Enantiopure 3-(arylmethylidene)aminoxy-2-methylpropionic acids: Synthesis and antiinflammatory properties

Article abstract of DOI:10.1016/S0223-5234(01)01275-2

Some optically active 3-(arylmethylidene)aminoxy- (3a-g, 4a-g) and fluorenylideneaminoxy-2-methylpropionic acids (5, 6), were prepared as analogues of the antiinflammatory arylpropionic acids of type B, in which the aromatic group is substituted by an MAOM moiety. Some of the new compounds, tested in vivo for their antiinflammatory properties by means of the carrageenan-induced paw edema method in rats, exhibited activity indices similar to that shown in the same test by ibuprofen. Compounds 3a,b and 4a,b, for which at least one of the two enantiomers had shown an inhibition value higher than 40% in the in vivo test, were assayed for their in vitro enzymatic inhibitory activity, showing percentage inhibition values between 40 and 50% at a concentration of 10 μM against COX-2; at the same concentration, they appeared to be devoid of any activity towards COX-1. Compounds 3a,b and 4a,b also proved to possess a similar toxicity. The lack of enantioselectivity shown by compounds 3-6 was tentatively explained in terms of a conformational freedom of the enantiomers which allows their quasi-superimposition.

Full text of DOI:10.1016/S0223-5234(01)01275-2

Eur. J. Med. Chem. 36 (2001) 799–807
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01275-2/FLA
Original article
Enantiopure 3-(arylmethylidene)aminoxy-2-methylpropionic acids: synthesis and
antiinflammatory propertiesꢀ
Aldo Balsamoa,*, Simone Bertinib,1, Giambattista Gervasic, Annalina Lapuccia,
Susanna Nencettia, Elisabetta Orlandinia, Simona Rapposellia, Armando Rosselloa,
Giulio Soldanib
aDipartimento di Scienze Farmaceutiche, Facolta` di Farmacia, Universita` di Pisa, Via Bonanno, 6, I-56100 Pisa, Italy
bDipartimento di Clinica Veterinaria, Facolta` di Veterinaria, Universita` di Pisa, Via delle Piagge, 2, I-56100 Pisa, Italy
cLaboratori Baldacci S.p.A., Via San Michele degli Scalzi, 73, I-56100 Pisa, Italy
Received 9 February 2001; revised 18 July 2001; accepted 26 August 2001
Abstract – Some optically active 3-(arylmethylidene)aminoxy- (3a–g, 4a–g) and fluorenylideneaminoxy-2-methylpropionic acids (5,
6), were prepared as analogues of the antiinflammatory arylpropionic acids of type B, in which the aromatic group is substituted
by an MAOM moiety. Some of the new compounds, tested in vivo for their antiinflammatory properties by means of the
carrageenan-induced paw edema method in rats, exhibited activity indices similar to that shown in the same test by ibuprofen.
Compounds 3a,b and 4a,b, for which at least one of the two enantiomers had shown an inhibition value higher than 40% in the in
vivo test, were assayed for their in vitro enzymatic inhibitory activity, showing percentage inhibition values between 40 and 50% at
a concentration of 10 mM against COX-2; at the same concentration, they appeared to be devoid of any activity towards COX-1.
Compounds 3a,b and 4a,b also proved to possess a similar toxicity. The lack of enantioselectivity shown by compounds 3–6 was
tentatively explained in terms of a conformational freedom of the enantiomers which allows their quasi-superimposition. © 2001
´
Editions scientifiques et me´dicales Elsevier SAS
antiinflammatory drug / NSAID / 3-aminoxy-2-methylpropionic acid
1. Introduction
(COX) [1, 2]. This enzyme exists as two isoforms, one
constitutive (COX-1), and the other inducible (COX-
2) [2, 3]. The COX-1 isoform is constitutively ex-
pressed in most tissues, and is involved in the
physiological production of prostaglandins, which are
responsible for gastric cytoprotection. The COX-2
isoform is induced by cytokines, mitogens and endo-
toxins [4–6] in inflammatory cells, and is responsible
for the elevated production of prostaglandins during
inflammation. Most of the non-steroidal antiinflam-
matory drugs (NSAIDs) act by reducing
prostaglandin biosynthesis through the inhibition of
the cyclooxygenase (COX) reaction [7, 8]. The
NSAIDs more commonly used in the therapy of
inflammatory conditions belong to the class of ary-
lacetic (A) or arylpropionic acid (B) derivatives (figure
1) and lack selectivity towards the two types of COX
Prostaglandins are endogenous substances involved
in different processes of a physiological nature, and
are potent mediators of inflammation. Prostaglandins
are produced, together with other prostanoids, in the
arachidonic acid metabolism, whose first step, consist-
ing of the oxidative conversion of arachidonic acid
into prostaglandin H2, is catalysed by cyclooxygenase
Abbreviations: NSAID, non-steroidal antiinflammatory drug; COX,
cyclooxygenase; MAOMM, methyleneaminoxymethyl moiety.
ꢀ A part of this work was presented at the XIII Convegno
Nazionale della Divisione Chimica di Farmaceutica della Societa`
Chimica Italiana, Paestum, Italy, 23–27 September 1996.
* Correspondence and reprints
E-mail address: balsamo@farm.unipi.it (A. Balsamo).
1 Present address: Istituto di Ispezione degli Alimenti di Origine
Animale, Facolta` di Medicina Veterinaria, Universita` di Parma, Via
del Taglio, 8, I-43100 Parma, Italy.

800
A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
[9–11]. Furthermore, NSAIDs of type B usually ex-
hibit a higher activity than the corresponding type A
compounds [12, 13]; this has been attributed to the
conformational
differences
in
the
common
ArꢀCHꢀCOOH portion, due to the different steric
interactions between the remaining hydrogen (for A)
or the methyl (for B) with the adjacent aryl and the
carboxyl group [13]. Compounds B exist in two enan-
tiomeric forms, of which the S one is, in the majority
of cases, more active than the R form, probably
because it gives a better fit with the COX active site
[12, 13].
In previous papers of ours [14, 15], we have de-
scribed the synthesis and the antiinflammatory prop-
erties, evaluated by the carrageenan-induced edema
test in rats, of a series of b-aminoxypropionic acids of
type C, some of which are shown in formulas 1 and 2,
designed as analogues of arylacetic type A NSAIDs,
on the basis of the hypothesis that a methyle-
neaminoxymethyl moiety (CꢁNOCH2, MAOMM)
may act as a bioequivalent of aryl groups in this class
of drugs.
Figure 1. General structure of arylacetic (A) and arylpropionic
(B) NSAIDs and their corresponding methyleneaminoxy-
methyl (MAOM) analogues C and D.
Some of the new type C compounds, either com-
pletely aliphatic or substituted on the oximic carbon
with an aryl, have been shown to possess an apprecia-
ble antiinflammatory activity, especially in the case of
certain arylmethylidene derivatives, such as the un-
substituted (1a) and m-chloro-substituted (1b) ones
[14, 15] (figure 2).
On the basis of these results, bearing in mind the
positive effects on the pharmacological properties in-
duced by the introduction of a methyl in the a-posi-
tion with respect to the carboxylic group of the
arylacetic drugs of type A [12, 13], leading to drugs of
type B, it appeared to be of interest to study a series
of N-substituted-3-aminoxy-2-methylpropionic acids
of type D which are the a-methyl-substituted homo-
logues of compounds of type C, with the aim of
verifying whether this type of substitution may im-
prove the level of activity also for the class of the
aminoxypropionic acids. Furthermore, as the com-
pounds of type D can be viewed as MAOM-substi-
tuted analogues of arylpropionic compounds of type
B in which the type of chirality influences the trend of
activity, we planned the preparation of both enan-
tiomers of compounds of type D through an enan-
tioselective synthesis starting from commercially
available chiral compounds (figure 1).
Figure 2. Structures of the previously described arylmethyli-
dene (1) and fluorenylidene (2) derivatives and of the corre-
sponding enantiopure a-methyl-substituted analogues 3, 4 and
5, 6, showing their absolute configuration.
Consequently,
the
arylmethylideneaminoxy-2-
methylpropionic acids with the R (3a–g) and the S

A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
801
Figure 3. Synthesis of methylpropionic MAOM derivatives 3–6.
(4a–g) (Fig. 2) configurations were prepared, in
which the substituents on the phenyl ring correspond
to those present in some of the corresponding type C
compounds which in previous studies have been
shown to possess different antiinflammatory activity
levels [14, 15]. In addition, the fluorenylidene com-
pounds 5 and 6 were prepared as the only oximethers
of a ketone, whose corresponding analogue (2)
proved to be completely inactive [14].
[16] of the appropriate optically active methyl-2-hy-
droxymethyl propionate 7 and 8 with the R or S
configuration, respectively, in the presence of N-hy-
droxyphthalimide,
diethylazodicarboxylate
and
triphenylphosphine in THF, yielded the correspond-
ing O-phthalimido derivatives 9 and 10, which by
hydrazinolysis with NH2NH2·H2O afforded the hy-
droxylaminethers 11 and 12. The reaction of 11 and
12 with the appropriate aldehydes 13a–g, or
fluorenone 14 in MeOH, gave the corresponding
oximethers 15–19 which, in the cases in which an E/Z
isomerism is possible, exclusively possessed the E
configuration, with the only exception of the benzyli-
dene derivative unsubstituted on the phenyl ring,
which gave an 85:15 mixture of E 16a and Z 17a
2. Chemistry
The 3-aminoxy-2-methylpropionic acids 3–6, were
prepared as shown in figure 3. Mitsunobu reaction

802
A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
isomers. The oximethers 15, 16, 18 and 19 were
purified by column chromatography, and were then
saponified with 1 N NaOH in THF–H2O to yield the
corresponding optically active arylmethylidene acids 3
and 4, and fluorenylidene acids 5 and 6. The acids
were purified by crystallisation from the appropriate
solvents (for 3a–c,f,g, 4a–c, 4f,g, 5 and 6) or by
bulb-to-bulb distillation (for 3d,e and 4d,e).
The assignment of the type of configuration around
the CꢁN double bond of the benzylidene-substituted
esters with the S configuration (E 16a and Z 17a) was
made unequivocally by means of 1H-NMR spec-
trometry, seeing that in the oximethers with the E
configuration, the oximic proton resonates at lower
fields than the same hydrogen of the Z isomer.
Once the E configuration had been assigned to
oximether 16a, it was possible to attribute the same
type of configuration also to esters 15a–g and 16b–g
and to the corresponding acids 3 and 4, on the basis
of the observation that their proton linked to the
oximic carbon resonates at chemical shift values prac-
tically identical or very close to those of the same
proton of ester 16a.
paw edema test in rats. The results obtained for 3–6
and for ibuprofen, taken as a reference drug, are
shown in table I, together with those previously ob-
tained in the same type of test with the corresponding
arylmethylideneaminoxypropionic acids of type C,
1a–g and 2. All compounds (3–6 and ibuprofen) were
administered intraperitoneally at a dose of 50 mg
kg−1. The activity was expressed as percentage inhibi-
tion values.
Among the compounds with the R configuration
(3a–g, 5), those unsubstituted on the phenyl ring (3a)
or m-halosubstituted (3b and 3c) exhibited the highest
activity, with percentage inhibition values ranging
between 52% of 3a and 43% of 3c. A slightly lower
activity was shown by the fluorenyl derivative 5, while
the other type 3 compounds (3d–g) appeared to pos-
sess a lower degree of activity (percentage inhibition
values <20%).
As regards the compounds with the S configuration
(4a–g, 6), the m-fluorosubstituted compound 4c and
the fluorenyl derivative 6 showed the highest activity,
with percentage inhibition values close to 40%; the
other type 4 compounds, with the exception of 4e
which was found to be practically inactive, exhibited
inhibition indices ranging from 28 to 22%.
As regards the configurational situation of esters 15
and 16 and of the corresponding acids 3 and 4 at the
level of their chiral carbon, the values of the specific
rotatory power in the couples of enantiomers, practi-
cally equal in terms of their absolute value, but with
the opposite sign, indicate that the single enantiomers
might possess an analogous optical purity.
For the compounds unsubstituted (3a, 4a) or m-
chlorosubstituted (3b and 4b) on the phenyl ring, for
which at least one of the two enantiomers had shown
an inhibition value higher than 40% in the in vivo
test, also the inhibition properties towards COX-1
and COX-2 were evaluated. The results obtained in
vitro by measuring the prostaglandin (PGE2) produc-
tion in U937 cell lines for COX-1 and activated J774.2
macrophages for COX-2, at a concentration of 10 and
1 mM are reported in table II; this table reports the
results obtained in the same test for celecoxib, one of
the most selective NSAIDs recently introduced into
clinical use, taken as a reference drug for this in vitro
test. The four compounds 3a,b and 4a,b exhibited a
similar trend of inhibitory activity; in particular, these
compounds showed percentage inhibition values be-
tween 40 and 50% at a concentration of 10 mM
towards COX-2, while at the same concentration,
they appeared to be completely devoid of any activity
towards COX-1. In the same experiment, celecoxib,
tested at a concentration of 0.1 mM, exhibited a
percentage inhibition value of about 50% towards
COX-2 and a practically complete inactivity towards
COX-1.
1
An H-NMR analysis carried out on the couple of
m-fluoro-substituted acids, 3c and 4c, using R(−)-cin-
conidine as the chiral resolving agent showed that
both enantiomers possess an enantiomeric purity
>96%. Assuming that the type of substituent on the
phenyl ring of the remaining couples of acids does not
affect the chiral stability of these compounds, an
analogous optical purity was assigned also to the
remaining acids of types 3 and 4, and therefore also to
the corresponding esters, 15 and 16.
As regards the absolute configuration of the inter-
mediate and final products, this derives from a knowl-
edge of the structure of the starting hydroxymethyl-
propionic methylesters 7 and 8.
3. Biopharmacological results and discussion
Compounds 3–6 were assayed for their antiinflam-
matory activity by means of the carrageenan-induced
In the case of compounds 3a,b and 4a,b, the ap-
proximate LD50 was also evaluated in the mouse by

A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
803
A comparison of the data obtained for the 2-
methylpropionic acids 3–6 with those found for the
analogues lacking the 2-methyl group (1 and 2) shows
that in the case of the compounds unsubstituted,
m-chloro- or m-fluoro-substituted on the phenyl ring,
the presence of the methyl group does not appear to
play an important role, with the exception of the S
stereoisomers 4a and 4b, for which the activity de-
creased with respect to 1a and 1b. For the p-EtO-sub-
stituted compounds (3e and 4e), the same methyl
group leads to a marked decrease in the activity of
both stereoisomers, compared with the corresponding
type 1 analogues, while for the m-MeO- (3d and 4d)
and p-CF3-substituted (3g and 4g) compounds, the
methyl does not contribute to any substantial modifi-
i.v. administration. The values obtained ranged from
697 mg kg−1 of 3b to 795 mg kg−1 of 3a, thus
indicating a similar toxicity.
Results obtained for compounds with the R
configuration (3a–f and 5), and for compounds with
the S configuration (4a–f and 6) in the in vivo test,
shows that for these compounds in which the R
substituent is hydrogen or m-chloro, the R isomer
exhibits a better activity, while for the m-fluoro-sub-
stituted compounds, or for the 9-fluorenyl (5 and 6)
derivatives, the activity shown by the two isomers is
practically the same. Instead, for the compounds in
which R is m-MeO, p-Me or p-CF3, the isomers with
the S configuration appear to be slightly more active
than the R ones.
Table I. Chemical and pharmacological data of compounds 1–6.
Compound
R
[h]2D2 (c)
m.p. or b.p.
[mmHg] (°C)
Recrystallisation
solvent a
Inhibition activity on carrageenan
paw edema (50 mg kg−1 i.p.) b
inhibition at 4 h
%
(R)-3a c
(S)-4a c
1a d
(R)-3b
(S)-4b
1b e
(R)-3c
(S)-4c
1c e
(R)-3d
(S)-4d
1d e
(R)-3e
(S)-4e
1e e
(R)-3f
(S)-4f
1f e
(R)-3g
(S)-4g
1g e
H
H
H
m-Cl
m-Cl
m-Cl
m-F
m-F
m-F
m-MeO
m-MeO
m-MeO
p-EtO
p-EtO
p-EtO
p-Me
p-Me
p-Me
p-CF3
p-CF3
p-CF3
−53.7 (1.38)
+53.6 (1.38)
67–69
67–69
A
A
52** (10)
28 (10)
66* (10)
46** (10)
22 (10)
53* (10)
43** (10)
40** (10)
39* (10)
16* (10)
26 (10)
10 (10)
16 (10)
9 (10)
52* (10)
16 (10)
−45.6 (0.99)
+45.2 (1.01)
64–66
63–65
A
A
−48.9 (1.00)
+48.7 (1.01)
57–59
57–59
A
A
+44.8 (1.14)
−43.3 (1.14)
250 (0.3)
250 (0.3)
−42.9 (1.13)
+43.3 (1.53)
230 (0.3)
230 (0.3)
−53.7 (1.18)
+52.2 (1.14)
80–82
75–78
A
A
28** (10)
0 (10)
12 (10)
−39.5 (1.10)
+40.6 (2.11)
32–34
33–35
B
B
22 (10)
15* (10)
37** (10)
37** (10)
0 (10)
(R)-5
(S)-6
2
−29.8 (1.09)
+30.1 (1.41)
102–104
103–105
A
A
(R,S)-Ibuprofen
42** (10)
Results marked with an asterisk or two asterisks are percentage reductions which are significant (*, PB0.05; **, PB0.01,
respectively). The number of animals used per groups is reported in parentheses.
a A, hexane; B, AcOEt–hexane.
b A Student’s t-test was carried out.
c See Ref. [17].
d See Ref. [14].
e See Ref. [15].

804
A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
Table II. In vitro inhibitory activity of 3a,b and 4a,b towards
group gives rise to a lower degree of steric interac-
tions with the adjacent groups compared with those
found in compounds of type B, and should thus exert
less influence on their conformational situation, and
consequently on their ability to interact more effec-
tively with the enzymic active site.
COX-1 and COX-2.
Compound
R
Cyclooxygenase inhibitory activity
(% inhibition) a
Dose (mM)
COX-1 COX-2
As regards the influence of the chirality on the
activity, the pharmacological data do not show a
homogeneous trend, seeing that the R enantiomers
may appear to be more active (as in the case of
compounds 3a–c), or less active (as in the case of
3d,f,g) than the corresponding S enantiomers (4a–
d,f,g), or they may possess a very similar activity (as
in the case of the remaining compounds), depending
on the type of substituent on the oximic carbon.
The apparent absence of enantioselectivity of com-
pounds 3–6 may be tentatively attributed to the high
conformational freedom of the 3-methyleneaminoxy-
2-methylpropionyl chain, which makes it possible to
find low-energy conformations which allow a quasi-
superimposition of the enantiomers. This kind of ex-
planation has also been advanced to justify the lack
of selectivity in the field of MAOM-type analogues of
adrenergic b-blocking drugs [18, 19], which, however,
possess a propanolaminic structure instead of the
methylpropionic one of compounds 3–6.
(R)-3a
(S)-4a
H
10
1
0
0
39
23
H
10
1
14
0
41
30
(R)-3b
(S)-4b
m-Cl
m-Cl
10
1
0
0
50
24
10
1
0
0
42
5
Celecoxib
0.1
12
52
a S.E.M. was lower than 10% in all experiments.
cation of the antiinflammatory activity. Only in the
case of the p-Me-substituted (3f and 4f) compounds
and the fluorenyl derivatives (5, 6) for which the
corresponding analogues lacking the methyl group (1f
and 2) are inactive, did the presence of this sub-
stituent in the 2-position with respect to the carboxyl
group, determine the appearance of an antiinflamma-
tory activity, especially in the case of both the
fluorenyl enantiomers 5 and 6.
As far as the in vitro activity of the compounds
tested (3a,b and 4a,b) is concerned, this appeared to
be relatively modest, and directed exclusively towards
COX-2; moreover, for the same compounds, this ac-
tivity proved to be practically independent of their
configuration and of the substituent (hydrogen or
chlorine), on the aromatic ring.
5. Experimental protocols
5.1. Chemistry
Melting points were determined on a Kofler hot-stage
apparatus and are uncorrected. Boiling points refer to
the air bath temperature of bulb-to-bulb distillation
using a Bu¨chi GKR-51 apparatus and are uncorrected.
IR spectra for comparison of compounds were taken as
paraffin oil mulls or as liquid films on a Unicam
4. Conclusions
1
Mattson 1000 FT-IR spectrometer. H-NMR spectra of
An analysis of the pharmacological data reveals
that, contrary to what might be expected on the basis
of the positive effect of the introduction of a methyl
group on the carbon adjacent to the carboxyl of
compounds A leading to compounds B, in the major-
ity of cases the same type of substitution in
aminoxypropionic acids of type C does not affect the
antiinflammatory activity of the corresponding
methylpropionic analogues of type D. A possible ex-
planation of these results may be sought in the fact
that in compounds of type D, the additional methyl
all compounds were obtained with a Varian CFT 20
instrument or a Bruker AC-200 spectrometer operating
at 80 or 200 MHz, respectively, in ca. 2% solution of
CDCl3. Analytical TLCs were carried out on 0.25 mm
layer silica gel plates containing a fluorescent indicator;
spots were detected under UV light (254 nm). Column
chromatography was performed using 70–230 mesh sil-
ica gel. GLC analyses were performed on a Perkin–
Elmer model 8410 apparatus with a flame ionisation
detector, using a 1.5 m×2.5 mm column packed with
methylsilicone OV-1 3% on Supelcoport 80–100 mesh.

A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
805
Evaporation was performed in vacuo (rotating evapora-
tor). Na2SO4 was always used as the drying agent.
Elemental analyses were performed in our analytical
laboratory and agreed with the theoretical values to
within 0.4%.
An analytical sample of (R)-11 or (S)-12 as a free
base was obtained by dissolving 50 mg of the corre-
sponding oxalate salt in H2O (2 mL), adding solid
K2CO3 (100 mg), extracting with CHCl3, and then evap-
1
orating the filtered organic extracts: H-NMR l 1.17 (d,
3H, J=7.0 Hz, CHCH3), 2.90 (m, 1H, CHCH3), 3.75
5.1.1. Synthesis of (R)-9 and (S)-10 methyl
ppm (s, 3H, CO2CH3).
2-methyl-3-(2-phthalimidoxy)propionates
A solution of diethylazodicarboxylate (15 mL, 93
mmol) in anhydrous THF (20 mL) was added dropwise
to a stirred solution of the commercially available (R)-7
or (S)-8 methyl 2-methyl-3-hydroxypropionate (10 g, 85
mmol), triphenylphosphine (22.4 g, 85 mmol) and N-hy-
droxyphthalimide (85 mmol). After 24 h at room tem-
perature (r.t.) the solvent was evaporated to yield an
oily residue which was triturated with Et2O. The solid
precipitate was filtered and the Et2O solution was evap-
orated to give an oil which was chromatographed on a
silica gel column eluting with a 9:1 C6H6–Et2O mixture.
Evaporation of the fractions containing the main
product afforded practically pure 9 or 10, which was
5.1.3. General procedure for the synthesis of
(E,R)-15a–g, (E,S)-16a–g, (Z,S)-17a, methyl
3-aryl-methylideneaminoxy-2-methylpropionates and
(R)-18 and (S)-19 methyl fluorenylideneaminoxy-2-
methylpropionates
A solution of the appropriate aromatic aldehyde 13a–
g or fluorenone 14 (26 mmol) and of the aminoxyester
(R)-11 or (S)-12 (3.06 g, 26 mmol) in MeOH (60 mL)
was stirred at r.t. until the disappearance of the starting
carbonyl compound (TLC). The solvent was then evap-
orated and the residue was taken up with CHCl3 and
washed with 10% H3PO4 and H2O. Evaporation of the
filtered organic layers afforded an oily residue which
was purified by chromatography on a silica gel column
eluting with a 7:1:2 toluene–hexane–CH2Cl2 mixture.
Evaporation of the fractions containing the faster mov-
ing significant compound yielded the corresponding
practically pure (GLC) aldoximethers (E,R)-15a–g or
(E,S)-16a–g or fluorenonoximethers (R)-18 or (S)-19.
Only in the case of the chromatography of the crude
reaction mixture of aminoxyester (R)-11 with benzalde-
hyde 13a, was a second set of fractions obtained yielding
a small quantity of practically pure (Z,S) oximether 17a.
(E,R)-15a (63%); [h]D22= −49.1 (CHCl3, c=1.16). (E,R)-
15b (41%); [h]2D2= −51.8 (CHCl3, c=1.0). (E,R)-15c
(55%); [h]2D2= −49.4 (CHCl3, c=1.10). (E,R)-15d (46%);
[h]2D2= −47.1 (CHCl3, c=1.61). (E,R)-15e (49%);
[h]2D2= −46.9 (CHCl3, c=1.13). (E,R)-15f (51%);
[h]2D2= −51.8 (CHCl3, c=1.30). (E,R)-15g (48%);
[h]2D2= −41.5 (CHCl3, c=2.06). (E,S)-16a (58%):
[h]2D2= +51.4 (CHCl3, c=1.05). (E,S)-16b (54%);
[h]2D2= +50.8 (CHCl3, c=0.99). (E,S)-16c (50%);
[h]2D2= +49.9 (CHCl3, c=1.84). (E,S)-16d (44%);
[h]2D2= +46.6 (CHCl3, c=1.16). (E,S)-16e (43%);
[h]2D2= +47.4 (CHCl3, c=1.13). (E,S)-16f (56%);
[h]2D2= +52.0 (CHCl3, c=1.09). (E,S)-16g (34%);
[h]2D2= +42.0 (CHCl3, c=1.63). (Z,S)-17a: IR (Nujol) w
1749 cm−1 (CꢁO); [h]D22= +51.4 (CHCl3, c=1.05). (R)-
18 (23%); [h]2D2= −19.8 (CHCl3, c=1.09). (S)-19 (20%);
[h]2D2= +20.3 (CHCl3, c=1.10). IR (Nujol) and 1H-
NMR. (E,R)-15a and (E,S)-16a: w 1741 cm−1 (CꢁO); l
1.20 (d, 3H, J=7.0 Hz, CHCH3), 2.97 (m, 1H,
1
crystallised from Et2O–hexane: H-NMR l 1.36 (d, 3H,
J=7.0 Hz, CHCH3), 3.03 (m, 1H, CHCH3), 3.75 (s,
3H, CO2CH3), 4.41 ppm (m, 2H, OCH2). 9 (44%): m.p.
44–46 °C; IR (Nujol) w 1741 cm−1 (CꢁO); [h]D22= −9.86
(CHCl3, c=1.42). C13H13NO5 (C, H, N). 10 (48%): m.p.
44–46 °C; IR (Nujol) w 1733 cm−1 (CꢁO); [h]D22= +10.0
(CHCl3, c=1.03). C13H13NO5 (C, H, N).
5.1.2. Preparation of (R)-11 and (S)-12 methyl
2-methyl-3-aminoxypropionate
A solution of 98% H2NNH2·H2O (19 mmol) in anhy-
drous EtOH (5 mL) was added to a solution of the
phthalimido derivatives 9 or 10 (5.0 g, 19 mmol) in
anhydrous EtOH (35 mL). After 1 h at the reflux
temperature, the mixture was cooled at r.t. and then
filtered. The organic solution was evaporated and the
crude residue was dissolved in CHCl3, washed (3% aq.
K2CO3 and H2O), filtered and evaporated to dryness to
yield an oily residue consisting almost exclusively in the
aminoxyester 11 or 12. The crude product was dissolved
in anhydrous Et2O and the resulting solution was added
to a 7:3 Et2O–MeOH solution of oxalic acid. The solid
precipitate was collected by filtration and then crys-
tallised from EtOH to give the pure oxalate salt of 11 or
1
12: H-NMR (D2O) l 1.20 (d, 3H, J=7.0 Hz, CHCH3),
2.99 (m, 1H, CHCH3), 3.75 (s, 3H, CO2CH3), 4.20 ppm
(m, 2H, OCH2); (R)-11·H2C2O4 (58%): m.p. 92–94 °C;
C7H13NO7 (C, H, N). (S)-12·H2C2O4 (64%): m.p. 92–
94 °C; C7H13NO7 (C, H, N).

806
A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
CHCH3), 3.63 (s, 3H, CO2CH3), 4.22 (m, 2H, OCH2),
8.00 ppm (s, 1H, CHꢁN); (E,R)-15b and (E,S)-16b w
1741 cm−1 (CꢁO);l 1.23 (d, 3H, J=7.0 Hz, CHCH3),
2,98 (m, 1H, CHCH3), 3.70 (s, 3H, CO2CH3), 4.30 (m,
2H, OCH2), 8.10 ppm (s, 1H, CHꢁN); (E,R)-15c and
(E,S)-16c: w 1741 cm−1 (CꢁO); l 1.23 (d, 3H, J=7.0 Hz,
CHCH3), 3.08 (m, 1H, CHCH3), 3.70 (s, 3H, CO2CH3),
4.30 (m, 2H, OCH2), 8.10 ppm (s, 1H, CHꢁN); (E,R)-
15d and (E,S)-16d: w 1741 cm−1 (CꢁO); l 1.20 (d, 3H,
J=7.0 Hz, CHCH3), 2.88 (m, 1H, CHCH3), 3.67 (s,
3H, CO2CH3), 3.80 (s, 3H, OCH3), 4.23 (m, 2H, OCH2),
7.98 ppm (s, 1H, CHꢁN); (E,R)-15e and (E,S)-16e: w
1741 cm−1 (CꢁO); l 1.20 (d, 3H, J=7.0 Hz, CHCH3),
1.40 (t, 3H, J=7.0 Hz, OCH2CH3), 2.87 (m, 1H,
CHCH3), 3.67 (s, 3H, CO2CH3), 4.0 (q, 2H, J=7.0 Hz,
OCH2CH3), 4.20 (m, 2H, OCH2), 7.95 ppm (s, 1H,
CHꢁN); (E,R)-15f and (E,S)-16f: w 1741 cm−1 (CꢁO); l
1.25 (d, 3H, J=7.0 Hz, CHCH3), 2.40 (s, 3H, CH3),
2.96 (m, 1H, CHCH3), 3.76 (s, 3H, CO2CH3), 4.30 (m,
2H, OCH2), 8.12 ppm (s, 1H, CHꢁN); (E,R)-15g and
(E,S)-16g: w 1741 cm−1 (CꢁO); l 1.26 (d, 3H, J=7.0
Hz, CHCH3), 2.96 (m, 1H, CHCH3), 3.73 (s, 3H,
CO2CH3), 4.45 (m, 2H, OCH2), 8.13 ppm (s, 1H,
CHꢁN); (Z,S)-17a: l 1.20 (d, 3H, J=7.0 Hz, CHCH3),
3.03 (m, 1H, CHCH3), 3.70 (s, 3H, CO2CH3), 4.37 (m,
2H, OCH2), 7.93 ppm (s, 1H, CHꢁN); (R)-18 and
(S)-19: w 1741 cm−1 (CꢁO); l 1.28 (d, 3H, J=7.0 Hz,
CHCH3), 3.08 (m, 1H, CHCH3), 3.72 (s, 3H, CO2CH3),
4.57 ppm (m, 2H, OCH2).
C12H15NO3 (C, H, N). (E,R)-3g (44%); C12H12F3NO3
(C, H, N). (E,S)-4a (43%); C11H13NO3 (C, H, N).
(E,S)-4b (32%); C11H12ClNO3 (C, H, N). (E,S)-4c
(36%); C11H12FNO3 (C, H, N). (E,S)-4d (21%);
C12H15NO4 (C, H, N). (E,S)-4e (40%); C13H17NO4 (C,
H, N). (E,S)-4f (46%); C12H15NO3 (C, H, N). (E,S)-4g
(25%); C12H12F3NO3 (C, H, N). (R)-5 (20%);
C17H15NO3 (C, H, N). (S)-6 (57%); C17H15NO3 (C, H,
1
N). IR (Nujol) and H-NMR. (E,R)-3a and (E,S)-4a: w
1703 cm−1 (CꢁO); l 1.28 (d, 3H, J=7.0 Hz, CHCH3),
3.10 (m, 1H, CHCH3), 4.30 (m, 2H, OCH2), 8.15 ppm
(s, 1H, CHꢁN); (E,R)-3b and (E,S)-4b: w 1703 cm−1
(CꢁO); l 1.27 (d, 3H, J=7.0 Hz, CHCH3), 2.68 (m, 1H,
CHCH3), 4.42 (m, 2H, OCH2), 8.07 ppm (s, 1H,
CHꢁN); (E,R)-3c and (E,S)-4c: w 1703 cm−1 (CꢁO); l
1.32 (d, 3H, J=7.0 Hz, CHCH3), 3.03 (m, 1H,
CHCH3), 4.40 (m, 2H, OCH2), 8.03 ppm (s, 1H,
CHꢁN); (E,R)-3d and (E,S)-4d w 1718 cm−1 (CꢁO)
(liquid film); l 1.27 (d, 3H, J=7.0 Hz, CHCH3), 2,95
(m, 1H, CHCH3), 3.82 (s, 3H, OCH3), 4.33 (m, 2H,
OCH2), 8.03 ppm (s, 1H, CHꢁN); (E,R)-3e and (E,S)-
4e: w 1718 cm−1 (CꢁO) (liquid film); l 1.27 (d, 3H,
J=7.0 Hz, CHCH3), 1.38 (t, 3H, J=7.0 Hz,
OCH2CH3), 3.06 (m, 1H, CHCH3), 4.0 (q, 2H, J=7.0
Hz, OCH2CH3), 4.28 (m, 2H, OCH2), 7.99 ppm (s, 1H,
CHꢁN); (E,R)-3f and (E,S)-4f: w 1726 cm−1 (CꢁO); l
1.27 (d, 3H, J=7.0 Hz, CHCH3), 2.38 (s, 3H, CH3),
2.96 (m, 1H, CHCH3), 4.33 (m, 2H, OCH2), 8.10 ppm
(s, 1H, CHꢁN); (E,R)-3g and (E,S)-4g: w 1719 cm−1
(CꢁO); l 1.27 (d, 3H, J=7.0 Hz, CHCH3), 2.98 (m, 1H,
CHCH3), 4.33 (m, 2H, OCH2), 8.17 ppm (s, 1H,
CHꢁN); (R)-5 and (S)-6: w 1719 cm−1 (CꢁO); l 1.37 (d,
3H, J=7.0 Hz, CHCH3), 3.17 (m, 1H, CHCH3), 4.55
ppm (m, 2H, OCH2).
5.1.4. General procedure for the synthesis of
(E,R)-3a–g, (E,S)-4a–g, 3-arylmethylideneaminoxy-2-
methylpropionic acids and (R)-5 and (S)-6
fluorenylideneaminoxy-2-methylpropionic acids
A solution of 1 N NaOH (5.5 mL) was added drop-
wise to a cooled (0 °C) solution of the appropriate
methylester (15a–g, 16a–g, 18 or 19) (5.0 mmol) in
THF (25 mL) and the resulting solution was stirred at
the same temperature until the disappearance of the
starting ester (TLC). The solution was concentrated and
the aqueous residue was then washed with Et2O, aci-
dified to pH 3 with 10% aq. HCl and extracted with
CHCl3. Evaporation of the filtered organic layer af-
forded a residue which was crystallised from the appro-
priate solvent, in the case of the solids, or purified by
distillation at reduced pressure in the case of the liquids.
(E,R)-3a (27%); C11H13NO3 (C, H, N). (E,R)-3b (34%);
C11H12ClNO3 (C, H, N). (E,R)-3c (55%); C11H12FNO3
(C, H, N). (E,R)-3d (36%); C12H15NO4 (C, H, N).
(E,R)-3e (46%); C13H17NO4 (C, H, N). (E,R)-3f (51%);
An equimolar mixture of racemic m-F-substituted-3-
benzylideneaminoxy-2-methyl-propionic acids 3c and 4c,
1
with (R)-cinconidine in CDCl3 displayed in its H-NMR
spectrum two doublets at 1.24 and 1.22 ppm at-
tributable to 2-methyl protons of E,R-(3c) and E,S-(4c)
stereoisomers, respectively.
5.2. Biopharmacological methods
5.2.1. Carrageenan-induced paw edema
The procedure was the same already followed in
previous papers [14, 15]. Ten or 20 female Wistar rats
(200–250 g body weight) were used for each group.
Hind paw volumes were measured using a water plethys-
mometer (Basile) in accordance with the method de-

A. Balsamo et al. / European Journal of Medicinal Chemistry 36 (2001) 799–807
807
scribed by Winter et al. [20]. Then the compounds to be
tested were administered i.p. at a constant dose of 50 mg
kg−1, using a solution in aq. NaHCO3 (pH 7.4) at a
concentration of 10 mg mL−1. Control rats received the
same volume of the vehicle. Thirty min later, 0.05 mL of
a 1% solution of carrageenan (Sigma) was subcuta-
neously injected into the plantar surface of the right
hind paw. The increase in paw volume 4 h after the
injection of the phlogistic agent was adopted as a mea-
sure of the effect, and results were expressed as a
percentage reduction of control edema. Student’s t-test
for grouped data was used for statistical evaluation.
Differences were considered to be statistically significant
when P was 50.05.
dell’Universita` e della Ricerca Scientifica e Tecnologica.
We would also like to thank Dr C. Milanese for the
enzyme assays.
References
[1] Bayly C.I., Black W.C., Leger S., Ouimet N., Ouellet M.,
Percival M.D., Bioorg. Med. Chem. Lett. 9 (1999) 307–312.
[2] Kurumbail R.G., Stevens A.M., Gierse J.K., McDonald J.J.,
Stegerman R.A., Pak J.Y., Gildehaus D., Miyashiro J.M.,
Penning T.D., Seibert K., Isakson P.C., Stallings W.C., Nature
384 (1996) 644–648.
[3] Bakhle Y.S., Drugs Today 35 (1999) 237–250.
[4] Kuyubu D.A., Fletcher B.S., Varnum B.C., Lim R.W., Her-
schman H.R., J. Biol. Chem. 266 (1991) 12866–12872.
[5] Lee S.H., Soyoola E., Chanmugam P., Hart S., Sun W., Zhong
H., Liou S., Simmons D., Hwang D., J. Biol. Chem. 267 (1992)
25934–25938.
5.2.2. Approximate LD50
The compounds to be tested (3a,b and 4a,b) were
administered i.v. in aq. NaHCO3 solution (pH 7.5) at
increasing doses, to male albino Swiss mice from Nos-
san Farm. Approximate LD50 values were determined in
accordance with the method of Litchfield et al. [21].
[6] O’Sullivan M.G., Huggins E.M. Jr., Mccall C.E., Biochem.
Biophys. Res. Commun. 191 (1993) 1294–1300.
[7] Vane J.R., Nature 231 (1971) 232–235.
[8] Lombardino G., Nonsteroidal Antiinflammatory Drugs, John
Wiley and Sons Eds.; Wiley Interscience, New York, 1985.
[9] Mantri P., Witiak D.T., Curr. Med. Chem. 1 (1994) 328–355.
5.2.3. Enzyme assays
[10] Chan C.-C., Boyce S., Brideau C., Ford-Hutchinson A.W.,
Gordon R., Guay D., J. Pharm. Exp. Ther. 274 (1995) 1531–
1537.
Compounds 3a,b and 4a,b were tested in intact cell
assays to verify their capacity to inhibit PGE2 produc-
tion, considered as an index of activity on COX-1 and
COX-2 enzymes. For the COX-1 assay, 1.5×106 resting
U937 human cells were incubated with the test com-
pounds for 30 min in the presence of 10 mM arachidonic
acid. Tubes were then centrifuged and the PGE2 content
in the supernatant was measured by a commercial im-
munoenzymatic assay (Amersham). The COX-2 assay
was performed in accordance with the method described
by Mitchell et al. [22] with minor modifications, as
suggested by Grossman et al. [23]. Murine J774.2 cells
were pretreated for 1 h with 300 mM aspirin to inacti-
vate endogenous constitutive COX-1 and were then
stimulated with LPS to induce COX-2 expression. After
overnight incubation, cells were treated for 45 min with
the different test compounds. Supernatants were then
collected and PGE2 was measured as described above.
All compounds were tested in duplicate. For each
product, a stock solution was prepared in DMSO at a
concentration of 100 mM.
[11] Griswold D.E., Adams J.L., Med. Res. Rev. 16 (1996) 181–
206.
[12] Juby P.F., in: Scherrer R.A., Whitehouse M.W. (Eds.), Medici-
nal Chemistry. A series of monographs, Academic Press, New
York, 1974, pp. 92–127 chapter 4th.
[13] Shen T.Y., in: Wolff M. (Ed.), Burger’s Medicinal Chemistry,
Part III, 4th ed., John Wiley and Sons, New York, 1981, pp.
1212–1271.
[14] Macchia B., Balsamo A., Lapucci A., Macchia F., Martinelli
A., Nencetti S., Orlandini E., Baldacci M., Mengozzi G.,
Soldani G., Domiano P., J. Med. Chem. 33 (1990) 1423–1430.
[15] Lapucci A., Macchia M., Martinelli A., Nencetti S., Orlandini
E., Rossello A., Baldacci M., Soldani G., Mengozzi G., Eur. J.
Med. Chem. 29 (1994) 33–39.
[16] Mitsunobu O., Synthesis (1981) 1–28.
[17] Macchia B., Eur. Pat. Appl. (1986) 175304.
[18] Baldwin J.J., McClure D.E., Gross D.M., Williams M., J. Med.
Chem. 25 (1982) 931–936.
[19] Bouzoubaa M., Leclerc G., Rakhit S., Andermann G., J. Med.
Chem. 28 (1985) 896–900.
[20] Winter C.A., Risley E.A., Nuss G.W., Proc. Soc. Exp. Biol.
Med. 111 (1962) 544–547.
[21] Litchfield J.T., Wilcoxon F., J. Pharm. Exp. Ther. 96 (1949)
99–113.
[22] Mitchell J.A., Akarasereenont P., Thiemermann C., Flower
R.J., Vane J.R., Proc. Natl. Acad. Sci. USA 90 (1994) 11693–
11697.
Acknowledgements
This work was supported in part by a grant from the
Consiglio Nazionale delle Ricerche and the Ministero
[23] Grossman C.J., Wiseman J., Lucas F.S., Trevethick M.A.,
Birch P.J., Inflamm. Res. 44 (1995) 253–257.