Structural basis of substrate recognition in human nicotinamide N-methyltransferase

Article abstract of DOI:10.1021/bi2007614

Nicotinamide N-methyltransferase (NNMT) catalyzes the N-methylation of nicotinamide, pyridines, and other analogues using S-adenosyl-l-methionine as donor. NNMT plays a significant role in the regulation of metabolic pathways and is expressed at markedly high levels in several kinds of cancers, presenting it as a potential molecular target for cancer therapy. We have determined the crystal structure of human NNMT as a ternary complex bound to both the demethylated donor S-adenosyl-l-homocysteine and the acceptor substrate nicotinamide, to 2.7 A resolution. These studies reveal the structural basis for nicotinamide binding and highlight several residues in the active site which may play roles in nicotinamide recognition and NNMT catalysis. The functional importance of these residues was probed by mutagenesis. Of three residues near the nicotinamide's amide group, substitution of S201 and S213 had no effect on enzyme activity while replacement of D197 dramatically decreased activity. Substitutions of Y20, whose side chain hydroxyl interacts with both the nicotinamide aromatic ring and AdoHcy carboxylate, also compromised activity. Enzyme kinetics analysis revealed k_cat/K_m decreases of 2-3 orders of magnitude for the D197A and Y20A mutants, confirming the functional importance of these active site residues. The mutants exhibited substantially increased K_m for both NCA and AdoMet and modestly decreased k_cat. MD simulations revealed long-range conformational effects which provide an explanation for the large increase in K _m(AdoMet) for the D197A mutant, which interacts directly only with nicotinamide in the ternary complex crystal structure.

Full text of DOI:10.1021/bi2007614

Article
pubs.acs.org/biochemistry
Structural Basis of Substrate Recognition in Human Nicotinamide
N-Methyltransferase
Yi Peng, Davide Sartini, Valentina Pozzi, Dennis Wilk, Monica Emanuelli, and Vivien C. Yee*
†
‡
‡
†
‡
,†
†Department of Biochemistry, Case Western Reserve University, Cleveland, Ohio 44106, United States
‡Dipartimento di Biochimica, Biologia e Genetica, Universita
̀
Politecnica Marche, Ancona, Italy
ABSTRACT: Nicotinamide N-methyltransferase (NNMT) catalyzes the
N-methylation of nicotinamide, pyridines, and other analogues using S-adenosyl-
L-methionine as donor. NNMT plays a significant role in the regulation of
metabolic pathways and is expressed at markedly high levels in several kinds of
cancers, presenting it as a potential molecular target for cancer therapy. We
have determined the crystal structure of human NNMT as a ternary complex
bound to both the demethylated donor S-adenosyl-L-homocysteine and the
acceptor substrate nicotinamide, to 2.7 Å resolution. These studies reveal the
structural basis for nicotinamide binding and highlight several residues in the
active site which may play roles in nicotinamide recognition and NNMT catalysis.
The functional importance of these residues was probed by mutagenesis. Of three
residues near the nicotinamide’s amide group, substitution of S201 and S213 had
no effect on enzyme activity while replacement of D197 dramatically decreased
activity. Substitutions of Y20, whose side chain hydroxyl interacts with both the
nicotinamide aromatic ring and AdoHcy carboxylate, also compromised activity. Enzyme kinetics analysis revealed k /K decreases
cat
m
of 2−3 orders of magnitude for the D197A and Y20A mutants, confirming the functional importance of these active site residues.
The mutants exhibited substantially increased Km for both NCA and AdoMet and modestly decreased kcat. MD simulations revealed
long-range conformational effects which provide an explanation for the large increase in Km(AdoMet) for the D197A mutant, which
interacts directly only with nicotinamide in the ternary complex crystal structure.
icotinamide (NCA) is a compound essential for the
formation of NAD(H) and NADP(H), which are
PDB accession codes 2IIP and 2I62, respectively), they did not
reveal the NCA acceptor substrate binding site. We report here
the crystal structure of hNNMT bound to both AdoHcy and
NCA. The structure reveals the protein features important for
NCA binding and highlights several residues in the active site
which were subsequently tested by mutagenesis for their
possible importance in NCA recognition and NNMT catalysis.
N
involved in many important biological processes, including
energy production, cellular resistance to stress, and longevity.
Nicotinamide N-methyltransferase (NNMT) is a cytosolic
enzyme that catalyzes the N-methylation of NCA, pyridines,
and analogues using S-adenosyl-L-methionine (AdoMet) as the
1
methyl donor. NNMT plays a role in NCA metabolism and in
the detoxification of many xenobiotics. The enzyme was first
characterized in liver where it is mainly expressed; there its
activity varies 5-fold among individuals and has a bimodal
MATERIALS AND METHODS
2
■
Cloning, Expression, and Purification. Total RNA was
isolated from normal renal tissue using the SV Total RNA
Isolation System (Promega). RNA (1 μg) was reverse
transcribed with a first strand cDNA synthesis kit II (Bio
Basic) using oligo-dT18 primers. 0.5 μL of the reaction mixture
was then subjected to PCR with Taq polymerase (total volume
25 μL) using the primers 5′-tcacatatggaatcaggcttca-3′ (forward)
and 5′-ctaaagctttcacaggggtctg-3′ (reverse) to amplify the human
NNMT ORF and to insert NdeI and HindIII restriction sites.
The amplified and digested PCR product was cloned into a
pT7-7 plasmid vector to obtain the expression construct pT7-7-
wt-hNNMT, used to transform E. coli BL21 (DE3) cells that
were grown at 37 °C to an OD600 of 1.0 before induction with
1 mM IPTG and incubation overnight. Cells were harvested by
frequency distribution, which raises the possibility that a genetic
polymorphism may be associated with its enzyme activity. In
1
addition, in Parkinson’s disease, an enhanced NNMT activity
seems to produce toxic N-methylpyridinium compounds,
advanced as possible neurotoxins underlying nigrostriatal
3,4
degeneration. Lower expression levels of NNMT are also
found in kidney, lung, placenta, heart, brain, and muscle.
Interestingly, abnormally high expression of NNMT has been
identified in various cancers such as papillary thyroid carcinoma,
5−9
colorectal cancer, lung cancer, and oral carcinoma.
To
elucidate the structural basis of NNMT catalysis and function
and to aid the design of new NNMT inhibitors that may have
therapeutic use, crystal structures of NNMT bound to substrate
or inhibitor are desirable. Although the crystal structures of
human and mouse NNMT (hNNMT and mNNMT,
respectively) bound to the demethylated product S-adenosyl-L-
homocysteine (AdoHcy) have been determined (unpublished,
Received: May 17, 2011
Revised: August 3, 2011
Published: August 8, 2011
©
2011 American Chemical Society
7800
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
a
centrifugation and resuspended in lysis buffer (10 mM Tris-
HCl, pH 8.6, PMSF 1 mM, DTT 1 mM, and aprotinin 2 μg/
mL) before sonication and centrifugation. The supernatant was
loaded onto a hydroxyapatite column, equilibrated with 10 mM
sodium phosphate, pH 7.5, DTT 1 mM, and purified wt-
hNNMT eluted with a linear gradient of NaCl from 0 to 1 M in
the equilibration buffer.
Extensive crystallization trials failed to yield diffracting
crystals of wt-hNNMT. During the course of this work, the
crystal structure of a wt-like hNNMT K100A/E101A/E103A
triple mutant (“tm”) was deposited in the PDB (unpublished,
PDB accession code 2IIP). This mutant was designed to facilitate
crystallization by reducing the entropy of three surface-exposed
residues in a loop distant from the active site. This tm-hNNMT
was reproduced using the QuikChange kit (Stratagene) using the
primer 5′-ggagaagtggctgaaggcagcgccagcggcctttgactggtcc-3′ (sub-
stituted nucleotides are underlined) and its complement. Briefly,
Table 1. Data Collection and Refinement Statistics
data collection
cell dimensions
a, b, c (Å)
60.7, 61.8, 74.3
106.6, 104.0, 104.1
0.97857
α, β, γ (deg)
wavelength (Å)
resolution (Å)
Rsym (%)
50−2.7 (2.8−2.7)
9.1 (47.6)
I/σI
9.6 (1.9)
completeness (%)
redundancy
98.7 (97.5)
2.5 (2.4)
refinement
resolution (Å)
no. of reflections
Rwork/Rfree (%)
no. of atoms
50−2.7 (2.8−2.7)
23684
20.5/26.4
protein, ligand, water
8082, 140, 50
2
5 ng of pT7-7-wt-hNNMT template DNA was incubated with
av B-factors (Å2)
the mutagenic primers, dNTPs, and Pfu DNA polymerase
according to manufacturer’s recommendations. The full-length
tm-hNNMT coding region was then cloned into the NdeI and
HindIII sites of a pET-28a plasmid (Novagen) to express
N-terminally His6-tagged tm-hNNMT. This pET-28a-tm-
hNNMT vector was transformed into E. coli BL21(DE3) cells
that were grown at 37 °C to an OD600 of ∼1.0 before induction
with 1 mM IPTG and incubation overnight at 25 °C. Cells
were harvested by centrifugation and resuspended in lysis buffer
4
chains: protein, AdoHcy, NCA, water
37.0, 32.1, 48.7, 24.6
31.1, 31.3, 42.4, 43.4
26.7, 21.5, 41.8, 38.2
31.4, 38.7, 57.3, 67.3
protein: chains A, B, C, D
AdoHcy: chains A, B, C, D
NCA: chains A, B, C, D
b
rmsd
bond lengths (Å)
bond angles (deg)
Ramachandran
0.017
1.678
most favored (%)
additional allowed (%)
disallowed (%)
82.0
18.0
(50 mM Tris-HCl, pH 8.0, 0.5 M NaCl, 5 mM imidazole, 2 mM
β-mercaptoethanol, and 5% glycerol, 1 mM PMSF) before
sonication. After centrifugation, the supernatant was loaded onto
Ni-NTA resin (Qiagen) and washed with 50 mM Tris-HCl, pH
0.0
a
b
Numbers in parentheses refer to the highest resolution shell. rmsd =
root-mean-square deviation.
8.0, 0.5 M NaCl, 25 mM imidazole, and 5% glycerol. The
expressed tm-hNNMT protein was eluted with 50 mM Tris-
HCl, pH 8.0, 0.5 M NaCl, 250 mM imidazole, and 5% glycerol.
The combined eluted fractions were dialyzed against 50 mM
Tris-HCl, pH 8.0, 100 mM NaCl, 5% glycerol, and 1 mM DTT.
Final purification was by gel filtration on a Superdex 75 column
structure of tm-hNNMT bound to AdoHcy (PDB accession
code 2IIP). A 10-residue N-terminal extension, apparently a
cloning artifact, was removed from the search model since it is
not present in the construct used for crystallization in this work.
Iterative rounds of model building were performed with
(Amersham Biosciences) equilibrated with 20 mM Tris-HCl, pH
12
8.0, 50 mM NaCl, 5% glycerol, and 1 mM DTT.
COOT, and refinement calculations were carried out in
13 14
CNS and REFMAC. A Ramachandran plot calculated with
Crystallization and Data Collection. The purified tm-
15
PROCHECK placed 82.0% of the residues in the most
favored regions and none in disallowed regions. No density is
observed for a couple of residues at the N- and C-termini. The
final refined structure includes four molecules of NNMT
containing residues 3−261 (chains A−C) or 5−261 (chain D),
AdoHcy and NCA ligands bound to each, and 50 water
molecules. A summary of the refinement statistics is provided in
hNNMT was incubated with AdoHcy at 1:4 molar ratio before
concentration to 10 mg/mL, followed by incubation with
1
0 mM NCA for ∼30 min on ice before crystallization. Crystals
of tm-hNNMT bound to AdoHcy and NCA were grown by
sitting drop vapor diffusion at 20 °C using a 1:1 protein:-
precipitant volume ratio from two conditions: 25% PEG3350 in
0
.1 M bis-Tris (pH 5.5), and 25−27% PEG3350 with 0.2 M
16
Table 1. All figures were generated using PyMol.
NaCl in 0.1 M Bis-Tris (pH 5.5) or imidazole (pH 6.5).
Crystals grew in clusters which needed to be dissected to obtain
single crystals for data collection. Cryoprotection in artificial
mother liquor containing 35% PEG3350 was followed by flash
cooling of the crystals in liquid nitrogen. The best diffracting
crystal yielded a 2.7 Å resolution native diffraction data set
Preparation of Active Site Mutants. Site-directed muta-
genesis to change the selected active site residues was carried out
using the QuikChange kit (Stratagene), similarly as described
earlier to generate tm-hNNMT from wt-hNNMT. The
mutagenic primers used were 5′- atctaagccattttaaccctcgggatgccc-
tagaaaaatattacaagtttg-3′ and its complement for the
Y20A mutant,
5′- taagccattttaaccctcgggatttcctagaaaaatattacaagttt-3′ and its com-
plement for Y20F, 5′- gcagaagttctccgcc ctccccctgggc-3′ and its
complement for S213A, 5′-gatcatggatgcgctcaaggccagctactacat-
gattggt-3′ and its complement for S201A, and 5′- tcctggtgat-
catggctgcgctcaagagcag-3′ and its complement for D197A
(substituted nucleotides are underlined). The pET-28a-tm-
hNNMT plasmid was used as the template for the PCR
mutagenesis reaction.
10
measured at APS beamline 19ID and processed with HKL
Table 1). The crystal form is not isomorphous to either the
(
tm-hNNMT-AdoHcy or mNNMT-AdoHcy structures depos-
ited in the PDB (accession codes 2IIP and 2I62). The tm-
hNNMT ternary complex crystal has space group P1 and
contains four molecules in the asymmetric unit/unit cell.
Structure Determination and Refinement. The structure
was determined by molecular replacement using the program
1
1
MOLREP and the protein coordinates from the crystal
7
801
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
The active site mutants were expressed and purified similarly
as for tm-hNNMT, with a few modifications. Induction of
transformed bacteria was carried out at 18 °C, and harvested
cells were resuspended in a modified lysis buffer (50 mM
sodium phosphate, pH 8.0, 0.3 M NaCl, 10 mM imidazole,
tm-hNNMT exhibited similar activity, confirming that alter-
ation of the three surface-exposed residues ∼20 Å from the
active site did not compromise enzyme function.
Molecular Dynamics (MD) Simulations. Chain A of the
tm-hNNMT-AdoHcy-NCA crystal structure was used as the
starting point for modeling and MD simulations. Residues Y20
and D197 were computationally replaced with phenylalanine
or alanine to generate initial structural models of Y20F, Y20A,
and D197A mutants. AdoHcy, NCA, and crystallographic
solvent atoms were removed, and parallel MD simulations of
tm-hNNMT and the three mutants were performed using the
GROMACS-4.0.7 software package and the G43a1force
0
.5 mM DTT, 1 mM PMSF, and 2 μg/mL aprotinin) before
sonication. After centrifugation of lysed cells, the supernatant
was applied to a Ni-NTA Superflow Cartridge (Qiagen)
equilibrated with 50 mM sodium phosphate, pH 8.0, 0.3 M
NaCl, 10 mM imidazole, 0.5 mM DTT, and further washed
with 50 mM sodium phosphate, pH 8.0, 0.3 M NaCl, 20 mM
imidazole, 0.5 mM DTT. The tm-hNNMT mutants were
eluted by a linear gradient of imidazole from 20 to 250 mM in
18,19
field.
The protein structures were first solvated in a cubic
5
0
0 mM sodium phosphate, pH 8.0, containing 0.3 M NaCl and
.5 mM DTT and used to determine catalytic properties.
periodic box with walls extending 9 Å from the protein
20
coordinates using the SPC water model. Energy was minimized
in 400 steps using a steepest descents minimization procedure. In
order to solvate the system without distorting the protein
models, 30 ps of position-restrained MD was performed before
an extended 5 ns MD simulation with coordinates saved for
analysis every 1 ps. Polypeptide backbone rmsd values relative to
the starting structure were calculated for coordinates saved at
different time points during each simulation.
Enzyme Assays. Initial enzymatic activities were carried
out using a 0.45 mL standard assay mixture containing 50 mM
Tris-HCl, pH 8.6, 1 mM DTT, 5 mM NCA, 0.5 mM AdoMet
and 0.024−0.059 mg/mL of enzyme. Each reaction was started
by adding AdoMet, carried out at 37 °C for incubation times
ranging between 3 and 60 min, and stopped by adding 100 μL
assay mixture to 50 μL of ice-cold 1.2 M HClO . After 10 min
4
at 0 °C proteins were removed by centrifugation, and 130 μL of
the perchloric acid supernatant was then neutralized by adding
RESULTS
■
3
5 μL of 0.8 M K2CO3. The resulting KClO4 was removed by
Overall Structure of NNMT Ternary Complex. The
crystal asymmetric unit contains four conformationally similar
copies of hNNMT; pairwise superpositions of the Cα atoms
give overall rmsd values ranging from 0.4 to 0.7 Å.
Superposition with the previously determined hNNMT-
AdoHcy and mNNMT-AdoHcy structures gave similar rmsd
values of 0.4 and 1.0 Å, respectively. Electron density for both
AdoHcy and NCA was found in all four active sites, but since
the average B-factors for NCA and protein are slightly lower
for chain A than the other three (Table 1), structural details
provided in the following sections are for molecule A unless
otherwise noted.
centrifugation, and an aliquot of the neutralized supernatant
was injected into a Supelcosil LC-18 5 μm reversed phase
17
column (Sigma-Aldrich) and eluted as previously described.
1
The amount of N -methylnicotinamide produced was deter-
mined by the elution peak areas of the separated compound,
with 1 U of enzyme activity representing the formation of
1
1
3
nmol of N -methylnicotinamide per hour of incubation at
7 °C. For substrate kinetics studies, NCA concentrations were
varied from 0.1 to 100 mM, and AdoMet concentrations ranged
from 0.01 to 25 mM. Three independent sets of kinetic experi-
ments were carried out and in duplicate each time. Relative
enzyme specific activities and enzyme kinetics results are
reported in Tables 2 and 3, respectively. Both wt-hNNMT and
NNMT possesses the class I AdoMet-dependent methyl-
21−23
transferase (MTase) core fold,
seven-stranded β sheet flanked by α helices on both sides
Figure 1A). The NNMT structure has several modifications
which is comprised of a
a
Table 2. Enzyme Activity of hNNMT Proteins
(
3
to the MTase core, namely the addition of two α helices at its
N-terminus and a reverse β hairpin from Y203-S212 which
together form a “cap” over the active site, and two α-helices
inserted within W107-A134, adjacent to one set of flanking
helices.
specific activity (×10 U/mg of
relative specific activity
hNNMT
protein)
(%)
wt
9.1 ± 0.5
10.6 ± 0.5
0.064 ± 0.007
4.6 ± 0.5
86
100
0.6
43
tm-“wt”
Y20A
Y20F
24
A structural similarity search using DALI identified a large
number of known and putative MTase structures, of which
two stood out as the closest matches: human indolethylamine
N-methyltransferase (hINMT) and human phenylethanolamine
N-methyltransferase (hPNMT). The hINMT structure is
bound to AdoHcy (PDB accession code 2A14, 1.1 Å rmsd
D197A
S201A
S213A
0.250 ± 0.028
8.5 ± 0.4
12.3 ± 0.7
2
81
117
a
Values are given as the means ± standard deviation of three
independent sets of experiments, each carried out in duplicate.
a
Table 3. Enzyme Kinetics of hNNMT Proteins
kcat/Km (s− mM−1
1
)
Km (mM)
hNNMT
NCA
AdoMet
kcat (s−1)
NCA
AdoMet
b
wt
0.43−0.38
0.105 ± 0.009
6.3 ± 0.8
0.24 ± 0.04
11.8 ± 2.0
0.0018−0.0022
0.0050 ± 0.0006
3.5 ± 0.3
0.028 ± 0.004
2.73 ± 0.16
0.077
0.083
0.031
0.041
0.036
0.18−0.2
0.8
35−43
17
tm-“wt”
Y20A
0.005
0.17
0.009
1.5
Y20F
D197A
0.003
0.013
a
b
Values are given as the means ± standard deviation of three independent sets of experiments, each carried out in duplicate. From ref 1.
7
802
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
Figure 1. Crystal structure of hNNMT. (A) Ribbon diagram of the hNNMT-AdoHcy-NCA ternary complex. The hNNMT protein (chain A) is
colored cyan for the conserved class I AdoMet-dependent MTase core fold and light purple for the nonconserved inserted regions characteristic of
the NNMT structure, including the “cap” over the active site and the flexible F27−H31 loop (dark purple). AdoHcy (green) and NCA (orange) are
2
5
shown as space-filling spheres. (B) Ribbon diagram of the hPNMT-AdoHcy-NA ternary complex crystal structure (PDB accession code 3HCD).
AdoHcy (green) and the NA substrate (beige) are rendered as space-filling spheres, and the hPNMT with the conserved core fold in cyan and the
PNMT-characteristic inserted regions in light purple.
for 257 superimposed Cα atoms, 52% sequence identity) while
of the many hPNMT structures, the most relevant for com-
parison is that of the ternary complex bound to both AdoHcy
and a noradrenaline acceptor substrate (PDB accession code
variation in orientation, in the four hNNMT molecules in the
asymmetric unit. In chains A−C, the NCA ligands are in nearly
the same orientation, while in chain D the NCA is tilted ∼40°
(Figure 4A). The NCA density is clearest in chains A and B,
and the average refined B-factors for NCA in these two chains
are lower than those in the other two chains by almost 2-fold
(Table 1); thus, the NCA orientation as bound to chain A is
selected as the clearest representative of NCA binding to the
hNNMT active site.
3
HCD, 1.5 Å rmsd for 251 superimposed Cα atoms, 39%
25
sequence identity). The next most similar ternary structure is
that of a plant enzyme, 3,7-dimethylxanthine methyltransferase
(
DXMT), bound to both AdoHcy and a theobromine acceptor
substrate (PDB accession code 2EFJ, 3.0 Å rmsd for 200
26
superimposed Cα atoms, 16% sequence identity). Reflecting
their relative degrees of structural and sequence conservation,
NNMT, INMT, and PNMT share similar patterns of insertions
in the MTase core fold (Figure 1A,B) while DXMT has more
extensive additions, particularly in the region capping the active
site (not shown).
NNMT catalyzes the transfer of a methyl group from the
sulfur of AdoMet to the NCA ring N1 atom. The distance
between the NCA N1 and AdoHcy sulfur atom ranges from
3
.5 to 4.2 Å. This range is a little longer than the sum of S−C
and N−C single bonds and is consistent with the S 2-like
mechanism presumed for AdoMet-dependent MTases.
N
AdoHcy Binding to NNMT. Electron density correspond-
ing to AdoHcy is very well-defined in all hNNMT molecules
Y204, part of the constellation of conserved active site
aromatic residues, forms a hydrophobic clamp along with L164
to position the NCA ring. The side chains of three hNNMT
residues, D197, S201, and S213, are within hydrogen-bonding
distance of the NCA amide (Figure 2B). In addition, the Y20
hydroxyl group is close enough to the other end of the
substrate to form electrostatic interactions with the NCA ring.
Structure-Based Mutagenesis of NNMT. To probe the
functional importance of the four hydrophilic active site residues
interacting with NCA, Y20, D197, S201, and S213 were targeted
for mutagenesis. D197A, S201A, and S213A mutants were
designed to remove the side-chain hydrogen-bonding inter-
actions with the NCA amide. Y20F and Y20A mutants were
prepared to test whether the hydroxyl group or aromatic ring of
the side chain is important. Enzyme activity assays were carried
out, and the relative specific enzyme activities were compared
(Table 2). The Y20F mutant had only modestly diminished
activity (to 43%) while the Y20A mutant was much more
dramatically compromised (0.6% of wild-type activity). The
D197A mutant activity was substantially decreased, to 2.4% that
of wt. The other two polar side chains near the NCA amide do
not appear to play a role in function, since the S201A and
S213A exhibited near wt activity.
(
Figure 2A). AdoHcy is bound in a similar position and con-
formation in NNMT as seen in other class I MTase structures
Figure 1A,B), through extensive hydrogen-bonding and van
(
der Waals interactions. The adenine ring is sandwiched
between the hydrophobic side chains of Y86 and A169. In
addition, it forms hydrogen bonds to hNNMT, to the V143
main-chain amide via ring atom N1 (chains A, C, and D), and
to the D142 side chain via N6 (chain B). The ligand’s ribose
hydroxyl groups form hydrogen bonds with the D85 (all four
chains) and N90 (chain B) side chains. At the other end of
AdoHcy, the homocysteine amine group hydrogen bonds to
the carbonyl oxygen atoms of G63 and T163, and the two
carboxylate oxygen atoms interact with four side chain
hydroxyls, those of Y20, Y25, Y69, and T163 (Figure 2A). As
also seen in INMT and PNMT, a constellation of conserved
aromatic residues lines the active site. In hNNMT, these are
Y20, Y25, and Y69 (interacting with AdoHcy carboxylate), Y11
and F15 (near AdoHcy ribose), Y86 (packs against AdoHcy
adenine, is phenylalanine in PNMT and INMT), and Y204.
Nicotinamide Binding to NNMT. The electron density
for the NCA acceptor substrate is weaker than for AdoHcy
(
Figure 2B), consistent with significantly hNNMT higher
affinity for AdoMet (low μM) than for NCA (high μM) (ref 1
and Table 3). NCA is bound in the same position, with a small
Enzyme Kinetics. Detailed enzyme kinetics were then
carried out for the three active site mutants with significantly
7
803
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
Figure 2. hNNMT-AdoHcy-NCA active site (chain A). (A) Stereoview of the AdoHcy binding site. AdoHcy is shown as a ball-and-stick structure
with green carbon atoms; the nearby NCA substrate has orange carbon atoms. Hydrogen bonds involving several residues which interact with NCA
(
side chains of Y20, D197, S201, and S213 with gray carbon atoms) are represented as dashed lines. The 2.7 Å resolution simulated annealing
|
Fo| − |Fc| omit density is contoured at 3σ (blue mesh). (B) Stereoview of the NCA binding site. The 2.7 Å resolution simulated annealing |Fo| − |Fc|
omit density, contoured at 1.0σ (pink mesh), is weaker than that for AdoHcy but clearly shows the bound NCA. (C) Stereoview of hNNMT region
which is conformationally variable in D197A MD simulation. The D197 side chain forms H-bonds to both NCA and the D167 main chain. During
MD simulation, the V138-L154 region changes conformation most significantly. Residues in this region interact with the adenine moiety of AdoHcy,
explaining how the long-range conformational effects of D197A may influence AdoMet affinity.
decreased activity and compared with wt parameters (Table 3).
Values for kcat for the Y20A, Y20F, and D197A mutants were
decreased modestly by only ∼2-fold compared to wt, while Km
for both NCA and AdoMet were more substantially affected.
For Y20A and D197A, Km(NCA) increased 60- and 112-fold,
respectively, while the effect on Km(AdoMet) was even more
dramatic (690- and 546-fold increases, respectively). In
contrast, K (NCA) and K (AdoMet) increased only 2- and
MD Simulations. Since the Y20 side chain interacts with
both NCA and AdoMet (Figure 2A,B), the increases in both
Km(NCA) and Km(AdoMet) for the Y20A mutant are not
surprising. However, D197 interacts only with NCA (Figure
2
A); the >500-fold increase in K (AdoMet) is not easily
m
predicted by the hNNMT-AdoHcy-NCA crystal structure. MD
simulations were carried out to explore the possible structural
consequences of the active site mutants. During the course of
m
m
5
ns simulations, the D197A conformations deviated more from
6
-fold for Y20F. Because of the differences in K consequences,
m
the initial crystal structure-based model than either wt or the
Y20A and Y20F mutants (Figure 3A). When structures at the
end of the simulation runs were compared to the starting
models, the largest conformational shifts were found to involve
k /K decreases only modestly for Y20F (5- and 11-fold) and
cat
m
much more significantly for Y20A (159- and 1847-fold) and
D197A (258- and 1259-fold).
7
804
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
to three water molecules in the binary complex structures (each
crystal structure contains four protein molecules in its
asymmetric unit). The surrounding protein residues are in
the same conformation whether NCA or solvent are bound.
While many structures of AdoMet-dependent methyltrans-
ferases have been reported, very few are of enzyme-AdoHcy/
AdoMet-acceptor substrate ternary complexes. For comparison
with hNNMT-AdoHcy-NCA, the most similar structure is that
25
of hPNMT-AdoHcy-NA (noradrenaline). Structural super-
position shows that AdoHcy binds to both hNNMT and
hPNMT in the same position and orientation (Figure 4C),
which is consistent with the high degree of structural and
sequence similarity for the two enzymes. The NCA and NA
binding modes are less similar: the two acceptor substrates
approach the AdoHcy from the same general direction, but
their binding sites only partially overlap (Figure 4C). When the
structures are superimposed, NCA bound to hNNMT sterically
clashes with hPNMT residue E219, which is a much shorter
S201 in hNNMT. Similarly, NA bound to hPNMT sterically
clashes with hNNMT residue Y242. This bulky tyrosine residue
is at the end of a β-strand in hNNMT that bulges out into
solvent in hPNMT due to a two-residue insertion. Thus, while
hNNMT and hPNMT have very conserved AdoHcy (and thus
presumably AdoMet) binding sites and overall protein folds,
their methyl substrate acceptor binding sites have diverged to
accommodate very different molecules.
After hPNMT, the next most similar ternary complex is
26
DXMT-AdoHcy-TB (theobromine). Although DXMT and
hNNMT are only 16% identical in sequence, they share a similar
core protein fold and bind AdoHcy (and thus presumably
AdoMet) in similar orientations (Figure 4D). TB is bound to
DXMT in two orientations, at a site which only partially overlaps
with the NCA binding site in hNNMT (Figure 4D). When the
structures are superimposed, NCA bound to hNNMT sterically
clashes with DXMT residues Y157 and I226. The homologous
residues in hNNMT either have different local protein main
chain conformation (L164) or are smaller (A198) so that they
do not interfere with NCA binding. TB bound to DXMT
sterically clashes with hNNMT residues Y24 and Y204. These
are located in nonconserved insertions which are different sizes
and conformations in DXMT. Comparison of the hNNMT,
hPNMT, and DXMT ternary complexes illustrates the high
degree of variation in the acceptor binding sites of AdoMet-
dependent methyltransferases that provides for substrate
diversity and specificity within a shared protein core scaffold.
Figure 3. MD simulations of hNNMT. (A) Conformational shifts
from starting coordinates for wt and mutant hNNMT structures
sampled during the 5 ns simulations, expressed as rmsd for main-chain
atoms. The D197A mutant deviates more from the starting structure
than the wt or two Y20 mutants. (B) Conformational differences
between starting and final coordinates for wt and mutant hNNMT
structures for 5 ns simulations, expressed as rmsf (root-mean-square
fluctuations) for main-chain atoms for each residue. The largest shifts,
for the V138-L154 region, are observed only for the D197A
simulations.
residues V138-L154 and in only the D197A simulation (Figure
3
B). This segment is distant from the D197A mutation site
The S 2-like reaction mechanism of AdoMet-dependent
N
(
Figure 2C), but residues within it interact with the adenine
methyltransferases involves the direct transfer of a methyl
group from AdoMet to the acceptor substrate. The hNNMT-
AdoHcy-NCA ternary complex reveals NCA nitrogen to AdoHcy
sulfur atom distances ranging from 3.5 to 4.2 Å. These N−S
distances are slightly longer than the sum of carbon−sulfur
(donor) and carbon−nitrogen (acceptor) single bonds (1.8 and
1.5 Å, respectively) and consistent with what would be expected
for an SN2-like intermediate with a linear arrangement of the N
(NCA nucleophile)−C (methyl)−S (AdoHcy leaving group)
atoms. The hNNMT AdoHcy−substrate distances are compara-
ble to those seen in the hPNMT and DXMT ternary complexes.
In the hPNMT structure, the AdoHcy sulfur to NA nitrogen
distance is slightly longer, at 5.5 Å (Figure 4C). In DXMT, TB
serves as both substrate and product; DXMT methylates two
different TB nitrogen atoms in two separate methylation reactions
that combine to produce caffeine. As a result, there are two
ring of AdoHcy and thus presumably would with AdoMet.
DISCUSSION
■
Structural comparison of the hNNMT-AdoHcy-NCA ternary
complex with the hNNMT-AdoHcy and mNNMT-AdoHcy
binary complexes reveals only one major conformational
difference: the hNNMT binary complex is missing residues
F27-H31 and is presumably disordered in this region. This loop
is ordered in both the hNNMT ternary complex and mNNMT
binary complex, albeit in different conformations (Figure 4B).
Other than this loop, the three NNMT structures are very
similar, even though they crystallized in three nonisomorphous
crystal forms. AdoHcy is bound in essentially the same position
and conformation and interacts with the same protein residues,
in all three structures. The NCA binding site is occupied by one
7
805
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
Figure 4. AdoHcy and acceptor substrate binding in hNNMT and other MTases. (A) AdoHcy and NCA bound to hNNMT. The four AdoHcy
bound to the four superimposed hNNMT have essentially identical conformations (green carbon atoms). NCA is bound similarly in chains A
(
orange carbon atoms, shown), B and C (not shown), and in a second slightly different orientation in chain D (light gray carbon atoms). Dashed
lines represent the possible path of methyl transfer between the AdoHcy sulfur atom and NCA acceptor nitrogen atom. (B) A surface loop
containing residues F27−H31 in the hNNMT-AdoHcy-NCA ternary complex (purple Cα trace) adopts a different conformation than the
homologous loops in the mNNMT-AdoHcy (light blue) and hINMT-AdoHcy (light pink) structures. The rest of the protein structures are shown as
light cyan backbone coils. (C) Comparison of hNNMT and hPNMT ligands. AdoHcy binds to hNNMT and hPNMT similarly. hNNMT-bound
NCA (chain A orientation) and hPNMT-bound NA (beige carbon atoms) approach their respective AdoHcy from the same direction, but bind in
very different orientations. Dashed lines show the path between the AdoHcy sulfur atom and acceptor nitrogen atoms in the two substrates. (D)
Comparison of hNNMT and DXMT ligands. AdoHcy binds to hNNMT and DXMT similarly. The DXMT TB substrate binds in two orientations
26
in the active site (ball-and-stick with dark and gray carbon atoms, PDB accession code 2EFJ), both of which are different from the orientation of
hNNMT-bound NCA. (E) Comparison of NNMT and PNMT active sites. Mutagenesis of D197 in NNMT (side chain with orange bonds involving
carbon atoms) and E185 in PNMT (side chain with beige bonds involving carbon atoms) substantially decrease activity in their respective enzymes.
(
F) The F27−H31 loop conformational divergence when comparing the hNNMT and hPNMT ternary complex structures is greater since an
adjacent segment is helical in hNNMT (purple) but unwound in hPNMT (beige).
different TB binding modes in the DXMT crystal structure that
position the two different nitrogen atoms to approach the
AdoHcy (Figure 4D) with TB nitrogen atom to AdoHcy sulfur
atom distances of 3.6 and 4.5 Å, very similar to those seen in the
hNNMT-AdoHcy-NCA complex structure. While NCA is bound
to two slightly different orientations in hNNMT, both position
7
806
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
the same nucleophilic ring nitrogen atom to approach AdoHcy.
The more common NCA orientation has a shorter NCA
nitrogen−AdoHcy sulfur distance (3.5−3.8 Å, chains A−C) while
the less common NCA orientation (chain D) has a longer NCA−
AdoHcy distance of 4.2 Å (Figure 4A); the difference between
these two distances is not very significant given an estimated
average coordinate error of 0.4 Å based on the correlation of Rfree
with data resolution for the refined NNMT-AdoHcy-NNMT
structure. The two NCA orientations differ by about 40°, but
both allow reasonable bond angles about the AdoMet sulfur atom
consistent with tetrahedral geometry when a methyl group is
modeled in line between the AdoHcy sulfur and NCA nitrogen.
Thus, based on AdoHcy−NCA distances and modeled AdoMet
sulfur geometry, both NCA orientations are likely to be
productive for catalysis.
Inspection of the hNNMT-AdoHcy-NCA structure identi-
fied several hydrophilic residues, conserved in NNMT
sequences (Figure 2B). The Y20 side-chain hydroxyl group
interacts with the AdoHcy carboxylate and the NCA ring; Y20F
and Y20A mutants were prepared to investigate the functional
importance of the hydroxyl group. The S201 and S213 hydroxyl
groups, and the D197 carboxylate, interact with the NCA
amide; S201A, S213A, and D197A mutants were investigated
to probe the functional role of these side chains. Of the five
mutants, enzyme activity was dramatically diminished for only
Y20A and D197A (Table 2). Y20F had only modestly
compromised activity, suggesting that Y20s importance lies in
its aromatic ring, perhaps by packing against and orienting
AdoMet, NCA, and surrounding hydrophobic residues, rather
than primarily in the electrostatic interactions involving its
hydroxyl. Subsequent enzyme kinetics analysis revealed that
kcat/Km was diminished only modestly for Y20F, and much
more substantially for Y20A and D197A mutants, due to
dramatic increases in Km for both NCA and AdoMet, and only
a modest decrease in kcat (Table 3). These results suggest that
both residues are important for binding for both methyl donor
and acceptor.
of 8.4 for the hPNMT NA substrate, and thus a substrate
deprotonation step is unlikely to be important in the NNMT
reaction. Since hNNMT D197 and hPNMT E185 are not
located in homologous positions in the amino acid sequence of
their respective proteins, their spatial convergence is striking
considering the different roles they appear to play in acceptor
substrate binding and enzyme mechanism.
Comparison of the hNNMT, hPNMT, and DXMT ternary
complex structures reveals the details of how these three enzymes
share a common core protein fold and yet can accommodate
binding of diverse acceptor substrates in different binding modes.
Comparison of these, and the hINMT-AdoHcy binary structure,
highlights a flexible region which may contribute to allowing
diverse substrates to bind to different AdoMet-dependent
MTases. In hNNMT, a surface loop containing residues F27−
H31 is disordered in the absence of NCA but ordered when the
substrate is bound. This loop is ordered in the mNNMT binary
complex but adopts a different conformation although all the
loop residues are conserved (Figure 4B). It is interesting to note
that the corresponding loop in the hINMT binary complex is
in a third conformation which is similar to that in the hNNMT
ternary complex, despite the low 20% loop sequence con-
servation. The hNNMT F27−H31 loop is in an insertion region
that is not part of the conserved AdoMet-dependent MTase core
fold. The corresponding loop region in hPNMT shows even
greater conformational divergence: one turn of the adjacent
conserved helix is unwound relative to hNNMT, and the longer
loop adopts a very different conformation (Figure 4F). This
surface loop is one of only three significant differences in protein
main-chain conformation between hNNMT and hPNMT ternary
complex structures: the other two are the previously mentioned β-
bulge that accommodates acceptor substrate binding to hPNMT
and the N-terminus which forms part of the cap over the AdoHcy-
bound portion of the active site. This surface loop does not
directly interact with acceptor substrates in the ternary complexes,
but it is located ∼10−15 Å away in the various structures and is
part of the nonconserved insertions which cover the active site.
In conclusion, the hNNMT-AdoHcy-NCA ternary complex
structure identifies active site residues which define the acceptor
substrate binding pocket; of these, Y20 and D197 were shown
by mutagenesis to be important for function. These amino acids
could not have been predicted based on the available ternary
complex crystal structures of the closest AdoMet-dependent
homologues, due to sequence divergence and location of the
residues in conformationally variable segments. Comparison of
the hNNMT, mNNMT, hINMT, and hPNMT crystal
structures highlights a surface loop (F27-H31 in hNNMT)
that adopts very different conformations in the presence and
absence of acceptor substrate and among the different enzymes.
The location of this loop near the active site suggests that this
region may play a role in substrate access and/or selectivity and
might be usefully exploited in future drug design efforts.
Since the D197 side chain interacts directly with NCA, the
effect on K (NCA) is expected; the cause for the increased
m
Km(AdoMet) is less clear. MD simulations revealed that the
D197A structures deviate more from the starting crystal-
structure-based model than either wt or Y20A/F mutants
Figure 3A). The largest conformational shifts were seen only
during the D197A simulation and involved residues V138-L154
Figure 3B). The hNNMT-AdoHcy-NCA crystal structure
shows that residues C141-T144 in this segment pack against
AdoHcy, and the V143 main chain nitrogen is within
H-bonding distance with the AdoHcy adenine (Figure 2C).
These long-range conformational shifts seen only in the D197A
MD simulation explain how the D197A substitution leads to a
(
(
large increase in K (AdoMet) despite not directly interacting
m
with the methyl donor. It is interesting to note that when the
hNNMT and hPNMT ternary complex structures are super-
imposed, the carboxylate groups of the hNNMT D197 and
hPNMT E185 occupy approximately the same position in
space (Figure 4E). In hNNMT, the D197 carboxylate directly
interacts with the NCA amide (distances range from 3.0 to
ASSOCIATED CONTENT
■
Accession Codes
Coordinates and structure factors have been deposited in the
Protein Data Bank with accession code 3ROD.
3
.9 Å with no nearby solvent molecules in chains A−D) while
AUTHOR INFORMATION
in hPNMT, the E185 carboxylate is a more distant 5.2 Å with
an intervening solvent molecule and is proposed to carry out
water-mediated deprotonation of its substrate.
■
Corresponding Author
*Phone: (216) 368-1184. Fax: (216) 368-3419. E-mail: vivien.
yee@case.edu.
25,27
The pKa for
the hNNMT substrate, NCA, at 3.3 is much lower than the pKa
7
807
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808

Biochemistry
Article
Funding
(12) Emsley, P., and Cowtan, K. (2004) Coot: model-building tools
for molecular graphics. Acta Crystallogr. D60, 2126−2132.
(13) Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L.,
Gros, P., Grosse-Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M.,
Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., and Warren, G. L.
1998) Crystallography & NMR system: A new software suite for
macromolecular structure determination. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 54, 905−921.
This work was supported by National Institute of Health grant
U01 GM61388 (Y.P., V.C.Y.). Argonne is operated by
UChicago Argonne, LLC, for the U.S. Department of Energy,
Office of Biological and Environmental Research under
Contract DE-AC02-06CH11357
(
ACKNOWLEDGMENTS
■
(14) Murshudov, G. N., Vagin, A. A., and Dodson, E. J. (1997)
We thank S. Ginell for beamline support. Diffraction data were
measured at the Structural Biology Center beamline 19-ID at
the Advanced Photon Source, Argonne National Laboratory.
Refinement of macromolecular structures by the maximum-likelihood
method. Acta Crystallogr. D53, 240−255.
(15) Laskowski, R. A., MacArthur, M. W., Moss, D. S., and
Thornton, J. M. (1993) PROCHECK: a program to check the
stereochemical quality of protein structures. J. Appl. Crystallogr. 26,
ABBREVIATIONS
■
2
83−291.
16) DeLano, W. L. (2004) The PyMOL Molecular Graphics System,
DeLano Scientific LLC, San Carlos, CA (www.pymol.org).
17) Balducci, E., Emanuelli, M., Raffaelli, N., Ruggieri, S., Amici, A.,
NNMT, nicotinamide N-methyltransferase; wt-hNNMT, wild-
type human NNMT; tm-hNNMT, hNNMT K100A/E101A/
E103A triple mutant; AdoMet, S-adenosyl-L-methionine;
AdoHcy, S-adenosyl-L-homocysteine; NCA, nicotinamide;
MTase, AdoMet-dependent methyltransferase; INMT, indole-
thylamine N-methyltransferase; PNMT, phenylethanolamine
N-methyltransferase; DXMT, 3,7-dimethylxanthine methyl-
transferase.
(
(
Magni, G., Orsomando, G., Polzonetti, V., and Natalini, P. (1995)
Assay methods for nicotinamide mononucleotide adenylyltransferase
of wide applicability. Anal. Biochem. 228, 64−68.
(18) van der Spoel, D., Lindahl, E., Hess, B., Groenhof, G., Mark, A.
E., and Berendsen, J. J. C. (2005) GROMACS: Fast, flexible, and free.
J. Comput. Chem. 26, 1701−1718.
(19) van Gunsteren, W. F., Billeter, S. R., Eising, A. A., Hunenberger,
P. H., Kruger, P., Mark, A. E., Scott, W. R. P., and Tironi, I. G. (1996)
Biomolecular Simulation: the GROMOS96 Manual and User Guide,
Hochschulverlag AG an der ETZ Zurich, Zurich.
REFERENCES
■
(
1) Aksoy, S., Szumlanski, C. L., and Weinshilboum, R. M. (1994)
Human liver nicotinamide N-methyltransferase. cDNA cloning,
expression, and biochemical characterization. J. Biol. Chem. 269,
1
4835−14840.
2) Cantoni, G. L. (1951) Methylation of nicotinamide with soluble
enzyme system from rat liver. J. Biol. Chem. 189, 203−216.
3) Williams, A. C., Cartwright, L. S., and Ramsden, D. B. (2005)
(20) Berendsen, H. J. C., Postma, J. P. M., van Gunsteren, W. F., and
Hermans, J. (1981) Interaction models for water in relation to protein
hydration, in Intermolecular Forces (Pullman, B., Ed.), pp 331−342, D.
Reidel Publishing Co., Dordrecht.
(
(
(21) Martin, J. L., and McMillan, F. M. (2002) SAM (dependent) I
Parkinson’s disease: the first common neurological disease due to
auto-intoxication? QJM 98, 215−226.
AM: the S-adenosylmethionine-dependent methyltransferase fold.
Curr. Opin. Struct. Biol. 12, 783−793.
(
4) Williams, A. C., and Ramsden, D. B. (2005) Autotoxicity,
(22) Schubert, H. L., Blumenthal, R. M., and Cheng, X. (2003) Many
methylation and a road to the prevention of Parkinson’s disease.
J. Clin. Neurosci. 12, 6−11.
paths to methyltransfer: a chronicle of convergence. Trends Biochem.
Sci. 28, 329−335.
(
5) Sartini, D., Muzzonigro, G., Milanese, G., Pierella, F., Rossi, V.,
(23) Kozbial, P. Z., and Mushegian, A. R. (2005) Natural history of
and Emanuelli, M. (2006) Identification of nicotinamide N-
methyltransferase as a novel tumor marker for renal clear cell
carcinoma. J. Urol. 176, 2248−2254.
S-adenosylmethionine-binding proteins. BMC Struct. Biol. 5, 19.
(24) Holm, L., and Sander, C. (1993) Protein structure comparison
by alignment of distance matrices. J. Mol. Biol. 233, 123−138.
(
6) Xu, J., Moatamed, F., Caldwell, J. S., Walker, J. R., Kraiem, Z.,
Taki, K., Brent, G. A., and Hershman, J. M. (2003) Enhanced
expression of nicotinamide N-methyltransferase in human papillary
thyroid carcinoma cells. J. Clin. Endocrimol. Metab. 88, 4990−4996.
(25) Drinkwater, N., Gee, C. L., Puri, M., Criscione, K. R., McLeish,
M. J., Grunewald, G. L., and Martin, J. L. (2009) Molecular
recognition of physiological substrate noradrenaline by the adrena-
line-synthesizing enzyme PNMT and factors influencing its methyl-
transferase activity. Biochem. J. 422, 463−471.
(26) McCarthy, A. A., and McCarthy, J. G. (2007) The structure of
two N-methyltransferases from the caffeine biosynthetic pathway.
Plant Physiol. 144, 879−889.
(
7) Roessler, M., Rollinger, W., Palme, S., Hagmann, M. L., Berndt,
P., Engel, A. M., Schneidinger, B., Pfeffer, M., Andres, H., Karl, J.,
Bodenmuller, H., Ruschoff, J., Henkel, T., Rohr, G., Rossol, S., Rosch,
̈
̈
̈
W., Langen, H., Zolg, W., and Tacke, M. (2005) Identification of
nicotinamide N-methyltransferase as a novel serum tumor marker for
colorectal cancer. Clin. Cancer Res. 11, 6550−6557.
(27) Gee, C. L., Tyndall, J. D. A., Grunewald, G. L., Wu, Q.,
(
8) Sartini, D., Santarelli, A., Rossi, V., Goteri, G., Rubini, C.,
McLeish, M. J., and Martin, J. L. (2005) Mode of binding of methyl
acceptor substrates to the adrenaline-synthesizing enzyme phenyl-
ethanolamine N-methyltransferase: implications for catalysis. Bio-
chemistry 44, 16875−16885.
Ciavarella, D., Lo Muzio, L., and Emanuelli, M. (2007) Nicotinamide
N-methyltransferase upregulation inversely correlates with lymph node
metastasis in oral squamous cell carcinoma. Mol. Med. 13, 415−421.
(
9) Tomida, M., Mikami, I., Takeuchi, S., Nishimura, H., and
Akiyama, H. (2009) Serum levels of nicotinamide N-methyltransferase
in patients with lung cancer. J. Cancer Res. Clin. Oncol. 135, 1223−
1
229.
(
10) Otwinowski, Z., and Minor, W. (1997) Processing of X-ray
Diffraction Data Collected in Oscillation Mode. Methods in
Enzymology (Carter, C. W., Jr., and Sweet, R. M., Eds.) Vol. 276,
pp 307−326, Academic Press, New York.
(
11) Vagin, A., and Teplyakov, A. (1997) MOLREP: an automated
program for molecular replacement. J. Appl. Crystallogr. 30, 1022−
025.
1
7
808
dx.doi.org/10.1021/bi2007614|Biochemistry 2011, 50, 7800−7808