De novo design and synthesis of a γ-turn peptidomimetic scaffold and its application as JNK3 allosteric ligand

Article abstract of DOI:10.1002/asia.201403417

As a way to develop a neuroprotective agent for the JNK3-JIP1-binding site, peptidomimetics of JIP-1 as JNK3 allosteric regulators have been examined. The study consisted of in silico scaffold hopping, molecular docking, solution and solid-phase peptide s

Full text of DOI:10.1002/asia.201403417

CHEMISTRY
AN ASIAN JOURNAL
Accepted Article
Title: De Novo Design and Synthesis of a -Turn Peptidomimetic Scaffold and Its Ap-
plication as JNK3 Allosteric Ligand
Authors: Mi-hyun Kim; Junghun Lee; Jung-Mi Hah
This manuscript has been accepted after peer review and the authors have elected
to post their Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by using the Digital
Object Identifier (DOI) given below. The VoR will be published online in Early View as
soon as possible and may be different to this Accepted Article as a result of editing.
Readers should obtain the VoR from the journal website shown below when it is pu-
blished to ensure accuracy of information. The authors are responsible for the content
of this Accepted Article.
To be cited as: Chem. Asian J. 10.1002/asia.201403417
Link to VoR: http://dx.doi.org/10.1002/asia.201403417
A sister journal of Angewandte Chemie
and Chemistry – A European Journal
Supported by
Federation of
Asian Chemical
Societies

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
De Novo Design and Synthesis of a γ-Turn Peptidomimetic
Scaffold and Its Application as JNK3 Allosteric Ligand
Mi-hyun Kim,[a][b] Junghun Lee,[a] and Jung-Mi Hah*[a]
Dedication ((optional))
Abstract: As a way to develop a neuroprotective agent for the
JNK3-JIP1-binding site, peptidomimetics of JIP-1 as JNK3 allosteric
regulators have been conducted. The study consisted of in silico
scaffold hopping, molecular docking, solution and solid-phase
peptide syntheses, and Kd measurements using surface plasmon
resonance. As a peptidomimetic of JIP1, heptamer mimetics 16 (Kd
= 7.3 µM) presented higher affinity than decamer JIP1 (Kd = 27.7
µM). The high affinity of 16 implies that the characteristic γ-turn
mimetic structure, ““Φ–X–Φ” hydrophobic motif in 16, increased its
affinity toward the JIP-site of JNK3.
neurodegenerative disease states.[9] While JNK1 and JNK2 are
ubiquitously expressed in most tissues, JNK3 is primarily
expressed in the brain and, to a lesser extent, in the heart and
testes.[10] Many studies have shown that JNK3 plays a central
role in the brain in mediating neurodegeneration, such as β-
amyloid processing and neuronal apoptosis in Alzheimer’s
disease,[11] as well as mediation of neurotoxicity in rodent
models of Parkinson’s disease.[12]
Most of the known JNK3 inhibitors also bind the ATP-binding
pocket; however, JNKs have another druggable pocket named
JIP (JNK-interacting potein)-site. The JIP-site is associated with
JNK-interacting protein, a scaffold protein that enhances JNK
signaling by creating a proximity effect between JNK and
upstream kinases. More specifically, JIP protein helps
MAP2K7/MAP2K4 activate JNK3 by phosphorylation.[13]
Furthermore, many studies have tried to achieve JIP-mimetics,
especially using an epitope peptide approach to inhibit JNK
activation. The JNK-JIP interaction is mediated by a specific,
high-affinity D-domain on JIP, whose consensus sequence is
R/KXXXXLXL. Consequently, several peptide inhibitors of JNK
have been identified by investigating the minimal JNK-binding
region of JIP. A representative peptide is pepJIP1 (amino acids
153–163 on D-domain; Pro-Lys-Arg-Pro-Thr-Thr-Leu-Asn-Leu-
Phe) that inhibits JNK activity in vitro toward recombinant c-Jun,
Elk, and ATF-2, and displays a remarkable selectivity with little
inhibition of closely related Erk and p38 MAPKs.
Introduction
Protein kinases are now the second largest group of drug
targets, and most protein kinase inhibitors in clinical
development are directed toward the ATP-binding site. However,
these inhibitors must compete with high intracellular ATP
concentrations and must discriminate between the ATP-binding
sites of all other protein kinases as well as other proteins that
utilize ATP. It would therefore be beneficial to target sites on
protein kinases other than the ATP-binding pocket in terms of
potency and selectivity.
Like other valid kinase targets, the development of small
molecule inhibitors of c-Jun N-terminal kinases (JNKs) as
therapeutics for a variety of diseases has been a target of
In this paper, we describe a novel allosteric inhibitor toward the
JNK-JIP interaction site instead of conventional orthosteric
inhibitors that use the ATP-binding site. For this purpose, the
structure of a peptide decamer (JIP1) was examined (Figure
1).[14] The JIP site in JNK has already been reported, and some
peptide analogues of JIP1 have been presented as JNK
inhibitors.[15]
intense research. JNK is a well-known checkpoint that regulates
[1]
cell apoptosis and regulation of cell survival
by
phosphorylating the activator protein-1 (AP-1) transcription
factor family (such as c-Jun, ATF-2, NFAT-4), which is involved
in responses to cytokines, oxidative stress, neurotoxins, and
[2] [3]
fatty acids.
,
Studies of JNKs have focused on the
mechanisms of their related diseases (cancer, metabolic
diseases, autoimmune diseases, and neuronal diseases) and
direct or indirect regulation of JNKs have been investigated
based on these studies.[4] There are three human JNK isoforms:
JNK1, JNK2, and JNK3,[5]-[8] and of the three subtypes, JNK3,
has garnered our attention as a potent therapeutic target of
[a]
M. H. Kim, J. Lee, J. M. Hah
Department of Pharmacy
College of Pharmacy & Institute of Pharmaceutical Science and
Technology, Hanyang University
55 Hanyangdaehak-ro, Sangnok-gu, Ansan, Kyeonggi-do, 426-791,
Korea
E-mail: jhah@hanyang.ac.kr
[b]
M. H. Kim
Department of Pharmacy
Figure 1. X-ray Binding Mode of a Part of JIP1 with JNK3 (Reported PDB
College of Pharmacy & Gachon Institute of Pharmaceutical Science,
Gachon University
Code 4H39).[14]
155 Gaetbeol-ro, Yeonsu-gu, Incheon, Korea 406-840

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
Although there are reports on small peptides as JNK-JIP
binding-site modulators, their peptidomimetic modification is
relatively unknown from a medicinal chemistry standpoint.[16] In
particular, research into the peptidomimetic modification of JIP1
could be of great interest, since peptidomimetic drugs have lot of
advantages including stability (less steep rates of
biodegradation), bioavailability, and the consequent efficacy in
vivo.[17]
CombiGlide module in Schrodinger package (experimental
section). The hit structures were sorted by docking score and
also modified through rational design and redocking analysis,
considering hexahydropyrroloimidazole bicyclic ring was
selected as a candidate for synthesis as the γ-turn mimetics.
The binding feature of γ-turn incorporated JIP peptide on JIP site
of JNK3 was simulated in the Figure 4. Since the hydrophobic
residue Leu seems to be essential for JIP-JNK interaction, we
also planned to introduce Phe residue as a hydrophobic
replacement in either or both Φ of “Φ–X–Φ” hydrophobic motif.
Results and Discussion
γ- Turn mimetic strategy
Peptidomimetics can be classified into as α-helix, β-turn, β-
strand, γ-turn in terms of their secondary structure, which is
important for their recognition.[17] In the JIP-JNK interaction site,
the γ-turn Leu-X-Leu motif (“Φ–X–Φ” hydrophobic motif) in JIP1
(Figure 1) is known as the key interaction residue binding the
JNK3-JIP site. A γ-turn is a three amino acid sequence that
contains a seven-membered ring constructed by a hydrogen
bond between the carbonyl C=O of the ith residue and the amide
N–H of the (i + 2) residue (Figure 2). In order to design a JIP1
mimetic, we focused on the γ-turn of the key motif in the JIP1
peptide. While various β-turn mimetics including some innovative
scaffolds are reported, γ-turn mimetics are known comparatively
rare.[18] As shown in figure 2, the β-turn mimetics (e & f) are
quite different from the original template (d) making it is difficult
to predict β-turn in 2D structure without priori information.
However, in γ-turn mimetics, most mimetics (including hydrogen
bonding such as BMS-790052 or covalent bonding such as 1,4-
diazepin-2-one) retain the 7-membered ring of the template
(a).[19] We speculated whether an innovative scaffold from
modified γ-turn could contribute to increase in binding affinity to
inhibit JNK3 activity efficiently.
Figure 3. Screening of γ-turn Mimetics Using CombiGlide.
Synthesis of - turn mimetic scaffold
The synthetic route to compounds (8) is shown in the Scheme 1.
The synthetic scheme emphasized the use of amino acids as
synthons for an easy scale-up of the reaction. The amino group
of the C-terminal amino acid (Leu/Phe) reacted with 1,4-
dichloro-but-2-ene (1) to produce 3-pyrrolines (3).[20] The double
bond of (3) was dihydroxylated under 4% (w/w) osmium
tetroxide (as an oxidant) and N-methylmorpholine N-oxide (as a
co-oxidant) to produce a vicinal syn-type meso compound.[21]
Dihydroxypyrroline then reacted with mesyl chloride to provide
dimesylated products (4). The dimesyl groups were then
substituted with diazide groups under reflux conditions with
excess sodium azide to produce diazido compounds (5).
Compounds (5) were then reduced using Raney Ni to afford
diamines (6). Compounds (6) were cyclized to form an
imidazoline with imidates (7), which were synthesized from
another amino acid synthon.[22] The Kd values of cyclized
hexahydropyrrolo[3,4-d]imidazole bicyclic compounds (8) as
Figure 2. (a) γ-turn, (b) γ-turn mimetic with H-bond, (c) γ-turn mimetic without
H-bond, (d) β-turn, (e) β-turn mimetic with H-bond, (f) β-turn mimetic without
H-bond
In order to find eligible scaffold replacing the γ-turn motif, we
began scaffold hopping, which is a virtual screening of various
fragments in silico based on crystal structure of JIP-JNK (Figure
3). The effect of side-chain of Asn, (i+1) residue on γ-turn was
not considered during this screening since Asn is oriented
toward solvent exposable part (opposite to JNK3). Out of 285
protocore library, we could generate 5521 fragments library by
chopping and then they were screened for γ-turn mimetics using

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
trimer γ-turn mimetics of the “Φ–X–Φ” hydrophobic motif with
was synthesized from (10c), and introduced to the Phe-
preloaded Wang resin (Scheme 2). The N-Boc protection of
imidazoline was important for three purposes; first, to prevent
side reactions in the SPPS (for Fmoc amino acid to couple with
the nitrogen of the amidine ring); second, for regioselective
Fmoc protection after Cbz removal; finally, because of the
reactivity of the primary amine produced from the removal of
Cbz in compound 10c. Converting compound (10c) to
compound (12) efficiently proceeded under the one-pot reaction
conditions (hydrogenation and Fmoc protection), rather than
stepwise. To synthesize the heptamer peptidomimetic (16) from
compound (12), convergent solid-phase synthesis (C to N
growing CSPPS) was chosen to use efficient amounts of
compound (12). Under microwave conditions, excessive
amounts of coupling reagents (diisopropylcarbodimide (DIC) and
HOAt) accelerated the reaction rate without epimerization.
Through NMR and LCMS analysis, we confirmed that there was
no epimerization during the microwave reaction.[23] Finally, 16
was obtained and identified by ESI-HRMS, MALDI-TOF, and
NMR.
JNK3 were then measured.
Figure 4. (a) Predicted Docking of the Rationally Designed Heptamer (Thick
Tube, Glide Score = -5.98) on the JNK3 Surface (apo of 4H39)[14] (b) 2D Map
of Docking Pose: Red, Negatively Charged; Violet, Positively Charged; Cyan,
Polar; Pale Green, Hydrophobic; Pink Line, Hydrogen Bonding.
Scheme 2. Syntheses of CSPPS of Heptamer Peptidomimetic (16)a.
Scheme 1. Syntheses of γ-Turn Mimetic Motifs (8 & 10)a.
a Reagents and reaction conditions: (i) Pd/C, H2 (g), Fmoc-OSu (1
equiv.), MeOH:THF = 1:1, rt, 5 h; (ii) Wang resin with Phe (0.4
mmol/g, 100–200 mesh, 1% DVB), DIC (9 equiv.), HOAt (6 equiv.),
compound 12 (3 equiv.), DMF (1 mL), MW 50 W (40– 60 °C), 1 h;
(iii) 20% piperidine in DMF (3 mL) 5 min  3 times; (iv) compound 14
(Wang resin linked), DIC (10 equiv.), HOAt (6 equiv.), Boc-Pro-Thr(t-
Bu)-Thr(t-Bu)-CO2H (15, 2 equiv.), DMF (1.6 mL), MW 30 W (38 °C),
1 h and then rt, 12 h; (v) TFA:MC = 1:1 (1 mL), rt, 20 min  2 times;
DIC = diisopropyl carbodiimide, MW = microwave.
a
Reagents and reaction conditions: (i) 1 (1.0 equiv.), 2 (1.2 equiv.),
DIPEA (2.5 equiv.), CH2Cl2, RT, 2 day; (ii) 4% (w/w) OsO4 (0.04
equiv.) in DDW, NMO (1.1 equiv.), THF: DDW = 3: 1, RT, 12 h; (iii)
MsCl (2.2 equiv.), DIPEA (2.2 equiv.), CH2Cl2, from 0 °C to rt, 30
min; (iv) NaN3 (6 equiv.), DMF, 100 ℃, 8 h; (v) Raney Ni, H2 (g),
MeOH, rt, 1 h; (vi) 6 (1.5 equiv.), 7 (1 equiv.), CHCl3:MeOH = 4:3,
from 0 °C to rt, 36 h. (vii) Boc2O (1 equiv.), Na2CO3 (2 equiv.),
THF:DDW = 1:1, from 0 °C to rt, 4 h. Acidic work-up using citric acid
(pH 5–6) for extraction
Affinity of - turn mimetic scaffold
The affinities of the synthesized peptides and peptidomimetics
toward JNK3 were measured by surface plasmon resonance
(SPR) in the table 1. To compare the effect of peptide length
and amino acid side chain on the binding affinity, various peptide
fragments, and peptide including γ-turn mimetics, were
synthesized. First, heptamer (entry b) was synthesized for
comparison with decamer JIP1 (entry a) and heptamer mimetics
To determine key residues for interaction at the pepJIP1,
extended synthesis from compounds (8) was tried. But in case
of 8c, the major product was free carboxylic acid 8c instead of
methyl ester and carbamic acid. We believe that was caused by
ammonia, which was by-product of cyclization. The acid (8c)
was identified using LCMS and protected to produce compound
10c. Then, the mimetic of the “Φ–X–Φ” hydrophobic motif (12)

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
(entry h) to determine the efficiency of heptamer mimetics in
view of length. Second, the affinity presented by the key motif,
“Φ–X–Φ” was investigated through trimer γ-turn mimetics
(compound 8a, 8d, 10b, and 12). Lastly, to investigate the effect
of side chain, γ-turn mimetics with leucine residues (8a, 10b;
entry c, e) were compared with γ-turn mimetics with
phenylalanine residues (8d, 12; entry d, f).
hydrophobic interactions. Although differences in the protecting
groups (Fmoc, Boc, methyl ester) could affect the response unit
(RU) values in the SPR, an identical hydrophobic pocket motif
presented approximately similar Kd values. Most importantly, the
Kd values shown in entries e and g explain the effect of γ-turn
mimetics. In other words, compound (11b), which possesses a
similar structure as compound (10b) except for an imidazoline
ring, exhibited a 4-fold difference in the Kd value. In case of
compound 12, it was 20-fold stronger than compound (11b). In
summary, compound (16), which was a heptamer containing γ-
turn mimetics, showed the highest affinity among all the
compounds, with a much higher binding affinity than heptamer
and even the decamer JIP1 (entry a). Compound (16) was found
to show a similar inhibitory activity in vitro against JNK3 (data
not shown in this report) confirming that it is
a valuable γ-turn mimetic.
Since multivalent interactions show better affinities than
monovalent interaction, the longer peptide (entry a) had a higher
binding affinity than the shorter peptide (entry b). However,
shorter mimetics (16) presented higher affinity than the longer
peptide (entries a) as well as the shorter peptide (entry b),
implying that hydrophobicity in γ-turn mimetics results in
increased affinity. Replacing the Leu residue with a Phe residue
improved the binding affinity in all cases (entry c vs. entry d;
entry e vs. entry f). It was inferred that the phenylalanine residue
could contribute to the binding affinity resulting from increased
Table 1. Affinity (Kd) between JNK3 and Derivatives of Decamer JIP1[a]
Rmax
(RU)
Chi2
/Rmax (%)
Entry
Name
Structure
Kd (µM)
23.6
chi2 [b] (RU)
5.85
Pro-Lys-Arg-Pro-Thr-Thr-Leu-Asn-Leu-
Phe-CO2H
a
Decamer JIP
126.3
60.0
4.6
b
c
Heptamer
8a
Pro-Thr-Thr-Leu-Asn-Leu-Phe-CO2H
79.5
239
0.788
692
1.31
33.3
150
200
302
200
20.8[d]
5.38
0.75
2.11
d
e
f
8d
10b
12
50.8
234
44.7
896
8.06
1.5
1.94
10.0
g
11b
5.03
h
16
2.72
1.98
25
7.9
[a] Kd values were determined by SPR analysis. [b] RU (response unit) was measured under SPR and the chi-squared value was used to judge the quality of
fit between experimental data and that derived from fitted model. The smaller the chi-square, the smaller their variation. For reasonable fitting, the threshold
of a ratio between Rmax and the chi-squared value was determined to 10 % except for entry g. [d] “Chi2/Rmax (= 20.8)” means that Kd of entry g could be
explained with less reliable than other entries.

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
To 0.6 M-(L)-amino acid ester 2 (5.5 g, 38 mmol, 1.2 eq) in methylene
chloride solution, diethyldiisopropylamine (13 mL, 80 mmol, 2.5 eq) and
Conclusions
1,4-dichloro-but-2-ene
1 (3.4 mL, 32 mmol, 1 eq.) was added
sequentially. The reaction mixture was stirred during 2 days at ambient
temperature and then was concentrated without work-up process. The
concentrated mixture was purified by flash column chromatography on
silica gel using EA: Hex (1:1) as the eluent, affording the compound 3 as
white solid (ca. 65 %).
Protein kinases have been investigated intensely as drug targets,
with an increased emphasis on selectivity compared to other
targets; this was even more important in case of non-oncogenic
kinase target. In this context, JNK3, a molecular target for
neurodegenerative diseases, has an allosteric JIP site that could
be explored for highly selective modulation.
As an allosteric JNK3 ligand, γ-turn peptidomimetics of JIP1
were synthesized and their biological affinities were investigated.
For this purpose, the key interaction residues “Φ–X–Φ”
hydrophobic motif in the γ-turn of decamer JIP1 were scaffold-
hopped. Compounds were virtually screened via docking-scoring
to produce compounds (8) as a hit molecule. Compounds 8
were synthesized and their extension, compound (16) were
synthesized via convergent solid-phase peptide synthesis.
Among the synthesized compounds, compound (16; Kd = 7.3
µM) presented a higher affinity than decamer JIP1 (Kd = 27.7
(S)-Methyl 2-(2,5-dihydro-1H-pyrrol-1-yl)-4-methylpentanoate (3a)
The compound 3a was identical with already published data.[10]
(S)-tert-Butyl 2-(2,5-dihydro-1H-pyrrol-1-yl)-4-methylpentanoate (3b)
1H NMR (400 MHz, CDCl3) 5.73 (2H, s), 3.76- 3.81 (2H, m), 3.57-3.61
(2H, m), 3.36 (1H, t, J = 7.6 Hz), 1.57-1.70 (2H, m), 1.47-1.50 (1H, m),
1.42 (9H, s), 0.91 (3H, d, J = 6.4 Hz), 0.89 (3H, d, J = 6.4 Hz) ppm; 13C
NMR (100 MHz, CDCl3) 170.30, 126.57, 81.90, 62.92, 56.15, 39.85,
28.14, 25.16, 22.79, 22.24 ppm; IR (KBr) 3381, 2924, 2516, 1735, 1459
cm-1; HRMS (ESI) calcd for C14H26NO2 [M+H]+: 240.1958, found
240.1955.
(S)-Methyl 2-(2,5-dihydro-1H-pyrrol-1-yl)-3-phenylpropanoate (3c)
1H NMR (300 MHz, CDCl3) 7.16- 7.33 (5H, m), 5.77 (2H, s), 3.73- 3.79
(2H, m), 3.63-3.71 (3H, m), 3.59 (3H, s), 2.99- 3.11 (2H, m) ppm; 13C
NMR (100 MHz, CDCl3) 170.59, 136.31, 128.81, 128.54, 127.02, 63.57,
52.54, 39.48, 29.69 ppm; IR (KBr) 3332, 2917, 1742, 1638, 1457 cm-1;
HRMS (ESI) calcd for C14H18NO2 [M+H]+: 232.1332, found 232.1311.
µM) showing that compound (16) could be
a potential
neuroprotective agent targeting the JNK3-JIP binding site.
The selective inhibition of JNK3 with the ATP-competitive small
molecules would be very challenging owing to the similarities in
active site of all terminal protein kinase in signaling cascade (eg.
p38a/b, ERK1/2) and JNK isoforms. In this context, the allosteric
γ-turn peptidomimetics developed in this study are an excellent
platform for selective inhibition of JNK3 signaling. We expect
that peptidomimetics toward JIP site can be expanded to
bivalent ligand with high potency and selectivity by combining
them with ATP site inhibitors per se.
Di-substituted internal olefin of the compound 3 was oxidized under the
condition of 4 % (w/w) osmium tetroxide (5.1 mL, 0.8 mmol, 0.04 eq) as
an oxidant and NMO (2.6 g, 22 mmol, 1.1 eq) as a co-oxidant to produce
the di-hydroxylated meso compound as vicinal syn type. The compound
without any purification was dissolved in anhydrous methylene chloride
and became mesylated by adding diethyldiisopropylamine (7.7 mL, 44
mmol, 2.2 eq) and mesyl chloride (3.4 mL, 44 mmol, 2.2 eq) one after the
other to the reaction mixture. The reaction mixture was diluted with DDW
and methylene chloride and washed with aqueous NaHCO3 (sat) and
saturated aqueous sodium chloride solution. The organic phase was
dried over (drying agent: sodium sulfate) and concentrated in vacuo. The
crude product was purified by flash column chromatography on silicagel
using EA: Hex (3:1) as the eluent, affording the mesylated compound 4
as white solid (12 mmol, two step yield ca. 80 %).
The γ-turn peptidomimetics described in this study are also
applicable for general γ-turn peptidomimetics in protein-protein
interaction. Specially, the application would be versatile since
the mimetic can be readily synthesized from any amino acids.
Experimental Section
(S)-Methyl
methylpentanoate (4a)
2-(syn-3,4-bis((methylsulfonyl)oxy)pyrrolidin-1-yl)-4-
Instrumentation and chemicals
All chemicals (reagent grade) used were purchased from Aldrich and Alfa
aesar (USA). Separation of the compounds by column chromatography
was carried out with silica gel 60 (200–300 mesh ASTM, E. Merck,
Germany). The quantity of silica gel used was 50–100 times the weight
charged on the column. Thin layer chromatography (TLC) was run on the
silica gel coated aluminum sheets (silica gel 60 GF254, E. Merck,
Germany) and visualized in ultraviolet (UV) light (254 nm). 1H NMR and
13C NMR spectra were recorded on Varian unity Plus 300 MHz and
Brucker Avance 400 MHz and 600 MHz spectrometer at 25 oC, using
tetramethylsilane (TMS) as the internal standard. High-resolution MS
(HR/MS) experiments were conducted with a Finnigan LTQ Orbitrap
mass spectrometry (Thermo Fisher Scientific Inc, MA, USA) operated in
positive-ion electrospray mode. Infrared (IR) spectra were recorded on a
JASCO FT/IR-300E and Perkin-Elmer 1710 FT spectrometer. Waters
autopurification system was used for isolation of acid products
(compound 10c, compound 12 and compound 16) under the condition of
1H NMR (400 MHz, CDCl3) 5.11- 5.17 (2H, m), 3.70 (3H, s), 3.40- 3.45
(2H, m), 3.22-3.23 (2H, m), 3.05- 3.12 (1H, m), 3.07 (6H, s), 1.47-1.70
(3H, m), 0.91 (3H, d, J = 6.4 Hz), 0.88 (3H, d, J = 6.4 Hz) ppm; 13C NMR
(100 MHz, CDCl3) 170.30, 74.15, 65.52, 53.83, 52.93, 39.10, 38.54,
37.46, 25.22, 23.11, 21.14 ppm; IR (KBr) 3391, 2917, 1734, 1460, 1364,
1177, 758 cm-1; HRMS (ESI) calcd for C13H26NO8S2 [M+H]+: 388.1094,
found 388.1074.
(S)-tert-Butyl
2-(syn-3,4-bis((methylsulfonyl)oxy)pyrrolidin-1-yl)-4-
methylpentanoate (4b)
1H NMR (400 MHz, CDCl3) 5.10- 5.28 (2H, m), 3.41- 3.46 (1H, m), 3.25-
3.29 (1H, m), 3.21- 3.2 (2H, m), 3.09 (6H, s), 3.04- 3.12 (1H, m), 1.59-
1.67 (1H, m), 1.48-1.55 (2H, m), 1.45 (9H, s), 0.90 (3H, d, J = 6.4 Hz),
0.87 (3H, d, J = 6.4 Hz) ppm; 13C NMR (100 MHz, CDCl3) 171.62, 81.95,
76.27, 75.86, 61.86, 52.36, 51.43, 39.21, 38.53, 38.45, 28.11, 24.93,
22.43, 22.25 ppm; IR (KBr) 2920, 1721, 1467, 1365, 1176, 756 cm-1;
HRMS (ESI) calcd for C16H32NO8S2 [M+H]+: 430.1564, found 430.1545.
2545 pump, 3100 single quadrupole mass detector and
a 2998
photodiode array detector using a Waters Sunfire 5 μm (I.D. 150 × 2.1
mm) column at a flow rate of 20 mL min-1 (eluents: DDW and 0.035 %
formic acid MeOH solution). All solid phase synthesis was check by
KAISER kit. If it needed, small portion of reacted solid was cleaved by
TFA in CH2Cl2 and then was detected by LCMS (Waters autopurification).
(S)-Methyl
2-(syn-3,4-bis((methylsulfonyl)oxy)pyrrolidin-1-yl)-3-
phenylpropanoate (4c)
1H NMR (300 MHz, DMSO-d6) 7.18- 7.30 (5H, m), 5.16- 5.17 (2H, m),
3.70 (1H, t, J = 7.6 Hz), 3.57 (3H, s), 3.25 (3H, s), 3.24 (3H, s), 3.20- 3.33
(2H, m), 2.98- 3.04 (2H, m), 2.89- 2.97 (2H, m) ppm; 13C NMR (100 MHz,
Synthesis of -turn mimetic derivatives (8a~8c)

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
CDCl3) 171.51, 136.95, 128.77, 128.44, 126.73, 76.09, 75.66, 64.91,
52.38, 51.91, 51.64, 38.50, 38.42, 36.36; IR (KBr) 2918, 1730, 1455,
1357, 1174, 752 cm-1; HRMS (ESI) calcd for C16H24NO8S2 [M+H]+:
422.0938, found 422.0919.
(S)-tert-Butyl 2-((S)-1-((((9H-fluoren-9-yl)methoxy)carbonyl)amino)-3-
methylbutyl)-3a,4,6,6a-tetrahydropyrrolo [3,4-d]imidazol-5(1H)-yl)-4-
methylpentanoate (8b)
1H NMR (400 MHz, CDCl3) 7.71-7.73 (2H, m), 7.54-7.61 (2H, m), 7.35-
7.38 (2H, m), 7.24-7.28 (2H, m), 4.95 (1H, brs), 4.68 (1H, brs), 4.16- 4.26
(3H, m), 3.62- 3.80 (4H, m), 3.26 (1H, brs), 2.62 (1H, m), 1.85 (1H, brs),
1.74 (2H, brs), 1.54- 1.60 (3H, m), 1.44 (9H, s), 0.85- 0.93 (12H, m) ppm;
13C NMR (100 MHz, CDCl3) 172.99, 172.57, 171.75, 157.39, 143.57,
141.16, 141.09, 127.71, 127.17, 127.09, 125.54, 119.82, 81.46, 67.75,
61.45, 60.92, 60.68, 57.61, 52.90, 46.94, 46.68, 40.76, 39.25, 29.66,
28.15, 25.29, 24.55, 24.45, 22.62, 21.96, 21.25 ppm; IR (KBr) 3356,
2920, 1673, 1144, 1045 cm-1; HRMS (ESI) calcd for C35H49N4O4 [M+H]+:
589.3748, found 589.3706.
Mesyl functional group of the compound 4 (1.34 mmol, 1 eq) was
substituted with azide group under the condition of sodium azide (523 mg,
8.04 mmol, 6 eq) at 100℃. To avoid explosion, scraping diazide 5 was
restrained. The reaction mixture was filtrated, concentrated in vaccuo
and then purified by purified by flash column chromatography on silicagel
using EA: Hex (5:1) as the eluent, affording the product 5 as colorless oil
(1.07 mmol, 2 step yield ca. 80 %).
(S)-Methyl 2-(syn-3,4-diazidopyrrolidin-1-yl)-4-methylpentanoate (5a)
1H NMR (400 MHz, CDCl3) 3.96- 4.06 (2H, m), 3.68 (3H, s), 3.41 (1H, dd,
J = 8.0 Hz, J = 7.2 Hz), 3.27 (1H, dd, J = 9.6 Hz, J = 7.2 Hz), 3.09 (1H,
dd, J = 9.6 Hz, J = 6.0 Hz), 2.95 (1H, dd, J = 9.6 Hz, J = 4.4 Hz), 2.84
(1H, dd, J = 9.6 Hz, J = 6.0 Hz), 1.60-1.70 (1H, m), 1.47- 1.57 (2H, m),
0.91 (3H, d, J = 6.4 Hz), 0.87 (3H, d, J = 6.4 Hz) ppm; 13C NMR (100
MHz, CDCl3) 173.30, 61.77, 61.69, 61.06, 52.91, 51.38, 50.90, 39.44,
24.86, 22.71, 22.00; IR (KBr) 3317, 2917, 2849, 2101, 1731, 1464, 1259,
1158 cm-1; HRMS (ESI) calcd for C11H20N7O2 [M+H]+: 282.1673, found
282.1647.
(S)-tert-Butyl 2-((S)-1-((((9H-fluoren-9-yl)methoxy)carbonyl)amino)-
3-methylbutyl)-3a,4,6,6a-tetrahydropyrrolo [3,4-d]imidazol-5(1H)-yl)-
4-methylpentanoate (8b)
1H NMR (400 MHz, CDCl3) 7.71-7.73 (2H, m), 7.54-7.61 (2H, m), 7.35-
7.38 (2H, m), 7.24-7.28 (2H, m), 4.95 (1H, brs), 4.68 (1H, brs), 4.16- 4.26
(3H, m), 3.62- 3.80 (4H, m), 3.26 (1H, brs), 2.62 (1H, m), 1.85 (1H, brs),
1.74 (2H, brs), 1.54- 1.60 (3H, m), 1.44 (9H, s), 0.85- 0.93 (12H, m) ppm;
13C NMR (100 MHz, CDCl3) 172.99, 172.57, 171.75, 157.39, 143.57,
141.16, 141.09, 127.71, 127.17, 127.09, 125.54, 119.82, 81.46, 67.75,
61.45, 60.92, 60.68, 57.61, 52.90, 46.94, 46.68, 40.76, 39.25, 29.66,
28.15, 25.29, 24.55, 24.45, 22.62, 21.96, 21.25 ppm; IR (KBr) 3356,
2920, 1673, 1144, 1045 cm-1; HRMS (ESI) calcd for C35H49N4O4 [M+H]+:
589.3748, found 589.3706.
(S)-tert-Butyl 2-(syn-3,4-diazidopyrrolidin-1-yl)-4-methylpentanoate
(5b)
1H NMR (400 MHz, CDCl3) 3.96- 4.05 (2H, m), 3.25- 3.34 (2H, m), 3.09-
3.13 (1H, m), 2.97 (1H, dd, J = 9.6 Hz, J = 4.4 Hz), 2.84 (1H, dd, J = 9.6
Hz, J = 5.6 Hz), 1.60-1.68 (1H, m), 1.49- 1.53 (2H, m), 1.44 (9H, s), 0.91
(3H, d, J = 6.4 Hz), 0.87 (3H, d, J = 6.4 Hz) ppm; 13C NMR (100 MHz,
CDCl3) 170.62, 82.10, 64.47, 61.16, 60.59, 52.45, 51.85, 37.79, 27.98,
25.30, 23.14, 21.27 ppm; IR (KBr) 2916, 2104, 1732, 1465, 1376, 1148
cm-1; HRMS (ESI) calcd for C14H26N7O2 [M+H]+: 324.2142, found
324.2134.
(2S)-Methyl-2-(2-((S)-1-((tert-butoxycarbonyl)amino)-2-phenylethyl)-
3a,4,6,6a-tetrahydropyrrolo[3,4-d]imidazol-5(1H)-yl)-3-
phenylpropanoate (8d)
1H NMR (400 MHz, CDCl3) 7.05-7.24 (10H, m), 4.73 (1H, brs), 4.40- 4.50
(2H, m), 3.49- 3.65 (5H, m), 3.23- 3.25 (1H, m), 2.75- 3.04 (8H, m), 2.62
(1H, m), 1.22- 1.38 (9H, m) ppm; 13C NMR (100 MHz, CDCl3) 171.01,
169.81, 155.43, 136.95, 135.95, 129.15, 128.90, 128.46, 127.90, 126.31,
126.09, 118.08, 115.15, 79.19, 64.28, 61.69, 59.87, 57.76, 54.09, 50.74,
49.65, 38.24, 36.37, 27.77 ppm; HRMS (ESI) calcd for C28H37N4O4
[M+H]+: 493.2809, found 493.2773.
(S)-Methyl 2-(syn-3,4-diazidopyrrolidin-1-yl)-3-phenylpropanoate (5c)
1H NMR (300 MHz, DMSO-d6) 7.19- 7.28 (5H, m), 4.26- 4.27 (2H, m),
3.62- 3.66 (1H, t, J = 7.6 Hz), 3.55 (3H, s), 3.08- 3.16 (2H, m), 2.81- 2.94
(2H, m), 2.76- 2.82 (2H, m); 13C NMR (100 MHz, CDCl3) 171.84, 137.26,
128.71, 128.29, 126.51, 65.05, 61.54, 61.41, 52.63, 51.72, 51.27, 36.63;
IR (KBr) 2952, 2099, 1730, 1455, 1261 cm-1; HRMS (ESI) calcd for
C14H18N7O2 [M+H]+: 316.1516, found 316.1501.
Synthesis of -turn mimetic derivative (10b) and
compound (11b)
The compound 8b (mmol, 1eq) was added to 4M-HCl in dioxane or TFA
in methylene chloride (volume ration; 1:3) and the reaction was detected
by LCMS. When the compound 8b disappears in LCMS, all reagent and
solvent of the reaction mixture was evaporated in vacuo and removed.
The concentrated product was Boc protected under the next conditions
that (i) method A: Boc2O (1 eq), NaHCO3 (2 eq), THF: DDW =1:1 or (ii)
method B: Boc2O (1 eq), DIPEA (3 eq), CH2Cl2 (anhydrous). Until mass
peak of substrate decreased, the reaction was stirred and major product
was compound 11b instead of compound 10b. The reaction mixture was
purified by prep LCMS and product fractions were collected and
evaporated after adding benzene at ambient temperature. DDW of the
collected solution was removed under freezing dry to afford compound
11b, compound 10b each.
The di-azide compound 5 was reduced under the condition of Raney-
nickel of Raney-nickel and H2 (gas) to form 3,4-diamino-pyrrolidine core
of the compound 6. The compound 6 without any purification process
was used in the cyclization in order to form pyrrolo[3,4-d]imidazole
bicycle ring as a -turn mimetic scaffold. According to previous report, the
imidate compound 7 (0.71 mmol, 1 eq) synthesized was dissolved in
anhydrous chloroform. The crude diamine compound 6 (roughly 1.5 eq)
solution in anhydrous methanol was transferred to the imidate mixture by
cannula and stirred at ambient temperature until the product spot didn’t
increase any more in the thin layer chromatography. The reaction mixture
was concentrated and purified by prep LCMS (Waters autopurification,
eluent was DDW and MeOH containing TFA) to afford the final product 8.
(2S)-Methyl-2-(2-((S)-1-((tert-butoxycarbonyl)amino)-3-methylbutyl)-
3a,4,6,6a-tetrahydropyrrolo[3,4-d]imidazol-5(1H)-yl)-4-
tert-Butyl-2-((S)-1-((((9H-fluoren-9-yl)methoxy)carbonyl)amino)-3-
methylbutyl)-5-((S)-1-methoxy-4-methyl-1-oxopentan-2-yl)-4,5,6,6a-
tetrahydropyrrolo[3,4-d]imidazole-1(3aH)-carboxylate as the methyl
methylpentanoate (8a)
1H NMR (400 MHz, CDCl3) 6.92 (1H, brs), 4.69 (1H, brs), 4.51 (1H, brs),
3.68 (2H, s), 3.54 (3H, s), 2.86- 3.18 (4H, m), 2.00 (1H, brs), 1.87 (1H, s),
1.58 (4H, s), 1.43 (9H, s), 0.87- 0.92 (12H, m); 13C NMR (100 MHz,
CDCl3) 173.27, 171.95, 80.73, 60.88, 60.53, 53.53, 52.88, 51.61, 47.89,
46.63, 40.29, 38.90, 28.06, 24.58, 24.40, 22.62, 22.36, 22.00, 20.91; IR
(KBr) 2960, 2929, 1671, 1169 cm-1; HRMS (ESI) calcd for C22H41N4O4
[M+H]+: 425.3122, found 425.3089.
ester form of compound (10b)
Because it was difficult to assign H-NMR of compound 10b, N-Boc
protection of amidine ring in compound 8b was conducted to confirm H-
NMR assignment of compound 8b more confidently; 1H NMR (400 MHz,
CDCl3) 7.74-7.76 (2H, m), 7.63-7.66 (2H, m), 7.36- 7.40 (2H, m), 7.29-
7.32 (2H, m), 5.50 (1H, brs), 4.49 (2H, brs), 4.33- 4.49 (2H, m), 4.22-
4.26 (1H, m), 3.67 (3H, s), 3.46- 3.77 (2H, m), 2.80- 3.15 (4H, m), 1.83-

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
1.84 (2H, m), 1.59- 1.68 (4H, m), 1.53 (9H, s), 0.85-1.02 (12H, m) ppm;
IR (KBr) 2921, 2852, 1822, 1732, 1466, 1256, 1159, 741 cm-1; HRMS
(ESI) calcd for C36H49N4O6 [M+H]+:633.3647, found 633.3612.
129.81, 128.83, 128.77, 128.61, 128.56, 128.36, 128.32, 128.11, 127.59,
127.03, 126.93, 126.92 , 126.78, 125.36, 125.29, 124.01, 119.84, 81.46,
67.00, 66.85, 60.42, 57.99, 57.58, 51.28, 51.19, 51.14, 47.08, 42.43,
37.11, 36.99, 36.78, 32.76, 31.94, 29.72 ppm; IR (KBr) 3374, 1648,
1023, 990 cm-1; HRMS (ESI) calcd for C43H47N4O6 [M+H]+: 715.3490,
found 715.3472.
Compound 11b
1H NMR (400 MHz, DMSO-d6) 7.88 (2H, d, J= 7.6 Hz), 7.71 (2H, d, J=
4.8 Hz), 7.52-7.54 (1H, m), 7.39- 7.43 (2H, m), 7.31-7.33 (2H, m), 6.36-
6.42 (1H, m), 4.75 (1H, brs), 4.20- 4.28 (4H, m), 3.90- 4.03 (3H, m), 2.99-
3.16 (2H, m), 2.66-2.67 (1H, m), 2.54- 2.55 (1H, m), 1.90- 1.97 (1H, m),
1.59- 1.60 (1H, m), 1.40- 1.48 (4H,m), 1.31- 1.35 (9H, m), 0.82- 0.88
(12H, m); 13C NMR (150 MHz, DMSO-d6) 156.43, 156.31, 155.46, 155.43,
144.36, 144.10, 141.18, 128.10, 127.49, 125.79, 125.71, 120.57, 79.66,
79.44, 79.22, 70.23, 65.98, 63.23, 47.14, 29.46, 28.60, 28.55, 25.02,
24.94, 24.63, 23.64, 23.49, 22.73, 21.82 ppm; IR (KBr) 3339, 2920,
1687, 1527, 1451, 1366, 1248, 1165 cm-1; HRMS (ESI) calcd for
C36H51N4O7 [M+H]+: 651.3752, found 651.3716
Synthesis of -turn mimetic derivatives 16
Wang resin with pre-loaded phenylalanine (23 umol, 0.4 mmol/g, 100-
200 mesh, 1% DVB) was treated with 20% piperidine/ DMF (3 × 3 mL × 5
min) and subsequently washed with DMF (5 × 4 mL), CH2Cl2 (5 × 4 mL),
and DMF (5 × 4 mL) to remove Fmoc. A solution of compound 9d (69
umol, 3eq), diisopropyl carbodiimide (207 umol, 9eq) and HOAt (138
umol, 6eq) in DMF (1 mL) was pre-activated for 20 min with LCMS peak
check and then added to the resin. The suspension agitated for 1 h under
the microwave condition (fixing 50 w and continuous cooling in range of
40~ 60 ℃). The resin was then washed with DMF (5 × 4 mL), CH2Cl2 (5
× 4 mL), and DMF (5 × 4 mL). The compound 14 (Wang resin linked, 12
umol) was treated with 20% piperidine/ DMF (3 × 3 mL × 5 min) and
subsequently washed with DMF (5 × 4 mL), CH2Cl2 (5 × 4 mL), and DMF
(5 × 4 mL) to remove Fmoc. The customized peptide Pro-Thr(t-Bu)-Thr(t-
Bu)-CO2H (25 umol, 2 eq), diisopropyl carbodiimide (123 umol, 10eq)
and HOAt (74 umol, 6eq) in DMF (1.6 mL) was pre-activated for 30 min
with LCMS peak check and then added to the resin with compound 14.
The suspension agitated for 1 h under the microwave condition (fixing 30
w and continuous cooling to be at ca. 38℃) and then additional agitated
for 12 hr at ambient temperature without microwave. The resin was then
washed with DMF (5 × 4 mL), CH2Cl2 (5 × 4 mL), and DMF (5 × 4 mL)
and then treated with TFA: CH2Cl2= 1:1 (1mL) at ambient temperature (2
× 20 min × 1mL). The resin washed subsequently TFA in CH2Cl2 (4 × 1
mL) and the combined cleavage solution were concentrated under
reduced pressure at ambient temperature to afford the crude peptide,
which was subjected to purification by LCMS.
Synthesis of -turn mimetic derivative (12)
The di-azide compound 5c was reduced under the condition of Raney-
nickel and H2 (gas). The reduced crude product (roughly 1.5 eq) was
dissolved by anhydrous methanol and transferred to the imidate
compound 7c with Cbz group (0.71 mmol, 1 eq) in anhydrous chloroform
by cannula. The reaction mixture was stirred at ambient temperature until
the substrate ratio didn’t decrease any more in LCMS. The reaction
mixture was concentrated and dissolved in tetrahydrofuran and DDW
(1:1). Na2CO3 (1.4 mmol, 2 eq) was added to the solution and then Boc
anhydride (0.7 mmol, 1eq) was added. The reaction solution purified by
prep LCMS (Waters autopurification, eluent was DDW and MeOH
containing 0.035% formic acid) and product fractions were collected and
evaporated after adding benzene at ambient temperature. DDW of the
collected solution was removed under freezing dry to afford the
compound 10c. Because it was difficult to assign H-NMR of compound
10c, methyl ester of compound 10c was used to chracterize H-NMR
assignment of compound 10c more confidently.
(S)-2-((S)-2-(syn-2-((S)-1-((2S,3R)-3-hydroxy-2-((2S,3R)-3-hydroxy-2-
((R)-pyrrolidine-2-carboxamido)butanamido)butanamido)-2-
phenylethyl)-3a,4,6,6a-tetrahydropyrrolo[3,4-d]imidazol-5(1H)-yl)-3-
tert-butyl-2-((S)-1-(((benzyloxy)carbonyl)amino)-2-phenylethyl)-5-
((S)-1-methoxy-1-oxo-3-phenylpropan-2-yl)-4,5,6,6a-
phenylpropanamido)-3-phenylpropanoic acid (16)
tetrahydropyrrolo[3,4-d]imidazole-1(3aH)-carboxylate as the methyl
1H NMR (400 MHz, DMSO-d6) 9.83-9.98 (1H, m), 9.17-9.18 (1H, m),
8.52- 8.61 (1H, m), 8.31 (1H, d, J = 7.6 Hz), 7.77 (1H, d, J = 8.4 Hz),
7.11-7.30 (15H, m), 6.98 (1H, s), 5.12- 5.18 (2H, m), 4.50- 4.56 (2H, m),
4.41-4.45 (1H, m), 4.29-4.392 (2H, m), 4.22- 4.25 (1H, m), 4.01-4.02 (2H,
m), 2.92- 3.00 (3H, m), 2.71- 2.81 (3H, m), 2.67 (1H, m), 2.33- 2.40 (4H,
m), 1.77- 1.97 (4H, m), 0.98-1.07 (6H, m) ppm; IR (KBr) 2924, 1726,
1497, 1450, 1141, 1033, 700 cm-1; HRMS (ESI) calcd for C44H57N8O8
[M+H]+: 825.4294, found 825.4285.
ester form of compound (10c)
Despite methyl ester formation to acquire clear spectra, purity in H-NMR
was low. 1H NMR (600 MHz, CDCl3) 7.14-7.33 (15H, m), 5.57- 5.60 (1H,
m), 4.95- 5.10 (2H, m), 4.50-4.37 (2H, m), 3.54-3.72 (5H, m), 3.04- 3.11
(2H, m), 2.92- 2.99 (4H, m), 2.73- 2.85 (2H, m), 1.27 (9H, s) ppm; IR
(KBr) 3341, 2917, 1638 cm-1; HRMS (ESI) calcd for C36H43N4O6 [M+H]+:
627.3177, found 627.3172.
Cbz group of the compound 10c (0.30 mmol, 1 eq) was deprotected and
reprotected under the condition of catalytic Pd/C, H2 gas and Fmoc-OSu
(0.30 mmol, 1eq) without additional organic base. The reaction was
detected by LCMS and terminated until ratio of product UV peak doesn’t
any more increase. The reaction solution purified by LCMS (Waters
autopurification, eluent was DDW and MeOH containing 0.035% formic
acid) and product fractions were collected and evaporated after adding
benzene at ambient temperature. DDW of the collected solution was
removed under freezing dry to afford the final compound 12.
SPR measurement & analysis
JNK3 was cloned, expressed and purified as our previous publication.[13]
As a direct binding assay for JNK3, surface Plasmon resonance (SPR)
measurements were performed on a ProteOn™ XPR36 instrument (Bio-
Rad Laboratories Inc.). All reagents and buffers used were purchased
from Bio-Rad Laboratories Inc. JNK3 was immobilized (average
25,000RU) by standard amine coupling chemistry on a GLH chip. Binding
experiments were performed using PBST buffer (phosphate buffered
saline, pH 7.4, 0.005% Tween 20) as the running buffer. Compounds
were dissolved in 10 mM DMSO and subsequently diluted with PBST
buffer to produce a 1% DMSO solution. In order to calculate Kd value,
each 100 M test compound in 1% DMSO PBST solution was serially
diluted threefold from 100 to 0.01 M using 1% DMSO in PBST solution
to adjust the final DMSO content and the concentration of the respective
compound. Five concentrations were selected from among the nine serial
concentrations and blank PBST with 1% DMSO was injected into six
analytes in the SPR instrument. The optimal concentration range was
finally determined from at least three measurements. To obtain a reliable
deviation value of the response unit between apo protein and binding
tert-Butyl 2-((S)-1-((((9H-fluoren-9-yl)methoxy)carbonyl)amino)-2-
phenylethyl)-5-((S)-1-methoxy-1-oxo-3-phenylpropan-2-yl)-4,5,6,6a-
tetrahydropyrrolo[3,4-d]imidazole-1-carboxylate; as the methyl ester
form of compound 12
1H NMR (600 MHz, CDCl3) 7.73-7.70 (4H, m), 7.51-7.58 (4H, m), 7.26-
7.39 (10H, m), 5.62- 5.66 (1H, m), 4.46-4.59 (2H, m), 4.31- 4.40 (2H, m),
4.11- 4.23 (m, 3H), 3.72-3.82 (1H, m), 3.66 (3H, s), 3.56- 3.58 (1H, m),
3.22- 3.28 (1H, m), 3.10-3.18 (1H, m), 2.95- 3.03 (3H, m), 2.86- 2.88 (1H,
m), 2.79- 2.83 (1H, m), 1.26 (9H, s) ppm; 13C NMR (150 MHz, CDCl3)
171.83, 171.65, 148.99 144.02, 143.92, 141.22, 140.52, 130.22, 130.03,

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
protein with compound, the baseline was standardized using inter spot
references and a blank solution containing 1% DMSO. A curve was fit to
the raw data using the Langmuir equation in ProteOn Manager™ in next
figures. However, the kinetic fitting method for kd calculation presented
low reproducibility in this study as well as high square value. Therefore,
dynamic equilibrium fitting method was chosen for kd calculation.
replacing the ɤ-turn motif, we began scaffold hopping, which is a virtual
screening of various fragments in silico based on crystal structure of JIP-
JNK. Out of 285 protocore library, we could generate 5521 fragments
library by chopping and then they were screened for ɤ-turn mimetics
using CombiGlide module in Schrodinger package in the Figure 3. The
hit structures were sorted by docking score and also modified through
rational design and redocking analysis. The compound (8),
hexahydropyrroloimidazole bicyclic ring was selected as a first candidate
for synthesis as the ɤ-turn mimetics. The binging feature of ɤ-turn
incorporated JIP peptide on JIP site of JNK3 was considered. Since the
hydrophobic residue Leu seems to be essential for JIP-JNK interaction,
we also planned to introduce Phe residue as a hydrophobic replacement
in either or both of “X” hydrophobic motif. Molecular docking of
compound 16 into the structure of JNK3 (PDB code: 4H39) produced
predictive docking pose, (1) SP (standard precision) and XP (extra
precision) docking of compound 16 and peptides, (2) revision of the
docking poses through substructure energy minimization using
Schrodinger Macro Model, and (3) scoring of revised docking pose.
Docking simulations
The 3D X-ray protein structure of JNK3 as a complex with decamer JIP1
was obtained from the PDB (code: 4H39) and prepared using Protein
Preparation Wizard of the Schrodinger Maestro program. The structures
of new designed inhibitors were drawn using Chemdraw, and their 3D
conformations were generated using the Schrodinger LigPrep program
with the OPLS 2005 force field. Redocking of decamer JIP1 into the
structure of JNK3 (PDB code: 4H39) was carried out using Schrodinger
Glide (whole software package; Version 9.2). When making an optimal
grid file, a grid box was manually adjusted from a cube into a cuboid,
constraints (hydrogen bonding & hydrophobic cube) were added, and
unwanted pockets was treated as excluded volume. The optimal grid file
was validated by RMSD between x-ray decamer JIP1 and redocked pose
of the decamer. After making the optimal grid file, to find eligible scaffold
2, 198−206 ; c) C. E. Crocker, S. Khan, M. D. Cameron, H. A.
Robertson, G. S. Robertson, P. LoGrasso, ACS Chem. Neurosci. 2011,
2, 207−212.
Acknowledgments
[13] D. K. Morrison, R. J. Davis, Annu. Rev. Cell Dev. Biol. 2003, 19, 91-
118.
This research was supported by the Basic Science Research
Program through the National Research Foundation of Korea
(NRF) funded by the Ministry of Education, Science and
Technology (J-M. Hah: NRF-2012R1A1A2A10038773) (M.-h.
Kim: NRF-2012R1A6A3A04038302).
[14] J. D. Laughlin, J. C. Nwachukwu, M. Figuera-Losada, L. Cherry, K. W.
Nettles, P. V. LoGrasso, Structure 2012, 20, 2174-2184.
[15] a) J. M. Combette, C. Deloche, WO2013079213; b) C. H. Nijboer, H. J.
C. Bonestroo, J. Zijlstra, A. Kavelaars, C. J. Heijnen, Neurobiol. Dis.
2013, 54, 432-444; c) T. S. Kaoud, S. Mitra, S. Lee, J. Taliaferro, M.
Cantrell, K. D. Linse, C. L. Van Den Berg, K. N. Dalby, ACS Chem Biol.
2011, 6, 658-666; d) J. Pan, J. Qian, Y. Zhang, J. Ma, G. Wang, Q. Xiao,
S. Chen, J. Ding, Lab. Invest. 2010, 90, 156-167.
Keywords: peptidomimetic; gamma turn; allosteric; scaffold
[16] Merely BI-78D3 was published as a non-peptide drug that binds with JIP
binding site but doesn’t originate from structure of JIP. J. L. Stebbins, S.
K. De, T. Machleidt, B. Becattini, J. Vazquez, C. Kuntzen, L. H. Chen, J.
F. Cellitti, M. Riel-Mehan, A. Emdadi, G. Solinas, M. Karin, M.
Pellecchia, Proc. Natl. Acad. Sci. U S A 2008, 105, 16809-16813.
[17] a) F. A. Etzkorn, J. M. Travins, S. A. Hart in Advances in Amino Acid
Mimetics and Peptidomimetics, Vol. 2 (Ed.: A. Abell), Jai Press Inc.,
Stamford, CT, 1999, pp. 125-163; b) Adv. Drug. Res. 1997, 29, 1; c) K.
Eunhwa, J. Liu, L. M. Perez, G. Lu, A. Schaefer, K. Burgess, J. Am.
Chem. Soc. 2011, 133, 462.
hopping; JNK 3; kinase inhibitor;
[1]
[2]
A. M. Manning, R. J. Davis, Nat. Rev. Drug. Discov. 2003, 2, 554-565.
J. J. Ventura, N. J. Kennedy, J. A. Lamb, R. A. Flavell, R. J. Davis, Mol.
Cell Biol. 2003, 23, 2871–2882.
[3]
[4]
[5]
[6]
C. Tournier, et al. Science 2000, 288, 870–874.
K. Sabapathy, Prog. Mol. Biol. Transl. Sci. 2012, 106, 145-169.
A. A. Mohit, J. H. Martin, C. A. Miller, Neuron 1995, 14, 67−78.
B. Derijard, M. Hibi, I. H. Wu, T. Barrett, B. Su, T. Deng, M. Karin, R. J.
Davis, Cell 1994, 76, 1025−1037.
[18] a) L. R. Whitby, D. L. Boger, Acc. Chem Res. 2012, 45, 10, 1698–1709;
b) R. F. Hirschmann, K. C. Nicolaou, A. R. Angeles, J. S. Chen, A. B.
Smith, III, Acc. Chem. Res., 2009, 42, 10, 1511–1520.
[7]
[8]
M. Hibi, A.Lin, T. Smeal, A. Minden, M. Karin, Genes Dev. 1993, 7,
2135−2148.
a) J. M. Kyriakis, P. Banerjee, E. Nikolakaki, T. Dai, E. A. Rubie, M. F.
Ahmad, J. Avruch, J. R. Woodgett, Nature 1994, 369, 156−160.; b) L. F.
Chang, Y. Jones, M. H. Ellisman, L. S. B. Goldstein, M. Karin, Dev. Cell
2003, 4, 521−533.
[19] a) W. Chang, R. T. Mosley, S. Bansal, M. Keilman, A. M. Lam, P. A.
Furman, M. J. Otto, M. J. Sofia, Bioorg. Med. Chem. Lett. 2012, 22,
2938-2942; b) K. Guitot, S. Carboni, O. Reiser, U. Piarulli, J. Org. Chem.
2009, 74, 8433-8436; c) J. S. Park, K. R. Kim, H. Y. Nam, C. E. Yeom,
C. Chough, S. H. Kwon, S. Ro, D. K. Shin, B. M. Kim, Protein. Pept. Lett.
2008, 15, 980; d) J. Georgsson, C. Sköld, B. Plouffe, G. Lindeberg, M.
Botros, M. Larhed, F. Nyberg, N. Gallo-Payet, A. Gogoll, A. Karlén, A.
Hallberg, J. Med. Chem. 2005, 48, 21, 6620-6631; e) H. S. Iden, W.
D. Lubell, Org. Lett. 2006, 8, 3425-3428.
[9]
a) J. Peng, J. Andersen, IUBMB Life 2003, 55, 267-271; b) X. Antoniou,
M. Falconi, D. Di Marino, T. Borsello, J. Alzheimer’s Dis. 2011, 24, 633-
642; c) C. E. Crocker, S. Khan, M. D. Cameron, H. A. Robertson, G. S.
Robertson, P. LoGrasso, ACS Chem. Neurosci. 2011, 2, 207-212.
[10] a) E. K. Kim, E. J. Choi, Biochim. Biophys. Acta, Mol. Basis Dis, 2010,
1802, 396−405.; b) B. Salh, Expert Opin. Ther. Targets 2007, 11,
1339−1353.; c) D. D. Yang, C. Y. Kuan, A. J. Whitmarsh, M. Rincon, T.
S. Zheng, R. J. Davis, P. Rakic, R. A. Flavell, Nature 1997, 389,
865−870.
[20] F. Cuevas, C. García-Granda, H. Buschmann, A. Torrens, S. Yenes, M.
A. Pericàs, Synlett. 2007, 1, 119-122.
[21] D. Kuppert, J. Sander, C. Roth, M. Wörle, T. Weyhermüller, G. J. Reiss,
U. Schilde, I. Müller, K. Hegetschweiler, Eur. J. Inorg. Chem. 2001, 10,
2525-2542.
[11] S. O. Yoon, D. J. Park, J. C. Ryu, H. G. Ozer, C. Tep, Y. J. Shin, T. H.
Lim, L. Pastorino, A. J. Kunwar, J. C. Walton, A. H. Nagahara, K. P. Lu,
R. J. Nelson, M. H. Tuszynski, K. Huang, Neuron 2012, 75, 824−837.
[12] a) T. Borsello, G. Forloni, Cur. Pharm. Des. 2007, 13, 1875−1886. ;b) J.
W. Chambers, A. Pachori, S. Howard, M. Ganno, D., Jr. Hansen, T.
Kamenecka, X. Song, D. Duckett, W. Chen, Y. Y. Ling, L. Cherry, M. D.
Cameron, L. Lin, C. H. Ruiz, P. Lograsso, ACS Chem. Neurosci. 2011,
[22] B. Wagner, D. Schumann, U. Linne, U. Koert, M. A. Marahiel, J. Am.
Chem. Soc. 2006, 128, 10513-10520.
[23] X. Xie, Y. Gu, T. Fox, J. T. Coll, M. A. Fleming, W. Markland, P. R.
Caron, K. P. Wilson, M. S. Su, Structure 1998, 6, 983-991.

Chemistry - An Asian Journal
10.1002/asia.201403417
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Mi-hyun Kim, Junghun Lee, Jung-Mi
Hah*
Page No. – Page No.
De Novo Design and Synthesis of a γ-
Turn Peptidomimetic Scaffold and Its
Application as JNK3 Allosteric Ligand
As
a way to develop a neuroprotective agent for the JNK3-JIP1-binding site,
peptidomimetics of JIP-1 as JNK3 allosteric regulators have been conducted. The study
consisted of in silico scaffold hopping, molecular docking, solution and solid-phase
peptide syntheses, and Kd measurements using surface plasmon resonance. As a
peptidomimetic of JIP1, heptamer mimetics 16 (Kd = 7.3 µM) presented higher affinity
than decamer JIP1 (Kd = 27.7 µM). The high affinity of 16 implies that the characteristic
γ-turn mimetic structure, “Φ–X–Φ” hydrophobic motif in 16, increased its affinity toward
the JIP-site of JNK3.