Antitumor agents. 293. Nontoxic dimethyl-4,4′-dimethoxy-5,6,5′, 6′-dimethylenedioxybiphenyl-2,2′-dicarboxylate (DDB) analogues chemosensitize multidrug-resistant cancer cells to clinical anticancer drugs

Article abstract of DOI:10.1021/jm300378k

Novel dimethyl-4,4′-dimethoxy-5,6,5′,6′- dimethylenedioxybiphenyl-2,2′-dicarboxylate (DDB) analogues were designed and synthesized to improve their chemosensitizing action on KBvin (vincristine-resistant nasopharyngeal carcinoma) cells, a multidrug resistant cell line overexpressing P-glycoprotein (P-gp). Structure-activity relationship analysis showed that aromatic and bulky aliphatic side chains at the 2,2′-positions effectively and significantly sensitized P-gp overexpressing multidrug resistant (MDR) cells to anticancer drugs, such as paclitaxel (TAX), vincristine (VCR), and doxorubicin (DOX). DDB derivatives 16 and 23 showed 5-10 times more effective reversal ability than verapamil (VRP) for TAX and VCR. Analogue 6 also exhibited five times greater chemosensitizing effect against DOX than VRP. Importantly, no cytotoxicity was observed by the active DDB analogues against both non-MDR and MDR cells, suggesting that DDB analogues serve as novel lead compounds for the development of chemosensitizers to overcome the MDR phenotype. The mechanism of action studies demonstrated that effective inhibition of P-glycoprotein by DDB analogues dramatically elevated the cellular concentration of anticancer drugs.

Full text of DOI:10.1021/jm300378k

Article
pubs.acs.org/jmc
Antitumor Agents. 293. Nontoxic Dimethyl-4,4′-dimethoxy-5,6,5′,6′-
dimethylenedioxybiphenyl-2,2′-dicarboxylate (DDB) Analogues
Chemosensitize Multidrug-Resistant Cancer Cells to Clinical
Anticancer Drugs
Hsin-Yi Hung,† Emika Ohkoshi,† Masuo Goto,‡ Kenneth F. Bastow,§ Kyoko Nakagawa-Goto,*,†
and Kuo-Hsiung Lee*,†,⊥
†Natural Products Research Laboratories, UNC Eshelman School of Pharmacy, University of North Carolina, Chapel Hill, North
Carolina 27599-7568, United States
‡Cell and Developmental Biology, School of Medicine, University of North Carolina, Chapel Hill, North Carolina 27599-7090,
United States
§Division of Chemical Biology & Medicinal Chemistry, UNC Eshelman School of Pharmacy, University of North Carolina, Chapel
Hill, North Carolina 27599-7568, United States
⊥Chinese Medicine Research and Development Center, China Medical University and Hospital, Taichung, Taiwan
S
* Supporting Information
ABSTRACT: Novel dimethyl-4,4′-dimethoxy-5,6,5′,6′-dime-
thylenedioxybiphenyl-2,2′-dicarboxylate (DDB) analogues
were designed and synthesized to improve their chemo-
sensitizing action on KBvin (vincristine-resistant nasophar-
yngeal carcinoma) cells, a multidrug resistant cell line
overexpressing P-glycoprotein (P-gp). Structure−activity
relationship analysis showed that aromatic and bulky aliphatic
side chains at the 2,2′-positions effectively and significantly
sensitized P-gp overexpressing multidrug resistant (MDR)
cells to anticancer drugs, such as paclitaxel (TAX), vincristine (VCR), and doxorubicin (DOX). DDB derivatives 16 and 23
showed 5−10 times more effective reversal ability than verapamil (VRP) for TAX and VCR. Analogue 6 also exhibited five times
greater chemosensitizing effect against DOX than VRP. Importantly, no cytotoxicity was observed by the active DDB analogues
against both non-MDR and MDR cells, suggesting that DDB analogues serve as novel lead compounds for the development of
chemosensitizers to overcome the MDR phenotype. The mechanism of action studies demonstrated that effective inhibition of P-
glycoprotein by DDB analogues dramatically elevated the cellular concentration of anticancer drugs.
gp inhibitors.8 Verapamil (VRP) and cyclosporine A (CsA),
two first-generation chemosensitizers, were precluded from
clinical use because of significant toxicity but are used in
INTRODUCTION
Despite substantial biomedical research on cancer therapy,
cancers still remain the leading cause of death. Among all
■
experiments as positive controls. Second- and third-generation
factors resulting in the ultimate failure of cancer treatment,
chemosensitizers were developed subsequently; however,
chemotherapy resistance is a significant player, and multidrug
resistance (MDR), cross-resistance to different chemical drug
unsatisfactory toxicity and pharmacokinetic complications still
impeded drug candidate development. Although several third-
classes, occurs in various cancer types. Cellular mechanisms of
generation P-gp inhibitors, including tariquidar, are now in
MDR include decreased uptake of chemotherapeutic agents, via
expression of vacuolar ATPase (V-ATPase), or adaptation of
cancer cells to the cytotoxic ability of chemotherapeutic agents,
phase II cancer clinical trials,9 their clinical efficacies are not yet
clear. Thus, the discovery of safe and effective MDR
modulators is still attractive and greatly needed to overcome
via down-regulation of topoisomerase II and overexpression of
glutathione S-transferase-π.1−3 An emerging understanding of
the MDR issue in the field of cancer chemotherapy.
cancer resistance results from cancer stem cell-like features.4
Schisandrin B (Figure 1), the most abundant dibenzocy-
clooctadiene lignan from Schisandra chinensis, was found to
However, overexpression of drug efflux transporters, such as P-
inhibit P-gp/MDR1 and MRP1/ABCC1.10,11 Structurally
glycoprotein (P-gp) and MDR-associated protein (MRP), is the
primary cause leading to multidrug resistance.5 In order to
similar lignans, schisandrin A, schisandrol A, schisantherins A
surmount MDR, great efforts have been put into developing
clinically usable chemosensitizing agents, categorized as either
Received: March 19, 2012
Published: May 21, 2012
apoptosis modulators6,7 or MDR modulators, also known as P-
© 2012 American Chemical Society
5413
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
DOX-induced apoptosis of innate MDR cell line Bel7402 and
acquired MDR MCF-7/Adr.14 Since DDB acts against
multitargets, which is a common feature of natural products
or natural product-derived agents, ligand-based modification of
DDB was applied. Bicyclol bears a primary alcohol at the C-2′
position.17 Therefore, we selected the 2,2′- bishydroxymethyl
biphenyl 2 as a base scaffold to design various esters 3−28
(Scheme 1) in order to define substituent effects at the C2 and
C2′ positions. The ester groups (R in Scheme 1) were selected
by considering size, hydrophobicity, and electron density. The
diverse set included aliphatic acyclic (group I), cyclic (group
II), unsaturated aliphatic (group III), polar (group IV), and
aromatic (group V) groups, which could be transformed into
water-soluble salts, if necessary. Esterifications of 2 were
performed by treatment with the related acid (RCO2H), N-(3-
dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride
(EDCI), and 4-dimethylaminopyridine (DMAP) to produce
3, 7−12, 15−17, and 22, as well as by treatment with the
related acid chloride (RCOCl) and Et3N to give 4−6, 13−14,
19−21, and 23−28 (Scheme 1). Halogenated DDB analogues
with various functional groups were also designed to explore
their effects (Scheme 2) since 3,3′-dibromo-DDB has shown
good activity in anti-HIV research.19 3,3′-Dibromo-DDB (29)
was previously synthesized.19 Iodination of 1 was achieved by
treatment with silver trifluoacetate and iodine to provide 30.20
The diester groups of 29 were reduced with diisobutylalumi-
nium hydride (DIBAL) to afford bis-hydroxymethyl biphenyl
31 in good yield. The esters 32 and 33 were synthesized by
esterification of 31 with butyric and benzoic acids, respectively,
in the presence of EDCI and DMAP. Hydrolysis of dimethyl
ester 29 under basic conditions resulted in dicarboxylic acid 34.
Reflux of 34 in Ac2O gave anhydrous analogue 35, followed by
asymmetric cleavage of the carbonyl anhydrous bridge with
NaBH4, resulting in di-Br bicyclol analogue 36.21 All
synthesized analogues were prepared as the racemic mixtures.
Figure 1. Structures of DDB, bicyclol, and dibenzocyclooctadiene
lignans.
and B, also chemosensitized various anticancer drugs, including
vincristine (VCR), daunorubicin, and etoposide, in human
promyelocytic leukemia cell lines with overexpressed MRP1/
ABCC1.12 Dimethyl-4,4′-dimethoxy-5,6,5′,6′-dimethylenedioxy-
biphenyl-2,2′-dicarboxylate (DDB, 1, Figure 1), which was
discovered as a synthetic intermediate derivative of schisandrin
C,13 shares the biphenyl partial structure of dibenzocycloocta-
diene lignans. DDB (1) exhibited multidrug-resistant reversal
ability in MDR breast carcinoma MCF-7/Adr cells, KBv200,
and intrinsic MDR human hepatocarcinoma Bel7402 cells via
inhibition of P-gp and enhancement of apoptosis.14 However, a
very high concentration (50 μM) was required for effective
reversal action. Cotreatment of 1 with VCR using nude mice
with KBv200 xenografts also enhanced antitumor activity at
doses of 300 and 500 mg/kg.14 DDB (1) has been used to treat
chronic viral hepatitis B patients in China for more than 20
years, as well as in Korea and Egypt for more than 10 years,
without any significant adverse effects.15,16 This important fact
indicates that DDB analogues could be highly attractive MDR
reversal agents with significant clinical potential due to their
proven low toxicity. In addition, pharmacokinetic issues where
chemosensitizers would interfere with the clearance of
anticancer drugs often impede further development of an
effective chemosensitizer. DDB was found not to alter the
clearance of DOX by the evidence that plasma AUC0−24 of
DOX alone and DOX plus DDB were similar in ICR mice
bearing S180 sarcoma model.14 In 2006, Zhu et al. reported
that an asymmetric analogue of DDB, bicyclol (Figure 1), also
exhibited a chemosensitizing effect in two established MDR
cancer cell lines, VinRKB and AdrRMCF-7.17,18 Although DDB
and bicyclol have high potentials as MDR reversal agents and
various DDB analogues have been prepared, their MDR
reversal abilities and structure−activity relationship (SAR)
correlations have not been investigated. To explore more
potent nontoxic MDR reversal analogues with lower effective
dosing and to study SAR, we designed and synthesized
additional DDB analogues. Herein, we report the chemo-
sensitizing effects of newly synthesized DDB analogues.
RESULTS AND DISCUSSION
■
Evaluation of Cytotoxicity and Preliminary MDR
Reversal Activity Screening. All synthesized compounds
were evaluated in a cytotoxic activity assay using four tumor cell
lines, A549 (lung cancer), DU145 (prostate cancer), KB
(epidermoid carcinoma of the nasopharynx), and a resistant
subline, KBvin (overexpression of P-gp selected using
increasing concentrations of vincristine). Verapamil, the first
generation chemosensitizer precluded from clinical use due to
toxicity, showed some cytotoxicity especially to KB (IC50 ∼50
μM) and KBvin cells (IC50 ∼ 80 μM). Although compounds 28
and 32 were slightly cytotoxic, most of the DDB-derived
compounds did not exhibit significant cytotoxicity (IC50 > 100
μM), which implied low toxicity of these analogues
(Supporting Information Table S1).
For evaluating chemosensitizing activity, KB and KBvin cells
were cotreated with test compounds at 10 μM and the
anticancer drug paclitaxel (TAX) (Figure 2). As shown in
Figure 2B, DDB (1) and diol 2 did not exhibit MDR reversal
effect at a concentration of 10 μM in KBvin cells. 3,3′-
Dihalogenated DDB analogues with different 2,2′-functional
groups did not show potent activity, except for pentanoate 32
and benzoate 33, but 32 was slightly cytotoxic to KBvin cells,
which may result from bromo substitution. While a 2,2′-methyl
ester group [CH2OC(O)R] appeared to be critical for reversal
ability activity, 3,3′-halogenation reduced the potency (compare
33 with 19).
DESIGN AND SYNTHESES
On the basis of previous literature, DDB had effects on
inhibition of P-gp and activation of cellular caspase-3 leading to
■
5414
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
a
Scheme 1. Syntheses of Diester Derivatives 3−28
a
Reagents and conditions: (a) RCO2H, EDCI, DMAP, for 3, 7−12, 15−17, and 22, (b) RCOCl, Et3N for 4−6, 13, 14, 19−21, and 23−38, (c) acid
anhydride, DMAP for 18.
Many of the analogues with aliphatic ester substituents
(groups I and II in Figure 2) did show potent activity; however,
acetate 3 (R = methyl) and dodecanoate 12 (R = undecanyl)
were inactive, and 2-methylhexanoate 11 (R = 2-methybutyl)
was only moderately active. The screening results suggested a
rough SAR; at least a two-carbon linear ester chain (R = ethyl)
was necessary for chemosensitizing activity, but a long chain (R
> approximately five carbons) led to reduced or no activity.
5415
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
a
the tested analogues, including VRP, showed greater reversal
against VCR resistance compared with TAX resistance. The
reversal effects of compounds 6, 10, 13−16, 22, and 23 against
VCR were significantly better than that of VRP. Especially, 16
and 23 showed 560-fold reversal effect, which was 5.1 times
greater than that of VRP. The following SAR correlations were
proposed based on the chemosensitizing effects against TAX
and VCR. In compounds with aliphatic esters, unsaturated
group III compounds 15 and 16 were generally more potent
than saturated group I and II compounds. Compounds 15 and
16 displayed greater chemoreversal ability than 7 and 9, which
contain structurally related saturated groups. In the case of
compounds with aromatic esters (group V), an electron
donating group, such as methyl and methoxy, at the para-
position reduced the reversal ability, while additional methoxy
groups at the meta-position enhanced the ability. The following
rank order of potency was seen: 3,4,5-tri-OMe (23) > 3,4-di-
OMe (22) > H (19) > 4-Me (20) ≈ 4-OMe (21). Different
SAR correlations between TAX and VCR were found in the
saturated aliphatic group. Cyclic aliphatic side chains (group II,
13 and 14) were more effective than noncyclic aliphatic side
chains (group I, 4−11) in the case of VCR resistance, while this
difference was not present for TAX resistance. Within the
tested aliphatic acyclic group I compounds (4−11), 2-
methylbutyrate (R = isobutyl) 9 was most potent for TAX,
while pivalate (R = tert-butyl) 6 and petanoate (R = n-butyl) 10
were most potent for VCR. In addition, 10 exhibited greater
activity than the related unsaturated analogue 17 against VCR
resistance. All of the tested compounds were less effective at
reversing DOX resistance than against TAX and VCR. This
phenomenon might be correlated with the difference of efflux
pump on the cell membrane with that on the nuclear
membrane, because both TAX and VCR act by binding
tubulin, while DOX interacts with DNA by intercalation.22
However, compounds 6, 8, 11−14, and 20−23 still showed
greater reversal effects than VRP for DOX chemosensitivity in
MDR cells. Especially, pivalate 6, with a bulky and short
aliphatic ester chain, reversed the activity most effectively,
exhibiting moderately more sensitivity than VRP. SAR against
DOX was slightly different from that against VCR. Compounds
6, 8, 11, 13, and 14 with a branched substituent at the ester α-
position showed significant MDR reversal activity. Benzoates
with an electron-donating group, such as p-methyl (20) and p-
methoxy (21), had little effect. Although an additional methoxy
group at the meta-position increased the ability, no difference
was found between trimethoxy 23 and dimethoxy 22. The
effects of substituent on the phenyl ring resulted in the
following order of potency: 3,4-di-OMe (22) ≈ 3,4,5-tri-OMe
(23) > H (19) ≈ 4-Me (20) ≈ 4-OMe (21). In conclusion,
analogues 16 and 23 showed significant TAX and VCR
cytotoxic reversal ability, and analogue 6 exhibited the most
potent DOX cytotoxic reversal ability.
Scheme 2. Syntheses of Dihalogenated DDB Analogues
a
Reagent and conditions: (a) Br2, CHCl3, 0 °C, (b) CF3CO2Ag, I2, (c)
DIBAL, CH2Cl2, (d) EDCI, DMAP, CH2Cl2, CH3(CH2)2CO2H for
32, and PhCO2H for 33, (e) KOH, EtOH, reflux, (f) Ac2O, reflux, (g)
NaBH4, THF, MeOH.
Isobutyrate (R = isopropyl) 5 and pivalate (R = tert-butyl) 6
displayed the most significant survival rate (13−14%) among all
saturated group I and II compounds, and the unsaturated 3-
methylbut-2-enoate (prenyl-like substituent) 16 in the group
III compounds further increased the effect, showing a 5%
survival rate. Among the group IV compounds with polar ester
substituents, succinate 18 (R = CH2CH2COOH) was inactive,
while compounds with phenyl or pyridinyl aromatic rings
showed potent reversal ability, unless an electron-withdrawing
group, such as nitro and cyano, was present on the aromatic
ring. Especially, benzoate 19 and p-methoxybenzoate 21
exhibited less than 1% survival rate. From these findings, we
speculated that decrease in electron density on the aromatic
ring affected chemoreversal activity. Quinoline derivative 28
exhibited cytotoxicity against both KB and KBvin cells.
From these results, the group 1 compounds (4−11), except
for 3 and 12, the group II and III compounds (13−17), and the
most active group V compounds (19−23) were selected for
further investigation. The remaining compounds (1−3, 12, 18,
24−27, 29−32, 35, 36), which generally had weaker reversal
ability, as well as compounds with inherent cytotoxicity (28)
were eliminated from further testing.
Chemoreversal Ability of DDB Analogues with TAX,
VCR, and Doxorubicin (DOX). A quantitative evaluation of
the reversal ability of DDB analogues 4−11, 13−17, and 19−
23 was performed using MDR KBvin cells with various
concentrations of TAX, VCR, and DOX, which are clinically
used and known as significant P-gp substrates, partly
accounting for their resistance (Table 1). The IC50 value of
anticancer drugs in the presence of test compounds at 10 μM
concentration was calculated, and fold reversal was determined
by dividing the IC50 of anticancer drug alone by the IC50 of
anticancer drugs plus DDB analogue. For chemoreversal ability
against TAX resistance, compounds 9, 15, 16, 19, and 23 were
3−10 times more potent than the positive control VRP.
Especially, 16, with a prenyl-like ester substituent, and
trimethoxybenzoate 23 effectively reversed the sensitivity of
TAX in KBvin cells by 326- and 222-fold, respectively. Most of
On the basis of the above results, compounds 6 and 23 were
selected for further evaluation of chemosensitizing efficacy. The
dose−response proliferation inhibitory effects of TAX, VCR,
and DOX at 10 μM were analyzed against KB and KBvin cells
(Figure 3). In the absence of DDB compound or VRP, KBvin
cells were resistant to all three anticancer drugs, resulting in
IC50 values over 1000 nM. When 10 μM of compound 6 or 23
or VRP was added, the sensitivity of KBvin cells to each
anticancer drug was dramatically increased to at least the same
levels of KB cells. The chemosensitizing efficacy of 6 or 23 was
either similar or better than that of VRP. KBvin became even
5416
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
a
b
Figure 2. Screening of reversal abilities against KB (A) and KBvin (B). Note: Concentration of compounds: 10 μM; survival rate (%) was
measured by SRB method using KB and KBvin cells in the presence (+) or absence (−) of paclitaxel (TAX). Compounds with survival rates below
c
20% were considered very potent and moved to further experiments. Group (I) saturated acyclic alkyl; (II) cyclic alkyl ; (III) unsaturated; (IV)
polar; (V) aromatic; (VI) halogenated.
more sensitive to VCR than KB, when KBvin cells were
cotreated with 10 μM 23 and VCR. These results demonstrated
that 10 μM of 6 and 23 effectively chemosensitized MDR cells.
Dose−Response Effect of Compounds 6 and 23 on
Sensitization of KBvin to TAX. To evaluate the reversal
activity of 6 and 23 in a dose−response manner, KBvin cells
were cultured with nontoxic concentration of TAX (100 nM)
in the presence of various concentrations of compounds
(Figure 4). As we expected, compounds 6 and 23 exhibited
reversal activity in a dose-dependent manner. Although the
median effective concentration (EC50) value of 6 (2.81 μM)
was similar to that of VRP (2.71 μM), compound 23 (1.87 μM)
exhibited a lower EC50 than VRP (P values are given in
Supporting Information Table S3). These results demonstrate
that 23 could be more effective than VRP in chemosensitizing
the MDR cells to TAX.
of anticancer drugs in MDR cells, the effect of compounds 6
and 23 on P-gp function in KBvin cells using calcein-AM as a
fluorogenic P-gp substrate was investigated (Figure 5,
Supporting Information Table S4). Dose-dependent intra-
cellular accumulation of calcein was observed in the presence of
these compounds. Although 6 was slightly less potent than
VRP, 23 was up to 2-fold more potent than VRP, especially at
concentrations around the EC50 value (1.87 μM) of 23.
Therefore, these results clearly indicated that DDB analogues,
especially 23, are effective P-gp inhibitors.
To demonstrate the effective efflux inhibition of anticancer
drugs, direct measurement of cellular accumulation of DOX in
KBvin cells was studied as the intensity of intrinsic fluorescence
of DOX (Figure 6). KBvin cells were pretreated with
compounds followed by addition of DOX. Intracellular
accumulation of DOX was measured as the fluorescence
intensity and standardized as fold ratio. All DDB analogues
induced DOX accumulation in KBvin cells at 1.2- to 2.4-fold.
The cellular accumulation of DOX by DDB analogues was
The Effect of DDB Analogues on P-gp Function in
KBvin Cells. To confirm our hypothesis that DDB analogues
inhibit efflux activity of P-gp resulting in elevated concentration
5417
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
Table 1. Reversal Effects with Paclitaxel (TAX), Vincristine (VCR), and Doxorubicin (DOX) in KBvin
a
b
c
group
compd
IC50 of TAX (nM)
fold
IC50 of VCR (nM)
fold
IC50 of DOX (nM)
fold
976.8 (TAX alone)
82.93
46.52
26.15
30.50
37.06
8.72
−
12
2520.7 (VCR alone)
80.96
41.42
7.22
−
31
1942.3 (DOX alone)
507.5
284.0
42.6
−
4
I
4
5
21
61
7
6
37
349
50
46
4
7
32
50.81
35.92
31.75
15.65
47.31
8.14
510.9
93.9
8
26
70
21
6
9
112
24
79
339.7
221.5
61.9
10
11
13
14
15
16
17
19
20
21
22
23
VRP
40.74
39.11
48.65
39.89
9.55
161
53
9
25
31
23
30
4
II
20
310
239
230
560
54
86.3
24
10.55
10.98
4.50
65.8
III
102
326
21
547.8
291.5
573.9
157.7
130.8
131.4
68.80
84.2
2.99
7
47.45
7.65
46.53
20.21
39.61
40.53
10.10
4.45
3
V
167
27
101
51
9
46.82
47.23
14.25
4.41
11
11
21
23
9
27
50
89
202
566
109
222
31
31.87
23.04
219
a
b
c
Concentration of compound: 10 μM. SD is shown in Supporting Information. The reversal fold values were calculated as the following: reversal
fold = IC50 (anticancer drug alone)/IC50 (anticancer drug + test compound).
consistent with sensitization of KBvin cells to DOX (Table 1).
CONCLUSIONS
■
Thus, these data further support that DDB-derived chemo-
sensitizers function as P-gp inhibitors resulting in cellular
accumulation of anticancer drugs.
Multidrug resistance is still a serious barrier to successful cancer
chemotherapy. Among the possible reasons for multidrug
resistance, the reduced intracellular concentration of anticancer
drug caused by the drug efflux pumps, such as P-gp, is a major
cause of chemoresistance. Effective P-gp modulators are still
unavailable for clinical use, partly because of their toxicity. We
selected DDB, a clinically used hepatoprotective compound, as
a lead, and 33 new DDB analogues were newly designed and
synthesized. All synthesized analogues were evaluated for MDR
chemosensitizing effects on clinically used anticancer drugs,
such as TAX, VCR, and DOX. We succeeded to improve
chemoreversal action by introducing a 2,2′-methyl ester group
to DDB, and SAR studies are summarized in Figure 7. Insertion
of a halogen onto the 3-position reduced the ability. In the 2-
position of the ester side chain, bulky groups, such as pivalate,
2-methylbutanoate, cyclic aliphatic, and trimethoxyphenyl,
tended to enhance the reversal ability of DDB analogues.
Among all tested compounds, DDB derivatives 16 and 23 were
5−10 times more effective than VRP for TAX and VCR
reversal ability. Analogue 6 also showed 5-fold greater
chemosensitizing effect against DOX than VRP. Importantly,
active DDB analogues displayed no cytotoxicity against tumor
cells established from different tissues, suggesting that our novel
DDB analogues are significant lead compounds for further
clinical development to overcome the MDR phenotype.
Intracellular accumulation studies using calcein-AM and DOX
in KBvin cells clearly demonstrated that DDB analogues
interfere with the P-gp drug efflux pump. To conclude, newly
synthesized DDB analogues were identified as P-gp inhibitors
possessing a new scaffold for nontoxic chemosensitizer drug
development.
Further screening studies demonstrated that DDB analogues
sensitized NIH3T3-MDR cells (murine fibroblast NIH3T3
cells with overexpressing human P-gp protein) to TAX and
VCR (unpublished data). These results also support our
conclusion that DDB analogues interfere with the drug efflux
function of P-gp.
Hydrophobicity Evaluation of Active DDB Analogues.
P-gp contains a large drug-binding pocket with a volume of
around 6000 Å.23 The pocket includes predominantly hydro-
phobic and aromatic residues in its upper half and more polar
and charged residues in its lower half. Thus, because of
similarity, hydrophobic substrates would bind to the hydro-
phobic residues, and aromatic substrates would overlap with the
π-orbitals of aromatic residues in the binding pocket. Further
evaluation of the chemical structures of various known P-gp
inhibitors identified common features, including high hydro-
phobicity, two or more aromatic rings, a methoxy group on the
aromatic ring (hydrogen bond acceptor), and one or two
protonatable nitrogens.24,25 Because the upper half of the
presumptive drug-binding pocket is composed of mainly
hydrophobic and aromatic residues,23 it is possible that
hydrophobicity of DDB analogues could influence their MDR
reversal activity. Table 2 shows the clogP values of synthesized
compounds. Although a few exceptions were present, the
chemosensitizing effects of compounds were moderately
correlated with their clogP values. Active compounds had
clogP values of 4−8; those with clogP lower than 4 tended to
lose chemosensitizing activity. The clogP values of 6, 16, and
23, which were significantly active as described above, were
between 4.8 and 6.2, which is close to that of VRP. This fact
implied that hydrophobicity is an important parameter in P-gp
inhibition.
EXPERIMENTAL SECTION
■
General. 1H NMR (400 MHz) spectra were measured on a Varian
Inova spectrometer with TMS as the internal standard. Mass spectra
were measured on a Shimazu LCMS-IT-TOF. All reactions were
5418
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
Figure 3. Reversal of chemosensitivity of KBvin by 6 or 23. Chemoresistant KBvin cells were incubated with various concentrations of anticancer
drugs TAX (A), VCR (B), or DOX (C) in the presence of test compounds, as indicated, for 72 h to evaluate the effect on chemosensitization. KB
○
●
cells ( ) were sensitive to anticancer drugs, while KBvin ( ) were resistant in the absence of test compounds. Chemosensitization of KBvin cells
⧫
■
▲
was observed when the cells were cotreated with 10 μM of 6 ( ), 23 ( ), or VRP ( ).
5419
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
Figure 4. Dose−response effect of compounds 6 and 23 on sensitization of KBvin to TAX. Multidrug-resistant KBvin cells were treated with various
concentrations of compounds 6 or 23 in the presence of 100 nM TAX, an absolutely nontoxic concentration for KBvin. Data are expressed as mean
SE of three independent experiments. Calculated median effective concentration (EC50) of compounds 6, 23, or VRP was 2.81 μM, 1.87 μM, or
2.71 μM, respectively.
(CDCl3) δ 6.68 (s, 1H), 5.96 (d, J = 1.4 Hz, 2H), 5.95 (d, J = 1.4 Hz,
2H) 4.86 (s, 4H), 3.93 (s, 6H), 2.22 (t, J = 7.2, 4H), 1.59 (sex., J = 7.2
Hz, 4H), 0.90 (t, J = 7.2 Hz, 6H); HRMS calcd for C26H30NaO10 (M +
Na)+ 525.1737, found 525.1759.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(2-methylbutanoate) (8). Yield 83%; colorless oil ;
1H NMR (CDCl3) δ 6.68 (s, 2H), 5.97 (s, 2H), 5.95 (s, 2H), 4.91−
4.82 (m, 4H), 3.93 (s, 6H), 2.32 (sex., J = 6 Hz, 2H), 1.67−1.56(m,
2H), 1.48−1.36 (m, 2H), 1.09 (d, J = 7.2 Hz, 3H), 1.07 (d, J = 7.2 Hz,
3H), 0.84 (dd, J = 7.4, 14.6 Hz, 6H); HRMS calcd for C28H34NaO10
(M + Na)+ 553.2050, found 553.2063.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(3-methylbutanoate) (9). Yield 91%; colorless oil;
1H NMR (CDCl3) δ 6.68 (s, 2H), 5.96 (s, 2H), 5.95 (s, 2H), 4.86 (s,
4H), 3.93 (s, 6H), 2.12 (d, J = 6.4 Hz, 4H), 2.03 (m, 2H), 0.90 (dd, J
= 2.8, 6.6 Hz, 12H); HRMS calcd for C28H34NaO10 (M + Na)+
553.2050, found 553.2063.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Dipentanoate (10). Yield 99%; colorless oil; 1H NMR
(CDCl3) δ 6.68 (s, 2H), 5.96 (d, J = 1.5 Hz, 2H), 5.94 (d, J = 1.5 Hz,
2H), 4.85 (s, 4H), 3.93 (s, 6H), 2.23 (t, J = 7.6 Hz, 4H), 1.54 (pent, J
= 7.6 Hz, 4H), 1.29 (sex., J = 7.6 Hz, 4H), 0.88 (t, J = 7.6 Hz, 6H);
HRMS calcd for C28H34NaO10 (M + Na)+ 553.2050, found 553.2067.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(2-methylpentanoate) (11). Yield 85%; colorless
Figure 5. Effect of compounds on P-gp function in KBvin cells. KBvin
cells were pretreated with compounds followed by addition of calcein-
AM. The cellular accumulation of calcein is represented by the relative
fluorescent unit ( × 104 RFU). Cellular accumulation of calcein
demonstrates inhibition of efflux activity of P-gp. Data with mean
SD of three independent experiments are shown in Supporting
Information Table S2.
1
oil; H NMR (CDCl3) δ 6.68 (s, 2H), 5.97 (s, 2H), 5.95 (s, 2H)
monitored by thin-layer chromatography (TLC) on aluminum sheets
(silica gel 60 F254 plate, 20 × 20, Merck). Melting points were
recorded on a Fisher Johns melting apparatus without correction.
Medium-pressure column chromatography was used in Biotage Flash
and Isco companion systems with silica 40 μM columns from Grace
Inc. All final compounds are >95% pure based on HPLC. Anhydrous
solvents were purchased from commercial suppliers.
General Synthetic Procedure for Compounds 3, 7−12, 15−
17, and 22. To a solution of 2 in CH2Cl2 were added an appropriate
carboxylic acid (5 equiv mol), N-(3-dimethylaminopropyl)-N′-ethyl-
carbodiimide hydrochloride (5 equiv mol), and 4-(dimethylamino)-
pyridine (1 equiv mol) and stirred overnight. The reaction mixture was
then applied directly to preparative TLC (hexane−EtOAc) without
workup.
4.91−4.81 (m, 4H), 3.93 (s, 6H), 2.37 (dd, J = 7.2, 14 Hz, 2H), 1.64−
1.13 (m, 8H), 1.09 (d, J = 7 Hz, 3H), 1.07 (d, J = 7 Hz, 3H), 0.86 (t, J
= 6.8 Hz, 6H); HRMS calcd for C30H38NaO10 (M + Na)+ 609.2676,
found 609.2688.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Didodecanoate (12). Yield 88%; colorless oil; 1H
NMR (CDCl3, 400 MHz) δ 6.68 (s, 2H), 5.96 (d, J = 1.5 Hz, 2H),
5.94 (d, J = 1.5 Hz, 2H), 4.85 (s, 4H), 3.93 (s, 6H), 2.22 (t, J = Hz,
4H), 1.57−1.19 (m), 0.87 (t, J = 6.8 Hz, 6H); HRMS calcd for
C42H62NaO10 (M + Na)+ 749.4247, found 749.4225.
(2E,2′E)-(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-
2,2′-diyl)bis(methylene) Bis(but-2-enoate) (15). Yield 8%; colorless
1
oil; H NMR (CDCl3) δ 6.96−6.87 (m), 6.7 (s, 2H), 5.94 (s, 4H),
5.78 (d, J = 15.6 Hz, 2H), 4.94 (d, J = 12.4 Hz, 2H), 4.85 (d, J = 12.4
Hz, 2H), 3.93 (s, 6H), 1.85 (d, J = 6.8 Hz, 6H); HRMS calcd for
C26H26NaO10 (M + Na)+ 521.1424, found 521.1449.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Diacetate (3). Yield 85%; colorless oil; 1H NMR
(CDCl3) δ 6.69 (s, 2H), 5.97 (s, 2H), 5.95 (s, 2H), 4.85 (s, 4H), 3.95
(s, 6H), 1.99 (s, 6H); HRMS calcd for C22H22NaO10 (M + Na)+
469.1111, found 469.1116.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(3-methylbut-2-enoate) (16). Yield 51%; colorless
1
oil; H NMR (CDCl3) δ 6.70 (s, 2H), 5.94 (d, J = 1.5 Hz, 2H), 5.93
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
1
bis(methylene) Dibutyrate (7). Yield 97%; colorless oil; H NMR
(d, J = 1.5 Hz, 2H), 5.62 (s, 2H), 4.95 (d, J = 12.5 Hz, 2H), 4.82 (d, J
5420
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
Figure 6. Recovered DOX accumulation in DDB analogue-treated drug resistant KBvin cells. KBvin cells were incubated in DOX medium (final
concn 10 μM) for 3 h in the presence of 10 μM DDB analogues, and then cellular accumulation of DOX was measured as the intrinsic fluorescence
intensity of DOX. The fluorescence intensity of DOX was expressed as the ratio of effect of compound to negative control (Dox). Intracellular
accumulation of DOX was clearly observed in the presence of DDB-derived compounds. Data are represented as mean SD, n = 3. P-gp inhibitor
VRP (10 μM) or cyclosporine (CsA) (5 μM) was used as a positive control.
nitrogen. The reaction was warmed to room temperature gradually
and stirred for 1−3 h. After the reaction was completed, water and
saturated sodium carbonate solution were added and the reaction
mixture was extracted with dichloromethane, dried over sodium
sulfate, and concentrated. Further purification was performed by
Combiflash (hexane−EtOAc gradient).
Table 2. cLogP Values of Synthesized DDB-Derived
Compounds
a
compd
clogP
compd
clogP
1
4.27
1.26
2.98
4.03
4.65
5.45
5.09
5.71
5.89
6.15
6.77
N/A
5.92
7.04
5.24
6.04
6.51
2.40
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
VRP
6.93
7.93
7.28
6.69
5.93
6.42
5.80
7.31
6.11
6.21
4.68
5.02
2.60
7.49
8.27
1.96
5.34
2.28
4.58
2
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
3
1
bis(methylene) Dipropionate (4). Yield 99%; colorless oil; H NMR
4
(CDCl3) δ 6.68 (s, 2H), 5.95 (d, J = 6.5 Hz, 4H), 4.85 (s, 4H), 3.93 (s,
6H), 2.26 (q, J = 7.2 Hz, 4H), 1.08 (t, J = 7.6 Hz, 6H); HRMS calcd
for C24H27O10 (M + H)+ 497.1424, found 497.1432.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(2-methylpropanoate) (5). Yield 83%; colorless
oil; 1H NMR (CDCl3) δ 6.68 (s, 2H), 5.96 (dd, J = 9.2 Hz, 4H), 4.88
(d, J = 12.6 Hz, 2H), 4.84 (d, J = 12.6 Hz, 2H), 3.93 (s, 6H), 2.49
(sept, J = 7.2 Hz, 2H), 1.11 (t, J = 7.2 Hz, 12H); HRMS calcd for
C26H30NaO10 (M + Na)+ 525.1737, found 525.1754.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(2,2-dimethylpropanoate) (6). Yield 82%; color-
less oil; 1H NMR (CDCl3) δ 6.68 (s, 2H), 5.96 (d, J = 13.8 Hz, 4H),
4.91 (d, J = 12.8 Hz, 2H), 4.82 (d, J = 12.8 Hz, 2H), 3.93 (s, 6H), 1.15
(s, 18H); HRMS calcd for C28H34NaO10 (M + Na)+ 553.2050, found
553.2054.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Dicyclopentanecarboxylate (13). Yield 99%; color-
1
a
less oil; H NMR (CDCl3) δ 6.68 (s, 2H), 5.95 (d, J = 7.9 Hz, 4H),
clogP was calculated by ChemDraw Ultra Version 12.0.
4.88 (d, J = 12.6 Hz, 2H), 4.84 (d, J = 12.6 Hz, 2H), 3.93 (s, 6H),
2.71−2.65 (m, 2H), 1.88−1.52 (m, 16H); HRMS calcd for
C30H34NaO10 (M + Na)+ 577.2050, found 577.2052.
= 12.5 Hz, 2H), 3.93 (s, 6H), 2.12 (s, 6H), 1.86 (s, 6H); HRMS calcd
for C28H30NaO10 (M + Na)+ 549.1737, found 549.1761.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(hexa-2,4-dienoate) (17). Yield 55%; colorless oil;
1H NMR (CDCl3) δ 7.19 (d, J = 9.8 Hz, 1H), 7.15 (d, J = 9.8 Hz, 1H),
6.70 (s, 2H), 6.18−6.08 (m, 4H), 5.93 (d, J = 1.5 Hz, 4H), 5.71 (s,
1H), 5.68 (s, 1H), 4.96 (d, J = 12.3 Hz, 4H), 4.87 (d, J = 12.3 Hz, 4H),
3.93 (s, 6H), 1.83 (d, J = 5.2 Hz, 6H); HRMS calcd for C30H30NaO10
(M + Na)+ 573.1737, found 573.1746.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(3,4-dimethoxybenzoate) (22). Yield 90%; color-
less oil; 1H NMR (CDCl3) δ 7.55 (dd, J = 2, 8.4 Hz, 2H), 7.43 (d, J =
2 Hz, 2H), 6.78 (d, J = 8.4 Hz, 2H), 6.76 (s, 2H), 5.91 (s, 2H), 5.81 (s,
2H), 5.14 (d, J = 12.4 Hz, 2H), 5.04 (d, J = 12.4 Hz, 2H), 3.93 (s, 6H),
3.89 (s, 6H), 3.88 (s, 6H); HRMS calcd for C36H34NaO14 (M + Na)+
713.1846, found 713.1839.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Dicyclohexanecarboxylate (14). Yield 99%; colorless
1
oil; H NMR (CDCl3) δ 6.67 (s, 2H), 5.95 (d, J = 7.6 Hz, 4H), 4.87
(d, J = 12.6 Hz, 2H), 4.83 (d, J = 12.6 Hz, 2H), 3.93 (s, 6H), 2.27−
2.20 (m, 2H), 1.95−1.13 (m, 20H); HRMS calcd for C32H38NaO10
(M + Na)+ 605.2363, found 605.2387.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Dibenzoate (19). Yield 59%; colorless amorphous; 1H
NMR (CDCl3) δ 7.94−7.92 (m, 4H), 7.52−7.47 (m, 2H), 7.38−7.34
(m, 4H), 6.76 (s, 2H), 5.90 (s, 2H), 5.79 (s, 2H), 5.15 (d, J = 12.3 Hz,
2H), 5.06 (d, J = 12.3 Hz, 2H), 3.93 (s, 6H); HRMS calcd for
C32H26NaO10 (M + Na)+ 593.1424, found 593.1452.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(4-methylbenzoate) (20). Yield 34%; colorless
amorphous; 1H NMR (CDCl3) δ 7.81 (d, J = 8.0 Hz, 4H), 7.15 (d, J =
8.0 Hz, 4H), 6.75 (s, 2H), 5.91 (s, 2H), 5.81 (s, 2H), 5.12 (d, J = 12.5
Hz, 2H), 5.04 (d, J = 12.5 Hz, 2H), 3.92 (s, 6H), 2.35 (s, 6H) ; HRMS
calcd for C34H30NaO10 (M + Na)+ 621.1737, found 621.1761.
General Procedure for Compound 4−6, 13, 14, 19−21, and
23−28. To a flask with anhydrous dichloromethane were added
compound 2 and triethylamine (5−10 equiv mol) first, and then the
appropriate acyl chloride (2.2 equiv mol) was added at 0 °C under
5421
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(4-methoxybenzoate) (21). Yield 30%; colorless
amorphous; H NMR (CDCl3) δ 7.90−7.86 (m, 4H), 6.85−6.81 (m,
diisobutylaluminum hydride (DIBAL) (0.83 mL, 0.83 mmol)
dropwise. After 1 h, another portion of DIBAL (0.4 mL, 0.4 mmol
was added and stirring continued until starting material disappeared.
The reaction was quenched with MeOH (3 mL), 10% Rochelle salt
solution (3 mL) was added, and the mixture was stirred for 30 min.
Water was added to the mixture, which was then extracted with EtOAc
three times, dried over sodium sulfate, and concentrated. The
compound has low solubility in various solvents (EtOAc, CH2Cl2,
MeOH, acetone). A small portion was taken for further purification for
bioassay. The remaining portion was used in the next reaction without
1
4H), 6.75 (s, 2H), 5.91 (s, 2H), 5.82 (s, 2H), 5.12 (d, J = 12.3 Hz,
2H), 5.03 (d, J = 12.3 Hz, 2H), 3.93 (s, 6H), 3.80 (s, 6H); HRMS
calcd for C34H30NaO12 (M + Na)+ 653.1635, found 653.1615.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(2,3,4-trimethoxybenzoate) (23). Yield 98%;
colorless amorphous; 1H NMR (CDCl3) δ 7.18 (s, 4H), 6.73 (s,
2H), 5.90 (s, 2H), 5.77 (s, 2H), 5.16 (d, J = 12.3 Hz, 2H), 5.04 (d, J =
12.3 Hz, 2H), 3.92 (s, 6H), 3.87 (s, 6H), 3.86 (s, 12H); HRMS calcd
for C38H39O16 (M + H)+ 773.2058, found 773.2050.
1
further purification. Colorless amorphous solid; H NMR (CDCl3) δ
5.96 (s, 4H), 4.67 (d, J = 12.4 Hz, 2H), 4.26 (d, J = 12 Hz, 2H), 4.09
(s, 6H), 3.19 (bs, 2H); HRMS calcd for C18H16Br2NaO8 (M + Na)+
542.9089, found 542.9091.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(4-nitrobenzoate) (24). Yield 52%; colorless
amorphous; 1H NMR (CDCl3) δ 8.22−8.19 (m, 4H), 8.09−8.06
(m, 4H), 6.75 (s, 2H), 5.92 (s, 2H), 5.85 (s, 2H), 5.18 (d, J = 12.4 Hz,
2H), 5.12 (d, J = 12.4 Hz, 2H), 3.94 (s, 6H); HRMS calcd for
C32H25N2O14 (M + H)+ 683.1125, found 683.1119.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(4-cyanobenzoate) (25). Yield 38%; colorless
amorphous; 1H NMR (CDCl3) δ 8.02−7.99 (m, 4H), 7.69−7.66
(m, 4H), 6.73 (s, 2H), 5.91 (s, 2H), 5.81 (s, 2H), 5.16 (d, J = 12.4 Hz,
2H), 5.08 (d, J = 12.4 Hz, 2H), 3.94 (s, 6H); HRMS calcd for
C34H24N2NaO10 (M + Na)+ 643.1329, found 643.1314.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(benzo[d][1,3]dioxole-5-carboxylate) (26). Yield
85%; colorless amorphous; 1H NMR (CDCl3) δ 7.53 (m, 2H), 7.34 (s,
2H), 6.77 (s, 2H), 6.75 (s, 2H), 5.99 (s, 4H), 5.93 (s,2H), 5.88 (s,
2H), 5.10 (d, J = 12.4 Hz, 2H), 5.02 (d, J = 12.4 Hz, 2H), 3.93 (s,
6H); HRMS calcd for C34H26NaO14 (M + Na)+ 681.1220, found
681.1188.
(3,3′-Dibromo-4,4′-dimethoxy-5,6,5′,6′-dimethylenedioxybi-
phenyl-2,2′-diyl)bis(methylene) Dibutyrate (32). Compound 31
(16.8 mg, 0.033 mmol), butyric acid (0.02 mL, 0.21 mmol), N-(3-
dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (31.5 mg,
0.16 mmol), and 4-(dimethylamino)pyridine (4.1 mg, 0.033 mmol)
were mixed together in CH2Cl2 overnight. The reaction mixture was
subjected to preparative TLC to give the desired compound, which
was recrystallized from CH2Cl2−hexane. Yield 86%; colorless prisms;
mp: 109−110 °C; 1H NMR (CDCl3) δ 5.94 (d, J = 1.2 Hz, 2H), 5.92
(d, J = 1.6 Hz, 2H), 4.97 (d, J = 1.2 Hz, 2H), 4.89 (d, J = 1.2 Hz, 2H),
4.06 (s, 6H), 2.18 (t, J = 7.6 Hz, 6H), 1.57 (sex., J = 7.6 Hz, 4H), 0.88
(t, J = 7.2 Hz, 6H); HRMS calcd for C28H32Br2NaO10 (M + Na)+
682.9926, found 682.9916.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Dibenzoate (33). The same procedure as for
compound 32, but benzoic acid was used instead of butyric acid.
1
Yield 74%; colorless prisms; mp: 127−128 °C; H NMR (CDCl3) δ
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
7.92 (d, J = 7.6 Hz, 4H), 7.46−7.52 (m, 2H), 7.36 (t, J = 7.6 Hz, 4H),
5.88 (s, 2H), 5.70 (s, 2H), 5.31 (d, J = 11.6 Hz, 2H), 5.12 (d, J = 11.6
Hz, 2H), 4.07 (s, 3H); HRMS calcd for C32H24Br2NaO10 (M + Na)+
750.9613, found 750.9614.
bis(methylene) Bis[6-(trifluoromethyl)nicotinate] (27). Yield 61%;
1
colorless oil; H NMR (CDCl3) δ 9.15 (m, 2H), 8.38 (m, 2H), 7.71
(m, 2H), 6.74 (s, 2H), 5.93 (dd, J = 1.6, 8.8 Hz, 4H), 5.21 (d, J = 7.6
Hz, 4H), 3.93 (s, 3H); HRMS calcd for C32H22F6N2NaO10 (M + Na)+
731.1076, found 731.1076.
3,3′-Dibromo-4,4′-dimethoxy-5,6,5′,6′-bis(methylenedioxy)-
biphenyl-2-hydroxymethyl-2′-carboxylic Acid (36). To a stirred
solution containing anhydrous THF (5 mL) and 35 (69.4 mg, 0.13
mmol) under nitrogen was added sodium borohydride (15 mg, 0.39
mmol), followed by MeOH (0.1 mL), at room temperature. After the
reaction was completed, 2 N HCl solution was added to acidify the
solution to pH = 2. The solution was extracted with CH2Cl2 (with less
than 10% MeOH), and the organic layer was dried over sodium
sulfate. Flash chromatography (CH2Cl2−MeOH) was used to purify
the desired compound. Yield 88%; amorphous solid; 1H NMR
(DMSO, 400 MHz) δ 13.2 (bs), 6.04 (d, J = 34 Hz, 2H), 5.97 (d, J =
40 Hz, 2H), 4.55 (bs), 4.28 (dd, J = 11.2, 25.8 Hz, 4H), 3.99 (s, 6H),
3.95 (s, 6H); HRMS calcd for C18H14Br2NaO9 (M + Na)+ 582.9029,
found 582.9038.
Cell Culture. A549 (lung carcinoma), DU-145 (prostate cancer),
K562 (chronic myelogenous leukemia), and KB (epidermoid
carcinoma) cell lines were obtained from Lineberger Comprehensive
Cancer Center (UNC-CH). KBvin (vincristine-resistant KB subline)
was generously provided by Professor Y.-C. Cheng (Yale University,
CT). Cells were cultured in RPMI 1640 medium supplemented with
25 mM HEPES, 2 mM L-glutamine (Mediatech), 10% heat inactivated
fetal bovine serum (Hyclone), 100 IU of penicillin, 100 μg/mL
streptomycin, and 0.25 μg/mL amphotericin B (Mediatech). KBvin
cells were maintained in the culture medium containing 100 nM VCR.
Cells were maintained at 37 °C in a humidifier with 5% CO2
atmosphere.
(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-diyl)-
bis(methylene) Bis(quinoline-2-carboxylate) (28). Yield 97%; orange
1
needles; mp: 129−130 °C; H NMR (CDCl3) δ 8.25 (m, 2H), 8.16
(m, 2H), 7.98 (m, 2H), 7.77−7.70 (m, 4H), 7.59−7.55 (m, 2H), 6.86
(s, 2H), 5.88 (dd, J = 1.6, 19.6 Hz, 4H), 5.32 (d, J = 2.4 Hz, 4H), 3.93
(s, 6H); HRMS calcd for C38H29N2O10 (M + H)+ 673.1822, found
673.1798.
4,4′-[(4,4′-Dimethoxy-5,6,5′,6′-dimethylenedioxybiphenyl-2,2′-
diyl)bis(methylene)]bis(oxy)bis(4-oxobutanoic acid) (18). To a
solution of 2 (26 mg, 0.072 mmol) in anhydrous THF (2.0 mL)
were added succinic anhydride (9 mg, 0.090 mmol) and DMAP (5 mg,
0.040 mmol). The mixture was refluxed overnight. After cooling to rt,
the whole was acidified with 1 N HCl aq and partitioned with EtOAc.
The organic phase was concentrated. The residue was purified by
preparative TLC (CH2Cl2:MeOH:TFA = 95:5:0.25). Yield 99%;
1
colorless oil; H NMR (CDCl3) δ 6.67 (s, 2H), 5.96 (d, J = 16.8 Hz,
4H), 4.94 (d, J = 12.6 Hz, 2H), 4.89 (d, J = 12.6 Hz, 2H), 3.93 (s, 6H),
2.59−2.54 (m, 8H); HRMS calcd for C26H26NaO14 (M + Na)+
585.1220, found 585.1228.
Dimethyl 3,3′-Diiodo-4,4′-dimethoxy-5,6,5′,6′-dimethylenediox-
ybiphenyl-2,2′-dicarboxylate (30). Silver trifluoroactate (89 mg, 0.4
mmol) was added to a solution containing 1 (42.1 mg, 0.1 mmol) and
CHCl3 (1 mL). Then iodine (102.2 mg, 0.4 mmol) was poured into
the solution, and the reaction mixture was stirred overnight at room
temperature. Further isolation was performed by preparative TLC
(hexane−EtOAc: 7:3). Mono- and diiodo products were found under
these conditions. Yield 7%; colorless amorphous; 1H NMR (CDCl3) δ
6.01 (d, J = 1.6 Hz, 2H), 5.99 (d, J = 1.2 Hz, 2H), 4.05 (s, 6H), 3.68
(s, 6H); HRMS calcd for C20H16I2NaO10 (M + Na)+ 692.8731, found
692.8704.
Cytotoxicity Analysis (SRB assay). Cytotoxicity was determined
by the sulforhodamine B (SRB) colorimetric assay. Cells (3−5 × 103
cells/well) were seeded in 96-well plates filled with culture medium
containing various concentrations of samples for 72 h. At the end of
the exposure period, the supernatant was removed and cells were
washed with 100 μL of fresh culture medium. The proliferated cells
were fixed with 50% trichloroacetic acid for 30 min and stained with
0.04% SRB (Sigma Chemical Co.) for 30 min. Protein-bound SRB dye
was dissolved from stained cells in 10 mM Tris base, and absorbance
3,3′-Dibromo-4,4′-dimethoxy-5,6,5′,6′-bis(methylenedioxy)-
biphenyl-2,2′-dimethanol (31). To a stirred solution containing 1
(95.2 mg, 0.165 mmol) and 1 mL of anhydrous CH2Cl2 under
nitrogen at around −20 °C (ice with brine) was added
5422
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
was measured at 515 nm on Microplate Reader ELx800 (Bio-Tek
instruments, Winooski, VT) with a Gen5 software.
diimide hydrochloride; DMAP, 4-dimethylaminopyridine;
DIBAL, diisobutylaluminium hydride
MDR Reversal Activity. For screening of chemosensitizing ability
of test compounds, MDR and parental chemosensitive cells were
incubated with test compound in the presence of 100 nM TAX, which
did not affect the cell growth. Multidrug resistant KBvin and parental
KB cells were seeded at 5−7 × 103 cells/well into 96-well plates and
incubated with 10 μM test compound with 100 nM TAX for 72 h, and
cell density was determined by a SRB assay. To assess the reversal
activity of MDR by candidate compounds, it was evaluated by
comparing IC50 values of anticancer drugs (vincristine, paclitaxel, and
doxorubicin) in the absence or presence of 10 μM test compound.
IC50 values were calculated by log−linear interpolation of data points.
Verapamil, as a known P-gp inhibitor/modulator, was used as a
positive control in all experiments. The reversal fold value, as the
potency parameter of test compounds, was calculated as follows:
reversal fold = IC50 (anticancer drug alone)/IC50 (anticancer drug +
test compound). All experiments were performed at least three times.
Measurement of Intracellular Accumulation of Calcein or
Doxorubicin. KBvin cells (5 × 103 cell/well) were seeded in 96-well
plates and pretreated with samples for 1 h before calcein-AM or
doxorubicin was added to make final concentrations of 1 μM calcein-
AM and 10 μM doxorubicin. After 10 min for calcein-AM or 3 h for
DOX, the medium was removed by aspiration, and the cells were
washed with ice-cold PBS followed by lysing with PBS containing 1%
sodium dodecyl sulfate (SDS). The mean fluorescence intensity of
calcein or DOX was measured at Ex: 494 nm/Em: 517 nm or Ex: 488
nm/Em: 580 nm, respectively, by a fluorescence microplate reader
(Plate Chameleon Multilabel Detection Platform, Hidex Oy, Turku,
Finland) with MikroWin software. VRP or CsA was used as a positive
control of the P-gp modulator. All data for intracellular accumulation
of DOX were calculated as the fluorescence intensity and were
represented as fold changes by treatment with compound compared
with vehicle (DMSO) treatment.
REFERENCES
■
(1) Perez-Sayans, M.; Somoza-Martin, J. M.; Barros-Angueira, F.; Diz,
P. G.; Rey, J. M.; Garcia-Garcia, A. Multidrug resistance in oral
squamous cell carcinoma: The role of vacuolar ATPases. Cancer Lett.
2010, 295, 135−143.
(2) Zhou, T.; Shi, Q.; Bastow, K. F.; Lee, K. H. Antitumor agents.
286. Design, synthesis, and structure−activity relationships of 3′R,4′R-
disubstituted-2′,2′-dimethyldihydropyrano[2,3-f ]chromone (DSP) an-
alogues as potent chemosensitizers to overcome multidrug resistance.
J. Med. Chem. 2010, 53, 8700−8708.
(3) Fodale, V.; Pierobon, M.; Liotta, L.; Petricoin, E. Mechanism of
cell adaptation: when and how do cancer cells develop chemo-
resistance? Cancer J. 2011, 17, 89−95.
(4) Singh, A.; Settleman, J. EMT, cancer stem cells and drug
resistance: an emerging axis of evil in the war on cancer. Oncogene
2010, 29, 4741−4751.
(5) Torres-Romero, D.; Munoz-Martinez, F.; Jimenez, I. A.;
Castanys, S.; Gamarro, F.; Bazzocchi, I. L. Novel dihydro-β-agarofuran
sesquiterpenes as potent modulators of human P-glycoprotein
dependent multidrug resistance. Org. Biomol. Chem. 2009, 7, 5166−
5172.
(6) Das, S. G.; Doshi, J. M.; Tian, D.; Addo, S. N.; Srinivasan, B.;
Hermanson, D. L.; Xing, C. Structure-activity relationship and
molecular mechanisms of ethyl 2-amino-4-(2-ethoxy-2-oxoethyl)-6-
phenyl-4h-chromene-3-carboxylate (sha 14−1) and its analogues. J.
Med. Chem. 2009, 52, 5937−5949.
(7) Mor, G.; Montagna, M. K.; Alvero, A. B. Modulation of apoptosis
to reverse chemoresistance. Methods Mol. Biol. 2008, 414, 1−12.
(8) Zhang, P. Y.; Wong, I. L.; Yan, C. S.; Zhang, X. Y.; Jiang, T.;
Chow, L. M.; Wan, S. B. Design and syntheses of permethyl ningalin B
analogues: potent multidrug resistance (MDR) reversal agents of
cancer cells. J. Med. Chem. 2010, 53, 5108−5120.
(9) Robey, R. W.; Shukla, S.; Finley, E. M.; Oldham, R. K.; Barnett,
D.; Ambudkar, S. V.; Fojo, T.; Bates, S. E. Inhibition of P-glycoprotein
(ABCB1)- and multidrug resistance-associated protein 1 (ABCC1)-
mediated transport by the orally administered inhibitor, CBT-1®.
Biochem. Pharmacol. 2008, 75, 1302−1312.
(10) Sun, M.; Xu, X.; Lu, Q.; Pan, Q.; Hu, X.; Schisandrin, B. A dual
inhibitor of P-glycoprotein and multidrug resistance-associated protein
1. Cancer Lett. 2007, 246, 300−307.
(11) Qiangrong, P.; Wang, T.; Lu, Q.; Hu, X.; Schisandrin, B. A novel
inhibitor of P-glycoprotein. Biochem. Biophys. Res. Commun. 2005, 335,
406−411.
(12) Li, L.; Pan, Q.; Sun, M.; Lu, Q.; Hu, X. Dibenzocyclooctadiene
lignans -- A class of novel inhibitors of multidrug resistance-associated
protein 1. Life Sci. 2007, 80, 741−748.
(13) Sun, H.; Liu, G. T. Chemopreventive effect of dimethyl
dicarboxylate biphenyl on malignant transformation of WB-F344 rat
liver epithelial cells. Acta Pharmacol. Sin. 2005, 26, 1339−1344.
(14) Jin, J.; Sun, H.; Wei, H.; Liu, G. The anti-hepatitis drug DDB
chemosensitizes multidrug resistant cancer cells in vitro and in vivo by
inhibiting P-gp and enhancing apoptosis. Invest. New Drugs 2007, 25,
95−105.
ASSOCIATED CONTENT
■
S
* Supporting Information
Cytotoxicity of DDB analogues (Table S1). IC50
standard
deviation (SD) for reversal effects with paclitaxel (TAX),
vincristine (VCR), and doxorubicin (DOX) in KBvin (Table
S2). Standardized P values of compounds 6, 23, and VRP
(Table S3). Effect of compounds 6 and 23 on P-gp function in
KBvin cells (Table S4). This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*(K.H.L.): phone: 919-962-0066; fax: 919-966-3893; e-mail:
khlee@unc.edu. (K.N.G.): phone: 919-843-6325; e-mail:
goto@email.unc.edu.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This study was supported by grant CA-17625-32 from the
National Cancer Institute, NIH, awarded to K. H. Lee. Support
in part was also provided by the Cancer Research Center of
Excellence (CRC) (DOH-100-TD-C-111-005).
(15) Yu, H. Q.; Yang, X. Y.; Zhang, Y. X.; Shi, J. Z. Biphenyl-dimethyl
dicarboxylate in treating and preventing hepatitis due to drug
poisoning. Chin. Med. J. (Beijing, China, Engl. Ed.) 1987, 100, 122−123.
(16) Huber, R.; Hockenjos, B.; Blum, H. E. DDB treatment of
patients with chronic hepatitis. Hepatology 2004, 39, 1732−1733.
(17) Zhu, B.; Liu, G. T.; Zhao, Y. M.; Wu, R. S.; Strada, S. J.
Chemosensitizing multiple drug resistance of human carcinoma by
Bicyclol involves attenuated p-glycoprotein, GST-P and Bcl-2. Cancer
Biol. Ther. 2006, 5, 536−543.
ABBREVIATIONS USED
■
DDB, dimethyl-4,4′-dimethoxy-5,6,5′,6′-dimethylenedioxybi-
phenyl-2,2′-dicarboxylate; MDR, multidrug resistance; P-gp,
P-glycoprotein; MRP, multidrug resistance-associated protein;
TAX, paclitaxel; VCR, vincristine; DOX, doxorubicin; VRP,
verapamil; EDCI, N-(3-dimethylaminopropyl)-N′-ethylcarbo-
(18) Zhu, B.; Liu, G. T.; Wu, R. S.; Strada, S. J. Chemoprevention of
bicyclol against hepatic preneoplastic lesions. Cancer Biol. Ther. 2006,
5, 1665−1673.
5423
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424

Journal of Medicinal Chemistry
Article
(19) Xie, L.; Xie, J. X.; Kashiwada, Y.; Cosentino, L. M.; Liu, S. H.;
Pai, R. B.; Cheng, Y. C.; Lee, K. H. Anti-AIDS (acquired immune
deficiency syndrome) agents. 17. New brominated hexahydroxybi-
phenyl derivatives as potent anti-HIV agents. J. Med. Chem. 1995, 38,
3003−3008.
(20) Molander, G. A.; George, K. M.; Monovich, L. G. Total
synthesis of (+)-isoschizandrin utilizing a samarium(II) iodide-
promoted 8-endo ketyl-olefin cyclization. J. Org. Chem. 2003, 68,
9533−9540.
(21) Soai, K.; Yokoyama, S.; Mochida, K. Reduction of symmetric
and mixed anhydrides of carboxylic acids by sodium borohydride with
dropwise addition of methanol. Synthesis 1987, 647−648.
(22) Calcabrini, A.; Meschini, S.; Stringaro, A.; Cianfriglia, M.;
Arancia, G.; Molinari, A. Detection of P-glycoprotein in the nuclear
envelope of multidrug resistant cells. Histochem. J. 2000, 32, 599−606.
(23) Aller, S. G.; Yu, J.; Ward, A.; Weng, Y.; Chittaboina, S.; Zhuo,
R.; Harrell, P. M.; Trinh, Y. T.; Zhang, Q.; Urbatsch, I. L.; Chang, G.
Structure of P-glycoprotein reveals a molecular basis for poly-specific
drug binding. Science 2009, 323, 1718−1722.
(24) Pajeva, I. K.; Wiese, M. Pharmacophore model of drugs involved
in P-glycoprotein multidrug resistance: explanation of structural variety
(hypothesis). J. Med. Chem. 2002, 45, 5671−5686.
(25) Pajeva, I. K.; Globisch, C.; Wiese, M. Combined pharmacophore
modeling, docking, and 3D QSAR studies of ABCB1 and ABCC1
transporter inhibitors. ChemMedChem 2009, 4, 1883−1896.
5424
dx.doi.org/10.1021/jm300378k | J. Med. Chem. 2012, 55, 5413−5424