Conversion of isomeric 2:3 adducts (aminoacid-formaldehyde) to N-acyl-pseudoprolines derivatives

Article abstract of DOI:10.1016/j.tetlet.2004.11.134

Reaction of acyl chlorides or acid anhydrides with isomeric 2:3 adducts derived from condensation of l-serine (1a), l-threonine (1b) and l-cysteine (1c) methyl esters with formaldehyde afforded N-acyl-pseudoprolines 7-19 in various yields. These 2:3 adducts can be considered as synthetic equivalents of oxaproline and thiaproline moieties. The present work revealed the versatile behaviour of the two 2:3 adducts as a consequence of the ring-chain tautomerism occurring in the five- and/or seven-membered rings.

Full text of DOI:10.1016/j.tetlet.2004.11.134

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 615–617
Conversion of isomeric 2:3 adducts (aminoacid–formaldehyde) to
N-acyl-pseudoprolines derivatives
a,
Jimmy Selambarom, Jacqueline Smadja and Andre A. Pavia
a
b
*
´
´
a
´
Laboratoire de Chimie des Substances, Naturelles et des Sciences des Aliments, Faculte des Sciences and Technologies,
´
Universite de La Reunion, BP 7151, F-97715 Saint-Denis Messageries Cedex 9, France (DOM)
´
bEmeritus professor of the University Montpellier II, France
Received 1 October 2004;revised 24 November 2004;accepted 26 November 2004
Abstract—Reaction of acyl chlorides or acid anhydrides with isomeric 2:3 adducts derived from condensation of L-serine (1a), L-
threonine (1b) and L-cysteine (1c) methyl esters with formaldehyde afforded N-acyl-pseudoprolines 7–19 in various yields. These
2:3 adducts can be considered as synthetic equivalents of oxaproline and thiaproline moieties. The present work revealed the ver-
satile behaviour of the two 2:3 adducts as a consequence of the ring-chain tautomerism occurring in the five- and/or seven-membered
rings.
Ó 2004 Elsevier Ltd. All rights reserved.
Condensation of carbonyl compounds (R1 5 H and/or
R2
R2
R1
X
R1
HX
R2 5 H) with primary b-hydroxyamine moieties affor-
ded NH-free oxazolidines type 2 (X = O: oxaproline)
(Scheme 1), which are generally poorly stable and non-
isolable. The latter proved to exist as a tautomeric mix-
ture with the imine open-chain forms type 3,1 in contrast
to sulfur derivatives (X = S: thiaproline).2 Depending on
the experimental conditions, condensation of formalde-
hyde (R1 = H and R2 = H) has been reported to afford
isomeric 2:3 adducts:3 N,N-methylenebis(oxazolidine)
derivatives type 5 (X = O) and/or isomeric 1,6-diaza-
i.
NH
N
R
CO2CH3
R
CO2CH3
HX
R
NH2
2a-c
3a-c
CO2CH3
1a-c
a. X=O; R=H (Ser)
b. X=O; R=CH3 (Thr)
c. X=S; R=H (Cys)
CO2CH3
R
X
CO2CH3
R
X
NH
CO2CH3
4a-c
N
N
X
N
X
R
N
ii.
X
R
R
H3CO2C
5a-c
CO2CH3
6a-c
3,9-dioxabicyclo[4.4.1]undecane structures type
(X = O), instead of the expected oxazolidines type 4
6
iii.
X = O or S
O
(X = O).
X = S
iii.
R'
Common strategic routes to N-substituted pseudopro-
lines are prone to ring-chain tautomerism and/or com-
peting reactions.4 In a recent paper, we reported that
2:3 adducts 5b (X = O;R = CH 3) and/or 6b (X = O;
R = CH3) derived from L-threonine methyl ester 1b
could be converted into N-acyloxaprolines 13 and 14,5
which are usually prepared by cyclocondensation of N-
monoacylated aminoesters with carbonyl compounds.
Such observation prompted us to examine the scope of
this approach, since these N-acyl derivatives have poten-
X
R
N
X
N
R'
O
CO2CH3
R
CO2CH3
Anti
Syn
7-19
Scheme 1. Reagents and conditions: (i) R1R2C@O, Et3N, CH2Cl2, rt;
(ii) (HCHO)3, Et3N, CH2Cl2, rt;(iii) R COCl, Na2CO3, H2O/CH3CN.
0
tial in various fields6,7 including development of peptides
mimics.8 The procedure has now been investigated on
similar and easily accessible 2:3 adducts 6a9 (X = O;
R = H) and 5c10 (X = S;R = H) derived from L-serine
methyl ester 1a and L-cysteine methyl ester 1c,
respectively.
Keyword: N-Acyl-pseudoprolines.
*
Corresponding author. Tel.: +33 262 938 619;fax: +33 262 938
183;e-mail: j.selamb@univ-reunion.fr
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.11.134

´
J. Selambarom et al. / Tetrahedron Letters 46 (2005) 615–617
616
Table 1.
CO2CH3
R
Starting 2:3
adducts
X
R
R0
Product
Yield
(%)
N
N
O
HO
NHCOR'
CO2CH3
R'OCO
NHCOR'
CO2CH3
O
R
+
6a
O
O
H
H
CH3
C6H5
7
10
31
6
CO2CH3
9. R' = CH3
12. R' = C6 H5
8. R' = CH3
11.R' = C6 H5
6a. R = H
6b. R = CH3
5b and/or 6b
O
O
O
O
CH3
CH3
CH3
CH3
CH3
C6H5
13
14
15
16
705
605
54
R'COCl
p-OMe-C6H4
p-NO2-C6H4
R = H
58
CO2CH3
N
R
HO
R
NH2
CO2CH3
5c
S
S
S
H
H
H
CH3
C6H5
p-OMe-C6H4
17
18
19
71
83
56
O
- 3 HCHO
O
N
R
CO2CH3
1a. R = H
1b. R = CH3
O
CO2CH3
R
R'
The reaction was performed in water/acetonitrile at
room temperature overnight with various acylating
agents according to previous procedure5 and the results
listed in Table 1. All products were isolated on silica gel
column chromatography and gave satisfactory spectro-
scopic data.11 Mass spectrum in positive FAB ionization
mode displayed the molecular ion [M+H]+ and the frag-
ments [M+HÀR0C@O]+ and [R0C@O]+ resulting from
the amide bond scission. As already reported for com-
pounds 13 and 14,5 1H NMR spectra of the novel com-
pounds 7, 10 and 15–19 displayed the expected
deshielding of intracyclic protons that are closed to the
N–C@O grouping and the duplication of NMR signals
reflecting the existence of two syn/anti-rotational iso-
mers (Scheme 1). This was confirmed by NMR signals
coalescence upon raising temperature.
O
N
R'
O
N
O
N
O
O
N
R'COCl
R
CO2CH3
R
CO2CH3
R
H3CO2C
Syn
Anti
5a. R = H
5b. R = CH3
Scheme 2.
The present findings open an alternative way to prepare
N-acyl-pseudoprolines derived from L-threonine and
L-cysteine. Further investigations are in progress for
the particular but not definitive case of L-serine. These
results supported the expectations that N,N-methylene-
bis(oxazolidine) and 1,6-diaza-3,9-dioxabicyclo[4.4.1]-
undecane structures could be considered as synthetic
equivalents of the non-isolable oxaprolines 4a and 4b.
This procedure will be extended to the synthesis of
pseudopeptides containing pseudoproline residues.
As shown in Table 1, condensation with 2:3 adduct 5c
afforded compounds 17–19 in satisfactory yields (56–
83%). The latter were very closed to those obtained
when starting from the thiaproline 4c (X = S;R = H).
Therefore, the potential of this procedure seemed to be
confirmed.
Acknowledgements
The authorÕs are grateful to the Regional Council of
Reunion Island for financial support.
For L-threonine derivatives 5b/6b, the conversion
occurred with yields ranging from 54% to 70%. Similar
yields ( 5%) were obtained when starting from either
pure isomers 5b and 6b or from the equilibrium mixture,
thus allowing to bypass the separation step. In contrast,
using the same conditions, isomer 6a afforded a mixture
of N-acylated and N,O-diacylated by-products 8/9 and
11/12,12 besides the expected compounds 7 or 10, respec-
tively. Formation of these by-products gave clear-cut
evidence of the existence of a ring-chain tautomerism
occurring in the N–C–O grouping of the seven-mem-
bered rings (Scheme 2).
References and notes
1. Fulo¨p, F.;Pihlaja, K. Tetrahedron 1993, 49, 6701–6706.
¨
2. Fulo¨p, F.;Mattinen, J.;Pihlaja, K. Tetrahedron 1990, 46,
¨
6545–6552.
3. See for examples: (a) Aitken, D. J.;Guillaume, D.;
Husson, H.-P.;Chiaroni, A.;Riche, C. J. Heterocycl.
Chem. 1991, 28, 705–709;(b) Kostyanovsky, R. G.;
ElÕnatanov, Yu. I.;Krutius, O. N.;Chervin, I. I.;Za ddach,
H.;Koehler, K. F. Russ. Chem. Bull. 1994, 43, 321–322;(c)
Bolm, C.;Chuang, T.-H.;Raabe, G.;Fang, J.-M. Synth.
Commun. 1999, 29, 43–51;(d) Aitken, D. J.;Besson, L.;
Fournier, F.;Husson, H.-P.;Lemoine, P.;Lesage, D.;
`
The efficiency of the procedure in the case of pure
bicyclic derivative 6b should be ascribed to its rapid
isomerization10 into N,N-methylenebis(oxazolidine)
isomer 5b via transient open-chain forms. In contrast,
the slow conversion10 6a ! 5a should favour their
decomposition into their parent aminoester 1a as
demonstrated for similar derivatives.13 Subsequent acyla-
tion led to the by-products 8/9 or 11/12. The overall
experiments suggested that N,N-methylenebis(oxa- or
thiazolidine) structures type 5 are the more reactive
forms.
`
Libot, F.;Martin, P.-G.;Mellin-Morlie re, C.;Monnier,
Heterocycles 2004, 64,
V.;Tabet, J.-C.;Viossat, B.
published on line.
4. See for example: Luo, Y.;Evindar, G.;Fishlock, D.;
Lajoie, G. A. Tetrahedron Lett. 2001, 42, 3807–3809.
´
´
5. Selambarom, J.;Monge, S.;Carre , F.;Roque, J.-P.;Pavia,
A. A. Tetrahedron 2002, 58, 9559–9566.
6. Kanemasa, S.;Onimura, K. Tetrahedron 1992, 48, 8645–
8658.
7. Taylor, W. G.;Hall, T. S.;Schreck, C. E. Can. J. Chem.
1992, 70, 165–172.

´
J. Selambarom et al. / Tetrahedron Letters 46 (2005) 615–617
617
8. (a) Kern, D.;Schutkowski, M.;Drakenberg, T. J. Am.
Chem. Soc. 1997, 119, 8403–8408;(b) Nishi, T.;Fukaz-
awa, T.;Kurata, H.;Ishibashi, K.;Nakajima, K.;Yam-
aguchi, T.;Itoh, K. EP-776893-A1, 1996, Sankyo Co., Ltd;
(c) Nishi, T.;Fukazawa, T.;Ishibashi, K.;Nakajima, K.;
Sugioka, Y.;Iio, Y.;Kurata, H.;Itoh, K.;Mukaiyama,
O.;Satoh, Y.;Yamaguchi, T. Bioorg. Med. Chem. 1999, 9,
875–880.
3H, CH3CO);3.24 and 3.27 (each m, 2H, S–C H2–CH);
3.77 and 3.82 (each s, 3H, OCH3) ;4.61 and 4.64 (each m,
2H, S–CH2–N);4.77 and 5.12 (each dd, J = 5.7 Hz,
J = 3.81 Hz and J = 6.6 Hz, J = 4.02 Hz, 1H, S–CH2–CH).
12. For instance, compound 11: FAB+ MS (NBA) m/z: 224
1
M+H+ ;105 [Ph–C @O]+; H NMR (CDCl3;200 MHz) d:
3.21 (br s, 1H, OH);3.81 (s, 3H, OCH 3);4.06 (dd,
J = 11.2 Hz, J = 7.5 Hz, 2H, O–CH2–CH);4.85 (dt,
J = 7.5 Hz, J = 3.5 Hz, 1H, O–CH2–CH);7.26 (d,
J=11.2 Hz, 1H, NH);7.43–7.86 (m, 5H, H-ar). Compound
12: FAB+ MS (NBA) m/z: 328 [M+H]+;105 [Ph–C @O]+;
1H NMR (CDCl3;200 MHz) d: 3.86 (s, 3H, OCH3);4.80
(dd, J=7.5 Hz, J = 3.5 Hz, 2H, O–CH2–CH);5.21 (dt,
J=3.5 Hz, J=3.5 Hz, 1H, O–CH2–CH);7.14 (d, J=7.3 Hz,
1H, NH);7.40–7.94 (m, 5H, H-ar).
´
9. Selambarom, J.;Monge, S.;Carre , F.;Fruchier, A.;
Roque, J.-P.;Pavia, A. A. Carbohydr. Res. 2001, 330, 43–
´
51.
´
´
10. Selambarom, J.;Carre , F.;Fruchier, A.;Roque, J.-P.;
Pavia, A. A. Tetrahedron 2002, 58, 4439–4444.
11. For instance, compound 17: FAB+ MS (NBA) m/z: 190
M+H+;212 [M+Na] +;43 [CH 3–C@O]+;148 [M+H ÀCH3–
C@O]+;379 [2M+H]
(syn- and anti-rotational isomers): 2.05 and 2.19 (each s,
;
1H NMR (CDCl3;200 MHz) d:
13. Moloney, G. P.;Iskander, M. N.;Craik, D. J.
Reson. Chem. 1997, 35, 337–341.
Magn.
+