An NAD(P)H:Quinone Oxidoreductase 1 Responsive and Self-Immolative Prodrug of 5-Fluorouracil for Safe and Effective Cancer Therapy

Article abstract of DOI:10.1021/acs.orglett.8b01409

Tripartite prodrug 1, composed of an NAD(P)H:quinone oxidoreductase 1 (NQO1)-responsive trigger group, a self-immolative linker, and the active drug 5-fluorouracil (5-FU), was designed and synthesized for site-specific cancer therapy. Upon bioreductive ac

Full text of DOI:10.1021/acs.orglett.8b01409

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
An NAD(P)H:Quinone Oxidoreductase 1 Responsive and Self-
Immolative Prodrug of 5‑Fluorouracil for Safe and Effective Cancer
Therapy
Xian Zhang,†,§ Xiang Li,†,§ Zhihong Li,†,‡ Xingsen Wu,†,‡ Yue Wu,† Qidong You,*,†
and Xiaojin Zhang*,†,‡
†Sate Key Laboratory of Natural Medicines and Jiangsu Key Laboratory of Drug Design and Optimization, China Pharmaceutical
University, Nanjing 210009, China
‡Department of Organic Chemistry, School of Science, China Pharmaceutical University, Nanjing 211198, China
S
* Supporting Information
ABSTRACT: Tripartite prodrug 1, composed of an NAD(P)-
H:quinone oxidoreductase 1 (NQO1)-responsive trigger group, a
self-immolative linker, and the active drug 5-fluorouracil (5-FU),
was designed and synthesized for site-specific cancer therapy.
Upon bioreductive activation by NQO1, 1 can release the parent
drug 5-FU specifically in NQO1-overexpressing cancer cells. This
prodrug exerts comparable antitumor activity and a more
favorable safety profile compared with 5-FU both in vitro and
in vivo.
he general goal in anticancer drug development is to
shortcomings, and considering the combination of NQO1
Tachieve highly selective toxicity against cancer cells while overexpression in cancer cells together with the Q3PA-based
rapid chemical release upon NQO1 bioreduction, we rationally
designed a tripartite prodrug (1) that was composed of an
NQO1-responsive trigger group (Q3PA), a self-immolative
linker, and the active drug 5-FU (Figure 1A). Utilizing the
crystal structure of NQO1 (PDB ID: 3JSX),9 a molecular
docking study was performed using GOLD 5.1 to elucidate the
possible binding mode of 1 with NQO1. It was observed that 1
fit well with the catalytic pocket of NQO1 (Figure 1B, Figure
S1). The quinone trigger group lay deep into the catalytic sites
and was oriented above the bound cofactor FAD by π-stacking
interaction with a reasonable distance of 3.38 Å between
C(1)O and N(5)H for hydride transfer; meanwhile, the H-
bond interaction between the other C(4)O and Tyr128 can
probably facilitate the hydride transfer for reduction. The self-
immolative linker inserted into the pocket formed by Phe232,
Phe236, Met154, Tyr128, and His161, whereas the 5-FU
moiety was ideally extended toward the solvent region (Figure
1B). It was proposed that, in response to NQO1 bioreductive
activation, the Q3PA trigger group of 1 would first undergo
intramolecular cyclization, and then the linker could be
eliminated via a well-known self-immolative 1,6-elimination
reaction6,10 to spontaneously liberate the active agent 5-FU,
leading to cancer cell-specific targeting (Figure 1C). In
simultaneously sparing normal cells. Prodrug strategies are an
advanced research area in which nontoxic drugs are carried to
cancer tissues and activated by overexpressing enzymes
specifically, bringing about controllable drug release in target
sites.1 NAD(P)H:quinone oxidoreductase 1 (NQO1) is such a
cancer-specific cytosolic flavoenzyme that is highly elevated in
numerous cancer cells,2 especially in nonsmall cell lung cancer
(NSCLC), where it is constitutively overexpressed over 100-
fold greater than that in correlative normal tissues.2c NQO1 can
catalyze the direct two-electron reduction of a broad range of
substrates, including diverse antitumor quinones3 via a FAD-
mediated hydride (H−) transfer from cofactor NAD(P)H to
quinone substrate.4 Recently, McCarley and co-workers have
identified the “trimethyl lock” containing quinone propionic
acid (Q3PA) as an excellent NQO1-responsive trigger group5
and further developed Q3PA-based fluorescent probes for
detection and imaging of NQO1 in cancer cells.6 Interestingly,
the NQO1-mediated reduction of Q3PA ester or amide
derivatives bearing a “trimethyl lock” is accompanied by
spontaneous intramolecular cyclization to form lactone, which
can lead to the rapid release of a chemical moiety.7
5-Fluorouracil (5-FU), an antimetabolic agent, is still in first-
line chemotherapy for various cancers, including lung, breast,
liver, colorectal, and pancreatic cancers. However, poor
selectivity for cancer cells and high incidence of normal tissue
toxicity limits its therapeutic utility.8 To overcome these
Received: May 3, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b01409
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
carbamate bond in the presence of triphosgene, giving designed
prodrug 1. Meanwhile, compound 2, without the self-
immolative linker, was generated through condensation of 5-
FU with acid 5 using DCC. In addition, compound 3, missing
the NQO1-responsive group, was synthesized by acylation of 5-
FU with benzyl carbonochloridate. Compounds 2 and 3 were
designed as comparison controls for biological assessment.
We first evaluated whether the prodrug could be efficiently
reduced by NQO1 and could specifically release 5-FU. The
initial NQO1 reductive rates of 1 and 2 under various
concentrations were determined by NADPH assay.3b Then, the
Michaelis−Menten curves for 1 and 2 were generated and their
NQO1 kinetic parameters (kcat/KM), reflecting their catalytic
efficiency, were calculated (Figure 2A). It was observed that
Figure 1. (A) Rational design of NQO1-responsive prodrug 1. (B)
Predicted binding mode of 1 in the NQO1 catalytic pocket. (C)
Proposed process for NQO1-responsive drug release.
addition, the cyclized lactone is considered to be nontoxic in
cells,11,11a whereas the formed iminoquinone methide causes
toxicity just at a high concentration.11b We then synthesized
prodrug 1 and further evaluated its capability to release 5-FU in
response to NQO1, its stability in buffers or plasma, its
selective antiproliferative activity toward NQO1-overexpressing
cancer cells, and the antitumor effect as well as its short-term
safety profile in vivo.
The synthetic routes for prodrug 1 and its counterparts 2 and
3 are depicted in Scheme 1. The reaction between 2,3,5-
trimethylbenzene-1,4-diol and 3-methylbut-2-enoic acid in the
presence of methanesulfonic acid provided 4, which was further
treated with NBS to afford 5. Condensation of 5 with a TBS-
protected linker followed by deprotection using TBAF
provided 6. Consequently, 5-FU was incorporated into 6 by a
Scheme 1. Synthetic Routes for Prodrug 1 and its
Counterparts 2 and 3
Figure 2. (A) Michaelis−Menten curves for 1 and 2 with NQO1. (B)
Percentage of 5-FU (as determined by HPLC) released from 1 (20
μM) and 2 (20 μM), respectively, as a function of time in the presence
of NQO1 (14 μg/mL) and NADPH (100 μM) in PBS. (C) HRMS
analyses for 1 (20 μM) after coincubation with NQO1 (14 μg/mL)
and cofactor NADPH (100 μM) for 6 h.
both 1 and 2 were efficient substrates for NQO1, though 1
showed a better kcat/KM value as compared to 2; this suggests
that the prodrug can bind to NQO1 catalytic sites and be
reduced as expected. Subsequently, we set out to study the
release profiles of 1 and 2 in response to NQO1 and cofactor
NADPH using the HPLC assay (Figure 2B). As for prodrug 1,
the time dependence of the prodrug’s release was observed.
With increasing incubation time, the percentage of 5-FU
increased steadily and reached a plateau in 48 h, coming up to
∼92%. However, for its counterpart 2 missing the linker in its
structure, the release of 5-FU was extremely low even after 48 h
incubation, indicating that the self-immolative linker in prodrug
1 is essential for 5-FU release. Furthermore, the anticipated
release of the active agent 5-FU was also observed by HRMS
(ESI) assay (Figure 2C). After prodrug 1 was coincubated with
NQO1 and NADPH, the corresponding fragments including 5-
FU ([M + H]+ m/z = 131.0250) were detected, providing
sufficient evidence for the release of 5-FU in the presence of
NQO1. Importantly, 1 exhibited excellent stability at 37 °C in
B
DOI: 10.1021/acs.orglett.8b01409
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
phosphate buffer (pH 4−11) with cofactor NADPH (Figure
S2) as well as in rat or human plasma (Figure S3) even after 24
h incubation. The results suggest that 1 is relatively stable to
provide enough exposure time under physiological conditions
such that 1 can be specifically activated and release the
cytotoxic drug in NQO1-overexpressing cancer tissues; these
findings provide theoretical reference for further study in vivo.
Next, we investigated the antiproliferative activity of 1,
together with 2 and 3 as comparison controls, against the
NQO1-overexpressing NSCLC cell line (A549), Taxol-resistant
A549 cell line (A549/T), and NQO1-deficient normal human
hepatocyte cell line (L02) using the MTT assay3d (Figure S4).
The IC50 values obtained are given in Table 1. As compared to
Figure 3. In vivo effects of 1 on tumor growth against A549 xenografts
in BALB/c nude mice (iv, 15 mg/kg). 5-FU was selected as control.
(A) The tumor diameters were measured and used to calculate the
tumor volumes. (B) After 20 days, mice were sacrificed, and the
individual tumors were weighed. **p < 0.01; Student’s t test (n = 6).
Table 1. In Vitro Antiproliferative Activity, IC50 (μM)
administration. Food intake in the group that received 1 was
consistent with the control group, whereas the animals, both
male and female, that received 5-FU consumed relatively less
food (Figure S5A and B). As for body weight, no obvious
differences were observed between the 1 group and the saline
control group (Figure S5C and D). However, the weight of the
mice in the 5-FU group significantly decreased compared with
that of the control mice (Figure S5C and D). On the sixth day,
all the mice in the 5-FU group had died, and following
completion of the trial, all male mice died and the survival rate
of female mice was only 20% (Figure 4). After the two week
compd
A549
A549/T
L02
A549+DIC
1
6.6 1.1
39.6 1.1
>100
6.8 1.1
39.5 1.1
96.8 1.9
4.0 1.1
>100
>100
>100
>100
2
3
>100
>100
5-FU
3.4 1.1
7.6 0.9
4.5 1.2
5-FU, 1 exerted comparable antiproliferative activity toward
both A549 and A549/T cancer cells with IC50 values of 6.6 and
6.8 μM, respectively. Importantly, 1 was almost nontoxic
toward L02 cells (IC50 > 100 μM) as expected, whereas no such
difference in selectivity was observed for 5-FU, indicating that
prodrug 1 was selectively toxic to cancer cells that overexpress
NQO1 while causing little damage to normal cells. In addition,
1 had considerable inhibitory activity toward Taxol-resistant
A549 cells. In comparison, 2, without the self-immolative linker,
was relatively less active toward A549 and A549/T cells with
IC50 values of 39.6 and 39.5 μM, respectively; this may be
attributed to its weak ability to release of 5-FU (Figure 2B).
Furthermore, 3, missing the NQO1-responsive group in its
structure, had no obvious antiproliferative activity against the
NQO1-overexpressing cancer cells, demonstrating that 1
exhibited its therapeutic effect through the bioreduction of
the Q3PA trigger group catalyzed specifically by NQO1. To
further verify the role of NQO1 in the drug release process in
the cells, we also performed MTT assays for A549 cells
pretreated with dicoumarol (DIC), an inhibitor of NQO1.
Upon pretreatment with DIC, the cell inhibition decreased
significantly (Figure S4D), suggesting that the prodrug exerted
its cytotoxicity depending on NQO1 bioreductive activation.
Encouraged by the in vitro antitumor cell activity of 1, we
further evaluated its in vivo effect against A549 xenografts in
the BALB/c nude mice model; 5-FU was used as a comparison
control. After tumor xenografts were established, the nude mice
were randomly injected through the tail vein with vehicle, 1,
and 5-FU at a dose of 15 mg/kg once every 2 days for three
consecutive weeks. The results indicate that prodrug 1 is
capable of effectively inhibiting tumor growth in vivo (Figure
3A). The tumor inhibition rate of 1 was 60.5%, which was
comparable to that of 5-FU (68.9%) (Figure 3B).
Figure 4. Survival rate curves of the various groups of male and female
mice.
trial, mice were sacrificed by cervical dislocation and were
dissected surgically for evaluation of possible pathological
changes. Histological examination of heart, liver, spleen, lung,
and kidney tissue samples was performed using standard
hematoxylin-eosin staining protocols. Histologically, tissue
damage, including myocardial cell injury, liver damage, and
interstitial pneumonia, were observed only in mice that received
5-FU, whereas the 1 treatment group showed no histological
damage (Figure 5). In addition, compared to the control group,
no obvious abnormalities in organ coefficients of the 1 group
were observed (Figure S6). These results indicate that,
compared with the parent drug 5-FU, 1 had more favorable
drug safety, exhibiting reduced toxicity to normal tissues. This
prodrug would benefit from further research on its potential as
an effective targeted therapy.
Herein, we present a novel tripartite prodrug 1 composed of
an NQO1-responsive trigger group, a self-immolative linker,
and the active drug 5-FU. This prodrug can be bioreductively
activated by NQO1, and following intramolecular cyclization
and the 1,6-elimination reaction, it can effectively liberate the
parent drug 5-FU. Prodrug 1 is metabolically stable under
physiological conditions, which may assist in avoiding
unexpected off-target drug release, and shows highly selective
toxicity to cancer cells that overexpress NQO1, whereas no
obvious damage to normal cells has been observed. Further in
Consequently, we decided to study the short-term toxicity of
1 in vivo in normal mice to determine if prodrug 1 could cause
toxicity under high doses. Thirty mice were randomly
distributed into three groups (five female and five male in
each group): group I, saline control group; group II, 100 mg/kg
of 5-FU; group III, 100 mg/kg of 1. Mice were treated every
other day by tail vein injection for 2 weeks; body weights and
food intake were recorded every other day after drug
C
DOI: 10.1021/acs.orglett.8b01409
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Central Universities (2632017ZD03), and the “Qing Lan”
Project of Jiangsu Province. Part of the work was supported by
the National Major Science and Technology Project of China
(Nos. 2015ZX09101032 and 2017ZX09302003), the Open
Project of State Key Laboratory of Natural Medicines
(SKLNMZZCX201803), and the College Students Innovation
Project for the R&D of Novel Drugs (J1310032).
REFERENCES
■
(1) For selected reviews, see: (a) Rautio, J.; Kumpulainen, H.;
Heimbach, T.; Oliyai, R.; Oh, D.; Jarvinen, T.; Savolainen, J. Nat. Rev.
Drug Discovery 2008, 7, 255. (b) Ettmayer, P.; Amidon, G. L.;
Clement, B.; Testa, B. J. Med. Chem. 2004, 47, 2393. (c) Zhang, X.; Li,
X.; You, Q.; Zhang, X. Eur. J. Med. Chem. 2017, 139, 542.
(2) (a) Oh, E.; Park, H. J. BMB Rep. 2015, 48, 609. (b) Ma, X.;
Huang, X.; Moore, Z.; Huang, G.; Kilgore, J. A.; Wang, Y.; Hammer,
S.; Williams, N. S.; Boothman, D. A.; Gao, J. J. Controlled Release 2015,
200, 201. (c) Liu, F.; Yu, G.; Wang, G.; Liu, H.; Wu, X.; Wang, Q.; Liu,
M.; Liao, K.; Wu, M.; Cheng, X.; Hao, H. PLoS One 2012, 7, e42138.
(d) Bey, E. A.; Bentle, M. S.; Reinicke, K. E.; Dong, Y.; Yang, C. R.;
Girard, L.; Minna, J. D.; Bornmann, W. G.; Gao, J.; Boothman, D. A.
Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 11832.
Figure 5. Representative images of heart, liver, spleen, lung, and
kidney tissues stained by hematoxylin and eosin at 400×. First row:
group I (saline control group) with no histological abnormality.
Second row: group II (100 mg/kg of 5-FU) with tissue damage
including myocardial cell injury, liver damage, and interstitial
pneumonia. Third row: group III (100 mg/kg 1) with no histological
abnormality as compared with group I.
(3) (a) Colucci, M. A.; Couch, G. D.; Moody, C. J. Org. Biomol.
Chem. 2008, 6, 637. (b) Bian, J.; Deng, B.; Xu, L.; Xu, X.; Wang, N.;
Hu, T.; Yao, Z.; Du, J.; Yang, L.; Lei, Y.; Li, X.; Sun, H.; Zhang, X.;
You, Q. Eur. J. Med. Chem. 2014, 82, 56. (c) Parkinson, E. I.;
Hergenrother, P. J. Acc. Chem. Res. 2015, 48, 2715. (d) Zhang, X.;
Bian, J.; Li, X.; Wu, X.; Dong, Y.; You, Q. Eur. J. Med. Chem. 2017,
138, 616. (e) Bian, J.; Li, X.; Wang, N.; Wu, X.; You, Q.; Zhang, X.
Eur. J. Med. Chem. 2017, 129, 27.
(4) Li, R.; Bianchet, M. A.; Talalay, P.; Amzel, L. M. Proc. Natl. Acad.
Sci. U. S. A. 1995, 92, 8846.
(5) Mendoza, M. F.; Hollabaugh, N. M.; Hettiarachchi, S. U.;
McCarley, R. L. Biochemistry 2012, 51, 8014.
(6) (a) Silvers, W. C.; Prasai, B.; Burk, D. H.; Brown, M. L.;
McCarley, R. L. J. Am. Chem. Soc. 2013, 135, 309. (b) Hettiarachchi, S.
U.; Prasai, B.; McCarley, R. L. J. Am. Chem. Soc. 2014, 136, 7575.
(7) (a) Sung, Y.-M.; Gayam, S. R.; Hsieh, P.-Y.; Hsu, H.-Y.; Diau, E.
W.-G.; Wu, S.-P. ACS Appl. Mater. Interfaces 2015, 7, 25961. (b) Zhou,
W.; Leippe, D.; Duellman, S.; Sobol, M.; Vidugiriene, J.; O’Brien, M.;
Shultz, J. W.; Kimball, J. J.; DiBernardo, C.; Moothart, L.; Bernad, L.;
Cali, J.; Klaubert, D. H.; Meisenheimer, P. ChemBioChem 2014, 15,
670. (c) Volpato, M.; Abou-Zeid, N.; Tanner, R. W.; Glassbrook, L.
T.; Taylor, J.; Stratford, I.; Loadman, P. M.; Jaffar, M.; Phillips, R. M.
Mol. Cancer Ther. 2007, 6, 3122.
(8) Noordhuis, P.; Holwerda, U.; Van der Wilt, C. L.; Van
Groeningen, C. J.; Smid, K.; Meijer, S.; Pinedo, H. M.; Peters, G. J.
Ann. Oncol. 2004, 15, 1025.
(9) Nolan, K. A.; Doncaster, J. R.; Dunstan, M. S.; Scott, K. A.;
Frenkel, A. D.; Siegel, D.; Ross, D.; Barnes, J.; Levy, C.; Leys, D.;
Whitehead, R. C.; Stratford, I. J.; Bryce, R. A. J. Med. Chem. 2009, 52,
7142.
(10) (a) Roth, M. E.; Green, O.; Gnaim, S.; Shabat, D. Chem. Rev.
2016, 116, 1309. (b) Richard, J.; Jean, L.; Romieu, A.; Massonneau,
M.; Noack-Fraissignes, P.; Renard, P. Org. Lett. 2007, 9, 4853.
(c) Esser-Kahn, A. P.; Sottos, N. R.; White, S. R.; Moore, J. S. J. Am.
Chem. Soc. 2010, 132, 10266.
(11) (a) Weerapreeyakul, N.; Anorach, R.; Khuansawad, T.; Yenjai,
C.; Isaka, M. Chem. Pharm. Bull. 2007, 55, 930. (b) Kinski, E.;
Marzenell, P.; Hofer, W.; Hagen, H.; Raskatov, J. A.; Knaup, K. X.;
Zolnhofer, E. M.; Meyer, K.; Mokhir, A. J. Inorg. Biochem. 2016, 160,
218.
vivo antitumor evaluation revealed that 1 can effectively inhibit
tumor growth in the A549 xenograft nude mice model at a dose
of 15 mg/kg. Notably, as compared to 5-FU, 1 exerts a
remarkably improved safety profile; no cases of mortality and
no obvious tissue damage were observed in normal mice under
high-dose administration of 100 mg/kg. Thus, 1 has a wide
therapeutic window and exhibits reduced toxicity to normal
tissues; this prodrug would benefit from further research on its
potential as a specific cancer therapy. Additionally, as an
NQO1-responsive prodrug, 1 is expected to be promising in
personalized chemotherapy for NQO1-overexpressing cancers
when combined with NQO1 diagnostic reagents.6,12
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b01409.
Experimental procedures and compound characterization
for all new compounds (PDF)
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: youqd@163.com.
*E-mail: zxj@cpu.edu.cn.
ORCID
Qidong You: 0000-0002-8587-0122
Xiaojin Zhang: 0000-0002-1898-3071
Author Contributions
§Xian Zhang and Xiang Li contributed equally to this work.
Notes
The authors declare no competing financial interest.
(12) Bian, J.; Qian, X.; Deng, B.; Xu, X.; Guo, X.; Wang, Y.; Li, X.;
Sun, H.; You, Q.; Zhang, X. RSC Adv. 2015, 5, 49471.
ACKNOWLEDGMENTS
■
This work was supported by grants from the National Natural
Science Foundation of China (Nos. 81603025 and 81773571),
Jiangsu Province Funds for Excellent Young Scientists
(BK20170088), the Fundamental Research Funds for the
D
DOI: 10.1021/acs.orglett.8b01409
Org. Lett. XXXX, XXX, XXX−XXX