Furan-2-ylmethylene thiazolidinediones as novel, potent, and selective inhibitors of phosphoinositide 3-kinase γ

Article abstract of DOI:10.1021/jm0601598

Class I phosphoinositide 3-kinases (PI3Ks), in particular PI3Kγ, have become attractive drug targets for inflammatory and autoimmune diseases. Here, we disclose a novel series of furan-2-ylmethylene thiazolidinediones as selective, ATP-competitive PI3Kγ i

Full text of DOI:10.1021/jm0601598

J. Med. Chem. 2006, 49, 3857-3871
3857
Furan-2-ylmethylene Thiazolidinediones as Novel, Potent, and Selective Inhibitors of
Phosphoinositide 3-Kinase γ
Vincent Pomel,† Jasna Klicic,† David Covini,† Dennis D. Church,† Jeffrey P. Shaw,† Karen Roulin,†
Fabienne Burgat-Charvillon,† Delphine Valognes,† Montserrat Camps,‡ Christian Chabert,‡ Corinne Gillieron,‡ Bernard Franc¸on,§
Dominique Perrin,§ Didier Leroy,§ Denise Gretener,§ Anthony Nichols,§ Pierre Alain Vitte,# Susanna Carboni,#
Christian Rommel,‡ Matthias K. Schwarz,† and Thomas Ru¨ckle*,†
Departments of Chemistry, Signal Transduction, Biochemical Pharmacology, and Experimental Pharmacology, Serono Pharmaceutical
Research Institute, 14 Chemin des Aulx, CH-1228 Plan-les-Ouates, GeneVa, Switzerland
ReceiVed February 13, 2006
Class I phosphoinositide 3-kinases (PI3Ks), in particular PI3Kγ, have become attractive drug targets for
inflammatory and autoimmune diseases. Here, we disclose a novel series of furan-2-ylmethylene
thiazolidinediones as selective, ATP-competitive PI3Kγ inhibitors. Structure-based design and X-ray
crystallography of complexes formed by inhibitors bound to PI3Kγ identified key pharmacophore features
for potency and selectivity. An acidic NH group on the thiazolidinedione moiety and a hydroxy group on
the furan-2-yl-phenyl part of the molecule play crucial roles in binding to PI3K and contribute to class IB
PI3K selectivity. Compound 26 (AS-252424), a potent and selective small-molecule PI3Kγ inhibitor emerging
from these efforts, was further profiled in three different cellular PI3K assays and shown to be selective for
class IB PI3K-mediated cellular effects. Oral administration of 26 in a mouse model of acute peritonitis led
to a significant reduction of leukocyte recruitment.
Introduction
Phosphoinositide 3-kinases (PI3Ks) are pivotal kinases that
participate as lipid, as well as protein, kinases in numerous
intracellular signaling pathways. Class I PI3Ks are subdivided
into class IA and class IB. Class IA PI3Ks contain p110R,
p110â, and p110δ as catalytic subunits, and these are activated
in tyrosine kinase receptor signaling. Class IB PI3Ks contain
only p110γ as a catalytic subunit, which is mostly activated by
seven-transmembrane G-protein-coupled receptors (GPCRs) via
its regulatory subunit p101 and G-protein âγ subunits.1 While
PI3KR and PI3Kâ are ubiquitously expressed, PI3Kγ and PI3Kδ
expression is mainly restricted to the hematopoietic system. All
four catalytic subunits have been investigated by genetic
manipulation. PI3KR and PI3Kâ knockout (KO) mice revealed
Figure 1. Known PI3K inhibitors.12,13
early-stage lethality during embryonic development,2,3 while
mice lacking expression or activity of PI3Kγ and PI3Kδ did
was the first synthetic molecule known to inhibit class I PI3Ks
not exhibit any overt adverse phenotype.4 PI3Kγ KO mice
(see Figure 1). The poor biopharmaceutical properties and lack
showed reduced chemoattractant-induced neutrophil migration
of PI3K isoform selectivity of these first-generation inhibitors
to infection sites and respiratory burst.5-8 Furthermore, mast
have encouraged us, as well as others, to seek PI3Κγ selective
cell degranulation was severely impaired in these genetically
inhibitors with more appropriate profiles.14-29,58 From a struc-
modified mice.9 The aggregate of the genetic studies shows that
tural point of view, the PI3K inhibitors currently in the public
PI3Kγ plays a crucial role in mediating leukocyte chemotaxis
domain can be divided into two classes: derivatives of
as well as mast cell degranulation, making it a potentially
LY29400230,31 and nonrelated structures. All published structures
interesting target for autoimmune and inflammatory diseases.10,11
of inhibitors of the PI3Kγ isoform do not appear to be
At the same time, the genetic studies indicate that PI3K isoform
derivatives of LY294002 (see Figure 2).
selectivity will be a crucial factor determining success or failure
This report discusses the discovery of a new chemical series
of PI3K inhibitor programs on the way to the clinic.
of potent and selective PI3Kγ inhibitors, culminating in the
Several PI3K inhibitors are already known, including natural
development of compound 26. The first two molecular genera-
products such as staurosporine, quercetin, myricetin, and wort-
tions of this series were presented recently.10 This article will
mannin.12 Another widely used inhibitor is LY294002,13 which
describe the follow-up program, which had a primary aim to
improve PI3K class IB selectivity.
* To whom correspondence should be addressed. Address: Serono
The identification of an initial set of furan-2-ylmethylidene
Pharmaceutical Research Institute, Serono International S.A, 14 Chemin
4-oxo-2-thiothiazolidine or 2,4-dioxothiazolidine (also referred
to in the text as rhodanine or thiazolidinedione (TZD), respec-
tively) PI3Kγ inhibitors was achieved using a proprietary
substructure analysis method, as described previously.32,33 Thus,
a set of 59 000 compounds was selected from our corporate
des Aulx, CH-1228 Plan-les-Ouates, Geneva, Switzerland. Phone: +41 22
706 9848. Fax: +41 22 706 9565. E-mail: thomas.ruckle@serono.com.
† Department of Chemistry.
‡ Department of Signal Transduction.
§ Department of Biochemical Pharmacology.
# Department of Experimental Pharmacology.
10.1021/jm0601598 CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/03/2006

3858 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Pomel et al.
Figure 2. PI3Kγ Inhibitors from patent literature.14-29,58
Chart 1. Representative Active Compounds of Initial PI3Kγ High-Throughput Screening
Chart 2. Retrosynthetic Analysis Amenable to Parallel
Synthesis
compound collection and tested in a high-throughput (HT)
PI3Kγ enzyme inhibition assay. This led to the identification
of several inhibitors, all of which exhibited only low-micromolar
activity. Accordingly, the chemical structures of these com-
pounds were subjected to discrete substructure analysis,33 which
led to the identification of a generic substructure as being the
fragment that was most likely to be the basis of the observed
inhibitory activity. This substructure was subsequently used for
in silico screening of commercial databases, which in turn led
to the identification of a number of novel inhibitors whose
structures comprised thiazolidinedione or closely related motifs.
Among these, furan-2-ylmethylidene 4-oxo-2-thiothiazolidines
were particularly prominent (Chart 1), exhibiting IC50 values
in the 0.9-5 µM range.
After identification of compound I, already displaying
promising selectivity against PI3KR, a secondary focused
screening campaign involving rhodanine and thiazolidinedione
derivatives was carried out to further investigate the structure-
activity relationship (SAR). We noticed that rhodanine and
thiazolidinedione derivatives bearing the same peripheral sub-
stitution pattern were equipotent against PI3Kγ, which encour-
aged us to base the synthetic medicinal chemistry program
exclusively on thiazolidinedione derivatives. The objectives were
to (i) increase potency against PI3Kγ, (ii) improve selectivity
versus other PI3K isoforms, and ultimately, (iii) develop an
orally active molecule with an appropriate biopharmaceutical
profile.
lidinedione as structural motif are on the market as insulin-
sensitizers. In consequence, we assessed the binding of our
thiazolidinedione series in the early phase of our drug discovery
program to various PPAR subtypes (R, γ, δ) . Νone of our
inhibitors tested showed significant activity at concentrations
as high as 10 µM.
Results and Discussions
Chemistry. To gain quick access to a range of final molecules
for testing against h-PI3Kγ, a simplified parallel synthesis
strategy was developed consisting of an initial Suzuki cross-
coupling reaction followed by a Knoevenagel reaction, starting
from the appropriate furan or thiophene boronic acid derivatives
(Chart 2).
In the first campaign, special emphasis was given to the
development of a general synthetic route that could be applied
to a diverse set of aryl bromides. To cover the whole array of
chemical reactivity, two distinct Suzuki coupling procedures
were used, as described in procedure A of Scheme 1,38-41 using
toluene/ethanol as solvents, potassium carbonate as base, and
tetrakistriphenylphosphine palladium as catalyst, and in proce-
dure B, using water as solvent, potassium carbonate as base,
and palladium acetate as catalyst.42,43
We have previously shown that the thiazolidinedione core
plays an important role in conferring binding affinity to PI3K.34
This report, together with another recent one,35 describes the
SAR studies around the Western part, i.e., the heteroarylmeth-
ylene moiety, of this class of inhibitors. Thiazolidinediones are
well-established modulators of several biological targets, espe-
cially known as agonists of peroxisome proliferator-activated
receptors (PPARs).36,37 Several drugs containing a thiazo-
To access quickly a large number of closely related structures,
we did not optimize the Suzuki reaction for high yields and we

Furan-2-ylmethylene Thiazolidinediones
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3859
Scheme 1. Parallel Synthesis of Furan-2-ylmethylene Thiazolidinedionesa
a Reagents and conditions: (A) Pd(PPh3)4, K2CO3, toluene/EtOH, ∆; (B) Pd(OAc)2, K2CO3, H2O; (C) thiazolidinedione, â-ala, AcOH, ∆.
Scheme 2. Synthesis of Furan-2-ylmethylene Thiazolidinediones via Protection and Deprotection Stepsa
a Reagents and conditions: (a) benzylbromide, K2CO3, THF, ∆, 68%; (b) furfuralboronic acid, Pd(PPh3)4, K2CO3, toluene/EtOH, ∆, 56%; (c) BCl3,
DCM, -30 °C, 66%; (d) thiazolidinedione, â-ala, AcOH, ∆, 95%.
Table 1. Inhibition of h-PI3Kγ and h-PI3KR: Phenols
developed a short workup protocol consisting of a simple
filtration through a silica gel pad. The resulting intermediate
compounds (1a-41a), having average purities of 85-95%, were
directly used for the next step without further purification.
The Knoevenagel reaction was performed under acidic
conditions using 2,4-thiazolidinedione and â-alanine.44 The
advantages of these conditions were numerous: all starting
materials were readily soluble; the reaction was quick and clean;
the final insoluble compound was precipitated in the course of
the reaction, yielding essentially pure final compound (purities
>90%). At this stage, the products were fully analyzed and
characterized before entering the biological tests.
In parallel, a second synthetic approach was developed
amenable to producing multigram quantities required for the
later-stage experiments in vivo, in which the yield of the cross-
coupling reaction was optimized from 5% to 50-60%. This
second approach was based on the protocol used for the parallel
synthesis approach (Scheme 1, conditions A), but with an
intermediate protection/deprotection sequence of the phenolic
hydroxy group, as shown in Scheme 2.
Intermediate 26a could be accessed over three steps in 25%
yield. The phenol group was first protected by a benzyl group
and the resulting bromobenzyloxy derivative 26c was coupled
to furfuralboronic acid using a palladium catalyst. Intermediate
26b was then debenzylated with boron trichloride at -30 °C to
afford intermediate 26a. The improved yield of the Suzuki
reaction with protected phenols may be explained by the
reduction of the organopalladium intermediate in the presence
of free phenols. The Knoevenagel reaction afforded compound
26 in 95% yield with high purity (96%) requiring no further
purification. Compounds such as 26 that progressed into in vivo
experiments were transformed into the corresponding potassium
salts using KOH in THF/water.
compd
R
methoda
anal. datab
PI3Kγ (nM)c
PI3KR (nM)c
1
2
3
4
H
A
A
A
A
a
a
a
a
20000
nd
2-OH
3-OH
4-OH
30 ( 5
290 ( 50
380 ( 70
330 ( 50
1800 ( 300
3600 ( 700
a Method A: prepared according to Scheme 1, using conditions A.
b Analytical data: (a) NMR, LC/MS, mp, CHN, HPLC >90%. c All values
determined in triplicate. IC50 values in nM. nd: not determined.
an interesting enzymatic selectivity profile, cellular activity was
assessed, as well as complete class I PI3K enzymatic selectivity
(PI3KR,â,γ,δ), followed in some cases by pharmacological
evaluation in vivo.
Early on, we hypothesized that the phenol group was
important in conferring PI3K binding, based on an inspection
of some of the early compounds (such as II and III) and our
previously gained insights into the binding mode of this type
of inhibitors.10,34,35 The first series of compounds was made to
test this hypothesis and to probe how the position of the phenolic
hydroxyl group would affect the activity. The results, sum-
marized in Table 1, confirmed the importance of the phenolic
hydroxy group for PI3K inhibition (1 vs 2). Repositioning the
hydroxy group around the phenyl ring from position 2 to
positions 3 and 4 resulted in a 10-fold reduction of the activity.
Interestingly, the 10-fold selectivity for PI3Kγ versus PI3KR
was retained between the three regioisomers, encouraging us
to consider substitutions in all three positions in the design of
a follow-up library.
Biological Activity. Inhibitory activity against PI3Kγ was
determined by measuring the percentage of inhibition at 1 and
10 µM compound concentration as a primary screen. Com-
pounds with percentage of inhibition values (% INH) higher
than 50% were progressed into full IC50 value determination
and, depending on their activity, tested against PI3KR to
determine the selectivity ratio. For selected inhibitors displaying
Because it became apparent that the interaction of the hydroxy
group, probably via hydrogen bonding to an appropriate residue
in the enzyme, is crucial for PI3Kγ inhibition, we decided to
explore substituents other than hydroxy to probe what effects
the hydrogen-bonding capacity, charge, polarity, and size of a
substituent would have on PI3K binding. The goal was to retain
or improve the inhibitory potency against PI3Kγ and, at the

3860 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Table 2. Inhibition of h-PI3Kγ: 2- and 3-Substituted Phenylfurans
Pomel et al.
Table 3. Inhibition of h-PI3Kγ and h-PI3KR: Heteroaryl-Substituted
Furans
compd
X
R
methoda
anal. datab
PI3Kγ (nM)c
5
6
7
8
9
10
11
12
13
14
15
16
17
18
O
O
O
O
O
O
O
O
O
O
O
S
2-OMe
2-Cl
2-F
A
A
A
A
A
A
A
A
A
A
A
A
A
A
a
a
a
a
a
a
a
a
a
a
b
a
a
a
44% INH
20000
1200 ( 330
4% INH
2-CO2H
2-CO2Me
2-CH2-CO2H
3-CO2H
3-CO2Me
3-CH2-CO2H
2-NHAc
2-NH2-5-NO2
2-OH
2% INH
12% INH
4300 ( 170
13500 ( 1500
15% INH
24% INH
18% INH
27% INH
2600 ( 470
24% INH
O
O
3-OMe
3-NO2
a Method A: prepared according to Scheme 1, conditions A. b Analytical
data: (a) NMR, LC/MS, mp, CHN, HPLC >90%; (b) LC/MS, HPLC
>90%. c All values determined in triplicate. IC50 values in nM. % INH at
1 µM (INH ) inhibition).
same time, to maximize the selectivity against the other PI3K
isoforms. The results are summarized in Table 2.
Replacement of the 2-hydroxy group with small, polar
substituents with limited hydrogen-bond acceptor potential led
to significant reduction in activity (5-7). Repositioning the
methoxy group to position 3 did not lead to any improvement
in terms of activity (17). Similarly, substitution with larger polar
or charged groups (as in 8-10) caused a marked loss in activity.
The same set of substituents in position 3 produced compounds
with similarly low activities (11-13), as did the introduction
of a nitro group at the same position (18). Interestingly, when
the 2-hydroxy group was replaced by hydrogen-bond-donating
(but not hydrogen-bond-accepting) NH groups at the same
position (14, 15), no significant inhibition of PI3Kγ activity
was observed at 1 µM. From the above data, we inferred that
a small group capable of acting as a hydrogen-bond acceptor
and donor in position 2, and, to a lower extent, also in positions
3 and 4 is crucial for a strong interaction with the PI3Kγ binding
site. The importance of geometry and size is demonstrated by
the binding data of compound 16, where the central furan ring
is replaced by a thiophene, shifting the 2-hydroxy group away
from its optimal position in the protein-binding pocket (see
section Docking Studies and X-ray Crystallography). None of
the compounds shown in Table 2 (5-18) exhibited any
inhibitory activity against PI3KR at concentrations up to 20 µM.
Next, we explored a set of mono- and bicyclic heterocycles
containing oxygen or nitrogen atoms capable of functioning as
a hydrogen-bond acceptor, the results of which are summarized
in Table 3. In the case of oxygen-containing heterocycles, both
piperonyl and furan derivatives (19-21) showed reduced activity
probably because of steric hindrance and/or suboptimal posi-
tioning of the hydrogen-bond-accepting oxygen. On the other
hand, the 2- and 3-pyridine analogues (22 and 23, respectively)
maintained activities in the same range as the respective phenol
compounds (2 and 3, respectively), with a mere 3-fold loss in
IC50 values (30 vs 110 nM; 290 vs 950 nM). Nonetheless, the
total lack of selectivity observed for both 2- and 3-pyridines
suggests a special role of the hydroxy group in differential
binding to PI3Kγ and PI3KR isoforms. However, introduction
of larger pyridine analogues, such as 2-quinoline and 8-quinoline
(24 and 25, respectively), was no longer tolerated by the enzyme.
The reduced activity may possibly be explained by increased
a Method A: prepared according to Scheme 1, condition A. Method B:
prepared according to Scheme 1, condition B. b Analytical data: (a) NMR,
LC/MS, mp, CHN, HPLC >90%; (b) LC/MS, HPLC >90%. c All values
determined in triplicate. IC50 values in nM. % INH at 1 µM (INH )
inhibition). nd: not determined.
steric hindrance and/or by a poor positioning of the heterocyclic
nitrogen atom, preventing proper hydrogen bonding with the
enzyme.
Having thus established the importance of the 2-hydroxy
group for both potency and isoform selectivity, we set out to
explore the effects of additional substituents on those two key
parameters (see Table 4).
The introduction of a second substituent was found to have
little influence on the inhibition of PI3Kγ (all IC50 values in
the range 20-80 nM), with the only exception being the
carboxylate group in position 5 (compound 36), which signifi-
cantly reduced the activity. One explanation offered by the
inspection of the cocrystal structure argues that the carboxylate
group is positioned close to the Lys890 side chain, with which
it can form an ionic interaction. This stabilization of the negative
charge will increase the polarization of the π electron system
and lower the electron density of the 2-hydroxy through the
long-range resonance effect, thus reducing its effectiveness as
a hydrogen-bond acceptor. Furthermore, because of the interac-
tion between Lys890 and the carboxylate, 36 may slightly be
shifted, potentially resulting in weaker binding (for more details,
consult section Docking Studies and X-ray Crystallography).
In contrast to the potency against the PI3Kγ isoform, the
PI3Kγ/R selectivity ratio was found to depend considerably on
the nature and the position of the second substituents, ranging
from 13-fold (for 33) to 30-fold (for 26). Among the compounds
with markedly reduced activity against PI3KR (IC50 > 1 µM)
are compounds 35 and 36. It is interesting to note that,

Furan-2-ylmethylene Thiazolidinediones
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3861
Table 4. Inhibition of h-PI3Kγ and h-PI3KR: Substituted
2-Furan-2-ylphenols
Table 5. Inhibition of h-PI3Kγ and h-PI3KR: Fluorine Substituted
2-Furan-2-ylphenols
anal.
PI3Kγ
PI3KR
anal.
PI3Kγ
PI3KR
compd
X
R
methoda datab
(nM)c
(nM)c
compd
R
methoda datab
(nM)c
(nM)c
pKa
d
26
37
38
39
40
41
O
O
S
O
O
O
2-OH, 4-F
2-OMe, 4-F
2-OH, 4-F
4-OH, 2-F
2F, 4-F
C
A
A
B
A
B
a
a
a
a
a
a
33 ( 10
11% INH
4% INH
260 ( 30
935 ( 150
nd
nd
2
26
27
28
29
30
31
32
33
34
35
36
4-H
4-F
4-OH
4-CO2Me
5-F
5-Cl
5-CN
5-OCF3
5-Me
5-OH
3-NO2
5-CO2H
A
C
B
B
B
B
B
B
B
B
B
B
a
a
a
a
b
a
a
a
a
a
b
b
30 ( 5
33 ( 10
20 ( 8
29 ( 5
37 ( 10
43 ( 5
40 ( 10
56 ( 20
39 ( 6
20 ( 3
80 ( 20
330 ( 50
935 ( 150
190 ( 20
210 ( 25
455 ( 50
545 ( 75
220 ( 40
600 ( 40
500 ( 70
100 ( 10
9.5 ( 0.1
8.4 ( 0.1
8.7 ( 0.1
8.7 ( 0.1
9.1 ( 0.1
8.9 ( 0.1
7.3 ( 0.1
8.7 ( 0.1
9.7 ( 0.1
10e
1000 ( 70
3800 ( 420 nd
29 ( 3
2-OH, 3-F 5-F
340 ( 25
a Method A: prepared according to Scheme 1, condition A. Method B:
prepared according to Scheme 1, condition B. Method C: prepared
according to Scheme 2. b QC/analytical data: (a) NMR, LC/MS, mp, CHN,
HPLC >90%. c All values determined in triplicate. IC50 values in nM. %
INH at 1 µM (INH ) inhibition). nd: not determined.
1250 ( 250 6.6e
530 ( 30 1350 ( 150 8.6 ( 0.1
a Method A: prepared according to Scheme 1, condition A. Method B:
prepared according to Scheme 1, condition B. Method C: prepared
according to Scheme 2. b QC/analytical data. (a) NMR, LC/MS, mp, CHN,
HPLC >90%. (b) LC/MS, HPLC >90%. c All values determined in
triplicate. IC50 values in nM. % INH at 1 µM (INH ) inhibition). nd: not
determined. d Experimental pKa values. e Predicted pKa values (ACD soft-
ware,version 7.0).
as a hydrogen-bond donor and hydrogen-bond acceptor in the
present molecular context and/or that space is limited to
accommodate only a hydrogen group. Substitution of the
2-hydroxy group by fluorine produced a weak inhibitor, 40 (IC50
) 4 µM), with activity comparable to activities of other
compounds lacking the phenolic hydroxy group (see Table 2).
Switching the positions of the hydroxy group and the fluorine
atom resulted in a compound with similar activity (39) as the
corresponding nonfluorinated analogue 4 (see Table 1). Re-
placement of the 4-fluoro by a 3,5-difluoro substitution pattern
(compounds 26 and 41) maintained the activity against PI3Kγ
but decreased the selectivity against PI3KR. Together with a
similar result obtained from the 5-fluoro analogue 29 (see Table
4), this indicates that the beneficial influence of the 4-fluoro
substituent on the isoform selectivity is probably not only due
to σ-inductive effects but must depend on additional factors.
Of note, replacement of the central furan by a thiophene
produced an inactive compound (38), as previously seen for
compounds 2 and 16 (see Table 2 and section Docking Studies
and X-ray Crystallography).
considering that Lys890 in PI3Kγ is replaced by neutral Gln859
in PI3KR and hence has no possibility for ionic interaction of
the 5-carboxylate in 36, the adverse effect of this group on the
hydrogen-bonding capability of 2-hydroxy is somewhat dimin-
ished, resulting in the lowest R/γ selectivity among all 2-hydroxy
analogues tested. Another compound showing reduced potency
on PI3KR (but constant activity against PI3Kγ) is the 4-fluoro
derivative 26. Detailed inspection reveals that the replacement
of the hydroxy group by a fluoro atom in either the 4- or the
5-position (27 vs 26; 34 vs 29) leads to reduced activity on
PI3KR. A possible reason could be reduced fluorine hydration
of these solvent-exposed groups. However, while this effect is
less prominent for the more tightly binding PI3Kγ isoform
(roughly 2-fold), it leads to a 5-fold difference in binding to
the less optimal ATPa binding site of PI3KR, resulting overall
in an increased PI3Kγ/PI3KR selectivity.
In summary, the SAR information obtained from the com-
pounds presented here indicates that the 2-hydroxy group plays
a crucial role in conferring PI3K binding and inhibition,
probably through a bidentate HBA/HBD function, leading to
PI3Kγ inhibitory activities in the range 30-50 nM. Addition
of other substituents around the 2-hydroxyphenyl moiety has
no major influence on the inhibitory potency. However,
introduction of a fluorine atom specifically at the 4-position
leads to a notable increase in selectivity against PI3KR.
Modeling studies suggest that the fluorine atom at C-4 is located
in the proximity of a nonconserved area of the binding site,
where the nonpolar Ala885 (present in PI3Kγ) is replaced by
the polar Ser606 (present in PI3KR). This could account for
the increased isoform selectivity. Whether this hypothesis holds
up and hence enables medicinal chemists to design even more
selective PI3Kγ inhibitors in a rational way remains an open
question. An alternative hypothesis argues that the plasticity of
the class IA and class IB PI3K isoforms is different, in which
case the shape of the ATP-binding site would matter less than
flexibility and adaptability upon inhibitor binding. This would
render rational design of selective PI3Kγ inhibitors more
complex. Further work is required to shed more light on the
key molecular interactions between PI3Ks and small-molecule
inhibitors that determine isoform selectivity.
In an attempt to rationalize the influence of additional
substituents in the series of compounds shown in Table 4, we
asked whether the inhibitory potency against PI3Kγ and/or the
PI3KR/PI3Kγ selectivity would depend on the acidity of the
phenolic hydroxy group, the underlying hypothesis being that
with increasing acidity, the electron density on the oxygen atom
will decrease, hence weakening its hydrogen-bond-forming
capacity. The panel of 4- and 5-substituted 2-phenol derivatives
presented in Table 4 covers a pKa range from 7 to 10 for the
phenolic hydroxy group. Although we could observe a weak
tentative correlation between PI3Kγ inhibition (log IC50) and
pKa values (R2 ) 0.6), no significant correlation was observed
for PI3KR and hence the selectivity ratio.
Finally, a highly focused set of analogues of compound 26
was prepared, the results of which are summarized in Table 5.
Corroborating previous results, replacing the 2-hydroxy by
a methoxy group, while maintaining the fluorine in the 4-posi-
tion (compound 37), led to a significant reduction of the activity
against PI3Kγ. This confirms that the hydroxy group functions
a Abbreviations: ATP, adenosine triphosphate; HBD, hydrogen-bond
donor; HBA, hydrogen-bond acceptor; PKB, protein kinase B; SCF, stem
cell factor; MCP-1, monocyte chemotactic protein-1; CCR, chemokine
receptor; CSF-1, colony stimulating factor 1.

3862 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Pomel et al.
Table 6. Class I PI3K Selectivity Profile for Compounds 26 and 2 in
Comparison with Wortmannin and LY294002
of the other class IA PI3K members was only checked for a
few key compounds. As an illustration, Table 6 shows the entire
class I PI3K selectivity profile for compounds 2 and 26,
emphasizing the excellent isoform selectivity of the latter
compound. In addition to the selectivity profile for PI3Kγ vs
class IA PI3Ks, another important feature of this class of
compounds was found to be their excellent selectivity profile
against a large set of unrelated protein kinases. As an example,
compound 26 was screened against 80 different Ser/Thr and
Tyr kinases and shown not to significantly inhibit any of them
at 10 µM except for casein kinase 2 (see Table 7).
IC50 (nM)a
PI3K
R
â
γ
δ
Wortmannin
LY294002
2
26
1 ( 0.2
720 ( 200
330 ( 50
940 ( 150
10 ( 2
5 ( 1
7260 ( 2600
30 ( 5
9 ( 2
306 ( 40
1330 ( 260
20000
20000
> 5000
20000
30 ( 10
a All values determined at least in triplicate.
Table 7. Selectivity Profile of Compound 26 against 80 Protein
Kinasesa
remaining
enzymatic
remaining
Docking Studies and X-ray Crystallography. To better
understand the interactions of furan thiazolidinedione PI3Kγ
inhibitors with the target protein, we determined the crystal-
lographic structure of complexes between PI3Kγ and several
representative ligands at resolutions between 2.7 and 2.9 Å. The
unequivocal electron density at the ATP-binding site of the
protein clearly identified the binding mode of these furan
thiazolidinediones (Figure 3). In the case of 26 (Figures 3 and
4), the thiazolidinedione nitrogen, which is expected to be
deprotonated, forms a strong salt-bridge interaction with the
positively charged Lys833 side chain (2.5 Å), which itself also
forms another salt bridge with the side chain of Asp964 (2.8
Å). The 2-hydroxy group on the phenol ring acts as an important
hydrogen-bond acceptor by interacting with the backbone NH
of Val882 (2.5 Å), while the phenyl and furan rings form
hydrophobic interactions with several hydrophobic residues in
the active site.
enzymatic
activity of 26
activity of 26
protein kinase
Abl(h)
AMPK(r)
Arg(m)
Aurora-A(h)
Axl(h)
Blk(m)
(% of control) protein kinase (% of control)
92
90
85
85
74
70
102
97
111
80
84
93
117
83
89
77
94
82
88
MEK1(h)
MKK4(m)
MKK6(h)
MKK7â(h)
MSK1(h)
p70S6K(h)
PAK2(h)
PDGFRR(h)
PDGFRâ(h)
PDK1(h)
PI3Kγ(h)
PKA(h)
PKBR(h)
PKBâ(h)
PKBγ(h)
PKCR(h)
PKCâII(h)
PKCγ(h)
PKCδ(h)
PKCꢀ(h)
PKCη(h)
PKCι(h)
PKCµ(h)
PKCθ(h)
PKD2(h)
PRAK(h)
PRK2(h)
ROCK-II(h)
Rsk1(h)
Rsk2(h)
Rsk3(h)
SAPK2a(h)
SAPK2b(h)
SAPK3(h)
SAPK4(h)
SGK(h)
96
85
105
2 (>10 µMb)
72
190
178
85
Bmx(h)
CaMKII(r)
CaMKIV(h)
CDK1/cyclinB(h)
CDK2/cyclinA(h)
CDK2/cyclinE(h)
CDK3/cyclinE(h)
CDK5/p35(h)
CDK6/cyclinD3(h)
CDK7/cyclinH/MAT1(h)
CHK1(h)
CHK2(h)
CK1(y)
CK2(h)
c-RAF(h)
96
104
5 (20 nMb)
89
97
103
96
98
98
107
109
93
91
84
98
102
91
95
95
90
104
86
6 (20 nMb)
99
The crystallographic structure of the complex between PI3Kγ
and 2 (Figure 5a) was used to carry out molecular modeling by
docking, minimization, and induced-fit docking, which provided
additional information concerning the interaction of analogues
from this series with PI3Kγ and rationalizing the structure-
activity relationship.
CSK(h)
cSRC(h)
Fes(h)
FGFR3(h)
Flt3(h)
103
66
90
121
93
91
66
84
96
82
89
90
97
97
Fyn(h)
GSK3R(h)
GSK3â(h)
IGF-1R(h)
IKKR(h)
IKKâ(h)
IR(h)
JNK1R1(h)
JNK2R2(h)
JNK3(h)
Lck(h)
Lyn(h)
As an example, we present here a model of 4 bound to PI3Kγ,
which was generated via induced-fit docking (Figure 5b). This
protocol includes sampling of the conformation space of amino
acid side chains surrounding the bound inhibitor before redock-
ing of the inhibitor to the newly generated protein conformation.
Molecule 4 adopts the same binding mode as 2, forming an
ionic interaction with Lys833 and a lipophilic interaction
between benzene and furan rings and the cavity lined by
hydrophobic residues Trp812, Ile831, Ile881, Met953, and
Phe961. However, the different position of the hydroxy group
prevents it from interacting with the buried Val882 backbone
NH. Instead, a hydrogen bond with the Ala885 carbonyl oxygen
can be formed. Given that both groups are solvent-exposed, the
73
84
89
93
93
98
93
98
86
104
101
103
117
Syk(h)
TrkB(h)
Yes(h)
ZAP-70(h)
MAPK1(h)
MAPKAP-K2(h)
99
159
a Protein kinases were assayed with 10 µM of 26 in the presence of 10
µM of ATP. b IC50 value on the enzyme.
Throughout the entire project, the PI3Kγ/PI3KR selectivity
was routinely monitored for all compounds, while the inhibition
Figure 3. (a) Crystallographic structure of PI3Kγ with the calculated surface electrostatic potential (positive charge shown in blue and negative
charge in red), showing the location and size of the binding site of 26. (b) Close-up of the structure of the active site and binding pocket of 26.

Furan-2-ylmethylene Thiazolidinediones
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3863
Figure 4. Stereodiagram depicting the crystallographic structure of 26 bound to the active site of human PI3Kγ.
Figure 5. Stereodiagram depicting the crystallographic structure of 2 bound to the active site of human PI3Kγ (a) and stereodiagrams depicting
the active site of PI3Kγ with the optimized docked pose of 4 (b) and 16 (c).
stabilizing contribution of this interaction is reduced in com-
parison with a hydrogen bond to Val882, resulting in a roughly
10-fold drop in potency relative to 2.
Another example is the model of the thiophene analogue 16
bound to PI3Kγ, which was generated by additional geometry
minimization of the docked structure (Figure 5c). Compound

3864 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Table 8. Cellular Activity in Raw-264 Murine Macrophages
Pomel et al.
C5a-mediated PKB/Akt
phosphorylation in
compd
PI3Kγ (nM)a
Raw-264 macrophages (µM)b
LY294002
7260 ( 2600
30 ( 10
30 ( 5
10.0 ( 0.5
26
2
0.23 ( 0.08
0.81 ( 0.25
3.65 ( 1.75
3.49 ( 0.09
1.71 ( 0.38
1.39 ( 0.26
0.72 ( 0.25
0.88 ( 0.04
1.06 ( 0.45
22
27
29
41
28
32
33
110 ( 20
20 ( 8
37 ( 10
29 ( 3
29 ( 5
56 ( 20
39 ( 6
Figure 6. Inhibition of C5a-induced PKB/Akt phosphorylation. After
starvation and 30 min of pretreatment with the indicated concentrations
of 26 or LY294002, Raw-264 macrophages were activated with 50
nM of C5a for 5 min. PKB/Akt phosphorylation was determined using
a phospho-Ser-473 Akt specific antibody, and ELISA protocols, as
described in the section Biological Methods. The graph represents the
mean ( standard deviation of four independent experiments done in
triplicate.
a IC50 values on the isolated enzyme.54 b All values determined at least
in triplicate.
16 is almost 1 Å longer along its phenol thiazolidinedione axis
and has a more linear shape than the corresponding furane
compound 2. Compared to the binding mode of 2, 16 is pivoted
around the phenol ring and moved toward the outside of the
binding pocket in order to retain the hydrogen bond between
2-hydroxy and Val882 NH. As a result, the thiazolidinedione
group is displaced by 5 Å and it forms weakened Coulomb
interactions within the phosphate-binding site. In addition, the
thiophene ring is also significantly displaced from the hydro-
phobic binding pocket, leading to suboptimal interactions. This
example illustrates how the increase in the molecular size when
going from furan to thiophene can have as a consequence a
drastic reduction in potency.
Cell-Signaling Effects. We further investigated the cellular
potency, selectivity, and functionality of our PI3Kγ inhibitors,
using cellular assays previously described.10 In contrast to the
class IA PI3KR, PI3Kâ, and PI3Kδ isoforms, class IB PI3Kγ
signals in response to the activation of GRi-protein-coupled
receptors (GPCRs), including chemokine receptors or receptors
for the chemoattractants fMLP or complement 5a (C5a), leading
to the generation of phosphatidylinositol-3,4,5-triphosphate
(PtdIns(3,4,5)P3) and consequent phosphorylation of PKB/
Akt.6,45 On the basis of, we investigated the effect of our
compounds on C5a-mediated PKB/Akt phosphorylation in Raw-
264 murine macrophages. This assay was considered a primary
cellular filter and allowed us to assess the potency of this
inhibitor class on inhibition of PI3Kγ in cells. Compounds
inhibited C5a-mediated PKB/Akt phosphorylation in a concen-
tration-dependent manner with submicromolar or low-micro-
molar IC50 values (Table 8). As an illustration, Figure 6 shows
a dose response curve obtained for compound 26 in comparison
to LY294002.
To validate the selective inhibition of PI3Kγ, as opposed to
unselective effects on PI3Kδ or PI3KR/PI3Kâ activities at a
cellular level, we further selected the most potent and selective
inhibitors to compare their effects on cellular readouts specific
to PI3Kγ, PI3Kδ, or PI3KR/PI3Kâ signaling. As described,
class IA PI3Ks signal downstream of cytokine and growth factor
receptors. Activation of primary mast cells with SCF, the ligand
for the receptor tyrosine kinase c-kit, is mediated exclusively
by PI3Kδ, and in mast cells isolated from p110dD910A/D910A
mutant mice, SCF-induced PKB/Akt phosphorylation is com-
pletely abrogated.46 Encouragingly and in accordance with its
in vitro selectivity, compound 26 had no effect on SCF-induced
phosphorylation of PKB/Akt in bone marrow derived mast cells,
even at concentrations as high as 10 µM, confirming its lack of
activity on the PI3Kδ isoform (Figure 7, full symbols). As a
control, we used a reference compound derived from a pyri-
midinones series as described previously that selectively inhibits
Figure 7. Lack of inhibition of SCF-induced PKB/Akt phosphorylation
in mast cells by 26. After pretreatment with the indicated concentrations
of compound, primary mast cells derived from wild-type bone marrow
were activated with SCF as described in the section Biological Methods,
and PKB/Akt phosphorylation was determined by FACS using phospho-
Ser-473 Akt specific antibodies: (closed symbols) 26; (open symbols)
PI3Kδ selective reference.57
PI3Kδ (Figure 7, empty symbols).57 LY294002 abrogated
completely SCF-induced Akt phosphorylation at concentrations
as high as 20 µM.
To further assess the selectivity of those compounds for class
IB PI3Kγ, we examined the capability of 26 and 2 to inhibit
preferentially chemokine-induced (PI3Kγ-mediated) but not
cytokine-induced (class IA PI3K-mediated) PKB/Akt phospho-
rylation in the human monocytic cell line THP-1.10 In those
cells, MCP-1, binding to the GPCR chemokine receptor CCR2,
strongly induced phosphorylation of PKB/Akt, which was
effectively inhibited by 26 at IC50 values as low as 0.4 µM
(Figure 8). In contrast, induction of PKB/Akt phosphorylation
by colony stimulating factor (CSF-1), binding to the growth
factor receptor c-fms, was only blocked by 26 at IC50 values as
high as 4.7 µM (Figure 8). These data are consistent with the
in vitro selectivity profile of this inhibitor class, and in particular
for compound 26. None of the inhibitors tested showed any
effect on cell viability. LY294002, as an example of a
nonselective class I PI3K inhibitor, did not display, as expected,
any preferential inhibition in cells.
Cell-Functional Effects. Since chemokine-induced migration
of neutrophils, monocytes, and macrophages is, at least in part,
a PI3Kγ-dependent mechanism,5,6,47-49 the effects of blocking
PI3Kγ activity on MCP-1-induced chemotaxis in the THP-1
monocytic cell line and in primary mouse monocytes were
examined. Compound 26 inhibited MCP-1-mediated chemotaxis
in wild-type primary monocytes in a concentration-dependent

Furan-2-ylmethylene Thiazolidinediones
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3865
Table 9. Pharmacokinetic Parameters of 26 after Single Dose of 10
mg/kg in Rats
CL
AUC
Cmax
t1/2
(h)
Vd
(L/kg)
route
(L kg-1 h-1
)
(h ng mL-1
)
(ng mL-1
)
iv
po
2.27
4401
847
11399
309
0.65
1
25.3
followed by monocytes/macrophages and lymphocytes, recruit-
ment of neutrophils was monitored 4 h after thioglycollate
injection.50
Oral administration of 26 at 10 mg/kg resulted in moderate
reduction of neutrophil recruitment (35% ( 14%), almost
matching the result observed in PI3Kγ-deficient mice.10 Given
the short oral half-life of 26 (t1/2 ) 1 h) and relative high
clearance (2.25 L kg-1 h-1), investigations at later time points
(24-48 h) to assess macrophage and monoycyte recruitment
were not undertaken (see Table 9). The modest pharmacokinetic
properties did not appear to be caused by rapid oxidative
metabolism (microsomal metabolism after 1 h: 16% (rat), 10%
(human)). We suspect that the phenolic hydroxy group might
be subject to glucoronidation and hence that compound 26 might
be cleared via nonoxidative metabolism. Current efforts are
focusing on generating analogues (including prodrugs) with
more suitable pharmacokinetic profiles.
Figure 8. Preferential inhibition of MCP-1- over CSF-1-induced PKP/
Akt phosphorylation in cells. After starvation and 20 min of pretreat-
ment with the indicated concentrations of 26 (squares) or LY294002
(circles), THP-1 monocytes were activated by MCP-1 (closed symbols)
or CSF-1 (open symbols) and PKB/Akt phosphorylation was monitored
by Western blot using phospho-Ser-473 and wild-type specific Akt
antibodies and quantified as described in the section Biological Methods.
Conclusions
We have designed a new chemical series of potent PI3Kγ
inhibitors. Through the use of X-ray crystallography and iterative
docking experiments, the key molecular features of this series
necessary for inhibitory activity against PI3Kγ have been
identified to be the NH group of the thiazolidinedione core and
the 2-hydroxy group on the phenyl ring. Introduction of an
additional 4-fluoro substituent generated a key compound (26)
with increased selectivity ratio between PI3Kγ and the class
IA PI3K isoforms (R, â, δ) . Notably, the isoform selectivity
determined for compound 26 at the enzymatic level could
subsequently be reproduced in cellular assays, where we show,
for the first time, a compound inhibiting PI3Kγ (but not PI3KR,
PI3Kâ, and PI3Kδ) dependent pathways inside cells. Compound
26, when tested in a murine peritonitis model, produced a similar
decrease in leukocyte infiltration as previously observed with
PI3Kγ-deficient mice.
General Experimental Methods
Procedures. All chemicals were purchased from Fluka-Aldrich,
Buchs (CH), unless otherwise stated. Melting points were measured
with a Bu¨chi melting point apparatus B-545 and are not corrected.
NMR spectra were recorded on a Bruker DPX 300 MHz spec-
trometer. Data were reported as follows: chemical shift δ in ppm
using either residual DMSO (2.49 ppm) or CHCl3 (7.19 ppm) as
internal standards, multiplicity (s ) singlet, d ) doublet, t ) triplet,
q ) quadruplet, m ) multiplet), integration, and coupling constants
(J) in hertz. MS data were obtained using a Perkin-Elmer API 150
EX (APCI) mass spectrometer. The analytical HPLC was performed
using an HPLC Waters Symmetry C8 50 mm × 4.6 mm column
with the following conditions: MeCN/H2O, 0.09% TFA, 0% to
100% (8 min), max plot 230-400 nm. Elemental analyses were
performed on an Erba Science 11108 CHN analyzer.
Figure 9. Blockade of MCP-1-induced monocyte chemotaxis. Graphs
show inhibition curve of 26 in comparison with LY294002. After
starvation of THP-1 (a) or primary bone marrow derived monocytes
(b) and pretreatment with inhibitor or DMSO, as indicated, chemotaxis
was induced with MCP-1 (10-9 M) as described in the section
Biological Methods.
General Procedures for the Suzuki Reaction. Procedure A.
The bromo derivative (100 mg, 1 equiv) was dissolved in a mixture
(7/3) of dry toluene and ethanol (10 mL) and purged with inert
argon gas for 10 min. Tetrakis(triphenylphosphine)palladium(0)
(0.05 equiv) was then added followed by the corresponding boronic
acid (1.1 equiv) and a 2 M aqueous potassium carbonate solution
(3 equiv). The reaction mixture was refluxed for 12 h, then allowed
to cool to room temperature The organic solvents were removed
manner with an IC50 value of 52 µM, as well as in the monocytic
cell line THP-1 with an IC50 value of 53 µM (Figure 9).
In Vivo Pharmacology. To evaluate the efficacy of 26 to
block leukocyte migration in vivo, it was tested in a mouse
model of thioglycollate-induced peritonitis. Since neutrophils
are the predominant cell type in the initial influx of leukocytes,

3866 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Pomel et al.
under vacuum, and the residue was taken up in ethyl acetate and
washed with water and brine. The organic phases were finally dried
over magnesium sulfate and evaporated to dryness to give a crude
compound that was purified through a silica gel pad using
dichloromethane as the eluent. The crude compounds were analyzed
by LC/MS and used in the next step without further purification.
Procedure B. The corresponding bromo derivative (100 mg, 1
equiv), boronic acid (1.1 equiv), palladium(II) acetate (0.1 equiv),
and 2 M aqueous potassium carbonate (3 equiv) were stirred in
water (5 mL) at room temperature for 12 h under a nitrogen
atmosphere. The reaction mixture was then neutralized by adding
citric acid (10%) and extracted with ethyl acetate. The organic phase
was washed with water and brine and dried over magnesium sulfate.
It was then concentrated under vacuum to give the corresponding
crude compound that was purified through a silica gel pad with
dichloromethane as the eluent. The desired crude compound was
analyzed by LC/MS and used in the next step without further
purification.
(HPLC: 98%, tR ) 2.77 min). 1H NMR (CDCl3, 300 MHz) δ 9.80
(m, 1H), 9.50 (s, 1H), 7.70-7.60 (m, 1H), 7.15 (d, 1H, J ) 3 Hz),
6.85 (d, 1H, J ) 3 Hz), 6.50-6.35 (m, 2H). MS (ES), m/z: 207.2
[M + H]+, 205.3 [M - H]-. The intermediate aldehyde 26a can
also be synthesized in one step according to procedure A in 4%
yield.
5-[5-(4-Fluoro-2-hydroxyphenyl)furan-2-ylmethylene]thiazo-
lidine-2,4-dione (26). Compound 26a (175 mg, 0.85 mmol, 1
equiv), â-alanine (151 mg, 1.7 mmol, 2 equiv), and 2,4-thiazo-
lidinedione (200 mg, 1.7 mmol, 2 equiv) were heated at 100 °C
for 1 h in acetic acid (3 mL). Water (0.5 mL) was added, and the
resulting precipitate was filtered off to give 245 mg (95%) of the
expected compound as a yellow solid. Treatment with 2 equiv of
KOH (1 M) gave the desired bis-potassium salt as a red powder
(HPLC: 96.20%, tR ) 3.44 min), mp >400 °C. 1H NMR (DMSO-
d6) δ 7.38 (d, 1H, J ) 9 Hz), 7.28 (d, 1H, J ) 3 Hz), 7.02 (s, 1H),
6.58 (d, 1H, J ) 3 Hz), 5.78-5.73 (m, 2H). MS (ES), m/z: 304.09
[M - H]-. Anal. (C14H6FK2NO4S) C, H, N.
5-[5-Phenyl)furan-2-ylmethylene]thiazolidine-2, 4-dione (1).
Aldehyde 1a is commercially available from Aldrich. The Knoev-
enagel reaction was performed starting from 100 mg of 1a. The
title compound (150 mg) was obtained in 95% yield following the
general procedure for the Knoevenagel reaction (HPLC: 99.9%,
General Procedure for the Knoevenagel Reaction. The cor-
responding aldehyde (1 equiv), â-alanine (2 equiv), and 2,4-
thiazolidinedione (2 equiv) were heated at 100 °C for 1 h in glacial
acetic acid (2 mL). Upon completion of the reaction, the mixture
was cooled, the reaction was quenched with water, and the
precipitate was filtered off. The solid was washed with water and
methanol and then finally dried with diethyl ether to give the desired
compound in high purity (>90%).
Preparation of 5-[5-(4-Fluoro-2-hydroxyphenyl)furan-2-yl-
methylene]thiazolidine-2,4-dione (26) According to Scheme 2.
2-Benzyloxy-1-bromo-4-fluorobenzene (26c). 2-Bromo-5-fluo-
rophenol (1.16 mL, 10.47 mmol, 1 equiv) was dissolved in THF
(25 mL) in the presence of potassium carbonate (2.90 g, 20.94
mmol, 2 equiv). The reaction mixture was stirred at room
temperature for 10 min. Benzyl bromide (1.50 mL, 12.57 mmol,
1.2 equiv) was added, and the reaction mixture was heated at 60
°C for 12 h before the reaction was quenched with water (20 mL).
Ethyl acetate was added, the desired compound was extracted, and
organic phases were washed several times with an aqueous
potassium carbonate solution. The organic layer was dried over
magnesium sulfate, evaporated under vacuum, and purified by flash
chromatography using cyclohexane as eluent. An amount of 2 g of
the expected compound (68%) was recovered as a white oil
(HPLC: 95%, tR ) 4.51 min). 1H NMR (CDCl3, 300 MHz) δ 7.52-
7.36 (m, 6H), 6.74-6.59 (m, 2H), 5.16 (m, 2H). MS (ES), m/z:
282.1 [M + H]+.
5-(2-Benzyloxy-4-fluorophenyl)furan-2-carbaldehyde (26b).
Compound 26c (1.5 g, 5.34 mmol, 1 equiv) was dissolved in a
mixture (7/3) of dry toluene and ethanol (25 mL) and purged with
argon gas for 10 min. Tetrakis(triphenylphosphine)palladium(0)
(590 mg, 0.53 mmol, 0.1 equiv) was then added followed by
5-formyl-2-furanboronic acid (821 mg, 5.87 mmol, 1.1 equiv) and
a 2 M aqueous potassium carbonate solution (3 mL, 6 mmol). The
reaction mixture was refluxed for 4 h before being cooled. The
organic solvents were evaporated under vacuum, and the residue
was redissolved in ethyl acetate, washed with water and brine, and
finally dried over magnesium sulfate. Ethyl acetate was evaporated
under vacuum to give a crude compound that was purified on a
silica gel using cyclohexane/ethyl acetate (9/1) as eluent to give
880 mg of the expected compound (56%) as a white oil (HPLC:
94%, tR ) 4.30 min). 1H NMR (CDCl3, 300 MHz) δ 9.54 (s, 1H),
8.0-7.95 (m, 1H), 7.39-7.37 (m, 5H), 7.17 (d, 1H, J ) 3 Hz),
6.91 (d, 1H, J ) 3 Hz), 6.78-6.70 (m, 2H). MS (ES), m/z: 297.3
[M + H]+, 295.3 [M - H]-.
5-(4-Fluoro-2-hydroxyphenyl)furan-2-carbaldehyde (26a). Com-
pound 26b (650 mg, 2.19 mmol, 1 equiv) was dissolved in DCM
(25 mL) under an inert atmosphere and cooled to -70 °C. Boron
trichloride (1 M) (5.3 mL, 5.3 mmol, 2.4 equiv) was added slowly,
and the reaction mixture was left at -70 °C for 30 min before the
reaction was quenched slowly with water. The aqueous phase was
made neutral to pH 7 by addition of a saturated potassium carbonate
solution. The organic phase was decanted, dried over magnesium
sulfate, and evaporated to give 300 mg (66%) of a yellow oil
1
tR ) 3.74 min), mp 262-264 °C. H NMR (DMSO-d6) δ 12.48
(m, 1H), 7.82 (m, 2H), 7.65 (s, 1H), 7.53 (m, 2H), 7.41 (m, 1H),
7.28 (d, 1H, J ) 3 Hz), 7.24 (d, 1H, J ) 3 Hz). MS (ES), m/z:
272.8 [M + H]+, 270.16 [M - H]-. Anal. (C14H9NO3S) C, H, N.
5-[5-(2-Hydroxyphenyl)furan-2-ylmethylene]thiazolidine-2, 4-di-
one (2). Intermediate aldehyde 2a (11 mg) was synthesized
according to procedure A in 10% yield. MS (ES), m/z: 187.2 [M
- H]-. The title compound (6 mg) was obtained in 35% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 91.97%, tR ) 3.45 min), mp 290-292 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 1H), 9.80 (m, 1H), 7.70 (m, 1H), 7.60 (m,
2H), 7.50 (s, 1H), 6.80 (m, 2H), 6.60 (d, 1H, J ) 3 Hz). MS (ES),
m/z: 288.1 [M + H]+, 286.10 [M - H]-. Anal. (C14H9NO4S) C,
H, N.
5-[5-(3-Hydroxyphenyl)furan-2-ylmethylene]thiazolidine-2,4-
dione (3). Intermediate aldehyde 3a (15 mg) was synthesized
according to procedure A in 14% yield. MS (ES), m/z: 187.2 [M
- H]-. The title compound (5 mg) was obtained in 22% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 90.4%, tR ) 3.32 min), mp 280-281 °C. 1H NMR
(DMSO-d6) δ 12.45 (m, 1H), 9.80 (m, 1H), 7.63 (s, 1H), 7.31-
7.20 (m, 5H), 6.80 (d, 1H, J ) 3 Hz). MS (ES), m/z: 288.7 [M +
H]+, 286.7 [M - H]-. Anal. (C14H9NO4S) C, H, N.
5-[5-(4-Hydroxyphenyl)furan-2-ylmethylene]thiazolidine-2, 4-di-
one (4). Intermediate aldehyde 4a (11 mg) was synthesized
according to procedure A in 10% yield. MS (ES), m/z: 187.2 [M
- H]-. The title compound (12 mg) was obtained in 68% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 96.5%, tR ) 3.62 min), mp 269-270 °C. 1H NMR
(DMSO-d6) δ 12.55 (m, 1H), 9.80 (m, 1H), 7.65 (d, 2H, J ) 9
Hz), 7.60 (s, 1H), 7.19 (d, 1H, J ) 3 Hz), 7.05 (d, 1H, J ) 3 Hz),
6.90 (d, 2H, J ) 9 Hz). MS (ES), m/z: 288.1 [M + H]+, 286.10
[M - H]-. Anal. (C14H9NO4S) C, H, N.
5-[5-(2-Methoxyphenyl)furan-2-ylmethylene]thiazolidine-2,4-
dione (5). Intermediate aldehyde 5a (20 mg) was synthesized
according to procedure A in 18% yield. MS (ES), m/z: 201.2 [M
- H]-. The title compound (12 mg) was obtained in 41% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 93.25%, tR ) 3.74 min), mp 270-272 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 1H), 7.81 (d, 1H, J ) 3 Hz), 7.64 (s, 1H),
7.43-7.38 (m, 1H), 7.23-7.14 (m, 4H), 3.95 (s, 3H). MS (ES),
m/z: 302.3 [M + H]+, m/z: 300.3 [M - H]-. Anal. (C15H11NO4S)
C, H, N.
5-[5-(2-Chlorophenyl)furan-2-ylmethylene]thiazolidine-2,4-di-
one (6). Aldehyde 6a is commercially available from Aldrich. The
Knoevenagel reaction was performed starting from 100 mg of 6a.
The title compound (134 mg) was obtained in 90% yield following
the general procedure for the Knoevenagel reaction (HPLC: 99.3%,

Furan-2-ylmethylene Thiazolidinediones
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3867
1
tR ) 4.08 min), mp 268-270 °C. H NMR (DMSO-d6) δ 12.53
(DMSO-d6) δ 12.60 (m, 2H), 7.72 (m, 2H), 7.64 (s, 1H), 7.47 (m,
1H), 7.31 (m, 1H), 7.27 (d, 1H, J ) 3 Hz), 7.26 (d, 1H, J ) 3 Hz),
3.67 (s, 2H). MS (ES), m/z: 328.3 [M - H]-. Anal. (C16H11NO5S)
C, H, N.
(m, 1H), 7.92 (dd, 1H, J ) 9, 3 Hz), 7.68 (s, 1H), 7.70-7.55 (m,
2H), 7.47 (m, 1H), 7.42 (d, 1H, J ) 3 Hz), 7.27 (d, 1H, J ) 3 Hz).
MS (ES), m/z: 306.10 [M + H]+, 304.10 [M - H]-. Anal. (C14H8-
ClNO3S) C, H, N.
5-[5-(2-Fluorophenyl)furan-2-ylmethylene]thiazolidine-2,4-di-
one (7). Intermediate aldehyde 7a (53 mg) was synthesized
according to procedure A in 50% yield. MS (ES), m/z: 191.2 [M
+ H]+. The title compound (55 mg) was obtained in 69% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 96.2%, tR ) 4.09 min), mp 240-242 °C. 1H NMR
(DMSO-d6) δ 12.53 (m, 1H), 7.88 (m, 1H), 7.84 (s, 1H), 7.45 (m,
3H), 7.26 (d, 1H, J ) 3 Hz), 7.12 (d, 1H, J ) 3 Hz). MS (ES),
m/z: 288.2 [M - H]-. Anal. (C14H8FNO3S) C, H, N.
N-[2-(5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl)phe-
nyl]acetamide (14). Intermediate aldehyde 14a (72 mg) was
synthesized according to procedure A in 68% yield. MS (ES), m/z:
230.2 [M + H]+. The title compound (49 mg) was obtained in
48% yield following the general procedure for the Knoevenagel
reaction (HPLC: 93.5%, tR ) 3.65 min), mp 251-253 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 1H), 9.71 (m, 1H), 7.78 (m, 1H), 7.76 (s,
1H), 7.46-7.40 (m, 3H), 7.24 (d, 1H, J ) 3 Hz), 7.03 (d, 1H, J )
3 Hz), 2.08 (s, 3H). MS (ES), m/z: 329.3 [M + H]+, 327.3 [M -
H]-. Anal. (C16H12N2O4S) C, H, N.
2-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]benzo-
ic Acid (8). Intermediate aldehyde 8a (30 mg) was synthesized
according to procedure A in 27% yield. MS (ES), m/z: 215.2 [M
- H]-. The title compound (14 mg) was obtained in 32% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 94.6%, tR ) 2.20 min), mp 220-222 °C. 1H NMR
(DMSO-d6) δ 13.20 (m, 1H), 12.46 (m, 1H), 7.75 (m, 2H), 7.65
(m, 1H), 7.63 (s, 1H), 7.53 (m,1H), 7.22 (d, 1H, J ) 3 Hz), 7.01
(d, 1H, J ) 3 Hz). MS (ES), m/z: 314.3 [M - H]-. Anal. (C15H9-
NO5S) C, H, N.
2-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]benzo-
ic Acid Methyl Ester (9). Intermediate aldehyde 9a (12 mg) was
synthesized according to procedure A in 10% yield. MS (ES), m/z:
231.2 [M + H]+. The title compound (6 mg) was obtained in 36%
yield following the general procedure for the Knoevenagel reaction
(HPLC: 93.0%, tR ) 3.24 min), mp 190-192 °C. 1H NMR
(DMSO-d6) δ 12.48 (m, 1H), 7.75-7.70 (m, 3H), 7.62 (s, 1H),
7.56 (m,1H), 7.22 (d, 1H, J ) 3 Hz), 7.02 (d, 1H, J ) 3 Hz), 3.79
(s, 3H). MS (ES), m/z: 330.3 [M + H]+, 328.3 [M - H]-. Anal.
(C16H11NO5S) C, H, N.
2-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]phen-
ylacetic Acid (10). Intermediate aldehyde 10a (43 mg) was
synthesized according to procedure A in 41% yield. MS (ES), m/z:
229.2 [M - H]-. The title compound (15 mg) was obtained in
25% yield following the general procedure for the Knoevenagel
reaction (HPLC: 97.4%, tR ) 2.70 min), mp 232-234 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 2H), 7.76 (d, 1H, J ) 9 Hz), 7.64 (s, 1H),
7.47-7.41 (m, 3H), 7.24 (d, 1H, J ) 3 Hz), 6.99 (d, 1H, J ) 3
Hz), 3.91 (s, 2H). MS (ES), m/z: 328.3 [M - H]-. Anal. (C16H11-
NO5S) C, H, N.
3-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]benzo-
ic Acid (11). Intermediate aldehyde 11a (48 mg) was synthesized
according to procedure A in 45% yield. MS (ES), m/z: 215.2 [M
- H]-. The title compound (31 mg) was obtained in 45% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 94%, tR ) 2.15 min), mp 358-360 °C. 1H NMR (DMSO-
d6) δ 13.15 (m, 1H), 12.51 (m, 1H), 8.36 (m, 1H), 8.05 (m, 1H),
7.94 (m, 1H), 7.66 (m, 2H), 7.41 (d, 1H, J ) 3 Hz), 7.24 (d, 1H,
J ) 3 Hz). MS (ES), m/z: 314.3 [M - H]-. Anal. (C15H9NO5S)
C, H, N.
3-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]benzo-
ic Acid Methyl Ester (12). Intermediate aldehyde 12a (12 mg)
was synthesized according to procedure A in 10% yield. MS (ES),
m/z: 231.2 [M + H]+. The title compound (5 mg) was obtained in
26% yield following the general procedure for the Knoevenagel
reaction (HPLC: 92.0%, tR ) 3.31 min), mp 320-322 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 1H), 8.33 (m, 1H), 8.03 (dd, 1H, J ) 9,
3 Hz), 7.88 (d, 1H, J ) 9, 3 Hz), 7.64 (m, 1H), 7.26 (d, 1H, J )
3 Hz), 7.15 (s, 1H), 6.80 (d, 1H, J ) 3 Hz), 3.91 (s, 3H). MS (ES),
m/z: 330.3 [M + H]+, 328.3 [M - H]-. Anal. (C16H11NO5S) C,
H, N.
3-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]phe-
nyl Acetic Acid (13). Intermediate aldehyde 13a (33 mg) was
synthesized according to procedure A in 30% yield. MS (ES), m/z:
229.2 [M - H]-. The title compound (17 mg) was obtained in
36% yield following the general procedure for the Knoevenagel
reaction (HPLC: 94.3%, tR ) 2.72 min), mp 210-212 °C. 1H NMR
5-[5-(2-Amino-5-nitrophenyl)furan-2-ylmethylene]thiazolidine-
2,4-dione (15). Intermediate aldehyde 15a (11 mg) was synthesized
according to procedure A in 10% yield. MS (ES), m/z: 233.2 [M
+ H]+. The title compound (1.5 mg) was obtained in 10% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 95.5%, tR ) 2.51 min). MS (ES), m/z: 332.3 [M + H]+,
330.3 [M - H]-.
5-[5-(2-Hydroxyphenyl)thiophen-2-ylmethylene]thiazolidine-
2,4-dione (16). Intermediate aldehyde 16a (13 mg) was synthesized
according to procedure A in 10% yield. MS (ES), m/z: 203.2 [M
- H]-. The title compound (8 mg) was obtained in 40% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 98%, tR ) 3.29 min), mp 270-272 °C. 1H NMR (DMSO-
d6) δ 12.51 (m, 1H), 10.69 (m, 1H), 8.03 (s, 1H), 7.79-7.75 (m,
2H), 7.64 (d, 1H, J ) 3 Hz), 7.24-7.18 (m, 1H), 6.97-6.90 (m,
2H). MS (ES), m/z: 304.2 [M + H]+, 302.2 [M - H]-. Anal.
(C14H9NO3S2) C, H, N.
5-[5-(3-Methoxyphenyl)furan-2-ylmethylene]thiazolidine-2,4-
dione (17). Intermediate aldehyde 17a (21 mg) was synthesized
according to procedure A in 19% yield. MS (ES), m/z: 201.2 [M
- H]-. The title compound (17 mg) was obtained in 55% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 99.78%, tR ) 3.90 min), mp 240-243 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 1H), 7.65 (s, 1H), 7.46-7.35 (m, 4H),
7.23 (d, 1H, J ) 3 Hz), 7.05 (d, 1H, J ) 3 Hz), 3.84 (s, 3H). MS
(ES), m/z: 302.2 [M + H]+, 300.2 [M - H]-. Anal. (C15H11NO4S)
C, H, N.
5-[5-(3-Nitrophenyl)furan-2-ylmethylene]thiazolidine-2,4-di-
one (18). Intermediate aldehyde 18a (20 mg) was synthesized
according to procedure A in 20% yield. MS (ES), m/z: 218.2 [M
+ H]+. The title compound (22 mg) was obtained in 75% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 98.17%, tR ) 3.85 min), mp 230-232 °C. 1H NMR
(DMSO-d6) δ 12.54 (m, 1H), 8.57 (m, 1H), 8.22 (m, 2H), 7.82 (m,
1H), 7.67 (s, 1H), 7.56 (d, 1H, J ) 3 Hz), 7.28 (d, 1H, J ) 3 Hz).
MS (ES), m/z: 315.3 [M - H]-. Anal. (C14H8N2O5S) C, H, N.
5-(5-Benzo[1,3]dioxol-5-yl-furan-2-ylmethylene)thiazolidine-
2,4-dione (19). Intermediate aldehyde 19a (11 mg) was synthesized
according to procedure A in 10% yield. MS (ES), m/z: 217.2 [M
+ H]+. The title compound (5 mg) was obtained in 34% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 98.9%, tR ) 4.40min), mp 295-298 °C. 1H NMR (DMSO-
d6) δ 12.45 (m, 1H), 7.61 (s, 1H), 7.37 (m, 2H), 7.20 (d, 1H, J )
3 Hz), 7.16 (d, 1H, J ) 3 Hz), 7.10 (d, 1H, J ) 9 Hz), 6.11 (s,
2H). MS (ES), m/z: 316.1 [M + H]+, 314.3 [M - H]-. Anal.
(C15H9NO5S) C, H, N.
5-[2,3′]Bifuranyl-5-ylmethylenethiazolidine-2,4-dione (20). In-
termediate aldehyde 20a (43 mg) was synthesized according to
procedure B in 40% yield. MS (ES), m/z: 163.2 [M + H]+. The
title compound (38 mg) was obtained in 55% yield following the
general procedure for the Knoevenagel reaction (HPLC: 96.2%,
1
tR ) 4.53 min), mp 255-257 °C. H NMR (DMSO-d6) δ 12.43
(m, 1H), 8.22 (m, 1H), 7.83 (m, 1H), 7.60 (s, 1H), 7.18 (d, 1H, J
) 3 Hz), 6.95 (d, 1H, J ) 3 Hz), 6.91 (m, 1H). MS (ES), m/z:
262.2 [M + H]+, 260.2 [M - H]-. Anal. (C12H7NO4S) C, H, N.
5-(5′-Nitro[2,2′]bifuranyl-5-ylmethylene)thiazolidine-2,4-di-
one (21). Intermediate aldehyde 21a (42 mg) was synthesized

3868 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Pomel et al.
according to procedure B in 40% yield. MS (ES), m/z: 208.1 [M
+ H]+. The title compound (34 mg) was obtained in 55% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 95.4%, tR ) 4.83 min), mp 328-330 °C. 1H NMR
(DMSO-d6) δ 12.59 (m, 1H), 7.87 (d, 1H, J ) 3 Hz), 7.67 (s, 1H),
7.41 (d, 1H, J ) 3 Hz), 7.28 (m, 2H). MS (ES), m/z: 307.2 [M +
H]+. Anal. (C12H6N2O6S) C, H, N.
5-(5-Pyridin-2-ylfuran-2-ylmethylene)thiazolidine-2,4-dione
(22). Intermediate aldehyde 22a (25 mg) was synthesized according
to procedure A in 23% yield. MS (ES), m/z: 174.1 [M + H]+.
The title compound (20 mg) was obtained in 52% yield following
the general procedure for the Knoevenagel reaction (HPLC:
96.77%, tR ) 3.69 min), mp 225-227 °C. 1H NMR (DMSO-d6) δ
12.50 (m, 1H), 8.70 (dd, 1H, J ) 6, 3 Hz), 8.10 (m, 1H), 7.90 (m,
1H), 7.65 (m, 1H), 7.27 (d, 1H, J ) 3 Hz), 7.10 (s, 1H), 6.80 (d,
1H, J ) 3 Hz). MS (ES), m/z: 273.2 [M + H]+, 271.2 [M - H]-.
Anal. (C13H8N2O3S) C, H, N.
207.1 [M + H]+. The title compound (2 mg) was obtained in 5%
yield following the general procedure for the Knoevenagel reaction
(HPLC: 98.2%, tR ) 3.51 min). MS (ES), m/z: 306.4 [M + H]+,
304.4 [M - H]-.
5-[5-(2-Hydroxy-5-chlorophenyl)furan-2-ylmethylene]thiazo-
lidine-2,4-dione (30). Intermediate aldehyde 30a (42 mg) was
synthesized according to procedure B in 40% yield. MS (ES), m/z:
221.5 [M - H]-. The title compound (27 mg) was obtained in
45% yield following the general procedure for the Knoevenagel
reaction (HPLC: 97.2%, tR ) 4.10min), mp 320-322 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 1H), 10.84 (m, 1H), 7.66 (s, 1H), 7.63 (d,
1H, J ) 3 Hz), 7.25 (m, 3H), 7.02 (d, 1H, J ) 3 Hz). MS (ES),
m/z: 322.5 [M + H]+, 320.5 [M - H]-. Anal. (C14H8ClNO4S) C,
H, N.
3-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]-4-hy-
droxybenzonitrile (31). Intermediate aldehyde 31a (68 mg) was
synthesized according to procedure B in 65% yield. MS (ES), m/z:
212.1 [M - H]-. The title compound (33 mg) was obtained in
33% yield following the general procedure for the Knoevenagel
reaction (HPLC: 95.7%, tR ) 3.25 min), mp 317-320 °C. 1H NMR
(DMSO-d6) δ 12.51 (m, 1H), 11.78 (m, 1H), 7.68 (m, 2H), 7.27
(m, 3H), 7.18 (m, 1H). MS (ES), m/z: 313.3 [M + H]+, 311.3 [M
- H]-. Anal. (C15H8N2O4S) C, H, N.
5-(5-Pyridin-3-ylfuran-2-ylmethylene)thiazolidine-2,4-dione
(23). Intermediate aldehyde 23a (32 mg) was synthesized according
to procedure A in 30% yield. MS (ES), m/z: 174.1 [M + H]+.
The title compound (20 mg) was obtained in 40% yield following
the general procedure for the Knoevenagel reaction (HPLC: 93.7%,
1
tR ) 3.80 min), mp 332-335 °C. H NMR (DMSO-d6) δ 12.52
(m, 1H), 9.05 (d, 1H, J ) 3 Hz), 8.58 (dd, 1H, J ) 6, 3 Hz), 8.16
(m, 1H), 7.67 (s, 1H), 7.56 (m, 1H), 7.44 (d, 1H, J ) 3 Hz), 7.26
(d, 1H, J ) 3 Hz). MS (ES), m/z: 273.2 [M + H]+, 271.2 [M -
H]-. Anal. (C13H8N2O3S) C, H, N.
5-[5-(2-Hydroxy-5-trifluoromethoxyphenyl)furan-2-ylmeth-
ylene]thiazolidine-2,4-dione (32). Intermediate aldehyde 32a (20
mg) was synthesized according to procedure B in 18% yield. MS
(ES), m/z: 271.2 [M - H]-. The title compound (20 mg) was
obtained in 71% yield following the general procedure for the
Knoevenagel reaction (HPLC: 95.9%, tR ) 3.58 min), mp 304-
5-(5-Quinolin-2-ylfuran-2-ylmethylene)thiazolidine-2,4-di-
one (24). Intermediate aldehyde 24a (14 mg) was synthesized
according to procedure A in 13% yield. MS (ES), m/z: 224.2 [M
+ H]+. The title compound (6 mg) was obtained in 33% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 98%, tR ) 3.15 min), mp 302-305 °C. 1H NMR (DMSO-
d6) δ 12.56 (m, 1H), 8.55 (d, 1H, J ) 9 Hz), 8.05-7.98 (m, 3H),
7.80 (m, 1H), 7.72 (s, 1H), 7.63 (m, 1H), 7.55 (d, 1H, J ) 3 Hz),
7.31 (d, 1H, J ) 3 Hz). MS (ES), m/z: 323.2 [M + H]+, 321.2 [M
- H]-. Anal. (C17H10N2O3S) C, H, N.
5-(5-Quinolin-8-ylfuran-2-ylmethylene)thiazolidine-2,4-di-
one (25). Intermediate aldehyde 25a (12 mg) was synthesized
according to procedure A in 11% yield. MS (ES), m/z: 224.2 [M
+ H]+. The title compound (5 mg) was obtained in 26% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 99%, tR ) 3.28 min), mp 295-297 °C. 1H NMR (DMSO-
d6) δ 12.50 (m, 1H), 9.05 (dd, 1H, J ) 6, 3 Hz), 8.48 (dd, 1H, J
) 9, 3 Hz), 8.27 (dd, 1H, J ) 9, 1 Hz), 8.08 (d, 1H, J ) 6 Hz,),
8.05 (m, 1H), 7.83 (m, 1H), 7.71(s, 1H), 7.65 (dd, 1H, J ) 9, 6
Hz), 7.31 (d, 1H, J ) 6 Hz). MS (ES), m/z: 323.2 [M + H]+,
321.2 [M - H]-. Anal. (C17H10N2O3S) C, H, N.
5-[5-(2,4-Dihydroxyphenyl)furan-2-ylmethylene]thiazolidine-
2,4-dione (27). Intermediate aldehyde 27a (34 mg) was synthesized
according to procedure B in 32% yield. MS (ES), m/z: 205.2 [M
+ H]+. The title compound (26 mg) was obtained in 52% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 95.2%, tR ) 3.05 min), mp 296-300 °C. 1H NMR
(DMSO-d6) δ 12.36 (m, 1H), 10.36 (m, 1H), 9.82 (m, 1H), 7.60
(s, 1H), 7.55 (d, 1H, J ) 9 Hz), 7.18 (d, 1H, J ) 3 Hz), 6.98 (d,
1H, J ) 3 Hz), 6.46 (m, 2H). MS (ES), m/z: 304.3 [M + H]+,
302.3 [M - H]-. Anal. (C14H9NO5S) C, H, N.
1
305 °C. H NMR (DMSO-d6) δ 12.50 (m, 1H), 10.97 (m, 1H),
7.67 (s, 1H), 7.59 (d, 1H, J ) 3 Hz), 7.24 (m, 3H), 7.08 (d, 1H, J
) 3 Hz). MS (ES), m/z: 372.3 [M + H]+, 370.3 [M - H]-. Anal.
(C15H8F3NO5S) C, H, N.
5-[5-(2-Hydroxy-5-methylphenyl)furan-2-ylmethylene]thiazo-
lidine-2,4-dione (33). Intermediate aldehyde 33a (52 mg) was
synthesized according to procedure B in 50% yield. MS (ES), m/z:
201.2 [M - H]-. The title compound (31 mg) was obtained in
40% yield following the general procedure for the Knoevenagel
reaction (HPLC: 97.9%, tR ) 3.79 min), mp 275-277 °C. 1H NMR
(DMSO-d6) δ 12.45 (m, 1H), 10.25 (m, 1H), 7.65 (s, 1H), 7.52 (d,
1H, J ) 3 Hz), 7.22 (d, 1H, J ) 3 Hz), 7.17 (d, 1H, J ) 3 Hz),
7.05 (dd, 1H, J ) 9, 3 Hz), 6.90 (d, 1H, J ) 9 Hz), 2.30 (s, 3H).
MS (ES), m/z: 302.3 [M + H]+, 300.3 [M - H]-. Anal. (C15H11-
NO4S) C, H, N.
5-[5-(2,5-Dihydroxyphenyl)furan-2-ylmethylene]thiazolidine-
2,4-dione (34). Intermediate aldehyde 34a (36 mg) was synthesized
according to procedure B in 34% yield. MS (ES), m/z: 205.2 [M
+ H]+. The title compound (10 mg) was obtained in 19% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 94.7%, tR ) 2.95 min), mp 263-265 °C. 1H NMR
(DMSO-d6) δ 12.42 (m, 1H), 9.78 (m, 1H), 9.17 (m, 1H), 7.63 (s,
1H), 7.20 (m, 3H), 6.81 (d, 1H, J ) 3 Hz), 6.65 (d, 1H, J ) 3 Hz).
MS (ES), m/z: 304.3 [M + H]+, 302.3 [M - H]-. Anal. (C14H9-
NO5S) C, H, N.
5-[5-(2-Hydroxy-3-nitrophenyl)furan-2-ylmethylene]thiazoli-
dine-2,4-dione (35). Intermediate aldehyde 35a (33 mg) was
synthesized according to procedure B in 32% yield. MS (ES), m/z:
234.1 [M + H]+. The title compound (9 mg) was obtained in 20%
yield following the general procedure for the Knoevenagel reaction
(HPLC: 97.1%, tR ) 3.51 min). MS (ES), m/z: 333.3 [M + H]+,
331.3 [M - H]-.
3-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]-4-hy-
droxybenzoic Acid (36). Intermediate aldehyde 36a (50 mg) was
synthesized according to procedure B in 47% yield. MS (ES), m/z:
231.2 [M - H]-. The title compound (54 mg) was obtained in
75% yield following the general procedure for the Knoevenagel
reaction (HPLC: 90.5%, tR ) 3.52 min), mp 310-311 °C. 1H NMR
(DMSO-d6) δ 13.40 (m, 1H), 12.50 (m, 2H), 8.50 (m, 1H), 8.00
(m, 1H), 7.70 (s, 1H), 7.53 (m, 1H), 7.29 (d, 1H, J ) 3 Hz), 7.20
(d, 1H, J ) 3 Hz). MS (ES), m/z: 332.1 [M - H]-.
4-[5-(2,4-Dioxothiazolidin-5-ylidenemethyl)furan-2-yl]-3-hy-
droxybenzoic Acid Methyl Ester (28). Intermediate aldehyde 28a
(11 mg) was synthesized according to procedure B in 10% yield.
MS (ES), m/z: 245.2 [M - H]-. The title compound (10 mg) was
obtained in 62% yield following the general procedure for the
Knoevenagel reaction (HPLC: 98.5%, tR ) 3.22 min), mp 340-
1
345 °C. H NMR (DMSO-d6) δ 12.49 (m, 1H), 10.94 (m, 1H),
7.82 (d, 1H, J ) 9 Hz), 7.66 (s,1H), 7.60 (m, 2H), 7.33 (d, 1H, J
) 3 Hz), 7.24 (d, 1H, J ) 3 Hz), 3.85 (s, 3H). MS (ES), m/z:
346.3 [M + H]+, 344.3 [M - H]-. Anal. (C16H11NO6S) C, H, N.
5-[5-(5-Fluoro-2-hydroxyphenyl)furan-2-ylmethylene]thiazo-
lidine-2,4-dione (29). Intermediate aldehyde 29a (21 mg) was
synthesized according to procedure B in 20% yield. MS (ES), m/z:

Furan-2-ylmethylene Thiazolidinediones
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3869
5-[5-(4-Fluoro-2-methoxyphenyl)furan-2-ylmethylene]thiazo-
lidine-2,4-dione (37). Intermediate aldehyde 37a (12 mg) was
synthesized according to procedure A in 11% yield. MS (ES), m/z:
219.2 [M - H]-. The title compound (10 mg) was obtained in
55% yield following the general procedure for the Knoevenagel
reaction (HPLC: 98.9%, tR ) 3.74 min), mp 263-265 °C. 1H NMR
(DMSO-d6) δ 12.46 (m, 1H), 7.84 (m, 1H), 7.64 (s, 1H), 7.22 (d,
1H, J ) 3 Hz), 7.10-7.15 (m, 2H), 7.05 (m, 1H), 3.97 (s, 3H).
MS (ES), m/z: 320.3 [M + H]+, 318.3 [M - H]-. Anal. (C15H10-
FNO4S) C, H, N.
5-[5-(4-Fluoro-2-hydroxyphenyl)thiophen-2-ylmethylene]thi-
azolidine-2,4-dione (38). Intermediate aldehyde 38a (14 mg) was
synthesized according to procedure A in 9% yield. MS (ES), m/z:
221.5 [M - H]-. The title compound (10 mg) was obtained in
45% yield following the general procedure for the Knoevenagel
reaction (HPLC: 95.1%, tR ) 3.64 min), mp 250-251 °C. 1H NMR
(DMSO-d6) δ 12.50 (m, 1H), 11.23 (m, 1H), 8.02 (s, 1H), 7.83
(m, 1H), 7.74 (d, 1H, J ) 3 Hz), 7.64 (d, 1H, J ) 3 Hz), 6.78-
6.74 (m, 2H). MS (ES), m/z: 322.5 [M + H]+, 320.5 [M - H]-.
Anal. (C14H8FNO3S2) C, H, N.
M KCl before use. All titrations were performed under argon
atmosphere at 25 ( 0.5 °C. Typically, an amount of 20 µL of 10
mM DMSO stock solution of test sample was dissolved in 15 mL
of 0.15 M KCl solutions containing 60, 50, 40, and 30 wt % MeOH
prior to titration. Standardized 0.5 M KOH was added to increase
the pH to 11.0, and the solutions were titrated with 0.5 M HCl to
pH 4. The pH change per titrant addition was limited to 0.15 pH
units. After each titrant addition, the pH was measured and the
UV spectrum was recorded from 200 to 500 nm. All experiments
were carried out in the presence of 0.25 mM potassium dihydrogen
phosphate. The data processing, including Yasuda-Shedlovsky
extrapolation to cosolvent zero, was carried out using pKaUV
software (version 1.001, Sirius Ltd.).
Biological Methods. In Vitro PI3Kγ Lipid Kinase Assay. A
PI3Kγ lipid kinase assay, based on the neomycin-coated scintillation
proximity assay (SPA) bead technology (Amersham), was per-
formed in 384-well plates using ATP/[γ33P]ATP and PtdIns (Sigma)
as substrates, as previously described.10,54 Kinase assays for IC50
value determinations with PI3KR, PI3Kâ, and PI3Kδ were carried
out, as previously described.10
5-[5-(2-Fluoro-4-hydroxyphenyl)furan-2-ylmethylene]thiazo-
lidine-2,4-dione (39). Intermediate aldehyde 39a (15 mg) was
synthesized according to procedure B in 13% yield. MS (ES), m/z:
205.2 [M - H]-. The title compound (5 mg) was obtained in 21%
yield following the general procedure for the Knoevenagel reaction
(HPLC: 94.9%, tR ) 3.46min), mp 298-300 °C. 1H NMR (DMSO-
d6) δ 12.45 (m, 1H), 10.47 (m, 1H), 7.68 (m, 1H), 7.63 (s, 1H),
7.22 (d, 1H, J ) 3 Hz), 6.91-6.83 (m, 2H), 6.77 (dd, 1H, J ) 15,
3 Hz). MS (ES), m/z: 306.2 [M + H]+, 304.2 [M - H]-. Anal.
(C14H8FNO4S) C, H, N.
5-[5-(2,4-Difluorophenyl)furan-2-ylmethylene]thiazolidine-
2,4-dione (40). Intermediate aldehyde 40a (68 mg) was synthesized
according to procedure A in 65% yield. MS (ES), m/z: 209.2 [M
+ H]+. The title compound (69 mg) was obtained in 69% yield
following the general procedure for the Knoevenagel reaction
(HPLC: 100%, tR ) 4.14 min), mp 287-290 °C. 1H NMR (DMSO-
d6) δ 12.52 (m, 1H), 7.90 (m, 1H), 7.67 (s, 1H), 7.50 (m, 1H),
7.35 (dd, 1H, J ) 9, 3 Hz), 7.26 (d, 1H, J ) 3 Hz), 7.02 (d, 1H,
J ) 3 Hz). MS (ES), m/z: 306.2 [M - H]-. Anal. (C14H7F2NO3S)
C, H, N.
5-[5-(3,5-Difluoro-2-hydroxyphenyl)furan-2-ylmethylene]thi-
azolidine-2,4-dione (41). Intermediate aldehyde 41a (7 mg) was
synthesized according to procedure B in 6% yield. MS (ES), m/z:
223.2 [M - H]-. The title compound (5 mg) was obtained in 52%
yield following the general procedure for the Knoevenagel reaction
(HPLC: 95.2%, tR ) 3.75 min), mp 319-325 °C. 1H NMR
(DMSO-d6) δ 12.53 (m, 1H), 10.67 (m, 1H), 7.67 (s, 1H), 7.35
(m, 1H), 7.30 (d, 1H, J ) 3 Hz), 7.27 (m, 1H), 7.24 (d, 1H, J )
3 Hz). MS (ES), m/z: 324.2 [M + H]+, 322.3 [M - H]-. Anal.
(C14H7F2NO4S) C, H, N.
Crystallization of PI3Kγ and Soaking of Inhibitors. A
pVL1393-based vector expressing aa 144-1′102 of human PI3Kγ,
in fusion with N-terminal 6 × His-tag, kindly provided by R.
Williams (MRC, Cambridge, U.K.), was used for expression in TN5
cells. The protein was purified essentially as described.12 PI3Kγ,
in 20 mM Tris-HCl, pH 7.2, 50 mM (NH4)2SO4, 1% betaine, 1%
ethylene glycol, 0.02% CHAPS, and 5 mM DTT, was crystallized
by sitting drop vapor diffusion. An amount of 5 µL of protein
solution was mixed with 5 µL of a solution containing 20-22%
PEG 4000, 100 mM Tris-HCl, pH 7.2, and 200 mM (NH4)2SO4.
Crystals were harvested directly into a soaking solution consisting
of 25% PEG 4000, 20 mM (NH4)2SO4, 100 mM Tris-HCl, pH 7.2,
10% (v/v) glycerol, and 2 mM of inhibitor and soaked for 24 h.
X-ray diffraction data were collected at 100 K at the X06SA
beamline at the Swiss Light Source and processed using MOSFLM
and programs from the CCP4 package.51 The atomic models were
manually built using O and refined with CNX.52,53
C5a-Mediated PKB/Akt Phosphorylation in Macrophages.
After 3 h of starvation in serum-free medium, Raw-264 macro-
phages were pretreated with inhibitors or DMSO for 30 min and
stimulated for 5 min with 50 nM C5a (Sigma). We monitored PKB/
Akt phosphorylation using phospho-Ser-473 Akt specific antibody
(CST) and standard ELISA protocols.
MCP-1/CCL2 and CSF-1-Mediated PKB/Akt Phosphoryla-
tion in Monocytes. THP-1 monocytes, starved in serum-free
medium for 3 h, were pretreated with inhibitors or DMSO for 15
min and stimulated with 100 nM MCP-1/CCL2 (R&D) for 30 s or
50 ng/mL CSF-1 (Peprotec) for 5 min. PKB/Akt phosphorylation
was monitored using phospho-Ser-473 Akt specific antibody, as
previously described.10
SCF-Induced PKB/Akt Phosphorylation in Mast Cells. Pri-
mary bone marrow cells were isolated from 4- to 8-week-old wild-
type mice as previously described45 and derived to mast cells by
incubation with medium containing 20 ng/mL of stem cell factor
(SCF) (PeproTech, Switzerland) and 20 ng/mL of IL-3 (PeproTech,
Switzerland) for at least 4 weeks. Confirmation of the expression
of mast cell specific surface markers was done by FACS analysis
using antibodies against c-kit (cKit-PE mouse antibody, Pharmin-
gen). Cells were maintained in culture in the presence of SCF and
IL-3. To induce PKB/Akt phosphorylation, mast cells were
resuspended at 2.5 × 106 cells/mL and starved in medium
containing no SCF or IL-3 for 24 h. After preincubation with
compounds or 1% DMSO for 20 min, cells were activated with 20
ng/mL of SCF for 15 min at 37 °C, fixed in 1.5% paraformaldehyde
for 20 min, and permeabilized with 0.2% Triton X-100 for 10 min,
at room temperature. PKB/Akt phosphorylation was visualized using
phospho-Ser-473 specific Akt antibodies (Cell Signaling) and
standard FACS protocols.
Preparation of Mouse Bone Marrow Derived Monocytes
(BMDM) and in Vitro Chemotaxis. Bone marrow cells were
isolated, prepared, and stimulated, as previously described.45 For
in vitro chemotaxis, 107 5-day-derived BMDMs were suspended
in 1 mL of medium containing 0.5% BSA and inhibitors or DMSO
and applied to the upper chamber of a transwell, 5 µm pore size,
COSTAR chemotaxis plate. An amount of 600 µL of medium
containing MCP-1/CCL2 and inhibitors or DMSO was added to
the lower chamber. After 3 h of incubation at 37 °C and 5% CO2,
the number of cells in the lower chamber was quantified with a
Beckman Coulter AcT 5diffTM.
Computational Methods. Docking. For all modeling studies
we used the crystal structure of PI3Kγ in complex with 2. All
docking simulations were performed using the program Glide.56
The center and the size for for the docking grid box were selected
with respect to the cocrystallized ligand, and the default parameters
were used. We performed flexible ligand docking using the SP level
of accuracy and typically saved five docked poses per ligand for
pKa Measurements. Acid-base ionization constants were
measured by UV spectroscopy using the GlpKa automatic titrator
equipped with the DPAS module (Sirius Ldt, East Sussex, U.K.).
The pH electrode was calibrated in the pH range 1.8-12.2 in 0.15

3870 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Pomel et al.
(12) Walker, E. H.; Pacold, M. E.; Perisic, O.; Stephens, L.; Hawkins, P.
T.; Wymann, M. P.; Williams, R. L. Structural determinants of
phosphoinositide 3-kinase inhibition by wortmannin, LY294002,
quercetin, myricetin, and staurosporine. Mol. Cell 2000, 6, 909-
919.
(13) Gammill, R. B.; Judge, T. M.; Morris, J. Preparation of antiathero-
sclerotic and antithrombotic 1-benzopyran-4-ones and 2-amino-1,3-
benzoxazine-4-ones. PCT Int. Appl. WO9119707, 1991.
(14) Gogliotti, R. D.; Muccioli, K. L.; Para, K. S.; Visnick, M. Preparation
of benzoxazines and related compounds as inhibitors of PI3Ks. PCT
Int. Appl. WO2004056820, 2004.
subsequent analysis. For docking of molecules containing the
2-hydroxy group, a hydrogen-bonding constraint to Val882 NH was
requested.
Minimization. Geometry minimization was carried out using
the program MacroModel from Schro¨dinger, LLC, New York. In
a typical simulation, a 5 Å shell of completed residues around the
ligand was fully minimized using MMFFs and the default settings.
Induced-Fit Docking. The induced-fit docking protocol used is
based on docking by Glide and on side chain optimization by Prime
(Schro¨dinger, LLC, New York). We used the default settings, which
sampled side chain conformation space of residues within 5 Å from
the ligand. Final complex structures were generated by Glide
redocking into newly generated protein conformers.
(15) Barvian, N. C.; Kolz, C. N.; Para, K. S.; Patt, W. C.; Visnick, M.
Preparation of benzoxazin-3-ones and derivatives as inhibitors of
PI3K kinase for treating inflammations, cardiovascular diseases and
cancers. PCT Int. Appl. WO2004052373, 2004.
(16) Bruendl, M. M.; Connolly, M. K.; Goodman, A. P.; Gogliotti, R.
D.; Lee, H. T.; Plummer, M. S.; Sexton, K. E.; Reichard, G. A.;
Visnick, M.; Wilson, M. W. Preparation of N-tetrazolyl benzo[b]-
thiophenecarboxamides as phosphoinositide-3-kinase (PI3K) inhibi-
tors for the treatment of cancer, inflammatory and cardiovascular
diseases. PCT Int. Appl. WO2004108715, 2004.
(17) Connolly, M. K.; Cogliotti, R. D.; Hurt, C. R.; Reichard, G. A.;
Visnick, M. Preparation of halo-substituted N-tetrazolylbenzo[b]-
thiophenecarboxamides with PI3K inhibitory activity as therapeutic
agents. PCT Int. Appl. WO2005023800, 2005.
(18) Connolly, M.; Gogliotti, R. D.; Lee, H. T.; Plummer, M. S.; Sexton,
K. E.; Visnick, M. Preparation of N-tetrazolyl benzo[b]thiophen-
ecarboxamides as phosphoinositide-3-kinase (PI3K) inhibitors for the
treatment of cancer, inflammatory and cardiovascular diseases. PCT
Int. Appl. WO2004108713, 2004.
(19) Gogliotti, R. D.; Lee, H. T.; Sexton, K. E.; Visnick, M. Preparation
of tetrazolyl benzofurancarboxamides as phosphoinositide-3-kinase
(PI3K) inhibitors for the treatment of cancer, inflammatory and
cardiovascular diseases. PCT Int. Appl. WO2004108709, 2004.
(20) Para, K. S.; Stankovic, C. J.; Visnick, M.. Preparation of 3-substituted
indoles as inhibitors of phosphoinositide-3 kinases (PI3Ks). PCT Int.
Appl. WO2004108708, 2004.
(21) Connolly, M. K.; Gogliotti, R. D.; Plummer, M. S.; Visnick, M.
Preparation of morpholinyl-pyrimidine derivatives as inhibitors of
phosphoinositide-3-kinases. PCT Int. Appl. WO2005042519, 2005.
(22) Bruce, I.; Finan, P.; Leblanc, C.; McCarthy, C.; Whitehead, L.; Blair,
N. E.; Bloomfield, G. C.; Hayler, J.; Kirman, L.; Oza, M. S.; Shukla,
L.. 5-Phenylthiazole derivatives and their use as phosphatidylinositol
3-kinase (PI3K) inhibitors for the treatment of allergic and inflam-
matory diseases. PCT Int. Appl. WO2003072557, 2003.
(23) Shimada, M.; Murata, T.; Fuchikami, K.; Tsujishita, H.; Omori, N.;
Kato, I.; Miura, M.; Urbahns, K.; Gantner, F.; Bacon, K. Preparation
of fused azole-pyrimidine derivatives as PI3K inhibitors with
therapeutic uses. PCT Int. Appl. WO2004029055, 2004.
(24) Bloomfield, G. C.; Bruce, I.; Leblanc, C.; Oza, M. S.; Whitehead,
L.. Preparation of 5-phenylthiazoles as phosphatidylinositol 3-kinase
(PI3 kinase) inhibitors. PCT Int. Appl. WO2004078754, 2004.
(25) Bruce, I.; Cuenoud, B.; Keller, T. H.; Pilgrim, G. E.; Press, N.; Le
Grand, D. M.; Ritchie, C.; Valade, B.; Hayler, J.; Budd, E. Preparation
of pyrazinyl/pyridinyl thiazolylamines as inhibitors of phosphati-
dylinositol 3-kinase. PCT Int. Appl. WO2004096797. 2004.
(26) Bloomfield, G. C.; Bruce, I.; Hayler, J.; Leblanc, C.; Le Grand, D.
M.; McCarthy, C. Preparation of phenylthiazolylureas as inhibitors
of phosphatidylinositol 3-kinase. PCT Int. Appl. WO2005021519,
2005.
(27) Quattropani, A.; Ru¨ckle, T.; Schwarz, M.; Dorbais, J.; Sauer, W.;
Cleva, C.; Desforges, G. Preparation of thiazole derivatives as
modulators of the phosphoinositide 3-kinases (PI3Ks). PCT Int. Appl.
WO2005068444, 2005.
(28) Ru¨ckle, T.; Shaw, J.; Church, D.; Covini, D. Preparation of 2-imino-
4-(thio)oxo-5-polycyclovinylazolines as PI3 kinase inhibitors. PCT
Int. Appl. WO2005011686, 2005.
(29) Ru¨ckle, T.; Jiang, X.; Gaillard, P.; Church, D.; Vallotton, T.
Preparation of azolidinone-vinyl fused-benzene derivatives for thera-
peutic uses as PI3 kinase inhibitors. PCT Int. Appl. WO2004007491,
2004.
(30) Robertson, A. D.; Jackson, S.; Kenche, V.; Yaip, C.; Parbaharan,
H.; Thompson, P. Preparation and formulation of morpholino-
substituted heterocycles as phosphoinositide 3-kinase inhibitors for
therapeutic use. PCT Int. Appl. WO2001053266, 2001.
(31) Jackson, S. P.; Robertson, A. D.; Kenche, V.; Thompson, P.;
Prabaharan, H.; Anderson, K.; Abbott, B.; Goncalves, I.; Nesbitt,
W.; Schoenwaelder, S.; Saylik, D. Preparation of morpholinyl- and
pyridinyl-substituted heterobicyclic ketones as selective inhibitors of
phosphoinositide 3-kinase â for use against thrombosis. PCT Int.
Appl. WO2004016607, 2004.
Acknowledgment. We thank the Analytical Chemistry group
at SPRI for support in physicochemical parameters determina-
tion. We also thank the DMPK science group at Serono Ivrea
for providing biopharmaceutical measurements. We further
thank people from CCG at SPRI for compound support. We
are grateful to R. Williams for advice and technical support for
crystallization. We further express our gratitude to the protein
chemistry group at SPRI for providing biological material for
enzyme assays and crystallization.
Supporting Information Available: Elemental analysis results
for compounds 1-14, 16-28, 30-34, 37-41. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Vanhaesebroeck, B.; Leevers, S. J.; Ahmadi, K.; Timms, J.; Katso,
R.; Driscoll, P. C.; Woscholski, R.; Parker, P. J.; Waterfield, M. D.
Synthesis and function of 3-phosphorylated inositol lipids. Annu. ReV.
Biochem. 2001, 70, 535-602.
(2) Bi, L.; Okabe, I.; Bernard, D. J.; Wynshaw-Boris, A.; Nussbaum, R.
L. Proliferative defect and embryonic lethality in mice homozygous
for a deletion in the p110alpha subunit of phosphoinositide 3-kinase.
J. Biol. Chem. 1999, 274, 10963-10968.
(3) Bi, L.; Okabe, I.; Bernard, D. J.; Nussbaum, R. L. Early embryonic
lethality in mice deficient in the p110beta catalytic subunit of PI
3-kinase. Mamm. Genome 2002, 13, 169-172.
(4) Okkenhaug, K.; Bilancio, A.; Farjot, G.; Priddle, H.; Sancho, S.;
Peskett, E.; Pearce, W.; Meek, S. E.; Salpekar, A.; Waterfield, M.
D.; Smith, A. J.; Vanhaesebroeck, B. Impaired B and T cell antigen
receptor signaling in p110delta PI 3-kinase mutant mice. Science
2002, 297, 1031-1034.
(5) Sasaki, T.; Irie-Sasaki, J.; Jones, R. G.; Oliveira-dos-Santos, A. J.;
Stanford, W. L.; Bolon, B.; Wakeham, A.; Itie, A.; Bouchard, D.;
Kozieradzki, I.; Joza, N.; Mak, T. W.; Ohashi, P. S.; Suzuki, A.;
Penninger, J. M. Function of PI3Kgamma in thymocyte development,
T cell activation, and neutrophil migration. Science 2000, 287, 1040-
1046.
(6) Hirsch, E.; Katanaev, V. L.; Garlanda, C.; Azzolino, O.; Pirola, L.;
Silengo, L.; Sozzani, S.; Mantovani, A.; Altruda, F.; Wymann, M.
P. Central role for G protein-coupled phosphoinositide 3-kinase
gamma in inflammation. Science 2000, 287, 1049-1053.
(7) Li, Z.; Jiang, H.; Xie, W.; Zhang, Z.; Smrcka, A. V.; Wu, D. Roles
of PLC-beta2 and -beta3 and PI3Kgamma in chemoattractant-
mediated signal transduction. Science 2000, 287, 1046-1049.
(8) Condliffe, A. M.; Davidson, K.; Anderson, K. E.; Ellson, C. D.;
Crabbe, T.; Okkenhaug, K.; Vanhaesebroeck, B.; Turner, M.; Webb,
L.; Wymann, M. P.; Hirsch, E.; Ruckle, T.; Camps, M.; Rommel,
C.; Jackson, S. P.; Chilvers, E. R.; Stephens, L. R.; Hawkins, P. T.
Sequential activation of class IB and class IA PI3K is important for
the primed respiratory burst of human but not murine neutrophils.
Blood 2005, 106, 1432-1440.
(9) Laffargue, M.; Calvez, R.; Finan, P.; Trifilieff, A.; Barbier, M.;
Altruda, F.; Hirsch, E.; Wymann, M. P. Phosphoinositide 3-kinase
gamma is an essential amplifier of mast cell function. Immunity 2002,
16, 441-451.
(10) Camps, M.; Ru¨ckle, T.; Ji, H.; Ardissone, V.; Rintelen, F.; Shaw, J.;
Ferrandi, C.; Chabert, C.; Gillieron, C.; Francon, B.; Martin, T.;
Gretener, D.; Perrin, D.; Leroy, D.; Vitte, P. A.; Hirsch, E.; Wymann,
M. P.; Cirillo, R.; Schwarz, M. K.; Rommel, C. Blockade of PI3Kγ
suppresses joint inflammation and damage in mouse models of
rheumatoid arthritis. Nat. Med. (N. Y.) 2005, 11, 936-943.
(11) Wymann, M. P.; Bjoerkloef, K.; Calvez, R.; Finan, P.; Thomast, M.;
Trifilieff, A.; Barbier, M.; Altruda, F.; Hirsch, E.; Laffargue, M.
Phosphoinositide 3-kinase γ: a key modulator in inflammation and
allergy. Biochem. Soc. Trans. 2003, 31, 275-280.

Furan-2-ylmethylene Thiazolidinediones
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3871
(32) Cleva, C.; Domine, D.; Merlot, C.; Bunn, J.; Sebille, E.; Sauer, W.;
Church, D. Priviledged substructure searching for focused set design.
Abstracts of Papers, 224th National Meeting of the American
Chemical Society, Boston, MA, August 18-22, 2002; American
Chemical Society: Washington, DC, 2002.
(33) Church, D.; Colinge, J. Method of operating a computer system to
perform a discrete substructural analysis. PCT Int. Appl. WO200233-
596, 2002.
(34) Ru¨ckle, T.; Grippi-Vallotton, T.; Maio, M.; Covini, D.; Pomel, V.;
Burgat-Charvillon, F.; Church, D.; Shaw, J.; Sauer, W.; Klicic, J.;
Sherbukhin, V.; Quattropani, A.; Dorbais, J.; Desforges, G.; Valognes,
D.; Camps, M.; Chabert, C.; Gillieron, C.; Francon, B.; Perrin, D.;
Leroy, D.; Gretener, D.; Nichols, A.; Ji, H.; Rintelen, F.; Ardissone,
V.; Ferrandi, C.; Vitte, P. A.; Carboni, S.; Cirillo, R.; Schwarz, M.
K.; Rommel, C. Identification and development of azolidinone vinyl-
fused benzene derivatives, as potent and selective PI3Kγ inhibitors,
orally active in models of rheumatoid arthritis. Abstracts of Papers,
230th National Meeting of the American Chemical Society, Wash-
ington, DC, August 28 through September 1, 2005; American
Chemical Society: Washington, DC, 2005.
(45) Weiss-Haljiti, C.; Pasquali, C.; Ji, H.; Gillieron, C.; Chabert, C.;
Curchod, M. L.; Hirsch, E.; Ridley, A. J.; van Huijsduijnen, R. H.;
Camps, M.; Rommel, C. Involvement of phosphoinositide 3-kinase
gamma, Rac, and PAK signaling in chemokine-induced macrophage
migration. J. Biol. Chem. 2004, 279, 43273-43284.
(46) Okkenhaug, K.; Bilancio, A.; Farjot, G.; Priddle, H.; Sancho, S.;
Peskett, E.; Pearce, W.; Meek, S. E.; Salpekar, A.; Waterfield, M.
D.; Smith Andrew, J. H.; Vanhaesebroeck, B. Impaired B and T cell
antigen receptor signaling in p110delta PI 3-kinase mutant mice.
Science 2002, 297, 1031-1034.
(47) Jones, G. E.; Prigmore, E.; Calvez, R.; Hogan, C.; Dunn, G. A.;
Hirsch, E.; Wymann, M. P.; Ridley, A. J. Requirement for PI 3-kinase
gamma in macrophage migration to MCP-1 and CSF-1. Exp. Cell
Res. 2003, 290, 120-131.
(48) Johnson, Z.; Power, C. A.; Weiss, C.; Rintelen, F.; Ji, H.; Ruckle,
T.; Camps, M.; Wells, T. N. C.; Schwarz, M. K.; Proudfoot, A. E.
I.; Rommel, C. Chemokine inhibitionswhy, when, where, which and
how? Biochem. Soc. Trans. 2004, 32, 366-377.
(49) Schwarz, M. K.; Wells, T. N. New therapeutics that modulate
chemokine networks. Nat. ReV. Drug DiscoVery 2002, 1, 347-58.
(50) Henderson, R. B.; Hobbs, J. A.; Mathies, M.; Hogg, N. Rapid
recruitment of inflammatory monocytes is independent of neutrophil
migration. Blood 2003, 102, 328-335.
(51) Leslie, A. G. Integration of macromolecular diffraction data. Acta
Crystallogr. 1999, D55, 1696-1702.
(52) Jones, T. A.; Zou, J. Y.; Cowan, S. W.; Kjeldgaard, M. Improved
methods for building protein models in electron density maps and
the location of errors in these models. Acta Crystallogr. 1991, A47,
110-119.
(53) Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros,
P.; Grosse-Kunstleve, R. W.; Jiang, J. S.; Kuszewski, J.; Nilges, M.;
Pannu, N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren, G. L.
Crystallography & NMR system: A new software suite for macro-
molecular structure determination. Acta Crystallogr., Sect. D.: Biol.
Crystallogr. 1998, 54, 905-921.
(54) Camps, M.; Chabert, C.; Martin, T.; Perrin, D.; Rommel, C.;
Wymann, M. P. Use of a scintillating solid support coated with an
aminoglycoside for identifying and/or quantifying a radiolabeled
aminoglycoside binding molecule such as mono- or polyphosphated
phosphoinositide in a sample. PCT Int. Appl. WO2002101084, 2002.
(55) Chatham, W. W.; Swaim, R.; Frohsin, H., Jr.; Heck, L. W.; Miller,
E. J.; Blackburn, W. D., Jr. Degradation of human articular cartilage
by neutrophils in synovial fluid. Arthritis Rheum. 1993, 36, 51-58.
(56) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic,
J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry,
J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: A new approach
for rapid, accurate docking and scoring. 1. method and assessment
of docking accuracy. J. Med. Chem. 2004, 47, 1739-1749.
(57) Sadhu, C.; Dick, K.; Treiberg, K.; Sowell, G.; Kesicki, E.; Oliver,
A. Inhibitors of human phosphatidyl-inositol 3-kinase delta. PCT Int.
Appl. WO200181346, 2001.
(35) Pomel, V.; Klicic J.; Covini, D.; Church, D.; Shaw, J.; Valognes,
D.; Camps, M.; Chabert, C.; Gillieron, C.; Franc¸on, B.; Perrin, D.;
Leroy, D.; Gretener, D.; Nichols, A.; Vitte, P. A, Carboni, S.;
Rommel, C.; Schwarz, M. K.; Ru¨ckle, T. Design and synthesis of a
novel class of potent and highly isoform selective PI3Kγ inhibitors.
Abstracts of Papers, 230th National Meeting of the American
Chemical Society, Washington, DC, August 28 through September
1, 2005; American Chemical Society: Washington, DC, 2005.
(36) Campbell, I. W. The clinical significance of PPAR gamma agonism.
Curr. Mol. Med. 2005, 5, 349-363.
(37) Yki-Jaervinen, H. Thiazolidinediones. N. Engl. J. Med. 2004, 351,
1106-1118.
(38) Wiesner, J.; Mitsch, A.; Jomaa, H.; Schlitzer, M. Structure-activity
relationships of novel anti-malarial agents. Part 7: N-(3-Benzoyl-4-
tolylacetylaminophenyl)-3-(5-aryl-2-furyl)acrylic acid amides with
polar moieties. Bioorg. Med. Chem. Lett. 2003, 13, 2159-2161.
(39) Hosoya, T.; Aoyama, H.; Ikemoto, T.; Kihara, Y.; Hiramatsu, T.;
Endo, M.; Suzuki, M. Dantrolene analogues revisited: general
synthesis and specific functions capable of discriminating two kinds
of Ca2+ release from sarcoplasmic reticulum of mouse skeletal
muscle. Bioorg. Med. Chem. 2003, 11, 663-673.
(40) Ismail, M. A.; Brun, R.; Wenzler, T.; Tanious, F. A.; Wilson, W.
D.; Boykin, D. W. Dicationic biphenyl benzimidazole derivatives as
antiprotozoal agents. Bioorg. Med. Chem. Lett. 2004, 12, 5405-5413.
(41) Chittiboyina, A. G.; Reddy, C.; Watkins, E. B.; Avery, M. A. First
synthesis of antimalarial Machaeriols A and B. Tetrahedron Lett.
2004, 45, 1689-1691.
(42) Dupuis, C.; Adiey, K.; Charruault, L.; Michelet, V.; Savignac, M.;
Genet, J. P. Suzuki cross-coupling of arylboronic acids mediated by
a hydrosoluble Pd(0)/TPPTS catalyst. Tetrahedron Lett. 2001, 42,
6523-6526.
(43) Miyaura, N.; Suzuki, A. Palladium-catalyzed cross-coupling reactions
of organoboron compounds. Chem. ReV. (Washington, D.C.) 1995,
95, 2457-2483.
(44) Unangst, P. C.; Connor, D. T.; Cetenko, W. A.; Sorenson, R. J.;
Kostlan, C. R.; Sircar, J. C.; Wright, C. D.; Schrier, D. J.; Dyer, R.
D. Synthesis and biological evaluation of 5-[[3,5-bis(1,1-dimethyl-
ethyl)-4-hydroxyphenyl]methylene]oxazoles, -thiazoles, and -imida-
zoles: novel dual 5-lipoxygenase and cyclooxygenase inhibitors with
antiinflammatory activity. J. Med. Chem. 1994, 37, 322-328.
(58) Breitfelder, S.; Maier, U.; Grauert, M.; Brandl, T.; Hoffmann, M.;
Hoenke, C.; Pautsch, A.; Kalkbrenner, F.; Joergensen, A.; Schaenzle,
G.; Peters, S.; Buettner, F.; Bauer, E. PI3 kinases. PCT Int. Appl.
WO2006040279, 2006.
JM0601598