Characterization of marmoset CYP2B6: CDNA cloning, protein expression and enzymatic functions

Article abstract of DOI:10.1016/j.bcp.2013.01.024

The common marmoset is a promising species for evaluating the safety of drug candidates. To further understand the capacity for drug metabolism in marmosets, a cDNA encoding a CYP2B enzyme was cloned from the total RNA fraction of marmoset liver by 3′- and 5′-RACE methods. Nucleotide and deduced amino acid sequences showed 90.8 and 86.2% identity, respectively, with human CYP2B6. The marmoset CYP2B6 (marCYP2B6) protein was expressed in insect cells, and its enzymatic properties were compared with those of human (humCYP2B6) and cynomolgus monkey (cynCYP2B6) orthologs in liver and insect cell microsomes. Enzymatic functions were examined for the oxidation of 7-ethoxy-4-(trifluoromethyl)coumarin (7-ETC), bupropion (BUP) and efavirenz (EFV). The kinetic profiles for the oxidation of the three substrates by liver microsomal fractions were similar between humans and cynomolgus monkeys (biphasic for 7-ETC and monophasic for BUP and EFV), but that of marmosets was unique (monophasic for 7-ETC and biphasic for BUP and EFV). Recombinant enzymes, humCYP2B6 and cynCYP2B6, also yielded similar kinetic profiles for the oxidation of the three substrates, whereas marCYP2B6 showed activity only for 7-ETC hydroxylation. In silico docking simulations suggested that two amino acid residues, Val-114 and Leu-367, affect the activity of marCYP2B6. In fact, a marCYP2B6 mutant with substitutions V114I and L367V exhibited BUP hydroxylase activity that was 4-fold higher than that of humCYP2B6, while its EFV 8-hydroxylase activity was only 10% that of the human enzyme. These results indicate that the amino acids at positions 114 and 367 affect the enzymatic capacity of marmoset CYP2B6.

Full text of DOI:10.1016/j.bcp.2013.01.024

Biochemical Pharmacology 85 (2013) 1182–1194
Contents lists available at SciVerse ScienceDirect
Biochemical Pharmacology
journal homepage: www.elsevier.com/locate/biochempharm
Characterization of marmoset CYP2B6: cDNA cloning, protein
expression and enzymatic functions
Kei Mayumi a, Nobumitsu Hanioka a, Kazufumi Masuda b, Akiko Koeda c, Shinsaku Naito d,
Atsuro Miyata e, Shizuo Narimatsu a,
*
a Laboratory of Health Chemistry, Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, Okayama, Japan
b Graduate School of Clinical Pharmacy, Shujitsu University, Okayama, Japan
c Ina Research Inc., Nagano, Japan
d Otsuka Pharmacutical Factory, Inc., Tokushima, Japan
e Department of Pharmacology, Graduate School of Medical and Dental Sciences, Kagoshima University, Kagoshima, Japan
A R T I C L E I N F O
A B S T R A C T
Article history:
The common marmoset is a promising species for evaluating the safety of drug candidates. To further
understand the capacity for drug metabolism in marmosets, a cDNA encoding a CYP2B enzyme was
cloned from the total RNA fraction of marmoset liver by 30- and 50-RACE methods. Nucleotide and
deduced amino acid sequences showed 90.8 and 86.2% identity, respectively, with human CYP2B6. The
marmoset CYP2B6 (marCYP2B6) protein was expressed in insect cells, and its enzymatic properties were
compared with those of human (humCYP2B6) and cynomolgus monkey (cynCYP2B6) orthologs in liver
and insect cell microsomes. Enzymatic functions were examined for the oxidation of 7-ethoxy-4-
(trifluoromethyl)coumarin (7-ETC), bupropion (BUP) and efavirenz (EFV). The kinetic profiles for the
oxidation of the three substrates by liver microsomal fractions were similar between humans and
cynomolgus monkeys (biphasic for 7-ETC and monophasic for BUP and EFV), but that of marmosets was
unique (monophasic for 7-ETC and biphasic for BUP and EFV). Recombinant enzymes, humCYP2B6 and
cynCYP2B6, also yielded similar kinetic profiles for the oxidation of the three substrates, whereas
marCYP2B6 showed activity only for 7-ETC hydroxylation. In silico docking simulations suggested that
two amino acid residues, Val-114 and Leu-367, affect the activity of marCYP2B6. In fact, a marCYP2B6
mutant with substitutions V114I and L367V exhibited BUP hydroxylase activity that was 4-fold higher
than that of humCYP2B6, while its EFV 8-hydroxylase activity was only 10% that of the human enzyme.
These results indicate that the amino acids at positions 114 and 367 affect the enzymatic capacity of
marmoset CYP2B6.
Received 17 December 2012
Accepted 28 January 2013
Available online 5 February 2013
Keywords:
Marmoset CYP2B6
cDNA cloning
Bupropion
Efavirenz
Docking simulation
ß 2013 Elsevier Inc. All rights reserved.
using experimental animals. The animal data obtained can be
extrapolated into humans. Therefore, the choice of experimental
animals is a key point to obtaining a reliable estimation of the
1. Introduction
In the drug development process, it is important to understand
the toxicities of a candidate as well as its pharmacological effects in
the early stages. Though the availability of various kinds of
recombinant human drug-metabolizing enzymes makes it possible
to predict a fairly accurate metabolic profile for a drug candidate, a
total safety evaluation of the candidate should be performed in vivo
possible toxicity of drug candidates. From this view point,
monkeys are thought to be appropriate animal models of humans.
The common marmoset is a promising experimental animal
species for safety evaluation and metabolism of drug candidates
because of its small body size, and easy of handling and breeding,
clear advantages over cynomolgus monkeys and rhesus monkeys
which are too big to handle and have poor fertility. However,
cumulative data on drug-metabolizing enzymes are rather scarce
for common marmosets.
Cytochrome P450 (CYP) is a key enzyme in the oxidation of a
number of exogenous and endogenous compounds including drugs
[1,2]. Major human drug-metabolizing type CYP enzymes are
CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6,
CYP2E1 and CYP3A4, accounting for more than 90% of oxidative
Abbreviations: CYP, cytochrome P450; marCYP2B6, marmoset CYP2B6; hum-
CYP2B6, human CYP2B6; cynCYP2B6, cynomolgus monkey CYP2B6; 7-ETC, 7-
ethoxy-4-(trifluoromethyl)coumarin; BUP, bupropion; EFV, efavirenz; RACE, rapid
amplification of cDNA ends; ss-cDNA, single strand cDNA; fp2, NADPH-cytochrome
P450 reductase.
* Corresponding author at: 1-1-1 Tsushima-naka, Kita-ku, Okayama 700-8530,
Japan. Tel.: +81 86 251 7942; fax: +81 86 251 7942.
E-mail address: shizuo@pharm.okayama-u.ac.jp (S. Narimatsu).
0006-2952/$ – see front matter ß 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.bcp.2013.01.024

K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
1183
drug metabolism [3,4]. In cynomolgus monkeys, many kinds of
drug-metabolizing type CYP enzymes are expressed in various
organs or tissues, and a total of 23 CYP cDNAs in the CYP1 to 4
families have been registered in GenBank [5]. In marmosets, cDNA
nucleotide sequences and deduced amino acid sequences for eight
drug-metabolizing type CYP enzymes [CYP1A2 (accession number
D86475), CYP2C8 (AB242600), CYP2D19 (D29822), CYP2D30
(AY082602), CYP2E1 (D86477), CYP3A4 (D31921), CYP3A5
(EF589801) and CYP3A90 (EF589800)] have been registered in
GenBank to date. Previous studies have characterized the
enzymatic functions of marmoset CYP1A2 [6,7], CYP2C8 [8],
CYP2D19 [9–11] and CYP2D30 [10] expressed in various hetero-
logus expression systems, but no reliable experimental data have
been published on the enzymatic functions of recombinant
marmoset CYP2E1 or CYP3A enzymes so far.
(Tokyo, Japan); Bac-to-Bac Baculovirus Expression System, Cellfec-
tion II, High Five cells, 50-RACE system for rapid amplification of
cDNA ends, version 2.0 from Invitrogen (Carlsbad, CA): pGEM-T
Easy Vector system from Promega (Madison, WI); Quantum Prep
Plasmid Miniprep kit from BIO-RAD (Hercules, CA); cytochrome c
(type V) from Sigma–Aldrich (St. Louis, MO); aprotinin solution
and leupeptin hemisulfate monohydrate and b-NADPH from Wako
Pure Ind. (Osaka, Japan); and Big Dye Terminator v. 3.1 Cycle
Sequencing kit from Applied Biosystems (Foster City, CA). Primers
used in PCR were synthesized by Sigma–Aldrich Japan (Tokyo,
Japan). Male adult common marmoset livers were supplied by
Professor A. Miyata, Kagoshima University (Kagoshima, Japan). The
study was approved by the Institutional Animal Care and Use
Committee of Kagoshima University. Other reagents used were of
the highest quality commercially available.
In the present study, we focused on CYP2B enzymes in
marmoset livers, because only very little information has been
reported on the marmoset CYP2B family. Among the nine major
hepatic human drug-metabolizing type CYP enzymes, CYP2B6
accounts for 2–6% of total hepatic CYP content and 8–10% of drug
oxidation catalyzed by CYP enzymes [12]. Human CYP2B6
(humCYP2B6) catalyzes the oxidation of clinically prescribed
drugs such as bupuropion (BUP) [13], an antidepressant, efavirenz
(EFV) [14], an anti-HIV drug, and cyclophosphamide [15], an
anticancer drug. Furthermore, insecticides such as chlorpyrifos
2.2. Sequencing of marmoset CYP2BX cDNA
Total RNA was extracted from marmoset liver with the RNeasy
Mini kit and QIA shredder according to the manufacturer’s
directions. A portion of the total RNA was reverse-transcribed
using the RNA PCR kit (AMV) ver 3.0 to obtain single-strand cDNA.
A partial fragment of cDNA (250 ng) encoding the marmoset CYP2B
enzyme was amplified by polymerase chain reaction (PCR) from
the single-strand cDNA using the forward primer, 50-CTGCAGATG-
GATAGAAGAGGCCTAC-30 and the reverse primer, 50-GTATTTGAG-
CTTGAGCAGGAAGCCG-30, designed based on two regions in CYP2B
cDNA that are highly homologous in different species, such as
humans (accession number, AC023172), cynomolgus monkeys
(DQ074793) and rhesus monkeys (MN_001040212) in the
GenBank database. PCR by KOD-Plus DNA polymerase consisted
of initial denaturation at 94 8C for 120 s followed by 30 cycles of
denaturation at 94 8C for 15 s, annealing at 58 8C for 60 s and
extension at 68 8C for 50 s. The PCR product was introduced into
the pGEM-T Easy vector using TA cloning, and sequenced in both
forward and reverse directions.
30-RACE for marCYP2BX was performed using the 30-Full Race
Core Set according to the manufacturer’s instructions. Gene-
specific primers (GSP1 and GSP2 for 30-RACE, Table 1) were
designed based on the nucleotide sequence of a partial fragment of
marCYP2BX cDNA amplified by PCR using the primers for the
previous step. Single-strand cDNA was synthesized from marmo-
set liver total RNA using the Oligo(dT) 30 site adaptor primer (SAD,
Table 1). Amplification in the first PCR was primed using first-
strand cDNA as a template with GSP1 and the 30 site adaptor primer
(Table 1). PCR by Ex Taq DNA Polymerase consisted of initial
denaturation at 94 8C for 120 s followed by 30 cycles of
denaturation at 94 8C for 15 s, annealing at 56 8C for 50 s and
extension at 68 8C for 240 s. The PCR product was used as a
template for nested PCR, and the amplification was primed using
the GSP2 and 30 site adaptor primer under the same PCR conditions.
The PCR product was introduced into the pGEM-T Easy vector
using TA cloning and sequenced in both forward and reverse
directions.
[16] and endosulfan-a [17] are also oxidized by CYP2B6. Therefore,
humCYP2B6 is a unique enzyme that contributes to the oxidation
not only of clinically important drugs but also of agricultural
chemicals [12].
In contrast, the enzymatic properties of cynologus monkey
CYP2B6 (cynCYP2B6) have not been fully elucidated, though its
nucleotide and deduced amino aid sequences have been revealed
(GenBank accession No. DQ074793). Moreover, there is little
information about a marmoset ortholog of humCYP2B6 except for
the existence of proteins in marmoset liver microsomal fractions
revealed by immunoblot analyses [9,18]. The characterization of
marmoset CYP enzymes as orthologs of human enzymes is a great
help to further understanding the drug-metabolizing capacity of
marmosets, which promotes their usefulness as an experimental
animal in the research field of drug metabolism and toxicity
especially in drug development. If the enzymatic functions of a
marmoset enzyme are different from those of the human ortholog,
the search for the molecular mechanism(s) causing the difference
may bring about the elucidation of enzyme reaction mechanism(s).
We thus conducted the present study to clone a cDNA encoding a
novel marmoset ortholog of humCYP2B6, express its protein in
insect cells and characterize its enzymatic properties, which were
compared with those of humCYP2B6 and cynCYP2B6.
2. Materials and methods
2.1. Materials
7-Ethoxy-4-(trifluoromethyl)coumarin (7-ETC) was obtained
from Anaspec Inc. (Fremont, CA); 7-hydroxy-4-(trifluoromethyl)-
coumarin was from Sigma–Aldrich (St. Louis, MO); BUP hydro-
chloride and hydroxybupuropion were from Toronto Research
Chemicals (North York, ON, Canada); and EFV and 8-hydroxyefa-
virenz were from Santa Cruz Biotechnology (Santa Cruz, CA).
Pooled liver microsomal fractions from humans, cynomolgus
monkeys and common marmosets were purchased from BD
Biosciences (San Jose, CA). The RNeasy Mini kit, QIA shredder, and
MiniElute gel extraction kit were obtained from Qiagen (Valencia,
CA). RNA PCR kit (AMV) ver. 3.0, Ex Taq DNA Polymerase HS, DNA
ligation kit ver. 2.1, 30-Full RACE Core set and HindIII were from
TaKaRa Bio (Shiga, Japan); KOD-Plus DNA Polymerase from Toyobo
50-RACE for marCYP2BX was performed using the 50-RACE
System according to the manufacturer’s directions. Gene-specific
primers (GSP1, GSP2 and GSP3 for 50-RACE, Table 1) were designed
based the nucleotide sequence of a partial fragment of marCYP2BX
cDNA amplified by PCR using the primers for the previous step.
Single-strand cDNA was synthesized from marmoset liver total
RNA using GSP1, and the oligo(dC) tail was introduced at the 30-
end. Amplification in the first PCR was primed using oligo(dC)
tailed single-strand cDNA with the abridged anchor primer (AAP,
Table 1) and GSP2 primer. PCR by Ex Taq DNA Polymerase
consisted of initial denaturation at 94 8C for 120 s followed by 30
cycles of denaturation at 94 8C for 35 s, annealing at 55 8C for 60 s

1184
K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
Table 1
Oligonucleotide primers used for cDNA cloning of marmoset CYP2BX and human CYP2B6.
Primer
Sequence
Target
Homo-CYP2BX-FP
Homo-CYP2BX-RP
50-CTGCAGATGGATAGAAGAGGCCTAC-30
50-GTATTTGAGCATGAGCAGGAAGCCG-30
Partial fragment
30-GSP1
30-GSP2
30-SAD
50-CATGGAAAAGGAGAAATCCAACCCAC-30
50-CCACACAGTGAATTCCAACCACCAG-30
50-CTGATCTAGAGGTACCGGATCC-30
30-RACE
50-RACE
50-GSP1
50-GSP2
50-GSP3
50-AAP
50-GGTTTTTACGTATTTG-30
50-GAGAAGAGCTCGAACAGCTGGCTGAATAAAG-30
50-CTCTTTGTGTCTTGGTACTCAAAGCGTTTTCC-30
50-GGCCACGCGTCGACTAGTACGGGIIGGGIIGGGIIG-30
50-GGCCACGCGTCGACTAGTAC-30
50-AUAP
marCYP2BX-FP
marCYP2BX-RP
50-GGATCCAAAATGGAGCTCACCGTCTTC-30
Full-length
Full-length
50-GGATCCTCAGCGGGCAGGAAGC-30
humCYP2B6-FP
humCYP2B6-RP
50-GGATCCAAAAAAATGGAACTCAGCGTCCTC-30
50-GGATCCGCGGGCAGGATC-30
Underlined and italic letters indicate the restriction enzyme (BamH I) site and Kozak sequence, respectively.
and extension at 72 8C for 210 s with the final extension at 72 8C for
360 s. The PCR product was used as a template for nested PCR, and
the amplification was primed using the abridged universal
amplification primer (AUAP, Table 1) and GSP3 primer. PCR by
Ex Taq HS DNA Polymerase consisted of initial denaturation at
94 8C for 90 s followed by 30 cycles of denaturation at 98 8C for
35 s, annealing at 57 8C for 60 s and extension at 72 8C for 210 s.
The PCR product was introduced into the pGEM-T Easy vector
using TA cloning and sequenced in both forward and reverse
directions.
A recombinant baculovirus carrying human, cynomolgus
monkey or marmoset CYP2B6 cDNA was prepared employing
the Bac-to-Bac Expression System according to the manufac-
turer’s instructions. For protein expression, High Five cells
(4 ꢀ 108 cells/200 mL) were infected with recombinant baculo-
viruses at a multiplicity of infection of 2.0, followed by cultivation
for 72 h with fortifying 50 mM d-aminolevulinic acid and 50 mM
ferric citrate 24 h after infection. The cells were then harvested,
and suspended in 50 mM Tris–HCl buffer (pH 7.4) containing
250 mM sucrose, 0.1 mM EDTA, 0.1 mM DTT, 0.1% aprotinin and
1
m
g/mL leupeptin, and stored at ꢁ80 8C prior to use. For the
2.3. Cloning of human fp2, human and cynomolgus monkey CYP2B6
and marCYP2BX cDNAs
preparation of microsomal fractions, the cells were quickly
thawed in a water bath at 37 8C, sonicated 40 times with 5-s
bursts in ice, and centrifuged at 600 ꢀ g for 10 min. The
supernatant was successively centrifuged at 9000 ꢀ g for
20 min and then at 105,000 ꢀ g for 60 min at 4 8C. The resultant
pellet was resuspended in 50 mM potassium phosphate buffer
(pH 7.4) containing 1 mM DTT, 1 mM EDTA and 20% glycerol, and
stored at ꢁ80 8C until used. Protein concentrations were deter-
mined by the method of Lowry et al. [21] using bovine serum
albumin as a standard.
The full-length cDNA encoding human fp2 was amplified by PCR
from human fp2 cDNA cloned into the pGEM-3Z vector [19] as a
template using the forward primer 50-CCCAAGCTTGG-
GAAAAAAATGGGAGACTCC-30 and the reverse primer 50-
GGGGTACCCCCTAGCTCCACACGTCC-30 (underlined letters are the
HindIII sites). The PCR product was directly introduced into the
pGEM-T Easy vector using TA cloning and sequenced in both
forward and reverse directions to confirm that there were no PCR
errors. The DNA fragment corresponding to human fp2 was
inserted into the cloning site for the P10 promoter of the pFastBac
Dual vector.
2.4. Assay for recombinant CYP2B6 and fp2 contents
Microsomal fractions from insect cells expressing each primate
CYP2B6 was diluted to 2.5–3.0 mg protein/mL with 100 mM
potassium phosphate buffer (pH 7.4) containing 20% (v/v) glycerol,
and functional CYP content was measured spectrophotometrically
by the method of Omura and Sato [22] using 91 mMꢁ1 cmꢁ1 as the
absorption coefficient. Using the same microsomal fractions, the
reduction of cytochrome c by fp2 in the presence of NADPH was
measured according to the method of Thor et al. [23] using
21 mMꢁ1 cmꢁ1 as the absorption coefficient.
humCYP2B6 cDNA was amplified by PCR from humCYP2B6
cDNA cloned into the p-ENTR/D-TOPO vector [20] as a template
employing humCYP2B6-1 FP and RP as forward and reverse
primers, respectively in Table 1. The same primers were also
employed in the amplification of cynCYP2B6 cDNA. Single-strand
cDNAs were synthesized from cynomolgus monkey and marmoset
liver total RNA using the oligo(dT) adaptor primer. Full-length
CYP2B6 cDNAs of the cynomolgus monkey and marmoset were
amplified by PCR using Ex Taq DNA Polymerase from the respective
liver single-strand cDNAs as a template using forward and reverse
primers (Table 1). The PCR products of humCYP2B6, cynCYP2B6
and marCYP2BX cDNAs were directly introduced into the pGEM-T
Easy vector using TA cloning and sequenced in both forward and
reverse directions to confirm that there were no PCR errors. The
cDNA fragments corresponding to humCYP2B6, cynCYP2B6 and
marCYP2BX cDNAs were cut from the pGEM-T Easy plasmids with
BamH I and subsequently subcloned into a pFastBac Dual vector
containing human fp2 cDNA digested with BamH I. The expression
plasmids were sequenced to verify the correct orientation with
respect to the promoter (polyhedrin promoter) for the pFastBac
Dual vector.
Pooled microsomal fractions from human, cynomolgus monkey
and marmoset livers (each 60
mg protein) and insect cell
microsomal fractions (each 3 g) were subjected to 10% sodium
m
dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
[24]. Proteins on the gel were electroblotted to a polyvinylidene
difluoride membrane (Millipore, Billerica, MA), and analyzed by
Western blotting according to a published method [25] using
sheep anti-human CYP2B6 peptide (Ile265-Lys276) antibody
(BIOMOL, Hamburg, Germany) as
a primary antibody and
horseradish peroxidase-conjugated rabbit anti-sheep IgG (H + L)
as a secondary antibody. Proteins were visualized with chemi-
fluorescence using the ELC Plus Western blotting Detection System
(GE Healthcare Japan, Tokyo, Japan).

K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
1185
2.5. Enzymatic assay
San Jose, CA). All values are represented as the mean ꢂ S.D. for three
separate experiments. Statistical comparisons were made with
7-ETC O-deethylation activity was determined by the HPLC
Student’s t-test, and a difference was considered statistically
method of Jinno et al. [20] with a slight modification. The ice-cold
significant when the p value was <0.05.
reaction mixture (500
dehydrogenase, 5 mM MgCl2, 0.2 mM EDTA, 0.5 mM NADP+,
microsomes [HLM (200 g protein), CLM (60 g protein, MLM
(20 g protein) and recombinant enzymes (each 200 g protein)]
and 7-ETC (0.1–50 M) in 50 mM potassium phosphate buffer (pH
ml) contained 10 mM G-6-P, 2 IU of G-6-P
2.8. Docking simulation
m
m
m
m
The homology model of marCYP2B6 was constructed by Swiss-
Model (http://swissmodel.expasy. org/) using the crystallographic
data for human CYP2B6 (3IBD) obtained from the Protein Data Bank
(http://www.rcsb.org/pdb/) and the primary amino acid sequence of
marCYP2B6 determined in this study. Hydrogen atoms were added
for the homology model using the Biopolymer module of the Insight
II software package (Molecular Simulations Inc., San Diego, CA). Six
peptides of marCYP2B6 (Arg-97 to Asn-116, Met-198 to Ser-209,
Phe-234 to Leu-239, Gly-289 to Ser-303, Ile-359 to His-368, and Thr-
469 to Ser-478) were extracted as substrate recognition sites (SRSs)
[26]. The active-site cavities of human CYP2B6 and marCYP2B6 were
m
7.4). After preincubation at 37 8C for 1 min, the reaction was
initiated by adding the NADPH-generating system, followed by
incubation at 37 8C for 10 min. The reaction was stopped by adding
ice-cold 10% phosphoric acid, and the reaction tube was
centrifuged at 12,000 ꢀ g for 10 min at 4 8C. The supernatant
was passed through a PTFE membrane (pore size 0.45
mm,
Milipore), and a part of the filtrate was subjected to HPLC under
the conditions described below.
BUP hydroxylation activity was measured by the HPLC method
of Hesse et al. [13] with a slight modification. The ice-cold reaction
˚
made manually above thesixth ligand of hemeat a resolutionof1.0 A
mixture (500
7-ETC oxidation assay except the enzyme sources (liver micro-
somes HLM, CLM and MLM 200 g protein each; recombinant
enzymes 200 g protein each) and substrate (BUP 5–2000 M).
After preincubation at 37 8C for 1 min, the reaction was initiated by
adding the NADPH-generating system, followed by incubation at
37 8C for 20 min. The reaction mixture was treated as described
above for the following HPLC analysis.
m
l) contained the same ingredients described for the
using a homemade CG program working on a Windows PC as
described elsewhere [8]. The substitution of amino acid residues was
performed using Swiss-pdb Viewer 3.7 (GlaxoSmithKline, http://
www.genebee.msu.su/spdbv/mainpage.htm). The amino acid resi-
dues at the active sites of hum- and marCYP2B6s were drawn using
Accelrys ViewerLite 5.0 (Accelrys Inc., San Diego, CA). The making of
the three-dimensional structures of the substrates (7-ETC, BUP and
EFV) and their energy minimization were carried out with
Winmostar V3.806c (http://winmostar.com/). Docking simulation
of the substrates within the active-site cavities of CYP2B6s were
performed using Autodock 4.0 and Autodock Tools (ADT) v1.5.4
(Scripps Research Institute, http://autodock.scripps.edu/) and
Insight II.
m
m
m
EFV 8-hydroxylaionactivitywas determined by the HPLC method
of Ward et al. [14] with a slight modification. The ice-cold reaction
mixture (200
ETC oxidation assay except the enzyme sources [liver microsomes
HLM and MLM (500 g protein each), CLM (200 g protein);
recombinant enzymes, humCYP2B6 and cynCYP2B6 (500 g protein
each), marCYP2B6 (1 mg protein)] and substrate (EFV 0.5–200 M).
ml)contained the sameingredients described for the 7-
m
m
m
m
3. Results
After preincubation at 37 8C for 2 min, the reaction was initiated by
adding the NADPH-generating system, followed by incubation at
37 8C for 20 min. The reaction mixture was treated as described
above for the following HPLC analysis.
3.1. cDNA cloning
cDNA encoding humCYP2B6 contained in the pENTR/D-TOPO
plasmid was subcloned into the pGEM-T Easy vector, while cDNA
encoding cynCYP2B6 was cloned into the pGEM-T Easy vector from
total RNA of the liver of a male adult cynomolgus monkey. The
nucleotide sequences were confirmed via sequencing to be the
same as those of humCYP2B6 (GenBank accession number
AC023172.1) and cynCYP2B6 (DQ074793). Because there was no
information on a marmoset ortholog of human CYP2B6, we
employed 30- and 50-RACE methods using primer sets for highly
conserved nucleotide sequences of human and cynomolgus
monkey CYP2B6 cDNAs in combination with nested PCR. As a
result, a single strand cDNA encoding the marmoset CYP2B enzyme
(tentatively named CYP2BX) was obtained. The cDNA nucleotide
sequence and its deduced amino acid sequence are shown in Fig. 1.
Fig. 2 shows a comparison of the amino acid sequences of CYP2BX,
cynCYP2B6 and humCYP2B6. Table 2 summarizes the homology
between the nucleotide and amino acid sequences of the three
proteins. marCYP2BX and humCYP2B6 exhibited 90.8 and 86.2%
identity in nucleotide and amino acid sequences, respectively.
These sequences were registered with the Cytochrome P450
Nomenclature Committee via Dr. David Nelson of The University of
Tennessee Health Science Center and DNA Data Base of Japan
(DDBJ). The Committee named the protein as marmoset CYP2B6,
and DDBJ gave the accession number AB574423.
2.6. HPLC analysis
HPLC conditions were: a Shimadzu LC-9A liquid chromatograph
equipped with a SIL-6B autoinjector, SCL-6B system controller,
CTO-10AS column oven and SPD10A UV detector or a Hitachi
L2480 fluorescence detector with Smart Chrom data processor
(KYA Technologies, Tokyo, Japan). For 7-ETC O-deethylation:
column, Inertsil ODS-SP (5
m
m, 4.6 mm i.d. ꢀ 150 mm, GL
Sciences, Tokyo, Japan); mobile phase, 20 mM NaClO4 (pH 2.5)/
CH2CN (48:52, by volume); flow rate, 1.0 mL/min; detection,
fluorescence excitation/emission wavelengths 342/495 nm; injec-
tion volume, 20
mL; column temperature, 40 8C. For BUP hydrox-
ylation: column, Inertsil ODS-SP (5
mobile phase, 50 mM KH2PO4 (pH 3.0)/CH3CN (80:20, by volume);
flow rate, 1.0 mL/min; detection, UV 214 nm; injection volume,
m
m, 4.6 mm i.d. ꢀ 150 mm);
20
mL; column temperature, 40 8C. For EFV 8-hydroxylation:
column, Inertsil ODS-SP (5 mm, 4.6 mm i.d. ꢀ 150 mm); mobile
phase, 50 mM KH2PO4 (pH 2.4)/CH3CN (54:46, by volume); flow
rate, 1.0 mL/min; detection, UV 214 nm; injection volume, 50
column temperature, 40 8C.
mL;
2.7. Data analysis
Kinetic parameters (apparent Km and Vmax values) for the
oxidation of 7-ETC, BUP and EFV were calculated by Michaelis–
Menten and Eadie–Hofstee plots using Prism v. 5.02 (Graph Pad
Software, San Diego, CA) and SigmaPlot v.8.02 (Systat Software,
3.2. Expression of CYP2B proteins in insect cells
The cDNAs encoding human, cynomolgus monkey and marmo-
set CYP2B6s were inserted into the expression vector pFastBac

1186
K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
1
CAGGGCAGTCAGACCAGGACCATGGAGCTCACCGTCTTCCTCTTCTTTGCACTCCTCACA
1
M
E
L
T
V
F
L
F
F
A
L
L
T
13
120
33
61
GGCCTTTTGCTTCTCCTGCTCCGGCGTCACCCTAAGGCCCATGGCCGCCTCCCACCAGGC
14
G
L
L
L
L
L
L
R
R
H
P
K
A
H
G
R
L
P
P
G
121
CCCCGCCCTCTGCCCCTTTTGGGGAACCTTCTGCAGCTGGAGAGAAGAGGCCTACTCAAA
180
53
34
P
R
P
L
P
L
L
G
N
L
L
Q
L
E
R
R
G
L
L
K
181
54
TCCTTTCTGAAATTCCGAGAGAAATACGGGGATGTCTTCACGGTACACCTGGGACCGAGG
240
73
S
F
L
K
F
R
E
K
Y
G
D
V
F
T
V
H
L
G
P
R
241
74
CCCGTGGTCATGCTGTGTGGAGTGGATGCCATACGGGAGGCCCTGGTGGATCAGGCTGAG
300
93
P
V
V
M
L
C
G
V
D
A
I
R
E
A
L
V
D
Q
A
E
301
94
GTCTTCTCTGGCCGGGGCAAAATTGCCATCGTTGACCCAGTCTTCCAGGGCTACGGCGTG
360
113
420
133
480
153
540
173
600
193
660
213
720
233
780
253
840
273
900
293
960
313
1020
333
1080
353
1140
373
1200
393
1260
413
1320
433
1380
453
1440
473
1500
491
1560
V
F
S
G
R
G
K
I
A
I
V
D
P
V
F
Q
G
Y
G
V
361
114
421
134
481
154
541
174
601
194
661
214
721
234
781
254
841
274
901
294
961
314
GTCTTTGCCAATGGGGACCGCTGGAAGGCGCTGCGGCGATTCTCTCTGACCACCATGAGG
V
F
A
N
G
D
R
W
K
A
L
R
R
F
S
L
T
T
M
R
GACTTTGGGATGGGAAAGCGGAGTGTGGAGGAGCGGATCAAGGAGGAGGCTCAGTGTCTG
D
F
G
M
G
K
R
S
V
E
E
R
I
K
E
E
A
Q
C
L
GTGCAGGAGCTTCGGAAATACAAGGGAGCCCTCGTGGACCCCACCTTCTTCTTCCATTCC
V
Q
E
L
R
K
Y
K
G
A
L
V
D
P
T
F
F
F
H
S
ATCACCGCCAACATCATCTGCTCCATCGTCTTTGGAAAACGCTTTGAGTACCAAGACAAA
I
T
A
N
I
I
C
S
I
V
F
G
K
R
F
E
Y
Q
D
K
GAGTTCCTGAAGCTGCTGCGCTTGTTCTACCAGTCTTTTTCACTTGTCAGCTCTTTATTC
E
F
L
K
L
L
R
L
F
Y
Q
S
F
S
L
V
S
S
L
F
AGCCAGCTGTTCGAGCTCTTCTCTGATTTCCTAAAATACTTTCCTGGGGCGCACAGGCAA
S
Q
L
F
E
L
F
S
D
F
L
K
Y
F
P
G
A
H
R
Q
ATACGTAAAAACCTGCAGGAAATCGGTGCTTTCATTGGCCACAGTGTGGAGAAGCACCGT
I
R
K
N
L
Q
E
I
G
A
F
I
G
H
S
V
E
K
H
R
GAAGCCCTGGACCCCAGCTCCCCCCAGGACCTCATCGACACCTACCTGCTCCACATGGAA
E
A
L
D
P
S
S
P
Q
D
L
I
D
T
Y
L
L
H
M
E
AAGGAGAAATCCAACCCACACAGTGAATTCAACCACCAGAATCTCATCTTCAACACGCTC
K
E
K
S
N
P
H
S
E
F
N
H
Q
N
L
I
F
N
T
L
TCGCTCTTCTTTGCCGGCACCGAGACCACCAGTACCACTCTCCGCTACGGCTTCCTGCTC
S
L
F
F
A
G
T
E
T
T
S
T
T
L
R
Y
G
F
L
L
ATGCTCAAATACCCTCATGTGGCAGAGAGAGTCTACAAGGAGATTGAACAGGTGATTGGC
M
L
K
Y
P
H
V
A
E
R
V
Y
K
E
I
E
Q
V
I
G
1021 CCACATCGCCCTCCAGCATTAGATGACCGAGCCAAAATGCCATACACAGATGCAGTCATC
334
1081 CATGAGATCCAGAGAATTGCTGACCTTCTCCCCATGGGTTTGCCCCACATTGTCACACAA
354
1141 CACACGAGCTTCCGAGGGTACACCATCCCCAAGGGCACGGAAGTATTTCCCATCCTGAGC
374
1201 ACTGCTCTCAACGACCCACACTACTTTGAAAAACCAGACACCTTCAATCCTGACCACTTT
394
1261 CTGGATACCAATGGGGCACTGAAGAAGAATGAAGCTTTTATCCCCTTCTCCTTAGGGAAG
414
1321 CGCATCTGTCTTGGTGAAGGCATCGCCCGCACCGAATTGTTCCTCTTCTTCACCACCATC
434
1381 CTCCAAAACTTCTCCGTGGCCAGCCCCGTGGCTCCTGAAGACATCGACCTGACACCCCAG
454
1441 GAGAATGGTGTGGGCAAACTACCCCCAGCATACCAGATCCGCTTCCTGCCCCGCTGAAGG
474
1501 GGCTAAGAAGAGGGGGTCAAGAGATTCCGGGTCATCCAGTTGTCCCCACCTCTGTAGACA
P
H
R
P
P
A
L
D
D
R
A
K
M
P
Y
T
D
A
V
I
H
E
I
Q
R
I
A
D
L
L
P
M
G
L
P
H
I
V
T
Q
H
T
S
F
R
G
Y
T
I
P
K
G
T
E
V
F
P
I
L
S
T
A
L
N
D
P
H
Y
F
E
K
P
D
T
F
N
P
D
H
F
L
D
T
N
G
A
L
K
K
N
E
A
F
I
P
F
S
L
G
K
R
I
C
L
G
E
G
I
A
R
T
E
L
F
L
F
F
T
T
I
L
Q
N
F
S
V
A
S
P
V
A
P
E
D
I
D
L
T
P
Q
E
N
G
V
G
K
L
P
P
A
Y
Q
I
R
F
L
P
R
*
1561 ATGACAATTCCTCCAAAACTTTTGACTGCCCCCTGCAACTTTCTGTTT 1608
Fig. 1. Nucleotide and deduced amino acid sequences of marCYP2B6. The numbers of the deduced amino acid and nucleotide sequences are shown on the right and left sides,
respectively.
Dual containing human fp2 cDNA, and each CYP protein was co-
expressed with fp2 in insect High Five cells. Microsomal fractions
were obtained and assayed for CYP proteins and functional CYP
contents by Western blotting (Fig. 3) and reduced carbon
monoxide-difference spectroscopy, respectively. As shown in
Fig. 3, single protein bands cross-reacting with anti-humCYP2B6
antibody were observed in the lanes containing pooled human
(HLM), cynomolgus monkey (CLM) and marmoset liver micro-
somes (MLM) in the Western blotting. Furthermore, protein bands
were observed also in the lanes containing microsomal fractions of
insect cells expressing each of the three recombinant enzymes. The
functional CYP contents were 27.7 ꢂ 2.6, 65.2 ꢂ 23.6 and
50.7 ꢂ 11.3 pmol/mg protein for hum-, cyn- and marCYP2B6-
expressing insect cell microsomal fractions, respectively. The amount
of fp2 varied from 10.4 ꢂ 3.5 to 26.7 ꢂ 2.4 pmol/mg protein, and the
ratio of CYP to fp2 contained in microsomal fractions was calculated to
be 0.2–1.0.
3.3. 7-ETC O-deethylation by liver microsomes and recombinant
enzymes
When 7-ETC (0.1–50
mM) was used as a substrate, all of the
three liver microsomal fractions examined exhibited oxidative
activity as shown in Michaelis–Menten plots (Fig. 4A). Eadie–
Hofstee plots yielded biphasic kinetics for human and cynomolgus
monkey liver microsomes, but gave monophasic kinetics for
marmoset liver microsomes (Fig. 4B). Table 3 (upper column)
summarizes kinetic parameters. Kinetic profiles in this index are
similar between humans and cynomolgus monkeys, that is, their
apparent Km values for low- and high-Km phases were close, while

K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
1187
Fig. 2. Alignment of three amino acid sequences of CYP2B6s. humCYP2B6 is from humans (GenBank Accession No. AC023172) and cynCYP2B6 from cynomolgus monkeys (No.
DQ074793). (*) Amino acid residues conserved among the three CYP2B6s.
the Vmax and CLint values were significantly higher in cynomolgus
monkeys than in humans. The Km value for marmoset liver
Kinetic parameters are listed in Table 3. The apparent Km value was
smallest for the low-Km phase of marmoset liver microsomes
microsomes (0.32
humans and cynomolgus monkeys (0.12–0.18
m
M) was close to that for the low-Km phase of
M).
(2.8
(116
m
M), followed by cynomolgus monkey (13
mM) and human
m
mM) liver microsomes. The Vmax value was highest for
Fig. 5A shows Michaelis–Menten plots for 7-ETC O-deethylation
where insect cell microsomes instead of liver microsomes were
used as enzyme sources. Eadie–Hofstee plots gave monophase
kinetics for all of the three recombinant enzymes (Fig. 5B). The
cynomolgus monkey (1540 pmol/min/mg protein) followed by
human and marmoset liver microsomes.
When the enzyme source was changed from liver microsomes
to recombinant enzymes, human and cynomolgus monkey CYP2B6
enzymes exhibited BUP hydroxylase activity, whereas marmoset
CYP2B6 did not show any detectable activity under the conditions
used (Fig. 5C and D). The Apparent Km value of humCYP2B6
apparent Km values were marCYP2B6 (1.1
mM) < cynCYP2B6
(9.7 M) 2 humCYP2B6 (13 M), whereas Vmax values were
m
m
marCYP2B6 (0.5 pmol/min/pmol CYP) ꢃ cynCYP2B6 (0.6 pmol/
min/pmol CYP) < humCYP2B6 (1.5 pmol/min/pmol CYP), resulting
in the highest CLint values for marCYP2B6, followed by humCYP2B6
and cynCYP2B6 (Table 4, upper column).
(116
mM) was 3-fold higher than that (39 mM) of cynomolgus
monkey, while the Vmax value of the human enzyme was about half
that of the cynomolgus monkey enzyme (Table 4, middle column).
3.4. BUP hydroxylation by liver microsomes and recombinant
enzymes
3.5. EFV 8-hydroxylation by liver microsomes and recombinant
enzymes
In BUP hydroxylation at a substrate concentration ranging from
When EFV (0.5 to 200 mM) was employed as a substrate, human
5 to 2000
m
M by liver microsomal fractions, different kinetic
and cynomolgous monkey liver microsomes gave monophasic
profiles were observed (Fig. 4C). That is, human and cynomolgus
monkey liver microsomes showed monophasic kinetics, whereas
marmoset liver microsomes yielded biphasic kinetics in Michae-
lis–Menten plots (Fig. 4C) and Eadie–Hofstee plots (Fig. 4D).
Table 2
Identities of nucleotide and deduced amino acid sequences of human and
cynomolgus monkey CYP2B6 and marmoset CYP2BX.
Fig. 3. Western blot analysis of microsomes prepared from human, cynomolgus
humCYP2B6
cynCYP2B6
94.4
marCYP2BX
monkey and marmoset livers, and from insect cells expressing each CYP2B6 protein.
humCYP2B6
cynCYP2B6
marCYP2BX
90.8
90.9
Each liver (60
mg) and insect cell (3 mg) microsomal protein sample was applied to a
91.4
86.2
well, and resolved by SDS-PAGE. The proteins were then transblotted to a PVDF
membrane and immunochemically probed with polyclonal antibodies raised
against human CYP2B6. HLM, CLM and MLM represent human, cynomolgus
monkey and marmoset liver microsomes, respectively.
87.4
Upper-right values, percent identities of the nucleotide sequences; lower left
values, percent identities of the deduced amino acid sequences.

1188
K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
Fig. 4. Typical Michelis–Menten plots (left panels) and Eadie–Hofstee plots (right panels) for the oxidation of 7-ETC (upper panles), BUP (middle panels) and EFV (lower
panels) by microsomes from human, cynomolgus monkey and marmoset livers. Closed circles, human; closed triangles, cynomolgus monkey; closed squares, marmoset.
kinetics, whereas EFV oxidation by marmoset liver microsomes
was found to be biphasic (Fig. 4E and F). The low-Km enzyme,
which was mainly responsible for EFV 8-hydroxylation in
marmoset liver microsomes, had the lowest apparent Km value
in Michaelis–Menten (Fig. 5E) and Eadie–Hofstee plots (Fig. 5F).
Apparent Km values were comparable between human and
cynomolgus monkey enzymes, whereas the Vmax values were
11-fold higher for humCYP2B6 than for cynCYP2B6, resulting in a
13-fold higher CLint value for the human enzyme than for monkey
enzyme (Table 4, the lower column).
(0.5
m
M) and the cynomolgus monkey (5.5
mM) and human
enzymes (11.5
m
M) had 11- and 24-fold, respectively, higher Km
values (Table 3, the lower column). The Vmax values ranked as
follows: marmoset high-Km enzyme > cynomolgus monkey enzy-
me > human enzyme > marmoset low-Km enzyme, resulting in
the CLint values: cynomolgus monkey > marmoset 3 human.
Similar to the case of BUP hydroxylation, marCYP2B6 did not
show any detectable EFV 8-hydroxylase activity, while the activity
of humCYP2B6 was much higher than that of cynCYP2B6 as shown
3.6. In silico search for structural factors affecting enzymatic function
of marCYP2B6
The three-dimensional structure of marCYP2B6 was automati-
cally constructed via the web site of Swiss-model (http://
swissmodel.expasy.org/) employing the crystal structure of

K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
1189
Table 3
˚
˚
˚
˚
L367V-marCYP2B6, i.e., 4.6 A ! 2.7 A for S-BUP and 9.2 A ! 4.0 A
for S-EFV. On the basis of the results obtained from the in silico
simulation, we then prepared cDNAs encoding marCYP2B6
mutants in which Val-114 and/or Leu-367 were substituted with
human type Ile-114 and/or Val-367 by site-directed mutagenesis,
and expressed their proteins in the insect cell system.
Kinetic parameters for 7-ETC O-deethylation, BUP hydroxylation and EFV 8-
hydroxylation by human, cynomolgus monkey and marmoset liver microsomes.
a
b
Km
(
m
M)
Vmax
CLint
(1) 7-ETC O-deethylation
HLM
CLM
MLM
Low-Km phase
High-Km phase
Low Km phasec
High-Km phased
0.115 ꢂ 0.072
5.92 ꢂ 1.30
24.8 ꢂ 8.2
94.8 ꢂ 5.8
245 ꢂ 3.5**
982 ꢂ 5.7**
424 ꢂ 14
0.215 ꢂ 0.14
0.016 ꢂ 0.003
1.38 ꢂ 0.13**
0.169 ꢂ 0.003**
1.36 ꢂ 0.24**
0.178 ꢂ 0.035
5.80 ꢂ 0.99
3.7. The oxidation of BUP and EFV by insect cell microsomal fractions
expressing marCYP2B6 mutants
0.323 ꢂ 0.063
(2) BUP hydroxylation
Fig. 8 shows Michaelis–Menten plots for BUP hydroxylation (a
left panel) and EFV 8-hydroxylation (a right panel) by marCYP2B6
mutants. Eadie–Hofstee plots for both oxidation reactions gave
monophasic profiles (Figures not shown). Of particular interest, all
of the single and double amino acid-substituted marCYP2B6
mutants had BUP hydroxylase activity, which was undetectable for
the wild-type enzyme. In EFV 8-hydroxylation, however, the single
amino acid substituted mutant, V114I-marCYP2B6 did not show
detectable activity under the conditions used. Table 5 lists the
kinetic parameters calculated. For BUP hydroxylation, apparent Km
HLM
CML
116 ꢂ 13
13.2 ꢂ 2.8**
2.79 ꢂ 0.29
160 ꢂ 27
434 ꢂ 16
1540 ꢂ 40**
26.6 ꢂ 8.6
318 ꢂ 15
3.80 ꢂ 0.56
121 ꢂ 21**
MLM
Low-Km phase
High-Km phase
9.50 ꢂ 3.00
0.503 ꢂ 0.060
(3) EFV 8-hydroxylation
HLM
CML
11.5 ꢂ 0.07
5.48 ꢂ 0.76**
0.485 ꢂ 0.089
621 ꢂ 164
96.0 ꢂ 2.1
160 ꢂ 8.4**
4.21 ꢂ 0.93
355 ꢂ 67
8.36 ꢂ 0.31
29.9 ꢂ 5.0**
9.17 ꢂ 3.4
MLM
Low-Km phase
High-Km phase
0.588 ꢂ 0.062
HLM, human liver microsomes; CLM, cynomolgus monkey liver microsomes; MLM,
marmoset liver microsomes.
values of marCYP2B6 mutants ranged from 68 to 99
mM, which
a
pmol/min/mg protein.
b
were similar to that of humCYP2B6 (116 M). Based on Vmax
m
nL/min/mg protein. Each value represents the mean ꢂ S.D. for three separate
experiments.
values of the mutants ranked V114I/L367V > L367V > V114I, in
which the value for the L367V-mutant was comparable to that for
the human enzyme. As a result, the CLint values of the L367V- and
V114I/L367V-mutants were the same as and 3.5-fold higher than,
respectively, that of humCYP2D6, while V114I-mutant was much
less active than the human enzyme. For EFV 8-hydroxylation, the
V114I-mutant did not exert any detectable activity under the
conditions employed, while the L367V- and V114I/L367V-mutants
were much less active than the human enzyme (Table 5).
c
Compared with HLM-low-Km phase.
d
Compared with HLM-high-Km phase.
**
Significantly different from the human values (p< 0.01).
humCYP2B6 (3IBD) as a template. Within substrate recognition
sites 1–6, we carefully checked amino acid residues, having side-
chains oriented to the active-site cavity, that are different between
marmoset and human CYP2B6 proteins. As a result, there is a
possibility that two amino acid residues at positions 114 (valine for
marmoset and isoleucine for human) and 367 (leucine for
marmoset and valine for human) near the heme moiety may
affect the shape of the active-site cavity (Fig. 6). We then prepared
marCYP2B6 mutants having human-type amino acid residue(s) at
positions 114 and/or 367 in silico by using Swiss PDB Viewer, and
performed a docking simulation study of BUP or EFV within the
active-site cavities of the marCYP2B6 wild-type and mutants, and
compared the results with those for the human CYP2B6 wild-type
protein.
Fig. 7 shows the simulated orientations of 7-ETC, S-BUP and S-
EFV above the heme moiety in the active-site cavity of the
humCYP2B6 wild-type (A, C and F), the marCYP2B6 wild-type (B, D
and G) and a mutant having double amino acid substitutions
(V114I/L367V-marCYP2B6) (E and H). The distances between the
heme iron and the oxidation site of each substrate are also shown.
The distance between the oxidation site of 7-ETC and the heme
4. Discussion
The present study focused on the properties of marmoset and
cynomolgus monkey CYP2B6s. As a first step we conducted to
clone a cDNA encoding a marmoset ortholog of human CYP2B6
employing 30- and 50-RACE methods, succeeding in the cDNA
cloning and the protein expression in the insect cells. When we
consulted with Dr. Nelson of the Cytochrome P450 Nomenclature
Committee on the naming of a novel marmoset CYP2B enzyme, he
recommended us to use the name of ‘‘CYP2B6’’. In the home page of
Dr. Nelson (http://drnelson.uthsc.edu/biblioA.html#2B), the same
nomenclature, ‘‘CYP2B6’’, has been given to seven mammalian
CYPs; human, chimpanzee, rhesus monkey, cynomolgus monkey,
marmoset (this study), cow and cat. This may be a little confusing
when the genes and enzymatic functions of several CYP2B6s are
discussed, so we used the terms of humCYP2B6, cynCYP2B6 and
marCYP2B6 to distinguish them in the present study. However, it
would be convenient for us that each CYP has its own nomencla-
ture to avoid confusion.
As a second step, we kinetically analyzed the oxidation of three
substrates, 7-ETC, BUP and EFV, by monkey liver microsomal
fractions, and compared the results with those of human liver
microsomes. 7-ETC O-deethylation was analyzed to be biphasic for
human and cynomolgus monkey liver microsomal fractions. Code
et al. [27] reported that 7-ETC O-deethylation in the human liver
was mediated by multiple CYP enzymes including CYP1A2
˚
iron was 3.5 A in the active-site cavity of wild-type humCYP2B6
˚
(Fig. 7A), while it was 2.6 A in the active-site cavity of wild-type
marCYP2B6 (Fig. 7B). It is reasonable to assume that the oxidation
by CYP that occurs between the heme iron and the oxidation site of
˚
the substrate is within 4–5 A. From this assumption, 7-ETC can be
oxidized by both human and marmoset CYP2B6s, and the distance
between the heme iron and the oxidation site was shorter in
marCYP2B6 than in humCYP2B6, which is well consistent with the
experimental results revealing the higher 7-ETC O-deethylase
capacity for the marmoset enzyme than human enzyme (Table 4).
The oxidation profiles for BUP and EFV are different. That is, the
distances between the oxidation site of both substrates and the
heme iron in the active-site cavity are larger for the marmoset
enzyme than human enzyme. Notably, the distance between the
8-position of S-EFV and the heme iron is 9.2 A for the marCYP2B6
wild-type and 3.8 A for the human enzyme. Interestingly, the
distances between the oxidation sites of S-BUP and S-EFV
have become much shorter in the active-site cavity of V114L/
(Km < 0.1
mM) and CYP2B6 (2.9 mM) as major enzymes, being
consonant with the present results. Considering that the Km values
in biphasic kinetics were similar between human and cynomolgus
monkey liver microsomal fractions (Table 3), 7-ETC O-deethylation
in cynomolgus liver microsomes might be mediated by CYP1A2 as
a low-Km enzyme and CYP2B6 as a high-Km enzyme, respectively.
In contrast, marCYP2B6 could be mainly responsible for marmoset
liver microsomal 7-ETC O-deethylation.
˚
˚

1190
K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
Fig. 5. Typical Michaelis–Menten plots (left panels) and Eadie–Hofstee plots (right panels) for the oxidation of 7-ETC (upper panels), BUP (middle panels) and EFV (lower
panels) by microsomes from insect cells expressing humCYP2B6, cynCYP2B6 or marCYP2B6. Closed circles, human; closed triangles, cynomolgus monkey; closed squares,
marmoset.
BUP hydroxylation [13,28] and EFV 8-hydroxylation [14,29]
have been demonstrated to be catalyzed mainly by CYP2B6 in
human liver microsomes. Consonant with the previous findings,
both BUP hydroxylation and EFV 8-hydroxylation gave mono-
phasic profiles in human liver microsomes in the present kinetic
study. The Km values for BUP hydroxylation were in good
agreement between the human liver microsomes and hum-
CYP2B6, but the values for EFV 8-hydroxylation were consider-
ably different between the human liver microsomes and
recombinant enzyme. It is unclear at present what causes the
difference.
Cynomolgus monkey liver microsomes also showed mono-
phasic profiles for the oxidation of these substrates. Employing LC/
MS analysis, Mutlib et al. [30] observed a similarity in the
metabolic profiles of EFV between humans and cynomolgus
monkeys. Uno et al. [31] cloned a cDNA encoding cynomolgus
monkey CYP2B6 and expressed its proteins in Escherichia coli. They
found some oxidation activities for testosterone 16a-hydroxyl-
ation and BUP hydroxylation by the recombinant enzyme, but did
not perform kinetic analysis [31]. Therefore, the present study is
the first report describing the kinetic property of cynomolgus
monkey CYP2B6. Judging from the Km values, cynCYP2B6 is

K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
1191
Table 4
Table 5
Kinetic parameters for 7-ETC O-deethylation, BUP hydroxylation and EFV 8-
hydroxylation by human, cynomolgus monkey and marmoset CYP2B6s.
Kinetic parameters for BUP hydroxylation and EFV 8-hydroxylation by marCYP2B6
and three mutants.
a
b
a
b
Km
(
m
M)
Vmax
CLint
Km
(1) BUP hydroxylation
(
mM)
Vmax
CLint
(1) 7-ETC O-deethylation
humCYP2B6
cynCYP2B6
marCYP2B6
12.7 ꢂ 3.5
1.51 ꢂ 0.13
0.55 ꢂ 0.03**
0.51 ꢂ 0.07**
126 ꢂ 30
humCYP2B6
marCYP2B6
V114I
116 ꢂ 39
N.D.
26.5 ꢂ 2.5
<0.176
251 ꢂ 91
N.D.
9.65 ꢂ 1.73
52 ꢂ 9**
1.05 ꢂ 0.31**##
526 ꢂ 151**##
98.9 ꢂ 24.9
87.2 ꢂ 4.6
67.8 ꢂ 15.0
3.17 ꢂ 0.13*
34.0 ꢂ 7.9
250 ꢂ 44
882 ꢂ 191**yy##
L367V
21.8 ꢂ 4.2
(2) BUP hydroxylation
V114I/L367V
58.5 ꢂ 12.0**yy##
humCYP2B6
cynCYP2B6
marCYP2B6
116 ꢂ 39
26.5 ꢂ 2.5
49.5 ꢂ 5.8**
N.D.
251 ꢂ 91
1270 ꢂ 91**
N.D.
38.9 ꢂ 1.8*
(2) EFV 8-hydroxylation
N.D.
humCYP2B6
marCYP2B6
V114I
1.66 ꢂ 0.20
1.83 ꢂ 0.34
<0.05
1092 ꢂ115
N.D.
N.D.
(3) EFV 8-hydroxylation
N.D.
<0.05
N.D.
humCYP2B6
cynCYP2B6
marCYP2B6
1.66 ꢂ 0.20
1.83 ꢂ 0.34
0.167 ꢂ 0.021**
N.D.
1092 ꢂ 115
85 ꢂ 14**
N.D.
L367V
4.56 ꢂ 0.84**
0.143 ꢂ 0.002**
32.9 ꢂ 8.9**
2.04 ꢂ 0.53
V114I/L367V
1.93 ꢂ 0.24##
0.218 ꢂ 0.052**
117 ꢂ 37**
N.D.
N.D., not detectable. Significantly different from humCYP2B6 (*p < 0.05, **p < 0.01),
from V114I (yyp < 0.01), or from L367 V (##p < 0.01).
N.D., not detectable. Significantly different from humCYP2B6 (*p < 0.05 and
**p < 0.01). Significantly different from cynCYP2B6 (##p < 0.01).
a
pmol/min/pmol CYP.
a
pmol/min/pmol CYP.
b
nL/min/pmol CYP. Each value represents the mean ꢂ S.D. for three separate
b
nL/min/pmol CYP. Each value represents the mean ꢂ S.D. for three separate
experiments.
experiments.
Fig. 6. Comparison of the active-site cavity of humCYP2B6 (A) and marCYP2B6 (B) and speculated amino acid residues which may affect the substrate specificity of the two
enzymes by changing the shape of the active-site cavity (C).

1192
K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
Fig. 7. Orientations of the three substrates within the active-site cavities of humCYP2B6, marCYP2B6 and the V114I/L367V-marCYP2B6 mutant. The optimized orientation for
each substrate was obtained by docking simulations using ADT with default genetic algorithm parameters except for individual GA runs, 100. Arrows in left panels, oxidation
˚
sites. The value (A) in each panel showing the orienation of each substrate indicates the distance between the oxidation site of the substrate and the heme iron.
thought to be mainly responsible for the oxidation of BUP and EFV
in cynomolgus mokey liver microsomes.
The question arises, ‘‘why does not the marmoset enzyme
catalyze the oxidation of BUP or EFV?’’ In search of the molecular
mechanism(s) preventing marCYP2B6 from catalyzing BUP or EFV
oxidation, we prepared conformations of marCYP2B6 in silico via
Swiss Model employing the crystal structure of hum CYP2B6 as a
template. We carefully compared the shapes of active-site cavities
and amino acid residues whose side-chains extrude into the active-
site cavities between marCYP2B6 and humCYP2B6, and chose two
amino acids, Val-114 and Leu-367, as possible residues that affect
the shape of the active-site cavity of the marmoset enzyme, maybe
resulting in the change to its catalytic properties.
Marmoset liver microsomes, on the other hand, clearly
exhibited an oxidative capacity for BUP and EFV, though
marCYP2B6 did not show any oxidative activity toward these
substrates under the experimental conditions used. These results
indicate that some CYP enzyme(s) other than CYP2B6 are
responsible for marmoset liver microsomal BUP and EFV
oxidation, which are catalyzed mainly by CYP2B6s in human
and cynomolgus monkey liver microsomes. On the basis of the
CLint values, the oxidation capacity of liver microsomal fractions
ranked as follows: cynomolgus monkey ꢃ marmoset > humans
for 7-ETC, cynomolgus monkey ꢄ marmoset > humans for BUP,
and cynomolgus monkey > marmoset ꢃ humans for EFV
(Table 3).
In the docking simulation employing Autodock 4.0, we used
three substrates, 7-ETC, S-BUP and S-EFV, while we used BUP and
EFV racemates as substrates in the in vitro enzyme assay. The
reason why we employed S-BUP and S-EFV in the in silico study is
Fig. 8. Typical Michaelis–Menten plots for the oxidation of BUP (left panels) and EFV (right panels) by microsomes from insect cells expressing marCYP2B6 and its three
mutants having single and double amino acid substitutions. Open circles, V114I-mutant; open squares, L367V-mutant; open triangles, V114I/L367V-mutant.

K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
1193
that a BUP racemate is used clinically and the hydroxylation of S-
BUP was 1.5- and 3-fold higher than that of R-BUP by human liver
microsomes and humCYP2B6, respectively [32] and that S-EFV but
not R-EFV is clinically used as an anti-HIV drug [33].
In summary, the present study focused on the enzymatic
properies of marmoset CYP2B6 and compared them with those of
human and cynomolgous monkey orthologs using 7-ETC, BUP and
EFV as substrates. The kinetic profiles for the oxidation of the three
substrates by liver microsomal fractions were similar between
humans and cynomolgus monkeys (biphasic for 7-ETC and
monophasic for BUP and EFV), but that of marmosets was unique
(monophasic for 7-ETC and biphasic for BUP and EFV). Recombi-
nant enzymes, humCYP2B6 and cynCYP2B6, also yielded similar
kinetic profiles for the oxidation of the three substrates, whereas
marCYP2B6 showed activity only for 7-TEC hydroxylation. In silico
docking simulations suggested that two amino acid residues, Val-
114 and Leu-367, affect the activity of marCYP2B6. In fact, a
marCYP2B6 mutant with the substitutions of V114I and L367V
exhibited BUP hydroxylase activity that was 4-fold higher than
that of humCYP2B6, while its EFV 8-hydroxylase activity was only
10% that of the human enzyme. These results indicate that the
amino acids at positions 114 and 367 affect the enzymatic capacity
of marmoset CYP2B6.
As shown in Fig. 7, the oxidation sites of BUP and EFV are rather
˚
˚
far from the heme iron (4.6 A and 9.2 A for BUP and EFV,
respectively) in the active-site cavity of the marCYP2B6 wild-type,
˚
˚
whereas the distances became closer (2.7 A and 4.0 A for BUP and
EFV, respectively) in the active-site cavity of the V114I/L367V-
mutant. The distance between the oxidation site of BUP and the
˚
heme iron (2.7 A) of the mutant is shorter than that in the
˚
humCYP2B6 (3.2 A), which may result in the greater BUP oxidative
capacity of the mutant compared to humCYP2B6 (Table 5). Only
the single substitution of Val-114 with isoleucine or of Leu-367
with valine rendered marCYP2B6 the capacity to mediate BUP
hydroxylation, though the effect of the latter, in which the CLint
value was comparable to that of humCYP2B6, is much greater than
the former, in which the CLint value was only 14% that of the human
enzyme. Furthermore, the double subsitutions of amino acids at
positions 114 and 367 increased the CLint value 3.5-fold compared
to those of humCYP2B6 and the L367V-marCYP2B6 mutant. These
results may mean that the substitution of Leu-367 with valine is
mainly responsible for the increased BUP oxidizing capacity of
marCYP2B6, but some concerted effect is produced by the double
amino acid substitution, though the effect of the single substitution
of Val-114 with isoleucine is rather small.
Conflict of interest
The authors declare that there are no conflicts of interest.
References
For EFV 8-hydroxylation, on the other hand, the substitution of
amino acids at positions 114 and 367 did not have such a
remarkable effect. That is, the V114I-mutant did not exhibit any
detectable activity, and the CLint value of the L367V-mutant was
only 3% that of humCYP2B6. Even the V114I/L367V-mutant yielded
only 10% of the capacity of the human enzyme, though the
speculated distance between the oxidation site of EFV and the
[1] Nebert DW, Russell DW. Clinical importance of cytochrome P450. Lancet
2002;360:1155–62.
[2] Guengerich FP. Cytochrome P450 and chemical toxicology. Chem Res Toxicol
2008;21:70–83.
[3] Evans WE, Relling MV. Pharmacogenomics: tralslating functional genomics
into rational therapeutics. Science 1999;286:487–91.
[4] Rendic S. Summary of information on human CYP enzymes: human P450
metabolism data. Drug Metab Rev 2002;34:83–448.
˚
heme iron is 4.0 A in the active-site cavity (Fig. 7). It seems that the
[5] Uno Y, Iwasaki K, Yamazaki H, Nelson DR. Macaque cytochrome P450: no-
menclature, transcript, gene, genomic structure, and function. Drug Metab Rev
2011;43:346–61.
˚
distance ‘‘4 A’’ is close to the limit for the oxidation of the substrate
to occur in the active-site cavity of marCYP2B6. There is another
possibility that some other amino acid residue(s) as well as Val-
114 and Leu-367 in the active-site cavity may be involved in the
oxidation of EFV by the marmoset enzyme. It should be noted that
there has been no reports describing on human CYP2B6 alleles
involving amino acid substitutions at positions 114 or 367 [34].
In the present study, various recombinant CYP2B6 proteins were
co-expressed with fp2 in insect cells where the ratios of CYP to fp2
varied from 0.2 to 1.0 as described in Section 3. Therefore, some
activities of the recombinant CYP2B6s could be changed under the
conditions where fp2 is overexpressed to a similar extent in each
microsomal preparation. Moreover, we employed Swiss Model and
Insight II for making the homology models and Autodock Tools for
docking simulation, and found out Val-114 and Leu-367 to be key
residues for elevation in BUP and EFV oxidation activities of mar-
CYP2B6 in the present study. However, the system that we used has
some limitations. For example, though S-EFV takes similar special
orientations in the active-site cavities between hum-CYP2B6 and
V114I/L367V-marCYP2B6 in this model (Fig. 7), the activity of the
humanenzyme was much higher than that of the marmoset mutant.
Additionally, affinity between substrates and enzyme proteins
cannot be predicted by this system.
For further elucidation of molecular mechanism(s) involved in
the interesting phenomenon found in this study, highly efficient
tools are necessary to simulate possible interactions of substrates
with amino acid residues in the active-site cavity and to predict the
affinity of substrates for CYP enzymes, resulting in more accurate
prediction of enzyme activity and selectivity. Cumulated informa-
tion including the present results increases the usefulness of
marmosets in the research of drug metabolism, pharmacokinetics
and toxicokinetics for the development of new drug candidates.
[6] Sakuma T, Igarashi T, Hieda M, Ohgiya S, Isogai M, Ninomiya S, et al. Marmoset
CYP1A2: primary structure and constitutive expression in livers. Carcinogen-
esis 1997;18:1985–91.
[7] Narimatsu S, Oda M, Hichiya H, Isobe T, Asaoka K, Hanioka N, et al. Molecular
cloning and functional analysis of cytochrome P450 1A2 from Japanese
monkey liver: comparison with marmoset cytochrome P450 1A2. Chem Biol
Interact 2005;152:1–12.
[8] Narimatsu S, Torigoe F, Tsuneto Y, Saito K, Hanioka N, Masuda K, et al. Cloning
of a cDNA encoding a novel marmoset CYP2C enzyme, expression in yeast cells
and characterization of its enzymatic functions. Biochem Pharmacol
2006;72:1738–48.
[9] Igarashi T, Sakuma T, Isogai M, Nagata R, Kamataki T. Marmoset liver cyto-
chrome P450s: study for expression and molecular cloning of their cDNAs.
Arch Biochem Biophys 1997;339:85–91.
[10] Hichiya H, Kuramoto S, Yamamoto S, Shinoda S, Hanioka N, Narimatsu S, et al.
Cloning and functional expression of a novel marmoset cytochrome P450 2D
enzyme. CYP2D30: comparison with the known marmoset CYP2D19. Biochem
Pharmacol 2004;68:165–75.
[11] Narimatsu S, Nakata T, Shimizudani T, Nagaoka K, Nakura H, Masuda K, et al.
Regio- and stereoselective oxidation of propranolol enantiomers by human
CYP2D6, cynomolgus monkey CYP2D17 and marmoset CYP2D19. Chemico-
Biol Interact 2011;189:146–52.
[12] Wang H, Tompkins LM. CYP2B6: new insights into a historically overlooked
cytochrome P450 isozyme. Curr Drug Metab 2008;9:598–610.
[13] Hesse LM, Venkatakrishnan K, Court MH, von Moltke LL, Duan SX, Shader RI,
et al. CYP2B6 mediates the in vitro hydroxylation of bupuropion: potentioal
drug interactions with other antidepressants. Drug Metab Dispos
2000;28:1176–83.
[14] Ward BA, Gorski JC, Jones DR, Hall SD, Flockhart DA, Desta Z. The cytochrome
P450 2B6 (CYP2B6) is the main catalyst of efavirenz primary and secondary
metabolism: implication for HIV/AIDS therapy and utility of efavirenz as a
substrate marker of CYP2B6 catalytic activity.
2003;306:287–300.
J Pharmacol Exp Ther
[15] Roy P, Yu LJ, Crepsi C, Waxman DJ. Development of a substrate-activity based
approach to identify the major human liver P-450 catalists of cyclophospha-
mide and ifosphamide activation based on cDNA-expressed activities and liver
microsomal P-450 profiles. Drug Metab Dispos 1999;27:655–66.
[16] Tang J, Cao Y, Rose RL, Brimfield AA, Dai D, Goldstein JA, et al. Metabolism of
chlorpyrifos by human cytochrome P450 isoforms and human, mouse, and rat
liver microsomes. Drug Metab Dispos 2001;29:1201–4.

1194
K. Mayumi et al. / Biochemical Pharmacology 85 (2013) 1182–1194
[17] Casabar RC, Wallace AD, Hodgson E, Rose RL. Metabolism of endosulfan-alpha
by human liver microsomes and its utiliyu as a simultaneous in vitro probe for
CYP2B6 and CYP3A4. Drug Metab Dispos 2006;34:1779–85.
[26] Gotoh O. Substrate recognition sites in cytochrome P450 family 2 (CYP2)
proteins inferred from comparative analyses of amino acid and coding nucle-
otide sequences. J Biol Chem 1992;267:83–90.
[18] Schulz TG, Thiel R, Neubert D, Brassil P, Schulz-Utermoehl T, Boobis AR, et al.
Assessment of P450 induction in the marmoset monkey using targeted anti-
peptide antibodies. Biochim Biophys Acta 2001;1546:143–55.
[19] Hanioka N, Yamamoto M, Iwabu H, Jinno H, Tanaka-Kagawa T, Naito S, et al.
Functional characterization of human and cynomolgus money cytochrome
P450 2E1 enzymes. Life Sci 2007;81:1436–45.
[27] Code EL, Crespi CL, Penman BW, Gonzalez FJ, Chang TH, Waxman DJ. Human
cytochrome P4502B6 Interindividual hepatic expression, substrate specificity,
and role in procarcinogen activation. Drug Metab Dispos 1997;25:985–93.
[28] Faucette SR, Hawke RL, Shord SS, Lecluyse EL, Lindley CM. Evaluation of the
contribution of cytochrome P450 3A4 to human liver microsmal bupropion
hydroxylation. Drug Metab Dispos 2001;29:1123–9.
[20] Jinno H, Tanaka-Kagawa T, Ohno A, Makino Y, Matsushita E, Hanioka N, et al.
Functional characterization of cytochrome P450 2B6 allelic variants. Drug
Metab Dispos 2003;31:398–403.
[29] Saussele T, Ward B, Blievernicht J, Li L, Klein K, Flockhart DA, et al. Impact of
CYP2B6 polymorphism on hepatic efavirenz metabolism in vitro. Pharmaco-
genomics 2007;8:547–58.
[21] Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the
Folin phenol reagent. J Biol Chem 1951;193:265–75.
[30] Mutlib AE, Chen H, Nemeth GA, Markwalder JA, Seitz SP, Gan LS, et al. Drug
Metab Dispos 1999;27:1319–33.
[22] Omura T, Sato R. The carbon monoxide-binding pigment of liver microsomes. I.
Evidence for its hemoprotein nature. J Biol Chem 1964;239:2370–8.
[23] Thor H, Smith M, Hartzell P, Bellomo G, Jewell SA, Orrenius S. J Biol Chem 1982
1982;257:12419–24.
[24] Laemmli UK. Cleavage of structural proteins during the assembly of the head
of bacteriophage T4. Nature 1970;227:680–5.
[31] Uno Y, Matsuno K, Nakamura C, Utoh M, Yamazaki H. Identification and
characterization of CYP2B6 cDNA in cynomolgus macaques (Macaca facicu-
laris). J Vet Med Sci 2009;71:1653–6.
[32] Coles R, Kharasch ED. Stereoselective metabolism of bupuropion by cyto-
chrome P4502B6 (CYP2B6) and human liver microsomes. Pharm Res
2008;25:1405–11.
[25] Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from
polyacrylamide gels to nitrocellulose sheets: procedure and some applica-
tions. Proc Natl Acad Sci USA 1979;76:4350–4.
[33] http://www.healthmash.com/health/efavirenz.html.
[34] http://www.cypalleles.ki.se/cyp2b6.htm.