Developing a new drug from original ideea to the launch of a final product is a complex investigation, which can lead to failures if the new chemical… [618405]

1
I. GENERAL PART
I.1. Introduction
Developing a new drug from original ideea to the launch of a final product is a
complex investigation, which can lead to failures if the new chemical entity (NCE) lack s
efficiency or safety [1]. For this reason , preclinical stage of this process has a great
importance, consisting of different steps : initial target identification and validation, assay
development, high troughput screening, hit identification, lead optimization and selection of a
candidate molecule for clinica l development (Figure 1 .1.).

Since the efficacy and toxicity of drugs is closely related to their pharmacokinetics, a
better understanding of their metabolics pathways during the hit to lead optimisation stage is
essential [2]. In this context, in vitro methods for drug metabolism studies became mandatory.
Firstly, they can be used in the early preclinical stages and secondly, it is possible to use
human enzymes, cells and liver fractions, making the data more relevant for the human in vivo
studies [3].
Moreover, for a better understanting and a good quantification of all the processes
involved, a potent analitical methodology needs to be employed [2]. Nowayadays, there are
multiple techniques available. Among them, capillary electrophor esis offers fast analysis, with
very high peak efficiencies, while respecting the critical trends of bioanalytical assays,
automation and miniatu rization.

Basic
Research

Research Lead
Discovery Preclinical
Development Clinical
Development FDA
Filing T\Target ID
& Selection Candidate
selection IND
filing NDA
filing
Years 3 1 6 1.5
Figure 1 .1. Drug discovery process [1]

2
I.2. Metabolisation studies
I.2.1. Drug metabolism
The study of absorption, distribution, metaboli sm, excretion and pharmacokinetics
(ADME/PK) has developed into a relatively mature discipline in drug discovery . Drug
metabolism is a biochemical process of conversion of a lipophilic compound to a more
hydrophilic one, which can be easily eliminated fro m the human or animal body [3].

In general, drug metabolism can be divided i nto Phase I and Phase II (Fig ure 1 .2.).
Phase I involves reduction , oxydation and hydrolysis reactions mediated mainly by
cytochromes P450 ( CYPs) and to a lesser extent by f lavin -containing monooxygenases
(FMOs) [4]. Phase II metabolic enzymes, as UDP -glucuronyltransferases or sulfotransferases,
catalyze conjugation reactions of lipophilic chemicals with endogenous co -factors, like
uridine diphosphate -glucuronic acid [3]. T his leads to a increased solubility of drug
metabolite in water, hence making them more conveniently excreted from the organism [4].
However, CYPs are the dominant enzyme systems that control drug metabolism and
clearance, accounting 75% of the biotransfo rmation of marketed pharmaceuticals. They
belong to a superfamily of heme -containing proteins that generally catalyze the transfer of a
single oxygen atom onto a xenobiotic or an endogenous subtrate, e.g. steroids, fatty accids or
prostaglandins [5]. At th e same time, they transport electrons from nicotinamide adenine
dinucleotide phosphate -oxidase (NADPH) to adrenodoxin in microsome [6]. Due to the
content in human liver and high contribution to drug metabolism, only few isoforms of CYP
are of great import ance for must drug metabolism assays. Figure 1.2. Drug metabolism [4]

3
The most studied ones are: 1A family, 2A6, 2B6, 2C family, 2D6, 2E1 and 3A family
[3]. It is of great importance to notice that genetic polymorphism of some of these isoformes,
e.g., CYP2D6, CYP2C9 or CYP2C19, along w ith factors like age,gender and nutritional
habits, are at the base of the observed variability in CYP450 -mediated drug clearance [5].
I.2.2. CYP 1 family
In the past few years, a new CYP family presented interest for the scientific world,
regarding it’s polymorphism and regulation activity. CYP 1 family consists of CYP1A1,
CYP1A2 and CYP1B1. In human body, CYP 1A1 is found mainly in extrahepatic tissues, as
pancreas, thymus, uterus or small intestine, with a great role in the metabolism of xenobiotics
[6]. It is also known as one of the most important enzymes in the bioactivation of
procarcino genes to generate reactive metabolites [7]. In addition, CYP 1B1 is expressed in
extrahepatic tissues such as the prostate, breast and uterus and overexpressed in t umor tissue
[8].
Their transcription is regulated through aryl hydrocarbon receptor, who is acting like a
ligand -activated transcription factor. A series of xenobiotics acts as binding ligands:
halogenated aromatic hydrocarbons, benzo(α)pyrene and polycycli c aromatic hydrocarbons
[6]. This generates the increase of transcription , CYP 1A1 mRNA levels being very high in
the lung cells of smokers, comparing with non -smokers [7]. It is overexpressed also in breast
cancer, liver, bladder and colon [9,10].
Conseq uently, CYP1A1 in now considered an interesting target with respect at cancer
therapy.
I.2.3. In vitro metabolism studies
In order to be able to detect which enzyme is directly involved in the metabolisation
process, the in vitro approach uses different models, working with: fresh and cryopreserved
hepatocytes, tissue slices human liver microsomes [3] or recombinant expressed enzymes
(supersomes ) [11].
Tissue slices, along with fresh or cryopreserved hepatocytes, are the most complex in
vitro model, reuni ting phase I and phase II metabolic enzymes and co -factors . If the first ones
are really useful in the study formation of metabolites, the second ones can estimate and
predict the clearance and drug -drug interactions [3].

4
Human liver microsomes provide a more convenient way to study CYP -mediated
metabolism. They are a subcellular fraction of tissue obtained by differential high -speed
centrifugation [12] , coming from single or multiple individuals [11] . As a result, all of C YP
enzymes are to be found in th e microsomial fraction , but not phase II activities [13] . An
NADPH -regenerating system must be added to this system, in order to preserve the co -factor
function. By comparison with liver slices, they are more rapidly available, although the last
ones gives the most closely model to the in vivo situation [12]. Commercial microsomes can
be purchased from different companies, but a significant variability of their activity has been
noticed from vendor to vendor or batch to batch [14]. Therefore, it is recommen ded to check
the quality of the purchased batches. One simple method to verify their conformity is to run
an UV spectrum for carbon monoxide binding difference. The appearance of a peak at 420
nm, instead at 450 nm, shows the starting degradation of the product [15].
In order to prevent a decrease in the enzymatic activity, human liver microsomes
should be stored at -80șC [16].Those that are thawed and maintained on ice for less than 2
hours, can be re -frozen and re -used. The major disadvantage is that aft er 2 hours of incubation
at 37 șC, their enzymatic activity is lost [3] and the results obtained are not representatives.
Supersomes, named also enzyme -only systems, allow the study of a single metabolic
enzyme . As in the case of human liver microsomes, t he specificity is increased, this time, at a
maximum level. If at the begging, they were obtained from recombinant proteins in insect
cells [11], molecular biology techniques allows nowadays the successful cloning and
expression of a large number of human CYP isoenzymes, commercially available [3].
Another important characteristic of microsomes and supersomes is their compatibility
with modern analyzing techniques. There are two types of approaches: off -line biocatalytical
systems, based on, e.g. fluoresce nt and radioactive labeling of metabolites, immunoenzymatic
techniques and on -line biocatalytical systems. The last ones are performed in flow -through
system, the reaction being directly coupled with separation and detection of reactants by
liquid chromato graphy (LC) or capillary electrophoresis (CE) [4].

5
I.3. Capillary electrophoresis
I.3.1. Introduction
One of the earliest demonstrations of the advantages of CE was provided by Virtanen
(Helsinki University of Technology, Finland), in 1974. The author described a potentiometric
detection method for zonal capillary electrophoresis , applied to the quantitative analysis of the
alkali cations Li+, Na+ sau K+. The application area grew since then, including nowadays ionic
or non -ionic compounds, organic or inorganic, regardless of their molecular weight [17]. This
physical method of analysis is based on the migration, inside a capillary, of charged analytes
dissolved in an electrolyte solution, under the influence of a direct -current electric field [18].
Nowadays, CE has interesting apliquations in pharmacology -related assays, enabling the
study of proteins and proteins interaction under their native forme, in physiological conditions
[2]. Because this technique allows miniaturiz ation, CE -based enzyme assays are extensively
employed for enzyme kinetics and enzyme inhibition studies [20]. It is also a great alternative
to LC, offering uniques advantages : short analysis time, low reagent cost, minimal sample
requirement [19], high efficiency separations and ab ility to use several detection methods
[21].

I.3.2. In-capillary assays

Initially, CE was employed only as a separation tool in enzyme assays. Typically, the
enzyme -catalyzed reactions were performed off -line, the incubation of substr ate, cofactor and
enzyme being performed out of the capillary, in a separate vial. However, there are some
limitations. First of all, there is a time delay between the time of the reaction and the time of
the analysis. The enzymatic reaction must be stopped, by adding a new reagent to the
solution. In addition, to avoid capillary clogging or protein adsorption onto the capillary wall,
deproteinization prior to sample injection must be effectuated. For this purpose, centrifugation
and precipitation of the proteins are mandator y. Secondly, even though only a nanoliter scale
sample is needed for CE analysis, much larger amounts of starting materials are used [19, 22,
23].
More recently CE was employed as an on -line microreactor. In this kind of assays the
capillary has multiple functions; it serve s as a minireactor at a nanoliter scale, as a separation
column for both reaction products and remaining products and as a detection cell. Moreover,
the enzymatic reaction is conduct ed directly inside the capillary.

6

I.3.3. Electrophore tically mediated microanalysis

I.3.3.1. Theoretical backround

There are two methodologies to study on -line drug metabolism: the heterogeneous
mode and the homogenous mode [4]. The heterogeneous mode contains one immobilized
reactant, most often the enzym e, on the capillary wall while the others components of the
reaction mixture are present in the liquid phase or background electrolyte (BGE). On the
contrary, in homogeneous assays, all reactants are present in solution [23].
The first in -line capillary e lectrophoretic enzymatic assays were described by Bao and
Regnier [ 24]. In their study, a plug of enzyme (glucose -6-phosphate dehydrogenase) was
injected into a capillary filled with the substrate, glucose -6-phosphate and the cofactor
nicotinamide adenine dinucleotide phosphate (NADP) . The product, NADPH, was formed as
the enzyme migrated under the influence of the electric field applied, and was detected at a
downstream detector at 340 nm. This type of assay gained the name of “electrophoretically
mediated microanalysis” or EMMA. It is important to notice that the enzyme and the substrate
injected had different electrophoretic mobilities, which made possible their mixing, their
separation and quantification. By all means, EMMA defines herself as a homogeneo us
enzyme assay. This approach offers automation, minituarization and eliminates samples
handling.

I.3.3.2. EMMA modes

In general, two major EMMA formats are used to load and mix reagents in a capillary,
namely continuous engagement EMMA or long contact mode and transient engagement
EMMA, known also as plug-plug format . In the first one, the capillary is initially completely
filled with one of the reactants, whereas the ot her reactant is introduced as a plug (zonal
sample introduction , Fig ure 1.3. A.) or continuously from the inlet vial (moving boundary
sample introduction) [23]. Product formation begins under the influence of a voltage mixing,
being maintained until the active enzyme is present in the capillary. It is important to
emphasize that in the in let vial is maintained the faster migrating reactant, so that he can
interpenetrate with the slower migrating reactant, present inside of the formed micro -reactor
[22].

7
The plug -plug mode ( or short contact mode, Fig ure 1. 3.B.) consists in the introduction
of the enzyme and substrate as distinct plugs, the one with a lower electrophoretic mobility
being injected first. The enzymatic reaction takes place also at the application of an electric
field, when the zones overlap. As a result, the product(s) and the unreacted substrates are
electrophoretically transported toward the detector. There is one issue with this methodology
because the electrophoretic conditions (as the composition and pH of the BGE) must be
favorable for both the enzymatic reaction and the s eparation itself. If not, the enzyme may
lose its activity when in contact with the buffer [19].
To overcome this limitation, Van Dyck and coworkers [2 6] introduced the partial
filling technique (Figure 1.3. C.). In this setup, the capillary has two filli ngs: a buffer for the
enzymatic reaction and one for the separation and migration of the resulted products.
Moreover, there are three ways of mixing the reagent plugs. Firstly, if we deal with an
enzyme that is not resistant at an electric field, the reac tion takes place under no applied
voltage. The reactants are injected in a sandwich plug mode (substrate -enzyme -substrate) and
their mixing is realized by longitudinal diffusion (Figure 1. 3.D.). But, when working with
multiple plugs, this method seems to n ot be efficient. Hereby, mixing by transverse dif fusion
Figure 1.3 . EMMA modes [25]

8
of laminar flow profiles (TDLFP) has been proposed (Fig ure 1.3.E.) [27 ,28]. Solution s of
reactants are injected in the capillary by pressure as a series of consecutive plugs. Due to the
laminar nature of flow inside the capillary, the non –diffused plugs have parabolic profiles
with predominantly longitudinal interfaces between them. After injection, the plugs are mixed
by transverse diffusion and not by longitudinal diffusion.
In addition, to further increase the efficiency of in -capillary EMMA assays, Sanders et
al. [29 , 30] described a procedure based on the plug -plug mode, which allows successive
mixing by rapid polarity switches (RPS). Post -injection, a series of positive and negative
potentials ar e applied, the backward and forward movements being analogous to a mechanical
shaking. When carried out with attention to the injected plug size and to the rate of
electroosmotic flow, RPS approach gives rise to reproducibly larger product peaks and
accura te determination of molecules involved .

I.3.3.3. Development and optimization

The development of an EMMA method is challenging, due to the fact that it implies
optimized procedures for both the enzymatic reaction and the analytes separation. Although
there are a series of papers describing EMMA enzymatic assays, little work has been done on
the study and optimization of the specific parameters of an in -line reaction [2].

I.3.3.3.1. Key factors

Experimental factors, such as mixing voltage, enzyme concent ration and mixing time
of reaction at the applied voltage, should be considered when perfecting a method. Stahl et al.
[31] proved the importance of a design of experiment, investigating the effects of conductivity
on zone overlap, by computer simulation a nd by experiment. The difference between the two
approaches was less than 10%, meaning that the influence of each factor is unique and really
important to study.
Another group [32 ] studied the effects of reagent zone a nd buffer conditions for the
determination of creatinine in urine and blood sample. Through judicious selection of reagent
plug length, BGE concentration and pH, they improved the sensitivity of the method.
Furthermore, an important issue to consider when working with proteins in CE is their
adsorption to the capillary wall, which can lead to the clogging of the capillary [23]. Luckily,
artificial microsomes containing recombinant DMEs are not as protein -rich as purified

9
preparations of natural hepatic cells. A good rinsing step between runs h as turned out to be
sufficient to avoid disturbance of results. Yet, some authors have suggested other possible
ways to avoid alteration of the capillary. Pawels et al. [32 ] used multi -walled carbon
nanotubes (MWNTs) to improve capillary lifetime. This see ms to have great application when
working with large amounts of proteins or long incubation times. MWNTs not only prevented
the interaction between silanol groups and proteins, but they also improved the repeatability
of peak’s symmetry .
Another important factor to look after is the temperature of the electrolyte [23, 32]. If
the most conventional off -line assays are taking place at room temperature, 25 șC, in the case
of on -line systems, temperature adjustment is more complicated. Commonly, enzymatic
react ions should take place at physiological temperature, 37 șC , but in order to be fully
productive, sometimes they demand higher temperatures. Even if the capillary, for the most of
its length, is placed in contact with a thermostabilized heat exchanger, ther e are parts that are
not actively temperature -controlled , as the inlet . An ingenious solution has been proposed for
this problem [33] , moving the sample through the non -controlled -temperature inlet into the
controlled -temperature region by pressure or by a low-strength electric field. In their study,
the tray was also thermostated, so that the molecules avoid the exposure to the elevated
temperatures.

I.3.3.3.2. Injection, separation and detection

The duration of the analysis in CE is given by the migrati on time of the analyte s [34].
There are several approaches that can be used to increase the throughput of this technique:
– Reduc ing the length of the capillary;
– Increasing the separation voltage;
– Modifying the electrophoretic and electroosmotic mobilities via the composition of
the buffer , or by dynamic coating of the capillary or , as presented in Figure 1.4. , by
increasing the temperature.
The first option is the most commonly used, a capillary length of 25 -30 cm being the
inferior limit at which an analys is can be performed. However, there is an asymmetrically
positioned detector closer to the outlet part. This shorter part, of 7 -10 cm, can be used for
different procedures. Electrokinetic or hydrodynamic injection can be performed,
electrophoresis being re alized at reverse polarity on the sample placed at the outlet end of the
capillary (Fig ure 1. 4.).

10

Figure 1.4. Short end injection vs. normal injection [34]

All other procedures accompanying analysis are the same as for the normal injection
mode. Thi s so called “short -end injection procedure” can be integrated in EMMA [34, 35],
with the purpose of decreasing the analysis time.
As for the separation method , the choice depends on the kinds of reactants and
products to be separated. As in other CE applic ation fields, capillary zone electrophoresis
(CZE) is the most frequently used, followed by micellar electrokinetic capillary
electrophoresis (MECK).

Figure 1. 5. Micellar Electrokinetic Capillary [36]

11

MECK is particularly interesting, being based on the addition to the buffer solution of
a micellar “pseudostationary” phase. This last one interacts with the analytes, just like in a
chromatographic metho d, according to a partitioning mechanism. The “pseudostationary”
phase is in fact a surfactant, add ed to the BGE in a concentration above its critical micellar
concentration (CMC). In the system created, the EOF acts as the promoter that moves the
analytes through the capillary [36,37]. The most commonly used surfactant is sodium dodecyl
sulfate (SDS), an anionic surfactant . The formed micelles are electrostatically attracted
towards the anode ( Fig. 5) . The EOF transports the bulk solution towards the negative
electrode due to the negative charge on the internal surface of the silica capillaries. But th e
EOF is usually stronger than the electrophoretic migration of the micelles and therefore the
micelles will migrate also toward the negative electrode with a retarded velocity . This
approach is really useful when neutral compounds have to be analyzed .
Finally, for the detection step, when working with MECK, photometric detection is
generally used when the surfactant solution does not significantly absorb ultraviolet light. The
detection sensitivity of a photometric detector is not high in terms of concentr ation (i.e., above
the micromolar range). Laser -induced fluorescence (LIF) detection is ve ry sensitive, and it
can detect concentrations down t o the nanomolar scale. The micelle can enhance this LIF
sensitivit y. Electrochemical detection is another sensiti ve technique that is particularly
suitable for narrow -bore capillaries measuring less than 10 μm i.d. Mass spectrometry (MS) is
an indispensable dete ction method for CE . Several interfaces for CE -MS are available, but
significant work has to be done for th e development of reproductible CE -MS, taking into
consideration the effect of high concentrations of nonvolatile surfactant molecules [37].

12
II. EXPERIMENTAL PART
II.1. Introduction
In vitro studies for d rug metabolism have a growing importance in the characterization
of new chemical enteties. EMMA has proven to be an advantageous technique for this
type of studies , considering the high efficiency and short analysis time that can be attained
[38]. The capillary becomes a microreactor, where analytes re act, being also separated and
detected, minimizing the sample consumption.
Using this approach, Curcio et al. described the activity of CYP 450 using
dextromethorphan as substrate [5] . Zhang et al. [39] also performed an experiment for the
activity characte rization of CYP3A4, testosterone and nifedipine being tested as
substrates. In addition, the enantio selective metabolization of fluoxetine by CYP2D6 [40],
the stereoselective N-demethylations of verapamil[41] and ketamine[42] were
investigated, proving the versatility of EMMA. In addition, some factors that influence the
electrophoretic process were tested, by univariate approaches, like the effects of reagent
zone and buffer conditions [43]. Another group [30] performed a computer simulation in
order to pr edict zone overlap , while others [28] described the importance of rapid polarity
switches.
In this paper, a fully automated in -capillary system for studying the activity of
CYP1A1 was developed. 7-ethoxycoumarin (7 -EC) was chosen as the substrate, showing
the best Km value for our enzyme (10 ± 0.1µM) [44]. In the presence of nicotinamide
adenine dinucleotide phosphate reduced (NADPH), as a cofactor, 7 -EC undergoes a
reaction of O-deethylation, giving rise to 7 -hydroxycoumarin (7 -HC), the molecule that
will be assayed in our study. Optimization by multivariate approach was employed, using
a design of experiment (DoE) . To the best of out knowledge, is for the first time that a
DoE is performed, in order to gain a better understanding of factors that influence the
metabolic reaction.
The best results obtained in optimal conditions were compared with:
– offline metabolization ;
– human liver microzomes.

13
Finally, we demonstrated the versatility of our method by monitorizing CYP1A1
inhibition, using apigenin as a pot ent inhibitor.
II.2. Materials and methods
II.2.1. Instrumentation and capillaries
An HP 3D CE system (Agilent Technologies, Waldbronn, Germany) was used for all
the experiments. It was equipped with an autosampler, an oncolumn DAD and a temperature
contro l system. For instrument control, data acquisition and analysis, a chemstation (Hewlett –
Packard, Palo Alto, CA, USA) was employed.
Uncoated bared fused -silica capillaries were provided by ThermoSeparation Products
(San Jose, CA, USA). A total length of 48. 5 cm ( 8.5 cm effective length) was used , while the
internal and external diameter were 50 and 375 µm. The capillary was thermostated at 37 șC
by high -velocity air stream, whereas the external water bath was kept at 25 șC, unless
specified otherwise.
Capil lary preconditioning was realized daily, by flushing at 37 șC in the following
order:
– 1 M NaOH for 10 minutes;
– H2O for 10 minutes;
– BGE for 10 minutes.
Between runs, another rinsing procedure was followed:
– 1 M NaOH for 5 minutes;
– H2O for 5 minutes;
– BGE for 5 minutes.
To prevent carryover, the capillary ends were dipped into water, after each injection
step.

II.2.2 . Reagents and chemicals
The electrolyte s: sodium dihydrogen phosphate (NaH2PO4), disodium h ydrogen
phosphate (Na2HPO4) and magnesium chloride (Mg Cl2), were purchased from Merck

14
(Darmstadt, Germany). Dimethyl sulfoxide (DMSO) and sodium hydroxide (NaOH) were
from BDH Prolabo Chemicals ( Leuven, Belgium), whereas sodium chloride (NaCl) was
obtained from Acros Organics (NJ, USA).
The reactants used f or the enzymatic procedure: 7 -EC and 7 -HC were acquired from
Alfa Aesar (Karlsruhe, Germany); NADPH was purchased from Panreac AppliCHem
(Darmstadt, Germany); Supersomes CYP1A1 (1 nmol/mL) were supplied from Corning
Supersomes (Woburn, MA, USA), while meth ylcholantrene -induced (MC -induced) rat liver
microsomes (20mg/mL) were kindly provided by Advanced Technology Corporation
(Liege,Belgium).
The surfactant, SDS, was from Fischer Scientific (Loughborough, UK). The enzymatic
inhibitor, apigenin, was purchased from Extrasynthese (Genay, France). The organic solvent,
methanol, was provided by J.T. Baker Chemicals (Deventer, the Netherlands). Ultrapure
water was supplied by a Milli -Q equipment (Millipore, Bedford, MA, USA).
All reagents and chemicals were of ana lytical grade.

II.2.3 . Buffer and sample preparation
The incubation buffer (IB) was prepared by adding a solution of 50 mM NaH 2PO 4 to a
solution of 50 mM Na 2HPO 4, under continous stirring, until pH 7.4 . In order to obtain the
BGE, 18 mM of SDS were mixed with the IB.
Stock solutions of 10 mM EC and 10 mM HC were prepared in methanol, while a
stock solution of apigenin was prepared in DMSO. They were stored at 8 șC, protected form
light. To reach the appropriate work concentrations, they were subsequentl y diluted with IB.
As for the cofactor, a solution of 5 mM of NADPH was prepared freshly each day, in 0.05 M
MgCl 2.
The supersomes were kept at -80 șC, thawed rapidly a t 37 șC, in a water bath and the n
diluted with the IB to reach appropriate final concent rations. Also, they were stored on ice
until use.

15
II.2.4. EMMA procedure
The partial -filling technique, namely sandwich mode, was chosen for the EMMA
procedure. Different plugs of reactants were injected (outlet injection) , following the next
protocol:
– IB plug ( -50 mbar for 10 s);
– CYP 1A1 supersomes plug ( -50 mbar for 6 s);
– NADPH and 7 -EC plug ( -50 mbar for 6 s);
– CYP 1A1 supersomes plug ( -50 mbar for 6 s);
– IB plug ( – 50mbar for 10 s).
The reaction (Figure 2.1. ) was initiated by the application of a voltage switch ( –
0.2/0.2/ -0.2/0.2 kV, eac h for 10 s). The voltage was the n turned off during 15 minutes,
allowing the metabolic reaction to take place. The separation of the components was
performed at -25 kV , the detection being achieved at 320 nm.

O O H3C O
O O HO

Figure 2.1. Enzymatic reaction

The experimental design and the subsequent analysis were carried out using JMP
software version 10.0 ( SAS Institute, Cary, NC, USA). Thirty experiments were defined ,
repeating the central point five times, while the others were repeted three times. The statistical
significant threshold was established at p value < 0.05. 7- Ethoxycoumarin (EC)
7- Hydroxycoumarin (HC) CYP1A1 + NADPH, 37°C

16
In order to find the reaction rate (nmoles of H C/min/nmoles of CYP1A1), the
following mathematical equation was used for the calculation of the volume of each plug:
,
where V (m3) is the injected volume, ΔP (Pa) is the applied pressure, d (m) is the
capillary internal diameter, t (s) is the duration of pressure applications, ɳ is the solution
viscosity and L (m) is the capillary length.

II.2.5. Offline incubation procedure
The metabolization took place in a n eppendorf placed in a water bath (Heidolph
Instruments, Schwaback, Germany), during 15 minutes at 37șC. 10µL of 1 pmol/L
supersomes were mixed with 20 µL of IB, 10 µL of 5mM NADPH and 10 µL of 1.25 mM
EC. After the given time, the reaction was stopped by addition of 10 µL of a mixture of 0.05
M NaOH/NaCl.
Using a centrifuge (Eppendorf, Hamburg, Germany) , at a working speed of 13 400
rpm, for 5 minutes, the r esulted supernatant was injected in our system. The outlet injection
was performed also in this assay, at -50mbar for 6 seconds. The separation was carried out at
-25 kV, the amount of HC being spectrophotometrically determined.

II.2.6. Inhibition assay procedure
The inhibitor, apigenin (100 µM ), was preincubated for 30 minutes with our enzyme,
200 pmol/L CYP1A1 or 2 mg/mL MC -induced rat liver microsomes. The exact protocol as
for the EMMA procedure was consequently respected. The obtained results were co mpared
with the ones obtained after a blank EMMA procedure. Here, the supersomes were
preincubated with the solvent (DMSO 1% in BGE) employed for 100 µM apigenin
preparation.

17

II.3. Results and discussion
II.3.1. Preliminary studies – inlet injection opt imization
II.3.1 .1. Development of a CE method for EC and HC separation
The first step in the development of an automated EMMA procedure is to develop a
simple CE method for the substrate (EC) and product (HC) separation .
It is really important to notice that, at physiological pH, neither EC or HC don’t possess
an electrophoretic mobility, migrating with the electro -osmotic flow (EOF). A s a result,
MECK had to be used. For this purpose, a surfactant, namely SDS at a concentration of 10
mM was used .

Figure 2.2 . The electropherogram obtained for the separation of HC and EC standar ds
Time(min)0 2 4 6 8 10Absorbance
-202468101214HC EC
BGE: Sodium phosphate buffer 50 mM at pH 7.4 + 10 mM SDS;Inlet injection 50 mbar for 3 sec;
Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm total length (40 cm effective length); Voltage: 25 kV; Capillary
temperature: 37 °C;Detection: 320 nm.

18

As seen in the Figure 3.1., a good resolution (R S =16) and a short analysis time were
obtained for the separation of the two standards. The same IB w as maintained for all the
following experiments.
II.3.1.2. Off line metabolization
The next step was to investigate the behaviour of our substrate in the presence of the
enzyme. For this reason, an ATC protocol was followed, performing an off -line
metaboli zation assay , using microsomes . After a waiting time of 15 minutes, the reaction was
stopped using a mixture NaOH/NaCl and subjected to a centrifugation step. The resulted
supernatant was injected in the CE, obtaining the adjacent electropherogram (Figure 2.3.).
.

Figure 2.3. The el ectropherogram obtained for off line metabolization – preliminary studies

Time(min)0 1 2 3 4 5 6 7Absorbance
-1012345HC EC
NADPH
BGE: Sodium phosphate buffer 50 mM at pH 7.4 + 10 mM SDS; WT: 15 mi n; Inlet injection 50 mbar for 3 sec;
Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm total length (40 cm effective length); Voltage: 25 kV;
Capillary temperature: 37 °C;Detection: 320 nm.

19

As expected, the peak of HC emerged, along with the one of the cofactor. By comparison with
Figure 2.2. , the retention times w ere similar for substrate and reaction product, proving that
the choi ce of this reactants is justified . In addition, there are no interferences at the chosen
wave length.
II.3.1.3. On line metabolization
To perform inline assays, the EMMA approach was sele cted, allowing the enzymatic
reaction to take place directly inside the capillary. A sequential injection protocol was used, in
order to reduce the manipulation steps. In a previously BGE -filled capillary, 4 different plugs
were injected at inlet: BGE with out SDS, microsomes, EC and NADPH in BGE and again,
BGE without SDS and a voltage switch was applied.

Time(min)0 2 4 6 8 10 12Absorbance
-2024681012BGE: Sodium phosphate buffer 50 mM at pH 7.4 + 10 mM SDS; Uncoated fused silica capillaries : 50 μm i.d. and
48,5cm total length (40 cm effective length);Sequential injection : (50 mbar for 10s), (50 mbar for 6s), (50 mbar for
6s), (50 mbar for 10s);Voltage s witch : (+0,5 kV for 7s), ( -0,5 kV for 7 s), (+0,5 kV for 7s), ( -0,5 kV for 7s); Waiting
time: 15 min; Migration voltage: 25 kV; Capillary temperature: 37 °C ; Detection: 320 nm.
HC NADPH EC

20
Figure 2.4. The e lectropherogram obtained for on line metabolization – preliminary studies
The on -line metabolization gave rise to a s maller amount of metabolite, emphasizing
the need of optimization of various key parameters involved in the metabolic reaction.
II.3.1 .4. Use of internal standards
In order to decrease the variability of the CE analysis, the use of an internal standard
(IS) is justified. The first two tested were atenolol (Figure 2.5. A. ), a selective β1 receptor
antagonist or beta-blocker , and cetirizine (Figure 2.5. B. ), a new gen eration drug of H1 –
antihistamine s.

Figure 2.5. Chemical structures of atenolol and cetirizine

Their behaviour have been observed by an off -line procedure, after a waiting time of
30 minutes . The major difference between this two assays and the previous ones is the
working wave length of 235 nm ( Figure 2.6.) . If atenolol shows weak resolution (Figure
2.7.A.), cetirizine provides a better separation (Figure 2.7 .B.), bu t with a notably increased
migration time.

A.
B.

21

Figure 2.7. Atenolol (A.) and cetirizine (B.) as internal standards
Time0 2 4 6 8 10Absorbance
0246810121416
Time(min)0 2 4 6 8 10 12 14Absorbance
024681012(min) A.
B. Atenolol
Cetirizine
BGE: Sodium phosphate buffer 50 mM at pH 7.4 + 10 mM SDS; WT :30 min; Inlet injection 50 mbar for 3
sec; Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm total length (40 cm effective length); Voltage: 25
kV; Capillary temperature: 37 °C; Detection: 235 nm.

22
Despite of the prolonged time of analysis, cetirizine was further used in a series of off –
line control tests, in order to check the stability and possible interactions of our compounds.
Firstly, all the components of the enzymatic reaction were separately injected, after an
off-line incubation. The electropherogram (Figure 2.8.) offers the certai nty that all the
molecules are stable during the analysis, having distinct migration times .

Figure 2.8. Off -line stability control test
Secondly, another control test was performed, this time only NADPH and cetirizine
being compared. As shown both in Figure 2.8. and Figure 2.9., cetirizine is not metabolized ,
not even in the presence of microsomes. On the other hand, NADPH is consumed and
degraded, but only in the presence of the enzyme.
Therefore, the use of an internal standard proves to be useful tool when motinoring the
substrate.
Time(min)0 2 4 6 8 10Absorbance
0510152025Neat bu ffer Microso mes 0.1 mg/mL HC 0.1 mM EC 0.1 mM NADPH 0.4 mg/mL Cetirizine 0.05 mg/mL
BGE: Sodium phosphate buffer 50 mM at pH 7.4 + 10 mM SDS; WT: 30 min; Inlet injection 50 mbar for
3 sec; Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm total length (40 cm effective length);
Voltage: 25 kV; Capillary temperature: 37 °C; Detection: 235 nm.

23

Figure 2.9. Off-line c ontrol test

II.3.1. 5. Injection order
In a CE analysis, a particular aspect of the process is represented by the introduction of
the sample into the capillar y [45]. The right approach for the right application can lead to
significant improvements in performance, particularly with regard to senzitivity. Injection
order seems to be crucial, specially when performing an on -line reaction.
To demonstrate this princ iple, m ethylcholanthrene -induced microsomes (MC) 0.5
mg/mL were injected before and after the substrate, as different plugs. After an waiting time
of 10 minutes, with no voltage switch, the normalized areas (NA) obtained of HC were
compared. For the MC+sub strate, NA was 0.026 (n=3), while for substrate+MC, NA was
0.068 (n=3).
Taking into account that all the other experimental conditions were maintained
constant, it is clearly that injection order is a key factor for on -line CE metabolization.
Time0 2 4 6 8 10Absorbance
0510152025
NADPH+CT
NADPH(min) NADPH
related NADPH NADPH
related CT
BGE: Sodium phosphate buffer 50 mM at pH 7.4 + 10 mM SDS; WT: 15
min; Inlet injection 50 mbar for 3 sec; Uncoated fused silica capillaries : 50
μm i.d. and 48,5cm total length (40 cm effective length);
Voltage: 25 kV; Capillary temperature: 37 °C; Detection: 235 nm.

24
II.3.1. 6. Microsomes concentration
The single most important property of enzymes is the ability to increase the rates of
reactions occurring in living organisms, a property known as catalytic activity [46]. This
activity is greatly influenced by pH, temperature, subs trate concentration or enzyme
concentration. Norma lly, the reaction rate increases as the concentration of the catalyst is
increased.
In our case, at a concentration of MC 1 mg/mL versus a concentration of MC 0.1
mg/mL, a visible difference is spotted on the electropherogram (Figure 2.10.). This leads to a
essential increase of MC concentration for the on -line metaboli sation.

Figure 2.10. Impact of the microsomes concentration

Time(min)0 2 4 6 8 10Absorbance
-20246810121416MC 0.1 mg/mL
MC 1 mg/mL EC
EC
HC
BGE: Sodium phosphate buffer 50 mM at pH 7.4 + 10 mM SDS; Inlet injection 50 mbar for 3 sec;
Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm total length (40 cm effective length); No voltage
switch; Waiting time: 10 min; Voltage: 25 kV; Capillary temperature: 37 °C; Detection: 320 nm.

25
II.3.2. Preliminary studies – outlet injection optimiza tion
Even though an inlet injection produced a very short time of CE analysis (2.40 min) ,
an outlet injection may permit decreasing even more this parameter. Also, the total time of
analysis is reduced, with no need of sample manipulation, in -line metaboli sation was the
chosen protocol.
Another series of preliminary studies was carried out by optimizing more key factors ,
presented below . Some of them will be next included as variables in a design of experiment.
Extremely important to mention that another IS was employed, namely mesytil oxide
(MO) 0.005% , visible in the ultraviolet light at 235 nm (Figure 2.11.), since it migrates faster
than cetirizine: 0.37 minutes for MO versus 11.91 minutes for cetirizine.

Figure 2.11. MO as interna l standard

Time(min)0,0 0,5 1,0 1,5 2,0 2,5 3,0Ansorbance(mAU)
0510152025BGE: Sodium phosphate buffer 50 mM at p H 7.4 + 10 mM SDS; MO 0.005% ; Inlet injection -50 mbar for
3 sec; Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm total length ( 8.5 cm effective length); No
voltage switch; Migration voltage: -25 kV; Capillary temperature: 37 °C; Detection: 235 nm.
MO HC
EC

26
II.3.2.1. Injection order and voltage switch
As exposed before, injection order and mixing voltage switch are crucial parameters
for the obtainment of good reaction rates . Three alternatives were proposed (Figure 2.12 .),
applying or not a volta ge switch of (+0,5 kV for 7s), ( -0,5 kV for 7 s), (+0 ,5 kV for 7s), ( -0,5
kV for 7s).

Figure 2.12 . Alternatives of outlet injection
The third approach , named also the „sandiwch mode”, combined with rapid polarity
switches , generated the biggest amount of HC (Tabl e 2.1).
Tabel 2.1. Outlet injection modes
Area
HC TM
(min) NA HC
(RDS) Area
EC TM
(min) NA EC
(RDS) Ratio

S+M (-Vs) 2.31
0.64 0.046
(6.3)
155.2
1.02 1.96
(8.5) 42.2
(4.0)

S+M (+Vs)
2.46
0.64 0.04
(11.0)
154.8
1.03 1.55
(9.2) 39.1
(5.3)

M+S (-Vs)
1.87
0.59 0.05
(5.1)
92.4
0.79 1.85
(5.2) 37.12
(10.0)

M+S (+Vs)
2.13
0.61 0.04
(11.5)
127.1
0.86 1.85
(5.8) 42.8
(17.1)

M+S+M ( -Vs)
4.63
0.62 0.124
(15.9)
97.6
0.81 2.02
(11.8) 16.4
(8.1)

M+S+M (+Vs)
5.56
0.55 0.132
(4.07)
112.4
0.71 2.08
(3.7) 15.7
(2.02)
BGE
without
SDS MO+EC
+NADPH Microsomes BGE
without
SDS

BGE
without
SDS
MO+EC
+NADPH
MC BGE
without
SDS
BGE
without
SDS
MO+EC
+NADPH
Microsomes
BGE
without
SDS

MC
+
+
+ –
–
– BGE pH 7.4 (+ 15mM SDS)
50 mM phosphate buffer

BGE pH 7.4 (+ 15mM SDS)
50 mM phosphate buffer

BGE pH 7.4 (+ 1 5mM SDS)
50 mM phosphate buffer

27
II.3.2.2. Incubation time
After performing the RPS, a limited period of time is necessary for the enzymatic
reaction to take place. As generaly known , the incubation time must be chosen with respect to
the total time of analysis. In addition, this variable must permit the reaction to reach a high
conversion rate for the substrate.
In our case, a maximum quantity of HC is obtained after 30 mi nutes of incubation
(Figure 2.13. ). But in order not to increase excesively the analysis and to remain in the linear
part of the curve , the value of 10 minutes was chosen and set for all the others further
experiments .
II.3.2.3. Microsomes concentration
Once again, the impact of the enzyme conc entration over the HC obtained
concentration was tested. It appears that at 4 mg/mL microsomes, the plateau is reached and
further increase of the enzyme concentration produces no increase in the HC amount (Figure
2.14.) .
In order to avoid capillary overload , but also to remain i n the linear zone of the curve,
the microsomes concentration was set at 2 mg/mL.

Figure 2.13. Impact of incubation time on hydroxycoumarin CA and NA
C
A N
A
Incubati on time(min) Incubation time(min)
BGE: Sodium phosp hate buffer 50 mM at pH 7.4 + 12.5 mM SDS; Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm
total length (8.5 cm effective length); Sandwich mode injection ; Voltage switch : +0,5 kV/ -0,5 kV/+0,5 kV/ -0,5 kV each for
7s; Migration voltage: – 25 kV; Capillary temperature: 37 °C ; Detection: 320 nm.

28

Figure 2.14 . Impact of MC on hy droxycoumarin CA and NA

II.3.2.4. Voltage switch and switch time influence
In attempt to investigate the most important variables interfering with our experiment,
voltage switch influence generated interesting results. At first glance (Table 2.2.) , these data
may appear to indicate that, for a constant number of RPS events, higher potentials give rise
to lower yield . This migth be due to the fact that our reactants do not interact properly, EC
migrating further than our enzyme, while having different elec trophoretic mobilities.
Probably, the time they spend in contact decreases with the increase of the voltage.
As a results, one might expect that at a higher switch time, the NA of HC would
decrease. Observing the data (Table 2.3.) , it becomes clear that t his hypothesis is valid. In
addition, it’s certain that this two parameters should be considered as one, their influence over
the metabolization process being significant.

C
A N
A
MC concentration (mg/mL) MC concentration (mg/mL)
BGE: Sodium phosp hate buffer 50 mM at pH 7.4 + 12.5 mM SDS; Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm
total length ( 8.5 cm effective length); Sandwich mode injectio n; Voltage switch : +0,5 kV/ -0,5 kV/+0,5 kV/ -0,5 kV each for
7s; Wainting time: 10 min; Migration voltage: – 25 kV; Capillary temperature: 37 °C ; Detection: 320 nm.

29
Table 2.2. Voltage switch influence
Area HC TM(min) NA HC Area EC TM(min) NA EC Ratio
0.5 kV 5.56 0.55 0.13
(4.1) 112.4 0.71 2.08
(3.7) 15.7
(2.0)
1 kV 4.34 0.57 0.07
(7.0) 112.6 0.73 1.43
(2.9) 20.1
(4.7)
2.5 kV 3.93 0.54 0.07
(7.5) 106.2 0.68 1.58
(2.6) 21.4
(4.9)
5 kV 5.08 0.57 0.08
(5.6) 118.9 0.74 1.43
(2.3) 18.0
(7.6)

Table 2.3. Voltage switch time impact
Area HC TM(min) NA HC Area EC TM(min) NA EC Ratio
2 sec 4.19 0.56 0.065
(8.1) 113.97 0.72 1.37
(7.4) 21.0
(1.9)
6 sec 4.67 0.54 0.093
(4.8) 112.74 0.69 1.77
(6.5) 19.1
(2.4)
10 sec 5.10 0.59 0.075
(6.2) 122.61 0.79 1.36
(6.4) 18.02
(2.8)

II.3.3 . Optimization – final steps
As observed from the preliminary studies, some of the most important factors to be
taken into consideration are: incubation time, injection mode, concentration of the enzyme
and substrate, voltage swit ch and voltage switch time.
The main difference from the preliminary studies is the use of supersomes in t he
process of metabolization. In this manner, we monitorize the activity and behaviour of a
single enzyme, CYP1A1. We started by k eepping the same inj ection protocol, the „sandwich
mode”. Three distinct plugs of reactants were employed: first, the enzyme (CYP 1A1
supersomes), secondly, the substrate and the cofactor ( EC and NADPH) and again, the
enzyme. Once again, before and after the CYP1A1 plugs, tw o plugs of BGE without SDS
were injected, to protect the supersomes from surfactant denaturation. The hydrophobic tails
of the SDS may disrupt the hydrophobic interactions in the interior of the protein, which
favorize protein denaturation.

30
II.3.3.1. Optimization of incapillary metabolization procedure
EMMA procedure is compatible with the use of supersomes and a s seen above, RPS
was the method of choice for plug mixing.
Four rapid polarity switches were carried out, the overlap being obtained due to te
differences in the electrophoretic mobilities of the reactants. At pH of 7.4, EC is a neutral
molecule, migration with the EOF (µ ep=0 ; µ ap = µ EOF = (565.9 ± 2.4) × 10-6 cm2V-1s-1). On
the other hand, the enzyme has its own electrophoretic mobility (µ ep= (-162.7 ± 2.8) × 10-6
cm2V-1s-1, µ ap= 403.2 × 10-6 cm2V-1s-1).
The incubation time was set tled at 15 minutes, even though the preliminary studies
showed a good conversion rate for 10 minutes. During this time, no voltage was applied (zero
potential amplifi cation step) . A constant voltage of -25 kV was applied afterwards for the
separation step .
Also , the capillary temperature was investigated. No significant differ ence was
observed (Figure 2.15. ); the CA for HC at 25°C was14.5, while at 37°C was found to be 15.2.
But it is noteworthy to mention that the RDSs obtained using 37°C (14.3%) were higher than
those observed at 25°C (3.1%), showing that at 37°C, the stability of the enzyme is probably
affected.
Moreover, the final concentration of the organic solven t has to be kept as lower as
possible in the microreactor, in order to maintain the enzymatic activity. Methanol was , in this
case, limited at 2.5% , to conserve the supersomes function.
The next step was the study of the influence of injected plug length o n the
metabolization rate. The best results were obtained by injecting at -50 mbar for 6 seconds
each plug (Figure 2.16.B.). With longer plugs (Figure 2.16.C), the shape of the peak was
affected, while with shorter plugs (Figure 2.16.A), the amount of HC lowered.

31

Figure 2.15 . Influence of capillary temperature

Figure 2.16 . Influence of the plugs length on EC metabolization
Time(min)0,0 0,5 1,0 1,5 2,0 2,5 3,0Absorbance(mAU)
01020304025 °C 37 °C
BGE: Sodium phosp hate buffer 50 mM a t pH 7.4 + 18 mM SDS; Uncoated fused silica capillaries : 50 μm
i.d. and 48,5cm total length ( 8 cm effective length); Sandwich mode injection CYP:S:CYP, each for 6 s ;
Voltage switch : +0,5 kV/ -0,5 kV/+0,5 kV/ -0,5 kV each for 7s; Wainting time: 15 min; Migration
voltage: – 25 kV; Detection: 320 nm.

Time (min)0,0 0,5 1,0 1,5 2,0 2,5 3,0Absorbance (mAU)
0102030405060
BGE: Sodium phosp hate buffer 50 mM at pH 7.4 + 18 mM SDS; Uncoated fused silica capillaries : 50
μm i.d. and 48,5cm total length ( 8.5 cm effective length); Sandwich mode injection ; Temperature: 25
°C; Voltage switch : +0,5 kV/ -0,5 kV/+0,5 kV/ -0,5 kV each for 7s; Wainting time: 15 min; Migration
voltage: – 25 kV; Detection: 320 nm.
A C
B EC
HC NADPH

32
II.3.3.2. Study of the enzymatic parameters
When working with metabolic studies, the conversion rate of the substrate by the
enzymatic system must be optimized [3]. So it becomes crucial to perform the experiments
under conditions that ensure linearity regarding incubation time and enzyme concentration.
To begin wit h, the relation between incubation time and reaction velocity was studied,
maintaining the substrate and enzyme concentration constant (250 µM and 200 pmol/mL).
Four alternatives were compared, varying the incubation time at 10, 15 , 30 and 60 minutes
(Figu re 2.17 .). On t he other hand, the relation between the reaction velocity and CYP1A1
concentration was determing by maitaining constant the substrate concentration and the
incubation time (250 µM and 15 minutes). The enzyme concentrations employed were 100,
150, 200 and 250 pmol/mL (Figure 2. 18.). The best compromise between short time of
analysis and a good turnover for the substrate was achieved by choosing an incubation time of
15 minutes and a CYP1A1 concentration of 200 pmol/mL for further studies.
Time (min)

Figure 2.17. Effect of the incubation time on HC formation
BGE: Sodium phosp hate buffer 50 mM at pH 7.4 + 18 mM SDS; Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm total
length ( 8.5 cm effective length); Sandwich mode injection ; Temperature: 25 °C; Voltage switch : +0,5 kV/ -0,5 kV/+0,5 kV/ –
0,5 kV each for 7s ; Migration voltage: – 25 kV; Detection: 320 nm.

CA HC

33

Figure 2.1 8. Effect of the CYP1A1 concentration on the product formation

II.3.3.3. Design of experiment
The final step in our development was the optimization of the electrophoretic mixing,
a key factor in EMMA procedures. Instead of using a mathematical method to predict zone
overlaps, the use of a design of experiment was chosen. This alternative is due to a difference
in cond uctivity between the reagent zones and BGE, the electric field not being the same
across all the capillary.
A few number of parameters influencing the metabolization rate were thus
investigated: the voltage mixing value ( X1 ; three levels: 0.1, 0.55 and 1 kV) and the voltage
mixing time ( X2; three levels: 2, 6 and 10 seconds). HC CA was proposed as a response to
maximize, Y. The response was modeled by the following equation:
Y= β 0 + β 1X1 + β 2X2+ β 12X1X2+ β 11X11+ β 22X22 + ε,
Where β 0 is the intercept, β 1 and β2 are the main effect terms, β 1β2 is the interaction term, β 11
and β 22 are the quadratic terms and ε the error term.
CA HC
BGE: Sodium phosp hate buffer 50 mM at pH 7.4 + 18 mM SDS; Uncoated fused silica capillaries : 50 μm i.d. and 48,5cm
total length ( 8.5 cm effective length); Sandwich mode injection ; Temperature: 25 °C; Voltage switch : +0,5 kV/ -0,5 kV/+0,5
kV/-0,5 kV each for 7s ; WT: 15 min; Migration voltage: – 25 kV; Detection: 320 nm.
CYP1A1 concentration (mg/mL)

34
For a better visualisation of the results, the respo nse was plotted against the two
studied variables (Figure 2.19 .). The optimal mixing conditions were found to be 0.22 kV for
the voltage mixing and, as for the voltage mixing time, the best value was superior of the
tested . The limit of 10 seconds was chos en with respect to the total time of analysis.

Figure 2.1 9. Response surface of the effect of voltage mixing value and voltage mixing time
on HC CA and prediction profiles
The predicted condition was performed in triplicate and a value of 16.9 ± 0.4 was obtained for
the HC CA (Figure 2. 20.). The quality of this predictive model was underlined by the
experimental value, which falls into the confidence interval of 17.2 ± 0.7. In addition, the
RDSs for all the registred parameters ( HC CA, HC migration tim e, EC CA, EC migration
time) were lower than 2.1% , an excellent value considering the complexity of our system
(Table 2.4 .).
Table 2.4. Repeatability in the optimal conditions
Area
HC TM HC CA HC Area
EC TMEC CA EC Ratio
(C)
1st injection 10.25 0.59 17.28 50.99 0.82 62.11 3.59
2nd injection 10.04 0.60 16.88 50.87 0.82 62.11 3.68
3rd injection 10.13 0.61 16.58 49.20 0.80 61.27 3.69
Average 10.14 0.60 16.91 50.35 0.81 61.83 3.66
SD 0.10 0.01 0.35 1.00 0.01 0.48 0.05
RDS 1.00 1.65 2.07 1.98 1.21 0.78 1.48

35
Using a standard calibration curve, the 7 -HC peak areas were converted into
concentrations, in order to determine the amount produced and the conversion rate. Over the
range of concentrations employed, 20 -100 µM, the regression equation was found line ar
(y=0.2452 x + 0.2837) , with a really high coefficient of determination ( r2 = 0.9999). With a
measured concentration of HC (67.8 ± 1.4 µM), a volume of injected plugs calculated
according to the equation found in chapter II.2.4., an incubation time of 15 minutes and a
CYP1A1 concentration of 200 pmol/mL, the reaction rate was found to be 11.3 nmol of
HC/min/nmoles of CYP1A1 in our optimal conditions.
To test our system efficiency, a comparison with the conventional offline procedure
was performed, main taining all the experimental conditions identical (Figure 2. 20.). A slightly
higher HC CA was obtained after offline metabolization ( 20.0 ± 0.1), compared with our
inline value (16.9 ± 0.4). The difference may result from the fact that the mixing is perfo rmed
by RPSs and therefore may be not complete. However, out method is fully automated, saving
time and reagents.

II.3.4. Applicability
II.3.4.1. Applicability to microsomes in vitro model
The developed inline system may by useful for a more complex in v itro assay. For this
purpose, a test was performed using MC -induced liver microsomes (2mg/mL). A triplicate
was effectuated and the obtained results were comparable taking into consideration the
metabolite rate formation ( Figure 2. 20.). Once again, the se lectivity of our method was
confirmed .
II.3.4.2. Applicability to CYP 1A1 inhibitors screening
Knowing that CYP1A1 is a possible target for cancer treatement, our system potency
was tested regarding inhibition studies. A known inhibitor, apigenin, was sele cted and a
significant reduction of HC amount was noticed when our enzyme was preincubated with 100
µM apigenin, for 30 minutes at room temperature (Figure 2.21.).

36

Figure 2. 20. CYP1A1 vs. MC vs. Off -line metabolizatio n

Figure 2.21. Apigenin as CYP1A1 inhibitor
Time (min)0,0 0,5 1,0 1,5 2,0 2,5 3,0Absorbance (mAU)
01020304050CYP EMMA OFF-LINE
MC EMMA
BGE: Sodium phosp hate buffer 50 mM at pH 7.4 + 18 mM SDS; Uncoated fused silica capillaries : 50 μm i.d.
and 48,5cm total length ( 8.5 cm effective length); Temperature: 25 °C; Voltage switch : +0,22 kV/ -0,22
kV/+0,22 kV/ -0,22 kV each for 10s ; WT: 15 min; Migration voltage: – 25 kV; Detection: 320 nm.

Time (min)0,0 0,5 1,0 1,5 2,0 2,5 3,0Absorbance (mAU)
010203040
Without apigenin With apigenin HC NADPH
Apigenin
related EC
BGE: Sodium phosp hate buffer 50 mM at pH 7.4 + 18 mM SDS; Uncoated fused silica capillaries : 50 μm i.d.
and 48,5cm total length ( 8.5 cm effective length); Sandwich mode injection ; Temperature: 25 °C; Voltage
switch : +0,22 kV/ -0,22 kV/+0,22 kV/ -0,22 kV each for 10s ; WT: 15 min; Migration voltage: – 25 kV;
Detection: 320 nm.
NADPH EC
HC

37
CONCLUSION S

Since the efficacy and toxicity of drugs are closely related to their pharmacokinetics, a
good understanting of metabolic pathways is important at an early stage of development. F or
a good quantification of all the processes involved, a potent analitical methodology needs to
be employed.
The present study describes the development of a fully automated incapillary method
to monitor CYP1A1 activity, a possible target for cancer thera py. After two series of
preliminary studies, the key factors were identified: incubation time, injection mode,
concentration of the enzyme and substrate, voltage switch and voltage switch time. S atisfying
results were obtained using short -end injection pro cedure, employing microsomes (2 mg/mL)
in sandwich mode and applying rapid polarity switches for a determined period of time (0.5
kV for 6 seconds) .
The optimization stage gave rise to a baseline separation of the molecules of interest
(7-ethoxycoumarin , 7-hydroxycoumarin and nicotinamide adenine dinucleotide phosphate
reduced ) in less than 2 minutes. The optimization followed two pathways: the improvement of
incapillary metabolization procedure and the study of enzymatic parameters.
Incapillary metabolizat ion was significa ntly influenced by the length of the injected
plugs, the best results being obtained after injecting at -50 mbar for 6 seconds . Moreover, a
capillary temperature of 25 °C and a final concentration of organic solvent of 2.5% were
settled, in order to maintain enzyme activity.
The turnover of the substrate was also optimized, by studying the impact of the
incubation time and the CYP1A1 concentration over the amount of produced metabolite. An
incubation time of 15 minutes and a CYP1A1 concet ration of 200 pmol/mL were chosen for
further studies.
In addition, the electrophoretic mixing proved to be another key issue of integrated
microanalysis. To the best of our knowledge, it was for the first time that the mixing
parameters were studied by a fully factorial design of experiment. The voltage mixing value
and the voltage mixing time were selected as parameters. The optimal value predicted for the
voltage mixing value was 0.22 kV, for a period of 10 seconds. The experimental value (16.9 ±

38
0.4) w as found to fall within the confidence interval (17.2 ± 0.7), which underlines the quality
of the predictive model. Taking into account the measured concentration of hydroxycoumarin
(67.8 ± 1.4 µM) , along with all the others values presented above, the re action rate was found
to be 11.3 nmoles of hydroxycoumarin/min/nmoles of CYP1A1 in our optimal conditions.
Interestingly, the amount of 7 -hydroxycoumarin obtained in the optimal conditions
(16.9 ± 0.4) was found to be comparable to the one detected after c onventional offline
metabolization (20.0 ± 0.1 µM). Finally, the potency of our system to perform inhibition
studies was demonstrated using 100 µM apigenin as CYP1A1 inhibitor, a significant
reduction of hydroxycoumarin amount being noticed.
It is notewort hy that the compatibility of our system with the use of surfactants
(MECK mode), ensures its applicability to a large range of molecules. In addition, the
miniaturization and the automatization of the process are the main advantages of performing
inline me tabolization, since the enzymatic reaction, the separation and the detection are all
performed in a single capillary. Besides, the reagents consumption is drastically reduced due
to the injection of few tens of nanoliters, with a very short time of analysi s.

39
REFERENCES
1. Hughes JP, Rees S, Kalindjian SB et al – Principles of early drug discovery , Br J
Pharmacol, 2011, 162: 1239 -1249.
2. Farca ș E, Servais AC, Lamalle C et al – Fully automated electrophoretically
mediated microanalysis for CYP1A1 activity monitoring optimized by multivariate approach ,
Electrophoresis, 2016, 37: 248 -255.
3. Baranczewski P, Stanczak A, Sundberg K et al – Introduct ion to in vitro estimation
of metabolic stability and drug interactions of new chemical entities in drug discovery and
development , Pharmacol Rep, 2006, 58: 453 -472.
4. Nowak P, Woźniakiewicz M, Kościelniak P – Simulation of drug metabolism ,
Trends in Ana lytical Chemistry, 2014, 59: 42 -49.
5. Curcio R, Nicoli R, Rudaz S et al – Evaluation of an in -capillary approach for
performing quantitative cytochrome P450 acitivity studies , Anal Bioanal Chem, 2010, 398:
2163 -2171.
6. Go RE, Hwang KA, Choi KC – Cythochr ome P450 1 family and cancers , J Steroid
Biochem Mol Biol, 2015, 147: 24 -30.
7. Walsh AA, Szklarz GD, Scott EE – Human cytochrome P450 1A1 structure and
utility in understanding drug and xenobiotic metabolism , J. Biol. Chem, 2013 , 288: 12932 –
12943.
8. Sank hwar M, Sankhwar SN – Variations of CYP isoforms and bladder cancer: a
superfamily paradigm , Urol Oncol, 2014, 32: 33 -40.
9. Androutsopoulos VP, Spyrou I, Ploumidis A et al – Expression profile of CYP1A1
and CYP1B1 enzymes in colon and bladder tumors , Plos One, 2013, 8 : e82487.
10. Choudhury JH, Singh SA, Kundu S et al – Influence of the CYP1A1 T3801C
Polymorphism on Tobacco and Alcohol -Associated Head and Neck Cancer Susceptibility in
Northeast India , Asian Pac J Cancer Prev, 2015, 16: 6953 -6961.
11. Plant N – Strategies for using in vitro screens in drug metabolism , Drug Discov,
2004, 9: 328 -336.

40
12. Jia L, Liu X – The Conduct of Drug Meta bolism Studies Considered Good
Practice (II) : In vitro experiments , Curr Drug Metab, 2007, 8: 822 -829.
13. Donato MT, Castell JV – Strategies and Molecular Probes to Investigate the Role
of Cytochrome P450 in Drug Metabolism , Clinical Pharmacokinetics , 2003, 42: 153-178.
14. Di L , Kerns EH , Hong Y et al – Optimization of a higher throughput microsomal
stability screening as say for profiling drug discovery candidates , J Biomol Screen , 2003, 8 :
453-462.
15. Pearce RE , McIntyre CJ , Madan A et al – Effects of freezing, thawing, and storing
huma n liver microsomes on cytochrome P450 activity , Arch Biochem Biophys, 1996 , 331:
145-69.
16. Bjornsson TD , Callaghan JT , Einolf HJ et al – The conduct of in vitro and in vivo
drug -drug interaction studies: a Pharmaceutical Research and Manufacturers of America
(PhRMA) perspective , Drug Metab Dispos, 2003 , 31: 815-32.
17. Camilleri P – Capillary electrophoresis. Theory and practice , CRC press, Florida,
1993 , 5-6.
18. European Phar macopoeia 8th Edition, Chapter 2.2.47. Capillary Electrophoresis ,
2014, 83 -88.
19. Wang X, Li K, Adams E et al – Recent advances in CE -mediated microanalysis
for enzyme study , Electrophoresis, 2014, 35: 119 -127.
20. Iqbal S, Rahman N, Iqbal J – A capill ary electrophoresis -based enzyme assay for
kinetics and inhibition studies of carbonic anhydrase , Anal Biochem, 2014, 444: 16 -21.
21. Zhang J, Hoogmartens J, Schepdael AV – Advances in CE -mediated
microanalysis: An uptade , Electrophoresis, 2008, 29: 56 -65.
22. Zhang J, Hoogmartens J, Schepdael AV – Advances in capillary
electrophoretically mediated microanalysis: An update , Electrophoresis, 2006 , 27: 35 -43.
23. Hai X, Yang B, Schepdael AV – Recent developments and applications of EMMA
in enzymatic and der ivatization reactions , Electrophoresis, 2012, 33: 211 -227.

41
24. Bao J , Regnier FE – Ultramicro enzyme assays in a capillary electrophoretic
system , J Chromatogr A, 1992, 608: 217-224 .
25. Nowak P, Wozniakiewicz M, Wozniakiewicz P – An overview of on -line systems
using drug metabolizing enzymes integrated into capillary electrophoresis , Electrophoresis,
2013, 34: 2604 -2614.
26. Van Dyck S , Van Schepdael A , Hoogmartens J – Michaelis -Menten analysis of
bovine plasma amine oxidase by capillary electrophoresis using electrophoretically mediated
microanalysis in a partially filled capillary , Electrophoresis, 2001, 22: 1436 -1442.
27. Okhonin V , Wong E , Krylov SN – Mathematical model for mixing reactants in a
capillary microreactor by transverse di ffusion of laminar flow profiles , Anal Chem, 2008,
80: 7482 -7486.
28. Krylova SM ,Okhonin V, Krylov SV – Transverse diffusion of laminar flow
profiles -a generic method for mixing reactants in capillary microreactor , J Sep Sci, 2009, 32:
742-756.
29. Sanders BD, Slotcavage RL, Scheerbaum DL et al – Increasing efficiency of in –
capillary electrophoretically mediated microanalysis reactions via rapid polarity switching ,
Anal Chem, 2005, 77: 2332 -2337.
30. Merola ET , Catherman AD , Yehl JB et al – Determination of total antioxidant
capacity of commercial beverage samples by capillary ele ctrophore`sis via inline reaction
with 2,6 -dichlorophenolindophenol , J Agric Food Chem, 2009, 57: 6518 -6523.
31. Stahl JW, Catherman AD, Sampath RK et al – Investigating the effects of
conductivity on zone overlap with EMMA : Computer simulation and experiment ,
Electrophoresis, 2011, 32: 1492 -1499.
32. Pauwels J, Hoogmartens J, Schepdael AV – Application of carbon nanotubes for
in-capillary incubatio ns with cytochrome P450 enzymes , Electrophoresis, 2010, 31: 3867 –
3873.
33. Hai X, Konečny J, Zeisbergercva M et al – Development of electrophoretically
mediated microanalysis method for kinetic study of flavin -containing monooxygenase in a
partially filled capillary , Electrophoresis, 2008, 29: 3817 -3824.

42
34. Glatz Z – Applica tion of short -end injection procedure in CE , Electrophoresis,
2013, 34: 631 -642.
35. Nemec T, Glatz Z – Integration of short -end injection mode into
electrophoretically mediated microanalysis , J Chromatogr A, 2007, 1155: 206 -213.
36. Terabe S – Capillary Separation: Micellar Electrokinetic Chromatography , Anal
Chem, 2009, 2: 99 -120.
37. Hancu G, Simon B, Rusu A et al – Principles of Micellar Electrokinetic Capillary
Chromatography Applied in Pharmaceutical Analysis , Advanced Pharmaeutical Bulletin,
2013, 3: 1-8.
38. Hai X, Nauwelaers T, Busson R et al – A rapid and sensitive CE method with
field-enhanced sample injection and in -capillary derivatization for selenomethionine
metabolism catalysed by flavin -containing monooxygenases , Electrophoresis, 2010, 31:
3352 -3361.
39. Zhang J, Hoogmartens J, Schepdael AV – Kinetic study of cytochrome P450 by
capillary electrophoretically mediated microanalysis , Electrophoresis, 2008, 29: 3694 -3700.
40. Asensi BL, Martín BY, Escuder GL et al – Fast evaluation of enantioselective
drug metabolism by electrophoretically mediated microanalysis: application to fluoxetine
metabolism by CYP2D6 , Electrophoresis , 2013, 34: 3214 -3220.
41. Asensi BL, Martín BY, Escuder GL et al – In-line capillary electrophoretic
evaluation of the enantioselective metabolism of verapamil by cytochrome P3A4 , J
Chromatog r, 2013, 1298: 139-145.
42. Ying KH, Thormann W – Enantioselective capillary electrophoresis for the
assessment of CYP3A4 -mediated ketamine demethylation and inhibition in vitro ,
Electrophoresis , 2012, 33: 3299 –3305.
43. Mason PE, Schildt DC, Strein TG – In-capillary determination of creatinine with
electrophoretically mediated microanalysis: characterization of the effects of rea gent zone
and buffer conditions , J Chromatogr A, 2009, 1216: 154 -158.

43
44. Shimada T, Tsumura F , Yamazaki H – Prediction of hu man liver microsomal
oxidations of 7 -ethoxycoumarin and chlorzoxazone with kinetic parameters of recombinant
cytochrome P -450 enzymes , Drug Metab Dispos, 1999, 27: 1274 –1280.
45. Breadmore MC – Electrokinetic and hydrodynamic injection: making the right
choice for capillary electrophoresis , Bioanalysis, 2009, 1: 889-894.
46. Ball DW, Hill JW, Scott RJ – The Basics of General, Organic, and Biological
Chemistry, v. 1.0 , Chemistry D epartment Books, New York, 2011 , 1056 -1059.