Fused bicyclic Gly-Asp β-turn mimics with potent affinity for GPIIb-IIIa. Exploration of the arginine isostere

Article abstract of DOI:10.1016/S0960-894X(00)00008-1

6-[4-Amidinobenzoyl]amino]-tetralone-2-acetic acid is a potent antagonist of GPIIb-IIIa. Substitution in the meta position of the benzamidine, or replacement with a heteroaryl amidine was tolerated in this series. Use of an acyl-linked 4-alkyl piperidine

Full text of DOI:10.1016/S0960-894X(00)00008-1

Bioorganic & Medicinal Chemistry Letters 10 (2000) 385±389
Fused Bicyclic Gly-Asp ꢀ-Turn Mimics with Potent Anity for
GPIIb-IIIa. Exploration of the Arginine Isostere
Matthew J. Fisher, a,* Ulrich Giese, b Cathy S. Harms, a Michael D. Kinnick, a
Terry D. Lindstrom, a Je€erson R. McCowan, a Hans-Jurgen Mest, b
John M. Morin Jr., a Je€rey T. Mullaney, a Michael Paal, b Achim Rapp, b
Gerd Ruhter, b Ken J. Ruterbories, a Daniel J. Sall, a Robert M. Scarborough, c
Theo Schotten, b Wolfgang Stenzel, b Richard D. Towner, a Suzane L. Um, a
Barbara G. Utterback, a Virginia L. Wyss a and Joseph A. Jakubowski a
aEli Lilly and Company, Indianapolis, Indiana 46028, USA
bLilly Research Laboratories, Hamburg, Germany
cCOR Therapeutics, Inc., South San Francisco, California 94080, USA
Received 24 September 1999; accepted 17 December 1999
AbstractÐ6-[4-Amidinobenzoyl]amino]-tetralone-2-acetic acid is a potent antagonist of GPIIb-IIIa. Substitution in the meta posi-
tion of the benzamidine, or replacement with a heteroaryl amidine was tolerated in this series. Use of an acyl-linked 4-alkyl piper-
idine as an arginine isostere also provided active compounds. Compounds from this series provided substantial systemic exposure in
the rat following oral administration. # 2000 Elsevier Science Ltd. All rights reserved.
The search for a means to control thrombus formation
has consumed a signi®cant amount of resources during
the last decade. Two major themes have emerged from
this e€ort: modulation of the coagulation cascade,1 and
inhibition of platelet aggregation.2 The latter approach
has sought to control thrombosis by minimizing the
platelet's ability to copolymerize with the soluble
plasma protein ®brinogen.3 The interaction between
platelets and ®brinogen is known to be mediated by the
membrane bound integrin GPIIb-IIIa, and it is the
activated form of this receptor that binds to one of sev-
eral RGD sequences located on ®brinogen and other
proteins.4 A wide variety of peptides that contain an
RGD sequence can e€ectively inhibit the binding of
®brinogen to GPIIb-IIIa, and as a result, this tripeptide
motif has become the lead structural element used in
the design of novel non-peptide inhibitors of platelet
aggregation.5
of a potent cyclic RGD containing peptide. The data
indicated that the RGD sequence for the peptide stu-
died was constrained in a well de®ned Gly-Asp type II0
b-turn.6,7 We have shown that this structural element
can be e€ectively mimicked with an appropriately sub-
stituted 6,6-benzofused ring system (Fig. 1). Structure±
activity relationships de®ned for our early leads indi-
cated that e€ective antagonists could be obtained when
an ether-linked benzamidine was incorporated at the 6-
position and an acetic acid residue was attached at the
2-position of a variety of ring-systems.8 Linkage of the
benzamidine to the bicycle with an amide provided the
most potent derivatives as exempli®ed by compounds 1
and 2 (Fig. 1). A common feature found in the structure
activity evaluation of our Gly-Asp mimics was the use
of a benzamidine as an arginine isostere.9 While this
functional group provided the desired anity for
GPIIb-IIIa, we did not want to be limited to only one
e€ective arginine isostere. Therefore, we sought to
develop a series of chemically varied arginine isosteres
that would also provide equivalent or enhanced activity.
Towards this end, we investigated the e€ects that alter-
ing the phenyl portion of the benzamidine had on bio-
logical activity. We initially targeted compounds that
contained substitution meta to the amidine moiety since
the required intermediates were synthetically more
We have recently described our initial e€orts towards
the design and optimization of non-peptide antagonists
of GPIIb-IIIa which began with the structural evaluation
*Corresponding author. Tel.: +1-317-276-0632; fax: +1-317-276-
1417.
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00008-1

386
M. J. Fisher et al. / Bioorg. Med. Chem. Lett. 10 (2000) 385±389
Figure 1. Gly-Asp b-turn and 6,6-based mimics.
accessable, and ortho substituents (particulary halo10
groups) were not generally compatible with our syn-
thetic approach to amidine formation. Replacement of
the benzamidine with a heteroaryl amidine was also of
interest.
the known aniline 38 with cyano-arylcarboxylic acids
4a±f yielding amides 5a±f. The nitrile moiety in these
molecules was then transformed into a Boc-protected
amidine (6a±f) using a modi®ed thio-Pinner sequence
that we have recently described.6 These compounds
could either be N-deprotected with neat TFA, providing
the amidino-esters 7, or completely deprotected by ®rst
hydrolyzing the ethyl ester with ethanolic NaOH and
then N-deprotecting with neat TFA a€ording the
desired parent amidino-acids 8a±f as the TFA salts. The
required aryl carboxylic acids 4a±f were either commer-
cially available or prepared using conventional methods
as shown in Scheme 2. Compounds that contain an
acyl-linked, 4-substituted piperidine (13a±d) were pre-
pared by coupling the desired acid (9a±c) or chloro-
formate (10) with 3 in the presence of EDCI or pyridine,
respectively (Scheme 1). The formed adducts 11a±d were
deprotected with methods analogous to those described
for the preparation of compounds 7 and 8. The 4-sub-
stituted piperidines required for the preparation of these
compounds were either known in the literature11 or
prepared as outlined in Scheme 3. One example of an
Arginine isosteres other than benzamidine have also
found use in the preparation of RGD mimics. It has
been shown that 4-alkyl piperidines can be employed as
arginine isosteres with excellent results.11 While we have
previously shown that 6-oxo-linked 4-alkyl piperidines
could be employed in this series,6 the observed potency
a€orded by this structural motif was less than the cor-
responding benzamidine. Since the piperidine moiety
o€ers the possibility of providing an arginine isostere
with chemical features distinct from benzamidine, we
chose to reinvestigate the use of this structural motif in
the more potent amide linked series.
The general synthesis for the compounds disclosed in
this report is outlined in Scheme 1.12 Compounds that
contain an aryl amidine were prepared by ®rst coupling
Scheme 1. (a) EDCI; (b) H2S; MeI; NH4OAc; Boc2O; (c) TFA; (d) NaOH; TFA; (e) NaH; (f) H2/PdC.

M. J. Fisher et al. / Bioorg. Med. Chem. Lett. 10 (2000) 385±389
387
Scheme 2. (a) RuO4, NaOCl; (b) BuLi, CO2; (c) HONO, CuCN; (d) tBuLi, CO2; (e) TMSCl, NaCN.
amine-linked, 4-alkyl piperidine was prepared as shown
in Scheme 1. Alkylation of the Cbz protected analogue
of 3 (14) with the known bromide 15 provided inter-
mediate 16 in good yield. Removal of the Cbz group
with hydrogen and palladium provided secondary
amine 17, which was transformed into the target mole-
cule 18 using chemistry similar for the preparation of
compounds 8.
tetralone 1 (Table 1). Replacement of the chlorine with
¯uorine a€orded compound 8b, which was equipotent
with 1 in the ELISA (0.003 mM) and slightly less active
(0.1 mM) in PRP. Incorporation of a second ¯uorine
atom (8c) negatively impacted activity in PRP (0.24 mM)
while binding anity (ELISA) was una€ected. Incor-
poration of a tri¯uoromethyl group (8d) caused sub-
stantial erosion on both ELISA and PRP activity.
Exchange of the phenyl moiety for the similary sized
thiophene (8f) was detrimental for potency in general.
The pyridyl amidine (8e) a€orded good potency in
the ELISA (0.009 mM) and acceptable activity in PRP
(0.22 mM).
Biological activity was assessed in vitro by measuring
the agents' ability to inhibit the binding of ®brinogen to
puri®ed human GPIIb-IIIa in an ELISA type assay.13
Functional activity was subsequently determined by
measuring the inhibition of ADP (5 mM) induced plate-
let aggregation in human platelet-rich plasma (PRP).14
Analogues with sucient functional activity (PRP IC50
<0.5 mM) were candidates for oral evaluation in the
rat. For this assay, we were primarily interested in
determining the peak blood level (Cmax) and area
under the curve (AUC) a€orded by these compounds
for the 5 hour time course following a single oral dose of
10 mg/kg. Since we had previously demonstrated that
systemic exposure levels for related compounds in this
series were signi®cantly enhanced by conversion of the
parent amidino-acids to ethyl ester prodrugs, we chose
not to evaluate the corresponding carboxylates.8
Comparison of the data contained in Table 1 indicates
that substitution meta to the benzamidine (8a±d) gen-
erally results in a diminution of activity. Of the sub-
stituents examined, ¯uorine had the smallest e€ect and
slightly decreased activity in PRP relative to the non-
substituted tetralone 1. Introduction of a second ¯uoro
group had a greater negative impact on PRP activity
than the ELISA, suggesting that this modi®cation hin-
dered the translation of receptor anity into plasma
activity, possibly by increasing non-speci®c interactions
with other plasma constituents.7 Substitution of larger
groups such as chloro and tri¯uoromethyl, resulted in
more substantial losses in overall activity which may
suggest that steric bulk in this area of the antagonist
may not be tolerated by the receptor. The trend towards
lower activity is also observed for the heteroaryl
Analogue 8a, which contains a single chlorine atom
meta to the amidine residue, was 3-fold less active in
the ELISA and 6-fold less potent in PRP than the lead
Scheme 3. (a) Triethyl phosphonoacetate; (b) NaOH; (c) phosgene.

388
M. J. Fisher et al. / Bioorg. Med. Chem. Lett. 10 (2000) 385±389
Table 2. Piperidine analogues 18 and 13
Table 1. Amidine analogues 8
Compound
R
IC50 mMa (ELISA) IC50 mMb (PRP)
Compound
R
IC50 mMa (ELISA) IC50 mMb (PRP)
18
0.61
5.5
8a
0.011
0.003
0.005
0.072
0.4
0.1
0.24
2.0
13a
0.071
0.91
8b
8c
8d
13b
13c
13d
0.026
1.2
0.36
18
0.33
1.8
8e
8f
0.009
0.21
0.22
9.0
a,bAs in Table 1.
in these piperidine analogues. Introduction of an oxy-
gen atom into the tether connecting the piperidine and
tetralone moieties as in carbamate 13c and ether 13d
resulted in a decrease in activity relative to 13a.
aConcentration required to reduce binding of ®brinogen to puri®ed
human GPIIb-IIIa by 50%. The IC50 values are expressed as the
average of at least two determinations. The average error for the IC50
determinations was 15%.
bConcentration required to reduce ADP-induced human platelet
aggregation response by 50%. The IC50 values are expressed as the
average of at least two determinations. The average error for the IC50
determinations was 16%.
Compounds 8a±c, 8e and 13b displayed sucent activity
in vitro to warrant further examination in vivo. As dis-
cussed, we chose to evaluate the ethyl ester derivatives
of these compounds (7a±c, 7e and 12b, respectively)
orally in the rat. As shown in Table 3, all of the com-
pounds a€orded good blood levels of parent amidino-
acid following an oral dose of 10 mg/kg of the corre-
sponding ester prodrug. The data demonstrate that
incorporation of a single halogen meta to the amidine
moiety was well tolerated and compounds with this
modi®cation provided exposure levels similar to the
non-substituted parent. Di¯uoro compound 8c pro-
vided approximately half of the exposure of the mono-
halo and unsubstituted analogues. Heteroamidine 8e
provided the least exposure of the amidine series with
Cmax and AUC values approximately 3-fold less than
tetralone 1. Piperidine analogue 12b provided moderate
blood levels but overall exposure was less than the best
benzamidine containing analogues. While the exposure
levels varied with substitution, it is of interest to note
that each analogue a€orded peak concentration levels
signi®cantly greater than the corresponding IC50 values
recorded in human plasma. This suggests that if these
compounds provided similar exposure in humans,
pharmacologicaly e€ective concentrations of active
metabolite (acid) in plasma could be achieved following
an oral dose of less than 10 mg/kg.
amidines 8e±f. Of the two examined, the 6-membered
pyridyl analogue was preferred to the 5-membered
thiophene.
Consistent with earlier ®ndings, replacement of the
benzamidine moiety with a more simple, amine-linked
4-alkyl piperidine (18) caused a substantial loss of
activity (Table 2). The amide-linked congener 13a was
8- and 6-fold more potent in the ELISA and PRP
assays, respectively. Introduction of unsaturation into
the tether connecting the piperidine and tetralone moi-
eties further increased activity in the ELISA (0.026 mM)
and PRP (0.36 mM), which gives this analogue similar
activity to some of the amidine containing compounds
listed in Table 1. The increased potency a€orded by
amide 13a and unsaturated amide 13b relative to amine
18 demonstrates that incorporation of conformational
constraint in the tether can be used to enhance activity

M. J. Fisher et al. / Bioorg. Med. Chem. Lett. 10 (2000) 385±389
389
Table 3. Blood levels obtained after a single oral dose of compounds
1, 8a±c, 8e and 13b
2. Samanen, J. Ann. Rep. Med. Chem. 1996, I, 91.
3. Pytela, R.; Pierschbacher, M. D.; Ginsberg, M. H.; Plow, E.
F.; Rouslahti, E. Science 1986, 231, 1559.
4. Bennett, J. S.; Shattil, S. J.; Power, S. W.; Gartner, T. K. J.
Biol. Chem. 1988, 263, 12948.
5. Ogima, I.; Chakravarty, S.; Dong, Q. Bioorg. Med. Chem.
1995, 3, 337.
6. Fisher, M. J.; Gunn, B.; Harms, C. S.; Kline, A. D.; Mul-
laney, J. T.; Nunes, A.; Scarborough, R. M.; Skelton, M. A.;
Um, S. L.; Utterback, B. G.; Jakubowski, J. A. J. Med. Chem.
1997, 40, 2085.
Compound
Cmax (ng/mL)
AUC (ng*h/mL)
1
3630
3108
4020
2025
1174
2700
11243
10391
11868
6414
3565
8137
8a
8b
8c
8e
13b
7. Fisher, M. J.; Gunn, B. P.; Harms, C. S.; Kline, A. D.;
Mullaney, J. T.; Scarborough, R. M.; Skelton, M. A.; Um, S.
L.; Utterback, B. G.; Jakubowski, J. A. Bioorg. Med. Chem.
Lett. 1997, 7, 2537.
In summary, we have investigated modi®cations of the
arginine isostere contained in lead tetralone 1. Sub-
stitution of a single ¯uorine atom meta to the benzami-
dine a€orded an analogue with similar activity to the
parent while substitution of additional ¯uorines, or lar-
ger functional groups resulted in compounds with
diminished activity. The trend towards lower activity
was also seen with the two heteroaryl amidines studied.
Conformationally restricted 4-alkyl piperidines were
found to provide good activity while the more ¯exible
analogues were less potent. Benzamidine and piperidine
containing compounds from this study provided sub-
stantial systemic exposure in rats after an oral dose of
10 mg/kg. The maximum concentrations found in
plasma for all of the compounds evaluated were well
above what would be required to inhibit human platelet
aggregation.
8. Fisher, M. J.; Arfstan, A. E.; Bruce, U.; Gunn, B. P.;
Harms, C. S.; Khau, V.; Kinnick, M. D.; Lindstrom, T. D.;
Martinelli, M. J.; Mest, H.-J.; Mohr, M.; Morin, Jr., J. M.;
Mullaney, J. T.; Nunes, A.; Paal, M.; Rapp, A.; Ruhter, G.;
Ruterbories, K. J.; Sall, D. J.; Scarborough, R. M.; Schotten,
T.; Sommer, B.; Stenzel, W.; Towner, R. D.; Um, S. L.;
Utterback, B. G.; Vasile€, R. T.; Voelkers, S.; Wyss, V. L.;
Jakubowski, J. A. J. Med. Chem. 1999, 42, 4875.
9. Zablocki, J. A.; Miyano, M.; Garland, R. B.; Pireh, D.;
Schretzman, L.; Rao, S. N.; Lindmark, R. J.; Panzer-Knodle,
S. G.; Nicholson, N. S. J. Med. Chem. 1993, 36, 1811.
10. Sawyer, J. S.; Schmittling, E. A.; Palkowitz, J. A.; Smith
III, W. J. J. Org. Chem. 1998, 63, 6338.
11. Hartman, G. D.; Egbertson, M. S.; Halczenko, W.; Las-
well, W. L.; Duggan, M. E.; Smith, R. L.; Naylor, A. M.;
Manno, P. D.; Lynch, R. J.; Zhang, G.; Cheng, C. T.-C.;
Gould, R. J. J. Med. Chem. 1992, 35, 4640.
12. All new compounds provided satisfactory spectroscopic
and analytical data.
13. Scarborough, R. M.; Rose, J. W.; Naughton, M. A.; Phil-
lips, D. R.; Nannizzi, L.; Arfsten, A.; Campbell, A. M.;
Charo, I. F. J. Biol. Chem. 1993, 268, 1058.
References and Notes
1. Wiley, M. R.; Fisher, M. J. Exp. Opin. Ther. Patents 1997,
7, 1265.
14. Jakubowski, J. A.; Maraganore, J. M. Blood 1990, 75, 399.