Synthesis, physicochemical characterization, and biological evaluation of 2-(1'-hydroxyalkyl-3-hydroxypyridin-4-ones: Novel iron chelators with enhanced pFe3+ values

Article abstract of DOI:10.1021/jm991080o

The synthesis of a range of 2-(1'-hydroxyalkyl)-3-hydroxypyridin-4-ones as bidentate iron(III) chelators with potential for oral administration is described. The pK(a) values of the ligands and the stability constants of their iron(III) complexes have been determined. Results indicate that the introduction of a 1'-hydroxyalkyl group at the 2-position leads to a significant improvement in the pFe³⁺ values. Such an effect was found to be greater with the hydroxyethyl substituent than with the hydroxy-methyl substituent, particularly in the cases of 1-ethyl-2(1'-hydroxyethyl)-3- hydroxypyridin-4-one (pFe³⁺ = 21.4) and 1,6-dimethyl-2-(1'-hydroxyethyl)3- hydroxypyridin-4-one (pFe³⁺ = 21.5) where an enhancement on pFe³⁺ values in the region of two orders of magnitude is observed, as compared with Deferiprone (1,2-dimethyl-3-hydroxypyridin-4-one) (pFe³⁺ = 19.4). The ability of these novel 3-hydroxypyridin-4-ones to facilitate the iron excretion in bile was investigated using a [⁵⁹Fe] ferritin-loaded rat model. Chelators and prodrug chelators possessing high pFe³⁺ values show great promise in their ability to remove iron under in vivo conditions.

Full text of DOI:10.1021/jm991080o

4814
J . Med. Chem. 1999, 42, 4814-4823
Syn th esis, P h ysicoch em ica l Ch a r a cter iza tion , a n d Biologica l Eva lu a tion of
2-(1′-Hyd r oxya lk yl)-3-h yd r oxyp yr id in -4-on es: Novel Ir on Ch ela tor s w ith
En h a n ced p F e3+ Va lu es†
Zu D. Liu, Hicham H. Khodr, Ding Y. Liu, Shu L. Lu, and Robert C. Hider*
Department of Pharmacy, King’s College London, 150 Stamford Street, London SE1 8AW, U.K.
Received May 17, 1999
The synthesis of a range of 2-(1′-hydroxyalkyl)-3-hydroxypyridin-4-ones as bidentate iron(III)
chelators with potential for oral administration is described. The pKa values of the ligands
and the stability constants of their iron(III) complexes have been determined. Results indicate
that the introduction of a 1′-hydroxyalkyl group at the 2-position leads to a significant
improvement in the pFe3+ values. Such an effect was found to be greater with the hydroxyethyl
substituent than with the hydroxymethyl substituent, particularly in the cases of 1-ethyl-2-
(1′-hydroxyethyl)-3-hydroxypyridin-4-one (pFe3+ ) 21.4) and 1,6-dimethyl-2-(1′-hydroxyethyl)-
3-hydroxypyridin-4-one (pFe3+ ) 21.5) where an enhancement on pFe3+ values in the region of
two orders of magnitude is observed, as compared with Deferiprone (1,2-dimethyl-3-hydroxy-
pyridin-4-one) (pFe3+ ) 19.4). The ability of these novel 3-hydroxypyridin-4-ones to facilitate
the iron excretion in bile was investigated using a [59Fe]ferritin-loaded rat model. Chelators
and prodrug chelators possessing high pFe3+ values show great promise in their ability to
remove iron under in vivo conditions.
In tr od u ction
>44% and >85% of the administered dose is recovered
in the urine as the non-chelating 3-O-glucuronide
conjugate.5 This extensive biotransformation limits its
ability to mobilize excess body iron in thalassemic
patients.
The most frequent treatment of inherited hemato-
logical diseases such as â-thalassemia major is to
maintain high levels of hemoglobin by regular blood
transfusion. Repeated transfusion leads to elevated body
iron levels due to the inability of humans to excrete iron
via the kidney. Excess iron is mainly located within the
liver and other highly perfused organs leading to tissue
damage, organ failure, and eventually death.1 Compli-
cations associated with elevated iron levels can be
largely avoided by the use of iron-specific chelating
agents and in particular desferrioxamine (DFO). Un-
fortunately, DFO lacks oral activity, and this has a
dramatic influence on patient compliance.2
3-Hydroxypyridin-4-ones (HPOs) are currently one of
the main candidates for the development of orally active
iron chelators.3 Indeed, the 1,2-dimethyl derivative
CP20 (Deferiprone, L1) (1) (Table 1) is currently in
clinical trials. Unfortunately, the dose required to keep
a previously well-chelated patient in negative iron
balance appears to be relatively high. Not surprisingly
therefore, side effects have been observed in some
patients receiving CP20.4 One of the major reasons for
the limited efficacy of CP20 in clinical use is that it
undergoes extensive phase II metabolism in the liver.
The 3-hydroxyl functionality, which is crucial for scav-
enging iron, is also a prime target for phase II conjuga-
tion. Urinary recovery studies conducted on CP20 in
both rats and humans have shown that respectively
The 1,2-diethyl analogue CP94 (2) (Table 2) has also
been widely investigated.6-9 This chelator was found to
be much more efficient at iron removal than CP20 in
several mammalian species, e.g. mice,6 rat,7 and pri-
mate,8 whereas it was ineffective at mobilizing iron in
guinea pig7 and human9 due to extensive 3-O-glucu-
ronide conjugation. One of the presumed reasons for the
greater efficacy of CP94 in the rat is its unusual phase
I metabolic pathway which leads to the formation of the
2-(1′-hydroxyethyl) metabolite 3 (Scheme 1).5 This me-
tabolite does not undergo further phase II metabolism
to form a glucuronide conjugate and hence retains the
ability to chelate iron. However such an explanation
may well be an oversimplification since other 1-hy-
droxyalky HPOs, namely 4 and 5 (Table 1), are also not
extensively metabolised via phase II reactions10 and yet
their iron mobilization efficacies in rats are much less
than that of CP94.11 The same observation holds for
their ester prodrugs such as 6 and 7 (Table 1). Thus it
was decided to investigate the properties of the 2-(1′-
hydroxyethyl) metabolite of CP94. In this work we
describe the synthesis, physical-chemical properties,
and in vivo iron mobilization efficacies of a range of
novel 2-(1′-hydroxyalkyl) derivatives of 3-hydroxypyri-
din-4-ones.
* Corresponding author: Robert C. Hider. Tel: 20 7848 4646. Fax:
20 7848 4195. E-mail: robert.hider@kcl.ac.uk.
Ch em istr y
† Abbreviations: pFe3+, the negative logarithm of the concentra-
tion of the free iron(III) in solution, calculated for total [ligand] ) 10-5
M, total [iron] ) 10-6 M at pH 7.4; DFO, desferrioxamine; HPO,
hydroxypyridinone; CP20, 1,2-dimethyl-3-hydroxypyridin-4-one; CP94,
1,2-diethyl-3-hydroxypyridin-4-one; D7.4, distribution coefficient at pH
7.4; MOPS, 4-morpholinopropanesulfonic acid;
A furfuryl alcohol-based method has been developed
for the preparation of the key starting material, py-
romeconic acid (12) (Scheme 2). Bromination of furfuryl
alcohol (8) in buffered methanol leads to the formation
10.1021/jm991080o CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/28/1999

2-(1′-Hydroxyalkyl)-3-hydroxypyridin-3-ones
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23 4815
Ta ble 1. Chemical Structure of Selected 3-Hydroxypyridin-4-ones
compd
R1
R2
MW
D7.4
log P
efficacy (%)a
CP20 (1)
CP94 (2)
CH3
CH2CH3
CH2CH3
CH2CH2OH
(CH2)3OH
CH3
139
167
183
183
183
267
287
0.17
1.78
0.26
0.22
0.13
14.5
17.6
-0.77
0.25
-0.58
-0.66
-0.89
1.16
9.5
55.8
50.6
12.9
26.0
19.1
32.5
CH2CH3
CH(OH)CH3
CH2CH3
CH3
CH2CH3
CH3
3
4
5
6
7
CH2CH2OOC(CH3)3
(CH2)3OOCC6H5
1.25
a
Iron mobilization efficacy (%) of the chelators was measured on the [59Fe]ferritin-loaded rat model.11
Sch em e 1. Metabolism of CP94 via Phase I and Phase
Sch em e 3. Synthesis of Allomaltol (15)
II Metabolic Pathwaysa
Sch em e 4. Synthesis of
2-(1′-Hydroxyalkyl)-3-hydroxypyran-4-ones 16a -16e
6-methoxy-2H-pyran-3(6H)-one (10), which is then
smoothly converted to the 4-bromo-6-methoxy-2H-py-
ran-3(6H)-one (11) in a good yield (82%) by treatment
with 1 equiv of bromine and 1 equiv of an organic base,
e.g. triethylamine. The conversion of 10 to 11 may
proceed via the labile 4,5-dibromo derivative 11a which
has not been isolated in this sequence. Pyromeconic acid
(12) can then be prepared from the acid-induced rear-
rangement of 11 using concentrated trifluoroacetic acid
(yield: 80%).
a
The 2-1′-hydroxylated metabolite (3) is the major metabolite
in rat, whereas the phase II glucuronidation is the major metabo-
lism of CP94 in humans.
The synthesis of pyromeconic acid (12) by one-pot
reaction starting from furfuryl alcohol (8) has also been
attempted; however extensive decomposition was ob-
served, resulting in poor yields (less than 20%) (unpub-
lished data).
Sch em e 2. Synthesis of Pyromeconic Acid (12) from
Furfuryl Alcohol
The synthesis of allomaltol (15) can be readily ac-
complished using commercially available kojic acid (13)-
in a two-step reaction (Scheme 3). The 2-position of both
pyromeconic acid (12) and allomaltol (15) can be func-
tionalised in an analogous fashion to the aldol conden-
sation whereby the pyrone anion aldehyde, under
alkaline aqueous conditions, furnishes, on acidic work-
up, the 2-(1′-hydroxyalkyl)-3-hydroxypyran-4-one (16)
in high yields (80-90%) (Scheme 4). The pH of the
reaction solution was found to be critical, since highly
alkaline conditions resulted in extensive aldehyde po-
lymerization. The optimal pH for this reaction was
found to be approximately 10.5.
of a mixture of the cis and trans isomers of 2,5-dihydro-
2,5-dimethoxy-2-furanylmethanol (9). Mild acid hy-
drolysis of 9 in formic acid brings about cleavage of the
acetal bond and subsequent ring expansion to form

4816 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23
Liu et al.
Sch em e 5. Synthesis of 2-(1′-Hydroxyalkyl)-3-hydroxypyridin-4-ones 19
Sch em e 6. Synthesis of the Benzoyl Ester Derivative
21
(III) so that aqueous solutions equilibrate to give a
mixture in which the speciation depends on the metal
ion, ligand, and hydrogen ion concentration. In order
to investigate the influence of the 2-(1’-hydroxyalkyl)
substituent on ligand affinity, both the pKa values of
the ligands and the stability constants of their iron(III)
complexes were evaluated using a combined automated
spectrophotometric and potentiometric titration sys-
tem.12,13 Distribution coefficients of the chelators were
measured using an automated continuous flow tech-
nique.12,13
Biologica l Exp er im en ts
In vivo iron mobilization efficacy of all ligands was
evaluated in a non-iron-overloaded rat model originally
developed by Pippard et al.15 [59Fe]Ferritin is used to
label the liver iron pool, and this is followed by a
challenge with a test chelator at a time when the iron
released by lysosomal degradation of ferritin is maxi-
mally available.11,15 Since the liver is the major iron
storage organ under iron-overloaded conditions, this
method comes close of being an ideal animal model to
assess oral bioavailability and to compare the ability
of chelators to remove iron from liver. In this model,
DFO, Deferiprone, and diethylenetriaminepentaacetic
acid (DTPA) all behave in a manner similar to that
observed clinically.11,15 Furthermore this method pro-
duces highly reproducible data of the type necessary for
dose-response investigations and structure-activity
studies.15
The general methodology adopted for the synthesis
of 2-(1′-hydroxyalkyl)-3-hydroxypyridin-4-ones is sum-
marised in Scheme 5. The protection of the 2-(1′-
hydroxyl) function proved to be essential since without
protection, marked decomposition was observed, which
consequently resulted in low yields (<10%). In the
current work, both the 3-hydroxyl and the 2-(1′-hy-
droxyl) groups were protected in one step by reacting
corresponding pyran-4-ones 16 with benzaldehyde di-
methyl acetal in N,N-dimethylformamide in the pres-
ence of a catalytic amount of toluene-p-sulphonic acid.
The desired pyridinone products 19 were then prepared
by the reaction of the protected pyranone 17 with a
range of primary amines, followed by hydrogenation to
remove the protecting group. Generally, 2-monosubsti-
tuted pyran-4-one derivatives can react with various
primary amines under mild conditions, e.g. stirring at
ambient temperature or refluxing in aqueous alcohol,
resulting with the conversion to the corresponding
pyridin-4-one analogous in good yields.12,13 However, the
analogous reaction between 2,6-disubstituted pyran-4-
ones and amines is limited to primary amines of short
chain length (such as methylamine or ethylamine) since
longer chain lengths or branching results in yields of
less than 5% (unpublished data). This may be attributed
to the flanking 2- and 6-substituents providing a steric
barrier to ring closure resulting in high yields of poly-
meric side products.
Resu lts
Liga n d p Ka Va lu es. All ligands were investigated
by simultaneous spectrophotometric and potentiometric
titration. The optimized pKa values obtained from non-
linear least-square regression analysis are shown in
Table 4. The pKa1 values correspond to the deprotona-
tion of the 4-hydroxy group and the pKa2 values to the
dissociation of the 3-hydroxy group (Scheme 7). The pKa
values obtained from spectrophotometric titration are
similar and comparable to those calculated from the
potentiometric titration (Table 4). This indicates that
it is valid to use the pKa values obtained by simulta-
neous titration to support the experimental work. The
introduction of a 1′-hydroxyalkyl function at the 2-posi-
tion results in clear reduction in the pKa values of both
the 4-hydroxyl and 3-hydroxyl groups, as compared with
simple alkyl-substituted HPOs such as CP20 and CP94
(Table 4). Such an effect was found to be greater with
the hydroxyethyl substituent than with the hydroxy-
methyl substituent, particularly in the cases of ligands
3 (pKa2 ) 8.76 ( 0.02) and 19e (pKa2 ) 8.71 ( 0.01),
where a reduction in pKa values of the 3-hydroxyl group
in the region of one order of magnitude was observed,
as compared with CP20 (pKa2 ) 9.71 ( 0.06).
A benzoyl ester prodrug has been prepared by react-
ing the protected alcohol 18d with benzoic acid in the
presence of triphenylphosphine and diethyl azodicar-
boxylate (Scheme 6).14 After hydrogenation, the final
compound 21 was isolated as the hydrochloride salt.
P h ysicoch em ica l Ch a r a cter iza tion
3-Hydroxypyridin-4-ones possess two pKa values, one
corresponding to the 4-hydroxyl function and the second
corresponding to the 3-hydroxyl function. These biden-
tate ligands also form a number of complexes with iron-

2-(1′-Hydroxyalkyl)-3-hydroxypyridin-3-ones
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23 4817
Ta ble 2. Chemical Structure of Compounds 17, 18, and 20
expected to be almost identical to their partition coef-
ficients. The resulting values are presented in Table 3.
As expected the introduction of a hydroxyl function
results in a decrease in lipophilicity relative to the
simple alkyl derivatives. The distribution coefficient
values of the 2-(1′-hydroxyalkyl) HPOs cover the range
0.048-5.61. The most hydrophobic compound was 21,
the benzoyl ester derivative of 19d , which possesses a
log P value of 0.75.
Ir on Mobiliza tion Effica cy in Ra ts. In vivo iron
scavenging ability of the 2-(1′-hydroxyalkyl) HPOs was
compared with CP20 and CP94 in a [59Fe]ferritin-loaded
rat model.11 All chelators were administered orally, and
the dose of chelator was 450 µmol/kg. Several 2-(1′-
hydroxymethyl) derivatives such as 19a and 19f were
found to be poor scavengers of iron under in vivo
conditions, with efficacies of 4.6% and 0.6%, respec-
tively. However with the substituted 2-(1′-hydroxyalkyl)
derivatives, 3, 19h , and 19i were found to be as equally
effective as CP94 (Table 5). The benzoyl ester prodrug
21 was also found to be equally effective as CP94. This
ester prodrug (D7.4 ) 5.61) provided a significant
improvement in the ability of facilitating iron excretion
over that of the parent hydroxyl compound (19d ) (D7.4
) 0.056), the iron mobilization efficacy increasing from
7.8% to 52.1%.
compd
X
R1
R
R6
17a
17b
17c
17d
17e
18a
18b
18c
18d
18e
18f
18g
18h
18i
20
O
O
O
O
O
N
N
N
N
N
N
N
N
N
N
H
CH3
H
CH3
C2H5
H
H
CH3
H
CH3
H
H
H
H
CH3
CH3
CH3
H
H
H
H
H
CH3
CH3
CH3
CH3
H
CH3
C2H5
C2H5
CH2CH2CH2OH
CH2CH2CH2OH
CH3
C2H5
CH3
CH3
CH3
C2H5
H
CH2CH2CH2OOCPh
Sta bility Con sta n ts of Ir on (III) Com p lexes. In a
multicomponent solution the absorbance at a single
wavelength is given by the sum of the absorbances of
the individual species present. When full spectra are
recorded the observation of a shift in λmax at differing
metal/ligand equilibrium conditions is the most direct
evidence of a multicomponent system. Analysis of the
resultant absorbance spectra obtained from a solution
of metal/ligand at various pH values permits the
calculation of the concentration of each species as a
function of pH and hence the stability constant. In this
current work, the stability constants of all bidentate
HPO ligands were evaluated using an automated spec-
trophotometric titration system. The pH titration curve
for ligand 3 in the presence of iron(III) is presented as
an example in Figure 1. A clear shift in λmax is observed
in the full speciation spectra of ligand 3:iron(III) com-
plexes over pH range 1.59-10.05 (Figure 1) which
displays the pH dependence of the different metal/ligand
equilibrium. At low pH values the λmax of the iron(III)
complex (545 nm) corresponds to the [FeIIIL]2+ complex,
and the [FeIIIL2]+ complex (λmax, 510 nm) dominates over
the pH range 3.0-5.0. Above pH 6.5 the neutral
[FeIIIL3]0 complex (λmax, 460 nm) is the major species.
For bidentate ligands the logarithm of the cumulative
stability constants was determined. This value, log â3,
is obtained by summation of the logarithms of three
stepwise equilibrium constants (K1, K2, and K3) corre-
sponding to Scheme 8. The optimized values are pre-
sented in Table 4. The introduction of a 2-(1′-hydroxy-
alkyl) substitutent leads to a reduction in both log K
values and log â3 values, as compared with simple alkyl-
substituted HPOs (Table 4). The log â3 values for 2-(1′-
hydroxyalkyl) HPOs fall within the range of 34.7-35.6,
as compared to 36.4 for CP20 and 36.7 for CP94.
Discu ssion
In order to understand the greater efficacy of CP94
in rats, the 2-(1′-hydroxyethyl) metabolite of CP94 (3)
and seven analogues have been synthesized. Both the
pKa values of the ligands and the stability constants of
their iron(III) complexes have been evaluated. Gener-
ally, simple ring N-alkylation of 3-hydroxypyridin-4-
ones results in little change in the affinity for protons
as well as for iron(III).12,13 However 2-alkyl substitu-
tion tends to elevate the pKa values in the order ethyl
> methyl > hydrogen12 by virtue of the positive induc-
tive effect. An additional increment has also been
observed with 6-alkyl substitution (unpublished data).
However the introduction of a 1′-hydroxyalkyl group at
the 2-position significantly reduces the pKa values
(Table 4). This effect results from stabilising the ionised
species due to the combined effect of intramolecular
hydrogen bonding between the 2-(1′-hydroxyl) group and
the adjacent 3-hydroxyl function and the negative
inductive effect of the 2-(1′-hydroxyl) group from the
pyridinone ring (Figure 2). Although such an effect
reduces the overall stability constants for iron(III)
(Table 4), it also reduces the affinity of the chelating
function for protons. These changes result in an in-
crease in the corresponding pFe3+ values at pH 7.4
(Table 4).
In designing iron chelators for clinical application,
metal selectivity and ligand-metal complex stability are
paramount.16 The most suitable comparison standard
for ligands is the pFe3+ value. pFe3+ is defined as the
negative logarithm of the concentration of the free iron-
(III) in solution. Typically pFe3+ values are calculated
for total [ligand] ) 10-5 M, total [iron] ) 10-6 M at pH
7.4. The comparison of ligands under these conditions
is more meaningful, since pFe3+, unlike the stability
constants, takes into account the effects of ligand
basicity, ligand protonation, and metal hydrolysis as
Distr ibu tion Coefficien ts. The distribution coef-
ficients (D) of the ligands between 1-octanol and MOPS
buffer (pH 7.4) were determined using an automated
filter-probe system.12,13 Since the degree of ionization
of the two functional groups on the pyridinone ring is
relatively small at pH 7.4, the neutral species of the
compounds is the predominant form (>98%). Hence the
distribution coefficients of the compounds at pH 7.4 are

4818 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23
Liu et al.
Ta ble 3. Synthesis of 2-(1′-Hydroxyalkyl)-3-hydroxypyridin-4-one Hydrochlorides and Their Distribution Coefficients between an
Aqueous Phase Buffered at pH 7.4 and Octanol (n ) 5)
% yield
compd
R1
R
R6
mp, ˚C
(from 15)
formula
anal.
D7.4
log P
19a
19b
3 (19c)
19d
19e
19f
19g
19h
19i
CH3
C2H5
C2H5
CH2CH2CH2OH
CH2CH2CH2OH
CH3
C2H5
CH3
CH3
H
H
CH3
H
CH3
H
H
CH3
C2H5
H
H
H
H
H
157-159
168-169
139-140
138-139
117-120
140-143
160-162
208-212
221-223
142-143
56.6
61.4
65.5
64.7
47.2
57.4
57.8
46.6
31.7
76.2a
C7H10NO3Cl
C8H12NO3Cl
C9H14NO3Cl
C9H14NO4Cl
C10H16NO4Cl
C8H12NO3Cl‚1/2H2O
C9H14NO3Cl
C9H14NO3Cl
C10H16NO3Cl
C16H18NO5Cl
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
0.048 ( 0.003
0.140 ( 0.007
0.266 ( 0.003
0.056 ( 0.005
0.075 ( 0.007
0.098 ( 0.006
0.223 ( 0.018
0.248 ( 0.012
0.734 ( 0.013
5.610 ( 0.264
-1.319
-0.854
-0.575
-1.252
-1.125
-1.009
-0.652
-0.606
-0.134
0.749
H
CH3
CH3
CH3
CH3
H
21
CH2CH2CH2OOCPh
a
Percent yield from 18d .
Ta ble 4. Spectrophotometric and Potentiometric Determined pKa Values for Ligands and Affinity Constants for Fe(III) Complexes
pKa1 pKa2 affinity constants for Fe(III)
compd potentiometric spectrophotometric potentiometric spectrophotometric log K1 log K2 log K3 log â3 (at pH 7.45)
pFe3+
CP20
CP94
19a
19b
3
19d
19e
19f
3.68
3.81
2.92
2.80
3.03
3.02
2.98
3.32
3.29
3.54
3.76
3.56
9.77
9.93
9.11
9.27
8.77
9.29
8.72
9.44
9.45
8.99
9.00
9.64
14.56
15.20
14.47
14.66
15.21
15.67
14.93
14.39
14.37
15.38
14.50
12.19
11.76
11.27
11.16
10.97
10.01
10.87
11.57
11.60
11.09
11.48
9.69
9.84
9.58
9.43
8.90
9.57
8.87
9.55
9.64
9.05
9.05
36.4
36.8
35.3
35.3
35.1
35.3
34.7
35.5
35.6
35.5
35.0
19.4
19.7
20.9
21.0
21.4
20.7
21.5
20.4
20.4
21.5
21.0
2.93
2.88
3.11
2.87
2.96
3.37
3.28
3.55
3.78
9.11
9.05
8.74
9.14
8.69
9.42
9.38
8.97
8.98
19g
19h
19i
Sch em e 7. Protonation Equilibra of
3-Hydroxypyridin-4-ones
well as differences in metal-ligand stoichiometries.
Generally, ligands containing an identical chelating
group possess similar pFe3+ values: for instance, 3-hy-
droxypyridin-2-one ligands possess pFe3+ values of 16,
whereas the pFe3+ values for 3-hydroxypyridin-4-ones
are around 20.17 However the introduction of adjacent
functions can have a dramatic effect on pFe3+ values.
It has been previously established that the introduction
of an amide function at the 4-position of 3-hydroxypy-
ridin-2-ones reduces the pKa values and hence increases
the pFe3+ values, due to the formation of a stable
intramolecular hydrogen bond between the amide pro-
ton and the adjacent oxygen donor.18 In this current
study, a similar reduction on the pKa values and hence
enhancement on the pFe3+ values were also observed
by the introduction of a 1′-hydroxyalkyl group at the
2-position of 3-hydroxypyridin-4-ones. Such enhance-
ment on the pFe3+ values is typified by CP20 and 19a ,
where a differential of almost 1.5 log units in pFe3+
values is observed (Table 4). Similarly comparison of
CP94 with 3, which actually is the major metabolite of
CP94 in rats, demonstrates a positive increment of 1.7
log units. This difference could account for the extremely
F igu r e 1. pH dependence of the spectrum of ligand 3 in the
presence of iron(III) over the pH range 1.59-10.05: [Fe(III)]
) 1.8 × 10-4 M; [3] ) 1.8 × 10-3 M.
high efficacy of CP94 in the rat. In rats CP94 acts as a
“prodrug” of 3, a chelator with a higher pFe3+ value.
The introduction of a 1′-hydroxyalkyl group at the
2-position also has a dramatic effect on the speciation
plot of the iron(III) complexes as presented in Figure 3
which indicates that 3 will dissociate less readily leading

2-(1′-Hydroxyalkyl)-3-hydroxypyridin-3-ones
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23 4819
Sch em e 8. Complex Formation of Iron(III) with
Bidentate 3-Hydroxypyridin-4-ones
Ta ble 5. Iron Mobilization Efficacy Studies of
2-(1′-Hydroxyalkyl)-3-Hydroxypyridin-4-ones in the
[
59Fe]Ferritin-Loaded Rat Modela
chelator
iron mobilization (%)
efficacy (%)
F igu r e 3. Comparison of the speciation plots between CP20
and 3 in the presence of iron(III): L ) CP20; L′ ) 3; [Fe(III)]
) 1 × 10-6 M; [ligand] ) 1 × 10-5 M.
control
19a
19b
3
19d
19e
19f
19g
19h
19i
3.87 ( 1.0
8.44 ( 3.6
33.3 ( 6.7
54.5 ( 9.9
11.7 ( 4.1
14.5 ( 4.5
4.5 ( 1.1
12.9 ( 2.3
48.4 ( 7.2
60.4 ( 15.6
56.0 ( 6.0
4.6
29.4
50.6
7.8
10.6
0.6
values associated with the substituted 2-(1′-hydroxy-
alkyl) derivatives, 3, 19h , and 19i were found to be as
equally effective as CP94. They have a clear advantage
over CP94 as their respective log P values, -0.6, -0.59,
and -0.14, are significantly lower than that of CP94,
namely 0.26. They are therefore predicted to penetrate
the blood-brain barrier and the placental barrier and
enter peripheral cells (such as the heart and bone
marrow) much less readily than CP94. Furthermore
these compounds (3, 19h , and 19i) may not undergo
glucuronidation in vivo by virtue of the presence of the
1′-hydroxyalkyl function at the 2-position.5 Such mol-
ecules have immense clinical potential for the treatment
of iron overload.
9.0
44.5
56.5
52.1
21
a
All chelators (450 µmol/kg) were given orally, and control rats
were administered with an equivalent volume of water. Values
are expressed as means ( SD (n ) 5).
In order to improve both drug absorption and hepatic
extraction, a benzoyl ester derivative 21 has also been
synthesized. This ester prodrug provided a significant
improvement in the ability of facilitating iron excretion
over that of the parent hydroxyl compound 19d (Table
5). Such enhancement is almost certainly due to the
increased lipophilicity of this ester prodrug as compared
to the parent compound, which indicates that selective
delivery of the drug to the liver has been achieved (Fig-
ure 4). This molecule, due to its relatively high lipophi-
licity (log P ) 0.75), will be expected to be extracted by
the liver more efficiently. On entry to the liver, it may
undergo rapid hydrolysis to generate the hydrophilic
compound 19d (log P ) -1.25) which can be further
metabolised to the even more hydrophilic carboxylate
metabolite.10 Such molecules will severely limit the
distribution of the chelator, which in turn could possibly
minimise the potential toxicity. This ester is the most
effective hydroxypyridinone prodrug synthesized to date
(Tables 1 and 5).
F igu r e 2. Proposed intramolecular hydrogen bonding be-
tween the 2-(1′-hydroxyl) group and the adjacent 3-hydroxyl
function and the negative inductive effect of the 2-(1′-hydroxyl)
group from the pyridinone ring.
to lower concentrations of the L2Fe+ complex compared
with CP20. The partial dissociation of iron complexes
formed from either bi- or tridentate ligands generates
a water-coordinated iron surface which may interact
with oxygen and hydrogen peroxide thereby generating
hydroxyl radicals.3 Thus chelators with high pFe3+
values are predicted to scavenge iron more effectively
at low ligand concentrations and to dissociate less
readily, leading to lower concentrations of the partially
co-ordinated iron complexes.
In vivo iron mobilization efficacy of this series of
compounds has been investigated in the [59Fe]ferritin-
labelled rat model. The introduction of a 1’-hydroxy-
methyl function alone does not improve efficacy; thus
19a and 19f are poor scavengers of iron under in vivo
conditions (Table 5). This may be attributed to the low
log P values associated with these molecules (Table
3). Such low lipophilicities may lead to poor oral
absorption and liver extraction. Consequently, inef-
ficient iron mobilization would be expected. However
with the higher lipophilicities and improved pFe3+
In summary, the introduction of a 1′-hydroxyalkyl
group at the 2-position of 3-hydroxypyridin-4-ones will
lead a dramatic enhancement on the pFe3+ values. Such
novel HPOs show great promise in their ability to
remove iron under in vivo conditions and thereby may
offer a realistic approach to the design of orally active
iron chelators for clinical use.

4820 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23
Liu et al.
F igu r e 4. Schematic representation of the use of hydrophobic esters to enhance both the absorption from the gastrointestinal
tract (GIT) and the hepatic extraction of the chelator. Subsequent intracellular hydrolysis occurs yielding a hydrophilic chelator
which can undergo further metabolism to the extremely hydrophilic negatively charged 1-carboxyalkyl metabolite in the liver.
Recrystallization from ethyl acetate afforded a white crystal-
line solid (17g, 82%): mp 74-75 °C. H NMR (CDCl3) δ: 3.5
(s, 3H, OCH3), 4.4 (q, 2H, 2,2′-H, AB center, J 22′ ) 14.5 Hz,
∆δ22′ ) 18.5 Hz), 5.05 (d, 1H, 6-H), 7.25 (d, 1H, 5-H). MS (EI):
m/z 207 [M+•]. Anal. (C6H7O3Br) C, H.
Exp er im en ta l Section
1
Gen er a l P r oced u r e. Melting points were determined using
an Electrothermal IA 9100 digital melting point apparatus and
are uncorrected. 1H NMR spectra were recorded using a
Perkin-Elmer (60 MHz) NMR spectrometer. Chemical shifts
(δ) are reported in ppm downfield from the internal standard
tetramethylsilane (TMS). Mass spectra (EI) were recorded on
a J oel AX505W. Elemental analyses were performed by Micro
Analytical Laboratories, Department of Chemistry, The Uni-
versity of Manchester, Manchester M13 9PL, U.K.
2,5-Dih yd r o-2,5-d im eth oxy-2-fu r a n ylm eth a n ol (9). To
a solution of furfuryl alcohol (8) (98 g, 1.0 mol) in methanol
(800 ml) was added sodium hydrogen carbonate (185 g, 2.2
mol), and the mixture was cooled to -30 °C. To this well-
stirred solution was added a solution of bromine (176 g, 1.1
mol) in methanol (300 mL). The temperature of the reaction
was maintained between -30 and -10 °C. After the addition
period the reaction mixture was allowed to warm to room
temperature and stirred for 2 h. The reaction was then filtered
and concentrated. The residue was taken up into toluene (300
mL) and filtered through aluminium oxide. Evaporation of the
solvent and distillation at 74 °C/0.4 mmHg (lit. value19 71 °C/
1.0 mmHg) gave a colorless liquid (115.5g, 72.2%). 1H NMR
(CDCl3) δ: 2.33 (br, s, 1H, CH2OH), [3.25 (s, isomer A) & 3.3
(s, isomer B), 3H, 5-OCH3], 3.6 (s, 3H, 2-OCH3), 3.7 (br, s,
2H, CH2OH), [5.55 (s, isomer B) & 5.8 (s, isomer A), 1H, 5-H],
5.9-6.4 (m, 2H, 3,4-H).
3-Hyd r oxyp yr a n -4(1H)-on e (P yr om econ ic a cid ) (12).
5.8 g (28 mmol) of 11 was added to trifluoroacetic acid (30 mL)
and refluxed for 2 h. Concentration to dryness in vacuo
followed by repeated re-evaporations from methanol afforded
a brown crystalline residue. The solid was treated with
activated carbon and recrystallized from toluene to yield light
yellow plates (2.5 g, 80%): mp 114-115 °C (lit. value20 113-
1
115.5 °C). H NMR (CDCl3) δ: 6.48 (d, 1H, 5-H), 6.3-7.4 (br,
1H, OH), 7.76 (d, 1H, 6-H), 7.88 (s, 1H, 2-H).
2-Meth yl-5-h yd r oxyp yr a n -4(1H)-on e (a llom a ltol) (15)
was synthesized from kojic acid (13) in two steps using the
established procedure.21
2-Hyd r oxym eth yl-3-h yd r oxyp yr a n -4(1H)-on e (16a ). So-
dium hydroxide (4 g, 100 mmol) dissolved in distilled water
(10 mL) was added to a solution of 12 (8.96 g, 80 mmol, 1 equiv)
in methanol (50 mL) and allowed to stir at room temperature
for 5 min; 35% formaldehyde solution (16 mL) was added
dropwise over 15 min and the solution was stirred overnight.
After adjustment to pH 1 with 37% hydrochloric acid, the
reaction mixture was concentrated in vacuo to dryness and
the resulting solid was extracted with isopropanol (2 × 100
mL) at 90 °C. The isopropanol extracts were concentrated to
yield the crude products. Recrystallization from isopropanol
afforded a white crystalline solid (9.7 g, 85.4%): mp 154-156
°C (lit. value20 154-155 °C). 1H NMR (DMSO-d6) δ: 4.4 (s,
2H, 2-CH2OH), 4.6-5.7 (br, 1H, 2-CH2OH), 6.34 (d, 1H, 5-H),
8.1 (d, 1H, 6-H), 9.0 (br, s, 1H, 3-OH).
6-Meth oxy-2H-p yr a n -3(6H)-on e (10). A solution of 9 (80
g, 0.5 mol) in methanol (30 mL) was added during 15 min at
25 °C to a well-stirred solution of methanol (20 mL) in formic
acid (400 mL). The mixture was then poured into water (1000
mL) and extracted with chloroform (3 x 500 mL). The extracts
were combined, washed successively with saturated sodium
hydrogen carbonate solution and brine, dried with anhydrous
sodium sulfate, and concentrated in vacuo to yield a light
brown oil (54g, 84.4%). Distillation at 47-48 °C/0.5 mmHg (lit.
value19 76-81 °C/13 mmHg) afforded a clear, sharp-smelling
2-(1′-Hyd r oxyeth yl)-3-h yd r oxyp yr a n -4(1H)-on e (16b).
Pyromeconic acid (12) (5.6 g, 50 mmol) was added to water
(50 mL) and the pH of the solution was adjusted to 10.5 using
50% aqueous sodium hydroxides. Acetaldehyde (2.64 g, 60
mmol) dissolved in water (20 mL) was slowly added dropwise
over 1 h and the solution allowed to stir overnight. The
reaction mixture was acidified to pH 1 with 37% hydrochloric
acid and concentrated in vacuo to dryness. The residue was
extracted with isopropanol (2 × 70 mL) at 90 °C. The
isopropanol extracts were combined and concentrated to yield,
after recrystallization from toluene, a pale yellow crystalline
solid (3.7 g, 47.4%): mp 131-132 °C (lit. value20 130-131 °C).
1H NMR (DMSO-d6) δ: 1.3 (d, 3H, 2-CHCH3), 5.03 (q, 1H,
2-CHCH3), 6.38 (d, 1H, 5-H), 8.2 (d, 1H, 6-H).
1
oil. H NMR (CDCl3) δ: 3.5 (s, 3H, OCH3), 4.18 (q, 2H, 2,2′-
H, AB center, J 22′ ) 16.5 Hz, ∆δ22′ ) 20.5 Hz), 5.02 (d, 1H,
6-H), 6.03 (d, 1H, 4-H), 6.81 (q, 1H, 5-H, J 56 ) 3.3 Hz, J 45
)
10.2 Hz).
4-Br om o-6-m eth oxy-2H-p yr a n -3(6H)-on e (11). To a so-
lution of 10 (12.8 g, 0.1 mol) in dichloromethane (40 mL) at 0
°C was added 16.0 g (0.1 mol) of bromine in dichloromethane
(10 mL). Then triethylamine (14 mL) was added dropwise at
0 °C and the reaction was allowed to warm to room temper-
ature and stirred for 2 h. The reaction was then diluted with
toluene (200 mL). After filtration, the organic solution was
then washed with 5% sodium hydrogen carbonate solution and
brine, dried with anhydrous sodium sulfate, filtered, and
concentrated to yield the crude product as a light brown solid.
2-H yd r oxym e t h yl-3-h yd r oxy-6-m e t h ylp yr a n -4(1H )-
on e (16c), 2-(1′-h yd r oxyeth yl)-3-h yd r oxy-6-m eth ylp yr a n -
4(1H)-on e (16d ), a n d 2-(1′-h yd r oxyp r op yl)-3-h yd r oxy-6-

2-(1′-Hydroxyalkyl)-3-hydroxypyridin-3-ones
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23 4821
m eth ylp yr a n -4(1H)-on e (16e) were synthesized from allo-
maltol (15) by following the methodology as described by Ellis
et al.21
to afford a yellow oil (79.1%). 1H NMR (CDCl3) δ: 1.2-2.2 (m,
6H, CHCH3 & N-CH2CH3), 3.4-4.0 (m, 2H, N-CH2CH3), 4.8-
5.4 (m, 1H, CHCH3), [5.6 (s, isomer A) & 6.0 (s, isomer B);
1H, CHPh], 6.3 (d, 1H, 7-H(pyridinone)), 7.0-7.7 (m, 6H, Ar
& 6-H(pyridinone)).
8-Oxo-4,8-d ih yd r o-2-p h en yl-5,6-d im eth yl-4H-p yr id in o-
[3,2-d ]-m -d ioxin (18f): 66%; mp 256-258 °C. 1H NMR
(methanol-d4) δ: 2.2 (s, 3H, 6-CH3), 3.35 (s, 3H, N-CH3), 4.95
(s, 2H, CH2O), 5.8 (s, 1H, CHPh), 6.5 (s, 1H, 7-H(pyridinone)),
7.0-7.5 (m, 5H, Ar ).
8-Oxo-4,8-d ih yd r o-2-p h e n yl-4H -p yr a n o[3,2-d ]-m -d i-
oxin (17a ). A solution of 16a (2.84 g, 20 mmol), benzaldehyde
dimethyl acetal (6.08 g, 40 mmol), and toluene-p-sulfonic acid
monohydrate (0.04g, cat.) in N,N-dimethylformamide (50 mL)
was rotated under aspirator pressure at 80 °C for 3 h. The
solvent was removed under high vacuum and the residue
taken up into dichloromethane (100 mL). The organic solution
was washed successively with 5% sodium hydrogen carbonate
solution and brine. After drying over magnesium sulfate, the
solvent was removed to give the crude product. Recrystalliza-
tion form dichloromethane/petroleum ether (40-60 °C) af-
forded a white crystalline solid (3.77 g, 82%): mp 141-143
°C. 1H NMR (CDCl3) δ: 4.7 (d, 2H, CH2O), 5.88 (s, 1H, CHPh),
6.35 (d, 1H, 7-H(pyranone)), 7.2-7.9 (m, 6H, Ar & 6-H(pyra-
none)); MS (EI): m/z, 230 [M+•]. Anal. (C13H10O4) C, H.
Analogous syntheses starting with 16b-16e gave com-
pounds 17b-17e as shown in Table 2.
8-Oxo-4,8-d ih yd r o-2-p h en yl-5-eth yl-6-m eth yl-4H-p yr i-
d in o[3,2-d ]-m -d ioxin (18g): purification by column chroma-
tography on silica gel (eluant:methanol:chloroform, 20:80 v/v)
1
(62.3%); mp 201-203 °C. H NMR (DMSO-d6) δ: 1.25 (t, 3H,
N-CH2CH3), 2.3 (s, 3H, 6-CH3), 3.85 (q, 2H, N-CH2CH3), 5.02
(s, 2H, CH2O), 5.9 (s, 1H, CHPh), 6.45 (d, 1H, 7-H (pyridi-
none)), 7.1-7.6 (m, 5H, Ar ).
8-Oxo-4,8-dih ydr o-2-ph en yl-4,5,6-tr im eth yl-4H-pyr idin o-
[3,2-d ]-m -d ioxin (18h ): purification by column chromatog-
raphy on silica gel (eluant:methanol:chloroform, 20:80 v/v)
1
8-Oxo-4,8-d ih yd r o-4-m et h yl-2-p h en yl-4H -p yr a n o[3,2-
d ]-m -d ioxin (17b): mp 112-113 °C. 1H NMR (CDCl3) δ: 1.55
(d, 3H, CHCH3), 5.0 (q, 1H, CHCH3), 5.8 (s, 1H, CHPh), 6.25
(d, 1H, 7-H(pyranone)), 7.1-7.75 (m, 6H, Ar & 6-H(pyranone)).
MS (EI): m/z 244 [M+•]. Anal. (C14H12O4) C, H.
(56.8%); mp 199-201 °C. H NMR (DMSO-d6) δ: 1.7 (dd, 3H,
CHCH3), 2.35 (s, 3H, 6-CH3), [3.44 (s, isomer B) & 3.5 (s,
isomer A), 3H, N-CH3], 4.9-5.4 (m, 1H, CHCH3), [5.75 (s,
isomer A) & 6.05 (s, isomer B), 1H, CHPh], 6.35 (s, 1H,
7-H(pyridinone)), 7.2-7.9 (m, 5H, Ar ).
8-Oxo-4,8-d ih yd r o-6-m et h yl-2-p h en yl-4H -p yr a n o[3,2-
d ]-m -d ioxin (17c): mp 91-94 °C. 1H NMR (CDCl3) δ: 2.25
(s, 3H, 6-CH3), 4.75 (d, 2H, CH2O), 5.9 (s, 1H, CHPh), 6.18 (s,
1H, 7-H(pyranone)), 7.2-7.8 (m, 5H, Ar ). MS (EI): m/z 244
[M+•]. Anal. (C14H12O4) C, H.
8-Oxo-4,8-d ih yd r o-4,6-d im et h yl-2-p h en yl-4H -p yr a n o-
[3,2-d ]-m -d ioxin (17d ): mp 120-122 °C. 1H NMR (CDCl3)
δ: 1.6 (d, 3H, CHCH3), 2.25 (s, 3H, 6-CH3), 5.08 (q, 1H,
CHCH3), 5.9 (s, 1H, CHPh), 6.18 (s, 1H, 7-H(pyranone)), 7.2-
7.8 (m, 5H, Ar ). MS (EI): m/z 258 [M+•]. Anal. (C15H14O4) C,
H.
8-Oxo-4,8-d ih yd r o-4-et h yl-6-m et h yl-2-p h en yl-4H -p yr -
a n o[3,2-d ]-m -d ioxin (17e): mp 111-114 °C. 1H NMR (CDCl3)
δ: 1.0 (t, 3H, CHCH2CH3), 1.6-2.1 (m, 2H, CHCH2CH3), 2.2
(s, 3H, 6-CH3), 4.7-5.0 (m, 1H, CHCH2CH3), 5.8 (s, 1H,
CHPh), 6.1 (s, 1H, 7-H (pyranone)), 7.15-7.7 (m, 5H, Ar ). MS
(EI): m/z 272 [M+•]. Anal. (C16H16O4) C, H.
8-Oxo-4,8-d ih yd r o-2-p h en yl-5-m eth yl-4H-p yr id in o[3,2-
d ]-m -d ioxin (18a ). To a solution of 17a (2.3 g, 10 mmol) in
ethanol (10 mL)/water (10 mL) was added 40% aqueous
methylamine (2.5 mL) followed by 2 N sodium hydroxide
solution until pH 12.5 was obtained. The reaction mixture was
sealed in a thick-walled glass tube and stirred at 70 °C for 3
h. After adjustment to pH 1 with concentrated hydrochloric
acid, the solvent was removed by rotary evaporation prior to
addition of water (50 mL) and washing with diethyl ether (3
× 50 mL). Subsequent adjustment of the aqueous fraction to
pH 7 with 10 N sodium hydroxide solution was followed by
extraction into dichloromethane (4 × 50 mL), the combined
organic layers then being dried over anhydrous sodium sulfate,
filtered, and rotary-evaporated to give a yellow solid. Recrys-
tallization from methanol/diethyl ether afforded a light yellow
crystalline solid (1.6 g, 65.8%): mp 210-211 °C. 1H NMR
(DMSO-d6) δ: 3.55 (s, 3H, N-CH3), 5.08 (s, 2H, CH2O), 5.92
(s, 1H, CHPh), 6.12 (d, 1H, 7-H(pyridinone)), 7.25-7.85 (m, 6H,
Ar & 6-H(pyridinone)).
Analogous syntheses starting with reaction of 17a -17e with
methylamine or ethylamine gave compounds 18b, 18c, and
18f-18i as shown in Table 2.
8-Oxo-4,8-d ih yd r o-2-p h en yl-5-et h yl-4H -p yr id in o[3,2-
d ]-m -d ioxin (18b): purification by column chromatography
on silica gel (eluant: methanol:chloroform, 15:85 v/v) (69.7%);
mp 177-179 °C. 1H NMR (DMSO-d6) δ: 1.3 (t, 3H, N-CH2CH3),
3.8 (q, 2H, N-CH2CH3), 5.12 (s, 2H, CH2O), 5.9 (s, 1H, CHPh),
6.25 (d, 1H, 7-H (pyridinone)), 7.1-7.8 (m, 6H, Ar & 6-H(py-
ridinone)).
8-Oxo-4,8-d ih yd r o-2-p h en yl-4-et h yl-5,6-d im et h yl-4H -
p yr id in o[3,2-d ]-m -d ioxin (18i): purification by column chro-
matography on silica gel (eluant:methanol:chloroform, 20:80
v/v) (39.8%); mp 185-187 °C. 1H NMR (DMSO-d6) δ: 0.8-1.4
(m, 3H, CHCH2CH3), 1.5-2.2 (m, 2H, CHCH2CH3), 2.3 (s, 3H,
6-CH3), [3.38 (s, isomer B) & 3.45 (s, isomer A), 3H, N-CH3],
[4.5-4.8 (m, isomer B) & 4.9-5.4 (m, isomer A), 1H, CHCH2-
CH3], [5.68 (s, isomer A) & 5.95 (s, isomer B), 1H, CHPh], 6.25
(s, 1H, 7-H(pyridinone)), 7.2-7.8 (m, 5H, Ar ).
8-Oxo-4,8-d ih yd r o-2-p h en yl-5-(3′-h yd r oxyp r op yl)-4H-
p yr id in o[3,2-d ]-m -d ioxin (18d ). To a solution of 17a (3.45
g, 15 mmol) in ethanol (50 mL)/water (50 mL) was added
3-amino-1-propanol (2.25 g, 30 mmol) followed by 2 N sodium
hydroxide solution until pH 12.5 was obtained. The reaction
mixture was refluxed for 3 h. TLC analysis (eluant:methanol:
chloroform, 10:90 v/v) showed that no starting material was
present. After removal of solvent by rotary evaporation, the
residue was purified by column chromatography on silica gel
(eluant:methanol:chloroform, 20:80 v/v) to afford a yellow
crystalline solid (3.35 g, 77.8%): mp 73-76 °C, 1H NMR
(CDCl3) δ: 1.5-2.1 (m, 2H, N-CH2CH2CH2O), 3.2-4.0 (m, 4H,
N-CH2CH2CH2O), 4.0-5.2 (br, 1H, OH), 4.8 (s, 2H, CH2O),
5.7 (s, 1H, CHPh), 6.2 (d, 1H, 7-H(pyridinone)), 7.0-7.8 (m,
6H, Ar & 6-H(pyridinone)).
Analogous reaction of 17b gave compound 18e (Table 2).
8-Oxo-4,8-dih ydr o-2-ph en yl-4-m eth yl-5-(3′-h ydr oxypr o-
p yl)-4H-p yr id in o [3,2-d ]-m -d ioxin (18e): 57.6%; oil. 1H
NMR (CDCl3) δ: 1.5 (d, 3H, CHCH3), 1.5-2.1 (m, 2H,
NCH2CH2CH2O), 3.2-4.0 (m, 4H, NCH2CH2CH2O), 4.0-5.2
(br, 1H, OH), 5.28 (q, 1H, CHCH3), 5.58 (s, 1H, CHPh), 6.2
(d, 1H, 7-H(pyridinone)), 7.0-7.8 (m, 6H, Ar & 6-H(pyridi-
none)).
1-Met h yl-2-h yd r oxym et h yl-3-h yd r oxyp yr id in -4(1H )-
on e Hyd r och lor id e (19a ). A solution of 18a (1.22 g, 5mmol)
in ethanol (30 mL) was adjusted to pH 1 with concentrated
hydrochloric acid prior to hydrogenolysis for 12 hours in the
presence of 5% Pd/C catalyst (0.2g). Filtration followed by
rotary evaporation gave the crude product as a white solid.
Recrystallization from methanol/diethyl ether gave a white
crystalline solid (0.82 g, 86%): mp 157-159 °C. 1H NMR
(DMSO-d6) δ: 4.18 (s, 3H, N-CH3), 4.8 (s, 2H, 2-CH2OH), 7.4
(d, 1H, 5-H(pyridinone)), 8.3 (d, 1H, 6-H (pyridinone)), 7.6-
9.3 (br, 3H, OH). MS (EI): m/z 156 [(M - Cl)+•]. Anal. (C7H10
NO3Cl) C, H, N.
-
Analogous reaction of 18b-18i gave compounds 19b-19i,
respectively, as shown in Table 3.
8-Oxo-4,8-d ih yd r o-2-p h en yl-4-m eth yl-5-eth yl-4H-p yr i-
d in o[3,2-d ]-m -d ioxin (18c): purification by column chroma-
tography on silica gel (eluant:methanol:chloroform, 15:85 v/v)
1-E t h yl-2-h yd r oxym e t h yl-3-h yd r oxyp yr id in -4(1H )-
on e h yd r och lor id e (19b): 73%; mp 168-169 °C. 1H NMR
(D2O) δ: 1.45 (t, 3H, N-CH2CH3), 4.4 (q, 2H, N-CH2CH3), 4.88

4822 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23
(s, 2H, 2-CH2O), 7.1 (d, 1H, 5-H(pyridinone)), 8.1 (d, 1H, 6-H
Liu et al.
8.1 (m, 6H, Ar & 5-H(pyridinone)), 8.3 (d, 1H, 6-H(pyridi-
none)), 8.5-10.2 (br, 3H, OH). MS (EI): m/z 304 [(M - Cl)+•].
Anal. (C16H18NO5Cl) C, H, N.
(pyridinone)). MS (EI): m/z, 170 [(M - Cl)+•]. Anal. (C8H12
-
NO3Cl) C, H, N.
1-E t h yl-2-(1′-h yd r oxyet h yl)-3-h yd r oxyp yr id in -4(1H )-
on e h yd r och lor id e (3) (19c): 82.8%; mp 139-140 °C. 1H
NMR (DMSO-d6) δ: 1.3-1.9 (m, 6H, CHCH3 & N-CH2CH3,
4.6 (q, 2H, N-CH2CH3), 5.55 (q, 1H, CHCH3), 7.4 (d, 1H,
5-H(pyridinone)), 8.25 (d, 1H, 5-H(pyridinone)), 8.5-10.5 (br,
3H, OH). MS (EI): m/z 184 [(M - Cl)+•]. Anal. (C9H14NO3Cl)
C, H, N.
Det er m in a t ion of P h ysicoch em ica l P r op er t ies of
Liga n d s. 1. p Ka Deter m in a tion . Equilibrium constants of
protonated ligands were determined using an automated
computerized system, consisting of a Metrohm 665 dosimat, a
Perkin-Elmer Lambda 5 UV/vis spectrophotometer, a Corning
Delta 225 pH meter, and a 286 Opus PC which controlled the
integrated system.22 A combined Sirius electrode was used to
calibrate the electrode zero and to measure pH values. This
system is capable of undertaking simultaneous potentiometric
and spectrophotometric measurements. A blank titration of
0.1 M KCl (25 mL) was carried out to determine the electrode
zero using Gran’s plot method.23 The solution (0.1 M KCl, 25
mL), contained in a jacketed titration cell, was acidified by
0.15 mL of 0.2 M HCl. Titrations were carried out against 0.3
mL of 0.2 M KOH using 0.01-mL increments dispensed from
the dosimat. Solutions were maintained at 25 ( 0.5 °C under
an argon atmosphere. The above titration was repeated in the
presence of ligand. The data obtained from titrations were
subjected to non-linear least-square regression analysis using
the NONLINM1 program.22 The pKa values were obtained to
an accuracy of (0.02 pH unit.
2. Ir on (III) Affin ity Con sta n t Deter m in a tion . The sta-
bility constants were optimized from the spectrophotometric
titration of the metal-ligand system using the pKa values,
electrode slope, and electrode zero determined above. The
analytical equipment used was similar to that used for pKa
determination. The electrodes were calibrated by titrating a
volumetric standard strong acid with a volumetric standard
strong base under argon gas. In order to maintain sufficient
sensitivity, a 10-mm UV flow-cell was utilised. After electrode
calibration, the solution was reacidified by adding concentrated
hydrochloric acid and the pH of the solution was adjusted to
1.5-2.0. Iron(III) stock solution (atomic absorption standard;
Aldrich) and the test ligand were then added to give a final
ligand:iron(III) ratio of about 10:1. A preadjusted programmed
autoburette was used for the addition of a 0.2 M KOH solution.
The resulting spectrophotometric titration curve was then
subjected to non-linear least-square regression analysis.22
1-(3′-Hyd r oxyp r op yl)-2-h yd r oxym eth yl-3-h yd r oxyp y-
r id in -4(1H)-on e h yd r och lor id e (19d ): 83.2%; mp 138-139
°C. 1H NMR (D2O) δ: 1.9-2.6 (m, 2H, N-CH2CH2CH2O), 3.75
(t, 2H, N-CH2CH2CH2O), 4.6 (m, 4H, N-CH2CH2CH2O), 5.08
(s, 2H, CH2O), 7.25 (d, 1H, 5-H(pyridinone)), 8.2 (d, 1H,
6-H(pyridinone)). MS (EI): m/z 200 [(M - Cl)+•]. Anal. (C9H14
NO4Cl) C, H, N.
-
1-(3′-H yd r oxyp r op yl)-2-(1′-h yd r oxyet h yl)-3-h yd r oxy-
p yr id in -4(1H)-on e h yd r och lor id e (19e): 82%; mp 117-120
°C. 1H NMR (DMSO-d6) δ: 1.5 (d, 3H, CHCH3), 1.65-2.45 (m,
2H, N-CH2CH2CH2O), 3.45 (t, 2H, N-CH2CH2CH2O), 4.65 (m,
4H, N-CH2CH2CH2O), 5.5 (s, 2H, CHCH3), 7.3 (d, 1H, 5-H(py-
ridinone)), 8.18 (d, 1H, 6-H(pyridinone)), 7.3-9.4 (br, 4H, OH).
MS (EI): m/z 214 [(M - Cl)+•]. Anal. (C10H16NO4Cl) C, H, N.
1,6-Dim eth yl-2-h ydr oxym eth yl-3-h ydr oxypyr idin -4(1H)-
1
on e h yd r och lor id e (19f): 87.5%; mp 140-143 °C. H NMR
(DMSO-d6) δ: 2.7 (s, 3H, 6-CH3), 4.06 (s, 3H, N-CH3), 4.86 (s,
2H, 2-CH2OH), 7.4 (s, 1H, 5-H(pyridinone)), 6.4-8.7 (br, 3H,
OH). MS (EI): m/z 170 [(M - Cl)+•]. Anal. (C8H12NO3Cl‚1/2H2O)
C, H, N.
1-Eth yl-2-h yd r oxym eth yl-3-h yd r oxy-6-m eth ylp yr id in -
1
4(1H)-on e h yd r och lor id e (19g): 85%; mp 160-162 °C. H
NMR (DMSO-d6) δ: 1.3 (t, 3H, N-CH2CH3), 2.5 (s, 3H, 6-CH3),
4.3 (q, 2H, N-CH2CH3), 4.6 (s, 2H, 2-CH2O), 7.1 (s, 1H,
5-H(pyridinone)), 7.8-10.0 (br, OH). MS (EI): m/z 184 [(M -
Cl)+•]. Anal. (C9H14NO3Cl) C, H, N.
1,6-Dim et h yl-2-(1′-h yd r oxyet h yl)-3-h yd r oxyp yr id in -
4(1H)-on e h yd r och lor id e (19h ): 82%; mp 208-212 °C. 1H
NMR (DMSO-d6) δ: 1.4 (d, 3H, CHCH3), 2.5 (s, 3H, 6-CH3),
4.04 (s, 3H, N-CH3), 5.65 (q, 1H, CHCH3), 7.3 (s, 1H, 5-H(py-
ridinone)), 7.5-10.0 (br, 3H, OH). MS (EI): m/z 184 [(M -
Cl)+•]. Anal. (C9H14NO3Cl) C, H, N.
3. Deter m in a tion of Distr ibu tion Coefficien ts. Distri-
bution coefficients were determined using an automated
continuous flow technique.12,13 The system comprised an IBM
compatible PC running the Omniferous Personal Computer
Auto-Titrator “TOPCAT” program, which controlled both a
Metrohm 665 dosimat autoburette and a Pye-Unicam Lambda
5 UV/vis spectrophotometer, as well as performing all calcula-
tions of distribution coefficients. All distribution coefficient
determinations were performed using AnalaR grade reagents
under a nitrogen atmosphere using a flat-based glass vessel
equipped with a sealable lid at 25 °C. The aqueous and octanol
phases were presaturated with respect to each other before
use. The filter probe consisted of a polytetrafluoroethylene
(PTFE) plunger associated with a gel-filtration column. The
aqueous phase (50 mM MOPS buffer, pH 7.4, prepared using
Milli-Q water) was separated from the two-phase system (1-
octanol/MOPS buffer, pH 7.4) by means of a hydrophilic
cellulose filter (5 µm diameter, 589/3 Blauband filter paper;
Schleicher and Schuell) mounted in the gel-filtration column
adjuster (SR 25/50, Pharmacia). A known volume (normally
25-50 mL) of MOPS buffer (saturated with octanol) is taken
in the flat base mixing chamber. After a base line was obtained
the solution was used for reference absorbance. The compound
to be examined was dissolved in buffer (saturated with octanol)
so as to give an absorbance of between 0.5-1.5 absorbance
units at the preselected wavelength (∼280 nm). The flow rate
of the aqueous circuit was limited to 1 mL. The computer
program calculates the distribution coefficient (D7.4) for each
octanol addition.
1,6-Dim eth yl-2-(1′-h yd r oxyp r op yl)-3-h yd r oxyp yr id in -
4(1H)-on e h yd r och lor id e (19i): 79.7%; mp 221-223 °C. H
NMR (DMSO-d6) δ: 0.8 (t, 3H, CHCH2CH3), 1.3-2.1 (m, 2H,
CHCH2CH3), 2.43 (s, 3H, 6-CH3), 3.94 (s, 3H, N-CH3), 5.3 (t,
1H, CHCH2CH3), 7.15 (s, 1H, 5-H(pyridinone)), 7.5-10.5 (br,
3H, OH). MS (EI): m/z 198 [(M - Cl)+•]. Anal. (C10H16NO3Cl)
C, H, N.
1
8-Oxo-4,8-d ih yd r o-2-p h en yl-5-[3′-ben zoyloxy)p r op yl]-
4H -p yr id in o[3,2-d ]-m -d ioxin (20). A solution of triphen-
ylphosphine (3.46 g, 13.2 mmol) and 18d (3.3 g, 12 mmol) in
dry tetrahydrofuran (100 mL) was added dropwise to a solution
of diethyl azodicarboxylate (2.3 g, 13.2 mmol) and benzoic acid
(1.5 g, 12 mmol) in dry tetrahydrofuran (30 mL) at room
temperature. After the mixture stirred overnight at room
temperature, the solvent was removed under reduced pressure.
The residue thus obtained was purified by column chroma-
tography on silica gel (eluant:methanol:chloroform, 12:88 v/v)
1
yielding a light yellow oil (4.1 g, 89.7%). H NMR (CDCl3) δ:
1.95-2.55 (m, 2H, N-CH2CH2CH2O), 3.82 (t, 2H, N-CH2CH2-
CH2O), 4.34 (t, 2H, N-CH2CH2CH2O), 4.9 (s, 2H, CH2O), 5.8
(s, 1H, CHPh), 6.3 (d, 1H, 7-H(pyridinone)), 7.0-8.2 (m, 11H,
Ar & 6-H(pyridinone)).
1-[(3′-(Be n zoyloxy)p r op yl]-2-h yd r oxym e t h yl-3-h y-
d r oxyp yr id in -4(1H)-on e Hyd r och lor id e (21). A solution of
20 (4.1 g, 10 mmol) in N,N-dimethylformamide (50 mL) was
adjusted to pH 1 with concentrated hydrochloric acid prior to
hydrogenolysis for 6 h in the presence of 5% Pd/C catalyst
(1.0g). Filtration followed by rotary evaporation in vacuo gave
a white solid. Recrystallization from methanol/diethyl ether
gave a white crystalline solid (2.9 g, 85%): mp 142-143 °C.
1H NMR (DMSO-d6) δ: 1.9-2.8 (m, 2H, N-CH2CH2CH2O),
4.0-5.0 (m, 4H, N-CH2CH2CH2O), 4.8 (s, 2H, CH2O), 7.2-
Biologica l Meth od : Ir on Mobiliza tion Effica cy Stu d y
in [59F e]F er r itin -Loa d ed Ra ts. Hepatocytes of normal fasted
rats (190-230 g) were labelled with 59Fe by administration of
[59Fe]ferritin from tail vein.11 One hour later, each rat was

2-(1′-Hydroxyalkyl)-3-hydroxypyridin-3-ones
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 23 4823
(9) Porter, J . B.; Singh, S.; Hayes, K. P.; Epemolu, O.; Abeysinghe,
R. D.; Hider, R. C. Lessons from preclinical and clinical studies
with 1,2-diethyl-3-hydroxypyridin-4-one, CP94 and related com-
pounds. Adv. Expt. Med. Biol. 1994, 356, 361-370.
(10) Singh, S.; Choudhury, R.; Epemolu, R. O.; Hider, R. C. Metabo-
lism and pharmacokinetics of 1-(2′-hydroxyethyl)- and 1-(3'-
hydroxypropyl)-2-ethyl-3-hydroxypyridin-4-ones in the rat. Eur.
J . Drug Metab. Pharmacokinet. 1996, 21, 33-41.
(11) Liu, Z. D.; Lu, S. L.; Hider, R. C. In vivo iron mobilisation
evaluation of hydroxypyridinones in 59Fe-ferritin loaded rat
model. Biochem. Pharmacol. 1999, 57, 559-566.
(12) Dobbin, P. S.; Hider, R. C.; Hall, A. D.; Taylor, P. D.; Sarpong,
P.; Porter, J . B.; Xiao, G.; van der Helm, D. Synthesis, physico-
chemical properties, and biological evaluation of N-substituted
2-alkyl-3-hydroxy-4(1H)-pyridinones: Orally active iron chela-
tors with clinical potential. J . Med. Chem. 1993, 36, 2448-2458.
(13) Rai, B. L.; Dekhordi, L. S.; Khodr, H.; J in, Y.; Liu, Z.; Hider, R.
C. Synthesis, Physicochemical properties and evaluation of
N-substituted-2-alkyl-3-hydroxy-4(1H)-pyridinones. J . Med. Chem.
1998, 41, 3347-3359.
(14) Mitsunobu, O. The use of diethylazodicarboxylate and triphen-
ylphosphine in synthesis and transformation of nature products.
Synthesis 1981, 1, 1-28.
(15) Pippard, M. J .; J ohnson, D. K.; Finch, C. A. A rapid assay for
evaluation of iron chelating agents in rats. Blood 1981, 58, 685-
692.
administered orally with chelator (450 µmol/kg). Control rats
were administered with an equivalent volume of water. The
rats were placed in individual metabolic cages and urine and
feces collected. Rats were allowed access to food one hour after
oral administration of chelator. There was no restriction of
water throughout the study period. The investigation was
terminated 24 h after the [59Fe]ferritin administration, rats
were sacrificed, and the liver and gastrointestinal tract
(including its content and feces) were removed for gamma
counting. The “iron mobilization” and “efficacy” were calculated
according to following equations:
iron mobilization (%) )
59Fe-activity(gut&feces)
× 100% (1)
59Fe-activity(gut&feces)
+
59Fe-activity(liver)
efficacy (%) ) iron mobilization (%) - control (%) (2)
Ack n ow led gm en t. The authors thank Apotex Re-
search Inc., Canada, and Biomed EC Grant BMH4-
CT97-2149 for supporting this research project.
(16) Hider, R. C.; Khodr, H.; Liu, Z. D.; Tilbrook, G. Optimisation of
pFe3+ values of iron(III) ligands with clinical potential. In Metal
Ions in Biology and Medicine Volume 5; Collery, Ph., Bra¨tter,
P., Negretti de Bra¨tter, V., Khassanova, L., Etienne, J .C., Eds.;
J ohn Libbey Eurotext: Paris, 1998; pp 51-55.
(17) Hider, R. C.; Epemolu, O.; Singh, S.; Porter, J . B. Iron chelator
design. In Progress in Iron Research; Hershko, C., Konijn, A.
M., Aisen, P. M., Eds.; Plenium Press: New York, 1994; Vol.
356, pp 343-349.
(18) Xu, J . D.; Kullgren, B.; Durbin, P. W.; Raymond, K. N. Specific
sequestering agents for the actinides. 28: Synthesis and initial
evaluation of multidentate 4-carbamoyl-3-hydroxy-1-methyl-
2(1H)-pyridinone ligands for in vivo plutonium(IV) chelation. J .
Med. Chem. 1995, 38, 2606-2614.
Refer en ces
(1) Brittenham, G.M. Disorders of iron metabolism: deficiency and
overload. In Hematology: Basic Principles and Practice; Hoff-
man, R., Benz, E., Shattil, S., Furie, B., Cohen, H., Eds.;
Churchill Livingstone: New York, 1991; pp 327-349.
(2) Hershko, C.; Konijn A.M.; Link, G. Iron chelators for thala-
saemia. Br. J . Haematol. 1998, 101, 399-406.
(3) Tilbrook, G. S.; Hider, R.C. Iron Chelators for clinical use. In
Metal Irons in Biological Systems. Vol. 35: Iron Transport and
Storage in Microorganisms, Plants and Animals; Sigel, A., Sigel,
H., Eds.; Marcel Dekker: New York, 1998; pp 691-730.
(4) Brittenham, G. M. Development of iron-chelating agents for
clinical use. Blood 1992, 80, 569-574.
(5) Singh, S.; Epemolu, O.; Dobbin, P. S.; Tilbrook, G. S.; Ellis, B.
L.; Damani, L. A.; Hider, R. C. Urinary metabolic profiles in man
and rat of 1,2-dimethyl- and 1,2-diethyl substituted 3-hydroxy-
pyridin-4-ones. Drug Metab. Dispos. 1992, 20, 256-261.
(6) Porter, J . B.; Morgan, J .; Hoyes, K. P.; Burke, L. C.; Huehns, E.
R.; Hider, R. C. Relative oral efficacy and acute toxicity of
hydroxypyridin-4-one iron chelators in mice. Blood 1990, 76,
2389-2396.
(7) Porter, J . B.; Abeysinghe, R. D.; Hoyes, K. P.; Barra, C.; Huehns,
E. R.; Brooks, P. N.; Blackwell, M. P.; Araneta, M.; Brittenham,
G.; Singh, S.; Dobbin, P.; Hider, R. C. Contrasting interspecies
efficacy and toxicology of 1,2-diethyl-3-hydroxypyridin-4-one,
CP94, relates to differing metabolism of the iron chelating site.
Br. J . Haematol. 1993, 85, 159-168.
(19) Achmatowicz, J r., O.; Bukowski, P.; Szechner, B.; Zwierzchows-
ka, Z.; Zamojski, A. Synthesis of methyl 2,3-dideoxy-DL-alk-2-
enopyranosides from furan compounds; A general approach to
the total synthesis of monosaccharides. Tetrahedron 1971, 27,
1973-1996.
(20) Tate, B. E.; Miller, R. L. U.S.A. Patent 3 130 204, 1964.
(21) Ellis, B. L.; Duhme, A. K.; Hider, R. C.; Hossain, M. B.; Rizvi,
S.; van der Helm, D. Synthesis, physicochemical properties, and
biological evaluation of hydroxypyranones and hydroxypyridi-
nones: Novel bidentate ligands for cell-labeling J . Med. Chem.
1996, 39, 3659-3670.
(22) Taylor, P. D.; Morrison, I. E. G.; Hider, R. C. Microcomputer
application of nonlinear regression analysis to metal-ligand
equilibria. Talanta 1988, 35, 507-512.
(23) Gran, G. Determination of the equivalence point in potentio-
metric titration. Analyst 1952, 77, 661-671.
(8) Bergeron, R.J .; Streiff, R.R.; Wiegand, J .; Luchetta, G.; Creary,
E.A.; Peter, H.H. A comparison of the iron-clearing properties
of 1,2-dimethyl-3-hydroxypyrid-4-one, 1,2-diethyl-3-hydroxy-
pyrid-4-one, and Deferoxamine. Blood 1992, 79, 1882-1890.
J M991080O