Novel 4,5-diaryl-3-hydroxy-2(5H)-furanones as anti-oxidants and anti-inflammatory agents

Article abstract of DOI:10.1016/S0968-0896(02)00053-6

In order to study the effect of phenol moieties on biological activities of ascorbic acid derivatives, we synthesized 13 novel 4,5-diaryl-3-hydroxy-2(5H)-furanones 5a-m with various substitution patterns. Compound 5g bearing a 2,3-dihydroxy phenyl ring on

Full text of DOI:10.1016/S0968-0896(02)00053-6

Bioorganic & Medicinal Chemistry 10 (2002) 1647–1658
Novel 4,5-Diaryl-3-hydroxy-2(5H)-furanones as Anti-Oxidants
and Anti-Inflammatory Agents
Valerie Weber,a Pascal Coudert,a,* Catherine Rubat,b Eliane Duroux,b
Danielle Vallee-Goyet,c Daniel Gardette,c Marc Bria,d Eliane Albuisson,e
Fernand Leal,a Jean-Claude Gramain,c Jacques Couqueleta and Michel Madesclairea
a
´
´
Laboratoire de Chimie Therapeutique, Faculte de Pharmacie, 28, place Henri Dunant, 63001, Clermont-Ferrand Cedex, France
b
´
Laboratoire de Pharmacologie, Faculte de Pharmacie, 28, place Henri Dunant, 63001, Clermont-Ferrand Cedex, France
c
`
´
`
Equipe Synthese Organique UMR 6504, Universite Blaise Pascal, 63177, Aubiere Cedex, France
dCentre commun de mesure RMN/RPE, USTL, 59655 Villeneuve d’Ascq, France
e
´
Laboratoire de Biomathematiques et Informatique Medicale, Faculte de Medecine, 28, place Henri Dunant,
´
´
´
63001, Clermont-Ferrand Cedex, France
Received 22 May 2001; revised 20 June 2001; accepted 28 January 2002
Abstract—In order to study the effect of phenol moieties on biological activities of ascorbic acid derivatives, we synthesized 13
novel 4,5-diaryl-3-hydroxy-2(5H)-furanones 5a–m with various substitution patterns. Compound 5g bearing a 2,3-dihydroxy
phenyl ring on the 5-position of the heterocycle appeared to be the most powerful anti-oxidant furanone with reducing activity
against DPPH (IC50=10.3 mM), superoxide anion quenching capacity (IC50=0.187 mM) and lipid peroxidation inhibitory effect
(IC50=0.129 mM). To ascertain determinant molecular features for anti-oxidant activities, structure–activity relationships were
studied. Lipophilicity and molecular parameters related to electron distribution and structure (difference in heats of formation
between the compound and its radical or its cation radical, energy of the highest occupied molecular orbital, HOMO) were found to
correlate with the anti-oxidant action of compounds 5 in the different tests used. Oxygen-derived free radicals are known to con-
tribute to inflammatory disorders; therefore we have investigated effects of compounds 5 in two models of inflammation: phorbol
ester-induced ear edema in mice (TPA-test) and carrageenan-induced paw edema in rat. At 100 mg/kg ip in the TPA-test, the anti-
inflammatory activity of compounds 5 was potent compared with that of indomethacin and ketorolac and all the results suggested a
cyclooxygenase inhibition in the emergence of such properties. The combined pharmacological actions of compounds 5 associated
with a favorable therapeutic index prompt with interesting perspectives for their use in heart and brain disorders as well as in
inflammatory diseases. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
scavengers such as a-tocopherol (vitamin E), b-caro-
tene, glutathione and ascorbic acid (vitamin C).3
Ascorbate is one of the most potent naturally occuring
anti-oxidants because it works at the same time directly
by reaction with aqueous peroxyl radicals and indirectly
by restoring the anti-oxidant properties of fat-soluble
vitamin E.4 Furthermore, it plays an important part in
the prevention of a large number of chronic diseases.5,6
However, the low solubility of ascorbic acid in lipophilic
environments as well as its susceptibility to thermal and
oxidative degeneration has led to interest in the synth-
esis of derivatives with better lipophilicity and increased
stability.7ꢀ17
Multiple medical conditions1,2 such as myocardial
infarction, carcinogenesis and neurologic trauma, to
name a few, could be heightened by the presence of
oxygen free radicals. These radicals appear to be toxic
mainly because they initiate the chain reaction of lipid
peroxidation. These small molecular units could split
DNA, RNA, proteins, cellular membranes and cellular
organization. Under normal conditions, aerobic organ-
isms have evolved anti-oxidant defences to protect them
against free radicals: enzymes such as superoxide dis-
mutase, glutathione peroxidase, anti-oxidants and radical
Due to the biological interest of this class of compounds,
we previously reported the syntheses and the anti-oxidant
properties of lipophilic ascorbic acid analogues 1 and 2
*Corresponding author. Tel.: +33-4-7360-8016; fax: +33-4-7328-
2707; e-mail: pascal.coudert@u-clermont1.fr
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00053-6

1648
V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
Chart 1.
Scheme 1. Reagents and conditions: (i) DBU, CH3I, DMF, OꢂC;
Chart 2.
(ii) DBU, DMF, 0 ꢂC; (iii) HCl.
(Chart 1).18,19 Although these furanone derivatives were
more effective than ascorbic acid both as superoxide
anion scavengers and as lipid peroxidation inhibitors,
the substitution of the furan nucleus by either a benzoyl
or an acetyl group in the 4-position appeared to enhance
toxicity (LD50 ꢁ200 mg/kg ip) with regard to vitamin C
(LD50 >800 mg/kg ip). Consequently, we investigated
here the anti-oxidant properties of a series of novel
3-hydroxy furanones 5 (Chart 2) having aromatic sub-
stituents directly linked to the heterocycle in the 4- and
5-positions, in order to identify suitable features for
optimal activity and minimal toxic effects.
Thus, anti-oxidant effects of these new derivatives have
been evaluated in vitro by means of three tests:
measurement of the reducing activity on the stable
2,2-diphenyl-1-picrylhydrazyl radical (DPPH), super-
oxide anion scavenging assay and lipid peroxidation
assay.
Scheme 2. Reagents and conditions: (i) DBU, DMF, 0 ꢂC; (ii) Pd/C,
H2, EtOH/THF.
Furthermore, free radicals are well known to play an
important part in the inflammatory process.20 Super-
oxide anion radicals, hydrogen peroxide and hydroxyl
radicals, produced by activation of phagocytes, are
considered to be implicated in inflammation and tissue
destruction. Free radicals are also involved in the bio-
synthesis of prostaglandins, which are important medi-
ators in inflammatory process.21 Therefore, the
evaluation of anti-inflammatory properties of furanones
5 seemed to be of great interest.
1
determination of furanones 5 was achieved by IR, H
and 13C NMR spectral studies and purity of products
was confirmed by elemental analysis. All compounds 5
were obtained as racemates.
Results and Discussion
In vitro anti-oxidant activities
Anti-oxidants can act at several different levels in the
oxidative sequence and their mechanisms of action can
be multiple.25 So, we chose to investigate anti-oxidant
activities of furanones 5 by means of three different tests
(Table 1).
Chemistry
The synthesis of 4,5-diaryl-3-hydroxyfuranones 5a–c, e–j,
l–m was accomplished by reaction of methyl arylpyr-
uvates 4 with appropriate benzaldehydes in the presence
of 1,5-diazabicyclo[5.4.0]undecene (DBU) in dimethyl-
formamide,22ꢀ24 as shown in Scheme 1. Methyl aryl-
pyruvates 4 were prepared by methylating arylpyruvic
acids 3 according to the procedure described by Namiki
et al.22 Preparation of 5d and 5k from arylpyruvates 4
failed so these compounds were then obtained by
hydrogenolysis of 7a and 7b (Scheme 2). Structural
First, the free radical scavenging properties of tested
compounds were evaluated by decrease in the absorp-
tion of the stable 2,2-diphenyl-1-picrylhydrazyl radical
(DPPH) at 517 nm. This bleaching of DPPH absorption
occurs when the odd electron of the radical is paired.
Thus, it is representative of the capacity of furanones 5 to
scavenge free radicals, independently from any enzymatic

V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
Table 1. Anti-oxidant activities and lipophilicity of derivatives 5
1649
Compd
Ar1
Ar2
Reducing
activity
on DPPH
IC50 (mM)
Superoxide
anion
scavenging
activity
IC50 (mM)
Lipid
peroxidation
IC50 (mM)
log k0
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
C6H5
C6H5
C6H5
C6H5
C6H5
C6H5
C6H5
C6H5
C6H5
C6H5
2-OH C6H4
3-OH C6H4
970.0
34.0
698.0
586.0
165.0
762.0
10.3
12.2
20.4
23.2
39.5
20.5
22.0
46.6
19.6
NT
0.499
0.277
0.705
NT
NT
1.219
0.436
28.0ꢃ2.1a
NT
1.75
1.59
1.46
1.45
1.80
1.90
1.43
1.07
1.58
1.54
1.15
1.44
1.53
—
4-OH C6H4
2-OCH3 C6H4
4-OCH3 C6H4
2,3-diOH C6H3
2,5-diOH C6H3
2-OH, 3-OCH3 C6H3
2-OH, 5-OCH3 C6H3
3,5-diOCH3, 4-OH C6H2
C6H5
NT
NT
0.187
0.559
>0.8
0.277
0.886
NT
0.129
0.210
0.413
0.152
0.170
NT
NT
Pro-oxidant
NT
C6H5
C6H5
4-OH C6H4
4-OH C6H4
4-OCH3 C6H4
NT
24.0ꢃ1.5b
Ascorbic acid
Vitamin E
NT
—
NT, not tested.
a% inhibition at 0.25 mg/mL.
b% inhibition at 1 mg/mL.
Scheme 3.
activity. The presence of substituents on either aromatic
ring bound to the furan skeleton led to more potent
derivatives 5b–m (10.3ꢁIC50ꢁ762 mM) than the
unsubstituted parent compound 5a (IC50=970 mM).
Obviously, the increased anti-oxidant activity of
derivatives 5b, 5g–m is essentially due to the presence of
a phenol moiety on the olide ring. The presence of a
hydroxyl group at C-20 (5b) in ring Ar2 is associated
with a high reducing activity against DPPH (IC50=34
mM), which is confirmed by the weak one of the corre-
sponding compound 5e (IC50=165 mM) having a meth-
oxy group at this position. Moreover, the IC50 values of
5c (698 mM) and 5d (586 mM) were far higher than that
of 5b, suggesting that a meta- or para-substituent on the
phenyl nucleus is unfavourable for activity. In addition,
the IC50 values of diphenols 5g, 5h (10.3 and 12.2 mM,
respectively) were found to be 2 to 3 times weaker than
that of monophenol 5b or polysubstituted compounds
5i–k (20<IC50 values <40 mM). The potent activity of
compound 5g could be due to the weakness of OH
linkages related to the mesomeric electron-donating
effect of the OH substituent in the ortho-position with
regard to the phenoxyl radical. Thus, 5g could form a
semiquinonic intermediate stabilized by electronic de-
localization and, consequently, easily oxidized to pro-
duce a stable orthoquinonic compound as previously
observed with catechin (Scheme 3) by Sawaı and
Sakata.26 In addition, this semiquinonic intermediate is
stabilized by hydrogen bonding which prevents such a
radical to generate secondary radical reactions. The
formation of the semiquinonic radical of 5g was corro-
borated by electron spin resonance (ESR) measurement.

1650
V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
Scheme 4.
Oꢄ2ꢀ could be protecting agents against cellular injury
during reperfusion of ischemic tissues. Therefore,
superoxide anion scavenging activity of furanones 5 was
explored using a non-enzymic biological generator of
superoxide anion radical, for instance phenazine
methosulfate NADH system.30 This test was limited by
the solubility of products in aqueous solution. So, it was
realized on compounds 5b, c, d, g, h, j, k, whose solubi-
lity is remarkably increased by the introduction of a
hydroxyl group on the phenyl nucleus in the 5-position.
Unfortunately, it could not have been done on deriva-
tives 5a, e, f, l and 5m due to their poor solubility and 5i
was investigated only at doses up to 0.25 mg/mL.
Except compound 5i, all tested furanones exhibited
noteworthy scavenging activity with IC50 values ranging
from 0.187 to 0.886 mM (Table 1). Although clear
structure–activity relationships are not evident, the
presence of a hydroxyl group in the 3-position on Ar2
appeared to increase the efficiency of molecules (5c vs
5b, 5d and 5g vs 5h, 5i). Furanone 5g was again found to
be the most active compound in this test, what could be
explained by the capacity of superoxide anions to
deprotonate diphenolic compounds,31 as shown in
Scheme 4 for furanone 5g. In addition, all furanones
exhibited a better superoxide anion scavenging activity
than ascorbic acid considered as a reference, with only
24% inhibition at a concentration of 1 mg/mL. Vitamin
E was not tested because of its insolubility in aqueous
medium.
Figure 1. ESR spectrum of stable semiquinone radical generated from
compound 5g. (A) Experimental spectrum recorded after NaOH
addition. Spectrometer setting: microwave power, 10 mW, modula-
tion amplitude, 0.5 gauss, scan range, 100 gauss, frequency, 9.42
GHz. (B) Simulated ESR spectrum.
Thus, addition of NaOH to aerated solution of 5g gen-
erated a stable radical, which is presented in Figure 1
under our experimental conditions, 5g with a catechol
function was sensitive to autoxidation and gave rise to a
stable free radical. The ESR spectrum consists of a six-
line spectrum: a small hyperfine coupling constant of
1.38 gauss and two coupling constant of 3.75 gauss. The
result clearly indicated the formation of the semi-
quinonic radical of 5g.27 Similarly, 2,5-diOH substitu-
tion on the phenyl ring in derivative 5h could lead to a
stable parasemiquinonic derivative which find expres-
sion in potent reducing activity. The presence of a
hydroxyl group in the 4-position on either Ar1 or Ar2
seems to differently affect anti-oxidant properties in this
series of derivatives. Thus, the superior activity of 5l, 5m
versus 5d could be explained by the fact that these two
compounds are vinylogous analogues of the enolic form
of ascorbic acid. Finally, compared with conventional
anti-oxidants, diphenols 5g and 5h were slightly super-
ior to vitamin C and vitamin E (IC50=46.6 and 19.6
mM, respectively).
Excessive lipid peroxidation in biomembranes is
thought to be related to several abnormal states produ-
cing peroxyl radicals ROOꢄ resulting from chain reac-
tions carried by free radicals.32ꢀ35 Because phenolic
compounds and ascorbic acid have ROOꢄ trapping and
z ꢄOH scavenging abilities,34,35 the effects of furanones 5
(Table 1) were investigated on the Fe2+/ADP-ascorbate
induced lipid peroxidation of rat liver microsomes. We
observed potent antioxidative activities for most com-
pounds (5g–k). Therefore, di- or trisubstitution on Ar2
seems to be required for efficient free radical quenching
by furanones as it is proved by comparing 5g–k (IC50
values ꢁ0.413 mM) with 5b–d (IC50 values ꢅ0.436
mM). Methylating one hydroxyl group of 5g, providing
5i, decreased the efficiency of scavenging activity,
The first product of the univalent reduction of oxygen is
the superoxide anion, Oꢄ2ꢀ, which is generated in many
biological processes.28,29 So, compounds able to scavenge

V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
Table 2. Computed molecular properties of derivatives 5
1651
nevertheless, 5i remained much more effective than 5b.
An electron-donating methoxy group in the ortho-
position could increase the stabilization of the aryloxyl
radical, formed by the abstraction of hydroxyl hydro-
gen, through electron delocalization. This hypothesis
appears again consistent when comparing 5k
(IC50=0.170 mM) and 5d (28% inhibition at 0.25 mg/
mL). However, the nature of substituent (hydroxy or
methoxy group) in the 5-position on Ar2 seems to
moderately influence anti-oxidant effect in the test (5j
versus 5h). As previously reported,35ꢀ37 vitamin C
underwent autoxidative destruction in presence of iron
initiating lipid peroxidation. However, this ‘pro-oxi-
dant’ activity in vitro has obviously no significance in
vivo and ascorbic acid works as anti-oxidant under
most circumstances.36,37
Compd
ÁHf
molecule
(kcal
ÁÁHf
enol
(kcal
ÁÁHf
phenol
(kcal
ÁÁHox
(kcal
E HOMO
molecule
(eV)
molꢀ1
)
molꢀ1
)
molꢀ1
)
molꢀ1
)
5a
5b
5c
5d
5e
5f
ꢀ39.74
ꢀ81.37
ꢀ83.66
ꢀ84.08
ꢀ74.73
ꢀ77.78
ꢀ127.94
25.22
24.90
25.13
25.23
24.63
25.24
25.11
—
192.85
187.90
193.31
192.99
188.16
192.94
190.58
ꢀ9.05
ꢀ9.04
ꢀ9.05
ꢀ9.05
ꢀ8.88
ꢀ9.02
ꢀ8.96
38.66
39.31
39.52
—
—
5g
32.25 (OH en 2)
33.49 (OH en 3)
33.58 (OH en 2)
36.28 (OH en 5)
36.10
33.43
34.58
37.44
37.47
5h
ꢀ127.25
25.22
187.39
ꢀ9.01
5i
5j
5k
5l
5m
ꢀ121.67
ꢀ121.03
ꢀ156.98
ꢀ84.26
25.05
25.11
25.03
25.63
24.56
190.33
194.94
197.20
187.22
186.54
ꢀ8.93
ꢀ8.92
ꢀ8.98
ꢀ8.77
ꢀ8.76
ꢀ122.46
The hydrophilic or lipophilic character of anti-oxidants
greatly influenced them to act as radical scavengers in
aqueous phase or as chain-breaking anti-oxidants in
biological membranes.38,39 Therefore, we chose to
determine capacity factors (log k0) from reversed-phase
high-performance liquid chromatography, which esti-
mate the hydrophobicity of compounds. A mobile phase
composition containing 50% MeOH appeared to be
most reliable to predict the conventionally hydrophobic
parameter log P value for molecules possessing hydro-
gen-bonding substituents.40 Although lipophilicity has a
marked impact on activity, no clear correlation was
found between lipophilicity and anti-oxidant properties
whatever the test used (Table 1).
were considered for the study. The energy of electron
abstraction (ÁÁHox) was calculated as the difference in
heats of formation between the cation radical and the
parent compound. This parameter was used because
anti-oxidant action could involve single-electron trans-
fer followed by proton transfer as an alternative or in
addition to H-abstraction. The energy of the HOMO
(highest occupied molecular orbital) was computed and
determines the ionization potential and, therefore, will
correlate with the ability of enol and/or phenol moieties
to donate electrons.
Compound 5g appears to possess ideal structural fea-
tures for anti-oxidant activities, considering its reactiv-
ity as hydrogen or electron-donating agent, its
superoxide anion scavenging property and its possible
interaction with hydroxyl and lipid peroxyl radicals. In
addition, its propensity for metal chelations, particu-
larly iron and copper, supports the role of any poly-
phenolic derivative41 and, thus, the role of 5g as
preventive anti-oxidant in terms of inhibiting transition
metal catalysed free radical formation.
First, a number of statistical investigations were con-
ducted with the aim of finding correlations between
DPPH reactivity on the one hand and calculated
molecular and reactivity properties as well as lipophili-
city on the other. Eq 1 was obtained from the 13 tested
furanones:
À
Á
log ð1=IC50Þ ¼ À 0:027 ðÆ0:020Þ ÁÁHf phenol Ar1
À
Á
þ 0:015 ðÆ0:012Þ ÁÁHf phenol Ar2
Structure–activity relationships
À 0:074 ðÆ0:048Þ ÁÁHox
þ 7:269 ðÆ2:512Þ EHOMO
In order to find a possible explanation for the observations
described above and to provide a better understanding
of the influence of structure on the anti-oxidant activity
in a molecular way, different physicochemical para-
meters were calculated using Sybyl 6.6 software42 and
examined, in particular the enthalpy of homolytic H–O
bond cleavage (ÁÁHf), the energy of electron abstrac-
tion (ÁÁHox) and the energy of the HOMO (Table 2).
The enthalpy of homolytic H–O bond cleavage (ÁÁHf)
was determined as the difference in heats of formation
between the radical formed after abstraction of
hydroxyl enolic or phenolic hydrogen by drugs and the
parent compound. The more stable the radical, the
weaker will be the H–O bond in the enol and/or phenol
groups. The weaker this bond, the more readily it will
be cleaved by an attacking active oxygen species and
stronger will be the activity. Thus, in the case of diphe-
nols 5g and 5h, only the lowest ÁÁHf phenol values
ð1Þ
À 1:279 ðÆ0:828Þ log k0 þ 85:101 ðÆ23:482Þ
n ¼ 13; r2 ¼ 0:774; s ¼ 0:470; F ¼ 4:798; p ¼ 0:0318
In this equation, the number in parentheses is the 95%
confidence interval associated with the coefficients, n is
the number of compounds used, r2 is the squared cor-
relation coefficient corresponding to the fraction of
observed variance accounted for, s is the standard
deviation and F (as also p) the statistical significance of
fit. Calculated electron abstraction energy and ionization
potential are important electronic parameters to predict
the reactivity towards DPPH, although lipophilicity also
plays a determinant part. Surprisingly, ÁÁHf (enol) did
not interfere in eq 1, contrary to ÁÁHf (phenol), what

1652
V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
was explained by the fact that this former parameter was
practically constant in the studied molecules (range
24.25–25.63). This last result also corroborates the
importance of phenol moieties and supports our
hypothesis about a part of the action mechanism evoked
for 5g and 5h.
Finally, the best correlation for lipid peroxidation is
given by eq 3:
log ð1=IC50Þ ¼ 1:160 ðÆ0:284Þ ÁÁHf ðenolÞ
À
Á
À 0:086 ðÆ0:010Þ ÁÁHf phenol Ar2
However, the modest quality of eq 1 suggests that
other complex interactions, not encoded in the para-
meters used (e.g., intermolecular and intramolecular
H-bonds),43,44 could influence the reaction.
þ 0:037 ðÆ0:007Þ ÁÁHox
À 29:574 ðÆ7:511Þ
ð3Þ
n ¼ 7; r2 ¼ 0:985; s ¼ 0:061; F ¼ 64:939; p ¼ 0:0032
Concerning superoxide anion scavenging activity,
regression analysis resulted in non-significance of
equations with the same set of variables as for DPPH
study. From these data, the activity of 5c appeared to
be fundamentally different from the calculated value
and is certainly notably influenced by other factors
than those considered in establishing correlations.
Consequently, excluding 5c, a predictive model descri-
bed by eq 2 was found using stepwise multilinear
regression:
Examination of eq 3 leads to the conclusion that activity
is governed by the enthalpy of homolytic H–O bond
cleavage. Contrary to all expectations, no or less sig-
nificant correlations were found with lipophilicity.
Despite of the useful indications furnished by eqs 1–3,
the traditional QSAR approach has revealed its limi-
tations on the key structural elements that can modulate
the anti-oxidant activity of furanones. For example, as
far as ÁÁHf values are concerned in the emergence of
scavenging activity, the difference in sign between the
slopes of the equations gives evidence for a complex
multifactorial mechanism of action, whatever the test
considered. However, similarly to a study published by
Ancerewiz et al.,45 it is interesting to note that the
potency of 5 to inhibit the oxidation of lipids can be
usefully predicted, using a simple in vitro reactivity test,
as exemplified by eq 4:
À
Á
log ð1=IC50Þ ¼ À 0:131 ðÆ0:027Þ ÁÁHf phenol Ar2
À 3:291 ðÆ1:531Þ EHOMO
þ 1:375 ðÆ0:221Þ log k0
À 23:494 ðÆ13:097Þ
ð2Þ
n ¼ 6; r2 ¼ 0:971; s ¼ 0:071; F ¼ 22:722; p ¼ 0:0424
log ð1=IC50Þlipo ¼ À 0:423 ðÆ0:043Þlog ð1=IC50ÞDPPH
À
Á
This analysis reveals three parameters to be the best
predictors of superoxide anion scavengers for the stud-
ied compounds. Eq 2 shows that ÁÁHf (phenol Ar2) is
the less influencial property whereas the ability of the
molecule to donate electrons predominates.
À 0:187ðÆ0:010Þ ÁÁHf phenol Ar2
ð4Þ
þ 12:055ðÆ0:513Þ
n ¼ 7; r2 ¼ 0:991; s ¼ 0:041; F ¼ 211:572; p < 0:0001
Table 3. Anti-inflammatory activity of derivatives 5
Compound
TPA-induced ear edema test
% inhibition of edema at
Carrageenin-induced paw edema test
% inhibition of edema at
100 mg/kg ip
200 mg/kg po
100 mg/kg ip
200 mg/kg po
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
48.4ꢃ5.6b
NT
NT
NT
NT
35.5ꢃ0.4a
NT
NT
NT
NT
NT
23.8ꢃ4.5a
24.7ꢃ5.5a
NT
52.4ꢃ5.4b
39.4ꢃ0.4b
48.4ꢃ6.5b
NT
58.3ꢃ7.1b
5.8ꢃ5.5 (NS)
29.8ꢃ1.8 (NS)
29.0ꢃ10.3a
48.3ꢃ1.4b
NT
NT
8.7ꢃ5.5 (NS)
5.3ꢃ5.7 (NS)
NT
NT
NT
52.6 ꢃ7.5b
73.8ꢃ6.5b
NT
Inactive
Inactive
NT
NT
NT
62.8ꢃ9.9b
6.2ꢃ7.2 (NS)
27.7ꢃ12.5 (NS)
44.2ꢃ8.4b
NT
NT
NT
21.7ꢃ4.5a
NT
NT
77.5ꢃ7.7b
5.6ꢃ5.5 (NS)
18.8ꢃ6.4 (NS)
9.3ꢃ7.3 (NS)
83.9ꢃ4.1b,d
59.7ꢃ4.9b
54.8ꢃ0.4b,c
53.3ꢃ9.1b,c
76.5ꢃ5.4b,d
Inactive
52.6ꢃ8.8a
45.6ꢃ5.3a
49.1ꢃ10.5a
Indomethacin
Ketorolac
Dexamethasone
13.8ꢃ10.1 (NS)
10.6ꢃ8.2 (NS)
93.3ꢃ2.0b,d
NT, not tested; NS, not significant.
a0.001<p <0.05.
bp <0.001.
cTested at 15 mg/kg ip.
dTested at 1 mg/kg ip.

V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
1653
In vivo anti-inflammatory activities
of action resulting from both anti-oxidant and anti-
inflammatory activities. However, compounds 5 did not
exhibit significant anti-inflammatory effects when admi-
nistered by oral route, suggesting a possible low bio-
availability due to glucuronidation, as was classically
observed for phenol derivatives.56
Reactive oxygen species are well known to be implicated
in the induction and prolongation of inflammatory
process.20,21,46,47 In addition, a number of commercially
available non steroidal anti-inflammatory drugs
(NSAIDs) and analgesics such as salicylates, acetamino-
phen, indomethacin and nimesulide, have been shown
to possess radical scavenging properties.48ꢀ51 All these
observations prompted us to investigate derivatives 5 as
anti-inflammatory drugs.
Conclusions
Our study provides evidence that furanones 5 exhibit
interesting anti-oxidant properties expressed either by
their reducing activity against the stable radical DPPH
or their capacity to scavenge superoxide anion and to
inhibit lipid peroxidation. Appropriate modifications of
previous structures 1 and 2 lead to products with a
favorable therapeutic index. In this series, free radical
scavenging activity is favored by the presence at the
same time of enol and phenol moieties as shown by
compound 5g. Moreover, the introduction of two aro-
matic rings on the furanone skeleton enhances the lipo-
philicity of compounds 5 compared to vitamin C and,
thus, is expected to contribute to their effective access,
retention and interaction with biological membranes.
So, derivatives 5 appeared to be related to tocopherol,
especially with respect to lipid peroxidation. This fact
suggests interesting perspectives for the use of com-
pound 5g in the treatment of heart and brain disorders
involving free radicals, such as myocardial infraction or
stroke. Preliminary studies indicate that some furanones
5 possess, by ip route, anti-inflammatory activity com-
parable with that of classical NSAIDs and even higher
in the TPA-induced ear edema test. The combined anti-
oxidant, anti-inflammatory and lipophilic effects of
compound 5g could also have therapeutic value in
inflammatory diseases, such as rheumatoid arthritis, in
which synovial lipid hydroperoxides provide a reservoir
of potential oxygen-derived free radicals.
At first, behavioral effects and intraperitoneal acute
toxicity of compounds 5 were examined in mice. Neither
of tested drugs produced significant behavioral effects,
even at doses up to 800 mg/kg ip and all animals were
still alive after an observation period of 1 week. Conse-
quently, these results enhance the interest of ascorbic
acid derivatives 5 as anti-oxidant drugs compared to
previously described products 1 and 2, the therapeutic
margin of which was fairly weak.
With regard to evaluation of furanones 5 as anti-
inflammatory compounds, they were first administered
ip at the dose of 100 mg/kg in the phorbol 12-myristate
13-acetate (TPA)-induced ear edema test (Table 3).
Only six furanones, 5a, 5d, 5e, 5g, 5h, 5m, exhibited a
noteworthy activity and produced more than 45% inhi-
bitory effects. If compounds 5e, 5g, 5h, 5m, which
demonstrated anti-oxidant properties in either of the
tests used, were active to diminish TPA inflammation,
surprisingly, the unsubstituted parent compound 5a, the
least efficient to reduce DPPH, exhibited significant
anti-inflammatory activity. Moreover, lipophilicity did
not seem to play a major role since one of the highest
lipophilic compound 5e (log k0=1.80) and the lowest
one 5h (log k0=1.07) were equipotent with about 60%
inhibition of edema. In contrast, classical NSAIDs
indomethacin and ketorolac, failed to produce sig-
nificant effect after ip administration, since maximum
reduction of ear thickness was only 13.8% at 100 mg/
kg. These results, consistent with literature data,52
demonstrate the particular interest of furanones 5 as
potent anti-inflammatory compounds, even though
their effectiveness remained clearly less important than
that of steroid dexamethasone (93.3% inhibition at 1
mg/kg ip). Considering that the TPA-induced ear edema
test is primarily mediated by PGE2,53,54 derivatives 5
could be assumed to possess cyclo-oxygenase inhibitory
activity. However, to confirm this hypothesis, the in
vivo anti-inflammatory efficacy of furanones 5a, 5d, 5e,
5g, 5h and 5m was assessed by using the functional
model of carrageenan-induced paw edema in the rat.55
This test is particularly efficient in detecting compounds
whose activity is the result of inhibition of pros-
taglandin amplification. At 100 mg/kg ip compounds
5g, 5h and 5m demonstrated a significant inhibition of
the edema similar to that of indomethacin administered
at the low dose of 15 mg/kg ip (Table 3). These results
corroborate those obtained in the TPA-induced ear
edema test in the mouse. Potency of furanones 5 in the
TPA-induced ear edema test compared to that of refer-
ence compounds is probably due to a multiple mechanism
Experimental
General chemical methods
All starting materials used for synthesis were reagent-
grade from Janssen Chimica (Noisy-le-Grand, France).
Reagents for biochemical assays were purchased from
Sigma (Montlucon, France). Melting points were deter-
mined on a Reichert apparatus (Isi, Paris, France) and
were uncorrected. Column chromatography was carried
out on SDS silica gel 60 (70–200 mesh). The infrared
(IR) spectra were recorded with a FTIR-Nicolet Impact
410 spectrophotometer (Thermo Optek, Montigny le
Bretonneux, France). The proton and carbon nuclear
magnetic resonance (1H NMR and 13C NMR) spectra
were recorded on a Bruker AC-400 (400 MHz) spectro-
meter (Bruker, Wissembourg, France). Chemical shifts
are reported in parts per million (d ppm) relative to
tetramethylsilane used as an internal standard. Abbre-
viations used for signal patterns are: br s, broad sing-
let; s, singlet; d, doublet; m, multiplet. The ultra violet
(UV) measurements were carried out with a DU-70
spectrometer (Beckman Instruments, Fullerton, USA).

1654
V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
Elemental analysis were done by the Service Central
d’Analyses, CNRS, Vernaison, France.
EtOAc/cyclohexane (2/8), gave compound 7a in 34%
yield; mp 223 ꢂC; IR (KBr) n 3306, 1747, 1681, 1608,
1
1512 cmꢀ1; H NMR (DMSO-d6) d 11.02 (br s, 1H),
4,5-Diaryl-3-hydroxy-2(5H)-furanones (5). General pro-
cedure. To a cold (0 ꢂC) solution of a mixture of methyl
2-oxo-3-phenylpropionate 4 (5.27 mmol) and appro-
priate benzaldehyde (1.3 equiv, 6.85 mmol) in dry DMF
(24mL) was added DBU (0.80 mL, 5.35 mmol) under
stirring. The mixture was stirred for 3–5 h at 0 ꢂC and
then poured into a mixture of ethyl acetate (10 mL) and
HCl 1M (50 mL). The organic layer was separated and
the aqueous layer was extracted with AcOEt. The
organic layers were combinated, dried over Na2SO4 and
evaporated in vacuo. The oily residue was then purified
either by column chromatography (AcOEt/cyclohexane)
or by precipitation in water, filtration and crystal-
lization from ethyl ether.
7.61 (d, 2H), 7.42–7.20 (m, 10H), 6.99 (d, 2H), 6.52 (s,
1H), 5.06 (s, 2H); 13C NMR (DMSO-d6) d 169.2 (C-2),
159.0 (C-OBn), 138.8 (C-3), 130.7 (C-4) 136.8, 128.7,
127.8 (Cipso), 129.4, 128.6, 128.5, 128.3, 127.9, 127.8,
127.5 (CH aromatic), 79.5 (C-5), 69.3 (CH2).
3-Hydroxy-5-(4-hydroxyphenyl)-4-phenyl-2(5H)-furanone
(5d). 0.10 g of 10% Pd/C was added to 5-(4-benzyloxy-
phenyl)-3-hydroxy-4-phenyl-2(5H)-furanone 7a (0.40 g,
1.1 mmol) dissolved in THF/EtOH and submitted to a
hydrogen atmosphere for 4 h. After filtration of the
catalyst and evaporation of the solvent, crystallization
from ethyl ether afforded 0.26 g (88%) of 5d; mp 232 ꢂC;
IR (KBr) n 3254, 1732, 1679, 1614, 1519, 1449 cmꢀ1; 1H
NMR (DMSO-d6) d 11.05 (br s, 1H), 9.73 (br s, 1H),
7.63 (d, 2H), 7.36 (t, 2H), 7.28 (t, 1H), 7.21 (d, 2H), 6.77
(d, 2H), 6.48 (s, 1H); 13C NMR (DMSO-d6) d 169.2 (C-
2), 158.2 (C-OH), 138.8 (C-3), 130.7 (C-4), 129.3, 128.4,
128.2, 127.4, 115.6 (CH aromatic), 127.7, 126.6 (Cipso),
79.8 (C-5). Anal. calcd for C16H12O4: C, 71.64; H, 4.51.
Found: C, 71.43; H, 4.50.
4,5-Diphenyl-3-hydroxy-2(5H)-furanone (5a). This was
prepared from methyl 2-oxo-3-phenylpropionate and
benzaldehyde. Precipitation in water and crystallization
from ethyl ether led to compound 5a in 72% yield; mp
ꢂ
1
202 C; IR (KBr) n 3285, 1731, 1676, 1454 cmꢀ1; H
NMR (DMSO-d6) d 11.09 (br s, 1H), 7.63 (d, 2H), 7.42–
7.23 (m, 8H), 6.58 (s, 1H); 13C NMR (DMSO-d6) d
169.1 (C-2), 138.8 (C-3), 136.6, 128.0 (Cipso), 130.5 (C-
4), 129.3, 129.0, 128.5, 128.4, 127.9, 127.4 (CH aro-
matic), 79.8 (C-5). Anal. calcd for C16H12O3: C, 76.18;
H, 4.79. Found: C, 76.32; H, 4.78.
3-Hydroxy-5-(2-methoxyphenyl)-4-phenyl-2(5H)-furanone
(5e). This was prepared from methyl 2-oxo-3-phenyl-
propionate and o-anisaldehyde. Precipitation in water
and crystallization from ethyl ether led to compound 5e
in 22% yield; mp 217 ꢂC; IR (KBr) n 3274, 1736, 1676,
1
3-Hydroxy-5-(2-hydroxyphenyl)-4-phenyl-2(5H)-furanone
(5b). This was prepared from methyl 2-oxo-3-phenyl-
propionate and 2-hydroxybenzaldehyde. Precipitation
in water and crystallization from ethyl ether afforded
compound 5b in 56% yield; mp 166 ꢂC; IR (KBr) n
1600, 1584, 1490, 1461 cmꢀ1; H NMR (DMSO-d6) d
10.99 (br s, 1H), 7.57 (d, 2H), 7.34–7.24 (m, 4H), 7.10–
7.03 (m, 2H), 6.88–6.86 (m, 1H), 6.74 (s, 1H), 3.89 (s,
3H); 13C NMR (DMSO-d6) d 169.4 (C-2), 157.7 (C–
OCH3), 139.0 (C-3), 130.8, 128.5, 128.3, 128.1, 127.0,
120.8, 111.9 (CH aromatic), 127.6 (C-4), 127.9, 123.9
(Cipso), 74.3 (C-5), 55.9 (OCH3). Anal. calcd for
C17H14O4: C, 72.33; H, 5.00. Found: C, 72.61; H, 4.99.
3268, 1750, 1674, 1605, 1497, 1459 cmꢀ1 1H NMR
;
(DMSO-d6) d 10.94 (br s, 1H), 10.10 (br.s, 1H), 7.62 (d,
2H), 7.34 (t, 2H), 7.25 (t, 1H), 7.15 (t, 1H), 6.97 (d, 1H),
6.89 (d, 1H), 6.73 (d, 1H), 6.71 (s, 1H); 13C NMR
(DMSO-d6) d 169.6 (C-2), 155.9 (C-OH), 138.8 (C-3),
130.9 (C-4), 127.9, 122.1 (Cipso), 130.3, 128.5, 128.2,
128.0, 127.0, 119.4, 116.0 (CH aromatic), 74.6 (C-5).
Anal. calcd for C16H12O4: C, 71.64; H, 4.51. Found: C,
71.39; H, 4.50.
3-Hydroxy-5-(4-methoxyphenyl)-4-phenyl-2(5H)-furanone
(5f). This was prepared from methyl 2-oxo-3-phenyl-
propionate and p-anisaldehyde. Precipitation in water
and crystallization from ethyl ether gave compound 5f
in 32% yield; mp 188 ꢂC; IR (KBr) n 3293, 1742, 1674,
1
1615, 1517 cmꢀ1; H NMR (DMSO-d6) d 11.07 (br s,
3-Hydroxy-5-(3-hydroxyphenyl)-4-phenyl-2(5H)-furanone
(5c). This was prepared from methyl 2-oxo-3-phenyl-
propionate and 3-hydroxybenzaldehyde. Purification by
column chromatography, eluting with EtOAc/cyclohex-
ane (3/7), gave compound 5c in 65% yield; mp 215 ꢂC;
1H), 7.63 (d, 2H), 7.34–7.27 (m, 5H), 6.93 (d, 2H), 6.54
(s, 1H), 3.73 (s, 3H); 13C NMR (DMSO-d6) d 169.5 (C-
2), 159.9 (C–OCH3), 138.8 (C-3), 130.6 (C-4), 129.3,
128.5, 128.3, 127.4, 114.3 (CH aromatic), 128.4, 128.0
(Cipso), 79.5 (C-5), 55.1 (OCH3). Anal. calcd for
C17H14O4: C, 72.33; H, 5.00. Found: C, 72.53; H, 4.98.
IR (KBr) n 3392, 1726, 1682, 1604, 1458 cmꢀ1 1H
;
NMR (DMSO-d6) d 11.05 (br s, 1H), 9.54 (s, 1H), 7.62
(d, 2H), 7.34 (t, 2H), 7.27 (t, 1H), 7.17 (t, 1H), 6.86 (d,
1H), 6.73 (m, 2H), 6.48 (s, 1H); 13C NMR (DMSO-d6) d
169.2 (C-2), 157.7 (C-OH), 138.7 (C-3), 137.9, 128.0
(Cipso), 130.6 (C-4), 130.6, 128.5, 128.4, 127.4, 118.8,
116.4, 114.0 (CH aromatic), 79.7 (C-5). Anal. calcd for
C16H12O4: C, 71.64; H, 4.51. Found: C, 71.39; H, 4.49.
5-(2,3-Dihydroxyphenyl)-3-hydroxy-4-phenyl-2(5H)-fura-
none (5g). This was prepared from methyl 2-oxo-
3-phenylpropionate and 2,3-dihydroxybenzaldehyde.
Purification by column chromatography, eluting with a
gradient of EtOAc/cyclohexane, afforded compound 5g
in 44% yield; mp 189 ꢂC; IR (KBr) n 3357, 1720, 1672,
1
1604 cmꢀ1; H NMR (DMSO-d6) d 10.89 (br s, 1H),
5-(4-Benzyloxyphenyl)-3-hydroxy-4-phenyl-2(5H)-furanone
(7a). This was prepared from methyl 2-oxo-3-phenyl-
propionate and 4-benzyloxybenzaldehyde 6a.57,58 Puri-
fication by column chromatography, eluting with
9.64 (br s, 1H), 9.02 (br s, 1H), 7.64 (d, 2H), 7.34 (t,
2H), 7.25 (t, 1H), 6.77 (d, 1H), 6.73 (s, 1H), 6.56 (t, 1H),
6.43 (d, 1H); 13C NMR (DMSO-d6) d 169.6 (C-2),
145.6, 144.6 (C–OH), 138.8 (C-3), 131.0 (C-4), 128.5,

V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
1655
128.2, 127.1, 119.3, 117.8, 115.8 (CH aromatic), 128.1,
122.9 (Cipso), 74.9 (C-5). Anal. calcd for C16H12O5:
C, 67.60; H, 4.25. Found: C, 67.45; H, 4.26.
128.0, 128.1, 127.9, 127.7, 127.3, 105.2 (CH aromatic),
80.0 (C-5), 73.9 (CH2), 56.1 (OCH3).
5-(3,5-Dimethoxy-4-hydroxyphenyl)-3-hydroxy-4-phenyl-
2(5H)-furanone (5k). This was synthesized like 5d in
97% yield from 5-(4-benzyloxy-3,5-dimethoxyphenyl)-
3-hydroxy-4-phenyl-2(5H)-furanone 7b (0.5 g); mp
197 ꢂC; IR (KBr) n 3342, 1747, 1672, 1620, 1573, 1449
cmꢀ1; 1H NMR (DMSO-d6) d 10.97 (br s, 1H), 8.65 (br
s, 1H), 7.65 (d, 2H), 7.38 (t, 2H), 7.29 (t, 1H), 6.67 (s,
2H), 6.45 (s, 1H), 3.73 (s, 6H); 13C NMR (DMSO-d6) d
169.1 (C-2), 148.1, 136.5 (C–OH, C–OCH3), 138.7 (C-
3), 130.8 (C-4), 128.4, 128.2, 127.4, 105.5 (CH aromatic),
127.7, 126.2 (Cipso), 80.4 (C-5), 56.1 (OCH3). Anal. calcd
for C18H16O6: C, 65.85; H, 4.91. Found: C, 66.02; H, 4.90.
5-(2,5-Dihydroxyphenyl)-3-hydroxy-4-phenyl-2(5H)-fura-
none (5h). This was prepared from methyl 2-oxo-
3-phenylpropionate and 2,5-dihydroxybenzaldehyde.
Purification by column chromatography, eluting with
a gradient of EtOAc/cyclohexane, led to compound
5h in 49% yield; mp 176 ꢂC; IR (KBr) n 3365, 1696,
1
1664, 1521, 1455 cmꢀ1; H NMR (DMSO-d6) d 10.99
(br s, 1H), 9.47 (br s, 1H), 8.78 (br s, 1H), 7.64 (d,
2H), 7.36 (t, 2H), 7.27 (t, 1H), 6.76 (d, 1H), 6.67 (s,
1H), 6.59 (d, 1H), 6.36 (s, 1H); 13C NMR (DMSO-
d6) d 169.7 (C-2), 150.1, 148.4 (C–OH), 138.8 (C-3),
130.9 (C-4), 128.6, 128.4, 127.1, 117.4, 117.0, 113.2
(CH aromatic), 128.2, 122.6 (Cipso), 74.25 (C-5). Anal.
calcd for C16H12O5: C, 67.60; H, 4.25. Found: C,
67.82; H, 4.25.
3-Hydroxy-4-(4-hydroxyphenyl)-5-phenyl-2(5H)-furanone
(5l). This was prepared from methyl 2-oxo-3-(4-hydroxy-
phenyl)propionate and benzaldehyde. Precipitation in
water and crystallization from ethyl ether gave com-
pound 5l in 21% yield; mp 258 ꢂC; IR (KBr) n 3338,
3-Hydroxy-5-(2-hydroxy-3-methoxyphenyl)-4-phenyl-
2(5H)-furanone (5i). This was prepared from methyl
2-oxo-3-phenylpropionate and 2-hydroxy-3-methoxy-
benzaldehyde. Purification by column chromatography,
eluting with a gradient of EtOAc/cyclohexane, gave
compound 5i in 39% yield; mp 202 ꢂC; IR (KBr) n 3280,
1740, 1709, 1664, 1641, 1607, 1581 cmꢀ1 1H NMR
;
(DMSO-d6) d 10.61 (br s, 1H), 9.79 (br s, 1H), 7.48 (d,
2H), 7.37–7.35 (m, 5H), 6.72 (d, 2H), 6.46 (s, 1H); 13C
NMR (DMSO-d6) d 169.4 (C-2), 157.8 (C–OH), 137.0
(C-3), 136.5, 121.5 (Cipso), 129.2, 129.1, 128.9, 127.8,
115.3 (CH aromatic), 79.6 (C-5). Anal. calcd for
C16H12O4, 1/2H2O: C, 69.31; H, 4.72. Found: C, 69.45;
H, 4.71.
1
1735, 1677, 1450, 1484 cmꢀ1; H NMR (DMSO-d6) d
10.91 (br s, 1H), 9.32 (br s, 1H), 7.61 (d, 2H), 7.31 (m,
2H), 7.24 (t, 1H), 6.92 (d, 1H), 6.74 (s, 1H), 6.69 (t, 1H),
6.55 (d, 1H), 3.80 (s, 3H); 13C NMR (DMSO-d6) d 169.5
(C-2), 147.8, 145.2 (C-OH, C–OCH3), 138.7 (C-3), 130.9
(C-4), 128.5, 128.2, 127.0, 119.2, 112.1 (CH aromatic),
128.0, 122.5 (Cipso), 74.4 (C-5), 55.8 (OCH3). Anal. calcd
for C17H14O5: C, 68.45; H, 4.73. Found: C, 68.57; H,
4.72.
3-Hydroxy-4-(4-hydroxyphenyl)-5-(4-methoxyphenyl)-
2(5H)-furanone (5m). This was prepared from methyl
2-oxo-3-(4-hydroxyphenyl)propionate and 4-methoxy-
benzaldehyde. Precipitation in water and crystallization
from ethyl ether gave compound 5m in 44% yield; mp
256 ꢂC; IR (KBr) n 3333, 1705, 1667, 1608, 1582 cmꢀ1
;
3-Hydroxy-5-(2-hydroxy-5-methoxyphenyl)-4-phenyl-
2(5H)-furanone (5j). This was prepared from methyl
2-oxo-3-phenylpropionate and 2-hydroxy-5-methoxy-
benzaldehyde. Precipitation in water and crystallization
from ethyl ether afforded compound 5j in 72% yield;
mp 126 ꢂC; IR (KBr) n 3282, 1738, 1650, 1505 cmꢀ1; 1H
NMR (DMSO-d6) d 10.94 (br s, 1H), 9.65 (br s, 1H),
7.63 (d, 2H), 7.33 (t, 2H), 7.26 (t, 1H), 6.82 (m, 2H),
6.68 (s, 1H), 6.50 (s, 1H), 3.73 (s, 3H); 13C NMR
(DMSO-d6) d 169.5 (C-2), 152.1, 149.8 (C–OH, C–
OCH3), 138.8 (C-3), 130.8 (C-4), 128.5, 128.3, 127.0,
116.8, 115.5, 113.1 (CH aromatic), 127.9, 122.8
(Cipso), 74.5 (C-5), 55.3 (OCH3). Anal. calcd for
C17H14O5, H2O: C, 64.55; H, 5.10. Found: C, 64.68;
H, 5.11.
1H NMR (DMSO-d6) d 10.58 (br s, 1H), 9.79 (br s, 1H),
7.46 (d, 2H), 7.28 (d, 2H), 6.91 (d, 2H), 6.72 (d, 2H),
6.41 (s, 1H), 3.72 (s, 3H); 13C NMR (DMSO-d6) d 169.4
(C-2), 159.8, 157.7 (C–OH, C–OCH3), 136.5 (C-3),
129.0 (C-4), 129.2, 115.4, 114.3 (CH aromatic), 128.8,
121.6 (Cipso), 79.4 (C-5), 55.1 (OCH3). Anal. calcd for
C17H14O5: C, 68.45; H, 4.73. Found: C, 68.67; H, 4.74.
Anti-oxidant assays
Measurement of the reducingactivity against the stable
radical DPPH. A solution of 4 mL of DPPH (10ꢀ4 M,
2 mg) in EtOH (50 mL) was added to 1 mL of a solution
of the test compound in DMF, to prepare a 5ꢆ10ꢀ3
M
sample solution. The absorbance at 517 nm was mea-
sured each 5 mn during 3 h. The absorbance of the
control sample without test compound was measured
simultaneously. The difference in absorbance between
the control and the test compound was taken as the
reducing activity.
5 - (4 - Benzyloxy - 3,5 - dimethoxyphenyl) - 3 - hydroxy - 4 -
phenyl-2(5H)-furanone (7b). This was prepared from
methyl 2-oxo-3-phenylpropionate and 4-benzyloxysyr-
ingaldehyde 6b.59,60 Precipitation in water and crystal-
lization from ethyl ether led to compound 7b in 52%
yield; mp 166 ꢂC; IR (KBr) n 3302, 1750, 1674, 1592,
Superoxide anion scavenging assay. The technique of
Slater and Eakins30 utilizing the interactions of NADH,
phenazine methosulfate (PMS), molecular oxygen and
nitro blue tetrazolium (NBT) was used for evaluating
superoxide anion scavenging. The NADH/PMS/O2/
NBT system involves the intermediate formation of the
1
1504 cmꢀ1; H NMR (DMSO-d6) d 11.05 (br s, 1H),
7.65 (d, 2H), 7.46–7.26 (m, 8H), 6.73 (s, 2H), 6.48 (s,
1H), 4.87 (s, 2H), 3.71 (s, 6H); 13C NMR (DMSO-d6) d
169.1 (C-2), 138.9 (C-3), 153.3, 153.4 (C–OBn, C–OCH3),
137.7, 136.9, 132.2 (Cipso), 130.7 (C-4), 128.5, 128.3,

1656
V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
superoxide anion radical (Oꢄ2ꢀ) from the interaction of
reduced PMS with O2; the superoxide anion radical
then reduced NBT to the highly colored formazan. The
reaction can be followed by measuring the absorbance
of formazan at 578 nm. The incubation mixture con-
tained disodium hydrogen phosphate buffer (200 mL, 19
mM, pH=7.4), PMS (200 mL, 10.8 mM), NBT (200 mL,
172 mM) and NADH+H+ (200 mL, 360 mM). In order to
estimate their activity, the compounds were dissolved in the
buffer at different concentrations and 200 mL of buffer
in the reaction mixture was replaced by 200 mL of these
solutions of compounds. In parallel, blanks were realized
by remplacing 200 mL of NADH+H+ by 200 mL of buffer.
Each assay was performed after 5 min incubation.
procedures.65,66 A Varian 5000 liquid chromatograph
equipped with a detector operating at 254 nm was
employed. A Varian CDS 111L integrator was used for
peak registration and calculation of retention times. An
ODP column (15 cm ꢆ 6 mm ID), prepacked with
octadecyl copolymer gel, particle size 5 mm, was used as
a non-polar stationary phase. Mobile phases were pre-
pared volumetrically from 65:35 to 95:5 combinations
of methanol/acetic acid (0.085 N, pH=3). The flow rate
was 1 mL/min. Isocratic capacity factors ki were defined
as ki=(trꢀt0)/t0 where tr is solute retention time and t0 is
column dead time, determined by using methanol as the
non-retained compound. Log k0, used as the lipophilic
index, was obtained by linear extrapolation of log ki to
50% water.
Lipid peroxidation assay. Iron-dependent peroxidation
of male rat liver microsomes was assayed as previously
described.61 Liver microsomes were prepared according
to the technique of May and McCay.62 Lipid peroxi-
dation was initiated by the ADP-Fe2+/ascorbate sys-
tem63 which produced hydroxyl radical (ꢄOH).
Incubations of 200 mL microsomal fractions (1 mg pro-
tein/mL) were carried out at 37 ꢂC with 200 mL potas-
sium phosphate buffer (0.1 M, pH=7.4), 200 mL of
ADP solution in buffer (2 mM), 200 mL of aqueous
FeSO4 solution (16 mM). Lipid peroxidation was started
by adding 200 mL aqueous ascorbate solution (0.5 mM)
to the above incubation mixture. The extent of poly-
unsaturated fatty acid peroxidation was measured spec-
trophotometrically by the rate of malondialdehyde
formed after 90 min microsomial incubation. The reac-
tion was quenched by the addition of 0.5 mL of 20%
trichloroacetic acid; after centrifugation, the super-
natant fraction was collected and then boiled with 0.5
mL of 0.67% thiobarbituric acid solution for 20 min.
After cooling, samples were diluted with 1 mL of dis-
tilled water and extinction was read at 535 nm. In order
to estimate their activity, the compounds were dissolved
in disodium hydrogen phosphate buffer (19 mM,
pH=7.4) at different concentrations and 200 mL of
buffer in the reaction mixture was replaced by 200 mL of
these solution of compounds. In parallel, blanks were
realized by replacing 200 mL of ascorbate by 200 mL of
buffer.
Molecular modeling
All computations were performed using SYBYL 6.6
software package on a Silicon Graphics SGI Power
Indigo R 8000 workstation. Structures were built within
SYBYL and minimized with the Tripos force field,
Maximin 2, in vacuo conditions, to provide reasonable
standard geometries. Molecules were deemed to be
minimized until there was a minimum energy change of
less than 5ꢆ10ꢀ3 kcal molꢀ1 for one iteration. The
BFGS method was used for minimization.
The conformational spaces of 5a–m were explored using
the SYBYL search facility. Torsion angles were defined
around the single bonds C4–CipsoAr1 and C5–CipsoAr2 of
5, and a gridsearch was performed allowing these bonds
to rotate with a 180ꢂ and a 360ꢂ revolution by 15ꢂ
increments for unsubstituted and substituted aromatic
rings, respectively. The lowest energy conformers thus
obtained were submitted to AM1 calculations
(MOPAC version 6.0)67 to optimize their geometry
and calculate physicochemical parameters: enthalpy of
formation ÁHf and energy of the highest occupied
molecular orbital (HOMO). The radical corresponding
to each lowest energy conformer was obtained by
removing the enolic or phenolic hydrogen. The calcu-
lations of open shell systems were performed using
the Unrestricted Hartree-Fock method (UHF key-
word). To avoid spin contamination and because the
UHFenergy cannot be compared with the restricted
Hartree Fock (RHF) energy of the mother structure, a
SCF(1SCFkeyword) Restricted Open Shell calculation
was performed on the UHF-optimized radical struc-
ture. The calculated enthalpies of formation were
used to calculate the enthalpies of reaction of H-
atom abstraction (ÁÁHf). Similarly, the enthalpies of
reaction of single electron transfer (ÁÁHox) were cal-
culated after geometry optimizations of radical
cations using AM1 method (Charge +1, UHFkey-
word and then 1SCFcalculation on the UHF-optimized
geometry).
Electron spin resonance measurement
ESR spectrum was recorded with a Varian (USA) E-109
spectrometer operating at 9.42 GHz with a 100 kHz
high-frequency modulation amplitude of 0.5 gauss. The
sample solution was examined in a flat quartz cell
inserted in an E-238 cavity operating in the TM110
mode. The semiquinonic radical was obtained spon-
taneously by addition of NaOH (120 mL, 2 M) to 1 mL
aerated solution of tested compound 5g in methanol (1
mM). The spectrum was simulated with the public
WinSim EPR software.64 Coupling constants and signal
intensity were adjusted using simplex method.
Pharmacological assays
Lipophilicity measurements
All tests were performed in accordance with the recom-
mendations and policies of the International Associ-
68
ation for the Study of Pain (IASP).
Lipophilicity was determined by reversed-phase high-
performance liquid chromatography using literature

V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
1657
Acute toxicity in mice. The compounds were admini-
References and Notes
stered ip as suspensions in saline (0.9% NaCl) at doses
of 200, 400, 600 and 800 mg/kg. Swiss male mice pur-
chased from Elevage Depre (Saint-Doulchard, France)
weighting 18–22 g were used. Mice were kept in groups
of 10 in a temperature controlled room with a 12 h light/
dark cycle. Food and water were available ad libitum
during the time of the experiment. The animals were
observed for 8 days in order to detect any sign of toxicity.
1. Cockrane, C. G. Am. J. Med. 1991, 91 (Suppl. 3C), 23S.
2. Gutteridge, J. M. C. Free Radic. Res. Commun. 1993, 19,
141.
3. Halliwell, B. Drugs 1991, 42, 569.
4. Bendich, A.; Machlin, L. J.; Scandura, O.; Burton, G. W.;
Wayner, D. D. M. Adv. Free Radic. Biol. Med. 1986, 2, 419.
5. Boeing, H.; Rausch, E. Sub-cell. Biochem. 1996, 25, 117.
6. Jacob, R. A.; Burri, B. J. J. Clin. Nutr. 1996, 63, 985S.
7. Kato, K.; Terao, S.; Shimamoto, N.; Hirata, M. J. Med.
Chem. 1988, 31, 793.
TPA-induced ear edema. Male Swiss mice (20–22 g)
were used (Elevage Depre, Saint-Doulchard, France).
The anti-inflammatory activity of the drugs studied was
evaluated according to the procedure of Carlson et al.52
Mice were divided into groups of 10. Phorbol 12-myri-
state 13-acetate (TPA) (Sigma-Aldrich, Saint-Quentin
Fallavier, France) was dissolved in 80% aqueous etha-
nol at concentration of 250 mg/mL. Drugs or saline were
given intraperitoneally (0.25 mL/20 g) or orally (0.50
mL/20 g). Thirty minutes after intraperitoneal drug
administration or 1 h after oral drug administration,
TPA was applied topically by an automatic pipette to
the right ear (10 mL) and left ear only received the vehi-
cle (10 mL of ethanol/water 80:20). Inflammation was
allowed to develop for 3 h 30 min after TPA. Edema
measurements were taken with an Oditest Calipers
(Oditest S 5010-Kreoplin, Gmbh, Germany). The
thickness of the right and left ears was measured in units
of 0.01 mm. Ear edema was calculated by substracting
the thickness of the left ear (vehicle control) from the
right one (treated ear).
8. Nagao, A.; Terao, J. Biochem. Biophys. Res. Commun.
1990, 172, 385.
9. Nihro, Y.; Miyataka, H.; Sudo, T.; Matsumoto, H.; Satoh,
T. J. Med. Chem. 1991, 34, 2152.
10. Nihro, Y.; Sogawa, S.; Sudo, T.; Miki, T.; Matsumoto,
H.; Satoh, T. Chem. Pharm. Bull. 1991, 39, 1731.
11. Nihro, Y.; Sogawa, S.; Izumi, A.; Sasamori, A.; Sudo, T.;
Miki, T.; Matsumoto, H.; Satoh, T. J. Med. Chem. 1992, 35,
1618.
12. Schmid, E.; Figala, V.; Roth, D.; Ullrich, V. J. Med.
Chem. 1993, 36, 4021.
13. Grisar, J. M.; Marciniak, G.; Bolkenius, F. N.; Verne-
Mismer, J.; Wagner, E. R. J. Med. Chem. 1995, 38, 2880.
14. Morisaki, K.; Ozaki, S. Chem. Pharm. Bull. 1996, 44,
1647.
15. Andersson, C. M.; Hallberg, A.; Hogberg, T. Adv. Drug.
Res. 1996, 28, 65.
16. Perricone, N.; Nagy, K.; Horvath, F.; Dajko, G.; Uray, I.;
Nagy, I. Z. Biochem. Biophys. Res. Commun. 1999, 262, 661.
17. Beifuss, U.; Kunz, O.; Aguado, G. P. Synlett 1999, 1, 147.
18. Coudert, P.; Leal, F.; Duroux, E.; Rubat, C.; Couquelet,
J. Biol. Pharm. Bull. 1996, 19, 220.
19. Weber, V.; Coudert, P.; Rubat, C.; Duroux, E.; Leal, F.;
Couquelet, J. J. Pharm. Pharmacol. 2000, 52, 523.
20. Dowling, E. J.; Symons, A. M.; Parke, D. V. Agents
Actions 1986, 19, 203.
21. Halliwell, B.; Hoult, J. R.; Blake, D. R. FASEB J. 1988, 2,
2867.
22. Namiki, T.; Baba, Y.; Suzuki, Y.; Nishikawa, M.;
Sawada, K.; Itoh, Y.; Oku, T.; Kitaura, Y.; Hashimoto, M.
Chem. Pharm. Bull. 1988, 36, 1404.
23. Ricca, J. M.; Crout, D. H. G. J. Chem. Soc., Perkin
Trans. 1 1993, 11, 1225.
24. Parisi, M. F.; Gattuso, G.; Notti, A.; Raymo, F. M.;
Abeles, R. H. J. Org. Chem. 1995, 60, 5174.
25. Cadenas, E. Biofactors 1997, 6, 391.
26. Sawai, Y.; Sakata, K. J. Agric. Food Chem. 1998, 46, 111.
27. Holton, D. M.; Murphy, D. J. Chem. Soc., Faraday Trans
1 1982, 78, 1223.
28. Halliwell, B. Am. J. Med. 1991, 91 (suppl 3C), 14S.
29. Werns, S. W.; Lucchesi, R. TIPS 1990, 11, 161.
30. Slater, T. F.; Eakins, M. N. Int. Symp. Tirrenia 1974, 84
(Karger: Basel, 1975).
Carrageenan-induced paw edema in rat. Male Sprague–
Dawley rats (120–140 g) were used (Elevage Depre,
Saint-Doulchard, France). The anti-inflammatory
activity of drugs studied was evaluated according to the
procedure of Winter et al.55 Rats were divided into
groups of six and treated intraperitoneally (0.25 mL/100
g) or orally (0.50 mL/100 g) with compounds or saline.
Thirty min after intraperitoneal drug administration or
1 h after oral drug administration, 0.05 mL of a w/v 1%
suspension of carrageenan (Sigma-Aldrich, Saint-Quen-
tin Fallavier, France) in deionized water was given by
intraplantar injection into the right paw. Animals were
returned to the cage and 3 h later, edema was measured
with a plethysmometer (V.G.O. Basile-Apelex 05.7150).
The edema means of the compound-treated groups were
compared to the vehicle to give the percentage of inhi-
bition of paw swelling.
Data analysis. Statistical analysis of the results was per-
formed using the method of Schwartz.69 Data were
analyzed by using the Student’s t-test. All values were
expressed as meanꢃSD.
31. Poupko, R.; Rosenthal, I. J. Phys. Chem. 1973, 77, 1722.
32. Burton, G. W.; Ingold, K. U. Science 1984, 224, 569.
33. Slater, T. F. Biochem. J. 1984, 222, 1.
34. Hashizume, H. Jpn. J. Conserv. Dent 1999, 36, 534.
35. Cheeseman, K. H.; Burton, G. W.; Ingold, K. U.; Slater,
T. F. Toxicol. Pathol. 1984, 12, 235.
36. Levine, M. N. Engl. J. Med. 1986, 314, 892.
37. Halliwell, B. Free Radic. Res. Commun. 1990, 9, 1.
38. Ohkawa, S.; Terao, S. S.; Terashita, Z. I.; Shibow, Y.;
Nishikawa, K. J. Med. Chem. 1991, 34, 267.
39. Yu, M. J.; McCowan, J. R.; Phebus, L. A.; Towner, R. D.;
Ho, P. P. K.; Keith, P. T.; Luttman, C. A.; Saunders, R. D.;
Ruterbories, K. J.; Lindstrom, T. D.; Wikel, J.; Morgan, E.;
Hahn, R. A. J. Med. Chem. 1993, 36, 1262.
Acknowledgements
The authors would like to thank Prof. Philip Hoggan
from Universite Blaise Pascal, Clermont-Ferrand, for
fruitful discussions, and Anne Marie Privat from
Laboratoire de Pharmacologie, Universite d’Auvergne
for technical assistance.

1658
V. Weber et al. / Bioorg. Med. Chem. 10 (2002) 1647–1658
40. Yamagami, C.; Masaki, Y. Chem. Pharm. Bull. 1995, 43,
2238.
55. Winter, C. A.; Risley, E. A.; Nuss, G. W. Proc. Soc. Exp.
Biol. Med. 1962, 111, 544.
41. Bu-Abbas, A.; Clifford, M. N.; Ioannides, C.; Walker, R.
Food. Chem. Toxicol. 1995, 33, 27.
56. Pan, M. H.; Huang, T. M.; Lin, J. K. Drug Metab. Disp.
1999, 27, 486.
42. Tripos Associates, Inc., 1699 S. Hanley, St. Louis, MI
63144, USA.
57. Li-Ming, Y.; Lin, S. J.; Yang, T. H.; Lee, K. H. Bioorg.
Med. Chem. Lett. 1996, 6, 941.
43. Macfaul, P. A.; Ingold, K. U.; Lusztyk, J. J. Org. Chem.
1996, 61, 4245.
58. Zwaagstra, M. E.; Timmerman, H.; Tamura, M.; Tohma,
T.; Wada, Y.; Onogi, K.; Zhang, M. Q. J. Med. Chem. 1997,
40, 1075.
59. Tanoguci, M.; Kashima, T.; Saika, H.; Inoue, T.; Ari-
moto, M.; Yamaguchi, H. Chem. Pharm. Bull. 1989, 37, 68.
60. Cushman, M.; Nagarathnam, D.; Gopal, D.; He, H. M.;
Lin, C. M.; Hamel, E. J. Med. Chem. 1992, 35, 2293.
44. Avila, D. V.; Ingold, K. U.; Lusztyk, J.; Green, W. H.;
Procopio, D. R. J. Am. Chem. Soc. 1995, 117, 2929.
45. Ancerewicz, J.; Migliavacca, E.; Carrupt, P. A.; Testa, B.;
Bree, F.; Zini, R.; Tillement, J. P.; Labidalle, S.; Guyot, D.;
Chauvet-Monges, A. M.; Crevat, A.; Le Ridant, A. Free
Radic. Biol. Med. 1998, 25, 113.
46. Flohe, L.; Beckman, R.; Giertz, H.; Loschen, G. Oxygen-
Centred Free Radicals as Mediators of Inflammation; Sies H.,
Ed.; Oxidative Stress; Academic: London, 1985; p 403.
47. Weiss, S. J. N. Engl. J. Med. 1989, 320, 365.
48. Udassin, R.; Ariel, I.; Haskel, Y.; Kitrossky, N.; Chevion,
M. Free Radic. Biol. Med. 1991, 10, 1.
49. Tsujimoto, Y.; Saitoh, K.; Kashima, M.; Shiozawa, A.;
Kozuka, M.; Hashizume, H.; Kimura, K.; Yamazaki, M.;
Fuju, A. Gen. Pharmacol. 1998, 31, 405.
50. Prasad, K.; Laxdal, V. A. Mol. Cell. Biochem. 1994, 136, 139.
51. Maffei Facino, R.; Carini, M.; Aldini, G.; Saibene, L.;
Morelli, R. Arzneim.-Forsch./Drug Res. 1995, 45, 1102.
52. Carlson, R. P.; O’Neill-Davis, L.; Chang, J.; Lewis, A. J.
Agents Actions 1985, 17, 197.
´
61. Meunier, M. T.; Duroux, E.; Bastide, P. Plant. Med.
´
Phytother. 1989, 23, 267.
62. May, H. E.; McCay, P. B. J. Biol. Chem. 1968, 243,
2288.
63. Muller, A.; Cadenas, E.; Graf, P.; Sies, H. Biochem.
Pharmacol. 1984, 33, 3235.
64. Public EPR Software Tools: WinSim. (online). Dec 11,
1999 (cited Jan 21, 2002); available from: URL:http://epr.-
niehs.nih.gov/download.html
65. El Tayar, N.; Van de Waterbeemd, H.; Testa, B. J. Chro-
matogr. 1985, 320, 293.
66. Bechalany, A.; Rothlisberger, T.; El Tayar, N.; Testa, B.
J. Chromatogr. 1989, 473, 115.
67. Merz, K. M., Besler, B. H. MOPAC, version 6.0 ESC;
QCPE no. 455, with AM1 force field parameters.
68. Zimmermann, M. Pain 1983, 16, 109.
53. Marks, F.; Furstenberger, G.; Kownatzki, E. Cancer Res.
1981, 41, 696.
54. Galey, C. I.; Ziboh, V. A.; Marcelo, C. I.; Voorhees, J. J.
J. Invest. Dermatol. 1985, 85, 319.
´
69. Schwartz, D. In Methodes statistiques a l’Usage des
Medecins et des Biologistes; Editions medicales; Flammarion:
`
´
Paris, 1984; p 263.