www.electrophoresis-journal.com
Page 1
Electrophoresis
Combination of on-line capillary electrophoretic assay with mass spectrometry detection for the study of drug metabolism by cytochromes P450
Monika Langmajerová, Roman Řemínek, Marta Pelcová, František Foret and Zdeněk Glatz
Department of Biochemistry, Faculty of Science and CEITEC – Central European Institute of Technology, Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic Correspondence: Professor Zdeněk Glatz, Department of Biochemistry, Faculty of Science and CEITEC – Central European Institute of Technology, Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic. E-mail: [email protected]. Fax: +420-549492690 List of abbreviations: BPE – base peak electropherogram; CYP – cytochrome P450 enzymes; CYP2C9 – cytochrome P450 2C9 isoform; DC – diclofenac; E – enzyme solution; EIE – extracted ion electropherogram; HDC – 4´-hydroxydiclofenac; HLM – human liver microsomes; IB – incubation buffer; Km´ – apparent Michaelis constant; LADME/Tox – liberation, absorption, distribution, metabolism, excretion, toxicity; NADPH – βnicotinamide adenine dinucleotide phosphate reduced; SL – sheath liquid; S – substrate solution; TB – tolbutamide; HTB – hydroxytolbutamide; TDLFP – transverse diffusion of laminar flow profiles; TIE – total ion electropherogram; Vmax´ – apparent maximum reaction velocity. Keywords: Capillary electrophoresis; Cytochrome P450; Drug metabolism; In-capillary reaction; Mass spectrometry Total number of words: 5934
Received: 12-Aug-2014; Revised: 09-Jan-2015; Accepted: 29-Jan-2015 This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/elps.201400394. This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 2
Electrophoresis
Abstract A new CE-MS method with enzymatic reaction inside the capillary was developed for the study of drug metabolism by cytochromes P450. This automated method, based on the transverse diffusion of laminar flow profiles methodology, is comprised of the injection of substrates and enzyme, their mixing, incubation and separation of the reaction products, all performed by CE, and their detection, identification and quantification by MS. The developed and validated method was finally used to conduct a kinetic study of cytochrome P450 isoform 2C9 or human liver microsomes with diclofenac in order to demonstrate its practical functionality. All the estimated kinetic values – apparent Michaelis constants and apparent maximum reaction velocities were in agreement with literature data obtained using other techniques. In addition, the consumption of reactants was in the tens of nL per analysis. The method’s usability was further demonstrated on tolbutamide, the other probe substrate of cytochrome P450 isoform 2C9. As a result, the method is conceptually applicable for the screening of any other cytochrome P450 isoform and its substrates and inhibitors after adapting the incubation and separation conditions.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 3
Electrophoresis
1. Introduction Drug discovery is a very complicated, time- and money-consuming process [1]. It starts with the identification of a given biological target for a specified disease, mostly an enzyme, with the development of a corresponding assay to evaluate the drug’s action. Progress in organic chemistry based on parallel syntheses produces libraries of thousands of promising hits that should enter the pre-clinical stage. However it has been estimated that almost 70 % of the compounds that reach this stage fail due to basic properties such as liberation, absorption, distribution, metabolism, excretion and toxicology (LADME/Tox) even if they pass in terms of their efficacy [2, 3]. As a result, pharmaceutical companies are moving the LADME/Tox tests to the early stages of drug discovery. This has resulted in a need for innovative highthroughput methods that can provide high-quality and reliable information [4]. One of the important challenges in drug discovery is therefore the study of lead metabolism that can be responsible for the problems with their bioavailability, inter-individual variation and especially drug-drug interactions. First of all, the specific enzymatic systems involved in these processes are identified. The drug metabolising enzymes can be separated into two groups based on the chemical nature of the reaction, namely oxidative and conjugative [5]. Oxidative enzymes involved in so-called Phase I metabolism are mainly comprised of cytochrome P450 enzymes (CYP) and flavin monooxygenases. They catalyse the introduction of a reactive oxygen atom into the mostly lipophilic drug molecule. Phase II enzymes
–
sulfotransferases,
glucuronyltransferases,
glutathione-S-transferases,
etc.
conjugate this modified molecule with a group or molecule to further increase its polarity for better excretion from the body.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 4
Electrophoresis
As a result, the CYP superfamily of enzymes has received the most attention to date, since it is one of the most important systems involved in drug metabolism [6, 7]. According to the literature, the five main isoforms (1A2, 2C9, 2C19, 2D6 and 3A4) are responsible for the major route of metabolism for more than 75 % of commonly prescribed drugs [8]. Their reactions typically involve CYPs, NADPH and oxygen in addition to NADPH-reductase for the electron transfer from NADPH to CYPs. Various in vitro assays of drug metabolism by CYPs have been developed [9]. They differ both in their CYP source – human liver microsomes, liver slices, primary hepatocytes or recombinant enzymes; and in the technique used for metabolite analysis. Numerous methods have been used for this purpose: radiometric, spectrometric and chromatographic with UV, fluorescence or MS detection. From the pharmacological point of view LC with its recent advances, especially columns packed with porous particles ≤ 2 µm combined with ultra-high pressure (UHLC), is a key methodology due to its high-throughput capacity, selectivity, resolution and sensitivity [10]. In the last two decades, CE has become an alternative to LC in drug metabolism studies, in particular due to its minute sample volume requirement, high-throughput capabilities and the possibility of automation [11, 12]. In addition, CE can be easily connected with various detectors – UV-VIS, laser-induced fluorescence, MS, electrochemical and conductometric. Similarly to other enzymatic reactions, two basic CE approaches can be used for CYP reaction monitoring [13]. In a classic off-line CE assay a complete reaction mixture is incubated in the vial and then analysed by CE. In contrast the on-line CE assay utilises the capillary not only as a separation column but also as a reaction chamber. In general, there are two ways to mix the reagents inside the capillary, differing in the principle used for the mixing. In electrophoretically mediated microanalysis (EMMA) the reaction mixture is formed due to differences in the electrophoretic mobilities of its compounds and the application of an electric potential for their interpenetration [14]. Diffusion is used for the This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 5
Electrophoresis
same purpose in at-inlet [15] and transverse diffusion of laminar flow profiles (TDLFP) [16] methodologies. The TDLFP methodology, primarily introduced by Krylov et al., was recently modified by our group to provide the reaction mixture with a uniform concentration of reactants. Solutions are injected by hydrodynamic pressure as a series of repeated consecutive plugs with parabolic profiles due to the laminar nature of the flow inside the capillary. The reactants are then rapidly mixed mainly by transverse diffusion and their concentrations seem to be uniform over most of the reaction mixture volume [17]. The reaction is terminated by the application of voltage and separation of the reaction mixture components. The approach described thus represents a generic method facilitating the rapid nanoliter mixing of the reagents with different physicochemical properties in contrast with EMMA, where many CE parameters have to be modified due to the differing mobilities of the potential reaction partners. In this consequence the main aim of this study was to incorporate MS detection in the TDLFP-based setup, since CE combined with MS or even MS/MS enables more sensitive detection and also provides the option of target identification and quantification and represents thus a universal tool for microscale enzyme assay. Similarly the pharmacologically important cytochrome P450 isoform 2C9 (CYP2C9) and diclofenac (DC), the nonsteroidal anti-inflammatory drug, known as a specific substrate for the CYP2C9 isoform were used. The application of the developed method was further demonstrated using human liver microsomes (HLM) as a source of CYP2C9 and tolbutamide (TB) – the first-generation potassium channel blocker and sulfonylurea oral hypoglycemic drug, the other probe substrate of CYP2C9.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 6
Electrophoresis
2. Materials and methods 2.1. Chemicals and reagents Recombinant CYP2C9 microsomes with P450 reductase, expressed in baculovirus-infected Sf9 cells (protein concentration 3.2 mg·mL-1, diclofenac 4´-hydroxylase activity 18.8 nmol·min-1·nmol-1); HLM, pooled adult male and female (protein concentration 20 mg·mL-1, diclofenac 4´-hydroxylase activity 2.5 nmol·min-1·mg-1); DC, TB, β-nicotinamide adenine dinucleotide phosphate reduced (NADPH), ammonium acetate, sodium phosphate, ammonia solution (28 – 30 % NH3, v/v, ACS reagent), acetic acid (≥99.8%, puriss. p.a., ACS reagent), methanol, isopropanol, acetonitrile (all LC-MS grade) were purchased from Sigma (St. Louis, MO, USA). The 4´-hydroxydiclofenac (HDC) standard was obtained from Becton Dickinson International (Erembodegem, Belgium). All reagents used (when not specified) were analytical reagent grade or the highest purity available and were used without any pretreatment. Millipore Direct Q 5 UV system (Merck, Milford, MA, USA) deionised water was used to prepare all solutions. The incubation buffer (IB) and background electrolyte (BGE) were prepared fresh every 3 days and stored at room temperature. DC and TB were prepared as a 60 mM and 50 mM stock solution in methanol, respectively and stored at -20 °C. The solutions used in the method optimization process and for the kinetic studies were obtained by diluting the stock solution to the corresponding concentration with IB. These procedures ensured that the methanol content in the incubation mixture was below 1 % (v/v), so as to have no effect on CYP2C9 activity. Working solutions of CYP2C9, HLM and NADPH were prepared fresh every day by diluting or dissolving their commercial available preparations with the IB. All working solutions were kept at 4 °C before use.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 7
Electrophoresis
2.2. Instrumentation Analyses were performed in an Agilent 7100 CE System (Waldbronn, Germany) coupled to a Bruker maXis impact QTOF MS (Bruker Daltonics, Bremen, Germany). An approximately 5cm length of the polyimide coating at the end of the bare fused-silica capillary (75 cm length, 75 µm id, 363 µm od) (PolymicroTechnologies, Phoenix, AZ, USA) was burned and removed and the capillary protruded approximately 0.1 mm from the sprayer tip. The connection between CE and MS was achieved via the sheath liquid (SL) using the co-axial ESI interface from Agilent (part no. G1607-60002) and SL was delivered by a Dionex UltiMate 3000 isocratic pump with a degasser (Thermo Scientific, Dreieich, Germany) via a 1:100 sheath flow splitter. Nitrogen 5.0 (nitrogen of a purity of at least 99.999 %) from a gas cylinder served as the collision gas. Nitrogen for nebulization of the sample and drying of the formed droplets was delivered by a Genius NM32LA generator (Peak Scientific, Inchinnan, Scotland, UK). 2.3. Experimental conditions 2.3.1. CE conditions At the beginning of each day, the capillary was flushed with deionised water, 20 % (v/v) ammonia solution, water and finally with the BGE for 1.5, 1.5, 2 and 3 min, respectively. The pre-conditioning before each analysis consisted of flushing with water, 20 % ammonia solution, water and BGE for 1, 1.5, 2 and 5 min. The post-conditioning treatment after analysis was composed of flushing with water for 0.5 min. All conditioning steps were performed with a pressure of about 950 mbar at 37 °C. 30 mM ammonium acetate (pH 8.7) was used as the BGE. The complete injection procedure is described in Section 2.3.2 In-capillary reaction. The capillary temperature was maintained at 37 °C during all steps. Separation was accomplished by the application of a voltage of -22 kV (293 V·cm-1). A linear voltage gradient from 0 kV to -22 kV during the first 0.2 min and a positive pressure of 0.1 bar after 6.5 min of analysis was applied. Under these conditions, a current of 48 – 50 µA was observed.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 8
Electrophoresis
2.3.2. MS conditions The MS detection under optimised conditions (for complete optimisation see Section 3.2) was performed using ESI in positive mode. The ESI needle was grounded and the ESI voltage -5 kV was applied on the MS sampling capillary entrance. The flow rate of the isopropanol-water (1:1, v/v) containing 0.15% ammonia SL was set to 2.5 µL·min -1 (after splitting). The drying gas flow rate was set to 5 L·min-1, the drying temperature to 180 °C, and nebulization gas pressure to 0.2 bar. One precaution was taken to prevent unnecessary contamination of the MS instrument during preconditioning of the capillary – the ESI voltage was switched off during this step until the beginning of the analysis. In addition, since the nebulization gas pressure affected injection and incubation in the capillary (see below), it was reduced to 0.1 bar during these processes. Both these parameters were changed to the general conditions after the first 30 seconds of electropherogram acquisition. The MS/MS experiments were performed using the collision energy of 20 eV for DC and HDC and 25 eV for TB and HTB, and the isolation width was set at 1 m/z. The MS data were acquired in the mass range from 50 to 1600 m/z at a spectra rate of 1 Hz and were processed with Bruker Compass Data Analysis software version 4.1. Analytes were detected as protonated molecular ions [M+H]+ with monoisotopic masses calculated by the IsotopePattern tool – 296.024, 312.019, 271.111 and 287.106 for DC, HDC, TB and HTB, respectively. Quantification was done from extracted ion electropherograms (EIE) with width accuracy ± 0.005 m/z. To achieve this mass accuracy, m/z internal mass recalibration needed to be performed for each analysis. Commonly occurring salt clusters originating from acetic acid present in the SL or phosphate salts from the IB were used. Two types of reference mass lists were used depending on the type of SL. For acidic SL sodium acetate clusters [Na(CH3COONa)n+] were used, which were observed at the time corresponding to the EOF; phosphoric acid clusters [(H3PO4)nH+] were used for alkali SL that were recorded after all reaction mixture compounds. External mass calibration was only performed when the capillary had to be replaced and thus the electrospray was dismounted. Data from kinetic measurements were evaluated by means of the MS Excel (Microsoft, Redmond, WA, USA) and the SigmaPlot 8.02 (SPSS, Chicago, USA) software.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 9
Electrophoresis
2.3.3. In-capillary reaction The conditions of in-capillary reaction such as the composition of IB, the injection procedure, incubation time and concentration of enzyme were carefully optimised (see below) to get the highest yield of the products of CYP2C9 reaction. The solutions of substrates DC or TB and cofactor NADPH – S and enzyme CYP2C9 or HLM – E prepared in the IB (25 mM phosphate buffer (pH 7.4)) were maintained in a cooled tray at 4 °C and transferred to CE sample tray before the start of the analysis as described in the previous published work [17]. The solutions S and E were injected with a hydrodynamic pressure of 15 mbar as consecutive plugs into a capillary previously filled with BGE in this order – S for 3, 3, 3 and 6 s and E for 3 s between S plugs. After each injection, the inlet tip of the capillary was immersed into the IB to prevent contamination between the vials of reactants. The resulting reaction mixture containing 17.5 nM CYP2C9 or 0.5 mg·mL-1 HLM, 1 mM NADPH and the corresponding concentration of DC or TB was incubated for 10 or 5 min in the capillary with a vial of IB on the inlet end. This inlet vial was then replaced with a vial containing the BGE and incubation was terminated by the application of separation voltage. All steps are visualised in the Fig. 1. 3. Results and discussion So far many detection techniques including direct and indirect UV-VIS, laser-induced fluorescence, chemiluminiscence and contactless conductivity detection were combined with the on-line CE enzymatic assays, whereas MS was only used twice. Veuthey et al. even aimed to develop an on-capillary approach for automated CYP assay based on MS, however their methodology only used the capillary as a reaction vessel [18]. The reaction products were pushed out of the capillary and off-line analysed by UHPLC TOF-MS. The real-time combination of EMMA with MS was first described by Van Schepdael et al. who used this approach for an on-line screening of matrix metalloproteinase inhibitors by CE coupled to ITMS [19]. Since the TDLFP methodology represents more versatile tool then EMMA, its possible combination with MS detection was conscientiously optimised in this study for such important enzymatic system such as CYPs.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 10
Electrophoresis
3.1. CE optimisation The original arrangement of the most frequently utilised on-line CE enzymatic assay methodology – EMMA, that is employing the same buffer for the enzymatic reaction and the electrophoretic separation, results in a fundamental restriction because the composition and pH of the BGE have to be optimised for both the enzymatic reaction and the subsequent separation of substrate and product. To overcome this limitation Van Dyck et al. [20] introduced the combination of EMMA with a partial filling technique which can be easily adopted in the TDLFP methodology used. In this setup, part of the capillary is filled with a buffer optimised for the enzymatic reaction whereas the rest of the capillary is filled with the optimal BGE for the separation of substrate and product. This possibility is even more crucial in this study aiming to combine TDLFP with MS detection, since the choice of appropriate BGE is typically limited to volatile buffers. As was mentioned above, the main aim was not only to increase the sensitivity of the previously developed TDLFP-based method [17] by introducing MS detection, but also to extend its application potential in terms of metabolite identification and characterization. Since the previous method used classical UV detection in combination with a 20 mM sodium dihydrogen phosphate, 20 mM disodium tetraborate buffer (pH 8.6) as the BGE and a 50 mM phosphate buffer (7.40) as the IB, it was primarily necessary to find both an appropriate BGE compatible with MS detection and the combination of BGE-IB allowing the separation of reaction mixture compounds, as well. Consequently the buffers composed of acetic acid, ammonium acetate or ammonium formate at various concentrations were tested as the BGE. Based on the results of the IB composition effect evaluation on activity CYP2C9 performed in our previous study [17] 50 mM and 25 mM phosphate buffers (pH 7.4) were tested as the IB. The injection procedure was simplified to a single zone, 15 mbar for 6 s of the sample containing the substrate (20 µM DC), product (10 µM HDC) and cofactor (1 mM NADPH) This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 11
Electrophoresis
was used and the separation was started immediately after sample injection. Non-optimised MS detection was used. ESI in positive mode, 3.5 kV. The flow rate of the methanol-water (1:1 v/v) SL was set to 4 µL·min-1 (after splitting). The drying gas flow rate was set to 5 L·min-1, the drying temperature to 180 °C, and nebulisation gas pressure to 0.4 bar. The application of low pH buffers prepared from acetic acid resulted in the co-migration of DC and HDC. The ammonium formate buffers provided unstable current; the analyses were frequently interrupted by current drops. Therefore further optimisation of the resolution between DC and HDC, and total analysis time was done using ammonium acetate as the BGE at various concentrations (15 – 50 mM) and pH (8.50 – 9.20) which provided stable current. The best results were obtained using 30 mM ammonium acetate (pH 8.70) as the BGE; a new solution was used for every run. 25 mM phosphate (pH 7.4) was chosen as the IB in regard to its lower ionic strength, which is more suitable for both for the given BGE in terms of separation resolution and for MS detection 3.2. MS optimisation To achieve the highest ionisation efficiency of the corresponding analytes, which also leads to their lower detection limits, the effect of several parameters was studied. The simplified injection procedure mentioned above was used with an identical sample as for the BGE optimisation.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 12
Electrophoresis
Preliminary experiments were performed to select the suitable ionisation polarity, since DC contained both acidic and basic groups [21, 22]. Simultaneously the effect of SL composition was studied, because it comprises most of the spraying volume and thus has a major effect on the ionisation. MS analyses were accomplished in both ionisation modes; protonated molecular ions [M+H]+ or deprotonated molecular ions [M-H]- were observed depending on the mode. The signal-to-noise values for DC and HDC as model analytes were taken as a measure of ionisation response. A clearly higher ionisation was obtained for positive mode. After the selection of the appropriate ionisation mode, the influence of the type and volume of modifier added to the SL was investigated while the type of organic solvent and its ratio to water were not changed. A methanol-water solution (1:1) (v/v) containing acetic acid or ammonia in the range of 0.0 – 0.7 % and 0.0 – 0.25 %, respectively was tested. The SL with 0.15 % ammonia resulted in the highest ionisation response (see Supplementary material). Subsequently the nature and proportion of the organic solvent was examined while keeping the concentration of basic modifier constant. Methanol, isopropanol and acetonitrile solutions with water in a 1:1 and 1:2 ratios were tested. The ionisations of DC and HDC were the highest for the isopropanol-water based SL (1:1 v/v). Acetonitrile solutions provided the worst results; in addition it caused changes in the polyimide coating at the end of the capillary which was in contact with the SL. The polyimide coating was extended and pulled apart, which was caused by its swelling as described by Baeuml and Welsch [23]. Lastly, the variation on the SL flow rate was studied since the stability of the spray, the solvent evaporation and the formation of small droplets depend on it [24]. The influence of this parameter was measured in the range 1 – 4 µL·min-1 with 0.5 µL·min-1 increments and no significant differences between the ionisation intensities were observed. The finally selected flow rate of 2.5 µL·min-1 of the SL composed of isopropanol-water 1:1 containing 0.15% (v/v) ammonia was sufficient to keep a stable spray and a smooth TIE baseline, therefore it was used for subsequent MS optimising experiments (see Supplementary material).
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 13
Electrophoresis
The following optimisation step was focused on two important parameters in the ESI settings – the nebulisation gas pressure and the ESI voltage. Both these parameters were changed during the analysis, because different conditions were needed for precondition, injection and incubation, and for separation. Therefore two time segments in the complete analytical procedure were created. The first time segment was used for precondition, injection, incubation and for the first 30 s of separation. The second time segment was used for the rest of the analysis. The simplified injection procedure with the identical sample was used again and the experiments were performed without incubation. First the influence of nebulisation gas pressure (0.0 – 0.6 bar) applied in the first segment, e.g. during precondition and injection, was studied. The application of this pressure caused the formation of hydrodynamic bulk flow, as it was described by Huikko et al. [24], which led to an enhancement of the injected volume and also affected the shape of the sample zone. Therefore increasing the nebulisation gas pressure resulted in higher but also broadened peaks. Next the nebulisation gas pressure was optimised in the second segment in the range of 0.2 – 0.6 bar, which mainly affected the total analysis time, resolution and spray stability. The TIE baseline was unstable at a nebulisation gas pressure greater than 0.3 bar. As a result the first segment with no pressure and the second segment with a pressure of 0.2 bar gave the best results in terms of resolution and total analysis time (see Supplementary material). Similarly the ESI voltage was turned off in the first segment. This arrangement reduced the amount of salts that were introduced into the MS during the preconditioning of the capillary and prevented also the electrokinetic introduction of the analytes. Therefore the ESI voltage was optimized only in the second segment in the range of 3 – 5.5 kV (positive polarity), the highest ionisation was achieved at 5 kV. As a final point the drying gas temperature (from 150 to 250 °C) and flow rate (from 4 to 8 L·min-1) were optimised to obtain the highest ionisation intensities for DC and HDC. The optimal values were found to be 180 °C and 5 L·min-1.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 14
Electrophoresis
3.3. In-capillary reaction optimisation Even though the in-capillary reaction procedure was taken from a previously published work [17] its modification was essential due to the MS detection applied that involved the increase in the capillary length and the change in the used BGE (see above). In addition the irreproducible migration times were observed when an incubation interval of 10 minutes (CYP2C9) or 5 minutes (HLM) was added between the injection and application of voltage. It can be assumed that this was caused by a movement of the injected reactants inside the capillary by siphoning effect that can be easily eliminated in the case of UV by adjusting the liquid levels in inlet and outlet vials. In contrast this solution is not possible with MS detection, since only inlet vial is used. After short optimisation it was found that the application of a minimal nebulization pressure of 0.1 bar prevented the movement of the injected reactants inside the capillary and provided very stable migration times with an RSDs value below 1 %. What is more since CE-MS combination required longer capillary than was necessary for separation using UV detection, external pressure 0.1 bar was applied on the inlet vial from 6.5 minute to the end of analysis. This precaution brought reduction in analysis time about 4 minutes and resulted in the migration times of DC and HDC 7.0 and 7.2, respectively. In addition we found that this pressure had a positive effect on the intensities in MS (see Supplementary material). Typical EIEs obtained under optimised conditions and in-capillary incubation 10 minutes for CYP2C9 or 5 minutes for HLM, are shown in Fig. 2. The substrate – DC and the product – HDC were identified using standards and also using the precise mass of parent and fragment ions from MS/MS spectra [25] (Fig. 3). The formation of the homogeneous reaction mixture was confirmed by simulation using a mathematical model of TDLFP proposed by Krylov et
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 15
Electrophoresis
al. [26]. Using the optimal conditions obtained from the above investigations, the method was also used to evaluate the dependences of HDC production on incubation time. The dependencies found were linear over the tested incubation time from 1 to 10 min for a CYP2C9 concentration of 17.5 nM with correlation coefficient 0.985 (data not shown). To demonstrate the versatility of the developed CE-MS method the metabolism study of another compound was also tested. TB was selected for this purpose since it is the CYP2C9 probe substrate and its metabolism by this isoform was thus guaranteed. The complete analysis of on-capillary incubation with TB including the MS/MS spectra are shown in Fig.4. 3.4. Validation of the developed method After the method optimisation, the validation was performed under the conditions summarised in Section 2.3. Method repeatability was determined by conducting on-line analyses of the complete CYP2C9 and HLM reactions with 5 µM DC and 10 and 5 min incubations, respectively. Linearity and sensitivity were tested by analysing the sample of CYP2C9 or HLM, NADPH and the standard of HDC added; the mixture was analysed immediately without the incubation. Each sample was analysed in triplicate. The external quantification was done without a classic or isotopically labelled internal standard, the calibration graphs were based on the peak areas from the EIEs obtained. The results of validations are summarized in Tab. 1.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 16
Electrophoresis
An excellent repeatability for migration times (RSD = 0.3 % for CYP2C9, RSD = 0.7 % for HLM; n = 10) was estimated, the repeatability for peak areas (RSD = 8.9 % for CYP2C9, RSD = 13.4% for HLM; n = 10) was lower because this parameter is affected not only by the CE-MS analysis but also by the enzymatic reaction itself. Very good linearity of the calibration graphs for HDC were obtained in the range 25 – 250 nM (R2 = 0.999) for CYP2C9 and 25 – 600 nM (R2 = 0.994) for HLM. LOD and LOQ were 7 nM and, 25 nM respectively for CYP2C9, and 6 nM and 23 nM for HLM. The recovery values estimated by adding a known amount of HDC to a sample containing a known concentration of this compound were in an acceptable range (96 – 103 %). In contrast with the former CE-UV method [17], a new CE-MS one achieved LOD 40 times and LOQ 30 times lower, and also slightly better repeatability of migration times. The RSD of the repeatability of peak areas published by Wang et al. who performed the in-capillary reaction of matrix metalloproteinase using the EMMA methodology combined with MS detection were better 3.2 % for classic and 4.8 % for pressure-mediated EMMA [19]. But it is necessary to point out that the CYP represents more complicated enzymatic system in contrast with the matrix metalloproteinase. 3.5. Application to kinetic study of CYP2C9 reaction Since one of the most important factors for the in vitro characterisation of drug biotransformation and their interaction is the kinetic analysis of metabolite formation, the basic kinetic parameters of the CYP2C9 and HLM reaction with DC were also evaluated. Determination of the Michaelis constant (Km) and maximum reaction velocity (Vmax) was performed by changing the concentration of DC in the reaction mixture from 0 to 50 μM for CYP2C9 and from 0 to 100 μM for HLM, the incubation was performed for 10 min with 17.5 nM CYP2C9 or for 5 min with 0.5 mg·mL-1 HLM in the reaction mixture. Each
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 17
Electrophoresis
concentration was analysed in duplicate. Fig. 5 shows the Michaelis-Menten plots of the initial velocities (recalculated on the CYP2C9 or the protein concentration) versus DC concentrations. The apparent kinetic parameters for CYP2C9 Km´ 4.9 ± 0.9 µM, Vmax´ 0.9 ± 0.1 nmol·min-1·nmol-1 and for HLM Km´ 29.2 ± 4.5 µM, Vmax´ 1.3 ± 0.1 nmol·min-1·mg-1 obtained by this method are in relative agreement with previously published values for CYP2C9 – Km´ 2.66 ± 0.18 µM, Vmax´ 7.91 ± 0.22 nmol·min-1·nmol-1 CYP2C9 [17] and Km´ 3.44 ± 0.45 µM, Vmax´ 19.78 ± 0.76 nmol·min-1·nmol-1 CYP2C9 [28], and for HLM – Km´ 4.4 – 28.5 µM, Vmax´ 0.39 – 4.70 nmol·min-1·mg-1 [29]. The better agreement can be seen in the case of Km´, the values are in the same concentration order. The larger variation in the Vmax´ values could be caused by using different enzyme batch or even by using enzyme from different supplier. 4. Concluding remark A new on-line CE-MS method enabling the creation, identification and quantification of CYP reaction products within a single analysis was introduced. The principle of reactant mixing inside the capillary based on TDLFP and the employment of tandem MS detection guarantee the generic applicability of the method regardless of the tested CYP isoform and substrate. Utilisation of MS greatly enhances the effectiveness of CE by more sensitive detection as well as the target identification. What is more the lower LOD and LOQ allow reduction of the incubation time, which fulfils the high-throughput requirement. As a result the promising tool for the screenings of potential drug candidates in the early stages of new drug development is presented. It is characterised by a minimal consumption of enzyme and other chemicals, fully automated analyses and high-throughput capabilities. Acknowledgment This work was supported by grant GBP206/12/G014 from the Czech Science Foundation. The authors have declared no conflict of interest.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 18
Electrophoresis
References [1] Buonansegna, E., Salomo, S., Maier, A. M., Li-Ying, J., R & D Manage. 2014, 44, 189– 202. [2] DiMasi, J. A., Feldman, L., Seckler, A., Wilson, A., Clin. Pharmacol. Ther. 2010, 87, 272–277. [3] Bugrim, A., Nikolskaya, T., Nikolsky, Y., Drug Discov. Today 2004, 9, 127–135. [4] Saunders, K. C., Drug Discov. Today 2004, 1, 373–380. [5] King, R. S., in: Nassar, A. F., Hollenberg, P. F., Scatina, J. (Eds.), Drug metabolism handbook : concepts and applications, John Willey & Sons, Hoboken 2009, pp. 17–40. [6] Guengerich, F. P., in: Ortiz de Montellano, P. R. (Ed.), Cytochrome P450 : Structure, Mechanism, and Biochemistry, Kluwer Academic, New York 2005, pp. 377–530. [7] Nebert, D. W., Russell, D. W., Lancet 2002, 360, 1155–1162. [8] Guengerich, F. P., Chem. Res. Toxicol. 2008, 21, 70–83. [9] Zlokarnik, G., Grootenhuis, P. D. J., Watson, J. B., Drug Discov. Today 2005, 10, 1443– 1450. [10] Youdim, K. A., Saunders, K. C., J. Chromatogr. B 2010, 878, 1326–1336. [11] Shanmuganathan, M., Britz-McKibbin, P., Anal. Chim. Acta 2013, 773, 24–36. [12] Nowak, P., Wozniakiewicz, M., Koscielniak, P., Electrophoresis 2013, 34, 2604–2614. [13] Glatz, Z., J. Chromatogr. B 2006, 841, 23–37. [14] Bao, J. M., Regnier, F. E., J. Chromatogr. 1992, 608, 217–224. [15] Van Dyck, S., Vissers, S., Van Schepdael, A., Hoogmartens, J., J. Chromatogr. A 2003, 986, 303–311. [16] Okhonin, V., Liu, X., Krylov, S. N., Anal. Chem. 2005, 77, 5925–5929. [17] Reminek, R., Zeisbergerova, M., Langmajerova, M., Glatz, Z., Electrophoresis 2013, 34, 2705–2711. This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 19
Electrophoresis
[18] Nicoli, R., Curcio, R., Rudaz, S., Veuthey, J. L., J. Med. Chem. 2009, 52, 2192–2195. [19] Wang, X., Dou, Z. Y., Yuan, Y. Z., Shuli, M. L., Wolfs, K., Adams, E., Van Schepdael, A., J. Chromatogr. B 2013, 930, 48–53. [20] Van Dyck, S., Van Schepdael, A., Hoogmartens, J., Electrophoresis 2001, 22, 1436– 1442. [21] Ahrer, W., Buchberger, W., Monatsh. Chem. 2001, 132, 329–337. [22] Kosjek, T., Heath, E., Perez, S., Petrovic, M., Barcelo, D., J. Hydrol. 2009, 372, 109– 117. [23] Baeuml, F., Welsch, T., J. Chromatogr. A 2002, 961, 35–44. [24] Nilsson, S. L., Bylund, D., Jörntén-Karlsonn, M., Petersson, P., Markides, K. E., Electrophoresis 2004, 25, 2100–2107. [24] Huikko, K., Kotiaho, T., Kostiainen, R., Rapid Commun. Mass Sp. 2002, 16, 1562–1568. [25] Galmier, M. J., Bouchon, B., Madelmont, J. C., Mercier, F., Pilotaz, F., Lartigue, C., J. Pharmaceut. Biomed. 2005, 38, 790–796. [26] Wong, E., Okhonin, V., Berezovski, M. V., Nozaki, T., Waldmann, H., Alexandrov, K., Krylov, S. N., J. Am. Chem. Soc. 2008, 130, 11862–11863. [27] Tozuka, Z., Aoyama, S., Nozawa, K., Akita, S., Oh-Hara, T., Adachi, Y., Ninomiya, S. I., J. Pharm. Sci. 2011, 100, 4024–4036. [28] Konecny, J., Jurica, J., Tomandl, J., Glatz, Z., Electrophoresis 2007, 28, 1229–1234. [29] Yang, J. L., He, M. M., Niu, W., Wrighton, S. A., Li, L., Liu, Y., Li, C., Brit. J. Clin. Pharmaco. 2012, 73, 268–284.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 20
Electrophoresis
Table 1. Validation parameters of developed on-line methods.
4´-hydroxydiclofenac CYP 2C9
HLM
Migration time repeatability (RSD, n = 10)
0.3 %
0.7 %
Peak area repeatability (RSD, n = 10)
8.9 %
13.4 %
25 – 250 nM
25 – 600 nM
Correlation coefficient
0.999
0.994
LOD (S/N ˃ 3)
7 nM
6 nM
LOQ (S/N ˃ 10)
25 nM
23 nM
Linearity
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 21
Electrophoresis
Fig. 1 Analysis procedure scheme. A – Injection of substrate (S) and enzyme (E) zones by pressure. B – Creation of reaction mixture (RM) by transverse diffusion. C – Separation of product (P), S and E. D – Detection of formed product by MS.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 22
Electrophoresis
Fig. 2 Record of analysis after in-capillary incubation with DC. A – 10 minutes incubation with CYP. B – 5 minutes incubation with HLM. EIE for DC – green trace, for HDC – red trace. BPE – blue trace is shown in the insert. The concentration of DC was 5 μM. Asterisks indicate unknown components, which were also found in the blank analysis (incubation time = 0).
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 23
Electrophoresis
Fig. 3 MS/MS spectrum of DC (above) and HDC (below) obtained by QTOF after incapillary incubation. Calculated monoisotopic mass is indicated in blue-colored font. The measured mass is written in black numbers.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 24
Electrophoresis
Fig. 4 Record of analysis after in-capillary incubation with TB and MS/MS spectra of HTB. A – 5 minutes incubation with HLM. EIE for TB – green trace, for HTB – red trace. BPE – blue trace is shown in the insert. The concentration of TB was 200 µM. Asterisks indicate unknown components, which were also found in the blank analysis (incubation time = 0). B – MS/MS spectrum of HTB is accompanied by proposed structural formulae [27]. Calculated monoisotopic mass is indicated in blue-colored font. The measured mass is written in black numbers.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 25
Fig. 5 Michelis-Menten plots for CYP2C9 (A) and HLM (B) reactions with DC.
This article is protected by copyright. All rights reserved.
Electrophoresis
Page 1
Electrophoresis
Combination of on-line capillary electrophoretic assay with mass spectrometry detection for the study of drug metabolism by cytochromes P450
Monika Langmajerová, Roman Řemínek, Marta Pelcová, František Foret and Zdeněk Glatz
Department of Biochemistry, Faculty of Science and CEITEC – Central European Institute of Technology, Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic Correspondence: Professor Zdeněk Glatz, Department of Biochemistry, Faculty of Science and CEITEC – Central European Institute of Technology, Masaryk University, Kamenice 5, 625 00 Brno, Czech Republic. E-mail: [email protected]. Fax: +420-549492690 List of abbreviations: BPE – base peak electropherogram; CYP – cytochrome P450 enzymes; CYP2C9 – cytochrome P450 2C9 isoform; DC – diclofenac; E – enzyme solution; EIE – extracted ion electropherogram; HDC – 4´-hydroxydiclofenac; HLM – human liver microsomes; IB – incubation buffer; Km´ – apparent Michaelis constant; LADME/Tox – liberation, absorption, distribution, metabolism, excretion, toxicity; NADPH – βnicotinamide adenine dinucleotide phosphate reduced; SL – sheath liquid; S – substrate solution; TB – tolbutamide; HTB – hydroxytolbutamide; TDLFP – transverse diffusion of laminar flow profiles; TIE – total ion electropherogram; Vmax´ – apparent maximum reaction velocity. Keywords: Capillary electrophoresis; Cytochrome P450; Drug metabolism; In-capillary reaction; Mass spectrometry Total number of words: 5934
Received: 12-Aug-2014; Revised: 09-Jan-2015; Accepted: 29-Jan-2015 This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/elps.201400394. This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 2
Electrophoresis
Abstract A new CE-MS method with enzymatic reaction inside the capillary was developed for the study of drug metabolism by cytochromes P450. This automated method, based on the transverse diffusion of laminar flow profiles methodology, is comprised of the injection of substrates and enzyme, their mixing, incubation and separation of the reaction products, all performed by CE, and their detection, identification and quantification by MS. The developed and validated method was finally used to conduct a kinetic study of cytochrome P450 isoform 2C9 or human liver microsomes with diclofenac in order to demonstrate its practical functionality. All the estimated kinetic values – apparent Michaelis constants and apparent maximum reaction velocities were in agreement with literature data obtained using other techniques. In addition, the consumption of reactants was in the tens of nL per analysis. The method’s usability was further demonstrated on tolbutamide, the other probe substrate of cytochrome P450 isoform 2C9. As a result, the method is conceptually applicable for the screening of any other cytochrome P450 isoform and its substrates and inhibitors after adapting the incubation and separation conditions.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 3
Electrophoresis
1. Introduction Drug discovery is a very complicated, time- and money-consuming process [1]. It starts with the identification of a given biological target for a specified disease, mostly an enzyme, with the development of a corresponding assay to evaluate the drug’s action. Progress in organic chemistry based on parallel syntheses produces libraries of thousands of promising hits that should enter the pre-clinical stage. However it has been estimated that almost 70 % of the compounds that reach this stage fail due to basic properties such as liberation, absorption, distribution, metabolism, excretion and toxicology (LADME/Tox) even if they pass in terms of their efficacy [2, 3]. As a result, pharmaceutical companies are moving the LADME/Tox tests to the early stages of drug discovery. This has resulted in a need for innovative highthroughput methods that can provide high-quality and reliable information [4]. One of the important challenges in drug discovery is therefore the study of lead metabolism that can be responsible for the problems with their bioavailability, inter-individual variation and especially drug-drug interactions. First of all, the specific enzymatic systems involved in these processes are identified. The drug metabolising enzymes can be separated into two groups based on the chemical nature of the reaction, namely oxidative and conjugative [5]. Oxidative enzymes involved in so-called Phase I metabolism are mainly comprised of cytochrome P450 enzymes (CYP) and flavin monooxygenases. They catalyse the introduction of a reactive oxygen atom into the mostly lipophilic drug molecule. Phase II enzymes
–
sulfotransferases,
glucuronyltransferases,
glutathione-S-transferases,
etc.
conjugate this modified molecule with a group or molecule to further increase its polarity for better excretion from the body.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 4
Electrophoresis
As a result, the CYP superfamily of enzymes has received the most attention to date, since it is one of the most important systems involved in drug metabolism [6, 7]. According to the literature, the five main isoforms (1A2, 2C9, 2C19, 2D6 and 3A4) are responsible for the major route of metabolism for more than 75 % of commonly prescribed drugs [8]. Their reactions typically involve CYPs, NADPH and oxygen in addition to NADPH-reductase for the electron transfer from NADPH to CYPs. Various in vitro assays of drug metabolism by CYPs have been developed [9]. They differ both in their CYP source – human liver microsomes, liver slices, primary hepatocytes or recombinant enzymes; and in the technique used for metabolite analysis. Numerous methods have been used for this purpose: radiometric, spectrometric and chromatographic with UV, fluorescence or MS detection. From the pharmacological point of view LC with its recent advances, especially columns packed with porous particles ≤ 2 µm combined with ultra-high pressure (UHLC), is a key methodology due to its high-throughput capacity, selectivity, resolution and sensitivity [10]. In the last two decades, CE has become an alternative to LC in drug metabolism studies, in particular due to its minute sample volume requirement, high-throughput capabilities and the possibility of automation [11, 12]. In addition, CE can be easily connected with various detectors – UV-VIS, laser-induced fluorescence, MS, electrochemical and conductometric. Similarly to other enzymatic reactions, two basic CE approaches can be used for CYP reaction monitoring [13]. In a classic off-line CE assay a complete reaction mixture is incubated in the vial and then analysed by CE. In contrast the on-line CE assay utilises the capillary not only as a separation column but also as a reaction chamber. In general, there are two ways to mix the reagents inside the capillary, differing in the principle used for the mixing. In electrophoretically mediated microanalysis (EMMA) the reaction mixture is formed due to differences in the electrophoretic mobilities of its compounds and the application of an electric potential for their interpenetration [14]. Diffusion is used for the This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 5
Electrophoresis
same purpose in at-inlet [15] and transverse diffusion of laminar flow profiles (TDLFP) [16] methodologies. The TDLFP methodology, primarily introduced by Krylov et al., was recently modified by our group to provide the reaction mixture with a uniform concentration of reactants. Solutions are injected by hydrodynamic pressure as a series of repeated consecutive plugs with parabolic profiles due to the laminar nature of the flow inside the capillary. The reactants are then rapidly mixed mainly by transverse diffusion and their concentrations seem to be uniform over most of the reaction mixture volume [17]. The reaction is terminated by the application of voltage and separation of the reaction mixture components. The approach described thus represents a generic method facilitating the rapid nanoliter mixing of the reagents with different physicochemical properties in contrast with EMMA, where many CE parameters have to be modified due to the differing mobilities of the potential reaction partners. In this consequence the main aim of this study was to incorporate MS detection in the TDLFP-based setup, since CE combined with MS or even MS/MS enables more sensitive detection and also provides the option of target identification and quantification and represents thus a universal tool for microscale enzyme assay. Similarly the pharmacologically important cytochrome P450 isoform 2C9 (CYP2C9) and diclofenac (DC), the nonsteroidal anti-inflammatory drug, known as a specific substrate for the CYP2C9 isoform were used. The application of the developed method was further demonstrated using human liver microsomes (HLM) as a source of CYP2C9 and tolbutamide (TB) – the first-generation potassium channel blocker and sulfonylurea oral hypoglycemic drug, the other probe substrate of CYP2C9.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 6
Electrophoresis
2. Materials and methods 2.1. Chemicals and reagents Recombinant CYP2C9 microsomes with P450 reductase, expressed in baculovirus-infected Sf9 cells (protein concentration 3.2 mg·mL-1, diclofenac 4´-hydroxylase activity 18.8 nmol·min-1·nmol-1); HLM, pooled adult male and female (protein concentration 20 mg·mL-1, diclofenac 4´-hydroxylase activity 2.5 nmol·min-1·mg-1); DC, TB, β-nicotinamide adenine dinucleotide phosphate reduced (NADPH), ammonium acetate, sodium phosphate, ammonia solution (28 – 30 % NH3, v/v, ACS reagent), acetic acid (≥99.8%, puriss. p.a., ACS reagent), methanol, isopropanol, acetonitrile (all LC-MS grade) were purchased from Sigma (St. Louis, MO, USA). The 4´-hydroxydiclofenac (HDC) standard was obtained from Becton Dickinson International (Erembodegem, Belgium). All reagents used (when not specified) were analytical reagent grade or the highest purity available and were used without any pretreatment. Millipore Direct Q 5 UV system (Merck, Milford, MA, USA) deionised water was used to prepare all solutions. The incubation buffer (IB) and background electrolyte (BGE) were prepared fresh every 3 days and stored at room temperature. DC and TB were prepared as a 60 mM and 50 mM stock solution in methanol, respectively and stored at -20 °C. The solutions used in the method optimization process and for the kinetic studies were obtained by diluting the stock solution to the corresponding concentration with IB. These procedures ensured that the methanol content in the incubation mixture was below 1 % (v/v), so as to have no effect on CYP2C9 activity. Working solutions of CYP2C9, HLM and NADPH were prepared fresh every day by diluting or dissolving their commercial available preparations with the IB. All working solutions were kept at 4 °C before use.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 7
Electrophoresis
2.2. Instrumentation Analyses were performed in an Agilent 7100 CE System (Waldbronn, Germany) coupled to a Bruker maXis impact QTOF MS (Bruker Daltonics, Bremen, Germany). An approximately 5cm length of the polyimide coating at the end of the bare fused-silica capillary (75 cm length, 75 µm id, 363 µm od) (PolymicroTechnologies, Phoenix, AZ, USA) was burned and removed and the capillary protruded approximately 0.1 mm from the sprayer tip. The connection between CE and MS was achieved via the sheath liquid (SL) using the co-axial ESI interface from Agilent (part no. G1607-60002) and SL was delivered by a Dionex UltiMate 3000 isocratic pump with a degasser (Thermo Scientific, Dreieich, Germany) via a 1:100 sheath flow splitter. Nitrogen 5.0 (nitrogen of a purity of at least 99.999 %) from a gas cylinder served as the collision gas. Nitrogen for nebulization of the sample and drying of the formed droplets was delivered by a Genius NM32LA generator (Peak Scientific, Inchinnan, Scotland, UK). 2.3. Experimental conditions 2.3.1. CE conditions At the beginning of each day, the capillary was flushed with deionised water, 20 % (v/v) ammonia solution, water and finally with the BGE for 1.5, 1.5, 2 and 3 min, respectively. The pre-conditioning before each analysis consisted of flushing with water, 20 % ammonia solution, water and BGE for 1, 1.5, 2 and 5 min. The post-conditioning treatment after analysis was composed of flushing with water for 0.5 min. All conditioning steps were performed with a pressure of about 950 mbar at 37 °C. 30 mM ammonium acetate (pH 8.7) was used as the BGE. The complete injection procedure is described in Section 2.3.2 In-capillary reaction. The capillary temperature was maintained at 37 °C during all steps. Separation was accomplished by the application of a voltage of -22 kV (293 V·cm-1). A linear voltage gradient from 0 kV to -22 kV during the first 0.2 min and a positive pressure of 0.1 bar after 6.5 min of analysis was applied. Under these conditions, a current of 48 – 50 µA was observed.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 8
Electrophoresis
2.3.2. MS conditions The MS detection under optimised conditions (for complete optimisation see Section 3.2) was performed using ESI in positive mode. The ESI needle was grounded and the ESI voltage -5 kV was applied on the MS sampling capillary entrance. The flow rate of the isopropanol-water (1:1, v/v) containing 0.15% ammonia SL was set to 2.5 µL·min -1 (after splitting). The drying gas flow rate was set to 5 L·min-1, the drying temperature to 180 °C, and nebulization gas pressure to 0.2 bar. One precaution was taken to prevent unnecessary contamination of the MS instrument during preconditioning of the capillary – the ESI voltage was switched off during this step until the beginning of the analysis. In addition, since the nebulization gas pressure affected injection and incubation in the capillary (see below), it was reduced to 0.1 bar during these processes. Both these parameters were changed to the general conditions after the first 30 seconds of electropherogram acquisition. The MS/MS experiments were performed using the collision energy of 20 eV for DC and HDC and 25 eV for TB and HTB, and the isolation width was set at 1 m/z. The MS data were acquired in the mass range from 50 to 1600 m/z at a spectra rate of 1 Hz and were processed with Bruker Compass Data Analysis software version 4.1. Analytes were detected as protonated molecular ions [M+H]+ with monoisotopic masses calculated by the IsotopePattern tool – 296.024, 312.019, 271.111 and 287.106 for DC, HDC, TB and HTB, respectively. Quantification was done from extracted ion electropherograms (EIE) with width accuracy ± 0.005 m/z. To achieve this mass accuracy, m/z internal mass recalibration needed to be performed for each analysis. Commonly occurring salt clusters originating from acetic acid present in the SL or phosphate salts from the IB were used. Two types of reference mass lists were used depending on the type of SL. For acidic SL sodium acetate clusters [Na(CH3COONa)n+] were used, which were observed at the time corresponding to the EOF; phosphoric acid clusters [(H3PO4)nH+] were used for alkali SL that were recorded after all reaction mixture compounds. External mass calibration was only performed when the capillary had to be replaced and thus the electrospray was dismounted. Data from kinetic measurements were evaluated by means of the MS Excel (Microsoft, Redmond, WA, USA) and the SigmaPlot 8.02 (SPSS, Chicago, USA) software.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 9
Electrophoresis
2.3.3. In-capillary reaction The conditions of in-capillary reaction such as the composition of IB, the injection procedure, incubation time and concentration of enzyme were carefully optimised (see below) to get the highest yield of the products of CYP2C9 reaction. The solutions of substrates DC or TB and cofactor NADPH – S and enzyme CYP2C9 or HLM – E prepared in the IB (25 mM phosphate buffer (pH 7.4)) were maintained in a cooled tray at 4 °C and transferred to CE sample tray before the start of the analysis as described in the previous published work [17]. The solutions S and E were injected with a hydrodynamic pressure of 15 mbar as consecutive plugs into a capillary previously filled with BGE in this order – S for 3, 3, 3 and 6 s and E for 3 s between S plugs. After each injection, the inlet tip of the capillary was immersed into the IB to prevent contamination between the vials of reactants. The resulting reaction mixture containing 17.5 nM CYP2C9 or 0.5 mg·mL-1 HLM, 1 mM NADPH and the corresponding concentration of DC or TB was incubated for 10 or 5 min in the capillary with a vial of IB on the inlet end. This inlet vial was then replaced with a vial containing the BGE and incubation was terminated by the application of separation voltage. All steps are visualised in the Fig. 1. 3. Results and discussion So far many detection techniques including direct and indirect UV-VIS, laser-induced fluorescence, chemiluminiscence and contactless conductivity detection were combined with the on-line CE enzymatic assays, whereas MS was only used twice. Veuthey et al. even aimed to develop an on-capillary approach for automated CYP assay based on MS, however their methodology only used the capillary as a reaction vessel [18]. The reaction products were pushed out of the capillary and off-line analysed by UHPLC TOF-MS. The real-time combination of EMMA with MS was first described by Van Schepdael et al. who used this approach for an on-line screening of matrix metalloproteinase inhibitors by CE coupled to ITMS [19]. Since the TDLFP methodology represents more versatile tool then EMMA, its possible combination with MS detection was conscientiously optimised in this study for such important enzymatic system such as CYPs.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 10
Electrophoresis
3.1. CE optimisation The original arrangement of the most frequently utilised on-line CE enzymatic assay methodology – EMMA, that is employing the same buffer for the enzymatic reaction and the electrophoretic separation, results in a fundamental restriction because the composition and pH of the BGE have to be optimised for both the enzymatic reaction and the subsequent separation of substrate and product. To overcome this limitation Van Dyck et al. [20] introduced the combination of EMMA with a partial filling technique which can be easily adopted in the TDLFP methodology used. In this setup, part of the capillary is filled with a buffer optimised for the enzymatic reaction whereas the rest of the capillary is filled with the optimal BGE for the separation of substrate and product. This possibility is even more crucial in this study aiming to combine TDLFP with MS detection, since the choice of appropriate BGE is typically limited to volatile buffers. As was mentioned above, the main aim was not only to increase the sensitivity of the previously developed TDLFP-based method [17] by introducing MS detection, but also to extend its application potential in terms of metabolite identification and characterization. Since the previous method used classical UV detection in combination with a 20 mM sodium dihydrogen phosphate, 20 mM disodium tetraborate buffer (pH 8.6) as the BGE and a 50 mM phosphate buffer (7.40) as the IB, it was primarily necessary to find both an appropriate BGE compatible with MS detection and the combination of BGE-IB allowing the separation of reaction mixture compounds, as well. Consequently the buffers composed of acetic acid, ammonium acetate or ammonium formate at various concentrations were tested as the BGE. Based on the results of the IB composition effect evaluation on activity CYP2C9 performed in our previous study [17] 50 mM and 25 mM phosphate buffers (pH 7.4) were tested as the IB. The injection procedure was simplified to a single zone, 15 mbar for 6 s of the sample containing the substrate (20 µM DC), product (10 µM HDC) and cofactor (1 mM NADPH) This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 11
Electrophoresis
was used and the separation was started immediately after sample injection. Non-optimised MS detection was used. ESI in positive mode, 3.5 kV. The flow rate of the methanol-water (1:1 v/v) SL was set to 4 µL·min-1 (after splitting). The drying gas flow rate was set to 5 L·min-1, the drying temperature to 180 °C, and nebulisation gas pressure to 0.4 bar. The application of low pH buffers prepared from acetic acid resulted in the co-migration of DC and HDC. The ammonium formate buffers provided unstable current; the analyses were frequently interrupted by current drops. Therefore further optimisation of the resolution between DC and HDC, and total analysis time was done using ammonium acetate as the BGE at various concentrations (15 – 50 mM) and pH (8.50 – 9.20) which provided stable current. The best results were obtained using 30 mM ammonium acetate (pH 8.70) as the BGE; a new solution was used for every run. 25 mM phosphate (pH 7.4) was chosen as the IB in regard to its lower ionic strength, which is more suitable for both for the given BGE in terms of separation resolution and for MS detection 3.2. MS optimisation To achieve the highest ionisation efficiency of the corresponding analytes, which also leads to their lower detection limits, the effect of several parameters was studied. The simplified injection procedure mentioned above was used with an identical sample as for the BGE optimisation.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 12
Electrophoresis
Preliminary experiments were performed to select the suitable ionisation polarity, since DC contained both acidic and basic groups [21, 22]. Simultaneously the effect of SL composition was studied, because it comprises most of the spraying volume and thus has a major effect on the ionisation. MS analyses were accomplished in both ionisation modes; protonated molecular ions [M+H]+ or deprotonated molecular ions [M-H]- were observed depending on the mode. The signal-to-noise values for DC and HDC as model analytes were taken as a measure of ionisation response. A clearly higher ionisation was obtained for positive mode. After the selection of the appropriate ionisation mode, the influence of the type and volume of modifier added to the SL was investigated while the type of organic solvent and its ratio to water were not changed. A methanol-water solution (1:1) (v/v) containing acetic acid or ammonia in the range of 0.0 – 0.7 % and 0.0 – 0.25 %, respectively was tested. The SL with 0.15 % ammonia resulted in the highest ionisation response (see Supplementary material). Subsequently the nature and proportion of the organic solvent was examined while keeping the concentration of basic modifier constant. Methanol, isopropanol and acetonitrile solutions with water in a 1:1 and 1:2 ratios were tested. The ionisations of DC and HDC were the highest for the isopropanol-water based SL (1:1 v/v). Acetonitrile solutions provided the worst results; in addition it caused changes in the polyimide coating at the end of the capillary which was in contact with the SL. The polyimide coating was extended and pulled apart, which was caused by its swelling as described by Baeuml and Welsch [23]. Lastly, the variation on the SL flow rate was studied since the stability of the spray, the solvent evaporation and the formation of small droplets depend on it [24]. The influence of this parameter was measured in the range 1 – 4 µL·min-1 with 0.5 µL·min-1 increments and no significant differences between the ionisation intensities were observed. The finally selected flow rate of 2.5 µL·min-1 of the SL composed of isopropanol-water 1:1 containing 0.15% (v/v) ammonia was sufficient to keep a stable spray and a smooth TIE baseline, therefore it was used for subsequent MS optimising experiments (see Supplementary material).
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 13
Electrophoresis
The following optimisation step was focused on two important parameters in the ESI settings – the nebulisation gas pressure and the ESI voltage. Both these parameters were changed during the analysis, because different conditions were needed for precondition, injection and incubation, and for separation. Therefore two time segments in the complete analytical procedure were created. The first time segment was used for precondition, injection, incubation and for the first 30 s of separation. The second time segment was used for the rest of the analysis. The simplified injection procedure with the identical sample was used again and the experiments were performed without incubation. First the influence of nebulisation gas pressure (0.0 – 0.6 bar) applied in the first segment, e.g. during precondition and injection, was studied. The application of this pressure caused the formation of hydrodynamic bulk flow, as it was described by Huikko et al. [24], which led to an enhancement of the injected volume and also affected the shape of the sample zone. Therefore increasing the nebulisation gas pressure resulted in higher but also broadened peaks. Next the nebulisation gas pressure was optimised in the second segment in the range of 0.2 – 0.6 bar, which mainly affected the total analysis time, resolution and spray stability. The TIE baseline was unstable at a nebulisation gas pressure greater than 0.3 bar. As a result the first segment with no pressure and the second segment with a pressure of 0.2 bar gave the best results in terms of resolution and total analysis time (see Supplementary material). Similarly the ESI voltage was turned off in the first segment. This arrangement reduced the amount of salts that were introduced into the MS during the preconditioning of the capillary and prevented also the electrokinetic introduction of the analytes. Therefore the ESI voltage was optimized only in the second segment in the range of 3 – 5.5 kV (positive polarity), the highest ionisation was achieved at 5 kV. As a final point the drying gas temperature (from 150 to 250 °C) and flow rate (from 4 to 8 L·min-1) were optimised to obtain the highest ionisation intensities for DC and HDC. The optimal values were found to be 180 °C and 5 L·min-1.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 14
Electrophoresis
3.3. In-capillary reaction optimisation Even though the in-capillary reaction procedure was taken from a previously published work [17] its modification was essential due to the MS detection applied that involved the increase in the capillary length and the change in the used BGE (see above). In addition the irreproducible migration times were observed when an incubation interval of 10 minutes (CYP2C9) or 5 minutes (HLM) was added between the injection and application of voltage. It can be assumed that this was caused by a movement of the injected reactants inside the capillary by siphoning effect that can be easily eliminated in the case of UV by adjusting the liquid levels in inlet and outlet vials. In contrast this solution is not possible with MS detection, since only inlet vial is used. After short optimisation it was found that the application of a minimal nebulization pressure of 0.1 bar prevented the movement of the injected reactants inside the capillary and provided very stable migration times with an RSDs value below 1 %. What is more since CE-MS combination required longer capillary than was necessary for separation using UV detection, external pressure 0.1 bar was applied on the inlet vial from 6.5 minute to the end of analysis. This precaution brought reduction in analysis time about 4 minutes and resulted in the migration times of DC and HDC 7.0 and 7.2, respectively. In addition we found that this pressure had a positive effect on the intensities in MS (see Supplementary material). Typical EIEs obtained under optimised conditions and in-capillary incubation 10 minutes for CYP2C9 or 5 minutes for HLM, are shown in Fig. 2. The substrate – DC and the product – HDC were identified using standards and also using the precise mass of parent and fragment ions from MS/MS spectra [25] (Fig. 3). The formation of the homogeneous reaction mixture was confirmed by simulation using a mathematical model of TDLFP proposed by Krylov et
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 15
Electrophoresis
al. [26]. Using the optimal conditions obtained from the above investigations, the method was also used to evaluate the dependences of HDC production on incubation time. The dependencies found were linear over the tested incubation time from 1 to 10 min for a CYP2C9 concentration of 17.5 nM with correlation coefficient 0.985 (data not shown). To demonstrate the versatility of the developed CE-MS method the metabolism study of another compound was also tested. TB was selected for this purpose since it is the CYP2C9 probe substrate and its metabolism by this isoform was thus guaranteed. The complete analysis of on-capillary incubation with TB including the MS/MS spectra are shown in Fig.4. 3.4. Validation of the developed method After the method optimisation, the validation was performed under the conditions summarised in Section 2.3. Method repeatability was determined by conducting on-line analyses of the complete CYP2C9 and HLM reactions with 5 µM DC and 10 and 5 min incubations, respectively. Linearity and sensitivity were tested by analysing the sample of CYP2C9 or HLM, NADPH and the standard of HDC added; the mixture was analysed immediately without the incubation. Each sample was analysed in triplicate. The external quantification was done without a classic or isotopically labelled internal standard, the calibration graphs were based on the peak areas from the EIEs obtained. The results of validations are summarized in Tab. 1.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 16
Electrophoresis
An excellent repeatability for migration times (RSD = 0.3 % for CYP2C9, RSD = 0.7 % for HLM; n = 10) was estimated, the repeatability for peak areas (RSD = 8.9 % for CYP2C9, RSD = 13.4% for HLM; n = 10) was lower because this parameter is affected not only by the CE-MS analysis but also by the enzymatic reaction itself. Very good linearity of the calibration graphs for HDC were obtained in the range 25 – 250 nM (R2 = 0.999) for CYP2C9 and 25 – 600 nM (R2 = 0.994) for HLM. LOD and LOQ were 7 nM and, 25 nM respectively for CYP2C9, and 6 nM and 23 nM for HLM. The recovery values estimated by adding a known amount of HDC to a sample containing a known concentration of this compound were in an acceptable range (96 – 103 %). In contrast with the former CE-UV method [17], a new CE-MS one achieved LOD 40 times and LOQ 30 times lower, and also slightly better repeatability of migration times. The RSD of the repeatability of peak areas published by Wang et al. who performed the in-capillary reaction of matrix metalloproteinase using the EMMA methodology combined with MS detection were better 3.2 % for classic and 4.8 % for pressure-mediated EMMA [19]. But it is necessary to point out that the CYP represents more complicated enzymatic system in contrast with the matrix metalloproteinase. 3.5. Application to kinetic study of CYP2C9 reaction Since one of the most important factors for the in vitro characterisation of drug biotransformation and their interaction is the kinetic analysis of metabolite formation, the basic kinetic parameters of the CYP2C9 and HLM reaction with DC were also evaluated. Determination of the Michaelis constant (Km) and maximum reaction velocity (Vmax) was performed by changing the concentration of DC in the reaction mixture from 0 to 50 μM for CYP2C9 and from 0 to 100 μM for HLM, the incubation was performed for 10 min with 17.5 nM CYP2C9 or for 5 min with 0.5 mg·mL-1 HLM in the reaction mixture. Each
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 17
Electrophoresis
concentration was analysed in duplicate. Fig. 5 shows the Michaelis-Menten plots of the initial velocities (recalculated on the CYP2C9 or the protein concentration) versus DC concentrations. The apparent kinetic parameters for CYP2C9 Km´ 4.9 ± 0.9 µM, Vmax´ 0.9 ± 0.1 nmol·min-1·nmol-1 and for HLM Km´ 29.2 ± 4.5 µM, Vmax´ 1.3 ± 0.1 nmol·min-1·mg-1 obtained by this method are in relative agreement with previously published values for CYP2C9 – Km´ 2.66 ± 0.18 µM, Vmax´ 7.91 ± 0.22 nmol·min-1·nmol-1 CYP2C9 [17] and Km´ 3.44 ± 0.45 µM, Vmax´ 19.78 ± 0.76 nmol·min-1·nmol-1 CYP2C9 [28], and for HLM – Km´ 4.4 – 28.5 µM, Vmax´ 0.39 – 4.70 nmol·min-1·mg-1 [29]. The better agreement can be seen in the case of Km´, the values are in the same concentration order. The larger variation in the Vmax´ values could be caused by using different enzyme batch or even by using enzyme from different supplier. 4. Concluding remark A new on-line CE-MS method enabling the creation, identification and quantification of CYP reaction products within a single analysis was introduced. The principle of reactant mixing inside the capillary based on TDLFP and the employment of tandem MS detection guarantee the generic applicability of the method regardless of the tested CYP isoform and substrate. Utilisation of MS greatly enhances the effectiveness of CE by more sensitive detection as well as the target identification. What is more the lower LOD and LOQ allow reduction of the incubation time, which fulfils the high-throughput requirement. As a result the promising tool for the screenings of potential drug candidates in the early stages of new drug development is presented. It is characterised by a minimal consumption of enzyme and other chemicals, fully automated analyses and high-throughput capabilities. Acknowledgment This work was supported by grant GBP206/12/G014 from the Czech Science Foundation. The authors have declared no conflict of interest.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 18
Electrophoresis
References [1] Buonansegna, E., Salomo, S., Maier, A. M., Li-Ying, J., R & D Manage. 2014, 44, 189– 202. [2] DiMasi, J. A., Feldman, L., Seckler, A., Wilson, A., Clin. Pharmacol. Ther. 2010, 87, 272–277. [3] Bugrim, A., Nikolskaya, T., Nikolsky, Y., Drug Discov. Today 2004, 9, 127–135. [4] Saunders, K. C., Drug Discov. Today 2004, 1, 373–380. [5] King, R. S., in: Nassar, A. F., Hollenberg, P. F., Scatina, J. (Eds.), Drug metabolism handbook : concepts and applications, John Willey & Sons, Hoboken 2009, pp. 17–40. [6] Guengerich, F. P., in: Ortiz de Montellano, P. R. (Ed.), Cytochrome P450 : Structure, Mechanism, and Biochemistry, Kluwer Academic, New York 2005, pp. 377–530. [7] Nebert, D. W., Russell, D. W., Lancet 2002, 360, 1155–1162. [8] Guengerich, F. P., Chem. Res. Toxicol. 2008, 21, 70–83. [9] Zlokarnik, G., Grootenhuis, P. D. J., Watson, J. B., Drug Discov. Today 2005, 10, 1443– 1450. [10] Youdim, K. A., Saunders, K. C., J. Chromatogr. B 2010, 878, 1326–1336. [11] Shanmuganathan, M., Britz-McKibbin, P., Anal. Chim. Acta 2013, 773, 24–36. [12] Nowak, P., Wozniakiewicz, M., Koscielniak, P., Electrophoresis 2013, 34, 2604–2614. [13] Glatz, Z., J. Chromatogr. B 2006, 841, 23–37. [14] Bao, J. M., Regnier, F. E., J. Chromatogr. 1992, 608, 217–224. [15] Van Dyck, S., Vissers, S., Van Schepdael, A., Hoogmartens, J., J. Chromatogr. A 2003, 986, 303–311. [16] Okhonin, V., Liu, X., Krylov, S. N., Anal. Chem. 2005, 77, 5925–5929. [17] Reminek, R., Zeisbergerova, M., Langmajerova, M., Glatz, Z., Electrophoresis 2013, 34, 2705–2711. This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 19
Electrophoresis
[18] Nicoli, R., Curcio, R., Rudaz, S., Veuthey, J. L., J. Med. Chem. 2009, 52, 2192–2195. [19] Wang, X., Dou, Z. Y., Yuan, Y. Z., Shuli, M. L., Wolfs, K., Adams, E., Van Schepdael, A., J. Chromatogr. B 2013, 930, 48–53. [20] Van Dyck, S., Van Schepdael, A., Hoogmartens, J., Electrophoresis 2001, 22, 1436– 1442. [21] Ahrer, W., Buchberger, W., Monatsh. Chem. 2001, 132, 329–337. [22] Kosjek, T., Heath, E., Perez, S., Petrovic, M., Barcelo, D., J. Hydrol. 2009, 372, 109– 117. [23] Baeuml, F., Welsch, T., J. Chromatogr. A 2002, 961, 35–44. [24] Nilsson, S. L., Bylund, D., Jörntén-Karlsonn, M., Petersson, P., Markides, K. E., Electrophoresis 2004, 25, 2100–2107. [24] Huikko, K., Kotiaho, T., Kostiainen, R., Rapid Commun. Mass Sp. 2002, 16, 1562–1568. [25] Galmier, M. J., Bouchon, B., Madelmont, J. C., Mercier, F., Pilotaz, F., Lartigue, C., J. Pharmaceut. Biomed. 2005, 38, 790–796. [26] Wong, E., Okhonin, V., Berezovski, M. V., Nozaki, T., Waldmann, H., Alexandrov, K., Krylov, S. N., J. Am. Chem. Soc. 2008, 130, 11862–11863. [27] Tozuka, Z., Aoyama, S., Nozawa, K., Akita, S., Oh-Hara, T., Adachi, Y., Ninomiya, S. I., J. Pharm. Sci. 2011, 100, 4024–4036. [28] Konecny, J., Jurica, J., Tomandl, J., Glatz, Z., Electrophoresis 2007, 28, 1229–1234. [29] Yang, J. L., He, M. M., Niu, W., Wrighton, S. A., Li, L., Liu, Y., Li, C., Brit. J. Clin. Pharmaco. 2012, 73, 268–284.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 20
Electrophoresis
Table 1. Validation parameters of developed on-line methods.
4´-hydroxydiclofenac CYP 2C9
HLM
Migration time repeatability (RSD, n = 10)
0.3 %
0.7 %
Peak area repeatability (RSD, n = 10)
8.9 %
13.4 %
25 – 250 nM
25 – 600 nM
Correlation coefficient
0.999
0.994
LOD (S/N ˃ 3)
7 nM
6 nM
LOQ (S/N ˃ 10)
25 nM
23 nM
Linearity
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 21
Electrophoresis
Fig. 1 Analysis procedure scheme. A – Injection of substrate (S) and enzyme (E) zones by pressure. B – Creation of reaction mixture (RM) by transverse diffusion. C – Separation of product (P), S and E. D – Detection of formed product by MS.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 22
Electrophoresis
Fig. 2 Record of analysis after in-capillary incubation with DC. A – 10 minutes incubation with CYP. B – 5 minutes incubation with HLM. EIE for DC – green trace, for HDC – red trace. BPE – blue trace is shown in the insert. The concentration of DC was 5 μM. Asterisks indicate unknown components, which were also found in the blank analysis (incubation time = 0).
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 23
Electrophoresis
Fig. 3 MS/MS spectrum of DC (above) and HDC (below) obtained by QTOF after incapillary incubation. Calculated monoisotopic mass is indicated in blue-colored font. The measured mass is written in black numbers.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 24
Electrophoresis
Fig. 4 Record of analysis after in-capillary incubation with TB and MS/MS spectra of HTB. A – 5 minutes incubation with HLM. EIE for TB – green trace, for HTB – red trace. BPE – blue trace is shown in the insert. The concentration of TB was 200 µM. Asterisks indicate unknown components, which were also found in the blank analysis (incubation time = 0). B – MS/MS spectrum of HTB is accompanied by proposed structural formulae [27]. Calculated monoisotopic mass is indicated in blue-colored font. The measured mass is written in black numbers.
This article is protected by copyright. All rights reserved.
www.electrophoresis-journal.com
Page 25
Fig. 5 Michelis-Menten plots for CYP2C9 (A) and HLM (B) reactions with DC.
This article is protected by copyright. All rights reserved.
Electrophoresis
