Parallel solid-phase synthesis of vitronectin receptor (αvβ3) inhibitors

Article abstract of DOI:10.1016/S0960-894X(00)00319-X

A combinatorial approach for rapid optimization of a vitronectin receptor (αvβ3) inhibitor lead was accomplished by solid-phase synthesis. Orthogonally bis protected 2,3-diaminopropionic acid was used to immobilize the C-terminus of the molecule. Selective deprotection and functionalization of the α-amino group followed by acyl resorcinol scaffold attachment and N-terminus diversification was used to explore structure-activity relationship (SAR). (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00319-X

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1715±1718
Parallel Solid-Phase Synthesis of Vitronectin Receptor
(ꢀvꢁ3) Inhibitors
Ariamala Gopalsamy,a,* Hui Yang,a John W. Ellingboe,a Kenneth L. Kees,b
Jeanne Yoonb and Richard Murrillsb
aWyeth-Ayerst Research, Pearl River, NY 10965, USA
bWyeth-Ayerst Research, Radnor, PA 19087, USA
Received 21 February 2000; accepted 30 May 2000
AbstractÐA combinatorial approach for rapid optimization of a vitronectin receptor (avb3) inhibitor lead was accomplished by
solid-phase synthesis. Orthogonally bis protected 2,3-diaminopropionic acid was used to immobilize the C-terminus of the mole-
cule. Selective deprotection and functionalization of the a-amino group followed by acyl resorcinol sca€old attachment and N-
terminus diversi®cation was used to explore structure±activity relationship (SAR). # 2000 Elsevier Science Ltd. All rights reserved.
Ligation of integrin avb3 within a 3-dimensional dermal
collagen matrix prevents apoptosis1 and promotes mela-
noma cell growth.2 Vitronectin receptor antagonists
have been shown to inhibit the growth of various solid
tumors of human origin.3 More recently, avb3 has been
shown to be involved in liver metastasis.4 avb3 Has been
shown to play a pivotal role in the proliferation and
migration of both smooth muscle and vascular endothelial
cells, a pathological process leading to restenosis after
balloon angioplasty.5 Various bone diseases involve bone
resorption that is mediated by only one known class of
cells, the osteoclasts. When activated for resorption, these
motile cells intially bind to bone, a process well known to
be mediated by avb3.6 It is also well known that blockade
of avb3 with antibodies or RGD containing peptides
block osteoclast cell adhesion and bone resorption in
vitro.7 More recently, several RGD peptidomimetics have
likewise been shown to inhibit osteoclasts in vitro and in
vivo and block bone loss in an ovariectomized rat.8
independently varied. Most of the reported10 potent
integrin inhibitors are RGD mimetics and a subset of
these inhibitors use the 2,3-diaminopropionic acid as the
C-terminus. While a cyclic or acyclic guanidino moiety is
preferred for the N-terminus, other groups such as ureas11
and amidines11 have been used as well. The central
sca€old, connecting these two pieces, can be varied12 also.
By developing a convenient route to appropriately pro-
tected fragments and a mild solid-phase synthesis that
incorporates all the components in an independent
fashion, it is possible to prepare combinatorial libraries
of this important class of integrin inhibitors.
Corbett et al.13 reported the ®rst solid-phase synthesis
of integrin antagonists, however, this synthesis does not
provide a means of varying the substitutions on the C-
terminus a-amino group of the 2,3-diaminopropionic
acid and is limited to the commercially available a-N-
CBZ-2,3-diaminopropionic acid as the only fragment. A
more recent solid-phase synthesis14 had employed the
a-N-Alloc-b-Fmoc-l-diaminopropionic acid to incorpo-
rate sulfonamide group at the a-position. However, we
were interested in a true combinatorial optimization of
the N-terminus region and the a-amino group of the 2,3-
diaminopropionic acid. The lead chosen for optimization
by solid phase parallel synthesis is shown in (Fig. 1).
Combinatorial chemistry is becoming an important tool
for drug discovery and lead optimization.9 A combina-
torial synthesis requires that at least two components of
the molecule be independently variable, so that all combi-
nations of these components can be prepared. Thus, to
prepare a combinatorial library of integrin inhibitors with
a high degree of potential diversity for rapid SAR gen-
eration using solid-phase techniques, it is important to
identify a synthesis in which various components can be
Chemistry
To facilitate selective functionalization of the a-amino
group the 2,3-diaminopropionic acid was orthogonally
*Corresponding author. Tel.: +1-914-732-2841; fax: +1-914-732-3045;
e-mail: gopalsa@war.wyeth.com
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00319-X

1716
A. Gopalsamy et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1715±1718
bis protected as b-N-dde-a-N-Fmoc-l-diaminopropionic
acid 1.15 It was attached to the Wang resin using HOBT/
HBTU as coupling reagents to give 2 and the a-amino
group was deprotected to the free amine 3 using 20%
piperidine in DMF. The amine 3 was reacted with a
number of chloroformates, isocyanates and acids or acid
chlorides to give the a-derivatized amine 4 as correspond-
ing carbamates, ureas or amides.16 The dde group on
the b-amino group was deprotected using 2% hydrazine
in DMF to give the free amine 5. The sca€old 6 was
synthesized as shown in Scheme 2 and coupled to amine
5 using DIC as the coupling agent to give the Fmoc pro-
tected acyl resorcinol17 derivative 7. The common inter-
mediate amine 8 was obtained by further deprotection
using 20% piperidine in DMF (Scheme 1).
The free amine 8 was used for diversi®cation of the N-ter-
minus. As shown in Scheme 3, it was converted to guani-
dine 9 by reaction with 1,3-bis-Boc-2-methyl-2-thiopseudo-
urea or to the dihydroimidazole derivative 10 using 2-(3,5-
dimethylpyrazolyl)-4,5-dihydroimidazole hydrobromide
salt or to amidine 11 by reaction with 1-aza-2-methoxy-
1-cycloheptene.
The intermediate amine 8 was also converted to a number
of ureas 12 by activation with p-nitrophenyl chloroformate
followed by treatment with amines. However, the N-pyr-
imidino derivatives 15 and N-tetrahydropyrimidino deri-
vatives 16 were obtained by direct coupling with amine
intermediate 5 with 13 or 14 (Scheme 4).
Figure 1.
Structure±Activity Relationship
Having established a means of diversifying both the C-
terminus amino substitution and the N-terminus, a matrix
of 112 carbamate analogues were generated18 to obtain a
rapid SAR. The ®rst array incorporated 14 di€erent car-
bamates for the C-terminus amino substitution and eight
di€erent N-termini. All the analogues were tested in avb3
binding assay19 and results are plotted in Chart A.
From Chart A it is very clear that cyclic guanidines and
guanidine seem to be the best N-terminus groups. While
N-terminus ureas are generally less potent, within the
ureas there seems to be a substitutent dependence.
Among the N-terminus ureas pyridyl ureas or simple
benzyl urea were found to be better than substitued ben-
zyl ureas. Among the a-substituents sterically crowded
substituent like neopentyl carbamate seems to be
more active than smaller methyl or ethyl carbamate
analogues.
Scheme 1. Reagents: (a) HOBT, HBTU, DIEA, DMF, rt, 2Â4 h; (b)
20% piperidine in DMF; (c) R1OCOCl, Et3N, DCM or R1NCO,
DCM or R1COOH, DIC, DMAP, DMF; (d) 2% hydrazine in DMF;
(e) 6, DIC, DMAP, DMF.
Similar arrays were generated using ureas and amides as
C-terminus a-amino substitution and the same set of N-
terminus substituents. Overall it was observed that
amides were less potent than carbamates and ureas were
less potent than the amides. However the trend within
the N-substituents was constant with the tetra-
hydropyrimidine being the best. All the active compounds
Scheme 2. Reagents: (f) DEAD, Ph3P, THF; 72% (g) KOH, dioxane:
water (1:1), 1N HCl; 88% (h) 4M HCl in dioxane, rt, 6 h; Fmoc-Osu,
Na2CO3, acetone-water (1:1), rt, 18 h; 92%.
Scheme 3. Reagents: (i) 1,3-bis-Boc-2-methyl-2-thiopsudourea, DIEA, DMF, rt, 18 h; (j) 2-(3,5-dimethylpyrazolyl)-4,5-dihydroimidazole.HBr,
DIEA, DMF; (k) 1-aza-2-methoxy-1-cycloheptene, rt, 18 h; (l) p-Nitrophenyl chloro-formate, DIEA, DCM:THF (1:1), 0.5 h, rt; (m) RNH2, Et3N,
DMF, 2 h, rt; (n) 50% TFA in DCM, 0.5 h, rt.

A. Gopalsamy et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1715±1718
1717
Scheme 4. Reagents: (o) 2-bromopyrimidine, TMSCl, DIEA, dioxane, 2 days, 80 ꢀC; 55% (p) H2, Pd/C, HCl-HOAc; 80% (q) DIC, DMAP, DMF.
showed >100-fold selectivity over the related integrin
GPIIb/IIIA in a platelet aggregation assay.19
smaller or unbranched alkyl groups. Of the N-terminus
groups examined, cyclic guanidines like tetra-
hydropyrimidine and dihydroimidazole were more potent
than guanidine itself. N-terminus ureas were found to be
less active than cyclic or acyclic guanidines. N-terminus
amidines and pyrimidines were found to be very weak
inhibitors. Further study in this area of integrin inhibitors
are in progress and will be reported in the future.
In summary, we have developed a solid-phase method for
generating libraries for RGD mimetic class of integrin
inhibitors. The method provides a facile means of varying
the substitution on the a-amino group as well as the N-
terminus groups. The methodology was used for generat-
ing lead optimization libraries of a vitronectin inhibitor
lead molecule using an acyl resorcinol sca€old. Structure±
activity information gathered from these libraries indicates
that a carbamate is more potent than a urea which in turn
is better than an amide for the a- substitution at the car-
boxy terminus. Of the lipophilic groups used, a sterically
crowded neopentyl group was found to be preferred over
Acknowledgements
We would like to thank Ms. G. Kha®zova and Mr. H.
Fletcher III for their help in scale up of some of the
intermediates.

1718
A. Gopalsamy et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1715±1718
33887. (b) Stilz, H.; Beck, G.; Radau, M. EP 811616. (c) Pitts,
W.; Jadhav, P.K. PCT Int. Appl. WO 99/50249.
References and Notes
1. Seftor, R. E. B.; Seftor, E. A.; Gehlsen, K. R.; Stetler-Ste-
venson, W. G.; Brown, P. D.; Ruoslahti, E.; Hendrix, M. J. C.
Proc. Natl. Acad. Sci. USA 1992, 89, 1557.
2. Montgomery, A. M.P; Reisfeld, R. A.; Cheresh, D. A.
Proc. Natl. Acad. Sci. USA 1994, 91, 8856.
13. Corbett, J. W.; Graciani, N. R.; Mousa, S. A.; DeGrado,
W. F. Bioorg. Med. Chem. Lett. 1997, 7, 1371.
14. Rockwell, A. I.; Rafalski, M.; Pitts, W. J.; Batt, D. G.;
Petraitis, J. J.; DeGrado, W. F.; Mousa, S.; Jadhav, P. K.
Bioorg. Med. Chem. Lett. 1999, 9, 937.
3. Brooks, P. C.; Montgomery, A. M. P.; Rosenfeld, M.;
Reisfeld, R. A.; Hu, T.; Klier, G.; Cheresh, D. A. Cell 1994,
79, 1157.
15. Compound 2 is commercially available from Nova-
Biochem. However, it was readily prepared from a-N-Fmoc-l-
diaminopropionic acid as shown below:
4. Yun, Z.; Menter, D. G.; Nicolson, G. L. Cancer Res. 1996,
56, 3103.
5. Choi, E. T.; Engel, L.; Callow, A. D.; Sun, S.; Trachtenberg,
J.; Santoro, S.; Ryan, U. S. J. Vasc. Surgery 1994, 19, 125.
6. Davies, J.; Warwick, J.; Totty, N.; Philp, R.; Helfrich, M.;
Horton, M. J. Cell. Biol. 1989, 109, 1817.
7. Horton, M. A.; Taylor, M. L.; Arnett, T. R.; Helfrich, M.
H. Exp. Cell Res. 1991, 195, 368.
16. When the reaction was carried out with some unreactive
or sterically hindered chloroformates or isocyanates migration
of the dde protecting group to the a-position leading to the b-
acylation was observed. This was however overcome by using
bulky ddiv following the recent report: Chabra, S. R.; Hothi,
B.; Evans, D. J.; White, P. D.; Bycroft, B. W.; Chan, W. C.
Tetrahedon Lett. 1998, 39, 1603.
8. (a) Engleman, V. W.; Nickols, G. A.; Ross, F. P.; Horton,
M. A.; Griggs, D. W.; Settle, S. L.; Ruminski, P. G.; Tei-
telbaum, S. L. J. Clin. Invest. 1997, 99, 2284. (b) Miller, W. H.;
Bondinell, W. E.; Ku, T. W. F. PCT Int. Appl. WO 98/30542.
(c) Drake, F.H. PCT Int. Appl. WO 98/15278.
9. (a) Lorsbach, B. A.; Kurth, M. J. Chem. Rev. 1999, 99,
1549. (b) Hermkens, P. H. H.; Ottenheijim, H. C. J.; Rees, D.
Tetrahedron 1997, 53, 5643. (c) Hermkens, P. H. H.; Otten-
heijim, H. C. J.; Rees, D. Tetrahedron 1996, 52, 4527. (d)
Booth, S.; Hermkens, P. H. H.; Ottenheijm, H. C. J.; Rees, D.
C. Tetrahedron 1998, 54, 15385. (e) Ellman, J. A. Acc. Chem.
Res. 1996, 29, 132.
10. (a) Duggan, M. E.; Hartman, G. D. PCT Int. Appl. WO
98/18460. (b) Duggan, M. E.; Hartman, G. D.; Ho€man, W.F.
PCT Int. Appl. WO 98/18461. (c) Hartman, G. D.; Duggan, M.
E. PCT Int. Appl. WO 98/18461. (d) Wehner, V.; Knolle, J.;
Stilz, H. U.; Gadek, T.; McDowell, R. EP 796855. (e) Chan-
drakumar, N.; Clare, M. PCT Int. Appl. WO 97/36861.
11. Ruminski, P. G.; Clare, M.; Collins, P. W. Desai, B. N.;
Lindmark, R. J.; Rico, J. G.; Rogers, T. E.; Russell, M. A.
PCT Int. Appl. WO 97/08145.
17. Acyl resorcinol was chosen as the sca€old since the pre-
sence of the -OH group ortho to the carbonyl group forms a
strong H bond (shown by NMR studies) it may obviate 2 of
the hydrating water molecules and improve bioavailability.
18. Isolated yield range 55±88% (5±22 mg) based on loading
as determined by the elemental analysis of resin bound amino
acid 2. All the compounds were puri®ed by reverse-phase
HPLC and further characterized by LC and MS for >90%
purity. LC Conditions: HP 1100, 23 ꢀC, 10mL injected; Col-
umn: YMC-ODS-A 4.6 Â 50 5m Gradient A: 0.05% TFA/
Water, B: 0.05% TFA/Acetonitrile; Time 0 & 1 min: 98%A &
2%B; 7 min: 10%A & 90%B; 8 min: 10%A & 90%B; 8.9 min:
98%A & 2%B; Post time 1 min; Flow rate 2.5 mL/min;
Detection: 215 and 254 nm, DAD.
19. For details on the binding and selectivity assay format see-
Kees, K. L.; Garrick, L. M.; Gopalsamy, A. PCT Int. Appl.
WO 99/52879.
12. (a) Jadhav, P.K.; Smallheer, J. PCT Int. Appl. WO 97/