Purification and characterization of coclaurine N-methyltransferase from cultured Coptis japonica cells

Article abstract of DOI:10.1016/S0031-9422(00)00481-7

S-Adenosyl-L-methionine (SAM): coclaurine N-methyltransferase (CNMT), which catalyzes the transfer of a methyl group from S-adenosyl-L-methionine to the amino group of the tetrahydrobenzylisoquinoline alkaloid coclaurine, was purified 340-fold from Coptis japonica cells in 1% yield to give an almost homogeneous protein. The purified enzyme, which occurred as a homotetramer with a native M_r of 160 kDa (gel-filtration chromatography) and a subunit M_r of 45 kDa (SDS-polyacrylamide gel electrophoresis), had an optimum pH of 7.0 and a pI of 4.2. Whereas (R)-coclaurine was the best substrate for enzyme activity, Coptis CNMT had broad substrate specificity and no stereospecificity; CNMT methylated norlaudanosoline, 6,7-dimethoxyl-1,2,3,4-tetrahydroisoquinoline and 1-methyl-6,7-dihydroxy-1,2,3,4,-tetrahydroisoquinoline. The enzyme did not require any metal ion. p-Chloromercuribenzoate and iodoacetamide did not inhibit CNMT activity, but the addition of Co²⁺, Cu²⁺ or Mn²⁺ at 5 mM severely inhibited such activity by 75, 47 and 57%, respectively. The substrate-saturation kinetics of CNMT for norreticuline and SAM were of the typical Michaelis-Menten-type with respective K_m values of 0.38 and 0.65 mM.

Full text of DOI:10.1016/S0031-9422(00)00481-7

Phytochemistry 56 (2001) 649±655
www.elsevier.com/locate/phytochem
Puri®cation and characterization of coclaurine N-methyltransferase
from cultured Coptis japonica cells
Kum-Boo Choi a, Takashi Morishige a, Fumihiko Sato a,b,
*
aDivision of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Sakyo, Kyoto 606-8502, Japan
bDivision of Integrated Life Science, Graduate School of Biostudies, Kyoto University, Sakyo, Kyoto 606-8502, Japan
Received 11 July 2000; received in revised form 26 October 2000
Abstract
S-Adenosyl-l-methionine (SAM): coclaurine N-methyltransferase (CNMT), which catalyzes the transfer of a methyl group from
S-adenosyl-l-methionine to the amino group of the tetrahydrobenzylisoquinoline alkaloid coclaurine, was puri®ed 340-fold from
Coptis japonica cells in 1% yield to give an almost homogeneous protein. The puri®ed enzyme, which occurred as a homotetramer
with a native Mr of 160 kDa (gel-®ltration chromatography) and a subunit Mr of 45 kDa (SDS-polyacrylamide gel electrophoresis),
had an optimum pH of 7.0 and a pI of 4.2. Whereas (R)-coclaurine was the best substrate for enzyme activity, Coptis CNMT had broad
substrate speci®city and no stereospeci®city; CNMT methylated norlaudanosoline, 6,7-dimethoxyl-1,2,3,4-tetrahydroisoquinoline and
1-methyl-6,7-dihydroxy-1,2,3,4,-tetrahydroisoquinoline. The enzyme did not require any metal ion. p-Chloromercuribenzoate and
iodoacetamide did not inhibit CNMT activity, but the addition of Co2+, Cu2+ or Mn2+ at 5 mM severely inhibited such activity by 75,
47 and 57%, respectively. The substrate-saturation kinetics of CNMT for norreticuline and SAM were of the typical Michaelis±Men-
ten-type with respective Km values of 0.38 and 0.65 mM. # 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Alkaloid biosynthesis; Berberine; Coptis japonica; Ranunculaceae; Isoquinoline alkaloid; S-Adenosyl-l-methionine; Coclaurine N-
methyltransferase
1. Introduction
Zenk, 1985) and a tetrahydroprotoberberine oxidase
(Yamada and Okada, 1985; Galneder et al., 1988). Only
norcoclaurine 6-O-methyltransferase (6OMT) (Rue€er
et. al., 1983; Sato et al., 1994), 30-hydroxy-N-methyl-
coclaurine 40-O-methyltransferase (40OMT) (Frenzel and
Zenk, 1990b) and (S)-scoulerine 9-O-methyltransferase
(SMT) (Sato et al., 1993) have been highly puri®ed and
characterized.
The enzymological characterization of secondary
metabolism and the biotechnological application of
these biosynthetic enzymes should prove useful for the
biotransformation of ®ne chemicals. One diculty in
purifying the enzymes in the berberine biosynthetic
pathway is that they have similar enzymological prop-
erties due to the similarity of their reaction mechanisms
and substrates (Sato et al., 1994; Morishige et al., 2000).
Recent isolation of the cDNAs of methyltransferases and
the investigation of their expression in Escherichia coli
have provided further information about OMTs (Frick
and Kutchan, 1999; Morishige et al., 2000). However,
there is little information available on coclaurine N-
methyltransferase (CNMT), a unique N-methyltransferase
Berberine, a benzylisoquinoline alkaloid obtained
from Berberis and Coptis species, is an important phar-
maceutical alkaloid with antibacterial, stomachic and
anti-in¯ammatory activity (Sato et al., 1994). The bio-
technological importance of this alkaloid and the basic
interest in elucidating the steps involved in its biosynth-
esis have led to the complete characterization of this
pathway in plant cell cultures (Sato et al., 1993, 1994;
Kutchan, 1998). Berberine is derived from l-tyrosine via
13 di€erent enzyme reactions involving an N-methyl-
transferase (NMT) (Frenzel and Zenk, 1990a; Wat et al.,
1986), three O-methyltransferases (OMTs) (Rue€er et al.,
1983; Muemmler et al., 1985; Frenzel and Zenk, 1990b;
Sato et al., 1993, 1994), a hydroxylase (Loe‚er and Zenk,
1990), a berberine bridge enzyme (Ste€ens et al., 1984), a
methylenedioxy ring-forming enzyme (Rue€er and
*Corresponding author. Tel.: +81-75-753-6381; fax: +81-75-753-
6398.
E-mail address: fumihiko@kais.kyoto-u.ac.jp (F. Sato).
0031-9422/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0031-9422(00)00481-7

650
K.-B. Choi et al. / Phytochemistry 56 (2001) 649±655
in berberine biosynthesis (Fig. 1) (Frenzel and Zenk,
1990a). CNMT catalyzes the transfer of a methyl group
from S-adenosyl-l-methionine (SAM) to coclaurine
with the formation of N-methylcoclaurine. So far, of the
N-methyltransferases involved in secondary plant meta-
bolism, only the putrescine N-methyltransferase in nico-
tine biosynthesis has been well characterized, and its
cDNA has been isolated (Hibi et al., 1994). Characteriza-
tion of the CNMT in berberine biosynthesis should help
us understand the molecular aspects of NMT in secondary
metabolism, especially alkaloid biosynthesis. We report
here the puri®cation and characterization of CNMT from
cultured Coptis japonica cells and compare it with OMTs
in berberine biosynthesis. Its broad substrate speci®city
and biotechnological applications are discussed.
activity of the ®rst CNMT fraction was only 25% that of
the second fraction, the latter was used for further pur-
i®cation. Although ®nal isoelectric focusing chromato-
graphy on a Mono P column considerably decreased the
recovery of CNMT activity, this step was essential for
removing contaminating OMTs and purifying the enzyme
to near homogeneity. Sodium dodecyl sulfate-poly-
acrylamide gel electrophoresis (SDS-PAGE) analysis
and silver staining showed that the fraction puri®ed
through the Mono P column gave only two protein bands
which had similar mobilities (silver staining) (Fig. 2).
Whereas it was not possible to separate these proteins
further, careful analysis by SDS-PAGE and enzyme assay
2. Results and discussion
2.1. Puri®cation of coclaurine N-methyltransferase
(CNMT)
An outline of the puri®cation of coclaurine N-
methyltransferase (CNMT) from cultured Coptis cells is
shown in Table 1. The enzyme was enriched 343-fold
and gave a yield of 1% after ammonium sulfate frac-
tionation, Phenyl Sepharose, Q-Sepharose, Mono Q
anion-exchange and Mono P column chromatographies
of the crude extract. Preliminary experiments showed
that chromatography of the crude enzyme extracts of
Coptis cells in 20 mM Tris±HCl bu€er on a Q-Sephar-
ose column gave two distinct active fractions. Since the
Fig. 2. SDS-PAGE analysis of puri®ed coclaurine N-methyltransfer-
ase. (a) Analysis of proteins at various stages of puri®cation. Fractions
from each puri®cation step were separated by SDS-PAGE and then
stained with Coomassie Brilliant Blue: Lane 1, crude extract; Lane 2,
30±50% (NH4)2SO4; Lane 3, Phenyl Sepharose; Lane 4, Q-Sepharose;
Lane 5, Mono Q; Lane 6, Mono P column puri®ed fraction. (b) Ana-
lysis of puri®ed coclaurine N-methyltransferase by SDS-PAGE. Frac-
tions were separated from the Mono P column, analyzed by SDS-
PAGE in 10% polyacrylamide gel and made visible by silver staining.
Fig. 1. Reaction catalyzed by coclaurine N-methyltransferase (CNMT).
SAH, S-adenosyl-l-homocysteine; SAM, S-adenosyl-l-methionine.
Table 1
Puri®cation of coclaurine N-methyltransferase from Coptis japonica cellsa
Puri®cation step
Total protein
(mg)
Total activity
(pkat)
Speci®c activity
(pkat/mg)
Puri®cation
fold
Yield
(%)
Crude extract
30±50% Ammonium sulfate
Phenyl Sepharose
Q-Sepharose
512
151
428
240
369
204
34.4
5.7
0.83
1.59
5.66
15.4
45.3
285
1.0
1.91
6.81
18.6
54.5
343
100
54
82
45
8
65.1
13.2
0.76
0.02
Mono Q
Mono P
1.2
a
Two hundred grams of cultured Coptis cells were homogenized for puri®cation of the enzyme.

K.-B. Choi et al. / Phytochemistry 56 (2001) 649±655
651
indicated that the larger 45-kDa protein corresponded to
CNMT activity (data not shown). Since the CNMT pur-
i®ed in the Mono P column showed no O-methyltransfer-
ase activity and no positive protein band was detected by
Western blotting with the anti-Coptis 40OMT and SMT
antibodies, we used this fraction as the puri®ed CNMT
fraction (data not shown).
To our knowledge, this is the ®rst study in which
coclaurine N-methyltransferase has been obtained in high
purity and characterized. Since four methyltransferases
(40OMT, 6 OMT, SMT and CNMT) with similar reac-
tions and substrates are involved in berberine biosynthesis
(Rue€er et al., 1983; Muemmler et al., 1985; Frenzel
and Zenk, 1990a,b; Sato et al., 1993, 1994; Morishige et
al., 2000), it is not easy to purify each to homogeneity.
In fact, preparation of homogenous 40OMT and 6OMT
required the heterologous expression of enzymes in E.
coli (Morishige et al., 2000). In our study, also, 40OMT
and SMT could not be separated from CNMT during
the early puri®cation steps, even though their primary
structures di€er considerably (Takeshita et al., 1995;
Morishige et al., 2000).
Superose 12 column chromatography showed that the
molecular mass of native CNMT was approximately 160
kDa. These results suggest that the native enzyme is a
homotetramer. Chromatofocusing of CNMT on a Mono
P column gave an isoelectric point (pI) of 4.2, and exam-
ination of the pH-dependence of enzyme activity showed
that the highest activity was at pH 7.0, whereas 50% of the
activity was at pH 6.0 and 9.0 (data not shown).
Examination of the e€ects of metal ions (Ca2+, Co2+
,
Cu2+, Fe2+, Li+, Mg2+, Mn2+, Ni2+, Zn2+ at 5 mM)
and potential inhibitors (EDTA, iodoacetamide, p-
chloromercuribenzoate at 1 or 5 mM) on CNMT activ-
ity indicated that the enzyme did not require a divalent
cation and the addition of EDTA did not inhibit activity,
whereas the addition of Ca2+ at 5 mM slightly increased
activity by 25%. The addition of some divalent cations
(i.e. Co2+, Cu2+ or Mn2+), however, severely inhibited
enzyme activity (by 75, 47 and 57%, respectively).
Whereas some O-methyltransferase activities are inhibited
by the addition of SH-directed inhibitors, e.g., p-chlor-
omercuribenzoate and iodoacetamide at 5 mM, these
chemicals did not inhibit CNMT activity.
2.2. CNMT characterization
2.3. Substrate speci®city of CNMT
SDS-PAGE analysis showed that the molecular mass of
the CNMT monomer is 45 kDa, whereas gel-®ltration by
The substrate speci®city of CNMT was examined for
a wide range of isoquinoline alkaloid substrates (Fig. 3).
Fig. 3. Structures of isoquinoline alkaloids and chemicals tested as CNMT substrates. The benzylisoquinoline substrates in Table 2 are coclaurine,
norreticuline, norlaudanosoline and 6-O-methylnorlaudanosoline.

652
K.-B. Choi et al. / Phytochemistry 56 (2001) 649±655
The incorporation of radioactivity from S-adenosyl-l-
[methyl-3H] methionine into alkaloid products was used
to measure enzyme activity, since the radioactivities mea-
sured corresponded to those of HPLC measurement for
several substrates (data not shown). When (S)-coclaurine
at 1 mM was chosen as the control substrate (i.e. the rela-
tive incorporation was taken as 100%) for Coptis CNMT,
the respective relative activities with (R)-coclaurine, norre-
ticuline, norlaudanosoline and 6-O-methylnorlaudanoso-
line were 122, 55, 48 and 38% (Table 2). Interestingly, 6,7-
dimethoxy-1,2,3,4-tetrahydroisoquinoline was also a
good substrate, whereas 1,2,3,4-tetrahydroisoquinoline
was not methylated. Coptis CNMT transferred the methyl
group to a fairly broad range of isoquinoline substrates;
i.e. from simple dimethoxy-tetrahydroisoquinoline to
more complicated benzylisoquinoline substrates. The
similar relative activities of (R)-coclaurine and (S)-
coclaurine suggests that Coptis CNMT was not stereo-
speci®c. However, the lack of methylation activity with
4-O-methyldopamine, tryptamine, histamine and purine
suggests that the isoquinoline ring is important for the
Coptis CNMT reaction (data not shown).
Coptis CNMT has a substrate speci®city similar to
that of Berberis NMT (Frenzel and Zenk, 1990a), in
that both enzymes show no stereospeci®city and have
relatively broad substrate speci®city for (R)- and (S)-
coclaurine, norreticuline and 6-O-methylnorlaudanoso-
line. Coptis CNMT also N-methylates norlaudanosoline,
whereas Berberis NMT does not. These results indicate
that Coptis CNMT has a much broader substrate speci®-
city; O-methylation of the isoquinoline moiety of the
benzylisoquinoline substrate has little e€ect on Coptis
CNMT activity. The fact that even 6,7-dimethoxy-
1,2,3,4-tetrahydroisoquinoline was methylated, as well as
norreticuline, suggests that the benzyl moiety of norreti-
culine is not essential for the CNMT reaction. However,
the fact that there was no CNMT activity with emetine
indicates that replacement of this benzyl moiety with a
large group makes this compound non-functional as a
substrate. The lack of methylation activity with 1,2,3,4-
tetrahydroisoquinoline suggests that hydroxylation of
the isoquinoline ring is important for the Coptis CNMT
reaction. The methylation of 1-methyl-6-7-dihydroxy-
1,2,3,4-tetrahydroisoquinoline supports this idea.
2.4. Kinetics
Although (R)-coclaurine was the best substrate, we
also determined the enzyme characteristics of CNMT
with norreticuline, the most readily available substrate,
and its similar product formation. CNMT from Coptis
cells showed Michaelis±Menten-type kinetics for norre-
ticuline and SAM (Fig. 4). Fig. 4 shows that Coptis
CNMT followed a sequential bi-substrate mechanism
(ordered bi bi or rapid equilibrium random bi bi), but
further experiments on product inhibition are needed to
determine the actual mechanism. The kinetic constants
of CNMT (Km and Vmax) were calculated from double-
reciprocal plots. The Vmax of CNMT was estimated
from Lineweaver±Burk double-reciprocal plots, and the
respective apparent Km values for norreticuline and S-
adenosyl-l-methionine (SAM) were 0.38 and 0.65 mM
(data not shown).
Although the Km values of Coptis CNMT for norre-
ticuline (0.38 mM) and SAM (0.65 mM) are somewhat
larger than those of Berberis CNMT (ca.0.2 mM for (R)-
norreticuline and 0.04 mM for SAM) and Sanguinaria
NMT (0.02 mM for (R,S)-tetrahydroberberine and 0.012
mM for SAM) (Frenzel and Zenk, 1990a, O'Keefe and
Beecher, 1994), the Coptis CNMT values are still smaller
than those of 6OMT (2.23 mM for (R,S)-norlaudanoso-
line and 3.95 mM for SAM ) (Sato et al., 1994). Enzymo-
logical characterization of the puri®ed CNMT protein
should be bene®cial in the future use of this enzyme with
other O-methyltransferases to convert various isoquino-
line derivatives into N-methylated products by enzymo-
logical biotransformation; e.g., from norlaudanosoline
to reticuline.
Table 2
Substrate speci®city of puri®ed Coptis coclaurine N-methyltransferasea
Substrate
Relative activity
R1b
R2
R3
R4
R5
(R)-Coclaurine
(S)-Coclaurine
122
100
55
48
38
0
OMe
OMe
OMe
OH
OH
OH
OH
OH
OH
H
OH
OH
OMe
OH
OH
H
H
H
H
H
H
(R,S)-Norreticuline
(R,S)-Norlaudanosoline
OH
OH
OH
(R,S)-6-O-Methylnorlaudanosoline
(R,S)-Scoulerine
OMe
6,7-Dimethoxy-1,2,3,4-tetrahydroisoquinoline
1,2,3,4-Tetrahydroisoquinoline
1-Methyl-6,7-dihydroxy-1,2,3,4-tetrahydroisoquinoline
1,2,3,4-Tetrahydro-3-isoquinolinecarboxylic acid
(+)-Emetine
39
0
10
0
0
The enzyme reaction mixture was incubated for 1 h at 30ꢀC; total volume 50 ml containing 100 mM potassium phosphate (pH 7.0), 2.5 mM
a
3
sodium ascorbate, 1 mM H-SAM and 20 ml of the puri®ed Mono-P fraction.
b
See Fig. 3 for the numbering system.

K.-B. Choi et al. / Phytochemistry 56 (2001) 649±655
653
(SAM) was purchased from NENTM Life Science, and
the other alkaloids were gifts from Mitsui Petrochemical
Industries Ltd. The other chemicals used were of the
highest purity available.
3.3. Enzyme assay and protein determination
CNMT enzyme activity was measured by HPLC or
by the incorporation of [methyl-3H] into the product
from [methyl-3H] SAM. Enzyme activities during pur-
i®cation were analyzed by HPLC. The standard assay
solution consisted of 0.1 mM norreticuline, 100 mM
potassium phosphate bu€er (pH 7.0), 2.5 mM sodium
ascorbate, 1 mM SAM and 20 ml of the enzyme solution.
Norreticuline was selected as the standard substrate
because of its availability and similarity of product for-
mation to that of coclaurine, the native substrate. The
assay mixture was incubated at 30ꢀC for 60 min, after
which the reaction was stopped by adding 50 ml of
methanol. After protein precipitation at 10,000Âg for
10 min at 4ꢀC, the amount of reticuline formed was
determined by the following HPLC separation: mobile
phase, 22% acetonitrile and 1% acetic acid; column,
LiChrosper 100RP-18 (250Â4 mm, Cica Merck); ¯ow
rate, 1.0 ml/min; detection, at 280 nm.
Transfer of the [methyl-3H] group of [methyl-3H]
SAM was investigated for the enzymological character-
ization of CNMT. The standard assay solution con-
sisted of 1 mM norreticuline, 100 mM potassium
phosphate bu€er (pH 7.0), 2.5 mM sodium ascorbate, 1
mM [methyl-3H] SAM (1.5 MBq/mmol) and 20 ml of the
puri®ed enzyme fraction; total volume 50 ml. This mix-
ture was incubated at 30ꢀC for 20 min, after which the
reaction was terminated by adding 150 ml of Na2CO3
(0.5 M) and 400 ml of isoamylalcohol. After vigorous
shaking, the mixture was centrifuged at 10,000Âg for 5
min at room temperature, and 150 ml of the organic
phase was then used to determine the radioactivity. An
assay mixture without the enzyme fraction was used as
the control. The radioactivities incorporated corre-
sponded to the activities measured by HPLC assay.
Protein concentration was determined with a Bio-Rad
protein assay kit according to the manufacturer's pro-
tocols.
Fig. 4. Double reciprocal plots of CNMT activity. (a) Plots of 1/v
versus 1/[SAM] at various constant concentrations of norreticuline. (b)
Plots of 1/v versus 1/[norreticuline] at various constant concentrations
of SAM. ^, &, ~ and * respectively represent 0.125, 0.25, 0.5 and
1.0 mM of these substrates.
3. Experimental
3.1. Plant material
Cultured Coptis japonica cells with high berberine
productivity were used (Sato and Yamada, 1984). Cells
were subcultured every 3 weeks on a rotary shaker (ca.
100 rpm) in the dark in Linsmaier-Skoog liquid medium
containing 10 mM naphthalene acetic acid and 0.01 mM 6-
benzyladenine as described elsewhere (Sato and Yamada,
1984), and 14-day-old cultured cells were used for enzyme
puri®cation. Harvested cells were immediately frozen in
liquid nitrogen and stored at À80ꢀC until use.
3.2. Chemicals
3.4. Puri®cation of coclaurine N-methyltransferase
(R,S)-1-[(3,4-Dihydroxyphenyl)methyl]-1,2,3,4-tetra-
hydro-6,7-isoquinolinediol (norlaudanosoline or tetra-
hydropapaveroline), 6,7-dimethoxy-1,2,3,4-tetrahydro-
isoquinoline hydrochloride, 1,2,3,4-tetrahydroisoquino-
line, 1-methyl-6,7-dihydroxy-1,2,3,4-tetrahydroisoquino-
line hydrobromide and 1,2,3,4-tetrahydro-3-isoquino-
line carboxylic acid hydrochloride were purchased from
Aldrich. (R)-Coclaurine and (S)-coclaurine were gifts
from Dr. N. Nagakura, Kobe Women's College of
All operations were performed at 0±4ꢀC, and all the
bu€ers contained 20 mM b-mercaptoethanol and 10%
glycerol unless stated otherwise. Typically, about 200 g
of frozen cells were homogenized in 450 ml 0.2 M Tris±
HCl bu€er (pH 7.5) in a Waring blender at top speed
for 2 min, and then sonicated for 20 min and ®ltered.
The ®ltrate was centrifuged at 10,000Âg for 50 min.
Part of the supernatant which was desalted through an
NAP-10 column (Pharmacia) was designated the crude
Pharmacy.
[Methyl-3H]
S-adenosyl-l-methionine

654
K.-B. Choi et al. / Phytochemistry 56 (2001) 649±655
extract. The enzyme which precipitated between 30 and
50% (NH4)2SO4 saturation was dissolved in 60 ml of the
extraction bu€er and then centrifuged. The protein solu-
tion obtained was desalted in a PD-10 column (Pharma-
cia) and then applied to a Phenyl Sepharose CL-4B
column (Pharmacia, 50 ml of bed volume) that had been
equilibrated with a starting bu€er (200 mM Tris±HCl (pH
7.5) containing 30% (NH4)2SO4). After the column had
been washed suciently with the same bu€er to remove
any remaining yellow alkaloids, CNMT was eluted with
120 ml of a linear gradient (30±0%) of (NH4)2SO4
solution. The CNMT fractions were then subjected to
ion-exchange chromatography in a Q-Sepharose FF
column (Pharmacia, 50 ml of bed volume) that had been
pre-equilibrated with 200 mM Tris±HCl bu€er (pH 7.5).
After the column was washed with the same bu€er,
CNMT was eluted with a linear gradient (0±0.5 M) of
NaCl solution (120 ml). Active fractions were collected,
desalted and puri®ed in an FPLC system with a Mono
Q column (HR 5/5, Pharmacia) that had been pre-equi-
librated with 20 mM Tris±HCl bu€er (pH 7.5). CNMT
was then eluted stepwise with a gradient (0±0.35 M) of
NaCl solution (40 ml). The active fractions were com-
bined and chromatofocused in a Mono P column (HR 5/
20, Pharmacia) in an FPLC system: starting bu€er; 25
mM Bis Tris±IDA (iminodiacetic acid), pH 7.1, eluting
bu€er; 10% Polybu€er 74 adjusted to pH 4 by IDA. The
pHs of the eluted fractions were immediately adjusted
with 100 mM Tris±HCl (pH 7.5). The puri®ed enzyme was
kept at À20ꢀC in 50% glycerol. Since repeated freezing
and thawing markedly reduced CNMT activity, thawed
samples were not repeatedly frozen. CNMT was active
after 1 year at À20ꢀC when frozen together with 50%
glycerol. All FPLC steps were done at room temperature.
chromatography on a calibrated Superose 12 column
(HR 10/30, Pharmacia).
3.7. Kinetic properties
Kinetic constants for CNMT activity were determined
for various amounts of norreticuline and SAM with 0.2
mg of puri®ed CNMT in a 50 ml reaction mixture.
Kinetic data were ®tted to the standard equation.
Acknowledgements
We thank Dr. N. Nagakura and Mitsui Petrochemical
Industries Ltd. for their generous gifts of alkaloids. This
research was supported in part by a Grant-in-Aid B
(08456172) from the Ministry of Education, Science,
Sports and Culture, Japan.
References
Frenzel, T., Zenk, M.H., 1990a. Puri®cation and characterization of
three isoforms of S-adenosyl-l-methionine: (R,S)-tetrahydrobenzyl-
isoquinoline-N-methyltransferase from Berberis koetineana cell
cultures. Phytochemistry 29, 3491±3497.
Frenzel, T., Zenk, M.H., 1990b. S-Adenosyl-l-methionine: 30hy-
droxylN-methyl-(S)-coclaurine-40-O-methyltransferase, a regio- and
stereoselective enzyme of the (S)-reticuline pathway. Phytochem-
istry 29, 3505±3511.
Frick, S., Kutchan, T.M., 1999. Molecular cloning and functional
expression of O-methyltransferases common to isoquinoline alkaloid
and phenylpropanoid biosynthesis. Plant Journal 17 (4), 329±339.
Galneder, E., Rue€er, M., Wanner, G., Tabata, M., Zenk, M.H.,
1988. Alternative ®nal steps in berberine biosynthesis in Coptis
japonica cell cultures. Plant Cell Reports 7, 1±4.
Hibi, N., Higashiguchi, S., Hashimoto, T., Yamada, Y., 1994. Gene
expression in tobacco low-nicotine mutants. Plant Cell 6, 723±735.
Kutchan, T.M., 1998. Molecular genetics of plant alkaloid biosynth-
esis. In: Cardell, G. (Ed.), The Alkaloids, Vol 50. Academic Press,
San Diego, pp. 257±316.
3.5. Characterization of coclaurine N-methyltransferase
Loe‚er, S., Zenk, M.H., 1990. The hydroxylation step in the biosyn-
thetic pathway leading from norcoclaurine to reticuline. Phy-
tochemistry 29, 3499±3503.
The CNMT fraction puri®ed on the Mono P column
was used for further enzymological characterization.
E€ects of pH, metal ions and chemicals on CNMT activ-
ity were determined with norreticuline as the substrate.
Divalent cations were added as their chloride salts. The
optimum pH of CNMT activity was determined with the
following bu€ers at concentrations of 100 mM: 2-(N-
morpholino)ethane sulfuric acid (MES) (pH 5.4, 6.2 and
6.6), potassium phosphate (pH 5.8, 6.2, 6.6, 7.0, 7.4 and
7.8), N-2-hydroxyethypiperazine-N-2-ethane sulfonic acid
(HEPES) (pH 6.5, 7.0, 7.5, 8.0 and 8.5), Tris±HCl (pH
7.5, 8.0, 8.5 and 9.0), glycine-NaOH (pH 9.0 and 10.0).
Morishige, T., Tsujita, T., Yamada, Y., Sato, F., 2000. Molecular
characterization of the S-adenosyl-l-methionine: 30-hydroxyl-N-
methylcoclaurine 40-O-methyltransferase involved in isoquinoline
alkaloid biosynthesis in Coptis japonica. Journal of Biological
Chemistry 275, 23398±23405.
Muemmler, S., Rue€er, M., Nagakura, N., Zenk, M.H., 1985. S-
Adenosyl-l-methionine: (S)-scoulerine 9-O-methyltransferase, a
highly stereo- and regio-speci®c enzyme in tetrahydroprotoberber-
ine biosynthesis. Plant Cell Reports 4, 36±39.
O'Keefe, B.R., Beecher, C.W.W., 1994. Isolation and characterization
of S-adenosyl-l-methionine: tetrahydroberberine-cis-N-methyl-
transferase from suspension cultures of Sanguinaria canadensis L.
Plant Physiology 105, 395±403.
3.6. Molecular weight determination
Rue€er, M., Nagakura, N., Zenk, M.H., 1983. Partial puri®cation and
properties of S-adenosylmethionine: (R,S)-norlaudanosoline-6-O-
methyltransferase from Argemone platyceras cell cultures. Planta
Medica 49, 131±137.
The subunit molecular weight of CNMT was deter-
mined by SDS-PAGE (10%) analysis with molecular
weight standards (Bio-Rad). The native molecular
weight of CNMT was determined by gel ®ltration
Rue€er, M., Zenk, M.H., 1985. Berberine synthase, the methylene-
dioxy group-forming enzyme in berberine synthesis. Tetrahedron
Letters 26, 201±202.

K.-B. Choi et al. / Phytochemistry 56 (2001) 649±655
655
Sato, F., Yamada, Y., 1984. High berberine-producing cultures of
Coptis japonica cells. Phytochemistry 23, 281±285.
Takeshita, N., Fujiwara, H., Mimura, H., Fitchen, J.H., Yamada, Y.,
Sato, F., 1995. Molecular cloning and characterization of S-ade-
nosyl-l-methionine: scoulerine-9-O-methyltransferase from cul-
tured cells of Coptis japonica. Plant and Cell Physiology 36 (1), 29±
36.
Sato, F., Takeshita, N., Fitchen, J.H., Fujiwara, H., Yamada, Y.,
1993. S-Adenosyl-L-methionine: scoulerine-9-O-methyltransferase
from cultured Coptis japonica cells. Phytochemistry 32, 659±664.
Sato, F., Tsujita, T., Katagiri, Y., Yoshida, S., Yamada, Y., 1994.
Puri®cation and characterization of S-adenosyl-l-methionine: nor-
coclaurine 6-O-methyltransferase from cultured Coptis japonica.
European Journal of Biochemistry 225, 125±131.
Yamada, Y., Okada, N., 1985. Biotransformation of tetrahydro-
berberine to berberine by enzymes prepared from cultured Coptis
japonica cells. Phytochemistry 24, 63±65.
Wat, C.K., Ste€ens, P., Zenk, M.H., 1986. Partial puri®cation and
characterization of S-adenosyl-l-methionine: norreticuline N-methyl-
Ste€ens, P., Nagakura, N., Zenk, M.H., 1984. The berberine bridge-
forming enzyme in tetrahydroprotoberberine biosynthesis. Tetra-
hedron Letters 25, 951±952.
transferase from Berberis cell suspension cultures. Zeitschrift fur
Naturforschung 41c, 126±134.
È