Discovery of Nonquinone Substrates for NAD(P)H: Quinone Oxidoreductase 1 (NQO1) as Effective Intracellular ROS Generators for the Treatment of Drug-Resistant Non-Small-Cell Lung Cancer

Article abstract of DOI:10.1021/acs.jmedchem.8b01424

The elevation of oxidative stress preferentially in cancer cells by efficient NQO1 substrates, which promote ROS generation through redox cycling, has emerged as an effective strategy for cancer therapy, even for treating drug-resistant cancers. Here, we described the identification and structural optimization studies of the hit compound 1, a new chemotype of nonquinone substrate for NQO1 as an efficient ROS generator. Further structure-activity relationship studies resulted in the most active compound 20k, a tricyclic 2,3-dicyano indenopyrazinone, which selectively inhibited the proliferation of NQO1-overexpressing A549 and A549/Taxol cancer cells. Furthermore, 20k dramatically elevated the intracellular ROS levels through NQO1-catalyzed redox cycling and induced PARP-1-mediated cell apoptosis in A549/Taxol cells. In addition, 20k significantly suppressed the growth of A549/Taxol xenograft tumors in mice with no apparent toxicity observed in vivo. Together, 20k acts as an efficient NQO1 substrate and may be a new option for the treatment of NQO1-overexpresssing drug-resistant NSCLC.

Full text of DOI:10.1021/acs.jmedchem.8b01424

Subscriber access provided by Gothenburg University Library
Article
Discovery of Non-Quinone Substrates for NAD(P)H: Quinone
Oxidoreductase 1 (NQO1) as Effective Intracellular ROS Generators
for the Treatment of Drug-Resistant Non-Small Cell Lung Cancer
Xingsen Wu, Xiang Li, Zhihong Li, Yancheng Yu, Qi-Dong You, and Xiaojin Zhang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01424 • Publication Date (Web): 03 Dec 2018
Downloaded from http://pubs.acs.org on December 3, 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 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Non-Quinone Substrates for
NAD(P)H: Quinone Oxidoreductase 1 (NQO1) as
Effective Intracellular ROS Generators for the
Treatment of Drug-Resistant Non-Small Cell Lung
Cancer
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
Xingsen Wu, †,‡ Xiang Li, †,§ Zhihong Li, †,‡ Yancheng Yu, †,‡ Qidong You, *,† Xiaojin Zhang *,†,‡
†
State Key Laboratory of Natural Medicines and Jiangsu Key Laboratory of Drug Design and
Optimization, China Pharmaceutical University, Nanjing, 210009, China
‡
Department of Chemistry, School of Science, China Pharmaceutical University, Nanjing,
2
1119, China
§
Department of Pharmaceutical Engineering, China Pharmaceutical University, Nanjing, 21119,
China
*
Corresponding authors. E-mail: youqd@163.com (Q. You), zxj@cpu.edu.cn (X. Zhang)
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 67
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
ABSTRACT.
The elevation of oxidative stress preferentially in cancer cells by efficient NQO1 substrates
which promote ROS generation through redox cycling has emerged as an effective strategy for
cancer therapy, even for treating drug-resistant cancers. Here, we described the identification and
structural optimization studies of the hit compound 1, a new chemotype of non-quinone substrate
for NQO1 as an efficient ROS generator. Further structure-activity relationship studies resulted in
the most active compound 20k, a tricyclic 2,3-dicyano indenopyrazinone, which selectively
inhibited the proliferation of NQO1-overexpressing A549 and A549/Taxol cancer cells.
Furthermore, 20k dramatically elevated the intracellular ROS levels through NQO1-catalyzed
redox cycling, and induced PARP-1-mediated cell apoptosis in A549/Taxol cells. In addition, 20k
significantly suppressed the growth of A549/Taxol xenograft tumors in mice with no apparent
toxicity observed in vivo. Together, 20k acts as an efficient NQO1 substrate and may be a new
option for the treatment of NQO1-overexpresssing drug-resistant NSCLC.
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
KEYWORDS
NQO1, Non-quinone substrates, ROS, Drug-resistant, NSCLC
ACS Paragon Plus Environment
2

Page 3 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
. INTRODUCTION
Non-small cell lung cancer (NSCLC) is one of the common aggressive malignant tumor
accounting for approximately 85% of human lung cancer, a disease which remains the leading
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
cause of cancer-related mortality worldwide. The overall 5-year survival rate for NSCLC (all
stages combined) is roughly 17% and the 5-year survival rate in stage IV is only about 2%.2
Although the treatment of NSCLC has evolved over the past two decades resulting in different
therapeutic options including chemotherapy,3 tyrosine kinase inhibitors,4 monoclonal
antibodies,5,6 and some combinations, great challenges still exist due to the intrinsic or acquired
7
8
drug resistance and genomic instability in cancers. For instance, taxol, a chemotherapy drug for
the first-line treatment of NSCLC, almost universally fail for late-stage NSCLC patients owing to
9
drug resistance. Therefore, new drugs and new strategies for effective treatment of drug resistant
NSCLC are urgent needs to develop.
Modulation of the unique biochemical alterations in cancer cells provides a new approach
1
0
for cancer therapy, and probably prevents the development of drug resistance. Reactive oxygen
species (ROS) at low levels play important roles as second messengers in several cellular processes
such as proliferation, angiogenesis, and metastasis.11,12 However, when present at high levels, ROS
1
3
irreversibly damage DNA, lipids, and proteins, and ultimately cause apoptosis of cancer cells. It
is worth noting that cancer cells are usually exposed to a relatively high level of ROS compared to
that for normal cells.14 Such biochemical differences make cancer cells more susceptible to
induced oxidative stress, exceed the threshold that they can endure and lead to ROS-mediated cell
death.15-17 Therefore, pharmacological elevation of intracellular ROS level by small molecules has
been regarded as an effective strategy for cancer therapy.18-20 Moreover, the efficacy of this ROS-
2
1
generating method has been demonstrated in some drug-resistance cancers. Cotreatment of the
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 67
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
molecules that mediate their effects through a ROS based mechanism with anticancer drugs has
significantly sensitized the cells to drugs by overcoming the resistance.22,23 Thus, the strategy to
develop ROS-generating molecules could be promising in treating drug-resistance cancers.
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
A crucial issue in developing ROS-generating strategy is how to elevate the levels of ROS
specifically in cancer cells while simultaneously sparing normal cells. In this respect, NAD(P)H:
quinone oxidoreductase (NQO1) and its redox substrates are being paid close attention to
gradually.24-26 NQO1 is such a cancer-specific cytosolic flavoenzyme that is highly elevated in
27
multiple cancer cells, especially in NSCLC cells, where it is constitutively overexpressed over
1
00-fold greater than that in correlative normal tissues.28 NQO1 uses the reduced pyridine
nucleotide NADH or NADPH as cofactor to catalyze the direct two-electron reduction of a wide
variety of quinones to give hydroquinones (Figure 1A).29,30 Many efficient NQO1 substrates have
been reported to be potential antitumor drugs and the three main mechanisms are described as
follows: 1) a direct alkylation of DNA by the corresponding hydroquinones formed by NQO1
bioreduction (e.g., mitomycin C); 2) hydroquinones as potent Hsp90 inhibitors (e.g., geldamycin
and its derivative 17-AAG); 3) rapid generation of ROS through redox cycling between
hydroquinones and its quinone substrates, such as β-lapachone (β-lap, Figure 1B) (an antitumor
3
1
32
drug candidate in multiple clinical trials), deoxynyboquinone (DNQ, Figure 1B), streptonigrin
33
34
(
STN), and their derivatives. The last mode-of-action by redox cycling is the most substantiated
NQO1-dependent mechanism for anticancer quinones. Upon bioreductive activation of certain
quinones, the corresponding unstable hydroquinones rapidly react with physiological oxygen (O
2
)
in cells to provide two equivalent of superoxide (O
•‒), the main constitute of ROS, and regenerate
2
35
the quinones (Figure 1A). Due to the fact that these redox substrates rapidly and catalytically
generate a great amount of toxic ROS specifically in NQO1-overexpressing cells, they have
ACS Paragon Plus Environment
4

Page 5 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
considerable potential as efficient ROS generators for cancer treatment. However, what cannot be
ignored is that all the present NQO1-mediated ROS generators via redox cycling, structurally,
contain a quinone moiety (Figure 1B), which is often considered as a structural alert as applied in
medicinal chemistry for the risk of idiosyncratic drug toxicity, owing to its reactivity toward
cellular nucleophiles through Michael addition.36,37
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
Figure 1. (A) Redox cycling of quinone substrates for NQO1, and the principle of NADPH-based
NQO1 assay for screening new redox cycling NQO1 substrates; (B) Representative NQO1-
mediated redox cycling quinones as antitumor ROS generators; (C) Structure of hit compound 1
as non-quinone redox cycling substrate for NQO1 in this work.
Here, we report our efforts on the discovery of non-quinone substrates for NQO1 as
effective intracellular ROS generators for the treatment of drug-resistant NSCLC. Through
screening of an in-house collection of 7500 compounds by using an NADPH-based NQO1 assay,
fortunately, compound 1 was identified as an efficient ROS generator through NQO1-catalyzed
redox cycling with a tetracyclic non-quinone scaffold. It exerted significant antitumor activities in
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 67
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
vitro toward both A549 and taxol-resistant A549/Taxol lung cancer cells, and showed considerable
selectivity to NQO1-rich A549/Taxol cancer cells over NQO1-deficient L02 normal hepatic cells.
Thus, tetracyclic compound 1 was selected as a starting point for optimization. Then, compound
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
0k with a simplified tricyclic scaffold, which was more selective toward A549/Taxol cancer cells,
was revealed from the structure-activity relationship (SAR) study by a step-by-step modification
strategy. Pharmacological study demonstrated that 20k dramatically elevated the intracellular ROS
levels through NQO1-catalyzed redox cycling, and induced poly(ADP-ribose)polymerase-1
(PARP-1)-mediated cell apoptosis in an NQO1- and ROS-dependent manner in A549/Taxol cells.
It also significally suppressed tumor growth in A549/Taxol cell-xenografted mouse model with no
apparent toxicity observed in vivo. Therefore, we provide a new chemotype of non-quinone
substrates for NQO1 as effective and selective intracellular ROS generators for developing
promising new drugs for the treatment of NQO1-overexpresssing drug-resistant NSCLC.
2
. RESULTS AND DISCUSSION
2.1. Hit Identification from an In-House Compound Collection Using an NADPH-
Based NQO1 Assay. Screening for Non-quinone Redox Cycling NQO1 Substrates as ROS
Generators. With the aim to discover new chemical types of NQO1 substrates with redox cycling
property, we screened experimentally an in-house compound collection containing 7500
3
8
compounds by using an NADPH-based NQO1 assay. As illustrated in Figure 1A, given that the
tested compound acted as an NQO1 substrate and formed redox cycling, the amount of consumed
NADPH would be much more than that of tested compound. N (cycles), defined as the equiv of
consumed NADPH divided by the equiv of tested compound, was used as an index to the character
of redox cycling. Compounds with N (cycles) > 10 (half of the N value of positive control quinone
substrate β-lap) were selected as hit compounds. Fortunately, non-quinone compound 1 (Chemdiv
ACS Paragon Plus Environment
6

Page 7 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ID: 000A-0055, Figure 1C), was revealed as an efficient redox cycling NQO1 substrate with an N
(cycles) value of 23, which was comparable to that of well-known quinone substrate β-lap (Figure
2A). Moreover, when compound 1 was tested in the absence of NQO1 or with the addition of an
NQO1 inhibitor dicoumarol (DIC), no or little amount of NAPDH was consumed, indicating that
the redox cycling of 1 relied on NQO1 bioreduction (Figure 2A). Further, by quantifying NADPH
oxidation at 2 s intervals in the initial 5 min under different concentrations of 1, the NQO1
reduction rates (Figure 2B) and Michaelis-Menten curves (Figure 2C) were generated, and the
apparent catalytic efficiencies were calculated. It was observed that compound 1 was a highly
efficient substrate and redox cycler with a reduction rate of 1412 ± 64 μmol NADPH/min/μmol
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
6
-1 -1
NQO1 at 10 μM, and a kcat/K
M
value of 5.8 ± 1.0 × 10 M s , which was processed much faster
than β-lap (reduction rate = 1155 ± 44 μmol NADPH/min/μmol NQO1 at 10 μM, and kcat/K
M
= 4.4
6
-1 -1
±
0.9 × 10 M s ). Considering that the redox cycling could lead to ROS generation, we thus set
•‒) was measured
out to identify the ability of 1 to generate ROS. The generation of superoxide (O
2
3
9
by a spectrophotometric assay using cytochrome c as the terminal electron acceptor. As shown
in Figure 2D, non-quinone compound 1 showed rapid rate of ROS generation (ROS generation
rate = 1092 ± 64 μmol cyt c/min/mg NQO1), being much more efficient than β-lap (ROS
generation rate = 724 ± 7 μmol cyt c/min/mg NQO1). Additionally. coincubation with DIC
dramatically reduced the rate of ROS generation by 1 (Figure 2D). These results indicated that the
hit compound 1 could generate a high level of ROS via NQO1-directed redox cycling.
Hit Compound 1 Exhibited Potent and Selective Cytotoxicity toward NQO1-
Overespressing Drug Resistant NSCLC cells In Vitro. Cytotoxicity studies were performed on hit
compound 1 with cell survival being determined by the 3-(4, 5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) colorimetric assay against the human NSCLC cells, A549
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 67
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
(
NQO1-rich) and drug-resistant A549/Taxol (NQO1-rich), and L02 normal hepatic cells (NQO1-
deficient) (Figure S1). β-Lap was used as a positive control. Compound 1 showed potent
cytotoxicity against NQO1-overexpressing A549 and A549/Taxol cells with IC50 values of 1.0 ±
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
0
.2 μM (Figure 2E) and 0.6 ± 0.1 μM (Figure 2F), respectively, which was more active than β-lap.
Whereas in the case of L02 cells, compound 1 was less sensitive with an IC50 of 14.3 ± 1.6 μM
Figure 2G). Of particular note was that non-quinone compound 1 showed considerable selectivity
(
to drug-resistant NQO1-rich A549/Taxol cancer cells over NQO1-deficient L02 normal hepatic
cells with a high selectivity ratio of 23.8, being much more selective than quinone compound β-
lap (Figure H). Thus, compound 1 could be a promising hit for further optimization.
Figure 2. (A) N (cycles) represents that the equiv of NADPH consumed relatives to the equiv of
substrates (10 μM) over the course of the NADPH-based NQO1 assay (in 5 min) monitored at 340
nm. (B) The initial reduction rates of non-quinone compound 1 and β-lap (10 μM) by NQO1. (C)
Michaelis-Menten curves for 1 and β-lap with NQO1. (D) Rate of ROS production of 1 and β-lap
in the presence and absence of the NQO1 inhibitor (DIC, 25 μM) with NQO1. (E) Cell death of
NQO1-rich A549 cancer cells with 1 and β-lap for 72 h by MTT assay. (F) Cell death of NQO1-
ACS Paragon Plus Environment
8

Page 9 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
rich A549/Taxol drug-resistant cancer cells with 1 and β-lap for 72 h by MTT assay. (G) Cell death
of NQO1-deficient L02 normal hepatic cells with 1 and β-lap for 72 h by MTT assay. (H) The
selectivity ratio (SR) toward NQO1-rich A549/Taxol cancer cells over NQO1-deficient L02
normal hepatic cells; SR = IC50 (L02)/IC50 (A549/Taxol).
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
2
.2. SAR and Structural Modification of Compound 1 Utilizing a Step-By-Step
Strategy. SAR Studies Focusing on Rings C and D of Tetracyclic Compound 1 Resulting in
Tricyclic Compound 10. To further optimization of hit compound 1, a detailed SAR study was
performed based on their NQO1 reduction rates in vitro. The target compounds 2-10 focusing on
Rings D and C of 1 were designed as illustrated in Figure 3. Their reduction rates and N (cycles)
by NQO1 were determined at the concentration of 10 μM by using the aforementioned NADPH-
based assay, and the results were shown in Table 1. First, the oxadiazole ring D was converted into
other aromatic rings such as benzene (2-4) and thiophene (5) as well as aliphatic ring such as
cyclohexane (6), however, their NQO1 activities were dramatically decreased as compared to 1
with reduction rates ranging from 254 to 362 μmol NADPH/min/μmol NQO1. Removal of ring D
as in compound 7 was also detrimental for its reduction activity by NQO1 (356 ± 41μmol
NADPH/min/μmol NQO1). These results suggested that ring D of 1 was of great importance for
its NQO1 reduction activity. Then, the pyrazine ring C of 7 was converted to 1,2,4-trazine as in
compound 8; no increased activity was observed. Moreover, the replacement of pyrazine ring C
with benzene ring as in compound 9 led to a complete loss in activity, indicating that the nitrogen
atoms in C ring were responsible for NQO1 reduction. Further, considering that the oxadiazole
aromatic ring D of 1 probably acted as a delocalization moiety in the process of NQO1 reduction,
we designed compound 10 with two cyano substitution directly linked to ring C by a ring opening
strategy (Figure 3). The sp hybridized atoms in cyano group were expected to act similar function
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 67
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
to the oxadiazole ring D for electron delocalization during NQO1 reduction. Encouragingly,
compound 10 exerted considerable activity to 1 as expected, with a reduction rate of 1362 ± 38
μmol NADPH/min/μmol NQO1, which was slightly more active than β-lap (1155 ± 44 μmol
NADPH/min/μmol NQO1). Compound 10 was also an efficient redox cycler, showing twenty
redox cycles in the presence of NQO1 in 5 min. It must be emphasized that compound 10 with a
simplified indenopyrazinone tricylic scaffold, structurally, was more beneficial for further
optimization than the tetracylic hit compound 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
Figure 3. Design of the Target Compounds 2-10 Focusing on the Rings C and D of 1.
Table 1. Reduction rates of the Target Compounds 1-10 Catalyzed by NQO1.
ACS Paragon Plus Environment
1
0

Page 11 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reduction rate by NQO1
μmol NADPH/min/μmol NQO1)
compd
N (cycles) a
(
1
2
3
4
5
6
7
8
9
1412 ± 64
362 ± 30
300 ± 22
320 ± 35
316 ± 21
254 ± 34
356 ± 41
23
5
5
4
4
4
5
5
‒
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
347 ± 65
b
NA
10
1362 ± 38
1155 ± 44
20
20
β-lap
a
b
The equiv of NADPH utilized relatives to the equiv of substrates over the course of the assay (in 5 min).
No activity.
SAR Studies Focusing on Rings B and C of Tricyclic Compound 10. Next, for further
exploring the SAR, the target compounds 11-19 focusing on Rings B and C of 10 were designed
as illustrated in Figure 4. Their reduction rates and N (cycles) by NQO1 were determined, and the
results were shown in Table 2. Removal of the carbonyl group in B ring as in 11 led to a complete
loss in activity, suggesting that the carbonyl group was crucial for NQO1 reduction. Compounds
12-16 were designed by replacing the cyano group at C3 with groups containing lone pair electrons,
such as oxygen-containing compounds 12-13, nitrogen-containing compounds 14-15, and
halogen-containing compound 16. However, compounds 12-16 showed extremely low reduction
rates ranging from 117 to 260 μmol NADPH/min/μmol NQO1 (Table 2). Further, the target
compounds 17-19 were generated by changing the linear cyano group into similarly linear alkynyl
groups with sp hybridized carbon atoms. Nevertheless, they all exerted significantly decreased
activity when compared to 10. In addition, compound 19 with bulky benzyl substitution to the end
of alkynyl moiety could hardly be reduced by NQO1. The results demonstrated that the cyano
groups substituted at ring C played a crucial role for NQO1 reduction activity and these sites were
not suitable for further modification.
ACS Paragon Plus Environment
1
1

Journal of Medicinal Chemistry
Page 12 of 67
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
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
Figure 4. Design of the Target Compounds 11-19 Focusing on the Rings B and C of 10.
Table 2. Reduction Rates of the Target Compounds 11-19 Catalyzed by NQO1.
reduction rate by NQO1
compd
N (cycles) a
(
μmol NADPH/min/μmol NQO1)
NA b
‒
2
11
12
151 ± 33
183 ± 42
149 ± 34
260 ± 45
117 ± 30
332 ± 63
363 ± 92
NA b
13
3
1
1
4
5
2
4
16
2
1
1
1
7
8
9
4
5
‒
10
1362 ± 38
1155 ± 44
20
20
β-lap
a
b
The equiv of NADPH utilized relatives to the equiv of substrates over the course of the assay (in 5 min).
No activity.
SAR Studies Focusing on A Ring of Tricyclic Compound 10. According to the SAR
information above, the following moieties including the carbonyl group in ring B, the nitrogen-
containing pyrazine ring C, and the cyano groups substituted on C, were revealed to be important
ACS Paragon Plus Environment
1
2

Page 13 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pharmacorphoric features responsible for 10 by NQO1 reduction. Thus, we retained these features
in the subsequent modification and focused on the ring A of 10. The target compounds 20a-m
were designed by introducing different groups, mainly concerning the electronic effects, such as
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
weak electron-donating group (–CH
3
), strong electron-donating groups (–OCH
3
and –OH), and
electron-withdrawing groups (F, Cl, and –CF
3
) (Table 3). As shown in Table 3, mono-substituted
compounds 20a-g were all efficient NQO1 substrates and redox cyclers, possessing reduction rates
ranging from 1104 to 1376 μmol NADPH/min/μmol NQO1 that were comparable to 10.
Interestingly, mono-substitutions with electron-withdrawing groups (-F and –CF
found to be more beneficial for improving NQO1 reduction activity as compared to mono-
substitutions with electron-donating groups such as –CH , –OCH , and –OH as in compounds 20a-
e. Further, introducing di-substitutions with electron-donating –CH (20h) and –OCH (20i-j)
3
) as in 20f-g were
3
3
3
3
significantly led to decrease in reduction rates by NQO1 (656-902 μmol NADPH/min/μmol NQO1)
when compared to 10. Notably, among this series of target compounds, 20k with difluoro
substitution at C6 and C7 sites of ring A revealed to be the most efficient NQO1 substrate with a
reduction rate of 1471 ± 42 μmol NADPH/min/μmol NQO1, as well as the most efficient redox
cycler, showing 25 redox cycles in the presence of NQO1 in 5 min (Table 3), which was much
more efficient than 10 and β-lap. Changing the positions of difluoro groups in 20k to C5 and C8
as in 20l, and replacing the difluoro with dichloro groups as in 20m both resulted in decreased
activity when compared with 20k.
In addition, representative compounds 1, 10, 20i, and 20k were selected to calculate their
catalytic efficiency. The reduction rates by NQO1 at various concentrations (3.1-70 μM) were
determined and the corresponding Michaelis-Menten curves were subsequently generated (Figure
5
). It was observed that compound 20k was the most efficient substrate (kcat/K
= 6.2 ± 1.1 × 106
M
ACS Paragon Plus Environment
1
3

Journal of Medicinal Chemistry
Page 14 of 67
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 -1
6
-1 -1
M s ), possessing much better kinetic parameters than 10 (kcat/K = 5.1 ± 0.9 × 10 M s ) and
M
6
-1 -1
β-lap (kcat/K = 4.4 ± 0.9 × 10 M s ).
M
Table 3. The Target Compounds 20a-m Focusing on Ring A of 10, and Their Reduction Rates
Catalyzed by NQO1.
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
reduction rate by NQO1
μmol NADPH/min/μmol NQO1)
compd
0a
R1
R2
R3
R4
N (cycles) a
(
2
H
OCH
H
H
H
H
H
H
H
H
H
OCH
H
H
H
H
CH
OCH
H
F
CH
H
H
OCH
OH
F
3
H
H
H
H
H
H
H
H
H
1197 ± 44
1298 ± 58
1104 ± 43
1118 ± 36
1141 ± 39
1376 ± 65
1365 ± 53
902 ± 43
20
20
18
18
18
21
20
12
10
10
25
20
18
20
20
2
2
2
2
2
2
0b
0c
0d
0e
0f
0g
3
3
3
3
CF
CH
3
20h
3
3
2
0i
3
OCH
H
F
656 ± 88
768 ± 78
1471 ± 42
20j
OCH
H
3
OCH
H
3
2
0k
0l
0m
2
2
F
H
H
F
1302 ± 69
H
H
Cl
H
Cl
H
H
H
1180 ± 72
1362 ± 38
1
0
β-lap
1155 ± 44
a
The equiv of NADPH utilized relatives to the equiv of substrates over the course of the assay (in 5 min).
ACS Paragon Plus Environment
1
4

Page 15 of 67
Journal of Medicinal Chemistry
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
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
Figure 5. Michaelis-Menten curves for 1, 10, 20i, 20k and β-lap with NQO1.
Synthesis of the Target Compounds. The target compounds 2-8 and 10 were synthesized
using the routes summarized in Scheme 1. Briefly, diverse diamine materials (22a-c, 23-27)
reacted with ninhydrin hydrate (21) in the presence of acetic acid using ethanol as solvent to give
efficiently the target compounds 2-8 and 10 in one step. Target compound 9 was commercially
available and was purchased directly. The rest target compounds 11-19 and 20a-m were
synthesized using the routes summarized in Scheme 2. Similarly, indanedione 28 reacted with 2,3-
diaminomaleonitrile (27) in isopropanol at room temperature gave 11 in one step with a yield of
8
2%. Treatment of 10 with corresponding nucleophiles such as sodium alkoxides and amines gave
2-15 in moderate yields of 37-87%. Compound 16 was synthesized by heating 14 in acetonitrile
1
in the presence of tert-butyl nitrite and CuCl
2
in 58%. Subsequent treatment of 16 with
corresponding alkynes through Sonogashira coupling reaction gave 17-19 in acetonitrile in 24-
3
6%. Compounds 20a-m were obtained by cyclization of different substituted indanediones (29a-
ACS Paragon Plus Environment
1
5

Journal of Medicinal Chemistry
Page 16 of 67
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
m) with 2,3-diaminomaleonitrile (27) in ethanol and followed by oxidation using potassium
bichromate in a mixture solvent of acetic acid and water in 42-66% over two steps.
Scheme 1. Synthetic Routes for the Target Compounds 2-8, and 10.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
a
o
Reagents and conditions: (a) AcOH, EtOH, rt, 2 h, 73-86%; (b) AcOH, EtOH, 40 C, 4 h, 28-30%; (c) AcOH,
o
EtOH, 70 C, 4 h, 66%.
Scheme 2. Synthetic Routes for the Target Compounds 11-19, and 20a-m.a
ACS Paragon Plus Environment
1
6

Page 17 of 67
Journal of Medicinal Chemistry
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
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
a
Reagents and conditions: (a) 27 (1 equiv), isopropanol, rt, 2 h, 82%; (b) corresponding sodium alkoxides (1
equiv) or amines (1-3 equiv), THF, rt, 2 h, 37-87%; (c) tert-butyl nitrite (5 equiv), CuCl
2
(5 equiv), CH CN, 65
3
oC, 4 h, 58%; (d) corresponding alkynes (1.5 equiv), Pd(dppf)
Cl
2
(0.1 equiv), CuI (0.1 equiv), Et
Cr (0.6 equiv), AcOH, H
N (1.1 equiv),
2
3
o
o
CH
3
CN, 80 C, 3 h, 24-36%; (e) 27 (1 equiv), EtOH, rt, 6 h; (f) K
2
2
O
7
2
O, 100 C, 1 h,
42-66% over two steps. The detailed R groups for 20a-m see Table 3.
SAR Analysis by Electrochemical Property. Electrochemistry is the standard method for
studying chemical redox reactions, and electrochemical techniques has been considerably applied
to biology.40 We have investigated and reported on the electrochemical behavior of ortho-
quinones.41 As for redox cycling substrates for NQO1, undoubtedly, the electrochemical
parameters were always correlated with their biological activities. In this work, the electrochemical
studies were consequently performed to investigate the electrochemical behavior of the non-
quinone redox cycler and to further explain their SAR.
ACS Paragon Plus Environment
1
7

Journal of Medicinal Chemistry
Page 18 of 67
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
Representative compounds (Figure 6) were selected to contain each of the structural classes.
The Eredox values, with reference to ferrocene (F 0/+), were determined and shown in Figure 6 and
c
the cyclic voltammograms (CV) of the representative compounds were shown in Figure S2.
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
As illustrated in Figure 6, these investigated non-quinone compounds possessed Eredox
values in the range of -0.918 V to -1.500 V. It has been suggested that the Eredox values of successful
NQO1 quinone substrates should be in the range of -0.70 V to -1.20 V, and the compounds were
4
2
hard to be bioreduced by NQO1 if the Eredox values were beyond the range. The non-quinone
compounds with an indenopyrazinone scaffold in this work were also found to meet the
requirement of the rule. Compounds (20k, 1, and 10) with rapid NQO1 reduction rates (> 1300
μmol NADPH/min/μmol NQO1) showed higher Eredox values (> -0.1V) ranging from -0.918 V to
-
0.994 V, while compounds (12, 7, 17, and 2) with very slow NQO1 reduction rates exhibited
much lower Eredox values (< -1.2 V). These results could well explain why the replacement by a
benzene ring (as in 2) or removal of ring D (as in 7) of compound 1, and changing of the cyano
group of compound 10 (as in 12 and 17) were all detrimental for reduction by NQO1. Furthermore,
compound 20k, which was derived by introducing additional electron-withdrawing difluoro
substituents to 10, exerted a slightly increased Eredox value, resulting in improvement of NQO1
reduction rate when compared to 10. Similarly, in the case of 20i, possessing additional electron-
donating methoxy groups, led to a decrease in Eredox value and a significant decline in NQO1
reduction activity. In summary, as for the non-quinones with an indenopyrazinone scaffold in this
work, the electrochemical reduction potentials could correlate well with their reduction rates by
NQO1, and the preferable Eredox values fell within -0.70 V and -1.0 V. These findings provided us
a complementary method to further design and optimization of this new chemotype of redox
cycling substrates by NQO1 for developing novel ROS generating antitumor agents.
ACS Paragon Plus Environment
1
8

Page 19 of 67
Journal of Medicinal Chemistry
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
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
Figure 6. Eredox values and NQO1 rates for representative target compounds.
Binding Mode and Catalytic Analysis by Molecular Modelling. Utilizing the crystal
structure of NQO1 (PDB ID: 1H69),43 molecular modelling was performed using GOLD 5.1
software44 and depicted using PyMOL to elucidate the possible binding mode of NQO1 with
tetracyclic hit compound 1 (Figure 7A) and the most active, representative tricyclic compound 20k
(Figure 7B), and the other compounds used in electrochemistry study (Figure S3). It was shown
that they lay deep into the catalytic sites of NQO1 protein and was oriented above the isoalloxazine
ring of bound cofactor FAD by -stacking interaction. This was similar to the reported quinone
4
5
substrates. The carbonyl group and N1 atom of 1 and 20k could interact with the Tyr126 and
Tyr128 residues by hydrogen bonding, respectively (Figure 7A and 7B). The residues including
Trp105, Phe106, and Phe178 formed a hydrophobic pocket of sufficient size to allow only small
substituents to interact favorable. Besides of matching for the catalytic binding site of NQO1, the
substrates to be bioreduced by NQO1 required an appropriate hydride donor-acceptor distance
between the acceptor site in substrate and the FAD cofactor (atom N5 which transfer the hydride)
4
6
in NQO1. Thus, the distance between the carbon atom of C(9)=O in 20k and the nitrogen atom
of N(5)H in FAD was calculated as 3.425 Å, which was within a reasonable distance (<4 Å) for
ACS Paragon Plus Environment
1
9

Journal of Medicinal Chemistry
Page 20 of 67
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
hydride transfer, the key process for initiating NQO1 bioreduction. Based on this, the catalytic
process for the reduction of the new chemotype of non-quinone substrates by NQO1, as
exemplified by 20k, was proposed as illustrated in Figure 7C. The C(9)=O moiety acted as 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
electrophilic site for accepting hydride (H ) from the reduced FAD (FADH
2
), which is similar to
that of quinone substrate β-lap (Figure S4). The two hydrogen bonding interaction of
C(9)=O…Tyr126 and N(1)….Tyr128 were suggested to reduce the electron density of the C9
-
hydride acceptor, thus facilitating hydride (H ) transfer from FADH
2
. Meanwhile, the Tyr126 and
Tyr128 residues were also responsible for promoting substrate 20k (oxidized form) to accept
+
another proton (H ), thus completing the reduction process to give 20k-H2 (reduced form).
Notably, it is the N(1) atom that plays as the proton acceptor in nonquinone substrate 20k, while
it is the C=O moiety in quinone substrate β-lap (Figure S4). These computational findings were
also in accordance with the key pharmacophoric features aforementioned. In addition, the efficient
recovery of substrate 20k after NQO1-mediated redox cycling has also been observed
experimentally in an HPLC-based assay (Figure S5).
ACS Paragon Plus Environment
2
0

Page 21 of 67
Journal of Medicinal Chemistry
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
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
Figure 7. Docked poses of compounds 1 (A) and 20k (B) into the catalytic sites of NQO1 (PDB
code: 1H69). (C) Proposed redox cycling process for 20k by NQO1 reduction.
2.3. In Vitro Antitumor Evaluation. On the basis of the enzymatic assays and
electrochemical studies, compounds (1, 10, and 20a-20m) that possessed good NQO1 reduction
rates and preferable redox cycling properties, were selected for cellular assays to investigate their
potency against NQO1-rich NSCLC cells (A549 cells and Taxol-resistant A549/Taxol cells) and
ACS Paragon Plus Environment
2
1

Journal of Medicinal Chemistry
Page 22 of 67
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
NQO1-deficient normal hepatic cells (L02) by the MTT assay. β-Lap and Taxol were used as
comparison controls. As shown in Table 4, all the target compounds were sensitive to NQO1-rich
and Taxol-resistant A549/Taxol cancer cells with IC50 values ranging from 0.6 to 6.8 μM, which
were much more potent than Taxol (IC50 = 27.0 μM). These compounds were also active to NQO1-
rich A549 cancer cells with IC50 values ranging from 0.8 to 5.9 μM. Whereas in the case of NQO1-
deficient normal L02 cells, the compounds were much less sensitive, exerting IC50 values ranging
from 14.3 to 35.9 μM, showing a desirable selectivity between cancer and normal cells. Notably,
all of the tricyclic target compounds (10 and 20a-m) were less toxic to normal L02 cell when
compared to the tetracyclic hit 1. Relative high selectivity ratios in compounds 20f and 20k-m
were observed, being much greater than hit 1 and positive control β-lap. Interestingly, these
compounds with high selectivity were found to be substituted with fluoro or chloro in the ring A.
Among all the compounds tested, 20k, the most excellent redox cycling substrate for NQO1,
turned out to be the most active and selective tricyclic non-quinone compound to drug resistant
A549/Taxol cancer cells with an IC50 value of 0.6 μM and a selectivity ratio of 50.3. Besides, as
shown in Figure 8, 20k was also found to be active toward other NQO1-overexpressing human
cancers (Figure S1) such as multidrug resistant A549/VCR lung cancer cells, MCF-7 breast cancer
cells, and MIA PaCa-2 pancreatic cancer cells. As for NQO1-deficient human lung cancer H596
cells, 20k was much less sensitive (Figure 8).
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 addition, when the A549/Taxol cells were pretreated with either 50 μM dicoumarol (DIC)
as a NQO1 inhibitor or 10 mM N-acetylcysteine (NAC) as an antioxidant prior to treatment with
2
0k, compound 20k exhibited a 42 to 55-fold less antiproliferative activity against A549/Taxol
cells (A549 + DIC and A549 + NAC, IC50 = 25.6 μM and 33.2 μM, respectively) (Figure 9). The
ACS Paragon Plus Environment
2
2

Page 23 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
results indicated that 20k exhibited antitumor activity and overcame drug resistance through
NQO1-dependent and ROS-mediated pathways.
Table 4. Cytotoxicity of Representative Non-Quinone Compounds toward NQO1-Rich A549 and
A549/Taxol Cancer Cells, and NQO1-Deficient L02 Normal Cells.
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
cytotoxicity IC50 (μM)
selectivity ratio
compd
A549/Taxol
A549
(NQO1-rich)
1.0 ± 0.2
1.4 ± 0.3
1.5 ± 0.3
0.8 ± 0.2
2.3 ± 0.5
2.1 ± 1.5
3.5 ± 0.8
1.4 ± 0.4
5.2 ± 1.0
4.4 ± 0.5
4.5 ± 1.2
5.9 ± 0.4
1.0 ± 0.3
1.9 ± 0.3
2.7 ± 1.0
0.7 ± 0.2
3.1 ± 0.8
L02
(NQO1-deficient)
14.3 ± 1.6
21.3 ± 1.8
26.0 ± 1.5
22.6 ± 1.6
21.6 ± 1.3
22.0 ± 2.4
25.0 ± 1.4
26.1 ± 0.8
28.8 ± 2.3
30.9 ± 1.4
28.1 ± 0.8
35.9 ± 0.9
30.2 ± 1.3
23.0 ± 1.2
35.8 ± 2.3
5.6 ± 2.1
IC50 (L02)/ IC50 (A549/Taxol)
(
NQO1-rich)
0.6 ± 0.1
2.9 ± 0.1
2.7 ± 0.3
2.4 ± 0.3
4.9 ± 0.1
2.2 ± 0.7
6.7 ± 0.6
0.8 ± 0.2
4.1 ± 1.5
6.3 ± 0.6
5.1 ± 0.2
6.8 ± 0.1
0.6 ± 0.1
0.9 ± 0.2
1.3 ± 1.5
27.0 ± 1.6
1.9 ± 0.5
1
23.8
7.3
9.6
9.4
4.4
10.0
3.7
32.6
7.0
4.9
5.5
10
2
2
0a
0b
2
0c
0d
0e
0f
2
2
2
2
2
0g
0h
2
0i
0j
0k
0l
0m
2
2
2
2
5.3
50.3
25.6
27.5
0.2
Taxol
β-lap
11.9 ± 1.4
6.3
Figure 8. Cytotoxicity of 20k toward NQO1-rich A549/Taxol, A549/VCR, MCF-7, and MIA
PaCa-2 cancer cells and NQO1-deficient H596 lung cancer cells.
ACS Paragon Plus Environment
2
3

Journal of Medicinal Chemistry
Page 24 of 67
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
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
Figure 9. Cytotoxicity of 20k toward A549/Taxol cancer cells in the presence and absence of the
NQO1 inhibitor (DIC, 50 μM) and ROS Scavenger (NAC, 10 mM).
2.4. Determination of ROS Generation in Cell-Free and Cell-Based Assays.
Determination of ROS Generation in a Cell-Free Assay. Considering that the redox cycling could
lead to rapid ROS generation, we thus selected representative compounds (1, 10, and 20k) with
efficient NQO1 reduction rates to further identify their abilities to generate ROS. Compound 20k
coincubated with NQO1 inhibitor DIC, and compounds 9 and 11 with no NQO1 reduction activity
were selected as negative controls. β-Lap was chosen as a positive control. The production of
superoxide anion, the main constitute of ROS, was measured by a cell-free spectrophotometric
assay using cytochrome c as the terminal electron acceptor. As shown in Figure 10, the non-
quinones 1, 10, and 20k showed rapid rates of ROS generation, being much more efficient than β-
lap. Compound 20k with the highest metabolic rate by NQO1 (1471 ± 42 μmol NADPH/min/μmol
NQO1) (Table 3) also exhibited the highest rate of ROS production (1208 ± 45 μmol cyt c/min/mg
NQO1) (Figure 10). Further, coincubation with DIC dramatically reduced the rate of ROS
ACS Paragon Plus Environment
2
4

Page 25 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
production by 20k. In addition, compounds 9 and 11 without NQO1 reduction activity also failed
to generate ROS by NQO1.The results indicated that these non-quinone substrates could rapidly
generate a high level of ROS through NQO1-dependent redox cycling.
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
Figure 10. Rate of ROS production of representative compounds (1, 10, and 20k), 20k + NQO1
inhibitor DIC (25μM), 9, and 11 were used as negative controls. β-Lap was used as a positive
control.
Determination of ROS Generation Induced by 20k in A549/Taxol Cells. It was supposed
that 20k exerted its antitumor activity by generating ROS through futile redox cycles catalyzed by
intracellular NQO1. We then monitored intracellular ROS level using a ROS-sensitive fluorogenic
dye (2′,7′-dichlorodihydrofluorescein diacetate, DCFH-DA) after treating the A549/Taxol cells
with compounds 20k and β-lap at 1 and 10 μM for 3 h. As shown in Figure 11A, 20k strongly
ACS Paragon Plus Environment
2
5

Journal of Medicinal Chemistry
Page 26 of 67
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
generated excessive ROS in A549/Taxol cells and the ROS levels were even higher than the
control β-lap in A549/Taxol cells. Considering that DCFH-DA detects a broad range of ROS
including hydrogen peroxide, hydroxyl radical, and peroxy radicals, subsequently, we employed
two other ROS sensors that probe more specific ROS in cells. APF (3′-(p-aminophenyl) fluorescein)
mainly senses hydroxyl radical, whereas dihydroethidium (DHE) selectively detects superoxide
anion in cells.47,48 Fluorescence imaging with the three diff erent ROS probes showed 20k
generated excessive ROS that mainly included superoxide anion rather than hydroxyl radical
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
(Figure 11B). Notably, the level of ROS generated by 20k outcompeted that induced by β-lap.
When we pretreated the A549/Taxol cells with either 50 μM DIC as a NQO1 inhibitor or 10 mM
NAC as an antioxidant prior to treatment with the compound 20k (10 μM), ROS generation was
greatly attenuated. The results indicated that 20k induced ROS generation in an NQO1-dependent
manner in A549/Taxol cells.
ACS Paragon Plus Environment
2
6

Page 27 of 67
Journal of Medicinal Chemistry
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
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
Figure 11. (A) Compound 20k generated high levels of ROS in A549/Taxol cancer cells.
Fluorescence was detected using a fluorescence microscope. (B) Similar experiments were done
in A549 cells using three diff erent ROS probes (DCFH-DA, DHE, APF) with pretreatment with
5
0 μM DIC or 10 mM NAC.
ACS Paragon Plus Environment
2
7

Journal of Medicinal Chemistry
Page 28 of 67
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
2
.5. Pharmacological Studies on Compound 20k in A549/Taxol Cells. Induction of DNA
Damage in A549/Taxol Cells by 20k. Dramatic ROS induction by 20k suggested that 20k exposure
may cause damage and breaks in DNA. DNA damage, measured by comet tail formation, was
detected in A549/Taxol cells after the treatment with 20k (10 μM, 2 h) (Figure 12A and 12B),
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
which was comparable to the positive control H
2
2
O (2 mM). Furthermore, DNA damage was
greatly attenuated in DIC and NAC pretreated A549/Taxol cells. Thus, the results demonstrated
that NQO1-mediated ROS formation by 20k led to significant DNA breaks.
Figure 12. Compound 20k induced DNA damage in A549/Taxol cells in a NQO1, ROS-
dependent manner. Cells were also exposed to 2 mM H
2
2
O for 2 h as positive controls. Comet tail
lengths of A549/Taxol cells were measured by using Image J software (a.u., arbitrary unit).
ACS Paragon Plus Environment
2
8

Page 29 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Effect of 20k on Nucleotides Depletion by PARP-1 Hyperactivation. Excessive ROS
production and significant DNA breaks were observed after the treatment with 20k in A549/Taxol
cells. Poly(ADP-ribose)polymerase-1 (PARP-1), a critical DNA repair enzyme, is immediately
activated after genotoxic stress and catalyze the formation of short-lived negatively charged
Poly(ADP-ribose) (PAR) polymers. we examined the role of this repair protein in 20k-induced
lethality. PAR accumulation, an indicator of PARP-1 hyperactivation, was detected in the
A549/Taxol cells in 10 min of the treatment of 20k, then reduced rapidly in a time-dependent
manner and disappeared after 60 min of the treatment (Figure 13A). Furthermore, PAR
accumulation was suppressed in DIC and NAC pretreated A549/Taxol cells (Figure 13A).
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 addition, rapid PAR formation correlated well with loss of essential metabolic
+
nucleotides (i.e., ATP and NAD ) in the A549/Taxol cells (Figure 13 and 14). Addition of DIC
and NAC prevented 20k-induced PAR formation (Figure 13A) as well as nucleotide loss (Figure
+
1
3B). Inhibition of PARP-1 activity using 3-aminobenzamide (3-AB), an NAD analog, spared
A549/Taxol cells from 20k-induced dose-dependent PAR formation (Figure 14A) and attenuated
nucleotide depletion (Figure 14B). Thus, PARP-1 inhibition protected cells from 20k-induced
PARP-1 hyperactivation. These results indicated that 20k induced PARP-1 hyperactivation and
depletion of essential nucleotides in an NQO1- and ROS- dependent manner.
ACS Paragon Plus Environment
2
9