Binding mode prediction of aplysiatoxin, a potent agonist of protein kinase C, through molecular simulation and structure–activity study on simplified analogs of the receptor-recognition domain

Article abstract of DOI:10.1016/j.bmc.2016.07.011

Aplysiatoxin (ATX) is a naturally occurring tumor promoter isolated from a sea hare and cyanobacteria. ATX binds to, and activates, protein kinase C (PKC) isozymes and shows anti-proliferative activity against human cancer cell lines. Recently, ATX has attracted attention as a lead compound for the development of novel anticancer drugs. In order to predict the binding mode between ATX and protein kinase Cδ (PKCδ) C1B domain, we carried out molecular docking simulation, atomistic molecular dynamics simulation in phospholipid membrane environment, and structure–activity study on a simple acyclic analog of ATX. These studies provided the binding model where the carbonyl group at position 27, the hydroxyl group at position 30, and the phenolic hydroxyl group at position 20 of ATX were involved in intermolecular hydrogen bonding with the PKCδ C1B domain, which would be useful for the rational design of ATX derivatives as anticancer lead compounds.

Full text of DOI:10.1016/j.bmc.2016.07.011

Bioorganic & Medicinal Chemistry xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Binding mode prediction of aplysiatoxin, a potent agonist of protein
kinase C, through molecular simulation and structure–activity study
on simplified analogs of the receptor-recognition domain
c
b
d
Yoshiki Ashida a, Ryo C. Yanagita b, , Chise Takahashi , Yasuhiro Kawanami , Kazuhiro Irie
⇑
a Department of Applied Bioresource Science, The United Graduate School of Agricultural Sciences, Ehime University (Kagawa University), Kagawa 761-0795, Japan
b Department of Applied Biological Science, Faculty of Agriculture, Kagawa University, Kagawa 761-0795, Japan
c Division of Applied Bioresource Science, Graduate School of Agriculture, Kagawa University, Kagawa 761-0795, Japan
d Division of Food Science and Biotechnology, Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Aplysiatoxin (ATX) is a naturally occurring tumor promoter isolated from a sea hare and cyanobacteria.
ATX binds to, and activates, protein kinase C (PKC) isozymes and shows anti-proliferative activity against
human cancer cell lines. Recently, ATX has attracted attention as a lead compound for the development of
novel anticancer drugs. In order to predict the binding mode between ATX and protein kinase Cd (PKCd)
C1B domain, we carried out molecular docking simulation, atomistic molecular dynamics simulation in
phospholipid membrane environment, and structure–activity study on a simple acyclic analog of ATX.
These studies provided the binding model where the carbonyl group at position 27, the hydroxyl group
at position 30, and the phenolic hydroxyl group at position 20 of ATX were involved in intermolecular
hydrogen bonding with the PKCd C1B domain, which would be useful for the rational design of ATX
derivatives as anticancer lead compounds.
Received 13 June 2016
Revised 5 July 2016
Accepted 6 July 2016
Available online xxxx
Keywords:
Aplysiatoxin
Tumor promoter
Protein kinase C
Molecular dynamics simulation
Structure–activity study
Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction
isozymes.8 While ATX and debromoaplysiatoxin (DAT) showed
both anti-proliferative and tumor-promoting activities, their sim-
Aplysiatoxin (ATX, Fig. 1) is a polyacetate isolated from sea hare
Stylocheilus longicauda1 and cyanobacteria.2 ATX showed potent
tumor-promoting activity in vivo,3 but also showed anti-prolifera-
tive activity against several human cancer cell lines.4 The main cel-
lular target of ATX is protein kinase C (PKC),5 which is a serine/
threonine kinase family that plays important roles in cellular signal
transduction such as proliferation, differentiation, and apoptosis.6
ATX binds to tandem C1 domains (C1A and C1B) of eight isozymes
plified analogs showed anti-proliferative activity comparable to
DAT but showed little tumor-promoting activity.9 Knowledge
about the precise binding mode of ATX and PKC C1 domains is
needed to rationally design ATX derivatives as anticancer drugs.
Although several groups have reported comparisons of pharma-
cophoric elements between ATX and other PKC ligands,10,11 a pre-
cise binding mode of ATX with the PKC C1 domains has not yet
been proposed.
of conventional PKC (
a
, bI, bII,
c
) and novel PKC (d,
e
,
g
, h), result-
Although X-ray crystallography and NMR-based methods are
‘gold standards’ for analysis of protein–ligand interaction and
binding mode, it is difficult to apply these methods to the PKC
ligand/C1 domain system because holo-C1 domain exists as a tern-
ary complex of protein/ligand/phospholipid bilayer membrane.12
To overcome this experimental limitation, we in the present study
carried out molecular docking simulation, molecular dynamics
simulation of ATX/C1 domain complex in phospholipid membrane
environment, and structure–activity study on the receptor recogni-
tion domain of ATX.
ing in translocation of these isozymes to the cellular membrane
fraction to activate the enzymes (Fig. 2).7
Recently, ATX has attracted attention as a lead compound for
the development of novel anticancer drugs targeting PKC
Abbreviations: ATX, aplysiatoxin; PKC, protein kinase C; DAT, debromoaplysi-
atoxin; DAG, 1,2-diacyl-sn-glycerol; RD, recognition domain; MM, molecular
mechanics; DFT, density functional theory; MD, molecular dynamics; POPS, 1-
palmitoyl-2-oleoyl-phosphatidylserine; Kd, dissociation constant; Ki, binding inhi-
bition constant.
⇑
Corresponding author. Tel./fax: +81 87 891 3092.
E-mail address: charlesy@ag.kagawa-u.ac.jp (R.C. Yanagita).
http://dx.doi.org/10.1016/j.bmc.2016.07.011
0968-0896/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

2
Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
Figure 1. Structure of aplysiatoxins.
2. Results and discussion
2.1. Conformational search of ATX
ATX is composed of a 14-membered bis-macrolactone ring and
a side chain containing an aromatic ring. In terms of receptor
recognition and conformational control, the structure of ATX can
be divided into three regions (Fig. 3):13 the recognition domain
(RD) at positions 1 and 27–31, the conformation-controlling unit
from position 2 to 11, and the side chain at position 11. RD contains
two ester groups and a hydroxyl group that could play some roles
in the receptor recognition in a manner similar to those of an
endogenous second messenger 1,2-diacyl-sn-glycerol (DAG).10
The conformation-controlling unit contains a [6,6]-spiroketal moi-
ety and is thought to keep RD in its active conformation.10 There-
fore, ATX can be viewed as a conformation-constrained analog of
flexible DAG, whereas the direction of ester linkages at position
27 of ATX and sn-1 position of DAG are opposite. Finally, the side
chain may play roles similar to those of hydrophobic part of natu-
rally-occurring PKC activators such as phorbol esters (Fig. 3) and
teleocidins.11
Although some protein–ligand docking programs can take into
account flexibility of cyclic parts of a molecule in an indirect
way, it may dramatically increase the number of rotatable bonds
and computational complexity. Thus, prior to docking simulation,
we performed conformational search for the macrolactone part of
ATX.
Conformational search was carried out using simulated anneal-
ing (SA) method to generate 1000 geometries. The side chain at
position 11 was replaced with an isopropyl group to simplify the
calculation. The CHARMM general force field (CGenFF) was
employed to calculate potential energy of the system. Among the
Figure 3. Structures and pharmacophoric elements of ATX, DAG, TPA, and
bryostatin-1.
structures generated, 351 out of 1000 took chair–chair [6,6]-
spiroketal conformation as with the experimental structures in
solution and crystal.14 They were classified into six groups (A–F)
based on the lactone ring conformation. Table 1 lists size and aver-
age molecular mechanics (MM) energy of each group. While con-
formers B and C were identical to RD-I and RD-II conformers
calculated by Knust and Hoffmann for 3-deoxyaplysiatoxin,13
respectively, conformers A, D, E, and F were newly identified ones.
Conformer A with the lowest energy was conformationally close to
conformer B. Conformational difference between them was only
orientation of the carbonyl group at position 27; the dihedral angle
difference at C9–O9 was approximately 40 degrees (Fig. 4).
Although average MM energies for conformers C and D were
within 5 kcal/mol from that of conformer A, the 30-OH group in
these conformers formed intramolecular hydrogen bonding with
other parts of the molecule, lowering the potential energy of the
conformers. However, this 30-OH group was predicted to form
hydrogen bondings with the PKCd C1B domain. Therefore, we con-
sidered that binding of conformers C and D to the PKC C1 domain is
energetically unfavorable. Conformers E and F were poor in terms
of either group size or energy. Since conformer B was almost iden-
tical to the crystal conformation (Fig. 4), we postulated that con-
formers
A and/or B are active conformers. To estimate the
relative stabilities of and energy barrier between conformers A
and B, we performed density functional theory (DFT) calculations.
The molecular geometry optimization of these conformers and
Table 1
Conformational analysis of macrocyclic core of ATX
Conformer No. in
cluster
Average MM energy
Relative MM energy
(kcal molÀ1
)
(kcal molÀ1
)
A
B
C
D
E
F
57
167
63
55
7
À127.61
À125.21
À123.44
À122.75
À122.41
À104.05
0
2.40
4.17
4.86
5.20
23.56
1
Figure 2. Activation mechanism of protein kinase C (PKC).
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
3
nated structures. ATX was described by the CGenFF force filed
(version 3.0.1), which could reproduce conformational energy
landscape of ATX relatively well. During the 10 ns of the produc-
tion run, root mean square deviation (RMSD) of backbone atoms
of the protein and all atoms of the ligand from the initial structure
remained at values below 2 and 1 Å (Supplemental Fig. S1), respec-
tively, suggesting adequacy of the stability of the model.
Figure 5 shows an overview of the system (Fig. 5A), interactions
between ATX/PKCd C1B domain and the phospholipids (Fig. 5B),
and the representative binding mode between ATX and the recep-
tor (Fig. 5C). ATX and the hydrophobic residues forming the bind-
ing cleft were located in a hydrophobic core of the PS membrane
and were fully dehydrated (Fig. 5A and B).
In the binding model (Fig. 5C), four hydrogen bonds were
formed between ATX and d-C1B: In addition to three hydrogen
bonds observed in the docking simulation, a hydrogen bond
between phenolic 20-OH group of ATX and C@O group of Met-
239 was observed. This result is consistent with a report that 20-
O-methyldebromoaplysiatoxin was less active in induction of
mouse-ear irritation, induction of ornithine decarboxylase (ODC),
and the inhibition of specific 12-O-tetradeconylphorbol 13-acetate
(TPA) binding than in debromoaplysiatoxin (DAT).21 The C@O
group of Met-239 was also predicted to form a hydrogen bond with
9-OH group of bryostatin-1.22 The average distances among these
hydrogen-bonding pairs, rAH, during the MD simulation are listed
in Table 2. These distances can be categorized as moderate (1.5–
2.2 Å) and weak (2.2–3.2 Å).23 The standard deviation for these dis-
tances suggests that these hydrogen bonds were stably maintained
during the simulation.
Figure 4. Possible conformers of a macrocyclic core of ATX and their relative energy
at the
B97X-D/6-31G⁄ level.
x
intermediate conformers between them was performed at the
B97X-D/6-31G⁄ level of theory. Figure 4 shows optimized geome-
x
tries and relative energies of conformers A and B. The difference in
energy between conformers A and B was 0.331 kcalÁmolÀ1, and
energy barrier was only 0.602 kcalÁmolÀ1. The small energy differ-
ence and barrier indicate that these conformers readily intercon-
vert with each other at room temperature and that the
orientation of the carbonyl group at position 27 is not restricted
to its local minima, conformers A and B.
During the MD simulation, two conformations for the side chain
at position 11 of ATX were observed, where dihedral angles for
C11–C15 were T–T–T–GÀ or T–G+–T–G+ (T, trans; G, gauche;
Fig. 6), though the hydrogen bond between the 20-OH of ATX
and Met-239 was maintained in both conformations.
2.2. Docking simulation of ATX with PKCd C1B domain
Since the DFT calculation indicated that the difference in
energy between conformers A and B is very small, we performed
docking simulation of the both conformers with a crystal struc-
ture of PKCd C1B domain (PDB code: 1PTR)15 using the
AutoDock16 program (version 4.2.6). The macrolactone ring of
ATX was treated as rigid, and the other parts of ATX and side
chain of Leu-254 of the receptor were treated as flexible. From
two hundred docking results for each conformer, binding modes
where 30-OH group of the ligand forms hydrogen bonds with
Thr-242 and Leu-251 of the receptor were selected as valid bind-
ing modes because the 30-OH group of ATX was postulated to
play the same role as 20-OH group of phorbol esters that formed
hydrogen bonds with the PKCd C1B domain in the crystal struc-
ture (Fig. 3).15 Three docking results were selected for both con-
formers, in all of which the carbonyl group at position 27 formed
a hydrogen bond with the N–H group of Gly-253. This binding
mode corresponds to the binding mode of DAG predicted by Mar-
quez et al.17 and Li et al.,18 where the sn-1 carbonyl group of DAG
formed a hydrogen bond with the Gly-253.
2.4. Molecular electrostatic potential calculation of ligand/PKCd
C1B domain complex
The MD simulation provided clear insights into the role of the
20-OH, 27-C@O, and 30-OH groups of ATX in the PKC binding.
However, the ester group at position 1 was not involved in any
intermolecular hydrogen bond with the receptor and its role
remained unclear, despite its presence in three major classes of
PKC activators: aplysiatoxins, DAG, and bryostatins (Fig. 3). In
order to predict the role of the ester group at position 1, we
focused on molecular electrostatic potentials of ATX and the
receptor. Electrostatic potential complementarity as well as shape
complementarity is a key determinant of molecular recognition.24
Figure 7A shows electrostatic potential surface of the PKCd C1B
domain alone or in complex with phorbol 13-acetate (PDB code:
1PTR). A ligand binding cleft of the PKCd C1B domain was char-
acterized by a marked positive potential of the back wall and the
bottom as well as neutral to negative potential of the left frontal
part. As shown in Figure 7B, ATX had an electrostatic potential
surface complementary to the binding cleft. The polarization of
the ester group at position 1 of ATX could contribute to this com-
plementary electrostatic potential. Moreover, we found that
phorbol 13-acetate had a potential surface similar to that of
ATX (Fig. 7C), supporting the predicted binding mode of ATX with
the PKCd C1B domain.
2.3. Molecular dynamics simulation of ATX/PKCd C1B domain
complex in phospholipid membrane environment
Next, we carried out molecular dynamics (MD) simulation of
ATX/PKCd C1B domain complex in phospholipid membrane envi-
ronment to refine the binding model from the docking simulation.
At first, ATX/PKCd C1B domain complex was inserted into 1-palmi-
toyl-2-oleoyl-phosphatidylserine (POPS) bilayer membrane, then
the system was equilibrated followed by 10 ns of production MD
simulation. We chose phosphatidylserine (PS) as a lipid component
because binding of the PKC ligands to the C1 domains shows speci-
ficity for PS.19 The PKCd C1B domain and the lipids were described
by the charmm36 force field20 with modification for zinc-coordi-
2.5. Synthesis and PKC binding of acyclic analogs of the
recognition domain of ATX
Although the effect of modification of 20-OH and 30-OH groups
of ATX on the PKC binding have been studied previously, the role of
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

4
Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
Figure 5. Structures of a ternary complex of POPS membrane/ATX/PKCd C1B domain obtained from the MD simulation. (A) Overall structure of the POPS membrane/ATX/
PKCd C1B domain complex in water after the 10 ns of production run. Protein is depicted in surface; ATX is in spheres colored green (carbon), red (oxygen), firebrick red
(bromine), and white (hydrogen); phosphorous atoms of the lipids are in spheres (orange); water layer is depicted as surface (blue). (B) Interaction between POPS and ATX/
PKCd C1B domain complex. ATX is in spheres and POPS are in stick model. The carbon atoms of POPS from upper and lower leaflet were colored by cyan and slate,
respectively. (C) Front views of structure of ATX/PKCd C1B domain complex. ATX is in the stick model. The residues near the ligand are represented in the stick model colored
purple (carbon), red (oxygen), blue (nitrogen), and white (polar hydrogen) with the labels of their residue names, while other residues are shown only in backbone. Black
dashed lines represent hydrogen bondings.
Table 2
the two carbonyl groups in the recognition domain (RD) has not
been experimentally examined. Thus, we examined the role of
Hydrogen-bonding parameters derived from MD simulation of the ATX/PKCd C1B
complex
the carbonyl groups of ATX by employing ester-to-ethylene (–CH2-
H-bond donors
H-bond acceptors
rAH (Å)a
CH2–) and ester-to-ether modifications. In order to avoid possible
conformational changes in the macrocyclic ring caused by these
modifications, we selected a simple acyclic analog of RD of ATX
(1, Fig. 8) as a reference compound because Kishi et al. reported
that the acyclic analog of RD of ATX was capable of activating
PKC.10 The lengths of acyl and alkyl chains were set to C8, which
could provide sufficient hydrophobicity to ligands as exemplified
in 1,2-dioctanoyl-sn-glycerol (diC8), a commercially available syn-
thetic DAG. The carbon numbering of acyclic analogs is based on
those of parent compounds.
20-OH (ATX)
NH (Thr-242)
30-OH (ATX)
NH (Gly-253)
C@O (Met-239)
30-OH (ATX)
C@O (Leu-251)
27-C@O (ATX)
1.86 (0.16)b
2.27 (0.24)
1.91 (0.17)
2.29 (0.38)
a
Distance between an acceptor atom and a hydrogen atom.
Standard deviations from 1000 trajectories.
b
We synthesized three acyclic analogs (1–3) from known com-
pounds 4, 6, and 8, respectively (Scheme 1), and evaluated their
ability to bind to the PKCd C1B domain. In 2, the ester group at
position 27 was replaced with an ethylene group. On the other
hand, in 3, the carbonyl group at position 1 was replaced with a
methylene group. In 2, we removed both the carbonyl and the
sp3 oxygen atoms of the 27-ester group because the ring-opening
of macrocycle would allow a free rotation of a C27–O27 bond,
which allows the sp3 oxygen atom to act as a hydrogen-bonding
acceptor in the absence of the carbonyl group. We used the syn-
thetic PKCd C1B peptide25 (d-C1B) to precisely evaluate the affinity
of ligands for the C1 domain. The concentration required to cause
50% inhibition (IC50) of [3H]phorbol 12,13-dibutyrate (PDBu) was
measured using a competitive binding assay.26 Affinity for d-C1B
Figure 6. Two conformations of the side chain at position 11 of ATX during the MD
simulation.
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
5
3. Conclusion
In conclusion, the MD simulation of the ternary complex of
POPS membrane/ATX/PKCd C1B domain suggests that the 20-OH,
27-C@O, and 30-OH groups of ATX are involved in the intermolec-
ular hydrogen bondings, and that the ester groups at position 1
might contribute to generate molecular electrostatic potential
complimentary to that of the receptor rather than hydrogen bond-
ing. The predicted binding mode of ATX with d-C1B was consistent
with previous structure–activity studies on aplysiatoxins and pro-
vided a deeper understanding of receptor-recognition by ATX. MD
simulation in explicit membrane using the binding model estab-
lished in this study would be useful for the rational design of
ATX derivatives as an anticancer seed with isozyme selectivity
and less side effects.
4. Experimental
4.1. General remarks
The following spectroscopic and analytical instruments were
used: Digital Polarimeter, Jasco P-1010 (Jasco, Tokyo, Japan); 1H
and 13C, JOEL JNM-ECA 600 (Jeol, Japan, reference TMS); HPLC,
JASCO PU-980 Intelligent HPLC pump with a JASCO PV-970 Intelli-
gent UV/VIS Detector (JASCO, Tokyo, Japan); HR-ESI-TOF-MS, Xevo
G2-XS (Waters, Tokyo, Japan). HPLC was carried out on a YMC-Pack
ODS-AM AM12S05-2510WT (YMC, Kyoto, Japan). WakogelÒ C-200
and C-300 (silica gel, Wako Pure Chemical Laboratory, Osaka,
Japan) were used for column chromatography. [3H]PDBu
(17.16 Ci/mmol) was purchased from PerkinElmer Life Science
Research Products (Boston, MA, US). The PKC C1 peptide was syn-
thesized as reported previously.25 All other chemicals and reagents
were purchased from chemical companies and used without fur-
ther purification.
Figure 7. Molecular electrostatic potential surface representation of the PKCd C1B
domain, ATX, and phorbol 13-acetate. (A) Crystal structure of the PKCd C1B
domain (PDB code: 1PTR) in complex with phorbol 13-acetate. (B) ATX in the
bound state. (C) Phorbol 13-acetate in the bound state (PDB code: 1PTR). Surfaces
with negative, neutral, and positive potentials are shown in red, white, and blue,
respectively.
4.2. Molecular modeling
4.2.1. Conformational search
The structures of ATX was built using Avogadro (version
1.0.3).28 Topology of ATX for MD simulation was generated via
the CGenFF server.29 The simulated annealing was carried out
using the GROMACS software package (version 5.1.2).30 The side
chain at C11 was replaced with a isopropyl group. All bonds were
constrained using the LINCS algorithm. The annealing temperature
was initially set to 1500 K and the temperature was kept constant
for 1 ps. The temperature was linearly dropped to 100 K over 1 ps
and then to 0 K over 1 ps, and kept at the same temperature for
1 ps. This 5-ps cycle was repeated 1000 times to give the con-
former library. Conformers with chair–chair spiroketal conforma-
tion were selected and classified into six groups (A–F). Structures
of conformers A and B were optimized using the DFT method at
was expressed as a Ki value calculated from the IC50 value of each
ligand and the dissociation constant (Kd) of [3H]PDBu, as reported
by Goldstein and Barrett.27
The affinity of 3 lacking the carbonyl group at position 1 for d-
C1B (Ki, 2900 nM) was approximately eight times lower than that
of 1 (Ki, 370 nM) (Table 3). On the other hand, 2 lacking the ester
group at position 27 showed little binding to d-C1B (Ki,
>20,000 nM). These results suggest that the carbonyl group at posi-
tion 27 was essential for receptor-recognition and the carbonyl
group at position 1 also contribute to receptor recognition, which
are consistent with ATX binding model from the docking simula-
tion, the MD simulation, and the molecular electrostatic potential
calculation.
the level of
mediate conformers between conformers A and B were optimized
at the level of
B97X-D/6-31G⁄ with a dihedral restraint at C9–O9.
x
B97X-D/6-31G⁄31 employing Gaussian09.32 The inter-
x
The obtained geometries were characterized as minimum struc-
tures on the basis of their harmonic vibrational frequencies and
number of imaginary frequencies.
4.2.2. Docking simulation
(a) Protein preparation: The crystal structure of the PKCd C1B
domain/phorbol 13-acetate complex (PDB code: 1PTR)15 was
obtained from the Protein Data Bank. (b) Ligand preparation: The
partial atomic charge of the ligands was calculated with
MOPAC201233 using PM6 Hamiltonian. (c) Docking simulations:
The docking experiment was performed with AutoDock16 4.2.6
Figure 8. Structures of acyclic analogs of the recognition domain (RD) of ATX (1–3).
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

6
Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
Scheme 1. Syntheses of 1–3. Reagents and conditions: (a) (i) 2,4,6-trichlorobenzoyl chloride, Et3N, n-octanol, DMAP, toluene; (ii) DDQ, CH2Cl2, H2O, (iii) n-octanoyl chloride,
DMAP, Et3N, CH2Cl2, 34% in three steps; (b) Pd/C, H2, EtOH, 77%; (c) (i) (1-hepthyl)triphenylphosphonium bromide, KHMDS, THF; (ii) DDQ, CH2Cl2, H2O; (iii) n-octanoic
anhydride, DMAP, pyridine, 7% in three steps; (d) Pd/C, H2, EtOH, 65%; (e) (i) n-octyl bromide, NaH, DMF; (ii) DDQ, CH2Cl2, H2O, 75% in two steps; (f) (i) SO3Ápyridine, Et3N,
DMSO, CH2Cl2; (ii) MeMgBr, MgBr2ÁEt2O, CH2Cl2, 52% in two steps; (g) (i) BnBr, NaH, DMF; (ii) MeI, NaHCO3, MeCN, H2O; (iii) NaClO2, 2-methyl-2-butene, satd NaH2PO4 aq,
t-BuOH, 30% in three steps; (h) n-octanol, 2,4,6-trichlorobenzoyl chloride, Et3N, DMAP, toluene, 50%; (i) Pd/C, H2, EtOH, 85%.
Table 3
force field were used to describe the system. Coordination of cys-
teine and histidine residues to zinc ions was treated as single
Ki values for inhibition of specific binding of [3H]PDBu by 1–3
bonds with tetrahedral geometry.
Compound
Ki for d-C1B (nM)
All MD simulations were performed using the GROMACS soft-
ware package30 (version 5.1.2). The geometry for the system was
a rectangular box with periodic boundary conditions. The system
was gradually relaxed according to CHARMM-GUI-generated posi-
tion and angle restraint conditions to reach equilibrium (300 K,
1 atm). Then, 40 ns NPT (constant number of atoms, pressure,
and temperature) simulation without any position restraint with
2 fs time step was performed. Production NPT simulations were
performed for 10 ns with a 2 fs time step. In NPT simulation, tem-
perature and pressure were regulated using the Nose–Hoover ther-
mostat algorithm and the Parinello–Rahman barostat algorithm,
respectively. The time constant for the temperature and pressure
coupling was kept at 0.5 and 5 ps, respectively. The pressure was
coupled with semi-isotropic scheme with isothermal compressibil-
ity of 4.5 Â 10À5 barÀ1. The short-range nonbonded interactions
were computed for the atom pairs within the cutoff of 1.2 nm,
while the long-range electrostatic interactions were calculated
using particle-mesh-Ewald summation method with fourth-order
cubic interpolation and 0.16 nm grid spacing. All bonds were con-
strained using the parallel LINCS method. The center of mass trans-
lations of membrane, ligand/protein, and water/ion were removed
every time step.
1
2
3
370 (280–450)a
>20,000
2900 (2300–3500)
a
95% confidence interval from at least six experiments.
and MGLTools 1.5.7 RC1. Lamarckian Genetic Algorithm was
employed as the docking algorithm. A docking grid was set around
the ligand-binding cleft of the receptor. (d) Docking parameters:
Number of Genetic Algorithm (GA) runs: 200, Population size:
150, Maximum number of evaluation: 25,000,000, Maximum num-
ber of generations: 27,000.
4.2.3. MD simulation
Missing side-chains (Lys-234, Arg-274, and Glu-274) were auto-
matically added to the crystal structure of the PKCd C1B domain
using the Swiss PDB Viewer.34 Glycine residues were added to N-
and C-termini. An initial model of POPS—ATX—PKCd-C1B ternary
complex in rectangle box filled by TIP3P water were constructed
using the CHARMM-GUI35 Bilayer Builder and the PPM Server36
as reproducing the membrane binding model of the C1 domain cal-
culated by Lomize et al.37 Z-Length of simulation box was deter-
mined by water thickness (minimum water height on top and
bottom of the system was set to 20 Å), and initial XY-lengths were
set to 61 Å. The net charge on the system was neutralized by add-
ing Na+ ions. The numbers of lipid molecule in upper and lower
leaflets were 64 and 58, respectively. The charmm36 force field20
with modification for zinc-coordinated residues and the CGenFF
4.2.4. Molecular electrostatic potential calculation
Molecular electrostatic potentials of the PKCd C1B domain, ATX,
and phorbol 13-acetate were calculated using the APBS (Adaptive
Poisson–Boltzmann Solver) program (version 1.4.2)38 and visual-
ized using PyMOL (version 1.7.6; Schrödinger, LLC). Molecular (sol-
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
7
vent-excluded) surface was colored by potential on solvent acces-
sible surface. PQR files required for the calculations were generated
from a GROMACS portable binary run input file (a tpr file).
(1.42 mg, 3.81
l
mol, 77%) as a colorless oil. [
a
]
+9.6° (c 0.071,
D
MeOH, 25.9 °C). 1H NMR (600 MHz, 296 K, CDCl3, 7.62 mM): d
0.88 (6H, m), 1.20 (3H, d, J = 6.4 Hz), 1.24–1.34 (18H, m), 1.62
(4H, m), 1.87 (1H, d, J = 7.1 Hz), 2.33 (2H, m), 2.64 (1H, dd,
J = 15.8, 7.6 Hz), 2.72 (1H, dd, J = 15.8, 5.6 Hz), 3.93 (1H, m), 4.06
(2H, t, J = 6.8 Hz), 5.16 (1H, m) ppm. 13C NMR (150 MHz, 296 K,
CDCl3, 7.62 mM): d 14.1, 14.1, 19.3, 22.6, 22.6, 25.0, 25.9, 28.6,
28.9, 29.1, 29.2, 29.2, 31.7, 31.8, 34.3, 35.8, 65.1, 68.3, 73.3,
170.7, 173.2 ppm. HR-ESI-MS: m/z = 395.2773 ([MNa]+, calcd for
4.3. Synthetic procedures
4.3.1. Synthesis of 5
To a solution of a known carboxylic acid (4)39 (192.0 mg,
0.557 mmol) and Et3N (217 lL, 1.56 mmol) in toluene (10 mL)
was added 2,4,6-trichlorobenzoyl chloride (95.8 mL, 0.618 mmol)
at rt under an Ar atmosphere. After 2.5 h of stirring, a supernatant
of the resulting suspension was added to a solution of n-octanol
C21H40O5Na, 395.2773).
4.3.3. Synthesis of 7
To suspension of heptyltriphenylphosphonium bromide
(107.7 mg, 0.244 mmol) in THF (1.5 mL) was added a 0.5 M solu-
tion of KHMDS in toluene (390
(174
l
L, 1.11 mmol) and N,N-dimethyl-4-aminopyridine (DMAP)
a
(204.0 mg, 1.67 mmol) in toluene (10 mL) at rt. The resulting mix-
ture was stirred for 1.5 h and then poured into water and EtOAc.
The mixture was extracted with EtOAc (50 mL Â 3). The combined
organic layers were washed with brine, dried over Na2SO4, filtered,
and concentrated in vacuo. The residue was purified by column
chromatography (silica gel, 8% EtOAc in hexane) to afford a mix-
ture of an octyl ester and a carboxylic acid (257.7 mg) as a colorless
oil, which was taken to the next step without further purification.
To a solution of the mixture octyl ester (257.7 mg) in CH2Cl2
(30 mL) and H2O (4.6 mL) was added DDQ (256.2 mg, 1.13 mmol)
at rt under an Ar atmosphere. After 1 h of stirring at rt, the reaction
was quenched with satd NaHCO3 aq (40 mL). The resulting mixture
was extracted with CH2Cl2 (25 mL Â 3). The combined organic lay-
ers were washed with brine, dried over Na2SO4, filtered, and con-
centrated in vacuo. The residue was purified by column
chromatography (silica gel, 8–15% EtOAc in hexane) to afford a
mixture of an alcohol and p-anisaldehyde (141.0 mg) as a colorless
oil, which was taken to the next step without further purification.
To a solution of the mixture alcohol (136.0 mg) and Et3N
l
L, 0.195 mmol) at À78 °C under
an Ar atmosphere. The resulting mixture was stirred at rt for
40 min. The reaction mixture was recooled to À78 °C, and then a
solution of a known aldehyde (6)40 (40.1 mg, 0.122 mmol) in THF
(800 lL) was added. After 18 h of stirring at rt, the reaction was
quenched with H2O (1.5 mL), and then was extracted with EtOAc
(6 mL Â 3). The combined organic layers were washed with brine,
dried over Na2SO4, filtered, and concentrated in vacuo. The residue
was purified by column chromatography (silica gel, 10–20% EtOAc
in hexane) to afford a mixture of an olefin and the by-product
(24.0 mg), which was taken to the next step without further
purification.
To a solution of the mixture olefin (24.0 mg) in CH2Cl2 (3.1 mL)
and H2O (480 lL) was added DDQ (19.9 mg, 0.876 mmol) at rt. The
reaction mixture was stirred for 3 h, and then DDQ (19.9 mg,
0.876 mmol) was added. After 1 h of stirring, the reaction was
quenched with satd NaHCO3 aq (10 mL). The resulting mixture
was extracted with CH2Cl2 (15 mL Â 3). The combined organic lay-
ers were washed with brine, dried over Na2SO4, filtered, and con-
centrated in vacuo. The residue was purified by column
chromatography (silica gel, 10% EtOAc in hexane) to afford a mix-
ture of an alcohol and p-anisaldehyde (4.6 mg) as a pale yellow oil,
which was taken to the next step without further purification.
To a solution of the mixture alcohol (4.6 mg) in pyridine
(226
ride (206
under an Ar atmosphere. The resulting mixture was stirred for
2 h, and then n-octanoyl chloride (68.8 L) and Et3N (111 L) were
l
L, 1.62 mmol) in CH2Cl2 (9 mL) were added n-octanoyl chlo-
l
L, 1.21 mmol) and DMAP (24.2 mg, 0.202 mmol) at rt
l
l
added. After 30 min of stirring, the reaction was quenched with
H2O (30 mL). The resulting mixture was extracted with EtOAc
(30 mL Â 3). The combined organic layers were washed with brine,
dried over Na2SO4, filtered, and concentrated in vacuo. The residue
was purified by column chromatography (silica gel, 2–5–15–30–
50–75–100% EtOAc in hexane) to afford an octanoyl ester
(92.2 mg) as a colorless oil.
A solution of the octanoyl ester (7.7 mg) in MeOH was filtered
and purified by HPLC (column: YMC-Pack ODS-AM AM12S05-
2510WT, solvent: 95–100% MeOH/H2O, gradient 30 min, flow rate:
8 mL/min, UV detector: 254 nm, retention time: 19.1 min) to afford
(300
DMAP (small amount) at rt under an Ar atmosphere. After 3 h of
stirring, the reaction was quenched with H2O (300 L). The result-
lL) were added octanoic anhydride (5.7 lL, 19.2 lmol) and
l
ing mixture was extracted with EtOAc (3 mL Â 3). The combined
organic layers were washed with 5% (w/v) NaHCO3 aq and brine,
dried over Na2SO4, filtered, and concentrated in vacuo. The residue
was purified by column chromatography (silica gel, 5% EtOAc in
hexane) to afford an octanoyl ester as a colorless oil. A solution
of the octanoyl ester in MeOH was filtered and purified by HPLC
(column: YMC-Pack ODS-AM AM12S05-2510WT, solvent: 95%
MeOH/H2O, flow rate: 8 mL/min, UV detector: 254 nm, retention
an octanoyl ester (5) (7.0 mg, 15.1 lmol, 49% in three steps) as a
colorless oil.
[
a]
À1.2° (c 0.23, CHCl3, 26.4 °C). 1H NMR
D
(600 MHz, 299 K, CDCl3, 9.94 mM): d 0.88 (6H, m), 1.16 (3H, d,
J = 6.4 Hz), 1.21–1.33 (18H, m), 1.60 (4H, m), 2.30 (2H, t,
J = 7.6 Hz), 2.61 (1H, dd, J = 15.9, 8.6 Hz), 2.71 (1H, dd, J = 15.9,
4.8 Hz), 3.74 (1H, qd, J = 6.4, 4.0 Hz), 4.02 (2H, t, J = 6.8 Hz), 4.52
(1H, d, J = 11.9 Hz), 4.62 (1H, d, J = 11.9 Hz), 5.39 (1H, m), 7.26–
7.35 (5H, m) ppm. 13C NMR (150 MHz, 299 K, CDCl3, 9.94 M): d
14.1, 14.1, 14.9, 22.6, 22.6, 25.0, 25.9, 28.6, 28.9, 29.1, 29.2, 29.2,
31.7, 31.8, 34.4, 34.8, 64.9, 71.2, 71.3, 73.8, 127.7, 127.8 (2C),
128.4 (2C), 138.3, 170.8, 173.0 ppm. HR-ESI-MS: m/z = 485.3247
([MNa]+, calcd for C28H46O5Na, 485.3243).
time: 26.8 min) to afford the octanoyl ester (7) (3.6 mg, 8.64
7% in three steps) as a colorless oil. [
lmol,
a]
À11.8° (c 0.507, MeOH,
D
26.8 °C). 1H NMR (600 MHz, 296 K, CDCl3, 0.0173 M): d 0.87 (3H,
t, J = 7.1 Hz), 0.88 (3H, t, J = 7.0 Hz), 1.16 (3H, d, J = 6.4 Hz), 1.21–
1.36 (16H, m), 1.61 (2H, m), 2.02 (2H, br q, J = 6.9 Hz), 2.30 (2H,
t, J = 7.6 Hz), 2.37 (2H, t, J = 7.2 Hz), 3.64 (1H, m), 4.51 (1H, d,
J = 12.0 Hz), 4.63 (1H, d, J = 12.0 Hz), 4.97 (1H, m), 5.28 (1H, m),
5.45 (1H, m), 7.26–7.43 (5H, m) ppm. 13C NMR (150 MHz, 296 K,
CDCl3, 0.0173 M): d 14.1, 14.1, 15.4, 22.6, 22.6, 25.1, 27.4, 27.8,
29.0, 29.0, 29.1, 29.6, 31.7, 31.8, 34.5, 71.2, 74.5, 75.2, 124.3,
127.5, 127.6 (2C), 128.3 (2C), 132.8, 138.6, 173.6 ppm. HR-ESI-
MS: m/z = 439.3187 ([MNa]+, calcd for C27H44O3Na, 439.3188).
4.3.2. Synthesis of 1
To a solution of 5 (2.3 mg, 4.97 lmol) in EtOH (900 lL) was
added Pd/C (small amount). After 20 h of stirring at rt under H2
atmosphere, the reaction mixture was filtered, and the filtrate
was concentrated in vacuo. The residue was purified by column
chromatography (silica gel, 15% EtOAc in hexane) to afford 1
4.3.4. Synthesis of 2
To a solution of 7 (3.6 mg, 8.64 lmol) in EtOH (700 lL) was
added Pd/C (small amount). After 4 h of stirring under a H2 atmo-
sphere, the reaction mixture was filtered, and the filtrate was
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

8
Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
concentrated in vacuo. The residue was purified by column chro-
matography (silica gel, 10% EtOAc in hexane) to afford
(1.84 mg, 5.60 mol, 65%) as colorless oil. [ +9.6° (c 0.0920,
hexane) to afford an alcohol (10) (9.9 mg, 30.9
lmol, 52% in two
2
steps) as a pale yellow oil. [
a]
D
+1.9° (c 0.97, CHCl3, 26.3 °C). 1H
l
a]
NMR (600 MHz, 298 K, CDCl3, 0.0618 M): d 0.88 (3H, t, J = 7.1 Hz),
1.20 (3H, d, J = 6.4 Hz), 1.25–1.40 (10H, m), 1.55–1.61 (2H, m),
1.84–1.90 (1H, m), 1.90–1.96 (2H, m), 2.09–2.26 (1H, m), 2.22
(1H, d, J = 5.0 Hz), 2.78–2.95 (4H, m), 3.38 (1H, dd, J = 12.0,
5.5 Hz), 3.54 (1H, dt, J = 9.1, 6.6 Hz), 3.66 (1H, m), 3.71 (1H, m),
4.15 (1H, J = 7.3 Hz) ppm. 13C NMR (150 MHz, 298 K, CDCl3,
0.0618 M): d 14.1, 19.3, 22.7, 25.9, 26.2, 29.3, 29.5, 30.2, 30.3,
30.5, 31.8, 37.2, 43.9, 69.4, 71.8, 80.3 ppm. HR-ESI-MS: m/
z = 343.1733 ([MNa]+, calcd for C16H32O2NaS2, 343.1741).
D
MeOH, 20.2 °C). 1H NMR (600 MHz, 296 K, CDCl3, 0.0184 M): d
0.88 (6H, m), 1.17 (3H, d, J = 6.4 Hz), 1.22–1.33 (22H, m), 1.59
(2H, m), 1.64 (2H, m), 1.68 (1H, br d, J = 6.9 Hz), 2.35 (2H, t,
J = 7.6 Hz), 3.80 (1H, m), 4.78 (1H, dt, J = 7.9, 5.0 Hz) ppm. 13C
NMR (150 MHz, 296 K, CDCl3, 0.0184 M): d 14.1, 14.1, 19.7, 22.6,
22.7, 25.2, 25.3, 29.1, 29.0, 29.3, 29.5 (3C), 30.6, 31.7, 31.9, 34.5,
68.8, 77.5, 173.9 ppm. HR-ESI-MS: m/z = 351.2862 ([MNa]+, calcd
for C20H40O3Na, 351.2852).
4.3.5. Synthesis of 9
4.3.7. Synthesis of 11
To a suspension of 60% NaH in oil (382.0 mg, 9.55 mmol) in
DMF (10 mL) was added a solution of a known alcohol (8)39
(1201.3 mg, 3.82 mmol) in DMF (24 mL). After 30 min of stirring
at 0 °C, 1-bromooctane (1.3 mL, 7.6 mmol) was added dropwise.
The mixture was stirred for 30 min, and then warmed to rt. After
3.5 h of stirring, the reaction was quenched with H2O (25 mL),
and the resulting mixture was diluted with EtOAc (25 mL). The
organic layer was separated, and the aqueous layer was extracted
with EtOAc (50 mL Â 3). The combined organic layers were washed
with brine, dried over Na2SO4, filtered, and concentrated in vacuo.
The residue was purified by column chromatography (silica gel, 7%
EtOAc in hexane) to afford a mixture of an octyl ether and an alco-
hol (1.477 mg) as a colorless oil, which was taken to the next step
without further purification.
To a solution of the mixture octyl ether (1307 mg) in CH2Cl2
(162 mL) and H2O (24.8 mL) was added DDQ (1.4 g, 6.1 mmol) at
rt. After 1.5 h of stirring, the reaction was quenched with satd
NaHCO3 aq (160 mL). The resulting mixture was extracted with
CH2Cl2 (150 mL Â 3). The combined organic layers were washed
with brine, dried over Na2SO4, filtered, and concentrated in vacuo.
The residue was purified by column chromatography (silica gel,
15–30% EtOAc in hexane) to afford an alcohol (9) (778.8 mg,
To a suspension of 60% NaH in oil (103.2 mg, 2.58 mmol) in
DMF (5 mL) was added a solution of 10 (413.5 mg, 1.29 mmol) in
DMF (10 mL) at 0 °C. After 30 min of stirring, BnBr (230 lL,
1.94 mmol) was added dropwise. The reaction mixture was stirred
for 30 min, and then warmed to rt. After 4 h of stirring, the reaction
was quenched with satd NH4Cl aq (12 mL). The resulting mixture
was extracted with EtOAc (20 mL Â 3). The combined mixture
was washed with brine, dried over Na2SO4, filtered, and concen-
trated in vacuo. The residue was purified by column chromatogra-
phy (silica gel, 2–10–15% EtOAc in hexane) to afford a mixture of a
benzyl ether and a reagent (252.0 mg) as a colorless oil and a
recovered substrate (235.1 mg, 0.733 mmol, 57%). The benzyl ether
mixture was taken to the next step without further purification.
To a solution of the mixture benzyl ether (235.1 mg) in CH3CN
(5.3 mL) and H2O (1.3 mL) were added NaHCO3 (142 mg,
1.69 mmol) and CH3I (1.4 mL, 22.5 mmol) at rt. After 18 h of stir-
ring, the reaction mixture was extracted with EtOAc (10 mL Â 3).
The combined organic layers were washed with brine, dried over
Na2SO4, filtered, and concentrated in vacuo. The residue was puri-
fied by column chromatography (silica gel, 5–15% EtOAc in hexane)
to afford a mixture of an aldehyde and the by-product (149.0 mg)
as a colorless oil, which was taken to the next step without
purification.
To a solution of the mixture of the aldehyde (117.6 mg) in
t-BuOH (2.9 mL) and satd NaH2PO4 aq (1.5 mL) was added a solution
of 2-methyl-2-butene (276 lL) and NaClO2 (66.2 mg, 0.512 mmol)
in satd NaH2PO4 aq (1.4 mL) at 0 °C. After 10 min of stirring, the
reaction was quenched with 5% H3PO4 aq (7 mL). The resulting mix-
ture was extracted with EtOAc (7 mL Â 3). The combined organic
layers were washed with brine, dried over Na2SO4, filtered, and con-
centrated in vacuo. The residue was purified by column chromatog-
raphy (silica gel, 100% CHCl3) to afford a carboxylic acid (11)
2.54 mmol, 75% in two steps) as a colorless oil. [
a
]
À3.6° (c 1.00,
D
CHCl3, 26.1 °C). 1H NMR (600 MHz, 300 K, CDCl3, 0.0568 M): d
0.88 (3H, t, J = 7.0 Hz), 1.27–1.39 (10H, m), 1.54–1.61 (2H, m),
1.83–1.94 (3H, m), 2.05 (1H, m), 2.12 (1H, m), 2.81–2.95 (4H, m),
3.47–3.57 (3H, m), 3.64–3.68 (1H, m), 3.75 (1H, dt, J = 11.4,
4.2 Hz), 4.14 (1H, m) ppm. 13C NMR (150 MHz, 300 K, CDCl3,
0.0568 M): d 14.1, 22.7, 25.9, 26.2, 29.3, 29.4, 30.1 (2C), 30.4,
31.9, 37.2, 43.8, 63.9, 70.1, 76.2 ppm. HR-ESI-MS: m/z = 329.1580
([MNa]+, calcd for C15H30O2NaS2, 329.1585).
4.3.6. Synthesis of 10
To a solution of 9 (178.6 mg, 0.583 mmol) in CH2Cl2 (2.4 mL)
and DMSO (611 lL) were added Et3N (568 lL, 4.08 mmol) and SO3-
(121.7 mg, 0.367 mmol, 30% in three steps) as a colorless oil. [a]
D
À0.4° (c 0.67, CHCl3, 26.4 °C). 1H NMR (600 MHz, 300 K, CDCl3,
0.0297 M): d 0.88 (3H, m), 1.17 (3H, d, J = 6.4 Hz), 1.21–1.30 (10H,
m), 1.52 (2H, m), 2.49 (1H, dd, J = 15.8, 8.1 Hz), 2.68 (1H, dd,
J = 15.8, 4.3 Hz), 3.52 (2H, t, J = 6.6 Hz), 3.70 (1H, m), 3.84 (1H, m),
4.51 (1H, d, J = 11.9 Hz), 4.61 (1H, d, J = 11.9 Hz), 7.26–7.35 (5H,
m) ppm. 13C NMR (150 MHz, 300 K, CDCl3, 0.0297 M): d 14.1,
14.4, 22.7, 26.1, 29.3, 29.4, 30.0, 31.8, 35.1, 71.1, 71.2, 74.6, 77.7,
127.6, 127.7 (2C), 128.4 (2C), 138.4, 176.3 ppm. HR-ESI-MS:
m/z = 359.2190 ([MNa]+, calcd for C20H32O4Na, 359.2198).
Ápyridine (370.8 mg, 2.33 mmol) at 0 °C. After 2.5 h of stirring at rt,
the reaction was quenched with satd NH4Cl aq (5 mL). The result-
ing mixture was extracted with CH2Cl2 (10 mL Â 3). The combined
organic layers were washed with brine, dried over Na2SO4, filtered,
and concentrated in vacuo. The residue was purified by column
chromatography (silica gel, 7–10% EtOAc in hexane) to afford a
mixture of an aldehyde and the by-product (120.1 mg) as a pale
yellow oil, which was taken to the next step without further
purification.
4.3.8. Synthesis of 12
To a solution of the mixture aldehyde (12.1 mg) in CH2Cl2
To a solution of 11 (57.2 mg, 0.170 mmol) and Et3N (66.3
0.476 mmol) in toluene (3 mL) was added 2,4,6-trichlorobenzoyl
chloride (29.1 L, 0.187 mmol) at rt. After 2.5 h of stirring, a super-
natant of the resulting suspension was added to a solution of
n-octanol (53.3 L, 0.34 mmol) and DMAP (62.3 mg, 0.51 mmol) in
toluene (10 mL) at rt. The resulting mixture was stirred for 1.5 h
and then poured into water and EtOAc. The mixture was extracted
with EtOAc (15 mL Â 3). The combined organic layers were washed
with brine, dried over Na2SO4, filtered, and concentrated in vacuo.
lL,
(300
The mixture was stirred for 1 h, and then 3 M CH3MgBr in diethyl
ether (199 L, 59.6
l
L) was added MgBr2ÁEt2O (12.3 mg, 47.6
l
mol) at À78 °C.
l
l
l
mol) was added dropwise at À78 °C. After
1.5 h of stirring, the reaction was quenched with satd NH4Cl aq
(2 mL). The resulting mixture was extracted with CH2Cl2
(10 mL Â 3). The combined organic layers were washed with brine,
dried over Na2SO4, filtered, and concentrated in vacuo. The residue
was purified by column chromatography (silica gel, 15% EtOAc in
l
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
9
The residue was purified by column chromatography (silica gel, 3%
EtOAc in hexane) to afford a mixture of an octyl ester and 2,4,6-
trichlorobenzoic acid. The mixture was further purified by column
chromatography (silica gel, 1–2% EtOAc in hexane) to afford a mix-
ture of the ester and the carboxylic acid (85.8 mg), and a pure ester
(44.1 mg) as a colorless oil, respectively. A pure portion of the ester
(11.0 mg) was purified by HPLC (column: YMC-Pack ODS-AM
AM12S05-2510WT, solvent: a linear gradient of 95–100% MeOH/
H2O over 30 min, flow rate: 8 mL/min, UV detector: 254 nm, reten-
tion time: 22.5 min) to afford an octanoyl ester (12) (9.6 mg,
ucts’ [number 23102011] (to K.I. and R.C.Y.) from The Ministry of
Education, Culture, Sports, Science, and Technology, Japan.
The docking and MD simulations were carried out at the
Research Center for Computational Science, Okazaki Research
Facilities, National Institutes of Natural Sciences (NINS).
Supplementary data
Supplementary data (root mean square deviation (RMSD) of the
PKCd C1B domain and ATX during MD simulation. NMR spectra of
1–3) associated with this article can be found, in the online version,
at http://dx.doi.org/10.1016/j.bmc.2016.07.011.
21.3
lmol, 50%) as a colorless oil. [
a]
+1.2° (c 0.37, CHCl3,
D
26.6 °C). 1H NMR (600 MHz, 300 K, CDCl3, 0.0165 M): d 0.88 (6H,
m), 1.16 (3H, d, J = 6.4 Hz), 1.22–1.36 (20H, m), 1.50 (2H, m), 1.61
(2H, m), 2.45 (1H, dd, J = 15.5, 8.8 Hz), 2.59 (1H, dd, J = 15.5,
4.1 Hz), 3.49 (2H, m), 3.68 (1H, m), 3.88 (1H, quint, J = 4.3 Hz),
4.05 (2H, m), 4.52 (1H, d, J = 11.8 Hz), 4.66 (1H, d, J = 11.8 Hz),
7.26–7.33 (5H, m) ppm. 13C NMR (150 MHz, 300 K, CDCl3,
0.0165 M): d 14.1 (2C), 14.5, 22.7 (2C), 26.0, 26.1, 28.7, 29.2,
29.3, 29.3, 29.5, 30.1, 31.8, 31.9, 35.6, 64.7, 71.1, 71.2, 74.8, 78.0,
127.5, 127.7 (2C), 128.3 (2C), 138.7, 172.4 ppm. HR-ESI-MS:
m/z = 471.3454 ([MNa]+, calcd for C28H48O4Na, 471.3450).
References and notes
1. Kato, Y.; Scheuer, P. J. J. Am. Chem. Soc. 1974, 96, 2245.
2. Moore, R. E. Pure Appl. Chem. 1982, 54, 1919.
3. Fujiki, H.; Sugimura, T. Adv. Cancer Res. 1987, 49, 223.
4. Ramos, O. F.; Masucci, M. G.; Klein, E. Cancer Res. 1984, 44, 1857.
5. Nishizuka, Y. Nature 1984, 308, 693.
6. Mackay, H. J.; Twelves, C. J. Nat. Rev. Cancer 2007, 7, 554.
7. Arcoleo, J. P.; Weinstein, I. B. Carcinogenesis 1985, 6, 213.
8. Nakagawa, Y.; Yanagita, R. C.; Hamada, N.; Murakami, A.; Takahashi, H.; Saito,
N.; Nagai, H.; Irie, K. J. Am. Chem. Soc. 2009, 131, 7573.
9. Irie, K.; Yanagita, R. C. Chem. Rec. 2014, 14, 251.
10. Kong, F.; Kishi, Y.; Perez-Sala, D.; Rando, R. R. Proc. Natl. Acad. Sci. U.S.A. 1991,
88, 1973.
4.3.9. Synthesis of 3
To a solution of 12 (13.3 mg, 29.6 lmol) in EtOH (1.5 mL) was
added Pd/C (1.3 mg) at rt. After 21 h of stirring under a H2 atmo-
sphere, the reaction mixture was filtered. The filtrate was purified
by column chromatography (silica gel, 5–8% EtOAc in hexane) to
11. Nakamura, H.; Kishi, Y.; Pajares, M. A.; Rando, R. R. Proc. Natl. Acad. Sci. U.S.A.
1989, 86, 9672.
12. Giorgione, J. R.; Lin, J. H.; McCammon, J. A.; Newton, A. C. J. Biol. Chem. 2006,
281, 1660.
13. (a) Knust, H.; Hoffmann, R. W. Chem. Rec. 2002, 2, 405; (b) Knust, H.; Hoffmann,
R. W. Helv. Chim. Acta 2003, 86, 1871.
afford 3 (9.04 mg, 25.2
lmol, 85%) as a colorless oil. [
a]
+4.9°
D
(c 0.452, MeOH, 19.4 °C) 1H NMR (600 MHz, 295 K, CDCl3,
0.0428 M): d 0.88 (6H, t, J = 7.0 Hz), 1.20 (3H, d, J = 6.4 Hz), 1.26–
1.35 (20H, m), 1.51–1.66 (4H, m), 2.35 (1H, s), 2.51 (1H, dd,
J = 15.5, 6.4 Hz), 2.55 (1H, dd, J = 15.6, 5.5 Hz), 3.46 (1H, dt, J = 9.0,
6.7 Hz), 3.55 (1H, m), 3.61 (1H, m), 3.72 (1H, quintet, J = 6.0 Hz),
4.09 (2H, m) ppm. 13C NMR (150 MHz, 295 K, CDCl3, 0.0428 M): d
14.1 (2C), 18.9, 22.6 (2C), 25.9, 26.1, 28.6, 29.2, 29.2, 29.3, 29.4,
30.1, 31.8, 31.8, 36.5, 64.9, 69.4, 71.2, 80.4, 171.8 ppm. HR-ESI-MS:
m/z = 381.2980 ([MNa]+, calcd for C21H42O4Na, 381.2981).
14. Moore, R. E.; Blackman, A. J.; Cheuk, C. E.; Mynderse, J. S.; Matsumoto, G. K.;
Clardy, J.; Woodard, R. W.; Craig, J. C. J. Org. Chem. 1984, 49, 2484.
15. Zhang, G.; Kazanietz, M. G.; Blumberg, P. M.; Hurley, J. H. Cell 1995, 81, 917.
16. Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew, R. K.; Goodsell, D.
S.; Olson, A. J. J. Comput. Chem. 2009, 30, 2785.
17. Sigano, D. M.; Peach, M. L.; Nacro, K.; Choi, Y.; Lewin, N. E.; Nicklaus, M. C.;
Blumberg, P. M.; Marquez, V. E. J. Med. Chem. 2003, 46, 1571.
18. Li, J.; Ziemba, B. P.; Falke, J. J.; Voth, G. A. J. Am. Chem. Soc. 2014, 136, 11757.
19. Dries, D. R.; Gallegos, L. L.; Newton, A. C. J. Biol. Chem. 2007, 282, 826.
20. (a) Best, R. B.; Zhu, X.; Shim, J.; Lopes, P. E.; Mittal, J.; Feig, M.; MacKerell, A. D.,
Jr. J. Chem. Theory Comput. 2012, 8, 3257; (b) Klauda, J. B.; Venable, R. M.;
Freites, J. A.; O’Connor, J. W.; Tobias, D. J.; Mondragon-Ramirez, C.; Vorobyov, I.;
MacKerell, A. D., Jr.; Pastor, R. W. J. Phys. Chem. B 2010, 114, 7830.
21. Suganuma, M.; Fujiki, H.; Tahira, T.; Cheuk, C.; Moore, R. E.; Sugimura, T.
Carcinogenesis 1984, 5, 315.
4.4. Inhibition of specific binding of [3H]PDBu to the PKCd C1B
peptide
22. Keck, G. E.; Poudel, Y. B.; Rudra, A.; Stephens, J. C.; Kedei, N.; Lewin, N. E.;
Peach, M. L.; Blumberg, P. M. Angew. Chem., Int. Ed. 2010, 49, 4580.
23. Jeffrey, George A. An Introduction to Hydrogen Bonding; Oxford University Press:
Oxford, 1997.
24. Náray-Szabó, G. J. Mol. Recognit. 1993, 6, 205.
25. Irie, K.; Oie, K.; Nakahara, A.; Yanai, Y.; Ohigashi, H.; Wender, P. A.; Fukuda, H.;
Konishi, H.; Kikkawa, U. J. Am. Chem. Soc. 1998, 120, 9159.
26. Sharkey, N. A.; Blumberg, P. M. Cancer Res. 1985, 45, 19.
27. Goldstein, A.; Barrett, R. W. Mol. Pharmacol. 1987, 31, 603.
28. Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.; Zurek, E.;
Hutchison, G. R. J. Cheminform. 2012, 4, 17.
The binding of [3H]PDBu to the PKCd C1B peptide was evaluated
by the procedure of Sharkey and Blumberg27 with modifications as
reported previously41 using 50 mM Tris–maleate buffer (pH 7.4 at
4 °C), 13.8 nM PKCd C1B peptide, 5 (for 2 and 3) or 20 (for 1) nM
[3H]PDBu (17.16 Ci/mmol, Perkin–Elmer Life Science), 50
l
g/mL
-serine (Funakoshi), 3 mg/mL
-globulin (Sigma), and various concentrations of inhibitors.
1,2-dioleoyl-sn-glycero-3-phospho-
L
bovine
c
Binding affinity was evaluated on the basis of the concentration
required to cause 50% inhibition of the specific binding of [3H]PDBu,
IC50, which was calculated by logit analysis using Microsoft Excel.
When the inhibition rate of specific binding of [3H]PDBu did not fall
below 50%, IC50 was extrapolated from regression line. The binding
inhibition constant, Ki, was calculated by the formula of Goldstein
and Barrett,28 Ki = IC50/(2[L50]/[L0] À 1 + [L50]/Kd), where [L50] and
[L0] are the free concentration of [3H]PDBu at 50% and 0% inhibition,
respectively. Although we used PKCd C1B peptide in 13.8 nM, the
concentration of the properly folded peptide was estimated to be
about 3–4 nM on the base of Bmax value of Scatchard analysis
reported previously.40 Therefore, the concentration of free PDBu
did not markedly vary over the dose–response curve.
29. Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.;
Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, I.; MacKerell, A. D., Jr. J. Comput.
Chem. 2010, 31, 671.
30. van der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark, A. E.; Berendsen, H. J.
C. J. Comput. Chem. 2005, 26, 1701.
31. Chai, J. D.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2008, 10, 6615.
32. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.;
Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.;
Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven,
T., ; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.;
Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.;
Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.;
Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas,
Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09. Revision
D.01; Gaussian: Wallingford CT, 2009.
Acknowledgements
33. MOPAC2012, Stewart, J. J. P.; Stewart Computational Chemistry; Colorado
Springs: CO, USA, 2012. http://OpenMOPAC.net.
This work was supported by a Grant-in-Aid for Scientific
Research on Innovative Areas ‘Chemical Biology of Natural Prod-
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011

10
Y. Ashida et al. / Bioorg. Med. Chem. xxx (2016) xxx–xxx
34. Guex, N.; Peitsch, M. C. Electrophoresis 1997, 18, 2714.
39. Hanaki, Y.; Kikumori, M.; Ueno, S.; Tokuda, H.; Suzuki, N.; Irie, K. Tetrahedron
2013, 69, 7636.
40. Wender, P. A.; Brabander, J. D.; Harran, P. G.; Jimenez, J. M.; Koehler, M. F. T.;
Lippa, B.; Park, C. M.; Shiozaki, M. J. Am. Chem. Soc. 1998, 120, 4534.
41. Shindo, M.; Irie, K.; Nakahara, A.; Ohigashi, H.; Konishi, H.; Kikkawa, U.;
Fukuda, H.; Wender, P. A. Bioorg. Med. Chem. 2001, 9, 2073.
35. Jo, S.; Kim, T.; Iyer, V. G.; Im, W. J. Comput. Chem. 2008, 29, 1859.
36. Lomize, M. A.; Pogozheva, I. D.; Joo, H.; Mosberg, H. I.; Lomize, A. L. Nucleic
Acids Res. 2012, 40, D370.
37. Lomize, A. L.; Pogozheva, I. D.; Lomize, M. A.; Mosberg, H. I. BMC Struct. Biol.
2007, 7, 44.
38. Baker, N. A.; Sept, D.; Joseph, S.; Holst, M. J.; McCammon, J. A. Porc. Natl. Acad.
Sci. U.S.A. 2001, 98, 10037.
Please cite this article in press as: Ashida, Y.; et al. Bioorg. Med. Chem. (2016), http://dx.doi.org/10.1016/j.bmc.2016.07.011