NQO1-selective activated prodrugs of combretastatin A-4: Synthesis and biological evaluation

Article abstract of DOI:10.1016/j.bioorg.2020.104200

Tumor-specific prodrug treatment renders the exclusive delivery of antitumor agents with the lowest untoward effects. In this work, we reported the synthesis and biological assessment of four NQO1-activatable combretastatin A-4 prodrugs constituted by active drug CA-4, different self-immolating linkers, and NQO1-responsive trigger groups. The in vitro antiproliferative activities showed that prodrug 4 displayed greater selective toxicity toward the tumor cells that overexpressed NQO1, taxol-resistant A549 cells, hypoxia-exposed A549 and HepG2 cells, and incurred lower damage to normal cells in comparison with combretastatin A-4, prodrugs 1, 2, and 3. Moreover, based on a mechanistic study, NQO1 triggered prodrug 4 to effectively liberate the parent drug combretastatin A-4 and kill tumor cells. Furthermore, we also demonstrated that prodrug 4 exerted a stronger anticancer effect and greater safety than combretastatin A-4 under in vivo conditions. Hence, from the above results, NQO1 can be used as a specific delivery system for releasing anticancer agents; besides, prodrug 4 can serve as a candidate lead for developing specific anticancer agents.

Full text of DOI:10.1016/j.bioorg.2020.104200

BioorganicChemistry103(2020)104200
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
NQO1-selective activated prodrugs of combretastatin A-4: Synthesis and
biological evaluation
⁎
Chong Zhanga, Yan Qub, Xin Maa, Manping Lib, Sen Lia, Yue Lib, Liqiang Wua,
b School of Basic Medical Sciences, Xinxiang Medical University, Xinxiang 453003, China
a School of Pharmacy, Xinxiang Medical University, Xinxiang 453003, China
A R T I C L E I N F O
A B S T R A C T
Keywords:
Tumor-specific prodrug treatment renders the exclusive delivery of antitumor agents with the lowest untoward
effects. In this work, we reported the synthesis and biological assessment of four NQO1-activatable com-
bretastatin A-4 prodrugs constituted by active drug CA-4, different self-immolating linkers, and NQO1-re-
sponsive trigger groups. The in vitro antiproliferative activities showed that prodrug 4 displayed greater se-
lective toxicity toward the tumor cells that overexpressed NQO1, taxol-resistant A549 cells, hypoxia-exposed
A549 and HepG2 cells, and incurred lower damage to normal cells in comparison with combretastatin A-4,
prodrugs 1, 2, and 3. Moreover, based on a mechanistic study, NQO1 triggered prodrug 4 to effectively liberate
the parent drug combretastatin A-4 and kill tumor cells. Furthermore, we also demonstrated that prodrug 4
exerted a stronger anticancer effect and greater safety than combretastatin A-4 under in vivo conditions. Hence,
from the above results, NQO1 can be used as a specific delivery system for releasing anticancer agents; besides,
prodrug 4 can serve as a candidate lead for developing specific anticancer agents.
Combretastatin A-4
NQO1-activatable prodrug
NQO1
Antitumor
1. Introduction
fluorouracil, and a self-immolating linker. Suebsakwong et al. [15]
synthesized a bioreductive prodrug of cucurbitacin B and the prodrug
Chemotherapy is a mainstream anticancer therapeutic modality.
However, its application is restricted because of serious adverse effects
on the surrounding non-carcinoma cells and tissues resulting from its
high toxicity and lack of specificity [1,2]. On the contrary, cancer-
specific prodrugs can be targeted accurately at the tumor site in the
unique cancer microenvironment conditions of low pH [3,4], high
glutathione [5,6], varied enzyme concentrations [7,8], or reactive
oxygen species (ROS). This type of prodrug targeting can markedly
improve the therapeutic index and decrease off-target toxicity [9,10].
NAD(P)H:quinone oxidoreductase 1 (NQO1) is a two-electron re-
ductase responsible for detoxification of quinones and also bioactiva-
tion of certain quinones [11]. NQO1 is over-expressed in many solid
tumors and the enzyme is a potential molecular target in tumor treat-
ment [12]. Besides its role as the target of the action, NQO1 is also used
to activate quinone prodrugs or fluorescent probes. For instance,
McCarley's group [13] utilized the quinone propionic acid moiety as a
trigger group in creating self-immolating profluorophores, this pro-
fluorophore was efficiently degraded by NQO1 for releasing the fluor-
escent probes within the NQO1-positive tumor cells. Zhang et al. [14]
proposed a prodrug showing high specificity to cancer, which contained
a quinone propionic acid moiety as the trigger group, an active drug 5-
showed a satisfactory and comparable effectiveness in controlling
tumor growth as that by tamoxifen in the 4 T1 xenograft mice model.
Typically, NQO1 efficiently and selectively reduced the prodrug to re-
lease the active drug in NQO1-overexpressing cancer cells. According to
the above results, it appears feasible to design NQO1-responsive pro-
drugs that can identify specific anticancer drugs.
Combretastatin A-4 (CA-4) is collected from the bark of Combretum
caffrum (South African bush willow tree) and can destabilize micro-
tubules by combining with tubulin and bypassing the division of cells
[16]. Previous reports have demonstrated that CA-4 effectively fights
against many malignant cancers, including lung cancer, breast cancer,
and liver cancer [17]. The efficacy of CA-4 as an antitumor therapeutic
agent is greatly limited by its low solubility and adverse effects. Con-
sequently, some small molecular prodrugs of CA-4 are proposed to
overcome these limitations (Fig. 1) [18–21]. One of them, CA-4P, is a
distinct phosphate drug, which is tested against various malignancies in
clinical trials. CA-4P can be solubilized in water and broken down
through endogenous phosphatases for producing CA-4, the active
compound. CA-4P shows moderate effect in clinical trials, but it can
sometimes lead to adverse events in grades 3 or 4 [22]. As a result, new
prodrug treatments should be developed to decrease the toxicity of the
⁎ Corresponding author.
E-mail address: wliq1974@163.com (L. Wu).
https://doi.org/10.1016/j.bioorg.2020.104200
Received 25 February 2020; Received in revised form 28 July 2020; Accepted 31 July 2020
Availableonline26August2020
0045-2068/©2020ElsevierInc.Allrightsreserved.

C. Zhang, et al.
BioorganicChemistry103(2020)104200
Fig. 1. Structures of the representative small molecule prodrugs of CA-4.
antitumor agents.
2. Results and discussion
The prodrug strategy is a practical approach in development of
novel antitumor agents. Many reported NQO1-selective activated pro-
drugs are composed of three moieties: a parent drug, a self-immolating
linker and a NQO1-responsive trigger group. The trigger group, as a
substrate for NQO1, is often connected to the parent drug via a self-
immolating linker. The linker is incorporated to promote enzymatic
cleavage of the trigger group, and must spontaneously eliminate to
release the desired active agent after trigger group. Carbamate and
carbonate functionality confers chemical and proteolytic stability and a
great capacity to pass through cell membranes, Therefore, carbamate
and carbonate prodrug strategy with attachment of an eligible self-
immolative spacer has been well utilized in the development of novel
NQO1-selective activated prodrugs [12]. In this work, we designed and
synthesized four NQO1-activatable CA-4 prodrugs that were constituted
by an active drug CA-4, a different self-immolating linker, and an
NQO1-responsive trigger group (Fig. 2). Based on the outcomes of the
preliminary tumor cell selection, prodrug 3 with the highest potency
and selectivity was assessed for its ability to release CA-4 in response to
NQO1, its antitumor effect and stability within plasma or buffers.
2.1. Chemistry
The synthetic routes for these prodrugs are depicted in Scheme 1.
Compound 5, 6a, 6b and 10 was prepared according to literature [23].
Compound 7a was prepared according to literature [24]. Compound 8
was prepared according to literature [25]. First of all, the quinone
propionic acid 5 was synthesized by lactonizing the commercial tri-
methylhydroquinone and 3-methylbut-2-enoic acid in the presence of
dry methanesulfonic acid, followed by oxidation with N-bromosucci-
nimide (NBS). Compound 1, which contained no self-immolating linker,
was produced by condensing CA-4 and 5 in the presence of EDCl
(Scheme 1A). The syntheses of compounds 2 and 3 are shown in
Scheme 1B. Compound 5 was activated with isobutyl chloroformate
(IBCF) and coupled with p-aminobenzyl alcohol or p-methylamino-
benzyl alcohol to give compounds 6a or 6b. These compounds were
activated with p-nitrobenzyl chloroformate, followed by treatment with
CA-4, to yield the desired products 2 or 3. Meanwhile, compound 5 was
activated with IBCF and used in combination with mono-Boc-protected
Fig. 2. Structures for NQO1-activatable CA-4 prodrugs.
2

C. Zhang, et al.
BioorganicChemistry103(2020)104200
O
O
O
O
O
A.
O
O
O
EDCl, DMAP
rt, 24 h
OH
+
O
O
O
O
OH
O
O
O
CA-4
1
5
O
OH
B.
O
O
NO2
O
O
N
R
i)
0 oC, 30 min
IBCF,
OH
,
DIPEA
Cl
O
O
OH
O
OH
rt, 6 h
, rt, 24 h
HN
or
ii)
H2N
6a, R = H
6b, R = CH3
5
NO2
O
O
O
R
O
O
O
O
N
O
N
R
O
O
O
O
CA-4
, DMAP
O
O
O
O
rt, 24 h
7a, R = H
7b, R = CH3
2, R = H
3, R = CH3
C.
NO2
,
O
O
O
O
O
DIPEA
NO2
Cl
O
O
rt, 3 h
O
O
O
O
OH
O
O
CA-4
8
O
N
O
O
H
N
O
O
N
O
O
N
i)
0 oC, 30 min
IBCF,
OH
TFA
TFA
O
O
O
O
N
O
ii)
, rt, 24 h
N
H
O
10
9
5
O
O
O
O
O
8
N
N
O
, rt, 24 h
DIPEA, DMAP
O
O
O
4
Scheme 1. Route for synthesizing CA-4 prodrugs 1–4.
diamine for producing compound 9. TFA was used to remove Boc to
yield product 10. When CA-4 was first activated with p-nitrobenzyl-
chloroformate, followed by treatment with 10, the desired product 4
was obtained (Scheme 1C).
water molecule. The prodrug 2 exhibited two pair of double-peaks at
δ = 7.32 and 7.43 ppm, one single at δ = 7.23 ppm for NH proton, and
one single at δ = 5.17 ppm for CH2O protons, which showed that
prodrug 2 formed a linker of NH–(C6H4)–CH2O. The 1H NMR spectrum
of prodrug 2 displayed three singlet at δ = 3.69, 3.79 and 3.84 ppm
(due to OCH3 protons of CA-4), characterising the attachment of the
CA-4 unit in prodrug 2. The prodrug 3 exhibited two pair of double-
peaks at δ = 7.23 and 7.50 ppm, one single at δ = 3.16 ppm for N-CH3
All these prodrugs were identified through 1H NMR, 13C NMR and
HRMS. The HRMS of prodrug 1 displayed the pseudo-molecular ion
([M + Na] +) peak at m/z = 571.2302, which was consistent with the
1: 1 adduct of quinone propionic acid 5 and CA-4 with the loss of one
3

C. Zhang, et al.
BioorganicChemistry103(2020)104200
Table 1
The antiproliferative activities of CA-4 prodrugs 1–4.
Compd.
IC50 (nM)
HepG2
A549
LO2
HEK-293
HepG2/Hyp
A549/Hyp
A549/T
1
31
70
23
10
9
5
59
14
64
7
35
71
44
85
14
4
153
353
141
56
17
47
13
202
381
167
74
33
40
21
150
220
170
35
40
23
2
17
4
107
34
11
105
34
15
10
8
3
7
4
3
4
9
4
4
2
26
72
11
3
8
12
88
17
CA4
1
21
18
112
11
139
16
158
25
protons, and one single at δ = 5.27 ppm for CH2O protons, which
showed that prodrug 3 formed a linker of N(CH3)–(C6H4)–CH2O. The
1H NMR spectrum displayed two singlet at δ = 3.71 and 3.84 ppm (due
to OCH3 protons of CA-4), characterising the attachment of the CA-4
unit in prodrug 3. The prodrug 4 exhibited two pair of multiple-peaks at
δ = 3.38–3.59 ppm for CH2-CH2, and 2.97–3.06 for NCH3 protons,
which showed that prodrug 4 formed a N(CH3)–CH2–CH2–N(CH3) unit.
The 1H NMR spectrum displayed three singlet at δ = 3.72, 3.80 and
3.84 ppm (due to OCH3 protons of CA-4), characterising the attachment
of the CA-4 unit in prodrug 4.
2.2. In vitro antiproliferative study
We first evaluated the antiproliferative activities of prodrugs 1–4,
against the NQO1-overexpressing A549 and HepG2 cells, hypoxia-ex-
posed A549 (A549/Hyp) and HepG2 (HepG2/Hyp) cells, taxol-resistant
A549 cells (A549/T), normal human liver LO2 cells, and HEK293 cells
by the MTT assay. As Table 1 shown, in comparison with CA-4, com-
pound 4 showed a similar antiproliferative effect on the HepG2 and
A549 tumor cells, and the IC50 values were 10 nM and 26 nM, re-
spectively. In comparison, compounds 1–3 were relatively less active
toward the two cancer cells. It was possibly because of the rapid re-
ducibility of compound 4 that resulted in the unfavorable degradation
within the extracellular environment. On the contrary, for normal liver
cells, compound 1 had an IC50 value of 72 nM, which was 7-times less
active than the anti-HepG2 effect (IC50 = 10 nM) and 4-times less ac-
tive than that of the CA-4 (IC50 = 18 nM); for human embryonic
kidney, compound 1 had an IC50 value of 85 nM, which was 8.5-times
less active than the anti-HepG2 effect (IC50 = 10 nM) and 6-times less
active than that of the CA-4 (IC50 = 14 nM). This observation revealed
the wide therapeutic index of prodrug 4. For prodrugs 1–3, their cy-
totoxicity in the LO2 and HEK293 normal cells were equal to that in the
A549 cells; thus, these prodrugs were highly toxic to normal cells and
had no therapeutic value. Additionally, compound 4 showed a re-
markable inhibitory effect on A549/T cells relative to that shown by
CA-4 and compounds 1–3. It is noteworthy that, compared with CA-4
and compounds 1–3, compound 4 displayed a stronger ability to
overcome the hypoxia-induced tolerance in the HepG2/Hyp and the
A549/Hyp cells. Based on these observations of preliminary tumor cell
screening, compound 4 was found to be the most potent and selective.
Hence, compound 4 was further evaluated for its effect on CA-4 release
in response to NQO1, its stability within plasma or buffers, and its
antitumor effect.
Fig. 3. Proposed binding model for 4 (pink) binding with NQO1. (For inter-
pretation of the references to colour in this figure legend, the reader is referred
to the web version of this article.)
the H-bonding interaction of C (4) = O with Tyr126 and Phe106 pos-
sibly facilitates the hydride transfer for the sake of reduction. In addi-
tion, the self-immolating linker is inserted into the pocket constituted
by Gln233, His194, Tyr128, and Met154, while the leaving group CA-4
was left out of the substrate-binding pocket. This indicated that the
introduction of CA-4 into the quinone caused no difference in the in-
teraction between the quinone and NQO1.
2.4. In vitro drug release profiles of CA-4
As suggested, the “trimethyl lock” containing the quinone propionic
acid (Q3PA) trigger group in compound 4 firstly experiences in-
tramolecular cyclization in response to the bioreductive activation of
NQO1; thereafter, the N-methyl carbamate linkage self-cyclizes for the
spontaneous liberation of the CA-4 active compound when the end
amine gets exposed. The release profile of compound 4 in response to
NQO1 and the cofactor NADPH was investigated through the HPLC
assay (Fig. 4A). We observed the prodrug release time-dependence for
compound 4 (Fig. 4B). As the incubation time increases, the proportion
of CA-4 steadily increases, with the plateau being observed at ~70%
after 9 h. To determine whether compound 4 was dependent on NQO1
for release, compound 4 was tested in combination with dicoumarol
(DIC, the NQO1 inhibitor). The results suggested that the transforma-
tion of compound 4 into CA-4 was suppressed when the DIC content
increased to 50 µM. The CA-4 formation within the recombinant NQO1
at different concentrations of the prodrug 4 was measured through the
2.3. Molecular modelling study
Molecular docking analysis was carried out by the Autodock Vina
1.1.2. software using the NQO1 crystal structure (PDB ID: 3JSX) to
examine the potential binding pattern between compound 4 and NQO1.
In compound 4, the quinone moiety is bound to the NQO1 active site
residue as well as to the flavin adenine dinucleotide (FAD) residue,
which is necessary for the catalytic action of NQO1 (Fig. 3). Moreover,
4

C. Zhang, et al.
BioorganicChemistry103(2020)104200
Fig. 4. (A) Degradation scheme of 4. (B) HPLC spectrum of 4 in the presence of NQO1. C) CA-4 released from prodrug in the presence of NQO1 (red), NQO1 + DIC
(blue) respectively. D) Michaelis − Menten curves for 4 with NQO1. (For interpretation of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
HPLC assay. Later, Michaelis − Menten curves were plotted. The re-
combinant NQO1 transformed compound 4 into CA-4 with the re-
quirement of NADPH, and the apparent Km and Vmax were 12.41 µM
and 37.16 nmol/mg/min, respectively (Fig. 4C). Further, an HR-MS
(high resolution-mass spectrometry) assay was also carried out to ob-
serve the CA-4 release (Fig. 5). When the compound 4 was subjected to
coincubation with NADPH and NQO1, the resultant fragments that in-
cluded CA-4 ([M + Na]+ m/z = 339.1203) were measured. Together,
these findings indicated the important role of NQO1 in facilitating the
transformation of compound 4 to CA-4.
2.5. Stability of prodrug 4
A majority of anticancer or antitumor agents show no tissue speci-
ficity and get easily removed from the liver and plasma. Therefore, only
a fraction of the agents penetrate the site of treatment, and this leads to
Fig 5. HR-MS analyses for 4 following co-incubation with NQO1 for 30 min.
5

C. Zhang, et al.
BioorganicChemistry103(2020)104200
Fig. 6. (A) The stability of prodrugs 4 in in phosphate buffer. (B) The stability of prodrugs 4 in phosphate buffer rat plasma and human liver microsomes.
reduced efficacy. Our experiments examined compound 4 for its me-
tabolic stability in phosphate buffer, human liver microsomes, and rat
plasma. Subsequent to incubation with cofactor NADPH for 9 h under in
vitro conditions, compound 4 displayed a small amount of degradation,
suggesting its high complex stability under the tested conditions
(Fig. 6).
2.7. Cell cycle analysis
For determining the effect of compound 4 on cell cycle arrest,
HepG2 cells were treated for 24 h with 10 nM of compound 4 and then
stained with propidium iodide (PI). The nuclear DNA level was de-
termined by flow cytometry, and the non-treated cells were used as
controls. The percentage of cells at the G2/M phase was elevated from
12.71% (control group) to 33.49% (Fig. 8). The result revealed that
compound 4 caused the G2/M phase arrest of the HepG2 cells.
2.6. Immunofluorescence and inhibition of microtubule assembly analysis
CA-4 can target the colchicine site to suppress the polymerization of
tubulin. Hence, to investigate whether compound 4 was targeted to
tubulin, we evaluated the role of compound 4 in the microtubule cy-
toskeleton within the NQO1-overexpressing HepG2 cells. Based on our
results, the DMSO-treated microtubule network was normally arranged
and organized. Meanwhile, treatment with 10 nM CA-4 markedly de-
stroyed the microtubule network. Compound 4 had similar effects on
the microtubule network as obtained with CA-4 (Fig. 7A). Next, we
evaluated the effect of compound 4 on tubulin polymerization activity
(Fig. 7B). The results indicated that compound 4 had no inhibitory ef-
fects on tubulin polymerization. Following the coincubation of com-
pound 4 with NADPH and NQO1, the product was capable of inhibiting
tubulin polymerization. These results provided sufficient evidence that
NQO1 facilitated the transformation of compound 4 into CA-4, and this
effectively inhibited microtubule assembly and displayed greater anti-
tumor activity.
2.8. Apoptosis of HepG2 cells
Mitotic arrest in cancer cells through tubulin-targeted drugs is
considered to be related to cell apoptosis. For assessing the effect of
compound 4 on inducing cell apoptosis, the HepG2 cells were subjected
to 24 h treatment with 10 nM of compound 4. Thereafter, annexin-V
combined with PI was used for staining; flow cytometry was conducted
to determine the apoptotic cell proportion. As shown in Fig. 9, the
apoptotic rate of the HepG2 cells was elevated from 6.54% (control) to
23.93% after treatment with compound 4. The results proved that
compound 4 was capable of causing remarkable apoptosis of the HepG2
cells.
2.9. Antitumor activity in vivo
Based on the favorable observations from the in vitro experiments,
the anticancer effect of the prodrug 4 on HepG2 xenograft-bearing
Fig. 7. (A) Effects of 4 on microtubule network organization, HepG cells were subjected to 48 h of DMSO, 10 nM of 4, and 10 nM of CA-4 treatments. (B) Effects of 4
on tubulin polymerization in vitro.
6

C. Zhang, et al.
BioorganicChemistry103(2020)104200
Fig. 8. 4 effects on the progression of HepG2 cell cycle. Cells were subjected to 24 h of 4 (10 nM) and CA-4 (10 nM) treatments.
BALB/c mice was investigated under in vivo conditions (Fig. 10).
HepG2 cells were subcutaneously injected into the right armpit of the
mice to establish the liver cancer xenograft model. Mice were rando-
mized into four groups when the tumor volume grew to 70–100 mm3:
Group 1 was treated with CA-4, groups 2 and 3 were treated with two
different doses of compound 4, and group 4 served as the vehicle
control group. Each group comprised of six mice. The drugs or vehicle
were administered intraperitoneally once a day for 17 consecutive days.
The vehicle (5% DMSO + 40% PEG300 + 5% Tween 80 + 50%
normal saline) was given to the control mice. Group 1 received 20 mg/
kg of CA–4, groups 2 and 3 received compound 4 at the concentrations
of 20 mg/kg and 40 mg/kg, respectively. On day 17, following the
initial administration, each mouse was sacrificed to harvest the tumor.
Subsequently, the livers were sliced and stained with hematoxylin and
eosin (H&E). The results are summarized in Figs. 10 and 11. Relative to
the vehicle group, intraperitoneal injection with 20 mg/kg of com-
pound 4 resulted in a 47.49% decrease in tumor growth. This was
comparable to the tumor suppression rate of 52.21% observed in the
CA-4-treated group 1. Group 3 mice [injected with 40 mg/kg of com-
pound 4 (an equimolar dose to CA-4)] showed a marked reduction of
54.87% in the tumor growth at the end of the observation period; this
was remarkably potent compared with the effect of CA-4 (Fig. 10B). In
the entire treatment process, no obvious change in the body weight was
observed among the treated mice, indicating minimal or no toxicity
(Fig. 10D). Under the light microscope, the morphology and structure
of the hepatocytes were normal in the control group. Structural disorder
of the hepatic lobules and mild hydropic degeneration of the hepato-
cytes occurred in the CA-4 group, whereas the groups treated with
compound 4 showed no significant histological damage (Fig. 11). Ac-
cording to the above findings, relative to CA-4 (the parent drug),
compound 4 exhibited a higher drug safety and lower toxicity toward
non-carcinoma tissues. Therefore, the possible role of compound 4 as an
efficient target should be further investigated.
3. Conclusions
In summary, to enhance the efficacy of chemotherapy, we devel-
oped four NQO1-activatable CA-4 prodrugs that were constituted by an
active drug CA-4, a different self-immolating linker, and an NQO1-re-
sponsive trigger group. In vitro study on the antiproliferative activity
showed that compound 4 displayed high selective toxicity to NQO1-
overexpressing tumor cells and lower injuries to non-carcinoma cells.
The compound 4 considerably suppressed the hypoxia-induced A549
and HepG2 cells and A549/T cell activities in comparison with CA-4
and compounds 1–3. According to further investigation, NQO1 con-
tributed to the bioreductive activation of compound 4 and effectively
liberated the parent drug CA-4 and killed tumor cells. Further, in vivo
antitumor evaluation showed that compound 4 reduced the xenograft
tumor size in the absence of minimized adverse effects. Consequently,
this NQO1-responsive prodrug might potentially serve as a promising
novel chemotherapeutic target for anticancer treatment.
4. Experimental section
4.1. Chemistry
4.1.1. General
The 98% CA-4 was provided by Aladdin Chemistry Co., Ltd.
(Shanghai, China). NQO1 human with the purity of > 90% was pro-
vided by Sigma-Aldrich China Co., Ltd. (Shanghai, China). The re-
maining analytically pure solvents and reagents were obtained from the
market. The XT-4 binocular microscope was used to determine the
melting points with no any correction. The Bruker AV-400 spectrometer
was used to record the 13H NMR and 1H NMR spectra under ambient
temperature, DMSO-d6 was used as the solvent, whereas tetra-
methylsilane (TMS) was used as the internal reference. Chemical shifts
(d) are expressed in parts per million (ppm) and J values are given in
Fig. 9. (A) Compound 4 led to apoptosis of HepG2 cells under 4 (10 nM) and CA-4 (10 nM) treatments.
7

C. Zhang, et al.
BioorganicChemistry103(2020)104200
Fig.10. 4 suppressed the growth of liver
cancer xenograft in vivo. Following ve-
hicle, CA-4 (20 mg/kg/day), 4 (20 mg/
kg/day) and 4 (40 mg/kg/day) treat-
ments for 17 consecutive days, all ani-
mals were sacrificed to harvest and
weigh the tumors. (A) Tumor images
acquired from mice on 17 days fol-
lowing treatment initiation. (B) The ex-
cised tumor weights in each group.
**P
< 0.05 compared with control
group. (C) Changes in mouse tumor vo-
lume in the treatment process. (D)
Changes in mouse body weight in the
process of treatment.
Fig. 11. Representative images of liver tissue stained by H&E.
hertz (Hz). The Bruker micrOTOF-QIII mass spectrometer was em-
ployed to record the HRMS spectra. The ACS Accurasil C18 column
(dimension, 4.6 mm × 150 mm, 5 μm) was utilized for HPLC analysis
by the use of the solvent acetonitrile /water mixture at the 0.5 mL/min
flow rate, and the UV absorption was observed at the wavelength of
292 nm.
NMR (100 MHz, DMSO‑d6) δ: 168.7, 149.8, 143.15, 128.25, 124.0,
122.2, 121.3, 45.6, 35.4, 27.6 (2C), 15.4, 13.3, 12.8; HR-TOF-MS with
ESI as ionization technique (m/z): calcd for C14H18NaO3 [M + Na]+
257.1148, found: 257.1148.
To a mixture of the above intermediate (15 g, 64 mmol), acetonitrile
(600 mL) and water (300 mL), N-bromosuccinimide (11.4 g, 64 mmol)
dissolved in 50 mL of acetonitrile was added to the suspension slowly.
The mixture was stirred at room temperature for 30 min. After com-
pletion of the reaction, the organic solvent was evaporated to give 16 g
(100%) of 5 as a yellow power, m.p.105–106 °C; 1H NMR (400 MHz,
DMSO‑d6) δ: 12.14 (s, 1H), 2.83 (s, 2H), 2.07 (s, 3H), 1.89–1.88 (m,
6H), 1.36 (s, 6H); 13C NMR (100 MHz, DMSO‑d6) δ: 190.8, 187.2,
174.1, 153.6, 143.3, 137.8, 137.7, 47.6, 28.6 (2C), 14.3, 12.9, 12.2;
HR-TOF-MS with ESI as ionization technique (m/z): calcd for
4.1.2. Synthesis of 3-methyl-3-(2,4,5-trimethyl-3,6-dioxocyclohexa-1,4-
dien-1-yl) butanoic acid 5
To a mixture of 3-methylbut-2-enoic acid (7.5 g, 65.7 mmol) and
methyl sulfonic acid (50 mL), trimethylhydroquinone (10 g,
65.7 mmol) was added. The mixture was stirred at 70 °C for 2 h. After
completion of the reaction, the mixture was then cooled to room tem-
perature, added 100 mL water and extracted with ethyl acetate
(3 × 100 mL). The organic phase was washed with water, dried with
anhydrous Na2SO4 and removed most of solvent. The precipitate was
collected by filtration to afford 11.1 g (72%) of the pure product as a
white power, m.p.116–117 °C; 1H NMR (400 MHz, DMSO‑d6) δ: 8.01 (s,
1H), 2.57 (s, 2H), 2.27 (s, 3H), 2.10–2.09 (m, 6H), 1.34 (s, 6H); 13C
C
14H18NaO4 [M + Na]+ 273.1097, found: 273.1123.
4.1.3. Synthesis of prodrug 1
A mixture containing 5 (140 mg, 0.6 mmol), EDCl (115 mg, 0.6 mol)
and CH2Cl2 (10 mL) was incubate at 0 °C. After stirring for 30 min, 6 mg
8

C. Zhang, et al.
BioorganicChemistry103(2020)104200
DMAP (0.05 mmol) and 158 mg CA-4 (0.5 mmol) were put into the
resultant mixture for 24 h reaction under ambient temperature. After
reaction completion, the mixed product was washed by the NaHCO3
saturated solution and brine, dried with anhydrous Na2SO4. After fil-
tration, the solution was concentrated in vacuo. The residue was pur-
ified by silica gel column chromatography (petroleum ether/ethyl
acetate = 3:1) to give 1 as a yellow oil (187 mg, 68%); 1H NMR
(400 MHz, DMSO‑d6) δ: 7.13 (dd, J = 2.0, 8.4 Hz,1H), 7.03 (d,
J = 8.4 Hz,1H), 6.86 (d, J = 2.0 Hz,1H), 6.51 (s, 2H), 6.47 (s, 2H),
3.68 (s, 3H), 3.63 (s, 3H), 3.57 (s, 6H), 3.10 (s, 2H), 2.07 (s, 3H), 1.84
(s, 3H), 1.82 (s, 3H), 1.41 (s, 6H); 13C NMR (100 MHz, DMSO‑d6) δ:
190.7, 187.2, 170.3, 153.1 (2C), 152.3, 150.4, 143.0, 139.0, 138.8,
138.2, 137.2, 132.4, 129.9, 129.8, 128.6, 128.1, 122.8, 113.1, 106.2
(2C), 60.5, 56.2, 55.9 (2C), 46.8, 38.6, 28.7 (2C), 14.3, 127.7, 12.2; HR-
TOF-MS with ESI as ionization technique (m/z): calcd for C32H36NaO8
[M + Na]+ 571.2300, found: 571.2302.
CDCl3) δ: 191.6, 187.5, 170.3, 153.1 (2C), 152.9, 150.2, 143.3, 139.7,
138.4, 138.2, 138.0, 137.0, 132.4, 130.6, 130.0, 129.6, 129.5, 129.4
(2C), 128.4, 128.0, 122.7, 119.7 (2C), 112.1, 105.7 (2C), 70.0, 60.9,
56.0, 55.9 (2C), 50.4, 38.3, 29.1 (2C), 14.2, 12.7, 12.2; HR-TOF-MS
with ESI as ionization technique (m/z): calcd for C40H43NNaO10
[M + Na]+ 720.2779, found: 720.2776.
4.1.7. Synthesis of 4-methylaminobenzylalcohol
Methyl 4-(methylamino)benzoate (3.1 g, 18.6 mmol) were dissolved
in dry THF (20 mL), then added THF solution of LiAlH4 (18.6 mL,
1 mmol), and the reaction mixture was stirred for overnight at room
temperature. After completion of the reaction, the reaction solution was
treated with water (3 mL), 10% NaOH (3 mL) and water (3 mL) suc-
cessively, and then the reaction mixture was stirred for 30 min at room
temperature, filtered, treated with CH2Cl2 (2 × 200 mL), dried over
anhydrous Na2SO4. After the solvents was removed by rotary eva-
poration, the crude product was purified by silica gel column chro-
matography using petroleum ether and ethyl acetate (v:v = 3:2) as
eluent to provide 4-methylaminobenzylalcoholas a yellow solid (1.4 g,
55%), m.p.68–69 °C; 1H NMR (400 Hz, CDCl3) δ: 7.02 (d,
J = 8.4 Hz,2H), 6.47 (d, J = 8.4 Hz,2H), 5.51 (d, J = 4.0 Hz, 1H), 4.81
(t, J = 5.6 Hz,1H), 4.30 (d, J = 5.6 Hz,2H), 2.64 (d, J = 4.0 Hz, 3H);
13C NMR (100 MHz, CDCl3) δ: 149.4, 129.8, 128.4 (2C), 111.7 (2C),
63.6, 30.4; HR-TOF-MS with ESI as ionization technique (m/z): calcd
for C8H11NNaO [M + Na]+ 138.0913, found: 138.0915.
4.1.4. Synthesis of 6a
To the solution of 5 (1.75 g, 7.46 mmol) and isobutyl chloroformate
(1.11 g, 8.12 mmol) in 20 mL of dry THF was added N-methyl mor-
phorline (0.82 mL) at 0° C. The resultant mixture was stirred for 30 min
at 0 °C, 4-aminobenzylalcohol (1.38 g, 11.2 mmol) was added, and the
resultant mixture was stirred overnight. After filtration, the solution
was concentrated in vacuo. The residue was purified by silica gel
column chromatography (petroleum ether/ethyl acetate = 3:2) to give
6a as a yellow power (1.85 g, 70%), m.p.166–167 °C; 1H NMR
(400 MHz, DMSO‑d6) δ: 9.90 (s, 1H), 7.43 (d, J = 8.4 Hz,2H), 7.20 (d,
J = 8.4 Hz,2H), 5.09 (s, 1H), 4.41 (d, J = 5.6 Hz,2H), 2.95 (s, 2H),
2.05 (s, 3H), 1.90–1.89 (m, 6H), 1.40 (s, 6H); 13C NMR (100 MHz,
DMSO‑d6) δ: 190.8, 187.3, 170.7, 154.7, 144.0, 137.9, 137.8, 137.1,
136.5, 127.3 (2C), 119.4 (2C), 63.1, 49.3, 38.1, 28.6 (2C), 14.2, 13.1,
12.2; HR-TOF-MS with ESI as ionization technique (m/z): calcd for
C21H25NNaO4 [M + Na]+ 378.1676, found: 378.1678.
4.1.8. Synthesis of 6b
To the solution of 5 (1.75 g, 7.46 mmol) and isobutyl chloroformate
(1.11 g, 8.12 mmol) in 20 mL of dry THF was added N-methyl mor-
phorline (0.82 mL) at 0 °C. The resultant mixture was stirred for 30 min
at 0 °C, 4-methylaminobenzylalcohol (1.53 g, 11.2 mmol) was added,
and the resultant mixture was stirred overnight. After filtration, the
solution was concentrated in vacuo. The residue was purified by silica
gel column chromatography (petroleum ether/ethyl acetate = 3:2) to
give 6b as a yellow power (1.87 g, 69%), m.p.153–154 °C; 1H NMR
4.1.5. Synthesis of 7a
Compound 6a (355 mg, 1 mmol) was dissolved in dry CH2Cl2
(30 mL), and 4-nitrophenyl chloroformate (425 mg, 2 mmol) was
added. To the above solution, DIPEA (260 mg, 2 mmol) was added
dropwise. The resultant mixture was kept stirring for 6 h at room
temperature. After reaction completion, the mixed product was washed
by the water, dried with anhydrous Na2SO4. After filtration, the solu-
tion was concentrated in vacuo. The residue was purified by silica gel
column chromatography (petroleum ether/ethyl acetate = 3:1) to give
7a as a yellow soild (364 mg, 70%), m.p.127–128 °C; 1H NMR
(400 MHz, CDCl3) δ: 8.19 (d, J = 7.2 Hz,2H), 7.38 (d, J = 8.4 Hz,2H),
7.29 (d, J = 8.4 Hz,4H), 7.15 (s, 1H), 5.16 (s, 2H), 2.97 (s, 2H), 2.09 (s,
3H), 1.89 (s, 3H), 1.88 (s, 3H), 1.42 (s, 6H); 13C NMR (100 MHz, CDCl3)
δ: 190.6, 186.5, 169.4, 154.5, 151.7, 151.4, 144.3, 142.1, 137.5, 137.4,
137.3, 128.8, 128.7 (2C), 124.3 (2C), 120.8 (2C), 118.8 (2C), 69.6,
49.4, 37.4, 28.1 (2C), 13.2, 11.7, 11.2; HR-TOF-MS with ESI as ioni-
zation technique (m/z): calcd for C28H28N2NaO8 [M + Na]+ 543.1738,
found: 543.1747.
(400 MHz, DMSO‑d6) δ: 7.42 (d, J = 8.0 Hz,2H), 7.23 (d,
J = 8.0 Hz,2H), 5.30 (s, 1H), 4.54 (d, J = 5.2 Hz, 2H), 3.04 (s, 3H),
2.64 (s, 2H), 2.02 (, 3H), 1.93 (, 3H), 1.90 (, 3H),1.22(s, 6H); 13C NMR
(100 MHz, DMSO‑d6) δ: 190.8, 187.3, 171.4, 155.4, 143.9, 142.6,
137.1, 135.6, 128.2 (2C), 127.5 (2C), 112.2, 62.8, 47.3, 38.2, 37.2, 28.5
(2C), 14.2, 13.1, 12.2; HR-TOF-MS with ESI as ionization technique (m/
z): calcd for C22H27NNaO4 [M + Na]+ 392.1832, found: 392.1831.
4.1.9. Synthesis of 7b
Compound 6a (360 mg, 1 mmol) was added into 30 mL dry CH2Cl2,
followed by the addition of 425 mg 4-nitrophenyl chloroformate
(2 mmol). Later, 260 mg DIPEA (2 mmol) was put into the resultant
mixture drop by drop, followed by 6 h stirring under ambient tem-
perature. After reaction completion, the mixed product was washed by
the water, dried with anhydrous Na2SO4. After filtration, the solution
was concentrated in vacuo. The residue was purified by silica gel
column chromatography (petroleum ether/ethyl acetate = 3:1) to give
7b as a yellow oil (363 mg, 68%); 1H NMR (400 MHz, CDCl3) δ: 8.29 (d,
J = 9.2 Hz, 2H), 7.55 (d, J = 8.0 Hz, 2H), 7.43 (d, J = 9.2 Hz, 2H),
7.28 (d, J = 8.0 Hz, 2H), 5.36 (s, 2H), 3.18 (s, 3H), 2.78 (s, 3H), 2.11 (s,
3H), 2.01 (s, 3H), 1.96 (s, 3H), 1.42 (s, 6H); 13C NMR (100 MHz, CDCl3)
δ: 191.2, 187.6, 172.0, 155.4, 154.6, 152.4, 145.4, 144.5, 143.5, 137.8,
136.3 (2C), 133.9, 130.1 (2C), 127.9, 125.3 (2C), 121.8 (2C), 70.1,
47.7, 38.1, 37.1, 28.5 (2C), 14.1, 12.7, 12.1; HR-TOF-MS with ESI as
ionization technique (m/z): calcd for C29H30N2NaO8 [M + Na]+
557.1900, found: 557.1890.
4.1.6. Synthesis of prodrug 2
Compound 7a (520 mg, 1 mmol) and DMAP (122 mg, 1 mmol) were
dissolved in dry CH2Cl2 (20 mL), then added CA-4 ((316 mg, 1 mmol),
and the reaction mixture was stirred for overnight at room temperature.
After reaction completion, the mixed product was washed by water,
dried with anhydrous Na2SO4. After filtration, the solution was con-
centrated in vacuo. The residue was purified by silica gel column
chromatography (petroleum ether/ethyl acetate = 3:1) to give 2 as a
yellow soild (293 mg, 42%), m.p.102–103 °C; 1H NMR (400 Hz, CDCl3)
δ: 7.43 (d, J = 8.4 Hz,2H), 7.32 (d, J = 8.4 Hz,2H), 7.23 (s, 1H),
7.14–7.09 (m, 2H), 6.85 (d, J = 8.4 Hz, 1H), 6.49 (s, 2H), 6.44 (s, 2H),
5.17 (s, 2H), 3.84 (s, 3H), 3.79 (s, 3H), 3.69 (s, 6H), 3.04 (s, 2H), 2.16
(s, 3H), 1.96 (s, 3H), 1.94 (s, 3H), 1.50 (s, 6H); 13C NMR (100 MHz,
4.1.10. Synthesis of prodrug 3
Compound 7b (534 mg, 1 mmol) and DMAP (122 mg, 1 mmol) were
dissolved in dry CH2Cl2 (20 mL), then added CA-4 ((316 mg, 1 mmol),
and the reaction mixture was stirred for overnight at room temperature.
9

C. Zhang, et al.
BioorganicChemistry103(2020)104200
After reaction completion, the mixed product was washed by water,
dried with anhydrous Na2SO4. After filtration, the solution was con-
centrated in vacuo. The residue was purified by silica gel column
chromatography (petroleum ether/ethyl acetate = 3:1) to give 3 as a
yellow oil (392 mg, 55%); 1H NMR (400 Hz, CDCl3) δ: 7.50 (d,
J = 8.0 Hz,2H), 7.24–7.13 (m, 4H), 6.88 (d, J = 8.0 Hz, 1H), 6.50 (s,
2H), 6.47 (s, 2H), 5.28 (s, 2H), 3.84 (s, 6H), 3.71 (s, 6H), 3.16 (s, 3H),
(m, 6H); 13C NMR (100 MHz, CDCl3) δ: 191.2, 187.7, 172.5, 172.4,
163.0, 154.6, 154.5, 152.9, 150.7, 143.2, 143.1, 139.9, 139.8, 138.1,
137.0, 136.2, 132.6, 129.9, 129.3, 129.2, 128.7, 128.6, 127.4, 127.3,
126.0, 123.7, 115.4, 111.9, 111.7, 105.7, 60.9, 55.9, 55.8, 47.7, 47.6,
47.2, 47.0, 46.6, 45.2, 37.4, 37.3, 36.6, 36.2, 35.9, 35.6, 28.5, 14.2,
12.7, 12.6, 12.1; HR-TOF-MS with ESI as ionization technique (m/z):
calcd for C37H46N2NaO9 [M + Na]+ 685.3096, found: 685.3098.
2.76 (s, 2H), 2.11 (s, 3H), 2.01 (s, 3H), 1.96 (s, 3H), 1.31 (s, 6H); 13
C
NMR (100 MHz, CDCl3) δ: 191.3, 187.7, 172.0, 154.7, 153.2, 153.0
(2C), 150.1, 144.2, 143.6, 139.7, 137.8, 137.2, 136.3, 134.6, 132.3,
130.1, 129.8 (2C), 128.3 (2C), 128.1, 127.7 (2C), 122.7 (2C), 112.2,
105.8, 69.5, 60.9, 56.0, 55.9 (2C), 47.7, 38.1, 37.1, 28.4 (2C), 14.1,
12.8, 12.1; HR-TOF-MS with ESI as ionization technique (m/z): calcd
for C41H45NNaO10 [M + Na]+ 734.2932, found: 734.2936.
4.2. Antiproliferative assay in vitro
The antiproliferative activities of the compounds were determined
using MTT assay. HepG2, A549, LO2 and HEK293 cells were inoculated
into the high-glucose DMEM; whereas taxol resistant A549 (A549/
Taxol) cells were cultured within RPMI1640, supplemented with 10%
FBS. Then, all cells were incubated in a humid incubator under 37 °C
and 5% CO2 conditions. The 3% as well as 21% O2 content was used to
be the hypoxic and normoxic conditions, separately. Then, cells were
plated into the 96-well plates, followed by incubation under different
circumstances such as hypoxia and normoxia. Then, the indicated
compounds were diluted with respective vehicle control serially within
the culture medium, which were later used to incubate cells for an
additional 48 h. After that, MTT was put into every well, and the cul-
ture plates were cultured for further 4 h, followed by supernatant re-
moval and dissolution of formazon crystals into DMSO. The absorbance
was recorded at the wavelength of 570 nm.
4.1.11. Synthesis of compound 8
4-nitrophenyl chloroformate (207 mg, 1 mmol), as well as 0.12 mL
pyridine, was added into 10 mL anhydrous THF, followed by dropwise
addition of 150 mg THF solution of CA-4 (0.5 mmol). Later, that ob-
tained mixture was subjected to 3 h stirring under ambient tempera-
ture. After filtration, the solution was concentrated in vacuo. The re-
sidue was purified by silica gel column chromatography (petroleum
ether/ethyl acetate = 3:1) to give 8 as a light yellow oil (180 mg, 75%);
1H NMR (400 Hz, CDCl3) δ: 8.30 (d, J = 9.2 Hz,2H), 7.45 (d,
J = 9.2 Hz, 2H), 7.18 (s, 1H), 6.93–6.91 (dd, J = 5.2, 7.6 Hz, 2H),
6.53–6.46 (m, 4H), 3.89 (s, 3H), 3.84 (s, 3H), 3.70 (s, 6H); 13C NMR
(100 MHz, CDCl3) δ: 155.4, 153.0 (2C), 149.9, 145.5, 139.2, 137.1,
132.4, 130.0, 128.6, 128.2, 126.2, 125.4 (2C), 122.4, 121.7 (2C),
115.6, 112.3, 105.7 (2C), 60.9, 56.1, 55.9 (2C); HR-TOF-MS with ESI as
ionization technique (m/z): calcd for C25H23NNaO9 [M + Na]+
504.1265, found: 504.1269.
4.3. Molecular docking
The binding pattern of compounds with human NQO1 was in-
vestigated through molecular docking by the use of Autodock vina
1.1.2. The RCSB Protein Data Bank (http://www.rcsb.org/) was applied
in obtaining the NQO1 3D structure (PDB ID: 3JSX). Meanwhile,
compound 3D structures were drawn through ChemBio3D Ultra 14.0
and ChemBioDraw Ultra 14.0. Docking input files were generated
through employing AutoDockTools 1.5.6 package. For NQO1, its search
grid was determined as follows, center_x: 11.108, center_y: −3.473,
and center_z: 30.598 with dimensions size_x: 20, size_y: 20, and size_z:
20. The exhaustiveness value was 20. To carry out Vina docking, default
parameters were applied unless otherwise specified. The best-scoring
pose determined based on the score of Vina docking was selected, and
PyMoL 1.7.6 software (http://www.pymol.org/) was used for visual
analysis.
4.1.12. Synthesis of compound 10
To the solution of 5 (1.5 g, 5.99 mmol) and isobutyl chloroformate
(0.90 g, 6.59 mmol) in 30 mL of dry THF was added N-methyl mor-
phorline (0.67 g, 6.59 mmol) at 0 °C. The resultant mixture was stirred
for 30 min at 0 °C. then mono-Boc-protected diamine (1.45 g,
7.19 mmol) was added, and the resultant mixture was stirred overnight.
After filtration, the solution was concentrated in vacuo. The residue
was purified by silica gel column chromatography (petroleum ether/
ethyl acetate = 3:2) to give 10 as a yellowoil (1.44 g, 75%). The pro-
duct was mixture of cis- and trans- isomers oncarbamate bonds; 1H
NMR (400 Hz, CDCl3) δ: 3.42–3.25 (m, 4H), 3.02–2.83 (m, 8H), 2.13 (s,
3H), 1.94–1.91 (m, 6H), 1.49–1.42 (m, 15H); 13C NMR (100 MHz,
CDCl3) δ: 191.3, 187.7, 172.2, 172.1, 155.8, 154.7, 143.3, 137.9, 136.1,
79.6, 79.5, 47.9, 47.2, 46.7, 45.8, 45.1, 37.7, 37.4, 36.1, 35.0, 29.7,
28.6, 28.5, 28.4, 14.2, 14.1, 12.7, 12.6, 12.1; HR-TOF-MS with ESI as
ionization technique (m/z): calcd for C23H36N2NaO5 [M + Na]+
443.2516, found: 443.2514.
4.4. In vitro drug release profiles of CA-4
Incubation mixtures contain 10 mM PBS, 100 μM NADPH and
20 μM prodrug 4. Recombinant NQO1 (10 µg/mL) was directly put to
that incubation mixture for reaction initiation and then incubated at
37 °C for appropriate time intervals. The reaction was terminated by
adding cold acetonitrile (100 μL). After centrifugation in a refrigerated
centrifuge, 20 µL of the liquid were loaded and analyzed by HPLC and
HRMS analysis. To investigate the dicoumarol effect, NQO1 was pre-
viously incubated using 50 μM dicoumarol for 20 min.
4.1.13. Synthesis of prodrug 4
10 (421 mg, 1 mmol) were dissolved in 10 mL of CH2Cl2 and added
5 mL of TFA. The solution was stirred at room temperature for 2 h and
the solvent was evaporated in vacuum. The crude product was re-dis-
solved in 20 mL of anhydrous CH2Cl2, 8 (243 mg, 0.5 mmol), DIPEA
(1.0 mL), DMAP (10 mg, 0.08 mmol) were added to the reaction mix-
ture. The reaction mixture was stirred at room temperature for 24 h.
After completion of the reaction, the reaction solution was treated with
water (2 × 20 mL), dried using anhydrous Na2SO4. After filtration, the
solution was concentrated in vacuo. The residue was purified by silica
gel column chromatography (petroleum ether/ethyl acetate = 2:1) to
give 1 as a yellow oil (185 mg, 56%). The product was mixture of cis-
and trans- isomers on carbamate bonds; 1H NMR (400 Hz, CDCl3) δ:
7.13–7.04 (m, 2H), 6.84–6.80 (m, 1H), 6.53 (s, 2H), 6.45 (s, 2H), 3.84
(s, 3H), 3.80 (s, 3H), 3.72 (s, 6H), 3.59–3.38 (m, 4H), 3.06–2.97 (m,
6H), 2.13–2.12 (m, 2H), 1.92–1.90 (m, 6H), 1.70 (m, 3H), 1.42–1.41
4.5. Formation rates of CA-4 in recombinant NQO1 from 4
The assay mixtures contain NADPH (100 μM), recombinant human
NQO1 (10 μg/mL) and 4 (40, 20, 10, 5, 2.5, 1.25 and 0.625 μM) within
100 μL PBS (10 mmol, pH 7.4). Reactions were carried out at 37 °C for
appropriate time intervals. The reaction was terminated by adding
100 μL cold acetonitrile. After centrifugation in a refrigerated cen-
trifuge, 20 µL of the liquid were loaded and analyzed by HPLC.
Enzymatic parameters (Km and Vmax) were calculated from the kinetic
curves using GraphPad Prism 7.0.
10

C. Zhang, et al.
BioorganicChemistry103(2020)104200
4.6. Stability of 4 in vitro
the FITC Annexin V Apoptosis Detection Kit (KeyGen Biotech) in ac-
cordance with manufacturer instruction. The apoptosis levels were
evaluated by flow cytometry.
The in vitro stability of 4 was studied in PBS buffer with different
pH values, in rat plasma and pooled Human Liver S9 Fraction. 0.1 M
NaOH and 0.1 M HCl solutions were utilized to prepare PBS buffer with
different pH values (pH 5–9). The assays were performed at 37 °C and
conducted in triplicate. In the absence of NQO1, the prodrug (10 μM)
was co-incubated with cofactor NADPH (100 μM) in PBS at different pH
values. After incubation for 9 h, the reaction was terminated by adding
cold acetonitrile (100 μL). After centrifugation in a refrigerated cen-
trifuge, 20 µL of the liquid were loaded and analyzed by HPLC.
Rat plasma were subjected to vortex mixing and then centrifugation
for 5 min at 10000 rpm to deproteinize and the resulting supernatants
were withdrawn, prodrug 4 (10 μM) was co-incubated with super-
natants in 37 °C for at different times. After centrifugation in a re-
frigerated centrifuge, 20 µL of the liquid were loaded and analyzed by
HPLC.
4.11. Antitumor activity in vivo
The 6-week-old BALB/c male nude mice provided by Vital River
Laboratory Animal Technology Co. Ltd were injected with 5 × 106
HepG2 cells subcutaneously. The tumor size was determined at regular
period by the use of calipers, whereas tumor volume was measured
according to the formula of volume (mm3) = length × width × width/
2. When the tumor grew to 70–100 mm3, the animals were randomLy
separated as 4 groups (with 5 in every group): control, 20 mg/kg CA-4,
20 mg/kg 4 and 40 mg/kg 4, and the compounds were administrated
once daily for 17 days by intraperitoneal injection. All the compounds
were prepared with the formulation in 5% DMSO, 40% PEG 300, 5%
tween-80 and 50% saline. Tumor volumes, as well as mouse weights,
were monitored and measured at an interval of 2 days; while tumor
weight were measured after the mice were euthanized at the end of the
experiments. Liver tissue samples were obtained after execution and
fixed in 4% paraformaldehyde solution, dehydrated in ethanol, em-
bedded in paraffin, followed by HE staining and microscopic observa-
tion. Every animal experiment was carried out according to animal
protocols that had been approved by the Institutional Animal Use and
Care Committee of Xinxiang Medical University.
The assay mixtures contain NADPH (100 μM), pooled Human Liver
S9 Fraction (0.5 mg) and 4 (10 μM) within100 μL PBS (10 mmol, pH
7.4). Reactions were carried out at 37 °C for appropriate time intervals,
and terminated by adding cold acetonitrile (100 μL). After centrifuga-
tion in a refrigerated centrifuge, 20 µL of the liquid were loaded and
analyzed by HPLC.
4.7. Immunofluorescence staining
HepG2 cells were inoculated within the 12-well plates, incubated
for 48 h using compounds 4, CA-4 at 10 nM concentrations and the
vehicle control, respectively. Later, cells were subjected to 4% paraf-
ormaldehyde fixation as well as permeabilized using PBS containing
Triton X-100. After that, cells were blocked and incubated within that
diluted α-tubulin antibody (proteintech, 11224-1-AP) overnight at 4 °C.
And the secondary antibody were used to incubate the cells and DAPI to
counterstain the nucleus. The immunofluorescence staining was vi-
sualized under the reverse fluorescence microscopy.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgements
This work was supported by NSFC-Henan Joint Fund (No.
U1604164).
4.8. Tubulin polymerization assay
Appendix A. Supplementary material
The 400 μL reaction mixture supplemented with 10 mM phosphate
buffer (pH 7.4), 100 μM NADPH and 3 μM 4 with or without 10 μg/mL
NQO1. The reaction was carried out within the oscillator under con-
stant temperature of 37 °C for 9 h. The component was extracted with
ethyl acetate for subsequent use.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bioorg.2020.104200.
References
Tubulin polymerization test in vitro was carried out using the HTS-
Tubulin Polymerization Assay Biochem Kit™ (BK004P, Cytoskeleton,
Inc) in accordance with manufacturer instructions. The absorbance
were recorded for 1 h at 340 nm wavelength every 30 s, to obtain the
kinetic curves of polymerization and depolymerization of microtubules.
The reaction groups for these experiments included: vehicle control, 4
(3 μM), 4 (3 μM) + NQO1, CA-4 (3 μM) and Paclitaxel (3 μM).
[1] B.A. Chabner, T.G. Roberts Jr, Chemotherapy and the war on cancer, Nat. Rev.
Cancer 5 (1) (2005) 65–72.
[2] R. Mahato, W. Tai, K. Cheng, Prodrugs for improving tumor targetability and effi-
ciency, Adv. Drug Deliv. Rev. 63 (8) (2011) 659–670.
[3] Y. Chan, T.o. Wong, F. Byrne, M. Kavallaris, V. Bulmus, Acid-labile core cross-linked
micelles for pH-triggered release of antitumor drugs, Biomacromolecules 9 (7)
(2008) 1826–1836.
[4] S.-Y. Li, L.-H. Liu, H.-Z. Jia, W.-X. Qiu, L. Rong, H. Cheng, X.-Z. Zhang, A pH-
responsive prodrug for real-time drug release monitoring and targeted cancer
therapy, Chem. Commun. 50 (80) (2014) 11852–11855.
4.9. Cell cycl analysis
[5] M. Ye, X. Wang, J. Tang, Z. Guo, Y. Shen, H.e. Tian, W.-H. Zhu, Dual-channel NIR
activatable theranostic prodrug for in vivo spatiotemporal tracking thiol-triggered
chemotherapy, Chem. Sci. 7 (8) (2016) 4958–4965.
10 nM Compound 4, 10 nM CA4, and the vehicle control were used
to incubate HepG2 cells for 24 h, respectively. Then the cells were
collected, followed by 70% ethanol fixation overnight, and propidium
iodide/RNase A liquid staining in accordance with manufacturer in-
structions (Cell Cycle Detection Kit KGA511, KeyGEN BioTECH,). Cell
cycle was estimated by flow cytometry.
[6] Y. Liu, S. Zhu, K. Gu, Z. Guo, X. Huang, M. Wang, H.M. Amin, W. Zhu, P. Shi, GSH-
activated NIR fluorescent prodrug for podophyllotoxin delivery, ACS Appl. Mater.
Interfaces 9 (35) (2017) 29496–29504.
[7] P. Liu, J. Xu, D. Yan, P. Zhang, F. Zeng, B. Li, S. Wu, A DT-diaphorase responsive
theranostic prodrug for diagnosis, drug release monitoring and therapy, Chem.
Commun. 51 (46) (2015) 9567–9570.
[8] K. Sharma, A. Iyer, K. Sengupta, H. Chakrapani, INDQ/NO, a bioreductively acti-
vated nitric oxide prodrug, Org. Lett. 15 (11) (2013) 2636–2639.
[9] K. Ou, Y. Kang, L. Chen, X. Zhang, X. Chen, Y. Zheng, J. Wu, S. Guan, H 2 O 2
-responsive nano-prodrug for podophyllotoxin delivery, Biomater. Sci. 7 (6) (2019)
2491–2498.
4.10. Cell apoptosis assay
HepG2 cells (3 × 105 cells/well) were inoculated within the 6-well
plates. Following cell adherence overnight, 10 nM compounds 4 and
10 nM CA4 were used to treat cells for 24 h. Later, cells were collected
and stained with Annexin V and fluorescein isothiocyanate (FITC) using
[10] O. Redy-Keisar, S. Ferber, R. Satchi-Fainaro, D. Shabat, NIR fluorogenic dye as a
modular platform for prodrug assembly: real-time in vivo monitoring of drug re-
lease, ChemMedChem 10 (6) (2015) 999–1007.
[11] D. Siegel, D.L. Gustafson, D.L. Dehn, J.Y. Han, P. Boonchoong, L.J. Berliner, D. Ross,
11

C. Zhang, et al.
BioorganicChemistry103(2020)104200
D. NAD(P)H: quinone oxidoreductase 1: role as a superoxide scavenger, Mol.
[19] D.W. Siemann, D.J. Chaplin, P.A. Walicke, A review and update of the current status
of the vasculature disabling agent combretastatin-A 4 phos- phate (CA4P), Expet
Opin. Invest. Drugs 18 (2009) 189–197.
Pharmacol. 65 (2004) 1238–1247.
[12] K. Zhang, D. Chen, K. Ma, X. Wu, H. Hao, S. Jiang, NAD(P)H: Quinone oxidor-
eductase 1 (NQO1) as a therapeutic and diagnostic target in cancer, J. Med. Chem.
61 (2018) 6983–7003.
[20] Y. Kong, J. Smith, K. Li, J. Cui, J. Han, S. Hou, M.L. Brown, Development of a novel
near-infrared fluorescent theranostic combretastain A-4 analogue, YK-5-252, to
target triple negative breast cancer, Bioorg. Med. Chem. 25 (2017) 2226–2233.
[21] T. Doura, K. Takahashi, Y. Ogra, N. Suzuki, Combretastatin A4-β-Galactosyl con-
jugates for ovarian cancer prodrug monotherapy, ACS Med. Chem. Lett. 8 (2017)
211–214.
[13] M.F. Mendoza, N.M. Hollabaugh, S.U. Hettiarachchi, R.L. McCarley, Human NAD
(P)H:quinone oxidoreductase type I (hNQO1) activation of quinone propionic acid
trigger groups, Biochemistry 51 (2012) 8014–8026.
[14] X. Zhang, X. Li, Z. Li, X. Wu, Y. Wu, Q. You, X. Zhang, An NAD(P)H: Quinone
oxidoreductase 1 responsive and self-immolative prodrug of 5-fluorouracil for safe
and effective cancer therapy, Org. Lett. 20 (2018) 3635–3638.
[22] B. Venegas, W. Zhu, N.B. Haloupek, J. Lee, E. Zellhart, I.P. Sugár, M.F. Kiani,
P.L. Chong, Cholesterol superlattice modulates CA4P release from liposomes and
CA4P cytotoxicity on mammary cancer cells, Biophys. J. 102 (2012) 2086–2094.
[23] W. Zhou, D. Leippe, S. Duellman, M. Sobol, J. Vidugiriene, M. O'Brien, J.W. Shultz,
J.J. Kimball, C. DiBernardo, L. Moothart, L. Bernad, J. Cali, D.H. Klaubert,
P. Meisenheimer, Self-immolative bioluminogenic quinone luciferins for NAD(P)H
assays and reducing capacity-based cell viability assays, ChemBioChem 15 (2014)
670–675.
[15] P. Suebsakwong, J. Wang, P. Khetkam, N. Weerapreeyakul, J. Wu, Y. Du, Z.-J. Yao,
J.-X. Li, A. Suksamrarn, A bioreductive prodrug of cucurbitacin B significantly in-
hibits tumor growth in the 4T1 xenograft mice model, ACS Med. Chem. Lett. 10
(2019) 1400–1406.
[16] G.R. Pettit, S.B. Singh, M.R. Boyd, E. Hamel, R.K. Pettit, J.M. Schmidt, F. Hogan,
Antineoplastic agents. 291. Isolation and synthesis of combretastatins A-4, A-5, and
A-6, J. Med. Chem. 38 (1995) 1666–1672.
[24] Q. Fei, L. Zhou, F. Wang, B. Shi, C. Li, R. Wang, C. Zhao, Rational construction of
probes rendering ratiometric response to the cancer-specific enzyme NQO1, Dyes
Pigments 136 (2017) 846–851.
[17] A.T. Mc Gown, B.W. Fox, Differential cytotoxicity of combretastatins A1 and A4 in
two daunorubicin-resistant P388 cell lines, Cancer Chemother. Pharmacol. 26
(1990) 79–81.
[25] E. Ragozin, A. Hesin, A. Bazylevich, H. Tuchinsky, A. Bovina, T. Shekhter Zahavi,
M. Oron-Herman, G. Kostenich, M.A. Firer, T. Rubinek, I. Wolf, G. Luboshits,
M.Y. Sherman, G. Gellerman, New somatostatin-drug conjugates for effective tar-
geting pancreatic cancer, Bioorg. Med. Chem. 26 (2018) 3825–3836.
[18] K. Jaroch, M.P. Gorski, A. Jaroch, A. Krajewski, A. Ilnicka, A. Sloderbach,
T. Stefanski, S. Sobiak, Combretastatins: In vitro structure-activity relationship,
mode of action and current clinical status, Pharmacol. Rep. 68 (2016) 1266–1275.
12