Rational drug design of 6-substituted 4-anilino-2-phenylpyrimidines for exploration of novel ABCG2 binding site

Article abstract of DOI:10.1016/j.ejmech.2020.113045

In the search for novel, highly potent, and nontoxic adjuvant chemotherapeutics to resolve the major issue of ABC transporter-mediated multidrug resistance (MDR), pyrimidines were discovered as a promising compound class of modern ABCG2 inhibitors. As ABC

Full text of DOI:10.1016/j.ejmech.2020.113045

European Journal of Medicinal Chemistry 212 (2021) 113045
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Rational drug design of 6-substituted 4-anilino-2-phenylpyrimidines
for exploration of novel ABCG2 binding site
2
2
2
**, 1
Katja Silbermann , Jiyang Li , Vigneshwaran Namasivayam , Sven Marcel Stefan
Michael Wiese
,
*
Pharmaceutical and Cellbiological Chemistry, Pharmaceutical Institute, Rheinische Friedrich-Wilhelms-University of Bonn, An der Immenburg 4, 53121,
Bonn, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 September 2020
Received in revised form
In the search for novel, highly potent, and nontoxic adjuvant chemotherapeutics to resolve the major
issue of ABC transporter-mediated multidrug resistance (MDR), pyrimidines were discovered as a
promising compound class of modern ABCG2 inhibitors. As ABCG2-mediated MDR is a major obstacle in
leukemia, pancreatic carcinoma, and breast cancer chemotherapy, adjuvant chemotherapeutics are
highly desired for future clinical oncology. Very recently, docking studies of one of the most potent
reversers of ABCG2-mediated MDR were reported and revealed a putative second binding pocket of
ABCG2. Based on this (sub)pocket, a series of 16 differently 6-substituted 4-anilino-2-phenylpyrimidines
was designed and synthesized to explore the potential increase in inhibitory activity of these ABCG2
inhibitors. The compounds were assessed for their influence on the ABCG2-mediated pheophorbide A
transport, as well as the ABCB1- and ABCC1-mediated transport of calcein AM. They were additionally
evaluated in MDR reversal assays to determine their half-maximal reversal concentration (EC50). The 6-
substitution did not only show increased toxicity against ABCG2-overexpressing cells in combination
with SN-38 but also a negative influence on cell viability in general. Nevertheless, several candidates had
EC50 values in the low double-digit nanomolar concentration range, qualifying them as some of the most
potent reversers of ABCG2-mediated MDR. In addition, five novel multitarget ABCB1, ABCC1, and ABCG2
inhibitors were discovered, four of them exerting their inhibitory power against the three stated
transporters at least in the single-digit micromolar concentration range.
18 November 2020
Accepted 21 November 2020
Available online 3 December 2020
Keywords:
ABCG2 (BCRP)
ABCB1 (P-gp)
ABCC1 (MRP1)
Multidrug resistance (MDR)
Dual inhibition
Promiscuous inhibitor
Multitarget antagonist
Broad-spectrum antagonism
©
2020 Elsevier Masson SAS. All rights reserved.
1. Introduction
The ABC transport protein breast cancer resistance protein
Abbreviations: A2780/ADR, adriamycin-(doxorubicin)-selected A2780 cells;
ABC, ATP-binding cassette; ATP, adenosine-triphosphate; BCRP, breast cancer
resistance protein; calcein AM, calcein acetoxymethyl ester; CsA, cyclosporine A;
cryo-EM, cryogenic electron microscopy; EC50, half-maximal reversal concentra-
tion; GFP, green fluorescent protein; GI50, half-maximal growth inhibition con-
centration; Imax, maximal inhibition level; MDCK, madin-darby canine kidney; MTT,
(ABCG2, BCRP) is recognized as a negative marker in (relapsed/re-
fractory) cancer in general. It is associated with decreased pro-
gression- or disease-free survival and increased mortality [1e4], as
well as the development of metastases [5]. Its expression was
determined in several cancers, such as leukemia [3,4,6], pancreatic
carcinoma [2,6], or breast cancer [6]. Like ABCB1 (P-glycoprotein, P-
gp) and ABCC1 (multidrug resistance-associated protein 1, MRP1),
ABCG2 recognizes applied antineoplastic agents inside the cancer
cell as well as in the cell membrane and extrudes these into the
exterior at the expense of the energy source ATP [7]. This results in
the resistance of the respective cancer cells to chemotherapeutic
drugs. Moreover, resistance to other first- and second-line anti-
neoplastic agents is observed that were not used a priori in anti-
cancer regimens. This so-called cross-resistance is concomitant
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; MDR, multidrug
resistance; MRP1, multidrug resistance-associated protein 1; P-gp, P-glycoprotein;
ꢀ
s, degree of sensitization; SEM, standard error of the mean; sr, selectivity ratio; TKI,
tyrosine kinase inhibitor; tr, therapeutic ratio; vmax, maximal transport velocity.
*
Corresponding author.
** Corresponding author.
E-mail addresses: s.m.stefan@medisin.uio.no (S.M. Stefan), mwiese@uni-bonn.
de (M. Wiese).
1
Present author address: Sven Marcel Stefan, Translational Neurodegeneration
Research and Neuropathology Lab, Department of Neuro-/Pathology, University of
Oslo, Oslo, Norway.
2
These three authors contributed equally to this work.
https://doi.org/10.1016/j.ejmech.2020.113045
0223-5234/© 2020 Elsevier Masson SAS. All rights reserved.

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
with the overexpression of ABC transport proteins, and the asso-
ciated phenomenon was designated as ‘multidrug resistance’
In case of both compound classes quinazolines and pyrimidines,
the inhibition type could be determined as competitive with
respect to the ABCG2-mediated transport of Hoechst 33342
[17,26,37], and our studies with structure-based drug design ap-
proaches proposed the binding behavior of both scaffolds [17,26].
Interestingly, a binding (sub)pocket has opened up in the case of
compound 8 [17], which forms the basis for the current study. In
order to evaluate the (sub)pocket we designed and synthesized
compounds to enhance their potency and selectivity against ABCG2
as well as their capability to reverse ABCG2-mediated resistance
against SN-38.
(
MDR). With respect to ABCG2, cross-resistance against several
antineoplastic agents of structural diversity has been described,
such as camptothecins, epipodophyllotoxins, mitoxantrone, or
tyrosine kinase inhibitors (TKIs) [6e8].
The latter compound class comprises of compounds with com-
mon structural features, such as quinazoline (e.g., canertinib [9], 1),
quinoline (e.g., pelitinib [10], 2), pyrimidine (e.g., ceritinib [11], 3),
or pyridine (e.g., apatinib [12], 4). TKIs have a special standing with
regard to ABCG2 inhibition, which becomes clear when focusing on
the stated examples, which are depicted in Fig. 1: In 2001, can-
ertinib was the very first TKI proven to interact with ABCG2 in an
inhibitory way, which made it one of the first ABCG2 inhibitors [9];
pelitinib is one of only around 75 multitarget reversers of ABCB1-,
ABCC1-, and ABCG2-mediated MDR known until today [10,13e24];
ceritinib is the most potent TKI-related (selective) reverser of
ABCG2-mediated MDR [11]; and apatinib is a very rare example of
an TKI that made it into the clinic in combination with another
antineoplastic agent [25].
Starting from the above stated compounds, our workgroup was
able to develop several new potent inhibitors of ABC transporters in
general, and of ABCG2 in particular. Until today, compound 5 with
its quinazoline scaffold is one of the most potent inhibitors of
ABCG2 [26]; Compound 6 with its quinoline scaffold is one of the 20
most potent multitarget ABCB1, ABCC1, and ABCG2 inhibitors
2. Results and discussion
2.1. ABCG2ecompound 8 complex analysis
Recently, on the basis of the cryo-electron microscopy (cryo-
EM) structure of human ABCG2 and docking studies [38,39], the
putative binding mode of Hoechst 33342 was reported [26]. The
studies were extended to explore the binding mode of the potent
quinazoline inhibitors of ABCG2 and to explain the mechanism of
inhibition [17,26]. Very recently, pyrimidine based ABCG2 in-
hibitors showed excellent inhibitory potency and antiproliferative
efficacy [17]. The docking studies explained the binding mode and
their interaction with the amino acid residues in the pocket, which
was in agreement with the competitive mode of inhibition against
the substrate Hoechst 33342 (Fig. 3A and B). The pyrimidine scaf-
fold formed interactions with Asn387 and Asn391, the amino linker
formed an interaction with Glu451, the substituted phenyl ring was
positioned in the hydrophobic (sub)pocket, and the cyano substit-
uent formed strong electrostatic interaction with Arg191 (Fig. 3C).
The binding pose of the potent pyrimidine scaffold 8 suggested a
[
11,15,27e36]; the consequent downsizing of the main scaffold to a
pyrimidine led to the discovery of compound 7 as a novel ABCG2
inhibitor [37]; and subsequent development of small-molecule
pyrimidines led to one of the most potent reversers of ABCG2-
mediated MDR, compound 8 [17]. Fig. 2 shows the stated com-
pounds 5e8.
Fig. 1. 2D representation of famous TKIs reported in the context of ABCG2 inhibition. (A) The very first TKI associated with ABCG2 inhibition, canertinib (1) [9]; (B) one of only
around 75 known (multitarget) reversers of ABCB1-, ABCC1-, and ABCG2-mediated MDR, pelitinib (2) [10]; (C) the most potent TKI-related selective reverser of ABCG2-mediated
MDR, ceritinib (3) [11]; one of the rare TKIs used in combination with another antineoplastic agent (and ABCG2 substrate), apatinib, in an anticancer regimen (4) [25].
2

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Fig. 2. Depiction of recent advances in the development of ABCG2 inhibitors. (A) One of the most potent ABCG2 inhibitors, quinazoline 41 (5; IC50 (ABCG2) ¼ 23.4 nM) [26]; (B) one of
M) [27]. (C) one of the first
the 20 most potent multitarget ABCB1, ABCC1, and ABCG2 inhibitors, quinoline 29 (6; IC50 (ABCB1) ¼ 1.42
m
M; IC50 (ABCC1) ¼ 2.98
m
M; IC50 (ABCG2) ¼ 1.90
m
reported ABCG2 inhibitors with pyrimidine core structure, pyrimidine 50 (7; IC50 (ABCG2) ¼ 71.4 nM) [37]; and one of the most potent reversers of ABCG2-mediated MDR as reported
recently, pyrimidine 19 (8; EC50 (ABCG2) ¼ 11.0 nM) [17].
(
sub)pocket lined by Gln18, Phe182, and Glu446 closer to the
standard ABCG2 inhibitor Ko143 [(3S,6S,12aS)-1,2,3,4,6,7,12,12a-
0
0
methyl substitution at position 6 (Fig. 3C). In this work, the (sub)
pocket was explored with different substitutions including alkyl,
aromatic, and substituted aromatic side chains at position 6 of the
pyrimidine core structure.
octahydro-9-methoxy-6-(2-methylpropyl)-1,4-dioxopyrazino[1 ,2 :
1,6]pyrido[3,4-b]indole-3-propanoic acid 1,1-dimethylethyl ester;
28] was used as reference to determine the maximum inhibition
level (Imax). Table 1 summarizes the obtained half-maximal inhi-
bition concentrations (IC50) and Imax values. In addition, the cor-
responding IC50 values of compounds 7 and 8 are shown for
comparison as taken from the literature [17,37].
2
.2. Chemistry
On the basis of the docking studies [17,26] we synthesized
compound set (i) with different alkyl residues linked to the py-
rimidine moiety by an amino linker (12e15). Here, the methyl-
The 6-substituted 4-anilino-2-phenylpyrimidine derivatives
were synthesized in four steps. First, benzamidine was condensed
with diethyl malonate in a freshly prepared sodium ethoxide so-
lution to form the pyrimidine scaffold (9). Second, intermediate 9
was refluxed in phosphorous oxychloride to yield the chlorinated
precursor 10. Third, compound 10 was treated with one equivalent
of 4-cyanoaniline to yield the monosubstituted product 11 under
catalyzation with 0.1 equivalent of concentrated hydrochloric acid.
Finally, the desired compounds were synthesized by exposing
precursor 11 with differently substituted amines (12e15), anilines
amino derivative 12 (IC50 ¼ 0.141
m
M) exceeded the inhibitory
M) as well as the
M) and 28 (IC50 ¼ 0.274 M),
activity of the parent compound 7 (IC50 ¼ 0.189
m
reference compounds 8 (IC50 ¼ 0.175
m
m
but with a slight decrease of Imax. Larger residues such as iso-
propylamino (13) or butylamino (15) were less preferred, while on
the other hand a n-propyl substituent (14) resulted in notable
inhibitory activity (IC50 ¼ 0.293
that of compound 7, while Imax equaled that of reference 28
Fig. 4A).
Possibly the aliphatic side chains open up only limited potential
mM), which was slightly inferior to
(
17e27) or phenylboric acid (via Suzuki coupling in the presence of
a palladium catalyst and base; 16). Scheme 1 summarizes the
prevailing reaction steps.
(
for lipophilic interactions with the target and can adopt different
orientations in the binding pocket due to their high flexibility. This
questions whether aromatic and/or aromatic-aliphatic side chains
would be beneficial for inhibitory activity against ABCG2. There-
fore, compound set (ii) was synthesized (16e20). Interestingly, a
phenyl ring without amino linker (16) resulted in great loss of
2
2
.3. Biological investigation
.3.1. Determination of inhibitory activity against ABCG2
The inhibitory activity of the synthesized compounds 12e27
against ABCG2 was investigated in a pheophorbide A assay using
ABCG2-overexpressing MDCK II BCRP cells as already described
earlier [15e17,40,41].
In this assay, blockage of ABCG2-mediated pheophorbide A
transport leads to the enhanced intracellular presence of this
fluorescence dye proportionate to the degree of inhibition. The
inhibitory activity. Its amino analog 17 (IC50 ¼ 0.169
mM) was nearly
as potent as the corresponding methylamino derivative 12, and
slightly superior to the parent compound 7. While the insertion of a
methylene linker (18) resulted in a weaker inhibitory activity, a
propylene linker boosted the inhibitory potency, making the
3

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Fig. 3. Putative binding modes of Hoechst 33342 and compound 8. The docked pose of (A) the ABCG2 substrate Hoechst 33342 (carbon colored orange, represented as spheres) and
B) the ABCG2 inhibitor compound 8 (carbon colored green, represented as spheres) in the binding pocket of human ABCG2 represented in cartoon and colored chain A and B in grey
(
and cyan, respectively. The chemical structures of (A) Hoechst 33342 and (B) compound 8 are provided. (C) The binding pocket of compound 8 with the important amino acid
residues (label colored red) interacting with compound 8 (colored in green) and the residues in the additional (sub)pocket (label colored blue) represented as sticks and colored in
cyan are shown. The interactions of compound 8 with amino acid residues in the binding pocket are colored in red dashed lines. Oxygen atoms are colored in red, nitrogen atoms in
blue, and polar hydrogen atoms of compound 8 are colored in white.
corresponding compound 19 the most potent within this manu-
M; Fig. 4B). This shows that there is a need for
an optimal linker to position the larger aromatic substituent into
the binding pocket.
counterpart 17 as well as reference compounds 7 and 8, while the
3- (22) and 4-pyridylamino (23) derivatives showed less pro-
nounced inhibition. This suggests that the positioning of the ni-
trogen atom is important and it could form an interaction with
Gln181 when it was placed at position 2 of the aromatic ring.
Since hydroxy- and cyano-anilino pyrimidine derivatives
showed a marked inhibitory activity against ABCG2 in our previous
studies [17,37], these substituents were inserted at position 6 to
further evaluate the new (sub)pocket [compound set (iv); 24e27].
Comparable to the correlation given for the pyridine derivatives
21e23, the 2-hydroxyaniline derivative 24 was the most potent
script (IC50 ¼ 0.0960
m
Interestingly, compound 20 with its phenylbutylamino substit-
uent was only marginally weaker than compound 19, which ap-
pears to contradict the finding that a butyl chain is much less
favorable than a propyl chain (14 vs 15). This could be explained by
two things: first, compound 19 found good interaction with the
residues in the binding pocket (possibly Phe182) according to the
chain length and the resulting orientation of the phenyl ring; sec-
ond, the butyl chain indeed hampers interaction, but subsequently
leads to an even better orientation of the phenyl ring, thus, leading
to a compensation of these two effects. However, it should be noted
that the Imax value of both compounds 19 and 20 was rather low
representative of this sub-set (IC50 ¼ 0.223
m
M) followed by its 4-
0.387 M) and 3-
M). Interestingly, the 4-
hydroxyaniline derivatives 26 (IC50
hydroxyaniline analog 25 (IC50 ¼ 0.608
¼
m
m
cyano derivative 27 was found to be one of the most potent com-
pounds within the investigated compounds (IC50 ¼ 0.133). It
exceeded not only the inhibitory activities of the parent and
reference compounds 7, 8, and 28, but also the lead molecules of the
corresponding compound classes, 12 (i), 17 (ii), and 21 (iii). It was
only slightly less potent than compound 19, but had an inferior Imax
compared to the latter. This is possible as the hydroxy substitution
(
61% each).
In the next step, the effect of a nitrogen-containing aromatic
ring system was investigated. Therefore, compound set (iii)
comprising of pyrimidines 21e23 was synthesized. The 2-
pyridylamino derivative 21 had a notable inhibitory activity
(
IC50 ¼ 0.210
mM) against ABCG2 comparable to its phenyl
4

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Scheme 1. General synthesis of 6-substituted 4-anilino-2-phenylpyrimidine derivatives. Reagents and conditions: (a) NaOEt/EtOH, reflux, overnight; (b) POCl
3
, reflux, 8e9 h; (c) i-
ꢀ
ꢀ
PrOH, HCl-catalyzed, reflux, overnight; (d) 1. i-PrOH, microwave-assisted, 110 C, 0.5 h; 2. dioxane, 2 M K CO (3:1), Pd(t-Bu P) -catalyzed, argon, 90 C, 7 h.
2 3 3 2
of compounds 24 and 26 at positions 2 and 4, respectively, and the
-cyano (27) might introduce hydrogen bond interactions with
Glu446 or Gln181, while the hydroxyl substitution at position 3 (25)
While compound set (ii) contained no compound with note-
worthy inhibitory power against ABCC1, compounds 17 and 18
were shown to be dual ABCB1 and ABCG2 inhibitors, but as already
stated for compound 15, both the benzyl and phenethyl derivatives
4
might alter the orientation of the compound in the binding pocket.
1
7 (IC50 (ABCB1) ¼ 5.38
m
M) and 18 (IC50 (ABCB1) ¼ 4.47
and 16 times weaker with regard to ABCB1 inhibition.
The 6-pyridyl pyrimidines 21e23 were found to be dual ABCB1
and ABCG2 inhibitors with a 16 (21; IC50 (ABCB1) ¼ 3.40 M), 15 (22;
M) times
mM) were 32
2.3.2. Determination of selectivity against ABCB1 and ABCC1
The synthesized compounds were evaluated for their inhibitory
m
activity against two other major important targets in ABC
transporter-mediated MDR, namely ABCB1 and ABCC1. For this
purpose, a calcein AM assay was conducted using either ABCB1-
overexpressing A2780/ADR or ABCC1-overexpressing H69AR cells
as described earlier [15e17,40,41]. In short, the highly lipophilic
calcein AM enters the cell by passive diffusion and gets effluxed by
either ABCB1 or ABCC1. In case of ABCB1 or ABCC1 inhibition, the
calcein AM persists inside the cell and gets cleaved by unspecific
intracellular esterases to the fluorescence dye calcein. As
mentioned above for the pheophorbide A assay, the degree of in-
hibition of either ABCB1 or ABCC1 is proportionately correlated to
the amount of measured fluorescence of calcein. The compounds
IC50 (ABCB1) ¼ 5.41
mM), and 10 (23; IC50 (ABCB1) ¼ 4.26
m
higher inhibitory activity against ABCG2, respectively. Indeed,
compound 22 had slight inhibitory power against ABCC1 (IC50
(
ABCC1) ¼ 20.7
mM), which made it a rare example of a triple ABCB1,
ABCC1, and ABCG2 inhibitor, but considering its 56 times lower
affinity toward ABCC1, it is de facto a dual ABCB1 and ABCG2
inhibitor.
Compound set (iv) was quite diverse according to its interaction
profile: While the 2-hydroxyphenylamino derivative 24 was shown
to be a dual ABCB1 and ABCG2 inhibitor with 13 times higher af-
finity to the latter transporter (IC50 (ABCB1) ¼ 2.98
5 (IC50 (ABCB1) ¼ 3.52 M; IC50 (ABCC1) ¼ 4.08
ABCB1) ¼ 4.54 M; IC50 (ABCC1) ¼ 4.04
m
m
M); compounds
M) and 26 (IC50
mM) are special representatives
2
m
have been evaluated at a concentration of 10
sentatives with effect values ꢁ 20% (þSEM) compared to the
standard ABCB1 and ABCC1 inhibitor, cyclosporine A (CsA; 10 M),
mM, and for repre-
(
m
of broad-spectrum ABCB1, ABCC1, and ABCG2 inhibitors in low
single-digit and submicromolar concentration range, belonging to
the 22 most potent of its kind [11,15,27e36]; and the second most
potent ABCG2 inhibitor 27 can be considered ABCG2-selective as its
m
complete concentration-effect curves have been generated and
their IC50 values were calculated. The results of the corresponding
screening can be seen in Fig. 5A (ABCB1) and 5B (ABCC1), while the
associated values are presented in Table 2.
affinity to ABCB1 is over 50 times less (IC50 ¼ 6.75
mM).
In summary, the presented results highlight the most potent
ABCG2 inhibitors 12, 17, 19e20 and 27 as selective, while com-
pounds 13e14 and 25e26 represent limited examples of broad-
spectrum ABCB1, ABCC1, and ABCG2 inhibitors. Compound 24 is a
good representative of a dual ABCB1 and ABCG2 inhibitor.
Compared to the results in Table 1, the methylamino derivative
12 showed only minor inhibitory activity against ABCB1 (IC50
(
ABCB1) ¼ 25.9
m
M) and none against ABCC1, suggesting that this
compound is a selective and highly potent ABCG2 inhibitor. The
inhibitory activities of compounds 13 and 14 revealed both as
moderately potent multitarget ABCB1, ABCC1, and ABCG2 in-
hibitors. The butylamino derivative 15 was discovered as a dual
2.3.3. Determination of intrinsic toxicity in MDCK II BCRP and
MDCK II cells
ABCB1 and ABCG2 inhibitor (IC50 (ABCB1) ¼ 6.62
mM), although it
must be stated that its affinity to ABCG2 was over 13 times higher
The intrinsic cytotoxicity of the cells was evaluated by per-
than to ABCB1.
forming
a
MTT-(3-(4,5-demethyl-2-thiazolyl)-2,5-diphenyl-2H-
5

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Table 1
Inhibitory activity of 6-substituted 4-anilino-2-phenylpyrimidine derivatives ob-
tained in a pheophorbide A assay using ABCG2-overexpressing MDCK II BCRP cells as
described earlier [15e17,40,41]. IC50 and
Imax values are expressed as
mean ± standard error of the mean (SEM) of at least three independent experiments.
No
R
IC50 ± SEM [
m
M]
Imax ± SEM [%]
MDCK II BCRP
MDCK II BCRP
a
7
8
0.189 ± 0.022a
0.175 ± 0.011
87 ± 8
b
76 ± 12b
68 ± 8
101 ± 6
89 ± 8
73 ± 9
100c
84 ± 14
99 ± 11
61 ± 7
61 ± 1
88 ± 4
91 ± 4
78 ± 11
86 ± 10
66 ± 9
76 ± 4
60 ± 3
100
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
methylamino
i-propylamino
n-propylamino
n-butylamino
phenyl
0.141 ± 0.029
0.763 ± 0.088
0.293 ± 0.043
0.500 ± 0.034
4.41 ± 0.13
0.169 ± 0.010
0.283 ± 0.025
0.0960 ± 0.0162
0.147 ± 0.021
0.210 ± 0.004
0.373 ± 0.023
0.435 ± 0.001
0.223 ± 0.013
0.608 ± 0.028
0.387 ± 0.036
0.133 ± 0.014
0.274 ± 0.014
anilino
benzylamino
phenylpropylamino
phenylbutylamino
2-pyridylmethylamino
3-pyridylmethylamino
4-pyridylmethylamino
2-hydroxyanilino
3-hydroxyanilino
4-hydroxyanilino
4-cyanoanilino
a
Compound
7
has been reported before in
a
Hoechst 33342 assay
(
IC50 ¼ 71.4 nM) [37]; the inhibitory potency has been determined in the pheo-
phorbide A assay.
b
Data for compound 8 as reported in the literature [17].
c
Constrained to Imax of compound 28.
Fig. 4. Concentration-effect curves of compounds 14 (A; closed squares;
IC50 ¼ 0.293
m
M) and 19 (B; closed squares; IC50 ¼ 0.0960
mM) in comparison to
M) applying a pheophorbide A
tetrazolium bromide)-based cell viability assay applying the same
cells that were used for the inhibition assays, MDCK II BCRP and
MDCK II, as already described [15e17,41]. Pure culture medium
served as a reference for 100% cell viability and 10% DMSO was
chosen as a positive control (0% cell viability). To determine the
reference inhibitor 28 (A,B; closed circles; IC50 ¼ 0.274
m
accumulation assay. The data are shown as mean ± SEM of at least three independent
experiments.
half-maximal growth inhibition (GI50
)
values, complete
2.3.4. Determination of intrinsic toxicity using A2780/ADR, A2780,
concentration-effect curves have been generated. Table 3 summa-
rizes the obtained results including the therapeutic ratio (tr) using
the inhibition values as reported in Table 1.
In general, the compound class of 6-substituted 4-anilino-2-
phenylpyrimidines seems to impair cell viability as can be visual-
ized from the rather low GI50 values in the mid-single-digit con-
centration range. Considering the very selective and highly potent
representatives 12, 17, 19e21, 24, and 27, only compounds 12, 20,
and 27 had acceptable therapeutic ratios of over 50 (180, 65.0, and
and H69AR cells
As compound 24 was found to be a moderately good dual ABCB1
and ABCG2 inhibitor, and compounds 13e14 as well as 25e26 were
discovered as multitarget ABCB1, ABCC1, and ABCG2 inhibitors,
their intrinsic toxicity has also been assessed using A2780/ADR,
A2780, and H69AR cells applying the MTT-based cell viability assay
as described above [15e17,40,41]. The obtained GI50 values are
presented in Table 4.
The GI50 values of the dual ABCB1 and ABCG2 inhibitor 24 and
triple ABCB1, ABCC1, and ABCG2 inhibitors 13e14 and 25e26 are
rather low and can be compared to the values determined for the
MDCK II BCRP and MDCK II cells (Table 3). Since the inhibition
values of compound 24 against ABCB1 and compounds 13e14 as
well as 25e26 against ABCB1 and ABCC2 were in the mid-single-
digit micromolar concentration range, it could be expected that
their tr is suboptimal.
5
8.7, respectively), which gives these three compounds a special
standing among the presented compounds. Unfortunately, the
most potent ABCG2 inhibitor 19 turned out to be one of the most
toxic compounds within the four sets, not only regarding its tr
(
38.4), but especially due to its absolute GI50 values of 3.64
mM
(
MDCK II BCRP) and 5.03 M (MDCK II), which are amongst the
m
lowest measured values of all. Both extremes, the rather nontoxic
compound 12 and the most potent but most toxic compounds 19
are shown in Fig. 6.
6

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Fig. 5. Screening of 6-substituted 4-anilino-2-phenylpyrimidine sets (i)-(iv) at 10
mM against ABCB1, using the ABCB1-overexpressing cell line A2780/ADR (A), and ABCC1, applying
the ABCC1-overexpressing cell line H69AR (B), conducted with a calcein AM assay as described earlier [15e17,40,41]. CsA (10 mM) has been used as a reference which defined 100%
inhibition, while buffer medium served as a negative control (0%). Shown is mean ± SEM of at least three independent experiments.
Table 2
Inhibitory activity of 6-substituted 4-anilino-2-phenylpyrimidine derivatives obtained in a calcein AM assay using either ABCB1-overexpressing A2780/ADR or ABCC1-
overexpressing H69AR cells as described earlier [15e17,40,41]. IC50 values are expressed as mean ± SEM of at least three independent experiments.
No
R
IC50 ± SEM [
m
M]
IC50 ± SEM [mM]
A2780/ADR
H69AR
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
methylamino
i-propylamino
n-propylamino
n-butylamino
phenyl
25.9 ± 10.2
6.42 ± 0.40
5.90 ± 0.42
6.62 ± 0.69
n. e.a
9.05 ± 1.15
9.58 ± 0.64
a
n. e.
n. e.
n. e.
n. e.
n. e.
n. e.
n. e.
a
a
n. e.
a
anilino
benzylamino
5.38 ± 0.39
4.47 ± 0.31
a
a
a
phenylpropylamino
phenylbutylamino
2-pyridylmethylamino
3-pyridylmethylamino
4-pyridylmethylamino
2-hydroxyanilino
3-hydroxyanilino
4-hydroxyanilino
4-cyanoanilino
n. e.
n. e.
a
a
a
3.40 ± 0.26
5.41 ± 0.46
4.26 ± 0.18
2.98 ± 0.36
3.52 ± 0.45
4.54 ± 0.07
6.75 ± 0.60
1.20 ± 0.03
20.7 ± 3.6
a
n. e.
n. e.
a
4.08 ± 0.11
4.04 ± 0.11
a
n. e.
CsA
2.95 ± 0.19
a
n. e. ¼ no effect.
2
.3.5. Reversal of ABCG2-mediated MDR
While IC50 values give proof of their inhibitory power against
herein reported compounds were evaluated in a MTT-based MDR
reversal assay as reported earlier [15e17,41]. Using a dilution series
of the ABCG2 substrate SN-38, different concentrations of the target
compounds were added. The greater the MDR reversing property of
the compound is, the lower is the concentration needed to shift the
concentration-effect curve of SN-38 toward the left. This specifies
that the amount of SN-38 required to half-maximally inhibit
growth of ABCG2-overexpressing cells decreases. From plotting the
ABCG2 e and therefore of their primary mode of action e half-
maximal sensitization concentrations (EC50) reveal the actual
antiproliferative capability of the compounds. It can be considered
as the most important value regarding the reversal of MDR, but has
only very recently started to be addressed and highlighted properly
[
15e17,41e43]. In order to address this very important aspect, all
7

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Table 3
Intrinsic toxicity of the 6-substituted 4-anilino-2-phenylpyrimidine derivatives obtained in a MTT-based cell viability assay using MDCK II BCRP and MDCK II cells as described
earlier [15e17,41]. GI50 values are expressed as mean ± SEM of at least three independent experiments. The tr has been calculated by dividing the GI50 values of a compound by
the corresponding IC50 value of the very same compound as shown in Table 1.
No
R
GI50 ± SEM [
m
M]
GI50 ± SEM [
m
M]
tr
MDCK II BCRP
MDCK II
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
methylamino
i-propylamino
n-propylamino
n-butylamino
phenyl
25.0 ± 3.7
5.06 ± 0.14
6.56 ± 0.20
6.08 ± 0.21
21.9 ± 1.6
4.05 ± 0.31
10.7 ± 0.8
3.64 ± 0.16
9.48 ± 0.05
8.56 ± 0.51
5.54 ± 0.09
10.0 ± 0.1
4.61 ± 0.34
5.12 ± 0.07
5.15 ± 0.10
7.78 ± 0.36
34.1 ± 4.2
180
5.58 ± 0.11
8.34 ± 0.21
8.32 ± 0.35
25.1 ± 1.2
4.95 ± 0.24
4.19 ± 0.08
5.03 ± 0.11
10.6 ± 0.37
9.44 ± 0.24
7.47 ± 0.39
10.7 ± 0.2
6.68
22.7
12.2
4.96
24.0
38.0
38.4
65.0
40.8
14.9
23.1
20.7
8.42
13.4
58.7
anilino
benzylamino
phenylpropylamino
phenylbutylamino
2-pyridylmethylamino
3-pyridylmethylamino
4-pyridylmethylamino
2-hydroxyanilino
3-hydroxyanilino
4-hydroxyanilino
4-cyanoanilino
4.23 ± 0.13
5.24 ± 0.06
5.30 ± 0.06
7.09 ± 0.29
gives these compounds an exceptional standing, as there have
hardly been reported compounds with equal or superior efficacy
[
17,42].
2.3.6. Reversal of ABCB1- and ABCC1-mediated MDR
From this study, one compound (24) was found to be a moder-
ately good dual ABCB1 and ABCG2 inhibitor and four compounds
13e14, 25e26) were multitarget ABCB1, ABCC1, and ABCG2 in-
(
hibitors. Especially the latter finding is a rather unusual observation
in literature [43]. Hence, these agents were assessed for their
capability to reverse ABCB1- and ABCC1-mediated MDR applying a
MTT-based cell viability assay as described earlier [15e17,40,41].
The intrinsic toxicity of compound 24 prevented test concen-
trations higher than 10
ABCB1 substrate daunorubicin dependent on the presence of either
.5 M, 1.0 M, 5.0 M, or 10 M of compound 24. Fig. 8B provides
mM. Fig. 8A shows the GI50 values of the
0
m
m
m
m
the resistance factors plotted against the tested concentrations of
compound 24. As can be seen from the plot, a tendency to a shift to
lower resistance factors as a function of the used compound con-
centration could be determined. Compound 24 had an EC50 of
Fig. 6. MTT-based cell viability assays of compounds 12 (circles) and 19 (squares) using
MDCK II BCRP (closed) and MDCK II (open) cells. While cell culture medium was taken
as reference for full cell viability (100%), 10% DMSO defined 0% cell viability. Shown is
mean ± SEM of at least three independent experiments.
34.4 mM, however, considering its high intrinsic toxicity as outlined
in Table 2, this value must be handled with care.
resistance factors of each experiment against the concentration of
the compound tested, the EC50 can be deduced. Table 5 gives the
GI50 values obtained for all compounds at different concentrations,
while Table 6 provides the EC50 values of these compounds. Fig. 7A
and B shows the results obtained for the most potent reverser of
ABCG2-mediated MDR within this manuscript, compound 14.
Strikingly, the obtained EC50 values of compounds 13e14 and 21
were in the low double-digit nanomolar concentration range. This
Unfortunately, the intrinsic toxicity and low potency of com-
pounds 13e14 and 25e26 made it impossible to analyze their
capability to reverse ABCB1- as well as ABCC1-mediated MDR (data
not shown). Nevertheless, as the field of multitarget ABC trans-
porter inhibitors emerged as a new potential therapeutic approach
to target multidrug-resistant cancer cells [43,44,45,46], and the
amount of potent broad-spectrum inhibitors is highly limited
[
11,15,27e35], the finding of these structures is of great interest
Table 4
Intrinsic toxicity of the 6-substituted 4-anilino-2-phenylpyrimidine derivatives obtained in a MTT-based cell viability assay using A2780/ADR, A2780, and H69AR cells as
described earlier [15e17,40,41]. GI50 values are expressed as mean ± SEM of at least three independent experiments. The tr has been calculated by dividing the GI50 value of a
compound by the corresponding IC50 value of the very same compound as shown in Table 2.
No
R
GI50 ± SEM [
m
M]
tr
GI50 ± SEM [
m
M]
GI50 ± SEM [
m
M]
tr
A2780/ADR
A2780
H69AR
13
14
24
25
26
i-propylamino
8.09 ± 0.50
12.3 ± 0.5
9.96 ± 0.29
5.91 ± 0.23
9.21 ± 0.19
1.26
2.08
3.34
1.68
2.03
13.3 ± 0.6
15.1 ± 0.4
9.07 ± 0.34
3.55 ± 0.31
8.02 ± 0.31
21.4 ± 0.3
23.0 ± 0.4
2.36
2.40
n-propylamino
2-hydroxyanilino
3-hydroxyanilino
4-hydroxyanilino
11.0 ± 0.1
13.7 ± 0.2
2.70
3.39
8

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Table 5
Shift of GI50 values of SN-38 in the presence of different concentrations of the evaluated 6-substituted 4-anilino-2-phenylpyrimidines at 0.01 mM, 0.10 mM, and 1.00 mM. Data
obtained using MDCK II BCRP as well as MDCK II cells. Shown is mean ± SEM of at least three independent experiments.
No
R
GI50
resistant cells
no compound
[
m
M] ± SEM
GI50
resistant cells
0.01 M compound
[
m
M] ± SEM
GI50
resistant cells
0.10 M compound
[
m
M] ± SEM
GI50
resistant cells
1.00 M compound
[
m
M] ± SEM
GI50 [mM] ± SEM
sensitive cells
no compound
m
m
m
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
methylamino
i-propylamino
n-propylamino
n-butylamino
phenyl
5.06 ± 0.38
4.29 ± 0.26
4.31 ± 0.30
7.23 ± 1.48
5.38 ± 0.40
5.09 ± 0.17
5.02 ± 0.47
7.40 ± 1.92
5.56 ± 0.38
4.56 ± 0.25
5.48 ± 0.26
4.67 ± 0.35
4.88 ± 0.15
4.74 ± 0.51
3.85 ± 0.45
5.34 ± 0.64
3.62 ± 0.19
2.93 ± 0.18
2.78 ± 0.10
5.81 ± 1.16
4.44 ± 0.41
4.13 ± 0.16
3.62 ± 0.24
6.55 ± 2.12
4.84 ± 0.40
2.96 ± 0.08
4.66 ± 0.15
4.15 ± 0.25
4.62 ± 0.27
4.24 ± 0.97
3.15 ± 0.43
5.12 ± 0.57
0.917 ± 0.026
0.717 ± 0.045
0.603 ± 0.035
1.68 ± 0.44
1.72 ± 0.16
1.64 ± 0.19
1.11 ± 0.05
3.18 ± 1.00
3.75 ± 0.34
0.616 ± 0.028
1.73 ± 0.05
2.09 ± 0.30
3.07 ± 0.27
3.28 ± 0.74
2.57 ± 0.52
4.08 ± 0.51
0.407 ± 0.027
0.314 ± 0.040
0.334 ± 0.022
0.731 ± 0.224
0.367 ± 0.041
0.333 ± 0.024
0.421 ± 0.025
0.727 ± 0.323
0.494 ± 0.012
0.334 ± 0.029
0.415 ± 0.022
0.365 ± 0.030
0.441 ± 0.028
0.563 ± 0.110
0.325 ± 0.023
0.563 ± 0.080
0.169 ± 0.021
0.232 ± 0.044
0.191 ± 0.021
0.265 ± 0.098
0.149 ± 0.015
0.187 ± 0.024
0.190 ± 0.018
0.255 ± 0.140
0.139 ± 0.014
0.231 ± 0.038
0.154 ± 0.014
0.256 ± 0.031
0.158 ± 0.012
0.265 ± 0.058
0.195 ± 0.025
0.175 ± 0.028
anilino
benzylamino
phenylpropylamino
phenylbutylamino
2-pyridylmethylamino
3-pyridylmethylamino
4-pyridylmethylamino
2-hydroxyanilino
3-hydroxyanilino
4-hydroxyanilino
4-cyanoanilino
Table 6
determine preferable (green) or less preferable (red) properties
were set according to critical markers. Regarding ABCG2 inhibition
Quantification of the compound’s ability to reverse ABCG2-mediated MDR deter-
mined with a MTT-based viability assay. The resistance factors derived from GI50
values from Table 5 were plotted against the corresponding concentration of the
investigated compound. The nonlinear regression of the obtained curve gave the
half-maximal reversal concentration (EC50). Shown is mean ± SEM of at least three
independent experiments.
(IC50), 0.25 mM was chosen as this represents superiority compared
to the gold standard for ABCG2 inhibition, compound 28. The Imax
cut-off value was set at 75% as previously reported [15]. Selectivity
and therapeutic ratios were considered to be preferable at ꢁ10.
Since double-digit nanomolar sensitizers were reported before
[17,37,42], a fairly low EC50 of 20 nM was chosen to classify the
compounds. Finally, the maximal degree of sensitization was
considered ꢁ50% to be preferable. As can be seen from Table 7,
compound 21 was found as the only representative that fulfilled the
requirements regarding the defined positive compound charac-
teristics. Hence, it was used first to evaluate the mode of inhibition
regarding the BCRP-mediated Hoechst 33342 transport, and second
to validate the computational model by docking studies as reported
earlier [17,26].
No
R
EC50 [mM] ± SEM
MDCK II BCRP
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
methylamino
i-propylamino
n-propylamino
n-butylamino
phenyl
0.0217 ± 0.0017
0.0180 ± 0.0024
0.0153 ± 0.0012
0.0298 ± 0.0028
0.0405 ± 0.0025
0.0435 ± 0.0024
0.0255 ± 0.0021
0.0670 ± 0.0041
0.187 ± 0.005
0.0155 ± 0.0006
0.0443 ± 0.0012
0.0725 ± 0.0094
0.147 ± 0.008
anilino
benzylamino
phenylpropylamino
phenylbutylamino
2-pyridylmethylamino
3-pyridylmethylamino
4-pyridylmethylamino
2-hydroxyanilino
3-hydroxyanilino
4-hydroxyanilino
4-cyanoanilino
3.2. Determination of mode of inhibition of compound 21
0.181 ± 0.023
0.171 ± 0.024
0.231 ± 0.006
As pyrimidines were shown before to inhibit ABCG2-mediated
Hoechst 33342 transport competitively [17,26], compound 21 was
evaluated to confirm the same mode of action as basis for ongoing
in silico experiments. The Lineweaver-Burk analysis as depicted in
Fig. 9 underpins the expected mode of competitive inhibition of the
compound class of the 6-substituted 4-anilino-2-phenylpyrimidine
derivatives in general and compound 21 in particular.
[
43,47] and might help in the development of further, more potent,
and less toxic therapeutics.
3
. Binding mode analysis of 6-substituted 4-anilino-2-
3.3. Molecular docking studies of compound 21
phenylpyrimidine derivatives
In order to visualize the hypothesis of the new binding (sub)
pocket in the binding site of ABCG2 we took compound 21 for
ongoing in silico experiments. As shown in Fig. 10, the binding pose
of 21 was similarly oriented as compound 8 [17]. The pyrimidine
scaffold formed hydrogen bond interactions with Asn387 and
Asn391, the 4-cyanophenyl group at position 2 was oriented in the
hydrophobic pocket formed by Val129, Val130, Met131, Gly132,
Val178, Thr180 and Val450 and the cyano group formed strong
electrostatic interaction with Arg191, while the phenyl substituent
at position 4 was positioned in a pocket formed by Leu388, Leu447
and Asp477. The methyl-2-pyridyl substituent connected by an
amino linker at position 6 of the pyrimidine scaffold oriented in the
identified (sub)pocket and possibly formed interactions with
Phe182 and Glu446 located at a distance of 3.2 Å and 3.6 Å,
3
.1. Classification of 6-substituted 4-anilino-2-phenylpyrimidine
derivatives
Table 7 gives a summary of the obtained data of all compounds,
which was taken in consideration for further validation of the
developed computational model. For simplicity, positive and
negative characteristics of the compounds were marked with green
or red, respectively. The IC50 and Imax values were taken from
Table 1. The selectivity ratios (sr) were calculated from values of
Tables 1 and 2 The therapeutic ratios (tr) were assembled from
Table 3. The EC50 values were compiled from Table 4, and the
ꢀ
maximal degree of sensitization ( s) was calculated from the values
of Table 5 as previously reported [15]. The cut-off values to
9

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Fig. 7. Reversal of ABCG2-mediated MDR by compound 14 determined with a MTT-
based cell viability assay. (A) Sensitization of ABCG2-overexpressing MDCK II BCRP
cells with regard to SN-38. Compound 14 was used at concentrations of 0
mM (C),
0
.010 M (-), 0.100 M (:), as well as 1.000 M (;), and was compared to the wild
m
m
m
type cell line without compound addition (B). Shown is mean ± SEM of at least three
independent experiments. (B) Plotted resistance factors derived from GI50 values from
Table 5 against applied concentrations of compound 14. Nonlinear regression of the
Fig. 8. Reversal of ABCB1-mediated MDR by compound 24 determined with a MTT-
based cell viability assay. (A) Sensitization of ABCB1-overexpressing A2780/ADR cells
obtained curve resulted in the half-maximal reversal concentration (EC50
) of
with regard to daunorubicin. Compound 24 was used at concentrations of 0
mM (C),
15.3 ± 1.2 nM.
0
.5 M (-), 1.0 M (:), 5.0 M (;), as well as 10.0 M (A), and was compared to
m
m
m
m
the wild type cell line without compound supplementation (B). Shown is mean ± SEM
of at least three independent experiments. (B) Plotted resistance factors derived from
GI50 values from (A) against applied concentrations of compound 24. Nonlinear
regression and extrapolation of the obtained curve resulted in the half-maximal
respectively. The amino linker between might introduce flexibility
for the different substitutions at position 6. The hypothesis is
supported by compound 16, where a phenyl substitution resulted
in severe loss of activity against ABCG2 possibly due to steric
clashes in the binding pocket. Furthermore, the substituted aro-
matic residues, in particular at position 2, favored the interaction
with the amino acid residues in the binding (sub)pocket. Addi-
tionally, it opened up the possibility for further exploration of the
reversal concentration (EC50) of 34.4 ± 4.9 mM.
herein presented 6-substituted 4-anilino-2-phenylpyrimidines
have three major properties: (i) they are like their parent com-
pounds highly potent regarding ABCG2 inhibition, which is re-
flected in seven compounds (12, 17, 19e21, 24, 27) having IC50
values of less than 250 nM e making them significantly more
potent than the gold standard compound 28; (ii) similar to their
parent compounds, their efficacy regarding reversal of ABCG2-
mediated MDR is some of the highest reported in the literature
(
sub)pocket with different substitutions to improve the activity and
properties of ABCG2 inhibitors in general.
4
. Conclusions
-Anilino-6-methyl-2-phenylpyrimidines were discovered as
[
17,23,42]. This accounts especially for compounds 13e14, and 21
4
with EC50 values below 20 nM; (iii) due to their 6-substitution,
these compounds inherited exceptional cell toxicity, with single-
digit GI50 values for most of the evaluated compounds e inde-
pendent from the used cell lines. However, as their inhibitory
highly potent inhibitors of ABCG2 [17], with an efficacy regarding
reversal of ABCG2-mediated MDR that has barely been reported
before [17,23,42]. Molecular docking studies revealed a binding
(
sub)pocket which was explored as part of the current study. The
10

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Table 7
Summary of obtained biological results of tested 6-substituted 4-anilino-2-phenylpyrimidine derivatives as taken and/or calculated from Tables 1e6 Positive and negative
compound characteristics are depicted in green and red, respectively.
a ¼ the half maximal-inhibition concentration (IC50) values were taken from Table 1.
b ¼ the maximal inhibition (Imax) values were taken from Table 1.
c ¼ the selectivity ratios over ABCB1 were calculated by taking the IC50 (ABCB1) of the compounds from Table 2 divided by the IC50 (ABCG2) values of the corresponding
compounds from Table 1.
d ¼ the selectivity ratios over ABCC1 were calculated by taking the IC50 (ABCC1) of the compounds from Table 2 divided by the IC50 (ABCG2) values of the corresponding
compounds from Table 1.
e ¼ the therapeutic ratios (tr) were taken from Table 3.
f ¼ the half-maximal sensitization concentrations (EC50) were obtained from Table 4.
ꢀ
g ¼ the maximal degree of sensitization ( s) was calculated as reported before [15] from the potentiation factor of the compound at 1
mM divided by the resistance factor of the
resistant cell line without treatment. The potentiation factor was calculated taking the GI50 value of the untreated resistant cells (Table 5) divided by the GI50 values of the
resistant cells treated with 1 M of the compound (Table 5). The resistance factors were calculated by taking the GI50 values of the resistant cells without treatment (Table 5)
m
divided by the GI50 values of the sensitive cell line (Table 5).
h ¼ n. e. ¼ no effect.
i ¼ Imax at highest tested concentration.
power and efficacy were determined in a much lower concentra-
tion range, it can be concluded that the compounds exerted their
reversal of ABCG2-mediated MDR due to ABCG2 inhibition.
Nevertheless, their intrinsic toxicity is an undesired side effect,
making these compounds generally less applicable for anticancer
therapy.
The most significant finding of this work is that the previously
discovered (sub)pocket [17] could be used to engage the herein
presented compound 21 by strong interactions with its molecular
target, human ABCG2. To our best knowledge, this is the very first
time that a structure-based drug design approach successfully
discovered new and highly potent ABCG2 inhibitors and elucidated
in parallel the drug-target interaction.
Until now, less than 140 compounds have been reported with this
ability. Among these, only around 50 have been found to exert their
inhibitory power against ABCB1, ABCC1, and ABCG2 below 10 mM
[11,15e17,27e36,42,48e55] Hence, the sole discovery of com-
pounds 13e14 and 25e26 is very important for ongoing research to
understand the function of ABC transporters in general and the
necessities for inhibition of the particular representatives, as these
compounds belong to the 50 most potent triple ABCB1, ABCC1, and
ABCG2 inhibitors ever reported. Strikingly, compounds 25e26
belong even to the 22 most potent multitarget inhibitors that have
IC50 values against ABCB1, ABCC1, and ABCG2 ꢂ5
mM [11,15,27e36].
As broad-spectrum reversers of ABCB1-, ABCC1-, and ABCG2-
mediated MDR have barely been reported [10,13e24], this discov-
ery might be of great importance in the field of multitarget ABC
transporter inhibition [43,47].
Interestingly, the herein evaluated compounds contained five
multitarget ABCB1, ABCC1, and ABCG2 inhibitors, four of them
exerting their inhibitory power against the three stated trans-
porters at least in the single-digit micromolar concentration range.
11

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Fig. 9. Lineweaver-Burk plot of compound 21 using ABCG2-overexpressing MDCK II BCRP cells. Hoechst 33342 was used at 0.3, 0.4, 0.5, 0.6, 0.7, and 1.0
m
M. Compound 21 was used
at concentrations of 0
experiments.
mM (B), 0.00560 mM (C), 0.0100 mM (-), 0.0178 mM (:), 0.0316 mM (;), and 0.0562 (A). Depicted is a representative experiment out of three independent
Fig. 10. The docked pose of 8 (carbon colored green) and 21 (carbon colored magenta) as observed in the binding pocket of human ABCG2 (PDB ID: 6FFC). The amino acids in the
binding pocket are shown as line representation and in particular the important residues that possibly form interaction with the compounds are indicated in stick representation.
The interactions with the compounds are shown in red dashed lines. Oxygen atoms are colored in red, nitrogen atoms in blue, and polar hydrogen atoms are colored in white. (For
interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
12

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
5
. Experimental section
Germany), Th. Geyer GmbH Co KG (Renningen, Germany) and
Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Storage of the
5
.1. Molecular docking
used compounds (10 mM stock solution) was performed in dime-
ꢀ
thylsulfoxide (DMSO) at ꢃ20 C. Krebs-HEPES buffer [KHB; CaCl
2
The human ABCG2 transporter was crystallized using cryo-
(1.3 mM)],
hydroxyethylpiperazin-N -2-ethansulfonic acid; 10.0 mM), KCl
(4.7 mM), KH PO (1.2 mM) MgSO (1.2 mM) NaCl (118.6 mM)
NaHCO
(4.2 mM), and NaOH (adjusting to pH ¼ 7.4) were used for
dilutions. Before use, KHB was sterile filtered with membrane fil-
ters (Whatman FP 30/0.2 M CA-S filter units, GE Healthcare UK
D-glucose monohydrate (11.7 mM), HEPES (N-2-
0
electron microscopy (cryo-EM) in complex with the fumi-
tremorgin C-related inhibitor, MZ29 (PDB ID: 6FFC) was down-
loaded from Research Collaboratory for Structural Bioinformatics
Protein Data Bank (PDB) [23]. As an initial step, the structure was
prepared and protonated using the protein structure preparation
and Protonate3D modules implemented in Molecular Operating
Environment (MOE) 2018.01, respectively [56]. Then the receptor
structure of the human ABCG2 was applied for flexible ligand
docking using LeadIT from BioSolveIT, GmbH Germany [57]. During
the docking simulations, the ligands were fully flexible while the
residues of the receptor were treated as rigid. For the docking
calculations, the binding site of the receptor was defined in 20 Å
spacing of the amino acid residues centered based on the cocrys-
tallized ligand, MZ29. Then the selected compounds were docked
with default parameters and predicted the binding pose using the
FlexX module implemented in LeadIT [58]. From the resulted
binding poses, the top 100 highest scoring docked poses were
selected for the analysis. Then the docked pose having the lowest
binding free energy for interactions was visualized and selected as
the putative binding mode of the compound using MOE2018.01
2
4
4
3
m
Limited, Buckinghamshire, UK) using a Braun Injekt 20 mL syringe
(ALMO-Erzeugnisse, Erwin Busch GmbH, Bad Arolsen, Germany)
and stored in cellstar 50 mL tubes (Greiner Bio-One, Frickenhausen,
Germany).
5.2.2. Synthesis
5.2.2.1. Preparation of 2-phenylpyrimidine-4,6-diol (9). 5.56
g
(241.7 mmol) sodium were dissolved in 200 ml ethanol to yield
sodium ethoxide solution. Subsequently, 12.18 g (66 mmol) ben-
zamidine hydrochloride hydrate and 11.79 g (89 mmol) ethyl
dimethyl malonate were added into sodium ethoxide solution
respectively. The resulting mixture was heated to reflux overnight
and cooled down to room temperature. After solvent was removed
under reduced pressure, 50 ml of water were added to the solid and
the mixture was acidified with concentrated HCl to pH 5e6. The
formed precipitate was vacuum filtered and washed with water
three times. Finally, the obtained crude residue was recrystallized
[
56].
5
5
.2. Chemistry
from DMF/H
5.8 g; 46.7%). H NMR (500 MHz, DMSO‑d
8.11e8.07 (m, 2H), 7.60e7.56 (m, 1H), 7.54e7.50 (m, 2H), 5.35 (s,
2
O to give 2-phenylpyrimidine-4,6-diol as orange solid
1
(
6
) d: 11.80 (s, 2H),
.2.1. Materials
13
Chemicals were purchased from Acros Organics (Geel, Belgium),
1H). C NMR APT (126 MHz, DMSO) d: 131.81, 128.69, 127.87, 88.46.
Alfa Aesar (Karlsruhe, Germany), SigmaAldrich (Steinheim, Ger-
many), Carl Roth GmbH (Karlsruhe, Germany), Calbiochem (San
Diego, CA, USA), Frontier Scientific Inc. (Logan, UT, USA), or Merck
5.2.2.2. Preparation of 4,6-dichloro-2-phenylpyrimidine (10).
5.8 g (30.8 mmol) 2-phenylpyrimidine-4,6-diol was added to 20 ml
(
1
Darmstadt, Germany). Microwave reactions were performed in
0 mL vials with a CEM Discover SP (CEM GmbH, Kamp-Lintfort,
POCl
Excess POCl
3
. The solution was refluxed until the reaction was complete.
was removed under reduced pressure. The oily res-
3
Germany). The reaction progress was monitored by thin layer
chromatography (TLC) using an aluminum plate coated with silica
gel 60 F254 (Merck Millipore, Billerica, MA, USA). Dichloro-
idue was transferred onto crushed ice portion by portion while
keeping the temperature of solution below 10 C. Ethyl acetate was
used to extract products 3 times. The organic layer was washed
3
with saturated NaHCO , saturated NaCl solution and dried over
ꢀ
ꢀ
methane/methanol and ethyl acetate/petroleum ether (40e60 C)
were used as eluents. The identity of the compounds was
confirmed by NMR. H- and C spectra were obtained on Bruker
Advance 500 MHz (500/126 MHz) or Bruker Advance 600 MHz
magnesium sulfate. After filtration, ethyl acetate was removed
under reduced pressure. Final purification was performed by col-
umn chromatography with DCM/MeOH ¼ 50:1 as eluent to yield
1
13
1
(
600/151 MHz). The chemical shifts (
d
) are expressed in ppm. For
white needle crystals (yield: 4.1 g; 59.4%). H NMR (600 MHz,
13
13
the assignment of some C signals, attached proton test (APT) was
used. Multiplicity of the signals is indicated as singlet (s), doublet
CDCl
3
)
d
: 8.44 (d, J ¼ 7.9 Hz, 2H), 7.56e7.48 (m, 3H), 7.28 (s, 1H).
C
3
NMR (126 MHz, CDCl ) d: 165.91, 162.18, 135.02, 132.40, 129.02,
(
d), triplet (t), quartet (q), quintet (p) and multiplet (m). Coupling
128.88, 118.91.
constants J are given in Hz. The purity and molecular mass of
compounds were determined by LC-MS (Applied Biosystems API
5.2.2.3. Preparation of 4-((6-chloro-2-phenylpyrimidin-4-yl)amino)
benzonitrile (11). 3.04 (13.5 mmol) 4,6-dichloro-2-
2
000 LC-MS/MS, HPLC Agilent 1100). Krebs-HEPES-buffer [KHB;
CaCl (1.3 mM), -glucose monohydrate (11.7 mM), HEPES
10.0 mM), KCl (4.7 mM), KH PO (1.2 mM), MgSO (1.2 mM), NaCl
118.6 mM), NaHCO
(4.2 mM), and NaOH (adjusting to pH ¼ 7.4)
g
2
D
phenylpyrimidine was added to 30 ml isopropanol and this
mixture was refluxed until all 4,6-dichloro-2-phenylpyrimidine
was dissolved. 1.57 g (13.3 mmol) 4-cyanoaniline was dissolved in
30 ml isopropanol and this solution was added dropwise to
refluxed 4,6-dichloro-2-phenylpyrimidine solution. If reaction did
not initiate, 0.1 equivalent of conc. HCl was added as catalyst. The
precipitate was filtered off and washed with saturated sodium bi-
carbonate solution 3 times followed by distilled water 3 times.
Further purification was performed by column chromatography
(
(
2
4
4
3
was filtered sterile and used for dilutions. The reference com-
pounds cyclosporine (ABCB1) and Ko143 [(3S,6S,12aS)-
A
1
,2,3,4,6,7,12,12a-octahydro-9-methoxy-6-(2-methylpropyl)-1,4-
0
dioxopyrazino[1 ,2’:1,6]pyrido[3,4-b]indole-3-propanoic acid 1,1-
dimethylethylester; compound 28; ABCG2] were delivered by
Tocris Bioscience (Bristol, IO, USA). Calcein AM (ABCC1 and ABCB1)
and pheophorbide A (ABCG2) were purchased from Calbiochem
ꢀ
with ethyl acetate/petroleum ether (40e60 C) ¼ 1:2 as eluent,
1
[EMD Chemicals (San Diego, CA, USA] and Frontier Scientific Inc.
yellow solid (3.6 g, 88.2%). H NMR (500 MHz, DMSO‑d
6
)
d: 10.34 (s,
(
Logan, UT, USA), respectively. Daunorubicin was delivered by
1H), 8.32e8.29 (m, 2H), 7.97e7.94 (m, 2H), 7.87e7.84 (m, 2H),
13
Sigma (Oakville, ON, Canada). SN-38 was supplied by Tocris
Bioscience (Bristol, IO, USA). Other chemicals were purchased from
Carl Roth GmbH (Karlsruhe, Germany), Merck KgaA (Darmstadt,
7.58e7.53 (m, 3H), 6.85 (s, 1H). C NMR (126 MHz, DMSO)
163.70, 160.98, 158.82, 143.56, 136.07, 133.31, 131.43, 128.72, 127.92,
119.67, 119.10, 104.52, 104.15. LC-MS (m/z) calcd. for C17 11ClN
d:
H
4
13

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
þ
[
Mþ1] : 307.07, Found: 307.2, Purity: 97.9%.
2H), 7.90 (d, J ¼ 8.9 Hz, 2H), 7.75e7.70 (m, 2H), 7.51e7.45 (m, 3H),
7
.14 (s, 1H), 5.83 (s, 1H), 1.61 (h, J ¼ 7.4 Hz, 2H), 0.95 (t, J ¼ 7.4 Hz,
13
5
.2.2.4. General procedure for the preparation of 4-anilino-2-
phenylpyrimidine derivatives (12e27)
.2.2.4.1. Method A. A mixture of 1 equivalent 4-((6-chloro-2-
phenylpyrimidin-4-yl)amino)benzonitrile, 10e20 equivalent of
amine in 0.5 ml isopropanol was heated to 120 C in microwave for
3H). C NMR (126 MHz, DMSO) d: 163.25, 162.20, 145.68, 138.36,
133.05, 130.00, 128.19, 127.52, 119.59, 118.24, 101.52, 42.20, 22.15,
þ
5
11.52. LC-MS (m/z) calcd. for C20
330.1, Purity: 99.3%.
H
19
N
5
[Mþ1] : 330.16, Found:
ꢀ
3
0 min. Further purification for certain compounds was performed
5.2.2.8. 4-((6-(butylamino)-2-phenylpyrimidin-4-yl)amino)benzoni-
trile (15). The title compound was synthesized from 4-((6-chloro-
2-phenylpyrimidin-4-yl)amino)benzonitrile (125 mg, 0.41 mmol)
and 1-butylamine (390 mg, 5.33 mmol) as described in the general
procedure of method A. Brown solid (66.5 mg, 47.4%). H NMR
6
(500 MHz, DMSO‑d ) d: 9.55 (s, 1H), 8.32e8.29 (m, 2H), 7.90 (d,
by column chromatography with ethyl acetate/petroleum ether
ꢀ
(
40e60 C) ¼ 1:2 as eluent.
5.2.2.4.2. Method B. To a three necked round-bottom flask were
1
added 112 mg 4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzo-
nitrile (0.37 mmol), 50 mg phenylboronic acid (0.41 mmol) and
8
ml Dioxane/2M K
uum/argon cycle three times. Catalyst 10 mg Pd(t-Bu
0.02 mmol) was added and the mixture was degassed twice more.
2
CO
3
¼ 3:1. The mixture was degassed by vac-
J ¼ 8.4 Hz, 2H), 7.75e7.71 (m, 2H), 7.51e7.46 (m, 3H), 7.14 (s, 1H),
3
P)
2
5.82 (s, 1H), 1.60e1.54 (m, 2H), 1.39 (q, J ¼ 7.4 Hz, 2H), 0.93 (t,
1
3
(
J ¼ 7.4 Hz, 3H). C NMR (126 MHz, DMSO)
d: 163.25, 162.22, 145.71,
ꢀ
The mixture was heated to 85 C for 7 h until TLC confirmed the
completion of the reaction and cooled to ambient temperature.
Afterward, the slurry was extracted with saturated NaHCO
138.39, 133.09, 130.05, 128.23, 127.54, 119.64, 118.27, 101.53, 19.68,
13.73. LC-MS (m/z) calcd. for C21
þ
H
21
N
5
[Mþ1] : 344.18, Found:
3
/ethyl
344.1, Purity: 98.9%.
acetate 3 times. The collected ethyl acetate solution was dried over
magnesium sulfate and concentrated under reduced pressure.
Further purification was performed by recrystallization from ethyl
acetate/n-hexane.
5.2.2.9. 4-((2,6-diphenylpyrimidin-4-yl)amino)benzonitrile
The title compound was synthesized as described in the general
procedure of method B. Yellow solid (63.0 mg, 49.6%). H NMR
(16).
1
5.2.2.4.3. Method C. A mixture of 1 equivalent 4-((6-chloro-2-
(500 MHz, DMSO‑d
6
) d: 10.26 (s, 1H), 8.51e8.48 (m, 2H), 8.21e8.18
phenylpyrimidin-4-yl)amino)benzonitrile, 10e15 equivalent of
(m, 2H), 8.07e8.04 (m, 2H), 7.87e7.84 (m, 2H), 7.62e7.55 (m, 6H),
7.29 (s, 1H). C NMR (126 MHz, DMSO) d: 163.02, 162.22, 161.01,
144.44, 137.68, 136.81, 133.31, 130.74, 130.63, 128.98, 128.61, 127.87,
13
substituted aniline in 3 ml isopropanol and 0.1 equivalent conc. HCl
ꢀ
was heated to 110 C in microwave for 30 min. The precipitate was
filtered off and washed with saturated sodium bicarbonate solution
126.61, 119.34, 119.16, 103.24, 101.47. LC-MS (m/z) calcd. for
þ
3
times followed by distilled water 3 times. Further purification was
C
H
23 16
N
4
[Mþ1] : 349.14, Found: 349.0, Purity: 97.1%.
performed by column chromatography with ethyl acetate/petro-
ꢀ
leum ether (40e60 C) ¼ 1:2 as eluent.
5.2.2.10. 4-((2-phenyl-6-(phenylamino)pyrimidin-4-yl)amino)ben-
zonitrile (17). The title compound was synthesized from 4-((6-
5
.2.2.5. 4-((6-(methylamino)-2-phenylpyrimidin-4-yl)amino)benzo-
nitrile (12). The title compound was synthesized from 4-((6-
chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (136 mg,
.44 mmol), methylamine hydrochloride (306 mg, 4.53 mmol) and
chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile
0.41 mmol) and aniline (508 mg, 5.46 mmol) as described in the
general procedure of method C. White solid (47.6 mg, 31.8%).
(127
mg,
1
H
0
NMR (500 MHz, DMSO‑d : 9.77 (s,1H), 9.37 (s,1H), 8.36e8.32 (m,
6
) d
N,N-diisopropylethylamine (0.8 ml, 4.48 mmol) as described in the
general procedure of method A. Yellow solid (20.7 mg, 15.5%). H
2H), 7.95e7.91 (m, 2H), 7.79e7.75 (m, 2H), 7.69e7.64 (m, 2H),
7.56e7.50 (m, 3H), 7.41e7.35 (m, 2H), 7.04 (tt, J ¼ 1.1, 7.3 Hz, 1H),
1
13
NMR (500 MHz, DMSO‑d
6
)
d
: 9.57 (s, 1H), 8.32 (dd, J ¼ 3.1, 6.6 Hz,
6.23 (s, 1H). C NMR (126 MHz, DMSO) d: 162.61, 161.09, 160.17,
2
H), 7.91 (d, J ¼ 8.6 Hz, 2H), 7.75e7.68 (m, 2H), 7.51e7.40 (m, 3H),
145.25, 140.23, 138.03, 133.13, 130.37, 128.79, 128.42, 127.61, 122.09,
1
3
7.07 (s, 1H), 5.80 (s, 1H), 2.87 (s, 3H). C NMR (126 MHz, DMSO)
d
:
120.08,119.49,118.59,102.11, 86.90. LC-MS (m/z) calcd. for C23
[Mþ1] : 364.15, Found: 364.0, Purity: 99.1%.
H
17
N
5
þ
163.78, 162.18, 159.71, 145.64, 138.29, 133.06, 130.03, 128.19, 127.55,
1
19.59, 118.26, 101.58, 27.35. LC-MS (m/z) calcd. for C18
H
15
N
5
þ
[Mþ1] : 302.13, Found: 302.0, Purity: 98.3%.
5.2.2.11. 4-((6-(nenzylamino)-2-phenylpyrimidin-4-yl)amino)benzo-
nitrile (18). The title compound was synthesized from 4-((6-
5
.2.2.6. 4-((6-(isopropylamino)-2-phenylpyrimidin-4-yl)amino)ben-
chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile
(76
mg,
zonitrile (13). The title compound was synthesized from 4-((6-
chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (141 mg,
.46 mmol) and isopropylamine (880 mg, 14.89 mmol) as described
in the general procedure of method A. Yellow solid (100.3 mg,
6.1%).
1H NMR (500 MHz, DMSO‑d
.89 (d, J ¼ 8.7 Hz, 2H), 7.74e7.71 (m, 2H), 7.50e7.46 (m, 3H), 6.99
0.25 mmol) and benzylamine (270 mg, 2.52 mmol) as described in
the general procedure of method A. Brown solid (25.3 mg, 27.3%).
1
0
H NMR (500 MHz, DMSO‑d
6
)
d
: 9.58 (s, 1H), 8.30 (dd, J ¼ 3.0,
6.7 Hz, 2H), 7.88 (d, J ¼ 8.8 Hz, 2H), 7.75e7.70 (m, 3H), 7.48 (dd,
J ¼ 1.9, 5.0 Hz, 3H), 7.40e7.37 (m, 2H), 7.34 (t, J ¼ 7.6 Hz, 2H), 7.24 (t,
6
13
6
)
d
: 9.54 (s, 1H), 8.32e8.29 (m, 2H),
J ¼ 7.3 Hz, 1H), 5.86 (s, 1H), 4.60 (s, 2H). C NMR (126 MHz, DMSO)
7
d: 163.19, 162.25, 159.77, 145.58, 140.13, 138.23, 133.08, 130.12,
13
(
d, J ¼ 7.4 Hz, 1H), 5.81 (s, 1H), 1.21 (d, J ¼ 6.5 Hz, 6H). C NMR
128.31, 128.25, 127.57, 127.17, 126.70, 119.60, 118.32, 101.65, 43.79.
þ
(
126 MHz, DMSO) : 162.49, 162.25, 145.73, 138.42, 133.07, 130.02,
d
LC-MS (m/z) calcd. for C24
Purity: 96.7%.
19
H N
5
[Mþ1] : 378.16, Found: 378.3,
1
28.22, 127.53, 119.64, 118.25, 101.50, 59.71, 41.74, 22.50, 14.06. LC-
þ
MS (m/z) calcd. for C20
19
H N
5
[Mþ1] : 330.16, Found: 330.1, Purity:
9
9.0%.
5.2.2.12. 4-((2-phenyl-6-((3-phenylpropyl)amino)pyrimidin-4-yl)
amino)benzonitrile (19). The title compound was synthesized from
4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (86 mg,
0.28 mmol) and 3-phenylpropylamine (759 mg, 5.61 mmol) as
5
.2.2.7. 4-((2-phenyl-6-(propylamino)pyrimidin-4-yl)amino)benzo-
nitrile (14). The title compound was synthesized from 4-((6-
chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (168 mg,
.55 mmol) and 1-propylamine (308 mg, 5.22 mmol) as described
in the general procedure of method A. White solid (126.4 mg,
described in the general procedure of method A. White solid
1
0
(70.3 mg, 61.9%). H NMR (500 MHz, DMSO‑d
6
)
d: 9.54 (s, 1H),
8.30e8.25 (m, 2H), 7.90 (d, J ¼ 8.6 Hz, 2H), 7.75e7.71 (m, 2H),
1
6
6
9.9%). H NMR (500 MHz, DMSO‑d ) d: 9.54 (s, 1H), 8.33e8.28 (m,
7.50e7.46 (m, 3H), 7.32e7.17 (m, 6H), 5.83 (s, 1H), 3.36 (d,
14

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
J ¼ 18.3 Hz, 0H), 2.69 (t, J ¼ 7.6 Hz, 2H), 1.90 (p, J ¼ 7.3 Hz, 2H). 13C
0.33 mmol) and 2-hydroxyaniline (222 mg, 2.03 mmol) as
NMR (126 MHz, DMSO)
d
: 163.21, 162.19, 159.55, 145.66, 141.72,
described in the general procedure of method C. Yellow solid
1
138.31, 133.05, 130.02, 128.28, 128.22, 128.18, 127.53, 125.67, 119.59,
(48.8 mg, 38.6%). H NMR (500 MHz, DMSO‑d
6
)
d: 9.60 (s, 1H), 8.36
1
18.27, 101.55, 32.62, 30.73. LC-MS (m/z) calcd. for C26
H
23
N
5
(dd, J ¼ 3.0, 6.6 Hz, 2H), 7.91e7.87 (m, 2H), 7.78e7.71 (m, 3H),
þ
[Mþ1] : 406.20, Found: 406.3, Purity: 98.0%.
7.52e7.47 (m, 3H), 7.36 (dd, J ¼ 1.3, 5.0 Hz, 1H), 7.07 (dd, J ¼ 1.3,
1
3
3.5 Hz, 1H), 6.97 (dd, J ¼ 3.4, 5.1 Hz, 1H), 5.90 (s, 1H), 4.78 (s, 2H).
C
5
.2.2.13. 4-((2-phenyl-6-((4-phenylbutyl)amino)pyrimidin-4-yl)
NMR (126 MHz, DMSO) d: 162.77, 162.20, 159.71, 145.54, 143.35,
amino)benzonitrile (20). The title compound was synthesized from
138.16, 133.36, 133.06, 130.15, 128.24, 127.64, 126.54, 125.19, 124.88,
4
0
-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (130 mg,
.42 mmol) and 4-phenylbutylamine (522 mg, 3.50 mmol) as
119.57, 118.35, 113.41, 101.70, 85.28. LC-MS (m/z) calcd. for
þ
C
H
23 17
N
5
O [Mþ1] : 380.14, Found: 380.1, Purity: 95.9%.
described in the general procedure of method A. Pale brown solid
76.8 mg, 43.3%). 1H NMR (500 MHz, DMSO‑d
: 9.55 (s, 1H),
.32e8.28 (m, 2H), 7.89 (d, J ¼ 8.4 Hz, 2H), 7.75e7.71 (m, 2H),
.50e7.45 (m, 3H), 7.26 (t, J ¼ 7.5 Hz, 2H), 7.20 (d, J ¼ 6.8 Hz, 2H),
(
8
7
6
)
d
5.2.2.18. 4-((6-((3-hydroxyphenyl)amino)-2-phenylpyrimidin-4-yl)
amino)benzonitrile (25). The title compound was synthesized from
4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (106 mg,
0.34 mmol) and 3-hydroxyaniline (290 mg, 2.65 mmol) as
7
2
.18e7.14 (m, 2H), 5.82 (s, 1H), 2.64 (t, J ¼ 7.5 Hz, 2H), 1.71e1.64 (m,
13
H), 1.64e1.57 (m, 2H). C NMR (126 MHz, DMSO)
d
: 163.23,
described in the general procedure of method C. Grey solid
1
162.24, 145.70, 142.11, 138.38, 133.09, 130.05, 128.25, 128.22, 128.18,
(67.1 mg, 51.4%). H NMR (500 MHz, DMSO‑d
6
)
d: 9.76 (s, 1H), 9.38
1
27.56, 125.61, 119.64, 118.27, 101.53, 55.99, 34.84, 28.46, 18.52. LC-
(s, 1H), 9.24 (s, 1H), 8.37e8.33 (m, 2H), 7.94e7.91 (m, 2H), 7.78e7.75
(m, 2H), 7.56e7.50 (m, 3H), 7.16e7.12 (m, 2H), 7.06 (ddd, J ¼ 1.0, 2.0,
þ
MS (m/z) calcd. for C27
H
25
N
5
[Mþ1] : 420.21, Found: 420.3, Purity:
13
9
6.3%.
8.1 Hz, 1H), 6.46 (ddd, J ¼ 1.0, 2.4, 8.0 Hz, 1H), 6.23 (s, 1H). C NMR
(126 MHz, DMSO)
d: 162.63, 161.13, 160.13, 157.75, 145.29, 141.25,
5
.2.2.14. 4-((2-phenyl-6-((pyridin-2-ylmethyl)amino)pyrimidin-4-
138.04, 133.13, 130.35, 129.41, 128.39, 127.70, 119.51, 118.57, 110.97,
yl)amino)benzonitrile (21). The title compound was synthesized
from 4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile
95 mg, 0.31 mmol) and 2-picolylamine (585 mg, 5.39 mmol) as
described in the general procedure of method A. Grey solid
109.45, 107.26, 102.07, 86.91. LC-MS (m/z) calcd. for C23
[Mþ1] : 380.14, Found: 380.1, Purity: 99.0%.
H
17
N
5
O
þ
(
5.2.2.19. 4-((6-((4-hydroxyphenyl)amino)-2-phenylpyrimidin-4-yl)
amino)benzonitrile (26). The title compound was synthesized from
4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (101 mg,
0.33 mmol) and 4-hydroxyaniline (390 mg, 3.57 mmol) as
1
(
(
7
22.0 mg, 18.8%). H NMR (500 MHz, DMSO‑d
d, J ¼ 4.3 Hz, 1H), 8.26 (dd, J ¼ 3.0, 6.6 Hz, 2H), 7.90e7.87 (m, 2H),
.80e7.71 (m, 4H), 7.48e7.44 (m, 3H), 7.39 (d, J ¼ 7.9 Hz, 1H), 7.26
6
) d: 9.60 (s, 1H), 8.54
13
(
dd, J ¼ 5.0, 7.3 Hz, 1H), 5.91 (s, 1H), 4.68 (s, 2H). C NMR (126 MHz,
described in the general procedure of method C. Grey solid
1
DMSO) : 163.17, 162.23, 159.81, 159.25, 148.85, 145.54, 138.15,
d
(22.7 mg, 18.2%). H NMR (500 MHz, DMSO‑d
6
)
d: 9.66 (s, 1H), 9.22
1
36.67, 133.05, 130.10, 128.21, 127.54, 122.00, 121.01, 119.56, 118.33,
(s, 1H), 8.98 (s,1H), 8.34e8.30 (m, 2H), 7.93e7.90 (m, 2H), 7.76e7.73
(m, 2H), 7.54e7.48 (m, 3H), 7.33 (d, J ¼ 8.3 Hz, 2H), 6.81e6.78 (m,
þ
101.69, 45.93. LC-MS (m/z) calcd. for C23
H
18
N
6
[Mþ1] : 379.16,
13
Found: 379.3, Purity: 98.5%.
2H), 6.02 (s, 1H). C NMR (126 MHz, DMSO) d: 162.54, 161.87,
160.13, 153.51, 145.41, 138.14, 133.09, 131.11, 130.22, 128.33, 127.58,
5
.2.2.15. 4-((2-phenyl-6-((pyridin-3-ylmethyl)amino)pyrimidin-4-
yl)amino)benzonitrile (22). The title compound was synthesized
from 4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile
86 mg, 0.28 mmol) and 3-picolylamine (687 mg, 6.35 mmol) as
described in the general procedure of method A. White solid
123.52, 119.54, 118.41, 115.41, 101.86, 85.41. LC-MS (m/z) calcd. for
þ
C
H
23 17
N
5
O [Mþ1] : 380.14, Found: 380.1, Purity: 96.5%.
(
5.2.2.20. 4,4’-((2-phenylpyrimidine-4,6-diyl)bis(azanediyl))dibenzo-
nitrile (27). The title compound was the side product in synthesis
1
(
(
2
7
80.3 mg, 75.8%). H NMR (500 MHz, DMSO‑d
s, 1H), 8.45 (d, J ¼ 3.3 Hz, 1H), 8.32e8.27 (m, 2H), 7.90e7.87 (m,
H), 7.79e7.71 (m, 4H), 7.50e7.46 (m, 3H), 7.36 (ddd, J ¼ 0.9, 4.7,
6
)
d
: 9.59 (s, 1H), 8.63
of 4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile. Yel-
1
low solid. H NMR (500 MHz, DMSO‑d
6
)
d
: 9.92 (s, 2H), 8.34 (dd,
J ¼ 3.0, 6.7 Hz, 2H), 7.96e7.92 (m, 4H), 7.81e7.77 (m, 4H), 7.55 (dd,
13
13
.9 Hz, 1H), 5.89 (s, 1H), 4.63 (s, 2H). C NMR (126 MHz, DMSO)
d
:
J ¼ 2.0, 5.0 Hz, 3H), 6.32 (s, 1H). C NMR (126 MHz, DMSO)
d:
163.00, 162.28, 159.75, 148.82, 148.00, 145.51, 138.15, 134.99, 133.07,
162.71, 160.33, 144.92, 137.71, 133.18, 130.64, 128.57, 127.72, 119.42,
118.88, 102.57, 88.87. LC-MS (m/z) calcd. for C24
389.14, Found: 389.2, Purity: 98.2%.
þ
130.15, 128.25, 127.56, 123.46, 119.56, 118.36, 101.73, 85.15, 41.45.
H
16
N
6
[Mþ1] :
þ
LC-MS (m/z) calcd. for C23
Purity: 98.9%.
H
18
N
6
[Mþ1] : 379.16, Found: 379.3,
5.3. Biological investigation
5
.2.2.16. 4-((2-phenyl-6-((pyridin-4-ylmethyl)amino)pyrimidin-4-
yl)amino)benzonitrile (23). The title compound was synthesized
from 4-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile
86 mg, 0.28 mmol) and 4-picolylamine (861 mg, 7.96 mmol) as
described in the general procedure of method A. White solid
5.3.1. Cell culture
The MDCK II BCRP cell line transfected with ABCG2 and the
corresponding wild type cell line (MDCK II) were a generous gift by
Dr. A. Schinkel (The Netherlands Cancer Institute, Amsterdam, The
Netherlands). Cultivation was carried out in Dulbecco’s modified
Eagle’s medium (DMEM; Sigma Life Science, Steinheim, Germany)
supplemented with FCS (10%; PAN-Biotech GmbH, Aidenbach,
(
1
(
8
7
7
74.2 mg, 69.5%). H NMR (500 MHz, DMSO‑d
.52e8.50 (m, 2H), 8.25 (dd, J ¼ 3.0, 6.3 Hz, 2H), 7.90e7.87 (m, 2H),
.80 (t, J ¼ 6.1 Hz, 1H), 7.75e7.71 (m, 2H), 7.48e7.44 (m, 3H),
6
) d: 9.60 (s, 1H),
13
.38e7.35 (m, 2H), 5.89 (s, 1H), 4.63 (s, 2H). C NMR (126 MHz,
Germany),
Germany), penicillin G (50 U/ml; PAN-Biotech GmbH, Aidenbach,
Germany), and streptomycin (50 g/mL; PAN-Biotech GmbH,
Aidenbach, Germany). The human ovarian carcinoma cell line
A2780/ADR (ECACC no. 931125120) overexpressing ABCB1 and the
non-selected wild type cell line A2780 (ECACC no. 93112519) were
purchased from European Collection of Animal Cell Culture
(ECACC). RPMI-1640 medium with additional FCS (10%; PAN-
L-glutamine (2 mM; PAN-Biotech GmbH, Aidenbach,
DMSO) : 163.09, 162.28, 149.51, 145.48, 138.09, 133.07, 130.15,
d
1
28.23, 127.54, 122.09, 119.55, 118.37, 101.76, 42.98. LC-MS (m/z)
m
þ
calcd. for C23
H
18
N
6
[Mþ1] : 379.16, Found: 379.3, Purity: 98.5%.
5
.2.2.17. 4-((6-((2-hydroxyphenyl)amino)-2-phenylpyrimidin-4-yl)
amino)benzonitrile (24). The title compound was synthesized from
-((6-chloro-2-phenylpyrimidin-4-yl)amino)benzonitrile (102 mg,
4
15

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
Biotech GmbH, Aidenbach, Germany),
cillin G (50 U/ml), and streptomycin (50
L
-glutamine (2 mM), peni-
g/mL) was used for
cultivation. The small-cell lung cancer cell line H69AR (ATCC CRL-
1351) and the corresponding parental H69 cell line (NCL-H69,
ATCC HTB-119) were provided by American Type Culture Collection
ATCC). Cultivation was performed using RPMI-1640 medium
supplemented with FCS (20%; PAN-Biotech GmbH, Aidenbach,
Germany), -glutamine (2 mM), penicillin G (50 U/ml), and strep-
tomycin (50 g/mL). For subculturing or biological assays, over 90%
confluence was required in T75 cell culture flasks (Greiner Bio-One,
wavelength of 520 nm. The initial linear increase in fluorescence
was calculated as a slope between minutes 5 and 30 and compared
to the slope of cyclosporine A at a concentration of 10 mM. Only
candidates that reached more than 20% inhibition level were
further evaluated with complete concentration-response curves. If
the curve lacked the top values, the maximal effect of cyclospine A
m
1
(
(10 mM) was used to constrain the concentration-response curve.
L
Four-parameter logistic equation with variable hill slope or three-
parameter logistic equation with fixed slope (¼1), whatever was
statistically preferred, was used to generate concentration-
response curves and IC50 values were calculated with GraphPad
Prism (version 6.0 for Windows; San Diego, CA, USA).
m
Frickenhausen, Germany) stored under a 5% CO
2
-humidified at-
ꢀ
mosphere at 37 C in an incubator (Münchener Medizin Mechanik
GmbH, Planegg, Germany). The cell layer was rinsed three times
with phosphate-buffered saline (PAN-Biotech GmbH, Aidenbach,
Germany) and subsequently exposed to trypsin/EDTA solution
5.3.4. Cytotoxicity assay
Intrinsic cytotoxicity of the compounds was examined with an
MTT-based assay as described before with minor deviations
[15e17,40,41]. A dilution series of the compounds in the range of
(
0.05%/0.02%; PAN-Biotech GmbH, Aidenbach, Germany). Detached
cells were collected in a 50 mL falcon tube and centrifuged (266ꢄg;
ꢀ
4
C; 4 min). The supernatant was replaced by fresh medium and
3.16 mM to 1 mM was prepared using culture medium as a solvent.
the cell count was determined by CASY TT (Sch a€ rfe System GmbH,
Reutlingen, Germany) equipped with a 150 m capillary. Long-term
storage of cells was conducted in a medium/DMSO mixture (90%/
0%) under liquid nitrogen.
Cells were harvested as described before and a cell suspension was
prepared using MDCK II (2000 cells per well), MDCK II BCRP
(2000 cells per well), A2780 (8000 cells per well), A2780/ADR
m
1
(8000 cells per well), or H69AR (20,000 cells per well). To the 20
mL
of the compound dilution, 180 L of cell suspension were added to a
m
5
.3.2. Pheophorbide A accumulation assay to investigate ABCG2
inhibition
Pheophorbide A assay was applied in order to determine the
96-well tissue-culture treated plate (Starlab GmbH, Hamburg,
ꢀ
Germany) and incubated for 72 h at 37 C and 5% CO
2
. Culture
medium and DMSO defined 100% and 0% of cell viability, respec-
tively. At the end of the incubation period, 40 L of a 5 mg/mL MTT
inhibitory potency of the compounds toward ABCG2 as described
earlier with minor deviations [15e17,40,41]. First, the dilution se-
ries of the tested compound was prepared using concentrations in
the range of 0.316e100
a flat-bottom 96 well plate (Thermo Scientific, Waltham, MA, USA).
After harvesting the MDCK II BCRP cells as described earlier, 160
of cell suspension in KHB buffer containing approximately
m
solution were added to each well and incubated for another 60 min
ꢀ
(37 C, 5% CO
2
). After removing the supernatant, the formazan
L of DMSO and the absorbance was
m
M. These solutions were then pipetted into
crystals were solved in 100
m
measured applying a Multiscan microplate photometer (Thermo
Fischer Scientific, Waltham, MA, USA) at 570 nm with a background
correction at 690 nm. Concentration-effect curves were obtained
by plotting the absorbance values against logarithmic concentra-
tions of the test compounds in GraphPad Prism and generating a
curve on the basis of the four-parameter logistic equation with
variable hill slope.
mL
4
2
5,000 cells were added to each well. Following the addition of
0
mL of 5
m
M pheophorbide A solution, the plate was kept in the
ꢀ
incubator for 2 h at 37 C and 5% CO
2
. Afterward, flow cytometry
(
Guava easyCyte™ HT) was applied to measure fluorescence at
excitation wavelength of 488 nm and detection wavelength of 695/
0 nm. Compound 28 was used as reference and for the level of
5
5.3.5. MDR reversal assay
possible total inhibition. Four-parameter logistic equation with
variable hill slope or three-parameter logistic equation with fixed
slope (¼1), whatever was statistically preferred, was used to
generate concentration-response curves and IC50 values were
calculated with GraphPad Prism (version 6.0 for Windows; San
Diego, CA, USA).
The ability of the test compounds to reverse MDR was investi-
gated with a different version of a MTT assay as described before
with minor deviations [15e17,40,41]. Test compounds were diluted
with culture medium to necessary concentrations (depending on
the inhibitory activity in the respective cell line) and pipetted into a
96-well tissue-culture plate (Starlab GmbH, Hamburg, Germany).
Cells were harvested as described before and 160
pension were added to the plate at a concentration of 2000 (MDCK
II), 8000 (A2780), or 20,000 (H69) cells per well. Additionally, 20
of a dilution series (100 nMe100 M) of a cytotoxic drug (SN-38 for
mL of cell sus-
5
.3.3. Calcein AM accumulation assay to investigate ABCB1 and
ABCC1 inhibition
mL
Calcein AM assay was applied to investigate inhibitory activity of
the compounds toward ABCB1 and ABCC1 using A2780/ADR and
H69AR cell lines, respectively. The accumulation assay was per-
formed as described before with minor deviations [15e17,40,41].
Different dilutions of the test compounds (20
for the screening or between 0.316 M and 100
m
ABCG2, daunorubicin for ABCB1 and ABCC1) were pipetted into the
ꢀ
solution and incubated for 72 h (37 C, 5% CO
2
). Culture medium
and DMSO (10%) defined 100% and 0% cell viability, respectively.
After the incubation period, the procedure was the same as
described for the cytotoxicity MTT assay.
mL) either at 100
mM
m
mM for the deter-
mination of the IC50 value were pipetted into a clear F bottom 96-
well microplate (Greiner bio one, Frickenhausen, Germany). Cell
Declaration of competing interest
suspension was obtained as described above and 160
mL were
added to the plate with either 30,000 (ABCB1) or 60,000 (ABCC1)
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
cells per well. After an incubation period of 30 min under CO
2
-
ꢀ
humidified atmosphere (5%/37 C), 20
mL of 3.125
mM calcein AM
solution were added to the wells and the plate was immediately
measured in a Fluostar Optima or Polarstar microplate reader
Acknowledgements
(
BMG-Labtech, Software versions 2.00R2/2.20 and 4.11-0, respec-
tively). The fluorescence was recorded every 60 s for a time period
of 1 h at the excitation wavelength of 485 nm and emission
JL was supported by the China Scholarship Council (CSC). VN
thanks BioSolveIT GmbH (Germany) for providing the evaluation
16

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
license of LeadIT. SMS receives a Walter Benjamin fellowship of the
Deutsche Forschungsgemeinschaft (DFG, German Research Foun-
dation; STE2931/2).
[18] J. Yuan, I.L. Wong, T. Jiang, S.W. Wang, T. Liu, B.J. Wen, L.M. Chow, B. Wan
Sheng, Synthesis of methylated quercetin derivatives and their reversal ac-
tivities on P-gp- and BCRP-mediated multidrug resistance tumour cells, Eur. J.
Med.
Chem.
54
(2012)
413e422,
https://doi.org/10.1016/
j.ejmech.2012.05.026.
[
19] L. Cheung, C.L. Flemming, F. Watt, N. Masada, D.M.T. Yu, T. Huynh, G. Conseil,
A. Tivnan, A. Polinsky, A.V. Gudkov, M.A. Munoz, A. Vishvanath, D.M.F. Cooper,
M.J. Henderson, S.P.C. Cole, J.I. Fletcher, M. Haber, M.D. Norris, High-
throughput screening identifies ceefourin 1 and ceefourin 2 as highly selective
inhibitors of multidrug resistance protein 4 (MRP4), Biochem. Pharmacol. 91
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2020.113045.
(2014) 97e108, https://doi.org/10.1016/j.bcp.2014.05.023.
[
20] Y.F. Fan, W. Zhang, L. Zeng, Z.N. Lei, C.Y. Cai, P. Gupta, D.H. Yang, Q. Cui,
Z.D. Quin, Z.S. Chen, L.D. Trombetta, Dacomitinib antagonizes multidrug
resistance (MDR) in cancer cells by inhibiting the efflux activity of ABCB1 and
ABCG2 transporters, Canc. Lett. 421 (2018) 186e198, https://doi.org/10.1016/
j.canlet.2018.01.021.
References
[
[
[
1] L. Amiri-Kordestani, A. Basseville, K. Kurdziel, A.T. Jojo, S.E. Bates, Targeting
MDR in breast and lung cancer: discriminating its potential importance from
the failure of drug resistance reversal studies, Drug Resist. Updates 15 (2012)
[21] W. Zhang, Y.F. Fan, C.Y. Cai, J.Q. Wang, Q.X. Teng, Z.N. Lei, L. Zeng, P. Gupta,
Z.S. Chen, Olmutinib (BI1482694/HM61713), a novel epidermal growth factor
receptor tyrosine kinase inhibitor, reverses ABCG2-mediated multidrug
resistance in cancer cells, Front. Pharmacol. 9 (2018) 1097, https://doi.org/
10.3389/fphar.2018.01097.
[22] P. Gupta, Y.K. Zhang, X.Y. Zhang, Y.J. Wang, K.W. Lu, T. Hall, R. Peng, D.H. Yang,
N. Xie, Z.S. Chen, Voruciclib, a potent CDK4/6 inhibitor, antagonizes ABCB1
and ABCG2-mediated multi-drug resistance in cancer cells, Cell. Physiol.
Biochem. 45 (2018) 1515e1528, https://doi.org/10.1159/000487578.
[23] X. Zhu, I.L.K. Wong, K.F. Chan, J. Cui, M.C. Law, T.C. Chong, X. Hu, L.M.C. Chow,
T.H. Chan, Triazole bridged flavonoid dimers as potent, nontoxic, and highly
selective breast cancer resistance protein (BCRP/ABCG2) inhibitors, J. Med.
Chem. 62 (2019) 8578e8608, https://doi.org/10.1021/acs.jmedchem.9b00963.
[24] Z.X. Wu, Q.X. Teng, C.Y. Cai, J.Q. Wang, Z.N. Lei, Y. Yang, Y.F. Fan, J.Y. Zhang,
J. Li, Z.S. Chen, Tepotinib reverses ABCB1-mediated multidrug resistance in
cancer cells, Biochem. Pharmacol. 166 (2019) 120e127, https://doi.org/
10.1016/j.bcp.2019.05.015.
[25] L. Wang, L. Liang, T. Yang, Y. Qiao, Y. Xia, L. Liu, C. Li, P. Lu, X. Jjiang, A pilot
clinical study of apatinib plus irinotecan in patients with recurrents high-
grade glioma, Medicine (Baltim.) 96 (2017) 1e5, https://doi.org/10.1097/
MD.0000000000009053, e9053.
[26] M.K. Krapf, J. Gallus, V. Namasivayam, M. Wiese, 2,4,6-Substituted quinazo-
lines with extraordinary inhibitory potency toward ABCG2, J. Med. Chem. 61
(2018) 7952e7976, https://doi.org/10.1021/acs.jmedchem.8b01011.
[27] M.K. Krapf, M. Wiese, Synthesis and biological evaluation of 4-anilino-qui-
nazolines and equinolines as inhibitors of breast cancer resistance protein
(ABCG2), J. Med. Chem. 59 (2016) 5449e5461, https://doi.org/10.1021/
acs.jmedchem.6b00330.
50e61, https://doi.org/10.1016/j.drup.2012.02.002.
2] S.H. Lee, H. Kim, J.-H. Hwang, H.S. Lee, J.Y. Cho, Y.-S. Yoon, H.-S. Han, Breast
cancer resistance protein expression is associated with early recurrence and
decreased survival in resectable pancreatic cancer patients, Pathol. Int. 62
(
2012) 167e175, https://doi.org/10.1111/j.1440-1827.2011.02772.x.
3] D. Damiani, M. Tribelli, E. Calistri, A. Geromin, A. Chiarvesio, A. Michelutti, M.
Cavallin R. Fanin, The prognostic value of P-glycoprotein (ABCB) and breast
cancer resistance protein (ABCG2) in adults with de novo acute myeloid
leukemia with normal karyotype, Haematologica 91 (2006) 825e828, https://
doi.org/10.3324/%25x.
[
[
4] D. Steinbach, W. Sell, A. Voigt, J. Hermann, F. Zintl, A. Sauerbrey, BCRP gene
expression is associated with a poor response to remission induction therapy
in childhood acute myeloid leukemia, Leukemia 16 (2002) 1443e1447,
https://doi.org/10.1038/sj.leu.2402541.
5] K. Matsuo, M.L. Eno, E.H. Ahn, M.M. Shahzad, D.D. Im, N.B. Rosenein, A.K. Sood,
Multidrug resistance gene (MDR-1) and risk of brain metastasis in epithelial
ovarian, fallopian tube, and peritoneal cancer, Am. J. Clin. Oncol. 34 (2011)
488e493, https://doi.org/10.1097/COC.0b013e3181ec5f4b.
[
[
[
[
6] R.W. Robey, O. Polgar, J. Deeken, K.W. To, S.E. Bates, ABCG2: determining its
relevance in clinical drug resistance, Canc. Metastasis Rev. 26 (2007) 39e57,
https://doi.org/10.1007/s10555-007-9042-6.
7] G. Szakacs, J.K. Paterson, J.A. Ludwig, C. Booth-Genthe, M.M. Gottesmann,
Targeting multidrug resistance in cancer, Nat. Rev. Drug Discov. 5 (2006)
219e234, https://doi.org/10.1038/nrd1984.
8] Q. Mao, J.D. Unadkat, Role of the breast cancer resistance protein (BCRP/
ABCG2) in drug transport e an update, AAPS J. 17 (2015) 65e82, https://
doi.org/10.1208/s12248-014-9668-6.
9] C. Erlichman, S.A. Boerner, C.G. Hallgren, R. Spieker, X.-Y. Wang, C.D. James,
G.L. Scheffer, M. Maliepaard, D.D. Ross, K.C. Bible, S.H. Kaufmann, The HER
tyrosine kinase inhibitor CI1033 enhances cytotoxicity of 6-ethyl-10-
hydroxycamptrothecin and topotecan by inhibiting breast cancer resistance-
mediated drug efflux, Canc. Res. 61 (2001) 739e748.
[28] C. Ozvegy-Laczka, T. Hegedus, G. Varady, O. Ujhelly, J.D. Schuetz, A. Varadi,
G. Keri, L. Orfi, K. Nemet, B. Sarkadi, High-affinity interaction of tyrosine ki-
nase inhibitors with the ABCG2 multidrug transporter, Mol. Pharmacol. 65
(2004) 1485e1495, https://doi.org/10.1124/mol.65.6.1485.
[29] N.A. Colabufo, V. Pagliarulo, F. Berardi, M. Contino, C. Inglese, M. Niso,
P. Ancona, G. Albo, A. Pagliarulo, R. Perrone, Bicalutamide failure in prostate
cancer treatment: involvement of Multi Drug Resistance Proteins, Eur. J.
Pharmacol. 601 (2008) 38e42, https://doi.org/10.1016/j.ejphar.2008.10.038.
[30] N.A. Colabufo, F. Berardi, M.G. Perrone, M. Cantore, M. Contino, C. Inglese,
M. Niso, R. Perrone, Multi-drug-resistance-reverting agents: 2-aryloxazole
and 2-arylthiazole derivatives as potent BCRP and MRP1 inhibitors, Chem-
MedChem 4 (2009) 188e195, https://doi.org/10.1002/cmdc.200800329.
[31] C.L. Dai, A.K. Tiwari, C.P. Wu, X.D. Su, S.R. Wang, D.G. Liu, C.R. Ashby Jr.,
Y. Huang, R.W. Robey, Y.J. Liang, L.M. Chen, C.J. Shi, S.V. Ambudkar, Z.S. Chen,
L.W. Fu, Lapatinib (Tykerb, GW572016) reverses multidrug resistance in
cancer cells by inhibiting the activity of ATP-binding cassette subfamily B
member 1 and G member 2, Canc. Res. 68 (2008) 7905e7914, https://doi.org/
10.1158/0008-5472.CAN-08-0499.
[
[
[
10] K.K.W. To, D.C. Poon, Y. Wei, F. Wang, G. Lin, L. Fu, Pelitinib (EKB-569) targets
the up-regulation of ABCB1 and ABCG2 induced by hyperthermia to eradicate
lung cancer, Br. J. Pharmacol. 172 (2015) 4089e4106, https://doi.org/10.1111/
bph.13189.
11] J. Hu, X. Zhang, F. Wang, X. Wang, K. Yang, M. Xu, K.K.W. To, Q. Li, L. Fu, Effect
of ceritinib (LDK378) on enhancement of chemotherapeutic agents in ABCB1
and ABCG2 overexpressing cells in vitro and in vivo, Oncotarget 6 (2015)
44643e44659, https://doi.org/10.18632/oncotarget.5989.
12] Y.-J. Mi, Y.-J. Liang, H.-B. Huang, H.-Y. Zhao, C.-P. Wu, F. Wang, L.-Y. Tao, C.-
Z. Zhang, C.-L. Dai, A.K. Tiwari, X.-X. Ma, K.K.W. To, S.V. Ambudkar, Z.-S. Chen,
L.-W. Fu, Apatinib (YN968D1) reverses multidrug resistance by inhibiting the
efflux function of multiple ATP-binding cassette transporters, Canc. Res. 70
(
2010) 7980e7991, https://doi.org/10.1158/0008-5472.CAN-10-0111.
[32] S.L. Ma, Y.P. Hu, F. Wang, Z.C. Huang, Y.F. Chen, X.K. Wang, L.W. Fu, Lapatinib
antagonizes multidrug resistance-associated protein 1-mediated multidrug
resistance by inhibiting its transport function, Mol. Med. 20 (2014) 390e399,
https://doi.org/10.2119/molmed.2014.00059.
[33] K. Juvale, K. Stefan, M. Wiese, Synthesis and biological evaluation of flavones
and benzoflavones as inhibitors of BCRP/ABCG2, Eur. J. Med. Chem. 67 (2013)
115e126, https://doi.org/10.1016/j.ejmech.2013.06.035.
[34] A. Krauze, S. Grinberga, L. Krasnova, I. Adlere, E. Sokolova, I. Domracheva,
I. Shestakova, Z. Andzans, G. Duburs, Thieno[3,2-b]pyridinesea new class of
multidrug resistance (MDR) modulators, Bioorg. Med. Chem. 22 (2014)
5860e5870, https://doi.org/10.1016/j.bmc.2014.09.023.
[35] E. Teodori, M. Contino, C. Riganti, G. Bartolucci, L. Braconi, D. Manetti,
M.N. Manetti, A. Trezza, A. Athanasios, O. Spiga, M.G. Perrone, R. Giampietro,
E. Gazzano, M. Salerno, N.A. Colabufo, S. Dei, Design, synthesis and biological
evaluation of stereo- and regioisomers of amino aryl esters as multidrug
resistance (MDR) reversers, Eur. J. Med. Chem. 182 (2019) 111655, https://
doi.org/10.1016/j.ejmech.2019.111655.
[
13] C.-L. Dai, Y.-J. Liang, Y.-S. Wang, A.K. Tiwari, Y.-Y. Yan, F. Wang, Z.-S. Chen, X.-
Z. Tong, L.-W. Fu, Sensitization of ABCG2-overexpressing cells to conventional
chemotherapeutic agent by sunitinib was associated with inhibiting the
function of ABCG2, Canc. Lett. 279 (2009) 74e83, https://doi.org/10.1016/
j.canlet.2009.01.027.
14] K.K.W. To, D.C. Poon, Y. Wei, F. Wang, G. Lin, L.-W. Fu, Vatalanib sensitizes
ABCB1 and ABCG2-overexpressing multidrug resistant colon cancer cells to
chemotherapy under hypoxia, Biochem. Pharmacol. 97 (2015) 27e37, https://
doi.org/10.1016/j.bcp.2015.06.034.
[
[
[
15] K. Stefan, S.M. Schmitt, M. Wiese, 9-Deazapurines as broad-spectrum in-
hibitors of the ABC transport proteins P-glycoprotein, multidrug resistance-
associated protein 1, and breast cancer resistance protein, J. Med. Chem. 60
(
2017) 8758e8780, https://doi.org/10.1021/acs.jmedchem.7b00788.
16] K. Silbermann, S.M. Stefan, R. Elshawadfy, V. Namasivayam, M. Wiese, Iden-
tification of thienopyrimidine scaffold as inhibitor of the ABC transport pro-
tein ABCC1 (MRP1) and related transporters using
screening approach, J. Med. Chem. 62 (2019) 4383e4400, https://doi.org/
0.1021/acs.jmedchem.8b01821.
a combined virtual
[36] F. Antoni, D. Wifling, G. Bernhardt, Water-soluble inhibitors of ABCG2 e a
fragment-based and computational approach, Eur. J. Med. Chem. (2020),
https://doi.org/10.1016/j.emech.2020.112958. In press.
[37] M.K. Krapf, J. Gallus, M. Wiese, 4-Anilino-2-pyridylquinazolines and
epyrimidines as highly potent and nontoxic inhibitors of breast cancer
resistance protein (ABCG2), J. Med. Chem. 60 (2017) 4474e4495, https://
1
[
17] K. Silbermann, J. Li, V. Namasivayam, F. Baltes, G. Bendas, S.M. Stefan,
M. Wiese, Superior pyrimidine derivatives as selective ABCG2 inhibitors and
broad-spectrum ABCB1, ABCC1, and ABCG2 antagonists, J. Med. Chem. 63
(
2020) 10412e10432, https://doi.org/10.1021/acs.jmedchem.0c00961.
17

K. Silbermann, J. Li, V. Namasivayam et al.
European Journal of Medicinal Chemistry 212 (2021) 113045
doi.org/10.1021/acs.jmedchem.7b00441.
38] N.M.I. Taylor, I. Manolaridis, S.M. Jackson, J. Kowal, H. Stahlberg, K.P. Locher,
Structure of the human multidrug transporter ABCG2, Nature 546 (2017)
[48] A. Pawarode, S. Shukla, H. Minderman, S.M. Fricke, E.M. Pinder,
K.L. O’Loughlin, S.V. Ambudkar, M.R. Baer, Differential effects of the immu-
nosuppressive agents cyclosporine A, tacrolimus and sirolimus on drug
transport by multidrug resistance proteins, Canc. Chemother. Pharmacol. 60
(2007) 179e188, https://doi.org/10.1007/s00280-006-0357-8.
[49] I. Ivnitski-Steele, R.S. Larson, D.M. Lovato, H.M. Khawaja, S.S. Winter,
T.I. Oprea, L.A. Sklar, B.S. Edwards, High-throughput flow cytometry to detect
selective inhibitors of ABCB1, ABCC1, and ABCG2 transporters, Assay Drug
Dev. Technol. 6 (2008) 263e276, https://doi.org/10.1089/adt.2007.107.
[50] A. Pick, W. Klinkhammer, M. Wiese, Specific inhibitors of the breast cancer
resistance protein (BCRP), ChemMedChem 5 (2010) 1498e1505, https://
doi.org/10.1002/cmdc.201000216.
[
[
504e509, https://doi.org/10.1038/nature22345.
39] S.M. Jackson, I. Manolaridis, J. Kowal, M. Zechner, N.M.I. Taylor, M. Bause,
S. Bause, R. Bartholomaeus, G. Bernhardt, B. Koenig, A. Buschauer,
H. Stahlberg, K.H. Altmann, K.P. Locher, Structural basis of small-molecule
inhibition of human multidrug transporter ABCG2, Nat. Struct. Mol. Biol. 25
(
2018) 333e340, https://doi.org/10.1038/s41594-018-0049-1.
[
[
40] S.M. Schmitt, K. Stefan, M. Wiese, Pyrrolopyrimidine derivatives as novel in-
hibitors of multidrug resistance-associated protein 1, J. Med. Chem. 59 (2016)
3018e3033, https://doi.org/10.1021/acs.jmedchem.5b01644.
41] K. Silbermann, C.P. Shah, N.U. Sahu, K. Juvale, S.M. Stefan, P.S. Kharkar,
M. Wiese, Novel chalcone and flavone derivatives as selective and dual in-
hibitors of the transport proteins ABCB1 and ABCG2, Eur. J. Med. Chem. 164
[51] Y.-J. Mi, Y.-J. Liang, H.-B. Huang, H.-Y. Zhao, C.-P. Wu, F. Wang, L.-Y. Tao, C.-
Z. Zhang, C.-L. Dai, A.K. Tiwari, X.-X. Ma, K.K.W. To, S.V. Ambudkar, Z.S. Chen,
L.-W. Fu, Apatinib (YN968D1) reverses multidrug resistance by inhibiting the
efflux function of multiple ATP-binding cassette transporters, Canc. Res. 70
(2010) 7981e7991, https://doi.org/10.1158/0008-5472.CAN-10-0111.
[52] K. Juvale, J. Gallus, M. Wiese, Investigation of quinazolines as inhibitors of
breast cancer resistance proteins (ABCG2), Bioorg. Med. Chem. 21 (2013)
7858e7873, https://doi.org/10.1016/j.bmc.2013.10.007.
(
2019) 193e213, https://doi.org/10.1016/j.ejmech.2018.12.019.
[
42] M.K. Krapf, J. Gallus, M. Wiese, Synthesis and biological investigation of 2,4-
substituted quinazolines as highly potent inhibitors of breast cancer resis-
tance protein (ABCG2), Eur. J. Med. Chem. 139 (2017) 587e611, https://
doi.org/10.1016/j.ejmech.2017.08.020.
[
43] S.M. Stefan, Multi-target ABC transporter modulators: what next and where to
go? Future Med. Chem. 11 (2019) 2353e2358, https://doi.org/10.4155/fmc-
[53] T.J. Mathias, K. Natarjan, S. Shukla, K.A. Doshi, Z.N. Singh, S.V. Ambudkar,
M.R. Baer, The FLT3 and PDGFR inhibitor crenolanib is a substrate of the
multidrug resistance protein ABCB1 but not does not inhibit transport func-
tion at pharmacologically relevant concentrations, Invest. N. Drugs 33 (2015)
300e309, https://doi.org/10.1007/s10637-015-0205.
2019-0185.
[
44] A. Ahmed-Belkacem, A. Pozza, S. Macalou, J.M. Perez-Victoria, A. Boumendjel,
A. Di Pietro, Inhibitors of cancer cell multidrug resistance mediated by breast
cancer resistance protein (BCRP/ABCG2), Anti Canc. Drugs 17 (2006) 239e243,
https://doi.org/10.1097/00001813-200603000-00001.
[54] M.K. Krapf, J. Gallus, A. Spindler, M. Wiese, Synthesis and biological evaluation
of quinazoline derivatives e a SAR study of novel inhibitors of ABCG2, Eur. J.
[
[
[
45] M. Contino, S. Guglielmo, M.G. Perrone, R. Giampietro, B. Rolando, A> Carrieri,
D. Zaccaria, K. Chegaev, V. Borio, C. Riganti, K. Zabielska-Koczywas,
N.A. Colabufo, R. Fruttero, New tetrahydroisoquinoline-based P-glycoprotein
modulators: decoration of the biphenyl core gives selective ligands, Med.
Chem. Comm. 9 (2018) 862e869, https://doi.org/10.1039/c8md00075a.
46] S. Ranjbar, R. Khonkarn, A. Moreno, H. Baubichon-Sortay, R. Miri,
M. Khoshneviszadeh, L. Saso, N. Edraki, P. Falson, O. Firuzi, 5-Oxo-hexahy-
droquinioline derivatives as modulators of P-gp, MRP1 and BCRP transporters
to overcome multidrug resistance in cancer cells, Toxicol. Appl. Pharmacol.
Med.
Chem.
161
(2019)
506e525,
https://doi.org/10.1016/
j.ejmech.2018.10.026.
[55] F. Antoni, M. Bause, M. Scholler, S. Bauer, S.A. Stark, S.M. Jackson,
I. Manolaridis, K.P. Locher, B. K o€ nig, A. Buschauer, G. Bernhardt, Tariquidar-
related triazoles as potent, selective and stable inhibitors of ABCG2 (BCRP),
Eur. J. Med. Chem. 191 (2020) 112133, https://doi.org/10.1016/
j.ejmech.2020.112133.
[56] Molecular Operating Environment (MOE), 2018.01; Chemical Computing
Group ULC: 1010 Sherbooke St. West, Suite #910, Montreal, QC, Canada, H3A
2R7, 2018.
[57] LeadIT, version 2.3.2; BioSolveIT GmbH: Sankt Augustin, Germany, 2017.
[58] M. Rarey, B. Kramer, T. Lengauer, G. Klebe, A fast flexible docking method
using an incremental construction algorithm, J. Mol. Biol. 261 (1996)
470e489, https://doi.org/10.1006/jmbi.1996.0477.
362 (2019) 136e149, https://doi.org/10.1016/j.taap.2018.10.025.
47] K. Stefan, L.Y.W. Leck, V. Namasivayam, P. Bascu n~ ana, M.L.-H. Huang, P.J. Riss,
J. Pahnke, P.J. Jansson, S.M. Stefan, Vesicular ATP-binding cassette transporters
in human disease: relevant aspects of their organization for future drug
development, Future Drug Discov. 2 (2020) FDD51, https://doi.org/10.4155/
fdd-2020-0025.
18