Technical scale synthesis of a new and highly potent thrombin inhibitor

Article abstract of DOI:10.1055/s-2004-831200

In this account, we describe the development of an efficient and convergent process for the peptidomimetic thrombin inhibitor 1 on production plant scale. Starting from nicotinonitrile (13), (2S,4R)-1-(tert-butoxycarbonyl)-4-hydroxy-2- pyrrolidinecarboxylic acid (5) and (2R)-2-amino-3-cyclohexylpropanoic acid (29) compound 1 was obtained in 16 chemical steps. New methods had been developed for the preparation of the key intermediate dehydroproline 22 and the transformation of nitriles into amidines. The thrombin inhibitor 1 was isolated by special techniques (nanofiltration and spray drying). Almost all salts of 1 are amorphous, however, a crystalline complex was obtained with 1,2-benzisothiazol-3(2H)-one 1,1-dioxide (Saccharin).

Full text of DOI:10.1055/s-2004-831200

PAPER
2367
Technical Scale Synthesis of a New and Highly Potent Thrombin Inhibitor
T
echnical Scale Sy
a
nthesis of
r
a
N
ewan
a
dHighlyP
l
otent
T
d
hrombinInhibitorBernard,a Gerd Bülow,a Udo E. W. Lange,b Helmut Mack,b Thomas Pfeiffer,a Bernd Schäfer,*a Werner Seitz,a
Thomas Zierkea
a
BASF AG, GVF/A-A30, 67056 Ludwigshafen, Germany
Fax +49(621)6042290; E-mail: bernd.schaefer@basf-ag.de
b
Abbott GmbH & Co. KG, 67061 Ludwigshafen, Germany
Received 2 April 2004; revised 3 June 2004
Abstract: In this account, we describe the development of an effi-
cient and convergent process for the peptidomimetic thrombin in-
hibitor 1 on production plant scale. Starting from nicotinonitrile
(13), (2S,4R)-1-(tert-butoxycarbonyl)-4-hydroxy-2-pyrrolidinecar-
boxylic acid (5) and (2R)-2-amino-3-cyclohexylpropanoic acid (29)
compound 1 was obtained in 16 chemical steps. New methods had
been developed for the preparation of the key intermediate dehydro-
proline 22 and the transformation of nitriles into amidines. The
thrombin inhibitor 1 was isolated by special techniques (nanofiltra-
tion and spray drying). Almost all salts of 1 are amorphous, howev-
er, a crystalline complex was obtained with 1,2-benzisothiazol-
3(2H)-one 1,1-dioxide (Saccharin®).
Key words: amino acids, drugs, enantiomeric resolution, heterocy-
cles, stereoselective synthesis
Figure 1 Low molecular weight thrombin inhibitors share the motif
of D-Phe-Pro-Arg
Thromboembolic diseases like myocard infarction, unsta-
ble angina, ischaemic stroke and venous thrombosis are a
major cause of morbidity and mortality in the industrial-
ized world. Common drugs for treatment and prophylaxis
of these diseases are vitamin K antagonists (e.g. Marcu-
mar®) and low molecular weight heparin. Due to safety is-
sues in the first case and low oral bio-availability in the
latter, there is still an urgent medical need for low molec-
ular weight, orally active thrombin inhibitors. Therefore
many research groups in academia and industry have fo-
cused their attention on the elaboration of these types of
synthetic thrombin inhibitors.1
been one of the most challenging parts of the manufactur-
ing process of the new thrombin inhibitor 1.
In this communication, we describe our efforts to develop
an efficient, convergent total synthesis of 1 suitable for
technical scale production. The final route meets all re-
quirements of production scale, e.g. no exotic starting ma-
terials, high yields and easy purification by crystallization
or salt precipitation.
First Laboratory Scale Synthesis
Thrombin is a multifunctional serine protease that plays a
pivotal role in the critical balance of hemostasis and
thrombosis.2 Derivatives of D-Phe-Pro-Arg can mimic the
binding region of thrombin’s natural substrate fibrinogen.
Inspired by this observation many, low molecular weight
thrombin inhibitors are derived from this motif
(Figure 1).3 Our efforts on structural optimization of this
small peptide, resulted in the highly potent, orally active
and selective thrombin inhibitor 1 with unique structural
characteristics.4
Initially we started with a linear synthesis, which had been
established from our structure optimization program
(Scheme 1).5 2,5-Pyridinedicarboxylic acid was convert-
ed regioselectively via the mono ester in 5 steps into ni-
trile 3. The P1 moiety 4·2 HCl was obtained by
hydrogenation. The coupling partner 7 was synthesized
from (2S,4R)-1-(tert-butoxycarbonyl)-4-hydroxy-2-pyr-
rolidinecarboxylic acid (5) by activation of the carboxylic
acid with 2,2-dimethylpropanoyl chloride and mesityla-
tion of the hydroxy group. After coupling, the P1–P2 pre-
cursor 8 was reacted with 1,2-diphenyldiselane in the
presence of sodium borohydride. Oxidation of the sele-
nium with concurrent syn-elimination gave the desired
double bond.6 Coupling with (2R)-2-[(tert-butoxycarbon-
yl)amino]-3-cyclohexylpropanoic acid, dehydration of
the carboxylic amide, removal of the Boc-group and intro-
duction of the tert-butyl acetate unit led to the carbon scaf-
fold of compound 1. In the final four steps the amidine
functionality was introduced by the hydrogen sulfide
The most significant modifications are found in the P1
and P2 moiety. Arginine is replaced by an amidinopyri-
dine moiety and proline by 3,4-dehydroproline. Both
building blocks are either not commercially available or
very expensive. The synthesis of both intermediates has
SYNTHESIS 2004, No. 14, pp 2367–2375
x
x.
x
x
.
2
0
0
4
Advanced online publication: 25.08.2004
DOI: 10.1055/s-2004-831200; Art ID: T03704SS
© Georg Thieme Verlag Stuttgart · New York

2368
H. Bernard et al.
PAPER
Scheme 1 Linear synthesis of thrombin inhibitor 1
Reagents and conditions: a) EtOH, reflux; b) 1,2-dichloroethane, SOCl2, DMF (cat.), reflux; c) NH3, THF, –10 °C; d) (CF3CO)2O, i-Pr2NEt,
CH2Cl2, 20 °C; e) NH3, EtOH, –10 °C; f) H2, Pd/C, 20 °C, HCl, H2O–n-PrOH; g) t-BuCOCl, CH2Cl2, 0 °C; h) MesCl, i-Pr2NEt, 0–5 °C; i) 5–
20 °C; j) (PhSe)2, NaBH4, EtOH, H2O, 30–40 °C, then 8, reflux; k) NaIO4, AcOH, MeOH, 5 °C; l) HCl, EtOH, 0–20 °C; m) (2R)-2-[(tert-but-
oxycarbonyl)amino]-3-cyclohexylpropanoic acid, i-Pr2NEt, CH2Cl2, propane phosphonic acid anhydride (50% in EtOAc), 0–5 °C; n)
(CF3CO)2O, i-Pr2NEt, CH2Cl2, 0–5 °C; o) HCl, Et2O, CH2Cl2; p) BrCH2CO2t-Bu, i-Pr2NEt, CH2Cl2, 20 °C; q) Boc2O, i-Pr2NEt, CH2Cl2, 20 °C,
chromatography; r) H2S, pyridine, Et3N, 20 °C; s) MeI, acetone, 20 °C; t) NH4OAc, MeOH, 20 °C, chromatography; u) HCl, Et2O, 20 °C, chro-
matography.
method7 and all protective groups were subsequently re- cused our efforts on the formation of the P1–P2 moiety
moved by acidic cleavage. The thrombin inhibitor 1·HCl and the final reaction steps.
was obtained in 21 reaction steps and an overall yield of
about 3% based on 2,5-pyridinedicarboxylic acid.
The convenience and reliability of the peptide coupling
with propane phosphonic acid anhydride8 (PPAA) (no ra-
cemization or epimerization) as well as our successful
protective group strategy prompted us to maintain this
methodology.
Conception of a New Synthesis
The most important strategic decision made at the outset
of this work was to build up the entire molecule by a con-
vergent route, which comprises P1–P2, prepared by sepa-
P1 Building Block
rate syntheses of readily functionalized P1 and P2, and
We chose nicotinonitrile (13) as an appropriate starting
P3–P4 as key intermediates. After peptide coupling of
material (Scheme 2, sequence A). It was converted in a se-
these building blocks, the final P1–P4 scaffold should be
quence of four reaction steps into P1 building block 16.
Nicotinonitrile (13) was hydrogenated and the resulting
converted into the drug in very few reaction steps.
Major issues were to exclude some critical chemicals amino function protected by known methods.9 Interest-
(1,2-diphenyldiselane, hydrogen sulfide and methyl io- ingly, in contrast to (1-oxido-3-pyridinyl)methanol, the
dide), to reduce the number of reaction steps, prefer one- N-protected (1-oxido-3-pyridinyl)methylamine10 15
pot reactions, and to simplify workup (no chromatogra- could be converted into 16 with high regioselectivity and
phy, purification by crystallization). Therefore we fo- in good yields. However, the drawback for large scale
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York

PAPER
Technical Scale Synthesis of a New and Highly Potent Thrombin Inhibitor
2369
synthesis was some hazard considerations with regard to role-2-carboxylic acid (22) (Scheme 3), which did not in-
the use of large quantities of 3-chloroperoxybenzoic acid volve chemicals from the chiral pool.
and sodium cyanide.
To circumvent these problems for production plant syn-
thesis we took advantage of Minisci’s method11 to intro-
duce a carboxylic amide in heterocycles (Scheme 2,
sequence B). In general, yields were not high but the reac-
tion sequence was simplified dramatically. The regioiso-
mer 3-cyanoisonicotinamide, formed as a minor by-
product under these reaction conditions, could easily be
removed by crystallization. The desired target molecule
16 was finally obtained by catalytic hydrogenation and
dehydration with trifluoroacetic anhydride.12
Scheme 3 Routes for the Synthesis of P2
Reagents and conditions: a) CCl3COCl, cyclohexane, 15 °C; b) NaO-
Me, MeOH; c) Boc2O, cyclohexane; d) Li, NH3, –78 °C; e) Novo-
zym®, NaHCO3, H2O; f) Cl2Ru(PCy3)2CHMe, CH2Cl2; g) LDA,
(MeO)2CO, THF, 0 °C; h) t-BuCOCl, Et3N; i) MesCl, Et3N; j)
NaOC2H4OC2H4NMe2, toluene–MeCN, –5 to 20 °C, HNEt2.
Scheme 2 Routes for the Synthesis of P1
Reagents and Conditions: a) Pd/C, AcOH; b) MCPBA, CH2Cl2; c)
NaCN, Me3SiCl, Et3N, DMF; d) HCl, i-PrOH; e) HCONH2, H2O2,
H2SO4/FeSO4; f) H2, Pd/C, n-PrOH, H2O, HCl; g) Boc2O, NaOH, Me-
OH, H2O; h) (CF3CO)2O, Et3N, CH2Cl2; i) HCl, i-PrOH.
In our first attempt we applied the Birch reduction of pyr-
roles, discovered by Donohoe (Scheme 3, sequence A).13
The required N-Boc pyrrole carboxylic ester 21 was readi-
ly available in a three step sequence from pyrrole.14–16 It
may be noted, that for safety reasons as well as to assure
high yield and product quality the sequence of addition
was crucial. Advantageously, pyrrole (19) was added to
trichloroacetic chloride, and a solution of the ketone 20 to
sodium methanolate. Isolation of the trichloromethyl ke-
tone 20 should be avoided for reasons of its skin irritant
properties and light sensitivity. The most suitable metals
for the Birch reduction were lithium and sodium. In all
cases the major side reaction was the reduction of the ester
P2 Building Block
Parallel to our work on P1, we considered also several ap-
proaches for enantiopure 1-(tert-butoxycarbonyl)-2,5-
dihydro-1H-pyrrole-2-carboxylic acid (N-Boc dehydro-
proline, 22). The major drawback of our initial synthesis
of the dehydroproline derivative 9 was the use of toxic and
expensive selenium compounds. We therefore explored
several alternative routes to this key-building block that
did avoid such chemicals.
At the beginning we investigated two de novo approaches
towards
1-(tert-butoxycarbonyl)-2,5-dihydro-1H-pyr-
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York

2370
H. Bernard et al.
PAPER
moiety. The best results were obtained with sterically hin-
dered esters at –78 °C.
Table 1 Optimization of the Elimination Reaction
Entry Base/Additive
22:26
95:5
(S:R)-22
90:10
For the final step we favored the enzymatic kinetic reso-
lution of the racemic methyl ester 21. An extensive
screening of enzymes and esters revealed that no ester
could be cleaved by enzymes other than methyl.17 This
limitation, but also the low temperature conditions re-
quired for the Birch reduction and a lack in selectivity in
the case of the methyl ester caused us to look for a superi-
or access to 21.
1
2
3
4
5
6
7
NaH/DME
a
NaH/i-PrOH
-
-
57:43
a
NaH/t-BuOH
57:43
a
NaOCMe2Et/THF
NaH/MeOH
16:84
-
67:33 93:7
72:28 99:1
78:22 99:1
Next, we anticipated a-methoxycarbonylation of tert-bu-
tyl 2,5-dihydro-1H-pyrrole-1-carboxylate (24), readily
available from diallylamine (23) by ring-closing metathe-
sis, would be a smooth approach (Scheme 3, sequence B).
Several carbonylation reagents, like chloroformates, cy-
anoformates or ethyl methoxy(methyl)carbamate were
checked, but afforded only complex mixtures of products.
Finally, the use of dialkyl carbonates gave the desired
product even at 0 °C in excellent yields.17 We judged this
approach to be very feasible for chemical production, but
a third synthesis had already succeeded.
NaOCMe2Et/2-(methoxy)ethanol, THF
NaOCMe2Et/2-[2-(methoxy)ethoxy]-
ethanol, THF
8
9
NaOCMe2Et/2-[2-(dimethylamino)-
ethoxy]ethanol, THF
80:20 99:1
NaOCMe2Et/2-[2-(dimethylamino)-
ethoxy]ethanol, toluene, MeCN
>80:20 99:1
a Not recorded.
An improvement in terms of configurational stability at
the a-carbon was obtained with sodium methanolate as
base, but the regioselectivity of the elimination was poor
(Table 1, entry 5). A breakthrough was finally achieved
when in situ generated sodium 2-methoxyethanolate was
used as base in the elimination affording the desired dehy-
droproline 22 with hardly any racemization and with good
regioselectivity (Table 1, entry 6).21 Self-complexation
was considered to be responsible for low nucleophilicity
and effecting kinetically controlled elimination selective-
ly in 3,4-position without racemization. An in-depth
screening of several other primary alcoholates with addi-
tional coordinating heteroatoms in the side-chain of the
base led to slightly better regioselectivity of the elimina-
tion (Table 1, entries 7 and 8). The best results were ob-
tained with sodium 2-[2-(dimethylamino)ethoxy]-
ethanolate in toluene–acetonitrile. Sodium 2-methyl-2-
butanolate could even be replaced by sodium methano-
late, if methanol was removed by azeotropic distillation
prior to the addition of mesylate 25. We considered sodi-
um 2-[2-(dimethylamino)ethoxy]ethanolate as bulky as
tert-alcoholates, but its basicity was similar to primary al-
coholates. Additionally this base was superior, because
the tertiary amine moiety allowed an easy removal of the
reagent during workup. Acidic workup also caused hydro-
lytic decomposition of the undesired isomer 26. Purifica-
tion of 22 was performed by salt formation with
diethylamine.
Almost at the same time, we had adapted the elimination
reaction under basic conditions from our first total synthe-
sis, starting from chiral pool-based (2S,4R)-1-(tert-but-
oxycarbonyl)-4-hydroxy-2-pyrrolidinecarboxylic acid (5)
(Scheme 3, sequence C). The main challenges of this
route were to avoid the formation of the thermodynami-
cally favored 1-(tert-butoxycarbonyl)-2,3-dihydro-1H-
pyrrole-2-carboxylic acid (26) and the racemization at the
a-carbon of the amino acid. Formation of the requisite
double bond of dehydroproline by literature known
Chugaev reactions18 were not considered an option due to
toxic sulfur or selenium compounds involved.
Hydrolysis of the mixed anhydride 6 after mesylation af-
forded the carboxylic acid 25, which was considered to be
better protected against racemization under basic condi-
tions than the corresponding esters. In the presence of a
base, a second deprotonation of the neighbouring a-car-
bon was diminished by the carboxylate moiety. Treatment
of mesylate 25 with a strong, non-nucleophilic base, like
sodium hydride in DMF, furnished the desired product 22
in modest yield and with moderate selectivity.19 Sodium
hydride in DMF should be handled with caution and is not
an ideal reagent for large scale synthesis.20 We therefore
replaced DMF by ethers, e.g. DME (Table 1, entry 1), but
did not obtain better yields or higher selectivity. Further-
more, on larger scale the reaction was found to be capri-
cious leading to incomplete conversion of the starting
material even when an excess of sodium hydride was
used.
Key Intermediates P1–P2, P3–P4 and P1–P4
Replacement of sodium hydride with secondary or tertiary
alcoholates, like sodium isopropanolate and sodium tert-
butanolate, prepared in situ from NaH and the correspond-
ing alcohol, led to complete racemization (Table 1, entries
2 and 3). Exposure to sodium 2-methyl-2-butanolate in
THF gave predominately the undesired isomer 26
(Table 1, entry 4).
In the next step we synthesized the key intermediates P1–
P2, P3–P4 and P1–P4 (Scheme 4). The P1–P2 building
block arose from a PPAA coupling of 22 and 16 followed
by an acidic N-deprotection and the formation of the hy-
drogen chloride salt. Isopropyl alcohol was the most ap-
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York

PAPER
Technical Scale Synthesis of a New and Highly Potent Thrombin Inhibitor
2371
propriate solvent because of negligible addition to the addition of (2R)-2-(acetylamino)-3-sulfanylpropanoic
cyano group of the P1 moiety. The process was robust and acid to the nitrile 11 was presumably the rate determining
delivered the desired product 28·HCl in 80% yield.
step. It may be noted that acetylcysteine was not only act-
ing as an organocatalyst30 but also stabilized the generated
amidine.31 Therefore we used one equivalent (2R)-2-
(acetylamino)-3-sulfanylpropanoic acid and obtained the
desired product as a salt. When the reaction was run cata-
lytically or on a small scale ammonium acetate was used
preferentially as an alternative to gaseous ammonia. The
catalyst could be removed from the crude product by
means of ion exchange chromatography32 or crystalliza-
tion. The resulting product was free of by-products with
an unpleasant odor.
Our approach to the P3–P4 moiety started with (2R)-2-
amino-3-cyclohexylpropanoic acid. Esterification with
benzylic alcohol was followed by N-alkylation and N-
protection. The carboxylic acid was liberated by hydroge-
nation. Purification of 31 was achieved by salt formation
with cyclohexylamine. The overall yield was 76%. Pro-
pane phosphonic acid anhydride was also used for the for-
mation of the P1–P4 precursor 11, which was an oil and
therefore handled as a solution in dichloromethane.
An alternative process, preferentially used in production-
scale synthesis, combined workup of the amidine 12 and
cleavage of the protective groups. This was accomplished
Amidine Syntheses and Workup of 1
Numerous synthetic methods for the preparation of in aqueous solution with hydrogen chloride at 65 °C fol-
amidines described in the literature involve highly acid- lowed by extraction of (2R)-2-(acetylamino)-3-sulfanyl-
ic,22 alkaline23,24 or strong reducing25 conditions or require propanoic acid and some impurities. The liberation was
high temperature26 and are less suitable to replace the hy- complete within 3 hours. The thrombin inhibitor 1 precip-
drogen sulfide/methyl iodide method.7
itated in DMF after neutralization with aqueous ammonia
at pH 8.75 as a betaine, forming a 1:1-adduct with the sol-
vent, which also triggered an efficient purification of the
final product. DMF was a very unique solvent for this pur-
pose, which could not be replaced by any other solvent.
Subsequently, we established special techniques to end up
with a DMF-free mono hydrochloride of the thrombin in-
hibitor 1. Therefore, the betaine was dissolved in water in
the presence of 1 equivalent hydrogen chloride and DMF
was removed by nanofiltration.33 Finally, the HCl-salt of
1 was obtained by spray drying34 or preferentially on
small scale by lyophilization (Scheme 4).
Nevertheless, inspired by a literature method23 we
checked in our first attempt the possibility to expose the
pyridine nitrile 11 to methanol and sodium methoxide at
20 °C followed by the addition of ammonium chloride
(Scheme 4). In a kinetic study, the addition of methanol
revealed to be the rate determining step, but was also ac-
companied by slow epimerization at the dehydroproline
residue. Both the reaction and epimerization rate were
correlated to the amount of sodium methoxide. Exposure
to one equivalent base after 100 minutes afforded 96%
conversion, but also 40% epimerization of the stereogenic
center at the P2 moiety. However, at 0 °C complete con- In summary, we have developed a highly convergent and
version was obtained after 6.5 hours without any epimer- efficient process for producing the peptidomimetic throm-
ization. The best results (complete conversion after 5 h bin inhibitor 1 suitable for production plant scale. Starting
without any racemization) were finally achieved with two from nicotinonitrile (13), (2S,4R)-1-(tert-butoxycarbon-
equivalents of sodium methoxide at 10–12 °C.
yl)-4-hydroxy-2-pyrrolidinecarboxylic acid (5) and (2R)-
2-amino-3-cyclohexylpropanoic acid (29) the thrombin
inhibitor 1·HCl was obtained in 16 chemical steps. Wher-
ever possible, we simplified workup and developed one-
pot reactions. All starting materials used in the process
were bulk chemicals. The optimized synthesis was devoid
of any chromatographic purification. The key intermedi-
ates were purified by salt formation and crystallization.
The active pharmaceutical ingredient (API, 1·HCl) was
isolated by special techniques (nanofiltration and spray
drying) and could be crystallized in the presence of 1,2-
benzisothiazol-3(2H)-one 1,1-dioxide (Saccharin®).35
The process depicted in Scheme 4 was already used to
produce nearly 100 kg of thrombin inhibitor 1.
For a more robust technical process we developed a new
method for the synthesis of amidines which did not have
the above mentioned drawbacks. (2R)-2-(Acetylamino)-
3-sulfanylpropanoic acid (N-acetylcysteine) or (2S)-1-
[(2S)-2-methyl-3-sulfanylpropanoyl]-2-pyrrolidinecarbo-
xylic acid (Captopril®) were used in a mild, organocata-
lytic process for the synthesis of highly functionalized
amidines.27 The catalytical properties of these thiols
proved to be very similar, but for large scale production
we chose acetylcysteine for economical reasons. Mecha-
nistically related to the thio-Pinner reaction,7 we proposed
the formation of an iminothioether, which was converted
to an amidine by nucleophilic attack of ammonia. In con-
trast, the use of cysteine, according to a similar method
described earlier,28 led as expected to the formation of a
mixture of the desired product 12 and the corresponding
thiazoline carboxylic acid derivative.29 The acetylcys-
teine-catalyzed synthesis exhibited a marked dependence
on the reaction temperature used. At elevated temperature
(60–65 °C) or at high concentration, the reaction rate in-
creased despite of the reduced solubility of ammonia. The
All starting materials are bulk chemicals and were used without fur-
ther purification. All reactions were run under N2. Due to its appli-
cation, thrombin inhibitor 1·HCl was produced fully in accordance
to GMP regulations.36 Melting points are uncorrected. IR spectra
were recorded on Bruker IFS-66. 1H and 13C NMR spectra were re-
corded on Bruker AC-200, AC-250, AC-400 and AMX-400 spec-
trometer and TMS was used as internal standard at 20 °C.
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York

2372
H. Bernard et al.
PAPER
150 kg) at 90 °C. The product crystallized on cooling to 5 °C. It was
filtered, washed with dil. aq HCl (5%, 100 kg) and H2O (100 kg)
and dried at 60 °C/30 mbar (16 kg, 27%); mp 278 °C (dec.).
IR (KBr): 3410, 3164, 2233, 1711, 1605, 1599, 1423, 1372, 872,
550, 527 cm–1.
1H NMR (250 MHz, DMSO-d6): d = 7.94 (s, 1 H, NH2), 8.18 (d,
J = 7.6 Hz, 1 H, CH), 8.33 (s, 1 H, NH2), 8.52 (dd, J = 7.6, 2.3 Hz,
1 H, CH), 9.12 (d, J = 2.3 Hz, 1 H, CH).
13C NMR (100.61 MHz, DMSO-d6): d = 111.19 (s, CCN), 116.59
(s, CN), 121.84 (d, CH), 141.74 (d, CH), 151.49 (d, CH), 152.76 (s,
CCONH2), 164.63 (s, CONH2).
HRMS: m/z calcd for C7H5N3O [M+]: 147.0433; found: 147.0426.
5-(Aminomethyl)-2-pyridinecarboxamide·2HCl
5-Cyano-2-pyridinecarboxamide (17; 40 kg, 272 mol) was dis-
solved in aq HCl (20%, 109.2 kg). After addition of H2O (400 L)
and propanol (400 L) it was hydrogenated in the presence of Pd/C
(5%, 20 kg) at 2 bar. After 6 h, the catalyst was filtered off. H2O was
removed by azeotropic distillation. The precipitated product was fil-
tered at 0 °C, washed with propanol (50 L) and dried by purging
with N2 (56.7 kg, 93%); mp 272 °C (dec.).
IR (KBr): 3107, 3065, 3038, 2981, 2902, 2855, 1706, 1597, 1428,
1409, 1217, 1138, 1115, 880 cm–1.
1H NMR (250 MHz, D2O): d = 4.67 (s, 2 H, CH2), 8.53 (d, J = 8.4
Hz, 1 H, CH), 8.82 (d, J = 8.4 Hz, 1 H, CH), 9.12 (s, 1 H, CH).
13C NMR (100.61 MHz, D2O): d = 40.00 (t, CH2), 125.33 (d, CH),
134.99 (s, CCH2), 144.13 (s, CCONH2), 145.03 (d, CH), 146.32 (d,
CH), 163.09 (s, CONH2).
HRMS: m/z calcd for C7H10N3O [(M + H)+]: 152.0824; found:
152.0817.
tert-Butyl[6-(aminocarbonyl)-3-pyridinyl]methylcarbamate
(18)
5-(Aminomethyl)-2-pyridinecarboxamide·2HCl (85.6 kg, 382 mol)
was dissolved in H2O (460 L) and MeOH (330 L), and neutralized
by the addition of an aq solution of NaOH (50%, 67.2 kg, 840 mol)
at 30 °C. Di(tert-butyl) dicarbonate (100 kg, 458 mol) was added at
50–55 °C. The mixture was kept overnight and the product formed
was filtered, washed with H2O (2 × 50 L) and MeOH (2 × 50 L) and
dried in a vacuum drying chamber. Product 18 (86 kg, 90%) was ob-
tained as a colorless solid.
1H NMR (200 MHz, DMSO-d6): d = 1.40 (s, 9 H, t-C4H9), 4.25 (d,
J = 6.0 Hz, 2 H, CH2), 7.50 (m, 1 H, NH), 7.60 (s, 1 H, NH2), 7.81
(d, J = 7.9 Hz, 1 H, CH), 8.00 (d, J = 7.9 Hz, 1 H, CH), 8.10 (s, 1 H,
NH2), 8.50 (s, 1 H, CH).
Scheme 4 Optimized total synthesis
Reagents and conditions: a) BnOH, p-TsOH, toluene; b) tert-butyl
bromoacetate, i-Pr2NEt, DMF; c) (Boc)2O, i-Pr2NEt, DMF; d) H2, Pd/
C; e) c-HexNH2, MTBE; f) aq HCl, CH2Cl2, then PPAA, i-Pr2NEt,
CH2Cl2; g) HCl, i-PrOH; h) PPAA, i-Pr2NEt, CH2Cl2; i) NH3, cata-
lyst, MeOH, 62 °C, nanofiltration; j) HCl, H2O, spray drying.
5-(Aminomethyl)-2-pyridinecarbonitrile·HCl (16·HCl)
Compound 18 (78 kg, 310 mol) and Et3N (81.5 kg, 806 mol) were
dissolved in CH2Cl2 (450 L) and trifluoroacetic anhydride (84.6 kg,
403 mol) was added at 20–25 °C. After 1 h, H2O (120 L) was added
and the pH was adjusted to 3.3–3.6 with aq HCl (1 N, about 50–60
L). The phases were separated and the organic layer was washed
with H2O (2 × 120 L). The solvent CH2Cl2 was removed by distilla-
tion and replaced by propanol (800 L). The product was precipitated
by addition of aq HCl (30%, 3 equiv). The suspension was stirred
for 8 h at 50 °C and then for 12 h at 20 °C. The product 16·HCl
(47.3 kg, 90%) was obtained by filtration and drying; mp 262 °C.
HRMS were recorded on a spectrometer from Micromass using the
LC-TOF or GC-TOF mode.
Scheme 2, Sequence B
5-Cyano-2-pyridinecarboxamide (17)
H2O (300 kg), FeSO4·7H2O (5.3 kg) and nicotinonitrile (13; 42 kg,
404 mol) were placed in a reaction vessel and formamide (109 kg,
2422 mol), H2SO4 (60.5 kg) and H2O2 (30%, 62.3 kg) were added
subsequently at 35–50 °C. After 1 h at 50 °C, the reaction mixture
was added to brine [NaCl (120 kg) dissolved in H2O (330 kg) and
ice (270 kg)]. The solid was filtered and washed with H2O (110 kg).
The wet filter cake was dissolved in H2O (900 kg) and aq HCl (36%,
IR (KBr): 3053, 3007, 2239, 1565, 1474, 1432, 1412, 1092, 1000,
895, 550 cm–1.
1H NMR (400 MHz, DMSO-d6): d = 4.28 (s, 2 H, CH2), 8.17 (d,
J = 8.5 Hz, 1 H, CH), 8.48 (d, J = 8.5 Hz, 1 H, CH), 8.96 (s, 1 H,
CH), 9.04 (s, 3 H, NH3).
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York

PAPER
Technical Scale Synthesis of a New and Highly Potent Thrombin Inhibitor
2373
13C NMR (100.61 MHz, DMSO-d6): d = 39.33 (t, CH2), 117.25 (s,
CN), 128.51 (d, CH), 132.05 (s, CCN), 134.27 (s, CCH2), 138.45 (d,
CH), 151.75 (d, CH).
125.25, 125.57 (d, CHCH=), 128.39, 128.67 (d, CH2CH=), 152.95,
153.52 (s, NCO), 172.72, 172.90 (s, COO).
HRMS: m/z calcd for C10H14NO4 [(M – H)–]: 212.0923; found:
212.0910.
HRMS: m/z calcd for C7H8N3 [(M + H)+]: 134.0718; found:
134.0751.
Scheme 4
Scheme 3, Sequence C
(2S,4R)-1-(tert-Butoxycarbonyl)-4-[(methylsulfonyl)oxy]-2-
pyrrolidinecarboxylic Acid (25)
(2S)-N-[(6-Cyano-3-pyridinyl)methyl]-2,5-dihydro-1H-pyr-
role-2-carboxamide·HCl (28·HCl)
Compound 22·HNEt2 (58.2 kg, 204 mol) was dissolved in H2O
(230 L) and the pH was adjusted to 0.5–1 by addition of aq HCl
(30%, 24.7 kg, 204 mol). (2S)-1-(tert-Butoxycarbonyl)-2,5-dihy-
dro-1H-pyrrole-2-carboxylic acid was extracted with CH2Cl2 (130
L and 2 × 40 L). The combined organic phases were dried by azeo-
tropic distillation. The reaction mixture was subjected consecutive-
ly to 5-(aminomethyl)-2-pyridinecarbonitrile·HCl (16·HCl; 34.5
kg, 204 mol), N-ethyl-N,N-diisopropylamine (143.3 kg, 1099 mol)
and propane phosphonic acid anhydride in EtOAc (50%, 168.3 kg,
264.7 mol) at –10 °C. After 1 h at –5 °C and 12 h at 20–25 °C, the
reaction was quenched by addition of H2O (173 L) at 10 °C. The
phases were separated and the aqueous layer was extracted with
CH2Cl2 (2 × 60 L). The combined organic phases were re-extracted
with dil. aq HCl (100 L) at pH 3.0–3.5 and H2O (60 L). The organic
phase was dried by azeotropic distillation and the solvent replaced
by i-PrOH (270 L). The product was precipitated by addition of HCl
in i-PrOH (23%, 98 kg) at 50 °C, stirred at 50 °C for 5 h and at
20 °C for 12 h. It was filtered at 0 °C, washed with i-PrOH (3 × 60
L) and dried by purging with N2 at 45 °C for 48 h (43 kg, 80%); mp
212 °C.
(2S,4R)-1-(tert-Butoxycarbonyl)-4-hydroxy-2-pyrrolidinecarboxy-
lic acid (5; 80 kg, 346 mol), CH2Cl2 (360 L), Et3N (39.2 kg, 384
mol) and 2,2-dimethylpropanoyl chloride (98%, 45.5 kg, 370 mol)
were mixed consecutively in a stainless steel vessel below 0 °C. Af-
ter 1 h, Et3N (39.2 kg, 384 mol) and methanesulfonyl chloride (45.6
kg, 394 mol) were added at 0 °C. The reaction was quenched by ad-
dition of H2O (480 L). The pH value was adjusted to 0.5 with aq
HCl (30%, 18.5 kg, 152 mol) and the layers were separated. The
product was extracted with aq 50% NaOH [58.2 kg, 726 mol in H2O
(320 L)] and re-extracted with EtOAc (2 × 320 L) after acidification
with aq HCl (30%, 85 kg) at pH 1–1.5. The organic phase was
washed with H2O (160 L) and dried by azeotropic distillation. The
product was crystallized by addition of heptane (480 L) to the re-
maining solution (284 L), filtered, washed with heptane (160 L),
and dried at 50 °C (85.5 kg, 80%); mp 103 °C.
IR (KBr): 1436, 1369, 1357, 1169, 914 cm–1.
1H NMR (200 MHz, CDCl3): d = 1.43 (2 s, 9 H, t-C4H9), 2.40 (m, 1
H, CH2), 2.66 (m, 1 H, CH2), 3.06 (s, 1 H, CH3), 3.82 (m, 2 H, CH2),
4.43 (m, 1 H, CHOMes), 5.23 (s, 1 H, CHCO2H), 11.30 (s, 1 H,
CO2H).
HRMS: m/z calcd for C11H18NO7S [(M – H)–]: 308.0804; found:
308.0839.
IR (KBr): 3179, 3051, 2884, 2490, 2232, 1668, 1556, 1394, 1251,
1027, 839, 656 cm–1.
1H NMR (250 MHz, DMSO-d6): d = 4.00 (d, J = 16.8 Hz, 2 H,
CH2), 4.50 (dd, J = 16.8, 5.3 Hz, 2 H, CH2), 5.11 (s, 1 H, CHCO),
6.05 (m, J = 8.4 Hz, 2 H, 2 CH), 7.95 (dd, J = 8.4, 2.1 Hz, 1 H, CH),
8.05 (d, J = 8.4 Hz, 1 H, CH), 8.75 (d, J = 2.1 Hz, 1 H, CH), 8.91
(2S)-1-(tert-Butoxycarbonyl)-2,5-dihydro-1H-pyrrole-2-car-
boxylic Acid·HNEt2 (22·HNEt2)
+
(br s, 1 H, NH2 ), 9.80 (t, J = 5.3 Hz, 1 H, NH), 10.90 (br s, 1 H,
A solution of NaOMe in MeOH (30%, 72.8 kg, 404 mol), toluene
(150 L), and 2-[2-(dimethylamino)ethoxy]ethanol (59.2 kg, 436
mol) were charged consecutively into a stainless steel vessel.
MeOH was removed by azeotropic distillation at 64 °C. The resi-
due, still containing toluene (about 30 L) was diluted with MeCN
(53 L) and a solution of 25 (50 kg, 162 mol) in MeCN (69 L) was
added. The reaction mixture was stirred for 20 h at 20 °C. After the
addition of H2O (125 L) at 0–15 °C, the phases were separated and
the organic layer was extracted with H2O (25 L). The combined
aqueous phases were washed with tert-butyl methyl ether (125 L)
and the pH was adjusted to 2.5 by addition of aq HCl (31%, 73.5
kg). tert-Butyl methyl ether (100 L) was added. After separation of
the layers, the aqueous phase was re-extracted with tert-butyl meth-
yl ether (33 L). The combined organic phases were dried by azeo-
tropic distillation, and EtOAc (42 L) and Et2NH (9.6 kg, 131.5 mol)
were added at <40 °C. After stirring at 20–25 °C for 12 h, the sus-
pension was filtered. The filter cake was washed with tert-butyl
methyl ether (3 × 30 L) and purged with nitrogen at 40 °C for 48 h.
The diethyl ammonium salt of (2S)-1-(tert-butoxycarbonyl)-2,5-di-
hydro-1H-pyrrole-2-carboxylic acid (29.5 kg, 60%) was obtained as
a colorless solid; Purity by HPLC: 95%; mp 143 °C.
+
NH2 ).
13C NMR (100.61 MHz, DMSO-d6): d = 40.02 (t, CH2Ar), 52.23 (t,
CH2N), 65.97 (d, CHN), 117.44 (s, CN), 125.27 (d, CHCH2),
127.16 (d, CHCH), 128.68 (d, CH), 131.06 (s, CCN), 136.38 (d,
CH), 138.77 (s, CCH2), 150.25 (d, CHN), 166.53 (s, C=O).
HRMS: m/z calcd for C12H13N4O [MH+]: 229.1089; found:
229.1084.
Benzyl (2R)-2-Amino-3-cyclohexylpropanoate·TosOH
(30·TsOH)
(2R)-2-Amino-3-cyclohexylpropanoic acid (29; 61.6 kg, 360 mol)
and 4-methylbenzenesulfonic acid monohydrate (82.2 kg, 432 mol)
were dissolved in toluene (620 L), and benzyl alcohol (58.4 kg, 540
mol) was added. H2O was removed by azeotropic distillation. The
product was crystallized by addition of seeding crystals and cooling
to 5 °C. The product was filtered, washed with tert-butyl methyl
ether (400 L) and dried by purging with N2 (140.5 kg, 90%).
1H NMR (250 MHz, DMSO-d6): d = 0.62 (m, 2 H, aliphatic H),
0.94 (m, 3 H, aliphatic H), 1.44 (m, 8 H, aliphatic H), 2.32 (s, 3 H,
CH3), 3.96 (m, 1 H, CH), 4.96 (d, J = 12.2 Hz, 1 H, CH2), 5.14 (d,
J = 12.2 Hz, 1 H, CH2), 7.06 (d, J = 7.6 Hz, 2 H, CH), 7.24 (s, 5 H,
CH), 7.77 (d, J = 7.6 Hz, 2 H, CH), 8.16 (s, 2 H, NH2).
13C NMR (100.61 MHz, CDCl3): d = 25.68 (t, CH2, c-Hex), 25.68
(t, CH2, c-Hex), 26.16 (t, CH2, c-Hex), 32.25 (t, CH2, c-Hex), 32.69
(t, CH2, c-Hex), 33.18 (d, CH, c-Hex), 38.00 (t, c-Hex–CH2), 51.09
(d, CHNH2), 67.58 (t, CH2Ph), 128.31 (d, CH, C6H5), 128.31 (d,
CH, C6H5), 128.31 (d, CH, C6H5), 128.47 (d, CH, C6H5), 128.47 (d,
CH, C6H5), 134.95 (s, C, C6H5), 169.81 (s, COO).
IR (KBr): 1698, 1640, 1398, 1362, 1119 cm–1.
1H NMR (250 MHz, DMSO-d6): d = 1.12 (t, J = 7.2 Hz, 6 H,
CH3CH2), 1.37 (2 s, 9 H, t-C4H9), 2.80 (q, J = 7.2 Hz, 4 H,
CH3CH2), 4.00 (m, 2 H, CH2), 4.55 (m, 1 H, CHCOO), 5.75 (m, 1
H, 2 CH).
13C NMR (100.61 MHz, DMSO-d6): d (2 rotamers: 2:1) = 11.20 (q,
NCH2CH3), 28.04, 28.18 [q, C(CH3)3], 40.98 (t, NCH2CH3), 53.11,
53.31 (t, CH2), 68.83, 68.95 (d, CH), 77.65, 77.76 [s, C(CH3)3],
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York

2374
H. Bernard et al.
PAPER
(2R)-2-[(tert-Butoxycarbonyl)(2-tert-butoxy-2-oxoethyl)ami-
no]-3-cyclohexylpropanoic Acid·c-HexNH2 (31·c-HexNH2)
Compound 30·TosOH (140 kg, 323 mol) was dissolved in DMF
(360 L) and treated with N-ethyl-N,N-diisopropylamine (125.2 kg,
969 mol). tert-Butyl bromoacetate (69.3 kg, 355 mol) was added
over a period of 4 h at 25 °C. After 20 h, di(tert-butyl) dicarbonate
(84.6 kg, 387 mol) was added. The reaction mixture was heated to
40 °C for 24 h. After the addition of H2O (500 L) and tert-butyl
methyl ether (500 L), the phases were separated. The organic layer
was washed with H2O (400 L), diluted with tert-butyl methyl ether
(100 L) and, after the addition of Pd/C (5%, 4 kg), hydrogenated at
1–2 bar. The catalyst was filtered off. The solution was diluted with
tert-butyl methyl ether (1000 L). By addition of cyclohexylamine
(32 kg, 323 mol) at 45–50 °C and cooling to 5 °C, the product was
crystallized. It was isolated by filtration, washed with tert-butyl
methyl ether (200 L) and dried by purging with N2 (133 kg, 85%).
It was recrystallized from i-PrOH (1330 L) by heating to 71 °C for
1 h and cooling to 5 °C. It was filtered, washed with tert-butyl
methyl ether (400 L) and dried by purging with N2 (120 kg, 77%);
mp 202 °C.
CH2CO2C(CH3)3], 40.40 (t, CH2–py), 45.90 (t, CH2–c-Hex), 52.56
(d, NCHCO), 54.49 (t, CH2CH=), 68.40 (d, CHCH=), 81.35 [s,
NCO2C(CH3)3], 81.77 [s, CH2COOC(CH3)3], 117.32 (s, CN),
126.89 (d, CHCH=), 126.95 (d, CH2CH=), 128.34 (d, CH, py),
132.05 (s, Cpy–CN), 135.27 (d, CH, py), 138.87 (s, CH2–Cpy),
149.87 (d, CH, py), 155.49 [s, NCO2C(CH3)3], 170.10 [s,
CH2CO2C(CH3)3], 170.41 (s, NCO), 170.98 (s, CHCON).
HRMS: m/z calcd for C32H46N5O6 [(M + H)+]: 596.3448; found:
596.3430.
{[(1R)-2-((2S)-2-{[({6-[Amino(imino)methyl]-3-pyridinyl}me-
thyl)amino]carbonyl}-2,5-dihydro-1H-pyrrol-1-yl)-1-(cyclo-
hexylmethyl)-2-oxoethyl]amino}acetic Acid·HCl (1·HCl)
CH2Cl2 from a solution of 11 (83.8 kg, 140 mol) prepared as above
was distilled off. Traces of CH2Cl2 were removed by azeotropic dis-
tillation with MeOH (350 L). The residue was dissolved in MeOH
(120 L) and (2R)-2-(acetylamino)-3-sulfanylpropanoic acid (24 kg,
147 mol) was added. The reaction mixture was purged with ammo-
nia (7.4 kg, 435 mol) over a period of 15 h at 62 °C. MeOH was re-
placed by H2O (250 L) and, after the addition of aq HCl (38%, 41.6
kg, 433 mol), the mixture was heated to 60–70 °C for 3 h. H2O (125
L) was distilled off. The residue was dissolved in DMF (1290 L)
and neutralized with aq ammonia (25%, 60 kg, 882 mol). The be-
taine-DMF complex was filtered and dried at 65 °C (53 kg). After
the addition of aq HCl (32%, 1 equiv), the product was dissolved in
H2O (500 L) and cleared by filtration. DMF was removed by nano-
filtration (Desal 5 DK®, 10–20 °C, 40 bar). The product was ob-
tained by spray drying (N2: 60 m3/h, 120 °C, 2.5 bar, aq solution: 13
L/h). The HCl-salt of 1 was obtained as a colorless solid (31 kg,
45%); mp 133 °C (dec.).
IR (KBr): 2928, 2920, 1741, 1684, 1560, 1179, 1173, 1152 cm–1.
1H NMR (500 MHz, DMSO-d6, 100 °C): d = 0.80–0.96 (m, 2 H, al-
iphatic H), 1.08–1.30 (m, 9 H, aliphatic H), 1.38 (s, 9 H, t-C4H9),
1.42 (s, 9 H, t-C4H9), 1.50–1.72 (m, 9 H, aliphatic H), 1.75–1.86 (m,
+
3 H, aliphatic H), 2.70 (m, 1 H, CHNH3 ), 3.72 (d, J = 17.9 Hz, 1 H,
CH2N), 3.86 (d, J = 17.9 Hz, 1 H, CH2N).
13C NMR (125.76 MHz, MeOH-d4): d (2 rotamers: 1:1) = 25.40 (t,
+
c-Hex-NH3 ), 25.42 (t, c-Hex), 25.96 (t, c-Hex), 27.42, 27.57 (t, c-
Hex), 27.85 (q, t-C4H9), 28.42 (q, t-C4H9), 28.47 (q, t-C4H9), 28.68
+
(q, t-C4H9), 28.80 (q, t-C4H9), 32.06 (t, c-Hex-NH3 ), 33.51, 33.73
IR (KBr): 2924, 1658, 1623 cm–1.
(t, c-Hex), 35.06, 35.20 (t, c-Hex), 35.25, 35.32 (d, c-Hex), 39.75,
40.08 (t, CH2–c-Hex), 48.01, 48.26 (t, CH2N), 51.46 (d, c-Hex-
1H NMR (400 MHz, DMSO-d6): d = 0.86–1.90 (m, 13 H, CH2–c-
Hex), 3.23 (d, J = 16.1 Hz, 1 H, NCH2CO2H), 3.33 (d, J = 16.1 Hz,
1 H, NCH2CO2H), 3.96 (m, 1 H, CHNH), 4.25 (d, J = 13.2 Hz, 1
H, =CHCH2N), 4.40 (dd, J = 16.5, 5.6 Hz, 1 H, CH2–py), 4.50 (dd,
J = 16.6, 6.1 Hz, 1 H, CH2–py), 4.63 (d, J = 13.6 Hz, 1
H, =CHCH2N), 5.06 (m, 1 H, =CHCHN), 5.86 (dd, J = 6.3, 2.1 Hz,
1 H, CH=CH), 6.10 (dd, J = 6.3, 1.9 Hz, 1 H, CH=CH), 7.96 (dd,
J = 8.1, 1.85 Hz, 1 H, CH, py), 8.23 (d, J = 8.1 Hz, 1 H, CH, py),
8.70 (d, J = 1.6 Hz, 1 H, CH, py), 9.00 (t, J = 5.9 Hz, 1 H, NHCH2),
+
NH3 ), 59.29, 60.32 (d, CHCOO–), 81.03, 81.33 (s, t-C4H9), 81.84,
81.98 (s, t-C4H9), 157.58, 157.71 (s, NCO), 171.17, 171.22 (s,
CH2CO), 179.15, 179.44 (s, COO–).
HRMS: m/z calcd for C20H34NO6 [(M – H)–]: 384.2386; found:
384.2387.
tert-Butyl {(tert-Butoxycarbonyl)[(1R)-2-[(2S)-2-({[(6-cyano-3-
pyridinyl)methyl]amino}carbonyl)-2,5-dihydro-1H-pyrrol-1-
yl]-1-(cyclohexylmethyl)-2-oxoethyl]amino}acetate (11)
+
+
9.58 [br s, 2 H, C(NH2)NH2 ], 10.20 [br s, 2 H, C(NH2)NH2 ].
Compound 31·c-HexNH2 (38.8 kg, 80 mol) was dissolved in H2O
(185 L) and CH2Cl2 (310 L) and adjusted to pH 2–3 by addition of
aq HCl (32%, about 9.2 kg). The layers were separated and the
aqueous phase was re-extracted with CH2Cl2 (100 L). The com-
bined organic phases were washed with H2O (100 L) and dried by
azeotropic distillation. To the reaction mixture were added subse-
quently N-ethyl-N,N-diisopropylamine (43.9 kg, 336 mol), 28·HCl
(21.2 kg, 80 mol) and propane phosphonic acid anhydride in EtOAc
(50%, 51 kg, 80 mol) at 0 °C. After 1 h at 0 °C and 12 h at 20–25 °C,
the reaction was quenched by the addition of H2O (150 L) at 10 °C.
The phases were separated, the aqueous phase was re-extracted with
CH2Cl2 (70 L) and the combined organic phases were washed with
dil. aq HCl (pH 3–4, 160 L) and H2O (70 L) and dried by azeotropic
distillation. The solution was directly used in the next step. For an-
alytical purposes a sample was taken and, after evaporation of the
solvent, subjected to HPLC (purity >94%, 45.3 kg, 95%) and to
spectroscopic analysis; mp 72 °C.
13C NMR (100.61 MHz, DMSO-d6): d = 25.38 (t, CH2, c-Hex),
25.59 (t, CH2, c-Hex), 25.81 (t, CH2, c-Hex), 31.91 (t, CH2, c-Hex),
33.22 (t, CH2, c-Hex), 32.85 (d, CH, c-Hex), 37.52 (t, CH2–c-Hex),
40.00 (t, CH2–py), 47.58 (t, CH2CO2H), 53.38 (t, CH2CH=), 55.02
(d, NCHCO), 68.09 (d, CHCH=), 122.78 (d, CH, py), 125.68 (d,
CH2CH=), 127.89 (d, CHCH=), 136.05 (d, CH, py), 139.99 (s,
+
CH2–Cpy), 142.47 [s, Cpy–C(NH2)NH2 ], 148.44 (d, CH, py),
+
162.28 [s, C(NH2)NH2 ], 168.58 (s, CHCON), 169.01 (s, NCO),
170.26 (s, CH2CO2H).
HRMS: m/z calcd for C23H33N6O4 [(M + H)+]: 457.2563; found:
457.2563.
Acknowledgment
We are indebted to Dr. Helmut Hagen, Dr. Stefan Koser, Dr. Stefan
Müller, Dr. Hartmut Riechers, Dr. Bernd-Peter Walther and Dr.
Volker Wolfart for valuable collaboration. The analytical and theo-
retical help of Karlheinz Rauwolf (HPLC), Dr. Norbert Zimmer-
mann (NMR), Dr. Walter Meyer (NMR), and Dr. Ansgar Schäfer
(quantum-mechanical calculations) are gratefully acknowledged.
We particularly thank Dr. Ernst Buschmann for countless, encoura-
ging discussions over the years.
IR (KBr): 2926, 1749, 1699, 1657, 1447, 1393, 1368, 1164 cm–1.
13C NMR (100.61 MHz, CDCl3): d = 26.06 (t, CH2, c-Hex), 26.26
(t, CH2, c-Hex), 26.33 (t, CH2, c-Hex), 27.98 [3 q,
CH2CO2C(CH3)3], 28.13 [3 q, NCO2C(CH3)3], 32.96 (t, CH2, c-
Hex), 33.83 (t, CH2, c-Hex), 33.98 (d, CH, c-Hex), 36.58 [t,
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York

PAPER
Technical Scale Synthesis of a New and Highly Potent Thrombin Inhibitor
2375
(19) Pfeiffer, T.; Seitz, W.; Mack, H.; Zierke, T.; Balkenhohl, F.;
Lange, U. German Patent 19630082, 1996; Chem. Abstr.
1998, 128, 154380.
References
(1) (a) Sorbera, L. A.; Bayes, M.; Castaner, J.; Silvestre, J.
Drugs Future 2001, 12, 1155. (b) Pfau, R. Curr. Opin. Drug
Disc. 2003, 6, 437. (c) Sanderson, P. E. J.; Stanton, M. G.;
Dorsey, B. D.; Lyle, T. A.; McDonough, C.; Sanders, W. M.;
Savage, K. L.; Naylor-Olsen, A. M.; Krueger, J. A.; Lewis,
S. D.; Lucas, B. J.; Lynch, J. J.; Yan, Y. Bioorg. Med. Chem.
Lett. 2003, 13, 795. (d) Olsen, J. A.; Banner, D. W.; Seiler,
P.; Obst Sander, U.; D’Arcy, A.; Stihle, M.; Müller, K.;
Diederich, F. Angew. Chem. Int. Ed. 2003, 42, 2507.
(e) Nilsson, J. W.; Kvarnström, I.; Musil, D.; Nilsson, I.;
Samulesson, B. J. Med. Chem. 2003, 46, 3985.
(2) Hemostasis and Thrombosis: Basic Principles and Clinical
Practice, 3rd ed.; Colman, R. W.; Hirsch, J.; Marder, V. J.;
Salzman, E. W., Eds.; Lippincott J. B.: Philadelphia, 1994.
(3) (a) Wiley, M. R.; Fisher, M. J. Exp. Opin. Ther. Patents
1997, 7, 1265. (b) Rewinkel, J. B. M.; Adang, A. E. P. Curr.
Pharm. Design 1999, 5, 1043.
(20) CAUTION! Thermal runaway reaction with NaH and DMF
are known in literature: (a) DeWall, G. Chem. Eng. News
1982, 60(37), 5. (b) DeWall, G. Chem. Eng. News 1982,
60(37), 43. (c) Buckley, J.; Webb, R. L.; Laird, T.; Ward, R.
J. Chem. Eng. News 1982, 60(28), 5. (d) Bretherick, L.
Handbook of Reactive Chemical Hazards, 4th ed.;
Butterworth-Heinemann Ltd.: Boston, 1990, 1181.
(21) (a) Jones, M. F. J. Chem. Soc., Perkin Trans. 1 1991, 2479.
(b) Rapoport, H. J. Org. Chem. 1994, 54, 394.
(22) (a) Pinner, A. Ber. Dtsch. Chem. Ges. 1885, 18, 2845.
(b) Stilz, H. U.; Jablonka, B.; Just, M.; Knolle, J.; Paulus, E.
F.; Zoller, G. J. Med. Chem. 1996, 39, 2118.
(23) Schaefer, F. C.; Peters, G. A. J. Org. Chem. 1961, 26, 412.
(24) (a) Medwid, J. B. J. Med. Chem. 1990, 33, 1230. (b) Soula,
C.; Marsura, A.; Luu-Duc, C. J. Pharm. Belg. 1987, 42, 293;
Chem. Abstr. 1988, 109, 109969. (c) Hullin, R. P.; Miller,
J.; Short, W. F. J. Org. Chem. 1947, 12, 394. (d) Thurkauf,
A. J. Labelled Compd. Radiopharm. 1997, 39, 123.
(e) Garigipati, R. S. Tetrahedron Lett. 1990, 31, 1969.
(f) Moss, R. A.; Ma, W.; Merrer, D. C.; Xue, S. Tetrahedron
Lett. 1995, 36, 8761. (g) Kirby, J. B.; van Dantzig, N. A.;
Chang, C. K.; Nocera, D. G. Tetrahedron Lett. 1995, 36,
3477.
(25) (a) Jendralla, H.; Seuring, B.; Herchen, J.; Kulitzscher, B.;
Wunner, J.; Stüber, W.; Koschinsky, R. Tetrahedron 1995,
51, 12047. (b) Judkins, B. D.; Allen, D. G.; Cook, T. A.;
Evans, B.; Sardharwalla, T. A. Synth. Commun. 1996, 26,
4351. (c) Bolton, R. E.; Coote, S. J.; Finch, H.; Lowdon, H.;
Pegg, N.; Vinader, M. V. Tetrahedron Lett. 1995, 36, 4471.
(26) Partridge, M. W.; Short, W. F. J. Org. Chem. 1947, 12, 390.
(27) Lange, U. E. W.; Schäfer, B.; Baucke, D.; Buschmann, E.;
Mack, H. Tetrahedron Lett. 1999, 40, 7067.
(28) Moser, H.; Fliri, A.; Steiger, A.; Costello, G.; Schreiber, J.;
Eschenmoser, A. Helv. Chim. Acta 1986, 69, 1224.
(29) (a) Bergeron, H. J.; Niegard, J.; Dichs, J. B.; Egli-Karmakka,
M.; Frei, J.; Huxley-Tencer, A.; Peter, H. H. J. Med. Chem.
1991, 34, 2072. (b) Kwiatkowski, S.; Crocker, P. J.; Chavan,
A. J.; Imai, N.; Haley, B. E.; Watt, D. S. Tetrahedron Lett.
1990, 31, 2093. (c) Ehrler, J.; Farooq, S. Synlett 1994, 702.
(30) Reviews: (a) Dalko, P. I.; Moisan, L. Angew. Chem. Int. Ed.
2001, 40, 3726; Angew. Chem. 2001, 113, 3840. (b) Jarvo,
E. R.; Miller, S. J. Tetrahedron 2002, 58, 2481.
(4) (a) Böhm, H.-J.; Hoeffken, H. W.; Hornberger, W.; Koser,
S.; Mack, H.; Pfeiffer, T.; Seitz, W.; Zierke, T. World Patent
9625426, 1995; Chem. Abstr. 1996, 125, 301605.
(b) Lange, U. E. W.; Baucke, D.; Hornberger, W.; Mack, H.;
Seitz, W.; Höffken, W. Bioorg. Med. Chem. Lett. 2003, 13,
2029. (c) Lange, U. E. W.; Baucke, D.; Hornberger, W.;
Mack, H.; Seitz, W.; Höffken, W., manuscript in
preparation.
(5) Lit. Ref.4a, Example 32 and 93.
(6) Rüeger, H.; Benn, M. H. Can. J. Chem. 1982, 60, 2918.
(7) Barker, P. L.; Gendler, P. L.; Rapoport, H. J. Org. Chem.
1981, 46, 2455.
(8) (a) Wissmann, H.; Kleiner, H.-J. Angew. Chem., Int. Ed.
Engl. 1980, 19, 133; Angew. Chem. 1980, 92, 129.
(b) Wissmann, H. Phosphorus and Sulfur 1987, 30, 645.
(c) Escher, R.; Bünning, P. Angew. Chem., Int. Ed. Engl.
1986, 25, 277; Angew. Chem. 1986, 98, 264.
(9) (a) Engbersen, J. F. J.; Koudijs, A.; Joosten, M. H. A.; van
der Plas, H. C. J. Heterocycl. Chem. 1986, 13, 989.
(b) Takamizawa, S.; Wakasa, N.; Fuchikami, T. Synlett
2001, 1623. (c) Bullock, M. W.; Hand, J. J.; Stokstad, E. L.
R. J. Am. Chem. Soc. 1956, 78, 3693. (d) Atkins, H.; Wolff,
I. A.; Pavlic, A.; Hutchinson, E. J. Am. Chem. Soc. 1944, 66,
1293. (e) Hilpert, K.; Ackermann, J.; Banner, D. W.; Gast,
A.; Gubernator, K.; Hadváry, P.; Labler, L.; Müller, K.;
Schmid, G.; Tschopp, T. B.; van der Waterbeemd, H. J. Med.
Chem. 1994, 37, 3889.
(31) Henecka, H.; Kurtz, P. In Houben-Weyl Methoden der
Organischen Chemie, 4th ed., Vol. 8; Thieme: Stuttgart,
1952, 703.
(10) Bouchet, M.-J.; Rendon, A.; Wermuth, C. G.; Goeldner, M.;
Hirth, C. J. Med. Chem. 1987, 30, 2222.
(11) (a) Minisci, F.; Galli, R. Tetrahedron Lett. 1965, 6, 433.
(b) Minisci, F. Synthesis 1973, 1. (c) Minisci, F.; Citterio,
A.; Vismara, E.; Giordano, C. Tetrahedron 1985, 41, 4157.
(12) Houssin, R.; Bernier, J.-L.; Hénichart, J.-P. Synthesis 1988,
259.
(13) Donohoe, T. J.; Guyo, P. M. J. Org. Chem. 1996, 61, 7664.
(14) Harbuck, J. W.; Rapoport, H. J. Org. Chem. 1972, 37, 3618.
(15) Bailey, D. M.; Johnson, R. E.; Albertson, N. F. Org. Synth.
1971, 51, 100.
(16) Grehn, L.; Ragnarsson, U. Angew. Chem. Int. Ed. Engl.
1984, 23, 296; Angew. Chem. 1984, 96, 291.
(17) Stürmer, R.; Schäfer, B.; Wolfart, V.; Stahr, H.; Kazmaier,
U.; Helmchen, G. Synthesis 2001, 46.
(32) For ion exchange chromatography an acetate ion exchanger
from BIO RAD was used: AG 1-X8 Resin, 100–200 mesh,
acetate form (catalog-Nr.: 140-1443) Address: BIO RAD
(BIO RAD laboratories, 2000 Alfred Nobel Dr., Hercules,
CA 94547. Due to the water content (48%) before
chromatography the ion exchanger was treated with an
excess of anhyd MeOH. Based on the content of acetate
more than two equivalents were used. The column should be
thin and long. For ion exchange chromatography MeOH was
used as the solvent.
(33) (a) Razdan, U.; Shah, V. J. J. Sci. Ind. Res. 2001, 60, 560.
(b) Dijkstra, H. P.; van Klink, G. P. M.; van Koten, G. Acc.
Chem. Res. 2002, 35, 798.
(34) Schäfer, B.; Harms, G.; Ascherl, H.; Krei, G. A. World
Patent 200068254, 1999; Chem. Abstr. 2000, 133, 366431.
(35) Schäfer, B. European Patent 1054017, 1999; Chem. Abstr.
2000, 133, 366468.
(18) (a) Dormay, J.-R.; Castro, B.; Chappuis, G.; Fritschi, U.-S.;
Grogg, P. Angew. Chem. Int. Ed. Engl. 1980, 19, 742;
Angew. Chem. 1980, 92, 761. (b) Dormay, J.-R. Synthesis
1982, 753.
(36) ICH Guideline Q7A: http://www.fda.gov/cder/guidance/
1289dft.pdf.
Synthesis 2004, No. 14, 2367–2375 © Thieme Stuttgart · New York