Restraining the flexibility of the central linker in terameprocol results in constrained analogs with improved growth inhibitory activity

Article abstract of DOI:10.1016/j.bmcl.2013.09.014

The semi-synthetic lignan terameprocol inhibits the transcription of several inflammatory and oncogenic genes and has been evaluated for its anti-cancer properties. Here we investigated the effect of restricting the flexibility of the carbon linker connecting the terminal rings of terameprocol on its growth inhibitory activity. Conformational restriction was explored by introducing unsaturation, inserting polar entities with limited flexibility and cyclization of the connecting linker. Twenty three compounds were synthesized and evaluated on a panel of malignant human cells. The most promising compounds were those with non-polar linkers, as seen in butadiene 1a and the cyclized benzylideneindane analog 7. Both compounds were more potent than terameprocol on pancreatic BxPC-3 cells with GI₅₀ values of 3.4 and 8.1 μM, respectively. Selected isomers of 1a (E,E) and 7 (Z) adopted low energy bent conformations that mimicked the low energy conformer of terameprocol. It is tempting to propose that conformational similarity to terameprocol may have contributed to their good activity. The scaffolds of 1a and 7 should be further investigated for their anticancer potential.

Full text of DOI:10.1016/j.bmcl.2013.09.014

Bioorganic & Medicinal Chemistry Letters 23 (2013) 6127–6133
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Restraining the flexibility of the central linker in terameprocol
results in constrained analogs with improved growth inhibitory
activity
⇑
Sherman Si Han Ho, Mei Lin Go
Department of Pharmacy, National University of Singapore, 18 Science Drive 4, Singapore 117543, Republic of Singapore
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 June 2013
Revised 12 August 2013
Accepted 5 September 2013
Available online 13 September 2013
The semi-synthetic lignan terameprocol inhibits the transcription of several inflammatory and oncogenic
genes and has been evaluated for its anti-cancer properties. Here we investigated the effect of restricting
the flexibility of the carbon linker connecting the terminal rings of terameprocol on its growth inhibitory
activity. Conformational restriction was explored by introducing unsaturation, inserting polar entities
with limited flexibility and cyclization of the connecting linker. Twenty three compounds were synthe-
sized and evaluated on a panel of malignant human cells. The most promising compounds were those
with non-polar linkers, as seen in butadiene 1a and the cyclized benzylideneindane analog 7. Both com-
pounds were more potent than terameprocol on pancreatic BxPC-3 cells with GI50 values of 3.4 and
Keywords:
Terameprocol
Constrained analogs
Growth inhibitory activity
Structure–activity relationship
8.1 lM, respectively. Selected isomers of 1a (E,E) and 7 (Z) adopted low energy bent conformations that
mimicked the low energy conformer of terameprocol. It is tempting to propose that conformational sim-
ilarity to terameprocol may have contributed to their good activity. The scaffolds of 1a and 7 should be
further investigated for their anticancer potential.
Ó 2013 Elsevier Ltd. All rights reserved.
Nordihydroguaiaretic acid (NDGA) is a naturally occurring lig-
nan found in the dried leaves and resin of the creosote bush (Larrea
divaricata). NDGA is associated with a wide spectrum of biological
activity1 and its structural analogs have likewise been found to
possess an impressive range of therapeutically relevant proper-
ties.2 Terameprocol (meso-tetra-O-methyl nordihydroguaiaretic
acid), also known as M4N or EM-1421, is arguably the best known
and most widely investigated NDGA analog.3–10 (Fig. 1) Terame-
procol inhibited the transcription of several genes that are involved
in carcinogenesis and inflammation. These include the Sp1-depen-
dent genes that transcribe the anti-apoptotic protein survivin and
the cell cycle protein cyclin dependent kinase cdc2.8,9 It also sup-
pressed the expression of inducible cyclooxygenase-2 (COX-2)5
deductions to be made.11,13 It is likely that the methyl groups on
the butane side chain linking the phenyl rings were not critical
for anti-cancer activity. Compounds I, II and III which lacked
methyl substituents on the connecting butane linker provided sup-
port for this notion (Fig. 1). Compound I (IC50 0.3
lM) was at least
10 times more potent than NDGA (IC50 4.4 M) on the small lung
l
cancer cell line H69.13 Compounds II and III were singled out for
detailed biological characterization in a patent on a large series
of tetra-O-substituted butane bridge-modified NDGA analogs,11
presumably due to their promising biological activities. Indeed,
III was cited as the most potent survivin inhibitor among the inves-
tigated compounds.11 It was also apparent from II and III that func-
tionalization of the bis-catechol moieties to give ethers was well
tolerated. Compound IV is a conformationally restricted analog of
NDGA in which the butane linker is locked in a tetrahydronaptha-
lene scaffold. This modification was deemed to have improved
selectivity: IV was reported to inhibit the breast cancer target
IGF-1R to a greater extent than 15-lipooxygenase, an off-target en-
zyme that was preferentially inhibited by NDGA.14
and inhibited nuclear factor kappa-B (NF-
jB)-dependent transcrip-
tion of several pro-inflammatory cytokines by preventing the bind-
ing of RelA, a transcription factor of the NF-
sites on DNA.6
jB family, to its cognate
There is limited information on the structure–activity relation-
ship (SAR) of terameprocol with regards to its anti-cancer activity.
Most investigations focussed on other NDGA analogs and activities
that were not related to carcinogenesis.11,12 Nonetheless, in those
cases where anticancer activity was monitored, the structural sim-
ilarity between terameprocol and NDGA allowed some broad
Conformational restriction is a widely employed optimization
strategy in medicinal chemistry. It minimizes the loss of conforma-
tional entropy that is incurred when a ligand binds to its target be-
cause the rotatable bonds in a conformationally restricted analog,
unlike its flexible counterpart, are pre-shaped for the binding inter-
action. Thus, there is a free energy advantage conferred by this
modification, which could translate to greater binding affinity or
⇑
Corresponding author. Tel./fax: +65 65162654.
E-mail address: phagoml@nus.edu.sg (M.L. Go).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.09.014

6128
S. S. H. Ho, M. L. Go / Bioorg. Med. Chem. Lett. 23 (2013) 6127–6133
O
O
OH
OH
O
O
HO
HO
Terameprocol
Nordihydroguaiaretic acid (NDGA)
O
O
N
OH
OH
N
O
O
HO
N
HO
I
Compound
II
Compound
N
IC50H69 0.3 µM
GI50 on a range of
cancer cell lines: 1-8 µM
O
HO
HO
N(C2H5)2
O
O
(C2H5)2N
O
N(C2H5)2
N(C2H5)2
Compound
IV
III
IC50 survivin: 25 µM
Compound
IC50 IGF1R: 1 µM
Figure 1. Structures of nordihydroguaiaretic acid (NDGA), terameprocol and compounds I–VI.
specificity. Besides IV, other conformationally restricted NDGA
analogs have been reported, but not always in the context of anti-
cancer activity.12 In these compounds, the butane linker was re-
strained in furan or tetrahydrofuran or lactone rings.12 Here, we
report the syntheses and antiproliferative activities of conforma-
tionally restricted terameprocol analogs. Broadly, conformational
restraint was imposed on the butane linker by (i) inclusion of con-
jugated double bonds (Series 1), (ii) insertion of a carbonylamino
or sulfonylamino or ureido functionality (Series 2–4) and (iii) cycli-
zation (Series 5, 6, compound 7 and ganschisandrine) (Fig. 2). The
compounds were designed to retain the bis-3,4-dimethoxy substi-
tuted phenyl rings present in terameprocol. Electron delocalization
in the polar carbonylamino, sulfonylamino and ureido moieties
would restrict flexibility of the linker besides serving to reduce
the lipophilic character of terameprocol (estimated ClogP 5.76).
Cyclization strategies were loosely based on IV, but included polar
entities (sulfonylamino, carbonylamino) within the cyclized scaf-
fold. A Scifinder Scholar search showed that all the compounds
have been reported but only 9 compounds have citations for bio-
logical activity. The relevant references are given in the Supple-
mentary data.
A total of 23 compounds were synthesized by conventional
methods. The diphenylbutadiene 1a was synthesized by a palla-
dium-catalyzed Heck reaction between 3,4-dimethoxy-b-bromo-
styrene (8) and 3,4-dimethoxystyrene as detailed in Scheme 1.15
The proton NMR spectra of 1a indicated the presence of a single
isomer, which was deduced to be E,E based on the symmetrical
appearance of the two singlets attributed to the ring methoxy pro-
tons. Coupling constants could not be used for structure assign-
ment of 1a because one pair of vinylic protons was embedded
with the aromatic protons while the two protons at the center of
the butadiene linker exhibited complex splitting patterns due to
long range coupling. The proton NMR spectra of 1a is shown in
Supplementary data. When 1a was analyzed by HPLC, only a single
peak was observed in the the chromatogram, which further
supported the presence of a single isomer. Compound 8 was syn-
thesized by a modified Hunsdiecker reaction from 3,4-dimethoxy-
cinnamic acid.16
The acrylamides 1b, 1c were obtained by amide coupling in the
presence
of
1-ethyl-3-(3-dimethylaminopropyl)carbodiimide
(EDC) and 1-hydroxybenzotriazole (HOBt) (Scheme 2). The same
reaction was used to prepare the amides 2a–2f (Scheme 2).
The symmetrical ureas 3a and 3c were synthesized by refluxing
urea with the corresponding amines,17 whereas 3b was afforded by
the condensation of an amine with an isocyanate (Scheme 3).
Compounds 4a–4c were obtained by reacting 3,4-dimethoxy-
benzenesulfonyl chloride with the respective amines (Scheme 4).
The constrained analogs of Series 5 were derived by incorporat-
ing the N of the carbonylamino (5a, 5b) or sulfonylamino (5c)
moiety in
a tetrahydroisoquinoline ring. In 5d, the sulfonyl
moiety of 5c was replaced by methylene. The Schotten–Baumann
reaction between an acyl chloride and 3,4-dimethoxy-tetra-
hydroisoquinoline was employed for the syntheses of 5a and
5b.19,20 The same tetrahydroisoquinoline was reacted with
3,4-dimethoxybenzenesulfonylchloride to give 5c, and 3,4-dime-
thoxybenzyl bromide (10) in the presence of diisopropylethyl-
amine (Huenig’s base)18 to give 5d (Scheme 5).
In Series 6, the linker was incorporated in an isoquinoline ring
as in Series 5, except that the dimethoxyphenyl ring was not at-
tached directly to the heterocyclic nitrogen but at the
a-carbon.
Subjecting 2c and 2f to phosphoryl chloride in a microwave as-
sisted Bischler–Naperialski reaction afforded dihydroisoquinolines
6a and 6b, respectively. Tetrahydroisoquinoline 6c was synthe-
sized by a microwave assisted Pictet–Spengler reaction with

S. S. H. Ho, M. L. Go / Bioorg. Med. Chem. Lett. 23 (2013) 6127–6133
6129
AND Enantiomer
O
O
O
O
MODIFICATION OF BUTANE LINKER OF TERAMEPROCOL
(A) UNSATURATION
(C) CYCLIZATION
O
O
X
O
O
O
O
O
O
N
n
O
O
O
O
N
O
O
X = SO2
X= CH2
n
Series 5
N
O
O
O
N
H
O
Series 1
O
O
O
O
O
NH
N
O
(B) POLAR SUBSTITUENTS
O
O
O
H
n
O
O
O
O
N
O
O
O
O
n
O
Series 6
O
Series 2
O
O
O
O
O
n
n
O
O
O
O
O
N
N
H
O
O
H
7
O
Ganschisandrine
Series 3
O
O
O
O
O
S
N
H
O
n
Series 4
Figure 2. Conformational restriction of the butane linker in terameprocol by (A) Introducing unsaturation (Series 1, n = 0, 1); (B) inserting electron delocalized polar entities
(Series 2–4, n = 0, 1, 2); (C) cyclization (Series 5, 6, compound 7, ganschisandrine).
O
O
O
O
O
O
Br
(a)
(b)
OH
O
O
O
1a
8
Scheme 1. Synthesis of 1a. Reagents and conditions: (a) N-bromosuccinimide, triethylamine, CH2Cl2, rt, 0.5 h; (b) 3,4-dimethoxystyrene, Pd(OAc)2, PPh3, triethylamine, DMF,
N2, 110 °C, 5 h.
trifluoroacetic acid (Scheme 6).19 The chiral forms of 6c were not
separated at this stage of the investigation.
(glioblastoma). Growth inhibitory activity was quantified as GI50
(concentration required to reduce viability to 50% of that observed
in untreated controls, with readings corrected for cell count at time
zero) using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoli-
um bromide (MTT) assay which is based on cellular redox enzyme
activity (Table 1).
Terameprocol curtailed the proliferation of only three of the five
cell lines investigated, namely pancreatic BxPC-3, renal RCC786-0
and breast T47D. Its GI50 on T47D and RCC786-0 were in the low
Compound 7 was prepared by the Wittig reaction between 5,6-
dimethoxy-1-indanone and the triphenylphosphonium salt of 3,4-
dimethoxybenzyl bromide (11) with sodium methoxide as base
(Scheme 7). Both E and Z isomers of 7 were obtained in a ratio that
was estimated from 1H NMR to be 7:1, in favour of the E isomer.
They were not separated at this stage of the investigation.
The compounds were evaluated for growth inhibitory activity
(72 h) on a panel of human malignant cells: BxPC-3 (pancreatic
adenocarcinoma), RCC786-0 (clear cell renal cell carcinoma),
T47D (breast cancer), HeLa (cervical cancer), T98G and U87
micromolar range (ꢀ5
BxPC-3 cells (GI50 17
occurring lignan structurally related to terameprocol as
l
M), but it was less active on the pancreatic
lM). Ganschisandrine (Fig. 2) is a naturally
a

6130
S. S. H. Ho, M. L. Go / Bioorg. Med. Chem. Lett. 23 (2013) 6127–6133
O
O
n
O
O
O
O
O
O
O
O
(a)
N
H
OH
+
H
N
2
n
1b: n=0
1c: n=1
O
O
O
O
O
O
O
N
n2
O
+
(a)
O
H
N
2
OH
n1
H
n1
n2
O
2a: n1=0; n2=0
2d n1=1; n2=0
2e n1=1; n2=1
2f n1=1; n2=2
2b: n1=0; n2=1
2c: n1=0; n2=2
Scheme 2. Synthesis of 1b, 1c, 2a–2f. Reagents and conditions: (a) EDC, HOBt, TEA, CH2Cl2, rt, 18 h.
O
O
O
O
O
(a)
H
N
H
N
H2N
O
n
n
n
O
3a: n=0
3c
: n=1
O
O
O
O
H
H
N
O
O
NH2
(b)
O
O
N
C
N
+
O
O
3b
Scheme 3. Synthesis of 3a–c. Reagents and conditions: (a) urea, 170 °C, 3.5 h; (b) THF, rt, 1 h.
O
S
O
O
O
n
O
O
O
O
O
O
S
O
O
(a)
Cl
N
H2N
H
+
n
4a: n=0
4b
: n=1
: n=2
4c
Scheme 4. Synthesis of 4a–4c. Reagents and conditions: (a) TEA, CH2Cl2, rt, 0.5 h.
O
O
n
n
n
O
O
OH
O
O
Cl
O
O
N
(a)
(b)
O
O
O
5a:
n=0
9a: n=0
5b: n=1
9b:
n=1
O
S
O
O
O
N
O
O
O
O
S
O
Cl
(c)
HN
O
+
O
O
5c
O
O
O
O
(e)
O
O
O
O
(d)
OH
Br
N
10
5d
Scheme 5. Synthesis of 5a–5d. Reagents and conditions: (a) SOCl2, CH2Cl2, N2, rt, 18 h; (b) 3,4-dimethoxy-1,2,3,4-tetrahydroisoquinoline HCl, NaOH, CH2Cl2, H2O, rt, 3 h; (c)
TEA, CH2Cl2, rt, 0.5 h; (d) PBr3, CH2Cl2, rt, 24 h; (e) 3,4-dimethoxy-1,2,3,4-tetrahydroisoquinoline HCl, DIPEA, CH3CN, N2, rt, 4 h.

S. S. H. Ho, M. L. Go / Bioorg. Med. Chem. Lett. 23 (2013) 6127–6133
6131
H
O
(a)
O
O
O
O
N
O
O
O
O
O
(a)
N
O
H
N
N
O
O
O
O
2f
6a
2c
6b
O
O
O
O
O
O
O
(b)
O
O
NH2
NH
HO
O
O
6c
O
O
Scheme 6. Synthesis of 6a–6c. Reagents and conditions: (a) POCl3, toluene, mw 140 °C, 0.5 h; (b) TFA, toluene, mw 140 °C, 0.5 h.
O
O
O
O
O
O
(a)
Br
O
O
(b)
OH
(c)
PPh3Br
O
O
10
7
11
Scheme 7. Synthesis of 5a–5d. Reagents and conditions: (a) PBr3, CHCl2, rt, 24 h; (b) TPP, toluene, 110 °C, 18 h; (c) 5,6-dimethoxy-1-indanone, NaOMe, THF, N2, 50 °C, 18 h.
Table 1
Growth inhibitory GI50 values of test compounds, terameprocol and ganschisandrine on malignant pancreatic BxPC-3, renal RCC786-0, breast T47D, cervical HeLa and
glioblastoma T98G, U87 cells
a
Compound
Antiproliferative activity GI50
(lM)
BxPC-3
RCC786-0
T47D
HeLa
>100
T98G
U87
1a
1b
3.41 0.42
>50
2.67 0.48
>50
>100
>50
>100
>50
>100
>50
>50
1c
2a
2b
2c
2d
2e
2g
3a
>75
>100
>100
>100
>100
>100
>100
>50
68.4 5.9
>100
>100
>100
>100
>100
>100
>50
>100
>100
>100
>100
>100
>100
>100
>50
>75
70.3 5.5
>100
>100
>100
>100
>100
>100
>50
57.1 1.0
>100
>100
>100
>100
>100
>100
>50
>100
>100
>100
>100
>100
>100
>50
3b
3c
4a
4b
4c
5a
5b
5c
>100
>100
>100
>100
>100
13. 6 2.8
>100
>100
>100
>100
>100
>100
24.1 4.1
>100
>100
>100
>100
>100
>100
11.9 2.9
>100
>100
>100
>100
>100
>100
8.23 0.32
>100
>50
>100
>100
>100
>100
>100
10.1 1.2
>100
>50
>100
>100
>100
>100
>100
20.6 3.3
>100
>50
>50
>50
>50
5d
6a
6b
>100
>100
>100
>100
>100
>100
5.50 1.22
4.54 0.98
>75
45.4 6.4
>100
16.1 1.5
>100
4.24 0.63
5.36 1.04
4.40 0.46
55.9 9.0
79.3 16.3
16.9 3.9
57.6 11.1
>25
>100
>100
>100
>100
>25
>100
>100
>100
>100
>25
26.3 5.5
>100
8.10 2.19
17.1 4.0
17.2 1.8
6c
7b
Terameprocol
Ganschisandrine
>50
>75
>50
58.3 1.1
>50
>75
a
Mean standard deviation of n P 3 determinations. GI50 was determined from the expression: T À To/C À To = 50% where T = absorbance at 72 h of cells exposed to test
compound, To = absorbance at 0 h, C = absorbance at 72 h of untreated cells. Absorbance readings were obtained from MTT assay.
b
Compound 7 had low solubility and antiproliferative activity could not be determined beyond 25 lM.
constrained analog. It retains several features of terameprocol,
namely the terminal dimethoxyphenyl rings, the four carbon sep-
aration between the phenyl rings and methyl substitution of the
ring/linker.It was equipotent to terameprocol on T47D and BxPC-
linker in a tetrahydrofuran ring did not abolish growth inhibitory
activity. We also found that the most stable conformer of terame-
procol, identified from a conformational search on MOE (Molecular
Operating Environment, 2011.10), was folded with the terminal
phenyl rings oriented orthogonally to each other (Fig. 3A). Thus,
cyclization of the linker may provide a means of mimicking the
3 but significantly less potent on RCC786-0 (GI50 >75
lM). The lat-
ter notwithstanding, it may be inferred that restraining the butane

6132
S. S. H. Ho, M. L. Go / Bioorg. Med. Chem. Lett. 23 (2013) 6127–6133
Figure 3. (A) Stable conformer of terameprocol identified from conformational search on MOE (Molecular Operating Environment, Version 2011.10). Method: LowModeMD,
RMS gradient 0.005, number of iterations 10,000. (B) Flexible alignment of terameprocol (red) and 7Z (yellow) by MOE (flexible alignment mode, 200 iterations, failure rate
20, energy cutoff 15 kcal/mol). Dotted lines depict Van der Waals interactions between the methyl and phenyl rings. (C) Overlapping stable conformers of 1a (E,E, ciscoid)
(blue), 7Z (yellow) and terameprocol (red) as determined by flexible alignment MOE. A conformational search showed that the energy of 1a (E,E ciscoid) was 98.37 kcal/mol
compared to 95.79 kcal/mol for the stable conformer of 1a (E,E transoid).
energetically more stable, and possibly, biologically relevant con-
former of terameprocol.
Aside from shape considerations, the lipophilicity of the synthe-
sized compounds was also considered to determine its contribu-
tion to activity. Based on their ClogP values (Supplementary
data), there was an apparent trend between lipophilicity and
growth inhibitory activity. Among the synthesized compounds,
the potent analogs 1a and 7 were the most lipophilic (ClogP 4.35
and 4.25, respectively) while the inactive Series 2–4 compounds
were at least a hundred fold less lipophilic (ClogP values <2.7).
As conformational restraint in the latter compounds was achieved
by incorporating polar carbonylamino, sulfonylamino and ureido
functionalities, the concurrent fall in lipophilicity may have had
an overriding adverse effect on activity. Nonetheless, it was grati-
fying to note that analogs (1a, 7) with greater potency and lesser
lipophilic character than terameprocol (ClogP 5.76) could be ob-
tained by the present approaches.
Of the synthesized compounds, only four compounds (1a, 5a, 6b
and 7) had one or more GI50 values that were lower than those of
terameprocol. Compounds in Series 2, 3 and 4 were devoid of
growth inhibitory properties. This may indicate that limiting the
rotational flexibility of the carbonylamino, sulfonylamino and ure-
ido groups in 2, 3 and 4, respectively, gave rise to unfavourable
conformations, or that the polarity introduced by these residues
had an adverse effect on activity.
Cyclization in 5a and 6b was achieved by incorporating a tetra-
hydro or dihydroisoquinoline ring in the linker. Compound 5a was
more potent than 6b on BxPC-3, T47D and Hela cells but it fared
less well when compared to 1a and 7. The most outstanding fea-
ture of 5a was its widespread growth inhibitory activity—it was ac-
tive on all 5 cell lines with GI50 values ranging from 8 to 24
l
M. A
Taken together, we have shown here that restricting the confor-
mational flexibility of the connecting linker in terameprocol is a
viable means of improving growth inhibitory activity against a pa-
nel of malignant cells. Restraining flexibility by incorporating
unsaturation (1a) and cyclization coupled with unsaturation (7)
yielded promising results. Selected isomers of 1a and 7 mimicked
the bent shape found in the lowest energy conformer of terame-
procol. Their scaffolds warrant further investigation to determine
if their biological targets coincided with that of terameprocol and
if additional functionalization could lead to improved in vitro
and in vivo growth inhibitory activities.
comparison with other compounds in Series 5 highlighted the
importance of the amide carbonyl in linking the two rings of 5a.
Notably, a one-carbon extension of the linker (5b), replacing car-
bonyl with the isosteric sulfonyl (5c) or methylene (5d) signifi-
cantly reduced growth inhibitory activity.
Among the actives with a cyclized feature, 7 had outstanding
growth inhibitory activity. It was more potent than terameprocol
on pancreatic BxPC-3 cells (GI50 8.1 vs 17.1
and equipotent on renal RCC786-0 and breast T47D cells (GI50 4–
M). Interestingly, the most stable conformer of terameprocol
lM for terameprocol)
5
l
and the Z isomer of 7 were remarkably well aligned when these
molecules were tested for overlap using the Flexible Alignment
module in MOE (Fig. 3A and B). As mentioned earlier, 7 was ob-
tained as a mixture of E and Z isomers, with E the dominant isomer.
Equilibria between E and Z isomers of 3-substituted indolin-2-ones
have been reported20 and it is conceivable that a similar E/Z isom-
erisation may have occurred in the structurally related 7. If this re-
sults in the enrichment of 7Z in the biological milieu, the good
activity of 7 may derive in part from its ability to mimic the bent
conformation of terameprocol.
Acknowledgments
Sherman S.H. Ho gratefully acknowledges National University
of Singapore for the President’s Graduate Fellowship. The work is
supported by a research grant from the National University of Sin-
gapore (RP 148 000 171 112) to M.L. Go.
Supplementary data
Supplementary data (details on the growth inhibitory assay,
spectral characteristics and ClogP values of synthesized compounds
are provided) associated with this article can be found, in the online
version, at http://dx.doi.org/10.1016/j.bmcl.2013.09.014.
The most active analog identified in this investigation was the
butadiene 1a which had outstanding growth inhibitory activities
on BxPC-3 and RCC786-0 cells (GI50 ꢀ 3
lM). Notwithstanding its
good activity profile, 1a is a structural anomaly as it is unique among
the other actives in lacking a cyclized feature. Conformational search
identified the linear transoid E,E isomer of 1a as the most stable con-
former (95.79 kcal/mol) but 1a had an energetically viable bent cis-
coid E,E form (98.37 kcal/mol). This conformer could be reasonably
aligned to terameprocol and 7Z, with the three structures retaining
the bent shape of the stable terameprocol conformer (Fig. 3C).
References and notes
1. Lambert, J. D.; Dorr, R. T.; Timmermann, B. N. Pharm. Biol. 2004, 42, 149.
2. Saleem, M.; Kim, H. J.; Ali, M. S.; Lee, Y. S. Nat. Prod. Rep. 2005, 22, 696.
3. Hwu, J. R.; Tseng, W. N.; Gnabre, J.; Giza, P.; Huang, R. C. C. J. Med. Chem. 1998,
41, 2994.

S. S. H. Ho, M. L. Go / Bioorg. Med. Chem. Lett. 23 (2013) 6127–6133
6133
4. Chen, H.; Teng, L.; Li, J. N.; Park, R.; Mold, D. E.; Gnabre, J.; Hwu, J. R.; Tseng, W.
N.; Huang, R. C. C. J. Med. Chem. 1998, 41, 3001.
5. Eads, D.; Hansen, R. L.; Oyegunwa, A. O.; Cecil, C. E.; Culver, C. A.; Scholle, F.;
Petty, I. T. D.; Laster, S. M. J. Inflammatory (Lond) 2009, 6.
6. Oyegunwa, A. O.; Sikes, M. L.; Wilson, J. R.; Scholle, F.; Laster, S. M. J.
Inflammatory (Lond) 2010, 7.
13. McDonald, R. W.; Bunkobpon, W.; Liu, T.; Fessler, S.; Pardo, O. E.; Freer, I. K. A.;
Glaser, M.; Seckl, M. J.; Robins, D. J. Anticancer Drug Des. 2001, 16, 261.
14. Blecha, J. E.; Anderson, M. O.; Chow, J. M.; Guevarra, C. C.; Pender, C.;
Penaranda, C.; Zavodovskaya, M.; Youngren, J. F.; Berkman, C. E. Bioorg. Med.
Chem. Lett. 2007, 17, 4026.
15. Dieck, H. A.; Heck, R. F. J. Am. Chem. Soc. 1974, 96, 1133.
16. Das, J. P.; Roy, S. J. Org. Chem. 2008, 67, 7861.
17. Baeyer, A. Justus Liebigs Ann. Chem. 1864, 131, 251.
7. Heller, J. D.; Kuo, J.; Wu, T. C.; Kast, W. M.; Huang, R. C. C. Cancer Res. 2001, 61,
5499.
8. Chang, C. C.; Heller, J. D.; Kuo, J.; Huang, R. C. C. Proc. Natl. Acad. Sci. USA 2004,
101, 13239.
9. Li, F.; Altieri, D. C. Cancer Res. 1999, 59, 3143.
10. Park, R.; Chang, C. C.; Liang, Y. C. Clin. Cancer Res. 2005, 11, 4601.
11. Huang, R. C.; Baltimore, M. D.; Kotohiko, K. US20110014192, 2011.
12. Chen, Q. Curr. Top. Med. Chem. 2009, 9, 1636.
18. Moore, J. L.; Taylor, S. M.; Soloshonok, V. A. ARKIVOC 2005, 6, 287.
19. Awuah, E.; Capretta, A. J. Org. Chem. 2010, 75, 5627.
20. Sun, L.; Tran, N.; Tang, F.; App, H.; Hirth, P.; McMahon, G.; Tang, C. J. Med. Chem.
1998, 41, 2588.