2,4,6-Substituted Quinazolines with Extraordinary Inhibitory Potency toward ABCG2

Article abstract of DOI:10.1021/acs.jmedchem.8b01011

Several members of the ABC transporter superfamily play a decisive role in the development of multidrug resistance (MDR) in cancer. One of these MDR associated efflux transporters is ABCG2. One way to overcome this MDR is the coadministration of potent in

Full text of DOI:10.1021/acs.jmedchem.8b01011

Subscriber access provided by Kaohsiung Medical University
Article
2,4,6-Substituted Quinazolines With
Extraordinary Inhibitory Potency Toward ABCG2
Michael K Krapf, Jennifer Gallus, Vigneshwaran Namasivayam, and Michael Wiese
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01011 • Publication Date (Web): 03 Aug 2018
Downloaded from http://pubs.acs.org on August 4, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Abstract
Several members of the ABC transporter superfamily play a decisive role in the development
of multidrug resistance (MDR) in cancer. One of these MDR associated efflux transporters is
ABCG2. One way to overcome this MDR is the coꢀadministration of potent inhibitors of
ABCG2.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In this study we identified novel inhibitors containing a 2,4,6ꢀsubstituted quinazoline scaffold.
Introduction of a 6ꢀnitro function led to extraordinarily potent compounds, that were highly
selective for ABCG2 and also able to reverse the MDR toward the chemotherapeutic drugs
SNꢀ38 and mitoxantrone. The binding of substrate Hoechst 33342 and the two potent
inhibitors 31 and 41 which differ in their mechanism of inhibition was rationalized using the
recently published cryoꢀEM structures of ABCG2. For a better understanding of the
interaction between the inhibitors and ABCG2, additional investigations regarding the
ATPase activity, the interaction with Hoechst 33342 and with the conformational sensitive
5D3 antibody were carried out.
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Introduction
ATPꢀbinding cassette transport proteins are widely expressed in different tissues where they
perform the efflux of structurally diverse substrates out of the cells. The excretion is powered
by ATP hydrolysis, forming ADP and inorganic phosphate, and occurs even against a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
concentration gradient. In humans 48 ABC transport proteins are expressed which can be
classified according to their phylogenetic similarity in seven subfamilies labelled ABC A to
2
, 3, 4, 5
ABC G.
ABC transport proteins are often expressed in tissues with a barrier function,
where they protect the cells from harmful substances, but can also effectuate the membrane
6
transport of nutrient. Unfortunately, many cytostatic drugs are substrates of ABC transporters
and their overexpression in cancer cells persists as a major obstacle in the chemotherapeutic
7
treatment. This soꢀcalled multidrug resistance (MDR) is mainly related to the expression of
the three major MDRꢀassociated transport proteins Pꢀglycoprotein (Pꢀgp, ABCB1), multidrugꢀ
resistance associated protein 1 (MRP1, ABCC1) and breast cancer resistance protein (BCRP,
8
,9,10
ABCG2).
ABCG2 was discovered last when the workgroup of Doyle et al. identified the
transport protein in a human breast cancer cell line, which was selected for doxorubicin
1
1,12,13
resistance in the presence of verapamil, an inhibitor of Pꢀgp.
High levels of ABCG2 are
often found in tissues performing a protective function or tissues which are associated with
1
4
the homeostasis of nutrients and hormone. Interestingly, the structure of ABCG2 differs
from ABCB1 and ABCC1: being a half transporter, ABCG2 is assumed to form dimers or
tetramers for function while ABCB1 and ABCC1 are expressed in their fully functional form.
1
5 , 16 , 17 , 18
ABCG2 is composed of 655 amino acids and consists of only one cytosolic
1
9
nucleotideꢀbinding domain (NBD) and one transmembrane domain (TMD). To date the
function of ABCG2 remains unclear and further investigations are important to effectively
20,21,22
treat ABCG2 overexpressing cancers.
As mentioned before, a common problem in the
chemotherapeutic treatment of cancers is that many cytostatic drugs are substrates of ABCꢀ
transport proteins like ABCG2: the anthracene derivative mitoxantrone (MX) for instance is a
3
ACS Paragon Plus Environment

Page 3 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
substrate of ABCG2 and can be used to select cells with an overexpression of ABCG2, based
2
3,24
on the resistance toward the cytostatic drug.
Another substrate of ABCG2 is the cytostatic
2
5,26
drug SNꢀ38, which is an active metabolite of irinotecan.
Both, MX and SNꢀ38, have
already been used in our prior studies to investigate the reversal of the MDR in MDCK II
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
27
BCRP cells by coꢀadministering inhibitors of ABCG2.
Although coꢀadministration of potent inhibitors is one way to overcome MDR, the scientific
community has become more cautious in terms of in vivo application of inhibitors of ABCG2,
since the promising in vitro results with inhibitors of ABCB1 led to disappointing outcomes
in the clinic. Nevertheless, several studies found a correlation between the overexpression of
ABCG2 and a worse clinical outcome, while other studies remain sceptical toward that topic.
28,29,30
The first potent inhibitor of ABCG2 was isolated from Aspergillus fumigatus and named
fumitremorgin C. Despite its high inhibitory potency the high neurotoxic effects prevented
3
1
any clinical use. The second generation inhibitor 59 was more promising, since it exhibited
less toxic effects, a high selectivity toward ABCG2 up to 1 µM and an considerably high
3
2,33
inhibitory potency.
Its application in photodynamic therapy for instance has already led to
promising results, where the efflux of phototoxic agents could be inhibited by coꢀ
29,34
administration of compound 59 and resulted in a higher effectiveness of the therapy.
Other
inhibitors have already been used in cancer therapy. For instance some compounds in the
class of tyrosine kinase inhibitors (TKIs) were found to be useful drugs in chemotherapeutic
treatment. Preliminary in vitro investigations of the workgroup of OzvegyꢀLaczka et. al.
found a high affinity interaction of several TKIs with ABCG2, some even at submicromolar
35,36
concentration.
The TKI gefitinib for instance is currently used as a drug for certain breast,
lung and other cancers. However, only very few selective and potent inhibitors of ABCG2
3
7
active at a submicromolar concentration have been reported.
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Our workgroup modified the basic quinazoline scaffold of gefitinib and carried out different
modifications: in our previous studies we found that an aromatic residue at position 2 and a
substituted aniline linker at position 4 led to a considerably higher inhibitory potency and
27,56
selectivity toward ABCG2 compared to gefitinib.
We also identified some structural
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
features that have an impact on the selectivity of a substance: in particular the presence of
methoxy substituents was found to increase the inhibitory potency toward ABCB1 and also
ABCC1. In contrast, potent and selective compounds often contained nitro functions at the
aniline moiety at position 4 of the quinazoline scaffold.
In the current work, we started from our 2,4ꢀdisubstituted quinazoline scaffold and introduced
nitro, amino and some other groups at position 6. These new derivatives were compared to
some of our previous compounds providing new insights into the structureꢀactivity
relationship of these novel derivatives. Most notably, substitution with nitro at position 6
resulted in extraordinarily potent and selective inhibitors of ABCG2.
Results and Discussion
Chemistry.
The synthetic route for all compounds of this work is illustrated in Scheme 1. First, 2ꢀaminoꢀ
5ꢀnitrobenzoic acid was reacted with phosgene to form a cyclic anhydride. Subsequently, the
anhydride function underwent cleavage when bubbling gaseous ammonia through the reaction
mixture to form 2ꢀaminoꢀ5ꢀnitrobenzamide 1, the precursor for the following steps. The 2ꢀ
phenylquinazolinone derivatives 2ꢀ7 were obtained by cyclisation of 1 with a substituted
benzaldehyde. Substitution of the carbonyl function with chlorine was achieved by reaction
3
with POCl to yield the 4ꢀchloroquinazoline derivatives 8ꢀ13. A nucleophilic substitution of
chlorine at position 4 was carried out with different aniline derivatives using microwave
radiation to yield compounds 14ꢀ42. Compound 20, which contains a 3',4'ꢀdimethoxy
residue, was treated with boron tribromide to obtain a 3',4'ꢀdihydroxy function at the aniline
5
ACS Paragon Plus Environment

Page 5 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
linker (16). Moreover, the nitro function at position 6 was converted into an amino group.
Therefor, a Pd/C catalysed hydrogenation was performed in a pressure vessel. The obtained 6ꢀ
amino derivatives 43ꢀ47 were then reacted in a nucleophilic substitution with different acid
chlorides to yield compounds 48ꢀ51.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Based on our potent 2ꢀphenylꢀ4ꢀanilinoquinazoline derivatives, we carried out further
modifications at position 2 and most notably at position 6. This idea was inspired by the fact
that many inhibitors containing a quinazoline structure like gefitinib possess substitution at
position 6 and sometimes also at position 7. Therefore, we introduced nitro, amino and
aromatic functions at position 6 to elucidate the impact of these modifications on the
inhibitory activity toward ABCG2. In general, we preferred substituents, such as cyano, nitro,
27
fluoro and hydroxy, which exhibited high potencies at position 4 in our previous work.
There we found that high potencies were obtained in either meta or para position depending
on the function introduced at the anilino moiety: functions with negative mesomeric effects
were in most cases more beneficial in para position whereas substituents with positive
mesomeric effects mostly led to high potencies in meta position. Thanks to a facile
comparison with our compounds investigated in earlier studies, we were able to find
structureꢀactivity relations that could be beneficial for subsequent synthesis.
1
13
Identity of all compounds was confirmed by H and C NMR spectra and the purity by
elemental analysis.
Biological evaluation.
Hoechst 33342 Accumulation Assay. The inhibitory potency of the test compounds toward
ABCG2 was determined by the wellꢀestablished Hoechst 33342 accumulation assay. The
principle of the assay is based on the efflux of the fluorescent dye Hoechst 33342, a substrate
of ABCG2, out of the cell. Hoechst 33342 only shows low fluorescence intensities in watery
environment, but fluorescence increases considerably upon binding to DNA or embedding in
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
8
a lipophilic environment like a cell membrane. Inhibition of the transport protein results in
an increased intracellular accumulation of Hoechst 33342 that leads to higher fluorescence
which can be correlated to the inhibitory potency of a compound. Structural formulas of
Hoechst 33342 and the other substrates and standard inhibitors of ABCG2 are given in Figure
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
for reference.
Inhibitory activities of the test compounds 14ꢀ58 from the Hoechst 33342 accumulation assay
are presented in Table 1. A quick overview illustrating the SAR of the investigated
compounds is provided at the end of this chapter (Figure 2).
3
Compounds 14ꢀ42 all contain a nitro group at R . At first, the unsubstituted basic structure 14
1
2
with phenyl at R and hydrogen at R was investigated. Interestingly, the resulting IC50 value
of 106 nM is considerably lower than its analogue 52 (IC50 = 882 nM) lacking the nitro
3
2
function at R . Subsequently, polar groups were investigated at R using a 3'ꢀhydroxy (15)
and 3',4'ꢀdihydroxy (16) substitution, of which the monoꢀsubstitution yielded a low IC50
value of 65.3 nM and the disubstitution an IC50 value of 119 nM. Even higher activities
2
resulted by a substitution with 3ꢀmethoxy (17) and 3',4'ꢀdimethoxy (18) at R , with
remarkably low IC50 values of 33.1 nM and 37.0 nM. In contrast, the wellꢀknown and highly
potent standard inhibitor 59 yielded a higher IC50 of 227 nM. In comparison to the 3',4'ꢀ
3
dimethoxy quinazoline analogue 53, lacking the nitro function at R , compound 18 is about
2
fourꢀfold more potent. Derivatives with 3'ꢀcyano (19) or 4'ꢀcyano (20) at R possessed IC50
values of 48.2 nM and 48.1 nM, respectively. The 2ꢀphenylquinazoline analog 54, which
showed the highest activity in our earlier study yielded in a slightly higher IC50 of 69.9 nM.
Similar activities resulted from a 3'ꢀfluoro (21: 51.5 nM) and a 3'ꢀtrifluoromethyl (22: 49.1
3
nM) substituent. Again, the 3'ꢀfluoro analogue 55, lacking the nitro function at R , showed a
2
considerably lower potency with an IC50 value of 355 nM. Also the 3'ꢀnitroꢀ4'ꢀhydroxy (R )
6ꢀnitro derivative 23 was more potent (IC = 37.4 nM) than its analogue 56 (IC = 80.0 nM).
5
0
50
Also compound 24 containing an acetamido group (IC50 = 78.7 nM) was significantly more
7
ACS Paragon Plus Environment

Page 7 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potent than its analog 57 without 6ꢀnitro group (IC = 278 nM). Surprisingly, compound 25
50
with a carboxyl residue led to an IC in the submicromolar range (904 nM), although it forms
50
an anionic species under physiological conditions decreasing the membrane permeability of
this derivative. Comparison with its analog 58 (IC50 = 6380 nM) once again illustrates the
beneficial impact of the 6ꢀnitro group.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
For the next compounds, the phenyl residue at R was replaced by a more hydrophilic pyridyl
residue. First, a 3ꢀpyridyl residue was combined with 3'ꢀmethoxy (26: IC50 = 47 nM) and 4'ꢀ
2
methoxy (27: IC50 = 57.7 nM) at R , respectively. In comparison to compound 27 the 4ꢀ
pyridyl analog 28 led to a comparable potency (IC50 = 75.5 nM). Also, the interchange of
phenyl with pyridyl had no major impact on the IC50 value, illustrated in the compound pairs
17/26 and 19/29.
1
In the next series a 3ꢀtrifluoromethylphenyl substituent was introduced at R . Interchange of
the trifluoromethyl group between the aromatic cores at position 2 and 4 resulted in similar
high potencies in the range of 50ꢀ60 nM (compare compounds 22 and 30). Extraordinarily
2
high inhibitory activities resulted from substitutions at R with 3'ꢀhydroxy (compound 31,
IC50 = 27.6 nM), 3',4'ꢀdimethoxy (33, IC50 = 30.2 nM), 3'ꢀfluoro (compound 35, IC50 = 44.8
nM), 3'ꢀnitroꢀ4'ꢀhydroxy (compound 38, IC50 = 33.0 nM) and 3'ꢀacetamido (compound 39,
IC50 = 44.8 nM). Substitution with 3'ꢀmethoxy or 3'ꢀthiomethyl as well as 3'ꢀ or 4'ꢀnitro
resulted in good activities between 60 and 100 nM (compare compounds 32, 34, 36 and 37).
Except for compound 32 all 3ꢀtrifluoromethylphenyl derivatives possessed a similar or even
higher potency than their corresponding phenyl analogs.
1
Additionally, two 3ꢀmethoxyphenyl (R ) derivatives were synthesized containing a 4'ꢀcyano
2
(
40) or 3'ꢀfluoro substituent (41) at R . Surprisingly, both derivatives yielded very high
inhibitory activities with IC values of 27.8 nM for compound 40 and 23.4 nM for 41, which
50
is the most potent compound in the test set. Compared to their phenyl analogs 20 and 21 they
are roughly twoꢀfold more potent, and compound 41 shows a nineꢀfold higher activity than
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the standard inhibitor compound 59. Concentrationꢀresponse curves of 40, 41 and 59 are
1
2
depicted in Figure 3. Moreover, the 3,4ꢀdimethoxyphenyl (R ), 3'ꢀtrifluoromethyl at R
derivative 42 yielded an IC value of 57.2 nM. Unfortunately, the 3,4ꢀdimethoxy substitution
5
0
did not increase the potency in comparison to compound 22, which contains a phenyl group at
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
R instead.
In the next series, the 6ꢀnitro functions of some derivatives were reduced to an amino group.
Free amino groups act as Hꢀdonor functions and increase the electron density at the attached
aromatic cores, whereas nitro groups are Hꢀacceptor functions and decrease the electron
density.
1
First, a 3ꢀpyridyl residue at R was combined with either a 3'ꢀmethoxy (43) or a 4'ꢀmethoxy
2
group (44) at R . The 3'ꢀmethoxy derivative yielded an IC50 value of 369 nM whereas the
para substituted compound had a twofold higher IC50 value of 646 nM. Three more 6ꢀamino
1
derivatives were investigated containing a 3ꢀtrifluoromethylphenyl residue at R : compound
2
45 with a no substituent at R possessed an IC50 of 726 nM, whereas substitution with 3'ꢀ
hydroxy and 3'ꢀmethoxy resulted in IC50 values of 799 nM and 835 nM for compounds 46
and 47, respectively. The comparison with their 6ꢀnitro analogs illustrates that the amino
function at position 6 lowers the inhibitory potency of all compounds to a considerable extent.
3
Subsequently, the amino function at R was modified by introducing aromatic groups and one
1
2
ester function. Initially, a substitution with 3ꢀpyridyl at R , 3'ꢀmethoxy at R together with Nꢀ
3
1ꢀiminoethyl acetate at R was carried out for compound 48. The resulting IC50 value of 5310
nM illustrates the strong loss of potency in comparison to compound 43 with the free amine at
3
R . Also, substitution with Nꢀ3ꢀnitrobenzamide (49: IC = 1150 nM) and Nꢀnicotinamide (51:
50
IC = 1040 nM) resulted in a decrease of inhibitory potency in comparison to compound 43.
5
0
2
Also the 4ꢀmethoxy (R ) analog of compound 49, namely compound 50, yielded a far higher
IC50 value of 1620 nM compared to its 3ꢀmethoxy counterpart.
9
ACS Paragon Plus Environment

Page 9 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In summary, the investigation of our new 2ꢀsubstitutedꢀ4ꢀanilinoquinazoline derivatives
containing additional substitution at position 6 revealed several findings: Extraordinary high
3
inhibitory activity toward ABCG2 resulted from nitro functions at R . Most notably, the
presence of methoxy and hydroxy groups led to very high potencies, as observed for
compounds 17, 18, 23, 31, 33, 38, 40 and 41. The unsubstituted compound 14 showed a lower
inhibitory activity than the substituted 6ꢀnitro derivatives, which is in good accordance to our
previous results for the 4ꢀsubstitutedꢀ2ꢀphenylquinazolines (compare compounds 52ꢀ58). The
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
most potent compound 41 containing a 3ꢀmethoxyphenyl group at R and 3'ꢀfluoro
2
substituent at R possessed a nearly tenꢀfold higher potency than compound 59, one of the
most potent inhibitors of ABCG2 found in literature to date. Reduction of the nitro function at
3
R resulting in an amine, decreased of the inhibitory activity significantly. Further reaction of
the free amines with acid chlorides yielded substituted imines and amides. Unfortunately, this
modification led to further reduction of the inhibitory activity.
Screening for ABCB1 and ABCC1 Inhibition. The selectivity toward ABCG2 was
investigated by determining the inhibitory activity of some of the most potent compounds in
the calcein AM assay using ABCB1 and ABCC1 overexpressing cell lines A2780adr and
H69AR, respectively. Initial screening was carried out at a compound concentration of 10 µM
measuring the increase of fluorescence resulting from intracellular calcein produced by the
cleavage of the calcein AM esters over time. The standard inhibitor cyclosporine A (CsA) was
used as positive control for both cell lines defining the state of “full inhibition”. All tested
compounds showed an inhibitory activity of less than 25% of the standard CsA and thus
demonstrated a high selectivity toward ABCG2.
Determination of the Intrinsic Cytotoxicity via MTT Assay. Determination of the intrinsic
cytotoxicity was carried out with parental and ABCG2 overexpressing MDCK II cells which
were incubated with different dilutions of selected compounds. Cell viability was measured
10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
after 72 h of incubation using 3ꢀ(4,5ꢀdimethylthiazolꢀ2ꢀyl)ꢀ2,5ꢀdiphenyltetrazolium bromide
(MTT) to detect the metabolic activity of the remaining viable cells. Reduction of MTT in
living cells leads to the formation of a non waterꢀsoluble formazan species which can be
detected colorimetrically. Further details are described in the experimental section. The half
maximal growth inhibition (GI50) is listed in Table 2 and representative concentration effect
curves are shown in Figure 4. The therapeutic ratio (TR = GI /IC ) was calculated from the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
50
GI value of a compound obtained with ABCG2 overexpressing cells.
50
1
2
For the unsubstituted basic structure 14, containing a phenyl group at R , hydrogen at R and
3
nitro at R , a GI50 of 18.3 µM and a TR of 173 was found. For the analogue 52, lacking the
3
nitro group at R , a GI50 of 26.6 µM and a TR of 30.1 resulted. Despite of the increased
toxicity caused by the 6ꢀnitro substitution, the benefit of the high inhibitory potency of
compound 14 becomes apparent yielding a TR that is more than threeꢀfold higher than that of
1
2
compound 59 (TR = 55.5). Compound 17 with phenyl group at R , 3'ꢀmethoxy at R and
3
nitro at R displayed a pronounced cytotoxicity of 5.50 µM but still accomplished a TR of 166.
2
Compound 18, which is analogous to 17, but contains 3',4'ꢀdimethoxy at R , yielded a
similar GI50 of 8.13 µM with a calculated TR of 220. In comparison, compound 53 without
3
substitution with nitro at R had a lower TR of 148. Another disubstituted derivative was
1
2
3
investigated containing a phenyl group at R , 3'ꢀnitroꢀ4'ꢀhydroxy at R and nitro at R
(
compound 23). Due to its high inhibitory potency together with a GI of 13.2 µM a TR of
50
353 was found. Its analog 56, without 6ꢀnitro group, possessed a TR of 655, which was the
27
highest ratio observed in our previous study. Interestingly, the 2ꢀpyridyl derivatives 28 and
9 showed almost no toxic effects with excellent GI50 values of 84.5 µM and 77.3 µM,
respectively. This in accordance to our findings in the investigation of the cytotoxic effects of
ꢀanilinoꢀ2ꢀpyridylquinazolines, where the pyridyl group was highly preferable over a phenyl
2
4
3
9
group. Moreover, for compound 29 the highest TR of 1212 among the investigated
compounds was found.
11
ACS Paragon Plus Environment

Page 11 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
3
Compounds 31, 33 and 38 all contain a 3ꢀtrifluoromethylphenyl residue at R and nitro at R .
Unfortunately, low GI50 values around 3.5 µM resulted for 31 and 33 with 3'ꢀhydroxy and
3
',4'ꢀdimethoxy substituents, respectively. The 3'ꢀnitroꢀ4'ꢀhydroxy derivative possessed a
considerably higher GI50 value of 27.4 µM resulting in a high TR of 830. In our previous
study, the benefit of a 3ꢀnitroꢀ4ꢀhydroxy substitution was already observed, illustrated by
compound 52. Also, the highly potent compounds 40 and 41, both containing 3ꢀ
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
3
methoxyphenyl at R and 3'ꢀnitro at R , were investigated. Compound 40 with 4'ꢀcyano
2
substituent at R yielded a GI50 value of 11.0 µM resulting in a TR of 395, whereas the most
2
potent compound 41 (3'ꢀfluoro at R ) unfortunately possessed a low GI50 value of 4.38 µM
1
2
giving a TR of 187. Finally, compound 44 with a 3ꢀpyridyl residue at R , 4'ꢀmethoxy at R
3
and amino at R yielded a GI50 of 19.1 µM and giving the lowest TR of 29.6, which is due to a
relatively high cytotoxicity along with a relatively low inhibitory potency.
3
Despite of the increased cytotoxicity caused by the nitro function in R , this modification
shows its advantage in the context of inhibitory potency: the very low IC50 values found for
the investigated 6ꢀnitro derivatives led to high therapeutic ratios. The highest TR was found
for compounds 28 and 29 containing a pyridyl group at position 2. It is notable, that a pyridyl
substituent at position 2 led to only negligible intrinsic cytotoxicity but retained high
inhibitory potency resulting in a remarkable TR.
Determination of the Ability to Reverse MDR via MTT Assay.
The ability to reverse the MDR toward the cytostatic drug SNꢀ38 was investigated for
compounds 23, 28, 29, 38 and 41. Different dilutions of the test compounds were added to the
cells inhibiting the efflux of the cytostatic drug to a different extent. Inhibition of the transport
protein therefore leads to a higher intracellular concentration of SNꢀ38 and thus to a decreased
cell viability. As positive control served MDCK II parental cells, representing the state of
“full inhibition” due to the lack of ABCG2 expression.
12
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Compounds 23 (IC = 37.4 nM), 28 (IC = 75.5 nM), 29 (IC = 63.8 nM), 38 (IC = 33.0
50
50
50
50
nM) and 41 (IC = 23.4 nM) were investigated in coꢀadministration with the cytostatic drug
50
SNꢀ38. They were chosen due to their high inhibitory potency in the Hoechst 33342
accumulation assay (compounds 23, 38 and 41) as well as their considerably low cytotoxic
effects in the cell viability assay (compounds 28 and 29). The shift of the pGI50 values toward
the sensitive cells with increasing compound concentrations is depicted in Figure 5 a), c) and
d) for selected compounds, indicated by the grey arrows. Full reversal of the MDR was
achieved by all compounds at a concentration of about 1 µM obtaining approximately the
same pGI50 value as the parental cells (∆). To preclude unwanted transporter unspecific
effects of the compound, an analogous assay was carried out using parental cells (Figure 5b).
The resulting pGI values for each compound were all comparable giving no indication of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
unspecific effects. Based on the data derived from the graphs presented in Figure 5, the power
of reversal of the compounds was calculated by plotting each pGI50 value of the
corresponding concentrationꢀviability curves against the logarithm of the compound
concentration (Figure 6). Since the parental cells show no resistance against SNꢀ38, they are
expected to show the highest pEC values independently from the compound concentration
5
0
with a low deviation among each concentration (○). In contrast, the BCRP overexpressing
cells show a resistance dependent on the degree of inhibition by the compound. ABCG2
overexpressing cells without compound should obtain the lowest pEC50 value, increasing
simultaneously with an increasing compound concentration (●). From the doseꢀeffect plot it
can be concluded, that all compounds reach a plateau at the same level as the sensitive cells.
The EC50 values were calculated by nonlinear regression using the logistic equation to
generate concentrationꢀeffect curves. By this method, for compound 23 an EC50 value of 34.0
nM, for 28 an EC value of 71.2 nM, for 29 an EC value of 48.3 nM for 38 an EC value
5
0
50
50
of 57.0 nM and for 41 an EC50 value of 51.1 nM were calculated.
13
ACS Paragon Plus Environment

Page 13 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A second MDR reversal assay was carried out coꢀadministering the cytostatic drug
mitoxantrone (MX), which is also a substrate of ABCG2, with the two most potent
compounds in the test set, namely 31 and 41. The assay was carried out similar to the assay
with SNꢀ38, but using only MDCK II ABCG2 overexpressing cells. Different compound
concentrations were investigated in the presence of 0.5 µM MX as well as in the absence of
MX. The cell viability after an incubation period of 72 h is depicted in Figure 7 as bar chart.
The control was performed using only cells without modulator in the absence and presence of
MX. Increasing compound concentrations are expected to lead to an increasing sensitization
of the cells toward MX, resulting in a decrease of the cell viability. At a concentration of 1
µM compounds 31 and 41 lead to full reversal of the MDR (Figure 7, light grey bars).
Analogously EC values were derived for both compounds resulting in EC values of 21.8
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
50
nM and 19.5 nM for compound 31 and 41, respectively. Cytotoxic effects of both compounds
in the absence of MX can be observed by the decrease of the cell viability with increasing
compound concentration (Figure 7, dark grey bars). However, cytotoxic effects are only
noticeable at concentrations ≥ 1 µM. Both compounds show an excellent correlation between
the EC50 resulting from the MDR reversal assay with MX and the Hoechst 33342
accumulation assay.
Investigation of ATPase activity. ATPase activity was measured using High Five insect cell
membranes, which have previously been infected by baculovirus containing ABCG2 cDNA.
The released inorganic phosphate, which is closely related to the transport activity of ABCG2
is detected by a colorimetric method. More details are provided in the experimental section.
Despite being an inhibitor of ABCG2, synthesized compounds could also be transported
substrates of ABCG2 and therefore stimulate ATPase activity. In earlier studies of our
working group we showed that closely related compounds displayed different behaviour in
ATPase activity and enzyme kinetic investigations. To investigate the behaviour of the 6ꢀnitro
14
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
derivatives we carried out vanadateꢀsensitive ATPase activity studies with highly potent
compounds in Hoechst 33342 inhibition assay and compounds displaying structural
similarities. The investigated compounds can be divided into three groups with different
behaviour in ATPase activity assay, as illustrated in Table 3. Eight of the tested compounds
had no observable effect on the ATPase activity in all tested concentrations (14, 17, 21, 31, 32,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
37, 41, and 44). This may be an indication that these compounds are not transported by
ABCG2 and represent only inhibitors of ABCG2 detected in Hoechst 33342 accumulation
assay. Another eight compounds showed a stimulation of ATPase activity, from which two
compounds (26 and 47) possessed a strong activating effect to nearly the same extent as the
standard activator quercetin. All other compounds were moderately stimulating ranging
between 35 % and 52%. With regard to the potency of activation, five compounds had EC50
values, in the low nanomolar range (18, 20, 28, 30, 40), while one compound (19) showed an
EC50 value in the range of quercetin and two compounds (26, 47) possessed even higher
values in the low micromolar range. Interestingly the two compounds with the highest
activation contained both a meta methoxy substituent at the phenyl ring at position 4 of the
quinazoline scaffold, but different substituents at positions 2 and 6.
Four compounds (22, 27, 33, 43) showed a bellꢀshaped concentrationꢀresponse curve with an
activating effect at low concentrations and an inhibiting effect at higher concentrations.
Inspection of the concentration of the maximal activity revealed one compound (33) in the
low nanomolar range, one compound (43) with a concentration seven times higher and two
compounds (22, 27), with an about 30 times higher maximal concentration in the micromolar
range. Inspection of the maximal stimulation (Vmax) caused by the biphasic compounds
showed that two compounds possessed a negligible effect of less than 30% (27, 43). The other
two compounds were ranging between 30% and 40% at the maximum. All biphasic
compounds were not stimulating the ATPase activity very well. With the exception of
compounds 26 and 47, all investigated compounds showed low ATPase activity stimulation
15
ACS Paragon Plus Environment

Page 15 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
or even no observable effect and can therefore be seen as very good inhibitors of ABCG2
with nearly no impact on ATPase activity. This is probably due to the nitro function at
position 6 of the quinazoline scaffold as most quinazolines investigate so far stimulated
ATPase activity. This modification leads to higher inhibitory potency in the Hoechst 33342
accumulation assay and no impact on ATPase activity.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Representative concentrationꢀresponse curves of compounds 22 (biphasic), 26 (stimulating)
and quercetin (activating standard) are illustrated in Figure 8. Compound 26 showed solely
stimulation of ATPase activity over the whole range of concentrations similar to the standard
activator quercetin. This activating effect on ATPase activity may point to substrate properties
of the compound despite the inhibitory activity in the Hoechst 33342 accumulation assay.
However, in earlier studies of our working group, the assumption of transport by ABCG2 in
relation to ATPase activity stimulation, was challenged by a fluorescent heterodimeric
ABCG2 inhibitor comprised of a quinazoline and chalcone moiety, which was also
stimulating ATPase activity to a high extent. In confocal microscopy studies, this compound
accumulated to the same extent in transfected MDCK II BCRP cells as in parenteral MDCK
4
0
cells without the transporter protein, pointing to absence of transport. For compound 22 a
bellꢀshaped ATPase activity curve was observed. The occurrence of such bellꢀshaped curves
is presumed to be due to two different binding sites for substrates. In most cases it is supposed
that the activating and in some cases also transporting site has a high affinity for substrates
4
1,42
and the inhibitory site has a low affinity for substrates.
The bellꢀshaped behaviour in
ATPase activity can be due to the assumption that different binding pockets for one
compound exist and the different behaviour of closely related structures with small
differences in their substitution pattern lead to the proposal that multiple binding sites in
ABCG2 protein must be present. This phenomenon can further be investigated by enzyme
kinetic studies, which is illustrated in the following chapter.
16
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Investigation of Interaction Type. Investigation of the interaction with the ABCG2 substrate
Hoechst 33342 was carried out with compounds 17, 31, 40 and 41 that were selected
according to their different ATPase profiles and their inhibitory potency in the Hoechst 33342
accumulation assay. For this purpose the ABCG2 overexpressing MDCK II cell line was used.
Representative LineweaverꢀBurk and CornishꢀBowden diagrams are depicted in Figure 9
together with a summary of the results from the LineweaverꢀBurk diagrams in Table 4.
With exception of compound 31, all other investigated compounds showed a nonꢀcompetitive
interaction with Hoechst 33342 according to the results of the LineweaverꢀBurk method.
Interestingly, all nonꢀcompetitive interactions are of the “mixedꢀtype”, illustrated by the
intersection of the lines in the second or third quadrant of the coordinate system. Although
such “mixedꢀtype” inhibitors bind differently from Hoechst 33342, they influence the active
site, for instance by conformational changes of the protein. Compound 31 showed a
competitive interaction with Hoechst 33342 for the same binding site, according to the
intersection of the lines at the Yꢀaxis of the plot. Representative LineweaverꢀBurk plots for
compounds 17, 31, 40 and 41 are illustrated in Figure 9 a, c, e and g. To confirm our results
the direct linear plot according to CornishꢀBowden was carried out, since this method has
been claimed to be less sensitive to outliers (Figure 9 b, d, f, h): a strong decrease of the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
maximum velocity Vmax and roughly constant K
1 with increasing concentrations, which is characteristic for a nonꢀcompetitive interaction. In
contrast, compound 31 displayed constant Vmax values and increasing K values, which is
M
values were found for compounds 17 and
4
M
characteristic for a competitive interaction. Simultaneously decreasing Vmax and increasing
K values confirmed the “mixedꢀtype” properties of compound 40. In our previous studies as
M
well as in our ongoing work we discovered that a nonꢀcompetitive interaction with Hoechst
39
3
3342 is the predominant mechanism for the quinazoline derivatives. However, some
substituents led to a rather competitive interaction with the substrate, which could be
17
ACS Paragon Plus Environment

Page 17 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
explained by the different physicochemical properties of the compounds and for some by the
ability to form hydrogen bonds via Hꢀdonor and Hꢀacceptorꢀfunctions.
5D3 antibody binding assay. To get further insights into the interaction between the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
compounds and ABCG2 we performed a binding assay with conformation sensitive antibody
TM
5
D3 for selected derivatives. Therefor the monoclonal antibody PerCPꢀCy 5.5 Mouse Antiꢀ
Human CD338 was used with the ABCG2 overexpressing PLBꢀ985 cell line. The antibody is
conjugated with a fluorescent dye and binds specifically to the CD338 epitope of ABCG2,
dependent on the conformational state of the protein. We used the standard ABCG2 inhibitor
5
9 at a concentration of 10 µM as a reference due to its significant effect on antibody binding
and set it as 100 % labelling of ABCG2. Furthermore, we used the ABCG2 substrate Hoechst
3342 for comparison, since it is known that substrates in most cases lead to a lower rate of
3
4
3
antibody binding than inhibitors. The unspecific binding of the antibody was taken into
account by subtracting the fluorescence detected with an isotypeꢀcontrol.
We selected our most active compounds together with some 6ꢀamino derivatives and
conducted a screening at a compound concentration of 10 µM. The results are presented in
Figure 10. The highest amount of labelling with the antibody was found for compound 17
(88%) which showed a nonꢀcompetitive interaction with Hoechst 33342 in the interaction
type studies. The lowest immunostaining with 5D3 antibody among the tested compounds
was found for the 6ꢀamino derivative 47 (47%), which is still higher than the labelling for the
substrate Hoechst 33342 (35%). Although the differences in labelling among the remaining
compounds were rather low, some tendencies were observable: compounds 43 and 44 which
already exhibited different IC50 values, showed a different rate of labelling with the antibody
of 72% and 84%, respectively. Compounds 31, 33, 40 and 41 gave a similar immunostaining
of 66%, 58%, 62%, and 64%, respectively. There was no correlation between the IC50 value
18
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of a compound and the rate of labelling with the antibody. However, the high rate of labelling
with the antibody suggests that the compounds are no substrates of ABCG2.
Molecular docking
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Docking of substrate Hoechst 33342. Very recently the cryoꢀEM structure of human
ABCG2 has been reported and additionally in complex with a structural homolog of the
4
4,45
standard inhibitor 59.
Additionally a thorough discussion of the results of mutagenesis
4
6
studies has been published. Therefore we performed docking studies of selected compounds
in order to gain insight into the molecular determinants involved in the formation of the
receptorꢀsubstrate complex. First the substrate Hoechst 33342 was docked into the cryoꢀEM
structure of the human ABCG2 and the interactions with the receptor were studied in detail.
The results from docking studies show that putative binding modes of Hoechst 33342 occupy
two binding sites S1 and S2 as depicted in Figure 11a. The site S1 is located at the interface
between the two transmembrane domains and the site S2 is between the transmembrane
regions. As shown in Figure 11b, 64% of docked poses of Hoechst 33242 occupied site S1
and the remaining 36% of poses were occupying site S2. The previously published literature
specified that the site S1 is highly promiscuous allowing substrates and nonꢀsubstrates to dock
4
7
in a wide variety of binding poses. On contrary, the site S2 is highly selective and favoring
substrates in comparison to nonꢀsubstrates or inhibitors. The new cryoꢀEM structure (Jackson
45
et. al) specified that the site S1 is occupied by the fumitremorgin Cꢀrelated inhibitor, MZ29.
According to interaction type analysis we had found a nonꢀcompetitive mixedꢀtype interaction
of compound 59 and Hoechst 33342. Thus, we considered the binding poses of Hoechst
3
3342 docked at the site S2 for our analysis.
The 2D interaction diagram of the putative binding pose of Hoechst 33342 is shown in Figure
1c. Hoechst 33342 has many functional groups that could potentially interact with the
receptor. The nitrogen in the methylpiperazine might acts as a donor and possibly interacts
1
19
ACS Paragon Plus Environment

Page 19 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with the main chain of the residue Gly179. The benzimidazole group attached to the
methylpiperazine is positioned at the edge of the transmembrane domain and either N1 or N3
of the benzimidazole group act as donor and acceptor and interact with Asn387. The amino
acid Ans387 is from connecting helix the interface between the transmembrane and
nucleotide binding domains and plays an important role for the efflux function of the receptor.
The finding was confirmed by the negative effects of N397A mutant on receptor efflux
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
8
activity with Hoechst 33342 as substrate. The mutagenesis study also suggests that the
positioning of Asn387 at the interface could be vital for communication between the two
domains. The second benzimidazole group is surrounded by hydrophobic residues Leu338,
Leu447, Val450 and polar residues Asn387, Asn391, Glu451 and Asp477. The ethoxyphenyl
group is located in a subpocket formed by Gln398, Ser443, Glu446, Met481, and Arg482.
Among the residues, the side chains of Gln398, Ser443 and Arg482 interact with the
ethoxyphenyl group of Hoechst 33342. The results from different mutagenesis studies
specified that the overall efflux activity was decreased for the mutants Q398A, R482G and
48, 49, 50 , 51
R482T.
The mutants also show a decreasing effect on the ATPase activity. An
aromatic residue Tyr489 located at a distance of approximately 6 Å might act as an edge for
the binding site and position the substrate Hoechst 33342 inside the binding pocket. The
mutagenesis studies by Tamura et al. show decrease in efflux and ATPase activity for the
5
2
F489L mutant. This indicates that an aromatic residue is required for strong hydrophobic
interaction between Hoechst 33342 and the amino acids inside the substrate binding pocket .
Docking of inhibitors. To explore binding modes for the quinazoline derivatives and identify
a possible explanation for the mechanism of inhibition, we selected compounds 31 and 41 for
performing molecular docking studies in the cryoꢀEM structure of the ABCG2 receptor.
Compound 31 was selected as it showed competitive type of interaction with Hoechst 33342
20
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
while Compound 41 was the most potent compound showing nonꢀcompetitive type of
interaction (table 4).
Similar to Hoechst 33342, compound 31 can be accommodated in the binding site S1 and S2
(Figure 12a). In our docking results, only 6% of poses of compound 31 occupied the site S1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and the remaining 94% of poses were occupied site S2. In particular, the best scoring binding
poses were bound near the connecting helix between the transmembrane and nucleotide
binding domain. The 2D interaction diagram of putative binding pose with the receptor is
shown in Figure 12b. The quinazoline moiety of 31 is placed below the ionic lock formed
3
between Asp127 and Lys453. The 3ꢀCF ꢀphenyl group is located in the subpocket formed by
the residues Val449, Val450, Lys452, Lys453 and Ser532. The NH of the 3ꢀhydroxyaniline
group possibly forms interaction with the main chain of Asp128. In particular, the hydroxyl
substituent at position 3 of the aniline group acts as hydrogen bond donor and acceptor which
forms strong interaction with the main chain of Val450 and Leu454. Additionally, the nitro
group at position 6 of the quinazoline moiety forms electrostatic interactions with the side
chain of Arg191 and the main chain of Asp128. The obtained docked poses suggest that
compound 31 possibly binds at the site S1 which is closer to the substrate banding site and
exhibits competitive type of inhibition.
As a next step, we analyzed the binding poses of compound 41 which displayed nonꢀ
competitive and mixed type of inhibition. The putative binding poses and the interaction with
the receptor are shown in Figure 12c and d. The results from docking show that all 100
binding poses were docked to the site S2. Similar to the bound inhibitor, MZ29 in the cryoꢀ
EM structure it occupied the binding site S2 interface between the two transmembrane
domains. The quinazoline moiety forms a strong aromatic πꢀπ interaction with Phe439. The
nitro group substituted at position 6 of the quinazoline moiety forms electrostatic interactions
with Asn436. The NHꢀgroup of the 3ꢀfluoroꢀaniline group forms hydrogen bond interaction
with Met549. The phenyl ring is positioned in a subpocket formed by the hydrophobic
21
ACS Paragon Plus Environment

Page 21 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
residues Phe431, Phe432, Thr435, Met549 and Leu555. The 3ꢀmethoxyphenyl group forms a
πꢀstacking with another aromatic residue Phe439. Based on the analysis of binding poses and
interaction from docking studies, compound 41 occupies the site S2 and shows nonꢀ
competitive or mixedꢀtype of interaction with the substrate Hoechst 33242.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Conclusions
This study is a continuation of our previous investigations of 2,4ꢀdisubstitutedꢀquinazolines.
Extraordinary low IC50 values resulted from introducing a nitro group at position 6 of a 2,4ꢀ
disubstituted quinazoline scaffold. Further substitution with methoxy residues at the aromatic
rings at position 2 and 4 of the scaffold increased the potency toward ABCG2 even more. The
highest potency in the Hoechst 33342 accumulation assay resulted for compound 41 from a
substitution with 3ꢀmethoxyphenyl at position 2 and 3'ꢀfluoroaniline at position 4. It was
determined to be about 10ꢀfold more potent than compound 59, one of the most potent
inhibitors of ABCG2 found in literature. Reduction of the nitro function at position 6 yielded
derivatives with an amino group with significantly lower potencies than the corresponding 6ꢀ
nitro derivatives. Reaction of the amino function at position 6 with acid chlorides yielded
aromatic amides with unfortunately even further decreased inhibitory potency. We concluded
that the nitro function is preferable over amino groups and amido groups.
Moreover, the screening of selected compounds in a calcein AM assay against the transport
proteins ABCB1 and ABCC1 showed their high selectivity toward ABCG2. Resulting
activities toward both transport proteins were below 25% of the response observed by the
control cyclosporine A confirming the specificity of the inhibitors. Investigation of the
intrinsic cytotoxicity showed slightly increased toxic effects for the 6ꢀnitro derivatives
compared to their analogs lacking the nitro group. Generally aromatic nitro groups are
considered as toxicologically problematic. Most nitroꢀcontaining compounds can cause
22
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
3
methaemoglobinemia and are potentially mutagenic. However, a prominent exception from
this rule are calcium channel blockers of the dihydropyridine type. Therefore additional
toxicological studies are necessary to explore the feasibility of future in vivo studies of the
ABCG2 inhibitors. The calculated therapeutic ratios (GI50/IC50) still were relatively high due
to the high inhibitory potencies of the compounds. Most notably a substitution with pyridyl at
position 2 turned out to be highly beneficial (see compounds 28 and 29), leading to a
therapeutic ratio which is up to 22ꢀfold higher than that of compound 59. Moreover, most
potent compounds were investigated for their ability to reverse MDR toward the cytostatic
drugs SNꢀ38 and mitoxantrone using ABCG2 overexpressing MDCK II BCRP cells. All
tested compounds were able to reverse the MDR restoring the sensitivity toward SNꢀ38 and
mitoxantrone at low compound concentrations. Full reversal of the MDR was achieved by all
compounds at compound concentrations of about 1 µM. Investigation of the ATPase activity
of selected compounds resulted in three different groups: more than one third of the
compounds showed no effect on ATPase activity. This may be an indication that these
compounds are not transported by ABCG2 and represent only inhibitors of ABCG2 as
detected in Hoechst 33342 accumulation assay. Four compounds showed biphasic ATPase
activity curves with an activating/stimulating site at low concentrations and an inhibitory site
at higher concentrations (see compounds 22, 27, 33 and 43). It is reasonable that different
binding regions with different affinities are present in ABCG2. The third group consists of
eight compounds with stimulation of ATPase activity, of which almost all showed low
stimulation with exception of compounds 26 and 47. These two compounds both contain a
meta methoxy group at the phenyl ring at position 4 of the quinazoline scaffold and showed
the same stimulating effect as the standard stimulator quercetin. The nitro substitution at
position 6 of the quinazoline scaffold seems to lead to higher inhibitory potencies in the
Hoechst 33342 accumulation assay and leads to low or no impact on ATPase activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
ACS Paragon Plus Environment

Page 23 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Furthermore, the type of interaction with Hoechst 33342 was determined for selected
compounds using the methods according to LineweaverꢀBurk and CornishꢀBowden.
Interestingly, the investigated compounds showed different types of interaction with Hoechst
33342 but were mostly nonꢀcompetitive, meaning that they bind apart from Hoechst 33342.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Some compounds showed competitive portions, which is characteristic for “mixedꢀtype”
inhibitors or pure competitive inhibitors. Here, one compound (compound 31) was found to
compete for the same binding site as Hoechst 33342.
Additionally the conformation sensitive 5D3 antibody binding assay was carried out to
monitor the conformational change of ABCG2 in the presence of selected compounds. All
compounds induced a higher level of immunostaining with the antibody than the ABCG2
substrate Hoechst 33342. Only small differences in the rate of labelling resulted for the
investigated compounds, whereas no correlation was found with the IC50 values from the
Hoechst 33342 accumulation assay.
As a logical continuation of our earlier work, the quinazoline scaffold was further modified
and functionalized giving more insights into the SAR of this compound class. The study
yielded several compounds with extraordinary high inhibitory potencies and selectivity
toward ABCG2. Based on the promising in vitro results, particularly their ability to reverse
MDR toward SNꢀ38 and MX, further in vivo studies should be interesting. Furthermore,
molecular modelling studies based on recently published cryoꢀEM structures of ABCG2
supported the results of the substrate and inhibitors obtained by biological and kinetic studies.
Experimental Section
Chemistry. Materials. Chemicals were purchased from Acros Organics (Geel, Belgium),
Alfa Aesar (Karlsruhe, Germany), Sigma Aldrich (Steinheim, Germany) or Merck (Darmstadt,
Germany) and used without further purification. Microwave reactions were carried out in 50
mL vials using a CEM Discover SP (CEM GmbH, KampꢀLintfort, Germany) and monitored
24
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
by thin layer chromatography (TLC) employing an aluminium plate coated with silica gel 60
F254 (Merck Millipore, Billerica, MA, USA). For the preparation of the TLC a mixture of
dichloromethane and methanol in the ratio of 9:1 was used as eluent. Following, the plates
were analyzed in a UV cabinet at an excitation wavelength of 254 nm. Identity of the test
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
13
compounds was confirmed by NMR with the Hꢀ and C spectra obtained on a Bruker
Advance 500 MHz (500/126 MHz) or Bruker Advance 600 MHz (600/151 MHz),
respectively. Chemical shifts (δ) relating to the internal standard DMSOꢀd6, also used as
1
3
solvent for the compounds, are expressed in ppm. For the assignment of the C signals
distortionless enhancement by polarization transfer (DEPT) as well as attached proton test
(APT) was applied. Multiplicity of the signals are given as singlet (s), doublet (d), doublet of
doublets (dd), triplet of doublets (td), triplet (t), doublet of triplets (dt), quartet (q) and
multiplet (m) whereas the coupling constants J are given in Hz. The purity of all biologically
evaluated compounds was determined to be >95% by elemental analysis recorded on a Vario
EL V24 CHN Elemental Analyzer (Elementar Analysensysteme GmbH, Hanau, Germany).
Found values were in the range of ±0.4% of the theoretical values, unless indicated
2
ꢀaminoꢀ5ꢀnitrobenzamide (1).
To a solution of 2ꢀaminoꢀ5ꢀnitrobenzoic acid (1.80 g, 10 mmol) in 100 mL dry
tetrahydrofuran was slowly added triphosgene (1.19 g, 4 mmol) while stirring at room
temperature. The mixture was then refluxed under moisture exclusion for 3 h. Under stirring,
a constant stream of ammonia gas was bubbled through the mixture for 30 min at 50 °C. Then,
50 mL of water was added and the precipitate filtered off. Additional product was obtained by
concentrating the mixture under reduced pressure to one half of the initial volume to yield 1
1
as bright yellow crystals (1.58 g, 87%). H NMR (500 MHz, DMSOꢀd
6
) δ 12.97 (s, 1H), 8.84
(d, J = 2.7 Hz, 1H), 8.56 (dd, J = 9.0, 2.7 Hz, 1H), 8.27 – 8.19 (m, 2H), 7.92 (d, J = 9.0 Hz,
13
1
H), 7.68 – 7.62 (m, 1H), 7.58 (dd, J = 8.3, 6.7 Hz, 2H). C NMR (126 MHz, DMSO) δ
25
ACS Paragon Plus Environment

Page 25 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
60.74, 155.16, 152.43, 144.86, 132.40, 128.86, 128.65, 128.24, 127.99, 126.66, 122.18,
121.22.
General Procedure for the Preparation of compounds 2ꢀ7. A mixture of 2ꢀaminoꢀ5ꢀ
nitrobenzamide 1 (3.62 g, 20 mmol), the corresponding aldehyde derivative (20 mmol), iodine
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(3.17 g, 25 mmol), anhydrous potassium carbonate (2.76 g, 20 mmol) and 20 ml DMF was
stirred at 90ꢀ110 °C for 8ꢀ24 h. The end of the reaction was monitored by TLC and the
mixture poured on crushed ice to form a precipitate. Incomplete precipitation can be
prevented by adjusting the pH with concentrated HCl solution to 7. After filtration of the
precipitate, it was thoroughly washed with 100 mL of a 20% sodium thiosulfate solution
followed by 100 mL of hot distilled water. Purification was performed by recrystallization
from ethanol.
6ꢀnitroꢀ2ꢀphenylquinazolinꢀ4(3H)ꢀone (2). The compound was synthesized from 1 (3.62 g,
20 mmol) and benzaldehyde (2.12 g, 20 mmol) as described in the general procedure for
1
compounds 2ꢀ7 to yield 2 as a yellow solid (3.85 g, 72%). H NMR (500 MHz, DMSOꢀd ) δ
6
1
7
2.97 (s, 1H), 8.84 (d, J = 2.7 Hz, 1H), 8.56 (dd, J = 9.0, 2.7 Hz, 1H), 8.27 – 8.19 (m, 2H),
13
.92 (d, J = 9.0 Hz, 1H), 7.68 – 7.62 (m, 1H), 7.58 (dd, J = 8.3, 6.7 Hz, 2H). C NMR (126
MHz, DMSO) δ 144.86, 132.40, 132.24, 128.86, 128.65, 128.39, 122.18.
6ꢀnitroꢀ2ꢀ(pyridinꢀ3ꢀyl)quinazolinꢀ4(3H)ꢀone (3). The compound was synthesized from 1
(3.62 g, 20 mmol) and nicotinaldehyde (2.14 g, 20 mmol) as described in the general
1
procedure for compounds 2ꢀ7 to yield 3 as a yellowꢀorange solid (3.70 g, 69%). H NMR (500
MHz, DMSOꢀd
6
) δ 9.58 – 9.44 (m, 1H), 8.82 (d, J = 2.8 Hz, 1H), 8.67 (dt, J = 7.9, 2.0 Hz,
1
9
H), 8.61 (dd, J = 4.9, 1.7 Hz, 1H), 8.24 (dd, J = 9.0, 2.8 Hz, 1H), 7.94 (s, 1H), 7.58 (d, J =
13
.0 Hz, 1H), 7.45 (ddd, J = 7.9, 4.7, 0.8 Hz, 1H). C NMR (126 MHz, DMSO) δ 171.52,
26
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
63.89, 162.42, 156.83, 150.48, 149.70, 141.89, 135.45, 135.21, 127.35, 125.06, 123.24,
120.82.
6ꢀnitroꢀ2ꢀ(pyridinꢀ4ꢀyl)quinazolinꢀ4(3H)ꢀone (4). The compound was synthesized from 1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(3.62 g, 20 mmol) and isonicotinaldehyde (2.14 g, 20 mmol) as described in the general
1
procedure for compounds 2ꢀ7 to yield 4 as a yellow solid (2.74 g, 51%). H NMR (500 MHz,
DMSOꢀd ) δ 13.18 (s, 1H), 8.82 (dd, J = 8.1, 4.1 Hz, 3H), 8.57 (dd, J = 9.2, 2.7 Hz, 1H), 8.16
6
1
3
– 8.07 (m, 2H), 7.95 (d, J = 9.3 Hz, 1H). C NMR (126 MHz, DMSO) δ 161.62, 154.18,
152.54, 150.51, 145.35, 139.58, 129.56, 128.73, 122.09, 121.96, 121.77.
6ꢀnitroꢀ2ꢀ(3ꢀ(trifluoromethyl)phenyl)quinazolinꢀ4(3H)ꢀone (5). The compound was
synthesized from 1 (3.62 g, 20 mmol) and 3ꢀ(trifluoromethyl)benzaldehyde (3.48 g, 20 mmol)
as described in the general procedure for compounds 2ꢀ7 to yield 5 as a yellow solid (5.16 g,
1
7
7%). H NMR (500 MHz, DMSOꢀd
6
) δ 13.18 (s, 1H), 8.82 (d, J = 2.7 Hz, 1H), 8.59 – 8.53
(m, 2H), 8.50 (ddd, J = 8.2, 2.0, 1.1 Hz, 1H), 8.00 (ddt, J = 7.9, 1.9, 1.0 Hz, 1H), 7.97 – 7.91
13
(m, 1H), 7.87 – 7.77 (m, 1H). C NMR (126 MHz, DMSO) δ 161.67, 154.51, 152.69, 145.07,
1
33.26, 132.33, 130.07, 129.61 (d, J = 32.3 Hz), 129.37, 128.71 (d, J = 3.5 Hz), 128.65,
27.25, 125.04 (d, J = 3.8 Hz), 122.92, 122.07, 121.36.
1
2ꢀ(3ꢀmethoxyphenyl)ꢀ6ꢀnitroquinazolinꢀ4(3H)ꢀone (6). The compound was synthesized
from 1 (3.62 g, 20 mmol) and 3ꢀmethoxybenzaldehyde (2.72 g, 20 mmol) as described in the
1
general procedure for compounds 2ꢀ7 to yield 6 as a yellow solid (4.88 g, 82%). H NMR
(500 MHz, DMSOꢀd
6
) δ 12.94 (s, 1H), 8.83 (d, J = 2.7 Hz, 1H), 8.52 (dd, J = 9.0, 2.8 Hz, 1H),
7.88 (d, J = 8.9 Hz, 1H), 7.85 (ddd, J = 7.8, 1.7, 0.9 Hz, 1H), 7.80 (t, J = 2.2 Hz, 1H), 7.47 (t,
1
3
J = 8.0 Hz, 1H), 7.18 (ddd, J = 8.2, 2.6, 0.9 Hz, 1H), 3.86 (s, 3H). C NMR (126 MHz,
27
ACS Paragon Plus Environment

Page 27 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DMSO) δ 161.65, 161.42, 159.51, 153.03, 144.84, 142.84, 133.43, 130.00, 128.61, 122.13,
120.75, 118.60, 113.11, 55.59.
2ꢀ(3,4ꢀdimethoxyphenyl)ꢀ6ꢀnitroquinazolinꢀ4(3H)ꢀone (7). The compound was synthesized
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
from 1 (3.62 g, 20 mmol) and 3,4ꢀdimethoxybenzaldehyde (3.32 g, 20 mmol) as described in
1
the general procedure for compounds 2ꢀ7 to yield 7 as a yellow solid (4.19 g, 64%). H NMR
(
500 MHz, DMSOꢀd ) δ 12.81 (s, 1H), 8.80 (d, J = 2.7 Hz, 1H), 8.50 (dd, J = 9.0, 2.8 Hz, 1H),
6
7.94 (dd, J = 8.5, 2.2 Hz, 1H), 7.88 – 7.82 (m, 2H), 7.13 (d, J = 8.7 Hz, 1H), 3.88 (s, 3H),
13
3.85 (s, 3H). C NMR (126 MHz, DMSO) δ 162.24, 155.65, 153.41, 152.49, 148.72, 144.28,
28.86, 128.42, 124.35, 122.24, 122.18, 120.68, 111.58, 111.20, 55.89, 55.87.
1
General Procedure for the Preparation of compounds 8ꢀ13. The corresponding 2ꢀ
substituted 6ꢀnitroquinazolinꢀ4(3H)ꢀone derivative 2ꢀ7 (10 mmol) was added to phosphorous
oxychloride (30 mL, 0.32 mol) and stirred for 10 min at room temperature. The mixture was
then refluxed for 4ꢀ12 h and the reaction monitored by TLC. After completion of the reaction,
excess POCl was removed under reduced pressure and 50 mL ice water added. Subsequently,
3
5
2
0 mL DCM was added while stirring and the pH of the mixture slowly adjusted to 7 with
5% ammonium solution. With a separatory funnel, the organic phase was collected, washed
4
with 50 mL brine and dried over MgSO . The solvent was removed under reduced pressure
and the obtained solid recrystallized from isopropanol.
3ꢀ(6ꢀnitroꢀ4ꢀoxoꢀ3,4ꢀdihydroquinazolinꢀ2ꢀyl)benzonitrile (8). The compound was
synthesized from 2 (2.67 g, 10 mmol) as described in the general procedure for compounds 8ꢀ
1
1
3 to yield 8 as a yellow solid (2.46 g, 86%). H NMR (500 MHz, DMSOꢀd
6
) δ 8.83 (d, J =
2.7 Hz, 1H), 8.55 (dd, J = 9.0, 2.8 Hz, 1H), 8.24 – 8.19 (m, 2H), 7.91 (d, J = 9.0 Hz, 1H), 7.67
28
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 93
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
3
–
7.62 (m, 1H), 7.60 – 7.55 (m, 2H). C NMR (126 MHz, DMSO) δ 161.74, 155.85, 152.97,
144.85, 132.41, 132.16, 129.19, 128.85, 128.64, 128.39, 122.16, 121.14.
4
ꢀchloroꢀ6ꢀnitroꢀ2ꢀ(pyridinꢀ3ꢀyl)quinazoline (9). The compound was synthesized from 3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
2.68 g, 10 mmol) as described in the general procedure for compounds 8ꢀ13 to yield 9 as an
1
orange solid (2.38 g, 83%). H NMR (500 MHz, DMSOꢀd ) δ 9.38 (dd, J = 2.3, 0.8 Hz, 1H),
6
8
.88 (dd, J = 5.0, 1.6 Hz, 1H), 8.82 (d, J = 2.7 Hz, 1H), 8.72 (ddd, J = 8.1, 2.3, 1.6 Hz, 1H),
13
8.60 – 8.54 (m, 1H), 7.94 (d, J = 8.8 Hz, 1H), 7.79 (ddd, J = 8.1, 5.1, 0.8 Hz, 1H). C NMR
(126 MHz, DMSO) δ 161.49, 153.56, 152.52, 150.29, 147.22, 145.21, 138.43, 129.36, 129.20,
128.77, 124.70, 122.09, 121.42.
4
ꢀchloroꢀ6ꢀnitroꢀ2ꢀ(pyridinꢀ4ꢀyl)quinazoline (10). The compound was synthesized from 4
(
2.68 g, 10 mmol) as described in the general procedure for compounds 8ꢀ13 to yield 10 as an
1
orange solid (2.12 g, 74%). H NMR (500 MHz, DMSOꢀd
6
) δ 13.21 (s, 1H), 8.85 – 8.81 (m,
1
3
2H), 8.57 (dd, J = 9.0, 2.7 Hz, 1H), 8.18 – 8.10 (m, 2H), 7.95 (d, J = 9.0 Hz, 1H). C NMR
(
126 MHz, DMSO) δ 161.55, 153.98, 152.46, 150.07, 145.44, 140.03, 129.63, 128.80, 122.10,
121.81.
4
ꢀchloroꢀ6ꢀnitroꢀ2ꢀ(3ꢀ(trifluoromethyl)phenyl)quinazoline (11). The compound was
synthesized from 5 (3.35 g, 10 mmol) as described in the general procedure for compounds 8ꢀ
1
13 to yield 11 as a yellow solid (3.18 g, 90%). H NMR (600 MHz, DMSOꢀd ) δ 8.81 (d, J =
6
2
.7 Hz, 1H), 8.59 – 8.52 (m, 2H), 8.52 – 8.46 (m, 1H), 8.03 – 7.97 (m, 1H), 7.94 (d, J = 8.9
13
Hz, 1H), 7.82 (t, J = 7.9 Hz, 1H). C NMR (151 MHz, DMSO) δ 161.68, 154.53, 152.67,
1
45.14, 133.24, 132.40, 130.14, 129.66 (d, J = 32.5 Hz), 129.42, 128.80 (d, J = 3.8 Hz),
28.74, 125.09 (t, J = 3.9 Hz), 123.16, 122.13, 121.40.
1
29
ACS Paragon Plus Environment

Page 29 of 93
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
ꢀchloroꢀ2ꢀ(3ꢀmethoxyphenyl)ꢀ6ꢀnitroquinazoline (12). The compound was synthesized
from 6 (2.97 g, 10 mmol) as described in the general procedure for compounds 8ꢀ13 to yield
1
1
8
7
3
1
2 as a yellow solid (2.81 g, 89%). H NMR (500 MHz, DMSOꢀd
6
) δ 8.83 (d, J = 2.7 Hz, 1H),
.55 (dd, J = 9.0, 2.8 Hz, 1H), 7.92 (d, J = 9.0 Hz, 1H), 7.83 (ddd, J = 7.7, 1.7, 0.9 Hz, 1H),
.78 (dd, J = 2.6, 1.7 Hz, 1H), 7.49 (t, J = 8.0 Hz, 1H), 7.20 (ddd, J = 8.2, 2.6, 0.9 Hz, 1H),
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
13
.87 (s, 3H). C NMR (126 MHz, DMSO) δ 161.74, 159.51, 155.56, 152.89, 144.87, 133.42,
30.02, 129.23, 128.65, 122.15, 121.18, 120.77, 118.63, 113.14, 55.61.
4ꢀchloroꢀ2ꢀ(3,4ꢀdimethoxyphenyl)ꢀ6ꢀnitroquinazoline (13). The compound was synthesized
from 7 (3.27 g, 10 mmol) as described in the general procedure for compounds 8ꢀ13 to yield
1
1
8
8
3 as a yellow solid (2.75 g, 71%). H NMR (500 MHz, DMSOꢀd ) δ 8.79 (d, J = 2.7 Hz, 1H),
6
.50 (dd, J = 9.0, 2.7 Hz, 1H), 7.92 (dd, J = 8.5, 2.2 Hz, 1H), 7.89 – 7.77 (m, 2H), 7.13 (d, J =
1
3
.6 Hz, 1H), 3.88 (s, 3H), 3.85 (s, 3H). C NMR (126 MHz, DMSO) δ 161.80, 155.29,
153.05, 152.61, 148.73, 144.43, 128.80, 128.59, 123.85, 122.26, 122.19, 120.66, 111.60,
111.20, 55.91, 55.89.
General Procedure for the Preparation of compounds 14ꢀ42. The corresponding 2
substituted 4ꢀchloroquinazoline 8ꢀ13 (1 mmol) and a substituted aniline (1 mmol) were added
to a 50 mL microwave tube and suspended in 25 mL isopropanol. The tube was sealed and the
reaction mixture stirred under 100 watt microwave irradiation at 110 °C for 20ꢀ40 min.
Completion of the reaction was monitored by TLC. After cooling, a precipitate was formed
and filtered off by suction. Recrystallization was carried out from ethanol.
6ꢀnitroꢀN,2ꢀdiphenylquinazolinꢀ4ꢀamine (14). The compound was synthesized according to
the general procedure for compounds 14ꢀ42 from 8 (2.86 g, 10 mmol) and aniline (0.93 g, 10
1
mmol) to yield 14 as an orange solid (2.40 g, 70%), mp 264ꢀ265 °C. H NMR (500 MHz,
30
ACS Paragon Plus Environment