Prodrug and covalent linker strategies for the solubilization of dual-action antioxidants/iron chelators.

Article abstract of DOI:10.1016/S0960-894X(02)00698-4

Water soluble prodrugs of hybrid free radical scavenger/iron chelating molecules, based on 3,5-disubstituted-4-hydroxyphenyl derivatives and 3-hydroxy-2-methyl-4(1H)-pyridinone (deferiprone), have been prepared. Related hybrid molecules containing a covalent poly(ethylene)glycol or an amine linker were also synthesized.

Full text of DOI:10.1016/S0960-894X(02)00698-4

Bioorganic & Medicinal Chemistry Letters 12 (2002) 3297–3300
Prodrug and Covalent Linker Strategies for the Solubilization of
Dual-Action Antioxidants/Iron Chelators
David Bebbington,y Claire E. Dawson,* Suneel Gaur and John Spencer*
Department of Chemistry, Vernalis Research Limited, Oakdene Court, 613 Reading Road, Winnersh, Wokingham, RG41 5UA, UK
Received 17 May 2002; revised 9 August 2002; accepted 19 August 2002
Abstract—Water soluble prodrugs of hybrid free radical scavenger/iron chelating molecules, based on 3,5-disubstituted-4-
hydroxyphenyl derivatives and 3-hydroxy-2-methyl-4(1H)-pyridinone (deferiprone), have been prepared. Related hybrid molecules
containing a covalent poly(ethylene)glycol or an amine linker were also synthesized.
# 2002 Elsevier Science Ltd. All rights reserved.
The role that oxidative stress plays in the onset of neuro-
degenerative disorders is now well established.1 The
brain is particularly susceptible to oxidative stress since
it has a high consumption of both energy and oxygen,
yet has a relatively limited antioxidant capacity, and
contains high levels of iron and polyunsaturated fatty
acids in its cell membranes. Both antioxidants and iron
chelating molecules have shown neuroprotective efficacy
in animal models of neurodegenerative disorders such as
Alzheimer’s disease, Parkinson’s disease and stroke.2
In an extension to this study we wished to prepare ana-
logues of 1 and 2 with improved water solubility, both
for systemic administration and to eliminate toxic
effects associated with the use of a non-aqueous vehicle,
as 1 (n=1) and 2 (n=0) show a water solubility of less
than 0.1 mg/mL.
Our strategy, summarized in Figure 2, was aimed at
targeting three specific regions of molecules 1 and 2.
Approach (i) would necessitate the synthesis of further
analogues of 1 by the development of a new solubilizing
linker to bridge the iron chelating and radical scaven-
ging entities,5 whereas approaches (ii) and (iii) would
involve the derivatization of either the phenolic or
hydroxy pyridone groups of 1 or 2 respectively via a
prodrug approach.6
Recently, we disclosed novel, dual action, neuroprotec-
tive, hybrid molecules such as 1 and 2, which synergis-
tically combine an iron chelating molecule, such as
deferiprone, with an antioxidant, e.g., BHT or a
hydroxybenzo-furan or -pyran derivative (see Fig. 1).3
Hybrid molecules 1 and 2 were shown to offer protec-
tion against oxidative stress in a model of iodoacetate-
induced cell toxicity in cerebellar granule cells, and in a
model of lipid peroxidation in rat brain homogenates.3,4
For example 1 (n=0) inhibited lipid peroxidation with
an IC50 of 0.3 mM, which represents at the very least a
ca. 10-fold improvement over its separate components,
that is 5.9 mMfor BHT and 3.9 mMfor deferiprone.
Results and Discussion
Poly(ethyleneglycol) (PEG) derivatives have been suc-
cessfully employed in medicinal chemistry, as esters or
ethers, to improve the aqueous solubility and adminis-
tration of drugs.7 To this end we decided to devise syn-
thetic routes to new analogues of 1 containing a
covalently-attached PEG ether linkage, that is defer-
iprone and BHT attached with z=PEG linker
(Approach (i), Fig. 2). One of the key steps in the
synthesis of compounds such as 1 involves the forma-
tion of a BHT derivative bearing a primary amine. The
amine is subsequently condensed with benzyl maltol,8 to
afford, after deprotection, the final hybrid BHT-defer-
iprone molecule. Using this approach, we synthesized
*Corresponding author at current address: The James Black Founda-
tion, 68 Half Moon Lane, Dulwich SE24 9JE, UK. Tel.: +44-207-737-
8282; fax: +44-207-274-9687; e-mail: john.spencer@kcl.ac.uk (J.
Spencer); c.dawson@vernalis.com (C. E. Dawson).
yCurrent address: Vertex Pharmaceuticals (Europe) Ltd., 88 Milton
Park, Abingdon, Oxfordshire OX14 4RY, UK.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00698-4

3298
D. Bebbington et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3297–3300
Figure 1. Hybrid iron chelator radical scavengers.
Figure 2. Derivatization of 1 and 2 using prodrug and new linker
strategies.
the PEG-containing amines 6 and 7 (see Scheme 1). This
protocol necessitated suitable O-protection of 3a, as the
free phenol itself did not react cleanly with poly
(ethyleneglycols). Gratifyingly, the use of a recently
disclosed procedure for the protection of hindered phe-
nols with a tert-butoxycarbonyl (Boc) protecting group
was successful,9 and the carbonate 3b was formed in
87% yield.
Scheme 1. Reagents and conditions: (a) (Boc)2O, DMAP, heptane,
87%; (b) NaH, DMF, HO(CH2CH2O)nCH2CH2OH, 0 ꢀC then 90 ꢀC,
37% (n=3), 41% (n=4); (c) (i) NEt3, MsCl, CH2Cl2; (ii) NaN3, DM F,
90 ꢀC; (iii) PtO2, EtOH, H2, 1 atm, 86% (n=3), 68% (n=4); (d) (i)
benzyl maltol, EtOH, H2O, reflux; (ii) Pd/C, EtOH, H2, 1 atm; (iii)
TFA, 7% (n=3), 9% (n=4) for last 3 steps.
The ensuing reactions were straightforward: displace-
ment of the iodide from 3b by the sodium salts of either
tetraethylene or pentaethylene glycol yielded the expec-
ted alcohols 4 and 5 respectively, along with some elim-
ination (-HI) product. Mesylation of alcohols 4 and 5,
followed by displacement with sodium azide and then
reduction furnished amines 6 and 7. These amines
underwent condensation reactions with benzyl maltol,
which after deprotection gave the requisite hybrid
molecules 8 and 9. Although the yields for the final
maltol condensation reactions were low (less than 10%),
ca. 100 mg samples of 8 and 9 were readily prepared.10
However, 8 and 9 were found to be largely insoluble in
water, despite the presence of up to five oxygens in the
linker group, so this led us to consider shorter linkers
with basic centres.
maltol and finally debenzylation. Compounds 12b and
13 showed excellent activities for lipid peroxidation
(IC50 of 0.2 and 0.1 mMfor 12b and 13 respectively), but
limited aqueous solubilities, although the former was
soluble in a DMSO/H2O (1:9) mixture.
Our next approach required the formation of ester pro-
drugs by functionalising a hydroxyl group of 1 or 2, that
is strategies (ii) and (iii) as outlined in Figure 2. The
ester prodrug approach is well established, as exempli-
fied by the formation of water soluble prodrugs of
Vitamin E.11 Thus, the phenolic hydroxyl of hybrid
molecules 2 was esterified to prepare aminoalkyl- and
aminoaryl-carboxylic acid esters. Initial attempts to
functionalise the phenolic hydroxyl of 14 (Scheme 3),
using DCC and N-Boc b-alanine in pyridine, as descri-
bed for the preparation of Vitamin E prodrugs, were
unsuccessful. However, treatment of 14 with potassium
carbonate in acetone followed by the addition of
N-hydroxysuccinimide (HOSu) esters of various N-Boc
protected amino acids resulted in esterification in yields
of greater than 76%. Following deprotection com-
pounds 15a–e were isolated. Compound 15f was pre-
pared by treating a solution of 14 in pyridine at reflux
The alternative system 12b, with a simple amine linker,
was synthesized by a low-yielding reductive amination
reaction of benzaldehyde 10 and 1-(2-aminoethyl)
deferiprone 11, followed by debenzylation (Scheme 2).
The related amine linked derivative 13 was synthesized
by a five-step process: a coupling reaction between Tro-
lox1 and a N-Boc protected diamine, deprotection,
amide reduction followed by condensation with benzyl

D. Bebbington et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3297–3300
3299
reaction of 16 with HOSu esters of N-Boc protected
amino acids, without the need for protection of the
phenolic group,13 except for 17c which was prepared by
treating 16 with acetic anhydride in pyridine.
The water solubilities of the prodrugs and the hetero-
atom linked hybrids are summarized in Table 1. All of
the hydrochloride salts of the aminoalkyl carboxylic
acid ester prodrugs (15a, 15c–f and 17a) showed good
solubility in water, that is at least 10 mg/mL. The two
aminoaryl carboxylic acid esters (15b and 17b) showed
little water solubility, as did the PEG linked hybrid
molecules 8 and 9. The simple acetate prodrug 17c was
also very sparingly soluble.
Having achieved our primary objective of synthesizing
water soluble, hybrid molecules, it was then necessary to
explore the usefulness of the prodrugs by measuring
their ease of hydrolysis under physiological conditions,
and their stability as an aqueous formulation. Pre-
liminary experiments have shown that 15c underwent
almost complete hydrolysis to 2 (n=0) on incubation in
rat plasma at 37 ꢀC for 3 h, and yet during this time
remained unchanged as a solution in phosphate buffer
(pH 7.4). Under the same conditions 15a and 15d were
hydrolysed to a lesser extent in rat plasma (approximately
20–30%), and were also stable to the phosphate buffer.
Scheme 2. Reagents and conditions: (a) NaCNBH3, MeOH, 13%; (b)
Pd/C, EtOH, H2, 1 atm, 96%; (c) (i) EDCI, HOBt, DIPEA,
BocHN(CH2)4NH2, DMF, 72%; (ii) TFA, CH2Cl2, 60%; (iii) LiAlH4,
THF, reflux, 57%; (d) (i) benzyl maltol, EtOH, H2O, reflux, 21%; (ii)
Pd/C, EtOH, H2, 1 atm, 96%.
Conclusions
with N,N-dimethylglycinyl chloride HCl and in a yield
of 54%. The water solubility of the hydrochloride salts
of 15a–f was determined.12
Three approaches to improve the water solubilization of
dual action, hybrid molecules 1 and 2 have been descri-
bed. Covalently-attached PEG and amine linkers
offered little improvement towards increasing the water
solubility. However, the prodrug strategy, as exempli-
fied by attaching amino acids via ester linkages, led to
substantial improvements in the water solubility of the
hybrid molecules. Early experiments in vitro have
demonstrated that the ester can be hydrolysed relatively
efficiently, and thus the prodrug approach is a viable
In the final approach, method (iii) in Figure 2, hybrid
derivatives 16 were functionalized by the chemoselective
acylation of the pyridone 3-hydroxyl group to prepare
esters 17 (Scheme 3). Esterification was also achieved by
Table 1. Water solubility
Compd
Acylating agentc
R
Solubility
in water
(mg/mL)
8a
—
—
—
—
—
—
—
—
<0.1
<0.1
10–20d
<0.1
10–20
<0.1
>20
10–20
>20
>20
9a
12bb
13b
15ab
15ba
15cb
15db
15eb
15fb
17ab
17ba
17ca
Boc-b-Ala-OSu
BocHNpC6H4C(O)OSu
Boc-Sar-OSu
Boc-Gaba-OSu
(S)-Boc-Val-OSu
DMG-Cl.HCl
Boc-b-Ala-OSu
BocHNpC6H4C(O)OSu
Ac2O
CH2CH2NH2
p-C6H4NH2
CH2NHCH3
CH2CH2CH2NH2
CH(i-Pr)NH2
CH2N(CH3)2
CH2CH2NH2
p-C6H4NH2
CH3
>20
<0.1
<0.1
amono.HCl salt. Boc removal by HCl treatments.
bbis.HCl salt. Boc removal by HCl treatments.
cAla, alanine; Sar, sarcosine; Gaba, g-aminobutyric acid; Val, valine;
DMG-Cl, N,N-dimethylglycinyl chloride.
dIn DMSO/H2O (1:9).
Scheme 3. Reagents and conditions: (a) K2CO3, acetone, acylating
agent (see Table 1), reflux; (b) Pd/C, EtOH, H2, 1 atm; (c) acylating
agent (see Table 1), pyridine, reflux, HCl treatment.

3300
D. Bebbington et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3297–3300
strategy as these molecules could be formulated as
10 mg/mL aqueous solutions suitable for administration
in vivo.14
Hider, R. C. J. Med. Chem. 1998, 41, 3347. (b) Dobbin, P. S.;
Hider, R. C.; Hall, A. D.; Taylor, P. D.; Sarpong, P.; Poter,
J. B.; Xiao, G.; van der Helm, D. J. Med. Chem. 1993, 36,
2448.
9. Hansen, M. M.; Riggs, J. R. Tetrahedron Lett. 1998, 39,
2705.
10. Full experimental details for the synthesis of 3–13 and
analytical data are provided in ref 3(b).
Acknowledgements
Dr. Alan Palmer is acknowledged for help in the
manuscript preparation. David Rush is thanked for the
stability and plasma hydrolysis studies. Members of the
Department of Analytical Chemistry, Vernalis Research
Ltd., are thanked for their assistance.
11. (a) Takata, J.; Ito, S.; Karube, Y.; Nagata, Y.; Matsush-
ima, Y. Biol. Pharm. Bull. 1997, 20, 204. (b) Takata, J.; Kar-
ube, Y.; Nagata, Y.; Matsushima, Y. J. Pharm. Sci. 1995, 84,
96.
12. Determination of water solubility. To 5 mg of the test
compound in a clear glass vial was added 0.25 mL HPLC
grade water and the vial was shaken vigorously for 1 h at room
temperature. The sample was viewed by visual inspection to
determine whether any particles of compound remained and
also the sample was viewed between crossed polarising filters
in front of a light source. If undissolved compound remained
then a second portion of 0.25 mL HPLC grade water was
added and the process repeated until complete dissolution was
apparent. Although a more accurate determination of water
solubility could be obtained from analysing aqueous solutions
by HPLC, the chromatograms obtained from these molecules
are characterised by broad, asymmetric peak shapes due to the
iron chelating ability of these molecules.
13. We did not attempt to derivatize the phenolic group in 14,
although there is precedent for prodrug formation at hindered
phenols as in the related anaesthetic 2,6-diisopropylphenol,
Propofol. See: Trapani, G.; Latrofa, A.; Franco, M.; Lope-
dota, A. Maciocco, Liso, G. Int. J. Pharm. 1998, 175, 195.
14. (a) The investigation of dual action molecules for neuro-
protection is a very active field of research at present: Jarrott, B.;
Callaway, J. K.; Jackson, W. R.; Beart, P. M. Drug Dev. Res.
1999, 46, 261. (b) Ohkawa, S.; Fukatsu, K.; Miki, S.; Hashi-
moto, T.; Sakamoto, J.; Doi, T.; Nagai, Y.; Aono, T. J. Med.
Chem. 1997, 40, 559. (c) Chabrier, P. E.; Auguet, A.; Spinne-
References and Notes
1. (a) Marciniak, G.; Petty, M. A. Drugs Future 1996, 21,
1037. (b) Butterfield, D. A.; Howard, B. J.; LaFontaine, M. A.
Curr. Med. Chem. 2001, 8, 815. (c) Pitchumoni, S. S.; Dor-
aiswamy, P. M. J. Am. Geriatrics Soc. 1998, 46, 1566.
2. See, for example: (a) Hunter, A. J.; Green, A. R.; Cross,
A. J. Trends Pharmacol. Sci. 1995, 16, 123. (b) Norris, J. W.;
Hachinski, V. Eds.; Stroke Prevention; Oxford, 1999. (c) Sano,
M.; Ernesto, C.; Thomas, R. G.; Klauber, M. R.; Schafer, K.;
Grundman, M.; Woodbury, P.; Growdon, J.; Cotman, C. W.;
Pfeiffer, E.; Schneider, L. S.; Thal, L. J. New Engl. J. Med.
1997, 336, 1216. (d) Clark, W. M.; Rinker, L. G.; Lessov,
N. S.; Lowery, S. L.; Cipolla, M. J. Stroke 2001, 32, 1000.
3. (a) Bebbington, D.; Monck, N. J. T.; Gaur, S.; Palmer,
A. M.; Benwell, K.; Harvey, V.; Malcolm, C. S.; Porter,
R. H. P. J. Med. Chem. 2000, 43, 2779. (b) See, for example:
EP 1006108A1; Chem. Abstr. 2000, 133, 17385. WO 9923075;
Chem. Abstr. 1999, 130, 338022.
4. Malcolm, C. S.; Benwell, K. R.; Lamb, H.; Bebbington, D.;
Porter, R. H. P. Free Radical Biol. Med. 2000, 28, 102.
5. Review: Wermuth, C. G. In The Practice of Medicinal
Chemistry; Wermuth, C. G., Ed.; Academic: London, 1996; p
755.
6. Review: Wermuth, C. G.; Gaignault, J.; Marchandeau, C.
In The Practice of Medicinal Chemistry; Wermuth, C. G., Ed.;
Academic: London, 1996; p 671.
wyn, B.; Auvin, S.; Cornet, S.; Demerle-Pallardy, C.; Guil-
´
mard-Favre, C.; Marin, J.-G.; Pignol, B.; Gillard Roubert, V.;
Roussillot-Charnet, C.; Schulz, J.; Viossat, I.; Bigg, D.; Mon-
cada, S. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 10824. (d)
Giblin, G. M. P.; Box, P. C.; Campbell, I. C.; Hancock, A. P.;
Roomans, S.; Mills, G. I.; Molloy, C.; Tranter, G. E.; Walker,
A. L.; Doctrow, S. R.; Huffman, K.; Malfroy, B. Bioorg. Med.
Chem. Lett. 2001, 11, 1367. (e) Naughton, D. P.; Grootveld,
M. Bioorg. Med. Chem. Lett. 2001, 11, 2573.
7. See, for example: WO9807713; Chem. Abstr. 1998, 128,
217533. WO9324476; Chem. Abstr. 1993, 120, 144136.
8. (a) Rai, B. L.; Dekhordi, L. S.; Khodr, H.; Jin, Y.; Liu, Z.;