Determination of the absolute configuration of an unexpectedly stable N- silylated isomer isolated en route to the trinem antibiotic GV129606

Article abstract of DOI:10.1016/S0957-4166(98)00310-3

The unexpected (3S,4R)-3-[(R)-1-(hydroxy)ethyl]-4-[(2'R,6'S)-1'-oxo-2'- (N-benzyloxy-N-methyl)aminocyclohexen-6'-yl]-1-(t-butyl- dimethylsilyl)azetidin-2-one was one of the main reaction products of the Lewis acid catalysed condensation of (3S,4R)-3-[(R)-

Full text of DOI:10.1016/S0957-4166(98)00310-3

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 9 (1998) 2787–2790
Determination of the absolute configuration of an unexpectedly
stable N-silylated isomer isolated en route to the trinem antibiotic
GV129606
Angelo Pecunioso,∗ Micaela Maffeis and Carla Marchioro
Glaxo Wellcome S.p.A., Medicines Research Centre, Via Fleming 4, 37135 Verona, Italy
Received 13 July 1998; accepted 31 July 1998
Abstract
The unexpected (3S,4R)-3-[(R)-1-(hydroxy)ethyl]-4-[(20R,60S)-10-oxo-20-(N-benzyloxy-N-methyl)aminocyclo-
hexen-60-yl]-1-(t-butyl-dimethylsilyl)azetidin-2-one was one of the main reaction products of the Lewis acid
catalysed condensation of (3S,4R)-3-[(R)-1-(t-butyl-dimethylsilyloxy)ethyl]-4-acetoxyazetidin-2-one with
1-trimethylsilyloxy-6-(N-benzyloxycarbonyl-N-methylamino)cyclohexene. Its absolute configuration was
established by NMR experiments on the corresponding, conformationally rigid, acetonide derivative. © 1998
Elsevier Science Ltd. All rights reserved.
The bicyclic compound GV158943X 1 (Fig. 1) is the key synthetic intermediate en route to
GV129606X 2,1 a potent and broad spectrum trinem2 antibiotic. As part of a wider program aimed
at the definition of a more practical and scaleable synthetic route to 2, we approached the synthe-
sis of the intermediate 1 via the single-step Lewis acid catalysed aldol-type condensation between
(3S,4R)-3-[(R)-1-(t-butyl-dimethylsilyloxy)ethyl]-4-acetoxyazetidin-2-one 3 and 1-trimethylsilyloxy-6-
(N-benzyloxycarbonyl-N-methylamino)cyclohexene 4 (Scheme 1).3
Preliminary results show that the Lewis acids effectively producing the desired isomer 1 also gave
variable amounts of an unexpected4 N-silylated isomer 55 (Scheme 1) and, in some cases, this latter
Fig. 1.
∗
Corresponding author. E-mail: AAP6071@Glaxowellcome.co.uk
0957-4166/98/$ - see front matter © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00310-3

2788
A. Pecunioso et al. / Tetrahedron: Asymmetry 9 (1998) 2787–2790
Scheme 1.
derivative was the predominant reaction product. The isomer 5 was the only N-silylated derivative
observed under different reaction conditions and its N–silicon bond proved to be unusually stable,
surviving both acid work-up conditions and elution on silica gel. The molar ratio 1:5 obtained was
essentially related to the nature of the catalyst employed in the reaction.3
For a better understanding of the factors affecting the diastereoselection of the condensation reaction,
we judged it necessary to determine the absolute configuration of this novel N-silylated derivative.
However, due to an unfavourable overlap of the signals, it was not possible to obtain a straightforward
assignment of the absolute configuration of the C-20 and C-60 stereocentres in 5 by means of NMR
techniques. Attempts to convert 5 into the corresponding, and possibly more easily interpretable, O-
silylated derivative 7 (Scheme 2) failed due to the surprising stability of the N–silicon bond in the bis-
silylated derivative 66 that caused a preferential or, at least simultaneous, removal of the protecting group
on the oxygen (Scheme 2).
Scheme 2.
The absolute configuration of 5 was eventually determined by converting it into the conformationally
more rigid acetonide derivative 10 (Scheme 3) to obtain an unambiguous NOE pattern. Consequently,
the reduction of 5 by means of NaBH4 gave the alcohol 8 as a single isomer, this was in turn desilylated
with tetrabutylammonium fluoride and reacted with 2,2-dimethoxypropane to form 10.7
The NOE data were obtained for compounds 8, 9 and 10. While for compounds 8 and 9 the measured
NOEs resulted in more than one solution, the NMR studies on compound 10 allowed an unambiguous
determination of the stereochemistry on the basis of the NOE effects observed between H-11, H-9 and
H-4 together with the enhancement from H-5 to H-10 (Fig. 2). As a consequence, the stereochemistry of
the C-20 and C-60 stereocentres in 5 (Fig. 2) were determined as 20R,60S, i.e. the opposite to that observed
in the isomer 1. A convincing rationale for the formation of a stable N-silylated derivative only in the
(20R,60S)-isomer has yet to be found and further work on this subject is in progress.

A. Pecunioso et al. / Tetrahedron: Asymmetry 9 (1998) 2787–2790
2789
Scheme 3.
Fig. 2.
Acknowledgements
We wish to thank the personnel of the mass spectrometry laboratory of GlaxoWellcome Research
Laboratories Verona for the support given to this work.
References
1. (a) Di Modugno, E.; Broggio, R.; Erbetti, I.; Lowther, J. Antimicrob. Agents Chemother. 1997 41, 2742; (b) Biondi,
S. Trinems: Synthesis and Antibacterial Activity of a New Generation of Antibacterial beta-Lactams. In Anti-Infective:
Recent Advances in Chemistry and Structure–Activity Relationship; Bentley, P. H. and O’Hanlon, P. J., Eds; The Royal
Society of Chemistry, 1997, p. 86.
2. This class of compounds was formerly referred to as tribactams. See Gaviraghi, G. Eur. J. Med. Chem. 1995, 30, 467.
3. Preliminary results on the preparation of 1 by single-step reaction of 3 with 4 will be published in due course.
4. The aldol-type condensations of the azetidinone 3 with different nucleophiles have been extensively used during the studies
related to the syntheses of several penem and carbapenem antibiotics (see: Wild, H. In The Organic Chemistry of Beta-
Lactams; Georg, G. I., Ed.; VCH: New York, NY, 1993, Chapter 2) but no products such as 5 were previously reported.
5. To a suspension of FeCl3 (324 mg, 2 mmol) in dry dichloromethane (10 ml), stirred under nitrogen and at 0°C, was added
a solution of the azetidinone 3 (574 mg, 2 mmol) and silyl enol ether 4 (2 g, 6 mmol) in dry dichloromethane (15 ml). The
reaction mixture was stirred for 30 min at 0°C, then quenched with a saturated solution of NaHCO3 (40 ml) and extracted
with AcOEt (2×40 ml). The combined organic extracts were dried over sodium sulphate, filtered and concentrated under
reduced pressure. The crude product was purified by flash chromatography (ethyl acetate:cyclohexane=1:1) yielding ca.
220 mg each of compounds 1 and 5. Compound 5 showed: IR (CDCl3, cm−1): 1738 C_O β-lactam; 1693 C_O; 1H-NMR
(CDCl3): 7.4 (5H, m, –Ph); 6.16 (1H, bs, –OH); 5.17, 5.07 (2H, m, –CH2Ph); 4.57 (1H, bm, H20); 4.16 (1H, m, H5); 3.81
(1H, m, H4); 2.89 (3H, s, –NCH3); 2.72 (1H, m, H3); 2.58 (1H, bm, H60); 2.18–1.80 (6H, m, H30+H40+H50); 1.25 (3H, d,

2790
A. Pecunioso et al. / Tetrahedron: Asymmetry 9 (1998) 2787–2790
J=6.0 Hz, Me); 0.88 (9H, s, –C(CH3)3); 0.09 (3H, s, –Si(Me)2); 0.07 (3H, s, –Si(Me)2); MS: 489 (M+H), 431; Anal. calcd
for C26H40N2O5Si: C, 63.17; H, 8.13; N, 5.84. Found: C, 63.89; H, 8.27; N, 5.73.
6. Compound 6 showed: IR (CDCl3, cm−1): 1738 C_O β-lactam; 1693 C_O; 1H-NMR (CDCl3): 7.34 (5H, m, –Ph); 5.11
(2H, bs, –CH2Ph); 4.56 (1H, bm, H20); 4.22–4.04 (2H, m, H5+H4); 2.94 (1H, m, H3); 2.87 (3H, s, –NCH3); 2.74 (1H,
bm, H60); 2.02–1.78–1.80 (6H, m, H30+H40+H50); 1.10 (3H, bd, Me); 0.95 (9H, s, –C(CH3)3); 0.87 (9H, s, –C(CH3)3);
0.20 (3H, s, –Si(Me)2); 0.17 (3H, s, –Si(Me)2); 0.06 (3H, s, –Si(Me)2); 0.04 (3H, s, –Si(Me)2); MS: 603 (M+H), 545.
7. The N-silylated derivative 5 (1 g) was dissolved at 0°C in 50 ml of an ethanol:water (9:1) mixture and treated portionwise
with 2.5 molar equiv. of NaBH4. After stirring for a further 1 hour at room temperature, the reaction mixture was diluted
with water (20 ml), concentrated to ca. 20 ml, acidified with 0.1 N HCl and extracted with ethyl acetate. The organic
layer was washed with brine (2×20 ml), dried over Na2SO4 and the solvent removed in vacuo. The crude product was
1
purified by silica gel chromatography (EtOAc, 100%) to give 8 in 75% yield. H-NMR (CDCl3): 7.4–7.3 (5H, m, –Ph);
6.30 (1H, bs, 10-OH); 5.15 (2H, m, –CH2Ph); 4.20 (1H, bm, H20); 4.23 (1H, m, H5); 3.89 (2H, m, H4+H10); 2.85 (3H,
s, –NCH3); 2.75 (1H, bm, H3); 2.2 (1H, m, H60); 1.80–1.40 (6H, m, H30+H40+H50); 1.23 (3H, d, J=6.2 Hz, Me); 0.88
(9H, s, –C(CH3)3); 0.08 (6H, s, –Si(Me)2); MS: 491 (M+H), 433. Anal. calcd for C26H42N2O5Si: C, 62.32; H, 8.59; N,
5.50. Found: C, 62.62; H, 8.64; N, 5.71. Compound 8 (50 mg, 0.1 mmol) was dissolved in dry tetrahydrofuran (3 ml) and
treated with 2 equiv. of tetrabutylammonium fluoride. The reaction mixture was stirred for 3 hours at room temperature,
the solvent was evaporated in vacuo and the crude mixture purified by chromatography on silica gel (EtOAc:MeOH=9:1)
to give 25 mg of 9 (yield=66%). IR (CDCl3, cm−1): 1749 C_O β-lactam; 1691 C_O; 1H-NMR (CDCl3): 7.4–7.3 (5H, m,
–Ph); 6.60 (1H, bs, NH); 5.13 (2H, m, –CH2Ph); 4.27 (1H, bm, H20); 4.15 (1H, m, H5); 3.9 (2H, m, H4+H10); 2.96 (1H,
bm, 10–OH); 2.84 (3H, s, –NCH3); 2.85 (1H, bm, H3); 2.2 (1H, m, H60); 1.80–1.40 (6H, m, H30+H40+H50); 1.32 (3H,
bm, Me); MS: 377 (M+H). Compound 9 (20 mg) was dissolved in 2,2-dimethoxypropane and the reaction mixture was
stirred at room temperature for 2 hours. The solvent was removed in vacuo and the residue diluted with ethyl acetate (20
ml) and washed successively with saturated NaHCO3 (20 ml) and brine (20 ml). The organic layer was dried over Na2SO4,
the solvent evaporated and the crude product purified by chromatography on silica gel (EtOAc 100%) to give 18 mg of the
compound 10 (yield=78%). IR (CDCl3, cm−1): 1738 C_O β-lactam; 1690 C_O; 1H-NMR (CDCl3): 7.3 (5H, m, –Ph);
5.14 (2H, m, –CH2Ph); 4.45 (1H, bm, H5); 4.17 (1H, bm, H12); 3.96 (1H, dd, J=5.4, 11.70 Hz, H4); 3.65 (1H, bm, H10);
2.88 (3H, s, –NCH3); 2.77 (1H, dd, J=5.8, 1.8 Hz, H11); 2.10 (1H, m, H9); 2.01 (1H, m, –OH); 1.78 (2H, m); 1.65 (3H, s,
–Me); 1.7–1.4 (3H, m); 1.30 (1H, m); 1.30 (3H, s, Me); 1.29 (3H, d, J=5.8 Hz, Me); MS: (M+H) 603.