Design, synthesis and in vitro evaluation of potent, novel, small molecule inhibitors of plasminogen activator inhibitor-1

Article abstract of DOI:10.1016/S0960-894X(02)00078-1

We have synthesized and evaluated a series of tetramic acid-based and hydroxyquinolinone-based inhibitors of plasminogen activator inhibitor-1 (PAI-1). These studies resulted in the identification of several compounds which showed excellent potency against PAI-1. The design, synthesis and SAR of these compounds are described.

Full text of DOI:10.1016/S0960-894X(02)00078-1

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1063–1066
Design, Synthesis and In Vitro Evaluation of Potent, Novel, Small
Molecule Inhibitors of Plasminogen Activator Inhibitor-1
Adrian Folkes,a,* S. David Brown,b Lynne E. Canne,b Jocelyn Chan,b Erin Engelhardt,b
Sergey Epshteyn,b Richard Faint,a Julian Golec,a Art Hanel,b Patrick Kearney,b
James W. Leahy,b Morrison Mac,b David Matthews,b Michael P. Prisbylla,b
Jason Sanderson,a Reyna J. Simon,b Zerom Tesfai,b Nigel Vicker,a Shouming Wang,a
Robert R. Webbb and Peter Charltona
aXenova Limited, 957 Buckingham Avenue, Slough, Berks SL1 4NL, UK
bMetaxen, 280 E. Grand Avenue, So. San Fransisco, CA 94080, USA
Received 9 November 2001; revised 24 January 2002; accepted 28 January 2002
Abstract—We have synthesized and evaluated a series of tetramic acid-based and hydroxyquinolinone-based inhibitors of plasmi-
nogen activator inhibitor-1 (PAI-1). These studies resulted in the identification of several compounds which showed excellent
potency against PAI-1. The design, synthesis and SARof these compounds are described. # 2002 Elsevier Science Ltd. All rights
reserved.
Plasminogen activators (PAs) are serine proteases that
control the activation of the zymogen, plasminogen, to
the active enzyme plasmin.1 Plasminogen is converted to
proteolytically active plasmin by either tissue-type plas-
minogen activator (tPA) or urokinase-type plasminogen
activator (uPA).1 Plasminogen activator inhibitor (PAI-
1), a member of the serpin superfamily of protease
inhibitors,2,3 is the major physiological inhibitor of
plasminogen activators. Several studies have linked
increased PAI-1 activity with thromboembolic disease4,5
and elevated levels of PAI-1 in transgenic mice are
associated with severe venous thrombosis.6 PAI-1 is also
associated with a poor prognosis in a variety of can-
cers,7 and is believed to play a role in angiogenesis,8
invasion8 and metastasis.9
We have previously reported XR11211, which demon-
strated excellent potency (0.20 mM) in a mechanistic
plasmin generation assay which used the S2251 tripep-
tide as the chromogenic substrate.13 XR11211 also
retained this level of activity in a functional fibrinolysis
assay, and prevented complex formation between tPA
and PAI-1.13 However poor physicochemical properties
were undesirable features of this series, and to overcome
this issue we sought to identify alternative novel small
molecule inhibitors that incorporated key structural
features of the diketopiperazine template. This work
resulted in the discovery of several novel templates,
which demonstrated PAI-1 inhibitory activity including
the tetramic acid series and the hydroxyquinolinone
series. We describe here the design, synthesis and biolo-
gical activities of these series.
PAI-1 is produced in its active conformation, but can
spontaneously convert into a latent, inactive form that
is unable to bind to PAs.10 PAI-1 inhibition of PAs is
mediated through
a bait peptide bond (Arg346-
Met347), present on the reactive center loop (RCL),
which mimics the natural substrate for PAs, plasmino-
gen.11 The RCL is exposed in the active form, but is
inserted into the major b-sheet A in the latent con-
firmation.12
Our search for novel templates initially focused on the
diketopiperazine nucleus. It was speculated that the
biological activity of these compounds was exerted
through one of diketopiperazine ring amides existing in
the enol tautomer, whilst the other remained as a carbox-
amide.14 This led to an extensive search for five-, six- and
*Corresponding author. Tel.: +44-1753-706762; fax: +44-1753-706607;
e-mail: adrian_folkes@xenova.co.uk
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00078-1

1064
A. Folkes et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1063–1066
seven-membered rings that contained both enol and
amide functionalities. We also wanted to replace the
exocyclic double bond with a group that retained a
degree of rigidity, and yet would not be prone to iso-
merisation. Several ring systems were identified which
satisfied these criteria, and low levels of activity were
seen in some of these series. The tetramic acid series (A)
and the hydroxyquinolinone series (B) appeared to be
the most attractive of these templates, and we sought to
further develop these compounds.
tion of 3 with the appropriate amine (R00NH2) in
refluxing xylene yielded the desired compounds 4.
Scheme 2 outlines the route used to prepare hydroxy-
quinolinones. Substituted methyl anthranilates 5 were
either commercially available, or were prepared by
Suzuki–Miyaura coupling of the 4- and 5-bromo or
iodo substituted methyl anthranilates with the appro-
priate arylboronic acid.15 The 4- and 5-bromo or iodo
substituted methyl anthranilates in turn were prepared
by treatment of the appropriate isatoic anhydrides with
methanol. Isatoic anhydrides were prepared from the
appropriately substituted isatins.16,17 Compounds 5
were then reacted with methyl malonyl chloride to yield
6 in good yields. Dieckmann cyclisation using sodium
methoxide in methanol afforded 7, and treatment of 7
with the appropriate amine (R00NH2) in refluxing xylene
yielded the desired compounds 8.
The general synthetic route used to prepare tetramic
acid derivatives is outlined in Scheme 1. Compounds 1
were prepared via derivatisation of the amine RNH2 or
from the appropriate amino acids. Reaction with methyl
malonyl chloride yielded 2 in good yields, and Dieck-
mann cyclisation using sodium methoxide in methanol
afforded the 3-methoxycarbonyl tetramic acids 3. Reac-
Schemes 3 and 4 describe the routes used to synthesise
anilines 9 and 10 used in this work. Reaction between 4-
nitrophenol and 8-bromo-octanoic acid methyl ester,
followed by reduction of the nitro group yielded 9 in
good yield. To prepare 10, 4-nitrophenol was alkylated
Scheme 1.
Scheme 2.
Scheme 3.
Scheme 4.

A. Folkes et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1063–1066
1065
with 1,7-dibromoheptane and then treated with potas-
sium cyanide to yield the corresponding nitrile. Reac-
tion with sodium azide followed by reduction of the
nitro group yielded 10 in satisfactory yield.
possibly due to a steric interaction affecting the con-
formation of the phenyl ring. 3- and 3,4-substituted
compounds (24, 29) gave intermediate activity.
Replacement of the terminal carboxyl acid group with
the isosteric tetrazole group resulted in compounds of
Compounds containing an ester functionality (e.g., 30)
were hydrolysed to their corresponding carboxylic acids
(e.g., 17) using NaOH in H2O/THF/MeOH.
Table 2. PAI-1 inhibitory activity for compounds 17–18, 30–32
Compounds were tested in a PAI-1 mechanistic assay in
which the ability of the compound to inhibit PAI-1 was
quantified by measuring the activity of tPA using a
chromogenic substrate.18,19
Compd
RY
IC
(mM)
50
The versatile routes utilised in the synthesis of these
series allowed for a diverse array of analogues to be
prepared. It was rapidly established that in each of the
series, introduction of the octanoic acid side chain, as
used in XR11211, resulted in an improvement in activ-
ity. This result suggests that the tetramic acid series and
the hydroxyquinolinone series inhibit PAI-1 in a similar
manner to XR11211.
17
30
31
18
32
4-Cl–Ph
4-Cl–Ph
4-Cl–Ph
4-Me–Ph
4-Me–Ph
CO2H
CO2Me
Tetrazole
CO2H
5.00ꢀ3.00
>50
0.69ꢀ0.06
1.95ꢀ0.37
0.89ꢀ0.17
Tetrazole
Table 3. PAI-1 inhibitory activity for compounds 33–56
Once it had been established that the octanoic acid side
chain was beneficial for activity within the tetramic acid
series, our investigations focused on exploration of
0
groups Rand R (Table 1). It was found that substitu-
tion at R0 was not well tolerated as compounds with
methyl (12), phenyl (11) and substituted benzyl (13) in
this position showed poor activity, regardless of the
substitution at R. At the R position, benzyl (15) was not
tolerated but introduction of a phenyl group (14) to this
position gave a compound of modest activity. This ring
was then extensively explored with a wide variety of
substituents. This work showed that substitution in the
4-position by small lipophilic groups gave a significant
improvement in activity (e.g., 16, 17, 18, 19, 21). Sub-
stitution in the 2-position was not well tolerated (25, 27)
Compd
X
IC50 (mM)
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Unsubstituted
1.85ꢀ0.21
7.0 (n=1)
7.9ꢀ0.28
4.25ꢀ1.1
3.0ꢀ1.4
5-I
5-NO2
5-NH2
5-Cl
6-Cl
6-F
6-NO2
6-NH2
6-I
6-nC5H11
1.95ꢀ1.34
4.9ꢀ1.27
3.6ꢀ0.57
>20
2.2ꢀ0.14
1.80ꢀ0.42
2.00ꢀ0.7
0.67ꢀ0.13
0.51ꢀ0.14
0.50ꢀ0.14
2.35ꢀ0.50
1.65ꢀ0.07
0.30ꢀ0.14
1.03ꢀ0.06
0.90ꢀ0.28
>30
Table 1. PAI-1 inhibitory activity for compounds 11–29
6-(4-Cl–Ph)
6-(2-Naphthyl)
6-(Benzo[b]thiophen-3-yl)
6-(benzo[b]thiophen-2-yl)
6,7-diF
6,7-diMeO
7-(Benzo[b]thiophen-2-yl)
7-CF3
0
Compd
RR
IC50 (mM)
7-CN
7-NH2
7-(2-Naphthyl)
8-(1-Naphthyl)
8-(4-MeO–Ph)
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
H
Ph
H
Ph
Ph
Me
49.0
>50
0.60ꢀ0.14
4-OH–PhCH2
>50
21ꢀ13
>50
30.0ꢀ7.1
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Bn
>50
4-OMe–Ph
4-Cl–Ph
4-Me–Ph
4-Br–Ph
4-NO2–Ph
4-iPr–Ph
4-Morpholino–Ph
4-NHAc–Ph
3,4-diCl–Ph
1-Naphthyl
2-Naphthyl
2-Cl–Ph
2-OMe
3.66ꢀ1.20
5.00ꢀ3.00
1.95ꢀ0.37
1.99ꢀ0.29
4.25ꢀ0.96
2.35ꢀ0.45
>50
Table 4. PAI-1 inhibitory activity for 47, 57
>50
13.0ꢀ7.0
>50
4.00ꢀ1.00
>50
>50
Compd
X
Y
IC50 (mM)
47
57
6-(Benzo[b]thiophen-2-yl)
6-(Benzo[b]thiophen-2-yl)
CO2H
Tetrazole
0.50ꢀ0.14
1.01ꢀ0.34
3-CO2Me–Ph
17.9

1066
A. Folkes et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1063–1066
Table 5. Comparison of selected compounds in the complex and
chromogenic assays
standing of PAI-1 inhibition through further mechan-
istic studies.
Compd
Complex assay
IC50 (mM)
Chromogenic assay
IC50 (mM)
Acknowledgements
XR11211
0.51ꢀ0.08
0.83ꢀ0.26
0.30 (n=1)
0.94ꢀ0.41
0.29ꢀ0.15
0.20ꢀ0.015
0.89ꢀ0.17
0.69ꢀ0.06
0.51ꢀ0.14
1.01ꢀ0.34
32
31
46
57
We are grateful for the excellent support we received
from the following colleagues: Stephen Wren, Luke
Burgess, Hayley Farmer, Debbie Bootle and Ajay Jain.
marginally improved potency (Table 2). Comparison of
17 with its corresponding methyl ester 30 clearly
demonstrates the importance of the acidic functionality.
References and Notes
1. Vassali, J. D.; Sappino, A. P.; Belin, D. J. Clin. Invest.
1991, 88, 1067.
2. Van Meijer, M.; Pannekoek, H. Fibrinolysis 1995, 9, 263.
3. Potempa, J.; Korzus, E.; Travis, J. J. Biol. Chem. 1994, 269,
15957.
4. Dawson, S.; Henney, A. Atherosclerosis 1992, 95, 105.
5. Aznar, J.; Estelles, A. Haemostasis 1994, 24, 243.
6. Erikson, L. A.; Fici, G. J.; Lund, J. E.; Boyle, T. P.; Polites,
H. G.; Marotti, K. R. Nature 1990, 346, 74.
7. Pappot, H.; Gardsvoll, H.; Rømer, J.; Pedersen, A.; Grøn-
Within the hydroxyquinolinone series it was also
quickly established that incorporation of the octanoic
acid side chain was beneficial for activity. Our efforts in
this series then focused on the introduction of various
groups to the fused aryl ring (Table 3).
The hydroxyquinolinone system tolerated a wide variety
of substituents in the 5-, 6- and 7-positions, although an
amino group in the 6- and 7-positions was not tolerated.
Substitution in the 8-position resulted in a loss of activ-
ity. Although many substituents were tolerated and
gave good activities, larger lipophilic groups such as
naphthyl (45, 54) and benzothiophene (46, 47, 50) in the
6- and 7-positions appeared to enhance potency.
Methylation of either amide NH abolished activity
(data not shown) while tetrazole replacement of the
carboxylic acid was also found to be tolerated in this
series (Table 4).
dahl-Hansen, J.; Pyke, C.; Brunner, N. Biol. Chem. Hoppe-
Seyler 1995, 376, 259.
¨
8. Bajou, K.; Noel, A.; Gerard, R. D.; Masson, V.; Brunner,
¨
N.; Holst-Hansen, C.; Skobe, M.; Fusenig, N. E.; Carmeliet,
P.; Collen, D.; Foidart, J. M. Nature Med. 1998, 4, 923.
9. Tsuchiya, H.; Katsuo, S.; Matsuda, E.; Sunamaya, C.;
Tomita, K.; Veda, Y.; Binder, B. Gen. Diagn. Pathol. 1995,
141, 41.
10. Sancho, E.; Declerck, P. J.; Price, N. C.; Kelly, S. M.;
Booth, N. A. Biochemistry 1995, 34, 1064.
11. Lindahl, T. L.; Ohlsson, P. I.; Wiman, B. Biochem. J.
1990, 265, 109.
12. Mottonen, J.; Strand, A.; Symersky, J.; Sweet, R. M.;
Danley, D. E.; Geoghegan, K. F.; Gerard, R. D.; Goldsmith,
E. J. Nature 1992, 355, 270.
13. Folkes, A.; Roe, M. B.; Sohal, S.; Golec, J.; Faint, R.;
Brooks, T.; Charlton, P. Bioorg. Med. Chem. Lett. 2001, 11,
2589.
14. The SARof the diketopiperazine series suggested a dis-
tinction between the roles of the two amides within the dike-
topiperazine ring, which led to our hypothesis. There was,
however, no direct evidence of enol formation within the
diketopiperazine ring.
15. Miyaura, N. Advances in Metal-Organic Chemistry, JAI:
Stamford, 1998, 6, 187.
16. Coppola, G. M. Synthesis 1980, 505.
17. Holt, S. J.; Sadler, P. W. Proc. R. Soc. London 1958,
148(B), 481.
18. The chromogenic substrate used was Pefachrome (Penta-
pharm).
19. Charlton, P. A.; Faint, R. F.; Bent, F.; Bryans, J.; Chi-
carelli-Robinson, I.; Mackie, I.; Machin, S.; Bevan, P.
Thromb. Haemost. 1996, 75, 808.
20. Nieuwenhuizen, W.; Laterveer, R.; Hoegee, B. Blood
Coagul. Fibrinolysis 1995, 6, 520.
The activities of 31, 32, 46 and 57 were confirmed in a
tPA/PAI-1 complex assay (Table 5). In the solid-phase
tPA/PAI-1 complex assay,20 pre-incubation of PAI-1
with an inhibitor prevents tPA/PAI-1 complex forma-
tion in a concentration dependant manner.
In summary, this work has shown that we have been
able to identify novel inhibitors of PAI-1 by incorpor-
ating certain key structural features present within the
diketopiperazine series. This has led to the discovery
and SARdevelopment of the tetramic acid series and
the hydroxyquinolinone series, several examples of
which (e.g., 31, 32, 46, 57) show high levels of potency
comparable to XR11211 in both the chromogenic and
complex assays. These novel series did show some
improvement in their physicochemical properties over
the diketopiperazine series, although this improvement
was not sufficient to warrant further development of
these series. However, the levels of potency attained
with these series have allowed us to increase our under-