Hypoxia-driven elimination of thiopurines from their nitrobenzyl prodrugs

Article abstract of DOI:10.1016/j.bmcl.2007.05.018

A novel bioreductive prodrug of 6-thioguanine, 2-amino-6-[2-(4-nitrophenyl)prop-2-ylsulfanyl]-9H-purine, containing a gem-dimethyl thioether linkage, was synthesised and compared with its unsubstituted analogue. In A549 whole cell experiments hypoxia sele

Full text of DOI:10.1016/j.bmcl.2007.05.018

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4320–4322
Hypoxia-driven elimination of thiopurines from
their nitrobenzyl prodrugs
a,
a
a
a
*
Peter Thomson, Matthew A. Naylor, Michael R. L. Stratford, Gemma Lewis,
a
a
a
b
Sally Hill, Kantilal B. Patel, Peter Wardman and Peter D. Davis
a
University of Oxford, Gray Cancer Institute, Mount Vernon Hospital, Northwood, Middlesex HA6 2JR, UK
b
Angiogene Pharmaceuticals Ltd, The Magdalen Centre, Oxford Science Park, Oxon OX4 4GA, UK
Received 23 March 2007; revised 5 May 2007; accepted 9 May 2007
Available online 13 May 2007
Abstract—A novel bioreductive prodrug of 6-thioguanine, 2-amino-6-[2-(4-nitrophenyl)prop-2-ylsulfanyl]-9H-purine, containing a
gem-dimethyl thioether linkage, was synthesised and compared with its unsubstituted analogue. In A549 whole cell experiments
hypoxia selective release of 6-thioguanine was observed with the substituted prodrug only.
ꢀ
2007 Elsevier Ltd. All rights reserved.
Tumour hypoxia is a potentially exploitable factor for
1
the treatment of many solid tumour types. Hypoxia
being reduced, and fragment upon reduction. There
has been much interest in the use of quinoid and nitro-
aromatic ‘triggers’ because such compounds possess the
occurs as a result of poor blood supply, and therefore
oxygenation, due to the chaotic network of the neovas-
culature in growing tumours. It is well documented that
tumour hypoxia is directly correlated with disease pro-
gression and patient survival, with higher levels of hy-
7
,10
necessary functionality for reduction.
reported the synthesis and biological properties of nitro-
Recently, we
1
1
thiophene derived prodrugs of combretastatin A4. Our
findings show that inclusion of a gem-dimethyl substi-
tuted linkage between the active drug and trigger im-
proves a number of properties including metabolic
stability, prodrug fragmentation rate and elimination
efficiency.
2
poxia leading to generally poorer prognoses.
Unfortunately, tumour hypoxia protects cells from
radiotherapy, photodynamic therapy and some cyto-
toxic reagents. There is thus a clear need for new agents
that can target hypoxic regions and even utilise hypoxia
to their advantage.
The antileukaemic drugs 6-thioguanine (6-TG) and 6-
mercaptopurine (6-MP) show little clinical activity
against solid tumours despite broad in vitro cytotoxic-
ity. Such limitations probably result from their short
elimination half-lives, rapid cellular metabolism and rel-
ative hydrophilicity, which together would restrict pene-
tration into solid tumour tissue. In humans, catabolism
takes place largely via an inactivating sulfur methyla-
Many studies have been carried out using bioreductive
3
–9
drug targeting.
One approach utilises a prodrug
which delivers an agent, for example a cytotoxin, via
reductive elimination from a non-toxic parent molecule.
Reduction and prodrug fragmentation occurs within
regions of low oxygen concentration but is absent in
healthy, aerobic tissues. This process can be driven by
a number of ‘reductase’ enzymes, including cytochrome
P450, xanthine oxidase and NADPH-cytochrome P450
reductase. The prodrug must therefore be a substrate
for the enzyme, possess a chemical function capable of
1
2,13
tion.
We reasoned that prodrugs with enhanced met-
abolic stability and increased lipophilicity might be
better able to penetrate tumour tissue and deliver the
active cytotoxic agents to tumour cells unreached by
systemic administration of the parent drugs.
Previously, Kirkpatrick et al. reported the synthesis of 6-
1
4
(
4-nitrobenzyl)thioguanine (1). Biological evaluation
Keywords: Hypoxia; Nitrobenzyl; Thiopurine; Targeting; Delivery;
Tumor; A549; Thioguanine; Mercaptopurine.
of this compound in aerobic and hypoxic EMT6 tumour
cells failed to show an advantage over 6-thioguanine.
They demonstrated, by isolating the reduction product
*
Corresponding author. Tel.: +44 01923 828611; fax: +44 01923
35210; e-mail: thomson@gci.ac.uk
8
0
960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.05.018

P. Thomson et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4320–4322
4321
Scheme 1. Synthesis of thiopurine prodrugs.
6
-(4-aminobenzyl)thioguanine via chemical hydrogena-
combined with stabilisation of the transition state elec-
tron density by the alkyl groups.
tion, that the thioether linkage was stable to fragmenta-
tion. We therefore sought to use our gem-disubstitution
technology to improve the delivery of thioguanine to tu-
mours whilst capitalising on the inherent lower toxicity
of 6-alkyl analogues.
Following successful radical fragmentation studies, the
6-TG prodrugs were evaluated in A549 whole cell
suspension experiments. These were performed under
anoxic (N ) and aerobic conditions to ascertain the rel-
2
During our studies with combretastatin A4, we synthes-
ised mono-methyl, gem-dimethyl and unsubstituted ana-
logues for comparison. In this thiopurine investigation,
we decided to omit the mono-methyl analogue as the
beneficial properties imbued by the gem-dimethyl linker
in the combretastatin A4 series, such as better elimina-
tion efficiency, lower radical half-life and absence of a
chiral centre, are all attractive properties for drug design
and biological evaluation. Prodrugs were synthesised in
variable yields (9–81%) by nucleophilic substitution of 4-
nitrobenzyl bromide or 2-bromo-2-(4-nitrophenyl)pro-
ative rates and efficiencies of bioreductive elimination of
thiopurines from the prodrugs. It is known that the
A549 cell line has an appreciable level of P450 reduc-
1
6,2
tase
and, as a human lung cancer cell line, affords
an attractive opportunity for eventual clinical targeting.
Table 2 summarises the improved properties of 2. A ten-
fold increase in prodrug metabolism under anoxia com-
pared to air is noted for the lead compound whereas a
much lower increase is seen for the unsubstituted 1.
Interestingly, the loss of 1 is not matched with produc-
tion of 6-TG, unlike 2 which demonstrates excellent pro-
duction under anoxia (Fig. 1). It is evident that the loss
of 1 coupled with zero production of 6-TG indicates a
metabolic pathway other than reductive elimination
may be occurring. This further demonstrates the impor-
tance of the gem-dimethyl linker for abrogating
unwanted metabolic events. As expected, both prodrugs
fail to be metabolised or release 6-TG under oxic
conditions.
1
5
pane with 6-thioguanine in DMF using sodium
hydride as the base (Scheme 1). The compounds were
1
identified by H NMR, mass spectrometry and elemen-
tal analysis.
Pulse and steady state c-radiolysis experiments were
used to measure the radical anion half-lives and frag-
mentation efficiency, respectively (Table 1). Incorpora-
tion of the gem-dimethyl linker significantly lowered
the radical half-life of the substituted prodrugs 2 and 3
compared to the unsubstituted 1 (half-life not measur-
able within the timeframe of the experiment). The
importance of the gem-dimethyl group for successful
drug release is underlined with the sixfold increase in
fragmentation efficiency over the des-methyl analogue.
These characteristics might be attributed to a drive to re-
lease steric strain in the fragmentation transition state
In conclusion, we have successfully synthesised thiopu-
rine prodrugs and tested our gem-disubstitution techno-
1
4
logy. Previous work has shown nitrobenzyl conjugated
6-TG prodrugs to be stable under hypoxia, but we have
demonstrated that, with our improved dimethyl linker
Table 2. Metabolism of prodrugs in A549 whole cells
Compound
Rate of loss of
prodrug/nM min
Rate of 6-TG
production/
À1
Table 1. Characterisation of radical fragmentation and drug release
efficiency using radiolytic production of radicals
À1
nM min
1
2
Compound
R
R
Drug Radical
% Fragmentation
Air
N
2
Air
N
2
a
half-life/ms efficiency
b
1
2
0
9.5 ± 2.5
39 ± 6
100 ± 6
1.7 ± 0.4
1.7 ± 0.4
0
72 ± 4
c
—
1
2
3
H
H
6-TG
11
67
66
Me Me 6-TG 29
Me Me 6-MP 18
Compounds 1 or 2 was dissolved in DMSO to a concentration of
25 lmol/L, and 80 lL was added to 10 mL A549 cells in Eagle’s
MEM (8.19 · 10 cells/mL) to give a final prodrug concentration of
5 lmol/L and incubated at 37 ꢁC. For anoxic experiments, the mixture
was degassed with N for 30 min before prodrug addition and then
2
overgassed with N during incubation. Samples (1 mL) were added to
6
a
6
Pulse radiolysis done in N
9.2) using a 6-MeV linear accelerator.
2
O-saturated 50% 2-propanol/water (pH
b
Steady-state c-radiolysis determined in N
panol/water (pH 7.4); dose rate, 3.26 Gy min ; % fragmentation
2
O-saturated 50% 2-pro-
À1
2
efficiency = G(drug)/G(Me
Not measurable.
2
COH) · 100.
100 mmol/L hydrochloric acid (0.2 mL), mixed and then extracted (via
solid-phase extraction) before HPLC analysis.
c

4322
P. Thomson et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4320–4322
(
water then brine), dried (MgSO ) and evaporated.
4
2
2
1
1
5
0
5
0
5
Flash chromatography, eluting with 50% ethyl acetate/
hexane then 100% ethyl acetate and finally 25% metha-
nol/ethyl acetate, afforded an oily, yellow solid. The res-
idue was triturated with acetonitrile and the suspension
filtered. The filtrate was concentrated in vacuo then trit-
urated with ether, and afforded an orange powder. This
was recrystallized from water/methanol and furnished a
white powder (150 mg, 11%); TLC R = 0.19, ethyl ace-
f
1
tate; mp 189–191 ꢁC; H NMR (500 MHz, DMSO) d
8
6
HPLC-RT 4.08 min (TFA 20–50%); MS (m/z, %) 330
(
.12 (2H, d, J = 7.0, ArH), 7.96 (1H, d, J = 7.0, ArH),
.13 (1H, s, N@CH), 2.51 (6H, s, 2 · CH ) ppm;
Prodrug 1
3
+
M , 4%), 167 (100%), 163 (13%), 133 (26%); analysis
C H N O S Æ1/2CH OHÆ1/2H O calcd C, 49.01; H,
1
4
14
6
2
1
3
2
4
.79; N, 23.66; found C, 48.88; H, 5.12; N 23.58.
0
5
2
2
1
1
Acknowledgment
The authors thank Angiogene Pharmaceuticals Ltd and
Cancer Research UK for generous financial assistance.
0
5
0
5
0
References and notes
1
2
. Hockel, M.; Vaupel, P. J. Natl. Cancer Inst. 2001, 93, 266.
. Yu, L. J.; Matias, J.; Scudiero, D. A.; Hite, K. M.; Monks,
A.; Sausville, E. A.; Waxman, D. J. Drug Metab. Dispos.
Prodrug 2
2
001, 29, 304.
3
4
. Hay, M. P.; Atwell, G. J.; Wilson, W. R.; Pullen, S. M.;
Denny, W. A. J. Med. Chem. 2003, 46, 2456.
. Hay, M. P.; Wilson, W. R.; Denny, W. A. Bioorg. Med.
Chem. 2005, 13, 4043.
5. Jaffar, M.; Abou-Zeid, N.; Bai, L.; Mrema, I.; Robinson,
I.; Tanner, R.; Stratford, I. J. Curr. Drug. Del. 2004, 1,
0
50
100
150
3
45.
Time / min
6. Mulchahy, R. T.; Gipp, J. J.; Schmidt, J. P.; Joswig, C.;
Borch, R. F. J. Med. Chem. 1994, 37, 1610.
7. Naylor, M. A.; Thomson, P. Mini-Rev. Med. Chem. 2001,
1, 17.
Figure 1. Loss of prodrug under air (s) and N
-TG under air (h) and N (j).
2
(d), and production of
6
2
8
. Parveen, I.; Naughton, D. P.; Whish, W. J. D.; Threadgill,
M. D. Bioorg. Med. Chem. Lett. 1999, 9, 2031.
. Nagasawa, H.; Uto, Y.; Kirk, K. L.; Hori, H. Biol. Pharm.
Bull. 2006, 29, 2335.
9
group incorporated into the prodrug, it is possible to
release the active agent using this drug delivery strategy.
We are currently evaluating this series of prodrugs in
further experiments.
1
0. Wardman, P. Curr. Med. Chem. 2001, 8, 739.
11. Thomson, P.; Naylor, M. A.; Everett, S. A.; Stratford, M.
R. L.; Lewis, G.; Hill, S.; Patel, K. B.; Wardman, P.;
Davis, P. D. Mol. Cancer Therap. 2006, 5, 2886.
1
1
2. Elgemeie, G. H. Curr. Pharm. Des. 2003, 9, 2627.
3. Coulthard, S.; Hogarth, L. Invest. New Drugs 2005, 23,
Synthetic example. 2-Amino-6-[2-(4-nitrophenyl)prop-2-
ylsulfanyl]-9H-purine (2). Sodium hydride (160 mg,
5
23.
4
4
.0 mmol) was added to 6-thioguanine (668 mg,
.0 mmol) in DMF (15 mL). After 1 h, 2-bromo-2-(4-
1
1
1
4. Kirkpatrick, D. L.; Johnson, K. E.; Sartorelli, A. C.
J. Med. Chem. 1986, 29, 2048.
5. Paine, S. W.; Ridd, J. H. J. Chem. Soc., Perkin Trans. 2
nitrophenyl)propane (1.08 g, 4.4 mmol) in DMF
5 mL) was added and the reaction was stirred for
4 h. The reaction mixture was partitioned (ethyl acetate
(
2
1
996, 2571.
6. Fitzsimmons, S. A.; Workman, P.; Grever, M.; Paull, K.;
Camalier, R.; Lewis, A. D. J. Natl. Cancer Inst. 1996, 88,
259.
and brine), the aqueous phase was extracted (ethyl ace-
tate); the organic phases were combined then washed