Spirocyclic nonpeptide glycoprotein IIb-IIIa antagonists. Part 3: Synthesis and SAR of potent and specific 2,8-diazaspiro[4.5]decanes

Article abstract of DOI:10.1016/S0960-894X(02)00095-1

The synthesis and biological activity of analogues containing spiro piperidinylpyridine and pyrrolidinylpyridine templates are described. The potent activity of these compounds as platelet aggregation inhibitors demonstrates the utility of the spiro structures as central template for nonpeptide RGD mimics.

Full text of DOI:10.1016/S0960-894X(02)00095-1

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1103–1107
Spirocyclic Nonpeptide Glycoprotein IIb–IIIa Antagonists.
Part 3: Synthesis and SAR of Potent and Specific
2,8-Diazaspiro[4.5]decanes
Mukund M. Mehrotra,a Julie A. Heath,a Jack W. Rose,a Mark S. Smyth,a
Joseph Seroogy,a Deborah L. Volkots,a Gerd Ruhter,b Theo Schotten,b
Lisa Alaimo,a Gary Park,a Anjali Pandeya and Robert M. Scarborougha,*
aCOR Therapeutics, Inc., Departments of Medicinal Chemistry and Biology, South San Francisco, CA 94080, USA
bLilly Research Laboratories, Hamburg, Germany
Received 29 November 2001; accepted 23 January 2002
Abstract—The synthesis and biological activity of analogues containing spiro piperidinylpyridine and pyrrolidinylpyridine tem-
plates are described. The potent activity of these compounds as platelet aggregation inhibitors demonstrates the utility of the spiro
structures as central template for nonpeptide RGD mimics. # 2002 Elsevier Science Ltd. All rights reserved.
Over the past several years, potent and selective GPIIb–
IIIa antagonists (peptides and nonpeptides) have been
identified and developed, such as eptifibatide, tirofiban,
and abciximab, and are currently being used as intrave-
nous agents in the treatment of acute thrombosis.1
Efforts to discover orally active GPIIb–IIIa inhibitors
with appropriate pharmaceutical properties have also
been extensive.1 As part of a program to identify oral
antagonists with improved pharmaceutical properties,
we have focused on spirocyclic scaffolds, and have
identified 3,9-diazaspiro[5,5]undecane, and 3-aza-
spiro[5,5] undecanes as constrained central templates
suitable for preparing potent, conformationally-limited
GPIIb–IIIa antagonists such as CT51688 and
CT51819.2
The focus of the current study was to evaluate two new
series of 6,5-diazaspirocyclic containing templates (Fig.
1) with basic and acidic pharmacophore groups alter-
nately attached to five- and six-membered ring amines.
The synthesis of the 2,8-diaza-8-(4-pyridyl)-spiro
[4.5]decane (piperidinylpyridine nucleus) and its isomer,
the 2,8-diaza-2-(4-pyridyl)spiro[4.5]decane (pyrrolidinyl
Figure 1. The 2,8-diazaspiro[4.5]decane series.
Based on our initial success with these templates we
have also explored the 6,5-spirocyclic system such as the
2,8-diazaspiro[4.5]decane template for its potential in
obtaining potent orally active GPIIb–IIIa antagonists.
Scheme 1. Synthesis of dicarboxylic acid 1: (a) ethyl cyanoacetate,
Et3N, CH2Cl2, rt; ( b) KCN, EtOH–H2O, reflux, 18 h; (c) concd HCl,
reflux, 24 h.
*Corresponding author. Tel.: +1-650-244-6822; fax: +1-650-244-
9287; e-mail: rscarborough@corr.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00095-1

1104
M. M. Mehrotra et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1103–1107
pyridine nucleus) are shown in Schemes 1 and 2.3 Our
first goal was to explore the optimal distance and type
of linkage between the pyridyl and carboxylate func-
tionalities. These templates were coupled to various
carboxylic acid bearing segments via amide and urea
linkages, following the protocols described in Schemes
3, 4, 5, and 7. In general, the amide-linked analogues
were synthesized by reacting the tricyclic-spirocyclics
with either suitably protected carboxylic acids (via
peptide coupling), or with acid chlorides to provide
compounds 4–9, 27–32, and 36–38 (Schemes 3, 7, and 8).
The urea-linked analogues were synthesized by reacting
the spirocyclics with either corresponding isocyanates,
or p-nitrophenylcarbamates; for example, compounds
7–8 (Scheme 3), and compounds 11–16 (Scheme 5).
To determine whether more rigid conformational
restrictions within the linker would further modulate the
activity of antagonists, the linear carboxylic acid linkers
were substituted with several rigid cyclic, and aryl-acid
units (Schemes 5 and 7). The syntheses of these rigid lin-
kers, bearing the required carboxylic acid functionality,
Scheme 2. Synthesis of 2,8-diaza-8-(4-pyridyl)spiro[4.5]decane, and
2,8-diaza-2-(4-pyridyl)spiro[4.5]decane: (a) SOCl2; (b) p-methoxy-
benzyl amine, DIEA, CH3CN, rt; (c) NaOAc, Ac2O, reflux; (d) BH3/
THF, reflux; (e) CAN, 5% H2O, CH3CN; (f) (Boc)2O, NaOH, diox-
ane; (g) Pd(OH)2/C, MeOH, 100 psi H2; (h) 4-bromopyridine,
Pd2(dba)3, S-BINAP, t-BuONa, toluene, 95 ꢀC; (i) 50% TFA, CH2Cl2;
(j) 4-aminopyridine, DIEA, DMF, 100ꢀC.
Scheme 5. General synthesis of analogues containing 2,8-diaza-2-(4-
pyridyl)spiro[4.5]decane nucleus: (a) PNP–NH(CH2)nCO2Et, DIEA,
DCM; (b) 2 N HCl; (c) p-nitrophenyl (PNP)-derivative.
Scheme 3. Synthesis of analogues containing 2,8-diaza-8-(4-pyr-
idyl)spiro[4.5]decane nucleus: (a) ClCO(CH2)4À6CO2Et, DIEA, DCM;
(b) 2 N HCl; (c) 4-(2-ethoxycarbonyl-acetyl)-cyclohexanecarboxylic
acid, HOBt, DIEA, DMF; (d) PNP-derivative, DIEA, DCM.
Scheme 6. Synthesis of carboxylic acid segments: (a) NaH, DMF,
Br(CH2)nCO2Et; (b) Pd(OH)2/C, EtOH, 50 psi H2; (c) p-nitrophenyl
chloroformate, DIEA, DCM; (d) ClCO(CH2)nCO2Et, DIEA, DMF;
(e) sarcosine ethyl ester, NaB(OAc)3H, AcOH, DCM; (f) 10% Pd/C,
100 psi H2; (g) Br (CH2)2CO2Et, DIEA, DCM; (h) 10% Pd/C, 1 atm
H2; (i) ethyl acetoacetate, NaB(OAc)3H, AcOH, DCM; (j) ethyl mal-
onyl chloride, DIEA, DCM.
Scheme 4. (a) Ethyldiazoacetate, {Rh(OAc)2}2, fluorobenzene, rt;
(b) TFA; (c) 4-nitrophenyl chloroformate, DIEA, CH2Cl2, rt; (d) tri-
cyclic piperidine hydrochloride, DIEA, DMF, 70 ꢀC; (e) 2 N HCl, rt.

M. M. Mehrotra et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1103–1107
1105
are shown in Scheme 6. Appropriately protected
2-piperazinone, piperazines, pyrrolidines, were alkylated
or acylated, then deprotected and converted to the
required reactive synthons, p-nitrophenyl carbamates.
The carboxylic acid moieties required for analogues 19
and 29 were synthesized using the previously described
procedures.2 These segments were then coupled with the
spirocyclic units via amide and urea linkages (Schemes
5–8).
normal-chain carboxylic acids of varying lengths were
appended to the spirocyclic nucleus via amide linkage.
From this initial series, alkyl-linked analogue
5
appeared to exhibit the optimum required distance, but
displayed only moderate potency (Table 1).
To determine whether more rigid conformational
restriction within the linker would improve further upon
the activity, various rigid cyclic linkers were explored.
The incorporation of piperazine, piperidine and piper-
azinone moieties indeed yielded compounds with slight
enhancement in activity, compared to the acyclic ana-
logue 5 (Table 2). Analogues 7–10 were significantly
more potent in the binding assay than in platelet aggre-
gation, which could well be due to plasma protein
binding in the PRP assay, as has also been noted for
other GPIIb-IIIa inhibitors.4
As a continuation of our previous work utilizing spiro-
cyclic structures as central template for nonpeptide
RGD mimics,2 we have now investigated the pyridine
head-group analogues in the 2,8-diazaspiro-[4.5]-decane
class.
Each of the synthesized spirocyclic analogues was
assayed for its ability to inhibit the aggregation of
ADP-stimulated human platelets and to block the
binding of fibrinogen to purified human GPIIb–IIIa.
Based on our previous work,2 we utilized the less basic,
direct linked pyridine group in this series of 2,8-diaza-
spiro[4.5]decanes. In both piperidinyl–pyridine and
pyrrolidinyl–pyridine series, with the foremost goal of
determining the optimum distance between ‘the basic pyr-
idyl nitrogen’ and ‘the acidic carboxylate functionality’
Table 1. In vitro activity of analogues with the piperdinylpyridine
nucleus
Compd
R
ELISA
Fg/GPIIb–IIa
IC50 (mM)a
Human
PRP
IC50 (mM)a
4
5
6
CO(CH2)4CO2H
CO(CH2)5CO2H
CO(CH2)6CO2H
>50
80
1.3
18.8
0.043
0.196
aIC50 values expressed as the average of at least two determinations. The
average error for the determinations was Æ15%.
Table 2. Compounds with rigid acidic moiety in 6,5-spiro series
Scheme 7. Synthesis of urea and amide linked analogues 17–32: (a)
PNP-derivative, DIEA, DCM; (b) 2 N HCl; (c) carboxylic acids;
HOBt, DIEA, DMF.
Analogue
R
X
Y
Z
ELISA
Fg/GPIIb–IIa
IC50 (mM)a
Human
PRP
IC50 (mM)a
7
8
9
10
CH2CO2H
(CH2)2CO2H
COCH2CO2H
OCH2CO2H
N
N
C
N
N
N
C
O
O
H2
H2
0.025
0.238
0.113
0.153
1.18
8.34
1.58
1.04
N
aSee Table 1.
Table 3. In vitro activity of compounds with the pyrrolidinyl–pyri-
dine template
Compd
R
ELISA
Fg/GPIIb-IIa
IC50 (mM)a
Human
PRP
IC50 (mM)a
Scheme 8. (a) (i) NaH (2.3 equiv), DMF, 0 ꢀC, 2 h; (ii) bromo tert-
butyl acetate (1.0 equiv), rt, 20 h; (b) tricyclic piperidine hydrochlor-
ide, DIEA, HBTU, DMF; (c) TFA, rt; (d) (i) NaH (2.3 equiv), DMF,
0 ꢀC, 2 h; (ii) bromo ethylacetate (1.0 equiv), rt; (e) tricyclic piperidine
hydrochloride, DIEA, HBTU, DMF, rt; (f) 2 N HCl, rt; (g) 3,5-di-
methylisoxazole-4-sulfonylchloride, 1 N NaOH, Na2CO3, THF–H2O;
(h) tricyclic piperidinehydrochloride, DIEA, HOBt, CH2Cl2, rt;
(i) TFA, rt.
11
12
13
14
CONH(CH2)2CO2H
CONH(CH2)3CO2H
CONH(CH2)4CO2H
CONH(CH2)5CO2H
28.9
>50
11.9
34.6
80
10.6
74.0
>50
aSee Table 1.

1106
M. M. Mehrotra et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1103–1107
Table 4. Compounds with urea and amide linked rigid acid moiety
Table 6. Compounds with rigid heteroaromatic carboxylic acid
moiety
Compd
R
X
Y
n
ELISA
Fg/GPIIb–IIa
IC50 (mM)a
Human
PRP
IC50 (mM)a
Compd
R
ELISA
Fg/GPIIb–IIa
IC50 (mM)a
Human
PRP
IC50 (mM)a
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
CO2H
CH2CO2H
OCH2CO2H
CH2OCH2CO2H
N(Me)CH2CO2H
CH2CO2H
(CH2)2CO2H
(CH2)2CO2H
COCH2CO2H
COCH2CO2H
CO(CH2)2CO2H
NHCOCH2CO2H
COCH2CO2H
SO2CH2CO2H
CH2CH2CO2H
CH(Me)CH2CO2H
CH2CO2H
C
C
C
C
C
N
N
N
N
N
N
C
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
N
C
1
1
1
1
1
1
1
2
1
2
1
0
1
1
1
1
1
4.2
55
1.2
3.8
2.8
12.4
8.9
3.2
3.7
0.68
0.70
24.1
5.9
0.26
1.1
1.0
3.3
0.115
0.138
0.106
1.2
0.252
0.087
0.254
0.136
0.025
1.5
0.447
0.004
0.090
0.017
0.225
0.320
34
35
(CH2)2CO2H
CH2CO2H
1.80
0.047
23.3
0.95
aSee Table 1.
Table 7. In vitro activity for compounds with a-substitution
C
C
C
C
Compd
R
ELISA
Fg/GPIIb–IIa
IC50 (mM)a
Human
PRP
IC50 (mM)a
9.5
aSee Table 1.
36
37
38
CBz
SO2-isoxazol-4-yl
CO2Butyl
0.025
0.002
0.017
4.1
0.27
4.0
Table 5. Compounds with rigid aromatic carboxylic acid moiety in
5,6-spirocyclic series
aSee Table 1.
in the aggregation and binding assays, respectively
(Table 6). Also, the incorporation of a-amino substitu-
tion within the carboxylic acid segment in acyclic ana-
logues was found to significantly enhance the activity.
Compounds 36–38 (Table 7) exhibit 20- to 400-fold
more potency than 11–14 (Table 3).
Compd
R
ELISA
Fg/GPIIb–IIa
IC50 (mM)a
Human
PRP
IC50 (mM)a
32
33
m-OCH2CO2H
p-OCH2CO2H
0.873
0.688
18.3
11.3
Incorporation of the 3,5-dimethylisoxazol-4-yl sulfon-
amide functionalty (analogue 37) displayed 300-fold
potency enhancement in the platelet aggregation assay
and 1000-fold more potent in the binding assay relative
to its unsubstituted analogue 11 (Table 3). The com-
pound 37 (CT51810), a [5,6]-spirocyclic, is a ring-
contracted version of the potent [6,6]-spirocyclic analogue
CT51688.2 Incorporation of this specific sulfonamide
residue has been shown previously also to yield
extremely potent, tight-binding GPIIb–IIIa antago-
nists.2,5 The ꢀ-amino substituted analogue 37 is 10-fold
less potent in platelet aggregation assays compared to
the corresponding [6,6]-spirocyclic analogues CT51688
and CT51819 in [6,6]-azaspiro series. All the active
compounds described here were found to have IC50
values >100 mM against avb3 and thus are highly selec-
tive inhibitors of GPIIb–IIIa.
aSee Table 1.
The other series of analogues that we have investigated
involves the direct-linked pyrrolidinyl–pyridine tem-
plate, since this template, although isomeric to the pre-
vious one, would be expected to exist in a different
conformation. In this series, the straight alkyl-chain
linked analogues, 11–14 (Table 3), exhibited diminished
inhibitory activity in both the binding and aggregation
assays, compared to 5.
However, when rigid-linkers were utilized, with both the
urea and amide attachments, a much more profound
enhancing effect on the potency was observed for sev-
eral of the analogues 15–31 (Table 4), compared to the
small enhancement in activity observed for compounds
7–10 in the piperidinyl–pyridine series (Table 2). Most
of these analogues displayed ca. 30- to 100-fold potency
enhancement in PRP assay, and ca. 500-fold enhance-
ment in the binding assay, relative to the acyclic analo-
gues, 11–14 (Table 3).
The two most potent analogues 27 and 37, were further
studied in vivo in Sprague–Dawley rats, both as free
acid, and as their ethyl ester prodrugs. The calculated
oral bioavailability following oral and iv administra-
tions (1 mg/kg) of 27 was 10% with a t1/2 0.98 h.
By contrast, rigid aryl-acid linked analogues (32, 33,
and 34) displayed only modest potency (Tables 5 and 6),
with the exception of the indoleacetic acid derivative 35
which exhibited submicromolar and nanomolar activity
The oral bioavailability of its ethyl ester prodrug of 27
was surprisingly less (5.5%). The oral bioavailability of
37 was 12% with a plasma t1/2=0.38 h, and that of its

M. M. Mehrotra et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1103–1107
1107
ethyl ester prodrug ꢁ5%. In summary, highly active
GPIIb–IIIa inhibitors containing the 2,8-diazaspiro[4.5]
decane scaffold as a new template have been discovered.
However, the marginal pharmacokinetic properties of
these compounds preclude their further development.
3. (a) Claremon, D. A.; Liverton, N.; Baldwin, J. J. US Patent
5,451,578, 1995; Chem. Abstr. 1995, 124, 145932. (b) Carrera,
G. M.; Garvey, D. S. J. Heterocycl. Chem. 1992, 29, 847.
4. Sall, D. J.; Arfsten, A. E.; Berry, D. R.; Denney, M. L.;
Harms, C. S.; McCowan, J. R.; Ray, J. K.; Scarborough,
R. M.; Um, S. L.; Utterback, B. G.; Jakubowski, J. A. Bioorg.
Med. Chem. Lett. 1996, 6, 81.
5. (a) Xue, C.-B.; Wityak, J.; Sielecki, T. M.; Pinto, D. J.;
Batt, D. G.; Cain, G. A.; Sworin, M.; Rockwell, A. L.;
Roderick, J. J.; Wang, S.; Orwat, M. J.; Frietze, W. E.;
Bostrom, L. L.; Liu, J.; Higley, A.; Rankin, F. W.; Tobin,
A. E.; Emmett, G.; Lalka, G. K.; Sze, J. Y.; Di Meo, V.;
Mousa, S. A.; Thoolen, M. J.; Racanelli, A. L.; Hausner, E. A.;
Reilly, T. M.; DeGrado, W. F.; Wexler, R. W.; Olson, R. E.
J. Med. Chem. 1997, 40, 2064. (b) Su, T.; Naughton, M. A.;
Smyth, M. S.; Rose, J. W.; Arfsten, A. E.; McCowan, J. R.;
Jakubowsky, J. A.; Wyss, V. L.; Ruterbories, K. J.; Sall, D.;
Scarborough, R. M. J. Med. Chem. 1997, 40, 4308.
References and Notes
1. (a) Scarborough, R. M. J. Med. Chem. 2000, 43, 3453. (b)
Scarborough, R. M. Curr. Med. Chem. 1999, 6, 971.
2. (a) Smyth, M. S.; Rose, J.; Mehrotra, M. M.; Heath, J.;
Ruhter, G.; Schotten, T.; Seroogy, J.; Volkots, D.; Pandey, A.;
Scarborough, R. M. Bioorg. Med. Chem. Lett. 2001, 11, 1289.
(b) Pandey, A.; Seroogy, J.; Smyth, M. S.; Rose, J.; Mehrotra,
M. M.; Heath, J.; Ruhter, G.; Schotten, T.; Volkots, D.;
Scarborough, R. M. Bioorg. Med. Chem. Lett. 2001, 11, 1293.