Functional characterization of allelic variants of polymorphic human cytochrome P450 2A6 (CYP2A6*5, *7, *8, *18, *19, and *35)

Article abstract of DOI:10.1248/bpb.35.394

Cytochrome P450 2A6 (CYP2A6) catalyzes important metabolic reactions of many xenobiotic compounds, including coumarin, nicotine, cotinine, and clinical drugs. Genetic polymorphisms of CYP2A6 can influence its metabolic activities. This study analyzed the functional activities of six CYP2A6 allelic variants (CYP2A6*5,*7,*8,*18,*19, and *35) containing nonsynonymous single-nucleotide polymorphisms. Recombinant variant enzymes of CYP2A6*7,*8,*18,*19, and *35 were successfully expressed in Escherichia coli and purified. However, a P450 holoenzyme spectrum was not detected for the CYP2A6*5 allelic variant (G479V). Structural analysis shows that the G479V mutation may alter the interaction between the A helix and the F-G helices. Enzyme kinetic analyses indicated that the effects of mutations in CYP2A6 allelic variants on drug metabolism are dependent on the substrates. In the case of coumarin 7-hydroxylation, CYP2A6*8 and *35 displayed increased K_m values whereas CYP2A6*18 and *19 showed decreased k_cat values, which resulted in lower catalytic efficiencies (k_cat/K_m). In the case of nicotine 5-oxidation, the CYP2A6*19 variant exhibited an increased K_m value, whereas CYP2A6*18 and *35 showed much greater decreases in k_catvalues. These results suggest that individuals carrying these allelic variants are likely to have different metabolisms for different CYP2A6 substrates. Functional characterization of these allelic variants of CYP2A6 can help determine the importance of CYP2A6 polymorphisms in the metabolism of many clinical drugs.

Full text of DOI:10.1248/bpb.35.394

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
—
394
Regular Article
Biol. Pharm. Bull. 35(3) 394 399 (2012)
Vol. 35, No. 3
Functional Characterization of Allelic Variants of Polymorphic Human
Cytochrome P450 2A6 (CYP2A6*5, *7, *8, *18, *19, and *35)
Songhee Han,a Seunghye Choi,a Young-Jin Chun,b Chul-Ho Yun,c Chang Hoon Lee,d
,a
Hee Jung Shin,e Han Sung Na,e Myeon Woo Chung,e and Donghak Kim*
a Department of Biological Sciences and Center for Biotechnology Research in UBITA (CBRU), Konkuk University;
b
c
Seoul 143–701, Korea: College of Pharmacy, Chung-Ang University; Seoul 156–756, Korea: School of Biological
d
Sciences and Technology, Chonnam National University; Gwangju 500–757, Korea: College of Pharmacy, Dongguk
University; Seoul 100–715, Korea: and National Institute of Food and Drug Safety Evaluation, Korea Food & Drug
e
Administration; Chungcheongbuk-do 363–951, Korea.
Received October 13, 2011; accepted December 13, 2011; published online December 22, 2011
Cytochrome P450 2A6 (CYP2A6) catalyzes important metabolic reactions of many xenobiotic com-
pounds, including coumarin, nicotine, cotinine, and clinical drugs. Genetic polymorphisms of CYP2A6 can
influence its metabolic activities. This study analyzed the functional activities of six CYP2A6 allelic vari-
ants (CYP2A6*5, *7, *8, *18, *19, and *35) containing nonsynonymous single-nucleotide polymorphisms.
Recombinant variant enzymes of CYP2A6*7, *8, *18, *19, and *35 were successfully expressed in Escherichia
coli and purified. However, a P450 holoenzyme spectrum was not detected for the CYP2A6*5 allelic vari-
ant (G479V). Structural analysis shows that the G479V mutation may alter the interaction between the A
—
helix and the F G helices. Enzyme kinetic analyses indicated that the effects of mutations in CYP2A6 al-
lelic variants on drug metabolism are dependent on the substrates. In the case of coumarin 7-hydroxylation,
CYP2A6*8 and *35 displayed increased Km values whereas CYP2A6*18 and *19 showed decreased kcat values,
which resulted in lower catalytic efficiencies (kcat/Km). In the case of nicotine 5-oxidation, the CYP2A6*19
variant exhibited an increased Km value, whereas CYP2A6*18 and *35 showed much greater decreases in kcat
values. These results suggest that individuals carrying these allelic variants are likely to have different me-
tabolisms for different CYP2A6 substrates. Functional characterization of these allelic variants of CYP2A6
can help determine the importance of CYP2A6 polymorphisms in the metabolism of many clinical drugs.
Key words P450; CYP 2A6; allelic variant; polymorphism; coumarin; nicotine
The cytochrome P450 (CYP, P450) family is composed of CYP2A6. The substrate preferences of these mutant enzymes
major phase I enzymes responsible for metabolizing xenobi- are different from those of the wild-type enzymes, and there-
otic chemicals including clinical drugs.1) CYP2A6 was first fore, the mutant enzymes can produce dramatic effects in
determined to catalyze coumarin 7-hydroxylation in the hu- CYP2A6 metabolism.
man liver.2) CYP2A6 also metabolizes many clinical drugs,
In this study, we analyzed the functional activities of six
including valproic acid, losigamone, chlormethiazole, letro- allelic variants of CYP2A6 (CYP2A6*5, *7, *8, *18, *19, and
zole, fadrozole, methoxyflurane, and tegafur.3) This enzyme *35) containing nonsynonymous SNPs. They contain substi-
is known to catalyze the metabolism of tobacco-specific com- tutions of the following amino acids: CYP2A6*5 (G479V);
pounds such as nicotine, cotinine, 4-(methylnitrosamino)-1-(3- CYP2A6*7 (I471T); CYP2A6*8 (R485L); *18 (Y392F; *19
pyridyl)-1-butanone, and N-nitrosonornicotine, and its role in (Y392F; I471T); and *35 (N438Y) (http://www.cypalleles.
the bioactivation of procarcinogens has also been intensively ki.se/). These variants are frequently found in Asian popula-
—
8)
investigated.4
tions. The allelic frequencies of CYP2A6*5, *7, *8, *18, and
—
Genetic variations in metabolic enzymes can cause dra- *35 in Japanese and Koreans were 0.5%, 9.8%, 1.4 2.2%,
—
13 19)
—
matic differences in the response to specific drugs. Some 0.5%, and 0.5 0.8%, respectively.
In view of the impor-
drug-metabolizing P450 enzymes are very polymorphic, and tance of CYP2A6 in the metabolism of many clinical drugs
the study of their polymorphisms is of particular interest be- and several tobacco-specific compounds, we determined the
cause it can help develop a rational means to optimize drug functional changes in these allelic variants.
therapy and ensure maximum efficacy with minimal adverse
effects. Previous studies reported the genetic polymorphisms
of CYP2A6 and proposed their relevance to cancer risk due
MATERIALS AND METHODS
—
12)
to variations in nicotine and N-nitrosoamine metabolism.9
Chemicals and Enzymes Coumarin, 7-hydroxycoumarin,
Interestingly, the ethnic distribution of CYP2A6 polymor- nicotine, imidazole, 3-[(3-cholamidopropyl)dimethylammonio]-
phisms is slightly different from that of other P450 enzymes. 1-propanesulfonate (CHAPS), dilauroyl-L-phosphatidylcholine
A very low frequency of poor metabolizers (PMs) has been (DLPC), NADP+, and NADPH were purchased from Sigma
reported in Caucasians and Africans, but the occurrence of Chemical (St. Louis, MO, U.S.A.) or Aldrich Chemical Co.
PMs is much more common in Asians.3) To date, at least 36 (Milwaukee, WI, U.S.A.). Ni2+-Nitrilotriacetate agarose
allelic variants of CYP2A6 have been identified, 26 of which was purchased from Qiagen (Valencia, CA, U.S.A.). Other
are located in the coding region of the CYP2A6 gene (http:// chemicals were of the highest commercially available grade.
www.cypalleles.ki.se/). Nonsynonymous single nucleotide Escherichia coli DH5α cells were purchased from Invitrogen
polymorphisms (SNPs) can alter the protein sequences of (Carlsbad, CA, U.S.A.). Rat NADPH-P450 reductase was het-
*To whom correspondence should be addressed. e-mail: donghak@konkuk.ac.kr
© 2012 The Pharmaceutical Society of Japan

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
March 2012
395
erologously expressed in E. coli (TOPP3 strain) and purified were determined using the P450/NADPH-P450 reductase/
as described elsewhere.20) phospholipid reconstituted systems.24) The reaction mixture
Construction of CYP 2A6 Variants We used a pCW included purified CYP2A6 proteins 50pmol, rat NADPH-P450
plasmid containing the coding sequence of CYP2A6 to reductase 100pmol, and DLPC 45μM in 0.50mL of potas-
construct expression vectors for the CYP2A6 variants. The sium phosphate buffer 100mM (pH 7.4), along with a speci-
general approach was described previously.21) Briefly, site- fied amount of the substrates. An NADPH-generating system
°
directed mutagenesis was performed using the QuikChange was used to start reactions after 3-min preincubation at 37 C.
°
mutagenesis kit (Stratagene, La Jolla, CA, U.S.A.) with the Incubation was generally for 10min at 37 C and terminated
following primers: 5′-CCCAAACACGTGGTCTTTGCC-3′, with 0.05mL of HCl 2M in the case of the coumarin oxidation
5′-GGCAAAGACCACGTGTTTGGG-3′
5′-CCTAAGGACACTGACGTGTC-3′,
(CYP2A6*5); reaction. CH2Cl2 was added to extract the oxidized products,
5′-GACACGTCAG- followed by vortex mixing and centrifugation at 2000rpm
5′-GCCACGATCCCA- for 10min. The organic (lower) layer was transferred to a
TGTCCTTAGG-3′
CTAAACTACACC-3′,
(CYP2A6*7);
5′-GGTGTAGTTTAGTGGGAT- clean test tube, and 1mL of sodium borate buffer 0.1M (pH
CGTGGC-3′
GCTGG-3′,
2A6*18);
(CYP2A6*8); 5′-CCGAAGTGTTCCCTAT- 9.6) was added. The production of 7-hydroxycoumarin in a
5′-CCAGCATAGGGAACACTTCGG-3′(CYP- microtiter plate was measured a fluorescence plate reader
5′-GGAAAGCGGTACTGTTTCGGAG-3′, (350nm/453nm). For nicotine oxidation reaction, the enzyme
and
5′-CTCCGAAACAGTACCGCTTTCC-3′ (CYP2A6*35). The reactions were terminated by vortexing after the addition
CYP2A6*19 variant was constructed using the primers for of 0.05mL of HClO4 and then placing the reaction mixture
CYP2A6*7 and *18. The open reading frames for CYP2A6 on ice for 1min. The reaction mixtures were centrifuged
°
variants included 6× His at the C-terminals. The constructed at 13000rpm for 30min at 4 C. To measure oxidized nico-
variant clones were verified by nucleotide sequencing analysis. tine, the reaction products were analyzed with HPLC using
Expression and Purification of CYP 2A6 Variants a reversed-phase ODS column (5μm, 150×4.6mm, Alltech
Expression and purification of CYP2A6 variants were carried Associates, Inc., Deerfield, IL, U.S.A.) equipped with a
out following the previously described methods with some YL9110 HPLC system (YoungLin, Korea). Isocratic flow
modifications.22) The E. coli DH5α strains transformed with was used with a mobile phase of potassium phosphate buffer
pCW vectors were inoculated into Luria–Bertani (LB) me- 20mM containing heptane sulfonate 3.3mM, 0.5% trieth-
dium containing ampicillin 50μg/mL and incubated overnight ylamine, and 10% acetonitrile (pH 5.07) at a flow rate of
°
at 37 C. LB cultures were then seeded into 500mL of Terrific 0.9mL/min, while monitoring A254. The oxidized nicotine
broth (TB) expression medium containing ampicillin 50μg/ formation was quantified by comparing HPLC peaks using
°
mL. The expression cultures were incubated at 37 C with an authentic standard. Kinetic parameters were estimated
shaking at 250rpm. When the OD600 of the cultures reached from the fitted curves using a computer program (GraphPad
0.5, supplements (5-aminolevulinic acid 0.5mM, isopropyl β-D- Prism 4) designed for nonlinear regression analysis. Statistical
thiogalactoside 1.0mM, thiamine 1.0mM, and trace elements) analyses of the kinetic parameters were performed using the
were added, and the expression cultures were incubated at two-tailed Student t-test.
°
28 C with shaking at 200rpm for an additional 24h.
NADPH Oxidation Assays The NADPH oxidation
Bacterial inner membrane fractions containing mutant rates were determined using phospholipid reconstituted
CYP2A6 were isolated and prepared from TB expression cul- systems containing CYP2A6 and NADPH-P450 reductase.
tures following the method described previously.21) Purification Reconstituted enzymes were preincubated for 5min at 37 C in
°
of CYP2A6 enzymes using a Ni2+-nitrilotriacetate column the presence or absence of coumarin (100μM). The reactions
was also performed by using a previously described method.23) were initiated by the addition of 10μL of NADPH 10mM, and
Briefly, the prepared membrane fractions were solubilized the reduction of A340 was monitored. The rates were calculated
overnight at 4 C in potassium phosphate buffer 100mM (pH using ε340=6.22mM−1 cm−1.
7.4) containing 20% glycerol, EDTA 1mM, β-mercaptoethanol
°
Δ
10mM, and 1.5% CHAPS (w/v). The solubilized fractions were
then loaded onto a Ni2+-nitrilotriacetate column and the pro-
teins were eluted with a buffer containing imidazole 300mM.
RESULTS
Expression and Purification of Recombinant CYP2A6
The eluted fractions containing CYP2A6 variant proteins were Variant Enzymes All six CYP2A6 variants were success-
°
subsequently dialyzed at 4 C against potassium phosphate fully constructed in the pCW expression vectors, and recom-
buffer 100mM (pH 7.7) containing 20% glycerol. binant CYP2A6 variant proteins were expressed in E. coli.
Substrate Binding Assays Purified wild-type and variant Recombinant CYP2A6 expression levels in the whole-cell cul-
CYP2A6 enzymes were diluted to 1μM in potassium phos- tures were determined spectroscopically. The expression levels
phate buffer 100mM (pH 7.4) and then divided between two of CYP2A6*1, *7, *8, *18, *19, and *35 ranged from 100 to
glass cuvettes. Spectroscopic changes from 350 to 500nm 200nmol P450 holoenzyme per liter of culture medium (Fig.
were recorded with subsequent additions of coumarin sub- 1). However, no P450 holoenzyme spectrum was observed for
strate using a CARY 100 spectrophotometer (Varian Inc., the CYP2A6*5 allelic variant (G479V) (Fig. 1). Immunoblot
Santa Clara, CA, U.S.A.). The difference in the absorbance analysis using anti-his-tag antibody in the membrane fraction
values between the maximum and minimum wavelengths was showed that the CYP2A6*5 variant protein was not detectable
plotted against the substrate concentration to estimate the sub- (data not shown). This result suggests that the G479V substitu-
strate binding affinities.
tion may interfere with the accurate expression of the enzyme.
Bacterial inner membrane fractions containing the CYP2A6
Catalytic Activity Assays Catalytic activities of CYP2A6
variants in the oxidation reactions of coumarin and nicotine variants (CYP2A6*7, *8, *18, *19, and *35) were isolated and

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
396
Vol. 35, No. 3
Fig. 1. Expression of Recombinant CYP2A6 Allelic Variant Enzymes in E. coli
CO-binding spectra of P450 enzymes in E. coli whole cells were measured, and the contents of the P450 holoenzyme were calculated.
Table 1. Coumarin Binding Affinities of Purified Wild-Type and
Variants of CYP2A6
Kd, μM
Variants
CYP2A6*1
Mutations
Coumarin
WT
I471T
3.5 0.1
4.7 0.4
8.1 0.7
6.4 0.7
9.6 0.7
7.9 0.4
CYP2A6*7
CYP2A6*8
CYP2A6*18
CYP2A6*19
CYP2A6*35
R485L
Y392F
Y392F/I471T
N438Y
prepared. Purified wild type and variant CYP2A6 proteins
were successfully obtained using the Ni2+-affinity column. All
the purified proteins showed single bands of 57kDa on SDS-
PAGE as expected from the molecular mass of the open read-
ing frame (supplementary Fig. 1).
Substrate Binding Affinities Titration of purified wild-
type and variant CYP2A6 enzymes showed a typical type I
spectral change, suggesting the loss of coordination of H2O
to the P450 heme. Binding affinities of purified wild type and
variant CYP2A6 enzymes for coumarin were determined (Fig.
2, Table 1). The binding of coumarin to wild type CYP2A6
yielded a Kd value of 3.5 0.1μM (Table 1). CYP2A6 variants
did not exhibit significantly altered binding affinities for cou-
marin, although a small increase in the Kd values was seen
(Table 1). These results suggest that the amino acid changes
in the five variants did not affect the architecture of the active
site of the enzymes.
Fig. 2. Coumarin Binding Affinities of CYP2A6 Allelic Variant
Enzymes
(A) Binding titration of CYP2A6 with coumarin. Increasing concentrations
of coumarin were added to both the sample and reference cuvettes. (B) Plots of
Enzymatic Activities of CYP2A6 Variants Catalytic
activities of the CYP2A6 variant enzymes were determined
Δ
A385
vs. the concentration of coumarin for the CYP2A6 allelic variants.
—
418nm

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
March 2012
397
Table 2. Kinetic Parameters of Coumarin 7-Hydroxylation and Nicotine 5-Hydroxylation by Purified Wild-Type and Variants of CYP2A6
Coumarin 7-hydroxylation
Nicotine 5-oxidation
Variants
Mutations
k
cat, min−1
Km, μM
kcat/Km
kcat, min−1
Km, μM
kcat/Km
CYP2A6*1
CYP2A6*7
CYP2A6*8
CYP2A6*18
CYP2A6*19
CYP2A6*35
WT
I471T
11.3 0.4
7.6 1.0
8.1 1.1
1.49 0.20
1.23 0.18
0.75 0.11**
1.15 0.10
0.99 0.14*
1.28 0.14
24.5 2.1
18.1 1.0*
21.3 1.3
145 39
157 26
142 28
119 25
248 32*
110 24
0.17 0.05
0.12 0.02
0.15 0.03
0.13 0.03
0.12 0.02
0.10 0.02
10.0 0.4*
15.5 0.8**
6.9 0.2**
8.8 0.3**
26.5 1.0**
R485L
20.6 2.8*
6.6 0.6
Y392F
15.3 1.0**
27.9 1.3
Y392F/I471T
N438Y
8.9 1.2
20.7 2.2**
11.2 0.7**
’
Results are presented as means S.D.; *p<0.05, compared with WT (CYP2A6*1); **p<0.005, compared with WT (CYP2A6*1) by Student s t-test.
Table 3. NADPH Oxidation with Purified Wild-Type and Variants of
CYP2A6
Coumarin
NADPH
Variants
Mutations
Oxidationa)
(1000μM)
CYP2A6*1
CYP2A6*7
CYP2A6*8
CYP2A6*18
CYP2A6*19
CYP2A6*35
WT
I471T
+
−
+
−
+
−
+
−
+
−
+
−
178
42
4
3
3
2
151
61
R485L
243 13
50 10
95 1
Y392F
35
0
Y392F/I471T
N438Y
112 1
26 1
108 1
25 1
a) nmol of reduced NADPH/min/nmol of purified P450 protein.
NADPH oxidation rates (Table 3). CYP2A6*18, *19, and *35
showed slow rates of NADPH oxidation (ca. 40%) in the pres-
ence of the substrate (Table 3). These results suggest that the
lower enzymatic activities of the allelic variants CYP2A6*18
and *19 in coumarin oxidation may be caused partly by a de-
crease in the utilization of electrons. However, the slow rate of
NADPH oxidation did not alter its catalytic efficiency (kcat/Km
value) in the variant CYP2A6*35 because both kcat and Km
Fig. 3. Steady-State Kinetic Analysis of Coumarin 7-Hydroxylation and
Nicotine 5-Oxidation by CYP2A6 Allelic Variant Enzymes
Each point presented is mean S.D. of triplicate assays.
by measuring the rates of coumarin 7-hydroxylation and values were increased. In addition, the faster rate of NADPH
nicotine 5-oxidation. Steady-state kinetic analysis of cou- consumption (ca. 37%) observed in the case of CYP2A6*8
marin 7-hydroxylation indicated that the catalytic efficien- may indicate the inefficient coupling of NADPH in the enzy-
cies (kcat/Km) of all the variants were lower than that of the matic reaction.
wild-type CYP2A6*1 (Fig. 3A, Table 2). CYP2A6*8 and *35
Positions of the Mutated Amino Acids in CYP2A6
showed greater increases in Km values (with some increases Variants The X-ray crystal structure of CYP2A6 in complex
in kcat values), whereas CYP2A6*18 and *19 showed lower with coumarin (PDB-entry code, 1Z10) was used to posi-
catalytic efficiencies (kcat/Km), mainly because of the decreased tion the mutated residues in the CYP2A6 enzyme (Fig. 4).25)
kcat values (Fig. 3A, Table 2). In addition, all variants exhibited Mutated residues in CYP2A6*5, *7, and *8 (G479V, I471T,
—
reduced catalytic efficiencies (kcat/Km) for nicotine 5-oxidation. R485L, respectively) are located in the β3 β4 hairpin loop
The CYP2A6*19 variant exhibited an increased Km value, region near the C-terminus of the enzyme, and the Y392 resi-
which resulted in lower catalytic efficiencies (kcat/Km), whereas due in CYP2A6*18 is located in the β2-sheet. These residues
CYP2A6*18 and *35 showed much greater decreases in the are all located outside the formal active site of P450 (Fig. 4).
kcat values than in the Km values. CYP2A6*7 and *8 showed a Among these mutations, the G479V mutation in CYP2A6*5 is
small decrease in the kcat values, with no significant change in located spatially between the C-terminus of the A helix and
—
G helices, and the mutation seems to affect both regions
their Km values (Fig. 3B, Table 2). These results suggest that
the effects of mutations in CYP2A6 allelic variants on drug of the helices (Fig. 4). The mutation in this residue may affect
metabolism are highly dependent on the substrates. the structural architecture of both sides of the A helix and
F
—
NADPH Consumption by CYP2A6 Variants When the the F G helices, thereby resulting in the overall structural
rates of NADPH consumption by the CYP2A6 variants were stability of the P450 enzyme. We also found that the N438Y
analyzed, CYP2A6*7 and wild-type CYP2A6 showed similar mutation in CYP2A6*35 is located next to the proximal Cys