Synthesis and evaluation of stereoisomers of methylated catechin and epigallocatechin derivatives on modulating P-glycoprotein-mediated multidrug resistance in cancers

Article abstract of DOI:10.1016/j.ejmech.2021.113795

P-glycoprotein (P-gp; ABCB1)-mediated drug efflux causes multidrug resistance in cancer. Previous synthetic methylated epigallocatechin (EGC) possessed promising P-gp modulating activity. In order to further improve the potency, we have synthesized some novel stereoisomers of methylated epigallocatechin (EGC) and gallocatechin (GC) as well as epicatechin (EC) and catechin (C). The (2R, 3S)-trans-methylated C derivative 25 and the (2R, 3R)-cis-methylated EC derivative 31, both containing dimethyoxylation at ring B, tri-methoxylation at ring D and oxycarbonylphenylcarbamoyl linker between ring D and C3, are the most potent in reversing P-gp mediated drug resistance with EC₅₀ ranged from 32 nM to 93 nM. They are non-toxic to fibroblast with IC₅₀ > 100 μM. They can inhibit the P-gp mediated drug efflux and restore the intracellular drug concentration to a cytotoxic level. They do not downregulate surface P-gp protein level to enhance drug retention. They are specific for P-gp with no or low modulating activity towards MRP1- or BCRP-mediated drug resistance. In summary, methylated C 25 and EC 31 derivatives represent a new class of potent, specific and non-toxic P-gp modulator.

Full text of DOI:10.1016/j.ejmech.2021.113795

Journal Pre-proof
Synthesis and evaluation of stereoisomers of methylated catechin and
epigallocatechin derivatives on modulating P-glycoprotein-mediated multidrug
resistance in cancers
Iris L.K. Wong, Xing-kai Wang, Zhen Liu, Wenqin Sun, Fu-xing Li, Bao-chao Wang,
Peng Li, Sheng-biao Wan, Larry M.C. Chow
PII:
S0223-5234(21)00644-9
DOI:
https://doi.org/10.1016/j.ejmech.2021.113795
EJMECH 113795
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 23 April 2021
Revised Date: 13 July 2021
Accepted Date: 20 August 2021
Please cite this article as: I.L.K. Wong, X.-k. Wang, Z. Liu, W. Sun, F.-x. Li, B.-c. Wang, P. Li, S.-b. Wan,
L.M.C. Chow, Synthesis and evaluation of stereoisomers of methylated catechin and epigallocatechin
derivatives on modulating P-glycoprotein-mediated multidrug resistance in cancers, European Journal of
Medicinal Chemistry (2021), doi: https://doi.org/10.1016/j.ejmech.2021.113795.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2021 Published by Elsevier Masson SAS.

Graphical abstract

1
2
Synthesis and evaluation of stereoisomers of methylated catechin and epigallocatechin
derivatives on modulating P-glycoprotein-mediated multidrug resistance in cancers
#
1
#2
1,2
1
2
2
3
4
Iris L. K. Wong , Xing-kai Wang , Zhen Liu , Wenqin Sun , Fu-xing Li , Bao-chao Wang ,
2
2
1
Peng Li , Sheng-biao Wan* and Larry M. C. Chow*
5
6
7
1Department of Applied Biology and Chemical Technology and State Key Laboratory of
Chemical Biology and Drug Discovery, Hong Kong Polytechnic University, Hong Kong SAR,
China.
8
9
0
2Key Laboratory of Marine Drugs, Ministry of Education, School of Medicine and Pharmacy,
Ocean University of China, and Laboratory for Marine Drugs and Bioproducts of Qingdao
National Laboratory for Marine Science and Technology, Qingdao, China
1
1
1
1
2
#These two authors contribute equally to this work.
*Corresponding authors: Sheng-biao Wan and Larry M. C. Chow
1
1
3
4
Corresponding authors: Larry M. C. Chow, Tel: 852-34008662; Fax: 852-23649932; E-mail:
larry.chow@polyu.edu.hk (L.M.C.C.).
1
1
5
6
Sheng-biao
Wan,
Tel.:
86-532-82031087;
Fax:
86-532-82033054;
E-mail:
biaowan@ouc.edu.cn (S.B.W.).
1

1
7
ABSTRACT
1
1
2
2
2
2
2
2
2
2
2
2
3
3
8
9
0
1
2
3
4
5
6
7
8
9
0
1
P-glycoprotein (P-gp; ABCB1)-mediated drug efflux causes multidrug resistance in cancer.
Previous synthetic methylated epigallocatechin (EGC) possessed promising P-gp modulating
activity. In order to further improve the potency, we have synthesized some novel
stereoisomers of methylated epigallocatechin (EGC) and gallocatechin (GC) as well as
epicatechin (EC) and catechin (C). The (2R, 3S)-trans-methylated C derivative 25 and the (2R,
3R)-cis-methylated EC derivative 31, both containing dimethyoxylation at ring B, tri-
methoxylation at ring D and oxycarbonylphenylcarbamoyl linker between ring D and C3, are
the most potent in reversing P-gp mediated drug resistance with EC50 ranged from 32 nM to 93
nM. They are non-toxic to fibroblast with IC50 > 100 µM. They can inhibit the P-gp mediated
drug efflux and restore the intracellular drug concentration to a cytotoxic level. They do not
downregulate surface P-gp protein level to enhance drug retention. They are specific for P-gp
with no or low modulating activity towards MRP1- or BCRP-mediated drug resistance. In
summary, methylated C 25 and EC 31 derivatives represent a new class of potent, specific and
non-toxic P-gp modulator.
3
2
3
3
3
4
Keywords: P-glycoprotein (P-gp); Epigallocatechin (EGC); Gallocatechin (GC); Catechin
(C); Epicatechin (EC)
2

3
5
1. INTRODUCTION
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
The multidrug resistance (MDR) in cancer cells has been a major obstacle to successful
cancer chemotherapy in clinic. An important mechanism for MDR is the enhanced cellular
efflux of anticancer drugs by over-expression of ATP-binding cassette (ABC) transporter
proteins in tumor cells.[1] So far, P-glycoprotein (P-gp; ABCB1; MDR1) is the most well-
characterized ABC transporter and can transport a broad range of structurally diverse
anticancer drugs. Therefore, P-gp is a good drug target for treating multidrug resistant cancers.
Numerous P-gp inhibitors have been studied, including calcium channel blocker
verapamil[2-4] or its derivative dexverapamil,[5] antimalarial drug quinidine,[6] calmodulin
antagonists,[7, 8] the immunosuppressant cyclosporine A[9-12] or its derivatives PSC833
(valspodar),[13] some steroids,[14-16] dexniguldipine,[17] VX-710 (biricodar),[18, 19] zosuquidar
LY335979, tariquidar XR9576, laniquidar R101933, elacridar GF120918 and the substituted
diarylimidazole ONT-090.[20, 21] Among them, only a very few were selected for clinical trial
and none of them has been approved yet for clinical application.[22-25] These failures may be
because the previous clinical trials did not include patient selection to evaluate the expression
of drug transporters in the tumors. P-gp inhibitors might fail to overcome MDR due to the
overexpression of other ABC transporters like MRP1 or BCRP. It is better to monitor the
expression of ABC transporters in patient tumors before using any P-gp modulators. Other
factors may enhance the toxicity including drug-drug interaction between the anticancer drugs
and inhibitors and low specificity of the inhibitors itself. Further improvement of inhibitors of
ABC transporters should focus on potency, specificity and safety.
3

5
5
5
5
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
P-gp can be modulated by natural compounds including flavonoid, curcumin,
ginsenosides, piperine, catechins and silymarin for the purpose of reversing MDR in tumor
cells.[26-32] We have previously found that methylation of polyphenolic compounds such as
ningalin B and quercetin was effective in improving their P-gp modulating activity.[33-35] The
presence of ring D, O-methylation and linker modification of epigallocatechin (EGC, with 2R,
3R configuration) have been demonstrated to significantly improve their P-gp inhibitory
activities (Figure 1).[32] The EC50 values of EGC 4 was at least 5-fold lower than EGCG and
EGC 1 (Figure 1).[32]
Up till now, there was still no report concerning the P-gp modulating activities of
methylated epicatechin (EC) and catechin (C) derivatives. Currently, EC and C are not
promising P-gp modulators because they are rare components in green tea and their P-gp
modulating activity was low with effective concentration at 10 µM.[36] Despite this, our
previous study suggested that structural modifications of EC and C including methylation of
all hydroxyl groups on the rings and varying the linker rigidity between ring D and C3 position
can significantly improve their P-gp modulating activities.[32] To further understand the effect
of stereochemistry on the P-gp modulating activity of catechins, we have designed, synthesized
and evaluated more novel catechin stereoisomers including methylated EGC, methylated GC,
methylated epicatechin (methylated EC) and methylated catechin (methylated C) for their P-
gp modulating activities in breast cancer cells.
4

7
7
5
6
7
7
7
8
7
8
9
0
Figure 1. Improvement of P-gp modulating activity of EGC by presence of D ring, O-
8
8
1
2
methylation and linker modification. EGC 1 and EGC 4 (was named as 8 and 36 in [32]
)
5

8
8
8
3 2. RESULTS
4 2.1 Chemistry
5
Scheme 1 Synthetic route of stereoisomers of methylated GC derivatives
8
6
8
8
8
9
9
7
8
9
0
1
Synthesis of methylated gallocatechin derivatives is shown in scheme 1. (2S, 3R)-
pentamethylated gallocatechin 12, which was obtained from methylation of commercial
available (2S, 3R)-gallocatechin, was hydrolyzed by K
2
CO to afford intermediate 13 (with 2S,
3
3R configuration). Catalyzed by EDCI and DMAP, esterification of 13 with (E)-3-(3,4,5-
trimethoxyphenyl)acrylic acid or 4-fluoro-3-(3,4,5-trimethoxybenzamido)benzoic acid
6

9
9
9
9
9
9
2
3
4
5
6
7
produced target compounds 14 or 15, respectively. Compound 13 was reacted with 3,4,5-
trimethoxybenzoic acid, catalyzed by PPh and DIAD, gave a configuration-inversion product
3
of (2S, 3S)-pentamethylated epigallocatechin gallate 16. Hydrolysis of 16 provided 17 (2S, 3S),
the diastereomer of intermediate 13 (2S, 3R). Esterification of 17 with (E)-3-(3,4,5-
trimethoxyphenyl)acrylic acid or 4-fluoro-3-(3,4,5-trimethoxybenzamido)benzoic acid,
catalyzed by EDCI and DMAP, produced target compounds 18 or 19, respectively.
9
8
Scheme 2 Synthetic route of methylated catechin and epicatechin derivatives
9
9
7

1
1
1
1
1
1
1
1
1
1
1
1
1
00
01
02
03
04
05
06
07
08
09
10
11
12
Synthetic route of methylated catechin and epicatechin derivatives was shown in scheme
2. (2R, 3S)-tetramethylated catechin 20 was produced by methylation of commercial available
(2R, 3S)-catechin. Catalyzed by EDCI and DMAP, compound 20 was coupled with 3,4,5-
trimethoxybenzoic
acid,
(E)-3-(3,4-dimethoxyphenyl)acrylic
acid,
(E)-3-(3,4,5-
trimethoxyphenyl)acrylic acid, 4-fluoro-3-(3,4-dimethoxybenzamido)benzoic acid, 4-fluoro-
3-(3,4,5-trimethoxybenzamido)benzoic acid, (E)-3-(3-(3,4-dimethoxyphenyl)acrylamido)-4-
fluorobenzoic acid, or (E)-3-(3-(3,4,5-trimethoxyphenyl)acrylamido)-4-fluorobenzoic acid
provided target compounds 21, 22, 23, 24, 25, 26 or 27, respectively. (2R, 3R)-tetramethylated
epicatechin 28 was obtained from methylation of commercial available (2R, 3R)-epicatechin,
a diastereomer of (2R, 3S)-epicatechin. Compound 28 was reacted with 3,4,5-
trimethoxybenzoic acid, (E)-3-(3,4,5-trimethoxyphenyl)acrylic acid, or 4-fluoro-3-(3,4,5-
trimethoxybenzamido)benzoic acid catalyzed by EDCI and DMAP produced target compounds
29, 30 or 31, respectively.
8

1
13
Scheme 3 Synthetic route of compounds 33-44
1
1
14
15
Eight basic groups containing epigallocatechin derivatives and four basic groups
containing catechin derivatives were prepared and their synthetic routes were shown in Scheme
3. Catalyzed by EDCI and DMAP, important intermediates 1, 20 and 28 were reacted with
substituted piperidine-4-carboxylic acid or substituted piperidine-3-carboxylic acid to produce
target compounds 33-44, respectively.
1
1
1
1
1
1
16
17
18
19
20
21
In Scheme 4, 3-(3-(benzyloxy)-4-methoxybenzamido)-4-fluorobenzoic acid was treated
with pentamethylated epigallocatechin 1 to produce compound 45. Hydrogenation of
9

1
1
1
1
1
1
1
22
23
24
25
26
27
28
compound 45 with Pd/C and H afforded compound 46, which subsequently reacted with 2-
2
iodoethan-1-ol in DMF to provide compound 47. 3-(3-(2-bromoethoxy)-4-
methoxybenzamido)-4-fluorobenzoic acid was coupled with pentamethylated epigallocatechin
1 to provide the key intermediate 48. Compound 48 was then dissolved in 1-methylpiperazine,
morpholine, or piperidine and stirred at room temperature to afford basic ring-containing
compounds 49, 50, and 51, respectively.
Scheme 4 Synthetic route of compounds 45-51
1
1
29
30
10

1
31
2.2 Biological evaluation
1
1
32
33
2.2.1 Structure-activity relationship study of the P-gp modulating activity of methylated
EGC, methylated GC, methylated EC and methylated C derivatives
1
1
1
1
1
1
1
1
1
34
35
36
37
38
39
40
41
42
In this study, we have used a P-gp-transfected breast cancer cell line
(MDA435/LCC6MDR) to study the structure activity relationship of how catechins modulate
the P-gp. MDA435/LCC6MDR cells were about 95.3-, 38.6-, 43.8- and 107.3-fold more
resistant to paclitaxel (PTX), DOX, vinblastine and vincristine than the non-transfected
parental cells (MDA435/LCC6) (Table 1 and Table S1). P-gp modulating activity was
measured using a parameter known as relative fold (RF) which is defined as the ratio of IC50
towards PTX in MDA435/LCC6MDR cells without 1 µM of modulator relative to that with
modulator. Higher RF means higher P-gp modulating activity. A known P-gp modulator
verapamil showed weak P-gp modulating activity with RF = 4.0.
1
1
1
1
43
44
45
46
A total of 39 methylated EGC, methylated GC, methylated EC and methylated C
derivatives were synthesized for studying their P-gp modulating activities (Table 1). These
new derivatives differ from each other at (1) stereoselectivity at C2 and C3 position of ring C;
(2) linker length and rigidity between C3 and ring D or (3) substitutions at ring D.
1
1
1
1
1
47
48
49
50
51
Natural (-)-EGCG (at either 1 or 10 µM) did not show any significant P-gp-modulating
activity with RF = 1.2 (Table 1). Peracetylation of EGCG at all OH groups in rings A, B and
D did not improve either (RF = 0.9) (Table 1). In contrast, permethylation of EGCG, resulted
in a significant improvement with RF =7.3 (Table 1). These data suggest that O-methylation
in rings A, B and D is crucial.
11

1
1
1
1
1
52
53
54
55
56
We first investigated if ring D is needed for the P-gp modulating activity. We have
compared a pair of catechins with or without ring D in all 6 series (I to VI) and found that
addition of ring D can significantly increase the P-gp modulating activity in all 6 series by 2.8-
fold (in series II) to 17.8-fold (series IV) (Table 2). This result is consistent with our previous
observation that ring D was important for the P-gp modulating activity in EGC and GC.[32]
1
1
57
58
2.2.1.1 Effect of linker length between C3 and ring D and stereochemistry at C2 and C3 on P-
gp modulating activity
1
1
1
1
1
1
1
1
1
1
1
1
1
1
59
60
61
62
63
64
65
66
67
68
69
70
71
72
Catechin is composed of rings A and B and a dihydropyran heterocycle C ring in between.
C2 and C3 in ring C contain chiral centers. Ring D is attached to C3 of ring C. The importance
of these 2 chiral centers in P-gp modulating activity has not been studied before. Here we have
attached a trimethoxylated ring D to C3 of ring C with all 4 possible stereoisomers of EGC
(2R, 3R in series I and 2S, 3S in series III) and GC (2R, 3S in series II and 2S, 3R in series IV).
The length of the linker between rings D and C was varied from 1 atom to 8 atoms generating
21 compounds (Table 3). The linker lengths are as follows: oxycarbonyl (with 1 atom) <
oxycarbonylvinyl (with 3 atoms) < oxycarbonylphenylcarbamoyl (with 6 atoms) <
oxycarbonylphenylcarbamoylvinyl (with 8 atoms). Their P-gp modulating activities are
summarized in Table 3. In general, the P-gp modulating activity increased with linker length,
up till the linker has reached 6 atoms. Afterwards, the activity would drop when the linker
reached 8 atoms in length. This can be observed in all 6 series. In series I: permethyl EGCG,
2, 4 and 5 with RF = 7.3, 41.2, 46.2 and 24.6. In series II: 7, 8, 10 and 11 with RF = 3.1, 13.1,
56.5 and 23.1. In series III: 16, 18 and 19 with RF = 14.4, 4.1 and 38.1. In series IV: 12, 14 and
12

1
1
1
1
1
73
74
75
76
77
15 with RF = 23.1, 29.9, 56.5. In series V: 29, 30 and 31 with RF = 12.8, 12.6, 69.3. In series
VI: 21, 23, 25 and 27 with RF = 10.3, 12.8, 84.7 and 58.7. Overall, the
oxycarbonylphenylcarbamyol linker with 6 atoms was the optimal length, yielding the most
potent methylated catechin derivatives (EGC, GC, EC and C) as P-gp modulators, irrespective
of their stereochemistry at C2 and C3 position.
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
To study the importance of stereochemistry at C2 and C3 position, we compared the 4
stereoisomers namely series I (2R, 3R-EGC), series II (2R, 3S-GC), series III (2S, 3S-EGC)
and series IV (2S, 3R-GC) and the 2 stereoisomers of series V (2R, 3R-EC) and series VI (2R,
3S-C) (Table 3). In general, we observed that stereochemistry at C2 and C3 position was
important in those weaker modulators with short oxycarbonyl or medium length
oxycarbonylvinyl linkers, but not in those with long linkers of oxycarbonylphenylcarbamoyl.
When short linker (oxycarbonyl, 1 atom) was used, series IV GC (2S, 3R) displayed the highest
P-gp modulating activity (RF=23.1). But when medium linker length (oxycarbonylvinyl, 3
atoms) was used, series I EGC (2R, 3R) exhibited the highest activity (RF=41.2) (Table 3).
When longest linker (oxycarbonylphenylcarbamoyl, 6 atoms) was used, all stereoisomers have
similar activity (RF = 38.1 to 56.5) (Table 3). These results suggested that the stereochemistry
at C2 and C3 position only matters when shorter linker was used. When longer linker was used,
the P-gp modulating activity of all stereoisomers were all highly potent and stereochemistry at
C2 and C3 position plays a lesser role. In series V and VI, trans-(2R, 3R)-EC and cis-(2R, 3S)-
C derivatives exhibited similar activity no matter what linker length was conjugated at C3 and
ring D (Table 3).
13

1
94
2.2.1.2 Effect of linker rigidity on P-gp modulating activity
1
1
1
1
1
2
2
2
2
2
95
96
97
98
99
00
01
02
03
04
To study the effect of linker rigidity on P-gp modulation, we designed stereoisomers
with various linker flexibility. All of them have the same optimal linker length of 6 atoms.
Three levels of linker rigidity were studied: oxycarbonylphenylcarbamyol > N-acyl-piperidine-
4-carboxylate > N-alkyl-piperidine-4-carboxylate (Table 4). The CO-NH- (amide bond) in
oxycarbonylphenylcarbamyol or N-acyl-piperidine-4-carboxylate linker is conformationally
rigid whereas N-alkyl can freely rotate. In addition, the planar phenyl ring is more constrained
than the saturated piperidine ring in the linker. It was found that the strongest linker rigidity
(oxycarbonylphenylcarbamoyl) yielded the highest P-gp modulating activity (RF =46.2 to
84.7) in series I, V and VI (Table 4). The most flexible linker (N-alkyl-piperidine-carboxylate)
caused the lowest activity with (RF=9.1 to 15.7) in series I, V and VI (Table 4).
2
2
2
05
06
07
Moreover, we found that N-atom at either para or meta positions in piperidine ring of
flexible linker had no effect on P-gp modulation, giving similar RF values such as 39 and 40;
41 and 42; and 43 and 44 (Table 1).
2
08
2.2.1.3 Effect of substitutions at phenyl ring D on P-gp modulating activity
2
2
2
2
2
2
09
10
11
12
13
14
Next, we determined if mono-, di- or trimethoxylation on ring D is preferred. Seven groups
of compounds with different number of methoxy group on ring D were compared (Table 5).
They had the same chiral configuration at C2 and C3 positions, linker length/rigidity and
number of methoxy groups on ring B (Table 5). The number of methoxy substituent at ring D
had no essential influence on P-gp modulating activity in those weaker modulators such as 36,
37, 38 (RF = 6.7 to 10.3 in series I); 33, 34, 35 (RF = 9.1 to 13.3 in series I) or 22, 23 (RF =
14

2
2
2
2
2
2
2
15
16
17
18
19
20
21
12.2 to 12.8 in series VI) (Table 5). For potent modulators with cis-configuration, 2R, 3R-EGC
3 (RF = 50.8) and 4 (RF = 46.2) in series I either with di or tri-methoxylation on ring D yielded
similar activity. In contrast, other potent modulators with trans-configuration in series II and
VI, trimethoxylated-substituted phenyl ring D displayed higher P-gp inhibitory activity than
dimethoxylated-substituted ring D when comparing 2R, 3S-GC 9 (RF = 33.2) and 10 (RF =
56.5); 2R, 3S-C 24 (RF = 32.4) and 25 (RF = 84.7); or 2R, 3S-C 26 (RF = 29.3) and 27 (RF =
58.7), respectively (Table 5).
2
2
2
2
2
2
2
2
2
22
23
24
25
26
27
28
29
30
Other than methoxy substitution at ring D, we also studied the effect of functional group
size and polarity in series I (Table 1). The 3-methoxy group in compound 3 (RF = 50.8) was
replaced by bulky benzyloxy group (45, RF = 8.9), there was about 5.7-fold reduction,
indicating that smaller substitution is preferred. When replacing by 3-OH group (46, RF =6.2),
8.2 folds of diminishment was noted, suggesting that polar functional group was not preferred.
These data suggests that methoxy group at ring D with smaller size and non-polarity is a good
pharmacophore. When comparing other substituents at C3 position of ring D, 2-bromoethoxy
(48, RF = 1.6) and 2-hydroxyethoxy (47, RF = 15.0), polar hydroxyl group is better than
bromide to improve the P-gp modulating activity (Table 1).
2
2
2
2
2
31
32
33
34
35
Not only the size, we also study the hydrophobicity effect on P-gp modulation.
Different heterocyclic rings were substituted at meta position of ring D and the order of
hydrophobicity is as follows: 2-(piperidin-1-yl)ethoxy (51) > 2-morpholinoethoxy (50) > 2-(4-
methylpiperazin-1-yl)ethoxy) (49) (Table 1). The potency of derivatives was positively
correlated with the hydrophobicity of heterocyclic ring. The piperidine ring (51 with RF = 56.5)
15

2
2
2
2
2
2
2
2
36
37
38
39
40
41
42
43
exhibited the highest hydrophobicity and casued the highest RF values, then morpholine (50
with RF = 37.2) and finally hydrophilic piperazine resulted in the lowest RF value (49 with RF
= 1.4). It is likely believed that more hydrophobic side chain would bind more easily to the
transmembrane domain of P-gp than the hydrophilic side chain and finally result in higher
potency. Nevertheless, active compound 51 with hydrophobic piperidine ring at the ring D also
displayed severe toxicity towards L929 cells (IC50 = 5.0 ± 1.7 µM, Table S2). We did not select
it for further characterization. So far, trimethoxylation at ring D is highly preferred because it
retain the high P-gp inhibitory potency of catechin derivatives and causes no toxic effect.
16

2
2
44
45
Table 1. P-gp-modulating activity of methylated epigallocatechin, methylated gallocatechin,
methylated epicatechin and methylated catechin derivatives.
No. of atom between P-gp modulating activity
Series
Cpds
R group
RF
C3 and ring D
mean IC50 of PTX (nM)
152.5 ± 9.7
LCC6MDR
LCC6
0.1% DMSO
0.1% DMSO
Verapmail
/
/
/
/
/
1.0
95.3
4.0
1.6 ± 0.3
38.0 ± 7.0
/
a
124.1 ± 13.7a
122.6 ± 29.0a
176.1 ± 31.7a
EGCG (1 µM)
/
1
1
1
1.2
1.2
0.9
a
EGCG (10 µM)
/
/
a
peracetyl EGCG
a
155.2 ± 28.1a
21.0 ± 2.8a
3.7 ± 0.9a
3.0 ± 0.6a
3.3 ± 0.6a
6.2 ± 0.7a
11.5 ± 1.2
16.8 ± 3.7
15.2 ± 1.2
14 ± 2.1
1
H
0
1
3
6
6
8
6
6
6
5
5
5
6
5
1.0
7.3
a
permethyl EGCG
a
2
41.2
50.8
46.2
24.6
13.3
9.1
a
3
I
a
4
(2R, 3R) cis-methylated
EGC derivatives
a
5
3
3
3
3
3
3
3
4
3
4
5
6
7
8
9
0
10.0
10.9
10.3
6.7
14.8 ± 1.4
22.9 ± 5.0
44.7 ± 4.2
49.1 ± 6.2
3.4
3.1
2
46
17
2
47

2
48
Table 1………. to be continued.
No. of atom between P-gp modulating activity
Series
Cpds
R group
RF
C3 and ring D
mean IC50 of PTX (nM)
4
4
5
6
6
17.1 ± 5.0
8.9
6
6
24.7 ± 3.0
10.2 ± 2.5
6.2
I
4
7
15.0
(2R, 3R) cis-methylated
EGC derivatives
4
4
5
5
8
9
0
6
6
6
93.7 ± 8.6
107.7 ± 15.3
4.1 ± 0.9
1.6
1.4
37.2
1
6
0
1
3
2.7 ± 0.7
135.9 ± 17.9a
49.0 ± 30.5a
11.6 ± 0.7a
56.5
1.1
a
6
7
8
H
a
a
3.1
II
13.1
(2R,3S) trans-methylated
GC derivatives
a
4.6 ± 0.5a
2.7 ± 0.6a
4.2 ± 0.7a
9
6
6
8
33.2
56.5
36.3
a
1
1
0
1
a
2
2
49
50
18

2
51
Table 1………. to be continued.
No. of atom between P-gp modulating activity
Series
Cpds
R group
RF
C3 and ring D
mean IC50 of PTX (nM)
1
1
1
7
6
8
H
0
154.2 ± 15.7
1.0
III
1
3
10.6 ± 1.4
36.9 ± 5.8
14.4
4.1
(
2S, 3S) cis-methylated
EGC derivatives
1
1
9
3
6
0
4.0 ± 0.3
116 ± 9.1
38.1
1.3
H
IV
1
1
2
4
1
3
6.6 ± 1.1
5.1 ± 0.7
23.1
29.9
(
2S, 3R) trans-methylated
GC derivatives
1
5
6
2.7 ± 0.4
56.5
2
2
3
3
4
4
8
9
0
1
1
2
H
0
1
3
6
6
5
120 ± 4.5
11.9 ± 1.5
12.1 ± 0.9
2.2 ± 0.1
16.6 ± 4
1.3
12.8
12.6
69.3
9.2
V
(
2R,3R) cis-methylated
EC derivatives
18.6 ± 2.5
8.2
2
52
19

2
53
Table 1………. to be continued.
No. of atom between P-gp modulating activity
Series
Cpds
R group
RF
C3 and ring D
mean IC50 of PTX (nM)
2
2
2
2
2
2
2
2
4
4
0
1
2
3
4
5
6
7
3
4
H
0
117.5 ± 9.4
1.3
1
3
3
6
6
8
8
6
5
14.8 ± 1.7
12.5 ± 0.4
11.9 ± 1.8
4.7 ± 1.1
1.8 ± 0.2
5.2 ± 0.6
2.6 ± 0.1
9.7 ± 2.3
8.4 ± 2.2
10.3
12.2
12.8
32.4
84.7
29.3
58.7
15.7
18.2
VI
(
2R,3S) trans-methylated
C derivatives
2
2
54
55
Methylated EGC, GC, EC and C derivatives are divided into six series with their R group indicated in
the Table. P-gp modulating activity was measured by determining IC50 towards PTX in P-gp
overexpressing LCC6MDR cells in the absence or presence of 1.0 µM of modulator. Relative Fold (RF)
reflects P-gp modulating activity and is calculated as [IC50 of PTX without modulator / IC50 with 1.0
µM modulator]. All modulators were dissolved in DMSO and used at 1 µM concentration. Each
experiment was repeated three times independently and average RF is presented. The IC50 presented as
mean  standard error of mean. IC50 values of these compounds had been published [32] and included
here for comparison. Compound 1 was named as [(8) in J Med Chem, 2015, 58, 4529−4549], 2(23),
2
2
2
2
2
2
2
56
57
58
59
60
61
62
a
20

2
2
2
63
64
65
3(35), 4(36), 5(31), 6(44), 7(43), 8(49), 9(50), 10(51) and 11(53).[32] LCC6 and LCC6MDR cells
incubated with 0.1% DMSO were solvent control. The chemical structures of compounds are shown in
Supporting Information.
21

2
66
Table 2. Effect of ring D on P-gp modulating activity methylated EGC, GC, EC and C derivatives.
Trans /Cis
Configuration
No. of methoxy No. of methoxy
group in ring B group in ring D
Series
Derivatives
Linker used
Cpds
RF
Series I
Series I
2R, 3R
2R, 3R
2R, 3S
2R, 3S
2S, 3S
2S, 3S
2S, 3R
2S, 3R
EGC
EGC
GC
3
3
3
3
3
3
3
3
2
2
2
2
No ring D
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
oxycarbonyl
1
1.0
7.3
3
permethyl EGCG
Series II
Series II
Series III
Series III
Series IV
Series IV
No ring D
6
1.1
GC
3
7
3.1
EGC
EGC
GC
No ring D
17
16
13
12
28
29
20
21
1.0
3
14.4
1.3
No ring D
GC
3
23.1
1.3
2
2
2
2
R, 3R
R, 3R
R, 3S
R, 3S
No ring D
Series V
Series V
Series VI
Series VI
EC
EC
C
3
No ring D
3
12.8
1.3
10.3
C
2
2
67
68
For easy analysis of effect of ring D on P-gp modulating activity of derivatives, the RF values of respective compounds were extracted from Table 1.
2
69
22

2
70
Table 3. Effect of linker length and stereochemistry on P-gp modulating activity of methylated EGC, GC, EC and C derivatives.
Different linker length between C3 and ring D
Series
Trans /Cis
Configuration
Derivatives No. of methoxy No. of methoxy
Oxycarbonyl
(1 atom)
Oxycarbonylvinyl
(3 atoms)
Oxycarbonyl-
phenylcarbamoyl
Oxycarbonyl-
RF phenylcarbamoylvinyl RF
(8 atoms)
group in ring B
group in ring D
RF
RF
(6 atoms)
Series I
Series II
Series III
Series IV
2R, 3R
2R, 3S
2S, 3S
2S, 3R
3
3
3
3
3
3
3
3
Permethyl EGCG
7.3
3.1
2
8
41.2
13.1
4.1
4
46.2
56.5
38.1
56.5
5
24.6
EGC
GC
7
10
19
15
11
23.1
16
12
14.4
23.1
18
14
/
/
EGC
GC
/
/
29.9
Series V
Series VI
2R, 3R
2R, 3S
2
2
3
3
29
21
12.8
10.3
30
23
12.6
12.8
31
25
69.3
84.7
/
EC
C
/
27
58.7
2
71
2
2
72
73
For easy analysis of effect of linker length and stereochemistry on P-gp modulating activity of derivatives, the RF values of respective compounds were
extracted from Table 1. /: not determined.
23

2
74
Table 4. Effect of linker rigidity on P-gp modulating activity of methylated EGC, GC, EC and C derivatives.
Different linker rigidity
Series
Trans /Cis
Configuration
Derivatives No. of methoxy No. of methoxy
***
Oxycarbonylphenyl
carbamoyl (6 stoms)
**
*
in ring B
in ring D
N-acyl-piperidine-
4-carboxylate (6 atoms)
N-alkyl-piperidine-
4-carboxylate (6 atoms)
RF
RF
RF
Series I
Series I
2R, 3R
2R, 3R
2R, 3R
2S, 3R
EGC
EGC
EC
3
3
2
2
2
3
3
3
3
4
50.8
46.2
69.3
84.7
/
39
/
/
3.4
/
34
35
41
43
9.1
10.0
9.2
Series V
Series VI
31
25
C
/
/
15.7
2
75
2
2
76
77
For easy analysis of effect of linker rigidity on P-gp modulating activity of derivatives, the RF values of respective compounds were extracted from Table
1. ***: strong linker rigidity, ** medium level of linker rigidity and * weak linker rigidity. /: not determined.
24

2
78
Table 5. Effect of methoxy substitution at ring D on P-gp modulating activity of methylated EGC, GC, EC and C derivatives.
Tans/Cis
Configuration
Series
Derivatives
Linker used
No. of methoxy in ring B No. of methoxy in ring D
Cpds
RF
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
R, 3R
R, 3R
R, 3R
R, 3R
R, 3R
R, 3R
R, 3R
R, 3R
R, 3S
R, 3S
R, 3S
R, 3S
R, 3S
R, 3S
R, 3S
R, 3S
N -alkyl-piperidine-4-carboxylate (5 atoms)
N -alkyl-piperidine-4-carboxylate (5 atoms)
N -alkyl-piperidine-4-carboxylate (5 atoms)
N -alkyl-piperidine-4-carboxylate (6 atoms)
N -alkyl-piperidine-4-carboxylate (6 atoms)
N -alkyl-piperidine-4-carboxylate (6 atoms)
Oxycarbonylphenylcarbamoyl (6 atoms)
Oxycarbonylphenylcarbamoyl (6 atoms)
Oxycarbonylphenylcarbamoyl (6 atoms)
Oxycarbonylphenylcarbamoyl (6 atoms)
Oxycarbonylvinyl (3 atoms)
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
2
1
2
3
1
2
3
2
3
2
3
2
3
2
3
2
3
36
37
38
33
34
35
3
10.3
10.3
6.7
Series I
Series I
EGC
EGC
EGC
EGC
EGC
EGC
EGC
EGC
GC
GC
C
Series I
13.3
9.1
Series I
Series I
10.0
50.8
46.2
33.2
56.5
12.2
12.8
32.4
84.7
29.3
58.7
Series I
Series I
4
Series I
9
Series II
Series II
Series VI
Series VI
Series VI
Series VI
Series VI
Series VI
10
22
23
24
25
26
27
Oxycarbonylvinyl (3 atoms)
C
Oxycarbonylphenylcarbamoyl (6 atoms)
Oxycarbonylphenylcarbamoyl (6 atoms)
C
C
C
Oxycarbonylphenylcarbamoylvinyl (with 8 atoms)
Oxycarbonylphenylcarbamoylvinyl (with 8 atoms)
C
2
79
2
2
80
81
For easy analysis of effect of methoxylation at ring D on P-gp modulating activity of derivatives, the RF values of respective compounds were extracted
from Table 1.
25

2
2
82
83
2.2.2 EC50 and selective index values of methylated GC, C and EC derivatives for reversing multidrug
resistance in LCC6MDR
2
2
2
2
2
2
2
84
85
86
87
88
89
90
Four potent compounds with RF > 50 were chosen for further characterization in terms of their
effective concentration (EC50) in reversing P-gp mediated drug resistance and their selective index
(Table 5) including (2S, 3R)-trans-methylated GC 15, (2R, 3R)-cis-methylated EC 31, and (2R, 3S)-
trans-methylated C 25, 27. It is desirable for modulators to affect only LCC6MDR, but not the normal
cells. Selective index may be used as a safety indicator of a new compound. We therefore determined
the selective index of modulators by dividing IC50 of modulators in L929 by the EC50 of modulator for
reversing drug resistance in LCC6MDR cells.
2
2
2
2
2
2
2
2
2
91
92
93
94
95
96
97
98
99
EC50 values for reversing P-gp mediated resistance towards PTX, vinblastine, vincristine and
DOX resistance in LCC6MDR cells ranged from 32 to 178 nM (Table 5). Their selective indices ranged
from > 563 to > 1112 which was higher than verapamil (selective index = 200). (2R, 3S)-trans-
methylated C 25 and (2R, 3R)-cis-methylated EC 31 with tri-methoxy substituents at ring D and
oxycarbonylphenylcarbamoyl linker between ring D and C3 position were the most potent with EC50
ranging from 32 nM to 93 nM. Cyclosporin A showed moderate cytotoxicity towards L929 cell, but our
compounds did not. After considering the toxicity itself, the selective indices of 25 (> 1112) and 31
(>1078) are highly comparable to cyclosporine A with the selective index of 934. Overall, our
modulators are non-toxic and effective P-gp modulators.
26

3
3
00
01
Table 6. EC50 of potent methylated GC, methylated EC and methylated C derivatives for reversing
multidrug resistance in LCC6MDR cells.
Selective index
Mean EC50 (nM) for reversing drug resistance using LCC6MDR cells
Cpds
L929 (IC , µM)
50
(
relative to EC50 of PTX)
PTX
DOX
ND
31.8 ± 10.9
ND
Vinblastine
Vincristine
15
25
27
31
>100
>100
>100
>100
>741
>1112
>563
135.0 ± 5.0
89.9 ± 3.5
177.5 ± 2.5
92.8 ± 5.4
ND
60.0 ± 15.1
ND
ND
66.0 ± 4.0
ND
>1078
37.3 ± 4.3
60.7 ± 5.5
77.7 ± 6.7
a
a
a
a
Verapamil
89.2±8.2
29.9±5.7
200
445.7 ± 40.7
254.4 ± 22.9
ND
502.5 ± 91.7
ND
385.0 ± 35.1
ND
a
a
Cyclosporin A
934
32.0 ± 1.0
3
02
3
3
3
3
03
04
05
06
EC50 values were presented as mean  standard error of mean. N= 3 - 8 independent experiments.
Selective index value = (IC50 of modulators towards L929 fibroblasts) / (EC50 of modulators for
a
reversing PTX resistance in LCC6MDR cells). ND = not determined. the IC50 values, EC50 values and
selective index values of verapamil and cyclosporin A had been published.[32]
27

3
3
07
08
2.2.3 Effect of number of methoxy group at rings B and D on P-gp modulating activity of EGC and
EC as well as GC and C derivatives
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
EGC 3, 4 and EC 31 (2R, 3R-configuration) as well as GC 9, 10 and C 25 (2R, 3S-configuration)
derivatives were structurally similar (Table 7). All of them possessed the optimal
oxycarbonylphenylcarbamyol linker and O-methylated A, B and D rings. Surprisingly, there was a
correlation between the EC50 for reversing PTX resistance and number of methoxy group at B and D
rings (Table 7): EC 31 (EC50 = 93 nM with 2 methoxy groups at B ring + 3 methoxy groups at D ring)
< EGC 3 (EC50 = 159 nM with 3 methoxy groups at B ring + 2 methoxy groups at D ring) < EGC 4
(EC50 = 214 nM with 3 methoxy groups at B ring + 3 methoxy at D ring). EC 31 was about 1.7- to 2.3-
fold more potent than EGC 3 and 4. Similarly, C 25 (EC50 = 90 nM with 2 methoxy groups at B ring +
3 methoxy groups at D ring) < GC 10 (EC50 = 140 nM with 3 methoxy groups at B ring + 3 methoxy
group at D ring) < GC 9 (EC50 = 171 nM with 3 methoxy groups at B ring + 2 methoxy groups at D
ring). C 25 displayed about 1.6- to 1.9-fold higher potency than GC 9 and 10, respectively. It suggests
that the number of methoxy group at B and D rings can affect the P-gp modulating activity. Derivatives
EC 31 and C 25 containing dimethoxylated B ring and trimethoxylated D ring gave the highest P-gp
inhibitory activity as compared to other EGC and GC 3, 4, 9 and 10 which had trimethoxylation at ring
B and either di- or tri-methoxylation at ring D. It suggests that dimethoxylation at ring B is an important
pharmacophore of catechins for strong P-gp modulation.
28