 2007), and importantly, less selective (and less physiologically relevant) pesticide binding to commercial (fatty acid free) albumin protein has been reported (Peeples et al., 2005). For the investigation of drug enantioselective protein binding, different approaches have been proposed, which can be applied a priori to the evaluation of their binding to other xenobiotics (Escuder-Gilabert et al., 2009). The majority of these methodologies include a first step in which the free xenobiotic is separated from the bound one once the xenobiotic–protein equilibrium has been reached using different membrane-based separation techniques, such as the traditional equilibrium dialysis, ultracentrifugation or
ultrafiltration. One of the most important disadvantages of these classical methodologies is the nonspecific adsorption of the xenobiotic to different components of the devices, such as plastics or membranes, which can strongly affect the results of the experiment. The second step includes chiral analysis of the xenobiotic enantiomers, usually on the unbound fraction. For this step, different analytical techniques such as gas chromatography, high-performance liquid chromatography or more recently capillary electrophoresis (CE) have been proposed. The separation of enantiomers is the result of the interaction between them and the chiral selector in a separation system. Cyclodextrins (CDs) are the most widely employed chiral selectors in capillary electrophoresis, owing to their good enantiorecognition abilities, good water solubility, UV transparency and the wide assortment of different neutral, cationic and anionic CDs with different functional groups that can be employed. Even though charged CDs were used initially for the enantioseparation of neutral compounds, they also have shown
L. Asensi-Bernardi et al. high enantioselectivity towards many basic and acidic compounds. It has been reported that maximum enantioselectivities are obtained when analytes and CDs have opposite mobilities (Asensi-Bernardi et al., 2011; Rudaz et al., 2004). Imazalil (IMA) is a widely used chiral fungicide of the family of imidazole derivatives that is employed as racemic mixture of enantiomers. It is extensively used in fruit plantations and post-harvest treatments (Ortelli et al., 2005; Zega et al. 2009). There is potential exposure to workers handling citrus fruits or treated seeds (Sisman and Türkez, 2010), and the general population is exposed to IMA via ingestion of contaminated food. It has a teratogenic effect over vertebrates and invertebrates (Sisman and Türkez, 2010), cytotoxic effect (Nakagawa and Moore, 1995), ability to induce cytochrome P450 isoforms (Muto et al., 1997) and an inhibitory activity against CYP19 aromatase (Vinggaard et al., 2000), which catalyzes the conversion of androgens to estrogens. Different CE methods have been proposed in the literature for the chiral analysis of IMA (Chankvetadze et al., 1995; Kodama et al., 2003; Chu et al., 2007) employing CDs as chiral selectors. The enantioselectivity of its degradation in oranges (Kodama et al., 2003) and soils (Chu et al., 2007) has been evaluated employing these methods. However, to knowledge there are no studies on the enantioselectivity of its toxic activity or its toxicokinetics. The aim of this work is to evaluate the enantioselective binding of imazalil to HSA, the most abundant plasma protein. For this purpose, a new and simple methodology is proposed. This methodology consists of the pre-incubation of IMA–HSA mixtures and their posterior analysis by chiral CE using the highly sulfated β-cyclodextrin (HS-β-CD) and the complete filling technique (CFT). The CFT, which involves the filling of the capillary completely with the chiral selector prior to the sample injection, presents several advantages over the conventional electrokinetic chromatography employed for enantioseparations, such as the extremely low consumption of cyclodextrin, since the inlet and outlet vials of the separation system are free of chiral selector. The HS-β-CD, with an average content of 12 sulfate groups per molecule (Chen et al., 2001), presents a high density of negative charge that makes it an excellent chiral selector for positively charged compounds, since it is negative in practically all the pH range
dihydrogen phosphate dihydrate from Fluka (Buchs, Switzerland), racemic IMA from Dr Ehrenstofer GmH (Augsburg, Germany) and HS-β-CD 20% m/v aqueous solution from Beckman Coulter (Fullterton, CA, USA). Sodium hydroxide and sodium chloride were from Scharlau (Barcelona, Spain). Ultra Clear TWF UV deionized water (SG Water, Barsbüttel, Germany) was used to prepare solutions. Background electrolyte (BGE; 50 mM NaH2PO4, pH 7.0) was prepared by weighing the appropriate amount of phosphate salt, dissolving it in deionized water and adjusting the pH with 1 M NaOH. Incubation buffer for the study of IMA–HSA interaction (50 mM NaH2PO4, pH 7.4 + 9.2 g/L NaCl) was prepared similarly by dissolving the appropriate amount of the solid reagents in deionized water and adjusting the pH to 7.4. This incubation buffer was chosen to take into account the degree of ionization of IMA and the saline concentration at physiological conditions (pH 7.4). An IMA 1000 μM stock solution was prepared in methanol and kept at 4°C. Working solutions were prepared by dilution of the stock one. For the IMA–HSA binding study, a 1200 μM HSA stock solution was prepared daily in the incubation buffer.
Capillary conditioning New capillaries were conditioned for 15 min flush with 1 M NaOH at 60°C. Then, they were rinsed for 5 min with deionized water and 10 min with separation buffer at 25°C. In order to obtain good peak shapes and reproducible migration data, the capillary was conditioned prior to each injection. In all cases, the conditioning run included the following steps: (a) 1 min rinse with deionized water; (b) 1 min rinse with 1 M NaOH, to prevent protein adsorption to the capillary wall; (c) 1 min rinse with deionized water; and (d) 2 min rinse with separation buffer. All steps were carried out at 20 psi.
Procedure for the enantioseparation of IMA by CE-CFT For all experiments, the 50 mM NaH2PO4 solution at pH 7.0 was employed as background electrolyte. A 0.5% w/v HS-β-CD solution obtained by dilution of the 20% m/v solution in the BGE was employed. IMA solutions (standard solutions for calibration and IMA–HSA mixtures) were injected hydrodynamically at 0.5 psi for 5 s. Before sample injection, the capillary was filled with the 0.5% w/v HS-β-CD solution by applying 10 psi for 1 min. Separation voltage was 15 kV (normal polarity), and the capillary temperature was 40°C. These separation conditions provide a maximum current of 65 μA during the CE measurements.
Procedure for the evaluation of protein binding
Experimental Instrumentation A Beckman P/ACE MDQ capillary electrophoresis system equipped with a diode array detector (Beckman Coulter, Fullterton, CA, USA), and 32Karat software version 8.0 was used throughout. A 50 μm i.d. fused-silica capillary with total and effective lengths of 60.2 and 50 cm, respectively, was employed (Beckman Coulter, Fullterton, CA, USA). The CE system is equipped with a temperature-controlled garage for sample trays, so samples can be warmed during all the experiences. Detection wavelength was fixed at 200 nm for all experiments. Electrophoretic solutions and samples were filtered through 0.45 μm pore size nylon membranes (Micron Separation, Westboro, MA, USA) and degassed in an ultrasonic bath ( JP Selecta, Barcelona, Spain) prior to use. A Crison Micro pH 2000 pH-meter from Crison Instruments (Barcelona, Spain) was employed to adjust the pH of buffer solutions.
An experimental design was planned to evaluate the enantioselective binding of IMA to HSA. In this design, the protein concentration was kept near physiological values and the IMA concentration was varied according to the detector capabilities. Mixtures containing five different racemic IMA concentrations, approximately in the 150–350 μM range, and a fixed HSA concentration of 530 μM were prepared by dilution of the stock solutions in the incubation buffer. All of these mixtures were mixed vigorously and transferred to CE microvials. Mixtures were incubated directly in the CE system at 36.5°C for 30 min, and then injected hydrodynamically in the capillary for the separation. Corrected peak area was employed as response variable for the quantification of the unbound fraction, and with this purpose a calibration curve was constructed for each IMA enantiomer.
Software and calculations From the experimental data, the enantiomers resolution (Rs) was calculated using the following expression:
Chemicals and reagents
1638
All reagents were of analytical grade. Human serum albumin fraction V (code no. A1653) was purchased from Sigma (St. Louis, MO, USA), sodium
wileyonlinelibrary.com/journal/bmc
Copyright © 2015 John Wiley & Sons, Ltd.
Rs ¼
1:18ðt2 t1 Þ w1 þ w2
(1)
Biomed. Chromatogr. 2015; 29: 1637–1642
Enantioselective binding of imazalil to HSA were t1 and t2 are, respectively, the migration times of the first (E1) and second (E2) eluted enantiomers, and w1 and w2 represent the peak widths at 50% peak heights. For the IMA–HSA binding study, routines made/adapted in MATLAB® version 4.2 were used for calculations. The results generated were verified using EXCEL® and STATGRAPHICS®. In some calculations a SIMPLEX algorithm routine was used (nonlinear fitting). To evaluate the IMA–HSA binding, simplifications of the general protein binding model (eqn (2)) that considers m classes of independent active sites on the HSA molecule, each with ni binding sites, were used (AsensiBernardi et al., 2010).
r¼
m b Dd X Kid ¼ ¼ ni P P 1 þ Kid i¼1
(2)
In this equation, r is the fraction of bound enantiomer per molecule of protein, b is the bound concentration of enantiomer and P is the total protein concentration. D and d are the total and free enantiomer concentrations, respectively. All the concentrations must be expressed in molar units (M). For the site i, ni represents the apparent site-stoichometry. Equation (2) assumes negligible nonspecific and noncooperative binding. The first simplification assumes m = 1 (one class of binding site for this molecule) and n1 = 1. Both assumptions should be reasonably valid for low D/P ratios (the usual situation in vivo; Asensi-Bernardi et al., 2010). Concretely, D/P ratios 2.5 was fixed to eliminate a point in a round.
Table 3. Apparent affinity constants, apparent protein binding and enantioselective of the binding of IMA enantiomers to human serum albumin (HSA)
IMA-E1
IMA-E2
Log K1, M1a
PBb
ES
3.4 ± 0.4
57.1 (D ≤ 1 μM) 56.8 (D = 10 μM) 54.4 (D = 100 μM) 40.0 (D ≤ 1 μM) 39.8 (D = 10 μM) 38.2 (D = 100 μM)
2.0
3.1 ± 0.3
a
Data estimated from eqn (3) (median ± MADe). Estimated for D = 0.01, 0.1, 1, 10, 100 μM (total enantiomer concentration) and total protein concentration = 530 μM. K1, Apparent affinity constant; PB, protein binding; ES, enantioselectivity. b
zi ¼
logK 1i median MADe
(5)
Table 3 provides robust estimates of log K1 values obtained from eqn (3); no outliers were found. To verify the parameters n1 and K1, a nonlinear approach based on eqn (4) was employed. This is an alternative to other linear models having recognized mathematical inconsistencies and risks (Barri et al., 2008; Fielding et al., 2005). The SIMPLEX search was started with n1 = 1 and log K1 = 3 (based on estimates on Table 3), and only integer n1 values were allowed. Estimated log K1 values were 3.19 and 3.08 for E1 and E2, respectively, which is in agreement with values in Table 3. Moreover n1 = 1 was estimated for both enantiomers, verifying the assumption of eqn (3). Figure 3 shows the d vs D models for both enantiomers, obtained with eqn (4) (values of R2 > 0.8 were obtained for the fitted nonlinear models). The results suggest the adequacy of the estimations performed with eqn (3). Enantioselectivity (ES) was estimated as the ratio of KE1 (2512) and KE2 (1259), providing an ES value of 2.0, in favour of the first eluted enantiomer. This result is the first evidence of a moderate enantioselective binding of IMA to HSA. From a pharmacokinetic point of view, an interesting parameter is the percentage of protein binding (PB, %). Using P = 530 μM, the K1 values in Table 3 and different D values according to different intoxication levels with the plaguicide, d values were estimated for each enantiomer by means of eqn (4) and used for the apparent PB estimations with the following equation: PB ¼ 100
Dd D
(6)
Apparent estimated PB values for the different D situations, as well as the estimation of ES, are also shown in Table 3. These results, which demonstrate a moderate enantioselective binding of the pesticide IMA to the HSA, show the possibility of using this protein as chiral selector for the enantioseparation of IMA in future applications using affinity electrokinetic chromatography. In this technique the xenobiotic–HSA interaction is a prerequisite for enantioresolution. When the analyte and the analyte–protein complex have different mobilities, the interaction of the compound with the protein results in a change in the net mobility of the analyte. (Martínez-Gómez et al., 2006)
Biomed. Chromatogr. 2015; 29: 1637–1642
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/bmc
1641
Figure 3. d vs D plots (a, b) for the binding of, respectively, IMA-E1 and IMA-E2 to human serum albumin. d vs D plots obtained by SIMPLEX optimization of data to eqn (4).
L. Asensi-Bernardi et al.
Conclusions In this work, the enantiomers of imazalil, a chiral plaguicide, were separated by CE-CFT employing HS-β-CD as chiral selector. The separation method provided a good enantioresolution of 4.4 in a short analysis time (
Revised: 4 February 2015,
Accepted: 4 March 2015
Published online in Wiley Online Library: 9 April 2015
(wileyonlinelibrary.com) DOI 10.1002/bmc.3472
Evaluation of the enantioselective binding of imazalil to human serum albumin by capillary electrophoresis Lucía Asensi-Bernardia, Yolanda Martín-Bioscaa, Laura Escuder-Gilaberta, Salvador Sagradoa,b and María José Medina-Hernándeza* ABSTRACT: In this work, a methodology for the evaluation of enantioselective binding of imazalil (IMA) enantiomers to human serum albumin (HSA) that does not require the separation of free and bound to HSA fractions is developed. This methodology comprises the incubation of IMA–HSA designed mixtures for 30 min directly in the capillary electrophoresis system and the subsequent direct injection and chiral separation of IMA employing highly sulfated β-cyclodextrin as chiral selector and the complete filling technique. Two mathematical approaches were used to estimate apparent affinity constants (K1), protein binding and enantioselectivity (ES) for both enantiomers of IMA. Moderate enantioselective binding of IMA enantiomers to HSA (ES = 2.0) was shown by the 1:1 stoichiometry and log K1 values of 3.4 ± 0.4 and 3.1 ± 0.3 for the first and second eluted enantiomers, respectively. Copyright © 2015 John Wiley & Sons, Ltd. Keywords: chiral capillary electrophoresis; enantioselective binding; human serum albumin; imazalil enantiomers; toxicokinetics
Introduction
Biomed. Chromatogr. 2015; 29: 1637–1642
* Correspondence to: M. J. Medina-Hernández, Departamento de Química Analítica, Facultat de Farmacia, Universitat de Valencia, C/Vicent Andrés Estellés s/n, E-46100 Burjassot, Valencia, Spain. Email: [email protected] a
Departamento de Química Analítica, Facultad de Farmacia, Universitat de València, Burjassot, Spain
b
Centro Interuniversitario de Reconocimiento Molecular y Desarrollo Tecnológico, Unidad Mixta Universidad Politécnica de Valencia-Universitat de València, Valencia, Spain Abbreviations used: BGE, background electrolyte; CD, cyclodextrin; CE, capillary electrophoresis; CFT, complete filling technique; ES, enantioselectivity; HS-β-CD, highly sulfated β-cyclodextrin; HSA, human serum albumin; IMA, imazalil; K1, apparent affinity constant; PB, protein binding; Rs, enantiomers resolution.
Copyright © 2015 John Wiley & Sons, Ltd.
1637
Pesticides, which are widely used worldwide to protect crops from pests, constitute a major source of potential environmental hazard to humans and animals since they are present and concentrated in the food chain (Sisman and Türkez, 2010). Approximately 25% of the agrochemicals employed are chiral, and several differences in the efficacy and safety profiles of their enantiomers have been found, suggesting that sometimes, when only one of the enantiomers is active or toxic, the use of the racemate (which is the usual procedure) leads to unnecessary contamination that could be avoided and/or reduced with the use of a single enantiomer. For this reason, the main effort of research involving this kind of products concerns the development of new single-enantiomer pesticides with the same effectiveness over pests but showing higher selectivity and lower side effects (Pérez-Fernández et al., 2011). Xenobiotic–protein interactions are determining factors in the toxicokinetics and toxicity properties of xenobiotics. Human serum albumin (HSA) is the most abundant protein in human plasma, and it presents a high degree of enantioselective interaction with drugs (Ascoli et al., 2006), so it may be also enantioselective in its interaction with other chiral xenobiotics such as agrochemicals. Some studies have shown differential binding of pesticides to serum albumin (Carter et al., 2007), and importantly, less selective (and less physiologically relevant) pesticide binding to commercial (fatty acid free) albumin protein has been reported (Peeples et al., 2005). For the investigation of drug enantioselective protein binding, different approaches have been proposed, which can be applied a priori to the evaluation of their binding to other xenobiotics (Escuder-Gilabert et al., 2009). The majority of these methodologies include a first step in which the free xenobiotic is separated from the bound one once the xenobiotic–protein equilibrium has been reached using different membrane-based separation techniques, such as the traditional equilibrium dialysis, ultracentrifugation or
ultrafiltration. One of the most important disadvantages of these classical methodologies is the nonspecific adsorption of the xenobiotic to different components of the devices, such as plastics or membranes, which can strongly affect the results of the experiment. The second step includes chiral analysis of the xenobiotic enantiomers, usually on the unbound fraction. For this step, different analytical techniques such as gas chromatography, high-performance liquid chromatography or more recently capillary electrophoresis (CE) have been proposed. The separation of enantiomers is the result of the interaction between them and the chiral selector in a separation system. Cyclodextrins (CDs) are the most widely employed chiral selectors in capillary electrophoresis, owing to their good enantiorecognition abilities, good water solubility, UV transparency and the wide assortment of different neutral, cationic and anionic CDs with different functional groups that can be employed. Even though charged CDs were used initially for the enantioseparation of neutral compounds, they also have shown
L. Asensi-Bernardi et al. high enantioselectivity towards many basic and acidic compounds. It has been reported that maximum enantioselectivities are obtained when analytes and CDs have opposite mobilities (Asensi-Bernardi et al., 2011; Rudaz et al., 2004). Imazalil (IMA) is a widely used chiral fungicide of the family of imidazole derivatives that is employed as racemic mixture of enantiomers. It is extensively used in fruit plantations and post-harvest treatments (Ortelli et al., 2005; Zega et al. 2009). There is potential exposure to workers handling citrus fruits or treated seeds (Sisman and Türkez, 2010), and the general population is exposed to IMA via ingestion of contaminated food. It has a teratogenic effect over vertebrates and invertebrates (Sisman and Türkez, 2010), cytotoxic effect (Nakagawa and Moore, 1995), ability to induce cytochrome P450 isoforms (Muto et al., 1997) and an inhibitory activity against CYP19 aromatase (Vinggaard et al., 2000), which catalyzes the conversion of androgens to estrogens. Different CE methods have been proposed in the literature for the chiral analysis of IMA (Chankvetadze et al., 1995; Kodama et al., 2003; Chu et al., 2007) employing CDs as chiral selectors. The enantioselectivity of its degradation in oranges (Kodama et al., 2003) and soils (Chu et al., 2007) has been evaluated employing these methods. However, to knowledge there are no studies on the enantioselectivity of its toxic activity or its toxicokinetics. The aim of this work is to evaluate the enantioselective binding of imazalil to HSA, the most abundant plasma protein. For this purpose, a new and simple methodology is proposed. This methodology consists of the pre-incubation of IMA–HSA mixtures and their posterior analysis by chiral CE using the highly sulfated β-cyclodextrin (HS-β-CD) and the complete filling technique (CFT). The CFT, which involves the filling of the capillary completely with the chiral selector prior to the sample injection, presents several advantages over the conventional electrokinetic chromatography employed for enantioseparations, such as the extremely low consumption of cyclodextrin, since the inlet and outlet vials of the separation system are free of chiral selector. The HS-β-CD, with an average content of 12 sulfate groups per molecule (Chen et al., 2001), presents a high density of negative charge that makes it an excellent chiral selector for positively charged compounds, since it is negative in practically all the pH range
dihydrogen phosphate dihydrate from Fluka (Buchs, Switzerland), racemic IMA from Dr Ehrenstofer GmH (Augsburg, Germany) and HS-β-CD 20% m/v aqueous solution from Beckman Coulter (Fullterton, CA, USA). Sodium hydroxide and sodium chloride were from Scharlau (Barcelona, Spain). Ultra Clear TWF UV deionized water (SG Water, Barsbüttel, Germany) was used to prepare solutions. Background electrolyte (BGE; 50 mM NaH2PO4, pH 7.0) was prepared by weighing the appropriate amount of phosphate salt, dissolving it in deionized water and adjusting the pH with 1 M NaOH. Incubation buffer for the study of IMA–HSA interaction (50 mM NaH2PO4, pH 7.4 + 9.2 g/L NaCl) was prepared similarly by dissolving the appropriate amount of the solid reagents in deionized water and adjusting the pH to 7.4. This incubation buffer was chosen to take into account the degree of ionization of IMA and the saline concentration at physiological conditions (pH 7.4). An IMA 1000 μM stock solution was prepared in methanol and kept at 4°C. Working solutions were prepared by dilution of the stock one. For the IMA–HSA binding study, a 1200 μM HSA stock solution was prepared daily in the incubation buffer.
Capillary conditioning New capillaries were conditioned for 15 min flush with 1 M NaOH at 60°C. Then, they were rinsed for 5 min with deionized water and 10 min with separation buffer at 25°C. In order to obtain good peak shapes and reproducible migration data, the capillary was conditioned prior to each injection. In all cases, the conditioning run included the following steps: (a) 1 min rinse with deionized water; (b) 1 min rinse with 1 M NaOH, to prevent protein adsorption to the capillary wall; (c) 1 min rinse with deionized water; and (d) 2 min rinse with separation buffer. All steps were carried out at 20 psi.
Procedure for the enantioseparation of IMA by CE-CFT For all experiments, the 50 mM NaH2PO4 solution at pH 7.0 was employed as background electrolyte. A 0.5% w/v HS-β-CD solution obtained by dilution of the 20% m/v solution in the BGE was employed. IMA solutions (standard solutions for calibration and IMA–HSA mixtures) were injected hydrodynamically at 0.5 psi for 5 s. Before sample injection, the capillary was filled with the 0.5% w/v HS-β-CD solution by applying 10 psi for 1 min. Separation voltage was 15 kV (normal polarity), and the capillary temperature was 40°C. These separation conditions provide a maximum current of 65 μA during the CE measurements.
Procedure for the evaluation of protein binding
Experimental Instrumentation A Beckman P/ACE MDQ capillary electrophoresis system equipped with a diode array detector (Beckman Coulter, Fullterton, CA, USA), and 32Karat software version 8.0 was used throughout. A 50 μm i.d. fused-silica capillary with total and effective lengths of 60.2 and 50 cm, respectively, was employed (Beckman Coulter, Fullterton, CA, USA). The CE system is equipped with a temperature-controlled garage for sample trays, so samples can be warmed during all the experiences. Detection wavelength was fixed at 200 nm for all experiments. Electrophoretic solutions and samples were filtered through 0.45 μm pore size nylon membranes (Micron Separation, Westboro, MA, USA) and degassed in an ultrasonic bath ( JP Selecta, Barcelona, Spain) prior to use. A Crison Micro pH 2000 pH-meter from Crison Instruments (Barcelona, Spain) was employed to adjust the pH of buffer solutions.
An experimental design was planned to evaluate the enantioselective binding of IMA to HSA. In this design, the protein concentration was kept near physiological values and the IMA concentration was varied according to the detector capabilities. Mixtures containing five different racemic IMA concentrations, approximately in the 150–350 μM range, and a fixed HSA concentration of 530 μM were prepared by dilution of the stock solutions in the incubation buffer. All of these mixtures were mixed vigorously and transferred to CE microvials. Mixtures were incubated directly in the CE system at 36.5°C for 30 min, and then injected hydrodynamically in the capillary for the separation. Corrected peak area was employed as response variable for the quantification of the unbound fraction, and with this purpose a calibration curve was constructed for each IMA enantiomer.
Software and calculations From the experimental data, the enantiomers resolution (Rs) was calculated using the following expression:
Chemicals and reagents
1638
All reagents were of analytical grade. Human serum albumin fraction V (code no. A1653) was purchased from Sigma (St. Louis, MO, USA), sodium
wileyonlinelibrary.com/journal/bmc
Copyright © 2015 John Wiley & Sons, Ltd.
Rs ¼
1:18ðt2 t1 Þ w1 þ w2
(1)
Biomed. Chromatogr. 2015; 29: 1637–1642
Enantioselective binding of imazalil to HSA were t1 and t2 are, respectively, the migration times of the first (E1) and second (E2) eluted enantiomers, and w1 and w2 represent the peak widths at 50% peak heights. For the IMA–HSA binding study, routines made/adapted in MATLAB® version 4.2 were used for calculations. The results generated were verified using EXCEL® and STATGRAPHICS®. In some calculations a SIMPLEX algorithm routine was used (nonlinear fitting). To evaluate the IMA–HSA binding, simplifications of the general protein binding model (eqn (2)) that considers m classes of independent active sites on the HSA molecule, each with ni binding sites, were used (AsensiBernardi et al., 2010).
r¼
m b Dd X Kid ¼ ¼ ni P P 1 þ Kid i¼1
(2)
In this equation, r is the fraction of bound enantiomer per molecule of protein, b is the bound concentration of enantiomer and P is the total protein concentration. D and d are the total and free enantiomer concentrations, respectively. All the concentrations must be expressed in molar units (M). For the site i, ni represents the apparent site-stoichometry. Equation (2) assumes negligible nonspecific and noncooperative binding. The first simplification assumes m = 1 (one class of binding site for this molecule) and n1 = 1. Both assumptions should be reasonably valid for low D/P ratios (the usual situation in vivo; Asensi-Bernardi et al., 2010). Concretely, D/P ratios 2.5 was fixed to eliminate a point in a round.
Table 3. Apparent affinity constants, apparent protein binding and enantioselective of the binding of IMA enantiomers to human serum albumin (HSA)
IMA-E1
IMA-E2
Log K1, M1a
PBb
ES
3.4 ± 0.4
57.1 (D ≤ 1 μM) 56.8 (D = 10 μM) 54.4 (D = 100 μM) 40.0 (D ≤ 1 μM) 39.8 (D = 10 μM) 38.2 (D = 100 μM)
2.0
3.1 ± 0.3
a
Data estimated from eqn (3) (median ± MADe). Estimated for D = 0.01, 0.1, 1, 10, 100 μM (total enantiomer concentration) and total protein concentration = 530 μM. K1, Apparent affinity constant; PB, protein binding; ES, enantioselectivity. b
zi ¼
logK 1i median MADe
(5)
Table 3 provides robust estimates of log K1 values obtained from eqn (3); no outliers were found. To verify the parameters n1 and K1, a nonlinear approach based on eqn (4) was employed. This is an alternative to other linear models having recognized mathematical inconsistencies and risks (Barri et al., 2008; Fielding et al., 2005). The SIMPLEX search was started with n1 = 1 and log K1 = 3 (based on estimates on Table 3), and only integer n1 values were allowed. Estimated log K1 values were 3.19 and 3.08 for E1 and E2, respectively, which is in agreement with values in Table 3. Moreover n1 = 1 was estimated for both enantiomers, verifying the assumption of eqn (3). Figure 3 shows the d vs D models for both enantiomers, obtained with eqn (4) (values of R2 > 0.8 were obtained for the fitted nonlinear models). The results suggest the adequacy of the estimations performed with eqn (3). Enantioselectivity (ES) was estimated as the ratio of KE1 (2512) and KE2 (1259), providing an ES value of 2.0, in favour of the first eluted enantiomer. This result is the first evidence of a moderate enantioselective binding of IMA to HSA. From a pharmacokinetic point of view, an interesting parameter is the percentage of protein binding (PB, %). Using P = 530 μM, the K1 values in Table 3 and different D values according to different intoxication levels with the plaguicide, d values were estimated for each enantiomer by means of eqn (4) and used for the apparent PB estimations with the following equation: PB ¼ 100
Dd D
(6)
Apparent estimated PB values for the different D situations, as well as the estimation of ES, are also shown in Table 3. These results, which demonstrate a moderate enantioselective binding of the pesticide IMA to the HSA, show the possibility of using this protein as chiral selector for the enantioseparation of IMA in future applications using affinity electrokinetic chromatography. In this technique the xenobiotic–HSA interaction is a prerequisite for enantioresolution. When the analyte and the analyte–protein complex have different mobilities, the interaction of the compound with the protein results in a change in the net mobility of the analyte. (Martínez-Gómez et al., 2006)
Biomed. Chromatogr. 2015; 29: 1637–1642
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/bmc
1641
Figure 3. d vs D plots (a, b) for the binding of, respectively, IMA-E1 and IMA-E2 to human serum albumin. d vs D plots obtained by SIMPLEX optimization of data to eqn (4).
L. Asensi-Bernardi et al.
Conclusions In this work, the enantiomers of imazalil, a chiral plaguicide, were separated by CE-CFT employing HS-β-CD as chiral selector. The separation method provided a good enantioresolution of 4.4 in a short analysis time (
