X-ray crystal structure and binding mode analysis of human S-adenosylhomocysteine hydrolase complexed with novel mechanism-based inhibitors, haloneplanocin A analogues

Article abstract of DOI:10.1021/jm1010836

The X-ray crystal structure of human S-adenosylhomocysteine (AdoHcy) hydrolase was first determined as a tetrameric form bound with the novel mechanism-based inhibitor fluoroneplanocin A (4b). The crystallized enzyme complex showed the closed conformation

Full text of DOI:10.1021/jm1010836

930 J. Med. Chem. 2011, 54, 930–938
DOI: 10.1021/jm1010836
X-ray Crystal Structure and Binding Mode Analysis of Human S-Adenosylhomocysteine Hydrolase
Complexed with Novel Mechanism-Based Inhibitors, Haloneplanocin A Analogues†
Kang Man Lee,‡ Won Jun Choi,§, Yoonji Lee,‡ Hyun Joo Lee,‡ Long Xuan Zhao,‡,^ Hyuk Woo Lee,§ Jae Gyu Park,‡
Hea Ok Kim,‡ Kwang Yeon Hwang,# Yong-Seok Heo,¥ Sun Choi,*,‡ and Lak Shin Jeong*,‡,§
‡College of Pharmacy, and §Department of Bioinspired Science, Ewha Womans University, Seoul 120-750, Korea, College of Pharmacy,
Dongguk University, Kyungki-do 410-774, Korea, ^College of Chemistry and Chemical Engineering, Liaoning Normal University,
Dalian 116-029, China, #College of Life Sciences and Biotechnology, Korea University, Seoul 136-701, Korea, and ¥Department of Chemistry,
Konkuk University, Seoul 143-701, Korea
Received August 21, 2010
The X-ray crystal structure of human S-adenosylhomocysteine (AdoHcy) hydrolase was first deter-
mined as a tetrameric form bound with the novel mechanism-based inhibitor fluoroneplanocin A (4b).
The crystallized enzyme complex showed the closed conformation and turned out to be the intermediate
of mechanism-based inhibition. It confirmed that the cofactor depletion by 30-oxidation of fluorone-
planocin A contributes to the enzyme inhibition along with the irreversible covalent modification of
AdoHcy hydrolase. In addition, a series of haloneplanocin A analogues (4b-e and 5b-e) were designed
and synthesized to characterize the binding role and reactivity of the halogen substituents and the
40-CH2OH group. The biological evaluation and molecular modeling studies identified the key
pharmacophores and structural requirements for the inhibitor binding of AdoHcy hydrolase. The
inhibitory activity was decreased as the size of the halogen atom increased and/or if the 40-CH2OH group
was absent. These results could be utilized to design new therapeutic agents operating via AdoHcy
hydrolase inhibition.
Introduction
site, resulting in the permanent irreversible inhibition of the
enzyme.3 Thus, the catalytic activity of the enzyme is not
restored even after dialysis or incubation with NADþ.
S-Adenosylhomocysteine (AdoHcya) hydrolase1 is an im-
portant cellular enzyme catalyzing the hydrolysis of S-adeno-
sylhomocysteine to adenosine and L-homocysteine and an
attractive target for the development of broad-spectrum anti-
viral agents.2 The antiviral activity resulting from the inhibi-
tion of AdoHcy hydrolase is associated with the feedback
inhibition of S-adenosyl-L-methionine-dependent transmethy-
lase, which is essential for the formation of the capped
methylated structure at the 50-terminus of viral mRNA.
There have been AdoHcy hydrolase inhibitors synthesized
and evaluated for their activities, and two types of mechan-
ism-based inhibitors of AdoHcy hydrolase are reported.3
Type I mechanism-based inhibitors inactivate the enzyme by
converting the cofactor NADþ to NADH, resulting in the
depletion of NADþ.3 This type of inhibition is reversible on
incubation with an excess amount of NADþ or on dialysis.
Type II mechanism-based inhibitors not only deplete the
cofactor NADþ as type I inhibitors do but also bind cova-
lently with the nucleophilic amino acid residue at the active
NeplanocinA (X =CH2OH, Y = H) is a highly potent and
reversible inhibitor of AdoHcy hydrolase.4 It is oxidized to its
30-keto form by NADþ bound to AdoHcy hydrolase, and it
maintains the cofactor permanently in the reduced form (i.e.,
NADH),3,4 causing type I mechanism-based inhibition. We
recently designed and synthesized fluoroneplanocin A (X =
CH2OH, Y = F), containing a leaving group fluorine atom in
place of hydrogen at C60, and it turned out to be a more potent
inhibitor of AdoHcy hydrolase than neplanocin A.5 The
dialysis, incubation with NADþ, and 19F NMR experiments
indicated that fluoroneplanocin A works as an irreversible
mechanism-based inhibitor. Interestingly, it showed the bi-
phasic kinetic behavior, and 19F NMR experiments revealed
that the ratio of the released fluoride anion to the used enzyme
was not stoichiometric, indicating that the conversion of the
intermediate 1 to the final adduct 3 (Scheme 1) might not be
completely irreversible when the deprotonation of the inter-
mediate 2 results in the elimination of the enzyme instead of
the fluoride anion. Thus, it was hypothesized that the addi-
tional type I reversible inhibition by intermediate 1, alongwith
type II irreversible inhibition, may be involved in the mechan-
ism-based inhibition of AdoHcy hydrolase.
†The atomic coordinates and structure factors for the crystal struc-
ture of human AdoHcy hydrolase complexed with compound 4b have
been deposited in the Protein Data Bank with accession code 3NJ4.
*To whom correspondence should be addressed. For S.C.: phone, 82-2-
3277-4503; fax, 82-2-3277-2851; e-mail, sunchoi@ewha.ac.kr. For L.S.J.:
phone, 82-2-3277-3466; fax, 82-2-3277-2851; e-mail, lakjeong@ewha.ac.kr.
a Abbreviations: AdoHcy, S-adenosyl-L-homocysteine; NAD, nico-
tinamide adenine dinucleotide; NFSI, N-fluorobenzenesulfonimide;
GOLD, genetic algorithm of ligand docking; rmsd, root means square
deviation.
In order to confirm the additional type I reversible inhibition,
we obtained the X-ray crystal structure of human AdoHcy
hydrolase complexed with the fluoroneplanocin A. On the
basis of this X-ray crystal structure of human AdoHcy
hydrolase, a series of haloneplanocin A analogues 4 were
r
pubs.acs.org/jmc
Published on Web 01/12/2011
2011 American Chemical Society

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4 931
Scheme 1. Proposed Mechanism-Based Inhibition of AdoHcy Hydrolase by Haloneplanocin A Analogues
adenine, as shown in Scheme 2. D-Ribose was efficiently
converted to the key intermediates, cyclopentenone derivatives
6 (X=CH2OBn) and 7 (X=H), according to our published
procedures.7,8 Halogenation of cylcopentenone derivatives 6
(X=CH2OBn)8 and 7 (X=H)7 using Cl2, Br2, or I2 in pyridine
afforded the halogenated cyclopentenone derivatives 8 (X=
CH2OBn, Y = Cl, Br, or I) and 9 (X =H, Y = Cl, Br,
or I), respectively. For the introduction of fluorine atom on
the R position of R,β-unsaturated ketone, iodocyclopentenones
8 (X = CH2OBn, Y = I) and 9 (X = H, Y = I) were
stereoselectively reduced with NaBH4 to give the iodocyclo-
pentenols, which were treated with TBDPSCl to give the
TBDPS protected derivatives 10 (X=CH2OBn, Y = I) and
11 (X=H, Y=I), respectively. Treatment of 10 (X=CH2OBn,
Y=I) and 11 (X=H, Y=I) with N-fluorobenzenesulfonimide
(NFSI) in the presence of n-BuLi at -78 ꢀC yielded
R-fluorinated derivatives 12 (X=CH2OBn, Y=F)5 and 139
(X = H, Y = F), respectively, after the removal of TBDPS
group. For the synthesis of other halogen derivatives (Y=Cl,
Br, or I), compounds 8 (X = CH2OBn, Y = Cl, Br, or I) and 9
(X=H, Y=Cl, Br, or I) were reduced with NaBH4 to give other
halogenated alcohols 12 (X=CH2OBn, Y=Cl, Br, or I) and 13
(X = H, Y = Cl, Br, or I), respectively. The allylic alcohol
derivatives 12 (X=CH2OBn, Y=F, Cl, Br, or I) and 13 (X=H,
Y=F, Cl, Br, or I) were converted to the glycosyl donors 14
(X=CH2OBn, Y=F, Cl, Br, or I) and 15 (X=H, Y=F, Cl, Br,
or I), which were condensed with adenine anion in DMF to give
the condensed products 16 (X=CH2OBn, Y=F, Cl, Br, or I)
and 17 (X=H, Y=F, Cl, Br, or I). Removal of benzyl and
acetonide protecting groups of 16 (X=CH2OBn, Y=F, Cl, Br,
or I) with boron trichloride afforded the final haloneplanocin A
analogues 4 (X=CH2OH, Y=F, Cl, Br, or I). Treatment of 17
(X = H, Y = F, Cl, Br, or I) with 70% TFA afforded the
truncated haloneplanocin A analogues 5 (X=H, Y=F, Cl,
Br, or I).
Enzyme Assay. All the final nucleosides 4 and 5 were
assayed for the AdoHcy hydrolase inhibitory activities using
pure recombinant human placental AdoHcy hydrolase
obtained from E. coli JM109 containing the plasmid
pPROKcd20.5 Compounds 4 and 5 were preincubated
with the AdoHcy hydrolase at various concentrations for
5-10 min at 37 ꢀC, and the residual activity of the enzyme
was measured in the synthetic direction toward S-adenosyl-
homocysteine using adenosine and L-homocysteine.
Most of the synthesized compounds showed potent in-
hibitory activity against the AdoHcy hydrolase, as shown
in Table 1, and all these active compounds (4b-d and 5b)
inhibited the AdoHcy hydrolase in concentration- and time-
dependent manners (Figure S1 and S2 in the Supporting
Figure 1. Haloneplanocin A (4) and truncated haloneplanocin A
(5) designed for the identification of the pharmacophores of
AdoHcy hydrolase.
designed and synthesized to determine the effect of halogen in
binding at the active site of AdoHcy hydrolase (Figure 1).
A series of truncated haloneplanocin A analogues 5 were also
designed and synthesized to identify if the 40-CH2OH group
serves as the key pharmacophore in the active site of AdoHcy
hydrolase. We also performed the molecular docking study of the
synthesized nucleosides to the X-ray crystal structure of AdoHcy
hydrolase to study the correlation between the AdoHcy hydro-
lase inhibitory activity (IC50) and binding energy. Herein, we
report the full accounts of design, synthesis, and molecular
modeling studies of haloneplanocin A (4) and truncated halone-
planocin A (5) to reveal the key pharmacophores and structural
requirements for the inhibitor binding of AdoHcy hydrolase.
Results and Discussion
Chemistry. A series of halogenated neplanocin A analo-
gues 4 (X = CH2OH, Y = F, Cl, Br, or I)6 were designed and
synthesized to characterize the binding role and reactivity
of the halogen substituents. The corresponding 40-CH2OH
truncated analogues 5 (X = H, Y = H, F, Cl, Br, or I)6 were
also prepared to find out the binding role of the 40-CH2OH
group at the active site of AdoHcy hydrolase.
Our synthetic strategy to haloneplanocin A analogues 4 and
truncated haloneplanocin A analogues 5 was to synthesize
the glycosyl donors 14 and 15 and then to condense them with

932 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4
Lee et al.
Scheme 2. Synthesis of the Final Nucleosides 4 and 5 from D-Ribose
Table 1. AdoHcy Hydrolase Inhibitory Activity of Neplanocin A (4a),
Haloneplanocin A Analogues 4b-e, Truncated Neplanocin A (5a),
and Truncated Haloneplanocin A Analogues 5b-e
compounds. The truncated neplanocin A (5a) and its fluori-
nated compound (5b) have low micromolar range of IC50
values, but chloro, bromo, and iodo derivatives (5c, 5d, and
5e) are unlikely to bind at all. Since the 40-CH2OH group
seems to be one of the key pharmacophores for the binding
with AdoHcy hydrolase, the substituent increasing a hydro-
gen bonding capacity may improve the inhibitory activity
against AdoHcy hydrolase. This assumption was confirmed
by the X-ray crystal structure of AdoHcy hydrolase com-
plexed with fluoroneplanocin A (5b) in the following part.
The irreversibility of inactivation for compounds 4c-e
and 5c-e was confirmed by dialysis as in the case of 4b5 and
5b.9 The level of inhibition of AdoHcy hydrolase obtained
after preincubation of the enzyme at various inhibitory con-
centrations of each compound remained unchanged after
dialysis (Figure S3 in the Supporting Information).
X-ray Crystal Structure of AdoHcy Hydrolase Complexed
with Fluoroneplanocin A. In order to understand the binding
mode of the inhibitor at the molecular level, human AdoHcy
hydrolase was crystallized with the most potent inhibitor,
fluoroneplanocin A, along with the cofactor NADþ. This is
the first reported X-ray crystal structure of human AdoHcy
hydrolase as a tetrameric form (Figure 2a).
Four subunits are connected around the pseudo-222 sym-
metry to form a tetramer, where each monomer is occupied
by the cofactor and the ligand. The monomeric subunit
of AdoHcy hydrolase is composed of the catalytic domain,
the cofactor-binding domain, and the C-terminal domain
(Figure 2b). The catalytic and cofactor-binding domains
are linked by the R-helix, which might serve as a molecular
hinge. The final refined X-ray crystal structure showed
that it formed a strong ternary complex of the inhibitor,
AdoHcy hydrolase, and NADH. Interestingly, the crystal-
lized enzyme-ligand complex turned out to be the inter-
mediate of the inhibition, bound with fluoroneplanocin A
oxidized in its 30-position (i.e., 1 in Scheme 1) by the cofactor
NADþ, and its electron density map is as shown in Figure 2c.
Taken together with our previous report that fluoronepla-
nocin A inhibits AdoHcy hydrolase irreversibly and its
fluorine atom is extracted after the incubation with the
enzyme,5 it might function by dual inhibition: the cofactor
depletion and covalent modification of the enzyme.
compd (X, Y)
IC50 (μM)
compd (X, Y)
IC50 (μM)
4a (X=CH2OH, Y = H)
4b (X=CH2OH, Y = F)
4c (X=CH2OH, Y = Cl)
0.87
0.48
36.46
5a (X=H, Y=H)
5b (X=H, Y=F)
5c (X=H, Y=Cl)
5.83
7.67
>1000
4d (X=CH2OH, Y = Br) 60.17
4e (X=CH2OH, Y = I)
5d (X=H, Y=Br) >1000
>1000
>1000 5e (X=H, Y=I)
Information). Fluoroneplanocin A (4b) exhibited the most
potent enzyme inhibitory activity (IC50 = 480 nM), which
is ∼2-fold more potent than neplanocin A (4a). Other
haloneplanocin A analogues were designed with better leav-
ing groups (i.e., Cl, Br, and I) and expected to have higher
activities because the carbon-halogen bonds of other halo-
neplanocin A analogues could be cleaved more easily than
the carbon-fluorine bond during the mechanism-based
inhibition of AdoHcy hydrolase if they bind at the active
site. Surprisingly, the activities of the halo analogues tend
to be gradually diminished as the size of halogen atom
increased in the order of Cl, Br, and I. Especially, the iodo
derivative (4e) showed that it is less likely to bind to AdoHcy
hydrolase despite having the best leaving group. Since the
biological activity of this series of mechanism-based inhibi-
tors could be the net effect of how well the ligand binds and
how reactive the bound ligand is, these results indicated that
the proper ligand binding seems to be more critical than the
chemical reactivity for the activity in this series of inhibitors.
In order to determine whether the 40-CH2OH group is
important for binding, the truncated haloneplanocin A 5b-e
were synthesized and tested. They showed much lower poten-
cy compared with the corresponding 40-CH2OH-containing

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4 933
Figure 2. X-ray crystal structure of AdoHcy hydrolase complexed with fluoroneplanocin A. (a) Homotetrameric topology of human AdoHcy
hydrolase. Each monomer is shown in the secondary structure colored by purple, cyan, orange, and green, with the water molecules displayed as
balls. The ligand and the cofactor molecules are represented in spacefill, and their carbon atoms are in white and green, respectively. (b)
Monomeric subunit consisting of the catalytic domain, the cofactor-binding domain, and the C-terminal domain. (c) σA-weighted 2Fo - Fc
electron density map calculated with a final refined model of the structure without fluoroneplanocin A and NADþ at a level of 1.2σ. The carbon
atoms of fluoronecplanocin A and NADH are in white and green, respectively, and the fluorine atom is in black. (d) Fluoroneplanocin A with
the key binding residues at the enzyme active site. The ligand is depicted as ball-and-stick with carbon atoms in white. Its van der Waals surface
was generated by MOLCAD and colored by hydrogen bonding property (red, H-bond donating regions; blue, H-bond accepting regions). (e)
Schematic diagram of the binding interactions of fluoroneplanocin A at the active site of AdoHcy hydrolase. Interacting residues in the
cofactor-binding domain and the catalytic domain are represented in purple and magenta, respectively. The hydrophobic interactions are
colored in orange, and the hydrogen bondings are displayed as dashed lines.
The binding interactions of fluoroneplanocin A at the active
site of AdoHcy hydrolase is as shown in Figure 2d. Its overall
binding mode is retained as that of neplanocin A,10 including
the hydrogen bondings with His55, Thr57, Glu59, Asp131,
Glu156, Thr157, Lys186, Asp190, His301, Met351, and
His353. In addition, there is a hydrophobic interaction between
the fluorine atom and the active site hydrophobic pocket,
contributed by the two Leu residues (i.e., Leu344 and Leu347).
It might be one of the reasons why fluoroneplanocin A is more
potent than neplanocin A. In the dimer structure, the cofactor-
binding domain of a monomer has an interface interacting with
the R-helix of its dimer partner. Binding of the cofactor NADþ
to AdoHcy hydrolase requires the contributions from both
dimer subunits, unlike NAD-dependent dehydrogenases where
each cofactor is typically bound in one monomer exclusively.11
Also, the C-terminal domains are exchanged between the
adjacent dimer partners and accommodate part of the cofac-
tor-binding sites.12 Overall, a dimer of the enzyme is required to
form a complete cofactor-binding site.11
structure. Besides, the catalytic domains are connected with
the cofactor-binding domains via hinge regions. They are
located on the periphery of the tetramer and have little
interaction with one another. So, architecturally, the tetra-
meric structure indicates that the catalytic domains are more
mobile compared with the cofactor-binding domains. In
fact, it was reported that the average temperature factor of
the catalytic domain is significantly higher than that of the
cofactor-binding domain,12 and our human AdoHcy hydro-
lase X-ray crystal structure also confirmed this finding.
Overall, the tetramer formation appears to allow the enzyme
to build a strong and rigid molecular framework as well as
flexible catalytic sites.
Regarding the conformation of AdoHcy hydrolase, there
are two distinct forms: open and closed. In the resting state,
the catalytic and cofactor-binding domains are opened
through 19ꢀ to expose the active site to the environment.10
Upon ligand binding, AdoHcy hydrolase makes a substan-
tial conformational change. The ligand seems to bind to the
catalytic domain first and then move along with the relatively
mobile catalytic domain to acquire the interactions with the
cofactor-binding domain as depicted in Figure 2e.
The dimer itself, however, is not sufficient to establish a
strong rigid core framework. Therefore, AdoHcy hydrolase
is assembled into a tetrameric form.12 The four cofactor-
binding domains are tightly linked together at the center
of the tetramer and make it possible to build up the core
As a result, the enzyme forms the closed conformation.
Compared with our open form homology model (Figure 3a),

934 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4
Lee et al.
the protonation state of the ionizable groups was chosen
appropriate to pH 7.0. According to Yamada et al.,15 His54
residue in rat appears to be protonated in its NE2 position
and the protonation state of this residue is less influenced by
the pH of solution. So His55 residue was prepared to be
protonated for the molecular modeling studies. Then the
flexible docking was performed using the GOLD (genetic
algorithm of ligand docking) program, version 4.1. To con-
sider the protein flexibility, 10 residues interacting with the
ligand at the active site were selected and set to be flexible.
Each inhibitor structure was prepared as the oxidized form in
its 30-position and docked into the protein structure in the
presence of NADH. The GOLD fitness scores of the docked
molecules had good correlation with their biological activ-
ities. Among the docking results, the reasonable complexes
with high GOLD fitness scores were selected and energy
minimized with the conjugate gradient algorithm using
CHARMm program. After the energy minimization, con-
formations of the docked ligands did not change so much but
the whole systems were relaxed to the more stable structures.
The docked and energy-minimized conformations of the
analogues were shown in Figure 4.
The adenine part of the analogues was docked almost
exactly the same way, maintaining the bidentate hydrogen
bonding with Thr57 and Glu59. However, the position of the
cyclopentenyl ring was prone to be pushed away from the
hydrophobic pocket, which could accommodate the halogen
atom, as the size of the halogen atom increased. The van der
Figure 3. Comparison of the open and closed forms of human
AdoHcy hydrolase. (a) Homology model of the open form human
AdoHcy hydrolase. (b) X-ray crystal structure of the closed form
human AdoHcy hydrolase with fluoroneplanocin A bound be-
tween the catalytic and cofactor-binding domains. Connolly sur-
face was generated and Z-clipped from both the front and back
sides (in the upper view) and from the front side only (in the bottom
view). In the open form structure, the catalytic domain is colored
by cyan and the cofactor-binding and C-terminal domains are in
sky blue. In the closed form structure, they are in magenta and
purple, respectively. Fluoroneplanocin A is displayed in spacefill
and the cofactor is in capped-stick with their carbon colors of white
and green-blue, respectively. The nonpolar hydrogen atoms are
not displayed for clarity.
˚
Waals radius of hydrogen is 1.20 A, and those of halogen
˚
˚
˚
atoms are 1.47 A for fluorine, 1.75 A for chlorine, 1.85 A for
bromine, and 1.98 A for iodine.16 The surface representation
˚
of each complex clearly showed the conformational adjust-
ment of the compounds to fit into the active site cavity
(Figure 4a). In the case of iodoneplanocin A, the bulkiness
of the iodine atom seemed to cause too much of shifting for
the cyclopentenyl ring and lose the hydrogen bonding with
His301 (Figure 4b). These results demonstrate that there is a
spatial limitation of the small and rigid hydrophobic pocket
where the halogen atoms could be positioned. Thus, unlike
fluoroneplanocin A (4b), iodoneplanocin A (4e) appears not
to maintain the appropriate conformation for binding at the
active site and shows no activity in spite of having the most
reactive leaving group in this series. Therefore, the ligand
binding might be the more important factor than the chemi-
cal reactivity for the inhibitory activity of this series of
haloneplanocin A analogues. Also, the 40-CH2OH group
played a crucial role in the binding of neplanocin A and
haloneplanocin A analogues through the key hydrogen
bondings with the active site His301. Consequently, the 40-
CH2OH truncated analogues lost the activity to 1 order of
magnitude or more compared with their counterparts
(Table 1).
In addition, the binding energy was calculated for each
ligand-enzyme complex. For the calculation of the relative
binding energy of a ligand, it is necessary to consider the
structural changes between the holoenzyme in the open con-
formation and the ligand-bound enzyme in its closed con-
formation. The binding energies of neplanocin A and halo-
neplanocin A analogues were calculated using the docked
and energy-minimized systems for the ligand-bound struc-
tures and the refined homology model for the holoenzyme
structure. The total energy of each system was calculated
using CHARMm energy function and the implicit solvent
model, generalized Born with molecular volume (GBMV).
the crystal structure of human AdoHcy hydrolase complexed
with fluoroneplanocin A has the closed conformation, mak-
ing the ligand completely embraced between the two do-
mains, and forms the strong ternary complex of the inhibitor,
AdoHcy hydrolase, and NADH (Figure 3b).
In addition, our X-ray crystal structure of human AdoHcy
hydrolase was compared with those of other species. Their
sequence identities and rmsd values were measured, and the
overall structures appeared to be very similar (Table S1 and
Figure S4 in the Supporting Information). The only differ-
ence is the additionally inserted helix-loop-helix structure,
which was reported to be present within the catalytic domain
in most of bacteria and some eukaryotes but not in mammals.13
The active site residues of AdoHcy hydrolases are highly
conserved, and their 3D structures are very similar (Figure
S5 in the Supporting Information). It would be challenging
to achieve the ligand selectivity among different species, but
it was reported that introducing the functional group at the
C2 position of adenine ring could achieve the selectivity due
to the difference of Thr60 in human and Cys59 in Plasmo-
dium falciparum AdoHcy hydrolases14 (Figure S6 in the
Supporting Information).
Molecular Modeling Analysis. To investigate the binding
modes of the series of haloneplanocin A analogues, flexible
docking studies were performed. While the X-ray crystal
structure was prepared, all hydrogen atoms were added and

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4 935
Figure 4. Binding modes of neplanocin A (4a) and haloneplanocin A analogues 4b-e. (a) Flexible docked and energy-minimized con-
formations of the analogues at the enzyme active site. (b) Comparison of the ligand conformations aligned by the ligand-bound enzyme
complexes (with His301 displayed). Neplanocin A (4a) and fluoroneplanocin A (4b) are displayed in capped-stick with its carbon color of sky
blue and in ball-and-stick with its carbon color of white, respectively. Chloro- and bromoneplanocin A (4c and 4d) are depicted in capped-stick,
and their carbons are in green and brown, respectively. Iodoneplanocin A (4e) is shown in ball-and-stick with its carbon color of purple. His301,
Leu344, and Leu347 residues in each complex are displayed in capped-stick with the carbons in gray. The distances between His301 (ND1) and
0
˚
the 5 -oxygens of fluoroneplanocin A and iodo-neplanocin A are 2.8 and 6.5 A, respectively. The nonpolar hydrogen atoms, except the
60-hydrogen of neplanocin A (4a), are not displayed for clarity. The fast Connolly surface of the protein and the van der Waals surface of the
ligands were generated by MOLCAD. The protein surface is colored by lipophilicity and Z-clipped, and each ligand surface is in its carbon
color.
Table 2. Calculated Average Binding Energies of Neplanocin A (4a)
and Haloneplanocin A Analogues 4b-e
very stable when these ligands are bound to the enzyme. On
the other hand, the binding energy of iodo derivative is very
compd
X
IC50 (μM)
binding energy (kcal/mol)
high, indicating that the enzyme complex with this ligand
would be energetically unstable.
4a
4b
4c
4d
4e
H
F
0.87
0.48
-53.24
-65.03
-43.23
-30.83
-8.75
Cl
Br
I
36.46
60.17
>1000
Conclusion
The X-ray crystal structure of human AdoHcy hydrolase,
complexed with the novel mechanism-based inhibitor fluoro-
neplanocin A, was first determined, and it confirmed the
additional type I reversible mechanism. Considering the irre-
versible nature of inhibition by fluoroneplanocin A, it was
elucidated that the enzyme inhibition of fluoroneplanocin A is
involved in the dual mechanism: the cofactor depletion and
covalent modification of the enzyme. In addition, the crystal-
lized protein-ligand complex shows that the inhibitor induces
the open-to-closed conformational change. It makes the ligand
possible to bind tightly between the cofactor-binding and
For more efficient calculation, binding energies were esti-
mated using the following formula:
binding energy ¼ Ecomplex - Eholoenzyme - Efree ligand ð1Þ
The binding energies of the tested compounds showed good
correlation with their IC50 values; as the binding energies
increased, the IC50 values also increased (Table 2).
For neplanocin A and fluoroneplanocin A, the binding
energies are very low and it means that the complexes are

936 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4
Lee et al.
catalytic domains and form the strong ternary complex of
the inhibitor, AdoHcy hydrolase, and NADH, allowing the
mechanism-based inhibition reaction to occur. Moreover, we
identified the key pharmacophores and structural requirements
for inhibitor binding through the biological evaluation and
molecular modeling studies of the synthesized haloneplanocin
A and the corresponding truncated analogues. This kind of
comprehensive study, based on chemistry, structural biology,
and molecular modeling, gave insight for the inhibitor design,
and these results will be utilized for the discovery of new
therapeutic agents.
343 (M þ Hþ); 1H NMR (CD3 OD) δ 8.18 (s, 1 H), 8.15 (s, 1 H),
5.55 (dt, J = 1.4, 6.0 Hz, 1 H), 4.83 (s, 1 H), 4.70 (t, J = 6.0 Hz,
1 H), 4.40 (d, J = 13.2 Hz, 1 H), 4.26 (dd, J = 2.8, 13.2 Hz, 1 H);
13C NMR (DMSO-d6) δ 156.0, 152.5, 149.8, 145.8, 140.4, 119.3,
119.1, 74.1, 70.9, 67.5, 57.0. Anal. Calcd for C11H12BrN5O3: C,
38.61; H, 3.54; N, 20.47. Found: C, 38.66; H, 3.58; N, 20.54.
5-(6-Aminopurin-9-yl)-4-iodo-3-hydroxymethylcyclopent-3-ene-
1,2-diol (4e, X=CH2OH, Y=I):17 yield = 62%; MS (FAB) m/z
390 (M þ Hþ); mp 168.2-172.5 ꢀC; [R]2D5 -100.6ꢀ ( c 0.83,
MeOH); UV (MeOH) λmax 260.0 nm; 1H NMR (CD3OD) δ 8.42
(s, 1 H), 8.41 (s, 1 H), 5.67 (d, J=5.0 Hz, 1 H), 4.85 (d, J = 6.0
Hz, 1 H), 4.69 (t, J = 6.0 Hz, 1 H), 4.34 (brd, J=13.0 Hz, 1 H),
4.29 (d, J=13.0 Hz, 1 H); 13C NMR (CD3OD) δ 154.5, 151.9,
150.8, 145.4, 145.1, 120.5, 98.2, 77.1, 73.6, 73.2, 62.3. Anal.
Calcd for C11H12IN5O3: C, 33.95; H, 3.11; N, 18.00. Found: C,
33.83; H, 3.18; N, 17.87.
Experimental Section
General Methods. 1H NMR spectra (CDCl3, CD3OD, or
DMSO-d6) were recorded on Varian Unity Invoa 400 MHz.
The 1H NMR data are reported as peak multiplicities: s for
singlet, d for doublet, dd for doublet of doublets, t for triplet, q
for quartet, br s for broad singlet, and m for multiplet. Coupling
constants are reported in hertz. 13C NMR (CDCl3, CD3OD, or
DMSO-d6) spectra were recorded on Varian Unity Inova
100 MHz. 19F NMR (CDCl3, CD3OD) were recorded on Varian
Unity Inova 376 MHz. The chemical shifts were reported as
parts per million (δ) relative to the solvent peak. Optical rotations
were determined on Jasco III in the appropriate solvent. UV
spectra were recorded on U-3000 made by Hitachi in methanol
or water. Infrared spectra were recorded on FT-IR (FTS-135)
made by Bio-Rad. Melting points were measured on B-540
made by Buchi. Elemental analyses (C, H, and N) were used
to determine the purity of all synthesized compounds, and the
results were within (0.4% of the calculated values, confirming
g95% purity. Reactions were checked with TLC (Merck pre-
coated 60F254 plates). Spots were detected by viewing under a
UV light, colorizing with charring after dipping in anisaldehyde
solution with acetic acid, sulfuric acid, and methanol. Column
chromatography was performed on silica gel 60 (230-400 mesh,
Merck). Reagents were purchased from Aldrich Chemical Co.
Solvents were obtained from local suppliers. All the anhydrous
solvents were distilled over CaH2, P2O5, or sodium/benzophe-
none prior to the reaction.
5-(6-Aminopurin-9-yl)-4-fluorocyclopent-3-ene-1,2-diol (5b, X =
H, Y = F):9 yield = 92%; MS (FAB) m/z 252 (M þ Hþ);
[R]2D5 -172.4ꢀ ( c 0.77, MeOH); UV (MeOH) λmax 259.0 nm; 1H
NMR (CD3OD) δ 8.19 (s, 1 H), 8.18 (s, 1 H), 5.63 (m, 1 H), 5.60
(m, 1 H), 4.75 (m, 1 H), 4.68 (td, J = 1.2, 6.0 Hz, 1 H); 13C NMR
(CD3OD) δ 162.1, 159.3, 157.4, 153.8, 151.2, 142.1, 120.7, 110.1
(d, J = 4.9 Hz), 76.6 (d, J = 4.2 Hz), 70.7 (d, J = 10.4 Hz), 62.9
(d, J = 18.2 Hz); 19F NMR (CD3OD) δ -121.6. Anal. Calcd for
C10H10FN5O2: C, 47.81; H, 4.01; N, 27.88. Found: C, 47.96; H,
4.08; N, 28.04.
5-(6-Aminopurin-9-yl)-4-chlorocyclopent-3-ene-1,2-diol (5c,
X = H, Y = Cl): yield = 83%; MS (FAB) m/z 269 (M þ Hþ);
UV (MeOH) λmax 260.0 nm; MS (FAB) m/z 268 (M þ Hþ); 1H
NMR (DMSO-d6) δ 8.22 (s, 1 H), 8.11 (s, 1 H), 7.29 (br s,
2 H), 6.28 (t, J = 2.4 Hz, 1 H), 5.44 (m, 2 H), 5.26 (d, J = 6.0 Hz,
1 H), 4.67 (dd, J = 5.6, 6.0 Hz, 1 H), 4.57 (m, 1 H); 13C NMR
(DMSO-d6) δ 156.7, 156.6, 153.2, 150.4, 141.3, 134.7, 132.6,
75.2, 71.3, 66.7. Anal. Calcd for C10H10ClN5O2: C, 44.87; H,
3.77; N, 26.16. Found: C, 44.71; H, 3.95; N, 26.03.
5-(6-Aminopurin-9-yl)-4-bromocyclopent-3-ene-1,2-diol (5d,
X = H, Y = Br): yield = 99%; MS (FAB) m/z 313 (M þ Hþ);
mp 115.1-120.4 ꢀC; [R]2D5 -118.7ꢀ ( c 1.20, MeOH); UV
1
(MeOH) λmax 260.0 nm; H NMR (CD3OD) δ 8.18 (s, 1 H),
8.16 (s, 1 H), 6.47 (dd, J = 2.0, 2.8 Hz, 1 H), 5.55 (ddd, J = 1.6,
2.0, 6.0 Hz, 1 H), 4.76 (t, J = 6.0 Hz, 1 H), 4.70 (ddd, J = 1.6,
2.8, 6.0 Hz, 1 H); 13C NMR (CD3OD) δ 157.4, 153.8, 151.1,
142.2, 137.7, 126.6, 120.6, 76.7, 73.7, 69.82. Anal. Calcd for
C10H10BrN5O2: C, 38.48; H, 3.23; N, 22.44. Found: C, 38.61; H,
3.34; N, 22.46.
Synthesis. See the Supporting Information for full experi-
mental procedure.
5-(6-Aminopurin-9-yl)-4-fluoro-3-hydroxymethylcyclopent-3-
ene-1,2-diol (4b, X = CH2OH, Y = F):5 yield = 77%; MS
(FAB) m/z 282 (M þ Hþ); mp 181-184 ꢀC; [R]2D5 -181.1ꢀ (c 0.62,
MeOH); UV (H2O) λmax 260.0 nm; 1H NMR (CD3OD) δ 8.17
(s, 1 H), 8.15 (s, 1 H), 5.56 (br s, 1 H), 4.79 (pseudo t, J = 5.2, 6.0
Hz, 1 H), 4.56 (t, J = 5.6 Hz, 1 H), 4.41 (d, J = 13.2 Hz, 1 H),
4.16 (td, J = 2.4, 13.2 Hz, 1 H);13C NMR (CD3OD) δ 157.7, 154.7
(d, J = 285.3 Hz), 154.2, 151.2, 142.1, 122.8 (d, J = 2.3 Hz), 120.8,
75.7 (d, J = 4.6 Hz), 71.2 (d, J = 8.5 Hz), 63.6 (d, J = 18.4 Hz),
54.7; 19F NMR (CD3OD) δ -133.1. Anal. Calcd for
C11H12FN5O3: C, 46.98; H, 4.30; N, 24.90. Found: C, 46.99;
H, 4.28; N, 25.20.
5-(6-Aminopurin-9-yl)-4-iodocyclopent-3-ene-1,2-diol (5e,
X=H, Y=I): yield = 96%; MS (FAB) m/z 360 (M þ Hþ);
mp 119.4-124.8 ꢀC; [R]2D5 -55.6ꢀ ( c 1.30, MeOH); UV (MeOH)
λ
max 260.5 nm; 1H NMR (CD3OD) δ 8.18 (s, 1 H), 8.15 (s, 1 H),
6.72 (dd, J = 2.0, 2.8 Hz, 1 H), 5.52 (ddd, J = 12, 2.0, 6.0 Hz,
1 H), 4.75 (t, J = 6.0 Hz, 1 H), 4.65 (ddd, J = 1.2, 2,8, 6.0 Hz,
1 H); 13C NMR (CD3OD) δ 157.3, 153.5, 151.0, 146.0, 142.3,
120.6, 101.9, 76.2, 75.1, 72.4. Anal. Calcd for C10H10IN5O2: C,
33.44; H, 2.81; N, 19.50. Found: C, 33.56; H, 2.99; N, 19.42.
Enzyme Assay. Materials. S-Adenosyl-L-homocysteine
(AdoHcy) and homocysteine thiolactone were purchased
from Sigma. Co. Adenosine was obtained from BDH
Chemical Co. Dr. Michael S. Hershfield (Duke Univ.) kindly
provided the plasmic pPROKcd20 containing cloned gene of
human placental AdoHcy hydrolase. E. coli JM109 was
obtained from KTCC (Korea Type Culture Collection).
Purification of AdoHcy Hydrolase. The cloned human pla-
cental AdoHcy hydrolase was purified from the cell-free extracts
of E. coli JM109 containing the plasmid pPROKcd20 grown in
2xYT medium with IPTG induction.18 The cell-free extracts
were obtained by sonication of a cell suspension of E. coli JM109
containing pPROKcd20 in 50 mM Tris-HCl (pH 7.5) containing
2 mM EDTA.
5-(6-Aminopurin-9-yl)-4-chloro-3-hydroxymethylcyclopent-3-
ene-1,2-diol (4c, X = CH2OH, Y = Cl): yield = 64%; MS
(FAB) m/z 299 (M þ Hþ); mp 250.7-251.0 ꢀC; [R]2D5 -20.0ꢀ
(c 1.14, DMSO); UV (MeOH) λmax 265.0 nm; MS (FAB) m/z
299 (M þ Hþ); 1H NMR (DMSO-d6) δ 8.18 (s, 1 H), 8.12
(s, 1 H), 7.25 (s, 1 H), 5.42 (d, J = 5.6 Hz, 1 H), 5.40 (d, J = 6.4 Hz,
1 H), 5.21 (d, J = 5.6Hz, 1 H), 5.05 (t, J = 5.6 Hz, 1 H), 4.63-4.60
(m, 2 H), 4.21 (d, J = 13.2Hz, 1 H), 4.06 (d, J = 13.2Hz, 1 H); 13
C
NMR (DMSO-d6) δ156.6, 153.2, 150.5, 143.0, 141.1, 127.9, 119.7,
74.4, 70.9, 67.0, 55.7. Anal. Calcd for C11H12ClN5O3: C, 44.38; H,
4.06; N, 23.53. Found: C, 44.40; H, 4.08; N, 23.40.
5-(6-Aminopurin-9-yl)-4-bromo-3-hydroxymethyl-cyclopent-3-
ene-1,2-diol (4d, X = CH2OH, Y = Br): yield = 87%; MS
(FAB) m/z 343 (M þ Hþ); mp 240.1-241.8 ꢀC; [R]2D5 -14.2ꢀ
( c 1.26, MeOH); UV (MeOH) λmax 260.5 nm; MS (FAB) m/z
The purification was carried out through DEAE-cellu-
lose column (2.8 cm ꢀ 6 cm), ammonium sulfate fractionation

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4 937
(35-60%), Sephacryl S-300HR (1.0 cm ꢀ 105 cm), and DEAE-
cellulose (2.8 cm ꢀ 24 cm). The protein homogeneity was
confirmed by 10% SDS-PAGE. The protein concentration was
determined by using Bradford method.19 Bovine serum albumin
was a standard material for protein assay.
hydrogen atoms and charges appropriate to pH 7.0, and
His55 residue was set to be protonated in its NE2 position. Then
the flexible docking study was performed using GOLD, version
4.1 (Cambridge Crystallographic Data Centre, U.K.), in
˚
the defined binding site at 10 A from the crystal ligand. Ten
Assay of AdoHcy Hydrolase for Synthetic Direction. The
enzyme was incubated with 0.2 mM Ado and 5 mM Hcy in
500 μL of 50 mM potassium phosphate buffer (pH 7.2) contain-
ing 1 mM EDTA (buffer A) at 37 ꢀC for 5 min. The reaction
was terminated by the addition of 25 μL of 5 N HClO4. The
terminated reaction mixture was kept in ice for 5 min and
microcenrifuged. The supernatant was analyzed for AdoHcy by
HPLC equipped C-18 reversed-phase column (Econosphere
C18, 5 μm, 250 mm ꢀ 4.6 mm, Alltech, Deerfield, IL). The elution
was carried out at a flow rate of 1 mL/min in two sequential linear
gradients: 6-15% A over 15 min, 15-50% A over the next
5 min, where mobile phase A was acetonitrile and B was 50 mM
sodium phosphate buffer (NaH2PO4/H3PO4), pH 3.2 (HPLC
buffer B). The peak of AdoHcy was monitored at 258 nm. The
concentration was determined by the peak area of AdoHcy.
Concentration-Dependent Inactivation of AdoHcy Hydrolase
(Figure S1). Various concentrations of each inhibitor were
incubated with AdoHcy hydrolase (400 nM) in buffer A at
37 ꢀC for 5 min, and the remaining activity was determined in the
synthetic direction as described above.
Time- and Concentration-Dependent Inactivations of AdoHcy
Hydrolase (Figure S2). Various concentrations of active com-
pounds (4b-d and 5b) were incubated with AdoHcy hydrolase
(184 nM) in buffer A at 37 ꢀC for different amounts of time (5, 8,
or 10 min), and the remaining activity was determined in the
synthetic direction as described above.
Determination of Irreversibility of AdoHcy Hydrolase Activity
(Figure S3). The enzyme (400 nM) was preincubated with active
compounds (4b-d and 5b) at 37 ꢀC for 5 min. It was dialyzed
through Centricon 30 up to 90 fold. During dialysis, AdoHcy
hydrolase activity was determined in the synthetic direction as
described above.
residues in the binding site were selected and set to be flexible
with “crystal mode”, defining a rotamer in which all rotatable
torsions in the side chain are allowed to vary over the range (δ -
χ) to (δ þ χ), where χ values are taken from the protein input file.
The selected 10 residues were His55, Glu156, His301, Leu344,
Asn346, and Leu347 with δ of 20 and Asp131, Thr157, Lys186,
and Asp190 with δ of 10. Other options were set to default
except GA (genetic algorithm) runs of 30. Among the resulting
docked conformations, reasonable complexes with high GOLD
fitness scores were selected and energy minimized using
CHARMm program, implemented in Discovery Studio, version
2.5. By application of harmonic restraints with a force constant
˚
of 10 to backbone atoms of the residues beyond 5 A from the
ligand and heavy atoms of the cofactor, the system was mini-
mized with the conjugate gradient algorithm until the rms
˚
gradient tolerance reached 0.01 kcal/(mol A) (max steps of
10 000).
3
Homology Modeling. Homology model of the open form
human AdoHcy hydrolase was built by MODELER implemen-
ted in Discovery Studio, version 2.5, using the open form struc-
ture of rat AdoHcy hydrolase (PDB code 1B3R)12 as a template.
Among the 30 models generated with the high optimization
level, the initial model was selected by the lowest probability
density function (PDF) total energy and refined using molecular
dynamics (MD) simulation and energy minimization. MD
simulation was conducted via standard dynamics cascade pro-
tocol in Discovery Studio, version 2.5, followed by 500 steps
of minimization with steepest descent (rms gradient of 0.1
˚
kcal/(mol A)), 500 steps of minimization with conjugate gradi-
3
˚
ent (rms gradient of 0.0001 kcal/(mol A)), 2000 steps of heating,
3
4000 steps of equilibration, and finally 10 000 steps of produc-
tion phase with an NVE ensemble at 300 K. During the simula-
tion, an integration time step of 0.001 ps (1 fs) was used, and
harmonic restraints with force constant of 10 were applied to the
backbone of the identical residues and heavy atoms of the
cofactor. In addition, harmonic restraints with a force constant
of 5 were also applied to side chain heavy atoms of the identical
residues located in the interface during heating, equilibration,
and production phase. The resulting conformation in final tra-
jectory was further energy minimized with the conjugate gra-
dient algorithm (max steps of 10 000; rms gradient of 0.01
Crystallization, X-ray Data Collection, and Structure Deter-
mination. For crystallization, we added fluoroneplanocin A and
NADþ to the concentrated protein solution to give a fluorone-
planocin A/NADþ/protein molar ratio of 10:5:1 and incubated
the mixture for 1 h at 277 K. The incubated protein was cry-
stallized at 20 ꢀC by vapor diffusion using the hanging-drop
method and a protein to reservoir solution ratio of 1:1, with the
reservoir solution containing 0.1 M lithium sulfate, 0.1 M
HEPES (pH 8.0), and 20% PEG400. Prior to X-ray data
collection, crystals were transferred to the cryoprotectant solu-
tion containing the reservoir solution plus 20% (v/v) glycerol for
a few seconds, then looped from the drop and flash-frozen in
liquid nitrogen. The X-ray diffraction data for the crystal were
collected at the 6B beamline of Pohang Light Source (PLS) and
processed using HKL2000.20 The crystal belongs to the space
˚
kcal/(mol A)), and the final refined model was validated by
the Verify Protein (Profiles-3D) module in Discovery Studio,
version 2.5.
3
Binding Energy Calculation. The potential energies of each
complex and the refined homology model were calculated, and
the relative binding energy was estimated by eq 1. Energy values
of the complexes and the holoenzyme were obtained from
the docked and minimized systems and the refined homology
model, respectively. For free ligands that are not bound in
the enzyme, their potential energies were calculated using con-
jugate gradient with max steps of 1 000 000 and rms gradient of
˚
group P21, and the unit cell parameters are a = 91.201 A, b =
˚
˚
75.80 A, c = 189.01 A, and β = 107.01ꢀ.
The structure was solved by molecular replacement (MR)
using the program CNS21 with the previously reported structure
(PDB code 1A7A)11 as a search model. Using the data with a
˚
0.01 kcal/(mol A).
˚
3
resolution range of 50.0-2.5 A, the MR solution structure was
All the computational studies were performed on a Linux
(Cent OS, release 4.6) workstation. During the minimization
and MD simulations, the CHARMm force field and implicit
solvent model, generalized Born with molecular volume
(GBMV), were used and the bond lengths involving bonds to
hydrogen atoms were constrained by the SHAKE algorithm.
The resulting figures were generated in the SYBYL molecular
modeling program, version 8.1.1 (Tripos International, St.
Louis, MO, U.S.).
refined with the CNS package and manual model building was
performed using the program Coot.22 During refinement, 5% of
reflection data chosen randomly from the observed data were
used for cross-validation with the Rfree value. The coordinates
and structure factors have been deposited in PDB (http://www.
rcsb.org/pdb) under the accession code of 3NJ4. The statistics
for data collection and structure refinement are given in Table
S2 in the Supporting Information.
Molecular Modeling. X-ray Crystal Structure Preparation
and Flexible Docking. X-ray crystal structure was prepared with
clean module in Discovery Studio, version 2.5 package
(Accelrys, San Diego, CA, U.S.) including the addition of
Acknowledgment. This work was supported by grants
from the WCU program (Grant R31-2008-000-10010-0) and

938 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 4
Lee et al.
the National Core Research Center (NCRC) program (Grant
R15-2006-020) of the National Research Foundation (NRF)
of Korea. H.W.L. acknowledges a Research Professor Fel-
lowship from Ewha Womans University (2009).
(9) Kim, H. O.; Yoo, S. J.; Ahn, H. S.; Choi, W. J.; Moon, H. R.; Lee,
K. M.; Chun, M. W.; Jeong, L. S. Synthesis of fluorinated cyclo-
pentenyladenine as potent inhibitor of S-adenosylhomocysteine
hydrolase. Bioorg. Med. Chem. Lett. 2004, 14, 2091–2093.
(10) Yang, X. D.; Hu, Y. B.; Yin, D. H.; Turner, M. A.; Wang, M.;
Borchardt, R. T.; Howell, P. L.; Kuczera, K.; Schowen, R. L.
Catalytic strategy of S-adenosyl-L-homocysteine hydrolase: transi-
tion-state stabilization and the avoidance of abortive reactions.
Biochemistry 2003, 42, 1900–1909.
(11) Turner, M. A.; Yuan, C. S.; Borchardt, R. T.; Hershfield, M. S.;
Smith, G. D.; Howell, P. L. Structure determination of seleno-
methionyl S-adenosylhomocysteine hydrolase using data at a
single wavelength. Nat. Struct. Biol. 1998, 5, 369–376.
Supporting Information Available: Additional experimental
procedures and characterization data for all unknown com-
pounds; X-ray crystallographic results; enzyme assay results.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(12) Hu, Y. B.; Komoto, J.; Huang, Y.; Gomi, T.; Ogawa, H.; Takata,
Y.; Fujioka, M.; Takusagawa, F. Crystal structure of S-adenosyl-
homocysteine hydrolase from rat liver. Biochemistry 1999, 38,
8323–8333.
(13) (a) Stepkowski, T.; Brzezinski, K.; Legocki, A. B.; Jaskolski, M.;
Bena, G. Bayesian phylogenetic analysis reveals two-domain to-
pology of S-adenosylhomocysteine hydrolase protein sequences.
Mol. Phylogenet. Evol. 2005, 34, 15–28. (b) Reddy, M. C. M.; Kuppan,
G.; Shetty, N. D.; Owen, J. L.; Ioerger, T. R.; Sacchettini, J. C. Crystal
structures of Mycobacterium tuberculosis S-adenosyl-L-homocysteine
hydrolase in ternary complex with substrate and inhibitors. Protein Sci.
2008, 17, 2134–2144.
(14) Tanaka, N.; Nakanishi, M.; Kusakabe, Y.; Shiraiwa, K.; Yabe, S.;
Ito, Y.; Kitade, Y.; Nakamura, K. T. Crystal structure of S-
adenosyl-L-homocysteine hydrolase from the human malaria para-
site Plasmodium falciparum. J. Mol. Biol. 2004, 343, 1007–1017.
(15) Yamada, T.; Takata, Y.; Komoto, J.; Gomi, T.; Ogawa, H.;
Fujioka, M.; Takusagawa, F. Catalytic mechanism of S-adeno-
sylhomocysteine hydrolase: roles of His 54, Asp130, Glu155,
Lys185, and Asp189. Int. J. Biochem. Cell Biol. 2005, 37, 2417–
2435.
(16) Rowland, R. S.; Taylor, R. Intermolecular nonbonded contact
distances in organic crystal structures: comparison with distances
expected from van der Waals radii. J. Phys. Chem. 1996, 100, 7384–
7391.
References
(1) (a) Turner, M. A.; Yang, X.; Yin, D.; Kuczera, K.; Borchardt,
R. T.; Howell, P. L. Structure and function of S-adenosylhomo-
cysteine hydrolase. Cell Biochem. Biophys. 2000, 33, 101–125.
(b) Cantoni, G. L. The Centrality of S-Adenosylhomocysteinase in
the Regulation of the Biological Utilization of S-Adenosylmethionine.
In Biological Methylation and Drug Design; Borchardt, R. T., Crevel-
ing, C. R., Ueland, P. M., Eds.; Humana Press: Clifton, NJ, 1986;
pp 227-238.
(2) (a) Snoeck, R.; Andrei, G.; Neyts, J.; Schols, D.; Cools, M.;
Balzarini, J.; De Clercq, E. Inhibitory activity of S-adenosylho-
mocysteine hydrolase inhibitors against human cytomegalovirus
replication. Antiviral Res. 1993, 21, 197–216. (b) Liu, S.; Wolfe,
M. S.; Borchardt, R. T. Rational approaches to the design of antiviral
agents based on S-adenosyl-L-homocysteine hydrolase as a molecular
target. Antiviral Res. 1992, 19, 247–265.
(3) Wolfe, M. S.; Borchardt, R. T. S-Adenosyl-L-homocysteine hydro-
lase as a target for antiviral chemotherapy. J. Med. Chem. 1991, 34,
1521–1530.
(4) Keller, B. T.; Borchardt, R. T. Metabolism and Mechanism of
Action of Neplanocin A: A Potent Inhibitor of S-Adenosylhomo-
cysteine Hydrolase. In Biological Methylation and Drug Design;
Borchardt, R. T., Creveling, C. R., Ueland, P. M., Eds.; Humana Press:
Clifton, NJ, 1986; pp 385-396.
(5) Jeong, L. S.; Yoo, S. J.; Lee, K. M.; Koo, M. J.; Choi, W. J.; Kim,
H. O.; Moon, H. R.; Lee, M. Y.; Park, J. G.; Lee, S. K.; Chun,
M. W. Design, synthesis, and biological evaluation of fluorone-
planocin A as the novel mechanism-based inhibitor of S-adeno-
sylhomocysteine hydrolase. J. Med. Chem. 2003, 46, 201–203.
(6) The preliminary accounts were published in the Proceedings of XV
International Roundtable, Leuven, Belgium, September 10-14,
2002. (a) Jeong, L. S.; Moon, H. R.; Park, J. G.; Shin, D. H.; Choi,
W. J.; Lee, K. M.; Kim, H. O.; Chun, M. W.; Kim, H. D.; Kim, J. H.
Synthesis and biological evaluation of halo-neplanocin A as novel
mechanism-based inhibitors of S-adenosylhomocysteine hydro-
lase. Nucleosides, Nucleotides Nucleic Acids 2003, 22, 589–592. (b)
Jeong, L. S.; Park, J. G.; Choi, W. J.; Moon, H. R.; Lee, K. M.; Kim,
H. O.; Kim, H. D.; Chun, M. W.; Park, H. Y.; Kim, K.; Sheen, Y. Y.
Synthesis of halogenated 9-(dihydroxycyclopent-40-enyl) adenines and
their inhibitory activities against S-adenosylhomocysteine hydrolase.
Nucleosides, Nucleotides Nucleic Acids 2003, 22, 919–921.
(7) Moon, H. R.; Choi, W. J.; Kim, H. O.; Jeong, L. S. Preparative
synthesis of the key intermediate, (4R,5R)-3-benzyloxymethyl-4,5-
isopropylidenedioxycyclopent-2-enone for carbocyclic nucleo-
sides. Chem. Lett. 2004, 33, 506–507.
(17) Park, A.-Y.; Kim, K. R.; Lee, H.-R.; Kang, J.-A.; Kim, W. H.;
Chun, P.; Ahmad, P.; Jeong, L. S.; Moon, H. R. Design and
synthesis of 500-iodoneplanocin A and its analogues as potential
S-adenosylhomocysteine hydrolase inhibitor. Bull. Korean Chem.
Soc. 2008, 29, 2487–2490.
(18) Yuan, C. S.; Yeh, J.; Liu, S.; Borchardt, R. T. Mechanism of
inactivation of S-adenosylhomocysteine hydrolase by (Z)-40,50-
didehydro-50-deoxy-50-fluoroadenosine. J. Biol. Chem. 1993, 268,
17030–17037.
(19) Bradford, M. M. A rapid and sensitive method for the quantitation
of microgram quantities of protein utilizing the principle of pro-
tein-dye binding. Anal. Biochem. 1976, 72, 248–254.
(20) Otwinowski, Z.; Minor, W. Processing of X-ray Diffraction Data
Collected in Oscillation Mode. In Macromolecular Crystallogra-
phy. Part A; Carter, C. W., Sweet, R. M., Eds.; Methods in Enzymol-
ogy, Vol. 276; Academic Press: San Diego, CA, 1997; pp 307-326.
(21) 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.; Warren, G. L.
Crystallography & NMR system: a new software suite for macro-
molecular structure determination. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 1998, 54, 905–921.
(8) Moon, H. R.; Choi, W. J.; Kim, H. O.; Jeong, L. S. Improved and
alternative synthesis of D- and L-cyclopentenone derivatives, the
versatile intermediates for the synthesis of carbocyclic nucleosides.
Tetrahedron: Asymmetry 2002, 13, 1189–1193.
(22) Emsley, P.; Cowtan, K. Coot: model-building tools for molecular
graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004, 60,
2126–2132.