A concise synthesis of α-amino acid N-carboxy anhydrides of (2S,3S)-β-substituted serines

Article abstract of DOI:10.1016/S0040-4039(01)01766-X

A general access to 3-hydroxy-4-(1-hydroxyalkyl)-β-lactams with a C₄-C₁′ relative configuration unlike is provided. The subsequent oxidative ring expansion of the corresponding 3-hydroxy β-lactam smoothly affords an α-amino acid N-ca

Full text of DOI:10.1016/S0040-4039(01)01766-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 8955–8957
A concise synthesis of a-amino acid N-carboxy anhydrides of
(2S,3S)-b-substituted serines
Claudio Palomo,* Mikel Oiarbide, In˜aki Ganboa and Jose´ I. Miranda
Departamento de Qu´ımica Orga´nica I, Facultad de Qu´ımica, Universidad del Pa´ıs Vasco, Apdo 1072, 20080 San Sebastia´n,
Spain
Received 3 July 2001; accepted 19 September 2001
Abstract—A general access to 3-hydroxy-4-(1-hydroxyalkyl)-b-lactams with a C4–C1% relative configuration unlike is provided. The
subsequent oxidative ring expansion of the corresponding 3-hydroxy b-lactam smoothly affords an a-amino acid N-carboxy
anhydride (NCA) formally derived from (2S,3S)-b-substituted serine, which upon sequential peptide coupling furnishes the
tripeptide segment 2, present in lysobactin. © 2001 Elsevier Science Ltd. All rights reserved.
a-Amino b-hydroxy acids, also termed b-substituted
serines, are frequently found in bioactive natural prod-
ucts,1 as in the macrocyclic peptide lactone antibiotic
lysobactin.2 The antibacterial mode of action of
lysobactin is comparable with the observed selectivity
and potency of the well known, clinically useful antibi-
otic vancomycin.3 In those infectious strains of van-
comycin-resistant bacteria, lysobactin might be the
alternative antibacterial agent of choice, in spite of its
toxicity. Consequently, the development of new pep-
tides of the lysobactin family with improved therapeutic
indexes, is of considerable interest. Necessarily, syn-
thetic or semi-synthetic approaches to lysobactin itself
and to the analogs should address the construction of
the a-amino b-hydroxy acid units present in their struc-
tures with a correct stereochemistry.4
Recently, we have reported a b-lactam approach to
a-amino acid N-carboxyanhydrides (NCAs) formally
derived from (2S,3R)-b-substituted serines, and the
subsequent synthesis of tripeptide 1.5 According to this
approach (Scheme 1) from the ring expansion of 3-
hydroxy b-lactams 4, NCAs 5 are obtained that can be
directly submitted to a peptide coupling reaction. The
traditionally required carboxy group activation of the
intermediate a-amino acid is thus avoided.6 The
method is still slightly restricted, because NCAs 6,
showing a relative configuration unlike, which often
* Corresponding author. E-mail: qoppanic@sc.ehu.es
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01766-X

8956
C. Palomo et al. / Tetrahedron Letters 42 (2001) 8955–8957
OTBS
R
OTBS
O
H
H H
HO
OTBS
R
BnO
[2+2]
NaOCl-TEMPO
ring expansion
R
O
+
C
O
cycloaddition
NBn
NBn
NBn
O
5
O
O
like
4
3
OP
O
OP
H
H H
H H
HO
HO
R
OH
R
O
NBn
NBn
NBn
unlike
O
O
O
6
Scheme 1. General strategies for access to NCAs formally derived from b-substituted serines: previously described approach for
access to the like series and the proposed route to the unlike series.
corresponds with that of the biologically active com-
pounds, are not accessible through this route. This is so
because the sense of asymmetric induction typically
imparted by chiral a-oxyaldehydes-derived imines 3
during the [2+2] cycloaddition reaction7 produces the
corresponding b-lactam adducts with a a,b-like stereo-
chemical relationship. With the aim of expanding the
usefulness of our b-lactam approach, we herein report
on the facile access to NCAs 6 and to peptides derived
therefrom through a modified strategy.
reduction of 9a with sodium borohydride in methanol
as solvent provided an almost equimolecular mixture of
carbinols 10a/11a, while the reduction of the same
compound carried out with L-Selectride in either diethyl
ether or anhydrous THF furnished 10a/11a in a ratio
15:85. Likewise, the reduction of 9b with DIBAL-H in
THF as solvent proceeded to give 10b/11b in a ratio of
45:55. The ratio of isomers was improved up to 30:70
by using toluene as solvent, but, once again, the best
result was attained when the reduction was carried out
with -Selectride. In this latter case, a mixture of 10b/
L
In the modified strategy, the exocyclic stereocenter (car-
binol) is formed in a later stage with the desired
configuration. This may be achieved by diastereofacial
discrimination across the two faces of the precursor
carbonyl compound. The general route begins with the
enantiomerically pure 3-benzyloxy-4-carboxy azetidin-
2-one 7 as the key starting material that acts as a novel
non-racemic a-aminomalonic acid surrogate.8,9 The free
11b was produced in a 7:93 ratio. In every case, from
the corresponding mixture of carbinols each isomer was
separated by column chromatography, and the minor
isomer could be reused,12,13 as shown in Scheme 2.
With these products in hand, the synthesis of the pep-
tide fragment 15 of lysobactin was undertaken (Scheme
3). To this end, the hydroxyl group in 11a was first
protected as tert-butyldimethylsilyl ether and the result-
ing intermediate was subjected to O-debenzylation to
afford 12 (oil, [h]D25=+31.0, c=1, CH2Cl2). Compound
12, upon treatment with a solution of commercial
bleach and a catalytic amount of TEMPO, furnished
the NCA 13 in 95% yield. Thus, the access to this
NCA, which traditionally would require the previous
synthesis of the corresponding a-amino acid, can now
be obtained from a non-a-amino acid precursor in a
concise fashion. Furthermore, this approach generates
the required amino acid as an active species, thereby
overcoming the need of additional protection and acti-
carboxy function in
7 allows for a correlative
organometallic addition–reduction process that results
in the formation of the exocyclic carbinol with the right
configuration. Further ring expansion and peptide cou-
pling culminate the synthetic plan.
The carboxylic acid 710 was first transformed into
Weinreb’s amide11 8. Subsequent treatment of 8 with
Grignard reagents provides ketones 9 in yields in the
range 65–75%. The reduction of ketones 9 to the
desired carbinols was found to be strongly sensitive to
the type of reducing agent employed. For example, the
O
H
H
H
H
BnO
OMe
BnO
CO2H
Cl2C2O2, DMF
RMgX
N
NBn Me
THF, 0ºC, 3h
CH2Cl2, r.t., 2h
then HNMeOMe
80%
NBn
O
Product Yield %a 10:11b
O
7
8
9a
9b
9c
9d
70
65
70
75
O
OH
OH
H
H
H
H
H
H
BnO
BnO
BnO
L-Selectride®
R
R
R
+
NBn
THF, -60ºC, 3h
NBn
11
NBn
O
O
O
11a
11b
11c
11d
15:85
7:93
65
65
70
65
10
9
(Cl3CO)2O, DMSO, Et3N
10:90
5:95
a R: Me c R: iPr
b R: Et d R: Ph
CH2Cl2, -78ºC, 30min
then r.t., 1h
a) Yield of pure product.
b) Determined by HPLC analysis.
Scheme 2. Stereoselective access to b-lactams 11 from enantiomerically pure 7.

C. Palomo et al. / Tetrahedron Letters 42 (2001) 8955–8957
8957
OTBS
1. tBuMe2SiCl, Imidazol
DMF, r.t., 11h, 70%
H
OTBS
O
H
H
1M NaOCl, TEMPO (1mol%)
KBr, NaHCO3, pH=7
HO
O
11a
NBn
NBn
2. H2, Pd/C, EtOH
r.t., 16h, 90%
CH2Cl2-H2O, 0ºC, 1-5min
O
O
95%
12
13
tBuO
NH
O
OTBS
Ph
O
tBuO
Cbz-(L)-Ile, DCC, HOBt
H2NCH2CO2C(CH3)3
CH2Cl2, r.t., 16h
OTBS
O
O
N
H
CH2Cl2, r.t., 16h
75%
HN
O
HN
14
75%
Cbz N
H
15
Scheme 3. Synthesis of the tripeptide segment 15 present in lysobactin.
vation steps for peptide coupling. Accordingly, the
coupling of 13 with glycine tert-butyl ester afforded the
dipeptide product 14 in 75% yield without traces of
epimerized product at Ca.14 Subsequent N-debenzyla-
tion and further coupling with N-Cbz-Ille-OH under
standard peptide coupling conditions provided 15 in
75% yield. In summary, the general route presented
here, that makes use of a new non-racemic amino
malonic acid surrogate, broadens the usefulness of ear-
lier methodology for the access to a-amino acid N-car-
boxyanhydrides (NCAs) from non-amino acid
precursors, and applications to the synthesis of non-
trivial peptides arise.
8. For other surrogates of a-aminomalonic acid, see: (a)
Garner, P. Tetrahedron Lett. 1984, 25, 5855; (b) Lubell, W.
D.; Jamison, T. F.; Rapoport, H. J. Org. Chem. 1990, 55,
3511.
9. Krysan, D. J. Tetrahedron Lett. 1996, 36, 3303.
10. (a) Palomo, C.; Cabre, F.; Ontoria, J. M. Tetrahedron Lett.
1992, 33, 4819. Also see: (b) Jayaraman, M.; Deshmukh,
A. R.; Bhawal, B. M. Tetrahedron 1996, 52, 8989.
11. Weinreib, S. M.; Nahm, S. Tetrahedron Lett. 1981, 22,
3815.
12. Representative data: 11a: oil, [h]2D5=+57.6 (c=1, CH2Cl2);
1H NMR (l, ppm, CDCl3): 7.36–7.20 (m, 10H), 5.00 (d,
1H, J=11.5 Hz), 4.78 (d, 1H, J=15.3 Hz), 5.74 (d, 1H,
J=11.5 Hz), 4.73 (d, 1H, J=4.9 Hz), 4.10 (d, 1H, J=15.3
Hz), 4.00 (m, 1H), 3.43 (dd, 1H, 4.9, 5.1 Hz), 2.86 (d, 1H,
J=7.1 Hz), 1.21 (d, 3H, J=6.5 Hz); 13C NMR (l, ppm,
CDCl3): 167.5, 136.3, 135.0, 128.7, 128.5, 128.1, 128.0,
127.9, 127.7, 81.5, 73.1, 66.0, 60.5, 44.3, 20.0; 11b: oil,
Acknowledgements
1
[h]2D5=+12.6 (c=1, CH2Cl2); H NMR (l, ppm, CDCl3):
This work was financially supported by Gobierno
Vasco, Universidad del Pais Vasco and by Ministerio
de Ciencia y Tecnolog´ıa.
7.33–7.19 (m, 10H), 4.99 (d, 1H, J=11.7 Hz), 4.79–4.71 (m,
3H), 4.07 (d, 1H, J=15.2 Hz), 3.70 (m, 1H), 3.46 (m, 1H),
2.69 (d, 1H, J=7.7 Hz), 1.47 (m, 2H), 0.91 (t, J=7.3 Hz);
13C NMR (l, ppm, CDCl3): 167.6, 136.3, 135.0, 128.7,
128.5, 128.1, 127.9, 127.7, 81.6, 73.2, 71.2, 59.3, 44.3, 27.0,
1
References
10.2; 11c: oil, [h]D25=+67.6 (c=1, CH2Cl2); H NMR (l,
ppm, CDCl3): 7.34–7.20 (m, 10H), 5.01 (d, 1H, J=11.72
Hz), 4.78 (m, 3H), 4.04 (d, 1H, J=15.34 Hz), 3.61 (m, 2H),
2.73 (d, 1H, J=7.55 Hz), 1.76 (m, 1H), 0.89 (s, 3H), 0.86
(s, 3H); 13C NMR (l, ppm, CDCl3): 167.4, 136.3, 135.0,
128.7, 128.4, 128.2, 127.9, 127.7, 81.7, 74.4, 73.1, 57.1, 44.2,
1. Barret, G. C. In Amino Acids, Peptides and Proteins:
Specialist Periodical Reports; Davies, J. S., Ed.; The Royal
Society of Chemistry: London, 1997; Vol. 28, p. 1.
2. Tymiak, A. A.; McCornick, T. J.; Unger, S. E. J. Org.
Chem. 1989, 54, 1149.
31.1, 19.7, 17.2; 11d: oil, [h]2D5=+28.6 (c=1, CH2Cl2); H
1
3. (a) Salase, D. M.; Marino, J.; Jacobs, M. R. Antimicrob.
Agents Chemother. 1984, 25, 527; (b) Horowitz, H. W.;
Handwerger, S.; Van Horn, K. G.; Wormser, G. P. Lancet
1987, 2, 1329.
4. For another approach to the a-amino b-hydroxy acids of
lysobactin, see: Armaroli, S.; Cardillo, G.; Gentilucci, L.;
Gianotti, M.; Tolomelli, A. Org. Lett. 2000, 2, 1105.
5. (a) Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Odriozola,
B.; Maneiro, E.; Miranda, J. I.; Urchegui, R. Chem.
Commun. 1996, 161; (b) Palomo, C.; Aizpurua, J. M.;
Ganboa, I.; Oiarbide, M. Pure Appl. Chem. 2000, 72, 1763.
6. For a review on NCAs, see: Kricheldorf, H. R. h-
Aminoacid N-Carboxy Anhydrides and Related Heterocy-
cles; Springer: Berlin, 1987.
NMR (l, ppm, CDCl3): 7.40–7.16 (m, 13H), 6.70 (m, 13H),
4.99 (d, 1H, J=11.5 Hz), 4.93 (dd, 1H, J=7.8, 4.4 Hz),
4.77 (d, 1H, J=11.5 Hz), 4.74 (d, 1H, J=4.7 Hz), 4.60 (d,
1H, J=14.8 Hz), 3.75 (dd, 1H, J=7.8, 4.8 Hz), 3.30 (d,
1H, J=14.8 Hz), 3.04 (d, 1H, J=4.6 Hz); 13C NMR (l,
ppm, CDCl3): 167.5, 140.5, 136.5, 134.7, 128.7, 128.5,
128.2, 128.1, 127.2, 81.8, 73.3, 73.0, 60.5, 44.2.
13. Prior attempts to effect carbinol inversion under the
Mitsonobu reaction conditions in related substrates invari-
ably led to yields below 50%. See: Mielgo, A., Doctoral
Thesis, San Sebastia´n, 1996, p. 106. In the present case,
direct conversion of 10 into 11 was not examined.
14. For some aspects on racemization during ring opening of
NCAs, see: (a) Sim, T. B.; Rapoport, H. J. Org. Chem.
1999, 64, 2532; (b) Palomo, C.; Oiarbide, M.; Landa, A.;
Linden, A. J. Org. Chem. 2001, 66, 4180.
7. Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M.
Eur. J. Org. Chem. 1999, 3223.