Structure and Biocatalytic Scope of Coclaurine N-Methyltransferase

Article abstract of DOI:10.1002/anie.201805060

Benzylisoquinoline alkaloids (BIAs) are a structurally diverse family of plant secondary metabolites, which have been exploited to develop analgesics, antibiotics, antitumor agents, and other therapeutic agents. Biosynthesis of BIAs proceeds via a common pathway from tyrosine to (S)-reticulene at which point the pathway diverges. Coclaurine N-methyltransferase (CNMT) is a key enzyme in the pathway to (S)-reticulene, installing the N-methyl substituent that is essential for the bioactivity of many BIAs. In this paper, we describe the first crystal structure of CNMT which, along with mutagenesis studies, defines the enzymes active site architecture. The specificity of CNMT was also explored with a range of natural and synthetic substrates as well as co-factor analogues. Knowledge from this study could be used to generate improved CNMT variants required to produce BIAs or synthetic derivatives.

Full text of DOI:10.1002/anie.201805060

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: Structure and Biocatalytic Scope of Coclaurine N-
Methyltransferase
Authors: Matthew Bennett, Mark Thompson, Sarah Shepherd, Mark
Dunstan, Abigail Herbert, Duncan Smith, Victoria Cronin,
Binuraj Menon, Colin Levy, and Jason Micklefield
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201805060
Angew. Chem. 10.1002/ange.201805060
Link to VoR: http://dx.doi.org/10.1002/anie.201805060
http://dx.doi.org/10.1002/ange.201805060

Angewandte Chemie International Edition
10.1002/anie.201805060
COMMUNICATION
Structure and Biocatalytic Scope of Coclaurine N-
Methyltransferase
Matthew R. Bennett, Mark L. Thompson, Sarah A. Shepherd, Mark S. Dunstan, Abigail J. Herbert,
Duncan R. M. Smith, Victoria A. Cronin, Binuraj R. K. Menon, Colin Levy and Jason Micklefield*
HO
4
Abstract: Benzylisoquinoline alkaloids (BIAs) are a structurally
diverse family of plant secondary metabolites which have been
exploited to develop analgesics, antibiotics, antitumor agents and
other therapeutic agents. Biosynthesis of BIAs proceeds via a
common pathway from tyrosine to (S)-reticulene at which point the
pathway diverges. Coclaurine N-methyltransferase (CNMT) is a key
enzyme in the pathway to (S)-reticulene, installing the N-methyl
substituent that is essential for the bioactivity of many BIAs. In this
paper, we describe the first crystal structure of CNMT which, along
with mutagenesis studies, defines the enzymes active site
architecture. The specificity of CNMT was also explored with a range
of natural and synthetic substrates as well as co-factor analogues.
Knowledge from this study could be used to generate improved
CNMT variants required to produce BIAs or synthetic derivatives.
RO
HO
6
H CO
3
3
NH2
HO
NH
N
NCS
7
1
CNMT
HO
CH3
1
2
1
'
CHO
4
'
HO
HO
5
3
R = H
HO
6OMT
bisbenzyl-
NMCH
4 R = CH3
isoquinolines
H3CO
N
morphinans
aporphines
HO
HO
CH3
protoberberines
RO
6
R = H
4
'OMT
7 R = OH
Figure 1. Early steps in the biosynthesis of benzylisoquinoline alkaloids (BIA)
are catalysed by norcoclaurine synthase (NCS), norcoclaurine 6-O-
methyltransferase (6OMT), coclaurine N-methyltransferase (CNMT), N-methyl-
coclaurine-3´-hydroxylase (NMCH), methylcoclaurine-4´-O-methyltransferase
Plants produce a wide range of bioactive benzylisoquinoline
alkaloids (BIA) including the analgesics morphine and codeine,
antitussive agent glaucine, muscle relaxant (+)-tubocurarine,
antitumor agent dauricine, as well as compounds that possess
(
4´OMT) providing (S)-reticulene 7 the main branching point.
towards improved activity or more stringent substrate specificity,
which could reduce kinetic bottle necks or side-product
[
1]
antimicrobial and other bioactivities. (Figures 1 & S1). Whilst
some medically important BIAs can be isolated in high yields
from plants, there are some BIAs of interest that are in limited
[
2,3]
formation in engineered microbial BIA producing strains.
Similarly, engineered variants of BIA enzymes, with specificity
for non-natural substrates, could open up new routes to
[
1,2]
supply. In an effort to generate more robust platforms for BIA
production, recent attention has focused on reconstituting BIA
biosynthetic pathways in microbial host strains, such as E. coli
and yeast, which are more genetically tractable and easily
[
4-7]
alternative bioactive THIQ scaffolds. In this paper we describe
the first crystal structure of CNMT, an essential enzyme in the
core pathway to BIAs, which introduces the amino methyl
functionality that is often essential for the bioactivity of most of
the medically important BIAs (Figure 1 & S1). Active site
mutagenesis and substrate screening provide further
characterization of CNMT, which can guide engineering of
CNMT for biosynthetic and synthetic applications.
[
2,3]
cultivated.
However, there are numerous challenges
associated with producing BIAs in microbial hosts, including the
low activity of some plant enzymes in microorganisms and the
toxicity of some BIA biosynthetic intermediates to microbial
[
2,3]
hosts.
Moreover, a number of the enzymes from BIA
pathways exhibit promiscuity leading to the production of
multiple side-products, which further reduce yields of desirable
Previous studies with CNMT from Coptis japonica showed
that this S-adenosyl-L-methionine (AdoMet) dependent enzyme
can N-methylate a number of natural THIQ, exhibiting highest
[
2]
products.
[
8]
In addition to their utility in microbial production platforms,
enzymes from BIA biosynthesis have been exploited as
biocatalysts to produce ‘non-natural’ products, which could
activity with coclaurine 4 and heliamine 8. To further
characterize this enzyme, the C. japonica CNMT was
overproduced in E. coli (Figure S2). Initial attempts to obtain
crystal structures in the presence of its natural
substrates (3 or 4) proved unsuccessful. However, crystals of
CNMT with S-adenosyl-L-homocysteine (AdoHcy) and N-
methylheliamine 8a bound (PDB 6GKV, Figure 2 & Table S1),
were obtained from co-crystallization experiments set up in the
presence of AdoMet and heliamine 8, with methylation of 8,
most likely occurring during the long incubation periods required
for crystallization. The CNMT structure (Figure 2 & S3) reveals a
typical class I methyltransferase α/β Rossmann fold forming the
AdoMet-binding domain, which is capped by a predominantly
alpha helical BIA substrate recognition domain. CNMT shows a
similar overall fold (RMSD of 1.1) to the recently determined
[
4]
provide additional bioactive scaffolds for drug development.
For example, NCS has been utilized to produce a wide range of
tetrahydroisoquinolines (THIQ) from simple synthetic
precursors. Similarly the berberine bridge enzyme has also
[
5]
[
6]
been used to produce non-natural scoulerine analogues.
Additionally, enzymes from BIA biosynthesis, have been
combined with enzymes from other pathways to create novel
[
7]
products. The future exploitation of BIA biosynthetic enzymes
for pathway engineering in vivo, or for enzymatic synthesis in
vitro, is largely dependent on our increasing knowledge of the
structure, mechanism and substrate scope of these important
enzymes. Such insights can guide the engineering of enzymes
[
9]
structure of the pavine N-Methyltransferase (PavNMT). CNMT
forms a dimeric interface made up of 13 hydrogen bonds and 3
salt bridges, similar to that seen in the PavNMT structure. The
main differences observed between CNMT and PavNMT
structures are located in close proximity to the active site.
Dr M. R. Bennett, Dr M. L. Thompson, Dr S. A. Shepherd, Dr M. S.
Dunstan, A. J. Herbert, Dr D. R. M. Smith, V. A. Cronin, Dr B. R. K. Menon,
Dr C. Levy, Prof. J. Micklefield - School of Chemistry, Manchester Institute
of Biotechnology, The University of Manchester, 131 Princess Street,
Manchester M1 7DN, UK. E-mail: jason.micklefield@manchester.ac.uk
Supporting information is given via a link at the end of the document.
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201805060
COMMUNICATION
A
C
D
AdoHcy
AdoHcy
AdoHcy
W329
E204
E204
E204
W329
F332
W329
E207
F332
E207
F332
E207
H208
H208
H208
8
a
(S)-4
Y328
(
R)-4
Y328
Y328
T261
T261
T261
B
E
AdoMet
AdoHcy
AdoHcy
E204
E204
OH
OH
O
O–
H
N
O
H
N
+
S
S
O
–
CH
N
3
H
H
N
H
H
3
CO
CO
W329
3
H C
W329
NH
8
R = H
H
N
+
+
N
8a R = CH3
8a
N
N
3
R
H208
F322
H208
+
F322
H
OH
OCH
H
OH
OCH
3
3
H CO
3
O
O
NH
T261
O
NH
T261
(
S)-4
(
S)-4
9
NH
H
3
CO
Figure 2. (A) The structure of CNMT (PDB 6GKV) with AdoHcy in red and the product N-methylheliamine 8a in green (ball and stick) showing key amino acid
residues involved in catalysis or substrate recognition. (B) mF -mF electron density maps (contoured at 3.5 σ) of AdoHcy and 8a in the CNMT active site.
o
c
Structure of substrate heliamine 8, N-methylated product 8a and substrate analog 9. (C) Model of the CNMT (R)-coclaurine 4 docked in the active site and (D) a
model derived from docking (S)-4. (E) Proposed mechanism of CNMT showing the putative role of key active site amino acid residues.
CNMT consists of a relatively smaller, more compact, active site
Consequently, E204, E207 and H208 could potentially stabilize
the substrate ammonium ion through H-bonding or a salt bridge.
One of these residues could also serve as a general base,
deprotonating the ammonium ion to generate a nucleophilic free
amine required for subsequent methylation. To explore these
possibilities, E204, E207 and H208 were each mutated to
2
2
(
CNMT 387
positioning an extended loop (residues 249-260) and residues
4-82 of helix α4 closer to the substrate binding pocket (Table
Å
/PavNMT 1081 Å ) which is shaped by
6
S2). An apo structure was also determined with AdoHcy, but no
substrate bound (PDB 6GKZ) (Figure S3 & Table S1). This
reveals that the helix α4 is disordered in the apo CNMT structure, alanine. The H208A mutant showed the most significant effect,
however, in the presence of substrate becomes ordered and
protrudes directly into the active site, forming a closed cap
covering the substrate entrance, suggesting the helix α4 may
control substrate accessibility and/or recognition by CNMT. A
third structure (PDB 6GKY) was determined for CNMT in
complex with a quinolinone substrate analog 9 (Figure S4). The
structure of CNMT with 9 is similar to the N-methyl-heliamine 8a
structure (Figure 2) except the two ligands are flipped relative to
one another in the active site, with the carbonyl group of 9 in the
same position as the amino group of 8a.
resulting in very low enzyme activity of 1 % and 4 % relative to
wild-type CNMT with substrates 8 and 3 respectively (Figure S6),
suggesting H208 plays a key role in catalysis, most likely
functioning as a general base to deprotonate the substrate
ammonium ion. Sequence alignments with related NMTs
[
9,11]
including the PavNMT,
indicate that the H208 and E207
residues are highly conserved. Furthermore, mutation of these
residues to alanine also results in a large decrease in the activity
of PavNMT. Given that the E207 carboxyl is 7 Å from H208
imidazole, which is in turn 5 Å from the N atom of 8a, in the
CNMT structure, H208 and E207 may function as a catalytic
diad. However, the E207A mutation did not have a significant
impact, with the mutant retaining 60-75 % activity relative to the
wild type (Figure S6). Moreover, the backbone carbonyl group of
T261 is within H-bonding distance of H208, which could similarly
assist H208 in functioning as a general base. In contrast to
E207A, the mutant E204A resulted in a significant drop in
activity (4-6 % relative to wild-type). The close proximity of E204
carboxyl to the N atom of 8a (4 Å) suggest that E204 may inter-
act electrostatically with the substrate/product ammonium ion.
The proximity of the residues Y328, W329, R330, G331
and F332 to 8a, also indicates these residues may be involved
in substrate binding. Single alanine mutations were carried out
for all five positions and these mutants were tested with
substrates 8 and 3. The R330A and G331A mutations showed
negligible effect on activity, whilst W329A and F332A resulted in
a more significant decrease in activity (Figure S6), suggesting
these aromatic residues may provide potential hydrophobic and
π stacking interactions. Interestingly, whilst F332A resulted in a
similar drop in activity with both substrates (8 & 3), the W329A
Modeling studies were also conducted for CNMT with the
natural substrate coclaurine 4, using the software ICM PRO
Molsoft and the ligand bound structure of CNMT to guide the
docking of both (S)- and (R)-4 (Figure 2 & S5). The top docking
hits for (S)- and (R)-4 returned docking energies of e = –19.39
and –9.98 kcal/mol respectively. In both of these lowest energy
structures the THIQ ring of (S)- and (R)-4 occupy the same
position as the product 8a, except the ring is flipped; C1 of 8a is
orientated towards H208, whilst the C1 p-hydroxybenzyl
substituent of (S) and (R)-4 is on the opposite side of the active
site orientated out of the active site. Overall the modeling shows
both enantiomers of coclaurine can be accommodated in the
CNMT active site, consistent with previous studies with this
[
8]
[10]
enzyme and related CNMTs which show that CNMTs exhibit
little stereoselectivity and can methylate both (S)- and (R)-4.
Three active site residues, E204, E207 and H208 are in
close proximity to the methyl group of the product 8a in the
CNMT structure. The amino group of substrates (3, 4 & 8) have
pK
a
values of ca. 8.7-8.8, and will exist predominately as
protonated ammonium ions in aqueous solution at neutral pH.
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201805060
COMMUNICATION
HO
HO
H3CO
HO
H3CO
Table 1. Kinetic parameters of the and CNMT wild-type and mutants enzymes.
N
N
N
R
R
H3CO
R
3
R = H
3a R = CH3
81%[a] (70%)[b]
4
R = H
8 R = H
Sub.
Enzyme
K
M
k
cat
kcat /K
M
-1
-
1
-1
–3
HO
5 R = CH3
8a R = CH3
90%[a] (75%)[b]
HO
(
µM)
(min )
(min µM ) x 10
120 ± 7.3
5.6 ± 0.43
5.2 ± 0.27
ND
2%[a]
9
3
3
b R = CH2CH3[c]
c R = CH2CH=CH2[c]
8
WT
311 ± 18
2096 ± 144
1565 ± 77
ND
35.9 ± 0.6
11.8 ± 0.4
8.1 ± 0.2
ND
Y328A
W329A
F332A
WT
H3CO
H3CO
3
H CO
3
H CO
N
N
N
R
R
3
H CO
R
H3CO
1
1
8
0 R = H
0a R = CH3
8% (66%)
CN
11 R = H
12 R = H
11a R = CH3
87% (56%)
12a R = CH3
3
265 ± 31
601 ± 84
1033 ± 64
1726 ± 142
31.9 ± 1.1
7.0 ± 0.3
0.6 ± 0.01
1.2 ± 0.04
120 ± 14
85% (60%)
Y328A
W329A
F332A
12 ± 1.7
H3CO
H3CO
O
O
H3CO
0.63 ± 0.045
0.70 ± 0.062
N
N
N
R
R
3
H CO
R
OH
14 R = H
15 R = H
1
1
3 R = H
1
4a R = CH3
15a R = CH3
5% (32%)
3a R = CH3
5
8% (40%)
mutant exhibited a more significant drop in activity with 3 than 8,
indicating that F332 may interact with the THIQ aryl group
common to both substrates, whilst W329 may be interact with
the p-hydroxybenzyl group of 3. The fact that F332 is closer to
the product 8a, in the CNMT structure, than W329, would also
support this hypothesis (Figure 2). To explore this further, kinetic
parameters of the WT along with mutants Y328A, W329A and
F332A were determined with substrates 8 and 3 (Table 1 &
85% (62%)
H3CO
H3CO
H3CO
3
H3CO
N
N
N
R
H CO
R
3
H CO
R
16 R = H
6a R = CH3
% (32%)
CO2H
17 R = H
18 R = H
1
17a R = CH3
54% (45%)
18a R = CH3
22% (28%)
4
Figure 3. Substrate scope of CNMT showing [a] % conversions determined by
HPLC 45 min assays [b] isolated yields from CNMT-CLEA reactions and [c]
products from alkylation with S-ethyl and S-allyl AdoMet (Figures S25 & S26).
m
Figures S7-S13). The mutants all showed increased K values
with the two substrates, signifying that these residues are
important in substrate binding. Moreover, W329A afforded far
slower turnover with 3, exhibiting a kcat value with 3 that is 13-
fold lower than with 8, which is consistant with W329
participating in π-π stacking or an edge to face interaction with
the p-hydroxybenzyl substituent of 3 that may stabilize the
orientation of 3, relative to AdoMet, for efficient methylation
to screen for alternative, non-natural substrates (Figures S15 &
S16). The conversions of the substrates to N-methylated
products (Figures 3 & S15), was then confirmed by HPLC and
LCMS (Figure S17-S25). Coclaurine 4 and heliamine 8 proved
to be the best substrates (90-92
% conversion), whilst
norcoclaurine 3 and C1-substituted THIQs 10-13 were turned
over to lesser extent (85-88 %). The 6,7-diethoxy derivative 14,
was also methylated at lower levels than the substrates
possessing 6,7-dimethoxy substituents. THIQ 16, possessing a
C1-acetate substituent was also a poor substrate, presumably
because the carboxyl side chain is destabilized in more
hydrophobic binding pocket that accommodates the p-
hydroxybenzyl substituent of the natural substrate. A series of
THIQs with C3- or C4-substituents were also synthesized
(
Figure 2). Similarly, the kinetic parameters for F332A with 3
indicated that the F332 residue may also be involved in
substrate binding through π-π stacking. Kinetic parameters
could not be determined for F332A with 8 due to lack of
substrate saturation, which further supports the hypothesis that
F332 interacts with the THIQ aryl ring common to 8 and 3. The
combined structural and mutagenesis studies, indicate that
substrate binding in CNMT occurs through hydrophobic aromatic
residues (F322, W329) as well as a salt bridge between E204
and the substrate ammonium ion, which is most likely
deprotonated by H208, possibly aided by the backbone carbonyl
group of T261, to generate the nucleophilic amine species
required for methylation by AdoMet (Figure 2E).
(
Figure S16) and tested as CNMT substrates. THIQs 17 and 18,
with C4-methyl substituents were N-methylated (54 % and 22 %
respectively, Figure 3). However, THIQs with larger C3- or C4-
substituents gave very low levels of methylation products (Figure
S15). Thus whilst CNMT can accommodate THIQs with various
C1-substituents, the enzyme is less tolerant to modifications at
C3 or C4. The CNMT structure (Figure 2) indicates that larger
substituents at C3 or C4 would clash with E204, E207, H208
and I234 residues of the CNMT active site.
Whilst earlier studies have shown that C. japonica CNMT
accepts a number of natural THIQ as substrates, the broader
[
8]
synthetic potential of this enzyme has not been explored.
Given that N-methyl THIQ is a privileged pharmacophore, found
in many pharmaceuticals, (Figure S1) we sought to explore the
substrate scope of CNMT with synthetic substrates. N-
methylation can be achieved using non-enzymatic chemistries,
but often these methods use toxic, or otherwise deleterious,
reagents and lack selectivity, which with more complex
substrates can lead to side products or the laborious use of
protective groups. Cleaner, milder and more selective
biocatalytic methods for THIQ N-methylation may thus be useful
in drug synthesis. Accordingly, a colorimetric assay with CNMT
and S-adenosylhomocysteine hydrolase (Figure S14) was used
Given that CNMT has relatively low catalytic activity even
with native substrates (Table 1), immobilization of CNMT
[
12]
through the formation of cross-linked aggregates (CLEAs)
was explored in order to obtain a more robust biocatalyst. The
CNMT-CLEAs proved to have improved stability and can be
recycled at least 3 times without significant loss of activity. In
addition substrate concentration can be increased to 3-4 mM,
with CNMT-CLEAs, enabling biotransformations to be achieved
on a preparative scale and methylated products to be isolated
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201805060
COMMUNICATION
in good yields (Figure 3), even with poor substrates (e.g. 15-18).
To further explore the biocatalytic scope of CNMT, AdoMet
analogs with alternative S-alkyl substituents were tested as co-
factors. AdoMet analogs have been used for alkyl diversification
of natural products in an effort to generate variants with
Keywords: Methyltransferase • structure • mechanism •
biosynthesis • biocatalysis
[
[
1]
2]
J. M. Hagel,P. J. Facchini, Plant Cell Physiol. 2013, 54, 647–672.
L. Narcross, E. Fossati, L. Bourgeois, J. E. Dueber, V. J. J. Martin,
Trends Biotechnol. 2016, 34, 228–241.
[
13]
improved properties. Such biocatalytic alkyl diversification of
THIQ is an attractive goal given the broad ranging bioactivity of
these compounds. In addition to CNMT there are several other
methyltransferases involved in biosynthesis of BIAs and related
alkaloids, which could facilitate regioselective alkyl derivatization
of these pharmacologically important compounds. To test this
approach with CNMT, ethyl- and allyl-AdoMet analogs were
generated in situ, from L-ethionine and S-allyl-L-homocysteine
using a mutant of the human methionine adenosyltransferase
[
3]
a) A. Nakagawa, C. Matsuzaki, E. Matsumura, T. Koyanagi, T.
Katayama, K. Yamamoto, F. Sato, H. Kumagai, H. Minami, Sci. Rep.
2
014, 4, 1–8; b) A. Nakagawa, H. Minami, J.-S. Kim, T. Koyanagi, T.
Katayama, F. Sato, H. Kumagai, Nat. Commun. 2011, 2, 326; c) E.
Fossati, A. Ekins, L. Narcross, Y. Zhu, J. Falgueyret, G. A. W.
Beaudoin, P. J. Facchini, V. J. J. Martin, Nat. Commun. 2014, 5, 1–11;
d) S. Galanie, K. Thodey, I. J. Trenchard, M. F. Interrante, C. D.
Smolke, Science 2015, 349, 1095–1100.
[
[
4]
5]
a) J. H. Schrittwieser et al. RSC Adv. 2013, 3, 17602; b) H. Kries, S. E.
O’Connor, Curr. Opin. Chem. Biol. 2016, 31, 22–30.
[
13]
(
hMAT2A I322V) as described previously.
Both co-factor
a) B. M. Ruff, S. Bräse, S. E. O’Connor, Tetrahedron Lett. 2012, 53,
analogues were accepted by CNMT, with 3 as a substrate,
affording N-ethyl and N-allyl norcoclaurine 3b and 3c (Figures 3,
S26 & S27).ꢀ
1
071–1074; b) B. R. Lichman, J. Zhao, H. C. Hailes, J. M. Ward, Nat.
Commun. 2017, 8, 14883.
[6]
V. Resch, H. Lechner, J. H. Schrittwieser, S. Wallner, K. Gruber, P.
MacHeroux, W. Kroutil, Chem. Eur. J. 2012, 18, 13173–13179.
V. Erdmann, B. R. Lichman, J. Zhao, R. C. Simon, W. Kroutil, J. M.
Ward, H. C. Hailes, D. Rother Angew. Chem. Int. Ed.
ꢀ
In summary, we have determined the first structure of
[
[
[
7]
8]
9]
CNMT, which combined with active site mutagenesis suggest
the likely to function of key active site residues. The catalytic
scope of CNMT was assessed with a range of natural and
synthetic substrates, indicating that the enzyme exhibits relaxed
substrate specificity with C1-substituted THIQs. Finally, analogs
of the AdoMet co-factor were shown to be accepted by CNMT
leading to N-alkylated products. Taken together, the knowledge
presented here will be valuable in structure-guided engineering
of CNMT to generate enzymes with improved properties
required for microbial production of BIAs. Similarly engineered
variants of CNMT with improved activity and specificity could
provide new routes to synthetic drugs containing the privileged
N-alkyl THIQ pharmacophore.
1
0.1002/anie.201705855.
a) K. B. Choi, T. Morishige and F. Sato, Phytochemistry 2001, 56, 649–
5; b) K. B. Choi, T. Morishige, N. Shitan, K. Yazaki and F. Sato, J. Biol.
5
Chem. 2002, 277, 830–835.
M. A. Torres, E. Hoffarth, L. Eugenio, J. Savtchouk, X. Chen, J. S.
Morris, P. J. Facchini and K. K. S. Ng, J. Biol. Chem., 2016, 291,
23403–23415.
[
[
10] T. Frenzel and M. H. Zenk, Phytochemistry 1990, 29, 3491–3497.
11] a) D. K. Liscombe, J. Ziegler, J. Schmidt, C. Ammer, P. J. Facchini,
Plant J., 2009, 60, 729–743; b) J. S. Morris, P. J. Facchini, J. Biol.
Chem. 2016, 291, 23416–23427.
[
[
12] R. A. Sheldon, Biochem. Soc. Trans. 2007, 35, 1583–1587.
13] a) B. J. C. Law, A. Struck, M. R. Bennett, B. Wilkinson, J. Micklefield,
Chem. Sci. 2015, 6, 2885–2892; b) B. J. C. Law, M. R. Bennett, M. L.
Thompson, C. Levy, S. A. Shepherd, D. Leys, J. Micklefield, Angew.
Chem. Int. Ed. 2016, 55, 2683–2687.
Acknowledgements: We thank BBSRC (grants BB/K00199X/1
&
BB/M017702/1 SynBioChem) and GSK for support.
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201805060
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
Coclaurine N-methyltransferase
Matthew R. Bennett, Mark L. Thompson,
Sarah A. Shepherd, Mark S. Dunstan,
Abigail J. Herbert, Duncan R. M. Smith,
Victoria A. Cronin, Binuraj R. K. Menon,
Colin Levy and Jason Micklefield*
(
CNMT) installs the N-methyl
AdoMet
substituent essential for bioactivity of
therapeutically important benzyl-
isoquinoline alkaloids (BIA). CNMT’s
active site architechture is defined
through structure guided mutagenesis
and its biocatalytic scope is explored
with synthetic substrates and co-factor
analogs. These studies provide
E204
OH
O
O–
H
N
+
S
CH3
N
H
W329
F322
Page No. – Page No.
H
N
+
N
H208
Structure and Biocatalytic Scope of
Coclaurine N-Methyltransferase
H
OH
OCH3
O
NH
T261
insights for engineering and exploiting
CNMT in the production of bioactive
BIA and synthetic derivatives.
This article is protected by copyright. All rights reserved.