Assessment of cytochrome P450 (1A2, 2B6, 2C9 and 3A4) induction in cryopreserved human hepatocytes cultured in 48-well plates using the cocktail strategy

Article abstract of DOI:10.3109/00498254.2012.719088

1. A fast, straightforward and cost-effective assay was validated for the assessment of CYP induction in cryopreserved human hepatocytes cultured in 48-well plates. The cocktail strategy (in situ incubation) was used to assess the induction of CYP1A2, CYP2B6, CYP2C9 and CYP3A4 by using the recommended probe substrate, i.e. phenacetin, bupropion, diclofenac and midazolam, respectively. 2. Cryopreserved human hepatocytes were treated for 72 h with prototypical reference inducers, β-naphthoflavone (25 μM), phenobarbital (500 μM) and rifampicin (10 μM) as positive controls for CYP induction. The use of a cocktail strategy has been validated and compared to the classical approach (single incubation). The need of using phase II inhibitor (salicylamide) in CYP induction assay was also investigated. 3. By using three different batches of cryopreserved human hepatocytes and our conditions of incubations, we showed that there was no relevant drug-drug interaction using the cocktail strategy. The same conclusions were observed when a broad range of enzyme activity has to be assessed (wide range of reference inducers, i.e. EC50-Emax experiment). In addition, the interassay reproducibility assessment showed that the day-to-day variability was minimal. 4. In summary, the study showed that the conditions used (probe substrates, concentration of probe substrate and time of incubation) for the cocktail approach were appropriate for investigations of CYP induction potential of new chemical entities. In addition, it was also clear that the use of salicylamide in the incubation media was not mandatory and could generate drug-drug interactions. For this reason, we recommend to not use salicylamide in CYP induction assay.

Full text of DOI:10.3109/00498254.2012.719088

Xenobiotica, 2013; 43(4): 320–335
2013 Informa UK, Ltd.
©
ISSN 0049-8254 print/ISSN 1366-5928 online
DOI: 10.3109/00498254.2012.719088
ReseaRꢀh aRTIꢀLe
Assessment of cytochrome P450 (1A2, 2B6, 2C9 and 3A4)
induction in cryopreserved human hepatocytes cultured
in 48-well plates using the cocktail strategy
1
1
2
1
1
Brigitte Gerin , Sylvie Dell’Aiera , Lysiane Richert , Steven Smith , and Hugues Chanteux
1
2
UCB Pharma SA, Non Clinical DMPK, B-1420 Braine l’Alleud, Belgium and Kaly-Cell, Bioparc, Boulevard
Sébastien Brant, 67400 Illkirch, France
abꢁtrꢂct
1
2
3
. A fast, straightforward and cost-effective assay was validated for the assessment of CYP induction in cryopreserved
human hepatocytes cultured in 48-well plates. The cocktail strategy (in situ incubation) was used to assess the
induction of CYP1A2, CYP2B6, CYP2C9 and CYP3A4 by using the recommended probe substrate, i.e. phenacetin,
bupropion, diclofenac and midazolam, respectively.
. Cryopreserved human hepatocytes were treated for 72h with prototypical reference inducers, β-naphthoflavone
(
25 µM), phenobarbital (500 µM) and rifampicin (10 µM) as positive controls for CYP induction. The use of a cocktail
strategy has been validated and compared to the classical approach (single incubation). The need of using phase
II inhibitor (salicylamide) in CYP induction assay was also investigated.
. By using three different batches of cryopreserved human hepatocytes and our conditions of incubations, we
showed that there was no relevant drug–drug interaction using the cocktail strategy. The same conclusions
were observed when a broad range of enzyme activity has to be assessed (wide range of reference inducers, i.e.
EC –E experiment). In addition, the interassay reproducibility assessment showed that the day-to-day variability
5
0
max
was minimal.
4
. In summary, the study showed that the conditions used (probe substrates, concentration of probe substrate and
time of incubation) for the cocktail approach were appropriate for investigations of CYP induction potential of
new chemical entities. In addition, it was also clear that the use of salicylamide in the incubation media was not
mandatory and could generate drug–drug interactions. For this reason, we recommend to not use salicylamide in
CYP induction assay.
Kꢃywordꢁ: CYP induction, cocktail strategy, human hepatocytes
Introduction
drugdiscoveryandshouldbethemorereliablepossibleto
Drug–drug interactions (DDIs) represent a serious
problem in clinical practice. To reduce its occurrence,
great importance has been placed on in vitro studies
as tools for predicting in vivo DDIs, particularly those
resultingfromcytochromeP450(P450)induction.Indeed,
the induction of P450 can reduce the exposure to parent
compound and increase formation of metabolite(s) and
as a consequence can alter the safety and efficacy profile
of a co-administered drug. In vitro screening for potential
induction of human P450s is therefore a crucial part of the
deselect drug candidate before they reach development
or clinical trials. Primary human hepatocytes are widely
recognized as the gold standard in vitro assay to study a
drug’s induction potential (LeCluyse et al. 2000; Huang
& Stifano 2006; Hewitt et al. 2007). Indeed, other test
systems that could be used as alternatives to hepatocytes
suffer from several drawbacks. For example, the receptors
PXR and CAR controlling the expression of CYP2 and
CYP3A families are strongly down-regulated or absent
in cell lines, such as HepG2 or Hepa-1c1c7 (Pascussi
Address for Correspondence: Hugues Chanteux, UCB Pharma, Chemin du Foriest, B-1420 Braine-l’Alleud, Belgium. Tel.: +32(0)2 386 3477,
Fax: +32(0)2 386 2798. E-mail: hugues.chanteux@ucb.com
(
Received 06 June 2012; revised 30 July 2012; accepted 03 August 2012)
3
20

CYP induction assessment using cocktail approach 321
et al. 2001). Moreover, Gerbal-Chaloin et al (2006) have
also demonstrated the presence of cross-talk between
PXR/CYP3A4 and AhR that could result in misleading
interpretation when using hepatoma cell lines for
induction assay. In addition, due to significant species
differences in CYP isoforms, substrate specificities and
induction response, human CYPs induction evaluation
can only be accurately performed with human tissues
that responses to CYP inducers were similar using both
approaches (Lecluyse 2001; Richert et al. 2010). e
use of the cocktail strategy consisted of a simultaneous
evaluation of selected CYP activities by concomitant
incubation of the selected probes. One of the main chal-
lenges of this method remains the biological validation to
ensure that no DDIs occurred during the incubation step
and also the development of a fast, sensitive and robust
analytical method allowing the simultaneous measure-
ment of each metabolite without any cross-interference
from parent compound and other metabolites present
in the samples. Selective probes for CYP1A2, CYP2B6,
CYP2C9 and CYP3A4 were selected to be part of the
cocktail. e advantage of this cocktail is twofold. First,
induction of P450s under the regulation of the three
major transcription factors, AhR (CYP1A), CAR and PXR
(CYP2 and CYP3A) will be detected (Hewitt et al. 2007)
and secondly, these four P450s enzymes belong to the
major P450s involved in drug metabolism. Indeed, 60%
of marketed drugs are metabolized by these enzymes
(Zuber et al. 2002; Guengerich 2006), the other 30%
being metabolized by CYP2D6, not known as an induc-
ible CYP (Li & Kedderis 1997; Rae et al. 2001; Westerink &
Schoonen 2007).
(Lu & Li 2001). e use of cryopreserved hepatocytes for
long-term studies has always been challenged compared
to the fresh hepatocytes. However, in the last couple of
years, cryopreservation techniques have been improved,
allowing a high percentage of viable and plateable
hepatocytes after thawing (Gómez-Lechón et al. 2006; Li
2
007). Cryopreserved hepatocytes are now accepted by
FDA as a valid alternative to fresh human hepatocytes
and they have been shown to be comparable to fresh
hepatocytes in terms of their response to prototypical
CYP1A2, CYP2B6 and CYP3A4 inducers (Huang & Stifano
2

006; Kafert-Kasting et al. 2006; Roymans et al. 2005).
e recent EMA guidance on drug interactions does not
make any distinction between the use of fresh versus
cryopreserved hepatocytes for CYP induction study.
However, it is clearly stated that cultured hepatocytes is
the preferred in vitro system for these studies. In addition
to the general consensus on the use of cryopreserved
hepatocytes for enzyme induction, the following
advantages over fresh hepatocytes are not negligible: (i)
ease of experimentation (planning of experiments), (ii)
repeat experimentation and historical data available (iii)
choice of donor.
is study summarizes the validation of a straightfor-
ward CYP induction assay using cryopreserved human
hepatocytes cultured in 48-well plates with in situ enzy-
matic activities determination using the cocktail strategy.
Mꢃtꢄodꢁ

e work described in this publication is part of all
Chemicals and reagents
the steps undertaken in the context of the CYP induc-
tion international validation project, managed by the
European Union Reference Laboratory for Alternative
Methods to Animal Testing (ECVAM) (2010). e cocktail
approach, used in this validation project for endpoint
measurement, was first described by Kanebratt and
Andersson in (2008). e basis for the general applicabil-
ity of the cocktail approach has been developed within
the EU FP6 project “Vitrocellomics” (Project No. 018940)
to which JRC-ECVAM participated together with the
Karolinska Institute, Pharmacelsus GmbH and Astra
Zeneca Sweden (Mandenius et al. 2011).
All compounds and reagents were of analytical grade.
e reference inducers, probe substrates and metabo-
lites used in the present study were purchased from
Sigma-Aldrich (Saint Louis, MO, USA) except phenobar-
bital which was obtained from Certa (Braine-L’Alleud,
Belgium), midazolam obtained from AAPIN Chemicals
(Oxfordshire, UK) and hydroxybupropion obtained
from BD Gentest (Franklin Lakes, NJ, USA). All the other
chemicals were purchased from different commercial
sources.
Hepatocyte culture and treatment with
More specifically, the main objective of the present
study was to validate the use of the cocktail strategy to
assess CYP P450 inducibility. CYP activities were mea-
sured at the end of the induction period by an in situ
approach (addition of substrates to the washed hepato-
cyte monolayer). Even if the in situ method has the main
disadvantage of a low sensitivity of the activity measure-
ment, this method required much fewer hepatocytes
than the measurement of CYP activities in microsomes
prepared from the monolayer and gives also a global
view of hepatocytes functionality following treatment
with an unknown test compound. In addition, several
studies have already compared CYP activities deter-
mined in microsomes and monolayers and have shown
CYP450 inducers
Cryopreserved human hepatocytes from 6 different
donors (see characteristics in Table 1) were purchased
from CellzDirect (Pittsboro, NC, USA) and were stored
in vapours of liquid nitrogen until use. e hepatocytes
were thawed using cryopreserved hepatocytes recovery
medium (CHRM™) and seeded (250 µL of a cell sus-
6
pension at 10 cell/mL) in a 48-well plate pre-coated
with a single film of collagen (in house coating, ca
50 µg/mL). e hepatocytes were allowed to attach for
ca. 6 h in cryopreserved hepatocytes plating medium
(CHPM™) in a 5% CO :95% air humidified atmosphere
2
at 37°C before starting an overnight pre-incubation
period in WBL medium (Williams E medium containing
©
2013 Informa UK, Ltd.

3
22 B. Gerin et al.
Table 1. Human hepatocytes donor demographics.
Cause of
Death
Batch
Age (years) Sex Race
Smoker Alcohol Use
Drug Use Medications
Serological Data
Hu4037
64
♀
Caucasian Stroke
Yes
No
No
Avonex
Negative except
CMV positive
yrosine
HCTZ
Hydroxycut
None
Hu4122
Hu4198
19
9
♂
♀
Caucasian GSW to head
Yes
No
Yes
No
Yes
No
All negative
Caucasian Anoxia/head
trauma
CMV positive
Hu4237
Hu8063
Hu8127
57
4
♀
♂
♀
Caucasian Anoxia
Caucasian Anoxia
No
No
Yes
No
No
Yes
No
No
Yes
Vitamins
None
CMV positive
All negative
31
Caucasian CV event
after heart
Anti-rejection
meds for heart
transplant
CMV positive
transplant
Glutamax I, 100 IU/mL penicillin, 100 µg/mL streptomy-
cin, 4 µg/mL bovine insulin and 50 µM hydrocortisone
hemissuccinate).
At the end of the overnight pre-incubation period, the
monolayers of cryopreserved human hepatocytes were
treated with the reference inducers β-naphtoflavone
conditions, the incubation was performed in the pres-
ence of 1 mM salicylamide as phase II inhibitor, that was
solubilized in HBSS. When different conditions were
tested (cocktail vs single incubation or with vs without
salicylamide), different wells were used. However, they
were treated in the same way up to the final incubation
step. At the end of the incubation period, the supernatants
were collected and centrifuged for 10 min at 10000×g. e
clear supernatants were stored at −80°C until analyzed.
(
β-NF, 25 µM), phenobarbital (PB, 500 µM) or rifampi-
cin (RIF, 10 µM). e reference inducers were dissolved
in DMSO, which reached a final concentration of 0.1%
in the culture medium. In control hepatocytes, culture
medium was spiked with 0.1% DMSO. e cells were
Protein determination
then incubated in a 5% CO :95% air humidified atmo-
sphere at 37°C for 72 h with medium renewal every ca.
e hepatocytes were dissolved in 0.1 N sodium hydrox-
ide for 2 h at room temperature and the protein content
in each well after the treatment period was determined
using a Pierce BCA (bicinchoninic acid) Protein Assay Kit
(Pierce Biotechnology, Rockford, IL, USA) using bovine
serum albumin as a standard.
2
2
4 h. All conditions were performed in triplicate.
Human liver microsomes
Mixed gender pooled (n = 50) human liver microsomes
(HLM) from Xenotech (Lenexa, KS, USA) were used. e
incubations of HLM (0.1 mg/mL) with phenacetin (60
µM) in presence or absence of salicylamide were carried
out in polypropylene containers at ca 37°C in a shaking
water bath, with 50 mM potassium phosphate buffer (pH
Analytical methods
e formation of acetaminophen, hydroxybupropion,
4’-hydroxydiclofenac and 1’-hydroxymidazolam was
measured by LC-MS/MS analytical method validated
according to Shah et al. (2000). Basically, three valida-
tion batches were processed on three separate days. In
each batch, three sets of 8-point standard curves were
constructed by plotting peak area ratios of analytes/IS
vs nominal analytes concentrations. Back-calculation
of the concentrations using the calibration curves and
linear regression analysis were conducted by using a
7
.4) containing a NADPH regenerating system [NADP
(
1.3 mM), glucose 6-phosphate (3.3 mM), MgCl (3.3 mM)
2
and glucose 6-phosphate dehydrogenase (0.4 U/mL)].
Incubations were performed in triplicate. Final solvent
concentration in incubates was ≤ 1%.
After 20-min incubation, reactions were stopped by
1

volume (buffer incubations) of ice-cold acetonitrile.
e tubes were then thoroughly mixed using a vortex
2
weighted 1/(concentration) . Regression parameters
of the slope, y-intercept, and correlation coefficient (r )
2
mixer and centrifuged (ca 10000×g, 4°C) for 10 min. Clear
supernatants were stored at ≤−80°C prior to analysis for
acetaminophen by LC/MS-MS.
were calculated to evaluate linearity and reproducibility.
In addition, three levels of QC samples (low, medium
and high) were used to evaluate intra-day and inter-day
assay accuracy and precision. Concentrations of QC
sample replicates (n = 5) were determined using a cali-
bration curve obtained on the same day. Relative errors
(REs [%]) and coefficients of variations (CVs [%]) of the
measured concentrations were calculated to evaluate the
validity of the method. e intraday and interday preci-
sions were required to be less than 15%, and the accu-
racy to be within ±15%. Validation parameters are shown
in Table 3. Preparation of sample was straightforward.
Cytochrome P450 activity assays
After a 72-h exposure to vehicle control (0.1% DMSO)
and reference inducers, the hepatocyte monolayers
(vehicle control and induced hepatocytes) were washed
twice with prewarmed PBS and incubated in situ in the
presence of phenacetin, bupropion, diclofenac and/or
midazolam in HBSS (0.4% MeOH final concentration),
either individually (determination of kinetic parameters)
or as a cocktail (final concentrations in Table 2). In some
Xenobiotica

CYP induction assessment using cocktail approach 323
Table 2. Marker reaction for CYP induction assay.
P450
Probe susbtrate
Phenacetin
Bupropion
Enzymatic reaction
Metabolite monitored
Acetaminophen
Concentration in cocktail (µM)
CYP1A2
CYP2B6
CYP2C9
CYP3A4
Phenacetin O-deethylation
Bupropion hydroxylation
Diclofenac hydroxylation
Midazolam hydroxylation
10
250
10
Hydroxybupropion
4’-Hydroxydiclofenac
1’-Hydroxymidazolam
Diclofenac
Midazolam
3
Table 3. Analytical method parameters.
Acetaminophen
Hydroxybupropion
1’-Hydroxymidazolam
4’-Hydroxydiclofenac
2
Linearity (r )
0.9974
0.696
6.63
0.9982
0.260
3.73
0.9960
0.0708
2.90
0.9981
1.38
LOD (nM)
LLOQ (nM)
ULOQ (nM)
Accuracy (%)*
Intraday
9.61
1330
747
579
1920
−3.7/6.6
−0.9/0.7
−7.0/4.6
−1.0/2.3
−15.0/3.7
−13.0/0.0
−9.8/0.6
−6.9/5.7
Interday
Precision (%)*
Intraday
4.0/4.2
6.2/7.1
2.2/4.5
4.0/6.0
1.8/4.2
2.8/4.3
3.2/5.5
3.2/5.5
Interday
LOD: Limit of detection; LLOQ: Lower limit of quantification; ULOQ: Upper limit of quantification.
Minimum and maximum values observed are reported in table.
*
Briefly, 50 µL of clear supernatant was added to 10 µL
of a mixture of acetaminophen–D4, hydroxybupropion–
D6, 4’-hydroxydiclofenac–13C6 and 1’-hydroxymid-
azolam–D4, used as internal standards. e final mixture
was directly injected into the HPLC system (Agilent
Technologies, Palo Alto, CA, USA) equipped with a
Zorbax Eclipse XDB-C18 (5 µm, 50 × 2.1 mm). Detection
and quantification of acetaminophen, hydroxybupro-
pion, 4’-hydroxydiclofenac and 1’-hydroxymidazolam
were performed with a Quattro Micro mass spectrometer
prot were plotted against the substrate concentration for
the determination of K and V according to the bipha-
m
max
sic saturation (CYP1A2 – vehicle control and treated cells
and CYP2B6, vehicle control cells) or Michaelis–Menten
models (all others) using Erithacus GraFit Software (East
Grinstead, UK). Results are presented as individual value
(kinetic parameters) or as the mean ± SD from triplicate
sample for each donor (effect of incubation conditions).
Statistical analyses were performed to compare both
approaches (cocktail and single incubation). T test was
used on log-transformed data to compare absolute enzy-
matic activities determined either by classical approach
(single incubation) or by cocktail approach. Two-factor
ANOVA with replication was used on log-transformed
data to compare both approach on the fold induction
parameter. Statistical analysis was performed using
Microsoft Excel 2003. ANOVA with Dunnett post test
were performed using GraphPad Prism (Version 5.02).
(Waters-Micromass, Manchester, UK). Elution was per-
formed using a mixture of solvent A consisting of water
containing 0.1% trifluoroacetic acid, adjusted to pH 2.4
with ammonium hydroxide and solvent B consisting of
acetonitrile. e proportion of solvent B in the mobile
phase was increased from 5 to 25% for 1 min, increased
from 25 to 50% from 1 to 4.5 min, increased from 50 to
9
1
0% from 4.5 to 5.5 min and decreased to 5% from 7.5 to
1 min. Mass spectrometric conditions for each analyte
and its internal standard are described in Table 4. e
chromatogram of probe substrates, their metabolites and
salicylamide is presented in Figure 1.
For assessment of acetaminophen formation in clear
supernatant coming from liver microsomes incubations,
internal standard were first added to samples and then
they were evaporated to dryness under nitrogen at 45°C.
Rꢃꢁultꢁ
e first work of the present study was the validation of
an analytical method allowing the quantification of the
four metabolites of interest in a same run using adequate
chromatographic conditions to obtain maximum separa-
tion of parent compound from their metabolites within
an acceptable run time. Probe substrate selected for the
assessment of CYP1A2, CYP2B6, CYP2C9 and CYP3A4
activities are described in Table 2 and are those recom-
mended by FDA guidance (Huang & Stifano 2006). e
chromatographic conditions, as described in Material
and Methods section, allowed the separation of all peaks
of interest as shown in the chromatogram presented in
Figure 1. Following the development of this analytical
method, it was validated according to FDA criteria and
used for the present study.

e residue was reconstituted in 60 µL using a mixture of
water containing 0.1% TFA pH 2.4/acetonitrile (95:5; v/v).
e mixture was then injected onto HPLC as described

above.
Data analysis
Metabolite formation rates (hepatocytes and micro-
somal incubation) were normalized to protein content
and incubation time and expressed as pmol/min/mg
protein. Enzyme activities expressed as pmol/min/mg
©
2013 Informa UK, Ltd.

3
24 B. Gerin et al.
Table 4. Mass spectrometric conditions.
Compound
Monitored transitions (Da)
256 → 139
Cone voltage (V)
Collision energy (eV)
Dwell time (s)
0.060
Hydroxybupropion
15
15
25
22
30
30
20
20
20
25
15
17
25
25
17
17
Hydroxybupropion – D6
Acetaminophen
262 → 139
0.060
152 → 110.1
156 → 113.7
342→ 203
0.060
Acetaminophen – D4
0.060
1
1
4
4
’-Hydroxymidazolam
0.060
’-Hydroxymidazolam – D4
’-Hydroxydiclofenac
346 → 203
0.060
312 → 231
0.060
’-Hydroxydiclofenac – 13C6
318 → 237
0.060
Figure 1. MRM chromatograms of a supernatant sample following the in situ incubation of probe substrates with cryopreserved human
hepatocytes seeded in a 48-well plate format (250,000 cells, 30 min incubation with 10 µM phenacetin, 250 µM bupropion, 10 µM diclofenac
and 3 µM midazolam, salicylamide = 10 µM diclofenac, 3 µM midazolam, and 1 mM salicylamide).
Preliminary development
hepatocytes have been investigated under linear condi-
tions as determined in the preliminary developments.
e two main objectives of this assay were (i) to check
that the fold induction is independent on the probe sub-
strate concentration and (ii) to select the concentration
of the probe substrate to be used in the cocktail accord-
ing to the limit of quantification of the analytical method.
To this aim, cryopreserved human hepatocytes from
one representative donor (Hu8063, see characteristics
in Table 1) were treated for 72 h (medium renewal every
24 h) with DMSO (0.1% v/v final concentration, acting
as dose vehicle control), β-NF (25 µM), PB (500 µM) or
RIF (10 µM), as reference inducers of CYP1A2, CYP2B6
A first set of experiments were conducted in order to assess
the linearity of metabolite formation in control hepato-
cytes as well as in induced hepatocytes (72h treatment
with reference inducers, i.e. β-naphthoflavone for CYP1A2,
phenobarbital for CYP2B6 and CYP2C9, and rifampicin for
CYP3A4 and CYP2C9). Probe substrates were incubated
from 15 to 60min at 37°C and metabolites were quantified.
Formation of the corresponding metabolite was linear up
to 60min for phenacetin, bupropion and diclofenac in
control and induced hepatocytes whereas for midazolam,
it was linear up to 60min only in control hepatocytes (lin-
ear up to 30min in induced hepatocytes) (data not shown).
(
and CYP2C9) and CYP3A4 (and CYP2C9), respectively.
Kinetic parameters of probe substrates
As second step, the estimation of the kinetic param-
eters for each probe substrate in control and induced
Following 72 h treatment, relevant CYP activities were
evaluated in situ using the specific probe substrate for
each P450 enzymes, i.e. phenacetin for CYP1A2 (vehicle
Xenobiotica

CYP induction assessment using cocktail approach 325
control and β-NF), bupropion for CYP2B6 (vehicle con-
trol and PB), diclofenac for CYP2C9 (vehicle control, PB
and RIF) and midazolam for CYP3A4 (vehicle control
and RIF). Figure 2 shows for each P450, the direct plot
of the enzymatic activity (expressed in pmol/min/mg
protein) in relation with the concentration of the probe
substrate. Table 5 describes the model used as well as the
kinetic parameters derived from the fitting of the data
presented in Figure 2. For the formation of acetamino-
phen (CYP1A2), the biphasic saturation model had to
be used to fit accurately the data. In addition, the fact
that this model had to be used to describe the forma-
tion of acetaminophen was expected and has already
been described previously (von Moltke et al. 1996).
For the formation of hydroxybupropion (CYP2B6), the
biphasic saturation model was also used but only in
control hepatocytes whereas in induced hepatocytes, the
classical Michaelis–Menten model better fitted the data.
For the formation of hydroxydiclofenac (CYP2C9) and
hydroxymidazolam (CYP3A4), both enzymatic kinetics
were perfectly fitted with the Michaelis–Menten equation
in both conditions (control and induced hepatocytes). As
expected, the K value for each reaction was quite similar
m
in control and induced hepatocyte whereas the Vmax and
CL were always increased in induced hepatocytes com-
int
pared to control hepatocytes. In in vitro CYP induction
studies, the “fold induction” parameters is the final end-
point classically used for the assessment of the potential
of a NCE to induce P450 enzymes. In Table 6, the fold
induction parameter has been calculated by using 4
Figure 2. Cryopreserved human hepatocytes from 1 representative donor (Hu8063) were treated with vehicle (0.1% DMSO; ○) or reference
inducer (●) for 3 days at 37°C with medium renewal every ca. 24 h. Reference inducer was β-NF 25 µM (panel A), PB 1 mM (panel B), RIF
1
0 µM (panel C and D). At the end of the treatment, human hepatocytes were incubated for 60 min in situ with phenacetin (panel A),
bupropion (panel B) and diclofenac (panel C) or 30 min with midazolam (panel D) as probe substrates of CYP1A2, CYP2B6, CYP2C9 and
CYP3A4, respectively.
Table 5. Determination of kinetic parameters.
Target CYP Treatment
Model used
Vehicle control Biphasic saturation
β-NF treated
Biphasic saturation
Vehicle control Biphasic saturation
K (µM) V (pmol/min/mg prot) CLint (µL/min/mg prot) CLint2 (µL/min/mg prot)
m max
CYP1A2
CYP2B6
8.8
7.7
15.5
169
1.76
21.9
0.0675
0.115
67
4.8
0.071
0.0038
–
PB treated
Vehicle control Michaelis–Menten
RIF Treated Michaelis–Menten
CYP3A4/5 Vehicle control Michaelis–Menten
RIF Treated Michaelis–Menten
Michaelis–Menten
31
13
48.8
116
240
19.9
91.6
1.55
8.85
CYP2C9
–
–
–
–
12
20.9
8.29
21.8
2.4
4.2
©
2013 Informa UK, Ltd.

3
26 B. Gerin et al.
different approaches: (i) using Vmax ratio, (ii) using CLint
ratio, (iii) using mean of all enzymatic activities ratio
(see characteristics in Table 1). e results for CYP1A2,
CYP2B6, CYP2C9 and CYP3A4 are shown in Figures 3–6,
respectively. In each figure and for each human donor,
the results were expressed as enzymatic activity in con-
trol and induced hepatocytes (panels A, B, C) and as fold
induction over control (panels D, E, F).
(calculated for each probe substrate concentration), and
(iv) using the enzymatic activity ratio at the anticipated
concentration of probe substrate in the cocktail. e data
showed that fold-induction parameters were similar
whatever the approach used for all P450 tested excepted
for CYP2B6 and CYP3A4 when the CLint ratio was used.
Indeed, as CL is V /K ratio, this discrepancy is easily
As shown in Figure 3, CYP1A2 activities were induced
4- to 35-fold, depending on the donor, after 72 h treat-
ment with 25 µM β-NF. Interestingly enough, the abso-
lute CYP1A2 activity determined in control hepatocytes
and in treated hepatocytes using the cocktail approach
(grey bar) were not statistically different from those
determined by the classical approach (open bars) except
for one donor (panel C) in treated hepatocytes where
the CYP 1A2 activity was 20% lower using the cocktail
approach. However, when looking at the final endpoint
for CYP induction (fold-induction parameter), the
results were similar among the three donors in the sense
that no statistical difference was observed between
the two approaches for the determination of CYP1A2
induction.
int
max
m
explained by the different K values observed in control
m
and induced hepatocytes (within twofold) for these two
P450 enzymes. ese results strongly confirmed that the
determination of the fold-induction parameter is inde-
pendent of the probe substrate concentration. erefore,
the only limitation for the selection of the probe substrate
concentration in the cocktail remained the sensitivity
of the analytical method that should be high enough to
detect the formation of the metabolite in control hepa-
tocytes. Indeed, the activity in control hepatocytes is
important to calculate the fold-induction parameter.
According to these values, the selection of probe sub-
strate concentration in the cocktail was mainly driven
by the sensitivity of the analytical method. Indeed, the
lowest concentration possible of probe substrate was
selected in order to reduce to a maximum the risk of DDI
between probe substrates. e final composition of the
cocktail is described in Table 2.
Validation of the cocktail strategy

e most important objective of the present study was
to validate the use of the cocktail strategy for CYP induc-
tion assessment with regards to the classical approach
(incubation with single substrate). To this aim, at the end
of the induction period (as described in Material and
Methods section), the P450 activities (CYP1A2, CYP2B6,
CYP2C9 and CYP3A4) were determined either individu-
ally (one CYP assessed per sample) or concomitantly
by using the cocktail strategy (four CYPs assessed in the
same sample). e concentration of probe substrate
was the same in both conditions (and described in
Table 2). In addition, to firm up the validation, the experi-
ment was performed on three different human donors
Table 6. Fold-induction calculation using several approaches.
Fold-induction calculated based on ratio
(
induced/control) of
Enzymatic activities
Enzymatic
Vmax CLint activities (mean)*
(concentration in
†
P450
cocktail)
CYP1A2
CYP2B6
CYP2C9
CYP3A4
11
10
12
22
9.1 ± 2.3
12 ± 2
13 ± 1
Figure 3. Cryopreserved human hepatocytes from 3 donors
12 ± 4
(
Hu4037: A and D; Hu4122: B and E; Hu8063: C and F) were
treated with vehicle (0.1% DMSO) or β-NF (25 µM) for 3 days at
7°C with medium renewal every ca. 24 h before the determination
2.1
4.6
2.4
2.6
2.2 ± 0.2
4.1 ± 1.0
2.0 ± 0.3
4.1 ± 0.4
3
*
For each substrate concentration, the ratio of enzymatic activities
in induced hepatocytes over control was calculated and the mean
of all these ratios is represented in this column.
e ratio of enzymatic activities in induced hepatocytes over
control at the intended probe substrate concentration in the
cocktail is represented.
of acetaminophen as marker of CYP1A2 after 30 min incubation
in situ with phenacetin (10 µM) (open bar: single incubation; grey
bar: incubation using the cocktail strategy [Table 2]). Results are the
mean of a triplicate determination and are expressed as enzymatic
activity (A, B, C) or fold-induction over control cells (D, E, F). NS:
Not significant (p > 0.05), **p < 0.01.
†
Xenobiotica

CYP induction assessment using cocktail approach 327
Figure 4. Cryopreserved human hepatocytes from 3 donors (Hu4037: A and D; Hu4122: B and E; Hu8063: C and F) were treated with vehicle
(
0.1% DMSO) or PB (500 µM) 3 days at 37°C with medium renewal every ca. 24 h before the determination of hydroxybupropion as marker of
CYP2B6 after 30 min incubation in situ with bupropion (250 µM) (open bar: single incubation; grey bar: incubation using the cocktail strategy
Table 2]). Results are the mean of a triplicate determination and are expressed as enzymatic activity (A, B, C) or fold-induction over control
[
cells (D, E, F). NS: Not significant (p > 0.05), *p < 0.05.
In Figure 4, the results showed that CYP2B6 activities
were induced 6- to 30-fold, depending on the donor fol-
lowing 72 h treatment with PB (500 µM). As for CYP1A2,
no statistical difference was observed between both
approaches, except in one donor in treated hepatocytes
only (20% of apparent decreased activity using the cock-
tail approach). However, as for CYP1A2, this statistical
difference becomes not significant when the fold-induc-
tion parameter was used as endpoint. Results for CYP2C9
are shown in Figure 5. Since both PB and RIF are known
as prototypical inducers of CYP2C, CYP2C9 activities
were reported following 72 h treatment with PB (500
µM) and RIF (10 µM). Panel D, E and F showed that the
level of CYP2C9 induction was roughly similar regardless
of the inducer used (PB or RIF) and reached twofold to
threefold induction over control. In some cases, a sig-
nificant statistical difference was observed between the
classical approach and the cocktail strategy. However, no
clear trend was observed and the statistical differences
were not always in the same direction (e.g. CYP activ-
ity higher with classical approach compared to cocktail
strategy). ese statistical differences probably result
from a lower experimental variability in these conditions
suggesting that these differences were likely not related
to a specific DDI occurring with cocktail approach. e
results for CYP3A4 are shown in Figure 6. For this P450
enzyme, for all conditions, no statistical difference was
observed between incubation with midazolam alone or
using the cocktail strategy whatever the endpoint used
(absolute enzyme activity or fold-induction). Regarding
the level of CYP3A4 induction, it reached 2- to 4-fold over
control depending on the human donor which was in the
usual range.
Having shown that the cocktail approach could be
used for CYP induction and that no significant DDI
occurred during the incubation step, additional valida-
tion studies were performed to further firm up the use of
the cocktail. First, we wanted to confirm that the cocktail
©
2013 Informa UK, Ltd.

3
28 B. Gerin et al.
Figure 5. Cryopreserved human hepatocytes from 3 donors (Hu4037: A and D; Hu4122: B and E; Hu8063: C and F) were treated with
vehicle (0.1% DMSO), PB (500 µM) or RIF (10 µM) for 3 days at 37°C with medium renewal every ca. 24 h before the determination of 4’-
hydroxydiclofenac as marker of CYP2C9 after 30 min incubation in situ with diclofenac (10 µM) (open bar: single incubation; grey bar:
incubation using the cocktail strategy [Table 2]). Results are the mean of a triplicate determination and are expressed as enzymatic activity (A,
B, C) or fold-induction over control cells (D, E, F). NS: Not significant (p > 0.05) *p < 0.05, **p < 0.01.
approach could also be used to derive EC and E val-
As last validation experiment, we evaluated the inter-
day reproducibility of the present CYP induction assay.
Hepatocytes from three different donors were treated
with reference inducers for 72 h according the classical
protocol. At the end of induction period, fold induction
was assessed using the cocktail approach only. e full
assay (from hepatocytes thawing up to enzymatic activi-
ties measurement) was repeated on three different occa-
sions to check for inter-day reproducibility. Results are
shown on Figure 8 and clearly demonstrated minimal
difference between assays for all CYPs as well as for the
three donors.
50
max
ues when different concentrations of reference inducers
are used. To this aim, hepatocytes were incubated with
increasing concentrations of reference inducers and then,
enzymatic activities were determined, either using cock-
tail approach or single incubation. As shown in Figure 7,
no difference between fold-induction determined using
cocktail approach (open symbol) and using single incu-
bation (closed symbol) was observed for CYP1A2 (panel
A), CYP2B6 (panel B), CYP2C9 (panel C) and CYP3A4
(
panel D). Even though final EC and E values could
50 max
not be determined accurately for CYP2B6 and CYP2C9
mainly because of high concentration of phenobarbital to
be used, the results showed no difference between both
approaches on a wide range of concentration of refer-
ence inducers. For CYP1A2, EC and E values were, for
Use of salicylamide in CYP induction assay
Salicylamide has sometimes been used in CYP induction
assay in order to inhibit phase II metabolism, especially
in the case of midazolam. Indeed, 1’OH-midazolam is
further metabolized by phase II enzymes to midazolam-
O-glucuronide. To prevent this phase II metabolism
and to obtain the true CYP3A4 activity, salicylamide has
been used by some authors (Ring et al. 2005). In this
paper, we investigated the impact of salicylamide on the
50
max
single incubation, 3.5±0.6 µM and 169±16 fold increase,
respectively and, for cocktail incubation, 3.6±0.4 µM and
1
34±9 fold increase, respectively. For CYP3A4, EC and
50
Emax values were, for single incubation, 1.6±1.3 µM and
6±4 fold increase, respectively and, for cocktail incuba-
tion, 0.8±0.3 µM and 18±4 fold increase, respectively.
2
Xenobiotica

CYP induction assessment using cocktail approach 329
Figure 6. Cryopreserved human hepatocytes from three donors (Hu4037: A and D; Hu4122: B and E; Hu8063: C and F) were treated with
vehicle (0.1% DMSO) or RIF (10 µM) for 3 days at 37°C with medium renewal every ca. 24 h before the determination of 1’- hydroxymidazolam
as marker of CYP3A4 after 30 min incubation in situ with midazolam (3 µM) (open bar: single incubation; grey bar: incubation using the
cocktail strategy [Table 2]). Results are the mean of a triplicate determination and are expressed as enzymatic activity (A, B, C) or fold-
induction over control cells (D, E, F). NS: Not significant (p > 0.05).
absolute CYP activity and on the fold-induction param-
eters to clearly identify if it was worth to use it, since, to
our knowledge, no available data showed the advantage
of salicylamide in CYP induction studies.
II enzymes. Interestingly enough, the results presented in
Figure 10 showed that salicylamide dose dependently
inhibits CYP1A2 activities with approximately 50% inhi-
bition at 2 mM which is in accordance with our results
obtained in hepatocytes. Regarding CYP2B6 (panels C
and G from Figure 9), no significant effect of salicylamide
was observed on CYP activities and on the fold induction
parameter. For CYP2C9 (panels C and G from Figure 9),
there was a trend of increased CYP activities in presence
of salicylamide. Even though this increase was moderate
(maximum 30%), it reached statistically significance only
in induced hepatocytes. However, when looking at the
fold induction parameter, this statistical difference disap-
peared. As expected for CYP3A4, the presence of salicyl-
amide increased the absolute CYP3A4 activities in both
control and induced hepatocytes. However, when results
were processed to obtain the fold induction parameters,
there was no effect of salicylamide.
To this aim, three donors (Hu4037, Hu4122, Hu8063)
were used and salicylamide (1mM) was added dur-
ing the incubation with probe substrates (classical and
cocktail approach were compared). For sake of clarity,
only the results from one representative donor using the
classical approach for the measurement of CYP activities
are represented in Figure 9. Results for CYP1A2, CYP2B6,
CYP2C9 and CYP3A4 are represented in panels A-E,
B-F, C-G, and D-H, respectively. In panels A, B, C and D,
results are expressed as absolute CYP activities (pmol/
min/mg prot) whereas in panels E, F, G and H, results
are reported in fold-induction. Results showed that
salicylamide significantly inhibited CYP1A2 activities in
control and induced hepatocytes (approximately 60%
inhibition, panel A). However, no significant difference
was observed between incubation with or without sali-
cylamide when the fold induction parameter was used as
endpoint (panel E). In order to confirm the inhibition of
CYP1A2 by salicylamide, human liver microsomes were
used to rule out potential effect of salicylamide on phase
Diꢁcuꢁꢁion
As of today, the scientific community (industry, regula-
tory, academia) unanimously accepted and considered
that primary human hepatocytes are the gold standard
©
2013 Informa UK, Ltd.

3
30 B. Gerin et al.
Figure7. CryopreservedhumanhepatocytesfromonerepresentativedonorHu4237weretreatedwithvehicleorwithincreasingconcentrations
of reference inducers, i.e. β-NF (1-40 µM, panel A), phenobarbital (20–2000 µM, panel B and C), rifampicin (0.03–30 µM, panel D). CYP1A2
(
panel A), CYP2B6 (panel B), CYP2C9 (panel C) and CYP3A4 (panel D) activities were determined following in situ incubation using cocktail
strategy (open symbol ○) or using single incubation (closed symbol ●). Data are expressed as fold induction over vehicle treated hepatocytes.
in vitro assay to assess the potential for drug candidates
to induce human CYP expression. Human hepatocytes
are a complete cellular system with properties analo-
gous the liver function, which is essential to effectively
model the inducibility of drug candidates. Indeed, pri-
mary human hepatocytes have been shown to efficiently
model human in vivo induction responses and are well
recognized by FDA and EMA as the preferred in vitro tool
for assessing induction potential (Huang & Stifano 2006;
European Medecines Agency and Committee for Human
Medicinal Products, 2010). Regarding the use of fresh or
cryopreserved hepatocytes for these CYP induction stud-
ies, the consensus has never been so clear-cut mainly
because of two major drawbacks, i.e. the platebility and
viability after thawing. However, these last years, con-
siderable improvements in the cryopreservation process
have been made making that cryopreserved hepatocytes
do no longer suffer from their historical inconveniences.
In addition to that, the convenience of cryopreserved
human hepatocytes resulted in considerable regain of
interest and demonstrated their ability to be an adequate
model for the in vitro assessment of CYP induction.
Among the different endpoints that could be used for
the evaluation of CYP induction, enzyme activity is by far
and in most cases, the most recommended approach for
a reliable prediction of the in vivo situation. In the pres-
ent study, in situ determination of the enzyme activity
has been preferred for two major reasons, (i) use of less
hepatocytes and (ii) increased throughput compared to
microsomes that required additional preparation steps.
In order to minimize the use of hepatocytes, a so-called
cocktail approach was investigated. In addition to the
lower total amount of hepatocytes required which is not
negligible knowing the price of hepatocytes, the cock-
tail approach has many other interesting advantages.
Indeed, since multiple CYPs are measured simultane-
ously in one samples, there is much less samples to pre-
pare and handle as well as to analyze, therefore saving
human ressources and reducing utilization time of mass
spectrometer which is critical in discovery.
Even though the use of cryopreserved human hepa-
tocytes is becoming more common these days, we did
not find robust validation in the literature for the use of
cocktail approach in CYP induction assay. Indeed, most
of the study described in the literature using the cocktail
approach, were for CYP inhibition experiment in liver
microsomes (Dierks et al. 2001; Testino & Patonay 2003;
Floby et al. 2004; Kim et al. 2005). Having a deeper look in
the CYP induction studies, it appeared that they all have
some gaps that the present study tries to fill. For example,
Dixit et al. (2007) determined the enzyme activity using
microsomes and had to use two different cocktails to
avoid DDI. In the study of Mohutsky et al. (2005), CYP2B6
was not included in the P450 enzymes investigated. In the
study from Lahoz et al. (2008), benzoxyresorufin was used
as probe substrate for CYP2B6 which is not considered by
regulatory agencies as preferred substrate for CYP2B6.
In addition, the incubation time used for the CYP activ-
ity determination was quite long (2h) even though they
claimed that it was under linear condition. is is in
Xenobiotica

CYP induction assessment using cocktail approach 331
concentration and time of incubation. Incubation time is
important because linearity of the enzymatic reaction is
not the same for all probes and a compromise between
linearity and detection should be made in the case of the
cocktail approach. In some studies, the linearity of the
enzymatic reaction was not taken into account in the val-
idation of the cocktail (Testino & Patonay 2003; Tolonen
et al. 2007). Based on our data, we selected an incuba-
tion time as shorter as possible for two reasons (i) first
to remains within the linear conditions for midazolam
and (ii) to limit phase II metabolism of 1’HO-midazolam.
Regarding the selection of the concentration of each
probe substrate, this was a compromise between limit
of detection of the analytical method and the risk of DDI
between probes. In order to select the concentration
that will be used in the cocktail, we first performed a Km/
Vmax experiment under control and treated conditions
(
with reference inducers). Our results (K values) were
m
in accordance with published values (von Moltke et al.
996; Faucette et al. 2000; Brown et al. 2007; Hallifax et
1
al. 2008). Enzymatic kinetic for acetaminophen forma-
tion was biphasic as expected due to the involvement
of low affinity P450 enzymes for this reaction (CYP2C9,
CYP2C19, CYP2E1, CYP2A6, CYP2D6) (Venkatakrishnan
et al. 1998; Kobayashi et al. 1999). For CYP3A4 and
CYP2C9, enzymatic kinetics was classical Michaelis–
Menten reaction whereas for CYP2B6, it was a biphasic
kinetic in control hepatocytes and Michaelis–Menten in
induced hepatocytes. On a first instance, this could be
surprising to observe two different kind of kinetics within
the same batch of hepatocytes. However, an easy expla-
nation is that under induced conditions, the expression
CYP2B6 has been so up-regulated by the inducer that it
masks the activity of the other low affinity enzymes mak-
ing that the biphasic model is no longer the adequate
model to fit the data. In addition, this K /V experiment
also clearly showed, as expected, that the fold-induction
parameters is independent of the substrate concentra-
tion. On that basis, concentration of each probe substrate
in the cocktail was selected to find a compromise between
the lower concentration of probe substrate achievable to
minimize the risk of DDI and a concentration allowing
detection of formation of metabolite in control condi-
tions to calculate the fold-induction parameters. Now
that the conditions for the use of the cocktail approach
were selected (time of incubation and concentration of
each probe substrate), the final step of the study was the
validation, strictly speaking, of this approach compared
to classical one (single incubation of probe substrate).
First, even though this was not the primary objective
of the present study, our results are in accordance with
other studies and confirmed that in the conditions we
used, cryopreserved human hepatocytes are a suitable
model to evaluate the potential of new chemical enti-
ties to induce the expression of P450 enzymes. Indeed,
β-NF elicited a 4- to 35-fold induction of CYP1A2, PB a
6- to 30-fold induction of CYP2B6 and RIF a 2- to 5-fold
induction of CYP3A4. ese values were obviously
Figure 8. Cryopreserved human hepatocytes from three different
donors (Hu4198, Hu4237 and Hu8127) were used for interassay
variability assessment. Human hepatocytes were incubated with
vehicle (0.1% DMSO), β-NF (25 µM – panel A), PB (500 µM – panel
B and C) or RIF (10 µM – panel D) for 3 days at 37°C with medium
renewal every ca. 24h. CYP450 activities were determined following
30min incubation in situ with cocktail solution (composition
described in Table 2). Results are expressed as fold induction over
vehicle treated hepatocytes for CYP1A2 (panel A), 2B6 (panel B),
2
C9 (panel C) and 3A4 (panel D). e experiment was repeated on
three different occasions represented by three different colour codes
■ assay 1, ■ assay 2, □ assay 3). Results are the mean of a triplicate
(
determination with error bars being the standard deviation.
contradiction with our data where we showed that in
induced hepatocytes, the formation of 1’OH-midazolam
was linear up to 30min. However, it was not clear if they
checked the linearity also in induced hepatocytes where
we observed that this could be different from the control
hepatocytes. Recently, Feidt et al. (2010) issued another
study with the use of cocktail for CYP induction. However,
in order to rule out potential risk of DDI during the devel-
opment of its cocktail, other probe substrate than those
classically used and recommended by FDA and EMA
were used (e.g. use of atorvastatin as marker for CYP3A4
instead of midazolam or testosterone, propafenone for
CYP2D6 instead of dextromethorphan or bufuralol).
Indeed, as already mentioned, the major issue with this
cocktail approach remained the potential cross-interac-
tion between probes during the incubation process.
Since the objective of the present study was the devel-
opment of a straightforward CYP induction assay that
could be used for submission, classical probes were used
and limited to those strictly required, i.e. phenacetin
for CYP1A2, bupropion for CYP2B6, and midazolam for
CYP3A4. Even if not strictly needed, diclofenac (CYP2C9)
was also added as a marker for CYP2C.
m
max
Another critical point with the use of cocktails is
the selection of the incubation conditions like probe
©
2013 Informa UK, Ltd.

3
32 B. Gerin et al.
Figure 9. Cryopreserved human hepatocytes from one representative donor Hu4037 were treated with vehicle (0.1% DMSO), β-NF (25 µM),
PB (500 µM) or RIF (10 µM) for 3 days at 37°C with medium renewal every ca. 24 h. CYP450 activities were determined following 30 min
incubation in situ with phenacetin (10 µM), bupropion (250 µM), diclofenac (10 µM) or midazolam (3 µM) (single incubation) as substrate
marker for CYP1A2, 2B6, 2C9 and 3A4, respectively, in the absence (open bar) and in the presence of 1 mM salicylamide (grey bar). Results
are the mean of a triplicate determination and are expressed as enzymatic activity (A, B, C, D) or fold-induction over control cells (E, F, G, H).
NS: Not significant (p > 0.05), *p < 0.05, **p < 0.01, ***p < 0.001.
dependent on the donor but remained in the classical
range of fold-induction usually described (Faucette et al.
specific P450 enzyme, the statistical difference observed
was not reproduced in all donors tested and was not in
the same direction between donors. ese observations
strengthen the hypothesis that these statistical differ-
ences were without or of low biological relevance. In
addition, the only P450 enzymes for which a statistical
difference was observed on the fold induction endpoint,
was CYP2C9 which is not per se, mandatory in the CYP
induction assay. Moreover, when observed, the differ-
ence never exceeded 30% which remained within the
variability of the assay.
2
000; Lu & Li, 2001; Madan et al. 2003; Zhang et al. 2010).
Most importantly are the results of the cocktail which
clearly and unambiguously showed that the use of cock-
tail strategy as defined in the present study is adequate
and optimal for CYP induction assay. Indeed, in these
conditions, no significant and biologically relevant
DDIs were observed between probe substrates since the
absolute CYP activities were similar using both meth-
ods. In addition, this was further confirmed by using
three different batches of cryopreserved human hepa-
tocytes. Even though if in some conditions, statistical
differences were observed, they were mainly ascribed
to a lower variability in those conditions. Indeed, for a
e adequacy of cocktail approach was also con-
firmed in an EC -E experimental setting where a huge
5
0
max
range of enzyme activity has to be assessed. Indeed,
the results showed no difference between cocktail and
Xenobiotica

CYP induction assessment using cocktail approach 333
on the screening strategy put in place in the company.
Even though the current recommendation in FDA DDI
guidance is adopted in the final document, we strongly
believe that measurement of CYP activities in CYP induc-
tion assay still deserve a place in the overall assessment
of DDI potential for all the reasons explained previously.
Even if mRNA data are used for CYP induction assay,
measuring enzymatic activities in parallel will provide
more mechanistic insight in the overall picture of DDI
and brings a better knowledge of the kind of interaction
that can be expected.
In order to close the debate on the real need to use
phase II inhibitor in CYP induction study, we evaluated
the effect of salicylamide on (i) absolute CYP activity
and (ii) on the fold induction parameters, the final end-
point used in this CYP induction study. Our data clearly
demonstrated that (i) at the concentration used (1 mM),
salicylamide inhibits CYP1A1/2, (ii), as expected, the
absolute CYP3A4 activities are enhanced due to decrease
of phase II metabolism of 1’HO-midazolam and (iii) the
use of salicylamide did not increase significantly the fold
induction parameter. To our knowledge, the inhibition of
phenacetin O-deethylation reaction by high concentra-
tion of salicylamide (≥1 mM) has never been reported.
MacDonald et al. (2004) reported that salicylamide (250
µM) had no effect on EROD activities (marker of CYP1A)
in HepG2 cells, which is in accordance with our results
but they did not investigate higher concentration of sali-
cylamide which are needed to inhibit phase II reaction.
In conclusion, the cocktail assay validated in the
present study using cryopreserved human hepatocytes
cultured in 48-well plate is a rapid, straightforward and
cost-effective tool to assess induction of CYP1A2, 2B6,
Figure10. Effectofsalicylamideontheformationofacetaminophen
in human liver microsomes. Human liver microsomes were
incubated 20 min at 37°C with 60 µM phenacetin and salicylamide.
Results are represented as mean ± SD (n = 3). ANOVA with Dunnett
post test (***p < 0.001).
single incubation whatever the concentration of refer-
ence inducers used.
Finally, the final assay (using cocktail incubation)
was shown to be reproducible as no significant differ-
ence were observed among results when performed on
three different occasions. To our knowledge, the present
study is the first validation of the cocktail strategy for its
application in CYP induction assay with the advantage of
using the FDA and EMA recommended probe substrate
for assessment of CYP1A2, 2B6, 2C9 and 3A4 activities.
However, the recent new FDA draft guidance (Huang
2
012) on DDI changes things since in the current ver-
2
C9 and 3A4. In addition, the use of salicylamide is not
sion, it is recommended to use mRNA as endpoint for
CYP induction. Compared to mRNA, the main drawback
of measuring CYP activities as endpoint in CYP induc-
tion is in the case where the test compound is a CYP3A4
mechanism-based inhibitor and therefore could gener-
ated erroneous conclusions that such compound will not
cause clinical induction, yet the other induced but non-
inactivated enzymes will still have increased expression
and activity. However, this scenario will only occurred if
the two following conditions are met: (i) the compound is
a potent mechanism-based inhibitor of CYP3A4 and (ii)
mechanism-based inhibition potential of the compound
is not known at the time of CYP induction. Usually, the
CYP inhibition profile of NCE is identified before CYP
induction meaning that, in the latter case, the right inter-
pretation of CYP induction data could be drawn using
enzymatic activities. In addition, mechanism-based
inhibition do not always translate into no observed
effect in CYP induction assay, it really depends on the
potency of inactivation and affinity for CYP3A4 as well
as on the affinity for PXR. Moreover, potent mechanism-
based inhibitor of CYP3A4 are dismissed during devel-
opment meaning that they should not be tested in CYP
induction assay. However, this scenario really depends
recommended in CYP induction because of it can gener-
ate some DDI without providing any added value.
Dꢃclꢂrꢂtion of intꢃrꢃꢁt
e authors report no declarations of interest
Rꢃfꢃrꢃncꢃꢁ
Brown HS, Chadwick A, Houston JB. (2007). Use of isolated hepatocyte
preparations for cytochrome P450 inhibition studies: comparison
with microsomes for Ki determination. Drug Metab Dispos
3
5:2119–2126.
Dierks EA, Stams KR, Lim HK, Cornelius G, Zhang H, Ball SE. (2001).
A method for the simultaneous evaluation of the activities of
seven major human drug-metabolizing cytochrome P450s using
an in vitro cocktail of probe substrates and fast gradient liquid
chromatography tandem mass spectrometry. Drug Metab Dispos
2
9:23–29.
Dixit V, Hariparsad N, Desai P, Unadkat JD. (2007). In vitro LC-MS
cocktail assays to simultaneously determine human cytochrome
P450 activities. Biopharm Drug Dispos 28:257–262.
European Medecines Agency and Committee for Human Medicinal
Products (CHMP). (2010). Guideline on the Investigation of Drug
Interactions.
Faucette SR, Hawke RL, Lecluyse EL, Shord SS, Yan B, Laethem RM,
Lindley CM. (2000). Validation of bupropion hydroxylation as a
©
2013 Informa UK, Ltd.

3
34 B. Gerin et al.
selective marker of human cytochrome P450 2B6 catalytic activity.
Drug Metab Dispos 28:1222–1230.
Li AP, Kedderis GL. (1997). Primary hepatocyte culture as an experimental
model for the evaluation of interactions between xenobiotics and
drug-metabolizing enzymes. Chem Biol Interact 107:1–3.
Feidt DM, Klein K, Hofmann U, Riedmaier S, Knobeloch D, Thasler
WE, Weiss TS, Schwab M, Zanger UM. (2010). Profiling
induction of cytochrome p450 enzyme activity by statins using
a new liquid chromatography-tandem mass spectrometry
cocktail assay in human hepatocytes. Drug Metab Dispos
Lu C, Li AP. (2001). Species comparison in P450 induction: effects of
dexamethasone, omeprazole, and rifampin on P450 isoforms 1A
and 3A in primary cultured hepatocytes from man, Sprague-Dawley
rat, minipig, and beagle dog. Chem Biol Interact 134:271–281.
MacDonald CJ, Ciolino HP, Yeh GC. (2004). e drug salicylamide is
an antagonist of the aryl hydrocarbon receptor that inhibits signal
transduction induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin.
Cancer Res 64:429–434.
Madan A, Graham RA, Carroll KM, Mudra DR, Burton LA, Krueger LA,
Downey AD, Czerwinski M, Forster J, Ribadeneira MD, Gan LS,
LeCluyse EL, Zech K, Robertson P Jr, Koch P, Antonian L, Wagner
G, Yu L, Parkinson A. (2003). Effects of prototypical microsomal
enzyme inducers on cytochrome P450 expression in cultured
human hepatocytes. Drug Metab Dispos 31:421–431.
3
8:1589–1597.
Floby E, Briem S, Terelius Y, Sohlenius-Sternbeck AK. (2004). Use of a
cocktail of probe substrates for drug-metabolizing enzymes for the
assessment of the metabolic capacity of hepatocyte preparations.
Xenobiotica 34:949–959.
Gerbal-Chaloin S, Pichard-Garcia L, Fabre JM, Sa-Cunha A, Poellinger
L, Maurel P, Daujat-Chavanieu M. (2006). Role of CYP3A4 in
the regulation of the aryl hydrocarbon receptor by omeprazole
sulphide. Cell Signal 18:740–750.
Gómez-Lechón MJ, Lahoz A, Jiménez N, Vicente Castell J, Donato MT.
(
2006). Cryopreservation of rat, dog and human hepatocytes:
Mandenius CF, Andersson TB, Alves PM, Batzl-Hartmann C, Björquist
P, Carrondo MJ, Chesne C, Coecke S, Edsbagge J, Fredriksson
JM, Gerlach JC, Heinzle E, Ingelman-Sundberg M, Johansson I,
Küppers-Munther B, Müller-Vieira U, Noor F, Zeilinger K. (2011).
Toward preclinical predictive drug testing for metabolism and
hepatotoxicity by using in vitro models derived from human
embryonic stem cells and human cell lines – a report on the
Vitrocellomics EU-project. Altern Lab Anim 39:147–171.
Mohutsky MA, Petullo DM, Wrighton SA. (2005). e use of a substrate
cassette strategy to improve the capacity and throughput of
cytochrome P450 induction studies in human hepatocytes. Drug
Metab Dispos 33:920–923.
Pascussi JM, Drocourt L, Gerbal-Chaloin S, Fabre JM, Maurel P, Vilarem
MJ. (2001). Dual effect of dexamethasone on CYP3A4 gene
expression in human hepatocytes. Sequential role of glucocorticoid
receptor and pregnane X receptor. Eur J Biochem 268:6346–6358.
Rae JM, Johnson MD, Lippman ME, Flockhart DA. (2001). Rifampin is a
selective, pleiotropic inducer of drug metabolism genes in human
hepatocytes: studies with cDNA and oligonucleotide expression
arrays. J Pharmacol Exp er 299:849–857.
influence of preculture and cryoprotectants on recovery,
cytochrome P450 activities and induction upon thawing.
Xenobiotica 36:457–472.
Guengerich FP. (2006). Cytochrome P450s and other enzymes in drug
metabolism and toxicity. AAPS J 8:E101–E111.
Hallifax D, Galetin A, Houston JB. (2008). Prediction of metabolic
clearance using fresh human hepatocytes: comparison with
cryopreserved hepatocytes and hepatic microsomes for five
benzodiazepines. Xenobiotica 38:353–367.
Hewitt NJ, Lecluyse EL, Ferguson SS. (2007). Induction of
hepatic cytochrome P450 enzymes: methods, mechanisms,
recommendations, and in vitro-in vivo correlations. Xenobiotica
3
7:1196–1224.
Huang S-M. (2012). Drug Interaction Studies
–
Study Design,
Data Analysis, Implications for Dosing, and Labeling
Recommendations – Draft Guidance. CDER (Center for Drug
Evaluation and Research) – FDA – http://www fda gov/downloads/
Drugs/GuidanceComplianceRegulatoryInformation/Guidances/
ucm292362 pdf
Huang S-M, Stifano T. (2006). Guidance for Industry: Drug interaction
studies - Study design, data analysis, and implications for dosing
and labelling. Draft Guidance, September, 2006. http://www fda
gov/cber/gdlns/interactstud htm
Kafert-Kasting S, Alexandrova K, Barthold M, Laube B, Friedrich G,
ArsenievL,HengstlerJG.(2006).Enzymeinductionincryopreserved
human hepatocyte cultures. Toxicology 220:117–125.
Kanebratt KP, Andersson TB. (2008). Evaluation of HepaRG cells as an
in vitro model for human drug metabolism studies. Drug Metab
Dispos 36:1444–1452.
Kim MJ, Kim H, Cha IJ, Park JS, Shon JH, Liu KH, Shin JG. (2005).
High-throughput screening of inhibitory potential of nine
cytochrome P450 enzymes in vitro using liquid chromatography/
tandem mass spectrometry. Rapid Commun Mass Spectrom
RichertL, AbadieC, BonetA, HeydB, MantionG, AlexandreE, Bachellier
P, Kingston S, Pattenden C, Illouz S, Dennison A, Hoffmann S,
Coecke S. (2010). Inter-laboratory evaluation of the response of
primary human hepatocyte cultures to model CYP inducers - a
European Centre for Validation of Alternative Methods (ECVAM) -
funded pre-validation study. Toxicol In Vitro 24:335–345.
Ring BJ, Patterson BE, Mitchell MI, Vandenbranden M, Gillespie J, Bedding
AW, Jewell H, Payne CD, Forgue ST, Eckstein J, Wrighton SA, Phillips
DL. (2005). Effect of tadalafil on cytochrome P450 3A4-mediated
clearance: studies in vitro and in vivo. Clin Pharmacol er 77:63–75.
Roymans D, Annaert P, Van Houdt J, Weygers A, Noukens J,
Sensenhauser C, Silva J, Van Looveren C, Hendrickx J, Mannens
G, Meuldermans W. (2005). Expression and induction potential
of cytochromes P450 in human cryopreserved hepatocytes. Drug
Metab Dispos 33:1004–1016.
1
9:2651–2658.
Kobayashi K, Nakajima M, Oshima K, Shimada N, Yokoi T, Chiba
K. (1999). Involvement of CYP2E1 as a low-affinity enzyme in
phenacetin O-deethylation in human liver microsomes. Drug
Metab Dispos 27:860–865.
Shah VP, Midha KK, Findlay JW, Hill HM, Hulse JD, McGilveray IJ,
McKay G, Miller KJ, Patnaik RN, Powell ML, Tonelli A, Viswanathan
CT, Yacobi A. (2000). Bioanalytical method validation–a revisit
with a decade of progress. Pharm Res 17:1551–1557.
Lahoz A, Donato MT, Picazo L, Castell JV, Gómez-Lechón MJ. (2008).
Assessment of cytochrome P450 induction in human hepatocytes
using the cocktail strategy plus liquid chromatography tandem
mass spectrometry. Drug Metab Lett 2:205–209.
Testino SA Jr, Patonay G. (2003). High-throughput inhibition screening
of major human cytochrome P450 enzymes using an in vitro
cocktail and liquid chromatography-tandem mass spectrometry.
J Pharm Biomed Anal 30:1459–1467.
LeCluyse E, Madan A, Hamilton G, Carroll K, DeHaan R, Parkinson A.
Tolonen A, Petsalo A, Turpeinen M, Uusitalo J, Pelkonen O. (2007).
In vitro interaction cocktail assay for nine major cytochrome
P450 enzymes with 13 probe reactions and a single LC/MSMS
run: analytical validation and testing with monoclonal anti-CYP
antibodies. J Mass Spectrom 42:960–966.
Venkatakrishnan K, von Moltke LL, Greenblatt DJ. (1998). Human
cytochromes P450 mediating phenacetin O-deethylation in vitro:
validation of the high affinity component as an index of CYP1A2
activity. J Pharm Sci 87:1502–1507.
(
2000). Expression and regulation of cytochrome P450 enzymes
in primary cultures of human hepatocytes. J Biochem Mol Toxicol
4:177–188.
1
LeCluyse EL. (2001). Human hepatocyte culture systems for the in vitro
evaluation of cytochrome P450 expression and regulation. Eur J
Pharm Sci 13:343–368.
Li AP. (2007). Human hepatocytes: isolation, cryopreservation and
applications in drug development. Chem Biol Interact 168:16–29.
Xenobiotica

CYP induction assessment using cocktail approach 335
von Moltke LL, Greenblatt DJ, Duan SX, Schmider J, Kudchadker
L, Fogelman SM, Harmatz JS, Shader RI. (1996). Phenacetin
O-deethylation by human liver microsomes in vitro: inhibition
by chemical probes, SSRI antidepressants, nefazodone and
venlafaxine. Psychopharmacology (Berl) 128:398–407.
Zhang JG, Ho T, Callendrello AL, Crespi CL, Stresser DM. (2010). A
multi-endpoint evaluation of cytochrome P450 1A2, 2B6 and 3A4
induction response in human hepatocyte cultures after treatment
with β-naphthoflavone, phenobarbital and rifampicin. Drug
Metab Lett 4:185–194.
Westerink WM, Schoonen WG. (2007). Cytochrome P450 enzyme levels
in HepG2 cells and cryopreserved primary human hepatocytes
and their induction in HepG2 cells. Toxicol In Vitro 21:1581–1591.
Zuber R, Anzenbacherová E, Anzenbacher P. (2002). Cytochromes P450
and experimental models of drug metabolism. J Cell Mol Med
6:189–198.
©
2013 Informa UK, Ltd.