2-Alkyloxyalkylthiohypoxanthines as new potent inhibitors of xanthine oxidase.

Article abstract of DOI:10.1016/S0014-827X(01)01160-0

The title compounds were prepared and tested as xanthine oxidase (XO) inhibitors. Results evidenced that potency was related to the position of the oxygen atom in the 2-linear chain and that it grew with distance from the sulfur atom until it became equipotent to 2-n-hexylthiohypoxanthine. Enzymatic oxidation on C(2) occurred in the 8-alkylthiohypoxanthines. On the contrary, oxidation on C(8) did not occur in the 2-alkythioderivatives, demonstrating that the chain forced these molecules to form a complex with molybdenum(VI) involving only the N(3) and N(9) nitrogen atoms.

Full text of DOI:10.1016/S0014-827X(01)01160-0

www.elsevier.com/locate/farmac
Il Farmaco 56 (2001) 809–813
2-Alkyloxyalkylthiohypoxanthines as new potent inhibitors of
xanthine oxidase
Giuliana Biagi, Irene Giorgi *, Federica Pacchini, Oreste Livi, Valerio Scartoni
Dipartimento di Scienze Farmaceutiche, Uni6ersita` di Pisa, 6ia Bonanno, 6, 56100 Pisa, Italy
Received 20 February 2001; accepted 30 May 2001
Abstract
The title compounds were prepared and tested as xanthine oxidase (XO) inhibitors. Results evidenced that potency was related
to the position of the oxygen atom in the 2-linear chain and that it grew with distance from the sulfur atom until it became
equipotent to 2-n-hexylthiohypoxanthine. Enzymatic oxidation on C(2) occurred in the 8-alkylthiohypoxanthines. On the
contrary, oxidation on C(8) did not occur in the 2-alkythioderivatives, demonstrating that the chain forced these molecules to
form a complex with molybdenum(VI) involving only the N(3) and N(9) nitrogen atoms. © 2001 Elsevier Science S.A. All rights
reserved.
Keywords: Xanthine oxidase; Inhibitors; 2-Thiohypoxanthine; 2-Mercaptohypoxanthine; 2-Thioxo-1,2,3,7(9)-tetrahydro-purin-6-one
was able to receive a linear alkyl chain; this goal was
1. Introduction
obtained through the same favourable interaction with
The enzyme xanthine oxidase (XO) catalyses the
both 2-n-alkylhypoxanthines and 8-n-alkylhypoxanthi-
hydroxylation of hypoxanthine and of xanthine to yield
uric acid and superoxide anions. These superoxide an-
nes [6–8]. In the first case, 2-substituted compounds
demonstrated good inhibitory activity whereas, in the
ions have been linked to postischaemic tissue injury and
second case, 8-substituted derivatives showed substrate
oedema [1] as well as to vascular permeability [2]. XO
behaviour. Further, the fact that 2-n-alkoxycarbonyl-8-
can also oxidise synthetic purine drugs, such as an-
azahypoxanthines were more active than the corre-
tileukaemic 6-mercaptopurine, with loss of their phar-
macological properties [3]. Then, the control of the
sponding 2-n-alkyl-8-azahypoxanthines [9] allowed us
to hypothesise that the part of this pocket nearest to the
action of XO may help the therapy of some diseases.
site of interaction of the enzyme with the purine nu-
Today, the therapy of gout makes use of allopurinol, a
cleus has some residues capable of polar interactions
potent inhibitor of XO known for a long time [4].
with atoms, bearing electron lone pairs, at the begin-
However, given its side effects and its inability to
ning of the chain. This concept was supported with the
prevent the formation of free radicals by the enzyme
preparation, among others, and biological evaluation of
[5], the research on new XO inhibitors is in progress.
In the past we reported the good activity as inhibitors
of XO of 2-alkylhypoxanthines and 2-alkyl-8-azahy-
poxanthines [6–10] and a type-II binding with enzyme
was hypothesised according to the scheme proposed by
Robins et al. [4]. We demonstrated that near the active
site of the enzyme only one lipophilic narrow pocket,
characterised by well-defined dimensions, existed which
2-n-alkylthio-hypoxanthines, which were very potent.
In fact, the 2-n-hexylthiohypoxanthine had a Ki value
of 9.8 nM, thus being 700 times more potent than
allopurinol [10].
Starting from this lead compound, we projected to
further explore the pocket facing the C(2) position of
the purine nucleus by the modification of the linear
chain bound to it. In the present paper we describe the
preparation and biological evaluation of the title com-
pounds with the aim of obtaining some other informa-
tion about the enzyme zone which interacts with the
side-chain.
* Corresponding author.
E-mail address: igiorgi@farm.unipi.it (I. Giorgi).
0014-827X/01/$ - see front matter © 2001 Elsevier Science S.A. All rights reserved.
PII: S0014-827X(01)01160-0

810
G. Biagi et al. / Il Farmaco 56 (2001) 809–813
with carbon disulfide in ethanol at 120 °C [20]. Com-
pound 9 was obtained by the reaction of 8 with bromo-
hexane, DBU and N,N-dimethylformamide [21];
compound 12 was obtained by the same chemical
method starting from 11, or by the enzymatic reaction
with XO from 8-hexylthio-hypoxanthine (9).
3. Experimental
3.1. Chemistry
Melting points were determined on a Kofler hot-
stage apparatus and are uncorrected. IR spectra in
Nujol mulls were recorded on a Perkin–Elmer Model
1
1310 spectrometer. H NMR spectra were recorded on
a Bruker AC 200 spectrometer in l units from TMS as
an internal standard; the compounds were dissolved in
the DMSO. HPLC were carried out on a Violet PM
900 with a PH-410 UV photometer as detector, and a
Nucleosil C18 column (15×4.8 mm) using as eluent
MeOH–phosphate buffer pH 7.5 70:30.
Scheme 1. Synthetic routes to compounds 3-6, 9, 12.
2. Chemistry
Mass spectra were performed on a Hewlett–Packard
GC/MS System 5988A. TLC was performed on pre-
coated silica gel F254 plates (Merck) or on precoated
RP-18 F254 plates (Merck). Flash-column chromatogra-
phies were performed using Merck Kieselgel 60 (230–
400 mesh). Microanalyses (C, H, N) were carried out
on a Carlo Erba elemental analyser (Model 1106) and
were within 90.4% of the theoretical values.
The synthetic routes are summarised in Scheme 1.
The 2-thiohypoxanthine (2) was obtained by cyclisation
of 4,5-diamino-6-hydroxy-2-thiopyrimidine (1) with an-
hydrous formamide, as reported by Beaman in 1954
[11] at very high temperatures (200–250 °C); the very
good yield obtained (94%) was higher than the one
reported in the original article.
Compound
2
reacted with the suitable bro-
3.1.1. General synthesis of
moalkylether, obtained by known methods [12–18], in
the presence of 1,8-diazabicyclo-(5.4.0)-undec-7-ene
(DBU) and N,N-dimethylformamide to give com-
pounds 3–6.
By fusion of 4,5-diamino-6-hydroxypyrimidine sul-
fate (7) or 4,5-diamino-2,6-dihydroxypyrimidine sulfate
(10) and thiourea, the 8-thio-hypoxanthine (8) [19] and
the 8-thio-xanthine (11) [20], respectively, were ob-
tained in a good yield. Compound 11 was also obtained
by heating 4,5-diamino-2,6-dihydroxypyrimidine sulfate
2-alkyloxyalkylthio-hypoxanthines (3–6)
To a solution of 2 (1.68 g, 10 mmol) in the minimum
amount of N,N-dimethylformamide, DBU (1.52 g, 10
mmol) and the suitable bromoalkylether (Table 1) (1.66
g, 10 mmol) were added. The mixture was stirred for
24–52 h (Table 1) at room temperature, then water was
slowly added until a solid precipitated. After filtration
the solid was flash-chromatographed using as eluent
CHCl3–MeOH 9:1 obtaining the pure product which
crystallised from MeOH (Tables 1 and 2).
Table 1
Reaction and physical data of compounds 3–6, 9 and 12
Comp.
Bromoalkylether
Reaction time (h)
Empirical formula
(molecular weight)
M.p. (°C)
Rf
Yield (%)
3
4
5
6
9
12
BrCH2O(CH2)3CH3
BrCH2CH2O(CH2)2CH3
BrCH2(CH2)2OCH2CH3
BrCH2(CH2)3OCH3
Br(CH2)5CH3
24
48
48
52
12
48
C10H14N4O2S (254)
C10H14N4O2S (254)
C10H14N4O2S (254)
C10H14N4O2S (254)
C11H16N4OS (252)
C11H16N4O2S (268)
275
280
277
265
268
268
0.28 a
0.27 a
0.30 a
0.25 a
0.37 b
0.31 b
42
74
48
35
56
53
Br(CH2)5CH3
a Silica gel, eluent CHCl3–MeOH 9:1.
b RP18, eluent MeOH–H2O 6:4.

G. Biagi et al. / Il Farmaco 56 (2001) 809–813
811
Table 2
Spectroscopic data for compounds 3–6, 9, and 12
Comp. 1H NMR
MS
Aromatic H Aliphatic H
Exchang. H m/z (%)
3
8.13 (s, 1H) 5.48 (s, 2H); 3.59 (m, 2H); 1.45 (m, 2H); 1.24 (m, 2H); 0.82 (t, 3H)
7.99 (s, 1H) 3.62 (s, 2H); 3.62 (m, 2H); 3.36 (m, 4H); 1.49 (m, 2H); 0.84 (t, 3H)
8.08 (s, 1H) 4.18 (s, 2H); 3.28 (m, 2H); 3.10 (m, 2H); 1.95 (m, 2H); 1.1 (t, 3H)
7.61 (s, 1H) 3.37 (m, 7H); 2.03 (m, 4H)
254 (M+, 17.3); 168 (21); 57
(100)
4
8.31 (1H)
12.2 (1H)
254 (M+, 6.8); 168 (100); 110
(23); 43 (38)
5
254 (M+, 1.1); 208 (65); 168
(100); 44 (97)
6
254 (M+, 5.1); 168 (100); 110
(43)
9
7.89 (s, 1H) 3.18 (t, 2H); 1.62 (m, 2H); 1.22 (m, 6H); 0.81 (m, 3H)
7.92 (s, 1H) 3.12 (t, 2H); 1.58 (t, 2H); 1.24 (m, 6H); 0.82 (t, 3H)
12.2 (1H)
11.3 (2H)
252 (M+, 5.5); 205 (27); 168
(100); 43 (29)
12
268 (M+, 15.6); 184 (100); 85
(58); 43 (42)
3.1.2. 8-Hexylthio-hypoxanthine (9)
tometrically in an air-saturated Tris buffer–HCl at pH
7.6, I=0.1 at 2590.2 °C using a Beckman DU 50
spectrophotometer with a thermostated cuvette holder.
The increase in uric acid (enzymatic reaction product)
concentration was evaluated at 295 nm with hypoxan-
thine as substrate (10 mM) and XO sufficient to obtain
an average reaction rate for the control reaction of
0.10090.005 absorbance units/min. The inhibitors in
question were dissolved in DMSO. DMSO concentra-
tion in all the assays was kept at 3% v/v, which has no
effect on XO activity. IC50 values (the concentration
required to produce 50% inhibition of the enzyme
catalysed reaction) were determined from least-square
analysis of the linear portion of log dose–inhibition
curves. Each curve was generated using at least three
concentrations of inhibitor producing an inhibition be-
tween 20 and 80%, with three replicates at each concen-
To a solution of 8-thio-hypoxanthine (8) (0.20 g, 1.19
mmol) in the minimum amount of N,N-dimethylfor-
mamide, DBU (0.18 g, 1.18 mmol) and 1-bromohexane
(0.19 g, 1.18 mmol) were added. The mixture was
stirred for 12 h at room temperature, then water was
slowly added until a solid precipitated. After filtration,
the solid was crystallised from MeOH to give pure 9
(0.168 g, 0.66 mmol) (Tables 1 and 2).
3.1.3. 8-Hexylthio-xanthine (12)
To a solution of 8-thioxanthine (11) (0.50 g, 2.72
mmol) in the minimum amount of N,N-dimethylfor-
mamide, DBU (0.41 g, 2.69 mmol) and 1-bromohexane
(0.41 g, 2.51 mmol) were added. The mixture was
stirred for 48 h at room temperature, then water was
slowly added until a solid precipitated. After filtration,
the solid was flash-chromatographed using as eluent
CHCl3–MeOH 9:1 obtaining pure product which crys-
tallised from MeOH to give 12 (0.386 g, 1.44 mmol)
(Tables 1 and 2). Compound 12 was obtained also by
the following enzymatic method. To a solution of 9 (10
mg, 0.04 mmol) in 1 ml of DMSO and 80 ml of
phosphate buffer at pH 7.6 was added XO (see Section
3.2) (40 ml); the mixture was stirred and air bubbled at
35 °C for 12 h, then was evaporated and the residue
was analysed by HPLC showing only a peak at the
same retention time as compound 12 obtained by chem-
ical reaction.
tration. KM determination for compound
9 was
performed by plotting 1/V versus 1/[S] (2–6 mM) read-
ing the absorbance change at 300 nm which evaluated
the increase in 12 (enzymatic reaction product)
concentration.
Kinetic studies were performed with at least three
different concentrations of inhibitor in the presence of
variable concentrations of hypoxanthine (6–10 mM). Ki
values (the dissociation constants of the enzyme–in-
hibitor complex) were determined from the slopes in
double reciprocal plots [22].
3.2. Biochemistry
4. Results and discussion
XO (from buttermilk, 1.36 U/mg) was purchased
from Boehringer (Boehringer Mannheim Italia S.p.A.)
and hypoxanthine from Fluka (Sigma–Aldrich s.r.l.)
and used as a 100 mM solution in a Tris buffer–HCl at
pH 7.6. Compound 9 was used as substrate in 100 mM
DMSO solution. XO activity was assayed spectropho-
Biological results showed that the introduction of an
oxygen atom in the linear chain leads to a modulation
of the inhibitory potency among tested compounds
(Table 3). In particular, IC50 values decreased little by
little with the increase of oxygen atom distance from
the sulfur atom, i.e. 6 (IC50 28 nM)B5 (IC50 58 nM)B

812
G. Biagi et al. / Il Farmaco 56 (2001) 809–813
thioalkyl chain in 2 position. The trend registered
among compounds 3, 4, 5, and 6 could be imputed at
the start to a different conformation of the chain within
the lipophilic pocket with respect to the usual descrip-
tion of the linear chains, especially in the case of
compound 3. In this case, in which a methylene unit
was interposed between the sulfur and oxygen atoms,
lone pairs of these two atoms, by their reciprocal
repulsive action, should force, in our opinion, the linear
chain to assume a conformation different from the
linear one. Instead, molecular modelling with minimisa-
tion of energy (Chemoffice and Insight II programs)
indicated that the chain retained a linear conformation
in all cases as in the thioalkyl one. Therefore, the low
potency of compound 3 could be explained taking into
account the repulsive effect provoked by a negative
charged group facing the oxygen atom of the chain.
This group could be the negative part of a protein
fragment with dipole characteristics, whose positive
part, near the narrow entrance of the lipophilic pocket,
influenced positively the bind with the sulfur atom.
Then as the SꢀO distance grew from 3 to 6 the repulsive
effect became progressively smaller and smaller.
Furthermore, to clarify the binding mechanism of
2-alkyloxy-alkylthiohypoxanthines with active site of
XO, we prepared 8-hexylthio-hypoxanthine (9) and its
2-hydroxy analogue (12).
4 (IC50 670 nM)B3 (IC50 20 560 nM), allopurinol IC50
4200 nM. Compound 6, Ki 19 nM, resulted almost
equipotent to compound 13, Ki 9.8 nM, prepared and
assayed in a previous paper [10], having a linear
Table 3
Biological results
Comp.
R1
R2
IC50 Ki or Km
(nM) (nM)
3
4
5
6
9
12
13
SCH2O(CH2)3CH3
SCH2CH2O(CH2)2CH3
SCH2(CH2)2OCH2CH3
SCH2(CH2)3OCH3
H
H
H
H
H
20 560
670
58 Ki: 23
28 Ki: 19
560 Km: 3860
660
S(CH2)5CH3
S(CH2)5CH3
H
OH
S(CH2)5CH3
Ki: 9.8 a
4200 Ki: 7000 b
Allopurinol
a See Ref. [10].
b See Ref. [4].
Both these compounds showed an inhibitory be-
haviour with IC50 560 and 660 nM, respectively. Com-
pound 9, contrary to the title compounds, resulted also
a substrate as it was transformed by enzyme (XO)
oxidation in a compound which resulted identical to
compound 12 by HPLC analysis. These results indi-
cated that the enzyme active site was able to accept
both compounds with the linear chain in position 2 or
8. However the enzyme was able to oxidise in position
2 only the compounds bearing a chain on C(8) like 9
(Fig. 1), whereas oxidation on C(8) in compound like 6,
having the chain in position C(2), was not allowed
probably owing to too long a distance between the
sulfur anion of the enzyme prosthetic group and C(8)
(Fig. 2).
Fig. 1. Type II binding of C(8) substituted compounds: oxidation on
C(2) is allowed.
This last result can be explained hypothesising that
only N(3) and N(9) atoms, according to the binding
type II proposed by Robins [4], participated in the
complex with molybdenum(VI) present in the enzyme
active site; molecules like 9 or 6 would orient them-
selves within it in a way suitable to insert the linear
chain inside the lipophilic pocket (Figs. 1 and 2). Oxi-
dation on C(8) in compounds with the chain in position
2 did not occur since the chain position did not allow
the formation of the complex of molybdenum(VI) with
O6 and N(7) according to Robins binding type I (Fig.
3).
Fig. 2. Type II binding for C(2) substituted compounds: oxidation on
C(8) is not allowed.

G. Biagi et al. / Il Farmaco 56 (2001) 809–813
813
inhibition: effect of an n-alkyl substituent on C-2 of the nucleus
of 8-azahypoxanthine, Farmaco 41 (1989) 843–849.
[7] G. Biagi, I. Giorgi, O. Livi, V. Scartoni, I. Tonetti, Xanthine
oxidase inhibition: effect of a linear carboalkoxy substituent on
C-2 of the nucleus of 8-azahypoxanthine, Farmaco 45 (1990)
979–988.
[8] G. Biagi, I. Giorgi, O. Livi, V. Scartoni, I. Tonetti, A. Lucac-
chini, Xanthine oxidase (XO): relative configuration of com-
plexes formed by the enzyme, 2- or 8-n-alkyl hypoxanthines and
2-n-alkyl-8-azahypoxanthine. XII, Farmaco 48 (1993) 357–374.
[9] G. Biagi, I. Giorgi, O. Livi, V. Scartoni, I. Tonetti, L.
Costantino, Xanthine oxidase (XO). 4(5)-Aminosubstituted-5(4)-
carboxyamido-1H-1,2,3-triazoles: a new class of monocyclic tria-
zole inhibitors, Farmaco 50 (1995) 257–264.
[10] G. Biagi, A. Costantini, L. Costantino, I. Giorgi, O. Livi, P.
Pecorari, M. Rinaldi, V. Scartoni, Synthesis and biological eval-
uation of new imidazole, pyrimidine, and purine derivatives and
analogs as inhibitors of xanthine oxidase, J. Med. Chem. 39
(1996) 2529–2535.
[11] A.G. Beaman, New syntheses of purine, J. Am. Chem. Soc. 76
(1954) 5633–5636.
[12] C.M. Blair, H.R. Henze, Bromomethyl normal alkyl ethers, J.
Am. Chem. Soc. 54 (1932) 399–401.
Fig. 3. Type I binding is not allowed for C(2) substituted compounds.
Acknowledgements
[13] A. Karvonen, Beitrage zur Kenntnis der Halogenather, Chem.
Zentralbl. II (1912) 1266–1271.
[14] M.H. Palomaa, A. Kenetti, Studien uber atherartige Verbindun-
gen, IV: Synthese und synthetische Verwendung der 2-Halo-
genather, Chem. Ber. 64 (1931) 797–804.
This research was supported by the Ministero della
Ricerca Scientifica e Tecnologica (MURST).
References
[15] A. Wolkow, B.N. Menschutkin, Zur Umwandlung des
Trimethylens in Propylen, Chem. Ber. 31 (1898) 3070.
[16] L. Wartski, Sur une pre´paration d’oxacyclanes. I. Synthe`se de
me´thoxy g et l haloge´nures, Bull. Soc. Chim. Fr. (1965) 3066–
3076.
[17] A. Kirrmann, N. Hamaide, Nouvelle pre´paration d’oxacyclanes,
Bull. Soc. Chim. Fr. (1957) 789.
[18] O. Riobe`, G. Geeraert, Symmetric nonanetriol, Compt. Rend.
231 (1950) 151–152.
[1] D.J. Hearse, A.S. Manning, J. Downey, D.M. Yellon, Xanthine
oxidase: a critical mediator of myocardial injury during ischemia
and reperfusion?, Acta Physiol. Scand. Suppl. 548 (1986) 65–78.
[2] P.H. Chan, J.W. Schimidley, R.A. Fishman, S.M. Longar, Brain
injury, edema and vascular permeability changes induced by
oxygen-derived free radicals, Neurology 34 (1984) 315–320.
[3] H.S.D. Naeff, M.C.R. Franssen, H.C. van der Plas, Quantitative
structure–activity relationship (QSAR) studies of the inhibition
of xanthine oxidase by heterocyclic compounds, Recl. Trav.
Pays-Bas 110 (1991) 139–150.
[19] R. Robins, Potenzial purine antagonists. XV. Preparation of
some 6,8-disubstituted purines, J. Am. Chem. Soc. 80 (1958)
6671–6679.
[20] T.L. Loo, M. Michael, E. Garceau, J.C. Reid, 6-Thiouric acid—
a metabolite of 6-mercaptopurine, J. Am. Chem. Soc. 81 (1959)
3039–3341.
[21] N. Ono, H. Miyake, T. Saito, A. Kaji, A convenient synthesis of
sulfides, formaldeide dithioacetals, and chloromethylsulfides,
Synthesis (1980) 952–953.
[22] L. Costantino, G. Rastrelli, A. Albasini, Inhibitory activity of
flavonols towards the xanthine oxidase enzyme, Int. J. Pharm. 86
(1992) 17–23.
[4] R.K. Robins, G.R. Revankar, D.E. O’Brien, R.H. Springer, T.
Novinson, A. Albert, K. Senga, J.P. Miller, D.G. Streeter,
Purine analog inhibitors of xanthine oxidase. Structure–activity
relationships and proposed binding of the molybdenum cofactor,
J. Heterocycl. Chem. 22 (1985) 601–634.
[5] A. Haberland, H. Luther, I. Schinker, Does allopurinol prevent
superoxide radical production by xanthine oxidase (XOD)?,
Agents Actions 32 (1991) 96–97.
[6] G. Biagi, I. Giorgi, O. Livi, V. Scartoni, Xanthine oxidase