Synthesis, pharmacological characterization, and docking analysis of a novel family of diarylisoxazoles as highly selective cyclooxygenase-1 (COX-1) inhibitors

Article abstract of DOI:10.1021/jm301905a

3-(5-Chlorofuran-2-yl)-5-methyl-4-phenylisoxazole (P6), a known selective cyclooxygenase-1 (COX-1) inhibitor, was used to design a new series of 3,4-diarylisoxazoles in order to improve its biochemical COX-1 selectivity and antiplatelet efficacy. Structure-activity relationships were studied using human whole blood assays for COX-1 and COX-2 inhibition in vitro, and results showed that the simultaneous presence of 5-methyl (or -CF₃), 4-phenyl, and 5-chloro(-bromo or -methyl)furan-2-yl groups on the isoxazole core was essential for their selectivity toward COX-1. 3g, 3s, 3d were potent and selective COX-1 inhibitors that affected platelet aggregation in vitro through the inhibition of COX-1-dependent thromboxane (TX) A₂. Moreover, we characterized their kinetics of COX-1 inhibition. 3g, 3s, and 3d were more potent inhibitors of platelet COX-1 and aggregation than P6 (named 6) for their tighter binding to the enzyme. The pharmacological results were supported by docking simulations. The oral administration of 3d to mice translated into preferential inhibition of platelet-derived TXA₂ over protective vascular-derived prostacyclin (PGI₂).

Full text of DOI:10.1021/jm301905a

Subscriber access provided by UQ Library
Article
Synthesis, Pharmacological Characterization and Docking Analysis of
a Novel Family of Diarylisoxazoles as Highly Selective COX-1 Inhibitors
Paola Vitale, Stefania Tacconelli, Maria Grazia Perrone, Paola Malerba, laura simone, Antonio Scilimati,
Antonio Lavecchia, Melania Dovizio, Emanuela Marcantoni, Annalisa Bruno, and Paola Patrignani
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm301905a • Publication Date (Web): 07 May 2013
Downloaded from http://pubs.acs.org on May 15, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis, Pharmacological Characterization and Docking Analysis of a Novel Family of
Diarylisoxazoles as Highly Selective COX-1 Inhibitors
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Paola Vitale, Stefania Tacconelli, Maria Grazia Perrone, Paola Malerba, Laura Simone, Antonio
Scilimati*, Antonio Lavecchia*, Melania Dovizio, Emanuela Marcantoni, Annalisa Bruno and
Paola Patrignani*.
Dipartimento di FarmaciaꢀScienze del Farmaco, Università degli Studi di Bari “A. Moro”, Via
Orabona 4, 70125 Bari, Italy (PV, MGP, PM, LS, AS); Dipartimento di Farmacia, “Drug
Discovery” Laboratory, Università di Napoli “Federico II”, via D. Montesano 49, 80131 Napoli,
Italy (AL); Department of Neuroscience and Imaging (ST, MD, PP), Medicine and Aging (EM,
AB), “G. d’Annunzio” University, and Ce.S.I., Chieti, Italy (ST, MD, EM, AB, PP).
PV, ST and MGP contributed equally
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
ꢀ(5ꢀChlorofuranꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀphenylisoxazole(P6), a known selective cyclooxygenase(COX)ꢀ
3
1 inhibitor, was used to design a new series of 3,4ꢀdiarylisoxazoles, in order to improve its
biochemical COXꢀ1 selectivity and antiplatelet efficacy. Structureꢀactivity relationships were
studied using human whole blood assays for COXꢀ1 and COXꢀ2 inhibition in vitro and results
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
showed that the simultaneous presence of 5ꢀmethyl(or ꢀCF ),4ꢀphenyl and 5ꢀchloro(ꢀbromo or ꢀ
methyl)ꢀfuranꢀ2ꢀyl groups on the isoxazole core was essential for their selectivity towards COXꢀ1.
3g, 3s, 3d were potent and selective COXꢀ1 inhibitors which affected platelet aggregation in vitro
through the inhibition of COXꢀ1ꢀdependent thromboxane(TX)A . Moreover, we characterized their
2
kinetics of COXꢀ1 inhibition. 3g, 3s, and 3d resulted more potent inhibitors of platelet COXꢀ1 and
aggregation than P6(named 6) for their tighter binding to the enzyme. The pharmacological results
were supported by docking simulations. The oral administration of 3d to mice translated into
preferential inhibition of plateletꢀderived TXA
2
over protective vascularꢀderived
prostacyclin(PGI2).
.
2
ACS Paragon Plus Environment

Page 3 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Cyclooxygenase (COX)ꢀ1 and COXꢀ2 catalyze the first step of the biosynthesis of prostanoids from
1
arachidonic acid (AA). Differently from COXꢀ2 gene that is mainly inducible, COXꢀ1 is a
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
housekeeping gene, constitutively expressed in almost all mammalian tissues and cells. However,
2
the expression of COXꢀ1 can be regulated in some circumstances, such as during development. In
3,4
physiological conditions, COXꢀ1 is highly expressed in the gastrointestinal(GI) tract and platelets,
2
where it is involved in the generation of cytoprotective prostaglandin (PG)E and platelet
1
2
proaggregatory thromboxane(TX)A , respectively.
COXꢀ1 plays a role in several pathological conditions, such as thrombosis, atherosclerosis and
5ꢀ10
tumorigenesis.
Importantly, platelet COXꢀ1 is the target of one of the most efficacious
11
antithrombotic agents used for prevention of vascular occlusive events, i.e. aspirin.
1
2
Aspirin irreversibly acetylates Ser529 and Ser516 of COXꢀ1 and COXꢀ2, respectively, leading to
irreversible enzyme inactivation. Due to pharmacokinetics features (i.e. short halfꢀlife: 20 minutes)
and pharmacodynamics features (higher potency to inhibit COXꢀ1 than COXꢀ2), lowꢀdoses of
aspirin (75ꢀ100 mg once daily) act by affecting platelet COXꢀ1 activity while they cause only a
marginal and transient inhibitory effect on COXꢀ2 and extraꢀplatelet cellular COXꢀ1. Lowꢀdoses of
aspirin cause an almost complete suppression (>95%) of platelet TXA
2
generation ex vivo,
11,13
persisting throughout dosing interval (i.e. 24h).
This is a fundamental requisite to obtain an
14
antiplatelet effect, since even tiny concentrations of TXA may activate platelets and importantly
2
1
5
they synergize with other platelet agonists. The antiplatelet effect of aspirin is strictly related to
11,12
platelet turnover.
Thus, enhanced platelet turnover rate detected in some cardiovascular(CV)
conditions, such as diabetes, might decrease the efficacy of the drug to halt almost completely
16
platelet TXA
2
generation. Moreover, we have recently shown that platelets contain COXꢀ1
mRNA and the enzymatic machinery for protein synthesis, thus, these anucleated megacaryocyte
17
fragments are able to synthesize de novo COXꢀ1, in response to platelet activation in vitro. Thus,
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
we hypothesized that a reversible inhibitor of platelet COXꢀ1 may be useful as antiplatelet agent in
some patients with CV disease where lowꢀdose aspirin failed to completely affect platelet TXA
2
generation.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Naproxen is a weak, timeꢀdependent reversible inhibitor of platelet COXꢀ1 and is the only
traditional(t) nonsteroidal antiꢀinflammatory drug(NSAID) (also known as nonselective NSAIDs)
which causes a profound and persistent COXꢀ1 inhibition due to its long halfꢀlife when given at
1
8,19
high doses.
This effect on platelet COXꢀ1 is plausibly involved in a better CV safety profile of
20
naproxen than other tNSAIDs. However, naproxen use in the general population is not
cardioprotective for two major reasons: i) it is associated with variability in causing complete
(
>95%) suppression of the maximal capacity of platelets to generate TXA at dosing interval (ie 12h
2
2
1,22
after dosing) which, however, can be reduced at high doses;
ii) it can cause a profound reduction
of the biosynthesis of vasoprotective prostacyclin (PGI
2
) by inhibiting COXꢀ2 activity in the
19,22
vasculature.
A selective inhibitor of COXꢀ1 may be associated with improved CV safety than naproxen because,
in front of a profound suppression of platelet COXꢀ1, it may leave PGI to act by restraining: i)
2
platelet activation induced by all platelet agonists, not only TXA
2
, ii) coagulation for its capacity to
2
3
induce thrombomodulin, iii) oxidative stress for its capacity to induce hemeꢀoxygenaseꢀ1. Finally,
23
2
PGI may cause vasodilation and inhibition of vascular smooth muscle cells proliferation.
Selective inhibitors of COXꢀ1 may be associated with enhanced risk of upper GI bleeding(UGIB)
due to the inhibition of COXꢀ1ꢀderived prostanoids which are involved in the constitutive defense
mechanism operating in physiological conditions. However, they could be better tolerated than
tNSAIDs. In fact, it has been shown that tNSAIDs, which cause a profound and coincident
24
inhibition of both COXꢀ1 and COXꢀ2, are associated with higher risk of UGIB. This is
mechanistically plausible with results obtained in the mouse showing that inhibition of both COXꢀ1
25
and COXꢀ2 is required for the formation of gastric lesions.
4
ACS Paragon Plus Environment

Page 5 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The development of reversible and highly selective COXꢀ1 inhibitors could provide valuable tools
2
6,27
for investigating the role of COXꢀ1 in tumorigenesis.
Recent studies have shown that lowꢀdose
aspirin which targets almost selectively platelet COXꢀ1 reduces longꢀterm incidence and mortality
2
8,29
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
due to colorectal cancer and other types of cancer.
The use of reversible, selective COXꢀ1
inhibitors will help to unravel the possible contribution of extraꢀplatelet sources of COXꢀ1 in
30
tumorigenesis.
In a previous study, we found that 3ꢀ(5ꢀchlorofuranꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀphenylisoxazole (P6) is a
31
highly selective COXꢀ1 inhibitor. Structureꢀactivity relationship (SAR) studies of a series of
diarylpyrazoles analogues of P6 (named 6) allowed us also to clarify that isoxazole core ring and
32
the presence of furanyl group are crucial determinants for high COXꢀ1 selectivity. In the present
study, we synthesized in fair to good yield a new series of diarylisoxazoles (Table 1). We compared
33,34
6 and novel compounds for: i) the degree of COXꢀ1 selectivity in human whole blood
; ii) the
inhibition of platelet aggregation and TXB generation, in platelet rich plasma (PRP) in response to
2
arachidonic acid (AA) or collagen. Since COXꢀ1 activity is regulated by substrateꢀdependent
cooperative activation which can influence the affinity of many cyclooxygenase inhibitors for
35
COXꢀ1, we compared concentrationꢀresponse curves for inhibition of COXꢀ1 activity in isolated
human platelets by 6 and the novel COXꢀ1 inhibitors, in the presence of low and high
concentrations of AA (i.e., 0.50 ꢀM, and 10 ꢀM, respectively). Moreover, we studied the
reversibility of the inhibition of COXꢀ1 and COXꢀ2 in two cell lines which selectively express only
COXꢀ1 or COXꢀ2, i.e. the THPꢀ1 human monocytic leukemia cell line and the MDAꢀMBꢀ231
breast cancer cell line, respectively. We present the results of docking simulation of diarylisoxazole
inhibitors at the COXꢀ1 active site to predict the molecular determinants of the interaction. Among
the novel diarylisoxazoleꢀderivatives, 3d was administered to mice to assess the inhibitory effect on
3
6
the systemic biosynthesis of TXA and vascular PGI .
2
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
RESULTS AND DISCUSSION
Rational design of novel diarylisoxazoles. 3ꢀ(5ꢀChlorofuranꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀphenylisoxazole
(herein 6) was used to design a set of novel diarylisoxazoles (Table 1). We projected chloroꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
substituted 3ꢀarylisoxazoles as 3a and 3b to evaluate the effect of the C ꢀsubstituent modification
3
both on the potency of these compounds to inhibit COXꢀ1 and COXꢀ2 activities and on the extent
of their COXꢀ1 selectivity, assessed in human whole blood assays. Moreover, to verify the role of
3
the chlorine atom and of the aromaticity of the furanyl group at C isoxazole ring position, we
prepared the compounds 3g, 3q, 3r and 3s. In addition, we studied the effects of the modifications
on the group at isoxazole C by using 3h, 3i, 3l and 3m; in particular, 3m is the acetoxy derivative
5
,
of 6. The introduction of a CF ꢀgroup in place of a methyl in 6, provided 3ꢀ(5ꢀchlorofuranꢀ2ꢀyl)ꢀ4ꢀ
3
phenylꢀ5ꢀtrifluoromethylisoxazole (3d), a more stable lipophilic compound and eventually more
3
stable from a metabolic viewpoint, because the CF ꢀgroup in place of a methyl does not undergo to
37
any transformation as shown in in vitro experiments with celecoxib thus, 3d should not have a
38
metabolism similar to valdecoxib (and potentially 6). The introduction of a fluorine atom at orthoꢀ
and paraꢀposition, respectively, of the aromatic ring bonded at isoxazole C of 6, gave 3e and 3f
4
synthesized to further improve 6 metabolic stability, being orthoꢀ and paraꢀposition of the phenyl
4
ring at isoxazole C susceptible of enzymeꢀmediated hydroxylation reaction. Similarly, (Z)ꢀ4ꢀ
aminoꢀ4ꢀ(5ꢀchlorofuranꢀ2ꢀyl)ꢀ3ꢀphenylbutꢀ3ꢀenꢀ2ꢀone (4) and (Z)ꢀ4ꢀaminoꢀ4ꢀ(furanꢀ2ꢀyl)ꢀ3ꢀ
phenylbutꢀ3ꢀenꢀ2ꢀone (5) were prepared, as “ringꢀopened” analogues and as potential metabolites of
6
, since it is known that reduction of the isoxazole ring yields the formation of an enaminoꢀketone
30
metabolite in phase I metabolic pathways of valdecoxib. Moreover, compounds having the aryl
groups positioned at C and C ꢀisoxazole ring (3t, 3u, 3w and 3v) and not at C /C , were designed
3
5
3
4
to evaluate the effect of their greater distance on the COXꢀ1 inhibitory activity with respect to 6. 3n
and 3p bearing a NOꢀreleasing group (morfoline and nitrate) linked through a spacer to the
isoxazole ring were also designed and synthesized.
6
ACS Paragon Plus Environment

Page 7 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1.
Synthesis of novel diarylisoxazoles. We have previously found that a number of 3,4ꢀ
diarylisoxazoles, structurally related to the valdecoxib, a selective COXꢀ2 inhibitor, could be
prepared by a two step methodology: a) 1,3ꢀdipolar cycloaddition of arylnitrile oxides to the lithium
enolate of phenylacetone, obtained by metallation with LDA at 0°C, and b) the dehydrationꢀ
aromatization reaction under basic (or acid) conditions of the 3ꢀarylꢀ5ꢀhydroxyꢀ5ꢀmethylꢀ4ꢀphenylꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
1
2
ꢀisoxazoline intermediates. By using the same synthetic methodology we synthesized the novel
compounds 3a-3g in high yields by a oneꢀpot procedure starting from arylnitrile oxides and the
enolates of 3ꢀarylꢀ2ꢀpropanones, easily obtained with sodium hydride or lithium diisopropylamide
at 0°C (Scheme 1 and Table 2).
Scheme 1.
Table 2.
We synthesized the 3ꢀ(5ꢀchlorofuranꢀ2ꢀyl)ꢀ5ꢀhydroxyꢀ4ꢀphenylꢀ4,5ꢀdihydroisoxazoleꢀ5ꢀcarboxylic
acid (2h) by 1,3ꢀdipolar cycloaddition of 5ꢀchlorofuranꢀ2ꢀyl nitrile oxide to the lithium dianion of
phenylpyruvic acid obtained at 0°C in the presence of LDA (2 equivalents), followed by final
treatment with HCl (Scheme 2).
Scheme 2
However, the isolation through silica gel column chromatography of 2h and its dehydrationꢀ
aromatization reaction into the corresponding 3ꢀ(5ꢀchlorofuranꢀ2ꢀyl)ꢀ4ꢀphenylisoxazoleꢀ5ꢀ
carboxylic acid (3h) under basic (or acid) conditions was unsuccessful. Thus, 2h was first converted
3
into its corresponding methyl ester 2i in the presence of BF /MeOH and after column
2 3
chromatography over silica gel, 2i was refluxed in MeOH and acqueous Na CO to give 3h in high
yield (Scheme 2) (Ar = 5ꢀchlorofuranꢀ2ꢀyl).
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
On the other hand, 3h was also used as the starting material for the preparation of its methyl ester
3
i, as well as of the alcohol 3l, in turn transformed into its acetoxymethyl derivative 3m (Scheme
3).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 3
Further, 3h could also be converted into isoxazoles 3n-p following the procedures depicted in
Scheme 4.
Scheme 4
2
4
3ꢀ(5ꢀChlorofuranꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀphenylisoxazole (6), obtained as previously reported, was used
as starting material for the preparation of 3q, 3r, 4 and 5, by hydrogenation performed in the
presence of a different catalyst (Scheme 5).
Scheme 5
Isoxazole 3q was transformed in high yield into 3s by bromination in the presence of NBS/DMF
(Scheme 6).
Scheme 6
Differently, 3t and 3u (Ar = 5ꢀchlorofuranꢀ2ꢀyl and phenyl, respectively) were synthesized in good
yields by the 1,3ꢀdipolar cycloaddition of the appropriate arylnitrile oxide to the methyl 3ꢀ
2 2
phenylpropiolate in anhydrous CH Cl (Scheme 7). From them, the corresponding carboxylic acid
derivatives 3v and 3w were then quantitatively obtained by hydrolysis under basic conditions (KOH
in THF).
Scheme 7
Screening strategy to identify and characterize novel highly selective COX-1 inhibitors. The
extent of selectivity towards COXꢀ1 by the novel diarylisoxazoles was assessed in human whole
33,34
blood, in vitro,
by calculating COXꢀ2/COXꢀ1 IC50 ratios (Table 1 and Figure S1). In the first
phase of the screening, COXꢀ1 and COXꢀ2 inhibition by the compounds was evaluated up to 100
ꢀM (Table 1).
8
ACS Paragon Plus Environment

Page 9 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
We selected 3 compounds (3g, 3s and 3d) with IC50 values for COXꢀ1<1 ꢀM and high COXꢀ1
selectivity (i.e. COXꢀ2/COXꢀ1 IC50 ratios >100) and the acetoxyꢀderivative of 6 (3m) (which might
be able to acetylate COXꢀ1 similarly to aspirin) for inꢀdepth characterization of COXꢀ1 selectivity
and type of inhibitory mechanism of COXꢀisozymes. In order to be able to define the degree of
COXꢀ1 selectivity in human whole blood of the selected compounds, the concentrationꢀresponse
curves for inhibition of whole blood COXꢀ2 were assessed up to 1000 ꢀM. Moreover, they were
assessed for: i) inhibition of platelet aggregation in PRP induced by AA or collagen and ii)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
inhibition of TXB generation in AAꢀ or collagenꢀstimulated PRP. Moreover, we evaluated the
2
concentrationꢀresponse curves for inhibition of platelet COXꢀ1 activity by 3s, 6 and 3m in human
washed platelets, in the presence of low and high concentrations of AA (0.50 and 10 ꢀM,
respectively). Reversibility of the inhibition of COXꢀ1 and COXꢀ2 was studied in two cell lines
which selectively expressed only COXꢀ1 or COXꢀ2, i.e. the THPꢀ1 human monocytic leukemia cell
line and the MDAꢀMBꢀ231 breast cancer cell line, respectively. In all these experiments, 6 was used
as reference compound.
Inhibitory effects of the diarylisoxazole derivatives on COX-1 and COX-2 activities, in human
whole blood. In human whole blood, allowed to clot at 37°C for 1 h, we detected 452±32 ng/ml of
34
TXB
2
(mean±SEM, n=28) which is generated by platelet COXꢀ1 activity. Fourteen, out of 21
compounds, inhibited platelet COXꢀ1 activity in a concentrationꢀdependent fashion (Figure S1) and
the IC50 values were reported in Table 1. In contrast, 7 compounds did not affect platelet COXꢀ1
activity up to 100 ꢀM (Table 1). All compounds were assessed for the capacity to inhibit LPSꢀ
induced COXꢀ2 activity in whole blood and the IC50 values were reported in Table 1. Fifteen
compounds, assessed up to 100 ꢀM, inhibited COXꢀ2 activity less than 50%. We selected three
compounds showing both an IC50 for COXꢀ1<1ꢀM and a COXꢀ2/COXꢀ1 selectivity >100 (i.e. 3s,
3
g and 3d).
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
s and 3g showed IC50 values for COXꢀ1 that were approximately 2ꢀfold lower than that of 6. 3d
showed an IC50 value for COXꢀ1 comparable to 6 (Table 1). In order to define the IC50 values for
inhibition of COXꢀ2 by these three compounds and 6, concentrationꢀresponse curves were
performed at concentrations up to 1000 ꢀM (Figure 1AꢀD). At 1000 ꢀM, 6 inhibited COXꢀ2
activity by 61.00±5.20% which was not statistically different from that caused by 3d (49±3%).
Altogether our data showed that 6 and 3d had similar potency to inhibit COXꢀ1 and COXꢀ2 in
whole blood and had a comparable COXꢀ1 selectivity >1000 (Table1 and Figure 1A and D,
respectively). Differently, 3s and 3g inhibited COXꢀ2 activity with the following IC50 values
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
[
expressed as mean and 95% confidence interval (CI)]: 34 (19ꢀ59) and 68.20 (41.00ꢀ114.50) ꢀM,
respectively (Table 1 and Figure 1B and C) and showed lower COXꢀ1 selectivity than 6 (Table 1,
Figure 1B and 1C) (COXꢀ2/COXꢀ1 IC50 ratios: 103.03 and 213.12, respectively). Thus, 3s and 3g
resulted more potent towards COXꢀ1 than 6 and they were characterized by high COXꢀ1 selectivity
which however was lower than that showed by 6.
The acetoxy derivative of 6 (i.e. 3m) inhibited platelet COXꢀ1 activity with IC50 of 5 (2ꢀ12) ꢀM
(Figure 1E and table 1) that resulted significantly higher (8ꢀfold) than the value obtained by 6. At
1
000 ꢀM, 3m inhibited COXꢀ2 activity by 40±4% and this value was lower than that caused by 6.
Thus, 3m was less potent than 6 to inhibit both COXꢀisozymes in human whole blood.
In order to verify whether different IC values for inhibition of COXꢀ1 and COXꢀ2 by these novel
5
0
compounds and 6 were driven by differences in binding to plasma proteins, we evaluated the degree
of binding to human serum albumin (HSA) and α ꢀacid glycoprotein (AGP) of 3d, 3g, 3s and 6. The
compounds showed high degree of protein binding (i.e. 94.30±0.86, 94.70±9.33, 96.10±1.27 and
5.70± 0.49%, respectively, mean±SEM, n=6) and the values were not significantly different from
1
9
each other.
Effects of selected compounds (3s, 3g, 3d, and 3m) and 6 on platelet aggregation and TXB
2
generation in PRP. The compounds 3s, 3g, 3d, and 3m were assessed for their inhibitory effects
1
0
ACS Paragon Plus Environment

Page 11 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
on platelet aggregation induced by AA 1mM (Figure 2A, C, E and G) in comparison to 6 (Figure
3
A). Up to 3ꢀM, 3s, 3g and 3d did not inhibit AAꢀinduced platelet aggregation. In contrast, at 10
M, they almost completely inhibited AAꢀinduced platelet aggregation (96.33±1.80%,
7.80±1.60% and 99.20±0.60%, respectively), in a statistically significant fashion (P<0.001 versus
ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
DMSO vehicle) (Figure 2A, C, E).
The three novel diarylisoxazole derivatives resulted more potent than 6 in inhibiting AAꢀinduced
platelet aggregation (Figure 2A, C, E and Figure 3A). In fact, 6 did not affect AAꢀinduced platelet
aggregation at 10 ꢀM but it was required a concentration of 30 ꢀM to almost completely abrogate
the platelet response (97.00±1.80%, P<0.001 versus DMSO vehicle) (Figure 3A). Similarly to 6,
3m significantly inhibited platelet response at 30 ꢀM (97.00±1.50%, P<0.001 versus DMSO
vehicle) (Figure 2G).
As shown in Figure 4(A, C, E), 3s, 3g and 3d caused a concentrationꢀdependent inhibition of
collagenꢀinduced platelet aggregation. Similarly to AAꢀinduced platelet aggregation, up to 3ꢀM, 3s,
3
g and 3d did not significantly inhibit collagenꢀinduced platelet aggregation. A profound and
statistically significant (P<0.001 versus DMSO vehicle) inhibition of collagenꢀinduced platelet
aggregation occurred at 10ꢀM (78.00±7.50%, 66±12% and 72±9%, respectively). Even in collagenꢀ
induced platelet aggregation, 6 and 3m required a concentration of 30ꢀM to cause a statistically
significant inhibition >50% (64.60±5.00% and 62±8%, P<0.01 and P<0.001 versus DMSO vehicle,
respectively) (Figure 3C and 4G).
Thus, despite the novel diarylisoxazole derivatives showed a similar (3d) or a small higher potency
(3s and 3g) than 6 to inhibit platelet COXꢀ1 activity in whole blood, they were all more potent than
6
to affect platelet aggregation in PRP induced by AA and collagen.
To clarify the reason of these results, we assessed TXB generation in AAꢀ and collagenꢀinduced
PRP before and after treatment with the different compounds. In PRP stimulated for 5 minutes with
2
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
AA 1 mM, we detected 1163+220 ng/ml of TXB
2
(n=13). The compounds 3s, 3g and 3d inhibited
TXB
2
generation in AAꢀstimulated PRP in a concentrationꢀdependent fashion (Figure 2B, D, F)
with comparable IC50 values of 3.40(2.20ꢀ5.30), 4.30(2.00ꢀ10.00), 3.70(2.00ꢀ7.00) ꢀM,
respectively, which resulted approximately 3ꢀfold lower than the IC50 value of 6 [12(7ꢀ20)ꢀM]
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Figure 3B). However, only the IC50 values of 3s and 3d were significantly different from 6.
In PRP stimulated for 5 minutes with collagen 2 ꢀg/ml, we detected 82+20 ng/ml of TXB (n=13).
2
As shown in Figure 4(B, D and F), 3s, 3g and 3d inhibited TXB
PRP in a concentration dependent fashion with comparable IC50 values of 7.60(6.60ꢀ9.60),
.40(4.50ꢀ12.00), 10(9ꢀ11) ꢀM, respectively. These values were lower than 6 IC50 value
18.60(11.00ꢀ32.00)ꢀM] (Figure 3D), but a statistically significant difference was detected only for
s and 3d versus 6.
2
generation in collagenꢀstimulated
7
[
3
Differently, 3m (Figure 2H and 4H) inhibited TXB generation in AAꢀand collagenꢀstimulated PRP
2
with IC50 values of 20(15ꢀ26) and 23.00(13.00ꢀ42.50) ꢀM, respectively, which were similar to 6
and significantly higher than 3s, 3g and 3d.
2
These results showed that all compounds inhibited TXB generation in AAꢀstimulated PRP and
collagenꢀstimulated PRP. However, we found that all compounds were less potent to inhibit TXB2
generation in PRP, in response to AA 1 mM and collagen 2 ꢀg/ml, than in whole blood. Among the
diarylisoxazole derivatives that we studied, 6 was the compound that has lost more potency to
inhibit COXꢀ1 in PRP versus whole blood.
Comparison of concentration-response curves for inhibition of platelet COX-1 by
diarylisoxazole derivative compounds in the presence of low and high concentrations of AA.
Mechanistically, COX inhibitors fall into two broad categories: timeꢀdependent/slowly reversible,
39ꢀ41
such as indomethacin, and timeꢀindependent/rapidly reversible, such as ibuprofen.
Moreover,
aspirin causes an irreversible inhibition of COXꢀisozymes. To verify the type of mechanism of
1
2
ACS Paragon Plus Environment

Page 13 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
COXꢀ1 inhibition by the novel diarylisoxazole derivatives and 6, we performed experiments in
washed human platelets and increasing concentrations of different compounds were preꢀincubated
with the cells, to allow the possible formation of tightly bound enzymeꢀinhibitor complexes, then
AA, at low (0.50ꢀM) and high concentrations (10ꢀM), was added to initiate the reaction that was
terminated after 30 minutes of incubation. We hypothesized that a rapidly reversible inhibitor, but
not a slowly reversible inhibitor, should be characterized by a right shift of the logꢀconcentrationꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
response curve for inhibition of TXB obtained in the presence of high concentration of AA versus
that obtained in the presence of low concentration of AA.
Human washed platelets preꢀtreated with vehicle and then incubated with AA 0.50 ꢀM for
additional 30 minutes at 37°C, generated 54.00±7.90 ng/ml of TXB
TXB by human washed platelets was significantly (P=0.008) higher in the presence of 10ꢀM of
AA (680.00±207.50 ng/ml, n=10).
First, we studied the inhibitory effects of ibuprofen, aspirin and indomethacin (Figure 5 AꢀC). At
0ꢀM of AA, ibuprofen caused a concentrationꢀdependent reduction of TXB generation with an
IC value that was significantly higher than that found at 0.50ꢀM of AA [1.86(0.88ꢀ3.90) versus
2
(n=11). The production of
2
1
2
5
0
0
.29(0.17ꢀ0.48)ꢀM, respectively] (Figure 5A). In contrast, the concentrationꢀresponse curves for
inhibition of platelet COXꢀ1 by indomethacin were not influenced by the concentration of AA
(Figure 5B). Aspirin, which is an irreversible inhibitor of COXꢀ1, inhibited platelet TXB
2
generation with comparable potency at the two different concentrations of AA used to initiate the
enzymatic reaction (Figure 5C). Once we had the demonstration that the use of this experimental
model is appropriate to discriminate between freely reversible and slowly reversible/irreversible
inhibitors of COXꢀ1, we tested the behaviour of 3s, 3g and 3d, 6 and the acetoxyꢀderivative of 6
(3m).
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
As shown in Figure 5DꢀF, 3s, 3g and 3d inhibited TXB
2
production in human washed platelets with
similar IC50 values at both AA concentrations, which may suggest a strong interaction of the drug
with platelet COXꢀ1, similarly to indomethacin. In contrast, in the presence of high concentrations
of AA, the concentration response curves for inhibition of platelet COXꢀ1 activity by 6 and 3m
were right shifted as compared to those obtained in the presence of low concentrations of AA
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Figure 5G, H). This finding may suggest a weaker interaction of the two compounds with COXꢀ1
than 3s, 3g and 3d.
Assessment of reversibility of THP-1 cell COX-1 inhibition by diarylisoxazole derivatives. As
shown in Figure 6A, THPꢀ1 cells, cultured in suspension, constitutively expressed COXꢀ1 but not
COXꢀ2. THPꢀ1 cells incubated with vehicle (DMSO) for 30 minutes (at room temperature) and then
stimulated with AA 20ꢀM for further 30 minutes (at 37°C), generated TXB (0.34±0.05 pg/ꢀg
2
protein). At 100ꢀM, 6, four selected diarylisoxazole derivatives (3s, 3g, 3d and 3m), aspirin,
indomethacin and ibuprofen profoundly (>90%) inhibited COXꢀ1 activity (P<0.0001 versus
DMSO) (Figure 6B).
The profound inhibitory effect of COXꢀ1 activity by aspirin, indomethacin, 3s, 3g and 3d persisted
after extensive washing of cells (Figure 6B). This is consistent with the irreversible inhibition of
COXꢀ1 by aspirin and slow reversibility of the inhibition of COXꢀ1 by indomethacin and the
diarylisoxazole derivatives 3s, 3g and 3d. Differently, the profound inhibitory effect of COXꢀ1 by
ibuprofen was completely reverted by cell washing (Figure 6B), consistent with its mechanism of
action of COXꢀ1 inhibition which is rapidly reversible. The inhibition of COXꢀ1 by 3m and 6 was
partially reversed (Figure 6B) by cell washing.
Altogether these results suggest that: i) similarly to indomethacin, 3s, 3g and 3d had a strong
interaction with COXꢀ1 which is slowly reversible; ii) the inhibition of COXꢀ1 by 3m and 6 was
reversible even if it occurred with a slower kinetics than that showed by ibuprofen (Figure 6B).
1
4
ACS Paragon Plus Environment

Page 15 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Assessment of reversibility of MDA-MB231 cell COX-2 inhibition by diarylisoxazole
derivatives. The reversibility of the interaction of the selected diarylisoxazole derivatives (3s, 3g,
3d, 3m) and 6 with COXꢀ2 was assessed in MDAꢀMB231 cells which expressed COXꢀ2, but not
COXꢀ1 (Figure 7A). The effects of the novel compounds were compared with those of aspirin,
indomethacin and ibuprofen. Cells, preꢀincubated with DMSO vehicle and then for further 30
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
minutes with AA 0.50ꢀM, generated 1.40+0.08 pg/ꢀg proteins of PGE .
We assessed the inhibitory effects on COXꢀ2 by 100ꢀM of the different compounds, in MDAꢀ
MB231 cells cultured in a medium containing very low concentrations of proteins (0.50% FBS).
The preꢀincubation of ibuprofen, indomethacin, 3s, 3g, 3d and 6, at the concentration of 100ꢀM,
with MDAꢀMB231 cells for 30 minutes caused a profound and significant inhibition of AAꢀinduced
PGE generation which ranged from 86 to 70% (Figure 7B). This concentration of diarylisoxazole
2
derivatives inhibited only marginally COXꢀ2 activity in whole blood (Figure 1), because in this
assay the presence of plasma proteins reduced the free drug fraction.
A lower degree of inhibition was caused by aspirin (60+4%) which is consistent with previous
27,42
results showing a lower potency of aspirin towards COXꢀ2 than COXꢀ1.
As shown in Figure
7B, 3m did not significantly affect COXꢀ2 activity.
The reversibility of the interaction between COXꢀ2 and the different compounds was evaluated by
assessing the recovery of COXꢀ2 activity by washing the cells previously preꢀincubated with them
for 30 minutes. Similarly to aspirin and indomethacin, the inhibition caused by 3g was not
significantly reversed by the washing procedure. In contrast, the inhibition caused by ibuprofen and
3
d was completely reversed. The inhibition of COXꢀ2 activity by 3s and 6 were only partially
reversed by washing the cells (Figure 7B).
Altogether these results showed that the diarylisoxazole derivatives may affect COXꢀ2 at high
concentrations and this effect was caused by different mechanisms. 3d resembles the mechanism of
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ibuprofen, i.e. fastly reversible inhibition of COXꢀ2, thus suggesting a weak binding to COXꢀ2; 3g
resembles the mechanism of indomethacin, i.e. slowly reversible inhibition, thus suggesting a
strong binding to COXꢀ2; 6 and 3s were characterized by intermediate responses, thus suggesting
that they bind COXꢀ2 with lower affinity than 3g but higher affinity than 3d.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Effects of the oral administration of 3d on in vivo markers of COX-1 and COX-2 activity in
mice. Among the novel diarylisoxazoleꢀderivatives, 3d represents an interesting candidate for
preclinical development as novel antithrombotic agent and further studies are needed to assess its
efficacy in animal models of thrombosis and atherosclerosis. Here we verified the capacity of the
oral administration of 3d [20mg/kg/day, p.o., in 1% carboxymethylcellulose (CMC) solution given
for three consecutive days to mice] to affect the systemic biosynthesis of TXA
urinary levels of an enzymatic metabolite of TXB , i.e. 11ꢀdehydroꢀTXB (TXꢀM) which is mainly
derived from plateletꢀCOXꢀ1. In fact, its urinary levels are reduced by 70ꢀ80% in mice and humans
2
by assessing the
2
2
1
9,43
after dosing with lowꢀdose aspirin.
70%, P=0.0007 versus CMC, n=4) of the urinary levels of TXꢀM.
Moreover, we assessed the impact of 3d administration on the urinary levels of PGI
by the measurement of the nonꢀenzymatic hydrolysis product 6ꢀketoꢀPGF1α) which is a marker of
As shown in Figure 8A, 3d caused a significant reduction
(
2
(as assessed
4
4,45
renal biosynthesis of PGI
did not significantly affect the urinary excretion of PGI
effective also in vivo to affect the biosynthesis of the platelet agonist TXA
2
, mainly derived from vascular COXꢀ2.
. Altogether these results showed that 3d is
while sparing COXꢀ2ꢀ
As shown in Figure 8B, 3d
2
2
dependent PGI
2
. Vascular PGI
2
is a potent vasodilator, an inhibitor of platelet function, but it is also
46
involved in the maintenance of cortical kidney functions.
Docking Studies. To get a better comprehension of the high COXꢀ1 inhibitory potency of the
newly synthesized isoxazole derivatives at a molecular level and shed light on the structural
determinants for their COXꢀ1 specificity, docking experiments were conducted using the ovine
47
COXꢀ1 in complex with celecoxib (PDB code: 3kk6). Docking investigations were carried out
1
6
ACS Paragon Plus Environment

Page 17 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
48
through the automated docking program GOLD 5.0.1 and the outcomes were rationalized using
49
the clustering algorithm AclAP. Alternative conformations of R120, Y355 and S530 have been
previously observed in the COX crystal structures complexed with several inhibitors, highlighting
50ꢀ53
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the innate plasticity of the active site.
Accordingly, the side chains of R120, Y355 and S530
were allowed to move during the docking experiments.
We performed docking of 3s and 3m which may affect COXꢀ1 activity by different mechanisms. In
fact, pharmacological findings have shown that: i) 3s inhibited COXꢀ1 through a slowly reversible
mechanism; ii) 3m inhibited COXꢀ1 through a rapidly reversible mechanism. Docking of 3s into the
COXꢀ1 active site revealed a very clear preference for two alternative poses (Figure 9A and B). In
ꢀ1
the topꢀranking pose A (GOLD fitness score = 58.70 kcal mol ), 3s lies on the internal side of the
constriction, at the base of the funnelꢀshaped entrance to the COXꢀ1 active site (Figure 9A). The
constriction, which consists in a Hꢀbonding network formed by R120, E524, and Y355, is proposed
5
4,55
to control the binding dynamics of some NSAIDs and selective COXꢀ2 inhibitors.
Both O1
furan oxygen and N2 isoxazole nitrogen of the inhibitor are within Hꢀbonding distance of the OH
group of Y355. The O1ꢁꢁꢁO and N2ꢁꢁꢁO atoms are separated by 2.70 Å and 2.90 Å, respectively, and
display a favourable geometry for Hꢀbonding. The phenyl ring at position 4 of isoxazole forms
hydrophobic interactions with I523, G526 and V349. Importantly, the 5ꢀbromofuranyl moiety of 3s
is oriented towards the side pocket (residues 513ꢀ520) of COXꢀ1 and makes hydrophobic contacts
with residues H90, L352, I517, F518 and I523. In the secondꢀranking pose B (GOLD fitness score
ꢀ1
=
53.80 kcal mol ), 3s is located at the apex of the COXꢀ1 channel instead of near the opening
(Figure 9B). The N2 isoxazole nitrogen and the O1 furan oxygen of the inhibitor engage a
bifurcated Hꢀbond with the γꢀOH of S530, displaying a favourable geometry for Hꢀbonding. The
NꢁꢁꢁHO bond is stronger than that OꢀHꢁꢁꢁO, being the N2ꢀO and O1ꢀO distance of 2.50 Å and 2.90
Å, respectively. The phenyl ring at position 4 of isoxazole is in close contact with various
hydrophobic side chains, including L352, I523 and Y355. Interestingly, this phenyl group engages
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
an edgeꢀtoꢀface interaction with Y355. The 5ꢀmethyl group is tucked into a small hydrophobic cleft
lined by V349, L359, V116 and L531 residues. Finally, the 5ꢀbromofuranyl moiety of 3s is oriented
towards the apex of the COXꢀ1 active site and forms van der Waals interactions with residues
M522, F518, L352, W387, Y385, L384, F381.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
When the inhibitor 3m was docked into the COXꢀ1 binding pocket, only 10% of the generated
conformations adopted the aboveꢀdescribed binding mode B of 3s (Figure 9A), whereas 90% were
located at the base of the active site. As illustrated in Figure 10, the N2 isoxazole nitrogen of the
inhibitor makes a Hꢀbond with the OH group of Y355, while the C=O oxygen of the 5ꢀacetoxy
substituent establishes a Hꢀbond with the guanidinium group of R120 typically seen with substrates
and carboxylic acidꢀcontaining inhibitors. The 4ꢀphenyl ring and the 5ꢀchlorofuranyl moiety of 3m
insert into the hydrophobic pocket framed of I523, G526 and V349 and the side pocket,
respectively. Taken together, these results suggest a rationale for the different kinetic behaviour of
the COXꢀ1 inhibitors, 3s, and 3m.
Recent evidences suggest that both COXꢀ1 and COXꢀ2 exhibit half of sites COX activity (i.e., the
56,57
COX active site of only one monomer of a COX dimer can function at a given time)
and that the
enzymes function as an allosteric/catalytic couple in which one COX active site has a catalytic
function (Ecat) that is modulated by the nature of the ligand occupying the COX site of the partner,
47,52,56,58,59
allosteric monomer (Eallo).
This conclusion comes from the recent discovery that the two
56
subunits communicate through the dimer interface. Timeꢀindependent COX inhibitors compete
reversibly with the natural substrate to form a complex enzymeꢀinhibitor (EI). These inhibitors can
6
0
function by binding Eallo and Ecat. On the contrary, timeꢀdependent COX inhibitors are generally
40
imagined as functioning in two steps, although there may be a third step with some inhibitors. The
first step is a rapid, reversible binding to the COX active site of either Eallo or Ecat to form a complex
(EI). The second step involves a slow conformational change to a tightly bound complex (EI*).
40,61,62
The latter process occurs in seconds or minutes and is even more slowly reversible.
1
8
ACS Paragon Plus Environment

Page 19 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Based on our docking results, it is possible to speculate that the binding mode A of the timeꢀ
dependent inhibitor 3s represents the initial enzymeꢀinhibitor complex (EI), where the channel Hꢀ
bonding network of R120, Y355, and E524 is perturbed by the interaction between the inhibitor and
Y355 residue (binding mode A, Figure 9B). Additional protein conformational changes may also
occur as part of this process. Crystal structures of NSAIDs bound to COXꢀ1 and COXꢀ2
demonstrate that protein conformational changes must occur during or after the binding of the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
50,63
inhibitor in the COX active site.
A reꢀarrangement of the inhibitor in the active site ultimately results in a stable complex with S530
residue (binding mode B, Figure 9B), which represents the enzymeꢀinhibitor complex (EI*). This
model is consistent with a timeꢀdependent inhibition mechanism as observed experimentally.
Support for this model comes from fluorescence studies of holoꢀCOXꢀ2, which suggest that for a
64
timeꢀdependent inhibitor, EI and EI* correspond to inhibitor binding at distinct sites.
3
m also disturbs the Hꢀbond network in the constriction region of COXꢀ1 engaging Hꢀbonds with
Y355 and R120 and it is consequently expected that the inhibitor forms a tight complex in the COX
binding site that could represent the EI state (Figure 10). Its status as relatively weak, rapidly
reversible COX inhibitor might derive from the fact that 3m has to bind to both subunits Eallo and
Ecat of COXꢀ1 to inhibit productive binding of AA and that binding in the second subunit is
competitive with AA.
CONCLUSIONS
Structural features of the highly selective COXꢀ1 inhibitor, 3ꢀ(5ꢀchlorofuranꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀ
31
phenylisoxazole (6) have been used to design a new series of diarylisoxazoles. By using the
human whole blood assays in vitro, the SAR study has evidenced that the simultaneous presence of
i) 5ꢀmethyl, ii) 4ꢀphenyl and iii) 5ꢀchlorofuranꢀ2ꢀyl groups on the isoxazole core is crucial to obtain
selective inhibition of COXꢀ1. The replacement of the chloro atom in the furyl core of 6 with the
bromo atom or with the methyl group (i.e., 3s and 3g, respectively) and the introduction of CF3ꢀ
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
group in place of the 5ꢀmethyl group of 6 (i.e. 3d) led to compounds that bind more tightly COXꢀ1
and its inhibition is slowly reversible. This translated into an improved potency to inhibit platelet
aggregation in response to AA and collagen, which occurs as a consequence of the rapid generation
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of TXA from elevated concentration of free AA. We synthesized the acetoxyꢀderivative of 6 (3m)
2
with the aim to acetylate COXꢀisozymes similarly to aspirin. However, this compound was not able
to cause an irreversible inhibition of COXꢀ1 and COXꢀ2, but similarly to 6 was a reversible
competitive inhibitor of COXꢀ1 and this mechanism of inhibition translated into a lower efficacy to
inhibit platelet function than the other diarylisoxazoleꢀderivatives.
Molecular docking provided insight into the binding mode of 3s and 3m into the COXꢀ1 binding
site and rationalized their different kinetic inhibition mechanisms.
Among the novel diarylisoxazoleꢀderivatives that we have synthesized, 3d represents a candidate
for preclinical development as novel antithrombotic agent. It is >1000ꢀfold more potent to inhibit
COXꢀ1 than COXꢀ2 showing a different mechanism of interaction with the two COXꢀisozymes, i.e.
a strong, slowly reversible interaction with COXꢀ1 and a rapidly reversible one with COXꢀ2. The
potency of this compound to inhibit human platelet aggregation correlated with its capacity to
inhibit the generation of TXA
action as antiplatelet agent through the inhibition of platelet COXꢀ1. The oral administration of 3d
to mice caused a selective inhibitory effect on TXB in vivo (mainly COXꢀ1ꢀderived), while sparing
PGI (mainly COXꢀ2ꢀderived). These results may implicate the induction of a protective phenotype
2
in response to proaggregatory stimuli, supporting its mechanism of
2
2
for the CV system by this compound which needs to be confirmed in appropriate animal models.
EXPERIMENTAL SECTION
1
13
Chemistry. Melting points taken on Electrothermal apparatus were uncorrected. H NMR and C
NMR spectra were recorded on a VarianꢀMercury 300 MHz, on a VarianꢀInovaꢀ400 MHz
spectrometer, on a BrukerꢀAspect 3000 consoleꢀ500 MHz spectrometer and chemical shifts are
2
0
ACS Paragon Plus Environment

Page 21 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
19
reported in parts per million (δ). F NMR spectra were recorded by using CFCl
3
as internal
standard. Absolute values of the coupling constant are reported. FTꢀIR spectra were recorded on a
PerkinꢀElmer 681 spectrometer. GC analyses were performed by using a HP1 column (methyl
siloxane; 30 m x 0.32 mm x 0.25 ꢂm film thickness) on a HP 6890 model, Series II. Thinꢀlayer
chromatography (TLC) was performed on silica gel sheets with fluorescent indicator, the spots on
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
the TLC were observed under ultraviolet light or were visualized by I vapour. Chromatography
was conducted by using silica gel 60 with a particle size distribution 40ꢀ63 ꢂm and 230ꢀ400 ASTM.
GCꢀMS analyses were performed on an HP 5995C model and elemental analyses on an Elemental
Analyzer 1106ꢀCarlo Erbaꢀinstrument. MSꢀESI analyses were performed on Agilent 1100 LC/MSD
trap system VL. All synthesized compounds were analyzed by CHN or by HPLC analysis
performed on an Agilent 1260 Infinity instrument equipped with a 1260 DAD VL+ detector, and
their purity is higher than 95%.
Materials. Tetrahydrofuran (THF) from commercial source was purified by distillation (twice)
from sodium wire under nitrogen. Dichloromethane from commercial source was purified by
distillation from CaH under nitrogen atmosphere. DMF from commercial source was purified by
2
distillation from CaH
2
under reduced pressure. Standardized (2.50 M) nꢀbutyllithium in hexanes
was purchased from Aldrich Chemical Co. and its titration was performed with Nꢀpivaloylꢀoꢀ
65
2
toluidine. Diisopropylamine from commercial source was purified by distillation from CaH under
reduced pressure and nitrogen atmosphere. All other chemicals and solvents of commercial grade
were further purified by distillation or crystallization prior to use. Arylnitrile oxides were prepared
from aldehydes through their conversion into the corresponding oximes and then into
31,66
benzohydroximinoyl chlorides.
These were finally converted into nitrile oxides by treatment
.
67
with NEt
3
3
, followed by vacuum filtration of NEt HCl from the solution. Oximes, prepared from
.
reaction of aldehydes/EtOH and NH OH HCl/aq.NaOH, had analytical and spectroscopic data
2
6
8ꢀ71
identical to those previously reported or commercially available.
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
General Procedure for the Synthesis of 5-Alkyl-3,4-diaryl-5-hydroxy-2-isoxazolines 2a-g by
reaction of arylnitrile oxides with enolate ion of 3-aryl-2-propanones 1a-d (Table 1). A solution
of the 3ꢀarylꢀ2ꢀpropanones 1a-d (105mg, 0.558mmol) in THF (3ml) was dropwise added to a
suspension of NaH (95% w/w, 28.20mg, 1.116mmol) in THF (6ml) at 0°C under nitrogen
atmosphere, using a nitrogenꢀflushed, three necked flask equipped with a magnetic stirrer, a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
2
nitrogen inlet and two dropping funnels. After the yellow mixture had been stirred for 1h, a
solution of arylnitrile oxide (0.558mmol) in THF (3ml) was added. The reaction mixture was
allowed to reach room temperature and stirred overnight, and then quenched by adding aqueous
NH Cl solution. The reaction products were extracted three times with ethyl acetate. The organic
4
phase was dried over anhydrous Na SO4 and then evaporated under vacuum. Column
2
chromatography (silica gel, petroleum ether : ethyl acetate= 15/1) of the residue affords the 5ꢀ
Alkylꢀ3,4ꢀdiarylꢀ5ꢀhydroxyꢀ2ꢀisoxazolines 2a-g in 23ꢀ71% yield (Table 2).
3-(3-chlorophenyl)-5-hydroxy-5-methyl-4-phenyl-2-isoxazoline (2a). 71 %. Mp = 124.50ꢀ
128.40°C (Diastereoisomeric mixture). FTꢀIR (KBr): 3600ꢀ3150, 3078, 3026, 3005, 2987, 2940,
1591, 1561, 1490, 1455, 1409, 1383, 1355, 1339, 1265, 1229, 1139, 1094, 1044, 925, 887, 873,
ꢀ1 1
809, 802, 750, 700, 679 cm . H NMR (300 MHz, CDCl , δ): 7.80ꢀ7.77 (m, 1H, aromatic proton,
3
minor isomer); 7.67ꢀ7.64 (m, 1H, aromatic proton, major isomer); 7.43ꢀ7.01 (m, 8H, aromatic
protons for each stereoisomer); 4.50 (s, 1H, major isomer); 4.40 (s, 1H, minor isomer); 3.35ꢀ3.26
(
bs, 1H, OH: exchanges with D O; 1H major isomer); 2.30ꢀ2.10 (bs, 1H, OH: exchanges with D O;
2
2
1
3
1H minor isomer); 1.77 (s, 3H, minor isomer); 1.27 (s, 3H, major isomer). C NMR (75 MHz,
CDCl , δ): 159.90, 134.90, 134.30, 130.90, 130.40, 130.20, 130.10, 130.10, 129.80, 129.60, 129.20,
3
+
1
28.60, 127.40, 127.10, 125.70, 125.40, 63, 60.70, 22.40. GCꢀMS (70 eV) m/z (int.rel.): 287 (M ,
0
.70), 269 (18), 229 (35), 228 (26), 227 (100), 226 (9), 193 (9), 192 (15), 165 (23), 137 (17), 134
(9), 116 (15), 91 (17), 90 (30), 89 (34), 77 (9), 75 (12), 43 (26). Anal. calcd for C H ClNO : C,
1
6
14
2
6
6.79; H, 4.90; N, 4.87. Found: C, 66.70; H, 4.93; N, 4.81.
2
2
ACS Paragon Plus Environment

Page 23 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
-(5-Chlorofuran-2-yl)-5-hydroxy-5-(trifluoromethyl)-4-phenyl-2-isoxazoline (2d). 71 %. FTꢀ
IR (KBr): 3600ꢀ3150, 3037, 2924, 2852, 1616, 1494, 1457, 1324, 1244, 1174, 1086, 1041, 1018,
ꢀ1 1
946, 857, 790, 701 cm . H NMR (300 MHz, CDCl , δ): 7.45ꢀ7.43 (m, 3H, aromatic protons, major
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
isomer); 7.34ꢀ7.32 (m, 3H, aromatic protons, minor isomer); 7.27ꢀ7.24 (m, 2H, aromatic protons for
each stereoisomer); 6.44ꢀ6.43 (d, J= 3.60 Hz, 1H, furylic proton, minor isomer); 6.23ꢀ6.22 (d, J=
3
.60Hz, 1H, furylic proton, major isomer); 6.17ꢀ6.16 (d, J= 3.60Hz, 1H, furylic proton, minor
isomer); 6.14ꢀ6.13 (d, J= 3.60Hz, 1H, furylic proton, major isomer); 4.91 (s, 1H, major isomer);
.78 (s, 1H, minor isomer); 3.42ꢀ3.30 (bs, 2H, OH: exchanges with D O; 1H for each stereoisomer).
, δ): 150.50, 142.60, 140.40, 129.90, 129.70, 129.40, 129.10, 122.30 (q,
19Fꢀ13C = 285 Hz), 116.60, 108.90 (minor diastereoisomer), 108.80 (major diastereoisomer), 103.90
4
2
13
C NMR (75 MHz, CDCl
3
1
J
2
+
(q, J19Fꢀ13C = 33.50 Hz), 57.70. GCꢀMS (70 eV) m/z (int.rel.): 331 (M , 92), 216 (24), 188 (100),
1
5
3
39 (20), 127 (7), 102 (31), 89 (22), 77 (25), 63 (2), 51 (9). Anal. calcd for C14
9 3 3
H ClF NO : C,
0.70; H, 2.74; N, 4.22. Found: C, 50.48; H, 2.60; N, 4.20.
-(5-Chlorofuran-2-yl)-4-(2-fluorophenyl)-5-hydroxy-5-methyl-2-isoxazoline (2e). 56% yield.
Mp 130ꢀ131°C. FTꢀIR (KBr): 3590ꢀ3200, 3131, 2964, 1613, 1582, 1547, 1490, 1458, 1387, 1378,
ꢀ1 1
1
229, 1206, 1145, 1100, 1015, 915, 901, 789, 758 cm . H NMR (300 MHz, CDCl , δ): 7.40ꢀ6.90
3
(
m, 8H, four aromatic protons of each stereoisomer); 6.51 (d, J= 3.6Hz, 1H, cis diastereoisomer);
.17 (d, J= 3.60Hz, 2H, one furyl proton of the cis diastereoisomer and one of the trans); 6.12 (d, J=
3.60Hz, 1H, furyl proton, trans diastereoisomer); 4.84 (s, 2H, isoxazoline protons, 1H of each
diastereoisomer); 3.36 (bs, 1H, OH: exchanges with D O; proton of cis diastereoisomer); 3.15ꢀ2.80
bs, 1H, OH: exchanges with D O, proton of trans diastereoisomer); 1.78 (s, 3H, CH of trans
, δ): 160.50
6
2
(
2
3
1
3
3 3
diastereoisomer); 1.33 (s, 3H, CH of cis diastereoisomer). C NMR (75 MHz, CDCl
1
(d, J1
= 247.40Hz), 152.10, 143.90 (minor diastereoisomer), 143.70 (major diastereoisomer),
9Fꢀ13C
3
1
39.80 (major diastereoisomer), 139.70 (minor diastereoisomer), 131.30, 130.70 (d, J19Fꢀ13C
=
3
8
.90Hz, minor diastereoisomer), 130.40 (d, J19Fꢀ13C = 8.30Hz, major diastereoisomer), 129.20,
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
1
25.20 (d, J19Fꢀ13C = 3.80Hz, major diastereoisomer), 124.80 (minor diastereoisomer), 121.40 (d,
2
2
J
19Fꢀ13C = 14.50Hz, minor diastereoisomer), 116.10 (d, J19Fꢀ13C = 22.00Hz, major diastereoisomer),
2
115.50 (major diastereoisomer), 115.10 (minor diastereoisomer), 108.90 (d, J19Fꢀ13C = 24.70Hz,
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
major diastereoisomer), 108.40 (d, J19Fꢀ13C = 16.90Hz, minor diastereoisomer), 66.20 (minor
diastereoisomer), 54.60 (major diastereoisomer), 25.70 (minor diastereoisomer), 21.40 (major
+
+
diastereoisomer). GCꢀMS (70 eV) m/z (int.rel.): 297 [(M +2), 10], 295 (M , 26), 278 (22), 253
20), 237 (11), 236 (19), 235 (22), 234 (13), 209 (12), 206 (15), 200 (12), 173 (19), 172 (100), 157
(13), 145 (13), 133 (10), 120 (11), 109 (13), 107 (13), 43 (25). Anal. Calc. for C14 : C,
(
H
11ClFNO
2
5
6.87; H, 3.75; N, 4.74. Found: C, 56.84; H, 3.72; N, 4.75.
-(5-Chlorofuran-2-yl)-4-(4-fluorophenyl)-5-hydroxy-5-methyl-2-isoxazoline (2f). 62% yield.
3
Mp 101ꢀ102°C. FTꢀIR (KBr): 3580ꢀ3100, 3050, 2999, 2926, 1602, 1512, 1495, 1442, 1392, 1374,
ꢀ1 1
1318, 1233, 1126, 1014, 945, 917, 898, 828, 786 cm . H NMR (300 MHz, CDCl , δ): 7.30ꢀ7.00
3
(
m, 8H, four aromatic protons for each stereoisomer); 6.47 (d, J= 3.50Hz, 1H, furyl proton, cis
diastereoisomer); 6.15 (d, J= 3.50Hz, 1H, furyl proton, cis diastereoisomer); 6.12 (d, J= 3.50Hz, 1H,
furyl proton, trans diastereoisomer); 6.09 (d, J= 3.50Hz, 1H, furyl proton, trans diastereoisomer);
4.40 (s, 1H, cis isoxazoline proton); 4.36 (s, 1H, trans isoxazoline proton); 1.75 (s, 3H, CH
3
of trans
13
diastereoisomer); 1.28 (s, 3H, CH of cis diastereoisomer). C NMR (125 MHz, CDCl , δ): 167,
3
3
1
1
62.70 (d, J
= 248.10Hz), 152.10, 143.40, 138.50 (major diastereoisomer), 138.40 (minor
9Fꢀ13C
1
3
3
diastereoisomer),131.70 (d, J
= 8.10 Hz, major diastereoisomer), 130.90 (d, J
= 8 Hz,
19Fꢀ13C
19Fꢀ13C
4
2
minor diastereoisomer), 125.30 (d, J19Fꢀ13C = 3.30 Hz, major diastereoisomer), 115.90 (d, J19Fꢀ13C
=
2
2
1
1
1.70 Hz, major diastereoisomer), 115.60 (d, J19Fꢀ13C = 21.40 Hz, minor diastereoisomer), 115,
13.80 (minor diastereoisomer), 113.40 (major diastereoisomer), 108.40 (minor diastereoisomer),
+
07.90 (major diastereoisomer), 62, 49.90, 11.10. GCꢀMS (70 eV) m/z (int.rel.): 297 [(M +2), 7],
+
295 (M , 18), 237 (12), 236 (21), 235 (25), 234 (13), 206 (13), 200 (13), 173 (19), 172 (100), 157
2
4
ACS Paragon Plus Environment

Page 25 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
2
11), 145 (11), 120 (12), 109 (11), 107 (12), 43 (17). Anal. Calc. for C14H11ClFNO : C, 56.87; H,
3
.75; N, 4.74. Found: C, 56.80; H, 3.74; N, 4.71.
5-hydroxy-5-methyl-3-(5-methylfuran-2-yl)-4-phenyl-2-isoxazoline (2g). 23% yield. Mp =
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
8
21.50ꢀ123.40 °C (Diastereoisomeric mixture). FTꢀIR (KBr):3478, 3117, 3084, 3028, 3005, 2989,
956, 2924, 2847, 1613, 1574, 1518, 1496, 1454, 1385, 1237, 1209, 1149, 1083, 1021, 950, 911,
ꢀ1 1
86, 871, 791, 753, 700 cm . H NMR (400 MHz, CDCl , δ): 7.38ꢀ7.26 (m, 3H, aromatic protons
3
for each stereoisomer); 7.25ꢀ7.13 (m, 2H, aromatic protons for each stereoisomer); 6.30 (d, J= 3.30
Hz, 1H, furylic proton, major isomer); 5.96 (d, J= 3.60 Hz, 1H, furylic proton, minor isomer); 5.93
(
d, J= 3.30 Hz, 1H, furylic proton, major isomer); 5.90 (d, J= 3.60 Hz, 1H, furylic proton, minor
isomer); 4.38 (s, 1H, major isomer); 4.33 (s, 1H, minor isomer); 3.62ꢀ3.52 (bs, 1H, OH: exchanges
with D O); 2.28 (s, 3H protons for each stereoisomer); 1.74 (s, 3H, minor isomer); 1.26 (s, 3H,
major isomer). CꢀNMR (100 MHz, CDCl3, δ): 155.10 (major isomer), 155 (minor isomer), 153
2
13
(
major isomer), 152.06 (minor isomer), 142.60 (minor isomer), 142.50 (major isomer), 134.70
major isomer), 132.20 (minor isomer), 129.30, 129, 128.90, 128.50, 128, 115 (major isomer),
(
1
14.90 (minor isomer), 108.70, 107.90, 107.72, 107.71, 63.20 (major isomer), 61.10 (minor
isomer), 25.70 (minor isomer), 21.80 (major isomer), 13.70 (major isomer), 13.68 (minor isomer).
+
GCꢀMS (70 eV) m/z (int.rel.): 257 (M , 23), 240 (18), 239 (38), 198 (28), 197 (100), 196 (26), 182
(8), 171 (9), 168 (9), 155 (7), 154 (22), 115 (8), 102 (41), 89 (9), 43 (131). Anal. calcd for
C H NO : C, 70.02; H, 5.88; N, 5.44. Found: C:69.98, H: 5.89, N: 5.46.
15
15
3
General Procedure for the Synthesis of Isoxazoles 3a-g from 5-Alkyl-3,4-diaryl-5-hydroxy-2-
67
2 3
isoxazolines 2a-g. A solution of Na CO (70 mg, 0.568mmol) in water (3ml) was added to a
solution of the 5ꢀalkylꢀ3,4ꢀdiarylꢀ5ꢀhydroxyꢀ2ꢀisoxazoline 2a-g (94mg, 0.284mmol) in THF (10ml).
The reaction mixture was then heated under reflux (reaction time reported in Scheme 1). The two
phases were separated and aqueous layer was extracted three times with ethyl acetate. The
2 4
combined organic extracts were dried over anhydrous Na SO and the solvent evaporated under
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
reduced pressure. Column chromatography (silica gel, petroleum ether : ethyl acetate= 19/1) of the
residue affords the 5ꢀalkylꢀ3,4ꢀdiarylisoxazoles in 42ꢀ90% yield (Table 2).
3-(3-chlorophenyl)-5-methyl-4-phenylisoxazole (3a). 72% yield (100mg). Mp 85ꢀ88°C (hexane),
white powder. FTꢀIR (KBr): 3079, 3065, 3028, 2920, 1633, 1598, 1570, 1494, 1453, 1426, 1397,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
1
370, 1299, 1234, 1178, 1132, 1091, 1074, 1042, 1018, 975, 914, 897, 790, 757, 688 cmꢀ . H
NMR (300 MHz, CDCl , δ): 7.52ꢀ7.48 (m, 1H, aromatic protons); 7.44ꢀ7.30 (m, 4H, aromatic
protons); 7.28ꢀ7.14 (m, 4H, aromatic protons); 2.45 (s, 3H). C NMR (75 MHz, CDCl
60.20, 134.70, 131.10, 130.10, 123, 129.90, 129.70, 129.10, 128.60, 128.10, 126.80, 116, 11.80.
3
1
3
3
, δ): 167.20,
1
+
+
+
GCꢀMS (70 eV) m/z (int.rel.): 271 [(M + 2), 37], 270 [(M + 1), 21], 269 (M , 9), 256 (9), 254
25), 229 (35), 228 (23), 227 (97), 226 (20), 191 (9), 190 (11), 165 (21), 103 (14), 89 (41), 77 (8),
(
75 (11), 63 (12), 43 (19). Anal. calcd for C16
H, 4.60; N, 5.20.
H12ClNO: C, 71.25; H, 4.48; N, 5.19. Found: C, 70.98;
3
-(2-chlorophenyl)-5-methyl-4-phenylisoxazole (3b). Yellow solid. FT-IR (KBr): 3431, 3070,
624, 1599, 1491, 1435, 1413, 1312, 1234, 1125, 1058, 1017, 963, 910, 768, 724, 696, 678, 572,
1
ꢀ1 1
519, 460 cm . H NMR (400 MHz, CDCl
3
, δ): 7.42ꢀ7.25 (m, 7H, aromatic protons); 7.09ꢀ7.06 (m,
13
2
H, aromatic protons); 2.55 (s, 3H). C NMR (100 MHz, CDCl
3
, δ): 165.80, 160.60, 133.80,
1
31.70, 130.70, 130.10, 129.90, 128.80, 128.60, 127.30, 126.70, 117.10, 11.80. GCꢀMS (70 eV) m/z
+
+
(int.rel.): 271 [(M + 2), 32], 269 (M , 93), 254 (28), 229 (34), 228 (21), 227 (100), 226 (21), 192
(9), 191 (9), 190 (14), 165 (19), 103 (9), 89 (41), 75 (7), 63 (8), 43 (9). Anal. calcd for C16
C, 71.25; H, 4.48; N, 5.19. Found: C, 70.74; H, 4.50; N, 5.07.
H12ClNO:
3
-(2-chlorophenyl)-4-(2-fluorophenyl)-5-methylisoxazole (3c). Colourless semiꢀsolid. FTꢀIR
(
KBr): 3059, 3033, 2953, 2924, 2852, 1630, 1607, 1579, 1516, 1492, 1437, 1414, 1376, 1315,
1
1268, 1240, 1212, 1159, 1136, 1105, 1059, 1018, 965, 904, 818, 764, 753, 745, 734, 796, 664 cmꢀ .
1
H NMR (400 MHz, CDCl , δ): 7.44ꢀ7.41 (m, 1H, aromatic proton); 7.35ꢀ7.24 (m, 4H, aromatic
3
13
protons); 7.09ꢀ6.93 (m, 3H, aromatic protons); 2.48 (d, 3H, J = 1.50 Hz). C NMR (100 MHz,
2
6
ACS Paragon Plus Environment

Page 27 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
4
CDCl
3
, δ): 167.30, 159.80 (d, J19Fꢀ13C = 247.50 Hz), 133.50, 131.60, 131.10 (d, J19Fꢀ13C = 3 Hz),
3
3
1
30.70, 129.80, 129.70 (d, J19Fꢀ13C = 8.40 Hz), 128.50, 126.70, 124.10 (d, J19Fꢀ13C = 3.60 Hz),
2
2
5
117.80 (d, J19Fꢀ13C = 15.60 Hz), 115.90 (d, J19Fꢀ13C = 22 Hz), 111.50, 11.90 (d, J19Fꢀ13C = 3.60 Hz).
+
+
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
GCꢀMS (70 eV) m/z (int.rel.): 289 [(M + 2), 24], 287 (M , 69), 272 (13), 247 (34), 246 (20), 245
100), 244 (14), 210 (7), 208 (9), 183 (14), 107 (34), 75 (6), 43 (8). Anal. calcd for C16 11ClFNO:
C, 66.79; H, 3.85; N, 4.87. Found: C, 66.78; H, 4.00; N, 4.90.
(
H
3-(5-Chlorofuran-2-yl)-4-phenyl-5-(trifluoromethyl)isoxazole (3d). 65% yield. Mp 55ꢀ57°C. FTꢀ
IR (neat): 3147, 3064, 2925, 2853, 1608, 1517, 1493, 1446, 1415, 1355, 1315, 1260, 1210, 1196,
ꢀ1
1
1
160, 1148, 1073, 1023, 992, 968, 943, 899, 775, 700 cm . H NMR (300 MHz, CDCl
3
, δ): 7.51ꢀ
7
.48 (m, 3H, aromatic protons); 7.34ꢀ7.32 (m, 2H, aromatic protons); 6.17 (d, J = 3.50Hz, 1H, furyl
13
2
proton); 6.15 (d, J = 3.50Hz, 1H, furyl proton C NMR (75 MHz, CDCl , δ): 155 (q, J
=
3
19Fꢀ13C
1
4
1
0.40Hz), 153.60, 141.80, 139.90, 129.90, 129.10, 125.90, 118.30 (q, J19Fꢀ13C = 270Hz), 119.40,
+
+
15.50, 108.50. GCꢀMS (70 eV) m/z (int.rel.): 315 [(M + 2), 37], 313 (M , 100), 250 (13), 244
(15), 216 (22), 190 (23), 189 (11), 188 (64), 153 (22), 152 (19), 129 (20), 127 (12), 89 (20), 77 (6),
73 (10). Anal. Calc. for C14 ClF NO : C, 53.61; H, 2.45; N, 4.46. Found: C, 53.60; H, 2.47; N,
H
7
3
2
4
3
3
.41.
-(5-Chlorofuran-2-yl)-4-(2-fluorophenyl)-5-methylisoxazole (3e). 90% yield. FTꢀIR (neat):
144, 3067, 2928, 2854, 1639, 1604, 1580, 1521, 1495, 1455, 1436, 1417, 1208, 1137, 1108, 1021,
ꢀ1 1
989, 943, 923, 899, 789, 763 cm . H NMR (400 MHz, CDCl
3
, δ): 7.43ꢀ7.37 (m, 1H, aromatic
proton); 7.27ꢀ7.14 (m, 3H, aromatic protons); 6.31 (d, J= 3.60 Hz, 1H, furyl proton); 6.13 (d, J=
13
3
.60 Hz, 1H, furyl proton); 2.34 (s, 3H, CH
3
). C NMR (100 MHz, CDCl
3
, δ): 167.80, 160.20 (d,
1
4
3
J19Fꢀ13C=247.9 Hz), 152.40, 143.30, 138.30, 132.10 (d, J19Fꢀ13C = 2.70 Hz), 130.60 (d, J
=8
19Fꢀ13C
3
2
2
Hz), 124.30 (d, J19Fꢀ13C =3.80 Hz), 117 (d, J19Fꢀ13C = 16 Hz), 115.90 (d, J19Fꢀ13C =21.70 Hz), 113,
5
+
1
08.30, 107.90, 11.20 (d, J19Fꢀ13C = 0.80 Hz). GCꢀMS (70 eV) m/z (int.rel.): 279 [(M + 2), 21], 277
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 78
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
+
(
M , 57), 237 (36), 236 (20), 235 (100), 234 (14), 206 (11), 172 (49), 171 (18), 145 (19), 43 (15).
Anal. Calc. for C14 ClFNO : C, 60.56; H, 3.27; N, 5.04. Found: C, 60.60; H, 3.30; N, 5.06.
3-(5-Chlorofuran-2-yl)-4-(4-fluorophenyl)-5-methylisoxazole (3f). 90% yield. Mp 103.70ꢀ
H
9
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
7
05.20°C. FTꢀIR (KBr): 3147, 3077, 2931, 1637, 1595, 1522, 1508, 1445, 1418, 1403, 1236, 1230,
ꢀ1
1
223, 1203, 1158, 1135, 1018, 987, 939, 921, 896, 846, 794 cm . H NMR (500 MHz, CDCl
3
, δ):
.28ꢀ7.26 (m, 2H, aromatic protons); 7.16ꢀ7.13 (m, 2H, aromatic protons); 6.29 (d, J= 3.50Hz, 1H,
1
3
furyl proton); 6.15 (d, J= 3.50Hz, 1H, furyl proton); 2.37 (s, 3H, CH
3
). C NMR (125 MHz,
1
3
CDCl
3
, δ): 167, 162.60 (d, J19Fꢀ13C=247.90 Hz), 152.20, 143.30, 138.50, 131.70 (d, J19Fꢀ13C =8.60
4
2
Hz), 125.30 (d, J
=2.90 Hz), 115.80 (d, J
=21 Hz), 113.80, 113.50, 107.90, 11.17. GCꢀ
1
9Fꢀ13C
19Fꢀ13C
+
+
MS (70 eV) m/z (int.rel.): 279 (M + 2), 16), 277 (M , 49), 237 (35), 236 (19), 235 (100), 234 (12),
206 (12), 172 (46), 171 (20), 170 (11), 147 (16), 145 (20), 89 (13), 43 (30). Anal. Calc. for
14 9 2
C H ClFNO : C, 60.56; H, 3.27; N, 5.04. Found: C, 60.58; H, 3.24; N, 5.01.
5
-methyl-3-(5-methylfuran-2-yl)-4-phenylisoxazole (3g). 65% yield. Yellow oil. FTꢀIR (KBr):
3
058, 2924, 2852, 1627, 1598, 1566, 1498, 1435, 1300, 1239, 1208, 1149, 1074, 1024, 922, 898,
ꢀ
1 1
791, 772, 701 cm . H NMR (400 MHz, CDCl
3
, δ): 7.44ꢀ7.40 (m, 3H, aromatic protons); 7.31ꢀ7.24
(
m, 2H, aromatic protons); 6.12 (d, 1H, J= 3.30 Hz, furylic proton); 5.94ꢀ5.92 (m, 1H, furylic
1
3
proton); 2.36 (s, 3H), 2.32 (s, 3H). C NMR (100 MHz, CDCl , δ): 166.30, 153.90, 153, 142.20,
3
130.10, 130, 128.60, 128.10, 114.10, 112.80, 107.30, 13.60, 11.20. GCꢀMS (70 eV) m/z (int. rel.):
+
239 (M , 48), 198 (15), 197 (100), 196 (23), 168 (5), 167 (6), 154 (7), 129 (7), 127 (5), 115 (5), 89
(
4), 43 (7), 102 (10), 261 (100), 234 (10), 232 (10), 154 (54), 153 (33), 129 (12), 127 (31), 77 (9),
3 (14). Anal. calcd for C15 : C, 75.30; H, 5.48; N, 5.85. Found: C, 75.08; H, 5.60; N, 5.90.
Synthesis of 3-(furan-2-yl)-5-methyl-4-phenylisoxazole (3q) and 5-methyl-4-phenyl-3-
(tetrahydrofuran-2-yl)isoxazole (3r). solution of 3ꢀ(5ꢀChlorofuranꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀ
phenylisoxazole (0.097g, 0.374mmol) in EtOH (2ml), was hydrogenated over Adams catalyst
0.005g, 0.022mmol) at room temperature and atmospheric pressure for 6h in a roundꢀbottom flask
4
H13NO
2
A
(
2
8
ACS Paragon Plus Environment

Page 29 of 78
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
equipped with magnetic stirring. The catalyst was removed by filtration and the solution was
concentrated in vacuo. Column chromatography (silica gel, petroleum ether : ethyl acetate= 12/1) of
the residue affords the 3ꢀ(furanꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀphenylisoxazole (3q) in 89% yield as a white solid
and the 5ꢀmethylꢀ4ꢀphenylꢀ3ꢀ(tetrahydrofuranꢀ2ꢀyl)isoxazole (3r) in 16% yield as a colourless
semiꢀsolid.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
-(furan-2-yl)-5-methyl-4-phenylisoxazole (3q). 89% yield. Mp 93ꢀ96 °C (hexane). White
powder. FTꢀIR (KBr): 3149, 3122, 3059, 2962, 2925, 2853, 1633, 1605, 1511, 1496, 1448, 1438,
ꢀ1
1
1417, 1239, 1225, 1176, 1070, 1020, 976, 923, 891, 782, 714, 700 cm . H NMR (400 MHz,
CDCl , δ): 7.47ꢀ7.41 (m, 4H, aromatic protons); 7.31ꢀ7.28 (m, 2H, aromatic protons); 6.37ꢀ6.35
dd, 1H, J= 3.30 Hz, J= 1.80 Hz, furylic proton); 6.34ꢀ6.33 (d, 1H, J= 3.30 Hz, furylic proton); 2.39
3
(
13
(s, 3H). C NMR (100 MHz, CDCl , δ): 166.60, 153.10, 144, 143.60, 130, 129.80, 128.70, 128.10,
3
+
1
1
1
14.90, 111.50, 111.10, 11.20. GCꢀMS (70 eV) m/z (int.rel.): 225 (M , 40), 184 (14), 183 (100),
54 (17), 127 (13), 102 (12), 43 (11). Anal. calcd for C H NO : C, 74.65; H, 4.92; N, 6.22; O,
14
11
2
4.21. Found: C, 74.48; H, 4.94; N, 6.19.
5-methyl-4-phenyl-3-(tetrahydrofuran-2-yl)isoxazole (3r). 16% yield. Colourless semiꢀsolid. FTꢀ
ꢀ1 1
IR (neat): 3058, 2924, 2852, 1629, 1600, 1500, 1444, 1237, 1057, 1012, 920, 771, 701 cm . H
NMR (500 MHz, CDCl , δ): 7.44ꢀ7.41 (m, 2H, aromatic protons); 7.37ꢀ7.34 (m, 3H, aromatic
protons); 4.97 (t, 1H, J= 6.90 Hz); 3.91ꢀ3.83 (m, 2H); 2.39 (s, 3H); 2.22ꢀ2.08 (m, 2H); 2.05ꢀ1.85
3
13
(m, 2H). C NMR (125 MHz, CDCl
3
, δ): 166.10, 162.60, 130, 129.70, 128.60, 127.60, 116.20,
+
7
2.70, 68.40, 30.20, 26, 11.40. GCꢀMS (70 eV) m/z (int.rel.): 229 (M , 12), 186 (6), 173 (6), 158
(4), 117 (10), 115 (6), 103 (7), 89 (9), 77 (5), 71 (100), 63 (4), 51 (5), 43 (30), 41 (9). Anal. calcd
for C H NO : C, 73.34; H, 6.59; N, 6.11.
14 15
2
Synthesis of 3-(5-bromofuran-2-yl)-5-methyl-4-phenylisoxazole (3s). 3ꢀ(furanꢀ2ꢀyl)ꢀ5ꢀmethylꢀ4ꢀ
phenylisoxazole (3q) (0.152g, 0.670mmol) dissolved in anhydrous DMF (6ml), contained in a
roundꢀbottom flask equipped with magnetic stirring, was cooled to 0°C. NBS (0.144g, 0.810mmol)
2
9
ACS Paragon Plus Environment