First enantioselective synthesis of the novel antiinfective TAN-1057A via its aminomethyl-substituted dihydropyrimidinone heterocycle

Article abstract of DOI:10.1016/j.tet.2004.06.034

Enantiomerically pure N²-Z-N²-MeAsnOH [(S)-14], prepared in 8 steps (23% overall yield) from asparaginic acid, was first subjected to a Hofmann degradation with PhI(OCOCF₃)₂ yielding (S)-N²-Z-N²-methyl-2,3-diaminopropanoic acid [N²-Z-N²-Me-L-A₂pr, (S)-15], and this in turn was protected to give N²-Z-N³-Boc-N²-Me-L- A₂pr [(S)-17]. Condensation of (S)-17 with HNC(SMe)NHCONH₂ followed by removal of the tert-butoxycarbonyl protecting group, cyclization and hydrogenolytic removal of the Z-group gave the heterocycle of TAN-1057A [(S)-1] with an e.e. of 87 in 36% yield [from (S)-14]. Coupling of (S)-1 with (S)-tris-Z-β-homoarginine (20a) in the presence of O-(7-azabenzotriazol-1- yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate (HATU) and iPr₂NEt in N,N-dimethylacetamide followed by hydrogenolysis afforded the most active A-diastereomer of the natural antibiotic TAN-1057 in 52% yield (from (S)-1 and 20a). Similarly, starting from (S)-1, a single diastereomer of the potent, less toxic TAN-1057A analogue 22b with a β-lysine side chain has been prepared. All described synthetic steps do not require column chromatography for purification of the products.

Full text of DOI:10.1016/j.tet.2004.06.034

Tetrahedron 60 (2004) 7579–7589
First enantioselective synthesis of the novel antiinfective
TAN-1057A via its aminomethyl-substituted
dihydropyrimidinone heterocycle
Vladimir N. Belov,a,b Michael Brands,c Siegfried Raddatz,d Jochen Kru¨ger,d Sofia Nikolskaya,a,e
Viktor Sokolova,e and Armin de Meijerea,b
,
*
a
¨ ¨ ¨
Institut fur Organische und Biomolekulare Chemie, Georg-August Universitat Gottingen, Tammannstrasse 2, D-37077 Gottingen, Germany
KAdemCustomChem GmbH, Brombeerweg 13, D-37077 Gottingen, Germany
¨
b
¨
cBAYER Corporation, Pharmaceutical Division, 400 Morgen Lane-B 27, West Haven, CT 06516-4175, USA
dPharma Research Center, Bayer HealthCare AG, D-42096 Wuppertal, Germany
eDepartment of Chemistry, St. Petersburg State University, Universitetskii Pr. 26, 198504 St. Petersburg, Petrodvorets, Russian Federation
Received 12 March 2004; revised 25 May 2004; accepted 3 June 2004
Available online 2 July 2004
Dedicated to Professor Dieter Seebach
Abstract—Enantiomerically pure N 2-Z-N 2-MeAsnOH [(S)-14], prepared in 8 steps (23% overall yield) from asparaginic acid, was first
subjected to a Hofmann degradation with PhI(OCOCF3)2 yielding (S)-N 2-Z-N 2-methyl-2,3-diaminopropanoic acid [N 2-Z-N 2-Me-L-A2pr,
(S)-15], and this in turn was protected to give N 2-Z-N 3-Boc-N 2-Me-L-A2pr [(S)-17]. Condensation of (S)-17 with HNvC(SMe)NHCONH2
followed by removal of the tert-butoxycarbonyl protecting group, cyclization and hydrogenolytic removal of the Z-group gave the
heterocycle of TAN-1057A [(S)-1] with an e.e. of 87 in 36% yield [from (S)-14]. Coupling of (S)-1 with (S)-tris-Z-b-homoarginine (20a) in
the presence of O-(7-azabenzotriazol-1-yl)-N,N,N0,N0-tetramethyluronium hexafluorophosphate (HATU) and iPr2NEt in N,N-dimethyl-
acetamide followed by hydrogenolysis afforded the most active A-diastereomer of the natural antibiotic TAN-1057 in 52% yield (from (S)-1
and 20a). Similarly, starting from (S)-1, a single diastereomer of the potent, less toxic TAN-1057A analogue 22b with a b-lysine side chain
has been prepared. All described synthetic steps do not require column chromatography for purification of the products.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
in the CD spectra which were recorded for the product
obtained by mild hydrolysis (2% aq. Na2CO3, 70 8C, 1 h) of
TAN-1057A (acyclic dipeptide 2), and the model compound
3 with its known (S)-configuration at C-2 (Fig. 1).2
A mixture of diastereomers of the very potent antibiotics
TAN-1057A/B was first isolated by Japanese scientists at
the Takeda company from the culture broth of Flexibacter
sp. PK-74.1 These compounds were found to be highly
active against methicillin-resistant Staphylococcus aureus
(MRSA) strains.1,2 Both main components of this mixture
were found to be dipeptides with an (S)-b-homoarginine
side chain attached to a dihydropyrimidinone heterocyclic
moiety.2 The Takeda group succeeded in separating the
natural diastereomeric mixture of TAN-1057A/B (6.0 g)
into two fractions and isolated TAN-1057A (1.56 g) and
TAN-1057B (5 mg) as dihydrochlorides.2
(S)-Configuration of C-5 in the heterocyclic portion of the
molecule was assigned to TAN-1057A, while (R)-con-
figuration of C-5 was attributed to TAN-1057B. Epimeriza-
tion at C-5 in TAN-1057A occurs in MeOH in the presence
of MeONa, or during heating in aqueous HCl. Acidic
degradation of the molecule yields derivatives of N-methyl-
2,3-diaminopropionic acid, a constituent amino acid of the
TAN-1057 heterocycle.
The first successful synthesis of TAN-1057A/B in which an
attempt to prepare TAN-1057A as a single diastereomer
was made, instead resulted in a diastereomeric mixture of
TAN-1057A and B due to the lability of the 2,3-diamino-
propionic acid to epimerization.3 Before embarking on the
present study, it was not clear, whether this stereogenic
center is configurationally stable enough to allow for a total
synthesis of the enantiomerically pure material in vitro.
The absolute configurations at C-5 for both epimers of
TAN-1057A/B were assigned on the basis of the similarities
Keywords: Total synthesis; Amino acids; Chiral pool; Nitrogen
heterocycles; Natural products.
*
Corresponding author. Tel.: þ49-551-393231; fax: þ49-551-399475;
e-mail address: armin.demeijere@chemie.uni-gottingen.de
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.06.034

7580
V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
Figure 1. Structures and absolute configurations of TAN-1057A/B and the dihydropyrimidinone derivative 1 contained therein.
Up to now, most of the unique biological activity data have
been determined for the diastereomeric mixture of TAN-
1057A,B. For example, no cross-resistance between TAN-
1057A,B and methicillin, erythromycin and gentamycin
was found. TAN-1057A was shown to be by a factor of 2–4
more potent against Gram-positive and Gram-negative
bacteria than TAN-1057B.1,2 TAN-1057A is efficient
against all of the MRSA strains evaluated, and was found
to compare very favorably to vancomycin in an in vivo
infection model.1 Therefore, for further tests it was deemed
necessary to elaborate a stereoselective access to either
diastereomer of TAN-1057. This synthetic procedure should
also be applicable for the facile preparation of dia-
stereomerically pure analogues of TAN-10574 with
improved properties (more active, less toxic) and therefore
might lead to the discovery of new therapeutically useful
drugs. We have reported convenient, convergent syntheses
of TAN-1057 A/B and its analogues with various side
chains4b–e utilizing the heterocycle 1 as a starting material.5
sation of N 2-Z-N 2-Me-A2prOMe [(RS)-4] with 2-methyl-2-
thiopseudobiuret hydroiodide (5) and subsequent removal
of the Z-protecting group from the resulting (RS)-6
(Scheme 1) by hydrogenolysis.5
With a similar approach to the enantiomerically pure
heterocycle (S)-1 in mind, optically pure N 2-Z-L-A2pr and
N 2-Boc-L-A2pr were prepared from the correspondingly
protected asparagines.7 In order to be able to create the
2,3-diaminopropionic acid fragment at a later step of the
synthesis, N 2-Z-N 2-Me-L-AsnOH was needed. Selective
N 2-methylation of asparagine is not possible, unless the
amido group (CONH2) is blocked. Therefore, the simplest
way to overcome this difficulty was to use the aspartic acid
with an appropriate protection which discriminates the two
carboxyl groups, and then, after performing the mono
N-methylation, to transform the terminal carboxyl group
into its corresponding amide. Thus, L-aspartic acid was first
converted into Z-L-Asp(OMe)OH [(S)-8]8 which was then
protected by the formation of the tert-butyl ester (S)-9
(Scheme 2).9
In designing a synthetic strategy towards the enantio-
merically pure heterocycle (S)-1, the known facile racemi-
zation of N 2-methyl-2,3-diaminopropionic acid derivatives
had to be taken into account,3 therefore a reliable approach
to the suitably protected N 2-methyl-2,3-diaminopropionic
acid (N 2-Me-L-A2pr) in enantiomerically pure form was the
key issue.
N-Methylation of (S)-9 was achieved with methyl iodide in
DMF in the presence of silver(I) oxide as a base according
to the known method.10 As stated in the original publication,
no racemization was detected under these conditions.
Nevertheless, due to the strong basicity of Ag2O, this step
as well as the next one, are the most dangerous in this
respect. Saponification of the methyl ester (S)-10 was
performed carefully by slow addition of a slight excess of
lithium hydroxide solution at þ5 8C. The acid (S)-11 may be
converted into the amide (S)-13 in a one pot operation by
treatment first with N-hydroxybenzotriazole (HOBt) and
N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide (EDC) in
THF at 0 8C, and then with conc. aq. NH3. However, it is
2. Results and discussion
2.1. Enantioselective synthesis of the heterocycle in
TAN-1057A
The racemic heterocycle (RS)-1 was prepared by conden-
Scheme 1. Improved conditions of the one-step cyclization to the racemic heterocycle (RS)-10: (a) AcONa, MeCN, 55 8C, 48 h.6

V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
7581
Scheme 2. Stereoselective synthesis of N 2-benzyloxycarbonyl-N 2-methyl-L-asparagine (N 2-Z-N 2-Me-L-AsnOH, (S)-14) and (S)-3-amino-2-(N-
benzyloxycarbonyl-N-methyl)aminopropionic acid (N 2-Z-N 2-Me-L-A2pr, (S)-15): (a) SOCl2, MeOH, 210!þ20 8C, 25 min; (b) ZCl, MgO, H2O/ether,
þ5 8C, 6 h; (c) isobutene, conc. H2SO4, CH2Cl2, 20 8C, 48 h; (d) MeI (7 equiv.), Ag2O (1.02 equiv.), DMF, 20 8C, 7 h; (e) LiOH (1.0 equiv.), MeOH/H2O,
þ5!20 8C; (f) N-hydroxysuccinimide (1.1 equiv.), EtOAc, DCC (1.1 equiv.) in dioxane, þ5 8C, overnight; (g) aq. NH3, THF, 0 8C; (h) TFA, CH2Cl2, 20 8C,
4 h; (i) PhI(OCOCF3)2, DMF/H2O, pyridine, 20 8C, 16 h.
better to first transform (S)-11 into the crystalline O-suc-
cinimidyl ester (S)-12. With this, two goals were achieved:
firstly, the ester (S)-12 could be easily purified by
recrystallization from a dioxane/ether mixture, and,
secondly, the v-carboxyl group was activated for subse-
quent conversion to the amide. Afterwards, the a-carboxyl
group in (S)-13 was deprotected under standard conditions,
and the key intermediate N a-Z-N a-MeAsnOH [(S)-14] was
isolated. In several experiments the enantiomeric excess of
the crude (S)-14 was always found to be about 77% (HPLC
analysis on a chiral stationary phase; see Section 4 for
details). Therefore, considerable racemization must have
taken place at an earlier stage. Luckily, the optical purity
could easily be restored after recrystallization from an
iPrOH/EtOAc mixture. The racemic mixture (RS)-14 (mp
160–162 8C) is much less soluble in iPrOH and water, and
therefore the soluble fraction contains mostly the material
with high enantiomeric excess. By this procedure, optically
pure (S)-14 (.99% e.e.) was obtained in 58% yield after one
or two recrystallizations (mp 129 8C). When this work was
completed, a general approach to N-methylamino acids by
way of intermediate 5-oxazolidinones appeared in print.11
By this published approach, compound (S)-14 has been
prepared in only 4 steps.11a However, two of these four steps
require column chromatography, certainly a drawback,
when scale-up would be required.12
Direct amidation of the methyl ester (S)-10 with methanolic
ammonia was found to be difficult, as it required prolonged
heating at about 100 8C, or the reaction with liquid ammonia
at room temperature.13 Due to the high probability of
racemization, these reactions were not tried on a preparative
scale. Hofmann degradation of the amide (S)-14 proceeded
under the same conditions as for the racemic compound5
and for N 2-Z(Boc)-L-AsnOH,7 and thus the 2,3-diamino-
propionic acid fragment in the target compound (S)-15 was
created.
An attempt to synthesize (S)-15 along a shorter route was
made (Scheme 3). Curtius degradation of the azide which
was generated from the acid (S)-11 and O,O-diphenylphos-
phoryl azide (DPPA) in the presence of tBuOH led to the
Boc-protected amino ester (S)-16. After purification by
chromatography on silica gel, it was isolated in moderate
Scheme 3. Preparation of N 2-Z-N 2-Me-L-A2pr [(S)-15] through Curtius degradation of (S)-11: (a) DPPA (1.1 equiv.), Et3N (1 equiv.), tBuOH, 80 8C, 17 h;
(b) TFA, 0 8C.

7582
V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
Scheme 4. Enantioselective synthesis of the heterocycle (S)-1: (a) 1.3 M aq. NaOH, Boc2O, tBuOH; (b) HOBt, EDC, N,N-diisopropylethylamine (DIEA),
CH2Cl2; (c) TFA, anisole, CH2Cl2, 20 8C; (d) Et3N, AcOH (pH 8–9); (e) H2, Pd/C, N,N-dimethylacetamide (DMAA).
yield and transformed directly to the amino acid (S)-15 by
simultaneous removal of both tert-butyl-containing protect-
ing groups.
as worked out for the preparation of the racemic heterocycle
(RS)-6 could not be used (Scheme 1). Therefore, a two-step
cyclization was employed (Scheme 4, cf.3).
The optical purity of (S)-15 obtained along this route was
lower: it was necessary to recrystallize it twice from an
EtOH/H2O mixture in order to get an optical rotation value
which was similar to that obtained earlier ([a]2D0¼243.5
versus 245.2 (Scheme 2), c¼1, H2O). Lower optical
purity, higher losses during recrystallizations, the moderate
yield of the intermediate (S)-16, and the necessity to
purify it by column chromatography render this route
inappropriate.
The monoprotected diamino acid (S)-15 was first orthogo-
nally bisprotected, and then coupled with easily available
S-methylisothiobiuret hydroiodide (5) to give the inter-
mediate (S)-15. In 8 the nitrogen next to the methylthio
group turned out to be more nucleophilic than any of the
other two, thus it is not necessary to use a protected equiva-
lent of the compound 5, as reported by Williams et al.3 In
monothiobiuret 5, the nucleophilicity of the nitrogen atom
next to the MeS-group proved to be much higher than that of
the other two nitrogens, and the selectivity of the coupling
was very high. Removal of the Boc-protecting group from
the coupling product (S)-18 followed by cyclization under
very mild basic conditions afforded (S)-6 (e.e.¼92%).
Hydrogenolysis under ordinary conditions gave the desired
heterocycle (S)-1 with an enantiomeric excess of 87%. This
means that in the six synthetic steps the enantiomeric excess
did decrease, but not drastically.
Surprisingly, esterification of (S)-15 in methanol with
thionyl chloride (HCl and (MeO)2SO) was accompanied
by considerable epimerization, and the e.e. of the amino
ester (S)-4 was found to be only 30–50%. Probably, the
strongly electron-withdrawing properties of the protonated
primary amino group enhance the a-CH acidity at the
stereogenic center adjacent to the ester group (which may
also be protonated in the strongly acidic medium), and
therefore, rapid racemization occurs.14
2.2. Diastereoselective synthesis of TAN-1057A as well as
an analogue with lower acute toxicity
Since the preparation of the amino ester (S)-4 with high
enantiomeric excess failed, the direct one-step cyclization
The last task to complete the synthesis of TAN-1057A was
Scheme 5. Diastereoselective synthesis of TAN-1057Ap2HCl and its analogue with a b-lysine side chain: (a) hn, dioxane/water, 30 8C, 2 h; (b) HATU, DIEA,
DMAA, 20 8C; (c) MeOH, H2, PdCl2, 25 8C.

V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
7583
to connect the (S)-tris-Z-b-homoarginine (20a) with the
enantiomerically enriched heterocycle (S)-1 (Scheme 5).
Towards this goal, the tris-Z-protected diazoketone 19,5
precursor to 20a, was irradiated with a daylight lamp in a
dioxane/water mixture to yield the acid 20a (47%).15
N-methyl group (near 3.1 ppm), as there is one signal for
each epimer.
As has been reported previously,4b – e several novel
analogues of TAN-1057A,B showed lower acute toxicity
compared to the natural product, yet with concomitant
retention of excellent anti-microbial activity. For example,
the analogue of the compound (S,S)-22b (Scheme 5) first
prepared from the racemic heterocycle (RS)-6 and bis-Z-b-
(S)-lysine (20b) as a 1:1 mixture of two diastereomers was
shown to be at least 4 times less toxic than TAN-1057A/B
(mice, i. p.), while in vivo anti-staphylococcal activity was
nearly the same.4b,d Therefore the pure diastereomer
(30S,5S)-21b was synthesized by coupling the enantiomeri-
cally enriched heterocycle (S)-1 (e.e.¼75–87%) with bis-Z-
b-(S)-lysine (20b). After recrystallization, the obtained
product 21b (55% yield) had de¼85–90%. Final deprotec-
tion of (S,S)-21b was performed under the same conditions
as for TAN-1057A. The CD-spectrum of (30S,5S)-22b is
very similar to that of TAN-1057A. The diastereomeric
purity of this sample was also confirmed by 1H NMR
spectroscopy. As for TAN-1057A, there is only one set of
signals including well-resolved sharp singlets of the N-Me
group at 3.16 and 2.89 ppm (the latter is much less intense
and indicates the presence of a small amount of the second
amide rotamer).
Several coupling reagents were tried. Treatment with EDC
(alone or in the presence of HOBt or 7-aza-1-hydroxyben-
zotriazole) failed to give the coupling product (S,S)-21a.
The desired compound was prepared in high crude yield
(84%) by using of 2 equiv. of HATU and DIEA. The
coupling reaction was not accompanied by any detectable
epimerization: the diastereomeric ratio of TAN-1057A/B in
the crude product was about 93:7, virtually the same as the
enantiomeric ratio of the starting heterocycle (S)-1 with an
e.e.¼87%. The minor epimer was removed completely by
recrystallization from dichloromethane. However, recrys-
tallization was accompanied by considerable losses, so that
the isolated yield of the pure A-diastereomer dropped to
52% from about 79% calculated from the crude yield of the
93:7 mixture. The purity of the recrystallized coupling
product (S,S)-21a was confirmed by LC–MS and the
elemental analysis.
Compound (S,S)-21a exists as a mixture of two amide
rotamers which (in [D6]DMSO) display two singlets of
1
N-methyl groups in the H NMR spectrum (600 MHz): a
low-field broad resonance with higher intensity at 2.85 ppm,
and another (sharp) resonance at 2.66 ppm. The 1:1 epi-
meric mixture (C-5) of the coupling product 21a prepared
from the racemic heterocycle (RS)-15 shows three singlets
3. Conclusion
The first synthesis of the chiral heterocycle of TAN-1057A
with high enantiomeric purity has been accomplished. The
overall yield was about 9% over 14 steps, starting from
L-aspartic acid. The same methodology is in principle
applicable for the synthesis of the (R)-enantiomer (R)-1.
Starting from (S)-1, diastereomerically pure TAN-1057A
was synthesized (5% yield over 16 steps), and thus the
absolute configurations at C-5 in both natural diastereomers
of TAN-1057 were rigorously confirmed. No chromato-
graphic separations were necessary, therefore these syn-
thetic procedures may easily be scaled-up. In addition, the
route to numerous analogues with various side-chains is
feasible (for example, compound (S,S)-22b). Enantiomeri-
cally pure orthogonally protected (S)-N 2-methyl-2,3-dia-
minopropionic acid (S)-17 also is a valuable intermediate
for the synthesis of other biologically active compounds.17
1
of N-methyl groups in the H NMR spectrum: a low field
resonance (2.85 ppm) and two closely positioned sharp
singlets with d¼2.65 and 2.66 ppm. Each of them is ca. two
times less intensive than the high-field singlet (2.66 ppm)
for the A-diastereomer. Therefore, the degree of dia-
1
stereomeric purity may easily be estimated by H NMR
spectroscopy. An attempted separation of 21a by HPLC
failed.
The final deprotection of (S,S)-21a by catalytic hydro-
genation with PdCl2 as a (pre)catalyst and a source of HCl
necessary for the salt formation, was also found to occur
without epimerization. The CD-spectrum of this sample was
identical to that of HPLC-purified TAN-1057A reported
earlier.2 This rigorously proves the assignment of the
absolute configuration at C-5 initially made by the Takeda
group.2 The absolute value of the optical rotation of pure
TAN-1057Ap2HCl found here is lower than that previously
reported ([a]D22¼222.7 versus 239.1,2 c¼0.53, H2O).16
However, this lower value cannot be explained with
epimerization, though the B-epimer has [a]2D2¼þ72.6
(c¼0.53, H2O).2 The diastereomeric purity of this current
sample was established unequivocally by comparison of its
1H NMR spectrum (600 MHz) with that of the epimeric
mixture. There is a striking contrast between them. The
synthetic sample has one set of signals; only the N-methyl
group displays two sharp singlets of the two rotamers: at
3.12 ppm and a much less intensive one at 2.88 ppm. In the
case of the epimeric mixture of TAN-1057A,B, nearly all
the signals are doubled. The most convenient way to
estimate the diastereomeric purity is by measuring the
relative intensities of the two resolved singlets of the
4. Experimental
4.1. General
Melting points (uncorrected) were determined in capillaries
using a Bu¨chi 510 apparatus. IR: Bruker IFS 66 (FT-IR)
spectrometer, measured as KBr pellets. H NMR: Bruker
1
AM 250 instrument (250 MHz), and VARIAN INOVA-600
spectrometer (600 MHz); 13C (and DEPT) NMR: Bruker
AM 250 at 62.9 MHz and Varian UNITY 300 at 75.5 MHz.
All spectra are calibrated against tetramethylsilane as an
internal standard (d¼0) or the signals of residual protons of
deuterated solvents: 7.26 for CHCl3, 2.50 for [D5]DMSO
and 3.30 for [D3]MeOH. Multiplicities of signals are
reported as follows: s¼singlet, d¼doublet, t¼triplet,

7584
V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
q¼quartet, quint¼quintet, mc¼centrosymmetrical multi-
plet. Coupling constants (J) are given in Hz. EI-MS:
Finnigan MAT 95 and Varian CH 5 spectrometers (70 eV).
HPLC-MS: Hewlett–Packard 1100 instrument (Micromass
Platform LCZ). HRMS: Micromass LCT (TOF MS,
electrospray ionization, positive and negative modes).
5.25 (m, 2H, OCH2), 7.28 (br. s, 5H); 13C NMR (62.9 MHz,
CDCl3, the signals of the major rotamer are marked with p)
d 27.7p/27.8 (Me), 33.5/34.4p (NMe), 34.9p/36.3 (CH2),
51.7p/51.8 (OMe), 57.6/58.2p (CHN), 67.1/67.4p (OCH2),
81.9p/82.1 (C–O), 127.7 (CH), 127.9 (CH), 128.3 (CH),
136.5 (C), 156.01/156.07p (NCvO), 168.85p/168.91
(CvO), 171.1/171.3p (CvO).
HPLC-MS parameters: column: Kromasil
C
18;
50£2.1 mm; eluent A: 5 mL 70% HClO4 in 1 L H2O;
eluent B. MeCN; 0–0.5 min: 98% Aþ2% B, 0.5–4.5 min:
2–90% B, 4.5–6.5 min: 10% Aþ90% B; UV-detection at
210 nm; column temp. 30 8C, flow rate 0.75 mL/min.
Analytical TLC: Macherey-Nagel ready-to-use plates
AluGram Sil G/UV254. Detection under a UV-lamp at
254 nm, development with molybdatophosphoric acid
solution (5% in EtOH) or 0.5% aq. KMnO4. Column
chromatography: Merck silica gel, grade 60, 230–400
mesh. Optical rotations were measured with a Perkin–
Elmer polarimeter, and CD-spectra were recorded with a
J-810 instrument (JASCO). Elemental analyses were
performed by the Mikroanalytisches Laboratorium des
Instituts fu¨r Organische und Biomolekulare Chemie der
4.1.3. 1-tert-Butyl (S)-N-benzyloxycarbonyl-N-methyl-
aspartate (N-Z-N-MeAspOtBu, (S)-11). To a solution of
diester (S)-10 (102 g, 0.29 mol) in MeOH (0.8 L) kept in
an ice-bath, was added dropwise within 4 h a solution of
LiOHpH2O (12.2 g, 0.29 mol) in water (200 mL). The
mixture was left to warm up to room temperature. Most of
the MeOH was evaporated in vacuo (bath temp. #35 8C),
the residue was diluted with water (0.5 L) and extracted
with ether (3£150 mL). The aqueous layer was acidified
with cold 6 M HCl up to pH 3–4, and extracted with ether
(3£200 mL). The organic layers were washed with brine,
dried and evaporated to give 75.9 g (78%) of (S)-11 as an
1
oil. H NMR (CDCl3, the signals of the major rotamer are
Georg-August Universitat Gottingen. Solvents were purified
according to standard procedures. Organic solutions were
dried over MgSO4. All reactions were carried out with
magnetic stirring, unless otherwise stated.
marked with p) d 1.34/1.40p (s, 9H), 2.66–2.89 (m, 1H,
CHH), 2.96 (s, 3H, NMe), 3.00–3.17 (m, 1H, CHH), 4.71
(q, 1H, J¼6.6 Hz, CH), 5.04–5.21 (m, 2H, OCH2), 7.31–
7.39 (m, 5H), ,9.9 (br. s, 1H, COOH); 13C NMR
(62.9 MHz, CDCl3, the signals of the major rotamer are
marked with p) d 27.7 (Me), 33.7 (NMe), 34.4p/34.9 (CH2),
57.5/58.2p (CHN), 67.4p/67.6 (OCH2), 82.3p/82.4 (C–O),
127.7 (CH), 127.9 (CH), 128.5 (CH), 130.4/136.4p (C),
155.9/156.3p (NCvO), 168.7/168.8p (CvO), 176.2/176.5p
(CvO).
¨
¨
4.1.1. L-Aspartic acid 3-methyl ester [(S)-7] and N-benzyl-
oxycarbonyl-L-asparaginic acid 3-methyl ester [(S)-8].
The title compounds were prepared as described in the
literature.8 Unlike the authors of the original publication, we
were not able to obtain a crystalline sample of (S)-8
(reported mp 98 8C). Diester (S)-9 was synthesized as
reported,9 [a]D20¼22.3 (c 1.75, EtOAc); lit.:9 [a]2D0¼21.7 (c
1.75, EtOAc). In the latter publication compound (S)-8 is
also described to be a colorless oil. The isothiuronium salt 5
was synthesized according to the known procedure.18 Bis-Z-
b-(S)-lysine 20b was purchased from EMKA Chemical
Enterprise, Ltd.
4.1.4. (S)-3-tert-Butoxycarbonyl-3-[(N-benzyloxycarbo-
nyl-N-methyl)amino]propanoylsuccinimide (N-Z-N-
MeAsp(OSu)OtBu, (S)-12). To a solution of the ester
(S)-11 (8.94 g, 26.5 mmol) in EtOAc (50 mL) was added at
þ5 8C N-hydroxysuccinimide (3.35 g, 29.2 mmol) and then
dropwise with ice-cooling a solution of N,N0-dicyclo-
hexylcarbodiimide (6.00 g, 29.1 mmol) in 50 mL of diox-
ane. The reaction mixture was kept overnight at þ5 8C.
N,N-Dicyclohexylurea was removed by filtration, washed
with dioxane (15 mL), and the filtrate was concentrated
in vacuo to give 11.2 g (97%) of (S)-12 as a solid. An
analytical sample was recrystallized from dioxane–ether,
mp 120–122 8C. Found C 57.81, H 6.19; calcd for
4.1.2. 1-tert-Butyl methyl (S)-N-benzyloxycarbonyl-N-
methylaspartate (N-Z-N-MeAsp(OMe)OtBu, (S)-10). To
a solution of diester (S)-9 (21.4 g, 63.4 mmol) and MeI
(28.0 mL, 63.6 g, 448 mmol) in anhydrous DMF (100 mL),
was added 15.0 g (64.7 mmol) of Ag2O, and the black
suspension was vigorously stirred at room temperature for
7 h. A small amount of the reaction mixture was worked-up
(see below) and analyzed by means of TLC (eluent EtOAc/
hexane, 1:4) or, better, NMR spectroscopy to determine,
whether the reaction was complete; compound (S)-9:
Rf¼0.53; compound (S)-10: Rf¼0.57. Dichloromethane
(700 mL) was added to the reaction mixture, and it was
filtered through Celitew. The filter cake was washed with
dichloromethane (2£100 mL), and the combined organic
solutions were washed with saturated aq. Na2S2O3 or 10%
aq. NaCN solution (2£100 mL) and water (8£100 mL).
After drying, they were concentrated in vacuo, and the oily
residue was kept at 0.01 mm Hg to remove traces of DMF,
and compound (S)-10 was isolated as a yellowish oil
(21.8 g, 98%) and was used in the next step without further
purification. 1H NMR (CDCl3, the signals of the major
rotamer are marked with p ) d 1.35/1.40p (s, 9H), 2.63–2.80
(m, 1H, CHH), 2.93/2.94p (s, 3H, NMe), 2.91–3.10 (m, 1H,
CHH), 3.64/3.66p (s, 3H, OMe), 4.75 (mc, 1H, CH), 5.03–
1
C21H26N2O8 (434.43) C 58.06, H 6.03; H NMR (CDCl3,
the signals of the major isomer are marked with p) d 1.37/
1.41p (s, 9H, Me), 2.81 (br. s, 3H, NMe), 3.00 (s, 4H, CH2),
2.91–3.41 (m, 2H, CH2), 4.61–4.75 (m, 1H, CH), 5.02–
5.28 (m, 2H, CH2O), 7.30–7.39 (m, 5H); 13C NMR
(62.9 MHz, CDCl3, the signals of the major isomer are
marked with p) d 25.5 (CH2), 27.7 (Me), 31.7p/32.2 (CH2),
34.2/34.4p (MeN), 57.5/58.3p (CHN), 67.4p/67.6 (CH2O),
82.7p/82.9 (C–O), 127.8 (CH), 127.9 (CH), 128.4 (CH),
136.4 (C), 155.6/156.1p (NCO), 166.2/166.6p (CO), 167.9/
168.0p (CO), 168.7/168.8p (CO).
4.1.5. N 2-Benzyloxycarbonyl-N 2-methyl-L-asparagine
tert-butyl ester (N 2-Z-N 2-Me-L-AsnOtBu, (S)-13). To a
suspension of compound (S)-12 (11.05 g, 25.4 mmol) in
THF (100 mL) was added dropwise at 0 8C 4 mL of 25% aq.
NH3. The reaction mixture was stirred at room temperature
overnight, filtered, and the filtrate was evaporated in vacuo.

V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
7585
The residue was dissolved in ether (200 mL), the solution
was washed with 0.5 M aq. HCl (20 mL), H2O (20 mL), sat.
aq. NaHCO3 (20 mL) and brine (20 mL). After drying and
evaporation of the solvent, 7.15 g (84%) of the title
compound was obtained as an oil which was used in the
next step without further purification. 1H NMR (CDCl3, the
signals of the major rotamer are marked with p) d 1.38 (s,
9H, Me), 2.48–3.04 (m, CH2), 2.96/2.99p (s, NMe, total
intensity 5H), 4.49p (dd, J¼5.8, 8.4 Hz, CH) and 4.68 (t,
J¼7.1 Hz, total intensity 1H, CH), 4.98–5.23 (m, 2H,
CH2O), 5.82/5.93p/6.00/6.22p (br. s, total intensity 2H,
NH2), 7.38 (br. s, 5H); 13C NMR (62.9 MHz, CDCl3, the
signals of the major isomer are marked with p) d 27.8 (Me),
34.7p/35.2 (CH2), 36.0p/36.5 (MeN), 58.3/59.5p (CHN),
67.2p/67.4 (CH2O), 82.0p/82.8 (C–O), 127.7 (CH), 128.0
(CH), 128.5 (CH), 136.4 (C), 156.4 (NCO), 169.4 (CO),
172.7 (CO).
under nitrogen DPPA (1.19 mL, 5.50 mmol). The reaction
mixture was stirred at 80 8C for 17 h, diluted with ethyl
acetate (200 mL), washed with 1% aq. solution of citric acid
(30 mL), sat. aq. NaHCO3 (50 mL), dried and evaporated.
Chromatography on SiO2 (60 g) eluting with an EtOAc/
hexane mixture (1:4) gave 1.00 g (49%) of (S)-16 as a
colorless oil; Rf¼0.75 (CH2Cl2/MeOH, 20:1). 1H NMR
(CDCl3, the signals of the major rotamer are marked with p)
d 1.40 (s, 18H, 2£tBu), 2.92 (s, 3H, NMe), 3.37/3.58p (m,
2H, CH2), 4.42 (m, 1H, CH), 4.79/4.83p (m, 1H, NH), 5.00–
5.23 (m, 2H, CH2O), 7.29–7.35 (m, 5H); 13C NMR
(62.9 MHz, CDCl3, the signals of the major rotamer are
marked with p) d 27.8 (Me), 28.2 (Me), 32.9p/33.2 (NMe),
39.2p/39.5 (CH2), 59.9/60.4p (CH), 67.2p/67.4 (CH2O),
79.3p/79.5 (CO), 82.0p/82.1 (CO), 127.6 (CH), 127.7 (CH),
128.4 (CH), 136.1 (C), 136.5 (C),155.8 (C), 155.9 (C), 156.2
(C), 157.0 (C), 168.7 (CO).
4.1.6. N 2-Benzyloxycarbonyl-N 2-methyl-L-asparagine
(N 2-Z-N 2-Me-L-AsnOH, (S)-14). To a solution of N 2-Z-
N 2-Me-L-AsnOtBu (16.0 g, 47.6 mmol) in CH2Cl2 (20 mL)
was added dropwise at 0 8C TFA (30 mL), and the solution
was stirred at room temperature for 4 h. Volatiles were
evaporated in vacuo, the residue was triturated with
anhydrous ether (200 mL) and kept at 0 8C overnight,
until it solidified completely. After washing with anhydrous
ether (3£100 mL) and drying in vacuo, 11.0 g of a colorless
solid was obtained; [a]D20¼247.0 (c 0.99, MeOH), e.e.¼
77% (HPLC on a chiral stationary phase: poly(N-acyloyl-
L-leucid-d-methylamide) grafted silica gel; column
250£20 mm, isohexane/THF¼3:7, 1 mL/min; (S)-isomer:
7.75 min, (R)-isomer: 9.77 min). The solid was powdered
and stirred in 2-propanol (100 mL) at 50–60 8C for 30 min
with occasional sonification. The suspension was left
overnight at room temperature and filtered through a
sintered glass filter (No. 4). The filter cake was washed
with EtOAc (20 mL), and the clear solution was evaporated
in vacuo. The residue was recrystallized from 2-propanol
and EtOAc mixture, and the solid which formed was
collected by filtration to give 7.71 g (58%) of the title
compound with [a]2D0¼259.2 (c 0.95, MeOH), e.e. $99%,
mp 129 8C (dec.). An analytical sample was recrystallized
once more from 2-propanol. Found C 56.08, H 5.75, N 9.84;
calcd for C13H16N2O5 (280.27): C 55.71, H 5.75, N 9.99;
[a]2D0¼260.9 (c 1.0, MeOH); lit.:11a [a]2D0¼260.8 (c 1.0,
MeOH), mp 134–136 8C. The other spectra of (S)-18 are
identical to those of the racemate5 and to the previously
published ones.11a
4.1.9. Reaction of compound (S)-16 with trifluoroacetic
acid. A solution of (S)-16 (1.00 g, 2.45 mmol) in TFA
(5 mL) prepared at 0 8C, was stirred at room temperature for
1.5 h. Volatiles were evaporated in vacuo at room
temperature, the residue was dissolved in water (10 mL),
and 25% aq. NH3 was added carefully until the pH value
reached 7. The suspension was evaporated to dryness, and
the solid residue was recrystallized from EtOH to give
0.423 g (68%) of (S)-15 with [a]2D0¼238.8 (c 0.98, H2O). In
another run the [a]2D0 was found to be 240.28 (c 0.99, H2O).
A second recrystallization from aq. EtOH afforded 0.280 g
(45%) of (S)-19 with [a]2D0¼243.58 (c 1.00, H2O).
4.1.10. Methyl (2)-3-amino-2-(N-benzyloxycarbonyl-N-
methyl)aminopropionate hydrochloride [(S)-4]. The title
compound was prepared from 0.475 g (1.88 mmol) of
(S)-15 in anhydrous MeOH (5.6 mL) in the presence of
0.49 mL SOCl2 as described before.5 The reaction mixture
was stirred at room temperature for 1.5 h and kept at þ5 8C
overnight. Crystallization from MeOH (5 mL) and ether
(20 mL) gave 0.490 g of (S)-4 (86%) with mp 171–172 8C
(the racemate has mp 175–176 8C5) and [a]D20¼251.8 (c
0.92, MeOH); e.e.<55%. HPLC on Gromchiral AD;
heptane/EtOH/TFA, 82:17:0.2, 0.2 mL/min; column
250£2 mm, Rt¼11.82 min (S-isomer) and 14.87 min
(R-isomer). The spectral data were found to be identical to
these of the racemic compound.5
4.1.11. (S)-2-(N-Benzyloxycarbonyl-N-methyl)amino-3-
(tert-butoxycarbonylamino)propionic acid [(S)-17, N 2-
Z-N 2-Me-N 3-Boc-L-A2pr].19 A suspension of (S)-15
(0.660 g, 2.62 mmol) in 10 mL of water was cooled to
0 8C, and 2.0 mL of 1.3 M aq. NaOH was added dropwise
followed by tBuOH (5 mL) and Boc2O (0.661 g,
3.02 mmol). The cold bath was removed, and stirring was
continued at room temperature. Gas evolution started after
about 15 min, and the pH value gradually decreased. It was
kept at about 9 by careful addition of sat. aq. Na2CO3
solution. The reaction was complete within 3.5 h. The
reaction mixture was diluted with H2O (20 mL), extracted
with hexane (2£20 mL), and acidified at þ5 8C with 5% aq.
KHSO4 (pH <2). Then it was extracted with ether
(4£20 mL); the combined ether layers were washed with
brine, dried, and evaporated to yield 0.90 g (97%) of (S)-17
as a glass-like foam. The racemate (RS)-17 was obtained
4.1.7. (S)-3-Amino-2-(N-benzyloxycarbonyl-N-methyl)-
aminopropionic acid (N 2-Z-N 2-Me-L-A2pr, (S)-15).
The title compound was prepared in 75% yield from
(S)-14 as described for the racemate (RS)-14;5 [a]D20¼245.2
(c 1.05, water), mp 204 8C (dec., aq. EtOH). Found C 57.43,
H 6.40, N 10.95; calcd for C12H16N2O4 (252.26) C 57.13, H
6.39, N 11.10. The spectra of (S)-14 were identical to those
of the racemate.5
4.1.8. tert-Butyl(S)-2-(N-benzyloxycarbonyl-N-methyl)-
amino-3-(tert-butoxycarbonyl)amino-propionate (N 2-Z-
N 2-Me-N 3-Boc-L-A2prOtBu, (S)-16). To a solution of
compound 11 (1.69 g, 5.01 mmol) and Et3N (0.51 g,
5.0 mmol) in 20 mL of anhydrous tBuOH was added

7586
V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
analogously from (RS)-155 and had mp 133–134 8C
(EtOAc/PE). IR: n¼3384, 3328, 2940, 2863, 1690, 1634,
1540, 1518, 1450, 1382, 1363, 1319, 1271, 1169 cm21; 1H
NMR (CDCl3, signals of the major isomer are marked with
(0.1 mm Hg) at room temperature, the residue was triturated
with anhydrous ether (2£15 mL, with decantation) and
dried in vacuo (0.01 Torr). Then it was suspended in
CH2Cl2, and Et3N (399 mg, 550 mL, 3.95 mmol) was added
at 0 8C. Slow gas evolution was observed, and a precipitate
started to form. A sheet of wet indicator paper introduced
into the head space of the reaction flask indicated that the pH
was about 11. A few small drops of glacial AcOH were
added, until the pH value (measured in the same way)
reached 8–9. Stirring at room temperature was continued
overnight. The solvent was evaporated in vacuo, 10 mL of
water was added, and the suspension was sonificated in an
ultra-sound bath (30 s). The precipitate was removed by
filtration, washed with H2O (10 mL), cold MeOH
(2£5 mL), ether (5 mL), and dried in vacuo to give
447 mg (55%) of (S)-6, [a]2D0¼2175 (c 0.525, DMF),
e.e.¼92%; column: Chiralpak AS (250£4.6 mm), eluent:
heptane/ethanol (1:1), 1 mL/min; (R)-isomer: 6.82 min,
(S)-isomer: 8.78 min. The spectral data were identical to
those of the racemate.5 In another run a product with
[a]2D0¼2157 (c 0.545, DMF) was obtained in 65% yield. A
sample of this product (464 mg) was suspended in 15 mL
of DMAA with sonification and warming at 30 8C. The
insoluble fraction was removed by filtration (35 mg after
washing with ether and drying), and the solution was used
in the next step (see below). (S)-1 was obtained with
practically the same e.e. (86%). This simple way to increase
the e.e. is based on the fact that the solubility of the racemate
(RS)-6 is much lower than that of (S)-6.
P
P
p) d 1.41p s/1.44 br. s ( 9H, tBu), 2.91/2.94p (s,
3H,
P
MeN), 3.40 br. m/3.62p m ( 2H, CH2N), 4.48p t/4.60 br. m
P
P
(
1H, CHN), 4.89/5.02p (br. t, 0.5H, NHCO), 5.12 br.
P
P
s/5.16p s ( 2H, CH2O), 6.16p/6.23 (br. s, 0.5H, NHCO),
7.23–7.39 m (5H), 9.63 br. s (1H, COOH); 13C NMR
(62.9 MHz, CDCl3, signals of the major isomer are marked
with p) d 28.3 (Me in tBu), 33.1/33.4p (MeN), 39.2/39.5p
(CH2N), 59.6/60.1p (CHN), 61.7p/67.8 (CH2O), 79.8
(C–O), 127.7, 127.9, 128.1, 128.5 (CH), 136.0/136.1p (C),
156.0, 156.9 (CONMe), 172.8 (COOH); ESI-MS (positive
mode), m/z (rel. int., %) 727 (100) [2MþNaþ], 375 (24)
[MþNaþ]. (S)-17pDCHA salt was prepared from (S)-17 and
dicyclohexylamine in EtOAc and precipitated at þ5 8C by
addition of hexane; mp 120–121 8C; found C 64.98, H 8.76,
N 7.61; calcd for C29H47N3O6 (533.71) C 65.26, H 8.88, N
7.87; [a]2D7¼23.9 (c 1.24, CHCl3).
4.1.12. (S)-5-(N-Benzyloxycarbonyl-N-methyl)amino-
3,4,5,6-tetrahydro-2-ureidopyrimidin-4-one
[(S)-6].
Coupling product 18. To a suspension of (S)-17 (0.900 g,
2.55 mmol) in CH2Cl2 (15 mL), was added N-hydroxyben-
zotriazole hydrate (0.430 g, 2.81 mmol), and the mixture
was cooled to 0 8C. EDC (0.434 g, 490 mL, 3.16 mmol) was
added dropwise, and the precipitate disappeared. Stirring
was continued for 20 min at 0 8C, DIEA (645 mg, 825 mL,
4.99 mmol) was added dropwise followed by the compound
5 (1.30 g, 4.98 mmol). Stirring at 0 8C was continued for
15 min, and the clear solution was kept at þ5 8C overnight.
Then, CH2Cl2 (200 mL) was added, and the solution was
washed with water (20 mL), 5% aq. KHSO4 (20 mL), water
(20 mL), 1% aq. NaHCO3 (20 mL), and brine (20 mL).
After drying and evaporation of the solvent, a semi-
crystalline mass of the crude coupling product 18 was
obtained (1.19 g, quantitative yield); Rf,0.5 (CH2Cl2/
MeOH, 20:1, UV-active spot). [a]2D0¼217.0 (c 1.14,
CHCl3); IR: n¼3353, 2978, 1700, 1570, 1456, 1400,
4.1.13. (S)-3,4,5,6-Tetrahydro-5-methylamino-2-ureido-
pyrimidin-4-one [(S)-1]. To a solution of (S)-6 (360 mg,
1.62 mmol, e.e.¼92%) in 15 mL of anhydrous N,N-
dimethylacetamide (DMAA) was added 200 mg of 10%
Pd/C. (Merck, oxidized form). The reaction mixture was
flushed with N2, then with hydrogen, and stirred at room
temperature. A rapid consumption of H2 (from a balloon
attached to the reaction flask) was observed. After 1.5 h, the
reaction was complete (detection by TLC (CH2Cl2/MeOH,
1:1); Rf of (S)-6,0.9, Rf of (S)-1,0.5). The suspension was
filtered through Celitew, the filter cake was washed with
DMAA (3£3 mL), and the filtrate was evaporated in vacuo
(bath temp. 30–40 8C, 0.01 mm Hg). The residue was
triturated with anhydrous ether (10 mL), the solid residue
was filtered off, washed with anhydrous ether (10 mL), and
dried to give 200 mg (96%) of (S)-1, [a]2D0¼2247 (c 0.635,
DMF), e.e.¼87%; column: Chiral AS (250£2 mm); hep-
tane/EtOH/Et2NH, 65:34.8:0.2; 0.25 mL/min; (S)-isomer:
6.63 min, (R)-isomer: 8.78 min.
1
1366, 1230, 1165, 1008 cm21; H NMR (CDCl3, signals
of the major isomer are marked with p) d 1.39 (s, 9H, tBu),
3H, MeN), 3.01 (s, 3H, MeS), 3.43/3.65p
P
2.28/2.32p (s,
(m, 2H, CH2N), 4.40 (br. t, 1H, CHN or NH), 4.57 (br. t, 1H,
NH or CHN), 4.98 (m, 2H, NH2), 5.08/5.20p (m, 2H,
CH2O), 5.42p/5.58 (br. s, 1H, NH), 7.32–7.39 (m, 5H); 13
C
NMR (75.5 MHz, CDCl3, the signals of the major isomer
are marked with p) d 14.3 (SMe), 28.26p/28.32 (Me), 33.2p/
34.1 (br. MeN), 38.8p/39.2p/39.7 (br. NCH2), 61.4p/63.3 (br.
CHN), 67.2/67.9p/68.0p (br. CH2O), 79.7/79.8 (br. C–O),
127.84, 127.9, 128.0, 128.2, 128.5p, 128.8 (CH), 135.8/
136.2p (C), 155.5/155.7p, 156.8 (NCO), 163.3, 168.2/168.6/
169.2p (CO). ESI-MS (positive mode), m/z (rel. int., %) 957
(100) [2MþNaþ], 490 (29) [MþNaþ].
4.1.14. (S)-N b,N 1,N v-Tris(benzyloxycarbonyl)-b-homo-
arginine [(S)-20a].15 To a solution of the diazoketone
(S)-19 (22.2 g, 37.0 mmol) in 250 mL of dioxane was added
18 mL of water. The solution was cooled in an ice bath and
irradiated for 2 h with a 300 W daylight lamp (the internal
temp. was about 30 8C). The solvents were removed in
vacuo, and the semi-solid residue was crystallized from
acetone/ether mixture to give a first crop (5.6 g) as a
colorless solid. The mother liquors were concentrated, and
the residue was purified by passing it through a short silica
gel pad (100 g) eluting with a gradient of CH2Cl2/EtOAc
(1:1) to CH2Cl2/EtOAc/MeOH (8:8:1). After evaporation,
the main fraction was crystallized from acetone/ether
Deprotection and cyclization of the coupling product 18 to
(S)-6
To a suspension of the semi-solid 18 (1.18 g, 2.55 mmol) in
anisole (5 mL), was added at room temperature a mixture of
CH2Cl2 (10 mL) and TFA (7.5 mL), and the solution was
stirred for 2.5 h. When the deprotection reaction was
complete (TLC), the solvents were evaporated in vacuo

V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
7587
yielding another 4.6 g of the title compound, total yield
10.2 g (47%). An analytical sample was recrystallized once
more from EtOH. Found C 62.28, H 6.29; calcd for
C31H34N4Op81/2H2O (599.62) C 62.09, H 5.88; [a]2D0¼þ0.8
(c 1.21, CHCl3); lit.20 [a]D20¼22.5 (c 1.1, CH2Cl2/MeOH);
30.4 (CH2), 1H NMR (CDCl3) d 1.42–1.65 (m, 4H, H-4 and
H-5), 2.55 (m, 2H, H-2), 3.88 (m, 3H, H-3 and H-6), 5.05 s,
5.17s and 5.25 s (6H, PhCH2O), 5.58 (d, J¼5 Hz, 1H, NH),
7.34–7.39 (m, 15H), 9.27 and 9.41 (br. s, 2H, NH); 13C
NMR (75.5 MHz, CDCl3) d 25.3 (CH2), 30.4 (CH2), 38.7
(CH2), 44.3 (CH2), 47.7 (CH), 66.6, 67.0 68.9 (CH2O),
127.8, 127.88, 127.94, 128.02, 128.11, 128.35, 128.4,
128.77, 128.82 (CH), 134.6, 136.4, 136.8 (C), 155.8, 156.0
(NCO), 160.5, 163.7 (CO), 175.9 (br. COOH); ESI-MS
(positive mode), m/z (rel. int., %) 1203 (64) [2MþNaþ], 613
(39) [MþNaþ], 591 (39) [MþHþ]; (negative mode), 1201
(34) [2M22HþNaþ], 589 (100) [M2Hþ].
[D6]DMSO, signals of the major rotamer are marked with p)
d 25.2 (CH2), 29.5/33.0 (NMe), 31.3 (CH2), 44.6 (CH2),
other signals of CH2-groups are masked by the signals of
[D6]DMSO, 47.8p/47.9 (CH), 52.1p/54.2 (CH), 65.1, 66.2,
68.2 (CH2O), 127.53, 127.66, 127.71, 127.79, 127.82,
127.88, 128.27, 128.32, 128.5 (CH), 135.3, 137.1, 137.2
(C), 155.0p/155.5, 155.6 (NCO), 159.7, 162.9, 170.9p/171.1
(CO); ESI-MS (positive mode), m/z (rel. int., %) 780 (100)
[MþNaþ], 758 (84) [MþHþ].
4.1.16. (30S,5S)-3,4,5,6-Tetrahydro-5-[N-methyl-N-(b-
homoarginyl)amino]-2-ureidopyrimidin-4-one dihydro-
chloride (TAN-1057Ap2HCl). A suspension of the coup-
ling product (S,S)-21a (205 mg, 0.271 mmol) and PdCl2
(48.5 mg, 0.274 mmol) in anhydrous MeOH (7 mL) was
flushed with nitrogen and then with hydrogen, and was
vigorously stirred under hydrogen (a balloon with H2
was attached). Gradually, the light-brown suspension turned
gray, and then the starting material dissolved. After stirring
for about 3–4 h at room temp. (28 8C), the reaction was
complete. After flushing with N2, the reaction mixture was
filtered through Celitew to remove the Pd-black. The filter-
cake was washed with MeOH (3£3 mL), and the filtrate
was evaporated in vacuo. The colorless solid residue was
triturated with anhydrous ether (10 mL), collected on a filter
under ether, quickly washed with ether once more (10 mL),
and dried in vacuo (0.01 mm Hg) overnight to give 122 mg
of the title compound (106% yield, 6% w/w of MeOH) as an
amorphous white powder. [a]2D2¼222.7 (c 0.6, water);16
lit.:2 [a]2D2¼239.1 (c 0.53, water); CD spectrum u (l, nm,
water)¼þ12,500 (215), 212,600 (233), 27500 (253),
214,200 (269); lit.2 u (l, nm, water)¼þ13,300 (215),
213,500 (231), 213,500 (267); IR: n¼3340, 3156, 1747,
1653, 1615, 1420, 1378, 1277, 1218, 1135, 1013 cm21; 1H
NMR ([D4]MeOH, 600 MHz) d 1.71, 1.80 (2 m, 4H, H-40
4.1.15. (30S,5S)-5-[N-Methyl-N-[N b,N 1,N v-tris(benzyl-
oxycarbonyl)-b-homoarginyl]amino]-5,6-dihydro-2-
ureidopyrimidin-4(1H)-one [(S,S)-21a]. To a suspension
of the acid (S)-20a (495 mg, 0.84 mmol) and (S)-1 (155 mg,
0.84 mmol) in anhydrous DMAA (4 mL) was added HATU
(646 mg, 1.70 mmol) in one portion at room temperature
followed by DIEA (0.29 mL, 227 mg, 1.75 mmol) which
was added dropwise without external cooling. A slightly
exothermic reaction was observed, and the solid heterocycle
(S)-1 gradually dissolved. After 4.5 h at room temperature,
the solvent was evaporated in vacuo (0.01 mm Hg, bath
temp. 30–40 8C), and the residue was dissolved in
dichloromethane (50 mL). The solution was washed with
water, 5% aq. KHSO4, water, sat aq. NaHCO3, brine (10 mL
each), and dried. The solvent was evaporated, and the
semi-solid residue was triturated with 10 mL of a mixture of
EtOAc and ether (1:1). After keeping this suspension
overnight at þ5 8C, the solid crude coupling product
(S,S)-21a was collected on a filter and washed with ether
(10 mL). After drying, the crude product (0.53 g, 84% yield)
was recrystallized from dichloromethane (45 mL). The
fraction, which was insoluble in boiling dichloromethane,
was removed by filtration (0.10 g after drying), the filtrate
was diluted with EtOAc (1:1), and kept at þ5 8C overnight.
The colorless solid of (S,S)-21a (0.33 g, 52%) was collected,
washed on a filter with ether (10 mL) and dried in vacuo
(0.01 Torr). Found C 58.16, H 5.81, N 16.28; calcd for
C37H43N9Op91/4H2O (761.8) C 58.30, H 5.75, N 16.54;
[a]2D0¼251 (c 0.26, DMF); HPLC-MS: Rt¼4.52 min (peak
area 100%); IR: n¼3388, 3266, 3133, 2946, 1718, 1610,
1575, 1507, 1456, 1377, 1257, 1097, 1029 cm21; 1H NMR
([D6]DMSO, 600 MHz, signals of the major rotamer are
marked with p ) d 1.32, 1.42, 1.51 and 1.58 (4 m, 4H, H-40
and H-50), 2.22 (dd, J¼16.2, 6.6 Hz, A-part of an
AB-system, H-20), 2.40p (dd, J¼15.3, 6.6 Hz, A-part of an
AB-system, H-20), 2.48p (dd, J¼[masked by the solvent] and
6.1 Hz, B-part of an AB-system, H-200), 2.56 (dd, J¼16.2,
6.6 Hz, B-part of an AB-system, H-2 ), 2.66/2.85p (s, 3H,
NMe), 3.35p (dd, J¼11, 8 Hz, H-6), 3.52p (t, J¼13.1 Hz,
H-6), 3.56–3.63 [0m, 2H (together with 3.35 and 3.52), H-6],
3.83 (m, 3H, H-3 and H-60), 4.73 (m, H-5), 4.95p [m, 3 H
(together with 4.73), H-5 and CH2O], 5.05 (s, 2H, CH2O),
5.22 (s, 2H, CH2O), <6. 8 (br. s, 1H, NH), 6.98/7.08p (d, 1H,
J¼5.5 Hz, ZNH–C-30), 7.23–7.40 (m, 15H), 9.11 (br. s, 2H,
NH) and <9.6 (br. s, 2H, NH); 13C NMR (75.5 MHz,
and H-50), 2.79 (dd, J¼17.4, 9 Hz, 1H, H-20), 2.97 (dd,
P
J¼17.4, 3.8 Hz, 1H, H-20), 2.88/3.12 (2 s, 3H, NMe), 3.25
(t, J¼6.8 Hz, 2H, H-60), 3.61 (m, 1H, H-30), 3.81 (dd,
J¼12.9, 7.9 Hz, 1H, H-6), 3.95 (t, J¼12.9 Hz, 1H, H-6),
5.19 (dd, J¼8.7, 12.7 Hz, 1H, H-5); 103C NMR (75.5 MHz
[D4]MeOH,) d 25.8 (C-50), 30.8 (C-4 ), 35.1 (NMe), 36.1
(C-20), 40.2 (C-6), 42.0 (C-60), 49.6 (C-30), 54.9 (C-5),
156.3, 156.7, 156.8 (NCO), 170.5, 173.0 (CO); ESI-MS
(positive mode), m/z (rel. int., %) 356 (100) [MþHþ].
4.1.17. (30S,5S)-5-[N-Methyl-N-[30,60-bis(benzyloxycar-
bonylamino)hexanoyl]amino]-5,6-dihydro-2-ureidopyri-
midin-4(1H)-one [(S,S)-21b]. The title compound was
obtained from the acid (S)-20b (317 mg, 0.77 mmol) and
(S)-1 (142 mg, 0.77 mmol; e.e.¼87%) in anhydrous DMF
(4 mL) with HATU (585 mg, 1.54 mmol) and DIEA
(199 mg, 1.54 mmol) as described for the coupling product
(S,S)-21a. After recrystallization from a mixture of CH2Cl2
and EtOAc, 370 mg (83%) of the title compound with
de¼90% was obtained; [a]2D0¼293 (c 1.03, DMF). In
another experiment, (S)-1 (1.20 g, 6.48 mmol) of lower
optical purity ([a]2D0¼2213 (c 0.48, DMF), e.e.<75%), acid
(S)-20b (2.69 g, 6.49 mmol), HATU (4.90 g, 12.9 mmol)
and DIEA (2.22 mL, 1.74 g, 13.5 mmol) in 15 mL of
DMAA gave 2.74 g (73%) of the crude (S,S)-21b, which
was twice recrystallized from a mixture of CH2Cl2 and
EtOAc to afford 2.09 g (55%) of the title compound with
de¼85%; (30S,5R)-isomer: Rt¼14.38 min, (30S,5S)-isomer:

7588
V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
Rt¼20.77 min (Chiracel OD-H, column 250£2 mm, heptane/
isopropanol, 1:1, 0.2 mL/min); (30S,5R)-isomer: Rt¼
15.99 min, (30S,5S)-isomer: Rt¼26.52 min (Chiralpak
OD-H, column 250£4.6 mm, methanol, 0.5 mL/min).
Found C 57.63, H 6.14, N 16.49; calcd for C28H35N7O7
(581.6) C 57.82, H 6.06, N 16.86; IR: n¼3335, 3148, 2940,
1716, 1698, 1643, 1610, 1576, 1528, 1455, 1404, 1347, 1271,
Dr. B. Knieriem for his careful reading of the final
manuscript.
References and notes
1
1141, 1069, 1026 cm21; H NMR ([D6]DMSO, 600 MHz,
1. Katayama, N.; Fukusumi, S.; Funabashi, Y.; Iwahi, T.; Ono,
H. J. Antibiot. 1993, 46, 606–613.
signals of the major rotamer ar0e marked with p ) d 1.37 and
1.45 (2 m, 4H, H-40 and H-5 ), 2.24 (dd, J¼14.5, 6.4 Hz,
A-part of an AB-system, H-20), 2.43p (dd, J¼15.6, 6.4 Hz,
A-part of an AB-system, H-20), 2.52p (dd, J¼15, 6 Hz, B-part
of an AB-system, H-20), 2.58 (dd, J¼15.7, 6 Hz, B-part of an
AB-system, H-20), 2.67/2.90p (s/br. sp, 3H, NMe), 2.97 (br. s,
2H, H-60), 3.40p (dd, J¼12.3, 7.8 Hz, H-6), 3.56p (t, J¼
12.9 Hz, H-6), 3.59–3.65 [m, 2H (together with 3.40 and
3.56), H-6], 3.80/3.84p (br. s, 1H, H-30), 4.78 (m, H-5), 5.05
[br. s, 5H (together with 4.78), 2£CH2O], <6.8 (br. s, 1H,
NH), 6.99/7.09p (d, 1H, J¼5.5 Hz, ZNH-C-30), 7.23–7.40
(m, 10H), <9.8 (br. s, 2H, NH). 13C NMR (75.5 MHz,
[D6]DMSO, signals of the major rotamer are marked with p)
d 26.3 (C-50), 29.5/33.1p (NMe), 31.6 (C-40), 47.8p/48.2
(C-30), 52.1p/54.1 (C-5), 65.2 (CH2O), 127.6, 127.7, 128.3
(CH), 137.3 (C), 155.7, 156.1 158.5 (br.) (NCO), 171.0p/
171.2, 173.1p/174.0 (CO). ESI-MS (positive mode), m/z (rel.
int., %) 1185 (100) [2MþNaþ], 604 (92) [MþNaþ]; (positive
mode) 580 (100) [M2Hþ].
2. Funabashi, Y.; Tsubotani, S.; Koyama, K.; Katayama, N.;
Harada, S. Tetrahedron 1993, 49, 13–28.
3. Williams, R. M.; Yuan, C. J. Am. Chem. Soc. 1997, 119,
11777–11784.
4. For the synthesis of analogues of TAN-1057A,B see:
(a) Williams, R. M.; Yuan, C.; Lee, V. J.; Chamberland, S.
J. Antibiot. 1998, 51, 189–201. (b) Brands, M.; Es-Sayed, M.;
¨
¨
Habich, D.; Raddatz, S.; Kruger, J.; Endermann, R.;
Gahlmann, R.; Kroll, H.- P.; Geschke, F.- U.; de Meijere,
A.; Belov, V. N.; Sokolov, V. V.; Kozhushkov, S. I.; Kordes,
M. TAN-1057 Derivatives. WO 00/12484.. (c) Brands, M.;
Endermann, R.; Gahlmann, R.; Kru¨ger, J.; Raddatz, S.;
Stoltefub, J.; Belov, V. N.; Nizamov, S.; Sokolov, V. V.; de
Meijere, A. J. Med. Chem. 2002, 45, 4246–4253. (d) Brands,
M.; Endermann, R.; Gahlmann, R.; Kru¨ger, J.; Raddatz, S.
Bioorg. Med. Chem. Lett. 2003, 13, 241–246. (e) Brands, M.;
Grande, Y. G.; Endermann, R.; Gahlmann, R.; Kru¨ger, J.;
Raddatz, S. Bioorg. Med. Chem. Lett. 2003, 13, 2641–2645.
5. For the synthesis of the racemic heterocycle (RS)-1 and TAN-
1057A/B from the racemic precursors (RS)-14, (RS)-15 and
(RS)-4, see: Sokolov, V. V.; Kozhushkov, S. I.; Nikolskaya, S.;
Belov, V. N.; Es-Sayed, M.; de Meijere, A. Eur. J. Org. Chem.
1998, 777–783.
4.1.18. (30S,5S)-3,4,5,6-Tetrahydro-5-[N-methyl-N-(3,6-
diaminohexanoyl)amino]-2-ureidopyrimidin-4-one
dihydrochloride [(S,S)-22b]. The coupling product (S,S)-
21b (268 mg, 0.461 mmol, de¼90%) with added PdCl2
(81.7 mmol, 0.461 mmol) in anhydrous MeOH (15 mL) was
hydrogenated and worked-up as described above for the
preparation of TAN-1057Ap2HCl to give 193 mg (108%
yield, 8% w/w of MeOH) of the title compound as an
amorphous powder. HRMS m/z (ESI) found 314.1913, calcd
for C12H24N7O3 [MþHþ] 314.1941; found 627.3810, calcd
for C24H47N14O6 [2MþHþ] 627.3810; [a]2D2¼þ0.6 (c 1.0,
water); CD spectrum u (l, nm, water)¼þ13,800 (215),
211,700 (235), 26900 (254), 212,000 (266); IR: n¼3343,
3127, 3010 sh, 2893, 1778, 1747, 1683, 1623, 1576, 1490,
6. The yield of (RS)-1 has been substantially improved (70% vs.
initially reported 35%), by using MeCN as a solvent instead of
iPrOH (8 mL per 1 mmol of (RS)-4 and 5), decreasing the
reaction temperature from 90 to 55 8C, and increasing the
reaction time up to 48 h. An amount of base (AcONa) was
the same as described previously.5 After 2 days, the solvent
was removed in vacuo, and the solid residue was shaken with
3% aq. NaHCO3 (1 mL per 1 mmol of the starting materials).
Further isolation procedure was identical to the already
described one.5
1380, 1284, 1212, 1123, 1085, 1012 cm21 1H NMR
;
7. (a) Waki, M.; Kitajima, Y.; Izumiya, N. Synthesis 1981,
266–268. (b) Otsuka, M.; Kittaka, A.; Iimori, T.; Yamashita,
H.; Kobayashi, S.; Ohno, M. Chem. Pharm. Bull. 1985, 33,
509–514.
([D4]MeOH, 600 MHz) d 1.81 (br. s, 4H, H-40 and H-50),
2.81 (dd, J¼17.6, 8.5 Hz, 1H, H-20), 2.98 (dd, J¼17.1,
P
P
4.1 Hz, H-20), 2.99 (br. s, 0 3H, H-60), 2.89/3.16 (2 s, 3H,
NMe), 3.62 (br. s, 1H, H-3 ), 3.89 (dd, J¼13.5, 7.9 Hz, 1H,
H-6), 4.02 (t, J¼12.9 Hz, 1H, H-6), 5.17 (dd, J¼7, 012 Hz,
H-5); 13C NMR ([D4]MeOH, 75.5 MHz) d 24.5 (C-5 ), 30.6
(C-40), 35.6 (NMe), 36.0 (C-20), 40.2 (C-6/C-60), 40.0 (C-6/
C-60), 49.4 (C-30), 55.4 (C-5), 154.1, 155.3 (NCO), 166.6,
172.8 (CO). ESI-MS (positive mode), m/z (rel. int., %) 627
(18) [2MþHþ], 314 (100) [MþHþ]; (negative mode) 348
(50) [MþCl2], 312 (100) [M2Hþ].
8. Schwarz, H.; Bumpus, F. M.; Page, I. H. J. Am. Chem. Soc.
1957, 79, 5697–5703.
9. Valerio, R. M.; Alewood, P. F.; Johns, R. B. Synthesis 1988,
786–789.
10. Olsen, R. K. J. Org. Chem. 1970, 35, 1912–1915.
11. (a) Aurelio, L.; Box, J. S.; Brownlee, R. T. C.; Hughes, A. B.;
Sleebs, M. M. J. Org. Chem. 2003, 68, 2652–2667. 5-Oxazo-
lidinones from the reaction of hexafluoroacetone and amino
acids (simultaneous protection of amino and COOH groups)
have already been widely used for the enantioselective
synthesis of N-methyl amino acids: (b) Burger, K.; Spengler,
J. Eur. J. Org. Chem. 2000, 199–204.
Acknowledgements
12. The overall yield over three of these four steps11a was 54%
The yield of one step was not specified. The authors11a did not
report any proof of the optical purity of N 2-Z-N 2-Me-L-
AsnOH. Identity of the optical rotation value of our sample
with .99% e.e. (established by HPLC on a chiral phase
This work was supported by the BAYER AG and the Fonds
der Chemischen Industrie. The authors are grateful to
Mrs. E. Pfeil for measuring optical rotation values,
Mr. R. Machinek for recording 600 MHz spectra and

V. N. Belov et al. / Tetrahedron 60 (2004) 7579–7589
7589
column) with the already reported value11a confirms that the
authors of the cited publication also managed to prepare the
enatiomerically pure compound.
R. E.; Sielecki, T. M.; Wityak, J.; Pinto, D. J.; Batt, D. G.;
Frietze, W. E.; Liu, J.; Tobin, A. E.; Orwat, M. J.; Di Meo,
S. V.; Houghton, G. C.; Lalka, G. K.; Mousa, S. A.; Racanelli,
A. L.; Hausner, E. A.; Kapil, R. P.; Rabel, H. R.; Thoolen,
M. J.; Reilly, T. M.; Anderson, P. S.; Wexler, R. R. J. Med.
Chem. 1999, 42, 1178–1192. (b) Larsen, S. D.; Connell, M. A.;
Cudahy, M. M.; Evans, B. R.; May, P. D.; Meglasson, M. D.;
O’Sullivan, T. J.; Schostarez, H. J.; Sih, J. C.; Stevens, F. C.;
Tanis, S. P.; Tegley, C. M.; Tucker, J. A.; Vaillancour, V. A.;
Vidmar, T. J.; Watt, W.; Yu, J. H. J. Med. Chem. 2001, 44,
1217–1230. (c) Xue, C.-B.; Wityak, J.; Sielecki, T. M.; Pinto,
D. J.; Batt, D. G.; Cain, C. A.; Sworin, M.; Rockwell, A. L.;
Roderick, J. J.; Wang, S.; Orwat, M. J.; Frietze, W. E.;
Bostrom, L. L.; Liu, J.; Higley, C. A.; Rankin, F. W.; Tobin,
A. E.; Emmett, G.; Lalka, G. K.; Sze, J. Y.; Di Meo, S. V.; De
Grado, W. F.; Wexler, R. R.; Olson, R. E. J. Med. Chem. 1997,
40, 2064–2084.
13. Goodmann, H.; Boardmann, F. J. Am. Chem. Soc. 1963, 85,
2483–2490.
14. Considerable racemization was observed in another attempt to
synthesize optically pure (S)-8 from the commercially
available Z-L-Asp(OtBu)OH (Nova-Biochem). It was methyl-
ated with MeI/Ag2O in DMF to yield Z-N-Me-L-Asp(OtBu)-
OMe. The tert-butyl ester was cleaved (Et3SiH/TFA in
CH2Cl2), and the v-COOH group in Z-N-Me-L-Asp(OH)OMe
was converted into the amide (1. EDC, HOBt, THF; 2. aq.
NH3). The N 2-Z-N 2-Me-L-AsnOMe thus obtained was
oxidatively degraded [Ph(OCOCF3)2, aq. DMF] to (S)-4.
The overall yield was good, but the enantiomeric excess was
found to be only about 40% ([a]2D0¼242 (c 1.1, MeOH)).
15. An alternative synthesis of the b-amino acid (S)-20a by Wolff-
rearrangement of the diazoketone (S)-19 catalyzed by silver
benzoate was described in the Supporting Information of
Ref. 3.
16. Our sample of TAN-1057Ap2HCl contains ca. 1 mol of MeOH
per 1 mol of the hydrochloride salt. This corresponds to ca. 7%
(w/w) of MeOH. Corrected [a]2D2¼224 (c 0.6, H2O). As
reported in Ref. 5 the residual amount of MeOH could not be
removed even by prolonged drying in high vacuum.
17. For recent data concerning the pharmacology of (S)-N 2-
methyl-2,3-diaminopropionic acid derivatives, see: (a) Olson,
18. (a) Parkanyi, C.; Yuan, H. L.; Cho, N. S.; Jaw, J.-H.;
Woodhaus, T. E.; Aung, T. L. J. Heterocycl. Chem. 1989, 26,
1331–1334. (b) Klaymann, D. L.; Shine, R. J.; Bower, J. D.
J. Org. Chem. 1972, 37, 1532–1537.
19. This compound (with an unknown degree of racemization and
an unknown optical rotation value) was also found to be an oil:
see Supporting Information of Ref. 3.
20. Data for the monohydrate of the title compound from the
Supporting information to Ref. 3. Relative amounts of the
solvents are not given.