Synthesis and evaluation of aryl thioxothiazolidinone inhibitors of ADAMTS-5 (Aggrecanase-2)

Article abstract of DOI:10.1016/j.bmcl.2006.12.027

5-Benzylidene-2-thioxo-thiazolidin-4-one inhibitors of ADAMTS-5 (Aggrecanase-2) have been prepared via commercially available starting materials. The identified compounds show micromolar ADAMTS-5 potency and demonstrate SAR trends for both the aryl group and thioxothiazolidinone zinc chelator. This series of compounds represents steps toward a metalloprotease inhibitor as a disease-modifying osteoarthritis drug.

Full text of DOI:10.1016/j.bmcl.2006.12.027

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1185–1188
Synthesis and evaluation of aryl thioxothiazolidinone inhibitors
of ADAMTS-5 (Aggrecanase-2)
Matthew G. Bursavich,a,* Adam M. Gilbert,a Sabrina Lombardi,a Katy E. Georgiadis,b
Erica Reifenberg,b Carl R. Flanneryb and Elisabeth A. Morrisb
aExploratory Medicinal Chemistry, Chemical and Screening Sciences, Wyeth Research, 401 North Middletown Road,
Pearl River, NY 10965, USA
bWomen’s Health and Musculoskeletal Biology, Wyeth Research, 200 Cambridge Park Drive, Cambridge, MA 02140, USA
Received 25 October 2006; revised 7 December 2006; accepted 8 December 2006
Available online 12 December 2006
Abstract—5-Benzylidene-2-thioxo-thiazolidin-4-one inhibitors of ADAMTS-5 (Aggrecanase-2) have been prepared via commercial-
ly available starting materials. The identified compounds show micromolar ADAMTS-5 potency and demonstrate SAR trends for
both the aryl group and thioxothiazolidinone zinc chelator. This series of compounds represents steps toward a metalloprotease
inhibitor as a disease-modifying osteoarthritis drug.
Ó 2006 Elsevier Ltd. All rights reserved.
Osteoarthritis (OA), the most common of musculoskel-
etal diseases, is rapidly becoming a significant medical
and financial burden upon the world.1 This debilitating
disease is characterized by progressive articular cartilage
breakdown leading to chronic pain, inflammation, and
reduced mobility in the affected joint. More than 20 mil-
lion Americans suffer the significant reduction in quality
of life associated with OA with this number expected to
grow with the aging population. Current treatments
provide only symptomatic relief (NSAIDs, intra-articu-
lar injections of hyaluronic acid conjugates, and surgical
joint replacement) with no therapy available to halt and/
or reverse the progression of this disease.2
Aggrecanases are members of the ADAMTS (a disinte-
grin and metalloprotease with thrombospondin motifs)
family of zinc metalloproteases. Aggrecanase-2 has been
designated ADAMTS-5 and shown to be responsible for
cleavage of aggrecan at the physiologically relevant
Glu373-Ala374 peptide bond.3 ADAMTS-5 knock-out
mice have been shown to have significantly reduced
OA severity in a surgically induced instability model.4
ADAMTS-5 has also been shown to be the major
ADAMTS in a mouse model of inflammatory arthritis.5
The inhibition of ADAMTS-5 may therefore protect
cartilage from damage and provide the first potential
therapy to halt and/or reverse the progression of OA.
Aggrecan, a multidomain proteoglycan, is a major com-
ponent of cartilage that interacts with hyaluronic acid
and link proteins to provide compressive resistance to
articular cartilage. An early and important characteristic
of the osteoarthritis process is loss of aggrecan from the
extracellular matrix with increased loss of aggrecan frag-
ments via aggrecanase proteolysis activity.3 If this degra-
dation is not halted or reversed the cartilage will be
subject to degradation by additional metalloproteinases
resulting in irreversible damage to the joint.
In our ADAMTS-5 HTS screening efforts several inhib-
itor scaffolds possessing zinc-chelating groups were iden-
tified.6 These included, among others, hydroxamic acids
(I), carboxylic acids (II), hydantoins (III), and thi-
oxothiazolidinones (IV) (Fig. 1). The significant prior
Zn
Zn
O
S
Zn
Zn
O
O
HN
R
S
N
N
O
R
R
N
O
H
R
O
O
I
II
III
IV
Keywords: Aggrecanase; ADAMTS-5; Thioxothiazolidinone; Metallo-
protease inhibitor; Zinc chelator.
Figure 1. Zinc binding modes associated with various known bidentate
zinc chelators: (I) hydroxamic acid, (II) carboxylic acid, (III) hydan-
toin, and (IV) thioxothiazolidinone.
*
Corresponding author. Tel.: +1 845 602 8143; fax: +1 845 602
5561; e-mail: Bursavm@wyeth.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.027

1186
M. G. Bursavich et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1185–1188
art describing hydroxamic acid, carboxylic acid, and
hydantoin-based metalloprotease inhibitors7 led us to
pursue a series of less-explored thioxothiazolidinones
as potential inhibitors of the zinc metalloprotease
ADAMTS-5. Whereas several general aggrecanase
inhibitors have been disclosed, to our knowledge, this
class of compounds8 represents the first disclosure of
ADAMTS-5 inhibitors.9–12
of the remaining p-benzyloxy analogs with micromolar
activity 5 and 6 were identified.
To follow up the initial SAR trends observed with the
benzyloxy analogs, a series of m-biphenyl ether thi-
oxothiazolidinones 7–14 were synthesized. Again vari-
ous substitutions on the biphenyl ether were
investigated and shown to play a significant role in
developing active analogs. Compounds with no substitu-
tion on the biphenyl ether were found to be inactive in 7,
whereas incorporation of p-chloro group in 8 provided a
low micromolar analog. Subsequent substitution of the
p-chloro for either a p-methoxy in 9 or p-methyl in 10
afforded a 3-fold reduction in potency. However increas-
ing the steric bulk of the p-substituent to a t-butyl in 11
led to an increase of potency. Keeping the p-chloro and
adding another m-chloro resulted in a small increase in
potency for 12. Switching from a 3,4-di-chloro substitu-
tion to a 3,5-di-chloro substitution in 13 also provided a
modest increase in potency. Moving the electron with-
drawing substitution around the ring proved unfruitful
except for the m-trifluoromethyl analog 14 that main-
tained low micromolar activity.
Initial efforts to explore the thioxothiazolidinone scaf-
fold began with the synthesis outlined in Scheme 1.13
Knoevenagel condensation14 of various substituted ben-
zaldehydes and 2-thioxo-thiazolidin-4-one in the pres-
ence of b-alanine in refluxing acetic acid15 afforded the
desired target thioxothiazolidinone compounds 1–14
(Table 1).
From the synthesis of the initial array it was apparent
that a benzyloxy substituted aryl thioxothiazolidinone
1–5 greatly enhanced activity as analogs bearing smaller
substituents showed no activity against ADAMTS-5
(data not shown). Whereas compound 1 possessing only
one meta-benzyloxy group showed modest activity, this
potency was increased 10-fold with the incorporation of
a second benzyloxy group in either the p-position 2 or
other meta-position 3. To reduce the molecular weight,
the replacement of one of the bis(benzyloxy) groups of
2 with a methoxy group was investigated with synthesis
of 4. This replacement leads to a significant drop in
activity (IC50 > 22 lM), but upon further substitution
To explore the importance of the thioxothiazolidinone-
binding motif, various other chelators were investigated.
This substitution was explored for the m-biphenyl ether
series of thioxothiazolidinones. Knoevenagel condensa-
tion of the desired m-biphenyl ether benzaldehydes with
either thiazolidine-2,4-dione, imidazolidine-2,4-dione,
or 1-methyl-imidazolidine-2,4-dione in the presence of
b-alanine in refluxing acetic acid afforded the desired
targets 15–22 as shown in Scheme 2.
O
S
a
S
H
N
From the focused array synthesis (Table 2) the impor-
tance of the thioxothiazolidinone chelator is evident.
Simply replacing the thioxothiazolidinone sulfur to
an oxygen 15–18 led to at least a 3-fold decrease in
activity (compared to 8 and 11–13). Similarly the
O
R
H
R
1-14
Scheme 1. Reagents and condition: (a) 2-Thioxo-thiazolidin-4-one,
b-alanine, AcOH, reflux.
Table 1. ADAMTS-5 inhibition for thioxothiazolidinones 1–14
R1
R2
R1
R2
O
S
S
S
N
H
S
N
O
O
H
R3
R3
A
B
R1
R2
R3
Scaffold
ADAMTS-5 IC50 (lM)
a
Compound
1
2
O–Bn
O–Bn
O–Bn
OMe
OMe
OMe
H
A
A
A
A
A
A
B
B
B
B
B
B
B
B
20.1
1.8
O–Bn
3
O–Bn
1.4
>22
3.5
4
O–Bn
p-Cl O–Bn
p-OMe O–Bn
5
6
11.6
>22
4.9
7
H
Cl
H
8
9
OMe
Me
t-Bu
Cl
16.2
14.8
1.7
10
11
12
13
14
Cl
Cl
3.2
2.5
Cl
CF3
5.3
a Values are means of two experiments, standard deviations are 10%.

M. G. Bursavich et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1185–1188
1187
O
R
R
O
R1
R2
O
R1
R2
O
X
CO2H
S
CO2H
R
H
N
O
a
O
N
S
a, b
c
OH
N
H2N
S
O
S
H
R3
R3
O
R
28-36
15-22
X = S, NH, NMe
R
Scheme 2. Reagents and condition: (a) thiazolidine-2,4-dione, imidaz-
olidine-2,4-dione, or 1-methyl-imidazolidine-2,4-dione, b-alanine,
AcOH, reflux.
Scheme 4. Reagents and condition: (a) CS2, KOH, H2O; (b) Na2CO3,
ClEtCO2H; (c) benzaldehyde derivative, b-alanine, AcOH, reflux.
the presence of b-alanine in refluxing acetic acid afforded
the desired target N-acid substituted thioxothiazolidi-
nones 28–36.
Table 2. ADAMTS-5 inhibition for substituted m-biphenyl ethers
thiazolidinedione, imidazolidinedione, or 1-methyl-imidazolidinedione
analogs 15–22
R1
O
From this focused array of N-substituted thi-
oxothiazolidinone analogs the importance of the che-
lator was once again evident (Table 3). Substitution
of the dibenzyloxy N-thioxothiazolidinone analogs
with alkyl, phenyl, benzyl or phenethyl groups 23–27
resulted in a complete loss of activity. Interestingly
substitution of the NH with an additional acid via
glycine 28 afforded a compound of modest activity
that could be improved slightly via chain extension
to the b-glycine amino acid analog 29. Substituted
amino acid derivatives 30–32 provided a further in-
crease of potency with the Ph-Gly analog 30 identified
with sub-micromolar potency against ADAMTS-5.
The activity trend was also evident in the biphenyl-
ether N-thioxothiazolidinone analogs 33–36. In this
scaffold substitution of the NH with an additional
acid via glycine afforded compound 33 of modest
activity that could be significantly improved via substi-
tuted amino acid derivatives 34–36.
X
O
N
R2
O
H
R3
Compound R1 R2
R3
Cl
X
ADAMTS-5 IC50 (lM)
a
15
16
17
18
19
20
21
22
Cl
Cl Cl
Cl
S
>22
10.0
9.2
S
S
t-Bu
Cl
S
NH
6.8
>67
>67
>67
Cl
Cl NH
NH
t-Bu
Cl
NMe >22
a Values are means of two experiments, standard deviations are 10%.
3,4-di-chloro and 3,5-di-chloro substituted analogs 16
and 17 were more potent than the single p-chloro ana-
log 15 (IC50 > 22 lM). The potency could also be in-
creased as the p-t-Bu analog 18 was more potent
than the p-chloro analog (IC50 > 22 lM). The substi-
tuted m-biphenyl ether imidazolidinedione analogs,
replacing the sulfur with nitrogen, were even worse,
as none of the substituted imidazolidinediones 19–21
or substituted N–Me imidazolidinedione 22 showed
any ADAMTS-5 inhibition.
Table 3. ADAMTS-5 inhibition for N-substituted thioxothiazolidi-
none analogs 2 and 23–36
BnO
BnO
O
S
S
S
S
N
N
Cl
To further explore the importance of the thioxothiazo-
lidinone chelator, an array of N-substituted thi-
oxothiazolidinone analogs were synthesized. Using
previously identified conditions, a variety of commer-
cially available N-substituted 2-thioxo-thiazolidin-4-
ones were condensed with 3,4 dibenzyloxy benzaldehyde
to afford the desired target thioxothiazolidinone com-
pounds 23–27 (Scheme 3). The desired N-acid substitut-
ed thioxothiazolidinones were synthesized (Scheme 4)
via reaction of the desired amino acid and carbon disul-
fide under basic conditions followed by ring closure with
chloroacetic acid.16 Knoevenagel condensation of the
thioxothiazolidinone with substituted benzaldehydes in
O
O
R
R
Cl
A
B
Compound
R
H
Scaffold
ADAMTS-5
a
IC50 (lM)
2
23
24
25
26
27
28
29
30
31
32
12
33
34
35
36
A
A
A
A
A
A
A
A
A
A
A
B
1.8
Me
Et
>67
>67
>67
>67
>67
6.7
Ph
Bn
Phenethyl
CH2CO2H
CH2CH2CO2H
CH(Ph)CO2H
5.3
0.9
CH(R-i-Pr)CO2H
CH(R-i-Bn)CO2H
H
1.1
1.3
O
3.2
S
S
a
S
S
CH2CO2H
B
11.1
1.3
H
N
N
BnO
CH(Ph)CO2H
CH(R-i-Pr)CO2H
CH(R-i-Bn)CO2H
B
O
O
R
R
BnO
OBn
B
1.8
1.6
23-27
OBn
B
a Values are means of two experiments, standard deviations are 10%.
Scheme 3. Reagents and condition: (a) b-alanine, AcOH, reflux.

1188
M. G. Bursavich et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1185–1188
studies were conducted using a fluorescence polarization
assay using a xanthene-ADAMTS-5 hydroxamic acid
inhibitor.
In summary, we have presented a series of aryl thi-
oxothiazolidinones with varying activities as inhibitors
of ADAMTS-5. This series of compounds has tractable
SAR for both the thioxothiazolidinone aryl substituents
and zinc chelator. The continued development of
ADAMTS-5 inhibitors using both zinc chelating and
non-chelating scaffolds is currently ongoing and will be
reported in due course.
7. Skiles, J. W.; Gonnella, N. C.; Jeng, A. Y. Curr. Med.
Chem. 2004, 11, 2911.
8. See the accompanying letter for the related 5-((1H-
Pyrazol-4-yl)methylene)-2-thioxothiazolidin-4-one varient
of this new class of ADAMTS-5 inhibitors: Gilbert, A. M.;
Bursavich, M. G.; Lombardi, S.; Georgiadis, K. E.;
Reifenberg, E.; Flannery, C. R; Morris, E. A. Bioorg.
Med. Chem. Lett. 2007, doi:10.1016/j.bmcl.2006.12.020.
9. Cherney, R. J.; Mo, R.; Meyer, D. T.; Wang, L.; Yao, W.;
Wasserman, Z. R.; Liu, R.-Q.; Covington, M. B.; Torto-
rella, M. D.; Arner, E. C.; Qian, M.; Christ, D. D.;
Trzaskos, J. M.; Newton, R. C.; Magolda, R. L.; Decicco,
C. P. Bioorg. Med. Chem. Lett. 2003, 13, 1297.
10. Xiang, J. S.; Hu, Y.; Rush, T. S.; Thomason, J. R.; Ipek,
M.; Sum, P.-E.; Abrous, L.; Sabatini, J. J.; Georgiadis, K.;
Reifenberg, E.; Majumdar, M.; Morris, E. A.; Tam, S.
Bioorg. Med. Chem. Lett. 2006, 16, 311.
Acknowledgments
We thank Dr. John Ellingboe for support of this work
and Dr. Jeremy I. Levin for useful discussion in prepa-
ration of this manuscript. We thank Dr. Kristina Cunn-
ingham for performing fluorescence polarization
binding to target studies.
11. Yao, W.; Chao, M.; Wasserman, Z. R.; Liu, R.-Q.;
Covington, M. B.; Newton, R.; Christ, D.; Wexler, R.
R.; Decicco, C. P. Bioorg. Med. Chem. Lett. 2001, 12,
101.
12. Yao, W.; Wasserman, Z. R.; Chao, M.; Reddy, G.; Shi,
E.; Liu, R.-Q.; Covington, M. B.; Arner, E. C.; Pratta, M.
A.; Tortorella, M.; Magolda, R. L.; Newton, R.; Qian,
M.; Ribadeneira, M. D.; Christ, D.; Wexler, R. R.;
Decicco, C. P. J. Med. Chem. 2001, 44, 3347.
13. All compounds were characterized by reversed phase-
HPLC/MS spectroscopy. Selected compounds were also
characterized by 1H NMR.
14. Knoevenagel condensation afforded the Z isomer as
shown. However 1H NMR spectroscopy studies demon-
strated a time dependent Z to E isomeration in DMSO
solution from the pure Z upto an 80:20 Z:E isomeric ratio.
15. Hu, B.; Malamas, M.; Ellingboe, J. Heterocycles 2002, 57,
857.
References and notes
1. Buckwalter, J. A.; Martin, J. A. Adv. Drug Delivery Rev.
2006, 58, 150.
2. Wieland, H. A.; Michaelis, M.; Kirschbaum, B. J.;
Rudolphi, K. A. Nat. Rev. Drug Disc. 2005, 4, 331.
3. Liu, R. Q.; Trzaskos, J. M. Curr. Med. Chem.—Anti-
Inflammatory and Anti-Allergy Agents 2005, 4, 251.
4. Glasson, S. S.; Askew, R.; Sheppard, B.; Carito, B. A.;
Blanchet, T.; Ma, H.-L.; Flannery, C. R.; Kanki, K.;
Wang, E.; Peluso, D.; Yang, Z.; Majumdar, M. K.;
Morris, E. A. Arthritis Rheum. 2004, 50, 2547.
5. Stanton, H.; Rogerson, F. M.; East, C. J.; Golub, S. B.;
Lawlor, K. E.; Meeker, C. T.; Little, C. B.; Last, K.;
Farmer, P. J.; Campbell, I. K.; Fourie, A. M.; Fosang, A.
J. Nature 2005, 434, 648.
6. ADAMTS-5 activity was determined using a Fluorescence
Resonance Energy Transfer assay. Binding to target
16. Zuber, F.; Sorkin, E. Helv. Chim. Acta 1952, 35, 1744.