From theoretical calculations to the enantioselective synthesis of a 1,3,4-trisubstituted Gly-derived 2-azetidinone

Article abstract of DOI:10.1016/j.tetlet.2007.11.099

Theoretical calculations on the transition states of the cyclization of 2S-chloropropionyl amino acid derivatives to the corresponding β-lactams have served to explain the high stereoselectivity of the reaction, and have been the driving force to extend t

Full text of DOI:10.1016/j.tetlet.2007.11.099

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 215–218
From theoretical calculations to the enantioselective synthesis
of a 1,3,4-trisubstituted Gly-derived 2-azetidinone
*
´
´
´
´
Paula Perez-Faginas, Ibon Alkorta, M. Teresa Garcıa-Lopez, Rosario Gonzalez-Muniz
˜
Instituto de Quı´mica Me´dica (CSIC), Juan de la Cierva 3, 28006 Madrid, Spain
Received 19 October 2007; accepted 19 November 2007
Available online 22 November 2007
Abstract
Theoretical calculations on the transition states of the cyclization of 2S-chloropropionyl amino acid derivatives to the corresponding
b-lactams have served to explain the high stereoselectivity of the reaction, and have been the driving force to extend the procedure to the
preparation of a Gly-derived 1,3,4-trisubstituted 2-azetidinone in enantiopure form.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The enormous interest of b-lactams in medicinal chem-
istry, as key structural feature of many antibiotics and
serine protease inhibitors,1 and as valuable synthetic inter-
mediates in organic chemistry,2 has triggered considerable
research efforts toward the enantioselective synthesis of
these compounds.3 Most outstanding methods include the
Staudinger reaction between ketenes and imines, with the
stereoselectivity controlled by the use of chiral auxiliaries,
and the Gilman–Speeter condensation between enolates
and imines, using an enantiomerically pure ester or imine
component.4,5 Alternatively, methods for the direct cata-
lytic enantioselective synthesis of b-lactams have already
been developed.6
In this respect, we have recently described a diastereo-
and enantioselective approach to 1,3,4,4-tetrasubstituted
b-lactams through the base-promoted cyclization of
optically pure N-(p-methoxybenzyl)-N-(2-chloro)propionyl
amino acid derivatives.7 We found that the stereochemical
control of the reaction is exclusively governed by the con-
figuration of the N-(2-chloro)propionyl moiety and totally
independent on the configuration of the starting amino
acid.
From a 2S-chloropropionyl amino acid derivative, two
possible intermediates can be envisaged, the pro-S,S and
the pro-SR enolates (Fig. 1). The existence of narrow desta-
bilizing contacts between the 20-methyl group and the
amino acid side-chain in the pro-SR enolate could hamper
H
R1
Me
Cl
(S)
N
CO2R
*
R2
O
B:
R1
OR
O-
H
(E)
R2
Cl
Me
R1
N
OR
O-
(E)
Me
H
N
O
O
(S)
R2
Cl
Enolate pro-SR
Enolate pro-SS
X
Me R1
Me R1
(R)CO2R
(S) CO2R
(S)
H
H
(S)
N
N
R2
R2
O
O
*
Fig. 1. Z-enolates from 2(S)-chloropropionyl amino acid derivatives as
precursors of b-lactams.
Corresponding author. Tel.: +34 91 562 29 00; fax: +34 91 5644853.
´
E-mail address: iqmg313@iqm.csic.es (R. Gonzalez-Muniz).
˜
0040-4039/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.11.099

216
P. Pe´rez-Faginas et al. / Tetrahedron Letters 49 (2008) 215–218
the formation of the trans 3S,4R-b-lactam, favouring the
generation of the cis b-lactam 3S,4S.
were considered for the formation of 3S,4S- and 3S,4R-
b-lactams.
To gain further insights into the formation of these b-
lactam, theoretical calculations on the possible transition
states leading to ring-closure were performed and described
here. To this end, three simplified models were constructed
(Table 1, Fig. 2): model I is an Ala derivative with no alkyl
group at the N atom, a requirement for the cyclization
occurs;8 model II is an N-MeAla analogue, which could
represent any amino acid derivative bearing R2 5 H; and
model III is the corresponding Gly derivative with
R2 = H. For each model, the initial conformation of the
chloropropionyl derivative approaching the reactive
groups, the transition state (TS) and the final b-lactam
The geometry of these three structures has been opti-
mized initially with the hybrid HF/DFT method B3LYP
and the 6-311++G** basis set within the GAUSSIAN-03
package.9–11 Frequency calculations have been carried
out at the same computational level to confirm that the
structure obtained corresponds to energetic minima or pure
transition states (zero and one imaginary frequencies,
respectively).
The energies of the stationary structures are gathered in
Table 1 and the geometrical results in the case of the N-
unsubstituted derivatives (model I) are shown in Figure
2. The most stable initial conformation has been used as
Table 1
Relative energies found for the three structures considered in each model (B3LYP/6-311++G**
)
R1
N
R1
HO2C
O
O
N
Cl-
+
+
Cl
CH3
H
HO2C
(S) (S)
R2
CH3
R2
H
R1
N
R1
O
O
N
Cl-
R2
R2
Cl
CH3
(R)(
S)
CH3
HO2C
H
H
CO2H
a
R1
R2
Config. 3,4
Relative energy (kJ/mol)
Distance (A)
˚
Model
Initial
TS
Final
I
H
Me
Me
H
S,R
S,S
13.87
0.00
77.65
63.85
ꢀ30.42
ꢀ31.33
2.284
2.295
II
Me
Me
S,R
S,S
13.00
0.00
65.73
52.72
ꢀ46.96
2.287
2.297
ꢀ48.20
III
S,R
S,S
6.45
0.00
61.92
56.52
ꢀ44.17
ꢀ41.75
2.251
2.280
a
Interatomic distance of the Cꢁ ꢁ ꢁC forming bond in the TS.
Fig. 2. Initial conformations, transition states and final b-lactams for model I.

P. Pe´rez-Faginas et al. / Tetrahedron Letters 49 (2008) 215–218
217
reference value to calculate the relative energy of the other
species in the reaction.
Cl
Me
N
Me
(S)
t
HN
CO2 Bu
(S)
t
Cl
CO2H
As shown in Table 1, the initial structure of the molecule
with S,S configuration is the most stable one in each
model. The difference of energy between S,S and R,S initial
structures was found to be highly dependent on the R2
group. Thus, when R2 is a methyl the difference is about
13 kJ/mol, while the relative energy is reduced to 6.5 kJ/
mol when it is a hydrogen. In contrast, the presence of a
methyl group or a hydrogen attached to the amide nitrogen
(R1) does not significantly change the energetic differences
obtained in the two reaction paths considered. A similar
tendency is observed for the corresponding TS structures,
with about 13 kJ/mol SS/SR difference in models I and
II and about 5 kJ/mol for model III. The negative values
of the final structures indicate that the reaction is exother-
mic. In addition, the relative stability of the S,S and S,R
species is reduced up to 1 kJ/mol, being the S,S isomer
more stable for models I and II and the S,R isomer pre-
ferred when R2 = H (model III).
O
CO2 Bu
Cl3CCN/Ph3P
OMe
2
1
O
OMe
THF
BTPP
MeCN
NOE 3.6%
H H
t
Me
O
O
O
Me (S)
(S)CO2 Bu
N
O
N
+
OMe
4
3
OMe
The results of the theoretical calculation are in agree-
ment with our previous experimental findings. Thus, the
cyclization of different N-alkyl-N-2S-chloropropionyl
amino acids always afforded the corresponding 3S,4S-b-
lactam, in agreement with model II (DETS = 13 kJ/mol).7
The absence of reaction with N-non-alkylated chloroacetyl
amino acids could be explained through the high energy of
the transition states in model I (R1 = H).8 However, the
possibility of obtaining b-lactams in a stereoselective way
from Gly, according to model III, remains to be explored.
To evaluate this possibility, N-(p-methoxybenzyl)-N-2S-
chloropropionyl-Gly-OtBu (2) was prepared and cyclized,
as shown in Scheme 1. This compound was synthesized
under rigorous neutral conditions to avoid racemization
of the enantiomerically pure 2S-chloropropionic acid.
Thus, 2S-chloropropionyl chloride was generated from
2S-chloropropionic acid with trichloroacetonitrile, as chlo-
rinating agent,12 and solid-supported triphenyl phosphine,
and then reacted with compound 1 and propylene oxide as
HCl scavenger. The BTPP-induced cyclization of com-
pound 2 afforded 2-azetidinone 3, as the only b-lactam,
along with a significant amount of the morpholinedione
derivative 4, the product of the O-alkylation (P10%).13
A coupling constant of 5.6 Hz between H-3-H-4 protons
1. TFA
1. TFA
2. H-D-Phe-OMe
BOP/TEA
2. H-L-Phe-OMe
BOP/TEA
O
O
H H
H H
Me (S)
(R)
(S)
CO2Me
CO2Me
N
H
N
H
Me (S)
(S)
(S)
N
N
O
O
6
5
OMe
OMe
Scheme 1.
1
In the H NMR spectra, the 3-methyl group of the azeti-
dine moiety in dipeptide 5 appeared 0.23 ppm more
shielded than the same signal in compound 6 (Table 2).
This shielding is indicative of a homochiral dipeptide, as
previously observed in related dipeptides containing Ala-,
Lys-, and Glu-derived azetidines.7 Moreover, the retention
time of dipeptide 6 is higher than that of analogue 5 and,
therefore, it was assigned as the heterochiral dipeptide.16
In short, it can be concluded that a 5.4 kJ/mol difference
of energy in the transition states of model III is sufficiently
high to allow the formation of 1,3,4-trisubstituted b-lac-
tams in an enantioselective way, enhancing the generality
1
in the H NMR spectrum of 3,14 and a 3.6% enhancement
of the H-3 signal upon saturation of the H-4 proton, along
with the absence of NOE between H-4 and 3-Me group, are
indicative of a cis-stereochemistry. This evidences that the
stereochemical course of the reaction is governed by the
different energy of the transition state (model III), while
the base used, BTPP, is not strong enough for the abstrac-
tion of H-3 or H-4 proton to afford the trans-b-lactam,
which according to the theoretical calculations is the most
stable.
Table 2
Selected chemical shifts and HPLC retention times for dipeptide deriva-
tives 5 and 6
Compd
d 4-H (ppm)
d 3-CH3 (ppm)
tRHPLCa (min)
5
6
a
3.79
3.74
0.85
1.08
24.24
25.97
The absolute configuration at C-3,4 in compound 3 was
established as S,S after the preparation of dipeptide deriv-
atives 5 and 6 and comparison with related dipeptides.15–17
Novapak C18 (3.9 · 150 mm), MeCN/H2O (0.05% TFA), (28:72).

218
P. Pe´rez-Faginas et al. / Tetrahedron Letters 49 (2008) 215–218
13. (3S,4S)-4-tert-Butoxycarbony-3-methyl-1-p-methoxy-benzyl-2-azetidi-
none (3): Oil. HPLC: tR = 8.56 min (MeCN/H2O (0.05% TFA) =
40:60). [a]D ꢀ25.5 (c 1.04, CHCl3). 1H NMR (300 MHz, CDCl3): d
of the method previously developed for the 1,3,4,4-tetra-
substituted analogues.
t
1.23 (d, 3H, J = 7.5, 3-CH3), 1.47 (s, 9H, CH3 Bu), 3.40 (m, 1H, 3-H),
Acknowledgements
3.79 (s, 3H, OMe), 3.84 (d, 1H, J = 5.6, 4-H), 4.05 (d, 1H, J = 4.6,
1-CH2), 4.77 (d, 1H, J = 14.6, 1-CH2), 6.86 (d, 2H, J = 8.7, C6H4),
7.14 (d, 2H, J = 8.7, C6H4). 13C NMR (75 MHz, CDCl3): d 9.9
We thank the Plan Nacional de Biomedicina (SAF
t
(3-CH3), 28.1 (CH3 Bu), 44.3 (1-CH2), 48.1 (3-C), 53.4 (4-C), 55.1
´
SAF2006-01205) and Ministerio de Ciencia y Tecnologıa
(OMe), 82.5 (CtBu), 114.1, 127.0, 129.8 y 158.2 (Ar), 168.7 y 169.3
(CO). EM (ES positive mode): 328.0 (M+Na)+. Anal. Calcd for
(Project No. CTQ2006-14487-C02-01/BQU) for financial
support. P.P.-F. thanks the CSIC for an I3P fellowship.
C
17H23NO4: C, 66.86; H, 7.59; N, 4.59. Found: C, 66.78; H, 7.74; N,
4.86.
14. (a) Alcaide, B.; Almendros, P.; Redondo, M.; Ruiz, P. J. Org. Chem.
References and notes
´
´
´
´
2005, 70, 8890–8894; (b) Dıez, E.; Fernandez, R.; Marques-Lopez, E.;
´
Martın-Zamora, E.; Lassaletta, J. Org. Lett. 2004, 6, 2749–2752.
1. For recent reviews and leading papers on bioactive b-lactams, see: (a)
Singh, G. S. Mini-Rev. Med. Chem. 2004, 4, 69–92 and pp 93–109; (b)
Veinberg, G.; Vorona, M.; Shestakova, I.; Kanepe, I.; Lukevics, E.
Curr. Med. Chem. 2003, 10, 1741–1757; (c) Clemente, A.; Domingos,
A.; Grancho, A. P.; Iley, J.; Moreira, R.; Neres, J.; Palma, N.;
Santana, A. B.; Valente, E. Bioorg. Med. Chem. Lett. 2001, 11, 1065–
15. (3S,4S,10S)-3-Methyl-1-p-methoxybenzyl-4-(10-methoxy-carbonyl-20-
phenylethyl)carbamoyl-2-azetidinone (5): Oil. HPLC: tR = 24.24 min
1
(MeCN/H2O (0.05% TFA) = 28:72). H NMR (300 MHz, CDCl3): d
0.92 (d, 3H, J = 7.5, 3-CH3), 2.89 (dd, 1H, J = 14.1, 7.3, 20-H), 3.14
(dd, 1H, J = 14.1, 5.3, 20-H), 3.36 (m, 1H, 3-H), 3.75 (s, 3H, OMe),
3.79 (s, 3H, OMe), 3.86 (d, 1H, J = 5.8, 4-H), 4.08 (d, 1H, J = 14.6, 1-
CH2), 4.58 (d, 1H, J = 14.6, 1-CH2), 4.87 (m, 1H, 10-H), 6.03 (d, 1H,
J = 8.1, 10-NH), 6.87 (d, 2H, J = 8.8, C6H4), 7.04 (m, 2H, C6H5), 7.81
(d, 2H, J = 8.8, C6H4), 7.23 (m, 3H, C6H5). EM (ES positive mode):
411.0 (M+1)+, 433 (M+Na)+. Anal. Calcd for C23H26N2O5: C, 67.30;
H, 6.38; N, 6.82. Found: C, 67.55; H, 6.19; N, 6.77. (3S,4S,10R)-3-
Methyl-1-p-methoxybenzyl-4-(10-metoxy-carbonyl-20-phenylethyl)carb-
amoyl-2-azetidinone (6): Oil. HPLC: tR = 25.97 min (MeCN/H2O
(0.05% TFA) = 28:72). 1H NMR (300 MHz, CDCl3): d 1.08 (d, 3H,
J = 7.7, 3-CH3), 2.89 (dd, 1H, J = 14.3, 7.7, 20-H), 3.13 (dd, 1H,
J = 14.3, 4.7, 20-H), 3.32 (m, 1H, 3-H), 3.53 (d, 1H, J = 14.8, 1-CH2),
3.68 (s, 3H, OMe), 3.71 (s, 3H, OMe), 3.74 (d, 1H, J = 5.8, 4-H), 4.52
(d, 1H, J = 14.8, 1-CH2), 4.83 (m, 1H, 10-H), 6.12 (d, 1H, J = 7.9,
10-NH), 6.74 (d, 2H, J = 7.8, C6H4), 6.89 (d, 2H, J = 7.8, C6H4), 7.04
(m, 2H, C6H5), 7.24 (m, 3H, C6H5). EM (ES positive mode): 411.0
(M+1)+, 433.0 (M+Na)+. Anal. Calcd for C23H26N2O5: C, 67.30; H,
6.38; N, 6.82. Found: C, 67.48; H, 6.56; N, 6.69.
´
´
´
1068; (d) Gerona-Navarro, G.; Perez de Vega, M. J.; Garcıa-Lopez,
´
M. T.; Andrei, G.; Snoeck, R.; De Clercq, E.; Balzarini, J.; Gonzalez-
Muniz, R. J. Med. Chem. 2005, 48, 2612–2621; (e) Sperka, T.; Pitlik,
˜
J.; Bagossi, P.; To¨zser, J. Bioorg. Med. Chem. Lett. 2005, 15, 3086–
3090; (f) Kvaerno, L.; Werder, M.; Hauser, H.; Carreira, E. M. J.
Med. Chem. 2005, 48, 6035–6053.
2. (a) Alcaide, B.; Almendros, P.; Aragoncillo, C. Chem. Rev. 2007,
doi:10.1021/cr0307300; (b) Alcaide, B.; Almendros, P. Curr. Med.
Chem. 2004, 11, 1921–1949; (c) Ojima, I.; Delaloge, F. Chem. Soc.
Rev. 1997, 26, 377–386.
3. Singh, G. S. Tetrahedron 2003, 59, 7631–7649.
4. Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M. Curr. Med.
Chem. 2004, 11, 1837–1872.
5. Benaglia, M.; Cinquini, M.; Cozzi, F. Eur. J. Org. Chem. 2000,
563–572.
6. (a) France, S.; Weatherwax, A.; Taggi, A. E.; Lectka, T. Acc. Chem.
Res. 2004, 37, 592–600; (b) Magriotis, P. A. Angew. Chem., Int. Ed.
2001, 40, 4377–4379.
16. In dipeptide derivatives composed by one aliphatic and one aromatic
amino acid, the dipeptide with higher retention time in HPLC and a
more shielded chemical shift of the b-H aliphatic side-chain protons is
´
´
7. Perez-Faginas, P.; O’Reilly, F.; O’Byrne, A.; Garcıa-Aparicio, C.;
´
´
´
´
´
Martın-Martınez, M.; Perez de Vega, M. J.; Garcıa-Lopez, M. T.;
´
assigned as the heterochiral dipeptide: (a) Gonzalez-Muniz, R.;
˜
Cornille, F.; Bergeron, F.; Ficheux, D.; Pothier, J.; Durieux, C.;
Roques, B. P. Int J. Pep. Prot. Res. 1991, 37, 331–340. (b) Fournie-
Zaluski, M. C.; Lucas-Soroca, E.; Devin, J.; Roques, B. P. J. Med.
Chem. 1986, 29, 751–757.
´
Gonzalez-Muniz, R. Org. Lett. 2007, 9, 1593–1596.
˜
´
´
8. Gerona-Navarro, G.; Bonache, M. A.; Herranz, R.; Garcıa-Lopez,
´
´
M. T.; Gonzalez-Muniz, R. J. Org. Chem. 2001, 66, 3538–3547.
˜
9. (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652; (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
10. Frisch, M. J.; Pople, J. A.; Krishnam, R.; Binkley, J. S. J. Chem. Phys.
1984, 80, 3265–3269.
11. Frisch, M. J. et al., GAUSSIAN 03, Gaussian: Pittsburgh PA, 2003.
12. Pluempanupat, W.; Chantarasriwong, O.; Taboonpong, P.; Jang, D.;
Chavarisi, W. Tetrahedron Lett. 2007, 48, 223–226.
17. The validity of the above dipeptide rules has unequivocally been
demonstrated for a dipeptide derivative composed by a Phe-derived
b-lactam, without the 3-methyl substituent, and L-Ala-OMe: Gerona-
´
´
´
Navarro, G.; Garcıa-Lopez, M. T.; Gonzalez-Muniz, R. J. Org.
Chem. 2002, 67, 3953–3956.
˜