Stereoselective synthesis of a potent thrombin inhibitor by a novel P2-P3 lactone ring opening

Article abstract of DOI:10.1021/jo035794p

The concise synthesis of a potent thrombin inhibitor was accomplished by a mild lactone aminolysis between an orthogonally protected bis-benzylic amine and a diastereomerically pure lactone. The lactone was synthesized by the condensation of L-proline met

Full text of DOI:10.1021/jo035794p

Ster eoselective Syn th esis of a P oten t Th r om bin In h ibitor by a
Novel P 2-P 3 La cton e Rin g Op en in g†
Todd D. Nelson,*,‡ Carl R. LeBlond,‡ Doug E. Frantz,‡ Louis Matty,§ J effrey V. Mitten,‡
Damian G. Weaver,‡ J effrey C. Moore,| J aehon M. Kim,| Russell Boyd,| Pei-Yi Kim,|
Kodzo Gbewonyo,| Mark Brower,| Michael Sturr,| Kathleen McLaughlin,|
Daniel R. McMasters, Michael H. Kress,‡ J ames M. McNamara,§ and Ulf H. Dolling§
Department of Process Research, Merck Research Laboratories, Merck & Co., 466 Devon Park Drive,
Wayne, Pennsylvania 19087, Department of Bioprocess R&D, Merck Research Laboratories, Merck & Co.,
Rahway, New J ersey 07065, Department of Process Research, Merck Research Laboratories, Merck & Co.,
Rahway, New J ersey 07065, and Department of Molecular Systems, Merck Research Laboratories,
Merck & Co., West Point, Pennsylvania 19486
todd_nelson@merck.com
Received December 8, 2003
The concise synthesis of a potent thrombin inhibitor was accomplished by a mild lactone aminolysis
between an orthogonally protected bis-benzylic amine and a diastereomerically pure lactone. The
lactone was synthesized by the condensation of L-proline methyl ester with an enantiomerically
pure hydroxy acid, which in turn was synthesized by a highly stereoselective (>500:1 er) and
productive (100000:1, S/C) enzymatic reduction of an R-ketoester. In addition, a second route to
the enantiomerically pure lactone was accomplished by a diastereoselective ketoamide reduction.
In tr od u ction
contains the common P1, P2, P3 structural motif found
in many potential thrombin inhibitors (Scheme 1).4
Thrombin is a trypsin-like protease enzyme that plays
a critical role in intrinsic and extrinsic blood coagulation.
As a result of the enzymatic activation of numerous
coagulation factors, thrombin is activated to cleave
fibrinogen, producing fibrin, which is directly responsible
for blood clotting. An imbalance between these factors
and their endogenous activators and inhibitors can give
rise to a number of disease states such as myocardial
infarction, unstable angina, stroke, ischemia, restenosis
following angioplasty, pulmonary embolism, deep vein
thrombosis, and arterial thrombosis. Thrombin is in-
volved in many of these processes and is the most potent
stimulator of platelet aggregation known. Thus, mol-
ecules that could selectively inhibit the formation of
thrombin or that could modulate the activity of thrombin
would have the potential to regulate many of the above-
mentioned disease states.1 Members of the class of
thromboembolitic agents (e.g., warfarin, hirudin, couma-
din, and bivalirudin2) are plagued by numerous limita-
tions. Consequently, the aggressive search for a potent,
selective, and bioavailable thrombin inhibitor is wide-
spread.3 An intensive effort by Merck has led to the
identification of thrombin inhibitor 1. This molecule
The existing route to 1 from the corresponding syn-
thons (P1, P2, and P3) resulted in a linear synthesis of
the target molecule and incorporated standard peptide
couplings and multiple protection/deprotection manipula-
tions.5 Accordingly, a more concise assembly of 1 was
desired, which was envisioned as arising via an aminoly-
sis of lactone P2-P3 with an appropriate P1 benzylic
amine. Herein, we describe the total synthesis of the
target thrombin inhibitor 1 utilizing this convergent
strategy.
Resu lts a n d Discu ssion
P 1 Syn th esis. The cornerstone of our strategy was
predicated on an efficient synthesis of the P2-P3 lactone
and its subsequent aminolysis with an appropriate P1
amine. This event would unmask the correct skeletal
framework of the target molecule. From the outset we
realized that one of the synthetic challenges would be
the differentiation between two benzylic amine positions
in P1. The existing synthetic approach to this orthogo-
nally protected, bis-benzylic amine counterpart utilized
a cyano ester, but the route was operationally cumber-
some.5 Thus, we initially chose to improve upon this
approach.
† Dedicated to Professor Gary H. Posner on the occasion of his 60th
birthday.
‡ Department of Process Research, Wayne, PA.
§ Department of Process Research, Rahway, NJ .
| Department of Bioprocess R&D.
(4) This nomenclature refers to the sequence D-Phe-Pro-Arg, which
mimics fibrinogen (thrombin’s natural substrate).
Department of Molecular Systems.
(5) Morrissette, M. M.; Stauffer, K. J .; Williams, P. D.; Lyle, T. A.;
Vacca, J . P.; Krueger, J . A.; Lewis, S. D.; Lucas, B. J .; Wong, B. K.;
White, R. B.; Miller-Stein, C.; Lyle, E. A.; Wallace, A. A.; Leonard, Y.
M.; Welsh, D. C.; Lynch, J . J .; McMasters, D. R. Bioorg. Med. Chem.
Lett. Manuscript submitted.
(1) (a) Vacca, J . P. Curr. Opin. Chem. Bio. 2000, 4, 394. (b)
Raghavan, S. A. V.; Dikshit, M. Drugs Future 2002, 27, 669.
(2) Gladwell, T. D. Clinical Therap. 2002, 24, 38.
(3) Fevig, J . M.; Wexler, R. R. Ann. Rep. Med. Chem. 1999, 34, 81.
10.1021/jo035794p CCC: $27.50 © 2004 American Chemical Society
Published on Web 05/05/2004
3620
J . Org. Chem. 2004, 69, 3620-3627

Stereoselective Synthesis of an Antithrombotic Agent
SCHEME 1
Realizing that palladium-based cyanations were prob-
lematic, we turned our attention to the stoichiometric
copper(I) cyanation (Rosenmund-von Braun reaction)10
of 2a . When 2a was heated with 1.1 equiv of CuCN in
DMF at 90 °C for 4 h, a clean transformation (>90%
assay yield) to 3a occurred with minimal over cyantion
to 3b (<0.4% by HPLC) on >200 g scale.11
Initial attempts to globally reduce a toluene stream of
cyano ester 3a to the corresponding amino alcohol 4
involved using excess lithium aluminum hydride in THF
at various reaction temperatures. In every case, these
reductions produced dark reaction mixtures with a
complex array of products including dechlorination of the
aromatic ring. Other reducing agents (i.e., BH3-DMS,
NaBH4, DIBAL-H) also provided unsatisfactory results.
Upon further investigation, we found that reductions that
12
utilized freshly prepared Zn(BH4)2 provided smooth
reduction of 3a to 4 with minimal byproduct formation
(eq 2).
The existing methodology to prepare cyano ester 3a
hinged on the palladium-catalyzed cyanation6 of methyl
2-bromo-5-chlorobenzoate (2a ) (eq 1). However, in our
hands, this was a capricious and tedious reaction that
resulted in significant amounts of over cyanated arene
3b. Although slight modifications offered initial improve-
ments and resulted in high yields of the corresponding
nitrile 3a (>92%), the final product was still contami-
nated with 3-5% of bis-nitrile 3b regardless of the source
of palladium used.7 In retrospect, this was not surprising
since the desired product 3a is a highly activated aryl
chloride itself, which is readily cyanated at the temper-
atures necessary for cyanation of the starting bromide
(80 °C).8
1
In this case, it was clear by H NMR and LC/MS data
of reaction mixtures that the ester was rapidly reduced
followed by slow reduction of the nitrile.13 Thus, addition
of a toluene stream of 3a to a THF solution of Zn(BH4)2
(2.0 equiv.) at 50 °C for 12-18 h gave >98% HPLC assay
yield of 4. Typical isolated yields starting were 60-65%
(from 2a ).14
Amino alcohol 4 was protected with Boc2O to afford
N-Boc arene 5 in a 93% yield (Scheme 2). Treatment of
(10) (a) Rosenmund, K. W.; Struck, E. Chem. Ber. 1919, 52, 1749.
(b) Von Braun, J .; Manz, G. Liebigs Ann. Chem. 1931, 448, 111. For
reviews see: (c) Ellis, G. P.; Romney-Alexander, T. M. Chem. Rev. 1987,
87, 779. (d) Lindley, J . Tetrahedron 1984, 40, 1433. (e) Ito, T.;
Watanabe, K. Bull. Chem. Soc. J pn. 1968, 41, 419.
(11) Two other impurities were identified as byproducts arising from
this process. One (0.4 LCAP) was identified as methyl 5-bromo-2-
cyanobenzoate. We attribute this impurity as a result of halogen
exchange of the desired product 4 with the CuBr formed during the
reaction. The other (1.2% LCAP) is the corresponding homocoupled
product 19.
In an attempt to circumvent this problem, we envi-
sioned that palladium-catalyzed cyanation of the corre-
sponding aryl triflate 2b would occur at a much lower
temperature and allow for selective monocyanation.9
Indeed, the cyanation of 2b proceeded well at 50 °C;
however, we were again plagued by over cyanation
(1-2%).
(12) For leading references on the synthesis and use of Zn(BH4)2,
see: (a) Narasimhan, S.; Balakumar, R. Aldrichim. Acta 1998, 31, 19.
(b) Narasimhan, S.; Madhavan, S.; Ganeshwar Prasad, K. J . Org.
Chem. 1995, 60, 5314.
(13) A significant byproduct in other reduction systems (especially
BH3-DMS) was the corresponding lactam 20, which we believe forms
from rapid nitrile reduction followed by ring closure on the unreduced
ester. Identifying a reducing agent that preferentially reduced the ester
functionality prior to the nitrile was vital in limiting formation of 20.
(6) (a) Tschaen, D. M.; Desmond, R.; King, A. O.; Fortin, M. C.; Pipik,
B.; King, S.; Verhoeven, T. R. Synth. Commun. 1994, 24, 887. (b)
Maligres, P. E.; Waters, M. S.; Fleitz, F.; Askin, D. Tetrahedron Lett.
1999, 40, 8193.
(7) Due to the fact that we were unable to reject the bis-nitrile 3b
impurity during the latter stages of the synthesis, a route that curbed
its formation was necessary.
(8) (a) Sundermeier, M.; Zapf, A.; Beller, M.; Sans, J . Tetrahedron
Lett. 2001, 42, 6707. (b) J in, F.; Confalone, P. N. Tetrahedron Lett.
2000, 41, 3271. (c) Akita, Y.; Shimazaki, M.; Ohta, A. Synthesis 1981,
974. (d) Anderson, Y.; Langstrom, B. J . Chem. Soc., Perkin Trans. 1
1994, 1395.
(9) For related procedures involving the Pd-catalyzed cyanation of
aryl triflates, see: (a) Drechsler, U.; Hanack, M. Synlett 1998, 1207.
(b) Anderson, B. A.; Bell, E. C.; Ginah, F. O.; Harn, N. K.; Pagh, L.
M.; Wepsiec, J . P. J . Org. Chem. 1998, 63, 8224.
(14) Crystallizing the amino alcohol at this point allowed us to
remove minor impurities that we were unable to reject downstream.
J . Org. Chem, Vol. 69, No. 11, 2004 3621

Nelson et al.
SCHEME 2
synthesized either by the addition of t-BuMgCl to di-
ethyl oxalate (10)23,24 or by the alkylation of 3,3-di-
methyl-2-oxobutanoic acid (11) with EtBr/DBU.25,26 Both
methods rapidly afforded ketoester 1227,28 in high yields
(Scheme 3).
The asymmetric reduction of 12 was accomplished with
the isolated enzyme KRED1001 (Scheme 4). KRED1001
(18) The conversion of 3,3-dimethyl-2-oxobutanoic acid (11) to (R)-
2-hydroxy-3,3-dimethylbutanoic acid has been reported to proceed with
a high level of enantioselectivity using the cell line Proteus vulgaris,
H2 gas, and benzyl viologen: (a) Schummer, A.; Yu, H.; Simon, H.
Tetrahedron 1991, 47, 9019. (b) Simon, H.; Bader, J .; Gu¨nther, H.;
Neumann, S.; Thanos, J . Angew. Chem., Int. Ed. Engl. 1985, 24, 539.
(19) The synthesis of (S)-2-hydroxy-3,3-dimethylbutanoic acid has
been accomplished by diazotization of L-tert-leucine: (a) Quast, H.;
Leybach, H. Chem. Ber. 1991, 124, 849. (b) Van Draanen, N. A.,
Arseniyadis, S.; Crimmins, M. T.; Heathcock, C. H. J . Org. Chem. 1991,
56, 2499. (c) Hartwig, W.; Schoellkopf, U. Liebigs Ann. Chem. 1982,
1952. This material can also be obtained from the racemate by the
classical resolution with either brucine (ref 10) or (S)-phenylethy-
lamine: (d) Zhang, W.-Y.; J akiela, D. J .; Maul, A.; Knors, C.; Lauher,
J . W.; Helquist, P.; Enders, D. J . Am. Chem. Soc. 1988, 110, 4652. (e)
Masamune, S.; Reed, L. A., III; Davis, J . T.; Choy, W. J . Org. Chem.
1983, 48, 4441. For other syntheses of racemic 2-hydroxy-3,3-dimeth-
ylbutanoic acid, see: (f) Enholm, E. J . Lavieri, S.; Cordo´va, T.;
Ghiviriga, I. Tetrahedron Lett. 2003, 44, 531.
an isopropyl acetate (IPAc) solution of 5 with TEA and
MsCl provided a 97:3 ratio of 6/7 after warming to room
temperature. The reaction was diluted with DMF, treated
with NaN3, and aged at ambient temperature for 24-48
h (as determined by the conversion of 6 and 7). Reaction
mixtures that contained elevated levels of 7 routinely
required longer reaction times (or higher DMF ratios) for
complete conversion. Washing the IPAc with water
afforded a clean stream of azide 8. The above one-vessel
protocol for transformation of alcohol 5 to azide 8 was
demonstrated on >200 g scale, affording a 93% overall
assay yield from 5 (8 in IPAc).
Azide 8 was next reduced to amine 9 with Lindlar’s
catalyst/H2.15 Initial hydrogenolysis screens were per-
formed in alcoholic solvents. Reductive dechlorination
was a significant competing side reaction under these
conditions. Reaction variables of solvent, temperature,
pressure, azide concentration, and catalyst loading were
screened for efficiency of transformation and suppression
of the des-Cl impurity. Optimal conditions were achieved
utilizing Lindlar’s catalyst (2-5 mol %) in IPAc, at 30-
40 °C, and 20-80 psi H2. The addition of TEA improved
the rate of reaction [1 equiv (no benefit observed at > 1
equiv)] as did the periodic purging of the reaction
atmosphere with fresh hydrogen (N2 removal). After
catalyst filtration, the direct isolation of 9 was ac-
complished by solvent removal and crystallization from
tri-n-propylamine.
P 2-P 3 Syn th esis. We have discovered two methods
for the stereoselective assembly of diastereomerically
pure lactone (S,R)-16. The first approach to lactone (S,R)-
16 centered on the synthesis of (R)-3,3-dimethyl-2-
hydroxybutyric acid [(R)-14]. Although this is a relatively
simple chiral hydroxy acid, and these types of molecules
have been commonly used as synthetic building blocks,16
only sparse reports on the synthesis of either antipode
of this molecule existed.17-21 To develop an efficient
approach to this compound, catalytic, asymmetric reduc-
tions of the corresponding R-ketoester 12 were investi-
gated.22 Ethyl 3,3-dimethyl-2-oxobutanoate (12) was
(20) Similarly, the asymmetric reduction of methyl 3,3-dimethyl-2-
oxobutanoate using stoichiometric amounts of borohydrides has been
used to form the corresponding (S)-carbinol: (a) Brown, H. C.; Cho, B.
T.; Park, W. S. J . Org. Chem. 1986, 51, 1. 3396. (b) Brown, H..; Pai, G.
G. J . Org. Chem. 1985, 50, 1384. This material has also been prepared
by a multistep synthesis: Ko, K.-Y.; Frazee, W. J .; Eliel, E. L.
Tetrahedron 1984, 40, 1333.
(21) For some leading references on other asymmetric syntheses of
simple monoalkylated hydroxy acids (and esters) of this type, see: (a)
Tang, L.; Deng, L. J . Am. Chem. Soc. 2002, 124, 2870. (b) Gawley, R.
E.; Campagna, S. A.; Santiago, M.; Ren, T. Tetrahedron Asymm. 2002,
13, 29. (c) Yu, H.; Ballard, E.; Wang, B.; Tetrahedron Lett. 2001, 42,
1835. (d) D´ıez, E.; Dixon, D. J .; Ley, S. V. Angew. Chem., Int. Ed. 2001,
40, 2906. (e) Aladro. F. J .; Guerra, F. M.; Moreno-Dorado, F. J .;
Bustamante, J . M.; J orge, Z. D.; Massanet, G. M. Tetrahedron Lett.
2000, 41, 3209. (f) Huerta, F. F.; Laxmi, Y. R. S.; Ba¨ckvall, J .-E. Org.
Lett. 2000, 2, 1037. (g) Chang, J .-W.; J ang, D.-P.; Uang, B.-J .; Liao,
F.-L.; Wang, S.-L. Org. Lett. 1999, 1, 2061. (h) Pansare, S. V.; Ravi, R.
G.; J ain, R. P. J . Org. Chem. 1998, 63, 4120. (i) Pandey, G.; Das, P.;
Reddy, P. Y. Tetrahedron Lett. 1998, 39, 7153. (j) Wang, Z.; La, B.;
Fortunak, J . M.; Meng, X.-J .; Kabalka, G. W. Tetrahedron Lett. 1998,
39, 5501. (k) Adam, W.; Lazarus, M.; Boss, B.; Saha-Mo¨ller, C. R.;
Humpf, H.-U.; Schreier, P. J . Org. Chem. 1997, 62, 7841. (l) Aggarwal,
V. K.; Thomas, A.; Schade. S. Tetrahedron, 1997, 53, 16213. (m)
Mashima, K.; Kusano, K.-h.; Sato, N.; Matsumura, Y.-i.; Nozaki, K.;
Kumobayashi, H.; Sayo, N.; Hori, Y.; Ishizaki, T.; Akutagawa, S.;
Takaya, H. J . Org. Chem. 1994, 59, 3064. (n) Davis, F. A.; Chen, B.-C.
Chem. Rev. 1992, 92, 919. (o) Corey, E. J .; Link, J . O. Tetrahedron
Lett. 1992, 33, 3431. (p) Potin, D.; Williams, K.; Rebek, J . Angew.
Chem., Int. Ed. Engl. 1990, 29, 1420. (q) Simon, E. S.; Plante, R.;
Whitesides, G. M. Appl. Biochem. Biotech. 1989, 22, 169. (r) Nakamura,
K.; Inoue, K.; Ushio, K.; Oka, S.; Ohno, A. J . Org. Chem. 1988, 53,
2589. (s) Evans, D. A.; Morrissey, M. M.; Dorow, R. L. J . Am. Chem.
Soc. 1985, 107, 4346. (t) Ramachandran, P. V.; Pitre, S.; Brown, H. C.
J . Org. Chem. 2002, 67, 5315.
(22) The asymmetric reduction of ethyl 3,3-dimethyl-2-oxobutanoate
using an (R)-diarylphosphine-ruthenium complex [i.e., SEGPHOS-
Ru(II)]/H2 (EtOH, 70 °C, 17 h) affords hydroxy ester (R)-13 in high
yield (99%) and ee (98.6%): Saito, T.; Yokozawa, T.; Ishizaki, T.; Moroi,
T.; Sayo, N.; Miura, T.; Kumobayashi, H. Adv. Synth. Catal. 2001, 343,
264.
(23) Rabjohn, N.; Harbert, C. A. J . Org. Chem. 1970, 35, 3726.
(24) J aeger, D. A.; Broadhurst, M. D.; Cram, D. J . J . Am. Chem.
Soc 1979, 101, 717.
(25) Rambaud, M.; Bakasse, M.; Duguay, G.; Villieras, J . Synthesis
1988, 566.
(26) Ono, N.; Yamada, T.; Saito, T.; Tanada, K.; Kaji, A. Bull. Chem.
Soc. J pn. 1978, 51, 2401
(27) For some alternative syntheses of keto ester 12, see: (a) Reetz,
M. T.; Heimbach, Schwellnus, K. Tetrahedron Lett. 1984, 25, 511. (b)
Seyferth, D.; Weinstein, R. M.; Hui, R. C.; Wang, W.-L.; Archer, C. M.
J . Org. Chem. 1991, 56, 5768. (c) de las Heras, M. A.; Vaquero, J . J .;
Garc´ıa-Navio, J . L.; Alvarez-Builla J . Org. Chem. 1996, 61, 9009. (d)
Wright, J . B. J . Org. Chem. 1955, 77, 4883.
(15) Corey, E. J .; Nicolaou, K. C.; Balanson, R. D.; Machida, Y.
Synthesis 1975, 590.
(16) Coppola, G. M.; Schuster, H. F. R-Hydroxy Acids in Enantiose-
lective Syntheses; VCH: Weinheim 1997.
(17) The classical resolution with cinchonidine has been reported,
but is low yielding and requires multiple crystallizations: Tanabe, T.;
Yajima, S.; Imaida, M. Bull. Chem. Soc. J pn. 1968, 41, 2178.
(28) For review articles on the synthesis of R-keto acids, see: (a)
Kova´cs, L. Recl. Trav. Chim. Pays-Bas 1993, 112, 471. ( b) Cooper, A.
J . L.; Ginos, J . Z.; Meister, A. Chem. Rev. 1983, 83, 321.
3622 J . Org. Chem., Vol. 69, No. 11, 2004

Stereoselective Synthesis of an Antithrombotic Agent
SCHEME 3
SCHEME 5
SCHEME 4
crystallized from heptane. This procedure was demon-
strated on > 200 g scale to provide lactone (S,R)-16 as a
white crystalline solid (73% isolated yield, >99 LCWP,
>99.5% de).
Lactone (S,R)-16 could also be accessed by a comple-
mentary method in which the carbinol stereochemistry
was installed by the substrate-controlled asymmetric
reduction of ketoamide (S)-17 (Scheme 5). Similar asym-
metric reductions of R-keto proline amides29 and R,â-
unsaturated proline amides30 have been reported to occur
with low levels of stereochemical induction.
Ketoamide (S)-17 was prepared in a single step (84%)
by coupling ketoacid 11 with L-proline methyl ester (EDC,
HOBT). An initial screen of catalytic hydrogenation
reducing agents indicated that Ru/C was superior to
Pt/C, Pd/C, or Rh/C for the substrate-controlled reduction
of the carbonyl group. The latter hydrogenation catalysts
resulted in decreased rate of carbonyl reduction.31 The
resulting mixture of diastereomeric hydroxy esters was
lactonized with catalytic TsOH to afford a 93:7 ratio of
lactones [(S,R)-16/(S,S)-16]. Moreover, it was observed
that the cyclization of hydroxy ester (S,R)-15 to the
is a commercially available ketoreductase that has been
used for the asymmetric reduction of â-ketoesters but,
to the best of our knowledge, has not been applied to the
asymmetric reduction of R-ketoesters. The reduction is
cofactor-dependent. Typical operating parameters uti-
lized ketoester 12 (45 mg/mL), NADPH (0.1-0.5 mg/mL),
glucose dehydrogenase (0.5-2 mg/mL), KRED1001 (0.1-2
mg/mL), and glucose (70 mg/mL) in an appropriate buffer
solution (e.g., phosphate, MOPS, etc.) at 25-40 °C with
the pH being maintained between 6 and 8. The source of
hydride for this ketone reduction was glucose. As moni-
tored by HPLC, the conversion is quantitative and occurs
with an extremely high degree of stereoselectivity (>500:
1; R/S). The hydroxy ester (R)-13 could be isolated as an
oil and then saponified to the corresponding enantio-
merically pure hydroxy acid (R)-14 without epimeriza-
tion. But in a more direct process, a simple pH adjust-
ment (pH >13) of the crude enzymatic reaction mixture
(post-reduction) immediately saponified the ethyl ester
(R)-13. After pH adjustment (ca. pH ) 2), the desired
product (R)-14 was extracted into ethyl acetate and the
KRED, GDH, NADPH, and glucose remained in the
aqueous phase. Removal of the EtOAc and crystallization
from heptane afforded the enantiomerically pure (R)-3,3-
dimethyl-2-hydroxybutyric acid (R)-14 in an 82% isolated
yield (>99.5% ee) and was successfully demonstrated on
a 200 g scale.
(29) (a) Munegumi, T.; Fujita, M.; Maruyama, T.; Shiono, S.;
Takasaki, M.; Harada, K. Bull. Chem. Soc. J pn. 1987, 60, 249. (b)
Boulmedais, A.; J ubault, M.; Tallec, A. Bull. Soc. Chim. Fr. 1989, 185.
(c) Byun, I. S.; Kim, Y. H. Synth. Commun. 1995, 25, 1963.
(30) Mallat, T.; Baiker, A. Appl. Catal. A: General 2000, 200, 3.
Prolinol-based keto amides (hemiacetals) have been reduced with good
stereoselectivity: Pansare, S. V.; Ravi, R. G. Tetrahedron 1998, 54,
14549.
(31) Leading references for examples of R-ketoamide reductions
include the following. Ru-clay: Aldea, R. Alper, H. J . Org. Chem. 1998,
63, 9425. Rh: Carpentier, J .-F.; Mortreux, A. Tetrahedron: Asymmetry
1997, 8, 1083. Pt/Al2O3: Wang, G.-Z.; Mallat, T.; Baiker, A. Tetrahe-
dron: Asymmetry 1997, 8, 2133. Ru: Chiba, T.; Miyashita, A.; Nohira,
H.; Takaya, H. Tetrahedron Lett. 1993, 34, 2351.
Amide formation between the enantiomerically pure
hydroxyacid and L-proline methyl ester (EDC, HOBT,
CH3CN) afforded a mixture of the hydroxy ester (S,R)-
15 and the lactone (S,R)-16 after aqueous workup.
Treatment of the mixture with catalytic TsOH in toluene
with the concomitant removal of MeOH from the system
afforded exclusively the lactone (S,R)-16, which was
J . Org. Chem, Vol. 69, No. 11, 2004 3623

Nelson et al.
F IGURE 1. Ester (S,R)-15 and (S,S)-1534 and lactone (S,R)-
16 and (S,S)-16 component profiles [HPLC area (mAU) vs
time].
F IGURE 3. Stereoscopic views of the structures of (S,R)-16
(top) and (S,S)-16 (bottom) optimized at the 6-31G** level.
SCHEME 6
F IGURE 2. Formation of lactones (S,R)-16 and (S,S)-16
[concentration (mM) vs time].
desired lactone (S,R)-16 occurred more rapidly than the
cyclization of the diastereomeric hydroxy ester (S,S)-15
to the lactone (S,S)-16.
To further probe this phenomenon, an equimolar
mixture of diastereomeric esters (S,R)-15 and (S,S)-15
was prepared by the EDC/HOBT coupling of rac-hydroxy
acid 14 and L-proline methyl ester HCl. These esters were
lactonized in toluene at 24 °C, with 10% TsOH. The
steady-state formation of lactone (S,R)-16 occurred in less
than 2 h.32 The molar rate of (S,R)-16 formation was more
rapid than the diastereomeric lactonization of hydroxy
ester (S,R)-15 to lactone (S,S)-16 [krel ) 17]. Component
profiles are provided in Figures 1 and 2. A similar bicyclic
lactam has been prepared by cyclization of an amino
a lower energy transition state or intermediate along the
lactonization pathway.35
La cton e Am in a tion a n d Com p letion of th e Syn -
th esis. The coupling of the two components [9 and (S,R)-
16] was accomplished in either triethylamine or 2-pro-
panol and occurred without the need for a catalyst or base
additive (Scheme 6). While many lactone aminolysis
reactions require more rigorous conditions or are facili-
tated by the addition of catalysts,36 the facile nature of
this amidation can be attributed to the inherent strain
in lactone (S,R)-16. Indeed, initial results indicate that
the rate of aminolysis with amine 9 is faster with lactone
(S,R)-16 than with the more stable diastereomeric lactone
(S,S)-16 [krel ) 2].
Other polar solvents screened (NMP, DMF, DIPEA,
CH3CN, THF) resulted in slightly slower coupling rates
or incompatibility with the lactone (MeOH, EtOH).
Alternatively, the lactone opening in THF could be
accelerated by performing the reaction at 40 °C in the
presence of HOAc. Regardless of the method for coupling
the two fragments, the subsequent workup included a
wash with 2 M citric acid to remove amine 9 followed by
33
group to a proline methyl ester tether.
A conformational search using molecular mechanics
(MMFFs, 4r distance-dependent dielectric) revealed two
conformers for each of the diastereomeric lactones; the
two conformers differed only in the five-membered ring
pucker. The lower-energy conformer of each stereoisomer
was optimized at the 6-31G** level of theory, and the
resulting structures are shown in Figure 3. The energy
of (S,R)-16 was found to be higher than that of (S,S)-16,
∆E ) +0.85 kcal/mol, which is probably due to the steric
repulsion caused by the axial t-Bu group.
The more rapid lactonization of the S,R isomer is
therefore likely to be due not to the relative stabilities of
the products, which in fact favors the S,S isomer, but to
(34) Ester (S,R)-15 and (S,S)-15 were not resolvable by HPLC.
(35) Consistent with this hypothesis is the finding that the lactone
hydrate, a model for a tetrahedral intermediate, is calculated at the
6-31G** level to be 2.2 kcal/mol lower in energy for the S,R isomer
than for the S,S isomer.
(36) Liu, W.; Xu, D. D.; Repic, O.; Blacklock, T. J . Tetrahedron Lett.
2001, 42, 2439.
(32) In this experiment, MeOH was not removed from the system;
therefore, the lactonization was not driven entirely to completion.
(33) Zhang, N.; Nubbemeyer, U. Synthesis 2002, 242.
3624 J . Org. Chem., Vol. 69, No. 11, 2004

Stereoselective Synthesis of an Antithrombotic Agent
added as antisolvent. Amino alcohol 4 was isolated by vacuum
filtration as an off-white crystalline solid (235 g, 70%): 1H
NMR (300 MHz, CDCl3) δ 7.35 (d, J ) 2.0 Hz, 1H), 7.25-7.17
(m, 2H), 4.59 (s, 2H), 3.96 (s, 2H) 4.10-2.60 (bs, NH2 and OH);
13C NMR (75.5 MHz, CDCl3) δ 143.3, 138.4, 133.7, 131.1, 130.2,
127.8, 64.2, 44.9. Anal. Calcd for C8H10ClNO: C, 55.99; H, 5.87;
N, 8.16; Cl, 20.66. Found: C, 55.95; H, 5.83; N, 8.18; Cl, 20.60.
a wash with 0.2 N NaOH (or saturated Na2CO3) to
hydrolyze and remove unreacted lactone (S,R)-16. This
process occurred without epimerization of either stereo-
center. Penultimate (S,R)-18 could be isolated as an
amorphous solid following solvent removal, but this
compound was typically utilized in the deprotection
without isolation. Numerous conditions were screened for
the unmasking of the benzylamine by N-Boc deprotection.
Problems ranged from sluggish reactivity to product
decomposition (spontaneous lactonization with extrusion
of the corresponding diamine). The optimized conditions
for N-Boc removal incorporated the addition of a 6 wt %
HBr (3 eq) solution to an anhydrous isopropyl acetate
(IPAc) solution of the substrate. This afforded the target
molecule HBr salt 1 as a white amorphous solid in an
overall 80% isolated yield from lactone (S,R)-16.
N-Boc Alcoh ol 5. To a stirred cooled (10 °C) slurry of amino
alcohol 4 (230 g, 1.4 mol) in toluene (2.5 L) was slowly added
a solution of Boc-anhydride (300 g, 1.4 mol) in toluene (500
mL). After complete addition, the reaction was allowed to
warm to ambient temperature and stirred until complete
consumption of the starting material was evident by HPLC
(ca. 6 h). The reaction was then concentrated to approximately
1/2 of its initial volume, at which point the product began to
precipitate. Heptane was added as antisolvent and the product
collected and dried by vacuum filtration as a crystalline white
1
solid (342 g, 93%): mp 92.1 °C; H NMR (300 MHz, CDCl3) δ
7.37 (bs, 1H), 7.33-7.20 (m, 2H), 5.50-4.80 (bs, 1H), 4.70 (s,
2H), 4.34 (s, 2H), 2.80-2.20 (bs, 1H), 1.44 (s, 9H); 13C NMR
(75.5 MHz, CDCl3) δ 156.1, 140.6, 135.5, 133.5, 130.7, 129.1,
128.3, 80.1, 62.7, 41.5, 28.4 (3C); IR (Nujol mull) 3387, 3316,
2925, 2854, 1682, cm-1. Anal. Calcd for C13H18ClNO3: C, 57.46;
H, 6.68; N, 5.15; Cl, 13.05. Found: C, 57.20; H, 6.71; N, 4.96;
Cl, 13.19.
Con clu sion s
In summary, the ring opening of a diastereomerically
pure bicyclic lactone with an orthogonally protected bis-
benzylic amine set the desired framework for a potent
thrombin inhibitor. The synthesis is highlighted by two
efficient syntheses of the diasteromerically pure lactone
(S,R)-16. The first synthesis relied on a highly stereose-
lective enzymatic reduction of an R-ketoester (>500:1 er)
while a complementary approach utilized a substrate-
controlled diastereomeric reduction of a proline keto-
amide.
N-Boc Am in e 9. To a solution of alcohol 5 (322.9 g, 1.19
mol) in isopropyl acetate (1292 mL) and triethylamine (241 g,
2.38 mol) at -15 °C was slowly added mesyl chloride (163.5 g,
1.43 mol) at such a rate as to maintain the internal temper-
ature below 0 °C. The resulting slurry was then warmed to
room temperature (HPLC assay of the slurry determined that
the reaction achieved a >99% conversion to 6 and 7). This
slurry was diluted with 1400 mL of DMF, and then sodium
azide (177.7 g, 2.73 mol) was added. After the mixture was
stirred at room temperature for 3 days, H2O (1000 mL) was
added and the resulting two-phase solution separated. The
organic layer was washed three additional times with water
and then dried. This solution of azide 8 was diluted to 2300
mL with isopropyl acetate followed by the addition of triethyl-
amine (111 g, 1.10 mol) and Lindlar’s catalyst (70.0 g) and
then hydrogenation (20-30 psig hydrogen, 35 °C, 23 h). The
catalyst was filtered through a plug of Solka Floc, and then
solvent was removed under vacuum (80-70 mmHg, 19-30 °C)
to about 0.5 the original volume. Tri-n-propylamine (350 mL)
was added and the concentration continued (50-59 mmHg,
38 °C) until distillation stopped. During the final distillation,
solids precipitated and the crystallization continued as the
IPAc was removed. Filtration afforded 242.1 g (75%) of amine
Exp er im en ta l Section
Am in o Alcoh ol 4.37 To a mechanically stirred heteroge-
neous mixture of CuCN (220 g, 2.4 mol) in DMF (2.5 L) under
an atmosphere of nitrogen was added methyl-2-bromo-5-
chlorobenzoate7 (550 g, 2.2 mol). The reaction was heated to
90 °C for 4 h and then cooled to 10 °C to give a bright yellow
heterogeneous mixture. Toluene (2.5 L) was added followed
by the slow addition of a 9:1 aqueous solution (2.5 L) of 10%
(w/v) NH4Cl/30% NH4OH to maintain the internal temperature
<25 °C. The flask was then exposed to air and the resulting
biphasic mixture stirred at ambient temperature for 12 h. The
deep blue aqueous layer was removed and an additional 2.5 L
of the 9:1 aqueous ammonia solution added and stirred for 2
h. This process was repeated until the aqueous layer was no
longer blue (typically three washes were needed). The clear
and colorless toluene layer was washed with brine (1.0 L) and
azeotropically dried (Kf < 100 ppm). HPLC assay of the final
toluene solution determined the yield to be 91% (390 g).
To a stirred slurry of ZnCl2 (260 g, 1.9 mol) in THF (2.0 L)
at ambient temperature was added a 2.0 M solution of LiBH4
in THF (2.0 L, 4.0 mol) under an atmosphere of nitrogen. A
mild exotherm was noticed during this addition. The resulting
mixture was then heated to 50 °C for 90 min followed by the
slow addition of the aforementioned toluene (2.0 L) solution
of methyl-5-bromo-2-cyanobenzoate (390 g, 2.0 mol) at a rate
to maintain the internal temperature <65 °C. Once addition
was complete, the reaction was allowed to proceed for 12 h at
60 °C. The resulting heterogeneous mixture was cooled to 10
°C at which point 3 N HCl (3.0 L) was added slowly to maintain
the internal temperature < 25 °C. This biphasic mixture was
then heated at 40 °C for 30 min and separated at that
temperature. Fresh toluene (2.5 L) was added to the acidic
aqueous layer and then cooled to 10 °C. Next, addition of 10
N NaOH was added slowly until the bottom aqueous layer was
at pH 12. The biphasic mixture was warmed to 40 °C, and the
layers were separated at this temperature. The toluene layer
was then concentrated in vacuo until a concentration of ca.
125 g/L was achieved and cooled to 5 °C, and heptane was
1
9 as a white solid: mp 77.7 °C; H NMR (300 MHz, CDCl3) δ
7.31-7.20 (m, 3H), 5.94 (bs, 1NH), 4.31 (bd, J ) 5.4 Hz, 2H),
3.89 (s, 2H), 1.58 (s, 2H), 1.44 (s, 9H); 13C NMR (75.5 MHz,
CDCl3) δ 155.8, 142.2, 135.9, 133.4, 130.8, 128.8, 127.5, 79.4,
43.9, 42.0, 28.5 (3C); IR (Nujol mull) 3358, 3291, 3169, 2925,
2854, 1696 cm-1. Anal. Calcd for C13H19ClN2O2: C, 57.67; H,
7.07; N, 10.35; Cl, 13.09. Found: C, 57.61; H, 7.19; N, 10.27;
Cl, 13.06.
Eth yl 3,3-Dim eth yl-2-oxobu ta n oa te (12) (Meth od A). To
a mixture of 3,3-dimethyl-2-oxobutanoic acid (11) (35 g, 83 wt
% by HPLC) and 260 mL of MTBE was slowly added DBU
(46.5 g). The addition was slightly exothermic. The reaction
was cooled to 37 °C, and bromoethane (57 g) was added and
stirred for 24 h at 38 °C. The reaction mixture was washed
with 1 N HCl and 10 wt % aq NaCl, dried over Na2SO4, and
concentrated in vacuo to afford 37.0 g of ketoester 12 as a light
yellow oil [97% corrected yield (93 wt % by 1H NMR, 92.5
LCWP), 89 LCAP]: bp 90 °C (40 mmHg); 1H NMR (300 MHz,
CDCl3) δ 4.30 (q, J ) 7.2 Hz, 2H), 1.34 (t, J ) 7.2 Hz, 3H),
1.24 (s, 9H); 13C NMR (75.5 MHz, CDCl3) δ 202.1, 163.9, 61.7,
42.6, 25.7 (3C), 14.1; IR (thin film) 2976, 1738 cm-1. Anal.
Calcd for C8H14O3: C, 60.74; H, 8.92. Found: C, 60.79; H, 8.79.
Eth yl 3,3-Dim eth yl-2-oxobu ta n oa te (12) (Meth od B).
To a solution of 498 g of diethyl oxalate in 2 L of toluene at
J . Org. Chem, Vol. 69, No. 11, 2004 3625

Nelson et al.
-78 °C was added tert-butylmagnesium chloride (1 M in THF)
(3.85 L) over 1 h while the internal temperature was main-
tained below -60 °C. After 1 h, the reaction mixture was
quenched with 3 N HCL (1 L) and water (1 L) and allowed to
warm to room temperature. The organic phase was separated
and washed with water. The solvent was removed in vacuo to
yield 598.6 g of crude ketoester, which was purified by vacuum
distillation (90 °C, 40 mm Hg) to produce 443.3 g (74% yield)
(300 MHz, CDCl3) δ 4.51 (s, 1H), 4.22 (dd, J ) 9.6, 6.4 Hz, 1
H), 3.80-3.68 (m, 1H), 3.59-3.52 (m, 1H), 2.54-2.42 (m, 1H),
2.11-1.86 (m, 3H), 1.12 (s, 9H); 13C NMR (75.5 MHz, CDCl3)
δ 167.8, 163.1, 89.6, 57.2, 45.8, 37.4, 30.5, 26.2 (3C), 21.9; IR
(Nujol mull) 2922, 2853, 1743, 1673 cm-1. Anal. Calcd for
C
11H17NO3: C, 62.54; H, 8.11; N, 6.63. Found: C, 62.57; H,
8.27; N, 6.51.
La cton e (S,S)-16. An authentic sample of this lactone was
1
of ketoester (90 LCWP, 73 LCAP, 95 wt % by H NMR).
prepared in an identical fashion to the preparation of the
diastereomeric lactone [(S,R)-16]: mp 162.5 °C; [R]25 -444
(R)-2-Hyd r oxy-3,3-d im eth ylbu ta n oic Acid [(R)-14]. The
following components were rapidly stirred at room tempera-
ture: 4.6 L of 400 mM phosphate buffer stock solution (see
the general Experimental Section (within the Supporting
Information) for the preparation of all stock solutions), 12 L
of the glucose stock solution, 320 mL of the GDH stock
solution, 73.6 mL of the NADP stock solution, 12.8 mL of the
KRED1001 stock solution, and 295.2 g of ketoester 12. The
pH was maintained at 7.0 by the addition of 5% NaOH during
the course of the reaction. [The enantiomerically pure ethyl
ester (R)-13 could be isolated at this stage, although the
saponification reaction was most frequently performed on the
crude reaction mixture: 1H NMR (300 MHz, CDCl3) δ 4.33-
4.21 (m, 2H), 3.80 (d, J ) 7.6 Hz, 1H), 2.83 (dd, J ) 7.6, 1.4
Hz, 1H), 1.32 (t, J ) 7.2 Hz, 3H), 0.981 (s, 9H); 13C NMR (75.5
MHz, CDCl3) δ 174.3, 78.4, 61.2, 35.2, 25.8, 14.2; IR (neat)
365
1
(c 1, MeOH); H NMR (300 MHz, CDCl3) δ 4.35 (s, 1H), 4.26
(t, J ) 7.7 Hz, 1 H), 3.64-3.53 (m, 2H), 2.41-2.32 (m, 2H),
2.01-1.96 (m, 2H), 1.28 (s, 9H); 13C NMR (75.5 MHz, CDCl3)
δ 169.8, 164.2, 84.6, 57.7, 45.1, 34.0, 28.3, 26.0 (3C), 23.3; IR
(Nujol mull) 2923, 2853, 1739, 1691 cm-1. Anal. Calcd for
C
11H17NO3: C, 62.54; H, 8.11; N, 6.63. Found: C, 62.45; H,
8.25; N, 6.75.
Ketoa m id e (S)-17. To a solution of D-proline methyl ester
hydrochloride (1.42 g, 8.5 mmol) in 26 mL of methylene
chloride at 0 °C was slowly added diisopropylethylamine (1.50
mL) over
a 5 min while the internal temperature was
maintained at <5 °C. To the reaction mixture were added
hydroxybenzotriazole hydrate (210 mg, 1.5 mmol), 3,3-di-
methyl-2-oxobutanoic acid (1.00 g, 7.7 mmol), and EDC hy-
drochloride (1.63 g, 8.5 mmol), and this was stirred overnight
at room temperature and then quenched with 3 N HCl. The
layers were separated, and the organic portion was washed
with 3 N HCl and saturated sodium bicarbonate, dried over
sodium sulfate, concentrated in vacuo, and purified by silica
gel chromatography to afford 1.56 g (84%) of ketoamide (S)-
17 as a white solid: mp 53.0 °C; [R]25365 -382 (c 1, MeOH); 1H
NMR (300 MHz, CDCl3) 3:1 rotameric mixture; major rotamer
δ 4.58-4.50 (m, 1H), 3.77 (s, 3H), 3.54-2.45 (m, 2H), 2.17-
1.90 (m, 4H), 1.30 (s, 9H); minor rotomer 4.68 (dd, J ) 8.4,
4.0 Hz, 1H), 3.74 (s, 3H), 3.70-3.63 (m, 2H), 2.30-2.17 (m,
4H), 1.30 (s, 9H); 13C NMR (75.5 MHz, C6D6) mixture of
rotamers δ 207.5, 207.2, 173.2, 172.2, 165.8, 163.9, 59.8, 58.6,
52.3, 52.1, 47.4, 47.1, 43.7, 43.6, 31.6, 29.1, 27.4, 26.7, 26.5,
25.0, 22.5; IR 2923, 2852, 1747, 1706, 1630 cm-1. Anal. Calcd
for C12H19NO4: C, 59.73; H, 7.94; N, 5.81. Found: C, 59.76;
H, 8.07; N, 5.80.
La cton e (S,R)-16 (Meth od B). Ketoamide (0.5 g, 2.07
mmol), 5% Ru/C (0.25 g), and methanol (50 mL) were combined
and degassed by three alternating nitrogen/vacuum cycles. The
mixture was heated to 50 °C and pressurized to 60 psig of
hydrogen pressure. After 72 h, no ketoamide remained (GC).
The mixture was cooled to room temperature, depressurized
and filtered through a Celite packed sintered glass funnel. The
methanol was removed in vacuo to yield the crude hydroxy
ester. The hydroxy ester was dissolved in 15 mL of toluene,
and PTSA (60 mg, 0.15 equivalents) was added. The mixture
was stirred at room temperature and vacuum applied (40
mmHg) to remove the methanol, which was formed as a
byproduct. After 3 h, HPLC analysis indicated that the
cyclization was complete. The ratio of (S,R)-lactone 16 to (S,S)-
lactone 16 was 10:1. The toluene was removed in vacuo, and
the (S,R)-lactone 16 was purified by silica gel chromatography
(4:1 dichloromethane/EtOAc) to yield 293 mg of lactone (S,R)-
16 [1.39 mmol, 67% yield from ketoamide (S)-9].
3515, 2960, 1727 cm-1; [R]25
-84 (c 1, MeOH); >95% ee]
365
After complete reduction (8 h, HPLC analysis), 720 mL of
NaOH (50% v/v) was added and stirred for 75 min to effect
complete saponification. The final hydrolyzed solu-
tion was neutralized to pH ) 2 with concentrated H2SO4 and
then extracted with EtOAc. The solvent was removed in
vacuo and the residue crystallized from heptane to afford 212.7
g of (R)-2-hydroxy-3,3-dimethylbutanoic acid (R)-14 as a white
solid (86% yield, >99.5% ee): mp 49.8 °C (lit.17 mp 51 °C);
[R]25 -46 (c 1, MeOH);38 1H NMR (300 MHz, CDCl3) δ 3.93
365
(s, 1H), 1.04 (s, 9H); 13C NMR (75.5 MHz, CDCl3) δ 178.7, 78.3,
35.2, 25.8 (3C); IR (Nujol mull) 3356, 2918, 2853, 1733 cm-1
.
Anal. Calcd for C6H12O3: C, 54.53; H, 9.15. Found: C, 54.16;
H, 9.09.
La cton e (S,R)-16 (Meth od A). A solution of L-proline
methyl ester hydrochloride (272.1 g, 1.64 mol) in 3 L of
acetonitrile was cooled to 0 °C, and diisopropylethylamine
(215.6 g, 1.67 equiv) was added. After 15 min, HOBT (61.5 g,
0.45 mol), hydroxy acid (R)-14 (200.1 g, 1.51 mol), and EDC
(350.9 g, 1.83 mole) were added sequentially, and the resulting
mixture was stirred at 0 °C for 5 h (90% HPLC assay yield).
The mixture was quenched with 1 L of 3 N HCl and diluted
with dichloromethane (3 L). The organic portion was separated
and washed with 3 N HCl, saturated NaHCO3, and then 10
wt % aqueous NaCl. The solvent was removed in vacuo to yield
a crude mixture of ester (S,R)-15 and lactone (S,R)-16 (376.5
g). This solid was dissolved in toluene (2 L) and placed in a 5
L three-neck round-bottomed flask that was equipped with a
short-path distillation head. To this was added PTSA (56.0 g,
0.30 mol) and the mixture heated to 45 °C at 100 mmHg. After
6 h at 45 °C, the conversion to lactone (S,R)-16 was complete
(100% by HPLC). During the distillation, approximately 500
mL of toluene was removed. The mixture was cooled to
ambient temperature, washed twice with saturated NaHCO3
and saturated NaCl, and dried over Na2SO4, and the solvent
was removed in vacuo. The residue was recrystallized from
heptane to afford 213.2 g of lactone (S,R)-16 as a white solid
Am id e 18. A slurry of the lactone (S,R)-16 (19.14 g, 90.60
mmol) and amine 9 in triethylamine (100 mL) was stirred
overnight. The initial slurry became homogeneous overnight.
The reaction mixture was diluted with isopropyl acetate and
cooled to 7 °C, and 1 M citric acid (250 mL) was added while
the internal temperature was maintained at less than 16 °C.
The mixture was separated, and the organic portion was
washed with 1 M citric acid (100 mL), water (100 mL), 1 M
Na2CO3 (150 mL), and water (20 mL). The solvent was
removed in vacuo to provide Boc 18 (45.42 g that contained
6.6 wt % (1H NMR) isopropyl acetate (3.02 g), corrected weight
) 42.4 g, 97% yield) as a glass. This material was used in the
subsequent step without further purification.
[67% isolated yield, 99.5 LCAP, 100 LCWP, 0.17 LCAP lactone
1
(S,S)-16]: mp 111.6 °C; [R]25 -214 (c 1, MeOH); H NMR
365
(37) Corral, C.; Madron˜ero, S.; Vega, S. Anal. Quim. 1972, 851.
(38) The literature value17 for the specific rotation of enantiomeri-
cally pure (R)-2-hydroxy-3,3-dimethylbutanoic acid is: [R]20 +65 [c 1,
aqueous ammonium molybdate (50 mg/10 mL)] and [R]20 D-4.3 (c 4,
D
H2O). Correspondingly, the specific rotation for (S)-2-hydroxy-3,3-
dimethylbutanoic acid is [R]20D -63 [c 1, aqueous ammonium molybdate
(48.5 mg/10 mL)]. The rotation of this (S)-antipode (purchased from
Aldrich Chemical Co., Milwaukee, WI), under our conditions, was
[R]20 +46.3 (c 1, MeOH).
365
3626 J . Org. Chem., Vol. 69, No. 11, 2004

Stereoselective Synthesis of an Antithrombotic Agent
HBr Sa lt of 1. A stock solution of 6.0 wt % HBr was
generated by sparging HBr gas through isopropyl acetate. In
a separate vessel, an ambient temperature solution of Boc
amine 18 (9.62 g, 20.0 mmol) in isopropyl acetate (100 mL)
was treated with the stock solution of HBr in isopropyl acetate
(80.6 g solution, 4.85 g HBr, 60.0 mmol) via addition fun-
nel over 30 min. The resultant slurry was aged for 1 h at
ambient temperature. The slurry was filtered under N2,
sequentially washed with isopropyl acetate, methyl acetate and
dried under vacuum to afford 8.12 g of an amorphous white
solid (85% isolated yield corrected for purity of 97 LCAP): 1H
NMR (300 MHz, CD3OD) δ 7.55-7.30 (m, 3H), 4.62 (d, AB,
1H, J ) 5.1 Hz), 4.33-4.12 (m, 4H), 4.09 (s, 1H), 3.86-3.61
(m, 2H), 2.28-1.82 (m, 4H) 0.93 (s, 9H); HRMS m/z calcd for
C19H29BrClN3O3 (M+) 382.1892, found 382.1891. Compound 1
that was prepared by this method was identical to an authentic
sample of 1.5
Ack n ow led gm en t. We gratefully acknowledge the
analytical support provided by Bing Mao and Mike
Miller (Merck).
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectral data for compounds 9, 12, (R)-13, (R)-14, (S,R)-16,
(S,S)-16, (R)-17. This material is available free of charge via
the Internet at http://pubs.acs.org.
J O035794P
J . Org. Chem, Vol. 69, No. 11, 2004 3627