Synthesis of novel amino acids, L-bis-tetrahydrofuranylglycines

Article abstract of DOI:10.1016/S0040-4039(02)00448-3

Novel amino acids, enantiomerically pure L-bis-tetrahydrofuranylglycines, were efficiently synthesized from dihydrofuran utilizing radical cyclization and chiral azidation as the key steps. The absolute stereochemistry of both diastereomers has been ascertained by their crystal structures.

Full text of DOI:10.1016/S0040-4039(02)00448-3

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 2931–2934
Synthesis of novel amino acids, -bis-tetrahydrofuranylglycines
L
Ei’ichi Ami,a S. Rajesh,a Jung Wang,a Tooru Kimura,a Yoshio Hayashi,a Yoshiaki Kisoa,* and
Toshimasa Ishidab
aDepartment of Medicinal Chemistry, Center for Frontier Research in Medicinal Science, Kyoto Pharmaceutical University,
Yamashina-ku, Kyoto 607-8412, Japan
bDepartment of Physical Chemistry, Osaka University of Pharmaceutical Sciences, Takatsuki, Osaka 569-1604, Japan
Received 24 January 2002; accepted 1 March 2002
Abstract—Novel amino acids, enantiomerically pure L-bis-tetrahydrofuranylglycines, were efficiently synthesized from dihydro-
furan utilizing radical cyclization and chiral azidation as the key steps. The absolute stereochemistry of both diastereomers has
been ascertained by their crystal structures. © 2002 Elsevier Science Ltd. All rights reserved.
Human immunodeficiency virus type
1
(HIV-1)
Our experience in this area7 suggested that
L-bis-tetra-
hydrofuranylglycine should fit in the S2 sub-site equal
protease, a virally encoded aspartic protease essential
for the production of infectious virions, remains an
important target of anti-AIDS (acquired immunodefi-
ciency syndrome) chemotherapeutic agents. Much
effort has been made to find potent, orally bioavailable
and peptidomimic inhibitors of this enzyme.1 In con-
nection with our work on HIV-1 protease inhibitors,2
we required a novel amino acid surrogate for aspargine
at the P2 site. Previously, the S2 sub-site of the protease
has been shown to favorably receive various types of
ligands such as tetrahydrofuranylcarbamates,3 bis-(tetra-
hydrofuranyl)carbamate,4 cyclosulfolanes,5 cyclic sul-
fone-3-carboxamides,5 and tetrahydrofuranylglycine.6
Particularly, tetrahydrofuranylglycine 1 and bis-(tetra-
hydrofuranyl)carbamate 2 ligands (Fig. 1) led to signifi-
cant improvement in the inhibitory potency by acting
as hydrogen acceptors from Asp-29 and Asp-30
residues of the HIV-1 protease active site.
to 1 and 2. In this report, we describe synthesis of novel
unnatural amino acids, N-Boc-protected
L-bis-tetra-
hydrofuranylglycines (3a and 3b) illustrated in Fig. 2.
The retrosynthetic analysis of the target amino acid
suggested the bis-(tetrahydrofuranyl) acetic acid as its
Figure 2. Structure of N-Boc-protected
L-bis-tetra-
hydrofuranylglycine.
Figure 1. Structure of HIV-1 protease inhibitors containing THF ligands.
Keywords: HIV-1 protease; HIV-1 protease inhibitor; radical cyclization; novel amino acid.
* Corresponding author. Tel.: +81-75-595-4636; fax: +81-75-595-4787; e-mail: kiso@mb.kyoto-phu.ac.jp
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00448-3

2932
E. Ami et al. / Tetrahedron Letters 43 (2002) 2931–2934
precursor, which in turn was proposed to be synthe-
sized from (E)-bromoether 6 or (Z)-bromoether 13.
(E)-Allyl alcohol 4 was obtained from the fumaric acid
monoethyl ester by the selective reduction of carboxylic
acid.8
in 63% yield after silica gel chromatography. Reaction
of 6 with tri-n-butyl tin hydride in refluxing benzene in
the presence of a catalytic amount of a,a%-azobisiso-
butyronitrile (AIBN) gave the corresponding 5-exo rad-
ical cyclization product 7 in 81% yield after column
chromatographic purification. In this step, three contin-
uous asymmetric centers of the carbon atoms were
spontaneously configured in the desired alignment.
Hydrolysis of bicyclic ester 7 in a mixture of aqueous
4N NaOH and EtOH provided the corresponding endo-
carboxylic acid 8 in 79% yield.
The synthesis of 3a and 3b is shown in Scheme 1.
Treatment of commercially available 2,3-dihydrofuran
5 with 1 equiv. of N-bromosuccinimide (NBS) and
ethyl (E)-4-hydroxy-2-butenoate 4 in CH2Cl2 at 0°C to
room temperature for 1 h, furnished the bromoether 6
Scheme 1. Synthetic route to L-bis-tetrahydrofuranylglycines 3a and 3b.

E. Ami et al. / Tetrahedron Letters 43 (2002) 2931–2934
2933
Alternatively, (Z)-alkene 13, prepared from 2,3-dihy-
drofuran 5 and monoacetate 12 by the same procedure
as that for bromo-ether 6, was subjected to the radical
cyclization reaction to yield bicyclic acetate 14 in 85%
after column separation. Compound 14 was hydrolyzed
by the treatment of aqueous 4N NaOH, and the result-
ing alcohol was subjected to Jones oxidation to give the
corresponding carboxylic acid 8 in 78% for two steps
(Scheme 2).
reduction. However, the reaction could be avoided in
10b with the use EtOH–THF. The absolute stereochem-
istry of isomer 11b was determined to be (S)-configura-
tion at C-3 by X-ray analysis (data not shown). The
C-terminal of these amino acids was then deprotected
through hydrolysis of the corresponding methyl ester
12a and oxazolidinone 11b using LiOH. Thus, we were
able to obtain both the threo-3a12 and erythro-3b13
forms of
L-bis-tetrahydrofuranylglycines in an optically
pure form.
We planned to separate both endo-enantiomers through
coupling to Evans’ oxazolidinone9 and to perform chi-
ral azidation of these two isomers separately. The car-
boxylic acid 8 was coupled to (S)-4-benzyl-2-oxazoli-
dinone using a mixed anhydride method.9 Unexpect-
edly, the oxazolidinones 9 could not be separated
through silica gel column chromatography. Hence, we
submitted the mixture to enantioselective introduction
of the azide group. Azidation using the standard Evans
protocol afforded a mixture of diastereomers (10a and
10b) in 62% yield after column chromatography. Fortu-
nately, one of the diastereomers could be crystallized
out of an ethyl acetate solution through slow diffusion
of n-hexane. The mother liquor was subjected to chro-
matographic purification again to separate the two
diastereomers. The diastereomeric azides also showed a
characteristic difference in their melting points. The
stereochemistry of the crystallized diastereomer 10a10
was determined to be (R)-configuration at C-3 by sin-
gle-crystal X-ray diffraction analysis (Fig. 3). However,
efforts to crystallize the other diastereomer 10b11 failed.
In conclusion, the novel amino acids were synthesized
efficiently and the enantiomers were easily purified by
crystallization. The importance and usefulness of these
amino acids in medicinal chemistry will be shown by
incorporation to the KNI series of HIV-1 protease
inhibitors.7c Further studies on the structural modifica-
tion as well as their SAR are actively under way.
Both azides 10a and 10b were reduced using SnCl2 in
MeOH–THF and EtOH–THF, respectively, and subse-
quently were protected as N-Boc derivatives (11a and
11b). Interestingly, there was concomitant transesterifi-
cation in the case of the (R)-isomer as this was largely
insoluble in MeOH–THF and hence necessitated the
use of an excess of MeOH and a small amount of
CHCl3. After 3 h reduction, and subsequent overnight
reaction with Boc2O, a 45% yield of transesterified
product 12a and 20% of the N-Boc protected oxazolidi-
none 11a were obtained. The use of large amounts of
CHCl3 to dissolve the azide retarded the reduction and
led to the recovery of 18% of the starting material. This
transesterification was also observed in the other
diastereomer 11b with increased reaction time of SnCl2
Figure 3. X-Ray crystal structure of 10a.
Scheme 2. Alternative route to 8.

2934
E. Ami et al. / Tetrahedron Letters 43 (2002) 2931–2934
Terashima, K.; Misawa, S.; Fukazawa, T.; Ueno, T.;
Acknowledgements
Sato, H.; Shintani, M.; Kiso, Y.; Hayashi, H. J. Med.
Chem. 1999, 42, 1789–1802.
This research was supported in part by the Frontier
Research Program of the Ministry of Education, Sci-
ence and Culture of Japan, and the Japan Health
Science Foundation. We thank Dr. T. Ogawa for sin-
gle-crystal X-ray diffraction for 11b and 2D NMR for
some of the compounds, Ms. M. Tsushima and Ms. K.
Oda for mass spectrometry and Mr. K. Hidaka for
technical assistance.
8. Fujisawa, T.; Mori, T.; Sato, T. Chem. Lett. 1983, 835–
838.
9. Evans, D. A.; Britton, T. C.; Ellman, J. A.; Dorow, R. L.
J. Am. Chem. Soc. 1990, 112, 4011–4030.
10. Selected data for tert-butyl-N-{(1S)-1-[(3R,3aR,6aS) per-
hydrofuro [2,3-b] furan-3-yl]-2-[(4S)-4-benzyl-2-oxo-1,3-
oxazolan-3-yl]-2-oxoethyl}
carbamate
(10a):
mp
1
165–168°C. H NMR (400 MHz/CDCl3) l 7.38–7.21 (m,
5H, aromatic protons), 5.78 (d, 1H, J=4.9 Hz), 5.25 (d,
1H, J=10.1 Hz), 4.80–4.74 (m, 1H), 4.34–4.26 (m, 2H),
4.16 (dd, 1H, J=7.3 Hz), 3.94 (ddd, 1H, J=6.0, 7.3, 7.5
Hz), 3.90–3.79 (m, 2H), 3.30 (dd, 1H, J=3.3, 13.6 Hz),
3.32–2.95 (m, 1H), 2.93–2.85 (m, 2H), 1.79 (d, 2H, J=7.5
Hz). HR-MS (EI+) calcd for C18H20N4O5 (M+) 372.1434,
found m/z 372.1436. [h]D=+24.84 (c=0.8 in THF).
11. Selected data for tert-butyl-N-{(1S)-1-[(3S,3aS,6aR) per-
hydrofuro [2,3-b] furan-3-yl]-2-[(4S)-4-benzyl-2-oxo-1, 3-
References
1. Kohl, N. E.; Emini, E. A.; Schlif, W. A.; Davis, L. J.;
Heimbach, J.; Dixon, R. A. F.; Herber, W. K.; Sigal, I.
S.; Darke, P. L.; Springer, J. P. Proc. Natl. Acad. Sci.
USA 1988, 85, 4686–4690.
2. Recent reviews, see: (a) Kempf, D. J.; Sham, H. L. Curr.
Pharm. Des. 1996, 2, 225–246; (b) Kiso, Y. Biopolymers
1996, 40, 235–244; (c) Lebon, F.; Ledecq, M. Curr. Med.
Chem. 2000, 7, 455–477; (d) Huff, J. R.; Kahn, J. Adv.
Protein Chem. 2001, 56, 213–251.
3. Ghosh, A. K.; Thompson, W. J.; Mckee, S. P.; Duong, T.
T.; Lyle, T. A.; Chen, J. C.; Darkle, P. L.; Zugay, J. A.;
Emini, E. A.; Schleif, W. A.; Huff, J. R.; Anderson, P. L.
J. Med. Chem. 1993, 36, 292–294.
4. Ghosh, A. K.; Thompson, W. J.; Fitzgerald, P. M. D.;
Culberson, J. C.; Axel, M. G.; McKee, S. P.; Huff, J. R.;
Anderson, P. S. J. Med. Chem. 1994, 37, 2506–2508.
5. Ghosh, A. K.; Thompson, W. J.; Lee, H. Y.; Mckee, S.
P.; Munson, P. M.; Duong, T. T.; Darkle, P. L.; Zugay,
J. A.; Emini, E. A.; Schleif, W. A.; Huff, J. R.; Anderson,
P. L. J. Med. Chem. 1993, 36, 924–927.
6. Thompson, W. J.; Ghosh, A. K.; Holloway, M. K.; Lee,
Y. L.; Munson, P. M.; Schwering, J. E.; Wai, J.; Darkle,
P. L.; Zugay, J. A.; Emini, E. A.; Schleif, W. A.; Huff, J.
R.; Anderson, P. L. J. Am. Chem. Soc. 1993, 115, 801–
803.
7. (a) Mimoto, T; Imai, J.; Tanaka, S.; Hattori, N.; Taka-
hashi, O.; Kisanuki, S.; Nagano, Y.; Shintani, M.;
Hayashi, H.; Sasikawa, H.; Akaji, K.; Kiso, Y. Chem.
Pharm. Bull. 1991, 39, 2465–2467; (b) Mimoto, T.; Imai,
J.; Kisanuki, S.; Enomoto, H.; Hatori, N.; Akaji, K.;
Kiso, Y. Chem. Pharm. Bull. 1992, 40, 2251–2253; (c)
Mimoto, T.; Kato, R.; Takaku, H.; Nojima, S.;
1
oxazolan-3-yl]-2-oxoethyl} carbamate (10b): mp 91°C. H
NMR (400 MHz/CDCl3) l 7.38–7.21 (m, 5H, aromatic
protons), 5.79 (d, 1H, J=4.8 Hz), 5.19 (d, 1H, J=11.0
Hz), 4.75–4.69 (m, 1H), 4.31–4.24 (m, 2H), 4.02–3.89 (m,
2H), 3.84 (dd, 1H, J=7.7, 7.5 Hz), 3.59 (dd, 1H, J=8.2
Hz), 3.30 (dd, 1H, J=3.3, 13.6 Hz), 3.10–3.04 (m, 1H),
3.03–2.96 (m, 1H), 2.87 (dd, 1H, J=9.0, 13.6 Hz), 2.11–
2.00 (m, 2H). HR-MS (EI+) calcd for C18H20N4O5 (M+)
372.1434, found m/z 372.1437. [h]D=+37.73 (c=0.8 in
THF).
12. Selected data for (2S)-2-[(3R,3aR,6aS) perhydrofuro [2,3-
b] furan-3-yl]-2-[(tert-butoxycarbonyl) amino] ethanoic
acid (3a): syrup, 1H NMR (300 MHz/CDCl3) l 6.67
(broad s, 1H), 5.74 (d, 1H, J=5.0 Hz), 5.38 (d, 1H,
J=9.0 Hz), 4.40–4.36 (m, 1H), 4.10–3.80 (m, 4H), 2.94–
2.82 (m, 1H), 2.12–2.07 (m, 2H), 1.44 (s, 9H). HR-MS
(FAB+) calcd for C13H21NO6 (M+) 287.1369, found m/z
310.1263 (M++Na). [h]D=−3.06 (c=0.6 in CHCl3).
13. Selected data for (2S)-2-[(3S,3aS,6aR) perhydrofuro [2,3-
b] furan-3-yl]-2-[(tert-butoxycarbonyl) amino] ethanoic
acid (3b): syrup, 1H NMR (300 MHz/CDCl3) l 6.44
(broad s, 1H), 5.72 (d, 1H, J=4.8 Hz), 5.21 (d, 1H,
J=8.1 Hz), 4.30–4.15 (m, 1H), 4.03–3.85 (m, 3H), 3.76
(dd, 1H, J=9.5 Hz) 2.991–2.87 (broad s, 1H), 2.59–2.52
(m, 1H), 1.97 (broad s, 2H), 1.45 (s, 9H). HR-MS (FAB+)
calcd for C13H21NO6 (M+) 287.1369, found m/z 310.1261
(M++Na). [h]D=−5.29 (c=0.9 in CHCl3).