Exploring the chiral space within the active site of α-thrombin with a constrained mimic of D-Phe-Pro-Arg - Design, synthesis, inhibitory activity, and X-ray structure of an enzyme-inhibitor complex

Article abstract of DOI:10.1016/S0960-894X(99)00669-1

An indolizidinone motif with strategically placed substitutents was designed and synthesized as a constrained mimic of D-Phe-Pho-Arg. Low nanomolar inhibition of α-thrombin validates the design elements in this inhibitor which also exhibits a 20-fold selectivity for thrombin versus trypsin. An X-ray crystal structure of the inhibitor with α-thrombin shows the expected interactions with key amino acids within the active site and some notable changes in positions. (C) Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(99)00669-1

Bioorganic & Medicinal Chemistry Letters 10 (2000) 243±247
Exploring the Chiral Space within the Active Site of ꢀ-Thrombin
with a Constrained Mimic of D-Phe-Pro-Arg Ð Design, Synthesis,
Inhibitory Activity, and X-ray Structure of an Enzyme±Inhibitor
Complex
Stephen Hanessian, a,* Elise Balaux, a Djorde Musil, b Lise-Lotte Olsson c
and Ingemar Nilsson d,y
a
Â
Â
Â
Â
Department of Chemistry, Universite de Montreal, CP 6128, Succursale Centre-ville, Montreal, Quebec, Canada H3C 3J7
È
bAstraZeneca R&D, Molndal, Structural Chemistry Laboratory, Molndal, Sweden
È
cDepartment of Inorganic Chemistry and Center for Structural Biology, Gothenburg University, Gothenburg, Sweden
dAstraZeneca R&D, Molndal, Medicinal Chemistry, Molndal, Sweden
È
È
Received 7 October 1999; accepted 22 November 1999
AbstractÐAn indolizidinone motif with strategically placed substitutents was designed and synthesized as a constrained mimic of
d-Phe-Pro-Arg. Low nanomolar inhibition of a-thrombin validates the design elements in this inhibitor which also exhibits a
20-fold selectivity for thrombin versus trypsin. An X-ray crystal structure of the inhibitor with a-thrombin shows the expected
interactions with key amino acids within the active site and some notable changes in positions. # 2000 Elsevier Science Ltd. All
rights reserved.
The enzyme thrombin, a member of the serine protease
family,1,2 plays a major role in thrombosis, which is one
of the leading causes of cardiovascular disease and
morbidity.3,4 Thrombin is involved in the blood coagu-
lation cascade following vascular injury, eventually
leading to cleavage of ®brinogen to ®brin and the acti-
vation of platelets.5 Cognizant of the potential bene®ts
in inhibiting thrombin, intense e€orts have been made
in search of e€ective inhibitors (for recent reviews, see
refs 6±8).
the S1 pocket, and an electrophilic serine-susceptible
group which mimics the oxyanion hole, and interacts
with the Ser-His-Asp catalytic triad of the enzyme. The
proximal S2-pocket, created primarily by Tyr 60A and
Trp 60D of the thrombin speci®c insertion loop, and the
distal S3-pocket, created mainly by the side-chains of
Leu 99, Ile 117 and Trp 215, are both of hydrophobic
character. Consequently potency is gained by the
presence of complementary lipophilic P2 and P3-resi-
dues of the inhibitors. These interactions are deduced
from X-ray crystal structure data for the complex of
thrombin with the chloromethylketone analogue of
d-Phe-Pro-Arg (PPACK) 1, which is believed to mimic
a thrombin-sensitive sequence in ®brinogen (Fig. 1).9
Of particular signi®cance was the elucidation of the
X-ray crystal structure of a number of enzyme±inhibitor
complexes.9±13 This invaluable information, which
allows synthetic chemists to literally `see' the interac-
tions of speci®c inhibitors within the active site of the
enzyme, has instigated a more logic-based approach to
the design of new analogues.1,2,14 In general, inhibitors
are designed to incorporate a basic guanidine-like
group, which interacts with Asp 189 in the bottom of
We reasoned that a constrained mimic consisting of an
indolizidinone nucleus15 with strategically situated sub-
stituents as in 2 would be an interesting analogue to
probe the ®delity of interactions within the enzyme
active site in spite of the absence of the activated elec-
trophilic site as in PPACK (Fig. 1). Superposition of the
energy-minimized structure 2 on a truncated d-Phe-Pro-
Arg backbone as in 3 showed excellent congruence.
Although numerous inhibitors of a-thrombin have been
synthesized with impressive activities,6±8 only a few
Keywords: thrombin inhibitor; enzyme-inhibitor complex; X-ray.
*Corresponding author. Tel.: +1-514-343-6738; fax: +1-514-343-
5728; e-mail: hanessia@ere.unmontreal.ca
ySecond corresponding author.
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(99)00669-1

244
S. Hanessian et al. / Bioorg. Med. Chem. Lett. 10 (2000) 243±247
Reformation of the enolate under the same conditions
and trapping with the Davis oxaziridine reagent27,28
led to 9 as a single isomer having the desired con®gura-
tion at C-3 (vide infra). Cleavage of the PMB group
with DDQ, Barton±McCombie deoxygenation,29 and
acetylation gave the indolizidinone 10. Deprotection of
the primary silyl ether, and oxidation of the alcohol
to the corresponding acid 11 proceeded uneventfully.
Deacylation and amide formation gave the desired
prototype 2.
In an e€ort to probe the chiral space at C-5 in the
context of structure 2, we considered maintaining the
oxygen functionality as a methyl ether as in 12. The
synthesis of this analogue was accomplished essentially
as described for 2 with equally good stereoselectivity
utilizing the methoxy group (Scheme 1). The structure
of 12 was con®rmed by single crystal X-ray analysis.
Figure 1. Structures of PPACK 1 (top left); prototype constrained
indolizidinone, 2 (top right); and a truncated PPACK (d-Phe-Pro-Arg,
3).
constrained bicyclic heterocyclic analogues have been
recently reported.10,16±22 The synthesis of 2 starts with
the readily available l-pyroglutamic acid 4, which was
transformed via six high-yielding steps23±25 into the
enantiopure precursor 5 (Scheme 1). Hydrogenation,
removal of the N-Boc group with B-bromocatechol
borane,26 lactam formation, and protection of the
resulting hydroxyl group led to the desired indolizidinone
7. The next challenge was to introduce the C-benzyl and
the tertiary hydroxyl groups at C-3 with the desired
stereochemistry. The a-orientation of the 5-p-methoxy-
benzyl (PMB) ether group was expected to play a stereo-
directing role in electrophilic reactions of the lactam
enolate from its convex face. Indeed, after investigating
several bases, it was found that t-butyl lithium23±25 and
benzyl bromide gave the product 8 as a single isomer.
In order to study the importance of the hydrophobic
interaction of the P3 subsite, we chose to replace the
benzyl group by a methyl group as shown in Scheme 2.
Thus, 7 was subjected to a C-methylation±hydroxyl-
ation sequence in good overall yield to give 14. In this
series, hydroxylation of the lithium enolate was
achieved in the presence of trimethylphosphite and
molecular oxygen.30,31 The same procedure gave a
modest yield in the C-benzyl series (ex. 8!9, Scheme 1),
hence the use of the Davis oxaziridine reagent.27,28 The
subsequent steps followed the original protocol, which
consisted of deprotection, deoxygenation to give 15,
oxidation to 16, and coupling to give the intended
amide 17. The structures of compounds in this series
were corroborated by X-ray crystal analysis.
Scheme 1. (a) Ref 11, six steps, (85% overall); (b) 10% Pd/C, H2, EtOAc, 96%; (c) B-Bromocatecholborane, CH2Cl2, 71% (threo); (d) NaOMe,
MeOH, 98%; (e) PMBCl, Bu4NI, THF, 86%; (f) t-BuLi, BnBr, THF, 78 ꢀC, 88%; (g) t-BuLi, Davis' achiral oxaziridine, THF, 78 ꢀC, 63%; (h)
DDQ, CH2Cl2, H2O, 79%; (i) NaH, CS2, MeI, THF, then Ph3SnH, AIBN, toluene, re¯ux 30 min, 48%; (j) Ac2O, Pyr., DMAP, 82%; (k) TBAF,
THF, 60%; (l) TPAP, NMO, CH2Cl2, 60%; (m) t-BuOH, NaClO2, NaH2PO4, H2O, quant.; (n) Cbz-protected 4-aminomethylbenzamidine, BOP,
i-Pr2EtN, CH2Cl2, 80%; (o) MeONa, MeOH, quant.; (p) H2, 10% Pd/C, EtOAc, HCl, quant.; (q) MeI, Ag2O, MeCN, 86%.

S. Hanessian et al. / Bioorg. Med. Chem. Lett. 10 (2000) 243±247
245
Scheme 2. (a)t-BuLi, MeI, THF, 78 ꢀC, 97%; (b) t-BuLi, (MeO)3P, O2, THF, 78 ꢀC, 64%; (c) DDQ, CH2Cl2, H2O, 84%; (d) NaH, CS2, MeI,
THF, then Ph3SnH, AIBN, toluene, re¯ux, 30 min, 67%; (e) Ac2O, Pyr., DMAP, 92%; (f) TBAF, THF, 75%; (g) Swern oxidation, 84%; (h) t-
BuOH, NaClO2, NaH2PO4, H2O, 93%; (i) LiOH, H2O2, THF, 63%; (j) Cbz-protected 4-aminomethylbenzamidine, BOP, i-Pr2EtN, CH2Cl2, 82%;
(k) H2, 10% Pd/C, EtOAc, HCl quant.
Scheme 3.
In the quest for ®nding simpler and less substituted
analogues, we reasoned that an a-oriented 5-benzyl
ether group could possibly simulate the hydrophobic
site occupied by the 2-benzyl group in 2.32 In this
regard, we prepared the 5- and 3,5-substituted indolizi-
dinones 18, 19 and 20, 21, respectively, as shown in
Scheme 3 adopting the methods already described
above.
the a-thrombin±2 and PPACK X-ray structures. It is of
interest that whereas there is excellent overlap for the
indolizidinone and the benzamidine portion with
PPACK, there is considerable deviation of the C-benzyl
group in 2. Consequently, the Ile 174 residue deviates
from its original position compared to the PPACK-a-
thrombin structure. The deviation of the Glu 192
residue is due to the larger size of the benzamidine
group in 2 compared to the aliphatic counterpart in
PPACK (Figure 3, dotted arrows for PPACK, solid
arrows for a-thrombin±2).
Enzymatic Activity
The amide derivatives in the three series shown in
Schemes 1±3 were tested for their ability to inhibit a-
thrombin.33 While compounds 18±21 were less active
(pIC50=<4.9), it was most gratifying to ®nd that the
`designed' constrained analogue 2 was highly active
(pIC50=7.7; Ki thrombin=9 nM). The methoxy analo-
gue 13 and the 3-methyl analogue 17 showed pIC50=6.6
(Ki thrombin=125 nM).34 Preliminary tests showed a
20-fold selectivity for thrombin over trypsin with the
inhibitor 2. It is noteworthy to point out that the acyclic
dipeptide analogue d-Phe-Pro-p-amindinobenzylamine
showed excellent activity (K1 1.5 nM), and a 130-fold
selectivity for thrombin over trypsin.34
X-ray Crystallography of Thrombin±Inhibitor
Complexes
As an important structural feature, we were able to
obtain X-ray quality crystals of the a-thrombin±2 com-
plex. Shown in Figure 2, is the Connolly surface map of
the X-ray structure of the a-thrombin±2 complex at
1.95A resolution.35 Figure 3 depicts the superposition of
Figure 2. Connolly surface map of the X-ray structure of a-thrombin±
2 complex at 1.95 A resolution.

246
S. Hanessian et al. / Bioorg. Med. Chem. Lett. 10 (2000) 243±247
5. Colman, R. W.; Marder, V. J.; Salzman, E.; Hirsch, W.
In Hemostasis and Thrombosis: Basic Principles and Clinical
Practice; 3rd ed.; Colman, R. W.; Hirsch, J.; Marder, V. J.;
Salzman, E. W., Eds. J. B. Lippincott, Philadelphia, 1994,
p 3.
6. Sanderson, P. E. J.; Naylor-Olsen, A. M. Curr. Med. Chem.
1998, 5, 289.
7. Ripka, W. C. Curr. Opin. Biol. 1997, 1, 242.
8. Ripka, W. C.; Vlasuk, G. P. In Ann. Rep. Med. Chem.;
Birstol, J. A., Ed., Academic: San Diego, 1997; vol. 32, p 71.
9. Bode, W.; Mayr, I.; Baumann, U.; Uber, R.; Stone, S. R.;
Hofskeenge, J. Embo J. 1989, 8, 3467.
10. St-Charles, R.; Matthews, J. H.; Zhang, E.; Tulinsky, A. J.
Med. Chem. 1999, 42, 1376.
11. Wagner, J.; Kallen, J.; Ehrhardt, C.; Evenou, J.-P.;
Wagner, D. J. Med. Chem. 1998, 41, 3664.
12. Plummer, J. S.; Berryman, K. A.; Cai, C.; Cody, W. L.;
DiMaio, J.; Doherty, A. M.; Edmunds, J. J.; He, J. X.;
Holland, D. R.; Levesque, S.; Kent, D. R.; Narasimhau, L. S.;
Rubin, J. R.; Rapundalo, S. T.; Siddiqui, M. A.; Susser, A. J.;
St-Denis, Y.; Winecour, P. D. Bioorg. Med. Chem. Lett. 1998,
8, 3409.
13. Lee, K.; Jung, W.-H.; Park, C. W.; Hong, C. Y.; Kim, I.
C.; Kim, S.; Oh, Y. S.; Kwon, O. H.; Lee, S.-H.; Park, H. D.;
Kim, S. W.; Lee, Y. H.; Yoo, Y. J. Bioorg. Med. Chem. Lett.
1998, 8, 2563.
Figure 3. Connolly surface map of the X-ray structure of a-thrombin±
2 complex (yellow) aligned with PPACK (purple). Dotted arrows
indicate original positions with PPACK. Solid arrows show deviations
with 2.
14. Jiang, H.; Chen, K.; Tang, Y.; Chen, J.; Li, Q.; Wang, Q.;
Ji, R. J. Med. Chem. 1997, 40, 3085.
15. For a recent review, see Hanessian, S.; McNaughton-
Smith, G.; Lombart, H.-G.; Lubell, W. D. Tetrahedron 1997,
53, 12789.
16. Bachand, B.; DiMaio, J.; Siddiqui, M. A. Bioorg. Med.
Chem. Lett. 1999, 9, 913.
17. Boatman, P. D.; Ogbu, C. O.; Eguchi, M.; Kim, M.;
Nakanishi, H.; Cao, B.; Shea, J. P.; Kahn, M. J. Med. Chem.
1999, 42, 1367.
18. Obst, U.; Banner, D.; Weber, L.; Diederich, F. Chem. Biol.
1997, 4287.
19. Iwanowicz, E. J.; Lau, W. F.; Lin, J.; Roberts, D. G. M.;
Seiler, S. M. Bioorg. Med. Chem. Lett. 1996, 6, 1339.
20. Chirgadze, N. Y.; Sall, D. J.; Klimkowski, V. J.; Clawson,
D. K.; Briggs, S. L.; Hermann, R.; Smith, G. F.; Gi€ord-
Moore, D. S.; Wery, J.-P. Protein Sci. 1997, 6, 1412.
21. Tamura, S. Y.; Goldman, E. A.; Brunck, T. K.; Ripka, W.
C.; Semple, J. E. Bioorg. Med. Chem. Lett. 1997, 7, 331.
22. Salimbeni, A.; Paleari, F.; Canevotti, R.; Criscuoli, M.;
Lippi, A.; Angiolini, M.; Belvisi, L.; Scolastico, C.; Colombo,
L. Bioorg. Med. Chem. Lett. 1997, 7, 2205.
An examination of the crystal structure of the a-throm-
bin±2 complex reveals the importance of the functional,
topological, and stereochemical features of the designed
inhibitor (Figs 2 and 3). The choice of a slightly convex
indolizidinone motif with the b-hydrogen at C-6 based
on modeling studies is also validated. Unlike many
other thrombin inhibitors, compound 2 does not incor-
porate a `serine trap' as a critical functional motif for
activity. The presence of a tertiary a-hydroxy lactam
group rather than an amine as a H-donor±acceptor unit
for Gly 216 is another noteworthy functional feature.
Further re®nements can now be considered in the light
of these results, and they will be reported in due course.
Acknowledgements
23. Hanessian, S.; McNaughton-Smith, G. Bioorg. Med.
Chem. Lett. 1996, 6, 1567.
24. Hanessian, S.; Ronan, B.; Laoui, A. Bioorg. Med. Chem.
Lett. 1994, 4, 1397.
25. Hache, B. M.Sc. thesis, Universite de Montreal, 1996.
26. Boeckman, R. K.; Potenza, J. C. Tetrahedron Lett. 1985,
26, 1411.
27. Vishwakarma, L. C.; Stringer, O. D.; Davis, F. A. Org.
Syn. 1988, 66, 203.
We thank NSERCC for a scholarship to E. B. and
AstraZeneca Molndal for generous ®nancial support
through the Medicinal Chemistry Chair Program, and
Dr. Michel Simard for X-ray structure analyses of
compounds 12, 15 and 16. We thank Dr. Olle Karlsson
for the synthesis of some amidine reagents, and Dr. Jan
Christer Ulvinge and co-workers, AstraZeneca R&D,
Molndal for the biological tests.
28. Davis, F. A.; Vishwakarma, L. C.; Billmers, J. M.; Finn, J.
J. Org. Chem. 1984, 49, 3241.
29. Barton, D. H. R.; McCombie, S. W. J. Chem. Soc. Perkin
1, 1975, 1575.
References and Notes
30. Hartwig, W.; Born, L. J. Org. Chem. 1987, 52, 4352.
31. Wasserman, H. H.; Lipshutz, B. H. Tetrahedron Lett.
1975, 38, 1731.
1. Bohm, M.; Sturzebecher, J.; Klebe, G. J. Med. Chem. 1999,
42, 458.
2. Bode, W.; Brandstetter, H.; Mather, T.; Stubbs, M. T.
Thromb. Haemostasis 1997, 78, 501.
3. Stone, S. R. Trends Cardiovasc. Med. 1995, 5, 134.
4. Tapparelli, C.; Metternich, R.; Ehrhardt, C.; Cook, N. S.
Trends Pharm. Sci. 1993, 147, 366.
32. For examples of thrombin inhibitors based on 3-benzyloxy
proline as a bioisostere, see Klein, S. I.; Dener, J. M.; Molino,
B. F.; Gardner, C. J.; D'Alisa, R.; Dunwiddie, C. T.;
Kasiewski, C.; Leadley, R. J. Bioorg. Med. Chem. Lett. 1996,
6, 2225.

S. Hanessian et al. / Bioorg. Med. Chem. Lett. 10 (2000) 243±247
247
33. Nilsson, T.; Sjoling-Ericksson, A.; Deinum, J. J. Enzyme
Inhibition 1998, 13, 11.
34. Wiley, M. R.; Chirgadze, N. Y.; Clawson, D. K.; Craft, T.
J.; Gi€ord-Moore, D. S.; Jones, N. D.; Olkowski, J. L.; Weir,
L. C.; Smith, G. F. Bioorg. Med. Chem. Lett. 1996, 6, 2387.
35. The hirugen-thrombin structure was used in the re®nement
of the 2±thrombin structure, see Skrzypczak-Jankun, E.;
Carperos, V. E.; Ravichandran, K. G.; Tulinsky, A. J. Mol.
Biol. 1991, 221, 1379. Details available upon request.