Studies relating to the synthesis, enzymatic reduction and cytotoxicity of a series of nitroaromatic prodrugs

Article abstract of DOI:10.1016/j.bmcl.2016.11.024

A series of N-nitroarylated-3-chloromethyl-1,2,3,4-tetrahydroisoquinoline derivatives, several of which also possessed a trifluoromethyl substituent, were prepared and assessed as potential nitroaromatic prodrugs. The enzymatic reduction of these compound

Full text of DOI:10.1016/j.bmcl.2016.11.024

Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Studies relating to the synthesis, enzymatic reduction and cytotoxicity of
a series of nitroaromatic prodrugs
Philip J. Burke a, Lai Chun Wong b, Terence C. Jenkins a, Richard J. Knox a, Ian T. Meikle b,
Stephen P. Stanforth b,
⇑
a Morvus Technology Ltd, Llanvetherine Court, Llanvetherine, Abergavenny NP7 8NL, UK
b Department of Applied Sciences, University of Northumbria, Newcastle-upon-Tyne NE1 8ST, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of N-nitroarylated-3-chloromethyl-1,2,3,4-tetrahydroisoquinoline derivatives, several of which
also possessed a trifluoromethyl substituent, were prepared and assessed as potential nitroaromatic
prodrugs. The enzymatic reduction of these compounds and their cytotoxicities were studied. The
compounds were cytotoxic, but this is probably not related to their enzymatic reduction.
Ó 2016 Published by Elsevier Ltd.
Received 17 October 2016
Revised 10 November 2016
Accepted 10 November 2016
Available online xxxx
Keywords:
Nitrogen mustards
NQO1
Nitroreductase
CB 1954
Nitroaromatic prodrugs
The reductive activation of the nitroaromatic prodrug CB 1954 1
(Fig. 1) produces a bifunctional DNA-alkylating agent that is cap-
able of producing DNA-DNA interstrand crosslinks.1–3 In rats,
reductive metabolism of the 4-nitro-group by the enzyme NAD
(P)H: quinone oxidoreductase 1 (NQO1, also known as DT-diaphor-
ase) resulted in the formation of the corresponding 4-hydroxy-
directed enzyme prodrug therapy (ADEPT) or virus/gene-directed
enzyme prodrug therapy (VDEPT/GDEPT) as activation protocols
for CB 1954 1 and related structures in tumours.9–17 The reduction
of the 2-nitro-group in CB 1954 1 also occurred in the presence of
E. coli NR resulting in the ultimate formation of amine derivative 4,
a monofunctional alkylating agent which exhibited a significant
bystander effect.18 Analogues of CB 1954 1 have also been prepared
and studied as potential cytotoxic agents19 as have the structurally
related nitrogen-mustard derivatives SN 23862 5 and its ana-
logues.20–26 The 2-nitro-group in SN 23862 5 is reduced by E. coli
NR producing the amine derivative 6 thus facilitating the forma-
tion of an aziridinium species 7 from the mustard moiety.
lamine derivative
2 that subsequently underwent acylation
generating the cytotoxic species 3.2,3 DNA alkylation then occurred
through the acylated hydroxylamine-group (via a putative nitre-
nium species) and presumably the aziridine moiety, thus creating
the DNA crosslinks.4 Since the highest levels of NQO1 are often
found in tumour tissues (breast, colon, lung, and liver), with lower
levels detected in bone marrow, this enzyme became an attractive
target for nitroaromatic-prodrug therapies in humans.5 CB 1954 1
has previously been shown to exhibit substantial and selective
cytotoxicity against rat Walker 256 carcinomas but, disappoint-
ingly, human cell lines, even those cells expressing high levels of
NQO1, were unresponsive towards this agent. A change in the
amino acid residue 104 (tyrosine in the rat enzyme and glutamine
in the human enzyme) was attributed to the poor catalytic
response of human NQO1 towards CB-1954 1.6,7 CB 1954 1 was,
however, reduced more efficiently by E. coli nitroreductase (NR)8
and this property has stimulated interest in using anti-body
In this Letter, we report the synthesis and evaluation
(enzymatic and cytotoxicity) of
a series of N-nitroarylated
1,2,3,4-tetrahydroisoquinoline derivatives with a core structure
represented by formula 8 as potential nitroaromatic prodrugs. In
view of the current interest in fluorinated compounds in medicinal
chemistry,27–30 structures 8b–8d which possess the strongly elec-
tron-withdrawing trifluoromethyl group31 have been prepared
and compared with the non-trifluoromethylated mono- and
di-nitro compounds 8a and 8e respectively. It was anticipated that
if metabolic reduction of the nitro-group occurred in these mole-
cules 8, the resulting hydroxylamine (or amine) derivative would
facilitate the formation of an aziridinium ion 9 (i.e. a similar activa-
tion process of transforming SN 23862 5 into the aziridinium ion
7). With compounds 8d and 8e (which are both associated with
⇑
Corresponding author.
E-mail address: steven.stanforth@northumbria.ac.uk (S.P. Stanforth).
http://dx.doi.org/10.1016/j.bmcl.2016.11.024
0960-894X/Ó 2016 Published by Elsevier Ltd.
Please cite this article in press as: Burke P.J., et al. Bioorg. Med. Chem. Lett. (2016). http://dx.doi.org/10.1016/j.bmcl.2016.11.024

2
P.J. Burke et al. / Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
X
O
NH2
O
NO2
O
NH2
O
NH2
NH2
NH2
NH2
O2N
Cl
O2N
R
O2N
+
N
N
N
_
Cl
N
Cl
Cl
1
5
R = NO2 (CB 1954)
2 R = NHOH
R = NHOCOMe
4
X = NO2 (SN 23862)
6 X = NH2
7
3
_
Cl
Cl
NHOH (NH2)
R
N
+
NO2
R
N
8
R = H, NO2, CF3
9
Fig. 1. Nitroaromatic prodrugs and their active metabolites.
R1
H
CF3
H
NO2
NO2
R2
NO2
NO2
NO2
CF3
R3
H
H
CF3
H
H
CO2H
CO2H
(i)
a
b
c
d
e
NH
N
56-73%
Ar
11
10
NO2
(ii)
49-84%
(iii) 37-61%
CO2Me
Cl
(iv)
OH
(v)
N
R3
R2
N
N
43-100%
65-67%
Ar
Ar
8
R1
13
12
Scheme 1. Synthesis of the prodrugs 8a–e. Reagents and conditions: (i), ArF, DMSO, K2CO3, 80 °C then dil. HCl; (ii) Me2SO4, acetone, reflux; (iii) EtOCOCl, Et3N, THF, À15 °C
then NaBH4, MeOH, 10 °C; (iv) LiBH4, B(OMe)3 (cat.), Et2O, reflux; (v) SOCl2, CH2Cl2, reflux.
Table 1
R1 = NO2), subsequent acylation of the hydroxylamine-group (if
Specific activities of CB 1954 1 and prodrugs 8a–e.
formed) might then afford a potential bifunctional alkylating
agent, structurally similar to the CB 1954 metabolite 2. Compounds
8a–8e (in which R2 = NO2) would not be expected to produce
bifunctional alkylating species, but their corresponding amines (if
formed), may exhibit a bystander effect similar to amine 4.18
Compounds 8a–e were therefore prepared from racemic
1,2,3,4-tetrahydroisoquinoline 10 as outlined in Scheme 1 (see
Supplementary information for experimental details). Thus, com-
pound 10 was reacted with an appropriate arylfluoride in warm
DMSO solution in the presence of K2CO3 yielding, after acidifica-
tion, the arylated carboxylic acid derivatives 11a–d. Compound
11e was prepared using a similar procedure except that boiling
aqueous EtOH was used as the solvent. These products (with the
exception of compounds 11d and 11e) were converted into their
corresponding methyl esters 12 by treatment with dimethyl sul-
phate under basic conditions. Reduction of these esters 12 with
LiBH4 in the presence of a catalytic quantity of B(OMe)3 afforded
the alcohols 13.32 The alcohols 13 could also be prepared directly
from the carboxylic acids 11 by formation of a mixed anhydride
with ethyl chloroformate under basic conditions followed by
NaBH4 reduction.33,34 The required chloromethyl derivatives 8
Compound
Human NQO1
mol/min/mg)
E. coli NR
mol/min/mg)
(l
Relative to
CB 1954 1
(l
Relative to
CB 1954 1
CB 1954 1
0.0062
<0.0001
0.0270
0.0120
0.0177
0.0033
1.000
<0.01
4.355
1.936
2.855
0.532
1.860
<0.01
0.166
0.106
<0.01
0.254
1.000
<0.001
0.089
0.057
<0.001
0.137
8a
8b
8c
8d
8e
Table 2
IC50 values (
l
mol) of prodrugs 8a–e and CB 1954 1.
Compound
Cytotoxicity (3 days exposure): IC50 values (lmol)
Control
F179
Human NQO1
hDT7
E. coli NR
T116
Rat NQO1
186/6
CB 1954 1
195.9
3.3
37.8
7.3
36.6
49.0
1.5
2.8
27.5
5.9
31.2
43.1
0.03
3.1
3.1
2.9
39.1
1.2
0.05
2.9
22.6
5.8
34.4
36.6
8a
8b
8c
8d
8e
Please cite this article in press as: Burke P.J., et al. Bioorg. Med. Chem. Lett. (2016). http://dx.doi.org/10.1016/j.bmcl.2016.11.024

P.J. Burke et al. / Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
3
NO2
NO2
NO2
_
Cl
X
X
X
+
NO2
R
N
+
N
N
N
_
Cl
Cl
Cl
15
14
16
X = H or NO2
17 R = H, NO2, CF3
Scheme 2. N-Nitroaryl aziridinium intermediates as potential mono-alkylating agents.
were prepared from the alcohols 13 by their reaction with SOCl2 in
CH2Cl2 solution at reflux.
The series of prodrugs 8 were assessed against the enzymes
human NQO1 and E. coli NR and their specific activities (calculated
by dividing the initial rate of reaction by the concentration of the
transfected cell lines has been attributed to the reduction potential
of these pro-drugs.19 The cytotoxicities of the mono and dinitro
prodrugs 8a and 8e respectively are not correlated with their per-
ceived reduction potentials because compound 8a is significantly
more cytotoxic than compound 8e in the control, human NQO1
and rat NQO1 cell lines and broadly similar to compound 8e in
the E. coli NR cell line. This evidence would also support the
hypothesis that the series of prodrugs 8a–8e may be acting as
mono-alkylating agents.
In conclusion, it is clear that all of the prodrugs 8a–8e examined
in this study are significantly less cytotoxic than CB 1954 1 in the
E. coli NR and rat NQO1 cell-lines. Compounds 8a and 8c displayed
IC50 values that are reasonably aligned to that of CB 1954 1, but
only in the human NQO1 cell-line. The cytotoxicity studies
suggest that these prodrugs may be functioning as mono-
functional alkylating agents.
enzyme used and quoted as
lmoles of compound reduced per min-
ute per mg of protein ( mol/min/mg)) have been compared to CB
l
1954 1 (Table 1). Interestingly, all the trifluoromethylated deriva-
tives 8b–d exhibited higher specific activities with human NQO1
than CB 1954 1. The mono-nitro derivative 8a was a poor substrate
for human NQO1, and in the absence of a second electron-with-
drawing group located on the N-aryl-substituent, this observation
was not unexpected. The specific activity of the dinitro-derivative
8e was lower than the nitro/trifluoromethylated derivatives 8b–
d. All of the series of prodrugs 8a–8e showed poor specific activi-
ties with E. coli NR compared to CB-1954 1. Noteworthy is the
observation that the prodrug 8d, which lacks a para-nitro sub-
stituent in the N-aryl ring, exhibits very little specific activity with
E. coli NR compared to prodrugs 8b, 8d and 8e. This correlates with
the E. coli NR-induced reduction of the nitro-group in both CB 1954
1 and SN 23862 5 (i.e. the nitro-group para to the aziridine/mus-
tard moieties is reduced).
Acknowledgements
We thank the EPSRC for a CASE award (to I.T.M.) and Morvus
Technology and Northumbria University for a studentship (to L.C.
W.). We also thank the EPSRC national mass spectrometry facility,
Swansea, for high resolution mass spectra.
In order to assess the cytotoxicities of the potential prodrugs
8a–e, their IC50 values were determined against constructed cell-
lines that expressed the relevant enzyme against a null background
using a conventional sulforhodamine-B (SRB) assay.35 Examination
of the cytotoxicity data (Table 2) revealed that prodrugs 8a and 8c
exhibited broadly similar IC50 values across the four cell-lines and
that there was no clear differentiation between the control line and
the three nitroreductase-expressing cell-lines. Additionally, both
compounds displayed a greater cytotoxicity in the control cell-line
than CB 1954 1. These observations suggested that compounds 8a
and 8c are associated with a cytotoxic effect that is not related to
their nitroreductase activity despite prodrug 8c showing a higher
specific activity to human NQO1 than CB 1954 1 (Table 1). A pos-
sible explanation for this observation is that these prodrugs are
behaving as mono-functional alkylating agents, either as alkyl
chlorides or via aziridinium intermediates. In support of this
hypothesis, we have recently shown that the N-nitroaryl-3-chloro-
piperidine derivatives 14 are converted via aziridinium intermedi-
ates 15 into the N-nitroaryl-2-chloropyrrolidines 16 (Scheme 2).36
Hence compounds 8a and 8c may be forming aziridinium interme-
diates 17 (rather than aziridiniums 9, Fig. 1) that might be capable
of functioning as mono-alkylating agents. Prodrug 8d also showed
broadly similar, but significantly higher IC50 values across the four
cell-lines compared to compounds 8a and 8c. Prodrugs 8b and 8e
displayed broadly similar IC50 values across three of the cell-lines
(control, human NQO1 and rat NQO1), but both of these com-
pounds are associated with significantly lower IC50 values in the
E. coli NR cell-line for reasons that are as yet unclear. The relatively
high cytotoxicity observed for CB 1954 1 compared to the poor
cytotoxicities seen for the mono-nitro analogues (i.e. structure 1
with one nitro-group replaced by hydrogen) in nitroreductase-
A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.11.
024.
References
1. Knox RJ, Burke PJ, Chen S, Kerr DJ. Curr Pharm Des. 2003;9:2091.
2. Knox RJ, Friedlos F, Jarman M, Roberts JJ. Biochem Pharmacol. 1988;37:4661.
3. Boland M, Knox R, Roberts J. Biochem Pharmacol. 1991;41:867.
4. Knox RJ, Friedlos F, Marchbank T, Roberts JJ. Biochem Pharmacol. 1991;42:1691.
5. Cresteil T, Jaiswal AK. Biochem Pharmacol. 1991;42:1021.
6. Chen S, Knox R, Wu K, et al. J Biol Chem. 1997;272:1437.
7. Skelly JV, Sanderson MR, Suter DA, et al. J Med Chem. 1999;42:4325.
8. Knox RJ, Friedlos F, Sherwood RF, Melton RG, Anlezark GM. Biochem Pharmacol.
1992;44:2297.
9. Denny WA, Wilson WR. J Pharm Pharmacol. 1998;50:387.
10. Denny WA. Eur J Med Chem. 2001;36:577.
11. Denny WA. Curr Pharm Des. 2002;8:1349.
12. Palmer DH, Milner AE, Kerr DJ, Young LS. Br J Cancer. 2003;89:944.
13. Johansson E, Parkinson GN, Denny WA, Neidle S. J Med Chem. 2003;46:4009.
14. Atwell GJ, Yang S, Pruijn FB, et al. J Med Chem. 2007;50:1197.
15. Mitchell DJ, Minchin RF. Cancer Gene Ther. 2008;15:758.
16. Prosser GA, Copp JN, Syddall SP, et al. Biochem Pharmacol. 2010;79:678.
17. Williams EM, Little RF, Mowday AM, et al. Biochem. J.. 2015;471:131.
18. Helsby NA, Ferry DM, Patterson AV, Pullen SM, Wilson WR. Br
2004;90:1084.
19. Helsby NA, Atwell GJ, Yang S, et al. J Med Chem. 2004;47:3295.
20. Palmer BD, Wilson WR, Pullen S, Denny WA. J Med Chem. 1990;33:112.
21. Palmer BD, Wilson WR, Atwell GJ, Schultz D, Xu XZ, Denny WA. J Med Chem.
1994;37:2175.
J Cancer.
22. Palmer BD, van Zijl P, Denny WA, Wilson WR. J Med Chem. 1995;38:1229.
23. Anlezark GM, Melton RG, Sherwood RF, et al. Biochem Pharmacol. 1995;50:609.
Please cite this article in press as: Burke P.J., et al. Bioorg. Med. Chem. Lett. (2016). http://dx.doi.org/10.1016/j.bmcl.2016.11.024

4
P.J. Burke et al. / Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
24. Palmer BD, Wilson WR, Anderson RF, Boyd M, Denny WA.
1996;39:2518.
25. Friedlos F, Denny WA, Palmer BD, Springer CJ. J Med Chem. 1997;40:1270.
26. Helsby NA, Wheeler SJ, Pruijn FB, et al. Chem Res Toxicol. 2003;16:469.
27. Purser S, Moore PR, Swallow S, Gouverneur V. Chem Soc Rev. 2008;37:320.
28. Wang J, Sánchez-Roselló M, Aceña JL, et al. Chem Rev. 2014;114:2432.
J
Med Chem.
30. Meyer F. Chem Commun. 2016;52:3077.
31. Roberts JD, Webb RL, McElhill EA. J Am Chem Soc. 1950;72:408.
32. Piers E, Chong J. J Org Chem. 1982;47:1604.
33. Soai K, Yokoyama S, Mochida K. Synthesis (Stuttgart). 1987;647.
34. Ishizumi K, Koga K, Yamada S. Chem Pharm Bull. 1968;16:492.
35. Li Z, Han J, Jiang Y, Browne P, Knox RJ, Hu L. Bioorg Med Chem. 2003;11:4171.
36. Burke PJ, Chun Wong L, Clegg W, et al. Tetrahedron Lett. 2010;51:3918.
29. Gillis EP, Eastman KJ, Hill MD, Donnelly DJ, Meanwell NA.
2015;58:8315.
J Med Chem.
Please cite this article in press as: Burke P.J., et al. Bioorg. Med. Chem. Lett. (2016). http://dx.doi.org/10.1016/j.bmcl.2016.11.024