Convenient syntheses of (3S,5S)-carbapenam-3-carboxylates and their biosynthetic relevance

Article abstract of DOI:10.1016/S0040-4039(02)02731-4

Starting from readily available glutamate derivatives, facile stereoselective syntheses of (3S,5S)-carbapenam-3-carboxylates have been developed. Their significance in confirming the absolute configuration of the natural material is discussed.

Full text of DOI:10.1016/S0040-4039(02)02731-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 973–976
Convenient syntheses of (3S,5S)-carbapenam-3-carboxylates and
their biosynthetic relevance
Barrie W. Bycroft,* Siri Ram Chhabra, Barrie Kellam and Paul Smith
School of Pharmaceutical Sciences, University of Nottingham, University Park, Nottingham NG7 2RD, UK
Received 8 November 2002; accepted 29 November 2002
Abstract—Starting from readily available glutamate derivatives, facile stereoselective syntheses of (3S,5S)-carbapenam-3-carboxy-
lates have been developed. Their significance in confirming the absolute configuration of the natural material is discussed. © 2003
Published by Elsevier Science Ltd.
Some years ago we reported the isolation, from vari-
ous bacterial species of Erwinia and Serratia, of the
known antibiotic, (R)-carbapen-2-em-3-carboxylic acid
1 together with both the cis- and trans-isomers of
carbapenam-3-carboxylic acid.1 The cis-isomer 2 was
assigned the (3S,5R) configuration by direct chemical
correlation with 1, the chirality of which had previ-
ously been established.2 The major trans-isomer which
was obviously epimeric with 2 at either C-3 or C-5
was provisionally assigned the (3R,5R) configuration,
not least because all naturally occurring carbapenem
antibiotics known at the time possessed the R-
configuration at the ring junction.3 However, synthe-
This observation has been particularly relevant to our
subsequent extensive biosynthetic studies. Initially we
established that 1, 2 and 3a were all derived from an
acetate unit, or an equivalent, and
L
-glutamate.5 The
genes encoding the elements required for their pro-
duction have been mapped, sequenced and
analysed.6,7 Those designated Car A–E within the cas-
sette were shown to be essential to the pathway and
we8 and others9 have proposed that the products of
these genes mediate the unusual sequence outlined in
Scheme 1. Intermediate 3a, the product of the CarA
protein, upon derivatisation to its PNB ester was
shown to be identical by 1H NMR spectroscopy to
sis of the trans-isomer starting from
D-glutamic acid,
that synthesised from
L
-glutamate.9 However the chi-
to confirm the assigned R-stereochemistry at C-3, fol-
lowed by direct comparison of the circular dichroism
(CD) plots of the p-nitrobenzyl (PNB) esters of both
the synthetic and natural material surprisingly
revealed a mirror image relationship.4 It followed
therefore that the natural trans-isomer possessed the
unique (3S,5S) configuration 3a.
rality of the natural material was not definitively
established but inferred on the basis of our original
data for 3d.
Recently the assignment of absolute configuration to
3a has been questioned by Tanaka and co-workers.10
This was based on their lengthy synthesis of the
methyl ester 3c and a comparison of its [h]D value
with that of our reported synthetic material but not
with the p-nitrobenzyl ester derived from natural
sources. While we were confident that our assignment
based on the CD data described above was correct,
the importance with respect to the biosynthetic pro-
posals required that this discrepancy was resolved. In
relation to our ongoing interest in the biosynthetic
pathways for carbapenem antibiotics, we have already
adapted our original synthesis to provide a more ver-
satile and efficient procedure. We now report its
application for the diastereoselective syntheses of 3a–d
Keywords: carbapenam-3-carboxylate; biosynthesis; absolute configu-
ration; b-lactam; tris(1,3-dihydro-2-benzoxazolin-3-yl)phosphine
oxide.
from
L-glutamate, which unequivocally confirms our
original assignment and resolves the discrepancy.
* Corresponding author. E-mail: barrie.bycroft@nottingham.ac.uk
0040-4039/03/$ - see front matter © 2003 Published by Elsevier Science Ltd.
PII: S0040-4039(02)02731-4

974
B. W. Bycroft et al. / Tetrahedron Letters 44 (2003) 973–976
Scheme 1. Proposed biosynthetic pathway to (R)-carbapen-2-em-3-carboxylic acid 1.
As shown in Scheme 2, aldehyde 9, which exists pre-
dominantly as the protected cyclic hemi-aminal 10, was
readily prepared in good yield from the corresponding
tion conditions to obtain the corresponding free acid
3a. The apparent instability of 3a to these mild condi-
tions has also been noted by others for related carbape-
nam-3-carboxylic acids.14 For our biosynthetic studies
we have only been able to generate aqueous solutions
of 3a via enzymatic hydrolysis of the corresponding
esters, preferably 3c. This led to a modification of the
initial step of the synthesis, which is shown in Scheme
3.
protected
conditions previously described.3 Wittig olefination
using t-butyl (triphenylphosphoranylidene)acetate
L
-glutamate derivative 8 using the reductive
afforded the trans-alkene 11. Following removal of the
t-butyl and Boc protecting groups and subsequent
treatment with base, the product cyclised to give almost
exclusively the trans-pyrrolidine derivative 12 in
approximately 50% overall yield from 8.
Dimethyl N,N-bisBoc-protected
prepared from commercially available
L
-glutamate 14 was
-glutamic acid
L
Earlier studies by others and ourselves used Mukaiya-
ma’s reagent to afford esters of 3a from the correspond-
ing b-amino acid, e.g. 12, albeit in poor yield. We now
report that after evaluating a range of other reagents,
tris(1,3 - dihydro - 2 - oxobenzoxazolin - 3 - yl)phosphine
oxide11 13 has proved to be the only method by which
3b12 can be obtained in acceptable yields (up to 70%).
The reagent 13 had previously been shown to cyclise
b-amino acids to the corresponding b-lactams and had
been reported for the cyclisation of the racemic cis-iso-
mer of 12.13
dimethyl ester hydrochloride as previously reported.15
DIBAL-H reduction of 14 afforded aldehyde 15 in high
yield, which was then transformed into the desired
methyl (3S,5S)-1-carbapenam-3-carboxylate 3c in good
yield as described earlier.16 Enzymatic hydrolysis of 3c
using pig liver esterase (PLE) on Eupergit® afforded 3a,
its presence in solution being confirmed by treatment
with 4-nitrobenzyl bromide followed by the isolation
and purification of p-nitrobenzyl (3S,5S)-1-carbape-
nam-3-carboxylate 3d.17 The CD curve obtained from
3d was superimposable on that of the PNB ester of the
material derived from natural sources (Dm −3.46 at 230
nm), thus confirming our original assignment of the
absolute configuration of 3a found in Erwinia and
While Scheme 2 provides ready access to esters such as
3b, surprisingly it has not been possible via hydrogena-
Scheme 2. Reagents and conditions: (i) (COCl)2, DMF, MeCN, THF, −30°C, 1 h., then 1 M LiAlH(OtBu)3, −78°C, 1 h, 82%; (ii)
t
Ph3PCHCO2 Bu, toluene, 110°C, 12 h, 86%; (iii) HCl/Et2O, 5 h, Et3N, THF, 48 h, 69%, 92% de; (iv) 13, Et3N, MeCN, 24 h, 67%.

B. W. Bycroft et al. / Tetrahedron Letters 44 (2003) 973–976
975
t
Scheme 3. Reagents and conditions: (i), DIBAL-H, Et2O, −78°C, 30 min, 85%; (ii) Ph3PCHCO2 Bu, toluene, 110°C, 12 h., 85%;
(iii) HCl/Et2O, 5 h, then Et3N, THF, 48 h, 60%, 92% de; (iv) tris(2,3-dihydro-2-oxobenzoxazol-3-yl) phosphine oxide, Et3N,
MeCN, 24 h, 45%; (v) PLE immobilised on Eupergit®, phosphate buffer pH 8.0, 3.5 h; then Aliquat® 336, DCM, 5 min; then
4-nitrobenzyl bromide, DCM, 20 h.
Serratia. The (3S,5S)-methyl ester 3c, [h]2D5 −194.0 (c
0.30, CHCl3), obtained from the above synthesis pos-
sessed a sign of optical rotation consistent with that
reported by Tanaka and co-workers10 for the enan-
tiomer. We have carefully checked the [h]D values of
the intermediates in our original communication, and
all of which are correct with exception of 3c. It appears
that we inadvertently misreported the sign of rotation
for the synthetic (3R,5R)-methyl ester.4 However, this
error has no impact on the assignment of the absolute
stereochemistry of the naturally occurring material.
4. Bycroft, B. W.; Chhabra, S. R. J. Chem. Soc., Chem.
Commun. 1989, 423–425.
5. Bycroft, B. W.; Maslen, C.; Box, S. J.; Brown, A. G.;
Tyler, J. W. J. Antibiot. 1988, 41, 1231–1242.
6. McGowan, S. J.; Sebaihia, M.; Porter, L. E.; Stewart, G.
S. A. B.; Williams, P.; Bycroft, B. W.; Salmond, G. P. C.
Mol. Microbiol. 1996, 22, 415–426.
7. McGowan, S. J.; Sebaihia, M.; O’Leary, S.; Hardie, K.
R.; Williams, P.; Stewart, G. S. A. B.; Bycroft, B. W.;
Salmond, G. P. C. Mol. Microbiol. 1997, 26, 545–556.
8. McGowan, S. J.; Bycroft, B. W.; Salmond, G. P. C.
Trends Microbiol. 1998, 6, 203–208.
In summary, we have developed efficient enantioselec-
tive syntheses of 1-carbapenam-3-carboxylate esters and
a convenient means of generating the unstable free acid
in solution. CD measurement on the PNB ester confi-
rms our original assignment of the absolute configura-
tion for the natural material as (3S,5S). Application of
the new synthetic routes for the preparation of substi-
tuted carbapenams is underway and will be reported in
due course.
9. Li, R.; Stapon, S.; Blanchfield, J. T.; Townsend, G. A. J.
Am. Chem. Soc. 2000, 122, 9296–9297.
10. Tanaka, H.; Sakagami, H.; Ogasawara, K. Tetrahedron
Lett. 2002, 43, 93–96.
11. Nagamatsu, T.; Kunieda, T. Chem. Pharm. Bull. 1988,
36, 1249–1251.
12. Benzyl (3S,5S)-1-carbapenam-3-carboxylate 3b: mp 32–
33°C; m/z (+ES) 263.9 (MNa+ requires 264.3); [h]2D0
−202.4 (c 0.25, CHCl3); wmax (KBr): 1746 (b-lactam and
ester CꢀO) cm−1; lH (CDCl3): 1.48–1.63 (1H, m, 1b-H),
2.17–2.39 (2H, m, 1a- and 2a-H), 2.53–2.60 (1H, m,
2b-H), 2.65 (1H, dd, J 15.8 and 2.1 Hz, 6b-H), 3.30 (1H,
dd, J 15.8 and 5.0 Hz, 6a-H), 3.85–3.93 (1H, m, 5b-H),
4.48 (1H, t, J 7.0 Hz, 3a-H), 5.20 (2H, s, PhCH2), 7.40
(5H, s, Ph); lC (CDCl3): 31.0 (1-C), 35.5 (2-C), 42.7
(6-C), 53.1 (5-C), 59.2 (3-C), 67.0 (PhCH2), 128.2, 128.4,
128.6 (phenyl C), 135.4 (benzyl i-C), 171.3 (ester CO),
176.2 (b-lactam CO).
Acknowledgements
We would like to thank the BBSRC for a studentship
to PS and Drs Mark Searle and Stephen Simpson of the
School of Chemistry, University of Nottingham for CD
measurements.
13. Gilchrist, T. L.; Lemos, A.; Ottaway, C. J. J. Chem. Soc.,
Perkin Trans. 1 1997, 3005–3012.
14. Buchi, M. D.; Breiman, R.; Meshulam, H. J. Org. Chem.
1983, 48, 1439–1444.
15. Padron, J. M.; Kokotos, G.; Martin, T.; Markidis, T.;
Gibbons, W. A.; Martin, V. S. Tetrahedron: Asymmetry
1998, 9, 3381–3394.
16. Methyl (3S,5S)-1-carbapenam-3-carboxylate 3c: mp 32–
33°C; [h]2D5 −194.0 (c 0.30, CHCl3) {lit.10 [h]2D5 +199.1 (c
0.20, CHCl3) for (3R,5R) enantiomer}; m/z (+ES) 170.0
(MH+ requires 170.1); wmax (KBr): 1762 (b-lactam CꢀO),
1735 (ester CꢀO) cm−1; lH (CDCl3): 1.48–1.62 (1H, m,
References
1. Bycroft, B. W.; Maslen, C.; Box, S. J.; Brown, A. G.;
Tyler, J. W. J. Chem. Soc., Chem. Commun. 1987, 1623–
1625.
2. (a) Ueda, Y.; Damas, C. E.; Vinet, V. Can. J. Chem.
1983, 61, 2257–2263; (b) Miyashita, M.; Chida, N.;
Yoshikoshi, A. J. Chem. Soc., Chem. Commun. 1984,
195–196.
3. Southgate, R.; Elson, S. Fortschr. Chem. Org. Naturst.
1985, 47, 1–106.

976
B. W. Bycroft et al. / Tetrahedron Letters 44 (2003) 973–976
1b-H), 2.18–2.36 (2H, m, 1a- and 2a-H), 2.52–2.61 (1H,
m, 2b-H), 2.65 (1H, dd, J 15.8 and 2.0 Hz, 6b-H), 3.29
(1H, dd, J 15.8 and 4.9 Hz, 6a-H), 3.75 (3H, s, Me),
3.84–3.90 (1H, m, 5b-H), 4.41 (1H, t, J 7.7 Hz, 3a-H); lC
(CDCl3): 31.1 (1-C), 35.4 (2-C), 42.6 (6-C), 52.4 (Me),
53.1 (5-C), 59.1 (3-C), 171.9 (ester CꢀO), 176.3 (b-lactam
CꢀO)
17. p-Nitrobenzyl (3S,5S)-1-carbapenam-3-carboxylate 3d: lH
(CDCl3): 1.48–1.62 (1H, m, 1b-H), 2.27–2.33 (2H, m, 1a-
and 2a-H), 2.59–2.71 (2H, m, 2b-H and 6b-H), 3.31 (1H,
dd, J 15.9 and 4.8 Hz, 6a-H), 3.85–3.89 (1H, m, 5b-H),
4.49 (1H, t, J 7.6 Hz, 3a-H), 5.26 (2H, s, CH2Ar), 7.52
(2H, d, J 8.8 Hz, Ar 2%,6%-H2), 8.24 (2H, d, J 8.8 Hz, Ar
3%,5%-H2).